UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Development of a novel liquid chromatography based tool to study post-translational modifications Lam, Wing Kai Edgar 2008

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

Item Metadata

Download

Media
24-ubc_2008_fall_wing_kai_edgar_lam.pdf [ 7.94MB ]
Metadata
JSON: 24-1.0066384.json
JSON-LD: 24-1.0066384-ld.json
RDF/XML (Pretty): 24-1.0066384-rdf.xml
RDF/JSON: 24-1.0066384-rdf.json
Turtle: 24-1.0066384-turtle.txt
N-Triples: 24-1.0066384-rdf-ntriples.txt
Original Record: 24-1.0066384-source.json
Full Text
24-1.0066384-fulltext.txt
Citation
24-1.0066384.ris

Full Text

DEVELOPMENT OF A NOVEL LIQUID CHROMATOGRAPHY BASED TOOL TO STUDY POSTTRANSLATIONAL MODIFICATIONS  by WING KAI EDGAR LAM B.Sc., The University of British Columbia, 2004  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in  THE FACULTY OF GRADUATE STUDIES (Experimental Medicine)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2008 © Wing Kai Edgar Lam, 2008  ABSTRACT There are many tools available for the study of post-translational modifications. The majority of these tools is specific towards the individual modification and involves separation of modified proteins from non-modified ones. The drawback of using a modification specific method is that there is a lack of flexibility in its usage for other modifications. The goal of these studies was to investigate the possibility of obtaining a similar separation effect by fractionating post-translationally modified proteins based on the physical properties of proteins. The posttranslational modification chosen to be the basis of this study was the O-GlcNAc modification. Using the C2C12 mouse myoblast cell line, it was determined that the optimal conditions for producing lysates containing increased yields of O-GlcNAc modified proteins was to treat differentiated C2C12 cells with 10nM insulin, 12g/L glucose and 2mM of the OGlcNAcase inhibitor Streptozotocin for 24 hours. Using the optimized lysis buffer, it was shown that protein separation by surface charge using standard anion exchange separation did not provide enough resolution or material to obtain any identifications of modified proteins. However, when a chromatofocusing method which separates proteins on the basis of their isoelectric points was used, a separation scheme with larger capacity and higher resolution was possible. Using this separation method followed by gel electrophoresis of individual fractions, proteins which are potentially O-GlcNAc modified were identified by mass spectrometry. It was evident from the number of protein bands observed per fraction on the Coomassie stained gels and the number of proteins identified per protein band by mass spectrometry that further  ii  reduction in sample complexity was required to assist in the positive identification of O-GlcNAc modified proteins. Among the identified proteins, 32 percent were metabolic proteins, 21 percent were protein processing proteins, 16 percent were structural proteins and the remainder a mix of other proteins. Unfortunately, it was not possible to validate the presence or absence of the OGlcNAc  modification  on  these  proteins  using  available  methodologies  such  as  immunoprecipitation. As such, further work is required to optimize the separation strategy and to verify the usefulness of this separation strategy in identifying O-GlcNAc/post-translationally modified proteins.  iii  TABLE OF CONTENTS Abstract ................................................................................................................................ ii Table of Contents................................................................................................................. iv List of Tables .........................................................................................................................vi List of Figures .......................................................................................................................vii List of Abbreviations ............................................................................................................ ix Acknowledgements ............................................................................................................. xi CHAPTER I – Introduction ...................................................................................................... 1 1.1 The O-GlcNAc modification ..........................................................................................2 1.1.1 Functional roles of the O-GlcNAc modification .............................................2 1.1.2 The structure of O-GlcNAc .............................................................................5 1.1.3 Enzymes involved in O-GlcNAc modification .................................................5 1.1.4 Inhibitors of the O-GlcNAc enzymes ..............................................................6 1.1.5 Current methods available for the study of O-GlcNAc modified proteins ....8 1.1.5.1 Galactosyltransferase-based methods ...........................................8 1.1.5.2 Antibody-based methods................................................................9 1.1.5.3 Lectin-based methods................................................................... 10 1.1.5.4 Chemical derivatisation-based methods ...................................... 10 1.2 Chromatographic separation of proteins................................................................... 12 1.2.1 Size exclusion chromatography ................................................................... 12 1.2.2 Reversed phase chromatography ................................................................ 13 1.2.3 Ion exchange chromatography .................................................................... 14 1.2.4 Chromatofocusing........................................................................................ 17 1.3 Towards an alternative O-GlcNAc detection strategy ............................................... 21 CHAPTER II – Materials and Methods ...................................................................................22 2.1 Methods for chapter III ............................................................................................. 22 2.1.1 General cell culture conditions .................................................................... 22 2.1.2 Determination of optimal STZ concentration .............................................. 22 2.1.3 Optimization of detergent free lysis ............................................................ 23 2.1.4 Determination of optimal treatment length ............................................... 23 2.1.5 Electrophoresis and immunoblotting – optimization steps......................... 24 2.1.6 Sample preparation for anion exchange chromatography ......................... 25 iv  2.1.7 Anion exchange chromatography ............................................................... 25 2.1.8 Electrophoresis and immunoblotting – post fractionation......................... 26 2.1.9 In-gel digestion and mass spectrometric analysis....................................... 27 2.2 Methods for chapter IV ............................................................................................. 28 2.2.1 Optimization of alloxan treatments ............................................................ 28 2.2.2 ICF Sample preparation and isoelectric chromatofocusing ........................ 29 2.2.3 Electrophoresis and immunoblotting ......................................................... 31 2.2.4 In-gel digestions and mass spectrometric analysis ..................................... 31 2.2.5 Immunoprecipitation .................................................................................. 31 CHAPTER III – Modulation of O-GlcNAc Yields and Anion Exchange Chromatography ............33 3.1 3.2 3.3 3.4  Determination of treatment conditions ................................................................... 33 Determination of lysis conditions ............................................................................. 36 Anion exchange chromatography ............................................................................. 38 Discussion .................................................................................................................. 43 3.4.1 Creating a model system for the study of O-GlcNAc modified proteins .... 43 3.4.2 Optimizing lysis conditions .......................................................................... 46 3.4.3 Anion exchange chromatography ............................................................... 47  CHAPTER IV – The Isoelectric Chromatofocusing (ICF) Platform.............................................52 4.1 4.2 4.3 4.4 4.5 4.6  Initial ICF studies ...................................................................................................... 52 ICF fractionation of insulin/STZ/glucose treated lysates .......................................... 59 ICF fractionation of untreated lysates ...................................................................... 64 ICF fractionation of alloxan treated lysates .............................................................. 72 Validation of method ................................................................................................ 79 Discussion .................................................................................................................. 81 4.6.1 Targeted improvements in chromatographic resolution ............................. 81 4.6.2 Analysis of insulin/STZ/glucose treated lysates............................................ 84 4.6.3 Usage of untreated and alloxan inhibited lysates as controls...................... 86 4.6.4 Problems which still need to be resolved ..................................................... 89 4.6.4.1 Sample complexity ........................................................................ 89 4.6.4.2 Sample yield .................................................................................. 91 4.6.5 Validation of results ...................................................................................... 92  CHAPTER V – General Conclusions and Future Directions ......................................................94 References ...........................................................................................................................97 v  LIST OF TABLES CHAPTER III Table 3.1 Lysis buffer compositions ............................................................................................. 37  CHAPTER IV Table 4.1 Proteins identified from insulin/STZ/ glucose treated C2C12 samples ....................... 62 Table 4.2 Proteins identified from ICF fractionation of untreated C2C12 samples .................... 70 Table 4.3 Proteins identified from ICF fractionated alloxan treated C2C12 samples ................. 77 Table 4.4 Comparative list of proteins identified between the three sample types................... 88  vi  LIST OF FIGURES CHAPTER I Figure 1.1  From the hexosamine biosynthesis pathway (HBP) to O-GlcNAc modified proteins .................................................................................................................................... 4  Figure 1.2  Structure of the O-GlcNAc modification .................................................................... 5  Figure 1.3  Inhibitors and substrates of O-GlcNAcase and OGT .................................................. 7  CHAPTER III Figure 3.1  Reduction of STZ concentration does not affect its efficacy ................................... 34  Figure 3.2  Time course to determine optimal treatment time................................................. 35  Figure 3.3  Time Course examining the effects of STZ addition................................................. 36  Figure 3.4  Comparison of differing lysis conditions on the recovery of O-GlcNAc modified proteins .................................................................................................................... 38  Figure 3.5  Anion exchange chromatography of treated C2C12 lysate ..................................... 39  Figure 3.6  Blot analysis of anion exchange separated fractions ............................................... 40  Figure 3.7  Determination of post anion exchange sample complexity via coomassie stained gels and western blotting ........................................................................................ 42  CHAPTER IV Figure 4.1  Isoelectric chromatofocusing with a standard gradient .......................................... 53  Figure 4.2  Isoelectric chromatofocusing with a concave gradient ........................................... 56  Figure 4.3  Isoelectric chromatofocusing with an extended gradient ....................................... 58  Figure 4.4  Analysis of selected fractions from a standard ICF separation ................................ 60  Figure 4.5  Functional breakdown of identified proteins ........................................................... 64  vii  Figure 4.6  Comparison between ICF chromatograms from O-GlcNAc enhanced and untreated samples .................................................................................................................... 66  Figure 4.7  Analysis of selected fractions from standard ICF of untreated C2C12 samples ...... 67  Figure 4.8  Dot blot analysis of analysis of ICF fractionated untreated C2C12 samples ............ 69  Figure 4.9  Time course and dose response experiments with alloxan ..................................... 73  Figure 4.10 ICF separations of treated, untreated and alloxan treated samples ....................... 74 Figure 4.11 Dot blot analysis of ICF fractionated alloxan treated C2C12 cells ........................... 75 Figure 4.12 Analysis of selected fractions from ICF fractionated alloxan treated samples ....... 76 Figure 4.13 Test immunoprecipitations with vinculin ................................................................ 80  viii  LIST OF ABBREVIATIONS  ALX  Alloxan  BAP  Biotin pentylamine  BCA  Bicinchoninic acid  BSA  Bovine serum albumin  CTD  Carboxy terminal domain  DTT  Dithiothreitol  ECL  Enhanced chemiluminesence  GAC  Streptocccal group A carbohydrate  GalNAc  N-acetly galactosamine  GFAT  Glutamine-fructose-6-phosphate amidotransferase  GlcNAc  N-acetyl glucosamine  HBP  Hexosamine biosynthesis pathway  HRP  Horse radish peroxidase  IB  Immunoblot  ICF  Isoelectric chromatofocusing  IP  Immunoprecipitation  LC-MS/MS  Liquid chromatography – tandem mass spectrometry  MCE  Muscle cell extract  MS  Mass spectrometry  OGT  O-GlcNAc transferase  O-GlcNAc  O-linked N-acetylglucosamine ix  O-GlcNAcase  N-acetyl-β-D-glucosaminidase  PBS  Phosphate buffered saline  PBS-T  Phosphate buffered saline + 0.1% tween20  PTM  Post-translational modification  PUGNAc  O-(2-acetamido-2-deoxy-Dglucopyranosylidene)amino-N-phenylcarbamate  PVDF  Polyvinylidene fluoride  QUIC-Tag  Quantitive isotopic and chemoenzymatic tagging  SDS-PAGE  Sodium dodecylsulphate polyacrylamide gel electrophoresis  STZ  Streptozotocin  TAS  Tagging-via-substrate  UDP-GlcNac  Uridine diphosphate N-acetylglucosamine  UMCE  Untreated muscle cell extract  WGA  Wheat germ agglutinin  x  ACKNOWLEDGEMENTS Firstly, I would like to acknowledge the guidance and support that I have received over the past three years from my supervisor Dr. Juergen Kast. I would like to thank him for introducing me to the wonderful world of protein chemistry. I would also like to thank Dr. Peter Schubert for providing my initial training and “showing me the ropes”. In addition, I would like to acknowledge the support of the Kast lab, in particular Dr. Brent Sutherland for providing thoughtful discussions and Jason Rogalski for providing assistance with the mass spectrometers. I would also like to send a big thank you to our collaborators Derek Choy and Dr. Charles Haynes for providing such a fantastic separation method. Finally, I would like to thank the wonderful support staff at the Biomedical Research Center for ensuring that I had all the supplies I required whenever I needed it.  xi  CHAPTER I – INTRODUCTION The study of Post-translational modifications (PTMs) has been and remains a major interest in the life sciences. Studies have shown that PTMs play an important role in mediating the biological activity of proteins within a cell. The modification of proteins with simple chemical groups can lead to a profound influence on the conformation of proteins and in most cases their functions. Examples of this include the conformational changes induced by Snitrosylation of myoglobin in blackfin tuna (1), the phosphorylation of the S6 kinase in mTOR signaling (2) and O-linked N-acetylglucosamine (O-GlcNAc) modification of IRS1/2 which leads to attenuated insulin signaling (3). Unfortunately such modifications are dynamically regulated and rapidly changing thus making their study difficult at best. This highlights the need to develop new methods that allow for better detection and identification of proteins which are post-translationally modified. Among many methods used in the study of PTMs the most commonly used include: antibody-based methods to detect and enrich modified proteins (4, 5), chemical derivatisation of the modification (6) and mass spectrometric methods (7). Each of these methods has its own advantages and disadvantages. However, one feature which is common to all these methods is that they predominantly target the modification itself. The goal of this thesis was to evaluate strategies which separate/enrich post-translationally modified proteins by targeting physical properties of the protein rather than the modification itself. The following sections will provide an overview of current literature concerning the modification of interest (O-GlcNAc), the  1  methods available to study this modification and the basics of the chromatographic techniques used in this thesis. 1.1 The O-GlcNAc modification 1.1.1 Functional roles of the O-GlcNAc modification Interest in understanding the O-linked N-acetylglucosamine (O-GlcNAc) modification, first discovered in 1984 by Torres and Hart (8), is due to the types of proteins on which it has been found. Since its discovery, O-GlcNAc modifications have been found on transcription factors, nuclear pore proteins, RNA-binding proteins, cytoskeletal proteins, stress induced proteins and other proteins involved in various cellular processes (9). The diversity of protein types implicates the functional importance of the O-GlcNAc modification in regulating cellular processes. Furthermore, the O-GlcNAc modification has been implicated in both diabetes and neurodegenerative diseases such as Alzheimer’s disease. In the context of diabetes, the O-GlcNAc modification plays a role in nutrient sensing and its dysregulation could potentially play a role in the pathology of diabetes. One relevant nutrient sensing pathway involved in glucose metabolism is the hexosamine biosynthesis pathway (HBP) (Figure 1.1). There has been a link established between the HBP and one of the hallmark symptoms of diabetes, insulin resistance (10). This link is important since one of the substrates involved in the addition of O-GlcNAc to a protein is Uridine Diphosphate Nacetylglucosamine (UDP-GlcNAc). This compound is the end product of the HBP and as such the amount of UDP-GlcNAc is dependent on the input of nutrients into the system. More importantly, the rate-limiting step of fructose-6-phosphate conversion to glucosamine-62  phosphate by glutamine-fructose-6-phosphate amido-transferase (GFAT) is what dictates the cellular levels of UDP-GlcNAc (11). UDP-GlcNAc has been shown to provide feedback inhibition to GFAT in its role in nutrient sensing (11). It has also been shown that insulin resistance is the result of a combination of high glucose, insulin and glutamine. There is evidence that, at least in some cases, this resistance might abolished by inhibiting GFAT (which uses glutamine and fructose-6-phosphate to form glucosamine-6-phoshate) (10). It is important to note that concerning insulin resistance, the level of UDP-GlcNAc produced is reduced and as such the OGlcNAc modification of proteins will be affected. Proteins such as IRS-1 (12), eNOS (13) and Sp1 (14) have all been shown to have their O-GlcNAc modifications affected by either insulin stimulation or the result of insulin resistance.  3  Glucose-6-phosphate isomerase  Glucose  Glucose-6-phosphate  Fructose-6-phosphate Glutamine  Glucosamine-fructose 6-phosphphate aminotransferase (GFAT)  Glutamate  Glucosamine-6-phosphate Glucosamine 6-phosphphate N-acetyl transferase  Acetyl-CoA CoA  N-acetylglucosamine-6-phosphate Phosphoacetylglucosamine mutase  N-acetylglucosamine-1-phosphate UDP-N-acetylglucosamine pyrophosphorylase  UTP PPi  GlcNAcstatin, PUGNAC, STZ  UDP-GlcNAc O-GlcNAc transferase  Alloxan  O-GlcNAcase  O-GlcNAc-Protein  GlcNAc + Protein  Figure 1.1 From the Hexosamine Biosynthesis Pathway (HBP) to O-GlcNAc Modified proteins. The HBP is a small branch involved in glucose metabolism. What is shown here is a simplified version the pathway indicating the position of the rate limiting step of GFAT (bold), the feedback inhibition due to UDP-GlcNAc, the enzymes involved (italicized) and their inhibitors (PUGNAC: O-(2-Acetamido-2-deoxy-D-glucopyranosylidene)amino N-phenyl carbamate and STZ: Streptozotocin) which are boxed in red. Figure adapted from (11, 15)  In the case of Alzheimer’s disease, it has been shown that the majority of proteins involved in this disease are O-GlcNAc modified. These proteins include tau (16) and the betaamyloid precursor protein (17). It has been suggested that the formation of tau tangles is the result  of  a  reduction  in  O-GlcNAc  modification  on  tau  and  the  subsequent  hyperphosphorylation of tau (18). The severity of Alzheimer’s disease has also been correlated  4  to glucose metabolism where decreased metabolism results in increased symptoms of the disease (19). Along with that, it has been shown that over-expression of O-GlcNAc Transferase increases the amount of GlcNAc modification while decreasing phosphorylation on tau (18) suggesting the importance of the O-GlcNAc modification and glucose metabolism in the development of Alzheimer’s disease. 