<%BANNER%>

Site-Directed Spin Labeling EPR Studies of Conformational Dynamics in the Flap Region of HIV-1 Protease

Permanent Link: http://ufdc.ufl.edu/UFE0022439/00001

Material Information

Title: Site-Directed Spin Labeling EPR Studies of Conformational Dynamics in the Flap Region of HIV-1 Protease
Physical Description: 1 online resource (230 p.)
Language: english
Creator: Galiano, Luis
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: biophysical, conformational, deer, electron, epr, flaps, flexibility, heterogeneity, hiv, label, lineshapes, mutagenesis, paramagnetic, protease, pulsed, resonance, spin
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Human Immunodeficiency Virus Type 1 (HIV-1) protease is an essential component in the processing of viral proteins encoded in the HIV viral genome. To allow full access of the substrate to the active site of the protease, two beta hairpin regions termed 'flaps', have to undergo a large conformational change, and it has been postulated that these flaps move in a segmental motion, with three major conformations (open, semi-open and closed or bound to inhibitor). Site-Directed spin labeling (SDSL) coupled with Electron Paramagnetic Resonance (EPR), is a powerful technique that can be utilized to measure distances and distance distributions associated with conformational changes. Here, we explored the use of SDSL-EPR in order to elucidate the distances and distance distributions accessible to the flap region in the protease. First, we studied the distances and distance distributions in a construct termed PMPR (PentaMutated Protease), which possesses wildtype-like structure and kinetic properties. From the distance distribution profiles obtained by utilizing double electron-electron resonance (DEER) spectroscopy, we established that the motion of the flaps in the absence of inhibitor is continuous, rather than segmental, and that it spans a set of distances between 24 - 48 Angstrom. Upon introduction of the inhibitor, the average distance is narrowed by ~3 Angstrom, which is consistent with the picture of the flaps closed on top of the inhibitor upon active site binding. Second, the drug-resistance problem in HIV-1 protease was investigated. When patients undergo Highly Active Antiretroviral Therapy (HAART), the drug pressure induced in the patient selects those mutant strains that have inhibitor resistance. Our results show that the distance distribution profile for two drug resistant strains, namely V6 and MDR769, did not significantly change for the protease in the presence and absence of inhibitor. We also found that the distance distributions accessible to the spin labels were narrower than in the PMPR strain, which would indicate a reduced flexibility of the flap in these mutant strains. The results obtained from this work will undoubtedly serve as a stepping stone for further research on the fields of HIV-1 protease, HIV protease drug resistance and SDSL-EPR.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Luis Galiano.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Fanucci, Gail E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022439:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022439/00001

Material Information

Title: Site-Directed Spin Labeling EPR Studies of Conformational Dynamics in the Flap Region of HIV-1 Protease
Physical Description: 1 online resource (230 p.)
Language: english
Creator: Galiano, Luis
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: biophysical, conformational, deer, electron, epr, flaps, flexibility, heterogeneity, hiv, label, lineshapes, mutagenesis, paramagnetic, protease, pulsed, resonance, spin
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Human Immunodeficiency Virus Type 1 (HIV-1) protease is an essential component in the processing of viral proteins encoded in the HIV viral genome. To allow full access of the substrate to the active site of the protease, two beta hairpin regions termed 'flaps', have to undergo a large conformational change, and it has been postulated that these flaps move in a segmental motion, with three major conformations (open, semi-open and closed or bound to inhibitor). Site-Directed spin labeling (SDSL) coupled with Electron Paramagnetic Resonance (EPR), is a powerful technique that can be utilized to measure distances and distance distributions associated with conformational changes. Here, we explored the use of SDSL-EPR in order to elucidate the distances and distance distributions accessible to the flap region in the protease. First, we studied the distances and distance distributions in a construct termed PMPR (PentaMutated Protease), which possesses wildtype-like structure and kinetic properties. From the distance distribution profiles obtained by utilizing double electron-electron resonance (DEER) spectroscopy, we established that the motion of the flaps in the absence of inhibitor is continuous, rather than segmental, and that it spans a set of distances between 24 - 48 Angstrom. Upon introduction of the inhibitor, the average distance is narrowed by ~3 Angstrom, which is consistent with the picture of the flaps closed on top of the inhibitor upon active site binding. Second, the drug-resistance problem in HIV-1 protease was investigated. When patients undergo Highly Active Antiretroviral Therapy (HAART), the drug pressure induced in the patient selects those mutant strains that have inhibitor resistance. Our results show that the distance distribution profile for two drug resistant strains, namely V6 and MDR769, did not significantly change for the protease in the presence and absence of inhibitor. We also found that the distance distributions accessible to the spin labels were narrower than in the PMPR strain, which would indicate a reduced flexibility of the flap in these mutant strains. The results obtained from this work will undoubtedly serve as a stepping stone for further research on the fields of HIV-1 protease, HIV protease drug resistance and SDSL-EPR.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Luis Galiano.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Fanucci, Gail E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022439:00001


This item has the following downloads:


Full Text

PAGE 1

1 SITE-DIRECTED SPIN LABELING ELECTRON PARAMAGNETIC RESONANCE STUDIES OF CONFORMATIONAL DYNAMIC S IN THE FLAP REGION OF HIV-1 PROTEASE By LUIS GALIANO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Luis Galiano

PAGE 3

3 To my wife, Susan; and to my parents, Concepcin Lafuente and Luis Galiano

PAGE 4

4 ACKNOWLEDGMENTS First and foremost, I would like to thank my parents, Concepcin Lafuente and Luis Galiano; and my wife, Susan Galiano, for their sacrifices, suppor t, patience and love. Without them, none of this work would have been possible. I want to express my deepest gratitude to my mentor, Dr. Gail E. Fa nucci for her support, allowing me to pursue my research ideas under her guidance. Her support has allowed me to present my work at major international conferen ces, and has given me the opportunity to interact with many scientists of different disciplines. I would also like to thank all the professors throughout my career, specifically Dr. Manuel Yez, my M.S. advisor Dr. Manuel Alcam, w ho encouraged me to pursue a Ph.D. at the University of Florida, as well as Dr. Rodney J. Bartlett for offering me the opportunity to study under his supervision. I would like to thank ou r collaborators in this project, specifically Dr. Marco Bonora and Dr. Peter Fajer for valuable advice and instrume nt time at the NHMFL, Dr. Keith Earle and Dr. Jack Freed for their work on High Field High Fr equency cw-EPR on HIV-1 protease samples, Dr. Carlos Simmerling and Ding Fangyu for Molecu lar Dynamics simulations on spin-labeled HIV-1 protease, and Dr. Ben D unn and Dr. Roxana Coman for the DNA for preliminary studies on HIV-1 protease and their insight on kinetic characterization of aspa rtic proteases. I would also like to thank Mrs. Dawn Zbell-Herrick, a Ph.D student at Dr. David Cafisos group at the University of Virginia, as well as Dr. Ralph We ber (Bruker Biospin) for their help with the DEER experiments. I would like to express my gratitude to my supervisory committee (Dr. Arthur Edison, Dr. Ben Dunn, Dr. John Eyler, Dr. Philip Brucat, Dr. Alex Angerhofer and Dr. Adrian Roitberg) for many valuable discussions and support.

PAGE 5

5 I also wish to express my gratitude to the members of the Fanucci research group, especially Jordan Mathias, Mandy Blackburn, Ch ad Mair and Thomas Frederick for their friendship and patience with my sometimes l ong-winded explanations about obscure topics. I would also like to thank Amer ican Heart Association for a pr edoctoral fellowship for this project; their continuous support offered me the funds to pursue this research. Finally, I am forever in debt to all the scientists that devote their life to understanding the complexity of HIV-1 virus; without their work, th is dissertation would not have been possible. As John of Salisbury said: we are like dwarfs on the shoulders of giants, so that we can see more than they, and things at a greater distan ce, not by virtue of any sharpness of sight on our part, or any physical distincti on, but because we are carried hi gh and raised up by their giant size.

PAGE 6

6 TABLE OF CONTENTS Page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......11 ABSTRACT....................................................................................................................... ............17 CHAPTER 1 INTRODUCTION TO HIV-1................................................................................................19 Introduction................................................................................................................... ..........19 HIV-1 Viral Life Cycle......................................................................................................... ..20 Drug Therapy Against HIV-1 Vi rus: Current Approaches.....................................................21 Viral Polymorphism: Reverse Transcript ase Lacks Proofreading Capabilities.....................22 The Role and Structure of HIV-1 Protease.............................................................................24 Previous Studies Perfor med on HIV-1 Protease.....................................................................25 Introduction................................................................................................................... ..25 The LAI Subtype B Consensus Sequence.......................................................................26 2 INTRODUCTION TO EPR...................................................................................................36 Introduction to Spectroscopy..................................................................................................3 6 The Static Spin Hamiltonian...................................................................................................3 7 Electron Zeeman Interaction...........................................................................................37 Hyperfine Interaction.......................................................................................................38 Transition Selection Rules in EPR Spectroscopy............................................................41 Nuclear Zeeman Interaction............................................................................................42 Nuclear Quadrupole Interaction......................................................................................43 Zero-Field Splitting.........................................................................................................43 Systems With More Than One Interacting Electron Spins.....................................................44 Dipolar Interaction...........................................................................................................4 4 Exchange Interaction.......................................................................................................46 Introduction to Pulsed EPR....................................................................................................4 7 Density Operator Formalism...........................................................................................47 Double Electron-Electron Resonance..............................................................................49 T1 and T2 (Tm) Measurements.........................................................................................52 Spin-lattice relaxation time (T1)...............................................................................52 Spin-spin relaxation time (T2 or Tm)........................................................................55 Pulsed EPR Instrumentation...................................................................................................57 Data Analysis.................................................................................................................. ........58 Spectral Parameters from cw-EPR Lineshapes...............................................................58 Double Electron-Electron Resonance Data Analysis......................................................60

PAGE 7

7 3 MATERIALS AND METHODS...........................................................................................80 Protein Expression............................................................................................................. .....80 Protein Purification........................................................................................................... ......81 Spin Labeling.................................................................................................................. .84 DEER Experiments.........................................................................................................84 Circular Dichroism Experiments.....................................................................................85 cw-EPR Experiments.......................................................................................................86 UV/Vis Substrate Degrad ation and Inhibition................................................................86 Osmolality and Viscosity Experiments...........................................................................87 4 CONTINUOUS-WAVE AND PULSED ELECTRON PARAMAGNETIC RESONANCE STUDIES OF THE SUBTYP E-B LAI CONSENSUS SEQUENCE OF HIV-1 PROTEASE................................................................................................................. 93 Introduction................................................................................................................... ..........93 Stability of Spin Labeled HIV-1 Protease............................................................................100 Effects of the Spin Label on the Struct ure and Activity of HIV-1 Protease.........................103 Effects of Spin-Labeling on th e Structure of the Protease............................................103 Effect of Spin Labeling on the Activity of the Protease................................................106 Osmolality and Viscosity......................................................................................................1 09 5 DOUBLE ELECTRON-ELECTRON RES ONANCE MEASUREMENTS ON THE FLAP REGION OF HIV-1 PROTEASE..............................................................................139 Preliminar y Studies............................................................................................................ ...139 Double Electron-Electron Resonance Experiments on PMPR+D25N+K55(SL)................141 6 PULSED EPR CHARACTERIZATION OF THE DRUG RESISTANT MUTANTS V6 AND MDR769 OF HIV-1 PROTEASE...............................................................................164 Drug Induced Mutations in HIV-1 Protease.........................................................................164 Nomenclature for Drug-Select ed Mutations in HIV-1 PR............................................165 Multidrug Resistant HIV-1 Protease Variants...............................................................166 Multidrug resistant V6...........................................................................................167 Multidrug resistant MDR769.................................................................................169 cw-EPR and Pulsed-EPR Studies of V6 and MDR769 Drug Resistant Strains...................171 7 FUTURE DIRECTIONS......................................................................................................191 Inhibitor and Substrate Binding Studies for D25N Mutants of HIV-1 Protease..................191 DEER Experiments Performe d With HIV-1 Inhibitors........................................................191 Single-Point Mutant Strains..................................................................................................19 2 High Field/High Frequency EPR Studies on the Dynamics of Spin Labels.........................192 Spin Label Dynamics from MD Simulations.......................................................................193

PAGE 8

8 APPENDIX A COMMON DRUG-SELECTED MUTATI ONS IN HIV-1 PROTEASE............................194 B SAMPLE DEER RUN DATA..............................................................................................198 C FDA APPROVED PROTEASE INHIBI TORS FOR HIV-1 TREATMENT......................205 D INTERACTIONS BETWEEN HIVPR AND RITONAVIR...............................................207 E PCR PRIMERS AND SITE-DIRECTED MUTAGENESIS PARAMETERS....................209 F CIRCULAR DICHROISM AND CW-EPR PARAMETERS.............................................212 G SPECTRAL PARAMETERS FOR THE EPR SPECTRA REPORTED IN THIS WORK........................................................................................................................... .......213 H 3D PLOTS FOR SPECTRAL PARAMETER S AND CW-EPR LINESHAPES AS A FUNCTION OF COSOLUTES FOR OS MOLALITY/VISCOSITY STUDIES.................216 LIST OF REFERENCES............................................................................................................. 221 BIOGRAPHICAL SKETCH.......................................................................................................230

PAGE 9

9 LIST OF TABLES Table Page 1-1 FDA approved drugs for highly activ e antiretroviral therapy (HAART)..........................22 1-2 HIV-1 protease proce ssing sequences in HIV...................................................................24 2-1 Parameters used in Tm measurements................................................................................56 2-2 Distance distribution profile used to analyze noise eff ects on dipolar evolution data: analysis for a 1 Gaussian distribution................................................................................64 2-3 Distance distribution profile used to analyze noise eff ects on dipolar evolution data: analysis for a 2 Gaussian distribution................................................................................65 4-1 HIV-1 PR flap mutants with WT activity..........................................................................94 4-2 Spectral parameters for the lineshapes of PMPR*SL........................................................99 4-3 Secondary structure assignments for HIVPR (PDBID 1HVI) and LAI consensus.........105 4-4 CD secondary structure assignments for PMPR*SL (MTSL, MSL, IAP, IASL)...........106 4-5 Osmolality and viscosity of solutions of cosolutes employed in this study (T=24C)....112 5-1 EPR distance measurements for K55SL-K55SL HIV protease.....................................148 6-1 Main and accessory mutations that a ppear in patients under protease inhibitor therapy........................................................................................................................ ......166 6-2 Common primary and secondary mutations in HIV-1 protease in similar positions to the V6 strain.................................................................................................................. ...168 6-3 Kinetic parameters for V6 strain and mutations in V6....................................................168 6-4 Susceptibility of WT and MDR769 isolates to protease inhibitors.................................169 6-5 2-Gaussian fit parameters employed to fit distance dist ribution profiles for PMPR*MTSL in the presence of Indi navir or chromogenic substrate............................174 6-6 Average distance distributions for PMPR,V6 and MDR769*MTSL..............................178 6-7 FWHM distributions for PMPR,V6 and MDR769*MTSL.............................................178 A-1 Common drug-induced muta tions in HIV-1 protease......................................................194 B-1 cw-EPR parameters for DEER experiment......................................................................198

PAGE 10

10 B-2 Field-swept echo paramete rs for DEER experiment.......................................................199 B-3 Echo decay (Tm) experiment............................................................................................200 B-4 Inversion decay (T1) experiment......................................................................................201 B-5 4p-DEER setup.............................................................................................................. ..202 B-6 Field swept 4p-DEER......................................................................................................20 3 B-7 4p-DEER experiment.......................................................................................................20 4 D-1 Summary of closest neighbor interacti ons between Ritonavir and HIV-1 protease........208 E-1 PCR primers utilized to introduce mutations in HIVPR..................................................209 E-2 Non codon-optimized PMPR DNA and protein sequences.............................................209 E-3 Codon-optimized PMPR DNA and protein sequences....................................................210 E-4 CO-PMPR+D25N DNA and protein sequences..............................................................210 E-5 CO-PMPR+K55C DNA and protein sequences..............................................................210 E-6 CO-PMPR+D25N+K55C DNA and protein sequences..................................................210 E-7 MDR769 DNA and protein sequences............................................................................211 E-8 V6 DNA and protein sequences.......................................................................................211 E-9 Thermal cycling parameters for HIV1 protease site-direc ted mutagenesis....................211 F-1 Typical parameters used for circular dichroism experiments..........................................212 F-2 Typical cw-EPR parameters used in this study................................................................212 G-1 Spectral parameters for the EPR spectra for the osmolality/viscosity effects.................213 G-2 Lineshape spectral parameters fo r the degradation of PMPR+K55MSL........................215

PAGE 11

11 LIST OF FIGURES Figure Page 1-1 Complete genome of HIV-1 virus......................................................................................31 1-2 Immature and mature form of the virus.............................................................................31 1-3 HIV-1 viral life cycle..................................................................................................... ....32 1-4 Drug targets in the HIV-1 virus life cycle.........................................................................32 1-5 Schematic phylogeneti c tree of HIV virus.........................................................................33 1-6 Circulant recombinant forms (CRF) of HIV-1 virus, group M.........................................33 1-7 X-ray crystal structure of HIV-1 protease bound to CA/p2 cleavage site.........................34 1-8 HIV-1 protease ptructure................................................................................................... 34 1-9 HIV-1 protease publication record/year.............................................................................34 1-10 HIV-1 protease NOE values..............................................................................................35 2-1 Electromagnetic spectrum showing various bands of EPR operation............................66 2-2 Energy level diagram for a free electron in an applied magnetic field..............................66 2-3 Comparison between cw-EPR and optical spectroscopy instrumentation.........................67 2-4 Typical energies (in Hz) of elect ron and nuclear spin interactions...................................68 2-5 Energy splitting diagram and typical cw -EPR lineshape for a nitroxide moiety...............68 2-6 Two-electron spin density matrix......................................................................................69 2-7 Pulsed ELDOR sequences.................................................................................................69 2-8 4-pulse DEER dipol ar evolution traces..............................................................................70 2-9. 4-pulse DEER sequence..................................................................................................... 70 2-10 Complete 4-pulse DEER scheme incl uding a representation of the magnetization vectors for each step in the pulse sequence.......................................................................71 2-11 Positions of the observer and the pu mp pulses for the DEER experiment........................72 2-12 Measurement of the sp in-lattice relaxation time (T1) by saturation recovery...................72

PAGE 12

12 2-13 Pulse sequence for the 2-pulse Hahn Echo T2/Tm measurements......................................73 2-14 Measurement of the spin-spin relaxation time (Tm) by a Hahn echo pulse sequence.......73 2-15 Comparison between instrumentation fo r nuclear magnetic resonance and pulsed electron paramagnetic resonance.......................................................................................74 2-16 Graphical determination of commonl y used cw-EPR spectral parameters.......................74 2-17 Background correction of the partially refocused echo in DEER spectroscopy................75 2-18 Data analysis algorithms for DEER dipolar evolution signal............................................76 2-19 Screenshot capt ure of DeerSim..........................................................................................77 2-20 TKR vs. MC analysis for a single Gaussi an distribution as a function of the average position and experimental noise.........................................................................................78 2-21 Montecarlo and Tikhonov regularization an alysis for a 2-Gaussian distribution..............79 3-1 Pilot expression of HIV-1 proteas e (PMPR) in LB media at 37C...................................89 3-2 DNA sequence encoding HIV-1 protease in pET-23a vector............................................89 3-3 Decomposition of urea into ammonium and cyanate ions.................................................90 3-4 Cartoon showing the labeling product from the chemical reaction between cyanate and sulfhydryl group..........................................................................................................9 0 3-5 Reducing Tris-Tricine gel showing the different steps in the purification of HIVPR+PMPR+D25N+K55C...........................................................................................91 3-6 Reducing Tris-Tricine gel showing the di fferent steps in the purification of active HIVPR+PMPR+K55C.......................................................................................................92 3-7 4-pulsed double electron-elec tron resonance (DEER) sequence.......................................92 4-1 Detailed molecular structure of the flap region of HIV-1 protease.................................121 4-2 Spin labels used in this study...........................................................................................12 1 4-3 Structure for the spin label MTSL appended to the cysteine residue..............................122 4-4 Scheme representing the factors i nvolved in the lineshape mobility...............................122 4-5 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease PMPR*SL in 2 mM NaOAc pH 5.0 as a function of spin label......................................123

PAGE 13

13 4-6 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease PMPR*SL in 2 mM NaOAc pH 5.0 in the presence and absence of inhibitor................124 4-7 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease PMPR*MTSL as a function of salt concentration...........................................................125 4-8 Area normalized 100 G X-Band EPR sp ectra showing the effects of acid concentration (HCOOH) on the cw -EPR lineshape of PMPR*SL..................................126 4-9 Graphical representation of spectral parameters obtaine d from the area normalized 100 G X-Band EPR spectra corresponding to the acidification of PMPR* (IAP/MTSL)..................................................................................................................... 127 4-10 Possible protonation reaction of the acetamido group in IAP.........................................127 4-11 Molecular orbital diagram for the acetamido moiety in IAP...........................................127 4-12 Scheme for light polarizati on and differential light absorp tion in the CD experiment...128 4-13 Sample circular dichroism spectra...................................................................................128 4-14 Comparison between circular dichroism spectra of HIVPR LAI consensus sequence and PMPR*SL.................................................................................................................129 4-15 Area normalized 100 G X-Band EP R spectra of HIV-1 protease...................................129 4-16 Degradation of chromogenic substrat e by active HIV-1 protease (PMPR+K55MSL) followed by UV/Vis spectroscopy...................................................................................130 4-17 Area normalized 100 G X-Band EPR spectra of HIVPR PMPR+K55MSL undergoing self-proteolysis followed by cw-EPR...........................................................131 4-18 Intensity ratio of the high field lin e/low field line for PMPR+K55MSL undergoing self-proteolysis............................................................................................................... ..132 4-19 Area normalized 100 G X-Band EPR spect ra of HIVPR PMPR+K55MSL on day 0 and day 46 of the self-degradation kinetic study.............................................................132 4-20 Structure of three compounds utilized to study viscosity and osmolality effects............133 4-21 Plot of the viscosity and osmolality as a function of percent content for the four solutes under study...........................................................................................................1 33 4-22 Osmolality vs. viscosity plot for solutes commonly used in cw-EPR.............................134 4-23 Changes in central field linewidth ( Hpp) for HIV-1 PMPR*SL as a function of cosolute species and concentration..................................................................................135

PAGE 14

14 4-24 Changes in spectral second moment

for HIV-1 PMPR*SL as a function of cosolute species and concentration..................................................................................136 4-25 Changes in LF/CF intensity ratio for HIV-1 PMPR*SL as a function of cosolute species and concentration................................................................................................137 4-26 DEER spectra of PMPR*MTSL as a func tion of cryoprotectant added (Glycerol and Ficoll 400).................................................................................................................... ....138 4-27 Tm measurements for HIV-1 PMPR*MTSL as a function of the composition of the cryoprotectant solution.....................................................................................................138 5-1 Distance analysis between residues K55-K55 of HIV-1 protease from X-ray crystal structures in the PDB databank........................................................................................153 5-2 Comparison between the struct ures of Lys55 and Cys55MTSL.....................................154 5-3 Area normalized 100 G X-Band EPR lines hape for PMPR*MTSL in the presence and absence of inhibitor...................................................................................................154 5-4 DEER results for HIVPR PMPR*MTSL.........................................................................155 5-5 DEER results for HIVPR PMPR*IAP.............................................................................156 5-6 DEER results for HIVPR PMPR*MSL...........................................................................157 5-7 DEER results for HIVPR PMPR*IASL..........................................................................158 5-8 Comparison between the distance dist ribution profiles functions obtained by DA2006 and DeerFit from the DEER dipolar evolution signal for PMPR*SL...............159 5-9 Average distance and distance distribution confidence interval plot for PMPR*MTSL..................................................................................................................160 5-10 Average distance and distance distribu tions for PMPR*SL in the presence and absence of inhibitor..........................................................................................................1 61 5-11 Correlation between spectral parameters obtained from cw-EPR lineshape analysis and FWHM of the distance dist ribution profile for PMPR*SL.......................................162 5-12 Distance distribution prof iles obtained by MD simulations between K55C labeled sites on HIV PMPR+D25N..............................................................................................163 6-1 Natural polymorphisms occurring in PI nave patients for all HIV-1 protease subtypes (A-AG)..............................................................................................................18 1 6-2 Drug-selected mutations occurring in PI exposed patients for all HIV-1 protease sybtypes (A-AG)..............................................................................................................18 2

PAGE 15

15 6-3 Mutation prevalence (%) difference be tween nave and PI exposed patients, by subtype........................................................................................................................ .....183 6-4 Mutation prevalence for subtype B as a function of residue number..............................184 6-5 Ribbon diagram of HIV-1 protease, V6 variant...............................................................184 6-6 Side and top views of HIV-1 proteas e LAI and MDR769 crystal structures..................185 6-7 Crystal structure of MDR769 includ ing nearest lattice symmetry neighbors..................185 6-8 X-ray crystallography structures of MDR769.................................................................186 6-9 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease drug resistant strains.............................................................................................................. ...186 6-10 DEER spectra of HIVPR PMPR*MTSL.........................................................................187 6-11 DEER spectra of HIVPR V6*MTSL...............................................................................188 6-12 DEER spectra of HIVPR MDR769*MTSL.....................................................................189 6-13 2-Gaussian fit to the distance di stribution profile fo r HIV PMPR*MTSL......................190 6-14 Comparison between distances obtain ed by DEER spectroscopy for HIVPR-PMPR, HIVPR-V6 and HIVPR-MDR769...................................................................................190 B-1 Sample cw-EPR of spin labeled HIVPR at 65 K.............................................................198 B-2 Sample field-swept echo spectra of spin labeled HIVPR at 65 K...................................199 B-3 Sample Hahn echo experiment of spin labeled HIVPR at 65 K......................................200 B-4 Sample inversion recovery experime nt of spin labeled HIVPR at 65 K.........................201 B-5 Sample refocused spin echo parameters on HIVPR sample............................................202 B-6 Sample field-swept spectra on a refocu sed echo on spin labeled HIVPR at 65 K..........203 B-7 Sample dipolar evolution in the DEER e xperiment of spin labeled HIVPR at 65 K......204 D-1 HIVPR crystal struct ure with Ritonavir...........................................................................207 D-2 Interactions between Rit onavir and HIV-1 protease........................................................207 H-1 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR-PMPR*MTSL as a function of cosolute concentration, a nd 3D plot showing pe rcent change in the spectral parameters...........................................................................................................2 17

PAGE 16

16 H-2 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR PMPR*MSL as a function of cosolute concentration, and 3D plot showing percent change in the spectral parameters...........................................................................................................2 18 H-3 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR PMPR*IAP as a function of cosolute concen tration, and 3D plot showi ng percent change in the spectral parameters...........................................................................................................2 19 H-4 Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR-PMPR*IASL as a function of cosolute concentration, and 3D plot showing percent change in the spectral parameters...........................................................................................................2 20

PAGE 17

17 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SITE-DIRECTED SPIN LABELING EPR STUDIES OF CONFORMATIONAL DYNAMICS IN THE FLAP REGION OF HIV-1 PROTEASE By Luis Galiano August 2008 Chair: Gail E. Fanucci Major: Chemistry Human Immunodeficiency Virus Ty pe 1 (HIV-1) protease is an essential component in the processing of viral proteins encoded in the HIV viral genome. To allow full access of the substrate to the active site of the protease, two beta hairpin regions term ed flaps, have to undergo a large conformational chan ge, and it has been postulated that these flaps move in a segmental motion, with three major conformati ons (open, semi-open and closed or bound to inhibitor). Site-Directed spin labeling (SDSL) coupled with Electron Paramagnetic Resonance (EPR), is a powerful technique that can be utilized to measure distances and distance distributions associated with conformational changes. Here, we explored the use of SDSL-EPR in order to elucidate the distances and distance distributions accessible to the flap region in the protease. First, we studied the distances and distance distributions in a c onstruct termed PMPR (PentaMutated Protease), which pos sesses wildtype-like structure and kinetic properties. From the distance distribution prof iles obtained by utilizing double electron-electron resonance (DEER) spectroscopy, we established that the motion of the flaps in the ab sence of inhibitor is continuous, rather than segmental, and that it spans a set of distances between 24 48 Upon

PAGE 18

18 introduction of the inhibitor, the average distance is narrowed by ~3 which is consistent with the picture of the flaps closed on top of the inhibitor upon active site binding. Second, the drug-resistance problem in HIV-1 protease was investigat ed. When patients undergo Highly Active Antiretrovi ral Therapy (HAART), the dr ug pressure induced in the patient selects those mutant strains that have inhibitor resistance. Our results show that the distance distribution profile for two drug resist ant strains, namely V6 and MDR769, did not significantly change for the protease in the presen ce and absence of inhibitor. We also found that the distance distributions accessible to the spin la bels were narrower than in the PMPR strain, which would indicate a reduced flexibility of the flap in these mutant strains. The results obtained from this work will undoubt edly serve as a stepping stone for further research on the fields of HIV-1 protease, HIV protease drug resist ance and SDSL-EPR.

PAGE 19

19 CHAPTER 1 INTRODUCTION TO HIV-1 Introduction Acquired Immunodeficiency Syndrome (AIDS) caused by the Human Immunodeficiency Virus (HIV), is a global health crisis with approximately 40 m illion people affected worldwide. The infection is characterized by an acquired, irreversible profound immunosuppresion that predisposes patients to multiple opportunistic in fections and progressive dysfunction of multiple organ systems. HIV is a virus of the retroviral family (viral genetic material is encoded as a positive-sense single stranded RNA, or +(m)RNA, and the virus contains the enzyme Reverse Transcriptase (RT) in order to generate a DNA in termediate). HIV is a member of the subfamily of the Lentiviruses, characterized by their long incubation periods. Figure 1-1 shows the complete genome of HIV-1 Virus. In pink, the Group-specific Antigen (or gag ) gene is responsible for th e structural elements that form the virus, encoding for the proteins matrix (MA), capsi d (CA) and nucleocapsid (NC). In light blue, the polymerase gene ( pol ) encodes for three enzymes that play a fundamental role in the replication and maturation of the virus, namely protease (PR), re verse transcriptase (RT) and integrase (IN). In orange, proteins responsible for viral recogni tion and binding to the host are encoded in the envelope gene ( env ); these proteins are Surface (SU), and transmembrane (TM). Other accessory proteins are Rev (responsibl e for the upregulation of gag pol and env ), Tat (transcriptional regulator that binds to the TAR sequence in the RNA), Vpr (involved in the formation and localization of the Pre-integra tion Complex, or PIC), Vpu (degradation of CD4 and enhancement of virion release) and Nef (suppression of the cell immune response by downregulation of CD4 and MHC Class I cell-surface proteins). Figure 1-2 shows the immature and mature form of the virus.1

PAGE 20

20 HIV-1 Viral Life Cycle The HIV-1 life cycle shares common elements with other members of the retrovirus family. Five principal stages can be distinguished within this cycle: 1) targeting and fusion, 2) uncoating, reverse transcription a nd formation of the integration complex, 3) nuclear import and integration in the host DNA, 4) transcription and translation of HIV1 polyproteins and 5) membrane localization, budding and pol yprotein processi ng (maturation). The infection process begins with the rec ognition and binding of the surface protein gp41 (SU)2,3 to CD4 receptors in the cell. The membra ne fusion process is mediated by TM (gp120)4-7, releasing the viral core, which cont ains the viral RNA as well as accessory proteins necessary for the next viral steps, into the cytoplasm. After th e core release, the uncoating process begins with the disassembly of the capsid (CA)8,9, and nucleocapsid (NC)10,11 proteins. This step in the viral maturation is poorly understood, a lthough it has been hypothesized th at the process involves the phosphorylation of matrix (MA)12,13 by a MAP kinase. The uncoati ng step releases the RNA and other proteins that were packed in the core such as RT, IN, Vpr and Vpu into the host cytoplasm. Viral RNA is reverse transcribed by the pol protein RT14-17 and the viral RNA is processed by RNAse-H to allow the formation of a complementary DNA strand, thus creating a double stranded viral DNA. A high molecular wei ght complex formed by viral cDNA, IN, MA, Vpr and host cell proteins, known as pre-integrati on complex (PIC) is then shuttled inside the nucleus. Afterwards, integrase (IN)18 catalyzes the integration of the viral DNA into the host DNA by splicing the viral DNA ends, the host DNA, and using the cell repair machinery to perform the integration. Depending on the integration posit ion of the viral DNA in the host, viral latency times vary. Upon transcription of the vira l DNA, viral mRNA travels back to the cytoplasm and the polyproteins gag, gag-pol and env, as well as accessory proteins, are translated. It is noteworthy

PAGE 21

21 to mention that the pol polyprotein never gets produced by itself, but ra ther as gag-pol fusion construct, due to a stem-loop that forms in th e C-terminal region of gag which causes a -1 ribosomal frameshift (the ratio gag/ga g-pol has been estimated to be 20:1)19. Env polyproteins are split into SU and TM by hos t cell proteases, and travel to the cell membrane. Gag and gagpol polyproteins target the cell membrane as we ll, initiating the budding of the immature virus. The viral maturation process concludes when HIV-1 protease splices the gag and gag-pol polyproteins into its constitutive proteins, namely matrix (MA), capsid (CA), nucleocapsid (NC), reverse transcriptase (RT), prot ease (PR) and integrase (IN). Drug Therapy Against HIV1 Virus: Current Approaches Antiretroviral therapy (ART) has been one of the mo st important areas in medicine in the last 25 years. The first attempts on HIV treatments date from 1987 with the introduction of AZT20, although it did not seem to provide durable efficacy. It wasnt until 1995 when the first combination therapies using nucleosid e analogs to inhibit RT appeared21,22, quickly followed by the first protease inhibitors (PIs)23. Currently, there are 4 typical st ages in the viral lifecycle that are being used as targets for drug therapy, namely reverse transcription (with nucleoside (NRTIs) and non-nucleoside (NNRTIs) reverse transcriptase inhibitors), viral matu ration ( with protease inhibitors (PIs)) and, more recentl y, viral fusion (entry inhibitors (EI)) and viral DNA integration in the host DNA (integrase inhibitors (II)). Table 1-1 shows the current FDA approved drug s for HIV treatment. It can be seen that HIV-1 protease has been extensively targeted as a candidate for drug therapy, which resulted in the approval of 10 different pr otease inhibitors. Despite the high number of PIs and other inhibitors for different targets in the HIV lifecycle, drug toxic ity and drug-selected mutations remain the biggest problem for a successful HIV1 treatment. Figure 1-4 shows the targets of different inhibitors in different stages of the viral lifecycle. Fusion inhibitors target the fusion

PAGE 22

22 step mediated by TM, SU and the host CD4 re ceptors and other correceptors, NRTIs and NNRTIs block the reverse transcription step, integr ase inhibitors block the viral DNA integration into the host cell DNA, and PI inhibitors bloc k the maturation process of newly synthesized virions. Table 1-1. FDA approved drugs for highly active antiretroviral therapy (HAART) NRTIs NNRTIs PIs Entry inhibitors Integrase inhibitors Zidovudine (AZT) Delavirdine (DLV) Amprenavir (APV) Enfuvirtide (ENF) Raltegravir Emtricitabine (FCT) Efavirenz (EFV) Tipranav ir (TPV) Celsentri Lamivudine (3TC) Nevirapine (NPV) Indinavir (IDV) Stavudine (d4T) Saquinavir (SQV) Abacavir (ABC) Lopinavir LPV) Didanosine (DDL) Fosamprenavir (FPV) Tenofovir (TDF) Ritonavir (RTV) Darunavir (DRV) Atazanavir (ATZ) Nelfinavir (NFV) Viral Polymorphism: Reverse Transcript ase Lacks Proofreading Capabilities Reverse transcriptase (RT), also known as RNA-dependent DNA polymerase, is an enzyme responsible for the production of a double-stranded DNA from an RNA template. As opposed to DNA-polymerases, which often possess proof-reading ability (proof-reading enzymes are those that can self-correct its own polymeriza tion errors as the transc ription occurs), HIV-1 RT lacks proofreading capability, and thus is unable to correct randomly introduced mutations while reverse transcribing the viral RNA to viral DNA within the hos t cytoplasm. Current in-vivo studies of HIV-1 RT estimate a mutation rate of 3.4 x 10-5 mutations/ (bp x cycle)24. Comparing this rate with the rate obtained for human DNA polymerases (5 x 10-11mutations/(bp x cycle))25, the mutation rate ratio for HIV(RT)/DNApol is 680.000:1. HIV mutations are classified by their proximity with a common sequence, or subtype (each subtype being defined by a consensus sequence). Figure 1-5 shows the cl assification of the HIV virus into different groups, subtypes

PAGE 23

23 and CRFs (Circular Recombinant Forms). Although a detailed description of the phylogenetic classification of the virus is outside the scope of this wor k, a brief introduction to the phylogenetic tree of the virus will be given. HIV-1 virus is the most prevalent form of HI V, and it is what is popularly known as the HIV Virus. The other HIV viru s variant, the less known HIV-226,27, has a decreased sexual transmission probability and decreased infectivities w ith longer life expectancies for hosts. It has been hypothesized that the orig in of HIV-1 and HIV-2 is di fferent, with HIV-1 being a descendent of the SIVcpz (Simian Immunodeficiency Virus Chimpanzee) whereas HIV-2 might have evolved from SIVMM (the SIV variant from Sooty Mangabey monkey) 28,29. HIV-1 Groups 30-33 refer to the very distinctive HIV1 lineages, being Group M (main) the most prevalent. The other two groups, O (outli er) and N (non-M/non-O) are far less common and only existent in Cameroon. HIV-1 Subtypes refer to the major clades (a clade is a taxonomic group comprising a single common ancestor and all the descendants of that ancestor) within subgroup M. Circulant Recombinant Forms (CRFs) describe a recombinant lineage that plays an important role in the HIV-1 pandemic. Curren tly, there are only 4 CRFs recognized, termed CRF01_AE, CRF02_AG, CRF03_AB and CRF04_cpx, de pending on the discovery date and the virus strains that form the CRF. It is importa nt to note that natura lly occurring sequence polymorphisms exist within each of the subtypes of HIV protease. As we shall see, in patients who have not been previously exposed to drug th erapy (nave patients), th is sequence variability is confined to specific sites within the viral genome. Figure 1-6 shows the viral composition of the 4 CRFs that have been recognized to date. It can be seen that each of the CRFs is formed by recombination of different viral strains that coexist in the same organism (For example, CRF01_AE is a combination of subtypes A a nd E, whereas CRF04_cpx combines subtypes A,

PAGE 24

24 G, K and H, as well as long segments that are not classified). The existence of multiple subtypes with naturally occurring polymorphisms, as well as CRFs poses a great cha llenge in the design of new inhibitors, as each of these subtypes po ssesses different characteristics in terms of infectivity, latency, viral viability and viral fitness. The Role and Structure of HIV-1 Protease As mentioned previously, HIV1 protease plays a fundamental role in the HIV-1 viral life cycle; it is responsible for the cleaving the gag a nd gag-pol polyproteins th at occur in the viral maturation stage. It has also been shown that inhibition of HIV-1 protease leads to immature, non-infectious virions.34,35 Table 1-2 summarizes the processing sites and cleavage sequences of HIV-1 protease. Table 1-2. HIV-1 protease pro cessing sequences in HIV Location Cutsite Sequence (HXB2 consensus) Pr55 Gag MA/CA VSQNY / PIVQN CA/p2 KARVL / AEAMS p2/NC TSAIM / MQRGN NC/p1 ERQAN / FLGKI P1/p6 Gag RPGNF / LQSRP Pr160 Gag-pol NC/TFP ERQAN / FLREN TFP/p6 Gag-pol EDLAF / LQGKA P6pol/PR TSFSF / PQITC PR/RTp51 CTLNF / PISPI RT/RTp66 GAETF / YVDGA RTp66/INT IRKVL / FLDGI Nef Nef AACAW / LEAQE The nomenclature for the amino acid cleavage positions is P4-P3-P2-P1*P1-P2-P3-P4, where the asterisk represents the scissile bond. Conversely, S4-S3-S2-S1-S1-S2-S3-S4 represents each of the bi nding pockets in the protease36-38.

PAGE 25

25 HIV-1 protease is a member of the aspartic protease family (A02.001)39. Structurally, it is a symmetric homodimer composed of two 99 ami no acid residue polypeptides. The structure of HIV-1 protease is shown on Figures 1-7 and 1-8. It can be seen that the monomer is formed by the repetition of four structural elements: a hairpin (containing loops A1 and A2), a wide loop (B1, containing the catalytic aspartic acid, and B2), an alpha helix (C 1 and C2) and a second hairpin (D1 and D2). Access to the floor of the active site formed by the B1 loop can be gained upon opening of the D1 hairpins, termed flaps. The linker regions between helix C1 and the flap and between the flap and the A2 loop are term ed hinges of the flap (Figure 1-8A). Figure 1-8B shows the structure of inhibited HIV-1 prot ease (1HVI), where the inhibitor is rendered in tube structure, and a volumetric rendering was ge nerated for all the residues within 5 of the inhibitor. From this representa tion it can clearly be seen that the access to the active site is blocked by the flaps (in blue), a nd that the flaps must open in order for the inhibitor or the natural substrate to access the ac tive site. Therefore, it is ex pected that modulation of the dynamics of aperture/closing of the flaps can have a dramatic effect on th e protein inhibition and catalytic activities. Previous Studies Performed on HIV-1 Protease Introduction HIV-1 protease remains one the most studied proteins, and without doubt, the most studied aspartic protease. As of January 2008, 207 struct ures were deposited in the Protein Databank (PDB), and another 72 structures were waiting to be released. It has been studied using a myriad of biophysical and biochemical techniques, su ch as X-ray crystallography, isothermal calorimetry (ITC), nuclear magnetic resonance (N MR), fluorescence and enzyme kinetics, as well as theoretical methods such as coarse-gra ined simulations and molecular dynamics (MD).

