Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Non-covalent control of mammalian acetyl-CoA carboxylase isoforms Lee, Weissy Michelle 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2010_spring_lee_weissy.pdf [ 2.32MB ]
Metadata
JSON: 24-1.0069141.json
JSON-LD: 24-1.0069141-ld.json
RDF/XML (Pretty): 24-1.0069141-rdf.xml
RDF/JSON: 24-1.0069141-rdf.json
Turtle: 24-1.0069141-turtle.txt
N-Triples: 24-1.0069141-rdf-ntriples.txt
Original Record: 24-1.0069141-source.json
Full Text
24-1.0069141-fulltext.txt
Citation
24-1.0069141.ris

Full Text

NON-COVALENT CONTROL OF MAMMALIAN ACETYL-COA CARBOXYLASE ISOFORMS  by  Weissy Michelle Lee  B.Sc., University of British Columbia, 2002  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in  The Faculty of Graduate Studies  (Biochemistry and Molecular Biology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  February 2010  © Weissy Michelle Lee, 2010  Abstract Acetyl-CoA carboxylase (ACC) plays several crucial roles in lipid metabolism and has been identified as a potential drug target for the treatment of obesity and cancer. Many aspects of ACC structure, function and control remain unclear and my general goal was to provide new insights into two unresolved areas. The first part of this thesis describes work to characterize the key allosteric site that is responsible for citrate-induced ACC activation. The second area addressed is the hypothesis that protein-protein interactions provide critical aspects of ACC control. The effects of citrate on ACC were probed by studying the effects of pyridoxal phosphate (PLP), modeled on earlier studies of phosphofructokinase. A comprehensive assessment of the inhibition kinetics of ACC by PLP provides support for the effects of PLP being non-competitive with respect to substrates and therefore perhaps mediated by binding to a non-catalytic site. Further studies established methods for covalent attachment of PLP via Schiff base reduction with sodium borohydride and set the stage for more precise localization of PLP binding. To define proteins that interact specifically with ACC, techniques were established to separate large polymeric forms of ACC from smaller dimeric forms of the enzyme by size exclusion chromatography. Analysis of the citrate-induced ACC polymers by tandem mass spectrometry led to the identification of a small number of proteins that consistently associated with ACC in a citrate-dependent manner, including fatty acid synthase (FASN) and tubulin. The interactions of ACC with FASN and tubulin were further explored by Western blotting, co-immunoprecipitation, ACC activity assays and immunofluorescence analysis of primary rat hepatocytes. The evidence from several complementary experimental approaches supports specific interactions between ACC and tubulin. Tubulin appears to have subtle effects on ACC catalytic activity and may be more critical for structural organization and cellular localization of ACC. Immunofluorescence microscopy of ACC within primary rat hepatocytes led to novel insights, notably that ACC is localized into discrete structures rather than being distributed through the cytosolic compartment. This work opens the way to further characterization of the relevant subcellular structures and the role this localization plays in the control of ACC isoforms.  ii  Table of contents Abstract.............................................................................................................................. ii Table of contents .............................................................................................................. iii List of tables....................................................................................................................... v List of figures.................................................................................................................... vi List of abbreviations ........................................................................................................ ix Acknowledgements ......................................................................................................... xii Chapter 1: Introduction ...................................................................................................... 1 1.1 Biotin-dependent carboxylases ................................................................................. 1 1.2 Acetyl-CoA Carboxylase .......................................................................................... 4 1.2.1 Discovery of ACC.............................................................................................. 4 1.2.2 ACC reaction mechanism and kinetics .............................................................. 6 1.2.3 Evolution of the structure and properties of ACC ........................................... 12 1.2.4 ACC structure .................................................................................................. 19 1.2.5 ACC isoforms in animal cells .......................................................................... 25 1.2.6 ACC regulation ................................................................................................ 29 1.2.6.1 Regulation of ACC expression ................................................................. 29 1.2.6.2 Allosteric regulation of ACC .................................................................... 33 1.2.6.3 ACC polymerization ................................................................................. 37 1.2.6.4 ACC regulation by phosphorylation ......................................................... 39 1.3 Biological relevance................................................................................................ 46 1.4 Thesis investigations............................................................................................... 51 Chapter 2: Experimental Procedures ................................................................................ 52 2.1 Materials ................................................................................................................. 52 2.2 Methods................................................................................................................... 53 2.2.1 Tissue isolation and preparation ...................................................................... 53 2.2.2 Tissue homogenization .................................................................................... 53 2.2.3 Purification of ACC ......................................................................................... 54 2.2.4 Purification of rat liver mitochondria .............................................................. 54 2.2.5 Purification of tubulin from rat brain............................................................... 55 2.2.6 Avidin affinity chromatography ...................................................................... 56 2.2.6.1 Preparation of tetrameric and monomeric avidin beads ........................... 56 2.2.6.2 Immobilization of ACC onto tetrameric avidin-agarose beads ................ 57 2.2.6.3 Purification of ACC using monomeric avidin-agarose beads................... 58 2.2.7 Labeling of ACC with PLP using [3H]-borohydride ....................................... 59 2.2.8 Sucrose gradient centrifugation ....................................................................... 60 2.2.9 Size exclusion chromatography using a BioGelA column .............................. 60 2.2.10 Preparation of acetyl-CoA ............................................................................. 61 2.2.11 ACC activity assays ....................................................................................... 62 2.2.12 SDS-PAGE .................................................................................................... 63 2.2.12.1 Sample preparation ................................................................................. 63 2.2.12.2 SDS-PAGE analysis ............................................................................... 63 2.2.13 Western blotting............................................................................................. 64 2.2.13.1 Transfer and blocking ............................................................................. 64 2.2.13.2 Incubations with primary antibody and streptavidin-HRP ..................... 65 2.2.13.3 Sample detection..................................................................................... 66  iii  2.2.13.4 Membrane stripping and staining............................................................ 66 2.2.14 Spectophotometric assays of Fatty acid synthase (FASN) ............................ 67 2.2.15 Mass spectrometry ......................................................................................... 67 2.2.15.1 Sample preparation and in-solution digestion......................................... 67 2.2.15.2 Sample preparation and in-gel digestion................................................. 68 2.2.15.3 LC MS/MS analysis................................................................................ 70 2.2.16 Immunocytochemistry ................................................................................... 71 Chapter 3: Inhibition of ACC by PLP............................................................................... 73 3.1 Rationale ................................................................................................................. 73 3.2 Non-competitive inhibition of ACC by PLP .......................................................... 76 3.3 Labeling of ACC with PLP and [3H]-borohydride ................................................. 81 3.4 Summary ................................................................................................................. 85 Chapter 4: Determining the presence of proteins associating with ACC ......................... 86 4.1 Rationale ................................................................................................................. 86 4.2 Purification of ACC “polymers” and ACC “dimers” through size exclusion chromatography ............................................................................................................ 93 4.3 Presence of associated proteins in the ACC polymeric fraction........................... 103 4.4 Analysis of ACC-associated proteins by mass spectrometry: in-solution digestions ..................................................................................................................................... 113 4.5 LC MS/MS analysis of ACC-associated proteins following SDS-PAGE and in-gel digestion...................................................................................................................... 122 4.6 Previously identified ACC-protein interactions.................................................... 133 4.7 Summary ............................................................................................................... 139 Chapter 5: Interactions of ACC with FASN ................................................................... 141 5.1 Rationale ............................................................................................................... 141 5.2 Co-migration of FASN in the ACC polymeric fraction........................................ 142 5.3 Co-immunoprecipitation ....................................................................................... 146 5.4 The effect of FASN on ACC activity ................................................................... 151 5.5 Summary ............................................................................................................... 152 Chapter 6: Interactions of ACC with tubulin.................................................................. 153 6.1 Rationale ............................................................................................................... 153 6.2 Co-migration of tubulin and ACC during size-exclusion chromatography.......... 154 6.3 The effect of tubulin on ACC activity .................................................................. 158 6.4 Co-immunoprecipitation ....................................................................................... 169 6.5 Studies of the intracellular localization of ACC in primary rat hepatocytes ........ 172 6.5.1 Establishing a method for the visualization of ACC and tubulin .................. 172 6.5.2 Localization of ACC in primary rat hepatocytes ........................................... 188 6.5.3 Hormonal influence on ACC localization and distribution in primary rat hepatocytes.............................................................................................................. 190 6.5.4 Relative localization of ACC and tubulin in primary rat hepatocytes ........... 200 6.6 Summary ............................................................................................................... 206 Chapter 7: Conclusion and future experiments............................................................... 209 References....................................................................................................................... 213 Appendices...................................................................................................................... 229 Appendix A................................................................................................................. 229  iv  List of tables Table 1.1: Some critical conserved residues of ACC . ..................................................... 18 Table 1.2: Amino acid sequence variants of ACC isoforms ............................................ 25 Table 1.3: Chromosomal location of the ACC gene in different organisms. ................... 26 Table 1.4: Identified and predicted phosphorylation sites for human and rat ACC isoforms............................................................................................................................. 43 Table 4.1: Distribution of ACC in different size ranges during size-exclusion chromatography. ............................................................................................................. 109 Table 4.2: Proteins identified from in-solution digestion of the polymeric fraction of rat liver ACC following size-exclusion chromatography. ................................................... 117 Table 4.3: Functions of the proteins identified by LC MS/MS analyses of rat liver ACC polymeric fractions. ........................................................................................................ 118 Table 4.4: Proteins identified by in-solution digestions and LC MS/MS of the polymeric fractions isolated from rat white adipose tissue.............................................................. 120 Table 4.5: Functions of the proteins identified by LC MS/MS analysis of white adipose tissue ACC polymeric fractions. ..................................................................................... 121 Table 4.6: Proteins from rat white adipose tissue that were detected in ACC polymers by MS/MS analysis of excised gel bands. ........................................................................... 125 Table 4.7: Yeast proteins that interact with ACC1 (FAS3). .......................................... 136 Table 4.8: List of predicted protein-binding partners of ACC, based on function of E. coli, S. cerevisiae, R. norvegicus, and human................................................................. 138 Table 5.1: Effects of citrate on the elution of ACC and FASN during size exclusion chromatography fractions following BioGelA-50M size exclusion chromatography was calculated as in figures 4.10 and 5.2 and presented as percentage total in each chromatogram. ................................................................................................................ 145 Table 6.1: Summary of distribution of ACC and tubulin during size-exclusion chromatography. ............................................................................................................. 157 Table 6.2: Effect of tubulin, colchicine and GTP on ACC at different stages of purification...................................................................................................................... 167 Table 6.3: Summary of controls performed to verify the specificity of the anti-tubulin, anti-ACC and anti-phospho ACC antibodies.................................................................. 176 Table 6.4: The dimensions of structures in rat primary hepatocytes visualized with antiACC antibodies............................................................................................................... 198 Table 6.5: The number of structures in rat primary hepatocytes visualized with anti-ACC antibodies. ....................................................................................................................... 199  v  List of figures Figure 1.1: The structure of biotin. ..................................................................................... 2 Figure 1.2: Reactions catalyzed by mammalian biotin-dependent carboxylases. .............. 3 Figure 1.3: Reaction scheme of ACC ................................................................................. 6 Figure 1.4: Two postulated kinetic mechanisms for the ACC reaction........................... 10 Figure 1.5: Domain structures of ACC from various organisms . .................................... 16 Figure 1.6: Key features conserved in ACC from human, rat, and S. cerevisiae. ............ 17 Figure 1.7: Crystal structures of biotin carboxylase. ........................................................ 21 Figure 1.8: X-ray crystal structure for the BCCP subunit of E coli ................................. 22 Figure 1.9: X-ray crystal structures for the CT subunit of S. cerevisiae ACC in complex ........................................................................................................................................... 24 Figure 1.10: Organization of ACC-1 gene and predicted mRNA transcripts .................. 31 Figure 1.11: Chemical structures of several allosteric regulators of ACC. ...................... 34 Figure 1.12: The metabolic roles of malonyl-CoA in mammalian cells.......................... 48 Figure 3.1: Chemical structure of citrate and pyridoxal 5-phosphate.............................. 74 Figure 3.2: Citrate protection of ACC decreases PLP inhibition..................................... 75 Figure 3.3: Effects of PLP on bicarbonate concentration dependence of the ACC reaction.............................................................................................................................. 78 Figure 3.4: Effects of PLP on ATP concentration dependence of the ACC reaction...... 79 Figure 3.5: Effects of PLP on acetyl-CoA concentration dependence of the ACC reaction.............................................................................................................................. 80 Figure 3.6: Incorporation of [3H] into ACC following incubation with PLP and [3H]borohydride. ...................................................................................................................... 83 Figure 3.7: Incorporation of [3H] into ACC following incubation with PLP and [3H]borohydride. ...................................................................................................................... 84 Figure 4.1: Evidence for an ACC-associated “regulator” protein. .................................. 88 Figure 4.2: Model for ACC activation by associated proteins......................................... 91 Figure 4.3: Model for ACC inhibition by associated proteins......................................... 92 Figure 4.4: Separation of the polymeric and dimeric forms of ACC by sucrose gradient centrifugation. ................................................................................................................... 94 Figure 4.5: Calibration of the BiogelA-50M size exclusion column................................ 97 Figure 4.6: Effect of citrate on elution of ACC during size-exclusion chromatography. 98 Figure 4.7: Effects of citrate on recovery of ACC in polymeric and dimeric forms. ...... 99 Figure 4.8: Size exclusion chromatography of rat liver ACC preparations................... 102 Figure 4.9: Citrate-dependent mobility shift of ACC revealed by size-exclusion chromatography. ............................................................................................................. 104 Figure 4.10: Citrate-dependent mobility size shift of ACC revealed by size-exclusion chromatography. ............................................................................................................. 105 Figure 4.11: Citrate-dependent mobility shift in ACC revealed by size-exclusion chromatography. ............................................................................................................. 106 Figure 4.12: Citrate-dependent mobility size shift of ACC revealed by size-exclusion chromatography. ............................................................................................................. 107 Figure 4.13: Protein composition of BioGelA-50M column fractions. ......................... 110 Figure 4.14: Sequence of ACC-1................................................................................... 114 Figure 4.15: Example of a fragment ion pattern of one peptide derived from ACC. ..... 115  vi  Figure 4.16: The protein composition of adipose tissue ACC polymeric fractions in the absence and presence of citrate....................................................................................... 123 Figure 4.17: Sequence of fatty acid synthase.................................................................. 126 Figure 4.18: Example of fragment ion pattern of one peptide derived from fatty acid synthase........................................................................................................................... 127 Figure 4.19: Sequences of rat tubulin isoforms. ............................................................ 128 Figure 4.20: Example of fragment ion pattern of one peptide derived from tubulin...... 129 Figure 5.1: Citrate-dependent size-shift of fatty acid synthase....................................... 143 Figure 5.2: Citrate-dependent size-shift of FASN. ........................................................ 144 Figure 5.3: Controls for the co-immunoprecipitation experiments. .............................. 147 Figure 5.4: Immunoprecipitation of ACC with anti-phospho ACC and anti-ACC1. .... 149 Figure 5.5: Attempts to immunoprecipitate proteins with anti-FAS antibody. ............. 150 Figure 6.1: Citrate-dependent co-migration of ACC and tubulin during size-exclusion chromatography. ............................................................................................................. 155 Figure 6.2: Citrate-dependent mobility size shift of tubulin revealed by size-exclusion chromatography. ............................................................................................................. 156 Figure 6.3: Isolation of microtubules from rat brain...................................................... 159 Figure 6.4: Effect of GTP on ACC activity. .................................................................. 161 Figure 6.5: Effect of colchicine on ACC activity. ......................................................... 162 Figure 6.6: The effects of GTP, colchicine, and tubulin on ACC activity. ................... 164 Figure 6.7: The effects of GTP, colchicine, and tubulin on ACC activity. ................... 166 Figure 6.8: Co-immunoprecipitation of tubulin with anti-ACC antibodies................... 170 Figure 6.9: Co-immunoprecipitation of ACC with anti-α-tubulin antibodies............... 171 Figure 6.10: The ACC protein distribution in primary rat hepatocytes and HepG2 cells. ......................................................................................................................................... 174 Figure 6.11: Specificity of secondary antibodies........................................................... 177 Figure 6.12: Immunofluorescence microscopy of primary rat hepatocytes cultured on matri-gel or fibronectin. .................................................................................................. 179 Figure 6.13: Comparison of fixation of primary hepatocytes with formaldehyde or paraformaldehyde. .......................................................................................................... 181 Figure 6.14: Nuclear staining of primary rat hepatocytes............................................... 183 Figure 6.15: Detection of tubulin in primary rat hepatocytes by immunofluorescence microscopy...................................................................................................................... 184 Figure 6.16: Detection of ACC-1 in primary rat hepatocytes by immunofluorescence microscopy...................................................................................................................... 186 Figure 6.17: Detection of ser-79 phosphorylated ACC in primary rat hepatocytes by immunofluorescence. ...................................................................................................... 187 Figure 6.18: Contrasting distribution of ACC within cultured cell lines and in primary rat liver hepatocytes. ............................................................................................................ 189 Figure 6.19: Effect of acute insulin and anoxic stress on ACC ser-79/ser-212 phosphorylation in primary rat hepatocytes.................................................................... 191 Figure 6.20: Effect of insulin or anaerobic stress on microtubule structure in primary rat hepatocytes...................................................................................................................... 193 Figure 6.21: Effect of insulin or anaerobic stress on ACC localization in primary rat hepatocytes...................................................................................................................... 194  vii  Figure 6.22: Multiple images of the effects of insulin or anaerobic stress on ACC localization...................................................................................................................... 196 Figure 6.23: Multiple images of the effects of insulin or anaerobic stress on ACC localization...................................................................................................................... 197 Figure 6.24: Cells with well-defined microtubule structures exhibit punctate localization of ACC. ........................................................................................................................... 202 Figure 6.25: Cells cultured on matri-gel lack microtubular structure and exhibit substantial cytosolic dispersion of ACC. ........................................................................ 203 Figure 6.26: Effects of colchicine on microtubule organization and ACC distribution in primary rat hepatocytes................................................................................................... 205  viii  List of abbreviations x  times  g  gravity  °C  degrees Celsius  %  percent  Å  angstrom  aa  amino acid  ACC-1 or 2  acetyl-CoA carboxylase-1 or 2  AMPK  AMP-activated protein kinase  ATP  adenosine 5’-triphosphate  BC  biotin carboxylase  BCCP  biotin carboxyl carrier protein  BRCA1  breast cancer-associated protein 1  BSA  bovine serum albumin  CaMK-II  calcium/calmodulin-dependent protein kinase-II  CHGB  chromogranin B  CKII  casein kinase II  CoA  coenzyme A  CPT-I  carnitine palmitoyl transferase I  CT  carboxytransferase  dpm  disintegrations per minute  DPNH  reduced diphosphopyridine nucleotide (see NADH)  DTNB  5,5’-dithiobis(2-nitrobenzoic acid)  DTT  dithiothreitol  EDTA  ethylenediaminetetraacetic acid  EGTA  ethylene glycol tetraacetic acid  FASN  fatty acid synthase  FPLC  fast protein liquid chromatography  FT  flowthrough  GFP  green fluorescent protein  GTP  guanosine 5’-triphosphate  ix  HEPES  4-(2-hydroxyethyl)-1-piperazine ethane sulfonic acid  HFHC  high fat high carbohydrate  HRP  horseradish peroxidase  IP  immunoprecipitation  kDa  kDaltons  LAS  liver ammonium sulfate  LC MS  liquid chromatography mass spectrometry  mA  milliamps  MCC  β-methycrotonyl-CoA carboxylase  MOPS  3-(N-morpholino)propane sulfonic acid  MW  molecular weight  NADH  nicotinamide adenine dinucleotide (reduced)  NADPH  nicotinamide adenine dinucleotide phosphate (reduced)  OAA  oxaloacetate  PBS  phosphate buffered saline  PC  pyruvate carboxylase  PCC  propionyl-CoA carboxylase  PDA  piperazine diacrylamide  PFA  paraformaldehyde  PFK-1 or -2  phosphofructokinase-1 or -2  PKA  cyclic AMP-dependent protein kinase  PLP  pyridoxal phosphate  PMSF  phenylmethanesulphonylfluoride  PVDF  polyvinylidene fluoride  rpm  revolutions per minute  SDS-PAGE  sodium dodecyl sulfate polyacrylamide gel electrophoresis  SREBP  sterol regulatory element binding protein  TCA  trichloroacetic acid  TFA  trifluoroacetic acid  TPN  triphosphopryidine nucleotide (see NADP)  TX-100  triton X-100  x  UV  ultraviolet  v/v  volume per volume  w/v  weight per volume  xi  Acknowledgements  I would like to thank my supervisor, Dr. Roger Brownsey for all the support, ideas, and help that he has given me over the last few years. I would like to express my gratitude to the following agencies that have provided financial support: University Graduate Fellowship, and the Natural Sciences and Engineering Research Council (NSERC). Special thanks go to the Canadian Institutes of Health Research (CIHR) for providing support for the laboratory in which the experiments were completed. I would like to thank Anna Maria Salzano, Shouming He, and especially Suzanne Perry for help with protocols, troubleshooting and data analysis for the mass spectrometry experiments. Thank you to Dr. Thomas Chang and Tony Kiang for the preparation of primary rat hepatocytes. I express my gratitude to Dr. Michel Roberge and the people in his laboratory for their help in the development of the immunofluroescence method. I would like to thank Dr. Robert Nabi and Pascal St-Pierre for their help with the confocal microscope. I would like to also thank my committee members Dr. Chris McIntosh and Dr. Pieter Cullis for their contributions. Much gratitude goes to Jerzy Kulpa for all his help in the laboratory. Thank you to all the people that I have worked with in the laboratory, past and present, for making these past few years much more enjoyable. Last, but not least, I would like to thank my husband Ken, and my two daughters Cassandra and Adrienna, for their continual patience and love.  xii  1  Chapter 1: Introduction  1.1 Biotin-dependent carboxylases Biotin, also known as vitamin H (for heat stable) or more commonly as vitamin B7 (figure 1.1) is an essential coenzyme that functions as a prosthetic group in a number of biological carboxylation reactions [1]. In humans, there are five known biotindependent carboxylases that use biotin as a cofactor: pyruvate carboxylase (PC), propionyl-CoA carboxylase (PCC), β-methylcrotonyl-CoA carboxylase (MCC), acetylCoA carboxylase-1 (ACC-1), and acetyl-CoA carboxylase (ACC-2). In each of these reactions, biotin is used to transfer a carboxyl moiety from one molecule to another and the reactions have a wide range of functions in metabolism. PC catalyzes the carboxylation of pyruvate to form oxaloacetate (figure 1.2b) and is therefore critical most directly for the generation of the TCA cycle intermediate (anaplerosis) and subsequently for further metabolism, especially for glucogenesis [2]. PCC catalyzes the carboxylation of propionyl-CoA to form methylmalonyl-CoA (figure 1.2c), an important intermediate in the catabolism of fatty acids containing an odd number of carbons, as well as in the catabolism of branched chain amino acids. MCC catalyzes the carboxylation of 3-methylcrotonyl-CoA to form 3-methylglutaconyl-CoA (figure 1.2d) which is involved in leucine catabolism. Finally, ACC-1 and ACC-2 are responsible for the carboxylation of acetyl-CoA to form malonyl-CoA (figure 1.2e) which is important in the synthesis of fatty acids, as well as the control of fatty acid oxidation. All five carboxylases require ATP, Mg2+ and bicarbonate as substrates. Of the five carboxylases, PC, PCC, and MCC are localized exclusively within the mitochondrial matrix, while ACC-1 and ACC-2 are mainly localized to the cytoplasm; a more detailed discussion of the localization of ACC is given later in this chapter.  1  Figure 1.1: The structure of biotin. The structures of biotin in three possible forms are shown: a) free biotin, b) carboxylated biotin, and c) biotinyl-lysine (also known as biocytin). a)  O HN  NH O S OH  b)  O O  -  O N  NH O S OH  c)  O HN  NH O NH  OH  S O  H2N  2  Figure 1.2: Reactions catalyzed by mammalian biotin-dependent carboxylases. The reactants and products are shown for a) the general acceptor, and for the reactions catalyzed by b) pyruvate carboxylase, c) propionyl-CoA carboxylase, d) βmethylcrotonyl-CoA carboxylase, and e) acetyl-CoA carboxylase 1 and 2. a)  ATP, HCO3-  ADP, Pi A-COO-  A b)  O  O H3C  O  -  -  -  O  OOC O Oxaloacetate  O Pyruvate c)  O  O H3C  H3C S-CoA  COO  Propionyl-CoA  -  S-CoA  D-methylmalonyl-CoA  d)  O  CH3  -  H3C  S-CoA  3-methylcrotonyl-CoA  CH3 OOC  O S-CoA  3-methylglutaconyl-CoA  e)  O  O H 3C  -  OOC  S-CoA  S-CoA  Acetyl-CoA  Malonyl-CoA  3  1.2 Acetyl-CoA Carboxylase 1.2.1 Discovery of ACC Acetyl-CoA carboxylase (ACC, EC 6.4.1.2) was discovered by Wakil et al in 1957 as a result of their studies targeted at elucidating the mechanism for long chain fatty acid synthesis. After determining that the formation of palmitate was not a reversal of βoxidation [3], they prepared four different enzyme fractions (R1, R2, R3, and R4) from pigeon liver, that upon complete recombination, were found to facilitate the synthesis of palmitate from small precursors [4]. In assessing the non-enzyme cofactors that might be involved in fatty acid synthesis, these authors found that ATP was absolutely required, together with Coenzyme A, DNPH (diphosphopyridine nucleotide, or NADH), TPN (triphosphopyridine nucleotide, or NADPH), manganese ions, isocitrate, and glutathione. Subsequent purification of the R1 enzyme fraction using gel adsorption chromatography and ammonium sulfate precipitation, was followed by the identification of biotin in this particular fraction [5]. The presence of biotin was unexpected and its role unclear as the fatty acid synthase enzymes were known not to require biotin. It was soon concluded that a separate biotin-dependent enzyme was involved in fatty acid synthesis [6] and that this new enzyme produced the malonyl-CoA required by fatty acid synthase for the chain elongation steps that would lead to the formation of palmitate. This biotin-containing enzyme was also found to require acetyl-CoA, Mn2+, ATP, and bicarbonate. In light of this sequence of discoveries, re-analysis of the pigeon liver protein factors indicated that the presence of only two of the original four enzyme fractions (R1 and R2) were essential for the synthesis of fatty acids from acetyl-CoA [7]. The importance of biotin in fatty acid synthesis was further explored by Wakil et al, one important observation being that the specific activity of fatty acid synthesis was directly proportional to biotin content. Furthermore, biotin was found to be covalently linked to one of the proteins in the enzyme fraction as tryptic digestion led to peptidebound forms of biotin, whereas boiling or acid digestion did not release free biotin. Ultimately, the discovery of biocytin (biotinyl-lysine) suggested that biotin might be attached to the epsilon amino group of a critical lysine residue of ACC and other biotindependent carboxylases [8]. Significantly, the addition of the biotin-binding protein avidin markedly inhibited fatty acid synthesis [9]. The avidin-biotin interaction, one of  4  the strongest of known biological interactions (Kd = 10-15 M) is now widely used to achieve desired interactions and is also extremely useful in the purification or recovery of ACC, as will be clear from experimental approaches used in this thesis. Another critical discovery made by Kallen and Lowenstein, and Vagelos et al, was that isocitrate and other carboxylic acids had an activating effect in the synthesis of palmitate from acetyl-CoA in tissue extracts, but no effect on the synthesis of palmitate from malonyl-CoA [10, 11]. It was therefore deduced that the enzyme that produced malonyl-CoA likely accounted for the activating effect of isocitrate on fatty acid synthesis. This enzyme was eventually named acetyl-CoA carboxylase and was further characterized by Waite and Wakil [12]. These authors confirmed that ATP and a divalent ion such as Mn2+ were required for ACC activity and that one mole of ATP was consumed per mole of malonyl-CoA formed. The addition of isocitrate also led to a five to six fold increase in the rate of malonyl-CoA production, confirming that isocitrate affected ACC rather than FASN. Further studies of the relative activities of ACC and FASN led to the conclusion that the activity of ACC contributed more strongly than that of FASN to control of the rate of overall fatty acid synthesis. ACC is therefore often considered to be “rate limiting” for fatty acid synthesis [13]. Following these early pioneering studies, many features of the structure, function and regulation of ACC have since emerged as outlined in more detail in this chapter. Some of the complexity and particular interest in ACC arises from the fact that the mammalian forms of the enzyme contain three functional units including the biotin cofactor and two active sites, one acting as the biotin carboxylase (BC), and the other as the carboxytransferase (CT) [14]. In ACC from animals, plants and yeast, all these functions are contained within a single multifunctional polypeptide [15]. The emergence of the multifunctional forms of ACC from the multi-component bacterial forms of the enzyme has provided an interesting model of molecular evolution [16].  5  1.2.2 ACC reaction mechanism and kinetics The carboxylation of acetyl-CoA occurs in two separate steps or partial reactions [17] catalyzed by two different domains of ACC (figure 1.3). The first step is catalyzed by biotin carboxylase and involves the ATP-dependent addition of a carboxyl group from bicarbonate to a ring nitrogen of the biotin attached to ACC (figure 1.1b). In the second step, the carboxyl group on biotin is transferred to acetyl-CoA to form the final product of malonyl-CoA. ACC can also catalyze the reverse reaction of malonyl-CoA decarboxylation and this can occur in vitro independently of biotin and proceeds with no re-synthesis of ATP [18]. Within intact cells, however, it is not clear to what extent ACC contributes to this reverse reaction and it is generally considered that a discrete malonyl-CoA decarboxylase (MCD) is used to degrade malonyl-CoA to form acetyl-CoA and carbon dioxide [19]. The enzyme MCD and its access to malonyl-CoA is highly regulated, but this is not completely understood [20].  Figure 1.3: Reaction scheme of ACC HCO3-  O O  -  ATP  ADP  Pi  ACC  Biotin Carboxylase  ACC  Biotin  Carboxyltransferase  Biotin  O  O  S-CoA Malonyl-CoA  COO-  H3C  S-CoA  Acetyl-CoA  The role of biotin raises a number of important and interesting points. The biotin is attached to apo-ACC (and other apo-carboxylases) through the ε-amino group of a critical lysine residue (figure 1.1c) which in mammalian ACC is located about one third of the way through the primary sequence and within a very characteristic “MKM” motif [21]. In human ACC-1, the biotinylated residue is Lys-786, and within the sequence  6  EIEVMKMVMTL, while in human ACC-2, Lys-929 occurs within the sequence EMEVMKMIMTL [22]. In the rat ACC-1 and ACC-2 isoforms, the corresponding lysines are 785 and 925 within the sequences EIEVMKMVMTL and EMEVMKMIMTL, respectively [8]. The attachment of biotin to ACC and other carboxylases is catalyzed posttranslationally by the mammalian enzyme biotin protein ligase, also referred to as holocarboxylase synthetase [22]. In a study by Shriver et al, in which rats were fed a biotin-deficient diet for approximately 50 days, there was no significant change in levels of cytosolic liver holo-ACC protein and activity over the whole experimental period [23]. In fact, there was only a ten percent decrease in the level of total biotinylated proteins in the liver, indicating an efficient mechanism for biotin recycling. In adipose tissue, however, Jacobs et al demonstrated a decrease in holo-ACC, with an accumulation of apo-ACC over a period of 14 days [24]. Biotin deficiency through poor nutrition is rare, as the requirements are rather low in humans (30μg/day) and biotin is absorbed readily from a variety of sources [25]. The biotinyl arm of ACC is generally considered to have substantial mobility in that it must be able to interact with the two active sites located within the BC and CT domains. The respective binding sites for d-biotin have been defined in both BC and CT domains and in fact, free biotin or carboxybiotin can also be used as a substrate for the separate BC and CT structures of E. coli ACC. It is interesting to note that it is generally much easier to assess the partial reactions of bacterial forms of ACC because the multisubunit proteins readily dissociate upon cell disruption rendering full ACC function and production of malonyl-CoA difficult to determine. Despite the potential movement of the biotin arm, an early study by St. Maurice et al indicates that in PC, biotin movement is quite limited [26]. This may be an important reason for ACC to exist in dimeric form, as biotin may visit BC on one subunit and CT on the other to minimize the distance needed for the biotin arm to travel. Details of the BC reaction mechanism have emerged from extensive studies of acetyl-CoA carboxylase from E. coli [27] and also from the study of other biotin enzymes, notably of pyruvate carboxylase, for which biotin carboxylation is a common first step regardless of the ultimate carboxyl acceptor [28]. ATP and bicarbonate bind in  7  order, to their respective binding sites within the BC active site of ACC. The gamma phosphate of ATP is then added to bicarbonate to form carboxyphosphate which remains bound within the BC active site of ACC. Based on the properties of ACC, pyruvate carboxylase and of the related carbamoyl-phosphate synthetase (CPS), it is likely that the carboxyl group of carboxyphosphate is then transferred to the 1’ nitrogen of biotin. Indeed, carboxybiotin has been “trapped” chemically and its structure has been confirmed [29]. Following the transfer of the carboxyl group, ADP and Pi are released from the BC active site. The sequence of steps involved in the binding of substrates and release of products during biotin carboxylation has been determined by a combination of kinetic measurements, assessment of exchange reactions (exchange of 14C between bicarbonate and carboxybiotin, and exchange of 32P between ATP and phosphate) and by studying the effects of inhibition with modified forms of biotin, such as phosphonoacetyl biotin [30]. In comparison to BC, much less is known about the kinetic reaction mechanism of CT, although it appears from studies of the reverse reaction that binding and release of substrates is also ordered; the release of malonyl-CoA being followed by release of biocytin [31]. In E. coli ACC, based on the pH dependence of the CT reaction, it was concluded that a critical cysteine residue may be involved in catalysis [31]. The kinetics of several animal forms of ACC have been studied, including studies of partial reactions as well as the overall composite reactions. Because of the susceptibility of animal ACCs to proteolysis, early characterization of the enzyme has to be interpreted with some caution – at least in those cases where there is no evidence to confirm that the ACC subunits remain fully intact. In particular, in early studies, ACC was reported to have subunit sizes of approximately 120-125 kDa, clearly representing cleaved forms of the enzyme [21]. In addition, loss of an N-terminal portion was found to significantly enhance enzyme activity suggesting an inhibitory function for that subdomain [32]. Using rat liver as their primary source of ACC, Hashimoto and Numa used three different assays to elucidate the reaction mechanism of ACC [33]. The first was a coupled assay in which the forward reaction of ACC was coupled with pyruvate kinase and lactate dehydrogenase, to allow monitoring of the oxidation of NADH. The same coupled assay was used for the reverse reaction of ACC, following the reduction of NAD+ to NADH. For certain experiments involving ATP hydrolysis, [32P]-Pi levels were  8  measured. Based on the initial rates of reaction measured from the initial slopes of NADH oxidation (or NAD+ reduction) it was concluded that a bi bi uni uni ping pong (double displacement) mechanism best fit the obtained results (figure 1.4a). In this particular reaction mechanism, ATP and bicarbonate bind first to ACC, and ADP and Pi are released following carboxylation of the biotin arm. Following the release of ADP and Pi, acetyl-CoA then binds to the enzyme prior to carboxylation to form malonyl-CoA. In 1982, Beaty and Lane used steady state kinetics to further characterize the kinetic mechanism of ACC isolated from chicken liver. Instead of using a coupled assay, they used a [14C]-HCO3 fixation assay that measured the incorporation of [14C]-HCO3 into malonyl-CoA. Their results led them to conclude that the reaction mechanism of ACC fits a ter ter mechanism with the formation of a quaternary complex (figure 1.4b), rather than the double displacement. In this concerted reaction mechanism, all three substrates (ATP, bicarbonate, and acetyl-CoA) bind to ACC in no specific order, forming a quaternary complex, followed by release of the products after the carboxylation of acetyl-CoA (malonyl-CoA, ADP, Pi). These authors also pointed out that the difference in reaction mechanism could be a result of difference in tissue source, or that difference in isolation methods may lead to versions of ACC that can have different kinetic mechanisms. In view of the potential for proteolysis, this caution is important.  9  Figure 1.4: Two postulated kinetic mechanisms for the ACC reaction. a) bi bi uni uni ping pong and (b) ordered ter ter mechanism.  a)  b)  A similar debate has taken place in considering the reaction mechanism of pyruvate carboxylase (PC). McClure et al used a coupled assay to determine the substrate kinetics and the effects of a competitive product inhibitor of PC isolated from rat liver [28] and concluded that the reaction mechanism is bi bi uni uni ping pong. Later, Ashman and Keech used an isotope exchange assay to confirm the same kinetic mechanism using PC from sheep mitochondria [34] and Barden et al confirmed the same mechanism for chicken liver [35]. In contrast, based on studies of the substrate kinetics of PC and the activating effect of acetyl-CoA and free Mg2+ ions, Warren and Tipton concluded that the reaction mechanism of PC from pig liver was most likely a sequential ter ter mechanism [36]. Later, Easterbrook-Smith et al performed a series of kinetic experiments on PC that also provided support for a sequential mechanism [2]. Studies of the PC reaction mechanism can provide insights into the possible reaction mechanisms of ACC since PC and ACC catalyze very similar reactions but a  10  definitive mechanism for ACC is still not absolutely agreed upon. In their discussion, Beaty and Lane suggested that the differences in reaction mechanisms for ACC may be due to studies done on enzyme from different sources as they used chicken liver, while Hasimoto and Numa used rat liver. This is not a convincing argument because similar mechanisms have been deduced for PC from several sources. On balance, it seems likely that detailed structural studies of ACC with various substrate combinations will be required to resolve this debate. It is worth considering that the kinetics of ACC may potentially be more complicated than those of PC, especially considering the range of allosteric ligands and the role of multiple site phosphorylation that may influence ACC.  11  1.2.3 Evolution of the structure and properties of ACC ACC is found in all free-living organisms that have been investigated, ranging from bacteria to higher-order organisms including plants and mammals [14]. The only organisms so far found to lack ACC are certain obligate parasites that are able to bypass the metabolic requirements for fatty acid biosynthesis. In free-living bacteria and unicellular organisms such as yeast, ACC is essential for normal growth and replication [37-39]. In bacteria, ACC exists as a multi subunit complex made up of two to four discrete polypeptides encoded by separate genes. E. coli ACC is composed of four different subunits: one subunit contains the BC component, one is known as the biotin carboxyl carrier protein (BCCP), and the two remaining subunits, CT-α and CT-β, make up the carboxyltransferase component [16]. In the native state, the biotin carboxylase is actually a homodimer of two BC subunits, the native form of carboxyl transferase is an α2β2 tetrameric complex, and BCCP also forms a dimer [40]. Therefore, overall, E. coli ACC contains eight polypeptide chains and the molecular mass is on the order of 270 kDa. The genes for bacterial ACC subunits often occur in a gene cluster or operon. In E coli, the BC gene (accC) is immediately downstream of the BCCP gene (accB) and hence both genes are co-transcribed. The CT-α and CT-β genes (accA and accD, respectively) are found at a different genetic locus. The transcription of the four ACC subunits is tightly regulated, and the proteins are expressed at equal molar ratios but the mechanisms which regulate this transcription and subunit synthesis are currently unknown. In E. coli, it appears that ACC activity is controlled mainly through gene transcription, although some experiments have shown that the acyl carrier protein, a key component for fatty acid biosynthesis, can inhibit ACC activity. A key cysteine residue exists in the CT subunit of ACC, and this can potentially be regulated through redox activity. The main role of bacterial ACC is to form malonyl-CoA for fatty acid synthesis to support membrane lipid production. Although ACC was first discovered in animals, many groups have studied the bacterial form of ACC for a variety of practical reasons, not least being the ability to express quantities for enzymatic and structural studies, for genetic manipulation and because the individual components can be studied in isolation as well as in combinations.  