USF Libraries
USF Digital Collections

Structure-function studies of conserved sequence motifs of cytochrome b5 reductase

MISSING IMAGE

Material Information

Title:
Structure-function studies of conserved sequence motifs of cytochrome b5 reductase
Physical Description:
Book
Language:
English
Creator:
Crowley, Louis J
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Flavoprotein
Transhydrogenases
Oxidoreductases
Methemoglobinemia
Mutagenesis
Dissertations, Academic -- Molecular Medicine -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: NADH:Cytochrome b5 Reductase (cb5r) catalyzes the two electron reduction of the iron center of the heme cofactor found within cytochrome b5 (cb5) utilizing reducing equivalents of the nicotinamide adenine dinucleotide (NADH) coenzyme. Cb5r is characterized by two domains necessary for proper enzyme function: a flavin-binding domain and a pyridine nucleotide-binding domain. Within these domains are highly conserved "motifs" necessary for the proper binding and orientation of both the NADH coenzyme and the FAD cofactor. To address the importance of these conserved motifs site-directed mutagenesis was utilized to generate a series of variants upon residues found within the motifs to allow for the full characterizations. Second, naturally occurring recessive congenital methemoglobinemia (RCM) mutants that are found within or in close proximity to these highly conserved motifs were analyzed utilizing site-directed mutagenesis.^ The flavin-binding motif "91RxYSTxxSN97" was characterized by the generation of variants T94H, T94G, T94P, P95I, V96S, and S97N. In addition to this, the naturally occurring double mutant P92H/E255- was fully characterized to establish a role of the P92 residue giving rise to RCM. The role of the "124GRxxST127" was determined by the introduction of a positive charge, charge reversal, and conserved amino acid mutations through site-directed mutagenesis of the G124, K125, and M126 residues. Based on the data presented here, each of the residues of the GRxxST motif are directly involved in maintaining the proper binding and orientation of the cb5r flavin prosthetic group. Analysis of the NADH-binding motif "273CGxxx-M278" was accomplished through the characterization of the type II RCM variant M272- and the type I RCM variant P275L. This demonstrates that the deletion of the M272 residue causes a frame shift leading to the inability of the NADH substrate to bind.^ ^The introduction of the P275L variant showed that substrate affinity was diminished, yet turnover was comparable to wild-type cytochrome b5 reductase, indicating that although P275 is required for proper substrate binding it is not essential for overall catalytic function. Finally, analysis of the naturally occurring double mutant G75S/V252M provided the first insight into a methemoglobinemia variant that possessed mutations in both the FAD-binding and NADH-binding domains.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Louis J. Crowley.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 197 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001926553
oclc - 191885154
usfldc doi - E14-SFE0001913
usfldc handle - e14.1913
System ID:
SFS0026231:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001926553
003 fts
005 20080208142040.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080208s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001913
035
(OCoLC)191885154
040
FHM
c FHM
049
FHMM
090
QR46 (ONLINE)
1 100
Crowley, Louis J.
0 245
Structure-function studies of conserved sequence motifs of cytochrome b5 reductase
h [electronic resource] /
by Louis J. Crowley.
260
[Tampa, Fla] :
b University of South Florida,
2007.
3 520
ABSTRACT: NADH:Cytochrome b5 Reductase (cb5r) catalyzes the two electron reduction of the iron center of the heme cofactor found within cytochrome b5 (cb5) utilizing reducing equivalents of the nicotinamide adenine dinucleotide (NADH) coenzyme. Cb5r is characterized by two domains necessary for proper enzyme function: a flavin-binding domain and a pyridine nucleotide-binding domain. Within these domains are highly conserved "motifs" necessary for the proper binding and orientation of both the NADH coenzyme and the FAD cofactor. To address the importance of these conserved motifs site-directed mutagenesis was utilized to generate a series of variants upon residues found within the motifs to allow for the full characterizations. Second, naturally occurring recessive congenital methemoglobinemia (RCM) mutants that are found within or in close proximity to these highly conserved motifs were analyzed utilizing site-directed mutagenesis.^ The flavin-binding motif "91RxYSTxxSN97" was characterized by the generation of variants T94H, T94G, T94P, P95I, V96S, and S97N. In addition to this, the naturally occurring double mutant P92H/E255- was fully characterized to establish a role of the P92 residue giving rise to RCM. The role of the "124GRxxST127" was determined by the introduction of a positive charge, charge reversal, and conserved amino acid mutations through site-directed mutagenesis of the G124, K125, and M126 residues. Based on the data presented here, each of the residues of the GRxxST motif are directly involved in maintaining the proper binding and orientation of the cb5r flavin prosthetic group. Analysis of the NADH-binding motif "273CGxxx-M278" was accomplished through the characterization of the type II RCM variant M272- and the type I RCM variant P275L. This demonstrates that the deletion of the M272 residue causes a frame shift leading to the inability of the NADH substrate to bind.^ ^The introduction of the P275L variant showed that substrate affinity was diminished, yet turnover was comparable to wild-type cytochrome b5 reductase, indicating that although P275 is required for proper substrate binding it is not essential for overall catalytic function. Finally, analysis of the naturally occurring double mutant G75S/V252M provided the first insight into a methemoglobinemia variant that possessed mutations in both the FAD-binding and NADH-binding domains.
502
Dissertation (Ph.D.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 197 pages.
Includes vita.
590
Adviser: Michael J. Barber, D. Phil.
653
Flavoprotein.
Transhydrogenases.
Oxidoreductases.
Methemoglobinemia.
Mutagenesis.
690
Dissertations, Academic
z USF
x Molecular Medicine
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1913



PAGE 1

Structure-Function Studie s of Conserved Sequence Motifs of Cytochrome b 5 Reductase: by Louis J. Crowley A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Biochemist ry and Molecular Biology College of Medicine University of South Florida Major Professor: Michael J. Barber, D.Phil. Ronald K. Keller, Ph.D. Gloria C. Ferreira, Ph.D. Larry P. Solomonson, Ph.D. Craig A. Doupnik, Ph.D. Date of Approval: April 11, 2007 Keywords: flavoprotein, transhydrogenase s, oxidoreductases, methemoglobinemia, mutagenesis. Copyright 2007, Louis J. Crowley

PAGE 2

This work is dedicated to my mother, Joanne and to the loving memory of my father, Louis J. Crowley, Sr.

PAGE 3

ACKNOWLEDGEMENTS I would first like to thank Professor Michael J. Barber D.Phil. for providing me with the opportunity to achieve this goal. His guidan ce, input, direction, and most importantly, his support, have allowed me to succeed. A special thank you to my dissertation committee members, Professor Gl oria Ferreira, Ph.D., Professor Larry P. Solomonson, Ph.D., Professor Craig A. D oupnik, Ph.D., and my committee chairperson Professor Ronald K. Keller, Ph.D., for all of your constructive criticism, ideas, input, and always offering your encouragement to me. To Christopher C. Marohnic, Ph.D. and Christopher A. Davis, Ph.D., who welcomed me into Dr. Barbers laboratory and to this day continue to be invaluable mentors and friends. I can not tha nk the two of you enough for all of your guidance, support, friendship, and welcoming me into your families while I was away from mine. To my friend and lab mate Glenn W. Roma for your constant willingness to assist me, provide input, honesty, and keeping my mind on track even when I think Ive lost it. To my brother N eal who I will always look up to for always being there for me, being the role model in my life, and for keeping me in line while still being my friend. To my amazing Mother who always made sacrifices for me, provided me with a wonderful life, love, and supported every dream Ive ever ha d. Finally, to my wonderful fiance Heather who has helped me find the love, friendship, and support I have always been looking for and look forward to enjoying for the rest of my life.

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES iii LIST OF FIGURES v LIST OF ABBREVIATIONS x ABSTRACT xiv 1. INTRODUCTION 1 Cytochrome b 5 Reductase 1 Crystal Structure of Cytochrome b 5 Reductase 8 Recessive Congenital Methemoglobinemia 14 The FNR Family of Flavoprotein Oxidoreductases 16 Research Aims and Approaches 20 2. MATERIALS AND METHODS 25 Materials 25 Molecular Biology Reagents 25 Microbiology and Protein Purification Reagents 26 Enzyme Assay and Spectroscopy Reagents 26 Methods 27 Protein Expression and Purification 27 Site-Directed Mutagenesis 29 Ultra-violet and Visible Absorbance Spectroscopy 32 Ultra-violet and Visible Circul ar Dichroism Spectroscopy 32 Fluorescence Spectroscopy 33 Steady-State Enzyme Activities 33 Charge-Transfer Complex Determination 34 Spectral Binding Constant Determination by Differential Spectroscopy 34 Thermal Stability Measurements 36 Determination of Flavin Oxidation-Reduction Potential 36

PAGE 5

ii 3. RESULTS AND DISCUSSION 37 Analysis of the conserved FAD binding motif 91 RxY S / T xx S / N 97 37 Type I recessive congenital methemoglobinemia associated with a double mutation, the novel Pro92His and previously described Glu255-, in the cytochrome b 5 reductase gene 60 The role of the FAD/FMN specificity binding motif 124G / R xx S / T 127 79 Analysis of the effects of the RCM variant M272and generated variants M272A/I/L/R on the CGxxxM motif 116 Characterization of the type I recessive congenital methemoglobinemia Mutant P275L 141 Identification and characterizati on of the novel FAD-binding lobe G75S mutation in cytochrome b 5 reductase: An aid to determine recessive congenital methemoglobinemia status in an infant 157 4. CONCLUSIONS AND FUTURE AIMS 179 REFERENCES 188 PUBLICATIONS 196 ABOUT THE AUTHOR End Page

PAGE 6

iii LIST OF TABLES Table 1. Mutagenic Methemoglobinemia Mutant Oligonucleotide Primers 30 Table 2. NADH:FR and NADH:BR Kine tic Constants of T94H, P95G, P95I, V96S, S97N, and Wild Type Cytochrome b 5 Reductase 42 Table 3. Spectral Binding Consta nts Obtained for Wild Type c b 5 r and the T94H, P95G, P95I, V96S, and S97N Variants 45 Table 4. T 50 Values of the Wild-Type Cytochrome b 5 Reductase, T94H, P95G, P95I, V96S, and S97N Variants 50 Table 5. Flavin Midp oint Potentials (E ) Obtained for Wild-Type, T94H, P95G, P95I, V96S, and S97N c b 5 rs 53 Table 6. Kinetic Constants Obtained for the Different P92 and P95 c b 5 r Mutants 71 Table 7. Spectral Binding Constants Obtain ed for Wild Type c b 5 r and the P92A, P92H, P95A, P95H, and P92H/E255Variants 73 Table 8. T 50 Values of the Wild-Type Cytochrome b 5 Reductase, G124A, G124H, G124K, G124R, K125A, K 125D, and K125E Variants 90 Table 9. NADH:FR and NADH:BR Kinetic Constants Obtained for G124A, G124H, G124K, G124R, K125A, K125D, K125E, and Wild-Type cb 5 rs 90 Table 10. Spectral Binding Consta nts Obtained for Wild Type c b 5 r and the G124A, G124H, G124K, G124R, K125A, K125D, and K125E Variants 95 Table 11. Flavin Midp oint Potentials (E ) Obtained for Wild-Type, G124A, G124H, G124K, G124R, K125A, K125D, and K125E c b 5 rs 98 Table 12. NADH:FR and NADH:BR Ki netic Constants Obtained for M126C, M126F, M126G, M126P, and M126S 104

PAGE 7

iv Table 13. Spectral Binding Consta nts Obtained for Wild Type c b 5 r and the M126C, M126F, M126G, M126P, and M126S Variants 108 Table 14. T 50 Values of the Wild-Type Cytochrome b 5 Reductase, M126C, M126F, M126G, M126P, and M126S Variants 110 Table 15. Flavin Midp oint Potentials (E ) Obtained for M126C, M126F, M126G, M126P, M126S, and Wild-Type Cytochrome b 5 Reductases 112 Table 16. Comparison of the Kinetic Constants Obtained for Wild-Type Cytochrome b 5 Reductase, M272-, M272A, M272I, M272L, and M272R 127 Table 17. Spectral Binding Consta nts Obtained for Wild Type c b 5 r and the M272-, M272A, M272I, M272L, and M272R Variants 130 Table 18. Comparison of the Kinetic a nd Spectroscopic Binding Constants Obtained for the P275L Variant of c b 5 r 153 Table 19. Comparison of the Kinetic Constants Obtained for the G75S, V252M, G75S/V252M and Wild-Type Cytochrome b 5 Reductases 165 Table 20. Spectral Binding Consta nts Obtained for Wild Type c b 5 r and the G75S, V252M, G75S/V252M Variants 169

PAGE 8

v LIST OF FIGURES Figure 1. Structure of the Human DIA1 Gene that Encodes Cytochrome b 5 Reductase 2 Figure 2. Mechanism of Cytochrome b 5 Reductase Mediated Cytochrome b 5 Reduction 4 Figure 3. New Model of the Reaction Sequence of Cytochrome b 5 Reductase 4 Figure 4. Structure of Cytochrome b 5 Generated in silico 6 Figure 5. Structure of Cytochrome b 5 Reductase Generated in silico 7 Figure 6. The Tertiary Structure of Rat Cytochrome b 5 Reductase 10 Figure 7. The Structure of the F AD-Binding Domain of Cytochrome b 5 Reductase 11 Figure 8. The Structure of the NA DH-Binding Domain of Cytochrome b 5 Reductase 12 Figure 9. The Structure of the Hinge Region of Cytochrome b 5 Reductase 13 Figure 10. Alignment of Conser ved Flavin and NADH-Binding Motifs of Various Enzymes Belonging to the FNR Family of Flavoprotein Transhydrogenases 18 Figure 11. Ultra-Violet, Visible, and Circular Dichroism spectra of Wild Type, T94H, P95G, P95I V96S, and S97N Proteins of Cytochrome b 5 Reductase 39 Figure 12. Spectroscopic Titrations of H 4 c b 5 r and the T94H, P95G, P95I, V96S, and S97N Variants in the Presence of H 4 NAD 43

PAGE 9

vi Figure 13. Differential Spectroscopy of H 4 c b 5 r, T94H, P95G, P95I, V96S, and S97N c b 5 rs in the Presence of H 4 NAD 44 Figure 14. Enzyme Inhibition Assessment with NAD + for the T94H, P95G, and P95I Variants 48 Figure 15. Temperature Stability of c b 5 r and the T94H, P95G, P95I, V96S, and S97N Mutants 49 Figure 16. Potentiometric Titrations of the T94H, P95G, P95I, V96S, S97N, and Wild-Type Cytochrome b 5 Reductases 52 Figure 17. Electrostatic In teraction of Amino Acid Residue T94 with the FAD Cofactor of Cytochrome b 5 Reductase 56 Figure 18. Struct ure of the Wild-type c b 5 r, T94H, P95G, and P95I Variants Generated in silico 58 Figure 19. X-ray-Crystallographi c Structure of Cytochrome b 5 Reductase 63 Figure 20. Detection of the G255and P92H Mutations in the DIAI Gene 65 Figure 21. SDS-PAGE Analysis of the Different P92, P95 and E255 c b 5 r Variants 66 Figure 22. UV/Visible Absorp tion and CD Spectra of c b 5 r and the Various P92, P95 and E255 Mutants 67 Figure 23. Thermal Stability Profiles for the Various P92, P95 and E255 c b 5 r Mutants 70 Figure 24. Flavin Difference Spectr a Observed Following Binding of Pyridine Nucleotides to the Different P92, P95 and E255c b 5 r Variants 72 Figure 25. Oxidation-Reduction Midpoint Potentials for the FAD Prosthetic Group in the Different P92, P95 and E255 c b 5 r Variants 74 Figure 26. Structural Models of the P92H and P95H c b 5 r Variants 78 Figure 27. Structural Repres entation Displaying the 124G R xx S T 127 Motif by the FAD Cofactor Ligand Binding Plot and Secondary Structure of Wild-Type Cytochrome b 5 Reductase 82 Figure 28. Ultra-Violet and Visible Absorbance, and Circular Dichroism of

PAGE 10

vii Wild-Type, G124A, G124H, G124K, G124R, K125A, K125D, and K125E Cytochrome b 5 Reductases 86 Figure 29. Temperature Stability of Wild-Type c b 5 r and the G124A, G124H, G124K, G124R, K125A, K125D, and K125E Mutants 89 Figure 30. Spectroscopic Titrations of H 4 c b 5 r, G124A, G124H, G124K, G124R, K125A, K125D, and K125E c b 5 rs in the Presence of H 4 NAD 93 Figure 31. Spectroscopic Titrations of H 4 c b 5 r, G124A, G124H, G124K, G124R, K125A, K125D, and K125E c b 5 rs in the Presence of NAD + 94 Figure 32. Potentiometric Titratio ns of the G124A, G124H, G124K, G124R, K125A, K125D, K125E and Wild-Type Cytochrome b 5 reductases 97 Figure 33. SDS Poly-Acrylamide Gel Electrophoresis of Wild-Type, M126C, M126F, M126G, M126P, M126S, and M126V Cytochrome b 5 Reductases 101 Figure 34. Ultra-Violet and Visible Absorbance, and Circular Dichroism of Wild-Type, M126C, M126F, M126G, M126P, and M126S Cytochrome b 5 Reductases 102 Figure 35. Spectroscopic Titrations of H 4 c b 5 r, M126C, M126F, M126G, M126P, and M126S c b 5 rs in the Presence of H 4 NAD 105 Figure 36. Spectroscopic Titrations of H 4 c b 5 r, M126C, M126F, M126G, M126P, and M126S c b 5 rs in the Presence of NAD + 106 Figure 37. Temperature Stability of c b 5 r and the M126C, M126F, M126G, M126P, and M126S Mutants 109 Figure 38. Potentiometric Titratio ns of the M126C, M126F, M126G, M126P, M126S, and Wild-Type Cytochrome b 5 Reductases 111 Figure 39. Multiple Sequence Alignment of c b 5 r Primary Structures 119 Figure 40. Graphical Sequence Alignment and Structural Representation of Amino Acid Residue M272 120 Figure 41. SDS-PAGE Analysis of the Different M272 c b 5 r Variants 122

PAGE 11

viii Figure 42. UV/Visible Absorp tion and CD Spectra of c b 5 r and the Various M272 Mutants 123 Figure 43. Thermal Stability Profiles for the Various M272 c b 5 r Mutants 125 Figure 44. Flavin Difference Spectra Ob tained Following Binding of Either H 4 NAD or NAD + to the Various M272 c b 5 r Mutants 129 Figure 45. Comparison of the Oxidatio n-Reduction Midpoint Potentials for the FAD Prosthetic Group in the Di fferent M272 Variants 133 Figure 46. Comparison of a Structural Model of the M272Mutant and the Wild-Type cb 5 r Diaphorase Domains 136 Figure 47. Difference Spectra and Spectra l-Binding Constant Determination for Wild Type Cytochrome b 5 Reductase and the M272Variant with Various Pyridine Nucleotide Analogs 138 Figure 48. Detection of the C27161T and G27209A Mutations in the DIA1 Gene 144 Figure 49. SDS-PAGE Analysis of the P275L c b 5 r Variant 146 Figure 50. UV/Visible Absorption a nd CD Spectra of the P275L c b 5 r Mutants 147 Figure 51. Temperature Stabilities of the P275L c b 5 r Mutant 149 Figure 52. Potentiometric Titrations of the P275L c b 5 r Mutant 151 Figure 53. Spectroscopic Titrations of the P275L Mutant with H 4 NAD and NAD + 154 Figure 54. Homology Model of the P275L c b 5 r Variant 156 Figure 55. Detection of the G15,635A and G27,091A Mutation in the DIA1 Gene 161 Figure 56. X-Ray Crystallographic Structure of Cytochrome b 5 Reductase 162 Figure 57. SDS-PAGE Analysis of Wild Type Cytochrome b 5 Reductase and the G75S, V252M, and G75S/V252M Variants 163 Figure 58 Comparison of the UV/Visible Spectroscopic Properties of Wild-Type Cytochrome b 5 Reductase and the G75S, V252M,

PAGE 12

ix and G75S/V252M Variants 164 Figure 59. Flavin Difference Spectra Obtained Following Binding of Either H 4 NAD or NAD + to Wild-Type Cytochrome b 5 Reductase or the Various G75S and V252M Mutants 168 Figure 60. Comparison of the Th ermostability Properties of Wild-Type Cytochrome b 5 Reductase and the G75S, V252M, G75S/V252M Variants 170 Figure 61. Comparison of the Therm odynamic Properties of Wild-Type Cytochrome b 5 Reductase and the V252M Variant 171 Figure 62. Multiple Sequence Alignment of Selected Cytochrome b 5 Reductase Primary Structures 176

PAGE 13

x LIST OF ABBREVIATIONS A alanine angstrom (1 = 0.1 nm) ADP adenosine-5-diphosphate ADPR adenosine-5-diphosphoribose Amp ampicillin ATP adenosine-5-triphosphate C cysteine CaCl 2 calcium chloride c b 5 cytochrome b 5 c b 5 r cytochrome b 5 reductase CD circular dichroism CPK Corey, Pauling, and Kultun (molecular coloring scheme) D aspartate DEAE diethylaminoethyl DTT dithiothreitol E glutamate EDTA ethylenediaminetetraacetic acid EtBr ethidium bromide

PAGE 14

xi F phenylalanine FAD flavin adenine dinucleotide FADH flavin adenine dinucleotide, semiquinone FADH 2 flavin adenine dinucleotide, dihydroquinone FMN flavin mononucleotide FNR ferredoxin:NADP + reductase FPLC fast protein liquid chromatography g gram G glycine H histidine H 4 NAD tetrahydronicotinamide adenine dinucleotide I isoleucine IPTG isopropyl-thio-galactoside IR infrared K lysine k cat turnover number k 1 first order rate constant K d dissociation constant K m Michaelis constant, pseudofirst order binding constant K s spectral binding constant Kan kanamycin KPi potassium phosphate L leucine

PAGE 15

xii LB Luria broth M methionine MgCl 2 magnesium chloride MOPS 3-( N -Morpholino)propanesulfonic acid mV millivolt MW molecular weight m/z mass/charge ratio M micro molar N asparagine NaCl sodium chloride NAD + -nicotinamide adenine dinucleotide (oxidized) NADH -nicotinamide adenine dinucleotide (reduced) NADH:BR NADH:cytochrome b 5 reductase NADH:CR NADH:cytochrome c reductase NADH:FR NADH:ferricyanide reductase NCR NADH:cytochrome c reductase NH 3 ammonia (NH 4 ) 2 SO 4 ammonium sulfate nm nanometer NOS nitric oxide synthase NR nitrate reductase P proline

PAGE 16

xiii PCR polymerase chain reaction pCMB p-chloromercuribenzoic acid PDR phthalate dioxygenase reductase Pfu Pyrococcus furiosus pI isoelectric point Q glutamine R arginine RCM recessive congenital methemoglobinemia S serine SDS sodium dodecyl sulfate SOB sterile osmotic broth T threonine TB terrific broth T m thermal stability constant Tris-HCl tris(hydroxymethyl) am inomethane-hydrochloric acid UV ultraviolet V valine W tryptophan wt wild type Y tyrosine

PAGE 17

xiv Structure-Function Studies of Conserved Sequence Motifs of Cytochrome b 5 Reductase Louis J. Crowley ABSTRACT NADH:Cytochrome b 5 Reductase (c b 5 r) catalyzes the two electron reduction of the iron center of the heme co factor found within cytochrome b 5 (c b 5 ) utilizing reducing equivalents of the nicotinamide adenin e dinucleotide (NADH) coenzyme. C b 5 r is characterized by two domains necessary for proper enzyme functi on: a flavin-binding domain and a pyridine nucleotide-binding do main. Within these domains are highly conserved motifs necessary for the prope r binding and orientation of both the NADH coenzyme and the FAD cofactor. To address the importance of these c onserved motifs site-directed mutagenesis was utilized to generate a series of vari ants upon residues found within the motifs to allow for the full characterizations. Sec ond, naturally occurring recessive congenital methemoglobinemia (RCM) mutants that are f ound within or in close proximity to these highly conserved motifs were analyzed ut ilizing site-directed mutagenesis. The flavin-binding motif 91 RxY S T xx S N 97 was characterized by the generation of variants T94H, T94G, T94P, P95I, V96S, and S97N. In addition to this, the naturally occurring double mutant P92H/E255was fully characterized to establish a role of the P92 residue giving ri se to RCM.

PAGE 18

xv The role of the 124G R xx S T 127 was determined by the introduction of a positive charge, charge reversal, and conserved am ino acid mutations through site-directed mutagenesis of the G124, K125, and M126 resi dues. Based on the data presented here, each of the residues of the G R xx S T motif are directly involved in maintaining the proper binding and orientation of the c b 5 r flavin prosthetic group. Analysis of the NADH-binding motif 273 CGxxx-M 278 was accomplished through the characterization of the type II RCM variant M272and the type I RCM variant P275L. This demonstrates that th e deletion of the M272 residue causes a frame shift leading to the inability of the NADH substrate to bind. The introduction of the P275L variant showed that substrate affi nity was diminished, yet turnover was comparable to wild-type cytochrome b 5 reductase, indicatin g that although P275 is required for proper substrate binding it is no t essential for overall catalytic function. Finally, analysis of the naturally occurring double mutant G75S/V252M provided the first insight into a methemoglobinemia variant that possessed mutations in both the FAD-binding and NADH-binding domains.

PAGE 19

1 1. INTRODUCTION Cytochrome b 5 Reductase NADH: Cytochrome b 5 reductase (c b 5 r, EC 1.6.6.2) is a member of the flavoprotein transhydrogenases [1] which catalyze the transf er of reducing equivalents between the two electron carrying nicotinamide dinucleotides and one electron carriers. Other members of this family include ferrodoxin:NADP + reductase (FNR) [2], plant and fungal NAD(P)H: nitrate reductases [3], NADPH: cytochrome P450 reductase [4], NADPH: sulphite reductase, phthalate dioxygena se reductase [5], nitric oxide synthase [6], and various other flavoproteins. The enzymes found within this flavoprotein family have been demonstrated to comprise bo th flavin-binding domains and pyridine nucleotide-binding domains. The overall function of cytochrome b 5 reductase is to serve as an electron donor for cytochrome b 5 through a hydride ion tran sfer from the reduced coenzyme, NADH, to the non-covalently bound cytochrome b 5 reductase flavin adenine dinucelotide (FAD) cofactor. Cytochrome b 5 reductase is encoded by the DIA1 gene (Figure 1), located on the human chromosome 22q13-qter [7, 8] and transcribes two isoforms of c b 5 r produced by the use of alternative promoters [9, 10]. Each isoform is localized to separate regions of the cell and thus, performs separate functi ons. The first isoform is the microsomal isozyme which comprises the majority of transcribed c b 5 r and is generated through the

PAGE 20

1a1b23456789 35 1a23456789 3 23456789 3 55A.B.C. Membrane-bound cb5r (ubiquitous)Soluble cb5r (erythrocyte) Figure 1. Structure of the Human DIA1 Gene that Encodes Cytochrome b 5 Reductase. The DIA1 gene (A) is composed of exons 1-9 [11]. There are two alternate transcription start sites, 1a and 1b, which yield the two isoforms of cb 5 r. (B) represents the membrane-bound cb 5 r isoform containing exon 1a-9, but not 1b, and consists of 300 amino acids (GAQLCFVF), including the myristoylation signature in exon 1a. (C) represents the soluble cb 5 r isoform containing exons 2-9, but not exon 1, and consisting of the 275 amino acids that comprise the diaphorase domain (LFQRCFVF). transcription of all 9 exons of the DIA1 gene. The microsomal isozyme comprises 300 amino acids residues with a molecular weight of ~34 kDa, and contains a myristoylated 25 amino acid N-terminal hydrophobic region that serves as the membrane anchoring domain [12-14]. This membrane bound isozyme is localized to the cytosolic face of the endoplasmic reticulum as well as the mitochondrial, nuclear, and plasma membranes of somatic cells. The electron transfer process between the microsomal cb 5 r and cytochrome b 5 are required for a variety of metabolic transformations that include: steroid biosynthesis [15], desaturation and elongation of fatty acids, cholesterol biosynthesis [16, 17], and P450-dependent reactions [18]. More recently, it has also been shown to be part of the microsomal cytochrome b 5 reductase-cytochrome b 5 pathway directly involved in the reductive detoxification of arylhydroxylamine carcinogens in human liver [19]. In the second isoform, only exons 2-9 are transcribed from the DIA1 gene producing a 275 amino acid residue soluble form of cb 5 r. Soluble cytochrome b 5 2

PAGE 21

3 reductase is comprised of an N-terminal FAD-binding domain and a C-terminal NADHbinding domain [20, 21]. The NAD-binding domain forms a three-layered fold, which is common for pyridine dinucleotide -binding regions, while the FAD-binding domain has an antiparallel -barrel topology. This isozym e is a flavoprotein with a molecular mass of ~31 kDa is primarily found within the circulating erythrocytes where it is involved in the reduction of methemoglobin to hemoglobin [22, 23] through an interaction with cytochrome b 5 The mechanism of electron transf er for the soluble form of cb 5 r from the physiological substrate, NADH and the FAD cofactor of c b 5 r is shown in Figure 2. In this schematic, two electrons are first tran sferred from NADH to FAD by hydride ion (H ) transfer. The two-electron reduced enzyme-NAD + complex (E-FADH NAD + ) sub sequentially transfers two electrons to two molecules of cytochrome b 5 by the anionic red semiquinone form (E-FAD NAD + ), returns the reduced enzyme to its oxidized state [24]. The reduction of FAD by NADH has been determined to be the rate-limiting step in the electron transfer pro cess catalyzed by cytochrome b 5 reductase [25-27]. In the porcine variant of cytochrome b 5 reductase, it was also determined that the anionic red semiquinone of FAD is stabiliz ed by the binding of NAD + [28, 29]. Further studies by Meyer et al. [30] confirmed that the binding of NAD + is involved in the stabilization of the red semiquinone of c b 5 r and also showed that it modul ates the electron transfer to cytochrome b 5 [30], thus demonstrating the impor tance of the anionic red semiquinone form of c b 5 r (E-FAD NAD + ) in the catalytic process. Recent studies by Iyanagi et al [24] have generated a new scheme for the electron transfer sequence of cytochrome b 5 reductase involving the neutral blue semiquinone form and the oxidized

PAGE 22

NADH + 2cb5(Fe3+) cb5rox(FAD)NAD++ H++ 2cb5(Fe2+) cb5rred (FADH2)A.B.NADH E-FADE-FAD--NAD+ H+e-E-FAD-.NAD+ e-NAD+ (1) Figure 2. Mechanism of Cytochrome b 5 Reductase Mediated Cytochrome b 5 Reduction. (A) represents the chemical reaction (equation 1) catalyzed by cb 5 r. (B) illustrates the mechanism of cb 5 r transfer where each electron (e ) reduces one molecule of cb 5 [31]. E-FAD-NADH-NAD+E-FAD-NADHE-FAD-NADH*E-FADH--NAD+NAD+E-FADNADHH-transferNADHE-FAD-NAD+E-FAD---NAD+E-FADH-NAD+e-H+e(i)(ii)(iii)(iv)(v)(vi)(vii)(viii) Figure 3. New Model of the Reaction Sequence of Cytochrome b 5 Reductase. This new model represents the reaction sequence of cytochrome b 5 reductase containing the neutral blue semi-quinone form and the oxidized enzyme-NADH-NAD + ternary complex as intermediates [24]. 4

PAGE 23

enzyme-NADH-NAD + ternary complex (Figure 3). This scheme involves the following seven-step process: (i) formation of the oxidized enzyme-NADH complex (E-FAD-NADH), (ii) conversion of E-FAD-NADH to a form that has the ability to transfer a hydride ion (E-FAD-NADH), (iii) hydride (H ) transfer from NADH to FAD, (iv) the first one-electron transfer from the two-electron reduced enzyme complex (E-FADH -NAD + ), (v) rapid conversion of the neutral blue semiquinone form (E-FADH-NAD + ) to the anionic red semiquinone form (E-FAD-NAD + ), (vi) the second one-electron transfer from E-FAD -NAD + (vii) formation of the oxidized enzyme-NADH-NAD + ternary complex (E-FAD-NADH-NAD + ) by binding of NADH, and (viii) release of NAD + Similar to the binding interaction between NADH and cytochrome b 5 reductase, the binding interaction between cytochrome b 5 reductase and its one electron acceptor, cytochrome b 5 has also been extensively studied [26, 35-38, 112]. Cytochrome b 5 is also transcribed as both a soluble and microsomal isozyme, comprised of 97 and 133 amino acid residues, respectively, by the CYB5 gene located on human chromosome 18q23 and has a molecular weight of ~11 kDa. The electron acceptor of cytochrome b 5 is the iron center of a single protoheme IX prosthetic group which also gives cb 5 its characteristic red color. Cytochrome b 5 is an acidic molecule with 21 surface accessible carboxylic acid residues and possesses an isoelectric point (pI) of 4.90 [32], whereas the isoelectric point (pI) of cytochrome b 5 reductase is 7.15, thus suggesting a role for charge complimentarity between cytochrome b 5 reductase and cytochrome b 5 complex formation. This was confirmed through studies by Salemme [33] utilizing a best-fit computer program based on the crystallographic data for cytochrome b 5 indicated that 5

PAGE 24

GLU48GLU47GLU52ASP64GLU60Heme propionate group Figure 4. Structure of Cytochrome b 5 Generated in silico. The molecular model of cytochrome b5 displaying the negatively charged residues and exposed heme propionate group determined to be involved in the binding interaction with cytochrome b 5 reductase was generated utilizing the molecular modeling software Web Lab Viewer Pro [34]. This model represents a surface representation of cb 5 (1CYO ) where the acidic residues are red, basic residues are blue, and neutral residues are displayed as white clouds. 6

PAGE 25

LYS162LYS163LYS125LYS41 Figure 5. Structure of Cytochrome b 5 Reductase Generated in silico. The molecular model of cytochrome b5 reductase displaying the positively charged residues K41, K125, K162, and K163that have been determined to be involved in the binding interaction with cytochrome b 5 was generated utilizing the molecular modeling software Web Lab Viewer Pro [34]. This model represents a surface representation of cb 5 reductase (1IB0 ) where the acidic residues are red, basic residues are blue, and neutral residues are displayed as white clouds. 7

PAGE 26

8 electron transfer involved in teractions through the char ge pairing of the surface accessible carboxyl residues E-47, E-48, E-52, E-60, D-64 (shown in Figure 4) and the exposed heme propionate group of the active site. Further confirmation was established by Dailey and Strittmatter [35] utilizing se lective carboxylate group modification to reveal that these same carboxyl groups, lo cated in the active site of cytochrome b 5 were involved in the electron transfer interaction between cytochrome b 5 reductase, NADPHcytochrome-P450, and stearyl-CoA desaturase [ 35, 36]. In order to establish the charge pairing of cytochrome b 5 reductase residues, Ozols and Strittmatter determined that the lysyl residues of cytochrome b 5 reductase played a significant role in the interaction between cytochrome b 5 and cytochrome b 5 reductase. Modification of only seven lysyl residues of cb 5 r resulted in a loss of catalytic activity utilizing c b 5 as the final electron acceptor, whereas no change in catalytic activity was demonstrated through the modification of the lysyl residues on cytochrome b 5 Furthermore, through the selective acylation of several lysyl residues on cytochrome b 5 reductase, cytochrome b 5 reduction was inhibited [37]. Sequence analyses of protease-generated cross-linked peptides demonstrated that four essentia l lysine residues in cytochrome b 5 reductase involved in forming charge pairs with the active site carboxyl groups of cytochrome b 5 and included: K-163, K-41, K-125, and K-162 [38] (Figure 5). Crystal Structure of Cytochrome b 5 Reductase High resolution crystal stru ctures of the soluble, diaphorase domain of R. norvegicus cytochrome b 5 reductase was recently solved in the absence (2.0 PDB 1I7P) and presence (2.3 PDB 1IB0) of bound NAD + [21] (Figure 6). The structures

PAGE 27

9 displayed a classical bi-domain arrangement comprised of an amino terminal FADbinding lobe (amino acid residues I33-R142) and a carboxy-terminal NADH-binding lobe (amino acid residues K172-F300) Residues G148-V171 compri se what has recently been termed the linker or hinge region. This 28 amino acid residue segment forms a three-stranded, anti-parallel -sheet that has been demonstrated to be of critical importance in determining the correct orient ation and modulation of electron transfer between that of the FADand NADH-binding domain [39, 40]. The amino-terminal FAD-binding domain (Fig ure 7) is comprised of a seven-stranded antiparallel -barrel, labeled as F 1-F 7, oriented towards the si -face of the flavin isoalloxazine ring and capped by a single -helix found in the FAD-binding domain. A majority of the interactions with the FADprosthetic group occur via the adenine dinucleotide moiety of the fl avin and with a long loop of the FAD-binding domain which is comprised of amino acid residues K110-K 125 forming a lid and is situated between F 6 and the -helix. The NADH binding domain (Figure 8) is composed of a canonical Rossman fold [41] formed by 3 / / layers arranged into a five-stranded parallel -sheet that is oriented towards the re -face of the isoalloxazine ring of the FAD prosthetic group. The isoalloxazine ring is the only region of the FADcofactor that forms interactions via hydrogen bonds between the NADH-binding site and the isoalloxazine ring. This occurs through the FAD-binding domain loop strands F 3 and F 4 (amino acid residues P92K110) and the NADH-binding domain N 1-N A (amino acid residues T181-T184). The hinge or linker region (Figure 9) is situated be tween the FAD-binding and NADHbinding domains and is comprised of a three-stranded anti-parallel -sheet structure designated as H 1-H 3. Studies of this linker region by Davis et al. [40] demonstrated

PAGE 28

Figure 6. The Tertiary Structure of Rat Cytochrome b 5 Reductase. The 2.0 resolution X-ray crystal structure derived from soluble recombinant rat liver cb 5 r (PDB 1IB0) [21] is illustrated as a schematic model where the red portions represent -helices, the yellow ribbons with directional arrows represent -sheet structures, and the green tubes are indicative of random coils and flexible loop structures. The FAD cofactor (left) and NAD + (right) are shown in stick configurations with standard CPK coloration. 10

