Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

L-ribulose-5-phosphate 4-epimerase: epimerisation through carbon-carbon bond cleavage Johnson, Anne Elizabeth 1998

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

Item Metadata

Download

Media
831-ubc_1998-271692.pdf [ 8.03MB ]
Metadata
JSON: 831-1.0061643.json
JSON-LD: 831-1.0061643-ld.json
RDF/XML (Pretty): 831-1.0061643-rdf.xml
RDF/JSON: 831-1.0061643-rdf.json
Turtle: 831-1.0061643-turtle.txt
N-Triples: 831-1.0061643-rdf-ntriples.txt
Original Record: 831-1.0061643-source.json
Full Text
831-1.0061643-fulltext.txt
Citation
831-1.0061643.ris

Full Text

L-RIBULOSE-5-PHOSPHATE 4-EPIMERASE: EPIMERISATION T H R O U G H CARBON-CARBON B O N D C L E A V A G E  by  A N N E ELIZABETH JOHNSON B S c , University of Toronto, 1992  A THESIS SUBMITTED I N P A R T I A L F U L F I L L M E N T O F T H E REQUIREMENTS FOR T H E D E G R E E OF DOCTOR OF PHILOSOPHY in T H E F A C U L T Y O F G R A D U A T E STUDIES Department of Chemistry  We accept this thesis as conforming tojthe requked_slandard:  T H E U N I V E R S I T Y O F BRITISH C O L U M B I A April, 1998 © A n n e E.Johnson, 1998  In  presenting  degree  this  thesis  in  at the University of  partial  fulfilment  British Columbia,  of  the  requirements  for  of  department  this or  thesis by  for scholarly  his  publication of this thesis  or  her  Department The University of British Columbia Vancouver, Canada  DE-6 (2/88)  1^1  m  may be  representatives.  It  is  granted  for extensive  by the head  understood  that  for financial gain shall not be allowed without  permission.  Date  purposes  advanced  I agree that the Library shall make it  freely available for reference and study. I further agree that permission copying  an  of  my  copying  or  my written  ii  Abstract  L-Ribulose-5-phosphate 4-epimerase, a bacterial enzyme which catalyses the final step in L-arabinose metabolism, interconverts L-ribulose-5-phosphate and D-xylulose-5-phosphate independently of N A D . The labile stereocentre does not bear an acidic proton; therefore, +  a simple deprotonation-reprotonation mechanism cannot be followed. The epimerase was cloned from Escherichia coli into a highly efficient overexpression vector.  The recombinant enzyme was found to contain a mixture of divalent zinc,  manganese and copper ions. A preparation of a homogeneous sample of the Z n  2 +  form of  the recombinant enzyme displayed kinetic constants similar to those of the naturally abundant epimerase from E. coli. Amino acid sequence similarity between the epimerase and the Class II L-fuculose-1phosphate aldolase suggests that these two enzymes may be evolutionarily related and that the epimerisation, which is metal-dependent, may occur through carbon-carbon bond cleavage and reformation. Three conserved residues (H95, H97 and D76) which are thought to be the metal ion ligands in the epimerase have been independently altered by site directed mutagenesis to asparagine. The resulting mutant epimerases exhibit low k  Cit  H95N and H97N epimerases have a reduced affinity for Z n  2 +  values. The  and lose metal readily, while  Ill  the D 7 6 N epimerase which has an affinity for Z n  2 +  comparable to that of the wild type  epimerase loses metal upon extended dialysis. These observations serve to establish a structural link between the active sites of the epimerase and the aldolase. The H 9 7 N epimerase was found capable of catalysing the aldol addition between dihydroxyacetone and glycolaldehyde phosphate (the unbound forms of the proposed reaction intermediates) to form an equilibrium mixture of L-ribulose-5-phosphate and D xylulose-5-phosphate. In addition, the epimerase was able to release dihydroxyacetone from an equiHbrating pool of L-ribulose-5-phosphate and D-xylulose-5-phosphate.  These  observations of aldolase activity establish that the active site of the epimerase is capable of catalysing carbon-carbon bond cleavage, and support the notion that the epimerase and the aldolase are evolutionarily related. Glycolaldehyde phosphate was shown to be a competitive inhibitor of the H 9 7 N enzyme with a K of 0.37 m M . The wild type enzyme was not T  significandy inhibited at 5 m M . The H97N mutation appears to have created a "leaky" epimerase which can bind to the normal reaction intermediates and generate them from the unbound aldol cleavage products.  iv  Table of Contents  Abstract  ii  Table of Contents  iv  List of Tables  viii  List of Figures  ix  Abbreviations and Symbols Acknowledgements  xii xvii  Chapter I: Enzyme Catalysed Stereochemical Inversion and Carbon-Carbon Bond Cleavage Stereochemical Inversion Carbohydrate Epimerases L-Ribulose-5-Phosphate 4-Epimerase Role and Distribution Previous Mechanistic Studies Proposed Mechanisms Relationship with Other Enzymes Aldolases Class I Aldolases Class II Aldolases Summary Chapter II: Cloning, Overexpression and Characterization of L-Ribulose-5-Phosphate 4-Epimerase from Escherichia coli Introduction Overexpression Systems The araBAD Operon ,  1 1 3 8 8 10 11 13 15 16 17 22  24 24 24 28  L-Ribulose-5-Phosphate 4-Epimerase: A n Enigmatic Reaction Mechanism L-Ribulokinase: A Means of Producing L-Ribulose-5Phosphate Results and Discussion Preparation of L-Ribulose-5-Phosphate Attempted Cloning of L-Ribulokinase Cloning, Expression, Purification and Characterization of the LRibulose-5-Phosphate 4-Epimerase of Escherichia coli Determination of Metal Ion Content EquiHbrium Conclusion Experimental Methods Bacterial Strains and Plasmids Preparation of L-Ribulose-5-Phosphate Attempted Cloning of L-Ribulokinase Cloning of the araD gene of Escherichia coli Attempts to Subclone the araD gene oi Escherichia coli Calculation and Determination of Subunit Molecular Mass D N A sequencing Enzyme Purification Protein Determination Determination of Metal Ion Content Preparation of Metal-Free Buffers and Glassware Preparation of Zn-Substituted Enzyme and Analysis of Zinc Content Measurement of Enzyme Activity G A P D H Assay for Enzyme Activity Assay for Apoenzyme Activity Measurement of Equilibrium Chapter III: Nature of the Metal Ion Ligands of L-Ribulose-5-Phosphate 4-Epimerase Introduction Zinc Ions in Enzyme Catalysis Characterization of Zinc Binding Sites The Zinc-Binding Site of L-Fuculose-1-Phosphate Aldolase Design of Mutant L-Ribulose-5-Phosphate 4-Epimerases Results and Discussion Protein Sequence Alignments Site Directed Mutagenesis Purification of the Mutant Epimerases  29 32 33 34 35 37 45 47 48 50 50 51 52 54 56 57 57 58 60 61 62 62 63 64 65 66  67 67 68 70 75 77 83 83 84 86  vi Physical Characterization of the Mutant L-Ribulose-5-Phosphate 4Epimerases 87 Activity of the Wild Type and Mutant Enzymes 90 Zinc Ion Content of the Z n Reconstituted Epimerases 94 Activity of the Cobalt-Substituted Epimerases 95 UV-Visible Spectra of the Co(II) Substituted Enzymes 97 Co -Modi£ication of the Co -Substituted Wild Type 4-Epimerase . . . 97 Conclusion 99 Experimental Methods 101 Data Base Searches and Protein Sequence Alignments 101 Bacterial and Phage Strains 101 ssDNA Rescue of p R E l 102 Site-Directed Mutagenesis 102 H95N 103 H97N 106 D76N 107 H171N 107 D N A sequencing 108 Enzyme Purification 108 Calculation and Determination of Mutant Enzymes'Subunit Masses . 109 Determination of Tetrameric Structure of the Native Enzymes 109 Preparation of Metal-Free Buffers and Glassware 110 Preparation of Apoenzyme and Reactivation with Zn(II) or Co(II) . . . 110 Circular Dichroism and Thermal Stability Ill Determination of Zinc(II) Content 112 Assay for L-ribulose-5-phosphate 4-epimerase activity 112 Assay for Apoenzyme Activity 113 UV-Visible Spectra of the Co(II)-Substituted 4-Epimerases 113 Co(III) Modification of the Co(II)-Substituted Wild Type L-Ribulose5-Phosphate 4-Epimerase 114 2 +  3+  2+  Chapter IV: L-Ribulose-5-Phosphate 4-Epimerase Follows a Retroaldol/Aldol Mechanism Introduction Results and Discussion Intermediate Identification with the Wild Type L-Ribulose-5Phosphate 4-Epimerase Intermediate Identification Using Mutant 4-Epimerases Detection of Intermediate Release A n Aldol-Like Mechanism for L-Ribulose-5-Phosphate 4-Epimerase Conclusion  115 115 119 119 122 131 133 135  vii Experimental Methods 136 Preparation of Glycolaldehyde Phosphate 136 A) Bis(cyclohexylammonium) allyl phosphate 137 B) Sodium Glycolaldehyde Phosphate 138 Continuous Assay for Dihydroxyacetone Release 139 Enzymatic Coupling of Dihydroxyacetone and Glycolaldehyde Phosphate 140 Assay for Formation of Ketopentose Phosphates in Coupling Experiments 140 Identification of Ketopentose Phosphates in Coupling Experiments . . 142 Stopped Assays for Dihydroxyacetone Release 142 Assay for Enzyme Activity 143 Assay for Inhibition and Determination of Kj 144 Chapter V: Proposed Exploration of the Active Site of L-Ribulose-5-Phosphate 4-Epimerase Rationale Small Molecule Analogues Inactivating Agents Epimerase Structure and Site-Directed Mutagenesis Studies Related Enzymes Conclusion  145 145 146 149 151 152 153  Appendix A: D N A Sequences AraD of Escherichia coli¥L12 AraB of Escherichia coli Sequence of p R E l Some Unique Restriction Sites on p R E l Oligonucleotides Used in This Study  154 154 155 157 160 161  Appendix B: N M R Spectra  162  Appendix C: Protein Sequence Alignments Complete Amino Acid Sequence Alignment Members of the A r a D / F u c A Family  167 167 168  Appendix D : Graphical Representation of Enzyme Kinetics  169  References  172  List of Tables Table 2.1. Predicted Subunit Molecular Masses of L-Ribulose-5-Phosphate 4Epimerase 41 Table 2.2. Comparison of K and Constants Determined for L-Ribulose-5Phosphate 4-Epimerase 47 Table 2.3. Primers Used for P C R Amplification of the araB Gene 53 Table 2.4. Primers Used for P C R Amplification of the araD Gene 55 Table 2.5. Primers Used in Sequencing the araD Gene 58 Table 3.1. Molecular Masses of Mutant 4-Epimerase Subunits 86 Table 3.2. Melting Temperature of Wild Type and Mutant 4-Epimerases 89 Table 3.3. Kinetic Parameters of the Zinc-Reconstituted Wild Type and Mutant L-Ribulose-5-Phosphate 4-Epimerases 92 Table 3.4. Kinetic Parameters of the Zinc-Reconstituted Wild Type and Mutant L-Ribulose-5-Phosphate 4-Epimerases In the Presence of Additional Z n . . . . 93 Table 3.5. Zinc Content of Wild Type and Mutant 4-Epimerases 94 Table 3.6. Kinetic Parameters of the Cobalt-Substituted Wild Type and Mutant LRibulose-5-Phosphate 4-Epimerases 96 Table 3.6. Primers Used for Site-Directed Mutagenesis of the araD Gene 104 u  2 +  ix  List of Figures Figure 1.1. Stereochemical Inversion Figure 1.2. The Mechanism of Epimerisation of D-Ribulose-5-Phosphate 3Epimerase Figure 1.3. The Mechanism of Epimerisation of UDP-Glucose 4-Epimerase Figure 1.4. Proposed Mechanism of GDP-D-Mannose 3,5-Epimerisation Figure 1.5. Proposed Mechanism for Epimerisation by UDP-NAcetylglucosamine 2-Epimerase Figure 1.6. Interconversion of L-Ribulose-5-Phosphate and D-Xylulose-5Phosphate Figure 1.7. Bacterial L-Arabinose Utilization Pathway Figure 1.8. Proposed Pathways for L-Ribulose-5-Phosphate 4-Epimerase Consistent with Observations Figure 1.9. Reaction Catalysed by D-^r)///6ra-Dihydroneopterin Triphosphate 2'Epimerase Figure 1.10. Mechanism of Class I Aldolases Figure 1.11. Mechanism of Class II Aldolases Figure 1.12. Coordination of the Active Site Zn Ion by Phosphoglycolohydroxamate and the Enediolate of Dihydroxyacetone Phosphate Figure 1.13. A Proposed Mechanism for Catalysis by the Class II Aldolase LFuculose-1-Phosphate Aldolase Figure 1.14. Epimerisation of D-Tagatose-l,6-diphosphate Catalysed by DTagatose-l,6-Diphosphate Aldolase Figure 1.15. Parallel Pathways of L-Fucose and L-Rhamnose Dissimilation Figure 2.1. A genetic map of the E. coli araC and araBAD operons indicating the proteins encoded by the araBAD operon and the reactions which these proteins catalyse Figure 2.2. L-Ribulose-5-Phosphate 4-Epimerase Interconverts L-Ribulose-5Phosphate and D-Xylulose-5-Phosphate Figure 2.3. Phosphorylation of L-Ribulose by L-Ribulokinase Provides L-Ribulose5-Phosphate, the Substrate for L-Ribulose-5-Phosphate 4-Epimerase Figure 2.4. Proton-NMR Spectrum of Purified L-Ribulose-5-Phosphate (free acid) at 500 M H z (in D 0 )  2 4 4 5 7 8 9 12 14 16 17  2 +  2  19 19 20 21  29 30 32 36  X  Figure 2.5. S D S - P A G E of Crude and Purified Protein Figure 2.6. A Coupled Assay for L-Ribulose-5-Phosphate 4-Epimerase Activity. . Figure 2.7. A n Alternate Assay for L-Ribulose-5-Phosphate 4-Epimerase Activity Figure 2.8. Equilibrium between L-Ribulose-5-Phosphate and D-Xylulose-5Phosphate as Determined by ' H - N M R (500 MHz) Figure 3.1. ' H - N M R of the Co -substituted B.fragilis metallo-P-lactamase (1 mM) atpH7.4 Figure 3.2. Coordination of Z n in the Active Site of L-Fuculose-1-Phosphate Aldolase Figure 3.3. Rotation of E73 of L-Fuculose-1-Phosphate Aldolase Away from Z n on Binding of Phosphoglycolohydroxamate Figure 3.4. Amino Acid Residues Figure 3.5. The R C P C R Method of Site-Directed Mutagenesis Figure 3.6. Creation of a Megaprimer Prior to R C P C R Figure 3.7. Amino Acid Sequence Alignment Figure 3.8. C D Spectra of the Wild Type and Mutant L-Ribulose-5-Phosphate 4Epimerases Figure 3.9. Native Structure of L-Ribulose-5-Phosphate 4-Epimerases Figure 3.10. Deconvoluted Electrospray Ionization Mass Analysis of the Co(III) Wild Type 4-Epimerase Figure 3.11. A Schematic Diagram of p R E l Indicating Relative Primer Binding Sites and Locations of Restriction Sites on p R E l and Restriction Sites Introduced by Site Directed Mutagenesis Figure 4.1. Class II L-Fuculose-1-Phosphate Aldolase (A) and L-Ribulose-5Phosphate 4-Epimerase (B) Can Utilise Similar Mechanistic Strategies Figure 4.2. Reduction of Dihydroxyacetone to Glycerol by Glycerol Dehydrogenase Figure 4.3. Formation of A n Epimeric Mixture of L-Ribulose-5-Phosphate and DXylulose-5-Phosphate From Dihydroxyacetone and Glycolaldehyde Phosphate (5 mM) Figure 4.4. ' H - N M R (500 M H z , D 0 , solvent suppressed) of the H97N Catalysed Coupling Reaction Products at 72 h Figure 4.5. Glycolaldehyde Phosphate is a Competitive Inhibitor of Epimerisation by H97N L-Ribulose-5-Phosphate 4-Epimerase Figure 4.6. K Determination for H97N L-Ribulose-5-Phosphate 4-epimerase Inhibition by Glycolaldehyde Phosphate Figure 4.7. Dihydroxyacetone Release by H97N and Wild Type L-Ribulose-5Phosphate 4-Epimerases Equilibrating 10 m M L-Ribulose-5-Phosphate and D-Xylulose-5-Phosphate Figure 4.8. Synthetic Route to Glycolaldehyde Phosphate from Allyl Alcohol. . . . Figure 4.9. Determination of Sugar Content  38 . 43 44 49  2+  73  2 +  76  2 +  77 78 81 82 84 88 91 98  105 116 117  124  2  127 129  x  130  132 136 141  XI  Figure 5.1. Hydroxamate Derivatives Phosphate 4-Epimerase  Which  Could  Inhibit L-Ribulose-5147  Figure 5.2. Preparation of Pentitol Phosphates as Potential Inhibitors of LRibulose-5-Phosphate 4-Epimerase Figure 5.3. Addition of Glycolate to Glycidol Phosphate Figure 5.4. Potential Active Site Labelling Agents Figure B . l . 500 M H z Proton-NMR Spectrum of Commercial D-Xylulose-5Phosphate (Sodium Salt) in D 0 (Solvent Suppressed) Figure B.2. 500 M H z D Q F C O S Y of L-Ribulose-5-Phosphate (Free Acid) in D 0 (Solvent Suppressed) Figure B.3. 500 M H z D Q F C O S Y of Commercial D-Xylulose-5-Phosphate (Sodium Salt) in D 0 (Solvent Suppressed) Figure B.4. 200 M H z Proton-NMR of Allyl Phosphate (Cyclohexylammonium Salt)inD 0 Figure B.5. 200 M H z Proton-NMR Spectrum of Glycolaldehyde Phosphate (Ca Salt) in Acidic D 0 Figure C l . Proteins Sharing a Homologous Domain with E. coli L-Ribulose-5Phosphate 4-Epimerase Figure D . l . Direct Plot of Kinetic Data from Zinc(II)-Substituted L-Ribulose-5Phosphate 4-Epimerases Figure D.2. Direct Plots of Kinetic Data for Zinc(II)-Substituted L-Ribulose-5Phosphate 4-Epimerases Assayed in the Presence of 0.1 m M Z n Figure D.3. Direct Plots of Kinetic Data for Cobalt(II)-Substituted L-Ribulose-5Phosphate 4-Epimerases 2  148 149 150 162  2  2  2  163 164 165  2+  2  2 +  166 168 169 170 171  xu  Abbreviations and Symbols  aGDH 6 8 S 8 d A A A A ADP araA AraA araB AraB  a-glycerophosphate dehydrogenase chemical shift (in N M R ) extinction coefficient extinction coefficient of denatured protein extinction coefficient of native protein observed ellipticity Angstrom(s) (deoxy)adenosine (in nucleic acids); alanine (Ala) (in proteins); absorbance absorbance of a solution of denatured protein absorbance of a solution of native protein adenosine diphosphate gene encoding L-arabinose isomerase L-arabinose isomerase gene encoding L-ribulokinase L-ribulokinase  araBAD  the group of araA,  araC araD AraD Asn Asp ATCC ATP  gene encoding the regulatory protein for the araBAD gene encoding L-ribulose-5-phosphate 4-epimerase L-ribulose-5-phosphate 4-epimerase asparagine; N aspartic acid; D American Type Culture Collection adenosine triphosphate  B. fragilis  Bacteroides fragilis  B. subtilis  Bacillus subtilis  BSA C ca. CD CoA  bovine serum albumin (deoxy) cy to sine (in nucleic acids); cysteine (Cys) (in proteins) circa circular dichroism coenzyme A  d  n  obs  d  n  araB and araD genes.  operon  xiii d D D 0 D76 D76N Da dd DHAP DNA 2  DR5P ds dt dTDP DTT  dut DXu5P E  doublet aspartic acid; Asp deuterium oxide aspartate at position 76 of the AraD sequence AraD in which D76 has been replaced with an asparagine dalton(s) doublet of doublet dihydroxyacetone phosphate deoxyribonucleic acid D-ribose-5-phosphate double stranded doublet of triplet deoxy thymidine diphosphate chlhiothreitol gene encoding dUTPase (duf indicates an inability to make active dUTPase) D-xylulose-5-phosphate glutamic acid; Glu  E73  glutamate at position 73 of the FucA sequence  E. coli  Escherichia coli  EC  Enzyme Commission (classification number) of the International Union of Biochemistry ethylenediamine tetraacetic acid, disodium salt electrospray ionization mass spectrometry a bacteriophage  EDTA ESIMS Fl FucA Fuel FucK G GAP GAPDH ccGDH GDP Glu h H H92 H94 H95 H95N H97 H97N  L-fuculose-l-phosphate aldolase L-fucose isomerase L-fuculose kinase (deoxy)guano sine (in nucleic acids); or glycine (Gly) (in proteins) glycer aldehyde- 3 -pho sphate glyceraldehyde-3-phosphate dehydrogenase oc-glycerophosphate dehydrogenase guanosine diphosphate glutamic acid; E hour(s) hitidine; His histidine at position 92 of FucA sequence histidine at position 94 of FucA sequence histidine at position 95 of AraD sequence AraD in which H95 has been replaced with an asparagine histidine at position 97 of AraD sequence AraD in which H97 has been replaced with an asparagine  H155 H171  J  histidine at position 155 of FucA sequence histidine at position 171 or the AraD sequence AraD in which H I 71 has been replaced with an asparagine N-[2-hydroxyethyl]piperazine-IY-[2-ethane sulphonic acid] histidine; H high pressure/performance liquid chromatography Hertz isoleucine; He inductively coupled plasma mass spectrometry inner diameter coupling constant (in NMR)  K. pneumoniae  Klebsiella pneumoniae  KM  catalytic rate constant (turnover number) kilodalton(s) equilibrium constant dissociation constant for an enzyme-inhibitor complex Michaelis constant gene encoding the lac repressor  H171N HEPES His HPLC Hz I ICPMS ID  kDa K,  M  K  lad lacZ  gene encoding P-galactosidase  L. pentoaceticus  Lactobacillus  pentoaceticus  L,. pentosus  Lactobacillus  pentosus  L.  Lactobacillus  plantarum  LB LRu5P LSIMS m M13K07 mdeg mRNA N NAD NADH NMR NMWL nt dNTP PAGE pAJl pAJ2 +  PAJ3 pBS pBSTIM  plantarum  Luria Bertani medium L-ribulose-5-phosphate liquid soft ionization mass spectrometry multiplet a "helper" phage millidegree(s) messenger R N A asparagine; Asn nicotinamide adenine dinucleotide nicotinamide adenine dinucleotide, reduced form nuclear magnetic resonance nominal molecular weight limit nucleotide (s) deoxynucleotide triphosphate polyacrylamide gel electrophoresis plasmid which overexpresses the H 9 5 N epimerase plasmid which overexpresses the H 9 7 N epimerase plasmid which overexpresses the D 7 6 N epimerase a protein overexpression vector a modified pBS plasmid which overexpresses triosephosph  XV  isomerase phenol: chloroform: isoamyl alcohol polymerase chain reaction Protein Data Bank polyethylene glycol a protein overexpression vector a plasmid containing the polB and araD genes parts per billion parts per million  P/C/I PCR PDB PEG pET-lla pHC5 ppb ppm pREl psi RBS RCPCR RhaA RhaB RhaD RNA rpm s S7P  a overexpression plasmid for L-ribulose-5-phosphate 4-epimerase pounds per square inch ribosomal binding site recombinant circle polymerase chain reaction L-rhamnose isomerase L-rhamnulose kinase L-rhamnulose-1 -phosphate aldolase ribonucleic acid rotations per minute singlet sedoheptulose-7-phosphate  S. typhimurium  Salmonella  SD SDS ss t T T4 T7 TIM Tn  Shine-Dalgarno sodium dodecyl sulphate single stranded triplet  typhimurium  (deoxy)thymidine (in nucleic acids); threonine (Thr) (in proteins) a bacteriophage a bacteriophage triosephosphate isomerase transposon  vol0 U  gene encoding triosephosphate isomerase thiamine pyrophosphate; cocarboxylase tris (hydroxy) amino methane tryptophan triplet of triplets tyrosine valine; Val initial reaction rate (by) volume unit (of enzyme activity), which is the amount of enzyme required  UDP  to catalyse formation of 1 umol of product per minute uridine diphosphate  tpi  TPP Tris Trp. tt Tyr V  v  xvi ung UV Vis  gene encoding uracil-N-glycosylase (ung indicates the inability to make the active enzyme) ultraviolet visible  XVII  Acknowledgements  I would like to thank Mum, Dad, and Jean for their support and encouragement throughout the duration of my doctoral studies. I cherish the wonderful friendship Claire Johnson and I have had during the last SVz years.  Finally, I thank my husband Philip  Johnson for his love and encouragement, and for running N M R experiments for me.  Chapter I  Enzyme Catalysed Stereochemical Inversion and Carbon-Carbon Bond Cleavage  Stereochemical Inversion  Bacteria benefit from their ability to utilize a wide variety of sugars and amino acids of unusual stereochemistry that are not metabolized by animals. These compounds are often enantiomers or diastereomers of common biomolecules and can serve as unique bio synthetic building blocks or energy sources. Thus, stereochemical control is an important feature in many biosynthetic and metabolic pathways, and often involves the inversion of configuration about an asymmetric carbon centre catalysed by enzymes known as racemases and epimerases (Adams, 1976; Glaser, 1972). Racemases are enzymes that act on substrates with only one asymmetric centre (interconverting enantiomers), while epimerases catalyse  1  Chapter I: Stereochemical Inversion  stereochemical  inversion on  •  substrates  with more  than one  asymmetric  2  centre  (interconverting diastereomers). These enzymes can be divided into three groups based on their substrates (see Figure 1.1).  The first group, where X = O H and Y^carbonyl or carboxylate, includes the  H  X  X H  F i g u r e 1.1. Stereochemical Inversion.  carbohydrate epimerases and oc-hydroxyacid racemases. A second group in which X = N H  2  or N H R ' and Y = C O O H , comprises the amino acid racemases and epimerases, and a third group, where X = C O O H , aryl or alkyl and Y=acyl-CoA includes the acyl-CoA racemases. N o enzymatic inversion of stereochemistry is known around stereocentres that are not hydrogen-substituted (Adams, 1976). For comprehensive reviews on epimerases and their mechanisms, see the works of Glaser (1972), Adams (1976), and Tanner & Kenyon (1998). Inversion could, in principle, occur by breaking and reforming any of the bonds at the carbon centre; however, almost all racemases and epimerases ultimately operate by a proton-transfer pathway (Adams, 1976; Barber, 1979; Faraci & Walsh, 1988; Frey, 1987; Glaser, 1972; McDonough & Wood, 1961; Melo & Glaser, 1968).  The enzyme first  deprotonates the substrate and then reprotonates the resulting intermediate on the opposite face, bringing about inversion. Often, the centre is "activated" toward pro ton-transfer by  Chapter I: Stereochemical Inversion  3  an adjacent carbonyl or imminium functionality, which increases the acidity of the labile proton (Faraci & Walsh, 1988; McDonough & Wood, 1961). If the centre is not already activated, an activating group can be introduced by the enzyme with the use of a N A D  +  or  pyridoxal phosphate cofactor.  Carbohydrate Epimerases  Carbohydrates are densely functionalised, and therefore have an abundance of stereochemical configurations. In order to make more use of these molecules, a number of carbohydrate epimerases have evolved that interconvert sugars by stereochemical inversion. These epimerases can be categorized according to whether or not they use a N A D The  vast majority of cofactor-independent  +  cofactor.  carbohydrate epimerases  invert  stereochemistry at a position adjacent to a carbonyl or carboxylate functionality, using a deprotonation-reprotonation mechanism. A n example of epimerisation at a centre activated by an adjacent carbonyl group is the interconversion of D-ribulose-5-phosphate and D-xylulose-5-phosphate by D-ribulose-5phosphate 3-epimerase. This epimerisation (Figure 1.2) occurs by deprotonation at C-3 to form the enediol(ate) intermediate, followed by reprotonation on the opposite face to form the epimeric product (Adams, 1976; Glaser, 1972).  Chapter I: Stereochemical Inversion  4  CH OH ENZ HB  CH OH  ?  r B: I  ENZ  H  N  CH OH  Z  ENZ  2  OH'  —OH  \—OH  -OH  I  ENZ  3  HO-  OH  BH H-  CH OP0 = 2  T  2  B  CH OP0 = 2  3  I  ENZ  H-  -H -OH CH OP0 = 2  3  D-xylulose 5-phosphate  D-ribulose 5-phosphate  F i g u r e 1.2. The Mechanism of Epimerisation of D-Ribulose-5-Phosphate 3-Epimerase.  Alternatively, epimerisation may occur at an "unactivated" centre by oxidation of that centre with a cofactor such as N A D , followed by re-reduction of the intermediate species +  from the opposite face (Figure 1.3). Oxidation and re-reduction occur by hydride-transfer,  UDP-galactose F i g u r e 1.3. The Mechanism of Epimerisation of UDP-Glucose 4-Epimerase  Chapter I: Stereochemical Inversion  such as in UDP-glucose 4-epimerase. enzyme-bound N A D  +  5  In this enzyme, the C-4 hydride is transferred to  to produce the intermediate UDP-4-ketopyranose. Rotation about  the bond connecting the glycosyl anomeric oxygen and the U D P moiety allows the intermediate to become reoriented in the active site so that the hydride can add back to C-4 on the opposite face of the carbohydrate ring (Frey, 1987). NAD  +  can also be used to activate an adjacent centre: GDP-D-mannose epimerase  of Chlorellapyrenoidosa  appears to operate in this manner (Barber, 1979). In this mechanism  (Figure 1.4), the epimerase transiently oxidizes the C-4 hydroxyl group to a keto functionality.  It is thought that the epimerisation then occurs through two sequential  deprotonation/reprotonation events, with two enediols as intermediates. The catalytic cycle  GDP-L-galactose  Figure 1.4. Proposed Mechanism of GDP-D-Mannose 3,5-Epimerisation.  Chapter I: Stereochemical Inversion  6  is completed by stereospecific reduction of the C-4 keto group. A n epimerase in dTDP-Lrhamnose synthetase of Pseudomonas aeruginosa also appears to share this mechanism, although the reductase activity is found in an associated protein (Glaser, 1972; Glaser etal., 1972; Melo & Glaser, 1968; Tanner & Kenyon, 1998).  This enzyme epimerises the C-3 and C-5  positions of dTDP-4-keto-6-deoxy-D-glucose. Epimerases and racemases which appear to invert unactivated stereocentres without the use of any cofactors and which, therefore, must employ unique reaction mechanisms are of  great interest.  interconversion  of  Bacterial UDP-AT-acetylglucosamine 2-epimerase UDP-AT-acetylglucosamine and  catalyses  the  UDP-AT-acetylmannosamine by  epimerisation at C-2, an unactivated stereocentre. Solvent isotope was incorporated into C-2 of the equilibrating epimers (Sala etal., 1996; Salo, 1976). Furthermore, a primary kinetic isotope effect is observed upon epimerisation of [2- FfJ-UDP-iV-acetylglucosamine (Morgan 2  etal., 1997). These results suggest that epimerisation occurs through removal of the proton at C-2 followed by its replacement in an opposite stereochemical sense. This epimerisation must occur without NAD -induced transient oxidation at C-3 since addition of N A D +  the epimerase did not have an effect on its activity and a tightly-bound N A D  +  +  to  molecule was  not found in association with the enzyme (Morgan et al., 1997). Studies on the epimerase have provided evidence that the anomeric carbon-oxygen bond is broken (and reformed) during the course of the reaction (Sala et al., 1996). These observations point toward a mechanism involving /ratf^elimination of U D P to generate  an enzyme-bound 2-  acetamidoglucal intermediate, followed by the ^-addition of U D P to the C-1 - C-2 double  Chapter I: Stereochemical Inversion  7  bond with reprotonation of the opposite face to form UDP-iV-acetylmannosamine (Figure 1.5). Further evidence for this mechanism is that on extended incubation in the presence of the enzyme, an equilibrating pool of the epimers was converted into free U D P and 2acetamidoglucal (Morgan et al, 1997).  ENZ  UDP-N-Acetylglucosamine  F i g u r e 1.5. Epimerase.  ENZ  2-Acetamidoglucal  UDP  UDP-N-Acetylmannosamine  Proposed Mechanism for Epimerisation by UDP-AT-Acetylglucosamine 2-  Another enzyme which operates at an "unactivated" stereocentre without the aid of a cofactor is L-ribulose-5-phosphate 4-epimerase, which plays a key role in bacterial Larabinose utilization, interconverting L-ribulose-5-phosphate and D-xylulose-5-phosphate by epimerisation at C-4 (Figure 1.6) (Burma & Horecker, 1958b; Lee et al, 1968; Wolin et al, 1957; Wolin et al, 1958). Determining the mechanism of this enzyme will be the focus of this thesis.  Chapter I: Stereochemical Inversion  8  CH OH  CH OH  2  2  =0 HO-  -H  HO-  -H  L-ribulose-5-phosphate 4-epimerase  HOH-  CH OP0 = 2  =o  3  L-ribulose5-phosphate (LRu5P)  -H -OH CH OP0 = 2  3  D-xylulose5-phosphate (DXu5P)  Figure 1.6. Interconversion of L-Ribulose-5-Phosphate and D-Xylulose-5-Phosphate.  L-Ribulose-5-Phosphate 4-Epimerase  Role and Distribution L-Arabinose is an unusual sugar, having an L- rather than a D- stereoconfiguration, and is found in plant pectins, gums and complex polysaccharides. Various bacteria are capable of utilizing this sugar as an energy source through a pathway elucidated in 1958 (Burma & Horecker, 1958b; Simpson etal., 1958). In this pathway (Figure 1.7), L-arabinose is first isomerized to L-ribulose by a specific isomerase. L-Ribulose is then phosphorylated with A T P by a typical Mg -requiring kinase. A key step in L-arabinose metabolism is the 2+  interconversion of L-ribulose-5-phosphate (LRu5P) and D-xylulose-5-phosphate (DXu5P) by epimerisation at C-4. DXu5P can then be shuttled into the pentose-phosphate pathway.  Chapter I: Stereochemical Inversion  CHO  CH OH  CH OH  CH OH  =o  =o  =o  2  -OH  H-  9  isomerase  HO-  -H  HO-  -H ~~  HO-  -H  HO-  -H  CH OH L-arabinose  ATP A D P kinase  HO-  -H  HO-  -H ~"  CH OH  2  2  2  CH OP0 =  2  L-ribulose  epimerase  2  3  L-ribulose5-phosphate (LRu5P)  -H  HOH-  -OH CH OP0 = 2  3  D-xylulose5-phosphate (DXu5P)  F i g u r e 1.7. Bacterial L-Arabinose Utilization Pathway  L-Ribulose-5-phosphate  4-epimerase has been isolated from Klebsiella  pneumoniae  (WoUn et al., 1958) (formerly Aerobacter aerogenes (Orskov, 1984)), Escherichia coli (Lee et al., 1968), and Lactobacillusplantarum  (Burma & Horecker, 1958b), while its activity has been  detected in a number of other bacteria including L. pentosus (Burma & Horecker, 1957), and L. pentoaceticus (Rappoport et al., 1951).  Furthermore, the D N A sequence of the gene  encoding this epimerase, araD, has been determined for E. coli (Bonner et al., 1990; Chen et al., 1990; Iwasaki etal., 1991; Lee etal, 1986; Mineno etal., 1990; Yura etal, 1992), Salmonella typhimurium  (Lin et al, 1985), Haemophilus  influenzae (Fleischmann et al., 1995), Mycoplasma  pneumoniae (Himmefreich et al, 1996) and most recently, Bacillus subtilis (Sa-Nogueira et al., 1997). Undoubtedly, with more bacterial genome sequencing, further examples will be discovered.  Chapter I: Stereochemical Inversion  10  Previous Mechanistic Studies A n examination of the substrate's structure clearly indicates that this enzyme does not employ a simple proton-transfer mechanism since the proton at the unactivated C-4 stereocentre is quite non-acidic. The mechanistic studies to date show that there is no appreciable kinetic isotope effect upon epimerisation of D-[4- H]-Xu5P or of L-[4- H]-Ru5P 3  2  (this effect is expected in both proton and hydride transfer pathways), and that epimerisation does not cause exchange of either hydrogen or oxygen isotopes with solvent (Davis et al., 1972; McDonough & Wood, 1961; Salo et al., 1972). All known epimerases and racemases that operate by proton transfer pathways show either complete or some degree of proton exchange with solvent. Exchange of oxygen with solvent could occur if epimerisation was achieved through direct displacement of the C-4 hydroxyl group by solvent. Furthermore, careful studies have indicated that this enzyme does not use the cofactor N A D  +  to oxidize  the substrate transiently (several known sugar epimerases employ this strategy). particularly elegant experiment showed the lack of N A D  +  A  by isolating the epimerase from  a culture of a K. pneumoniae nicotinic acid auxotroph which was supplemented with Relabelled nicotinic acid. N o radioactivity was associated with the purified enzyme, which was fully active (Deupree & Wood, 1970). The epimerase was, however, found to require divalent metal ions for activity (Deupree & Wood, 1972). The epimerase was completely inhibited by treatment with E D T A . Activity of the resultant apoenzyme could be restored to various extents by addition  Chapter I: Stereochemical Inversion  of divalent metal ions to the assay cuvettes, in the order M n > Mg  2 +  11 2 +  > Co  2 +  > Ni  2 +  > Ca  2 +  > Zn  2 +  (greatest to least degree of reactivation).  