12  From an evolutionary perspective, the ACC of some bacterial species such as mycobacterium tuberculosis, corynebacterium glutamicum, saccharopolyspora erythraea, and streptomyces venezuelae presents an interesting case as the BC and BCCP subunits are fused to form one subunit designated α, while the carboxyltransferase forms another subunit designated β [41, 42]. One small subunit, labelled ε, is also present which acts to enhance the overall activity of the carboxylases [43]. The subunit composition and regulation of the ACC of these bacterial species is not well-understood and will not be further discussed in this thesis. The yeast form of ACC is expressed as a single large polypeptide containing all functional domains [15]. In solution, yeast ACC forms a homo-oligomer, most likely an octamer with a molecular mass on the order of 2000 kDa. In addition to the ACC that functions in classical cytosolic fatty acid synthesis in yeast, an additional mitochondrial form of ACC has also been recognized that is encoded by the Hfa1 gene. The functions of this mitochondrial ACC, for which no counterpart has been reported in animal tissues, remains to be fully established. While both yeast forms of ACC contain all the important functional domains, the two proteins are not overlapping in functions and Hfa1 cannot compensate for a lack of ACC function [39]. As in bacteria, regulation of the expression of yeast ACC is coordinated with the rate of cell growth and therefore the need for membrane phospholipid biosynthesis, falling under the control of the transcription factors Ino2/Ino4 and Opi1. In addition to control of protein expression, yeast ACC may also be subject to acute controls, being susceptible to histone acetylation and also able to undergo homo-oligomerization upon activation. The complex allosteric controls found with mammalian ACC isoforms have not yet been demonstrated for yeast ACC and although citrate can cause some aggregation of yeast ACC, no corresponding effect on ACC activity has been found [15]. The yeast ACC protein can also be phosphorylated, notably by Snf-1, which is the yeast counterpart of mammalian AMPK and also, in vitro at least, by mammalian AMPK and PKA [44]. However, a subsequent study demonstrated that yeast ACC does not contain the AMPK phosphorylation site [45], indicating the possibility of an alternate AMPK site on yeast ACC. The expression of ACC in plants presents a very complex picture, particularly as plants of different genera exhibit different profiles of both eukaryotic (homomeric) and  13  prokaryotic (heteromeric) forms of ACC [46]. The term homomeric is used to describe forms of ACC that are expressed as single large polypeptides that contains all functional domains and are typically located in the cytosol of plant cells. The term heteromeric refers to the multi-subunit forms of ACC, comparable to the multi-subunit bacterial forms of ACC, and these are typically located in the plastids of plant cells. The heteromeric forms of plant ACC mainly function to form malonyl-CoA for fatty acid synthesis, while the homomeric form of ACC provide malonyl-CoA for a number of cellular processes such as elongation of fatty acids and malonylation of a variety of carbon structures by the polyketide synthases that catalyze the formation of a broad range of “secondary metabolites” [47]. Purification of the heteromeric form of ACC from plants has so far not been achieved and the mechanism of regulation of gene expression has yet to be elucidated. Depending on the plant species, the expression of the homomeric forms of ACC depends on various metabolic cues. The catalytic activity of plant ACCs is sensitive to pH, Mg2+ ions, redox of key amino acids but no effects of carboxylic acids such as citrate have been found. Intriguingly, plant ACC is also controlled according to a circadian rhythm geared especially to the light/dark cycle and corresponding signals. In particular, there is evidence that light can stimulate a kinase that can phosphorylate the homomeric ACC. It is interesting to consider if this fundamental biological control might also apply to animal forms of ACC which also appear to respond to circadian controls [48]. For example, lipogenesis levels may vary according to dietary habits, and thus experiments involving animals should be timed according to their dietary habits. In animals, ACC exists in a number of isoforms, all of which are expressed as large multifunctional polypeptides which also associate into higher oligomeric structures. Animal ACCs are highly regulated by a number of mechanisms including allosteric and covalent mechanisms, including reversible multiple-site phosphorylation that changes dynamically in response to hormones. The allosteric control of animal ACCs is particularly striking, the activation that follows exposure to citrate being accompanied by a dramatic polymerization. A detailed description of the regulation of the animal forms of ACC will be given in subsequent sections of this chapter.  14  Despite the vast differences in ACC characteristics in different organisms, the functional domains of ACC are highly conserved. Figure 1.5 shows the alignment of the functional domains of ACC from various organisms. The biotin carboxylase domain is highly conserved, typically extends approximately 500 amino acid residues and is located within the N-terminal third of the protein. This BC domain also contains sub-domains characteristic of ATP binding (ATP grasp) and associated specific residues as well as bicarbonate binding and catalytic residues. The attachment of biotin occurs at a highly conserved sub-domain and within a specific amino acid context. The ACC central region is also highly conserved and can be found only in the eukaryotic forms of ACC. While the role of the central region is unclear, it is possible that this region links the BC and CT domains together as this region is not present in prokaryotic forms of ACC. Figure 1.6 shows the amino acid alignment of the putative ATP binding site, the BC active site, and the biotinylated lysine, and the surrounding sequence, for human, rat and yeast ACC. Interestingly, the actual sites and their surrounding sequences are quite highly conserved between all three species Table 1.1 lists the location for substrate or cofactor binding for human, rat, E. coli, and yeast forms of ACC that have been determined experimentally and/or predicted based on sequence alignment and function association.  15  Figure 1.5: Domain structures of ACC from various organisms [49]. ACC from Homo sapiens (human), Rattus norvegicus (rat), Zea mays (sweet corn), Drosophila melanogaster (fruit fly), Saccharomyces cerevisiae (yeast), and Escherichia coli (bacteria). The colour scheme indicates functional domains of ACC and as follows: red – ATP-binding site, blue – ADP formation site, green – biotin carboxylase C-terminal domain, black – biotin binding site, and yellow – ACC central region. Letters indicate protein domains as follows: A – AccC (biotin carboxylase), B – AccB (biotin carboxyl carrier protein), and C – carboxyl transferase. Note that E. coli ACC is expressed in four separate polypeptides, the products of four genes: AccA, AccB, AccC, and AccD.  16  Figure 1.6: Key features conserved in ACC from human, rat, and S. cerevisiae. Partial sequences of the ATP binding site, ADP formation, and biotin binding subdomains are shown in the same colour scheme as figure 1.5 [50]. In particular, important residues highlighted include: ATP binding site (single underline), BC active site (double-underline), and biotinylated lysine (*). Grey represents sequences that are highly conserved but do not belong to a specific subdomain. human rat yeast  301  human rat yeast  361  human rat yeast  421  human rat yeast  …772  301 241  361 301  421 362  …771 …721  AEEVGYPVMIKASEGGGGKGIRKVNNADDFPNLFRQVQAEVPGSPIFVMRLAKQSRHLEV EEVGYPVMIKASEGGGGKGIRKVNNADDFPNLFRQVQAEVPGSPIFVMRLAKQSRHLEVQ KAKRIGFPVMIKASEGGGGKGIRQVEREEDFIALYHQAANEIPGSPIFIMKLAGRARHLE QILADQYGNAISLFGRDCSVQRRHQKIIEEAPATIATPAVFEHMEQCAVKLAKMVGYVSA ILADQYGNAISLFGRDCSVQRRHQKIIEEAPAAIATPAVFEHMEQCAVKLAKMVGYVSAG VQLLADQYGTNISLFGRDCSVQRRHQKIIEEAPVTIAKAETFHEMEKAAVRLGKLVGYVS GTVEYLYSQDRSFYFLELNPRLQVEHPCTEMVADVNLPAAQLQIAMGIPLYRIKDIRMMY… TVEYLYSQDGSFYFLELNPRLQVEHPCTEMVADVNLPAAQLQIAMGIPLFRIKDIRMMY… GTVEYLYSHDDGKFYFLELNPRLQVEHPTTEMVSGVNLPAAQLQIAMGIPMHRISDIRT… HVLAGQCYAEIEVMK*MVMTLTAVESGCIHYVKRPGAALDPGCVLAKMQLDNPSKVQQAE HVFAGQCYAEIEVMK*MVMTLTAVESGCIHYVKRPGAALDPGCVIAKMQLDNPSKVQQAE HIIKGQPYAEIEVMK*MQMPLVSQENGIVQLLKQPGSTIVAGDIMAIMTLDDPSKVKHAL  17  Table 1.1: Some critical conserved residues of ACC [51]. ATP  BC Active  Mn2+  Biotinylated  CT, CoA  binding site  site  binding sites  Lysine  binding site  315-320*  441*  424 or 437*,  786  1823, 2127,  Human ACC-1  437 or 439* ACC-2  458-463*  584*  567 or 580*,  2129* 929*  580 or 582*  1934, 2238, 2240*  Rat ACC-1  314-319*  440  423 or 436*,  785  436 or 438* ACC-2  1822, 2126, 2128  ND1  ND  ND  925*  ND  256-261*  383*  365 or 379*,  735  1731, 2034,  S. cerevisiae ACC-1  379 or 381* Hfa1  332-337*  459*  ND  2036 804  1776, 2080, 2082  E. coli BC  116, 201,  292*  ND  236* 1  (ND) indicates sequence features that are unknown.  *  Residues are predicted but not confirmed experimentally  122  ND  (of BCCP)  Comparison of the primary sequences of ACC-1 from different animal species reveals approximately 90% amino acid sequence identity (including a comparison between human ACC-1 and rat ACC-1, for example), while there is a 35% amino acid sequence identity between animal ACC-1 and yeast ACC [52]. The homology between forms of ACC-1 is even higher when allowing for conservative changes in which amino acids are replaced with a chemically similar amino acid.  18  1.2.4 ACC structure Due to the large size of mammalian ACC and the inability to express large amounts of recombinant protein, the complete three dimensional protein structure has not yet been established. However, the high resolution structures of specific domains of individual subunits of bacterial ACC have been reported based on the results of X-ray crystallography. For example, high resolution structures have been reported for the BC domain of E. coli [53], yeast [54], and most recently of human ACC2 [45]. The BCCP subunit structure has been determined in E. coli [55] and human, by NMR [56] and the structure of the CT domain has been determined for yeast ACC [57]. Studies of E. coli ACC have been more tractable because of the smaller size of the individual subunits. The first high resolution structural information was therefore obtained for the biotin carboxylase domain of E. coli ACC (figure 1.7a). This first report in 1994 by Waldrop et al, showed that the BC subunits associated as a dimer in the crystal structure. The BC subunits of E. coli ACC contain at least three major domains, designated “A”, “B”, and “C”. The A domain has a dinucleotide binding motif (ATP grasp fold) containing two sets of α-helices sandwiching a parallel β-sheet. The B domain was found to be very flexible and changed conformation upon ATP binding to the A domain. The C-domain is a globular portion of the subunit and contains one eightstranded antiparallel β-sheet, an additional three-stranded antiparallel β-sheet and seven α-helices. The corresponding structure of the yeast BC domain was reported in 2004 by Shen et al (figure 1.7b). E. coli and yeast BC have only a 35% amino acid sequence identity, and attempts to use the E. coli structure to elucidate the yeast BC domain structure through molecular replacement were unsuccessful. The general structures of the two BC domain/subunits are similar, each having three subdomains, with the ATP grasp fold in the A subdomain, and a conformational change of the B subdomain being observed upon ATP binding. However, there were many structural differences between the two BC domain/subunits, mostly in the C subdomain. In yeast BC, there are nine βsheets surrounded by 9 α-helices, while in E. coli, there are eight β-strands surrounded by 7 α-helices.  19  The structure of the recombinant form of the BC domain of human ACC2 was reported in 2008 by Cho et al (figure 1.7c). These authors found that the structure of the BC domain of human ACC2 was similar to that of the corresponding yeast structure. As with the two earlier structures, the BC domain of human ACC2 also contains three subdomains (A, B, and C). Again, the ATP grasp fold is in the A subdomain and the B subdomain showed indications of conformational change upon ATP binding. Despite the similarities, there are, however, some potentially important differences between the Nterminal domain of the yeast and human BC structures. Most notably, the N-terminal 200 amino acids of the human ACC2 BC domain are not present in the yeast ACC structure.  20  Figure 1.7: Crystal structures of biotin carboxylase. a) E. coli [53], b) yeast [54], and c) human ACC2 [45]. E. coli ACC is shown as a dimer, whereas the other two structures are shown as monomeric. a)  b)  c)  21  Currently, the structure for the BCCP subunit has been determined only for E. coli ACC (figure 1.8) [55]. The BCCP subunit contains two sets of four antiparallel βstrands. The lysine to which biotin is added to the subunit is found on a hairpin turn between β-strands 4 and 5. While the structure of BCCP in E. coli can give some insight to the other forms of ACC, it is important to keep in mind that the E. coli enzyme lacks a large portion of the ACC central region, found in eukaryotic forms of ACC, a region that has a major influence on the corresponding BCCP structure.  Figure 1.8: X-ray crystal structure for the BCCP subunit of E coli [55]. Biotin (ball and stick format) is represented in green, while orange indicates the β-sheet structure.  22  The structures of E. coli ACC CT subunits have been more difficult to elucidate because they have been found to be quite labile and therefore difficult to purify and to crystallize. In contrast to other subunits of ACC, the CT domain structure has only been determined for yeast ACC (figure 1.9) [57]. Based on these studies, the yeast CT domain appears to contain two subdomains, labeled “N” and “C”, that show sequence homology to the β and α subunits, respectively, of the bacterial ACC. In the crystal structure, the CT domain exists as a dimer, with the N subdomain of one subunit interacting closely with the C subdomain of the partner subunit. Significantly, a key cysteine residue thought to be involved in catalysis based on studies of bacterial ACC is not found in the yeast CT structure. It has been of considerable interest to define how these herbicides act on the CT domain and to determine the mechanism of inhibition. Herbicides, such as haloxyfop, act mainly on the plastid or heteromeric form of ACC [58]. Several studies have addressed the inhibition mechanism through structural determination. For example, a structure was reported in 2004 for the complex formed between yeast CT and haloxyfop, one of the potent herbicides [59]. In this particular structure, the binding of haloxyfop with the CT subunit led to a conformation change in Tyr-1738 and Phe-1956 that led to formation of a hydrophobic pocket to which the herbicide was bound. The formation of this pocket has a profound effect on the organization of key catalytic residues, explaining the profound effect of the herbicides on ACC catalytic activity. Due to the significant conformational change required for this pocket to form, it was thought that factors that can regulate this conformation change would alter the sensitivity of ACC to this herbicide.  23  Figure 1.9: X-ray crystal structures for the CT subunit of S. cerevisiae ACC in complex with: a) coenzyme A [57], and b) the herbicide, haloxyfop [59]. In panel a, two subunits (Mol1 and Mol2) are shown and each contains two subdomains (N and C). Green and purple indicate subdomains of one subunit, and cyan and yellow represent the subdomains of the second subunit. In panel b, only one subunit is shown. a)  b)  24  1.2.5 ACC isoforms in animal cells In view of the focus of the work described in this thesis, most of the subsequent introduction will focus on the two major forms of rat ACC. The isoform of ACC that was discovered first is now referred to as ACC-1, ACC-α, ACACA, or ACC-265 because its subunit size is 265 kDa [60]. The second major isoform, ACC-2, ACACB or ACC-β was discovered in 1990, and is sometimes referred to as ACC-280 because its subunit size is 280 kDa [61]. The sequence of the human form of ACC-2 was reported in 1996 [62]. Variants of the ACC isoforms have also been found [63, 64] (see table 1.2). In humans, there are four variants of ACC-1 arising from alternative promoters and alternative splicing, and similarly, two minor isoforms of ACC-2 due to alternative mRNA splicing. In rats, there are two variants of ACC-1 due to alternative mRNA splicing [51]. Interestingly, the two variants in rats differ in a critical coding sequence for the phosphorylation site of Ser-1200, one isoform lacking this specific site [65]. These splice variants were determined at the gene level, and little is known about the expression levels of these splice variants.  Table 1.2: Amino acid sequence variants of ACC isoforms [51]. ACC type Variants1 Human ACC-1  Residues 1-75 replaced with 17 amino acids Residues 1-78 are missing Addition of 37 amino acids between residue 1 and 2  Human ACC-2  Residues 1118-1187 are missing  Rat ACC-1  Residues 1189-1196 are missing  1  relative to the “canonical” sequence of the dominant form of ACC.  Both ACC isoforms have so far been determined to be mainly cytosolic when tissues or cells are homogenized and cell fractions prepared by centrifugation [61]. ACC is unique in that it is the only biotinylated carboxylase to be cytosolic, while other biotinylated carboxylases are found in the mitochondrial matrix [1]. The situation is more complex in yeast, however, which contains a mitochondrial ACC, the product of the HFA1 gene [15]. The situation in plants is even more complex because of the expression  25  of both eukaryotic and prokaryotic forms of ACC and localization either in the cytosol or within plastids [46]. Upon discovery of the second animal ACC isoform, it was postulated that the two isoforms might be transcribed from the same gene through differential splicing and alternative mRNA species. However, it was subsequently discovered in mammals that these two isoforms come from distinct genes on different chromosomes. This was also predicted by Winz et al when they showed that ACC-1 and ACC-2 differ very substantially in sequence based on mass spectrometry and peptide sequencing [66]. The chromosomal locations of the ACC gene in different organisms can be found in table 1.3 [50].  Table 1.3: Chromosomal location of the ACC gene in different organisms. Species  ACC-1  ACC-2  Human  17q21  12q24  Rat  10q26  12q16  Mouse  11C  5F  Yeast  XIV  XIII  The human ACC isoforms exhibit 60% nucleotide sequence identity, and 80% amino acid sequence identity [52] and when the two isoforms are aligned based on amino acid identity, the major functional sites are also spaced at similar distances. The N-terminal 140 amino acids in ACC2 represents a sequence not present at all in ACC-1 [52, 67], and this largely accounts for the additional 15 kDa size difference between the two isoforms. This unique N-terminal sequence of ACC-2 is especially enriched in amino acids containing 19 prolines, 108 hydrophobic amino acids, and 19 basic amino acids. The first twenty-five amino acids are mostly hydrophobic residues surrounded by positively charged amino acids and may serve as a mitochondrial targeting and/or anchoring sequence: MVLLLFLTYLVFSCLTISWKIWGK Through the use of fluorescence antibodies, Abu-Elheiga et al have found that ACC-2 was localized to the mitochondrial outer membrane [68], mostly dictated by the first 20  26  amino acid residues of the N-terminal end of ACC-2. In their experiments, they developed a construct in which the N-terminal 416 residues of ACC-2 was linked to GFP, and then expressed the construct in human HepG2 cells. Antibodies specific to ACC-2 as well as GFP fluorescence were used to visualize the location of ACC-2 in the cells by fluorescence microscopy. After confirmation with organelle markers, it was demonstrated that ACC-2 was targeted to the mitochondrial membrane. Similar experiments with ACC-1 showed that it remained in the cytosol. It has been argued that ACC-1 and ACC-2 have different functions and the different tissue distribution seems to support this idea. ACC-1 is found predominantly in lipogenic tissues such as liver, adipose tissue, and mammary glands [69]. In rat liver, ACC-2 represents only 20-25% of total ACC [61], so that most hepatic malonyl-CoA is produced by ACC-1 [38]. In rat mammary glands, brown [70] and white adipose tissue [61] ACC-1 is the major or sole isoform expressed. ACC-2, on the other hand, is the predominant isoform expressed in oxidative, non-lipogenic tissues such as skeletal muscle and the heart ([38, 70]. In these oxidative tissues, ACC-2 accounts for approximately ninety percent of total ACC [70]. For heart tissue, therefore, most of the malonyl-CoA is produced by ACC-2. Brain appears to contain ACC-1 and its expression and control is poorly understood [71]. In general, lipogenic tissues such as liver, mammary gland and white adipose tissue total ACC expression is relatively high (approximately 10-50 μg/g wet tissue), while expression of ACC in skeletal muscle and heart is much lower (approximately 1-2 μg/g wet tissue) [72]. Based on these expression levels, it is interesting to note that although it represents a lower proportion of hepatic ACC, ACC-2 is actually expressed at a much higher level in rat liver than in rat muscle or heart [61]. The possibility that the two ACC isoforms produce malonyl-CoA for different purposes has also been addressed in a series of experiments by Wakil et al in ACC-2 knockout mice in which basal rates of fatty acid oxidation were enhanced and the animals exhibited a leaner body mass despite being hyperphagic [38]. Levels of malonyl-CoA of these knockout mice were decreased in the heart and skeletal muscle, where ACC-2 is normally the predominant isoform, whereas in the liver and adipose tissue, where ACC-1 is the predominant isoform, there was no change in malonyl-CoA levels or in total ACC  27  activity. This indicates that malonyl-CoA produced by ACC-2 does have an appreciable overall effect on whole-body beta-oxidation. Interestingly, somewhat conflicting evidence was reported in shorter-term studies by Lee et al. in which it was shown that when ACC-2 expression was reduced using antisense RNA molecules in H9c2 cells, there was a decrease in muscle cell differentiation but no change in fatty acid oxidation [73]. Further experiments with the ACC-2 knockout mice revealed that despite feeding a high fat and high carbohydrate (HFHC) diet, these mice did not become obese and diabetic, unlike their wildtype counterparts. Correspondingly, the HFHC diet led to suppression of beta-oxidation and enhanced lipid storage in wild-type mice, but not in ACC-2 knockout mice [74]. These results provide further support for the idea that the two ACC isoforms produce malonyl-CoA for different purposes. The effects of HFHC diet on adipose tissue of the mice were also interesting. In fact, HFHC diets led to reduced fatty acid oxidation in adipose tissue of both wild-type and ACC-2 knockout mice but absolute levels were higher in the ACC-2 knockout mice on both diets. Other regulators, such as leptin, were thought to be responsible for the change in FA oxidation levels in the adipose tissue. An attempt to produce ACC-1 knockout mice revealed that the deletion of the ACC-1 gene cannot be tolerated and leads to embryonic lethality [37]. ACC-1 (+/-) heterozygous mice were shown to be similar to their wildtype counterparts in terms of tissue malonyl-CoA concentrations, ACC protein, and FA oxidation levels [37]. Interestingly, mice with liver specific ACC-1 knockout were viable under normal feeding conditions [75]. When subjected to starvation followed by refeeding for 48 hours, these mice had lower triglyceride amounts in the liver, as well as up to 70% lower ACC activity and malonyl-CoA levels, in comparison to their wildtype counterparts. However, there was no change in FA oxidation rates and a HFHC diet in these liver specific ACC-1 knockout mice still led to a fatty liver and insulin resistance. Again, these results demonstrate a difference in the mechanism of action of the malonyl-CoA produced from the two ACC isoforms, although malonyl-CoA from ACC-2 can clearly be used for fatty acid synthesis in the absence of ACC-1.  28  1.2.6 ACC regulation 1.2.6.1 Regulation of ACC expression Consistent with the importance of controlling ACC activity, expression of the ACC genes and proteins is highly regulated by a number of different mechanisms. Early studies on ACC expression in which protein and mRNA levels were measured following dietary changes (e.g., fasting or refeeding; high or low fat feeding, etc.) showed that changes in mRNA levels can be detected within an hour. The half-life of ACC mRNA has been measured to be in the range of four to six hours [76]. Early studies of ACC protein synthesis and degradation were reported by several groups including Nakanishi et al and Majerus et al [77, 78]. Using isotopic leucine incorporation and immunochemical titration experiments, they showed that changes in total ACC activity correlated well with immunodetectable protein, and they inferred there were no changes in ACC catalytic efficiency. Upon examination of the effects of different dietary conditions on the levels of ACC protein, they showed that in insulin-deficient diabetic rats, there was a decreased rate of synthesis, while there was an increased rate of synthesis in rats that were re-fed after fasting. In both conditions, there was no change in rate of ACC degradation. However, in rats that were starved, there was no change in ACC synthesis, but there was an increase in ACC degradation. These studies therefore provided evidence that synthesis and degradation of the ACC protein were both controlled and contributed differently in different physiological settings. The half-life of the ACC protein in fed rats was found to be approximately 48 hours, while in rats that were starved, the half-life of the ACC protein was only 18 hours, confirming enhanced rates of degradation in starvation. The control of rat ACC gene transcription has also been found to be complex. The rat gene for ACC-1 contains at least two identified promoters, PI and PII which form class I and class II transcripts, respectively and that lead to the formation of various combinations of the five possible exons (figure 1.10a) [69]. While both classes of transcripts lead to the generation of the same-sized protein, they are involved in control of ACC expression under different physiological conditions. The PI promoter, which contains a TATA box and a CAAT box, is known to be stimulated under lipogenic conditions. There is also an intrinsic repressor found near the PI promoter and as a result,  29  the expression of PI under non-stimulating conditions has been measured to be about 70 percent of the maximal rate. The PII promoter does not contain TATA or CAAT boxes, and has been found to be more responsive to glucose stimulation. PII is also responsive to the sterol regulatory element binding protein (SREBP-1) and both promoters can be stimulated by insulin. The control of the human ACC-1 gene is likely even more complex than that of the rat gene as it contains three promoters and at least 6 major exons (figure 1.10b) [79]. These three promoters are designated as PI, PII, and PIII, but differ from the PI and PII promoters of the rat ACC gene. For human ACC, PI is constitutively active, PII is regulated by hormones (e.g., leptin) and by SREBP, and PIII plays a different tissuespecific role. In the experiments outlined by Mao et al, the PIII promoter activity was very low in HepG2 cells, whereas in the human mammary cancer cell line T-47D, the PIII promoter was highly active indicating that the PIII promoter might play a role in the generation of milk fat during lactation [79]. Not surprisingly, in contrast to the extensive studies of ACC-1 expression, relatively few studies have so far been done on the control of ACC-2 expression. Both the rat and human ACC-2 genes contain two promoters, the relative importance of which is tissue-specific. The PI promoter appears to play a dominant role in the control of ACC-2 expression in both rat and human heart. This PI promoter is activated by cardiac transcription factors such as GATA4 and CSX/NKX2.5 [80]. In skeletal muscle, rat ACC-2 gene transcription appears to be largely under the control of the PI promoter, while human ACC-2 expression is driven by both PI and PII promoters, suggesting that in humans, ACC-2 can also be regulated by changes in nutritional status. The PI promoter in skeletal muscle can be activated by myogenic regulatory factors (e.g., MRF4, myoD) and retinoic acid receptors (RAR and RXR) [80, 81]. The PII promoter dominates in the control of ACC-2 expression in rat and human liver, being activated by some MRFs and controlled by changes in nutrition via SREBP [82].  30  Figure 1.10: Organization of ACC-1 gene and predicted mRNA transcripts [79-81]. The 5’ end of the unspliced mRNA transcripts are shown for rat (a) and human (b) ACC1 genes (not drawn to scale). Known mRNA transcripts are listed below the corresponding genes and transcripts in bold are typically the most abundant. The grey boxes represent the potential exons with their corresponding numerical designations. a) PI  PII AUG  1 Class I:  2  3  Class II:  1:4:5 1:5  4  5  2:4:5 2:3:4:5 2:5  b) PI  PII  1  Class I:  1  1  1  1:4:5:6 1:5:6 1:4:5:5B:6 1: 1:1B: 1:1A:1B:  PIII  2  3  4  5  5A’  5A 5B  6  Class II: 2:3:5:6 2:4:5:5B:6 1C:3:4:5:6 1C:4:5:6 1C:3:4:5:5A’:5B:6 1C:3:4:5:6  Class III: 5A:6 5A:5B:6  31  The various mRNA transcripts all appear to give rise to the same protein, the main differences in the currently known mRNA transcripts being in the 5’ untranslated regions. Alternative splicing can also lead to different protein products, as discussed earlier. This indicates the possibility of regulation of ACC abundance at the translational level and although this is currently not well understood, it is possible that the 5’ untranslated region can contribute to translational control [69]. The mechanism by which ACC protein subunits are degraded has only been recently addressed. A link between ACC-1 and the pseudokinase TRB3, the human homolog of the drosophila protein tribbles, has recently been found by Qi et al [83]. In their studies, they found that ACC-1 co-immunoprecipitated with TRB3 in extracts from human embryonic kidney cells (HEK293T). Using truncated mutants, they found that ACC1 was associated with the N-terminal region of TRB3. Interestingly, binding of ACC-1 to TRB3 appeared to be linked to the inactivation of ACC. TRB3 was also found to form a complex with the E3 ubuiquitin ligase mCOP1, known as the “constitutive photomorphogenic protein 1”, through the C-terminal region of TRB3. These observations led to further studies that supported the idea that TRB3 inhibits ACC activity by binding to ACC-1, recruiting COP1 and thereby triggering the ubiquination and proteasomal degradation of ACC protein subunits. It has also been shown that TRB3 binds preferentially to ACC that has been phsophorylated at the AMPK-dependent site(s) and that this interaction inhibits subsequent dephosphorylation of ACC, thereby keeping the enzyme in the less active phosphorylated form. This action therefore acutely suppresses ACC activity as well as promoting longer-term down regulation through protein degradation. It must be mentioned however, that a contrary report has also appeared [84]. In this latter study, it was found that the loss of TRB3 through genetic knock-out did not have any appreciable effects on ACC or lipid metabolism in vivo, questioning the physiological relevance of the prior studies.  32  1.2.6.2 Allosteric regulation of ACC ACC is subject to the effects of several potentially important physiological allosteric activators and inhibitors. Allosteric activators include di- and tri-carboxylic acids such as citrate, isocitrate and glutamate [14, 85]. Citrate is generally considered to be more important as an ACC activator, from a physiological perspective and also because citrate leads to greater ACC activation in vitro [86]. The effects of di- and tri-carboxylic acids (e.g., isocitrate, citrate, and malonate) on ACC were inadvertently discovered based on the fact that these carboxylic acids activated fatty acid synthesis when acetate or acetyl-CoA were used as a substrate, but not if the substrate was supplied as malonyl-CoA. These observations actually preceded the discovery of ACC itself, but direct effects of the carboxylic acids on purified ACC have since been amply demonstrated [10]. After the discovery of an intermediate step involving the carboxylation of acetyl-CoA (namely the ACC reaction), it was demonstrated that citrate and isocitrate greatly activated this intermediate step. However, it was still unclear whether citrate was involved in the formation of a necessary intermediate, or whether it was a direct activator. This was clarified in studies by Martin and Vagelos, in which they indubated ACC with radioactive citrate and showed that the citrate was not metabolized during this incubation period [87]. Citrate was also found to be more effective in ACC activation than other di- and tri-carboxylic acids the structures of which are shown in figure 1.11. Further work by Vagelos et al eliminated the possibility that citrate activated ACC through the chelation of metals, and also demonstrated that citrate leads to a conformational change in the ACC enzyme causing an effect on the active site and on the sedimentation characteristics of ACC [11]. It was later demonstrated that the activation of ACC by citrate was associated with the polymerization of the enzyme [85], a process that will be further discussed in the next section.  33  Figure 1.11: Chemical structures of several allosteric regulators of ACC. Citrate  O  O  Isocitrate  -  O O  O O  -  -  O  O O  OH  O  -  O  -  O  -  OH Malonate  O O  Glutamate  O  O  -  O  -  O  O  -  O  -  +  NH3 Through kinetic analysis of the effects of citrate on ACC, it has been proposed by Hashimoto and Numa that an equilibrium exists between an inactive carboxylated enzyme and an active carboxylated enzyme, and that citrate drives the equilibrium to the active form. These authors also proposed that the site of citrate binding on ACC is most likely on the biotin carboxylase subdomain of ACC, but the exact site of citrate binding has not yet been identified. Despite the dramatic in vitro effects, the extent to which citrate controls ACC in vivo is not so clear because the calculated Ka for citrate is generally higher than typical cellular citrate concentrations. Furthermore, ACC activity has been shown to change under conditions in which cellular citrate concentrations are rather constant. For example, the activation of ACC by insulin occurs with no change in cellular citrate levels [88]. Also, cell citrate concentrations do not always change in parallel with ACC activity and rates of fatty acid synthesis whereas the Ka observed in vitro is often substantially higher than 1 mM. Furthermore, Halestrap and Denton, measured ACC activity from fat  34  pads stimulated with insulin and found that ACC activity was enhanced by insulin treatment even when assays were performed in the absence of citrate [89]. It should be considered that not only citrate concentration, but also that the citrate sensitivity of ACC may vary. For example, citrate sensitivity may depend on phosphorylation or the effects of inhibitory ligands or association with other proteins. Some of these points will be discussed below and through the thesis. In addition to uncertainty about the physiological effects of citrate, the two ACC isoforms may differ in this respect, particularly considering the different tissue types in which their roles predominate. For example, it appears likely that citrate more clearly links to ACC function in skeletal muscle [90]. In rats, the effects of fasting and refeeding on the citrate concentration required for half-maximal activation (Kact) was determined using ACC purified from liver (ACC1) or skeletal muscle (ACC-2). The Ka for citrate activation of liver ACC was higher following fasting than in rats that fed ad libitum and then decreased upon subsequent refeeding for 48 hours. However, there was no change in Kact for ACC extracted from skeletal muscle [91]. A study by Belke et al on Richardson’s ground squirrels compared the amounts of ACC-1 and ACC-2 protein in the hearts of hibernating and nonhibernating animals. The levels of ACC-1 remained the same in the hearts of both groups, while the levels of ACC-2 were 41% lower in the hearts of the hibernating animals [92]. Both of these studies demonstrate that these two isoforms are regulated by different mechanisms. Glutamate is another di-carboxylic acid that has been found to activate ACC. In initial studies by Baquet et al designed to investigate effects of hepatocyte cell volume on metabolism, it was found that treatment of hepatocytes with glutamate led to swelling and ACC activation [93]. They also demonstrated that glutamate was the likely intracellular mediator and that microcystin-LR, a phosphatase inhibitor, blocked the glutamateinduced stimulation of ACC that they later determined was mediated by a type 2A protein phosphatase [94]. Further studies in my laboratory led to the demonstration thtat the effects of glutamate on ACC are more complex and also involve direct allosteric activation [95]. While the effects of glutamate on ACC activity are not as dramatic as citrate  35  (approximately 60% of the activation seen with citrate), it is worth nothing that intracellular glutamate concentrations are substantially higher than citrate. Therefore, whereas citrate concentrations in the cytosol of hepatocytes are generally in the range of 0.1 to 0.5 mM, the corresponding concentration of glutamate can be more than one hundred-fold higher (10-30 mM). Similar glutamate-mediated activation was seen with ACC from liver, heart and adipose tissues. Through size exclusion chromatography, it was also shown that, like citrate, glutamate binds to ACC dimers and promotes polymerization. It was also demonstrated that glutamate is a more effective activator when ACC is dephosphorylated [95]. In counterbalance to effects of carboxylic acids, a number of physiological allosteric inhibitors of ACC have been identified. Acyl-CoA esters such as palmitoyl CoA and malonyl-CoA are known to inhibit ACC activity, presumably providing a negative feedback mechanism. ACC is sensitive to low μM concentrations of palmitoylCoA and, contrary to initial concerns, the inhibitory effects are reversible. High concentrations of 10 μM or more, palmitoyl-CoA caused denaturation and aggregation of ACC, similar to the effects of detergents such as dodecylsulfate [96]. However, Halestrap and Denton demonstrated that even in the presence of high albumin concentrations (10 mg/mL), ACC is reversibly inhibited by concentrations of palmitoylCoA in the micromolar range or lower. After incubation with palmitoyl-CoA, dilution of the enzyme did not rapidly reverse the effects of palmitoyl-CoA, indicating that palmitoyl-CoA binds to ACC and potentially dissociates very slowly from the enzyme. Interestingly, incubation of freshly-isolated extracts of fat cells with serum albumin leads to a time-dependent ACC activation, suggesting that endogenous CoA esters or other hydrophobic chemicals restrict ACC activity prior to albumin treatment. Independently, Goodridge demonstrated that palmitoyl-CoA induces a conformational change which leads to dissociation of citrate from ACC. Separate studies of the effects of palmitoylCoA and malonyl-CoA on ACC kinetics have given conflicting results and it is not certain if the effects are competitive or non-competitive with respect to acetyl-CoA [33, 70]. Considering that high levels of these CoA esters were required to competitively inhibit ACC, it is possible that palmitoyl-CoA and malonyl-CoA can inhibit ACC through complex mechanisms. For example, it is possible that the CoA esters act  36  allosterically at low concentrations, but act competitively with acetyl-CoA at high concentrations. Conflicting results have also been obtained for free coenzyme A. Yeh et al found that coenzyme A allosterically activated ACC, even after prior activation by exposure to saturating levels of citrate, concluding that the binding site for CoA is different from the citrate binding site. CoA was also found to act competitively with respect to palmitoylCoA but not with respect to acetyl-CoA [97]. In contrast, Moule et al found that coenzyme A was not an activator at all, but rather was a potent inhibitor of ACC [98]. They showed that coenzyme A induced depolymerization of ACC and that this effect could be reversed with citrate. They also found that the presence of MgATP is essential for the CoA-dependent inactivation. The difference in the results from these two groups is difficult to reconcile although it is worth noting that in the initial experiments performed during the discovery of ACC, CoA was found to be beneficial at low concentrations, but was found to be inhibitory at high concentrations [4].  1.2.6.3 ACC polymerization Under normal physiological conditions in animal cells, the inactivated form of ACC exists as a dimer, traditionally called a protomer [63]. When both ACC isoforms are present, it is likely that homodimers and heterodimers can exist. Upon activation in vitro, ACC polymerizes into a filamentous structure of about 10-20 dimers [99] with a molecular weight of about 5-10 x 106 [86]. The polymerization is readily reversible upon removal of allosteric activators and/or reducing the incubation temperature and depolymerization of the enzyme leads to inactivation. Electron microscopy has shown that the ACC filaments are twisted helical structures with a length up to 5000 Å, and a width of 70-100 Å [85, 100]. The sedimentation coefficient of dimers is about 20 S, while the polymerized form has sedimentation coefficients measured to be in the range between 30-42 S [86, 101]. Studies on citrate activation and polymerization of ACC have revealed what appears to be a two step process. Sedimentation velocity centrifugation has revealed that citrate activation of ACC can occur rapidly even while the enzyme is still in the dimeric form. Based on these studies, it was concluded that citrate is first needed to bind to the  37  ACC dimers leading to a conformation change that facilitates rapid activation that precedes enzyme polymerization [99]. Binding experiments involving the use of [14C] citrate and either rapid filtration or ultracentrifugation with subsequent Scatchard analysis demonstrated that one citrate molecule is bound per dimer in a polymeric filament [18]. Steady state kinetic analysis by Beaty et al lends further support to these observations. In their work, they compared the time taken to reach steady-state biotin carboxylation in assay buffer for enzyme that was pre-incubated with citrate versus enzyme that had not been pre-incubated. A difference was observed between these two conditions with the pre-incubated ACC reaching steady-state carboxylation faster and with a shorter lag-time than non-incubated enzyme. These observations led to the conclusion that ACC activation and polymerization are two separate processes and that activation is the ratelimiting step in the reaction and occurs faster than polymerization [102]. The rate of citrate-induced polymerization of chicken liver ACC is particularly rapid [99]. In contrast, rat liver ACC requires a longer incubation with citrate, usually about ten to twenty minutes [86]. In fact, the polymerization and activation of mammalian ACC from rat liver, fat and muscle is consistently much slower, with a half time of ten to fifteen minutes. This suggests either intrinsic differences between avian and mammalian ACCs or the fact that association with proteins/tissue factors may influence polymerization. Considering all the in vitro evidence for ACC polymerization, it has been of considerable interest to determine if ACC polymerization also occurs in intact cells. A series of experiments by Meredith et al involving digitonin treatment of chick liver cells appears to confirm the polymerization of ACC with intact cells. The use of digitonin leads to permeabilization of the cell membrane, with minimal effects on intracellular organelles, allowing the release of cytosolic components, while larger cytoplasmic components remain in the cell. In these experiments, the dimeric form of ACC was released, while the polymerized form remained in the cell. When cell citrate concentrations were rapidly decreased in the permeabilized cells, the rate of ACC release also increased, indicating that ACC polymers can be destabilized through a decrease in citrate concentration [103].  38  In the same series of experiments with digitionin-treated chick liver cells, the addition of avidin prevented subsequent ACC polymerization, even in the presence of citrate. However, if the enzyme was first incubated with citrate, allowing polymerization to occur, the effects of avidin were diminished, leading to reduced rates of ACC release [103, 104]. In related studies, allosteric inhibitors of ACC, such as palmitoyl-CoA, were found to prevent polymerization and enhanced the rates of digitonin-induced release [96]. Size exclusion chromatography has also been employed to separate the polymeric, dimeric and any intermediate-sized forms of ACC. These results have confirmed the effects of citrate on the polymerization of ACC. This method also showed that preincubation of ACC with PKA led to decreased, polymerization of ACC in vitro [105]. Although ACC-2 appears to show clear citrate-induced activation, the extent to which this isoform also polymerizes has not been reported. Preliminary experiments in my laboratory have demonstrated that ACC-2 polymerizes to a lesser extent than ACC-1, and that in the presence of both isoforms, ACC-1 may induce the polymerization of ACC-2 [unpublished work]. However, recent work with the recombinant form of ACC using dynamic light scattering showed that the two recombinant isoforms of ACC were differentially activated by citrate, where ACC-2 was activated and aggregated more than ACC-1 by citrate and Mg [106].  1.2.6.4 ACC regulation by phosphorylation Changes in ACC expression and degradation are highly regulated and can have an important bearing on ACC activity. Perhaps the situation in which changes in ACC expression are most definitive occurs in models of insulin-deficiency such as STZinduced or alloxan-induced diabetes. In these settings, ACC gene and protein expression and corresponding activity declines to less than 10 percent of normal in adipose tissue and liver. In this, as in other settings, changes in ACC protein expression typically take many hours to take effect. In many other situations, however, ACC expression change does not fully account for changes in ACC activity. For example, in rats starved for 48 hours, the total activity of ACC is reduced to about 40% of control activity in fat and liver but rates of fatty acid synthesis are much more profoundly reduced. After refeeding, especially with a carbohydrate-rich, fat-free diet, total ACC activity rebounds to  39  normal levels or even supra-normal within 24 hours and indeed clear activation is observed within 4-6 hours, well before changes in total ACC protein expression [107]. These experiments show that other mechanisms must be available to regulate ACC activity and especially to allow rapid responses to acute changes in nutrition and the associated changes in hormone concentrations. The phosphorylation/de-phosphorylation of several critical serine residues of ACC provides what appears to be the major mechanism to account for rapid ACC control, in combination with effects of allosteric ligands. The phosphorylation of ACC was discovered by Lowenstein when he demonstrated that purified rat liver ACC contained 2 moles of phosphate per mol of ACC subunit [108]. Following this, Carlson and Kim demonstrated that they could label rat liver ACC with [32P]-ATP [109]. This labeled ACC was also rendered inactive, and the label remained with ACC through ammonium sulfate precipitation, gel filtration, and ion exchange chromatography. The use of [14C]-ATP did not lead to incorporation, indicating that the labeled phosphate group of ATP was used to phosphorylate ACC, rather than the incorporation by binding of ATP as a substrate or regulator. The removal of the phosphate group, and thus reactivation of the enzyme, was achieved via incubation of the labeled ACC with a magnesium-dependent phosphatase. The in vivo phosphorylation studies were first initiated by Brownsey and Denton, when they demonstrated for the first time that ACC could be labeled with 32P in intact fat cells [110]. Following the discovery of ACC phosphorylation, a relationship between hormone regulation of activity and the phosphorylation/dephosphorylation was hypothesized. Several hormones were found to have effects on ACC activity. For example, insulin treatment leads to rapid activation of ACC in vivo in fat and liver [111], in isolated fat tissue and cells [88, 89], and in isolated hepatocytes [101]. Effects of insulin on ACC polymerization have not been uniformly observed. In early studies by Halestrap et al, insulin treatment of adipose tissue led to greater ACC sedimentation by centrifugation [88, 89]. Borthwick et al also observed effects of insulin on the properties of ACC recovered in polymeric forms by FPLC size-exclusion chromatography [105]. In these studies, the effects of insulin were not as dramatic as the subsequent activation and polymerization induced by in vitro addition of citrate. However, others have reported  40  conflicting results. For example, in a study by Buechler et al, insulin treatment of hepatocytes did not lead to a change in the sedimentation behavior of ACC [101]. Hormones that lead to increases in cellular cAMP levels such as glucagon and catecholamines have a rapid inactivating effect on ACC and counteract the effects of insulin. For example, ACC that has been isolated from rats injected with glucagon is less citrate-sensitive and is more phosphorylated than the enzyme from rats injected with insulin [101, 112]. Epinephrine also has an inhibitory effect on ACC and leads to increased ACC phosphorylation both in isolated fat tissue [113] and in liver following in vivo injection [112]. Epinephrine treatment also leads to disassociation of ACC polymers into an intermediate size species that is citrate-insensitive [114]. Other hormones may also influence ACC phosphorylation and activity including thyroid hormone [115] and leptin [116], but the physiological relevance of these has not been as firmly established. The relationship between the hormonal effects on ACC activity, polymerization and phosphorylation were first tested in rat epididymal fat tissue treated with insulin and epinephrine. In these studies, metabolic labeling with [32P]-phosphate led to the demonstration that ACC-1 is phosphorylated on at least four hormone-responsive sites in fat cells. Epinephrine was found to increase the phosphorylation of ACC in parallel with inactivation while insulin treatment led to ACC activation and an apparently paradoxical specific site phosphorylation [117, 118]. Studies by Lee and Kim revealed that the administration of epinephrine in 32Pinjected rats led to an increased incorporation of 32P into liver ACC and this was associated with ACC inactivation [119] and in the same study, it was found that ACC has multiple-phosphorylation sites, some of which were hormone-sensitive. In related studies of isolated hepatocytes, Witters et al found that glucagon led to the inactivation of ACC, while increasing levels of cAMP and ACC phosphorylation [120]. Based on these studies, it was assumed that cAMP dependent protein kinase probably mediated the phosphorylation of ACC. Several years later, the major phosphorylation sites on rat ACC were identified. The first sites were identified at serine residues 79, 1200, and 1215 [121]. Despite the discovery of the role of cAMP, it was also found that the phosphorylation of ACC-1 at these identified serine residues was mediated mainly by AMP-activated protein kinase (AMPK), rather than cyclic AMP-dependent protein kinase  41  (PKA) [122]. Considerable subsequent work has shown that AMPK plays a crucial role in mediating many stress responses and the inhibition of ACC is just one of many outcomes of AMPK activation [123]. To this date, several more ACC phosphorylation sites have been identified, either through specific experimental study or through genome-wide protein phosphorylation analysis [124] (table 1.4). However, the two kinases that still appear to be the major players in ACC phosphorylation are AMPK and PKA. The phosphorylation sites on ACC-1 have been determined more clearly than those of ACC-2 for historical reasons and because of the abundance of ACC-1. While both isoforms have sites for both PKA and AMPK, AMPK appears to be the major kinase for ACC-1 while evidence suggests PKA may have a greater role in phosphorylation of ACC-2 [125, 126]. The effects of PKA on ACC-2 have been demonstrated on purified rat liver ACC [66] and in intact cardiac myocytes that were treated with isoproterenol [126] or in endothelial cells treated with leptin [127]. Interestingly, several phosphorylation sites on ACC have been identified to be “silent” as they have no known effect on ACC activity. For example, the phosphorylation of ACC by CKII is stimulated by insulin and appears to have no apparent effect on ACC activity [128]. Similarly, phosphorylation of ACC by CaMK-II also has no apparent effect on ACC activity [129]. Recently, the interaction of BRCA1 with ACC has been identified, and phosphorylation site Ser-1262 appears to be important in this interaction [130]. However, this phosphorylation site has only been found in studies of breast cancer cells, and not yet in fat or liver cells. Several other kinases have been identified to phosphorylate ACC, albeit minimally, such as PKB and PKC, but the physiological role of these kinases in ACC phosphorylation has not been established [125].  