1.1.2 The structure of O-GlcNAc The O-GlcNAc modification is unique compared to other types of sugar modifications found on intracellular and extracellular proteins. Most sugar modification of proteins are highly complex glycans such as those found on the mucin family of glycoproteins, while the O-GlcNAc modification is a single acetylated glucose molecule found on serine/threonine residues of proteins (Figure 1.2). Most glycans found on proteins are relatively static and unresponsive to environmental changes while the O-GlcNAc modification is dynamically regulated. OH  O  HO  O  HO AcNH  X-X-X-Ser/Thr-X-X-X  Figure 1.2 Structure of the O-GlcNAc modification. The O-GlcNAc modification is an acetylated glucosamine molecule typically found on the Serine/Threonine residues of proteins.  1.1.3 Enzymes involved in O-GlcNAc modification In contrast to protein phosphorylation which is regulated by an estimated 700 different kinases and phosphatases (20), there are only two enzymes which mediate the addition and 5  removal of the O-GlcNAc modification. One is N-acetyl-β-D-glucosaminidase (O-GlcNAcase) which catalyzes the removal of the O-GlcNAc modification from proteins. The second is the opposing enzyme O-GlcNAc transferase (OGT) which catalyzes the addition of UDP-GlcNAc to Ser/Thr residues on proteins. O-GlcNAcase is a heterodimer composed of a 54kDa (α) and a 51kDa (β) subunit (21). The α-subunit is considered to be the functional component with the Nterminus containing the catalytic domain and the C-terminus containing an acetyltransferase. This enzyme is highly specific towards GlcNAc and is nonreactive towards the GlcNAc epimer GalNAc (21). However, not much is known about the function of the acetyltransferase and how the enzyme is regulated. The OGT is a heterotrimer composed of two 110kDa (α) subunits and a single 78kDa (β) subunit (22). The α-subunit is considered to be the functional portion of this complex. Structurally, the N-terminus of this subunit contains 11 tetratricopeptide repeats (TPR) while the C-terminus contains the catalytic domain (23). It is believed that the TPR domain mediates protein specificity since it contains multiple hydrophobic sites which can be used for protein-protein interactions (24, 25). OGT itself is O-GlcNAc modified and phosphorylated. It is believed that these modifications play a role in its regulation (26). However, there is also not much known about the overall regulation of OGT, nor is there a known consensus sequence which can be used to predict potential modification sites. 1.1.4 Inhibitors of the O-GlcNAc enzymes The three known inhibitors of O-GlcNAcase are GlcNAcstatin (27), O-(2-acetamido-2deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate (PUGNAc) (28) and Streptozotocin (STZ), the inhibitor used in this study. STZ was chosen since it was the most cost effective of the  6  three inhibitors for use in large scale experiments. Traditionally, STZ has successfully been used to induce diabetes in animal models (29, 30). However, within the last ten years it has become evident that STZ treatment also induces an increase in O-GlcNAc modified proteins (31). Structurally, STZ is an analogue of GlcNAc as shown in figures 1.2a and 1.2b. It was shown that STZ does not competitively inhibit O-GlcNAcase but forms a transition state analogue which is more stable than that of GlcNAc and irreversibly modifies O-GlcNAcase (32). It was also shown that this inhibition lasts for more than one day under experimental conditions indicating the usefulness of this compound as an O-GlcNAcase inhibitor (33-35).  a)  b) OH  OH NH  O HO  CH3 O  OH  CH3 NH O  HO  Streptozotocin  N-acetyl Glucosamine d)  O  Uracil  O HN  NH N H  N O  OH  OH  c)  N  O  O  NH  O  O O Alloxan  Figure 1.3 Inhibitors and substrates of O-GlcNAcase and OGT. a) The O-GlcNAcase substrate NAcetylglucosamine (GlcNAc), b) the inhibitor Streptozotocin, c) the OGT substrate uracil, and d) the inhibitor alloxan. Note the structural similarities between the inhibitors and substrates. It is likely this similarity which contributes to the inhibitory effects observed.  7  Besides its use as an OGT inhibitor, alloxan is also a known diabetogenic drug and has been used to generate models of Type I diabetes in mice due to its implicated role in the destruction of β-cells (36). Alloxan is an analogue of uracil (Figure 1.2c) which is a component of the OGT substrate UDP-GlcNAc and the actions of alloxan are believed to be competitive (37). It is interesting to note that alloxan has been shown to be an inhibitor of O-GlcNAcase. However, it is not clear if the method of inhibition is either due to alloxan covalently modifying the active site of O-GlcNAcase or is just due to a non-covalent occupation of the active site (38). 1.1.5 Current methods available for the study of O-GlcNAc modified proteins The major tools which are currently available for the study of O-GlcNAc modifications are galactosyltransferase-based GlcNAc end-capping, antibody-based detection and purification, lectin-based purification and detection and chemical derivitisation of the modification for simpler detection and/or purification. As mentioned above, the target of these methods is the modification itself. In addition, no single method currently available is capable of analyzing OGlcNAc modified proteins with high specificity without altering the proteome in some fashion. 1.1.5.1 Galactosyltransferase-based methods This method involves the attachment of a radio-labeled galactose to the GlcNAc moiety using a bovine milk galactosyltransferase. Subsequently proteins containing this galactose can be detected using radiographic film. This was the method used by Torres and Hart to identify the presence of O-GlcNAc modified proteins (8). The weakness of this method however, is that galactosyltransferase is not specific towards the O-GlcNAc modification since it attaches galactose to any terminal GlcNAc (39). As such, the results obtained could potentially be 8  proteins that contain a terminal GlcNAc on a complex carbohydrate and not a true O-GlcNAc modification. 1.1.5.2 Antibody-based methods The simplest and most frequently used methods are antibody-based, using one of three antibodies available which recognize this modification. The antibodies are designated as HGAC85 (40), RL-2 (41) and CTD110.6 (42) and the latter two being the most commonly used. The antigen used in the production of the HGAC85 antibody was the streptococcal group A carbohydrate (GAC) (40). However, the antibody-O-GlcNAc interaction observed is the effect of cross-reactivity between the O-GlcNAc modification and the antibody which was originally developed to study GAC. In addition, this antibody has only been used in one study where the cross-reactivity was characterized (40). The RL-2 antibody was originally developed to study nuclear envelope proteins and it was later discovered that the epitope was specific towards the O-GlcNAc modification on these nuclear envelope proteins (43). Since then this antibody has been used in multiple studies to detect O-GlcNAc modified proteins (44-47). The CTD110.6 antibody is the only antibody available which was specifically generated against the O-GlcNAc modification. The antigen used to create this antibody was a synthetic peptide based on the carboxy terminal domain motif (CTD) of RNA polymerase II containing a single O-GlcNAc modified serine. The CTD110.6 antibody possesses high specificity towards the O-GlcNAc modification and shows no cross-reactivity with the epimer O-GalNAc. This antibody has also been used in western blots and protein purifications (48-50). Comer et al. also noted that this antibody detects a wider range of O-GlcNAc modified proteins than the RL-2 antibody, as was 9  seen in western blots comparing the two antibodies (42). The antibodies are highly specific when used for western blotting. However when used to purify modified proteins, as specificity is dependent on the stringency of the washing conditions, there is a possibility of false positives due to non-specific interactions with the modified proteins. 1.1.5.3 Lectin-based methods A third method used for the study of glycoproteins uses lectins, which are naturally occurring proteins that bind specific sugar moieties. For O-GlcNAc, this lectin is Wheat Germ Agglutinin (WGA). WGA has been used in many studies to detect O-GlcNAc modified proteins in addition to purification/enrichment of modified proteins (51, 52). However, the major problem with WGA is its selectivity. Although WGA recognizes and is fairly specific for GlcNAc moieties, it is not very selective for the location in which these moieties are located. As long as the target is a terminal uncapped glucose residue, it will bind to WGA. This moiety is also found on the ends of the branched glycosylations of some membrane or membrane associated proteins and thus these proteins will also be recognized by WGA. In addition, the avidity of WGA for GlcNAc moieties increase as the number of surrounding moieties increases. This implies that there is an increased possibility of false positives due to the fact that proteins containing large groupings of complex GlcNAc moieties in close proximity to each other will be preferentially purified rather than proteins containing a simple O-GlcNAc modification. 1.1.5.4 Chemical derivatisation based methods The final type of tool available for the study of O-GlcNAc modified proteins discussed here is the chemical derivatisation of the O-GlcNAc modification. The available strategies 10  includes β-elimination followed by Michael addition with dithiothreiotol (BEMAD) (53), taggingvia-substrate (TAS) (54, 55) and quantitive isotopic and chemoenzymatic tagging (QUIC-Tag) (56) Briefly, the BEMAD method subjects peptides from a digest of a protein preparation to gentle β-elimination which removes the O-GlcNAc group from the serine or threonine residue. Subsequently the peptides are subjected to a Michael addition which adds either dithiothreitol (DTT) or biotin pentylamine (BAP) to the β-eliminated peptide. The peptides are then purified by affinity chromatography using a thiol column or avidin column respectively, and the modified peptides are identified by LC-MS/MS. By looking for peptides which differ from the unmodified ones by a characteristic mass shift (136.2 Da for DTT, 310.5 Da for BAP), proteins which are O-GlcNAc modified can be identified. The advantage of this method is that it allows for relatively straightforward identification of modified proteins using mass spectrometry. However, the β-elimination step is still not specific for O-GlcNAc modified serine/threonine residues but will also remove serine/threonine phosphorylations. As such, the possibility of a false positive still exists (53). The TAS method involves culturing of cells in the presence of an azide-derivatised version of peracetylated GlcNAc which results in metabolically labeling O-GlcNAc modified proteins with azido-GlcNAc. The azido-GlcNAc is converted within the cell to UDP-azido-GlcNAc which is then transferred by OGT on to proteins. The proteins which contain the azido-GlcNAc can then be conjugated to a biotinylated phosphine capture reagent via a Staudinger ligation. The resulting proteins can either be purified using an avidin column or detected in a western blot using avidin-HRP. This “tag” allows for a highly specific purification since extremely  11  stringent washing conditions can be used due to the high affinity of the biotin-avidin interaction. As this method uses the endogenous cellular machinery, the proteins identified are most likely O-GlcNAc modified. However, since chemical reactions are not always complete, there might be proteins which are not labeled by the Staudinger ligation and therefore escape identification. QUIC-tagging is a chemical derivatisation method which can be used to identify and quantify O-GlcNAc modified proteins under different stimulation states. This method uses a genetically engineered galactosyltransferase to label O-GlcNAc modified proteins with ketonecontaining galactose. Proteins containing this group can then be biotinylated and purified using an avidin column. After this process, these proteins are digested and the peptides are labeled via a reductive amidation of primary amines and ε-amino groups on lysines with either a deuterated or non-deuterated form of formaldehyde/sodium cyanoborohydride. The peptides obtained from two different conditions (e.g. stimulated vs. unstimulated) are mixed and purified using an avidin column. Peptides which are O-GlcNAc modified can then be identified by LC-MS/MS and the relative amounts of modification can then be quantified in a manner similar to Stable Isotope Labeling by Amino Acids in Cell Culture (SILAC) (57). The disadvantage of this system however, is that it uses a galactosyltransfersase which transfers galactose to any terminal GlcNAc resulting in a low specificity. 1.2 Chromatographic separation of proteins 1.2.1 Size exclusion chromatography Chromatography is widely used in the life sciences to separate proteins based according to physical or chemical properties. Size exclusion, reversed phase and ion exchange 12  chromatography are among the most commonly used tools. Size exclusion chromatography separates proteins based on their size and shape. In this method, protein movement through the column is dependent on the pore size of the resin. As proteins pass through the column, smaller proteins elute from the column later because they are retained in the pores while larger proteins elute earlier because they are not retained in the pores. One advantage of this method is that the column conditions used are very gentle and results in proteins retaining their native forms and biological activity during separation. However, one disadvantage is that size exclusion chromatography is considered to be a low resolution technique as it is unable to separate proteins which are very similar in size (58). 1.2.2 Reversed phase chromatography Unlike size exclusion chromatography, reversed phase chromatography separates proteins on the basis of their surface hydrophobicity. The columns used are packed with beads that have hydrophobic groups coupled to their surfaces to provide the interaction sites. Proteins generally have pockets of hydrophobicity, the more hydrophobic pockets there are, the stronger the interaction of proteins with the beads. Thus when eluting with an increasing concentration of organic solvent, proteins which are less hydrophobic elute early on in the gradient while more hydrophobic proteins elute at a later time point. The hydrophobicity of the column is dependent on the length of the carbon chain group bound to the beads, longer chains being more hydrophobic than shorter chains. For protein separations a shorter chain (C4) is normally used while a longer chain (C18) is used for peptide separation. In terms of resolution and usage, reversed phase chromatography is the highest resolution and the most prevalent of  13  the chromatographic methods. Based on a literature search, the primary use of reversed phase chromatography ranges from small molecule separations (59) to peptide separations (60). Unfortunately in contrast to size exclusion chromatography, the usage of organic solvents in reversed phase chromatography leads to the denaturation of proteins. 1.2.3 Ion exchange chromatography Ion exchange chromatography is another chromatographic separation method which is commonly used in the biological sciences. It separates molecules on the basis of their overall surface charge and has been used in studies to purify small molecules and in the separation of proteins (61, 62). There are two major kinds of ion exchange chromatography primarily based on the charge of the chemical group bound on the beads in the column: cation and anion exchange. The chemistry behind cation exchange typically consists of groups which are negatively charged. As such, proteins which are to be bound need to have an overall positive charge when the sample is injected. This is typically achieved by adjusting the pH of the sample so that it is below the pI of the majority of proteins. What this does is create an environment where an excess of protons is present. Proteins under these conditions will be protonated and contain positive charges. The positively charged proteins will then bind to the negatively charged beads. In addition, the starting conditions must be at a low ionic strength regardless of the pH required or proteins will not bind. In anion exchange chromatography, the opposite holds true. The charges of the chemical groups covalently bound to the beads are positive. As such, protein interactions with 14  these chemical groups require that proteins be at a pH above the pI of the majority of proteins. These two types of ion exchange are further divided into strong and weak types. The difference between a strong and weak ion exchanger is the degree of change in ionization of the chemical groups under varying pH conditions. A strong ion exchanger will have a more constant charge state over a wider range of pH than a weak ion exchanger. In many cases, it is preferable to use a strong ion exchange column due to the charge stability of the chemical group on the beads. A typical ion exchange experiment consists of four main phases: column equilibration, sample loading, sample elution and column washing. The column equilibration step is to ensure that the column conditions are correct for sample loading. Typically, these conditions include the pH of the column environment and the salt/buffer concentrations within the column all of which are important for proper binding of the sample. In the sample application stage, the sample is loaded on the column and allowed to bind to the beads. In this phase, unbound proteins are washed away by passing several column volumes of start buffer through the column. The elution phase is when the proteins are removed from the column sequentially based on charge. Typically, there are two major ways of eluting proteins from an ion exchange column. The first method is by changing the pH within the column. This method is useful for the purification of proteins with a known pI. To do so, the pH of the elution buffer is changed so that proteins with pI values lower or higher (dependent on the column type) are eluted first. Then the pH can be adjusted so that only the protein of interest is eluted. This type of elution strategy however, is not useful when the goal of the experiment is to achieve separation of all the proteins in a proteome. The second type of elution method is competitive and based on increasing the concentration of a neutral salt (typically NaCl). Before the gradient begins, the 15  ionic strength of the buffer environment is low due to the absence of the salt. By applying a linear gradient of increasing salt concentration, the ions in the elution buffer compete with the bound proteins for the charged sites on the beads. In addition, the counter ions in the elution buffer will shield the protein from rebinding to the beads. As such, proteins with a low net charge will elute first as they are displaced by the counter ions and ones with higher net charge states at later stage. In the final step of washing, a high ionic strength wash buffer is applied to remove any proteins which may still be very tightly bound to the column. When evaluating the results of a separation, there are several terms to describe the quality of the separation which need to be defined. Resolution in chromatography is a measure of the relative separation between the maxima of two peaks compared to the average width of the two peaks. This value is governed by a combination of column selectivity and efficiency (63). Column selectivity is defined as the ability of a column to separate similar species from each other. Thus a column with high selectivity will show a greater separation between the maxima of the peaks observed. In addition, a highly selective column will have symmetrical peaks with very little overlap. The selectivity of the column is affected by the chemistry of the beads and the type of eluent used. The bead chemistry affects the selectivity since different chemistries will bind proteins in a slightly different manner. Some will bind preferentially to one form while others will have similar or less binding capacities to both affecting the elution characteristics. The eluent used also affects the selectivity in ion exchange primarily due to its pH and ionic strength. An eluent with a stronger ionic strength will be less selective than one with a weaker ionic strength. This is because the stronger eluent will overcome the ionic  16  interaction between the protein and column faster than a weaker eluent, and is unable to discriminate between small differences in ionic interactions. Column efficiency is defined by the ability to elute proteins into narrow peaks as indicated by a chromatogram. A highly efficient column will have narrow peak widths while a less efficient column will have wider peak widths. Efficiency is affected by the flow rate, column length and bead size. The flow rate affects the efficiency of the column since a lower flow rate allows for more time for a protein to pass through the column and with an increasing eluent concentration will lead to more sample component mixing and reduced efficiency. In a longer column the proteins in a peak take a longer time to exit resulting in greater peak overlap due to axial dispersion. This overlap results in mixing of the components in the two peaks and thus results in the reduced ability of the column to separate components. The bead size determines the amount of diffusion that can occur according to the amount of space between beads. A smaller bead size will pack more efficiently and thus decreases the space in which diffusion can occur. Bead packing also affects efficiency in that an unevenly packed column will allow for buffer channeling (uneven flow rates throughout the column) which will lead to peak broadening since proteins which should have been in the same peak are eluted at different rates (63). 