PAGE 26

26 Figure 1-9 reflects the increase in the number of publications on HI V-1 protease alone, per year since the first pape r was published in 1988. Despite the large number of publications in th e subject (around 1600 in 20 years), the most important questions remain elus ive: How does the substrate a ccess the active site? How does HIV-1 protease become resistant to inhibitors? Which are the main processes that are involved in drug resistance? The first characterization of HIV-1 proteas e was performed in 1988 by total chemical synthesis. It was shown that the protease belongs to the aspartic protease family, and substrate specificity for processing of gag and gag-pol po lyproteins was characteri zed. Shortly after, the first recombinant HIV-1 protease was expressed in E. coli leading to the first X-ray crystallography studies, and the firs t complete mutagenesis studies were performed in an attempt to characterize well conserved regions in the prot ein that might serve as drug targets. At this point, the search for new inhibitors by struct ure-based drug design star ted. Today, development of drug therapy and protease inhibi tors is largely based on a rat ional approach to drug design, combining information from the system obtained from diverse biophysical/biochemical and computational techniques to gain a basic understa nding of the protease structure and function at the atomic level. In this chapter, we will tr y to summarize the principal structural studies on HIV-1 protease. The LAI Subtype B Consensus Sequence The B.FR.83.HXB2 sequence is the primary re ference for HIV-1 at the Los Alamos HIV Database. HXB2 is a specific clone from the French isolate LAI40 (also known as BRU). The first crystallographic studies of recombinant native HIV-1 protease were first reported by scientists from Merck41, and later confirmed by crystal stru ctures from recombinant and total chemical synthesis of the protease42. There are roughly 200 X-ra y crystallography studies

PAGE 27

27 published for HIV-1 protease; in this chapter we will try to summarize only a few of those studies due to their relevance in the present study or future work. In 1990, Gustchina et al. performed a comparison between HIV-1 protease bound to an inhibitor and non-viral aspartic proteases. In their study they c oncluded that HIV-1 protease shows more interac tions between a peptidic in hibitor vs. their non-viral counterparts. The existence of two flaps that cover the active site re stricts the access and conformation of the polypeptide precurso r of the protease to the active site.43 X-ray crystallography on HIV1 protease has been widely used to design and improve new protease inhibitors. A clear example of this is the success in structure-based inhibitor design.44-46 X-ray crystallography has been recently applied to the study of non-subtype B proteases, and the structural similarities and differenc es with HIV subtype B, as well as other retroviral proteases associat ed to immunodeficiency in animals, such as FIV (Feline Immunodeficiency Virus) 47-49, SIV (Simian Immunodeficiency Virus)33,50,51 or EAIV (Equine Anemia immunodeficiency virus)50,52. Undoubtedly, the biggest contri bution of X-ray crystallograp hy in the field of HIV-1 protease is the possibility to study single-point mu tations as well as multidrug resistant proteases, and derive useful atomistic information from crystal structures. Alt hough the picture provided from x-ray crystallography is that of a static pr otein, neglecting protein dy namics, it is the most powerful tool used to date to obtain structural details at the atom level. Nuclear Magnetic Resonance (NMR) is a powerful technique to elucidate protein structure in solution. D.A. Torchias group has pioneered the NMR research on HIV-1 protease dimer, as well as extensive studies on HIV-1 protease monomer. We will focus on their studies of the protease dimer free in solution and bound to inhibitors, substrates a nd substrate analogs. For their work on HIV-1 protease monomer, th e reader is referred to Ref. 53-55 and references therein. In the unliganded HIVPR NMR characterization, Torchia et al. performed an analysis based on 1H and 15N relaxation rates of the backbone amides56. This study provided direct evidence of the millisecond-microsecond timescale fo r the motions of the flap. NOE spectra of the free protease indicates that the conformations of the flap are best described as an ensemble of

PAGE 28

28 semi-open conformations, which might include sm all populations of an open form and the closed form, all in dynamic equilibrium. Upon subs trate binding, data obtained from model-free S2 order parameters suggest that flap motion in the sub-ns timescale is not completely quenched in the substrate-bound protease, bu t the flaps still present a clos ed conformation similar to the inhibitor-bound protease. Upon inhibition, for which they choose DMP323 as the inhibitor, they report that residues i and i+100 (same residues in different monomers, for which the notation (i, i) is also commonly used) have identical chemical shifts, showing that the average solution structure of the protease is symmetric on the mss timescale. Figure 1-10 shows the 1H-15N NOE values for HIV-1 free an d bound to inhibitor. From the comparison between the red (bound) and the blue (free) data points, it can be seen that the flap region (46-56) undergoes an unordered (low NOE values) to ordered (higher NOE values) transition upon inhibition. The main conclusions that can be drawn fr om NMR experiments are summarized below: The flap region is flexible on the mss timescale, with sub-ns motions in the tips of the flaps. The flap maintains the hydrogen bondi ng pattern even when the protease is uninhibited. The flap region becomes more rigid upon protease inhibition. In the presence of substrate, the flap motion is not completely quenched, al lowing for ns motions in the entire flaps. From the protease chemical shifts, th e monomers are symmetric on the mss timescale. The uninhibited conformation of the protein ca n be described as an ensemble of semiopen conformations, which might include some closed and fully ope n structures, in dynamic equilibrium. Isothermal Calorimetry (ITC) has been extens ively used by the group of Ernesto Freire to characterize HIV-1 protease fo lding, drug-selected mutations and binding of substrate and inhibitors to the active site of the protease. ITC is a powerfu l technique to investigate the thermodynamic properties of molecular interactio ns, by accurately determining the enthalpy of

PAGE 29

29 binding as well as the binding affinity and th e stoichiometry in a single experiment. This technique can be used for interactions with low, moderate and high binding affinities, making it a powerful technique for studying the interaction between HIVPR and substrate/inhibitors. In 1998, Freire et al. published the first thermodynamic characterization of the protease57, and determined the values for the thermal denatu ration temperature of the protease (59 C at 25 M protease, pH 3.4), as well as the urea denatura tion curves (full denatu ration occurs at 6 M urea). Utilizing structure-based thermodynamic ch aracterization, they concluded that the most stable region of the protease is the dime rization region, between residues 1-7 and 92-99, followed by the active site floor (22-28) and the ti p of the flaps (47-52). The region that had the least contribution to the dimer stabi lization energy was the protease flap. ITC has been widely applied to study the eff ects of mutations on the binding affinities and the thermodynamic design of inhibitors. For furt her reviews, the reader is referred to Refs.57-60 HIV-1 protease has also been widely studied by theoretical m eans, such as MD simulations and coarse-grained simulations. MD simulations have been proven to be a valuable tool in structure-based inhibitor design, as well as understanding the internal dynamics of the flaps and the nature of the inhibitor binding process, as well as the effects of point mutations on the protease structur e and dynamics. In 2005, Simmerling et al. presented the first unrestrained, all-atom MD simulation of the protease for which the flaps spontaneously open and close61,62. This is an important result since all MD simulations previous to 2005 used steered MD or activated dynamics to be able to open the flaps, without achieving a re-closi ng event in the experimental time. From these MD simulations, Simmerling et al. report that, in the pr esence of inhibitor, only the closed conformation of the flaps is presen t, in agreement with crystallographic data, and

PAGE 30

30 that the flap flexibility in th e inhibited (closed) conformation is reduced, as previously found by NMR. The most stable structure for the protease in the absence of inhibitor is similar to the semi-open form reported by crystallography, bu t this semi-open form undergoes several conformational changes towards a more open" conformation. It was also shown that, upon addition of inhibitor to the open conformati on, the flaps close, and the handedness of the tips of the flaps reverse with respect to that of the semi-open forms. Other studies on the protease using MD have been focused on the role of molecular crowding in flap opening events63, the role of water in the peptide bond cleavage reaction64,65, and the protonation states of the catalytic aspartates.64,66-69 From the structural point of view, we can summarize the most pertinent results obtained from MD as follows: In the presence of inhibitor, the flexibility of the flaps is greatly reduced and the flaps remain in a closed conformation In the absence of inhibitor, a structure that resembles the semi-open form as obtained by crystallography is the predominant form but the flap undergoes several opening events which are compatible with an open conformation which allows for substrate/inhibitor binding to the active site Upon addition of inhibitor to the open c onformation, the flaps close and the handedness of the flaps is reversed with respect to that of the semi-open conformation The sub-nanosecond motion of the tips of the flaps might be involve d in flap-opening and the well conserved sequence of the flap tips (IGGIGG) might be exploited in new drug development.70 Multiscale simulations (coarse-grained a nd all-atom MD) were employed to study binding of inhibitors to HIV1 protease. It was shown that inhibitors can bind the open HIV protease, and small urea cyc lic inhibitors may be able to bind the active site in a closed conformation of the protease, but with a large inhibitor/activ e site reorganization penalty.71-73

PAGE 31

31 nefvpunef vpr tat rev LTR 3 Env MA NC CAp 6 MA NC CA LTR 5 pol gag MANC CAp 6PRIN RT MA CA PR IN RT MA CA PRIN RT SUTM SUTM SU TMGag polyprotein Gag-Pol polyprotein Env polyprotein protease protease cellular protease nefvpunef vpr tat rev LTR 3 Env MA NC CAp 6 MA NC CA LTR 5 pol gag MANC CAp 6PRIN RT MA CA PR IN RT MA CA PRIN RT SUTM SUTM SU TMGag polyprotein Gag-Pol polyprotein Env polyprotein protease protease cellular protease Figure 1-1. Complete geno me of HIV-1 virus. Figure 1-2. Immature and mature form of the vi rus.[Adapted from Louis E. Henderson and Larry Arthur, NIH, NIAIDS]

PAGE 32

32 CD4RT-RNAseH Complex PICIntegration Transcription Translation Uncoating Rev. Transcription Envproteins (SU, TM) Budding Maturation Binding Fusion Import Gag, gag/pol precursors HIV-1 Life Cycle CD4RT-RNAseH Complex PICIntegration Transcription Translation Uncoating Rev. Transcription Envproteins (SU, TM) Budding Maturation Binding Fusion Import Gag, gag/pol precursors HIV-1 Life Cycle Figure 1-3. HIV-1 viral life cycle. Integrase Inhibitors CD4Integration Rev. Transcription Maturation Fusion NRTI & NNRTI Fusion Inhibitors PIs Integrase Inhibitors CD4Integration Rev. Transcription Maturation Fusion NRTI & NNRTI Fusion Inhibitors PIs Figure 1-4. Drug targets in th e HIV-1 virus life cycle.

PAGE 33

33 HIV-1HIV-2 ABCDFGHCRFs AEAG Group MGroup NGroup OVirus Group Subtype CRFsHIV-1HIV-2 ABCDFGHCRFs AEAG Group MGroup NGroup OHIV-1HIV-2 ABCDFGHCRFs AEAG Group MGroup NGroup OVirus Group Subtype CRFs Figure 1-5. Schematic phylogene tic tree of HIV virus. vifvpunef vpr tat rev LTR 3 Env LTR 5 pol gag MANC CAp 6PRIN RT SU TM A BEG H K U (Unclassified)CRF01_AE CRF03_AB CRF02_AG CRF04_cpx vifvpunef vpr tat rev LTR 3 Env LTR 5 pol gag MANC CAp 6PRIN RT SU TM A BEG H K U (Unclassified)CRF01_AE CRF03_AB CRF02_AG CRF04_cpx Figure 1-6. Circulant recombinant forms (CRF) of HIV-1 virus, gr oup M. Each of the colors in the figure represents a different subtype of the virus.

PAGE 34

34 A B Figure 1-7. X-ray crystal structur e of HIV-1 protease bound to CA/p2 cleavage site (PDB 1nqs). In red, sites P1-P1; in dark orange, sites P2 -P2; in light orange, sites P3-P3 and, in green, sites P4-P4.A) Front view of HIVPR. B) Top view of HIVPR. B1 B1 D2 D2 C2 C2 D1 D1 C1 C1 A2 A1 A1 A2 B1 B1 D2 D2 C2 C2 D1 D1 C1 C1 A2 A1 A1 A2A B B1 B1 D2 D2 C2 C2 D1 D1 C1 C1 A2 A1 A1 A2 B1 B1 D2 D2 C2 C2 D1 D1 C1 C1 A2 A1 A1 A2A B Figure 1-8. HIV-1 protease struct ure. A) Left, cartoon representa tion of HIV-1 protease in the closed conformation (inhibitor was omitted for clarity). B) Right, space filling model of residues within 5 of th e active site. The flap region is indicated in blue. (PDB 1HVI). 1990199520002005 0 20 40 60 80 100 120 140 # PublicationsYear Figure 1-9. HIV-1 protease publication record/year.

PAGE 35

35 0102030405060708090100 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 NOEResidueBound Free 0102030405060708090100 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 NOEResidueBound Free Figure 1-10. HIV-1 protease NOE values. NOEs in red correspond to the inhibited form of HIV1 protease and, NOEs in blue correspond to the uninhibited form. [ Values obtained from Ishima, R.; Freedberg, D. I.; Wang, Y. X.; Louis, J. M.; Torchia, D. A. Structure 1999 7 1047-55 and Freedberg, D. I.; Ishi ma, R.; Jacob, J.; Wang, Y. X.; Kustanovich, I.; Louis, J. M.; Torchia, D. A. Protein Sci. 2002 11 221-32]

PAGE 36

36 CHAPTER 2 INTRODUCTION TO EPR Introduction to Spectroscopy Spectroscopy is classically defined as the study of the absorption and emission of light and other radiation by matter, as related to th e dependence of these processes on the wavelength of the radiation ., and it is classified by the different types of wavelengths at which it is operating. Figure 2-1 shows a graphical represen tation of the electromagnetic spectrum as a function of wavelength, showing the various bands of EPR operation. Electron Paramagnetic Resonance (EPR) is a magnetic resonance technique that studies systems with unpaired electrons and their intera ctions in an external applied magnetic field75,76. In the simplest case, for a free electron in a ma gnetic field, the electron magnetic moment aligns itself parallel or antiparallel to the applied magne tic field, therefore breaki ng the spin degeneracy in what is known as the Zeeman Effect77. The difference between the energy levels is proportional to the magnitude of the applied fiel d and the spectroscopic g-factor, or Land gfactor (Figure 2-2). eEhgB (2-1) This proportionality constant is known as the Bohr magneton e, defined as 2e ee m (2-2) In Equation (2-2), e is the electric charge carried by a proton (1.6021 19 C), is the Plank constant divided by 2 (1.054 34 Js) and me is the electron rest mass (9.109 kg). The value of e is 9.274-24 JT-1.

PAGE 37

37 The resonant condition, a common denominator in all forms of spectroscopy, occurs when the applied energy ( E) is equal to the splitting between levels. In the particular case of continuous-wave EPR (cw-EPR), the resonant c ondition is achieved by varying the magnetic field at a constant frequency. As it can be seen in Figure 2-3, whic h shows block diagrams for both a cw-EPR spectrometer and a typical optical spectroscopy sp ectrometer, the basic layout of a spectrometer is similar regardless of the wavelength used. Th e main differences are the radiation source, the sample cell and the detector used. The Static Spin Hamiltonian The spin Hamiltonian describes the energies of the different states within the ground state of the paramagnetic species (in our case, ni troxides). We can write a generalized spin Hamiltonian for a system with electron spin S and k nuclei with spin I as: 0kk,0kkkk 111/2kEZZFSHFNZNQNN ik eNnkik kkIikg B g SSDSSAIBIIPIIdI (2-3) In Equation (2-3), the followi ng terms can be distinguished:EZ: electron Zeeman interaction, Z FS: Zero-field splitting, H F: Hyperfine interaction,NZ: nuclear Zeeman,NQ: nuclear Quadrupole and NN: nuclear-nuclear spin interaction. Figure 2-4 represents the typical energies of the electron and nuclear sp ins interactions in the Hz-THz range. Electron Zeeman Interaction The interaction between the electron and the a pplied external magnetic field B is described by the electron Zeeman term. EZe0B g S (2-4)

PAGE 38

38 Where e is the Bohr magneton, B0 is the external magnetic applied field, g is the spectroscopic splitting factor, S is the total spin of the system under study and is the reduced Planks constant. It can be shown that the g tensor can be diagonalized by m eans of a similarity transform and expressed in terms of its principal values gxx, gyy and gzz. 1 () diag EZexxxxeyyyyezzzzgHSgHSgHS LgLg (2-5) The value of g for the free electron is 2.0023 ( it deviates from the theoretical value of 2 due to spin-orbit coupling effects) In organic radicals tumbling isot ropically in solution, only the average value is observed. Hyperfine Interaction The hyperfine interaction represents the coupling between an electron and a nuclear spin. It can be written as: 1 HF k kkSAI (2-6) where S is the vector for the spin system under consideration, A is the hyperfine interaction matrix and Ik is the vector for the nuclear spin system k, and the index k in Equation (2-6) runs over all nuclei. The hyperfine interaction H F in Equation (2-6) can be expressed as a sum of an isotropic component (known as Fermi Contact term, orF) and the anisotropic part given by the dipolar interaction between the electron and the nuclear spin ( D D). H FFDD (2-7)

PAGE 39

39 In 1930, Fermi showed that, for systems with one electron, the isotropic part of the hyperfine interaction can be written as78: 2 08 0 3Fisoxxyyzz isoeeNNaISISIS agg (2-8) where 2 00is the electron spin de nsity at the nucleus, ge is the electron spectroscopic splitting factor, gN the spectroscopic splitting factor fo r the nucleus under consideration, e is the Bohr (electron) magneton and N is the nuclear magneton (9.274-24 JT-1). In the high field approximation, which assumes that the perpendi cular components of the applied external magnetic field are negligible in compar ison to the perpendicular component (Bz>>Bx, By), the Fermi Contact term further reduces to: FisozzaIS (2-9) If we now consider the dipolar co mponent of the hyperfine interaction D D, the quantum analog of the classical dipolar interaction between the electron and the nucleus can be readily derived from Equation (2-10)79. 533E rr eN eN r r (2-10) where e and N are the electron and nuclear magnetic mo ments, and r is the distance that separates the electron from the nucleus. The co rresponding quantum mechanical expressions in Equation (2-11) an be substituted by their classical analogs in Equation (2-10): ee NNg g e N S I (2-11)

PAGE 40

40 0 533 4DDeeNNgg rr SrIr SI (2-12) Expanding now the productsSI ,Sr and Ir: x xyyzz xyz xyzSISISI SxSySz I xIyIz SI Sr Ir (2-13) Substituting Eq. (2-13) in Eq. (2-12), and rewriting the expression in matrix form: 22555 52255 55225333 333 333 x D DeNeNxyz y zrxrxyrxzr I ggSSSxyrryryzrI I xzryzrrzr STI (2-14) Lets consider now the partic ular case of a single unpaired electron interacting with a nucleus of spin I=1. This example, for which th e assumption of isotropic g and A will be made, can be used to obtain a simplified model for the inte ractions in spin label nitroxides. Spin labels that contain the nitroxide moiety present an unpaired electron (Ms=1/2) coupled to a 14N nucleus (MI=1). Therefore, a simplified Hamiltonian of the form of Eq. (2-15) can be written. ezNNzisozzHgBSgBIaIS (2-15) The coupling between the electron and the 14N nucleus given isozzaIS splits each of the Zeeman interaction levels into three di fferent energy levels corresponding to MI=0, From the diagram in the Figure 2-5, the energi es corresponding to the different levels can be computed by taking into account the Zeem an and the hyperfine interaction terms. s I M M (2-16)

PAGE 41

41 We can represent the spin states in Equation(2-16) for Ms =1/2 and MI=1 by Equation (2-17). 11 ,1,1 22 11 ,0,0 22 11 ,1,1 22 (2-17) We now apply the Hamiltonian of E quation (2-15) acting upon each of the bra of Equation (2-17) to obtain the energy levels in Equation (2-18). 1010 ,1,1 22 11 ,0,0 22 1010 ,1,1 221111 2222 11 22 1111 2222eNNeNN ee eNNeNNEgBgBAEgBgBA EgBEgB EgBgBAEgBgBA (2-18) As we shall see in the next section, the spect roscopic selection rules indicate that the only allowed transitions between different energy levels are those that follow Ms=1, MI=0. 110 00 110 e e eEgBA EgB EgBA (2-19) where 0 isoeeAag and a0 is a constant known as the hyperfine splitting. Transition Selection Rules in EPR Spectroscopy As previously mentioned, Figure 2-5 shows the EPR lineshape corresponding to the three allowed transitions for a system with Ms=1/2 and MI=1. Here, we will briefly analyze the origin of the selection rules and the observed transitions.

PAGE 42

42 Suppose that an oscillating magnetic field 1'2cos HHt is applied in the x direction. The transition probability depends on the integral ,, f fxiiSISSI. In this expression, Sf and If correspond to the final electronic and nuclear spin states, and Si and Ni to the initial electronic and nuclear spin states. The electron resonant transitions are caused by the interaction of the oscillating magnetic field with the magnetic mo ment of the electron spins, and as such the nuclear spins I (nuclear transitions) can be i gnored. Consider a system for which S=1/2 and I= 1/2. The allowed electronic and nu clear spin states are, therefore, Ms=/2 and MI=/2. The positive +(1/2) states will be denoted as e and N, whereas the negative (1/2) states will be denoted as e and N. The elements of the integral ,, f fxiiSISSI can be evaluated, leading to the following selection rules: 1 2 111 0 222 11 0 22 1111 2222 1111 2222x eNeNeeNNee eNeNeeNN eNeNeeNNee eNeNeeNNeeSSS SSS SSS SSS SSS (2-20) The non-vanishing terms in Equation (2-20) are those corresponding to the transition selection rules of Equation (2-21). 10sImm (2-21) Nuclear Zeeman Interaction In an analogous manner to the electron Zeeman interaction, the corresponding term for the interaction between the nuclear spin an d the applied external magnetic field B can be written as:

PAGE 43

43 1NZNnk kg0kBI (2-22) where the nuclear spin quantum number I and the nuclear g-factor ar e nuclei-dependant, and cover the range of 1/2 I 6 and 0.097 gN 5.58. In most EPR experiments, NZ has little influence on the EPR spectrum due to th e difference between the electron magneton e (9.274-24 JT-1) and the nuclear magneton N (5.051-27 JT-1) Nuclear Quadrupole Interaction The non-spherical charge distribu tion present on nuclei with I 1 can be described by a nuclear electrical quadrupole moment Q In its general form, kkk 1/2kNQ I IPI (2-23) where P is the nuclear quadrupole tensor. In EPR, nuclear quadrupole co ntributions represent small second-order effects, which are difficu lt to observe. For a deta iled description of quadrupole interactions, the reader is referred to the general NMR literature, such as Refs. 80,81 and references therein. Zero-Field Splitting In spin systems with S>1/2, the dipole-dip ole coupling between electron spins removes the spin degeneracy of the ground st ate in the absence of an applie d magnetic field. The Zero-field splitting term of the Hamiltonian can be written as ZFS SDS (2-24) For nitroxide spin labels S = 1/2, and theref ore do not present ZFS effects. ZFS is often important in EPR spectroscopy of metals, whic h often present spin systems with S>1/2. A detailed description of the effects of ZFS on th e electronic structure and magnetic properties are beyond the scope of this introduction and can be found elsewhere.

PAGE 44

44 Systems with More Than One Interacting Electron Spins For the particular case of systems which contai n more than one interacting spin, the spin Hamiltonian can be written as: 1212S,S D DexchSS (2-25) In Equation (2-25), H(S1) and H(S2) represent the spin Hamiltonian for each of the noninteracting spins S1 and S2. The terms D D (dipole-dipole interaction) and exch (spin exchange) account for the interactions between S1 and S2. Dipolar Interaction For the particular case of the interaction betw een two electron spins, we can further reduce Eq. (2-12) to 2 0 121212112212 32 121213 4DDegg rr SDSSSSrSr (2-26) We can expand now the products S1S2 and Sr12 in an analogous manner as in Eq. (2-13): 121212 111 222 x xyyzz xyz xyzSSSSSS SxSySz SxSySz 12 112 212SS Sr Sr (2-27) Rewriting Eq. (2-26) in matrix form: 22555 2 252255 0 12111 2 55225 2333 333 4 333 x D Dexyz y zrxrxyrxzr S ggSSSxyrryryzrS S xzryzrrzr STS (2-28)

PAGE 45

45 To obtain a simplified expression for the dipola r interaction we will consider the high field approximation, for which the terms containing co mponents in the x and y planes vanish. Making the substitution cos zr in Eq. (2-28) under the high field approximation: 2 2 0 1212 33cos1 4 D DezzggSS r (2-29) Equation (2-29) shows the dependency of the dipole-dipole interaction with the relative orientation between both dipolar vectors () and the distance between them (r3). In order to aid in the explanation of pulsed EPR experiments, a more rigorous and general expansion can be made of Equation (2-28) by transforming the full T tensor from Cartesian to Polar coordinates, as follows: sincos sinsin cosxr yr zr (2-30) The operators Sx and Sy can be written in terms of the ladder operators I+ and I12 12xy xy x ySSiS SSiS SSS SSS (2-31) Substituting in the dipolar Hamiltonian of E quation (2-28), the following expression is obtained, which is known as dipolar alphabet (Equation (2-32))79. In Equation (2-32), term B corresponds to zero quantum coherences (ZQ), terms C and D to single quantum coherences (SQ), and terms E and F to double quantum coherences (DQ). At this point, it is useful to introduce the concept of density matrix. The density matrix of a quantum mechanical system is defined as a He rmitian matrix of trace one that describes the statistical state of a system.

PAGE 46

46 2 12 3 2 12 2 1212 1212 1212 22 12 22 1213cos 1 13cos 4 3 sincos 2 3 sincos 2 3 sin 4 3 sin 4e DD zz i zz i zz i igg ABCDEF r ASS BSSSS CeSSSS DeSSSS EeSS FeSS (2-32) In order to illustrate the different trans itions associated with particular quantum coherences, the density matrix for a two-electron system can be written. (Figure 2-6). Figure 2-6 shows the energy level populations in th e diagonal elements of the matrix ( P ), as well as the Single, Double and Zero quantum cohe rences as off-diagonal elements. Exchange Interaction Exchange interaction (or Heisenberg spin ex change) appears when both interacting spins are close enough to each other for their orbitals to overlap (below 15 in solids, and below 7-8 in solution). In general, the exchange Hamiltonian can be written as: 12 exchSJS (2-33) with J being the Heisenberg spin exchange matrix. Previous attempts of using Hexch to determine distances below 7 showed a strong dependence of the lineshape on the orbital overlap and therefore, a general distance ruler cannot be developed accurately82.

PAGE 47

47 Introduction to Pulsed EPR Density Operator Formalism Pulsed EPR experiments are based on time-depe ndant manipulations of a particular spin system. These experiments can be described as a sequence of pulses and time intervals characterized by different time-evolved Hamiltoni ans. As previously mentioned, the density matrix of a quantum mechanical system describe s the statistical state of the system. In this section we will introduce a more rigorous explana tion of the density matrix, and its usefulness in understanding pulsed magneti c resonance techniques. We will start by expanding a wavefunctio n in a basis set, in Equation (2-34) 1N i itcti (2-34) The density matrix can be de fined as the product of the ket and the bra expansion (Equation (2-35)) 11 NN ij ijtttctctij (2-35) For the particular case of a mixed state, we need to introduce the population ensemble of Equation (2-36). 11 kkk k NN kkk ij kijtptt p ctctij (2-36) where k denotes the number of elements, and p the weight of each of the elements in the ensemble. The time evolution of the density matrix at a pa rticular time can be written as a product of time-independent propagators and the initial density matrix 0 of Equation (2-37)

PAGE 48

48 21120NNtUUUUUU (2-37) where UN represents the propagator N in the time sequence. The evolution of a particular operator A under operator B can be written as: iieeBBAC (2-38) Equation (2-38) can be expanded using the Baker-Hausdorff formula as follows83,84: 23,,,,, 2!3!iii eeAiBABBABBBA BBA (2-39) We can further simplify this expression grou ping even and odd terms with the relation [B,[B,A]]: 24351, 2!4!3!5!iieeAiBA BBA (2-40) which, in turn, can be simplified in terms of sine and cosine functions cos,siniieei B BBAACABA (2-41) For the particular case in which B = A (Equation (2-42)) iiee B BB BAAAAA (2-42) In the case of a spin Hamiltonian for which the perturbation operators are time independent, the operators can be applied consecutively: 1122 NNN B BBAC (2-43) In the particular case of electr on spin operators, such as those which will be considered for the description of distance measurements usin g Double Electron-Electron Resonance (DEER) or Double Quantum Coherence (DQC), these operators (Sx, Sy and Sz) belong to a subgroup termed cyclic commutators, which obey the rules in Equation (2-44).

PAGE 49

49 ,;, ,;, ,;, x yzyxz yzxzyx zxyxzySSiSSSiS SSiSSSiS SSiSSSiS (2-44) The cyclic commutator property is very useful to calculate the time propagation of the spin density operator. For example, we can study wh at is the result of a spin population (using Sz as the equilibrium spin state) evolving under a +Sx /2 pulse ( =90) as shown in Equation (2-45). 2cos(),sin() 22 (1)xS zzxz y ySSiSS iiS S (2-45) Double Electron-Electron Resonance Double electron-electron resonance (DEER)85 is a pulsed EPR techni que that can separate dipolar electron-electron couplings from other interactions in th e Hamiltonian. The basic DEER sequence consists of a twopulse echo-forming sequence ( /2) separated by a constant interpulse delay ( ) at a frequency 1 and a pump pulse ( ) at a frequency 2 with a variable time t with respect to the first observer pulse. The first pulsed ELDOR sequences developed were the three-pulse DEER and the 2+1 pulse D EER, and can be seen in Figure 2-7. A technical limitation of the conventional 3pulse DEER as well as the +1 experiment86 is that the experimental deadtime (appr oximately 64ns) does not allow accurate signal measurements for broad electronelectron couplings. (Deadtime is defined as the time after an event for which the system can not record data). In order to avoid this experimental deadtime problem, Pannier et al. introduced a 4-pulse version of DEER87. This was accomplished by

PAGE 50

50 introducing an additional refocusing ( ) pulse in the observer freque ncy, which allows the access to the first few ns of the dipolar evolution. Figure 2-8 shows the dipolar evolution traces for a broad distance di stribution (upper) and a monoradical in solution (TEMPO, lower). It can clearly be seen that the most prominent spectral features are hidden under the deadtime of the conventional 3pulse DEER (shadowed region). Currently, 4-pulse DEER is the most common pulse sequence used for distance measurements by pulsed EPR (Figur e 2-9). However, other pulse sequences to measure distance distributions based on double quantum c oherences (DQC) have been developed.88,89 In order to better understand the 4p-DEER se quence, Figure 2-10 shows the evolution of the magnetization of the spin subsets A and B as a function of the position in the pulse sequence. Figure 2-10 shows the pulse sequence utili zed for deadtime-free double electron-electron resonance (DEER). In the DEER experime nt, the observer pulse at the frequency (1)mw is applied at the frequency of the central resonance. The pulse sequence 112 generates a Hahn spin echo at position 21 .After a time evolution the following pulse refocuses the subset of spins A at a time after the pulse. The dipolar evoluti on can be studied by introducing an additional pulse at the microwave frequency(2)mw This additional pulse flips the spin subset B, affecting the local magnetic field of spin subset A. Because both subsets of spins are coupled through the dipolar term in the Hamiltonian (Eqn.( 2-46)), the electron spin echo oscillates with the dipolar coupling frequency ee as a function of the interpulse distance t. Figure 2-11 shows the positioning of the pump pu lse with respect to the observe pulse in the field-swept echo experiment. It can be seen that the pump pulse is positioned in the central

PAGE 51

51 resonance line due to its intensity, as a larger population of excited spins will produce a better S/N ratio. It is important to remember th at for proteins that contain two spin labels with similar g values, the spectral features of each of the indivi dual labels overlap, and selective excitation of one particular label is not feasible. Therefore, the observer pulse is posi tioned in the second most intense feature of the spectra, namely, the low field line, which, for nitroxide spin labels, lays ~26 G (72MHz) below the central resonance line. In order to describe the DEER experiment a simplified spin Hamiltonian that only includes the Zeeman terms for the spin subsets A and B, as well as the dipolar coupling between A and B can be written as87,90: ABAB ABAzBzeezzSSSS (2-46) Utilizing the density operator formalism, and propagating the time evolution of the density matrix in the DEER sequence, we arrive at the following expression for the DEER signal cosDEEReeVtt (2-47) where ee modulates the refocused spin echo and can be written as: 2 2 0 31 3cos1 4ABe eeddAB ABgg JJ r (2-48) dd represents the dipole-dipole coupling and AB corresponds to the angle between the static field and the vector connecting both dipoles. For a non-magnetically aligned frozen sample, the angle AB takes all possible orientations due to the existence of macroscopic disorder in the protein system, and therefore the angular dependence can be averaged out.

PAGE 52

52 T1 and T2 (Tm) Measurements In order to set-up the DEER experiment, th e spin-lattice and spin-spin relaxation times must be known. Hence, the following section prov ides a brief introduction to the concepts of T1 (spin-lattice) and T2 (spin-spin) relaxation, the implications for the DEER experiment as well as experimental techniques for the measurement of these values. Spin-lattice relaxation time (T1) In order to describe the spin-lattice rela xation time, the description provided by the Bloembergen-Purcell-Pound theory (BPP)91 will be followed. Consider a two-level system, with I=1/2 for which the population of th e lower level is expressed as 1 2N and the population for the higher level is expressed as 1 2N. We can therefore write the surp lus population of the lower level over the higher level as in Equation (2-49). 11 22nNN (2-49) Let n0 be the spin populations at thermal equilibri um at the temperature of the lattice. In the absence of an applied field (whether rf or mw), we can monitor the evolution of the surplus population (Equation (2-50)). 101 dn Tnn dt (2-50) The constant for the decay process of the spin magnetization in the z-axis towards the equilibrium is known as spin-lat tice relaxation time, or longit udinal spin relaxation time (T1). From this relaxation phenomenon, we can now investigate the e ffects of applying a radiation field to the system. The transition between two states im and 'im is given by the expression in Equation (2-51).

PAGE 53

53 '2 32'83iiii mmWhmMm (2-51) with M being the magnetic transition moment and the energy density. In this model we will work under the assumption of a normalized absorption line g and isotropic radiation density written as: 2 134 Hg (2-52) where H1 in Equation (2-52) repres ents the effective magnetic field inside the EPR cavity. Combining Equations (2-50) and (2-51), the time evolution of th e system can be written as Equation (2-53). 1011 2212 dn TnnnW dt (2-53) Substituting Eqs. (2-51) and (2-52) into Equatio n (2-53) and in equilibrium conditions for which 0 dn dt the expression in Equation (2-54) can be obtained. 1 22 0111 1 2 snnHTg (2-54) where s is referred to as the saturation factor and as the gyromagnetic ratio After a phenomenological description of the spin-lattice relaxation time T1, we can proceed to examine the different experimental techniques for T1 determination and its advantages and pitfalls. The two most commonly used techniques to measure T1 relaxation rates are using inversion recovery pulse sequences or saturati on recovery techniques. Due to the somewhat specialized instrumentation needed to perform saturation recovery, only the inversion recovery pulse sequence will be discussed here. Noneth eless, it is noteworthy to mention that T1 measurements obtained by saturati on recovery tend to be more precise due to broad excitation

PAGE 54

54 pulses and access to spectral diffu sion levels that are otherwise inaccessible to pulsed techniques which utilize harder pulses. The determination of T1 relaxation rates using conventiona l pulsed-EPR instrumentation is relatively straightforward. The T1 constant can be determined by the pulse sequence /2observe which measures the integrated signal intensity as a function of the interpulse delay time The main advantage is that the hard pulses used in pulsed-EPR excite all spins within a several Gauss window, therefore improving th e signal/noise ratio. Nonethele ss, the short pulses used do not saturate spin levels accessible by spectral diff usion, strongly impacting the recovery curve. Data analysis of T1 experiments (Figure 2-12) is perfor med by least squares fitting to an exponential function of the form provided by Equation (2-55). 011exp y AxxT (2-55) In practice, for double electronelectron resonance experiments, it is not necessary to know the precise value of T1. As it was mentioned before, T1 relaxation determines the time for the spin magnetization to return to equilibrium. Ther efore, for pulse sequences in which the signal is collected as a function of a particular interpulse delay, the value of T1 will determine the waiting time between consecutive experiments (also known as Shot Repetition Time (SRT) or Recycle Delay). Commonly, the Shot Repetition time is quickly determined by estimating an SRT 5* T1. Experimentally, an estimation of the SR T can be obtained by performing a Hahn echo experiment for which the intensity of the echo as a function of SRT is monitored. The optimal value for the SRT is obtained when increasing th e SRT will not increase the intensity of the echo anymore.

PAGE 55

55 Spin-spin relaxation time (T2 or Tm) In an analogous manner to T1, which was defined as the characteristic decay time along the z-axis of the magnetic field, T2 is the characteristic decay time for the transverse components Mx and My after the application of a microwav e or RF pulse. For this reason, T2 is also known as transverse relaxation time or spin-spin relaxation time In the absence of inhomogeneous broade ning effects, such as magnetic field inhomogeneities and other extraneous processes that might contribute to spin-spin relaxation, T2 can be seen as a measurement of the width of the spectral lineshape. In practice, the spin-spin relaxation description is very complex, due to the many factors involved in the process. The factors included in this dephasing include local spin concentrations, li brational motions, dynamic averaging of inequivalent nucle i and dipolar interactions with nuclear spins. Thus, the experimental time obtained from a Hahn echo experiment is not T2, but rather Tm, known as phase memory The phase memory includes all processes that contribute to homogeneous and inhomogeneous spectral broadening. The analysis of spin echo decays can be performed by fitting experimental data to a stretched exponential 2 2exp x mE T (2-56) where is the time between pulses, Tm is the dephasing time constant (phase memory) and x is a phenomenological parameter that depends on th e dephasing mechanism (values of x range between 0.5 and ~2.5). The experimental data for Tm is obtained using a pulsed Hahn echo experiment with a /2-observe sequence. In this sc heme, shown in Figure 2-13, the refocused echo intensity or the echo area is meas ured as a function of the interpulse delay It is

PAGE 56

56 noteworthy to mention that spins whose precessi on frequency do not remain constant during the length of the experiment (2 ), the second pulse ( ) does not exactly reverse the precession that occurred during the first part of the pulse se quence, and therefore doe s not contribute to echo formation. Table 2-1. Parameters used in Tm measurements Parameter Value /2 16 ns 32 ns 200 ns 8 ns PG ~130 ns Time Base 4 ns Bandwidth 20 MHz SRT 2000-4000 s Table 2-1 shows common parameters used in the determination of Tm. represents the interpulse distance, represents the increment of th e pulse distance in each successive experiment, PG is the width of the integrator ga te and the Time base is the integration time resolution. As previously menti oned in the text, the Shot Repeti tion Time (SRT) should be set to approximately 5*T1. In order to maximize the longest distance available to the DEER experiment, as well as to obtain a good S/N ratio, measurements of Tm are required. The value of Tm determines the maximum length after the initial Hahn echo generated by the observer frequency ( /2-echo) for which the intensity of the refocused echo will be measurable. To obtain a numerical value for Tm, the data collected from the spin echo decay experiment can be fitted to a decreasing exponential function, as show in Figure 2-14. The small oscillations that are visible at short times (<3 s) correspond to solvent protons m odulating the spin echo, in what is known as ESEEM (Electron Sp in Echo Envelope Modulation).

PAGE 57

57 In order for the refocused echo to be measurable, Tm should be at least 0.5 s. As a reference, in order to obtain di stances of 2 nm from the analysis of the dipolar evolution curves obtained in the DEER experiment, Tm should reach values around 1 s, and distances >5nm are only accessible for Tm>5-6 s. Pulsed EPR Instrumentation Those readers familiar with nuclear magnetic resonance instrumentation will find great similarities amongst NMR and EPR. The main diffe rence is that, in NMR, the magnetic field B is held constant, whereas in EPR the frequency is held constant and the field is swept. Figure 215A shows a block diagram for an NMR system and Figure 2-15B shows the block diagram for a pulsed-EPR instrument. The basic pulsed-EPR instru mentation consists of a microwave generating source (Klystron or Gunn diode), a pulse shaper and a set of microwave pulse forming units which control the position, length and phase of the pulses, a traveling wave tube amplifier (TWT) in order to increase the powe r of the applied pulses (from approx. 2 mW to ~1 kW output power), a microwave circulator and the sample cavity. In the receiver side, a signal am plifier, the receiver, an analog-to-digital c onverter (ADC) and phase adjustment controllers (Re and Im) are found. There are several different resonators that can be used for pulsed-EPR spectroscopy, being the most used ones the MS-2 and the MD-5 re sonator. The MS-2 resonator is a split-ring resonator with a 2 mm OD sample size that accommodates samples up to 10 L. The MD-5 resonator is a dielectric resonator that accommodat es the larger 5 mm OD tubes, with a volume of approx. 100 L. It has been shown that, when sample volume is not an issue, the S/N ratio in the MD5 resonator scales as a func tion of sample volume, up to 100 L, providing a better S/N ratio in comparison to that of the MS2.92

PAGE 58

58 Data Analysis Spectral Parameters from Cw-EPR Lineshapes In order to quantitatively compare different cw -EPR spectra, a series of spectral parameters must be calculated. The most common parameters us ed in EPR are (1) the peak to peak width of the central resonance in the nitroxide spectrum ( Hpp), (2) the ratio of the intensities of the low and the center field (LF/CF), (3)the spectral second moment (

) and in particular cases, (4) the spin label correlation time c, which can be obtained from spectral simulations93,94 Figure 2-6 shows the three most comm on spectral characterization methods. Figure 2-16 A) shows a graphical repres entation of the determination of Hpp from a cwEPR lineshape. As the spin label and/or the protein backbone becomes more immobilized, the central resonance of the spectrum becomes broa der, and the low-field peak split into two different peaks. Therefore, Hpp can be correlated with the b ackbone/label mobility; the broader the central resonance line, the higher the value of Hpp The determination of the width of the central resonance is tightly linked to the protein of choice. In order to compare results obtained in different spectrometers, as well as different samples, a parameter known as scaled mobility is used. The scaled mobility is simply a normalization of the central resonance linewidth us ing the linewidths of the most mobile and the most immobilized proteins characterized by EPR to date. 11 exp 11 i s miM (2-57) where i, m and exp are the most immobilized, mo st mobile and experimental Hpp. The ratio between the intensities of the center-fie ld resonant line and the low-field resonant line (Figure 2-16B) can be utilized as a measure of spin label/backbone mobility. As the mobility

PAGE 59

59 decreases, the cw-EPR spectrum becomes broa der and the intensity of the low-field line decreases faster than the inte nsity of the central linewidth, making the ratio LF/CF smaller. Last, the analysis of the cw-EPR spectral second moment75,79,80 (Figure 2-16C) is not always straightforward, and a very precise ba seline correction of th e absorption spectra is needed. In the case of a symmetric lineshape, the selection of the center field is trivial as it coincides with the maximum of the absorption lin eshape. For the particular case of asymmetric lineshapes, a correction factor must be included to account for the asymmetry of the lineshape. In the general case, the nth moment of a resonant absorption lineshape can be defined as in Equation (2-58) 1 0 1 n m jj n jj jHH HHHy A (2-58) where Hj Hj-1 is the step size, Hj is the field value for a given point j H0 is the proper choice of center field for which the first moment vanishes and yj is the intensity at point j For the appropriate choice of H0 in an asymmetric lineshape, the first moment of the spectral lineshape vanishes. 0 1 1 10m jj j m j jHHy HB y (2-59) Therefore, the true nth moment of a nonsymmetrical spec tral lineshape can be obtained from subtraction of the first moment (Equation (2-59)) from Equation (2-58): 1 0 1 n m jj n jj corr jHH HHHBy A (2-60) As we mentioned before, immobilized bac kbone/spin label is reflected in broader lineshapes, and therefore, higher second moments.