42  Table 1.4: Identified and predicted phosphorylation sites for human and rat ACC isoforms [51]. Human ACC-1  Rat ACC-2  ACC-1  Kinase ACC-2  Ser-5  Ser-5  Ser-23  Ser-231  unknown  Ser-25  Ser-25  CKII  Ser-29  Ser-29  CaMKII  Ser-48  Ser-47  Ser-50  Ser-49  Ser-53  Ser-52  Ser-56  Ser-55  Thr-58  Ser-57  Ser-60  Ser-59  Ser-78  Ser-220  Ser-77  Ser-219  PKA, PKC  Ser-80  Ser-222  Ser-79  Ser-221  AMPK  Ser-95  PKC  Ser-488 Thr-1042  Ser-1041  Ser-1201  Ser-1200  PKA, AMPK  Ser-1215  AMPK  Ser-1263  Ser-1262  Ser-1844  Ser-1843  Ser-2099 Tyr-2108 1  Ser-2107  Shaded residues indicate phosphorylation sites that have been directly confirmed  through experiments using purified proteins or in intact cells.  43  While the effects of stress hormones on ACC have been relatively well defined, the effects of insulin are still unresolved. Studies by Buechler et al and Borthwick et al indicate that the action of insulin on ACC is more indirect and transient than effects of catecholamines or glucagon. Studies of ACC in Fao Reuber hepatoma cells and perfused hearts treated with insulin have shown a decrease in ACC phosphorylation coincident with activation and that ACC dephosphorylation occurred at certain sites. From these studies, it was concluded that insulin activates ACC by either inhibiting the activity of a protein kinase, most likely AMPK, or stimulating the activity of a protein phosphatase [131]. However, these effects were not observed in primary fat or liver cells, and the effects of insulin in the hepatoma cells were independent of changes in cyclic AMP or 5’AMP concentrations. Interestingly, as noted above, it has been demonstrated that the presence of insulin leads to increased ACC phosphorylation at a specific site, as demonstrated by Brownsey and Denton [118]. Based on these results it was suggested that insulin leads to an activation of serine-kinase which leads to the activation of ACC [132].  44  1.2.6.5 Integrating effects of allostery and phosphorylation on ACC polymerization While the effects of allostery and phosphorylation on ACC have so far been discussed separately, the two regulatory mechanisms are intricately linked to one another. In fat, liver and muscle cells, cytosolic citrate concentrations typically range from 0.1 to 1 mM [133]. Cell citrate concentrations are therefore often sub-maximal for ACC activation and generally do not change in parallel with ACC and FASN activity, or rates of fatty acid synthesis. Most likely then, there are other mechanisms in place that influence ACC activity, both by influencing citrate sensitivity and by other means. For example, Halestrap et al demonstrated that insulin treatment led to enhanced citrate sensitivity of ACC [89]. Conversely, phosphorylated ACC is less citrate-sensitive either following incubation of purified ACC with PKA or AMPK or following isolation of ACC from glucagon or adrenaline-treated cells [113]. In a series of in vivo experiments by Mabrouk et al, ACC isolated from the livers of insulin-treated rats was found to be both dephosphorylated and also more sensitive to citrate [112]. From these experiments, it can be seen that the citrate sensitivity of ACC is substantially influenced by the degree of phosphorylation of ACC. Considering that ACC has several critical phosphorylation sites, it is likely that these may influence properties in addition to citrate sensitivity. For example, phosphorylation by AMPK has a significant effect on ACC Vmax as well as Ka for citrate, although effects on substrate Km values have not been observed. Another potential role for phosphorylation is to facilitate protein-protein interactions. This phenomenon has broad importanct in many cell signaling mechanisms and is also specifically demonstrated for ACC in that phosphoylation of Ser-1262 enhances binding to BRCA1 while AMPK phsophorylation facilitated binding of TRB3. In both cases, ACC-protein interactions appear to reduce the rate of ACC dephosphorylation and may also promote subsequent degradation.  45  1.3 Biological relevance Malonyl-CoA has a number of important metabolic functions in cells (figure 1.12). Firstly, it is used as a substrate by fatty acid synthase (FASN), the major product of which is palmitate (C16:0) [134]. The formation of palmitate is initiated with acetylCoA and requires seven cycles of a series of condensation, reduction and dehydration reactions that use seven molecules of malonyl-CoA. As mentioned earlier, the reaction catalyzed by ACC is generally thought to be the rate-limiting step for fatty acid biosynthesis [5, 13]. This idea is further compounded by results of Ha et al, who found that when ACC mRNA levels were decreased by an ACC specific ribozyme, FASN levels were also decreased accordingly [135]. It is important to qualify the term “ratelimiting enzyme” and to account for the notion of distributed metabolic control as outlined in metabolic control theory [136]. According to this concept, we should rather consider that ACC contributes (perhaps strongly) to flux control. The degree to which control of ACC is reflected in the overall pathway may vary according to physiological conditions and also how the pathway itself is defined. For example, defining “fatty acid synthesis” as the pathway from glucose to palmitate will require a different analysis than if the pathway is defined from acetate to palmitate. In this case, if the pathway for FA synthesis began with glucose uptake, key control steps such as glucose transport, phosphofructokinase (PFK), pyruvate kinase, and PDH should be considered and ACC would have a lower level of control over FA synthesis. However, if FA synthesis began at the level of acetate molecules, ACC would have a greater level of control over FA synthesis. In addition to its role in de novo lipid synthesis, malonyl-CoA is also an important regulator of carnitine palmitoyl transferase I (CPT-I), especially in tissues where there are low levels of fatty acid synthase [137]. The carnitine palmitoyl transferase system is required for the movement of palmitate into the mitochondria for subsequent fatty acid oxidation. CPT-I is located on the outer mitochondrial membrane and catalyzes the formation of fatty acyl-carnitine from free carnitine and fatty acyl-CoA. Carnitine/acylcarnitine translocase is an integral inner membrane protein while CPT-II is located on the inner mitochondrial membrane and catalyzes the formation of palmitoylCoA in the mitochondrial matrix. Malonyl-CoA inhibits CPT-I allosterically, thus  46  inhibiting the movement of fatty acids into the mitochondria and subsequent fatty acid oxidation [138, 139]. Kinetically, malonyl-CoA increases the Km of CPT-I for fatty acylCoA and L-carnitine. There are two binding sites for malonyl-CoA on CPT-I. The first is a low-affinity binding site at the active site where malonyl-CoA competes with the fatty acyl-CoA esters. The second, a high-affinity binding site for malonyl-CoA, is a true allosteric site located away from the active site. Currently, a detailed structural model to explain the mechanism by which malonyl-CoA inhibits CPT-I is not available. In recent years, studies of pancreatic beta cells have investigated the possibility that inhibition of CPT-I by malonyl-CoA might be involved in the control of insulin secretion [140]. According to this concept, malonyl-CoA could inhibit CPT-I and by reducing the rate of β-oxidation, thereby enhances fatty acid/CoA levels in the cytosol. It is then pictured that the elevated cytosolic levels of fatty acids and/or CoA esters may in some way influence insulin secretion. Finally, malonyl-CoA is used in microsomal fatty acid chain elongation systems [112], specifically by elongases that consists of four enzymes: β-keotoacyl CoA synthase, β-ketoacyl CoA reductase, β-hydroxyacyl CoA dehydrase, and trans-2-enoyl CoA reductase. Malonyl-CoA is used as a two carbon unit donor in the first reaction of elongation, catalyzed by β-keotoacyl CoA synthase. The carboxyl group of malonylCoA is used as a leaving group with the two acetyl carbons being used to extend the fatty acid chain. The next three enzymatic reactions then convert the initial beta-ketoacyl intermediate to a C:18 or longer fatty acid [141].  47  Figure 1.12: The metabolic roles of malonyl-CoA in mammalian cells. (1) Substrate for FASN, (2) Inhibition of CPT-I, and (3) Substrate for chain elongation.  48  Over the last two decades, ACC has been identified as a potential target for the actions of chemicals used as herbicides, anti-obesity drugs, and anti-cancer agents. As a result, there have been intensive efforts by many groups to search for ACC inhibitors, with these various applications in mind. In principle, ACC might be targeted through the inhibition of ACC expression, enhanced degradation, or through direct or indirect effects on allosteric or kinetic properties. Interestingly, one class of widely-used herbicides has been found to act by inhibiting one of the plant forms of ACC. In this context, the multi-subunit plastid ACC is inhibited by two classes of herbicides: aryloxyphenoxypropionates and cyclohexanediones [142]. Recently, it has been found that a single point mutation in the CT subunit of ACC can lead to increased resistance to the herbicide [143]. Of particular continuing interest is the fact that a growing tolerance to these herbicides is emerging, allowing ACC to regain substantial activity even in the presence of herbicides [143]. Considering that ACC plays an important role in fatty acid metabolism, many groups have targeted ACC to control obesity and associated diseases such as cardiovascular disease and type II diabetes. In one major pharmaceutical effort, Harwood et al found a group of N-substitued bipiperidylcarboxamide compounds that inhibited ACC [144]. The most promising early lead compound, CP610431 inhibited ACC activity with no apparent differential effect on the two isoforms. The same group went on to develop a related compound, CP640186 that when injected in rats, led to increased fatty acid oxidation, decreased malonyl-CoA levels, reduced body weight, decreased fat storage and improved insulin sensitivity. The concept of developing isoform-specific inhibitors is also worth considering, especially in light of the work of Wakil et al whose studies show that specific deletion of ACC-2 in mice leads to increased basal rates of fatty acid oxidation and decreased body mass despite 20-30% increased food intake. The ACC-2 null mice have similar life spans to their wildtype counterparts as well. In a subsequent study [74], the same authors showed that the ACC-2 null mice were protected from insulin resistance induced by feeding a high fat/high carbohydrate diet. It seems probable that ACC-2 may be a more important target than ACC-1 considering that its primary role is evidently to produce malonyl-CoA to inhibit fatty acid oxidation.  49  The effects of isoform-specific inhibition were also reported in 2006 by Savage et al who used a rat model for insulin resistance to test the effectiveness of antisense oligonucleotides against ACC-1 and ACC-2. In their studies, they found that knocking down expression of ACC-1 led to decreased lipogenesis, while knockdown of ACC-2 had no effect on lipogenesis. Interestingly, when the expression of both isoforms were simultaneously reduced, the result was increased whole body fatty acid oxidation, lowered hepatic lipid synthesis and export, and improved insulin sensitivity [145]. The concept that ACC might be useful as a target for cancer chemotherapy is supported by several lines of evidence. For example, BRCA1, a key protein linked to susceptibility to breast and ovarian cancer in humans has been found to directly associate with ACC [146]. Further studies on BRCA1 binding to ACC have demonstrated that the phosphorylation site Ser-1262, previously considered to be a silent phosphorylation site, is important in mediating the association between these two proteins [130]. In a more general sense, ACC may be particularly important in cancer cells in supplying the key substrate to enhance rates of fatty acid synthesis for membrane biogenesis in rapidly dividing cells. Accordingly, inhibition of ACC or FASN appears to slow the growth of several cancer cell lines, including LNCAP and PC-3 cells [147]. While the end application can vary, many groups have searched for ACC inhibitors. Ohmori et al found a synthetic benzoic acid derivative that inhibited HMGCoA reductase and acetyl-CoA carboxylase [148]. The use of polyketide natural products have also been studied, and one compound, soraphen A, has been identified to inhibit the biotin carboxylase component in mammalian ACC [54]. Inhibitors from the fungus Gongronella butreli have also been found to inhibit rat liver ACC [149]. Chloride ions have also been discovered to inhibit ACC activity [150].  50  1.4 Thesis investigations Many properties of ACC remain to be fully described or understood. For example, details of the allosteric binding sites of mammalian ACCs are not at all defined. Attempts to clone and express ACC isoforms have until quite recently been unsuccessful and the specific activity and citrate-sensitivity of these expressed forms differ from that of ACC extracted from animal tissues. Despite the fact that the effects of citrate on ACC have been known since the discovery of the enzyme, little is known about the citrate-binding site. At the initiation of the work outlined in this thesis, there was only a limited understanding of the effects of PLP on ACC. Preliminary studies indicated that PLP was a potent inhibitor of ACC and some aspects of the specificity of PLP binding to ACC were determined [151]. My aim was to fully define the kinetic effects of PLP and to seek evidence to test whether this involved direct binding and, if so, whether this occurred at a catalytic site or an allosteric site. It was further anticipated that studies of the effects of PLP might shed light on the actions of citrate. The second general feature of ACC control that I investigated was the extent to which and the possible significance of interactions of ACC with other cellular proteins. The possibility that protein-protein interactions may significantly affect ACC function arose from several considerations. In fact, interactions of ACC, for example with FASN, have been the subject of speculation for many years and represent a possible example of a functional “metabolon”. Previous studies in this laboratory by Katherine Quayle had provided direct evidence for interaction of rat liver ACC with one or more endogenous liver proteins that influenced ACC activity and sensitivity to citrate. Building on these earlier studies and recognizing the presence of emerging databases of protein-protein interactions, I set out to design approaches to search for ACC-interacting proteins with a major focus on proteins that might associate with ACC polymers.  51  2  Chapter 2: Experimental Procedures  2.1 Materials Male Wistar rats were supplied by the University of British Columbia Animal Care Facility and the rats were maintained under the care of the technicians in the Department of Cellular and Physiological Sciences for approximately one to three days. The rats were fed laboratory chow and kept on a 12-hour light-dark cycle with light from 8 am to 8 pm. All animal procedures were done according to the guidelines of the Canadian Council for Animal Care and were approved by the UBC Animal Care Committee. Primary liver hepatocytes were obtained from Dr. Tom Chang, Faculty of Pharmaceutical Sciences at the University of British Columbia. Laboratory solvents and most standard laboratory chemicals were from Fisher Scientific, while specific reagents were obtained from a variety of suppliers as specified below. Pepstatin A, glutathione, PMSF, avidin, GTP, ATP, PLP, NaBH4, biotin, EDTA, EGTA, benzamidine, MOPS, BSA, sodium citrate, HEPES, MgSO4, sodium bicarbonate, potassium bicarbonate, Tween-20, Tris, glycine, DTT, NADPH, iodoacetamide, ammonium bicarbonate, fibronectin, insulin, colchicine, calcium chloride, and magnesium chloride were all obtained from Sigma. BSA (fatty acid free) was from ICN Biomedicals Inc. Sequencing grade trypsin (porcine) was from Promega. BioGelA resin, urea, piperazine diacrylamide, and affigel-10 were from BioRad. Coenzyme A was from Roche. Leupeptin was from Peptides International (Louisville, KY). [3H]-NaBH4 was from Perkin Elmer Life Sciences. Matri-gel was from VWR. Centrifugal filters (10 kDa molecular weight cut-off) and PVDF membrane were from Millipore. Streptavidin horseradish peroxidase conjugate, [14C]-NaHCO3, SDS-PAGE high molecular weight rainbow markers, and ECL DualVue Western blot markers were from Amersham. Immunopure immobilized avidin, stripping buffer, SuperSignal West Pico peroxide/luminal, and immunopure antibody goat anti rabbit HRP were all from Pierce. Anti-FAS (H-300) and Anti-α-tubulin (B7) were from Santa Cruz. Anti-phospho ACC (ser-79) and anti-ACC1 were from Upstate Biotechnology. Tubulin (primary antimouse) for immunofluorescence was from Abcam. Fluorescently labeled secondary antibodies, including Alexa 568 (anti-rabbit) and Alexa 488 (anti-mouse) were from 52  Molecular Probes (Invitrogen). Purified and pressurized N2 gas and a mixture of O2:CO2 (95:5 v:v), as well as liquid nitrogen were from Praxair.  2.2 Methods 2.2.1 Tissue isolation and preparation Male Wistar rats weighing 120-140 g were killed by CO2 asphyxiation, usually between 8 am and 9 am. The liver was removed and immediately placed onto ice. Epididymal and perirenal fat pads (white adipose tissue) were removed and rinsed with pre-gassed (O2:CO2; 95:5 v/v) and pre-warmed (37°C ) Krebs-Henseleit buffer (25 mM NaHCO3, 1.2 mM KH2PO4, 120 mM NaCl, 5 mM KCl, 1.2 mM MgSO4, 1.3 mM CaCl2 and 11 mM glucose). Rinsed fat pads were incubated in the same buffer for 30 minutes at 37°C to allow the effect of any residual endogenous hormones to subside. Following incubation, the fat pads were blotted on Whatman 3MM paper, frozen in liquid nitrogen and stored on dry ice prior to be homogenization. To accumulate tissues for the preparation of tubulin, brains were removed, frozen in liquid nitrogen and stored at 80°C.  2.2.2 Tissue homogenization A pre-chilled Teflon-glass Potter-Elvehjem homogenizer was used to homogenize liver in 4-6 volumes of homogenization buffer (20mM MOPS, 0.25 M sucrose, 2 mM EDTA, 2 mM EGTA, 2.5 mM benzamidine, 3 μM pepstatin A, 5 μM leupeptin, 2.5 μM glutathione and 25 mM PMSF; pH 7.2). Stock solutions of the homogenization buffer were kept at 4°C, pepstatin A, leupeptin, and glutathione being added the morning of the tissue homogenization. Finally, PMSF was added immediately prior to use due to its short half-life. Following freezing in liquid nitrogen, the white adipose tissue and brain tissue were powdered while frozen, and then disrupted with a Polytron homogenizer for 2-3 seconds with 4-6 volumes of ice-cold homogenization buffer.  53  2.2.3 Purification of ACC Unless stated, all procedures were carried out to keep samples at 4°C. Homogenates were centrifuged for 90 seconds at 1,000 rpm (1,000 x g) using a VWR silencer centrifuge to remove cell debris (pellet). The 1000 x g supernatant was then filtered through glass wool and centrifuged at 12,000 rpm (11,000 x g) using a Sorvall RC5 centrifuge and SS34 rotor for 20 minutes to remove mitochondria, lysosomes and other dense membrane fractions (pellet). The 11,000 x g supernatant was again filtered through glass wool and centrifuged at 55,000 rpm (215,000 x g) for 90 minutes to remove remaining “light” microsomal and other membranes and thereby generating the final “high speed supernatant”. The high speed supernatant was next treated with powdered ammonium sulfate, which was slowly added over a period of twenty minutes to give a final saturation of 40%. The mixture was then stirred on ice for another one to two hours, before centrifugation at 12,000 rpm (11,000 x g) for 30 minutes. Following centrifugation, the ammonium sulfate supernatant was discarded and the protein pellet was resuspended using buffer that varied according to the intended use of the protein extract. Samples used for ACC activity assays were resuspended in a minimal volume of homogenization buffer as described above. Samples for further purification on the BioGelA size exclusion chromatography were resuspended in a minimal volume of homogenization buffer supplemented with 100mM NaCl. Samples for further purification by avidin affinity chromatography were resuspended in homogenization buffer supplemented with 500 mM KCl. When necessary, samples were stored at either -20°C (2-4 months) or 80°C (1 year).  2.2.4 Purification of rat liver mitochondria Rat liver tissue was disrupted gently with a hand-driven pre-chilled Teflon-glass Potter-Elvehjem homogenizer in 5-6 volumes of homogenizing buffer as used for ACC purification (see above). Homogenates were centrifuged for 90 seconds at 1,000 rpm (1000 x g) using a VWR silencer centrifuge to remove cell debris (pellet). The 1,000 x g supernatant was then filtered through glass wool and centrifuged at 12,000 rpm (11,000 x g) using a Sorvall RC5 centrifuge and SS34 rotor for 20 minutes to remove mitochondria,  54  lysosomes and other dense membrane fractions (pellet). The pellet was kept and the supernatant either discarded or used for further purification of ACC. The pellet was washed twice gently with homogenizing buffer, centrifuging using a Sorvall RC5 centrifuge and SS34 rotor for 10 minutes to bring down the mitochondria. Following isolation, the mitochondrial fraction was used immediately.  2.2.5 Purification of tubulin from rat brain The purification of tubulin was carried out essentially as described by Vallee, the technique being based upon repetitive centrifugation to recover tubulin alternately in the small soluble form or large polymeric form sedimented in the presence of GTP [152]. Brain homogenate was centrifuged in a Sorvall RC5 using a SS34 rotor at 12000 rpm (11000 x g) at 4°C and the pellet was discarded. GTP and ATP were added to the supernatant at final concentrations of 0.1 mM and 2.5 mM, respectively and the mixture incubated at 37°C for 30 minutes, with occasional swirling to induce tubulin polymerization. Following incubation, the mixture was poured into centrifuge tubes and a sucrose solution (10% w/v) introduced beneath the sample using a syringe. Following centrifugation at room temperature at 12,000 rpm for 45 minutes, the supernatant was discarded and the pellets (containing polymerized tubulin) were resuspended in homogenizing buffer supplemented with 1 mM GTP (approximately 125 μL for every g initial weight of tissue). The pellets were then homogenized on ice in a pre-chilled Teflon-glass Potter-Elvejhem, left on ice for 30 minutes to allow tubulin depolymerization, and then centrifuged at 4°C in the Sorvall at 12,000 rpm (11000 x g) for 30 minutes. The supernatant, containing depolymerized tubulin was then incubated at 37°C for 30 minutes. Following incubation, the supernatant was centrifuged at room temperature for 45 minutes. The supernatant was then discarded and the pellets were aspirated. The pellets were then frozen in liquid nitrogen and stored at -20°C. The pellets were stable for up to one year with no apparent loss of function or subunit degradation. Western blots and gels stained with Coomassie blue were used to confirm the presence and purity of tubulin protein in the pellets.  55  2.2.6 Avidin affinity chromatography 2.2.6.1 Preparation of tetrameric and monomeric avidin beads Avidin affinity chromatography was used to further purify ACC, both tetrameric and monomeric avidin beads being prepared by methods similar to those described by Kohanski and Lane [153]. Affi-gel 10 was first washed in 10 mM sodium acetate buffer (pH 4.5) and then incubated overnight with gentle shaking, at 4°C with an avidin solution (10 mg/mL in 100 mM MOPS, pH 7.5). Following incubation, the beads were allowed to settle by gravity or brief gentle centrifugation (approximately 2-3 minutes at 500 x g) and the absorbance at 280 nm of the supernatant was measured to confirm binding of avidin to the beads. The beads were then washed with several volumes of 100 mM MOPS (pH 7.2). Following the wash, the beads were incubated with 100 mM ethanolamine (pH 8.0) for two hours to block any residual reactive sites on the affi-gel and then further washed with 5 volumes of 500 mM NaCl. The beads were finally washed with PBS until the A280 was very low (less than 0.002) and stored in PBS buffer containing sodium azide (0.02% w/v). To prepare monomeric avidin-agarose beads the tetrameric avidin-agarose beads were first prepared as just described, and then the avidin was monomerized with chaotropic agents as described by Kohanski and Lane [153]. Briefly, monomerization was achieved by washing the tetrameric avidin beads first with 6 M guanidine hydrochloride and then with 3 M guanidine isothiocyanate 3 before finally washing with PBS and storing in PBS containing 0.02% sodium azide. Following monomerization, the beads were stored in PBS buffer containing sodium azide (0.02%). In some experiments, a commercial preparation of immobilized monomeric avidin was used (Immunopure immobilized avidin, Pierce).  56  2.2.6.2 Immobilization of ACC onto tetrameric avidin-agarose beads Following ammonium sulfate precipitation, redissolved protein samples were allowed to thaw, and incubated overnight at 4°C, while shaking, with the tetrameric avidin-agarose beads (10 μL packed volume of beads/g wet weight of tissue). Depending on the experiment, the ACC-containing sample was first diluted with homogenization buffer containing KCl (typically 250 to 500 mM). After overnight incubation, the beads were washed 5 times with the same buffer before using to recover ACC-binding proteins. All experiments using tetrameric avidin-agarose beads were done batch-wise in small microcentrifuge tubes (0.5 mL). It is worth nothing that the very high affinity for binding of biotin to tetrameric avidin (Kd = 10-15 M) means that the binding of ACC to tetrameric avidin is practically irreversible and therefore ideal for recovery of ACC-binding proteins. In fact, in order to recover ACC following binding to tetrameric avidin-agarose beads, the beads must be heated at 95°C for 10 minutes in SDS buffer containing β-mercaptoethanol and even then, full solubilization of ACC may not be achieved. The binding capacity of the avidin beads used was typically designed to be at least a two-fold excess over the maximum possible load of ACC applied. The respective amount of immobilized avidin was estimated based on several assumptions, beginning with the presumed 1:1 molar binding. Considering the relative molecular weights of tetrameric avidin (64 kDa) and ACC dimers (at least 530 kDa) 1 mg of tetrameric avidin should in principle bind at least 8 mg of ACC. Assuming the widely confirmed ACC specific activity of 2U/mg and an abundance in rat liver of 200 mU of ACC/g wet weight, the maximum yield of ACC would be about 100 μg per g wet weight of starting tissue. Therefore, 25 μg of tetrameric avidin would be more than adequate to bind the ACC recovered per g wet weight of liver. In the preparation of tetrameric avidin agarose beads, 5 mg of avidin was used per 1 mL of beads. Assuming 100% binding, 10 μL of beads should contain about 50 μg of avidin, which is twice the amount of required avidin to bind ACC in each g liver.  57  2.2.6.3 Purification of ACC using monomeric avidin-agarose beads Following ammonium sulfate precipitation (40% saturation) protein pellets were resuspended in homogenization buffer containing 500 mM KCl and centrifuged for three minutes at 11,000 x g to remove undissolved protein prior to application to a column (1.5 cm diameter x 15 cm height) containing monomeric avidin beads prepared as described above. All sample and column buffers used in this procedure were filtered prior to use (Millipore Filter Holder, GH Polypro Hydrophilic polypropylene membrane filters 47 mm, 0.2 μm). The monomeric avidin-agarose column was first washed with three to five column volumes of PBS and the protein sample was then loaded onto the column. Both ends of the column were sealed with parafilm and the protein sample was allowed to bind to the beads for at least 1 hour on ice while gently mixing. After incubation, the column was restored to a vertical position and the monomeric avidin-agarose beads were allowed to settle. The column was then slowly washed (0.5 – 1 mL/min) with avidin buffer (50 mM MOPS, 250 mM sucrose, 2 mM EDTA, 2 mM EGTA, 5% w/v glycerol, 3 μM pepstatin A, 5 μM leupeptin, 2.5 μM GSH; pH 7.5) until the UV absorbance at A280 was very low (less than 0.005) indicating essentially complete removal of unbound proteins. To elute ACC and any other biotinylated proteins, the avidin-agarose was incubated for 1 Hr on ice with occasional mixing with an equal volume of avidin buffer containing 2 mM biotin. The column was then washed (1 mL/min) with the PBS-biotin buffer and 1 mL fractions collected. The unbound flowthrough fractions and the biotin-eluted fraction were then concentrated using Millipore Microcon centrifugal concentrators. After each use, the monomeric avidin-agarose column material was regenerated by washing with 3-4 column volumes of 0.1 M glycine (pH 2.8) over a one hour period, followed by further washing with PBS containing sodium azide (0.02%, w/v).  58  2.2.7 Labeling of ACC with PLP using [3H]-borohydride ACC was first immobilized by binding to tetrameric avidin-agarose beads, the ACC-avidin-beads were washed five times with tissue homogenizing buffer and then incubated with PLP (0.5 mM to 1.5 mM, as indicated) for 30 minutes at room temperature in the dark. Following this incubation, sodium [3H]-borohydride was added and the whole mixture was allowed to incubate for another 30 minutes at room temperature, also in the dark. The concentration of sodium [3H]-borohydride was typically double the concentration of PLP that had been used in the first step of incubation. Following incubation, the beads were washed five times with homogenizing buffer to remove residual free [3H]-borohydride and then heated for 10 minutes at 95°C in SDS sample buffer. The beads were then removed by brief centrifugation using a microcentrifuge (1 min, 10,000 x g) and the supernatant was removed and subjected to SDS-PAGE analysis. Following SDS-PAGE analysis, the gels were stained and dried as outlined in a later section. Each gel lane was cut into 11 x 0.5 cm slices, and each slice was placed into a 3 mL scintillation vial. Approximately 200 μL of hydrogen peroxide (enough to cover the gel slice) was added to each vial, and the vials were capped and heated at 60°C for one hour. The remaining hydrogen peroxide was allowed to evaporate and then 400 mL of distilled water was added to each vial and the vials were allowed to shake for one hour. 4 mL of ACS scintillation fluid was then added to each vial prior to counting using a Beckman LS6000IC Scintillation Counter. Specific activities were determined in triplicate by adding a known volume of the [3H]-NaBH4 solution to separate vials containing 4 mL of ACS scintillation fluid.  59  2.2.8 Sucrose gradient centrifugation Following ammonium sulfate precipitation (40% saturation), the protein precipitate containing ACC was resuspended in a minimum volume of homogenizing buffer and incubated in the absence or presence of 20 mM citrate at 37°C. The sucrose gradient was prepared by layering decreasing concentrations of sucrose (60% to 20% in 10% increments, each layer being approximately 0.5 mL, w/v) in 3.2 mL TL-100 ultracentrifuge tubes. For ACC samples that were incubated in the presence of 20 mM citrate, the sucrose gradient also contained 20 mM citrate. The incubated ACC sample was then layered on top of the sucrose gradient solutions and then centrifuged in a fixed angle rotor (TLA110) in a benchtop Beckman ultracentrifuge for one hour at 100,000 rpm (290,000 x g). The sucrose gradients were spun at 4°C for ACC samples in the absence of citrate, and at room temperature if the ACC samples were incubated in the presence of 20 mM citrate. Following centrifugation, the tubes were separated into fractions by removing solution from the bottom of the tubes.  2.2.9 Size exclusion chromatography using a BioGelA column Size exclusion chromatography was performed using an AKTA FPLC (GE Health Sciences) operated through a workstation using Unicorn software (version 5.0). The column (dimensions 31 cm height x 0.9 cm diameter) was packed with BioGelA-50M beads (exclusion limit > 2000 kDa). Under typical running conditions, the operating pressure was close to 0.5 MPa at a flow rate of 20mL/hour, well within the maximum recommended operating pressure of 3.0 MPa. ACC was recovered from rat liver by ammonium sulfate precipitation as described previously (40% saturation), re-dissolved proteins were kept on ice, and centrifuged briefly (30 seconds at 11,000 x g) prior to column loading (typically 400 to 500 μL being applied). For analysis of non-polymerized ACC, chromatography was carried out in the absence of citrate and the protein samples were chromatographed at 4°C in BioGelA buffer (20 mM MOPS, 2 mM EDTA, 2% (w/v) glycerol and 100 mM NaCl). To examine polymeric forms of ACC, the same liver preparations were first incubated at 37°C for 30 minutes in the presence of 20 mM citrate, and centrifuged briefly (30  60  seconds at 11,000 x g) prior to chromatography at room temperature in the same BioGelA buffer that had been supplemented with 20mM sodium citrate.  2.2.10 Preparation of acetyl-CoA Acetyl-CoA was prepared by adding 10-50 mg free Coenzyme A (tri-lithium salt, from Sigma) to 1.5 mL distilled water. While stirring on ice and monitoring pH, solid NaHCO3 was added slowly to the mixture until the pH was approximately 7.5. 20 μL of acetic anhydride was then added using a syringe pre-washed with ethanol to remove traces of water. Over the next 3-5 minutes incubation with stirring on ice, solid NaHCO3 was added, if necessary, to maintain a pH of 7. After this 5-minute incubation. 5N HCl was added dropwise to the mixture until evolution of carbon dioxide had stopped and further dropwise addition of 1N HCl was used to bring the pH to a value between 4.5 and 5. The mixture was allowed to stir for another five minutes, and the volume of the mixture was brought up to 3 mL with distilled deionized water. To determine the final concentration of each acetyl-CoA preparation, a citrate synthase spectrophotometric assay was used, in the presence of DTNB to trap released free Coenzyme ASH. Briefly, the assay relies on the stoichiometric conversion of acetylCoA into free CoASH following the addition of oxaloacetate and citrate synthase: Oxaloacetate + Acetyl-CoA Æ Citrate + CoASH  The free coenzyme A produced then reacts with DTNB in the assay buffer to form 5merapto-2-nitrobenzoic acid, which absorbs strongly in the visible range at 410nm. The assay buffer comprised 50mM Tris buffer, pH 7.5, containing 0.1mM DTNB and to this was freshly added a final concentration of 0.25 mM oxaloacetate (also prepared as a stock solution in 50mM Tris buffer, pH 7.5), and 5 μL of the acetyl-CoA preparation. The reaction was initiated with 2 μg of citrate synthase (Sigma, specific activity 100 mU/μg) and the absorbance at 410 nm was monitored until it reached a stable plateau, usually within 2-3 min at room temperature. Acetyl-CoA concentration was calculated assuming the molar extinction coefficient of DTNB to be 13.6 M-1cm-1. Typically, the  61  final concentration of acetyl-CoA stock solutions was in the range 10 mM and these were stored in small aliquots at -20oC.  2.2.11 ACC activity assays The [14C]-HCO3 fixation method developed by Martin et al [87] and Halestrap et al [89] was used to determine ACC activity, one unit of activity being defined as the amount of ACC needed to convert one μmole of acetyl-CoA to malonyl-CoA in one minute. Assay buffer and assay incubation mixtures were kept on ice until pre-warming for assay and in tightly-capped microcentrifuge tubes to minimize loss of bicarbonate and exchange of labeled substrate with the atmosphere. Prior to assay, protein samples were pre-incubated for 20-30 minutes at 37°C, to allow allosteric activation of ACC, in homogenizing buffer containing citrate (varying concentrations) and BSA (2 mg/mL). Following pre-incubation, the assay was initiated by adding a 25 μL aliquot of the activated ACC preparation to 250 μL of pre-warmed assay buffer (37°C). The assay buffer consisted of 50 mM HEPES (pH 7.2), 10 mM MgSO4, 0.5 mM EDTA, 5 mM ATP, 7.5 mM glutathione, 2 mg/mL BSA, 150 μM acetyl-CoA, 7.5 mM KHCO3 and 7.5 mM [14C]-NaHCO3 (approximately 1000 dpm/nmole). Assays were terminated by adding 150 μL of 2M HCl, the tubes were allowed to incubate at room temperature for ten minutes and then centrifuged using a microcentrifuge at 13,200 rpm (16,000 x g) for three minutes. 300 μL of the assay mixture was then transferred to a liquid scintillation vial and evaporated to dryness either under a stream of air or under vacuum. Drying under acidic conditions leaves a residue containing [14C]-malonyl-CoA but allows essentially complete removal of unincorporated [14C]-HCO3, leaving background 14C levels of only 20-40 dpm. Dried assay residues were redissolved with 400 μL of distilled water with gentle shaking at room temperature for at least 1 hour. 4 mL of ACS scintillation fluid was then added to each vial prior to counting using a Beckman LS6000IC scintillation counter. The specific activity of [14 C]-bicarbonate was determined in triplicate each day by adding 10 μL of assay buffer to 4 mL ACS fluid containing 200 μL of the CO2-trapping reagent 2-phenylethylamine.  62  Blanks were also measured by adding 2M HCl to assay buffer before the addition of the pre-incubated enzyme sample. The effects of PLP on ACC activity were tested with slightly modified assay protocols. In some experiments the enzyme sample was first incubated with PLP (various concentrations) at 37°C for 30 minutes and then with citrate (varying concentrations) and 2 mg/mL BSA for another 30 minutes at 37°C prior to the assay. In another series of experiments, the incubation with PLP and citrate were reversed prior to assay. In both of these series of experiments, the concentrations of ATP, [14C]NaHCO3, and acetyl-CoA in the final assay buffers were varied as indicated to allow specific substrate kinetics to be evaluated.  2.2.12 SDS-PAGE 2.2.12.1 Sample preparation Protein samples were typically mixed directly with sample loading buffer, containing β-mercaptoethanol, in a 1:4 ratio, provided the initial protein concentration was adequate. TCA precipitation was used to concentrate samples or to remove problematic buffer components, particularly for protein samples in buffers with salt concentrations greater than 100 mM. An appropriate volume of TCA (100%, w/v) was added to the protein sample, to give a final concentration of TCA of 10% (w/v). After vortexing, the samples were incubated on ice for 30 minutes and then centrifuged at 4°C for 4 min at 13,200 rpm (16,000 x g). The supernatant was discarded, the pellet washed twice with acetone and then allowed to air dry before the addition of sample loading buffer. Proteins were fully dissolved in sample buffer by heating at 95°C for two to five minutes. The standard sample buffer was Tris-HCl, pH 6.8, containing sucrose (250 mM), bromophenol blue (0.2 mg/mL), and SDS (10% w/v).  2.2.12.2 SDS-PAGE analysis A discontinuous sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), based on Laemmli’s procedure [154] was used, in most cases with “mini”gels (10 cm x 10 cm). To separate and detect both ACC isoform bands, the concentration of acrylamide was only 3% (w/v) in stacking gels and 6% (w/v) in separating acrylamide  63  gels, so it was helpful to increase the gel strength by using piperazine diacrylamide as the crosslinker, instead of bis-acrylamide. For the separation of smaller proteins it was preferable to use acrylamide concentrations of 5% (w/v) in stacking gels and 8% (w/v) in separating gels. Stacking gels were formed by mixing tris pH 6.8 buffer (125 mM) and acrylamide:PDA (3%: 0.08%), followed by 50 μL APS (10% w/v) and 15 μL of TEMED to induce polymerization. Separating gels were formed by mixing Tris buffer, pH 8.8 (375 mM) and acrylamide: PDA (3%: 0.08%), followed by 70 μL APS (10% w/v) and 15 μL TEMED. Electrophoresis was typically carried out for 1.5 hours at 130 volts and 50 milliamps. Amersham Rainbow markers were used for gels that were stained and ECL DualVue Western blot markers were used for gels for subsequent transfer and immunoblotting. Following electrophoresis, the gels were removed from the glass plates, rinsed briefly in distilled water and stained with Coomassie Blue stain (0.25% w/v Coomassie Blue, 10% v/v acetic acid, and 45% v/v methanol). Following staining for approximately one hour, gels were destained with several changes of destaining solution (10% v/v acetic acid and 45% v/v methanol) and finally stabilized by soaking in aqueous 2% (v/v) glycerol. For permanent storage, the gels were placed between two sheets of cellulose clipped to a glass plate and allowed to air dry at room temperature.  2.2.13 Western blotting 2.2.13.1 Transfer and blocking Following electrophoresis, gels were rinsed briefly in distilled water and then incubated in Towbin buffer (48 mM Tris and 39 mM Glycine). A PVDF membrane was cut to the same size as the gel and prepared by soaking in 100% methanol for 15 seconds, followed by incubation for 2 min in distilled water. The membrane was then incubated in Anode II buffer (25 mM Tris, 10% methanol v/v, pH 10.4) for at least 5 minutes. Proteins were transferred from the gel to the membrane using a semi-dry transfer apparatus (BioRad). The transfer sandwich was assembled as follows: 2 sheets of Whatman 3MM filter paper (cut to the same size as the gel) soaked in Anode I buffer (300mM Tris, 10% v/v methanol, pH 10.4), 1 sheet of filter paper soaked in Anode II buffer, the pre-soaked gel, PVDF membrane, and 3 sheets of filter paper soaked in  64  Cathode buffer (25 mM Tris, 10% v/v methanol, pH 9.4). After assembly, the transfer was achieved by subjecting the sandwich to a current of 3mA/cm2 for 90 minutes. Following transfer, the membranes were soaked in Wash buffer (20mM Tris, 137 mM NaCl, 0.1% v/v Tween-20) for one minute. Membranes were then blocked in blocking buffer (5% w/v BSA, in Wash buffer) for 90 minutes. To check the success of the transfer, the gels were stained in Coomassie Blue stain, and destained as mentioned previously.  2.2.13.2 Incubations with primary antibody and streptavidin-HRP After blocking, the membranes were incubated for 30 minutes with the indicated primary antibody. Primary antibodies used at a dilution of 1:5,000 in blocking buffer included those able to recognize the Ser-79 phospho form of ACC-1 (Upstate, 07-303) and the “pan” anti-ACC antibody (Upstate, 04-322). It was found that antibody solutions could be used up to three times with little loss of signal intensity. Other antibodies included those directed against alpha tubulin and fatty acid synthase (FASN), both of which were from Santa Cruz (catalog numbers, SC-5286 and SC-20140, respectively) and were used at a 1:500 dilution in blocking buffer. Following incubation with primary antibodies, the membranes were washed for a total of 15 minutes with wash buffer, during which time the buffer was changed every 3 minutes. The membranes were then incubated for 30 min with secondary antibody (Goat anti-rabbit IgG-HRP, 1:50000 dilution in blocking buffer). Following incubation with the secondary antibody solution, the membranes were washed for a total of 40 minutes, the wash buffer being refreshed every 5 minutes. Alternatively, if streptavidin-HRP was used to detect all biotinylated protein subunits, the membranes were first washed for a total of 12 minutes, with the buffer being changed every 3 minutes. Then, the membranes were incubated for 2 hours in Amersham Streptavidin-HRP complex solution (1:5000 dilution, made up in 2% w/v BSA in wash buffer). After incubation, the streptavidin-HRP solution was removed and the membranes were washed for a total of 30 minutes, the wash buffer being changed every 5 minutes.  65  2.2.13.3 Sample detection The membranes were incubated for 5 minutes in a working solution containing substrates for HRP (Pierce SuperSignal West Pico peroxide/luminal) and then within the next hour, the membranes were exposed to photographic film (Kodak Biomax light film, Amersham Biosciences) for varying times as indicated and the films were developed with an automated film developer (SRX-101A Medical film processor, Konica Minolta Medical and Graphic Inc.).  2.2.13.4 Membrane stripping and staining In some experiments, membranes were stripped and re-probed with a different antibody. If the membrane had been previously probed with Streptavidin-HRP, then the membrane was incubated for 20 minutes with stripping buffer (commercially purchased from Pierce). If the membrane had been previously probed with primary and secondary antibodies, then the membrane was incubated for 10 minutes with stripping buffer, washed for 1 minute in wash buffer, and incubated with a fresh batch of stripping buffer for a further 10 minutes. Following either procedure, the membrane was washed for 12 minutes, the wash buffer being changed every 3 minutes. The membranes were then again treated with blocking buffer before starting the procedure with another antibody. Once a satisfactory film image had been obtained, the membranes were stained in Amido black solution (40% v/v methanol, 10% v/v glacial acetic acid, 0.1% w/v amido black) for four minutes, and then destained. The membranes were then rinsed in distilled water, allowed to dry and kept as a permanent record. To quantitate protein bands on the film images, film images were scanned using a computer scanner and the subsequent images were analyzed using Un-scan-it gel automated digitizing system (version 5.1, Silk scientific corporation).  66  2.2.14 Spectophotometric assays of Fatty acid synthase (FASN) The spectrophotometric assay of FASN is based on the consumption of NADPH (extinction coefficient 6.27 M-1 cm-1) [155]. The sample being tested for FASN activity was incubated in a buffer containing 200 mM KHPO4, 1 mM DTT, 1 mM EDTA, and 0.83 mM NADPH (to give a starting OD340 of 1.5) for one hour at 37°C. This incubation is considered important to allow the homodimeric FASN to assume a fully active conformation after storage. To initiate the reaction, a mixture of acetyl-CoA and malonyl-CoA (final assay concentrations of 30μM and 40μM respectively) was added to the incubation and the absorbance at 340 nm was monitored for 3 minutes at 37°C with a Lambda 35 UV/VIS spectrophotometer (Perkin Elmer Instruments) operated through a workstation using UV WinLab software (version 2.85.04). One unit of FASN activity is the activity required to produce one micromole NADPH per minute under the conditions described. This activity therefore represents the activity for each cycle of addition of a 2-carbon unit to the growing fatty acyl chain.  