PAGE 29

F1FADF1bF6 F1a F2F5LidF3F4 Figure 7. The Structure of the FAD-Binding Domain of Cytochrome b 5 Reductase. A view of the N-terminal FAD-binding domain of rat cb 5 r (PDB 1IB0) with FAD bound. The re-face of the isoalloxazine ring faces towards the right. FAD is shown in stick configuration with standard CPK coloration. The polypeptide backbone is shown in green with the single -helix in red and -sheets colored yellow and are labeled accordingly. 11

PAGE 30

NAD+NDNC N2NBN5N4NAN3N1 Figure 8. The Structure of the NADH-Binding Domain of Cytochrome b 5 Reductase. A view of the C-terminal NADH-binding domain of rat cb 5 r (PDB 1IB0) with NAD + bound. NAD + is shown in stick configuration with standard CPK coloration. The polypeptide backbone is shown in green with the single -helix in red and -sheets colored yellow and are labeled accordingly. 12

PAGE 31

F6 N1 H1H2H3 Figure 9. The Structure of the Hinge Region of Cytochrome b 5 Reductase. A view of the Hinge region of rat cb 5 r (PDB 1IB0) located between the FAD and NADH-binding domains. The polypeptide backbone is shown in green with the -sheets colored yellow and are labeled accordingly. 13

PAGE 32

14 that allows for efficient electron transfer to o ccur. These studies also demonstrated that it is this linker region which is actively i nvolved in properly orienting the FADand NADH-binding lobes of cb 5 r towards efficient electron transfer and suggested that any disruption within the conformation of the -sheet architecture would lead to a significant decrease in catalysis. Recessive Congenital Methemoglobinemia In 1943 Quenton Gibson of Queens Univer sity, Belfast, described the first hereditable trait that was associated with a sp ecific enzyme deficiency [43]. In patients with a familial incidence of idiopathic cyanosis, Hitzenberger later established a diagnosis of congenital, familial methemoglobinemia [42]. Based on this diagnosis, Gibson determined that the inability of circ ulating erythrocytes to reduce the pools of methemoglobin would give rise to familial id iopathic methemoglobinemia [43]. In 1948, Gibson further elucidated the pathway involve d in the actual reduction of methemoglobin which directly involved the enzyme methemogl obin reductase [44], now referred to as NADH: cytochrome b 5 reductase (c b 5 r). Methemoglobin is the derived form of hem oglobin in which the iron center of the heme prosthetic group is in its oxidized state. In the ferric heme state the iron center is unable to efficiently bind oxygen and the oxyge n affinity of the accompanying ferrous hemes in the hemoglobin tetramer is decr eased, thereby impairing oxygen delivery. Methemoglobin is constantly generated physio logically due to the deoxygenation of the molecule, however, endogenous enzymatic hemoglobin reduction pathways maintain methemoglobin levels at or below 3% of total hemoglobin. Inability to efficiently reduce

PAGE 33

15 the levels of methemoglobin gives rise to th e disease termed methemoglobinemia. The primary pathway that converts methemoglobin to hemoglobin involves the reduced form of NADH: cytochrome b 5 reductase. In this pathway, reducing equivalents are transferred from the pyridine dinucleotide, NADH to the FADprosthetic group of c b 5 r, followed by the reduction of tw o molecule of cytochrome b 5 In circulating erythrocytes, cytochrome b 5 directly transfers one electron to th e iron center of the heme prosthetic group in methemoglobin thereby reducing it to hemoglobin. The most common cause of recessive congenital methemoglobinemia (RCM) directly results from a deficiency of cytochrome b 5 reductase. Individuals with RCM have an inability to effectively reduce methemoglobin which is constantly being formed through the deoxygenation of hem oglobin in circulating erythrocytes. Defects in the expression of cytochrome b 5 reductase give rise to two forms of methemoglobinemia known as type I and type II which are classified based on the patient pathophysiology. In type I, the prevalent symptom is mild cyanosis which results in headaches, fatigue and shortness of breath during exercise and is restrict ed to the soluble isozyme of c b 5 r found in circulating erythrocytes. This form of methemoglobine mia is inherited in an autosomal recessive pattern and is found worldwide, however, it has been determined to be endemic in certain popula tions such as the Athabascan Alaskans, Navajo Indians, and Yakutsk natives of Siberia. Compound he terozygotes or homozyg otes that express type I RCM have methemoglobin concentrat ions ranging from 10 to 40% and are relatively asymptomatic where cyanosis is th e only symptom. Life expectancy is not affected in these patients. Type II methemoglobinemia has been established in 10 to15% of cases for RCM

PAGE 34

16 and is the more severe form of the disease in which the deficiencies of c b 5 r are present in the membrane-bound form of c b 5 r in somatic cells. The symptoms of type II are expressed with severe developmental abnormalities including mental retardation, microcephaly, opisthotonus, strabismus, and generalized hypertonia often leading to premature death [45-47]. Cytochrome b 5 reductase is encoded by the DIA1 gene. It is because of this that it is believed that patients expressing type I methemoglobinemia are producing the abnormal gene product at a norm al rate, but it is unstable. Thus, only the mature red blood cells, which can not synthesize proteins, are affected. In patients expressing type II methemoglobinemia, mutations in the DIA1 gene give rise to d ecreased expression or reduced catalytic activity of cytochrome b 5 reductase and the deficiency is ubiquitous. Thus far, about 40 amino acid mutations have been iden tified within the DIA1 gene giving rise to either type I or t ype II methemoglobinemia [10, 31, 48-50]. The Ferredoxin: NADP + Reductase Family of Flavoprotein Oxidoreductases. Through a sequence alignment utilizing the X-ray crystal structure of spinach ferredoxin-NADP + reductase [51], a family of flav in-dependent oxidoreductases have been identified which include: cytochrome b 5 reductase, cytochrome P-450 reductase, methionine synthase re ductase, ferredoxin:NADP + reductase, as well as many others. Of significant interest was the patt ern of conserved residues that were detected within the flavinand nucleotidebinding domains whic h were characteristic of the FNR family (Figure 10). These conserved sequences or motifs are essential for the proper characterization of FNR protei ns that share a common evolut ion, structure, or catalytic

PAGE 35

17 Figure 10. Alignment of Conserved FlavinBinding and NADH-Binding Motifs from Various Enzymes Belonging to th e FNR Family of Flavoprotein Transhydrogenases. Each of the published amino acid sequences was obtained from the GenBank protein database [ 52] with the corresponding acce ssion number and reference. The superscripted numbers indicate the amino ac id positions of the first and last residue of each motif within the full-length sequence of each protein. Abbreviations are: CB5R, cytochrome b 5 reductase, NR, nitrate reductase, FNR, ferredoxin:NADP+ reductase, CPR, cytochrome P450 reductase, NOS, n itric oxide synthase, PDR, phthalate dioxygenase reductase, MSR, methionine syntha se reductase, SR, sulfite reductase, FHG, flavohemoglobin, B5B5R, cytochrome b 5 -cytochrome b 5 reductase fusion protein, ETP, electron transfer protein, NDR, naphtha lene dioxygenase reductase, PH, phenol hydroxylase, NFR, NADPH:flavin reducta se, CDPGR, CDP-glucose dehydratase reductase, CMR, p-cymene monooxygenase reductase, XMR, xylene monooxygenase reductase, NMR, nitrotoluene monooxygenase reductase, DQMR, dihydorquinoline monooxgenase reductase, MMOR, meth ane monooxygenase reductase, ADR, anthranilate dioxygenase reduc tase, TDR, toluate dioxygenase reductase, BDR, benzoate dioxygenase reductase, HBDR, hydroxybenzoa te dioxygenase reductase, NQR1, Na + translocating NADH:quinone reductase 1, UNK1, unknown M. loti gene product, UNK2, unknown Halobacterium sp. NRC-1 gene product, HYPR, hypothetical Xanthobacter reductase.

PAGE 36

18 Enzyme Motif 1 Motif 2 Motif 3 Motif 4 GenBank Accession/Ref CB5R 91 RPYTPVS 97 125 GKMS 128 181 GTGIT-P 186 274 CGPPP-M 279 NP_620232 / [8] NR 722 RAYTPPS 728 755 GVMS 758 805 GTGIT-P 810 898 CGPPP-M 903 P23312 / [53] FNR 139 RLYSIAS 145 176 GVCS 179 217 GTGIA-P 222 318 CGLKG-M 323 P10933 / [54] CPR 454 RYYSIAS 460 488 GVAT 491 534 GTGVA-P 539 629 CGDARNM 635 P16435 / [55] NOS 1178 RYYSISS 1184 1214 GVCS 1217 1255 GTGIA-P 1260 1354 CGDVT-M 1359 P29475 / [56] PDR 51 RNYSLSN 57 83 RGGS 87 123 GIGIT-P 128 202 CGPRPLM 208 Q05182 / [5] MSR 478 RPYSCAS 484 514 GVCT 517 573 GTGIA-P 578 677 CGDKANM 683 Q9UBK8 / [57] SR 386 RLYSIAS 392 420 GASS 423 461 GTGIA-P 466 552 CGDANRM 558 P38038 / [58] FHG 204 RQYSLTR 210 227 GQVS 230 268 GVGQT-P 273 362 CGPVGFM 368 P24232 / [59] B5B5R 289 KPYTPVS 295 323 GLFT 326 364 GTGFT-P 369 458 CGPTPFT 464 NP_596918 / [60] ETP 56 RCYSITS 62 80 GRVS 83 120 GSGIA-P 125 203 CGPEPFM 209 BAA12809 / direct submission NDR 142 RPYSMAG 148 165 GRVT 168 206 GTGLA-P 211 295 CGAPA-M 300 AAD02134 / [61] PH 149 RAFSLAN 155 173 GAAT 176 214 GSGLSSP 220 306 CGPPP-M 311 AAA25944 / [62] NFR 58 RPFSMAS 64 77 GASE 80 123 GTGFSYA 129 212 AGRFE-M 217 AAN83224 / [63] CDPGR 141 RSYSIAN 147 164 GQMS 167 204 GTGFA-P 209 292 CGSPV-M 297 P26395 / [64] CMR 159 RSYSFAN 165 185 GEFT 188 225 GSGLA-P 230 315 CGPPP-M 320 AAB62300 / [65] XMR 160 RSYSFAT 166 184 GIFS 187 224 GTGLA-P 229 315 CGPPP-M 320 AAB70826 / direct submission NMR 161 RSYSFSA 167 185 GVFS 188 225 GTGLA-P 230 316 CGPPP-M 321 AAC38360 / [66] DQMR 155 RSYSPSS 161 179 GAMS 182 221 GTGLA-P 226 310 CGPQP-M 315 CAA73201 / [67] MMOR 159 RSYSPAN 165 183 GRFS 186 224 GTGLA-P 229 314 CGPPG-M 319 P22868 / [68] ADR 154 RSYSFAN 160 178 GVMS 181 218 GTGLS-A 223 307 CGPPP-M 312 AAC34815 / [69] TDR 153 RAYSFSS 159 176 GLMS 179 216 GTGLA-P 221 304 CGPPP-M 309 AAD31449 / direct submission BDR 166 RSYSFSS 172 189 GKMS 192 229 GTGIA-P 234 317 CGPVP-M 322 P07771 / [70] HBDR 152 RAYSYSS 158 175 GKMS 178 215 GTGLA-P 220 305 CGPPP-M 310 Q51603 / [71] NQR1 222 KAYSLAS 228 259 GVCS 262 298 GAGSSFG 304 401 CGPPLHN 407 Q9Z723 / [72] UNK1 168 RLYLVST 174 201 GSSP 204 383 GIGIT-P 388 471 SGPQA-M 476 NP_107088 / [73] UNK2 51 RYYTLSS 57 102 GEPS 105 112 GPGVG-P 117 188 CGAATDA 194 NP_279534 / [74] HYPR 142 RAYSVAN 148 166 GAGT 169 207 GSGLA-P 212 297 AGPAP-M 302 CAA09916 / [75] Consensus RxY T S xx S N Gxx S T (FAD) Rxx S N (FMN) GxGxxP CGxxx-M

PAGE 37

19 function. The structural motif of members of the FNR fam ily is that of a two-domain module with one binding the fl avin and the other binding the pyridine nucleotide. In regards to function, the FNR family is classifi ed by the two electron reduction of a flavin prosthetic group (FAD/FMN) by a pyridine dinucleotid e coenzyme (NAD(P)(H)), followed by the sequential transfer of electrons in one electron reac tions with carriers. An additional feature of the proteins found within the FNR superfamily is the varying levels of complexity based on the structural domains and modulation of electron transfer [2]. Cytochrome b 5 reductase and FNR are examples of the simplest members of the FNR family in that they each consist of only two domains and reduce molecules which are able to bind and then rapidly di sassociate following el ectron transfer. In contrast, cytochrome P450 re ductase, sulfite reductase, and PDR represent more complex enzymes. Although the structures also consist of two domain s, the one-electron acceptor molecule is actually fused to the two-domain core by a linker region. Finally, the most complex members of the FNR family contain additional sequences and prosthetic groups which is indicative of more than three structural domains, these enzymes include: nitrate reductase an d nitric oxide synthase [2]. The conservation of the RxY T S xx S N , G R xx S T , GxGxxP, and CGxxx-M sequence motifs found within the members of the FNR flavoprotein family, clearly demonstrates the importance of these motifs towards the interactions of flavincofactor and pyridine nucleotide binding and utilization.

PAGE 38

20 Research Aims and Approaches The general goal of this research was to gain further insight into the roles of the conserved motifs found within cytochrome b 5 reductase that are involved in either FAD or NADH-binding. Through the analyses of these motifs, naturally occurring methemoglobinemia variants which occur in relation to these motifs were also investigated. To achieve this goal, variants were generated utilizing site-directed mutagenesis of an affinity-tagged R. norvegicus cytochrome b 5 reductase construct [76]. Each mutant enzyme generated was then s ubsequently characteri zed through biophysical, potentiometric, and enzymo logical techniques. The initial specific aim was to furthe r establish the role of the conserved 91 RxY S T xx S N 97 motif in regards to binding of the FADcofactor and overall cytochrome b 5 reductase catalysis. Previous studies had already establis hed a role for the R91, P92, and Y93 amino acid residues [76, 77]. These authors concluded that none of the mentioned residues were essential for pr oper flavin incorpora tion but did in fact modulate the catalytic and thermodynamic properties within the active site of cytochrome b 5 reductase. In order to fully characterize th e remaining residues of this motif, a series of variants were generated co rresponding to amino acids that have been demonstrated to occur at the same residue positions within other members of the FNR family to include: T94H, T94G, T94P, P95G, P95I, V96S, and S97N Through the analysis of the crystal structure of cytochrome b 5 reductase (PDB 1IB0) we hypothesized that the introduction of a mutation to residue T94 would have the mo st significant impact due to the fact that the backbone nitrogen of this residue forms a hydrogen bond with the N5 atom of the FADcofactor isoalloxazine ring and is also positioned near the NADH-binding site. A

PAGE 39

21 second important feature of the T94 resi due was demonstrated in the porcine c b 5 r structure where it was shown that the co rresponding T66 residue was involved in the formation and stability of the NAD + -FAD complex [24]. The naturally occurring double mutant P92H and E255allowed for further characterization of the RxY S T xx S N motif as well as examinat ion of the effects of a methemoglobinemia double mutant occurring in both the F ADand NADH-binding domains. The E255variant was previously de tected [78] in the fa mily with cyanosis first described in 1943 [43] who were found to be deficient in cytochrome b 5 reductase [44]. In the P92H variant, homology modeli ng displayed that the side chain would be oriented towards the FAD isoalloxazine ring and the introduction of a histidine at this position would lead to a steric clash thereby re sulting in the decrease of catalytic function of the enzyme. E255 is positioned within the amino-terminus of an -helix that is involved in properly positioning the carboxyl-terminal region of c b 5 r that contains the CGxxxM motif which has been demonstr ated to be essential for NADH-binding. Deletion of this residue therefor e would result in a decrease in substrate bi nding affinity as shown previously [79]. It can be hypothe sized that the introduction of a double mutant occupying position in both the FADa nd NADH-binding lobes of cytochrome b 5 reductase would lead to the more severe type II RCM. The second aim focused on the highly conserved FAD/FMN specificity motif 124 G R xx S T 127 , specifically the residues G124, K12 5, and M126. The amino acid S127 has been previously characterized through the analysis of the natu rally occurring type II RCM variant, S127P. This mutation was the first RCM variant to be resolved by X-ray crystallography and demonstrated that the introduction of this mutation directly affects

PAGE 40

22 NADH substrate binding due a phenomenon known as flavin flexibility where the FADcofactor of cytochrome b 5 reductase is displaced into the NADH-binding cleft [80]. Through further analysis of the cr ystal structure of cytochrome b 5 reductase the critical importance of the G R xx S T motif can been visualized. Th is motif is located within the only -helix of the FADbinding lobe of cb 5 r and each residue of the motif forms either a hydrophobic or electrostatic interaction with the FAD prosthetic group. Thus, it is important to fully develop an understanding of the molecular ro le of each residue of the G R xx S T motif in regards to the flavin environm ent as well as proper cofactor binding and orientation. The third aim of this research involve d the investigation into the conserved 273 CGxxxM 278 motif which is located within th e NADH-binding domain of cytochrome b 5 reductase and is involved in the regulation of NADH-binding. A different approach was used for the analysis of this motif. Two methemoglobinemia variants have been previously described that ar e directly involved in the CGxxxM motif, the type II RCM variant M272[81] and type I RCM varian t P275L [78] and it is through these two mutations that the role of the motif was established. Although M272 is not technically considered to be part of the CGxxxM it has been found to be highly conserved in higher eukaryotes. The importance of studying the role of this re sidue was twofold, in that it directly precedes th e C273 residue which has been f ound to be essential towards NADH-binding affinity and orientation [12] and the deletion of M272 has also been shown to give rise to type II RCM. Upon the deletion of M272, a frame-shift is introduced to the motif that should cause a detrimental effect upon the affinity for the NADH substrate to the extent that it is capable of causing type II RCM. Further analysis

PAGE 41

23 of the molecular role of M272 involved the generation of variants of M272 based on amino acid residues that have been demonstr ated to occur at the equivalent position within other members of the FNR superfamily. Based on the crystal structure of cytochrome b 5 reductase in complex with NAD + a structural role for amino acid residue P275 was demonstrated [21]. It was shown that the conformation of the NAD + was primarily determined by the proline side chain with the pyrophosphate chain packed against P275 and the NAD + nicotinamide proximal ribose was packed parallel to the proline ring. We predicted that th e replacement of this proline side chain with a leucine would have a significant adverse effect upon the proper orientation and binding of the NADH substrate. The characterization of the P275L variant would therefore provide insight into the role of th e P275 residue in regards to NADH substrate binding and orientation as well as providing a molecular basis for disease. The final specific aim was based upon the identification and characterization of a novel FAD-binding domain variant, G75S. The G75S mutation was found in a patient who was compound heterozygous for the G75S mutation as well as the previously characterized V252M mutation [ 82-84] and exhibited a phenotype characteristic of type I methemoglobinemia. It was hypothesized th at possessing a mutati on in both the FADand NADH-binding domains of cytochrome b 5 reductase would result in the development of the more severe type II methemoglobinemia. Although th e patient was heterozygous for the double mutation, the G75S/V252M mutation was generated in order to investigate the synergistic effects of th e double mutant on the enzymes structural and catalytic properties. The generation of both the singl e and double mutants being associated with

PAGE 42

24 the FAD and NADH-binding domains could ther efore be utilized to provide further insight as an aid to determine the development of type I versus type II RCM status.

PAGE 43

25 2. MATERIALS & METHODS MATERIALS Molecular Biology Reagents Restriction enzymes were purchased from New England Biolabs (Bev erly, MA). Kits for plasmid preparation and agarose gel extr action were purchased from Qiagen Inc. (Valencia, CA). Oligonucleotide primers were obtained from Integrated DNA Technologies (Coralville, IA). Native Pfu and Pfu Turbo Polymerases as well as Epicurian coli BL21(DE3)-RIL cells were obtained fr om Stratagene (La Jolla, CA). The pET-23b vector was purchase d from Novagen (Madison, WI). T4 DNA ligase was purchased from Promega (Madison, WI). Rapi d DNA ligation kits were purchased from Roche Scientific (Palo Alto, CA). Triton X-100 and Hot Start Micro 50 PCR tubes were obtained from Molecular-Bio Products Inc. (San Diego, CA). DMSO and pCMB were obtained from Sigma Chemical Co. (St. L ouis, MO). The Molecular Biology Core Facility at the H. Lee Moffitt Cancer Center and Research Institute performed nucleotide sequencing.

PAGE 44

26 Microbiology and Protein Purification Reagents Plastic and glassware were purchased from Fisher Scientific (Pittsburg, PA), VWR International (Suwannee, GA), and Midwest Scientific (St. Louis, MO). Tryptone and yeast extract were obtained from EM Science (Gibbstown, NJ) and Ultrol grade MOPS buffer was purchased from Calbiochem (La Jolla, CA). IPTG and Ampicillin were obtained from Research Products International (Mt. Prospect, IL). Ni-NTA agarose was purchased from Qiagen Inc. (Valencia, CA). PMSF was purchased from Sigma Chemical Co. (St. Louis, MO). Enzyme Assay and Spectroscopy Reagents Chemicals including NADH, NAD + 5-ADP-agarose, 2-ADP, ADP-ribose, K 3 Fe(CN) 6 glucose, dithionite, FAD, cytochrome c (Type VI from horse heart), bovine serum albumin, potassium phosphate, riboflavi n, ferric citrate and trifluoroacetic acid were obtained from Sigma Chemical Co. (S t. Louis, MO). Tetrahydronicotinamide adenine dinucleotides (H 4 -NAD and H 4 -NADP) were synthesized according to the protocol described by Murata liev and Feyereisen [85] by bubbling molecular hydrogen through a stirred solution of 100 mM NAD + or NADP + contained in 50 mM Tris-HCl buffer, pH 8.0, in the presence of palladium catalyst (5 mg/mL). Sinapinic acid (3,5dimethoxy-4-hydroxy cinnamic acid) was purc hased from Aldrich Chemical Co. (Milwaukee, WI).

PAGE 45

27 METHODS Protein Expression and Purification Expression of the wild-type and mutant c b 5 r or NCR variants was accomplished using E. coli BL21(DE3)-RIL cells harboring either the pH4CB5R [76], or mutant constructs. Cells were grown aerobically in TB media supplemented with riboflavin (100 M), ferric citrate (100 M), and ampicillin (125 M) overnight at 37 C, and recombinant protein expression was induced by addition of IPTG (0.4 mM) followed by an additional 6-8 h incubation at 25 C. The cells were pell eted by centrifugation (5000xg, 10 min.), resuspended in lysis bu ffer (50 mM Tris-HCl, containing 300 mm NaCl and 5 mM imidazole, pH 8.0) and disrup ted by sonication in the presence of PMSF (1.2 mg/ml). Lysates were clarified by centrifugation (30,000xg, 30 min.) at 4 o C followed by incubation with Ni-NTA agarose with gent le agitation (1 ml matrix/10 ml lysate) for up to 1h at 4 C. The His-tagged-(c b 5 r or NCR)-Ni-NTA matrix suspension was collected by centrifugation (1000xg, 5 min.), washed twice with 25 mM phosphate buffer, containing 300 mM NaCl and 5 mM imid azole, pH 8.0 and transferred to a chromatography column (2.5 x 10 cm). Bound proteins were eluted with 25 mM phosphate buffer, containing 300 mM NaCl and 250 mM imidazole, pH 8.0. Fractions were pooled, assayed for NADH:FR activity a nd concentrated centr ifugally using 10,000 MWCO concentrators (Fischer Scientific, Pittsburg, PA). Fi nal purification was achieved by size-exclusion FPLC using either a Superdex 75 or Superdex 200 column (1 x 30 cm) equilibrated with 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. Fractions exhibiting NADH:FR activity were p ooled, concentrated, beaded and stored

PAGE 46

28 under liquid N 2 until required. Wild -type NCR concentrations were estimated using 413nm = 130 cm -1 mM -1 and wild-type c b 5 r concentrations were estimated using 461nm = 10.6 cm -1 mM -1 The recombinant soluble, heme domain of rat cytochrome b 5 was produced as described by Beck-von Bodman et al ., [86] using an expressi on construct provided by Dr. Steven Sligar (University of Il linois, Urbana, IL). Briefly, E. coli BL21(DE3) cells harboring the pUC18 expression construct, which contained a codon-optimized gene encoding the soluble heme-containing domain of rat c b 5 were grown overnight at 37 C in TB medium supplemented with ferric citrate (100 M) and ampicillin (125 M). Protein expression was induced at 25 C by addition of IPTG (0.4 mM), and the cells were harvested 8 hours later by centrifugation, yielding pink pellets. Lysis of the cells was achieved as described for c b 5 r. During fractionation of the cleared lysate, however, c b 5 was precipitated by 95% (w/v) (NH 4 ) 2 SO 4 The pellet was resuspended in the minimum volume of low ionic strength phosph ate buffer and dialy zed twice against a large volume the same buffer before separation on DE52 anion exchange chromatographic media. Elution was achieved by a gradient of 0-1M NaCl before final purification by FPLC on a Superdex 75 10/30 HR column. Red fractions containing c b 5 protein were pooled, concentrated, beaded, and stored under liquid N 2 SDS-polyacrylamide gel electrophoresis was performed as described by Laemmli [87] using 12.5% acrylamide/b is-acrylamide gels. 2-5 g of protein was solubilzed in up to 30 L of SDS loading buffer (10 mM Tris -HCl, pH 6.8, contai ning 1% SDS, 10% glycerol, 1 mM DTT, and 0.1% bromophenol blue) by boiling in water for 5 minutes before loading the gel. Gels were run at 150 volts in SDS running buffer (25 mM Tris-

PAGE 47

29 HCl, pH 8.3, containing 192 mM glycine, 0.1% SDS) until the blue dye front traversed the gel. Gels were stained with Coomassie Brilliant Blue dye to visualize protein bands. Site-directed Mutagenesis The pH4CB5R [76] expression construc t was specifically mutagenized using a modification of the Stratagene QuikChange (La Jolla, CA) protocol. Mutagenic Oligonucleotide Primer Design Complementary oligonucleotide primers (30-40 mers) containing the desired codon change as well as silent mutation (either inserting or deleting a restriction enzyme recognition sequence) were designed using the Primer Generator program ( http://www.med.jhu.edu/medcenter/primer/primer.cgi ) [88]. The sense strand annealing primers used in the construction of the vari ous mutant proteins ar e listed in Table 1. Vector Polymerase Chain Reaction Vector PCR was performed using Pfu Turbo polymerase (1.25 units) in the presence of cloned Pfu buffer (20 mM Tris-HCl, pH 8.8 containing 2 mM MgSO 4 10 mM KCl, 10 mM (NH 4 ) 2 SO 4 0.1% Triton X-100, and 0.1 mg/ml nuclease-free BSA) containing 10 ng vector DNA (pH4CB5R or pH6NCR), 125 ng of each synthetic primer, 5% DMSO, and 200 M dNTPs with cycling paramete rs (20 cycles) of 1 min at 94 C, 1 min at 55 C, 10 min at 68 C. DpnI restriction enzyme was added directly to the cooled PCR reaction tube following the PCR cycling reaction to cleave only the methylated template DNA. This step greatly redu ced the percentage of wild-t ype background transformants.

PAGE 48

Table 1. Mutagenic Oligonucleotide Primers. Nucleotides in bold encode the mutated bases generated to introduce the proper mutation. Silent mutations in italics add (+) or eliminate (-) the indicated restriction site. Mutation Primer Sequence (5 3) Restriction Site G75S CCT GGG CCT TCC TAT CAG TCA ACA CAT CTA CCT +HincII P92A GGC AAC TTG GTC GCC TAC ACC CCT GTG TCT -BsmFI P92H TGG GTG ATT CGT CAC TAC ACC CCT GTG TCT AGT +HinFI T94G GTC ATT CGT CCC TAC GGC CCT GTG TCT AGT GAT -MslI T94H GTC ATT CGT CCC TAC CAC CCT GTG TCT AGT GAT -MslI T94P GTC ATT CGT CCC TAC CCC CCT GTG TCT AGT GAT -MslI P95A CCC TAC ACC GCG GTG TCT AGT GAC GAT GAC AGG +MaeIII P95G ATT CGT CCC TAG ACC GGT GTG TCT AGT GAT GAT +BsaWI P95H GTC ATT CGT CCC TAC ACC CAC GTG TCT AGT GAT +AflIII P95I ATT CGT CCC TAC ACC ATT GTT TCT AGT GAT GAT -MslI V96S ATT CGT CCC TAC ACC CCT TCG TCT AGT GAT GAT -MslI S97N CGT CCC TAC ACC CCA GTG AAT AGT GAT GAT GAC +HincII G124A GTT TCC AGC CGG AGC GAA AAT GTC TCA GTA C -BpmI G124H GTT TCC AGC TGG ACA CAA AAT GTC TCA GTA CC -BpmI G124K GTT TCC AGC TGG AAA GAA AAT GTC TCA GTA CCT -BpmI G124R GTT TCC AGC TGG ACG GAA AAT GTC TCA GTA CC -BpmI K125A GCG GGA GGG GCA ATG TCT CAG TAC CTG -BpmI K125D CCC AAG TTT CCA GCA GGC GAC ATG TCT CAG TAC +NspI K125E CCA GCG GGA GGG GAA ATG TCT CAG TAC CTG +NspI M126C GCT GGA GGG AAA TGT TCT CAG TAC CTG GAA AAC +BsmAI M126F AAG TTT CCA GGG AGG GAA ATT CTC TCA GTA CCT +ApoI M126G GCT GGA GGG AAA GGG TCT CAG TAC CTG GAA AAC +BsaI M126P GCT GAA GGG AGG GAA ACC GTC TCA GTA CCT GGA +BsmBI M126S GCT GGA GGG AAG TCG TCT CAG TAC CTG GAA AAC +BsmBI E255TTC GTG AAT GAG ATG ATC AGG GAC CAT CTT CCA +HincII V252M GCC AAG GCT TCA TGA ATG AGG AGA TGA TCA GGG +BspHI M272GAG ACA CTG ATA CTG TGT GGA CCC CCA CCG NA* M272A GAG ACA CTG ATA CTG GCA TGT GGA CCC CCA CCG +NspI M272I CTG ATA CTG ATC TGT GGA CCC +TfiI M272L CTG ATT CTG CTG TGT GGA CCC +TfiI M272R CTG ATT CTG AGG TGT GGA CCC +TfiI P275L CTG ATG TGT GGA CTC CCA CCG ATG ATC CAG +PleI 30

PAGE 49

31 PCR products were subsequently purified us ing the Qiagen PCR Cl eanup Kit. Products were eluted in 30 L of sterile elution buffer and stored at -20 C until ready for use. Preparation of Competent Cells and Linear DNA Transformation Competent E. coli DH5 cells were prepared according to the protocol by Yuckenberg et al. [89] for storage at -70 C. In brief, an overnight culture of bacteria, grown in LB, was used to inoculate 250 mL of fresh LB medium. The culture was incubated with shaking until an O.D. 600nm of approximately 0.9 wa s reached. Cells were then harvested by centrifugation at 0 C and washed in a solution of ice-cold 100 mM MgCl 2 Cells were again harvested and resuspended in ice-cold 100 mM CaCl 2 where they were then incubated on ice for 90 minutes. Cells were finally harvested and resuspended in 85 mM CaCl 2 containing 15% glycerol before being snap frozen in liquid N 2 and stored at -80 C prior to use. This protocol was also used for the preparation of all other competent E. coli strains. Competent E. coli DH5 cells were transformed w ith linear DNA constructs according to the protocol provided with the Quikchange Mutagenesis Kit (Stratagene) Frozen cells were thawed slowly on ice and 50 L aliquots were transf erred to sterile 15 mL polypropylene conical tubes ma intained on ice. Up to 10 L of the purified linear DNA construct was mixed carefully with the cells by swirling the pipette tip as the DNA was expelled. The mixture was then incubate d on ice for 20 minutes before being heatshocked at 42 C for 45 seconds and then returned to ice for 2 more minutes. Finally, 1 mL of SOB media was added and the cells were allowed to recover for 1 hour at 37 C with gentle shaking before being plat ed on the appropriate selection media.

PAGE 50

32 Mutagenic Screen by DNA Restriction Digest Following overnight incubation at 37 C on ampicillin-containing SOB-agar plates, colonies were subcultured into 5 mL of liquid SOB media, cont aining ampicillin (125 M). Clones were analyzed and screened for the silent mutations, introduced during mutagenesis, by restriction enzyme digesti on using the appropriate enzymes listed in Table 3 and 4. Each digest was separated using polyacrylamide gel electrophoresis or agarose gel electrophoresis and the restriction pattern visualized by EtBr staining of the gel. Figure 14 illustrates the result of a typical gel observed during screening of mutants. Mutant constructs exhibiting the predicted patt ern that differentiated them from the wildtype construct were verified by nucleotide sequencing in both the forward and reverse directions. Positive constructs were subsequently used to transform competent E. coli BL21(DE3)-RIL cells to proliferate th e generation of the mutant enzymes. Ultraviolet and Visible Absorbance Spectroscopy Absorbance spectra in the far (190-250 nm) and near UV (250-350 nm), visible (360-750 nm), and near IR (750-1000 nm) wave length range were obtained for wild type and mutant enzymes using a Hewlett Packard (Agilent Technologies, Palo Alto, CA) 8453 diode-array spectrophotometer utiliz ing microcuvettes of either 90 L or 200 L capacity and with 1 cm pathlength. Ultraviolet and Visible Circu lar Dichroism Spectroscopy UV and visible CD spectra were obta ined using a JASCO (Easton, MD) J710 spectropolarimeter calibrated for both signal intensity and wavelength maxima using an

PAGE 51

33 aqueous solution of d-10-camphosulfonic acid [90]. UV CD spectra were obtained in 10 mM phosphate buffer, containing 0.1 mM EDTA pH 7.0 using a cylindrical quartz cell of 0.1 cm path length (300 l total volume) while visible CD spectra were obtained in 10 mM MOPS buffer, containing 0.1 mM EDTA, pH 7.0 using a 1 cm pathlength cell (90 l total volume). All spectra were corrected for the appropriate buffer contributions and are expressed in terms of molar ellipticities (M -1 cm -1 ). Fluorescence Spectroscopy Fluorescence spectra were obtained using a Shimadzu Scientific Inst. Inc. RF5301PC spectrofluorophotometer. Excitation and emission spectra were obtained using a slit width of 3 nm and emission and exci tation wavelengths of 520 nm and 450 nm, respectively. Following acquisition of the wild type and mutant spectra, the enzymes samples were heated to 100 o C for 30 min, centrifuged to rem ove protein aggregates and the corresponding spectra for the liberated F AD were subsequently recorded. All spectra were corrected for the appropr iate buffer contribution. Steady-State Enzyme Activities NADH:FR and NADH:CR activitie s were determined at 25 C under conditions of constant ionic strength and pH in 116 mM MOPS buffer ( = 0.05), containing 0.1 mM EDTA, pH 7.0, while NADH:BR activ ities were determined at 25 C under conditions of constant ionic strength and pH in 10mM potassium phosphate buffer ( = 0.01), containing 0.1mM EDTA, pH 7.0. NADH:FR activities were typically determined as the

PAGE 52

34 decrease in absorbance at 340 nm in the presence of NAD(P)H (250 M) and ferricyanide (500 M). NADH:BR activity was determined as the increase in absorbance at 423 nm in the presence of NADH (250 M) and rat c b 5 (30 M). All substrates were quantitated spectroscopically using the calculated molar extinction coefficients for NADH, H 4 NAD, NAD + c b 5 2-5, and ADP-ribose (Siegel 1959, Beck von Bodman 1986, The Merck Index 1989, Murataliev 2000). Activ ities are expressed as initial rates for the oxidation of NAD(P)H ( mol of NAD(P)H consumed/min/nmol FAD). Initial rate data at varying NAD(P)H, ferricyanide, or cb 5 concentrations were analyzed using the software ENZFIT (Elsevier Biosoft, Ferguson, MO) to determine apparent k cat and K m values. Charge-transfer Complex Determination Reductive titrations were performed under anaerobic conditions as described by Foust [91]. Enzyme samples (50 M FAD) in 116 mM MOPS bu ffer, containing 0.1 mM EDTA, pH 7.0, and NADH solutions (100 mM) we re prepared by repeated evacuation and flushing with oxygen-free argon. Enzyme samples were titr ated with NADH and monitored for increased absorbance in th e near IR wavelength range from 700-900 nm. Titrations were determined to be satura ting when the absorbance at 800 nm ceased to increase and no further bleaching of the flavin spectra occurred upon NAD(P)H addition. Spectral Binding Constant Determin ation by Differential Spectroscopy Spectral binding constants, K s for various NADH analogs were determined by differential spectroscopic titrations as desc ribed by Sancho and Gomez-Moreno [92] and

PAGE 53

35 Barber et al [93]. A Shimadzu UV2501-PC was used to measure the difference spectra of samples in matched split -cell quartz cuvettes with 0.439 cm pathlength and 2.5 ml capacity. Enzyme and NADH analogs were suspended in 10mM phosphate buffer, pH 7.0, containing 0.1mM EDTA for all spectral dete rminations. All analog concentrations were determined spectrophotometrically using previously published extinction coefficients. The reference cuvette contained 50 M enzyme (as determined by FAD concentration) in the front compartment and analog-titrated buffer in the rear compartment. The sample cuvette co ntained analog-titrated enzyme (50 M) in the front compartment and buffer in the rear compartm ent. The appropriate buffer was added to the enzyme compartment of the reference c uvette to correct for the dilution caused by analog addition to the enzyme compartment of the sample cuvette. Following sample addition, each compartment was stirred brie fly to insure thorough mixing. Spectra were then recorded over the wave length range from 300-800 nm de pending on the titration. The magnitude of the absorbance change was determined for each spectrum via peak to trough subtraction in order to compensate for any baseline drift between spectra. Absorbance changes were plotted versus nucle otide concentration and the resulting plots were fit to the hyperbolic equation: Absorbance = Absorbance max *[nucleotide] / [ K s + (nucleotide)] in order to determine the spectral binding constant ( K s ), which was defined as the concentration of analog at which half maximal spectral perturbation was observed.

