1 THE CONTRIBUTION OF THE NATURALLY OCCURRING POLYMORPHISMS IN ALTERING THE BIOCHEMICAL AND STRUCT URAL CHARACTERISTICS OF HIV-1 SUBTYPE C PROTEASE By ROXANA MARIA COMAN 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 2007
2 2007 Roxana Maria Coman
3 To my mother, Victoria Coman, and my father, Florin Coman, for their unwavering love and support
4 ACKNOWLEDGMENTS I thank my mentor, Dr. Ben M. Dunn, for allo wing me to pursue my doctorate training under his guidance. In his lab I have learned how to design a project, how to plan all the experimental steps, how to thi nk like scientist. I have been gi ven the opportunity to present my work at several national and inte rnational meetings, to interact and to exchange ideas with scientists from around the world. I can wholeheartedly say that the training in his lab eased my transition from a graduate studen t to a postdoctoral fellow to a professor managing my own lab. I would like to express my appreciation to the me mbers of my supervisory committee: Dr. Linda Bloom, Dr. Brian Cain, Dr. Richard Condit, and Dr. Robert McKenna for their patience and for their willingness to help and guide me through my graduate career. I would like to extend special thanks to Dr. Maureen G oodenow for being a true adviser and a true friend who always was available to answer my questions or give me guidance. I want to express my thanks to all my frie nds I have made while pursuing my doctorate, especially my colleagues, both graduate and undergraduate students in the Dunn lab and the McKennas lab. I thank my parents for their never-ending love and unconditional support. They have been behind me every step of the way, encouraging me to follow my aspirations. I would also like to thank my sister, Anca Cristina, for being a friend and for taking pride in all my accomplishments. I would like to thank my grandpare nts, Veta and Gheorghe Albulescu for their love and support. Most of all, I thank my husband, Altin Gjymis hka, for his endless love and friendship. He has always encouraged me to pursue my dreams. He has been my inspiration to accomplish what I have accomplished today and, without him, I could not have made it through it as well as I have.
5 Finally, my compassion goes out to all who are living with HIV/AIDS and fight it with so much courage and dignity. I extend my sincere th anks and admiration to all involved in fighting this disease and in bringing hope to the m illions afflicted by HIV/AIDS around the world.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............13 CHAPTER 1 HUMAN IMMUNODEFICIENCY VIRUS TYPE 1.................................................................15 Introduction on World HIV Epidemic Status.........................................................................15 HIV-1 Genome and Structure.................................................................................................15 Viral Life and Replication Cycles..........................................................................................16 Antiretroviral Therapy and Resistance...................................................................................19 Reverse Transcriptase Inhibitors (RTIs).........................................................................21 Protease Inhibitors (PIs)..................................................................................................22 Entry and Fusion Inhibitors.............................................................................................23 Highly Active Antiretroviral Therapy (HAART)............................................................23 HIV-1 Diversity................................................................................................................ ......24 Groups and Subtypes.......................................................................................................24 HIV-1 non-B Subtypes....................................................................................................25 HIV-1 Subtype C.............................................................................................................28 2 HUMAN IMMUNODEFICIENCY VI RUS TYPE 1 PROTEASE...........................................40 HIV-1 Protease Structure....................................................................................................... .40 Function and Substrate Specificity.........................................................................................41 Gag/pol Processing............................................................................................................. ....42 Matrix (MA)....................................................................................................................44 Capsid (CA).................................................................................................................... .45 p2............................................................................................................................. ........46 Nucleocapsid (NC)..........................................................................................................46 p1............................................................................................................................. ........46 p6GAG...............................................................................................................................47 Protease Inhibitors and Drug Resistance................................................................................47 Drug Resistance...............................................................................................................49 Active site mutations................................................................................................51 Non-active site mutations.........................................................................................51 Naturally Occurring Polymor phisms in HIV-1 Protease........................................................52
7 3 MATERIALS AND METHODS................................................................................................63 Sub-Cloning of HIV-1 Protease..............................................................................................63 Directional Cloning Di gestion and Ligation........................................................................64 Mutagenesis.................................................................................................................... ........65 Transformation................................................................................................................. ......66 Protein Expression and In clusion Body Extraction................................................................66 Protein Refolding.............................................................................................................. ......67 Ammonium Sulfate Precipitation...........................................................................................68 Size Exclusion Chromatography..........................................................................................68 Protease Kinetic Studies....................................................................................................... ..69 Determination of Michelis-Menten Constants................................................................69 Dissociation Constant (Ki) Determination......................................................................70 Enzyme Active Site Titration..........................................................................................70 Crystallization Studies........................................................................................................ ....71 Protein Sample Preparation.............................................................................................71 Crystal Preparation..........................................................................................................72 Data Collection and Processing.......................................................................................72 In Vitro Gag/pol Processing....................................................................................................73 Amplification and Cloning of the Gag/Pol DNA Sequence into the TNT Vector.........73 Site-Directed Mutagenesis to In troduce the Desired Mutations......................................74 In vitro Transcription-Tran slation Experiment...............................................................75 Separation of the Translation Products on an SDS-PAGE..............................................75 Autoradiography and Densitometric Analysis................................................................75 4 KINETIC STUDIES ON THE CONTRI BUTION OF NATURALLY OCCURRING POLYMORPHISMS IN ALTERING THE BI OCHEMICAL CHARACTERISTICS OF HIV-1 SUBTYPE C PROTEASE..........................................................................................85 Introduction................................................................................................................... ..........85 Results........................................................................................................................ .............87 Analysis of HIV-1 Subtype C Protease Sequence...........................................................87 Kinetic Analysis of HIV-1 Subtype B and C Proteases..................................................88 Kinetic Analysis of HIV1 Subtype B and C Proteases Harboring Drug-Resistant Mutations.....................................................................................................................89 Kinetic analysis of the single muta nts of subtype B and C proteases......................90 Kinetic analysis of the double muta nts of subtype B and C proteases.....................91 Kinetic analysis of the triple mu tants of subtype B and C proteases.......................92 Discussion..................................................................................................................... ..........92 5 STRUCTURAL ANALYSIS OF HIV-1 SUBTYPE C PROTEASE......................................113 Introduction................................................................................................................... ........113 Results........................................................................................................................ ...........114 Crystallization................................................................................................................114 Unbound HIV-1 subtype C protease......................................................................115 Drug-bound HIV-1 subtype C protease.................................................................115
8 Diffraction Data Collection, Processing and Scaling....................................................115 Unbound HIV-1 subtype C protease......................................................................116 Drug-bound HIV-1 subtype C protease.................................................................117 Molecular Replacement: Particle Orientation and Position..........................................117 Unbound HIV-1 subtype C protease......................................................................118 Drug-bound HIV-1 subtype C protease.................................................................118 Structure Refinement.....................................................................................................119 Unbound HIV-1 subtype C protease......................................................................120 Drug-bound HIV-1 subtype C protease.................................................................121 Structure Validation.......................................................................................................122 Unbound HIV-1 subtype C protease......................................................................122 Drug-bound HIV-1 subtype C protease.................................................................122 Structure Analysis.........................................................................................................122 Unbound HIV-1 subtype C protease......................................................................123 Drug-bound HIV-1 subtype C protease.................................................................126 Discussion..................................................................................................................... ........128 6 ANALYSIS OF THE PROCESSING EVENTS OF HIV-1 SUBTYPE A, B, AND C GAG POLYPROTEINS.................................................................................................................155 Introduction................................................................................................................... ........155 Results........................................................................................................................ ...........157 Polymorphic Sites within HIV-1 Subt ype A, B, and C Gag Polyproteins....................157 Cleavage Site Analysis..................................................................................................158 In Vitro Processing Studies of HIV-1 Subtype B, A, and C Gag Polyproteins.............159 In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype B protease..............................................................................................160 In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype A protease..............................................................................................161 In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype C protease..............................................................................................162 Analysis of MA/CA Cleavage Site................................................................................162 In Vitro Processing Studies of HIV-1 Subtype A and B Gag Polyproteins 124 and QV Variants...............................................................................................................162 In trans processing of HIV-1 subtype A gag polyprotein S124V and QV variants by HIV-1 subtype B protease.............................................................162 In trans processing of HIV-1 subtype B ga g polyprotein V124L, V124S and QV variants by HIV-1 subtype B protease...................................................163 Discussion..................................................................................................................... ........164 LIST OF REFERENCES.............................................................................................................185 BIOGRAPHICAL SKETCH.......................................................................................................208
9 LIST OF TABLES Table page 1-1 FDA approved individual anti-HIV drugs.........................................................................37 1-2 Summary of class and ag ent-specific side effects of antiretroviral agents........................39 4-1 The Michaelis-Menten constants fo r the HIV-1 subtype B and C proteases...................111 4-2 The Ki values for HIV-1 subtype B and C proteases.......................................................112 5-1 Data collection statistics................................................................................................. .152 5-2 Refinement steps in SHELX for the unbound form of HIV-1 subtype C protease.........153 5-3 Refinement statistics...................................................................................................... ..154
10 LIST OF FIGURES Figure page 1-1 Adults and children estimated to be living with AIDS in 2006.........................................31 1-2 The structure of the mature human immunodeficiency virus............................................32 1-3 The HIV life cycle......................................................................................................... ....33 1-4 Binding events and potential sites of ac tion for various viral-entry inhibitors..................34 1-5 HIV life cycle and drug targets..........................................................................................35 1-6 Genetic epidemiology of HIV............................................................................................36 2-1 HIV-1 protease............................................................................................................. ......56 2-2 Cryo-electron micrographs and schematic representations of HIV-1 particles.................57 2-3 Hydrogen bonding between HIV-1 pr otease and a modeled substrate..............................58 2-4 HIV-1 gag and ga g/pol polyproteins..................................................................................59 2-5 Structure of HI V-1 capsid monomer..................................................................................60 2-6 Structures of nine individual protease i nhibitors, approved for clinical use in HIVinfected patients.............................................................................................................. ...61 2-7 HIV-1 protease inhib itors resistance chart.........................................................................62 3-1 The expression vector pET23a...........................................................................................77 3-2 Cloning the protease gene into the expression vector pET23a..........................................78 3-3 HIV-1 protease expression in BL21 DE 3(Star) pLysS cells and inclusion bodies (IBs) extraction............................................................................................................... ...79 3-4 HIV-1 protease purification through size-exclusion column Superdex 75........................80 3-5 Kinetic constants determination.........................................................................................81 3-6 Inhibitor dissociation constant (Ki) determination.............................................................82 3-7 Enzyme titr ation curve..................................................................................................... ..83 3-8 Cloning of gag/pol sequence into the TNT vector.............................................................84 4-1 Polymorphic sites within the subtype B and C protease sequences.................................105
11 4-2 HIV-1 subtype C protease................................................................................................106 4-3 Kinetic analysis of HIV1 subtype B and C proteases.....................................................107 4-4 Kinetic analysis of the single muta nts of the subtype B and C proteases........................108 4-5 Kinetic analysis of the double mu tants of subtype B and C proteases............................109 4-6 Kinetic analysis of the triple mutants of subtype B and C proteases...............................110 5-1 Optical photographs of HIV1 subtype C protease crystals.............................................135 5-2 Packing diagram for HIV-1 subtype C protease unbound form...................................136 5-3 Packing diagram of HIV-1 s ubtype C protease IDV-bound form................................137 5-4 Ramachandran diagrams..................................................................................................138 5-5 The 2Fo-Fc electron density map of the unbound subtype C protease.............................139 5-6 The C-shaped electron density between the flaps............................................................140 5-7 The normalized mean B values for the main chain atoms of the wild type subtype C and multi-drug resistant subtype B proteases..................................................................141 5-9 Superimposition of the wild type subtype C protease with the wild type and multidrug resistant mutant of subtype B protease....................................................................143 5-10 The comparison between the flaps of subtype C and B proteases...................................144 5-11 The naturally occurring polym orphisms in subtype C protease......................................145 5-12 Superimposition of the IDV-bound subtype C protease with the IDV-bound subtype B protease..................................................................................................................... ....146 5-13 The normalized mean B values for the main chain atoms of the IDV-bound subtype C and B proteases.............................................................................................................147 5-14 IDV in the active site of HIV protease.............................................................................148 5-15 Ligplot analysis.......................................................................................................... ......149 5-16 Electron density map for NFV in the bound subtype C protease crystal structure..........151 6-1 HIV-1 gag processing......................................................................................................174 6-2 Alignment of the gag polyproteins of HIV-1 subtype B, C and A..................................175
12 6-3 Alignment of the cleavage sites within ga g polyproteins of HIV1 subtypes, B, C, and A.......................................................................................................................... ......177 6-4 The in vitro processing without the frames hift and without the addition in trans of the HIV protease..............................................................................................................178 6-5 In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype B protease...........................................................................................................179 6-6 In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype A protease...........................................................................................................180 6-7 In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype C protease...........................................................................................................181 6-8 In trans processing of HIV-1 subtype A gag pol yprotein S124V and QV variants by HIV-1 subtype B protease...........................................................................................182 6-9 In trans processing of HIV-1 subtype B gag polyprotein V124S and QV variants by HIV-1 subtype B protease........................................................................................183 6-10 HIV-1 matrix protein...................................................................................................... .184
13 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 THE CONTRIBUTION OF THE NATURALLY OCCURRING POLYMORPHISMS IN ALTERING THE BIOCHEMICAL AND STRUCT URAL CHARACTERISTICS OF HIV-1 SUBTYPE C PROTEASE By Roxana Maria Coman August, 2007 Chair: Ben M. Dunn Major: Medical SciencesBiochemistry and Molecular Biology The discovery of a very diverse panel of HI V-1 subtypes poses the question of whether the currently used therapeutic regimens designed ag ainst HIV-1 subtype B are as effective against non-B subtypes, including subtype C viruses. In our research, we focus on exploring the ro le of naturally occu rring polymorphisms in subtype C protease. Our goal is to understand ho w these sequence variations affect the overall characteristics of subtype C prot ease and ultimately how the res ponse to treatment is modulated by the preexistence of the baseline polymorphisms. We obtained a subtype C protease clone th rough the NIH-AIDS Reagent and Reference Program and we determined its kinetic parameters We then added several combinations of major drug resistance mutations: D30N, N88D, and L90M. Our data show that the naturally occurring polymorphisms by themselves, do not provide for a greater level of resistance against the clinically used protease inhib itors. However, we found that the baseline polymorphisms appear to have a role in preserving the catalytic efficiency of th e protease upon acquisition of major drug resistance mutations.
14 We were also able to obtain crystals of subtype C protease and to perform structural studies of this enzyme in the unbound form a nd complexed with two protease inhibitors: indinavir and nelfinavir. In our cr ystal structures we observed that the protease inhibitors bind in the same orientation and make the same contacts w ithin the active site as in subtype B protease. The results also showed that the naturally oc curring polymorphisms within subtype C protease might have a role in modulating the structural stability and flexibility of this enzyme. We also conducted in vitro gag processing studies, in which we compared the rates and order of processing among subtypes B, C, a nd A. The experiments were performed by expressing the gag polyprotein in a cell-free system and adding the active HIV protease in trans Our data reveal that th e gag amino acid sequence has a dominant effect on the order and the rate of cleavage. We also observed th at the rate of gag cleavage app ears to be modulated by both the cleavage site amino acid sequence and other determinants within gag polyprotein. It is widely known that clinical ly used protease inhibitors ha ve been designed using kinetic and structural information about HIV protease. The compilation and subsequent analysis of our data provide information about the biochemical and structural characteristics of subtype C protease, and its susceptibility to currently used protease inhibitors. This type of knowledge will hopefully aid in finding new drugs to combat HIV infection worldwide.
15 CHAPTER 1 HUMAN IMMUNODEFICIENCY VIRUS TYPE 1 Introduction on World HIV Epidemic Status Acquired immunodeficiency syndrome (AIDS) continues to spread largely unchecked since its first documentation in 1983 (Gottlieb et al. 1983; Groopman and Gottlieb 1983; Karpas 2004). Even though promising developments have been seen in the last year s in the global effort to control the AIDS epidemic, the number of pe ople living with HIV continues to grow as does the number of deaths due to AIDS. A total of 39.5 million people were living with HIV in 2006, but this number can go as high as 47.1 million, according to the figures released by the Joint United Nations Program on HIV/AIDS (UNAID S) and World Health Organization (WHO) (UNAIDS/WHO 2006) (Figure 1-1). One of the de veloping regions most devastated is SubSaharan Africa. Two thirds (63%) of all adults and children with HIV globally live in subSaharan Africa and almost three quarters (72%) of all adult and child deaths due to AIDS in 2006 occurred there. Overall, sub-Saharan Africa is home to an estimated 25 million adults and children infected with HIV. The most striking increases in the number of HIVpositive people have occurred in East Asia, Eastern Europe and Central Asia, where the number of people living with AIDS in 2006 was 21% higher than in 2004. Another important shift in the trend of 2006 AI DS epidemic is that, globally, and in every region, more adult women (15 years or older) than ever before are living now with HIV (UNAIDS/WHO 2006). HIV-1 Genome and Structure HIV is part of a family or group of viruse s called lentiviruses a nd it is now generally accepted that HIV is a descendant of simian (mon key) immunodeficiency virus (SIV) (Hirsch et
16 al. 1989). A further classification ca tegorizes HIV as a retrovirus that packages two copies of positive-sense RNA strands in its genome. HIV has a diameter of 100-120 nm, its genome is about 9 Kb in length and is flanked by two l ong terminal repeats (LTR) that are involved in integration and regulation of th e viral genome. The genome can be read in three frames and there are several overlaps of vi ral genes in different reading frames, allowing for the encoding of many proteins in a small genome. The viral genes encode for structural (gag and env), enzymatic (pol), accessory (vif, vpr, vpu, nef) and regul atory (rev, tat) protei ns (Saksena 1998). The outer shell of the virus, known as the viral envelope, consists of a lipid bilayer that is acquired as the virus buds from the cell surface. Embedded in the viral envelope is a complex protein known as env, which consists of an outer protruding cap that is a glycoprotein named gp120 and a stem glycoprotein called gp41. Inside the viral envelope is a protein called p17 (matrix) and within this is the viral core or capsi d, which is made of another viral protein p24 (core antigen). The major elements contained with in the viral core, besides the viral RNA, are protein p7 (nucleocapsid) that associates with the RNA molecules, and three enzyme proteins, p51 (reverse transcriptase), p11 (pro tease) and p32 (integrase). So me other regulatory proteins, such as nef, vpr and vif, are also packaged in the virion (Figure 1-2). Viral Life and Replication Cycles The HIV-1 life cycle is a complex multistage process involving interactions between HIV1 proteins and host macromolecules. The Early Phase of the virus life cycle comprises the infection of host cell and inte gration of viral genome, and the Late Phase comprises the regulation of the expression of the viral gene products and the producti on of viral particles (Turner and Summers 1999). Infection begins when an HIV part icle encounters a host cell with a surface receptor called CD4 (Figure 1-3). The cells mainly targeted by HIV are T-Helper lymphocytes, macrophages, and dendritic cells. The virus particle uses gp120 to attach itself to
17 the CD4 receptor and this is sufficient for bindi ng, but co-receptors are ne cessary for viral entry (Dragic et al. 1996; Alkhatib et al 1997; Deng et al. 1997). Several pr oteins have been identified as possible co-receptors but HIV, generally, uses mainly two co-receptors to enter a target cell: either CCR5 or CXCR4, depending on the strain of virus (Doranz et al. 1996; Samson et al. 1996; Clapham and Weiss 1997; Moore 1997). The stra ins of HIV most commonly seen early in HIV disease, known as macrophage-tropic (M-tropi c) viruses, use CD4 and CCR5 for cell entry. The importance of CCR5 has been demonstrated by the identification of polymorphisms within the CCR5 gene which affect transmission and/ or disease progression (Samson et al. 1996). The binding of viral gp120 env protein to the CD 4 receptor causes a structural change in gp120 exposing the binding site for the co-recep tors. Once the co-receptor is bound, further structural rearrangements occur, mostly in gp41 transmembrane protein, which lead to virus entry (Figure 1-4). Once within the cell, the vi rus particle releases its RNA, and the enzyme reverse transcriptase (RT) then makes a DNA co py of the viral RNA. Of importance in the process of retroviral DNA synthesis, for the purpo se of developing adequa te therapies, are the high degree of genome recombination and the hi gh error rate of RT wh ich provides means for a high level of viral genome variab ility. Recombination is facilita ted by the packaging into the virus particle of two copies of viral RNA and the ability of RT to jump from one copy of the genome to another. The high error rate is due to lack of proof reading ability of RT. It is estimated that HIV-1 RT has an average error rate of 3 X 10-5 errors per base per replication cycle (Mansky and Temin 1995). This value is an average of all types of point mutations, but deletions, insertions, and frameshift mutations are also commonly observed. The template viral genomic RNA is degrad ed by RNase H and the new HIV dsDNA then moves into the nucleus of the cell where, with the he lp of the enzyme integrase (IN), it is inserted
18 into the host cells DNA. This is the last step of the early phase of the HIV life cycle. Formation of the pre-integration complex must occur be fore the integration can take place. The preintegration complex carries sequences that interact with the cellular system and signal the transport of the viral proteins and nucleic acids in to the nucleus. Proteins such as matrix, vpr and IN have been postulated to be i nvolved in the integration of th e viral DNA into the host genomic DNA. This process involves not only the viral IN enzyme but also the host repair system. Initially it was believed that the site of in tegration was random. However, there are several reports that sites of HIV integration in th e human genome are not randomly distributed but instead are enriched in active ge nes and regional hotspots (Schr oder et al. 2002). Global analysis of cellular transcripti on indicated that active genes were preferential integration targets, particularly genes that were activated in cells after infection by HIV-1. Regional hotspots for integration were also found, including a 2.4 kb region containing 1% of sites. These data document unexpectedly strong biases in integrat ion site selection and suggest how selective targeting promotes aggressive HIV replication (Schroder et al. 2002). Once located in the genome of the cell, HIV DNA is called a provirus The HIV provirus is replicated by transcription into viral RNA, so me of which become new viral genomic material and some of which is needed to direct the synt hesis of viral polyprotei ns env, gag and gag/pol. After transcription, the viral mRNA, as any othe r cellular RNA, is modified by addition of a polyA tail. Also, in order to pr oduce proteins such as Rev and Tat, the mRNA must be properly spliced. However, genomic viral RNA has to be transported out of the nucleus without further splicing. To circumvent the strong aversion of the cellular machinery for transporting improperly spliced RNA molecules, the retrovir uses code for a constitutive transport element. This nucleotide sequence allows for the transpor t of the un-spliced mRNA, and in the case of
19 HIV, this is aided by the Rev pr otein (Bray et al. 1994) This creates a point of regulation where regulatory genes that are coded as nascent sequences and arise due to RNA splicing are transcribed and exported out of the nucleus for translation first. One of the these genes is rev that encodes the regulatory protein Rev. Once Rev re aches high enough concentrations it promotes the export of the intact mRNA for translation of gag pol and env genes (Pollard and Malim 1998). The final step in virus replic ation cycle is budding and matura tion. The association of viral RNA, gag and gag/pol polyproteins just underneath the cell membrane precedes budding. Aggregation of gag and gag/pol po lyproteins is mediated by both pr otein-protein interactions and protein-RNA interactions. Specifi cally, gag-gag interaction is me diated by nucleocapsid, matrix, and capsid interactions (Borsetti et al. 1998; Burniston et al. 1999; Gross et al. 2000; Ehrlich et al. 2001). The viral genomic RNA is recruited thr ough interactions with the nucleocapsid protein within gag. The accumulation of env, gag and gag/pol prot eins within the lipid rafts in the cell membrane induces changes that promote budding (Scarlata et al. 1998; Murakami and Freed 2000; Nguyen and Hildreth 2000; Liao et al. 2001). Antiretroviral Therapy and Resistance Access to treatment and care has greatly increased in the recent years. Even if the coverage is still low in many HIV/AIDS-stricken regions, the benefits are dramatic. Through the expanded provision of antiretroviral (ARV) treatment, an estimated two million life-years were gained from 2002 in lowand middlein come countries (UNAIDS/WHO 2006). The evolution of ARV therapy is an ongoi ng process aimed at discovering potent and tolerable drug regimens. The first ARV dr ug approved by the US-FDA in 1987 was zidovudine (AZT), a previously known potential anticancer ag ent. Shortly after, rapid progress in the
20 understanding of the structure and life cycle of the virus led to unprec edented development of other drugs targeted to a variety of viral protei ns. The retroviral enzymes RT, IN, and protease (PR) were the obvious targets fo r drug discovery (Figure 1-5). Th e first drugs to be identified were inhibitors of RT (DeVita et al. 1987), which were discovered and developed long before the structure of RT itself was solved (Jacobo-Mo lina et al. 1993). Newe r RT-targeted drugs, nonnucleoside inhibitors (NNRTI), and HIV prot ease inhibitors (PIs) have been developed bearing the enzyme structure in mind. The struct ure-assisted drug design and discovery process utilizes structural biochemical methods, such as protein crystallography, NMR, and computational biochemistry, to guide the synthe sis of potential drugs. This information can, in turn, be used to help explain the basis of their activity and to improve the potency and specificity of new lead compounds. Crystallography plays a partic ularly important role in this process. The past years have seen a virtual explosion of crystallographic studi es aimed at the characterization of the structures of HIV enzymes and of HIV en zyme/inhibitor complexes on an atomic level. Initially, the ARV drugs were given as monoor dualtherapy. Howe ver, this approach often led to treatment failures due to the development of resi stant virus. Patients who were receiving monotherapy or dual thera py initially experienced decrease s in viral loa d, increases in CD4 count, and improvement in quality of life; th en they experienced viral rebound and decrease CD4 counts further into therapy. The concept of highly active antiretroviral therapy, HAART, was introduced once the PIs were developed. The first PI approved in 1995 by the FDA to be administered to HIV-positive patients was saqui navir (SQV). To date, twenty-one individual ARV compounds within four cla sses have been approved for the treatment of HIV-1 infection: seven nucleoside (NRTIs) and one nucleotide anal og reverse transcriptase inhibitors (NtRTI),
21 three nonnucleoside RT inhibitors (NNRTIs), nine protease inhi bitors (PIs), and one fusion inhibitor (Table 1-1). Reverse Transcriptase Inhibitors (RTIs) The mode of action of NRTIs and NtRTIs is e ssentially the same; they are analogues of the naturally occurring deoxynucleotides needed to synthesize the vira l DNA and they compete with the natural deoxynucleotides for incorporation in to the growing viral DNA chain. The difference between these two subclasses of drugs is that NRTI s, in order to be inco rporated into the viral DNA, must be activated in the cell by the additio n of three phosphate groups to their deoxyribose moiety. This phosphorylation step, carried out by cellular kinase enzymes, converts NRTIs into NtRTIs. Taking NtRTIs directly allows conversi on steps to be skipped, causing less toxicity. Unlike the natural deoxynucleotides substrates NRTIs and NtRTIs lack a 3'-hydroxyl group on the deoxyribose moiety. As a result, following incor poration of an NRTI or an NtRTI, the next incoming deoxynucleotide cannot form the next 5'-3' phosphodiester bond needed to extend the DNA chain. Thus, when an NRTI or NtRTI is in corporated, viral DNA synthesis is halted, a process known as chain termination. All NRTIs and NtRTIs are classified as competitive substrate inhibitors. In contrast, NNRTIs have a completely di fferent mode of action. NNRTIs block RT by binding at a different site on the enzyme when compared to NRTIs and NtRTIs. NNRTIs are not incorporated into the viral DNA but instead inhibit the movement of protein domains of RT that are needed to carry out the process of DNA s ynthesis. NNRTIs are therefore classified as noncompetitive inhibitors of RT. The RTIs are also available as combination pills with the advantage of a reduced burden pill: zidovudine/lamivudine (Combivir), zi dovudine/lamivudine/abacavir (Trizivir),
22 lamivudine/abacavir (Epzicom), tenofovir/emtricitabine (Truvada), and efavirenz/tenofovir/emtricitabine (Atripla). The most common side effects of RTIs as a class are digestive problems (diarrhea, nausea, vomiting, abdominal pain) and constitutional problems (fatigue, fever) (Table 1-2). Protease Inhibitors (PIs) PIs work by inhibiting the prot eolytic cleavage of the structural and enzymatic proteins, which prevents the virus from maturing into an infectious virion. The addi tion of PIs to the ARV therapeutic regimens significantly improved the life expectancy of HIV-positive patients. Unfortunately, the older PIs came with multiple scheduling requirements and the need to take 10-16 capsules/day. Ritonavir (RTV) boosting is a relatively new con cept, and today is one of the mainstays of therapy. Boosting reduces the frequency of dosing and the number of required forms/day (Kilby et al. 2000; Veldkamp et al. 2001). The concept of boosting involves pharmacokinetic drug interactions between a PI, RTV as the boosting PI, and inhibition of the cytochrome P450 (CYP) 3A4 enzyme. Pharmacokine tic enhancement or boosting of PI serum levels and prolongation of their half-life can be achieved with subthera peutic doses of RTV. Metabolized mostly thro ugh the 2C19 enzyme of the cytochro me P450, nelfinavir (NFV) is the only PI not markedly boosted by RTV (Kurowski et al. 2002). The adverse effects of PIs are listed in Tabl e 1-2. Important to mention is that atazanavir (ATV), one of the newest and most potent PIs, is more tolerable that other PIs. However, it has some notable drug interactions: (1) its levels are decreased by 25% when combined with tenofovir, (2) it has significant interactions with H2 blockers and prot on pump inhibitors (Le Tiec et al. 2005).