1.2.4 Chromatofocusing The chromatofocusing method was first described in 1978 by Sluyterman et al. as a high resolution method of separating proteins in the liquid phase based on their isoelectric points (64, 65). This method is an adaptation of ion exchange chromatography whereby the salt 17  gradient is replaced with a pH gradient. This method works by changing the pH within the column gradually so when the pH within the column is at the isoelectric point of the protein, it will have a net charge of zero. As such, the protein will dissociate from the beads and move along with the buffer out of the column. If the protein migrates faster than the pH gradient, it will become negatively charged again and will be retained by the column until the section of the gradient with the proper pI reaches it (assuming the elution is from high pH to low pH). If the protein migrates slower than the gradient, it will become positively charged and will be repelled by the column and will migrate faster until it reaches its isoelectric point. This property of the separation strategy allows the proteins to “focus” into narrow peaks and thus a high resolution is obtained. The pH gradient can be generated by several methods. The most common method is to use polymeric ampholytes which provides a buffering capacity over a wide pH range. The most commonly available ampholyte for this purpose is the PolyBuffer system from GE Biosciences. As mentioned above, this is a proprietary blend of polymers which according to the manufacturer provides a linear pH gradient in the pH range of 4-7. This buffer has been used to separate and enrich proteins since the early 80’s (66-68). This buffer has also been used in a novel separation system in various studies to obtain protein profiles of cancer cells and in the search for serum biomarkers (69-71). As a multidimensional separation tool, this system separates on the basis of pI in the first dimension and hydrophobicity in the second. By having a two stage chromatographic separation system, the resulting samples are less complex and thus simpler to study. However, there is a significant disadvantage to using ampholytes. They have been found to associate with proteins, are expensive, and due to the complex synthesis process 18  have batch variations which makes reproduction of the separation difficult (72, 73). In addition, the pH buffering range provided by these ampholytes are typically limited to pH 7-4 and pH 9-7. In addition to the usage of ampholytes, traditional chromatofocusing uses a weak anion exchange column to assist in the generation of the pH gradient. The weak anion exchanger plays an important role in the generation of the pH gradient as the chemical groups on the beads are weakly charged and contain buffering capabilities. As such, the formation and properties of the pH gradient will be partially dependent on the buffering capacity of the resin. In this case, the gradient is formed by changing the pH in steps while allowing the buffering capacity of the column to assist in the gradient generation. This is referred to as an internally formed gradient. The disadvantage of this method is that low buffer concentrations are required since high buffer concentrations would overwhelm the buffering capacity of the column. This characteristic also leads to two potential shortcomings, one of which is the difficulty in which the gradient slope can be changed. By increasing buffer concentration, the slope can be increased. However, as mentioned previously, there is a limit to the amount of increase possible before the buffering effect of the column is overwhelmed. Due to this limitation, there is a limit to the amount of optimization which can be done (74). In addition, it has also been reported that linear pH gradients are difficult to generate at the low buffer concentrations required (75). As a result, the generated pH gradients often contain regions which are not linear. The use of a small molecule based buffer system would alleviate many of the problems associated with using ampholytes. Several studies have shown that it is possible to generate a  19  linear gradient without the use of these ampholytes. Bates and Frey showed that it is possible to generate a quasi-linear gradient by using stepwise increase in the elution buffer containing simple buffering species (72). However, due to the characteristics of the column packing material used, the generated pH ranges are quite limited in width (pH 9-7). Liu and Anderson showed that it is possible to combine an externally generated gradient with a column packing material which has a relatively wide buffering capacity to obtain a wider (pH 8-3.5) linear pH gradient capable of separating and focusing proteins (76). The Haynes group in the Department of Chemical Engineering at the University of British Columbia has developed an approach towards chromatofocusing which solves some of the aforementioned limitations (77). The method developed is slightly different in that it uses a strong anion exchange column rather than the commonly used weak anion exchange columns. By using a strong anion exchange column, the column is no longer has any buffering effect and the gradient is entirely external. The advantage of having a completely external gradient is that complete control over all aspects of the gradient is possible. Since there are fewer limitations on buffer concentration, it is possible to optimize slope of the gradient by adjusting the buffer composition and even develop a custom gradient shape. In addition, this method uses a cocktail of simple buffering species which are all volatile and easily removed to generate a linear pH gradient from pH 10 - 3.5. This means that proteins are not exposed to a denaturing environment and biological activity is retained with minimal processing required if further analysis is required to remove any potentially incompatible reagents from the sample.  20  1.3 Towards an alternative O-GlcNAc detection strategy This thesis will examine a novel approach towards the study of post-translationally modified proteins. The strategy is a two stage separation process designed to evaluate the possibility of identifying post-translationally modified proteins on the basis of the physical properties of proteins. The first stage is a means of proteome complexity reduction via physical properties of the proteins itself rather than interactions between the modification and a capture reagent. This is achieved using liquid chromatography. Following this, the fractions collected are screened to identify ones which contain O-GlcNAc modified proteins. Positive fractions are then further separated by molecular weight using gel electrophoresis. Subsequently, these gels are examined to identify individual O-GlcNAc containing bands. To develop this strategy, the following components were examined: Conditions required to increase the yield of O-GlcNAc modified proteins, classical anion exchange separation of proteins, conditions required to inhibit the addition of O-GlcNAc to proteins and the isoelectric chromatofocusing technique.  21  CHAPTER II – MATERIALS AND METHODS All chemicals and reagents unless noted otherwise were obtained from Sigma (St. Louis, MO) 2.1 Methods for chapter III 2.1.1 General cell culture conditions Mouse myoblast cells (C2C12) were kindly provided by Dr. Fabio Rossi (BRC, Vancouver, BC). Cells were cultured in DMEM (Invitrogen, Carlsbad, CA) supplemented with 2% Penicillin/Streptomycin and 10% Fetal Bovine Serum (37°C and 5% CO2). Cells were either passaged at 70% confluence or differentiated at 90% confluence with differentiation medium [MEMα (Invitrogen, Carlsbad, CA) supplemented with 5% Heat-Inactivated Horse Serum (Invitrogen, Carlsbad, CA) and 20% Penicillin/Streptomycin]. Cells were differentiated for three days before being subjected to various treatments as described below. 2.1.2 Determination of optimal STZ concentration Differentiated cells were washed two times with PBS (137mM NaCl, 10mM Phosphate 2.7mM KCl pH 7.4) and serum starved overnight in MEMα. Insulin (10nM), Streptozotocin (STZ) (2mM or 5mM), and glucose (12 g/L) were added and cells were incubated for 24 hours at 37°C 5% CO2. Cells were lysed with 200µL/dish in the V2 lysis buffer (50mM Tris pH 7.5, 150mM NaCl, 5mM MgCl2 and 0.1% v/v NP40). Whole cell lysates were then centrifuged at 14,000 RPM for 15 minutes to clarify the lysate. Protein concentrations were then determined using the BCA assay kit (Pierce) with a BSA standard curve. Subsequently 10, 30, and 50µg of the sample were analyzed by western blot as described in 2.1.5. 22  2.1.3 Optimization of detergent free lysis Similar to the previously described protocol (see 2.1.2), serum starved cells were treated in the presence of 2mM STZ and several different lysis conditions were tested. Standard lysis using buffer V2 (50mM Tris pH 7.5, 150mM NaCl, 5mM MgCl2 and 0.1% v/v NP40) plus a protease inhibitor cocktail (Roche, Laval, QC), detergent free lysis with normal salt levels using buffer V2F (50mM Tris pH 7.5, 150mM NaCl, 5mM MgCl2) plus a protease inhibitor cocktail and detergent free lysis conditions using buffer V3F (50mM Tris pH 7.5, 400mM NaCl, 5mM MgCl2) plus a protease inhibitor cocktail were tested. Cells lysed with the detergent free lysis buffer were homogenized in a Dounce Homogenizer using a type A pestle (100 strokes). Whole cell lysates were centrifuged at 14,000 RPM for 15 minutes to remove cell debris. Protein concentrations were determined using the BCA assay and 30µg of each sample were subjected to electrophoresis and immunoblotting as in 2.1.5. 2.1.4 Determination of optimal treatment length To determine the optimal length of time required for the treatment, cells were prepared as in 2.1.1 then treated with 10nm insulin, 2mM STZ and 66mM glucose for the following time points: No Treatment, 15 minutes, 30 minutes, 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 7 hours, 8 hours, 24 hours, 48 hours and 72 hours. After each time point, the cells were lysed using V3F lysis buffer and homogenized. 30µg of each sample were run using the electrophoresis and immune blotting conditions described in 2.1.5. Additionally these blots were stripped using a stripping buffer (100mM 2-mercaptoethanol, 2% w/v SDS and 62.5mM Tris pH 6.8) at 50°C for 30 minutes and followed by extensive washing with PBS plus 0.1% 23  Tween20 (PBS-T). The stripped membrane was blocked overnight at 4°C using a 5% BSA solution and re-probed using a 1:1000 dilution of an anti-β-actin (Sigma, St. Louis, MO) antibody for 1 hour at room temperature. The secondary antibody used was a 1:10000 dilution of a Goatanti-mouse IgG (Dako, Glostrup, Denmark) in PBS-T and was incubated for 1 hour at room temperature and then developed as described in 2.1.5. 2.1.5 Electrophoresis and immunoblotting – optimization steps Proteins along with the See-Blue Plus2 pre-stained marker (Invitrogen, Carlsbad, CA) were separated on 10% Tris-Glycine Polyacrylamide gels in 1X running buffer (25mM Tris, 192mM Glycine and 0.1M SDS) at 70 volts until the dye front passed the stacking gel, then the voltage was increased to 90 volts and run until the dye front reached the bottom of the gel. Proteins were transferred onto PVDF (Pall, East Hills, NY) at 23 volts for 1 hour in a 1X Semi-dry transfer buffer (48mM Tris, 39mM Glycine, 13mM SDS and 20% v/v Methanol) using a semi-dry transfer apparatus (Bio-Rad, Hercules, CA). Following transfer, the membranes were blocked with a 5% BSA solution in PBS-T overnight at 4°C. The primary antibody CTD110.6 (1:5000 dilution) (Covance, Denver, PA) was diluted in PBS-T. Membranes were incubated with rocking in the primary antibody for 1 hour at room temperature. Membranes were rinsed 5 times for 5 minutes in PBS-T. Membranes were incubated with horseradish-peroxidase coupled goat anti-mouse IgM secondary antibody (1:5000) (Sigma, St. Louis, MO) for 1 hour with rocking at room temperature, washed 5 times for 5 minutes in PBS-T. The membranes were developed using an Enhanced Chemiluminesence (ECL) (GE Healthcare, Buckinghamshire, UK) kit and subsequently exposed to high performance 24  chemiluminesence film (GE Healthcare, Buckinghamshire, UK) for 10s, 30s, 1 minute and overnight exposures. 2.1.6 Sample preparation for anion exchange chromatography Differentiated cells treated with 10nM insulin, 2mM STZ and 66mM of glucose for 24 hours were lysed using V3F lysis buffer. Whole cell lysates were homogenized using 100 strokes in a dounce homogenizer and clarified by centrifuging at 14,000 RPM for 15 minutes. Protein concentrations were determined using the BCA assay. To prepare samples for anion exchange chromatography, 2mg of the clarified lysate were desalted using a Pall NanoSep Omega microcentrifugal (Pall, East Hills, NY) spin column with a 10kDa molecular weight cutoff. Sample volumes were reduced until there was approximately 100µL of sample remaining in the well and 400µL of 20mM Tris at pH 8.2 was added. This process was repeated twice to ensure that the sample had been completely buffer exchanged. The final volume of the protein sample was 100µL. 2.1.7 Anion exchange chromatography Following desalting and concentration, samples were injected onto a MiniQ 4.6/50 anion exchange column (GE Healthcare, Buckinghamshire, UK). An Ettan MDLC (GE Healthcare, Buckinghamshire, UK) system was used for the anion exchange separations. Flow-through from the first five minutes after injection was collected and immediately following, a 20 column volume (20ml) gradient from 20mM Tris pH 8.2 (Buffer A) to 20mM Tris plus 0.5M NaCl pH 8.2 (Buffer B) was collected. Fractions were collected every 20 seconds for an approximate fraction  25  size of 333µL at a flow rate of 1mL/min. Fractions were collected into a 96 well plate. At the end of the gradient, the buffer was kept at 100% buffer B for 5 minutes so that proteins still bound to the column were removed. 2.1.8 Electrophoresis and immunoblotting – post fractionation Following IEX separation, individual fractions were tested for the presence of O-GlcNAc modification by dot blot. 12µL of each fraction was spotted onto a pre-wetted PVDF membrane and blocked with 5% BSA overnight at 4°C. Blots were then probed with the CTD110.6 antibody (1:5000 dilution) for 1 hour at room temperature, washed 5 times at 5 minutes each with PBS-T, probed with anti-mouse IGM-HRP for 1 hour at room temperature, developed with ECL and exposed to film (GE Healthcare, Buckinghamshire, UK) for 10s, 30s, 1 minute and overnight exposures. Fractions were transferred into 1.5mL Eppendorf tubes and concentrated using a heated vacuum centrifuge at 60°C. Sample volumes were reduced to 40µl to which 10µl of 5X Laemmelli’s sample buffer (10% SDS, 50% Glycerol, 200mM Tris, 752mM 2-Mercaptoethanol and Bromophenol Blue) was added to the sample. Samples were boiled for 10 minutes prior to loading on a 10% Tris-Glycine Polyacrylamide gel. Protein samples were split into a 60:20:20 ratio with 60% of the sample used for Coomassie staining, 20% for western blotting and the remaining 20% for a secondary antibody only control. Electrophoresis was performed as described in 2.1.5. The gel with 60% of the sample was stained with Coomassie Blue (0.1% Coomassie Brilliant Blue R250, 40% v/v methanol and 1% v/v Acetic Acid) for 1 hour and destained with 10% v/v Acetic Acid until the gel background was clear. 26  2.1.9 In-gel digestion and mass spectrometric analysis Gel bands were excised and cut into 1mm cubes. Gel particles were washed with water and shrunk with acetonitrile. Shrunken gel bands were swollen in 10mM dithiothreitol/0.1M ammonium bicarbonate and incubated at 56°C for 30 minutes to reduce any remaining disulphide bonds. Following reduction, gel particles were shrunk with acetonitrile and reswelled with 55mM iodoacetamine/0.1M ammonium bicarbonate. Gel particles were incubated in iodoacetamide for 30 minutes at room temperature in the dark to alkylate proteins in the gel bands. Following alkylation, gel particles were washed with 0.1M ammonium bicarbonate prior to being shrunk with acetonitrile again. To the shrunken gel bands, trypsin (Promega, Madison, WI) suspended in 50mM ammonium bicarbonate was added at a concentration of 0.1µg per band. This mixture was incubated at 37°C overnight. To extract the peptides, any digestion liquid remaining in the tubes were collected in separate tubes and the gel particles were washed with 20µL of 25mM ammonium bicarbonate for 15 minutes at 37°C with shaking. Following washing, 80µL of acetonitrile was added incubated at 37°C for 15 minutes. The acetonitrile/ammonium bicarbonate mixture was then removed and transferred to a separate tube. 40µL of 5% formic acid was added to the gel particles, and incubated at 37°C for 15 minutes. Subsequently 100µL of acetonitrile was added and incubated for 15 minutes. Following acetonitrile treatment, the liquid in the tubes were then pooled with the previously collected liquids and dried in a vacuum centrifuge at 55°C. Extracted peptides were then reconstituted with 7µL of 5% formic acid in preparation for MS analysis.  27  Reconstituted samples were analyzed by LC-MS/MS using the Q-Star (Applied Biosystems, Foster City, CA) hybrid quadrupole time of flight instrument. Peptides were separated by reversed phase chromatography using an Famos/Ultimate LC system (Dionex, Oakville, ON) with the following gradient conditions: (Buffer A, 0.1% trifluroacetic acid 5% acetonitrile; Buffer B 0.1% trifluoroacetic acid 80% acetonitrile) Wash at 2% buffer B for 10 minutes, elute from 2-20% buffer B for 50 minutes, increase buffer B to 90% over 5 minutes, hold at 90% buffer B for 5 minutes at a flowrate of 0.2µL/min prior to being analyzed by the mass spectrometer using a nanospray source. The MS/MS sequencing results were then matched to theoretical sequencing data using the Mascot database with a peptide tolerance of ± 0.15Da, MS/MS tolerance of ± 0.5Da, 1 missed cleavage site allowed, carbamidomethyl as a fixed modification and oxidation as a variable modification. 2.2 Methods for chapter IV 2.2.1 Optimization of alloxan treatments To determine the optimal amount of alloxan required, cells were cultured as in 2.1.1. Differentiated C2C12 cells were then serum starved overnight in MEMα (37°C and 5% CO2) prior treatment with 0, 10µM, 100µM, 500µM, 1mM 5mM and 10mM of alloxan for 1 hour. Cells were lysed with 200µL of V3F lysis buffer/dish. Whole cell lysates were homogenized in a dounce homogenizer (100 strokes, type A pestle) and clarified by centrifuging at 14,000 RPM for 15 minutes. Protein concentrations were determined using the BCA assay kit. 30µg of each sample was analyzed by SDS-PAGE using the CTD110.6 antibody as described in 2.1.5.  28  Additionally, membranes were stripped and re-probed using the anti-β-actin antibody as in 2.1.4. To determine the optimal length of treatments differentiated C2C12 cells were treated with 5mM alloxan for 0 minutes, 15 minutes, 30 minutes, 1 hour, 1.5 hours, 2 hours, 4 hours and 24 hours. Cells were then lysed with 200µL of V3F lysis buffer. Clarified samples were then treated and analyzed by immunoblotting as in 2.1.5. 2.2.2 ICF Sample preparation and isoelectric chromatofocusing Differentiated C2C12 cells were treated with insulin, STZ and glucose as described in 2.1.6. Additional treatments were as follows: differentiated C2C12 cells were serum starved in MEMα (37°C and 5% CO2) overnight and either left untreated and lysed or treated with 5mM alloxan for 4 hours prior to lysis with 200µL of V3F lysis buffer/dish. Both additional treatments were clarified by centrifugation at 14,000 RPM for 15 minutes and protein concentrations determined using the BCA assay. To prepare the sample for chromatography, samples were desalted and buffer exchanged into start buffer by applying 16mg of clarified sample onto a PD-10 desalting column (GE Healthcare, Buckinghamshire, UK) according to manufacturer’s instructions using isoelectric chromatofocusing (ICF) start buffer as the eluent. Following desalting, samples were concentrated to a final volume of 10mL using a 50mL Millipore centrifugal device (Millipore, Billerica, MA) with a 10kDa protein cutoff membrane. Following concentration, samples were first passed through a 0.45µm syringe filter (Pall, East Hills, NY) to remove the large particulates,  29  then passed through a 0.22µm syringe filter (Pall, East Hills, NY) to remove any remaining particulate matter. Samples were then injected onto a MonoQ 10/100GL Strong anion exchange column (GE Healthcare, Buckinghamshire, UK) for isoelectric chromatofocusing (ICF). An ÄKTA Explorer (GE Healthcare, Buckinghamshire, UK) system was used for the separations. Initial characterization studies used a linear pH gradient from pH 10 to 3.5 established by mixing Buffer A (10mM 1,3 Diaminopropane, 10mM Diethanolamine, 10mM Tris, 10mM Imidazole, 10mM Bis-Tris and 10mM Piperazine pH 10.0) and Buffer B (10mM 1,3 Diaminopropane, 10mM Diethanolamine, 10mM Tris, 10mM Imidazole, 10mM Bis-Tris, 10mM Piperazine 10mM Acetic Acid, and 10mM Lactic Acid pH 3.5) at a flow rate of 1mL/min and over a period of 60mL. The chromatography program was as follows: one 20mL fraction was collected post-injection, ninety-six 900µL fractions was collected for the duration of the gradient and one 20mL fraction was collected from a high salt wash after the gradient was completed. After initial characterization with the linear gradient, a concave gradient (Buffer A: 5mM 1,3 Diaminopropane, 5mM Diethanolamine, 5mM Tris, 10mM Imidazole, 5mM Bis-Tris, 5mM Piperazine Buffer B: 50mM Diaminopropane, 50mM Diethanolamine, 50mM Tris, 50mM Imidazole, 50mM Bis-Tris, 50mM Piperazine, 10mM Acetic Lactic) was tested to improve resolution in areas of interest. The same chromatography program as in the linear gradient was used. In addition, an extended gradient using standard gradient buffers but with the gradient length doubled was also examined.  30  2.2.3 Electrophoresis and immunoblotting Dot blots were completed as in 2.1.8. Subsequently individual samples were concentrated and processed for electrophoresis and immunoblotting as in 2.1.8. 