2.2.15 Mass spectrometry 2.2.15.1 Sample preparation and in-solution digestion For mass analysis, samples containing ACC were obtained from rat liver or adipose tissue by ammonium sulfate precipitation followed by two consecutive size exclusion chromatography steps using the BioGelA-50M column. The size exclusion steps were carried out consecutively in the absence and then in the presence of 20mM citrate. The high molecular weight ACC polymeric fractions were pooled and concentrated with buffer exchange to 50mM ammonium bicarbonate containing 6M urea. The protein concentration was determined using the Bradford assay [156] and the sample then reduced with 10mM DTT (final concentration) for one hour at room temperature. Following reduction, the protein sample was alkylated with 40mM iodoacetamide (final concentration) for one hour at room temperature after which unreacted iodoacetamide was removed by incubation for one hour with 40 mM DTT. Following reduction and alkylation, the protein sample was diluted to lower the urea concentration to approximately 0.6M and then trypsin was added to give a protease  67  to substrate ratio of 1 to 50. This mixture was allowed to incubate overnight at 37°C and the reaction stopped with sufficient concentrated acetic acid to reduce the pH below 6. Following trypsin digestion, peptides were recovered by binding to C18 Zip-tips (Millipore) to concentrate and washed to remove all buffer components. The bound peptides were washed with 0.1% TFA (v/v) and eluted first with 50% acetonitrile (v/v), followed by a second elution with 80% acetonitrile (v/v). The combined eluted samples were then submitted to the MSL/LMB Proteomics Core Facility at UBC for tandem MS/MS analysis.  2.2.15.2 Sample preparation and in-gel digestion In addition to the procedure just described, in which protein mixtures were analyzed directly after size-exclusion chromatography, a second complementary approach was used in which proteins were subjected to separation by SDS-PAGE prior to trypsin digestion. In this second approach, the ACC polymeric fraction recovered following ammonium sulfate precipitation and the two-step size exclusion chromatography method were also subjected to SDS-PAGE so that individual protein bands could be selected for MS/MS analysis. Following size-exclusion chromatography, the polymeric fractions were pooled, concentrated and subjected to SDS-PAGE using acrylamide concentrations of 5% (w/v) in the stacking gels and 8% (w/v) in the separating gels. The gels were then stained using Simply Blue stain (Invitrogen) and appropriate bands carefully excised; each band typically being 1-2 mm long, 1 cm wide and 1.5 mm thick. Each band was then cut into approximately 1 mm in cubes and sequentially incubated at room temperature first with acetonitrile (20% v/v, 1M ammonium bicarbonate) and then with methanol (50% v/v, 5% v/v acetic acid) each for one hour. The two washes were repeated until all the stain was removed from the gel pieces. Two cycles of washes was usually sufficient to remove the stain. After washing, the gel pieces were dehydrated by incubation with 100% acetonitrile for five minutes, dried under a stream of nitrogen gas and then rehydrated with a minimum volume of 10 mM DTT for 30 minutes at room temperature. Following rehydration in DTT, a volume of 100 mM iodoacetamide sufficient to cover the gel pieces was added and the incubation continued for another 30 minutes at room  68  temperature. Following this alkylation step, the gel pieces were once again dehydrated by incubation for five minutes with 100% acetonitrile, rehydrated by incubation with 100 mM ammonium bicarbonate for ten minutes and finally dried under a stream of nitrogen gas. The dried gel pieces were then rehydrated on ice with a minimum volume of trypsin solution (20 ng/μL, in 50mM ammonium bicarbonate) and a further volume of 5 mM ammonium bicarbonate was added, equivalent to 1/5th of the volume of trypsin solution used and the digestion was allowed to continue overnight at 37°C. Following trypsin digestion, an equal volume of 50mM ammonium bicarbonate was added and ten minutes later the mixture was centrifuged and the supernatant containing the peptides was retained. The gel pieces were re-extracted twice with a minimum volume of 10% (v/v) formic acid. The two formic acid washes were then added to the initial peptide sample volume. The peptides in the combined sample and washes were then recovered using C18-zip-tips as described before.  69  2.2.15.3 LC MS/MS analysis Liquid chromatography and tandem mass spectrometry (LC MS/MS) analysis of the samples for peptide identification were done at the University of British Columbia (UBC) Michael Smith Laboratory/Laboratory for Molecular Biophysics Proteomics Core. Liquid chromatography was carried out using an LC packings system and this was coupled to a tandem quadrupole time-of-flight mass spectrometer (QSTAR Pulsar, Applied Biosystems, Foster City, CA). Peptide samples were resuspended in 2% (v/v) acetonitrile, 0.1% (w/v) formic acid and loaded onto a small pre-clearing column (inside diameter 300 μm, length 5 mm) that contained PepMap C18-100 micro resin (LC Packings). The outflow from the precolumn was fed directly into the main separating column (inside diameter 20μm, length 65 cm), that was packed with PepMap C18 resin (LC Packings). Column equilibration and initial washing was carried out using a mixture of 95% Buffer A (2% acetonitrile, 0.1% formic acid) and 5% Buffer B (85% acetonitrile, 0.1% formic acid). The peptides were eluted with a gradient of 5%-40% buffer B. The mass spectrometry data were acquired twice using the Analyst QS software (Applied Biosystems) for data acquisition based on a 1 second MS survey scan in the ranges 300 to 700 m/z and 7001500 m/z, followed by up to three MS/MS scans of three seconds each. Nitrogen gas was used as the collision gas and the ionization tip voltage was 2200 volts. The error window for MS was set to 1.2 Da and for MS/MS was set to 0.6 Da. Following acquisition, the data was searched using Mascot [157], which encompasses three methods for peptide mass fingerprint, sequence query, and MS/MS ion search. Within Mascot, the experimentally derived values are matched to the MSDB and NCBI databases and mowse scores, for significance of results, were calculated.  70  2.2.16 Immunocytochemistry Primary rat hepatocytes were kindly supplied by the laboratory of Dr. Thomas Chang, Faculty of Pharmaceutical Sciences, UBC. The hepatocytes were cultured on plastic microscope cover slips that had been pre-coated with either matri-gel (VWR) or fibronectin (Sigma). After pre-coating, the cover slips were given three one-minute washes with PBS and then stored in PBS at 4°C for up to 24 hours until use. Hepatocytes were recovered from rat liver during the morning as outlined by Chang et al [158] and cultured overnight at a density of approximately 3.0 x 105 cells per well in 12 well culture dishes for use the following day. During culture, the cells were kept in a humidified 37°C incubator with a gas mixture containing 95% O2:5% CO2. The time between hepatocyte isolation and immunocytochemistry experiments was therefore in the range 15-20 hours. Just prior to fixing, the cells were treated under conditions designed to influence ACC activity, including incubation with colchicine, insulin, or anoxic stress to activate AMPK. The cells were then fixed with 3.7% (v/v) formaldehyde and allowed to incubate in fixing solution for 10-15 minutes at 4°C. Following fixing, the cover slips were washed four times with PBS-CM (PBS containing 0.1 mM CaCl2 and 1.0 mM MgCl2) each wash lasting ten minutes. The cover slips were then incubated in blocking solution (1% w/v BSA, 0.1% v/v Triton X-100, in PBS-CM) for 30 minutes. Following blocking, the cover slips were incubated with required primary antibodies (diluted in blocking buffer) for 30 minutes. Primary antibodies included anti-acetyl-K40 tubulin (Abcam, ab11323), anti-ACC1 (Upstate), or anti-phospho ACC (Upstate). The cover slips were given two ten-minute washes with PBS-CM and then incubated with the secondary antibodies (anti-mouse Alexa 488 or anti-rabbit Alexa 568, Abcam) and Hoechst stain for 30 minutes. The cover slips were then given two final 10-minute washes with PBS-CM, mounted onto microscope slides and sealed with nail polish. Once mounted, the cells were then visualized using a Nikon Elipse E400 system fluorescence microscope or an Olympus FV1000 Confocal system. Cells were examined typically at 60 x magnification with an oil immersion lens. Cell images were typically attained at a speed of 8-12 μs/pixel and with a resolution of 1024 x 1024. The background for each cell image was also adjusted for each laser to improve  71  the cell image. Following confocal microscopy, the images were viewed using FV10ASW viewer (version 1.6).  72  Results and Discussion 3  Chapter 3: Inhibition of ACC by PLP  3.1 Rationale Even before the definition of the enzymes involved, it had been recognized that di- and tri-carboxylic acids, especially isocitrate and citrate, could stimulate rates of fatty acid synthesis in cell extracts using acetate or acetyl-CoA as the initial substrate. Subsequently, it was revealed that the effects of citrate were accounted for by the ability of the carboxylic acid to allosterically activate ACC. Although the effects of citrate have been known and studied for many years, the citrate binding site of ACC has not yet been defined, mainly due to lack of understanding of mammalian ACC structure. It should be noted that prokaryotic and yeast ACCs are apparently not citrate sensitive and therefore knowledge of these structures is not necessarily instructive. To obtain further insight into the ACC citrate-binding site, approaches used to examine the citrate-binding sites of other proteins were evaluated. In fact, relatively few mammalian proteins are known to utilize or show sensitivity to citrate. Examples of these include citrate synthase, ATP-citrate lyase, citrate transporter proteins and phosphofructokinase (PFK). PFK exists in two cytosolic isoforms (PFK-1 and PFK-2), that catalyze the formation of fructose 1,6-bisphosphate or fructose 2,6-bisphosphate from fructose 6-phosphate [159]. PFK was of particular interest because, like ACC, it has a citrate-binding site, as citrate is an allosteric regulator. Of particular interest were studies carried out by Colombo et al [159], and Uyeda [160], in which it was shown that PFK is inhibited by PLP and because the enzyme could be protected from PLP by pre-incubating with citrate, it was hypothesized that PLP was able to bind to the citrate binding site of the enzyme. This hypothesis was supported by labeling PFK with PLP and reducing the Schiff base with sodium borohydride. Using this approach, PLP was found to specifically bind to Lys-529 in PFK, which is also known to be within the citrate-binding site [161].  73  Figure 3.1: Chemical structure of citrate and pyridoxal 5-phosphate. The structures of citrate (a) and PLP (b) are oriented to show possible chemical analogy. a)  O  -  O  O O  -  OH  O  b)  O  O  O  O  -  O  -  P O  -  H3C  -  O  N  Based on those previous studies of PFK, the possibility that PLP might also be used to probe the citrate-binding site of ACC was explored. Investigations of the effects of PLP on ACC were first initiated by Jason Elliott [151], who found that PLP indeed had an inhibitory effect on ACC activity. He also found that the inhibitory effect of PLP was decreased when ACC was first incubated with citrate (figure 3.2). When PLP was added before citrate, ACC activity was inhibited by more than 80 percent, but following citrate pre-incubation, PLP reduced ACC activity by only 20 to 30 percent. Based on these preliminary studies, the kinetics of PLP inhibition with respect to each substrate of the ACC reaction were studied. The purpose of this approach was to rigorously test the possibility that PLP might interact with a substrate-binding site rather than with the citrate-binding site of ACC. Following the kinetic analysis, approaches were developed to achieve irreversible binding of PLP to ACC based on the assumption that the interaction likely involved the formation of a Schiff base between PLP and a lysyl amino group of ACC. If correct, the Schiff base could then be reduced with sodium borohydride to produce a covalent adduct and furthermore allow the introduction of a radiolabel by using [3H]-borohydride. The aim here was to assess the specificity of PLP interaction, as ACC contains so many lysine residues that might conceivably interact with PLP. This work would also potentially set the stage to determine the exact site of PLP binding.  74  ACC Activity (% Maximum)  Figure 3.2: Citrate protection of ACC decreases PLP inhibition. Liver ACC was prepared by ammonium sulfate precipitation (40% saturation) and incubated with PLP (0.5 mM) for the indicated times prior to the assay before ( z ) or after ( { ) incubation for 30 minutes with 20 mM citrate. Results were obtained and analyzed by Jason Elliott and are from one experiment that was repeated with very similar results. 100  A 80 60 40 20 0 0  5  10  15  20  25  PLP Incubation (min)  75  3.2 Non-competitive inhibition of ACC by PLP The nature and extent of PLP inhibition was studied with respect to the three substrates of the ACC reaction. ACC that had been purified from rat liver through ammonium sulfate precipitation was incubated with varying concentrations of PLP followed by incubation with 20 mM citrate. The incubated ACC preparations were then assayed using buffer in which the concentrations of two substrates remained constant while the concentration of the third was varied. Primary and secondary plots were then derived for each varying substrate: bicarbonate (figure 3.3), ATP (figure 3.4), and acetylCoA (figure 3.5). Analysis of the primary and secondary plots shows that for all three substrates, the Vmax is lowered in a concentration-dependent manner following incubation with PLP, while the Km values were not significantly altered by PLP treatment. This set of results is indicative of non-competitive or mixed inhibition, meaning that PLP inhibits ACC by binding to a site other than that required for substrate binding. Interestingly, the secondary plots with respect to varying ATP concentration gave non-linear results, in accordance with previous studies [18, 33]. The kinetic constants for these reactions were determined manually through Lineweaver-Burke secondary plots and by using SigmaPlot 9.01, both methods giving similar results. The inhibition constants for the effects of PLP were calculated to be 167 ± 32 μM (with respect to bicarbonate), 133 ± 19 μM (for ATP), and 135 ± 69 μM (for acetyl-CoA). It is important to notice that these Ki values are very similar when calculated with respect to each of the three ACC substrates. If PLP were to exert its effects by binding to one of the substrate-binding sites, then one might have expected to see differences in the Ki values. Specifically, one might expect that the Ki with respect to acetyl-CoA would differ from the Ki values with respect to bicarbonate and ATP because acetyl-CoA must bind to a discrete active site of the carboxyltransferase domain, which is quite separate from the biotin carboxylase domain that must bind the other two substrates. The results of the enzyme kinetic analysis as well as previous findings concerning the effect of PLP on PFK indicate that PLP may conceivably bind to the citrate-binding site of ACC. The results are further supported by the fact that the effects of PLP are diminished by prior exposure to citrate (figure 3.2) [162]. These latter results were 76  published in the same JBC paper although those particular studies of the effect with citrate prior to PLP were carried out by Jason Elliott. Taken together, these experiments demonstrate that PLP is a potent inhibitor of ACC, exerting effects on ACC activity that can be seen after only a few minutes of incubation. Importantly, pre-incubation of ACC with citrate significantly diminished the effect of PLP on enzyme activity. In previous studies of the effects of PLP on PFK [161], the authors had achieved covalent coupling between PLP and the enzyme using sodium borohydride. Therefore, the next step was to attempt to label ACC with PLP in an attempt to determine the location of its binding to ACC.  77  Figure 3.3: Effects of PLP on bicarbonate concentration dependence of the ACC reaction. ACC was partially purified from rat liver by ammonium sulfate precipitation and incubated in the absence (○) or presence of PLP at 0.05mM (□), 0.1mM (▲), 0.25mM (■), or 0.5mM (♦) for 30 minutes, followed by a second 30 minute incubation with 20mM citrate. ACC activity was then determined with the indicated varying concentrations of bicarbonate. The primary (A) and double reciprocal (B) plots represent the results of single experiments that were repeated three times with similar results. 2500  ACC Activity (mU/mL)  A 2000  1500  1000  500  0 0  2  4  6  8  10  Bicarbonate (mM)  B 0.006  1/v  0.003  0 -1  0  1  2  -0.003 1 / [Bicarbonate]  78  Figure 3.4: Effects of PLP on ATP concentration dependence of the ACC reaction. ACC was partially purified from rat liver by ammonium sulfate precipitation and incubated in the absence (○) or presence of PLP at 0.05mM (□), 0.1mM (▲), 0.25mM (■), or 0.5mM (♦) for 30 minutes, followed by a second 30 minute incubation with 20mM citrate. ACC activity was then determined with the indicated varying concentrations of ATP. The primary (A) and double reciprocal (B) plots represent the results of single experiments that were repeated three times with similar results. 1200  ACC Activity (mU/mL)  A 1000 800 600 400 200 0 0  1  2  3  4  5  ATP (mM)  0.02  B  1/v  0.01  0 -15  -5  5  15  25  -0.01 1 / [ATP]  79  Figure 3.5: Effects of PLP on acetyl-CoA concentration dependence of the ACC reaction. ACC was partially purified from rat liver by ammonium sulfate precipitation and incubated in the absence (○) or at varying PLP concentrations of 0.05mM (□), 0.1mM (▲), 0.25mM (■), or 0.5mM (♦), followed by a second 30 minute incubation with 20mM citrate. ACC activity was then determined with the indicated varying concentrations of acetyl-CoA. The primary (A) and double reciprocal (B) plots represent the results of single experiments that were repeated three times with similar results. 1200  800 600 400 200 0 0  100  200  300  Acetyl-CoA (μM)  0.015  B  0.01  1/v  ACC Activity(mU/mL)  A 1000  0.005  0 -0.03  0  0.03  0.06  -0.005 1 / [Acetyl-CoA]  80  3.3 Labeling of ACC with PLP and [3H]-borohydride It is well established that the Schiff base formed upon interaction between PLP and free amino groups such as the ε-amino group of protein lysyl side chains can be reduced with sodium borohydride. With this approach, previous studies have demonstrated that the use of sodium borohydride can lead to irreversible coupling of PLP to a specific lysine residue of PFK [161]. Once the initial Schiff base had been reduced to the corresponding amide linkage, forming a covalent bond between PLP and PFK, the enzyme was then digested with trypsin and PLP-labeled peptides were isolated. From this method, it was determined that PLP was attached primarily to lysine-529 of PFK-1 [163]. Based on these prior related studies, an attempt was made to determine whether ACC could also be covalently labeled by PLP and several approaches were considered. It was felt that the most straightforward method would be to inhibit ACC with PLP and then to incubate with sodium [3H]-borohydride. If PLP could be successfully coupled, then ACC would also become radiolabeled and when subjected to SDS-PAGE analysis would allow assessment of [3H] incorporation into ACC subunits. In addition to this basic strategy, it was also considered that a “solid” phase method would be advantageous because it would provide an opportunity to thoroughly wash the preparation at various stages and also potentially increase overall yield. A solid phase method that had initially been developed by Jason Elliott was therefore used. In this method, ACC was first immobilized by binding to tetrameric avidin-agarose beads. This allowed essentially irreversible binding of ACC through interaction with the intrinsic biotin to avidin (Kd ~ 10-15 M). Once bound, the ACC was stringently washed to remove any other residual proteins and then treated sequentially with PLP and [3H]-borohydride, finally being washed again to remove reagents prior to SDS-PAGE. In the rat liver preparations used, only ACC and pyruvate carboxylase remained bound to the tetrameric avidin beads as confirmed by western blotting. Immobilized ACC bound to avidin-agarose was then incubated with various concentrations of PLP, followed by incubation with [3H] borohydride. Following incubation, the beads were thoroughly washed to remove reagents and then digested in SDS sample buffer and subjected to SDS-PAGE. After staining to visualize the protein  81  bands, the gel was dried, and then each lane was cut into small slices and each slice was digested individually with hydrogen peroxide and the 3H-incorporation was measured (figure 3.6). As shown in figure 3.6, the ACC protein band consistently showed the highest 3  H-incorporation. In the absence of PLP, there was no 3H-incorporation indicating that  3  H-incorporation occurred only in the presence of PLP. It is important to note the  absence of any detectable 3H-incorporation into pyruvate carboxylase subunits that migrate at approximately 120 kDa (slices 6, 7). This is an important negative control because PC is closely related to ACC, particularly in the highly conserved biotin carboxylase catalytic center and biotin attachment sites. If PLP were binding to a lysine residue at or near the highly-conserved active site, one would therefore expect PC to show similar labeling. Significantly, PC lacks any citrate-binding site or citrate sensitivity [164]. Further studies were carried out to determine the PLP concentration-dependence of ACC labeling (figure 3.7). From these experiments, it can be seen that at higher concentrations of PLP (over 1 mM), the efficacy of PLP labeling is somewhat decreased. The optimum PLP concentration was found to be 0.75 to 1.0 mM.  82  Figure 3.6: Incorporation of [3H] into ACC following incubation with PLP and [3H]borohydride. ACC was immobilized onto tetrameric avidin-agarose beads and then incubated in the absence of PLP (♦), or in the presence of PLP at concentrations of 0.5 mM (■), 0.75 mM (▲), 1 mM (×), 1.25 mM (○), or 1.5 mM(●). Beads were then treated with [3H]borohydride and finally washed, digested with SDS-sample buffer prior to SDS-PAGE. Gels slices were digested with hydrogen peroxide and [3H] incorporation was determined. Gel slices are numbered from origin (1) to tracking dye (11). Numbered arrows indicated the migration of ACC subunits, 265 kDa and 280 kDa (1), myosin, 220 kDa (2), phosphorylase a, 98 kDa (3), and serum albumin, 68 kDa (4). Results are typical of three independent experiments. 1  4500  2  3  4  4000  Incorporated dpm  3500 3000 2500 2000 1500 1000 500 0 1  3  5  7  9  11  Gel Slice  83  Figure 3.7: Incorporation of [3H] into ACC following incubation with PLP and [3H]borohydride. ACC was immobilized on tetrameric avidin-agarose beads and then incubated in the presence of the indicated concentrations of PLP. [3H] incorporation was determined after SDS-PAGE and excision of the ACC subunits (ACC-1 and ACC-2 combined). 4500 4000 Incorporated dpm  3500 3000 2500 2000 1500 1000 500 0 0  0.5  1  1.5  PLP concentration (mM)  The molar stoichiometry of PLP incorporation into ACC was calculated based on the specific activity of [3H]-NaBH4 (80-100 dpm/pmol) and a specific activity of homogenous ACC of 2U/mg protein (determined in a large number of preparations). On this basis, the incorporation of PLP into ACC was calculated to be in the range of 0.2 to 0.4 mol/mol ACC subunit. This means that the labeling of ACC subunits is somewhat sub-stoichiometric. Considering that there are 112 lysines in the primary sequence of ACC-1, this incorporation level of PLP indicates that PLP binding to ACC is very specific, potentially occurring at a single lysyl site. The fact that substantial inhibition of ACC is observed under conditions that lead to less than 100% modification of the protein subunits may be consistent with the idea of “half site reactivity”. Thus, the fact that the minimal catalytic unit of ACC is a dimer, it is possible that inhibition of just one of each two subunits is sufficient to cause enzyme inactivation.  84  3.4 Summary Kinetic analysis with respect to all three ACC substrates revealed that PLP inhibition of ACC was mixed or non-competitive. Binding of ACC to tetrameric avidinagarose beads provided a convenient solid phase method to label ACC with PLP and sodium borohydride followed by extensive washing to remove reagents prior to SDSPAGE. With this method, it was found that PLP binds specifically to ACC and not to the related pyruvate carboxylase, which is also present and binds to tetrameric avidin-agarose beads. By testing a range of PLP concentrations (0 mM to 1.5 mM), the optimum concentration for labeling ACC with PLP was found to be 0.75 mM. The specificity of PLP interactions with ACC was further underlined by the fact that although ACC contains more than one hundred lysyl residues, PLP labeling occurred at a stoichiometry of somewhat less than 1 mol PLP per mol ACC subunits. Considering that PLP interferes with the citrate-binding site of other proteins, especially of PFK, PLP has the potential to be a useful tool for elucidating the citratebinding site of ACC [162]. Based on the [3H] labeling studies, I had planned to extend this work by applying mass spectrometry methods to try to define the PLP-binding site on ACC. This plan turned out to require the development of several techniques that could not be taken to completion in the time available. Two of the limiting factors proved to be the ability to generate sufficient quantities of PLP-labeled peptides and also the availability of a new high resolution mass spectrometry system (Qstar) that came online in the Proteomics Core centre only in 2009. Two key approaches that have therefore been significantly advanced but not completed at this time are a) to purify specific PLPlabeled peptides and b) to establish a protein ”footprinting” method to assess PLP binding sites. With respect to PLP-peptide purification, the key technique being developed is to use IMAC chromatography based on the assumption that the phosphate of PLP will facilitate binding to the metal ion matrix (Fe3+, Ga3+, or TiO2). With respect to the footprinting method, we have now been able to successfully demonstrate that the new MS/MS system provides a method to substantially sequence the entire ACC-1 and ACC2 isoforms in a single analysis (greater than 80% sequence coverage). This should now allow us to explore the ability of PLP to block one or more trypsin cleavage sites and thereby disrupt the pattern of tryptic peptides and sequences derived by LC MS/MS.  85  4  Chapter 4: Determining the presence of proteins associating with ACC  4.1 Rationale Based on in vivo analysis [111], it is clear that ACC must become highly active in intact tissues in order to account for observed rates of fatty acid synthesis. This essentially full enzyme activation occurs even though tissue citrate concentrations are sub-optimal (0.1-0.5 mM). Indeed, this level of activity of ACC is also achieved despite the presence of at least minimal levels of inhibitory ligands such as CoA esters and at least some level of phosphorylation of inhibitory sites. Evidently, other cellular factors are extremely important in facilitating ACC activation. Other specific observations further support this general idea. For example, cellular citrate concentrations appear not to change in a number of situations in which ACC activity changes [113]. For example, in earlier studies, Halestrap and Denton demonstrated in rat epididymal fat pads that insulin-induced ACC activation occurred in the absence of any change in citrate concentration [89]. In other studies, dramatic increases in intracellular citrate concentrations, following incubation of fat pads with pyruvate or fluoroacetate, were not accompanied by any corresponding changes in ACC activity [88]. Similarly, studies of ACC activity in rat liver during the starving/refeeding transition by Moir and Zammit led them to conclude that changes in cellular citrate or of ACC phosphorylation could not account for changes in ACC activity [165]. For example, at the first 2-4 hours of refeeding, there was a steady increase of ACC activity, in the absence of changes in citrate. Even 24 hours after refeeding, total ACC activity levels remained 50% lower than in livers of normally-fed (non-fasted) rats although similar cellular citrate concentrations were observed in tissues for all groups of rats. The authors further concluded that the observed changes in ACC phosphorylation also could not explain their results. The results of these experiments lead to three possible conclusions: ACC activity can change even when cellular citrate concentrations remain constant, changes in citrate concentration are not necessarily accompanied by changes in ACC activity, and the sensitivity of ACC to citrate can be altered both in response to phosphorylation and independently of phosphorylation.  86  Based on the foregoing discussion, it was considered that one reasonable hypothesis is that ACC activity might also be controlled through physical association with other cellular proteins. This general idea is in keeping with the metabolon concept introduced by Paul Srere in the late 1980’s. The term “metabolon” was used to define the organization of sequential enzymes of a metabolic pathway in the form of a noncovalently bound protein complex. It was predicted that these complexes would provide a mechanism to increase the efficiency of substrate channeling (passing the product of one enzyme to another enzyme) as well as opportunities for integrated control. Enzymes of the TCA cycle provided the earliest example of a metabolon [166]. Whether ACC is part of a metabolon, or not, it is possible that ACC is part of a protein complex that modulates its citrate sensitivity or some other property that enhances overall catalytic activity. In comparison to the other biotinylated carboxylases, ACC is a much larger and more complex protein and much of the ACC primary sequence has no obvious defined function. It is therefore not unreasonable to postulate that such “unassigned” domains might be important. A number of years ago, initial attempts were made to address this possibility and evidence for an ACC “regulator protein” that modulated ACC citrate sensitivity was produced from studies of rat liver ACC by Quayle et al [167]. In these studies, it was discovered that purifying ACC by monomeric avidin affinity chromatography resulted in lowering of citrate sensitivity compared to that of the less purified enzyme. Citrate sensitivity could then be restored by reconstituting ACC with proteins that had been removed in the flowthrough (FT) fractions of the avidin affinity purification column (figure 4.1). Based on these studies, it was hypothesized that an endogenous rat liver protein, or proteins, associate with ACC and increases its citrate sensitivity. Further studies led to initial characterization of the “ACC regulator” which was determined to be an approximately 75 kDa protein. Studies with 32P-labelled ACC demonstrated that this protein regulator did not lead to dephosphorylation of ACC. The reversibility of the effects of this protein regulator also demonstrated that this regulator protein did not lead to ACC proteolysis (an important safeguard given the proteolytic sensitivity of ACC). Ultimately, with few probes other than ACC activity profiles, the identification of the ACC regulator was not possible at that time.  87  Indirectly, studies involving the expression of recombinant forms of rat and human ACC first reported in 2007, showed that these recombinant forms of ACC also exhibit low citrate sensitivity and corresponding high Ka for citrate activation (see figure 4.1, red line) [168]. In conclusion, highly-purified preparations of native ACC from rat liver, rat adipose tissue, as well as expressed recombinant enzyme all have notably lower citrate sensitivity than forms of ACC in less-purified forms from rat tissues. As the effects of allosteric ligands and phosphorylation have been largely eliminated in these various preparations, it is most likely that the absence of other key protein has a major impact on the altered kinetic properties of ACC. Figure 4.1: Evidence for an ACC-associated “regulator” protein. This figure illustrates the results for two previous studies by others. Black lines (Quayle et al [167]) show citrate sensitivity of rat liver ACC before (○) and after (●) avidin affinity chromatography and also following recombination of avidin-purified ACC with the avidin column flowthrough (▲). The citrate sensitivity of a highly-purified preparation of recombinant human ACC-2 (red, Cheng et al [168]) is also markedly lower that that of ACC in less-purified forms.  ACC Activity (% max)  100  80  60  40  20  0 0  5  10  15  Citrate Concentration (mM)  88  Based on the earlier studies described in the rationale, general models were developed to test the interaction between ACC and ACC-associated proteins (figures 4.2 and 4.3). For example, ACC-associated proteins might promote the polymerization and activation of ACC (figure 4.2). According to this model, ACC is predicted to be activated and polymerized to a greater extent and with higher citrate sensitivity in the presence of these proteins (figure 4.2 b versus figure 4.2a). Within this general model, ACC might be activated by proteins that associate with ACC dimers and that positively facilitate polymerization, perhaps by a chaperone-like function. Such proteins may or may not remain in close association after polymerization but the model illustrates the case in which association does persist after polymerization. Activating proteins might also be visualized with the potential to bind only to ACC oligomers, thereby affecting the dimer and polymer equilibrium. In principal, associated proteins might also play roles in inhibition of ACC, so a model was also developed to cover this scenario (figure 4.3). For example, it is possible that ACC-associated proteins may block ACC polymerization by binding to ACC dimers, rendering the enzyme less active and refractory to the effects of citrate or more sensitive to the effects of allosteric inhibitors (figure 4.3). It is also possible that both activating and inhibitory mechanisms may exist, implying a multi-protein complex is formed between ACC and proteins that bind either to the dimeric form, to the polymeric form, or both. The presence of two ACC isoforms further complicates the proposed models. For example, distinct proteins may bind to the different ACC isoforms and the type and number of ACC-associated proteins may change depending on the proportion of the isoforms present. It is not known if ACC-2 can polymerize to the same extent or with the same ligand sensitivity as ACC-1 and the extent, properties and mechanism of ACC polymerization in the presence of both isoforms has not been evaluated. For example, it is unknown whether hetero-dimers are formed in which subunits of ACC-1 and ACC-2 combine, or if only homo-dimers exist. My initial goal was to develop methods to search for ACC-associated proteins that might lead to the activation of ACC. First, the purification of ACC using a monomeric avidin-agarose column was attempted, as outlined by Quayle et al. However,  89  these attempts with several commercial preparations of monomeric avidin-agarose were unsuccessful and led to recovery of very limited quantities of ACC or any other proteins. An alternative method was therefore essential and was developed based on the unique citrate-induced activation and polymerization of ACC. The main aim was to examine the possibility that ACC polymers might involve the association of ACC with other proteins as well as the established association between ACC dimers. In general terms, the approach that was planned was to isolate ACC polymers using sedimentation or size exclusion chromatography and to analyze the protein composition of the ACC polymeric fractions.  90  Figure 4.2: Model for ACC activation by associated proteins. In the absence of any associated proteins, there is minimal ACC polymerization (a). Associated proteins might bind only to the polymeric form of ACC thereby stabilizing the polymeric form of ACC (b). Associated proteins might instead bind to ACC dimers and thereby positively facilitate polymerization (c). Associated proteins could in principle influence the Vmax (d, black +) or the Ka for citrate (d, red +), or substrate kinetics (not shown).  d)  91  Figure 4.3: Model for ACC inhibition by associated proteins. In the absence of any associated proteins, ACC polymerizes (a). Associated inhibitory proteins might bind to the polymeric form of ACC to promote depolymerization or to the dimeric form of ACC thus inhibiting polymerization (b). The impact of the inhibitory proteins might be reflected in the Vmax (c, black +), or Ka for citrate (c, red +) or substrate kinetics (not shown).  c)  92  4.2 Purification of ACC “polymers” and ACC “dimers” through size exclusion chromatography Originally, the identification of a “regulator protein” by Quayle et al was demonstrated through the use of monomeric avidin-agarose column. Several attempts were made to repeat these experiments but technical difficulties meant that these attempts were unsuccessful. In addition, it was already known that the avidin method would require the purification of putative ACC-associated proteins from the very complex mixture that was removed from ACC during affinity chromatography. Such a complex mixture would therefore make identification very difficult. For these reasons, alternative techniques for ACC purification independent of the avidin purification approaches were chosen. In considering avidin-independent purification approaches, I chose to exploit the citrate-induced polymerization of ACC that is such a unique characteristic. The initial working hypothesis was that the regulatory protein(s) would bind to the polymeric form of ACC and so the purification of the polymeric fraction would also in principal provide a method to purify the regulatory protein. Additionally, this approach would also simplify the mixture, potentially making identification of the regulatory protein more feasible. To separate the polymeric and dimeric forms of ACC, sucrose gradient centrifugation was first investigated, due to its ease and speed of separation within a few hours. Sucrose gradient centrifugation has also been used extensively by many previous investigators to examine the citrate-induced polymerization characteristics of ACC. For example, in studies by Vagelos et al [11], and Gregolin et al [100], sucrose gradients were relatively shallow (0 to 20% w/v), and were done at the relatively low speeds of 30,000 to 40,000 rpm for about three hours. Since these earlier studies, developments in ultracentrifuge design have allowed for higher speeds and smaller sample volumes, so a step sucrose gradient method was developed to separate the polymeric and dimeric forms of ACC that was effective with relatively short centrifugation time. This sucrose step gradient was formed and centrifuged as outlined in Methods. Following centrifugation, six fractions were removed, with the top sample layer of the step gradient being designated fraction #1, and  93  fraction 6 being at the bottom of the step gradient. The ACC activity of each fraction was then measured (figure 4.4). It was expected that in the absence of citrate, ACC would remain largely dimeric (near the top of the tube), while in the presence of citrate, ACC would polymerize and would sediment lower in the tube. Indeed, this was demonstrated as can be seen in the activity profile. Figure 4.4: Separation of the polymeric and dimeric forms of ACC by sucrose gradient centrifugation. ACC was prepared from rat liver through ammonium sulfate precipitation, and incubated in the absence or presence of 20 mM citrate. Following incubation, ACC was layered onto a sucrose step gradient and centrifuged at 100,000 rpm (290,000 x g) for one hour at 4°C in the absence (♦) or at room temperature in presence of 20 mM citrate (■). The sucrose gradient was then separated into six fractions, ACC activity was determined in each fraction and expressed as percent total ACC activity per tube. The experiment was repeated twice with the average of the two results shown. 45 40  % ACC Activity  35 30 25 20 15 10 5 0 1  Top 10% sucrose  2  3  4  Fraction Number  5  6  Bottom 60% sucrose  94  While sucrose gradient centrifugation was successful in separating the polymeric and dimeric form of ACC, this method was inconvenient for several reasons. One of the disadvantages of this method is that the separation between the polymeric and dimeric fractions is limited and subject to disruption with relatively subtle movements of the tube. As a result, there may still be some crossover in the protein composition between layers. More importantly, attempts to scale this method to a larger volume, and thus a larger separation between the two fractions, were unsuccessful. For these reasons, size exclusion chromatography was explored as an alternative method that might be suitable for small amounts of material and might give a more effective and complete separation of ACC polymers from small proteins. The use of size-exclusion chromatography to separate ACC dimers and polymers has been reported in previous studies using FPLC with Superose 6 size exclusion columns. This work had led to confirmation of citrateinduced polymerization with the corresponding large size shift as well as a depolymerizing effect of ACC phosphorylation [105]. Since those earlier studies had been carried out, a variety of size-exclusion materials have subsequently been developed with improved properties. Among several commercial size-exclusion resins available, BioGelA-50M was chosen because of its large exclusion limit and the fact that it is available in bulk loose form allowing for custom columns to be packed. To provide the necessary capacity, a 31 cm x 0.9 cm (height x diameter) column was packed with BioGelA-50M as outlined in Materials and Methods, and this column was calibrated with several pure standard proteins (figure 4.5). The fractionation range of BioGelA-50M is 100 kDa to 50,000 kDa and as seen in the calibration (figure 4.5) the thyroglobulin standard (approximately 660 kDa), which is a good marker for ACC dimers (approximately 530-580 kDa) is very well separated from Blue Dextran (approximately 2000 kDa) which is somewhat smaller than the size range of ACC polymers containing 10-20 dimers. As the BioGelA-50M resin had an unusually large predicted exclusion size, it was uncertain whether Blue Dextran would be suitable to define the exclusion volume of the column. To test this, the mitochondrial fraction was prepared from rat liver and used to further test the exclusion volume of the column. As can be seen, the mitochondrial preparation and Blue Dextran both began to elute from the column at the same volume (8 mL). In contrast, the thyroglobulin standard eluted  95  over the range of 13 to 20 mL with peak elution at approximately 16 mL. Finally, cytochrome C was used to determine the “included” column volume and this 12 kDa protein eluted between 19 to 29 mL, with a peak elution at about 22 mL. Based on these calibration studies, the “polymeric” forms of ACC were predicted to elute in the range of 8 to 12 mL, while fractions 16 to 21 should contain the “dimeric” forms of ACC.  96  Figure 4.5: Calibration of the BiogelA-50M size exclusion column. The indicated standards were run individually through the column (31 cm x 0.9 cm) at a flow rate of 20 mL/hour with a fraction size of 1 mL: Rat liver mitochondria (♦), Blue Dextran, 2000 kDa (■), thyroglobulin, 660 kDa (▲), and cytochrome C, 12.4 kDa (×). For blue dextran, thyroglobulin, and cytochrome C, the applied sample contained 0.5 mg in 0.5 mL. Mitochondria were prepared from rat liver and 0.2 mg was applied to the column. 1.2  Relative Absorbance  1 0.8 0.6 0.4 0.2 0 5  10  15  20  25  Fraction number (Eluted volume)  97  After size standard calibration of the column, ACC was purified from rat liver by ammonium sulfate precipitation, and then subjected to chromatography on the BioGelA50M column (a) at 4°C in the absence of citrate or (b) at 20°C in the presence of 20 mM citrate (figure 4.6). These two separate column runs were carried out as soon as possible, either the same day as the ACC preparation had been started or after overnight storage of the ammonium sulfate fraction on ice. In all cases, columns were run at 20mL/hour and fractions (1 mL) were assayed for total ACC activity (20 mM citrate). It was predicted that in the absence of citrate, there would be more ACC activity in the dimeric fractions eluting at approximately 16 mL while there would be more ACC activity in the polymeric fractions, eluting at approximately 8 mL after incubation with 20 mM citrate. Figure 4.6: Effect of citrate on elution of ACC during size-exclusion chromatography. ACC was partially purified from rat liver through ammonium sulfate precipitation and then subjected to a pre-clearing BioGelA-50M column in the absence of citrate. After pre-clearing, fractions 12-20 were concentrated and divided into two aliquots and incubated and subjected to BioGelA-50M chromatography in the absence (♦) or presence of 20 mM citrate (■). Fractions (1 mL) were assayed for ACC activity. The activity profile represents the results of one experiment which was repeated six times. The overall recovery of ACC activity in fractions 8 – 20 was typically greater than 90% of the activity applied to the column.  % Max ACC Activity  100  80  60 40  20  0 5  7  9  11  13  15  17  19  Fraction Number  98  Figure 4.6 shows the elution of ACC activity following chromatography using the BioGelA-50M size exclusion chromatography. As predicted, there is dramatic citrateinduced shift in ACC elution with much higher ACC activity in the dimeric fractions in the absence of citrate, and correspondingly much higher ACC activity in the polymeric fractions in the presence of citrate. This activity profile also demonstrates that the BioGelA-50M column was able to very convincingly separate the polymeric and dimeric forms of ACC. Figure 4.7: Effects of citrate on recovery of ACC in polymeric and dimeric forms. ACC was partially purified from rat liver through ammonium sulfate precipitation and then subjected to pre-clearing chromatography followed by two subsequent column fractionations in the absence (black bars) or presence of 20 mM citrate (red bars). For full details, see legend to figure 4.6 and Methods. Each value is significantly different from all other values P < 0.01, n = 6. 100 90  % ACC Activity  80 70 60 50 40 30 20 10 0 Polymeric  Dimeric Fraction  99  The experiment illustrated in figure 4.6 was repeated six times and the results summarized in figure 4.7. ACC activity was determined in all column fractions and then expressed as the percent of total eluted in polymeric fraction (8-12) and dimeric fraction (15-20). In these experiments, ACC prepared from rat liver through ammonium sulfate precipitation was pre-cleared and then chromatographed in the absence and presence of 20 mM citrate. The small error bars and dramatic difference in the ACC activity levels in each fraction in the absence and presence of citrate demonstrates that the BioGelA-50M column is an effective method to separate the dimeric and polymeric fractions of ACC. As noted above, a pre-clearing step was used prior to loading the extracts onto the BioGelA-50M size exclusion column, so a brief description of the effects of this step is given in figures 4.8. In the absence of a pre-clearing column, (figure 4.8) direct BioGelA-50M chromatography of rat liver ammonium sulfate fraction reveals that a significant amount of protein eluted in the “polymeric” fraction even in the absence of citrate. This shows that the non-cleared sample contains a significant quantity of protein that would migrate with ACC in a citrate-independent manner and thereby complicate further analysis. Following ammonium sulfate precipitation, it is very likely that some protein aggregates fail to fully re-solubilize and this would lead to observed high molecular weight mixture. Although this did not affect the ACC activity profile, these protein aggregates could cause a problem during the process of identifying putative ACCassociated proteins. The pre-clearing step was therefore necessary to remove these protein aggregates. The pre-clearing step was initially carried out using sucrose step gradient centrifugation, in the absence of citrate as discussed earlier, fractions 2 and 3 being pooled and concentrated. It was presumed that all protein aggregates would be excluded in fractions 5 and 6. This concentrated fraction was then subjected to BioGelA-50 M size exclusion chromatography. Subsequently, and to maintain consistency, the pre-clearing was carried out using the BioGelA-50M column. In this method of pre-clearing, the rat liver ammonium sulfate (0-40%) fraction was loaded onto the BioGelA-50M column and run in the absence of citrate. The fractions corresponding to the dimeric form of ACC (fractions 16 to 21) were pooled and concentrated. The efficacy of this column pre-  100  clearing step is demonstrated in figure 4.8 so that in the absence of citrate, the polymeric fraction contains very little protein.  