PAGE 54

36 Thermal Stability Measurements Thermal stabilities of wild type and mutant enzymes were determined as described by Trimboli et. al [67] by monitoring both the re lease of the FAD prosthetic group, as indicated by the change in flav in fluorescence, and the loss of NADH:FR activity over the temperature range from 0-100 C. Determination of Flavin Midpoint Potential Flavin midpoint potentials were determin ed by dye equilibration using the method of Massey [94]. Xanthine (30 M) and xanthine oxidase (50 nM) were used to reduce a mixture of enzyme (40 M FAD) and the indicator dye, phenosafranine (15 M) in 100 mM phosphate buffer containing 0.1 mM EDTA first made anaerobic by repeated evacuation and flushing with oxyge n-free argon. Benzyl viologen (6 M) and methyl viologen (1 M) were included to facilitate e quilibration of the system. Visible absorbance spectra were collected over the course of each 3-hour equilibration. Flavin reduction was monitored at 410 nm while pheno safranine reduction was monitored at 530 nm. E 0 values were calculated by graphical analysis of the plot [log (ox)/(red)] phenosafranine versus [log (ox)/(red)] FAD using the published midpoint pot ential of phenosafranine of 252 mV [94].

PAGE 55

37 3. RESULTS AND DISCUSSION Analysis of the conserved RxY S / T xx S / N FAD binding motif Multiple sequence alignments of the members of the flavoprotein transhydrogenases family have revealed four distinct regions, ea ch containing a high level of sequence and structural conservati on. These regions are found within both the flavinand pyridine nucleotide-b inding regions of the ferrodoxin:NADP + reductase (FNR, EC 1.18.1.2) superfamily (Figure 10) [5]. The first of these conserved sequences, corresponding to RxY S T xx S N , is a flavin-binding motif comprised of the residues 91 RPYTPVS 97 in R. norvegicus cytochrome b 5 reductase sequence. It is considered to be one of the most highly conserved of the known flavin-b inding motifs. Site-directed mutagenesis studies of R 91, P92, and Y93 have demonstrated these residues to be non-essential for flavin cof actor incorporation. However, amino acid residues R91 and Y93 do modulate the spect roscopic, catalytic, and thermodynamic properties of c b 5 r, and are critical in maintaining pr oper structure and or ientation of the active site of the enzyme [76, 77]. Additi onally, formation and stability of the NAD + FAD semiquinone complex during NADH turnover has also recently been shown to be controlled by the fourth residue in the RxY S T xx S N motif, corresponding to T66 in porcine cytochrome b 5 reductase, through the generation of the T66S and T66V variants [24].

PAGE 56

38 To further probe the structural and functional characteristics of this highly conserved flavin-binding motif, a series of variants were created through site-directed mutagenesis, utilizing the rat cytochrome b 5 reductase expression system previously generated by Marohnic et al. [76]. The variants generated corresponded to alternate amino acid residues that occur at the same residue position within additional GenBank sequences of other members of the FNR s uperfamily and include: T94H, T94G, T94P, P95G, P95I, V96S, and S97N. The fidelity of each variant was confirmed by nucleotide sequencing in both forward and reverse directions and the mu tant proteins were then recombinantly expressed in BL21 (DE3)-RIL cells. Purifi cation to homogeneity was carried ou t through the utilization of a co mbination of Ni-NTA affinity and gel filtration chromatography for the mutant proteins and with the exception of T94G and T94P, all were expressed at levels comparable to that of the wild-type domain as observed by the appearance of single protein bands following SDS-PAGE analys is. Neither the T94G nor T94P mutants yielded viable proteins for subsequent characterization, in each case the flavin cofactor failed to properly incorporate to yield a stable flavoprotein. UV/visible absorption spectra were obtained for oxidized samples of each of the purified variants together with wild-type cytochrome b 5 reductase and are presented in Figure 11A. The P95G, P95I, V96S, and S97N mutants each displayed spectra comparable to that of the wild-type domain, attributable to protei n-bound flavin with an aromatic absorption maxima detected at 272 nm in the UV region of the spectrum and a peak at 461 nm with a pronounced shoulder with in the range of 485-500 nm in the visible region of the spectrum. The T94H mutant however, produced visible absorption spectra

PAGE 57

260300340380420460500540580 Absorbance 0.00.20.40.60.81.01.21.41.61.8 420440460480500520 0.000.150.30 190250 Molar Ellipticity (x106) -4-2024 Wavelength (nm) 300350400450500550 Molar Elli p ticit y ( x104 ) -4-2024 ABC Figure 11. Ultra-Violet, Visible, and Circular Dichroism Spectra of Wild Type, T94H, P95G, P95I, V96S, and S97N Proteins of Cytochrome b 5 Reductase. (A) Oxidized samples of wild-type and mutants cb 5 rs (20M), (B) 7M, and (C) 60M FAD in 10mM phosphate with 0.1mM EDTA, pH 7.0 buffer. The inset shows an expanded region of the visible spectrum where the flavin cofactor makes a major contribution. Individual spectra correspond to ( ____ ) H 4 cb 5 r; ( ____ ____ ) T94H; ( __ __ __ ) P95G; ( _ _ ) P95I; ( . . ) V96S; ( __.__.__ ) S97N and free flavin ( _____ ). 39

PAGE 58

40 that was significantly altered with a red-shifted max of 467 nm compared to that of the wild-type cytochrome b 5 reductase visible max of 461 nm. Shifts within the visible absorption spectra of flavoproteins have prev iously been attributed to changes in the hydrophilicity of the flavin e nvironment near the N(5) locu s of the isoalloxazine ring [76]. The absorption ratios (A 276 /A 461 ) for each viable mutant were within a range of 5.5 to 5.9 which indicated a full flavin complement. To examine the effects of each mutation upon the secondary structure of the protein, UV circular dichroism spectra were determined in the UV wavelength range (190 300 nm). As shown in Figure 11B, all of the mutants generated displayed line shape comparable to wild-type cytochrome b 5 reductase with positive CD from 190-210 nm and negative spectra from 210-250 nm. Although th ese results were in good agreement with the wild-type domain, both the T94H and V96S variants displayed a negative CD intensity greater than that of the wild-type protein at 222 nm. The alteration displayed for the T94H and V96S UV CD sp ectra indicated an adverse structural change within the overall -helical or -sheet components of the two variants. In order to examine the flavin environment of each variant, visible CD spectra were obtained as shown in Figure 11C. Representative spectra demonstrated that each variant retained a similar line shape as that of wild type cytochrome b 5 reductase with positive maxima at 310 and 390 nm and negative maxima at 460 and 485 nm, re spectfully. However, both the T94H and P95G variants displayed an altered flavin environment within the region of 450 485 nm which is in agreement with the observation from the x-ray crystal structure, that T94 forms a hydrogen bond with the isoalloxazine ring of the FADprosthetic group and the mutation of P95 to a glycine could lead to flexibility within the secondary structure

PAGE 59

41 altering the FADcofactor c onformation. No significant ch anges were observed in the visible CD spectra for either the V96S or S97N variants demonstrating that neither residue is required for corre ct flavin orientation. To establish if the alterations observed in the flavin environment from the UV/Vis and CD spectra affected the catalytic function of the enzyme, initial-rate kinetic parameters for NADH:FR and NADH:BR assays were determined for the wild-type cytochrome b 5 reductase and for each purified variant. The values obtained for the NADH:FR and NADH:BR assays are presented in Table 2. NADH catalytic efficiencies ( k cat / K m NADH ) for all variants ranged between 22 and 62% compared to wild-type cytochrome b 5 reductase, with T94H yielding th e lowest catalytic efficiency. Interestingly, it was only the k cat for each of the variants that lead to the decrease observed in the overall catalytic efficiency, while the K m NADH of each mutant was either comparable to the wild-type domain or slightly enhanced as seen in the T94H variant indicating a higher affinity for the NADH substr ate. Similar results were observed in the NADH:BR assays in which the overall catalytic efficiency was decreased primarily due to the decreased rate of turnover ( k cat ) for each variant. This was to be expected since the residues within the RxY S T xx S N motif are not involved in or located within close proximity to the amino acids that participate in cytochrome b 5 reductase cytochrome b 5 complex formation. From the kinetic data obtained it was demonstrated that each of the variants generated displayed a K m NADH that was highly comparable to wild-type cytochrome b 5 reductase or, as mentioned previously, in the case of the T94H the affinity for substrate was enhanced almost 2-fold over the wild-typ e domain. To further establish the binding

PAGE 60

42 Table 2. NADH:FR and NADH:BR Kinetic Co nstants of T94H, P95G, P95I, V96S, S97N, and Wild Type Cytochrome b 5 Reductase. Mutant NADH:FR NADH:BR k cat K m NADH K m FeCN6 k cat / K m NADH (s -1 ) (M) (M) (s -1 M -1 ) k cat K m NADH K m Cyt b 5 k cat / K m Cyt b 5 (s -1 ) (M) (M) (s -1 M -1 ) H 4 c b 5 r T94H P95G P95I V96S S97N 800 21 6 1 8 1 1.4 0.3 x 10 8 102 21 4 1 7 1 2.8 0.1 x 10 7 326 25 7 1 8 1 4.2 0.2 x 10 7 232 10 6 1 8 1 4.0 0.1 x 10 7 637 27 8 1 7 1 8.0 0.6 x 10 7 296 5 5 1 8 1 5.7 0.6 x 10 7 600 10 2 1 13 1 4.7 0.5 x 10 7 94 2 2 1 3 0.3 3.0 0.3 x 10 7 244 15 3 1 11 2 2.3 0.5 x 10 7 137 6 3 1 12 1 1.2 0.2 x 10 7 465 20 2 1 16 1 3.0 0.4 x 10 7 567 36 3 1 24 1 2.4 0.4 x 10 7 affinity of the substr ate NADH and the product NAD + spectroscopic binding constants were determined utilizing differential spectroscopy. As stated in Methods and Materials, the isoteric analog 1,4,5,6-tetrahydro-NAD (H 4 NAD) was utilized as an alternative substrat e analog to evaluate the affinity for the physiological substrate NADH [85]. Differential spectroscopy was used to monitor complex formation during titrations with either H 4 NAD or the physiological product NAD + as shown in Figure 12 and Figure 13. Spectra obtained for each of the titrations carried out in the presence of H 4 NAD displayed a line shape comparable to wild-type c b 5 r with the exception of the P95I variant. This indicated that for P95I, the isoteric substrate analog bound in an altered conformatio n due to the distorted environment of the FADprosthetic group as obser ved in the visible CD spectra. Values observed for the spectroscopic binding constant ( K s ) for each of the H 4 NAD titrations, shown in Table 3, demonstrated that the binding affinity for each of the variants was not significantly

PAGE 61

Absorbance -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 [H4NAD] (mM) 0.00.10.20.3 (A 450 -A 400 ) 0.000 0.025 0.050 380435490545600 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 [H4NAD] (mM) 0.00.30.6 (A 480 -A 395 ) 0.00 0.01 0.02 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 [H4NAD] (mM) 000.40.8 (A 410 -A 460 ) 0.000 0.025 0.050 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 [H4NAD] (mM) 000.30.6 (A 450 -A 395 ) 0.00 0.01 0.02 Wavelength (nm) 380435490545600 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 [H4NAD] (mM) 000.40.8 (A 480 -A 395 ) 0.000 0.011 0.022 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 [H4NAD] (mM) 0.00.40.8 (A 480 -A 395 ) 0.00 0.02 0.04 A B D E C F Figure 12. Spectroscopic Titrations of H4c b5r and the T94H, P95G, P95I, V96S, and S97N Variants in the Presence of H4NAD. Titrations of all mutants (50mM) were performed in split cell op tical cuvettes in 10mM phospha te buffer containing 0.1mM EDTA, pH 7.0 at 23 C. Difference spectra were recorded following the addition of solution containing H4NAD (30M). The inset panel co rresponds to a plot of the magnitude of the spectral perturbations at the indicated wavelengt hs versus pyridine nucleotide concentration where a difference spec trum was observed. Plots of the relative absorbance changes observed are as follows: (A) H4c b5r; (B) T94H; (C) P95G; (D) P95I; (E) V96S; and (F) S97N. 43

PAGE 62

-0.04 0.00 0.04 0.08 0.12 Absorbance -0.06 0.00 0.06 0.12 0.18 [NAD+] (mM) 0.00.10.20.3 (A 410 -A 460 ) 0.00 0.04 0.08 B A Wavelength (nm) 380435490545600 -0.04 0.00 0.04 0.08 0.12 [NAD+] (mM) 0.01.22.4 (A 410 -A 460 ) 0.00 0.03 0.06 C -0.06 0.00 0.06 0.12 [NAD+] (mM) 000.51.0 (A 410 -A 460 ) 0.000 0.035 0.070 -0.02 0.00 0.02 0.04 0.06 [NAD+] (mM) 012 (A 410 -A 460 ) 0.000 0.012 0.024 380435490545600 -0.02 0.00 0.02 0.04 0.06 [NAD+] (mM) 0.01.22.4 (A 410 -A 460 ) 0.000 0.015 0.030 D E F [NAD + ] (mM) 024 (A 410 -A 460 ) 0.00 0.04 0.08 Figure 13. Differential Spectroscopy of H4c b5r, T94H, P95G, P95I, V96S, and S97N c b5rs in the Presence of H4NAD. Titrations of all mutants (50mM) were performed in split cell optical cuvettes in 10mM phosphate buffer containing 0.1mM EDTA, pH 7.0 at 23 C. Difference spectra were recorded following the addition of solution containing NAD+ (30M). The inset panel corresponds to a plot of th e magnitude of the spectral perturbations at the indicated wavelengths versus pyridine nucleotide concentration where a difference spectrum was observed. Plots of the relative absorbance changes observed are as follows: (A) H4c b5r; (B) T94H; (C) P95G; (D) P95I; (E) V96S; and (F) S97N. 44

PAGE 63

45 Table 3. Spectral Binding Constants Obtained for Wild Type c b 5 r and the T94H, P95G, P95I, V96S, and S97N Variants. Mutant K s H4NAD (M) K s NAD+ (M) K i NAD+ H 4 c b 5 r 45 10 760 30 675 23 T94H 122 16 13 2 23 2 P95G 53 13 491 15 360 16 P95I 116 4 196 14 225 14 V96S 62 8 1503 140 ND S97N 114 11 775 121 ND altered, with the P95G and V96S mutants yielding values compar able to wild-type cytochrome b 5 reductase and the T94H, P95I, and S97N displaying K s values ~ 2-fold greater than that of the w ild-type domain values. These results indicated that the a ffinity for the substrate analog H 4 NAD is not significantly affected through the perturbations observed in the spectroscopic properties of the FAD prosthetic group. Un like the spectra seen in the H 4 NAD titrations, all of the generated variants retained a line shape for the NAD + titrations compared to that of wildtype cytochrome b 5 reductase displaying positive maxima at 410 and 510 nm, and negative maxima at 430, 460, and 485 nm. However, K s values were markedly altered for the P95G, P95I, and T94H variants displa ying an enhanced binding affinity for the NAD + product with the T94H variant displaying a K s value of 13 M compared to that of the wild-type domain value of 800 M (Tab le 3). Values obtained for the V96S and S97N variants were ~ 2-fold greater than or comparable to wild-type cytochrome b 5 reductase, respectively. These results are in good agreement with the kinetic values obtained in that the substrate H 4 NAD is still able to bind efficiently as is the product NAD + However, the enhanced binding affinity of NAD + for the T94H, P95G, and P95I

PAGE 64

46 variants combined with the reduced k cat kinetic values, indicated that the NAD + is unable to dissociate from the NADH binding domain effi ciently which has been demonstrated to be a critical step in the react ion mechanism of cytochrome b 5 reductase [24]. To further probe the properties of produc t inhibition, initial rate enzyme kinetics were performed, as described previously, in the presence of vary ing concentration of NAD + to determine if the competitive inhibition properties of the NAD + product was capable of giving rise to the alterations obser ved in the overall catalytic efficiencies and K s values of the T94H, P95G, and P95I variants. As shown in Figure 14, the Lineweaver-Burke plots obtaine d for wild-type cytochrome b 5 reductase and the T94H, P95G, and P95I mutants all displayed a mode l of competitive inhibition represented by a single intercept at the y -axis showing no change in V max in the presence of varying concentrations of NAD + K i values [ x -intercept = -( K i )] were established for the wildtype cytochrome b 5 reductase and corresponding varian ts through the secondary plot of the K m values for each assay against that of the concentration of the NAD + inhibitor. Table 3 shows the K i values obtained which corresponded to 675, 23, 360, and 225 M for the wild-type domain, T94H, P95G, and P95I respectively. These results are in good agreement with the cata lytic efficiencies and K s values obtained and confirmed that NAD + is not able to disassociate from the co mplex efficiently and thus acts as an inhibitor in the cytochrome b 5 reductase reaction pathway. Protein stability was evaluated by monitoring the thermal NADH:FR inactivation profile coupled with the increase in intrinsi c flavin fluorescence and emission intensity of each of the variants which was compared to values obtained for wild-type cytochrome b 5 reductase. The results obtained for the ther mal denaturation profiles and change in

PAGE 65

47 Figure 14. Enzyme Inhibi tion Assessment with NAD + for the T94H, P95G, and P95I Variants. Panels A-D represent the Lineweaver-B urke plots obtained from the initial rate kinetic analyses in the presence of various concentrations of NAD + Concentrations of NAD + correspond to ( ) 0 M; ( ) 50 M; ( ) 100 M; ( ) 250 M; ( ) 500 M; and ( +) 1000 M for panel (A) [wild-type c b 5 r], ( ) 0 M; ( ) 25 M; ( ) 50 M; ( ) 100 M; and ( ) 500 M for panel (D) [P95I], and ( ) 0 M; ( ) 25 M; ( ) 50 M; ( ) 100 M; () 250 M; and ( + ) 500 M for panels [B and C] [T94H and P95G], respectively. Panel (E) repres ents a replot of the observed K m NADH determined from each Lineweaver-Burke plot in (A-D) versus the concentration of inhibitor (NAD + ) where the inhibition constant ( K i ) can be determined by the equation: x -intercept = -( K i ), and the y intercept = K m

PAGE 66

1/[NADH] ( M -1 ) -0.4-0.3-0.2-0.10.00.10.20.30.40.50.6 0.0 0.2 0.4 0.6 0.8 1.0 1.2 -0.4-0.3-0.2-0.10.00.10.20.30.40.5 0.0 0.3 0.6 0.9 1.2 1.5 1.8 -0.2-0.10.00.10.20.3 0.0 0.1 0.2 0.3 0.4 0.5 -0.2-0.10.00.10.20.3 1/V max ( M min -1 nmol -1 FAD) 0.00 0.04 0.08 0.12 0.16 D [NAD + ] (mM) -0.9-0.6-0.30.00.30.60.91.2 Observed K m ( M) 0 25 50 75 100 A B C E 48

PAGE 67

Temperature (oC) 0102030405060708090100 Residual NADH:FR Activity 0255075100% Fluorescence Intensity 0255075100 Figure 15. Temperature Stability Profiles of cb 5 r and the T94H, P95G, P95I, V96S, and S97N Mutants. Oxidized samples of T94H, P95G, P95I, V96S, S97N, and wild-type H 4 cb 5 r (5 M FAD) were incubated at the indicated temperatures, and aliquots were withdrawn and assayed for both residual NADH:FR activity (closed symbols) and intrinsic flavin fluorescence (open symbols) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0 using excitation and emission wavelengths of 450 nm and 523 nm, respectively. Points correspond to: (,) H 4 cb 5 r; (, ) T94H; (, ) P95G; (, ) P95I; (, ) V96S; and (+) S97N. 49

PAGE 68

50 Table 4. T 50 Values of the Wild-Type Cytochrome b 5 Reductase, T94H, P95G, P95I, V96S, and S97N Variants. Mutant T 50 (C) H 4 c b 5 r 56 T94H 50 P95G 43 P95I 47 V96S 50 S97N 50 intrinsic flavin fluorescence are shown in Figure 15. T 50 values (the temperature at which 50% of maximum fluorescence and 50% retention of NADH:FR activity was detected) for each of the mutants indicated a decreased protein thermostability compared to the T 50 value of 56 C for wild-type cytochrome b 5 reductase. Both the P95G and P95I variants showed the greatest decrease in thermostability with T 50 values of 43 C and 47 C, respectively, while the remaining T94H, V96S, and S97N variants yielded a T 50 value of 50 C. These results indicated that any s ubstitution of the P95 residue may lead to conformational changes in the proteins secondary structure resulting in protein instability. To examine the effects of potential stru ctural changes on th e properties of the flavin prosthetic group, oxidation-reduction potentials for the FAD cofactor were determined utilizing dye-equilibration titrations in the presence of phenosafranine (E = 252 mV) since the variants generated we re based on residues within the RxY S N xx S T motif and thus could affect th e flavin reduction potentials due to their close proximity to the FADcofactor. Flav in midpoint potentials (E n = 2) for the FAD/FADH 2 couple

PAGE 69

51 were determined for the variants alone and in the presence of NAD + Spectra obtained during a representative titration of th e T94H variant are shown in Figure 16. The flavin redox potentials (n = 2) for the wild-type cytochrome b 5 reductase and generated variants for the FAD/FADH 2 couple were determined from the Nernst semi-log plots of the log ([oxidized]/[reduced]) FAD versus potential (mV) and are shown in Figure 16. The standard midpoint potentia l values obtained are shown in Table 5. For the standard flavin midpoint potentials in the ab sence and presence of NAD + all potentials were nearly identical for the wild-type domain, P95G, and S97N variants, indicating that the introduction of structural altera tions of these two variants did not have a dramatic effect on the flavin midpoint potential. The V96S va riant displayed a flav in midpoint potential comparable to the wild-type domain in the absence of NAD + indicating that the mutation of V96S had no effect on the electron transfer pot ential. Upon the addition of the product NAD + however, a negative shift was observe d yielding a midpoint potential -217 mV, comparable to the value obtained for the midpoint potential of free flavin (-220 mV) [95]. Much the opposite was observed with the P95I variant, in that in the presence of NAD + a midpoint potential of -193 mV was established which is in good agreement with the wildtype cytochrome b 5 reductase value of -190 mV. In the absence of NAD + the P95I mutant displayed a positive shift by 65 mV to -207 mV, a shift that, as mentioned, demonstrates a midpoint potential like that s een in the potential of free flavin. T94H yielded the most significant resu lts by displaying a midpoint pot ential of -219 mV both in the absence and presence of NAD + indicating the introduction of a T94H variant causes a structural displacement of the flavin prosth etic group that leads to standard midpoint potential corresponding to that of free flavin.

PAGE 70

Potential (mV) -280 -260 -240 -220 -200 -180 -160 -140 -1.0 -0.5 0.0 0.5 1.0 -300 -280 -260 -240 -220 -200 log (oxidized/reduced) [FAD] -1.0 -0.5 0.0 0.5 1.0 A B Figure 16. Potentiometric Titrations of the T94H, P95G, P95I, V96S, S97N, and Wild-Type Cytochrome b5 Reductases. Reductive dye titrations were performed at 25 C as described in Materials and Methods using phenosafranine as the indicator dye in 100 mM phosphate buffer containi ng 0.1 mM EDTA, pH 7.0. Nernst plots in the absence (A) and presence (B) of 2 mM NAD+. Plots correspond to ( ) H4c b5r; ( ) T94H; ( ) P95G; ( ) P95I; ( ) V96S; and (+) S97N. 52

PAGE 71

53 Table 5. Flavin Midpoint Potentials (E ) Obtained for Wild-Type, T94H, P95G, P95I, V96S, and S97N c b 5 rs. E FAD/FADH2 (mV) Slope (mV) E FAD/FADH2 (mV) Slope (mV) Enzyme -NAD + +NAD + H 4 c b 5 r T94H P95G P95I V96S S97N -272 -30 -218 -60 -267 -61 -207 -57 -276 -30 -272 -35 -190 -30 -219 -31 -201 -50 -193 -32 -217 -23 -197 -55 Summary of Analysis of the conserved FAD binding motif 91 RxY S / T xx S / N 97 Previous studies based on the cr ystal structure of NADH-cytochrome b 5 reductase derived from pig liver micr osomes described by Nishida et al. [39] revealed a highly conserved -barrel structure for the binding motif of the flavin prosthetic group through the structural homology comparisons of the porcine c b 5 r, ferredoxin-NADP + reductase (FNR) [96], phthalate dioxygenase reductas e (PDR) [5] and the FADcontaining fragment of nitrate reductase [6]. Through these comparisons it was revealed that the overall barrel foldings are similar to each othe r and that three conser ved residues, (R, Y, and S/T) were in a specific arrangement a nd necessary for correct flavin binding and orientation. These three residues correspond to R91, Y93, and S127 within the rat model of cytochrome b 5 reductase and provided the first insight into the highly conserved 91 RxY S T xx S N 97 flavin-binding motif. Since the desc ription of this motif, mutagenesis studies have been performed in order to elucidate the roles of these residues in maintaining proper flavin incorporation and overall function. Previous studies by Marohnic et al [76] demonstrated that although R91 is not essential for flavin

PAGE 72

54 incorporation, the removal of the positively ch arged side chain results in the phenomenon known as flavin flexibility. In this case, production of the R91A and R91L variants resulted in the displacement of the ADP por tion of the flavin pros thetic group into the ADP region of the NADH-binding si te, thus acting as its own inhibitor. Analysis of the P92 and Y93 variant indicated the roles in which these residues modulated the flavin midpoint potential, spectrosc opic properties of the prot ein-bound cofactor, and the catalytic efficiencies of cytochrome b 5 reductase [77]. Mutageni c studies of the porcine T66V and T66S variants (corresp onding to T94 in rat cytochrome b 5 reductase) by Kimura et al [24] established that this residue is involved in the stabilization of semiquinone intermediates within the reaction mechanism of cytochrome b 5 reductase. In order to fully elucidate the role of each residue, a series of variants were generated corresponding to naturally occurr ing substitutions observed within the FNR superfamily [5] that included T94G, T94H, T94P, P95G, P95I, V96S, and S97N. The results presented here give fu rther insight into the role of this motif towards correct orientation and binding of the FAD prosthetic group. While the V96S and S97N variants gave results that were comparable to that of wild-type cytochrome b 5 reductase, the T94H, P95G, and P95I variants all displayed significant deviations in the biophysical properties of the enzyme. T94 has been considered to be a critical residue within the 91 RxY S T xx S N 97 motif due to its location within the FAD-binding lobe of cytochrome b 5 reductase. As shown in Figure 17, T94 forms a backbone hydrogen bond to the N5 of the isoalloxazine ring of the FAD prosthetic group and is situated near the potential binding site of the nicotinamide ring of NADH [97, 98]. T94 also lies within 4.02 of residue T184 of the

PAGE 73

55 Figure 17. Electrostatic Interaction of Amino Acid Residue T94 with the FAD Cofactor of Cytochrome b 5 Reductase. (A) Ligplot [99] of 1IB0. C, O, N, S, and P atoms are represented as white, blue, red, yellow, and violet spheres, respectively, while covalent bonds are green sticks within FAD and blue sticks within amino acid residues of the FAD-binding lobe. Hydrogen bonds are dr awn as dashed lines with distances between atoms labeled. Residues cont ributing to hydrophobic interactions are represented as arcs with rays and colored blue. (B) Structural m odel of residue T94 and its orientation and hydrogen bonding interact ion to FAD (atoms colored in CPK).

PAGE 74

A 3.11 T94 B 56

PAGE 75

57 NADH-binding lobe that forms essential hydr ogen bond contacts from the OG1 atom of T184 to both the O4 and N5 of the isoallo xazine ring of the FADcofactor. The importance of the P95 residue has b een demonstrated through the resolved crystal structure of cytochrome b 5 reductase [21]. P95 is lo calized in a hydrophobic core of the FAD-binding domain, together with L72 and V105. This triad is approximately 6 away from the active site making it a cr itical area in regard s to enzyme function. Homology modeling software (GENO3 D) [100] was utilized to generate a structural model of the T94H, P95G, and P95I variants in order to compare the structural models with the resu lts obtained Figure 18. As described previously, various attempts were made to generate the T 94G and T94P variants which re sulted in limited success. The introduction of a glycine residue into a protein has been shown to relax the backbone secondary structure, thus the P95G variant may result in the perturbed flavin environment as observed in the visible CD sp ectra. The introduction of a proline causes the opposite effect, introducing a kink into the backbone. Both of these mutations may lead to an altered conformation of the protein. This demonstr ates that the position of the T94 and P95 residues at the carboxy terminus of the F 4 strand and the hydrogen bonding of the T94 residue with the FAD isoa lloxazine ring is necessary for correct flavin incorporation. Although there is no side chain interac tion involved with the T94 residue the introduction of a histidine imidizole group removes the polarity of the T94 hydroxyl group and replaces it with a positively charged, bulky side chain. We predicted that the effects of this would have a si gnificant effect on the environment of the flavin prosthetic group and results obtained from the red shift seen in the Uv/visible spectra and the

PAGE 76

P95 T94 V96 AT94H BP95I DP95G C Figure 18. Structure of the Wild-type c b5r, T94H, P95G, and P95I Variants Generated in silica. The models of wild-type cytochrome b5 reductase with T94, P95, and V96 and the T94H, P95G, and P95I mu tants were generated using the modeling software GENO3D and SWISS-MODEL [100, 101] Panel (A) is a structure of 1IB0 [21] focused in on th e region of the RxYS TxxS N motif, specifically on amino acid residues T94, P95 and V96, where the polypeptide backbone is cyan and the FAD cofactor and amino acid residues are colored by element (N=blue and O=Red). Panel B is the GENO3D theoretical structure of the T94H variant, where the FAD cofactor is colored in yellow and T94H in green. Panel C represents the P95G variant co lored in orange, and Panel (D) represents the P95I variant colored in red. 58

PAGE 77

59 perturbed visible CD spectral da ta clearly demonstrated this. We proposed that the T94H variant generated a steric hindrance that affect s the interaction of the T94 residue with the FAD isoalloxazine ring as well as the hydrogen bonding interaction of the T184 residue with the N5 atom of the FADprosthetic group isoalloxazine ring. The alteration in the FAD environment was further confirmed by the results of the oxidation/reduction potential of the FAD/FADH 2 couple which yielded a midpoint potential comparable to that of free flavin (-220 mV) in the absence and presence of NAD + suggesting an unfavorable electron transfer efficiency. As mentioned previously, T94 and P95 are located near the poten tial binding site of the nicotinamide ring of NADH and the effects of the T94H, P95G, and P95I variants upon substrate binding yielded in teresting results. The spectra obtained for both the isoteric analog H 4 NAD and the product NAD + displayed a line shape comparable to that of wild-type cytochrome b 5 reductase indicating that no severe conformational changes occurred in the binding of either H 4 NAD or NAD + However, upon examination of the spectroscopic binding constants ( K s ) for the T94H and P95I variants, the binding affinity for H 4 NAD had increased almost 3-fold compar ed to that of the wild-type domain showing a reduced binding affin ity for the analog, whereas the K s values obtained for the product NAD + were significantly decreased for all three variant with the greatest impact observed of the T94H variant yielding a K s value of 13 M compared to that of 667 M obtained for wild-type cytochrome b 5 reductase. Thus the sp ectral binding constants suggest that the NADH substrate bi nds with an altered efficiency compared to that of the wild-type domain yet the NAD + product is not able to read ily disassociate from the NADH-binding site and therefore acts as a potential inhibitor in the cytochrome b 5

PAGE 78

60 reductase reaction pathway. The diaphorase ac tivities associated w ith the T94H, P95G, and P95I variants were the fi rst indication that this hypothe sis was correct, in that the overall catalytic efficiency ( k cat / K m ) was only 22%, 32%, and 31%, respectively, that of wild-type cytochrome b 5 reductase, yet this was only due to the poor rate of turnover ( k cat ) while the K m NADH was comparable to that of wild-type cytochrome b 5 reductase for each mutant. Competitive inhibition assays fu rther confirmed this by determining initial rate enzyme kinetics in va rying concentrations of NAD + Lineweaver-Burke plots generated from these assays revealed a model of competitive inhibition for the three variants as indicated by the increase in K m NADH with no change in the V max K i values obtained from these assays were also in agreem ent, being significantly lower than that of wild-type. These results are in good agreem ent with those in the previously reported T94V variant [24], in that the blue neut ral semiquinone of FAD (FADH) has a hydrogen atom on the N5 atom of the isoalloxazine ring, and release of a proton from the N5 position is required for conversion of the bl ue neutral semiquinone to the anionic red semiquinone [102] and the introduction of the T9 4H, P95G, and P95I variants would lead to a conformational change and hindrance towa rds the release of th e proton from the N5 position of the FADH semiquinone thus causing a perturbed effect on the cytochrome b 5 reductasecytochrome b 5 electron transfer process. Type I recessive congenital methemoglobinem ia associated with a double mutation, the novel P92H and previously de scribed E255-, in the cytochrome b 5 reductase gene Recently, a family was identified wher e two siblings suffered from type I RCM suggesting that both parents were carriers. Screening of the DIA1 gene for molecular

PAGE 79

61 defects in the 2 year old sister detected a novel homozygous C to A mutation in exon 4 and a heterozygous 3 bp in frame deletion of GAG in exon 9, causing a loss of glutamic acid at position 255. Thus, one allele uniquely possessed a double mutation. The exon 4 mutation predicted a proline to histidine exch ange at residue 92, co rresponding to a P92H variant. In addition, the sa me amino acid change at a ne arby residue, corresponding to P95H, had previously been described in a pati ent with type II RCM [103]. The relative locations of both of these residues in the c b 5 r structure, with respect to the FAD prosthetic group and the complexed NAD + together with indicati ons of the degrees of conservation of these residues in different cb 5 rs, is shown in Figure 19. The E255mutation has been previously de tected [78] in the family fi rst diagnosed with cyanosis described in 1943 [104], who were subsequently found by Gibson in 1948 to be deficient in cb 5 r [44] Since half of the c b 5 r enzyme synthesized would ha ve mutations in both the flavinand substratebinding domains, it would suggest a severe cl inical phenotype type I. With this knowledge, utilizing the previous ly characterized hete rologous rat expression system [105] could provide insight into the function of naturally occurring variants of c b 5 r [106] [79] [107], we prep ared both the P92H and doubl e P92H/E255variants. By performing initial-rate kineti c, thermostability, substrat e affinity, and thermodynamic studies we were able to assess the impact of each mutation on the function of c b 5 r. In addition, we also characterized the P95H vari ant and compared the results to the data previously generated for the E255mutant [79]. Sequencing of the DIA1 gene in a cyanotic patient detected a heterozygous deletion of GAG, bases 27,100 to 27,102 (N CBI accession number NT_011520) in exon 9 resulting in the lo ss of E255 (Figure 20). This mutation has been previously described

PAGE 80

62 Figure 19. X-ray Crystallographic Structure of Cytochrome b 5 Reductase. (A) A ribbon diagram of the st ructure of the rat c b 5 r diaphorase domain in complex with NAD + (PDB=1IB0) is depicted showing the structur al elements that comprise the FADand NADH-binding lobes and the connecting hi nge region. The FAD prosthetic group and the bound NAD + are shown in the stick representation with the individual atoms colored using the standard CPK coloring scheme. The positions of the three mutated residues, corresponding to P92, P95 and E255, are shown in ball and stick representation. (B) Gra phical representation (sequence lo go) of the amino acid residues present at positions 89 to 99 in the primary sequences of 56 cb 5 r sequences deposited in GenBank. A multiple sequences alignment was constructed using ClustalX [108], manually adjusted for maximum sequence conservation and used for the logo generation. The logo consists of stacks of symbols (one stack for each position in the sequence) with the overall height of the stack indicating th e sequence conservation at that position and the height of the symbols within the stack indicating the relative frequency of each amino acid at that position [109]. (C) Graphical representation of the amino acid residues present at positions 251 to 259 in the primary sequences of 56 c b 5 r sequences deposited in GenBank.