2.2.4 In-gel digestions and mass spectrometric analysis From the Coomassie stained gels corresponding to the insulin/STZ/glucose treated samples, regions equivalent to regions where there were signals on the western blot were excised, digested and analyzed by mass spectrometry as described in 2.1.9. Regions equivalent to the ones excised from the insulin/STZ/glucose gels were also excised and processed from untreated samples and alloxan treated samples as well. 2.2.5 Immunoprecipitation Differentiated C2C12 cells were treated for 24 hours with 10nm insulin, 2mM STZ and 12g/L glucose before being lysed with 200µL of V3F lysis buffer/dish. Protein concentrations were determined using the BCA assay kit. Prior to the addition of the antibody, protein samples were diluted with a salt free version of the V3F lysis buffer so that the final salt concentration was at 100mM NaCl. 5µg of anti-vinculin antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was added to 1mL of the diluted protein sample (1mg) and 5µg of isotype matched 9E10 (BRC, Vancouver, BC) antibody was added to another tube with 1mL of the protein sample. Tubes were incubated overnight at 4°C while rotating. To the incubated lysate, 20µL of packed Protein G sepharose (Roche, Laval, QC) was added and incubated for 4 hours at 4°C. After incubation, beads were spun down at 14000 RPM for 30 seconds and the supernatant removed. Beads  31  were then washed three times with the V2F lysis buffer at 1mL per wash and 10 minutes each. After the washes, beads were spun down and the supernatant removed. To the beads, 40µL of 2.5X Laemmelli’s Sample buffer was added and the samples boiled for 5 minutes. Samples were then divided evenly into two portions and resolved on a 10% gel. Proteins were transferred to a PVDF membrane, and blocked overnight with a 1:5000 dilution of CTD110.6 and subsequently with a goat-anti-mouse IgM HRP secondary antibody at a 1:5000 dilution. Following developing, membranes were stripped using stripping buffer as in 2.1.4 and re-probed with a 1:1000 dilution of the anti-vinculin antibody and a goat-anti-mouse IgG HRP (1:10,000 dilution) secondary was used.  32  CHAPTER III – MODULATION OF O-GLCNAC YIELDS AND ANION EXCHANGE CHROMATOGRAPHY 3.1 Determination of treatment conditions Under standard physiological conditions, the levels of O-GlcNAc modified proteins are substoichiometric. Therefore, to study the feasibility of using a chromatographic separation scheme to identify O-GlcNAc modified proteins, a model system to artificially increase the amount of O-GlcNAc modified proteins within cultured muscle cells was designed and optimized. Previous studies indicated that the presence of high glucose and insulin can stimulate O-GlcNAc modification of proteins within cultured muscle cells (14). Furthermore, it has been established that streptozotocin (STZ), an enzyme inhibitor, can inhibit O-GlcNAcase by preventing the removal of the GlcNAc group from modified proteins (35). As such, determination of optimal conditions for the usage of a combination of these three reagents was conducted and results are presented in the following sections. The typical concentration of STZ used in inhibition studies is 5mM. However, STZ is a costly reagent. As such, it was preferable to determine if it was possible to achieve a similar result with lower concentrations of of inhibitor and the addition of insulin and high glucose concentrations. Differentiated C2C12 cells were treated with 10nM insulin, 66mM glucose and either 2mM or 5mM of STZ for 24 hours prior to being lysed and analyzed. As shown in Figure 3.1, the reduction of STZ from 5mM to 2mM does not show an appreciable decrease in the amount of proteins which are O-GlcNAc modified. More importantly, this data also shows an appreciable increase in both the intensity and number of O-GlcNAc modified proteins detected  33  in STZ, insulin and glucose treated cells as compared to untreated cells. Because of these findings, subsequent experiments employed 2mM STZ rather than 5mM STZ.  kDa 250 148 98 64 50  IB: CTD110.6  36  22  IB: β-actin Figure 3.1 Reduction of STZ concentration does not affect its efficacy. Differentiated C2C12 cells were treated with either 2mM or 5mM STZ in addition to 10nM insulin and 66mM glucose. Cells were lysed using the following buffer 50mM Tris pH 7.5, 150mM NaCl, 5mM MgCl2 and 0.1% v/v NP40. Equal amounts of lysates were probed using an O-GlcNAc specific antibody (CTD110.6). Blots were then stripped and re-probed with an anti-β-actin antibody to verify equal loading. Blots were replicated at n=2 with similar results observed.  To examine how exposure time influenced O-GlcNAc modifications, differentiated C2C12 cells were serum starved overnight and the cells treated with insulin, STZ, and glucose for 15 minutes, 30 minutes, 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 7 hours, 8 hours, 24 hours, 48 hours and 72 hours or left untreated as a control. As shown in Figure 3.2 a gradual increase in the amount of O-GlcNAc modified proteins from 0 to 24 hours was seen. However, longer treatment times resulted in minimal differences in the banding patterns when compared 34  to those at 24 hours. Equal input was verified using β-actin on the membrane that had been stripped re-probed using an anti-β-actin antibody.  kDa 250 148 98 64 50  IB: CTD110.6 36  22  IB: β-actin  Figure 3.2 Time Course to determine optimal treatment time. Differentiated C2C12 cells were treated with 2mM STZ, 10nM insulin and 66mM Glucose for 15’, 30’, 1hr, 2hr, 3hr, 4hr, 5hr, 6hr, 7hr, 8hr, 24hr, 48hr, 72hrs or not treated. 30 µg of each sample was resolved by SDS-PAGE (10%) and probed using an O-GlcNAc specific antibody (CTD110.6). Blots were then stripped and re-probed with an anti-β-actin antibody to verify equal loading. Blots were replicated at n=2 with similar results observed.  To examine how effective STZ was at inhibiting O-GlcNAcase, differentiated C2C12 cells were serum starved as in previous experiments and treated with 10nM insulin, 66mM glucose for 0 to 72 hours in a) the presence of 2mM STZ, b) the absence of STZ. From the western blot (Figure 3.3) it was determined that without STZ, the insulin and glucose caused an accumulation 35  of O-GlcNAc modified proteins over a period of 24 hours. However, as shown in Figure 3.3, there was a noticeable increase in the intensity and number of O-GlcNAc modified proteins in all time points that contained STZ when compared to the ones that did not receive any STZ. As in previous experiments, actin was used to verify equal loading.  b)  a)  kDa 250 148  kDa 250 148  98 64 50  98 64  36  36  22  22  IB: CTD110.6  50  IB: β-actin  Figure 3.3 Time Course examining the effects of STZ addition. Differentiated C2C12 cells were treated with 10nM insulin, 66mM glucose and a) with 2mM STZ for the indicated times and b) without 2mM STZ for the indicated times. 30µg of each sample was resolved by SDS-PAGE (10%) and the blots probed using the O-GlcNAc specific antibody CTD110.6. Blots were subsequently stripped and re-probed using an anti-β-actin antibody to verify equal loading. Blots were replicated at n=2 with similar results observed.  3.2 Determining lysis conditions Certain types of detergents are compatible with ion exchange columns but often times its usage requires additional sample processing or column cleaning after usage. As such, conditions which would allow for cell lysis while avoiding the use of detergents in the lysis buffer were tested. The downside to using a detergent free lysis method is that it is generally  36  less efficient than one which uses detergents. The lysis conditions used in previous experiments were the standard V2 buffer (refer to Table 3.1 for compositions) used by the Kast group which contains 0.1% NP40 and 150mM NaCl. Initially it was thought that by removing detergent from this lysis buffer to create the V2F lysis buffer it would be possible to obtain similar yields of OGlcNAc modified proteins. This proved to be not the case as seen in Figure 3.4a. There was a noticeable decrease in the band intensity when comparing the detergent free lysis to the detergent containing conditions. To circumvent this problem, the concentration of NaCl was increased to 400mM to create the V3F lysis buffer and an additional homogenization step was added to aid in disruption of the cell membrane. In this case, it was observed that although there was a slight difference in band intensity when comparing the V2 lysed sample to the V3F lysed sample (Figure 3.4b), the overall decrease is minimal when compared to the difference observed in Figure 3.4a. This indicated that by increasing the salt concentration and the addition of a homogenization step, it was possible to obtain results similar to previous experiments without resorting to the use of detergents in the lysis buffers.  Table 3.1 Lysis buffer compositions V2  V2F  V3F  15% Glycerol 0.1% NP40 50mM Tris pH 7.5 150mM NaCl 5mM MgCl2 Protease inhibitors  15% Glycerol 0% NP40 50mM Tris pH 7.5 150mM NaCl 5mM MgCl2 Protease inhibitors  15% Glycerol 0% NP40 50mM Tris pH 7.5 400mM NaCl 5mM MgCl2 Protease inhibitors  37  a)  kDa 250 148  b)  kDa 250 148  98  98  64  64  50  IB: CTD110.6  IB: CTD110.6  50  36 36 22 22  IB: β-actin  IB: β-actin  Figure 3.4 Comparison of differing lysis conditions on the recovery of O-GlcNAc modified proteins. a) Treated C2C12 cells were lysed with either detergent containing lysis buffer with 150mM NaCl and 0.1% v/v NP40 (V2) or a detergent free lysis buffer with 150mM NaCl (V2F). 30µg of each sample was resolved by SDS-PAGE (10%) and probed with the O-GlcNAc specific antibody CTD110.6 and re-probed using a β-actin antibody to verify equal loading. b) Cells were treated as in panel a) but were lysed with the V2 lysis buffer and a high salt (400mM) containing version of the detergent free V2F lysis buffer we label as V3F. 30µg of each sample was resolved by SDS-PAGE and probed with the O-GlcNAc specific antibody CTD110.6 and re-probed using a β-actin antibody to verify equal loading. Blots were replicated at n=2 with similar results observed.  3.3 Anion exchange chromatography The initial separation studies were conducted on a standard anion exchange column using a Tris buffer base and sodium chloride as an eluent. Figure 3.5 shows a representative chromatogram obtained from anion exchange runs. During the 5 minute sample application phase (buffer concentration is 100% A) a peak indicating proteins in the sample which do not bind to the column these conditions was seen. From the 5 minute to 25 minute time point (salt gradient begins and ends), there are seven observable peaks within the chromatogram 38  indicating the fractions which contain proteins. From the 25 to 30 minute time point (100 percent B) proteins which may still bound to the column were washed off. As seen in figure 3.5 there were no significant peaks in this region indicating that all proteins had been eluted. Comaprison between runs of C2C12 Anion Exchange Separations 0.22  Washing  Sample Elution  Binding 0.2  Run 1 Run 2 Fractions  Absorbance (UV 562nm)  0.18  0.16  0.14  0.12  0.1  A1  B1  C1  D1  E1  F1  H1  G1  0.08 0  5  10  15  20  25  30  Time (minutes)  Figure 3.5 Anion exchange chromatography of treated C2C12 lysate. (a) 2mg of treated C2C12 lysate was processed and injected onto a MiniQ strong anion exchange column and the following program was used following sample injection: Buffer A (20mM Tris pH 8.2) for 5 minutes a gradient of 0-100% Buffer B (20mM Tris, 0.5M NaCl pH 8.2) over 20 minutes 100% Buffer B for 5 minutes to wash out unbound material all at a flowrate of 1mL/min and fractions were collected every 20 seconds. Proteins were detected post separation using the BCA assay to generate the above UV trace measured at 562nm. Chromatograms from two runs completed under identical conditions are overlaid to illustrate the reproducibility of the separations. The three distinct phases in the chromatography program are shaded in grey. Runs were replicated at n=3.  39  Portions of the individual fractions were spotted onto PVDF to be analyzed using a dot blot. As shown in Figure 3.6 there were three to four fractions in the sample application phase where O-GlcNAc modified proteins did not bind to the column (fractions A2 –A5). Modified proteins in the elution phase were observed eluting one minute after the gradient started (fraction B7) and continued until the gradient ended (fraction G6). Furthermore, consistent with the lack of UV signals in the high salt wash region, fractions G7 – H9 did not show evidence of modified proteins. 1  2  3  4  5  6  7  8  9  10  11  12  A B C D E F G H  Figure 3.6 Dot Blot analysis of anion exchange separated fractions. To determine which fraction contains O-GlcNAc modified proteins, 12µl of each fraction was spotted and probed with the O-GlcNAc specific antibody CTD110.6. O-GlcNAc positive fractions show up as a spot. The binding phase fractions are boxed in ( ), the sample elution phase fractions are boxed with ( ) and the washing phase fractions are boxed in ( ). Blots were replicated at n=3 with similar results observed.  40  Given that modified proteins were evident throughout the entire range of the salt gradient, all the protein fractions were concentrated for analysis by Coomassie staining and western blot. From the Coomassie stained gels (Figure 3.7a) the groupings of bands which show a rise and drop in intensity at identical molecular weights were spread out over 5-7 fractions indicating a large peak width. From the corresponding western blots (Figure 3.7b), a similar distribution of O-GlcNAc modified proteins was observed. However, the signals from the western blots required a long exposure to observe any signal. These levels proved to be too low to provide suitable amounts of protein for mass spectrometric analysis  41  a)  b) kDa 200 116.3 97.4 66.3 55.4 36.5  kDa 250 148 98 64 50 36  31.0 22 kDa 200 116.3 97.4 66.3 55.4 36.5  kDa 250 148 98 64 50 36  31.0 22  kDa 200 116.3 97.4 66.3 55.4 36.5  kDa 250 148 98 64 50 36  31.0 22  kDa 200 116.3 97.4 66.3 55.4  kDa 250 148 98 64 50 36  36.5 31.0 22  Figure 3.7 Determination of post anion exchange sample complexity via Coomassie stained gels and western blotting. a) 60% of each concentrated fraction was resolved by SDS-PAGE (10%) and stained for 1 hour using Coomassie blue. Boxed in red are a series of protein bands which correspond to the average peak width as observed from the gel. Fractions E9 – H12 are not shown since there are no proteins detectable by Coomassie staining. b) Western blots probed with the CTD110.6 antibody corresponding to the Coomassie stained gels are shown. Boxed in red are a series of O-GlcNAc bands which correspond to the ones boxed in panel a). An extended exposure was required to obtain the O-GlcNAc signals shown. Experiments were replicated at n=2. 42  3.4 Discussion 3.4.1 Creating a model system for the study of O-GlcNAc modified proteins Previous studies have shown that the addition of insulin causes an enhanced uptake of glucose by skeletal muscle cells in vivo through the GLUT4 transporter (78). The uptake of glucose by the cell sets into action a variety of biochemical pathways and of particular interest for this study, the hexosamine biosynthetic pathway where the end product is UDP-GlcNAc. Since UDP-GlcNAc is the substrate which O-GlcNAc transferase uses as the GlcNAc donor group, the addition of insulin and the subsequent glucose uptake is one of the key components required to obtaining increased yields of O-GlcNAc modified protein within the cell by increasing the available levels of OGT substrate. However, the addition of insulin alone was not sufficient to obtain increased yields of modified proteins; the glucose concentration present in the culture medium was also important. The MEM-α media used during differentiation and treatment contained a physiological glucose concentration of 5.5mM, and so the glucose concentration was increased to create a hyperglycemic environment. What this should have done in conjunction with the effects of insulin, was to stimulate glucose uptake. However, previous studies have shown that the C2C12 cell line was not responsive to insulin mediated GLUT4 translocation (79) and correspondingly glucose uptake through the GLUT4 transporter. While a later study showed that differentiated C2C12 cells do have the correct machinery to properly translocate GLUT4 when treated with insulin albeit at a lower level (80). In the same study, Nedachi and Kanzaki also suggested that this translocation was difficult to detect due to a masking effect by other glucose transporters 43  when using a traditional 2-deoxyglucose assay. It was also suggested that although GLUT4 translocation is less in C2C12 cells when compared to in vivo systems, it still plays a role in glucose transport within the cell. It appeared that this reported reduction of glucose transport did not have an effect on intracellular O-GlcNAc modified proteins as was shown in figure 3.3b. Although there appeared to be increased levels of O-GlcNAc modified proteins obtained with just insulin and glucose treatments, the O-GlcNAcase inhibitor STZ was additionally used as a means to maximize these levels. STZ has traditionally been considered a β-cell toxin and used in studies to artificially induce diabetes in animal models. It was discovered later that STZ is also an O-GlcNAcase inhibitor and that it can be used to promote accumulation of O-GlcNAc modified proteins within cells. However, most studies using STZ to study O-GlcNAc modifications have used primary rat β-cells and only a few studies have used cultured cell lines such as liver cells (81), glial cells (82, 83), and mesangial cells (84) but never muscle cells. Common to those studies was that they all successfully used a concentration of 5mM STZ to obtain accumulation of O-GlcNAc modified proteins. It must also be noted that in β-cells, O-GlcNAc accumulation was observed even with 1mM STZ. This is quite surprising as βcells are known to have the highest intracellular concentration of OGT observed and it would be expected that a higher concentration would be required to obtain adequate amounts of inhibition as STZ is not a very strong inhibitor when compared to the other OGT inhibitors (31,  85). What this indicated however, is that 5mM STZ would be a suitable starting concentration and that reducing the STZ concentration to 2mM would not be a problem. The reduction was preferred as it would be more cost effective to use a lower dose if it could provide a similar  44  effect as a higher dose. Indeed, as seen in figure 3.1, similar levels of modification were obtained when comparing the effects of 5mM STZ versus 2mM STZ. The observed increase in levels of O-GlcNAc modified proteins was the effect of the three different reagents combined. The effects of insulin and glucose may have been minimal but there are several factors which can account for the effects seen in figure 3.3. The first being that serum starvation prior to treatment will induce glucose uptake when insulin is added (86). Since the protocol called for serum starvation of differentiated cells for 18 hours prior to treatment, this would likely elevate the amount of glucose uptake as reported. It is also possible that the starvation was creating a nutrient depleted environment similar to one which muscle fibres would experience post-exercise. In this type of environment, as discussed by Hamada et al, a sub-maximal dose of insulin causes an increase in glucose transport; an effect which was required by these studies (87). As differentiated cells were treated with a low dose of insulin (10nM), it was possible that this effect contributed to the accumulation of modified proteins which was observed. In addition, it was speculated that in the in vitro culture system, insulin which was added is not degraded as it would be in the circulation. Therefore, insulin would stay within the medium until it was used up or removed. Therefore, it was likely that a lingering sub-maximal dose of insulin was causing a minimal but sustained stimulatory effect on the cells. As seen in figure 3.3 the removal of STZ during treatment of differentiated C2C12 with insulin and glucose in a time course, resulted in an observable accumulation of modified proteins over the course of the treatment times while loading controls verified equal loading.  45  This suggested that insulin and high glucose treatment activated some unknown mechanism which increased the amount of O-GlcNAc modified proteins. While the addition of STZ aided in the accumulation of modified proteins 15 minutes post-addition and the effect persisted throughout the time course as the intensity of the western blot signals were noticeably stronger. This indicated that the addition of STZ as part of the treatment was indeed useful towards maximizing the amount of available O-GlcNAc modified proteins.  3.4.2 Optimizing lysis conditions Various lysis conditions were tested so that maximal recovery of O-GlcNAc modified proteins could be obtained while avoiding the use of detergents within the lysis conditions. This was done since the use of detergents within an ion exchange is, although allowed, not preferred since it necessitates extra clean up procedures. The advantage of the V3F lysis buffer adopted (containing 400mM NaCl and no detergent) was that a nucleocytoplasmic preparation was obtained (42, 54). This was the ideal type of sample for these studies as the majority of OGlcNAc modified proteins are typically found either in the cytoplasm or nuclei (88). What became evident when comparing the amount of O-GlcNAc modified proteins obtained using the V2 lysis buffer containing 150mM NaCl and 0.1% NP-40 detergent and the V2F buffer which did not contain 0.1% NP-40, there was a noticeable difference in western blot signal intensity between the two samples. In addition, the appearance of a few extra bands with the detergent containing sample was also observed. It was likely that the difference was due to inefficient lysis with the detergent-free buffer. With the detergent, cell membranes were easily disrupted and as such, the recovery of potentially modified proteins was greater. However, with the high  46  salt lysis condition, more modified proteins were observed as a result of the hypertonic conditions and mechanical breakage which has been previously used to extract nuclear proteins (89-91). By using high salt conditions as the lysis method of choice, mixtures of both nuclear and cytoplasmic proteins were obtained. This type of sample gave the widest range of modified proteins for use as starting material in the separation studies.  