PAGE 60

60 Double Electron-Electron Resonance Data Analysis As we stated previously, the 4p-DEER sequence is able to separate the dipolar interaction between two spins by using a dipolar-modulated refocused Hahn echo from the rest of the interactions in the spin Hamiltonian. Nonethel ess, the dipolar evolu tion signal is not only influenced by specific A-B spin pair interacti ons, but also by the bulk di stributions (nonspecific interactions) of spins in the sample. Therefore, the DEER signal can be written as a product of the specific intramolecular and the isotropic non-speci fic intermolecular dipolar interaction. intraintercosexpDEER eeBVVV tkCFt (2-61) The intermolecular contribution to the signa l has the form of the decaying exponential function of Equation (2-62) inter 22exp 8 93B eABVtkCFt gg k (2-62) where C denotes the spin concentration, FB the fraction of spins excited by the pump pulse and k is a constant that includes th e Bohr magneton constant and th e spectroscopic g-factors for both electrons. Given a long phase memory time (Tm), subtraction of the decaying exponential background can be easily performed by fitting a decaying exponential or a polynomial function to the region for which dipolar oscillations are negligible. Figure 2-17A shows the contribution of the background to the dipolar evolution in the DEER experiment. If we consid er the dipolar evolution sign al after subtraction of the background corresponding to nonspeci fic intermolecular dipolar c ouplings (Figure 2-17B), the dipole-dipole interactions can be written as in Equation (2-63) max min,R RVtrtPrdr (2-63)

PAGE 61

61 The kernel adopts the form of Equation (2-64) 2 1 2 3 0,cos13coserttd r (2-64) Inversion of this Fredholm equa tion of the first kind requires the use of a discretized form of the integral equation, which can be written in its matrix form as: S=KP (2-65) where S is the experimental signal, K corresponds to the kernel and P to the distance distribution associated to S. Because inversion of V(t) is an ill-define d problem (which means that there might be multiple functions P(r) that yield a predicted V(t) that agrees with the experimental V(t)), we have used two different met hods to obtain a reliable P(r). The first approach, termed Tikhonov regularization (TKR)95-97, is based on the construction of a functional of the form: 22 2PKP-SLP (2-66) In this equation, K represents the kernel in matrix form, P is the distance distribution, S corresponds to the experimental dipolar evolution signal, is known as th e regularization parameter and L, known as Tikhonov matrix, is us ually the identity operator I when preference to solutions with small norms is pref erred, or the second derivative operator 2 2r to enforce smoothness for those cases for which P is known to be continuous. By minimizing P a solution P that depends on the regularization parameter can be found. The first term in the func tional calculates the difference be tween the experimental and the

PAGE 62

62 generated signal corresponding to a given P (KP product), whereas the s econd term ensures a smooth P function. In order to obtain the optimal regularization pa rameter, several methods have been applied. One of these methods relies on numer ically minimizing the functional P as a function of using steepest-descent or Newton-Raphson methods The main inconvenience of this technique is that global minimization techniques are highl y sensitive to the star ting conditions, and often present local minima in the parameter hypersurf ace. The most widely used method is known as the L-curve criterion. In the L-curve plot, the residual norm KP-S is plotted vs. the residual norm LP. The corner of the L-curve represents the balance in ke eping both norms small. This corner point can be eas ily found by eye inspection, or by s earching for the point of maximum curvature, in what is known as the -Curve98. A second method that can be used to extract distance distributions fr om DEER signals is based on a Metropolis Montecarlo approach92. In this case, a distan ce distribution shape is assumed (e.g. Gaussian). Starting from an initial guess, the dipolar evolution is calculated for a particular distance distribution, and the differe nce between experimental and calculated dipolar evolution is calculated. The pr ocedure is iterated modifying th e position and the width of the generating function until a minimum is found. Th e advantage of Montecarlo methods is that given an analytical form of the kernel that is in expensive to compute, they are very time efficient and therefore allow for a bigger sample of the parameter space, overcoming the difficulties found in conventional minimization methods (local minima). To summarize, Tikhonov regularization requir es the calculation of a regularization parameter to ensure solution smoothness, but does not assume a particular shape for the distance distribution, whereas Montecarlo methods provide an exhaustive exploration of the

PAGE 63

63 solution parameter space, but are constrained to a specific shape for the distance distribution. Figure 2-18 summarizes the different data analysis algorithms for the DEER dipolar evolution signal. To obtain an initial estimate on the effect of the distance distribution on the dipolar evolution signal, an existing Matlab program created by the group of Dr. P. Fajer (NHMFL, Tallahassee), was modified in order to simulate dipolar evolution from a given set of distances. Currently this program, named D eerSim, uses a set of Gaussian or Lorentzian distance distributions with known width, breadth and percentage population. This program has also been recently modified to use custom user-generated di stance distributions as an input (for example, results of molecular dynamics trajectories). Figur e 2-19 shows a screenshot capture for DeerSim. Using this software, we performed stability calculations for the two most common methods to obtain distance distributions from dipolar evolution functions namely Montecarlo minization and Tikhonov regularization. As it was stated previously, Montecarlo methods assume a known function for the distance distribution (eg. Gaussian, Lorentzian), whereas Tikhonov regularization methods do not assume a particul ar function but a regularization (smoothing) parameter must be chosen. In order to study the performance of both Montecarlo and Tikhonov regularization, we generated a dipolar evolution f unction based on one single Gaussian function at different average positions (for parameters, see Table 2-2). In or der to study the effect of noise on the distance distribution solution, noise was added to each of the dipolar evolution function in increased percentages. Figure 2-20 shows the results the data generate d for a 1-Gaussian distance distribution profile, and the resu lts of analyzing the distance di stribution by either TKR of MC. From the generated distance dist ribution, we utilized DeerSi m to reconstruct the dipolar

PAGE 64

64 evolution signal. To analyze the effect of noise on both algorithms, the amount of noise in the signal was increased, and both MC and TKR met hods were utilized to obtain a distance distribution. Table 2-2. Distance distribution profile used to analyze noise effects on dipolar evolution data: analysis for a 1 Gaussian distribution Parameter Gaussian 1 Gaussian 2 Gaussian 3 Distance 25 35 45 Width 3.4 3.4 3.4 Percentage 100 100100 It can be seen that, for the case of a si ngle population, MC data analysis seems to reconstruct the signal without noise artifacts. In the case of TKR, an increase in the amount of noise generates wider distance distributions, as well as minor populations of smaller distances (which result from noise eigenvalu es incorporated into the soluti on eigenvalues). It can also be noted distance distributions with shorter averag e distance (Figure 2-20A) are more sensitive to the effects of noise in the DEER signal, wh ereas distance distributions with long average distances (Figure 2-20C) are less sensitive to noise. In order to analyze the limitations of bot h methods when more than one distance distribution is present, we gene rated a system with two differe nt average distances, modeled by two Gaussian functions, with different average di stances, different width of the distributions and relative population percentages. This system w ill provide further information to understand how sensitive TKR and MC are when more than one di stinct population is pres ent in the sample. The distance distribution generated for this system is shown in Figure 2-21A). Reconstructed (noiseless) DEER dipolar evolu tion function is shown in Figure 2-21B), and dipolar evolution function with increasing amounts of noise is shown in Figure 2-21C ). The results of analyzing

PAGE 65

65 the dipolar evolution functions obtained in 221C by TKR and MC are shown in Figures 2-21D and 2-21E, respectively. From the distance distributions obtained by TK R and MC from the dipolar evolution data of Figure 2-21C, it can be seen that for this particular choice of distance distributions, Montecarlo minimization performs poorly in the presence of noise when compared to Tikhonov regularization. Figure 2-21 D shows that, upon in creasing amounts of signal noise, the second Gaussian distribution correspondi ng to a function centered at 38 with a 5 width and 20% contribution to the total distan ce distribution profile disappears in the background noise. On the other hand, Tikhonov regularization is able to extract both distan ce distributions, even in the presence of very large amounts of noise. (F igure 2-21 E). Table 2-3 shows the parameters employed to generate the 2-Gaussian dist ribution used in this stability study. Table 2-3. Distance distribution profile used to analyze noise effects on dipolar evolution data: analysis for a 2 Gaussian distribution Parameter Gaussian 1 Gaussian 2 Distance 32 38 Width 3 5 Percentage 80 20 In a nutshell, we believe that in order to analyze the results obtained from double electronelectron resonance, an agreement must be found between the results obtained by Tikhonov regularization as well as MC techniques. This will ensure that the distribution function is not artificially broadened in the TK R analysis due to the regulariza tion parameter choi ce and that all possible distinct distance distributions are accounted for.

PAGE 66

66 10-1210-1010-810-610410-21 10-14 -RayX-RayUVIR Radio 102 10-4 500 L S X Ku Q W140 Ghz 240 Ghz1510 20 30 100 EPR NMR ( m) (mm) 10-1210-1010-810-610410-21 10-14 -RayX-RayUVIR Radio 102 10-4 500 L S X Ku Q W140 Ghz 240 Ghz1510 20 30 100 500 L S X Ku Q W140 Ghz 240 Ghz1510 20 30 100 EPR NMR ( m) (mm) Figure 2-1. Electromagnetic spectrum showing various bands of EPR operation. 1 2sm 0 ee E h g B 1 2 E 1 2 B0=0 B0 1 2sm 0 ee E h g B 1 2 E 1 2 B0=0 B0 Figure 2-2. Energy level diagram for a free electron in an applied magnetic field.

PAGE 67

67 Attenuator Klystron/ Gunn Diode Magnet Ref. Arm Detector SampleIsolator Monochromator Source Sample Cell Partially Reflecting mirrorRotating sector Detector Monitoring DetectorA B Attenuator Klystron/ Gunn Diode Magnet Ref. Arm Detector SampleIsolator Monochromator Source Sample Cell Partially Reflecting mirrorRotating sector Detector Monitoring Detector Attenuator Klystron/ Gunn Diode Magnet Ref. Arm Detector SampleIsolator Attenuator Klystron/ Gunn Diode Magnet Ref. Arm Detector SampleIsolator Monochromator Source Sample Cell Partially Reflecting mirrorRotating sector Detector Monitoring Detector Monochromator Source Sample Cell Partially Reflecting mirrorRotating sector Detector Monitoring DetectorA B Figure 2-3. Comparison between cw-EPR and optic al spectroscopy instrumentation. A) cw-EPR scheme and B) Optical spectroscopy setup [Adapted from Weil, J. A.; Bolton, J. R.; Wertz, J. E. Electron Spin Resonance: Elementary Theory and Practical Applications ; John Wiley & Sons, Inc., 1972.]

PAGE 68

68 10010410210610810101012 10010410210610810101012 SXQW 1mK 4.2 K 48 K Thermal energy Electron Zeeman interaction Zero-field splitting Hyperfine interaction Pulsed EPR excitation bandwidth Nuclear Zeeman interaction Nuclear quadrupoleinteraction 1 103 5 nm Electron dipole-dipole interaction Homogeneous EPR linewidths Homogeneous ENDOR/ESEEM Nuclear dipole-dipole interactionsHz Hz 10010410210610810101012 10010410210610810101012 10010410210610810101012 10010410210610810101012 SXQW 1mK 4.2 K 48 K Thermal energy Electron Zeeman interaction Zero-field splitting Hyperfine interaction Pulsed EPR excitation bandwidth Nuclear Zeeman interaction Nuclear quadrupoleinteraction 1 103 5 nm Electron dipole-dipole interaction Homogeneous EPR linewidths Homogeneous ENDOR/ESEEM Nuclear dipole-dipole interactionsHz Hz Figure 2-4. Typical energies (in Hz) of electron and nuclear spin interactions. [Modified from Schweiger, A.; Jeschke, G. Principles of pulse el ectron paramagnetic resonance ; Oxford University Press: London, 2001] 1 2sm 0 eeEhgB 1 2 E 1 2 B0=0 B0 -1 -1 0 0 1 1 klm B0 (1,1) (0,0) (-1,-1)AB 1 2sm 0 eeEhgB 1 2 E 1 2 B0=0 B0 -1 -1 0 0 1 1 klm B0 (1,1) (0,0) (-1,-1)AB Figure 2-5. Energy splitting diag ram and typical cw-EPR lineshape for a nitroxide moiety. A) Hyperfine interaction diagram for a system with Ms=1/2 and MI=1. B) EPR lineshape corresponding to a system with Ms=1/2 and MI=1.

PAGE 69

69 PPPPSQ ZQ DQ SQ SQ SQ SQSQ SQ SQ ZQ DQ PPPPSQ ZQ DQ SQ SQ SQ SQSQ SQ SQ ZQ DQ Figure 2-6. Two-electron spin density matrix. The diagonal terms correspond to the different energy level populations, whereas offdiagonal terms correspond to quantum coherences. (2) mw (1) mw 2 t 2 (1) mw 1t 1 1 3 A) B) (2)mw (1)mw 2 t t 2 (1)mw 1t 1 1 3 A) B) A B (2)mw (1)mw 2 t 2 (1)mw 1t 1 1 3 A) B) (2)mw (1)mw 2 t t 2 (1)mw 1t 1 1 3 A) B) A B Figure 2-7. Pulsed ELDOR sequences. A) Th ree-pulse DEER. Time t is varied and is fixed. B) +1 experiment (all pulses are applied at the same frequency).

PAGE 70

70 -1500-75007501500 0.0 0.5 1.0 1.5 Intensity [A.U]Time (ns) -1500-75007501500 0.0 0.5 1.0 1.5 Intensity [A.U]Time (ns) Figure 2-8. 4-pulse DEER dipol ar evolution traces. The upper trace corresponds to a broad distance distribution centered at 60. Th e lower trace corresponds to the dipolar evolution of the TEMPO monoradical. The shaded region corres ponds to the threepulse DEER experiment deadtime [Figure modi fied from Pannier, M.; Veit, S.; Godt, A.; Jeschke, G.; Spiess, H. W. J. Magn. Reson. 2000, 142 Page 339]. (2) mw (1) mw 1 1 2 2 2 t2= 1.5 s 3 s ( )= 32 ns 1= 200 ns (2)mw (1) mw 1 1 1 1 2 2 2 2 2 t t2= 1.5 s 3 s ( )= 32 ns 1= 200 ns 2= 1.5 s 3 s ( )= 32 ns 1= 200 ns Figure 2-9. 4-pulse DEER sequence. The obser ver sequence (top) consists of a Hahn echo sequence ( /2) followed by a refocusing pulse The pump pulse (bottom) at frequency 2mw excites a fraction of the spins coupled throughee changing the sign of the dipolar interaction.

PAGE 71

71 2 2 -Sz Sy Sx Sx Sx -Sz Sy Sx -Sz Sy Sx Sx (B) T=0T=0 2 2 -Sz Sy Sx Sx Sx -Sz Sy Sx -Sz Sy SxA B Sx (B) t T=t(1) mw (2) mw (1) mw (2) mw 2 2 -Sz Sy Sx -Sz Sy Sx Sx Sx -Sz Sy Sx -Sz Sy Sx -Sz Sy Sx -Sz Sy Sx Sx (B) T=0T=0 2 2 -Sz Sy Sx -Sz Sy Sx Sx Sx -Sz Sy Sx -Sz Sy Sx -Sz Sy Sx -Sz Sy SxA B Sx (B) t T=t(1) mw (2) mw (1) mw (2) mw Figure 2-10. Complete 4-pulse DEER scheme including a representation of the magnetization vectors for each step in the pulse sequence. In the top sequence, the pump pulse is applied at t=0, when the obs erver Hahn echo is at its maximum. The bottom sequence represents a particular time T=t for which th e refocusing of the observer spins after the pump pulse at time t leads to a partial refocusing of the spin echo.

PAGE 72

72 3400344034803520 Field (G) Pump Observe 26G 72MHz3400344034803520 Field (G) Pump Observe 26G 72MHz Figure 2-11. Positions of the observer and the pump pulses for the DEER experiment. The pump pulse is applied in the region with the larg est number of spins, corresponding to the central resonance line, whereas the observe r pulse is placed on the low field line which, on average, lays ~26 G (~72M Hz) below the central resonance. 02004006008001000 -300 -200 -100 0 100 200 300 400 500 600 Intensity (a.u.)Time (s) Figure 2-12. Measurement of th e spin-lattice relaxation time (T1) by saturation recovery. In this example, the data was fitted to a grow th exponential. The obtained value for T1 on this sample was ~380 s.

PAGE 73

73 2 -Sz Sy Sx 2 Sx Sx 2 -Sz Sy Sx 2 Sx Sx Figure 2-13. Pulse sequence for the 2-pulse Hahn Echo T2/Tm measurements. 02468 0 100 200 300 Intensity (a.u.)Time (s) Figure 2-14. Measurement of the spin-spin relaxation time (Tm) by a Hahn echo pulse sequence. In this example, the data was fitted to a decaying exponential. The obtained value for Tm on this sample was ~4 s. The oscillations in the first 3 s are due to proton modulation via ESEEM.

PAGE 74

74 Pulse Programmer ReceiverRe Re Re Im Im ImSynthesizer Phase ShifterPhase Adj. ADCsSig. Amp AmpPulse Gate Duplexer Magnet Probe Magnet Power Supply MPFU -1 MPFU -2 MPFU -3 TWT Amp Sig. AmpReceiver Pulse Shaper Source Klystron/ GunnRe Re Re Im Im Im Current Modulation Pulse ProgrammerADCs Phase Adj. Magnet CirculatorA B Pulse Programmer ReceiverRe Re Re Im Im ImSynthesizer Phase ShifterPhase Adj. ADCsSig. Amp AmpPulse Gate Duplexer Magnet Probe Magnet Power Supply MPFU -1 MPFU -2 MPFU -3 TWT Amp Sig. AmpReceiver Pulse Shaper Source Klystron/ GunnRe Re Re Im Im Im Current Modulation Pulse ProgrammerADCs Phase Adj. Magnet Circulator Pulse Programmer ReceiverRe Re Re Im Im ImSynthesizer Phase ShifterPhase Adj. ADCsSig. Amp AmpPulse Gate Duplexer Magnet Probe Magnet Power Supply MPFU -1 MPFU -2 MPFU -3 TWT Amp Sig. AmpReceiver Pulse Shaper Source Klystron/ GunnRe Re Re Im Im Im Current Modulation Pulse ProgrammerADCs Phase Adj. Magnet CirculatorA B Figure 2-15. Comparison between instrumentation for A) nuclear magnetic resonance, and B) pulsed electron paramagnetic resonance. HppA)A1) HppA)A1) LF CF LF CF 02004006008001000 0 2 4 6 8 Field PositionIntensity (au) B0 r I 02004006008001000 0 2 4 6 8 Field PositionIntensity (au) B0 r r I IB) C) Figure 2-16. Graphical determination of comm only used cw-EPR spectral parameters. A) Hpp. B) LF/CF ratio and C) second moment (

).

PAGE 75

75 01230.7 0.8 0.9 1.0 Intensity (a.u)Time (s)01230.7 0.8 0.9 1.0 Intensity (a.u)Time (s) Background Exptl. Data Data Fit Exptl. Data AB01230.7 0.8 0.9 1.0 Intensity (a.u)Time (s) 01230.7 0.8 0.9 1.0 Intensity (a.u)Time (s) Background Exptl. Data Data Fit Exptl. Data 01230.7 0.8 0.9 1.0 Intensity (a.u)Time (s) 01230.7 0.8 0.9 1.0 Intensity (a.u)Time (s) Background Exptl. Data Background Exptl. Data Data Fit Exptl. Data Data Fit Exptl. Data AB Figure 2-17. Background correction of the partially refocused echo in DEER spectroscopy. A) DEER dipolar evolution da ta (black) and choice of background (red). B) Background subtracted DEER data (black) a nd dipolar evolution fit (blue).

PAGE 76

76 15202530354045505560 P(r)Distance (nm) FT TKR MC -Gaussian 1/r3 2 22 x bcfxae 15202530354045505560 P(r)Distance (nm)01234 Intensity (a.u.)Time (S)15202530354045505560 P(r)Distance (nm)15202530354045505560 P(r)Distance (nm) FT TKR MC -Gaussian 1/r3 2 22 x bcfxae 15202530354045505560 P(r)Distance (nm)01234 Intensity (a.u.)Time (S)15202530354045505560 P(r)Distance (nm) Figure 2-18. Data analysis algorithms for DEER dipolar evolution sign al. Direct FT of the dipolar evolution, Montecarlo fitting of the background subtracted dipolar evolution data using a predefined distance distribu tion function and Tikhonov regularization of the background subtracted dipolar evolution.

PAGE 77

77 Figure 2-19. Screenshot capture of DeerSim. This program simu lates DEER dipolar evolution curves based on a give distance distributi on (Gaussian or Lorentzian). Also, the program performs the Fourier transform of the dipolar evolution and further depakeing of the Pake spectrum.

PAGE 78

78 01000200030004000 Time(s) 01000200030004000 Time(s) 01000200030004000 Time(s) 23456 Distance ()23456 P(r)Distance () 23456 P(r)Distance () 23456 Distance (). 23456 Distance () 23456 Distance () 23456 Distance ()23456 P(r)Distance () 23456 Distance ().Distance Profile Dipolar Evolution TKR Distances MC Distances 01000200030004000 Time(s) 01000200030004000 Time(s) 01000200030004000 Time(s) 23456 Distance ()23456 P(r)Distance () 23456 P(r)Distance () 23456 Distance (). 23456 Distance () 23456 Distance () 23456 Distance ()23456 P(r)Distance () 23456 Distance ().Distance Profile Dipolar Evolution TKR Distances MC DistancesA B C Figure 2-20. TKR vs. MC analysis for a single Ga ussian distribution as a function of the average position and experimental noise A) TKR and MC analysis of the dipolar evolution generated by a single Gaussian centered at 25 with a FWHM of 3.4 B) TKR and MC analysis of the dipolar evolution generated by a singl e Gaussian centered at 35 with a FWHM of 3.4 C) TKR and MC an alysis of the dipolar evolution generated by a single Gaussian centered at 45 with a FWHM of 3.4

PAGE 79

79 2345678 P(r) (a.u.)Distance (nm)01000200030004000 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Echo IntensityTime (s) 01000200030004000 Time (s)0% ~ 5% ~ 10% ~ 15%Noisea b c 2345678 Distance (nm) 2345678 Distance (nm). MC DistancesTKR Distances de2345678 P(r) (a.u.)Distance (nm) 01000200030004000 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Echo IntensityTime (s) 2345678 P(r) (a.u.)Distance (nm) 01000200030004000 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Echo IntensityTime (s) 01000200030004000 Time (s)0% ~ 5% ~ 10% ~ 15%Noisea b c 2345678 Distance (nm) 2345678 Distance (nm) 2345678 Distance (nm). 2345678 Distance (nm). MC DistancesTKR Distances de Figure 2-21. Montecarlo and Tikhono v regularization analysis for a 2-Gaussian distribution A) 2-Gaussian distance distributi on utilized in this study. B) DeerSim generated dipolar evolution for the distance dist ribution in A. C) Dipolar evolution with added noise. D) Distance distribution reconstr ucted from dipolar evoluti on using Montecarlo fitting and E) distance distributi on reconstructed from dipol ar evolution using Tikhonov regularization.

PAGE 80

80 CHAPTER 3 MATERIALS AND METHODS Protein Expression The gene encoding the sequences for HIV-1 protease LAI consensus sequence was obtained from NIH reagents program, indire ctly through Dr. Ben Dunns lab. The main disadvantage of using this clone is that the protease comes from a viral system, and thus is not optimized for E. coli expression. Figure 3-1 shows the comparison between non codon-optimized protease (left) and codon optimized protease (right). Although the samples werent OD600 normalized, Figure 3-1 shows a great improvement on the yield of HIV-1 protease by moving to a codon optimized system. Based on these results, all th e protease strains were pur chased codon and expression optimized from DNA2.0 to obtain the necessary amounts required to perform EPR and future NMR experiments. It is well know n that HIV-1 protease undergoes se lf-proteolysis, and that the sites at which proteolytic cleavage occurs ar e positions 6-7, 33-34 and 63-64. Hence, in agreement with previous results from the liter ature, mutations Q7K, L33I and L63I were introduced in the sequence in order to avoid self-proteolysis when working with active recombinant forms of the protease. Mutations C67A and C95A were introduced as well. The removal of these naturally occurring cysteines th at are not involved in di sulfide bonds stabilizes the protein against unspecific disulfide bond formati on as well as unspecific labeling by the spin probe. This mutated sequence will be referr ed to as PentaMutated Protease (PMPR) For the initial studies of the protease for which a pulsed-EPR machine was not readily available, mutation D25N was engineered in the protease. This muta tion corresponds to the removal of the catalytic aspartic acid residue s and has been previously shown by X-ray and

PAGE 81

81 NMR not to perturb the over all protease structure. 99-104 In order to simplify the notation of different HIV protease mutant stra ins utilized in this study, the follow nomenclature will be used: PMPR+D25N+K55SL PMPR*SL V6+D25N+K55MTSL V6*MTSL MDR769+D25N+K55MTSL MDR769*MTSL The LAI, V6 and MDR769 genes were cloned into the pET-23a vector (Novagen) under the control of the T7 promoter, between NdeI and BamHI cutsites (see Figure 3-2). Mutations D25N and K55C, as well as other mutations wh ich were prepared for future studies (F53C, T74C, M46I and I54V) were introduced using th e QuikChange site-directed mutagenesis kit (Stratagene). The sequences for all genes were confirmed by DNA sequencing. Further information on construct sequences, thermal cycl ing reaction parameters and PCR primers used for site-directed mutagenesis is included in appendix E. HIV-1 protease constructs were expresse d in BL21*(DE3) pLysS cells (Invitrogen). Overnight cultures in 100 g/mL ampicillin were grown overnight, and 2.8 L Fernbach flasks containing 1 L of LB me dia and 1 mL Amp at 100 g/mL were inoculated with 2% of the overnight cell growth. After approx. 2-3 h of in cubation at 37 C with shaking at 300 rpm, protease expression was indu ced by adding 1 mM isopropyl-D-thiogalactoside (IPTG) when the culture density (as determined by absorbance at 600 nm (OD600) was >1.0. After induction, the temperature was dropped to 20 C to prom ote formation of insol uble inclusion bodies. Protein Purification Before describing the protein purification process, a special note must be made about buffers containing urea. It is well known that urea solutions decompose over time105-108 according to the reaction in Figure 3-3.

PAGE 82

82 The newly formed cyanate group can react with sulfhydryl groups forming a C-S bond, rendering the cysteine residues unable to react with the spin label as illustrated in Figure 3-4. Therefore, to avoid carbamylation of the cy steines, all buffers c ontaining urea were ionexchanged using a mixed-ion bed resin (AG 501-X8, BioRad) by adding 5 g resin/100 mL urea solution and stirring for ~5 h; the buffers cont ained a small amount of glycylglycine (diGly), commonly used as an ion scavenge r for the urea decomposition products. Plasmid-encoded HIV-1 protease constructs were expressed as in clusion bodies in E. coli as described in the literature. In brief, cells were pelleted (7500 g 20 min, 4 C), resuspended in Buffer A (20 mM Tris, 1 mM EDTA, 10 mM BME pH 7.5) and lysed with three passes through a 32 mL French pressure cell ( 16000 psi) (Thermo Scientific, Cat # FA-032). Cell debris and protease-containing inclusion bodies were collected by centr ifugation in an Eppendorf 5810R centrifuge (18500 g 25 min, 4 C). Inclusion bodies were washed in three steps using different buffers (Buffer 1: 25 mM TrisHCl, 2.5 mM ED TA, 0.5 M NaCl, 1 mM diGly, 50 mM BME, pH 7.0; Buffer 2: 25 mM TrisHCl, 2.5 mM EDTA, 0.5 M NaCl, 1 mM diGly, 50 mM BME pH 7.0, 1 M Urea; Buffer 3: 25 mM TrisHCl, 1 mM ED TA, 1 mM diGly, 50 mM BME, pH 7.0). Each of the steps included washing, homogenization/sonication (Fisher Sonic Dismembrator, Model 100, 50% duty cycle) an d centrifugation to collect inclusion bodies (18500 g 20 min, 4 C). HIV-1 protease inclusion bodies were solubilized (25 mM TrisHCl, 5 mM NaCl, 1 mM EDTA, 1 mM diGly, 50 mM BME, 9 M urea), clarified by centrifugation, (18500 g 25 C), filtered through a 0.22 M filter, and applied directly to a 10 mL anionexchange Q-Sepharose column (HiTrap Q Sepharose HP, Amersham), previously equilibrated in the same buffer. Protease was collected as fl owthrough; protease-conta ining fractions were pooled (typical volumes of 30 mL), and fo rmic acid (HCOOH) was added to a final

PAGE 83

83 concentration of 25 mM to precipitate contaminants and to lo wer the pH for the refolding step. Protease was stored overnight at 4 C and precipitated co ntaminants were removed by centrifugation in a Sorvall RC6 equippe d with a fixed-angle SS-34 rotor (39000 g 4 C, 20 min) with typical volumes of ~30 mL. It has been shown in the literat ure that the folding efficiency of the protease increases at low pH and low salt concentration57; therefore, HIV-1 protease wa s refolded in approx. 4 hours by 10-fold dilution into 10 mM formic ac id (pH 3.0) at 0 C (approx. 300 mL). After refolding, the temperature was raised to 25 C and pH was increased to 5.0 by dropwise addition of 2.5 M NaOAc (pH 5.5). Precipitated contaminants were removed by centrifugation (18500 g 25 min, 4 C). Typical volumes were ~330 mL. Refolded protease samples were concentrated using an Amicon 8200 stirred cell equipped with a Millipore polyethersufone membrane with nominal mol ecular weight cutoff of 10,000 and desalted (HiPrep 26/10 desalting column, Amersham). At this point, between 4-12 mL of refolded protease at concentrations ~100 M were obtained, depending on protease strain. Figure 3-5 shows a Tris-Tricine reduci ng gel of the HIV-1 PMPR+D25N+K55C purification process. The removal of the vast majority of contaminants is accomplished by the use of ion-exchange chromatography (Lane 8) Lanes 9 and 11 show the effect of HCOOH and NaOAc addition to the sample to precipitate c ontaminants. The resulting fraction after final desalting step is shown in Lane 10. Sample concentrations were 15 L of protein and 15 L of protein loading gel containi ng BME, to a total of 30 L sample per well. Figure 3-6 shows a reducing Tris-Tri cine gel for the purification of active HIVPR+PMPR+K55C. The gel shows that great purity in the purification process can be achieved in the initial steps of the purification (Lane 2, be fore running the ion-exchange

PAGE 84

84 chromatographic step). Final removal of a contaminant band around 55 kD is accomplished by the precipitation of contaminants using NaOAc. (Lanes 7 to 9). Spin Labeling Purified and desalted HIV-1 protease cyst eine mutant stocks were diluted from ~OD280=1.4 to OD280=0.6, in order to avoid precipitation of the protein in the desalting step necessary to perform the spin labeling reaction. When the OD280 of the protein was ~0.6, the protease was desalted using a Hi Prep 26/10 desalting column (Amersham) into 10 mM Tris, pH 6.9 for the MTSL spin labeling reaction, and in to 10 mM HEPES, pH 7.6 for IAP, IASL and MSL. 20-fold molar excess spin la bels were dissolved into 200 L of 100% EtOH and added to the protein solution in 20:1 label:protein concen tration to ensure full labeling. Labeling reaction proceeded in the dark at 4 C overnight. After labeling, spin labeled protease samples were desalted using a HiTrap 26/10 desalting colu mn into 2 mM NaOAc buffer at pH 5.0. HIV-1 protease constructs we re concentrated using a 15 mL Amicon concentrator and stored at ~100 M at -20 C. Th e protease was estimated to be 95% pure by SDS-PAGE. DEER Experiments DEER data were collected at three different institutions: (1) Nati onal High Magnetic Field Laboratory (NHMFL, Tallahassee) using a Bruke r EleXsys E580/E680 equipped with the ER 4118X-MD5 Dielectric Ring Resonator under conditi ons of strong overcoupli ng at a temperature of 65 K at X-Band frequencies, (2 ) University of Virginia (Chartlottesville, VA) using a Bruker EleXsys E580 equipped with the ER 4118X-MD5 Diel ectric Ring Resonator as well as with an MS2 resonator (preliminary studies) at a temper ature of 65 K at X-Band frequencies, and (3) Bruker BioSpin (Billerica, MA) using an El eXsys E680 equipped with the ER 4118 X-MD5 resonator, at a temperature of 80 K at X-Band frequencies.

PAGE 85

85 Figure 3-7 illustrates the sequence for the f our-pulse DEER experiment. The integrated intensity of the echo was recorded as a func tion of the delay time for the pump pulse ( ) 2. For our experiments, data were typically reco rded at steps of 12-16 ns, ranging from 0 t 3 ms. The pulse lengths for /2 and were 16 ns and 32 ns, respectivel y. The long interpulse delay, 2 (typically 1.5 3 s for our samples) was optimized for each sample by recording a spin-echo decay experiment and evaluating the phase memory time, Tm. In all cases, a 1 of 200 ns was used. The pump frequency was set to the globa l maximum of the nitroxide spectrum (central peak) and the observer frequency wa s set to the relative maximum of the low field line (typically at a distance of 70-72 MHz from the central peak). Accumulation times for the different datasets varied between 2 24 hours, depending on protei n concentration and inte rpulse delay. Further information on the DEER data collecti on setup can be found in Appendix B. Circular Dichroism Experiments Circular dichroism experiments were perf ormed on an Aviv 400 spectrometer. Sample conditions were ~0.3 mg/mL protein concen tration in 2 mM NaOAc pH 5.0. Protein concentration was determined by absorption at 2 80 nm, using an extinction coefficient of 1.15 mg cm-1 ml-1.The parameters used for the CD experiments are summarized in Table F-1 in Appendix F. The CD signal from the buffer was subtracted from the protein signal, and converted to mean residue ellipticity (degrees cm2 dmol-1 residue-1) using Equation (3-1). 10 NCl (3-1) In Equation (3-1), [ ] is the mean residue ellipticity, is the ellipticity (mdeg), N is the number of residues in the protein (198), C is the protein concentration and l the path length in cm (0.1 cm for the Hellma CD cuvettes used in this work).

PAGE 86

86 Cw-EPR Experiments cw-EPR experiments were collected using a modified ESP200/ESP300 Bruker spectrometer equipped with a loop-gap resona tor (Molecular Specialtie s, Milwaukee, WI). Typical parameters for cw-EPR spectra were : 1 G (uncalibrated) modulation amplitude, 20-100 G sweep widths, 2 mW incident power a nd 40.6 sec acquisition time. For a complete table of parameters used in the cw-EPR spectra acquisition, the reader is referred to Appendix F, Table F-2. It is important to note that, for those samp les that required an accurate measure of the Hpp spectral parameter, the central linewidth cw -EPR spectra was taken using 20 G scans, 1024 points. This provides a resolution of th e peak-to-peak distance of ~0.02 G. Samples for EPR were prepared from a 100 M protease stock in 2 mM NaOAc pH 5.0. Typical sample volumes of 10 L were loaded into 0.60 I.D x 0.84 O.D. capillary tubes purchased from Fiber Optic Center (Cat # CV 6084). Unless stated ot herwise, all cw-EPR experiments were performed at 24C To study the effects of the addition of inhibi tors in both cw-EPR and pulsed-EPR (DEER) experiments, stock solutions of Ritonavir in 100% DMSO and Indinavir in 2 mM NaOAc were prepared at concentrations of 4 mM (Ritonavir) and 8 mM (Indi navir). Stock solutions were diluted to 300 M final concentration of inhibitor in the protease sample. UV/Vis Substrate Degrad ation and Inhibition In order to analyze the effect of the spin label on the HIV-1 protease structure, we performed chromogenic substrate degradati on of active PMPR+K55MSL followed by UV/Vis. The catalytic activity of PMPR+K55SL wa s monitored following the hydrolysis of chromogenic substrate VI (Lys-Ala -Arg-Val-Nle-nPhe-Glu-Ala-Nle-NH2). (American Peptide

PAGE 87

87 Company, Inc.)109 This substrate resembles the cleavag e site ARVL/AEAM between the capsid and p2 polypeptide sequences in the gag polyprotein of HIV virus. The solubility of chromogenic s ubstrates in water is very limited. Hence, an appropriate amount of substrate to yield a stock concentration of 10 mM was first dissolved in 100% HCOOH. The amount of substrate dissolved was dependent on the molecular weight of the substrate and the purity percent of the peptide (usually synthesized peptides include 10-25% salts in the lyophilized product). If quant itative kinetics are to be perf ormed, the exact concentration of peptide after stock preparation shou ld be determined by peptide analysis. 2-5 L HIV-1 protease were adde d to a 125 L microcuvette (Varian, Inc) containing chromogenic substrate at 37 C. Final concentr ations in the assay buffer were: 50 mM NaOAc pH 4.7, 150 mM NaCl, 4 mM EDTA, 100 M chrom ogenic substrate, ~ 40nM concentration of protein. The absorbance was m onitored at nine wavelengths (296-304 nm) using a Varian Cary 50 spectrometer (Varian Inc.) equipped with a Pe ltier water bath and an 18-cell holder, and corrected for spectrometer drift by subtracti ng the absorbance at 446-454 nm. Osmolality and Viscosity Experiments For the osmolality and viscosity experiments, the preparation of the cosolute stocks and final sample solutions were made as follows: (1) Glycerol samples were made weighting 12.6, 25.2, 37.8 and 50.4 g. of glycerol (liquid, =1.261) and adding 2 mM NaOAc pH 5.0 to 100 mL, corresponding to samples 10, 20, 30 and 40% (v/v) of glycerol. Due to the sample size in EPR (usually 10 L), EPR samples were prepar ed by adding 1, 2, 3 and 4 L of glycerol to the corresponding amount of protein and 2 mM NaOAc buffer for a final volume of 10 L.(2) A stock solution of Ficoll400 30% (w/v) was made by weighting 3 g. of Ficoll400 and adding 2 mM NaOAc pH 5.0 to 10 mL. From this stock so lution, samples containing 3, 6 and 9% (v/v) were made by dilution of the Ficoll400 stock in the appropriate amount of protease sample. (3) A

PAGE 88

88 stock solution of 60% sucrose (w /v) was made by weighting 6 g. of sucrose and adding 2 mM NaOAc pH 5.0 to 10 mL. From this stock soluti on, samples containing 6, 12 and 18% (v/v) were made by dilution of the sucrose stock in the appr opriate amount of protease sample. (4) A stock solution of 60% PEG3000 (w/v) was made by we ighting 6 g. of PEG3000 and adding 2 mM NaOAc pH 5.0 to 10 mL. From this stock soluti on, samples containing 6, 12, 18 and 24% (v/v) were made by dilution of the PEG3000 stock in the appropriate amount of protease sample.

PAGE 89

89 200kD 116kD 97kD 66kD 45kD 31kD 21.5kD 123456789101112 200kD 116kD 97kD 66kD 45kD 31kD 21.5kD 123456789101112 Figure 3-1. Pilot expression of HI V-1 protease (PMPR) in LB medi a at 37C. Lanes 1-7 refer to non-codon optimized(non-CO) protease, whereas lanes 8-12 correspond to codonoptimized(CO) protease. Lane 1: Broad mol ecular weight marker (BioRad). Lane 2: non-CO PMPR before induction. Lane 3: non-CO PMPR 30 min after induction. Lane 4: non-CO PMPR 60 min after inducti on. Lane 5: non-CO PMPR 120 min after induction. Lane 6: non-CO PMPR 180 min after induction. Lane 7: non-CO PMPR 240 min after induction. Lane 8: CO PMPR before induction. Lane 9: CO PMPR 30 min after induction. Lane 10: CO PMPR 60 min after induction. Lane 11: CO PMPR 120 min after induction. Lane 12: CO PMPR 180 min after induction. NdeI BamHI pET-23a pET-23a HIV-1 Protease Gene NdeI BamHI pET-23a pET-23a HIV-1 Protease Gene Figure 3-2. DNA sequence encodi ng HIV-1 protease in pET-23a vector In blue, pET-23a sequence before and after the protease gene. In red, restriction enzyme cutsites for the HIV-1 protease gene (NdeI and BamHI). In green, HIV-1 PMPR/D25N/K55C gene.