PAGE 81

63 B P92 E255 P95 NAD + FAD 89 90 91 92 93 94 95 96 97 98 99 Residue Number C A 251 252 253 254 255 256 257 258 259 Residue Number

PAGE 82

64 as compound heterozygous with the G291D muta tion in the original family with type I RCM disease [78] as described by Gibson [44]. In addition, a homozygous C to A change at base 16,076 (C16,076A) in exon 4 predicting an amino acid change of to histidine at codon 92 (Figure 20). Uniquel y, in this case one allele carried a double mutation. The P92H mutation is located in the FADbinding lobe of c b 5 r while the E255mutation is present in the carboxy-terminal NADH-binding domain. Consequently, 50% of cb 5 r synthesized would have mutati ons in both the prosthetic groupand the substrate binding lobes of th e enzyme. To investigate how the double mutation would impact on the function of c b 5 r, the P92H and E255variants were synthesized individually a nd as a P92H/E255double muta nt using a heterozygous expression system. The plasmid encoding either the wild-type or the variant forms of the soluble, diaphorase domain of rat c b 5 r, corresponding to residues I33 to F300, was used to transform E. coli BL21 (DE3)-RIL cells. The cells we re disrupted by sonication and the five (P92A/H, P95A/H, and P92H/E255-) c b 5 r variants were purified to homogeneity by a combination of metal-affinity chromatography and gel filtration FPLC as previously described [21]. Evaluation of the expression yiel d of the various P92 and 95 variants indicated that all the proteins were efficiently expressed with yields comparable to those obtained for the wild-typ e domain. All five c b 5 r variants were purifi ed to homogeneity as evident by the presence of single protein bands following SDS-PAGE analysis of the final FPLC fractions as shown in (Figure 21), which also confirmed molecular masses (M r ) of approximately 32 kDa for all five variants, identical to the wild-type c b 5 r. UV/visible absorbance spectra were obt ained for oxidized samples of the five

PAGE 83

Figure 20. Detection of the G255and P92H Mutations in the DIAI Gene. Sequencing of exon 9 of the DIA1 gene revealed a heterozygous deletion of GAG from bases 27,100 to 27,102 (NT_011520) in the child (A) as indicated by arrow in chromatogram, when compared to the wild type sequence (B). In addition, a homozygous base change of C to A at nucleotide 16,076 was detected in exon 4 in the child (C) indicated by arrow in chromatogram when compared to normal sequence (D). Bases are as follows: G = black; A = green; T = red; C = blue. 65

PAGE 84

ABCDEF50372575 Figure 21. SDS-PAGE Analysis of the Different P92, P95 and E255 cb 5 r Variants. Samples (2 g protein) obtained following isolation of wild-type cb5r (A), the P92A (B), P92H (C), P95A (D), P85H (E) and the P92H/E255double mutant (F) were analyzed using a 12.5 % polyacrylamide gel as described in Methods. The outer lanes correspond to protein molecular weight markers with the indicated molecular masses (kDa). purified P92 and P95 variants and were compared with the spectra obtained for the corresponding wild-type domain in (Figure 22A). The P92A and H and the P95A and H cb 5 r variants exhibited spectra identical to that of the wild-type domain that were characterized by and absorption maximum detected at 273 nm in the UV range of the spectrum, and a peak at 461 nm with an associated pronounced shoulder in the range of 485-500 nm in the visible region of the spectrum, the latter peak attributable to proteinbound flavin. The A 273 nm/461 nm absorbance ratios of the P92 and P95 variants were within the range of 5.5 0.2 which was comparable to values previously obtained for wild-type rat cb 5 r of 5.6 0.2 [77], indicating a full complement of the FAD prosthetic group. 66

PAGE 85

Wavelength (nm) 250300350400450500550600 Absorbance 0.00.20.40.60.81.01.21.41.6 A 190260 Molar Ellipticity ( x 106) -4-2024 Wavelength (nm) 300350400450500550600 Molar Ellipticity ( x 104) -4-2024 CB 320355390425460495530 0.000.150.30 Figure 22. UV/Visible Absorption and CD Spectra of cb 5 r and the Various P92, P95 and E255 Mutants. (A) UV/visible absorption spectra were obtained for oxidized samples of cb 5 r and the various P92, P95 and E255 mutants at equivalent flavin concentrations (1.7 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. The inset shows an expanded region of the visible spectrum where the flavin prosthetic group makes a major contribution. Individual spectra correspond to P92A ( ----); P92H ( .. ); P95A ( -.-.); P95H ( -); P92H/E255(-..-..-) and wild-type cb 5 r ( ____ );. (B) UV CD spectra were recorded using enzyme samples (7 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. (C) Visible CD spectra were recorded using enzyme samples (50 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. Line styles shown in B and C are the same as those depicted in A. 67

PAGE 86

68 To assess any alterations in the secondary structural c ontent of the different P92 and P95 c b 5 r variants, circular dichroism spectra were recorded in the UV wavelength range (190-300 nm). As shown in Figure 22B, all five P92 and P 95 variants exhibited positive CD spectra from 190-210 nm and nega tive CD spectra from 210-250 nm with the spectra retaining both positiv e and negative intensities very similar to that of the wildtype domain. The absence of any significant differences between the spectra of various mutants and the wild type domain suggested conservation of the secondary structure architecture and that none of the amino acid substitutions or deletions had any substantial effects on the folding of the protein. Visi ble CD spectroscopy was utilized to examine the environment of the FAD prosthetic group. As shown in Figure 22C all five variants generated exhibited visible CD spectra that were virtually i ndistinguishable from that of wild-type cb 5 r, indicating that none of the amino aci d substitutions or deletions had any significant effect on the conformatio n of the bound FAD prosthetic group. To examine the influence of the various amino acid substituti ons on the stability of the resulting proteins, thermal denaturation profiles were generated for the P92A and H, P95A and H and P92H/E255variants by measuring both changes in the intrinsic flavin fluorescence emission intensity ( ex = 450 nm, em = 523 nm) and retention of NADH:FR activity following incubation of the protein at temperat ures ranging from 0100 C (Figure 23). Changes in the intrinsic fluores cence of the cofactor or the retention of NADH:FR activity following thermal denatu ration was an effective indicator of the stability of the core stru cture of the protein. T 50 values (the temperat ure at which 50% of maximum fluorescence and 50% retention of NADH:FR activity was detected) ranged from a low of 41 C for the P 92H/E255mutant to a high of 57 C for the P92A variant.

PAGE 87

69 These values can be compared to a T 50 of 56 C obtained for wild-type c b 5 r. Initial-rate kinetic analyses we re performed for the P92 and P95 c b 5 r variants to evaluate the effects of the various residue substitutions and deletions on both NADH and cytochrome b 5 utilization. Values derived for the various kinetic constants for both the NADH:FR and NADH:BR activities are given in Table 6. The single variants of P92A and H and P95A and H demonstr ated a minor reduction in the overall catalytic efficiency compared to that of wild-type c b 5 r for both the NADH:FR and NADH:BR assays whereas in the P92H/E255variant the k cat (s -1 ) was decreased four fold for each assay and the K m for NADH was increased 33 fold. Thes e results for the P92H/E255variant correlated with the previously published data on the E255mutant [79] indicating that a deletion of residue E255 in the carboxy terminus of c b 5 r significantly affects NADH substrate utilization. To compare the interaction of the different P92 and P95 c b 5 r variants with either H 4 NAD or NAD + differential spectroscopy was utilized to monitor complex formation. Examples of alterations of the flavin visible absorbance difference spectrum following nucleotide binding are shown in Figure 24. The formation of spectrally-detectable complexes was observed for the diaphorase doma in during titrations of the majority of the mutants with both H 4 NAD and NAD + For H 4 NAD, the tetrahydronicotinamide derivative did not function as a hydride donor when substituted for NADH in either the NADH:FR or NADH:BR assays bu t provided a valuable tool for estimating the binding affinity for NADH. The H 4 -nucleotide is a close isosteri c analog and is assumed to involve the same contacts at the active site as NADH, but lacks the positive charge on nicotinamide ring that is present on NAD + The spectral changes observed following

PAGE 88

Temperature (oC) 0102030405060708090100 % Residual Activity (NADH:FR) 0255075100Fluorescence Intensity (%) 0255075100 Figure 23. Thermal Stability Profiles for the Various P92, P95 and E255 cb 5 r Mutants. Oxidized samples of the various P92, P95 and E255 variants (5 M FAD) together with wild-type cb 5 r in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0, were incubated at the indicated temperatures, aliquots withdrawn and assayed for both residual NADH:FR activity (closed symbols) and intrinsic flavin fluorescence (open symbols), the latter using excitation and emission wavelengths of 450 nm and 523 nm, respectively. The plots correspond to, P92A (+, +; T 50 =55+ 1 o C), P92H (, ; T 50 =45+ 1 o C), P95A (, T 50 =47+ 1 o C), P95H (, T 50 =46+ 1 o C), P92H/E255(, T 50 =42+ 1 o C) and wild type cb 5 r (, ; T 50 =55+ 1 o C). NADH:FR activities are relative to a sample of each protein maintained at 0 o C. Excitation and emission spectra were scaled relative to that of a sample of free FAD (5 M) which was assigned a fluorescence intensity of 100 %. 70

PAGE 89

Table 6. Kinetic Constants Obtained for the Different P92 and P95 cb 5 r Mutants. NADH:FR activity NADH:BR activity k cat K m NADH K m FeCN 6 k cat /K m NADH k cat K m b5 k cat /K m b5 Mutant (s -1 ) (M) (M) (s -1 M -1 ) (s -1 ) (M) (s -1 M -1 ) cb 5 r 800 + 21 6 + 1 8 + 1 1.4 + 0.3 x 10 8 600 + 16 13 + 3 4.9 + 1.3 x 10 7 P92A P92H P95A P95H P95H/E255 458 + 7 571 + 6 420 + 9 367 + 6 190 + 15 9 + 1 12 + 1 12 + 1 28 + 1 293 + 44 8 1 7 1 8 1 7 1 7 1 5.0 + 0.5 x 10 7 4.6 + 0.2 x 10 7 3.7 + 0.4 x 10 7 1.3 + 0.1 x 10 7 7.5 + 1.8 x 10 5 404 + 7 470 + 12 375 + 7 275 + 8 112 + 6 14 1 15 1 14 1 13 1 15 1 2.9 + 0.3 x 10 7 2.7 + 0.2 x 10 7 2.1 + 0.2 x 10 7 1.6 + 0.2 x 10 7 7.5 + 0.9 x 10 6 H 4 NAD binding were identical to those previously detected for the corresponding titrations of the rat enzyme and yielded positive absorbance changes with maxima at 421, 449, and 481 nm and negative absorbance changes at 396, 466, and 497 nm, respectively. The value obtained for the spectroscopic binding constant (K s ), shown in Table 7, was comparable to that obtained from wild-type cb 5 r and indicated that H 4 NAD bound to the domain twelve times stronger than NAD + For the binding of NAD + differential flavin spectra were observed for all of the variants with the exception of the P92H/E255double mutant, which showed no detectable spectral changes at NAD + concentrations up to 5 mM. In contrast, the remaining mutants exhibited spectral changes that were similar to those of the wild-type domain, with positive absorbance maxima at 407 nm and 509 nm and negative absorbance maxima at 456 nm and 487 nm respectively. The K s values (NAD + M) were comparable to that obtained for the rat domain for each of the single mutants. To examine whether the thermodynamic properties of the flavin prosthetic group were similar to those of the corresponding wild-type domain, potentiometric titrations 71

PAGE 90

-0.050.000.050.100.15 [NAD+] (mM) 0.01.22.4 Absorbance(A410-A460) 0.000.040.08 Wavelength (nm) -0.030.000.030.060.09 [H4NAD] (mM) 0.00.40.8 Absorbance(A395-A480) 0.0000.0150.030 380435490545600 Absorbance -0.030.000.030.060.09 [H4NAD] (mM) 0.00.20.4 Absorbance(A395-A480) 0.0000.0150.030 380435490545600 -0.030.000.030.060.09 [NAD+] (mM) 048 Absorbance(A410-A460) 0.0000.0170.034 -0.030.000.030.060.09 [NAD+] (mM) 012 Absorbance(A410-A460) 0.000.030.06 -0.030.000.030.060.09 [H4NAD] (mM) 0.00.30.6 Absorbance(A395-A480) 0.0000.0150.030 -0.030.000.030.060.09 -0.030.000.030.060.09 ADCBGFHE Figure 24. Flavin difference spectra observed following binding of pyridine nucleotides to the different P92, P95 and E255cb 5 r variants. Difference spectra were obtained for the various P92, P95 and E255 cb 5 r mutants at equivalent flavin concentrations (50 M FAD) in 20 mM MOPS buffer, containing 0.1 mM EDTA, pH 7.0 following titrations with either H 4 NAD (left panels) or NAD + (right panels) as described in Methods. (A and E) P92H; (B and F) P95H; (C and G) P92H/E255-; (D and H) wild-type cb 5 r. The insert panels correspond to plots of the magnitudes of the observed spectral perturbations (peak to trough measurements at the indicated wavelengths) versus ligand concentration. The corresponding Ks values are given in Table 7. 72

PAGE 91

73 Table 7. Spectral Binding Constants Obtained for Wild Type c b 5 r and the P92A, P92H, P95A, P95H, and P92H/E255Variants. Mutant K s H4NAD (M) K s NAD+ (M) H 4 c b 5 r 45 10 760 30 P92A 58 5 800 40 P92H 93 5 764 44 P95A 54 4 594 39 P95H 115 8 1873 230 P92H/E255ND ND were performed using the dye equilibration method for the different P92, P95 and E255 c b 5 r variants in the presen ce of phenosafranine (E = -252mV) as an indicator. Spectra obtained during a representative titration of the P92A mutant are shown in Figure 25. Qualitative analysis of the individual spectra obtained from the various titrations indicated that the majority of phenosafranin e was reduced prior to FAD reduction in the majority of the c b 5 r variants, suggesting that the fl avin midpoint potential was more negative than that of phenosafranine. In c ontrast, the titration spectra obtained for the P92H variant suggested that the enzyme-bound flavin was reduced in tandem with the dye, indicating a positive shift in the potential for the flavin prosthetic group in the P92H variant. The flavin redox poten tials (n=2) for the different c b 5 r variants were determined from the Nernst semi-log plots, shown in Figure 25. The standard midpoint potentials obtained for the FAD/FADH 2 couple in the native enzyme (E = -272 mV) was comparable to values obtained for the P92A, P95A and P95H variants. Estimates of the flavin potential in the P92H/E255double mutant could not be obtained owing to the significant instability of the enzyme.

PAGE 92

Potential (mV) -300-280-260-240-220-200 Log (oxized/reduced) [FAD] -1.0-0.50.00.51.0 Wavelength (nm) 300400500600700 Absorbance 0.00.20.40.60.8 AB Figure 25. Oxidation-Reduction Midpoint Potentials for the FAD Prosthetic Group in the Different P92, P95 and E255 cb 5 r Variants. Reductive dye-equilibration titrations of the different P92, P95, E255 mutants and wild-type cb 5 r (40 M FAD) were performed as described under Methods in 100 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0 in the presence of phenosafranine (15 M, E o =-252 mV) [Marohnic et al, 2003]. Individual spectra were collected at 2-3 min intervals during the time course of the titrations. A. Selected spectra obtained during the dye-mediated redox titration of the P92A variant in the presence of phenosafranine are shown. B. The corresponding Nernst plots are shown for the titrations of the various mutants and correspond to P92A (, E o =-278+ 5 mV), P92H (, E o =-240+ 5 mV), P95A (, E o =-265+ 5 mV), P95H (, E o =-265+ 5 mV), P92H/E255(E o =ND) and wild-type cb 5 r (, E o =-272+ 5 mV). 74

PAGE 93

75 Summary of the RCM Variants Pro92His and E255Our comparative biophysical studies of the P92H, P95H, and the P92H/E255double mutant have yielded results that could be accurately predicted from the knowledge of the detailed structural or ganization of the enzyme. Re sidues P92 and P95 are both members of a conserved RxY T S xx S N flavin-binding sequence motif that has been identified in all cb 5 r sequences deposited to date in addition to the other members of the FNR superfamily of flavoprotein transhydrogena ses. While the side-chains of P92 and P95 do not contribute any direct contacts to the flavin prosthe tic group, in contrast to the side-chains of other motif members, such as R91 and Y93 that provide direct hydrogenbonding interactions with oxygen atoms that are part of the fl avin pyrophosphate and ribityl moieties, respectively, P92 is invol ved in a back-bone hydrogen bond between the carbonyl oxygen and the oxygen atom of the rib ityl C2 and participates in providing a portion of the hydrophobic netw ork that is also impor tant for flavin binding. A multiple sequence alignment revealed that while P92 is not conserved in all the known c b 5 r primary structures, serine [22] and alanine [103] residue s are the only known substitutions, it is present in thirty-three of the fifty-seven various sequences that have been deposited in Genbank for that portion of the polypeptide chain. However, in contrast, P95 is conserved together with the other motif members, Y93 and T94. The results of homology modeling studi es, shown in Figure 26, illustrate the effects of the P92H and P95H substitutions on the respective protein structures. In contrast to the orientation of the side ch ain of P92 in the native enzyme, where the proline side-chain is oriented parallel to the dimethylbenzyl moiety of the flavin isoalloxazine ring (Figure 26), in the P92H va riant, modeling suggests that the imidizole

PAGE 94

76 side chain would be oriented directly towa rds the dimethylbenzyl ring producing a steric clash that could be alleviated by a modest displacement of th e isoalloxazine ring into the solvent cleft between the flavinand NADHbinding lobes. This displacement would be expected to result in a decrea se in the efficiency of substr ate utilization, as shown by the ten-fold decrease in the P92H specificity co nstant, while the incr eased exposure of the isoalloxazine ring to solvent w ould be expected to result in an increase in redox potential of the flavin prosthetic group, toward s that exhibited by free flavin (E =-219 mV) [111], as shown by the +32 mV increase for the mutant compared to wild-type c b 5 r. The modeling studies suggest that substitu tion of histidine for proline at residue 95 would be expected to have only a modest impact on the functi onal properties of the enzyme. P95 is situated at the C-terminal end of -strand F 4 which is part of a sixstranded antiparallel -barrel that comprises the flavin binding lobe. The proline amide nitrogen is located 6.1 distant from the N5 of the isoalloxazine ring and the side chain is oriented away from the flavin prosthe tic group. While the histidine substitution at residue 95 would not be expected to have any significant impact on the properties of the flavin, as confirmed by the absence of any changes in redox potential, the decrease in activity and increase in the K m for NADH may be due to small local changes in the folding of the polypeptide backbone. In contrast to P92 and P95, which are pa rt of a well establis hed flavin-binding motif, E255 does not comprise pa rt of a specific sequence mo tif of defined function and is located within a relati vely non-conserved sequence re gion. E255 is a solventaccessible surface residue located at the amino terminus of a short -helical segment (H7) that participates in correctly orienting the carboxyl-terminal portion of the protein that

PAGE 95

77 Figure 26. Structural Models of the P92H and P95H cb 5 r Variants. A. The structures of the P92H and P95H variants were modeled using SWISS-MODEL [101] and are shown superimposed on the X-ra y structure of wild-type c b 5 r (PDB=1IB0). The structures are shown as ribbon models with the P92H shown in bl ue, the P95H in red and the wild-type c b 5 r in yellow. Deviations (RMS) be tween the P92H a nd P95H models and the wild-type structure corresponded to 0.96 A o and 0.97 A o respectively. The FAD prosthetic group and complexed NAD+ in the wild-type structure are shown in stick configuration with CPK coloring. B. An enla rged view of the superimposed structures in the region of the flavin prosthetic group showing the P92H and P95 substitutions. C. A portion of the superimposed P92H and wild-type structures showing a ribbon diagram of the region of the sequence between residues R91 and S97, corresponding to the RxY T S xx S N conserved motif, together with the sid echains of the wild-type proline, the mutant histidine and the FAD prosthetic group. D. As for C, except the sidechains of P95 and the histidine substitution are shown.

PAGE 96

B C D A 78

PAGE 97

79 contains the CGxxxM motif and which is critical to NADH/NAD + -binding. Thirty-two of the fifty-seven c b 5 r sequences deposited in Genbank share a glutamate at this position while another eighteen retain an aspartate re sidue. The three base pair deletion of the codon corresponding to E255 would be anticipated to have a substantial impact on the interaction of c b 5 r with either NADH or NAD + since the carboxy-terminal portion of the protein is primarily involved in interactions with the reducing substrate. Removal of any residue prior to and in close proximity to the conserved 273 CGxxxM 278 motif, which provides a number of critical contacts for the reducing subs trate, would significantly decrease substrate affinity, by as much as two-orders of ma gnitude, as has been previously shown [79] for the discrete E255variant. In addition, the resulting incorrect orientation of the bound substrate would redu ce the catalytic efficiency by over 50%. Thus, combining the effects of P92H on the pr operties of the flavin prosthetic group with the results of the E255 deletion on substrate a ffinity and utilization would be anticipated to significantly impact enzyme functionality, as evident by the 10 3 decrease in the specificity constant ( k cat / K m NADH ) for the P92H/E255double mutant. The role of the FAD/FMN sp ecificity binding motif 124G R xx S T 127 Within the c b 5 r primary structure is a four residue sequence motif 124G R xx S T 127 which, through the analysis of structural a nd sequence alignments within the flavoprotein transhydrogenase family, has demonstrated a high level of conservation. A primary feature of this motif is the proposed regulation of FAD or FM N cofactor specificity [5]. This specificity is potentia lly regulated by resi due 124 where, in th e majority of FNR family members, including cytochrome b 5 reductase, FAD is incorporated as the cofactor

PAGE 98

80 and contains the motif Gxx S T whereas those members of the FNR family that utilize FMN, such as aniline dioxyge n reductase, exhibit the Rxx S T motif. This motif is located within the only -helix in the flavin-bin ding domain of cytochrome b 5 reductase known to be directly involved with correct bi nding and orientation of the flavin prosthetic group (Figure 27B). In viewing the ligand binding plot for cytochrome b 5 reductase and FADbinding (Figure 27A) the importance of each amino acid within the G R xx S T motif is demonstrated with residue G124 havi ng a hydrophobic interaction and residues K125, M126, and S127 all forming a backbone hydrogen bond with the pyrophosphate moiety of the FADprosthetic group. Thus, it can be theorized that any mutations introduced into this motif could result in perturbations of the flavin environment and disruption of correct flavin binding and orientation. This was previously demonstrated by Bewley et al [80] in the type II methemoglobinemia variant S127P, which was the first RCM variant structure to be solved by X-ray crystallography and revealed that a mutation within the FAD-binding domain can lead to alterations in the NADH substrate binding behavior. The X-ray structur e of S127P revealed that th e ADP moiety of the FADcofactor was displaced into the corresponding ADP substrate binding si te, thus inhibiting efficient substrate binding. This phenomenon has been referred to as flavin flexibility. A second observation in the S127P crystal structure related to the amino acid residue K125. K125 has been demonstrated to be a critical residue within the Gxx S T motif in that the backbone nitrogen of K125 forms a hydrogen bond contact with the adenine moiety of FAD, through an ionic interaction K125 forms contacts in binding between the active sites of cytochrome b 5 reductase and cytochrome b 5 [112] and K125 comprises part of a 13 residue (F113 K125) surface exposed loop region which is part

PAGE 99

81 Figure 27. Structural Repres entation Displaying the 124G R xx S T 127 Motif by the FAD Cofactor Ligand Binding Plot and Seconda ry Structure of Wild-Type Cytochrome b 5 Reductase. (A) LIGPLOT [99] of 1IB0. C, O, N, S, and P atoms are represented as white, blue, red, yellow, and violet spheres, respectively, while c ovalent bonds are green sticks within FAD and blue sticks within amino acid residues of the FAD-binding lobe. Hydrogen bonds are drawn as dashed lines with the distances between atoms labeled. Residues contributing to hydrophobic interactions are repres ented as arcs with rays and colored blue. Panel (B) is a WirPlot rendering of rat cytochrome b 5 reductase with corresponding labels for secondary structur es pertaining to th e NADH-binding domain (blue), the FAD-binding domai n (red), and also the h inge region (green) for completion. The location of the 124G R xx S T 127 motif is highlighted in red.

PAGE 100

B A 82

PAGE 101

83 of the amino terminal flavin binding lobe which form a lid structure capping a portion of the cleft between the FADand NADHbi nding domains. In the crystal structure of the S127P variant this regi on underwent a conformational ch ange and was displaced by 6 and led to perturbations in the adenin e binding pocket and pyr ophosphate binding site of the flavin prosthetic group demonstrating the importance of proper conformation of the lid structure. Residue M126 of cytochrome b 5 reductase forms a hydrogen bond with the backbone nitrogen and the O1P of the FMN pyr ophosphate moiety of the FADcofactor. Previous clinical information diagnosed a patient possessing the mutation of M126V, corresponding to a A G transversion of codon 126 (ATG GTG), as a mutation capable of giving rise to type II methemoglobinemia. However, this patient was compound heterozygous and a se cond mutation in the DIA1 gene revealed the presence of an intronic A G (ISV4-2A G) alteration which lead to the skipping of exon 5 [84]. Biochemical characterization of the M 126V variant by Davis [40] has recently established that the mutation introduced by M126V would only result in type I methemoglobinemia and the type II RCM observed in the patient was direct result of the intronic alteration leading to the skippi ng of exon 5 or a comb ination of the two mutations. Through a direct hydrogen bond of the M126 nitrogen backbone with the pyrophosphate of the FMN moiety of the FA Dprosthetic group the M126 residue is believed to be a key residue in maintaining the integrity of the FAD binding pocket together with residues R91 and S127P [76]. The introduction of a valine substitution at residue 126 did not exemplify this. However, both methionine and valine are non-polar and neutral amino acids, thus, the replacem ent of methionine to valine would be a

PAGE 102

84 conservative mutation and more extensive and non-conservative mutations would need to be introduced in order to gain a full understand ing of the role of th e M126 residue and its interaction with the FADcofactor. In order to fully elucidate th e role of each residue in the 124G R xx S T motif, a different approach was directed towards each of the residues, based on their biochemical properties within cytochrome b 5 reductase. As mentioned, FAD/FMN specificity is proposed to be regulated by residue 124 through a G to R substitution. To properly examine the effects of flavin binding and specificity on residue G124, a mutation of G124R was generated along with G124A G124H and G124K, corresponding to a conservative uncharged control substitution and two other positively charged residues, respectively. In the case of the K125 residue, specific charge reve rsal variants, K125D and K125E, and a control mutation, K125A, were introduced to establish the effects of charge in relation to correct flavin binding and orientation as well as the ionic interaction of cytochrome b 5 reductase cytochrome b 5 binding. Finally, for the M126 residue, a series of variants were cr eated which corresponded to am ino acids that have been demonstrated to occur within other memb ers of the FNR superfamily and included: M126C/F/G/P/S. Characterization of G124A/H/K/R and K125A/D/E Mutant constructs enco ding for the variants G 124A/H/K/R and K125A/D/E corresponding to the introduction of a positive charge at position 124 and charge reversal of residue 125, respectively, we re generated through site -directed mutagenesis as described in Methods and Ma terials utilizing the original four-histidine tagged

PAGE 103

85 cytochrome b 5 reductase expression construct a nd the corresponding oligonucleotide primers listed in Table 1. Nucleotide sequenc ing in both directions confirmed the correct sequence and proper introduction of each mutation, and each variant was further expressed in the E. coli strain BL21 (DE3)-RIL and purified to homogeneity via Ni-NTA agarose chromatography and size exclusion FPLC. Expression yields for each of the variants were comparable to that of wild-type cytochrome b 5 reductase and all were yellow in color indicating the stability of each mutant generated and stable incorporation of the flavin prosthetic group. Purificati on to homogeneity was further confirmed by SDS-PAGE analysis as demonstrated by the appearance of single protein bands, which displayed molecular masses comparable to that of wild-type cytochrome b 5 reductase (M r = 31 kDa). UV/visible absorption spectra were obtained for oxidized samples of wild type cytochrome b 5 reductase and each of the variants ge nerated and are shown in Figure 28A. Each of the variants displayed an absorption spectra comparable to that of the wild-type domain by an aromatic absorption maxima at 272 nm in the UV region and peaks in the visible region at 386, 460, and a pronounced s houlder at 485 nm, all characteristic of simple flavoproteins. This initial data sugge sted that none of the variants generated had any significant effects on the sp ectroscopic properties of the flavin prosthetic group. In order to establish the eff ects of each mutation upon the secondary structure of the oxidized protein, far UV CD sp ectra were recorded in th e range of 190 nm for each variant and wild-type cytochrome b 5 reductase. As shown in Fi gure 28B spectra of all of the G124 and K125 mutants correlated well to the wild-type domain by displaying positive CD maxima at 196 nm and negative maxima at 222 nm, indicating that the

PAGE 104

260300340380420460500540580 Absorbance 0.00.20.40.60.81.01.21.41.6 420440460480500520 0.000.050.100.150.200.250.30 A Wavelength (nm) 190250 Molar Ellipticity (x 106) -4-2024 300350400450500550 Molar Ellipticity (x 104) -4-2024 BC Figure 28. Ultra-Violet and Visible Absorbance, and Circular Dichroism of Wild-Type, G124A, G124H, G124K, G124R, K125A, K125D, and K125E Cytochrome b 5 Reductases. (A) Oxidized samples of wild-type and mutant cb 5 rs (10 M), (B) (7 M), and (C) (60 M) FAD in 10 mM phosphate with 0.1 mM EDTA, pH 7.0 buffer. The inset shows and expanded region of the visible spectrum where the flavin cofactor makes a significant contribution. Individual spectra correspond to ( ____ ) H 4 cb 5 r; ( ____ ____ ) G124A; ( __ __ __ ) G124H; ( _ _ ) G124K; ( . . ) G124R; ( __.__.__ ) K125A; ( __. .__. .__ ) K125D; and ( ______ ) K125E. 86

PAGE 105

87 mutations introduced had no eff ect on the overall global fold ing of the protein. In contrast, the visible CD spectra obtained displayed perturbations for each variant showing that the introduction of a positive charge at residue 124 and the charge reversal of K125 resulted in alterations in the environment of the FADcofactor. As illustrated in Figure 28C, the spectra for the G124 variants each di splayed an increased intensity within the region of 310-370 nm with the line shape of G124H and G124K sh owing a reversed polarity. Each of the G124 va riants also showed an alte ration in the 460-485 nm region with a slight blue shift for each mutation indi cating a change in the hydrophilicity of the flavin environment [77]. For the K125 variants, K125A only displayed a minor perturbance with a minor bl ue shift in the 450485 nm region, whereas the K125D and K125E variants showed a significant alteration with a decreased intens ity in the region of 350-410 nm in addition to a minor alterati on in line shape from 470-485 nm. These findings are not unexpected since all of the mutations introduced are involved with either a hydrophobic or electrostatic inte raction with the FAD cofactor. To analyze the overall protein stabi lity of each of the variants, thermal denaturation profiles were generated utilizing thermal NADH:FR profiling in tandem with the loss of intrinsic flavin fl uorescence emission quenchi ng, following incubation of the proteins at temperatures ranging from 0-100 C. Changes in the intrinsic fluorescence of the FAD prosthetic group monitored togeth er with the retention of NADH:FR activity following thermal denaturation yielded a T 50 value, the temperature at which there is 50% of maximum fluorescence or 50% retention of NADH:FR activity. T 50 values obtained are represented in Figure 29. All G124 variants demonstr ated decreased temperature stabilities with T 50 values of 47, 49, and 50 C for the G124R, G124A/K, and G124H

PAGE 106

88 variants, respectively, while each of the K125 mutations yielded T 50 values of 53 and 56 C for the K125E and K125A/D variants, comparable to the wild-type cytochrome b 5 reductase T 50 of 55 C. These results indicated th at the introduction of a charge reversal mutation at residue K125 did not significantly affect the thermostability of the protein in regards to flavin binding, whereas the intr oduction of a positive char ge at residue 124 resulted in a lowered T 50 value with G124R causing the most dramatic decrease to 47 C. Thus, the substitution of G124 with the positivel y charged arginine, lysine, and histidine has a more significant effect on the overall protein stability indicating the most altered flavin environment in regards to binding and orientation. In order to determine how the G124 a nd K125 variants affected the overall catalytic efficiency, in itial-rate kinetic analyses were performed as described in Methods and Materials. Kinetic values obtained from these assays are reported in Table 9. As expected, all of the generated variants resu lted in a decreased specific activity when compared to wild-type cytochrome b 5 reductase. For the NADH:FR assays the catalytic efficiencies, k cat / K m NADH for the G124A, G124H, G124K, and G124R variants were determined to be only 1.2 to 2.6% that of the catalytic efficien cy of the wild -type domain. These results were due to both a dramatic increase in the K m NADH and decrease in the k cat indicating that the introduction of a positive charge at position 124 leads to a decreased affinity for the NADH substrate resulting in decreased turnover, possibly due to the displacement of the FAD cofactor. Likewise, the K125 charge reve rsal variants, K125D and K125E displayed a significant increase in the K m NADH however, only a modest decrease in k cat was observed for each variant, showing that turnover was not significantly perturbed in these variants upon substrate binding. Results for the NADH:BR assays also

PAGE 107

Temperature (oC) 0102030405060708090100 Residual NADH:FR Actvity 0255075100% Fluorescence Intensity 0255075100 Figure 29. Temperature Stability of Wild-Type cb 5 r and the G124A, G124H, G124K, G124R, K125A, K125D, and K125E Mutants. Oxidized samples of G124A, G124H, G124K, G124R, K125A, K125D, K125E, and wild-type H 4 cb 5 r (5 M FAD) were incubated at the indicated temperatures, and aliquots were withdrawn and assayed for both residual NADH:FR activity (closed symbols) and intrinsic flavin fluorescence (open symbols) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0 using excitation and emission wavelengths of 450 nm and 523 nm, respectively. Points correspond to: (,) H 4 cb 5 r; (, ) G124A; (, ) G124H; (, ) G124K; (, ) G124R; (hexagon) K125A; () K125D; and (+) K125E. 89

PAGE 108

90 Table 8. T 50 Values of the Wild-Type Cytochrome b 5 Reductase, G124A, G124H, G124K, G124R, K125A, K125D, and K125E Variants. Mutant T 50 (C) H 4 c b 5 r 55 G124A 49 G124H 50 G124K 49 G124R 47 K125A 56 K125D 56 K125E 53 Table 9. NADH:FR and NADH:BR Kinetic Constants Obtained for G124A, G124H, G124K, G124R, K125A, K125D, K125E, and Wild-Type c b 5 rs. Mutant NADH:FR NADH:BR k cat K m NADH K m FeCN6 k cat / K m NADH (s -1 ) (M) (M) (s -1 M -1 ) k cat K m NADH K m Cyt b 5 k cat / K m Cyt b 5 (s -1 ) (M) (M) (s -1 M -1 ) H 4 c b 5 r G124A G124H G124K G124R K125A K125D K125E 800 21 6 1 8 1 1.4 0.3 x 10 8 355 33 226 30 7 1 1.6 0.4 x 10 6 185 4 55 3 8 1 3.4 0.3 x 10 6 145 5 88 6 8 1 1.7 0.2 x 10 6 260 10 115 1 7 1 2.3 0.3 x 10 6 871 16 27 1 8 1 3.3 0.3 x 10 7 697 16 20 2 8 1 3.4 0.4 x 10 7 658 30 423 1 8 1 1.6 0.2 x 10 6 600 10 2 1 13 1 4.7 0.5 x 10 7 226 12 4 1 19 0.3 1.2 0.2 x 10 7 160 15 5 1 9 2 1.8 0.4 x 10 7 259 22 3 1 21 3 1.2 0.3 x 10 7 321 15 3 1 18 2 1.8 0.3 x 10 7 55 2 2 1 15 1 3.7 0.5 x 10 6 308 12 4 1 56 9 5.6 1.1 x 10 6 89 9 5 1 82 12 1.1 0.3 x 10 6

PAGE 109

91 correlated well to what we had hypothesized. As mentioned, K125 is not only involved in a electrostatic inte raction with the pyrophosphate moiety of the flavin prosthetic group but is also one of the critical positively-charged lysine residues that are involved in the charge pairing ionic inte raction in cytochrome b 5 reductase-cytochrome b 5 complex formation. The effects of the K125 charge reversal variants are reflected in the k cat / K m values retaining only 2, 8, and 12% of wild-type c b 5 r efficiency for the K125E, K125A, and K125D variants, respectively. The re sults shown in the decreased catalytic efficiencies for the K125 variants ar e best reflected in the increased K m c b 5 leading to the significant decrease in k cat Correlating well with the values obtained for the NADH:FR activities, overall catalytic efficiencies ranged between 26 to 38% of wild-type cytochrome b 5 reductase for each of the G124 varian ts which was only due to a decrease in the k cat values while the K m values for cytochrome b 5 binding were unaffected. This was to be expected since residue G 124 is not involved in the cytochrome b 5 reductasecytochrome b 5 binding interaction so that the lowere d catalytic efficiencies observed for the G124 variants was the direct result of the decreased affinity for NADH. As demonstrated through the kinetic analyses, each of the G124 and K125 variants displayed a decreased affinity for the NADH substrate. Therefore, spectral binding constants were determined for eac h of the mutants utilizing differential spectroscopy in order to compare the affinities for both the substrate NADH and product NAD + As described in Methods, the NADH isosteric analog H 4 NAD was utilized to monitor complex formation and establish a sp ectral binding constant [85]. Differential spectroscopy was utilized to monitor complex formation during titrations with either H 4 NAD or NAD + and is illustrated in Figure 30 and Figure 31. The G124 variant

PAGE 110

92 titrations performed in the presence of H 4 NAD yielded spectra with a line shape similar to that of wild-type cytochrome b 5 reductase (Figure 30) with a negative maxima at 395 nm and a shifted positive maxima at 495 nm compared to the wild-type spectra possessing positive maxima spectra at 410, 450, 480, and 515 nm and negative maxima at 395 and 500 nm. In contrast to the similar complex formation observed for the analog H 4 NAD, all titrations performed for the G124 variants in the presence of the product NAD + produced a spectra more comparable to the spectral changes observed in the H 4 NAD titrations (Figure 31). The values obtained for the respective binding constants ( K s ) for all G124 variants are given in Table 10. Interestingly, the K s values obtained for the G124A, G124H, G124K, and G124R variants in the presence of H 4 NAD were 23, 14, 46, and 31 M, respectively, compared to the wild-type c b 5 r value of 45 M, indicating a mode of binding comparable to or more efficient that the wild-type domain. Much of the same effect was witnessed in the titra tions carried out in the presence of NAD + with the G124 variants, yielding K s values of 534, 436, 794, 651 M, respectively, all being lower or comparable to the wild-type value of 760 M. With regard to the K125 charge reversal variants, both K125A and K125D displayed differential spectra for H 4 NAD comparable to that of wild-type cytochrome b 5 reductase. Titrations performed in the presence of H 4 NAD generated spectra with positive maxima at 410, 450, 480, and 515 nm, and negative maxima at 395 and 500 nm. Howeve r, the K125E variant displayed only minor spectroscopic changes, suggesting an altered affinity for H 4 NAD. This data was confirmed in the determination of the sp ectral binding constants for the K125A, K125D, and K125E variants yielding K s values of 76, 65, and 380 M, respectively, demonstrating a slightly altered binding affinity for the K125A and K125D mutations while the

PAGE 111

Absorbance -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 [H4NAD] (mM) 0.00.40.8 (A 495 -A 395 ) 000 002 004 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.00.30.6 (A 400 -A 500 ) 0.000 0.012 0.024 [H4NAD] (mM) Wavelength (nm) 380435490545600 -0.02 0.00 0.02 0.04 0.06 0.00.10.2 (A 400 -A 500 ) 0.000 0.005 0.010 [H4NAD] (mM) -0.02 0.00 0.02 0.04 0.06 0.0030.6 (A 500 -A 400 ) 0.00 0.01 0.02 [H4NAD] (mM) -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.0020.4 (A 480 -A 395 ) 0.000 0.015 0.030 [H4NAD] (mM) -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0000.080.16 (A 450 -A 395 ) 0.000 0.008 0.016 [H4NAD] (mM) 380435490545600 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.0020.4 (A 500 -A 400 ) 0.0000 0.0035 0.0070 [H4NAD] (mM) -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.00.40.8 (A 395 -A 480 ) 0.000 0.018 0.036 [H4NAD] (mM)A B C D E F G H Figure 30. Spectroscop ic Titrations of H4c b5r, G124A, G124H, G124K, G124R, K125A, K125D, and K125E c b5rs with H4NAD. Titrations of all mutants (50mM) were carried out as previously described in 10mM phosphate buffer containing 0.1mM EDTA, pH 7.0 at 23 C. Difference spectra were recorded following the addition of solution containing H4NAD (30M). The inset panel corresponds to a plot of the magnitude of the spectral perturbations at the indicated wavelengths versus pyridine nucleotide concentration where a difference spectrum was observed. Plots of the relative absorbance changes observed are as follows: (A) H4c b5r; (B) G124A; (C) G124H; (D) G124K; (E) K125R; (F) K125A; (G) K125D; and (H) K125E. 93

PAGE 112

-0.040.000.040.080.12 [NAD+] (mM) 024 (A410-A460) 0.000.040.08 Absorbance -0.020.000.020.040.06 0123 (A500-A400) 0.0000.0060.012 [NAD+] (mM) -0.020.000.020.040.06 0123 (A500-A400) 0.0000.0070.014 [NAD+] (mM) Wavelength (nm) 380435490545600 -0.020.000.020.040.06 024 (A410-A450) 0.00000.00120.0024 [NAD+] (mM) -0.020.000.020.040.06 0123 (A500-A450) 0.0000.0040.008 [NAD+] (mM) -0.040.000.040.080.12 0.00.81.6 (A410-A460) 0.000.030.06 [NAD+] (mM) -0.020.000.020.040.06 0.01.12.2 (A410-A460) 0.000.020.04 [NAD+] (mM) 38043549054560 0 -0.020.000.020.040.06 0.02.55.0 (A500-A450) 0.0000.0050.010 [NAD+] (mM) ABEFCDHI Figure 31. Spectroscopic Titrations of H 4 cb 5 r, G124A, G124H, G124K, G124R, K125A, K125D, and K125E cb 5 rs with NAD + Titrations of all mutants (50mM) were carried out as previously described in 10mM phosphate buffer containing 0.1mM EDTA, pH 7.0 at 23 C. Difference spectra were recorded following the addition of solution containing NAD + (30M). The inset panel corresponds to a plot of the magnitude of the spectral perturbations at the indicated wavelengths versus pyridine nucleotide concentration where a difference spectrum was observed. Plots of the relative absorbance changes observed are as follows: (A) H 4 cb 5 r; (B) G124A; (C) G124H; (D) G124K; (E) K125R; (F) K125A; (G) K125D; and (H) K125E. 94

PAGE 113

95 Table 10. Spectral Binding Constants Obtained for Wild Type c b 5 r and the G124A, G124H, G124K, G124R, K125A, K125D, and K125E Variants. Mutant K s H4NAD (M) K s NAD+ (M) c b 5 r 45 10 760 30 G124A 23 2 534 84 G124H 14 2 436 57 G124 K 46 7 994 113 G124 R 31 6 651 121 K125A 76 11 451 70 K125D 65 8 820 86 K125E 380 60 1182 142 introduction of the K125E variant displayed a bi nding constant 8-fold greater than that of the wild-type domain. Similar behavior was observed for the differential spectra generated in the ti trations performed with the product NAD + Again, the K125A and K125D variants displayed differential spec tra comparable to wild-type cytochrome b 5 reductase, with positive maxima at 410 and 510nm, and negative maxima 430, 460, and 490nm. This indicated that the K125A and K125D variants exhibited similar complex formation upon the binding of NAD + However, the differential spectra generated from the K125E variant with NAD + displayed absorbance chan ge magnitudes which were greatly reduced indicating a decr eased binding affinity. The K s values reported in Table 10 agreed well with the differential spectra obt ained in that the kinetic binding constants increased on the order of 451, 820, and 1183 M for the K125A, K125D, and K125E variants, respectively, compared to the value of 760 M for wild-type cytochrome b 5 reductase. Based on the data and structural modeli ng of the G124 and K125 variants we

PAGE 114

96 hypothesized that the mutations introduced woul d lead to alterations within the flavin environment of the enzyme. To examine the ef fects of these predicted structural changes on the flavin prosthetic group of each variant, oxidation-reduction potentials for the FAD cofactor were determined utilizing the dye-e quilibration potentiometric titration method for wild-type cytochrome b 5 reductase and the G124 and K125 mutants in the presence of phenosafranine (E = -252 mV). Flavin mi dpoint potentials (E n=2) for the FAD/FADH 2 couple were determined both in the absence and presence of NAD + from the linear Nernst plots of the log [FAD ox /FAD red ] versus the potential (mV) and are shown in Figure 32. Values established for all redox potentials are shown in Table 11. Analysis of the midpoint potentials of the G 124 variants in the ab sence of any pyridine nucleotide indicated that FAD reduction occu rred after the reduction of phenosafranine which is represented by a positive shift in the midpoint potential values obtained corresponding to -252 mV for the G124K variant and -260 mV for the G124A, G124H, and G124R variants compared to the sta ndard midpoint potential of -272 mV for cytochrome b 5 reductase. Correlating well with the standard redox potentials, midpoint titrations carried out in the presence of NAD + all displayed a more negative shift in the redox behavior for each of the G124 varian ts yielding values of -222, -219, -209, and 227 mV for the G124A, G124H, G1 24K, and G124R variants, respectively. These values obtained in the presence of NAD + are similar to the reported value of -220 mV for that of free flavin [95]. Another intere sting aspect of the xanthine ox idase titrations for all of the G124 variants with NAD + was observed in the analysis of the slopes of each Nenrst plot. The Nernst plot for cytochrome b 5 reductase oxidationreducti on reaction in the presence of NAD + yielded a slope of -30mV, representative of a n=2, twoelectron reduction.