23 Entry and Fusion Inhibitors This new class of ARV targets prevention of the fusion of HIV and CD4 cell, and prevents passing the viral genome to the CD 4 cell. The only FDA approved fusion inhibitor, Enfuvirtide, is a structural analog of HR2 domain of gp41 and binds to the HR1 region preventing the change in conformation that allows the viral entry in to the cell. Enfuvirtide is a synthetic peptide recommended for use in patients who are treatment experienced or who have had multiple treatment failure regimens and in those with ongoi ng viral replication or vi ral load greater than 400 copies/mL (Piacenti 2006). Highly Active Antiretroviral Therapy (HAART) HAART is a combination of three or more drugs from two drug classes and has improved significantly the prognosis of HI V-infected individuals (Colli er 1996; Carpenter et al. 1997). Since the advent of HAART, the following improve ments have been noted: better patient quality of life, increased survival, slowed disease progression, decreased oppor tunistic infections, decreased viral loads, and incr eased CD4 counts. Many studies have shown that HAART is the most powerful and essential combination to treat HIV infection effectively and decrease the likelihood of emergence of resist ant viruses. However, even w ith this complex therapeutic approach, the development of drug resistance poses the greates t challenge for treating HIV infection and the margin of success for achieving and maintaining virus suppression is narrow. The evolution of HIV-1 drug resi stance within an individual de pends on three main variables: different immunologic stat us and response to therapy, HIV-1 ge netic variability that allows for rapid development of drug resistance mu tations, and the adherence to therapy. Adherence to a complex regimen is often a si gnificant barrier to trea tment success. It is known that HAART is expensive, requires the patient to ingest a large number of pills under a
24 complex dosing schedule and specific food requirement s. It is also associated with various and severe side-effects (Schieferstein and Buhk 2005). The occurrence of the drug-induced resistance mutations results from the inability of HAART to totally eradicate viral replication. With the extremely fa st replication rate of HIV, which can reach 1010 viral particles/day, the hi gh error rate of RT that incorporates, on average, one error per ten thousand base s, and the capacity for genomi c recombination, the virus can quickly develop genomic variatio ns that translate into protei n structural changes. These mutations tend to decrease the affinity of the drug for the target enzyme promoting therapeutic failure (Erickson and Burt 1996). As of Ma y 2006, the HIV guidelines recommend resistance testing in patients with acute or chronic HIV inf ection before therapy is started (U.S. Department of Health and Human Services 2006). This may help detect a virus that is resistant to initial treatment regimens, and regimens can be adju sted based on sensitivity to ARV agents. HIV-1 Diversity Groups and Subtypes The large genomic diversity of viral subtype s in different geogra phical regions is the consequence of the high mismatch error ra te of the HIV RT enzyme (between 1x10-4 and 5x10-5) (Mansky and Temin 1995; Wainberg et al. 1996) coupled with the absence of an exonuclease proofreading activity. Other factors that contribute to the rapid pace of genetic diversification include the replicative rate of each viral subtype, the number of mutations arising in each replicative cycle, the viral propensity for genomi c recombination, and viral fitness. In addition, high rates of genomic evolution may result from host, environment, and/or therapeutic selection pressure (Simon et al. 1998). HIV is characterized by significant genetic diversity among distinct types, groups and subtypes, (Brodine et al. 1995; Fleury et al 2003; Kantor and Katzenstein 2004) and this
25 variability has implications in prevention, diag nostic tests, therapy response, and vaccine development (Peeters and Sharp 2000b; Romano et al. 2000; Holguin et al. 2002; Kantor and Katzenstein 2004). This variability of HIV has led to the developm ent of various distinct groups and subtypes of HIV virus (Figure 1-6) (Peeters and Sh arp 2000a; Peeters 2001). HIV-1 group M, which accounts for most infections worldwide (Osmanov et al. 2002), is classified into 9 distinct subtypes (A-D, F-H, J, and K) and many circulating and unique recombinant forms. These subtypes and intersubtype recombinants differ from one another by 25-30% in the env gene (Robertson et al. 2000; Gaschen et al. 2002) and by 10-15% in the pol gene (Shafer et al. 1999; Robertson et al. 2000). HIV-1 non-B Subtypes While HIV-1 subtype B has been the most widely studied, subtypes A, C, and D predominate worldwide. Among the approximately 40 million people living with HIV/AIDS in 2003, more than 80% are infected with HI V-1 non-B subtypes (Qui nn 2002; Kantor and Katzenstein 2003; Kantor and Katzenstein 2004). In the absence of any drug exposure, RT a nd PR sequences from B and non-B HIV-1 are polymorphic among about 40% of the first 240 RT amino acids and 30% of the 99 PR amino acids. Based on the observation of differences be tween sequences from untreated and treated persons, mutations at PR positions 10, 20, 36, 63, 71, 77, 93 are characterized as secondary resistance mutations in subtype B PR (Hirsch et al. 2000; Shafer et al. 2000). These mutations are known to contribute to drug re sistance when present together with certain primary protease mutations, or have been shown to compensate for the decrease in catalytic efficiency caused by primary protease mutations selected by protease inhibitors. Substitutions at many of the abovementioned positions occur at high rates in certain non-subtype B viruses, and are designated as
26 naturally occurring or baseline polymorphisms. Whereas in treatment-nave patients, many of these baseline polymorphisms do not confer resistance to drugs per se among different subtypes, they may facilitate the development of drug resi stance (Spira et al. 2003). Recent studies have shown that non-B isolates were statistically associated with a more rapid progression to resistance after antiretroviral therapy and they had different mutational patterns when compared to those of B isolates (Pillay et al. 2000; Loemba et al. 2002). Th e combined effects of naturally existing polymorphisms and drug resistant mutations might have important consequences on the feasibility of continuing to use current HI V-1 PIs for non-subtype B infections. A study presented in 2006 found that Ugandans infected with subtype D or recombinant strains incorporating subtype D developed AIDS sooner th an those infected with subtype A, and also died sooner. The study's authors suggested that su btype D is more virulent because it is more effective at binding to immune ce lls (Laeyendecker et al. 2006) This result was supported by another study presented in 2007, which found that Kenyan women infected with subtype D had more than twice the risk of d eath over six years compared with those infected with subtype A (Baeten 2007). An earlier study of sex workers in Senegal, published in 1999, found that women infected with subtype C, D or G were more likely to develop AIDS within five years of infection than those infected with s ubtype A (Kanki et al. 1999). It has also been hypothesized that the HIV genetic vari ability might modulate the transmisibility rates and modes of different viral subtypes or CR Fs. It has been observed that certain subtypes/CRFs are predominantly associat ed with specific modes of transmission. In particular, subtype B is spread mostly by homosexual contact and intravenous drug use (essentially via blood), while s ubtype C and CRF A/E tend to fuel heterosexual epidemics (via a mucosal route) (Bhoopat et al. 2001). However, whether there are biological causes for the
27 observed differences in transmission routes remain the subject of debate. Some scientists, such as Dr. Max Essex of Harvard, believe such causes do exist. However, this theory has not been conclusively proven (Pope et al. 1997). More re cent studies have looked for variation within subtypes in rates of mother-to-child transmissi on. One of these found that such transmission is more common with subtype D than subtype A (Yang et al. 2003). Another reached the opposite conclusion (A worse than D), and also found th at subtype C was more often transmitted than subtype D (Blackard et al. 2001). A third study concluded that subt ype C is more transmissible than either D or A (Renjifo et al. 2004). Othe r researchers have found no association between subtype and rates of mother-to-child transmi ssion (Murray et al. 2000; Tapia et al. 2003). The main concern, however, remains how these various subtypes will respond to the current therapeutic strategies. Most current HIV-1 ARV drug re gimens were designed for use against subtype B, and so hypothetically might not be equally effective in Africa or Asia where other strains are more common. At present, ther e is no compelling evidence that subtypes differ in their sensitivity to ARV drugs. However, so me subtypes may occasionally be more likely to develop resistance to certain drugs. In some situ ations, the types of muta tions associated with resistance may vary and the drug resistance th reshold might be lowe r due to preexistent polymorphisms that could act as secondary resi stant mutations, increasing the drug resistance while conserving the catalytic activity and fitness of the protease. Up to date there are no long term studies analysing the speed of acquiring dr ug resistance in HIV1 non-B subtypes versus B subtype. This is an important subject for future research. The effectiveness of HIV-1 treatment is mon itored using viral load tests. It has been demonstrated that some such tests are sensitiv e only to subtype B and can produce a significant underestimate of viral load if us ed to process other strains. Th e latest tests do claim to produce
28 accurate results for most Group M subtypes, though not necessarily for Group O. It is important that health workers and patients are aware of the subtype/CRF th ey are testing for and of the limitations of the test they are applying. Important to mention is that HIV-2, even if it accounts for a lower percentage of HIVinfected indivuduals, has several specific traits. Fo r example, not all of the drugs used to treat HIV-1 infection are as effective against HIV-2. In particular, HIV-2 has a natural resistance to NNRTI antiretroviral drugs and they are ther efore not recommended to treat individuals harboring this HIV type As yet there is no FDA-licensed vi ral load test for HIV-2 and those designed for HIV-1 are not reliable for monitoring the other type. Instead, response to treatment may be monitored by following CD4+ T-cell co unts and indicators of immune system deterioration. More research and clinical experi ence is needed to determine the most effective treatment for HIV-2 HIV-1 Subtype C HIV-1 subtype C is one of th e nine HIV-1 subtypes but not eworthy are the facts that subtype C has the highest global prevalence, currently estimated at 42% of all HIV infections worldwide, and is now the dominant subtype in sub-Saharan Africa (McCormack et al. 2002; Osmanov et al. 2002). Recent studies (Walker et al. 2005) demonstrated that the growth rate of subtype C infections in southe rn Africa shows no evidence of declining. In addition, the subtype C epidemic has also spread to South and Central China, India a nd Brazil (Yu et al. 1998; Soares et al. 2003). The sequence of HIV-1 subtype C viruses shows the same genetic organization as the other HIV-1 subtypes. Characteristics th at are shared only by subtype C viruses include three copies of the NF-kB binding site in the LTR promoter enhancer region, an insertion in the vpu transmembrane domain as well as a truncation in the rev gene of the virus. From a biological
29 point of view, subtype C viruses differ from other subtypes in the choice of receptor used to enter the host. HIV-1 subtype C viruses have pref erential use of CCR5 over CXCR4 (Zhang et al. 2002). Centlivre et al. demonstrated that at the peak of primary in fection, preferential replication of HIV-1 subtype C was supported by the gut-ass ociated lymphoid tissue (GALT), an IL-7 rich microenvironment. This was shown by the correl ation of the RNA viral genotype in blood and stools, compartments directly draining virions from the GALT (Centlivre et al. 2006). These data show that the GALT cytokine network may we ll favor HIV-1 subtypes C replication during primary infection, and this could result in enhanced transmission. Another difference between B and C subtypes is that the main transmission route for subtype C viruses appears to be heterosexual co ntact (Buve et al. 2001), unlike subtype B that was primarily transmitted through homosexual sex and injecting drug use (UNAIDS/WHO 2006) but this could merely be due to the specif ic sexual practices of the different population groups The documentation of these dissimilarities adva nces the hypothesis that differences in response to antiretroviral therapy that was develo ped based on studies made with HIV-1 subtype B might occur as well. Initial reports suggest th at HIV-1 subtype C viruses respond equally well to therapy (Shafer et al. 1999). However, it wa s found that some mutatio ns (e.g. D30N against NFV) appear in very low rates in drug treated subtype C isolates wher eas the same mutations appear more often in subtype B (Grossman et al. 2004). Another example of differences is the RT V106M resistance mutation in HIV-1 subtype C patients treated with efavirenz as compared to a V106A mutation that occurs in HIV-1 subtype B patients treate d with nevirapine (Brenner et al. 2003). With the possibility of complicated mutation patterns ar ising in ARV treated isolates,
30 the impact of such and other signature mutations in HIV-1 subtype C needs to be investigated further to be able to design any salvage therapy as required.
31 Figure 1-1. Adults and children estimated to be living with AIDS in 2006. Total: 39.5 (34.1 47.1) million. Accessed and adapted on June 2007 from www.unaids.org Joint United Nations Programme on HIV/AIDS. Sub-Saharan Africa 24.7 million (21.8-27.7 million)
32 gag/pol Figure 1-2. The structure of the mature human immunodeficiency virus. Accessed and adapted on June 2007 from www.avert .org AVERTing HIV and AIDS.
33 CD4 host cell Figure 1-3. The HIV life cycle. (a) HIV (yellow) attaches to two cell-su rface receptors (the CD4 antigen and a specific chemokine receptor). (b) The virus and cell membrane fuse, and the virion core enters the cell. (c) Th e viral RNA and core proteins are released from the virion core. (d) Th e viral RNA genome is c onverted into double-stranded DNA through an enzyme unique to viruses, reverse transcriptase (red dot). (e) The double-stranded viral DNA moves into the cell nucleus. (f) Using a unique viral enzyme called integrase, the viral DNA is integrated into the cellular DNA. (g) Viral RNA is synthesized by the cellular enzyme RNA polymerase II using integrated viral DNA as a template. Two types of RNA transcripts, shorter spliced RNA (h) and fulllength genomic RNA (j) are produced. (h) S horter spliced RNAs are transported to the cytoplasm and used for the production of several viral proteins that are then modified in the Golgi apparatus of the cell (i). (j) Full-length genomic RNAs are transported to the cytoplasm (k). (l) New vi rion is assembled and then buds off. (m) Mature virus is released. Accessed and adapted on June 2007 from www.hhmi.com HHMI, Immunology and Infectious Diseases.
34 Figure 1-4. Binding events and potential sites of action for various viral-entry inhibitors. The viral-entry process consists of a series of coordinated interactions binding to two different receptors (Panel A) and membra ne fusion (Panel B). The viral envelope glycoproteins are synthesized as a single polyprotein that assembles into a trimer and then is broken down by host protease into surface glycoprotein subunits (gp120) and transmembrane glycoprotein subunits (gp41). Each gp120 monomer is a complex, folded structure, consisting of a series of variable loops formed by disulfide bonds, with noncontiguous segments brought togeth er to form three-dimensional binding sites for the CD4 receptor and a chemokine receptor (either CCR5 or CXCR4). After CD4 binding, each gp120 undergoes a confor mational change exposing the region that will bind to a seven-transmembrane chemokine receptor. Viral isolates have varying affinities for CCR5 or CXCR4 r eceptors. Gp41 contains a repeat domain which folds upon itself, thus bringing the tw o membranes in close proximity resulting in fusion. Reprinted and adapted from Kilby and Eron 2003.
35 NRTI, NtRTI, NNRTI interrupt transcription of viral RNA into viral DNA Entry and Fusion Inhibitors prevent passing the viral g enome CD4 host PIs inhibit HIV protease-mediate cleavage of the structural and enzymatic proteins Figure 1-5. HIV life cycle a nd drug targets. Accessed and adapted on June 2007 from ww.roche.com Roche.
36 Figure 1-6. Genetic epidemiology of HIV. Classification of HIV in types, groups, subtypes, subsubtypes. HIV-1 recombinants are categorized in two classes: circulating recombinant forms (CRFs) and unique recombinant form s (URFs). Reprinted and adapted from Takeb et al., 2004 (Takeb et al. 2004). Recombinant Forms
37 Table 1-1. FDA approved indi vidual anti-HIV drugs. Generic name Alternative name(s) Brand name (s) Comments Reverse Transcriptase Inhibitors Nucleoside Analog Reverse Transcriptase Inhibitors Zidovudine AZT, ZDV, azidothymidine Retrovir First ARV drug approved by the FDA for the treatment of HIV Didanosine ddI Videx Sec ond FDA-approved ARV drug Zalcitabine ddC, dideoxycytidine Hivid Due to lower potency and serious side effects, it is now rarely used for the treatment of HIV Stavudine d4T Zerit Lamivudine 3TC Epivir It is also used for the treatment of chronic hepatitis B. Abacavir ABC Ziagen The most powerful NRTI to treat HIV Emtricitabine FTC Emtriva It is the newest NRTI and is very similar to 3TC Nucleotide Analog Reverse Transcriptase Inhibitors Tenofovir tenofovir disoproxil fumarate, PMPA, TDF Viread It is also tested for treatment of hepatitis B Non-nucleoside Reverse Tr anscriptase Inhibitors Nevirapine NVP Viramune The first NNRTI approved by the FDA Delavirdine DVD Rescriptor It is now rarely used to its lower potency and complex drug interactions Efavirenz EFV Sustiva/Stocrin It is always given in combination with other ARV drugs
38 Table 1-1. Continued Generic name Alternative name(s) Brand name (s) Comments Protease Inhibitors Saquinavir SQV Invirase The fi rst PI approved by the FDA Ritonavir RTV Norvir It is widely used as a booster for other PIs Indinavir IDV Crixivan It re quires very precise dosing schedule Nelfinavir Nelfinavir mesylate, AG1343, NFV Viracept The only PI approved to treat HIV in pregnant women Amprenavir APV Marketed as the prodrug fosamprenavir (Lexiva), Lopinavir ABT-378, LPV Due to its insufficient bioavailability, it is marketed only as a co-formulation with ritonavir (Kaletra) Atazanavir ATV Reyataz The first PI approved for oncedaily dosing Tipranavir TPV Aptivus The first non-peptidic PI FDAapproved for HIV treatment Darunavir DNV Prezista The latest ARV drug on the market Entry and Fusion Inhibitors Enfuvirtide INN Fuzeon Due to its high cost ($25,000/year/patient) and inconvenient dosing schedule, it is used for salvage therapy only
39 Table 1-2. Summary of class and agent-specif ic side effects of antiretroviral agents. Antiretroviral Agent Adverse Effects Reverse Transcriptase Inhibitors Nucleoside analog reverse transcriptase inhibitors Class effects: lactic acidosis, hepatic steatos is, pancreatitis, bone marrow toxicity, rash Zidovudine Severe headache, nausea, hepatotoxicity Didanosine Pancreatitis, peripheral neuropathy Zalcitabine Peripheral neuropathy, stomatitis Stavudine Peripheral neuropat hy, lipodystrophy, pancreatitis, hyperlipidemia Lamivudine Minimal toxicity Abacavir Severe hypersensitivity, nausea, diarrhea Emtricitabine Minimal toxicity, palmar discoloration Nucleotide analog reverse transcriptase inhibitors Tenofovir Headache, diarrhea, naus ea, vomiting, renal insufficiency Non-nucleoside reverse transcriptase inhibitors Class effects: rash, elevated transaminase levels, nausea, abdominal pain, fatigue Nevirapine Steven-Johnson syndrome, hepatitis Delavirdine Efavirenz Insomnia, abnormal dreams, confusion, impaired concentration Protease Inhibitors Class effects: nausea, vomiting, diarrhea, hyperlipidemia (except atazanavir), fat maldistribution, hyperglycemia, possible increased bleeding in patients with hemophilia, elevated transaminase levels Saquinavir Headache Ritonavir Abdominal pain, periph eral and peri-oral parasthesias Indinavir Nephrolithiasis indirect hyperbilirubin emia, metallic taste, alopecia Nelfinavir Severe diarrhea Amprenavir Skin rash Lopinavir Pancreatic toxicity Atazanavir Indirect hyperbilirubinemia, can cause the prolongation of the PR interval on ECG Tipranavir Increased risk of intracranial hemorrhage Darunavir Cold-like symptoms Entry and Fusion Inhibitors Enfuvirtide Skin reactions at the inje ction site, severe allergic reaction, renal toxicity, paralysis,
40 CHAPTER 2 HUMAN IMMUNODEFICIENCY VIRUS TYPE 1 PROTEASE HIV-1 Protease Structure The HIV protease (PR) is a member of the cla ss of aspartic proteases. The classification is based on a number of criteria: (1) homology alig nment of HIV protease w ith cellular aspartic PRs reveals the Asp-Thr-Gly conserved sequence co mmon to all aspartic PR s; (2) several studies have shown that substituting th e aspartic residue in the Asp-Thr-Gly sequence removes all catalytic activity; (3) the HIV PR is inhibited by pepstatin, a natural inhibitor that is selective for aspartic PRs; (4) crystal structures of HIV PR show a homodimer with each monomer providing a copy of the Asp-Thr-Gly sequence Structural studies on HIV-1 PR were initially hampered by the fact that it constituted a minor component of mature virions (Henderson et al. 1988b), which ther efore necessitated the use of recombinant and synthetic technologies to produce the milligram quantities needed for structural investigations. The HIV-1 PR is an obligatory homodimer th at consists of two identical 99-residue subunits (Wlodawer et al. 1989). The dimer is stabilized by a -sheet formed between the Nand Ctermini of each subunit (Weber et al. 1989; Gu stchina and Weber 1990). The active site of the enzyme is located directly above the dimer inte rface and contains two aspartic residues, one Asp residue is provided by each monomer (Navia et al. 1989). The carboxylate groups of Asp25 from both chains are nearly coplanar and show close contacts. The netw ork is quite rigid due to the interaction (called firemans grip) in whic h each side-chain oxygen of the Thr26 accepts a hydrogen bond from the Thr26 main-chain NH of the opposing loop. Thr26 also donates a hydrogen bond to the carbonyl oxygen atom of re sidue 24 on the opposite loop. These aspartic residues are responsible fo r the catalytic function of the enzyme. As oriented in the figure 2-1, at
41 the bottom of the active site are the catalytic aspartic acids and at the top are two -hairpins, one from each monomer, called flaps. The flaps are thought to undergo a large conformational change to open up and allow access to the active s ite (Rick et al. 1998). It is believed that the flaps need to open as much as 20 to allow the substrate to enter the active site (Rick et al. 1998). The crystal structures of HIV-1 PR submitted to the protein data bank reveal a 7 change in the flap orientation in the unbound form of th e enzyme (Navia et al. 1989; Wlodawer et al. 1989). Function and Substrate Specificity Several studies (Kohl et al. 1988; McQuade et al. 1990) confirmed the necessity of an active PR in order to produce mature infectious vi rions (Figure 2-2). The f unction of the PR is to cleave gag and gag/pol polyproteins into th e independent viral st ructure components and enzymes during maturation. This process is ab solutely essential for viral maturation and rendering the PR inactive completely abolishes the viral infectivity. The cleavage of gag and gag/pol follows a sequential order, most likely due to both substrate specificity and structur al accessibility of the cleavag e sites (Wiegers et al. 1998). Unexpectedly, HIV-1 PR has a broad specificity. Studies done using non-viral proteins as substrate showed that the PR ha s a broad preference for cleavage site amino acids (Tomasselli et al. 1993; Tomasselli and Heinrikson 1994). Despite th is broad substrate preference, it has been possible to assign some general f eatures to PR substrate specific ity. The substrate must be at least 7 residues in length, and must be able to bind in an extended conformation (Figure 2-3). Hydrophobic-hydrophobic or aromatic-proline (Pro) residues seem to be preferred at position P1-P1flanking the scissile bond. Hydrophobic or Gl n/Glu residues appear to the most common residues at position P2. Aromatic residues are a lmost never seen at position P3. Small residues are preferred at position P2 (Dunn et al. 1994; Tomasselli and Heinrikson 1994). Also, several
42 studies showed that a residue in one subsite can influence the sp ecificity for the residue found in the adjoining subsites (Dauber et al. 2002). Binding of a substrate or an inhibitor introdu ces substantial conforma tional changes to the enzyme. The overall movement of the subunits can be described as a rota tion of up to about 2 o around a hinge axis located in the subunit -sheet interface. This moti on, which slightly tightens the cavity of the active site, seems to be also accompanied by a large motion of the flap region (Wlodawer et al. 1989). Gag/pol Processing The precursors gag (p55) and gag/pol (p160) are translated from the gag open reading frame and, together with the regulatory protein nef, serve as the subs trate for PR activity. An essential step in the viral life cycle is th e processing of these tw o polyproteins: gag and gag/pol. The gag precursor contains the structural proteins of th e viral core: matrix (MA), capsid (CA) and nucleocapsid (NC) (Henderson et al. 1 990; Oroszlan and Luftig 1990) and as well as regulatory proteins (p1, p2, p6) (P ettit et al. 1994). The gag/po l precursor itself is a fusion between gag and pol polyproteins an d it is translated as a result of a translational frameshift (Jacks et al. 1988) at the NC/p1 junction, occurring with a frequency of about 5%. While the gag domain of gag/pol encodes for a pproximately the same protein as gag precursors, the pol domain of gag/pol contains PR, reverse transcriptase (RT) and integrase (IN) (Figure 2-4) The role of the frameshift mechanism is to maintain the corr ect ratio of structural proteins to enzymatic proteins (20 : 1). The frameshi ft site contains the slippery sequence (UUUUUUA) followed by a stem loop. During translation, the ribosome reach es the stem loop and stalls on the slippery sequence approximately 5% of the time. The ribosome than slips back one nucleotide and translates the pol domain of ga g/pol polyprotein (Jacks et al. 1988; Somogyi et al. 1993). Any change to the frameshifting site reduces the efficiency of fram eshifting, thereby altering the gag
43 to gag/pol ratio. Numerous studies have shown that the correct gag to gag/ pol ratio is necessary for both viral assembly and replication (Karacost as et al. 1993; Hung et al. 1998; Shehu-Xhilaga et al. 2001). The gag polyprotein is synthesized from the viral mRNA on the cytosolic ribosomes. During this step a myristic acid moiety, which increases gags affinity for the cell membrane, is added through a cotranslational modification (Gottlinger et al 1989; Bryant and Ratner 1990). There are conflicting reports as to whether ga g multimerizes prior to or following membrane targeting. However, the process of viral maturati on involves a series of or dered cleavages and the processing of gag and gag/pol precursors is accomplished by the viral PR encoded within the pol domain of gag/pol fusion prot ein, without assistance from cellular PRs. Both the order and kinetics of cleavage as well as the extent of precurs or processing appear to be critical steps in the generation of fully infectious, a ppropriately assembled viral particles (Mervi s et al. 1988; Pettit et al. 1994; Wiegers et al. 1998). It has been show n that the initial cleavage takes place at the p2/NC cleavage site for both gag and gag/pol pol yprotein through trans and cis mechanisms respectively (Pettit et al. 1994; Wiegers et al. 1998). This step releases NC and allows the condensation of the viral core. Cl eavage between MA and CA then allows CA to separate from the membrane, and the cleavage at CA/p2 site re leases the p2 spacer peptide and allows for CA condensation (Wiegers et al. 1998). It is obvious that at least the first cleavage occurs when the PR is embedded within gag/pol fusion protein. Recent studies demonstrate that the location of the embedded PR limits the number of cleavages that can be made (Pettit et al. 2005). In fact, the Pro located at the first position in the PR is responsible for preventing PR from cleaving downstream sites (Pettit et al.