PAGE 90

90 H2N C NH2 O NH3HCNO + NCO-+ NH4 + Figure 3-3. Decomposition of urea in to ammonium and cyanate ions. HCNO +HS C H2 S C H2 C H2N O Protein ProteinHCNO +HS C H2 S C H2 C H2N O Protein Protein Protein Protein Figure 3-4. Cartoon showing the labeling product from the chemi cal reaction between cyanate and sulfhydryl group in the ne wly engineered cysteine.

PAGE 91

91 Figure 3-5. Reducing Tris-Trici ne gel showing the different steps in the purification of HIVPR+PMPR+D25N+K55C. Lane 1: Broad molecular weight marker (BioRad). Lane 2: Supernatant after wash 0. Lane 3: Supernatant after wash 1. Lane 4: Supernatant after wash 2. Lane 5: Supern atant after Wash 3 Lane 6: Supernatant before loading into Q column. Lane 7: Protease fraction off Q-column. Lane 8: Contaminants bound to Q-column. Lane 9: Pr ecipitants after HCOOH step. Lane 10: Supernatant after HCOOH preci pitation. Lane 11: Precipita nts after NaOAc addition. Lane 12-13: Fractions after protease desa lting. Lane 14: Purified IAP labeled protease. Lane 15: Purified MTSL labele d protease. Lane 16: purified IAP labeled protease without the addition of reducing BME. Lane 17: purified MTSL labeled protease without the addition of reducing BME. Lane 18: Peptide molecular marker (BioRad).

PAGE 92

92 Figure 3-6. Reducing Tris -Tricine gel showing the different steps in the purification of active HIVPR+PMPR+K55C. Lane 1: Broad molecula r weight marker (BioRad). Lane 2: Supernatant before loading into Q column. Lane 4-6: Protease fractions off the Qcolumn. Lane 7: Pooled fractions after the Q-column step. Lane 8-11: Fractions after protease desalting. Lane 12: Pep tide molecular marker (BioRad). (2) mw (1) mw 1 1 2 2 2 t2= 1.5 s 3 s ( )= 32 ns 1= 200 ns (2)mw (1) mw 1 1 1 1 2 2 2 2 2 t t2= 1.5 s 3 s ( )= 32 ns 1= 200 ns 2= 1.5 s 3 s ( )= 32 ns 1= 200 ns Figure 3-7. 4-pulse double electron-el ectron resonance (DEER) sequence.

PAGE 93

93 CHAPTER 4 CONTINUOUS-WAVE AND PULSED ELE CTRON PARAMAGNE TIC RESONANCE STUDIES OF THE SUBTYPE-B LAI CONSEN SUS SEQUENCE OF HIV-1 PROTEASE Introduction In Chapter 1 the biology of HIV virus was intr oduced and in Chapter 2, the experimental techniques for measuring distances by pulsed-EPR were explained. In this chapter, the choice of spin label reporter site will be discussed and expl ained. In addition, a va riety of other biophysical characterizations of the spin la beled HIV-1 protease samples are presented. These experimental results assess the effects of the introduction of the label on the proteas e structure and activity. In order to perform Electron Paramagnetic Resonance (EPR) measurements on HIV-1 protease, a spin label must be incorporated into the protein. The chemi cal modification occurs via a covalent C-S or S-S bond with a cysteine re sidue introduced at the desired location for the label. In this work, we were interested in meas uring distances between selected residues in the flap region of HIV-1 protease in order to char acterize the conformational heterogeneity and distances between the flaps, both when the pr otease is free in solution and when bound to inhibitors. Because many amino acid substitution s in the flap region are known to alter the substrate and/or inhibitor binding affinities, special ca re needed to be taken when choosing the residue in the flap to serve as the reporter site. Table 4-1 summa rizes the results of a previous study performed by saturation mutagenesis on the flap of HIV-1 protease, showing only those substitutions that yielded WT-like activity.110 Table 4-1 shows that positions M46, G48, F53 and K55 tolera te a large variety of amino acid substitutions while maintaining close to WT activity. As we shall see, many amino acids in the flap region are related to dr ug resistance or are involved in side-chain contacts when the inhibitor is present in the activ e site, so special care should be taken when choosing a reporter site.

PAGE 94

94 Table 4-1. HIV-1 PR flap mutants with WT activity M46 I47 G48 G49 I50 G51 G52 F53 I54 K55 V56 Val Trp Phe Leu Ser Thr His Gln Arg Ala Lys Leu Tyr Leu Ile Val Trp Met Leu Val Thr Asn Ile Arg Gln Thr Cys Table modified from Shao, W.; Everitt, L.; Manc hester, M.; Loeb, D. D.; Hutchison, C. A., Swanstrom, R. Proc. Natl. Acad. Sci. U S A 1997, 94 Page 2246. As is shown in Figure 4-1, three main groups of residues can be dist inguished in the flap region. The first group, highlight ed in blue, is composed of those amino acids with solvent exposed side chains (M46, F53 and K55). The s econd group, highlighted in red, comprises the residues at the tips of the flaps, formed by the well-conserved sequence 47-IGGIG-52 G. The final group, highlighted in grey includes those amino acids with side-chains extending inward toward the active site cavity (I47, I50, I54 and V56). From these sets of residues, it is notew orthy to mention the following key points: M46I/L is a common mutation that appears in patients receiving one of the following protease inhibitors: ATV, FPV IDV, LPV, NFV and TPV. 111-113 I47A/V develops in patients treated with ATV, DRV, FPV, I DV, LPV, NFV or TPV.114116 The Glycine rich 48-GGIGG sequence of the tips of the flaps is well conserved amongst different HIV-1 subtypes, with the exception of small amino acid substitutions in G48, and the mutations I50L/V. 117,118 F53 has been shown to tolerate a wide range of mutations (Table 41), but F53 seems to be involved in hydrophobic inte ractions and it has been hypothesized from NMR data and MD simulations that the F53 ring might be coupled to backbone fluctuations119,56. I54V/A/L/M mutations are associated with drug resistance in patients under any PI treatment118,120-123. K55 has not been associated with drug-resistan ce and seems to tolerate a large variety of amino acid side chains w ithout loss of activity.

PAGE 95

95 From the saturation mutagenesis results, as well as other results in the literature using drug-resistant variants, position K55 seems to be th e only available residue w ithin the flap region that tolerates large amino acid side chains (s olvent exposed site, no t involved in hydrophobic interactions) and is not involved in the drug resistance process. It is also noteworthy to remember that all amino acid substitutions at site K55 (Table 4-1) from saturation mutagenesis studies retained WT protease activity. Based on these prev ious studies, residue K55 was chosen as the reporter site to be mutated to the cy steine residue needed for SDSL-EPR. One aspect of SDSL is that the resultant EPR lineshape is a convolution of protein backbone motions coupled to the spin label side chain rotamers, which makes the interpretation of cw-EPR lineshapes in term s of biomolecule dynamics comp licated, as it is not always straightforward to decouple backbone from label mo tion. In addition, the distances obtained from pulsed-EPR provide interspin distan ces, which then need to be correlated to protein backbone distances. In an attempt to decouple spin label conformers from protein conformational changes, we performed all of our studies using 4 spin labels which differ in the nitroxide moiety attachment and the length of the linker chain. The 4 spin labels used in this work are (1-Oxyl2,2,5,5-Tetramethyl3-Pyrroline-3-Methyl) Methanethios ulfonate (MTSL), 4-MaleimidoTEMPO (MSL), 3-(2-Iodoacetamido)-PR OXYL (IAP) and 4-(2-Iodoacetamido)-TEMPO (IASL), and all of them are commercially availa ble from Toronto Research Chemicals or SigmaAldrich. Figure 4-2 shows a representation of the four most commonly used spin labels and the resultant covalent linkage that occurs upon chemical modification of the cysteine residue. From a molecular structure perspective, it can be seen that the IAP and IASL spin labels, namely those who have an acetamido group, possess one more atom in the linker region, increasing the

PAGE 96

96 number of degrees of freedom of the system. The linker region is defined as the group of atoms which link the ring in which the nitroxide group sits, to the C carbon of the cysteine. On the other hand, the maleimido-based label MSL ha s overall less rotable bonds (n=4), but its bulkiness may affect the interactions of the spin label with the protei n backbone and neighboring amino acid side chains, most commonly resulting in multiple trapped rotamers of the spin label. Last, MTSL has been extensively used in cw-EPR site-directed spin labeling as a reporter label due to a proposed 1,4interaction between the hydrogen in position C and the S atom. (Figure 4-3). Figure 4-3 shows the commonly used nomencla ture for each of the dihedral bonds of MTSL appended to the cysteine amino acid. Th e current model for spin label motion obtained from crystal structures of T4 lysozyme with MTSL attached to 3 different cysteine residues indicates that, when MTSL is incorporated in an alpha-helix environment, the electrostatic interaction between S and the hydrogen attached to the cysteine C reduces the mobility of the spin label around the angles 1, 2 and 3. This model was developed in W. Hubbells lab, and it is known as the 4/5 model124, indicating that spin label dihedral rotations are only allowed around the 4/5 angles. From this model it has been hypothesized that due to an existing 1,4interaction125 between S and H-C, which leads to a constrained conformational mobility of MTSL, coupling the motion of the label to the mo tion of the protein back bone and making it an ideal reporter for studies on protein dynamics. It is important to reiterate that this model is based on a spin label located in a solvent-ex posed site located on an alpha helix. The mobility of a given spin label attached to a cysteine residue in a protein depends on three main factors: (1) Protein correlation time ( R). As we will see throughout this chapter, the overall rotational correlation time of the protein (tumbling) is dependent on the radius, the

PAGE 97

97 temperature and the viscosity of this solution. As we shall see, it is possible to add cosolutes which produce high viscosity soluti ons in order to slow down th e protein rotational correlation time. (2) Backbone fluctuations ( B). Because the spin label is attached to a cysteine in the backbone of the protein, the motion of the bac kbone influences the shape of the spin label spectra. Highly mobile spectra show sharp res onance transitions, whereas immobilized spectra show a high degree of spectral broadening. (3) Spin label conformati onal heterogeneity and diffusion rate ( I). Due to the existence of a linker region that attaches the nitroxide moiety of the spin label to the protein bac kbone, the internal dynamics of the label (rotameric states and conformational switching) greatly impact the cw -EPR lineshape. On the one hand, spin label conformational heterogeneity is related to the torsional energi es around each of the dihedral bonds of the spin label tether; high torsional energy barriers for these dihedral angles will restrict the motion of the spin label in space, confining th e nitroxide moiety in a particular region of the conformational space. Therefore, high torsional energy barriers are likely to produce spin label lineshapes that are indicative of backbone fluctu ations as opposed to label motions. When the rotational barrier betwee n different conformers is low, conformational exchange between different rotamers occurs and the motion of th e spin label is decoupled from the backbone. On the other hand, but also related to the torsional energy barrier of the dihedral angles, is the rate of motion of the label. It is not straightforward to link the rate of motion of the label to the relative torsional energy barriers of its conformers, as geometric transition-states and transition-state barriers do not necessarily correspond to the kinetic transition state. A detailed description of the different transition state theories (TST) is outside the scope of this work, and the reader can refer to the work of D. Truhlar for further references. Nonetheles s, the rate of motion of the label can be rationalized by defining a rate of motion in te rms of a macroscopic Brownian diffusion tensor

PAGE 98

98 that describes the diffusion propertie s of the spin label as a whole in a particular reference system that can be associated to the laboratory frame by means of an Euler rotation. Therefore, we can conclude that the intrinsic mobility of the spin label reflected by the EPR lineshape is a product of two different eff ects: torsional (1) energy barriers for each of the dihedral angles in the spin labe l and (2) diffusion rate of the spin label as a whole, influenced by the rate of motion of each of the di hedrals that form the linker region. From this discussion, it seems evident that the ability of the spin label to report backbone motion is heavily dependent on the chemical nature of the spin label, as well as the environment around the label once attached to th e protein structure. It is impor tant to note that preferential conformations around each of the dihedral angles exist independently of the position of the labeling in a particular protein (e.g. -helix vs. -sheet), but the interaction of the label with neighboring residues, as well as elec trostatic interactions of the la bel with other residues and the protein backbone, will determine the relative torsiona l energy barrier as well as the exchange rate between different conformers. Figure 4-4 schematically represents the factors that play a key role in lineshape mobility. As we saw in the introduction of this chapter, we are limited in the positions in which mutations are tolerated in the flap region, ma king K55C the only amino acid substitution in the flap region that does not alter pr otease activity. HIV-1 protease has a molecular weight of 18k D, and there is the possibility of including high visc osity solutions to eliminate protein tumbling ( R) from the possible factors affecting the cw-E PR lineshape. In an attempt to decouple the remaining two factors, namely backbone motion a nd spin label motion, as well as to provide a comprehensive study on the different spin labels as backbone motion report ers, cw-EPR spectra

PAGE 99

99 of PMPR*SL both in the presence and in the absence of inhibitor (Ritonavir) were collected using four different spin la bels (Figures 4-5 and 4-6). Figure 4-5 shows the results obtained for all different spin labels in Figure 4-2. The changes in the high field spectral line and the spec tral parameters in Tabl e 4-2 indicate that the mobility trend from the cw-EPR spectra, obtained from the intensity ratio between the centerfield and low-field resonant transitions (LF/CF), correlates with what sh ould be expected from pure geometrical considerati ons, with the most mobile spectra corresponding to the iodoacetamido-based labels (IASL and IAP). The most immobilized lineshape corresponds to MSL, for which the low spectral mobility might be related to interactions between the protease flap amino acid sidechains and the bulky label. Last, the MTSL lineshape shows a somewhat restricted motion, which is both compatible with the 4/5 model and the fact that MTSL lacks one CH2 group in the side chain. Table 4-2 summarize s the spectral paramete rs obtained for all the spin labels in HIV-1 PMPR *SL utilized in this study. Table 4-2. Spectral parameters for the lineshapes of PMPR*SL Label Hpp Second Moment (

) LF/CF MSL 2.91 2160.54 MTSL 1.99 1840.62 IAP 1.86 1770.67 IASL 2.21 1980.96 Figure 4-6 shows the area normalized 100 G scan X-Band cw-EPR lineshapes corresponding to four different spin labels used in this study in the presence (Ritonavir, red) and absence of inhibitor (10% DMSO, black). It has been previously shown in the literature utilizing several biophysical characteriza tion techniques that in the WT protease, upon addition of inhibitor, the flaps close and lo ck on top of the inhibitor, a nd the only backbone motion observed is that of the small amplitude fluctuations of the protein backbone. Surprisingly, Figure 4-6

PAGE 100

100 reveals no significant differences in the EPR lin eshapes between the inhibited and uninhibited forms of the protease. The most plausible expl anation is that, at XBand (9-10 GHz), the EPR lineshapes may only reflect the lo cal modes of motion of the spin label and are not reporting backbone motion. For the particul ar case of the MTSL label, these results are surprisingly inconsistent with the 4/ 5 model proposed by Hubbell et al. High Field-High Frequency (HFHF) spectra at 170 GHz and 240 GHz for MTSL labele d protease in the presence and absence of inhibitor are being collected in collaboration w ith K. Earle and J. Freed at ACERT (Cornell University) in order to study whethe r differences in the lineshape of the spin label in the presence and absence of inhibitor (Ritonavi r) can be obtained from cw-EPR. The results of this study will be reported elsewhere. Stability of Spin Labeled HIV-1 Protease During the development of the purification scheme, it was observed that HIV-1 PR was highly sensitive to salt concentration. To fu rther determine which salt concentrations are compatible with the protease concentrations need ed for EPR/NMR, a series of EPR experiments were performed with 100 M HIV-1 PR labeled with MTSL a nd increasing concentrations of NaCl. The resultant X-Band EPR sp ectra are shown in Figure 4-7. Figure 4-7 shows the area normalized X-Band EPR spectra for 100 M HIV-1 protease PMPR+D25N+K55C in the presence of increasing concentrations of NaCl. These results show that HIV-1 protease EPR lineshape is not affect ed by concentrations up to 50 mM NaCl (Figure 4-7B), whereas the lineshape is broadened when higher concentrations of salt are added (500 mM NaCl, figure 4-7C). In the extreme case of 2. 5M NaCl (Figure 4-7D), the spectral lineshape is completely broadened. These results are consis tent with protease precipitation/aggregation at high salt concentrations, and this was corroborated by inspection of the sample capillary tube, in which white flakes of precipitate d proteins appeared as soon as 500 mM NaCl was added to the

PAGE 101

101 purified protease solution. To reduce protein a ggregation/precipitation du ring the experiments, we deemed that the maximum safe concentra tion of salt in the prot ease samples was ~50-150 mM for EPR and pulsed-EPR as well as for future NMR experiments. To continue with the character ization of the stability of the protease and the label in different conditions, we explored the effects of acid concentration (pH) on the EPR lineshape for two labels which differ in the structure of the li nker with the protease (S-S for the MTSL case vs. C-S for the IAP case). Figure 4-8 shows stack pl ots of X-Band EPR spectra of PMPR*SL as a function of formic acid concentration. For th e case of MTSL (Figure 4-8A), the spectra systematically broaden as the formic acid concentr ation is increased, whereas in the case of IAP, the high field peak in the cw-EPR lineshape firs t narrows as the acid concentration is increased (to 190 mM HCOOH), then the spectral lineshapes br oaden as the acid con centration is increased (from 190 mM to 3.8M). These ch anges in the EPR lineshapes at higher concentrations likely reflect precipitated protein, wh ereas at low acid concentratio ns (between 38 mM and 190 mM) the narrowing of the lineshape reflects label mo tion. Changes in EPR lineshapes as a function of acid concentration can be compared using th e second moment of the corresponding spectral lineshape (Figure 4-9A) and the ratio of the cen ter field/low field lines of the EPR lineshapes (Figure 4-9B). Figure 4-9A shows the change on the second moment of PMPR* for IAP (red) and MTSL (blue) labels as a function of acid concentration. This figure shows that for the case of MTSL there is a constant decrease in the second moment upon HCOOH addition, whereas for the case of IAP the spectral second moment first increases at low acid concentrations (up to 190 mM HCOOH), and decreases at concen trations higher than 200 mM. Th is decrease at moderate to high concentrations of acid is most likely due to protein aggregation/pr ecipitation, as we saw

PAGE 102

102 before that HIV protease tends to precipitate at moderate concen trations of salt (over 150 mM). Figure 4-9B shows the change on th e LF/CF ratio for IAP (red dots, ) and MTSL (blue dots, ) as a function of HCOOH concentration. It can be seen that the LF/CF ratio follows the same trend as the second moment, with an increase at low acid concen trations followed by a decrease in LF/CF ratio at high acid concentrations. The trend observed for the case of MTSL is co nsistent with what one should expect for HIV-1 protease labeled with MTSL: as the acid /salt concentration increases, the lineshape broadens as the protein starts to precipitate. Nonetheless, the increase in intensity and narrowing of the high field line for the cas e of IAP was not expected. Ou r current hypothesis for this narrowing effect of the IAP label at low acid con centrations is related to the protonation of the acetamido group in the linker region of IAP. Upon acidification of the IAP label, the carbonyl group in the acetamido moiety is protonated, and this protonation causes the orbital overlap between the C=O bond and the lone electron pair of the nitrog en to break, decreasing the rotational energy barrie r around the C-N bond, making the IAP label more mobile and therefore narrowing and increasing the intensity of the IA P lineshapes. The proposed reaction scheme is shown in Figure 4-10. In order to obtain a clear picture of the or bital overlap, we perf ormed Density Functional Theory (DFT) calculations using the Becke-3126 Lee-Yang-Parr127,128 (B3LYP) functional using a 6-31 G split valence basis set. Figure 4-11 (left) shows the molecular or bital for the N-C=O bonds. Upon protonation (left) the C=O orbital disappears and the rotationa l energy barrier for the C-N bond is lowered. The possibility of lines hape narrowing due to backbone mo tions is ruled out; this effect

PAGE 103

103 would be present regardless the spin label of choice. Furthermore, this effect is present only for IAP, and not for MTSL. Effects of the Spin Label on the Struct ure and Activity of HIV-1 Protease In the previous section we discussed the stabili ty of the spin labeled protease as a function of salt and acid concentration. In this section, th e effects of the spin labe l on the structure and the activity of HIV-1 protease are presented and di scussed, both in an active and inactive (D25N) HIVPR construct. Effects of Spin-Labeling on the Structure of the Protease In order to study the effects of the incorporation of the spin la bel probe in the flap region of HIV-1 protease, we utilized circul ar dichroism as an indicator of secondary structure in proteins. Circular Dichroism is a spectroscopic techni que based on the differential absorption of circularly polarized light which has been proven very useful in analyzing the secondary structure of proteins129. Figure 4-12A shows a scheme for the genera tion of in-plane polarized light (equal amplitude components from the left and right polariz ed light). As it can be seen in Figure 4-12B and Equation (4-1), differential absorption of left and right circularly polarized light generates elliptically polarized light. L R A AA (4-1) Generally, conventional circular dichrois m instrumentation repor ts the differential absorption in terms of the ellipticity ( ) which is defined in Equation (4-2) 1tan ba (4-2) where a and b are the major and minor axes of the resulting elliptically polarized light.

PAGE 104

104 Often, the units reported in raw data by the instrument in are mdeg ( ) and reporting of CD spectroscopy data in sc ientific journals occurs usually in units of mean residue ellipticity with units of [deg cm-2 dmol-1 residue-1]. 100 []r A M clN (4-3) In Equation (4-3), [ ] is the mean residue ellipticity, is the ellipticity (in mdeg), Mr is the protein molecular weight, c is the protein concentration (in mg/ml), l is the cuvette path length and NA the number of amino acids in the protein. Usually, CD spectra are collected in the region of 180-260 nm. The left vs. right absorption diffe rence by the protein pe ptide bond in different secondary structures can be used to distingui sh between alpha helices, beta sheets and random coils in proteins. Figure 4-13A shows typical CD spectra for an all-alpha helix protein, generated from the CDDATA.43 circular dichroism dataset. It can be seen that the spectrum possess two minima, at 214nm and 222nm. On the other hand the CD spectrum for a mostly -sheet protein (Figure 413B) shows a single minimum at 216 nm. In order to probe the secondary structure of the spin labeled HIV constructs, we performed circular dichroism experiments on PMPR*SL (S L=MTSL/ MSL/ IAP/ IASL) in 2 mM NaOAc and pH 5.0. (Figure 4-14). Experimental conditio ns as well as spectroscopic parameters used in this study are reported in Appendix F. From comparison with published data on th e active LAI variant of HIV-1 protease130, no significant differences can be found in the circul ar dichroism spectra of PMPR inactive (D25N) labeled proteases and th e LAI consensus sequence. In order to obtain a semi-qua ntitative analysis on secondary st ructure, CD spectra for all HIV-1 PMPR*SL (SL=MTSL, MSL, IAP and IA SL) were analyzed using different CD

PAGE 105

105 secondary assignment methods. The resu lts obtained for the percentages of -helix, -sheet and random coil for each sample are summarized in Tables 4-3 and 4-4. Table 4-3. Secondary structur e assignments for HIVPR (PDBID 1HVI) and LAI consensus Method Alpha Beta Other Method Alpha Beta Other Structure LAI DSSP 4 2 5640 K2D 2 47 50 Stride 7 2 5439 Selcon2 11.9 45.2 42.9 GOR 14 4144 Selcon3 5 41.4 53.6 Average 5.5 5539 220nm 7.9 Average 6.7 44.5 49.2 Table 4-2 shows the results obtained for HIVPR LAI consen sus sequence from secondary structure analysis on structures deposited in the PDB databa nk (1HVI) and from CD spectra analysis. In order to extract the secondary struct ure information from crystal structures, two main algorithms131 have been utilized: DSSP131,132 and Stride133. DSSP is an algorithm based on pattern recognition of possible hydrogen bonds in structures de posited in the PDB databank. STRIDE is an algorithm based on known atomic coordinates deposited in the databank, and makes combined use of hydrogen bond energies and backbone torsional angle potentials to determine the protein secondary structure. Current implementations of STRIDE and D SSP are reported to agr ee in up to 95% of cases134. In order to analyze the seco ndary structure reported in crystal structures on the PDB database, as well as to ensure a good estimate of the protein secondary structure, we performed both DSSP and STRIDE analysis, averaging the re sults for both methods to obtain a consensus %content in alpha helix and beta strands. The errors for the determination of the secondary structure from experimental st ructures of the protease deposit ed in the PDB databank were estimated assuming that no more than two residues in the protease were erroneously assigned to either alpha helix or beta sheet.

PAGE 106

106 The programs utilized to determine secondary structure from the experimental CD data were K2D, which is based on a self-organ izing map (SOM) neural network algorithm135 using a basis set of 43 proteins (CDDATA.43) with known structure and CD spectra, and Selcon2136, which is a simpler algorithm based on Singular Value Decomposition (SVD) of a similar protein dataset. Table 4-4. CD secondary structure assignmen ts for PMPR*SL (MTSL, MSL, IAP, IASL) Method Alpha Helix Beta Sheet Random Coil Method Alpha Helix Beta Sheet Random Coil CD MTSL CD IAP K2D 2 5147 K2D2 51 47 Selcon2 2.5 52.545 Selcon26.1 50.4 43.7 220nm 6.2 220nm5.6 Average 3.6 51.846 Average4.6 50.7 45.25 CD MSL CD IASL K2D 3 5047 K2D3 50 47 Selcon2 5.4 48.945.7 Selcon27.7 46.9 45.4 220nm 6.3 220nm6.3 Average 4.9 49.546.4 Average5.6 48.5 46.2 From the comparison between the calculated s econdary structure averages using different methods (K2D, Selcon2) and the predicted seco ndary structure based on the DSSP and STRIDE algorithms (Table 4-2), it can be seen that the percent of alpha helix, be ta sheet and random coils in the labeled PMPR* protease are in good agreement with th e expected results based on assignments made on PDB-deposited crystal structures. Effect of Spin Labeling on th e Activity of the Protease As mentioned previously, rem oval of the active site in HI V-1 protease by mutagenesis of residue 25 from an aspartic acid residue to as paragine (D25N) has been shown in previous literature reports not to affect protein folding or binding of an inhibitor/substrate to the active site. In addition, saturation muta genesis studies showed that resi due K55 tolerate s a wide range

PAGE 107

107 of amino acid substitutions without affecting pr otease activity. In order to ensure that the introduction of the spin label in positions 55-55 (PMPR+K55SL) doe s not perturb protease activity, kinetic analysis of active HIV-1 pr otease PMPR +K55MSL was performed, and cwEPR spectra between active and inactive MSL labeled protease were compared. Figure 4-15 shows the comparison between PMPR+K55MSL and PMPR*MSL. It is noteworthy to mention that MSL was chosen as the spin label for the kinetic assay of active HIV-1 protease, as the bond formed by the label with the engineered cysteine (C-S bond) is less labile than the one formed by MTSL (S-S bond). As it can be seen in Figure 4-15, there ar e no differences in the EPR lineshape between active and inactive protease. In order to ensure that the labeling does not interfere with the activity of the protease and to make sure that the inhibitor is able to bind the active site and inhibit protease activity, the sample used to obt ain the spectra in Figure 4-15 was diluted to a protein concentration of ~40 nM, and chromogeni c substrate degradation analysis followed by UV/Vis was performed as described in Materials and Methods chapter. Shown in Figure 4-16 are plots of the decr ease in UV absorbance by the chromogenic substrate monitored at nine wavelengths (296-304 nm) as a function of time for ~40 nM HIV-1 PMPR+K55MSL with 100 M chromogenic substrate VI (Lys-Ala-Arg-Val-Nle-nPhe-Glu-AlaNle-NH2) (black trace) and ~40nM HIV-1 PMPR+K55MSL+ 100 M chromogenic substrate. with 4:1 Ritonavir:Protease excess (red trace). In the ab sence of inhibitor, th ere is a decrease in the UV absorbance in the 296-304 nm range, indicatin g that the chromogenic substrate is being cleaved. When inhibitor is present (red trace) the UV absorbance maintains a constant value over the 15 minute period. Hence, we can conclu de that the absence of degradation of the chromogenic substrate in the presence of inhibito r (Ritonavir) indicates that Ritonavir is binding

PAGE 108

108 effectively to HIV-1 PMPR+K55MSL. Another impor tant feature of HIV-1 protease is its selfproteolytic activity. It has been shown in the literature by HPLC chromatography that HIV-1 protease self-cleavage occurs between positions 7-8, 33-34 and 63-64 137 In order to test for selfproteolytic activity, HIV1 protease PMPR (100 M) was incubated at 37 C for a period of 46 days (Figure 4-17). It can be seen in the lineshapes for PMPR+K 55MSL in Figure 4-17 that, over time, a sharp signal appears at the high-field transition, and that the intens ity of this narrow component increases in intensity over time. The appearance of this peak over time is consistent with the generation of a small peptide fragment (positions 34-63) with the spin labe l attached to it, as a product of autoproteolysis. This high field peak corresponds to a new motionally narrowed spectral component with fast rotational correla tion time. The evolution of the high field peak (marked in Figure 4-17 with a green dashed line ), corresponding to the time course generated small peptide fragment 34-63, can be monitored by plotting the ratio between the intensity of the high-field peak and the low field peak. (Figure 4-18). Figure 4-18 gives an indication of th e stability to self-proteolysis of the pentamutated protease at ~ 100 M concentrations in 2 mM NaOAc pH 5.0. It is noteworthy to mention that degradation of the PMPR protease occurs on a longer timescale than its LAI counterpart due to the incorporation of the Q7K, L33I and L63I muta tions that stabilize the protease against selfproteolysis. The intensity ratio data were fitted to a growth exponential of the form of Equation (4-4) where A =0.399.005, x0=-0.3.3 and T =12.4.7 days. 01exp()/ y AxxT (4-4) Another factor that accounts for the longer time scale is the low salt concentration of the buffer (2 mM). It has been shown in the literature that HIV-1 protease activ ity increases with salt

PAGE 109

109 concentration138 (up to 1 M), but high salt concentrat ions are incompatible with high (100 M) protein concentrations. In c onclusion, the time evolution of the cw-EPR lineshape of PMPR+K55MSL in Figure 4-17 and the HF/LF ratio in Figure 4-18 show that HIV-1 protease undergoes autoproteolysis even wh en mutations Q7K, L33I and L 63I are incorporated. Hence, for preliminary studies in which pulsed-EPR inst rumentation was not readily available, and for which acquisition times in pulsed-EPR were on th e order of 24 h, introducing a mutation which completely removes the active site, making the protease stable against proteolysis was necessary (D25N). Figure 4-19 shows a snapshot of the cw-EPR lines hape at day 0 and day 46. In this figure, it can be clearly appreciated that the spectru m has two components, one corresponding to the same spectra as Day 0 (uncleaved proteas e) and one motionally narrowed component corresponding to the cleavage product of the protease. Osmolality and Viscosity As mentioned in the beginning of this chap ter, the X-Band EPR lineshapes reported by the spin label probe correspond to a convolution of three different motions: (1) Protein rotational correlation time, r (which is related to the size of the protein and the viscosity of the media through the Stokes-Einstein equation), (2) sp in label internal dynamics correlation time i (corresponding to the different rotamers of the la bel when attached to a protein backbone) and (3) protein backbone fluctuations timescale B (associated to conformational heterogeneity or conformational changes upon substr ate/inhibitor binding). Although it is not straightforward to deconvolute all motions from the spectral line shapes from a single experiment, individual parameters that affect each of the motions can be manipulated. For example, if we wish to reduce the protein correlation time to avoid a lineshape that is motionally narrowed, cosolutes such as sucrose, glycerol or Ficoll 400 can be added to increase the viscosity of the solution. Equation

PAGE 110

110 (4-5) shows the Stokes-Einstein equation, wh ich relates the radius of the particle ( r ), the temperature ( T ) and the viscosity ( ) to the diffusion coefficient (D). 6BkT D r (4-5) In this model, the diffusing particle is assu med to have a spherical shape. It has been shown that by incorporating gl ycerol or sucrose into EPR sa mples (i.e., T4 Lysozyme, C2A domain, PKC, RNA aptamers), the rotational corr elation time of the biom olecule can be slowed by increasing the solution viscosity such that the EPR lineshape reflects only the internal modes of motion of the spin label and the protein backbone conformati onal changes. However, it has also been shown that solutions of glycerol, PEG3000 and sucrose can alter protein conformation via osmolality effects139. It is also likely that the osmola lity alters the conformation and the mobility of the spin label itself. In order to understand the possibl e effects of cosolutes that have been previously used in cw-EPR and pulsed-EPR to reduce protein rotational co rrelation time, and which often present high osmolalitys and viscosities, we present a syst ematic study of the effects of four of these cosolutes (namely Glycerol, PEG3000, sucrose a nd Ficoll 400) on the cw-EPR lineshapes from HIV-1 PMPR*SL (SL=MTSL, MSL, IAP, IASL). The spin labels which will be used in this wo rk differ from each other in the length of the tether to the protein, the natu re of the cysteine coupling bond a nd the ring to which the nitroxide moiety is attached to (Figure 4-2). Figure 4-20 shows the structures of gycerol (4-19A, mw =92.1 g mol-1, =1.26 g cm-3), PEG3000 (4-19B, mw ~3000) and sucrose (4-19C, 342.29 g mol-1, =1.587 g cm-3). Ficoll 400 is not represented, as it is a hi ghly branched crosslinked polymer of epichlorohydrin and sucrose (mw~4 x 105).

PAGE 111

111 Solution osmolalities were measured using a Wescor 5500 vapor pressure osmometer (Wescor, Inc) with a 10 L sample size. Measurements were re peated in triplicate to ensure reproducibility. Viscosity measur ements were performed using a Cannon-Fenske viscometer suspended in a water bath at 24 C. Measuremen ts were repeated in triplicate to ensure reproducibility. The results obt ained from viscosity measurements are expressed as kinematic viscosity in units of centistokes (cS, or cSt) and are related to the value of the dynamic viscosity (usually in mPas, or cPoise) by the density of the solu tion (Equation (4-6)) (4-6) where is the kinematic viscosity, is the dynamic (absolute) viscosity and is the solution density. Because the solution densities are not kn own for different percentages of the cosolutes in the final solution, the viscos ity will be given in terms of Table 4-5 shows the results obtained from th e osmolality and viscosity measurements for solutions of the cosolutes employed. For the part icular case of Glycerol, the high osmolality of these solutions precluded the use of vapor pressure osmometry and the values of the osmolality were obtained from a technique kno wn as vapor pressure deficit. Hence, the data presented for glycerol is taken from the literature140. For further information about how the solutions were prepared, the reader is referred to Chapter 3. The data in Table 4-5 can be plotted as osmo lality vs. % ( w/v) of cosolute (Figure 4-20A) and viscosity vs. % ( w/v) of cosolute (Figure 4-20B).

PAGE 112

112 Table 4-5. Osmolality and viscosity of solutions of cosolutes employed in this study (T=24C) Cosolute % (w/v) Osmolality (mmol/kg) Viscosity (cS) (.01) Cosolute % (w/v) Osmolality (mmol/kg) Viscosity (cS) (.01) Ficoll 400 Sucrose 3 80 1.53 6 201 1.06 6 80.7.8 2.38 12 356 1.22 9 80 3.63 18 533 1.40PEG3000 Glycerol(a) 6 83.3 .3 1.71 12.6 1486 1.25 12 114 2.80 25.2 3473 1.73 18 172 4.17 37.8 5961 2.54 24 369 5.87 50.5 8950 3.93(a) Glycerol osmolality values were obtained by in terpolation from Davis, D. J., Burlak, C., Money, N.P Mycol. Res. 2000, 7 800-804 Figure 4-21A plots the dependence of the soluti on viscosity (cS) as a function of % (w/v) of cosolute added to the soluti on. This dependence can be fitted to a single exponential function of the form of Equation (4-7). 1/ 1 x tyAe (4-7) For the solutions of cosolutes used in this study, the values of 1/t1 follow the trend Ficoll (6.8 0.2) < PEG 3000 (13.0 0.9) < glycerol ( 36 1) < sucrose (43.8 0.6). This indicates that the change in viscosity as a function of % (w/v) cosolute follows the trend: Ficoll > PEG 3000> glycerol > sucrose. Small changes in th e % (w/v) of Ficoll400 ge nerate highly viscous solutions, which can be understood in terms of the highly branched polymeric nature of Ficoll. It is also noteworthy to mention the fact that th e dependence of the viscosity on the % (w/v) for glycerol and sucrose is almost identical. Figure 4-21B plots the dependence of the solution osmolality (mmol kg-1) as a function of % (w/v) of cosolute added to the solution. Three main points can be drawn from this graph: (1) Ficoll400 osmolality does not increase as a f unction of % (w/v) of Ficoll400 added to the solution. (2) The osmolality for glycerol is roughly 1-2 orders of magnitude higher than that of

PAGE 113

113 any of the other cosolute s under study, with osmolalities as high as 1 x 105 mmol kg-1 for 40% (v/v) glycerol added to the solution. (3) PEG 3000 and sucrose are in an intermediate regime between glycerol and Ficoll400, with sucrose having a higher osmolality for the full range of cosolute added to the solution under study. From direct observation of Fi gure 4-21B, a trend for the osmolality of the solutions of cosolutes under study can be established. This trend follows glycerol> sucrose > PEG3000 > Ficoll400, which is the inverse of the trend observed for the viscos ity change with % (w/v) in the cosolute solutions. Finally, we can combine the graphical repres entations of Figures 4-21A and 4-21B in a plot that relates the osmolality vs. viscosity for al l concentrations of the cosolutes in the solutions under study (Figure 4-22). Figure 4-22 shows the dependence of the solution osmolality vs. the viscosity for the four cosolute solutions utilized in this work. This fi gure shows that, in the case of Ficoll solutions, the osmolality and the viscosity are independent whereas in the case of glycerol, the increment in viscosity is coupled to a large change in osmo lality. Sucrose and PEG 3000 change osmolality as the solution viscosity increases, although this effect is not as pronounced as in the case of glycerol. The main conclusions that we can extract from the measurements of osmolality and viscosity are that, for those studies that requir e low osmolality and high viscosity to reduce the overall rotational correlation time of a protein, Ficoll 400 would be the cosolute of choice. On the other hand, if the stability of the protein under osmolality c onditions is to be tested, both glycerol and sucrose are good cosolutes to induc e osmolality effects. Because glycerol and sucrose are commonly employed as cosolutes for cw-EPR and DEER spectroscopy samples, for

PAGE 114

114 which experimental conditions re quire the addition of cryoprotectan ts, we were interested in relating how changes in viscosity and osmolality are reflected into the spin label lineshape as a function of cosolutes added to the solution, as well as to different spin labels. In order to understand the e ffects of different cosolutes on spin label conformational heterogeneity, we investigated the lineshapes of HIV-1 PMPR*SL (SL= MTSL, MSL, IAP and IASL), as a function of four different cosolute s (glycerol, sucrose, PEG3000 and Ficoll 400) in different %(v/v) concentrations. Th e results of this work are summ arized in Figures 4-23 to 4-25. Figure 4-23 plots the % change in the central linewidth ( Hpp) as a function of concentration for each of the cosolutes for PMPR*SL (SL=MTSL, MSL, IAP, IASL). Accordi ng to the description at the beginning of this ch apter, the central linewidth Hpp increases as the mobility of the label/backbone decreases. In order to measure accurately the Hpp, the central linewidth cw-EPR spectra was taken using 20 Gauss scans and 1024 points. This provi des a resolution of the peakto-peak distance of ~0.02 G. The first thing that can be noted from these plots is that the Hpp value for all spin label EPR lineshapes increases linearly as the concentration of solute is increased, with the exception of Ficoll. As we discussed previously, for the concentrations under study, Ficoll 400 has the lowest osmolality of all the cosolutes and th e osmolality does not ch ange in the range of concentrations of Ficoll utili zed in this study. From the Hpp analysis of the cw-EPR lineshapes of PMPR*SL it can be extracted that, within error, the Hpp value for the spin labels under study does not change with increasing Ficoll concentration in the solutions (Fig ure 4-23B). A second point that is reflected in these plots is that the lineshapes for MTSL (vs. MSL, IAP and IASL) appear to have the most dramatic changes in Hpp as the solute concentration increases.

PAGE 115

115 From purely geometrical constraints, it can be argued that the mobility of the label should follow the trend MTSLIASL>IAP, with the effects of cosolute following the trend Glycerol>PEG3000>Su crose>Ficoll 400. Taken together with the data in Figure 4-22, these results suggest that os molality plays an important role in the changes of the spectral lineshape. Noting that Fi coll does not induce si gnificant changes in Hpp, yet it has the highest viscosity, the possibility of a process dominated by slowing down the overall protein correlation time is di scarded. If the process was dominated by protein rotational correlation time, an increase in the viscosity shou ld slow down protein tumbling, making the cwEPR lineshape less mobile, and therefore incr easing the value of the central linewidth Hpp,. Nonetheless, there are no changes in Hpp (within error) for all concentrations of Ficoll in the solution, and therefore we can conc lude that the changes in the cen tral lineshape width are due to osmolality and not viscosity changes. This also implies that the changes in the lineshape are not due to R, but rather to C and/or B. We will proceed to analyze now the changes in spectral lineshape using the spectral second moment t echnique. As we mentioned previously, second moment analysis is not straight forward, in the sense that a go od S/N ratio and finely adjusted baseline are needed in order to obtain accura te results. Figure 4-24 shows the first-moment corrected second moment

parameter. Figure 4-24A shows the change s in the second moment

when the cosolute added is sucrose, as a function of spin label. This figur e shows that there are no systematic trends for

PAGE 116

116

as a function of sucrose concentration, but th e largest changes occur for the particular case of PMPR*MTSL. Figure 4-24B plots the change s in the second moment

when Ficoll 400 is added to the solution in different concentrat ions, as a function of spin label. This plot shows that the value of the second moment

barely changes for any of the labels under study when Ficoll is added to the solution in different percentages. This result is consis tent with the picture offered by Figure 4-22B, for which the changes in Hpp as a function of Ficoll we re minimal. Figure 4-24C shows the changes in the va lue of the second moment

as a function of th e concentration of glycerol added to the solution. In this case, PMPR*MTSL seems to have the largest changes in spectral parameter. Last, Figure 4-24D plots the changes of the value of the second moment

as a function of the concentration of PEG3000 added to the solution. In this case, a growth exponential trend can be seen in the case of MTSL and IASL labeled protease, whereas the rest of the spin labels (MSL, IAP) do not seem to follow a particular trend. In the case of IASL labeled protease, the relative changes in

are too small to be able to make any claims about its meaning. Nonetheless, for th e case of MTSL, the changes in

are in the order of ~10% and well outside the error region (%) indicating that measurements of

for MTSL are sensitive to increasing con centrations of PEG3000. As shown in Figures 4-23A, C and D and Figur es 4-24A, C and D, the largest changes for spectral parameters, whether central linewidth ( Hpp) or second moment (

) occur for PMPR*MTSL spin labeled protease. This is indica tive of a higher sensibility of the MTSL spin label vs. MSL, IAP and IASL to perturbations in the local environment.