PAGE 115

-300-285-270-255-240 Log (oxized/reduced) [FAD] -1.0-0.50.00.51.0 Potential (mV) -260-240-220-200-180-160 -1.0-0.50.00.51.0 AB Figure 32. Potentiometric Titrations of the G124A, G124H, G124K, G124R, K125A, K125D, K125E and Wild-Type Cytochrome b 5 reductases. Reductive dye titrations were performed at 25 C as described in Materials and Methods using phenosafranine as the indicator dye in 100 mM phosphate buffer containing 0.1 mM EDTA, pH 7.0. Nernst plots in the absence (A) and presence (B) of 2 mM NAD + Plots correspond to () H 4 cb 5 r; () G124A; () G124H; () G124K; () G124R; (hexagon) K125A; ( . . ) K125D; and (+) K125E. 97

PAGE 116

98 Table 11. Flavin Midpoint Potentials (E ) Obtained for Wild-Type, G124A, G124H, G124K, G124R, K125A, K125D, and K125E c b 5 rs. E FAD/FADH2 (mV) Slope (mV) E FAD/FADH2 (mV) Slope (mV) Enzyme -NAD + +NAD + c b 5 r G124A G124H G124K G124R K125A K125D K125E -272 -30 -260 -30 -260 -30 -252 -30 -260 -30 -272 -30 -286 -20 -270 -30 -190 -30 -222 -10 -219 -15 -209 -20 -227 -10 -201 -38 -200 -30 -220 -42 However, the slopes obtained for the G124 va riants all displayed a shift ranging between 10 to 20mV slopes as shown in Figure 32. The negative potentia l shifts displayed for the G124 mutants in the presence of NAD + the alterations in the slopes, and negative potentials in the absence of product indicate d that the introduction of a positive charge resulted in a structural displacement of the FAD cofactor leading to a destabilization of substrate binding and inhibition of efficient electron transfer from the NADH substrate to the flavin prosthetic group. In contrast to the results obtained fo r the G124 variants, the K125 mutations exhibited a different behavior In the absence of NAD + the standard redox potentials and Nernst plot slopes obtained we re all comparable to the wild-type domain with standard redox potential values of -272, -282, and 270mV for the K125A, K125D, and K125E variants, respectively, with slopes of 30mV. In the presence of NAD + however, both the K125A and K125D exhibited midpoint potentials with only a minor shift compared to the potential for wild-type cytochrome b 5 reductase yielding potential values of -201 and

PAGE 117

99 200 mV, respectively. The slopes derived for each mutant were also comparable to the -30 mV slope of the wild type domain A more dramatic shift was observed for the K125E variant with a midpoint potential of -220 mV whic h is in agreement of the K m value obtained in the NADH:FR assay for this variant indicating that the low affinity for NADH also had an effect on the efficiency of electron transfer represented by the midpoint potential in the presence of NAD + The alteration in the mode of electron transfer for K125E was also exemplified by a slope of -42 mV which is similar to that of an n=1 system represented by a slope of -60 mV These results provided evidence that the conformation change introduced into the flavin environment caused by the charge reversal variant K125E results in a decrease in the efficiency of electr on transfer from the NADH substrate to the FAD cofactor. Characterization of the amino acid residue M126 In order to establish the molecular basis of the disease methemoglobinemia for the previously characterized RCM variant M126V [ 40], five more variants were generated to complement the M126V analysis. The varian ts were chosen base d on residues that naturally occur at position 126 in other members of the ferredoxin:NADP + reductase superfamily and corresponded to M126C, M126F, M126G, M126P, and M126S. It was hypothesized that alterations of the M126 residue would lead to a conformational disruption of the flavin environment due to the hydrogen bonding of the M126 backbone with the AMP phosphate moiety of the FAD cofactor, possibly resulting in the displacement of the flavin prosthetic group in to the substrate binding cleft as observed in the type II methemoglobinemia variant S127P.

PAGE 118

100 The M126 variants were generated thr ough site-directed mutagenesis of the rat H4CB5R expression construct as described in Methods. Dideoxy sequencing in both directions confirmed the identity of each mutant construct. The generated variants were subsequently expressed in the E. coli strain BL21 (DE3)-RILP and purified by Ni-NTA metal chelate chromatography and size exclusion gel filtration FPLC. All five variants were purified to homogeneity as demonstrat ed by the appearance of single protein bands with a molecular mass of 31 kDa following SDS-PAGE (Figure 33) Representative UV/visible absorption spect ra were obtained for oxidized samples of wild-type cytochrome b 5 reductase and each of the M126 variants (Figure 34A). The spectra for each mutant was comparable to that of the wild-type domain with an aromatic absorption maxima at 270 nm in the UV region of the spectrum, and peaks at 385, 461, and a pronounced shoulder at 485-500 nm in th e visible region all being signature peaks of protein-bound flavin. Further examinati on of the global folding of the polypeptide backbone and the environment of flavin pros thetic group was carried out for each M126 variant and wild-type cytochrome b 5 reductase using UV and visible CD spectroscopy, respectively (Figure 34 B and Figure 34C). All of the M126 variants displayed spectra in the UV range (Figure 34B) that were represen tative of the wild-type domain, displaying a positive CD maxima at 196 nm and negative maxima at 222 nm demonstrating that the helix and -sheet architecture of the protein were not affected by the M126 mutations. Alternatively, in the visible range of the CD spectra, only the M126C variant retained visible CD spectra that was iden tical to wild-type cytochrome b 5 reductase with positive CD maxima at 310 and 390 nm and negative maxima at 460 and 485 nm and followed an identical line shape. Although the M126F vari ant displayed both negative and positive

PAGE 119

S A B C D E F G S503725 Figure 33. SDS Poly-Acrylamide Gel Electrophoresis of Wild-Type, M126C, M126F, M126G, M126P, M126S, and M126V Cytochrome b 5 Reductases. 2g each of purified enzyme samples were resolved on a 15% polyacrylamide gel. M126V was utilized as a positive control in the M126 series of variants. Lanes correspond to (S) protein molecular weight markers labeled accordingly (kDa); (A) H 4 cb 5 r; (B) M126C; (C) M126F; (D) M126G; (E) M126P; (F) M126S; (G) M126V. CD maxima comparable to the wild-type domain as well as a similar line shape a minor alteration in the flavin environment was observed with a blue-shift in the450-500 nm region. A more significant perturbation in the flavin environment was observed for the M126S, M126G, and M126P mutants in that the intensity of the M126S and M126G spectra was diminished within the region of 400-500 nm while the M126P variant yielded a reversed polarity, as shown in Figure 34C. Similar observations were witnessed in previous studies of the S127P [80] and R91 [76] variants where the AMP moiety of the FAD cofactor was displaced into the AMP pocket of the NADH-substrate binding site, a term referred to as flavin flexibility. These results are not unexpected since M126 occupies a positively charged FAD cofactor binding pocket with both the amino acid residues S127 and R91. 101

PAGE 120

Wavelength (nm) 190250 Molar Ellipticity (x 106) -4-2024 300350400450500550 Molar Ellipticity (x 104) -4-2024 Wavelength (nm) 260310360410460510560 Absorbance 0.00.20.40.60.81.01.21.41.61.8 320370420470520 0.000.150.30 ABC Figure 34. Ultra-Violet and Visible Absorbance, and Circular Dichroism Spectra of Wild-Type, M126C, M126F, M126G, M126P, and M126S Cytochrome b 5 Reductases. (A) Oxidized samples of wild-type and mutant cb 5 rs (10 M), (B) (7 M), and (C) (60 M) FAD in 10 mM phosphate with 0.1 mM EDTA, pH 7.0 buffer. The inset shows and expanded region of the visible spectrum where the flavin cofactor makes a significant contribution. Individual spectra correspond to ( ____ ) H 4 cb 5 r; ( ____ ____ ) M126C; ( __ __ __ ) M126F; ( _ _ ) M126G; ( . . ) M126P; and ( __.__.__ ) M126S. 102

PAGE 121

103 To examine the effects of the M126 vari ants on catalytic f unction, initial rate kinetic analyses were carried out for NADH:FR and NADH:BR assays and correlated well with the data obtained from the circul ar dichroism spectra. Values obtained for k cat and K m NADH are represented in Table 12. NADH catalytic efficiencies, k cat / K m NADH for the M126 variants decreased in th e order of M126F>M126C>M126S>M126G>M126P with each retaining 24, 18, 7, 6 and <1%, respectively, of wild -type cytochrome b 5 reductase activity. The effect s of the perturbations upon the flavin environment seen in the visible CD spectra are represented well within the K m values established in the NADH:FR assays for the M126S, M 126G, and M126P variants yielding K m NADH values of 62, 72, and 751 M, respectively. With respect to the NADH:BR assays, the catalytic efficiencies were all reduced in the order of WT c b 5 r>M126>M126F>M126S>M126G>M126P with only the k cat being affected with the exception of M126P having a K m 46 M compared to the wi ld-type value of 13 M. Differential spectroscopy was utilized to determine if the effects observed in the spectral and kinetic analyses of the M126 varian ts were reflected in substrate utilization through monitoring complex formations duri ng titrations with the NADH-substrate isoteric analog H 4 NAD or product NAD + (shown in Figure 35 and Figure 36). In the titrations performed with H 4 NAD the K s values, shown in Tabl e 13, all decreased on the order of WT c b 5 r>M126F>M126C>M126S>M126G. A value for M126P was not able to be established and no spectroscopic changes were observed even upon the titration with 1mM H 4 NAD. In addition to the decreased affinity displayed in the K s values, the M126C variant was the only muta tion that retained the identical spectroscopic complex formation like that of wild-type cytochrome b 5 reductase with a positive maxima at 410,

PAGE 122

104 Table 12. NADH:FR and NADH:BR Kineti c Constants Obtained for M126C, M126F, M126G, M126P, and M126S. Mutant NADH:FR NADH:BR k cat K m NADH K m FeCN6 k cat / K m NADH (s -1 ) (M) (M) (s -1 M -1 ) k cat K m NADH K m Cyt b 5 k cat / K m Cyt b 5 (s -1 ) (M) (M) (s -1 M -1 ) H 4 c b 5 r M126C M126F M126G M126P M126S 800 21 6 1 8 1 1.4 0.3 x 10 8 641 16 27 2 7 1 2.4 0.3 x 10 7 531 13 18 2 7 1 3.1 0.4 x 10 7 550 21 72 6 7 1 7.7 0.9 x 10 6 46 3 751 66 8 1 6.2 0.9 x 10 4 527 13 62 4 7 1 8.6 0.7 x 10 6 600 10 2 1 13 1 4.7 0.5 x 10 7 444 24 3 1 15 2 2.9 0.5 x 10 7 384 26 3 1 15 2 2.7 0.6 x 10 7 443 43 2 1 26 4 1.8 0.5 x 10 7 54 7 5 1 46 5 1.2 0.3 x 10 6 386 37 4 1 19 1 2.1 0.6 x 10 7 450, 480, and 515 nm, and negative maxima at 395 and 500 nm. In contrast, the introduction of a phenylalanine mutation ge nerated a spectra that corresponded to a complex formed in the titration performed in the presence of NAD + with a positive maxima of 410 and 510 nm and negative maxi ma at 430, 460, and 485 nm. Correlating with the perturbations observed in the decreased visible CD spectra of the flavin environment, spectral changes for the M126G and M126S variants demonstrated an altered complex formation with positive maxima at 400 and 500 nm and negative maxima begin represented by a negative maxima between 410 and 490 nm. Unlike the altered complex formations observed for the titrations carried out in the presence of H 4 NAD, the spectral changes observed for the di fferential spectroscop y in the presence of NAD + yielded positive maxima at 410 and 510 nm and negative maxima at 430, 460, and 485 nm, identical to that of wild-type cytochrome b 5 reductase. Values for the titrations carried out in the presence of NAD + are listed in Table 13. K s values obtained

PAGE 123

-0.04 0.00 0.04 0.08 [H4NAD] (mM) 0.00.20.4 Absorbance (A 395 A 480 ) 000 002 004 Absorbance -0.04 0.00 0.04 0.08 0.00.510 0.000 0.013 0.026 Wavelength (nm) 380435490545600 -0.04 0.00 0.04 0.08 0.00.306 000 002 004 -0.02 0.00 0.02 0.04 0.0051.0 0.000 0.007 0.014 380435490545600 -0.02 0.00 0.02 0.04 0.0020.4 0.000 0.006 0.012 [H4NAD] (mM) [H4NAD] (mM) [H4NAD] (mM) [H4NAD] (mM) Absorbance (A 395 A 480 ) Absorbance (A 410 A 460 ) Absorbance (A 410 A 450 ) Absorbance (A 410 A 450 ) -0.04 0.00 0.04 0.08 A B D E C F Figure 35. Spectroscopic Titrations of H4c b5r, M126C, M126F, M126G, M126P, and M126S c b5rs with H4NAD. Titrations of all mutants (50mM) were carried out as previously described in 10 mM phosphate buffer containing 0.1mM EDTA, pH 7.0 at 23 C. Difference spectra were recorded fo llowing the addition of solution containing H4NAD (30M). The inset panel corresponds to a plot of the magnitude of the spectral perturbations at the indicated wavelengths versus pyridine nucleotide concentration where a difference spectrum was observed. Plots of the relative absorbance changes observed are as follows: (A) H4c b5r; (B) M126C; (C) M126F; (D) M126G; (E) M126P; and (F) M126S. 105

PAGE 124

-0.04 0.00 0.04 0.08 0.12 [NAD + ] (mM) 024 Absorbance (A 410 A 460 ) 0.00 0.04 0.08 Absorbance -0.04 0.00 0.04 0.08 0.12 00255.0 0.00 0.04 0.08 Absorbance (A 410 A 460 )[NAD + ] (mM) Wavelength (nm) 380435490545600 -0.06 0.02 0.10 0.18 00255.0 0.00 0.06 0.12 [NAD + ] (mM) Absorbance (A 410 A 460 ) -0.04 0.00 0.04 0.08 0.12 002.55.0 0.00 0.04 0.08 Absorbance (A 410 A 460 )[NAD + ] (mM) 380435490545600 -0.04 0.00 0.04 0.08 0.12 036 0.00 0.03 0.06 [NAD + ] (mM) Absorbance (A 410 A 460 ) -0.04 0.00 0.04 0.08 0.12 Figure 36. Spectroscopic Titrations of H4c b5r, M126C, M126F, M126G, M126P, and M126S c b5rs with NAD+. Titrations of all mutants (50mM) were carried out as previously described in 10 mM phosphate buffer containing 0.1mM EDTA, pH 7.0 at 23 C. Difference spectra were recorded fo llowing the addition of solution containing NAD+ (30M). The inset panel corresponds to a plot of th e magnitude of the spectral perturbations at the indicated wavelengths versus pyridine nucleotide concentration where a difference spectrum was observed. Plots of the relative absorbance changes observed are as follows: (A) H4c b5r; (B) M126C; (C) M126F; (D) M126G; (E) M126P; and (F) M126S. 106

PAGE 125

107 with NAD + correlated exactly with the K s values for H 4 NAD with the spectroscopic binding constants showing a decreased affinity in the order of WT c b 5 r>M126F>M126C>M126S>M126S>M126G>M126P, with no appreciable value obtained again for the M126P variant even upon the addition of up to 5mM NAD + a concentration ~8-fold greater than that for wild-type cytochrome b 5 reductase. These results indicate that the conformational changes in the environment of the flavin prosthetic group result in a decreased affinity for both H 4 NAD and NAD + providing evidence for a potential role for the M126 residue in catalysis. Thermal denaturation profiles were al so obtained for wild-type cytochrome b 5 reductase and the M126 variants and are shown in Figure 37. The T 50 values shown in Table 14 clearly demonstrate a direct correlation with the subsequent da ta shown thus far. The effect of the mutation upon the flavin environment with a de crease in catalytic efficiency, reduced substrate binding affinit y, and perturbed visible CD spectra for the M126 variants is also reflected in a decrea sed thermostability when compared to the wild-type cytochrome b 5 reductase T 50 value of 55 C. The M126 variant T 50 values decreased in the order of 52, 51, 47, 45, and 44 C for the M126F, M126C, M126S, M126P, and M126G mutants, respectively. The decrease in thermostability demonstrated that the degree at which there is a disruption in the environment of the flavin prosthetic group directly correlated to the stability of the protein. Finally, to asses the effects of the c onformational disrupti ons of the flavin environment upon the flavin reduction potentia l, oxidation-reduction potentials for the FAD prosthetic group were established usi ng the dye equilibration method both in the presence and absence of NAD + (shown in Figure 38). The standard redox potential (E )

PAGE 126

108 Table 13. Spectral Binding Constants Obtained for Wild Type c b 5 r and the M126C, M126F, M126G, M126P, and M126S Variants. Mutant K s H4NAD (M) K s NAD+ (M) H 4 c b 5 r 45 10 760 30 M126C 139 13 1071 32 M126F 66 1 781 15 M126G 242 9 4216 139 M126P ND ND M126S 214 6 2966 109 for the FAD/FADH 2 couple was derived from the Nern st plot and values obtained are listed in Table 15. Both the M126C and M126F variants exhibited a midpoint potential comparable to that of the wild-type domain (-272 mV) with values of -270 and -274 mV and a slope of -30 mV, representative of an n=2 reaction, respectivel y, and indicating that the substitution of cysteine or phenylalanine at position 126 had no significant effect on the standard redox potential. However, th e flavin redox behavior exhibited a more dramatic shift for the M126S, M126G, a nd M126P variants all yielding midpoint potentials of -258 mV although a slope of -30 mV was observed for each mutant. This variation can be attributed to the alterations within the FAD cofactor environment observed in the preceding data. Flavin midpoint potentials pe rformed in the presence of NAD + were all dissimilar compared to the value obtained for wild-type cytochrome b 5 reductase (-190 mV), yielding potentials of -206 -180, -198, -209, and -220 mV for the M126C, M126F, M126G, M126S, and M126P varian ts, respectively. Each variant with the exception of the M 126P mutant displayed potentials comparable to that of

PAGE 127

Temperature (OC) 0102030405060708090100 % Residual Activity (NADH:FR) 0255075100Flourescence Intensity % 0255075100 Figure 37. Temperature Stability of cb 5 r and the M126C, M126F, M126G, M126P, and M126S Mutants. Oxidized samples of M126C, M126F, M126G, M126P, M126S, and wild-type H 4 cb 5 r (5 M FAD) were incubated at the indicated temperatures, and aliquots were withdrawn and assayed for both residual NADH:FR activity (closed symbols) and intrinsic flavin fluorescence (open symbols) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0 using excitation and emission wavelengths of 450 nm and 523 nm, respectively. Points correspond to: (,) H 4 cb 5 r; (, ) M126C; (, ) M126F; (, ) M126G; (, ) M126P; and (+) M126S. 109

PAGE 128

110 Table 14. T 50 Values of the Wild-Type Cytochrome b 5 Reductase, M126C, M126F, M126G, M126P, and M126S Variants. Mutant T 50 ( C) H 4 c b 5 r 55 M126C 51 M126F 52 M126G 44 M126P 45 M126S 47 cytochrome b 5 reductase demonstrating the intro duction of the nonpolar bulky side chain of proline resulted in a greater perturbation of the flavin environment. Interestingly, the slope derived for each variant in the presence of NAD + also differed when compared to the wild-type domain, producing slopes of -60 mV for the M126C, M126F, and M126G mutants and -42 and -45 mV for the M126P and M126S variants, respectively, which were indicative of an n=1 reaction. Summary of the 124G R xx S T 127 Motif Overall, these results provide direct ev idence to support the importance of the 124G R xx S T 127 motif for proper FAD cofactor incor poration and orientation. Residues G124, K125, M126, and S127 are all located in the N-terminal portion of -helix F A, the only -helix found within the FAD bi nding domain of cytochrome b 5 reductase. Each residue also forms a structural framewor k for the proper positioning of the flavin

PAGE 129

-300-280-260-240-220-200 -1.0-0.50.00.51.0 Potential (mV) -260-240-220-200-180-160-140-120 log (oxidized/reduced) [FAD] -1.0-0.50.00.51.0 AB Figure 38. Potentiometric Titrations of the M126C, M126F, M126G, M126P, M126S, and Wild-Type Cytochrome b 5 Reductases. Reductive dye titrations were performed at 25 C as described in Materials and Methods using phenosafranine as the indicator dye in 100 mM phosphate buffer containing 0.1 mM EDTA, pH 7.0. Nernst plots in the absence (A) and presence (B) of 2 mM NAD + Plots correspond to () H 4 cb 5 r; () M126C; ()M126F; () M126G; () M126P; and (+) M126S. 111

PAGE 130

Table 15. Flavin Midpoint Potentials (E) Obtained for M126C, M126F, M126G, M126P, M126S, and Wild-Type Cytochrome b 5 Reductases. E FAD/FADH2 (mV) Slope (mV) E FAD/FADH2 (mV) Slope (mV) Enzyme -NAD + +NAD + H 4 cb 5 r M126C M126F M126G M126P M126S -272 -30 -270 -30 -274 -30 -258 -30 -258 -30 -258 -30 -190 -30 -206 -60 -190 -60 -198 -60 -209 -42 -220 -45 prosthetic group via a hydrophobic interaction with G124 and electrostatic interactions of the backbone nitrogen of K125, M126, and S127 with the AO2, O1P, and O2P of the FAD pyrophosphate moiety, respectively (shown in Figure 27A). Of the four residues, S127 also possessed a side chain hydrogen bond with the FAD cofactor from the OG of S127 to the O2P pyrophosphate moiety of FAD. Detailed analysis has demonstrated the importance of the S127 residue through the crystal structure of the type II RCM mutant S127P [80]. The introduction of a proline at position 127 leads to the displacement of the ADP moiety of the FAD cofactor into the corresponding ADP binding site of the NADH-binding domain. Through this displacement, the FADcofactor acts as an effective substrate inhibitor. This suggested that any other mutations introduced into the G R xx S T motif would also lead to an alteration in the flavin environment causing the displacement of the flavin prosthetic group and result in a significant decrease in the overall catalytic efficiency of the protein. The initial evidence that suggested that each of the residues within the G R xx S T motif was involved in maintaining proper flavin incorporation was obtained from the 112

PAGE 131

113 circular dichroism studies. Results from the spectra obtained in the far-UV region of the CD (190-300 nm) indicated that none of th e mutations introduced had any significant effect on the overall global folding of the prot ein. However, the spectra displayed in the near-visible region representing the environm ent of the flavin prosthetic group showed significant perturbations. For the G124 an d K125 variants, alterations were more prominent within the region of 300-400 nm and all displaye d a blue-shift at 450-485 nm, representative of an alterati on in the hydrophilicity of the fl avin environment. For the M126 variants, minor perturbations were observed as a blue-shift for the M126C and M126F mutants, whereas the M126S, M126G, and M126P variants all displayed a decrease in spectral intensity from the region of 400-500 nm which was in good agreement with the spectra previously publishe d for the R91 and S127P variants [76, 80]. This indicated that the mutations introduced at position 126 not only led to an alteration of the flavin environment but also to a possi ble displacement of the FADcofactor. This can be attributed to the fact that, as mentioned, the M126 residue is located in the only helical segment in the FADbinding domain and it has been shown that the introduction of a glycine mutation into an -helix can result in a change in the backbone structure through the removal of any side chain, whil e the introduction of a proline generates a kink in the backbone of an -helix. The perturbations observed in the visibl e CD spectra were all in good agreement with the results obtained for the kinetic and thermodynamic properties of the various mutants. Comparisons of the overall catalytic efficiencies revealed that the decreased turnover was directly attributab le to the reduced affinity, ( K m ), for the nucleotide NADH. This was in good agreement with the kinetic results obtained for the S127P variant which

PAGE 132

114 displayed a K m value 9-fold greater than that of wild-type cytochrome b 5 reductase. As anticipated the results obtai ned from the NADH:BR assays demonstrated that the mutations introduced at the G124 and M126 residues had li ttle to no effect on the K m cb5 whereas the charge reversal va riants K125D and K125E showed a significant decrease in their overall catalytic efficienc y. These results were in good agreement with previously published studies that had defined K125 as one of four positively-charged lysine residues that were involved in an ioni c interaction with cytochrome b 5 [112]. For the G124 variants, diffe rential binding constants, ( K s ), provided unexpected results in regards to substrate binding affinity based on the increased K m values obtained from the NADH:FR assay. In the titrations performed with the substrate analog H 4 NAD, all differential spectra displayed a line shape similar to that of wild-type cytochrome b 5 reductase, however, spectral binding constant s derived from each spectra were lower than that of wild-type cytochrome b 5 reductase showing an enhanced binding affinity for each G124 variant. In the titr ations that utilized NAD + the spectral binding constants for the G124 mutants followed similar trends; in which all but G124K yielded K s values lower than that of the wild-type dom ain. However, the differentia l spectra displayed for each variant were representative of the line shape observed in the presence of H 4 NAD. These results indicated an altered m ode of binding for the product, NAD + that was a direct result of the perturbations ge nerated by the G124 variants within the environment of the flavin prosthetic group. It can also be concluded that th e decrease in the K s values is not indicative of enhanced binding of H 4 NAD and NAD + but rather that, the disruptions of the flavin environment prevent efficient s ubstrate disassociation. The results for the differential spectroscopy were as expected for the K125 and M126 series of variants

PAGE 133

115 where each displayed a decreased binding affini ty which was in direct agreement with the K m values obtained for the activity assays. The predicted alterations of the FAD environment were also evident by the perturbations detected in the oxidati on-reduction potentials of the FAD/FADH 2 couple both in the absence and presence of NAD + Each of the G124 variants displayed a positive shift in the flavin redox potential in the absence of NAD + and a negative shift in the presence of NAD + A slope of -30 mV was also obt ained for each of the G124 Nernst plots which was similar to that determined for the wild-type cytochrome b 5 reductase oxidation-reduction potential, indicating a similar n=2 electron transfer process. However, the slopes derived from the Nernst plots in the presence of NAD + were severely compromised for the G124 series yi elding slope of -10, -12, -15, and -20 mV for the G124A, G124R, G124H, and G 124K variants respectively. In contrast, the midpoint potentials obtained for the K125A, K125D, and K125E variants were all comparable to that of wild-type cytochrome b 5 reductase with the exceptio n of K125E which yielded a midpoint potential of -200 mV in the presence of NAD + Finally, the results from the M126 series of mutants yielded oxidation-reduc tion potentials directly correlated with the kinetic and thermodynamic data already obtained for the M126 variants. The M126 variants catalytic efficiencies, thermostab ility, visible-CD, and differential binding constants all decreased in the order of M126C and M126F being comparable to the wildtype domain and then followed by M126S>M126G>M126P for each biophysical parameter. This trend was also observed for the standard oxidation-reduction potentials of each mutant, with M126C and M126F gene rating potentials comparable to wild-type cytochrome b 5 reductase and a positive shift observed in the titrations performed in the

PAGE 134

116 absence of NAD + for M126S, M126G, and M126P. Data obtained for the M126G mutant did not conform to this pattern yielding a potential in the presence of NAD + of 198 mV while the M126S and M126P displa yed potentials of -210 and -220 mV, respectively. The midpoint potentials obtaine d were most likely a direct result of the perturbations of the flavin environment and reflect potential displacement of the adenine moiety of the flavin prosthetic group as suggested by the visible-CD spectra. These observations, in conjunction with the poor substrate affinity, as demonstrated by the initial rate kinetic analyses would result in a highly unfa vorable electron transfer and flavin reduction process. Analysis of the effects of the RCM variant M272and generated variants M272A/I/L/R on the CGxxxM motif. Within the c b 5 r primary structure, several sequence motifs have been identified that are involved in either flavin binding ( 91 RxY T S xx S N 97 ) or FAD/FMN specificity ( 124G R XX S T 127 ) or function in modulating re duced pyridine nucleotide binding ( 180 GxGxxP 185 and 273 CGxxxM 278 ). The second of the conserved motifs that are involved in regulating pyridin e nucleotide affinity, corre sponds to a six amino acid 273 CGpppM 278 motif that comprises the resi dues C273 to M278 in the carboxylterminal lobe of rat c b 5 r (Figure 40). Preceding this motif in the sequences of the majority of higher eukaryote c b 5 rs is a conserved methionine residue, M272, that has been previously identified as absent in a patient with type II RCM [81]. High resolution X-ray structures have been obtained for the soluble forms of rat [12], human [28], and pig [113] c b 5 r. Within the rat c b 5 r diaphorase domain (PDB

PAGE 135

117 ID=1I7P), M272 has been shown to be lo cated at the terminus of the strand N together with the conserved essential cysteine residue, C273. This strand is part of a fivestranded anti-parallel -barrel structure which contains a canonical Rossman fold and comprises the pyridine nucleotide-binding lobe (residues K172 to F300) [12]. Analysis of the structure of the diaphorase domain complexed with NAD + (PDB ID=1IB0) has revealed that while M272 provides no direct electrostatic or hydrogen bond contacts with either th e nicotinamide or ribose moieties of the bound NAD + the subsequent residues, corresponding to C273 to M278, form a loop connecting N 4 with the final helical return in the NADH binding domain and provide an extensive framework of hydrophobic contacts that are involved in binding the reducing substrate and specifically orienting the nicotinamide portion of the reduced pyridine nucleotide for subsequent efficient hydride transfer to the FAD prosthetic gr oup. Within the CGxxxM motif, the conserved glycine, corresponding to G274, forms two critical hydrogen bonds. To probe the role of M272 in c b 5 r structure and function and to examine the effects of deleting this residue as previously identifie d in the type II methemoglobinemia variant, we utilized site-directed mutage nesis as a tool to either dele te the methionine residue at position 272 in the rat c b 5 r diaphorase domain or substitute the methionine with the amino acid residues, A, I, L and R. Residues that have been shown to commonly occur at the equivalent position within other members of the FNR superfamily. Subsequently, we primarily examined the effects of these dele tions and substitutions on the spectroscopic and thermodynamic properties of the FAD pros thetic group and the interactions the mutant with the physio logical reductant, NADH. Mutant constructs enco ding the five different c b 5 r variants, M272-, A, I, L and R,

PAGE 136

118 Figure 39. Multiple Sequence Alignment of c b 5 r Primary Structures. (A) Selected eukaryotic c b 5 r amino acid sequences were retrie ved from GenBank, aligned using the ClustalX algorithm [108] and the ali gnment optimized for maximum sequence conservation. Only the portions of the se quences surrounding the conserved pyridine nucleotide-binding 273 CGPPPM 278 motif (underlined) are sh own for clarity. Beginning and ending residue numbers with in the translated GenBank sequences are indicated by superscripts while the methionine residue corresponding to M272 in R. norvegicus c b 5 r is shown in bold face. The consensus sequence is also shown below with identical residues indicated by * and simila r residues indicated by : .