Proposed Mechanisms  This enzyme is unique among epimerases in that it presumably operates without breaking the carbon-hydrogen bond at the stereolabile position of the sugar. Two alternative pathways (see Figure 1.8) involving carbon-oxygen or carbon-carbon bond breakage and reformation are consistent with the present mechanistic information. It was postulated (Deupree & Wood, 1972) that the reaction catalysed by L-ribulose-5-phosphate 4-epimerase may proceed via a dehydration/rehydration mechanism with a sequestered water molecule (Path A) or via a retroaldol/aldol mechanism involving carbon-carbon bond breakage (Path B). Path A is an elimination/readdition mechanism which proceeds with a sequestered water molecule. The first step in this reaction would be a base-catalysed removal of the C-3 proton, which is activated by the adjacent carbonyl group at C-2. This would result in a carbanion at C-3 that could eliminate the C-4 hydroxyl group to form a double bond between C-3 and C-4. It is not clear how the enzyme would achieve the syn elimination which must occur in one direction, and the anti elimination in the other direction, without a major conformational change. The presence of a divalent metal ion in the active site could facilitate removal of hydroxide and prevent its exchange with bulk solvent.  The enone  Chapter I: Stereochemical Inversion  12  N Z  m—UJ II CO  N  z tu- rn  o X  I  o I  o& X  o  OJ  o-  °  Q_  OJ  L O  X  Z!  -o  o Q_  CL  X  •o  X  X  C  Q  JC  <  <_>  <D  I  1  "o 32  -LU  rt  C  <U i_ CQ  o  1_ +-> (1)  o—fc  o X  "D C O OQ  c (D OuO > v  X  o  I  c  c  X  X  o  X  L O  -o  ZS  CL  c  o-  o o I  N Z  »—LU  O  N X Z m—UJ  CD cz o XI  rt  <  U•  II CO  o  +-J  rt Q_  CD "O  rt U  1  x  X  L_  • II  o rt (D i_  o  Q_  o o  X  rt  c rt  Q  o  LO  o CM  o  9  II ^  c  CL  O  O  OJ  Q_  X  LO  Z3 cL  -o  _Q S_  rt U OQ  N OQ—LU  rt Q_  F i g u r e 1.8. Proposed Pathways for L-Ribulose-5-Phosphate 4-Epimerase Consistent with Observations.  Chapter I: Stereochemical Inversion  13  intermediate species would be transformed back to a ketopentose by addition of the hydroxyl group to C-4 on the opposite face of the double bond from which it was removed. Path B is an unprecedented means of achieving stereochemical inversion. In this retroaldol/aldol mechanism, deprotonation of the C-4 hydroxyl group leads to breakage of the C-3 - C-4 bond and to formation of the intermediates glycolaldehyde-2-phosphate and an enediolate anion. The carbonyl group of the substrate may be chelated to the metal ion and facilitate enolate formation. The carbonyl group of the aldehyde intermediate must rotate in the active site before the same face of the enediolate adds back to the opposite face of the aldehydic carbonyl group in an aldol addition. Thereby the bond between C-3 and C-4 is reformed with the geometry differing only at C-4 in the product. A tiiird possibility, not shown in Figure 1.8, involves a series of proton transfer steps that result in the isomerization of the substrate to a C-3 ketose. This carbonyl group at C-3 would acidify the proton at C-4 and allow a deprotonation/reprotonation event to occur, inverting the stereochemistry at this position. Finally, the product would be isomerized back to a C-2 ketose. This seems unlikely because all of the steps of this mechanism would have to occur without incorporation of solvent hydrogen atoms.  Relationship with Other E n z y m e s  A n enzyme which may have a similar mechanism to L-ribulose-5-phosphate 4epimerase is dihydroneopterin triphosphate 2'-epimerase, which interconverts  D-erythro-  dihydroneopterin triphosphate to L-^rra-dihydromonapterin triphosphate (Figure 1.9). This  Chapter I: Stereochemical Inversion  14  epimerase also acts at an "unactivated" centre and has a requirement for divalent metal ions, Mn  2 +  and M g  2 +  being the most effective (Heine & Brown, 1975). This epimerase shares  amino acid similarity to a putative dihydroneopterin aldolase domain but lacks aldolase activity (Haussmann et al., 1997).  O  OH OH  'N'  X  H N'  H  o H'N k  H NT 2  X  - o  H  /  P  "O  O "O  \ //  ^ o  /  P  O  \ //  ^ o  /  ^ o -  P  D-e/yffrro-dihydroneopterin triphosphate  H  2  -Q o \ //  OH H  V"1H H  "O  O "O  O "O  O  O  \ //  OH  O  \ //  O  \ //  O"  L-toreo-dihydromonapterin triphosphate  F i g u r e 1.9. Reaction Catalysed by D-^j/Aro-Dihydroneopterin Triphosphate 2'-Epimerase.  Recently, there was an observation of significant sequence homology between L-ribulose-5phosphate 4-epimerase and bacterial L-fuculose-1-phosphate aldolase (a member of a group of enzymes which catalyse metal-dependent carbon-carbon bond breaking reactions) (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996b).  It seems likely that the epimerase has  evolutionarily diverged from the aldolases. For this reason we favour mechanistic Path B,  Chapter I: Stereochemical Inversion  15  in which the enzyme generates a metal-bound enolate and instead of protonating and releasing it, adds it back to either face of the bound glycolaldehyde phosphate.  Aldolases  Aldolases are enzymes which catalyse aldol cleavages (or aldol condensations in the reverse direction) (Horecker etal, 1972; Morse & Horecker, 1968) yielding an aldehyde and a carbonyl-containing molecule for which the enzyme has very high specificity (Feingold & Hoffee, 1972; Gijsen et al, 1996; Wood, 1972).  Aldol cleavage reactions involve the  formation of a resonance-stabilized enolate intermediate. Aldolases catalyse this reaction by enhancing the stability of the enolate structure either through Schiff base formation or by use of a divalent metal ion which coordinates to the carbonyl oxygen atom. These two classes of aldolase may represent analogous proteins which have presumably arisen by convergent evolution (Rutter, 1964; Rutter, 1965). Aldolases are currendy of interest to synthetic chemists as catalysts, especially because of their ability to form enantiomerically pure vicinal diols (Gijsen et al., 1996).  Chapter I: Stereochemical Inversion  16  Class I Aldolases  Class I aldolases are found in plants and animals and use a Schiff base mechanism for catalysis. This group of enzymes, in particular fructose-bisphosphate aldolase (often referred to as "aldolase") are well-studied and have been extensively reviewed (Horecker et al, 1972; Morse & Horecker, 1968). This class of aldolase does not require divalent metal ions for catalysis and is not inhibited by treatment with E D T A or other metal chelator. They are,  CH OP0 = H 2  CH OP0 =  CH OP0 = 2  :NH  k=0 H0-  HH-  -H  2  :B  I  -OH  ENZ  CH 0P0 = 3  3  ©H =N—Lys-ENZ  H,0  V_7  Lys-ENZ  -OH  2  2  3  HH-  H —OH  ENZ  =0 H,0  H  3  :NH,  CH 0H  |  2  Lys-ENZ  B  I ,  ENZ  0  CHO  -OH  CH OP0 = 2  2  . N—Lys-ENZ HO-\-H  -H  CH 0P0 =  3  3  CH OP0 = 2  ENZ  -OH 3  CH 0P0 = 2  3  Figure 1.10. Mechanism of Class I Aldolases.  however, inhibited by reduction with N a B H in the presence of substrate. In these enzymes, 4  the e-amino group of an active site lysine residue reacts with the carbonyl group of the substrate to form an imminium cation or protonated Schiff base. This is illustrated in Figure 1.10 using fructose-l,6-diphosphate aldolase as an example. In this aldolase, base-catalysed deprotonation of the C-4 hydroxyl group leads to formation of a carbonyl group at C-4 and cleavage of the carbon-carbon bond between C-3 and C-4.  The aldehydic moiety,  Chapter I: Stereochemical Inversion  17  glyceraldehyde-3 phosphate (GAP) is then released from the enzyme's active site. The reaction is completed by protonation of the dihydroxyacetone phosphate (DHAP) anion by an active site proton donor, followed by hydrolysis of the Schiff base and release of D H A P .  Class II Aldolases Class II aldolases, found in microorganisms, use a divalent metal ion to polarize the carbonyl oxygen of the substrate and stabilize the enolate intermediate in the reaction. These enzymes are not inactivated by reduction with N a B H  4  in the presence or absence of  substrate. They are, however, inhibited by metal chelating agents such as E D T A , indicating their requirement for a divalent metal ion for catalysis. This metal ion is most often a tightlybound Z n  ion.  2 +  In the catalytic reaction, the divalent metal ion coordinates to and polarizes the carbonyl oxygen of D H A P or the substrate, and may also orient the C-1-phosphate group. In this mechanism (Figure 1.11), base-catalysed deprotonation of the C-4 hydroxyl group leads to formation of a carbonyl group at C-4 and cleavage of the bond between C-3 and C-4  Y  CH OPO " 2  3  -ZrT  CH OP0 2  0  HO BH  I  ENZ  -OH^B ENZ  HO  BH ENZ  BH  I  R  | ENZ  Figure 1.11. Mechanism of Class II Aldolases.  _  CH 0P0 "  3  2  "A  I  ENZ  -  = 0  Zn  3  Zn  +  CH OH  H  2  HB  I  +  ENZ  I ENZ  HB  +  I  ENZ  Chapter I: Stereochemical Inversion  in a retro-aldol reaction.  '  18  Polarization of the carbonyl bond at C-2 by Z n  2 +  facilitates  formation of a metal-bound enolate anion. The aldehydic product is then released from the enzyme's active site. The reaction is completed by protonation of the D H A P anion by an active site base, thereby regenerating active enzyme in a protonation state suitable for the retro-aldol reaction direction. Crystallographic work on two DHAP-dependent Class II aldolases,  fructose-  bisphosphate aldolase (Cooper et al, 1996) and fuculose-1-phosphate aldolase (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996b), indicates that the DHAP-dependent Class II aldolases should be further divided into two groups based on their structure - those which are homodimers and have two Z n  2 +  ions per subunit (eg. fructose-bisphosphate aldolase),  and those which are homotetramers with one Z n  2 +  ion per active site (eg. fuculose-1-  phosphate aldolase) (Cooper et al, 1996). There is little amino acid sequence similarity between these two groups of Class II aldolases. Recent work on L-fuculose-l-phosphate aldolase has led to the proposal of a new mechanism. This was the result of the observation of the structure of this aldolase with the bound inhibitor phosphoglycolohydroxamate (Figure 1.12). As a potent inhibitor (Xj = 5 u,M), this molecule niimics the enediolate intermediate or perhaps a transition state of the Class II aldolases. Phosphoglycolohydroxamate forms a bidentate chelate in which both the enol and hydroxamate ligands are coordinated to the Z n  2 +  ion (Dreyer & Schulz, 1996a).  Chapter I: Stereochemical Inversion  His  19  0  s  ° 3  His  _  If  His—-Zn" His  P  ^—OPO3-  v  His—Zn' His  cr  Bound phosphoglycolohydroxamate  QT  Bound dihydroxyacetone phosphate (enediolate)  F i g u r e 1.12. Coordination of the Active Site Z n Ion by Phosphoglycolohydroxamate and the Enediolate of Dihydroxyacetone Phosphate. 2 +  In the new mechanism for Class II aldolases (Figure 1.13), polarization of the carbonyl bond increases the acidity of the protons on C-3 of D H A P , one of which is abstracted by a general base, most likely the carboxylate group of E73, resulting in a metal-bound enediolate intermediate. Nucleophilic attack by this species on the carbonyl group of the incoming Llactaldehyde forms a new carbon-carbon bond. The aldehydic oxygen atom subsequently accepts a proton from a tyrosine hydroxyl group. This residue is thought to help stabilize the developing negative charge and to control the stereochemistry of the nucleophilic attack by hydrogen bonding to the aldehyde (Fessner etal, 1996).  Glu—COOH  Glu—COOH  F i g u r e 1.13. A Proposed Mechanism for Catalysis by the Class II Aldolase L-Fuculose-1Phosphate Aldolase.  Chapter I: Stereochemical Inversion  20  Although aldolases are noted for their exceptional stereospecificity, Class II tagatose1,6-diphosphate aldolase occasionally binds the aldehydic moiety inversely, and condenses it to D H A P (see Figure 1.14). Conversely, this enzyme is able to cleave the epimeric sugar, fuculose-l,6-diphosphate, although at a much slower rate (1%) than it cleaves its normal substrate, tagatose-l,6-diphosphate. The net effect is an overall epimerization (Fessner & Eyrisch, 1992).  CH 0P0 ~ 2  3  -H  HO-  -H  H-  normal reaction  CH 0P0 "  3  2  CH 0P0 ~  -OH  D-tagatose-1,6-diphosphate  H0-  -H  H-  -OH  H-  -OH CH 0P0 ~  3  2  CH 0P0 2  3  =0 side reaction  CH,0H  CHO  -OH 2  2  =0  =0 H0-  CH 0P0 ~  3  = 3  D-fructose-1,6-diphosphate  F i g u r e 1.14. Epimerisation of D-Tagatose-l,6-diphosphate Catalysed by D-Tagatose-1,6Diphosphate Aldolase.  L-Fuculose-l-phosphate aldolase, which has amino acid homology (34% identity) (Dreyer & Schulz, 1996b) to L-ribulose-5-phosphate 4-epimerase, is a metalloenzyme (Ghalambor & Heath, 1962; Ghalambor & Heath, 1966) and is the third enzyme in the bacterial L-fucose metabolic pathway (Figure 1.15) (Chen et al, 1987). Rhamnulose-1phosphate aldolase and fuculose-1 -phosphate aldolase, despite the similarity between the reactions they catalyse and their roles in the parallel metabolic pathways (see Figure 1.15) for bacterial rhamnose and fucose utilization, have only a small amount of homology in their  Chapter I: Stereochemical Inversion  21  amino acid sequences. Together with some additional genetic evidence, these two aldolases are thought to have evolved convergendy (Moralejo et al, 1993).  CHO H0-  -OH  H-  -OH  H0-  =0 isomerase  Fucl  -H CH  H-  HO-  -OH  HO-  -H  HO-  -H CH  3  L-rhamnose  J .  H-  -OH -OH  kinase  H-  -H  FucK  H0-  CH OP0 2  2  isomerase RhaA  H-  -OH  H0-  -H  HO-  -H CH  3  L-rhamnulose  CH 0P0 "  FucA  2  3  =0  L-fuculose1-phosphate  CH 0H =0  L-fuculose1 -phosphate aldolase  -H CH  3  L-fuculose  CHO  H-  V  3  r=0  ATP ADP  -OH  CH  L-fucose  -OH  -OH  H-  3  H-  2  2  -H  H-  CH 0P0  CH OH  J  kinase RhaB  H-  CHO 3  H0-  -H  HO-  -H CH  3  -H CH  -OH  H0-  DHAP  2  =0  ATP ADP  V  CH 0H  3  3  L-lactaldehyde L-rhamnulose1-phosphate aldolase RhaD  L-rhamnulose1-phosphate  Figure 1.15. Parallel Pathways of L-Fucose and L-Rhamnose Dissimilation. These sugars are degraded to dihydroxyacetone phosphate and L-lactaldehyde, which are further metabolized. The enzymes in each pathway are L-fucose isomerase (Fuel), L-fuculose kinase (FucK), L-fuculose-1-phosphate aldolase (FucA), L-rhamnose isomerase (RhaA), Lrhamnulose kinase (RhaB), and L-rhamnulose-1 -phosphate aldolase (RhaD).  Chapter I: Stereochemical Inversion  22  Summary  Enzymes are among the most efficient catalysts known. Therefore, an understanding of their mechanisms can provide valuable insights into the nature of chemical catalysis and shed Hght on the course of natural evolution. This thesis strives to discover the nature of the mechanism used by L-ribulose-5-phosphate 4-epimerase, whether it follows a dehydration/rehydration pathway or a retroaldol/aldol mechanism like that used by Lfuculose-l-phosphate aldolase. This work necessitates a plentiful source of the epimerase. Therefore, L-ribulose-5-phosphate 4-epimerase was cloned from Escherichia coli (Chapter II) and the recombinant enzyme was shown to have comparable properties to the naturally abundant E. coli enzyme. The epimerase has been shown to be dependent on metal ions for activity (Deupree & Wood, 1972) and have amino acid sequence similarity to L-fuculose-1phosphate aldolase (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996b). The metal ion ligands of the epimerase are determined (Chapter III) and are shown to be the same as those in the aldolase. After establishing a structural link between the aldolase and L-ribulose-5-phosphate 4-epimerase, it was thought that these two enzymes likely share a mechanistic strategy. Therefore, the epimerase and mutant epimerases were assayed for aldolase activity (Chapter IV). Low levels of activity were observed, supporting the notion that this enzyme utilizes a retroaldol/aldol mechanism. A number of possible future experiments are described in Chapter V which will provide further evidence for a retroaldol/aldol mechanism, identify  Chapter I: Stereochemical Inversion  23  the catalytic residues in the epimerase and show the differences in the active sites of the epimerase and the aldolase.  Chapter II  Cloning, Overexpression and Characterization of L-Ribulose-5-Phosphate 4-Epimerase from Escherichia coli  Introduction  Overexpression Systems A vital component in any study of a protein or enzyme is a source of homogeneous, soluble, and active material, preferably in milligram quantities.  Many small proteins or  peptides may be purified from their normal biological source, or, if small enough, peptide synthesis may be feasible. Frequently, however, overexpression in a suitable biological host system is required to obtain sufficient quantities of material to carry out the planned studies. The term overexpression applies to a technique in which the D N A encoding a protein is introduced into bacteria and the amount of this protein that is produced is significantly higher than that of a typical endogenous protein. The theoretical maximum  24  Chapter II: Cloning, Overexpression and Characterisation yield of a protein from one litre of an Escherichia  25  coli culture at 10 cells per m L is 75 mg, or 9  50% of the total protein (New England Biolabs, 1997 catalogue). Besides providing larger quantities of the desired protein, use of an overexpression system often provides a more economical means of obtaining the protein. Many natural sources contain only small quantities of the desired protein, and much time is spent purifying it from the necessarily large quantities of source material. A variety of overexpression systems are available to suit various needs: bacterial (usually  E. coli), yeast, insect cell and mammalian (tissue culture) are currently in use. The  molecular biology of  E. coli is well understood, and consequently, many overexpression  systems are available for this host. In addition, J B .  coli is inexpensive to grow, and it grows  quickly. Methods leading to the optimization of these systems are numerous and generally exploit every aspect of protein biosynthesis. A major disadvantage with the use of bacterial overexpression systems is their inability to perform many of the post-translational modifications found in eukaryotic proteins. In order to achieve overexpression, the gene encoding the protein of interest must first be cloned into an overexpression vector. A vector is a small piece of closed circular extrachromosomal, double stranded D N A and normally confers on the host cell antibiotic resistance which helps in vector selection and propagation. Vectors, have an origin of replication which determines the number of copies of the plasmid in the cell and which is where replication begins. Once the gene of interest is inserted into the vector, the resulting construct is called a plasmid. Not all vectors can be used for protein expression. Besides  Chapter II: Cloning, Overexpression and Characterisation  26  the elements found on cloning vectors, overexpression vectors also have a promoter, a ribosomal binding site (RBS), a Shine-Dalgarno sequence (SD) and a translation terminator. Many of these features have been optimized in commercially available overexpression vectors. The first step in protein expression is transcription of D N A into R N A (by R N A polymerase); the resulting transcript is called m R N A . A promoter is a sequence of D N A which enhances transcription of the m R N A for a gene.  The promoter is positioned  approximately 10 to 100 base pairs upstream of the ribosomal binding site and is normally under the control of a regulatory gene. This gene may be either on the plasmid itself or integrated into the host chromosome. E.  promoters consist of a hexanucleotide sequence  located about 35 base pairs upstream of the transcription initiation codon (the "-35 region") which is separated from another hexanucleotide sequence (the "-10 region") by a short nucleotide spacer. Promoters useful for overexpression should be sufficiently strong that at least 10% of the total soluble cellular protein is from overexpression. If the protein to be overexpressed is toxic to the host cell, it is essential that the control over gene transcription and translation be tight. The second step in protein biosynthesis is translation of the m R N A message into a polypeptide by ribosomes.  The ribosomal binding site is a region of D N A sequence  downstream from the promoter (at the 5' end of the mRNA) and spans up to approximately 54 nucleotides between positions -35 and +22 of the m R N A coding sequence. The ShineDalgarno (SD) is a sequence which interacts with the complementary 3' end of the 16S  Chapter II: Cloning, Overexpression and Characterisation r R N A (part of the ribosome) during translation initiation. Spacing between the SD and the start codon varies from 5 to 13 nucleotides, and influences the efficiency of translation initiation. The sequence of the spacer should eliminate the potential of any secondarystructure formation in the m R N A transcript, as these structures can reduce the efficiency of translation. A stop codon at the 3'-end of the coding sequence is the signal for termination of translation. A number of stem-loop structures in the m R N A transcript (of the D N A ) loccated downstream of the stop codon protect the m R N A from exonucleolytic degradation and extend the m R N A half-life. A n excellent review by Makrides (1996) on overexpression in  E. coli provides more detail on considerations in choosing and designing overexpression  vectors for this host. The overexpression vector into which the  araD gene was cloned is a modified version  of the pBS vector available from Stratagene. Modification of this vector in the laboratory of Jeremy Knowles (Hermes etal, 1990) removed a portion of the multiple cloning site and rendered the regulatory genes  lad and lacZ non-functional. In addition, the T7 promoter  sequence was replaced with a trc promoter. This promoter is a hybrid between the lac and tac promoters and is much stronger than either of them (Amann & Brosius, 1985; Amann etal, 1983). This plasmid requires no induction because it is not repressed, and was found to express triosephosphate isomerase (TIM) at roughly 100 mg of protein per litre of cell culture (Hermes et al, 1990). This level of overexpression appears to be higher than the theoretical yield, but could arise from cultures which were grown to higher cell density.  27  Chapter II: Cloning, Overexpression and Characterisation  28  Because of the high levels of overexpression achieved with pBSTIM, araD was cloned into this vector in place of tpi, the gene encoding T I M .  T h e araBAD Operon Mammals neither metabolize nor intestinally absorb the plant sugar L-arabinose. Therefore, the bacteria which inhabit the intestine are periodically presented with a feast of this pentose. The araBAD operon is a transcriptional unit which encodes the three structural genes for bacterial L-arabinose chssimilation (Figure 2.1). The operon can be transcribed (ie. m R N A is formed) under conditions of low glucose concentration and in the presence of Larabinose (Lee & Bendet, 1967; Lee et al, 1986; Sa-Nogueira et al, 1997). Even when grown in minimal medium and fully induced with L-arabinose, these enzymes, which are present in approximately equimolar amounts (Lin et al, 1985), each comprise only three to four percent of the total cellular protein (Lee & Bendet, 1967). Therefore, in order to obtain large quantities of any of these enzymes, and to allow for future site-directed mutagenesis studies, their genes must be cloned into an overexpression vector.  Chapter II: Cloning, Overexpression and Characterisation  29  Structural genesaraC  \ara0 \ 2  L araO,  CAR aral  araB  araA  araD  •*- Regulatory gene-*araBAD mRNA-  L-arabinose isomerase L-arabinose  L-ribulose  L-ribulokinase  L-ribulose-5-P 4-epimerase L-ribulose-5-P  D-xylulose-5-P  F i g u r e 2.1. A genetic map of the E. coli araC and araBAD operons indicating the proteins encoded by the araBAD operon and the reactions which these proteins catalyse. The end product, D-xylulose-5-phosphate, is further metabolized by the pentose phosphate pathway.  After entering the cell, L-arabinose is sequentially converted to L-ribulose, L-ribulose5-phosphate, and D-xylulose-5-phosphate by the action of the three structural gene products (see Figure 2.2):  L-arabinose isomerase (AraA, E C 5.3.1.4), L-ribulokinase (AraB, E C  2.7.1.16), and L-ribulose-5-phosphate 4-epimerase (AraD, E C 5.1.3.4), respectively (Burma & Horecker, 1957; Lee et al, 1986; Sa-Nogueira et al, 1997; Simpson et al, 1958). DXylulose-5-phosphate is further metabolized through the pentose phosphate pathway (Heath et al., 1958; Rappoport et al, 1951; Sa-Nogueira et al, 1997; Simpson et al, 1958). The two enzymes of particular interest to this study are L-ribulose-5-phosphate 4-epimerase, and Lribulokinase.  Chapter II: Cloning, Overexpression and Characterisation  30  L-Ribulose-5-Phosphate 4-Epimerase: A n E n i g m a t i c R e a c t i o n M e c h a n i s m  This epimerase is of particular interest because, unlike the vast majority of epimerases and racemases, this enzyme appears to invert the stereochemistry (Figure 2.2) at an unactivated centre without the aid of a cofactor. Therefore, this enzyme must use a unique mechanism which does not involve direct deprotonation of the non-acidic proton at the C-4 stereocentre.  CH OH  CH OH  2  2  =0 HO-  -H  HO-  -H  L-ribulose-5-phosphate 4-epimerase  =0 HO-  H-  -H -OH  CH OP0 = 2  3  L-ribulose5-phosphate  (LRu5P)  D-xylulose5-phosphate  (DXu5P)  F i g u r e 2.2. L-Ribulose-5-Phosphate 4-Epimerase Interconverts L-Ribulose-5-Phosphate and D-Xylulose-5-Phosphate.  Several studies were performed on this enzyme, but a clear description of the mechanism remains elusive. The enzyme was definitively demonstrated not to use a N A D  +  cofactor (Deupree & Wood, 1970). Furthermore, experiments using isotopically-labelled LRu5P, DXu5P or solvent (Davis etal, 1972; McDonough & Wood, 1961; Salo etal, 1972) showed that epimerisation appears to occur without breaking the carbon-hydrogen bond between C-4 and C-3 (see Chapter I for details).  Chapter II: Cloning, Overexpression and Characterisation  31  A divalent metal ion was required for activity. The naturally occurring metal ion was not determined; however, in the presence of at least 1 m M E D T A , the enzyme was rendered inactive. Various degrees of activity could be restored by addition of divalent metal ions (1 uM to 10 mM) to the assay cuvettes in the order M n  2 +  > Co  2 +  > Ni  2 +  > Ca  2+  > Zn  2 +  > Mg  2 +  (greatest to least degree of reactivation). These studies led to the proposal of two possible reaction mechanisms in which either the carbon-oxygen bond or a carbon-carbon bond at C-4 was broken and then reformed since the epimerase appears not to break the carbon-hydrogen bond at C-4 (Deupree & Wood, 1972). These mechanisms are discussed and illustrated in Chapter I. The native enzyme was found to have a molecular mass of about 110 kDa (Deupree & Wood, 1975; Gielow & Lee, 1975) as determined by high speed sedimentation equihbrium analysis and, until the gene was sequenced, was thought to have three subunits of about 35 kDa each (Gielow & Lee, 1975). After the  araD gene was sequenced and analysed, the  enzyme was determined to have four subunits of about 25.5 kDa, as calculated from the predicted amino acid sequence of the epimerase from E. coli (Lee 1990),  etal., 1986; Mineno et al.,  Bacillus subtilis (Sa-Nogueira et al., 1997), and Salmonella typhimurium (Lin et al, 1985,  corrected by A . Bairoch, unpublished data, Swiss Protein Data Bank, 1995). Kinetic parameters were determined for the enzyme, and p H rate profiles established that it is most active between p H 7 to 9 (Burma & Horecker, 1958b; Lee et al., 1968; Wolin  etal, 1958).  Chapter II: Cloning Overexpression and Characterisation  32  L-Ribulokinase: A M e a n s of Producing L-Ribulose-5-Phosphate  L-Ribulokinase is a typical magnesium-requiring kinase (Burma & Horecker, 1958a; Lee & Bendet, 1967; Lee & Englesberg, 1966; Simpson & Wood, 1956; Simpson & Wood, 1958; Wolin  etal, 1958) which phosphorylates L-ribulose with adenosine triphosphate (ATP)  to form L-ribulose-5-phosphate and adenosine diphosphate (ADP) (Figure 2.3). Like the epimerase, this enzyme has been purified from a number of bacterial sources, including E .  coli (Lee & Bendet, 1967; Lee & Englesberg, 1966), Klebsiella pneumoniae (formerly Aero bacter aerogenes) (Simpson & Wood, 1958), and Lactobacillus plantarum (Burma & Horecker, 1958a).  CH OH  CH OH  2  =0 HO-  -H  HO-  -H CH OH 2  L-ribulose  2  ATP  =0  ADP  L-ribulokinase  HO-  -H  HO-  -H  CH OP0 = 2  3  L-ribulose5-phosphate (LRu5P)  F i g u r e 2.3. Phosphorylation of L-Ribulose by L-Ribulokinase Provides L-Ribulose-5Phosphate, the Substrate for L-Ribulose-5-Phosphate 4-Epimerase.  Since the substrate for L-ribulose-5-phosphate 4-epimerase is not commercially available, it was necessary to prepare the substrate ourselves.  Using L-ribulokinase to  phosphorylate L-ribulose is a convenient method of preparing this phosphopentoketose. Although the kinase has been purified to homogeneity, this is not necessary for production of L-ribulose-5-phosphate.  According to the method of R.L. Anderson (1966), L-  Chapter II: Cloning, Overexpression and Characterisation  33  ribulokinase may be partially purified from an L-arabinose induced strain of bacteria which cannot produce active L-ribulose-5-phosphate 4-epimerase. When the bacteria are grown on minimal media in the presence of arabinose the levels of expression of the  araBAD gene  products increase so that the bacteria may utilize this sugar as an energy source. This crude enzyme preparation can then be used as a catalyst for phosphorylation of L-ribulose with ATP.  In the absence of the epimerase, arabinose can proceed only to LRu5P, which  accumulates. It is important to maintain the p H of this reaction mixture between 6 and 7.5; the enzyme is unstable below p H 6 and above p H 7.5 (Burma & Horecker, 1958a) and the maximal activity lies around p H 7.6 (Lee & Bendet, 1967). Even when fully induced, this enzyme is estimated to comprise only 3 or 4% of the total cellular protein (Lee & Bendet, 1967). For our purposes this was sufficient, although it would be more convenient to have this enzyme overexpressed. Besides phosphorylating L-ribulose, L-ribulokinase can phosphorylate its enantiomer, D-ribulose (Burma & Horecker, 1958a; Lee & Bendet, 1967; Lee & Englesberg, 1966; Simpson & Wood, 1958), as well as adonitol (also known as ribitol) and L-arabitol (Simpson & Wood, 1958).  Chapter II: Cloning, Overexpression and Characterisation  34  Results and Discussion  Preparation ofL-Ribulose-5-Phosphate L-Ribulokinase was initially obtained from L-arabinose induced E. coli Y1090 (modified) (see the experimental section for details), a strain which lacks a functional Lribulose-5-phosphate  4-epimerase.  Initial difficulties in preparing L-ribulokinase for  phosphorylation were overcome by maintaining the temperature of the cell paste and the crude cell lysate between 10 and 4°C. Although the purified kinase is reported to be heatstable (Lee & Bendet, 1967; Lee & Englesberg, 1966), in the cell lysate it was much less stable, perhaps due to proteolytic degradation. After a number of successful preparations, we were unable to obtain L-ribulokinase activity from this strain, possibly as a result of a spontaneous mutation or from movement of the transposon (TnlO) in the genome. At this point we tried cloning the L-ribulokinase, but our attemps have thus far been unsuccessful. Ultimately we found that another 4-epimerase-less strain, E. <W/MC4100, worked very well as a stable source of L-ribulokinase when used in Anderson's procedure. L-Ribulose-5-phosphate was purified by anion exchange chromatography and the fractions greater than 90% pure (by ' H - N M R spectroscopy; see Figure 2.4) from a number of preparations were pooled and lyophilized. The concentration of a stock solution of the substrate was accurately determined by using a sample of known dilution to initiate a cuvette (incubated at 37°C) containing all the  Chapter II: Cloning, Overexpression and Characterisation  35  normal assay components except LRu5P. A large excess of enzyme was used to epimerise the sugar rapidly.  Attempted Cloning of L-Ribulokinase The difficulties in preparing active L-ribulokinase from  E. coli Y1090 (modified) led  us to try cloning this enzyme for overexpression. The araB gene of  E. « ? / / D H 5 a P , encoding  L-ribulokinase (EC 2.7.1.16) was amplified directely by PCR (40 cycles) on whole cells. The P C R product, when thoroughly digested by Nde I proved to have a restriction site for this enzyme at nt 827 giving fragments ca. 827 nt and ca. 874 nt long. It was thought that partial digestion by Nde I might yield a product which was digested at the 5' end of the gene (where this site had been introduced on the primers) but not in the middle. The products of partial Nde I digestion were separated on an agarose gel and the large (ca. 1701 nt) fragment isolated. Ligation of this fragment into a p E T - l l a vector prepared for cloning by Nde I and BamW I digestion was not successful. isolated from  By this time, preparation of LRu5P using L-ribulokinase  E. coli MC4100 had proven successful, so further attempts at cloning  ribulokinase were abandoned.  L-  Chapter II: Cloning, Overexpression and Characterisation  36  H-C5  A. 7  A.6  4.5  A.3  A. A 1  Figure  4.2  4.1  A.O  H(ppm)  2.4. Proton-NMR Spectrum of Purified L-Ribulose-5-Phosphate (free acid) at  500 MHz (in D 0). 