101  Figure 4.8: Size exclusion chromatography of rat liver ACC preparations. (a) ACC was partially purified from rat liver through ammonium sulfate precipitation and then pre-incubated and subjected directly to chromatography on BioGelA-50M in the absence (⎯⎯⎯) and presence (⎯⎯⎯) of 20 mM citrate. (b) An additional preclearing step was employed in which ACC extracts were subjected to BioGelA-50M in the absence of citrate. Fractions containing the dimeric form of ACC (fractions 16 to 21) were pooled, concentrated and separated into two aliquots. Each aliquot was preincubated and subjected to chromatography for a second time on BioGelA-50M in the absence (⎯⎯⎯) or presence (⎯⎯⎯) of 20 mM citrate. The illustrated UV absorbance curve represents the results of one experiment which was repeated four times. a)  % Max Absorbance  100 80 60 40 20 0 5  10  15  20  25  20  25  Fraction Number  b)  % Max Absorbance  100 80 60 40 20 0 5  10  15  Fraction Number  102  4.3 Presence of associated proteins in the ACC polymeric fraction Western blot analysis of the BioGelA-50M column fractions using antibodies against ACC-1 (figure 4.9) showed a citrate-dependent shift of ACC from the dimeric size range (fractions 16 to 21) to polymeric size range (fractions 8 to 12). In the absence of citrate, ACC immunoreactive bands are found predominantly in the dimeric fractions, while in the presence of 20 mM citrate, ACC is found predominantly in the polymeric fractions confirming the results obtained by measuring the distribution of ACC activity. The results illustrated in the western blots were quantitated by scanning and the summary presented in figure 4.10. As with many antibody preparations, the anti-ACC antibody was not absolutely monospecific. Consequently, some bands in addition to the ACC 265 kDa subunit were detected. It is also possible that during the course of the experiment, the ACC protein subunits became partially degraded, leading to multiple bands in the film image. To confirm the results obtained with anti-ACC1 antibodies, streptavidin-HRP was also used to detect ACC and other biotinylated proteins (figure 4.11). Experiments were carried out exactly as described in the legend to figure 4.10 with the exception that final detection was based on binding of HRP-streptavidin rather than anti-ACC1 antibodies.  103  Figure 4.9: Citrate-dependent mobility shift of ACC revealed by size-exclusion chromatography. ACC was partially purified from rat liver by ammonium sulfate precipitation (0 – 40% saturation), pre-cleared and then subjected to size exclusion chromatography (BioGelA50M) at 4°C in the absence of citrate (a) or at room temperature in the presence of 20 mM citrate (b). Column fractions (1 mL) were collected and 250 μL of each fraction was concentrated through TCA precipitation and the proteins subjected to SDS-PAGE (acrylamide concentration of 4.5% in the stacking gel and 6% in the separating gel) followed by a transfer to a PVDF membrane. The ACC-1 band (arrow) was detected with anti-ACC and HRP-secondary antibodies (see Methods). Bars on the vertical axis represent the migration of molecular weight standards.  Polymers  Dimers  a) 9 10 11 12 15 16 17 18 19 20 21 22  220 kDa 97 kDa 65 kDa 45 kDa  b)  220 kDa 97 kDa 65 kDa 45 kDa  104  Figure 4.10: Citrate-dependent mobility size shift of ACC revealed by size-exclusion chromatography. The experiment described in figure 4.9 was repeated three times and resulting images were scanned and quantitated. The intensity of ACC staining in the polymeric fractions (8-12) and dimeric fractions (16-21) is expressed relative to total in each chromatogram. Pre-cleared rat liver ammonium sulfate fractions (0-40% saturation) were subjected to chromatography on BioGelA-50M either at 4°C in the absence of citrate (black bars) or at room temperature in the presence of 20 mM citrate (red bars) (n = 3). 100 90 % ACC protein  80 70 60 50 40 30 20 10 0 polymeric  dimeric Fraction  105  Figure 4.11: Citrate-dependent mobility shift in ACC revealed by size-exclusion chromatography. As in figure 4.9, pre-cleared ACC preparations were subjected to BioGelA-50M chromatography either at 4°C in the absence of citrate (a) or at room temperature in the presence of 20 mM citrate (b). Other experimental conditions are similar to the ones described in figure 4.9 with the exception that final detection was based on binding of HRP-streptavidin. The arrows indicate the ACC bands and the side bars indicate the migration of molecular weight standards. Polymers  Dimers  a) 8 9 10 11 12 15 16 17 18 19 20 21  220 kDa 97 kDa 65 kDa 45 kDa  b)  220 kDa 97 kDa 65 kDa 45 kDa  106  Figure 4.12: Citrate-dependent mobility size shift of ACC revealed by size-exclusion chromatography. The experiment described in figure 4.11 was repeated and resulting images were scanned and quantitated. The intensity of ACC staining in the polymeric fractions (8-12) and dimeric fractions (16-21) is expressed relative to total in each chromatogram. Pre-cleared rat liver ammonium sulfate fractions (0-40% saturation) were subjected to chromatography on BioGelA-50M either at 4°C in the absence of citrate (black bars) or at room temperature in the presence of 20 mM citrate (red bars). Results are the average of two independent experiments. 80 70 % ACC protein  60 50 40 30 20 10 0 polymeric  dimeric Fraction  107  The distribution of ACC protein was found to be very similar whether detection was based on the reaction with anti-ACC1 antibodies or with HRP-streptavidin (table 4.1). It should also be noted that the streptavidin-HRP detects both ACC isoforms as well as additional biotinylated proteins, while the anti-ACC1 antibody detects the ACC-1 isoform only. However, because ACC-1 is the predominant isoform expressed in hepatocytes (approximately 75% of total hepatic ACC), and from the results shown, the overall conclusion is supported by the use of both detection methods. Admittedly, there is a limitation of the maximum value obtained by using the Un-Scan-It software to quantitate the amount of protein found in each band. Therefore, relative values were obtained and an actual amount of protein could not be determined. Comparison of the various methods of assessing the citrate-dependent shift in the ACC elution during size exclusion chromatography provides an important perspective. Specifically, the measurement of ACC enzyme activity in column fractions reveals a much more dramatic citrate-induced mobility shift. Therefore, the increase in polymeric ACC enzyme activity of approximately 80% greatly exceeds the increase in ACC protein amounts by western blotting (approximately 30%). This demonstrates that polymeric ACC has a higher specific activity than the dimeric form of the enzyme, perhaps attributable to the presence of the other proteins in the polymeric fraction. Surprisingly, the analysis based on HRP-streptavidin also revealed an apparent size shift of pyruvate carboxylase (PC) in the presence of citrate. The reason for this apparent shift is not clear because PC is not known to oligomerize. It is also unlikely that PC and ACC could form a physiologically relevant functional protein complex because of their different cellular localizations, ACC being the only biotinylated carboxylase to be found in the cytosol [1] while PC is found exclusively in the mitochondrial matrix. Since PC is not known to polymerize upon incubation with citrate, there must be another explanation for the citrate induced size shift of PC especially as the presence of PC and “ACC polymer” fractions was observed quite consistently. Further experiments will be required to establish any direct or indirect link between ACC and PC.  108  Table 4.1: Distribution of ACC in different size ranges during size-exclusion chromatography. Data from experiments described in figures 4.6 to 4.12 were summarized for this composite table. Method of detection  Citrate  ACC activity (mU)  0 mM  9.0%  91.0%  20 mM  60.9%  39.1%  0 mM  21.5%  78.5%  20 mM  55.1%  44.9%  Streptavidin-HRP  0 mM  29.4%  70.6%  blot  20 mM  50.4%  49.6%  Anti-ACC1 blot  Polymeric ACC (fractions 8-12) as percent total  Dimeric ACC (fractions 16-21) as percent total  Having established the size-exclusion method to provide separation between large (polymeric) and small (dimeric) forms of ACC, I then set out to establish if other proteins migrated together with ACC in a citrate-dependent manner. To address this question, BioGelA-50 column fractions were subjected to a SDS-PAGE (4.5% stacking gel, 6% separating gel), followed by protein transfer to PVDF membranes and detection by staining with amido black to determine the protein composition of each column fraction (figure 4.13).  109  Figure 4.13: Protein composition of BioGelA-50M column fractions. Rat liver protein preparations recovered by ammonium sulfate precipitation (0 – 40% saturation) were pre-cleared by BioGelA-50M chromatography in the absence of citrate and the dimeric fractions (16-21) were concentrated and divided into two equal portions. These portions were again chromatographed through the BioGelA-50M column a) at 4°C in the absence of citrate or b) at room temperature in the presence of 20 mM citrate. The column fractions were subjected to SDS-PAGE (acrylamide concentration of 4.5% in the stacking gel and 6% in the separating gel), transferred onto a PVDF membrane and stained with amido black. Red boxes give an indication that several proteins may comigrate with ACC in a citrate-dependent manner. The letters A and B refer to proteins that will be discussed. Polymers Dimers a) 8 9 10 11 12 15 16 17 18 19 20 21 220 kDa  66 kDa  B A  b)  220 kDa  66 kDa  B A  110  Size exclusion chromatography was chosen because it is a relatively nondisruptive procedure that should allow preservation of protein complexes. As ACC is the only protein known to polymerize upon citrate treatment, the high molecular weight fractions collected after chromatography in the presence of citrate might ideally be expected to contain only ACC. However, this was not the case as the ACC polymer fractions also contained other proteins (figure 4.13). Since the pre-clearing step removed protein aggregates, as indicated by the UV absorbance profile, these proteins either spontaneously form larger aggregates in the presence or absence of citrate or in some way associate with ACC polymers. In the absence of citrate, no ACC is expected to migrate in the size range of polymers. However, this is not absolutely the case despite efforts to pre-clear the preparations prior to size-exclusion chromatography. Despite the fact that some proteins are still present may not necessarily mean that the pre-clearing step is ineffective. Since ACC polymers and dimers exist in equilibrium, it may be impossible to completely remove all polymeric form of ACC in vitro, even in the absence of citrate. Conversely, the presence of the putative ACC regulatory protein might influence the equilibrium of ACC in favor of the formation of polymers. Another possible explanation is that overnight storage of the partially purified ACC preparation on ice could lead to slight protein aggregation, although this is unlikely because centrifugation immediately prior to loading onto the column did not rectify the situation. It was accepted that the level of citrate-independent polymerization of ACC was tolerable considering the much larger citrate-induced phenomenon. The use of polymerization inhibitors, such as acyl-CoA esters might be contemplated in future experiments to further reduce the citrateindependent large forms of ACC. Accepting the modest limitations inherent with the low levels of ACC in the polymeric fraction in the absence of citrate, nevertheless, important new information is provided by these experiments. Specifically, several important observations can be made. First, in addition to ACC, several other proteins appear to show a citratedependent shift to higher molecular weight (figure 4.13, red boxes). Pyruvate carboxylase has already been mentioned and this latest approach revealed several other proteins in the 30 to 65 kDa range and perhaps other proteins not detectable at this level  111  with amido black stain. The second observation is that a number of proteins that are extremely abundant in the dimeric fraction size range (e.g., figure 4.13, A) show no evidence of migrating at higher molecular weight either in the absence or presence of citrate. These proteins therefore provide an important negative control indicating selectivity of the procedure. Finally, several proteins (e.g., figure 4.13, B) seem to show a distribution between large and small forms regardless of the presence or absence of citrate.  112  4.4 Analysis of ACC-associated proteins by mass spectrometry: in-solution digestions The work in the previous sections established techniques to separate polymeric from dimeric forms of ACC. Furthermore, these experiments established that several proteins appeared to co-migrate with acetyl-CoA carboxylase polymers in a citratesensitive manner. Based on these results, mass spectrometry was employed to identify the proteins that co-migrated with ACC polymers. Two general approaches were taken. In one approach, the ACC polymeric fractions were subjected directly (“in solution”) to tryptic digestion and then the peptides were analyzed by liquid chromatography coupled with tandem mass spectrometry (LC MS/MS). In a second approach, the proteins present in ACC polymeric fractions were first separated on SDS-PAGE gels and then specific stained bands excised, digested with trypsin and subjected to analysis by LC MS/MS. In addition to the two approaches to protein analysis, studies were also carried out using ACC preparations for rat liver and white adipose tissue to take advantage of the different ACC isoform expression patterns. While rat white adipose tissue appears to express exclusively ACC-1, both isoforms are expressed in rat liver, the hepatic ACC-1: ACC-2 ratio being approximately 3:1. The purpose of examining ACC from different tissues was to determine whether the presence or absence of ACC-2 would lead to a difference in co-migrating proteins. For example, ACC-1 is considered to be “cytosolic” whereas ACC-2 may be associated with mitochondria so that different partnering proteins are quite plausible. As anticipated, mass spectrometry led to the identification of ACC as a major component of the citrate-induced ACC polymeric fraction. This analysis was performed a total of five times; three times using preparations from rat liver, two times with preparations from white adipose tissue, and ACC was identified in four out of the five trials (figure 4.16). The fragmentation pattern for one of the identified ACC peptides is shown together with a table of all potential charged forms of that particular peptide (figure 4.17). The b-series of charged peptide fragments represent peptide fragmentation from the N-terminus of the peptide. The y-series of charged peptide fragments represent peptide fragmentation from the C-terminus of the peptide. Within these b- and y-series are various charged forms of that particular fragment.  113  Figure 4.14: Sequence of ACC-1. Rattus norvegicus, accession number: P11497, 2345 residues, predicted mass of 265, 194 Daltons [169], with peptides identified by LC MS/MS in multiple experiments shown in red. 1 61 121 181 241 301 361 421 481 541 601 661 721 781 841 901 961 1021 1081 1141 1201 1261 1321 1381 1441 1501 1561 1621 1681 1741 1801 1861 1921 1981 2041 2101 2161 2221 2281 2341  MDEPSPLAKT LGISALQDGL LIANNGIAAV ANNNNYANVE DKIASSIVAQ EEVGYPVMIK ILADQYGNAI TVEYLYSQDG VSPWGDAPID SVAAAGGLHE SFQLNRIDTG PAHTLLNTVD SYTTYMKEEV IEVMKMVMTL TALRGEKLHR IMTSVSGRIP FMNTQSIVQL NTVLNYIFSH RQVLIASHLP NQVVRMAALE FASNLNHYGM QSPTFPESGH HGIRRLTFLV ELNRMRNFDL LQNEGERLLL RLWKLRVLQA GDKQGPLHGM LPSPPLPSDI NDITYRIGSF EDPYKGYKYL MIAGESSLAY VYTSNNQLGG IIEFVPTKAP PVGVVAVETR VFANWRGFSG PRHMEMYADR KELESKLKER LLLEDLVKKK DGVRSVIEEN DSPST  LELNQHSRFI AFHMRSSMSG KCMRSIRRWS LILDIAKRIP TAGIPTLPWS ASEGGGGKGI SLFGRDCSVQ SFYFLELNPR FENSAHVPCP FADSQFGHCF WLDRLIAEKV VELIYEGIKY DRYRITIGNK TAVESGCIHY VFHYVLDNLV LNVEKSIKKE VQRYRSGIRG AQVTKKNLLV SYDVRHNQVE VYVRRAYIAY THVASVSDVL TSLYDEDKVP AQKDFRKQVN TAIPCANHKM EAMDELEVAF ELKINIRLTT LINTPYVTKD LTYTELVLDD GPQEDLLFLR YLTPQDYKRV DEIITISLVT IQIMHNNGVT YDPRWMLAGR TVELSVPADP GMKDMYDQVL ESRGSVLEPE EEFLIPIYHQ IHSANPELTD IKYISRDYVL  IGSVSEDNSE LHLVKQGRDR YEMFRNERAI VQAVWAGWGH GSGLRVDWQE RKVNNADDFP RRHQKIIEEA LQVEHPCTEM RGHVIAARIT SWGENREEAI QAERPDTMLG VLKVTRQSPN TCVFEKENDP VKRPGAALDP NVMNGYCLPD MAQYASNITS HMKAVVMDLL TMLIDQLCGR SIFLSAIDMY ELNSVQHRQL LDNAFTPPCQ RDEPIHILNV CEVDQRFHRE HLYLGAAKVE NNTNVRTDCN TGKAIPIRLF LLQSKRFQAQ QGQLVHMNRL ASELARAEGI SALNSVHCEH CRAIGIGAYL HCTVCDDFEG PHPTQKGQWL ANLDSEAKII KFGAYIVDGL GTVEIKFRKK VAVQFADLHD GQIQAMLRRW KQIRSLVQAN  DEISNLVKLD KKIDSQRDFT RFVVMVTPED ASENPKLPEL NDFSKRILNV NLFRQVQAEV PAAIATPAVF VADVNLPAAQ SENPDEGFKP SNMVVALKEL VVCGALHVAD SYVVIMNGSC SVMRSPSAGK GCVIAKMQLD PFFSSKVKDW VLCQFPSQQI RQYLRVETQF DPTLTDELLN GHQFCIENLQ KDNTCVVEFQ RMGGMVSFRT AIKTDGDIED FPKFFTFRAR VGTEVTDYRF HIFLNFVPTV LTNESGYYLD SLGTTYIYDI PGGNEIGMVA PRIYVAANSG VEDEGESRYK VRLGQRTIQV VFTVLHWLSY SGFFDYGSFS QQAGQVWFPD RECSQPVMVY DLVKTMRRVD TPGRMQEKGV FVEVEGTVKA PEVAMDSIVH  LEEKEGSLSP VASPAEFVTR LKANAEYIKM LLKNGIAFMG PQDLYEKGYV PGSPIFVMRL EHMEQCAVKL LQIAMGIPLF SSGTVQELNF SIRGDFRTTV VNLRNSISNF VEVDVHRLSD LIQYIVEDGG NPSKVQQAEL VERLMKTLRD ANILDSHAAT QNGHYDKCVF ILTELTQLSK KLILSETSIF FMLPTSHPNR FEDFVRIFDE DRLAAMFREF DKFEEDRIYR FVRAIIRHSD IMDPSKIEES ISLYKEVTDS PEMFRQSLIK WKMSLKSPEY ARIGLAEEIR ITDIIGKEEG ENSHLILTGA MPKNVHSSVP EIMQPWAQTV SAFKTYQAIK IPPQAELRGG PVYIRLAERL INDILDWKTS YVWDNNKDLV MTQHISPTQR  ASVSSDTLSD FGGNKVIEKV ADHYVPVPGG PPSQAMWALG KDVDDGLKAA AKQSRHLEVQ AKMVGYVSAG RIKDIRMMYG RSNKNVWGYF EYLIKLLETE LHSLERGQVL GGLLLSYDGS HVFAGQCYAE HTGSLPQIQS PSLPLLELQD LNRKSEREVF ALREENKSDM TTNAKVALRA DVLPNFFYHS GNIPTLNRMS VMGCFCDSPP TQQNKATLVE HLEPALAFQL LVTKEASFEY VRSMVMRYGS RTAQIMFQAY LWESMSTQAF PDGRDVIVIG HMFHVAWVDS LGAENLRGSG GALNKVLGRE LLNSKDPIDR VVGRARLGGI DFNREGLPLM SWVVIDPTIN GTPELSPTER RTFFYWRLRR EWLEKQLTEE AEVVRILSTM  114  Figure 4.15: Example of a fragment ion pattern of one peptide derived from ACC. The upper panel shows the mass spectrum of the indicated ACC peptide fragment ions with y-axis showing m/z and x-axis showing the intensity. Mass peaks are labeled according to fragmentation from the C-terminus (y-series) or the N-terminus (b-series). The lower panel shows all possible fragments of the indicated peptide and in red are highlighted the fragments that were detected.  b++  b*  b*++  b0  b0++  Seq  y  y++  y*  y*++  y0  y0++  #  b  1  114.09  57.55  2  171.11  86.06  3  258.14 129.58  4  405.21 203.11  387.20 194.11 F 1334.71 667.86 1317.68 659.35 1316.70 658.85 11  5  462.24 231.62  444.22 222.62 G 1187.64 594.33 1170.62 585.81 1169.63 585.32 10  6  559.29 280.15  541.28 271.14 P 1130.62 565.81 1113.59 557.30 1112.61 556.81 9  7  687.35 344.18 607.32 335.66 669.34 335.17 Q 1033.57 517.28 1016.54 508.77 1015.56 508.28 8  8  816.39 408.70 799.36 400.19 798.38 399.69 E  905.51 453.25 888.48 444.75 887.50 444.25 7  9  931.42 466.21 914.39 457.70 913.41 457.21 D  776.47 388.73 759.44 380.22 758.46 379.73 6  I  # 14  G 1478.76 739.89 1461.74 731.37 1460.75 730.88 13 240.13 120.57 S  1421.74 711.38 1404.72 702.86 1403.73 702.37 12  10 1044.50 522.75 1027.47 514.24 1026.49 513.75 L  661.44 331.22 644.41 322.71  5  11 1157.58 579.30 1140.56 570.78 1139.57 570.29 L  548.36 274.68 531.33 266.17  4  12 1304.65 652.83 1287.63 644.32 1286.64 643.82 F  435.27 218.14 418.25 209.63  3  13 1417.74 709.37 1400.71 700.86 1399.73 700.37 L  288.20 144.61 271.18 136.09  2  14  175.12  1  R  88.06 158.09 79.55  115  Confidence in the identification of ACC and other proteins is based particularly on two of the major analytical parameters, namely the number of peptides that could be assigned on the basis of exact mass matching to predicted ACC peptides and the subsequent mass fragmentation pattern that led to peptide sequencing. A number of proteins were identified with high confidence from the LC MS/MS analysis of peptides detected in high molecular weight BioGelA-50M fractions (ACC polymeric fractions 8 to 12) following chromatography in the presence of citrate. Overall confidence in the protein identification is expressed in terms of the mowse score and the number of peptides identified for each protein (see table 4.2). The mowse score refers to the probability that the identified protein is not a false positive. This probability is calculated as P = 10-(mowse score). The mowse score is calculated through an algorithm which compares the experimental peptide masses with the peptide masses listed in the database. In each analysis, a lower limit for the mowse score is calculated below which the identification of proteins is considered to be unreliable or not significant. This lower limit is calculated assuming that the probability of a random match is less than 5 percent of the highest scoring protein. An identified protein with a mowse score higher than the lower limit shows significance in the match.  For the in-  solution digestions, the lower threshold limits for the three experiments to analyze rat liver ACC polymers were 36, 28 and 32.  116  Table 4.2: Proteins identified from in-solution digestion of the polymeric fraction of rat liver ACC following size-exclusion chromatography. Rat liver protein samples were purified through ammonium sulfate precipitation (0 – 40% saturation), pre-cleared, and chromatographed in the presence of 20 mM citrate on the BioGelA-50 size exclusion column. The polymeric fractions (8-12) were pooled, concentrated, and digested with trypsin. The resulting peptides were analyzed and identified by LC MS/MS. Protein Acetyl-CoA carboxylase-1 Tubulin – alpha Tubulin – beta Actin – beta Formyltetrahydrofolate dehydrogenase Fatty acid synthase Apolipoprotein E precursor Valosin-containing protein Clathrin, heavy polypeptide Methionine adenosyltransferase Heat shock protein 2 Cystathionine beta-synthase Alcohol dehydrogenase Fructose-1,6-bisphosphatase Heat shock cognate protein 70 Nucleoside-diphosphate kinase  1 209 161 109 42 349  Mowse score 2 3 109 53 204 982 236 1164 221  244  # Peptides identified 1 2 3 11 6 1 7 9 19 5 12 26 2 5 18 17  193 185 181  4 6 9  140 91 55  3 2 5  49  1  44  2  42  2 42  1  Table 4.2 is organized according to identification in 3, 2, or 1 of the three replicate experiments. Only three proteins were identified in all these trials: ACC-1 and two tubulin isoforms. Actin beta subunits were identified in two of the trials and the remaining proteins were identified in only one experiment. In terms of developing strategies to understand the significance of the apparent association of proteins with ACC, the top priority must be given to those proteins most consistently identified in the high molecular weight/polymeric fractions and with the highest confidence (mowse score). On this basis, tubulin and actin stood out as the most significant from this analysis of rat liver preparations. The functions of the other proteins present in the ACC polymeric fractions are summarized in table 4.3.  117  Table 4.3: Functions of the proteins identified by LC MS/MS analyses of rat liver ACC polymeric fractions. Protein Function Tubulin – alpha Polymerization of tubulin subunits to form microtubules, Tubulin – beta essential for cytoskeleton maintenance, cell division, and vesicle transport [170]. Actin – beta Essential for cytoskeleton maintenance, cell motility, and involved in muscle contraction in animals [171]. Formyltetrahydrofolate An oxidoreductase which converts formyltetrahydrofolate and dehydrogenase NADP+ to form tetrahydrofolate and NADPH. This enzyme is important for replenishing tetrahydrofolate levels [172]. Fatty acid synthase Formation of palmitate from acetyl-CoA and malonyl-CoA (FASN) [173]. Apolipoprotein E ApoE can be found in chylomicrons, VLDL, and HDL. precursor Specific isoforms of apoE have been linked to an increased risk for Alzheimer’s disease [174]. Valosin-containing Involved in vesicle transport in cells, VCP forms a complex protein with clathrin and heat shock cognate protein 70 (Hsc70) [175]. Clathrin, heavy Involved in the formation of vesicles [176]. polypeptide Methionine Catalyzes the formation of S-adenosylmethionine from adenosyltransferase methionine and ATP Heat shock protein 2 Related to Hsc70, specific for rat testis [177]. Cystathionine betaA key enzyme in sulfur metabolism, catalyzing the formation synthase of cystathionine from homocysteine and serine [178]. Alcohol dehydrogenase Catalyzes conversion of ethanol to acetyaldehyde [179]. Fructose-1,6A key enzyme in gluconeogensis that catalyzes the hydrolysis bisphosphatase of fructose-1,6-bisphosphate to fructose-6-phosphate [180]. Heat shock cognate Classified as a heat shock protein by homology, Hsc70 is protein 70 (Hsc70) constitutively expressed, and is involved in folding and trafficking of newly synthesized proteins [181]. Nucleoside-diphosphate Phosphorylates nucleoside diphosphates to the corresponding kinase triphosphates [182]. NDK has been found to associate with microtubules [183].  118  From a functional perspective, the identification of FASN and the heat shock proteins (HSP2 and Hsc70) were of most interest. As ACC and FASN catalyze consecutive steps in the de novo biosynthesis of long chain fatty acids, it is plausible for the two proteins to form a complex or “metabolon” as mentioned previously in the introduction. The heat shock protein 2 and hsc70 are involved in the folding and trafficking or proteins. ACC is a large and complex protein that would most likely need a chaperonetype protein to ensure proper folding during biosynthesis and importantly, the polymerization and depolymerization of ACC might involve molecular chaperones. The identification of tubulin and actin were among the most consistent of the findings of the analysis. In one sense, this presents a considerable challenge because tubulin and actin are both abundant cellular proteins and their apparent association with ACC might be an artifact. Nevertheless, this observation is intriguing in light of previous studies of others who have argued for such a functional association. Certainly as tubulin and actin are important cytoskeleton elements, an association between these cytoskeletal elements and ACC may be important for ACC localization, or even to assist in ACC polymerization. The identification of clathrin, VCP and hsc70 is also of interest because it has previously been demonstrated that these proteins can form a complex [175]. Although all three of these proteins were identified in the polymeric fraction it is possible that only one of these proteins is directly associated with ACC. Another hint that ACC might be one of several proteins that interact indirectly is NDK, which is known to associate with microtubules and might therefore associate with ACC because of the presence of tubulin. The analysis of rat liver preparations was extended by carrying out parallel analyses of ACC preparations from rat white adipose tissue. A comparison of the different tissues is interesting for several reasons. First, white adipose tissue contains little, if any, ACC-2 and therefore provides an opportunity to explore specifically the proteins that associate with ACC-1 polymers. Secondly, the overall protein expression profile of adipose tissue is much different from that of liver, consistent with the major role in metabolism. ACC therefore represents a much higher proportion of overall protein in adipose tissue than in liver, enhancing the potential “signal to noise” ratio of  119  protein identification. Another advantage, looking ahead, is that adipose tissue can be readily exposed to hormones in vitro, potentially facilitating studies of effects on ACCprotein interactions. Rat white adipose tissue was isolated and used for the same citrate-induced preparation of ACC polymers as described above for rat liver and in the methods. Proteins identified in the adipose tissue ACC polymers by LC MS/MS are listed in table 4.4. Table 4.4: Proteins identified by in-solution digestions and LC MS/MS of the polymeric fractions isolated from rat white adipose tissue. Protein samples were purified through ammonium sulfate precipitation (0 – 40% saturation), pre-cleared, and chromatographed in the presence of 20 mM citrate on the BioGelA-50 size exclusion column. The polymeric fractions (8-12) were pooled, concentrated, and subjected to tryptic digestion. The resulting peptides were analyzed and identified through LC MS/MS. Protein Acetyl-CoA carboxylase Fatty acid synthase Tubulin – alpha Tubulin – beta Actin – alpha Actin – beta Pyruvate carboxylase Vimentin Pyruvate dehydrogenase Glyceraldehyde-3-phosphate dehydrogenase Apolipoprotein A-I  Mowse score 1 2 187 1068 919 47 392 810 192 301 172 621 183 131 453 210 167 73  # Peptides identified 1 2 5 42 18 4 10 14 7 6 6 11 10 3 10 3 3 1  The lower threshold limits of the mowse scores in trials one and two were 35 and 33, respectively. In the WAT analyses, ACC-1 shows up in only one of the two trials. FASN, tubulin, actin and pyruvate carboxylase were all detected with high confidence in both trials. The functions of the proteins identified in the citrate-induced ACC polymer function are listed in table 4.5.  120  Table 4.5: Functions of the proteins identified by LC MS/MS analysis of white adipose tissue ACC polymeric fractions. Protein Pyruvate carboxylase Vimentin Pyruvate dehydrogenase Glyceraldehyde-3phosphate dehydrogenase Apolipoprotein A-I  Function Catalyzes the formation of oxaloacetate from pyruvate. An intermediate filament protein specific for loose connective tissue and smooth muscle. Catalyzes the oxidative decarboxylation of pyruvate to form acetyl-CoA [184]. Catalyzes the glycolytic conversion of glyceraldehyde-3phosphate to glycerate 1,3-bisphosphate. Also has roles in control of gene expression [185] with links to ovarian cancer ApoAI is an important component of HDL particles and therefore reverse cholesterol transport [186]. In particular, ApoAI is involved in the formation of cholesterol esters.  In contrast to the analysis of liver ACC preparations, very few proteins were confidently identified in the ACC polymer fraction from white adipose tissue. Significantly, FASN, tubulin and actin were again detected as they were in liver. The presence of pyruvate carboxylase, while apparent in both experiments, is difficult to explain considering its intramitochondrial location. From this list, PC and PDH are unlikely to be direct ACC binding partners because they are mitochondrial proteins, while ACC is a cytosolic protein.  121  4.5 LC MS/MS analysis of ACC-associated proteins following SDS-PAGE and in-gel digestion In the first series of analyses reported in section 4.4, fractions were trypsinized and then subjected to LC MS/MS analysis. To complement these studies with an alternative approach, the ACC polymeric fractions were recovered using the same sizeexclusion chromatography method and then the column fractions were subjected to SDSPAGE so that specific bands could be excised and treated with trypsin prior to peptide LC MS/MS. Accordingly, the high-speed supernatant fraction from rat epididymal adipose tissue was treated with ammonium sulfate (0-40% saturation), the re-dissolved proteins were pre-cleared by size-exclusion chromatography and then ACC polymeric fractions were prepared using size exclusion chromatography in the absence and presence of citrate. The proteins recovered in the polymeric fractions in the absence and presence of citrate were separated by SDS-PAGE (figure 4.16). The SDS-PAGE analysis confirmed that chromatography in the presence of citrate led to enrichment of the high molecular weight band containing ACC and also of several other bands stained with Simply Blue stain. With unlimited resources, it would be ideal to divide each gel lane into many slices and subject each slice to trypsin digestion and analyze by LC MS/MS. However, this was not economically feasible, so we initially decided to focus on the four most prominent protein bands enriched in the ACC polymer fractions obtained in the presence of citrate (figure 4.16)  122  Figure 4.16: The protein composition of adipose tissue ACC polymeric fractions in the absence and presence of citrate. ACC was recovered from rat white adipose tissue by centrifugation and ammonium sulfate precipitation (40% saturation) as described in Materials and Methods. Proteins were re-solubilized and “pre-cleared” by centrifugation prior to size-exclusion chromatography (BioGelA-50M). The pre-cleared ACC sample was divided into two aliquots which were incubated in the absence (-) or presence (+) of 20 mM citrate. Column fractions containing the polymeric form of ACC were eluted, pooled and loaded on a SDS-PAGE gel (acrylamide concentration of 5% in the stacking gel and 8% in the separating gel). The same volume of protein sample was loaded onto the gel but the noncitrate polymeric fraction contained 15 μg protein, while the plus citrate polymeric fraction contained 40 μg protein. The two lanes in each condition are replicates. Boxes indicate bands that were excised, trypsinized, and subsequently analyzed through LC MS/MS. - citrate 220 kDa  + citrate A B  97 kDa 66 kDa 45 kDa 30 kDa  C D  20.1 kDa  123  Following SDS-PAGE, four protein-stained bands appeared to be particularly enriched in polymeric fractions obtained in the presence of citrate (bands A-D). The highest molecular weight band (band A) migrated with an approximate molecular weight slightly higher than the 220 KDa standard and the presence of ACC-1 was confirmed by identification of 30 ACC-1 peptides and an overall mowse score of 1268 (table 4.6). As well as ACC-1, the high molecular weight band A also contained fatty acid synthase which was detected with an even more impressive number of peptides (71) and mowse score (2999). This is consistent with the known subunit size of FASN (approximately 270 kDa) and confirms the detection of FASN found by the previous analysis using insolution digestion. The FASN peptides that were detected by LC MS/MS are shown in the context of the full protein sequence (figure 4.17). The identification of each of these FASN peptides was based on the appearance of multiple fragment ions detected by LC MS/MS. This is illustrated in the case of one of the FASN peptides (figure 4.18) for which 19 fragment ions were detected that matched to within 1 mU of predicted m/z. This peptide sequence, together with that of 70 other FASN peptides detected, spanned a total of 752 amino acid residues (30% of the FASN sequence). In band B, only the pyruvate carboxylase was identified, while in band C, tubulin and vimentin were identified, consistent with their subunit molecular weights. The presence of tubulin also confirms its detection in the in-solution digestions. The sequence coverage and the fragmentation sequence of one of the identified peptides for tubulin are shown in figures 4.19 and 4.20. In band D, actin, GADPH, and PDH were identified.  124  Table 4.6: Proteins from rat white adipose tissue that were detected in ACC polymers by MS/MS analysis of excised gel bands. Protein samples were purified through ammonium sulfate precipitation (0 – 40% saturation), pre-cleared, and chromatographed in the presence of 20 mM citrate by BioGelA-50 size exclusion chromatography. The polymeric fractions (8-12) were pooled, concentrated, and subjected to SDS-PAGE. Citrate-enriched protein bands were excised and digested with trypsin and the resulting peptides were analyzed and identified by LC MS/MS. Bands are labeled as listed in figure 4.15. The lower threshold limit for the mowse score was 35.  Band  Protein  A  Fatty Acid Synthase ACC-1 Pyruvate carboxylase Tubulin Vimentin Actin Glycerylaldehyde-3-phosphate dehydrogenase Pyruvate dehydrogenase  B C D  Mowse score 2999 1268 1555 1150 552 716 539 429  # identified peptides 71 30 31 30 11 23 10 9  125  Figure 4.17: Sequence of fatty acid synthase. Rattus norvegicus, accession number P12785, 2505 amino acid residues, estimated mass 272, 650 Daltons [187] and the peptides identified by LC MS/MS shown in blue. 1 61 121 181 241 301 361 421 481 541 601 661 721 781 841 901 961 1021 1081 1141 1201 1261 1321 1381 1441 1501 1561 1621 1681 1741 1801 1861 1921 1981 2041 2101 2161 2221 2281 2341 2401 2461  MEEVVIAGMS DASFFGVHPK SRDPETLLGY PAAIVGGINL RVYATILNAG DPQELNGITR NPNPEIPALL PHLLHASGRT EVQQVPASQR TDEHTFDDIV LAAYWRGQCI FVEQLKQEGV QWQSSLARTS IIPLMKRDHK TWDIPVAEDF RSLSLSLEET WEDPDSKLFD LLWKDNWVTF TSRCLGVTVS LCKGLAKALQ RLLLPEDPLI LQLEYTATDR ALALDNMVAA HLVGLKKSFY CPTSGVVGLV DGAWGAFRHF YYASLNFRDI LLSPDFLWDV IALSLGCRVF SLAEEKLQAS WREVAELLKA GAQPTLISAI KHVREWRRQG LFQDVNKPKY QRRHDGLPGL SSFVLVEKKA LEREHDLVLP EGPTLTRLNS AYYIDCIKQV YTQSYRAKLT SHQSLDRRDL VCDGKVSVHI  GKLPESENLQ QAHTMDPQLR SMVGCQRAMM LLKPNTSVQF TNTDGCKEQG SLCAFRQSPL DGRLQVVDRP MEAVQGLLEQ PLWFICSGMG HSFVSLTAIQ KDANLPAGSM FAKEVRTGGL SAEYNVNNLV DNLEFFLTNL PNGSSSSSAT PVVFENVTFH HPEVPIPAES MDTMLQISIL GGVYISRLQT TKATQQGLKM SGLLNSQALK HPQALKDVQT LKDGGFLLMH GTALFLCRRL NCLRKEPGGH QLEQDKPEEQ MLATGKLSPD PSSWTLEEAA TTVGSAEKRA VRCLAQHGRF GIRDGVVKPL SKTFCPEHKS IHVLVSTSNV NGTLNLDRAT AVQWGAIGDV VAHGDGEAQR IREVRQLTLR VQSSERPLFL QPEGPHRVAG PGCEAEAEAE SFAAVSFYYK IEGDHRTLLE  EFWANLIGGV LLLEVSYEAI ANRLSFFFDF MKLGMLSPDG VTFPSGEAQE LIGSTKSNMG LPVRGGIVGI GRQHSQDLAF TQWRGMGLSL IALIDLLTSM AAVGLSWEEC AFHSYFMEGI SPVLFQEALW GKVHLTGIDI VYNIDASSES QATILPRTGT ESVSRLTQGE GFSKQSLQLP TATSRRQQEQ TVPGLEDLPQ ACIDTALENL KLQQHDVAQG TVLKGHALGE SPQDKPIFLP RIRCILLSNL TAHAFVNVLT AIPGKWASRD SVPVVYTTAY YLQARFPQLD LEIGKFDLSN KCTVFPKAQV YIITGGLGGF SSLEGARALI REACPELDYF GIILEAMGTN DLVKAVAHIL KLQEMSSKAG VHPIEGSITV YSFGACVAFE AICFFIKQFV LRAADQYKPK GRGLESIINI  DMVTDDDRRW VDGGINPASL KGPSIALDTA TCRSFDDSGN QLIRSLYQPG HPEPASGLAA NSFGFGGANV VSMLNDIAAT MRLDSFRESI GLKPDGIIGH KQRCPPGVVP APTLLQALKK HVPEHAVVLE NPNALFPPVE SDHYLVDHCI VPLEVRLLEA VYKELRLRGY TRVTAIYIDP LVPTLEKFVF HGLPRLLAAA STLKMKVVEV QWDPSGPAPT TLACLPSEVQ VEDTSFQWVD SSTSHVPKLD RGDLASIRWV CMLGMEFSGR YSLVVRGRIQ DTSFANSRDT NHPLGMAIFL EDAFRYMAQG GLELARWLVL AEATKLGPVG VAFSSVSCGR DTVVGGTLPQ GIRDLAGINL SDTELAAPKS FHSLAAKLSV MCSQLQAQQG DAEHSKVLEA AKYHGNVILL IHSSLAEPRV  KAGLYGLPKR RGTNTGVWVG CSSSLLALQN GYCRAEAVVA GVAPESLEYI LTKVLLSLEN HVILQPNTQQ PTAAMPFRGY LRSDEALKPL SLGEVACGYA ACHNSEDTVT VIREPRPRSA IAPHALLQAV FPVPRGTPLI DGRVLFPGTG SHAFEVSDSG DYGPHFQGVY ATHLQKVYML TPHVEPECLS CQLQLNGNLQ LAGEGHLYSH NLGALDLVVC PGPSFLSQEE SLKSILATSS PGSSELQKVL SSPLKHMQPP DKCGRRVMGL HGETVLIHSG SFEQHVLLHT KNVTFHGILL KHIGKVLVQV RGAQRLVLTS GVFNLAMVLR GNAGQSNYGF RISSCMEVLD DSSLADLGLD KNDTSLKQAQ PTYGLQCTQA PAPAHNNLFL LLPLKSLEDR RAKTGGTYGE SVREG  SGKLKDLSKF VSGSEASEAL AYQAIRSGEC VLLTKKSLAR EAHGTGTKVG GVWAPNLHFH APAPAPHAAL TVLGVEGHVQ GVKVSDLLLS DGCLSQREAV ISGPQAAVNE RWLSTSIPEA LKRGVKPSCT SPHIKWDHSQ YLYLVWKTLA NLIVSGKVYQ EATLEGEQGK EGDTQVADVT ESAILQKELQ LELGEVLARE ISALLNTQPM NCALATLGDP WESLFSRKAL SQPVWLTAMN ESDLVMNVYR SSSGAQLCTV VPAEGLATSV SGGVGQAAIS GGKGVDLVLN DALFEGANDS REEEPEAMLP RSGIRTGYQA DAMLENQTPE ANSTMERICE LFLNQPHAVL SLMGVEVRQI LNLSILLVNP APLDSIPNLA FDGSHTYVLA VAAAVDLITR DLGADYNLSQ  126  Figure 4.18: Example of fragment ion pattern of one peptide derived from fatty acid synthase. The upper panel shows the mass spectrum of the indicated FASN peptide with y-axis showing m/z and x-axis showing the intensity. Mass peaks are labeled according to fragmentation from the C-terminus (y-series) or N-terminus (b-series). The lower panel shows all possible fragment ions of the indicated peptide with the detected ions highlighted in red.  a) b) #  b  b++  b*  b*++  b0  b0++  Seq  y  y++  y*  y*++  98.02 49.52 D  y0  y0++  #  1  116.03 58.5207  12  2  230.08 115.54 213.05 107.03 212.07 106.54 N 1295.70 648.35 1278.67 639.84 1277.69 639.35 11  3  343.16 172.08 326.13 163.57 325.15 163.08 L 1181.66 591.33 1164.63 582.82 1163.65 582.33 10  4  472.21 236.61 455.17 228.09 454.19 227.60 E 1068.57 534.79 1051.55 526.28 1050.56 525.78 9  5  619.27 310.14 602.25 301.63 601.26 301.13 F  939.53 470.27 922.50 461.76 921.52 461.26 8  6  766.34 383.67 749.31 375.16 748.33 374.67 F  792.46 396.73 775.43 388.22 774.45 387.73 7  7  879.42 440.22 862.40 431.70 861.41 431.21 L  645.39 323.20 628.37 314.69 627.38 314.20 6  8  980.47 490.74 963.45 482.23 962.46 481.73 T  532.31 266.66 515.28 258.14 514.30 257.65 5  9 1094.52 547.76 1077.49 539.25 1076.50 538.76 N  431.26 216.13 414.23 207.62  4  10 1207.60 604.30 1190.57 595.79 1189.59 595.29 L  317.22 159.11 300.19 150.60  3  11 1264.62 632.81 1247.60 624.30 1246.61 623.81 G  204.13 102.57 187.11 94.06  2  12  147.11  1  K  74.06 130.09 65.55  127  Figure 4.19: Sequences of rat tubulin isoforms. Rattus norvegicus and peptides identified by LC MS/MS with peptides identified from LC MS/MS shown in green. a) tubulin, alpha 4 (accession number Q5XIF6, 448 amino acid residues, molecular weight 49, 924 Daltons) [188] 1 61 121 181 241 301 361 421  MRECISVHVG HVPRAVFVDL RIRKLSDQCT VVEPYNSILT SLRFDGALNV QMVKCDPRHG TVVPGGDLAK AREDMAALEK  QAGVQMGNAC EPTVIDEIRN GLQGFLVFHS THTTLEHSDC DLTEFQTNLV KYMACCLLYR VQRAVCMLSN DYEEVGIDSY  WELYCLEHGI GPYRQLFHPE FGGGTGSGFT AFMVDNEAIY PYPRIHFPLA GDVVPKDVNA TTAIAEAWAR EDEDEGEE  QPDGQMPSDK QLITGKEDAA SLLMERLSVD DICRRNLDIE TYAPVISAEK AIAAIKTKRS LDHKFDLMYA  TIGGGDDSFT NNYARGHYTI YGKKSKLEFS RPTYTNLNRL AYHEQLSVAE IQFVDWCPTG KRAFVHWYVG  TFFCETGAGK GKEIIDPVLD IYPAPQVSTA ISQIVSSITA ITNACFEPAN FKVGINYQPP EGMEEGEFSE  b) tubulin, beta 5 (accession number P69897, 444 amino acid residues, molecular weight 49,671 Daltons) [170] 1 61 121 181 241 301 361 421  MREIVHIQAG PRAILVDLEP RKEAESCDCL EPYNATLSVH RFPGQLNADL AACDPRHGRY LKMAVTFIGN EYQQYQDATA  QCGNQIGAKF GTMDSVRSGP QGFQLTHSLG QLVENTDETY RKLAVNMVPF LTVAAVFRGR STAIQELFKR EEEEDFGEEA  WEVISDEHGI FGQIFRPDNF GGTGSGMGTL CIDNEALYDI PRLHFFMPGF MSMKEVDEQM ISEQFTAMFR EEEA  DPTGTYHGDS VFGQSGAGNN LISKIREEYP CFRTLKLTTP APLTSRGSQQ LNVQNKNSSY RKAFLHWYTG  DLQLDRISVY WAKGHYTEGA DRIMNTFSVV TYGDLNHLVS YRALTVPELT FVEWIPNNVK EGMDEMEFTE  YNEATGGKYV ELVDSVLDVV PSPKVSDTVV ATMSGVTTCL QQVFDAKNMM TAVCDIPPRG AESNMNDLVS  c) tubulin, beta 2C (accession number Q6P8T8, 445 amino acid residues, molecular weight 49,801 Daltons) [189] 1 61 121 181 241 301 361 421  MREIVHLQAG PRAVLVDLEP RKEAESCDCL EPYNATLSVH RFPGQLNADL AACDPRHGRY LKMSATFIGN EYQQYQDATA  QCGNQIGAKF GTMDSVRSGP QGFQLTHSLG QLVENTDETY RKLAVNMVPF LTVAAVFRGR STAIQELFKR EEEGEFEEEA  WEVISDEHGI FGQIFRPDNF GGTGSGMGTL CIDNEALYDI PRLHFFMPGF MSMKEVDEQM ISEQFTAMFR EEEVA  DPTGTYHGDS VFGQSGAGNN LISKIREEYP CFRTLKLTTP APLTSRGSQQ LNVQNKNSSY RKAFLHWYTG  DLQLERINVY WAKGHYTEGA DRIMNTFSVV TYGDLNHLVS YRALTVPELT FVEWIPNNVK EGMDEMEFTE  YNEATGGKYV ELVDSVLDVV PSPKVSDTVV ATMSGVTACL QQMFDAKNMM TAVCDIPPRG AESNMNDLVS  128  Figure 4.20: Example of fragment ion pattern of one peptide derived from tubulin. The upper panel shows the mass spectrum of the indicated tubulin beta 2C peptide with y-axis showing m/z and x-axis showing the intensity. Mass peaks are labeled according to fragmentation from the C-terminus (y-series) or N-terminus (b-series). The lower panel shows all possible fragments of the indicated peptide, with the detected ions highlighted in red. a)  b) #  b  b++  b*  b*++  b0  b0++  Seq  y  y++  y*  y*++  y0  y0++  #  1  148.08  74.54  F  10  2  245.13 123.07  P  983.53 492.27 966.50 483.75 965.52 483.26 9  3  302.15 151.58  G  886.47 443.74 869.45 435.23 868.46 434.74 8  4  430.21 215.61 413.18 207.09  Q  829.45 415.23 812.43 406.72 811.44 406.22 7  5  543.30 272.15 526.26 263.64  L  701.39 351.20 684.37 342.69 683.38 342.20 6  6  657.34 329.17 640.31 320.66  N  588.31 294.66 571.28 286.15 570.30 285.65 5  7  728.37 364.69 711.35 356.18  A  474.27 237.64 457.24 229.12 456.26 228.63 4  8  843.40 422.20 826.37 413.69 825.39 413.20 D  403.23 202.12 386.20 193.61 385.22 193.11 3  9  956.48 478.74 939.46 470.23 938.47 469.74 L  288.20 144.61 271.18 136.10  2  10  R  175.12  1  88.06 158.09 79.55  129  In summarizing the results generated from the in-solution LC MS/MS analysis of liver and WAT, and the in-gel LC MS/MS, several important conclusions emerge. In addition to ACC itself, one of the major proteins consistently detected in ACC polymers was fatty acid synthase, the 272 kDa enzyme involved in the sequential seven step synthesis of palmitate (C:16) [173]. The complex process is NADPH-dependent and also requires ATP, acetyl-CoA and malonyl-CoA. Considering the sequential metabolic roles of ACC and FASN, the detection of FASN in ACC polymers may have functional significance. Notably, close association of these proteins would provide support for the concept of a lipogenic metabolon that would facilitate substrate supply to FASN. ACC is generally considered to be the “rate-limiting step” for lipid synthesis and changes in ACC activity generally correlate closely with changes in overall rates of fatty acid synthesis. For example, incubation of rat hepatocytes or adipocytes with insulin leads to parallel activation of ACC activity and increased rates of fatty acid synthesis [190]. Conversely, incubation of fat cells with adrenaline or other β-adrenergic agonists or corresponding treatment of hepatocytes with glucagon lead to ACC inhibition and suppression of de novo lipid synthesis. Similarly, small-molecule inhibitors of ACC lead to changes in rates of fatty acid synthesis and longer-term dietary changes in rates of lipid synthesis are also reflected by parallel changes in ACC. Studies in which ACC expression levels are manipulated indicate that FASN expression may also be coordinately affected. This occurs during fasting/starving cycles and also during transitions between low-fat and high-fat diets. Furthermore, the expression of a ribozyme that specifically cleaves ACC mRNA in a preadipocyte line (#0A5) showed not only that ACC gene expression was inhibited [135], but that FASN activity was also decreased 30-70% compared to the controls. Although co-expression studies do not provide evidence that these two proteins actually physically associate, these studies do show how intimately their activities are linked. Overall, the studies of ACC polymerization reported here provide support for the possibility that ACC and FASN might be closely associated. Indeed the concept that ACC and FASN associate is not new. For example, Gregolin et al suggested many years ago that ACC filaments could provide a basic structure onto which FASN molecules may be able to bind [86]. In addition, both ACC and FASN, along with other lipogenic  130  enzymes (such as malic enzyme and glucose-6-phosphate dehydrogenase) are induced coordinately during fat cell development (adipogenesis). Another major conclusion to emerge from overall consideration of proteins detected in ACC polymer fractions is the presence of cytoskeletal protein tubulin, actin and perhaps vimentin. The two main forms of tubulin, α and β, each have a subunit size of about 50 kDa [103] and they form heterodimers which can polymerize into microtubules. Microtubules are vital in many cellular processes, including cell shape maintenance, vesicle transport and cell division. Microtubule formation requires GTP and is a dynamic and reversible process which is analogous to ACC polymerization and depolymerization. There are a number of microtubule-associated proteins (MAPs) including some involved in the assembly and disassembly of microtubules, as well as those mediating the binding between tubulin and other proteins. Considering the control of ACC by insulin, it is intriguing to note that MAP-2 is phosphorylated in response to insulin, although the function of this is not known. Tubulin itself also undergoes a number of post-translational modifications and this is compounded by the fact that tubulin exists in various isoforms, levels of expression of each isoform being tissue specific. Actin was also identified as a component of ACC polymer fractions in several trials from both liver and white adipose tissue. Actin is an essential protein needed for cytoskeleton maintenance and cell motility [171]. It is found in all eukaryotes and actinlike proteins are also found in bacteria. Actin monomers have a mass of 43 kDa, and spontaneously polymerize in vitro in solutions of low ionic strength (50-100 mM). A major concern about these observations is that cytoskeletal proteins are generally very abundant and might therefore be detected purely because they co-migrate with ACC non-specifically. While this remains a possibility, the case for tubulin nevertheless merits further consideration for several reasons. First, further analysis has shown that tubulin not only associates with ACC polymer fractions but also is detected in association with ACC following avidin affinity chromatography in the presence of 250500 mM KCl. Clearly, the latter association is observed under very stringent conditions. Furthermore, there is some independent evidence for ACC-tubulin interactions although the limited literature is not consistent. For example, in one study, Buechler et al  131  investigated the effects on ACC of the microtubule effectors GTP and colchicine. GTP is required for tubulin polymerization [191], binding reversibly to the β subunit of tubulin and subsequently being hydrolyzed to GDP and Pi. These studies found that GTP led to ACC activation as well as tubulin polymerization. The other compound tested was colchicine, which binds to tubulin dimers, leading to a conformational change in tubulin, preventing further polymerization without promoting microtubule disassembly [191]. Buechler et al found that stabilization of microtubules with colchicine was associated with inhibition of ACC. In the use of colchicine, no evidence was found that this agent has any direct effect on ACC so the inhibitory effect on ACC observed by Buechler et al is presumably indirect and modified by changes in microtubules or tubulin. The case for GTP is less clear-cut, however, and at least one group reported direct effects of GTP on ACC. For example, Mick et al demonstrated that GTP analogs inhibited ACC activity [192] and that radiolabeled GTP (α32P-GTP) was able to bind to ACC, although not to the related pyruvate carboxylase. Furthermore, Mick et al suggest a structural homology between a portion of the ACC-1 sequence and the GTP binding domain on β-tubulin. The probability that this homology might be explained by chance was calculated to be less than 0.1 percent. The studies of Mick et al provide some support for the earlier evidence that GTP directly affects ACC activity but in light of our evidence, these earlier ACC preparations might well have contained tubulin. Overall, then in light of the confusing results of studies by Buechler and others, it is obviously important to establish if GTP and colchicine act directly on ACC. So far, no link has been found that would suggest a functional role for association between actin and ACC. Furthermore, considering that actin can spontaneously polymerize in buffers of relatively low ionic strength, it is possible that this might explain the co-elution of ACC and actin following size-exclusion chromatography. Due to the stronger rationale, I decided to further explore the possible significance of interactions between ACC, FASN and tubulin.  