PAGE 117

117 Nonetheless, it can clearly be seen that the second moment is a worse spectral parameter than Hpp to measure the changes in the lineshape for this particular system. Also, it can be seen from Figure 4-24 that neither the spin labels nor the cosolutes seem to follow a particular trend, which is an indication of the lower reliability of the spectra l second moment when subtle changes in the lineshape are to be reported. Figure 4-25 shows the changes in the center-field to low-fiel d intensity ratio. For this parameter, which we will refer as LF/CF, we s ee that the % change in LF/CF follows the same trend as Hpp, namely glycerol>PEG3000>sucrose>Fico ll400. The label trend for Sucrose follows MTSL>MSL>IAP>IASL, and this trend is conserved throughout all the cosolutes, with the exception of Glycerol, for which there is a small but noticeable inve rsion for the MTSL and MSL trends. Another important f act about the LF/CF parameter is that it seems to have the largest changes as the concentrations of solutes in crease. This is an indication of the sensitivity of LF/CF intensity ratio to small changes in the cw-EPR lineshape structure, and further analysis of the ratios of different inte nsities (ratios between LF, CF an d HF) should be considered for future studies in the area of data analysis for cw-EPR. Further information, including 3D plots of sp ectral parameters vs. osmolality and viscosity as well as all the cw-EPR lineshapes from which the parameters in Figures 4-23 to 4-25 were obtained, is given in Appendix H. Based upon the previous facts drawn from Fi gures 4-23 to 4-25, the following question about the nature of the interaction of the coso lutes with the label ar ises: Is the cosolute influencing the backbone motion of the protease, leading to a more immobilized structure of the flap or is the cosolute interacting with the la bel? To discard the firs t possibility of backbone

PAGE 118

118 modulation by cosolute, we performed DEER spec troscopy on PMPR*MTSL in the presence of 30% Glycerol and in the presence of 6% Fico ll. The results are summarized in Figure 4-26. Figure 4-26 shows the result of the DEER expe riment where equal vi scosity but not equal osmolality solutes were used (40% Glycerol and 6% Ficoll). It can be seen that there are no substantial differences due to the osmoprotect ant used, and the differences in the distance distributions likely arise from experimental noi se. Hence, the changes observed in the cw-EPR lineshapes in high osmolality solutions are likely due to interactions of the cosolutes with the nitroxide moiety itself, altering I. It is noteworthy to mention that the protease has several hydrophobic pockets, one of them locat ed in the flap region. From MD simulations performed by Simmerling et al. the MTSL label seems to make hydrophobic contacts with neighboring residue sidechains. When a cosolute that induces osmolality effect is a dded, the hydration layer of the label might be partially lost due to the chan ge in the activity of the water, and this effect might be responsible for the change s observed in the cw-EPR lineshape. These changes in the mobility of the cw-EPR lineshapes can have their origin in two different sources; as we mentioned before in this chapter, the cha nge in mobility of the label is due to (1) spin label diffusion a nd (2) spin label conformational he terogeneity, which results in a macroscopic order parameter C With the results obtained with solutes that present moderate to high osmolalities, we can only conclude that th e broadening of the cw-EPR lineshape in the presence of high osmolality solution is due to eith er conformational locking of the spin label, by modifying the electrostatic interactions of the sp in label ring in which the nitroxide moiety sits, or by modifying the rate at which the spin labe l switches between different conformers, without being able to distinguish be tween those two possibilities.

PAGE 119

119 In conclusion, in this work we have shown that solutions that present high osmolalities may also affect the spin-label conformational heterogeneity a nd dynamics, without affecting the protease backbone. These results add an extra laye r of complexity to cw-EPR studies performed in solutions such as glycerol or sucrose; literature examples have shown that addition of osmolites can induce protein backbone conformati onal changes; in this study we showed that osmolites can also modify the dynamics and conformations of the label. Hence, the process of choosing a cosolute of high viscosity to sl ow down protein tumbli ng (protein rotational correlation time, R) is a complex one, with the need to study what are the sp ecific interactions between the spin label in a particul ar site and the cosolute of inte rest. Nonetheless, we argue that, whenever possible, solutions of Ficoll 400 are recommended over those prepared using glycerol, sucrose or PEGs, as these can modify protein backbone conformations, sp in label conformational heterogeneity and spin label dynamics. Also, we determined that the HIV-1 protease PMPR*SL (SL=MTSL, MSL, IAP, IASL) flap region does not seem to be altered in the presence of osmolites, and the distance and distance distributions of the flap remain invari ant (within error) when di fferent cosolutes with high (glycerol) and low (Ficoll 400) osmolalities are utilized. Last, final notes on the peculiarities obs erved when performing DEER spectroscopy utilizing Ficoll are that: (1) Fi coll is a solid powder, and ther efore can be added as a solid without the need to dilute the protein sample, achieving higher protein concentrations, which is important in those samples which cant be concentr ated above a preset valu e (usually higher than 100-125 M). (2). Refocused Hahn echo measurements performed on 6% Ficoll/D2O seem to give longer Tm than those using Glycerol/D2O. (Figure 4-27). This can be of great utility in the

PAGE 120

120 future for those samples in which Tm is too short to perform distance measurements corresponding to long average distances and/or wide distan ce distributions.

PAGE 121

121 K55 F53 -IGGIGG47-52 M46 I54 K45 K55 F53 -IGGIGG47-52 -IGGIGG47-52 M46 I54 K45 Figure 4-1. Detailed molecular structure of the flap region of HIV-1 prot ease. In blue, residues that are solvent exposed. In grey, residues pointing towards the act ive site cavity and in red the glycine-rich region lo cated at the tips of the flaps. N H N I O O N H N O O S N N O O O N N O O O N SSO2CH3 O N S O N H N I O O H N O N O ProteinSH C C SC C SC C SC C ProteinSH ProteinSH ProteinSH A C D BN H N I O O N H N O O S N N O O O N N O O O N SSO2CH3 O N S O N H N I O O H N O N O ProteinSH C C SC C SC C SC C ProteinSH ProteinSH ProteinSH A C D B Figure 4-2. Spin labels used in this study. A) (1-Oxyl-2,2,5,5-Tetramethyl3-Pyrroline-3Methyl) Methanethiosulfonate (MTSL); B) 3-(2-Iodoacetamido)-PROXYL (IAP); C) 4-Maleimido-TEMPO (MSL) and D) 4-(2-Iodoacetamido)-TEMPO (IASL).

PAGE 122

122 H N C O C N H C O S N H3C H3C H3C H3C CH2 S O CH2 H d HS 45 321H N C O C N H C O S N H3C H3C H3C H3C CH2 S O CH2 H d HS 45 321 Figure 4-3. Structure for the spin label MTSL (R1, blue) appended to the cysteine residue (red). The previously proposed 1,4interaction is indicated by the dashed line and the the rotable dihedral angles are indicated by the greek letter Lineshape mobilityProtein Tumbling ( R)Viscosity ( ) Temperature (T) Radius (r)Backbone fluctuations ( B)Conformational changes Secondary structure Tertiary structureSpin label intrinsic mobility ( I)Label rotamericstates (MD) Torsional Energy barriers (Ab-initio?) Conformational switching rate (D) Figure 4-4. Scheme representing the factor s involved in the lineshape mobility.

PAGE 123

123 A B C D 20G A B C D 20G A B C D A B C D 20G 20G Figure 4-5. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease PMPR*SL in 2 mM NaOAc pH 5.0 as a func tion of spin label. All spectra were collected at 2 mW incident power and 100 G scans. A) MSL, B) MTSL, C) IAP, D) IASL.

PAGE 124

124 MTSLRitonavir DMSO Ritonavir DMSO IAP IASLRitonavir DMSO Ritonavir DMSO MSLA B C D 20G MTSLRitonavir DMSO Ritonavir DMSO Ritonavir DMSO Ritonavir DMSO IAP IASLRitonavir DMSO Ritonavir DMSO Ritonavir DMSO Ritonavir DMSO MSLA B C D 20G 20G Figure 4-6. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease PMPR*SL in 2 mM NaOAc pH 5.0 in the presence (red, 300 M Ritonavir) and absence (black, 10% DMSO) of inhibitor. A) MTSL, B) MSL, C) IAP and D) IASL.

PAGE 125

125 20GA B C D 20G 20GA B C D Figure 4-7. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease PMPR*MTSL as a function of salt concen tration.A) 2 mM NaOAc. B) 2 mM NaOAc + 50 mM NaCl. C) 2 mM NaOAc + 500 mM NaCl. D) 2 mM NaOAc + 2500 mM NaCl.

PAGE 126

126 [HCOOH] (mM)0 0.38 3.8 38 95 950 190 380 1900 2850 3800 0 38 95 950 190 380 1900 2850 3800MTSL IAP 20G 20GA B [HCOOH] (mM)0 0.38 3.8 38 95 950 190 380 1900 2850 3800 0 38 95 950 190 380 1900 2850 3800MTSL IAP [HCOOH] (mM)0 0.38 3.8 38 95 950 190 380 1900 2850 3800 0 38 95 950 190 380 1900 2850 3800 [HCOOH] (mM)0 0.38 3.8 38 95 950 190 380 1900 2850 3800 0 38 95 950 190 380 1900 2850 3800MTSL IAP 20G 20G 20G 20GA B Figure 4-8. Area normalized 100 G X-Band EP R spectra showing the effects of acid concentration (HCOOH) on the cw-EPR lineshape of PMPR*SL A).MTSL. B) IAP.

PAGE 127

127 2004006008001000 0.0044 0.0046 0.0048 0.0050 0.0052 0.0054 0.0056 -1 HCOOH Concentration (mM)0500100015002000 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 LF/CF Intensity (a.u.)HCOOH Concentration (mM)MTSL IAP A)B)MTSL IAP Figure 4-9. Graphical representa tion of spectral parameters obtained from the area normalized 100 G X-Band EPR spectra corresponding to the acidification of PMPR* (IAP/MTSL). A) Spectral second moment of IAP (red) and MTSL (blue) as a function of acid concentration. B) LF/CF intensity ratio of IAP ( ) and MTSL ( ) as a function of acid concentration. N H N H3C CH3 H3C CH3 O O NH C O HC NH C O CH2S N H N H3C CH3 H3C CH3 O O NH C O HC NH C O CH2S H+ H .N H N H3C CH3 H3C CH3 O O NH C O HC NH C O CH2S N H N H3C CH3 H3C CH3 O O NH C O HC NH C O CH2S H+ H Figure 4-10. Possible protonation reaction of the acetamido group in IAP. The proton preferentially attacks the oxygen in the C=O bond, breaking the electron delocalization between th e C=O group and the lone pair of the nitrogen. N C O N C O N C O H N C O H H+ Figure 4-11. Molecular orbital diagram for the acetam ido moiety in IAP. Molecular orbitals were calculated using B3LYP/6-31 G. Upon oxygen protonation, the orbital delocalization between the C=O and the lone pair of the n itrogen is broken, and the rotational barrier around the C-N bond is lowered, resulting in an increase in label mobility.

PAGE 128

128 L R L R L R L R L RA) B) Figure 4-12. Scheme for light pola rization and differential light absorption in the CD experiment. A) Equal amplitude components of left and right circularly polarized light generate plane polarized light. B) Differential absorpti on of left vs. right circularly polarized light generates elliptically polarized light. [Adapted from Kelly, S. M.; Jess, T. J.; Price, N. C. Biochim. Biophys. Acta 2005, 1751 119-39.] 180200220240260 [] deg cm2 dmol-1 (nm)180200220240260 [] deg cm2 dmol-1 (nm)A) B)180200220240260 [] deg cm2 dmol-1 (nm) 180200220240260 [] deg cm2 dmol-1 (nm)A) B) Figure 4-13. Sample circular dichroism spectra for A) alpha helix and B) beta sheet.

PAGE 129

129 200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm) MSL LAI200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm)MTSL LAI 200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm)IAP LAI 200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm)LAI IASL A D C B200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm) MSL LAI MSL LAI200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm)MTSL LAI MTSL LAI 200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm)IAP LAI IAP LAI 200210220230240250260 -8000 -4000 0 4000 8000 [] deg cm 2 dmol -1Wavelength (nm)LAI IASL LAI IASL A D C B Figure 4-14. Comparison between circular dichroism spectra of HIVPR LAI consensus sequence and PMPR*SL. A) MTSL, B) MSL, C) IAP and D) IASL. The black trace corresponds to each of the spin labeled prot eins and the red trace to CD spectra of active PMPR.130 Spectra were collected in 2 mM NaOAc pH 5.0 at 0.3 mg/mL protein concentration as determined by absorption at 280nm, using an extinction coefficient of 1.15 mg cm-1 ml-1. PMPR-K55MSL PMPR-D25N-K55MSL 20G PMPR-K55MSL PMPR-D25N-K55MSL PMPR-K55MSL PMPR-D25N-K55MSL 20G 20G Figure 4-15. Area normalized 100 G X-Band EPR spectra of HIV-1 protease Represented in black, PMPR+K55MSL (active protease). Repr esented in red, PMPR*MSL (inactive protease).

PAGE 130

130 0246810121416 0.34 0.36 0.38 0.40 0.42 0.44 0.46 Absorbance (au)Time (min) Figure 4-16. Degradation of chromogenic subs trate by active HIV-1 protease (PMPR+K55MSL) followed by UV/Vis spectroscopy. In red, ~40 nM PMPR+K55MSL+ 100 M chromogenic substrate. in the presence of 4: 1 Ritonavir:protease excess. In black, ~40 nM PMPR+K55MSL + 100 M chromogenic substrate.

PAGE 131

131 26 27 28 29 30 31 32 34 37 39 40 42 45t(days) t(days)0 6 7 8 10 13 11 12 14 20 18 22 23 25PQITLWKRPLVTIKIGGQLKEALLNTGADDTV IEEMSLPGRWKPKMIGGI GGFICVRQYDQII IEIAGHKAIGTVLVGPTPVNIIGRNLLTQIGATLNF 20G 20G 26 27 28 29 30 31 32 34 37 39 40 42 45t(days) t(days)0 6 7 8 10 13 11 12 14 20 18 22 23 25PQITLWKRPLVTIKIGGQLKEALLNTGADDTV IEEMSLPGRWKPKMIGGI GGFICVRQYDQII IEIAGHKAIGTVLVGPTPVNIIGRNLLTQIGATLNF 26 27 28 29 30 31 32 34 37 39 40 42 45t(days) t(days)0 6 7 8 10 13 11 12 14 20 18 22 23 25PQITLWKRPLVTIKIGGQLKEALLNTGADDTV IEEMSLPGRWKPKMIGGI GGFICVRQYDQII IEIAGHKAIGTVLVGPTPVNIIGRNLLTQIGATLNF 20G 20G 20G 20G Figure 4-17. Area normalized 100 G X-Band EP R spectra of HIVPR PMPR+K55MSL undergoi ng self-proteolysis followed by cwEPR. The appearance of a sharp feature in the high field side indicated by the green arrow is consistent with a new spectral component for which the spin label is attached to a small peptid e that appears as a result of the degradation of the protease in the flap region.

PAGE 132

132 01020304050 0.0 0.1 0.2 0.3 0.4 HF/LF RatioTime (days) Figure 4-18. Intensity ratio of th e high field line/low field line ( ) for the area normalized, 100 G X-Band cw-EPR spectra of HIV-1 pr otease PMPR+K55MSL undergoing selfproteolysis. In red, best fit to a growing exponential function. Day 0 Day 46 20G Day 0 Day 46 Day 0 Day 46 Day 0 Day 46 20G 20G Figure 4-19. Area normalized 100 G X-Band EPR spectra of HIVPR PMPR+K55MSL on day 0 and day 46 of the self-degradation kinetic study. Protease sample was kept at 37 C.

PAGE 133

133 OH HO OH O H OH H O H OH H OH CH2OH H CH2OH H CH2OH OH H H O HO O OH nA) B) C)OH HO OH O H OH H O H OH H OH CH2OH H CH2OH H CH2OH OH H H O HO O OH nA) B) C) Figure 4-20. Structure of three compounds utilized to study viscosity and osmolality effects. A) Glycerol, B) PEG3000, C) sucrose. Ficoll 400 is not represented because it is a highly branched epichlorohydrin-sucrose crosslinked polymer. 01020304050 1 2 3 4 5 6 Viscosity (cS)% Content (w/v) Ficoll 400 PEG 3000 Sucrose Glycerol01020304050 100 1000 10000 Osmolality (mmol kg-1)% Content (w/v) Ficoll 400 PEG 3000 Sucrose GlycerolA B01020304050 1 2 3 4 5 6 Viscosity (cS)% Content (w/v) Ficoll 400 PEG 3000 Sucrose Glycerol Ficoll 400 PEG 3000 Sucrose Glycerol01020304050 100 1000 10000 Osmolality (mmol kg-1)% Content (w/v) Ficoll 400 PEG 3000 Sucrose Glycerol Ficoll 400 PEG 3000 Sucrose GlycerolA B Figure 4-21. Plot of the viscosit y (cS) and osmolality (mmol kg-1) as a function of percent content (w/v) for the four solutes under study: glycer ol, Ficoll 400, sucrose, PEG3000. A) Viscosity (cS) vs. percent c ontent (w/v) for gly cerol, sucrose, PEG 3000 and Ficoll 400. B) Osmolality (mmol kg-1) vs. percent content (w/v) for glycerol, sucrose, PEG 3000 and Ficoll 400.

PAGE 134

134 123456 100 1000 10000 0 3 6 9 6 12 18 6 12 18 24 12.61 25.22 37.83 50.44Osmolality (mmol kg-1)Viscosity (cS) Ficoll 400 PEG 3000 Sucrose Glycerol123456 100 1000 10000 0 3 6 9 6 12 18 6 12 18 24 12.61 25.22 37.83 50.44Osmolality (mmol kg-1)Viscosity (cS) Ficoll 400 PEG 3000 Sucrose Glycerol Ficoll 400 PEG 3000 Sucrose Glycerol Figure 4-22. Osmolality (mmol kg-1) vs. viscosity (cS) for solutes commonly used in cw-EPR. The numbers on each data point represen t the percent (w/v) for each sample.

PAGE 135

13505101520 0 5 10 0246810 -4 -2 0 2 4 05101520253035404550 0 10 20 30 40 50 051015202530 0 5 10 15 20 25 Hpp % ChangeSucrose %(w/v) Hpp % ChangeFicoll %(w/v) Hpp % ChangeGlycerol %(w/v) Hpp % ChangeP3K %(w/v) Hppchanges as a function of cosoluteSucrose Ficoll GlycerolPEG3000Change MTSL Change MSL Change IAP Change IASL A B C D05101520 0 5 10 0246810 -4 -2 0 2 4 05101520253035404550 0 10 20 30 40 50 051015202530 0 5 10 15 20 25 Hpp % ChangeSucrose %(w/v) Hpp % ChangeFicoll %(w/v) Hpp % ChangeGlycerol %(w/v) Hpp % ChangeP3K %(w/v) Hppchanges as a function of cosoluteSucrose Ficoll GlycerolPEG3000Change MTSL Change MSL Change IAP Change IASL Change MTSL Change MSL Change IAP Change IASL A B C D Figure 4-23. Changes in cen tral field linewidth ( Hpp) for HIV-1 PMPR*SL (SL = MTSL( ), MSL( ), IAP( ), IASL( )) as a function of cosolute species and co ncentration. A) Sucrose. B) Fi coll400. C) Glycer ol. D) PEG3000.

PAGE 136

1360246810 -4 -2 0 2 4 6 8 10 12 14 05101520253035404550 0 5 10 15 051015202530 0 5 10 15 05101520 0 5 10 15

% ChangeFicoll %(w/v)

% ChangeGlycerol %(v/v)

% ChangeP3K %(w/v)

% ChangeSucrose %(w/v)Change MTSL Change MSL Change IAP Change IASL Change MTSL Change MSL Change IAP Change IASL

changes as a function of cosoluteSucrose Ficoll GlycerolPEG3000 AB C D Figure 4-24.Changes in spectral second moment

for HIV-1 PMPR*SL (SL = MTSL( ), MSL( ), IAP( ), IASL( )) as a function of cosolute species and co ncentration. A) Sucrose. B) Fi coll400. C) Glycer ol. D) PEG3000.

PAGE 137

137051015202530 0 5 10 15 20 25 02468101214 0 2 4 6 8 10 12 14 05101520253035404550 0 10 20 30 05101520 0 5 10 15 CF/LF % ChangeP3K %(w/v) CF/LF % ChangeFicoll %(w/v) CF/LF % ChangeGlycerol %(v/v) CF/LF % ChangeSucrose %(w/v)CF/LF Intensity ratio changes as a function of cosoluteChange MTSL Change MSL Change IAP Change IASL Sucrose Ficoll GlycerolPEG3000 AB C D051015202530 0 5 10 15 20 25 02468101214 0 2 4 6 8 10 12 14 05101520253035404550 0 10 20 30 05101520 0 5 10 15 CF/LF % ChangeP3K %(w/v) CF/LF % ChangeFicoll %(w/v) CF/LF % ChangeGlycerol %(v/v) CF/LF % ChangeSucrose %(w/v)CF/LF Intensity ratio changes as a function of cosoluteChange MTSL Change MSL Change IAP Change IASL Change MTSL Change MSL Change IAP Change IASL Sucrose Ficoll GlycerolPEG3000 AB C D Figure 4-25. Changes in LF/CF intensity ratio for HIV-1 PMPR*SL(SL = MTSL( ), MSL( ), IAP( ), IASL( )) as a function of cosolute species and concentration. A) Sucr ose. B) Ficoll400. C) Glycerol. D) PEG3000.

PAGE 138

138 0.00.51.01.52.0 0.75 0.80 0.85 0.90 0.95 1.00 Time (s)Echo Intensity (a.u.)PMPR+D25N+K55MTSL Glycerol Ficoll2030405060 P(r)Distance () Glycerol Ficoll0.00.51.01.52.0 0.75 0.80 0.85 0.90 0.95 1.00 Time (s)Echo Intensity (a.u.)PMPR+D25N+K55MTSL Glycerol Ficoll Glycerol Ficoll2030405060 P(r)Distance () Glycerol Ficoll2030405060 P(r)Distance () Glycerol Ficoll Glycerol Ficoll Figure 4-26. DEER spectra of PMPR*MTSL as a function of cryoprotectan t added (glycerol and Ficoll 400). 050001000015000 Intensity (a.u.)Time (ns) H2O+Ficoll D2O+Ficoll D2O+Glycerol 050001000015000 Intensity (a.u.)Time (ns) H2O+Ficoll D2O+Ficoll D2O+Glycerol H2O+Ficoll D2O+Ficoll D2O+Glycerol Figure 4-27. Tm measurements for HIV-1 PMPR*MTSL as a function of the composition of the cryoprotectant solution; represented in blue, H2O + Ficoll. In black, D2O + 30% glycerol and, in red, D2O+Ficoll.

PAGE 139

139 CHAPTER 5 DOUBLE ELECTRON-ELECTRON RESONANCE MEASUREMENTS ON THE FLAP REGION OF HIV-1 PROTEASE Preliminary Studies In this chapter, the re sults of using double electr on-electron resonance (DEER) experiments on HIV-1 PMPR*SL (SL=MTSL, MS L, IAP and IASL) to characterize the conformations of the flap region will be discusse d. As it was previously mentioned in Chapter 3, residue K55 tolerates a la rge variety of mutations without a ffecting the protease activity; hence, distance measurements between residues K55C-K 55C (one on each flap of the homodimer) will be obtained by Tikhonov regularizat ion (TKR) and Montecarlo (MC) analysis of the DEER dipolar evolution data for each of the spin labels under study. In an attempt to predict the distance distribut ion profiles that will be obtained via DEER spectroscopy, we performed a detailed anal ysis of the distances between all atoms (C, C, C C C and Nz) between the lysine residues K55-K55 of all the X-ray crystal structures and NMR structures of HIV-1 protease submitted to the PD B databank. The result of this analysis is graphically summarized in Figur e 5-1. Figure 5-1 shows the di stances between atoms in the lysine side chain at positions K55 and K55 It can be seen from figure 5-1A) that C C spread in distance value is relatively narrow, and cente red on 22. A few structures deviate from this distance and correspond to (1) crystal structures for the MDR769 mutant (multi-drug resistant) crystallized in what was hypothe sized to be an open confor mation of the flaps, and (2) structures corresponding to fullere nes and metal complexes bound to HIVPR. It can be seen that, as we move toward the nitrogen atom in the ly sine side-chain, the dist ance distribution becomes more heterogeneous and the average distance between atoms shifts to longer distances (22 for C to C, 24 for C to C, 25 for C to C, 27 for CC, 28 for CC and 30 for Nz to Nz). A comparison between the structure of K 55 and C55MTSL can be qualitatively made by

PAGE 140

140 assuming a fully extended chain for both lysine and MTSL-labeled cysteine. Figure 5-2 shows the comparison between the structures of Lysine and MTSL labeled cysteine. This figure shows that, assuming a fully extended chain model and ne glecting the difference in length between C-C and S-S bonds, the distances obtained between diffe rent atoms in the lysine residue can be compared to distances between different atoms in the spin-label tether. Nonetheless, the maximum distance obtained between Lysines K55-K55 (atoms Nz-Nz) only corresponds to the carbon attachment between the li nker region and the five memb ered ring that supports the nitroxide moiety. Therefore, assuming that the distance between said car bon and the nitrogen in the nitroxide group is roughly ~3 the average distances expected from a distance measurement experiment should be in the range of 30-36 Distances in the range between 7 -60 can be measured by EPR techniques. For distances <7 an accurate distance rule r is impossible to obtain due to the strong dependence of the cwEPR lineshape on the orbital overl ap (via Heisenberg exchange)82,141. In the range between 8-25 for which dipolar interactions dominate the cw-EPR lineshape, distance measurements can be performed by deconvolution-convol ution of the cw-EPR linesha pes of singly-labeled and doubly-labeled samples142. Distances between 25-60 can be measured by pulsed-EPR techniques, which are able to separate the cont ribution of the electronelectron dipolar coupling from the rest of the interactions in the spin Hamiltonian (Chapter 2). From our previous analysis of the X-ray crystal structure dist ances between Lysine55 residues (30 -36 ), we conclude d that pulsed-EPR was the technique of choice for the expected distances between probes. Experi mental evidence of distances greater than 25 can also be obtained from examination of the cw-EPR lin eshapes of HIVPR+PMPR*SL. In Chapter 4, Figure 4-5, we showed that the cw-EPR lines hape did not reflect any changes between unbound

PAGE 141

141 and inhibited (Ritonavir) protease. To remind the reader of the HIVPR spectral lineshapes in the presence and absence of inhibitor, the lines hape corresponding to PMPR*MTSL is shown in Figure 5-3.The absence of dipolar broadening in the cw-EPR lineshape in Figure 5-3 indicates that the distance between spin label probes in the doubly labeled PMPR*MTSL is greater than 25 even in the conformation in which the flaps are closed and locked on top of the inhibitor (Ritonavir). Double Electron-Electron Resonance Experiments on PMPR+D25N+K55(SL) We determined from comparison of the cw-EPR lineshapes in the presence and absence of inhibitor (Figures 4-5 and 53) that X-Band cw-EPR spectro scopy is unable to distinguish between protease in the inhibi ted state (flaps closed) and the uninhibited states (flap conformation unknown). As it was mentioned in Chapter 4, the cw-EPR lineshape is a convolution of three differe nt types of motion, name ly (1) protein tumbling ( R), (2) conformational changes in the backbone ( B) and (3), rotational correlation time of the spin label due to interconversion between different rotamers ( I). We also determined that the addition of a high viscosity and low osmolality solution of Fi coll 400 did not significan tly alter the cw-EPR lineshapes of spin-labeled HIV-1 prot ease, discarding the contribution of R to the lineshape. Hence, the resulting lineshape of HIV-1 PMPR *SL has to be a convolution of backbone motion and spin label motions. Nonetheless, it is known fr om X-ray crystal structures as well as NMR and MD that upon binding of inhibitor, the flaps lock on top of the inhibitor and the only flap motion observed are small backbone fluctuations Upon binding of inhibitor, no changes were observed in the cw-EPR lineshape, and therefore we can conclude that the cw-EPR lineshape is only reporting label motion and not backbone conforma tional changes. It is also noteworthy that, as we previously mentioned in this chapter, cw-E PR lineshape of spin label at distances below 25 should present dipolar broadeni ng effects. No dipolar broade ning was seen for either the

PAGE 142

142 unbound or the inhibitor (Ritonavi r) bound protease, therefore we conclude that, in either conformation, the spin labels ar e separated by distances >25 In order to obtain distance measurements between PMPR*SL spin labeled residues, a technique known as 4p-DEER was utilized. As reviewed in Chap ter 2, DEER spectroscopy is a pulsed-EPR technique which is able to extract distance information from the spin Hamiltonian by decoupling the dipolar interactio n from the rest of the intera ctions contained in the spin Hamiltonian. This is achieved by partially refocusing a Hahn spin echo, with the aid of a pump pulse which changes position over time. Analysis of the dipolar evolution signal obtained by DEER, using either Tikhonov regularization95,96 or Montecarlo fitting, results in average distances and distance distributions be tween both labeled cysteine residues. As we previously saw in Chapter 4, the cw-E PR lineshape is highly dependent on the spin label of choice. This indicates that different spin labels differ in terms of mobility and conformational heterogeneity, and th erefore we can expect that the distance distributions that can be obtained by DEER will be highly dependent on the choice of spin label. In an effort to decouple spin label conformationa l heterogeneity (due to intrin sic differences in the label chemical composition and structure) from protea se backbone motion, all distance measurements on PMPR*SL were performed with four different labels, namely MTSL, MSL, IAP and IASL.143 Typical experiment conditions cons isted of protein samples (70 20 M) containing glycerol, loaded into 4 mm O.D. EP R tubes and flash frozen in liquid N2. HIV-1 protease PMPR+D25N+K55C was chemically m odified at the cysteine in pos ition 55 using four different spin labels, namely 3-(2-Iodoacetamido)-P ROXYL (IAP), 4-Maleimido-TEMPO (MSL) ,(1Oxyl-2,2,5,5-Tetramethyl3-Pyrroline-3-Methyl) Methanethi osulfonate (MTSL) and 4-(2Iodoacetamido)-TEMPO (IASL). For further info rmation on the spin label reaction and spin

PAGE 143

143 labeling conditions, the reader is referred to Chapter 3. Cw-EPR, field-swept echo, Tm, DEER Setup, field swept DEER echo and DEER expe riments were performed sequentially at 65 K using a Bruker E580/E680 X-Band pulsed spectrom eter operating at frequencies near 9.7 GHz. For inhibited samples, a 1:4 molar ratio of pr otein: Ritonavir was used. (See Chapter 2 and Appendix B for further information about the D EER experiment). The data reported in this chapter were collected at the National High Magne tic Field Lab, with the collaboration of Marco Bonora and P. Fajer. Figures 5-2 to 5-5 plot the DEER dipolar evolution curves (A), distance distribution profiles (C) and L-curve plot s for the unbound and the inhibitor bound (D) for HIV-1 PMPR*SL for SL= MTSL, MSL, IAP and IASL. For these figur es, the dipolar evolution curves have been analyzed via the Tikhonov regularization pr ocedure (see Chapter 2 for full details). Figure 5-4 plots the results obtained for the distance distribution an alysis of PMPR*MTSL In figure 5-4A, the dipolar evolution curves for unbound (blue) and Ritonavir bound (red) are plotted. This figure shows that the dipolar evolution of the inhibited samples is dramatically different than that of the uni nhibited ones, with strong dipolar evolution oscillations, which indicate narrow distance distri bution profiles. Figure 5-4C sh ows the distance distributions obtained via TKR from the dipolar evolution data. It can be seen that the distance distribution for the inhibited protease (red) is na rrow and centered at 32.6 In th e absence of inhibitor, this distance distribution is broade ned (spanning distances between 24 48 ), and the average distance is shifted to longer distances by r oughly 3 This is consistent with the C-C distance changes observed in the X-ray cr ystal structures between the inhibited (e.g., 1HVI) and the uninhibited (1HHP)144 forms of the protease. Figure 5-4B s hows the L-curve plot for the sample in absence of inhibitor. The regularization parameter in this case is 104 vs. 101 for the inhibited

PAGE 144

144 protease. This large difference be tween regularization parameters is most likely due to the wide spread of the distance distributi on of the uninhibited protease and the narrow dist ribution of the Ritonavir-bound protease. In an analogous ma nner, Figure 5-5 shows the DEER results for HIVPR PMPR*IAP. In this case, the dipolar oscill ations for the inhibited protease (Figure 5-5A) are not as marked as the oscill ations for MTSL labeled protease. This is consistent with the distance distributions obtained by TRK regulariz ation (Figure 5-5C), for which the distance distribution in the Ritonavir bound form of the pr otease is wider than in the MTSL counterpart. Overall, less change in the dist ance distribution width is observed when IAP is used to label HIV protease. It is important to poi nt out the fact that the averag e distances for the inhibitor bound protease (32.4 ) shifts to longer distances (by 3 ) when inhibitor is absent from the sample (to 35.4 ). These results are again consistent with the distance changes between K55C-K55C obtained from X-ray crystal structures and with the DEER results obtained for MTSL label. Nonetheless, the change in the full width at ha lf max (FWHM) of the distance distribution is smaller for IAP (2.3 ) than for MTSL (7.4 ). Figure 5-6 plots the results obtained from the di stance analysis from dipolar evolution data for the sample PMPR*MSL. Figure 5-6A shows the same trend in the dipolar evolution data as MTSL and IAP. The dipolar evolution data co rresponding to the sample containing inhibitor presents stronger oscillations in the dipolar evolution than the uninhibited sample, and this inconsistent with the narrower distance distri bution obtained by TRK (Figure 5-6C). For MSL, the average distance distribution shift between th e inhibited (34.7 ) a nd the uninhibited (37.0 ) is ~2.3 This distance change is somewhat smaller than the change observed in MTSL and IAP, and might be related to the unusual structure of the MSL spin label. Figure 5-6C also shows a change in the FWHM between the inhibited an d uninhibited forms of 5.7 Again, L-curve

PAGE 145

145 values as a function of regular ization parameter obtained for the free (Figure 5-6B) and the inhibited (Figure 5-6D) present the same trend as of those of MTSL and IAP; the value of the regularization parameter is higher when the distance distribu tion is wider. (103 for the unbound vs. 101 for the Ritonavir bound protease). Last, Figure 5-7 plots the re sults obtained from the eval uation of the DEER dipolar evolution (5-7A) for PMPR*IASL. We saw at the beginning of Chapter 4 that, based on simple geometric considerations such as the length of the tether that links the five-membered ring in which the nitroxide moiety sits as well as th e number of dihedral rotable bonds, IASL was expected to be the most mobile of the spin labe ls. Indeed, Figure 5-7A shows a dipolar evolution function for the protease bound to inhibitor that la cks the strong oscillations of its MTSL, MSL and IAP counterparts. The analysis of this dipolar evolution by TKR (Figure 5-7C) shows that there is no change in the FWHM between the inhi bited and the uninhibited form of the protease. This indicates that the flexibility of the spin label is responsible for the width of the distance distribution, as opposed to conformational changes in the protease flap backbone. Nonetheless, it is noteworthy to mention that even in the very fl exible case of IASL, there is an average distance change between the uninhibited (34.3 ) and the inhibitor bound form (29.8 ) of the protease of approximately 4.5 which is also in agreement with the changes expected from considering the x-ray crystal structures of HIV-1 pr otease deposited in the PDB databank. The L-curves (Figure 5-7B and 5-7D) show a high value for the regularization parameter, consistent with wide distan ce distributions (inhibited, 103, uninhibited 104). Figures 5-4B to 5-7B show that all distance distributions for uninhi bited HIV protease are broad, spanning distances from 24 to 48 As it ha s been shown in Figure 5-2 to 5-5, there is a correlation between the value of the regulariza tion parameter and the width of the distance

PAGE 146

146 distribution. This can pose a problem when the S/N of the dipolar evolution is low, and can introduce errors in the choice of regularization parameter, generating an oversmoothed and artificially broadened distance dist ribution. In order to ensure that these distance distributions are not broadened by the choice in the Tikhonov regula rization parameter, we performed detailed analysis for each of the samples PMPR*SL (SL=MTSL, MSL, IAP, IASL) by obtaining a distribution profile from the di polar evolution functions usi ng both TKR and MC methods. The differences in the solutions (ave rage distances and distance distri butions) obtained via either MC or TKR for each DEER dipolar evolution, can be us ed as an indication of the error in the solution due to artifacts in the fitting procedure. A comparison between the so lutions obtained by TKR and MC for PMPR*SL for each of the labels used in this study can be seen in Figure 5-8 Figure 5-8 plots the results obtained from an alyzing DEER dipolar evolution functions from HIV-1 PMPR*SL (SL=MTSL, MSL, IAP and IASL) by using either Tikhonov regularization or Montecarlo fi tting. Slight discrepancies in the distances and distance distributions (such as the differe nces in the inhibited form of MSL) arise from the presence of noise in the TKR regularized so lution. It is important to remind the reader that Tikhonov Regularization does not assume a functional form of the distance distribution, but a regularization parameter must be chosen to ensure that the distance distri bution profile resulting from the regularization procedure is well-behave d. On the other hand, Montecarlo techniques are relatively computationally inexpe nsive and there is no need for distance distribution smoothing, but the functional form of the solution (e.g. Gaussian, Lorentzian) must be known or assumed beforehand. From Figure 5-8, it can be concluded that the dist ance distributions calculated by either TKR or MC are in good agreement, a nd both techniques to analyze results are complementary and can be used in order to esti mate the error due to the fitting procedure.

PAGE 147

147 Nonetheless, whereas comparing TKR with MC solutions might give an estimate of the goodness of fit, it does not account for the error due to the experimental S/N ratio for each sample. In an attempt to evaluate the error in th e distance distribution due to experimental noise, we used DeerSim to generate a set of dipolar evolution curves corresponding to a small shift in the optimal distance distribution (as determined by TKR), while maintaining the FWHM of the distance distribution profile. An example for th e regularization of PMPR*MTSL with Ritonavir is given in figure 5-9. Figure 5-9 shows the change in the dipolar evolution as a f unction of the center position in the distance distribution. It can eas ily be seen (Figure 5-9, top) th at changes in distances of 1.52.5 generate a dipolar evoluti on function that is not compatible with the experimental data. When the distance changes are more subtle (Figur e 5-7, bottom), the generated dipolar evolution curves are more compatible with the solution, although distances of 32. 0 significantly differ from the best fit (32.5 ). With these data and the error obtained from the regularization method (Figure 4-22), we can estimate an error in the average distance of <0.5 The distance distributions obtained for all spin labels, using either TKR or MC fitting are reported in Table 5-1. Table 5-1 shows distance s obtained from TKR and MC analysis of the DEER echo curves for spin labels attached to the flap positions K 55-K55 in PMPR* HIV-1 protease. The average distances and distance dist ributions in Table 5-1 for the inhibited and uninhibited forms can be plotted as a function of spin label and distance (Figure 5-10). Figure 510 plots the distance and distance di stributions for HIVPR PMPR*SL +300 M Ritonavir (Figure 5-10A), and HIVPR PMPR*SL in the absence of inhibitor.