44 2005). Thus, these data raise the possibility that the structure of gag/pol helps to determine the order of the processing by the PR. Several studies demonstrated that the PR, flanked by short sequences that correspond the native sequences in the gag/pol polyprotein, has a decreased c onformational stability (Wondrak and Louis 1996). These results sugg est that the subunit in teractions and, hence the dimer stability of the PR domain within gag/pol polyprotein, diffe r from those of the mature PR. This seems to have an effect on both the enzyma tic activity of the enzyme (Louis et al. 1994; L ouis et al. 1999) and the drug sensitivity of the HIV PR (Pettit et al. 2004). In fact, the mature, free PR is significantly more sensitive to the protease inhi bitor (PI) ritonavir (RTV) than the immature, embedded PR (Pettit et al. 2004). The first cleav age to release PR from gag/pol polyprotein occurs at the N-terminus of the PR (Wondrak and Louis 1996), precedes the cleavage at the Cterminus, is accompanied by a large increase in the enzymatic catalytic activity and seems to lead to the stabilization of the dimer structure (Louis et al. 1999). Overal l, PR release from the polyprotein appears to be a relati vely late event in virion assemb ly, most likely occurring after the virion has been released from the membrane of the host cell. Matrix (MA) MA is located at the N-terminus of gag polypr otein and has two main roles: (1) targeting of gag to the plasma membrane, and (2) incorpora tion of the envelope glyc oproteins into the new virus particles. Targeting of MA, thus of the gag or gag/pol polypr oteins, to the cell membrane is mediated by both a cluster of conserved basic re sidues located near the N-terminal region (Yuan et al. 1993; Zhou et al. 1994) a nd a myristyl group that is ad ded to the N-terminus (Lee and Linial 1994). In fact, Lee et al. showed that e fficient particle formation can occur if the MA domain of HIV-1 gag is substituted by a myristyl ation signal. Evidence su ggests that the myristyl moiety is exposed when MA is still part of th e gag precursor, but is se questered and hidden after
45 gag processing (Spearman et al. 1997; Hermid a-Matsumoto and Resh 1999; Paillart and Gottlinger 1999). This myristyl switch model would account for the association of the gag precursor with the membrane during budding, but, at the same time, allow MA in the mature viral particles to dissociate from the membrane after infection and enable the viral core to enter the cytoplasm of the newly infected cells. MA is necessary for the incorpor ation of the envelope glycoproteins during virus assembly, and a stable interaction between MA and the cytoplasmic domain of the envelope suggests th at gag and envelope interact pr ior to transport to the plasma membrane (Dorfman et al. 1994b; Wyma et al. 2000). Capsid (CA) HIV-1 CA protein is located between MA a nd p2 and consists of two domains separated by a flexible linker region (Figure 2-5) (Gambl e et al. 1996; Gitti et al. 1996). The C-terminal domain is necessary for particle assembly as well as core formation, whereas the N-terminal domain is only required for core formation (Do rfman et al. 1994a; McDermott et al. 1996; Reicin et al. 1996; Borsetti et al. 1998) During virion maturation, CA assembles into a closed conical shell that surrounds the NC/RNA complex. The C-terminal domain of CA mediates dime rization of gag/pol. Evidence suggests that the dimerization domain extends from CA into the spacer peptide p2, forming an -helix, which appears to be necessary for assembly (Gamble et al. 1997). The dimerization process is important in activating the PR and initiating the processi ng events. Studies showed that small molecules targeting the CA/p2 junction can slow PR-mediate d cleavage at this site, thereby delaying virion maturation and reducing the vira l infectivity (Zhou et al. 2004). The major homology region (MHR), which is a stretch of 20 amino acids in the C-terminal domain (Figure 2-5), does not seem to be involved in gag and gag/pol dimeri zation but is required for both viral particle assembly and the correct assembly of th e viral core (Mammano et al. 1994).
46 p2 The p2 domain is located between CA and NC and its location is well conserved among primate lentiviruses, but the sequence and leng th are not well conserved (Henderson et al. 1988a). It has been suggested that p2 plays a role in the regulati on of gag processing. Petit et al. demonstrated that cleavage at the CA/p2 site was accelerated significantly when cleavage at p2/NC site was blocked, suggesti ng that p2 may function to slow th e cleavage rate at CA/p2 site during the processing (Pettit et al. 1994). Further evidence by this group also indicates that p2 is necessary for the correct formation of the core of the virus, as virions produced from viruses with p2 deleted were not infectious and contained an improperly form ed core (Pettit et al. 1994). Nucleocapsid (NC) NC has highly conserved zinc binding motifs wh ose alteration results in decreased efficiency of reverse transcription and initia l integration process, decreased packaging of genomic RNA, as well as decreased viral inf ectivity (Poon et al. 1998; Buckman et al. 2003). Because of the important role NC plays in the vira l life cycle, there were numerous attempts to develop drugs that would remove the zinc atom, thus destabilizing the binding motif. These trials have been proven unsuccessful because these drugs also removed zinc atoms from the cellular proteins. Current studies are focu sing on the identification of co mpounds that bind reversibly and are non-chelating (Steph en et al. 2002). p1 p1 is located between NC and p6. It is 16 am ino acids long in HIV-1 subtype B and has a remarkable high sequence identity between all the HIV-1 subtypes. Research centered on p1 has been limited and as yet no function has been ascribed to this spacer peptide. Because of its location, p1 overlaps with the sli ppery site and the stem loop of the frameshift site.
47 Several studies (Hill et al. 2002; Hill et al. 2005) showed that the two highly conserved Pro residues within p1 are important for viral infec tivity. Also multiple other changes in the amino acid sequence of p1 resulted in altered protein processing, reduc ed viral genomic RNA dimer stability, and abolished viral in fectivity (Hill et al. 2002). p6GAG p6GAG is located at the C-terminus of gag pol yprotein and three func tional domains have been mapped to it. (1) The P(T/S)AP motif located at the N-terminus binds to the host cellular factor TSG101 and is required for efficient viru s release (Gottlinger et al. 1991; Huang et al. 1995; Martin-Serrano et al. 2001). (2) AN LXSLFG motif in the C-terminus binds vpr, and mutations in this region diminish vpr incor poration into virions (Kondo and Gottlinger 1996). (3) LYP motif is situated upstream to LXSLFG, and both are necessary for the binding to the cellular factor AIP1 and modulate virion budding (Strack et al. 2003). AIP1 directly binds to TSG101 and von Schwedler et al. suggested a mode l in which gag forms a ternary complex with both TSG101 and AIP1 during budding (von Schwedle r et al. 2003). In uninfected cell, these factors are involved in a pathway that sorts cellular proteins into vesicle that bud into multivesicular bodies, and the interaction of HI V-1 p6 with these proteins suggests that in infected cells, HIV-1 uses this cellular network to promote budding of new virions. Protease Inhibitors and Drug Resistance As the proteolytic activity of the PR is ab solutely required for th e production of mature, infectious virions, this enzyme was soon identi fied as an attractive target for therapeutic interventions, and the discovery and devel opment of PIs are a great success of modern pharmacology and structural biology. There are several criter ia that have to be taken in c onsideration when designing a PI. The basic design criterion relied on the observation th at HIV-1 PR, unlike other PRs, is able to cleave
48 Tyr-Pro or Phe-Pro sequences in the viral polypr otein. Because the amide bonds of Pro residues are not susceptible to cleavage by mammalian endopeptidases, the desi gn of HIV-1 PR inhibitors based on this criterion was expect ed to bring potential advantages of higher selectivity: the PIs should be able to bind the PR with high affinity but should also be able to avoid binding with high affinity structurally similar human enzymes, su ch as pepsin or cathepsin D, that can lead to various adverse effects. Due to the high polymorphi c nature of HIV PR, the PIs must be able to bind tightly to as many variants as possibly. This will also aid in evading the occurrence of resistance. Because the fight against HIV requires the use of multiple drugs targeted at both HIV and opportunistic diseases, the PIs must be able to function in the presence of other drugs circulating in the blood at high conc entration. If possible, it is prefer able that the PIs be able to access body compartments that tend to function as vi ral reservoirs such as the CNS. Also, the costs of producing the drug should be low e nough to allow open access for those in need of treatment. This is very important because HIV/ AIDS is a chronic disease requiring a lifetime of treatment with many drugs. PIs (Figure 2-6) are small pe ptidic and nonpeptidic molecules that mimic the natural substrate in the gag/pol polyprotein and ther efore compete for active-site binding. Once bound, the inhibitor cannot be cleaved and consequently inactivates the enzyme. Mo st of the inhibitors co-crystallized with HIV PR, including all pep tidomimetic inhibitors, are bound in the enzyme active site in an extended conformation and th e hydrogen bonds are made mostly between the main-chain atoms of both the enzyme and the in hibitor. The hydroxyl group at the non-scissile junction present in inhibitors is positioned between the D25/ D25 carboxyls of the PR, within hydrogen-bonding distance to at least one carbox ylate oxygen of each aspartate. A feature common to almost all complexes of HIV-1 PR is a buried water molecule that bridges the P2 and
49 P1 carboxyl groups of the inhibitor and Ile50 and Ile150 amino groups of the flaps. This water is approximately tetrahedrally coordinated and is completely separated from the bulk solvent. A number of distinct subsites or pockets that accommodate side chai ns of the inhibitors can be identified in HIV PR. Three subsites on each side of the non-scissile bond (S1 S3 and S1 S3) are very well defined, whereas more dist ant subsites are not as clear. The amino acid composition of these pockets define s the specificity and affinity of the enzyme for a particular inhibitor or substrate, as was descri bed previously in this chapter. Drug Resistance PI resistance, defined as the loss of virus su sceptibility to a PI, involves multiple gradual changes that accumulate within the PR sequen ce (Erickson et al. 1999b). This happens during continued viral replication in the presence of PIs that allows mutations to accumulate either at the active site or within the core of the PR. At the molecular level, resistance to PIs pre dominantly takes the form of mutations within the HIV PR molecule that preferentially lower the affinity of PIs with respect to PR substrates, while still maintaining a viable catalytic activ ity. Basically, resistant mutations arise because they provide the virus with an advantage to survive in the presence of the drug. There are numerous ways to assess drug resist ance. There are clinical studies, genotypic and phenotypic assays, and biochemical studies. Biochemical resistance is a measure of the reduction in drug binding affinity to a variant enzyme compared to the wild type enzyme. This measure is reported as a measure of the dissocia tion constant for a partic ular inhibitor with a particular enzyme variant. These are measurements conducted in our studies. It is important to distinguish the differences in classifying resist ance because clinical resi stance can be due to a number of factors, including viral and host genetics, biochemi cal, cellular, and pharmacological drug levels. Though most resistance can be trace d to genomic mutations, there have been
50 instances where resistance has occurred without de tectable mutations within the PR sequences. It is now understood that factors that influence dr ug absorption, distribution, metabolism, and excretion have a strong role in the development of resistance a nd can vary between individuals (Burger et al. 1997; Hugen et al. 1999). A number of resistance mutati ons have been characterized for all clinically used PIs, including tipranavir (TPV), one of the most recen t and most potent PI on the market (Figure 2-7). In fact, over 57 mutations have been observed in at least 27 positions in the 99-residue HIV PR monomer in response to drug selecti on pressure (Hammond et al. 1997). PI resistance mutations can be classified as shown in Figure 2-7, retrieved from HIV Drug Resistance Database. The mutations ar e categorized into five groups: ( I) Major mutations that, with the exception of L90M, are located in the substrate cleft. Each by itself is capable of reducing susceptibility to one or more PIs. With the excep tion of V82I, which does not cause drug resistance, mutations at these positions ar e nonpolymorphic in that they do not occur in PInave persons. ( II) Flap mutations which occur in the PR flap and are second in importance only to the major mutations. Some may cause resistan ce by themselves, but these mutations are more often accessory. They are also nonpolymorphic. ( III) Other nonpolymorphic PI-resistance mutations that usually indicate past PI exposure. Several cause resistance by themselves but more often they are accessory. ( IV) With the exception of L 10F/R, these mutations are polymorphic. They contribute to re sistance only when present in combination with one or more category I-III mutations. ( V) There are recently reported mutations that have been associated with resistance to ATV, TPV, or DRV (TMC114). Some are non-polymorphic and likely to be significant. Others are polymorphic a nd less likely to be significant.
51 PI resistance mutations can also be classified as active site versus non-active mutations according to whether they occur inside or outside the inhibitor binding subsites. Active site mutations The active site mutations are th e first mutations observed in response to therapy and have received a great deal of atten tion. They occur in the substratebinding cleft of the enzyme and usually directly reduce inhibito r binding to PR. Most notable are V82A/F/T/S, I84V, and D30N. Mutations at position 82 accumulate in response to APV, IDV, RTV, and SQV (Condra et al. 1995; Molla et al. 1996; Patick et al 1998). Mutations at residue 84 ar e seen in response to APV, IDV, and SQV therapy. The D30N mutation is ve ry specific, only seen in response to NFV therapy. Other residues commonly mutated during drug therapy ar e found in the flaps. These mutations, unlike the previous mutations discusse d, are more selective to specific PIs. The mutation I47V and I50V are select ed against APV, the mutation G48V is selected in response to SQV, and I50L arises in response to TPV treatment. Even if the active site mutations have vary ing degree of specificity for different PIs, substantial cross-resistance has b een reported for most of them (H irsch et al. 2000; Kantor et al. 2001). In general, mutations conferring significant resistance are absent in untreated populations; this is consistent with the idea that these substitutions are themselves deleterious to virus replication and are maintained only in the presence of the inhibitor. Non-active site mutations It was argued that the main role of the non-active site mutation is to compensate for reduced catalytic efficiency of the enzyme. However, not too long ago, non-active site mutations have been biochemically characterized as havi ng an active role in de creasing inhibitor binding affinity (Olsen et al. 1999; Perno et al. 2002; Muzammil et al. 2003). Most of the non-active site residues, which mutate in response to drug thera py, are buried in the PR co re; and changes in the
52 PR core interaction could, thr ough a cooperative effect and via l ong range interactions, translate to structural changes in the active site. Recen t studies confirm the fact that many of these substitutions that arise in re sponse to drug therapy in subt ype B PR, occur as baseline polymorphisms in non-B subtypes as discussed be low (Tanuri et al. 1999 ; Shafer et al. 2000; Vergne et al. 2000; Pires et al. 2004). One of the non-active site mutations analyzed in our work is L90M. It is considered a major mutation because its occurr ence decreases the binding affinity for almost all clinically used PIs without a significant effect on the catalytic efficien cy of the enzyme. The L90M mutation appears to exert its effect by affecting th e structural conformation of the active site loop containing the catalytic aspartic acids (Prabu-Jeyabalan et al. 2003). Mutations in the flaps at positions 46 and 54 could also affect the interacti ons of the flaps with the PI or change the dynamics of the opening and closing of the flaps (Clemente et al. 2004). It is worth noting that most non-active site mutations, similar to most of the active site mutations are not specific to any single PI. Naturally Occurring Polymorphisms in HIV-1 Protease These mutations are defined as position sp ecific differences from consensus sequence within the subtype being evaluated, occurring in more than 1% of the sequences isolated from drug-nave patients. Usually these polymorphisms ar e expected to appear at sites that are less critical for enzyme activity and where variati on within the protein ma y be driven by immune selection and functional adapta tions (Kantor and Katzenstei n 2003). These polymorphisms are far more frequent in non-B subtype than in subtype B PRs and seem to be located at positions known to be associated with PI resistance in s ubtype B PR (Tanuri et al 1999; Caride et al. 2001; Perez et al. 2001; Holguin et al. 2002; Gordon et al. 2003). Accessory/secondary mutations are defined as residue changes that arise in response to drug therapy. For certain positions,
53 depending on the subtype analyzed, these two te rms can be used interchangeably. Some other residues are known only as baseline polymorphisms, while other residues are considered to be only secondary mutations and othe rs are not categorized in one of the previously mentioned groups. The precise role of baseline polymorphisms in non-B subtypes is not well understood but a number of studies postulate that th ey can have various functions: Increase the catalytic efficiency of HIV protease. Recent in vitro studies performed on subtype A and C PR mutants (Velazquez-Campoy et al. 2001) showed that in the pr esence of PIs such as IDV, SQV, RTV, and NFV, PRs from subtypes A and C consistently perf ormed their catalytic functions better than PRs from subtype B. These results point to a greater biochemical fitness of the subtypes A and C PRs in the presence of the existing inhibitors. Result in development of diverse mutation al pathways during antiretroviral treatment (Dumans et al. 2004; Grossman et al. 2004). D30N is a primary mutation that renders the HIV-1 PR less susceptible to NFV, and does not exhibit cross-resistan ce with other PIs (Patick et al. 1998). Grossman et al. showed that there is a significant difference in the occurrence of the D30N mutation between subtype B and subtype C infected patients (P = 0.03) when treated with NFV (Grossman et al. 2004). The authors conclude that the rates at which these mutational pathways develop differ in subtype C and subtype B infected patients fa iling therapy, possibly due to di fferential impact of baseline polymorphisms. This has important implications because NFV was often the first-line drug of choice for subtype B infected patients, as the frequent emergence of the non-cross-resistant D30N mutant would not bar the us e of alternative drug combinati ons. This strategy has no such advantage over other PI in the treatment of HIV-1 s ubtype C infected patients.
54 Influence the speed of acquiring PI -related resistance mutations. It has been argued that PI resistance-associated accessory substitutions in subtype B that correspond to baseline polymorphisms in non-B subtypes, may not resu lt in a significant decrease in sensitivity to drugs but are associated with an increase in viral fitness (Erickson et al. 1999a; Hirsch et al. 2000). Thus, the appearan ce of a major mutation in a genome already containing accessory mutations could influence th e speed with which highly resistance viruses are selected during thera py (Vergne et al. 2000). Contribute to resistance and/or maintenance of viral fitness once primary resistance mutations occur. Rose et al. (Rose et al. 1996) showed that baseline polymorphisms or secondary mutations such as L10I are necessary to accommodate a co mbination of other three mutations responsible for drug resistance. Also, these naturally occurring polymorphisms could introduce differences in structural stability of the PR that may in fluence binding affinity due to th e required conformational change associated with substrate a nd inhibitor binding (V elazquez-Campoy et al. 2003). This is important because conformationally constrained liga nds, such as the PIs in clinical use, have little capacity to adapt to changes to the target site and they lose affinity significantly, even when confronted with conservative change s (Velazquez-Campoy et al. 2002). Promote a poorer response to therapy. Servais et al. provided evidence that response to triple-drug thera py including a PI is closely related with the overall number of base line polymorphisms. The main effects of the natural polymorphisms are most likely to manifest themselves after the ons et of drug resistant mutations associated with ARV therapy. If this is the case, the amplifi cation of drug resistance
55 effects might have serious conse quences on the long-term viability of PI therapy in non-subtype B patients (Velazquez-Campoy et al. 2003). In our work, we aim to explore the role of naturally occurring polymorphisms in subtype C PR. Our goal is to understand how these sequence variations affect the overall characteristics of subtype C PR and ultimately how the response to treatment is modulated by the preexistence of the baseline polymorphisms. Comb ining kinetic and structural data will give us a broad understanding of the interactions that are bei ng affected by residue changes and how different inhibitors or substrates interact with the sa me enzyme. This study will bring more information about the contribution the baseline polymorphisms have in m odulating the subtype C PR affinity for the substrates and inhibitors. If our study supp orts the theory that th e high rate of natural polymorphisms in the protease gene in subtype C viruses compromises the protease-inhibitor interactions, this might have important conse quences on the feasibility of continuing to use current HIV-1 PIs for subtype C in fections. Furthermore, our data will allow for prediction about how a certain viral variant, depending on the pol ymorphisms preexistent w ithin the PR sequence, will respond to drug therapy and what mutational pa thway is more likely to evolve. This sort of knowledge will provide clinicians with inform ation that will allow for optimization of therapeutic regimens. This will also provide rese archers with clues to designing inhibitors that will retard the evolution of resistant protease vari ants and exhibit less cross-resistance with other classes of PIs.