132  4.6 Previously identified ACC-protein interactions It is important to consider the analysis using mass spectrometry in the context of other studies, particularly in light of the constantly growing literature on protein-protein interactions. Based on specific studies as well as broader and even genome-wide analyses, many proteins have been identified that are predicted to associate with or even form direct binding partnerships with ACC. One protein, BRCA1, has been found to directly associate with ACC in human cells. BRCA1 is a 220 kDa protein that when mutated, leads to increased susceptibility in breast and ovarian cancer [146], its function being mainly involved with DNA repair. BRCA1 contains two tandem BRCA1 C-terminal (BRCT) domains and was discovered to associate with ACC through these BRCT domains. It was subsequently discovered that BRCA1 binds only to the phosphorylated form of ACC [193], consistent with the binding properties of the BRCT domains. Considering that ACC is an essential enzyme for lipogenesis, it was suggested that the binding of BRCA1 protein may act to keep ACC inactive. This concept was further supported when a crystal structure was obtained in 2008 revealing the binding of an ACC-1 phosphorylated peptide in a groove between the two BRCT domains of BRCA1 [130]. In light of this work, attempts were made to test for the presence of BRCA1 in the liver polymeric fractions. However, using anti-BRCA1 antibodies provided by Dr. Cal Roskelly (Cellular and Physiological sciences, UBC), no evidence could be found for the expression of BRCA1 in rat liver or white adipose tissue or in ACC polymeric fractions obtained by size exclusion chromatography. Positive controls using cancer cell lysates showed that the anti-BRCA1 antibody was able to detect expressed BRCA1. Overall, although ACC-BRCA1 interaction occurs in some cell types, it seems this interaction may not be physiologically important in fat or liver cells and the overall significance is not easy to predict. For example, binding of BRCA1 to phospho-ACC appears likely to “lock” ACC in an inactive form and might even promote its degradation. Just how the interaction between ACC and BRCA1 might be important in mammary cells and how this might relate to cancer susceptibility remains unclear. Another protein, Trb3 has also been reported to bind to ACC and in doing so, recruits COP1, which then triggers ubiquination and degradation of the ACC protein.  133  Trb3 is a “pseudokinase” and, like BRCA1, binds mainly to phosphorylated forms of ACC. The consequence of this binding is the recruitment of COP1 to Trb3/ACC complex and subsequent targeting of ACC for ubiquination and proteasomal degradation [83]. The model developed as a result of these studies, provides a mechanism to explain the degradation of ACC during nutrient starvation and therefore to the increase in turnover of ACC during catabolic stress. Even so, ACC turnover is known to be slow, with the half time of approximately 24 hours or greater even during starvation or insulin deficiency [77, 78]. This mechanism is unlikely, therefore, to contribute to short-term ACC control. Furthermore, a subsequent report reveals that deletion of the Trb3 gene does not have any appreciable metabolic phenotype bringing this whole mechanism into question [83] More recently, secretoneurin (chromogranin B, CHGB) has also been found to associate with the human form of ACC [194], but no functional outcome has been determined. As secretoneurin is mainly expressed in neural tissues, it is perhaps possible that this interaction is important in the CNS, especially considering the role of ACC and FASN in the hypothalamus [195]. In addition, synthesis of long chain fatty acids for myelination of axons might provide another scenario in which neuronal ACC-protein interactions are important, especially in neural development. In addition to the specific studies of ACC interactions with BRCA1, Trb3 and chromogranin B, a number of wide-ranging protein-protein interaction studies have been reported. For example, a number of groups have used high throughput techniques to examine protein-protein interactions in yeast, notably in S. cerevisae. The significance of these genome-wide databases are still emerging as specific follow-up studies continue, but where similar interactions are detected in multiple screens, some confidence in significance may be generated. A variety of techniques have been used to identify protein-protein interactions, including yeast two hybrid (Y2H) analysis and the use of bait proteins, such as protein kinases, phosphatases, regulatory proteins, and DNA damage response proteins to “capture” proteins [196]. Another important approach has been the tandem affinity purification (TAP) tag method [197]. In this method, two affinity tags were engineered into the C-terminal end of each protein in the yeast proteome. Each protein was then purified by the dual affinity purification methods, and  134  analyzed through mass spectrometry to detect any protein complexes that may have formed with the tandem tagged protein. Based on the combined studies of interacting proteins in S. cerevisae, twenty-two proteins have been predicted to interact with yeast ACC (table 4.7). Of these 22 proteins, thirteen are involved in DNA structure and maintenance, transcription or translation, perhaps reflecting the effect of nutrition on the cell cycle in unicellular organisms. On this basis, ACC, through its supply of fatty acyl esters for membrane biogenesis, may have a greater impact on the cell cycle in yeast than in multicellular organisms. Three regulatory proteins were also found to interact with yeast ACC, two of these, SIT4 and PSA1, being involved in cell wall maintenance. As ACC is involved in the synthesis of fatty acids for membrane lipids, it would be appropriate for ACC to be associated with proteins that maintain the integrity of the yeast cell wall. These proteins could effectively activate or inhibit ACC, leading to the production or decrease in fatty acids needed to supplement the cell wall. The third regulatory protein, SNF-1, is a yeast homolog of mammalian AMPK, arguably the major ACC kinase in mammalian cells Among yeast cytoskeletal proteins that might associate with yeast ACC, the interactive studies have identified CCT5 and VAC14, which are involved in cytoskeletal assembly and protein trafficking. This may indicate an indirect relationship between ACC and cytoskeletal proteins, but no direct “hits” to tubulin or actin have been reported. Finally, several metabolic proteins were identified to associate with yeast ACC. However, none of these proteins matched the list of proteins that were found in the mass spectrometry analyses done on rat proteins in this thesis.  135  Table 4.7: Yeast proteins that interact with ACC1 (FAS3) [196-198]. DNA structure and maintenance proteins YKU80, YMR106C Involved in telomere length maintenance, and DNA repair [199] DMC1, YER179W Involved in DNA repair [200] SHS1, YDL225W One of five septins involved in DNA replication stress [201] Proteins involved with transcription or translation UTP5, YDR398W Nucleolar protein involved in the processing of pre-18S rRNA [202] RPC40, RPC5, RNA polymerase subunit, common to RNA polymerase I and YPR110C III [202] DTD1, YDL219W A hypothetical protein, potentially involved in protein translation [203] APQ12, YIL040W Involved in nuclocytoplasmic transport of mRNA MRPL22, YNL177C Mitochodrial ribosomal protein of the large subunit HSP82, HSP90, Chaperone involved in mitochondrial preprotein delivery, also YPL240C interacts with a number of other proteins SNP1, YIL061C Involved with mRNA splicing, and possibly with mRNA polyadenylation PRE1, YER012W A subunit in the 20S proteasome RNA-dependent ATPase RNA helicase BRR2, PRP44, RSS1, SLT22, YER172C HPR1, TRF1, Involved with complexes that couple transcription elongation YDR138W with mitotic recombination, and cell lifespan Regulatory proteins SIT4, LGN4, A serine/threonine phosphatase involved in cell wall YDL047W maintenance PSA1, MPG1, SRB1, GDP-mannose pyrophosphorylase, needed for cell wall VIG9, YDL055C structure [204] SNF1, CAT1, CCR1, AMP-activated serine/threonine protein kinase involved in GLC2, YDR477W transcription of glucose-repressed genes and sporulation Cytoskeletal proteins CCT5, TCP5, YJR064W Involved in the assembly of actin and tubulins in vivo VAC14, YLR386W Involved with protein trafficking Metabolic proteins URA7, YBL039C A major CTP synthase isozyme, involved in the final step in de novo biosynthesis of pyrimidines and also in phospholipid biosynthesis [205] HFA1, YMR207C Mitochondrial acetyl-coenzyme A carboxylase YJL070C A hypothetical protein, potentially involved with AMP deamination MAE1, YKL029C Mitochondrial malic enzyme, converting malate to pyruvate Although the list of yeast ACC-interacting proteins can provide insight for potential identities of rat ACC-interacting proteins, there are many basic differences  136  between the respective forms of ACC, which may lead to differences in interacting proteins. For example, yeast ACC is insensitive to the effects of citrate, and does not show any tendency to polymerize. It is also necessary to consider that protein complexes may form upon certain environmental or nutritional states that have yet to be explored in the high-throughput analyses done so far. To resolve these issues, I examined another more broadly-based database designated STRING (search tool for the retrieval of interacting genes/proteins). This database is a composite that captures functional protein-protein interactions that are mapped and predicted based on several criteria including co-expression of genes, gene proximity in the genome and information from other protein-protein interaction databases [206]. Based on information derived from STRING, a summary of predicted binding partners for E. coli, S. cerevisiae, rat (R. norvegicus), and human ACC is shown in table 4.8. Interestingly, FASN is identified to be a potential binding partner for E. coli, yeast, and for human ACC. Other enzymes of particular interest are two involved in the production of acetyl-CoA, namely acetyl-coenzyme A synthetase and ATP-citrate lyase. Once again, AMPK was also shown to be a potential binding partner for ACC suggesting that AMPK is not only an important regulator of ACC activity, but might also form a functional complex with ACC. Malonyl-CoA decarboxylase is also another interesting result, as its reaction is the formation of acetyl-CoA from malonyl-CoA. However, malonyl-CoA decarboxylase is localized mainly within the mitochondria [207], which limits its interaction with ACC. Amongst the various proteins that could be a potential ACC binding partner, FASN appears to be a strong functional candidate. Since FASN was also identified in the LC MS/MS analyses, FASN was chosen to be the first ACC binding partner candidate for further experimentation.  137  Table 4.8: List of predicted protein-binding partners of ACC, based on function of E. coli, S. cerevisiae, R. norvegicus, and human. Protein E. coli fabB, fabD, fabF, fabH glcB acs Yeast, FAS3 SNF1 SNF4 HFA1 FAS1, FAS2 ACS2 HPR1 ERG10 PDA2 DMC1 Yeast, HFA1 FAS3 SNF1 SNF4 SIP2 GAL83 YHR067W ETR1 NOP58 LSM2 Rat, ACC-1 and ACC-2 Prka Acacb Dlat Acat1 Human, ACC-1 PRKA FASN RNF53 ACSS2 MCD ACLY SREBF2  Function Enzymes involved in the synthesis of fatty acids Malate synthase Acetyl-coenzyme A synthetase Carbon catabolite derepressing protein kinase (AMPK) Nuclear protein SNF4 (regulatory protein CAT3) Mitochondrial ACC Fatty acid synthase Acetyl-coenzyme A synthetase 2 THO complex subunit, HPR1 Acetyl-CoA acetyltransferase Pyruvate dehydrogenase Meiotic recombination protein Acetyl-CoA carboxylase Carbon catabolite derepressing protein kinase Nuclear protein SNF4 (regulatory protein CAT3) Protein SIP2 Glucose repression protein Uncharacterized protein Mitochondrial respiratory function protein 1 Nucleolar protein Small nuclear ribonucleoprotein D 5’-AMP-activated protein kinase ACC-2 Pyruvate dehydrogenase Acetyl-CoA acetyltransferase 5’-AMP-activated protein kinase catalytic subunits Fatty acid synthase Breast cancer type 1 susceptibility protein Cytoplasmic acetyl-coenzyme A synthetase Malonyl-CoA decarboxylase ATP-citrate lyase Sterol regulatory element-binding protein-2  138  Protein Human, ACC-2 PRKA FASN CPT1A,B ACSS2 MCD SREBF1  Function 5’-AMP-activated protein kinase catalytic subunits Fatty acid synthase Carnitine palmitoyltransferase 1A Acetyl-coenzyme A synthetase Malonyl-CoA decarboxylase, mitochondrial precursor Sterol regulatory element-binding protein-1  4.7 Summary Due to issues with the monomeric avidin-agarose column, attempts to repeat the experiments done by Quayle et al to purify the ACC protein regulator were unsuccessful. An alternative method was developed to identify this ACC protein regulator and to explore the more general possibility of ACC-associated proteins. The protein regulator identified previously was determined to increase the citrate sensitivity of ACC, and attempts were initially made to isolate ACC polymers and any potential protein binding partners. Sucrose gradient centrifugation was first used to separate the polymeric and dimeric forms of ACC, but for practical reasons, a BioGelA-50M size exclusion column was developed instead to increase confidence in full separation of ACC polymers and dimers. The citrate-dependent shift in the size of ACC was verified through enzyme activity and by western blotting analyses of the eluted fractions from the BioGelA-50M column. Citrate induced an obvious increase in ACC elution in high molecular weight “polymeric” forms based on all the above measures. In addition, high molecular weight forms of ACC appear to have substantially higher specific activity than lower molecular weight forms. SDS-PAGE analysis showed that a number of proteins co-migrated with ACC polymers in the presence of citrate, despite the fact that these proteins normally do not show a citrate-dependent size shift. To identify these proteins, mass spectrometry was employed following in-solution and in-gel trypsinzation of the citrate-induced polymeric fractions. Three proteins were consistently identified by LC MS/MS, including FASN, tubulin, and actin. A search of specific literature and of protein-protein interaction databases reveals a number of additional proteins that should be considered potential ACC-interacting proteins. The  139  STRING database of functional protein-protein interactions combined with the results revealed in this chapter indicated that fatty acid synthase is a strong candidate as an ACC binding partner. Based on the results from this chapter as well as studies from other groups, tubulin was also considered as a second candidate that merits further study as a potential ACC-binding protein. The work described in the two concluding chapters deals with attempts to test the significance of interactions between ACC and FASN or tubulin.  140  5  Chapter 5: Interactions of ACC with FASN  5.1 Rationale In the previous chapter, evidence was provided that the formation of ACC polymers upon citrate-induced activation, leads to the co-association of additional proteins with the ACC subunits. Among the ACC polymer-associated proteins, two emerged that were detected most often and with the highest degree of confidence, namely fatty acid synthase and tubulin. Therefore, potential association between ACC and FASN or tubulin was chosen for further characterization, beginning with studies to examine the association between ACC and FASN. FASN was considered of particular interest because it exhibited very high mowse scores in the mass spectrometry analyses and also from a functional perspective because the product of the ACC reaction, malonyl-CoA, is a key substrate for FASN. These two enzymes, and perhaps others, might be candidates to form a potential “fatty acid metabolon”. In this chapter, several experimental approaches were used to try to verify the extent and significance of interactions between ACC and FASN. Size exclusion column fractions were analyzed by western blot with antibodies against FASN to assess the presence of FASN in the ACC polymeric fractions prepared in the absence and presence of citrate. Procedures for co-immunoprecipitation were also developed using antibodies against ACC and FASN for the primary immunocapture, followed by western blotting using both antibodies. Finally, attempts were made to explore the distribution of FASN activity between free and ACC-associated forms as well as effects of anti-FASN on ACC activity.  141  5.2 Co-migration of FASN in the ACC polymeric fraction Having demonstrated the presence of FASN in the peak ACC polymeric fractions through mass spectrometry analysis, the extent of this co-association was assessed by western blotting to give a more quantitative analysis by assessing the abundance of FAS in all BioGelA-50M column fractions with anti-FASN (figure 5.1). In the absence of citrate, FASN is present only in the size range corresponding to ACC dimers, consistent with the similar mass of FASN dimers (approximately 540 kDa). In the presence of 20 mM citrate, a citrate-dependent shift in size for FASN was observed, so that a significant fraction of total FASN protein eluted in the same large molecular mass fraction as ACC polymers. As FASN itself is unaffected directly by citrate and is unable to polymerize alone, the observed size-shift indicates that FASN associates directly or indirectly with ACC. Some smearing of the FASN is present, indicating incomplete digestion, and also overexposure of the films. The FASN bands detected by Western blotting were quantitated in all BioGelA-50M column fractions allowing calculation of the proportion of FASN in large “polymeric” or small “dimeric” forms (figure 5.2). The values for size distribution of FASN were also compared to the values for distribution of ACC protein determined from the anti-ACC1 and streptavidin-HRP blots (table 5.1).  142  Figure 5.1: Citrate-dependent size-shift of fatty acid synthase. Rat liver preparations were purified through ammonium sulfate precipitation, pre-cleared through the BioGelA-50M column in the absence of citrate and then incubated in presence or absence of 20 mM citrate. The incubated extracts were then loaded onto a BioGelA-50M column in the absence (a) or presence (b) of 20 mM citrate. Fractions (1 mL) were collected and concentrated through TCA precipitation. Each fraction was subjected to SDS-PAGE (acrylamide concentration of 4.5% in the stacking gel and 6% in the separating gel) and then transferred onto a PVDF membrane. The membranes were then probed with an anti-FASN antibody. Column fraction numbers are indicated above the image and the migration of FASN subunits is indicated with the arrow. Polymers  Dimers  a)  8 9 10 11 12 15 16 17 18 19 20 21  b)  8 9 10 11 12 15 16 17 18 19 20 21  143  Figure 5.2: Citrate-dependent size-shift of FASN. Images obtained from Western blots as shown in figure 5.1 were scanned to quantify the FASN immunoreactive bands in each column fraction and the percent of FAS found in each fraction (polymeric or dimeric) was determined relative to the total for the entire chromatogram in the absence of citrate (black) or presence of citrate (red). Values indicate FASN protein content of high mass (polymer) column fractions 8 to 12 and low mass (dimeric) column fractions 15 to 21 as percentage total and are the average of two independent experiments. 120  % FAS protein  100 80 60 40 20 0 polymeric  dimeric Fraction  144  Table 5.1: Effects of citrate on the elution of ACC and FASN during size exclusion chromatography fractions following BioGelA-50M size exclusion chromatography was calculated as in figures 4.10 and 5.2 and presented as percentage total in each chromatogram.  Anti-ACC1  Anti-FASN  Citrate  ACC protein/activity in polymeric fractions as percentage total  ACC protein/activity in dimeric fractions as percentage total  0 mM  21.5%  78.5%  20 mM  55.1%  44.9%  0 mM  2.7%  97.3%  20 mM  36.5%  63.5%  Interestingly, the elution of FASN protein in the larger mass polymeric fraction was minimal in the absence of citrate, thus further demonstrating the effectiveness of the pre-clearing column step and shows that the co-elution of FASN and ACC is almost completely citrate-dependent. The elution of FASN protein in the polymeric fraction increased to approximately 37% of total in the presence of citrate, whereas the amount of ACC protein eluting in the high mass polymeric fraction increased from approximately 22% to more than 55% as a result of citrate treatment combining values for anti-ACC1 and streptavidin-HRP blots. These values are relative and therefore indicative of changes in protein distribution but do not give any indication of the stoichiometry of association between ACC and FASN. Nevertheless because the total expression of FASN and ACC activities in liver and adipose tissue differ at most by a factor of two, the relative association in the large mass polymeric fractions is likely to be appreciable in the presence of citrate.  145  5.3 Co-immunoprecipitation Size-exclusion chromatography, coupled with analysis of column fractions by mass spectrometry analysis and by western blotting provides strong evidence for citrateinduced association between ACC and FASN. The next step was to determine if the interaction between ACC and FAS could also be detected by co-immunoprecipitation, a process involving more stringent washing. The essence of co-immunoprecipitation methods is to determine if antibodies directed against one protein are able to allow simultaneous capture on immobilized beads of the cognate protein and other binding partners. Two control experiments were first performed to confirm the lack of non-specific binding to protein G agarose beads to rat liver protein fractions that were prepared by centrifugation and ammonium sulfate precipitation (0 – 40% saturation) as outlined in Methods. As can be seen in figure 5.3 (condition A), ACC protein was clearly detected in the supernatant but none was bound to protein G beads in the absence of antibodies. The second control experiment was carried out to confirm binding of antibody to protein agarose G beads. Protein G agarose beads were therefore incubated with a sample (5 μL) of anti-ACC-1 for two hours at 4°C. Following incubation, the beads were washed five times and bound proteins were eluted as before. Following SDS-PAGE and transfer, staining with amido black revealed the presence of the expected IgG bands. Curiously, IgG bands were also detected in blotting with HRP-streptavidin.  146  Figure 5.3: Controls for the co-immunoprecipitation experiments. To test for non-specific binding to protein G agarose beads, rat liver protein fractions were recovered by ammonium sulfate precipitation (0 – 40% saturation), pre-cleared using protein agarose G beads, and incubated with protein agarose beads for one hour in the presence of 20 mM citrate (condition A). To confirm IgG binding to protein G agarose, anti-ACC was incubated with protein agarose G beads in the absence of any other protein for one hour in the presence of 20 mM citrate (condition B). Following incubations, the beads were washed five times and boiled with SDS sample loading buffer to elute bound proteins. The SDS-eluted proteins were subjected to SDS-PAGE (acrylamide concentration of 3% in the stacking gel and 6 % in the separating gel), transferred to a PVDF membrane and probed with streptavidin-HRP.  elution  wash 5  supernatant  B  elution  wash 5  supernatant  markers  A  150 kDa 100 kDa 75 kDa  30 kDa  147  The key first set of experiments involved the use of anti-phospho-ACC and antiACC1 antibodies to immunoprecipitate the ACC protein and potentially to test for the presence of ACC and FASN by subsequent western blotting. As before, ACC and FASN were recovered from rat liver high speed supernatant fractions by precipitation with ammonium sulfate (0 – 40% saturation). The proteins were then re-dissolved and the sample was pre-cleared by incubation with protein G agarose beads in the absence of antibodies as mentioned previously. The supernatant from the pre-clearing was then incubated with either anti-phospho ACC or the anti-ACC1 antibodies that had previously been immobilized on protein G agarose beads. Following incubation, the beads were washed five times and the bound proteins were eluted by boiling in SDS sample loading buffer. Following elution, samples of each step were subjected to SDS-PAGE and western blotting as described in Methods. Both the anti-phospho-ACC and the anti-ACC1 antibody were successful in bringing down the ACC protein (figure 5.4). However, the anti-phospho-ACC antibody seemed to be more successful as the membrane from the ACC1 antibody coimmunoprecipitation had to be exposed to film for much longer to achieve comparable image intensity. In fact, the anti-ACC-1 antibody used in this particular experiment was recognized to be least effective for Western blotting, which would explain why it was less effective for co-immunoprecipitation. Subsequent probing of the membranes by Western blotting with anti-FASN antibodies showed that no FASN was pulled down by immunopreciptiation in the presence of the ACC antibody. Since the early studies of using anti-ACC antibodies to differentiate the two major isoforms of ACC [61], few experiments have been performed in which anti-ACC antibodies have been used to immunoprecipitate ACC and no experiments have been reported in which anti-ACC antibodies have been used in co-immunoprecipitation experiments. The only studies reported so far have been those using the anti-BRCA1 antibody, which was shown to pull down the BRCA1 protein, along with ACC-1.  148  Figure 5.4: Immunoprecipitation of ACC with anti-phospho ACC and anti-ACC1. Proteins recovered from rat liver high-speed supernatant fractions were by ammonium sulfate precipitation (0 – 40% saturation), and incubated with protein G agarose beads in the presence of 20 mM citrate. Following incubation, the beads were centrifuged (1 min at 11,000 x g) and the supernatant was removed. This supernatant was then incubated with protein G agarose beads that had been pre-loaded with either the anti-phospho ACC antibody (panel A) or the anti-ACC1 antibody (panel B). Following incubations, the beads were washed five times, and mixed with SDS sample loading buffer and boiled to elute any bound protein. Eluted proteins were then subjected to SDS-PAGE gel (acrylamide concentrations of 3% in the stacking gel and 6% in the separating gel), transferred to a PVDF membrane and probed with streptavidin-HRP. The illustrated experiments were repeated twice with similar results.  elution  wash 5  supernatant  LAS  markers  elution  wash 1  supernatant  LAS  B  markers  A  150 kDa 150 kDa 100 kDa 100 kDa 75 kDa 75 kDa 30 kDa  30 kDa  149  The converse experiment was performed to test if the anti-FAS antibody was able to pull down the FASN protein, and potentially ACC as well. However, the anti-FAS antibody was apparently not able to bring down the FASN protein itself. Figure 5.5 shows the image of the PVDF membrane probed with anti-FAS after using anti-FAS antibody to pull down proteins from the liver ammonium sulfate fractions. Even when the film was substantially overexposed, there was no evidence of FAS. Figure 5.5: Attempts to immunoprecipitate proteins with anti-FAS antibody. Proteins from rat liver high speed supernatant fractions were recovered by ammonium sulfate precipitation (0 – 40% saturation), pre-cleared by incubating with protein agarose G beads for at least one hour in the presence of 20 mM citrate. Following incubation, the beads were centrifuged (1 min at 11,000 x g) and the supernatant was removed and incubated with a second batch of protein G agarose that had been pre-loaded with antiFASN. Following incubations, the beads were washed thoroughly, and mixed with SDS sample loading buffer and boiled to elute any bound protein. The eluted proteins were then subjected to SDS-PAGE gel (acrylamide concentrations of 3% in the stacking gel and 6% in the separating gel), transferred to a PVDF membrane and probed with either streptavidin-HRP (condition A) or anti-FASN antibody (condition B). The experiment was repeated twice with similar results.  150 kDa  elution  wash 5  supernatant  markets  elution  wash 5  supernatant  B  markers  A  150 kDa  100 kDa 75 kDa 75 kDa 30 kDa 30 kDa  150  Under the outlined conditions, the FASN antibody did not appear to precipitate any biotinylated proteins, nor the FASN protein itself. To pursue this specific objective, different preparations of anti-FAS would have to be developed that are able to work in immunoprecipitation procedures. This work was not pursued further at the time. Perhaps one additional control interpretation is that at least no ACC was recovered using protein G agarose beads loaded with a non-functional IgG.  5.4 The effect of FASN on ACC activity Several attempts were made to characterize the effects of association between ACC and FASN on enzyme activity. In the first set of experiments, anti-FASN antibodies were tested for effects on ACC activity. The rationale behind this experiment was to determine whether the interaction between FASN and its cognate antibody was able to disrupt ACC activity. In this experiment, proteins were recovered from rat liver high speed supernatant fractions by ammonium sulfate precipitation (0 – 40% saturation) and pre-incubated in 20 mM citrate prior to addition of various concentrations of the FASN antibody. Following citrate and anti-FAS incubations, ACC activity was determined. Despite several attempts using a wide range of anti-FAS antibody concentrations, no evidence for any effect on maximal ACC activity was obtained. The lack of effect of the FASN antibody on ACC activity indicates that the association between ACC and FASN could be a structural association, and FASN may not have an effect on ACC activity. Attempts were therefore made to purify FAS to determine if the purified protein had any effect on ACC activity. To purify FASN, a method similar to the one described by Hardie and Cohen [32] was used. In this method, following ammonium sulfate precipitation, a step gradient PEG precipitation of the rat liver preparations was performed, which would effectively separate FASN and ACC. After the purification was completed, there was a complete loss of both FASN and ACC activity. Repeated attempts to purify FASN were unsuccessful.  151  5.5 Summary Analysis of ACC polymer fractions by mass spectrometry identified FASN as an “ACC-associated” protein and subsequent Western blotting of size-exclusion column fractions then confirmed the citrate-induced association of ACC and FASN. Analysis of ACC/FASN association during size-exclusion chromatography also provided the basis to calculate relative FASN elution in large and small molecular forms together with ACC in the absence and presence of citrate. In fact, a substantial proportion (up to 30%) of FAS co-eluted with ACC in the presence of citrate. Co-immunoprecipitation studies gave mixed results. Whereas FASN did co-immunoprecipitate with ACC using anti-phosphoACC, the anti-ACC-1 and the anti-FASN antibody preparations were less effective or non-effective at immunoprecipitation and no clear conclusions could be drawn from these experiments and further work is required using good immunoprecipitation antibodies. Additional future work might also exploit the phospho-ACC antibody, to test for the potential presence of other ACC-associated proteins. Direct addition of anti-FASN antibody to rat liver preparations that contained both ACC and FAS did not have any measurable effect on ACC activity. Attempts were made to purify a protein fraction containing FASN free from ACC, as based on a previous procedure but these were unsuccessful.  152  6  Chapter 6: Interactions of ACC with tubulin  6.1 Rationale The results shown in chapter 4 demonstrated that isoforms of tubulin were found to be associated with the ACC polymeric fraction during size-exclusion chromatography. It is important to recognize that tubulin is a ubiquitous and a highly-expressed protein and so it is possible that the apparent association between tubulin and ACC is non-specific. On the other hand, several published reports provide evidence to support the notion that ACC might be influenced by microtubule function. For example, microtubule effectors such as GTP and colchicine have been found to affect ACC activity [208] and others have provided evidence that the slow elution of ACC from digitonin-permeabilized hepatocytes might be accounted for by binding to the microtubule elements of the cytoskeleton [209]. One caveat to these studies is that some groups have reported that the effects of GTP might be mediated by direct binding to ACC [192]. However, in light of our studies, it is possible that the ACC preparations used in these earlier studies actually contained tubulin as well and that the observed effects of GTP might therefore be explained via binding to tubulin or ACC or both. More interestingly, recent work in our laboratory has demonstrated that tubulin is still present when ACC is highly purified through avidin-affinity chromatography, a procedure that involves exposure to high salt (0.5 M KCl) washes. Since tubulin is not known to bind to avidin and this observation provides further evidence for a strong interaction between ACC and tubulin. In this chapter, I describe several experimental approaches used to further test the association between ACC and tubulin, including approaches similar to those described in chapter 5 for the association between ACC and FASN. In the first set of experiments, column fractions obtained following BioGelA-50M size exclusion chromatography were analyzed by western blotting with antibodies for tubulin. Subsequently, microtubules were purified from rat brain to test in re-constitution assays with purified ACC in combination with microtubule activators and inhibitors. Anti-tubulin antibodies were further tested for their ability to influence ACC activity and to co-immunoprecipitate ACC. Finally, an immunocytochemistry approach was used to examine the cellular localization of ACC and tubulin in primary hepatocytes. We could find no published  153  evidence for immunocytochemical analysis of ACC at the ultra-structural level, so this represented a novel approach that demanded method development and optimization.  6.2 Co-migration of tubulin and ACC during size-exclusion chromatography Similar to experiments described in chapter 5, the distribution of tubulin in the BioGelA-50M column fractions in the absence and presence of 20 mM citrate was examined with an α-tubulin (B7) antibody (figure 6.1). In the absence of citrate, tubulin was detected only in the relatively low mass fractions that contained ACC dimers, notably fractions 16-21. When the size-exclusion analysis was carried out in the presence of 20 mM citrate however, a strong tubulin signal was detected in the high mass “polymeric” fractions, notably fractions 8 to 12. Tubulin is known to polymerize upon incubation with GTP but not with citrate, so the appearance of a significant proportion of tubulin in the ACC polymeric fraction suggests some form of association between these two proteins. The amount of tubulin in the polymeric and dimeric size-exclusion fractions was quantitated in the absence and presence of citrate (figure 6.2).  154  Figure 6.1: Citrate-dependent co-migration of ACC and tubulin during size-exclusion chromatography. Rat liver ACC was partially purified through ammonium sulfate precipitation, pre-cleared through the BioGelA-50M column in the absence of citrate and then incubated and chromatographed on a BioGelA-50M column in the absence (a) or presence (b) of 20 mM citrate. Fractions (1 mL) were collected, concentrated by TCA precipitation and subjected to SDS-PAGE (acrylamide concentrations of 5% in the stacking gel and 8% in the separating gel). The proteins were then transferred onto a PVDF membrane and probed with an anti-α-tubulin antibody. Column fraction numbers are indicated along the top margin and the arrow indicates the migration of tubulin (approximately 50 kDa). Polymers a)  Dimers  8 9 10 11 12 15 16 17 18 19 20 21 220 kDa 52 kDa  b)  8 9 10 11 12 15 16 17 18 19 20 21 220 kDa  52 kDa  155  Figure 6.2: Citrate-dependent mobility size shift of tubulin revealed by size-exclusion chromatography. The experiment described in figure 6.1 was repeated three times and resulting images were scanned and quantitated. The intensity of tubulin staining in the polymeric fractions (8-12) and dimeric fractions (16-21) is expressed relative to total in each chromatogram. Pre-cleared rat liver ammonium sulfate fractions (0-40% saturation) were subjected to chromatography on BioGelA-50M either at 4°C in the absence of citrate (black bars) or at room temperature in the presence of 20 mM citrate (red bars) (n = 3). 100 90  % tubulin protein  80 70 60 50 40 30 20 10 0 polymeric  dimeric Fraction  156  Table 6.1: Summary of distribution of ACC and tubulin during size-exclusion chromatography. The elution of ACC and tubulin in large (polymeric) and small (dimeric) fractions following BioGelA-50M size exclusion chromatography was calculated based on western blotting of column fractions as in figures 4.10 and 6.2 and the results are presented as percentage total in each chromatogram.  ACC  tubulin  Citrate  ACC or tubulin in polymeric fractions  ACC or tubulin in dimeric fractions  0 mM  9%  78.5%  20 mM  55.1%  44.9%  0 mM  9.5%  90.5%  20 mM  35.8%  64.2%  On average, citrate induced an approximate three-fold increase in the abundance of tubulin in the high molecular weight fractions, so that more than 30% of total tubulin migrated in the some fractions as polymeric ACC in the presence of citrate. In the absence of citrate, approximately 10% of tubulin eluted in the high molecular weight fractions. A statistical analysis (P < 0.05) indicates a 95% confidence that values are statistically different, confirming the results from the Western blots where tubulin shifts to the high molecular size fraction in the presence of citrate.  157  6.3 The effect of tubulin on ACC activity To determine if the association between ACC and tubulin had any influence on ACC activity, the effects of tubulin preparations and tubulin modulators on ACC activity were determined. Anti-tubulin antibodies were also tested to determine whether they had an effect on ACC activity To test the effects of anti-tubulin antibodies, ACC was partially purified from rat liver through ammonium sulfate precipitation (40% saturation) and pre-incubated in the presence of 20 mM citrate and then with increasing concentrations of α-tubulin antibody prior to assay of ACC activity. These experiments with several independent preparations of ACC showed that there was no effect of the tubulin antibody on ACC activity. To test the effects of tubulin on ACC, microtubules were purified from isolated rat brains according to a method described by Vallee [152]. This method proved successful and microtubules were isolated with the presence of several other microtubuleassociated proteins (figure 6.3). The final concentration of the purified microtubules was found to be 8 mg/mL.  158  A  2nd assembly, supernatant  2nd assembly, pellet  disassembly, supernatant  disassembly, pellet  1st assembly, supernatant  1st assembly, pellet  homogenate, supernatant  homogenate, pellet  markers  Figure 6.3: Isolation of microtubules from rat brain. Rat brains were homogenized, incubated in the presence of 0.1 mM GTP and 2.5 mM ATP, and allowed to go through one cycle of microtubule assembly and disassembly as outlined in Methods. Following one more assembly step, the microtubules were recovered in the pellet. Samples of each isolation step were subjected to SDS-PAGE (acrylamide concentrations of 5% in the stacking gel, and 10% in the separating gel) and stained with Coomassie Blue (panel A) or the proteins were transferred to a PVDF membrane and probed with an anti-α-tubulin antibody (panel B). The tubulin band is indicated with the arrow.  225 kDa 76 kDa 50 kDa 35 kDa  B 150 kDa 100 kDa 50 kDa 35 kDa 25 kDa  159  The rationale for purifying microtubules was prompted by work described in the introduction and particularly by the studies of Buechler et al, in which they demonstrated that microtubule effectors, such as GTP and colchicine had opposing effects on ACC activity. In their experiments, they found that GTP, which is required for microtubule assembly, also activated ACC while colchicine, a known microtubule inhibitor was found to decrease ACC activity. In these studies, the effects of GTP and colchicine were observed in the absence of citrate. In the presence of 20 mM citrate, the inhibitory effects of colchicine were overcome, and ACC activity was brought back to control levels. For the studies described here, ACC was purified from rat liver through ammonium sulfate precipitation (40% saturation) and resuspended in homogenizing buffer. In the first set of experiments, the effects on ACC activity of varying the concentration of citrate, GTP, or colchicine were determined. The results of the assays are shown separately for the effects of GTP in combination with citrate (figure 6.4) and for colchicine in combination with citrate (figure 6.5). GTP exerted a clear and concentration-dependent inhibitory effect at all concentrations of citrate. In proportional terms, the effect of GTP on ACC activity was less evident in the absence of citrate. This is somewhat surprising and likely reflects an interference with citrate-induced activation and/or polymerization rather than a catalytic effect. These results conflict with the activating effects of GTP observed by Buechler et al and it is not clear why the results differed. Although one possibility arises because of the different procedures for ACC preparation. The method that was used by Buechler et al involved precipitation with Triton X-100, was also used to purify cytoskeleton elements so it is probable that the method used by Buechler et al enhanced the recovery of tubulin with ACC. The effects of addition of tubulin to ACC will be addressed in the experiments described below. Another possibility is that the effects of GTP may differ according to the level of purification of the ACC enzyme. Another study by Witters et al also involved minimal ACC purifications and in this study, several guanine nucleotides, in particular 5’-GTP, had an activating effect on ACC that was purified through high-speed centrifugation [210]. However, these authors were unable to detect any effect of GTP on ACC purified through avidin affinity chromatography.  160  Figure 6.4: Effect of GTP on ACC activity. a) ACC was partially purified from rat liver preparations through ammonium sulfate precipitation and incubated with GTP (♦ - 0 mM, ■ – 0.5 mM, ▲ – 1 mM, × - 2 mM) for 20 minutes, and then for another 20 minutes with citrate, prior to assay of ACC activity. The experiment was repeated once, with similar results. b) The dose-dependent inhibition of ACC by GTP is shown in a derivative plot at 2 mM (■) or 20 mM ({) citrate. a)  ACC Activity (mU/mL)  4000 3500 3000 2500 2000 1500 1000 500 0 0  5  10  15  20  Citrate Concentration (mM)  b)  ACC Activity (mU/mL)  4000 3500 3000 2500 2000 1500 1000 500 0 0  0.5  1  1.5  2  GTP concentration (mM)  161  In order to test the effects of colchicine, rat liver ACC was partially purified through ammonium sulfate precipitation and then treated with various concentrations of colchicine at each of several citrate concentrations (figure 6.5). No clear effects of colchicine were observed at any citrate concentration. Figure 6.5: Effect of colchicine on ACC activity. The effect of colchicine on ACC activity was studied using liver ammonium sulfate extracts. Rat liver preparations were purified through ammonium sulfate precipitation and incubated with colchicine (♦ - 0 mM, ■ – 0.01 mM, ▲ – 0.1 mM, and × - 1 mM) for 20 minutes, and then for another 20 minutes with 20 mM citrate prior to assay of ACC activity. The experiment was repeated once with similar results. 4500  ACC Activity (mU/mL)  4000 3500 3000 2500 2000 1500 1000 500 0 0  5  10  15  20  Citrate concentration (mM)  162  To further explore the possible effects of GTP and colchicine on ACC activity, an additional study was performed to test the effects of these agents on ACC in the absence and presence of tubulin purified from rat brain. ACC was isolated from rat livers through ammonium sulfate precipitation (40% saturation) and resuspended in homogenizing buffer. The ACC preparations were then incubated with GTP or colchicine in the absence or presence of microtubules, followed by further incubation in the absence or presence of 20 mM citrate, prior to assay of ACC activity (figure 6.6).  163  Figure 6.6: The effects of GTP, colchicine, and tubulin on ACC activity. Rat liver ACC was partially purified through ammonium sulfate precipitation and incubated for 30 minutes at 37°C with the indicated additions of colchicine (Col, 1 mM), GTP + MgSO4 (GTP, 2 mM), tubulin (TUB, 2 μg/μL) or the indicated combinations. Incubations were then continued for an additional 30 minutes in the absence (black) or presence of 20mM citrate (red) prior to assay of ACC activity. Values are expressed relative to the maximum activity obtained at 20mM citrate in the absence of other additions. Values shown are the average of two independent experiments in which maximal ACC activities were 3670 and 2847 mU/mL.  ACC Activity (% maximum)  120 100 80 60 40 20 0 None  Col  GTP  Tub  Tub + Col  Tub + GTP  Additions  164  The experiments so far described to test the effects of GTP and colchicine had involved the use of rat liver ACC purified through ammonium sulfate precipitation. To test these agents on more purified ACC preparations, the ammonium sulfate-purified ACC was further purified either by BioGelA-50M size exclusion chromatography or by monomeric avidin-agarose affinity chromatography. Following these further purification steps, ACC was tested for response to GTP, colchicine and microtubules, and the results are shown in figure 6.7. In addition, table 6.2 gives a summary of the effects of GTP and colchicine on ACC preparations at the different levels of purification.  165  Figure 6.7: The effects of GTP, colchicine, and tubulin on ACC activity. Rat liver ACC was partially purified through ammonium sulfate precipitation, followed by BioGelA-50M size exclusion chromatography. The polymeric fraction (fractions 8 to 12) were pooled, concentrated, and incubated for 30 minutes at 37°C with the indicated additions of colchicine (Col, 1 mM), GTP + MgSO4 (GTP, 2 mM), tubulin (TUB, 2 μg/μL) or the indicated combinations. Incubations were then continued for an additional 30 minutes in the absence (black) or presence of 20mM citrate (red) prior to assay of ACC activity. Values are expressed relative to the maximum activity obtained at 20mM citrate in the absence of other additions. Values shown are the average of two independent experiments in which maximal ACC activity were 44.2 mU/mL and 243.7 mU/mL).  ACC Activity (% maximum)  160 140 120 100 80 60 40 20 0 None  Col  GTP  Tub  Tub + Col  Tub + GTP  Additions  166  Table 6.2: Effect of tubulin, colchicine and GTP on ACC at different stages of purification. Additions Citrate  None  Colchicine GTP  Tubulin  Tubulin +  Tubulin  Colchicine + GTP ACC LAS extracts 0 mM  0%  -4.4%  -6.0%  -0.4%  -5.5%  -8.1%  20 mM  0%  -21.8%  -29.9%  -2.2%  -27.3%  -40.5%  Highly purified ACC preparations 0 mM  0%  7.2%  14.9%  12.1%  3.0%  17.8%  20 mM  0%  52.2%  12.8%  45.4%  28.3%  -0.1%  1  Values represent the absolute difference between conditions, in comparison to ACC activity in the absence of any additions. A positive value indicates a change leading to increased activation of ACC, while a negative value indicates inhibition of ACC activity. 2 The results are an average of two separate experiments. In experiments with ACC purified partially through ammonium sulfate precipitation (figure 6.6), the effects of GTP, colchicine, and tubulin were rather modest (table 6.2). Changes in ACC activity in the absence or presence of citrate were generally in the range of 10-20%. The most convincing effects were inhibitory effects of GTP of approximately 30-40% in the absence or presence of tubulin. A substantially different picture emerged in studies of the actions of GTP, colchicine and tubulin on more highly purified preparations of ACC. In this case, tubulin more than doubled ACC activity at low citrate and also increased the maximal activity of ACC in the presence of 20 mM citrate by approximately 45%. This observation is consistent with the possibility that the less purified ACC already contained sufficient endogenous tubulin to exert a maximum effect. In general, the results obtained in these ACC activity studies are not in agreement with the results obtained from Buechler et al. In the citrate dose-dependent studies, GTP had an inhibitory effect in the absence and presence of 20 mM citrate. The colchicine studies were the one exception where there was a marginal inhibitory effect in the  167  absence of citrate, but in the presence of 20mM citrate, colchicine had an activating effect. The effects of GTP and colchicine were examined in conjunction with purified microtubules. As above, the effects were studied using ACC partially purified with ammonium sulfate and also with highly purified ACC through BioGelA-50M size exclusion chromatography. The results of this comparison showed very different effects on tubulin and microtubules on different preparations of ACC. GTP, colchicine and purified microtubules all led to inhibition of ACC that was partially purified with ammonium sulfate precipitation. In contrast, these agents led to activation of highly purified forms of ACC. GTP and colchicine were more effective with the purified ACC preparations in the absence of citrate, which again is in agreement with the results by Buechler et al, but the addition of 20 mM citrate did not abolish the effects of GTP or colchicine. With the results obtained in this thesis, it cannot be concluded whether GTP, colchicine, or microtubules have a defined effect on ACC activity. The effect of tubulin on highly purified ACC preparations indicate that even in the absence of any other tubulin effectors, tubulin alone can increase ACC activity. Also, the level of purification can have an effect on the effects of tubulin, GTP and colchicine on ACC activity, and can be pursued further. However, it is possible that the association between ACC and tubulin can be a structural association, with little effect on ACC activity. This will be investigated in the remaining sections of this chapter.  