PAGE 148

148 Table 5-1. EPR distance measuremen ts for K55SL-K55SL HIV protease Distance ( ) (.5 ) Distance () FWHM () FWHM () Label Uninhibited Ritonavir uninhibitedRitonavir uninhibited Ritonavir uninhibitedRitonavir MTSL 35.5 (35.5) 32.6 (32.8) 2.9 (2.7)10.4 (9.3)3.0 (2.7) 7.4 (6.6) MSL 37.0 (36.7) 34.7 (34.7) 2.3 (2.0)8.9 (8.5)4.2 (4.6) 4.7 (3.9) IAP 35.4 (35.3) 32.4 (31.8) 3.0 (3.5)6.8 (6.3)4.5 (4.5) 2.3 (1.8) IASL 34.3 (34.5) 29.8 (29.7) 4.5 (4.8)7.3 (8.6)7.2 (8.2) 0.1 (0.4)Note: Distances correspond to TKR solutions and, in parenthesis, to MC solutions Several key points can be made about Figure 5-10, and we will discuss each of them in detail: (1) For all spin labels under study, there is a conformational change of the flaps that is reflected by a shift in the averag e distance from the inhibited to the uninhibited form of ~3-4 This shift in distance occurs fo r all spin labeled samples under st udy, and is consistent with the C-C distances obtained from X-ray crystal struct ures deposited in the PDB databank. (2) For MTSL, MSL and IAP, the conformational change in the flaps between the uninhibited and the inhibited protease is also reflected by a narrowi ng of the distance distribution. For the samples under study, protease labeled with MTSL shows the largest change in FWHM (7.4 ), followed by MSL (4.7 ) and IAP (2.3 ). It is importa nt to note that change s in the FWHM of the distance distribution profile for the IASL spin-labeled protease sample were not present. A plausible explanation is the high mobility of the spin label even when the protease is in the inhibited state, with the motion of the spin la bel masking flap backbone motions. Nonetheless, IASL also shows a change in the average distance of 4 indicating that this change is due to conformational heterogeneity of th e backbone and not of the spin la bel. (3) Information about the conformational heterogeneity of th e spin label can be extracted from the width of the distance distribution profile in th e inhibited form. It has been show n in the literature that, for HIV protease bound to inhibitors, the flaps are confor mationally locked on top of the inhibitor, and the only backbone motions present are due to sma ll amplitude fluctuations. From the FWHM of

PAGE 149

149 Ritonavir-bound samples, we can establish that the label mobility follows the trend MTSL) and LF/CF intensity ratio, the label mobility follows the trend MSL, c and LF/CF intensity ratio (Figure 5-11). Figure 5-11A plots the FWHM obtained via TKR from DEER dipolar evolution curves for HIV-1 PMPR*SL (SL=MTSL, MSL, IAP, IASL), aligned at the average distance. As it was previously discussed, the width of the distance distribution for inhibited PMPR*SL follows the trend MTSL
PAGE 150

150 / 0 x t y yAe (5-1) The values for the parameters in Equation (5-1) were y0=0.59, A =0.004 and t =1.63. Figure 5-11D plots the FWHM obtained from TKR regularization of th e DEER dipolar evolution of PMPR*SL+Ritonavir and the spectral second moment (

). It can be seen that MSL deviates from the values obtained for MTSL, IAP and IA SL. Although the relative values of MTSL and IAP fit the overall mobility trend of the LF/CF inte nsity ratio, the value of IASL is too high. The results in Chapter 4 showed that second moment might not be sensitive enough for small changes in the dynamics and conformations of the spin label. Last, Figure 5-11E plots the FWHM obtained from TKR regularization of the DEER dipolar evolution of PMPR*SL+Ritonavir and the central peak-to-peak width ( Hpp, in Gauss). This plot shows that the Hpp obtained for IASL is higher than that of MTSL, which is not consistent with the mobility trend expected (IASL is more mobile than MTSL), whereas th e FWHM of the distance distribution profile of the inhibitor-bound protease is narrower for MTSL than for IAP. This indicates that the MTSL spin label is less mobile than its IA SL counterpart. It can be argued that Hpp might not be a reliable spectral parameter when different spin la bels are compared, but further research in this area will be needed to obtain reliable conclusions. We mentioned in Chapter 4 that the mobility of the spin label is related to two factors: (1) rate of motion of the label, which can be rationa lized in terms of a macr oscopic diffusion tensor, and (2) conformational heterogeneity and rotatio nal energy barriers for each of the dihedral angles in the spin label. Understanding the conf ormational heterogeneity of the different spin labels will undoubtedly help in understanding what role the spin label rotamers play in the mobility of the spin label.

PAGE 151

151 In order to validate our DEER measurements, as well as to validate the technique used to generate HIV structures usi ng molecular dynamics, Simmerling et al. performed calculations on HIV PMPR*MTSL.145 The results of this study can be s een in Figure 5-12. Figure 5-12 plots the comparison between HIVPR PMPR*MTSL distan ce distribution profil es obtained by TKR regularization of the dipolar evolution f unction obtained by DEER spectroscopy and the distances obtained from MD simulations. Fi gure 5-12A plots the co mparison between the distance distribution profiles for HIV-1 protease in the presence of inhibitor obtained by means of either DEER spectroscopy or MD simulations. It can be seen that the MD results are in very good agreement with the results obtained from TKR regularization of the dipolar evolution signal. In both cases, the distances are confined between 29 and 37 with the average distance for the TKR distribution profile centered at 32.6 and the distan ce from MD results centered at 33.7 It can also be seen that the widths of the distance distribu tion profiles are in good agreement, with a FWHM of 3.0 for the TKR results and 2.9 for the MD simulations. The small populations of distances below 28 were not observed in the DEER experiments, and this is most likely due to the low population of thos e states or inaccuraci es in the force-field parameters for the spin label. Figure 5-12B pl ots a comparison between the distance distribution profile of the uninhibite d conformation of HIV-1 protease as determined by TKR regularization of HIVPR PMPR*MTSL and a hypothe tical structure calculated by MD in which the inhibitor has been removed from the active site but in which the flaps are still closed on top of the active site cavity. Figure 5-12C compares the distance distribution profile obtained for HIVPR+PMPR*MTSL from MD simula tions in the absence of i nhibitor (black) and the TKR solution for the dipolar evoluti on obtained in the DEER expe riment. From MD simulations, Simmerling et al. observed that the ensemble of distan ces obtained by TKR of the uninhibited

PAGE 152

152 HIVPR sample corresponded to distances that were not reported by any crystal structure. Furthermore, this ensemble can be accurately represented by MD simulations that include in the conformational ensemble a fully open, active-s ite accessible HIV protease. The discrepancies arising at long distances for th e open conformation can be due to incomplete space sampling during the MD simulations and the parameters util ized to describe the spin label in the force field. Another possible explanation is the use of cryoprotectants in the EPR experiments, as well as numerical artifacts that might be in cluded in the Tikhonov re gularized solution.

PAGE 153

153050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #A B CD E F C-CC-CC-CC-CC-CNz-Nz C CH2 H2C CH2 H2C NH2 O OH H NH2 CNzCCCCLys55050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #A B CD E F C-CC-CC-CC-CC-CNz-Nz C CH2 H2C CH2 H2C NH2 O OH H NH2 CNzCCCC050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #A B CD E F C-CC-CC-CC-CC-CNz-Nz 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 050100150200250 20 25 30 35 Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #Distance ()Structure #A B CD E F C-CC-CC-CC-CC-CNz-Nz C CH2 H2C CH2 H2C NH2 O OH H NH2 CNzCCCCC CH2 H2C CH2 H2C NH2 O OH H NH2 CNzCCCCLys55 Figure 5-1. Distance analysis betw een residues K55-K55 of HIV-1 protease from X-ray crystal stru ctures in the PDB databank. A) C to C distances. B) C to C distances. C) C to C distances. D) CC distances. E) CC distances and F) Nz to Nz distances.

PAGE 154

154 C CH2 H2C CH2 H2C NH2 O OH H NH2 C CH2 S S H2C O OH H NH2 N O C CH2 H2C CH2 H2C NH2 O OH H NH2 C CH2 S S H2C O OH H NH2 N O K55 C55MTSL Figure 5-2. Comparison between the st ructures of Lys55 and Cys55MTSL. Ritonavir DMSO Ritonavir DMSO Ritonavir DMSO Figure 5-3. Area normalized 100 G X-Band EPR lineshape for PMPR*MTSL, in the presence (red) and absence (bl ack) of inhibitor.

PAGE 155

155 -3.40-3.35-3.30-3.25-3.20-3.15 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 25000 50000 75000 100000 solution norm ||x||2residual norm ||Ax-b||20.00.51.01.52.0 Echo Intensity (a.u.)Time (s)2030405060 P(r)Distance ()-4.0-3.8-3.6-3.4-3.2-3.0-2.8 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 50 100 250 500 750 1000 5000 10000 50000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir FreeA C B D-3.40-3.35-3.30-3.25-3.20-3.15 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 25000 50000 75000 100000 solution norm ||x||2residual norm ||Ax-b||20.00.51.01.52.0 Echo Intensity (a.u.)Time (s)2030405060 P(r)Distance () -4.0-3.8-3.6-3.4-3.2-3.0-2.8 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 50 100 250 500 750 1000 5000 10000 50000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir FreeA C B D Figure 5-4. DEER results for HIVPR PMPR*M TSL. A) Background subtracted dipolar echo evolution curves for K55-K55 MTSL in th e absence (blue) and the presence (red) of Ritonavir (inhibitor). The solid lines overlaid on the e xperimental data represent the regenerated echo curves from TKR analysis B) L-curve corresponding to the dipolar evolution of the sample in the absence of inhibitor. C) Distance distributions obtained from A). D) L-curve corresponding to the di polar evolution of the sample in the presence of inhibitor.

PAGE 156

156 Ritonavir-4.3-4.2-4.1-4.0-3.9-3.8 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 50 250 1000 5000 10000 50000solution norm ||x||2residual norm ||Ax-b||2Free-2.40-2.35-2.30-2.25 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 250 1000 5000 10000 50000residual norm ||Ax-b||2solution norm ||x||22030405060 P(r)Distance () 0.00.51.01.52.02.5 Echo Intensity (a.u.)Time (s)A C B DRitonavir-4.3-4.2-4.1-4.0-3.9-3.8 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 50 250 1000 5000 10000 50000solution norm ||x||2residual norm ||Ax-b||2Free-2.40-2.35-2.30-2.25 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 250 1000 5000 10000 50000residual norm ||Ax-b||2solution norm ||x||22030405060 P(r)Distance () 0.00.51.01.52.02.5 Echo Intensity (a.u.)Time (s)A C B D Figure 5-5. DEER results for HIVPR PMPR*I AP. A) Background subtracted dipolar echo evolution curves for K55-K55 IAP in the absence (blue) and the presence (red) of Ritonavir (inhibitor). The solid lines overlaid on the e xperimental data represent the regenerated echo curves from TKR analysis B) L-curve corresponding to the dipolar evolution of the sample in the absence of inhibitor. C) Distance distributions obtained from A). D) L-curve corresponding to the dipolar evolution of the sample in the presence of inhibitor.

PAGE 157

157 2030405060 P(r)Distance ()0.00.51.01.52.02.53.0 Echo Intensity (a.u.)Time (s)-3.9-3.8-3.7-3.6-3.5-3.4-3.3-3.2 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir-4.0-3.9-3.8-3.7-3.6 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2FreeA C B D2030405060 P(r)Distance () 0.00.51.01.52.02.53.0 Echo Intensity (a.u.)Time (s)-3.9-3.8-3.7-3.6-3.5-3.4-3.3-3.2 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir-3.9-3.8-3.7-3.6-3.5-3.4-3.3-3.2 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir-4.0-3.9-3.8-3.7-3.6 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Free-4.0-3.9-3.8-3.7-3.6 -24 -22 -20 -18 -16 -14 -12 -10 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2FreeA C B D Figure 5-6. DEER results for HIVPR PMPR*M SL. A) Background subtracted dipolar echo evolution curves for K55-K55 MSL in the absence (blue) and the presence (red) of Ritonavir (inhibitor). The solid lines overlaid on the e xperimental data represent the regenerated echo curves from TKR analysis B) L-curve corresponding to the dipolar evolution of the sample in the absence of inhibitor. C) Distance distributions obtained from A). D) L-curve corresponding to the dipolar evolution of the sample in the presence of inhibitor.

PAGE 158

158 -3.6-3.5-3.4-3.3 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Free2030405060 P(r)Distance ()-3.75-3.70-3.65-3.60-3.55-3.50-3.45 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir0.00.51.01.52.02.5 Echo Intensity (a.u.)Time (s)A C B D-3.6-3.5-3.4-3.3 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Free-3.6-3.5-3.4-3.3 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Free2030405060 P(r)Distance ()-3.75-3.70-3.65-3.60-3.55-3.50-3.45 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir-3.75-3.70-3.65-3.60-3.55-3.50-3.45 -24 -22 -20 -18 -16 -14 -12 -10 -8 1E-3 0.01 0.1 1 10 100 1000 10000 100000solution norm ||x||2residual norm ||Ax-b||2Ritonavir0.00.51.01.52.02.5 Echo Intensity (a.u.)Time (s)A C B D Figure 5-7. DEER results for HIVPR PMPR*I ASL. A) Background subt racted dipolar echo evolution curves for K55-K55 IASL in th e absence (blue) and the presence (red) of Ritonavir (inhibitor). The solid lines overlaid on the e xperimental data represent the regenerated echo curves from TKR analysis B) L-curve corresponding to the dipolar evolution of the sample in the absence of inhibitor. C) Distance distributions obtained from A). D) L-curve corresponding to the dipolar evolution of the sample in the presence of inhibitor.

PAGE 159

159 IAP MTSL TKR vsMC -Free0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)2030405060 Distance () P(r)TKR vsMC -Inhibitor0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)2030405060 P(r)Distance ()2030405060 P(r)Distance ()0.00.51.01.52.02.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)TKR vsMC -Free TKR vsMC -Inhibitor2030405060 P(r)Distance () 0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)2030405060 P(r)Distance ()IASL TKR vsMC -Free TKR vsMC -Inhibitor2030405060 P(r)Distance ()MSL TKR vsMC -Free TKR vsMC -Inhibitor2030405060 P(r)Distance ()2030405060 P(r)Distance ()0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)2030405060 P(r)Distance ()IASL TKR vsMC -Free TKR vsMC -Inhibitor2030405060 P(r)Distance ()0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)2030405060 P(r)Distance ()IASL TKR vsMC -Free TKR vsMC -Inhibitor2030405060 P(r)Distance ()MSL TKR vsMC -Free TKR vsMC -Inhibitor2030405060 P(r)Distance () 2030405060 P(r)Distance ()0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)MSL TKR vsMC -Free TKR vsMC -Inhibitor2030405060 P(r)Distance () 2030405060 P(r)Distance ()0.00.51.01.52.0 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s) Figure 5-8. Comparison between the distance distribution prof iles functions obtained by DA2006 (Tikhonov regularization, in red) and DeerFit (MC, in blac k) from the DEER dipolar evolution signal for PMPR*SL (SL=MTSL, MSL, IAP, IASL).

PAGE 160

160 0.00.51.01.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)0.00.51.01.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)2345 P(r)Distance ()2345 P(r)Distance ()31.0 32.5 35.0 31.0 32.5 35.0 32.0 32.5 33.0 32.0 32.5 33.0 0.00.51.01.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s) 0.00.51.01.5 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Echo Intensity (a.u)Time (s)2345 P(r)Distance () 2345 P(r)Distance ()31.0 32.5 35.0 31.0 32.5 35.0 31.0 32.5 35.0 31.0 32.5 35.0 32.0 32.5 33.0 32.0 32.5 33.0 32.0 32.5 33.0 32.0 32.5 33.0 Figure 5-9. Average distance and distance di stribution confidence interval plot for PMPR*MTSL. The best solution obtained from TKR corresponds to an average distance of 32.5

PAGE 161

161 2030405060 P(r)Distance ()2030405060 P(r)Distance () MTSL MSL IAP IASL MTSL MSL IAP IASLAB2030405060 P(r)Distance () 2030405060 P(r)Distance () MTSL MSL IAP IASL MTSL MSL IAP IASL MTSL MSL IAP IASL MTSL MSL IAP IASLAB Figure 5-10. Average distance and distance distributions for PMPR*SL in the presence and absence of inhibitor. SL= MTSL (black), MSL (red), IAP (blue) and IASL (green). A) Average distance and distan ce distribution corresponding to 100 M PMPR*SL and 300 M Ritonavir. B) Average distance and distance distribution corresponding to 100 M PMPR*SL without inhibitor added.

PAGE 162

162345678 0.5 0.6 0.7 0.8 0.9 1.0 MTSL MSL IAP IASL LF/CF Intensity Ratiofwhm DEER PMPR+D25N+K55SL+Ritonavir ()345678 180 190 200 210 220 MTSL MSL IAP IASL Second Moment (

)fwhm DEER PMPR+D25N+ K55SL+Ritonavir ()2345678 1.8 2.0 2.2 2.4 2.6 2.8 3.0 MTSL MSL IAP IASL Hpp (G)fwhm DEER PMPR+D25N+K55SL+Ritonavir () MTSL IASL MSL IAPA B CDE345678 0.5 0.6 0.7 0.8 0.9 1.0 MTSL MSL IAP IASL LF/CF Intensity Ratiofwhm DEER PMPR+D25N+K55SL+Ritonavir () 345678 180 190 200 210 220 MTSL MSL IAP IASL Second Moment (

)fwhm DEER PMPR+D25N+K55SL+Ritonavir () 2345678 1.8 2.0 2.2 2.4 2.6 2.8 3.0 MTSL MSL IAP IASL Hpp (G)fwhm DEER PMPR+D25N+K55SL+Ritonavir () MTSL IASL MSL IAP345678 0.5 0.6 0.7 0.8 0.9 1.0 MTSL MSL IAP IASL LF/CF Intensity Ratiofwhm DEER PMPR+D25N+K55SL+Ritonavir () 345678 180 190 200 210 220 MTSL MSL IAP IASL Second Moment (

)fwhm DEER PMPR+D25N+K55SL+Ritonavir () 2345678 1.8 2.0 2.2 2.4 2.6 2.8 3.0 MTSL MSL IAP IASL Hpp (G)fwhm DEER PMPR+D25N+K55SL+Ritonavir () 345678 0.5 0.6 0.7 0.8 0.9 1.0 MTSL MSL IAP IASL LF/CF Intensity Ratiofwhm DEER PMPR+D25N+K55SL+Ritonavir () 345678 180 190 200 210 220 MTSL MSL IAP IASL Second Moment (

)fwhm DEER PMPR+D25N+K55SL+Ritonavir () 2345678 1.8 2.0 2.2 2.4 2.6 2.8 3.0 MTSL MSL IAP IASL Hpp (G)fwhm DEER PMPR+D25N+K55SL+Ritonavir () MTSL IASL MSL IAP MTSL IASL MSL IAPA B CDE Figure 5-11. Correlation between spectral pa rameters obtained from cw-EPR lineshape analysis and FWHM of the distance distribution profile for PMPR*SL (S L=MTSL, MSL, IAP, IASL) obtained by DEER spectroscopy. A) Distance distributions obtained by DEER sp ectroscopy for HIV-1 PMPR*SL + 300 M Ritonavir. B) cw-EPR lineshapes for PMPR*MTSL (black), MSL (red), IAP (blue) and IASL (green). C) Plot of the LF/CF intensity ratio vs. FWHM of the spin labeled PMPR*SL. D) Plot of the spectral second moment (

) ratio vs. FWHM of the spin labeled PMPR*SL. E) Plot of the spectral central resonance line peak-to-peak distance ( Hpp, in Gauss) vs. FWHM of the spin labeled PMPR*SL.

PAGE 163

163 2030405060 P(r)Distance ()2030405060 Distance () P(r)2030405060 P(r)Distance ()A BC Free BoundMD EPR MD EPR MD EPR 2030405060 P(r)Distance ()2030405060 Distance () P(r)2030405060 P(r)Distance ()A BC Free BoundMD EPR MD EPR MD EPR MD EPR MD EPR MD EPR Figure 5-12. Distance distributi on profiles obtained by MD simulations between K55C labeled sites on HIV PMPR+D25N. A) PMPR*MTS L+Ritonavir distance distributions obtained from TKR regularization of the D EER dipolar evoluti on function (red) and from MD simulations (black). B) Comparis on of the distance distributions from MD simulations of HIV PMPR*MTSL with closed flaps in the absence of inhibitors (black) and distance distri butions obtained from TKR analysis of the dipolar evolution function obtained by DEER for HI V PMPR*SL in the ab sence of inhibitor (red). C) Comparison of the distance distri butions obtained for HIV PMPR*SL in the absence of inhibitor by TKR regularization from the DEER dipolar evolution (red) and from MD simulations (black).

PAGE 164

164 CHAPTER 6 PULSED EPR CHARACTERIZATION OF TH E DRUG RESISTANT MUTANTS V6 AND MDR769 OF HIV-1 PROTEASE Drug Induced Mutations in HIV-1 Protease As mentioned in Chapter 2, there are na turally occurring polymo rphisms (naturally occurring mutations) within the HIV-1 genome th at arise from the fact that the enzyme responsible for transcribing viral RNA into viral dsDNA (reverse tr anscriptase) lacks proofreading ability. HIV-1 protease DNA sequen ces, obtained from the Stanford HIV Drug Resistance Database, were analyzed in order to de termine preferential locations within the HIV-1 protease gene that are susceptible to mutati ons. The Stanford HIV Drug Resistance Database (http://hivdb.s tanford.edu/ ) is A curated public database designed to represent, store and analyze the divergent forms of data underlying HIV drug resistance In a nutshell, HIV-1 protease DNA sequence da ta from 13280 PI nave patients (patients that havent been previously e xposed to Protease Inhibitor ther apy) and 7819 PI exposed patients (patients undergoing treatment with one or more PIs) were obtained from the Stanford drug resistance Database (Appendix A), which reports the percentage of each single point mutation encountered in every residue for each subtype. In or der to obtain the variab ility of a particular residue, the data from the Stanford database were processed by adding all percentages of all possible mutations for a particular residue in the protease, for both PI nave and PI exposed patients. As we will see in the following sets of figures, well conserved regions in the HIV protease gene exist across subtypes. It is not eworthy to mention that, although HIV protease exhibits a great number of polymorphic mutations even in PI nave patients, the new mutations that appear in patients undergoing HAART as a consequence of drug pressure appear in welldefined regions of the protease. Figures 6-1 and 6-2 show the prevalence of naturally occurring polymorphisms in PI Nave patients, as well as drug-selected mutations for each of the HIV-1

PAGE 165

165 subtypes. Prevalence can be defined as a meas ure of variability for each protease sequence vs. the subtype consensus (e.g. 0% prevalence corresponds to a residue that is conserved in all sequences for a particular subtype). Figure 6-3 shows the % prevalence difference be tween PI nave and PI exposed patients. Regions with positive values indicate an increase in the number of sequences that present mutations with respect to the subtype consensus for a particular residue; regions with negative values indicate a decrease in the number of seque nces exhibiting mutations in particular residue. From direct comparison amongst the different subt ypes in PI-nave patients (Figure 6-1), there are clear regions in the protease amino acid sequen ce that remain invariant to mutations for the various subtypes/patients. These sequences comp rise the dimerization region (residues 1-9 and 94-99), the active site floor (resi dues 21-32), the flap region (4756), the S3-S3 binding pocket (78-81) and residues 83-89. Upon PI exposure, nu merous mutations develop in the flap region (with the exception of the flap tips, which are hi ghly conserved) and in the 83-89 region, while the dimerization region, the active site floor a nd the S3-S3 binding pocket remain invariant to mutations. These effects in the mutation pattern u pon exposure to protease i nhibitors can be hard to discern from the 3D plots in Figures 6-1 to 6-3. In order to visualize the changes that occur from PI nave to PI exposed patients, a 2D plot for the particular case of subtype B is shown in Figure 6-4. Nomenclature for Drug-Selected Mutations in HIV-1 PR Drug induced mutations can be categorized into two different classes; (1) primary mutations, which are those which di rectly alter the shape and/or electrostatic properties of the active site, thus leading to lower inhibitor affinities; (2) secondary mutations, which, by themselves, do not significantly alter protease inhibito r binding, but that usually appear in

PAGE 166

166 addition to primary mutations, increasing the protease resistance up to 1000-fold for a given antiviral drug120. In table 6-1, we summarize some of the mo st common PIs used in drug therapy and the main and secondary mutations th at are associated with prolo nged exposure to a particular inhibitor. Table 6-1. Main and accessory mu tations that appear in patients under protease inhibitor therapy Protease Inhibitor Main Mutati ons Accessory Mutation Residues Tipranavir (TPV) L33F, V82L/T, I84V10,13,20,35,36,43,46,47,54 58,69,74,83,90 Indinavir (IDV) M46I/L, V82A/F/T, I84V10,20,24,32,36,54,71,73,76 77,90 Saquinavir (SQV) G48V,L90M10,24,54,62,71,73,77,82,84 Lopinavir (LPV) V32I, I47V/A, V82A/F/T/S10,20,24,33,46,50,53,54,63 71,73,76,84,90 Fosamprenavir (FPV) I50V, I84V10,32,46,47,54,73,76,82,90 Ritonavir (RTV) Current ly used as booster Darunavir (DRV) I50V, I54M/L, I84V11,32,33,47,73,89 Atazanavir (ATZ) I84V,N88S10,16,20,24,32,33,34,36,46 48,53,54,60,62,64,71,73,82 85,90,93 Nelfinavir (NFV) D30N, L90M10,36,46,71,77,82,84,90 Data obtained from Stanford HIV Drug Resistance Database Multidrug Resistant HIV1 Protease Variants The development of more specific and more poten t viral inhibitors, comb ined with the lack of proofreading capabilities of Re verse Transcriptase (RT) has l ead to the selection of viral strains that incorporate mutations that confer resistance to current FDA approved inhibitors. These multidrug resistant (MDRs) variants are of great interest from a structural and biological point of view. Understanding how individual point mutations in HIV-1 protease are involved in the drug-selection process will undoubtedly aid in th e design of more specific inhibitors that can accommodate different mutations while retaining binding affinities.

PAGE 167

167 One of the proposed models for the drug resi stance occurring from primary mutations is known as the active site expansion model146,147. It is hypothesized that the introduction of single amino acid substitutions of residues in the active s ite cavity with long side chains for amino acids with shorter side chains induces a series of conf ormational changes that lead to an expansion of the active site (increased volume of the active site due to loweri ng of side chain crowding), with the loss of van der Waals (vdW) interactions and hydrogen bonds between the inhibitor and the protease. It has been proposed, as well, that introduc tion of secondary mutations (those mutations that do not occur within the active site of the protease), might have cooperative in teractions with other mutations occurring within the protease, leading to a dramatic increase of the drugresistance while maintaining protease viability. Examples of these types of secondary mutations are those occurring in positions M46 and I54, for which the flap conformational heterogeneity is thought to be altered, possibly modifying the popul ations and/or the distances of the open/closed conformations of the flaps120. Multidrug resistant V6 A multidrug resistant strain of HIV-1 protease, termed V6, is a clinical isolate from a pediatric patient after prolonge d treatment with Ritonavir120. The amino acid sequence of V6 contains mutations associated wi th Ritonavir resistance at residue s 20, 32, 33, 36, 63, 71, 82 and 90. In 1996, Molla et al. showed that mutations that appear in patients under Ritonavir therapy emerged in a particular sequence148. The initial loss of antiviral ac tivity is associated with the appearance of the mutation V82A followed by the mutations I 54V, A71V and M36I. Finally, mutations in positions I84V, K20R, M46I, L33F and L90M appeared in the protease. It is important to note that mutations at residues K20 (R), M36(IL) and L63(P) appear to act as

PAGE 168

168 general upregulators of HIV-1 pr otease activity, meaning that these mutations enhance HIV-1 protease activity. These mutati ons are highly polymorphic and represent the consensus amino acids for some subtypes. It can be seen from tabl e 6-2 that the mutations that appear in the V6 variant confer not only resistance to RTV ther apy, but also to a large number of PIs. Table 6-2. Common primary and secondary mutati ons in HIV-1 protease in similar positions to the V6 strain K20 V32 L33 M36 L63 A71 V82 L90 ATV/r T F IVL P VTI AF M DRV/r I F P VTI F M FPV/r I F IL PAV VTI F M IDV/r I IL P VTI AFTS M LPV/r T I F IL P VTI AFTS M NFV T F IL P VTI AFTS M SQV/r P VTI TF M TPV/r I F IL P VTI TFSL M RTV R I F IL PV VTI AFTSL M Notes: Red indicates phenotypic evidence for reduced susceptibility in vivo/in vitro. /r Represents Ritonavir booster therapy Table 6-3. Kinetic parameters for V6 strain and mutations in V6 LAI V6 V646 V654 V646/84 V654/84 Km ( M) 182 475274497434 488 kcat (sec-1) 212 2721211719.90.9 7.10.9kcat/Km ( M-1 sec-1) 1.20.2 0.580.070.430.070.340.060.230.03 0.150.02Ki RTV 0.70.1 698658109934540 93283Ki IDV 3.10.1 302196324239877622 4235506Ki NFV 1.20.2 17311614232522 1259117Note: Ki, kcat and Km values were extracted from Clemente, J. C.; Moose, R. E.; Hemrajani, R.; Whitford, L. R.; Govindasamy, L.; Reutze l, R.; McKenna, R.; Agbandje-McKenna, M.; Goodenow, M. M.; Dunn, B. M. Biochemistry 2004, 43 Pages 11243-11244 Table 6-3 shows various kinetic parameters (Km, kcat, kcat/Km) for the LAI and V6 variants, as well as mutations within the V6 sequence. Inhibition constants (Ki) are included for 3 commonly used PIs (Ritonavir, Indinavir and Nelfinav ir). It can be clearly seen that there is a large increase in the value of Ki for all the inhibitors (between 14 and 42 fold) from LAI to V6. Upon introduction of further mutations (46, 54 or combinations of both with mutations in residue

PAGE 169

169 84), the Ki increases up to 1300 fold. On the other hand, V6 presents a 50% reduction of kcat/Km when compared to LAI, due to a 3-fold increase in Km. From crystallographi c studies performed on V6, it was postulated that mutations in the fl ap regions affect the dynamics of flap opening and closing. The basis for this postulate is the formation of a highly hydrophobic pocket by the amino acids F33, I25 and K20 where I36 fits, stru cturally modifying the position of the 30s loop and the 10s loop. Multidrug resistant MDR769 Multidrug resistant 769 (MDR769) is a clinical is olate from a patient which presented high resistance to currently FDA-approved PI inhi bitors. This construct contains the following mutations: L10I, M36V, M46L, I54V, I62V L63P, A71V, V82A, I84V and L90M.149,150 Table 6-2 shows a comparison between WTNL 4-3, a reference lab isolate and the multidrug resistant protease MDR769. It can be s een that there is a large change in inhibitor resistance (>40 fold change) in the drug resistant mutant. Table 6-4. Susceptibility of WT and MDR769 isolates to protease inhibitors WTNL4-3 MDR76946,54,82,84,90 Inhibitor IC50 ( M) IC50 ( M) Fold Change (WT) Approved IDV 0.035.022.2.463 SQV 0.030.011.3.243 NFV 0.15.077.0.447 Experimental Vx-478 0.023.010.31.0614 BMS 232632 0.004.0020.32.180 DG-35 0.058.007ND DG-43 0.001.00010.50.08>100 Palinavir 0.027.01>10>100 Cyclic Sulfone GS 3333 0.030.013.4.2>100 Notes: Data values were taken from Palmer, S.; Shafer, R. W.; Merigan, T. C. AIDS 1999, 13 661-7

PAGE 170

170 X-ray crystallography studi es of MDR769 (Kovari et al. ) reported that the distance between the tips of the flaps in the crystal stru cture increased by ~8 w ith respect to the LAI consensus146,147. A ribbon representation of HIV-1 proteas e, which shows the distance changes between the tips of the flaps (from 4.3 in the inhibited form to 12.3 in the open MDR769 form) is shown in Figure 6-6. The authors hypothesized that this open form of the flap is stabilized by a network of water molecules hydrogen-bonded to the gl ycine-rich region of the flaps. This flap rearrangement is initiated by the introduction of mutation L46L, wh ich induces a displacemen t in the backbone of 2 by removing a Van der Waals (vdW) contact between residues M36 and I15. Furthermore, residue 36V now forms new vdW interactions with residues V36 and L38. These new vdW interactions, and new contacts formed with G16, move the residue G16 by 2 toward L38, shifting the G40 loop out of the pos ition in the WT. This network of collective interactions might be responsible for the stability of the open fo rm, as well as the stabilization of the open conformation. Nevertheless, MD simulations performed by Simmerling et al. on the MDR769 mutant suggest that the open flap c onformation reported by Kovari et al. might be a form of the protease stabilized by crystal contacts 151. Although LAI protease both in the inhibited and uninhibited form shows possible crystal contact interactions close to the flap region, MDR769 flap tips are buried between the elbow and the fulcrum region of a neighboring protease in the crystal structure. Figure 6-7 shows a ribbon representa tion of HIVPR MDR769 st rain including the nearest lattice symmetry neighbor s (in red), with the central un it cell of HIV-1 protease (in orange). From this representation, it is clear that the elbows of neighboring flaps interact with the tips of the flaps of the orange protease, and thus can stabilize an open conformation of the flaps.

PAGE 171

171 Furthermore, MD simulations on MDR769 showed that when the closest crystal contact neighbors are included, the protease flaps are unable to close. Upon removal of the crystal contact interactions, the flaps return to a structur e that resembles the semi-open form of the LAI. With the aid of pulsed-EPR and DEER, we i nvestigated whether the distance distribution profile between residues K55-K55 in the flaps of MDR769 differs from the distribution profile that we obtained for the LAI consensus sequence. The effects of the addition of inhibitor and chromogenic substrate were explored, where the ch romogenic substrate was used as a substitute for the natural sequence of the CA/p2 cutsite. Figure 6-8 shows the mutations introduced for the MDR769 EPR studies. In green, position K55 was utilized to introduc e the spin label. In orange, active site mutation D25N and, in red, natural cysteines C67 and C95 wh ich were mutated to alanine residues. Cw-EPR and Pulsed-EPR Studies of V6 and MDR769 Drug Resistant Strains In an analogous manner to the studies perfor med for the LAI consensus sequence of HIV-1 protease, X-Band cw-EPR spectroscopy was performed on the V6 and MDR769 multidrug resistant strains of HIVPR. The results are shown in Figure 6-9. Figure 6-9 A) shows area normalized 100 G X-Band cw-EPR spectra of 100 M V6*MTSL mutant in the presence (red) and ab sence (black) of inhi bitor (Indinavir). The resulting lineshapes from cw-EPR show no difference upon addition of inhibitor. Figure 6-9 B) shows the MDR769*MTSL mutant, in the presence (red) and abse nce (black) of Indinavir. Again, the lineshapes do not reflect any changes upon inhibitor binding, as it was seen earlier for the case of PMPR*MTSL. For this consensus se quence in the presence of inhibitor, the EPR lineshapes for all 4 spin labels (Chapter 4) di d not show significant changes from those of the uninhibited protease.

PAGE 172

172 From the DEER results for the same samples, the distance distribution profiles reflect a change in the average distance, as well as the FWHM of the distance distribution; hence, we know that the addition of inhibitor closes the prot ease flaps. Therefore, we can conclude that the cw-EPR lineshape is reporting intrinsic spin la bel motion, and not conformational changes in the protease backbone. In an attempt to use cw-EPR analysis to discriminate between inhibited and non-inhibited protease, a multifrequency approach is being used in collaboration with Prof. Keith Earle and Prof. Jack Freed at ACERT in Cornel l University. Currently, cw-EPR experiments at High Field High Frequency (HFHF), corres ponding to 170 GHz and 240 GHz, are being performed at ACERT for PMPR*M TSL, V6*MTSL and MDR769*MTSL. Although X-Band cw-EPR lineshapes do not report conformational changes in the protease backbone, we showed in Chapter 5 that, using DE ER spectroscopy, we were able to distinguish between inhibited and uninhibited protease, and determine changes in th e distance distribution profile widths and the average distance changes. In this chapter, in an analogous manner, the results of performing DEER spectroscopy for three different conditions are presented: (1) In the presence of inhibitor: for this work we chose In dinavir as the inhibitor, as it is easier to work with (it is soluble in aqueous solvents, elimina ting the need of DMSO or organic solvents, see Appendix C) and, for the particul ar case of V6, it exhibits lower dissociation constant values (Ki IDV = 4.9 nM, Ki RTV=30 nM)120. (2) Absence of inhibitor. We are interested in comparing inhibited and non inhibited protea se, as well as the differences in between the uninhibited distance distribution profiles of different strain s of HIV-1 protease, namely PMPR*, V6* and MDR769*. These results shed light on the changes in the flap conformati onal heterogeneity and flap dynamics for different drug resistant strains. (3) In the presence of chromogenic substrate. (Chromogenic substrate VI, Lys-AlaArg-Val-Nle-nPhe-Glu-Ala-Nle-NH2). The rationale

PAGE 173

173 behind adding a substrate is to explore if the closed conformation when a substrate is bound to inhibitor differs from the closed form in the pr esence of inhibitor. The results obtained from direct comparison between these two samples will be of great help when trying to understand the distance distribution profiles obtained for multidrug resistant constructs, as the resulting distances obtained by TKR can reflect a variety of factors, namely partial inhibition of the sample (where Ki for a given inhibitor play s an important role) and pa rtial opening/closing of the flaps. The results obtained for PMPR*MTSL, V 6*MTSL and MDR769*MTSL for these three conditions are shown in Figures 611 to 6-13. Plots D, E and F in Figures 6-11 to 6-13 show experimental DEER dipolar evolution curves and plots A, B and C show the distance distribution profiles obtained from the DEER dipolar evol ution by TKR regularizati on. Figure 6-10 shows the results obtained from DEER spectroscopy for HIV PMPR*MTSL. Figure 6-10A) shows the difference between the distance distribution prof iles of PMPR*MTSL in the absence (blue) and presence (black) of chromogenic substrate. This plot clearly shows that the distance distribution of the flaps is altered (the distance profile is narrowed by and the distance is shifted by ~2) by the addition of chromogenic substr ate, and this results are consistent with what it was expected: due to the mutation removing the active site (D 25N), the protease is unable to cleave the substrate and upon substrate binding, the flaps close, trapping the substrate in the active site. It can also be seen that the width of the dist ance distribution for the chromogenic substrate bound protease seems to have a shoulder at long distan ces, possibly correspondi ng to a second distance distribution. Figure 610B) shows the distance distributi on profile for PMPR*MTSL in the presence (red) and absence (blue) of Indinavir. It should be noted that the distance distribution profile of the Indinavir bound pr otease shows the same effect as the chromogenic substrate

PAGE 174

174 bound protease (shoulder at longe r distances, around 37 ) which can also correspond to two different conformations. Figure 6-10C) shows th at, within experimental error, there are no differences between the inhibi ted, Indinavir bound state, a nd the chromogenic substrate bound state. In order to further understand what the tw o possible conformations pr esent in the Indinavir and the chromogenic substrate bound samples may correspond to, we fitted the distance distribution profiles for the chromogenic s ubstrate bound and the Indinavir bound PMPR*MTSL samples to a two Gaussian distribution function. The distance distribution analysis is shown in Figure 6-13. Figure 6-13 shows a 2 Gaussian fit to th e distance distribut ion profile for HIV PMPR*MTSL in the presence of inhibitor (Indinavir, 6-12A) and chromogenic substrate (CS VI, 6-12B). Table 6-5 summarizes the values employed for the fit for each Gaussian function. Table 6-5. 2-Gaussian fit parameters employe d to fit distance di stribution profiles for PMPR*MTSL in the presence of Indi navir or chromogenic substrate xc1 () w1() A1 xc2 w2() A2() Indinavir 32.9 2.6 0.5 36.1 4.2 0.5 Chrom Sub 33.2 3.0 0.7 36.5 3.6 0.3 Note: The functional form of the fit is 22 12 1222 12 0 1222ccxxxx wwAA yyee ww The values obtained for both Gaussian distribut ions in the fit can be compared. Recalling the data obtained for PMPR*MTSL, the results obtaine d were of an average distance of 32.6 for the inhibited (Ritonavir) samples and 35.5 for the uninhibited protease From Figure 6-12 as well as from the numerical values in Table 6-5, it can be seen that the first Gaussian (represented by the parameters xc1, w1 and A1) is in good agreement with the va lues expected for the average distance distribution for the Ritonavir inhibite d form of the protease, whereas the average

PAGE 175

175 distance of the second Gaussian, corresponding to the shoulder in the distance distribution (represented by xc2, w2 and A2), corresponds with average dist ance of the uninhibited form. Hence, it is likely that the dist ance distributions obtained for th e chromogenic substrate/Indinavir correspond to a convolution of the closed and semiopen/open forms. Figure 6-11 shows DEER data and distance distribution profiles obtained by TKR for V6*MTSL in the presence and absence of chrom ogenic substrate and Indinavir. The Figure 611A plot shows that, when chromo genic substrate is added to the protease (black), the resulting distance distribution average shifts by 1 to shorter distances from that of the unbound protease (from 34.9 in the uninhibited to 33.9 in th e chromogenic substrate bound form). This result is again consistent with the hypothesis of a larger extent of flap closi ng upon substrate binding, but extreme caution should be exercised when cha nges in the distance distributions are so small, as it might be an artifact of the TKR analysis Figure 6-11B shows no change in the distance distribution in the unbound proteas e (34.9 ) and the protease upon addition of Indinavir (35.1 ). This might be due to a more open structure of the flaps even in the presence of inhibitor (somewhat of a semi-open form), or to part ial inhibition of the sample. Figure 6-12C shows the comparison between the chromogenic s ubstrate bound and the Indinavir bound form, and reveals that the substrate bound form distances acce ssible to the spin labe l are slightly shorter than in its Indinavir counterpart. This could be explained in term s of drug-resistance: Even when inhibitor is present, the flaps might not be fully closed, and therefore not trapped in a closed state, favoring the release of the inhibitor from the active site cav ity. When substrate is present, the flaps now access a more closed conformati on, making the necessary contacts with the substrate to perform its catalytic activity.