PAGE 137

119 SPECIES SEQUENCE H. sapiens 267 EPLVLMCGPPPM IQYAC 283 P. troglodytes EPLVL M CGPPPM IQYAP M. mulatta EPLVL M CGPPPM IQYAC M. fascicularis EPLVL M CGPPPM IQYAC B. tauros EPLVL MCGPPPM IQYAC S. scrufa EPLVL MCGPPPM IQYAC C. familiaris 267 EPLILM CGPPPM IQYAC 283 R. norvegicus 267 ETLILMCGPPPM IQFAC 283 M. musculus EPLIL M CGPPPM IQFAC G. gallus DVLIL M CGPPPM IQYAC X. laevis DVLIL M CGPPPM VQYAI F. rubripes DTLIL M CGPPPM IQFAC T. nigroviridis DTLIL MCGPPPM IQFAC D. rerio DSMIL M CGPPPM IQFAC X. tropicalis DVVVL M CGPPPM IQFAC M. domestica DTMIL MCGPPPM IQFAC C. briggsae DSAVL M CGPPPM INFAC A. mellifera DTMVL MCGPPPM INFAC A. gambiae NSLVL M CGPPPM VNYAC L. major KVMAL M CGPPPM VQMAI C. neoformans GHKVL MCGPPPM ITAMK D. discoideum QTMVI MCGPPMM NKAMT C. elegans DSAVL LCGPPPM INFAC D. melanogaster DTIVL LCGPPPM INFAC C. albicans DTNLL LCGPPPM VSAMK D. hansenii ATNLL L CGPPPM ISAMK Y. lipolytica NTKLL L CGPPPM ISALK N. crassa DVKIL L CGPPPM ISGLK E. gossypii SAQLL L CGPPPM VSSAK M. alpina DIKVL L CGPPPM VSAMS G. zeae DSKVF LCGPPGM VNASK P. falciparum DTLIL LCGPPPM TSSIK C. glabrata DVQLL V CGPPGM VSSVK K. lactis GVQLL V CGPPPM VSSIK C. maxima DALIL V CGPPGM MKHIC T. brucei KAILL V CGPPGF MKTIS S. pombe ETKVL I CGPTPM VNSLR S. japonicum DTITL I CGPPPF IEFAC S. cerevisiae NVQIL I CGPPAM VASVR V. fordii DIQIL R CGPPPM NKAMA N. tabacum DIQIL R CGPPPM NKAMA A. thaliana DIQIL R CGPPPM NKAMA Z. mays DIQIL R CGPPPM NKAMA O. sativa DIQIL R CGPPPM NKAMA C. comosus DIQIL R CGPPPM NKAMA

PAGE 138

A C273 FAD B M272 FAD C273 Figure 40. Graphical Sequence Alignment an d Structural Representation of Amino Acid Residue M272. (A) Graphical representation (s equence logo) of the amino acid residues present at positions 267 to 283 in the primary sequences of 55 c b5r sequences deposited in GenBank. The logo consists of stacks of symbols (one stack for each position in the sequence) with the overall he ight of the stack indicating the sequence conservation at that position a nd the height of the symbols within the stack indicating the relative frequency of each ami no acid at that position. (B) Sc hematic diagram of a portion of the R. norvegicus c b5r X-ray structure (PDB=1IB0) sh owing the arrangements of the amino acids that comprise the seque nce region containi ng the conserved 273CGxxxM278 motif and residue M272 that precedes the motif. Amino acid resi dues are shown in stick representation using the CPK color sc heme while the approp riate portion of the peptide backbone, corresponding to residues I 270 to C283, is displayed as a ribbon diagram. The location of the FAD prosthetic group and complexed NAD+ are also shown in stick representation using the CPK color scheme. 120

PAGE 139

121 the latter corresponds to the most frequently encountered amino acid residues occurring at positions equivalent to M272 in the various c b 5 r sequences identified in GenBank and also within other members of the FNR supe rfamily of pyridine nucleotide-dependent flavoprotein transhydrogenases, were gene rated through directed mutagenesis of the original four-histidine tagged c b 5 r expression construct. Nucleotide sequencing confirmed the fidelity of each construct and each of the mutant proteins was subsequently expressed in the E. coli strain BL21(DE3)-RIL and purified to homogeneity by a combination of Ni-chelate chromatography and gel filtration FPLC. Evaluation of the expression yields of the various mutants indi cated that the different M272 variants were expressed at levels comparable to that of the wild-type domain, with the exception of M272L which was obtained in significantly decreased yield (approximately 30% of wildtype), suggesting decreased stab ility. All five M272 variants were purified to apparent homogeneity as evident by the presence of single protein bands following SDS-PAGE analysis of the different mutants as shown in Figure 41, which also indicated molecular masses comparable to that of the native enzyme (M r approx. 32 kDa). The oxidized forms of all five purified M272 variants were yellow in color indicating the incorporation of a flavin pros thetic group and confirming that M272 did not provide any backbone or side-chain contac ts that were essential for protein folding and the stable incorporation of the flavin prosthetic group. UV/visible absorbance spectra were obtai ned for oxidized samples of each mutant and wild-type c b 5 r and are compared in Figure 42 A. The M272-, M272A, M272I, M272L and M272R variants each exhibited absorption spectra comparable to that of the wildtype enzyme with an aromatic absorption maximum detected at 272 nm in the UV region

PAGE 140

D 50 75 B 25 37 F C AE Figure 41. SDS-PAGE Analysis of the Different M272 c b5r Variants. Samples of the different M272 variants (2 g each) and wild-type c b5r, obtained from the terminal FPLC purification step, were resolved on a 12.5% polyacrylamide gel : Lane A, wild-type c b5r; lane B, M272-; lane C, M272A; lane D, M 272I; lane E, M272L; lane F, M272R. The arrows indicate the positions of selected mo lecular weight markers with the indicated molecular masses (kDa). of the spectrum, and a peak at 461 nm with a associated pronounced shoulder in the range of 485-500 nm in the visible region of the sp ectrum, attributable to protein-bound flavin. None of the visible spectra of any of the M272 variants were blueor red-shifted with respect that of the wild-type protein, as has been previously demonstrated for mutations of other residues, such as R91 and Y93 [ 77] [76], suggesting th at none of the M272 substitutions or its deletion had any significant influence on the spectroscopic properties of the FAD prosthetic group. Blue shifts in th e visible absorbance spectra of flavoproteins have previously been attributed to changes in the hydrophilicity of the flavin environment near the N(5) locus of the isoalloxazine ring [114] [115] while red shifts have been observed for other mutants such as the previously ch aracterized T94H. To assess the secondary structural cont ent of each of the M272 variants, CD spectra were recorded in the UV wavelengt h range (190-300 nm). As shown in Figure 122

PAGE 141

Wavelength (nm) 250300350400450500550600 Absorbance 0.00.20.40.60.81.01.21.41.6 320355390425460495530 0.00.10.20.3 Wavelength (nm) 190250 Molar Ellipticity (x 106) -4-2024 300350400450500550 Molar Ellipiticity (x 104) -4-2024 ABC Figure 42. UV/Visible Absorption and CD Spectra of cb 5 r and the Various M272 Mutants. (A) UV/visible absorption spectra were obtained for oxidized samples of the various G272 mutants together with wild-type cb 5 r at equivalent flavin concentrations (1.7 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. The inset shows an expanded region of the visible spectrum where the flavin prosthetic group makes a major contribution. (B) UV CD spectra were recorded using enzyme samples (7 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. (C) Visible CD spectra were recorded using enzyme samples (50 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. Individual spectra correspond to; M272( ___ ___ ___ ); M272A ( __ __ __ ); M272I (-), M272L ( . . ); M272R ( __.__.__ ) and wild-type cb 5 r( _____ ). 123

PAGE 142

124 42B, all of the c b 5 r variants exhibited positive CD from 190-210 nm and negative CD from 210-250 nm with all the spectra retaining both positive and nega tive intensities very similar to that of the wild-type domain. The absence of any significant differences between the UV CD spectra of the wild-type and mu tant proteins suggested conservation of the overall secondary structure architec ture and that none of the M272 residue substitutions or its deletion had any deleterious effects on the global folding of the diaphorase domain. To examine the influence of the various M272 residue substitutions or deletion on the stabilities of the resulting proteins, thermal denaturation profiles were generated for the various mutant forms and compared with those obtained for wild-type c b 5 r and are shown in Figure 43. Alterati ons in both the intrinsic flavin fluorescence emission intensity ( em =523 nm) and retention of NADH:FR activity following incubation of the proteins at temperat ures ranging from 0 C were monitored. Changes in the intrinsic fluorescence of the cofactor or the reten tion of NADH:FR activity following thermal denaturation has been shown to be an effective indicator of the stability of the core structure of the protein. The various T 50 values, the temperature at which 50% of maximum fluorescence or 50% retention of NA DH:FR activity was detected, increased in the order M272R < M272A < M272< WT < M272L < M272I, w ith all variants exhibiting T 50 values in the range between 46 and 57 C, which suggested that none of the substitutions had a dramatic effect on the thermal stability of flavin binding with the exception of the M272R variant. The M272I and L variants exhibited T 50 values of approximately 57 C, in good agreement with the value of 55 C obtained for wild-type c b 5 r. The deletion (M272-) and A varian ts exhibited slightly decreased T 50 values of 51

PAGE 143

Temperature (oC) 0102030405060708090100 % Residual Avtivity (NADH:FR) 0255075100Fluorescence Intensity % 0255075100 Figure 43. Thermal Stability Profiles for the Various M272 cb 5 r Mutants. Oxidized samples of the various M272 variants together with wild-type cb 5 r (5 M FAD) were incubated at the indicated temperatures and aliquots were withdrawn and assayed for both residual NADH:FR activity and intrinsic flavin fluorescence in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0, the latter using excitation and emission wavelengths of 450 nm and 523 nm, respectively. The plots correspond to M272(, ); M272A (,); M272I (, ); M272L (, ); M272R (+, +) and wild type cb 5 r (,). NADH:FR activities were scaled relative to a sample of each protein that had been maintained at 0 o C while intrinsic flavin fluorescence was scaled to a sample of free FAD (5 M) which was assigned a fluorescence intensity of 100%. 125

PAGE 144

126 C and 50 C respectively, while the R mutation cau sed a decrease to 47 C, indicating the most altered structure. These results suggested that the majority of the M272 substitutions had only modest effects on th e thermal stability of the different c b 5 r variants. Initial-rate kinetic analyses were performed on all five c b 5 r M272 variants to evaluate the effects of the various residue substitutions on NADH utilization. Values derived for k cat and K m for both NADH:BR and NADH:FR activities of the various mutants are given in Table 16, together with the corresponding values obtained for wildtype cb 5 r. With the exception of the M272L variant which reta ined 84% of wild-t ype activity with a corresponding two-fold decrease in affinity for NADH, the remaining four variants exhibited both substantially decreased NADH:FR activities together with either modest or substantial decreases NADH affinities with the M272varian t showing the most dramatic changes. NADH catalytic efficiencies, as indicated by k cat / K m NADH were observed to decrease in the order WT > M272L = M272I > M272A > M272R > M272with the latter variant retaining a level of NADH:FR efficiency that was five orders of magnitude less than that of the wild type protein. Of the five M272 variants, both the deletion mutant and the M272R variant were dete rmined to have the lowest affinities for NADH. To confirm that substitution or deleti on of the M272 residue primarily affected the proteins affinity for either the phys iological reducing substrate NADH, or the product of the redox half-reaction, NAD + we utilized differential spectroscopy of the flavin cofactor absorbance to determ ine spectroscopic binding constants, K s for the

PAGE 145

127 Table 16. Comparison of the Kinetic Constants Obtained for Wild-Type Cytochrome b 5 Reductase, M272-, M272A, M272I, M272L, and M272R Mutant NADH:FR NADH:BR k cat K m NADH K m FeCN6 k cat / K m NADH (s -1 ) (M) (M) (s -1 M -1 ) k cat K m NADH K m Cyt b 5 k cat / K m Cyt b 5 (s -1 ) (M) (M) (s -1 M -1 ) H 4 c b 5 r M272M272A M272I M272L M272R 800 21 6 1 8 1 1.4 0.3 x 10 8 0.4 0.1 70 11 7 1 5.9 1.2 x 10 3 406 11 11 1 8 1 3.7 0.6 x 10 7 503 10 9 1 8 1 5.5 0.6 x 10 7 670 5 12 1 7 1 5.5 0.2 x 10 6 478 13 6 1 8 1 1.4 0.3 x 10 6 600 10 2 1 13 1 4.7 0.5 x 10 7 444 24 3 1 15 2 2.9 0.5 x 10 7 384 26 3 1 15 2 2.7 0.6 x 10 7 443 43 2 1 26 4 1.8 0.5 x 10 7 54 7 5 1 46 5 1.2 0.3 x 10 6 386 37 4 1 19 1 2.1 0.6 x 10 7 NADHanalog, H 4 NAD and NAD + (Figure 44). Changes in the FAD absorption spectrum following binding of H 4 NAD to the respective proteins were observed for the M272L, M272I and M272A variants, however, no spectroscopic cha nges were detected for either the M272or M272R variants, sugge sting substantially altered affinities. For the M272L, I, and A variants, characterist ic spectral changes were observed with lineshapes comparable to those recorded for the wild-type enzyme following complex formation with H 4 NAD. Similar behavior was observed for the differe ntial spectra obtained from the titrations of the various mutants with NAD + While the M272A, I and L mutants generated spectra that were comparable both in lineshape a nd intensity to those obt ained for wild-type c b 5 r, no differential spectra were obtained for either the M272or M272R variants at

PAGE 146

128 concentrations of NAD + up to 10 mM, indicating either su bstantially decreased affinities for both mutants or an altered mode of NAD + binding (Table 17). Figure 44. Flavin Difference Spectra Obtained Following Binding of Either H 4 NAD or NAD + to the Various M272 cb 5 r Mutants Difference spectra were obtained for the different M272 variants and wild-type cb 5 r at equivalent flavin concentrations (50 M FAD) in 20 mM MOPS buffe r, containing 0.1 mM EDTA, pH 7.0 following titrations with either H 4 NAD (left panels) or NAD + (right panels) as described in Methods. The insert panels correspond to plots of the magn itudes of the observed spectral perturbations (peak to trough measurements at the indicated wavelengths) versus ligand concentration. Plots of the relative absorbance changes observed are as follows: A and B: H 4 c b 5 r; C and D: M272A; E and F: M272I; G and H: M272L; I and J: M272(Titrations of M272R yielded spectra identical to thos e of M272-). The corresponding K s values are given in Table 17.

PAGE 147

-0.04 0.00 0.04 0.08 [H4NAD] (mM) 0.0030.6 Absorbance (A 410 -A 460 ) 0.00 0.01 0.02 -0.04 0.00 0.04 0.08 0.12 024 Absorbance (A 410 A 460 ) 0.00 0.04 0.08 -0.04 0.00 0.04 0.08 [H4NAD] (mM) 0.0020.4 Absorbance (A 395 A 480 ) 0.000 0.015 0.030 -0.03 0.00 0.03 0.06 0.09 048 0000 0017 0034 -0.04 0.00 0.04 0.08 [H4NAD] (mM) 0.0020.4 Absorbance (A 410 A 460 ) 0.000 0.009 0.018 -0.04 0.00 0.04 0.08 0.12 024 Absorbance (A 410 A 460 ) 0.00 0.05 0.10 -0.04 0.00 0.04 0.08 0.0051.0 0.000 0.015 0.030 -0.04 0.00 0.04 0.08 0.12 [NAD + ] (mM) 024 Absorbance (A 410 A 460 ) 0.00 0.04 0.08 Absorbance (A 410 A 460 )[H4NAD] (mM) Absorbance (A 395 A 480 )[NAD + ] (mM) [NAD + ] (mM) [NAD + ] (mM) 380435490545600 -0.04 -0.02 0.00 0.02 0.04 Wavelength (nm) 380435490545600 Absorbance -0.04 -0.02 0.00 0.02 0.04 A G H F E D C B I J 129

PAGE 148

130 Table 17. Spectral Binding Constants Obtained for Wild Type cb 5 r and the M272-, M272A, M272I, M272L, and M272R Variants. Mutant K s H4NAD (M) K s NAD+ (M) H 4 c b 5 r 45 10 760 30 M272ND ND M272A 98 11 891 51 M272I 204 10 1078 66 M272L 106 7 2293 125 M272R ND ND *ND indicates that either there was no detectable activity or no detectable spectroscopic change associated with the purified c b 5 r variant. To examine whether the substitution or deletion of residue M272 influenced NADH utilization through modulation of the flavin oxidation-reduction midpointpotential, potentiometric titrations we re performed using th e dye equilibration method for wild-type c b 5 r and the different M272 variants in the presence of phenosafranine (E o = -252 mV) as indicator dye. Fl avin midpoint potentials (E o , n=2) were determined for the enzyme s alone and in complex with NAD + The spectra obtained during representative titrations of the M272A variant in the absence of NAD + are shown in Fig. 6A. Qualitative analysis of the individual spectra obtained from the various titrations indicated that the majority of th e phenosafranine was reduced prior to FAD reduction for the majority of the M272 variants and wild-type c b 5 r in the absence of any pyridine nucleotide, suggesting the flavin midpoint potentials were more negative than that of phenosafranine, for all the c b 5 r variants examined except for the M272mutant.

PAGE 149

131 For the M272variant, the spectra indica ted that both the pr otein and dye were effectively reduced simultaneously, suggestin g a significantly more positive potential for the FAD/FADH 2 couple in this mutant than observed for wild-type c b 5 r or any of the other M272 mutants. In contrast to the results obtained for the different M272 variants in the absence of NAD + titrations performed in the presence of NAD + revealed very different redox behavior. Analysis of the spectra obtained fo r the M272A, I and L vari ants together with wild-type c b 5 r indicated that all four proteins were reduced well in advance of the phenosafranine dye, suggesting that complex fo rmation significantly perturbed the flavin midpoint potentials to values more positive than that for phenosafranine for these three mutants. In contrast, the presence of NAD + had a markedly decreased effect on the potential of the FAD/FADH 2 couple in the M272R variant and effectively, no effect on the M272mutant, suggesting little pert urbation or modulation of the FAD redox potential in these two variants. The flavin redox potentials (n=2) for the different M272 varian ts and wild-type c b 5 r alone or as the enzyme-nucleotide comple x were determined from the Nernst semilog plots shown in Figure 45. The standard midpoint potentials obtained for the FAD/FADH 2 couple in both the wild-type enzyme (E o = -272 mV) and the M272A, I and L variants (E o = -271 mV to -275 mV) were appr oximately equivalent for all three proteins in the absence of any pyridine nucle otide, the values spanning a range of only 4 mV. In contrast, the flavin midpoint potentia l in the M272and M272R variants were determined to be 27 mV and 12 mV, respect ively, more positive than the corresponding midpoint potential observe d for the wild-type c b 5 r.

PAGE 150

132 Figure 45. Comparison of the Oxidation-Re duction Midpoint Potentials for the FAD Prosthetic Group in the Different M272 Variants. Reductive dye-equilibration titrations of wild-type and the different M272 variants of c b 5 r (40 M FAD) were performed as described under Methods in 100 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0 in the presence of phenosafranine (15 M, E o =-252 mV). Individual spectra were collected at 2-3 min intervals du ring the time course of the titrations. (A): Selected spectra obtained during the titration of the M272A mutant. (B): Nernst plots obtained for the FAD/FADH 2 couple (n=2) are shown for th e titrations of the various M272 c b 5 r mutants in the absence of NAD + and correspond to: M272(+, E o =-245+ 5 mV), M272A ( E o =-271+ 5 mV), M272I ( E o =-274+ 5 mV), M272L ( E o =275+ 5 mV), M272R ( E o =-258+ 5 mV) and wild-type c b 5 r ( E o =-272+ 5 mV). (C) Nernst plots obtaine d for the FAD/FADH 2 couple (n=2) are shown for the titrations of the various M272 cb 5 r mutants in the presence of NAD + and correspond to: M272(+, E o =-244+ 5 mV), M272A ( E o =-192+ 5 mV), M272I ( E o =-192+ 5 mV), M272L ( E o =-191+ 5 mV), M272R ( ,E o =-220+ 5 mV) and wild-type c b 5 r ( E o =-190+ 5 mV).

PAGE 151

Potential (mV) -280-260-240-220-200-180-160 Log (oxidized/reduced) [FAD] -1.0-0.50.00.51.0 C Wavelength (nm) 300400500600700 Absorbance 0.00.20.40.60.8 A -310-300-290-280-270-260-250-240-230 log (oxidized/reduced) [FAD] -1.0-0.50.00.51.0 B 133

PAGE 152

134 The decreased affinity for NAD + that was determined for both the M272R and M272mutants was also reflected in alterati ons in the shifts and slopes of the Nernst plots when compared to the wild-type domai n. In contrast, while the presence of NAD + resulted in no perturbation of the FAD/FADH 2 redox potential or change in slope for the M272mutant, for the M272R variant, the degree of perturbati on was substantially decreased which was coupled with no apparent change in slope of the Nernst plot suggested that nucleotide binding did not result in any appreciable semiquinone formation. In the presence of NAD + the redox potential of the FAD/FADH 2 couple in the wild-type enzyme was posit ively-shifted by +77 mV (E o = -195 + 5 mV) which was comparable to the magnitude of the positive shifts observed for the potentials of the M272A, I and L variants. In contrast, substa ntially small shifts were observed in the potentials of the FAD/FADH 2 redox couple in both the M272R and M272mutants, which corresponded to +37 mV and +1 mV, respectively. To assess the potential struct ural changes that would be anticipated to result from the deletion of the M272 resi due, we utilized homology modeli ng to construct a model of the M272variant using the experimental wild-type structures as templates. The superposition of the structures of the wild type enzyme and the M272deletion mutant are shown in Figure 46. The modified loop containing the CGxxxxM NAD(P)H binding motif was strongly anchored by its connection to an interior strand. In the wild type structure, M272 and C273 both occupy positions close to those expected for strand extensions, but the final flanking strand does not extend suffi ciently far enough to reach these residues and they are s lightly displaced relative to the neighboring interior strand residues, making any backbone hydrogen bonds very weak. The residue preceding M272

PAGE 153

135 Figure 46. Comparison of a Structural Model of the M272Mutant and the WildType c b 5 r Diaphorase Domains. The X-ray structures obtained for the rat c b 5 r diaphorase domain in the absence (PDB ID = 1I7P) and complex with NAD + (PDB ID = 1IB0) were used as templates to generate a structural model for th e M272variant. (A) The overall backbone architecture of the wild -type (yellow) and M272mutant (red) are superimposed and displayed as a cartoon ri bbon model. The FAD prosthetic group is shown in ball and stick representation w ith individual atoms colored using the CPK convention. (B) An expanded portion of the PDB 1IB0 structure is di splayed that shows the spatial relationship between the region of the polypeptide seque nce corresponding to residues M272 to M278 and the bound NAD +

PAGE 154

A B 136

PAGE 155

137 is a full participant in interior sheet hydrogen bonding. Thus, the removal of residue M272 in the deletion mutant forces the subsequent cysteine residue (C273) to occupy the position comparable to that of M272 in the wild-type stru cture. Consequently, this deletion shortens the loop connecting the strand with the following -helical return. A primary consequence of this structural alteration is represented in figure 46 where the displacement of the loop backbone is shown by the path of the green ribbon that corresponds to the mutant structure in comparison to the red ri bbon that corresponds to the experimental wild-type structure. The wild type polypeptide makes five hydrogen bonds with NADH (a sixth is supplied by bound water); two of these would be disrupted by displacement of the loop, and th e hydrophobic interact ions between NAD + and the proline residues would also be weakened. Po ssible secondary effects could also result from the structural alterations. To examine the predictions generated by the modeling studies of the M272variant and the comparison with the experimental structur es of the wild-type enzyme, we examined the binding of two NAD + analogs, 5-ADP-ribose and 5-ADP, which correspond to successively truncated derivatives of NAD + using differential spectroscopy. The structural model derived for the M272varian t indicated that if the loop region containing the CGxxxxM motif was displace d relative to its pos ition in the wild-t ype structure, the major predicted impact would be to decrease the affinities for both 5-ADP-ribose and 5ADP, since the X-ray structure of wild-type c b 5 r in complex with NAD + has revealed that the CGxxxxM motif is primarily involved in interactions w ith the adenosine portion of NAD + where it contributes an extensiv e network of hydrophobic contacts. The differential spectra shown in Figur e 47, revealed that while both NAD + fragments form

PAGE 156

-0.03 0.00 0.03 0.06 0.09 Absorbance (A 410 A 460 )[NAD + ] (mM) 036 0.000 0.017 0.034 -0.02 0.00 0.02 0.04 0.06 Absorbance -0.02 0.00 0.02 0.04 [ADP Rbose] (mM) 0.00.20.4 Absorbance (A 410 A 500) 0.000 0.008 0.016 -0.02 0.00 0.02 0.04 0.06 Wavelength (nm) 380435490545600 -0.02 0.00 0.02 0.04 0.06 [5' ADP] (mM) 036 Absorbance(A 400 A 501 ) 00000 00023 00046 380435490545600 -0.02 0.00 0.02 0.04 0.06 AB C D EF Figure 47. Difference Spectra and SpectralBinding Constant Determination for Wild Type Cytochrome b5 Reductase and the M272Variant with various Pyridine Nucleotide analogs. Enzyme samples (50 M FAD) in 10mM MOPS buffer, containing 0.1 mM EDTA, pH 7.0 were titrated with the in dicated pyridine nucleotides {Panel A wild-type cb5r + NAD+, Panel B M272+ NAD+, Panel C wild-type c b5r + ADP ribose, Panel D M272+ ADP ribose, Panel E wild-type c b5r + 5 ADP, Panel F M272+ 5 ADP} as described in Materia ls and Methods. The difference spectra obtained during the titrations are shown. The inset panels correspond to a plot of the spectral perturbations (peak to trough s ubtraction) versus pyridine nucleotide concentration. 138

PAGE 157

139 spectroscopically-detectable complexes with the M272-variant, suggesting either substantially-decreased affinities for both 5-ADP-ribose and 5-ADP or that both ligands were bound in very different conf ormations than their respectiv e orientations in the wildtype complexes, confirming the results that were predicted from the M272modeling studies that suggested that the displacement of the hydrophobic CGxxxxM loop would have a substantial adverse impact on the affinity for NADH and the various analogs. Summary of the Recessive Congenital Methemoglobinemia Variant M272An extensive number of mutations have been identified in the DIA1 gene that codes for c b 5 r and result in RCM. Previous studies of the majority of the known c b 5 r methemoglobinemia mutants that utilized a novel c b 5 /c b 5 r fusion protein have indicated that the M272variant was the most compromised mutant in terms of specific activity [40]. Davis et al have shown that using the NCR expression system, the M272variant only retained <0.01% of the w ild-type NADH:CR specific activity. The model structure described in the pr eceding results section demonstrates that effects involving the CGxxxM motif are unavoidable in the M272mutant. These effects could account for the substantially -weakened binding of NADH to the M272mutant, when compared to the NADH-affin ities of the other M272 variants, if the combination of hydrogen-bonding to the phosph ate groups and hydrophobic interactions between NAD(H) and this motif are a significant factor in NADH binding. In the wild type protein, the hydr ophobic C273 side chain is positioned approximately 4.2 from the si -face of the FAD isoalloxazine ring system, and is in Van der Waals contact with the nicotinamide moiety of NAD + This suggested the possibility

PAGE 158

140 that C273 is involved in cont rolling the equilibrium betwee n the unstacked configuration of bound NADH and the charge transfer comp lex. In the M272deletion mutant, the cysteine side chain would be displaced into approximately the position of the M272 wildtype side chain, and the next residue, G273, would be unable to compensate because it effectively lacks a side chain. Secondary effects could also play a role in the reduced affin ity of the deletion mutant for NADH. Such effects could be pr opagated through small displacements in the neighboring loops which make up nearby sections of the NADH binding site. For example, a small reorientation of the -helical region following the MCGxxxM motif could reduce the space available for bi nding the adenine moiety, forcing some reorganization of the loop containing F251, th e stacking residue for adenine in the wildtype structure. The possible contribution of secondary eff ects of this kind coul d be investigated by measuring the affinity of phosphorylated ad enenines for the site. In homologues such as cytochrome P450 reduc tase and Ferredoxin:NADP + reductase, partial NADPH analogs such as 2-AMP and 2-ADP are inhibitors (and in some cases activators) because they compete with NADPH. The FAD and pyridine nu cleotide binding sites in these enzymes have features which cb 5 r lacks, including a stacking residue supplied by the carboxylterminal region for the FAD isoalloxazine and interdomain hydrogen bonds to the pyridine nucleotide.

PAGE 159

141 Characterization of the type I Recessive Congenital Methemoglobinemia mutant P275L Using the high-resolution x-ray structural model of the soluble rat c b 5 r diaphorase domain, [21] which has been solved in both the native state and complexed with NAD + and shares 95% sequence conservation with th e human variant, it was possible to make predictions regarding the effects of natura lly occurring mutations on the function of c b 5 r. With the development of a heterologous expr ession system [21] [79] it was possible to confirm these predictions and to provide impor tant insights into the structure-function relationships of variants w ith a view to dissecting the mechanisms involved in the development of both types of RCM. For exampl e, previous studies of the S127P type II variant using the rat model were the first to provide detailed structural evidence to explain the effects of this mutation in terms of perturbation of the conformation of the bound FAD prosthetic group substantially decreas ing the enzymes catalytic efficiency [80]. We have identified an infant, cyanotic at birth, who was found to be compound heterozygous for a novel mutation, P275L, and the previously desc ribed G291D mutation [107] [85]. Expressing the G291D variant had indicated th at the enzyme activity and thermostability of the c b 5 r variant were decreased [79] but it was unknown if the two mutations together would have a synergistic effect. The rat model indicated that the P275 residue was located in a highly conserved re gion that interacted with the NADH reducing substrate [21]. To understand how substituti on of P275 would affect the function of c b 5 r, the P275L variant was generated using our he terologous expression system and enzyme activity, thermostability, and substr ate affinity were determined.

PAGE 160

142 The propositus was the first live child of non-consanguineous parents of English/Scottish and Northern European with Cherokee Indian descent. He was born at full term from spontaneous vaginal delivery to a 31-year-old mother with a history of depression, Crohns disease and anklyosin g spondylitis. Bextra for anklyosing spondylitis and Wellbutrin for de pression were ceased when the mother became pregnant but she remained on Prozac for depressi on and Asacol for Crohns disease. The pregnancy was otherwise unremarkable but a nuchal cord was present at delivery. Apgar scores were recorded as 7 at 1 minute and 8 at 5 minutes. At 14 hours of age, the infant was noted to be cyanotic even though his O 2 saturations for hemoglobin were in the range of 90% and the methemoglobin level was found to be 13.2%. The infant was administered a dose of intravenous met hylene blue and the methemoglobin level promptly fell to 1.8%. It slowly increased over the next two w eeks to reach 10.2%. Several checks over the next few months indi cated that he had a baseline methemoglobin level ranging between 10.2% and 11.2%. Measurement of c b 5 r activity detected a level of 0.5 IU/g Hb (normal range 8.2-19.2 IU/g Hb) at two and seven weeks and on subsequent measurement at 14 months was 0.6 IU/g Hb. Both parents were found to have normal c b 5 r activity with measurements of 13 and 18 IU/g Hb for the father and mother respectively. Clinically, he had some mild perioral cyanosis but remains well at 14 months without medication. The DIA1 gene of the cyanotic infant was sequenced and a novel C to T change in exon 9 (C27161T; NCBI accession number: NT_011520) and a G to A change, also in exon 9 (G27209A; NCBI accession number: NT_011520) were detected, as shown in Figure 48. These nucleotide changes predicted P275L and G291D mutations.

PAGE 161

143 Sequencing both phenotypically normal parents re vealed that neither mutation had arisen de novo in the infant with the father and th e mother being heterozygous for the P275L and G291D mutations, respectively. The G291D variant has been described before [78] in the original family described by Gibson [44] and a second Nort hern Ireland family [107]. It is likely that the G291D variant had arisen indepe ndently in Northern Ireland and America, but performing haplotype studies may indicate if ther e is indeed a common origin for the G291D variant. Both muta tions are located in the NADH-binding lobe of c b 5 r and it is unknown if the two mutations toge ther would have a synergistic effect. Using the rat model, it was possible to predic t that since P275 is located in a conserved sequence motif corresponding to 273 CGPPPM 278 it is important for binding and correct positioning of the NADH reducing substrate, thus the P L mutation would adversely affect functionality. Thus, substitution of proline with leucine w ould be expected to result in the leucine side chain extending in to the NADH-binding cavity thus sterically precluding NADH binding in the normal wild-type position. A mutant construct encoding the P275L c b 5 r variant, which corresponded to the third residue in the conserved CGPPPM sequence motif, was genera ted using site-directed mutagenesis of the original four-histidine tagged c b 5 r construct. Dideoxy sequencing confirmed that the fidelity of the construct and the mutant protein was subsequently expressed in the E. coli strain BL21(DE3)-RIL and purified by metal-chelate affi nity chromatography and gel filtration FPLC. Evaluation of the expressi on yield of the P275L variant indicated an expression efficiency comparable to that of the wild-type c b 5 r domain, indicating the production of a stable protein product which could be purif ied to apparent homogeneity as indicated by the presence of a single protein band following

PAGE 162

27,153 27,173 27,998 27,218C FGTGTGGCCNCCCACCCATGAT A DGGACCACGTGGNCCACCCCAC 27,153 27,173 27,998 27,218B EGTGTGGCCCCCCACCCATGAT GGACCACGTGGNCCACCCCAC GTGTGGCCNCCCACCCATGAT GGACCACGTGGGCCACCCCAC 27,153 27,173 27,998 27,218 Figure 48. Detection of the C27161T and G27209A Mutations in the DIA1 Gene. Sequencing of exon 9 of the DIA1 gene detected two heterozygous base changes of C to T at nucleotide 27,161 (A) and G to A at nucleotide 27,209 (D) in the propositus as indicated by arrow in chromatogram. The father (B and E) was heterozygous for the G27209A change while the mother (C and F) was heterozygous for the C27161T mutation. Bases are as follows: G = black; A = green; T = red; C = blue. 144

PAGE 163

145 SDS-PAGE analysis (Figure 49). The P275L c b 5 r variant and the native enzyme were of comparable molecular mass, while MALDI-TOF analyses revealed the presence of a characteristic peak in the low mass region of the spectrum with a m/z of 792, indicative of the presence of FAD as the sole prosthetic group. These findings s uggested that residue P275 did not provide backbone or side-chain co ntacts that were essential for the stable incorporation of the FAD prosthetic group into c b 5 r. UV/Visible absorbance spectra were obtai ned for both an oxidized sample of the P275L variant and wild-type c b 5 r as a control and are presented in Figure 50A The P275L variant exhibited a spectrum comparable to that of th e wild-type enzyme with an aromatic maximum observed at 273 nm in the UV region of the spectrum, and a peak at 461 nm with an associated pronounced shoulder in the range of 485-500 nm in the visible region of the spectrum, attributab le to protein-bound flavin. To compare the secondary structural conten t of the P275L variant with that of the wild-type protein, circular di chroism (CD) spectra were recorded in the UV wavelength range (190-300 nm). As shown in Figure 50B, the P275L variant exhibited positive CD from 190 to 210 nm and the negative CD from 210 to 250 nm with the spectrum retaining both negative and positive intens ities very similar to those of the wild-type domain. The absence of any significant differences between the spectra of the mutant and wild-type protein suggested secondary structural c onservation and that the P to L residue substitution had no deleterious effects on the folding of the diaphorase domain. Similar results were observed for the corresponding visible CD spectra. CD measurements were performed in the near UV/Vi sible range (300-600 nm) in or der to probe the possible effects of the P275L mutation on both flavin conformation and polarity of the prosthetic

PAGE 164

25 37 50 kDa S A B Figure 49. SDS-PAGE Analysis of the P275L cb5r Variant. Samples (2 g protein) obtained following isolation of th e P275L variant and wild-type c b5r were analyzed using a 15 % polyacrylamide gel as described in Methods. Individual lanes correspond to: S, protein molecular weight markers; A, P275L variant; B, wild-type c b5r. group microenvironment. As shown in Figure 50C, the spectrum of the P275L mutant was unperturbed when compared to that of wild-type c b5r, exhibiting positive and negative maxima at approximately 400 and 460 nm, respectively, suggesting no significant change in either the polarity of th e flavin environment or the conformation of the bound prosthetic group. To examine the influence of the P275L re sidue on the stability of the resulting protein, thermal denaturation profiles were generated for both the P275L and wild-type c b5r by measuring changes in both intrin sic flavin fluorescence emission ( ex = 450 nm and em = 523 nm) and retention of NADH:ferricyanide reductase (NADH:FR) activity following incubation of the proteins at temper atures ranging from 0 C to 100 C (Figure 51). Thetemperature ( T50) at which 50% of maximum fluo rescence was detected or 50% of the proteins NADH:FR activity was retained, were effective indicators of the stability of the core structure of the protein. The T50 value obtained for the P275L variant was 146

PAGE 165

Wavelength (nm) 250300350400450500550600 Absorbance 0.00.20.40.60.81.01.21.4 320355390425460495530 0.000.150.30 Wavelength (nm) 300350400450500550 Molar Ellipticity (X 104) -4-2024 190250 Molar Ellipticity (X 106) -4-2024 ABC Figure 50. UV/Visible Absorption and CD Spectra of the P275L cb 5 r Mutants. (A) UV/visible absorption spectra were obtained for oxidized samples of the P275L mutant and wild-type cb 5 r, as a control, at equivalent flavin concentrations (10 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. The inset shows an expanded region of the visible spectrum where the flavin prosthetic group makes a major contribution. (B) UV CD spectra were recorded using enzyme samples (7 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. (C) Visible CD spectra were recorded using enzyme samples (50 M FAD) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. Individual spectra correspond to P275L ( ); wild-type cb 5 r ( ___ ). 147

PAGE 166

148 approximately 2 C lower than that of wild-type c b 5 r, suggesting that the residue substitution had no significant effect on the thermal stability of the mutant protein. To further confirm that the proline to leucine substitution had no substantial effects on the flavin environment, oxidationreduction potentials were determined for the FAD/FADH 2 couple (E n = 2) in both the P275L va riant and the wild-type c b 5 r using dye-equilibration potentiometery in the presence of phenosafranine as a redox indicator. Representative spectra obtained during the re dox titration of the P 275L mutant are shown in Figure 52A, and provide an example of the natu re of the spectral changes observed during the dye-equilibration re dox titrations while the co rresponding Nernst plots obtained for both P275L and wild-type domai n, both in the absence and presence of NAD + are presented in Figure 52B. The redox potential of the P275L variant was within 10 mV of wild-type cb 5 r in the absence of NAD + confirming that the mutation had little effect on the thermodynamic properties of th e FAD prosthetic group. However, in contrast, the redox potential obtained for the P275L variant in the presence of NAD + was significantly more negative than that obtained for the wild-type domain in the presence of NAD + indicating a decreased affinity of the mu tant for binding the pyridine nucleotide. As a measure of catalytic efficien cy, both NADH:FR and NADH:BR activities were determined for the P275L variant a nd compared with th at of wild-type c b 5 r. Kinetic constants derived from these a ssays are reported in Table 18, and indicated that the P275L mutant exhibited a decr eased NADH:FR turnover number compared to that of wild-type c b 5 r.Comparison of the specific activities of the mutant and wild-type c b 5 r indicated that the mutant retained 92% of wild-type NADH:FR activity. However, the Michaelis constant ( K m ) for NADH utilization increased 437 fold, from 6 M to 2.6 mM,

PAGE 167

Temperature (oC) 020406080100 % Residual Activity (NADH:FR) 0255075100Fluoresecence Intensity (%) 0255075100 Figure 51. Temperature Stabilities of the P275L cb 5 r Mutant. Samples of the purified P275L mutant and wild-type cb 5 r (20 M) were incubated in microfuge tubes as described in Methods. Flavin fluorescence emission spectra were determined at each temperature and the emission maxima at 524 nm plotted as a percentage of total fluorescence. Total flavin fluorescence was determined by measuring a similar dilution of the remaining stock solution following heating at 100 C for 30 min. P275L (); wild-type cb 5 r (). Residual NADH:FR activities were determined at 340nm for each temperature as described in Methods and correspond to P275L () and wild-type cb 5 r (). 149

PAGE 168

150 Figure 52. Potentiometric Titrations of the P275L c b 5 r Mutant Reductive dyeequilibration titrations were performed at 25 o C as described under Methods using phenosafranine (E o =-252 mV) as the indicator in 100 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. (A) Representative spect ra obtained during the titration of P275L in the absence of NAD + are displayed. (B) The Nernst plots of log ([oxidized]/[reduced]) FAD versus potential correspond to P275L in the absence ( E o = 264+ 5 mV) and presence ( E o =-212+ 5 mV), respectively of 2 mM NAD + Also shown are the Nernst plots obtained for wild-type c b 5 r as a control in the absence ( E o =-272+ 5 mV) and presence ( E o =-192+ 5 mV), respectively, of 2 mM NAD + Flavin reduction was monitored at 410 nm, an isosbestic wavelength for phenosafranine reduction and phenosafranine reduction wa s monitored at 530 nm, an isosbestic wavelength for flavin reduction. Flavin midpoint potentials for the FAD/FADH 2 couple (n=2) were determined from the Nernst plots and yielded the indicate d standard reduction potentials (E o ). Only selected data points obtained from each of the titrations are shown for clarity.