2  Chapter II: Cloning, Overexpression and Characterisation  37  A superior cloning strategy for this gene would be to introduce a Nco I site at the 5'end of the gene during PCR, and then to clone this fragment into a compatible vector. Nde I is the preferred restriction site at the 5'-end of genes cloned into overexpression vectors because it contains A T G , the start codon, as the second half of its six-base recognition sequence ( C A T A T G ) . Many overexpression vectors which use this site have optimized distances and nucleotide sequences between the "-10" site and the start codon for maximal protein expression. Nco I can also be used in this manner, as its recognition sequence ( C C A T G G ) also contains A T G . However, use of this restriction enzyme for cloning the 5'end of the gene into an overexpression vector restricts the second amino acid of the protein sequence to one of five residues: V , A , D , E , or G . The second residue in the L-ribulokinase sequence happens to be alanine.  Ooning, Expression, Purification and Characterization ofthel^BJbidose~5-Phosphate 4-Epime Escherichia coli The  araD gene  encoding L-ribulose-5-phosphate 4-epimerase (EC 5.1.3.4) was  amplified directly from whole cells of  E. coli  D H 1 using 40 cycles of PCR.  The P C R  product was cloned into the modified pBS vector described earlier, creating the new construct, p R E l (this work was performed by Dr. M . Tanner). Prior to sequencing the  araD gene contained on p R E l ,  we were concerned that there  may be some PCR-derived errors in this gene because of the large number of P C R cycles used. The number of cycles necessary can be reduced by using a plasmid as a template as  38  Chapter II: Cloning, Overexpression and Characterisation  opposed to genomic D N A when using whole cells. We obtained a plasmid, pHC5, from the  E. coli genome including the  laboratory of Dr. R.E. Moses which contained a section of the complete  araD gene (Chen etal, 1990). We planned to use this plasmid as the template for  PCR amplification of the the  araD gene and to redone it into the modified pBS vector. While  araD gene contained on the plasmid pHC5 was  successfully amplified by P C R (30 cycles), this P C R product was not successfully cloned into the same modified pBS vector used in formation of p R E l . These difficulties most likely arose during isolation of D N A from agarose or during the ligation phase of cloning. Ultimately, the portion of plasmid p R E l containing the  araD gene was sequenced and  was shown by sequencing not to contain any PCR-derived errors. In order to avoid contamination by endogenous enzyme, p R E l was transformed into a strain of E. coli which , . . _ __, . „ , . lacks a functional L R u 5 P 4-epimerase. For this purpose, h. r  r  r  F i g u r e 2.5. S D S - P A G E of „ , Crude and Purified Protein.  coli Y1090 (modified) was used, although E. coli MC4100 °  v  T  n  e  crude  P  r o t e i n  and  . . . , „ „ . , , , epimerase should work equally well. High levels of overexpression gp j T T  c  S  a r a t e c  components of purified 4-  samples by 12°/  were SDS  , i/ i • r o cv J J were observed (as shown in Figure 2.5), and were estimated  PAGE. The molecular . , . . weight standards were: 66 c i /nr./ c I I I LI A kDa, BSA; 29 kDa, carbonic to account for at least 60% of the total soluble protein. A , ' , ._ , _ anhydrase; and 18.4 kDa, p, . lactoglobulin. cell paste containing the highly overexpressed epimerase was 0  r  Chapter II: Cloning, Overexpression and Characterisation  39  lysed by passage through a chilled French Press cell and clarified by ultracentrifugation. Highly negatively-charged impurities were removed from the clarified lysate by passage through a DE-52 column, from which the epimerase was eluted using 0.4 M NaCl in 10 m M potassium phosphate buffer (pH 7.6, containing 10% glycerol). The epimerase was then purified from other proteins by ion-exchange H P L C on a Q-Pak column with a linear gradient of 0 to 0.4 M NaCl in 10 m M potassium phosphate buffer (pH 7.6, containing 10% glycerol). The epimerase, which eluted at about 0.3 M NaCl, was collected as a highly concentrated fraction. The purity was estimated to be at least 95% by visual examination of a Coomassie Blue stained SDS-polyacrylamide gel (12%) (Figure 2.5). At least 100 mg of purified enzyme were obtained from one litre of bacterial culture.  This level of  overexpression corresponds to that observed for T I M in this vector (Hermes et al., 1990). The modified pBS vector used in this work previously had its site for transcriptional repression deleted, which means that transcription of the gene of interest on the vector is effectively always "on". We found that overnight broth cultures when used as the inoculum for large cultures were incapable of overexpressing the epimerase. In order to achieve the high levels of overexpression of which this vector is capable, the inoculum for cultures used to prepare the epimerase was a colony. The reason for this is not understood. The overexpressed enzyme has electrophoretic mobility corresponding to about 31 kDa, similar to that observed for the recombinant enzyme from  S.typhimurium(Lin et al,  1985). The expected mass is 25 kDa, and one would expect an electrophoretic mobility which agrees with this. The lower mobility might be caused by the epimerase having more  Chapter II: Cloning, Overexpression and Characterisation  40  basic residues than an average protein. Based on translations of the genetic code, amino acid sequences can accurately be predicted. These sequences are used to calculated subunit molecular weights based on amino acid composition. The predicted subunit masses can then be verified by electrospray ionization mass spectrometry (ESIMS) which has been shown to be (Ashton et al, 1994) an accurate method for determining protein subunit mass. The subunit mass of the recombinant epimerase was determined in this fashion and was consistent with the mass calculated from the expected amino acid sequence (calculated: 25520 Da; found: 25522 ± 4 Da). The entire enzyme, being a homotetramer, has a mass fourtimesthis value. Table 2.1 lists the subunit masses of LRu5P 4-epimerases from various sources. Protein concentration was required to be accurate for determination of the zinccontent of the zinc-reconstituted enzyme.  Therefore, the extinction coefficient of the  purified enzyme at 280 nm was determined first by the method of Gill & von Hippel (1989), using the relationship,  (2.1)  where 8,n  the extinction coefficient at 280 nm of the native protein,  A,"n  the absorbance at 280 nm of the native protein,  A  the absorbance at 280 nm of the denatured protein, and  d  the extinction coefficient at 280 nm of the denatured protein.  Chapter II: Cloning, Overexpression and Characterisation  41  T a b l e 2.1. Predicted Subunit Molecular Masses of L-Ribulose-5-Phosphate 4-Epimerase. Source Organism  Subunit Length  Predicted Subunit Mass (Da)  E. coiiKW  231  25520  231  25504^  E. coliV>/t  h  S. typhimurium B. subtilis  c  d  E. coli DH1  e  248  27059  1  229  ca. 25700  231  25520  Mineno et al, 1990.  a  ''Lee etal, 1986. ' Lin etal, 1985. Sa-Nogueira , 1997. This study. There is a probable frame-shift mutation here, likely from an error in D N A sequencing (Lee et al, 1986; and A . Bairoch, unpublished observation, Swiss Protein Data Bank). Bairoch predicts 231 amino acids with a mass of 25540 Da. There are three amino acid substitutions in this sequence compared to the E. coli K12 sequence: I instead of V at residue 50, A rather than T at residue 70, and E rather than D at position 216. Even with these substitutions, the calculated mass should be 25518.  rf  e  f  g  8 was calculated from the number of Trp, Tyr residues and disulphide bonds in the protein d  as described by Edelhoch (1967): 4 Trp + 9 Tyr = 4(5690 cm" M" ) + 9(1280 cm" M" ) = 34280 cm" M" . 1  1  1  1  1  1  Protein at several concentrations was mixed with an equal volume of either water or guanidinium'HCl, and the absorbance of each measured at 280 nm. Using this method, the native epimerase was calculated to have an extinction coefficient of 8 = 34314 cm" M" 1  1  (where M" refers to the subunit concentration), or 1.76 m L mg" cm" as the average of three 1  1  1  determinations, which differed from that given by Lee, Patrick & Masson (1968) who  Chapter II: Cloning, Overexpression and Characterisation  42  reported 1.57 m L mg" based on the dry weight of the enzyme. This method can have errors 1  of up to 30% (Gill & von Hippel, 1989), which we deemed too high. As a result the extinction coefficient was also calculated by total amino acid analysis on a pure protein solution of known absorbance to which 0.1 m M norleucine (an unnatural amino acid) had been added as an internal standard. The relative amounts of alanine and norleucine were used to determine the concentration of the protein sample (the enzyme contains 21 alanine residues per subunit). This resulted in the extinction coefficient 8 =1.73 m L mg" cm" 1  1  (33717 cm" M" per subunit) for the epimerase. This value was used for all determinations 1  1  of protein concentration measured at 280 nm. It is of interest to note that a commonly used colorimetric technique (Bradford, 1976) calibrated to B S A gave values two to three times greater than those calculated from direct measurement of the A  2 8 0 n m  .  In this method, the  protein solution is mixed with an acidic solution of methanol and Gommassie Brilliant blue. After a five minute incubation, the absorbance at 595 nm is measured, and a calibration curve using B S A as a standard is used to interpolate the protein concentration of the standard. B S A is often quoted in the literature as a standard, but this protein has been observed to develop a deeper colour than other proteins (Bio-Rad 1996 catalogue), and is therefore not the best standard to use. The epimerase must then develop an even deeper colour than BSA; perhaps BSA has a lower proportion of basic and aromatic residues than the epimerase. The purified epimerase was assayed in the L-ribulose-5-phosphate (LRu5P) to Dxylulose-5-phosphate (DXu5P) direction using the assay described by Davis et al. (1972). In  Chapter II: Cloning, Overexpression and  Characterisation CH OH  CH OH =0  L-ribulose-5-phosphate 4-epimerase  -H  HO-  CHO  2  2  HO-  43  HO-  =o  HO-  -H  H-  -OH  -OH  H-  -OH  -H CH op0 =  CH op0 =  L-ribulose5-phosphate (LRu5P)  D-xylulose5-phosphate (DXu5P)  2  3  2  -H  CH op0 ~ D-ribose5-phosphate (DR5P)  3  2  3  transketolase CH OH 2  =0  CHO triosephosphate isomerase  H-  -OH  -OH  HO-  CH OPQ D-glyceraldehyde 3-phosphate =  2  NAD  +  CH OH  NADH  2  -OH CH OPO 2  3  -H  H-  -OH -OH  H-  CH OPO " 2  a-glycerophosphate CH OPO dehydrogenase (a-GDH) dihydroxyacetone phosphate (DHAP)  3  2  3  sedoheptulose7-phosphate  =  3  F i g u r e 2.6. A Coupled Assay for L-Ribulose-5-Phosphate 4-Epimerase Activity.  this assay (see Figure 2.6), the epimerase converts LRu5P to DXu5P. DXu5P and ribose-5phosphate (contained in the assay mixture) are transformed into sedoheptulose-7-phosphate (S7P) and glyceraldehyde-3-phosphate by the action of transketolase. Triosephosphate isomerase (TIM) converts glyceraldehyde-3-phosphate into dihydroxyacetone phosphate (DFIAP), which is then reduced with N A D H by the action of a-glycerophosphate dehydrogenase ( a G D H ) . The oxidation of N A D H to N A D  +  is followed as a decrease in  absorbance at 340 nm. A small epimerase-independent background rate was observed and was determined for each cuvette individually. This rate was also observed by Davis et al.  Chapter II: Cloning, Overexpression and Characterisation  44  (1972) and by Deupree and Wood (1975). The background rate was seen in the absence of both LRu5P and 4-epimerase, resulting perhaps from impurities in the coupling enzymes and reagents. This rate was subtracted from the rate of N A D epimerase.  +  formation after addition of the  A n alternate assay ( G A P D H assay; Figure 2.7) in which glyceraldehyde-3-  phosphate is oxidized to l-arseno-3-phosphoglycerate  (an unstable molecule which  decomposes rapidly) by the action of glyceraldehyde-3-phosphate dehydrogenase (GAPDFT) could also be used.  This assay, which exhibited a very high background rate and was  therefore thought to be unreliable, was not used routinely.  CH OH  CHO  CH OH  2  2  =0  L-ribulose-5-phosphate 4-epimerase  HO-  -H  HO-  -H CH OP0 = 2  Kinetic constants for the  HO-  =0  HO-  -OH  -H H-  -OH CH OP0 =  3  2  -OH  CH 0P0 ~ D-ribose5-phosphate| (DR5P)  3  2  D-xylulose5-phosphate (DXu5P)  L-ribulose-. 5-phosphate (LRu5P)  -H  3  transketolase CH 0H 2  =0  OH H-  O^.OAs0 -OH  CH OP0 " +  arsenate  3  H,0  -OH 2  NADH  3  2  +  H-  -OH  CH OP0 ~ glyceraldehyde-3-phosphate dehydrogenase D-glyceraldehyde (GAPDH) 3-phosphate 2  CH OPO ~ 3  + arsenate  H-  -OH  HOH_  -H'  CHO  NAD  3  H  -OH -OH CH OP0 " 2  3  sedoheptulose7-phosphate  F i g u r e 2.7. A n Alternate Assay for L-Ribulose-5-Phosphate 4-Epimerase Activity.  Chapter II: Cloning Overexpression and Characterisation epimerase (determined using the assay described by Davis etal. (1972) were K  45 M  0.008 m M and k  CM  = 0.026 ±  = 13.5 ± 0.3 s\  Determination of Metal Ion Content L-Ribulose-5-phosphate 4-epimerase can accept a variety of divalent metal ions for activity (Deupree & Wood, 1972); however, it is not known which of these is present in the enzyme isolated from bacteria at natural abundance. A purified sample of the recombinant enzyme and a sample of its buffer were digested in 10% ultrapure nitric acid. These samples were analysed for metal ion content by inductively-coupled plasma mass spectrometry (ICPMS). The results indicated that the enzyme preparation contained a mixture of metal ions in order of decreasing abundance: Z n  2 +  > Mn  2 +  > C u . It is possible that the very high 2+  levels of overexpression resulted in the enzyme sequestering a variety of metal ions as it is formed. The metal ion distribution we observed may therefore vary depending on the metal ions present in the growth medium and probably does not reflect the metal ion preferences obtained under normal growth conditions. For further studies we wished to prepare an enzyme with homogenous metal ion content. Zinc is a reasonable choice because it is the divalent metal ion commonly used for electtophilic catalysis in many enzymes including Class II aldolases (see Chapter III for discussion). In order to prepare epimerase with a homogeneous metal ion content, apoenzyme was prepared by treatment of the epimerase with E D T A as described by Deupree and Wood  Chapter II: Cloning Overexpression and Characterisation  46  (1972), followed by dialysis against two changes of metal-free H E P E S / T r i s buffer (50 m M , p H 7.6, containing 10% glycerol). The resultant epimerase was inactive as determined using an "end-point" assay. For this assay, metal-free LRu5P (0.94 mM) was incubated in a buffered solution of the epimerase (at 0.01 mg/mL) for five minutes, at which time it was added to a pre-incubated cuvette containing the coupling system.  Apoenzyme was  reconstituted by incubation with 10 equivalents of ZnCl . Excess metal ions were removed 2  by dialysis against two changes of H E P E S / T r i s buffer (50 m M , p H 7.6, containing 10% glycerol). Inconsistent ICPMS results were obtained after extended dialysis. Therefore, an alternate method for removing metal ions, passage through a size exclusion column, was required to obtain a reasonable result from ICPMS. Incidentally, there was no change in the activity of the enzyme after passage through the size exclusion column. reconstitution, the enzyme was found to have 1.05 equivalents of Z n ICPMS) and kinetic parameters of  2 +  After Zn(II)  per subunit (by  k = 20.4 ± 0.9 s" and K = 0.087 ± 0.007 mM. The 1  czt  M  K of the enzyme was similar to those reported in the literature (Table 2.2) except for that M  of the  L. plantarum enzyme.  Chapter II: Cloning, Overexpression and Characterisation Table 2.2. Comparison of K and 4-Epimerase. M  Source Organism  K  E.coliK/t  0.095  c  M  Constants Determined for L-Ribulose-5-Phosphate  (mM)"  K  b eq  ca.\  K. pneumoniae  0.1  1.86  JL plantarum  1.1  1.2  E.colidm  0.087  1.2  d  e  1  47  for L-ribulose-5-phosphate *for [DXu5P]/[LRu5P] 'Lee 1968; Gielow & Lee, 1975. Wolin etal, 1958; Deupree & Wood, 1975. Burma & Horecker, 1958. ^This study. a  rf  e  Equilibrium Treatment of a commercial sample of DXu5P with the enzyme and subsequent analysis of the equilibrium mixture by ^ - N M R spectroscopy indicated that the equilibrium was slighdy in favour of the DXu5P epimer (see Figure 2.8), in agreement with a previous report by Burma & Horecker (1958b) (see Table 2.2).  The equilibrium constant was  calculated to be 1.2 from intergration of the peaks arising from the C-4 protons of DXu5P and LRu5P.  Chapter II: Cloning, Overexpression and Characterisation  48  Conclusion  The high level of overexpression of L-ribulose-5-phosphate 4-epimerase in  E. coli  transformed with p R E l provided a rich source of the epimerase, which after purification yielded about 100 mg of enzyme per litre of culture. The recombinant enzyme, after zinc ion replacement is indistinguishable from the wild-type  E. coli enzyme.  Large quantities of pure epimerase makes possible detailed studies on the enzyme, including studies on the nature of its reaction mechanism and possibly x-ray crystal structure determination. As the  araD gene is cloned, site-directed mutagenesis studies are also feasible.  Chapter II: Cloning, Overexpression and Characterisation  49  H-C1 DXu5P & LRu5P and H-C3 DXu5P H-C5 DXu5P & LRu5P H-C3 LRu5P  4.6  H-C4 DXu5P  4.4  4.2 1  H-C4 LRu5P  4.0  3.8  H(ppm)  Figure 2.8. Equilibrium between L-Ribulose-5-Phosphate and D-Xylulose-5-Phosphate as Determined by 'H-NMR (500 MHz). Integration of the area below the peaks at 3.94 ppm (H-C4 of LRu5P) and 4.12 ppm (H-C4 of DXu5P) provided relative concentrations of these two sugars.  Chapter II: Cloning Overexpression and Characterisation  50  Experimental Methods  Bacterial Strains and Plasmids E. coli strain D H 5 a F ' (Gibco/BRL) was used for all plasmid manipulations. E. coli strain Y1090 (modified to cure it of pMC9) (a gift from Dr. Neil Gilkes, Department of Microbiology, University of British Columbia) was used as the host for overexpression of recombinant L-ribulose-5-phosphate 4-epimerase. was used in the preparation of substrate. Both were  E. coli strain MC4100 (ATCC #35695)  E. coli Y1090 (cured) and E. coli MC4100  araD139 mutants. E. coli D H 1 was used as the source of genomic D N A for the P C R  amplification of  araD in construction of p R E l , while E. coli DH5ccF' was the source of  genomic D N A for  araB P C R amplification. E.  JM109 (DE3) was used as the host for  p E T - l l a overexpression vectors. Cells for general cloning and overexpression were grown in L B media supplemented with ampicillin (50 pg/mL). The vector used for cloning araD was a modified version of pBSTIM (Hermes  et al,  1990), in which the Nde I restriction site on the original pBS vector (Stratagene) was +  removed by changing nucleotide residue 186 from C to A , and by introduction of an Nde I site at the start codon in the cloning region by changing the bases at -1 and -2 (relative to the start codon of TIM) from C C to A T . These modifications were performed by Dr. S. Pollack in the laboratory of Dr. Jeremy Knowles, Harvard. p E T - l l a (Novagen) was the vector used for attempts at L-ribulokinase cloning. Both of these plasmids encode a gene for ampicillin resistance.  Chapter II: Cloning, Overexpression and Characterisation  51  Unless otherwise noted, general methods for handling and manipulating D N A are those described by Sambrook, Maniatis & Fritsch (1989) or Ausubel  (1992).  Preparation ofL-Ribulose-5-Phosphate L-Ribulose-5-phosphate was prepared according to the procedure of R.L. Anderson (1966).  E. coliMC4100, an araD139, or 4-epimeraseless, mutant of E. coli was grown in 2 L  of a casein hydrolysate-mineral medium (as described by Englesberg (1961): 1% potassium phosphate (pH 7.0), 0.01% M g S G y 7 H 0 , 0.1% ( N H ) S 0 , and 1% Casamino acids) 2  4  2  4  containing yeast extract (0.25 g/L) until the optical density at 600 nm was about 0.4.  L-  Arabinose (2 g) was added to a concentration of 0.4%, and the incubation continued for another 4 hours. L-Ribulokinase was isolated in an ammonium sulfate fraction of the crude cell lysate, and was used without further purification. The reaction mixture contained 4 mmole L-ribulose (601 mg, 429 uL), 5 mmole A T P (2.75 g), 5 mmole M g C l (476 mg), 5 mmole N a F (210 mg), 1 mmole reduced sodium 2  glutathione (310 mg), 2.5 mmole E D T A (12.5 uL of a 0.2 M solution), and L-ribulokinase in a total volume of 200 mL. N a O H was added periodically to maintain p H 7.5. After the reaction was complete, 2.29 m L of glacial acetic acid was added to a final concentration of 0.2 M , the solution suction filtered through Whatman #1 filter paper, and the filtrate neutralized with about 19 m L of 2 M N a O H . L-Ribulose-5-phosphate was purified from other reaction components by ionexchange chromatography on a 100 m L A G 1-X8 (formate) column (ID 19 mm), eluting  Chapter II: Cloning, Overexpression and Characterisation  52  with a stepwise gradient of formic acid: 6 x 100 m L deionized water; 6 x 80 m L I N formic acid; 6 x 80 m L 2 N formic acid, 12 x 40 m L 2.5N formic acid; and 6 x 80 m L 3N formic acid. LRu5P eluted between 2 and 3 N formic acid. Fractions that were greater than 90% pure by N M R were pooled and lyophilized, and brought to p H 7.0 with N a O H . 1  H - N M R (500 M H z , D 0 ) : 6 4.02 (apparent t, J= 5.41 H z , 2 H , H-C5), 6 4.11 (dt 2  J= 5.41 Hz, 1 H , H-C4), 6 4.43 (d, J= 5.41 Hz, 1 H , H-C3), 6 4.56 (d, J = 19.5 Hz, 1 H , H Cl), 6 4.63 (d, J = 19.5 Hz, 1 H , H'-Cl); see Figure 2.4. A C O S Y plot is corresponding to this sample and which aided in peak identification is shown in Appendix B. The concentration of LRu5P in the stock solution was determined by initiating an assay cuvette preincubated at 3 7 ° C and containing all the normal assay components except LRu5P, but which contained 50 ug of 4-epimerase, with 1 0 u L o f a l / 1 0 dilution of the stock solution. The concentration of LRu5P in the cuvette was determined from the absolute change in absorbance at 340 nm due to oxidation of N A D H to N A D . +  Attempted Cloning otL-Ribulokinase The araB gene of E. coli D H 5 a , encoding L-ribulokinase (Lee et al, 1986; see Appendix A), was amplified by 40 cycles of P C R between primers AJ8 (containing a Nde I restriction site) and AJ9 (containing Pst I and BamW I restriction sites) (see Table 2.3). Template D N A was the genomic D N A provided in whole D H 5 a F ' cells which had been harvested, washed and resuspended in an equal volume of sterile distilled water.  Each  reaction contained (in 100 uL): 5 uL D H 5 a F ' cell suspension, 0.1 u M each dNTP, 0.5 u M  Chapter II: Cloning Overexpression and Characterisation each primer, 1.5 m M MgCl , 50 m M KC1, 20 m M Tris-HCl (pH 8.4), and 2.5 U Taq D N A 2  polymerase (Gibco/BRL). 100 uL of mineral oil was added to prevent evaporation during thermocycling. A n initial melt at 95 ° C (10 min), annealing at 55 ° C (5 min) and chain extension at 72 ° C (2 min) was followed by 39 cycles of 9 5 ° C (1.5 min), 5 5 ° C (2 min) and 7 2 ° C (2 min). The reaction was ended with a final extension of 4 minutes at 7 2 ° C .  Table  2.3. Primers Used for P C R Amplification of the  araB Gene.  Primer  Sequence  AJ8  a  5'-GGGAATTCCMT^TGGCGATTGCAATTCGCCTCGAT-3'  AJ9  h  5'-CGGCGG^TCCrGC4GTTATAGAGTCGCAACGGCCTGGGC-3'  a  h  Primer AJ8 , whose 24 3'-end bases are complementary to the 5'-end of the araB gene has a Nde I restriction site (in italics). Primer AJ9 has overlapping BamU I and PstI restriction sites (5' to 3', in italics), and its 24 3'-end bases are complementary to the 3'-end of the araB gene.  The resultant D N A fragment was isolated from a 1.2% agarose gel and purified using a Wizard P C R Preps D N A Purification System (Promega). Both the vector ( p E T - l l a ; Novagen) and the purified P C R fragment were treated with Nde I and BamU I D N A restriction enzymes. The vector was isolated from a 1.2% agarose gel and purified using a Wizard P C R Preps Purification System (Promega). The restricted P C R fragment was not purified since there is an additional Nde I restriction site in the middle of the gene. Partial digestion of the PCR-amplified araB gene, choosing conditions under which approximately half of the D N A was restricted was also tried. These conditions were: 1 u.g P C R fragment (araB), 6 U BamH I (10 U / u L ) , 1 X Buffer D (Promega), water to 6.0 uL,  53  Chapter II: Cloning, Overexpression and Characterisation incubated at 37 ° C for 3 hours. 2 U Nde I, more Buffer D (to keep 1 X concentration), and water (to a total volume of 10 uL) were then added, and the reaction mixture incubated 15 minutes longer.  The mixture was then heated at 80 ° C for 5 minutes, lyophilized and  redissolved in a small volume of water. The restricted  araB gene was isolated and purified  from a 1.2% agarose gel using Wizard P C R Preps D N A Purification System (Promega), ligated into prepared p E T - l l a , and transformed into  E. coli DH5aF'. Resultant colonies  were checked for plasmid containing araB by 1.2% agarose gel electrophoresis of the uncut and restricted  (Pst I and BamW I) plasmid D N A . In addition, promising colonies were  checked for overexpression in JM109 (induced with 0.4 m M IPTG).  Cloning of the araD gene ofEscherichia coli The  araD gene was cloned from E. coli strain D H 1 by generating a P C R product from  50 uL whole cell suspension (harvested and resuspended in an equal volume of water) using Taq D N A polymerase and 40 cycles of thermal cycling ( 9 5 ° C for 1.5 min, 5 5 ° C for 2 min, and 7 2 ° C for 2 min) according to the method of Joshi, Baichwal & Ames (1991). The sequence-specific primers (MT04 and MT05) encompassed the complete gene and contained restriction sites for Nde I and Pst I for cloning into pBSTIM (see Table 2.4). The resulting P C R product was purified by passage through an Ultrafree-MC 10000 N M W L filter unit (Millipore), followed by extraction with phenol and " S E V A G " and then ethanol precipitated.  54  Chapter II: Cloning, Overexpression and Characterisation  55  Table 2.4. Primers Used for P C R Amplification of the araD Gene. Primer  Sequence  MT04  a  5'-CCGCGCGGGGG^TCCTGC^GTTACTGCCCGTAATATGCC-3'  MT05  b  5 -GGCCC<^TGGCCC4T^TGTTAGAAGATCTCAAACGCCAGG-3' ,  MT04 has overlapping BamH I and PstI restriction sites (5' to 3', respectively, in italics), ' and the 19 bases at its 3'-end are complementary to the 3'-end of the araD gene. * MT05 has Nco I and Nde I restriction sites, (5' to 3', respectively, in italics), and the 25 bases at its 3'-end are complementary to the 5'-end of araD.  a  Both the vector (pBSTTM) and the P C R product (from six pooled reactions) were digested with Nde I and Pst I. Additionally, the vector was treated with alkaline phosphatase. These samples were isolated by electrophoresis on a 1% low melt SeaPlaque agarose gel (FMC BioProducts) and the appropriate bands (as visualized with ethidium bromide and ultraviolet light) were excised. A n in-gel ligation was then performed using T4 D N A ligase for 30 h at 1 5 ° C . The ligation mixture was melted and immediately transformed into  E. coli  XLl-Blue by electroporation. Colonies were picked from L B agar plates containing 100 u g / m L ampicillin and used to inoculate 10 m L overnight cultures of L B containing 200 |ag/mL ampicillin. Plasmid p R E l was obtained from an overnight culture using the Magic Minipreps D N A Purification System (Promega). This work was performed by Dr. Martin Tanner in the laboratory of Dr. Jeremy Knowles (Harvard, MA).  Chapter II: Cloning, Overexpression and Characterisation Attempts to Subclone the araD gene of Escherichia The araD gene of E.  coli  coli was amplified by PCR from plasmid pHC5 (Chen etal, 1990)  kindly donated by Dr. R.E. Moses, and with the sequence-specific primers MT04 and MT05 used previously. The araD gene was amplified in a M J Research MiniCycler. The reaction mixture (100 uL) contained: 0.1 u M each dNTP, 0.5 uM each primer, 0.1 ug pHC5,1.5 m M M g C l , 50 m M KC1, 20 m M Tris-HCl (pH 8.4), and 2.5 U Taq D N A polymerase 2  (Gibco/BRL).  100 \xL of mineral oil was added to prevent evaporation during  thermocycling. A n initial melt at 9 4 ° C (5 minutes), annealing at 50 ° C and chain extension at 7 2 ° C (5 minutes) was followed by 28 cycles of 9 4 ° C (1 min), 5 0 ° C (2 min) and 7 2 ° C (2 min).  The final cycle was 9 5 ° C (1 min), 5 0 ° C (2 min) and 7 2 ° C (10 min). The resultant  P C R product was prepared for insertion into an overexpression vector by digestion with Nde I and Pst I and purified from a 1.2% agarose gel using a Nucleotrap Extraction Kit for Nucleic Acids (Macherey-Nagel). The vector, p R E l , was digested with Pst I and Nde I and either treated with calf alkaline phosphatase (Promega) or purified from a 1.2% agarose gel with the Nucleotrap Extraction Kit for Nucleic Acids. The amplified, restricted D N A fragment was ligated (16 h at 15°C) into the prepared vector and transformed into E.  coli D H 5 a . Resultant colonies  were checked for plasmid D N A of appropriate size and for protein overexpression. At the time we had no other overexpression vectors into which to insert the araD gene.  56  Chapter II: Cloning, Overexpression and Characterisation  Calculation and Determination ofSubunit Molecular Mass The subunit molecular mass was calculated to be 25520 D a based on the amino acid composition.  Verification was provided by electrospray ionization mass spectrometry  performed by Dr. Shichang Miao and David Chow, which determined the subunit molecular mass to be 25522 ± 4 Da. Electrospray ionization experiments used an H P L C - E S M S setup consisting of a microbore H P L C (Michrom U M A ) connected in-line to either a P E - S C I E X APIII or API 300 MS as described by Hess etal. (1993). Intact epimerase samples (10 uL, > lug/uL) were injected into a microbore PLRP column (1 x 50 mm) and eluted with a linear gradient of 16 to 80% acetonitrile in water (containing 0.05% trifluoroacetic acid) over three minutes and maintained at 80% acetonitrile for an additional seven minutes. The eluant was introduced into the spectrometer, which was operated in the single quadrupole mode. Protein molecular weights were determined from this data using deconvolution software supplied by Sciex.  DNA sequencing The  araD gene of p R E l was sequenced in entirety using the Sanger dideoxynucleotide  chain termination method and Sequenase version 2.0 D N A Sequencing Kit (United States Biochemical) to ensure that no errors had occurred during P C R amplification of the gene. Double-stranded plasmid D N A was isolated for this purpose using a Wizard Minipreps D N A Purification System (Promega). Primers used for sequencing included MT04 and MT05 as well as MT12, MT13, MT14 and MT15 (see Table 2.5).  57  Chapter II: Cloning, Overexpression and Characterisation  58  Table 2.5. Primers Used in Sequencing the araD Gene. Primer  Sequence  MT12  a  5'-GGCGTCGATTACAGCGTCATGA-3'  MT13  h  5'-GGGCGCAGGCGGGTCAGTCGAT-3'  MT14  c  5'-GGCGTTCTGGTCCATTCCCACGGCCCGT-3'  MT15  d  5'-CGTACCTTCAACCACTTCACCGGT-3'  " Identical to the sense strand of araD between nucleotides 133 and 154. Identical to the sense strand of araD between nucleotides 311 and 332. Identical to the sense strand of araD between nucleotides 493 and 520. Complementary to the sense strand of araD between nucleotides 210 and 187. h  c  d  Enzyme PuriGcation The procedure for purification of recombinant L-ribulose-5-phosphate 4-epimerase was achieved using a modification of the method described by Lee etal. (Gielow & Lee, 1975;  Lee etal, 1968). For purification of L-ribulose-5-phosphate 4-epimerase, p R E l was transformed into  E. coli Y1090 (modified), an araD mutant. L B broth (500 mL) supplemented with 50 ug/rnL ampicillin in a 1 L flask was inoculated with a colony from the transformation and incubated overnight at 37 °C with agitation. Cells were harvested by centrifugation and resuspended in 10 m M potassium phosphate buffer, p H 7.6 containing 10% glycerol, 1 |j.g/mL aprotinin and 1 |^g/mL pepstatin. Cells were lysed using a French Press at 20000 psi. Cell debris was removed by a 20 minute centrifugation at 5000 rpm in a G S A rotor (Sorvall) followed by  Chapter II: Cloning Overexpression and Characterisation ulttacentrifugation to remove insolube cell debris for 1 hour at 30000 rpm in a 50 T i rotor (Beckman). The supernatant was brought to 40% ammonium sulphate saturation at 4 ° C . The precipitate was collected by centrifugation (20 minutes at 5000 rpm in a Sorvall G S A rotor) then redissolved in fresh phosphate buffer (10 mM; p H 7.6, containing 10% glycerol, 1 pg/mL aprotinin and 1 ug/rnL pepstatin). Ammonium sulphate was removed by dialysis overnight against 2 L fresh buffer. The protein solution was diluted from 5 m L to approximately 50 m L with 10 m M potassium phosphate buffer (pH 7.6, conaining 10% glycerol, 1 u,g/mL aprotinin and 1 ug/mL pepstatin) and loaded onto a 20 m L DE-52 (Whatman) column, washed with 50 m L 10 m M potassium phosphate buffer (pH 7.6, containing 10% glycerol, 1 ug/mL aprotinin and 1 ug/mL pepstatin), then eluted with 25 m L 0.4 M NaCI in 10 m M phosphate buffer, p H 7.6 (containing 10% glycerol, 1 ug/mL aprotinin and 1 p.g/mL pepstatin). The eluent was concentrated using a Millipore Ultrafree-15 centrifugal filter device with a Biomax-IOK N M W L membrane. Excess NaCI was also removed by washing the protein several times with salt-free buffer. The enzyme solution was frozen in liquid nitrogen and stored at -76 ° C overnight. The protein sample was filtered through a 0.2 um syringe filter and about 30 mg were loaded onto a Waters Q-Pak (Q 8HR) H P L C column. The column was eluted with a hnear gradient of 0 to 0.4 M NaCI in 10 m M potassium phosphate, p H 7.6 containing 10% glycerol. Flash-frozen samples of purified epimerase were stored in aliquots at -76 °C.  59  Chapter II: Cloning, Overexpression and Characterisation  60  Purity was checked by 12% S D S - P A G E on a Bio-Rad M i n i - P R O T E A N II electrophoresis system. Protein bands were visualized by staining with Coomassie Brilliant Blue.  