PAGE 176

176 Figure 6-12 shows the results obtained for MDR769*MTSL in the presence and absence of chromogenic substrate and Indina vir. Figure 6-12A plots the di stance distribution profile for uninhibited protease and chromogeni c substrate bound protease. It ca n be seen that the distance distribution profiles are nearly identical, indicating that ther e is no change between the unbound protease and the protease when chromogenic substrate is added. Figure 6-12B shows, again, no change in the distance distri bution profiles between unbound proteas e and protease with inhibitor (Indinavir) added. Figure 6-12C shows littl e to no change between the protease with chromogenic substrate and the protease with Indina vir. One plausible explanation for this effect is that, as MDR769 is a highly resistant strain of HIV-1 protease to all in hibitors, it is possible that addition of inhibitor/chro mogenic substrate has little to no change on the flap distance distributions, and therefore the catalytic efficiency and inhibitor efficiency are severely handicapped as a result from these mutations. Un fortunately, there are no kinetic experiments performed on MDR769, and values of Ki, kcat and Km are undetermined. In order to evaluate the effect s of drug-selected mutations in the distance distributions of the flaps, plots that compare the distance dist ributions for all unbound prot eases, as well as those with chromogenic substrate or Indina vir added, are shown in Figure 6-14. Figure 6-14 shows a comparison between the D EER distances obtained for HIV-1 protease variants, as a function of whethe r substrate or inhibitor were a dded, as well as unbound protease. From Figure 6-14A, an interesting observati on about the degree of flap opening in the uninhibited proteases can be made. It can clearly be seen that th e average distances between spin labels significantly vary for each of the protea ses under study. PMPR presents a broader range of distances, spanning from 24-48 whereas both V6 and MDR769 show more restricted motion of the flaps, in the range of 30-40 The aver age distance is also shif ted, with V6 giving a

PAGE 177

177 slightly shorter distance than PMPR and MDR769 with a longer average distance (Table 6-6). These distance distribution changes can be at tributed to different degrees of backbone conformational heterogeneity. Figure 6-14B shows HIVPR in the presence of chromogenic substrate as a function of the three variants studied: HIVPR PMPR*MTSL (blue), V6*MTSL (red) and MDR769*MTSL (black). From the distance distributions, the follow ing conclusions can be drawn. First, from the average distance distribution we observe the trend PMPR*~V6*
PAGE 178

178 Table 6-6. Average distance distribu tions for PMPR,V6 and MDR769*MTSL. Strain PMPR () V6 () MDR769 () V6-PMPR () MDR-PMPR () MDRV6() Free 35.5 34.9 36.4-0.60.9 1.5 Substrate 33.6 33.9 36.10.42.5 2.1 IDV 32.9 35.9 35.1 36.52.2 -0.8 3.6 0.6 1.4 Note: All errors, unless menti oned otherwise, are 0.5 Table 6-7. FWHM distributions for PMPR,V6 and MDR769*MTSL Strain PMPR () V6 () MDR769 () V6-PMPR () MDR-PMPR () MDRV6() Free 10.5 5.7 3.9-4.8-6.6 -1.8 Substrate 4.0 5.9 3.41.9-0.6 -2.5 IDV 2.6 4.2 5.4 3.62.8 1.2 1 -0.6 -1.8 We can summarize the conclusions of the results obtained for PMPR*MTSL, V6*MTSL and MDR769*MTSL in the absence a nd presence of Indinavir/subs trate as follows: (1) in the absence of inhibitor, the flaps seem to adopt a more closed conformation for V6*MTSL (average distance of 34.9 ) vs. the reference PMPR*MTSL (35.5 ), whereas MDR769*MTSL shows a longer distance between spin labels(36.4 ), whic h could correlate to a structure in which the flaps are more open. (2) in the case for which the sample contains chromogenic substrate, an analog of the natural CA/p2 cu tsite, the flaps for both V6*MTSL (34.0 ) and MDR769*MTSL (36.1) seem to exhibit a longe r average interlabel distance th an the PMPR*MTSL(33.6 ). For the case of V6, it can be seen that the ch ange in the average distance between unbound and substrate bound protease is of ~1 whereas for the case of MDR769, that change is only 0.3 This could be an indication for a reduced binding affinity of MDR769 for the substrate, as the change between the unbound conformation and the substrate bound conformation is almost negligible. (3) For the case of the inhibitor (Indinavir) bound proteas es, it can be seen that, for PMPR*MTSL, we find that the di stance distribution pr ofile can be explained by a 2-Gaussian

PAGE 179

179 model, in which the first Gaussian component corresponds to an average distance of 32.9 (which is similar to that of PMPR*MTSL + Ritonavir (32.6 )), and the second Gaussian component corresponds to a form of the prot ease that remains unbound to inhibitor (35.9 ). This effect is possibly due to the reduced binding affinities of Indinavir for V6 vs. Ritonavir for PMPR. For the case of V6*MTSL and MDR769*MTSL the change in the average distance is within experimental error (34.9 vs. 35.1 for V6, 36.4 vs. 36.5 for MDR769). These results could be explained as a) the inhibitor is only partially binding, and only a minority of the sample in the DEER experiment is Indina vir bound or b) the unbound conformation and the closed conformation have similar interlabel distances. We will now discuss the effects in the dist ance distribution FWHM for the addition of Indinavir or chromogenic substrate for PM PR*MTSL, V6*MTSL and MDR769*MTSL. First, we noted in Chapter 5 that the distance distribu tion width for MTSL when inhibitor is added to the solution undergoes a large change from FWHM of 10.5 in the uninhibited case, to 3.0 in the Ritonavir-bound case. The same effect occurs when Ritonavir is substituted for Indinavir, yielding distance distributions with a FWHM of 2.7-4.2 In the case of V6*MTSL, the distance distribution width does not change upon addition of subs trate or IDV, reporting FWHM of 5.6.3 for all V6*MTSL samples. The same effect occurs in MDR760*MTSL, with even shorter distance distributi on widths (3.6.3 ). From these DEER results, we can conclude that both V6 and MDR769 present a higher degree of structural rigidity in the flap region when compared to PMPR, and that the addition of IDV does not seem to perturb the structure of the flaps inducing an open/closed conformational change, at least not as pronounced as in the case of PMPR.

PAGE 180

180 Nonetheless, there are potential problems to be considered in further studies for these two systems: Inhibitor/substrate binding studies. Although we saw that PMPR+K55MTSL (Chapter 4) was able to cleave chromogenic substrate, in dicating binding and pro cessing of substrate, there is no experimental evid ence on V6*MTSL or MDR769*MTSL. Due to the nature of the drug-resistant strains, the value of Ki for such drug-resistant strains is higher than the PMPR values. Ther efore, partial inhibition of the sample is possible, and that could lead to distance di stribution profiles that are a convolution of open/closed states when inhibitor is present. Experimental evidence of substrate/inhib itor binding for V6+D25N and MDR769+D25N is lacking, and a more detailed study u tilizing Surface Plasmon Resonance (SPR) or Isothermal Calorimetry (ITC) should be performed for V6/V6+D25N/V6+D25N+K55MTSL, as well as MDR769/ MDR769+D25N/ MDR769+D25N+K55MTSL. Determination of Ki, kcat and Km for MDR769 is needed in orde r to correlate inhibitor and substrate binding affinities with di stance distributions obtained by DEER.

PAGE 181

181 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 A 0 10 20 30 40 50 60 70 80 90 100C D F AE AG G A BMutation Prevalence (PI Nave patients)Residue (aa#)Prevalence (%)S u b t y p e 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 A 0 10 20 30 40 50 60 70 80 90 100C D F AE AG G A BMutation Prevalence (PI Nave patients) 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 A 0 10 20 30 40 50 60 70 80 90 100C D F AE AG G A B 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 A 0 10 20 30 40 50 60 70 80 90 100C D F AE AG G A BMutation Prevalence (PI Nave patients)Residue (aa#)Prevalence (%)S u b t y p e Figure 6-1. Natural polymorphisms o ccurring in PI nave patients for all HIV-1 protease subtypes (A-AG).

PAGE 182

182 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 A 0 10 20 30 40 50 60 70 80 90 100C D F AE AG G A BMutation Prevalence (PI Therapy patients)Residue (aa#)Prevalence (%)S u b t y p e 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 A 0 10 20 30 40 50 60 70 80 90 100C D F AE AG G A B 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 A 0 10 20 30 40 50 60 70 80 90 100C D F AE AG G A BMutation Prevalence (PI Therapy patients)Residue (aa#)Prevalence (%)S u b t y p e Figure 6-2. Drug-selected mutations o ccurring in PI exposed patients for all HIV-1 protease subtypes (A-AG).

PAGE 183

183 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 DifferenceA DifferenceAG DifferenceC DifferenceF -40 -30 -20 -10 0 10 20 30 40 50 60 A B C D F AE AG GMutation Prevalence (PI Therapy PI Nave patients)Residue (aa#)Prevalence (%)S u b t y p e 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 DifferenceA DifferenceAG DifferenceC DifferenceF -40 -30 -20 -10 0 10 20 30 40 50 60 A B C D F AE AG G 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 DifferenceA DifferenceAG DifferenceC DifferenceF -40 -30 -20 -10 0 10 20 30 40 50 60 A B C D F AE AG GMutation Prevalence (PI Therapy PI Nave patients)Residue (aa#)Prevalence (%)S u b t y p e Figure 6-3. Mutation prevalence pe rcent difference between nave and PI exposed patients, by subtype.

PAGE 184

184 0102030405060708090100 0 20 40 60 80 100 % Mutation PrevalenceResidue # 0102030405060708090100 0 20 40 60 80 100 % Mutation PrevalenceResidue #PI Nave (Subt. B)PI Exposed (Subt. B) 0102030405060708090100 0 20 40 60 80 100 % Mutation PrevalenceResidue # 0102030405060708090100 0 20 40 60 80 100 % Mutation PrevalenceResidue #PI Nave (Subt. B)PI Exposed (Subt. B) Figure 6-4. Mutation prevalence for subtype B as a function of residue number. A) PI nave patients B). PI exposed patients. Figure 6-5. Ribbon diagram of HI V-1 protease, V6 variant (PDB 1SGU). In green, residue K55 was chosen for spin label attachment. In red, naturally occurring cysteines C67 and C95. The orange residues represent the activ e site D25. Characteristic mutations of the V6 variants are represented in space filling model rendering.

PAGE 185

185 12.3 4.3 12.3 4.3 Figure 6-6. Side (left) and t op (right) views of HIV-1 protea se LAI (blue, 1HHP) and MDR769 (red, 1TW7) crystal structures. The top view clearly shows the reversal handedness of the flap tips and the greate r separation of the flap tip s in MDR769 vs. the WT LAI protease. Figure 6-7. Crystal structure of MDR769 including nearest la ttice symmetry neighbors. The space group is P41. The symmetry neighbor s were generated with DeepView/ SwissPDB and rendered using VMD.

PAGE 186

186 A) B)46 55 54 10 84 82 71 63 62 36 90 95 67 A) B) A) B)46 55 54 10 84 82 71 63 62 36 90 95 67 Figure 6-8. X-ray crystallography structures of MDR769 (PDBID: 1TW7). A) Left, side view with catalytic aspartic re sidues (D25) in orange. Nativ e cysteines C67 and C95 are represented in red and resi due K55 in green. Drug-selec ted mutations were rendered in vdW and volumetric surfaces. B) Right top side view of the MDR769 drugresistant strain. V6 V6 + 400 M Indinavir MDR769 MDR769 + 400 M Indinavir 20GAB V6 V6 + 400 M Indinavir V6 V6 + 400 M Indinavir MDR769 MDR769 + 400 M Indinavir MDR769 MDR769 + 400 M Indinavir 20G 20GAB Figure 6-9. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIV-1 protease drug resistant strains A)V6*MTSL in 2 mM Na OAc pH 5.0 in the presence (red) and absence (black) of Indinavir. B) MDR 769*MTSL in 2 mM NaOAc pH 5.0 in the presence (red) and absence (black) of Indinavir.

PAGE 187

1872030405060 P(r)Distance ()2030405060 P(r)Distance ()2030405060 P(r)Distance ()0.00.51.01.52.02.53. 0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)A C D B E F2030405060 P(r)Distance ()2030405060 P(r)Distance ()2030405060 P(r)Distance () 0.00.51.01.52.02.53. 0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)A C D B E F Figure 6-10. DEER spectra of HIVPR PMPR*MTSL. A) HIVPR PM PR*MTSL distance distribution in the presence (black) and absence (blue) of chromogenic substrate. B) HIVPR PMPR*M TSL distance distribution in the presence (red) and absence (blue) of Indinavir. C) HIVPR PMPR*MTSL distance distribu tion in the presence of chro mogenic substrate (black) and Indinavir (red). D) HIVPR PMPR*MTS L background subtracted dipolar evoluti on in the presence (black) and absence (blue) of chromogenic substrate. E) HIVPR PMPR*MTSL ba ckground subtracted dipolar evolution in the presence (red) and absence (blue) of Indinavir. F) HIVPR PMPR*MTSL b ackground subtracted dipolar e volution in the presence of chromogenic substrate (bl ack) and Indinavir (red).

PAGE 188

1880.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)2030405060 P(r)Distance ()2030405060 P(r)Distance ()2030405060 P(r)Distance ()0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)A C D B EF0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s) 2030405060 P(r)Distance ()2030405060 P(r)Distance ()2030405060 P(r)Distance () 0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s) 0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)A C D B EF Figure 6-11. DEER spectra of HIVPR V6*MTS L. A) HIVPR V6*MTSL distan ce distribution in the presence (black) and absence (blue) of chromogenic substrate. B) HIVPR V6*MTSL distance distri bution in the presence (red) and absence (blue) of Indinavir. C) HIVPR V6*MTSL di stance distribution in the presence of chromo genic substrate (black) and Indinavir (red). D) HIVPR V6*MTSL background subt racted dipolar evolution in the presence (b lack) and absence (blue) of chromogenic substrate. E) HIVPR V6*MTSL background subtracted dipolar evolution in the presence (red) and ab sence (blue) of Indinavir. F) HIVPR V6*MTSL background su btracted dipolar evolution in the pres ence of chromogenic substrate (black) and Indinavir (red).

PAGE 189

1890.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)2030405060 P(r)Distance ()2030405060 P(r)Distance ()2030405060 P(r)Distance ()A C D B E F0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)0.00.51.01.52.02.53.0 Echo Intensity (a.u)Time (s)2030405060 P(r)Distance ()2030405060 P(r)Distance ()2030405060 P(r)Distance ()A C D B E F Figure 6-12. DEER spectra of HIVPR MDR 769*MTSL. A) HIVPR MDR769*MTS L distance distribution in the presence (black) and absence (blue) of chromogenic substrate. B) HIVPR MDR7 69*MTSL distance distribution in the presence (red) and absence (blue) of Indinavir. C) HIVP R MDR769*MTSL distance dist ribution in the presence of chromogenic substrate (black) and Indinavir (red). D) HIVPR MDR769*MTSL bac kground subtracted dipolar evolut ion in the presence (black) and absence (blue) of chromogenic su bstrate. E) HIVPR MDR769*MTSL background subtracted dipolar evolution in the presence (red) and absence (blue) of Indinavir. F) HIVP R MDR769*MTSL background subtract ed dipolar evolution in the presence of chromogenic substrat e (black) and Indinavir (red).

PAGE 190

190 2030405060 P(r)Distance () 2030405060 P(r)Distance () Data Peak1 Peak2 Data Peak1 Peak2AB2030405060 P(r)Distance () 2030405060 P(r)Distance () Data Peak1 Peak2 Data Peak1 Peak2 Data Peak1 Peak2 Data Peak1 Peak2AB Figure 6-13. 2-Gaussian fit to the distance distribution profile fo r HIV PMPR*MTSL. A) PMPR*MTSL in the presence of Indinavi r. B) HIV*PMPR in the presence of chromogenic substrate. 2030405060 P(r)Distance () 2030405060 P(r)Distance () 2030405060 P(r)Distance ()ABCPMPR V6 MDR769 PMPR-CS V6-CS MDR769-CS PMPR-IDV V6-IDV MDR769-IDV 2030405060 P(r)Distance () 2030405060 P(r)Distance () 2030405060 P(r)Distance ()ABCPMPR V6 MDR769 PMPR V6 MDR769 PMPR-CS V6-CS MDR769-CS PMPR-CS V6-CS MDR769-CS PMPR-IDV V6-IDV MDR769-IDV PMPR-IDV V6-IDV MDR769-IDV Figure 6-14. Comparison between distances ob tained by DEER spectroscopy for HIVPR-PMPR (blue), HIVPR-V6 (red) and HIVPR-MDR769 (black). A) Non inhibited PMPR*MTSL(blue), V6*MTSL (red) and MDR769*MTSL (black). B) In the presence of chromogenic substrate: PMPR*MTSL(blue), V6*MTSL (red) and MDR769*MTSL (black). C) In the presen ce of Indinavir: PMPR*MTSL(blue), V6*MTSL (red) and MDR769*MTSL (black).

PAGE 191

191 CHAPTER 7 FUTURE DIRECTIONS Inhibitor and Substrate Binding Studies for D25N Mutants of HIV-1 Protease In Chapter 4, we introduced the mutation that removes the catalytic active site of HIV-1 protease (D25N) in order to perfor m the preliminary studies on HIV-1 protease, as well as to determine the feasibility of cw -EPR and pulsed-EPR to study the conformational heterogeneity of HIV-1 protease. In order to determine thermo dynamic parameters related to inhibitor/substrate binding, ITC should be utilized in the future to determine binding enthalpies and specific heat (Cp). T This information will be of great use in understanding whether the inhibitor of choice is effectively bound to the protease, in particular fo r the cases of the multidrug resistant variants V6 and MD769 for which the addition of inhibitor does not significantl y change the distance distribution profiles obtained by DEER. Also, in order to determine binding, Surf ace Plasmon Resonance (SPR) experiments can be performed. These experiments are based on the covalent binding of HIV-1 protease to a modified Biacore chip, and flowing inhibitor in the solution as the SPR experiment is performed. These experiments have already b een reported for HIV-1 protease by the group of Kovari for MDR769. DEER Experiments Performe d With HIV-1 Inhibitors In order to understand the po ssible effects of the choice of inhibitor in the cw-EPR lineshape and the pulsed-EPR distance distribu tion profile, experiments utilizing other FDA approved inhibitors should be performed. With the introduction of new inhibitors (such as Darunavir (DRV) or Tipranavir (TPV)), with st rong binding affinities (in the low nM range) for drug-resistant mutants, further information can be obtained regarding the m eaning of the distance distribution profiles obtained by DEER for the V6 and MDR769 protease strains.

PAGE 192

192 Single-Point Mutant Strains In this work, we studied drug-resistant cons tructs of patients that were exposed to antiretroviral therapy for prolonged periods of time, and hence accumulated a great number of single-point mutations in the protease structure. In order to further understand the effects of drug-selected resistance to pa rticular Protease Inhibitors, es pecially in the case of those secondary mutations that alter flap dynamics but do not interfere directly with inhibitor binding, mutations following the natural pr ogression of drug-resistance in PI exposed patients can be introduced sequentially. This should help us understand the key steps in secondary mutation drug-induced resistance. Some of those mutations include mutations in residues 82, 84, 46 and 54. In fact, mutants for PMPR+D25N+V46I+K55C and PMPR+D25N+I54V +K55C have already been expressed, purified and spin labeled, and the results of this study will be reporte d in the literature. High Field/High Frequency EPR Studies on the Dynamics of Spin Labels As it was shown in this work, one of the cu rrent issues in the SD SL community is the decoupling of the internal modes of the spin label and the bac kbone motions of the protein to which is tethered. We mentioned the fact that, in HIV-1 protease, for a beta hairpin solvent exposed site, the cw-EPR X-Band lineshape of MTSL/MSL/IAP/IASL does not report backbone motion, but rather spin label motion. This was ev ident when inhibitor was added to the protease and there was no change in the cw-EPR lineshape In order to further decouple the motions due to conformational heterogeneity of spin label rotamers from the backbone motion, 170 GHz and 240 GHz measurements (known as HFHF, or High Field High Frequency) are currently being performed at ACERT in Cornell, with the collabor ation of Profs. K. Earle and J. Freed. Some of the main advantages of performing HFHF e xperiments are (1) higher resolution of the g -tensor anisotropy, allowing for the possi bility of discriminating between different conformations with

PAGE 193

193 different g values and (2) the faster timescales accessible to the technique. We already obtained the first results from these experiments, showing that there are noticeable differences in the high field lineshape between PMPR*M TSL in its unbound and inhibitor bound forms. We expect that results obtained from these studies will further increase our knowledge on spin label conformers and conformer dynamics. Spin Label Dynamics from MD Simulations As a result of this work, a collaboration with Dr. Carlos Simmerling in order to study other protease systems with the aid of MD and pulsedEPR was established. The results obtained by his group145, indicate that pulsed-EPR distance meas urements and MD simulations are very powerful complementary techniques to study pr otein dynamics and conformational changes. From MD simulations, the dihe dral angle values and distri butions can be obtained as a function of simulation time. This angular distributi ons of the label are curre ntly a hot topic in the spin-labeling community, as a broader understandi ng of what are the possible conformations of the dihedral angles of the spin label will lead to better and more accurate representations of distance measurements by EPR, as well as cw-EPR experiments in which the motion of the label has been associated to report backbone motio ns. Currently, Dr. Simmerling is performing MD simulations for the V6 and MDR769 drug resist ant strains of the protease and a comparison between the results obtained by MD and pulsed-EP R for the multidrug resistant strains V6 and MDR769 will be repor ted elsewhere.

PAGE 194

194 APPENDIX A COMMON DRUG-SELECTED MUTATI ONS IN HIV-1 PROTEASE Table A-1. Common dr ug-induced mutations in HIV-1 protease Position LAI AA Comment 10 L IVFR Y L10I/V/F/R are associated with resist ance to each of the PIs when present with other mutations. L10I/V occur commonly in 5-10% of untreated persons. L10FRY are nonpolymorphic. 11 V I V11I is a nonpolymorphic accessory mutation weakly associated with PI therapy. It is one of the 11 mutations associated with decreased response to DRV in the POWER trials. 13 I V I13V is a common polymorphism that is slightly more common in treated compared with untreated subtype B is olates. In subtypes A, AE, AG, and G it is the consensus residue. I13V w as weakly associated with decreased virological response to TPV in the RESIST trials. 16 G EA G16E is a highly polymorphic mutation occurring in 2% to 20% of viruses depending on subtype. It may be w eakly associated with ATV therapy and response. 20 K RMI VT K20R/M/I/T are weakly associated with resistance to each of the PIs when present with other mutations. Many variants at this position occur commonly in non-subtype B viruses. 23 L I L23I is a rare substrate cleft mutation that causes low-level resistance to NFV. 24 L I L24I is associated with reduced susceptibility to IDV and possibly RTV, LPV, SQV, and ATV when present with other mutations. 24 L F L24F is a rare mutation at this position; its phenotypic effect is not known. 30 D N D30N causes intermediate resistance to NFV. 32 V I V32I is a substrate cleft mutation that reduces susceptibility to IDV, RTV, FPV, LPV, TPV, and DRV. 33 L F L33F are selected by FPV, DRV, LPV, ATV, and TPV, and slightly contribute resistance to these drugs. 33 L I L33F are selected by FPV, DRV, LPV, ATV, and TPV, and slightly contribute resistance to these drugs. L33I is less common than L33F but may have a similar effect. 33 L V L33V is a polymorphism that does not appear to be related to PI therapy or drug resistance. 35 E G E35 G is a nonpolymorphic mutation that is slightly more common in viruses from PI-treated (particularly NFVtreated) compared with untreated persons. It was weakly associated a d ecreased virological response to TPV in the RESIST trials. 36 M IVLT M36I/V are weakly associated with PI resistance when present with other mutations. M36I occurs commonly in certain non-subtype B viruses. M36L/T are rare variants of uncertain significance. 43 K T K43T is a nonpolymorphic mutation that is significantly more common in viruses from PI-treated compared with untreated persons. It was weakly associated with a decreased virologi cal response to TPV in the RESIST trials. 46 M IL M46I/L decreases susceptibility to IDV, NFV, FPV, LPV, ATV, TPV, and possibly SQV and DRV when present with other mutations.

PAGE 195

195 Table A-1. Continued Position LAI AA Comment 47 I V I47V decrease susceptibility to FPV, ATV, IDV, LPV, TPV, and DRV. 47 I A I47A usually occurs in combinati on with V32I and in this setting causes highlevel LPV and FPV resistance and probably intermediate DRV resistance. 48 G V G48V causes high-level SQV resistance, intermediate ATV and NFV resistance, and low-level IDV and LPV resistance. 48 G M G48M causes high-level SQV resistance and intermediate resistance to NFV, ATV, IDV, and NFV. 50 I L I50L causes intermediate-to-high level resistance to ATV and hypersusceptibility to the remaining PIs. 50 I V I50V causes intermediate-to-high-le vel resistance to FPV, intermediate resistance to LPV and DRV, and increased susceptibility to TPV. 53 F LY F53L occurs only in isolates from treated patients nearly always in combinations with other PI-resistance mutations. It is associated with decreased susceptibility to IDV, LPV and SQV and possibly other PIs. F53Y is a rare treatment-associated variant. 54 I V I54V contributes resistance to each of the PIs except possibly DRV when present with other mutations. 54 I M I54M occurs in patients receiving FPV, LPV, and DRV and reduces susceptibility to these drugs and to ATV and NFV. 54 I L I54L occurs in patients receiving FPV, LPV, and DRV and reduces susceptibility to these drugs and to NFV and possibly ATV. It is associated with increased susceptibility to TPV. 54 I ATS I54T/A/S are PI-related mutations typically observed in heavily treated persons and like I54V and I54M/L are likely to reduce susceptibility to most PIs. 58 Q E Q58E is a nonpolymorphic mutation that occurs more commonly in treated patients. It has been associated with decreased virological response to TPV but probably is associated with resistance to multiple PIs. 60 D E D60E is a polymorphic mutation that is slightly more common in viruses from PI-treated compared with untreat ed persons. It was weakly associated with a decreased virologic response to ATV in one of three retrospective analyses. 62 I V I62V is a highly polymorphic mutation that is more common in PI-treated compared with untreated persons. It was weakly associated with a decreased virologic response to ATV in one of three retrospective analyses. 63 L P L63P is a common polymorphism that becomes even more common in persons receiving PIs. 69 H K H69K is a highly polymorphic residue that was weakly associated with a decreased virologic response to TPV in the RESIST trials. 71 A VTIL A71T/V are polymorphisms that occur in 1-2% of untreated persons but which become much more frequent in persons receiving PIs. A71I is an uncommon mutation that appears to occur only with PI therapy; its significance is not known. 73 G STC A G73S/T/C/A are selected by and appear to be associated with decreased susceptibility to most, if not all, PIs.

PAGE 196

196 Table A-1. Continued Position LAI AA Comment 74 T SP T74S is associated with reduced NFV susceptibility. It o ccasionally occurs in untreated persons. T74P is a nonpolymorphic mutation that occurs more commonly in heavily treated patients. It has been associated with decreased virological response to TPV but probab ly is associated with resistance to multiple PIs. 74 T A T74A is a polymorphic mutation more common in viruses from persons receiving PIs (particularly NFV and SQV) than in PI-naive persons. 76 L V L76V reduces susceptibility to FPV, IDV, LPV, and DRV and increases susceptibility to SQV and ATV. 77 V I V77I is a common polymorphism that is associated with NFV therapy. 82 V A V82A reduces susceptibility to IDV and LPV. With other mutations it is associated with reduced susceptibility to NFV, ATV, SQV, FPV, and TPV. 82 V T V82T reduces susceptibility to IDV, LPV, and TPV. With other mutations it is associated with reduced susceptib ility to NFV, ATV, FPV, and SQV. 82 V F V82F reduces susceptibility to IDV, LPV, FPV, and DRV. With other mutations it is associated with redu ced susceptibility to NFV, ATV, SQV, and TPV. 82 V S V82S probably has a similar effect to V82T which reduces susceptibility to IDV, LPV, and TPV and with other mutations is associated with reduced susceptibility to NFV, ATV, FPV, and SQV. 82 V M V82M has been shown to cause IDV resistance in subtype G viruses in which it is a common mutation. Its effect on other PIs has not been studied. 82 V I V82I is a polymorphism that is common in some non-B subtypes; it has little if any effect on PI susceptibility. 82 V L V82L is a rare TPV-selected mutation. Its effect on other PIs has not been characterized. 82 V C V82C is a rare mutation that occu rs primarily in protease genes containing multiple other PI-resistance mutations. 83 N D N83D is a nonpolymorphic mutation that occurs more commonly in heavily treated patients. It has been associat ed with decreased virological response to TPV but probably is associated w ith resistance to multiple PIs. 84 I V I84V causes intermediate/high-level resistance to ATV, FPV, IDV, NFV, SQV, and TPV, and low-level intermed iate resistance to LPV and DRV. 84 I AC I84A/C are rare mutations that like I84V appear to reduce susceptibility to each of the PIs. 85 I V I85V is a nonpolymorphic PI-selected mutation. It was weakly associated with a decreased virologic response to ATV in one of three retrospective analyses. 88 N S N88S causes high-level resistance to NFV and ATV and low-level resistance to IDV; it increases su sceptibility to FPV. 88 N T N88T has not been studied but is likely to be similar to N88S which causes high-level resistance to NFV and ATV, low-level resistance to IDV, and increased susceptibility to FPV. 88 N G N88 G has not been studied but is likely to be similar to N88S which causes high-level resistance to NFV and ATV, low-level resistance to IDV, and increased susceptibility to FPV.

PAGE 197

197 Table A-1. Continued Position LAI AA Comment 90 L M L90M causes resistance to NFV, SQV, ATV, and IDV. When present with other mutations it also compromises the activity of FPV, LPV, and TPV. Its effect on DRV is not known. 93 I L I93L is a common polymorphism that becomes even more common in persons receiving PIs. 93 I M I93M is a rare PI-associated mutation; its effect on PI susceptibility is not known. Note: Table obtained from Stan ford HIV Resistance Database

PAGE 198

198 APPENDIX B SAMPLE DEER RUN DATA Table B-1. cw-EPR parameters for DEER experiment # Points: 1024 Mod. Amplitude (G): 1 Resonator: MD5 Dielectric Mod. Phase: 27 Center Field (G,GHz): 3449.6 G /9. 691696 Conversion Time (ms): 163.84 Power (mW): 0.00006346 Time Constant (ms): 81.92 Power Attenuation (dB): 55 Sweep Width (G): 200 Reciever Gain (dB): 60 Sweep Time (s): 167.77 33503400345035003550 -1 0 1 Intensity (a.u.)Field (G) CF33503400345035003550 -1 0 1 Intensity (a.u.)Field (G) CF33503400345035003550 -1 0 1 Intensity (a.u.)Field (G) CF Figure B-1. Sample cw-EPR of spin labeled HIVPR at 65 K.

PAGE 199

199 Table B-2. Field-swept echo parameters for DEER experiment Shot Repetition Time (S): 3000 Ce nter Field (G,GHz): 3450.5 / 9.692631 # Points: 1024 Sweep Width (G): 200 # Scans: 1 Pulsed Attenuation (dB): 6 Shots/Point: 50 Video Bandwidth (MHz): 20 Integrator Base (ns): 4 ns Video Gain (dB): 48 +x Pulse Acquisition Trigger 1 2 1 Position 0 200 Position 860 Pulse Length 16 32 Pulse Length 148 Pos. Display 0 0 Pos. Display 0 New Center Field: 3450.5 G 340034503500 0 500 1000 1500 2000 Intensity (a.u.)Field (G)CF 340034503500 0 500 1000 1500 2000 Intensity (a.u.)Field (G)CF 340034503500 0 500 1000 1500 2000 Intensity (a.u.)Field (G)CF Figure B-2. Sample field-swept echo spect ra of spin labeled HIVPR at 65 K.

PAGE 200

200 Table B-3. Echo decay (Tm) experiment Shot Repetition Time (S): 3000 Ce nter Field (G,GHz): 3450.5 / 9.692631 # Points: 1024 Pulsed Attenuation (dB): 6 # Scans: 1 Video Bandwidth (MHz): 20 Shots/Point: 50 Video Gain (dB): 48 Integrator Base (ns): 4 ns +x Pulse Acquisition Trigger 1 2 1 Position 0 200 Position 860 Pulse Length 16 32 Pulse Length 148 Pos. Display 0 8 Pos. Display 16 d2 value (ns) 3000 02000400060008000 -500 0 500 1000 1500 2000 2500 3000 3500 Intensity (a.u.)Time (ns) d2 02000400060008000 -500 0 500 1000 1500 2000 2500 3000 3500 Intensity (a.u.)Time (ns) d2 02000400060008000 -500 0 500 1000 1500 2000 2500 3000 3500 Intensity (a.u.)Time (ns) d2 Figure B-3. Sample Hahn echo experiment of spin labeled HIVPR at 65 K.

PAGE 201

201 Table B-4. Inversion decay (T1) experiment Shot Repetition Time (S): 3000 Ce nter Field (G,GHz): 3450.5 / 9.692631 # Points: 1024 Pulsed Attenuation (dB): 6 # Scans: 1 Video Bandwidth (MHz): 20 Shots/Point: 50 Video Gain (dB): 48 Integrator Base (ns): ---+x Pulse Acquisition Trigger 1 2 1 Position 0 400 Position 600 Pulse Length 32 16 Pulse Length --Pos. Display 0 2000 Pos. Display 4000 SRT Value (s) 3200 02004006008001000 -1500 -1000 -500 0 500 1000 1500 2000 2500 Intensity (a.u.)Time (s)02004006008001000 -1500 -1000 -500 0 500 1000 1500 2000 2500 Intensity (a.u.)Time (s) Figure B-4. Sample inversion recovery expe riment of spin labeled HIVPR at 65 K.

PAGE 202

202 Table B-5. 4p-DEER setup Shot Repetition Time (S): 3200 Ce nter Field (G,GHz): 3450.5 / 9.692631 # Points: 256 Pulsed Attenuation (dB): 6 # Scans: 1 Video Bandwidth (MHz): 20 Shots/Point: 100 Video Gain (dB): 69 Phase Cycling: 8-step ELDOR Frequency (GHz): 9.692647 /2 Pulse (ns) : 16 ELDOR Attn (dB): 30 ELDOR (ns): 32 DEER Parameters d1 (ns) 200 d2 (ns) 3000 PG (ns) 108 d3 (ns) 100 D0 (ns) 448 d2 (ns) 3000 02004006008001000 -40 -30 -20 -10 0 10 Intensity (a.u.)Time (ns) d0 PG02004006008001000 -40 -30 -20 -10 0 10 Intensity (a.u.)Time (ns) d0 PG02004006008001000 -40 -30 -20 -10 0 10 Intensity (a.u.)Time (ns) d0 PG Figure B-5. Sample refocused spin echo parameters on HIVPR sample.

PAGE 203

203 Table B-6. Field swept 4p-DEER Shot Repetition Time (S): 3200 Ce nter Field (G,GHz): 3450.5 / 9.692631 # Points: 1024 Pulsed Attenuation (dB): 6 # Scans: 1 Video Bandwidth (MHz): 20 Shots/Point: 100 Video Gain (dB): 69 Phase Cycling: 2-step ELDOR Frequency (GHz): 9.692647 /2 Pulse (ns) : 16 ELDOR Attn (dB): 30 Sweep Width (G): 160 ELDOR (ns): 32 DEER Parameters d0 (ns) 480 d2 (ns) 3000 d1 (ns) 200 d3 (ns) 100 PG (ns) 108 Det. Pulse Pos.(G/GHz) 3424 G ELDOR Pump Pos (G): 3450 G Pump-Detection Dist. (G): 26 G 340034503500 -600 -400 -200 0 200 Intensity (a.u.)Field (G) Pump Observe340034503500 -600 -400 -200 0 200 Intensity (a.u.)Field (G) Pump Observe340034503500 -600 -400 -200 0 200 Intensity (a.u.)Field (G) Pump Observe Figure B-6. Sample field-swept spectra on a refocused echo on spin labeled HIVPR at 65 K.

PAGE 204

204 Table B-7. 4p-DEER experiment Shot Repetition Time (S): 3200 Center Field (G,GHz): 3424.0 # Points: 129 Pulsed Attenuation (dB): 6 # Scans: 30000 Video Bandwidth (MHz): 20 Shots/Point: 100 Video Gain (dB): 69 Phase Cycling: 2-step ELDOR Frequency (GHz): 9.61947 /2 Pulse (ns) : 16 ELDOR Attn (dB): 0 ELDOR (ns): 32 DEER Parameters d0 (ns) 0 d2 (ns) 3000 d1 (ns) 200 d3 (ns) 100 PG (ns) 108 dx (ns) 12 050010001500200025003000 -1200 -1150 -1100 -1050 -1000 -950 -900 -850 Intensity (a.u.)Time (ns) p 050010001500200025003000 -1200 -1150 -1100 -1050 -1000 -950 -900 -850 Intensity (a.u.)Time (ns) p Figure B-7. Sample dipolar evolution in the DEER experiment of spin labeled HIVPR at 65 K.

PAGE 205

205 APPENDIX C FDA APPROVED PROTEASE INHIBI TORS FOR HIV-1 TREATMENT Table C-1. FDA approved protease inhi bitors for HIV-1 HAART treatment Inhibitor Structure Formula Mol. wt Solubility Saquinavir C38H50N6O5CH2Cl2 675.10 DMSO CH3OH Ritonavir C37H48N6O5S2 720.96 DMSO Chlorof. Toluene Atazanavir C38H52N6O7H2SO4 802.9 4-5 mg/ml H2O DMSO EtOH Indinavir C36H47N5O4H2SO4 711.88 H2O CH3OH Tipranavir C31H33F3N2O5S 602.7 EtOAc

PAGE 206

206 Table C-1. Continued Inhibitor Structure Formula Mol.wt. Solubility Darunavir (TMC114) C27H37N3O7S C2H5OH 593.73 0.15 mg/ml H2O Lopinavir C37H48N4O5 628.81 DMSO CH3OH EtOH CH2Cl2 MetCl Amprenavir C25H35N3O6S 505.63 DMSO CH3OH EtOH CH2Cl2 MetCl

PAGE 207

207 APPENDIX D INTERACTIONS BETWEEN HIVPR AND RITONAVIR Figure D-1. HIVPR crystal structure with Ritonavi r (blue, space fill rendering) and residues that interact directly with the i nhibitor in VdW representation. Figure D-2. Interactions between Ritonavir and HIV-1 protease.