56 Figure 2-1. HIV-1 protease (PBD code 1HVP). The catalytic as partic residues (D25/D25) are shown in salmon-colored sticks. In the act ive site, the protease inhibitor indinavir (IDV) is shown in gray. Figure render ed with Pymol (DeLano Scientific). flap dimerization region 80s elbow of the flap
57 A B Figure 2-2. Cryo-electron micrograp hs and schematic representations of HIV-1 particles. A) Immature HIV-1 virions containing the gag and gag/pol polyproteins assembled in spherical layer underneath the viral envelope. B) Once activ e, the viral PR cleaves gag and gag/pol into their constitutive components: ma trix (MA), capsid (CA), nucleocapsid (NC) and p6 proteins thus cau sing the formation of mature, infection virion. Characteristic for this stage is the c one-like shape of the viral core. The color code is as follows: blue viral envelope pr oteins, yellow lipid bi layer, black viral proteins, green viral RNA. Reprinted and adapted from Briggs et. al., 2004. Coned-shaped core Spherical layers of gag and gag/pol polyproteins
58 Figure 2-3. Hydrogen bonding between HIV-1 proteas e and a modeled substrate. Substrate and inhibitor residue side ch ains are designated as Pn for those at the N-terminus of the scissile bond and Pn for those at the C-terminus of the scissile bond as defined by Schechter and Berger, 1967. Reprinted and adapted from Wlodawer and Vondrasek, 1998.
59 Figure 2-4. HIV-1 gag and gag/pol polyproteins. The gag precur sor contains the structural proteins of the viral core: matrix (MA), capsid (CA) and nucleocapsid (NC) and as well as regulatory proteins (p1, p2, p6GAG). The gag domain of gag/pol encodes for approximately the same protein as gag pr ecursors and the pol domain of gag/pol contains PR, reverse transcriptase (RT) and integrase (IN). The HIV PR is represented as blue ribbon.
60 NTD CTD FLR MHR Figure 2-5. Structure of HIV1 capsid monomer (PDB code 1E6J). The color code and abbreviations are as follows: green NT D amino-terminal do main, blue CTD carboxyl-terminal domain, red FLR flex ible linker region, yellow MHR major homology region. Figure rendered with Pymol (DeLano Scientific).
61 Figure 2-6. Structures of nine individual protease inhibitors, a pproved for clinical use in HIVinfected patients. Atazanavir Darunavir Tipranavir Lopinavir Nelfinavir Amprenavir Indinavir Ritonavir Saquinavir
62 Figure 2-7. HIV-1 protease inhibito rs resistance chart. Color c ode: dark blue represents high level of resistance, lighter tones matc h lower levels of resistance, (*) hypersusceptibility, (?) unknown. Also see text. Accessed and adapted on June 2007 from http://hivdb.stanford.edu HIVDatabase, Stanford University.
63 CHAPTER 3 MATERIALS AND METHODS Sub-Cloning of HIV-1 Protease The HIV-1 subtype B protease LAI clone (w ild type) was acquired from Dr. Maureen Goodenow, Department of Pathology, University of Florida. The HIV-1 subtype C protease (PR) clone p94IN476.104 and subtype A PR clone p92UG037.1 were obtained through the AIDS Research and Reference Reagent program, Di vision of AIDS, NIAID, from Drs. Rodenburg, Gao, and Hahn (Rodenburg et al. 2001). Th e HIV-1 PR was sub-cloned from the gag/pol gene sequence using primers to either end of the PR gene sequence. The primers were engineered to insert the restriction site NdeI at the 5 end and BamHI at the 3 end of the PR for directional cloning into expression vector pET23a shown in figure 3-1. The PCR reaction was initiated by mixing together gag/pol DNA, forward and reverse prim ers, Taq-polymerase, 10X Taqpolymerase buffer, MgCl2, dNTPs and water to a final volume of 50 l. The PCR reaction protocol consists of 4 steps: (1) the reacti on was started by heating the PCR mixture to 95 C for 1 min; (2) the reaction wa s then cycled 18 times th rough a melting step at 95 C for 30 sec, an annealing step at 45, 51 or 55 C for 30 sec, and an elongation step at 72 C for 1 min; (3) the temperature of the reaction remained to 72 C for 10 min; (4) the temperature of the reaction was dropped to 4 oC until the PCR cycler was turned off. The results were checked on a 1% agarose gel (Figure 3-2A). The PCR product was cloned into TOPO-cloning vector, pCR2.1, using the TA Cloning Kit from Invitrogen. TA-c loning reaction entails combining 1 L PCR product, 1 L T-4 Ligase (4 Weiss Units), 1 L 10X Ligation buffer (Invitrogen), 2 L pCR 2.1 vector (25 ng/ L), and 5 L of H2O with an incubation step at room temperature (25 oC) for 1 h. One microliter of TA-cloning reaction was transf ormed into chemically competent Top10 cells (Invitrogen), using the transformation procedure described below. The Top10 cells harboring the
64 cloned PRgene into the pCR2.1 vector, 750 L of cell suspension, were stored at C in LB media (1 Liter preparation contains 10 gm of Tr yptone, 5 gm of yeast extract, and 10 g NaCl, pH 7) supplemented with 10% glycerol (120 L of 75% glycerol). Directional Cloning Digestion and Ligation The constructs containing the PR gene cl oned in pCR2.1 vector, in Top10 cells from glycerol stocks, were grown over night (16 hours) in 7 mL of LB with 50 g/ml of ampicillin. The cells were harvested using a Beckman GS-1 5R centrifuge with B eckman rotor S4180 at 4800 rpm for 10 minutes. The plasmid containing th e gene of interest was purified from the Top10 cells using a Qiagen Spin Mini-Prep Kit (Qia gen). In brief, the procedure involves the resuspension of pelleted bacterial cell in prechilled (4 oC) P1 buffer (Qiagen) in the presence of RNase A. The lysis solution, solu tion P2, is a mixture of NaOH and SDS (Qiagen). The lysate solution is neutralized and brought to a high salt concentration in one step using solution N3 (Qiagen). The high salt concentr ation causes cellular components, chromosomal DNA and SDS to precipitate leaving the plasmid DNA in the supernatant. The plasmid DNA is bound to a DNA binding column. The DNA on the column is washed with 75% ethanol to remove any salt. The DNA is eluted off the column using 35 50 l of water. The concentration of DNA prep was measured at 260 nm with a UV spectrophotom eter (CaryUV Bio50), using the formula: OD260DNA x 50 ng/ L x dilution factor, where the diluti on factor is the DNA : final solution ratio in microliters. The plasmid was then double digested with the restriction enzymes NdeI and BamHI. Plasmid pET23a was subjected to the sa me procedure. Digestions were done using 3-10 g of DNA, 1 L of NdeI (10 U/ L) and/or 1 L of BamHI (10 U/ L), 1 L of 10X BamHI Buffer, 0.5 L of BSA, and dH2O to total volume of 30 L, 37 oC for 1 h. The bands of the double digested samples at 300 ba se pairs (HIV-1 protease) a nd 3600 base pairs (pET23a) were
65 excised and gel purified using Spin Gel Purifica tion Kit (Qiagen). The in sert gene and plasmid DNA (3:1 or 10:1 weight ratio) were ligated together in a reaction mixture containing 1 L of T4 ligase (4 Weiss units), 1 mM dATP, and 1.5 L ligase buffer (10X Buffer), in a total volume of 15 L. One microliter of ligation product was transformed into Top10 cells. DNA was purified from these cells using the Qiagen Spin Mini -Prep procedure (Qiagen) and it was used to transform (see page 63) expression cells BL21 DE 3(Star) pLysS (Invitrogen). The presence of the insert in the expression vectors was veri fied by double digestion with NdeI and BamHI (Figure 3-2B). Two microgram s of DNA were also sent fo r DNA sequencing. DNA sequencing was performed by the UF-ICBR sequencing core. Mutagenesis Mutations D30N, N88D, and L90M were a dded to HIV-1 subtype B and subtype C proteases using the Quick-Change Site Directed Mutagenesis K it (Stratagene). Mutations are generated using two primers that are comple mentary to both the coding and the non-coding strands surrounding the base or bases to be mu tated. The reactions were initiated by mixing together 50-100 ng of plasmid DNA, 5 L of Pfu polymerase buffer, 100 ng of dNTPs, 125 ng of upper and lower primer, 1 L of Pfu polymerase (2.5 U), and brought to a final volume of 50 L with ddH2O in a 250 L PCR tube. The PCR reaction protocol consists of 4 steps: (1) the reaction was started by h eating the reaction to 95 C for 30 sec; (2) the reaction was then cycled 18 times through a melting step at 95 C for 30 sec, an annealing step at 51 or 55 C for one minute, and an extension step at 68 C for 4 minutes; (3) the temp erature of the reaction was dropped to 4 C and held for 7 minutes; (4) the temper ature of the reaction was held at 4 oC until the the PCR cycler was turned off. To remove the template, 1 L of restriction enzyme DpnI (10
66 U/ L) was added to all PCR reactions at the end of the cycle and placed at 37 C for two hours. 1.5 L of the PCR reaction was used to tran sform competent Top10 cells (Invitrogen). Transformation All transformations were done using chemically competent cells, which require heating to 42oC to uptake the DNA vector. The vector containing the DNA of in terest was mixed with 15 20 L of cell stock in a microcentrifuge tube an d placed on ice for 30 min. The reaction tube was then placed into a 42 C water bath for 45 60 sec. Immediatel y afterwards it is placed on ice for 10 min. One hundred microliters of SOC media (2.0% Tryptone, 0.5% Yeast Extract, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, and 20 mM glucose, pH 7) we re then added to the reaction tube and placed in an inc ubator rotating at 250 rpm at 37 C for 45 60 min. Following incubation 50-100 L of cell culture were spread over plates containing LB-ampicillin (50 g/mL) for Top10 cells and LB-ampicillin and chloramphenicol (34 g/mL) for BL21 Star DE3 pLysS cells. The plates were then incubated overnight in a 37 C incubator to promote colony growth. The next day a colony was randomly picked and grown overnight (16 hr) in 7 mL of LBampicillin (50 g/mL) or LB-ampicillin (50 g/mL) and chloramphenicol (34 g/mL) to make 10% glycerol stocks and stored at -80 C. Protein Expression and Inclusion Body Extraction The recombinant PR variants were purified fr om the over-expression of the insert gene in BL21 Star DE3 pLysS cells pET23a vector. Four 1 L expressions in M9 minimal media (in 1L: 6.8 g Na2HPO4. 3 g KH2PO4, 0.5 g NaCl, 1 gm NH4SO4, 5 g Casamino Acids autoclaved together in 987 mL of H2O, then 1 mL of 0.1 M CaCl2, 2 mL of 1.0 M MgSO4, and 10 mL of 20% glucose are added) were initiated with a 4% inoculation from overn ight cultures of cells grown in LB media supplemented with 50 g/mL of ampicillin and 34 g/mL of
67 chloramphenicol for BL21 Star DE3 pLysS cells. The expression cultures were grown to an OD600 of 0.8 in a 37 C incubator rotating at 250 rpm. Expres sion of the recombinant protein was induced with 1 mM IPTG. Two 1 mL samples we re taken at zero, one, two, and three hours post induction to monitor expression. One milliliter was used to determine the OD600 of the culture. The second milliliter was centrif uged at 14,000x g for 1 minute to harvest the cells. The supernatant was discarded and the cell pellet was resuspended in 200 L of IB buffer (50 mM Tris, pH 7.5, 3 mM DTT, 2 mM EDTA). 30 L of the resuspension and 6 L of LSB (Laemmli sample buffer) were mixed together and boiled for 3 5 minutes to lyse the cells and denature the proteins. 30 L sample for the zero hour and a standardized volume based on the OD600 for each sample at one, two, and three hours were lo aded into a 15% or 18% Tris-tricine SDS PAGE gel (Figure 3-3). After three hours after induc tion with IPTG the ce lls were harvested by centrifugation at 14,300x g for 10 mi n and resuspended in 30 50 mL IB buffer. Cells were lysed using an SLM-Aminco French Pressure cell at a 1000 psi. The inclusion body pellet was retrieved by centrifuging the cell lysate over 10 mL of 27% su crose in 30 mL Corex tubes in a JS-13.1 swinging bucket rotor at 12,700x g for 50 min. The procedure was repeated two times. After the last spin the wet weight of the inclus ion bodies was recorded and a sample was taken to run on a gel shown in Figure 3-3. Protein Refolding The inclusion bodies were dissolved at 1 mg/mL in 8 M urea supplemented with 300 mM -mercaptoethanol, 50 mM CAPS buffer pH 10, and 1 mM EDTA, and stirred at room temperature for 1 h. The urea solution was prepared in 1 L total volume and mixed with 5 g of Dowex ion exchange resin and stirred at room temperature for 1 h. The ion exchange resin was removed using a 2 m filter. The suspension of denatured protein was loaded into Spectra-Por
68 MWCO 6000-8000 dialysis membrane (Spectrum). The suspension was dialyzed against 4 L of sodium phosphate buffer (5 mM Na2HPO4, pH 7.0, 5 mM EDTA, 300 mM NaCl, and 1 mM DTT) in a 4-L graduated cylinder for 2 h at 4 oC. The buffer was then exchanged for fresh buffer dialyzed over night at 4 oC. The protein suspension was never more than 6% of the total volume of dialysis buffer. Ammonium Sulfate Precipitation To concentrate the HIV-1 protease, the post-di alysate was subjected to ammonium sulfate ((NH4)2SO4) precipitation. The pos t-dialysate was sti rred on ice and (NH4)2SO4 was slowly added to 30% saturation and allowed to stir for 1 h. The solution was centrifuged at 14,300x g for 60 min and the supernatant was returned to ice and stirred. The precipitate was dissolved in 5 10 mL of FPLC buffer (50 mM K2PO4, pH 6.5, 2 mM EDTA, 100 mM NaCl, 2 mM DTT, 5% glycerol) and transferred to a 15 mL conical tube. The above pr ocedure was repeated for 50% ammonium sulfate saturation. Th e resuspended precipitates fro m the 30% and 50% ammonium sulfate fractions in the 15 mL conical tubes were centrifuged in a Beckman GS-15R centrifuge with Beckman rotor S4180 at 4800 rpm to remove any precipitate that di d not re-dissolve. The supernatants were transferred to a clean 15 mL conical tube and checked for activity. Size Exclusion Chromatography The ammonium sulfate fractions that contai n activity were loaded onto a Hi-Load 16/60 Superdex 75 (Amersham Pharmacia) gel filtrati on column that was prewashed with 0.1 M NaOH and water and equilibrated with the FPLC buffer. The column was connected to an FPLC system (Amersham Pharmacia) driven by an LCC-500 Plus controller. Two mL frac tions were collected using a Frac 200 (Amersham Pharmacia). Fractio ns showing protein peaks by absorbance at 280 nm were tested for activity (Figure 3-4A) and fr actions containing active enzyme were verified
69 to be HIV-1 PR by SDS-PAGE, which also a llowed us to determin e the presence of any contaminants (Figure 3-4B). Protease Kinetic Studies Determination of Michelis-Menten Constants All HIV-1 subtype B and C variants were a ssayed kinetically in 50 mM sodium phosphate buffer, pH 4.7 containing 150 mM s odium chloride, 2 mM EDTA, at 37 C. Reactions were initiated when 230 L containing enzyme, buffer, and filtered water were added to 20 L of a chromogenic substrate givi ng a total volume of 250 L. Both the enzyme mixture and the substrate were pre-incubated at 37 C for 3 min. Variable chromoge nic substrate concentrations were used ranging from 5 120 M. The substrates were synthesized by the ICBR Protein Chemistry Core using the so lid phase method with an A pplied Biosystems Model 432A automated peptide synthesizer. A stock of the substrate was made at 10 mg/mL in 5% formic acid. The substrate stock was analyzed by the ICBR Protein Chemistry Core to determine an accurate substrate concentration. At least six diff erent substrate concentrations were used to determine the Michaelis-Menten constants (Mic haelis and Menten 1913) The cleavage of the substrate was monitored using a Cary 50 Bio Va rian spectrophotometer equipped with a 18-cell multitransport. Constant temperature of the reaction was maintained at 37 C in quartz cuvetts using a water pump. A plot of the initial rate data for each substrate concentration versus substrate concentration gives the characteristic Michaelis-Menten curve. F it of the curve to the following equation: ) ( *maxS K S V vm
70 using non-linear Marquardt analysis was used to determine the Km and Vmax for the substrate (Figure 3-5) (Marquardt 1963). The kcat values were determined us ing the following equation: t catE V kmax Dissociation Constant (Ki) Determination To assess the effect of various mutations on inhibitor binding the Ki (dissociation constant) values for various clinically used inhi bitors of the HIV-1 PR were determined for each variant. Inhibition is measured as the decrease of the rate of substrate clea vage in the presence of inhibitor. After acquiring the Mi chaelis-Menten curve in the absence of inhibitor the assay is repeated twice in the presence of two different concentrations of i nhibitor. The curves are then simultaneously fit to the following equation: ) 1 ( ) 1 (max i mK I S K V v to determine Ki values of classical competitive inhibitors (Figure 3-6). The following equation was used to determine the Ki values of tight binding inhibitors (Williams and Morrison 1979). t E K S K E Km S K I E K S i K t I t E v vm i t i t t m o* 2 ))) 1 ) (( ( 4 ))) 1 ) (( ( ( )) 1 ) (( (2 Enzyme Active Site Titration The active enzyme concentration in our assays is determined th rough active site titration of the enzyme using a tight binding inhibitor (Ki 1 nM). The assay cons isted of monitoring the rate of substrate cleavage with a single subs trate concentration in the presence of various inhibitor concentrations in a total volume of 250 L. Reactions are initiated when a series (12-18
71 different reactions) of 230 L containing buffer (same as used in Km and Ki determinations), fixed enzyme concentration, zero and increasing concentration of i nhibitor (until the rate of the reaction was inhibited totally) and water is mixed with 20 L of a fixed substrate concentration. Both the enzyme mixture and the su bstrate were pre-incubated at 37 oC for 3 min before mixing. Cleavage of the substrate in the presence of inhibi tor is monitored as described before. The assay is set up to have the condition where [E]>>>Ki for the inhibitor used. A plot of initial rate versus inhibitor concentration gives an exponential deca y curve that when fitted to the equation above gives the active enzyme concentration (Figure 3-7). Crystallization Studies Protein Sample Preparation The clone used to express HIV-1 subtype C PR was ordered through DNA2.0. The DNA composition of this clone was c odon-optimized: modified to harbor codons that are found with high frequency in the E. coli strains. Such cha nges allow for expression of higher quantities of protein, necessary for crystalliz ation studies. However, the amino acid sequence is similar to that used in our kinetic studies except three differe nces: Q7K, L33I, L63I. The HIV-1 subtype C PR was expressed, refolded, purified and c oncentrated as described previously. To set up crystallization trays, the HIV-1 s ubtype C PR was concentrated to 3 mg/mL (140 M) using a 5 kD VivaSpin 15R MWCO 3000 spin concentrator (VivaScience). Then the buffer was exchanged for 30 mM sodium acetate, pH 4.7 containing 2 mM DTT. To obtain the drugbound forms, the HIV-1 subtype C PR was incubated with either indinavir (IDV) or nelfinavir (NFV) (obtained from the NIH Re search and Reference Reagent Resources Program), two of the clinically used PR inhibitors, in a 3-fold mo lar excess, for 60 min at 4 C in 20 mM sodium acetate pH 4.5, 2 mM DDT prior to crystallizat ion. The enzyme-inhibitor mixture was then spun
72 at 14,000 rpm for 30 min to pellet any precipita tion that might have occurred during incubation step. Crystal Preparation Initial crystallization trials were conducted using th e hanging-drop vapor-diffusion method (McPherson 1982), screening variou s conditions from Crystal Scr een 1, Light and Cryo kits (Hampton Research). Two microliters of the unbound or the drug-complexed forms were mixed in 1: 1 ratio with the various reservoir solution on siliconi zed glass circle cover slides (Hampton). The droplet was suspended over 1 mL of the crystallization buffer. Based on the results of the crystallization scre ening, useful X-ray diffraction-qua lity crystals of HIV-1 subtype C PR were obtained by mixing 2 L of enzyme with 2 L of reservoir solution consisting of 30 mM citric acid, pH 5.0 and 1 M sodium chloride. The crystals formed overnight or within several days and grew at their full size in 7 10 days. Data Collection and Processing Data were collected using A MAR CCD 225 de tector at the SER-CAT beam line BM22 at the Advanced Photon Source, Argonne National Laborat ory, with the kind help of the scientists team, especially Dr. Dauter Zbigniew. The crys tal-to-detector distance was 200 mm. The crystals were soaked in 35% glycerol solu tion and flash cooled at 100 K. All diffraction data frames were collected usi ng a 0.5 oscillation angle with an exposure time of 5 sec per frame. The data sets wa s indexed and scaled with DENZO and HKL-2000 software (Otwinowski and Minor 1997). Refinement and Structure Analysis Cross-rotation and translational searches were performed using the program AmoRe (Navaza 2001) as implemented in the CCP4 suite (Collaborative Computational Project 1994), and rigid body refinement using the CNS p ackage (Brunger et al. 1998). Positional and
73 temperature factors refinement steps were pe rformed in CNS up to 1.6 resolution and in SHELX (Sheldrick 1997) when data were above 1. 6 resolution. Electron density maps with 2Fo-Fc and Fo-Fc coefficients were used to guide manual fitting of the PR and the bound inhibitor, when applicable. Interactive manual model building, using th e molecular graphics program O, version 7 (Jones et al 1991) allowed for a further impr ovement of the structures. The quality of the final refined structure was validat ed with the PROCHECK (Laskowski et al. 1993). In Vitro Gag/pol Processing Amplification and Cloning of the Gag/Pol DNA Sequence into the TNT Vector The subtype C gag/pol sequence was amplified using as template the subtype C p94IN476.104 clone ordered from NIH usi ng primers to either ends of the gag/pol gene sequence: forward primer: 5-AGGCTAGA AGGCTCGAGATGGGTGCGA-3, and reverse primer: 5-TGGTGTTGTACTACGCGTTTACTAGTTCTGATCCTC-3 The primers were engineered to in sert the restriction site XhoI at the 5 end and MluI at the 3 end of the sequence for directional cloni ng into the expression vector TNT. The PCR amplification step was performed as described above. The PCR product was resolved on a 1% agarose gel, cut out and purif ied using a QIAquick Gel Extraction Kit (Qiagen) (Fig ure 3-8A). The PCR product was th en double digested with the restriction enzymes XhoI and MluI. The TNT v ector was subjected to the same procedure. Digestions were done using 5-10 g of DNA, 0.3 L of XhoI (20 U/ L) and/or 0.3 L of MluI (10 U/ L), 1 L of NEB 3, 0.3 L of BSA, and ddH2O to total volume of 40 L at 37 oC for 1 h. The vector was further treated with Inte stinal Calf Alkaline Phosphatase (1 U/ L) (Novagen), 10X buffer and ddH2O to total volume of 20 L and incubated at 37 oC for 1.5 hours. The insert gene and plasmid DNA (3:1 weight ratio) were ligated together in a reaction mixture containing
74 1 L of T-4 ligase (4 Weiss units) (Novagen), 1 L ligase buffer (10X Buffer) in a total volume of 10 L, at room temperature, for 2 h. Two micro liters of the ligation product were transformed into Top10 cells, as it has been described prev iously. The DNA extracted from these cells was checked by digestion with KpnI or XbaI and yielded the ex pected digestion products: 3 fragments for KpnI (6200 bp, 900 bp and 200 bp) and 3 fragments for XbaI (5600 bp, 1100 bp, 300 bp) (Figure 3-8B). The presence of an insert in the TNT vect or was also verified by double digestion with XhoI and MluI. Two microgr ams of DNA were sent for DNA sequencing, performed by the UF-ICBR sequencing core. The identity of the gag/pol sequence was verified using T7-promoter and T7-terminator primers as well as primers complementary to the gag/pol DNA sequence. The purified DNA plasmid was concentrated down to 1 g/ L and directly used for in vitro transcription/translation reactions. Site-Directed Mutagenesis to In troduce the Desired Mutations A continuous gag/pol open r eading frame was created by s ite-directed mutagenesis to reproduce exactly the amino acid sequences of th e gag/pol product proteins found in virions. For the same purpose, another frameshift was intr oduced in the HIV-1 s ubtype C gag/pol fusion protein within RT gene. These mutations were engineered using Quick-Change Site Directed Mutagenesis Kit (Stratagene). Mutations ar e generated using two primers that are complementary to both the coding and non-coding strand surrounding the base, or bases, to be mutated. The reactions were initiated by mixing together 50-100 ng of plasmid DNA, 5 L of 10X Pfu polymerase buffer, 800 ng of dNTP s, 250 ng of upper and lower primer, 1 L of Pfu polymerase (2.5 U), and ddH2O to a final volume of 50 L. The PCR amplification cycle was started by heating the reaction to 95 C for 30 sec. The reaction was then cycled 18 times through a melting step at 95 C for 30 sec, an annealing step at 55 C for 1 min, and an extension
75 step at 68 C for 8 min. After the last cycle the te mperature of the reaction was dropped to 4 C. To remove the template, 1 L of restriction enzyme DpnI (10 U/ L) was added to all PCR reactions at the end of the cycle and placed at 37 C for 1 hour. Two microliters of the PCR reaction were used to transform chemi cally competent Top10 cells (Invitrogen). In vitro Transcription-Tran slation Experiment The gag/pol DNA sequences used for in trans transcription-transla tion experiments did not harbor the continuous open reading fram e. Two microliters of DNA sample (1 g/ L) were mixed with a solution containing 40 L TNT T7 Quick Master Mix (contains rabbit reticulocite lysate and a mixture of all the amino acids except methionine), 2 L 35S-Met (1000 Ci/mmol at 10 mCi/mL) and nuclease-free H2O up to 50 L final volume, and the reaction mixture was incubated at 37 oC. After 2 h, 10 50 nM of the activ e HIV-1 PR was added in the reaction mixture and two to five microliter samples were taken at different time points and quenched with 2x Laemmli buffer at 1:1 ratio and heated to 70 oC for 2 min. Separation of the Translatio n Products on an SDS-PAGE The samples were spun down for 30 60 s ec and loaded onto a 10-20% Tris-HCl SDSPAGE gel (BioRad) that was used for separation of the gag/pol products with different MW. The samples were run at 30 mA until the front of th e loading dye reached th e bottom of the gel. Autoradiography and Densitometric Analysis The gels were fixed in a solution containi ng 10% acetic acid and 5% hydrochloric acid for 30 min, and then soaked in 10% glycerol so lution for 5 min. The gels were then placed on a sheet of absorbant filter paper (B iorad), dried, and expos ed to either a XAR-5 film (Kodak) or a phospho-screen, at room temperature, for 12 24 h. The amount of labeled proteins was
76 quantified by scanning the screens using a Molecular Dynamics PhosphorImager, Storm 860 model.