Protein Determination The extinction coefficient of L-ribulose-5-phosphate 4-epimerase was determined according to the method described by Gill and von Hippel (1989). Various amounts of a protein solution were diluted to 500 uL and mixed with an equal volume of either 6 M guanidinium-HCl, p H 7.7 or potassium phosphate buffer, p H 7.7, and absorbances of these diluted proteins were measured at 280 nm. Using the relationship described in equation 2.1, e (the extinction coefficient of the native protein) was calculated. i \ , (the absorbance at 280 n  nm of the native protein) is the measurement in the absence of guanichnium-HCl, A (the d  absorbance at 280 nm of the denatured protein) is the measurement in the presence of guanidinium-HCl, and 8 (the extinction coefficient of the denatured protein) calculated from d  the number of Trp, Tyr residues and disulphide bonds in the protein as described by Edelhoch (1967): 4 Trp + 9 Tyr + 0 disulphide bonds = 4(5690 cm" M" ) + 9(1280 c m 1  1  1  M" ) = 34280 cm" M" . A (from the guanidinium-HCl containing solutions) and A were 1  1  1  d  n  measured for each protein concentration. The extinction coefficient determined in this manner, and averaged from three such determinations was £ = 34314 cm" M" per subunit 1  1  n  (1.76 m L mg" cm" ), but was variable, probably due to incomplete denaturation of the 1  protein.  1  Chapter II: Cloning, Overexpression and Characterisation  61  The extinction coefficient at 280 nm was also determined by measuring the absorbance at 280 nm and by analysis of the total amino acid composition. Amino acid analysis on a sample of the purified epimerase (in H 0 ) to which norleucine (0.1 mM) had 2  been added as an internal standard was performed by Suzanne Perry in the Nucleic Acid and Protein Service at the University of British Columbia. The amount of alanine in the sample was used to calculate the concentration of the protein in the solution of known absorbance. Then, the extinction coefficient was determined using the relationship A = eel. This method gave 8 = 33717 M" cm" per subunit, or 1.73 m L mg" cm" , which is in agreement with the 1  1  1  1  n  extinction coefficient  calculated above.  Calculations from the  spectrophotometric  measurements were performed using the extinction coefficient calculated from amino acid analysis of the protein (1.73 m L mg" cm" ). 1  1  Determination of Metal Ion Content The recombinant L-ribulose-5-phosphate 4-epimerase was analysed for metal ion content by Bert Mueller (Dept. of Earth and Ocean Sciences, University of British Columbia) using ICPMS. The enzyme sample was centrifuged briefly (10 minutes) in a Centricon-10 filter unit. The concentration of the resultant enzyme sample was determined. The concentrated enzyme and the filtrate from centrifugation were diluted 1/400 in 10% nitric acid (ultrapure (twice distilled on quartz) from Seastar, Sidney, B.C.). Sc was added to 20 ppb as an internal standard. The samples were digested at 50°C for 12 h prior to analysis. The zinc content of the filtrate was subtracted from that in the enzyme sample.  Chapter II: Cloning, Overexpression and Characterisation  62  Preparation ofMetal-Free Buffers and Glassware Metal-free buffer was prepared by passing a solution of 50 m M H E P E S (2 L , containing 10% glycerol) through a column of Chelex-100 resin (70 m L N a form, 50-100 +  mesh, changed to H form by washing with 2 volumes 1 N HC1 and rinsing with 5 volumes +  deionized H 0 ) . The p H was adjusted to 7.6 using solid Tris base. All the glassware and 2  plasticware used in this procedure was soaked overnight in 4 N HC1 and washed thoroughly with deionized water (17.8 M Q / c m at 25°C).  Preparation ofZn-Substituted Enzyme and Analysis ofZinc Content Apoenzyme was prepared using a modification of the method described by Deupree and Wood (1972): protein (5 mg) was incubated with 20 m M E D T A (2.5 mL) for 3 h at 25°C. The enzyme was then dialysed against metal-free H E P E S / T r i s buffer (50 m M , p H 7.6, 2 x 500 mL, containing 10% glycerol) at 4°C. A stopped assay for LRu5P was employed to ensure that the enzyme was inactive (less than 5% remaining activity; see the assay for apoenzyme activity). The apoenzyme was reconstituted by the addition of 99.99% pure Z n C l  2  (10  equivalents, 250 uL of a 3.9 m M solution) and incubated for 2 h at 25°C. The reconstituted enzyme was dialysed against metal-free H E P E S / T r i s buffer (50 m M , p H 7.6, 2 x 500 mL) containing 10% glycerol at 4 ° C and flash frozen in liquid nitrogen. Samples to be analysed by ICPMS were passed through a size exclusion column (Waters Protein Pak 125), eluted with the same H E P E S / T r i s buffer, and then diluted 1/400  Chapter II: Cloning, Overexpression and Characterisation in 10% nitric acid (ultrapure (twice distilled on quartz) from Seastar, Sidney, B.C.). The column had previously been treated with 200 m M E D T A and then equilibrated with the H E P E S / T r i s buffer. The protein concentration of the fraction' containing the epimerase was determined. Sc was added to 20 ppb as an internal standard. The samples were digested at 50°C for 12 h prior to analysis. These samples were analysed on a V G Elemental Plasma Quad mass spectrometer that had been calibrated with a standard curve of Z n (0, 5, 15, 40 ppb, prepared in 10% ultrapure nitric acid containing a 1/400 dilution of the elution buffer and 20 ppb Sc as an internal standard). The analyses were performed by Bert Mueller in the Department of Oceanography at the University of British Columbia.  Measurement of Enzyme Activity Enzyme activity was measured according to the method of Davis, Lee and Glaser (1972). In this enzyme assay each cuvette contained, in a total volume of 1.00 mL, 25 m M glycylglycine (pH 7.6), 5 m M ribose-5-phosphate, 0.1 m M TPP, 0.15 m M N A D H , LRu5P between 0.02 and 1 m M , 5 U a G D H , 50 U T I M , 0.25 U transketolase (Sigma T-6133 from Bakers yeast, dissolved in water).  A n a G D H / T L M mixture could be purchased in the  correct ratio (Sigma G-1881 from rabbit muscle), otherwise an equivalent mixture could be prepared from a G D H (Sigma G-6751 from rabbit muscle) and T I M (Sigma T-2391 from rabbit muscle).  In either case, these coupling enzymes were in a suspension containing  ammonium sulphate and E D T A .  Enough of this mixture for ten reactions (24 uL) was  diluted into water (2 mL) and centrifuged in a Millipore 10K N M W L filter unit for 20  63  Chapter II: Cloning Overexpression and Characterisation  64  minutes prior to use. The resulting enzyme solution (normally 50 uL) was diluted to 500 uL in water. Each cuvette contained 50 uL of the prepared a G D H / T I M mixture. The cuvettes were equilibrated at 3 7 ° C and the background rate was monitored for 5 minutes prior to initiation of epimerisation by addition of 0.05 ug of purified L-ribulose-5-phosphate 4epimerase. The reaction was followed by the decrease in absorbance at 340 nm on a Cary 3E spectrophotometer.  The rate of formation of DXu5P was calculated from the initial  slopes using an extinction coefficient of N A D H of 6220 M" c m . The background rates 1  1  were calculated in the same manner from data collected prior to addition of the epimerase. The background rates were subtracted from the rates after addition of epimerase to obtain the rate due to epimerase action. Kinetic parameters were determined from a direct fit of the data to an enzyme kinetics equation using the computer program GraFit (Erithacus Software, U K ) .  GAPDH Assay for Enzyme Activity A n alternate coupled assay for L-ribulose-5-phosphate 4-epimerase activity is described by Wolin, Simpson and Wood (1958). Each cuvette in this assay contained (in a total volume of 1.0 mL): 20 m M glycylglycine (pH 7.4), 0.15 m M N A D , 3.4 m M sodium +  arsenate (pH 7.5), 1 m M D T T , 0.1 m M TPP, 150 ug crystalline G-3-P dehydrogenase (Sigma G-2267, dissolved in water), 0.50 U transketolase (Sigma T-6133 from Bakers yeast, dissolved in water), 1 m M MgCl , 5 m M DR5P, and LRu5P between 0.02 and 1 mM. After 2  the cuvette had reached 37°C, the absorbance was monitored for 5 minutes immediately  Chapter II: Cloning, Overexpression and Characterisation  65  prior to initiation by addition of 0.05 ug of the purified 4-epimerase. Rates were determined in exactly the same manner as with the assay described above. The background rates tended to be extremely high.  Assay for Apoenzyme Activity Apoenzyme at 0.01 m g / m L was incubated at 3 7 ° C for 5 minutes in metal-free H E P E S / T r i s buffer, p H 7.6, containing 9.44 m M LRu5P.  Controls were water and  untreated (active) enzyme. Aliquots (10 uL) were immediately added to pre-equilibrated cuvettes at 37 ° C containing (in 990 uL) 25 umoles glycylglycine (pH 7.6), 0.15 umole N A D H , 5 umoles DR5P, 0.1 umole TPP, 5 U a G D H , 50 U T I M and 0.25 U transketolase. The coupling enzymes were prepared as described for measurement of enzyme activity. The cuvettes were monitored at 340 nm, and the drop in absorbance for the first minute immediately after addition of the enzyme-containing aliquots was measured. The average drop in absorbance per minute before addition of the aliquot was determined and was subtracted from the drop measured after addition of the aliquot. The amount of activity observed in the apoenzyme assayed in this manner was typically about 5% of that observed for the untreated enzyme.  Chapter II: Cloning, Overexpression and Characterisation  66  Measurement of Equilibrium The Z n  2 +  form of the epimerase (0.5 mg) was incubated at 3 7 ° C with 5 mg of the  sodium salt of D-xylulose-5-phosphate (Sigma) at p H 7.6 in 1 m L for 4 hours. A t this time the enzyme was removed from the equilibrium mixture of the sugars by passage through a Millipore Ultrafree-4 centrifugal filter device with a Biomax-IOK N M W L membrane . The nitrate was lyopHlized, exchanged with D 0 twice and finally dissolved in 500 uL of D 0. 2  This mixture was analysed by quantitative 500 M H z  2  'H-NMR at 2 5 ° C .  Integration of the  peaks arising from the C-4 protons of each epimer provided the relative concentration of each epimer.  Chapter III  Nature of the Metal Ion Ligands of L-Ribulose-5-Phosphate 4-Epimerase  Introduction  The  Escherichia coli L-ribulose-5-phosphate 4-epimerase (Deupree & Wood, 1972) and  L-fuculose-l-phosphate aldolase are metalloenzymes with a homologous N-terminal domain (Dreyer & Schulz, 1996b). Three of the four zinc ligands of the aldolase lie within this domain and are reported to be conserved in the epimerase (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996b). A greater understanding about how L-ribulose-5-phosphate 4-epimerase operates can be gained by characterizing its metal binding site both by metal ion substitution and by site directed mutagenesis of the putative metal binding ligands. These ligands can be identified by mutation to residues known to be poorer metal ion ligands. These changes would result in the subsequent loss of the ability to bind metal ions tightly and reduction in  67  Chapter III: Metal Ion Ugands  68  catalytic activity. Identification of the epimerase's metal binding ligands as those predicted from homology to the aldolase would serve to establish a structural link between these two enzymes.  Z i n c Ions i n E n z y m e Catalysis It has long been established that many enzymes make use of metal ions in promoting catalysis as well as in forming structural motifs (Mildvan, 1972; Ochiai, 1987). Zinc, which is the second most abundant metal in biological systems, is an important catalytic component in each of the six enzyme classes and across all phyla (Lipscomb & Strater, 1996; Vallee & Auld, 1990a; Vallee & Auld, 1990b). Zinc is thought to be preferred over other divalent cations for a number of reasons.  With a flexible coordination geometry, zinc can have  anywhere between four and six ligands.  Zinc can undergo fast ligand association and  dissociation, which also allows the substrates and products to leave the active site. Zinc can act as a Lewis acid, has intermediate polarizability, is readily available, and can be bound tightly into suitable sites on proteins. Furthermore, zinc does not readily participate in redox reactions (Lipscomb & Strater, 1996; Maret & Vallee, 1993). A role for catalytic Z n  2 +  ions in enzymes is to act as electrophilic catalysts. The Z n  2 +  ion acts as a Lewis acid (electrophile), which can coordinate to carbonyl groups. This catalyst stabilizes the developing negative charge during the course of a reaction. Coordination of a carbonyl group to the metal ion polarizes the substrate towards nucleophilic attack (by another group in the enzyme's active site) or enolization.  Chapter III: Metal Ion Ugands  69  A comparison of twelve zinc enzymes whose crystal structures are known has uncovered some common features of the ligands used for binding the catalytically active zinc ion (Vallee & Auld, 1990a; Vallee & Auld, 1990b).  In each case, the zinc ions are  coordinated by three amino acid residues in the enzyme's active site and by an activated water molecule. In contrast, structural zinc ions are almost universally coordinated to four cysteine residues in a tetrahedral arrangement. Zinc forms complexes with nitrogen and oxygen ligands as readily as it does with sulphur ligands. In the catalytic zinc sites, the imidazole group of histidine, the sulfhydryl group of cysteine, and the carboxyl groups of glutamic acid and aspartic acid are all possible ligands for zinc. Histidine is, however, the most predominant ligand in catalytic zinc sites, followed by glutamate, aspartate and cysteine. A combination of three of these residues is used to bind zinc tightly in the active site and leave an open coordination sphere. In the free form of the enzyme an "activated" water molecule fills and completes the coordination sphere of the zinc ion. The water molecule can be ionized, polarized, or displaced upon substrate binding. The first two ligands in catalytic zinc sites are separated by a "short spacer" consisting of one to three amino acid residues. These ligands are separated from the third ligand by a "long spacer" of between about 20 and 120 amino acid residues. The short spacer enables the formation of a primary bidentate zinc complex, whereas the long spacer permits flexibility of coordination number (this can change during the course of a reaction) and geometry.  A long spacer also allows for any conformational changes which may occur  Chapter III: Metal Ion Ugands  70  during the course of the catalytic reaction, and might be part of the substtate-binding pocket. Structural zinc binding sites have much shorter spacers, which imparts rigidity to the metal centres and is consistent with their role in stabilizing the overall structure of the protein and local conformation (Vallee & Auld, 1990a; Vallee & Auld, 1990b). Tetradentate  zinc-binding sites are inaccessible  to  solvent.  For example,  procollagenase has tetradentate zinc coordination, but upon conversion to collagenase loses the fourth amino acid ligand, which is replaced with a water molecule (Vallee & Auld, 1990b). During this process, a structural zinc ion becomes catalytic.  C h a r a c t e r i z a t i o n o f Z i n c B i n d i n g Sites  Site directed mutagenesis can be used to alter any amino acid residue of a cloned and overexpressed protein to any other amino acid and can lead to the identification of putative metal-binding Ugands. Targets for mutagenesis can be either amino acid residues conserved over a number of related structures or all potential zinc-binding Ugands (histidine, glutamate, aspartate or cysteine residues).  Recent examples of these kinds of mutations are the  identification of histidine Ugands in the binuclear metal-ion site of phosphotriesterase (Kuo & Raushel, 1994), identification of histidine zinc Ugands of Class II fructose-bis-phosphate aldolase (Berry & MarshaU, 1993), confirmation of the zinc-binding site residues of a metaUo-P-lactamase (Crowder etal., 1996), and the identification of potential zinc binding Ugands in VanX (McCafferty etal, 1997). In addition to Ugand identification, site-directed  Chapter III: Metal Ion Ugands  71  mutagenesis can be used to probe the structural and functional roles of known metal ion  Ugands (eg., Alexander etal, 1993; Kuo etal, 1997). While histidine, glutamate and aspartate residues are often changed to alanine or valine (and cysteine to serine), mutation of histidine or aspartate to asparagine is another viable alternative (Kuo & Raushel, 1994). This is a more conservative mutation in that there is a possibility that the new amino acid residue may be able to make the same hydrogen bonds that were made by the histidine residue it replaces without itself acting as an effective metal ion ligand. Thus, while catalytic activity and/or the ability to bind the metal ion may be affected, the structural effects should be less drastic than with other mutations. This approach has been used to identify ligands of a binuclear metal centre in phosphotriesterase (Kuo & Raushel, 1994) and the iron ligands of tyrosine hydroxylase (Ramsey etal, 1995), and will also be used in this study. As long as the mutant enzyme has similar physical characteristics to the wild type enzyme, differences in activity (/£ ), substrate binding (as approximated by cat  or metal  binding ability may be attributed to the amino acid residue that was changed. A number of other techniques can also provide information about the zinc-binding site, either in combination with studies on mutant enzymes, or alone. techniques are spectroscopic and require that the Z n  2 +  Many of these  ion, which is spectroscopically silent,  be removed and replaced with other divalent metal ions. These replacements, which can include C o , C u , N i , M n , or C d , often yield active, sometimes even hyperactive 2+  2 +  2 +  2 +  2 +  Chapter III: Metal Ion Ligands enzymes.  72  O f these divalent metal ions, C o  2 +  is most likely to replace Z n  2 +  efficientiy  (Lipscomb & Strater, 1996; Maret & Vallee, 1993). UV-visible spectra of cobalt(II)-substituted zinc enzymes can provide information on the environment in which the catalytic zinc ion lies. If the coordination geometry of the enzyme's metal-binding site is simple, then it may be possible to determine this geometry from visible spectra, based on the extinction coefficients of absorption features between 500 and 600 nm (Elgren et al, 1994).  For most enzymes this is not normally possible; the  coordination geometry of the metal ion in the binding site is complex or unsymmetrical. UV-visible spectra can be used to determine the presence and number of thiolate ligands (from cysteine) based on intense bands in the near-ultraviolet. In addition, the absorption maxima of the enzyme-complexed cobalt(II) ion can indicate the type of ligands: in the order of sulphur, nitrogen, and oxygen protein donor ligands, the absorption maxima are progressively shifted to higher energies (shorter wavelengths). In practice this does not allow for distinction between oxygen and nitrogen donor atoms (Maret & Vallee, 1993). Sulphur donor ligands are also known to increase the molar absorptivities of some Co(II) complexes. If thiolate ligands are mutated to nitrogen or oxygen ligands, the spectra of the resulting Co(II) complex may not be readily observable, as was observed with a cysteine to serine mutant of a metallo-p-lactamase (Crowder etal., 1996). N M R spectroscopy of proteins with a paramagnetic metal centre is a powerful technique for the detection of ligand binding, conformational changes of the protein induced by ligands, or ionizations near the metal site (Bertini etal, 1993; Maret & Vallee, 1993). The  Chapter III: Metal Ion Ligands  73  somewhat short electron relaxation time of the cobalt(II) ion leads to relatively well-resolved isotropically shifted resonances in the 'H-NMR spectra of metal-coordinating ligands. Proton-NMR of Co -substituted enzymes can be used to ascertain if the metal ion has any 2+  histidine ligands, and often the number of these ligands. The imidazole-NH proton of the histidine metal ligands of Co -substituted metallo-P-lactamase from Bacteroidesfragilis appear 2+  downfield between 40 and 60 ppm (Figure 3.1), and as there are three of these resonances, three histidine residues are implicated as metal ion ligands (Crowder etal, 1996).  1  I  I  |  I  I  I  so  90  Figure  I  |  I  I  I  |  *o  70  I  I  I  |  I  I  m  I  |  m  I  I  I  |  I  I  at  3.1. 'H-NMR of the Co -substituted B. fragilis metallo-P-lactamase (1 mM) at pH 2+  7.4. The three solvent-exchangeable histidinyl NH protons are marked with asterisks. Taken from Crowder etal., 1996.  In addition, if the enzyme is substituted with C d , the chemical shift of the metal m  2+  ion (by Cd-NMR) can indicate the number, type, geometry and dissociation constants of 113  the metal ion's ligands. Cd-NMR can therefore be used to provide insights into proteinm  Chapter III: Metal Ion Ugands  74  substrate interactions, conformational changes, and metal displacement reactions (Kanaori  etal, 1996; Summers, 1988). Oxidation of Co -substituted enzymes to Co -substituted enzymes can be used to 2+  3+  investigate the role which the metal ion plays in catalysis (Maret & Vallee, 1993).  Co 3 +  substituted enzymes will become inactive if substrate ligation to the metal ion is important to catalysis; C o  3 +  is "substitution-inert". That is, ligands cannot dissociate from C o  3 +  as  easily as they might from C o . 2+  The ligand exchange-inert properties of C o After  3 +  may be used in an additional manner.  in situ oxidation of a Co -substituted enzyme to a Co -substituted enzyme, limited 2+  3+  proteolysis and subsequent purification of the cobalt-bearing peptide can lead to identification of the metal ion ligands (Maret & Vallee, 1993). A recent study using a Co -substituted enzyme is that of Hlavaty and Nowak 3+  (Hlavaty & Nowak, 1997), who investigated the nature of the phosphoenolpyruvate carboxykinase active site. The Co -substituted enzyme was found to retain 15 to 25% of 3+  the activity of the Co -substituted enzyme, consistent with ligand exchange not being 2+  essential to catalysis in this enzyme. Subsequent proteolytic degradation and purification of the cobalt-containing peptide led to the identification of this peptide, which had two aspartic acid residues as the only feasible metal ion ligands.  Chapter III: Metal Ion Ligands  75  T h e Z i n c - B i n d i n g Site o f L - F u c u l o s e - l - P h o s p h a t e A l d o l a s e  Recent X-ray crystal structures of L-fuculose-1 -phosphate aldolase in both its normal Z n -form and the Co -substituted form have shown unambiguously that this enzyme binds 2 +  2+  its catalytic zinc ion with four ligands: E73, H92, H94 and H I 55 (Figure 3.2) (Dreyer & Schulz, 1993; Dreyer 8c Schulz, 1996b). The spacing of these ligands follows the pattern established by Vallee & Auld (1990a, 1990b): there is a short spacer of one amino acid residue between H92 and H94, and the spacing between these two Ugands and E73 is 20 residues. The spacing between these two histidine residues and H I 55 is 61 residues, which is less than the maximum long spacer length of 120 amino acid residues.  Further  crystaUographic studies on this enzyme and the inhibitor phosphoglycolohydroxamate (which mimics the intermediates in the aldolase's reaction mechanism) have provided evidence that the E73 zinc Ugand rotates out of the Ugand sphere upon binding of the substrate or inhibitor (Figure 3.3) and participates in catalysis as an active site base (Dreyer & Schulz, 1996a; Fessner et ai, 1996). While this enzyme is unique in that it binds the catalytic zinc ion with four residues, the glutamate Ugand appears to be acting Uke a water molecule in that it becomes displaced from the Ugand sphere upon substrate binding. Note, however, that glutamate is charged and bidentate, while water is neutral (and monodentate).  Chapter III: Metal Ion Ugands  76  Figure 3.2. Coordination of Z n in the Active Site of L-Fuculose-l-Phosphate Aldolase (PDB code: 1FUA). The catalytically active zinc ion is tightly coordinated by three N atoms of histidines 92, 94 and 155, and a bidentate contact from glutamic acid residue 73, resulting in a distorted tetrahedral coordination sphere. 2 +  €  Chapter III: Metal Ion Ugands  11  Glu73-  His94  Rotation of E73 of L-Fuculose-1-Phosphate Aldolase Away from Z n Binding of Phosphoglycolohydroxamate. F i g u r e 3.3.  D e s i g n of M u t a n t L-Ribulose-5-Phosphate  2+  on  4-Epimerases  In order to probe the metal binding site of L-ribulose-5-phosphate 4-epimerase it was decided to mutate each of the putative metal ion Ugands to an asparagine residue. The N terminal domain of L-ribulose-5-phosphate 4-epimerase was reported to have significant homology to the N-terminal domain of L-fuculose-1-phosphate aldolase (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996b). Three of the metal-binding residues of the aldolase (E73, H92 and H94), which Ue in the homologous domain, were conserved in the amino acid sequence of the 4-epimerase (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996b), and therefore the corresponding residues (D76, H95, and H97) were postulated to be the epimerase's metal ion Ugands. An aspartate residue has replaced the glutamate residue in the  Chapter III: Metal Ion Ugands aldolase.  78  Subsequendy, amino acid sequence alignments using a different alignment  algorithm to that used by Dreyer & Schulz has implicated HI 71 in the 4-epimerase as the fourth ligand (see the experimental and results sections later in this chapter). A structural link between the epimerase and the aldolase suggests that there may also be a mechanistic link. Therefore, identification of the metal ion ligands of the epimerase predicted from homology to the aldolase implies that the epimerase follows a mechanism similar to the aldolase. A first step in this study involves the design and construction of asparagine mutants of each of the putative metal ion ligands. Asparagine (Figure 3.4) was chosen for the same reason that Kuo and Raushel (1994) chose it: asparagine can participate in some of the hydrogen bonds in which histidine participates and is smaller in size, yet it is not an efficient metal ion ligand.  1  N ^ N H  OH  NH  2  Asparagine was also chosen as a  histidine aspartic acid (His) (Asp) (H) (D) Asparagine has a similar shape and F i g u r e 3.4. Amino Acid Residues. substitute for the aspartate ligand.  asparagine (Asn) (N)  size to aspartate, but it is neutral whereas aspartate is negatively charged. Therefore, the mutant L-ribulose-5-phosphate 4epimerases should better retain the active site structure and the overallfoldingof the protein should be relatively unaffected.  Chapter III: Metal Ion Ugands  79  Site-directed mutagenesis of histidine residues to asparagine involves only a modification of one nucleotide in the D N A sequence: histidine residues are encoded by CAT or CAC, while asparagine residues are encoded by A A T or AAC. Similarly, the change from an aspartic acid (encoded by G A T or GAC) to an asparagine residue is also a single base change in the nucleotide sequence. However, for easy screening of mutant plasmids, it is useful to make more changes and to introduce silent mutations which allow restriction enzymes to cut the mutant enzymes' plasmid D N A but not the plasmid encoding the wild type epimerase. Alternatively, if a unique restriction site exists near the site of the sitedirected mutagenesis, this site may be removed by a silent mutation. A number of different methods of site-directed mutagenesis are currently available. One with the greatestfidelityis the Kunkel method, and requires that the plasmid have an origin of replication recognized by the filamentous phage F l . If this site exists on the plasmid, which is then also called a phagemid, then one of the strands of the plasmid may be encapsulated within the protein coat of a helper phage instead of the phage's own DNA. Growth of this plasmid in a dut~ ung' strain of E. coli will produce vector with the occasional uracil in place of thymine in the newly synthesized DNA. This strain of E. coli is deficient in dUTPase (dut) and uracil-N-glycosylase (ung). Coinfection with a helper phage is required to isolate the single stranded phagemid. A mutagenic primer (a synthetic single stranded oligodeoxynucleotide containing the desired base changes) is annealed to a single-stranded form of the plasmid isolated from the helper phage. The complementary strand is then synthesized with D N A polymerase (either  Chapter III: Metal Ion LJgands  80  T4 or T7) and deoxynucleotides using the single-stranded plasmid as the template. Ligase is used to close the newly formed strand of D N A to the 5'-end of the mutagenic primer. When this double-stranded D N A plasmid is transformed into a host strain of E.  coli with an  active uracil-IV-glycosylase, the uracil-containing template strand is destroyed and the new strand containing the mutation becomes the template for production of the mutant plasmid. This mutated plasmid can then be replicated using normal cellular replication machinery (Kunkel, 1992; Zhou etal, 1990). Removing the  araD gene from p R E l (the overexpression plasmid containing the araD  gene encoding L-ribulose-5-phosphate 4-epimerase, discussed in Chapter II) and cloning new genes into the "empty" plasmid proved to be extremely difficult. Although we tried to solve this problem, it was unclear where the problem lay. Excision of the  araD gene, ligation of  new genetic material and restriction of this D N A in preparation for ligation are all possible sources of difficulty. Therefore, mutagenic methods involving recloning the mutated gene into an overexpression vector were not attractive. When the Kunkel method (as described above) cannot be used, other methods (almost all PCR-based) are available. A mutagenesis strategy involving PCR-amplification of the entire plasmid was designed for this study. One method which does not require recloning the gene to be mutated, is the "recombinant circle PCR" method of Jones and Winistorfer (1992). In this method (Figure 3.5), the plasmid is amplified in two separate reactions as two "halves". Each half has an end which is homologous to about 200 bases at one end of the other half. The second end covers the mutated site and also has about 20 bases homologous to the other P C R product.  Chapter III: Metal Ion Ugands  81  Figure 3.5. The RCPCR Method of Site-Directed Mutagenesis. Small triangles indicate the site of mutagenesis. When the products of these reactions are combined, denatured and reannealed, they can form circular D N A products, which after transformation, are extended and ligated by the host cell. Since the PCR products contained the sequences which are required for plasmids to be replicated and retained in the host cell, the plasmids isolated from the resulting bacterial cultures should contain the desired mutation and other features of the plasmid template. In order to avoid large contamination by the template plasmid, it is cut at a unique restriction site in a region which will not be amplified in the original PCR. It may be possible to avoid contamination by using a plasmid template which was isolated from a duf ung strain  Chapter III: Metal Ion Ugands  82  of E. coli and which therefore contained uracil, as is used in the Kunkel method. Upon transformation of the RCPCR product into a host strain of E. coli which had functional dut and ung gene products, the template D N A should be degraded and replaced by the host cell. Mutagenesis methods such as that described above require the use of two mutagenic primers. It is possible, however, to carry out site-directed mutagenesis using only one mutagenic primer (Chen & Przybyla, 1994; Sarkar & Sommer, 1990). In this case, a "megaprimer" is created by PCR using the mutagenic primer and another primer. Since the D N A product of this reaction is double stranded, it can be used (purified or not) as the primer in each of two other PCR amplifications (Figure 3.6).  Figure  3.6.  RCPCR. mutagenesis.  Creation of a Megaprimer Prior to Small triangles indicate the site of  Chapter III: Metal Ion Ugands  83  Results and Discussion  Protein Sequence Alignments Sequence alignments performed in our laboratory on the L-ribulose-5-phosphate 4epimerase and Class II L-fuculose-l-phosphate aldolase (FucA) from E.  produced results  similar to those reported by Dreyer & Schulz (1993, 1996b). Whereas Dreyer & Schulz report 34% identity in the N-terminal domains, we found that these enzymes share 38% identity and 43% similarity. In addition, alignment using the computer programme S E Q S E E (Wishart etal, 1994) showed that H171 of AraD aligned with H155 of FucA and is therefore a likely candidate as the fourth ligand to the metal ion in the epimerase (the other three are D76, H95 and H97). Furthermore, these residues were found to be conserved among all the known sequences in the A r a D / F u c A family (Figure 3.7).  