168  6.4 Co-immunoprecipitation Co-immunoprecipitation was used to further test for possible association between ACC and tubulin. In the first set of experiments, anti-ACC1 antibody was used in the immunoprecipitation step and following SDS-PAGE and transfer to PVDF, the membranes were probed with either streptavidin-HRP or anti-α-tubulin antibody (figure 6.8). The results show that the ACC1 antibody was able to precipitate the ACC protein (panel A) and the corresponding blot using anti-α-tubulin antibodies showed that tubulin was also associated with the immune complex. In the converse experiments, the anti-α-tubulin antibody was pre-bound to protein G agarose for the immunoprecipitation procedure prior to blotting with streptavidin-HRP or anti-α-tubulin (figure 6.9). The results indicate that the anti-α-tubulin antibody was successful in precipitating both ACC (panel A) and tubulin (panel B). These results confirm the association between ACC and tubulin based on coimmunoprecipitation of the two proteins using either anti-ACC or anti-α-tubulin antibodies. This result shows that the association between ACC and tubulin can withstand co-immunoprecipitation and the extensive washing procedure involved and suggests that the presence of tubulin in the polymeric fraction, as demonstrated by MS/MS analysis was indeed due to protein-protein interactions between these two proteins and not a non-specific phenomenon. In fact, recent results in the laboratory have revealed, through MS/MS analysis that tubulin remains associated with ACC even after purification through monomeric avidin-agarose affinity chromatography, another procedure that involves stringent washing in high salt. This indicates the possibility that the putative interaction between ACC and tubulin may occur at other conditions other than the ACC polymeric state. It would be interesting to repeat the coimmunoprecipitation experiments in the absence of citrate to observe whether the association between ACC and tubulin can be maintained. Indeed, all of the results that have been obtained so far demonstrate a relatively strong and likely direct association between ACC and tubulin.  169  Figure 6.8: Co-immunoprecipitation of tubulin with anti-ACC antibodies. ACC was partially purified from rat liver by ammonium sulfate precipitation and incubated with protein agarose G beads for at least one hour in the presence of 20 mM citrate. Following incubation, the beads were centrifuged (1 min at 11,000 x g) and the supernatant was removed. This supernatant was then incubated with protein agarose G beads pre-loaded with the anti-ACC1 antibody. Following incubations, the beads were washed thoroughly, and mixed with SDS-PAGE sample loading buffer and boiled to elute any bound protein. The extracts were then loaded onto a SDS-PAGE gel (acrylamide concentration of 3% in the stacking gel and 6% in the separating gel), transferred to a PVDF membrane and probed with either streptavidin-HRP (panel A) or tubulin antibody (panel B) after transfer onto a PVDF membrane. The black arrow indicates the ACC band, and the red arrow indicates the tubulin band. The experiment was repeated twice with similar results.  150 kDa 100 kDa 75 kDa  elution (bound)  wash 5  supernatant (unbound)  markers  elution (bound)  supernatant (unbound) wash 5  LAS (“input”)  B  markers  A  150 kDa 100 kDa 75 kDa 50 kDa  50 kDa  membrane probed with streptavidin-HRP  membrane probed with anti-α-tubulin  170  Figure 6.9: Co-immunoprecipitation of ACC with anti-α-tubulin antibodies. ACC was partially purified from rat liver by ammonium sulfate precipitation and incubated with protein agarose G beads for at least one hour in the presence of 20 mM citrate. Following incubation, the beads were centrifuged (1 min at 11,000 x g) and the supernatant was removed. This supernatant was then incubated with protein agarose G beads pre-loaded with the anti-α-tubulin antibody. Following incubations, the beads were washed thoroughly, and mixed with SDS-PAGE sample loading buffer and boiled to elute any bound protein. The extracts were then loaded onto a SDS-PAGE gel (acrylamide concentration of 3% in the stacking gel and 6% in the separating gel), transferred to a PVDF membrane and probed with either streptavidin-HRP (panel A) or tubulin antibody (panel B) after transfer onto a PVDF membrane. The black arrow indicates the ACC band, and the red arrow indicates the tubulin band. The experiment was repeated five times with similar results.  150 kDa  150 kDa  100 kDa  100 kDa  75 kDa  elution  wash 5  supernatant  markets  elution  wash 5  supernatant  B  markets  A  75 kDa 50 kDa  50 kDa  membrane probed with streptavidin-HRP  membrane probed with anti-α-tubulin  171  6.5 Studies of the intracellular localization of ACC in primary rat hepatocytes The fact that tubulin associates with ACC in a citrate-dependent manner that the two proteins can be co-immunoprecipitated and even remain associated during avidin purification in the presence of high salt concentrations provides evidence that the association between these proteins may be specific and quite robust. On the other hand, the studies of the microtubule activator GTP and the inhibitor colchicine led to modest and not entirely consistent results. Despite the lack of convincing effects of tubulin on ACC activity, it is still possible that association between ACC and tubulin could play a structural or perhaps a targeting role. To test this idea, immunofluorescence microscopy was employed to determine the cellular localization of ACC and tubulin. At the outset, little or no published evidence could be found that demonstrated the use of immunofluorescence microscopy for the visualization of ACC localization in cells. Therefore, it was necessary to establish the basic methods to visualize ACC and tubulin in order to determine if co-localization might occur. Among several possible options, the initial focus was to optimize conditions for successful fixing, staining and microscopic examination of ACC, especially in comparison to tubulin. Studying ACC in the context of tubulin was fortunate because of the extensive knowledge from literature of immunofluorescence microscopy with respect to tubulin. The aim was therefore to establish basic techniques with respect to ACC and tubulin detection and then, if successful, to establish whether ACC and tubulin co-localize within intact cells.  6.5.1 Establishing a method for the visualization of ACC and tubulin For the purposes of cell visualization, two microscopes were available: a conventional wide lens fluorescence microscope (Nikon Elipse E400 system), and a confocal microscope (Olympus FV1000 Confocal system). The wide lens microscope was readily available, and therefore, was used initially to establish basic aspects of the methods. However, the color of the images obtained from the wide lens microscope could not be captured in the image analysis system, which was set up for black/white images only. The confocal microscope, although less available, was therefore used when possible to examine the cells in greater detail. In most cases, the wide lens microscope was used first, until greater detail was required, in which case the cells were then  172  examined via confocal microscopy to provide a more detailed understanding of the intracellular distribution of ACC relative to cell organelles and other proteins such as tubulin. Basic procedures had previously been established by the neighboring research group of Dr. Michel Roberge for the staining of cells, including the use of anti-tubulin antibodies and nuclear staining, and are outlined in Methods. The choice of cells to study merits some discussion. Several cell types were available in the laboratory, including cell lines such as HepG2, 3T3-L1, and H9c2 cells. Also available were primary cell lines such as rat fat cells and cardiac myocytes. In the end, primary rat hepatocytes were chosen to examine the localization of ACC and tubulin for a number of reasons. Firstly, primary hepatocytes should provide the most direct parallels to extend the studies with rat liver ACC. Also these primary rat hepatocytes were prepared with an established method developed in Dr. Tom Chang’s lab, and were also available on a regular basis and therefore a reliable source of cells. The high expression of ACC in liver was also considered to be an advantage in increasing the chance of successful ACC staining. Primary rat hepatocytes are also hormone responsive, opening the possibility of exploring the effects of hormones on ACC expression, activity and localization all within a relatively physiologically relevant context. Finally, confocal microscopy of a primary rat hepatocyte (figure 6.10a) showed that ACC distribution in primary rat hepatocytes were clearer, whereas there was a higher background when staining for ACC in HepG2 cells (figure 6.10b). Although the basic procedures for tubulin and nuclear staining have been optimized for various cell lines, the methods had to be adapted for the purposes outlined here. The decision to use primary hepatocytes presented some challenges in that it was necessary to verify they could be obtained routinely in adherent form and sufficient yield, as well as adaptable to an immunofluorescence method.  173  Figure 6.10: The ACC protein distribution in primary rat hepatocytes and HepG2 cells. a) Primary rat hepatocytes were isolated and cultured on cover slips. Following an overnight incubation, the cells were fixed and then probed with anti-ACC1 antibody (1:30000 dilution). b) HepG2 cells were similarly grown to approximately 70% confluence on cover slips, fixed, and then probed with the ACC1 antibody (1:30000 dilution). Images of representative individual cells were obtained using an Olympus FV1000 Confocal system (magnification 60 times). a)  50 μm  b)  20 μm  174  Control experiments were performed to confirm the specificity of antibody binding (table 6.3). The anti-tubulin antibody that was chosen had been raised in mice and the anti-ACC or anti-phospho ACC antibodies had been raised in rabbits. Corresponding anti-mouse and anti-rabbit secondary antibodies were then selected that carried distinct fluorescent labels, respectively, alexa fluor 488 – green, and alexa fluor 568 – red. In control experiments, it was shown that the anti-rabbit secondary antibodies did not react with the primary anti-mouse tubulin antibodies, and conversely, that the anti-mouse secondary antibodies did not react with the primary rabbit anti-CCC or antiphospho ACC antibodies. The secondary antibodies therefore reacted with the appropriate primary antibodies with little cross-reaction (example shown in figure 6.11)  175  Table 6.3: Summary of controls performed to verify the specificity of the anti-tubulin, anti-ACC and anti-phospho ACC antibodies. Primary antibody  Secondary antibody 3  Result (Colour)  ---  anti-mouse  None  ---  anti-rabbit4  None  anti-tubulin1  ---  None  anti-tubulin  anti-mouse  Green  anti-tubulin  anti-rabbit  None  anti-ACC2  ---  None  anti-ACC  anti-mouse  None  anti-ACC  anti-rabbit  Red  anti-phospho ACC2  ---  None  anti-phospho ACC  anti-mouse  None  anti-phospho ACC  anti-rabbit  None  1  Anti-tubulin antibodies were raised in mice. Anti-ACC and anti-phospho ACC antibodies were raised in rabbits. 3 Anti-mouse secondary antibodies were labeled with alexa fluor 488 (green dye) 4 Anti-rabbit secondary antibodies were labeled with alexa fluor 568 (red dye) 2  176  Figure 6.11: Specificity of secondary antibodies. Primary rat hepatocytes were isolated, cultured overnight on cover slips, fixed, and then incubated with rabbit anti-ACC-1 antibody (1:10000 dilution). Following the primary antibody treatment, samples were then incubated with either anti-rabbit secondary antibody (a, red) or anti-mouse secondary antibody (b, green). Images of primary rat hepatocytes were obtained using an Olympus FV1000 Confocal system (magnification 60 times). a)  b)  177  In initial experiments, the recommended protocol that had been suggested involved culture of primary hepatocytes on cover slips coated with matri-gel. However, some aspects of the early experiments showed that matri-gel may not be the best choice of matrix with which to coat the cover slips. For example, while many of the cells adhered using matri-gel, the tubulin staining appeared to be dispersed within the cell, with a somewhat limited appearance of filamentous microtubule structures. In addition, recovery of cells through fixing and washing was far from complete following adhesion of cells to the matri-gel, suggesting a rather weak attachment of the cells to the matrix. Fibronectin was chosen as an alternative adhesion media, as fibronectin is often recommended for the visualization of cytoskeletal elements, especially microtubules. The effects of culturing primary rat hepatocytes on matri-gel and fibronectin were therefore directly compared (figure 6.12). These experiments showed that a far greater number of hepatocytes remained on the cover slips after fixing and antibody staining when fibronectin was used as the adhering medium. In fact, at least ten times more cells were recovered following culture with fibronectin than with matri-gel. This conclusion was evident following immunostaining and counting of cells in multiple fields of multiple cover slips. The quality of the cells was also clearly different with hepatocytes cultured using fibronectin, compared with matri-gel. As demonstrated in figure 6.12, cells cultured on fibronectin maintained a more uniform rounded shape and exhibited extensive and clearly-defined microtubule networks.  178  Figure 6.12: Immunofluorescence microscopy of primary rat hepatocytes cultured on matri-gel or fibronectin. Primary rat hepatocytes were isolated, and cultured overnight on cover slips coated with either a) fibronectin or b) matri-gel. Following an overnight incubation, the cells were fixed with formaldehyde and then probed to reveal: tubulin (green), ACC (red), and DNA (Hoescht, blue). Primary antibodies were used at a dilution of 1:30000 (ACC1), or 1:200 (tubulin). Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x). a)  b)  100 μm  50 μm  179  Cell fixing is another critical step in immunofluorescence analysis and several options were tested. A dilute solution of formaldehyde (3.7% v/v) had been recommended for use with the wide lens microscope while paraformaldehyde (PFA) solution (4% v/v) was recommended among standard operating procedures for use with the confocal microscope, particularly for the visualization of microtubules. As the hepatocytes were generally spherical in shape, a methanol: acetone (80% methanol, 20% acetone, v/v) fixing method was also tested because this was considered to be beneficial in reducing the thickness of the cell layer as a result of cell flattening. This was considered to be important because flattening of cells would allow for more effective confocal microscopy, especially facilitating pixel analysis to determine co-localization of proteins. As can be seen (figure 6.13), fixing primary hepatocytes with formaldehyde led to visualization of cells with the most clearly-defined microtubule structure. The use of PFA as the fixing method led to less well-defined hepatocytes, with relatively little clearly defined microtubule structures. The number of cells that survived fixation with paraformaldehyde was also appreciably smaller than with formaldehyde. Very few, if any, cells were visible on cover slips that had been fixed with methanol/acetone.  180  Figure 6.13: Comparison of fixation of primary hepatocytes with formaldehyde or paraformaldehyde. Primary rat hepatocyes were isolated, and cultured overnight on cover slips coated with fibronectin. Following an overnight incubation, the cells were fixed with either a) 3.7% formaldehyde (v/v) or b) 4% PFA (v/v) and then probed to reveal: tubulin (green), ACC (red), and DNA (Hoescht, blue). Primary antibodies were used at a dilution of 1:30000 (ACC1), or 1:200 (tubulin). Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x). a)  b)  100 μm  50 μm  181  To summarize the initial experiments in which primary rat hepatocytes were studied using confocal microscopy, several observations may be noted. Firstly, culture of cells on different extracellular matrices has a pronounced effect on the visualization of the microtubule cytoskeleton, with fibronectin being preferred over matri-gel. Secondly, cell fixation also affects the analysis and formaldehyde was preferred over PFA or methanol/acetone. Lastly, use of fibronectin and formaldehyde fixation gave relatively high density of cell recovery, uniform overall cell morphology and convincing staining of microtubules and punctate distribution of ACC. The use of the DNA Hoescht stain and the tubulin antibody had already been optimized by other research groups and found to be applicable to many cell types. The DNA Hoechst stain (figure 6.14) and the tubulin antibody (figure 6.15) were used at previously established concentrations with the primary rat hepatocytes to confirm the compatibility of these stains with the hepatocytes. Both stains were found to be effective with the hepatocytes. The tubulin in the primary hepatocytes was found to be either dispersed within the cell, or formed filamentous structures.  182  Figure 6.14: Nuclear staining of primary rat hepatocytes. Primary rat hepatocytes were isolated, cultured on cover slips coated with fibronectin overnight, fixed and stained with a DNA Hoechst stain at 1:50 dilution. Images were obtained using a) a Nikon Elipse E400 system, magnification 40x and b) Olympus FV1000 Confocal system, magnification 60x. a)  b)  183  Figure 6.15: Detection of tubulin in primary rat hepatocytes by immunofluorescence microscopy. Primary rat hepatocytes were isolated, cultured on cover slips coated with fibronectin overnight, fixed and stained with anti-tubulin antibody at 1:200 dilution and an antimouse secondary antibody conjugated to Alexa 488. Images were obtained using a) a Nikon Elipse E400 system, magnification 40x and b) Olympus FV1000 Confocal system, magnification 60x. a)  b)  184  To optimize the concentrations of anti-ACC antibodies, primary rat hepatocytes were prepared as described previously and stained with either anti-ACC1 antibody or an anti-phospho-ACC (ser-79) at fold dilutions of 50, 500, 1000, 5000, 10000, 30000 and 50000. The anti-ACC1 antibody (figure 6.16) and the anti-phospho-ACC antibody (figure 6.17) both gave clear staining with the primary hepatocytes, even at high dilutions, indicating good specificity. The experiments were repeated at least twice using different preparations of primary rat hepatocytes for each antibody with similar results.  185  Figure 6.16: Detection of ACC-1 in primary rat hepatocytes by immunofluorescence microscopy. Primary rat hepatocytes were isolated, cultured overnight on cover slips coated with fibronectin, fixed and stained with anti-ACC-1 followed by anti-rabbit secondary antibody conjugated to Alexa-568. Images were obtained using a Nikon Elipse E400 system (magnification 40 x) at (a) 1:10000 dilution or (b) 1:100000 dilution and c) an Olympus FV1000 Confocal system (magnification 60 x) at 1:10000 dilution. a)  b)  c)  50 μm  186  Figure 6.17: Detection of ser-79 phosphorylated ACC in primary rat hepatocytes by immunofluorescence. Primary rat hepatocytes were isolated, cultured overnight on cover slips coated with fibronectin, then fixed and stained with anti-phospho ACC antibody followed by antirabbit secondary antibody conjugated to Alexa-568. Images were obtained using a Nikon Elipse E400 system (magnification 40 x) at (a) 1:10 000 dilution or (b) 1:50 000 dilution and c) an Olympus FV1000 Confocal system (magnification 60 x) at 1:10 000 dilution. a)  b)  c)  50 μm  187  6.5.2 Localization of ACC in primary rat hepatocytes An unexpected and intriguing qualitative impression of ACC localization within primary rat hepatocytes was gained from these experiments. Specifically, the images of the primary rat hepatocytes stained with anti-ACC1 revealed very clear apparently punctate structures. This was unexpected because, as mentioned before, most classical analysis has led to the conclusion that ACC-1 is a cytosolic and soluble enzyme. While several studies have examined tissue sections to determine the cell types within which ACC is expressed, very few published reports have appeared in which attempts have been made to establish the subcellular localization of ACC. In one report, Abu-Elheiga et al used antibodies directed against ACC-2 in parallel with studies of the localization of a construct containing the N-terminal domain of ACC-2 linked to green fluorescent protein (GFP) [68]. In these studies, it was demonstrated that some ACC-2 localized to mitochondria in three different cell types: HepG2, T47D, and neonatal rat cardiomyocytes. Their images, as well as several others on corporate websites, using 3T3 cells and A431 breast carcinoma cells (figure 6.18a, b) show that both ACC-1 and ACC-2 are largely dispersed throughout the cell with no clear structural localization defined, except for the fraction of ACC-2 associated with the mitochondria [211-213]. However, the images shown in this thesis overwhelmingly indicate that ACC in primary rat heptatocytes and in H9c2 cells is not diffusely or uniformly distributed through the soluble phase of the cell but rather is best described as punctate, indicating that ACC is substantially localized within substructures that may reflect association with organelles and/or protein complexes (figure 6.18c). Interestingly, the hepatoma cell line, HepG2, also revealed at least some punctate pattern of staining for ACC-1, similar to that seen with primary hepatocytes (figure 6.17d). Nonetheless, primary hepatocytes and HepG2 cells do appear to differ in some respects in ACC immunostaining. For example, it appears that a greater proportion of the ACC staining is punctate in primary hepatocytes while images of HepG2 cells appear to show greater “background” of dispersed ACC staining. In any event, the overall impression for analysis of primary or derived liver cells as well as H9c2 myocytes is of some form of intracellular localization of ACC.  188  Figure 6.18: Contrasting distribution of ACC within cultured cell lines and in primary rat liver hepatocytes. Images (a) and (b) were taken from corporate sites [211, 212], in which the properties of commercial anti-ACC antibodies are documented. Images (c) and (d) were prepared with cells prepared as described previously using anti-ACC primary antibodies and images obtained using an Olympus FV1000 Confocal system at 60 x magnification. a)  b)  c)  d)  20 μm 50 μm  189  6.5.3 Hormonal influence on ACC localization and distribution in primary rat hepatocytes The next set of experiments was designed to test if changes in the physiological conditions of cells influenced ACC localization as detected by immunofluorescence microscopy using anti-phospho ACC antibodies. For these studies, primary hepatocytes were prepared and cultured overnight and then treated briefly prior to fixing and immunostaining. Hepatocytes were either fixed with no additional treatment (controls) or were incubated aerobically with 10 nM insulin for 10 minutes, or were incubated for 10 minutes in a low-oxygen atmosphere to induce “anaerobic stress” to activate AMPK, thereby promoting phosphorylation of ACC. Cells treated in these ways were fixed and probed with either anti-phospho ACC or anti-tubulin. It was anticipated that insulin and anaerobic stress would respectively decrease and increase ACC phosphorylation by AMPK. As shown in figure 6.19, the images obtained following immunostaining with anti-phospho ACC antibody indeed demonstrated anticipated changes in phosphorylation of ACC. After incubation with insulin, the intensity of the signal detected with antiphospho ACC was substantially lower than in the control, indicating significantly lower level of ACC phosphorylation. In contrast, following brief anaerobic stress, the signal intensity obtained following staining with anti-phospho ACC antibody was higher than in the control cells, indicating a higher degree of ACC phosphorylation. Interestingly, the punctate structures noted earlier become more apparent following anaerobic stress than in the control cells. In contrast to effects of cell treatment with anti-phospho ACC staining, the images obtained following staining with anti-tubulin did not change noticeably after cell treatment. This experiment demonstrates a number of points. First, the rat primary hepatocytes are sensitive to the use of hormones, or stress, which can be detected even at high dilutions of the antibodies. Second, the ACC punctate structures appear to be present regardless of changes in the levels of ACC phosphorylation, indicating that these ACC-containing structures are rather robust. Third, tubulin was not affected by different levels of ACC phosphorylation, demonstrating that changes in ACC phosphorylation are not associated with gross changes in microtubule structure.  190  Figure 6.19: Effect of acute insulin and anoxic stress on ACC ser-79/ser-212 phosphorylation in primary rat hepatocytes. Following isolation and overnight culture on cover slips, primary rat hepatocytes were incubated for 10 minutes with 10 nM insulin (a, b), no hormone (control – c, d) or at low oxygen tension (anaerobic stress – e, f). Following fixing, the cells were stained with either an anti-phospho ACC antibody (1:30000 dilution) (a, c, e) or an anti-tubulin antibody (1:200 dilution) (b, d, f). Images were obtained using a Nikon Elipse E400 system (magnification 40x) using identical exposure times of 400 msec (anti-phospho ACC) or 30 msec (anti-tubulin) to allow direct comparison of treatments. Results are typical of three independent experiments. anti-phospho ACC  anti-tubulin b)  c)  d)  e)  f)  ACC phosphorylation  ACC dephosphorylation  a)  191  To further explore the effects of insulin and anaerobic stress, the studies using conventional immunofluoresence microscopy were extended by using confocal microscopy. In these studies, cells were visualized, using anti-tubulin, anti-ACC and anti-phospho ACC antibodies, following treatments with insulin or anaerobic stress as before, treatments that are known to respectively enhance or suppress ACC activity and fatty acid synthesis. As with conventional immunofluoresence, there was no difference in overall intensity of tubulin staining following insulin treatment or anaerobic stress (figure 6.20), confirming the results shown in figure 6.19. It is possible that changes in microtubule structural organization do occur but this would require more detailed and sophisticated image analysis. As noted before, insulin treatment again led to a reduction in intensity of signal detected with anti-phospho ACC (figure 6.21b, d, f) , relative to control cells with no corresponding change in signal with anti-ACC (figure 6.21a, c, e). Results observed following brief incubation under anaerobic conditions were less dramatic than observed by using a conventional fluorescence microscope, but still revealed an increase in the number of punctate structures per cell. Upon closer examination of insulin-treated hepatocytes stained with anti-phospho ACC antibody, the ACC punctate structures appear to localize mainly near the periphery of the cells and very few punctate structures appear in the peri-nuclear regions (figure 6.21b). Conversely, following anaerobic stress, the ACC punctate structures appear to be more evenly distributed throughout the cell volume (figure 6.21f). These qualitative observations underline the importance of developing approaches to examine and quantify images in more detail because they may be indicative of some sort of ACC localization that is dependent on enzyme activation and may have functional significance.  192  Figure 6.20: Effect of insulin or anaerobic stress on microtubule structure in primary rat hepatocytes. Primary rat hepatocytes were isolated and cultured overnight on cover slips coated with fibronectin. Following overnight incubation, the cells were incubated with 80 nM insulin for 10 minutes (a, b), untreated (control – c, d), or subjected to anerobic stress (e, f). Cells were fixed and then probed with anti-tubulin (1:200 dilution) followed by antimouse secondary conjugated to Alexa-468. Each image highlights the same primary hepatocyte shown in the following figure 6.21. Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x). b)  ACC dephosphorylation  a)  50 μm  d)  e)  f)  ACC phosphorylation  c)  193  ACC dephosphorylation  Figure 6.21: Effect of insulin or anaerobic stress on ACC localization in primary rat hepatocytes. Primary rat hepatocytes were isolated and cultured overnight on cover slips coated with fibronectin. Following overnight incubation, the cells were incubated with 80 nM insulin for 10 minutes (a, b), untreated (control – c, d), or subjected to anaerobic stress (e, f). Cells were fixed and then probed with either anti-ACC1 antibody (a, c, e), or antiphospho ACC antibody (b, d, f) (both at 1:30000 dilution), followed by anti-rabbit secondary conjugated to Alexa-568 (red). Each image highlights the same corresponding primary hepatocyte shown in figure 6.20. Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x). anti-ACC anti-phospho ACC a) b)  50 μm  d)  e)  f)  ACC phosphorylation  c)  194  To confirm the observations made from figure 6.21, multiple images were compiled from three separate experiments to examine the ACC localization and distribution following insulin treatment or anaerobic stress to respectively activate or inhibit ACC activity. A selection of the multiple images is shown in figure 6.22 for antiACC1 staining, and figure 6.23 for anti-phospho ACC staining.  195  Figure 6.22: Multiple images of the effects of insulin or anaerobic stress on ACC localization. Primary rat hepatocytes were isolated and cultured overnight on cover slips coated with fibronectin. Following overnight incubation, the cells were incubated with 80 nM insulin for 10 minutes (a), untreated (control – b), or subjected to anerobic stress (c). Cells were fixed and then probed with anti-ACC1 anbitody (1:30000 dilution), followed by antirabbit secondary conjugated to Alexa-568 (red). Each image highlights one primary hepatocyte. Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x). a) Insulin incubation b) Control c) “Stressed” condition  50 μm  50 μm  50 μm  196  Figure 6.23: Multiple images of the effects of insulin or anaerobic stress on ACC localization. Primary rat hepatocytes were isolated and cultured overnight on cover slips coated with fibronectin. Following overnight incubation, the cells were incubated with 80 nM insulin for 10 minutes (a), untreated (control – b), or subjected to anerobic stress (c). Cells were fixed and then probed with anti-phospho ACC antibody (1:30000 dilution), followed by anti-rabbit secondary conjugated to Alexa-568 (red). Each image highlights one primary hepatocyte. Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x). a) Insulin incubation b) Control c) “Stressed” condition  50 μm  50 μm  50 μm  197  To quantify the results from the images in figures 6.22 and 6.23, the size and number of punctate structures were measured and compared for hepatocytes incubated under control conditions or following insulin or anaerobic stress. The approximate length of an ACC punctate structure was determined based on an internally calibrated function of the Olympus viewer program FV10-ASW 1.6 Viewer. From each replicate image, ten ACC punctate structures were randomly chosen and the size of each structure was measured. The results are tabulated in table 6.4. Based on the assumption that ACC undergoes polymerization coincident with activation, it was initially anticipated that the activation of ACC (insulin incubation) might conceivably lead to a visibly larger ACC structure, while the suppression of ACC activity (anaerobic stress) might lead to a corresponding smaller ACC structure. In general, there were no significant changes in sizes of the structures stained with anti-ACC antibodies detectable under the different conditions.  Table 6.4: The dimensions of structures in rat primary hepatocytes visualized with antiACC antibodies. The size of ten ACC punctate structures was determined in each of several separate cells in each condition and the average size calculated. Values are given as mean ± SEM (n = 10). Treatment of hepatocytes Insulin (10 nM for 10 min) Control Anaerobic stress (10 min)  anti-ACC 0.98 ± 0.12 μm 0.93 ± 0.11 μm 0.91 ± 0.09 μm  anti-phospho ACC 0.87 ± 0.08 μm 0.92 ± 0.13 μm 0.94 ± 0.14 μm  The number of ACC punctate structures was estimated by manually counting structures with a given microscopic field under each prescribed conditions. The results are tabulated in table 6.5.  198  Table 6.5: The number of structures in rat primary hepatocytes visualized with anti-ACC antibodies. The number of ACC structures was estimated from five cells in each of five different field views in each of the examined conditions and the results presented as the mean ± SEM (n = 25). Treatment of hepatocytes Insulin (10 nM for 10 min) Control Anaerobic stress (10 min)  anti-ACC 112 ± 7 88 ± 4 146 ± 5  anti-phospho ACC 77 ± 6 99 ± 5 112 ± 8  The average number of ACC structures per unit area detected with the anti-ACC antibodies was significantly different in insulin-treated cells and in oxygen-deprived cells than in the corresponding control. Since ACC expression or degradation is unlikely to change appreciably in such short incubations, the changes suggest some form of ACC redistribution following cell treatments. In hepatocytes stained with anti-phospho ACC antibody, the insulin treatment led to a decline in the number of detected structures while anoxic incubation led to an increase in number, results that were consistent with the earlier analysis of overall cell staining with anti-phospho ACC antibodies. These results are consistent with stimulation of AMPK-dependent phosphorylation of ACC in anoxic stress and an increase in ACC structures that are dephosphorylated at AMPK-dependent sites following insulin treatment. An important caveat to note is that this analysis was carried out with a manual counting procedure and it will be important to repeat and consolidate on these observations with further automated analysis once the techniques have been established for more comprehensive image analysis. Even so, by using repeated observations on several independent cell preparations reasonable statistical confidence was obtained and the indicated values in Table 6.5 were all statistically different from controls (P < 0.05). The challenge now will be to adapt this approach so that a more detailed analysis and quantification can be obtained by confocal microscopy. At present, this has been precluded because of the size and depth of primary hepatocytes.  199  6.5.4 Relative localization of ACC and tubulin in primary rat hepatocytes To further explore the possible interactions between ACC and tubulin, confocal microscopy was used in an attempt to examine the possible relationship between intracellular organization of tubulin and ACC in primary rat hepatocytes. As hinted in previous studies and mentioned here previously, it is possible that microtubules could act as a scaffold for ACC, thereby providing a mechanism for localization in different parts of the cell, and perhaps even a method to control ACC, for example, in response to hormones, cell stresses, and/or changes in nutrient supply. The first potentially significant observation arose by chance during early attempts to optimize conditions for the visualization of ACC in intact cells. In initial experiments involving the culture of primary rat hepatocytes on cover slips coated with matri-gel, probing with anti-tubulin antibodies gave very diffuse staining, with little evidence for organized microtubule structures. Eventually, the use of fibronectin, rather than matrigel, to coat the cover slips was key in achieving consistent microtubule cytoskeletal organization (figure 6.23). As mentioned previously, work of others had shown that fibronectin was beneficial for the study of the microtubule network of cells and this proved to be true also for primary hepatocytes. In retrospect, the variable results in microtubule organization provided a fascinating insight into ACC organization as well because the lack of organization of tubulin into a defined microtubule network was also associated with a substantial dispersal of ACC throughout the cytosol (figure 6.24). In contrast, when conditions were optimized to allow microtubule cytoskeletal organization, ACC distribution was also markedly different and appeared in distinct punctate patterns mentioned above. These contrasting results of distribution of ACC and tubulin were observed in three independent experiments in which the appearance of ACC punctate structures was seldom observed in the absence of microtubule filaments. Although the presence of ACC organization into punctate forms appears to be highly correlated with the presence of microtubules, the spatial distribution of microtubules and ACC is evidently not closely aligned. Immunostaining and the merging of images does not reveal a close juxtaposition of the two fluorescent labels. However, when the microtubule structure of primary hepatocytes is substantially disrupted by culture on matri-gel rather than on fibronectin, ACC  200  distribution also becomes substantially dispersed throughout the cytosolic volume and, intriguingly, substantially overlaps with the distribution of free, non polymerized tubulin. It therefore appears that ACC may either associate with free tubulin and or with another microtubule-associated structure. Alternatively, microtubules may play an indirect role in the organization and localization of ACC into certain parts of the cell perhaps involving intracellular traffic. Undoubtedly, the staining pattern of ACC is far more distinct when microtubule structure is also well defined.  201  Figure 6.24: Cells with well-defined microtubule structures exhibit punctate localization of ACC. Primary rat hepatocytes were isolated and cultured overnight on cover slips coated with fibronectin. Cells were fixed with formaldehyde and then stained with anti-tubulin antibody (1:200 dilution) and anti-ACC-1 antibody (1:30000 dilution), followed by appropriate secondary antibodies conjugated with fluorescent dyes. Panels show (a) tubulin, green (b) ACC, red or (c) merged images including nuclei stained with Hoescht stain. Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x) and are typical of many fields observed in each of four independent experiments. a) b)  c)  100 μm  202  Figure 6.25: Cells cultured on matri-gel lack microtubular structure and exhibit substantial cytosolic dispersion of ACC. Primary rat hepatocytes were isolated and cultured overnight on cover slips coated with matri-gel. Cells were fixed with formaldehyde and then stained with anti-tubulin antibody (1:200 dilution) and anti-ACC-1 antibody (1:30000 dilution), followed by appropriate secondary antibodies conjugated with fluorescent dyes. Panels show (a) tubulin, green (b) ACC, red or (c) merged images including nuclei stained with Hoescht stain. Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x) and are typical of many fields observed in each of four independent experiments. a)  b)  c) 100 μm  50 μm  203  To further explore the impact of microtubule structure on ACC organization and localization, the microtubule-disrupting agent, colchicine, was used (figure 6.25). Sustained incubation of hepatocytes with colchicine led to substantial loss of cells, presumable due to cytoxicity leading to cell death. Incubation of cells with colchicine were therefore kept at a minimum, but even so, cell responses to colchicine were far from uniform and some cells remained unaffected, while other cells exhibited a diffuse tubulin pattern in which the filamentous microtubule structure was indeed disrupted. The ACC in cells in which microtubule structures were clearly and substantially disrupted by colchicine, remained somewhat punctate, but also a higher proportion of ACC became dispersed. Previously, it has been demonstrated by Buechler et al that colchicine affects ACC activity as well as decreasing ACC polymerization [208], but I was unable to confirm these results. Currently, it remains unclear whether colchicine affects ACC directly, or if ACC is affected due to the disassembly of the microtubules. It is recognized that higher resolution techniques, including electron microscopy, will be needed to more accurately define the localization of ACC, nevertheless the experiments described here provide an important framework for further work. One important limitation has been that the primary rat hepatocytes remain relatively spherical during culture and do not lay in a sufficiently flattened form on the culture surface for optimal confocal microscopy. Although several attempts were made, the available imaging system could not be readily adapted to accommodate the primary hepatocytes in such a way as to allow adequate pixel analysis. For this reason, other cultured cell lines will have to be tested to facilitate more informative image analysis. To this end, attempts have been made to explore the use of HepG2 and H4 hepatoma cell lines, although neither has so far proved to be convenient in that they very rapidly overgrow to multiple cell depths and maintaining a convenient monolayer proved difficult.  204  Figure 6.26: Effects of colchicine on microtubule organization and ACC distribution in primary rat hepatocytes. Primary rat hepatocytes were isolated and cultured overnight on cover slips coated with fibronectin. Cells were then untreated (a-c) or treated with 1 mM colchicine for 10 minutes (d-f) prior to fixing with formaldehyde. Cells were then stained with anti-tubulin antibody (1:200 dilution) and anti-ACC1 antibody (1:30000 dilution), followed by appropriate secondary antibodies conjugated with fluorescent dyes. Panels show (a, d) tubulin, green (b, e) ACC, red or (c, f) merged images including nuclei stained with Hoescht stain. Images were obtained using an Olympus FV1000 Confocal system (magnification 60 x) and are typical of many fields observed in each of four independent experiments. a) d)  b)  e)  c)  f)  50 μm  50 μm  205  6.6 Summary Studies of ACC polymerization by size-exclusion chromatography led to the identification of tubulin as one of the few proteins that consistently associated with ACC polymers. Furthermore, tubulin co-purifies with ACC even through avidin-agarose affinity chromatography in the presence of 0.5 M KCl. The significance of the interaction between ACC and tubulin was therefore tested by additional approaches including western blotting, co-immunoprecipitation and testing the effects of anti-tubulin antibodies and microtubule-specific reagents. Western blotting showed that tubulin, like ACC, exhibited a citrate-dependent shift in molecular size when subjected to BioGelA50M size exclusion chromatography. Indeed, a significant proportion of the residual tubulin present in those partially-purified rat liver extracts was associated with ACC polymeric fractions. In the co-immunoprecipitation studies, it was demonstrated that tubulin was present in the complexes recovered with anti-ACC antibodies immobilized on protein-G agarose. Conversely, ACC was present in complexes recovered with immobilized anti-tubulin antibodies. These studies all provide evidence that ACC and tubulin are closely associated in rat liver extracts. The effects of colchicine and GTP, known modulators of microtubules, were examined on ACC purified through ammonium sulfate precipitation. In terms of ACC activity observed with maximal citrate activation, colchicine had no significant effect on partially-purified ACC but induced a modest activation of highly-purified preparations, both directly and in combination with purified tubulin. GTP had a direct and dosedependent inhibitory effect on partially-purified ACC but inconsistent effects on highlypurified ACC, contrary to the results found by Buechler et al. The addition of tubulin had no effect on partially-purified ACC but induced modest activation of highly-purified ACC, an effect that was attenuated by colchicine. Overall, the results with microtubule reagents GTP and colchicine do not lead to very clear-cut conclusions but are perhaps most consistent with a modest positive effect of non-polymerized tubulin on ACC function. Immunofluorescence microscopy in primary rat hepatocytes was used to further explore the interaction between ACC and tubulin. As this technique had not previously been used to assess ACC localization within cells, optimized methods had to be  206  developed and tested. Basic parameters that were tested included the nature of the extracellular matrix to which the cells were allowed to adhere in culture, the fixing method and the nature and concentrations of antibodies. Based on these studies, fibronectin was the favored extracellular matrix and formaldehyde the most effective fixing reagent. These approaches led to reliable staining of primary rat hepatocytes to visualize tubulin, ACC and ser-79 phospho-ACC. The use of immunofluorescence microscopy led to new insights into the organization of ACC in primary rat hepatocytes, the most critical general observation being that ACC is substantially organized in cellular structures that appear punctate by fluorescence microscopy. This distribution pattern is in sharp contrast to a general distribution throughout the cytosol that had been anticipated. The nature of the structures to which ACC is associated remains to be defined, although rough estimates of the size of the structures places them in the range of one micron. Interestingly, this is somewhat above the size range of fully extended ACC polymers – previously estimated in vitro to be on the order 0.1 – 0.5 microns – and instead approaches the scale of mitochondria. The punctate staining of ACC did not show a clear correlation with or direct overlap with that of microtubules and did not change dramatically in size or distribution when cells were stimulated with insulin (to activate ACC and fatty acid synthesis) or when cells were exposed to low oxygen tension (to induce activation of AMPK, ACC phosphorylation and de-activation). From a qualitative perspective it appears there may be some re-distribution of ACC structures following cell treatments, for example with greater or lesser amounts appearing in the cell periphery relative to the peri-nuclear regions. This phenomenon was perhaps most obvious when the microtubule organization was disrupted with colchicine. However, as noted elsewhere, a more sophisticated approach to image analysis will be required to fully explore this fascinating area and those approaches were not technically possible. Intriguingly, a dramatic result was obtained inadvertently as a result of the preliminary studies to compare different extracellular matrices. Specifically, allowing cells to adhere to matri-gel in culture led to almost complete disruption of the microtubular cytoskeleton and this was associated with an almost complete loss of ACC punctate organization and corresponding dispersal through the cytosol. Moreover, merging of images of ACC and tubulin localization in this study gave the most clear-cut evidence for “co-localization”,  207  both tubulin and ACC being substantially dispersed through the cytosolic volume in cells cultured on matri-gel. Looking ahead, the studies performed in this thesis can be used as a basis to examine ACC cellular localization with respect to various organelles and under a variety of stimulatory or stress conditions. The possibility of distinct cellular localization and responsiveness to external stimuli of the distinct ACC isoforms will also be important to examine. So far, very little is known about ACC localization in any cell type and while it is established that ACC is largely soluble or “cytosolic” following cell disruption, the limited evidence for intracellular localization is inconsistent. One group has reported that ACC-2 localizes to the cytoplasmic surface of the mitochondrial membrane [68] while yeast ACC has been found to associate to the cytoplasmic surface of the endoplasmic reticulum [214]. In another study of rat tissues, ACC was found to not associate with any classical subcellular organelles [209] but rather it was concluded that ACC associated with cytoskeletal elements, in particular with microtubules. Another indirect but potentially interesting link to the cytoskeleton emerged from studies of CPT-I by Geelen et al [215]. These authors suggested that CPT-I could be regulated by association with cytoskeletal elements, and could be conversely activated upon disruption of the cytoskeletal elements induced by phosphorylation by calcium/calmodulin-dependent protein kinase II [216]. While this is very indirect, the potent role of malonyl-CoA in the control of CPT-I does hint that close association between ACC and CPT-I might be important functionally and perhaps this may be achieved in the context of mutual association with cytoskeletal elements. Taken together, the evidence obtained so far indicates that ACC and tubulin associate very closely, perhaps directly, and that this association may play a significant role in the distribution and organization of ACC structures within hepatocytes.  