77 A B BB Figure 3-1. The expression vector pET23a. A) The pET23a vector containing the T7 promoter without the lac operator control and the am picillin resistance gene as a selection marker. B) The poly-cloning region for in serting gene of interest. Accessed and adapted on June 2007 from www.invitrogen.com Invitrogen.
78 A B Figure 3-2. Cloning the protease ge ne into the expression vector pET23a. A) HIV-1 subtype C PR gene amplified using a PCR protocol w ith three different annealing temperatures: 45 oC (lane 2), 51 oC (lane 3), and 55 oC (lane 4). The marker in lane 1 is 1 kb DNA ladder (BioRad). B) Double di gestion with NdeI and BamHI of the pET23a vector containing the 300 bp insert (H IV-1 subtype C PR gene). 1 2 3 4 300 bp 300 bp 3600 bp
79 Figure 3-3. HIV-1 protea se expression in BL21 DE3(Star) pLys S cells and inclusion bodies (IBs) extraction. The gel is Tris-glycine 15% polyacrylamide. The lanes are labeled as follows: 1-ladder (Broad, BioRad), 2-empt y, 3-immediately before induction with IPTG, 4-one hour after induction, 5-two hour s after induction, 6three hours after induction, 7-inclusion body sample. 7 kD 20 kD 1 2 3 4 5 6 7 11 kD
80 A B Figure 3-4. HIV-1 protease purification through size-excl usion column Superdex 75 (Amersham). A) Plot of the enzymatic rate for the fractions 37 48. B) The gel is Tris-glycine 18% polyacrylamide. The la nes are labeled as follows: 1-ladder (Precision Plus Protein Kalei doscope, BioRad), 2 9 lanes: fractions 37 to 44 eluted from the column. 10 kD 15 kD HIV protease (11 kD) 1 2 3 4 5 6 7 8 9 0 0.5 1 1.5 2 2.5 373839404142434445464748Fraction #Rate (deltaAU/sec)
81 Figure 3-5. Kinetic constant s determination. Michaelis-M enten curve fit of rate ( mol/min/mg) versus substrate concentration ( M) to determine Km and Vmax. [Substrate] (M) 020406080100Rate (mol/min/mg) 0 5e-6 1e-5 2e-5 2e-5 3e-5 3e-5
82 Figure 3-6. Inhibitor di ssociation constant (Ki) determination. Michaelis-Menten curve fit of rate ( mol/min/mg) versus substrate ( M) with increasing inhibitor concentration [I] in nM ( = 0 nM; = 20 nM; = 40 nM). [Substrate] (M) 01020304050607080Rate (mol/min/mg) 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040
83 A B Figure 3-7. Enzyme titration curve. A) Linear ex trapolation to the x-axis gives [I] = [E]; B) Dixon plot. [Inhibitor] (nM) 020406080 Rate (mmol/min/mg) 0 2e-5 4e-5 6e-5 8e-5 1e-4 [ Inhibitor ] ( nM ) 2 00204060801001/Rate (mol/min/mg) 200 400 600 800 1000 1200 1400
84 A B Figure 3-8. Cloning of gag/pol sequence into the TNT vector. A) Gag/pol sequence amplified by PCR and purified by gel extr action procedure (Qiagen). B) TNT vector containing the gag/pol insert cut with K pnI (lane 1), cut by XbaI (lane 2) and uncut (lane 3). 3000 bp 1 2 3 3000 bp HIV-1 gag/pol
85 CHAPTER 4 KINETIC STUDIES ON THE CONTRI BUTION OF NATURALLY OCCURRING POLYMORPHISMS IN ALTERING THE BIOC HEMICAL CHARACTERISTICS OF HIV-1 SUBTYPE C PROTEASE Introduction There is extensive and growing literature on sequence data from untreated and treated persons infected with HIV-1 subtype B virus (K antor and Katzenstein 200 4; Shafer et al. 2007). This has led to increasingly accurate, though comp lex, interpretations of HIV-1 subtype B drug resistance (Hirsch et al 2000; Vergne et al. 2006; Shafer et al. 2007). Such data are generally not available for non-B subtypes. More information need s to be gathered and analyzed in order to fully understand the important role that differenc es in protease (PR) sequence of non-B subtypes, including subtype C play in the interaction of enzyme with the substrates and the inhibitors (Figure 4-1). In this chapter, we investigate the role that the eight polymorphic re sidues have in altering the enzymatic and structural features of subtype C PR as compared to subtype B PR. These residues are located in th e elbow of the flaps (M36I, and S37A), at the base of the enzyme (H69K, and I93L), and within the PR core (T12S, I15V, L19I, and L89M). Our goal is to analyze whether these baseline polymorphisms engender a decrease in the affinity of the PR for the clinically used inhibitors, or confer an increa sed catalytic activity a nd thus possibly a greater biochemical fitness in the presence of the inhibitors. We also engineered combinations of three major mutations (L90M, D30N, and N88D) that arise mainly in response to nelf inavir (NFV) treatment. NFV is a PI preferably used, especially in developing countries because of relatively lower cost and fewer side effects. Also, its relatively safe biochemical and pharmacodynamic pr ofiles make NFV the first choice to treat HIV-positive pregnant women. Recent studies have shown that the most common primary
86 mutation observed in treated patients infected with HIV-1 subtype C was L90M (Cane et al. 2001), a major mutation that confers multi-PI resist ance especially to saquinavir (SQV), ritonavir (RTV) and NFV (Condra et al. 1996; Hertogs et al. 2000; Kempf et al. 2001). The leucine (Leu) at the position 90 is not located in the active site but the side chai ns lie on either side of the catalytic Asp25, so the catalytic activity may be more sensitive to substitutions at Leu90 (Mahalingam et al. 2001). Its posi tion in the core of the PR might affect the stability of the enzyme as shown by Xie et al. The L90M mutant seems to be more sensitive to urea denaturation than other mutants. Sedimentation equilibrium st udies have also shown that L90M has reduced dimer stability at pH 7.0 relative to the wild-type PR (Xie et al. 1999). D30N is a primary mutation that renders the HIV-1 PR less susceptible to NFV, and does not exhibit cross-resistan ce with other PIs (Patick et al. 1998). The crystal structure of the HIV-1 PR complexed with a peptide containing the wild type sequence of CA/p2 cleavage site from P5 to P5, Lys-Ala-Arg-Val-Leu*Ala-Glu-Ala-Met-Ser shows that the Asp30 residue side chain is involved in hydrogen bonding to the Gl uP2 side chain (Prabu-Jeyabal an et al. 2000). This is the only direct hydrogen bonding interaction seen be tween the enzyme and a side chain of the substrate. The D30N mutation of Asp to an Asn in the PR likely results in a weaker hydrogen bond that destabilizes NFV bi nding (Kolli et al. 2006). Mutati ons at position 88 (N88D and N88S) commonly occur in patients receiving NFV and occasionally in patients receiving indinavir (IDV). By itself, a mutation at this position causes low-level resistance to NFV. However, when associated with D30N or M46I, the D88N mutation causes high-level NFV resistance (Petropoulos et al. 2000; Ziermann et al. 2000). These three mutations are considered major mutations against NFV and three patterns of mutational associations were identified. First, D30N was positively associated with N88D but negatively associated with N88S. Second, D30N
87 and L90M were negatively associated except in th e presence of N88D, which facilitated the cooccurrence of D30N and L90M. Th ird, D30N + N88D + L90M form ed a stable genetic backbone for the accumulation of additional PI resistance mutations. In 16 patients having isolates with more than one combination of mutations at positi ons 30, 88, and 90, all exhibited one of the steps in the following progression: D30N D30N + N88D D30N + N88D + L90M D30N + N88D + L90M + (L33F +/I84V or M46I /L +/I54V) (Mitsuya et al. 2006). In this chapter, we analyze whether the preex istent baseline polymorphisms, by themselves or in combination with drug resistance mutations, differentially alter the bi ochemical features of the subtype C PR compared w ith that of subtype B PR. Our goal is to understand the effects of sec ondary mutations in conjunction with published data and to evaluate mutations /polymorphisms as a predictive m easure for therapy development and effectiveness. We performed kinetic studies for each variant. Michaelis-Menten constants and Ki values provide biochemical information about the differences in the binding affinity of different mutants to the substrates and clinically used inhibitors, and co uld represent a guide for further in vitro and in vivo studies, pinpointing which inhibitor works best for a certain mutation or combination of mutations. Results Analysis of HIV-1 Subtyp e C Protease Sequence The HIV-1 subtype C PR sequence analyzed in our study comes from an HIV-positive patient from India, and the cl one was provided by the NIH Research and Reference Reagent Resources Program (Rodenburg et al. 2001). This clone contains eight amino acid differences (T12S, I15V, L19I, M36I, S37A, H69K, L89M and I93L) from the subtype B PR (LAI sequence) (Figure 4-2A). Two of these polymorphi sms are located in the elbow of the flaps, three are located in the 10s loop, and three are situated within the loops at the base of the PR.
88 All these regions harboring the subtype C PR pol ymorphic differences are located in the outside regions of the protease avoidi ng the active site, the flaps and the dimerization region (Figure 42B). The initial clone contained one amino acid di fference, N88D, considered a major mutation for NFV resistance. In our study, this enzyme is designated HI V-1 subtype C N88D variant. Subtype C PR analyzed here was obtained by back -mutating the aspartate (D) at the position 88 to the wild-type residue, asparagine (N). Kinetic Analysis of HIV-1 Subtype B and C Proteases The kinetic constants were measured for the HIV-1 subtype B and C PRs using a synthetic substrate: Lys-Ala-Arg-Val-Nle-nPhe-Glu-Ala-Nle-NH2. The P1 position was replaced with nitrophenylalanine to permit spectrophotometric monitoring of the substrate cleavage. The P4 position was replaced with norleucine for stability. This peptide resembles the cleavage s ites between the capsid protein and p2 (KARVL/AEAM). This site is well conserved among different HIV-1 subtypes and it is essentially identical for subytpe B and C gag pol yproteins. Previously pu blished reports noted that the sequence of the capsid/p2 cleavage site is one of the best HIV-1 processing sites when assayed as a peptide at low pH (R ichards et al. 1990; Tozser et al. 1991). Several studies showed that CA/p2 site appears to be important in regul ating the formation of the viral core being the rate-limiting step during gag and gag/pol proces sing (Pettit et al. 1994; Wiegers et al. 1998). Michaelis-Menten constants (Km, kcat) As seen in Table 4-1, Km for the subtype C PR is similar to that of the B subtype PR, with values of 19.5 M and 17 M, respectively. However, the situation is not the same with kcat values. Subtype C PR showed a 2-fold decrease in kcat (5.6 sec-1) when compared to subtype B
89 PR (10 sec-1) (Figure 4-3A). Accordingly, the catalytic efficiency, defined as kcat/Km, of the C subtype PR is about 2 times lower than that of the B subtype PR. Inhibition constants (Ki) The Ki values for subtype B and C PRs were calcu lated for eight protea se inhibitors (PIs) used in clinical settings: SQV, RTV, IDV, NFV, amprenavir (A PV), lopinavir (LPV), atazanavir (ATV), tipranavir (TPV). The results obtained for all of the inhibitors are summarized in Table 4-2. There are three PIs: SQV, IDV and NFV that have slightly higher Ki values than the other five inhibitors (Figure 4-3B). Even if the Ki values are comparable between subtype B and subtype C PRs, the naturally occurring polymor phisms within subtype C PR seem to have differential effect for different inhibitors. They increase the Ki value in subtype C protease for IDV by 2-fold when compared to that of subtyp e B, and they have similar effect on NFV binding but to a lesser extent, with only 1.6-fold increase in Ki value. On the other hand, the influence on the change in the Ki values is in the opposite direction for RTV and TPV with a decrease in Ki values for subtype C PR by 2.6and 3.6fold, re spectively, when compared to those of subtype B PR (Figure 4-3B). However, even with the differences described above, the Ki values for these enzymes with all of the inhibitors tested were within sub-nanomolar to low nanomolar range, in agreement with other studies analyzing both s ubtype B and C PRs (Muzammil et al. 2003; Velazquez-Campoy et al. 2003). Kinetic Analysis of HIV-1 Subtype B and C Proteases Harboring Drug-Resistant Mutations The mutations we analyzed in this study were introduced within both subtype B and subtype C PR sequences through site-directed mu tagenesis technique. We engineered single,
90 double and triple mutants, harboring various co mbinations of D30N, N88D, and L90M. There were a total of 14 mutants we constructed and analyzed. Kinetic analysis of the single mu tants of subtype B and C proteases Michaelis-Menten constants (Km, kcat) The addition of D30N mutation increased the Km by to 2and 3-fold for subtype B and subtype C PRs, respectively. However, the effect on the kcat value was divergent with a 2-fold decrease for subtype B PR and 2-fold increase for subtype C PR (Figure 4-4A). However, despite these varied effects on Km and kcat values, the enzymatic catalytic efficiency (kcat/Km) showed similar calculated values for both D 30N subtype B and C mutants: 0.13 and 0.18 sec-1 M-1, respectively (Table 4-1). The addition of the N88D mutation resulted in relatively unchanged Km and kcat values for subtype B PR with a 1.4-fold increase in the catalytic efficiency of the mutant. The HIV-1 subtype C N88D PR variant showed a slight increase in Km and an approximate 2-fold decrease in the enzymatic catalytic efficiency. The variants harboring L90M mutation exhibited similar changes in their kinetic parameters wh en compared with the N88D variants. Inhibition constants (Ki) As expected, the most significant effect of D30N mutation was on NFV binding for both subtype B and C PRs, with a fold-increase of Ki values of 26 and 89, re spectively (Table 4-2). The other PI significantly affected by D30N is ATV with a 12-fold increase of Ki value for subtype C PR and only 3-fold increase for subtype B variant. However, both Ki values remain in subnanomolar range, indicating tight binding inhi bition. HIV-1 subtype C D30N variant also exhibited a 9-fold increase in Ki value against RTV, while the same mutant of subtype B PR showed only 2-fold increase.
91 The subtype B and C N88D mu tants exhibited a 2-fold and 4-fold increase of K i values when tested with NFV (Figure 4-4B). The N88D mu tation is not selected by other clinically used inhibitors but decreases the binding affinity fo r SQV (3-fold) and IDV (6-fold and 13-fold) for subtype B and C PRs, respectively. The L90M mutation alone introduces, in both subtypes, slight increases (between 1.5and 3.5-fold) in K i values for almost all clinically used inhibitors tested in this study, except RTV where ther e is a 6-fold and 11-fold increase in Ki for subtype B and C PRs, respectively. The newer PIs APV, LPV, and TPV seem una ffected by any single mutations introduced within subtype B and C PRs backbone. Kinetic analysis of the double muta nts of subtype B and C proteases Michaelis-Menten constants (Km, kcat) The subtype C PR double mutants D30N /N88D and D30N/L90M exhibited an approximate 2-fold decrease in their catalytic activity. On the other hand, the same double mutants of subtype B PR showed more significant changes, re taining only about 10% of the enzymatic activity of the wild type enzyme. The double mutant N88D/L90M of subtype B PR showed slight changes in the Km and kcat values: 24 M and 13 sec-1, with a catalytic activity similar to that of the wild type subtype B PR. The introduction of the same combination of mutations in HIV-1 subtype B PR engendered a 2-fold and 4-fold increase in Km and kcat values respectively, result ing in doubli ng the variant catalytic activity when compared to the wild type subtype C PR (Figure 4-5A). Inhibition constants (Ki) The addition of D30N and N88D of both subtype B and C PRs resulted, as expected, in a comparable increase in Ki values for IDV, NFV, and SQV. RTV showed 11-fold increase of Ki
92 value for subtype B PR and almost no change for subtype C variant. APV showed a 5and 2.5fold decrease in Ki values for subtype B and C vari ants, respectively (Figure 4-5B). The subtype B PR D30N/L90M double mutant exhibited an increased in Ki values for all PIs tested in this study when compared with the same variant of subtype C PR. The most significant differences were for SQV, I DV, and NFV with an increase of Ki values by 7-, 6-, and 52fold and 3-, 3-, 24fold for subtype B and C PRs, respectively. Upon the acquisition of N88D/L90M, there was noted an increase in Ki values for all inhibitors analyzed except TPV, which showed a decrease in Ki values for both subtypes, thus indicating an increased affinity for the enzy mes harboring the N8D/L90M combination. The highest fold-changes were observe d for SQV, RTV, IDV with a 22-, 10-, 10fold increase for subtype B PR and a 50-, 14-, 28fo ld increase for subtype C PR. Kinetic analysis of the triple mu tants of subtype B and C proteases Michaelis-Menten constants (Km, kcat) Both variants of subtype B and C PR s harboring the triple combination D30N/N88D/L90M showed a 2-fold increase in their catalytic efficiencies (Figure 4-6A). Inhibition constants (Ki) The increase in Ki values is comparable between the triple mutants with the most significant differences for RTV and NFV with 14and 52-fold increase and 31and 83-fold increase for HIV-1 subtype B and C va riants respectively (Figure 4-6B). Discussion The discovery of such a diverse panel of non-B subtypes of HIV-1 has posed the question if the currently used methods of prevention, diag nostic, and treatment are as effective as they are for HIV-1 subtype B, the subtype these methods have been orig inally developed to combat. A large number of laboratories and researchers and numerous resources have been allocated to
93 investigate the matter. One of the most researched topics is the efficacy of the treatment and the development of drug resistance to currently avai lable PIs, which have been designed based on biochemical and structural information on subt ype B PR. The main concern is the that the preexisting naturally occurring polymorphisms within non-B s ubtypes PRs might increase the speed of which highly resistant vi ruses are selected, promoting a poor er response to the clinically used PIs. The aim of this chapter is to analyze the c ontributions to the catalytic efficiency and inhibitor resistance by naturally occurring polymorphisms and th erapy-selected, active and nonactive site, residue changes in subtype C compared to subtype B PRs. As mentioned before, even though the baseline polymorphisms found in subtype C PR have ar isen in the absence of PI therapy, residue changes in C PR at positions 36, 69, and 89 have been associated with PI resistance in vivo and in vitro in B PRs (Gong et al. 2000; Rusconi et al. 2000; Clemente et al. 2003; Holguin et al. 2004). Thus, these polymor phisms can potentially influence substrate processing and the binding of cu rrently available PIs and/or facilitate the development of resistance. In the PR sequence analyzed in this study all naturally occurring polymorphism changes represent conservative amino acid changes, except at position 37 where a hydrophilic amino acid, serine, is replaced by a small hydrophobi c residue, alanine. Two polymorphic residues (M36I and S37A) are located in the flap maki ng contact with the solvent molecules. More precisely, M36I lies in the inte rface between the enzyme core a nd the flap and develops as a secondary drug resistance mutation in subtype B PR, decreasing the enzyme affinity for several clinically-used PIs (Clemente et al. 2003; de Mendoza et al 2007). The S37A is rarely encountered in clinical isolates (Stanford 2007) and has unknown role if any in the development
94 of drug resistance but in our in vitro experiments this mutation seems to decrease the enzyme solubility in polar solvents a nd to promote precipitation. Three other polymorphisms are located in the 10s loop: T12S, I15V, and L19I. This co mbination is highly specific for subtype C PR (Stanford 2007). The role of these polymorphism s was not fully explored but the I15V mutation has been described to arise in association with TPV treatment. However, there is no clear link between the occurrence of this polymorphism and treatment failure to TPV (Rusconi et al. 2000). H69K is located in the loop at the base of th e PR, and together with M36I was documented to predict decreased phenotypic suscep tibility or diminished antiviral responses to TPV (Baxter et al. 2006; de Mendoza et al. 2007). L 89M is located at the base of the PR and Sanches et al. hypothesized that this polymorphisms may lead to early development of drug resistance in patients infected with non-B HIV subtypes (Sanch es et al. 2007). Also, recent studies inferred that the L89M mutation in subtype F viruses is a high genetic barrier to the accumulation of the L90M resistance mutation and can function as a resistance mutation, depending on the presence of other polymorphisms in the subtype F PR backbone (Calazans et al. 2005). The I93L polymorphism is located within a hydrogen-bonded turn im mediately upstream from the PR/RT cleavage site, in close proximity L89M and H69K and the dimerization domain. It seems to play a dual role, depending on the ARV regimen administered to the HIV-positive patients. A higher relative risk for developing tr eatment failure was observed with the presence of I93L in a subgroup of patients treated with I DV, RTV or SQV in severa l studies (Harrigan et al. 1999; Servais et al. 20 01). Also, although the I93L substitution has been described as a naturally occurring polymorphism in subtype B PR as well, the frequency of this substitution appears to be substantially higher in the treated population than in protease inhibitor naive patients (40% vs. 20%) (Kozal et al. 1996). On the other hand, I93L seems to increase the
95 susceptibility of subtype C PR to LPV as obser ved by Gonzalez et al. (G onzalez et al. 2003). In this study, the authors noticed that the subtype C PRs harbori ng I93L polymorphism presented significant hypersusceptibility to LPV. The results of all these studi es, and of many others not me ntioned here, demonstrate the need for further investigating th e role of naturally occurring polymorphisms in the development of drug resistance and their cont ribution in evading drug inhibiti on and in preservation of the catalytic activity of the PR upon accumula tion of major drug resistance mutations. Our kinetic data on the wild t ype subtype B and C PRs showed that there are no significant differences in the substrate affinity consta nt between these two enzymes, both having Km values within 17 19 M range. This lack of a significant effect in Km probably reflects the fact that the sequence variations found in s ubtype C PR are located outside the active site and that the chemical environment surrounding the aspartyl diad remains unchanged. Also, even if the naturally occurring polymorphism, such as L89M and I93L, located in the hydrophobic core of the enzyme would change the shape of the substrat e-binding cleft, the flexibility of the substrate seems to overcome this effect. From our data it is also appa rent that the naturally occurr ing polymorphisms in subtype C PR by themselves do not provide an advantage for substrate catalys is when compared to subtype B PR, with the catalytic enzymatic efficiency of subtype C PR being 60% of that of subtype B PR. These results come in contrast to the da ta reported earlier for subtype C PR. VelazquezCampoy et al. showed that the subtype C and A PRs exhibit catalytic a dvantage over B PR, with the C subtype PR displaying lower Km values against two different substrates and resulting in a higher (2.4-fol d) catalytic efficiency than the subtype B PR (Velazquez-Campoy et al. 2001). These different results can be explained by the differences in the amino acid sequences between
96 the subtype C PRs analyzed in theses studies. Th e enzyme analyzed in our studies harbors five additional polymorphisms: T12S, I15V, L19I, S37A, and I93L, whose over all effects on the biochemical characteristic of the HIV PR have not been studied in detail. The inhibition kinetic studies c onducted with eight clinically used PIs showed that the Ki values for all the inhibitors te sted on subtype B and C wild ty pe PRs are comparable between theses two enzymes and have low nanomolar and sub-nanomolar ranges, i ndicating high affinity of currently used PIs for these enzymes. These data are in agreement with previous clinical studies that report that, at least at the initiati on at ARV treatment, drug na ve patients infected with subtype C viruses respond to ARV treatment as well as HIV-1 subtype B infected patients (Alexander et al. 2002; Frater et al. 2002; Pillay et al 2002; Weidle et al. 2002; Bocket et al. 2005; Wester et al. 2005). However, several studies showed that differences in response to ARV treatment, measured as vi ral load or CD4 cells count, seem to arise after 1-2 years of treatment, with non-B subtypes infected patients performing worse than subtype B infected patients (Caride et al. 2001; Loveday et al. 2001; De Wit et al. 2004; Atlas et al. 2005). But other factors, such as adherence to treatment, ethnicity, psychosocial conditions, should be taken into consideration as well when analyzing these differences. However, up to date there are no long-term clinical studies (more than 5 years) to e xploring the efficacy of the existing PIs in patients infected with HIV-1 non-B subtypes, including su btype C (Frater et al. 2002). Du e to the severity of this disease and its life-long implicat ions, together with the use of antiretroviral therapy becoming more widespread across Africa, it is imperative to characterize baseline mo lecular variability and subtype-specific peculiarities of drug ta rgets in non-subtype B HIV-1 infection. We hypothesize that the effect s of naturally occurring polymorphisms on the enzymatic catalytic activity and inhibitor binding affinity of subtype C PR are augmented by the acquisition