Note that L-rhamnulose-1-  phosphate aldolase, which was included in the alignment, did not align well, in agreement with a report by Moralejo etal. (1993) that this aldolase has evolved convergendy to FucA. It was therefore decided to prepare mutants of the epimerase in which each of these residues was independently converted to an asparagine residue by site directed mutagenesis. These mutations should direcdy perturb the coordination sphere of the bound metal ion and affect both the enzyme's affinity for the metal and its catalytic constants.  Chapter III: Metal Ion Ugands  arad__ e c o l i arad__ s a l t y arad__haein f uca__ h a e i n f uca__ e c o l i rhad__ e c o l i Consensus  84  IVHTHSR : KKPS SDTPTHRLLYQAFPSIGG IVHTHSR : KKPSSDTPTHRLLYQAFPTIGG : KKPSSDTPTHLELYRQFPHIGG IVHTHSR : KLPSSEWQFHLSVYHTRPEANA WHNHSI : KLPSSEWRFHMAAYQSRPDANA WHNHAV : AVPTSELPAHFLSHCERIATNGKDRVIMHCHAT VH-H— : K-PSS H Y P arad_ecoli: arad_salty: arad_haein: fuca_haein: fuca_ecoli: rhad_ecoli: Consensus :  VLVHSHGPFAWG VLVHSHGPFAWG VLVHSHGPFAWG ILLAHHGLITCG TLLQHHGLIACE VLWPFHGVFGSG -L HG G  99 99 99 96 96 143  177 177 177 161 161 216  Figure 3.7. Amino Acid Sequence Alignment. The conserved metal-binding residues are shown in red. Key: arad, L-ribulose-5-phosphate 4-epimerase; fuca, L-fuculose-l-phosphate aldolase; rhad, L-rhamnulose-1 -phosphate aldolase; ecoli, Escherichia coli; salty, Salmonella typhimurium; haein, Haemophilus influenzae.  Site Directed Mutagenesis Wishing to take advantage of the Kunkel method for site-directed mutagenesis, we investigated whether pREl could be used to propagate single stranded DNA. The vector portion of pREl is largely from the pBS phagemid, which allows ssDNA to be created and packaged into a phage head. Modifications of pBSTIM (the precursor to pREl; see Chapter II) may have altered the F l origin enough to render it useless. Therefore, it was necessary  Chapter III: Metal Ion Ligands  85  to determine if a strand of p R E l could be isolated using helper phage M13K07 before planning a mutagenesis strategy dependent on ssDNA. ssDNA was isolated from  E. coli  DH5ocF transformed with p R E l and infected with helper phage M13K07 and visualized on an ethidium bromide stained 1.2% agarose gel. Sequencing reactions (in either direction) did not yield readable sequence.  We were therefore unsure if a strand of p R E l had been  packaged into the phage since the amount of ssDNA from p R E l could have been too small to give readable sequence. Therefore, it was decided to use other means of site-directed mutagenesis. Site-directed mutagenesis was performed using the R C P C R technique on p R E l isolated from transformed E.  coli D H 5 a F ' , and the mutant plasmids p A J l , pAJ2 and pAJ3  were constructed, encoding the H95N, H 9 7 N and D 7 6 N epimerases, respectively. Approximately eight colonies per transformation were checked for the presence of mutant plasmids, 25 % of which contained the mutation. In each case the mutant enzymes were overexpressed at levels comparable to the wild-type enzyme. Attempts at preparing the H171N  mutant were not successful.  Both halves of the plasmid appeared to have been  created by PCR, as visualized on an ethidium bromide stained 1.2% agarose gel. It is not clear why mutant plasmid could not be obtained after recombination and transformation. The plasmids containing the genes for the mutant enzymes were sequenced over the  araD  region. Only one error was found: pAJ3 had a base change at position 25 from C to G , which altered nucleotide 303 of the  araD gene. This error was not from P C R but rather  Chapter III: Metal Ion Ugands  86  from an error in designing the primer AJ3. Fortunately, this base change did not cause any change in the amino acid sequence.  Purification of the Mutant Epimerases In order to avoid contamination by endogenous L-ribulose-5-phosphate 4-epimerase, the plasmids encoding the mutant (and wild type) enzymes were transformed into a strain of E.  coli which does not have a functional 4-epimerase. Therefore, E. coli Y1090 (modified),  which was used for preparation of the wild type epimerase in Chapter II, was used. As with the wild type enzyme, the mutant enzymes were expressed to a high level, accounting for at least 60% of the total soluble protein in crude cell extracts. After purification, approximately 100 mg of enzyme was obtained per litre of bacterial culture. The molecular weights of the mutant enzymes were confirmed to be the same as those expected from their gene sequences by electrospray mass spectrometry (Table 3.1).  Table 3.1. Molecular Masses of Mutant 4-Epimerase Subunits.  a  4-Epimerase  Calculated Molecular Mass  Determined Molecular Mass  H95N  25495  25499 ± 5  H97N  25495  25498 ± 5  D76N  25520  25524 ± 5  Wild Type  25520  25522 ± 5  In Da.  Chapter III: Metal Ion Ligands Since zinc is commonly used by enzymes (such as the Class II aldolases) to promote electrophilic catalysis and since the zinc form of the wild type enzyme displayed kinetic constants which agreed with those reported in the literature, it was decided to prepare and use homogeneous samples of the zinc form of the mutant enzymes for further studies. Apoenzymes were prepared by dialysis against E D T A using conditions reported previously (Deupree & Wood, 1972), and they were subsequendy reconstituted with excess zinc. The zinc enzymes were extensively dialysed against metal-free buffer to remove excess metal.  Physical Characterization of the Mutant L-Ribulose-5-Phosphate 4-Epimerases It is important to ascertain that the mutations have not drastically altered the structure of the mutant enzymes from that of the wild type enzyme. Kinetic and metal binding results will not be valid if the mutant enzymes do not fold into the proper structure. The mutant enzymes were analysed using circular dichroism (CD) spectroscopy and were found to possess spectra virtually indistinguishable from that of the wild-type enzyme (Figure 3.8). The differences might be due to small variations in concentration. Intense double minima at 208 and 222 nm ate characteristic of a-helical structures, as is an intense maximum at 192 nm (Johnson, 1990). The intense double mimima observed in the wild type and mutant enzymes indicates that these enzymes have significant a-helical character. Unfolded protein, which can be approximated by a random coil structure, would be expected to have a small maximum at 220 nm and an intense mimimum at 198 nm. This is not seen in the C D spectra of either the wild type or mutant enzymes.  87  Chapter III: Metal Ion Ugands  88  1  1  1  1 1  1  i  1  i  20  I  0  <3>  —  WT  V -40  H95N/  H97N  -60  i  200  D 7  6N 1 220  i  1  i  240  1 260  i  280  Wavelength (nm)  3.8. C D Spectra of the Wild Type and Mutant L-Ribulose-5Phosphate 4-Epimerases. Figure  'er III: Metal Ion  89  In addition, the melting temperatures were measured by following the loss of elipticity at 220 nm as a function of temperature. Each of the enzymes was found to have a melting temperature in the range of 49.5 to 5 4 . 5 ° C (Table 3.2). These observations support the notion that the mutants were fully folded proteins with a tertiary structure and stability comparable to that of the wild-type enzyme. It should be noted that these studies were performed on Z n  2 +  reconstituted enzymes that had been dialysed to remove excess zinc.  Subsequent studies showed that apoenzyme can form upon extended dialysis of the mutant enzymes.  It would appear that the absence of zinc does not dramatically affect the C D  spectra or melting temperatures of the mutant epimerases.  T a b l e 3.2. Melting Temperature of Wild Type and Mutant 4-Epimerases." 4-Epimerase  Melting Temperature (°C)  Wild Type  49.9 ± 0.3  H95N  49.5 ± 0.3  H97N  54.4 ± 0.3  D76N  52.2 ± 0.3  a  Determined from the change in C D signal at 220 nm as a function of temperature, and corrected to a calibration curve.  To investigate the quaternary structure of the proteins, size exclusion chromatography was employed. The wild type and mutant enzymes eluted with similar retention times. A comparison to molecular weight standards indicated that they had masses of approximately  Chapter III: Metal Ion Ugands  90  127 kDa. This shows that the mutant enzymes had retained the ability to form a quaternary structure similar to that of the wild type enzyme (Figure 3.9). Given a subunit molecular mass of 25.5 kDa (as determined from the amino acid sequence), the native protein appears to be pentameric. Size exclusion chromatography is not the most accurate method for determining native masses; globular proteins are not perfect uniformly packed spheres and exhibit different mobilities based on their shapes and volumes. Ideally the standards used should have the same shape as the sample being analysed. Molecular weight calculated by sedimentation equilibrium at high speed is more accurate because it depends only on density. This technique has been used to show that the native wild type enzyme has a mass of about 105000 ± 2000 D a (Gielow & Lee, 1975), which corresponds to a tetrameric structure.  Activity of the Wild Type and Mutant Enzymes The kinetics of the zinc-reconstituted enyzmes were measured in the L-ribulose-5phosphate to D-xylulose-5-phosphate direction as described by Davis et al. (1972) and in Chapter II (for the wild type enzyme). As discussed in Chapter II, the jf\ value for LRu5P M  with the wild type enzyme correlates well with those previously published after zincreconstitution (Deupree & Wood, 1975; Gielow & Lee, 1975; Lee etal, 1968; Wolin etal, 1958), but not before this treatment. The mutant enzymes had much lower kC3t values than the wild type epimerase while their K values remained similar (Table 3.3). This might be M  expected for the alteration of residues which are not direcdy involved in substrate binding, and may be due to intrinsic changes in the active site or due to loss of the metal ion during  91  Chapter III: Metal Ion Ligands  \l  MW  5.4 5.2  —  5  —  4.8  —  4.6  —  \o  1 ^  RPE  1  1  1  o  LDH  1  i  I  1 ' —  • X BSA  cn /—\  1  PK  —  O  N.  —  \  Oval  \ n  ADH  4.4 4.2  v Cyt-C ~ 1  4 0.9  ,  1  , 1.1  1  1 1.2  1 1.3  1  l\ 1.4  i  1.5  VeA/o  Figure  3.9. Native Structure of L-Ribulose-5-Phosphate 4-Epimerases.  PK = pyruvate kinase, Aid = aldolase, LDH = lactate dehydrogenase, BSA bovine serum albumin, Oval - ovalbumin, ADH = alcohol dehydrogenase, Cyt-C = cytochrome C, RPE = L-ribulose-5-phosphate 4-epimerases (wild type, H97N, H95N, and D76N), Ve = elution volume, Vo = void volume.  ter III: Metal Ion  92  dialysis. For this reason, the mutant enzymes were assayed in the presence of excess (0.1 mM) Z n  2 +  (Table 3.4).  Table 3.3. Kinetic Parameters of the Zinc-Reconstituted Wild Type and Mutant LRibulose-5-Phosphate 4-Epimerases."' b  4-Epimerase  K  Wild Type  0.087 ± 0.007  20.4 ± 0.9  2.3 x 10  5  H95N  0.14 ± 0.01  0.047 ± 0.002  3.4 x 10  2  H97N  0.10 ± 0.01  0.048 ± 0.001  4.8 x 10  2  D76N  0.11 ± 0.01  0.073 ± 0.003  6.6 x 10  M  (mM)  4  (s )  kjK  4  at  ( M s' ) 1  M  1  2  " N o additional Z n was added to the reaction cuvettes, nor were any metal ions removed from the solutions used for enzyme assays; transketolase, a vital component of the assay system also requires divalent metal ions. See Appendix D for a graphical representation of these kinetics. 2 +  b  The kinetic constants of the wild type enzyme were relatively unaffected by. the presence of additional zinc. This is consistent with previous reports that the metal is bound tightly enough to survive dialysis against metal-free buffer. It also shows that excess zinc does not significandy activate the enzyme. The data obtained with the mutant enzymes were somewhat different: some of the lost activity could be recovered by addition of excess zinc to the cuvettes. Particularly notable are the increases in the k  Cit  values which were observed  in the case of the H 9 7 N epimerase. The K values remained relatively unaffected, however. M  It would appear that varying amounts of bound metal were lost during dialysis of the mutant enzymes against metal-free buffer resulting in mixtures of the apoenzyme and holoenzyme.  •erIII: Metal Ion Ugands  93  This is consistent with the notion that these mutations have modified a residue involved in metal-ion complexation. Measurements made in the presence of 0.2 m M Z n further changes in the k  ZM  values indicating that 0.1 m M Z n  2 +  caused no  2 +  was sufficient to activate these  mutants fully.  Table 3.4. Kinetic Parameters of the Zinc-Reconstituted Wild Type and Mutant L Ribulose-5-Phosphate 4-Epimerases In the Presence of Additional Z n . 2 +  4-Epimerase  K  Wild Type  0.060 ± 0.005  22.3 ± 0.4  3.7 x 10  5  H95N  0.096 ± 0.009  0.099 ± 0.003  1.0 x 10  3  H97N  0.14 ± 0.02  7.3 ± 0.3  5.2 x 10  D76N  0.14 ± 0.01  0.162 ± 0.005  1.2 x 10  M  k  (mM)  '  (s") 1  C3t  KJKu  a , h  (M" s") 1  1  4  3  "0.1 m M Z n was added to each cuvette. See Appendix D for a graphical representation of these kinetics. 2 +  h  In each case the k  CSlt  values obtained with the mutants in the presence of 0.1 m M Z n  2 +  were lower than that of the wild type enzyme. With the histidine mutants, the lower k  C2t  values are consistent with the notion that the Ugand sphere about the metal has been altered and that its abiUty to act as an electrophiHc catalyst has been impaired. In the case of the aspartate residue (that is Ukely displaced from the metal upon substrate binding), the lower k  CM  value may be due to the modification of a catalytic base.  Chapter III: Metal Ion Ugands  94  Zinc Ion Content of the Zn * Reconstituted Epimerases 2  The  mutant enzymes were analysed for zinc ion content by ICPMS after  reconstitution and purification on a size exclusion column. A large decrease in the amount of bound metal was observed for the two histidine residues that were mutated, but not for the aspartate residue (Table 3.5). This observation supports the notion that the histidine residues serve as metal ligands. It is not clear why the zinc-reconstituted, dialysed, H95N and H97N enzymes still have measurable activity in the absence of added zinc. One possible explanation is that the assay solutions probably contain divalent metal ions; transketolase, one of the coupling enzymes, requires a divalent metal ion for activity. The mutant enzymes may be scavenging these metals and may have widely differing binding constants. This could explain why the H 9 7 N mutant had a great increase in activity in the presence of additional zinc but the H 9 5 N mutant did not.  T a b l e 3.5. Zinc Content of Wild Type and Mutant 4-Epimerases. " 4-Epimerase Wild Type  Zinc Ions per Subunit 1.05 ± 0 . 0 5  H95N  <0.01 ± 0.05  H97N  0.14 ± 0.05  D76N  0.95 ± 0.05  " As determined by ICPMS.  Chapter III: Metal Ion Ugands  95  The metal content of the D 7 6 N mutant epimerase remained at about the same level as the wild type enzyme, when purified by size exclusion chromatography. However, it seemed to lose metal ion during extensive dialysis as analysed by kinetics. Presumably the release of the metal ion occurs slowly and litde is lost during the column purification, which is relatively rapid. The reduced k  CM  of this mutant indicates that D76 does have a role in  catalysis. These observations might be consistent with the role of the corresponding amino acid residue (E73) in the aldolase, which rotates away from the zinc ion on binding of the substrate. Therefore, this residue might not be important to metal binding in the aldolase either.  It would be interesting to see how the activity and metal binding ability of the  aldolase would be affected by mutating this residue. It would be necessary to analyse the kinetics and thermodynamics of metal complexation before these questions can be answered.  values for metal binding could be  determined using isothermal titration calorimetry on the apoenzymes.  Activity of the Cobalt-Substituted Epimerases Zinc is spectroscopically invisible, so in order to study effects of the metal ligands on the metal-binding site, it was decided to prepare samples of the cobalt(II) form of the enzymes for these studies. As in preparation of the zinc forms, apoenzymes were prepared by dialysis against E D T A , and were subsequendy reconstituted with excess cobalt, which was removed by extensive dialysis against metal-free buffer.  'er III: Metal Ion  96  Kinetic studies o f the Co -substituted w i l d type enzyme showed that it had a higher 2+  k  than the Zn -substituted enzyme (Table 3.6). This is not an unusual effect o f cobalt2+  cat  substitution (Maret & Vallee, 1993), and has also been observed for L-fuculose-1-phosphate aldolase (Dreyer & Schulz, 1996b). U n l i k e the zinc form o f the enzymes, the K values o f M  the mutant enzymes were affected by the metal i o n substitution and are approximately twice the K value o f the Co -substituted w i l d type enzyme. T h e origin o f the higher K 2+  M  M  values  observed for the cobalt-substituted mutant enzymes is not k n o w n . T h e true extent o f the effects o f C o  2 +  substitution o n the kinetics o f the w i l d type and mutant enzymes is not  k n o w n since the amount o f b o u n d C o  2 +  was not determined, nor were the kinetics studied  i n the presence o f additional C o . 2 +  T a b l e 3.6. K i n e t i c Parameters o f the Cobalt-Substituted W i l d Type and M u t a n t L Ribulose-5-Phosphate 4-Epimerases. h  4-Epimerase  K  W i l d Type  0.093 ± 0.009  26.9 ± 0.8  2.9 x 10  5  H95N  0.28 ± 0.03  0.10 ± 0.004  3.6 x 10  2  H97N  0.20 ± 0.02  0.082 ± 0.003  4.2 x 10  2  D76N  0.23 ± 0.02  0.11 ± 0.003  4.6 x 10  2  u  (mM)  k  (s )  kJK  4  cM  M  (M  1  s") 1  " These were determined i n the same manner as the Zn -substituted enzymes without any addition o f extra divalent metal ions to the cuvettes. See A p p e n d i x D for a graphical representation o f these kinetics. 2+  h  Chapter III: Metal Ion Ugands  97  UV-Visible Spectra of the Co (II) Substituted Enzymes UV-visible difference spectra were collected of the wild type and mutant C o 2 +  substituted 4-epimerases.  There was no significant absorbance in wild type or mutant  enzymes in the 500 to 700 nm range, suggesting that no cysteine residues were acting as metal-ion ligands (none were expected). Unfortunately, little information could be obtained from the UV-visible spectra of the cobalt-substituted enzymes. This is not an unknown event; cobalt(II)-substituted phosphodiesterase had a similar UV-visible spectrum to that of the 4-epimerases until one of its metal ion ligands was mutated to methionine (a sulphur donor) (Kuo etal, 1997).  Co *-Modification of the Co *-Substituted Wild Type 4-Epimerase 3  2  Co(III)-modification of Co -substituted enzymes has been used to identify metal ion 2+  ligands (Hlavaty & Nowak, 1997). In an effort to see if this technique could be used to identify the metal ion ligands of L-ribulose-5-phosphate 4-epimerase, the C o wild type enzyme was modified to the C o  3 +  2 +  form of the  form.  After hydrogen peroxide oxidation of the Co -substituted 4-epimerase, the mass of 2+  the modified epimerase was found to be 25524 Da. The mass of the unmodified epimerase has been experimentally determined to be 25524 D a (Chapter II).  There was a small  shoulder peak at 25581 D a (Figure 3.10), which corresponds to an addition of 59 D a to the wild type epimerase; this is the mass of cobalt. It was not clear if this peak was indeed from  Chapter III: Metal Ion Ugands  98  Chapter III: Metal Ion Ugands  99  addition of Co(III) to the enzyme or an impurity in the enzyme preparation, or if it was an artifact of hydrogen peroxide oxidation of the enzyme. Pursuit of the Co(III)-modified wild type epimerase may be worthwhile for a definitive identification of the metal-ion ligands. Better modification may be achieved with a longer hydrogen peroxide incubation; times of up to twelve hours are not unknown (Van Wart, 1988), and thirty minutes might be a little short. In addition, it might be interesting to see how the activity of the epimerase is affected by oxidation of Co(II) to Co(III).  Conclusion  This work provides evidence that there is a structural link between the L-ribulose-5phosphate 4-epimerase f r o m K coli and the Class II L-fuculose-l-phosphate aldolase from E. coli. The aldolases have traditionally been characterized as either Class I in which a lysine residue forms a Schiff base with a carbonyl of the nucleophilic substrate or as Class II in which a metal ion serves to stabilize the enediolate intermediate (Horecker et al., 1972; see Chapter I). Mutation of three of the four putative metal ion ligands of L-ribulose-5-phosphate 4-epimerase caused drastic reductions in the activities of these enzymes. These ligands were postulated to be metal binding based on homology between the 4-epimerase and a bacterial L-fuculose-l-phosphate aldolase, whose crystal structure has been solved (Dreyer & Schulz,  Chapter III: Metal Ion Ugands  100  1993; Dreyer & Schulz, 1996b). Mutagenesis of these residues affected the metal-binding abilities of the enzyme, indicating that these residues are involved in ligation of the divalent metal ion known to be required for catalysis. A reduction in k  C3t  mutant enzymes.  was observed for all three  In the case of the histidine mutants, this can be explained by having  changed the coordination sphere of the metal ion. In the case of the aspartate mutant, this residue, in addition to being a metal ion ligand might also be a catalytic base in the active site. By analogy to X-ray crystal structures of the aldolase (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996a; Dreyer & Schulz, 1996b), this residue might bind the metal ion in the uncomplexed enzyme and then rotate away upon substrate binding, and could then participate in catalysis. Since the active site Ugands to the divalent metal ion in the epimerase and the aldolase are conserved, a mechanism very similar to that of the aldolase could be employed by the epimerase provided that the attack on the aldehyde is non-stereospecific and the bound enediolate is protected from any proton source (Path B; see Chapter I). It is of interest to note that recent additions to the Swiss Protein Data Bank have added an annotation that the D76, H95, H97, and H171 residues of L-ribulose-5-phosphate 4-epimerase are probable zinc-ion Ugands based on the similarity of the sequences of the 4epimerase to L-fuculose-l-phosphate aldolase. Furthermore, with increasing numbers of bacterial genome sequences being reported, new sequences of enzymes are continuaUy being added to the A r a D / F u c A family. The functions of these proteins are not known, but they are thought to be involved in pentose metaboUsm pathways. The N-terminal sequences of  'er III: Metal Ion  101  each of these proteins share homology in a "catalytic domain", and in each case the known zinc-ligands of L-fuculose-l-phosphate aldolase are conserved.  Experimental Methods  Data Base Searches and Protein Sequence Alignments The Swiss Protein Data Bank was searched for protein sequences similar to that of L-ribulose-5-phosphate 4-epimerase. S E Q S E E (Wishart etal., 1994) was used to align the protein sequences of the L-ribulose-5-phosphate 4-epimerases of  Escherichia coli, Salmonella  typhimurium, and Haemophilus influenzae, L-fuculose-l-phosphate aldolases of E. coli and S. typhimurium, and E. ^//L-rhamnulose-l-phosphate aldolase.  Bacterial and Phage Strains E. coli DH5<xF' was used for preparation of plasmid D N A for site-directed mutagenesis and sequencing. This strain was also used in the s s D N A rescue experiments.  E. coli Y1090 (modified) was used (as described in Chapter II) for mutant enzyme overexpression to avoid any contamination from low levels of endogenous L-ribulose-5phosphate 4-epimerase production. Helper phage M13K07 was used in the ssDNA rescue experiments.  Chapter III: Metal Ion Ugands  102  ssDNA Rescue ofpREl L B broth (25 mL) with 50 ug/mL ampicillin was co-inoculated with  E. coli D H 5 a F '  (transformed with p R E l ) and helper phage M13K07, and incubated in a 3 7 ° C shaker overnight. Cells were removed from the growth medium by centrifugation at 5000 rpm in a Sorvall G S A rotor for 10 minutes. T o the supernatant was added 2.5 m L P E G / N H O A c 4  (50:50 vol 50% P E G 8000: 7.5 M N H O A c ) . This mixture was incubated on ice for 30 4  minutes and then centrifuged for 20 minutes at 10000 rpm in a Sorvall SS34 rotor. The supernatant was removed (and discarded), and the precipitated phage pellet was redissolved in 400 uL sterile distilled water. The phage suspension was extracted twice with 400 uL P / C / I (25:24:1 vol phenol: chloroform: isoamyl alcohol) and then once with 400 uL chloroform. N H O A c (20 uL, 3 M) and isopropanol (1 mL) were added to the aqueous 4  phase and mixed.  After 15 minutes of incubation at room temperature, ssDNA was  collected by a 10 minute centrifugation in a microfuge at 4 ° C . Recovery of ssDNA was checked by 1.2% agarose gel electrophoresis. s s D N A was used in sequencing reactions with Sequenase version 2.0 D N A Sequencing Kit and methods suggested by the supplier (United States Biochemical) for dideoxynucleotide chain termination sequencing of single stranded D N A .  Site-Directed Mutagenesis Primers for each mutation (Table 3.6) were designed so as to include a silent mutation that introduces a new restriction site in the  araD gene. This allows for quick screening of  Chapter III: Metal Ion Ligands  103  putative mutants. Figure 3.11 shows the relative primer binding locations and the restriction sites which are present on p R E l and those which are introduced by mutagenesis. H 9 5 N . A megaprimer was formed according to the method of Chen and Przybyla (1994) in a first round of P C R with p R E l serving as the template and primers MT05 and AJ1.  AJ1 contained the desired mutation and a silent mutation introducing a Hpa I  restriction site. The first round of PCR contained 0.1 m M of each d N T P , 10 m M Tris-HCl (pH 8.4), 50 m M KC1, 1.5 m M MgCl , 0.5 u M of each primer, ca 10 ng of p R E l template 2  and 2.5 U Taq D N A polymerase (Gibco/BRL) in 100 uL. 100 uL mineral oil was added to prevent evaporation during thermocycling.  A n initial melt at 95 ° C for 1 minute was  followed by 25 cycles of 9 5 ° C (30 sec), 4 5 ° C (30 sec), and 7 2 ° C (1 min), and ended with a final extension of 7 minutes at 72 °C. The megaprimer was purified using Wizard PCR Preps D N A Purification System (Promega), then treated with Nde I to remove D N A from MT05 which was not complementary to the plasmid p R E l . The resulting D N A was used as the primer in each of two subsequent rounds of P C R to create two halves of the plasmid, as described by Jones and Winistorfer (1992). Reactions were run with the megaprimer and MT08 on p R E l (linearized with  PstT), and with MT09 (linearized with EcoR I). The second  round of PCR was similar to the first round except that there were 28 cycles of PCR, 0.5 u M of MT08 or MT09 was used with about 5% of the megaprimer from one PCR, treated with Nde I.  The two halves were then mixed together, thermally denatured ( 9 4 ° C , 3 min),  annealed ( 5 0 ° C , 2 h), and then transformed into competent  E. coli DH5aF'. Clones were  ter III: Metal Ion Table 3.6.  Primers Used for Site-Directed Mutagenesis of the araD Gene.  Primer  Sequence"  AJI  5'-GCGAGTGCGT<7rrA4CAATGCCGCCAATGGA-3'  b  AJ2'  S'-CGTGGCGCGA^TrCGTATGCACAATGCCG-S'  AJ3  d  AJ4  e  AJ5  /  5'-GTTGGCGTGTrCG^GAGGGCTTTrTCGTACC-3'  AJ6  &  5 -TGGTCCATTCGA4CGGCCCGTTTGCAT-3'  AJ7  h  MT08  5'-TGTGCATACGA47TCGCGCCACGCGACCA-3' 5*-AAAAGCCCTC 7TCGA4CACGCCAACTCACCGG-3'  ,  5'-AACGGGCCGTrCG^L4TGGACCAGAACG-3' 1  MT09 J a  104  5'-CGACGAGCGTGACACCACGATGCC-3' 5'-GCAGAGCGAGGTATGTAGGCGGTG-3'  Mismatches are shown in bold, new restriction sites in italics.  This primer, used to make the H95N mutation found in pAJl is complementary to the sense strand of araD between nt 265 and 295, and introduces a Hpa I restriction site. AJ2 is complementary to the sense strand of araD between nt 273 and 301, and introduces an EcoK I restriction site as well as the H 9 7 N mutation found in pAJ2. This primer was used in construction of the H 9 7 N mutation of pAJ2. It is identical to the sense strand of araD between nt 279 and 307, and, like AJ2, introduces an EcoK I restriction site and the H 9 7 N mutation. ' Identical to the sense strand of araD between nt 212 and 243, this primer introduces the D 7 6 N mutation and a G^45 I restriction site found in pAJ3. f This primer, which binds to the sense strand between nt 205 and 247, was used to create pAJ3. This plasmid encodes the D76N mutation of L-ribulose-5-phosphate 4-epimerase, and has a new C^45 I restriction site. AJ6 binds to the antisense strand of araD between nt 526 and 500, and introduces an H171N mutation along with a new Csp45 I restriction site. * Binding to the sense strand of araD between nt 495 and 521, AJ7 introduces a H171N mutation and a new Csp45 I restriction site. MT08 binds to the + strand of p R E l (contains the anti-sense strand of araD) between nt 3071 and 3095 (see Appendix A). MT09 binds to the - strand of p R E l (contains the sense strand of araD) between nt 2365 and 2388 (see Appendix A). h  c  d  g  1  j  Chapter III: Metal Ion Ugands  F i g u r e 3.11. A Schematic Diagram of p R E l Indicating Relative Primer Binding Sites and Locations of Restriction Sites on p R E l and Restriction Sites Introduced by Site Directed Mutagenesis.  105  Chapter III: Metal Ion Ligands  106  picked and grown up in 5 m L cultures of L B with 50 |og/mL ampicillin at 37 °C. Plasmids were purified and screened for the desired mutation by digestion with both Hpa I and EcoR I. The construct p A J l was found to contain the desired mutation by sequencing of the mutant araD gene. H97N.  A plasmid encoding the H 9 7 N mutant was not readily made using AJ2 and  MTO5 following the same method as for H95N.  Instead, two primers (AJ2 and AJ3)  complementary to each other and both encoding the desired mutation and a silent mutation introducing a new EcoR I restriction site were used. These primers were used in Jones and Winistorfer's (1992) recombinant circle P C R (RCPCR) technique. To obtain one half of the plasmid, p R E l (linearized with Pst I) was amplified by P C R using primers AJ2 and MT08, while the other half was obtained by P C R amplification of p R E l (linearized with EcoR I) and primers AJ3 and MT09. Each amplification went through 28 cycles of P C R as described for formation of the H95N mutant, and each reaction contained in 100 uJL: 0.5 u M of each primer, 0.1 m M each d N T P , ca. 10 ng of template D N A , 10 m M Tris-HCl (pH 8.4), 50 m M KC1, 1.5 m M M g C l , and 2.5 U Taq D N A polymerase. 100 uL of mineral oil was used to 2  prevent evaporation during thermocycling. As with H95N, the two plasmid halves were annealed and transformed into E. coli DH5aF'. Plasmids isolated from saturated (overnight) cultures of resulting colonies were screened for the desired mutation by digestion with EcoR I.  pAJ2 contained the desired mutation, and its sequence was confirmed by D N A  sequencing.  Chapter III: Metal Ion Ugands  107  D76N. The construct pAJ3, which contained the araD gene encoding the desired mutation was created using the same RCPCR method as was used for H97N. In this case, however, a silent mutation was used which introduced a Csp45 I restriction site. The two plasmid halves were created by PCR amplification of pREl (linearized with EcoK I) and primers AJ4 and MT09, and of pREl (linearized with Pst I) and primers AJ5 and MT08. Plasmids were screened by restriction digestion with both Csp45 I and EcoK I. The sequence of araD was verified by D N A sequencing. H171N. Primers encoding this mutation also included a silent mutation introducing a Csp45 I restriction site, and were used in the same RCPCR procedure that was used for creating H97N and D76N.  In this case primers AJ6 and MT09 were used for PCR  amplification of pREl (linearized with EcoK I), and primers AJ7 and MT08 were used to amplify pREl (linearized with PstT). Colonies resulting from the transformation of E. coli D H 5 a F ' with annealed plasmid halves (the two PCR products) were used to inoculate overnight cultures of LB broth containing 50 ug/mL ampicillin and grown overnight in a 37 °C shaker. Plasmids isolated from these cultures were screened for restriction by both EcoK I and Csp45 I.  Chapter III: Metal Ion Ugands  108  DNA sequencing Sequencing reactions were performed on the ssDNA (isolated as described above) using primers MT12 and MT15 to determine if single stranded p R E l (containing the  araD  gene) was among the ssDNA isolated. These reactions were performed in the same manner as those performed on plasmid D N A except that there was no denaturation step. The  araD  gene of each mutant plasmid was sequenced in entirety using the Sequenase version 2.0 D N A sequencing kit (United States Biochemical) to ensure that no spurious mutations had occurred during site-directed mutagenesis. Double-stranded plasmid D N A was isolated for this purpose using the Wizard Minipreps D N A Purification System (Promega). The primers used were the same as those used to sequence the wild-type earlier. In addition, the  araD gene of p R E l as described  araD gene of pAJ3 was sequenced by Debbie Neufeld in the Nucleic  Acid and Protein Service at the University of British Columbia using primer AJ10 (5'C G A A C A C G C C A A C T C A C C - 3 ' ) and Applied Biosystems Inc. PRISM technology.  Enzyme Purification Mutant L-ribulose-5-phosphate 4-epimerases were purified using a modification of the method described by Lee etal. (Gielow & Lee, 1975; Lee et al, 1968), and in the same manner as the wild-type enzyme (Chapter II). For the purification of the mutant enzymes, pAJl (for H95N), pAJ2 (for H97N), and pAJ3 (for D76N) were individually transformed into a functional 4-epimerase.  