3.4.3 Anion exchange chromatography Anion exchange chromatography is an extremely versatile separation method which is widely used in a variety of processes ranging from the separation of chemical compounds to protein and peptide separation/purification. The goal of this research project was to assess the possibility of a novel approach towards the study of the O-GlcNAc proteome by using anion exchange as a non-specific method to separate/enrich O-GlcNAc modified proteins. However, there were some problems which were encountered and led to the conclusion that standard anion exchange protocol was not suitable using the equipment available in the laboratory. One of the problems encountered in the course of this study was the lack of a real time readout by measuring the absorbance at 280nm as the separation was occurring. The reason for this was that the instrument used (ETTAN MDLC) is an instrument which was primarily designed for the separation of peptides. As such, the UV flow cell was rated for small volumes rather than the large volumes and high flow-rates which were required. To overcome the lack of a UV280 readout, an indirect method of detection using the BCA assay was used. This assay is typically used to determine protein concentration in a sample based on the amount of colour change that occurs when the peptide bond reduces Cu2+ to Cu1+ and the reduced copper binds 47  with bicinchoninic acid to form a characteristic purple colour, the absorbance of which can be measured at 562nm. Therefore, by plotting the measured absorbance, it was possible to obtain a readout of the relative amount of protein present in each fraction. Although the BCA assay is more sensitive (detection range of 0.2 - 50µg) than a UV measurement at 280nm (detection range of 20µg -3mg) (92), it is still discontinuous and time-consuming. Correspondingly, only the overall trends could be observed and is not an accurate portrayal of the protein profile since there could be other peaks which could be detected if more points were available. In addition, even with the added sensitivity of the BCA assay a low overall UV response was obtained which indicated that more starting material was required. From this readout, it was observed that there is a portion of the injected sample which was not binding to the column (Figure 3.5). This indicated that the starting pH was not high enough to bind all the proteins in the sample. The non-binding proteins still had a net positive charge at the starting pH and were therefore repelled by the positive charge on the resin. In addition, the presence of only 7 relatively distinct peaks indicated two major possibilities. Firstly, there could have been a large protein loss due to the non-binding proteins and the remaining mixture of proteins is therefore less complex. However, the Coomassie stained gel did not seem to provide any evidence that this reasoning was valid. Although proteins present were observed within the first few fractions of the flow-through, there was not enough protein complexity and intensity within these fractions to account for a large loss of proteins. A typical Coomassie stained gel containing a whole cell lysate would have a high density of intensely staining protein bands with molecular weights from 200 plus kDa to the smallest  48  protein resolvable on a 10% Acrylamide gel which in our case would be between the 16 to 22kDa range. The only intense and observable protein bands were found in the 36-55kDa range while the remainder of the proteins were not as intense. In addition, very low number of bands were visible. The other possibility was that the resin in the column was not suited to resolve a mixture of this complexity. In chromatography, the resolution of a column is defined by the two parameters of selectivity and efficiency. In this case, selectivity is defined to be the amount of separation that is observed between peaks and efficiency the width of the peak itself. As mentioned previously, the majority of proteins were eluting over a period of 5-7 fractions. This number of fractions represented approximately a two minute period in the separation. From the chromatogram, it can be seen that fractions collected over two minutes approximately span one peak indicating a lack of efficiency. This is because in an efficient column, peaks elute over as few fractions as possible. In addition, these peaks also overlapped each other thus indicating a lack of efficiency as well. The MiniQ column used in this study had the highest resolution of all the columns available in the GE line. This is because it is packed with the smallest bead size (3µm) available. Typically the smaller the bead size, the greater the efficiency because smaller beads pack more tightly and more evenly. This allows for even buffer flow throughout the column and the closeness of the beads minimizes the diffusion distance. By minimizing diffusion distance, the exchange between counter ions and solute during elution is faster and correspondingly the peak width will be narrower. As mentioned in 2.1.6, the sample used was highly complex, highly concentrated (20mg/mL) and in a small volume. This proved to be a problem with the MiniQ column which was used. When a sample is extremely concentrated, 49  the viscosity of the sample correspondingly increases. High viscosity leads to increased back pressure because a viscous sample will travel slower through the column at a given flow-rate (e.g. increased resistance). To counteract the increased backpressure, the flow-rate is decreased and the rate of elution is also decreased which leads to increased peak width. In addition, highly concentrated protein solutions also run the risk of precipitation. The presence of particulate matter due to protein precipitation leads to uneven buffer flow within the column. Regional differences in elution buffer flow-rates leads to different elution rates of even proteins with similar properties and thus decreased resolution. In our situation, the solution would be to use a larger column packed with larger beads so that a larger volume of a more dilute sample can be injected which would alleviate all the problems discussed. From the dot blots, a lack of separation in O-GlcNAc modified proteins was observed as there appeared to be O-GlcNAc signals present in every fraction. This result confirms the conclusion that the separation scheme used was not efficient enough to gain the type of resolution which is required to be able to complete a thorough study of the O-GlcNAc subproteome. This result was further confirmed by both the Coomassie stained gels and the Western blots of the complete fractionation. These results both showed very little difference between individual fractions. In addition, the O-GlcNAc blots all required overexposures to obtain any signal at all. This indicated that in addition to a lack of resolution, there is also a need to scale up experiments such that modified proteins are easily detected. It was the combination of these observations which led to the evaluation of an alternative separation strategy which was more effective for use in the study of O-GlcNAc modified proteins. In addition, there was no need to quantify the dot blots as its use in the separation strategy is to 50  determine which fractions contain proteins which are O-GlcNAc modified. The determination of the relative abundance of O-GlcNAc modified proteins in this case is not as important.  51  CHAPTER IV – THE ISOELECTRIC CHROMATOFOCUSING (ICF) PLATFORM Given that suitable levels of separation could not be achieved using a standard anion exchange method, it was decided that the ICF method might be a more suitable and effective platform on which to perform the chromatographic separations which are required. As discussed in 1.2.4 (pg. 17) ICF provides a high resolution method to separate proteins on the basis of their isoelectric points. This method as shown in the following sections has a narrow peak width, a higher column capacity and a great reduction in sample complexity was obtained.  4.1 Initial ICF studies Preliminary studies were carried out using a standard linear ICF gradient from pH 10 – 3.5 to determine the general level of separation that can be achieved using this method. A representative chromatogram of an insulin/STZ/glucose treated sample separated using ICF is shown in figure 4.1a. From the UV trace at 280nm it was readily apparent that the resolution was greatly improved when compared to our previous attempts (see Figure 3.5 pg 38). Here, using the ICF method, 4 intense peaks and multiple less intense but resolvable peaks were detected. The peak width was also reduced to approximately 3 to 4 fractions per peak.  52  C2C12 Mouse Myoblast Cell Extract, 10mL 16mg MCE1003 11.00  a)  240.00 pH UV 280nm MCE1003F007  10.00  UV 280nm MCE1003F009  210.00  180.00  8.00  150.00  7.00  120.00  6.00  90.00  5.00  60.00  4.00  30.00  pH  9.00  C1  B1  A1  X1 3.00 0.00  20.00  D1  40.00  E1  G1  F1  60.00  80.00  H1 100.00  UV 280nm (mAU)  Fractions  X2 0.00 120.00  Run Volume (mL)  1  b)  2  3  4  5  6  7  8  9  10  11  12  A B C D E F G H  Figure 4.1 Isoelectric Chromatofocusing with a standard gradient. a) 16mg of lysate from treated C2C12 cells was injected and subjected to separation using the standard ICF gradient of pH 10-3.5 over a 60mL gradient. 900µL fractions were collected and absorbance measured at 280nm. The chromatograms from two representative runs have been overlaid to illustrate the reproducibility of the method. b) 12µL of each fraction obtained by ICF was spotted onto a PVDF membrane and blocked with a 5% BSA solution prior to being probed with CTD110.6. Runs and blots were replicated at n=2 53  To determine which fractions contain O-GlcNAc modified proteins, aliquots of each fraction were examined by the dot blot method as previously described (2.1.8). As shown in figure 4.1b, the most O-GlcNAc reactivity (which corresponds to modified proteins) was found in fractions E7 – F3. These fractions correspond to a very narrow pH range from 5.5 – 4.9 which on the chromatogram is seen as a series of less resolved peaks. There were also other fractions which contain modified proteins but at higher pH ranges. Given that the observed patterns of spots on the dot blot exhibited varying intensities and since the amount of protein per fraction spotted onto the membrane was normalized by volume, relative amounts of modified protein present were defined. Fractions E7 – F1 were defined to contain the most reactivity, relative to that, fractions C3 – C11, D4 – E6 and F2 – F5 were considered to be of medium intensity while any remaining observable spots were classified as weak. However, once again, blots were not quantified as the dot blot was primarily used to identify O-GlcNAc modified protein containing fractions. As mentioned previously, the most intense signals in the dot blot were found in fractions E7 – F1, and when related back to the chromatogram, these fractions are seen as a series of peaks which are not very well separated. To determine if better resolution in this region of the separation could be obtained, a concave pH gradient was generated and the protein sample re-run using this formulation. It was determined that instead of increasing the resolution near the end of the run, the resolution actually decreased and proteins eluted earlier in the separation scheme as shown by the chromatogram (Figure 4.2a). When the fractionated  54  sample was analyzed by dot blot, it was observed that O-GlcNAc modified proteins also eluted earlier in the gradient (Figure 4.2b).  55  C2C12 Mouse Myoblast Cell Extract, 10mL 16mg (Formulation 100314)  a)  11.00  100.00 pH  90.00  UV 280nm  10.00  Fractions  80.00 9.00 70.00 8.00  50.00  pH  7.00  40.00  6.00  30.00  UV 280nm (mAU)  60.00  5.00 20.00 4.00  10.00 X1  3.00 0.00  A1 20.00  B1  C1  D1  40.00  E1  G1  F1  60.00  80.00  H1 100.00  X2 0.00 120.00  Run Volume (mL)  1 2 3 4  b)  5  6  7  8  9 10 11 12  A B C D E F G H  Figure 4.2 Isoelectric chromatofocusing with a concave gradient. a) 16mg of lysate from treated C2C12 cells was injected and subjected to separation using the concave ICF gradient of pH 10-3.5 over a 60mL gradient. 900µL fractions were collected and absorbance measured at 280nm. b) 12µL of each fraction obtained by ICF was spotted onto a PVDF membrane and blocked with a 5% BSA solution prior to being probed with CTD110.6. Runs and blots were replicated at n=2.  56  As the concave gradient was unsuccessful, an alternative method to increasing resolution was to double the gradient length which would increase peak separation globally. As seen in the chromatogram (Figure 4.3a), the separation between peaks in the area of interest has been improved; the corresponding region in the extended gradient extends from fraction 2C7-2F8. By doubling the gradient however, the sample was diluted by half. Therefore the signals from the dot blot (Figure 4.3b) were also correspondingly weaker and are spread over more fractions.  57  C2C12 Mouse Myoblast Cell Extract, 10mL 16mg MCE1003 Extended Gradient 104.00  11.00  a)  pH UV 280nm  10.00  91.00  78.00  8.00  65.00  7.00  52.00  6.00  39.00  5.00  26.00  4.00  13.00  pH  9.00  X1 1A1 3.00 0.00  b)  1B1  1C1  1D1  1E1  1F1 1G1  1H1 2A1  80.00  100.00  2C1  2B1  2D1  2E1  2F1  2G1  2H1  UV 280nm (mAU)  Fractions  X2 0.00  20.00  40.00  60.00  120.00  140.00  160.00  180.00  Run Volume (mL)  1 2 3 4  5  6  7  8  9 10 11 12  1 2 3 4  1A  2A  1B 1C  2B  1D  2D  1E 1F  2E  1G  2G  1H  2H  5  6  7  8  9 10 11 12  2C  2F  Plate 1  Plate 2  Figure 4.3 Isoelectric chromatofocusing with an extended gradient. a) 16mg of lysate from treated C2C12 cells was injected and subjected to separation using an extended ICF gradient of pH 10-3.5 over a 120mL gradient. 900µL fractions were collected and absorbance measured at 280nm. b) 12µL of each fraction obtained by ICF was spotted onto a PVDF membrane and blocked with a 5% BSA solution prior to being probed with CTD110.6. Runs and blots were replicated at n=2.  58  4.2 ICF fractionation of insulin/STZ/glucose treated lysates It was decided from the data presented in the previous sections that the standard gradient gave the best balance of resolution and protein concentration per fraction. To determine which proteins are O-GlcNAc modified, several fractions obtained from the standard gradient were selected for analysis. As seen in Coomassie stained gels (Figure 4.4a), proteins typically eluted over three to four fractions which correspond to the results shown by the chromatogram. In general, results from independent experiments were reproducible. However, there appeared to be an occasional shift in isoelectric point. This shift was minor as there was only a pH difference of less than 0.1 pH units. As indicated by the large number of protein bands contained in individual fractions (Coomassie stained gels), there still appeared to be a high level of complexity present even after two dimensions of separation.  59  a)  50  kDa 250 148 98 64 50  36  36  22  22  kDa 250 148 98 64  kDa 250 148 98  A8 A9 A10 A11 A12 B1 B2 B3 B4  D2 D3 D4 D5 D6 D7 D8 D9 D10  64 50 36  98 64 50 36  kDa 250 148 98 64 50 36  22  kDa 250 148  b)  22  D11 D12 E1 E2 E3 E4 E5 E6 E7  kDa 250 148 98 64 50 36 22  22  Figure 4.4 Analysis of selected fractions from a standard ICF separation. a) 60% of each concentrated fraction obtained by the standard ICF separation was resolved by SDS-PAGE (10%) and the gels stained with Coomassie blue. Bands excised which corresponded to equivalent regions in the western blots are boxed in ( ). b) The remaining 40% of sample from the concentration fractions was divided evenly with one half being probed with the CTD110.6 antibody and the other half used as a secondary antibody only control (not shown). Gels and blots were replicated at n=2.  60  With that said, a portion of each fraction which was examined by Coomassie staining was also analyzed by western blot using the O-GlcNAc specific antibody CTD110.6. Typically, an average of 3 – 4 discernable O-GlcNAc modified bands per fraction was observed (Figure 4.4b). Using the western blot as a guide, protein bands or regions on the corresponding Coomassie stained gel were excised and the proteins bands subjected to enzymatic digestion and mass spectrometric analysis. The mass spectrometric was completed using the Q-Star a hybrid triplequadrupole time-of-flight instrument under LC-MS/MS conditions. Using the MASCOT database, the peptide mass fingerprints of proteins found in each band was determined. Findings are summarized in table 4.1. The band numbering nomenclature found in the table is as follows, the first part before the dash indicates the lane and the number after the dash indicates the band number in the lane when counted from the top of the lane. In this table, proteins from regions which are found to be above and below the band of interest are also listed to account for slight mobility variations. In addition, bands corresponding to O-GlcNAc reactivity in the western blots are marked in bold. Proteins identified by previous studies to be O-GlcNAc modified are marked with a (*). Of the proteins identified, the functional breakdown is shown in figure 4.5. Briefly, it was found that the two major categories of proteins were metabolism (32%) and protein processing (21%).  61  62  Glucose 6 phosphate isomerase  Keratin type I Cytoskeletal 10 Glutathione S-transferase P 1* Keratin type I Cytoskeletal 18* Four and a half LIM domains protein 1  Four and a half LIM domains protein 1  Carbonic Anhydrase 3 GTP:AMP phosphotransferase mitochondrial Peptidyl-prolyl cis-trans isomerase A FK506-binding protein 3 (same as P17742)  Malate dehydrogenase mitochondrial precursor Annexin A2 Annexin A2* Malate dehydrogenase motochonrial precursor Annexin A2  Cytosol Aminopeptidase*  Cytosol Aminopeptidase* Glucose 6 phosphate isomerase Moesin  No Matches  Adenosylhomocysteinase Putative GTP-binding protein 9  A9-1  A9-2  A10-1  A10-2  A10-3  A11-1  A11-2  D2-1  D2-2  D2-3  D2-4  A11-3  Protein ID  Band  P50247 Q9CZ30  Q9CPY7 P06745 P26041  Q9CPY7  P07356  P07356 P08249  P08249 P07356  P16015 Q9WTP7 P17742 Q62446  P97447  P02535 P19157 P05784 P97447  P06745  Accession #  Metabolism Metabolism  Protein Processing Protein Processing Metabolism Structural  Other  Other Metabolism  Metabolism Signaling Protein processing Protein Processing Metabolism Other  Other (Diff)  Structural Metabolism Structural Other (Diff)  Metabolism  Function  48.1 44.7  56.5 62.9 67.8  56.5  38.9  38.9 36.0  36.0 38.9  29.6 25.4 18.1 25.1  33.8  57.9 23.7 47.5 33.8  62.9  MW (kDa)  47  55  55  34-35  35  36 36  29  33  33  47-48  48-49  MW Found (kDa)  Table 4.1 Proteins identified from insulin/STZ/ glucose treated C2C12 samples  6.08 7.64  7.61 8.18 6.24  7.61  7.53  7.53 8.55  8.55 7.53  6.97 8.88 7.88 9.29  8.76  5.04 8.13 5.22 8.76  8.18  Predicted pI  7.0  7.0  7.0  9.2  9.2  9.2  9.3  9.3  9.3  9.4  9.4  Experimental pI  1150 418  798 216 144  217  348  882 117  581 306  268 150 133 124  148  172 110 88 418  275  Mascot Score  134 31  54 11 10  7  12  79 4  39 14  30 11 8 7  8  4 6 3 29  24  Peptides Matched  63  Protein ID  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  Moesin Phosphoglucomutase-1 Moesin  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  No Significant Matches  Adenosylhomocysteinase  Adenosylhomocycteinase  Vinculin* Elongation Factor 2  Vinculin* Elongation Factor 2  Elongation Factor 2 Vinculin* Glycogen Phosphorylase  T-complex Protein 1 Subunit Beta Cytosol aminopeptidase* T-complex Protein 1 Subunit Eta  T-complex Protein 1 Subunit Beta RuvB-Like 1 Septin 11  Band  D5-1  D5-2  D5-3  D5-4  D5-5 D5-6  D7-1  D7-2  D7-3  D7-4  D7-5  D7-6  D7-7  D7-8  D7-9  D7-10  D7-11  D7-12  E6-1  E6-2  E6-3  E6-4  E6-5  Table 4.1 Continued  P80314 P60122 Q8C1B7  P80314 Q9CPY7 P80313  P58252 Q64727 Q9WUB3  Q64727 P58252  Q64727 P58252  P50247  P50247  P26041 Q9D0F9 P26041  Accession #  Protein Processing Protein Processing Protein Processing Protein Processing Transcription/ Translation Cell Cycle  Structural Transcription/ Translation Structural Transcription/ Translation Transcription/ Translation Structural Metabolism  Metabolism  Metabolism  Structural Metabolism Structural  Function  53  57.7 56.5 60.1  52  98  96.2 117.2 97.68  57.7 50.5 50  100  123  47  48  62 60  MW Found (kDa)  117.2 96.2  117.2 96.2  48.1  48.1  67.8 61.7 67.8  MW (kDa)  5.98 6.02 6.26  5.98 7.61 7.95  6.42 5.77 6.65  5.77 6.42  5.77 6.42  6.08  6.08  6.24 6.32 6.32  Predicted pI  5.8  5.8  5.8  5.8  5.8  6.8  6.8  6.8 6.8  Experimental pI  779 246 237  779 433 408  1196 706 174  1269 775  1905 128  294  427  356 413 411  Mascot Score  25 10 8  25 14 16  56 26 10  39 30  68 6  10  15  20 22 21  Peptides Matched  Cell Cycle 5%  Signaling 5%  Transcription/Translation 11%  Metabolism 32%  Structural 16% Protein Processing 21% Other 10%  Figure 4.5 Functional breakdown of identified proteins. Proteins which were identified by mass spectrometry were classified by their function and plotted to determine the percentage composition of each functional group.  