97 of major drug resistance mutations As a first step to explore th is possibility we engineered a series of single mutants containing one P I-induced mutation: D30N, N88D or L90M. For most PIs analyzed in this study, when co mparing the effects of various combinations of mutations between subtype B and C PRs, we observed that the trend in the changes in Ki values are similar, exhibiting either concurrent in creases or decreases against the clinically used PIs analyzed in this study. The difference consis ts in the magnitude of the change that varies between subtype B and C PRs. The mo st significant fold-increase in Ki values was noted for NFV when tested against PRs ha rboring D30N mutation. Subtype B variants showed a 26-fold increase when compared to the wild type, while D30N subtype C PR exhi bited 89-fold increase, and thus a significant decrease in binding affinity towards NFV. The catalytic efficiencies for both variants were similar when compared as ab solute values. However, when compared to the wild type enzyme, D30N subtype B PR preserve d only 25 % of the catalytic efficiency of the wild type, while D30N subtype C PR maintained 50% catalytic efficiency. These results do not bring a biochemical explanation to the previ ous observation that HI V-1 subtype C viruses preferentially select L90M over D30N when e xposed to NFV treatment (Grossman et al. 2004). However, other clinical studies also could not corroborate with Gr ossman et al. findings, as no differences were observed in th e frequency of development of resistance mutations L90M and D30N in B and non-B viruses, including subtype C, upon NFV treatment (Pillay et al. 2002; Doualla-Bell et al. 2006; Stanford 2007). Subtype C PR harboring D30N al so exhibited larger decrease in the susceptibility to PIs, such as RTV and ATV inhibition when compared to similar mutant of subtype B PR. Although this might predict a predispositi on to developing resistance to thes e two inhibitors by the subtype
98 C variants, it does not seem to be a clear marker of resistance, because both PIs maintained a binding strength sufficient for effective inhibition of the PR. Both single mutants, N88D and L90M, of subt ype B PR showed a slig ht decrease in the Km values and a preservation of the enzymatic catal ytic efficiency when compared to the wild type subtype B PR. The similar mutants of the subtype C PR exhibited a slight decrease in substrate affinity and in enzymatic catalytic e fficiency. Both mutations are located on each side of the 89 position occupied by a Leu in subt ype B PR and a Met in subtype C PR. The differences in enzyme activity could be explaine d by slight distortions th at a larger 89M could induce in the subtype C PR core. The same hypoth esis was recently advan ced by Sanches et al. after structural characterization of subtype F PR which also contains the L89M polymorphisms (Sanches et al. 2007). Upon acquisition of N88D, the only significant ch ange in binding affinity was observed for IDV, with a higher increase in Ki values for subtype C PR (13-fold increase in Ki value, compared to 6-fold increase for subtype B PR mutant). Our study demonstrates that N88D by itself does not provide higher level of resistance for NFV in both subtype B and C PRs. L90M mutants exhibited slight decrease in binding affinity for SQV, IDV and RTV, for both subtype B and C PRs. So far, our data on single mutants for s ubtype B and C PRs confirmed some of the previous studies on subtype B PR and highlighted several differ ences in the kinetic profile between these two enzymes. These differences are minor for newer PIs such as APV, LPV, ATV, and TPV and of greater amplitude for SQV, RTV, IDV, and NFV. The variants harboring D30N/N 88D combination of mutations exhibited similar kinetic behavior for both subtype B and C PRs. The subt ype B mutant maintained 70% of the catalytic
99 efficiency when compared with the wild type, much better than the mutant harboring D30N alone. Previous studies showed that there is a strong association betw een D30N and N88D in subtype B PR, with the latter probable having the role to stabilize the enzyme by compensating for the negative charge lost with D30N subs titution (Roge et al. 2003; Mitsuya et al. 2006). The D30N/N88D subtype C PR preserved 55% of the wild type enzymatic activity. The Ki value profile is comparable between th ese two mutants with, as expect ed, a significant decrease in binding affinity for NFV. For the D30N/L90M double mutants, we no ticed a more notable increase in Ki value for NFV for subtype B PR (52-fold), compared to s ubtype C mutant (24-fold ). However, we also observed a significant decrease of almost 10-fold in the catalytic efficiency of the D30N/L90M subtype B mutant. On the other hand, upon acquisiti on of this combination of mutation, subtype C PR exhibited only a 50% decrease in its enzymatic activity. Our kinetic results appear to be in accordance with previous studies that showed that subtype B mutants carrying D30N/L90M combination are only rarely found in vivo and generally have a dditional, potentially compensatory, mutations. Interestingly, a major loss of replicative capacity for a mutant clone harboring both D30N and L90M was also descri bed (Martinez-Picado et al. 1999; Sugiura et al. 2002; Perrin and Mammano 2003). It can be inferred th at in the context of this combination of major resistance mutations, the naturally occurr ing polymorphisms within subtype C PR could act as compensatory mutations having role in pres erving the catalytic effici ency of the enzyme. N88D/L90M combines two non-active site muta nts: N88D, which has high stability with L90M, which shows poor structural stability (X ie et al. 1999; Mahalingam et al. 2002). Both N88D and L90M are not part of the substrate-bi nding site, but they can influence the catalytic process through indirect interact ions. This combination of mutations appeared to be very
100 favorable in the context of the naturally o ccurring polymorphisms of subtype C PR when compared to subtype B PR. The subtype C mutant exhibited increased catalytic efficiency when compared with both the wild type subtype C PR and the similar variant of subtype B PR. The inhibition profile showed higher Ki values for subtype C PR when compared to subtype B PR mutant for all PIs tested in this study Not surprisingly the fold-increase in Ki value for NFV was modest, but the Ki values for SQV and IDV showed 50a nd 28fold increase, respectively for N88D/L90M subtype C PR and 22and 10fold increase, respectively for subtype B variant. Surprisingly, the mutants harboring the tr iple combination D30N/N88D/L90M showed improved catalytic efficiencies for both subtypes and increased Ki values for NFV, SQV, and RTV. The fold-change was higher for NFV in s ubtype C PR (83-fold for NFV versus 52-fold increase in subtype B PR). Our data also demonstrated that the D30N mutation in combination with D88N and L90M provides a higher level of cross-resistance for su btype C PR, specifically to RTV. This is of interest as no such combination was described in response to RTV in s ubtype B PR. This result could argue for a new mutational pathway in de veloping resistance to RTV in subtype C PR. However, despite the high fold-increase, the absolute Ki value remains low, indicating that RTV maintains sufficient binding stre ngth to effectively inhibit th e D30N/N88D/L90M mutant of subtype C PR. For these reasons th e use of pretherapy and therapy resistance testing, especially in cases of early failure, is crucial for providing optimum treatment potency. An interesting observation is that the kineti c parameters of subtype C PR harboring the D30N mutation do not improve in the same wa y as happens for subtype B PR upon acquisition of other drug resistance mu tations. Specifically, D30N subtype C PR does not improve its catalytic efficiency upon acquis ition of N88D, as happens for subtype B PR. Also, the Ki values
101 for NFV for D30N subtype C PR do not change after combination with N88D and L90M. A possible interpretation of these results is that the naturally occu rring polymorphisms in subtype C PR have already set a favorable context in which the acquisition of D30N by itself is sufficient enough to attain a higher level of resistance, wh ile in subtype B PR other major mutation are required to develop a similar degree of resistance. We have to mention that this effect can be limited to the combination of baseline polymorphi sms in subtype C PR and the type of major mutations induced by the ARV treatment. We do not attempt to make the same assumption for other subtype PRs, due to the fact that they ha rbor different combinations of polymorphisms that might overall not have the same effect on the biochemical and stru ctural characteristics of the PR. This is another reason why the exploration of each subtype PR is important in the context of designing the best ARV treatment strategy, especially for patients who have failed multiple drug regimens. In our study the binding affinities of the ne wer PIs, APV, LPV, ATV, TPV, seem less affected by the introduction of three mutations studi ed here. We also have noticed that most of the subtype C PR mutants analyzed in this st udy appear to exhibit a modest increase in the susceptibility to TPV, as shown by the lower Ki values for this PI. It might be due to the fact the newly designed PIs are able to evade drug resistan ce mutations better that the older PIs. Also, the mutations we studied are quite specific for these PIs. On the other hand, we have not studied mutations that specifically aris e upon treatment with these inhib itors, such as I50L for ATV or I84 V for LPV but such studies are under way in our laborator y (data not shown). Also, the viruses isolated from patients that failed ARV treatment usua lly have PRs harboring a much larger number of mutations. So, in the long term the minor effects introduced by the baseline polymorphisms, exemplified in our studies by modest changes in Ki values, could be even
102 further accentuated by the multiplicative effect of addition of multiple drug resistance mutations (Wu et al. 2003; Clemente et al. 2006). We ex pect to see even larg er differences in Ki values between subtype B and C PRs upon acquisition of a larger number of drug resistance mutations, as usually happens during ARV therapy. Gonzalez et al. showed th at the addition of three or four IDV drug resistance mutations has led to a highly resistant subtype C virus compared with its subtype B counterpart (Gonzalez et al 2003). The same phenomenon was observed with NFV. This means that the IDV drug resistance mutations impact was much more noticeable in subtype C than in subtype B viruses. Differences in behaviors between the subtype C and subtype B vi ruses observed in vitro in this study and in our study may have a clinical relevance, suggesting a lower genetic barrier in subtype C viruses than in subtype B isolates. The PR clone used in our studies does not harbor a frequent polymorphism found in subtype C PRs: R41K. It is one of the residues that appears to be involved in preserving or augmenting the catalytic efficiency of the subt ype C PR and enhancing the viral fitness when compared to subtype B viruses (Ziermann et al 2000; Velazquez-Campoy et al. 2001). It also appears to decrease viral susceptibility to RT V and APV treatment. Our intention was not to study the effects of each individual polymorphic residue but their influence as a group within the subtype C PR context. In conclusion, the differences observed in our study between various mutants of subtype B and C PRs are due to the presence of the pree xisting polymorphisms. Our study showed that the presence at the start of thera py of the naturally occu rring polymorphisms could give the virus an advantage in the rapid development of drug resist ance while preserving the virus viability in the presence of a specific inhibitor. We also belie ve that the effects of the naturally occurring
103 polymorphisms in subtype C PR are influenced by the PI choice for treatment and by the type of major mutations acquired upon drug treatment. The same is probably true for all other HIV-1 subtypes. These mutation patterns are complex and fr equently overlapping. It is necessary to test the major groups of naturally occurring polymor phisms for each subtype in combination with several major drug resistant mutations to further be able to predict th e response to treatment based on the combination of baseline polymorphi sms. Also, determining the biochemical and biophysical properties of enzymes with these pa tterns of mutations will be important for designing new PIs that are less lik ely to trigger resistance or ar e effective against already drugresistant isolates. It is important to mention that studying the addition of PI-resistance mutations within the PR is limited only to their effect on the biochemi cal characteristics of the enzyme. The set up of our experiments prevents us from studying the e ffects of the naturally occurring polymorphisms within the gag or the viral context. For exampl e, several studies showed that there is an interrelation between ac quisition of D30N and N88D and cha nges occurring within the rest of the gag and gag/pol polyproteins It has been shown that mu tations in the HIV-1 protease substrate cleavage-site p1p6 covary with th e D30N/N88D protease mutations. Aspartyl at position 30 is important both to the binding of NFV and also likel y to the recognition of the p1 p6 cleavage site. Structural anal ysis shows that both NFV and p1p6 have atoms that protrude beyond the substrate envelope and contact As p30. Thus, both the i nhibitor and the p1p6 substrate are likely to be affect ed by D30N mutation. This likely explains the particular coevolution of the p1p6 cleavage site with the D30N-resistant muta tion and also why no other coevolution with any of the other subs trates occurs (Kolli et al. 2006).
104 All these studies highlight the importance of fu rther characterization of the development of resistance in non-subtype B viruses. For so me drugs, the differences seen in baseline polymorphisms between subtypes may influence which mutational patterns develop. Subtle effects of such polymorphisms on drug susceptibili ty and replicative capacity may underlie such changes. Despite their limitations, our studies veri fy and complement data obtained through phenotypic studies. While experime nts involving virus cultures might require more time and are more expensive, these biochemical studies can be done quickly and at lower cost. Also biochemical studies are performed in a controlled environment, in which differences in response can be tracked down to a single variable. The resu lts of our studies should prove useful in the design of new and continuing therapy.
105 Figure 4-1. Polymorphic sites within the s ubtype B and C protease sequences. Accessed and adapted on June, 2007 from http://hivdb.stanford.edu /pages/barChart_MutPrevB ySubtype/PR_untreated.html Stanford Database
106 A B 1 PQITLW Q RPL V T IK I GGQ L K EALLDTGADD TV L EE MS LPG RWKPKMIGGI 50 C 1 PQITLW K RPL V S IK V GGQ I K EALLDTGADD TV I EE IA LPG RWKPKMIGGI 50 B 51 GGFIKVRQYD QI L IEICG H K AIGTVLVGPT PVNIIGRN L L TQ I GCTLNF 99 C 51 GGFIKVRQYD QI I IEICG K K AIGTVLVGPT PVNIIGRN M L TQ L GCTLNF 99 B Figure 4-2. HIV-1 s ubtype C protease. A) Pair-wise sequence alignment of HIV-1 subtype B PR LAI strain (designated as B) and subt ype C PR (designated as C). B) Cartoon representation of HIV-1 subtype C PR. Th e naturally occurring polymorphisms in subtype C PR, represented as red spheres, are superimposed onto the crystal structure of HIV-1 subtype B PR (b lack ribbon). Amino acid positions are as numbered.
107 A B Figure 4-3. Kinetic analysis of HIV-1 subtype B and C protea ses. A) The Michaelis-Menten constants: kcat/Km. The dark gray and light gray ba rs designate subtype B PR and C PR, respectively. B) The Ki values for subtype B and C wild type PRs. The blue and red bars designate subtype B PR and C PR, respectively. 0 0.2 0.4 0.6 0.8 1.0 PR / B PR /C kcat/Km (sec-1 M-1) Ki values (nM) 0 0.5 1 1.5 2 2.5 3 3.5 4 RTV IDV NFV SQV APV LPV ATV TPV PR B PR C
108 A B Figure 4-4. Kinetic analysis of the single muta nts of the subtype B and C proteases. A) The Michaelis-Menten constants: kcat/Km. B) The Ki values for subtype B and C PRs. PR/C N88D PR/C L90M RTV IDV NFV SQV APV LPV ATV TPV PR/B-D30N PR/B-N88D PR/B-L90M PR/C-D30N -10 0 10 20 30 40 50 60 70 80 90 Ki values (fold change) kcat/Km (sec-1 M-1) 0 0.2 0.4 0.6 0.8 1 PR/ B PR/BD30N PR/BN88D PR/BL90M PR/C PR/CD30N PR/CN88D PR/CL90M
109 A B Figure 4-5. Kinetic analysis of the double mu tants of subtype B and C proteases. A) The Michaelis-Menten constants: kcat/Km. B) The Ki values for subtype B and C PRs. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 PR/B D30N/N88D D30N/L90M N88D/L90M PR/C D30N/N88D D30N/L90M N88D/L90M kcat/Km (sec-1 M-1 ) RTV IDV NFV SQV APV LPV ATV TPV PR/B-D30N/N88D PR/B-D30N/L90M PR/B-N88D/L90M PR/C-D30N/N88D PR/C-D30N/L90M PR/C-N88D/L90M -10 0 10 20 30 40 50 60 Ki values (fold change)
110 A B Figure 4-6. Kinetic analysis of the triple mu tants of subtype B and C proteases. A) The Michaelis-Menten constants: kcat/Km. B) The Ki values for subtype B and C PRs. kcat/Km (sec-1 M-1) 0 0.3 0.6 0.9 1.2 1.5 PR/B PR / B30/88/90 PR /C PR /C 30/88/90 Ki values (fold change) RTV IDV NFV SQV APV LPV ATV TPV PR/B-30/88/90 PR/C-30/88/90 -10 0 10 20 30 40 50 60 70 80 90
111 Table 4-1. The Michaelis-Menten constants for the HIV-1 subtype B and C proteases Enzyme/Subtype Km (mM) kcat (sec-1) kcat/ Km (sec-1mM-1) PR/B 19.5 2 10 1 0.55 0.06 PR/B-D30N 38.6 2 5 0.5 0.13 0.02 PR/B-N88D 16 1.5 12 1.1 0.75 0.10 PR/B-L90M 10 1 5.8 0.5 0.58 0.06 PR/B-D30N/N88D 35 3.5 12.4 1 0.4 0.04 PR/B-D30N/L90M 61 6 3.5 0.4 0.06 0.008 PR/B-N88D/L90M 30 2 10.2 1 0.34 0.04 PR/B-30/88/90 30 2.3 30.6 7.5 1.0 0.3 PR/C 17 1 5.6 0.2 0.32 0.02 PR/C-D30N 64 6 12 1 0.18 0.02 PR/C-N88D 22 3 4.3 0.4 0.19 0.03 PR/C-L90M 23 3 4.8 0.3 0.2 0.04 PR/C-D30N/N88D 19 1.6 3.2 0.3 0.17 0.02 PR/C-D30N/L90M 23 3 4 0.2 0.17 0.02 PR/C-N88D/L90M 24 3 13 1.7 0.51 0.09 PR/C-30/88/90 35.5 3 22 2.4 0.6 0.1
112 Table 4-2. The Ki values (M) for HIV-1 subt ype B and C proteases. The Ki values are in bold font and the errors are in paranthesis. Subtype-Mutant RTV IDV NFV SQV APV LPV ATV TPV B 0.07 (0.01) 1.8 (0.1) 1.7 (0.4) 2.2 (0.3) 0.4 (0.1) 0.11 (0.03) 0.07 (0.01) 0.4 (0.04) B-D30N 0.15 (0.01) 9.4 (1.1) 44.0 (5.5) 3.8 (0.3) 0.4 (0.02) 0.4 (0.05) 0.2 (0.02) 0.2 (0.03) B-N88D 0.24 (0.05) 9.9 (0.8) 3.4 (0.3) 5.7 (0.7) 0.39 (0.05) 0.08 (0.02) 0.21 (0.02) 0.8 (0.1) B-L90M 0.42 (0.07) 5.0 (0.4) 5.4 (0.5) 7.5 (0.9) 0.36 (0.05) 0.10 (0.03) 0.17 (0.04) 0.23 (0.04) B-D30N/L90M 0.32 (0.05) 10 (1) 88 (10) 15 (2) 0.47 (0.07) 0.44 (0.07) 0.28 (0.04) 0.5 (0.1) B-D30N/N88D 0.87 (0.17) 6.9 (0.6) 68 (6) 20 (2) 0.08 (0.02) ND 0.09 (0.03) ND B-N88D/L90M 0.71 (0.13) 17.7 (1.3) 12 (1) 49 (4) 0.89 (0.11) 0.15 (0.02) 0.35 (0.05) 0.38 (0.04) B-30/88/90 1.0 (0.2) 5 (1) 88 (7) 47 (6) 0.61 (0.07) 1.0 (0.2) 0.37 (0.05) 1.9 (0.2) C 0.27 (0.06) 3.3 (0.3) 2.7 (0.3) 2.6 (0.2) 0.29 (0.03) 0.19 (0.02) 0.13 (0.01) 0.11 (0.02) C-D30N 0.6 (0.1) 10.4 (0.3) 151 (4) 9.0 (0.1) 0.6 (0.1) 0.3 (0.03) 0.8 (0.1) 0.1 (0.01) C-N88D 0.1 (0.02) 24 (2) 6.2 (0.7) 5.7 (0.1) 0.5 (0.2) 0.15 (0.01) 0.5 (0.1) 0.2 (0.03) C-L90M 0.78 (0.05) 4.6 (0.4) 5.2 (0.5) 7.1 (0.8) 0.36 (0.05) 0.48 (0.04) 0.27 (0.04) 0.09 (0.02) C-D30N/L90M 0.27 (0.06) 4.8 (0.1) 41.6 (4.3) 6.5 (0.2) 0.18 (0.02) 0.19 (0.05) 0.08 (0.01) 0.4 (0.08) C-D30N/N88D 0.04 (0.01) 5.8 (0.4) 40.1 (3.4) 10.9 (0.8) 0.16 (0.03) 0.18 (0.03) 0.06 (0.01) 0.19 (0.03) C-N88D/L90M 1.0 (0.15) 50 (4) 14 (2) 109 (13) 1.5 (0.2) 0.43 (0.08) 0.54 (0.08) 0.15 (0.05) C-30/88/90 2.2 (0.3) 8.1 (0.1) 141 (17) 47 (4) 0.39 (0.08) 0.8 (0.2) 0.29 (0.03) 0.39 (0.06)
113 CHAPTER 5 STRUCTURAL ANALYSIS OF HIV-1 SUBTYPE C PROTEASE Introduction Extensive structural studies have been done w ith HIV-1 protease (PR) in an attempt to better understand the molecular mechanisms that govern the interactions between this enzyme and substrates or inhibitors. HIV-1 subtype B PR three-dimensional structure has been determined both alone and complexed with different inhibitors (Clemente et al., 2006; Clemente et al., 2004; Logsdon et al., 2004; Ringhofer et al., 1999; Ru tenber et al., 1993). Crystal structures of HIV-1 PR were first reported in 1989 (Miller et al., 1989; Navia et al., 1989; Wlodawer et al., 1989) and their av ailability had a major role in the process of drug development (Vondrasek and Wlodawer, 2002). Cr ystal structures show that HIV-1 PR forms a binding site that consists of subsites S4-S4, which span ab out eight residues (P4-P4) of a peptide substrate (Figure 2-3). Many HIV-1 PR muta nts have also been crystalli zed alone or complexed with peptido-mimetic or non-peptido-mimetic inhibito rs (Clemente et al., 2003; Mahalingam et al., 2001; Mahalingam et al., 2004). Some of these mutant s show structural cha nges consistent with the differences seen in their enzymatic activity (Mahalingam et al., 2002; Mahalingam et al., 2001; Mahalingam et al., 2004). In a recent study, Pr abu-Jeyabalan et al. (P rabu-Jeyabalan et al., 2003) solved the crystal structures of an inactiv e PR bearing the mutations D25N and V82A in complex with three substrates and two inhibito rs, (saquinavir (SQV) and ritonavir (RTV)). The study analyzes the mechanisms through which mutati ons in the active site can drastically reduce the affinity of the PR for the inhibitors, but at the same time, exert a smaller effect on the enzyme ability to cleave the substrate. These findings su ggest that future inhibitor design may reduce the probability of the appearance of drug-resistant muta tions by targeting residues that are essential
114 for substrate recognition. Perhaps this type of directed inhibitor-des ign will make it more difficult for further drug resistance to evolve. This wealth of information regarding inter actions between HIV-1 PR and substrates or inhibitors is available only fo r subtype B PR. The first crystal structure of a non-B subtype PR (subtype F PR) was reported recently (Sanches et al., 2007; Sanches et al., 2004). In addition, Coman et al. reported the crysta llization of the HIV-1 subtype C PR complexed with indinavir (IDV) and nelfinavir (NFV) (Coman et al., 2007). In this chapter we analyze and compare the differences and similarities between subtype B and C PRs. We believe that the successful crystallizati on of the protein-inhi bitor complex is of great significance. It will provide us with insight on how this PR interacts with clinically used inhibitors at each individual subs ite at the atomic level, especially for those residues presumably critical for substrate and inhibi tor binding. Thus, the structural information obtained from the crystallographic analysis can not only verify th e data observed in kinetic and functional studies but also provide structural reference for substrat e and inhibitor design. Furthermore, it will reveal instructive information about th e effect of baseline polymorphism s on the overall structure of PR, and will help identifying th e long-range interactions thro ugh which these non-active site mutations can affect the binding of subs trates/inhibitors in the active site. Results Crystallization X-ray crystallography is one of the few methods available to study the atomic threedimensional structures of pr oteins and macromolecular complexes. The first step in macromolecular structure determination is th e production of highly purified sample (95-99% homogenous) in high enough amounts. This is a crucial process, and often can mean the difference between diffraction-quality crystals and amorphous precipita te. The next most
115 important, and the most time-consuming and pa tience-demanding step is the finding of the condition(s) for crystallization of the protein under study. The overa ll approach is an iterative one. An initial set of crystallization conditions are screened, the outcome is observed, and, if crystals do not immediately form, the conditions are then modified taking into account what was learned during the previous attempt. The most common approach is to use commercially available crystallization kits. These sample k its contain a range of previously successful crystallization conditions with varying precipitant c oncentrations as well as pH ranges. Unbound HIV-1 subtype C protease Crystals of the unbound subtype C PR appeared under the following conditions: 1 M NaCl, 30 mM citric acid, pH 5.0 with Triton-X 100 as additive (Hampton). The in itial inspection of the crystallization drop, after 12 hour s following the equilibration agai nst the precipitant solution at room temperature, revealed fine precipitation that increased in amount over the next 48 hours. At the beginning of the third day se veral diamond-shape crystals app eared and they increased their size over the next five days (Figure 5-1A). Drug-bound HIV-1 subtype C protease We obtained crystals of the s ubtype C PR bound to Indinavir (IDV) or Nelfinavir (NFV). The crystallization conditions for both were: 1 M NaCl, 30 mM citric acid, pH 5.0. Crystals suitable for X-ray diffraction studies of the dr ug-bound subtype C PR appeared approximately 24 hours after setting up the drop. These crystals grew as thin, rectangular plates, stacked in top of each other, with approximate dimensions of 0.4 x 0.3 x 0.1 mm (Figure 5-1B). Diffraction Data Collection, Processing and Scaling The degree of oscillation data that should be collected from a crystal depends on the space group of the crystal. Accord ing to Friedels law, 180 o of data is the most, theoretically, that need to be collected if no other symmetry exists within the crystal system (triclinic system). Thus,
116 higher symmetry space groups require less data to be collected, this being due to the occurrence of more symmetry equivalent plan es due to the lattice symmetry. Once a complete data set has been collected, th e initial diffraction images can be used for space group determination, followed by processing of the complete set of diffraction images, reduction of data to only unique reflections base d on the assigned space group, and scaling of the data for intensity normalization and calculation of Rsym. The Rsym value measures how well the collected data fits th e imposed space group symmetry elemen ts. The collection of high quality data is of utmost importance for structure determination as well as refinement of the macromolecular model. The calculation of the Rwork and Rfree values (two important determinants of crystal structure quality) will all be dependent u pon the initial measurements of the intensities. Unbound HIV-1 subtype C protease A total of 180 o of data were collected using a 0.5 o oscillation angle. This resulted in a total 360 images and a total of 279351 reflections measured. A comp lete data set was collected to 1.2 resolution. Initially, the unbound form of the subtype C PR crystals were believ ed to belong to the tetragonal symmetry group P422, with unit-cell parameters a = b = 46.7 c = 100.8 The data were scaled in the tetragonal space group P422 and were merged and reduced to a set of 35611 independent reflections (99.3% completeness, 95.7% in the outer resolution shell) resulting in a scaling Rsym of 0.076 (0.276 in the outer resolution sh ell). But, due to the fact that a good electron density map could not be obtained, as explained later in this chapter, during the refinement the data were rescaled in the space group P41 and were merged and reduced to a set of 66386 independent reflections ( 98.9% completeness, 91.2% in the outer resolution shell) resulting in a scaling Rsym of 0.073 (0.246 in the outer reso lution shell) (Table 5-1). The Rsym values are very similar between the two scaled da ta sets processed in two different space groups.