PAGE 169

Potential (mv) -300-280-260-240-220-200-180-160 Log (Oxidized/Reduced) [FAD] -1.0-0.50.00.51.0 Wavelength (nm) 300400500600700 Absorbance 0.00.20.40.60.8 AB 151

PAGE 170

152 indicating that the dominant effect of the mutation was a substantial decrease in the affinity for the reducing substrate. In contra st, the Michaelis constant for the artificial electron acceptor ferricyanide was similar to that of wild-type c b 5 r, suggesting that the mutation did not affect the binding of this artificial oxidizing s ubstrate. The effect of the proline to leucine substitution on the overa ll NADH:FR catalytic efficiency of the enzyme, reflected in the k cat / K m NADH value, was to retain only 0.2% of wild-type NADH:FR catalytic efficienc y. These results suggested that the lower catalytic efficiency observed for the mutant was th e result of the decrea sed affinity for NADH, rather than any decreased affinity for c b 5 Assays were also performed to ascertain if the proline to leucine substitution had any effect on the selectivity of the enzyme for either NADH or NADPH. As shown in Table 18, the NADPH-dependent diaphorase activity decreased together with the affinity for NADPH. However, the specificity constant, defined as the ratio of the catalytic efficiencies with NADPH and NADH, increased 5-fold when compared to the value obtained for the wild-type enzyme, indicating th at P275 plays only a very modest role in regulating NADH/NADPH specificity. To confirm the effects of the P275L substitution on the affinities for either NADH or NAD + spectral binding constants ( K s ) were determined using differential spectroscopy. Representative spectra obtaine d from the P275L mutant and the wild-type enzyme with either H 4 NAD or NAD + is shown in Figure 53. While perturbations of the flavin visible absorbance spectrum were de tected for the wild-type enzyme, no complex formation could be detected for the P275L variant with either compound, reflecting the substantially decreased affinity for either H 4 NAD or NAD +

PAGE 171

153 Table 18. Comparison of the Kinetic and Spe ctroscopic Binding Constants Obtained for the P275L Variant of c b 5 r. Mutant NADH:FR NADH:BR NADPH:FR Specificity Constant a k cat K m NADH k cat / K m NADH (s -1 ) (M) (s -1 M -1 ) k cat K m Cyt b5 (s -1 ) (M) k cat K m NADPH k cat / K m NADPH (s -1 ) ( M) (s -1 M -1 ) P275L Wild Type 733+ 7 2623+ 460 2.8+ 0.1x10 5 800+ 21 6+ 1 1.4+ 0.3x10 8 67 + 18 4 + 1 600 + 17 12 + 1 0.45 1123 4.0x10 2 33 924 3.6x10 4 1.4x10 -3 2.7x10 -4 Summary of they Type I RCM Variant P275L The identification of the novel P275L RCM variant represents the discovery of the eighth mutation in exon 9 of the DIA1 gene. Five variants (V252M, R258W, P264L, P275L, and G291D) and a deletion mutant (E 255-) [82-85, 106, 116] ha ve been found to be associated with type I RCM while two other deletion mutants (M272and F298-) [31, 81] 273 CGPPPM 278 that is present in the majority of known c b 5 r sequences and is also found in other members of the ferredoxin:NADP + reductase superfamily of flavoprotein transhydrogenases [5]. The X-ray structure of the rat c b 5 r diaphorase domain complexed with NAD + (PDB code: 1IB0) has revealed that P 275 is part of a solvent-accessible loop region located between sheet N 5 and helix N 4 within the carboxy-terminal pyridine nucleotide-binding lobe [21]. Within this structure, NAD + has been shown to bind in an extended conformation along a plateau that is formed by three proline residues, P275P277. The conformation of the bound NAD + is shaped largely by P275 and bends around it with the pyrophosphate moiety packed agai nst it and with the nicotinamide proximal ribose packed parallel to the proline ring (Figure 54B), suggesting that substitution of

PAGE 172

380435490545600 -0.03 0.00 0.03 0.06 0.09 [H4NAD] (mM) 0.0020.4 Absorbance (A 395 -A 480 ) 0.000 0.015 0.030 Wavelength (nm) 380435490545600 -0.03 0.00 0.03 0.06 0.09 [NAD+] (mM) 048 000 002 004 Absorbance (A 410 -A 460 )C D Absorbance -0.03 0.00 0.03 0.06 0.09 -0.03 0.00 0.03 0.06 0.09 B A Figure 53. Spectroscopic Titrations of the P275L Mutant with H4NAD and NAD+. Titrations of the P275L c b5r variant (50 M FAD) were performed in split cell optical cuvettes in 10 mM phosphate buffer, co ntaining 0.1 mM EDTA, pH 7.0, at 23 C. Difference spectra were recorded, following additions of a concentrated solution of either H4NAD (50 mM) or NAD+ (100 mM) to both sample and reference cuvettes, after the complete cessation of all absorbance changes. The inset panels correspond to plots of the magnitude of the spectral perturbations at the indicated wavelengt hs versus pyridine nucleotide concentration where a difference spec trum was observed. The plots of relative absorbance changes observed du ring the titrations are as fo llows: (A and B), P275L in the presence of H4NADH and NAD+. (C and D) wild-type cb5r in the presence of H4NADH and NAD+. 154

PAGE 173

155 P275 would be anticipated to have profound effects on both binding and subsequent orientation of the nucleotide. The results of the biophysical characte rization of the P275L variant are in full agreement with this predic tion. The proline to leucine substitution was observed to have no detectable effects on ei ther the spectroscopic or thermodynamic properties of the flavin prosth etic group in th e absence of NAD + In addition, protein folding and stability were unaffected by the mu tation. In contrast, the proline to leucine substitution exhibited a profound e ffect on both the affinity of c b 5 r for the reducing substrate and its catalytic efficiency. Homology modeling of the P275L structure suggested that s ubstitution of the hydrophobic leucine side chain would not be expected to perturb the overall folding and structural organization of the protein as shown in the overlay of the ribbon diagrams for the P275L and wild-type proteins in Figure 54A. However, the linear isobutyl side chain would be anticipated to protrude into the NADH-binding site when compared to that of the wild-type proline structure, as shown in the electrostatic surface plots in Figure 54B and Figure 54C, resulting in steric hindrance and d ecreased binding affinity. These predictions are in agreement with the results of both the initial-rate kinetic and pyridine nucleotide K s analyses. The absence of any spect ral changes observed in either the H 4 NAD or NAD + titrations, suggested that the K s s for these two compounds would be expected to be greater than approximate ly 1 and 5 mM respectively, which are substantially greater than the values of 46 and 663 M, respectively, obtained for the wild-type domain. The molecular mechanisms responsible for the striking difference in phenotype between type I and type II RCM have not yet been satisfactorily explained. Studies of

PAGE 174

156 P275 A B C D Figure 54. Homology Model of the P275L c b5r Variant. (A) Superposition of the rat c b5r X-ray structure (yellow; PDB code 1IB0) and the homology modeled P275L (red) structure. Both structures are shown as ri bbon diagrams that represent the regions of secondary structure. The FAD prosthetic group, complexed NAD+, and residue P275, indicated by the arrow, are shown in s tick representation using the CPK coloring scheme. (B) Rendering of the sequence re gion, corresponding to I270 to F281, that surrounds residue P275. The peptide backbon e is drawn in ribbon configuration while the proline (gray) and leucine (magenta) si de chains at position P275 in the position 275 in the wild-type and mutant proteins are show n in stick representations. Also shown is the complexed NAD+ in stick configuration with CPK coloring. (C and D) Renderings of the sequence region, corres ponding to I270 to F281, which surrounds residue P275. The peptide backbone is show n in ribbon configuratio n while the proline (C) and leucine (D) side chains at position 275 in the wild-type and mutant proteins together with the complexed NAD+ are shown in an electrostatic surface representation with negatively and positively charged regi ons colored in red and blue, respectively.

PAGE 175

157 naturally occurring mutations of cb 5 r using the heterologous expression system will permit accumulation of data on the structure-func tion relationships of the enzyme. It is of interest that some mutations have been found in both types of disease, while the majority appear to be confined to one type The novel P275L is the fifth single amino acid substitution in the protein domain enc oded by exon 9. In general these mutations affect the affinity and activity of c b 5 r for NADH and may primarily or uniquely affect the heme iron reductase function of the enzyme. However, c b 5 r has pleiotropic functions and mutations in the more 5 regions of the gene may disrup t some of these thereby causing neurological impairment. Identification and characterization of the novel FAD-binding lobe G75S mutation in cytochrome b 5 reductase: An aid to determine recessive congenital methemoglobinemia status in an infant. Recently, we identified a four-month old infa nt with a novel G to A change in the last base of exon 3 in the DIA1 gene. This resulted in the substitution of glycine with serine at residue 75 in the F AD-binding lobe of cytochrome b 5 reductase. The infant was compound heterozygous for the previously ch aracterized V252M mutation [82-84] which is located in the C-terminal NADH-binding lobe of cytochrome b 5 reductase. Being compound heterozygous for mutations that would impair both the FAD and NADH function of cytochrome b 5 reductase would suggest the mo re severe clinical form of RCM. Thus, to provide insight at the molecu lar level and potentially to assist in the diagnosis of recessive congenital methem oglobinemia status, we generated and characterized the G75S, V 252M and the corresponding doubl e G75S/V252M cytochrome

PAGE 176

158 b 5 reductase variants using the rat heterologous expression sy stem [79, 105]. The patient was compound heterozygous for the G75S and V252M cytochrome b 5 variants and thus synthesized both individual va riants. However, while th e homozygous double mutant did not occur in the patien t, we were interested in genera ting the double mutant as an initial approach to investigate potential synergistic effects on either the structural or functional properties of the enzyme when both mutations occurred together in the same protein. Although the human erythrocyt ic form of cytochrome b 5 reductase was recently described [20], the rat system is the only vari ant that has been used to characterize all the known methemoglobin amino acid mutations in te rms of their effects on the expression, relative activities, and stability of the various proteins [79]. In addition, there is 95% sequence similarity between these two sp ecies with any non-conserved amino acid substitutions occurring in functionall y insignificant regions of the protein. The kinetic, protein st ability, and cofact or oxidation-reduction potential properties were determined for all three varian ts and compared to the properties of wild type cytochrome b 5 reductase and other methemoglobinemia-causing mutations, including E255-, G291D, and D239G [79, 105]. The proposita was the second child of non-consanguineous parents of Mediterranean descent. Cya nosis was noted at birth ac companied by reduced oxygen saturation determined by pulse oximetery but normal pO 2 on arterial blood gas analysis and no evidence of cardiopulmonary abnormality. Methemoglobinemia was detected at a level of 14.4%. At the age of 4 months, the infant was referred for assessment of persistent methemoglobinemia. Physical ex amination revealed no developmental delay or microcephaly. Erythrocyte cytochrome b 5 reductase activity was assayed

PAGE 177

159 spectrophotometrically at 340 nm as the rate of oxidation of NADH in the presence of potassium ferricyanide as previously de scribed [117]. Meas uring cytochrome b 5 reductase activity indicated that the infant showed a marked reduction with an activity of only 0.5 U/gHb while both the father (4.9 U/gH b) and the mother (5.0 U/gHb) had ~50% of the residual enzyme activity as define d by the adult reference range of 7.2-13.9 U/gHb. Sequencing of the DIA1 gene in the infant revealed a novel G to A change in the last base of exon 3 (G15,635A; genomic sequence NCBI NT_011520) and a G to A change in exon 9 (G 27,091A ; genomic sequence NCBI NT_011520), resulting in amino acid exchanges of glycine to se rine at residue 75 and valine to methionine at residue 252, respectively (Figure 55). Screening the parents indicated that the father was heterozygous for the G75S variant, and th e mother was heterozygous for V252M. The G75S mutation is located in the amino-te rminal FAD-binding lobe of cytochrome b 5 reductase, as shown in Figure 56, while the V252M mutation is present in the carboxyterminal NADH-binding domain. Thus, the infant was compound heterozygous for mutations in both the prosthetic group and substrate binding lobes of cytochrome b 5 reductase. In contrast to the G75S muta tion, the V252M variant has been previously described either in the homoz ygous state in both types I and II patients [82, 84] or in combination with the I215T mutation [83], which is also located in the NADH-binding lobe. To determine how these two radically different mutations would affect the function of cytochrome b 5 reductase, both the G75S and V252M variants and the corresponding double mutant were generated using our heterologous expression system. E. coli BL21 RIL cells harboring the expre ssion plasmids for either rat G75S, V252M or G75S/V252M cytochrome b 5 reductase variants were grown, disrupted and the

PAGE 178

160 three mutant proteins were purified using a combination of Ni-NTA agarose affinity and gel filtration chromotographies as previously described [76]. Qualitative analysis indicated that all three mutant s were obtained in substantial quantities and were expressed with efficiencies comparable to that of wild-type cytochrome b 5 reductase. All three proteins were yellow in color indicating fl avin incorporation and were purified to apparent homogeneity as indicated by the appearance of single protein bands following SDS-PAGE analysis as shown in Fi gure 57. The three cytochrome b 5 reductase variants exhibited molecular masses of approximately 32 kDa confirming the production of full length proteins. Spectroscopic analysis, including UV/visible absorption and CD spectra, revealed that G75S, V252M, and G75S/V252M variants retained identical spectroscopic properties to those of the wild-type protei n with UV/visible absorption maxima at 276, 386, and 461 nm, respectively, th at indicated a full complement of FAD incorporation and that the flavin was present in an environment spectroscopically indistinguishable from that of the native domain (Figure 58A). UV and CD spectra indicated an absence of any significant change in pr otein architecture suggesting comparable protein folding (Figure 58B and Figure 58C). Thus, neither the G75S nor V252M, or a combination of both mutations did not appear to adversely a ffect the global folding of the respective proteins while the substantial similarities in the visible CD spectra indicated that the flavin prosthetic groups were in a similar environment in all the proteins examined. In contrast, both NADH:FR and NADH:BR activity assays revealed that theG75S, V252M and G75S/V252M double mutant exhibited substa ntially decreased activity, as shown by the kinetic constant s given in Table 19. The G75S variant

PAGE 179

Figure 55. Detection of the G15,635A and G27,091A Mutation in the DIA1 Gene. Sequencing of exon 3 of the DIA1 gene revealed a heterozygous change of G to A at nucleotide 15,635 in the father (A) and daughter (C) indicated by an arrow in the chromatogram on the negative strand, when compared to normal sequence of the mother (B). Sequencing of exon 9 of the DIA1 gene detected a heterozygous base change of G to A at nucleotide 27,091 in the mother (E) and daughter (F) indicated by an arrow in the chromatogram when compared to the normal sequence of the father (D). Bases are as follows: G Black; A green; T red; C blue. exhibited a specific activity or catalytic constant, k cat corresponding to approximately 17% of the wild-type domain activity, while the affinity for the reduced pyridine dinucleotide was only slightly decreased as indicated by the 1.5 fold increase in the Michaelis Constant (K m ) for NADH, indicting that the G75S variant primarily exhibited a defect in the rate of hydride transfer rather than substantially impaired NADH binding. Overall, the G75S variant retained only 11% of the catalytic efficiency, as indicated by thespecificity constant (k cat /K m NADH ), of native cytochrome b 5 reductase. These results could be contrasted with those obtained for the V252M variant which exhibited only a modest decrease in specific activity, corresponding to an approximate 20% reduction, 161

PAGE 180

Figure 56. X-Ray Crystallographic Structure of Cytochrome b5 Reductase. A ribbon diagram of the structure of the R. norvegicus cytochrome b5 reductase diaphorase domain (PDB = 1IB0) complexed with NAD+ is depicted showing the structural elements that comprise the FADand NADH-bindi ng lobes and the hinge interconnecting region. The FAD-binding lobe is depicted in blue, the NADH-binding lobe is depicted in red with the hinge region depicted in gr een. The FAD-prosthetic group and the bound NAD+ molecule are shown in ball and stick representation using the Corey-PaulingKoltun convection. Te two residues that have been identified in the methemoglobinemia mutants, G75 and V252 are also shown in ball and stick representation. 162

PAGE 181

Figure 57. SDS-PAGE Analysis of Wild Type Cytochrome b 5 Reductase and the G75S, V252M, and G75S/V252M Variants. Purified proteins (2 g each protein) obtained from the final FPLC purification step of all three mutants together with a sample of wild-type cytochrome b 5 reductase were resolved on a 12.5% polyacrylamide gel. Lane A, wild-type cytochrome b 5 reductase; Lane B, G75S; Lane C, V252M; Lane D, G75S/V252M; Lane S, molecular weight standards. The positions of selected molecular weight markers are indicated by molecular masses (kDa). however, as anticipated for a mutation in the NADH-binding lobe, the affinity (K m ) for NADH was decreased approximately 9-fold, indicating that the V252M retained only 9% of the catalytic efficiency of wild-type cytochrome b 5 reductase. The results obtained for the G75S/V252M double mutant were intermediate between those of the two single mutants suggesting the effects of the two individual mutations were not synergistic. Comparison of the k cat and K m values obtained for all three mutants revealed that while the primary effect of the G75S mutation was to decrease the catalytic efficiency of the enzyme, in contrast, the V252M mutation primarily decreased the affinity for the reducing substrate NADH, which resulted in approximately equivalent specificity constants for both mutants. It should be noted that the values obtained for the specificity 163

PAGE 182

190250 Molar Ellipticity (x106) -4-2024 Wavelength (nm) 300350400450500550 Molar Ellipticity (x104) -4-2024 C B 250300350400450500550600 Absorbance 0.00.20.40.60.81.01.21.4 320355390425460495530 0.000.050.100.150.200.250.30 AWavelength (nm) Figure 58. Comparison of the UV/Visible Spectroscopic Properties of Wild-Type Cytochrome b 5 Reductase and the G75S, V252M, and G75S/V252M Variants. Oxidized samples of G75S, V252M, G75S/V252M, and wild-type cytochrome b 5 reductase (10-60 M FAD) were analyzed using ultra violet/visible absorbance (A) and circular dichrosim (B and C) spectroscopies, respectively, in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0. Individual spectra correspond to G75S ( ____ ____ ), V252M ( __ __ __ ), G75S/V252M ( ), and wild-type cytochrome b 5 reductase ( ______ ). The insert shows expanded portions of the visible regions of the spectra. 164

PAGE 183

165 Table 19. Comparison of the Kinetic Constants Obtained for the G75S, V252M, G75S/V252M and Wild-Type Cytochrome b 5 Reductases. Mutant NADH:FR NADH:BR k cat K m NADH K m FeCN6 k cat / K m NADH (s -1 ) (M) (M) (s -1 M -1 ) k cat K m NADH K m Cyt b 5 k cat / K m Cyt b 5 (s -1 ) (M) (M) (s -1 M -1 ) H 4 c b 5 r G75S V252M G75S/V252M 800 21 6 1 7 1 1.4 0.3 x 10 8 133 26 9 1 7 1 1.5 0.1 x 10 7 633 14 52 5 7 1 1.2 0.2 x 10 7 187 17 17 1 8 1 1.1 0.1 x 10 7 600 10 2 1 13 1 4.7 0.5 x 10 7 60 3 4 1 14 3 1.5 0.1 x 10 7 285 14 3 1 13 2 1.2 0.2 x 10 7 85 3 3 1 14 3 1.1 0.1 x 10 7 constants of all three mutants were an orde r of magnitude lower than the corresponding value obtained for the wild-type enzyme, reflect ing the significant effect of the mutations of the catalytic functionality of the enzyme. The modest increase in the K m for NADH observed for the G75S variant was confirmed upon the measurement of the spectral binding constants ( K s ) [77] for both H 4 NAD and NAD + In contrast to the results previously obtained for the wild-type domain which have yielded K s values of 45 and 760 M for H 4 NAD and NAD + respectively, identical experiments perf ormed with the G75S variant, yielded K s values of 59 M and 85 M, respectively, confirming the modest decrease in affinity for the reduced form of the pyridin e nucleotide (Figure 59). However, the significantly decreased K s value obtained for NAD + indicated that the mutant exhibited a substantially greater affinity for the reaction product, whic h would be expected to result in decreased enzyme activity owing to a decreased rate of product disassociation. In contrast, values obtained for both H 4 NAD and NAD + were significantly elevated for the V252M mutant, confirming the results obtained for the activity assays.

PAGE 184

166 To assess the impact of the G75S and V252M substitutions on protein folding, thermal stability studies were performed monitoring both the loss of NADH:FR activity and the increase in intrinsic flavin fluores cence for both the singl e and double mutants. Both parameters are indicative of relative prot ein unfolding and can be effectively used to monitor alterations in protei n stability. As shown in Figure 60, the G75S, V252M and G75S/V252M variants exhibited T 50 values of approximately 48 C, 53 C, and 45 C, respectively, which may be compared to the value of 57 C the wild-type domain, suggesting that both the glycine to serine and valine to methionine substitutions resulted in significant changes in the stability of th e mutant proteins when compared to native cytochrome b 5 reductase. Oxidation-reduction potential measurements for the FAD/FADH 2 couple (n = 2), were determined using th e xanthine/xanthine oxidase dye equilibration method utilizing phenosafranine ( E = -252 mV) as the indicator dye. While we were unable to determine the redox pot ential for the flavin prosthetic group in either the G75S or the G75S/V252M variants owing to substantial protein instability and aggregation during the titration experiments, a value for the V252M mutant was obtained reflecting the stability of this variant. The spectral changes accompanying reduction of the V252M variant during the course of the dye mediated titration were directly comparable to those previously published for the wild-type cytochrome b 5 reductase suggesting very similar flavin redox potenti als for the two proteins. The corresponding Nernst plots, shown in Figure 61, which relate to the concen trations of the oxidized and reduced forms of the flavin cofactor to the ob served solution potential [77], revealed that the flavin midpoint potential for the FAD/FADH 2 couple ( n = 2) in the V252M variant was -276 mV, which was essentially unchanged from the midpoint potential of -274 mV

PAGE 185

167 Figure 59. Flavin Difference Spectra Obtained Following Binding of Either H 4 NAD or NAD + to Wild-Type Cytochrome b 5 Reductase or the Various G75S and V252M Mutants. Difference Spectra were obtained for both wild-type cytochrome b 5 reductase (A and B) and the G75S (C and D), V252M (E and F), and G75S/V252M (G and H) variants at equivalent flav in concentrations (50 M FAD) in 20 mM MOPS buffer, containing 0.1 mM EDTA, pH 7.0 fo llowing titrations with either H 4 NAD (left panels) or NAD + (right panels) as described in Materials and Methods. The insert panels correspond to plots of the magnitudes of th e observed spectral pert urbations (peak to trough measurements at the indicated wavelengths) versus ligand co ncentration. The corresponding K s values are given in Table 20.

PAGE 186

-0.06 0.00 0.06 0.12 [H4NAD] (mM) 0.00.306 Absorbance (A 410 -A 460 ) 000 005 0.10 -0.03 0.00 0.03 0.06 0.09 [H4NAD] (mM) 0.00.20.4 Absorbance (A 395 -A 480 ) 0.000 0.015 0.030 Absorbance -0.02 0.00 0.02 0.04 [H4NAD] (mM) 0.00.71.4 Absorbance (A 395 -A 480 ) 0.000 0.006 0.012 380435490545600 -0.06 0.00 0.06 0.12 [H4NAD] (mM) 0.00.612 000 004 008 -0.03 0.00 0.03 0.06 0.09 [NAD+] (mM) 048 0.000 0.017 0.034 -0.08 0.00 0.08 0.16 0.24 [NAD+] (mM) 00061.2 0.00 0.09 0.18 -0.02 0.00 0.02 0.04 0.06 [NAD+] (mM) 036 Absorbance (A 410 -A 460 ) 0.00 0.01 0.02 Wavelength (nm) 380435490545600 -0.08 0.00 0.08 0.16 0.24 [NAD+] (mM) 024 0.00 0.09 0.18 Absorbance (A 410 -A 460 ) Absorbance (A 410 -A 460 ) Absorbance (A 410 -A 460 ) Absorbance (A 410 -A 460 ) A B C D E F G H 168

PAGE 187

169 Table 20. Spectral Binding Constants Obtained for Wild Type c b 5 r and the G75S, V252M, G75S/V252M Variants. Mutant K s H4NAD (M) K s NAD+ (M) H 4 c b 5 r 45 10 760 30 G75S 59 4 85 6 V252M 428 55 14,001 2251 G75S/V252M 134 7 956 83 obtained for the wild-type cytochrome b 5 reductase. However, in the presence of NAD + or NADP + the redox potential of the FAD/FADH 2 couple was shifted to -232 mV and 190 mV, respectively, which reflected and c onfirmed the decreased affinity of the V252M mutant for NAD + compared to the native enzyme. Summary of the novel FAD-binding lobe G75S mutation in cytochrome b 5 reductase The G75S and V252M mutation identified in exons 3 and 9 re spectively represent radical amino acid substitutions in distinctly different re gions of the pol ypeptide chain that are primarily responsible for eith er FAD or NADH binding, respectively. Identification of the G75S variant represen ts the discovery of a new mutation in the DIA1 gene, the V252M variant has been previously reported [82-84] in both type I and type II individuals, but biophysical st udies on a recombinant form of the V252M variant have not been previously described. To understand the effect of these mutations on the function of cytochrome b 5 reductase and the development of recessive congenital methemoglobinemia, we generated and characte rized the G75S variant together with the V252M and the G75S/V252M double mu tant using a rat cytochrome b 5 reductase

PAGE 188

Temperature (oC) 0102030405060708090100 % Residual Activity (NADH:FR) 0255075100% Fluorescence Intensity 0255075100 Figure 60. Comparison of the Thermostability Properties of Wild-Type Cytochrome b 5 Reductase and the G75S, V252M, G75S/V252M Variants. Oxidized samples of G75S, V252M, G75S/V252M and wild-type cytochrome b 5 reductase (5 M FAD) were incubated at the indicated temperatures, and aliquots were withdrawn and assayed for both residual NADH:FR activity (open symbols) and intrinsic flavin fluorescence (closed symbols) in 10 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0 using excitation and emission wavelengths of 450 nm and 523 nm, respectively. The plots correspond to (,) H 4 cb 5 r; (, ) G75S; (, ) V252M; and (, ) G75S/V252M. 170

PAGE 189

Fig. 61. Comparison of the thermodynamic properties of wild-type cytochrome b 5 reductase and the V252M variant. Reductive dye-equilibration titrations of both wild-type and the various of cytochrome b 5 reductase mutants (40 M FAD) were performed as described under Materials and Methods in 100 mM phosphate buffer, containing 0.1 mM EDTA, pH 7.0 in the presence of phenosafranine (15 M, E o = 252 mV) as indicator dye. Individual absorption spectra were collected at 2to 3-min intervals during the time course of the titration and the corresponding Nernst plot obtained for the FAD/FADH 2 couple (n = 2) is shown for the titration of the V252M variant (O, E o = 276 mV). Also shown is the corresponding plot for wild-type cytochrome b 5 reductase ( E o = 272 mV). Redox potentials are given with respect to the Standard Hydrogen Electrode and are considered accurate to mV. 171

PAGE 190

172 heterologous expression system The double mutant does not represent the situation in the patient, but we were curious to examine the impact of both mu tations on the protein structure of cytochrome b 5 reductase. Although the G75S mutation is located in the FAD-binding lobe, it did not affect the global folding of the protein or the incorporation of the flavin cofactor. Instead the G75S mutation caused reduced protein stability and impaired NADH activity, thus resulting in type I recessive congenital methemoglobinemia. The G75S variant is the fourth example of an amino acid substitution that results in a mutation (GGC to AGC) in exon 3, wh ich codes for a porti on, comprising residues I52 to G75, of the amino-terminal FAD-binding lobe of cytochrome b 5 reductase. Other mutations previously identified within th e FAD-binding lobe include R58Q, P65L and L72P [31, 48, 82, 118]. A ll three mutations in exon 3 give rise to the type I form of RCM which is the more benign form of the disease. The x-ray structure of the sol uble domain of rat cytochrome b 5 reductase [21], both in the absence and presence of NAD + indicated that residue G75 is situated at the apex of a broad loop structure that precedes the F 3 strand which comprises part of a FAD-binding six strand ed, anti-parallel -barrel structure. Inspection of the rat cytochrome b 5 reductase structure revealed that while G75 makes no direct contacts with either the FAD prosthetic group or the bound NAD + nucleotide, the residue is situated at a relatively solvent-accessible cleft within th e structure and is contacted by a number of bound water molecules. A multiple sequence alignment, shown in Figure 62, has revealed that G75 is part of a highly conserved region of the amino acid sequence that corresponded to residues 70 LGLPxGxH 77 that is present in the majority of cytochrome

PAGE 191

173 b 5 reductase sequences identified so far and also includes H 77 which provides hydrophobic interactions with the flavin isoall oxazine ring. In a ddition, modeling studies of the interaction of human cytochrome b 5 reductase with cytochrome b 5 have suggested that the conserved histidin e 77 is hydrogen bonded to the b 5 propionate group in the cytochrome b 5 reductasecytochrome b 5 complex [20]. Homology modeling of the G75S variant suggested that substitution of the more bulky serine side chain for the glycine hydroge n atom would be expected to result in steric constraints which would require displacement of the N 7 strand that forms the initial part of the linker or hinge region that joins the FADand NADH-binding lobes. Perturbation of the or ganization of this linker region would be anticipated to result in decreased diaphorase activity owing to a change in registration between the two structural lobes as previously reported for other linker domain RCM mutants, such as P144L [82], P144S [119], and L148P [48]. In addition, alterations in the organization of the two lobes would be expected to decrease the stability of the protein which has also been previously documented for the P144L and L148P variants [105]. The relative importance of a glycine residue at position 75 in the primary structure is also reinforced by its conservation in the majority of the other protein sequences that form part of the ferredoxin:NADP + reductase (FNR) super-family of flavoprotein transhydrogenases, including FNR, phthalate dioxygenase reductase and assimilatory NADH:nitrate reductase [5]. In contrast to the G75S mutation, the V252M substitution was determined to have a markedly less severe impact on either th e structure or function of cytochrome b 5 reductase. Unlike G75, residue V252 is not s ituated in a region of particularly well

PAGE 192

174 conserved primary sequence and can be found to occur as valine, isoleucine or proline residues in cytochrome b 5 reductase homologues, although the marked preference appears to be for the presence of an alipathi c hydrocarbon side chain at this position in this sequence, as shown in Figure 62. Within the rat cytochrome b 5 reductase structure, V 252 is part of a surfaceaccessible loop region located midway between strand N 4 and helix N 3 [21]. While V252 makes no direct backbone or side chain contacts with either the FADor bound NAD + this loop region forms part of the NAD + binding site and aids in correctly orienting the side chain of F251 over the adenin e moiety of the substr ate. Carbon CG2 of the isopropyl side chain of V252 is only 5.3 away from the adenine O 2 atom of the NAD + In addition, the V252 amide nitrogen at om participates in a water-mediated hydrogen bond to the NAD + AO 2 atom. Thus, substitution of V252 by the considerably more bulky side-chain of methionine would be unlikely to have any dramatic effects on either the properties of the FAD prosthetic grou p or the specific activ ity of the protein but could decrease the affinity for NADH due to potential steric interactions which would disrupt the structure of the loop region and decrease the in teraction with the adenine moiety. This prediction is in good agreem ent with the results obtained from the NADH:FR activity studies which indicated a nearly 10-fold increase in the K m for NADH. Previously, we have id entified other mutations in the carboxy-terminal NADH-binding lobe of cytochrome b 5 reductase, including D239G, E255-, and G291D [78, 105, 107], whose biochemical properties can be compared with those of the V252M variant. The V252M variant was found to reta in the spectroscopic pr operties typical of

PAGE 193

175 Fig. 62. Multiple Sequence Alignm ent of Selected Cytochrome b 5 Reductase Primary Structures. Published amino acid sequences for cytochrome b 5 reductases from various species were obtained from GenBank using the corresponding accession numbers, aligned using the CLUSTAL X al gorithm [108] and the alignment manually adjusted for maximum sequence conservation. Only portions of the sequences surrounding the residues equiva lent to G75 and V252 in H. sapiens cytochrome b 5 reductase are shown with the conserved G and V/I residues underlined. Superscripts indicate the positions of the starting residu es for the two sequence portions within the respective primary sequence. Shown below the alignment is a sequence logo [109] that consists of stacks of symbols (one stack for each position in the sequence) with the overall height of the stack indicating the sequence conservation at that position and the height of the symbols within the stack indicating the relative frequency of each amino acid at that position.

PAGE 194

176

PAGE 195

177 the wild-type protein and to exhibit a full complement of FAD incorporation with the flavin prosthetic group pres ent in an environment spectr oscopically indistinguishable from that of the native domain. This was similar to the spectroscopic properties and FAD complement of the other NADH-binding lobe variants [105, 107]. In contrast, NADH:FR activity assays reveal ed the V252M mutant retain ed substantial diaphorase activity, as indicated by a specific activity correspondin g to 80% of the wild-type domain, similar to D239G at 94%, while th e E255and G291D variants only retained 38% and 58% of wild-typ e activity, respectively. There was reduced affinity towards the NADH substrate but no change in cofactor selectivity as with D239G [107 ] and no significant change in the thermostability of the V252M variant when compared to wild-type cytochrome b 5 reductase. Thus, loss of the non-polar side chain at residue 252 does not affect thermostab ility or enzyme activity but reduces affinity towards the NADH substrate, indicating the key role that this region of the primary structure plays in substrate affinity. It is interest ing to note that a change in the K m for NADH alone is sufficient to cause the development of type I RCM. Since the NADH cofactor is required for th e reduction of soluble cytochrome b 5 and in turn methemoglobin reduction. In summary, characterization of the i ndividual G75S and V252M cytochrome b 5 reductase variants together with the G 75S/V252M double mutant indicated decreased catalytic activity rather than total loss of enzyme function as predicted by the suggested type I recessive congenital methemoglobinemia phenotype of the infant In addition, the results for the G75S/V252M double mutant were intermediate between those of the two individual variants suggesting that the effects of the two individual mutations were not

PAGE 196

178 synergistic. All three cytochrome b 5 reductase variants displayed an approximately 10fold decrease in catalytic efficiency combin ed with decreased protein stability, albeit slight for V252M, which would be expected to manifest as potentially increased methemoglobin concentrations. In conclusi on, these studies provi de insight into the development of type I as opposed to type II di sease and may be a poten tial diagnostic tool in young infants to aid in defining recessive congenital methemoglobin status.