4.3 ICF fractionation of untreated lysates From the gel bands identified, there were a few in which it was not very clear as to which protein is O-GlcNAc modified. This was due to multiple proteins being identified during the mass spectrometric analysis. To further distinguish the modified proteins from the unmodified proteins in these cases, identical ICF separations on samples which were not treated with insulin, STZ and glucose were completed. This was done to see if there were any differences between the treated and untreated samples which would allow for a positive 64  identification of modified proteins by a subtractive method. When comparing the chromatograms from the treated versus the untreated samples (Figure 4.6), it was observed that the overall patterns had three major features which remained constant, the first being that the first peak is more intense relative to the following 4-5 peaks. Secondly, three intense peaks were found in fractions D2, D7, and E3 and finally the series of less resolved peaks found at fractions E12 to G1. These data indicated that there should be very little differences between the two different samples when the overall composition is compared. Indeed, when the Coomassie stained gels of selected fractions from the untreated sample were examined, they were found to also contain similar banding patterns (Figure 4.7a). Proteins were also found to elute over 3-4 fractions which corresponds to the patterns shown in the chromatogram.  65  Overlays of Treated and Untreated treated C2C12 Mouse Myoblast Cell Extract ICF Runs 11.00  160.00 pH UV 280nm MCE1003  10.00  UV 280nm UMCE1003  140.00  120.00  8.00  100.00  7.00  80.00  6.00  60.00  5.00  40.00  4.00  20.00  pH  9.00  A1  X1 3.00 0.00  20.00  B1 40.00  C1  D1  E1  60.00  F1 80.00  G1  H1 100.00  UV 280nm (mAU)  Fractions  X2 0.00 120.00  Run Volume (mL)  Figure 4.6 Comparison between ICF chromatograms from O-GlcNAc enhanced and untreated samples. 16mg of each sample was separated using the standard ICF method and their respective chromatograms overlaid to determine if there are any differences in their UV280nm absorbance. Treated (MCE1003) and Untreated (UMCE1003). Runs were replicated at n=2.  66  a)  kDa 250 148  A8  A9 A10 A11 A12 B1 B2 B3 B4  b) kDa 250 148 98  98 64 50  64 50  36  36  22 kDa 250 148 98 64  D2 D3 D4 D5 D6 D7 D8 D9 D10  kDa 250 148 98  50  64 50  36  36  22  22  kDa 250 148 98 64 50 36  D11 D12 E1  E2 E3 E4  E5  E6  E7  kDa 250 148 98 64 50 36  22 22  Figure 4.7 Analysis of selected fractions from standard ICF of untreated C2C12 samples. a) 60% of each concentrated fraction obtained by the standard ICF separation was resolved by SDS-PAGE and the gels stained with Coomassie blue. Bands excised which corresponded to equivalent regions in the western blots are boxed in ( ). b) The remaining 40% was divided evenly with one half being immunoblotted with the CTD110.6 antibody and the other half used as a secondary antibody only control. Blots and gels were replicated at n=2.  67  Upon dot blot analysis of the untreated sample, it was evident that no detectable OGlcNAc reactivity from any of the fractions are present (Figure 4.8a). However, western blots of the pre-fractionated sample showed signs of modified proteins (Figure 4.8b). Fractions analyzed by western blot were found to contain proteins which are O-GlcNAc modified (Figure 4.7b). In addition, differences in the banding patterns of the treated and untreated samples were observed. Therefore, an extra set of gel bands from each sample type was excised from the D7 fraction and the proteins identified by trypsin digestion and subsequent mass spectrometric analysis. This fraction corresponded to the apex of one of the major peaks in both samples. When the list of proteins found by mass spectrometry (Table 4.2) in the untreated sample was compared to the ones found in the treated sample (Table 4.1), it was determined there are a few proteins which are identical to ones found in the treated sample at the same molecular weight. These proteins include cytosol aminopeptidase in fraction D2-2 and vinculin found in fraction E6-3. The overlapping proteins were found in regions which contained O-GlcNAc reactivity on the western blot (D2-2) or no reactivity (E6-3). The similarities indicated the possibility that proteins identified are either modified under both conditions or modified when treated and non-modified when untreated. More importantly, this indicated the possibility that the protein isoelectric point was not affected by the presence or absence of the O-GlcNAc modification.  68  a)  1  2  3  4  5  6  7  8  9  10  11  12  b)  Untreated C2C12  A B C  kDa 250 148 98 64  D  50  E 36 F G  22  H  Figure 4.8 Dot blot analysis of ICF fractionated untreated C2C12 samples. a) 12µL of each fraction obtained from ICF fractionation of untreated C2C12 samples were spotted onto PVDF and probed with the CTD110.6 antibody (1:5000) b) Untreated C2C12 Lysate was resolved by SDS-PAGE and probed with CTD110.6 prior to injection to determine the presence or absence of the O-GlcNAc modification. Blots were replicated at n=2.  69  70  No Significant Matches  Keratin Type I Cytoskeletal 10 Keratin Type I Cytoskeletal 15 Keratin Type I Cytoskeletal 15 Keratin Type II Cytoskeletal 6G Keratin Type I Cytoskeletal 10 Malate Dehydrogenase  A9-1  A9-2  No Significant Matches  Moesin  Cytosol Aminopepetidase  Glutathione Reductase  Adenosylhomocysteinase Putative GTP-binding protein 9  A11-3  D2-1  D2-2  D2-3  D2-4  No Significant Matches  A11-2  A11-1  A10-3  A10-2  A10-1  Protein ID  Band  P50247 Q9CZ30  P47791  Q9CPY7  P26041  P14152  Q9R0H5  Q61414  P02535 Q61414  Accession #  Metabolism Metabolism  Metabolism  Protein Processing  Structural  Metabolism  Structural  Structural  Structural Structural  Function  48.1 44.7  54.2  56.5  67.8  36  49.2  57.9 49.2  MW (kDa)  47  53  53  55  36  33  48  MW Found (kDa)  6.08 7.64  6.99  7.61  6.24  8.55  4.79  5.04 4.79  Predicted pI  Table 4.2 Proteins identified from ICF fractionation of untreated C2C12 samples  7.0  7.0  7.0  7.0  9.2  9.3  9.4  Experimental pI  726 202  89  222  80  116  159  234 112  Mascot Score  27 5  3  8  5  4  4  6 4  Peptides Matched  71  Moesin  Moesin  Moesin Ezrin Moesin  Phosphoglucomutase-1  Phosphoglucomutase-1  Leukotriene A-4 Hyrdolase  No Significant Matches  Elongation Factor 2  Ras-specific guanine nucleotide-releasing factor 1 No Significant Matches  Collagen Alpha-1(XXV) chain  6-phosphogluconate dehydrogenase Adenosylhomocysteinase  Adenosylhomocysteinase  D5-1  D5-2  D5-3  D5-5  D5-6  D7-1  D7-2  D7-3  D7-4  D7-6  D7-7  D7-9  Adenosylhomocysteinase  Adenosylhomocysteinase  Vinculin  Vinculin Elongation Factor 2  Vinculin Elongation Factor 2  Septin 11 Elongation Factor 2  Septin 11 Septin 8 Moesin GDP-fucose protein Ofucosyltransferase 2 precursor  D7-11  D7-12  E6-1  E6-2  E6-3  E6-4  E6-5  D7-10  D7-8  D7-5  D5-4  Protein ID  Band  Table 4.2 Continued Accession #  Q8C1B7 Q8CHH9 P26041 Q8VHI3  Q8C1B7 P58252  Q64727 P58252  Q64727 P58252  Q64727  P50247  P50247  P50247  P50247  Q9DCD0  Q99MQ5  P27671  P58252  P24527  Q9D0F9  Q9D0F9  P26041 P26040 P26041  P26041  P26041  Structural Transcription/ Translation Structural Transcription/ Translation Cell Cycle Transcription/ Translation Cell Cycle Cell Cycle Structural Metabolism  Structural  Metabolism  Metabolism  Metabolism  Metabolism  Metabolism  Other  Transcription/ Translation Signaling  Metabolism  Metabolism  Metabolism  Structural Structural Structural  Structural  Structural  Function  MW (kDa)  50.05 50.1 67.8 49.7  50.05 96.2  117.2 96.2  117.2 96.2  117.2  48.1  48.1  48.1  48.1  53.7  65.8  145.2  96.2  69.6  61.7  61.7  67.8 69.5 67.8  67.8  67.8  MW Found (kDa)  56  57  119  119  120  47  48  51  52  53  68  90  98  100  61  61  62  66  70  71  Predicted pI  6.26 5.68 6.24 5.95  6.26 6.42  5.77 6.42  5.77 6.42  5.77  6.08  6.08  6.08  6.08  6.88  8.7  6.74  6.42  5.98  6.32  6.32  6.24 5.83 6.24  6.24  6.24  Experimental pI  5.7  5.7  5.7  5.7  5.7  6.8  6.8  6.8  6.8  6.8  6.8  6.8  6.8  6.8  6.9  6.9  6.9  6.9  6.9  6.9  Mascot Score  555 196 140 135  249 82  1551 238  2225 78  450  132  630  155  539  205  32  29  110  169  221  390 144 78  273  78  Peptides Matched  98 51 10 15  25 6  188 16  370 8  31  11  51  7  24  9  2  3  4  14  16  34 16 5  20  3  4.4 ICF fractionation of alloxan treated lysates Given that several potential O-GlcNAc modified proteins were identified, a method of validating these proteins was required. This could be done by removing the stimulus that causes the addition of O-GlcNAc to proteins and seeing if the same proteins are still modified. Since there were modified proteins found within the untreated samples, a third sample type was introduced using the OGT inhibitor alloxan. Differentiated C2C12 cells were treated with the inhibitor alloxan (ALX). It was found that a treatment time of 4 hours with 5mM ALX was sufficient to reduce the amount of O-GlcNAc modified proteins within the lysate (Figure 4.9). Inhibition was not complete and proteins which are still modified after treatment are evident. By examining an overlay of the chromatograms from all three sample types (Untreated, insulin/STZ/glucose treated, and alloxan inhibited) (Figure 4.10) it was seen they all have very similar elution profiles.  72  a)  b)  kDa 250 148 98 64 50  kDa 250 148 98 64 50  IB: CTD110.6  36 36 22 22  IB: β-actin  Figure 4.9 Time course and dose response experiments with alloxan. a) Differentiated C2C12 cells were treated with 5mM Alloxan for 15’, 30’, 1 hour, 1.5 hours, 2 hours, 4 hours and 24 hours. 30µg of each time point along with an untreated and a STZ, Insulin, and glucose treated sample was resolved by SDS-PAGE and probed with the CTD110.6 antibody. b) Differentiated C2C12 cells were treated with 10µM, 100µM, 500µM, 1mM, 5mM and 10mM alloxan for 24 hours. 30µg of each sample along with an untreated and a STZ, insulin, and glucose treated sample was resolved by SDS-PAGE and probed with the CTD110.6 antibody. Both a) and b) were stripped and re-probed with an actin antibody to verify equal loading. Blots were replicated at n=2.  73  Overlays of Treated,Untreated and alloxan treated C2C12 Mouse Myoblast Cell Extract ICF Runs 11.00  160.00 pH UV 280nm MCE1003  10.00  UV 280nm UMCE1003  140.00  UV 280nm BMCE1003  8.00  100.00  7.00  80.00  6.00  60.00  5.00  40.00  4.00  20.00  pH  120.00  A1  X1 3.00 0.00  20.00  B1 40.00  C1  D1  E1  60.00  F1 80.00  G1  H1 100.00  UV 280nm (mAU)  Fractions  9.00  X2 0.00 120.00  Run Volume (mL)  Figure 4.10 ICF separations of Treated, Untreated and Alloxan treated samples. The chromatograms from Treated, Untreated and Alloxan treated samples were overlaid to determine if there are significant differences between the three different sample types. Treated (MCE1003), Untreated (UMCE1003), and Alloxan inhibited (BMCE1003).  Dot blots of the ALX treated samples were examined and it was found that there was no signal to be seen at all (Figure 4.11). In addition, western blots of the fractions in the ALX treated sample equivalent to the ones previously analyzed were also found to be negative (not shown). The Coomassie stained gels of these fractions when compared to the equivalent ones from the other two sample types looked visually similar (Figure 4.12). Not surprisingly, when the proteins in the gel bands were trypsin digested and identified by mass spectrometry (Table 4.3), proteins (where identifiable) such as vinculin and moesin were found in regions similar to  74  those in table 4.1 and 4.2. Once again, this indicated the possibility that the presence or absence of the modification did not affect the isoelectric point of the protein.  1  2  3  4  5  6  7  8  9  10  11  12  A B C D E F G H  Figure 4.11 Dot blot analysis of ICF fractionated Alloxan treated C2C12 cells. 12µL of each fraction from the ICF separation of Alloxan C2C12 samples was spotted onto PVDF and probed with the CTD110.6 antibody. Blot was replicated at n=2.  75  kDa 250 148 98 64  A8  A9 A10 A11 A12 B1  B2 B3  B4  50 36  22  kDa 250 148  D2 D3 D4 D5 D6 D7 D8 D9 D10  98 64 50 36  22  kDa 250 148  D11 D12  E1  E2  E3  E4  E5  E6  E7  98 64 50 36  22  Figure 4.12 Analysis of selected fractions from ICF fractionated Alloxan treated samples. 60% of each concentrated fraction obtained by the standard ICF separation was resolved by SDSPAGE and the gels stained with Coomassie blue. Bands which were excised for MS analysis are boxed ( ). Gels were replicated at n=2. 76  77  No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches No Significant Matches Vinculin  D5-1  D5-4  Elongation Factor 2  No Significant Matches No Significant Matches  E6-3  E6-4  E6-5  Vinculin  E6-2  E6-1  D7-12  D7-11  D7-10  D7-9  D7-8  D7-7  D7-6  D7-5  D7-4  D7-3  D7-2  D7-1  D5-6  D5-5  D5-3  D5-2  Protein ID  No Significant Matches No Significant Matches Moesin  Band  P58252  Q64727  Q64727  P26041  Accession #  Transcription/Transl ation  Structural  Structural  Structural  Function  96.2  117.2  117.2  67.8  MW (kDa)  98  100  123  68  MW Found (kDa)  6.42  5.77  5.77  6.24  Predicted pI  Table 4.3 Proteins identified from ICF fractionated alloxan treated C2C12 samples  5.8  5.8  5.8  6.8  Experimental pI  253  125  54  102  Mascot Score  29  10  2  6  Peptides Matched  78  Glucose 6 phosphate isomerase Glucose 6 phosphate isomerase Glucose 6 phosphate isomerase No Significant matches  A9-1  Carbonic Anhydrase 3  Glucose 6 phosphate isomerase No Significant Matches  Moesin  Keratin Type II Cytoskeletal 8  Moesin  Adenosylhomocys teinase  A11-1  A11-2  D2-1  D2-2  D2-3  D2-4  A11-3  Carbonic Anhydrase 3  A10-3  A10-2  A10-1  A9-2  Protein ID  Band  Table 4.3 Continued  P50247  P26041  P11679  P26041  P06745  P16015  P16015  P06745  P06745  P06745  Accession #  Metabolism  Structural  Structural  Structural  Metabolism  Metabolism  Metabolism  Metabolism  Metabolism  Metabolism  Function  48.1  67.8  54.5  67.8  62.9  29.63  29.63  62.9  62.9  62.9  MW (kDa)  44  52  53  55  44  45  30  48-49  48-49  48-49  MW Found (kDa)  6.7  6.24  5.7  6.24  8.18  6.9  6.97  8.18  8.18  8.18  Predicted pI  7.0  7.0  7.0  7.0  9.2  9.2  9.3  9.4  9.4  9.4  Experimental pI  383  124  179  245  257  217  131  63  128  1010  Mascot Score  47  15  22  26  40  468  17  5  16  160  Peptides Matched  4.5 Validation of method To validate the methods used in this study, vinculin was chosen as a candidate to be immunoprecipitated (IP) from lysates of STZ/Insulin/Glucose treated cells to confirm the presence of the O-GlcNAc modification. Initial IPs showed no evidence of O-GlcNAc modified proteins in the expected vinculin lane when probed with the CTD110.6 antibody (Figure 4.13a, left panel). However, there was the presence of an O-GlcNAc modified protein band in the control lane at the approximate molecular weight expected from vinculin. But when the blot was re-probed with the vinculin antibody, there was no sign of vinculin in either lane (Figure 4.13a, right panel). As the lysis conditions contained glycerol, the IP was attempted again using a glycerol free lysis buffer (Figure 4.13b, left panel). In this situation, the O-GlcNAc band previously seen in the control lane in the CTD110.6 blot was no longer present. Re-probing the blot with the vinculin antibody also showed no sign of purified vinculin (Figure 4.12b, right panel).  79  a) kDa 250 148 98 64  Antibodies Used for Immunoprecipitation: anti-vinculin Immunoblot: CTD110.6  Antibodies Used for Immunoprecipitation: anti-vinculin Vinculin Immunoblot: anti-vinculin  kDa 250 148 98  50  64 50  36  36  Heavy Chain  Light Chain 22  b)  22  Antibodies Used for Immunoprecipitation: anti-vinculin Immunoblot: CTD110.6  kDa 250 148  64  Antibodies Used for Immunoprecipitation: Vinculin anti-vinculin Immunoblot: anti-vinculin  50  50  Heavy Chain  36  36  kDa 250 148 98 64  98  Light Chain  Figure 4.13 Test immunoprecipitations with vinculin. 1mg of lysate from Insulin/STZ/Glucose treated cells were incubated with 5µg of either anti-vinculin antibody or the isotype matched 9E10 antibody overnight at 4 degrees. Completed IP samples along with a whole cell lysate control (WCL) were resolved by SDS-PAGE and probed with the CTD110.6 antibody to verify the presence of O-GlcNAc modification. Blots were subsequently stripped and re-probed with the anti-vinculin antibody as a control. a) Results from a glycerol containing sample. b) Results from a glycerol free sample. Blots were replicated at n=2.  80  4.6 Discussion The purpose of the data presented in this chapter was to determine if the ICF platform was a better separation solution in terms of resolution and the ability to identify O-GlcNAc modified proteins when compared to standard anion exchange. As seen from the chromatograms presented (Figure 4.1, pg. 53), the separation obtained by the ICF method was much greater in resolution than the one obtained by a standard anion exchange (Figure 3.5, pg. 39). It was observed in the ICF separated samples, that not only are there a larger number of peaks but the peak width was also much smaller in comparison. Rather than the 5-7 fraction peak width previously seen, the average peak width was 3-4 fractions. In addition, more protein could be separated with the larger ICF column than with the smaller anion exchange column. Using the ICF method, the MonoQ column has a separation capacity of at least 20mg of total protein, while the MiniQ column for anion exchange has a separation capacity of approximately 2mg of total protein. This increased sample capacity is a distinct advantage when searching for low abundance post-translationally modified proteins.  4.6.1 Targeted improvements in chromatographic resolution In the standard ICF method, fractions E7 – F3 gave the most intense O-GlcNAc signals as seen on the dot blot. However, the chromatogram indicated that these fractions were a series of poorly resolved peaks. As such, it was desirable to increase the separation between these peaks so that the modified proteins within these peaks could be identified. This could be achieved by optimizing the gradient used for elution.  81  Concave gradients or custom sculpted gradients are one way of optimizing the elution profile of a proteome. By controlling the slope of the gradient it should be possible to strategically improve the resolution of certain regions in the separation profile while maintaining a constant gradient length. This could be done by decreasing the slope of the gradient in the region of interest. By decreasing the slope, the time taken to elute a protein would increase. In this way, proteins could be separated into more fractions (larger peak width) and the distance between peaks is also increased. However, care must be taken so that the peak width does not increase so much that the resolution is adversely affected. As mentioned previously, it was observed that the region containing the most intense OGlcNAc signals were found in a series of less resolved peaks near the tail end of the separation. By generating a concave gradient, with a very shallow slope near the end, it should be sufficient to improve separation in the region of interest. The shallower slope (which results in a reduced rate of pH decrease) should have separated proteins in the target pH range over a larger number of fractions and correspondingly, an improved separation in our target area. Unfortunately in these experiments, proteins tended to elute much earlier than with the linear pH gradients. One explanation for this observation might lie within the elution buffer composition used to create this gradient. To generate a concave gradient, the concentrations of the individual buffer components in the elution buffer were increased to five times the concentration of the standard elution buffer while the start buffer concentrations were decreased by half. This resulted in a concentration difference between the start and elution  82  buffer of approximately 10 fold. Since the underlying column used in ICF is a strong anion exchange column, the protein-bead interaction is still sensitive to changes in ionic strength (93). By increasing the buffer concentration of the elution buffer compared to the start buffer, not only was a pH gradient established, but an ionic strength gradient which worked in parallel with the pH gradient was generated. This most likely was the cause of the observed early protein elution. An alternative approach for improving resolution was to double the gradient length. By doubling the length of the gradient, the slope of the pH gradient was decreased while the buffer composition remained constant. The decreased slope should allow for increased separation in the region of interest. Indeed, increased protein separation was observed but at the cost of protein concentration per fraction. The standard and doubled gradient both had 16mg of total concentration injected. However, when dot blots from the two different gradient lengths were compared, the intensity of the dot blot signals in the extended gradient decreased when compared to the standard one. In addition, it was determined from the chromatogram that the peak widths of the main fractions had correspondingly increased to six fractions. This was expected since the gradient length was doubled; the number of fractions was also doubled. From this, it would be expected that there would be a corresponding decrease in concentration by at least one half. Because of this dilution effect, there were two options to increase the signal intensity as observed in the dot blot. One of which was to double the amount of initial starting material or pool multiple runs, and the other option was to keep using the standard gradient and its more concentrated fractions. It was determined that in this situation, the standard gradient would provide sufficient resolution in the separation for at least the initial 83  studies. This was because the overall gain in separation in the acidic region when using the extended gradient did not provide any significant advantage over a standard length ICF separation due to the dilution effect observed.  