PAGE 208

208 Table D-1. Summary of closes t neighbor interactions betwee n Ritonavir and HIV-1 protease Atom1 Atom2 Distance Type RIT301:N5 B:ASP30:N 2.99Hydrophilic RIT301:N5 B:ASP30:O 3.14Hydrophilic RIT301:N20 A:GLY48:O 3.12Hydrophilic RIT301:O41 A:ASP25:OD1 2.66Hydrophilic RIT301:O41 A:ASP25:OD2 3.08Hydrophilic RIT301:O41 B:ASP25:OD1 2.65Hydrophilic RIT301:O41 B:ASP25:OD2 2.74Hydrophilic RIT301:N58 A:GLY27:O 3.18Hydrophilic RIT301:O76 A:ASP29:N 3.09Hydrophilic RIT301:C1 B:ALA28:CB 3.88Hydrophobic RIT301:C1 B:VAL32:CG2 3.29Hydrophobic RIT301:C31 A:ILE84:CD1 3.79Hydrophobic RIT301:C32 B:GLY49:CA 3.7Hydrophobic RIT301:C32 B:GLY49:C 3.38Hydrophobic RIT301:C32 B:ILE50:CA 3.76Hydrophobic RIT301:C33 A:VAL82:CG1 3.74Hydrophobic RIT301:C34 A:VAL82:CG1 3.64Hydrophobic RIT301:C45 B:VAL82:CG2 3.75Hydrophobic RIT301:C48 B:VAL82:CG2 3.88Hydrophobic RIT301:C50 A:GLY49:CA 3.52Hydrophobic RIT301:C50 A:GLY49:C 3.89Hydrophobic RIT301:C50 B:PRO81:CG 3.69Hydrophobic RIT301:C51 A:GLY49:CA 3.45Hydrophobic RIT301:C51 A:GLY49:C 3.43Hydrophobic RIT301:C51 B:PRO81:CG 3.71Hydrophobic RIT301:C52 B:VAL82:CG2 3.77Hydrophobic RIT301:C64 A:ILE84:CD1 3.76Hydrophobic RIT301:C77 B:ARG8:CZ 3.61Hydrophobic RIT301:C80 B:ARG8:CZ 3.54Hydrophobic RIT301:S3 HOH51:1H 3.05Ligand-H2O RIT301:O24 HOH1:O 2.7Ligand-H2O RIT301:O24 HOH1:1H 3.11Ligand-H2O RIT301:O61 HOH1:O 2.64Ligand-H2O RIT301:O61 HOH1:2H 3.08Ligand-H2O RIT301:N74 HOH36:2H 2.56Ligand-H2O RIT301:O76 HOH5:O 2.88Ligand-H2O RIT301:O76 HOH5:2H 3.28Ligand-H2O RIT301:N83 HOH10:O 2.94Ligand-H2O RIT301:N83 HOH10:1H 2.74Ligand-H2O A:ILE50:N HOH1:O 2.88Ligand-Protein (by H2O) B:ILE50:N HOH1:O 3.06Ligand-Protein (by H2O) A:GLY27:O HOH5:O 2.89Ligand-Protein (by H2O) A:ASP29:OD1 HOH5:O 2.61Ligand-Protein (by H2O) A:GLY27:O HOH5:2H 2.97Ligand-Protein (by H2O)

PAGE 209

209 APPENDIX E PCR PRIMERS AND SITE-DIRECTED MUTAGENESIS PARAMETERS Table E-1. PCR primers utilized to introduce mutations in HIVPR Mutation Primer (5 3) Tm ( C) % GC K55C Forward GTATCATCTGCACCGGTATTCAGCAGCGCTTCTTTCAG 65.8 50.0 Reverse GTATCATCTGCACCGGTATTCAGCAGCGCTTCTTTCAG 65.8 50.0D25N Forward CTGAAAGAAGCGCTGCTGAATACCGGTGCAGATGATACCG 67.5 52.5 Reverse CGGTATCATCTGCACCGGTATTCAGCAGCGCTTCTTTCAG 67.5 52.5T74C Forward CAAGGCAATTGGTTGCGTGCTGGTTGGCC 58.6 67.5 Reverse GGCCAACCAGCACGCAACCAATTGCCTTG 58.6 67.5F53C Forward GCGGTATTGGTGGTTGCATTAAAGTGCGCC 65.1 53.3 Reverse GGCGCACTTTAATGCAACCACCAATACCGC 65.1 53.3M46I Forward CGTTGGAAACCTAAAATTATTGGCGGTATTGGTGG 62.2 42.9 Reverse CCACCAATACCGCCAATAATTTTAGGTTTCCAACG 62.2 42.9I54V Forward CGGTATTGGTGGTTTCGTGTGTGTGCGCCAGTAC 67.3 55.9 Reverse GTACTGGCGCACACACACGAAACCACCAATACCG 67.3 55.9 Table E-2. Non codon-optimized PM PR DNA and protein sequences DNA CCTCAGATCACTCTTTGGAAACGACCCCTCGTCACAATAAAGATAGGGGGGCAAC TAAAGGAAGCTCTATTAGATACAGGAGCAGATGATACAGTAATAGAAGAAATGAG TTTGCCAGGAAGATGGAAACCAAAAATGATAGGGGGAATTGGAGGTTTTATCAAA GTAAGACAGTATGATCAGATAATCATAGAAATCGCGGGACATAAAGCTATAGGTA CAGTATTAGTAGGACCTACACCTGTCAACATAATTGGAAGAAATCTGTTGACTCA GATTGGTGCGACTTTAAATTTTTA Protein PQITLWKRPL VTIKIGGQLK EALLDTGADD TVIEEMSLPG RWKPKMIGGI GGFIKVRQYD QIIIEIAGHK AIGTVLVGPT PVNIIGRNLL TQIGATLNF

PAGE 210

210 Table E-3. Codon-optimized PMPR DNA and protein sequences DNA CCACAAATCACTCTGTGGAAACGTCCGCTGGTCACCATTAAAATTGGCGGTCAAC TGAAAGAAGCGCTGCTGGACACCGGTGCAGATGATACCGTTATCGAGGAAATGAG CCTGCCGGGTCGTTGGAAACCTAAAATGATTGGCGGTATTGGTGGTTTCATTAAA GTGCGCCAGTACGACCAGATCATTATCGAAATCGCCGGCCACAAGGCAATTGGTA CCGTGCTGGTTGGCCCGACCCCGGTTAACATCATCGGCCGCAACCTGCTGACTCA GATTGGCGCCACGCTGAACTTC Protein PQITLWKRPL VTIKIGGQLK EALLDTGADD TVIEEMSLPG RWKPKMIGGI GGFIKVRQYD QIIIEIAGHK AIGTVLVGPT PVNIIGRNLL TQIGATLNF Table E-4. CO-PMPR+D25N DNA and protein sequences DNA CCACAAATCACTCTGTGGAAACGTCCGCTGGTCACCATTAAAATTGGCGGTCAAC TGAAAGAAGCGCTGCTGAATACCGGTGCAGATGATACCGTTATCGAGGAAATGAG CCTGCCGGGTCGTTGGAAACCTAAAATGATTGGCGGTATTGGTGGTTTCATTAAA GTGCGCCAGTACGACCAGATCATTATCGAAATCGCCGGCCACAAGGCAATTGGTA CCGTGCTGGTTGGCCCGACCCCGGTTAACATCATCGGCCGCAACCTGCTGACTCA GATTGGCGCCACGCTGAACTTC Protein PQITLWKRPL VTIKIGGQLK EALLNTGADD TVIEEMSLPG RWKPKMIGGI GGFIKVRQYD QIIIEIAGHK AIGTVLVGPT PVNIIGRNLL TQIGATLNF Table E-5. CO-PMPR+K55C DNA and protein sequences DNA CCACAAATCACTCTGTGGAAACGTCCGCTGGTCACCATTAAAATTGGCGGTCAAC TGAAAGAAGCGCTGCTGGACACCGGTGCAGATGATACCGTTATCGAGGAAATGAG CCTGCCGGGTCGTTGGAAACCTAAAATGATTGGCGGTATTGGTGGTTTCATTTGT GTGCGCCAGTACGACCAGATCATTATCGAAATCGCCGGCCACAAGGCAATTGGTA CCGTGCTGGTTGGCCCGACCCCGGTTAACATCATCGGCCGCAACCTGCTGACTCA GATTGGCGCCACGCTGAACTTC Protein PQITLWKRPL VTIKIGGQLK EALLDTGADD TVIEEMSLPG RWKPKMIGGI GGFICVRQYD QIIIEIAGHK AIGTVLVGPT PVNIIGRNLL TQIGATLNF Table E-6. CO-PMPR+D25N+K55 C DNA and protein sequences DNA CCACAAATCACTCTGTGGAAACGTCCGCTGGTCACCATTAAAATTGGCGGTCAAC TGAAAGAAGCGCTGCTGAATACCGGTGCAGATGATACCGTTATCGAGGAAATGAG CCTGCCGGGTCGTTGGAAACCTAAAATGATTGGCGGTATTGGTGGTTTCATTTGT GTGCGCCAGTACGACCAGATCATTATCGAAATCGCCGGCCACAAGGCAATTGGTA CCGTGCTGGTTGGCCCGACCCCGGTTAACATCATCGGCCGCAACCTGCTGACTCA GATTGGCGCCACGCTGAACTTC Protein PQITLWKRPL VTIKIGGQLK EALLNTGADD TVIEEMSLPG RWKPKMIGGI GGFICVRQYD QIIIEIAGHK AIGTVLVGPT PVNIIGRNLL TQIGATLNF

PAGE 211

211 Table E-7. MDR769 DNA and protein sequences DNA CCTCAGATTACCCTGTGGCAACGTCCTATTGTGACTATCAAAATTGGTGGTCAGC TGAAAGAAGCGCTGCTGAACACTGGCGCTGATGATACGGTGCTGGAGGAAGTATC CCTGCCAGGCCGTTGGAAACCAAAACTGATTGGTGGTATCGGCGGCTTCGTTTGT GTTCGCCAGTACGACCAGGTCCCGATCGAGATTGCTGGCCACAAAGTTATTGGTA CTGTGCTGGTTGGTCCGACTCCGGCGAACGTGATCGGCCGTAATCTGATGACGCA AATTGGCGCCACTCTGAACTTC Protein PQITLWQRPI VTIKIGGQLK EALLNTGADD TVLEEVSLPG RWKPKLIGGI GGFVCVRQYD QVPIEIAGHK VIGTVLVGPT PANVIGRNLM TQIGATLNF Table E-8. V6 DNA a nd protein sequences DNA CCACAGATTACCCTGTGGCAGCGTCCACTGGTCACCATCAAAATTGGTGGTCAGC TGCGTGAAGCGCTGCTGAACACGGGTGCGGATGACACTATTTTCGAAGAAATCTC CCTGCCTGGTCGTTGGAAACCAAAAATGATTGGCGGCATCGGTGGTTTCATTTGC GTGCGCCAGTACGACCAGATTCCAATTGAGATCGCAGGCCATAAGGTTATCGGTA CCGTTCTGGTAGGTCCGACCCCGGCGAACATTATTGGTCGTAATCTGATGACTCA GATTGGTGCGACCCTGAACTTC Protein PQITLWQRPL VTIKIGGQLR EALLNTGADD TIFEEISLPG RWKPKMIGGI GGFICVRQYD QIPIEIAGHK VIGTVLVGPT PANIIGRNLM TQIGATLNF Table E-9. Thermal cycling parameters fo r HIV-1 protease site-directed mutagenesis Segment Cycles Temperature Time 1 1 95 C30 sec. 95 C30 sec. 55 C1 min. 2 16-18 68 C2 min./kb plasmid

PAGE 212

212 APPENDIX F CIRCULAR DICHROISM AND CW-EPR PARAMETERS Table F-1. Typical parameters used for circular dichroism experiments Parameter Value Experiment Type Wavelength Bandwidth 1 nm Temp. Setpoint 25.00 C Wavelength Start 250 nm Wavelength End 200 nm Wavelength Step 1.0 nm Averaging Time 1.000 sec Settling Time 0.333 sec Multi-scan Wait 1 sec Scans 4 Table F-2. Typical cw-EPR pa rameters used in this study Parameter Value # Points 1024 Center Field ~3450 G # Scans 1-32 Sweep Width 20-100 G Acquisition Time 40.63 sec Frequency ~ 9.6-9.7 GHz Power 20 dB 2 mW Reciever Gain 54 1105 Modulation Amplitude ~1 G (Uncalibrated) Time Constant 0.082-0.164 s Receiver Phase 100

PAGE 213

213 APPENDIX G SPECTRAL PARAMETERS FOR THE EPR SP ECTRA REPORTED IN THIS WORK Table G-1. Spectral parameters for the EPR spectra for the osmolality/viscosity effects Spin Label Sample Hpp

LF/CF MTSL 2 mM NaOAc 1.991840.62 10% Glycerol 2.151990.56 20% Glycerol 2.322010.50 30% Glycerol 2.642010.48 40% Glycerol 2.931980.47 6% Sucrose 2.111990.57 12% Sucrose 2.171990.55 18% Sucrose 2.211940.53 3% Ficoll 1.911920.61 6% Ficoll 2.011820.58 9% Ficoll 2.031850.57 6% PEG 3000 2.111920.56 12% PEG 3000 2.211990.51 18% PEG 3000 2.362030.47 24% PEG 3000 2.422050.47 MSL 2 mM NaOAc 2.912170.54 10% Glycerol 3.132200.47 20% Glycerol 3.342200.43 30% Glycerol 3.702210.39 40% Glycerol 3.962280.37 6% Sucrose 2.972140.51 12% Sucrose 3.052230.49 18% Sucrose 3.132190.46 3% Ficoll 2.972100.53 6% Ficoll 2.992200.50 9% Ficoll 2.972150.49 6% PEG 3000 2.932170.53 12% PEG 3000 3.092160.51 18% PEG 3000 3.092180.48 24% PEG 3000 3.152240.47 IAP 2 mM NaOAc 1.861780.67 10% Glycerol 1.911840.61 20% Glycerol 2.051800.55 30% Glycerol 2.151830.53 40% Glycerol 2.481970.47 6% Sucrose 1.901780.65 12% Sucrose 1.951870.62

PAGE 214

214 Table G-1.Continued. Spin Label Sample Hpp

LF/CF 18% Sucrose 1.991860.59 3% Ficoll 1.841820.66 6% Ficoll 1.881830.64 9% Ficoll 1.861770.63 6% PEG 3000 1.911900.63 12% PEG 3000 1.901790.61 18% PEG 3000 1.931750.60 24% PEG 3000 1.911870.59 IASL 2 mM NaOAc 2.211980.96 10% Glycerol 2.362000.93 20% Glycerol 2.482060.88 30% Glycerol 2.692080.87 40% Glycerol 2.892080.84 6% Sucrose 2.242030.95 12% Sucrose 2.262030.93 18% Sucrose 2.282040.91 3% Ficoll 2.251970.96 6% Ficoll 2.251990.95 9% Ficoll 2.252000.94 6% PEG 3000 2.232000.96 12% PEG 3000 2.292010.95 18% PEG 3000 2.322010.95 24% PEG 3000 2.402010.95

PAGE 215

215 Table G-2. Lineshape spectral parameters for the degradation of PMPR+K55MSL Time (days) Hpp

LF/CF HF/CF HF/LF 0 3.03 2130.5100 6 2.93 2030.480.0800.166 7 3.13 2210.450.0860.192 8 3.13 2300.440.0890.203 10 3.13 2210.450.1020.225 11 3.03 2190.470.1100.233 12 3.13 2120.470.1090.233 13 3.03 2190.480.1250.261 14 3.03 2210.470.1270.268 18 2.93 2140.490.1630.330 20 3.03 2220.490.1640.337 22 2.93 2160.510.1650.322 23 2.93 2190.510.1700.334 25 3.03 2200.520.1740.337 26 3.13 2180.510.1750.343 27 3.03 2260.510.1740.343 28 2.93 2190.520.1780.343 29 2.83 2200.530.1910.359 30 2.93 2240.540.2010.375 31 2.93 2140.560.2040.365 32 2.83 2210.540.2020.371 34 2.93 2250.550.2060.372 37 2.83 2200.570.2180.384 39 2.83 2160.600.2240.375 40 2.83 2200.580.2220.382 42 2.83 2240.580.2320.397 45 2.64 2210.610.2450.405

PAGE 216

216 APPENDIX H 3D PLOTS FOR SPECTRAL PARAMETER S AND CW-EPR LINESHAPES AS A FUNCTION OF COSOLUTES FOR OS MOLALITY/VISCOSITY STUDIES

PAGE 217

217 79 80 81 82 83 0 2 4 6 8 1 2 3 4 % C h a n g eV i s c o s i t y (c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% 2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 250 500 750 0 2 4 6 8 10 12 14 16 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 100 200 300 400 0 5 10 15 20 25 1 2 3 4 5 6 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 10 20 30 40 50 0 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CFA B D C 79 80 81 82 83 0 2 4 6 8 1 2 3 4 % C h a n g eV i s c o s i t y (c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% 2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 250 500 750 0 2 4 6 8 10 12 14 16 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 100 200 300 400 0 5 10 15 20 25 1 2 3 4 5 6 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 10 20 30 40 50 0 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CFA B D C2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% 2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 250 500 750 0 2 4 6 8 10 12 14 16 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 100 200 300 400 0 5 10 15 20 25 1 2 3 4 5 6 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 10 20 30 40 50 0 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CFA B D C Figure H-1. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR-PMPR*MTSL as a function of cosolute concentration, a nd 3D plot showing pe rcent change in the spectral parameters Hpp (red) ,

(blue) and LF/CF ratio (black) as a function of osmolality and viscosity for the corres ponding EPR spectra. A) Ficoll 400. B) Sucrose. C) PEG3000. D) Glycerol.

PAGE 218

218 2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% 79 80 81 82 83 -2 0 2 4 6 8 10 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 5 10 15 20 25 30 35 40 0 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) Hpp

LF/CF 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% 100 200 300 400 0 2 4 6 8 10 12 14 1 2 3 4 5 6 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) Hpp

LF/CF 2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% 250 500 750 0 5 10 15 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) Hpp

LF/CF Hpp

LF/CFA B D C2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% 79 80 81 82 83 -2 0 2 4 6 8 10 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 5 10 15 20 25 30 35 40 0 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) Hpp

LF/CF Hpp

LF/CF 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% 100 200 300 400 0 2 4 6 8 10 12 14 1 2 3 4 5 6 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) Hpp

LF/CF Hpp

LF/CF 2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% 250 500 750 0 5 10 15 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CFA B D C Figure H-2. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR PMPR*MSL as a function of cosolute concentration, and 3D plot showing percent change in the spectral parameters Hpp (red) ,

(blue) and LF/CF ratio (black) as a function of osmolality and viscosity for the corres ponding EPR spectra. A) Ficoll 400. B) Sucrose. C) PEG3000. D) Glycerol.

PAGE 219

219 79 80 81 82 83 0 2 4 6 8 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 10 20 30 40 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 250 500 750 0 2 4 6 8 10 12 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 100 200 300 400 0 2 4 6 8 10 12 14 1 2 3 4 5 6 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% 2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% 2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CFA B D C 79 80 81 82 83 0 2 4 6 8 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 10 20 30 40 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 250 500 750 0 2 4 6 8 10 12 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 100 200 300 400 0 2 4 6 8 10 12 14 1 2 3 4 5 6 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% 2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% 2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CFA B D C Figure H-3. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR PMPR*IAP as a function of cosolute concen tration, and 3D plot show ing percent change in the spectral parameters Hpp (red) ,

(blue) and LF/CF ratio (black) as a function of osmolality and viscosity for the corres ponding EPR spectra. A) Ficoll 400. B) Sucrose. C) PEG3000 D) Glycerol.

PAGE 220

220 79 80 81 82 83 0 1 2 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 5 10 15 20 25 30 35 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% Hpp

LF/CF 250 500 750 0 1 2 3 4 5 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% Hpp

LF/CF 100 200 300 400 0 2 4 6 8 10 1 2 3 4 5 6 V i s c o s i t y ( c S )% C h a n g eO s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% Hpp

LF/CF Hpp

LF/CFA B D C 79 80 81 82 83 0 1 2 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2500 5000 7500 10000 0 5 10 15 20 25 30 35 1 2 3 4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc Ficoll 3% Ficoll 6% Ficoll 9% Ficoll 12% 2mM NaOAc Glycerol 10% Glycerol 20% Glycerol 30% Glycerol 40% Hpp

LF/CF Hpp

LF/CF 250 500 750 0 1 2 3 4 5 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% Hpp

LF/CF 250 500 750 0 1 2 3 4 5 0.8 1.0 1.2 1.4 % C h a n g eV i s c o s i t y ( c S )O s m o t i c P r e s s u r e ( m m o l / k g ) 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% 2mM NaOAc Sucrose 6% Sucrose 12% Sucrose 18% Hpp

LF/CF Hpp

LF/CF 100 200 300 400 0 2 4 6 8 10 1 2 3 4 5 6 V i s c o s i t y ( c S )% C h a n g eO s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% Hpp

LF/CF 100 200 300 400 0 2 4 6 8 10 1 2 3 4 5 6 V i s c o s i t y ( c S )% C h a n g eO s m o t i c P r e s s u r e ( m m o l / k g )2mM NaOAc PEG 3000 6% PEG 3000 12% PEG 3000 18% PEG 3000 24% Hpp

LF/CF Hpp

LF/CF Hpp

LF/CF Hpp

LF/CFA B D C Figure H-4. Area normalized 100 G X-Band cw-EPR spectra of 100 M HIVPR-PMPR*IASL as a function of cosolute concentration, and 3D plot showing percent change in the spectral parameters Hpp (red) ,

(blue) and LF/CF ratio (black) as a function of osmolality and viscosity for the corres ponding EPR spectra. A) Ficoll 400. B) Sucrose. C) PEG3000. D) Glycerol

PAGE 221

221 LIST OF REFERENCES 1. Henderson, L. E.; Arthur, L. Research and Reference Reagent Program National Institute of Allergy and Infectious Dis eases, National Institutes of Health, 2007. 2. Shugars, D. C.; Wild, C. T.; Greenwell, T. K.; Matthews, T. J. J. Virol. 1996, 70 298291. 3. Dubay, J. W.; Dubay, S. R.; Shin, H. J.; Hunter, E. J. Virol. 1995, 69 4675-82. 4. Wei, X.; Decker, J. M.; Wang, S.; Hui, H. ; Kappes, J. C.; Wu, X.; Salazar-Gonzalez, J. F.; Salazar, M. G.; Kilby, J. M.; Saag, M. S.; Komarova, N. L.; Nowak, M. A.; Hahn, B. H.; Kwong, P. D.; Shaw, G. M. Nature 2003, 422 307-12. 5. Zhu, X.; Borchers, C.; Bienstock, R. J.; Tomer, K. B. Biochemistry 2000, 39 11194-204. 6. Zhou, T.; Xu, L.; Dey, B.; Hessell, A. J.; Van Ryk, D.; Xiang, S. H.; Yang, X.; Zhang, M. Y.; Zwick, M. B.; Arthos, J.; Burton, D. R.; Dimitrov, D. S.; Sodroski, J.; Wyatt, R.; Nabel, G. J.; Kwong, P. D. Nature 2007, 445 732-7. 7. Wyatt, R.; Kwong, P. D.; Desjardins, E.; Sweet, R. W.; Robinson, J.; Hendrickson, W. A.; Sodroski, J. G. Nature 1998, 393 705-11. 8. Tang, C.; Ndassa, Y.; Summers, M. F. Nat. Struct. Biol. 2002, 9 537-43. 9. Fitzon, T.; Leschonsky, B.; Bieler, K.; Paul us, C.; Schroder, J.; Wolf, H.; Wagner, R. Virology 2000, 268 294-307. 10. Summers, M. F.; Henderson, L. E.; Chance, M. R.; Bess, J. W., Jr.; South, T. L.; Blake, P. R.; Sagi, I.; Perez-Alvarado, G.; Sowder, R. C.; Hare, D. R.; et al. Protein Sci. 1992, 1 563-74. 11. Roques, B. P.; Morellet, N.; de Rocquigny, H.; Demene, H.; Schueler, W.; Jullian, N. Biochimie 1997, 79 673-80. 12. Tang, C.; Loeliger, E.; Luncsford, P.; Kinde, I.; Beckett, D.; Summers, M. F. Proc. Natl. Acad. Sci. U S A 2004, 101 517-22. 13. Cannon, P. M.; Matthews, S.; Clark, N.; Byles, E. D.; Iourin, O.; Hockley, D. J.; Kingsman, S. M.; Kingsman, A. J. J. Virol. 1997, 71 3474-83. 14. Chandra, A.; Gerber, T.; Kaul, S.; Wolf, C.; Demirhan, I.; Chandra, P. FEBS Lett. 1986, 200 327-32. 15. Farmerie, W. G.; Loeb, D. D.; Casavant, N. C.; Hutchison, C. A.; Edgell, M. H.; Swanstrom, R. Science 1987, 236 305-8.

PAGE 222

222 16. Das, K.; Bauman, J. D.; Clark, A. D., Jr.; Frenkel, Y. V.; Lewi, P. J.; Shatkin, A. J.; Hughes, S. H.; Arnold, E. Proc. Natl. Acad. Sci. U S A 2008, 105 1466-71. 17. Madrid, M.; Lukin, J. A.; Madur a, J. D.; Ding, J.; Arnold, E. Proteins 2001, 45 176-82. 18. Chiu, T. K.; Davies, D. R. Curr. Op. Med. Chem. 2004, 4 965-77. 19. Shehu-Xhilaga, M.; Crowe, S. M.; Mak, J. J. Virol. 2001, 75 1834-41. 20. Yarchoan, R.; Klecker, R. W.; Weinhold, K. J.; Markham, P. D.; Lyerly, H. K.; Durack, D. T.; Gelmann, E.; Lehrman, S. N.; Blum, R. M.; Barry, D. W.; et al. Lancet 1986, 1 575-80. 21. Bowersox, J. NIAID AIDS Agenda 1996, 10. 22. Kilby, J. M.; Saag, M. S. Infect. Agents. Dis. 1994, 3 313-23. 23. Baker, R. BETA 1995, 5 9. 24. Mansky, L. M.; Temin, H. M. J. Virol. 1995, 69 5087-94. 25. Drake, J. W.; Charlesworth, B.; Charlesworth, D.; Crow, J. F. Genetics 1998, 148 166786. 26. Duvall, M. G.; Lore, K.; Blaak, H.; Ambrozak, D. A.; Adams, W. C.; Santos, K.; Geldmacher, C.; Mascola, J. R.; McMichael, A. J.; Jaye, A.; Whittle, H. C.; RowlandJones, S. L.; Koup, R. A. J. Virol. 2007, 81 13486-98. 27. Reeves, J. D.; Doms, R. W. J. Gen. Virol. 2002, 83 1253-65. 28. Heeney, J. L.; Dalgleish, A. G.; Weiss, R. A. Science 2006, 313 462-6. 29. Tebit, D. M.; Nankya, I.; Arts, E. J.; Gao, Y. AIDS Rev. 2007, 9 75-87. 30. Simon, F.; Mauclere, P.; Roques, P.; L oussert-Ajaka, I.; Mull er-Trutwin, M. C.; Saragosti, S.; Georges-Courbot, M. C.; Barre-Sinoussi, F.; Brun-Vezinet, F. Nat. Med. 1998, 4 1032-7. 31. De Leys, R.; Vanderborght, B.; Vanden H aesevelde, M.; Heyndrickx, L.; van Geel, A.; Wauters, C.; Bernaerts, R.; Saman, E.; Nijs, P.; Willems, B.; et al. J. Virol. 1990, 64 1207-16. 32. Charneau, P.; Borman, A. M.; Quillent, C.; Guetard, D.; Chamaret, S.; Cohen, J.; Remy, G.; Montagnier, L.; Clavel, F. Virology 1994, 205 247-53.

PAGE 223

223 33. Gao, F.; Bailes, E.; Robertson, D. L.; Ch en, Y.; Rodenburg, C. M.; Michael, S. F.; Cummins, L. B.; Arthur, L. O.; Peeters, M.; Shaw, G. M.; Sharp, P. M.; Hahn, B. H. Nature 1999, 397 436-41. 34. Lambert, D. M.; Petteway, S. R.; McDanal, C. E.; Hart, T. K.; Leary, J. J.; Dreyer, G. B.; Meek, T. D.; Bugelski, P. J.; Bologne si, D. P.; Metcalf, B. W.; et al. Antimicrob Agents Chemother. 1992, 36 982-8. 35. Ashorn, P.; McQuade, T. J.; Thaisrivongs, S.; Tomasselli, A. G.; Tarpley, W. G.; Moss, B. Proc. Natl. Acad. Sci. U S A 1990, 87 7472-6. 36. Abramowitz, N.; Schechter, I.; Berger, A. Biochem. Biophys. Res. Commun. 1967, 29 862-7. 37. Schechter, I.; Berger, A. Biochem. Biophys. Res. Commun. 1967, 27 157-62. 38. Schechter, I.; Berger, A. Biochem. Biophys. Res. Commun. 1968, 32 898-902. 39. Barrett, A. J.; Rawlings, N. D.; Woessner, J. F. Handbook of Proteolytic Enzymes. Academic Press, London, 1998. 40. Barre-Sinoussi, F.; Chermann, J. C.; Rey, F. ; Nugeyre, M. T.; Chamaret, S.; Gruest, J.; Dauguet, C.; Axler-Blin, C.; Vezinet-Br un, F.; Rouzioux, C.; Rozenbaum, W.; Montagnier, L. Science 1983, 220 868-71. 41. Navia, M. A.; Fitzgerald, P. M.; McKeever, B. M.; Leu, C. T.; Heimbach, J. C.; Herber, W. K.; Sigal, I. S.; Darke, P. L.; Springer, J. P. Nature 1989, 337 615-20. 42. Miller, M.; Schneider, J.; Sathyanarayana, B. K.; Toth, M. V.; Marshall, G. R.; Clawson, L.; Selk, L.; Kent, S. B.; Wlodawer, A. Science 1989, 246 1149-52. 43. Gustchina, A.; Weber, I. T. FEBS Lett. 1990, 269 269-72. 44. Wlodawer, A.; Vondrasek, J. Annu. Rev. Biophys. Biomol. Struct. 1998, 27 249-84. 45. Gustchina, A.; Sansom, C.; Prevost, M.; Richelle, J.; Wodak, S. Y.; Wlodawer, A.; Weber, I. T. Protein Eng. 1994, 7 309-17. 46. Wlodawer, A.; Erickson, J. W. Annu. Rev. Biochem. 1993, 62 543-85. 47. Heaslet, H.; Lin, Y. C.; Tam, K.; Torb ett, B. E.; Elder, J. H.; Stout, C. D. Retrovirology 2007, 4 1. 48. Mak, C. C.; Le, V. D.; Lin, Y. C.; Elder, J. H.; Wong, C. H. Bioorg. Med. Chem. Lett. 2001, 11 219-22.

PAGE 224

224 49. Li, M.; Morris, G. M.; Lee, T.; Laco, G. S.; Wong, C. H.; Olson, A. J.; Elder, J. H.; Wlodawer, A.; Gustchina, A. Proteins 2000, 38 29-40. 50. Wlodawer, A.; Gustchina, A. Biochim. Biophys. Acta 2000, 1477 16-34. 51. Rose, R. B.; Craik, C. S.; Douglas, N. L.; Stroud, R. M. Biochemistry 1996, 35 1293344. 52. Kervinen, J.; Lubkowski, J.; Zdanov, A.; Bha tt, D.; Dunn, B. M.; Hui, K. Y.; Powell, D. J.; Kay, J.; Wlodawer, A.; Gustchina, A. Protein Sci. 1998, 7 2314-23. 53. Ishima, R.; Ghirlando, R.; Tozser, J.; Gronenbor n, A. M.; Torchia, D. A.; Louis, J. M. J. Biol. Chem. 2001, 276 49110-6. 54. Ishima, R.; Torchia, D. A.; Lynch, S. M.; Gronenborn, A. M.; Louis, J. M. J. Biol. Chem. 2003, 278 43311-9. 55. Louis, J. M.; Ishima, R.; Nesheiwat, I.; Pannell, L. K.; Lynch, S. M.; Torchia, D. A.; Gronenborn, A. M. J. Biol. Chem. 2003, 278 6085-92. 56. Ishima, R.; Freedberg, D. I.; Wang, Y. X.; Louis, J. M.; Torchia, D. A. Structure 1999, 7 1047-55. 57. Todd, M. J.; Semo, N.; Freire, E. J. Mol. Biol. 1998, 283 475-88. 58. Todd, M. J.; Luque, I.; Velazquez-Campoy, A.; Freire, E. Biochemistry 2000, 39 1187683. 59. Velazquez-Campoy, A.; Todd, M. J.; Freire, E. Biochemistry 2000, 39 2201-7. 60. Todd, M. J.; Freire, E. Proteins 1999, 36 147-56. 61. Hornak, V.; Okur, A.; Rizzo, R. C.; Simmerling, C. J. Am. Chem. Soc. 2006, 128 28123. 62. Hornak, V.; Okur, A.; Rizzo, R. C.; Simmerling, C. Proc. Natl. Acad. Sci. U S A 2006, 103 915-20. 63. Minh, D. D.; Chang, C. E.; Trylsk a, J.; Tozzini, V.; McCammon, J. A. J. Am. Chem. Soc. 2006, 128 6006-7. 64. Wittayanarakul, K.; Hannongbua, S.; Feig, M. J Comput. Chem. 2008, 29 673 685. 65. Lu, Y.; Yang, C. Y.; Wang, S. J. Am. Chem. Soc. 2006, 128 11830-9. 66. Trylska, J.; Grochowski, P.; McCammon, J. A. Protein Sci 2004, 13 513-28.

PAGE 225

225 67. Sirois, S.; Proynov, E. I.; Truchon, J. F.; Tsoukas, C. M.; Salahub, D. R. J Comput Chem 2003, 24 1110-9. 68. Trylska, J.; Bala, P.; Geller, M.; Grochowski, P. Biophys. J. 2002, 83 794-807. 69. Piana, S.; Sebastiani, D. ; Carloni, P.; Parrinello, M. J. Am. Chem. Soc. 2001, 123 8730-7. 70. Scott, W. R.; Schiffer, C. A. Structure Fold. Des. 2000, 8 1259-65. 71. Chang, C. E.; Trylska, J.; Tozzini, V.; McCammon, J. A. Chem. Biol. Drug. Des. 2007, 69 5-13. 72. Tozzini, V.; Trylska, J.; Chang, C. E.; McCammon, J. A. J. Struct. Biol. 2007, 157 60615. 73. Chang, C. E.; Shen, T.; Trylska, J.; Tozzini, V.; McCammon, J. A. Biophys. J 2006, 90 3880-5. 74. Freedberg, D. I.; Ishima, R.; Jacob, J.; Wang, Y. X.; Kustanovich, I.; Louis, J. M.; Torchia, D. A. Protein Sci. 2002, 11 221-32. 75. Poole, C. P. Electron Spin Resonance: A comprehe nsive Treatise on Experimental Techniques ; Dover Publications, USA, 1983. 76. Weil, J. A.; Bolton, J. R.; Wertz, J. E. Electron Spin Resonance: Elementary Theory and Practical Applications ; John Wiley & Sons, Inc., USA, 1972. 77. Zeeman, P. Astrophysical Journal 1897, 5 332. 78. Fermi, E. Z. Phys. 1930, 60, 642. 79. Carrington, A.; McLachlan, A. D. Introduction to Magnetic Resonance with applications to chemistry and chemical physics ; Harper and Row, New York and London, 1967. 80. Abragam, A. Principles of Nuclear Magnetism ; Oxford University Press, USA, 1961. 81. Ernst, R. R.; Bodenhausen, G.; Wokaun, A. Principles of Nuclear Magnetic Resonance in One and Two Dimensions; Oxford University Press, USA, 1990. 82. Coffman, R. E.; Buettner, G.R. J. Phys. Chem. 1979, 83 2392-2400. 83. Baker, H. Proc. Lond. Math. Soc. 1902, 1 347. 84. Campbell, J. Proc. Lond. Math. Soc. 1897, 28 381.

PAGE 226

226 85. Milov, A. D.; Salikhov, K. M.; Shirov, M. D. Sov. Phys. Solid State 1981, 23 565. 86. Kurshev, V. V.; Raitsimring, A. M.; Tsvetkov, Y. D. J. Magn. Reson. 1989, 81 441. 87. Pannier, M.; Veit, S.; Godt, A.; Jeschke, G.; Spiess, H. W. J. Magn. Reson. 2000, 142 331-40. 88. Saxena, S.; Freed, J. H. Chem. Phys. Lett. 1996, 251 102-110. 89. Borbat, P. P.; Freed, J. H. Chem. Phys. Lett. 1999, 313 145-154. 90. Schweiger, A.; Jeschke, G. Principles of pulse el ectron paramagnetic resonance ; Oxford University Press, London, 2001. 91. Bloembergen, N.; Purcell, E. M.; Pound, R. V. Phys. Rev. 1948, 73, 679. 92. Fajer, P. G.; Brown, L.; Song, L. Biological Magnetic Resonance, 2006, 1 95. 93. Budil, D. E.; Lee, S.; Saxena, S.; Freed, J. H. J. Magn. Res. A 1996, 120 155-189. 94. Schneider, D. J.; Freed, J. H. Spin Labeling: Theory and Applica tions, Vol. III, Biological Magnetic Resonance; 1989, 8 1-76. 95. Tikhonov, A. N. Dokl. Akad. Nauk USSR 1943, 39 195-198. 96. Chiang, Y. W.; Borbat, P. P.; Freed, J. H. J. Magn. Reson. 2005, 172 279-95. 97. Hansen, P. C. Rank-deficient and discre te ill-posed problems; SIAM, USA, 1998. 98. Santos, E. T. F.; Bassrei, A. Computers & Geosciences 2007, 33 618-629. 99. Prabu-Jeyabalan, M.; Naliv aika, E.; Schiffer, C. A. J. Mol. Biol. 2000, 301 1207-20. 100. Ishima, R.; Torchia, D. A.; Louis, J. M. J. Biol. Chem. 2007, 282 17190-9. 101. Katoh, E.; Louis, J. M.; Yamazaki, T.; Gr onenborn, A. M.; Torchia, D. A.; Ishima, R. Protein Sci. 2003, 12 1376-85. 102. Prabu-Jeyabalan, M.; Nalivaika, E. A.; King, N. M.; Schiffer, C. A. J. Virol. 2004, 78 12446-54. 103. Prabu-Jeyabalan, M.; Naliv aika, E.; Schiffer, C. A. Structure. 2002, 10 369-81. 104. Prabu-Jeyabalan, M.; Nalivaika, E. A.; King, N. M.; Schiffer, C. A. J. Virol. 2003, 77 1306-15.

PAGE 227

227 105. Stark, G. R. Biochemistry 1965, 4 1030-6. 106. Stark, G. R. Biochemistry 1965, 4 588-95. 107. Stark, G. R. Biochemistry 1965, 4 2363-7. 108. Lippincott, J.; Apostol, I. Anal. Biochem. 1999, 267 57-64. 109. Phylip, L. H.; Richards, A. D.; Kay, J.; K ovalinka, J.; Strop, P.; Blaha, I.; Velek, J.; Kostka, V.; Ritchie, A. J.; Broadhurst, A. V.; et al. Biochem. Biophys. Res. Commun. 1990, 171 439-44. 110. Shao, W.; Everitt, L.; Manchester, M.; Loeb D. D.; Hutchison, C. A.; Swanstrom, R. Proc. Natl. Acad. Sci. U S A 1997, 94 2243-8. 111. Muzammil, S.; Ross, P.; Freire, E. Biochemistry 2003, 42 631-8. 112. Piana, S.; Carloni, P.; Rothlisberger, U. Protein Sci. 2002, 11 2393-402. 113. Weber, I. T.; Harrison, R. W. Protein. Eng. 1999, 12 469-74. 114. Bandyopadhyay, P.; Meher, B. R. Chem. Biol. Drug. Des. 2006, 67 155-61. 115. de Mendoza, C.; Valer, L.; Bacheler, L.; Pattery, T.; Corral, A.; Soriano, V. AIDS 2006, 20 1071-4. 116. Kagan, R. M.; Shenderovich, M. D.; Heseltine, P. N.; Ramnarayan, K. Protein Sci 2005, 14 1870-8. 117. Liu, F.; Boross, P. I.; Wang, Y. F.; Tozser, J.; Louis, J. M.; Harrison, R. W.; Weber, I. T. J. Mol .Biol. 2005, 354 789-800. 118. Vergani, B.; Cicero, M. L.; Vigano, O.; Si rianni, F.; Ferramosca, S.; Vitiello, P.; Di Vincenzo, P.; Pasquale, M. P.; Galli, M.; Rusconi, S. J. Clin. Virol. 2007, 41, 154-159 119. Perryman, A. L.; Lin, J. H.; McCammon, J. A. Protein Sci. 2004, 13 1108-23. 120. Clemente, J. C.; Moose, R. E.; Hemrajani, R.; Whitford, L. R.; Govindasamy, L.; Reutzel, R.; McKenna, R.; Agbandje-McKenna, M.; Goodenow, M. M.; Dunn, B. M. Biochemistry 2004, 43 12141-51. 121. Jimenez, J. L.; Resino, S.; Martinez-Colom A.; Bellon, J. M.; Munoz-Fernandez, M. A. J Antimicrob. Chemother. 2005, 56, 1081-1086. 122. Murphy, M. D.; Marousek, G. I.; Chou, S. J. Clin. Virol. 2004, 30 62-67.

PAGE 228

228 123. Santoro, M. M.; Svicher, V.; Gori, C.; Zacca relli, M.; Tozzi, V.; Forb ici, F.; D'Arrigo, R.; Trotta, M. P.; Bellocchi, M. C.; Visco-Comandini U.; Cenci, A.; Bertoli, A.; Narciso, P.; Antinori, A.; Perno, C. F.; Ceccherini-Silberstein, F. New Microbiol. 2006, 29 89-100. 124. Langen, R.; Oh, K. J.; Cascio, D.; Hubbell, W. L. Biochemistry 2000, 39 8396-405. 125. Van Wart, H. E.; Scheraga, H. A. Proc. Natl. Acad. Sci. U S A 1977, 74 13-17. 126. Becke, A. D. J. Chem. Phys. 1993, 98 5648-5652. 127. Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989, 157 200-206. 128. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37 785. 129. Kelly, S. M.; Jess, T. J.; Price, N. C. Biochim. Biophys. Acta. 2005, 1751 119-39. 130. Xie, D.; Gulnik, S.; Gustchina, E.; Yu, B.; Shao, W.; Qoronfleh, W.; Nathan, A.; Erickson, J. W. Protein. Sci. 1999, 8 1702-7. 131. Frishman, D.; Argos, P. Proteins 1995, 23 566-79. 132. Carter, P.; Andersen, C. A.; Rost, B. Nucleic Acids Res. 2003, 31 3293-5. 133. Heinig, M.; Frishman, D. Nucleic Acids. Res. 2004, 32 W500-2. 134. Martin, J.; Letellier, G.; Marin, A.; Taly J. F.; de Brevern, A. G.; Gibrat, J. F. BMC Struct. Biol. 2005, 5 17. 135. Unneberg, P.; Merelo, J. J.; Chacon, P.; Moran, F. Proteins 2001, 42 460-70. 136. Sreerama, N.; Woody, R. W. Anal. Biochem. 2000, 287 252-60. 137. Tomasselli, A. G.; Mildner, A. M.; Rothrock D. J.; Sarcich, J. L.; Lull, J.; Leone, J.; Heinrikson, R. L. Adv. Exp. Med. Biol. 1995, 362 387-98. 138. Szeltner, Z.; Polgar, L. J. Biol. Chem. 1996, 271 5458-63. 139. Fanucci, G. E.; Lee, J. Y.; Cafiso, D. S. Biochemistry 2003, 42 13106-12. 140. Davis, D. J., Burlak, C., Money, N.P Mycol. Res. 2000, 7 800-804. 141. Eaton, G. R.; Eaton, S. S. Acct. Chem. Res. 1983, 21 107-113. 142. Rabenstein, M. D.; Shin, Y. K. Proc. Natl. Acad. Sci. U.S. 1995, 92 8239-3243. 143. Galiano, L.; Bonora, M.; Fanucci, G. E. J. Am. Chem. Soc. 2007, 129 11004-5.

PAGE 229

229 144. Spinelli, S.; Liu, Q. Z.; Alzari, P. M.; Hirel, P. H.; Poljak, R. J. Biochimie 1991, 73 1391-6. 145. Ding, F.; Layten, M.; Simmerling, C. J. Am. Chem. Soc. 2008, 130 7184 146. Martin, P.; Vickrey, J. F. ; Proteasa, G.; Jimenez, Y. L.; Wawrzak, Z.; Winters, M. A.; Merigan, T. C.; Kovari, L. C. Structure 2005, 13 1887-95. 147. Logsdon, B. C.; Vickrey, J. F.; Martin, P.; Proteasa, G.; Koepke, J. I.; Terlecky, S. R.; Wawrzak, Z.; Winters, M. A.; Merigan, T. C.; Kovari, L. C. J. Virol. 2004, 78 3123-32. 148. Molla, A.; Korneyeva, M.; Gao, Q.; Vasavanonda, S.; Schipper, P. J.; Mo, H. M.; Markowitz, M.; Chernyavskiy, T.; Niu, P.; Lyons, N.; Hsu, A.; Granneman, G. R.; Ho, D. D.; Boucher, C. A.; Leonard, J. M.; Norbeck, D. W.; Kempf, D. J. Nat. Med. 1996, 2 760-6. 149. Vickrey, J. F.; Logsdon, B. C.; Proteasa, G. ; Palmer, S.; Winters, M. A.; Merigan, T. C.; Kovari, L. C. Protein. Expr. Purif 2003, 28 165-72. 150. Palmer, S.; Shafer, R. W.; Merigan, T. C. AIDS 1999, 13 661-7. 151. Layten, M.; Hornak, V.; Simmerling, C. J. Am. Chem. Soc. 2006, 128 13360-1.

PAGE 230

230 BIOGRAPHICAL SKETCH Luis Galiano Lafuente was born in 1979 in Ma drid, Spain. He obtained his bachelors degree in chemistry from the Universidad Aut noma de Madrid in 2002. After a short-term scholarship at the University of Florida under Rodney J. Bartletts supervision in 2002, he returned to the University of Florida in 2003 as a Ph.D. student. He obtained his M.S. in computational chemistry from Spain at the Un iversidad Autnoma de Madrid in 2004. After almost two years in Prof. Bartletts group, he joined the newly formed group of Dr. Gail E. Fanucci at the end of 2004, and continued his Ph.D. track under superv ision of Dr. Fanucci.