117 From the unit-cell volume and the molecular weight of subtype C PR, Vm values (Matthews, 1968) of ~2.56 3/Da (52% solvent content) were calculated assuming four homodimers per asymmetric unit using CNS v.1.1 (Brunger et al., 1998). Drug-bound HIV-1 subtype C protease A total of 180 of data were collected (360 images) from each of two single crystals of subtype C PRs complexed with IDV or NFV and a total of 65956 and 61466 reflections were measured, respectively. Complete data sets were collected to 2.3 resolu tion for both structures. The data for the subtype C PR complexed with IDV were initially processed in the monoclinic space group P2. The diffr action data were subsequently processed and reduced with an overall Rsym of 0.084 (0.332 in the outer resolution sh ell), and completeness of 96.7 % (96.4% in the outer resolution shell). A similar data set for the subtype C PR complexed with NFV was processed as well in P2 space group and resulted in an Rsym of 0.113 (0.373 in the outer resolution shell) and completeness of 97.3 (93.7% in the outer resolution sh ell) Table 5-1 gives a full summary of the data collection statistics. On inspection of the intensities of the 0k0 refl ections, the existence of a twofold screw axis along the b direction could be inferred, theref ore implying that the space group was P21 for both crystals. From the unit-cell volume and the mol ecular weight of subtype C PR complexed with IDV and NFV, Vm values (Matthews, 1968) of ~2.9 3/Da (56% solvent content) for both crystals were calculated assuming two homodimers per asymmetric unit using CNS v.1.1 (Brunger et al., 1998). Molecular Replacement: Particle Orientation and Position Molecular replacement exploits the similarity of the tertiary fold of the protein or macromolecular complexes when structures are being determined. A clue to the degree of sequence and, thus, structural homology that ex ists between two prot ein molecules is an
118 alignment analysis or pair-wise BLAST search. It is generally agreed that if two proteins share more than 30 % sequence identity their three-di mensional structures will be superimposable to within few Angstroms. Structure determination using molecular replacement consists of two separate operations: a rotation s earch and a translation search. Both operations rely upon the three-dimensional Patterson function to derive a structure solution (Ro ssmann, 1990). There are many programs with slightly differe nt algorithms that have been developed to implement this process, such as those in CNS and CCP4 and molecular replacement has become the fastest and the most direct method of macromolecular stru cture determination (Bru nger et al., 1998). Unbound HIV-1 subtype C protease We decided to use 1HHP (Spinelli et al., 1991) as model to solv e the unbound subtype C PR molecule orientation and position. The cross-rotation function in the point group P422 searches provided one clear solu tion with the correlati on factor of 0.1503, with the next highest peak having a correlation coefficient of 0.0682. Using this orientation matrix, a translat ion function search in the Laue group P41212 for the non-crystallographic dimer rotation solution provid ed one peak with correlation coefficient of 0.654 and R-values of 0.587. We also perf ormed translation function searches for P42212 and P43212 but no peaks were found. Drug-bound HIV-1 subtype C protease Cross-rotation function searches with the subtype C PR IDV complex data in the point group P2, using the 1W5Y (Lindberg et al., 2004 ) polypeptide model provided four solutions, two of which reflected the molecular two-fold ro tation axis of the model. The four unique peaks were approximately twice the va lues of the next highest p eaks in the rotation function. Translation function search us ing the subtype C PR IDV co mplex data in the space group P21 for the two non-crystallographic dimer rotati on solutions provided pe aks with correlation
119 coefficients 0.380 and 0.287, and R-values of 0. 473 and 0.518, respectively. A second translation search was required to define th e relative orientation of the two dimers along the y direction, by constraining the position in the A dimer in a second translation search. This solution had a correlation coefficient and R-value of 0.675 and 0. 358, respectively. The fractional translations were Tx=0.4238, Ty=0.0000, and Tz=0.0486 for site A, and Tx=-0.4222, Ty=0.0245, and Tz=0.4449 for site B. The solution for the subtype C PR NFV complex data used the orientations and positions as defined from the subtype C PR IDV complex data. Structure Refinement Once an initial model has been constructed, ite rative refinement can be carried out either with CNS package or with the CCP4 program Refmac5. These algorithms will refine the model based upon accepted bond lengths and angles that ha ve been found in well-refined proteins and peptides solved to date. The goal of the refine ment is to improve the agreement between the observed amplitudes obtained from the diffraction data and calculated values for the model. To refine the HIV-1 subtype C PR structures, both bound and unbound forms, we employed the CNS refinement algorithm that consists of several steps: rigid-body, simulated annealing, B-factor refine ment and conjugate-gradien t energy minimization. For all three structures a ri gid body refinement was undertaken in order to optimize the positions of the molecules in the asymmetric unit. This is a preliminary step in refinement that does not take in the account the conformation of th e amino acid side-chains, but facilitates a finer treatment of the model in the subsequent refineme nt steps. Individual B f actor refinement allows for a better measurement of the movements and dyna mics of the molecules. This, combined with conjugate-gradient energy minimization, will ap ply restraints on bond lengths and angles to bring the molecule to the most energetically favor able conformation. Once these steps have been finished, electron density maps are calculated using these improved pha ses and the degree of
120 agreement with the model is examined manually and adjusted accordingly. These refinement steps are part of an iterative refinement loop a nd are repeated several ti mes to attain the best model. Unbound HIV-1 subtype C protease The solution obtained through ro tation and translation searches was further refined with cycles of rigid body, individual B-factor, energy minimization, a nd water-pick procedures in CNS interspersed with rounds of manual modeling using the program O (Jones, 1978). Initial Fo-Fc and 2Fo-Fc electron density maps were calcul ated using the phases from the model obtained by molecular replacem ent and were contoured at 3.0 and 1.5 respectively. The maps were of good quality, the main chain electron density was continuous and the side chains of the amino acids similar between the two protein molecules were well fitted into the density. The maps also clearly showed that the flap regions of the model were not fitting within the electron map, and further ex tensive rearrangements on the b ackbone and side chains were required. However, despite extensive cycles of refinement in CNS the Rwork and Rfree values did not converge to less than 20.6% and 21.5%. Also, at resolutions higher than 2.3 the Fo-Fc maps showed patches of positive elec tron density along the main chai n of the protein molecule. The results called for a re-examination of the assigned space group and data processing. The data were rescaled in P41 space group. Upon this reassignment, the R values and the quality of the electron density map were substantially improve d and further refinement was performed in CNS out to 1.6 resolution. Figure 5-2 shows the pa cking arrangement for the 8 monomers for the unbound HIV-1 subtype C PR in the space group P41. The PDB coordinate file obtai ned from the final cycle of CNS refinement was used to generate the fractional coordina tes and equivalent isotropic th ermal parameters for further
121 refinement using SHELX and SHLEXPRO (Sheldrick, 1997; Sheldrick and Schneider, 1997). The water and solvent molecules placed in the model during the CNS refinement steps were not removed prior to input into SHELX The first cycle of refinement resulted in an initial Rwork of 23.8% and Rfree of 26.4% (Table 5-2). Anisotropic refi nement of the protein atoms only during the second cycle in SHELX and removal of 50 w eaker water molecules in creased significantly the number of parameters (Npar) and brought the Rwork and Rfree down to 14.9% and 18.3% respectively. In the next two steps, the addition of more solvent molecules in the model did not significantly improved the R-values. The next st age of refinement involved the generation of 1618 hydrogen atoms (according to the riding H-atom model) resulting in a significant decrease in Rwork and Rfree to 13.9% and 16.9%, respectively. Full anisotropic refinement proceeded from this point given the high resolu tion (1.20 ) and yielded an Rwork of 12.4% and Rfree of 15.3%. The last step of refinement in SHELX was perf ormed using 100% of the data, including 5% of the data put aside for Rfree calculation and resulted a final Rwork of 12.3% (Table 5-3). Drug-bound HIV-1 subtype C protease Rigid body refinement of the solution for HIV1 subtype C PR IDV complex, using data between 15 and 4.0 resolution, converged at Rfree = 30.3 % and Rwork = 32.1 %. Figure 5-3 shows the packing arrangement fo r the two homodimers in the P21 space group for the subtype C PR IDV complex. Initial Fo-Fc electron density maps for both complexes, contoured at 3.0 showed clear interpretable density for the re spective inhibitors IDV and NFV present in both dimers. Topology and parameter files for IDV and NFV were obtained from the HIC-UP website. Iterative cycles of refinement were performed in CNS including data between 15 and 2.3 resolution. The final R values converged to Rwork of 25.3 % and Rfree of 28.3 % and Rwork of 21.1 % and Rfree of 23.7 % for NFV and IDV complexes, respectively (Table 5-3).
122 Structure Validation Structure validation of the refined model is an important final step in the structure determination and it can be perf ormed using programs such as PROCHECK (Laskowski et al., 1993). Validation is used to en sure that bond lengths and angl es are in accordance with acceptable values. Once validated, the model can be reliably interpreted to elucidate the biological implications the stru cture contains. Once structure refinement and validation is complete, final models are submitted to the Protein Data Bank (PDB) for later release to the public. Unbound HIV-1 subtype C protease The quality of the final refined structured of HIV-1 subtype C PR was verified with the PROCHECK program (Laskowski et al., 1993). 94.9 % of the dihedral angles were found to be located in the most favored regions, with all th e other in the additional allowed regions (Figure 5-4A). Drug-bound HIV-1 subtype C protease The final models for IDVand NFVbound s ubtype C PR showed 96.2 % and 95.9 % of the dihedral angles in the most favorable regions, respectively, with all the other in the additional allowed regions (Figure 5-4B, C). Structure Analysis As expected, given the high amino acid sequence identity, all three crystal structures solved in this study, the unbound form of subtyp e C PR and the enzyme complexed with IDV and NFV, had the overall shape and fold of HIV-1 subtype B PR. The PRs crystallized as homodimers with the substrate-binding cleft in the middle of the mo lecule, covered by two flexible arms known as the flaps.
123 Unbound HIV-1 subtype C protease This high-resolution structure showed excellent electron dens ity for all the PR atoms and solvent molecules. The electron density map of the naturally occurring polymorphisms at positions 36 and 37 is shown in the Figure 5-5. An unassigned density was located between the open flaps of the unbound subtype C PR. The Fo-Fc electron density has a C-shaped appearance and was initially interpreted and built as a string of disordered water molecules. Later, during the refinement, the density was interpreted and built as a fragment of Triton-X 100, the add itive added in the crys tallization drop (Figure 56). The average B values for the main-chain and the side-chain atoms were 13.92 and 16.1 2, respectively (Table5-3). The dist ribution of the B factors (atom thermal motion) for the main chain atoms showed similar peaks at the Cand Ntermini, at the elbow (34-38 residues), at the tip of 60s loop (65-70 residues) and in the activ e site (78-84) for both subunits for subtype C and subtype B multi-drug resistant PR (Figure 57). The B factors profile for the wild type subtype B PR was different from th e outer two structures in that it had a much lower peak for the active site residues 78-84. Interest ingly, the B factors for subtype C PR showed a peak in the 1220 -sheet region, where three of the naturally occurring polymorphisms are located. This peak does not appear in the multi-drug resistant B PR (PDB code 1RPI). Also, the B values for the flap regions in subtype B PR are below the aver age indicating that there is limited motion of the atoms within these structures. The B values of the same regions in subtype C PR show values slightly above the average and higher that those for subtype B PR, while the wild type subtype B PR showed higher B values for the flap region an d a slight shift of the peak B values for the residues located in the elbow of the flap.
124 The solvent for the final model included 341 wa ter and 4 glycerol mole cules, with average B values of 26.7 and 31.6 2, respectively (Table 5-3). Analysis of the anisotropy of the structure was performed using the final output file from SHELX Anisotropy is defined as the ratio betwee n the minimum and maximum eigenvalues of the matrix of anisotropic displacement parame ters (Merritt, 1999a). An analysis of the anisotropically refined structures in the Protein Data Bank perfo rmed by Merritt, 1999 indicates that the standard mean anisotr opy is 0.45 with a standard devia tion of 0.150 (Mer ritt, 1999b); the current model of the unbound subtype C PR confor ms to these values. We were particularly interested in the distribution of the anisotropic parameters of the atoms in the flaps as this would help in understanding the direction of the motion of these highly flexible structur es (Figure 5-8). A careful analysis of the ellips oidal atomic motions indicates th at a possible direction of the motion of the flaps is perpendicular on the activ e site. This motion is accompanied, at a lesser extent, by the similar movements in the elbow of the flap. High-resolution structural information allows for better interpretation of the structural disorder, including amino acid side chains that ex hibit alternate conformations (Esposito et al., 2000). In our model, several residues showed di sordered density for the side chain and had alternate conformations. The a lternate conformations for Glu21, Glu34, Glu35, Pro44, Arg57, Lys69, and Val82 were genera ted in the graphics program O version 10.0.1 (Jones, 1978). All of these residues are located in the outer loops of the enzyme, in contact with the solvent molecules. As expected, the HIV-1 subtype C PR structure is highly sim ilar in shape and fold with subtype B PR. A least-square superimposition of our model of subtype C PR in the unbound form and the previously reported wild type (PDB code 1HHP) and multi-drug resistant form (PDB code 1RPI) of subtype B PR was conducte d using the graphic software Pymol (DeLano
125 Scientific) (Figure 5-9A). The wild-type stru cture of subtype B PR (1HHP) known as the semiopen form, crystallized in the space group P41212 with cell constants similar to those of the subtype C PR and multi-drug re sistant mutant of subtype B PR (1RPI), which crystallized in P41 space group. The mean r.m.s. deviations for C atoms between 1HHP, 1RPI and our model were 1.09 and 0.65 respectively (Figure 5-9B). These valu es are above the ~0.5 that represents the crystallographic error margin when comparing to crystal structures The catalytic tr iplet residues 25-27 located in the active site showed values of 0.14 and 0.35 cons istent with the highly conserved core structure. However, an interestin g observation was that th e main chain atoms of the active site residue Val82 showed an r.m.s. de viation of 1.07 to 1.40 The highest difference was observed, as detailed below, between the flaps of the PR. The relatively high mean r.m.s.d between subt ype B and C PRs is the result of several differences of more than 1.0 that were observed for the main chain atoms of the residues: 3542 (the elbow of the flaps), 49-53 (the tip of th e flaps), 63-70 (the 60s loop), and 80-81. The highest r.m.s. deviations were located in the outer l oops, especially in the loops harboring the naturally occurring polymorphisms such as the elbow of the flap and the 60s loop. The most striking difference between the wild-type subtyp e C and B PRs is a very large conformational change in the flap region, with the distance be tween the tips of the flap s of the subtype B PR being 4.4 while the distance for subtype C PR is 12.2 (Figure 5-10A). This accounts for a movement of more than 8 of the flaps of the subtype C PR relative to the subtype B PR structure. Interestingly, our st ructure showed a similar distance between the flaps as that of a subtype B multi-drug resistant PR (1RPI) that sh ows an opening of the flaps of 12.53 (Figure 5-10B).
126 M36I is one of the baseline polymorphisms occu rring in subtype C PR that is considered a secondary drug resistance mutation in subtype B PR. Analyzing both PRs, we observed that the side chain of the smaller Ile residue make half of the van der Waals in teractions observed in subtype B PR harboring the Met residue. We also noted a downsh ift in the subtype C PR 36 41 loop with an average displacement of 1.6 and 1.3 when compared with the wild-type and multi-drug resistant subtype B PR, respectively (Figure 5-11A). Another polymorphism, known to occur in more than 95% of the subtype C strains, is L89M. This residue is located in the core of the PR and participates in an extensive network of hydrophobic interactions, about twice the number of interactions in subtype C PR when compared with subtype B PR. H69K is located at the base of the PR, within a loop that exhibited an average r.m.s.d of 1.7 and 1.0 when compared subtype C PR to the wild-type and multidrug resistant subtype B PR, respectively. The I9 3L polymorphism is located in close spatial vicinity of position 69 and, in both subtype B and C PRs, makes numerous interactions with the surrounding residues (Figure 5-11B). Drug-bound HIV-1 subtype C protease Both inhibitor-bound structures of su btype C PR were solved in the P21 space group with two homodimers in the unit cell. As for th e unbound PR, the crystallographic asymmetric unit had a dimer of subtype C PR. In both inhibito r-bound structures the PR exhibited a closed conformation with the inhibitor buried in the active site, covered by the two flaps. Both structures, solved at a resolution of 2.3 re vealed a good quality density map. The solvent for the final model included 169 water molecules and 2 sodium ions for IDV-bound PR and 136 water molecules for NFV-bound PR. The average B f actors for the main-chain of the IDVand NFVbound structures were 28.5 and 31.5 2, respectively (Table 5-3). In each structure the inhibitor, either IDV or NFV, was observed to have a single orientation.