4.6.2 Analysis of insulin/STZ/glucose treated C2C12 lysates From the SDS-PAGE separation of the ICF fractionated samples, a set of Coomassie stained gels and Western blots which correspond to regions in the basic, neutral and acidic regions of the pH gradient was obtained. From the Coomassie stained gels, there was a clearly observable trend in adjacent lanes, clusters of three to four fractions which contain bands that increase in intensity then decrease. Indeed, when compared to the chromatogram, the number of fractions corresponded to the peak width observed. These data indicated that in each of the peaks, the UV280 signal is primarily the result of a group of proteins found at a single molecular weight. When the corresponding Western blots were examined, O-GlcNAc signals often corresponded to regions of the Coomassie stained gels that contained either a faintly visible band or a region which appears blank. These observations were consistent with previous data. As reviewed by Love and Hanover, the largest percentage of proteins in the O-GlcNAc proteome are involved in transcription/translation (9). Transcription factors are low abundance modulators within the cell, thus usually require a concentration step to enrich for their detection in subsequent analyses (94, 95). This is likely why there was an observed difference between the Coomassie intensity and Western Blot signal intensity. In addition, the strength of the Western Blot signals also showed the effectiveness of the concentration effects provided by the ICF separation method.  84  The proteins identified by mass spectrometry in these studies were classified according to the categories defined by Love and Hanover (9). Functionally, the majority of proteins identified appeared to be primarily involved in metabolism (32%), followed by proteins involved in protein processing (21%) and finally structural proteins (16%). Interestingly, these findings differ from past studies (9). However, it must be noted that the proteins identified are only a small portion of the total protein which is available post-separation. As such, it was not surprising that certain categories of proteins were not identified. However, what should be noted is that the breakdown presented by Love and Hanover was compiled from multiple sources, representing a total of many different systems and it would be expected that distributions in individual systems will vary. In addition, the distribution of proteins could have been skewed due to the nature of the separation scheme. This is because the separation strategy used does not specifically extract O-GlcNAc modified proteins from the “background” proteins but relies on an indirect method (western blot) to determine which regions in the Coomassie stained gel was to be excised. As such, the possibility that the identified proteins are not modified was present. In light of the treatment conditions which the C2C12 cells were subjected to, it was not surprising that there was an abundance of metabolic and structural proteins identified. The experimental conditions used were designed to increase the levels of O-GlcNAc modification by treating the cells with insulin and glucose. Under these conditions, the cells will respond by upregulating metabolic processes to compensate for the sudden influx of glucose. Since the OGlcNAc modification has been shown to be involved in various metabolic pathways, the chances of finding modified proteins which are involved with metabolism also increases (96, 97). In 85  addition, the identification of a large proportion of structural proteins likely reflects the fact that a muscle cell line was used. These cell types typically contain large amounts of cytoskeletal proteins and their corresponding regulatory components, which are involved in forming contractile units and linking these units to other cellular organelles (98, 99). There are also reports in the literature which have shown that cytoskeletal components such as cytokeratins as well as contractile proteins like actin and myosin are O-GlcNAc modified (100, 101). These reports lend further weight to our structural and metabolic protein oriented results. However, from the table of proteins which were identified, there were a few bands (A10-3, D2-2, D2-4, etc) which contained multiple proteins, all of which appear to be positive identifications based on the number of peptides identified and the overall Mascot score obtained. This made it difficult to determine which protein, from the ones identified in a single gel band, was O-GlcNAc modified. As such, controls to determine which proteins are “background” was required.  4.6.3 Usage of untreated and alloxan inhibited C2C12 lysates as controls Two different controls were considered to be suitable for usage in the context of these studies. The first being the fractionation of a non insulin/STZ/glucose treated sample. This untreated sample should have been able to provide a fractionation profile of the basal levels of O-GlcNAc modified proteins, e.g. those which are modified even without treatment. While the second control was the use of the OGT inhibitor alloxan to prevent further O-GlcNAc modification of proteins upon addition to the cells. This inhibitory action combined with the uninhibited activity of O-GlcNAcase should have produced a fractionation profile of an 86  unmodified proteome. The proteins identified from selected fractions in the untreated samples corresponding to the ones identified in the treated samples were found to be similar. The results obtained from the selected fractions in alloxan treated samples were also similar. The goal of examining these additional sample types was to see if proteins which are common to all three sample types could be identified. These proteins are likely to be ones which are not modified. However, this was not what was observed as shown in Table 4.4 which compares a few selected proteins which were identified from the three different sample types as there was very little difference observed. It was originally speculated that there would be an observable pI shift between the different modification states since previous studies have shown that the pI of a protein is typically shifted upon being modified with a phosphate group (102). However, the pI of a protein is dependent on the composition of amino acids and correspondingly the overall charge of the protein. Therefore, the addition of the GlcNAc group onto serine and threonine residues will not affect the pI. This is because the side chains on serine and threonine residues are considered to be polar and uncharged, thus they do not contribute much to the overall pI of a protein. When the serine or threonine residues are modified with the GlcNAc group, the charge state of the protein is still unaffected since the NAcetylglucosamine group is uncharged and not ionisable. Therefore, the pI will not be affected since the charge state has not changed. Unlike a protein phosphorylation which does change the pI of the protein due to the additional ionisable charges which it adds. Even then, the study by Zhu et al showed that only in cases where the pI of the protein is greater than 6.4 were shifts in pI observed (102).  87  88  Glucose 6 phosphate isomerase  Four and a half LIM domains protein 1  Carbonic Anhydrase 3 GTP:AMP phosphotransferase mitochondrial Peptidyl-prolyl cis-trans isomerase A FK506-binding protein 3 (same as P17742) Annexin A2* Malate dehydrogenase motochonrial precursor Cytosol Aminopeptidase* Glucose 6 phosphate isomerase Moesin Adenosylhomocysteinase Putative GTP-binding protein 9 Moesin  Vinculin* Elongation Factor 2  A10-1  A10-3  E6-2  D5-5  D2-4  D2-2  A11-2  Insulin/TZ/Glucose treated  A9-1  Vinculin Elongation Factor 2  Adenosylhomocysteinase Putative GTP-binding protein 9 Phosphoglucomutase-1  Cytosol Aminopeptidase  No Significant matches  Keratin Type I Cytoskeletal 10  Keratin Type I Cytoskeletal 15  No Significant matches  Untreated  Table 4.4 Comparative list of proteins identified between the three sample types  Band  Vinculin  No significant matches  Adenosylhomocysteinase  Keratin Type II Cytoskeletal 8  Glucose 6 phosphate isomerase  Carbonic anhydrase 3  Glucose 6 phosphate isomerase  Glucose 6 phosphate isomerase  Alloxan treated  4.6.4 Problems which still need to be resolved Although proteins such as annexin and cytosol aminopeptidase which were identified in previous studies and the current study as potentially being O-GlcNAc modified, there are still two major experimental pitfalls which need to be addressed before further work can be completed. These problems are sample complexity and yield.  4.6.4.1 Sample complexity Sample complexity while reduced, still remains a concern since there were situations where protein identifications were ambiguous due to multiple proteins being present in a single band. This ambiguity existed even though theoretical pI and experimental pI information were available as an additional parameter alongside molecular weight, Mascot score and number of peptides matched to assign protein identifications. The use of pI information to assist in assigning protein identifications was not always possible. The reason being that the theoretical pI values calculated using the ProtParam software do not account for any annotated posttranslational modifications and as such do not account for any uncharacterized modifications as well (103). Thus, the values obtained can only be used as a reference value and this is likely the reason why the predicted pI values in the tables often do not match the experimental values exactly. As such, the experimental pI values in most cases do not provide any additional information towards confirming protein identifications. In addition to this, the multiple proteins identified within the bands often have excellent Mascot scores (above 100), match multiple peptides and have theoretical molecular weights that match closely to the approximate  89  molecular weight where the gel band was excised, thus making the assignment of a protein identification difficult. There are several ways that could be used to circumvent the complexity issue which still remains. One way would be to follow up the ICF separation with an additional chromatographic step which would separate proteins based on another physical property such as hydrophobicity (reversed phase chromatography). This approach of protein fractionation has been used by other groups using the Proteome Factory 2D system by Beckman Coulter (104, 105). However, the problem with this Beckman system is that the protocols used for the chromatofocusing only provide a limited number of runs before the back pressure increase prevents further separation (106). In addition, this system also uses proprietary buffer systems which are costly when compared to the simple buffer components used in the ICF platform. The most important point however, is that the additional separation step requires additional sample handling which could lead to sample loss. Thus the simplest method with the minimal amount of sample handling would be to complete a scouting run with a standard length gradient, and then redoing the fractionation with a reduced pH range gradient such that it only encompasses the region of interest. In this way, proteins with very minimal differences in pI values at the same molecular weight can be resolved and thus solving some of the protein complexity problems which were observed previously. However, increased peak widths will be observed due to the fact that proteins are now being fractionated over a narrow pH range while the gradient length remains the same. As such, the gradient length also needs to be optimized in addition to the pH ranges.  90  4.6.4.2 Sample yield Not withstanding the need to reduce sample complexity, increased recovery of proteins also appears to be a prerequisite for better identifications. As evident from the proteins identified (Table 4.1, pg. 60), there were multiple gel bands in which it was not possible to obtain protein identifications. Although current mass spectrometric instrumentation is highly sensitive with attomole detection limits (107), this level of detection is dependent on the complexity of the peptide mixture. A low abundant peptide mixed with a more abundant species is less likely to be detected. As such, what appears as a strong band in a western blot might not always be detectable by mass spectrometry since signal intensity is not a direct reflection of protein amount. In addition, many of the gel bands were excised from areas which did not appear to contain Coomassie stained protein bands. Generally speaking, it is difficult to identify proteins from gel regions which do not stain well due to the aforementioned sensitivity issue with mass spectrometers. However, by increasing the yield, so that less abundant proteins such as transcription factors are more prevalent in the sample, the probability of identifying proteins which are modified also increases. The results shown in this study show a strong trend towards identifying metabolic and structural proteins. These proteins are more abundant and were therefore easier to identify. However, previous studies have shown that the predominant types of modified proteins are transcription factors which only compose a very small portion of the proteome. Thus, simplest way of addressing the yield issue would be to pool multiple runs together so that the concentration of proteins is increased. This is possible since the method is highly reproducible  91  as indicated by the chromatograms and the identified proteins. In addition, by increasing overall protein concentration it will be possible to identify more proteins from the previously defined functional categories which are often at a lower overall concentration within the proteome. However, the complexity issue presented by using a standard gradient will not help, as such; the sample complexity still needs to be further reduced alongside an increase in protein yield.  4.6.5 Validation of results Of the proteins listed in Table 4.1, vinculin was chosen as a candidate for immunoprecipitation to verify the presence of O-GlcNAc modification. This protein was chosen since it is relatively abundant within the cell (simpler to optimize conditions) and had a commercial antibody which was suitable for immunoprecipitations. By purifying vinculin from the treated sample using a vinculin specific antibody and subsequently probing the purified vinculin with the CTD110.6 antibody, it should have been possible to verify the presence or absence of the modification. Unfortunately, the vinculin antibody chosen for this test was unable to bind enough vinculin by immunoprecipitation to be able to determine if vinculin is OGlcNAc modified. Initially, it was believed that the presence of glycerol in the lysis buffer prevented the proper binding of vinculin to the antibody. However, this was not the case as the removal of glycerol from the lysis buffer did not improve vinculin binding either. This was unexpected as the datasheet from the manufacturer indicated that this antibody was suitable for immunoprecipitation. However, further literature searches did not identify any studies which 92  used this particular antibody for immunoprecipitations. The O-GlcNAc signal seen in the glycerol containing sample was unlikely to be vinculin as the blot was blank when it was reprobed with the vinculin antibody. This signal was most likely to be the result of a non-specific protein interaction with the control antibody as a result of the glycerol in the lysis buffer. This is because when glycerol was removed from the lysis conditions, the signal which was previously seen did not appear again in the CTD110.6 blot. To verify the presence of the O-GlcNAc modification on the proteins which were identified would require antibodies for each of these proteins which are suitable for purifications of the proteins in question. More often than not, commercial antibodies which meet this criterion are unavailable. As such, to validate the method discussed in this thesis, alternative tools are required to verify the presence of the O-GlcNAc modification on the proteins identified.  93  CHAPTER V – GENERAL CONCLUSIONS AND FUTURE DIRECTIONS The studies conducted for this thesis were designed to determine the suitability of a generic chromatographic method to separate/enrich O-GlcNAc modified proteins from the proteome. From the data presented, it was determined that the usage of a MiniQ anion exchange column with a standard ion exchange protocol is unable to provide the resolution and capacity that is necessary to obtain the reduction of complexity required to identify proteins which are O-GlcNAc modified. While the Isoelectric Chromatofocusing platform (ICF) was found to have much higher resolution and capacity when compare to the standard anion exchange method. The proteins detected by ICF were found to be mostly comprised of metabolic and structural proteins. This was not surprising considering the conditions which the cells used to prepare the lysates were subjected to. Unfortunately, for the identified proteins, it was not possible to verify the presence or absence of the O-GlcNAc modification on them. Given these results, a need to increase the available resolution and resulting yield from an ICF separation is required. With that said, the usage of ICF method is advantageous in that the proteins obtained from fractionation are still native and are in a buffer system that is compatible with many downstream processes. This is because of the low salt and detergent free nature of the start and elution buffers used by the ICF method. In addition, the ICF system is easily scalable so that protein recovery can be increased. To further address the issue of validating the method, the Kast group has been developing a mass spectrometric method which will allow for the identification of O-GlcNAc 94  modified peptides. However, the sensitivity of this method is low and currently requires a large amount of peptide for it to work. In addition, this method has only been used to study model peptides so far. To truly validate this mass spectrometric method, biological samples are required. This is because physiological levels of modified peptides are much lower than the concentration of model peptides and this would allow the sensitivity issue to be addressed. It is proposed that purified vinculin from thrombin activated platelets will be a suitable candidate for these studies (108). Once this method has been established, it is envisioned that it will allow for the validation of many of the protein that were identified but not proven to be O-GlcNAc modified in these studies. To allow for a more definitive identification, it is proposed that the culture and treatment conditions be modified further. The proposed modification involves starving the cells in a glucose free medium to deplete all internal stores of UDP-GlcNAc, and then subsequently in the treatment phase replace the standard  12  C labeled glucose with  13  C labeled glucose. This  would allow for the mass spectrometric identification of O-GlcNAc modified peptides using the previously described MS method since any peptide which gives a neutral loss characteristic of 13  C labeled glucose will be O-GlcNAc modified as a result of the insulin treatment. However, the  main drawback to this method would be the expense of obtaining the 13C labeled glucose. Furthermore, it still remains to be determined whether it is possible to further reduce the complexity and increase the yield of biological samples after its proteome is fractionated by ICF. To address this, further experiments which test and optimize the length narrow pH range ICF separations to obtain maximal resolution are required. By doing so, the sample complexity  95  per fraction in the pH region of interest should be improved since the steepness of the gradient will be greatly reduced. Alternatively, the complexity might be reduced post-separation with a standard gradient by incorporating further chromatographic separations which target another physical property of the protein such as hydrophobicity. The addition of a second chromatographic step is similar to the PF2D platform sold by Beckman Coulter (104, 105). By providing a third dimension of separation in addition, to the two already provided by ICF and SDS-PAGE, the complexity per fraction should be greatly reduced. By pooling multiple runs together the amount of protein per fraction should increase and along with it the chances of detecting less abundant proteins such as modified transcription factors. If all these factors are addressed, this strategy can not only be applied to the study of OGlcNAc modified proteins but for other post-translational modifications as well. The reason why this strategy would work for other modifications easily is because the strategy only enriches/separates based on a physical property of the protein and not the modification itself. The only requirement however, is that there be a detection method specific to the modification of interest so that the fractions and proteins which are modified can be identified.  96  REFERENCES 1.  Schreiter, E. R., Rodriguez, M. M., Weichsel, A., Montfort, W. R., and Bonaventura, J. (2007) S-Nitrosylation-induced Conformational Change in Blackfin Tuna Myoglobin, J. Biol. Chem. 282, 19773-19780.  2.  Yang, Q., and Guan, K. L. (2007) Expanding mTOR signaling, Cell Res 17, 666-681.  3.  Federici, M., Menghini, R., Mauriello, A., Hribal, M. L., Ferrelli, F., Lauro, D., Sbraccia, P., Spagnoli, L. G., Sesti, G., and Lauro, R. (2002) Insulin-Dependent Activation of Endothelial Nitric Oxide Synthase Is Impaired by O-Linked Glycosylation Modification of Signaling Proteins in Human Coronary Endothelial Cells, Circulation 106, 466-472.  4.  Rush, J., Moritz, A., Lee, K. A., Guo, A., Goss, V. L., Spek, E. J., Zhang, H., Zha, X. M., Polakiewicz, R. D., and Comb, M. J. (2005) Immunoaffinity profiling of tyrosine phosphorylation in cancer cells, Nature biotechnology 23, 94-101.  5.  Li, S., Couvillon, A. D., Brasher, B. B., and Van Etten, R. A. (2001) Tyrosine phosphorylation of Grb2 by Bcr/Abl and epidermal growth factor receptor: a novel regulatory mechanism for tyrosine kinase signaling, Embo J 20, 6793-6804.  6.  Collins, M. O., Yu, L., and Choudhary, J. S. (2007) Analysis of protein phosphorylation on a proteome-scale, PROTEOMICS 7, 2751-2768.  7.  Metodiev, M. V., Timanova, A., and Stone, D. E. (2004) Differential phosphoproteome profiling by affinity capture and tandem matrix-assisted laser desorption/ionization mass spectrometry, PROTEOMICS 4, 1433-1438.  8.  Torres, C. R., and Hart, G. W. (1984) Topography and polypeptide distribution of terminal N-acetylglucosamine residues on the surfaces of intact lymphocytes. Evidence for O-linked GlcNAc, J. Biol. Chem. 259, 3308-3317.  9.  