PAGE 197

179 4. CONCLUSIONS AND FUTURE DIRECTIONS The studies presented here were intende d to provide additional structural and functional insight into th e highly conserved motifs found within cytochrome b 5 reductase, and other members of the FNR superfam ily, in respect to FAD-binding and NADHbinding as well as their catalyt ic function. This was acco mplished through two alternate approaches. The first approach involved the generation of variants of the amino acid residues found within the cons erved motifs based on alterna tive residues observed to occur at the same position in other members of the FNR family as well as within other species, charge reversal mutants, and the introduction of a positive charge as in the case of the G124 variants. The second approach incorporated the analyses of naturally occurring recessive co ngenital methemoglobinemia mutants that are located within, or in close proximity to, the conserved motifs. Through this method of characterization we were able to establish a role for each of the previously uncha racterized amino acid residues found in the FADand NADH-bindi ng motifs as well as determine the molecular basis for disease for each of the RCM variants. Previous studies based on the RxY S T xx S N motif had established it as being highly conserved within the -barrel structure in the FNR su perfamily of enzymes [39]. Work by Kirksey et al. determined that in monoamine oxi dase B the equivalent arginine residue beginning the motif possessed a necessa ry charge for proper incorporation of the

PAGE 198

180 FADcofactor [120]. More recently, a role for the R9 1, P92, and Y93 has been established [76, 77]. It was de monstrated that these residues were not in f act essential for flavin incorporation, yet R91 and Y93 were both requir ed for the modulation of the biophysical and catalytic pr operties of cytochrome b 5 reductase as well as maintaining the proper active site structure. The work presen ted here determined a role for the remainder of the motif to include: T94, P95, V96, and S97. While the V96S and S97N variants yielded results that were comparable to wild-type cytochrome b 5 reductase, indicating that these residues were not essential for prop er flavin incorporation or catalysis, the introduction of a T94H varian t involved the rem oval of the polarity of a hydroxyl group and caused a significant perturbation with re spect to the environment of the flavin prosthetic group. The introduction of the im idizole side chain resulted in a steric hindrance effect directly affecting the interact ion of the T94 residue with that of the FAD isoalloxazine ring and displaci ng it away from the FAD binding site. This was further confirmed in the analyses of the spectroscop ic binding constants, where the T94H variant displayed a greater affinity for the product NAD + but a reduced affinity for the H 4 NAD substrate analog. Inhibition studies conc luded that upon substrate binding the NAD + product was not able to disa ssociate efficiently from the NADH-binding site which may be due to the displacement of the flavin. Similar, yet not as dramatic, effects were observed for the P95I and P95G variants. Inte restingly, we were never able to generate a stable protein for either the T94G or T94P varian ts. We believe that this is due in part to the fact that the introduction of a glycine leads to a relaxa tion in the secondary structure of the protein and the introducti on of a proline causes a kink in the backbone structure. Due to its highly conserved position these dramatic alteration to the T94 residue would

PAGE 199

181 lead to the generation of an unstable form of the protein that was unable to properly incorporate the flavin prosthetic group. Further analysis of the RxY S T xx S N motif was accomplished through the characterization of the na turally occurring RCM double mutant P92H/E255-. The examination of this variant not only afforded an opportunity to determine the properties of a RCM mutant occurring within a highly conserved motif but also allowed us to investigate the effects of a mutation f ound in both the FAD-binding and NADH-binding domains of cytochrome b 5 reductase. The substitution of a proline by a histidine at residue position 92 leads to the introduction of an imidizole side ch ain that would result in a steric hindrance with the isoalloxazine ring of the flavin prosthetic group. The results of this steric hindrance lead to a signi ficant decrease in the efficiency of substrate utilization as well as a dramatic perturbati on of the environment of the flavin causing a partial displacement of the co factor as observed in the pot entiometric titration with a positive shift towards that of free flavin. The E255variant has been previously characterized and has been demonstrated to have a profound effect on the NADH-binding motif CGxxxM that results in a decrease in substrate a ffinity and a reduction of the catalytic efficiency of over 50% [79]. T hus, the combination of the P92H and E255variants in the double mutant combined th e effects of the P92H on the FAD-binding domain and the E255on the NADHbinding domain resulting in a 10 3 decrease in the overall catalytic efficiency. The analysis of the crystal structure of cytochrome b 5 reductase exposed the critical importance of the G R xx S T motif which has been proposed to be involved in the regulation of FAD or FMN specifi city [5]. Each of the four residues found within this

PAGE 200

182 motif are shown to be direc tly involved in proper FAD-bi nding and orientation through hydrophobic or electrostatic inte ractions. Additionally, the residues of this motif are located within the only -helical segment found within the FAD-binding lobe of cytochrome b 5 reductase. Therefore, any mutations of the residues of this motif would have a significant impact on the environment of the flavin prosthetic group. This was previously demonstrated through the characte rization of the type II RCM variant S127P [80], in which the ADP moiety of the flav in prosthetic group was displaced into the NADH-binding domain, thereby acti ng as its own inhibitor. This work represented the first application of site-directed mutage nesis to the G124, K125, and M126 residues found within the cytochrome b 5 reductase Gxx S T motif to gain insight into the roles of each of these residues in regulating flav in binding and orientation. Significant perturbations were first observed in the visibl e CD spectra for each of the variants. The most dramatic effect was determined for the M126 variants which displayed spectra comparable to that of the previously descri bed S127P variant, indicating that, much like the S127P mutation, alterations of the M126 resi due also lead to the displacement of the ADP moiety of the FADcofactor. The pe rturbations were also found to effect a decrease in the overall catalytic efficiency for all of the variants in regards to both turnover and substrate binding. Interestingly, in establishing the differential binding constants for the Gxx S T variants, the introduction of a positive charge at position 124 displayed an enhanced affinity for both H 4 NAD and NAD + yet showed an altered conformation of binding for the NAD + product. This indicated, th at much like that of the T94H variant, the substrate is not able to bind as efficiently yet is not able to disassociate due to the perturbation of the flavin prosthetic group caused by the introduction of a

PAGE 201

183 positive charge on residue G124. Finally, evid ence suggests that the disruption of the flavin environment leading to a displacement of the FADcofactor were supported by the shifts observed in the values of the midpoint potentials obtained in the presence of NAD + which all displayed a negative shift towards that of the potential of free flavin. The results from the kinetic, thermodynamic, and spectral data clearly demonstrate that any alterations introduced into the Gxx S T motif of cytochrome b 5 reductase lead to a partial displacement of the flavin pros thetic group resulting in a reduc tion in the overall catalytic efficiency and electron transfer. The characterization of the naturally-o ccurring RCM variants M272and P275L allowed for the determination of the im portance of the conserved NADH-binding motif CGxxxM motif as well as establishing a molecular basis for disease for each of the variants. The M272 residue ha s no direct interaction wi th the NADH substrate yet it directly precedes the highly conserved residue C273, which has been determined to be an essential residue within the CGxxxM motif The deletion of re sidue M272 has also been shown to give rise to type II RCM [81] and initial kinetic analyses has demonstrated that this mutation has the lowest cataly tic efficiency observed thus far for any methemoglobinemia variant [40]. The deletion of this residue results in a frame-shift of the entire nucleotide binding motif which disp layed a significant decrease in the substrate binding affinity as well as a reduction in the ca talytic efficiency. As demonstrated in the determination of the spectral binding constants, no spectr oscopic changes were observed for the M272variant even at very high concen trations of substrate. Interestingly, the same effects were witnessed for the M272R mutation. The introduction of the arginine side chain was hypothesized to protrude towards the C273 residue resulting in a

PAGE 202

184 disruption of the essential contacts that C 273 makes with the NADH substrate leading to the decreased substrate binding affinity. The effects upon substrate affi nity as well as the modulation of electron transfer by the M272and M272R variants was observed in the potentiometric titrations performed in the presence of NAD + For the wild-type cytochrome b 5 reductase a shift of +77 mV was observed for the FAD/FADH 2 couple in the presence of NAD + however, this shift was significan tly decreased for the M272and M272R variants, further confirming that substrate binding ha d been severely compromised. Residue P275 is a critical residue found within the CGxxxM motif that functions in proper NADH-binding and orienta tion of the pyridine nucleotide. Through analysis of the crystal structure of cytochrome b 5 reductase in complex with NAD + it was observed that the bound NAD + conforms itself and actually bends around the P275 residue. The mutation of the proline to that of a leucine, as determined in a patient exhibiting a type I RCM phenotype, demonstrated the critical role that this amino acid residue plays. The initial rate kinetic analys es provided the first evidence of the effects of the introduction of a leucine at position 275. Surprisingly, the P275L variant displayed an activity that was comparable to that of wild-type cytochrome b 5 reductase, however, the K m for NADH utilization was increased 437 fold compared to that of wild-type c b 5 r, indicating a dramatic decrease in the affinity for the NADH substrate. To further confirm the reduced affinity for NADH, spectral binding constants were determined for both the analog H 4 NAD and the reaction product NAD + In this experiment, no complex formation could be detected even upon the add ition of concentrations of substrate that were 10-fold greater than that required for saturation of the wild-type enzyme.

PAGE 203

185 Additionally in the potentiometr ic titrations, the redox poten tial in the absence of NAD + was comparable to the value obtained for wild-type cytochrome b 5 reductase, indicating that the P275L variant had no effect on the FADcofactor. However, upon the addition of NAD + the potential was significantly more negative, confirming that the replacement of the proline residue with a leucine does in fact lead to a decreased affinity for the NADH substrate that is capable of giving rise to type I RCM. It is evident from these results that any alterations in the highly conserved CGxxxM motif can significantly affect the overall catalytic efficiency of cytochrome b 5 reductase and the correct maintenance of the motif is imperati ve in proper NADH substrate binding and orientation. Finally, the characterization of th e naturally occurring double mutant G75S/V252M again allowed us to examine the effects of a RCM mutation that occurs in both the FAD-binding and NAD H-binding domains and use this information as a potential aid to determine the methemoglobine mia status (type I or type II) upon the first onset of the disease. Although the patient was actually hetero zygous for the double mutation, we decided to generate the double mu tant as a homozygote in order to examine the possible synergistic effects. Previous studies have described the V252M as giving rise to type I RCM [40, 84]. It was then our prediction that the G75S variant would also be capable of demonstrating characteristics of that of type I RCM, yet the combination of the two mutations affecting both the NADHbinding and FAD-binding domains would give rise to the more severe type II RCM. However, al though both the single and the double mutants displayed a decrease in catalytic efficiency, the double mutant actually yielded results that were intermediate betw een that of the two single mutants. This

PAGE 204

186 information can therefore provide valuable insight towards the early diagnosis of RCM type I versus type II status in an infant. This data provides valuable insight into th e structural and func tional properties of the highly conserved motifs of cytochrome b 5 reductase that are involved in FADor NADH-binding and orientation as well as a determination of the molecular basis of the disease recessive congenital methemoglobi nemia. These results support logical conclusions based on the kinetic, spectr al, thermodynamic, and molecular modeling properties of each of the generate d variants. However, in orde r to complete this work and support our conclusions, crystal structures of the variants, both in the absence and presence of NAD + should be obtained. This would prov ide the most concre te evidence to support the conclusions presented in this wor k. The evaluation of the naturally occurring methemoglobinemia variants M272and P275L provided us with a clear definition of the importance of the CGxxxM motif in re gards to NADH substrate binding and orientation. In order to fully elucidate the role of each residue though, a project is already underway that will involve alan ine scanning of the entire motif The data presented here in this research provided a broad analysis of the motif whereas the alanine scanning data will provide information on the st ructural and catalytic role of each individual residue. Additional work could also be carried out involving the G R xx S T motif. The generation of the G124, K125, and M126 variants established the critic al role that each residue plays towards the pr oper incorporation of the c b 5 r flavin prosthetic group. To gain a broader understanding of the roles of these residues, a series of double and triple mutants could be generated and only the most interesting variants selected for further characterization. As previously mentioned, the G R 124 residue not only forms a

PAGE 205

187 hydrophobic interaction with th e FADcofactor of c b 5 r but also has been proposed to regulate FAD/FMN specificity within members of the FNR superfamily. To investigate this, flavin analog substitution could be performed utilizing the G124R variant. From this a K s for FAD, FMN, and various other analogs could be determined.

PAGE 206

188 REFERENCES 1. Hyde, G.E., N.M. Crawford, and W.H. Campbell, The sequence of squash NADH:nitrate reductase and its relationship to the sequences of other flavoprotein oxidoreductases. A family of flavoprotein pyridine nucleotide cytochrome reductases. J Biol Chem, 1991. 266(35): p. 23542-7. 2. Karplus, P.A. and C.M. Bruns, Structure-function rela tions for ferredoxin reductase. J Bioenerg Biomembr, 1994. 26(1): p. 89-99. 3. Campbell, W.H. and K.R. Kinghorn, Functional domains of assimilatory nitrate reductases and nitrite reductases. Trends Biochem Sci, 1990. 15(8): p. 315-9. 4. Wang, M., et al., Three-dimensional structure of NADPH-cytochrome P450 reductase: prototype for FM Nand FAD-containing enzymes. Proc Natl Acad Sci U S A, 1997. 94(16): p. 8411-6. 5. Correll, C.C., et al., Structural prototypes for an extended family of flavoprotein reductases: comparison of phthalate dioxygenase reductas e with ferredoxin reductase and ferredoxin. Protein Sci, 1993. 2(12): p. 2112-33. 6. Lu, G., et al., Crystal structure of the FAD-containing fragment of corn nitrate reductase at 2.5 A resolution: relations hip to other flavoprotein reductases. Structure, 1994. 2(9): p. 809-21. 7. Bull, P.C., et al., Cloning and chromosomal mapping of human cytochrome b5 reductase (DIA1). Ann Hum Genet, 1988. 52(Pt 4): p. 263-8. 8. Pietrini, G., P. Carrera, and N. Borgese, Two transcripts encode rat cytochrome b5 reductase. Proc Natl Acad Sci U S A, 1988. 85 (19): p. 7246-50. 9. Leroux, A., et al., Generalised deficiency of cytochrome b5 reductase in congenital methaemoglobinaemia with mental retardation. Nature, 1975. 258(5536): p. 619-20. 10. Kobayashi, Y., et al., Serine-proline replacemen t at residue 127 of NADHcytochrome b5 reductase causes heredi tary methemoglobinemia, generalized type. Blood, 1990. 75(7): p. 1408-13. 11. Tomatsu, S., et al., The organization and the complete nucleotide sequence of the human NADH-cytochrome b5 reductase gene. Gene, 1989. 80(2): p. 353-61. 12. Ozols, J., S.A. Carr, and P. Strittmatter, Identification of the NH2-terminal blocking group of NADH-cytochrome b5 reductase as myristic acid and the complete amino acid sequence of the membrane-binding domain. J Biol Chem, 1984. 259(21): p. 13349-54. 13. Spatz, L. and P. Strittmatter, A form of cytochrome b5 t hat contains an additional hydrophobic sequence of 40 amino acid residues. Proc Natl Acad Sci U S A, 1971. 68(5): p. 1042-6.

PAGE 207

189 14. Strittmatter, P., M.J. Rogers, and L. Spatz, The binding of cytochrome b 5 to liver microsomes. J Biol Chem, 1972. 247 (22): p. 7188-94. 15. Reddy, V.V., D. Kupf er, and E. Caspi, Mechanism of C-5 double bond introduction in the biosynthesis of c holesterol by rat liver microsomes. J Biol Chem, 1977. 252(9): p. 2797-801. 16. Strittmatter, P., et al., Purification and properties of rat liver microsomal stearyl coenzyme A desaturase. Proc Natl Acad Sci U S A, 1974. 71(11): p. 4565-9. 17. Keyes, S.R. and D.L. Cinti, Biochemical properties of cytochrome b5-dependent microsomal fatty acid elongati on and identification of products. J Biol Chem, 1980. 255(23): p. 11357-64. 18. Hildebrandt, A. and R.W. Estabrook, Evidence for the participation of cytochrome b 5 in hepatic microsomal mixed-function oxidation reactions. Arch Biochem Biophys, 1971. 143(1): p. 66-79. 19. Kurian, J.R., et al., Reductive detoxification of aryl hydroxylamine carcinogens by human NADH cytochrome b5 re ductase and cytochrome b5. Chem Res Toxicol, 2006. 19(10): p. 1366-73. 20. Bando, S., et al., Structure of human erythrocyt e NADH-cytochrome b5 reductase. Acta Crystallogr D Biol Crystallogr, 2004. 60(Pt 11): p. 1929-34. 21. Bewley, M.C., C.C. Marohnic, and M.J. Barber, The structure and biochemistry of NADH-dependent cytochrome b5 reductase are now consistent. Biochemistry, 2001. 40(45): p. 13574-82. 22. Hultquist, D.E. and P.G. Passon, Catalysis of methaem oglobin reduction by erythrocyte cytochrome B5 and cytochrome B5 reductase. Nat New Biol, 1971. 229(8): p. 252-4. 23. Kuma, F., Properties of methemoglobin reduc tase and kinetic study of methemoglobin reduction. J Biol Chem, 1981. 256 (11): p. 5518-23. 24. Kimura, S., M. Kawamura, and T. Iyanagi, Role of Thr(66) in porcine NADHcytochrome b5 reductase in catalysis and co ntrol of the ratelimiting step in electron transfer. J Biol Chem, 2003. 278(6): p. 3580-9. 25. Strittmatter, P., Direct hydrogen transfer from reduced pyridine nucleotides to microsomal cytochrome b5 reductase. J Biol Chem, 1962. 237 : p. 3250-4. 26. Strittmatter, P., The interaction of reduced pyridine-aldehyde adenine dinucleotide with cytochrome b-5 reductase. J Biol Chem, 1963. 238: p. 2213-9. 27. Strittmatter, P., The reaction sequence in electr on transfer in the reduced nicotinamide adenine dinucleotide-cy tochrome b5 reductase system. J Biol Chem, 1965. 240(11): p. 4481-7. 28. Iyanagi, T., Redox properties of microsomal reduced nicotinamide adenine dinucleotide-cytochrome b5 reductase and cytochrome b5. Biochemistry, 1977. 16(12): p. 2725-30. 29. Iyanagi, T., S. Watanabe, and K.F. Anan, One-electron oxidation-reduction properties of hepatic NADH-cytochrome b5 reductase. Biochemistry, 1984. 23(7): p. 1418-25. 30. Meyer, T.E., et al., Transient kinetics of intrac omplex electron transfer in the human cytochrome b5 reductase-cytochr ome b5 system: NAD+ modulates

PAGE 208

190 protein-protein binding and electron transfer. Arch Biochem Biophys, 1995. 318(2): p. 457-64. 31. Shirabe, K., et al., Enzymatic instability of NADH-cytochrome b5 reductase as a cause of hereditary methemoglobinemia type I (red cell type). J Biol Chem, 1992. 267(28): p. 20416-21. 32. Lederer, F., et al., Two homologous cytochromes b5 in a single cell. Eur J Biochem, 1983. 132(1): p. 95-102. 33. Salemme, F.R., An hypothetical structure for an intermolecular electron transfer complex of cytochromes c and b5. J Mol Biol, 1976. 102(3): p. 563-8. 34. Accelrys. DS Viewer Pro Software suite 2004 [cited; http://www.accelrys.com] 35. Dailey, H.A. and P. Strittmatter, Modification and identifica tion of cytochrome b5 carboxyl groups involved in protein-protein interaction w ith cytochrome b5 reductase. J Biol Chem, 1979. 254(12): p. 5388-96. 36. Dailey, H.A. and P. Strittmatter, Characterization of the interaction of amphipathic cytochrome b5 with stearyl coenzyme A desaturase and NADPH:cytochrome P-450 reductase. J Biol Chem, 1980. 255 (11): p. 5184-9. 37. Loverde, A. and P. Strittmatter, The role of lysyl residu es in the structure and reactivity of cytochrome B5 reductase. J Biol Chem, 1968. 243 (21): p. 5779-87. 38. Strittmatter, P., et al., Characterization of the covalent cross-links of the active sites of amidinated cytochrome b5 and NADH:cytochrome b5 reductase. J Biol Chem, 1990. 265(35): p. 21709-13. 39. Nishida, H., K. Inaka, and K. Miki, Specific arrangement of three amino acid residues for flavin-binding barrel struct ures in NADH-cytochrome b5 reductase and the other flavin-dependent reductases. FEBS Lett, 1995. 361(1): p. 97-100. 40. Davis, C.A., Sytematic Analysis of Recessi ve Congenital Methemoglobinemia: Molecular Basis for Disease 2004: University of South Florida. 41. Rossmann, M.G., D. Moras, and K.W. Olsen, Chemical and biological evolution of nucleotide-binding protein. Nature, 1974. 250(463): p. 194-9. 42. Hitzenberger, K., Autotoxische zyanose (intr aglubulare methamoglobinamie). Wien. Arch. Inn. Med., 1932. 23: p. 85. 43. Gibson, Q.H., The reduction of methaemoglobin by ascorbic acid. Biochem J, 1943. 37(5): p. 615-8. 44. Gibson, Q.H., The reduction of meth aemoglobin in red blood cells and studies on the cause of idiopathi c methaemoglobinaemia. Biochem J, 1948. 42(1): p. 13-23. 45. Worster-Drought, C., J.C. White, and F. Sargent, Familial idiopathic methaemoglobinaemia, associated with mental deficien cy and neurological abnormalities. Br Med J, 1953. 2(4828): p. 114-8. 46. Scott, E.M. and I.V. Griffith, The enzymic defect of hereditary methemoglobinemia: diaphorase. Biochim Biophys Acta, 1959. 34: p. 584-6. 47. Aalfs, C.M., et al., A case of methemoglobinemia type II due to NADHcytochrome b5 reductase deficiency: determination of the molecular basis. Hum Mutat, 2000. 16(1): p. 18-22. 48. Katsube, T., et al., Exonic point mutations in NADH-cytochrome B5 reductase genes of homozygotes for hereditary methemoglobinemia, types I and III: putative

PAGE 209

191 mechanisms of tissue-dependent enzyme deficiency. Am J Hum Genet, 1991. 48(4): p. 799-808. 49. Yubisui, T., et al., Structural role of serine 1 27 in the NADH-binding site of human NADH-cytochrome b5 reductase. J Biol Chem, 1991. 266(1): p. 66-70. 50. Jenkins, M.M. and J.T. Prchal, A novel mutation found in the 3' domain of NADHcytochrome B5 reductase in an African-Am erican family with type I congenital methemoglobinemia. Blood, 1996. 87(7): p. 2993-9. 51. Karplus, P.A., M.J. Dani els, and J.R. Herriott, Atomic structure of ferredoxinNADP+ reductase: prototype for a stru cturally novel flavoenzyme family. Science, 1991. 251(4989): p. 60-6. 52. Benson, D.A., et al., GenBank. Nucleic Acids Res, 1994. 22(17): p. 3441-4. 53. Prosser, I.M. and C.M. Lazarus, Nucleotide sequence of a spinach nitrate reductase cDNA. Plant Mol Biol, 1990. 15(1): p. 187-90. 54. Newman, D.J., J.N. Ihle, and L. Dure, 3rd, Chemical composition of a ferredoxin isolated from cotton. Biochem Biophys Res Commun, 1969. 36(6): p. 947-50. 55. Haniu, M., et al., Structural and functional analysis of NADPH-cytochrome P-450 reductase from human liver: complete sequence of human enzyme and NADPHbinding sites. Biochemistry, 1989. 28(21): p. 8639-45. 56. Nakane, M., et al., Cloned human brain nitric oxid e synthase is highly expressed in skeletal muscle. FEBS Lett, 1993. 316(2): p. 175-80. 57. Leclerc, D., et al., Cloning and mapping of a cDNA for methionine synthase reductase, a flavoprotein defective in patients with homocystinuria. Proc Natl Acad Sci U S A, 1998. 95 (6): p. 3059-64. 58. Ostrowski, J., et al., Characterization of the cysJIH regions of Salmonella typhimurium and Escherichia coli B. DN A sequences of cysI and cysH and a model for the siroheme-Fe4S4 active cente r of sulfite reductase hemoprotein based on amino acid homology w ith spinach nitrite reductase. J Biol Chem, 1989. 264(26): p. 15726-37. 59. Plamann, M.D. and G.V. Stauffer, Characterization of the Escherichia coli gene for serine hydroxymethyltransferase. Gene, 1983. 22(1): p. 9-18. 60. Davis, C.A., et al., Heterologous expression of an endogenous rat cytochrome b(5)/cytochrome b(5) reducta se fusion protein: identific ation of histidines 62 and 85 as the heme axial ligands. Arch Biochem Biophys, 2002. 400(1): p. 63-75. 61. Bosch, R., E. Garcia-Valdes, and E.R. Moore, Genetic characterization and evolutionary implications of a chromo somally encoded naphthalene-degradation upper pathway from Pseudomonas stutzeri AN10. Gene, 1999. 236(1): p. 149-57. 62. Nordlund, I., J. Powlowski, and V. Shingler, Complete nucleotide sequence and polypeptide analysis of multicomponent phenol hydroxylase from Pseudomonas sp. strain CF600. J Bacteriol, 1990. 172(12): p. 6826-33. 63. Welch, R.A., et al., Extensive mosaic structure revealed by the complete genome sequence of uropathogenic Escherichia coli. Proc Natl Acad Sci U S A, 2002. 99(26): p. 17020-4. 64. Jiang, X.M., et al., Structure and sequence of the rfb (O antigen) gene cluster of Salmonella serovar typh imurium (strain LT2). Mol Microbiol, 1991. 5(3): p. 695713.

PAGE 210

192 65. Eaton, R.W., p-Cumate catabolic pathway in Pseudomonas putida Fl: cloning and characterization of DNA carrying the cmt operon. J Bacteriol, 1996. 178(5): p. 1351-62. 66. James, K.D. and P.A. Williams, ntn genes determining the early steps in the divergent catabolism of 4-nitrotoluene and toluene in Pseudomonas sp. strain TW3. J Bacteriol, 1998. 180(8): p. 2043-9. 67. Trimboli, A.J., et al., Thiol modification and site di rected mutagenesis of the flavin domain of spinach NADH:nitrate reductase. Arch Biochem Biophys, 1996. 331(1): p. 117-26. 68. Stainthorpe, A.C., et al., The methane monooxygenase gene cluster of Methylococcus capsulatus (Bath). Gene, 1990. 91(1): p. 27-34. 69. Bundy, B.M., A.L. Campbell, and E.L. Neidle, Similarities between the antABCencoded anthranilate dioxygenas e and the benABC-encoded benzoate dioxygenase of Acinetobacter sp. strain ADP1. J Bacteriol, 1998. 180(17): p. 4466-74. 70. Neidle, E.L., et al., Nucleotide sequences of the Acinetobacter calcoaceticus benABC genes for benzoate 1,2-dioxygenase reveal evolutionary relationships among multicomponent oxygenases. J Bacteriol, 1991. 173(17): p. 5385-95. 71. Haak, B., S. Fetzner, and F. Lingens, Cloning, nucleotide sequence, and expression of the plasmid-encoded genes for the two-component 2-halobenzoate 1,2-dioxygenase from Pseudomonas cepacia 2CBS. J Bacteriol, 1995. 177 (3): p. 667-75. 72. Kalman, S., et al., Comparative genomes of Chlamydia pneumoniae and C. trachomatis. Nat Genet, 1999. 21(4): p. 385-9. 73. Kaneko, T., et al., Complete genome structure of the nitrogen-fixing symbiotic bacterium Mesorhizobium loti. DNA Res, 2000. 7(6): p. 331-8. 74. Ng, W.V., et al., Genome sequence of Halobacterium species NRC-1. Proc Natl Acad Sci U S A, 2000. 97 (22): p. 12176-81. 75. Zhou, N.Y., et al., The alkene monooxygenase from Xanthobacter strain Py2 is closely related to aromatic monoo xygenases and catalyzes aromatic monohydroxylation of benzene, toluene, and phenol. Appl Environ Microbiol, 1999. 65(4): p. 1589-95. 76. Marohnic, C.C. and M.J. Barber, Arginine 91 is not essential for flavin incorporation in hepatic cy tochrome b(5) reductase. Arch Biochem Biophys, 2001. 389(2): p. 223-33. 77. Marohnic, C.C., et al., Cytochrome b5 reductase: role of the si-face residues, proline 92 and tyrosine 93, in structure and catalysis. Biochemistry, 2005. 44(7): p. 2449-61. 78. Percy, M.J., et al., Familial idiopathic methemoglobinemia revisited: original cases reveal 2 novel mutations in NADH-cytochrome b5 reductase. Blood, 2002. 100(10): p. 3447-9. 79. Davis, C.A. and M.J. Barber, Cytochrome b5 oxidoreductase: expression and characterization of the original famili al ideopathic methemoglobinemia mutations E255and G291D. Arch Biochem Biophys, 2004. 425(2): p. 123-32.

PAGE 211

193 80. Bewley, M.C., et al., The structure of the S127P mutant of cytochrome b5 reductase that causes methemoglobinemi a shows the AMP moiety of the flavin occupying the substrate binding site. Biochemistry, 2003. 42(45): p. 13145-51. 81. Vieira, L.M., et al., Four new mutations in the NADH-cytochrome b5 reductase gene from patients with recessive congenital methemoglobinemia type II. Blood, 1995. 85(8): p. 2254-62. 82. Dekker, J., et al., Seven new mutations in the nicotinamide adenine dinucleotide reduced-cytochrome b(5) reductase gene leading to methemoglobinemia type I. Blood, 2001. 97(4): p. 1106-14. 83. Grabowska, D., et al., Compound heterozygosity of two missense mutations in the NADH-cytochrome b5 reductase gene of a Polish patient with type I recessive congenital methaemoglobinaemia. Eur J Haematol, 2003. 70 (6): p. 404-9. 84. Kugler, W., et al., Molecular basis of recessive congenital methemoglobinemia, types I and II: Exon ski pping and three novel missense mutations in the NADHcytochrome b5 reductase (diaphorase 1) gene. Hum Mutat, 2001. 17(4): p. 348. 85. Murataliev, M.B. and R. Feyereisen, Interaction of NADP(H) with oxidized and reduced P450 reductase during catalysis. Studies with nucleotide analogues. Biochemistry, 2000. 39(17): p. 5066-74. 86. Beck von Bodman, S., et al., Synthesis, bacterial expression, and mutagenesis of the gene coding for mammalian cytochrome b5. Proc Natl Acad Sci U S A, 1986. 83(24): p. 9443-7. 87. Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 1970. 227 (5259): p. 680-5. 88. Turchin, A. and J.F. Lawler, Jr., The primer generator: a program that facilitates the selection of oli gonucleotides for site -directed mutagenesis. Biotechniques, 1999. 26(4): p. 672-6. 89. Yuckenberg, P.D., Witney, F., Ge isselsoder, J., and McClary, J., Site-directed in vitro mutagenesis using uracil-containing DNA and phagemid vectors. Directed Mutagenesis: A Pratical Approach, ed. M.J. McPherson. 1991, New York: Oxford University Press. 90. Chen, G.C., and Yang, J.T., Anal. Letter, 1977. 10 : p. 1195-1207. 91. Foust, G.P., et al., An anaerobic titration assembly for spectrophotometric use. Anal Biochem, 1969. 27(3): p. 530-5. 92. Sancho, J. and C. Gomez-Moreno, Interaction of ferredoxin-NADP+ reductase from Anabaena with its substrates. Arch Biochem Biophys, 1991. 288(1): p. 2318. 93. Barber, M.J. and G.B. Quinn, Production of a recombinant hybrid hemoflavoprotein: engineering a f unctional NADH:cytochrome c reductase. Protein Expr Purif, 2001. 23(2): p. 348-58. 94. Massey, V., A simple method for the determ ination of redox potentials. in Flavins and Flavoproteins S.R. B. Curti, and G. Zanetti, Editor. 1991, de Gruyter: Berlin. p. 59-66. 95. Chen, L., et al., Purification and characteriza tion of an NADH-rubredoxin oxidoreductase involved in the utiliza tion of oxygen by Desulfovibrio gigas. Eur J Biochem, 1993. 216(2): p. 443-8.

PAGE 212

194 96. Karplus, P.A. and G.E. Schulz, Refined structure of glut athione reductase at 1.54 A resolution. J Mol Biol, 1987. 195(3): p. 701-29. 97. Nishida, H. and K. Miki, Electrostatic properties deduced from refined structures of NADH-cytochrome b5 reductase and th e other flavin-dependent reductases: pyridine nucleotide-binding and interac tion with an electron-transfer partner. Proteins, 1996. 26(1): p. 32-41. 98. Kimura, S., H. Nishida, and T. Iyanagi, Effects of flavin-binding motif amino acid mutations in the NADH-cytochrome b5 reductase catalytic domain on protein stability and catalysis. J Biochem (Tokyo), 2001. 130(4): p. 481-90. 99. Wallace, A.C., R.A. Laskowski, and J.M. Thornton, LIGPLOT: a program to generate schematic diagrams of protein-ligand interactions. Protein Eng, 1995. 8(2): p. 127-34. 100. Combet, C., et al., Geno3D: automatic comparativ e molecular modelling of protein. Bioinformatics, 2002. 18(1): p. 213-4. 101. Schwede, T., et al., SWISS-MODEL: An automated protein homology-modeling server. Nucleic Acids Res, 2003. 31(13): p. 3381-5. 102. Kyte, J., Mechanism in Protein Chemistry 1995, Garland Publishing, New York. p. 72-90. 103. Manabe, J., et al., Two novel mutations in the reduced nicotinamide adenine dinucleotide (NADH)-cytochrom e b5 reductase gene of a patient with generalized type, hereditary methemoglobinemia. Blood, 1996. 88(8): p. 3208-15. 104. Deeney, J., Murdock, E.T., Rogan, J.J., Familial idopathic methaemoglobinemia. British Medical Journal, 1943. I : p. 721-723. 105. Davis, C.A., L.J. Crowley, and M.J. Barber, Cytochrome b5 reductase: the roles of the recessive congenital methemoglobinemia mutants P144L, L148P, and R159*. Arch Biochem Biophys, 2004. 431(2): p. 233-44. 106. Percy, M.J., et al., Identification and characteriza tion of the novel FAD-binding lobe G75S mutation in cytochrome b(5) re ductase: an aid to determine recessive congenital methemoglobinemia status in an infant. Blood Cells Mol Dis, 2006. 36(1): p. 81-90. 107. Percy, M.J., et al., Recessive congenital methaem oglobinaemia: functional characterization of the novel D239G mutation in the NADH-binding lobe of cytochrome b5 reductase. Br J Haematol, 2005. 129(6): p. 847-53. 108. Thompson, J.D., et al., The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res, 1997. 25(24): p. 4876-82. 109. Crooks, G.E., et al., WebLogo: a sequence logo generator. Genome Res, 2004. 14(6): p. 1188-90. 110. Marohnic, C.C., M.C. Be wley, and M.J. Barber, Engineering and characterization of a NADPH-utilizing cytochrome b5 reductase. Biochemistry, 2003. 42(38): p. 11170-82. 111. Clark, W.M. and H.J. Lowe, Studies on oxidation-reduction. XXIV. Oxidationreduction potentials of fl avin adenine dinucleotide. J Biol Chem, 1956. 221(2): p. 983-92.

PAGE 213

195 112. Strittmatter, P., et al., Characterization of lysyl re sidues of NADH-cytochrome b5 reductase implicated in charge-pairing with active-site carboxyl residues of cytochrome b5 by site-directed mutage nesis of an expression vector for the flavoprotein. J Biol Chem, 1992. 267 (4): p. 2519-23. 113. Shirabe, K., T. Yubisui, and M. Takeshita, Expression of human erythrocyte NADH-cytochrome b5 reductase as an alphathrombin-cleavable fused protein in Escherichia coli. Biochim Biophys Acta, 1989. 1008(2): p. 189-92. 114. Adams, M.D., et al., The genome sequence of Drosophila melanogaster. Science, 2000. 287(5461): p. 2185-95. 115. Davis, C.A. and M.J. Barber, Heterologous expression of enzymopenic methemoglobinemia variants using a novel NADH:cytochrome c reductase fusion protein. Protein Expr Purif, 2003. 30(1): p. 43-54. 116. Maran, J., et al., Heterogeneity of the molecular biology of methemoglobinemia: a study of eight consecutive patients. Haematologica, 2005. 90(5): p. 687-9. 117. Beutler, E., Acetamonophen and G-6-PD deficiency. Acta Haematol, 1984. 72(3): p. 211-2. 118. Wu, Y.S., et al., Identification of a novel poin t mutation (Leu72Pro) in the NADH-cytochrome b5 reductase gene of a patient with hereditary methaemoglobinaemia type I. Br J Haematol, 1998. 102(2): p. 575-7. 119. Percy, M.J., et al., Congenital methaemoglobinaemia Type I in a Turkish infant due to a novel mutation, Pro144Ser, in NADH-cytochrome b5 reductase. Hematol J, 2004. 5(4): p. 367-70. 120. Kirksey, T.J., S.W. Kwan, and C.W. Abell, Arginine-42 and threonine-45 are required for FAD incorporation and ca talytic activity in human monoamine oxidase B. Biochemistry, 1998. 37(35): p. 12360-6.

PAGE 214

196 PUBLICATIONS Roma, G.W., Crowley, L.J., and Barber, M.J. (2006) Arch. Biochem. Biophys. 452(1) 6982. Expression and characte rization of a functi onal canine variant of cytochrome b5 reductase. Percy, M.J., Crowley, L.J., Boudrea ux, J., and Barber, M.J. (2006) Arch. Biochem. Biophys. 447(1) 59-67. Expression of a novel P275L variant of NADH:cytochrome b5 reductase gives functional insight into the conserved motif important for pyridine nucleotide binding. Percy, M.J., Crowley, L.J., Roper, D., Vullia my, T.J., Layton, D.M., and Barber, M.J. (2006) Blood Cells Mol. Dis., 36(1), 81-90. Identification and characterization of the novel FAD-binding lobe G75S mutation in cy tochrome b(5) reductase: an aid to determine recessive congenital methem oglobinemia status in an infant. Roma, G.W., Crowley, L.J., Davis, C.A., and Barber, M.J. (2005) Biochemistry, 44(41) 13467-13476. Mutagenesis of Glycine 179 modulates both catalytic efficiency and reduced pyridine nucleotide specific ity in cytochrome b5 reductase. Percy, M.J., Crowley, L.J., Davis, C.A., McMullin, M.F., Savage, G., Hughes, J., McMahon, C., Quinn, R.J., Smith, O., Barber, M.J., Lappin, T.R. (2005) Brit. J. Haematol, 129(6) 847-853. Recessive congenital methaemoglobinaemia: functional characterization of the novel D239G mu tation in the NADH-binding lobe of cytochrome b5 reductase. Marohnic, C.C., Crowley, L.J., Davis, C. A., Smith, E.T., and Barber, M.J. (2005) Biochemistry, 44(7) 847-853. Cytochrome b5 reductase : role of the si-face residues, proline 92 and tyrosine 93, in structure and catalysis. Davis, C.A., Crowley, L.J., and Barber, M.J., (2004) Arch. Biochem. Biophys., 431(2) 233-244. Cytochrome b5 reductase: the roles of the recessive congenital methemoglobinemia mutants P144L, L148P, and R159*. Crowley, L.J., Salerno, J.C., Roma G.W., and Barber, M.J. (2007) Analysis of the effects of the RCM variant M272a nd generated variants M272A/I/L/R on the CGxxxM motif. ( Submitted to Arch. Biochem. Biophys. )

PAGE 215

197 Crowley, L.J., Percy, M.J., Roper, D., Lappin, T.R.J., Layton, D.M., and Barber, M.J.. (2007) Type I recessive congenital meth emoglobinemia associated with a double mutation, the novel Pro92His and previously described Glu255-, in the cytochrome b 5 reductase gene. ( Submitted to Brit. J. Haematol )

PAGE 216

ABOUT THE AUTHOR Louis Joseph Crowley was born and raised in Boston, Massachusetts. Louis first aspirations were actually in music where he played guitar in several Boston bands and following high school began working and stu dying at Berklee College of Music in Boston, MA. After realizing th e difficulties involved in becoming a famous musician, Louis served four years in the U.S. Army in order to hopefully attend college. Following the military, Louis enrolled at Pikes Peak Community College, Colorado, as a physical therapy major. It was then in a general biolog y course that he reali zed that his interest was in biology and biochemist ry. A move back to Boston brought him to Bridgewater State College where he was given his firs t opportunity to conduct organic chemistry synthesis research under his ch emistry professor Dr. Edward J. Brush and this fueled his interest in scientific resear ch. Following graduation, Loui s then went on to work for Genzyme in Framingham, MA in the protein purification division. Realizing that only possessing a B.S. degree would not allow him to perform the level of research he wanted to do he entered the Biochemistry and Mol ecular Biology program at the University of South Florida and began to work in the lab of Michael J. Barber, D.Phil where he has had a very productive career as a graduate st udent thanks to Dr. Barbers guidance and support. Louis hopes to re-enter the scien tific industry market with the new found knowledge he has gained in the field of protein biochemistry.