E. coli Y1090 (modified), which lacks  The ammonium sulphate precipitation step was omitted with the  Chapter III: Metal Ion Ugands  109  H97N mutant as this step led to large losses of this mutant. Otherwise, the purification procedure used is described in detail in Chapter II. Protein determination was performed by measuring the absorbance of protein samples at 280 nm and using the extinction coefficient determined for the wild type enzyme (Chapter II), e = 1.73 mL mg" cm" , or e = 33717 M" cm" per subunit. 1  1  1  1  Calculation and Determination of Mutant Enzymes' Subunit Masses As with the wild-type enzyme, the masses of the mutant enzymes were calculated based on their amino acid content.  The masses were determined experimentally by  electrospray ionization mass spectrometry (performed by David Chow) as described in Chapter II.  Determination ofTetrameric Structure of the Native Enzymes The tetrameric nature of the mutant enzymes was established by gel filtration chromatography with a Waters HPLC and a Protein-Pak 300 SW column (Waters). A gel filtration calibration curve was constructed using pyruvate kinase (237 kDa tetramer), aldolase (158 kDa tetramer), rabbit muscle lactate dehydrogenase (140 kDa monomer, bovine serum albumin (66 kDa monomer), and ovalbumin (43 kDa monomer). Blue Dextran was used to determine the void volume of the column. Standard solutions were made in 100 mM phosphate buffer (pH 7.0) at concentration of 1 mg/mL. Wild type and mutant L-ribulose-5-phosphate 4-epimerase were also prepared in 100 mM phosphate buffer  Chapter III: Metal Ion Ugands  110  (pH 7.0) at 1 mg/mL. Injections of 50 uL of each sample and standard were made at a flow rate of 1 m L / m i n .  Preparation ofMetal-Free Buffers and Glassware Metal-free buffer was prepared by passing a solution of 50 m M H E P E S (2 L , containing 10% glycerol) through a column of Chelex-100 resin (70 mL, N a form, 50-100 +  mesh, changed to H form by washing with 2 volumes 1 N HC1 and rinsing with 5 volumes +  deionized H 0 ) . The p H was adjusted to 7.6 using solid Tris base. Adventitious metal ions 2  were removed from glassware and plasticware used in this procedure by soaking them overnight in 4 N HC1 and then washing them thoroughly with deionized water (17.8 M Q / c m at 25°C).  Preparation of Apoenzyme and Reactivation with Zn(II) or Co(II) Apoenzymes were prepared in the same manner as described for the wild type enzyme in Chapter II. 5 mg of the purified mutant (and wild type) enzymes were incubated with 20 m M E D T A in a total volume of 2.5 m L at room temperature for 3 hours. E D T A and any metal ions chelated to it were removed by dialysis against two 500 m L portions of metal-free H E P E S / T r i s buffer (pH 7.6) containing 10% glycerol. The apoenzymes were reconstituted by addition of 10 equivalents of Z n C l (99.99%) to a concentration of 0.4 m M 2  and incubating at room temperature for about 2 hours. Excess metal ions were removed by dialysis against two 500 m L portions of metal-free H E P E S / T r i s buffer.  Chapter III: Metal Ion Ugands  111  Alternatively, the apoenzymes  (of the wild type and mutant enzymes) at a  concentration of between 2 and 2.5 m g / m L were reconstituted with the addition of 10 equivalents of a solution of C o C l (99.999%, 250 uL of a 4 m M solution) to a concentration 2  of 0.4 m M and incubating the mixture at room temperature for about 2 hours. Excess metal ions were removed by dialysis against two 500 m L portions of metal-free 50 m M H E P E S / T r i s (pH 7.6) buffer.  Circular Dichroism and Thermal Stability Solutions of the zinc-substituted enzymes were exchanged into 10 m M potassium phosphate buffer, p H 7.6 using Amicon centricon-10 concentrators. These solutions were diluted to 0.17 m g / m L and were scanned in the far U V (300 to 190 nm) on a Jasco J-720 spectropolarimeter. The thermal stabilities of the wild-type and mutant epimerases were determined by observing the change in C D signal at 220 nm in 50 m M H E P E S / T r i s buffer, p H 7.6 as a function of temperature. The temperature of the sample was raised from 30 ° C to 75 ° C at a rate of 5 0 ° C / h o u r using a N E S L A B M-RS-232 bath/computer interface, a N E S L A B RTE-111 waterbath and a N E S L A B RS-2 remote sensor.  These experiments were  performed in the laboratory of Dr. Grant Mauk in the Department of Biochemistry at the University of British Columbia. The data were analysed by fitting them to an I C using the computer programme GraFit.  50  curve  Chapter III: Metal Ion Ligands Determination  ofZinc(II)  112  Content  As described for the wild type enzyme in Chapter II, samples of the dialysed Znreconstituted enzymes were passed through a size exclusion column (Waters Protein Pak 125) at 0.5 mL/min to remove any excess metals remaining after dialysis. These samples were diluted 1/400 in 10% nitric acid containing 20 ppb Sc as an internal standard, and allowed to sit at room temperature at least overnight. A sample of the column buffer was also treated in this way. These samples were analysed for Zn content by ICPMS on a V G Elemental Plasma Quad mass spectrometer which had been calibrated with a standard curve of Zn (0, 5, 15, 40 ppb, prepared in 10% nitric acid containing a 1/400 dilution of the elution buffer and 20 ppb Sc as an internal standard) by Bert Mueller in the Department of Oceanography at the University of British Columbia.  Assay for L-ribulose-5-phosphate 4-epimerase activity  The kinetic parameters for each of the Zn(II)- and Co(II)-reconstituted enzymes were measured by monitoring the absorbance at 340 nm on a Cary 3E spectrophotometer at 37°C, as described in Chapter II, but with some minor modifications. In these assays, performed in 0.5 mL quartz cuvettes, each reaction contained, in a total volume of 500 uL, 25 mM glycylglycine (pH 7.6), 5 mM ribose-5-phosphate, 0.1 mM TPP, 0.15 mM N A D H , LRu5P between 0.02 and 1 mM, 2.5 U a G D H , 25 U TIM, and 0.125 U transketolase. The reactions were initiated by addition of 0.025 ug of wild type or 0.25 ug of mutant 4-epimerase.  Chapter III: Metal Ion Ugands  113  Additionally, the kinetic parameters were determined for the Zn(II) reconstituted enzymes in the presence of 0.1 m M Z n  2 +  in each 0.5 m L cuvette. In this case, each reaction  was initiated by the addition of 0.025 u,g of wild type epimerase, 0.25 u,g of the H95N or D 7 6 N epimerases, or 0.050 ug or the H97N epimerase. Kinetic parameters were determined from a direct fit of the data to an enzyme kinetics equation using the computer program GraFit (Erithacus Software, UK).  Errors  were obtained using GraFit, which performs a non-linear regression using the method of Marquart (1963).  Assay for Apoenzyme Activity Apoenzyme at 1 m g / m L (mutant epimerases) or 0.01 m g / m L (wild type epimerase) was incubated at 3 7 ° C for 5 minutes in metal-free H E P E S / T r i s buffer, p H 7.6, containing 9.44 m M LRu5P.  Controls were water and untreated enzyme.  10 uL aliquots were  immediately added to pre-equilibrated cuvettes containing (in 990 uL) 25  umoles  glycylglycine (pH 7.6), 0.15 umole N A D H , 5 umoles DR5P, 0.1 umole TPP, 5 U a G D H , 50 U T I M and 0.25 U transketolase.  The activity of the apoenzyme was compared to the  activity of the untreated enzyme and water control over one minute of reaction in the cuvette, as descrbed in Chapter II.  Chapter III: Metal Ion Ugands  114  UV-Visible Spectra of the Co(II)-Substituted 4-Epimerases UV-visible spectra between 350 and 700 nm of the wild type and mutant enzymes were collected on a Cary 3 E UV-Vis spectrophotometer at 25°C. Difference spectra were obtained by subtracting the UV-Vis spectrum of the Zn(II) containing epimerase from that of the Co(II)-substituted epimerase. In addition, spectra were buffer subtracted. The protein solutions used were all at 200 u M (based on subunits) in 50 m M H E P E S / T r i s buffer, p H 7.6.  Co(III) Modification of the Co(II)-Substituted Wild Type L-Ribulose-5-Phosphate 4-Epimerase This procedure was a variation of the Co(III) modification of phosphenolpyruvate carboxykinase reported by Hlavaty & Nowak (1997). Hydrogen peroxide (3%; 22 uL) was added to Co(II) substituted wild type epimerase at 61.5 u M (1.57 mg/mL; 98 uL) in 50 m M H E P E S / T r i s buffer (pH 7.6, containing 10% glycerol) to give a final concentration of 20 mM H 0 . 2  2  This solution was incubated on ice for 30 minutes with occasional agitation.  After incubation, the reaction was diluted to 2 m L with 50 m M H E P E S / T r i s buffer (pH 7.6), and excess hydrogen peroxide was removed by passage through an Amicon centricon10 filter unit at 5000 rpm in a Sorvall G S A rotor for 1 h. The final enzyme concentration, as determined by the absorbance at 280 nm was 2.6 mg/mL. The mass of the modified enzyme was determined by electrospray ionization mass spectrometry performed by Mr. Shouming He.  Chapter IV  L-Ribulose-5-Phosphate 4-Epimerase Follows a Retroaldol/Aldol Mechanism  Introduction  A structural relationship between Escherichia and the Class II L-fuculose-l-phosphate aldolase of  coli L-ribulose-5-phosphate 4-epimerase E. coll was established in Chapter III.  In particular, the active site zinc-binding ligands were found to be conserved in these two enzymes. Given this similarity, it is reasonable to suspect that these enzymes also share a related mechanistic strategy. L-FuculoseT-phosphate aldolase uses an aldol (or retroaldol in the reverse direction) reaction to form L-fuculose-l-phosphate from dihydroxyacetone phosphate and L-lactaldehyde (Figure 4.1A). A similar mechanism can be drawn for the epimerase (Figure 4.IB) in which an enzyme-catalysed retroaldol reaction breaks the carboncarbon  bond  of  L-ribulose-5-phosphate  115  between  C-3  and  C-4,  creating  the  Chapter IV: A Retroaldol/Aldol Mechanism  116  F i g u r e 4.1. Class II L-Fuculose-1-Phosphate Aldolase (A) and L-Ribulose-5-Phosphate 4Epimerase (B) Can Utilise Similar Mechanistic Strategies.  Chapter TV: A. Retroaldol/'A.ldo7 Mechanism  117  enediolate of dihydroxyacetone and glycolaldehyde phosphate. If the enediolate adds back to the opposite face of the aldehyde in an aldol condensation, D-xylulose-5-phosphate is formed. The best proof of an enzymatic pathway is identification of the reaction intermediates (if there are any). Unfortunately it is not always possible to isolate and characterize the intermediates: they may not be released from the enzyme's active site; they may be unstable outside the enzyme's active site; or, if they are released, the equilibrium constant of their release may be sufficiendy unfavourable that only very small quantities are ever present in solution. A n alternative approach is to supply the proposed intermediates in solution to the enzyme. If the enzyme is able to bind these molecules into its active site, they may be treated by the enzyme in exacdy the same manner as the regularly formed intermediates and equilibrated into the normal ratio of substrate and product. In the case of L-ribulose-5-phosphate 4-epimerase, release of the intermediates formed during an aldol mechanism would result in the production of dihydroxyacetone and glycolaldehyde phosphate.  Dihydroxyacetone is the protonated form of the enediolate  Figure 4.2. Reduction of Dihydroxyacetone to Glycerol by Glycerol Dehydrogenase.  Chapter IV: A Retroaldol/Aldol Mechanism  118  intermediate, and can be detected by coupling the release reaction to glycerol dehydrogenase, an enzyme which reversibly reduces dihydroxyacetone to glycerol with concomitant oxidation of N A D H to N A D  +  (Burton & Kaplan, 1953; Spencer etal, 1989; Scharschmidt  etal., 1983) (Figure 4.2). It is feasible that a small fraction of the time the bound enediolate could accept a proton and the resulting dihydroxyacetone could be displaced from the catalytic Z n  2 +  ion and released from the enzyme.  Alternatively, the coupling of dihydroxyacetone and glycolaldehyde phosphate would produce the epimeric pentose phosphates. The formation of these sugars could be detected using a stopped form of the normal epimerase assay (described in Chapter II).  In coupling  dihydroxyacetone and glycolaldehyde phosphate to form an epimeric mixture of the ketopentose phosphates, the enzyme must overcome a number of obstacles. The aldehyde in solution is primarily in its hydrated form, and would have to be dehydrated before an aldol condensation could occur. A large proportion of dihydroxyacetone (60 to 80%) is also present in aqueous solution as the hydrate. Furthermore, the enzyme must have a means of deprotonating dihydroxyacetone to form the proposed enediolate intermediate.  119  Chapter IV: A Retroaldol/Aldol Mechanism  Results and Discussion (  Intermediate Identification with the Wild Type L-Ribulose-5-Phosphate 4-Epimerase Initial studies were directed toward investigating whether or not the wild-type epimerase was able to promote carbon-carbon bond cleavage by observing low levels of aldolase activity. T o begin with, ^ - N M R (200 MHz) was used to detect formation of LRu5P and D X u 5 P from dihydroxyacetone and glycolaldehyde phosphate. amounts  The very small  of sugar formed in the coupling reaction, relative to the amounts  of  dihydroxyacetone and glycolaldehyde phosphate, made it very difficult to be certain that coupling had occured. Additionally, these coupling reactions were performed in a phosphate buffer, and it appeared that the phosphate buffer itself may have catalysed the aldol reaction to some extent. ' H - N M R was also used to observe the release of dihydroxyacetone and glycolaldehyde phosphate from the enzyme with an equilibrating pool of LRu5P and DXu5P. Again, it was not certain that dihydroxyacetone and glycolaldehyde phosphate were observed.  It was decided that a coupled assay for release of dihydroxyacetone or  glycolaldehyde phosphate might be more sensitive.  The chance observation that  glycolaldehyde phosphate could act as a substrate (albeit a poor one) for OC-glyceraldehyde phosphate dehydrogenase (a G D H ) provided a means of observing the release of this proposed intermediate. These reactions were initially followed for a short time by observing a change in the absorbance at 340 nm, but no dihydroxyacetone or glycolaldehyde phosphate  Chapter IV: A Retroaldol IAldol Mechanism  120  could be detected. Further incubation, periodically measuring the absorbance at 340 nm, still did not detect formation of dihydroxyacetone or glycolaldehyde phosphate. Following the release of dihydroxyacetone or glycolaldehyde phosphate continuously for up to 12 hours still did not provide any evidence for intermediate release. Eventually it was decided that stopped assays to follow the aldol coupling of dihydroxyacetone and glycolaldehyde phosphate over a long time (up to 72 hours) might yield better results.  The wild type enzyme was therefore assayed for aldolase activity  (carbon-carbon bond formation) by extensive incubation of a large amount of enzyme (0.5 mg/mL) with dihydroxyacetone (20 and 50 mM), glycolaldehyde phosphate (5 mM) and Z n C l (0.1 mM) at p H 7.6. At timed intervals, aliquots were removed and the total amount 2  of LRu5P and DXu5P was determined using a stopped assay for DXu5P using the coupled system described for epimerase kinetics in Chapter II. This assay normally detects only DXu5P; however, the large amounts of epimerase present ensured rapid equilibration between the epimers and allowed determination of both epimers. This assay could not be run continuously since very long incubation times were employed and a small background rate was observed when glycolaldehyde phosphate itself was incubated in the presence of the coupling enzymes and N A D H . The results showed that low levels of the pentose phosphate epimers were produced under these assay conditions; however, it was difficult to demonstrate unambiguously that they were not a result of a background buffer-catalysed reaction.  Chapter IV: A Retroaldol/Aldol Mechanism  121  Eqirikbrium constants observed for related aldolase-catalysed reactions, where X  e q  =  [DHAP][aldehyde]/[sugar], are 1 x 10" M for fructose-l,6-bisphosphate aldolase (Wong et 4  al, 1995), 4.6 x IO" M for FucA (Ghalambor & Heath, 1962; Ghalambor & Heath, 1966), 4  and 1 x 10" M for tagatose-l,6-bisphosphate aldolase (Fessner & Eyrisch, 1992). If the 3  equilibrium constant for the aldolase activity of L-ribulose-5-phosphate 4-epimerase lies in this range, then far more than 0.4 m M ketopentose phosphate should have been formed from 50 m M dihydroxyacetone and 5 m M glycolaldehyde phosphate. It was therefore of interest to determine if the low conversion of dihydroxyacetone and glycolaldehyde phosphate to LRu5P and DXu5P could be a result of "product" inhibition. The aldol substrates did not appear to bind well to the epimerase since the epimerisation of L-ribulose5-phosphate (0.094 mM) was not significantly inhibited in the presence of 0.1 m M Z n  2 +  by  either dihydroxyacetone (50 mM) or glycolaldehyde phosphate (5 mM). The rate was within 5% of the rate of the enzyme assayed in the absence of dihydroxyacetone and glycolaldehyde phosphate. In addition, a combination of these molecules (at 50 and 5 m M , respectively) did not cause any synergistic inhibition of the epimerase in the presence of 0.094 m M L-ribulose5-phosphate and 0.1 m M Z n . 2 +  The enzyme therefore has relatively high K  u  values for  dihydroxyacetone and glycolaldehyde phosphate. The aldol products, Ru5P and Xu5P, are therefore good competitive inhibitors of this reaction and prevent the equiUbrium from being reached within 48 hours. In addition, the enzyme was found to lose activity with a half life of about 30 h under these conditions.  Chapter IV: A Retroaldol/Aldol Mechanism  122  Although L-ribulose-5-phosphate 4-epimerase and L-fuculose-l-phosphate aldolase were found to share a common metal binding motif, they do not necessarily implement similar mechanisms. The observation of any aldolase activity with the epimerase would provide evidence in support of the idea that this enzyme operates by catalysing carboncarbon bond breaking and making reactions. The wild type enzyme did exhibit aldolase activity, but at levels only slighdy greater than the background rates. If the enediolate of dihydroxyacetone and glycolaldehyde phosphate are intermediates in this reaction, the enzyme has evolved to keep themtighdybound in its active site. Otherwise, aldolase activity might have been the predominant function of this enzyme. It is not common for enzymes readily to release their intermediates (a notable exception is the E. coli UDP-IVacetylglucosamine 2-epimerase (Morgan et al, 1997; Sala et al, 1996)), and L-ribulose-5phosphate 4-epimerase is no exception to this rule.  Intermediate Identification Using Mutant 4-Epimerases The observation that the epimerase mutants still retained significant activity (Chapter III) despite having an altered metal binding site led to their being tested for aldolase activity as well. It was thought that a mutant enzyme might be able to permit protonation of the bound enediolate (or conversely deprotonation of dihydroxyacetone), which could result in the release of dihydroxyacetone and glycolaldehyde phosphate (or conversely, of their consumption in the formation of LRu5P and DXu5P). Each of the mutants was tested for  Chapter IV: A Retroaldol IAldol Mechanism  123  aldolase activity in the same manner as described for the wild type enzyme, always in the presence of 0.1 m M Z n  2 +  to ensure that they were fully active (Chapter III).  The H 9 7 N mutant clearly, did catalyse the formation of LRu5P and DXu5P from dihydroxyacetone (20 or 50 mM) and glycolaldehyde phosphate (5 mM) in the presence of 0.1 m M Z n . The aldolase activity exhibited by this mutant was quite low (approximately 2 +  0.004 units/mg) but was greater than that observed with the wild type epimerase. The reaction was clearly catalytic and, in the case of the 50 m M dihydroxyacetone reaction, was followed until 30% of the glycolaldehyde phosphate was consumed (Figure 4.3A). When the reaction contained 20 m M dihydroxyacetone, 15% of the glycolaldehyde phosphate was consumed (Figure 4.3B). In both cases, formation of ketopentose phosphates consumed about 2% of the dihydroxyacetone.  If the equilibrium is similar to that for aldolases  catalysing reactions with similar substrates and favours sugar formation with an equilibrium constant in the range of 10 to 10 M" (for i C = [sugar]/ [dihydroxyacetone][aldehyde]) 3  4  1  q  (Fessner & Eyrisch, 1992; Ghalambor & Heath, 1962; Ghalambor & Heath, 1966; Wong et al, 1995), then one would expect at least 4.5 m M sugar to be formed in an analogous situation. Since this enzyme forms two isomers of approximately equal energy, then entropy should favour carbon-carbon bond formation even more than usual. Equilibrium was presumably not attained by the H 9 7 N catalysed condensation reactions because the extremely long incubation times resulted in a loss of enzyme activity (t, ~ 30 h). /2  Chapter IV: A Retroaldol/Aldol Mechanism  Figure 4.3. Formation of An Epimeric Mixture of L-Ribulose-5Phosphate and D-Xylulose-5-Phosphate From Dihydroxyacetone and Glycolaldehyde Phosphate (5 mM). A: 50 mM dihydroxyacetone; B: 20 mM dihydroxyacetone. Closed squares, wild type enzyme; closed circles, H97N; open squares, D76N; open circles, H95N; open triangles, supernatant from heat-inactivated H97N.  124  Chapter IV: A Retroaldol/Aldol Mechanism  125  Control reactions performed with the supernatant obtained from heat denatured H97N enzyme samples showed no detectable aldol products above background, indicating that the reaction was not a result of a non-protein impurity in the enzyme preparation. As expected, a control reaction run with water in place of enzyme also had no detectable activity (data not shown); therefore, Z n extent.  2 +  itself cannot catalyse the aldol reaction to any appreciable  In addition neither the wild type enzyme nor the other two mutants showed  comparable activity (all were isolated in an identical manner and assayed in the presence of 0.1  m M Zn ). 2+  Non-enzymatic condensation of glycolaldehyde phosphate and  dihydroxyacetone activity has been observed to occur on strongly basic anion exchange resin (Morgenlie, 1987), so the aldol condensation could possibly occur slowly at p H 7.6 on the surface of the enzyme. The very low levels of coupling seen in solutions containing the (active) H 9 5 N and D 7 6 N 4-epimerases indicate that only insignificant amounts of LRu5P and DXu5P could be formed in this manner. The lack of aldolase activity observed in the D 7 6 N mutant epimerase is not completely unexpected. The aspartate residue, by analogy to the L-fuculose-l-phosphate aldolase, could act as a catalytic base in the aldolase reaction (as well as a metal ion ligand), removing the proton from C-3 of dihydroxyacetone to create the reactive anionic species. Therefore, mutation of this residue to a non-reactive, neutral asparagine residue would eliminate this activity entirely. The H 9 5 N mutant also did not exhibit any appreciable aldolase activity. This mutant's low aldolase activity could simply be a reflection of its low epimerase activity.  Chapter IV: A Retroaldol/Aldol Mechanism  126  In order to identify the products unambiguously as the epimeric pentose phosphates, 1.0 m L of a reaction mixture containing 5 m M glycolaldehyde phosphate, 50 m M dihydroxyacetone and 0.1 m M Z n C l was lyophilised after 48 h of incubation at 3 7 ° C with 2  0.5 mg of the H 9 7 N epimerase, dissolved in D 0 and analysed by ' H - N M R spectroscopy. 2  The resulting spectrum showed signals identical to those of an authentic sample of the epimeric mixture that had been independently prepared using the wild type epimerase and D-xylulose-5-phosphate (Figure 4.4). A control reaction, which had no enzyme in it, had no detectable peaks due to either LRu5P or DXu5P. The observation of aldolase activity indicates that the mutant enzyme is able to bind the cleaved products from solution. In fact, glycolaldehyde phosphate was found to act as a competitive inhibitor (Figure 4.5), with K = 0.37 m M (Figure 4.6) of the H 9 7 N enzyme Y  in an assay for L-ribulose-5-phosphate epimerisation in the presence of 0.1 m M Z n . 2 +  Dihydroxyacetone (50 mM) did not significandy inhibit the epimerisation, nor was there any synergistic inhibition with dihydroxyacetone (50 mM) and glycolaldehyde phosphate (5 mM). The other two mutant epimerases were not tested for inhibition since they did not exhibit any aldolase activity.  Chapter IV: A Retroaldol IAldol Mechanism  4.5  4.2 1  127  3.9  H(ppm)  'H-NMR (500 MHz, D 0, solvent suppressed) of the H97N Catalysed Coupling Reaction Products at 72 h. A, non-enzymatic control reaction; B, H97N catalysed reaction; C, equilibrium mixture of LRu5P and DXu5P. * indicates spinning side bands from the dihydroxyacetone peak. F i g u r e 4.4.  2  Chapter IV: A Retroaldol/Aldol Mechanism  128  The observation that the H 9 7 N epimerase exhibits significant aldolase activity provides more insight into the role of the divalent metal ion in the epimerisation reaction. This activity indicates that the active site of the epimerase is capable of promoting carboncarbon bond cleavage. The H 9 7 N mutation has probably lowered the rate of the carboncarbon cleavage reaction. In addition, this mutation has disrupted the ability of the enzyme to keep the enediolate intermediate from being protonated. It seems odd that glycolaldehyde phosphate is a competitive inhibitor of the mutant enzyme but not of the wild type enzyme. This could be a kinetic effect if the form of the enzyme which accepts the substrates is different from the form which sequesters the intermediates, and binding of the intermediates from solution by the wild type enzyme occurs very slowly. Alternatively, the mutations may simply have altered the enzyme's active site structure and the coordination properties of the divalent metal ion, which could allow the mutant enzyme to bind the aldehyde more tightiy. It should be noted that the free aldehyde exists primarily in the hydrated form (Miiller  etal.,  1990) and this may be the species causing the inhibition. If this is the case, perhaps the wild type enzyme is unable to bind the hydrated aldehyde.  Chapter IV: A Retroaldol/Aldol Mechanism  129  1/[LRU5P] (mM') 1  Figure 4.5. Glycolaldehyde Phosphate is a Competitive Inhibitor of  Epimerisation by H97N L-Ribulose-5-Phosphate 4-Epimerase. Closed circles, 0.1 mM inhibitor; open squares, 0.2 mM inhibitor; closed squares, 0.4 mM inhibitor; open triangles, 0.5 mM inhibitor; closed triangles, 1 mM inhibitor; open inverted triangles, 2 mM inhibitor.  Chapter IV: A. Retroaldol/Aldol Mechanism  [glycolaldehyde phosphate] (mM)  F i g u r e 4.6.  K Detennination for H97N L-Ribulose-5-Phosphate 4Y  epimerase Inhibition by Glycolaldehyde Phosphate. The data points are the K (apparent) values from Figure 4.5. M  130  Chapter IV: A Retroaldol IAldol Mechanism  131  Detection of Intermediate Release A continuous assay to detect release of either dihydroxyacetone (using glycerol dehydrogenase) or glycolaldehyde phosphate (using CC-glycerophosphate dehydrogenase) by the wild type and mutant enzymes did not demonstrate any significant activity. Presumably, release of the intermediates was sufficiently slow during the time frame of the continuous assay (4 hours).  Therefore a stopped assay was employed to observe release of  dihydroxyacetone over 72 hours. The wild type and H 9 7 N epimerases were assayed for release in the presence of 10 m M of the equilibrating ketopentose phosphates and 0.1 m M Zn  2 +  using a stopped assay for dihydroxyacetone. A t timed intervals, aliquots were removed  and the amount of dihydroxyacetone present was determined.  The H 9 7 N epimerase  catalysed the reaction slightly more efficiently than the wild type epimerase (Figure 4.7), and the wild type epimerase catalysed the reaction at a rate greater than the background. In this case the background was determined using the supernatant from a heat-inactivated sample of the H 9 7 N epimerase.  There are a few points which indicate negative amounts of  dihydroxyacetone as a result of random error. The method for determining the amount of dihydroxyacetone is limited by the small amounts released. Therefore, this experiment served only to check the results observed from the coupling experiment and gave only qualitative results. The rate of dihydroxyacetone formation by this mutant was too low to measure accurately, however, it was clear that the aldolase products were formed. The release was clearly catalytic with the H97N epimerase, since approximately three equivalents  Chapter IV: A Retroaldol/Aldol Mechanism  Figure 4.7. Dihydroxyacetone Release by H97N and Wild Type L Ribulose-5-Phosphate 4-Epimerases Equilibrating 10 m M L-Ribulose-5Phosphate and D-Xylulose-5-Phosphate. Closed circles, H97N; closed squares, wild type; open triangles, supernatant from heat-inactivated H97N.  Chapter IV: A Retroaldol IAldol Mechanism  133  of dihydroxyacetone were produced over 24 hours. Approximately 2.5 equivalents of dihydroxyacetone were released over 48 hours with the wild type epimerase. These reactions level off after about 40 hours, at dihydroxyacetone concentrations well below expected equilibrium values, presumably as a result of the loss of enzyme activity.  An Aldol-Like Mechanism for L-Ribulose-5-Phosphate 4-Epimerase The observed aldol reaction could occur during epimerisation by an infrequent protonation of the bound enediolate by an enzymatic acid/base residue. In the reverse of the aldol reaction, this acid/base residue would deprotonate dihydroxyacetone. A n active site residue which could act as the acid/base catalyst in the aldolase reaction is the displaced D76 Ugand, by analogy to the glutamate residue (E73) of FucA. Alternatively, the mutant epimerase may simply be much "leakier" than the wild type enzyme and able to release the enediolate into solution. In this case the enzyme would have to bind the enediolate or perhaps even a zinc-enediolate complex directly from solution in the reverse direction. The retroaldol/aldol mechanism requires that for stereochemical inversion to occur, the enediolate must be able to add to either face of the bound aldehyde. Although the aldolases are usuaUy thought to exhibit high stereoselectivity, this is not always the case. For example, the Class II tagatose-l,6-bisphosphate aldolase is able to cleave the C-4 epimer (Dfructose-l,6-bisphosphate) of its natural substrate, endowing it with what is effectively an epimerase activity (Fessner & Eyrisch, 1992).  Chapter IV: A Retroaldol IAldol Mechanism  134  The aldolase activity of the H 9 7 N mutant epimerase supports a mechanism for the reaction catalysed by L-ribulose-5-phosphate 4-epimerase involving carbon - carbon bond cleavage and strengthens the notion that the enzyme is evolutionarily linked to the Class II aldolases. One noticeable difference is that FucA and the related aldolases (rhamnulose-1phosphate, fructose-l,6-bisphosphate and tagatose-l,6-bisphosphate aldolases) are highly specific for dihydroxyacetone phosphate as a donor substrate (Gijsen etal., 1996; Wong  et  al, 1995). In the case of the epimerase, dihydroxyacetone must play this role. The threedimensional structure of FucA shows that the phosphate of the intermediate analogue lies in a pocket lined with neutral residues that participate in hydrogen bonds. These residues are conserved in the epimerase sequence even though there is no phosphate group at C-1 of LRu5P and DXu5P.  A possible explanation for this is that the residues have been  conserved in order to bind to the displaced D76 residue and prevent it from protonating the bound enolate. Alternatively, this pocket could bind the C-5 phosphate group of LRu5P and DXu5P. Another difference is that sequence alignments place a threonine residue of the epimerase at the position of the FucA tyrosine 113. Threonine residues are not generally thought of as catalytic residues. This tyrosine is thought to act as a general base during the carbon-carbon bond cleavage step (Figure 4.1A) in the reaction catalysed by L-fuculose-lphosphate aldolase. In the case of the epimerase, it is likely that two spatially distinct bases (in place of the aldolase's tyrosine residue) would be required in order to promote the cleavage of both epimers (the aldolase cleaves only one epimer). A number of epimerases require two bases for epimerisation (Tanner & Kenyon, 1998). The changes required to  Chapter IV: A Retroaldol I'Aldol Mechanism  135  permit this may have caused significant restructuring i n this p o r t i o n o f the active site. A possible residue w h i c h may serve as a catalytic base i n the epimerisation is the displaced D 7 6 , although the side chain o f this residue may be too short for it to reach and deprotonate the C-4 hydroxyl group o f either L R u 5 P or D X u 5 P . It is more likely that this residue lies near C-3 o f the sugar substrate, where it could occasionally act i n the same manner as does E 7 3 in  the  aldolase  (and  give  the  epimerase  some  aldolase  activity),  deprotonating  dihydroxyacetone or conversely, protonating its enediolate.  Conclusion  The  E. coli L-ribulose-5-phosphate 4-epimerase (and especially the H 9 7 N mutant) has  been shown to catalyse an aldol addition o f dihydroxyacetone to glycolaldehyde phosphate. This w o r k strongly supports a retroaldol/aldol mechanism, similar to the mechanism used by the  E. coli Class II L-fuculose-l-phosphate aldolase, as the mechanism used by L-ribulose-  5-phosphate-4-epimerase. Therefore, the similarities between these two enzymes extend beyond amino acid sequence homology and a c o m m o n metal-binding motif.  