208  Chapter 7: Conclusion and future experiments  Although citrate is the best characterized of the allosteric ligands of ACC, the corresponding binding site on the enzyme is still unknown. The work described in the first part of this thesis stemmed from previous studies in which pyridoxal phosphate (PLP) had been used as a structural analog of citrate to elucidate the citrate-binding site of phsophofructokinase-1. The possibility that PLP might similarly be used to probe the citrate-binding site of ACC was initiated by a former graduate student and my work continued this investigation with detailed kinetic studies and covalent labeling. The results of these kinetic analyses showed that PLP is a potent non-competitive or mixed inhibitor of ACC with respect to all three major substrates ATP, bicarbonate and acetylCoA. Coupled with the ability of citrate to provide some protection from PLP-induced inhibition, these results are consistent with the hypothesis that ACC inhibition by PLP is mediated by binding to a site distinct from either of the two ACC active sites and therefore that PLP may bind to the citrate-binding site. Methods were developed to covalently attach PLP to ACC via Schiff base reduction with sodium borohydride. The use of [3H]-borohydride led to somewhat sub-stoichiometric labeling of ACC, indicating reaction of PLP with only a very limited number of the more than one hundred lysyl residues in the ACC primary sequence. Further work is in progress aimed at isolation of the PLP-labeled ACC peptides using IMAC chromatography, with the objective of defining the primary sequence around the PLP-bound lysyl residue. Other methods to determine the location of citrate binding are also being developed, in particular, a “footprinting” method in which it is hoped that the binding of PLP will protect a specific trypsin cleavage site and thereby lead to a modified trypsin cleavage pattern. The identification of the citrate-binding site is clearly crucial in understanding the mechanism of citrate activation of ACC. The aim of the work described in the balance of the thesis had been to test the hypothesis that ACC associates with other cellular proteins in a specific and functional manner. This aim arose from previous work that had provided preliminary characterization of an “ACC regulator” protein from rat liver and from the growing appreciation of the role of macromolecular protein complexes coupled with specific and  209  genome-wide studies providing evidence for the involvement of ACC in various proteinprotein interactions. With this background, I set out to test the specific hypothesis that specific proteins might associate with ACC polymers. This idea was attractive because ACC undergoes an essentially unique citrate-dependent change in molecular structure that provides a solid basis for isolation. Accordingly, I established sucrose gradient centrifugation and size exclusion techniques to separate large polymeric forms of ACC from smaller oligomeric and dimeric forms of the enzyme. I found that size exclusion chromatography provided the best degree of separation and this technique was then used more extensively. Following citrate-induced activation and polymerization, ACCcontaining chromatographic fractions were subjected to trypsin digestion and tandem mass spectrometry of peptides. This approach was carried out with multiple preparations of ACC from rat liver and rat white adipose tissue. In all cases, two proteins were consistently found to be associated with ACC polymers – fatty acid synthase and tubulin. The interactions of these two proteins with ACC were therefore explored in more detail. In the case of fatty acid synthase the results obtained by mass spectrometry were confirmed by Western blotting of column fractions, showing a substantive citrate-induced association of the two proteins. Results of co-immunoprecipitation studies were not convincing and no effect of FASN or of anti-FASN antibodies on ACC activity were detected. Further work to explore more subtle effects on ACC activity or on ACC phosphorylation are probably merited, but I decided to direct more focus to the interactions between ACC and tubulin. As with the ACC-FASN interactions, the results of mass spectrometry analysis were confirmed by Western blotting. In these studies it was shown that a substantial proportion of the tubulin that had co-purified with ACC through ammonium sulfate precipitation subsequently co-eluted with ACC during size exclusion chromatography in the presence of citrate. The association between ACC and tubulin was also confirmed by co-immunoprecipitation using both antibodies, anti-ACC antibodies being able to coimmunoprecipitate tubulin while anti-tubulin antibodies led to co-immunoprecitation of ACC. Furthermore, tubulin was found to be associated with ACC even after purification to near-homogeneity by avidin affinity chromatography in the presence of 0.25 to 0.5M salt. The activity of highly purified ACC was slightly activated by the addition of  210  purified tubulin, while less enriched ACC preparations were unaffected by tubulin. It is possible that the less purified ACC still contained sufficient endogenous tubulin to exert a maximal effect on activity. The addition of GTP had little or no effect on highly purified ACC preparations but induced a modest dose-dependent inhibition of less enriched preparations. This result is consistent with the possibility that endogenous tubulin or other GTP-dependent proteins mediated the effect of GTP on ACC. Immunofluorescence microscopy of primary rat hepatocytes was undertaken to extend the study of ACC-tubulin interactions to the whole cell level and led to unexpected general insights into ACC organization. The presence of punctate structures in primary rat hepatocytes stained with anti-ACC and anti-phospho ACC antibodies was unexpected and quite inconsistent with the idea that ACC is distributed throughout the cytosolic compartment of the cell. The presence of the ACC-containing punctate structures appeared to be dependent on the presence of defined microtubule structures within the cell. Most dramatically, the loss of microtubule organization following culture of primary hepatocytes on matri-gel led to a complete loss of punctate organization of ACC and complete dispersal throughout the cytosol, overlapping in space with staining of free tubulin. The effects of insulin incubation (to activate ACC) and anaerobic stress (to inactivate ACC) did not lead to any noticeable change in ACC localization in comparison to microtubule organization. However, it would be worth examining the effects of these conditions in ACC localization in comparison to other structures within the cell. For example, it is important to assess the extent to which ACC might co-localize with mitochondrial or other organelle markers. Efforts to quantify various parameters of ACC localization within primary hepatocytes by confocal microscopy were unsuccessful, largely due to the physical characteristics of the primary rat hepatocytes. In particular, the primary cells failed to adhere sufficiently closely to the culture surface for optimal imaging, presumably retaining a natural rounded shape and considerable depth above the culture surface. It will be important to explore alternative culture conditions or perhaps alternative cell types to tackle this problem of quantitative image analysis. On another level, it will be important to obtain higher resolution images using electron microscopy to determine the real proximity of ACC to potential binding partners. It is worth noting that the possibility  211  of expressing fluorescently tagged forms of intact full-length ACC isoforms remains technically and economically challenging.  212  References  1.  2.  3. 4.  5.  6. 7.  8.  9.  10.  11.  12.  13.  14.  15.  Pacheco-Alvarez, D., R.S. Solorzano-Vargas, and A. Leon Del Rio, Biotin in Metabolism and Its relationship to human disease. Archives of Medical Research, 2002. 33: 439-447. Easterbrook-Smith, S.B., J.C. Wallace, and D.B. Keech, A reappraisal of the reaction pathway of pyruvate carboxylase. Biochemical Journal, 1978. 169: 225228. Wakil, S.J., J.W. Porter, and D.M. Gibson, Studies on the mechanism of fatty acid synthesis. Biochimica et Biophysica Acta, 1957. 24: 453-455. Porter, J.W., S.J. Wakil, A. Tietz, M.I. Jacob, and D.M. Gibson, Studies on the mechanism of fatty acid synthesis. II. Cofactor requirements of the soluble pigeon liver system. Biochimica et Biophysica Acta, 1957. 25: 35-49. Wakil, S.J., E.B. Titchener, and D.M. Gibson, Evidence for the participation of biotin in the enzymic synthesis of fatty acids. Biochimica et Biophysica Acta, 1958. 29(1): 225-226. Wakil, S.J., A malonic acid derivative as an intermediate in fatty acid synthesis. Journal of the American Chemical Society, 1958. 80: 6465. Gibson, D.M., E.B. Titchener, and S.J. Wakil, Studies on the mechanism of fatty acid synthesis. V. Bicarbonate requirement for the synthesis of long-chain fatty acids. Biochimica et Biophysica Acta, 1958. 30(2): 376-383. Bai, D.H., T.-W. Moon, F. Lopez-Casillas, P.C. Andrews, and K.H. Kim, Analysis of the biotin-binding site on acetyl-CoA carboxylase from rat. European Journal of Biochemistry, 1989. 182(2): 239-245. Wakil, S.J. and D.M. Gibson, Studies on the mechanism of fatty acid synthesis. VIII. The participation of protein-bound biotin in the biosynthesis of fatty acids. Biochimica et Biophysica Acta, 1960. 41: 122-129. Kallen, R.G. and J.M. Lowenstein, The stimulation of fatty acid synthesis by isocitrate and malonate. Archives of Biochemistry and Biophysics, 1962. 96: 188-190. Vagelos, P.R., A.W. Alberts, and D.B. Martin, Studies on the mechanism of activation of acetyl coenzyme A carboxylase by citrate. Journal of Biological Chemistry, 1963. 238(2): 533-540. Waite, M. and S.J. Wakil, Studies on the mechanism of fatty acid synthesis. XII. Acetyl coenzyme A carboxylase. Journal of Biological Chemistry, 1962. 237(9): 2750-2757. Smith, S., D.J. Easter, and R. Dils, Fatty acid biosynthesis. III. Intracellular site of enzymes in lactating-rabbit mammary gland. Biochimica et Biophysica Acta, 1966. 125(3): 445-455. Brownsey, R.W., R. Zhande, and A.N. Boone, Isoforms of acetyl-CoA carboxylase: structures, regulatory properties and metabolic functions. Biochemical Society Transactions, 1997. 25(4): 1232-1238. Tehlivets, O., K. Scheuringer, and S.D. Kohlwein, Fatty acid synthesis and elongation in yeast. Biochimica et Biophysica Acta, 2007. 1771(3): 255-270.  213  16. 17.  18.  19.  20. 21. 22. 23.  24.  25. 26.  27. 28.  29.  30.  31.  Cronan, J.E. and G.L. Waldrop, Multi-subunit acetyl-CoA carboxylases. Progress in Lipid Research, 2002. 41: 407-435. Guchhait, R.B., J. Moss, W. Sokolski, and M.D. Lane, The carboxyl transferase component of acetyl CoA carboxylase: Structural evidence for intersubunit translocation of the biotin prosthetic group. Proceedings of the National Academy of Science USA, 1971. 68(3): 653-657. Gregolin, C., E. Ryder, R.C. Warner, A.K. Kleinschmidt, H.C. Chang, and M.D. Lane, Liver acetyl coenzyme A carboxylase. II. Further molecular characterization. Journal of Biological Chemistry, 1968. 243(16): 4236-4245. Dyck, J.R., L.G. Berthiaume, P.D. Thomas, P.F. Kantor, A.K. Barr, D. Singh, T.A. Hopkins, N. Voilley, M. Prentki, and G.D. Lopaschuk, Characterization of rat liver malonyl-CoA decarboxylase and the study of its role in regulating fatty acid metabolism. Biochemical Journal, 2000. 350(Pt 2): 599-608. Saggerson, D., Malonyl-CoA, a key signaling molecule in mammalian cells. Annual Review of Nutrition, 2008. 28: 253-272. Lane, M.D., Acetyl coenzyme A carboxylase. Current topics in cellular regulation, 1974. 8(0): 139-195. Chapman-Smith, A. and J.E. Cronan, Molecular biology of biotin attachment to proteins. The Journal of Nutrition, 1999. 129(2): 477S-484S. Shriver, B.J., C. Roman-Shriver, and J.B. Allred, Depletion and repletion of biotinyl enzymes in liver of biotin-deficient rats: evidence of a biotin storage system. The Journal of Nutrition, 1993. 123: 1140-1149. Jacobs, R., E. Kilburn, and P.W. Majerus, Acetyl Coenzyme A Carboxylase: The effects of biotin deficiency on enzyme in rat liver and adipose tissue. Journal of Biological Chemistry, 1970. 245(23): 6462-6467. Zempleni, J. and D.M. Mock, Biotin biochemistry and human requirements. Journal of Nutritional Biochemistry, 1999. 10(3): 128-138. St. Maurice, M., L. Reinhardt, K.H. Surinya, P.V. Attwood, J.C. Wallace, W.W. Cleland, and I. Rayment, Domain architecture of pyruvate carboxylase, a biotindependent multifunctional enzyme. Science, 2007. 317(5841): 1076-1079. Knowles, J.R., The mechanism of biotin-dependent enzymes. Annual Review of Biochemistry, 1989. 58: 195-221. McClure, W.R., H.A. Lardy, M. Wagner, and W.W. Cleland, Rat liver pyruvate carboxylase. II. Kinetic studies of the forward reaction. Journal of Biological Chemistry, 1971. 246(11): 3579-3583. Guchhait, R.B., S.E. Polakis, D. Hollis, C. Fenselau, and M.D. Lane, Acetyl coenzyme A carboxylase system of Escherichia coli. Site of carboxylation of biotin and enzymatic reactivity of 1'N-(ureido)-carboxybiotin derivatives. Journal of Biological Chemistry, 1974. 249(29): 6646-6656. Blanchard, C.Z., D. Amspacher, R. Strongin, and G.L. Waldrop, Inhibition of Biotin carboxylase by a reaction intermediate analog: implications for the kinetic mechanism. Biochemical and Biophysical Research Communications, 1999. 266(2): 466-471. Blanchard, C.Z. and G.L. Waldrop, Overexpression and kinetic characterization of the carboxyltransferase component of acetyl-CoA carboxylase. Journal of Biological Chemistry, 1998. 273(30): 19140-19145.  214  32.  33.  34.  35.  36.  37.  38.  39.  40.  41.  42.  43.  44.  45.  Hardie, D.G. and P. Cohen, Purification and physicochemical properties of fatty acid synthetase and acetyl-CoA carboxylase from lactating rabbit mammary gland. European Journal of Biochemistry, 1978. 92(1): 25-34. Hashimoto, T. and S. Numa, Kinetic studies on the reaction mechanism and the citrate activation of liver acetyl coenzyme A carboxylase. European Journal of Biochemistry, 1971. 18: 319-331. Ashman, L.K. and D.B. Keech, Sheep kidney pyruvate carboxylase. Studies on the coupling of adenosine triphosphate hydrolysis and CO2 fixation. Journal of Biological Chemistry, 1975. 250(1): 14-21. Barden, R.E., C.H. Fung, M.F. Utter, and M.C. Scrutton, Pyruvate carboxylase from chicken liver. Steady state kinetic studies indicate a "two site" ping pong mechanism. Journal of Biological Chemistry, 1972. 247(4): 1323-1333. Warren, G.B. and K.F. Tipton, The role of acetyl-CoA in the reaction pathway of pig-liver pyruvate carboxylase. European Journal of Biochemistry, 1974. 47(3): 549-554. Abu-Elheiga, L., M.M. Matzuk, P. Kordani, W. Oh, T. Shaikenov, Z. Gu, and S.J. Wakil, Mutant mice lacking acetyl-CoA carboxylase 1 are embryonically lethal. Proceedings of the National Academy of Science USA, 2005. 102(34): 1201112016. Abu-Elheiga, L., M.M. Matzuk, K.A.H. Abo-Hashema, and S.J. Wakil, Continuous fatty acid oxidation and reduced fat storage in mice lacking acetylCoA carboxylase 2. Science, 2001. 291(March 30): 2613-2616. Hasslacher, M., A.S. Ivessa, F. Palauf, and S.D. Kohlwein, Acetyl-CoA carboxylase from yeast is an essential enzyme and is regulated by factors that control phospholipid metabolism. Journal of Biological Chemistry, 1993. 268(15): 10946-10952. Choi-Rhee, E. and J.E. Cronan, The biotin carboxylase-biotin carboxyl carrier protein complex of Escherichia coli Acetyl-CoA carboxylase. Journal of Biological Chemistry, 2003. 278(33): 30806-30812. Han, L., K. Yang, K. Kulowski, E. Wendt-Pienkowski, C.R. Hutchinson, and L.C. Vining, An acyl-coenzyme A carboxylase encoding gene associated with jadomycin biosynthsis in Streptomyces venezuelae ISP5230. Microbiology, 2000. 146: 903-910. Gande, R., L.G. Dover, K. Krumbach, G.S. Besra, H. Sahm, T. Oikawa, and L. Eggeling, The two carboxylases of Corynebacterium glutamicum essential for fatty acid and mycolic acid synthesis. Journal of Bacteriology, 2007. 189(14): 5257-5264. Oh, T.J., J. Daniel, H.J. Kim, T.D. Sirakova, and P.E. Kolattukudy, Identification and characterization of Rv3281 as a novel subunit of a biotin-dependent acylCoA carboxylase in Mycobacterium tuberculosis H37Rv. Journal of Biological Chemistry, 2006. 281(7): 3899-3908. Witters, L.A. and T.D. Watts, Yeast acetyl-CoA carboxylase: In vitro phosphorylation by mammalian and yeast protein kinases. Biochemical and Biophysical Research Communications, 1990. 169(2): 369-376. Cho, Y.S., J.I. Lee, D. Shin, H.T. Kim, Y.H. Cheon, C.I. Seo, Y.E. Kim, Y.L. Hyun, Y.S. Lee, K. Sugiyama, S.Y. Park, S. Ro, J.M. Cho, T.G. Lee, and Y.S.  215  46.  47. 48.  49. 50.  51. 52.  53.  54.  55.  56.  57.  58.  59.  Heo, Crystal structure of the biotin carboxylase domain of human acetyl-CoA carboxylase 2. Proteins, 2008. 70(1): 268-272. Konishi, T., K. Shinohara, K. Yamada, and Y. Sasaki, Acetyl-CoA carboxylase in higher plants: most plants other than gramineae have both the prokaryotic and the eukaryotic forms of this enzyme. Plant and Cell Physiology, 1996. 37(2): 117122. Nikolau, B.J., J.B. Ohlrogge, and E.S. Wurtele, Plant biotin-containing carboxylases. Archives of Biochemistry and Biophysics, 2003. 414(2): 211-222. Barnea, M., Z. Madar, and O. Froy, High-fat diet delays and fasting advances the circadian expression of adiponectin signaling components in mouse liver. Endocrinology, 2009. 150(1): 161-168. NCBI, NCBI Protein Database. 2009, National Center for Biotechnology Information. Maglott, D., J. Ostell, K.D. Pruitt, and T. Tatusova, Entrez Gene: gene-centered information at NCBI. Nucleic Acids Research, 2005. 33(Database issue): D54D58. UniProt, The Universal Protein Resource (UniProt), in Nucleic Acids Research. 2008. p. D190-D195. Abu-Elheiga, L., D.B. Almarza-Ortega, A. Baldini, and S.J. Wakil, Human acetyl-CoA carboxylase 2. Molecular cloning, characterization, chromosomal mapping and evidence for two isoforms. Journal of Biological Chemistry, 1997. 272: 10669-10677. Waldrop, G.L., I. Rayment, and H.M. Holden, Three-dimensional structure of the biotin carboxylase subunit of acetyl-CoA carboxylase. Biochemistry, 1994. 33(34): 10249-10256. Shen, Y., S.L. Volrath, S.C. Weatherly, T.D. Elich, and L. Tong, A mechanism for the potent inhibition of eukaryotic acetyl-Coenzyme A carboxylase by soraphen A, a macrocyclic polyketide natural product. Molecular Cell, 2004. 16: 881-891. Athappilly, F.K. and W.A. Hendrickson, Structure of the biotinyl domain of acetyl-coenzyme A carboxylase determined by MAD phasing. Structure, 1995. 3: 1407-1419. Lee, C.K., H.K. Cheong, K.S. Ryu, J.I. Lee, W. Lee, Y.H. Jeon, and C. Cheong, Biotinoyl domain of human acetyl-CoA carboxylase: structural insights into the carboxyl transfer mechanism. Proteins, 2008. 72(2): 613-624. Zhang, H., Z. Yang, Y. Shen, and L. Tong, Crystal structure of the carboxyltransferase domain of acetyl-coenzyme A carboxylase. Science, 2003. 299(5615): 2064-2067. Liu, W., D. Harrison, D. Chalupska, P. Gornicki, C. O'Donnel, S. Adkins, R. Haselkorn, and R. Williams, Single-site mutations in the carboxyltransferase domain of plastid acetyl-CoA carboxylase confer resistance to grass-specific herbicides. Proceedings of the national academy of science USA, 2007. 104(9): 3627-3632. Zhang, H., B. Tweel, and L. Tong, Molecular basis for the inhibition of the carboxyltransferase domain of acetyl-coenzyme-A carboxylase by haloxyfop and diclofop. Proceedings of the National Academy of Science USA, 2004. 101(16): 5910-5915.  216  60.  61.  62.  63.  64.  65.  66.  67.  68.  69. 70.  71.  72.  73.  Beaty, N.B. and M.D. Lane, Acetyl Coenzyme A Carboxylase - Rapid purification of the chick liver enzyme and steady state kinetic analysis of the carboxylasecatalyzed reaction. Journal of Biological Chemistry, 1981. 257(2): 924-929. Bianchi, A., J.L. Evans, A.J. Iverson, A. Nordlund, W.T. D, and L.A. Witters, Identification of an isozymic form of acetyl-CoA carboxylase. Journal of Biological Chemistry, 1990. 265: 1502-1509. Widmer, J., K.S. Fassihi, S.C. Schlichter, K.S. Wheeler, B.E. Crute, N. King, N. Nutile-McMenemy, W.W. Noll, S. Daniel, J. Ha, K.H. Kim, and L.A. Witters, Identification of a second human acetyl-CoA carboxylase gene. Biochemical Journal, 1996. 316(915-922). Abu-Elheiga, L., A. Jayakumar, A. Baldini, S.S. Chirala, and S.J. Wakil, Human acetyl-CoA carboxylase: characterization, molecular cloning, and evidence for two isoforms. Proceedings of the national academy of science USA, 1995. 92: 4011-4015. Lopez-Casillas, F., D.H. Bai, X. Luo, I.S. Kong, M.A. Hermodson, and K.H. Kim, Structure of the coding sequence and primary amino acid sequence of acetyl-coenzyme A carboxylase. Proceedings of the National Academy of Science USA, 1988. 85: 5784-5788. Kong, I.S., F. Lopez-Casillas, and K.H. Kim, Acetyl-CoA carboxylase mRNA species with or without inhibitory coding sequence for Ser-1200 phosphorylation. Journal of Biological Chemistry, 1990. 265(23): 13695-13701. Winz, R., D. Hess, R. Aebersold, and R.W. Brownsey, Unique structural features and differential phosphorylation of the 280-kDa component (isozyme) of rat liver acetyl-CoA carboxylase. Journal of Biological Chemistry, 1994. 269(20): 1443814445. Ha, J., J.K. Lee, K.S. Kim, L.A. Witters, and K.H. Kim, Cloning of human acetylCoA carboxylase-beta and its unique features. Proceedings of the National Academy of Science USA, 1996. 93: 11466-11470. Abu-Elheiga, L., W.R. Brinkley, L. Zhong, S.S. Chirala, G. Woldegiorgis, and S.J. Wakil, The subcellular localization of acetyl-CoA carboxylase 2. Proceedings of the national academy of science USA, 2000. 97(4): 1444-1449. Kim, K.H., Regulation of mammalian acetyl-coenzyme A carboxylase. Annual Review of Nutrition, 1997. 17: 77-99. Trumble, G.E., M.A. Smith, and W.W. Winder, Purification and characterization of rat skeletal muscle acetyl-CoA carboxylase. European Journal of Biochemistry, 1995. 231: 192-198. Spencer, E.B., A. Bianchi, J. Widmer, and L.A. Witters, Brain acetyl-CoA carboxylase: Isozymic identification and studies of its regulation during development and altered nutrition. Biochemical and Biophysical Research Communications, 1993. 192(2): 820-825. Thampy, K.G., Formation of malonyl coenzyme A in rat heart identification and purification of an isozyme of acetyl-coenzyme A carboxylase from rat heart. Journal of Biological Chemistry, 1989. 264(30): 17631-17634. Lee, J.K. and K.H. Kim, Roles of Acetyl-CoA Carboxylase B in Muscle cell differentiation and in mitochondrial fatty acid oxidation. Biochemical and Biophysical Research Communications, 1999. 254: 657-660.  217  74.  75.  76. 77.  78.  79.  80.  81.  82.  83.  84.  85.  Abu-Elheiga, L., W. Oh, P. Kordani, and S.J. Wakil, Acetyl-CoA carboxylase 2 mutant mice are protected against obesity and diabetes induced by high-fat/highcarbohydrate diets. Proceedings of the National Academy of Science USA, 2003. 100(18): 10207-10212. Mao, J., F.J. DeMayo, H. Li, L. Abu-Elheiga, Z. Gu, T. Shaikenov, P. Kordari, S.S. Chirala, W.C. Heird, and S.J. Wakil, Liver-specific deletion of acetyl-CoA carboxylase 1 reduces hepatic triglyceride accumulation without affecting glucose homeostasis. Proceedings of the National Academy of Science USA, 2006. 103(22): 8552-8557. Iritani, N., Nutritional and hormonal regulation of lipogenic-enzyme gene expression in rat liver. European Journal of Biochemistry, 1992. 205(2): 433-442. Nakanishi, S. and S. Numa, Purification of rat liver acetyl coenzyme A carboxylase and immunochemical studies on its synthesis and degradation. European Journal of Biochemistry, 1970. 16(1): 161-173. Majerus, P.W. and E. Kilburn, Acetyl coenzyme A carboxylase. The roles of synthesis and degradation in regulation of enzyme levels in rat liver. Journal of Biological Chemistry, 1969. 244(22): 6254-6262. Mao, J., S.S. Chirala, and S.J. Wakil, Human acetyl-CoA carboxylase 1 gene: presence of three promoters and heterogeneity at the 5'-untranslated mRNA region. Proceedings of the National Academy of Science USA, 2003. 100(13): 7515-7520. Oh, S.Y., M.Y. Lee, J.M. Kim, S. Yoon, S. Shin, Y.N. Park, Y.H. Ahn, and K.S. Kim, Alternative usages of multiple promoters of the acetyl-CoA carboxylase beta gene are related to differential transcriptional regulation in human and rodent tissues. Journal of Biological Chemistry, 2005. 280(7): 5906-5916. Kim, J.Y., J.J. Lee, and K.S. Kim, Acetyl-CoA carboxylase beta expression mediated by MyoD and muscle regulatory factor 4 is differentially affected by retinoic acid receptor and retinoid X receptor. Experimental and Molecular Medicine, 2003. 35(1): 23-29. Oh, S.-Y., S.-K. Park, J.-W. Kim, Y.-H. Ahn, S.-W. Park, and K.-S. Kim, AcetylCoA carboxylase beta gene is regulated by sterol regulatory element-binding protein-1 in liver. Journal of Biological Chemistry, 2003. 278(31): 28410-28417. Qi, L., J.E. Heredia, J.Y. Altarejos, R. Screaton, N. Goebel, S. Niessen, I.X. MacLeod, C.W. Liew, R.N. Kulkarni, J. Bain, C. Newgard, M. Nelson, R.M. Evans, J. Yates, and M. Montminy, TRB3 links the E3 ubiquitin ligase COP1 to lipid metabolism. Science, 2006. 312(5781): 1763-1766. Okamoto, H., E. Latres, R. Liu, K. Thabet, A. Murphy, D. Valenzeula, G.D. Yancopourlos, T.N. Stitt, D.J. Glass, and M.W. Sleeman, Genetic deletion of Trb3, the mammalian Drosophila tribbles homolog, displays normal hepatic insulin signaling and glucose homeostasis. Diabetes, 2007. 56(5): 1350-1356. Ryder, E., C. Gregolin, H.C. Chang, and M.D. Lane, Liver acetyl-CoA carboxylase: insight into the mechanism of activation by tricarboxylic acids and acetyl CoA. Proceedings of the National Academy of Science USA, 1967. 57(5): 1455-1462.  218  86.  87.  88.  89. 90.  91.  92.  93.  94.  95.  96.  97.  98.  99. 100.  101.  Gregolin, C., E. Ryder, A.K. Kleinschmidt, R.C. Warner, and M.D. Lane, Molecular characteristics of liver acetyl-CoA carboxylase. Biochemistry, 1966. 56: 148-155. Martin, D.B. and P.R. Vagelos, The mechanism of tricarboxylic acid cycle regulation of fatty acid synthesis. Journal of Biological Chemistry, 1962. 237(6): 1787-1792. Halestrap, A.P. and R.M. Denton, Hormonal regulation of adipose-tisue acetylCoenzyme A carboxylase by changes in the polymeric state of the enzyme. Biochemical Journal, 1974. 142: 365-377. Halestrap, A.P. and R.M. Denton, Insulin and the regulation of adipose tissue acetyl-Coenzyme A carboxylase. Biochemical Journal, 1973. 132: 509-517. Saha, A.K., D. Vavvas, T.G. Kurowski, A. Apazidis, L.A. Witters, E. Shafrir, and N.B. Ruderman, Malonyl -CoA regulation in skeletal muscle: its link to cell citrate and the glucose-fatty acid cycle. American Journal of Physiology, 1997. 272(35): E641-E648. Winder, W.W., P.S. MacLean, J.C. Lucas, J.E. Fernley, and G.E. Trumble, Effect of fasting and refeeding on acetyl-CoA carboxylase in rat hindlimb. Journal of Applied Physiology, 1995. 78(2): 578-582. Belke, D.D., L.C.H. Wang, and G.D. Lopaschuk, Acetyl-CoA carboxylase control of fatty acid oxidation in hearts from hibernating Richardson's ground squirrels. Biochimica et Biophysica Acta, 1998. 1391: 25-36. Baquet, A., V. Gaussin, M. Bollen, W. Stalmans, and L. Hue, Mechanism of activation of liver acetyl-CoA carboxylase by cell swelling. European Journal of Biochemistry, 1993. 217(3): 1083-1089. Gaussin, V., L. Hue, W. Stalmans, and M. Bollen, Activation of hepatic acetylCoA carboxylase by glutamate and Mg2+ is mediated by protein phosphatase-2A. Biochemical Journal, 1996. 316(Pt 1): 217-224. Boone, A.N., A. Chan, J. Kulpa, and R.W. Brownsey, Bimodal activation of acetyl-CoA carboxylase by glutamate. Journal of Biological Chemistry, 2000. 275(15): 10819-10825. Ogiwara, H., T. Tanabe, J. Nikawa, and S. Numa, Inhibition of rat-liver acetylcoenzyme A carboxylase by palmitoyl-coenzyme A - Formation of equimolar enzyme-inhibitor complex. European Journal of Biochemistry, 1978. 89: 33-41. Yeh, L.A. and K.H. Kim, Regulation of acetyl-CoA carboxylase: Properties of CoA activation of acetyl-CoA carboxylase. Proceedings of the National Academy of Science USA, 1980. 77(6): 3351-3355. Moule, S.K., N.J. Edgell, A.C. Borthwick, and R.M. Denton, Coenzyme A is a potent inhibitor of acetyl-CoA carboxylase from rat epididymal fat pads. Biochemical Journal, 1992. 283: 35-38. Beaty, N.B. and M.D. Lane, The polymerization of acetyl-CoA carboxylase. Journal of Biological Chemistry, 1983. 258(21): 13051-13055. Gregolin, C., E. Ryder, A.K. Kleinschmidt, R.C. Warner, and M.D. Lane, Molecular characteristics of liver acetyl CoA carboxylase. Proceedings of the National Academy of Science USA, 1966. 56(1): 148-155. Buechler, K.F., A.C. Beynen, and M.J.H. Geelen, Studies on the assay, activity and sedimentation behaviour of acetyl-CoA carboxylase from isolated  219  102.  103.  104.  105.  106.  107.  108. 109.  110.  111.  112.  113.  114. 115.  hepatocytes incubated with insulin or glucagon. Biochemical Journal, 1984. 221: 869-874. Beaty, N.B. and M.D. Lane, Kinetics of activation of acetyl-CoA carboxylase by citrate. Relationship to the rate of polymerization of the enzyme. Journal of Biological Chemistry, 1983. 258(21): 13043-13050. Meredith, M.J. and M.D. Lane, Acetyl-CoA Carboxylase - evidence for polymeric filament to protomer transition in the intact avian liver cell. Journal of Biological Chemistry, 1978. 253(10): 3381-3383. Moss, J. and M.D. Lane, Acetyl coenzyme A carboxylase. III. Further studies on the relation of catalytic activity to polymeric state. Journal of Biological Chemistry, 1972. 247(16): 4944-4951. Borthwick, A.C., N.J. Edgell, and R.M. Denton, Use of rapid gel-permeation chromatography to explore the inter-relationships between polymerization, phosphorylation and activity of acetyl-CoA carboxylase. Biochemical Journal, 1987. 241: 773-782. Locke, G.A., D. Cheng, M.R. Witmer, J.K. Tamura, T. Haque, R.F. Carney, A.R. Rendina, and J. Marcinkeviciene, Differential activation of recombinant human acetyl-CoA carboxylases 1 and 2 by citrate. Archives of Biochemistry and Biophysics, 2008. 475(1): 72-79. Munday, M.R., M.R. Milic, S. Takhar, M.J. Holness, and M.C. Sugden, The short-term regulation of hepatic acetyl-CoA carboxylase during starvation and re-feeding in the rat. Biochemical Journal, 1991. 280: 733-737. Inoue, H. and J.M. Lowenstein, Acetyl coenzyme A carboxylase from rat liver. Journal of Biological Chemistry, 1972. 247: 4825-4832. Carlson, C.A. and K.H. Kim, Regulation of hepatic acetyl coenzyme A carboxylase by phosphorylation and dephosphorylation. Journal of Biological Chemistry, 1973. 218(1): 378-380. Brownsey, R.W., W.A. Hughes, and R.M. Denton, Demonstration of the phosphorylation of ACC in intact rat epididymal fat cells. Biochemical Journal, 1977. 168(3): 441-445. Stansbie, D., R.W. Brownsey, M. Crettaz, and R.M. Denton, Acute effects in vivo of anti-insulin serum on rates of fatty acid synthesis and activities of acetylcoenzyme A carboxylase and pyruvate dehydrogenase in liver and epididymal adipose tissue of fed rats. Biochemical Journal, 1976. 160: 413-415. Mabrouk, G.M., T.M. Helmy, K.G. Thampy, and S.J. Wakil, Acute hormonal control of acetyl-CoA carboxylase. Journal of Biological Chemistry, 1990. 265(11): 6330-6338. Brownsey, R.W., W.A. Hughes, and R.M. Denton, Adrenaline and the regulation of acetyl-coenzyme A carboxylase in rat epididymal adipose tissue. Biochemical Journal, 1979. 184: 23-32. Lee, K.-H. and K.-H. Kim, Effect of epinephrine on acetyl-CoA carboxylase in rat epididymal fat tissue. Journal of Biological Chemistry, 1978. 253(22): 8157-8161. Kim, Y.J., M.S. Lee, H.J. Lee, Y. Wu, H.C. Freake, H.S. Chun, and Y.E. Kim, Hormones and nutrients regulate acetyl-CoA carboxylase promoter I in rat primary hepatocytes. Journal of Nutritional Science and Vitaminology, 2005. 51(2): 124-128.  220  116.  117.  118.  119.  120.  121.  122.  123.  124.  125.  126.  127.  128.  Gao, S., K.P. Kinzig, S. Aja, K.A. Scott, W. Keung, S. Kelly, K. Strynadka, S. Chohnan, W.W. Smith, K.L.K. Tamashiro, E.E. Ladenheim, G.V. Ronnett, Y. Tu, M.J. Birnbaum, G.D. Lopaschuk, and T.H. Moran, Leptin activates hypothalamic acetyl-CoA carboxylase to inhibit food intake. Proceedings of the National Academy of Science USA, 2007. 104(44): 17358-17363. Lee, K.H., T. Thrall, and K.H. Kim, Hormonal regulation of acetyl-CoA carboxylase: effect of insulin and epinephrine. Biochemical and Biophysical Research Communications, 1973. 54(3): 1133-1140. Brownsey, R.W. and R.M. Denton, Evidence that insulin activates fat-cell acetylCoA carboxylase by increased phosphorylation at a specific site. Biochemical Journal, 1982. 202(1): 77-86. Lee, K.H. and K.H. Kim, Stimulation by epinephrine of in vivo phosphorylation and inactivation of acetyl coenzyme A carboxylase of rat epidiymal adipose tissue. Journal of Biological Chemistry, 1979. 254(5): 1450-1453. Witters, L.A., E.M. Kowaloff, and J. Avruch, Glucagon regulation of protein phosphorylation. Identification of acetyl coenzyme A carboxylase as a substrate. Journal of Biological Chemistry, 1979. 254(2): 245-248. Munday, M.R., D.G. Campbell, D. Carling, and D.G. Hardie, Identification by amino acid sequencing of three major regulatory phosphorylation sites on rat acetyl-CoA carboxylase. European Journal of Biochemistry, 1988. 175(2): 331338. Haystead, T.A.J., S. Moore, P. Cohen, and D.G. Hardie, Roles of the AMPactivated and cyclic AMP-dependent protein kinases in the adrenaline-induced inactivation of acetyl-CoA carboxylase in rat adipocytes. European Journal of Biochemistry, 1988. 187: 199-205. Hardie, D.G., AMPK: a key regulator of energy balance in the single cell and the whole organism. International Journal of Obesity (London), 2008. 32(Suppl 4): S7-S12. Hardie, D.G. and P.S. Guy, Reversible phosphorylation and inactivation of acetylCoA carboxylase from lactating rat mammary gland by cyclic AMP-dependent protein kinase. European Biophysics Journal, 1980. 110(1): 167-177. Brownsey, R.W., A.N. Boone, J.E. Elliott, J. Kulpa, and W.M. Lee, Regulation of acetyl-CoA carboxylase. Biochemical Society Transactions, 2006. 34(2): 223227. Boone, A.N., B. Rodrigues, and R.W. Brownsey, Multiple-site phosphorylation of the 280 kDa isoform of acetyl-CoA carboxylase in rat cardiac myocytes: evidence that cAMP dependent kinase mediates effects of beta-adrenergic stimulation. Biochemical Journal, 1999. 341: 347-354. Yamagishis, S.I., D. Edelstein, X.L. Du, Y. Kaneda, M. Guzman, and M. Brownlee, Leptin induced mitochondrial superoxide production and monocyte chemoattractant protein-1 expression in aortic endothelial cells by increasing fatty acid oxidation via protein kinase A. Journal of Biological Chemistry, 2001. 276(27): 25096-25100. Haystead, T.A.J., D.G. Campbell, and D.G. Hardie, Analysis of sites phosphorylated on acetyl-Coa carboxylase in response to insulin in isolated adipocytes. European Journal of Biochemistry, 1988. 175(2): 347-354.  221  129.  130.  131.  132.  133.  134. 135.  136.  137.  138.  139.  140.  141. 142. 143.  Hardie, D.G., D. Carling, S. Ferrari, P.S. Guy, and A. Aitken, Characterization of the phosphorylation of rat mammary ATP-citrate lyase and acetyl-CoA carboxylase by Ca2+ and calmodulin-dependent multiprotein kinase and Ca2+ and phospholipid-dependent protein kinase. European Journal of Biochemistry, 1986. 157(3): 553-561. Shen, Y. and L. Tong, Structural evidence for direct interactions between the BRCT domains of human BRCA1 and a phospho-peptide from human ACC1. Biochemistry, 2008. 47(21): 5767-5773. Witters, L.A., T.D. Watts, D.L. Daniels, and J.L. Evans, Insulin stimulates the dephosphorylation and activation of acetyl-CoA carboxylase. Proceedings of the National Academy of Science USA, 1988. 85: 5473-5477. Heesom, K.J., S.K. Moule, and R.M. Denton, Purification and characterisation of an insulin-stimulated protein-serine kinase which phosphorylates acetyl-CoA carboxylase. Federation of Biochemical Studies letters, 1998. 422: 43-46. Munday, M.R. and C.J. Hemingway, The regulation of acetyl-CoA carboxylase a potential target for the action of hypolipidemic agents. Advances in Enzyme Regulation, 1999. 39: 205-234. Wakil, S.J., J.K. Stoops, and V.C. Joshi, Fatty acid synthesis and its regulation. Annual Review of Biochemistry, 1983. 52: 537-579. Ha, J. and K.H. Kim, Inhibition of fatty acid synthesis by expression of an acetylCoA carboxylase-specific ribozyme gene. Proceedings of the National Academy of Science USA, 1994. 91(21): 9951-9955. Heinrich, R. and T.A. Rapoport, A linear steady-state treatment of enzymatic chains: general properties, control and effector strength. European Journal of Biochemistry, 1974. 42: 89-95. Ruderman, N.B., A.K. Saha, D. Vavvas, and L.A. Witters, Malonyl-CoA, fuel sensing, and insulin resistance. American Journal of Physiology, 1999. 276(39): E1-E18. McGarry, J.D., The mitochondrial carnitine palmitoyl-transferase system: its broadening role in fuel homeostasis and new insights into its molecular features. Biochemical Society Transactions, 1995. 23: 321-324. McGarry, J.D., S.E. Mills, C.S. Long, and D.W. Foster, Observations on the affinity for carnitine, and malonyl-CoA sensitivity, of carnitine palmitoyltransferase I in animal and human tissues. Demonstration of the presence of malonyl-CoA in non-hepatic tissues of the rat. Biochemical Journal, 1983. 214(1): 21-28. Brun, T., E. Roche, F. Assimacopoulos-Jeannet, B.E. Corkey, K.H. Kim, and M. Prentki, Evidence for an anaplerotic/malonyl-CoA pathway in pancreatic betacell nutrient signaling. Diabetes, 1996. 45: 190-198. Leonard, A.E., S.L. Pereira, H. Sprecher, and Y.-S. Huang, Elongation of longchain fatty acids. Progress in Lipid Research, 2004. 43(1): 36-54. Tong, L. and H.J. Harwood, Acetyl-coenzyme A carboxylases: versatile targets for drug discovery. Journal of Cellular Biochemistry, 2006. 99(6): 1476-1488. Parker, W.B., L.C. Marshall, J.D. Burton, D.A. Somers, D.L. Wyse, J.W. Gronwald, and B.G. Gengenback, Dominant mutations causing alterations in acetyl-coenzyme A carboxylase confer tolerance to cyclohexanedione and  222  144.  145.  146.  147.  148.  149.  150. 151.  152.  153. 154. 155.  156.  aryloxyphenoxypropionate herbicides in maize. Proceedings of the National Academy of Science USA, 1990. 87(18): 7175-7179. Harwood, H.J., S.F. Petras, L.D. Shelly, L.M. Zaccaro, D.A. Perry, M.R. Makowski, D.M. Hargrove, K.A. Martin, W.R. Tracey, J.G. Chapman, W.P. Magee, D.K. Dalvie, V.F. Soliman, W.H. Martin, C.J. Mularsi, and S.A. Eisenbeis, Isozyme-nonselective N-substituted bipiperidylcarboxamide acetylCoA carboxylase inhibitors reduce tissue malonyl-CoA concentrations, inhibit fatty acid synthesis, and increase fatty acid oxidation in cultured cells and in experimental animals. Journal of Biological Chemistry, 2003. 278(39): 3709937111. Savage, D.B., C.S. Choi, V.T. Samuel, Z.X. Liu, D. Zhang, A. Wang, X.M. Zhang, G.W. Cline, X.X. Yu, J.G. Geisler, S. Bhanot, B.P. Monia, and G.I. Shulman, Reversal of diet-induced hepatic steatosis and hepatic insulin resistance by antisense oligonucleotide inhibitors of acetyl-CoA carboxylases 1 and 2. The Journal of Clinical Investigation, 2006. 116(3): 817-824. Magnard, C., R. Bachelier, A. Vincent, M. Jaquinod, S. Kieffer, G.M. Lenoir, and N.D. Venezia, BRCA1 interacts with acetyl-CoA carboxylase through its tandem of BRCT domains. Oncogene, 2002. 21(44): 6729-6739. Brusselmans, K., E. De Schrijver, G. Verhoeven, and J.V. Swinnen, RNA interference - mediated silencing of the acetyl-CoA carboxylase alpha gene induces growth inhibition and apoptosis of prostrate cancer cells. Cancer Research, 2005. 65: 6719-6725. Ohmori, K., H. Yamada, A. Yasuda, A. Yamamoto, and N. Matsuura, Kiniwa, M, Anti-hyperlipidemic action of a newly synthesized benzoic acid derivative, S-2E. European Journal of Pharmacology, 2003. 471: 69-76. Endo, A., H. Takeshima, and K. Kuwabara, Acetyl CoA carboxylase inhibitors from the fungus Gongronella butleri. Journal of Antibiotics (Tokyo), 1985. 38(5): 599-604. Allred, J.B. and K.L. Roehrig, Inhibition of rat liver acetyl CoA carboxylase by chloride. Journal of Lipid Research, 1980. 21: 488-491. Elliott, J.E., Interaction of pyridoxal phosphate with acetyl-CoA carboxylase, in Department of Biochemistry and Molecular Biology. 2001, University of British Columbia: Vancouver. Vallee, R.B., Reversible assembly purification of microtubules without assemblypromoting agents and further purification of tubulin, microtubule-associated proteins, and MAP fragments. Methods in Enzymology, 1986. 134: 89-104. Kohanski, R.A. and M.D. Lane, Monovalent avidin affinity columns. Methods in Enzymology, 1990. 184: 194-200. Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 1970. 227(5259): 680-685. Carey, E.M. and R. Dils, Fatty acid biosynthesis. V. Purification and characterisation of fatty acid synthetase from lactating-rabbit mammary gland. Biochimica et Biophysica Acta, 1970. 210(3): 371-387. Bradford, M.M., A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the priciple of protein-dye binding. Analytical biochemistry, 1976. 72: 248-254.  223  157.  158.  159.  160. 161.  162.  163.  164. 165.  166. 167.  168.  169.  170.  171.  172.  Perkins, D.N., D.J.C. Pappin, D.M. Creasy, and J.S. Cottrell, Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis, 1999. 20(18): 3551-3567. Chang, T.K.H., J. Chen, and W.T. Xiao, Distinct role of bilobalide and ginkgolide A in the modulation of rat CYP2B1 and CYP3A23 gene expression by Ginkgo biolba extract in cultured hepatocytes. Drug Metabolism and Disposition, 2006. 34(2): 234-242. Colombo, G. and R.G. Kemp, Specific modification of an effector binding site of phosphofructokinase by pyridoxal phosphate. Biochemistry, 1976. 15(8): 17741780. Uyeda, K., Reaction of phosphofructokinase with maleic anhydride, succinic anhydride, and pyridoxal 5'-phosphate. Biochemistry, 1969. 8(6): 2366-73. Kemp, R.G., R.W. Fox, and S.P. Latshaw, Amino acid sequence at the citrate allosteric site of rabbit muscle phosphofructokinase. Biochemistry, 1987. 26: 3443-3446. Lee, W.M., J.E. Elliott, and R.W. Brownsey, Inhibition of acetyl-CoA carboxylase isoforms by pyridoxal phosphate. Journal of Biological Chemistry, 2005. 280(51): 41835-41843. Pandey, A. and S.S. Katiyar, Essential lysine residue in glutathione reductase: chemical modification by pyridoxal 5'-phosphate. Biochemistry and Molecular Biology International, 1995. 36(2): 347-354. Jitrapakdee, S. and J.C. Wallace, Structure, function and regulation of pyruvate carboxylase. Biochemical Journal, 1999. 340: 1-16. Moir, A.M. and V.A. Zammit, Changes in the properties of cytosolic acetyl-CoA carboxylase studied in cold-clamped liver samples from fed, starved and starvedrefed rats. Biochemical Journal, 1990. 272(2): 511-517. Srere, P.A., B. Sumegi, and A.D. Sherry, Organizational aspects of the citric acid cycle. Biochemical Society Symposium, 1987. 54: 173-178. Quayle, K.A., R.M. Denton, and R.W. Brownsey, Evidence for a protein regulator from rat liver which activates acetyl-CoA carboxylase. Biochemical Journal, 1993. 292: 75-84. Cheng, D., C. Chu, L. Chen, J. Feder, G.A. Mintier, Y. Wu, J. Cook, M.R. Harpel, G.A. Locke, Y. An, and J.K. Tamura, Expression, purification, and characterization of human and rat acetyl coenzyme A carboxylase (ACC) isozymes. Protein Expression and Purification, 2007. 51: 11-21. Abe, K., Y. Shinohara, and H. Terada, Isolation and characterization of cDNA encoding rat heart type acetyl-CoA carboxylase. Biochimica et Biophysica Acta, 1998. 1398(3): 347-352. Usui, H., Y. Miyzazki, D. Xin, T. Ichikawa, and T. Kumanishi, Cloning and sequencing of the rat cDNAs encoding class I beta-tubulin. DNA sequence, 1998(5-6): 365-368. Levitsky, D.I., A.V. Pivovarova, V.V. Mikhailova, and O.P. Nikolaeva, Thermal unfolding and aggregation of actin: Stabilization and destabilization of actin filaments. Federation of Biochemical Studies Journal, 2008. 275(17): 4280-4295. Krupenko, S.A., FDH: an aldehyde dehydrogenase fusion enzyme in folate metabolism. Chemico-Biological Interactions, 2008. Epublication.  224  173.  174.  175. 176.  177.  178.  179.  180.  181.  182.  183.  184. 185.  186. 187.  Ronnett, G.V., A.M. Kleman, E.-K. Kim, L.E. Landree, and Y. Tu, Fatty acid metabolism, the central nervous sytem, and feeding. Obesity, 2006. 14(Supplement): 201S-207S. Ang, L.S., R.P. Cruz, A. Hendel, and D.J. Granville, Apolipoprotein E, an important player in longevity and age-related diseases. Experimental Gerontology, 2008. 43(7): 615-622. Pleasure, I.T., M.M. Black, and J.H. Keen, Valosin-containing protein, VCP, is a ubiquitous clathrin-binding protein. Nature, 1993. 365: 459-462. Pearse, B.M., Clathrin: a unique protein associated with intracellular transfer of membrane by coated vesicles. Proceedings of the National Academy of Science USA, 1976. 73(4): 1255-1259. Wisniewski, J., T. Kordula, and A. Krawczyk, Isolation and nucleotide sequence analysis of the rat testis-specific major heat-shock protein (HSP70)-related gene. Biochimica et Biophysica Acta, 1990. 1048(1): 93-99. Liao, D., X. Yang, and H. Wang, Hyperhomocysteinemia and high-density lipoprotein metabolism in cardiovascular disease. Chemisty and Laboratory Medicine, 2007. 45(12): 1652-1659. Jelski, W. and M. Szmitkowski, Alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) in the cancer diseases. Clinica Chimica Acta, 2008. 395(1-2): 1-5. Dzugaj, A., Localization and regulation of muscle fructose-1,6-bisphosphatase, the key enzyme of glyconeogenesis. Advances in Enzyme Regulation, 2006. 46: 51-71. Goldfarb, S.B., O.B. Kahlan, J.N. Watkins, L. Suaud, W. Yan, T.R. Kleyman, and R.C. Rubenstein, Differential effects of Hsc70 and Hsp70 on the intracellular trafficking and functional expression of epithelial sodium channels. Proceedings of the National Academy of Science USA, 2006. 103(15): 5817-5822. Gilles, A.M., E. Presecan, A. Vonica, and I. Lascu, Nucleoside diphosphate kinase from human erythrocytes. Structural characterization of the two polypeptide chains responsible for heterogeneity of the hexameric enzyme. Journal of Biological Chemistry, 1991. 266(14): 8784-8789. Nickerson, J.A. and W.W. Wells, The microtubule-associated nucleoside diphosphate kinase. Journal of Biological Chemistry, 1984. 259(18): 1129711304. Patel, M.S. and L.G. Korotchkina, Regulation of the pyruvate dehydrogenase complex. Biochemical Society Transactions, 2006. 34(Pt 2): 217-222. Zhou, Y., X. Yi, J.B. Soffer, N. Bonafe, M. Gilmore-Hebert, J. McAlpine, and S.K. Chambers, The multifunctional protein glyceraldehyde-3-phosphate dehydrogenase is both regulated and controls colony-stimulating factor-1 messenger RNA stability in ovarian cancer. Molecular Cancer Research, 2008. 6: 1375-1384. Hargrove, G.M., A. Junco, and N.C. Wong, Hormonal regulation of apolipoprotein AI. Journal of Molecular Endocrinology, 1999. 22(2): 103-111. Larter, C.Z., M.M. Yeh, W.G. Haigh, J. Williams, S. Brown, K.S. Bell-Anderson, S.P. Lee, and G.C. Farrell, Hepatic free fatty acids accumulate in experimental  225  188.  189.  190.  191.  192.  193.  194.  195.  196.  197.  198.  199. 200.  201.  steatohepatitis: role of adaptive pathways. Journal of Hepatology, 2008. 48(4): 638-647. Florea, L., V. Di Francesco, J. Miller, R. Turner, A. Yao, M. Harris, B. Walenz, C. Mobarry, G.V. Merkulov, R. Charlab, I. Dew, Z. Deng, S. Istrail, P. Li, and G. Sutton, Gene and alternative splicing annotation with AIR. Genome Research, 2005. 15(1): 54-66. Gerhard, D.S., L. Wagner, E.A. Feingold, C.M. Shenmen, L.H. Grouse, G. Schuler, and S.L. Klein, The status, quality, and expansion of the NIH full-length cDNA project: the Mammalian Gene Collection (MGC). Genome Research, 2004. 14(10B): 2121-2127. Geelen, M.J.H., A.C. Beynen, R.Z. Christiansen, M.J. Lepreau-Jose, and D.M. Gibson, Short-term effects of insulin and glucagon on lipid synthesis in isolated rat hepatocytes. Federation of Biochemical Studies letters, 1978. 95(2): 326-329. Wilson, L., D. Panda, and M.A. Jordan, Modulation of microtubule dynamics by drugs: a paradigm for the actions of cellular regulators. Cell Structure and Function, 1999. 24: 329-335. Mick, G.J., K.Y. Chun, T.L. VanderBloomer, C.-L. Fu, and K.L. McCormick, Inhibition of acetyl CoA carboxylase by GTPgS. Biochimica et Biophysica ActaProtein structure and molecular enzymology, 1998. 1384(1): 130-140. Moreau, K., E. Dizin, H. Ray, C. Luquain, E. Lefai, F. Foufelle, M. Billaud, G.M. Lenoir, and N.D. Venezia, BRCA1 affects lipid synthesis through its interaction with acetyl-CoA carboxylase. Journal of Biological Chemistry, 2006. 281(6): 3172-3181. Stelzl, U., U. Worm, M. Lalowski, C. Haenig, F. Brembeck, H. Goehler, M. Stroedicke, M. Zenkner, and A. Schoenherr, A human protein-protein interaction network - a resource for annotating the proteome. Cell, 2005. 122(6): 957-968. Cha, S.H., Z. Hu, S. Chohnan, and M.D. Lane, Inhibition of hypothalamic fatty acid synthase triggers rapid activation of fatty acid oxidation in skeletal muscle. Proceedings of the National Academy of Science USA, 2005. 102(41): 14557014562. Gavin, A.C., M. Bosche, R. Krause, G. P, M. Marzioch, A. Bauer, and J. Schultz, Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature, 2002. 415(January 10): 141-147. Ho, Y., A. Gruhler, A. Hellbut, G.D. Bader, L. Moore, and S.L. Adams, Systematic identification of protein complexes in Saccharomyces cervevisiae by mass spectrometry. Nature, 2002. 415(January 10): 180-183. Stark, C., B.J. Breitkreutz, T. Reguly, L. Boucher, A. Breitkreutz, and M. Tyers, BioGRID: a general repository for interaction datasets. Nucleic Acids Research, 2006. 34(Database issue): D535-D539. Patterson, E.E. and C.A. Fox, The Ku complex in silencing the cryptic mating-type loci of Saccharomyces cerevisiae. Genetics, 2008. 180(2): 771-783. Fukuda, T. and Y. Ohya, Recruitment of RecA homologs Dmc1p and Rad51p to the double-strand break repair site initiated by meiosis-specific endonuclease VDE (PI-SceI). Molecular Genetics and Genomics, 2006. 275(2): 204-214. Smolka, M.B., S.H. Chen, P.S. Maddox, J.M. Enserink, C.P. Albuquerque, X.X. Wei, A. Desai, R.D. Kolodner, and H. Zhou, An FHA domain-mediated protein  226  202.  203.  204.  205.  206.  207.  208.  209.  210.  211. 212. 213.  214.  215.  interaction network of Rad53 reveals its role in polarized cell growth. Journal of Cell Biology, 2006. 175(5): 743-753. Tarassov, K., V. Messier, C.R. Landry, S. Radinovic, M.M.S. Molina, I. Shames, Y. Malitskaya, J. Vogel, H. Busey, and S.W. Michnick, An in vivo map of the yeast protein interactome. Science, 2008. 320(5882): 1465-1470. Jacq, C., J. Alt-Morbe, B. Andre, W. Arnold, A. Bahr, J.P. Ballesta, and M. Barques, The nucleotide sequence of Saccharomyces cerevisiae chromosome IV. Nature, 1997. 387(6632 Supplementary): 75-78. Gorka-Niec, W., R. Bankowska, G. Palamarczyk, H. Krotkiewski, and J.S. Kruszewska, Protein glycosylation in pmt mutants of Saccharomyces cerevisiae. Influence of heterologously expressed cellobiohydrolase II of Trichoderma reesei and elevated levels of GDP-mannose and cis-prenyltransferase activity. Biochimica et Biophysica Acta, 2007. 1770(5): 774-80. Yazawa, H., H. Iwahashi, and H. Uemura, Disruption of URA7 and GAL6 improves the ethanol tolerance and fermentation capacity of Saccharomyces cerevisiae. Yeast, 2007. 24(7): 551-60. Jensen, L.J., M. Kuhn, M. Stark, S. Chaffron, C. Creevey, J. Muller, T. Doerks, P. Julien, A. Roth, M. Simonovic, P. Bork, and C. von Mering, STRING 8 - a global view on proteins and their functional interactions in 630 organisms. Nucleic Acids Research, 2009. 37(Database issue): D412-D416. FitzPatrick, D.R., A. Hill, J.L. Tolmie, D.R. Throburn, and J. Christodoulou, The molecular basis of malonyl-CoA decarboxylase deficiency. Americal Journal of Human Genetics, 1999. 65(2): 318-326. Buechler, K.F. and D.M. Gibson, Guanosine triphosphate and colchicine affect the activity and the polymeric state of acetyl-CoA carboxylase. Archives of Biochemistry and Biophysics, 1984. 233(2): 698-707. Geelen, M.J.H., C. Bijleveld, G. Velasco, R.J.A. Wanders, and M. Guzman, Studies on the intracellular localization of acetyl-CoA carboxylase. Biochemical and Biophysical Research Communications, 1997. 233: 253-257. Witters, L.A., S.A. Friedman, J.P. Tipper, and G.W. Bacon, Regulation of acetylCoA carboxylase by guanine nucleotides. The Journal of Biological Chemistry, 1981. 256(16): 8573-8578. Cell Signaling Technology, I. Acetyl-CoA Carboxylase #3662. 2009 [cited August 2009]; Available from: http://www.cellsignal.com/products/3676.html. Cell Signaling Technology, I. Acetyl-CoA Carboxylase #3676. 2009 [cited August 2009]; Available from: http://www.cellsignal.com/products/3662.html. Santa Cruz biotechnology, I. ACCalpha (T-18): sc-26817. 2009 [cited Aug 2009]; Available from: http://www.scbt.com/datasheet-26817-accalpha-t-18antibody.html. Ivessa, A.S., R. Schneiter, and S.D. Kohlwein, Yeast acetyl-CoA carboxylase is associated with they cytoplasmic surface of the endoplasmic reticulum. European Journal of Cell Biology, 1997. 74: 399-406. Guzman, M., G. Velasco, and M.J.H. Geelen, Do cytoskeletal components control fatty acid translocation into liver mitochondria? Trends in Endocrinology and Metabolism, 2000. 11(2): 49-53.  227  216.  Choudhary, S., K. Joshi, and K.D. Gill, Possible role of enhanced microtuble phosphorylation in dichlorvos induced delyed neurotoxicity in rat. Brain Research, 2001. 897(1-2): 60-70.  228  7  Appendices  Appendix A  229  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0069141/manifest

Comment

Related Items