Love, D. C., and Hanover, J. A. (2005) The Hexosamine Signaling Pathway: Deciphering the "O-GlcNAc Code", Sci. STKE 2005, re13-.  10.  Marshall, S., Bacote, V., and Traxinger, R. R. (1991) Discovery of a metabolic pathway mediating glucose-induced desensitization of the glucose transport system. Role of hexosamine biosynthesis in the induction of insulin resistance, J. Biol. Chem. 266, 47064712.  11.  Buse, M. G. (2006) Hexosamines, insulin resistance, and the complications of diabetes: current status, Am J Physiol Endocrinol Metab 290, E1-E8.  97  12.  Ball, L. E., Berkaw, M. N., and Buse, M. G. (2006) Identification of the major site of Olinked beta-N-acetylglucosamine modification in the C terminus of insulin receptor substrate-1, Mol Cell Proteomics 5, 313-323.  13.  Du, X. L., Edelstein, D., Dimmeler, S., Ju, Q., Sui, C., and Brownlee, M. (2001) Hyperglycemia inhibits endothelial nitric oxide synthase activity by posttranslational modification at the Akt site, The Journal of clinical investigation 108, 1341-1348.  14.  Walgren, J. L. E., Vincent, T. S., Schey, K. L., and Buse, M. G. (2003) High glucose and insulin promote O-GlcNAc modification of proteins, including alpha -tubulin, Am J Physiol Endocrinol Metab 284, 424-434.  15.  Johnston, W. L., Krizus, A., and Dennis, J. W. (2006) The eggshell is required for meiotic fidelity, polar-body extrusion and polarization of the C. elegans embryo, BMC Biol 4, 35.  16.  Arnold, C. S., Johnson, G. V. W., Cole, R. N., Dong, D. L. Y., Lee, M., and Hart, G. W. (1996) The Microtubule-associated Protein Tau Is Extensively Modified with O-linked Nacetylglucosamine, J. Biol. Chem. 271, 28741-28744.  17.  Griffith, L. S., Mathes, M., and Schmitz, B. (1995) Beta-amyloid precursor protein is modified with O-linked N-acetylglucosamine, Journal of neuroscience research 41, 270278.  18.  Robertson, L. A., Moya, K. L., and Breen, K. C. (2004) The potential role of tau protein Oglycosylation in Alzheimer's disease, J Alzheimers Dis 6, 489-495.  19.  Alexander, G. E., Chen, K., Pietrini, P., Rapoport, S. I., and Reiman, E. M. (2002) Longitudinal PET Evaluation of Cerebral Metabolic Decline in Dementia: A Potential Outcome Measure in Alzheimer's Disease Treatment Studies, Am J Psychiatry 159, 738745.  20.  Forrest, A. R. R., Ravasi, T., Taylor, D., Huber, T., Hume, D. A., Group, R. G., Members, G. S. L., and Grimmond, S. (2003) Phosphoregulators: Protein Kinases and Protein Phosphatases of Mouse, Genome Res. 13, 1443-1454.  21.  Dong, D. L., and Hart, G. W. (1994) Purification and characterization of an O-GlcNAc selective N-acetyl- beta-D-glucosaminidase from rat spleen cytosol, J. Biol. Chem. 269, 19321-19330.  22.  Haltiwanger, R. S., Blomberg, M. A., and Hart, G. W. (1992) Glycosylation of nuclear and cytoplasmic proteins. Purification and characterization of a uridine diphospho-Nacetylglucosamine:polypeptide beta-N-acetylglucosaminyltransferase, J. Biol. Chem. 267, 9005-9013. 98  23.  Lubas, W. A., Frank, D. W., Krause, M., and Hanover, J. A. (1997) O-Linked GlcNAc Transferase Is a Conserved Nucleocytoplasmic Protein Containing Tetratricopeptide Repeats, J. Biol. Chem. 272, 9316-9324.  24.  Slawson, C., Housley, M. P., and Hart, G. W. (2006) O-GlcNAc cycling: how a single sugar post-translational modification is changing the way we think about signaling networks, Journal of cellular biochemistry 97, 71-83.  25.  Hart, G. W. (1997) Dynamic O-Linked Glycosylation of Nuclear and Cytoskeletal Proteins, Annual Review of Biochemistry 66, 315-335.  26.  Hanover, J. A. (2001) Glycan-dependent signaling: O-linked N-acetylglucosamine, FASEB J. 15, 1865-1876.  27.  Dorfmueller, H. C., Borodkin, V. S., Schimpl, M., Shepherd, S. M., Shpiro, N. A., and vanAalten, D. M. F. (2006) GlcNAcstatin: a Picomolar, Selective O-GlcNAcase Inhibitor That Modulates Intracellular O-GlcNAcylation Levels, J. Am. Chem. Soc. 128, 1648416485.  28.  Haltiwanger, R. S., Grove, K., and Philipsberg, G. A. (1998) Modulation of O-Linked NAcetylglucosamine Levels on Nuclear and Cytoplasmic Proteins in Vivo Using the Peptide O-GlcNAc-beta -N-acetylglucosaminidase Inhibitor O-(2-Acetamido-2-deoxyDglucopyranosylidene)amino-N-phenylcarbamate, J. Biol. Chem. 273, 3611-3617.  29.  Raza, H., Ahmed, I., Lakhani, M. S., Sharma, A. K., Pallot, D., and Montague, W. (1996) Effect of bitter melon (Momordica Charantia) fruit juice on the hepatic cytochrome P450-dependent monooxygenases and glutathione S-transferases in streptozotocininduced diabetic rats, Biochemical Pharmacology 52, 1639-1642.  30.  Barnard, R. J., Youngren, J. F., Kartel, D. S., and Martin, D. A. (1990) Effects of streptozotocin-induced diabetes on glucose transport in skeletal muscle, Endocrinology 126, 1921-1926.  31.  Roos, M. D., Xie, W., Su, K., Clark, J. A., Yang, X., Chin, E., Paterson, A. J., and Kudlow, J. E. (1998) Streptozotocin, an analog of N-acetylglucosamine, blocks the removal of OGlcNAc from intracellular proteins, Proc Assoc Am Physicians 10, 422-432.  32.  Liu, K., Paterson, A. J., Konrad, R. J., Parlow, A. F., Jimi, S., Roh, M., Chin, E., Jr., and Kudlow, J. E. (2002) Streptozotocin, an O-GlcNAcase inhibitor, blunts insulin and growth hormone secretion, Molecular and cellular endocrinology 194, 135-146.  33.  Konrad, R. J., Mikolaenko, I., Tolar, J. F., Liu, K., and Kudlow, J. E. (2001) The potential mechanism of the diabetogenic action of streptozotocin: inhibition of pancreatic betacell O-GlcNAc-selective N-acetyl-beta-D-glucosaminidase, Biochem. J. 356, 31-41. 99  34.  Liu, K., Paterson, A. J., Zhang, F., McAndrew, J., Fukuchi, K., Wyss, J. M., Peng, L., Hu, Y., and Kudlow, J. E. (2004) Accumulation of protein O-GlcNAc modification inhibits proteasomes in the brain and coincides with neuronal apoptosis in brain areas with high O-GlcNAc metabolism, Journal of neurochemistry 89, 1044-1055.  35.  Toleman, C., Paterson, A. J., Shin, R., and Kudlow, J. E. (2006) Streptozotocin inhibits OGlcNAcase via the production of a transition state analog, Biochemical and biophysical research communications 340, 526-534.  36.  Rees, D. A., and Alcolado, J. C. (2005) Animal models of diabetes mellitus, Diabetic Medicine 22, 359-370.  37.  Konrad, R. J., Zhang, F., Hale, J. E., Knierman, M. D., Becker, G. W., and Kudlow, J. E. (2002) Alloxan is an inhibitor of the enzyme O-linked N-acetylglucosamine transferase, Biochemical and biophysical research communications 293, 207-212.  38.  Lee, T. N., Alborn, W. E., Knierman, M. D., and Konrad, R. J. (2006) Alloxan is an inhibitor of O-GlcNAc-selective N-acetyl-[beta]-d-glucosaminidase, Biochemical and biophysical research communications 350, 1038-1043.  39.  Brew, K., Vanaman, T. C., and Hill, R. L. (1968) The role of alpha-lactalbumin and the A protein in lactose synthetase: a unique mechanism for the control of a biological reaction, Proceedings of the National Academy of Sciences of the United States of America 59, 491-497.  40.  Turner, J. R., Tartakoff, A. M., and Greenspan, N. S. (1990) Cytologic Assessment of Nuclear and Cytoplasmic O-Linked N- Acetylglucosamine Distribution by Using AntiStreptococcal Monoclonal Antibodies, Proceedings of the National Academy of Sciences 87, 5608-5612.  41.  Snow, C. M., Senior, A., and Gerace, L. (1987) Monoclonal antibodies identify a group of nuclear pore complex glycoproteins, J. Cell Biol. 104, 1143-1156.  42.  Comer, F. I., Vosseller, K., Wells, L., Accavitti, M. A., and Hart, G. W. (2001) Characterization of a Mouse Monoclonal Antibody Specific for O-Linked NAcetylglucosamine, Analytical Biochemistry 293, 169-177.  43.  Holt, G. D., Snow, C. M., Senior, A., Haltiwanger, R. S., Gerace, L., and Hart, G. W. (1987) Nuclear pore complex glycoproteins contain cytoplasmically disposed O- linked Nacetylglucosamine, J. Cell Biol. 104, 1157-1164.  44.  Lefebvre, T., Pinte, S., Guerardel, C., Deltour, S., Martin-Soudant, N., Slomianny, M. C., Michalski, J. C., and Leprince, D. (2004) The tumor suppressor HIC1 (hypermethylated in 100  cancer 1) is O-GlcNAc glycosylated, European journal of biochemistry / FEBS 271, 38433854. 45.  Konrad, R. J., Tolar, J. F., Hale, J. E., Knierman, M. D., Becker, G. W., and Kudlow, J. E. (2001) Purification of the O-Glycosylated Protein p135 and Identification as O-GlcNAc Transferase, Biochemical and biophysical research communications 288, 1136-1140.  46.  Han, I., Roos, M. D., and Kudlow, J. E. (1998) Interaction of the transcription factor Sp1 with the nuclear pore protein p62 requires the C-terminal domain of p62, Journal of cellular biochemistry 68, 50-61.  47.  Robinson, K. A., Ball, L. E., and Buse, M. G. (2007) Reduction of O-GlcNAc protein modification does not prevent insulin resistance in 3T3-L1 adipocytes, Am J Physiol Endocrinol Metab 292, E884-890.  48.  Ludemann, N., Clement, A., Hans, V. H., Leschik, J., Behl, C., and Brandt, R. (2005) OGlycosylation of the Tail Domain of Neurofilament Protein M in Human Neurons and in Spinal Cord Tissue of a Rat Model of Amyotrophic Lateral Sclerosis (ALS), J. Biol. Chem. 280, 31648-31658.  49.  Taylor, R. P., Parker, G. J., Hazel, M. W., Soesanto, Y., Fuller, W., Yazzie, M. J., and McClain, D. A. (2008) Glucose deprivation stimulates O-GlcNAc modification of proteins through upregulation of O-linked N-acetylglucosaminyltransferase, J. Biol. Chem., M707328200.  50.  Wang, Z., Pandey, A., and Hart, G. W. (2007) Dynamic interplay between O-linked Nacetylglucosaminylation and glycogen synthase kinase-3-dependent phosphorylation, Mol Cell Proteomics 6, 1365-1379.  51.  Gao, Y., Miyazaki, J.-I., and Hart, G. W. (2003) The transcription factor PDX-1 is posttranslationally modified by O-linked N-acetylglucosamine and this modification is correlated with its DNA binding activity and insulin secretion in min6 [beta]-cells, Archives of Biochemistry and Biophysics 415, 155-163.  52.  Lefebvre, T., Alonso, C., Mahboub, S., Dupire, M.-J., Zanetta, J.-P., Caillet-Boudin, M.-L., and Michalski, J.-C. (1999) Effect of okadaic acid on O-linked N-acetylglucosamine levels in a neuroblastoma cell line, Biochimica et Biophysica Acta (BBA) - General Subjects 1472, 71-81.  53.  Wells, L., Vosseller, K., Cole, R. N., Cronshaw, J. M., Matunis, M. J., and Hart, G. W. (2002) Mapping Sites of O-GlcNAc Modification Using Affinity Tags for Serine and Threonine Post-translational Modifications, Mol Cell Proteomics 1, 791-804.  101  54.  Nandi, A., Sprung, R., Barma, D. K., Zhao, Y., Kim, S. C., Falck, J. R., and Zhao, Y. (2006) Global identification of O-GlcNAc-modified proteins, Analytical chemistry 78, 452-458.  55.  Vocadlo, D. J., Hang, H. C., Kim, E. J., Hanover, J. A., and Bertozzi, C. R. (2003) A chemical approach for identifying O-GlcNAc-modified proteins in cells, Proceedings of the National Academy of Sciences 100, 9116-9121.  56.  Khidekel, N., Ficarro, S. B., Clark, P. M., Bryan, M. C., Swaney, D. L., Rexach, J. E., Sun, Y. E., Coon, J. J., Peters, E. C., and Hsieh-Wilson, L. C. (2007) Probing the dynamics of OGlcNAc glycosylation in the brain using quantitative proteomics, Nat Chem Biol advanced online publication.  57.  Ong, S. E., Blagoev, B., Kratchmarova, I., Kristensen, D. B., Steen, H., Pandey, A., and Mann, M. (2002) Stable Isotope Labeling by Amino Acids in Cell Culture, SILAC, as a Simple and Accurate Approach to Expression Proteomics, Mol Cell Proteomics 1, 376386.  58.  Nelms, S. M. (2005) Trace Metal Speciation with ICP-MS detection, in Inductively Coupled Plasma Mass Spectrometry, p 279, Blackwell Publishing.  59.  Deye, J. R., Shiveley, A. N., Oehrle, S. A., and Walters, K. A. (2008) Separation of substituted fullerenes using non-aqueous reversed-phase liquid chromatography-mass spectrometry, Journal of Chromatography A 1181, 159-161.  60.  Ren, D., Pipes, G., Xiao, G., Kleemann, G. R., Bondarenko, P. V., Treuheit, M. J., and Gadgil, H. S. (2008) Reversed-phase liquid chromatography-mass spectrometry of sitespecific chemical modifications in intact immunoglobulin molecules and their fragments, Journal of Chromatography A 1179, 198-204.  61.  Feng, B., Hu, W., Ma, B. P., Wang, Y. Z., Huang, H. Z., Wang, S. Q., and Qian, X. H. (2007) Purification, characterization, and substrate specificity of a glucoamylase with steroidal saponin-rhamnosidase activity from Curvularia lunata, Applied Microbiology and Biotechnology 76, 1329-1338.  62.  Tang, W., and Ng, S. C. (2007) Synthesis of cationic single-isomer cyclodextrins for the chiral separation of amino acids and anionic pharmaceuticals, Nat. Protocols 2, 31953200.  63.  (2004) Principles of Ion Exchange, in Ion Exchange & Chromatofocusing Principles and Methods, pp 11-25, GE Healthcare.  64.  Sluyterman, L. A. A. E., and Elgersma, O. (1978) Chromatofocusing: Isoelectric focusing on ion-exchange columns : I. General Principles, Journal of Chromatography A 150, 1730. 102  65.  Sluyterman, L. A. A. E., and Wijdenes, J. (1978) Chromatofocusing: Isoelectric focusing on ion-exchange columns : II. Experimental verification, Journal of Chromatography A 150, 31-44.  66.  Roboti, A., Livaniou, E., Evangelatos, G. P., Tsoupras, G., Tsolas, O., and Ithakissios, D. S. (1994) Large-scale chromatofocusing-based method for isolating thymosin [beta]4 and thymosin [beta]9 from bovine tissues, Journal of Chromatography B: Biomedical Sciences and Applications 662, 27-34.  67.  Fountoulakis, M., Langen, H., Gray, C., and Takacs, B. (1998) Enrichment and purification of proteins of Haemophilus influenzae by chromatofocusing, Journal of Chromatography A 806, 279-291.  68.  Lee, T. H., Nagy, L., Nagakura, T., Walport, M. J., and Kay, A. B. (1982) Identification and partial characterization of an exercise-induced neutrophil chemotactic factor in bronchial asthma, The Journal of clinical investigation 69, 889-899.  69.  Kreunin, P., Yoo, C., Urquidi, V., Lubman, D. M., and Goodison, S. (2007) Proteomic profiling identifies breast tumor metastasis-associated factors in an isogenic model, PROTEOMICS 7, 299-312.  70.  Nagalla, S. R., Canick, J. A., Jacob, T., Schneider, K. A., Reddy, A. P., Thomas, A., Dasari, S., Lu, X., Lapidus, J. A., Lambert-Messerlian, G. M., Gravett, M. G., Roberts, C. T., Luthy, D., Malone, F. D., and D'Alton, M. E. (2007) Proteomic Analysis of Maternal Serum in Down Syndrome: Identification of Novel Protein Biomarkers, J. Proteome Res. 6, 1245-1257.  71.  Yan, F., Subramanian, B., Nakeff, A., Barder, T. J., Parus, S. J., and Lubman, D. M. (2003) A Comparison of Drug-Treated and Untreated HCT-116 Human Colon Adenocarcinoma Cells Using a 2-D Liquid Separation Mapping Method Based upon Chromatofocusing PI Fractionation, Anal. Chem. 75, 2299-2308.  72.  Bates, R. C., and Frey, D. D. (1998) Quasi-linear pH gradients for chromatofocusing using simple buffer mixtures: local equilibrium theory and experimental verification, Journal of Chromatography A 814, 43-54.  73.  Lopez, M. F. (1999) Advantages and Disadvantages of Carrier Ampholyte IEF, in 2-D Proteome Analysis Protocols, pp 109-110, Humana Press, Totowa.  74.  Shan, L., and Anderson, D. J. (2001) Effect of buffer concentration on gradient chromatofocusing performance separating proteins on a high-performance DEAE column, Journal of Chromatography A 909, 191-205.  103  75.  Li, C. M., and Hutchens, T. (1992) Practical Protein Chromatography, in Methods in Molecular Biology (Kenny, A., and Fowell, S., Eds.), p 246, Humana Press, Totowa.  76.  Liu, Y., and Anderson, D. J. (1997) Gradient chromatofocusing high-performance liquid chromatography. I. Practical aspects, Journal of chromatography 762, 207-217.  77.  Choy, D., and Haynes, C. A. (In preparation) A Novel Isoelectric Chromatofocusing Technique for the Fractionation of Complex Proteomes.  78.  Watson, R. T., and Pessin, J. E. (2001) Intracellular Organization of Insulin Signaling and GLUT4 Translocation, Recent Prog Horm Res 56, 175-194.  79.  Tortorella, L. L., and Pilch, P. F. (2002) C2C12 myocytes lack an insulin-responsive vesicular compartment despite dexamethasone-induced GLUT4 expression, Am J Physiol Endocrinol Metab 283, E514-524.  80.  Nedachi, T., and Kanzaki, M. (2006) Regulation of glucose transporters by insulin and extracellular glucose in C2C12 myotubes, Am J Physiol Endocrinol Metab 291, E817-828.  81.  Majumdar, G., Harmon, A., Candelaria, R., Martinez-Hernandez, A., Raghow, R., and Solomon, S. S. (2003) O-glycosylation of Sp1 and transcriptional regulation of the calmodulin gene by insulin and glucagon, Am J Physiol Endocrinol Metab 285, E584-591.  82.  Matthews, J., Belof, J., Acevedo-Duncan, M., and Potter, R. (2007) Glucosamine-induced increase in Akt phosphorylation corresponds to increased endoplasmic reticulum stress in astroglial cells, Molecular and Cellular Biochemistry 298, 109-123.  83.  Matthews, J. A., Acevedo-Duncan, M., and Potter, R. L. (2005) Selective decrease of membrane-associated PKC-[alpha] and PKC-[epsilon] in response to elevated intracellular O-GlcNAc levels in transformed human glial cells, Biochimica et Biophysica Acta (BBA) - Molecular Cell Research 1743, 305-315.  84.  Weigert, C., Brodbeck, K., Sawadogo, M., Haring, H. U., and Schleicher, E. D. (2004) Upstream Stimulatory Factor (USF) Proteins Induce Human TGF-{beta}1 Gene Activation via the Glucose-response Element-1013/-1002 in Mesangial Cells: UP-REGULATION OF USF ACTIVITY BY THE HEXOSAMINE BIOSYNTHETIC PATHWAY, J. Biol. Chem. 279, 1590815915.  85.  Hanover, J. A., Lai, Z., Lee, G., Lubas, W. A., and Sato, S. M. (1999) Elevated O-LinkedNAcetylglucosamine Metabolism in Pancreatic [beta]-Cells, Archives of Biochemistry and Biophysics 362, 38-45.  86.  Klip, A., Li, G., and Logan, W. J. (1984) Induction of sugar uptake response to insulin by serum depletion in fusing L6 myoblasts, Am J Physiol Endocrinol Metab 247, E291-296. 104  87.  Hamada, T., Arias, E. B., and Cartee, G. D. (2006) Increased submaximal insulinstimulated glucose uptake in mouse skeletal muscle after treadmill exercise, J Appl Physiol 101, 1368-1376.  88.  Hart, G. W., Housley, M. P., and Slawson, C. (2007) Cycling of O-linked [beta]-Nacetylglucosamine on nucleocytoplasmic proteins, Nature 446, 1017-1022.  89.  Andrin, C., and Hendzel, M. J. (2004) F-actin-dependent Insolubility of Chromatinmodifying Components, J. Biol. Chem. 279, 25017-25023.  90.  De Lucia, F., Lorain, S., Scamps, C., Galisson, F., MacHold, J., and Lipinski, M. (2001) Subnuclear localization and mitotic phosphorylation of HIRA, the human homologue of Saccharomyces cerevisiae transcriptional regulators Hir1p/Hir2p, Biochem. J. 358, 447455.  91.  Hideshima, T., Chauhan, D., Hayashi, T., Podar, K., Akiyama, M., Gupta, D., Richardson, P., Munshi, N., and Anderson, K. C. (2002) The Biological Sequelae of Stromal Cell-derived Factor-1{alpha} in Multiple Myeloma, Mol Cancer Ther 1, 539-544.  92.  Stoscheck, C. (1990) Quantitation of Protein, Methods in enzymology 182, 50-69.  93.  Cutler, P. (2004) Protein Purification Protocols, 2nd ed., Humana Press, Totowa.  94.  Forde, C. E., and McCutchen-Maloney, S. L. (2002) Characterization of transcription factors by mass spectrometry and the role of SELDI-MS, Mass Spectrometry Reviews 21, 419-439.  95.  Gadgil, H., Jurado, L. A., and Jarrett, H. W. (2001) DNA Affinity Chromatography of Transcription Factors, Analytical Biochemistry 290, 147-178.  96.  Yki-Jarvinen, H., Virkamaki, A., Daniels, M. C., McClain, D., and Gottschalk, W. K. (1998) Insulin and glucosamine infusions increase O-linked N-acetyl-glucosamine in skeletal muscle proteins in vivo, Metabolism 47, 449-455.  97.  Cieniewski-Bernard, C., Bastide, B., Lefebvre, T., Lemoine, J., Mounier, Y., and Michalski, J.-C. (2004) Identification of O-linked N-Acetylglucosamine Proteins in Rat Skeletal Muscle Using Two-dimensional Gel Electrophoresis and Mass Spectrometry, Mol Cell Proteomics 3, 577-585.  98.  Berendse, M., Grounds, M. D., and Lloyd, C. M. (2003) Myoblast structure affects subsequent skeletal myotube morphology and sarcomere assembly, Experimental Cell Research 291, 435-450. 105  99.  Capetanaki, Y., Bloch, R. J., Kouloumenta, A., Mavroidis, M., and Psarras, S. (2007) Muscle intermediate filaments and their links to membranes and membranous organelles, Experimental Cell Research 313, 2063-2076.  100.  Chou, C. F., Smith, A. J., and Omary, M. B. (1992) Characterization and dynamics of Olinked glycosylation of human cytokeratin 8 and 18, J. Biol. Chem. 267, 3901-3906.  101.  Hedou, J., Cieniewski-Bernard, C., Leroy, Y., Michalski, J.-C., Mounier, Y., and Bastide, B. (2007) O-Linked N-Acetylglucosaminylation Is Involved in the Ca2+ Activation Properties of Rat Skeletal Muscle, J. Biol. Chem. 282, 10360-10369.  102.  Zhu, K., Zhao, J., Lubman, D. M., Miller, F. R., and Barder, T. J. (2005) Protein pI Shifts due to Posttranslational Modifications in the Separation and Characterization of Proteins, Anal. Chem. 77, 2745-2755.  103.  Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S., M.R., W., Appel, R. D., and Bairoch, A. (2005) Protein Identification and Analysis Tools on the ExPASy Server, in The Proteomics Protocols Handbook (Walker, J. M., Ed.), pp 571-607, Human Press Inc, Totowa.  104.  Sheng, S., Chen, D., and Van Eyk, J. E. (2006) Multidimensional Liquid Chromatography Separation of Intact Proteins by Chromatographic Focusing and Reversed Phase of the Human Serum Proteome: Optimization and Protein Database, Mol Cell Proteomics 5, 2634.  105.  Shin, Y. K., Lee, H. J., Lee, J. S., and Paik, Y. K. (2006) Proteomic analysis of mammalian basic proteins by liquid-based two-dimensional column chromatography, PROTEOMICS 6, 1143-1150.  106.  Olivier Barré, M. S. (2006) Improved protocol for chromatofocusing on the ProteomeLab PF2D, PROTEOMICS 6, 5096-5098.  107.  Wolf-Yadlin, A., Hautaniemi, S., Lauffenburger, D. A., and White, F. M. (2007) Multiple reaction monitoring for robust quantitative proteomic analysis of cellular signaling networks, Proceedings of the National Academy of Sciences 104, 5860-5865.  108.  Vostal, J. G., and Krasnewich, D. M. (1994) Dynamic O-linked N-Acetylglucosamine Glycosylation of Platelet Vinculin, Mol Biol Cell 5(S), 263a.  106  

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}]}"
                            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:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0066384/manifest

Comment

Related Items