Chapter IV: A Retroaldol IAldol Mechanism  136  Experimental Methods  Preparation of Glycolaldehyde Phosphate Glycolaldehyde phosphate was prepared according to a modification of the method of Miiller et al. (1990).  In this procedure (outlined in Figure 4.8), alfyl alcohol is  phosphorylated and isolated as bis(cyclohexylammonium) allyl phosphate, which is then converted by ozonolysis and exchange of the counterion to calcium glycolaldehyde phosphate. Because this compound is not very soluble in aqueous solutions at neutral p H , calcium was exchanged for sodium, creating a hygroscopic solid.  a) H P 0 Et N CCI3CN, 75°C 3  4  H  3  H  2 HN-  Hl C = ^  +  3  2  CH OH  b) Cyclohexylamine  CH OP0 ~ 2  3  2  a) Dowex AG50W (H ) b) 2 equiv. Et N c) 0 /MeOH, -78°C-*- Me S, -20°C d) Dowex AG50W (H+) +  3  3  1.0 equiv. Ca(OAc) in H 0 , precipitated with acetone  2  2  H  2  C H O P 0 ~ Ca'.2+ 2  3  H CH OP0 H 2  3  2  Figure 4.8. Synthetic Route to Glycolaldehyde Phosphate from Allyl Alcohol.  Chapter IV: A Retroaldol I'Aldol Mechanism  137  A) Bis(cyclohexylammonium) allylphosphate T o a mixtxire of 1.90 g (19.4 mmol) H P 0 3  4  (crystalline) and 4.00 g (39.6 mmol)  triethylamine in 30.00 g (517 mmol) allyl alcohol was added 14.40 g (100 mmol) trichloroacetonitrile dropwise, and stirred for 4 hours at 75 °C.  After distilhng off the  trichloroacetonitrile (75 °C, ca. 15 mm Hg) the mixture was concentrated to 15 m L by rotary evaporation at room temperature, and then mixed with 200 m L water. The aqueous phase was extracted twice with 150 m L diethyl ether and then mixed with 15 m L (130 mmol) cyclohexylamine (Chx) and evaporated at room temperature.  The white powder was  dissolved in 50 m L water and mixed with acetone until it became turbid. After sitting overnight at 4 ° C , the product precipitated as a white amorphous powder. The precipitate was dried under high vacuum (0.3 mm Hg) at ambient temperature. The ' H - , C - , and P 1 3  N M R spectra agreed with those reported by Muller  31  etal (1990). ' H - N M R (200 M H z , D 0 ) : 2  6 1.07-1.87 (cluster of peaks, 20 H , C h x N H ) ; 6 2.99 (m, 2 H , C h x N H ) ; 8 4.11 (tt, J = 4.7, +  +  3  3  1.4 Hz, 2 H , H - C l ) ; 6 5.02 (dd, J = 10.4,1.5 Hz, 1 H , H-C3); 8 5.19 (dd, J = 17.2,1.8 Hz, 1 H , H-C3); 8 5.84 (m, 1 H , H-C2); see Appendix B.  1 3  C - N M R (50 M H z , D 0 ) : 8 24.50 2  ( C H , C h x N H ) ; 6 25.00 ( C H , C h x N H ) ; 8 31.05 ( C H , C h x N H ) ; 8 51.00 (CH, +  2  +  3  2  3  +  2  3  C h x N H ) ; 8 65.92 (CH , Cl); 8 116.71 (CH , C3); 8 136.20 (CH, C2). P - N M R (51 M H z , +  31  3  D 0): 2  8 3.84  2  (s, IP).  2  Chapter IV: A Retroaldol/Aldol Mechanism  138  B) Sodium Glycolaldehyde Phosphate T o convert bis(cyclohexylamrnonium) allyl phosphate to triethylammonium allyl phosphate, 5.4 g (16 mmol) of the former was suspended in 50 m L water and shaken for 30 minutes with 7.5 g washed Dowex A G 5 0 W (H ). The resin was filtered out and washed +  with 150 m L water. The combined filtrates were mixed with 250 m L diethyl ether and 20 m L (144 mmol) triethylamine. The aqueous phase was extracted and evaporated at room temperature and dried under high vacuum (0.3 mm Hg). The resulting oil was ozonolyzed in 130 m L of methanol at -78 ° C until the solution was an intense blue colour. After removal of excess 0  3  with an argon stream, 5.5 m L methyl sulphide (75 mmol; 5 equivalents) was  added, and the colourless mixture was left at -20 ° C for 24 h. This mixture was then mixed with 100 m L ice water and 33 g washed Dowex A G 5 0 W (H ) and stirred for 30 minutes. +  The ion exchanger was filtered out and washed with 400 m L ice water. The combined filtrates were concentrated at room temperature to 50 mL. The resulting mixture was then mixed with a solution of 2.5 g Ca(OAc) (15.0 mmol, 1 equivalent) in 25 m L water. Acetone 2  (50 mL) was dropped slowly into this solution at 4 ° C . This suspension was kept at 4 ° C for at least 14 h, at which time the precipitate was collected by centrifugation.  Additional  precipitate was obtained by mixing the mother liquor with 30 m L of acetone. The combined precipitates were suspended in water and lyophilized. The N M R data were consistent with those presented in Muller etal. (1990). H - N M R (200 M H z , D 0 / d r o p HC1): 8 3.65 (dd, J  2  J= 4.90, 6.87 Hz, 2 H , H-C2), 8 4.97 (t, J= 4.83 Hz, 1 H , H - C l ) ; see Appendix B.  1 3  C-NMR  Chapter IV: A Retroaldol/Aldol Mechanism  139  (50 M H z , D 0 / d r o p HC1): 8 68.84 (d, J = 6.74 Hz, C H , C2), 8 88.81 (d, J = 9.00 Hz, C H , 2  Cl).  3 1  2  P - N M R (51 M H z , D 0 / d r o p HC1): 8 0.00 (s, IP). 2  Calcium glycolaldehyde phosphate (1.0 g) was stirred with a weakly acidic cation exchanger (Amberlite DP-1, previously washed with distilled water) in N a form. The resin +  was filtered out and washed with distilled water. This solution was lyophilized under high vacuum (0.3 mm Hg) to yield a hygroscopic pale yellow solid, and stored at -20 ° C over desiccant. In this form the aldehyde was stable for at least two years as judged by ' H - N M R .  Continuous Assay for Dihydroxyacetone Release Release of dihydroxyacetone was monitored at 37 ° C in a coupled assay containing 50 m M glycylglycine (pH 7.6), 5 U glycerol dehydrogenase (Sigma G-6267 from Bacillus  megaterium, dissolved in water), 0.15 m M N A D H , 0.47 m M LRu5P, 0.1 m M Z n C l , and 1.25 2  mg of 4-epimerase (Zn -reconstituted) in a total of 1 mL. The absorbance of the cuvette 2+  at 340 nm was monitored as a function of time.  Glycolaldehyde Phosphate Release (A Continuous Assay) Release of glycolaldehyde phosphate was monitored at 37 ° C in a coupled assay containing 50 m M glycylglycine (pH 7.6), 5 U a G D H (Sigma G-6751 from rabbit muscle, in an ammonium sulphate and E D T A suspension, diluted with water and concentrated in a Millipore Ultrafree Centrifugal Device (10K N M W L ) to remove ammonium sulphate and E D T A immediately prior to use), 0.15 m M N A D H , 0.47 m M LRu5P, 0.1 m M Z n C l , and 2  Chapter IV: A Retroaldol/Aldol Mechanism  140  1.25 mg of 4-epimerase (Zn -reconstituted) in a total of 1 mL. The absorbance of the 2+  cuvette at 340 nm was monitored as a function of time.  Enzymatic Coupling ofDihydroxyacetone and Glycolaldehyde Phosphate Glycolaldehyde phosphate (5 m M , p H 7.6), dihydroxyacetone (either 20 or 50 mM) and 0.1 m M Z n C l were incubated at 3 7 ° C in a total of 1 m L with 0.5 mg of 4-epimerase 2  (Zn -reconstituted). Aliquots (60 uL) were taken periodically over the course of several 2+  days to assay for sugar content.  Assay for Formation ofKetopentose Phosphates in Coupling Experiments Sugar content was determined by putting a 50 uL aliquot of the coupling reaction into a cuvette (preincubated and monitored at 340 nm for 5 minutes at 37 °C) containing the normal assay components as described in Chapter II . The absorbance at 340 nm was monitored, and the rapid change in absorbance used to calculate the concentration of sugar in the coupling reaction. The time (T) after addition of each aliquot was noted (see Figure 4.9).  The  background rate (collected during the five minute pre-incubation) was used to extrapolate the absorbance (A) at time T. The observed decrease in absorption upon addition of 50 uL of water to cuvettes containing all of the assay components (after a 5 minute incubation at 37°C) was subtracted from A to account for a small dilution of the cuvette contents: this was used as the absorbance (B) immediately after addition of the aliquot. The rate between  Chapter TV: A Retroaldol/Aldol  Mechanism  141  16 and 18 minutes was used to extrapolate the absorbance (C) at time T. The difference in absorption between B and C was proportional to the decrease in N A D H concentration as a result of LRu5P and DXu5P in the aliquot. Therefore this difference could be used to calculate the concentration of these sugars in the assay and therefore also in the aliquot, using the extinction coefficient 8 = 6220 M" cm" for N A D H at 340 nm. 1  I  cu o c CO  1  I  I  1  1  1  1  I  1  I  I  1  I  1  1  0.8  JD  i  o  CO XI  <  0.6  0.4  J I I I L)|J I I 4  T  6  I  I  I  I  8 10 12 Time (min)  F i g u r e 4.9. Determination of Sugar Content.  I  I  14  i  I  16  I 18  I  20  Chapter IV: A Retroaldol IAldol Mechanism  142  Identification ofKetopentose Phosphates in Coupling Experiments Glycolaldehyde phosphate (5 m M , p H 7.6), 50 m M dihydroxyacetone and 0.1 m M Z n C l were incubated at 3 7 ° C in a total of 1 m L with 0.5 mg of H 9 7 N 4-epimerase ( Z n 2+  2  reconstituted). A control containing no enzyme was used. After 48 hours of incubation, the reactions were passed through Millipore Ultrafree Centrifugal Devices (10K N M W L ) and the filtrate was lyophilized. The residues were dissolved in 500 uL D 0 and lyophilized 2  twice, and finally redissolved in 500 uL of D 0 . These samples were analysed by 500 M H z 2  'H-NMR.  The sample which had been catalysed with the H 9 7 N mutant epimerase was  spiked with an authentic sample of the equilibrium mixture (prepared as described in Chapter II) to see if the peaks due to the sugars formed in the aldol addition would increase.  Stopped Assays for Dihydroxyacetone Release LRu5P (10 m M , p H 7.6) and 0.1 m M Z n C l were incubated at 3 7 ° C in a total of 1.0 2  mL with 0.5 mg of wild type or H 9 7 N 4-epimerase ( Z n  2+  reconstituted). Aliquots (50 uL)  were taken periodically over the course of several days to assay for dihydroxyacetone release. A negative control consisting of the supernatant of the heat-inactivated H 9 7 N 4-epimerase was also run. The amount of dihydroxyacetone was determined spectrophotometrically. The 50 uL aliquots were incubated at 37°C in cuvettes containing 25 m M glycylglycine (pH 7.6), and 0.15 m M N A D H . Addition of glycerol dehydrogenase (1 U ; Sigma G-6267 from Bacillus  megaterium, dissolved in water for use) to the cuvette initiated the reaction. The absorbance  ChapterTV: A Retroaldol/Aldol Mechanism  143  at 340 nm was monitored, and the change in absorbance was used to calculate the concentration of dihydroxyacetone in the equilibration of L-ribulose-5-phosphate and D xylulose-5-phosphate in a similar manner to that used in determination of the ketopentose phosphate content of the coupling reactions. The time (T) after addition of each aliquot was noted. The background rate (collected during the five minute preincubation) was used to extrapolate the absorbance (A) at time T. The observed decrease in absorption upon addition of 10 uL of water to cuvettes containing all the assay components (after a 5 minute incubation at 37°C) was subtracted from A to account for a small dilution of the cuvette contents: this was used as the absorbance (B) immediately after addition of the aliquot. The rate between 18 and 20 minutes was used to extrapolate the absorbance (C) at time T. The difference between B and C was proportional to the decrease in N A D H concentration as a result of oxidation of the dihydroxyacetone in the aliquot. This difference was therefore used to calculate the concentration of dihydroxyacetone in the assay cuvette and also in the aliquot using the extinction coefficient for N A D H (8 = 6220 M" cm" ) at 340 nm. 1  1  Assay for Enzyme Activity The activities of the enzymes in the intermediate release incubation mixtures were checked periodically and were determined by monitoring the decrease in N A D H concentration (at A  3 4 0 nm  ) in a coupled reaction in a Cary 3 E UV-VIS spectrophotometer at  37 °C. In this assay, each cuvette contained, in a total volume of 0.5 mL, the components of the normal assay reaction with 0.1 m M Z n C l , as described in Chapter III, and 0.24 m M 2  Chapter IV: A Retroaldol/'Aldol Mechanism  144  L-tibulose-5-phosphate. The reaction was initiated by addition of L-ribulose-5-phosphate 4-epimerase five minutes after beginning observation of the absorbance. The amounts of enzyme used in this assay were 0.025 ug wild type epimerase and 0.05 ug H 9 7 N epimerase.  Assay for Inhibition and Determination ofK  x  Zinc-reconstituted wild type and H 9 7 N enzyme activity was monitored in the usual manner in the presence of 0.1 m M Z n  2 +  at 0.0944 m M LRu5P both with and without 5 m M  glycolaldehyde phosphate, and both with and without 50 m M dihydroxyacetone. In determining K for glycolaldehyde phosphate inhibition of the H 9 7 N enzyme, the x  substrate concentration varied from 0.038 to 0.94 mM; while the glycolaldehyde phosphate (inhibitor) concentration varied from 0 to 2.0 mM. All other assay conditions were the same as those described for determining the enzyme's activity. Lineweaver-Burk plot, and the apparent K  M  Data were analysed using a  values replotted as a function of inhibitor  concentration. The resulting x-intercept provided the K value. The slopes of the lines in t  the Lineweaver-Burk plot were also plotted as a function of inhibitor concentration, the resulting x-intercept again providing the K value. Both K values were identical. x  x  Chapter V  Proposed Exploration of the Active Site of L-Ribulose-5-Phosphate 4-Epimerase  Rationale  In Chapters III and IV, the  Escherichia coli L-ribulose-5-phosphate 4-epimerase was  shown to have both a metal-binding motif and a mechanistic strategy in common with the E. tWz L-fuculose-1 -phosphate aldolase, with which it shares significant amino acid sequence homology (Dreyer & Schulz, 1993; Dreyer & Schulz, 1996b). The similarities between these two enzymes are intriguing because this is the only epimerase currently known to use an aldolase-like mechanism. It is therefore of interest to explore the nature of the active site of the 4-epimerase, and to generate an understanding of how the active site of the 4-epimerase is different from the active site of the aldolase.  145  Chapter V: Active Site Exploration  '  146  Two approaches to this understanding are through use of 1) substrate and intermediate analogues, which may act as alternative substrates, as inhibitors or as active site labelling agents, and 2) site-directed mutagenesis studies aimed at altering the activity and specificity of the epimerase.  Small Molecule Analogues  Stable molecules which mimic the transition state(s) and/or intermediates of a reaction pathway, and for which the enzyme has great affinity, are often bound tightly by the enzyme. Hydroxamic acid derivatives are potent inhibitors of a number of enzymes using a catalytic Z n  2 +  ion, including carbonic anhydrase II (Scolnick et al, 1997), thermolysin  (Holmes & Matthews, 1981; Izquierdo-Martin & Stein, 1992), and matrilysin (Browner etal, 1995).  Hydroxamates are stable analogues of enediolate intermediates of a number of  enzymes (Collins, 1974), and are known to chelate metal ions (Bauer & Exner, 1974). A potent  inhibitor  of  the  phosphoglycolohydroxamate  Class (Fessner  II  dihydroxyacetone-dependent  et al,  1996), which mimics  aldolases the  intermediates of dihydroxyacetone phosphate and bidentately chelates the Z n  2 +  is  enediolate ion in their  active sites. A similar molecule, glycolohydroxamate, could inhibit the 4-epimerase and provide further evidence that this epimerase follows a retroaldol/aldol reaction mechanism.  Chapter V: Active Site Exploration  147  A series of hydroxamate derivatives (Figure 5.1) could be designed to determine the requirements  on  substrates  to  fit  into  the  epimerase's  active  site.  Phosphoglycolohydroxamate itself may inhibit the epimerase.  O^  Y  HO  XH,OPOV  T  ^H  HO  phosphoglycolohydroxamate  HO.  N  ,CH OH  ° < y  3  H  0=  0=  HO-  HO-  HO-  H2  3  L-erythro-4-phosphate-hydroxamate  3  acetohydroxamate  HO.  CH OP0 "  H  HO^ ^ H  glycolohydroxamate  M  C  N  M  -OH CH OP0 2  = 3  D-threo-4-phosphate-hydroxamate  Figure 5.1. Hydroxamate Derivatives Which Could Inhibit L-Ribulose-5-Phosphate 4Epimerase.  Glycolohydroxamate and acetohydroxamate might exhibit synergistic inhibition in combination with glycolaldehyde phosphate; this aldehyde has been observed to inhibit the H97N epimerase competitively. Glycolaldehyde phosphate was found to inhibit the H97N mutant epimerase but not the wild type epimerase (Chapter IV). A consideration in testing the hydroxamate derivatives as inhibitors of L-ribulose-5phosphate 4-epimerase is that phosphoglycolohydroxamate is an inhibitor of triose  Chapter V: Active Site Exploration  148  phosphate isomerase (TIM) (Collins, 1974), an enzyme used in assays of epimerase activity, coupling the epimerization reaction to N A D H oxidation.  CH OH  CH OH  2  =0 H0-  -H  H0-  -H  NaBH,  CH 0P0 ~ 2  CH OH  2  -H  H-  HO-  -H  HO-  -H  HO-  -H  HO-  -H  2  L-ribulose5-phosphate (LRu5P)  CH 0P0 "  =  2  3  L-ribitol5-phosphate  CH 0H  2  -H  H0-  -H  H0-  -H CH 0H 2  L-ribitol (L-adonitol)  3  L-arabitol5-phosphate  CH 0H  HO-  -OH  HO-  CH 0P0  3  2  2  ATP  ADP  L-ribulokinase  HO-  -H  H0-  -H  H0-  -H CH 0P0 = 2  3  L-ribitol5-phosphate  F i g u r e 5.2. Preparation of Pentitol Phosphates as Potential Inhibitors of L-Ribulose-5Phosphate 4-Epimerase.  Another potential inhibitor of the epimerase is L-ribitol-5-phosphate. compound  This  can be prepared as a mixture of L-ribitol-5-phosphate and L-arabitol-5-  phosphate by N a B H reduction of L-ribulose-5-phosphate in a buffered solution (Figure 5.2) 4  (London & Hausman, 1982). Alternatively, L-ribitol-5-phosphate may be prepared by  Chapter V: Active Site Exploration  149  selective phosphorylation of ribitol by L-ribulokinase (Simpson & Wood, 1958) in a manner similar to L-ribulokinase mediated phosphorylation of L-ribulose (Anderson, 1966). Glycolic acid or glycolate may inhibit the epimerase. It might also be able to add to glycidol phosphate to form an ester (Figure 5.3). Synergistic inhibition is possible and should be investigated.  glycolate  0  H  CH 0H 2  =0  .0^0  o I  ^  CH 0  CH OP0 " 2  2  I CHOH CH OP0 " 2  3  3  glycidol phosphate  Figure 5.3. Addition of Glycolate to Glycidol Phosphate.  Inactivating Agents  Irreversible inhibitors are active site labelling agents which react with catalytic residues. Therefore they are useful in combination with proteolytic degradation and peptide purification in identifying which amino acid residue catalyses the reaction. The epimerase  Chapter V: Active Site Exploration  150  may be inhibited or inactivated by glycidol phosphate.  This molecule is an active site  labelling agent for T I M (Rose & O'Connell, 1969; Schray et al., 1973).  Haloacetol  phosphates (Figure 5.4) are electrophiles which are known to inhibit T I M irreversibly (Coulson  etal, 1970; de la Mare etal., 1971; Hartman, 1971; Norton & Hartman, 1972) in  addition to inactivating yeast aldolase (a Class II aldolase) (Hartman, 1970; Lin et al, 1971). Another potential inactivator of L-ribulose-5-phosphate 4-epimerase is phosphoglycolic acid chloride, which could be synergistic with dihydroxyacetone or glycolate.  The anhydride  shown in (Figure 5.4) is another potential active site labelling agent.  CH OH  Cl  2  O  CH OP0 2  o CH OP0 "  3  2  glycidol phosphate  chloroacetol phosphate  phosphoglycolic acid chloride  F i g u r e 5.4. Potential Active Site Labelling Agents.  3  phosphoglycolic anhydride  Chapter V: Active Site Exploration  151  Epimerase Structure and Site-Directed Mutagenesis Studies  In order to make rational decisions about which residues to alter and with what to replace them, an X-ray crystal structure of the epimerase is required. The epimerase has been crystallized and found to diffract X-rays (Andersson etal, 1995), the preliminary steps toward obtaining a crystal structure. In addition, a collaboration with Dr. Natalie Strynadka (Department of Biochemistry, University of British Columbia) has been established to determine the crystal structure of the wild type L-ribulose-5-phosphate 4-epimerase and eventually of the mutant 4-epimerases.  In addition, the free and inhibitor-complexed  structures could provide information about the residues involved in epimerisation and their roles. The D 7 6 N 4-epimerase was observed to retain the ability to bind the metal ion (to a similar degree as the wild type enzyme; see Chapter III) and also had some epimerase activity. It would be interesting to find out what the fourth ligand to the metal ion is in this mutant (instead of aspartate). If there is sufficient room in the active site, the activity of this mutant could be "rescued" by inclusion of formate in the assay buffer. The corresponding residue in L-fuculose-l-phosphate aldolase is glutamate. A D 7 6 E mutation may make the epimerase more like the aldolase, especially in combination with a H 9 7 N mutation. The ultimate goal of these studies is to determine which residues cause the active sites of L-ribulose-5-phophate 4-epimerase and L-fuculose-l-phosphate aldolase to catalyse different reactions. Eventually this could lead to the ability to alter enough residues in the  Chapter V: Active Site Exploration  152  epimerase's active site that aldolase activity can be optimized. A comparison of the crystal structures of the aldolase (this is known: Dreyer & Schulz, 1993; Dreyer & Schulz, 1996a; Dreyer & Schulz, 1996b) and of the epimerase should unambiguously reveal which residues define the active site specificity and chemistry.  Related Enzymes  Recent bacterial genome sequencing projects have uncovered a number of sequences which have amino acid sequence homology to both the  E. coli L-ribulose-5-phosphate 4-  epimerase and the E. ^//L-fuculose-l-phosphate aldolase (see Appendix C). These enzymes appear to have the same divalent metal-binding motif as the epimerase and aldolase, and likely use a similar mechanistic strategy. Neither their substates nor their metabolic roles are known. Studies on these enzymes to identify their metal binding ligands and to discover their mechanisms of catalysis could strengthen and provide further insight into the relationship between L-ribulose-5-phosphate 4-epimerase aldolase.  and L-fuculose-l-phosphate  Chapter V: Active Site Exploration Conclusion  The catalytic bases used by L-ribulose-5-phosphate 4-epimerase to catalyse the interconversion of L-ribulose-5-phosphate and D-xylulose-5-phosphate are not known. Possible future studies could be performed and are outlined in this chapter. A combination of the active-site labelling agents and site-directed mutagenesis studies presented here will be able to identify these bases. X-ray crystallography on the epimerase should verify the identity of these residues as well as show the similarities and differences between the active sites of L-fuculose-l-phosphate aldolase and L-ribulose-5-phosphate 4-epimerase.  153  Appendix A D N A Sequences  AraD of Escherichia coli K12 M  L  E  D  ATG  T T A GAA GAT  TAC  AAT  L  V  CTT T  L  K  R  CTC AAA  C T A GAG T T T L  T  Q  V  CAT AAT  G  V  W  N  G T C A C G C T C A C A T G G GGC A A C  GAC  CAG TGC GAG TGT A C C K  P  S  G  V  E  GCG GTC  CTG  I  L  A  N  L  CGC CAG GTA T T A GAA GCC AAC  S  CTT A  L  CGG T T G GAC CGC V  D  R  E  R  G  S  V  M  T  A  D  D  V  TAG  TTT  GGA AGG CCG CAG C T A A T G T C G CAG TAC TGG CGA CTG C T A T A C  E  T  TAG  CTT  R  L  G  E  V  GGT  GAA  GTG GTT  TGG CCA CTT L  Y  CTG CTC  GCC  GAC GAG A T A GTC  A  T  I  TAT  W  GCC  ACC ATC  CGG  TGG TAG ACC  Y  F  Y  CAC  Q  CGG  V  TAC AGC GTC A T G A C C GCT  M  CCT  GAA A C C  E  G  K  F  S  P  I  G  CCC TCC ATT  A  G  Q  I  P  S  CCA GTC AGC  T  ATA  CTC ACC  A  Q  M  CTT P  H  C  T  R  G  N  V  GGT A A C  V  A  K  T  '  M  T  D  A  I  V  GTC ATC  E  L  V  H  F  E  K  CTT  C T A CGC  R  Q  V  H  N  GCG GTG CAT A A C  L  S  A  I  CAT  CTT  H  TGG AAA  G  V  TCC  P  Q  G  H  T  E  P  F  L  Q  S  S  T  E  L  P  D  M  CTT Q  V  I  N  A  A  CTT G  H  100 CAC  A  D  G  E  W  G  T  L  I  D  A  160  GAT  Y  M  G  N  A  D  K  CCG TTT I  F  Y  200 TGC  L  220  GCA  GTC AAT  GGC  GTG A T A  GAC  K  H  G  A  K  A  Y  Y  AAG CAT  GGC GCG AAG GCA TAT  GCA  TTC GTA  CCG CGC T T C  G  Q  600  AAG ACG  CTG  TGC GAC GAC C T A T T T  540  T T A CGG C  CCC TAT H  GCC  CAC TAT  R  4 80  180  T A T A T G GGG A T A T T C  L  GCA  CCA TAG CTA CGT K  CGT A C C  4 20  TAG TTG CCG CTT  CCG GAT A T G CAG CAA ACG CTG CTG GAT AAA  CGT  3 60  140  GGC GAA  CAG T T A GCG CCG CAG T T A  CTA TAC GTC GTT  300  120  CAC GCC GAC  CGT  CGC GGC GTC AAT  240  GCG GTG  G C A T G G GGC A A A A A T  CTC CAG CGA A T A TAC Q  CAC  TGA GTG  R  H  180  80  CAC TCG CGC  GTC  TTT  CAC GGC CCG T T T  E  60  H  CTT  GCC A T C GTG C T G GAA GAG GTC GCT  CAC GTA T T G CGG T A G CAC GAC A  T  CAG GGT A T C  GGG C C G C A A GAC C A G G T A A G G G T G C C G GGC A A A  A  P  GAA AAA  GTT D  CTG GTC  T  GTA GAA ACC T T T  CGC  GAT  40 120  CAA TCG  CAG T G C GGT  T  TAC TGG CTG CGT  CCC GGC GTT  E  CAG CAC  T  60  CGT TGG CCG TGG TGG GTG CGG CTG  CAA ATG  GAA  D  GTG CAT ACG  GCG  TAC  V  CCC TGC ACC CGC AAA A T G ACC GAC GCA GAA A T C AAC  TGG CCA T T G CAG TAG CAT G  V  C C A GCA A C C GGC A C C A C C  T A A GGT  C C G T G G T A A GGG A C G T G G GCG T T T E  S  V  P  AAG ATG W  I  I  ATA  E  S  T T C GGG A G G AGG  CAG TCG ATT  GGC A C C A T T  GAG TGG GAA A C C  P  G  TTC TAC  Y  V  GAC GAT A T G GTC GTG GTT AGC  GGC GGC A T T  TAT  TAT  F  CGT A A G GGG A G G T A A C C G C C G T A A C A C G T A T G C G T G A G C  CGC GTC CGC T  K  CCA TGC TTT  T G G G C G C A G GCG GGT  F  T  CAA CTT  A  Q  20  GTG TTG  GAA GGT A C G A A A A A G C C C T C C T C C GAC A C G C C A A C T  CAG GCA T T C  A  TTT  V  N CAC AAC  CAC  AAA  I  H  CCG T T G CAG T C G CGG CAA C T A GCG C T C GCG CCG CAG A A A Y  GAT  GAC GGT  ATC  ATC  K  GTG  T C C GGC GTC GAT  GCC GTT  P  CGC GAG CGC GGC G T C T T T  D  GTC AGC  A  CTG GCG CTG CCA A A A  660  END  231  T A C GGG CAG T A A  696  CGT A T A A T G CCC GTC A T T  154  Appendix A: DNA Sequences  155  AraB of Escherichia coli ATGGCGATTG CAATTGGCCT CGATTTTGGC AGTGATTCTG TGCGAGCTTT GGCGGTGGAC TACCGCTAAC GTTAACCGGA GCTAAAACCG TCACTAAGAC ACGCTCGAAA CCGCCACCTG  60  TGCGCCAGCG GTGAAGAGAT CGCCACCAGC GTAGAGTGGT ATCCCCGTTG GCAAAAAGGG ACGCGGTCGC CACTTCTCTA GCGGTGGTCG CATCTCACCA TAGGGGCAAC CGTTTTTCCC  120  CAATTTTGTG ATGCCCCGAA TAACCAGTTC CGTCATCATC CGCGTGACTA CATTGAGTCA GTTAAAACAC TACGGGGCTT ATTGGTCAAG GCAGTAGTAG GCGCACTGAT GTAACTCAGT  180  ATGGAAGCGG CACTGAAAAC CGTGCTTGCA GAGCTTAGCG TCGAACAGCG CGCAGCTGTG TACCTTCGCC GTGACTTTTG GCACGAACGT CTCGAATCGC AGCTTGTCGC GCGTCGACAC  240  GTCGGGATTG GCGTTGACAG TACCGGCTCG ACGCCCGCAC CGATTGATGC CGACGGTAAC CAGCCCTAAC CGCAACTGTC ATGGCCGAGC TGCGGGCGTG GCTAACTACG GCTGCCATTG  300  GTGCTGGCGC TGCGCCCGGA GTTTGCCGAA AACCCGAACG CGATGTTCGT ATTGTGGAAA CACGACCGCG ACGCGGGCCT CAAACGGCTT TTGGGCTTGC GCTACAAGCA TAACACCTTT  360  GACCACACTG CGGTTGAAAG AAGCGAAGAG ATTACCCGTT TGTGCCACGC GCCGGGCAAT CTGGTGTGAC GCCAACTTTC TTCGCTTCTC TAATGGGCAA ACACGGTGCG CGGCCCGTTA  4 20  GTTGACTACT CCCGCTATAT TGGCGGTATT TATTCCAGCG AATGGTTCTG GGCAAAAATC CAACTGATGA GGGCGATATA ACCGCCATAA ATAAGGTCGC TTACCAAGAC CCGTTTTTAG  4 80  CTGCATGTGA CTCGCCAGGA CAGCGCCGTG GCGCAATCTG CCGCATCGTG GATTGAGCTG GACGTACACT GAGCGGTCCT GTCGCGGCAC CGCGTTAGAC GGCGTAGCAC CTAACTCGAC  54 0  TGCGACTGGG TGCCAGCTCT GCTTTCCGGT ACCACCCGCC CGCAGGATAT TCGTCGCGGA ACGCTGACCC ACGGTCGAGA CGAAAGGCCA TGGTGGGCGG GCGTCCTATA AGCAGCGCCT  600  CGTTGCAGCG CCGGGCATAA ATCTCTGTGG CACGAAAGCT GGGGCGGCTT GCCGCCAGCC GCAACGTCGC GGCCCGTATT TAGAGACACC GTGCTTTCGA CCCCGCCGAA CGGCGGTCGG  660  AGTTTCTTTG ATGAGCTGGA CCCGATCCTC AATCGCCATT TGCCTTCCCC GCTGTTCACT TCAAAGAAAC TACTCGACCT GGGCTAGGAG TTAGCGGTAA ACGGAAGGGG CGACAAGTGA  720  GACACCTGGA CTGCCGATAT TCCGGTGGGC ACCTTATGCC CGGAATGGGC GCAGCGTCTC CTGTGGACCT GACGGCTATA AGGCCACCCG TGGAATACGG GCCTTACCCG CGTCGCAGAG  7 80  Nde  I  GGCCTGCCTG AAAGCGTGGT GATTTCCGGC GGCGCGTTTG ACTGCCATAT GGGCGCAGTT CCGGACGGAC TTTCGCACCA CTAAAGGCCG CCGCGCAAAC TGACGGTATA CCCGCGTCAA  84 0  GGCGCAGGCG CACAGCCTAA CGCACTGGTA AAAGTTATCG GTACTTCCAC CTGCGACATT CCGCGTCCGC GTGTCGGATT GCGTGACCAT TTTCAATAGC CATGAAGGTG GACGCTGTAA  900  CTGATTGCCG ACAAACAGAG CGTTGGCGAG CGGGCAGTTA AAGGTATTTG CGGTCAGGTT GACTAACGGC TGTTTGTCTC GCAACCGCTC GCCCGTCAAT TTCCATAAAC GCCAGTCCAA  960  GATGGCAGCG TGGTGCCTGG ATTTATCGGT CTGGAAGCAG GCCAATCGGC GTTTGGTGAT CTACCGTCGC ACCACGGACC TAAATAGCCA GACCTTCGTC CGGTTAGCCG CAAACCACTA  1020  ATCTACGCCT GGTTCGGTCG CGTACTCAGC TGGCCGCTGG AACAGCTTGC CGCCCAGCAT TAGATGCGGA CCAAGCCAGC GCATGAGTCG ACCGGCGACC TTGTCGAACG GCGGGTCGTA  1080  Appendix A: DNA Sequences  156  CCGGAACTGA AAGCGCAAAT CAACGCCAGC CAGAAACAAC TGCTTCCGGC GCTGACCGAA GGCCTTGACT TTCGCGTTTA GTTGCGGTCG GTCTTTGTTG ACGAAGGCCG CGACTGGCTT  1140  GCATGGGCCA AAAATCCGTC TCTGGATCAC CTGCCGGTGG TGCTCGACTG GTTTAACGGT CGTACCCGGT TTTTAGGCAG AGACCTAGTG GACGGCCACC ACGAGCTGAC CAAATTGCCA  1200  CGTCGCTCGC CAAACGCTAA CCAACGCCTG AAAGGGGTGA TTACCGATCT TAACCTCGCT GCAGCGAGCG GTTTGCGATT GGTTGCGGAC TTTCCCCACT AATGGCTAGA ATTGGAGCGA  12 60  ACCGACGCTC CGCTGCTGTT CGGCGGTTTG ATTGCTGCCA CCGCCTTTGG CGCACGCGCA TGGCTGCGAG GCGACGACAA GCCGCCAAAC TAACGACGGT GGCGGAAACC GCGTGCGCGT  1320  ATCATGGAGT GCTTTACCGA TCAGGGGATC GCCGTCAATA ACGTGATGGC GCTGGGCGGC TAGTACCTCA GCAAATGGCT AGTCCCCTAG CGGCAGTTAT TGCACTACCG CGACCCGCCG  1380  ATCGCGCGGA AAAACCAAGT CATTATGCAG GCCTGCTGCG ACGTGCTGAA TCGCCCGCTG TAGCGCGCCT TTTTGGTTCA GTAATACGTC CGGACGACGC TGCACGACTT AGCGGGCGAC  14 4 0  CAAATTGTTG CCTCTGACCA GTGCTGTGCG CTCGGTGCGG CGATTTTTGC TGCCGTCGCC GTTTAACAAC GGAGACTGGT CACGACACGC GAGCCACGCC GCTAAAAACG ACGGCAGCGG  1500  GCGAAAGTGC ACGCAGACAT CCCATCAGCC CAGCAAAAAA TGGCCAGTGC GGTAGAGAAA CGCTTTCACG TGCGTCTGTA GGGTAGTCGG GTCGTTTTTT ACCGGTCACG CCATCTCTTT  15 60  ACCCTGCAAC CGCGCAGCGA ACAGGCACAA CGCTTTGAAC AGCTTTATCG CCGCTATCAG TGGGACGTTG GCGCGTCGCT TGTCCGTGTT GCGAAACTTG TCGAAATAGC GGCGATAGTC  1620  CAATGGGCGA TGAGCGCCGA ACAACACTAT CTTCCAACTT CCGCCCCGGC ACAGGCTGCC GTTACCCGCT ACTCGCGGCT TGTTGTGATA GAAGGTTGAA GGCGGGGCCG TGTCCGACGG  168 0  CAGGCCGTTG CGACTCTATA A GTCCGGCAAC GCTGAGATAT T  17 01  Appendix A.: DNA Sequences  157  Sequence of p R E l TCGCGCGTTT CGGTGATGAC GGTGAAAACC TCTGACACAT GCAGCTCCCG GAGACGGTCA AGCGCGCAAA GCCACTACTG CCACTTTTGG AGACTGTGTA CGTCGAGGGC CTCTGCCAGT  60  CAGCTTGTCT GTAAGCGGAT GCCGGGAGCA GACAAGCCCG TCAGGGCGCG TCAGCGGGTG GTCGAACAGA CATTCGCCTA CGGCCCTCGT CTGTTCGGGC AGTCCCGCGC AGTCGCCCAC  120  TTGGCGGGTG TCGGGGCTGG CTTAACTATG CGGCATCAGA GCAGATTGTA CTGAGAGTGC AACCGCCCAC AGCCCCGACC GAATTGATAC GCCGTAGTCT CGTCTAACAT GACTCTCACG  180  ACCATGTGCG GTGTGAAATA CCGCACAGAT GCGTAAGGAG AAAATACCGC ATCAGGCGAA TGGTACACGC CACACTTTAT GGCGTGTCTA CGCATTCCTC TTTTATGGCG TAGTCCGCTT  24 0  ATTGTAAACG TTAATATTTT GTTAAAATTC GCGTTAAATT TTTGTTAAAT CAGCTCATTT TAACATTTGC AATTATAAAA CAATTTTAAG CGCAATTTAA AAACAATTTA GTCGAGTAAA  300  TTTAACCAAT AGGCCGAAAT CGGCAAAATC CCTTATAAAT CAAAAGAATA GACCGAGATA AAATTGGTTA TCCGGCTTTA GCCGTTTTAG GGAATATTTA GTTTTCTTAT CTGGCTCTAT  360  GGGTTGAGTG TTGTTCCAGT TTGGAACAAG AGTCCACTAT TAAAGAACGT GGACTCCAAC CCCAACTCAC AACAAGGTCA AACCTTGTTC TCAGGTGATA ATTTCTTGCA CCTGAGGTTG  420  GTCAAAGGGC GAAAAACCGT CTATCAGGGC GATGGCCCAC TACGTGAACC ATCACCCTAA CAGTTTCCCG CTTTTTGGCA GATAGTCCCG CTACCGGGTG ATGCACTTGG TAGTGGGATT  4 80  TCAAGTTTTT TGGGGTCGAG GTGCCGTAAA GCACTAAATC GGAACCCTAA AGGGAGCCCC AGTTGAAAAA ACCCCAGCTC CACGGCATTT CGTGATTTAG CCTTGGGATT TCCCTCGGGG  54 0  CGATTTAGAG CTTGACGGGG AAAGCCGGCG AACGTGGCGA GAAAGGAAGG GAAGAAAGCG GCTAAATCTC GAACTGCCCC TTTCGGCCGC TTGCACCGCT CTTTCCTTCC CTTCTTTCGC  600  AAAGGAGCGG GCGCTAGGGC GCTGGCAAGT GTAGCGGTCA CGCTGCGCGT AACCACCACA TTTCCTCGAA CGCGATCCCG CGACCGTTCA CATCACCAGT GCGACGCGCA TTGGTGGTGT  660  CCCGCCGCGC TTAATGCGCC GCTACAGGGC GCGTCGCGCC ATTCGCCATT CAGGCTGCGC GGGCGGCGCG AATTACGCGG CGATGTCCCG CGCAGCGCGG TAAGCGGTAA GTCCTACGCG  720  Pst  I  AACTGTTGGG AAGGGCGATC GGTGCGGGCC TCTTCGCTAT TACGCCAGCT GCAGTTACTG TTGACAACCC TTCCCGCTAG CCACGCCCGG AGAAGCGATA ATGCGGTCGA CGTCAATGAC  780  CCCGTAATAT GCCAACGCGC CTAGCAATCG CAGATAGTGT TTATCCAGCA GCGTTTGCTG GGGCATTATA CGGTTGCGCG GATCGTTAGC GTCTATCACA AATAGGTCGT CGCAAACGAC  84 0  CATATCCGGT AACTGCGGCG CTAACTGACG GCAGAATATC CCCATATAAG CGACCTCTTC GTATAGGCCA TTGACGCCGC GATTGACTGC CGTCTTATAG GGGTATATTC GCTGGAGAAG  900  CAGCACGATG GCGTTATGCA CCGCATCTTC GGCATTTTTG CCCCATGCAA ACGGGCCGTG GTCGTGCTAC CGCAATACGT GGCGTAGAAG CCGTAAAAAC GGGGTACGTT TGCCCGGCAC  960  GGAATGGACC AGAACGCCGG GCATTTGCGC TGCATCGATA CCCTGTTTTT CAAAGGTTTC CCTTACCTGG TCTTGCGGCC CGTAAACGCG ACGTAGCTAT GGGACAAAAA GTTTCCAAAG  102 0  TACGATGACG TTACCGGTTT CCCACTCATA TTCGCCGTTG ATTTCTGCGT CGGTCATTTT ATGCTACTGC AATGGCCAAA GGGTGAGTAT AAGCGGCAAC TAAAGACGCA GCCAGTAAAA  1080  Appendix A.: DNA Sequences  158  GCGGGTGCAG GGAATGGTGC CGTAGAAATA GTCGGCGTGG GTGGTGCCGG TTGCTGGAAT CGCCCACGTC CCTTACCACG GCATCTTTAT CAGCCGCACC CACCACGGCC AACGACCTTA  114 0  CGACTGACCC GCCTGCGCCC AGATGGTGGC GTGGCGCGAG TGCGTATGCA CAATGCCGCC GCTGACTGGG CGGACGCGGG TCTACCACCG CACCGCGCTC ACGCATACGT GTTACGGCGG  1200  AATGGAGGGG AATGCCTGAT AGAGCAGCCG GTGAGTTGGC GTGACGGAGG AGGGCTTTTT TTACCTCCCC TTACGGACTA TCTCGTCGGC CACTCAACCG CACTGCCTCC TCCCGAAAAA  1260  CGTACCTTCA ACCACTTCAC CGGTTTCGAT GCTAACCACG ACCATATCGT CAGCGGTCAT GCATGGAAGT AGGTGAAGTG GCCAAAGCTA CGATTGGTGC TGGTATAGCA GTCGCCAGTA  1320  GACGCTGTAA TCGACGCCGG AAGGTTTGAT CACAAAGACG CCGCGCTCGC GATCAACGGC CTGCGACATT AGCTGCGGCC TTCCAAACTA GTGTTTCTGC GGCGCGAGCG CTAGTTGCCG  1380  GCTGACGTTG CCCCATGTGA GCGTGACCAG GTTGTGTTTT GGCAGCGCCA GGTTGGCTTC CGACTGCAAC GGGGTACACT CGCACTGGTC CAACACAAAA CCGTCGCGGT CCAACCGAAG  14 4 0  Nde I S D TAATACCTGG CGTTTGAGAT CTTCTAACAT ATGCTGTTTC CTGTGTGAAA TTGTTATCCG ATTATGGACC GCAAACTCTA GAAGATTGTA TACGACAAAG GACACACTTT AACAATAGGC  1500  Pribnow ( t r c promoter) -35 CTCACAATTC CACACATTAT ACGAGCCGGA TGATTAATTG TCAACAGCTC ATTTCAGAAT GAGTGTTAAG GTGTGTAATA TGCTCGGCCT ACTAATTAAC AGTTGTCGAG TAAAGTCTTA  15 60  ATTTGCCAGA ACCGTTATGA TGTCGGCGCA AAAAACATTA TCCAGAACGG GAGTGCGCCT TAAACGGTCT TGGCAATACT ACAGCCGCGT TTTTTGTAAT AGGTCTTGCC CTCACGCGGA  1620  TGAGCGACAC GAATTATGCA GTGATTTACG ACCTGCACAG CCAATCCACA GCTTCCGATG ACTCGCTGTG CTTAATACGT CACTAAATGC TGGACGTGTC GGTTAGGTGT CGAAGGCTAC  168 0  GCTGCCTGAC GCCAGAAGCA TTGGTGCACC GTGCAGTCGA TGATAAGCTG TCAAACATGA CGACGGACTG CGGTCTTCGT AACCACGTGG CACGTCAGCT ACTATTCGAC AGTTTGTACT  17 4 0  EcoR I GAATTCGGCG CTCTTCCGCT TCCTCGCTCA CTGACTCGCT GCGCTCGGTC GTTCGGCTGC CTTAAGCCGC GAGAAGGCGA AGGAGCGAGT'GACTGAGCGA CGCGAGCCAG CAAGCCGACG  1800  GGCGAGCGGT ATCAGCTCAC TCAAAGGCGG TAATACGGTT ATCCACAGAA TCAGGGGATA CCGCTCGCCA TAGTCGAGTG AGTTTCCGCC ATTATGCCAA TAGGTGTCTT AGTCCCCTAT  18 60  ACGCAGGAAA GAACATGTGA GCAAAAGGCC AGCAAAAGGC CAGGAACCGT AAAAAGGCCG TGCGTCCTTT CTTGTACACT CGTTTTCCGG TCGTTTTCCG GTCCTTGGCA TTTTTCCGGC  1920  CGTTGCTGGC GTTTTTCCAT AGGCTCCGCC CCCCTGACGA GCATCACAAA AATCGACGCT GCAACGACCG CAAAAAGGTA TCCGAGGCGG GGGGACTGCT CGTAGTGTTT TTAGCTGCGA  1980  CAAGTCAGAG GTGGCGAAAC CCGACAGGAC TATAAAGATA CCAGGCGTTT CCCCCTGGAA GTTCAGTCTC CACCGCTTTG GGCTGTCCTG ATATTTCTAT GGTCCGCAAA GGGGGACCTT  204 0  GCTCCCTCGT GCGCTCTCCT GTTCCGACCC TGCCGCTTAC CGGATACCTG TCCGCCTTTC CGAGGGAGCA CGCGAGAGGA CAAGGCTGGG ACGGCGAATG GCCTATGGAC AGGCGGAAAG  2100  TCCCTTCGGG AAGCGTGGCG CTTTCTCATA GCTCACGCTG TAGGTATCTC AGTTCGGTGT AGGGAAGCCC TTCGCACCGC GAAAGAGTAT CGAGTGCGAC ATCCATAGAG TCAAGCCACA  2160  Appendix A.: DNA Sequences  159  AGGTCGTTCG CTCCAAGCTG GGCTGTGTGC ACGAACCCCC CGTTCAGCCC GACCGCTGCG TCCAGCAAGC GAGGTTCGAC C.GGACACACG TGCTTGGGGG GCAAGTCGGG CTGGCGACGC  2220  CCTTATCCGG TAACTATCGT CTTGAGTCCA ACCCGGTAAG ACACGACTTA TCGCCACTGG GGAATAGGCC ATTGATAGCA GAACTCAGGT TGGGCCATTC TGTGCTGAAT AGCGGTGACC  2280  CAGCAGCCAC TGGTAACAGG ATTAGCAGAG CGAGGTATGT AGGCGGTGCT ACAGAGTTCT GTCGTCGGTG ACCATTGTCC TAATCGTCTC GCTCCATACA TCCGCCACGA TGTCTCAAGA  234 0  TGAAGTGGTG GCCTAACTAC GGCTACACTA GAAGGACAGT ATTTGGTATC TGCGCTCTGC ACTTCACCAC CGGATTGATG CCGATGTGAT CTTCCTGTCA TAAACCATAG ACGCGAGACG  24 00  TGAAGCCAGT TACCTTCGGA AAAAGAGTTG GTAGCTCTTG ATCCGGCAAA CAAACCACCG ACTTCGGTCA ATGGAAGCCT TTTTCTCAAC CATCGAGAAC TAGGCCGTTT GTTTGGTGGC  24 60  CTGGTAGCGG TGGTTTTTTT GTTTGCAAGC AGCAGATTAC GCGCAGAAAA AAAGGATCTC GACCATCGCC ACCAAAAAAA CAAACGTTCG TCGTCTAATG CGCGTCTTTT TTTCCTAGAG  2520  AAGAAGATCC TTTGATCTTT TCTACGGGGT CTGACGCTCA GTGGAACGAA AACTCACGTT TTCTTCTAGG AAACTAGAAA AGATGCCCCA GACTGCGAGT CACCTTGCTT TTGAGTGCAA  258 0  AAGGGATTTT GGTCATGAGA TTATCAAAAA GGATCTTCAC CTAGATCCTT TTAAATTAAA TTCCCTAAAA CCAGTACTCT AATAGTTTTT CCTAGAAGTG GATCTAGGAA AATTTAATTT  2 64 0  AATGAAGTTT TAAATCAATC TAAAGTATAT ATGAGTAAAC TTGGTCTGAC AGTTACCAAT TTACAACAAA ATTTAGTTAG ATTTCATATA TACTCATTTG AACCAGACTG TCAATGGTTA  27 00  GCTTAATCAG TGAGGCACCT ATCTCAGCGA TCTGTCTATT TCGTTCATCC ATAGTTGCCT CGAATTAGTC ACTCCGTGGA TAGAGTCGCT AGACAGATAA AGCAAGTAGG TATCAACGGA  27 60  GACTCCCCGT CGTGTAGATA ACTACGATAC GGGAGGGCTT ACCATCTGGC CCCAGTGCTG CTGAGGGGCA GCACATCTAT TGATGCTATG CCCTCCCGAA TGGTAGACCG GGGTCACGAC  282 0  CAATGATACC GCGAGACCCA CGCTCACCGG CTCCAGATTT ATCAGCAATA AACCAGCCAG GTTACTATGG CGCTCAGGGT GCGAGTGGCC GAGGTCTAAA TAGTCGTTAT TTGGTCGGTC  2880  CCGGAAGGGC CGAGCGCAGA AGTGGTCCTG CAACTTTATC CGCCTCCATC CAGTCTATTA GGCCTTCCCG GCTCGCGTCT TCACCAGGAC GTTGAAATAG GCGGAGGTAG GTCAGATAAT  2 94 0  ATTGTTGCCG GGAAGCTAGA GTAAGTAGTT CGCCAGTTAA TAGTTTGCGC AACGTTGTTG TAACAACGGC CCTTCGATCT CATTCATCAA GCGGTCAATT ATCAAACGCG TTGCAACAAC  3000  CCATTGCTAC AGGCATCGTG GTGTCACGCT CGTCGTTTGG TATGGCTTCA TTCAGCTCCG GGTAACGATG TCCGTAGCAC CACAGAGCGA GCAGCAAACC ATACCGAAGT AAGTCGAGGC  3060  GTTCCCAACG ATCAAGGCGA GTTACATGAT CCCCCATGTT GTGCAAAAAA GCGGTTAGCT CAAGGGTTGC TAGTTCCGCT CAATGTACTA GGGGGTACAA CACGTTTTTT CGCGAATCGA  3120  CCTTCGGTCC TCCGATCGTT GTCAGAAGTA AGTTGGCCGC AGTGTTATCA CTCATGGTTA GGAAGCCAGG AGGCTAGCAA CAGTCTTCAT TCAACCGGCG TCACAATAGT GAGTACCAAT  3180  TGGCAGCACT GCATAATTCT CTTACTGTCA TGCCATCCGT AAGATGCTTT TCTGTGACTG ACCGTCGTGA CGTATTAAGA GAATGACAGT ACGGTAGGCA TTCTACGAAA AGACACTGAC  324 0  GTGAGTACTC AACCAAGTCA TTCTGAGAAT AGTGTATGCG GCGACCGAGT TGCTCTTGCC CACTCATGAG TTGGTTCAGT AAGACTCTTA TCACATACGC CGCTGGCTCA ACGAGAACGG  3300  Appendix A.: DNA Sequences  160  CGGCGTCAAT ACGGGATAAT ACCGCGCCAC ATAGCAGAAC TTTAAAAGTG CTCATCATTG GCCGCAGTTA TGCCCTATTA TGGCGCGGTG TATCGTCTTG AAATTTTCAC GAGTAGTAAC  3360  GAAAACGTTC TTCGGGGCGA AAACTCTCAA GGATCTTACC GCTGTTGAGA TCCAGTTCGA CTTTTGCAAG AAGCCCCGCT TTTGAGAGTT CCTAGAATGG CGACAACTCT AGGTCAAGCT  3420  TGTAACCCAC TCGTGCACCC AACTGATCTT CAGCATCTTT TACTTTCACC AGCGTTTCTG ACATTGGGTG AGCACGTGGG TTGACTAGAA GTCGTAGAAA ATGAAAGTGG TCGCAAAGAC  3480  GGTGAGCAAA AACAG GAAG G CAAAATGCCG CAAAAAAGGG AATAAGGGCG ACACGGAAAT CCACTCGTTT TTGTCCTTCC GTTTTACGGC GTTTTTTCCC TTATTCCCGC TGTGCCTTTA  3540  GTTGAATACT CATACTCTTC CTTTTTCAAT ATTATTGAAG CATTTATCAG GGTTATTGTC CAACTTATGA GTATGAGAAG GAAAAAGTTA TAATAACTTC GTAAATAGTC CCAATAACAG  3600  TCATGAGCGG ATACATATTT GAATGTATTT AGAAAAATAA ACAAATAGGG GTTCCGCGCA AGTACTCGCC TATGTATAAA CTTACATAAA TCTTTTTATT TGTTTATCCC CAAGGCGCGT  3660  CATTTCCCCG AAAAGTGCCA CCTGACGTCT AAGAAAC CAT TATTATCATG ACATTAACCT GTAAAGGGGC TTTTCACGGT GGACTGCAGA TTCTTTGGTA ATAATAGTAC TGTAATTGGA  3720  ATAAAAATAG GCGTATCACG AGGCCCTTTC GTC TATTTTTATC CGCATAGTGC TCCGGGAAAG CAG  3753  This sequence was assembled from information about pBS provided by Stratagene (and annotated by members of the Knowles lab), araD sequence (nt 785 to 1470), and pKK233-2 sequence (nt 1471-1740) (from Genebank). S o m e U n i q u e R e s t r i c t i o n Sites o n p R E l  Aatll Bell Bglll Clal EcoKl Ndel Nrul Pstl Pvull Sea I  3788 1375 1475 995 1741 1469 1369 773 768 3236  Appendix A.: DNA Sequences  161  Oligonucleotides U s e d i n T h i s Study  Name  Sequence (5' to 3')  AJ1  GCGAGTGCGTGTTAACAATGCCGCCAATGGA  8  7  10  AJ2  CGTGGCGCGAATTCGTATGCACAATGCCG  6  8  AJ3  TGTGCATACGAATTCGCGCCACGCGACCA  7  AJ4  AAAAGCCCTCTTCGAACACGCCAACTCACCG G  AJ5  A  T  total  mol. wt.  6  31  9634  9  6  29  8970  10  7  5  29  8899  10  13  5  4  32  9740  GTTGGCGTGTTCGAAGAGGGCTTTTTCGTAC  4  6  11  11  32  9941  AJ6  TGGTCCATTCGAACGGCCCGTTTGCAT  4  8  7  8  27  8295  AJ7  AACGGGCCGTTCGAATGGACCAGAACG  8  7  9  3  27  8396  AJ8  GGGAATTCCATATGGCGATTGCAATTGGCCT CGAT  8  7  10  10  35  10847  AJ9  CGGCGGATCCTGCAGTTATAGAGTCGCAACG GCCTGGGC  7  11  14  7  39  12091  AJ10  CGAACACGCCAACTCACC  6  9  2  1  18  5445  MT04  CCGCGCGGGGGATCCTGCAGTTACTGCCCGT AATATGCC  6  13  12  8  39  12002  MT05  GGCCCCATGGCCCATATGTTAGAAGATCTCA AACGCCAGG  11  12  10  7  40  12315  MT08  C GAC GAGCGTGACAC CAC GAT G C C  6  9  7  2  24  7389  MT09  GCAGAGCGAGGTATGTAGGCGGTG  5  3  12  4  24  7595  MT12  GGCGTCGATTACAGCGTCATGA  5  5  7  5  22  6833  MT13  GGGCGCAGGCGGGTCAGTCGAT  3  5  11  3  22  6915  MT14  GGCGTTCTGGTCCATTCCCACGGCCCGT  2  11  8  7  28  8561  MT15  CGTACCTTCAACCACTTCACCGGT  5  10  3  6  24  7265  c  C  G  Appendix B N M R Spectra  H-C1  4 . 6  4 . 4  4 . 2 1  4 . 0  3 . 8  H(ppm)  500 MHz Proton-NMR Spectrum of Commercial D-Xylulose-5-Phosphate (Sodium Salt) in D 0 (Solvent Suppressed).  Figure B.l.  2  162  Appendix B: NMR Spectra  1  4.8  163  :  1  4.5  |  4.2 1  I  3.9  H (ppm)  Figure B.2. 500 MHz DQFCOSY of L-Ribulose-5-Phosphate (Free Acid) in D 0 (Solvent Suppressed). 2  Appendix B: NMR Spectra  4.5  164  1  4.2  H (ppm)  3.9  Figure B.3. 500 MHz DQFCOSY of Commercial D-Xylulose-5-Phosphate (Sodium Salt) in D 0 (Solvent Suppressed). 2  Appendix B: NMR Spectra  165  I  1 1  ! i  i  1  i•  i  T 1  i  i  lit  AI^^  i—i—i—i—i—i—i—i—i—i F i g u r6e 0B . 4 . 5.3 200  D 0. 2  i ,  11  M  1 1  ' i  l  ii  i - i • i • i  5.4 5.1 0 4.3 4.6 4.a a.2 Salt) in MHz 5.6 Proton-NMR of Allyl5 . Phosphate (Cyclohexylammonium  Appendix B: NMR Spectra  166  i  i i i  1  i  !  j  j  i  |  !  H  |  || ij  i l l  l If i!  V  c c a  1  ,  i i  ' i  Ijii i  |f r> r>  E  1  1  1  Figure  1  5.0  5.2  1  Ij i  i' % 1  i 1  .  1  4.8  1  1  4.6  .  1  4.4  .  1  4.2 PPM  .  1  4.0  1  1  3.8  1  1  3.6  1  1  1  3.4  B.5. 200 MHz Proton-NMR Spectrum of Glycolaldehyde Phosphate (Ca Salt) in 2+  Acidic D 0. 2  Appendix C Protein Sequence Alignments  ete A m i n o A c i d S e q u e n c e A l i g n m e n t arad_ecoli: ML EDLKRQVLEANLALPKHNL V arad_salty: ML EDLKRQVLEANLALPKHNL V arad_haein: ML AQLKKEVFEANLALPKHHL V fuca_haein: MNRAELSQKIIDTCLEMTKLGL N fuca_ecoli: MERNKLARQIIDTCLEMTRLGL N r h a d _ e c o l i : MQNITQSWFVGMIKATTDAWLKGWDERNGGNLTLRLDDADIAPYHDNFHQ Consensus: M L L L--  22 22 22 23 23 50  arad_ecoli: arad_salty: arad_haein: fuca_haein: fuca_ecoli: rhad_ecoli: Consensus :  TLTWGNVSAVDR ERGVFVIKPSGVDY SVMTADDMVWSIET G TLTWGNVSAVDR ERGVLVIKPSGVDY SVMTADDMVWSLES G TFTWGNVS AI DR EKNLWIKPSGVDY DVMTENDMVWDLFT G QGTAGNVS V R YKDGMLITPTGMPY HLMKTENIVYVD GN G QGTAGNVS V R YQDGMLITPTGIPY EKLTESHIVFID GN G QPRYIPLSQPMPLLANTPFIVTGSGKFFRNVQLDPAANLGIVKVDSDGAG —T-GNVS R I-P-G—Y V-V G  64 64 64 62 62 100  arad_ecoli: arad_salty: arad_haein: fuca_haein: fuca_ecoli: rhad_ecoli: Consensus:  EW EGTKKPSSDTPTHRLLYQAFPSIGG IVHTHSRHATIWAQ EW EGHKKPSSDTPTHRLLYQAFPTIGG IVHTHSRHATIWAQ NIV EGNKKPSSDTPTHLELYRQFPHIGG IVHTHSRHATIWAQ KH EENKLPSSEWQFHLSVYHTRPEANA WHNHSIHCAGLSI KHE EG KLPSSEWRFHMAAYQSRPDANA WHNHAVHCTAVSI YHILWGLTNEAVPTSELPAHFLSHCERIATNGKDRVIMHCHATNLIALTY E—K-PSS H Y P VH-H—H  106 106 106 103 103 150  arad_ecoli: arad_salty: arad_haein: fuca_haein: fuca_ecoli: rhad_ecoli: Consensus:  AGQSIPATGTTHADYFYGTIPCTRKMTDAEINGEYEWETGNVIVETFEKQ AGQPIPATGTTHADYFYGTIPCTRKMTEAEINGEYEWETGNVIVETFEKQ AGLDIIEVGTTHGDYFYGTIPCTRQMTTKEIKGNYELETGKVIVETFLSR LEKPIPAIHYMVAVSGTDHIPCVPYAT F GSHKLASYVAT LNRSIPAIHYMIAAAGGNSIPCAPYAT F GTRELSEHVAL VLENDTAVFT RQLWEGSTECLWFPDG VGILPWMVPGTDEIGQATAQ I-A IPC T G  156 156 156 142 142 197  arad_ecoli: arad_salty: arad_haein: fuca_haein: fuca_ecoli: rhad_ecoli: Consensus:  GIDAAQMPGVLVHSHGPFAWGKNAEDAVHNAIVLEEVA YMGIFCRQLA GIDAAQMPGVLVHSHGPFAWGKNAEDAVHNAIVLEEVA YMGIFCRHLA GIEPDNIPAVLVHSHGPFAWGKDANNAVHNAWLEEVA YMNLFSQQLN GIKESK AI LLAHHGLITCGENLDKALWLAQEVEVLASWYLKLLSTGLE ALKNRK ATLLQHHGLIACEVNLEKALWLAHEVEVLAQLYLTTLAITDP EMQKHSL VLWPFHGVFGSGPTLDETFGLIDTAEKSAQVLVKVYSMGGM L HG G A A E—A—Y  204 204 204 190 190 245  arad_ecoli: arad_salty: arad_haein: fuca_haein: fuca_ecoli: rhad_ecoli: Consensus :  PQLPDMQ QTLLDKHYLRKHGAKAYYGQ PQLPDMQ QSLLDKHYLRKHGAKAYYGQ PYLSPMQ KDLLDKHYLRKHGQNAYYGQ IPLLSKEQMQWLGK FHTYGLRIEES VPVLSDEEIAWLEK FKTYGLRIEE KQTISREEL IALGKRFGVTPLASALAL -P-L L-K G  231 231 231 216 215 272  Sequences were aligned using SEQSEE (Wishart et al, 1994). Key: arad, L-ribulose-5phosphate 4-epimerase; fuca, L-fuculose-l-phosphate aldolase; rhad, L-rhamnulose-1-  167  Appendix C: Protein Sequence Alignments  168  Members of the AraD/FucA Family  ARAD_BACSU  ^^XXXXXIIIIIIIIIIIIIHIIIIIIIB  SGAE_MYCPN  ^**X*XXM\\\\\\\m\\n\\Ml-  ARAD _HAEIN  ^XXXXXXailllllllllllllllHHIIII  YJFX._ECOLI  ^XXXXXXllllllllIIIIIIHill|n||  ARAD._SALTY  ARAD._EC0LI  ^xxKxxxaiiiiiiiiiiiiimiiimii  YIAS_.EC0LI  FUCA_.EC0LI  HXXXXXXXl  16727  FUCA_ HAEIN  XXXXXXXl  16727  YGBL_ EC0LI YE18_ METJA  -1  - ^ ^ x x x x  ******  YGBL_ HAEIN  Figure C.I. Proteins Sharing a Homologous Domain with E. coli L-Ribulose-5-Phosphate 4-Epimerase. Key: BACSU, Bacillus subtilir, MYCPN, Mycoplasma pneumoniae; HAEIN, Haemophilus influenzae; ECOLI, Escherichia coli; SALTY, Salmonella typhimurium; METJA,  Methanococcusjannaschii; ARAD, L-ribulose-5-phosphate 4-epimerase; FUCA, L-fuculose-l-phosphate aldolase; SGAE, probable sugar isomerase; YJFX, probable sugar isomerase SGAE; YIAS, probable sugar isomerase SGBE; YGBL, hypothetical 23.2 kDa protein; YE18, hypothetical protein.  Appendix D Graphical Representation of Enzyme Kinetics  169  7  Appendix  D: Graphical Representation of Enzyme  170  Kinetics  " i i  co o  i ' i ' i ' i r 1  co o  2  E  CN  o  I (S  OD  T-  o o  o o T -  o o  4  (D  o  o p  o q o  o  CM O O  00  CN  CD  CM  CN ( N CM CM  00 T -  CO  Tj* ~  (Ujoi/iftiri) °A  CN T  _  I ! I r -  o  i-  1  00 O  co o  co o  I  1  oo o o(fi o  (uiw/iflri) °A  (ujUJ/^rl) °A  CO  ,  CN  d  (u!LU/v\|r1) °A  d  4  Appendix D: Graphical Representation of Enzyme Kinetics  171  References  Adams, E . (1976). Catalytic Aspects of Enzymatic Racemization. Adv. En%ymol. Re/at. Areas Mol. Biol. 44, 69-138. Alexander, R.S., L . L . Kiefer, C A . Fierke, and D . W . Christianson (1993). Engineering the Zinc Binding Site of Human Carbonic Anhydrase II: Structure of the His94-Cys Apoenzyme in a New Crystalline Form. Biochemistry 32, 1510-1518. Amann, E . , and J. Brosius (1985). " A T G Vectors" for Regulated High-Level Expression of Cloned Genes in Escherichia coli. Gene 40, 183-190. Amann, E . , J. Brosius, and M . Ptashne (1983). Vectors Bearing a Hybrid trp-lac Promoter Useful of Regulated Expression of Cloned Genes in Escherichia coli. Gene 25,167-178. Anderson, R.L. (1966). L-Ribulose 5-Phosphate. Methods in Entomology 9, 48-51. Andersson, A . , G . Schneider, and Y . Lindqvist (1995). Purification and Preliminary X-ray Crystallographic Studies of Recombinant L-Ribulose-5-Phosphate 4-Epimerase from Escherichia coli. Protein Science 4, 1648-1650. Ashton, D.S., C R . BeddeU, B . N . Green, and R.W.A. Oliver (1994). Rapid Validation of Molecular Structures of Biological Samples by Electrospray-Mass Spectrometry. FEBSLett 342,1-6. Ausubel, F . M . , R. Brent, R . E . Kingston, D . D . Moore, J . G . Seidman, J.A. Smith, and K . Struhl. (1992). Short Protocols in Molecular Biology. (A Compedium of Protocolsfrom Current Protocols in Molecular Biology). (NY: Green Publishing Associates and John Wiley and Sons). Barber, G . A . (1979). Observations on the Mechanism of the Reversible Epimerization of GDP-D-mannose to GDP-L-galactose by an Enzyme from Chlorellapyrenoidosa. J. Biol. Chem. 254, 7600-7603.  172  173 Bauer, L., and O. Exner (1974). The Chemistry of Hydroxamic Acids and N-Hydroxyimides.  Angew. Chem. Int. Ed. Engl. 13, 376-384. Berry, A., and K . E . Marshall (1993). Identification of Zinc-Binding Ligands in the Class II Fructose-l,6-bisphosphate Aldolase of Escherichia coli. FEBS 318,11-16. Bertini, I., P. Turano, and A.J. Vila (1993). Nuclear Magnetic Resonance of Paramagnetic  Metalloproteins. Chem. Rep. 93, 2833-2932. Bonner, C.A., S. Hays, K . McEntee, and M . F . Goodman (1990). D N A Polymerase II is Encoded by the D N A Damage-Inducible dinA gene of Escherichia coli. Proc. Natl.  Acad. Sci. USA 87, 7663-7667. Bradford, M . M . (1976). A Rapid and Sensitive Method for the Quantitation of Microgram Quantities of Protein Utilizing the Principle of Protein-Dye Binding. Anal. Biochem. 72, 248-254. Browner, M.F., W.W. Smith, and A . L . Castelhano (1995). Matrilysin-Inhibitor Complexes: Common Themes among Metalloproteases. Biochemistry 34, 6602-6610. Burma, D.P., and B.L. Horecker (1957). L-Ribulokinase and the Formation of D-Xylulose Phosphate in Lactobacillus pentosus. Biochim. Biophys. Acta 24, 660-661. Burma, D.P., and B.L. Horecker (1958a). Pentose Fermentation by Lactobacillus plantarum.  III. Ribulokinase./. Biol. Chem. 231, 1039-1051. Burma, D.P., and B.L. Horecker (1958b). Pentose Fermentation by Lactobacillus plantarum.  IV. L-Ribulose-5-Phosphate 4-Epimerase./. Biol. Chem. 231, 1053-1064. Burton, R . M . , and N . O . Kaplan (1953). A D P N Specific Glycerol Dehydrogenase from Aerobacteraerogenes. J.Am. Chem. Soc. 75, 1005-1006. Chen, B., and A . E . Przybyla (1994). A n Efficient Site-Directed Mutagenesis Method Based  on PCR. BioTechniques 17, 657-659. Chen, H . , Y . Sun, T. Stark, W. Beattie, and R . E . Moses (1990). Nucleotide Sequence and Deletion Analysis of the polB Gene of Escherichia coli. DNA and Cell Biology 9, 631-635. Chen, Y . M . , Y . Zhu, and E . C . C . Lin (1987). The Organization of the/^Regulon Specifying L-Fuculose Dissimilation in Escherichia coliKA2 as Determined by Gene Cloning. Mol.  Gen. Genet. 210, 331-337.  174 Collins, K . D . (1974). A n Activated Intermediate Analogue. The Use of Phosphoglycolohydroxamate as a Stable Analogue of a Transiendy Occurring Dihydroxyacetone Phosphate-Derived Enolate in Enzymatic Catalysis. / . Biol. Chem. 249,136-142. Cooper, S.J., G . A . Leonard, S.M. McSweeney, A.W. Thompson, J . H . Naismith, S. Qamar, A . Plater, A . Berry, and W . N . Hunter (1996). The Crystal Structure of a Class II Fructose-l,6-Bisphosphate Aldolase Shows a Novel Binuclear Metal-Binding Active Stie Embedded in a Familiar Fold. Structure 4, 1303-1315. Coulson, A.F.W., J.R. Knowles, and R.E. Offord (1970). Active-Site-Directed Inhibition of  Triosephosphate Isomerase. / . Chem. Soc. D 1, 7-8. Crowder, M.W., Z . Wang, S.L. Franklin, E . P . Zovinka, and S.J. Benkovic (1996). Characterization of the Metal-Binding Sites of the P-Lactamase from Bacteroides fragilis. Biochemistry 35, 12126-12132. Davis, L . , N . Lee, and L . Glaser (1972). O n the Mechanism of the Pentose Phosphate  Epimerases. /. Biol. Chem. 247, 5862-5866. de la Mare, S., A.F.W. Coulson, J.R. Knowles, J.D. Priddle, and R.E. Offord (1971). ActiveSite Labelling of Triose Phosphate Isomerase. The Reaction of Bromohydroxyacetone Phosphate with a Unique Glutamic Acid Residue and the Migration of the Label to Tyrosine. Biochem. J. 129, 321-331. Deupree, J., and W. A . Wood (1975). L-Ribulose-5-Phosphate 4-Epimerase from Aerobacter  aerogenes. Methods En%ymol. 41, 412-419. Deupree, J.D., and W.A. Wood (1970). L-Ribulose 5-Phosphate 4-Epimerase of Aerobacter aerogenes. Evidence for Nicotinamide Adenine Dinucleotide-Independent 4-  Epimerization by the Crystalline Enzyme. / . Biol. Chem. 245, 3988-3995. Deupree, J.D., and W.A. Wood (1972). L-Ribulose 5-Phosphate 4-Epimerase from Aerobacter aerogenes. Evidence for a Role of Divalent Metal Ions in the Epimerization Reaction.  /. Biol. Chem. 247, 3093-3097. Dreyer, M . K . , and G . E . Schulz (1993). The Spatial Structure of the Class II L-Fuculose-1-  Phosphate Aldolase from Escherichia coli. J. Mol. Biol. 231, 549-553. Dreyer, M . K . , and G . E . Schulz (1996a). Catalytic Mechanism of the Metal-Dependent Fuculose Aldolase from Escherichia cok as Derived from the Structure. / . Mol. Biol. 259, 458-466.  175  Dreyer, M . K . , and G . E . Schulz (1996b). Refined High-Resolution Structure of the Metal-Ion Dependent L-Fuculose-1-Phosphate Aldolase (Class II) from Escherichia coli. Acta Cryst. D52, 1082-1091. Edelhoch, H . (1967). Spectroscopic Determination of Tryptophan and Tyrosine in Proteins. Biochemistry 6, 1948-1954. Elgren, T . E . , L.-J. Ming, and L . Jr. Que (1994). Spectroscopic Studies of Co(II)Reconstituted Ribonucleotide Reductase R2 from Escherichia coli. Inorg Chem. 33, 891894. Englelsberg, E . (1961). Enzymatic Characterization of 17 L-Arabinose Negative Mutants of Escherichia coli. J. Bacteriol. 81, 996-1006.  Faraci, W.S., and C T . Walsh (1988). Racemization of Alanine by the Alanine Racemases from Salmonella typhimurium and Bacillus stearothermophilus: Energetic Reaction Profile Biochemistry 27, 3267-3276. Feingold, D.S., and P.A. Hoffee. (1972). Other Sugar Deoxy Aldolases. In The Enzymes (3rd Ed.), Paul D . Boyer, ed. (New York: N Y : Academic Press), 7, 303-321. Fessner, W.-D., and O . Eyrisch (1992). One-Pot Synthesis of Tagatose-l,6-Bisphosphate by Diastereoselective Enzymatic Aldol Addition. Angew. Chem. Int. Ed. Engl. 31, 56-58. Fessner, W.-D., A . Schneider, H . Held, G . Sinerius, C. Walter, M . Hixon, and J.V. Schloss (1996). The Mechanism of Class II, Metal-Dependent Aldolases. Angew. Chem. Int. Ed. Engl. 35, 2219-2221. Fleischmann, R.D., M . D . Adams, O. White, R. Clayton, E.F. Kirkness, A.R. Kerlavage, C J . Bult, J.-F. Tomb, B.A. Dougherty, J.M. Merrick, K . McKenney, G . Sutton, W. FitzHugh, C. Fields, J.D. Gocayne, J. Scott, R. Shirley, L.-I. Liu, A . Glodek, J.M. Kelley, J.F. Weidman, C A . Phillips, T.Spriggs, E . Hedblom, M . D . Cotton, T.R. Utterback, M . C . Hanna, D . T . Nguyen, D . M . Sandek, R . C Brandon, L . D . Fine, J.L. Fritchman, J.L. Fuhrmann, N.S.M. Geoghagen, C L . Gnehm, L . A . McDonald, K . V . Small, C M . Fraser, H . O . Smith, and J.C. Venter (1995). Whole-Genome Random Sequencing and Assembly of Haemophilus influenzae Rd. Science 269, 496-512. Frey, P.A. (1987). Complex Pyridine Nucleotide-Dependent Transformations. In Pyridine Nucleotide Coenzymes, D . Dolphin, R. Poulson and O . Avramovic, eds. (New York: Wiley), pp. 461-511.  176 Ghalambor, M.A., and E . C Heath (1962). The Metabolism of L-Fucose. II. The Enzymatic  Cleavage of L-Fucose-1 -Phosphate. / . Biol. Chem. 237, 2427-2433. Ghalambor, M . A . , and E . C . Heath (1966). L-Fuculose 1-Phosphate Aldolase.  Methods  En%ymol. 9, 539-542. Gielow, W.O., and N . Lee (1975). L-Ribulose-5-Phosphate 4-Epimerase from Escherichia coli.  Methods Ensymol. 41, 419-423. Gijsen, H.J.M., L . Qiao, W. Fitz, and C . - H . Wong (1996). Recent Advances in the Chemoenzymatic Synthesis of Carbohydrates and Carbohydrate Mimetics. Chem. Rev. 96,443-473. Gill, S.C., and P.H. von Hippel (1989). Calculation of protein Extinction Coefficients from Amino Acid Sequence Data. Anal Biochem 182, 319-326. Glaser, L. (1972). Epimerases. In The Enzymes (3rd Ed.), Paul D . Boyer, ed. (New York, N Y : Academic Press), 6, 355-380. Glaser, L., H . Zarkowsky, and L. Ward (1972). Formation of TDP-L-Rhamnose from T D P D-Glucose. Methods En%ymol. 28, 446-454. Hartman, F.C. (1970). Haloacetol Phosphates. Potential Active-Site Reagents for Aldolase, Triose Phosphate Isomerase, and Glycerophosphate Dehydrogenase. I. Preparation and Properties. Biochemistry 9, 1776-1782. Hartman, F.C. (1971). Haloacetol Phosphates. Characterization of the Active Site of Rabbit Muscle Triose Phosphate Isomerase. Biochemistry 10,146-154. Haussmann, C , F. Rohdich, F. Lottspeich, S. Eberhardt, J. Scheuring, S. Mackamul, and A . Bacher (1997). Dihydroneopterin Triphosphate Epimerase of Escherichia coli: Purification, Genetic Cloning, and Expression./. Bacteriol 179, 949-951. Heath, E . C , J. Hurwitz, B.L. Horecker, and A . Ginsburg (1958). Pentose Fermentation by Lactobacillus plantarum. I. The Cleavage of Xylulose 5-Phosphate by Phosphoketolase. /. Biol Chem. 231, 1009-1029. Heine,  M . C . , and G . M . Brown (1975). Enzymatic Epimerization of T)-ErythroDihydroneopterin Triphosphate to L-T^o-Dihydroneopterin Triphosphate. Biochim. et Biophys. Acta 411, 236-249.  177 Hermes, J.D., S.C. Blacklow, and J.D. Knowles (1990). Searching Sequence Space by Definably Random Mutagenesis: Improving the Catalytic Potency of an Enzyme. Proc. Natl. Acad. Sci. USA 87, 696-700. Hess, D . , T.C. Covey, R. Winz, R. W . Brownsey, and R. Aebersold (1993). Analytical and Micropreparative Peptide Mapping by High Performance Liquid Chromatography/Electrospray Mass Spectrometry of Proteins Purified by Gel Electrophoresis. Protein Sci. 2, 1342-1351. Himmelreich, R., H . Hilbert, H . Plagens, E . Pirki, B.-C. L i , and R. Herrmann (1996). Complete Sequence Analysis of the Genome of the Bacterium Mycoplasma pneumoniae. Nucleic Acids Res. 24, 4420-4449. Hlavaty, J.J., and T. Nowak (1997). Formation and Characterization of an Active Phosphoenolpyruvate Carboxykinase-Cobalt(III) Complex. Biochemistry 36, 33893403. Holmes, A . , and B . W . Matthews (1981). Binding of Hydroxamic Acid Inhibitors to Crystalline Thermolysin Suggests A Pentacoordinate Zinc Intermediate in Catalysis. Biochemistry 20, 6912-6920. Horecker, B.L., O . Tsolas, and C Y . Lai. (1972). Aldolases. In The Enzymes (3rd Ed.), Paul D . Boyer, ed. (New York, N Y : Academic Press), 7, 213-258. Iwasaki, H . , Y . Ishino, H . Toh, A . Nakata, and H . Shinagawa (1991). Escherichia «?/*'DNA Polymerase II is Homologous to the a-like D N A Polymerases. Mol. Gen. Genet. 226, 24-33. Izquierdo-Martin, M . , and R . L . Stein (1992). Mechanistic Studies on the Inhibition of Thermolysin by a Peptide Hydroxamic Acid. / . Am. Chem. Soc. 114, 325-331. Johnson, W . C , Jr. (1990). Protein Secondary Structure and Circular Dichroism: A Practical Guide. Proteins: Structure, Function and Genetics 7, 205-214. Jones, D . H . , and S.C. Winistorfer (1992). Recombinant Circle P C R and Recombination P C R for Site-Directed Mutagenesis Without P C R Product Purification. BioTechniques 12, 528-534. Joshi, A . K . , V . Baichwal, and G . Ferro-Luzzi Ames (1991). Rapid Polymerase Chain Reaction Amplification Using Intact Bacterial Cells. BioTechniques 10, 42-45.  178 Kanaori, K . , N . Uodome, A . Nagai, D . Ohta, A . Ogawa, G . Iwasaki, and A . Y . Nosaka (1996). C d Nuclear Magnetic Resonance Studies of Cabbage Histidinol Dehydrogenase. Biochemistry 35, 5999-5954. 1 1 3  Kunkel, T.A. (1992). Oligonucleotide-Directed Mutagenesis Without Phenotypic Selection. In Current Protocols in Molecular Biology., Frederick M . Ausubel, Roger Brent, Robert E . Kingston, David D . Moore, J.G. Seidman, John A . Smith and Kevin Struhl, eds. (NY: Green Pulishing Associates and John Wiley and Sons.), pp. 8.3-8.5. Kuo, J.M., M . Y . Chae, and F . M . Raushel (1997). Perturbations to the Active Site of Phosphotriesterase. Biochemistry 36, 1982-1988. Kuo, J.M., and F . M . Raushel (1994). Identification of the Histidine Ligands to the Binuclear Metal Center of Phosphotriesterase by Site-Directed Mutagenesis. Biochemistry 33, 4265-4272. Lee, N . , and I. Bendet (1967). Crystalline L-Ribulokinase from Escherichia coll. J. Biol. Chem. IM, 2043-2050. Lee, N . , and E . Englesberg (1966). L-Ribulokinase. Methods En%ymoI. 9, 449-454. Lee, N . , W. Gielow, R. Martin, E . Hamilton, and A . Fowler (1986). The Organization of the araBAD Operon of Escherichia coli. Gene 47, 231-244. Lee, N . , J.W. Patrick, and M . Masson (1968). Crystalline L-Ribulose 5-Phosphate 4Epimerase from Escherichia coli. J. Biol. Chem. 243, 4700-4705. Lin, H . - C , S.-P. Lei, G . Studnicka, and M . Wilcox (1985). The araBAD Operon of Salmonella typhimurium LT2. III. Nucleotide Sequence of araD and its Flanking Regions, and Primary Structure of its Product, L-Ribulose-5-Phosphate 4-Epimerase. Gene 34,129134. Lin, Y . , R . D . Kobes, L L . Norton, and F.C. Hartman (1971). Affinity Labeling of Yeast Aldolase by Haloacetol Phosphates. Biochem. Biophys. Res. Commun. 45, 34-41. Lipscomb, W . N . , and N . Strater (1996). Recent Advances in Zinc Enzymology. Chem. Rev. 96, 2375-2433. London, J., and S. Hausman (1982). Xylitol-Mediated Transient Inhibition of Ribitol Utilization by Lactobacillus easel. J. Bacteriol. 150, 657-661.  179 Makrides, S.C. (1996). Strategies for Achieving High-Level Expression of Genes in Escherichia coli. Microbiol. Rev. 60, 512-538.  Maret, W., and B.L. Vallee. (1993). Cobalt as Probe and Label of Proteins. Methods En^ymo I. 158, 52-71. Marquart, D.W. (1963). An Algorithm for Least-Squares Estimation of Nonhnear Parameters. /. Soc. Ind. Appl. Math. 11, 431-441. McCafferty, D.G., I.A.D. Lessard, and CT. Walsh (1997). Mutational Analysis of Potential Zinc-Binding Residues in the Active Site of the Enterococcal D-Ala-D-Ala Dipeptidase VanX. Biochemistry 36, 10498-10505. McDonough, M.W., and W. A. Wood (1961). The Mechanism of Pentose Phosphate Isomerization and Epimerization Studied with T O and H 0 . /. Biol. Chem. 236, 1220-1224. 18  z  2  Melo, A., and L. Glaser (1968). The Mechanism of 6-Deoxyhexose Synthesis. II. Conversion of Deoxythymidine Diphosphate 4-Keto-6-Deoxy-D-Glucose to Deoxythymidine Diphosphate L-Rhamnose./. Biol Chem. 243, 1475-1478. Mildvan, A. S. (1972). Metals in Enzyme Catalysis. In The Enzymes (3rd Ed.), Paul D. Boyer, ed. (New York: Academic Press), 2, 445-536. Mineno, J., H. Fukui, Y. Ishino, I. Kato, and H. Shinagawa (1990). Nucleotide Sequence of the araD Gene of Escherichia coli K12 Encoding the L-Ribulose 5-Phosphate 4Epimerase. Nucleic Acids Res. 18, 6722. Moralejo, P., S.M. Egan, E. Hidalgo, and J. Aguilar (1993). Sequencing and Characterization of a Gene Cluster Encoding the Enzymes for L-Rhamnulose Metabolism in Escherichia coli. J. Bacteriol. 175, 5585-5594.  Morgan, P.M., R.F. Sala, and M.E. Tanner (1997). Eliminations in the Reactions Catalyzed by UDP-N-Acetylglucosamine 2-Epimerase. /. Am. Chem. Soc. 119, 10269-10277. Morgenlie, S. (1987). Aldol Reaction Between Small Sugars. Preparation of DL,-threo-2Pentulose and DL-Zyxo-S-Hexulose and their Isolation as OTsopropylidene Derivatives. Acta Chemica Scandinavica B41, 745-748.  Morse, D.E., and B.L. Horecker (1968). The Mechanism of Action of Aldolases. Adv. En^ymol. 31,125-181.  180 Muller, D., S. Pitsch, A. Kittaka, E. Wagner, CE. Wintner, and A. Eschenmoser (1990). Chemie von a-Aminonittilen. Aldomerisierung von Glycolaldehyd-Phosphat zu Racemischen Hexose-2,4,6-Triphosphaten und (in Gegenwart von Formaldehyd) Racemischen Pentose-2,4-Diphosphaten:ra^-Allose-2,4,6-Triphosphatund racRibose-2,4-Diphosphat sind die Reaktionshauptprodukte. Helv. Chim. Acta 73, 14101468. Norton, I.L., and F.C. Hartman (1972). Haloacetol Phosphates. A Comparative Study of the Active Sites of Yeast and Muscle Triose Phosphate Isomerase. Biochemistry 11, 44354441. Ochiai, E.-I. (1987). Principles of Biochemistry of the Elements. (New York: Plenum Press). Orskov, I.. (1984). Klebsiella. In Bergey's Manual of Systematic Bacteriology, J.G. Holt, ed. (Baltimore, MD: The Williams & Wilkins Company), pp. 461-465. Ramsey, A.J., S.C Daubner, J.I. Ehrlich, and P.F. Fitzpatrick (1995). Identification of Iron Ligands in Tyrosine Hydroxylase by Mutagenesis of Conserved Histidinyl Residues. Protein Science 4, 2082-2086.  Rappoport, D.A., H.A. Barker, and W.Z. Hassid (1951). Fermentation of L-Arabinose-l- C 14  by Lactobacillus pentoaceticus. Arch. Biochem. and Biophys. 31, 326.  Rose, I., A., and E.L. O'Connell (1969). Inactivation and Labeling of Triose Phosphate Isomerase and Enolase by Glycidol Phosphate. /. Biol. Chem. 244, 6548-6557. Rutter, W.J. (1964). Evolution of Aldolase. Federation Proc. 23, 1248-1257. Rutter, W.J. (1965). Enzymatic Homology and Analogy in Phylogeny. In Evolving Genes and Proteins., Vernon Bryson and Henry J. Vogel, eds. (New York: Academic Press), pp. 279-291. Sala, R.F., P.M. Morgan, and M.E. Tanner (1996). Enzymatic Formation and Release of a Stable Glycal Intermediate: The Mechanism of the Reaction Catlyzed by UDP-NAcetylglucosamine 2-Epimerase. /. Am. Chem. Soc. 118, 3033-3034. Salo, W.L. (1976). The Incorporation of Tritium from Tritium-Enriched Water into UDPIV-Acetylglucosamine and UDP-IV-Acetylmannosamine 2-Epimerase from Escherichia coli. Biochim. Biophys. Acta 452, 625-628.  181 Salo, W.L., D.D. Fossitt, R.D. Bevill, S. Kirkwood, and W.A. Wood (1972). L-Ribulose 5Phosphate 4-Epimerase from Aerobacter aerogenes. Kinetic Isotope Effect with Tritiated Substrate./. Biol. Chem. 247, 3098-3100. Sambrook, J., T. Maniatis, and E.F. Fritsch. (1989). Molecular Cloning. A Laboratory Manual (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory). Sa-Nogueira, I., T.V. Nogueira, S. Soares, and H. de Lencastre (1997). The Bacillus subtilis LArabinose (ard) Operon: Nucleotide Sequence, Genetic Organization and Expression. Microbiology 143, 957-969. Sarkar, G., and S.S. Sommer (1990). The "Megaprimer" Method of Site-Directed Mutagenesis. BioTechniques 8, 404-407. Scharschimdt, M., G. Pfleiderer, H. Metz, and W. Brummer (1983). Isolierung und Charakterisierung von Glycerin-Dehydrogenase aus Bacillus megaterium. Hoppe-Seyler's Z. Physiol. Chem. 364, 911-921.  Schray, KJ., E.L. O'Connell, and LA. Rose (1973). Inactivation of Muscle Triose Phosphate Isomerase by D- and L-Glycidol Phosphate./. Biol. Chem. 248, 2214-2218. Scolnick, L.R., A.M. Clements, J. Liao, L. Crenshaw, M. Hellberg, J. May, T.R. Dean, and D.W. Christianson (1997). Novel Binding Mode of Hydroxamate Inhibitors to Human Carbonic Anhydrase II./. Am. Chem. Soc. 119, 850-851. Simpson, F.J., M.J. Wolin, and W.A. Wood (1958). Degradation of L-Arabinose by Aerobacter aerogenes. I. A Pathway Involving Phosphorylated Intermediates. /. Biol. Chem. 230, 457-472. Simpson, F.J., and W.A. Wood (1956). L-Ribulose-5-Phosphate: Formation by Purified Kinase from Aerobacter Aerogenes. J. Am. Chem. Soc. 78, 5452-5453. Simpson, F.J., and W.A. Wood (1958). Degradation of L-Arabinose by Aerobacter aerogenes. II. Purification and Properties of L-Ribulokinase. /. Biol. Chem. 230, 473-486. Spencer, P., K.J. Bown, M.D. Scawen, T. Atkinson, M.G. Gore (1989). Isolation and Characterisation of the Glycerol Dehydrogenase from Bacillus stearothermophilus. Biochim. Biophys. Acta 994, 270-279.  Summers, M.F. (1988). Cd NMR Spectroscopy of Coordination Compounds and Proteins. 113  Coord. Chem. Rev. 86, 43-134.  182 Tanner, M.E., and G.L. Kenyon. (1998). Inversion of Stereocentres. In Comprehensive  Biological Catalysis. A Mechanistic Reference., Reactions ofNucleophilic/Carbanionoid Carb  M. Sinnott, ed. (New York: Academic Press) 2, 7-46. Vallee, B.L., and D.S. Auld (1990a). Active-site Zinc Ligands and Activated H 0 of Zinc 2  Enzymes. Proc. Natl. Acad. Sci. USA 87, 220-224.  Vallee, B.L., and D.S. Auld (1990b). Zinc Coordination, Function, and Structure of Zinc Enzymes and Other Proteins. Biochemistry 29, 5647-5659. Van Wart, H.E. (1988). Introduction of Exchange-Inert Metal Ions into Enzymes. Methods EnymoL 158, 95-110.  Wishart, D.S., R.F. Boyko, L. Willard, F.M. Richards, and B.D. Sykes (1994). SEQSEE: A Comprehensive Program Suite for Protein Sequence Analysis. Comput. Appl. Biosci. 10, 121-132. Wolin, M.J., F.J. Simpson, and W.A. Wood (1957). L-Ribulose-5-Phosphate - D-xylulose-5phosphate stereoisomerase and its role in L-arabinose fermentation. Biochim. Biophys. Acta 24, 635-638. Wolin, M.J., F.J. Simpson, and W.A. Wood (1958). Degradation of L-Arabinose by Aerobacter aerogenes. III. Identification and Properties of L-Ribulose-5-Phosphate 4-Epimerase. /. Biol. Chem. 232, 559-575.  Wong, C.-H., R. L. Halcomb, Y. Ichikawa, and T. Kajimoto (1995). Enzymes in Organic Synthesis: Application to the Problems of Carbohydrate Recognition (Part 1). Angew. Chem. Int. Ed. Engl. 34, 412-432.  Wood, W.A. (1972). 2-Keto-3-deoxy-6-phosphogluconic and Related Aldolases. In The Enzymes (3rd Ed.), Paul D. Boyer, ed. (New York: Academic Press), 7, 281-302. Yura, T., H. Mori, H. Nagai, T. Nagata, A. Ishihama, N. Fujita, K. Isono, K. Mizobuchi, and A. Nakata (1992). Systematic Sequencing of the Escherichia coli Genome: Analysis of the 0-2.4 min Region. Nucleic Acids Res. 20, 3305-3308. Zhou, C, L. Abaigar, and A.Y. Jong (1990). A Protocol for Using T7 DNA Polymerase in Oligonucleotide Site-Directed Mutagenesis. BioTechniques 8, 503.  

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.831.1-0061643/manifest

Comment

Related Items