127 The least square supe rimposition of the IDV-bound structur es of our model and subtype B PR (PDB code 1SDT) was performed in the program O (Jones, 1978) and revealed an r.m.s. deviation of 0.30 for the C atoms, with the highest values measured in the outer loops 65-69 of 1.3 (Figure 5-12A,B). In the IDV-bound subtype C PR, the profile of mean main-chain B factors is similar to that of bound subtype B PR indicating a similar beha vior of these two proteins upon inhibitor binding. However, there were several differences in the B value profiles of these two enzymes (Figure 5-13). We observed that monomer A in subtype C PR has significantly higher B factors for the elbow of the flap when compared with the B monomer. Also, the 60s loop in the A monomer of the bound subtype B PR showed highe r B factors when compared to the same region in B monomer. We also noticed that the flap regions in the bound subtype B PR showed a ~2-fold increase in the B values when compar ed with the bound subtype C PR. This is a perplexing finding, as we would expect that the closed flaps, due to de crease flexibility and atomic motion, would exhibit low B values. Th e observed difference between the B factors of these two structures might be due to crystalliz ation conditions, crystal quality and/or crystal packing environment. The bound subtype B PR structure was solved at 1.3 in the P212121 space group while our structure wa s solved at 2.3 in the P21 group. The IDV atoms were in well-defined electr on density, and the s uperimposition of the bound subtype B and C PRs showed that the IDV st ructures superposed ve ry well (Figure 5-14) and the number of subtype C PR IDV van der Waals contacts showed only a small variation among the subtype C and B PR crystal structures (Figure 5-15). Also these crystal structures showed a very similar arrangement of PR-IDV hydrogen bond interactions (H-bonds), including the same water-mediated interactions. Six water molecules that mediate interactions between
128 IDV and the subtype C PR were observed in our structure forming 14 Hbonds, while there are 7 water molecules that form 16 H-bonds in the subtype B PR. The number of H-bonds between the inhibitor and the active site residues is 3 for both structur es compared in our study. In our model, the NFV atoms are in welldefined electron density (Figure 5-16). We attempted to perform the same comparison of the NFV-bound subtype C PR with a similar subtype B PR. To our surprise, despite the fact that NFV was one of the protease inhibitors frequently used to treat HIV-positive patients and NFV-induced drug resistance mutations, such as D30N, L90M, N88D, have been extensivel y studied, only one NFVbound subtype B PR was deposited in the Protein Data Bank (PDB code 1OHR). This structure was deposited in 1997, crystallized in the P212121 space group and was solved at 2.1 resolution. However, when superimposed, the monomer B of the subtype B PR structure showed a spatial shift of ~0.5 relative to our structure, apparently due to a rotational movement at th e dimerization region at the base of the PR. Consequently a corre ct alignment could not be performed. Discussion We were able to obtain crystals of subtype C PR and have solved the first crystal structure of this enzyme, in both unbound and inhibito r-bound forms. The unbound form of the HIV-1 subtype C PR was solved at 1.20 resolution, re presenting the first structure of subtype C PR and the highest resolution solved structure of non-B subtype PR reported to date. Our research is of interest because many in vivo and in vitro studies (Clemente et al., 2006; Gonzalez et al., 2006; Kantor and Katzenstein, 2003; Peeters, 2001; Sanches et al., 2007; Tanuri et al., 1999; Velazquez-Campoy et al., 2003) ad vanced the hypothesis that the naturally occurring polymorphisms play a role in modulating antiretrovira l drug susceptibili ty with the possibility of faster development of drug resistan ce during therapy. In this study we analyzed the structural parameters of the subtype C PR, unbound and complexed with IDV and NFV, and we
129 also compared them to available structures of subtype B PR in an attempt to understand the structural differences due to the baseline polymor phisms and their implications in antiretroviral drug resistance/susceptibility. One of the PR regions believed to be involve d in modulating the affinity of the PR for inhibitors is the flap domain. Understanding the factors underlining the HIV PR flap mobility has profound implications in el ucidating the detailed mechanism of substrate/inhibitor binding of this enzyme and in the design of new therapeutic agen ts such as allosteric inhibitors intended to interfere with the flap opening and thereby w ith enzymatic function. The mechanisms and the factors involved in coordinating and modulating the motion of the flaps have been the focus of study for many researchers. Several studies showed that that flaps open upward and laterally (Ishima et al., 1999; Nicholson et al., 1995; Toth and Borics 2006; Wlodawer and Erickson, 1993) while others argued that th e tip of the flaps curl insi de, making hydrophobic contacts with the several residues located in th e active site (Scott and Schiffer, 2000). It is generally agreed that the large motion of the tip of the flap is accompanied by changes in the hinge and the elbow of the flap as well (Clemente et al., 2004; Pe rryman et al., 2006). Several NMR and molecular dynamics studies investigated the conversion betw een closed, semi-open, and fully open forms of HIV PR. These conformations appear to be in dynamic equilibrium, with the semi-open form being the most prevalent (Freedberg et al., 2002; Hamelberg and McCammon, 2005; Hornak et al., 2006b; Nicholson et al., 1995). The main focus of our study wa s to analyze the structural differences between subtype B and C PRs in an attempt to better understand the role of the naturally occurring polymorphisms. However, our 1.2 resolution structure allowed us to study the thermal motion and anisotropic parameters w ithin subtype C PR and make inferences about the possible mechanism of opening/closing of the flaps. In our structure the amplitude of the
130 atomic motion in the flaps does not appear to be significantly highe r than the core of the enzyme. These data, correlated with relatively low B values for the flaps (Figure 5-7), argue for a limitation of the flap movements, probably due to crystal contacts as has been proposed by Hornak et al. (Hornak et al., 2006a). Among the cr ystal contacts involved in holding the flap open are: H-bonding of the carbonyl of Gly49 w ith side chain amino group of Arg41 and hydrophobic interaction Ile50 with Pro81. In interpreti ng the factors involv ed in keeping the flaps open, we have also to consider the role of the C-shaped density obse rved in our structure. The multi-drug resistant form of subtype B PR co ntains about 100 water molecules in the active site cavity. Martin et al. proposed that these water molecules form a scaffold in the active site cavity, preventing the PR from collapsing in th e absence of a ligand (Martin et al., 2005). Also, an unbound structure of the wild-type subtyp e B PR has been recently released in the Protein Data Bank (code 2PC0) by Heaslet et al. (Heaslet et al., to be published). The structure crystallized in the P41212 space group and has widely open flaps. This new finding argues for crystal contacts having a prominent role in proppi ng open the flaps, when HIV PR crystallizes in P4 or P422 space groups. However, it could be that the unbound PR prefers the open conformation in solution and the pr evalence of this form induces th e enzyme to crystallize in the P4 or P422 space groups and, consequently, the crys tal contacts are formed due to the open form of the PR and are not alone the cau se of the flaps staying open. Other regions of interest th at could further elucidate th e changes in the flaps during binding/releasing the substrate/ inhibitor are the hinge and th e elbow of the flaps. When superimposing the unbound subtype C PR with either wild type or drug resistant subtype B PR, we observed that there are severa l interesting differences. Position 36 occupies a region in the PR that is highly mobile during flap opening and closing in the course of ligand binding. It has been
131 argued that the M36I mutation may promote long-ra nge structural changes in the active site or changes in the flexibility of the PR which may l ead to either the closed or the open conformation of the PR being dominant (Clemente et al., 2004). As seen in our structure and other several previous studies (Clemente et al ., 2004; Martin et al., 2005), Met 36 makes extensive interactions with residues located in the 10s and 60s loops. In subtype C PR the bulkier Met is exchanged for a smaller Ile and consequently there is a decrea se in the number of van der Waals interactions between the 10s and the 60s loops. This effect is augmented by the I15V polymorphisms where a smaller Val replaces the Ile residue. The overall e ffect is a decreased number of interactions between these two loops in subtype C PR, allowing for an increased flexib ility of the elbow of the flap. Previous studies of subtype B and F PRs have also argued that th is polymorphic change causes a collapse of the elbow of the flap, resulting in displacement of the main chain of this loop toward the loop 76 83, stab ilizing the catalytic S1/S1 pockets (Sanches et al., 2007). We observed a similar effect in our structure, wh ere the catalytic residue s Pro81 and Val82 are shifted towards the active site. The amino acid residue at position 89 is lo cated in the hydrophobic co re of the PR and, when mutated to a Met, it makes extensiv e hydrophobic contacts with neighboring residues, more than in subtype B PR which harbors a Leu at this position. Vari ation in the number of hydrophobic residues appears to be important for both maintaining the structural stability of the enzyme and allowing conformational changes. It has been hypothesi zed that the hydrophobic core residues slide by each other, exchanging one hydrophobic van der Waals contact for another, with little energy penalty, while ma intaining many structurally important hydrogen bonds. Such hydrophobic sliding may represent a ge neral mechanism by which proteins undergo conformational changes. Conseque ntly, mutation of these residues in HIV-1 PR would alter the
132 packing of the hydrophobic core, a ffecting the conformational flexib ility of the PR. It has been proposed that these residues impact the dynami c balance between processing substrates and binding inhibitors, and thus any change in this region coul d contribute to drug resistance/susceptibility (FoulkesMurzycki et al., 2007). The incr eased number of van der Waals interactions with the presence of L89M polymorphi sm might increase the stability of the subtype C PR and affect the dynamic properties of the PR and potentially affect its ability to bind inhibitors and substrates. Furt hermore, a previous study hypothesi zed that Met89 was assumed to mimic the role of the L90M mutation, by disp lacing Asp25 and thus c onstraining the S1/S1 pockets (Sanches et al., 2007). The subtype C PR harbors thr ee signature residues: 12, 15, 19 which are located in a sheet that forms what is called the 10s loop. The influence of the polymorphisms in this region was not widely studied, but there are two inte resting observations in our study. First, as mentioned above, the polymorphic change from a larger Ile15 to a smaller Val15 in subtype C PR further reduces the number of interactions between the 10s loop and the elbow of the flap, changing in this way the dynamics of the elbow of the flap. Second, the analysis of the B factors (Figure 5-7) showed that there is a significant difference between the main-chain B values of the 10-22 residues between subtype B and C PRs. In the unbound form of subtype C PR the 10s loop had B factors just below the average B va lue, while the unbound subtype B PR exhibited a ~2.5-fold increase in the B values in this region. This finding leads to the conclusion that the 10s loop is more ordered and probably less flexible in subtype C PR. A similar effect, but of a lesser magnitude, happens for 60s loop. All thes e data taken together, the increased hydrophobic contacts due to L89M polymorphisms and decrease d stability of the 10s and 60s loops, could suggest that, in subtype C, there is an increased stabil ity at the base of the PR. The large number
133 of van der Waals interactions forms a scaffold on which the flaps can swing open easier, with fewer energetic requirements. Also, this arrangem ent could change the size of the active site to an extent where, upon addition of major drug re sistance mutations, the inhibitor binding is hindered, while at the same time maintaining a reas onable affinity for the more flexible substrate. The role of polymorphisms might be that they stabilize the core of the HIV PR while maintaining the flexibility of the flaps, promo ting the open, flexible conformation of subtype C PR. Since inhibitors are rigid and are designed to bind the clos ed conformation, they would preferentially bind to enzymes that carry mutatio ns that favor the closed conformation (Clemente et al., 2004). Consequently, this open conformation of the HI V subtype C PR would be less favorable for inhibitor binding. Thes e results correlate with our kinetic data as well as with other recent structural and kinetic studies (Sanches et al., 2007) that showed that the naturally occurring polymorphisms in subtype F PR might amplify the effect of drug resistance mutations. The overall superimposition of IDV-bound structures of subtype B and C PRs showed that the most significant differences ar e located in the 60s loop, at th e base of the PR, with slight differences in the elbow of the flap. The flaps in both enzymes are firmly closed over the active site enclosing the inhibitor within. The LIGPLOT (Wallace et al., 1995) analysis of the IDV bound in the active site of subtype B and C PRs re vealed that there are no significant differences in the number of H-bonds and hydrophobic interactions between the inhibito r and the active site of the PR. These results correlate well with pr evious kinetic data that showed that the Ki values for IDV for subtype B and C PRs are comparable, indicating that both enzymes bind the inhibitor with similar affinity. Our structural study revealed se veral structural differences between subtype B and C PRs. Even though crystallography offers a static exploration of a structure, we were also able to make
134 several inferences about the dynami cs of the flaps. These results could add to the general effort in explaining if and how the naturally occu rring polymorphisms contribute to the mechanism through which subtype C PR could gain resistance to protease inhibitors. These data and subsequent studi es with other inhibitors will greatly aid in our efforts to understand the influence of baseline polymorphism s in modulating the en zyme sensitivity and resistance to current drug therapy and hopefully provide ne w insight into designing novel inhibitors less likely to promote deve lopment of drug-resistance mutations.
135 A B Figure 5-1. Optical photographs of HIV-1 subtype C protease crystals. A) Crystal of unbound HIV-1 subtype C PR. Approximate dimensions are 0.5 x 0.4 x 0.3. B) Crystal of the IDV-bound subtype C PR. Approximate dimens ions are 0.4 x 0.3 x 0.1 mm. NFV-PR complexes formed crystals similar shape and dimensions.
136 Figure 5-2. Packing diagra m for HIV-1 subtype C prot ease unbound form. Ribbon Packing Diagram in P41 Crystal Lattice. There are four s ubtype C PR homodimers in the unit cell (shown in red and light red, and green and light green). Th e box depicts the unit cell. Figure rendered with Pymol (Delano Scientific). o a c b
137 Figure 5-3. Packing diagram of HIV-1 subt ype C protease IDV-bound form. There are two subtype C PR homodimers in the unit cell (shown in red a nd light red, and green and light green). The box depicts the unit cel l. Figure rendered with Pymol (Delano Scientific).
138 A B C Figure 5-4. Ramachandran diagrams. A) U nbound subtype C PR. B) IDV-bound subtype C PR. C) NFV-bound subtype C PR. The color code is as follows: pink most favored regions; yellow additional allowed regions, gray disallowed regions. Plots created using Coot (Emsley and Cowtan, 2004).
139 Figure 5-5. The 2Fo-Fc electron density map of the unbound subtype C protease. The map is contoured at 3 using data to 1.2 resolution. Figure rendered with Pymol (Delano Scientific). Ala37 Leu38 Ile36
140 A B Figure 5-6. The C-shaped electr on density between the flaps. A) Ribbon representation of the subtype C PR with the naturally occurring polymorphisms showed as red spheres. The alternative structures of the Triton X 100 fragments are shown in light and dark yellow between the flaps of the PR. B) Cl ose-up view from above the PR molecule. The electron density map is contoured at 2 level, using data to 1.2 resolution. Figure rendered with Pymol (Delano Scientific). 3.7 4.2 G51 G51 37 37 36 36 12 12 1 5 1 9 1 9 89 93 93 69 69
141 Figure 5-7. The normalized mean B values for the main chain atoms of the wild type subtype C and multi-drug resistant subtype B proteases. The normalized B values are plotted for the residues of subtype C PR (red), multi-dr ug resistant B PR (gr een) and wild type subtype B PR (blue). Normalization was done by dividing the B values for the main chain of each residue by the average B va lues for the entire PR molecule. The residues in the two subunits ar e labeled 1-99 and 1-99. 0.00 0.50 1.00 1.50 2.00 2.50 3.00 020406080020'40'60'80' Monomer Monomer (') Residue #Normalized B factors (MC/meanMC) wt C PR mut B PR wt B PR
142 A B Figure 5-8. Thermal ellipsoid diagram for unbound HIV-1 subtype C protease. A) Thermal diagram representing the anisotropy for the 42-59 residues within the flaps, including the side chains. B) Thermal diagram representing the anisotropy for the main-chain atoms of the 42-59 residues within the flap s. C, O, N and S atoms are colored gray, red, blue and yellow, respectively. The a rrows indicate the direction of the flap motion. Figure created using RASTEP (Merritt, 1999a; Merritt and Bacon, 1997). I50 I50 M46 M46 F53 I54 G51 G49 G49 G51 I54 K43 R57 G52 P44 K43 W42 R57 V56 W42 V56 F53 I50 G49 K43 P44 V56 W42 G51 I54
143 A B Figure 5-9. Superimposition of the wild type s ubtype C protease with the wild type and multidrug resistant mutant of subtype B protease. A) Ribbon representations of wild type subtype C PR (red) superimposed over wild type subtype B PR (blue) or multidrug resistant mutant B PR (green). The natu rally occurring polymorphisms in subtype C PR are represented as red spheres. B) Th e r.m.s differences () per residues are plotted for the C atoms of wild type subtype B PR (blue) and mutant B PR (green) compared with the wild type subtype C PR. The residues in the two subunits are labeled 1-99 and 1-99. 0 1 2 3 4 5 6 7 8 9 020406080020'40'60'80'Monomer Monomer (') Residue #R.M.S.Deviation mut B PR wt B PR
144 A B Figure 5-10. The comparison between the flaps of subtype C and B proteases. A) Ribbon representation of the flap re gions of the wild type subt ype C PR (red) superimposed over the wild type subtype B PR (blue). B) Ribbon representation of the flap regions of the wild type subtype C PR (red) s uperimposed over the mutant subtype B PR (green). The gray surface represents th e active site. Figure rendered with Pymol (Delano Scientific). 12.2 4.4 12.5 12.2
145 A B Figure 5-11. The naturally o ccurring polymorphisms in s ubtype C protease. A) Ribbon representation of the flap and elbow regions of the wild type subtype C PR (red) superimposed over the wild type B PR (b lue). B) Ribbon representation of the 60s loop and residues 89 and 93 of the wild type subtype C PR (red) superimposed over the wild type B PR (blue). The amino acid residues are represented as sticks. Figure rendered with Pymol (Delano Scientific). 37S 37A 36M 36I 15I 15V 19L 19I 4.4 3.4 69K 69H 93L 93l 89L 89M 1.8
146 A B Figure 5-12. Superimposition of the IDV-bound s ubtype C protease with the IDV-bound subtype B protease A) Ribbon re presentations of IDV-bound subtype C PR (red) superimposed over IDVbound B PR (gray) The naturally occurring polymorphisms in subtype C PR are represented as red s pheres. B) The r.m.s differences () per residues are plotted for the C atoms of bound B PR co mpared with the bound subtype C PR. The residues in the two subunits are labeled 1-99 and B1-99. 36 36 37 37 19 12 69 93 93 69 89 15 0 0.5 1 1.5 2 2.5 020406080020'40'60'80' Monomer Monomer (') Residue #R.M.S.Deviation
147 Figure 5-13. The normalized mean B values for the main chain atoms of the IDV-bound subtype C and B proteases. The normalized B values ar e plotted for the resi dues of subtype C PR (red) and B PR (gray). Normalization was done by dividing the B values for the main chain of each residue by the average B values for the entire PR molecule. The residues in the two subunits ar e labeled 1-99 and 1-99. 0.0 0.5 1.0 1.5 2.0 2.5 020406080020'40'60'80' Monomer Monomer (') Residue #Normalized B factors (MC/meanMC) C PR B PR
148 A B Figure 5-14. IDV in the active si te of HIV protease. A) Electr on Density Map for IDV in the Bound Subtype C PR Crystal Struct ure. The contour level is 2 The pyridyl group of IDV is labeled. B) Stick representation of the IDV in the active site of the subtype C (red) and B (blue) PRs. Ball-and-stick re presentation of the amino acid residues in the subtype C (red) and subtype B (blue) surrounding the inhibito r. Figure rendered with Pymol (Delano Scientific). I47 V82 I84 V32 T80 I50 D25 T26 G27 G49 I47 V32 I84 D25 pyridyl group
149 A Figure 5-15. Ligplot analysis. A) Schematic representation of the H-bonds and hydrophobic interactions of the IDV bound in the activ e site of subtype C PR B) Schematic representation of the H-bonds and hydrophobic interactions of the IDV bound in the active site of subtype B PR. Figure made with LIGPLOT (Wallace et al., 1995).
150 B Figure 5-15. Continued
151 Error! 3-hydroxy-2methyl-benzoyl group decahydroiso quinoline group Figure 5-16. Electron density map for NFV in the bound subtype C protease crystal structure. The contour level is 2 Figure rendered with Pymol (Delano Scientific).
152 Table 5-1. Data collection statistics Subtype C protease with indinavir (IDV) Subtype C protease with nelfinavir (NFV) Unbound subtype C protease Space group P21 P21 P41 Unit-cell parameters (, ) a =46.7, b =59.5, c =87.4, =95.6 a =46.7, b=60.1, c =86.7, =94.7 a =46.7, b =46.7, c =100.8, =90.0 Unit cell volume (3) 241,915 242,491 219,737 Vm (3 Da-1) / solvent fraction (%) 2.90 / 55 2.90 / 55 2.56 / 52 Total reflections 65956 61466 279351 Unique reflections 21311 21028 35611 Crystal mosaicity () 0.9 0.9 0.4 Resolution range () 20.0 2.3 (2.38 2.3) 20.0 2.3 (2.38 2.3) 30.0 1.2 (1.24 1.2) Completeness (%) 96.7 (96.4) 97.3 (93.7) 99.3 (95.7) Rsym # (%) 8.4 (33.2) 11.3 (37.3) 7.6 (27.3) Redundancy 3.1 (3.0) 2.9 (2.6) 7.9 (4.7) Average I/ (I) 8.4 5.6 14.0 I/ (I) > 3 (%) 71.1 (36.8) 59.7 (21.0) 76.0 (41.7)
153 Table5-2. Refinement steps in SHELX for the unbound form of HIV-1 subtype C protease Job Action taken NP NH NW/NW1/2/NX Npar Rwork Rfree 1 First refinement in SHELX 1508 0 358 / 0 / 24 6855 23.8 26.4 2 Protein atoms only anisotropic 1560 0 306 / 0 / 36 15699 14.9 18.3 3 More solvent added 1560 0 250 / 25 / 56 15655 15.1 18.3 4 Minor adjustments 1560 0 250 / 25 / 56 15655 15.1 17.9 5 Riding hydrogens added 1560 1618 250 / 69 / 56 15831 13.9 16.9 6 All atoms anisotropic 1560 1618 243 / 94 / 44 17472 12.4 15.3 7 More disorder 1578 1618 237 / 93 / 48 17625 12.4 15.5 8 Minor adjustments 1578 1618 237 / 121 / 36 17661 12.9 16.5 9 Minor adjustments 1582 1618 241 / 116 / 24 17670 12.2 15.6 10 Minor adjustments 1583 1618 236 / 110 / 24 17580 12.1 15.8 11 Minor adjustments 1587 1618 231 / 109 / 24 17562 12.2 15.8 12 Fragments of TX-100 added 1583 1618 231 / 101 / 50 17688 12.2 15.8 13 SIMU changed from 0.1 to 0.05 1583 1618 243 / 98 / 50 17769 12.2 15.7 14 Minor adjustments 1583 1618 227 / 114 / 50 17652 12.2 15.6 15 WGHT changed from 0.1 to 0.2 1583 1618 243 / 98 / 50 17769 12.3 15.5 16 Further refinement 1583 1618 243 / 98 / 50 17895 12.3 15.5 17 100% of data refined 1583 1618 243 / 98 / 50 17895 12.3
154 Table 5-3. Refinement statistics. Note: Rwork = [ (Fo-Fc)/ Fo]x100; Rfree is identical to Rwork for 5% of data omitted from refinement. Subtype C protease with IDV Subtype C protease with NFV Unbound subtype C protease R values (%) Rwork 21.1 25.3 12.3 Rfree 23.8 28.4 15.5 No. of water molecules 169 136 341 No. of inhibitors per PR dimer 1 1 0 Ramachandran statistics (%) Most favored regions 96.2 95.9 94.9 Allowed regions 3.8 4.1 5.1 B factors (2) Average main chain atoms 28.5 31.5 13.3 Average, side-chain atoms 27.7 53.1 15.3 Solvent 35.9 32.8 29.1 Inhibitor 27.5 27.5
155 CHAPTER 6 ANALYSIS OF THE PROCESSING EVENTS OF HIV-1 SUBTYPE A, B, AND C GAG POLYPROTEINS Introduction Gag and gag/pol processing, an important step in the HIV life, is mediated by the viral protease (PR) and occurs at the plasma membrane of the infected cells during budding and release of progeny virions (Kapla n et al. 1994). It has been show n that this is a sequential process, regulated by the order of the cleavage and the rates of the pr oteolytic processing at individual cleavage sites (Wiegers et al. 1998). An effective appr oach to prevent HIV maturation and spread from cell to cell has proven to be the inhibition of gag a nd gag/pol cleavage. Several studies (Henderson et al. 1990; Carter and Zybarth 1994; Pettit et al. 1994; Louis et al. 1999; Goodenow et al. 2002; Iga et al. 2002; Liao et al. 2004; Liao and Wang 2004; Pettit et al. 2004) conducted in a number of in vivo and in vitro systems, have examined the order of cleavage for different sites within the HIV-1 ga g/pol precursors. There is a general agreement that in HIV-1 subtype B the first cleavage o ccurs between the p2 spacer peptide and the NC protein. Intermediate cleavages occur at the matrix/capsid (MA/CA) and transframe/p6POL (TF/p6POL) sites (Pettit et al. 2003) (Figur e 6-1). Not only the order of cleavage but also the rates of cleavage at different sites have been evaluate d and they are estimated to vary by as much as 400-fold between sites (Pettit et al. 1994). Baseline polymorphisms or mutations arising in response to protease inhibitor (PI) therapy can affect the order and the rates of cleavage. St udies have shown that resistance mutations in the protease of subtype B virus are associated with impaired proteolytic processing and decreased enzymatic activity (Zennou et al. 1998) and that compensatory mutations within gag and gag/pol cleavage sites can partially overcome these def ects (Mammano et al. 1998). Furthermore, recent in vitro studies showed not only that the HIV-1 PR embedded within the gag/pol polyprotein
156 seems to be less sensitive to the effect of clinic ally used PIs than is the processed PR, but also that the addition of PIs might pr omote a differential cleavage site affinity (Pettit et al. 2004). Detailed understanding of these events would make possi ble designing new drugs that would prevent gag/pol processing and thus the virus spreading from cell to cell. While a relatively larg e body of literature describes and interprets data about subtype B gag and gag/pol processing, there are no data ab out the order or the rate of gag and gag/pol cleavage in HIV-1 subtype C or any other subtypes. Sequence analysis of the cleavage sites within the gag and gag/pol precu rsors showed that these sites are relatively well conserved for HIV-1 group M subtypes. In a recent study perfor med on 84 full-length nucleotide sequences (de Oliveira et al. 2003) it has been shown that overall, 58.3% of the 12 HIV-1 cleavage sites are significantly more diverse in C th an in B viruses. The same study supports the theory that these cleavage sites have evolved in a subtype-specifi c manner, and that the variation of subtype C cleavage sites began early, prior to th e diversification of HIV-1 subtypes. There are sequence differences at the level of cleavage site among different subtype polyproteins studied. It has been shown that the sites situated upstream of the protease domain directly influence both the (auto) processing activity of PR a nd the order and efficiency of cleavage of gag and gag/pol precu rsors (Pettit et al. 1994; Zyba rth et al. 1994; Zybarth and Carter 1995; Louis et al. 1999). Natural variation at p2/nucleo capsid (p2/NC) cleavage sites may play an important role, not only in regulation of the vira l cycle but also in disease pr ogression and response to therapy (de Oliveira et al. 2003). Also, HIV-1 subtype C can be considered to be natural and viable variants of subtype B viruses. More data abou t processing events in subtype C would allow for evaluation of efficiency of current therapy and also would contribu te to the effort of finding new
157 drugs to inhibit this process. The identification of common patterns may facilitate the development of broad-based inhibitors with in creased specificity and improved binding to the mutated PR. These secondary inhibitors may pr eempt or delay the emergence of resistance. These studies will add to our growing understa nding of interaction within the gag/pol polyprotein pre-processing structure that will be an excelle nt target for new drug discovery. In this study we aim to analyze the proces sing events in HIV-1 s ubtype C and subtype A gag polyproteins and to bring qua litative and quantitative information regarding the order and the rate of processing in subtype C and subtype A gag polyproteins. We performed similar analysis for HIV-1 subtype B gag polyprotein and the data obtained were compared and interpreted. Our goals were to understand whether the naturally occurring polymorphisms within the PR or/and the cleavage sites modulate the processing in HI V-1 subtype C and subtype A gag polyproteins. Overall, the results of proposed experiments w ill aid in developing more effective therapeutic approaches for patients infect ed with subtype C viruses. Results Polymorphic Sites within HIV-1 Subtyp e A, B, and C Gag Polyproteins The alignment of the three ga g polyproteins was performed w ith ClustalW software. The HIV-1 subtype B gag polyprotein contains 512 amino acid residues, while HIV-1 subtype C and A gag polyproteins have only 491 and 496 amino acid residues, respectively. The three gag sequences had 74.7% identity, 9.9% strong si milarity, 4.1% weak similarity and 11.3% difference (Figure 6-2). The most conserved regions were capsid (CA) (with 95.2% identity and strong similarity and only 1% difference), nucle ocapsid (NC) (with 92.7% identity and strong similarity and 5.4% difference), and p1 (with 94.7% identity and strong similarity and 5.3% difference). The most polymorphic regions were p2 (with 53.3% identity and strong similarity and 33.3% difference) and p6 (with 47% identity and strong similarity and 45.5% difference).
158 Analysis of the matrix (MA) protein yielded intermediate va lues: 84.1% identity and strong similarity and 13.6% difference. Several small gaps and inserts within A and C gag sequences were observed as well. We considered gaps to be amino acid residues missing within subtype A and C gag polyproteins when compared with su btype B gag, used as a reference sequence. Inserts are defined as extra amino acid residues within subtype A and C gags when compared with subtype B gag. The largest gaps (3 ami no acid residues) were f ound in both subtype C and A gag polyproteins and they were located in the close vicinity of the MA/CA and p2/NC cleavage sites. We also noticed that there is a duplication of the FLQSRPEPTAPP sequence within p6 of subtype B gag polyprotein. It has been shown that the PTAP motif binds the host cellular factor TSG101 and is requir ed for efficient virus release (G ottlinger et al. 1991; Huang et al. 1995; Yu et al. 1998; Garrus et al. 2001; Demir ov et al. 2002). It is worth mentioning that p1 protein in al l three subtypes has all the Pro residues well conserved. Recent studies showed that these resi dues are important for the regulatory role of p1 in maintaining the viral infectivity (Hill et al. 2002). Cleavage Site Analysis There are 5 cleavage sites within gag pol yprotein: MA/CA, cap sid-p2 (CA/p2), p2/NC, nucleocapsid-p1 (NC/p1), and p1-p6 (p1/p6) (Figure 6-3). The cleav age sites were defined as being 10 amino acid residues long, with 5 amino acids on each side of the scissile bond. Two of the cleavage sites, CA/p2 and NC/p1, are very well conserved among subtype A, B, and C gag polyproteins, having 100% sequence identity. The MA/CA cleavage site is identical between subtype B and C gag polyproteins and differs in only one amino acid residue at position P5 (serine valine) in subtype A gag. The p2/NC cleavag e site differs by three residues, located at P3, P4, and P5, while the p1/p6GAG site differs by three residues as well, but these are situated on the carboxyl side of the cleavage site at P1, P3, and P5 positions.
159 In Vitro Processing Studies of HIV-1 Subtyp e B, A, and C Gag Polyproteins To study the influence that the naturally o ccurring polymorphisms have on the rate and order of cleavage in subtype A, B, and C gag a nd gag/pol polyproteins processing, we employed an assay in which labeled gag pr ecursors is processed sequentially in vitro after the addition of HIV-1 PR (Erickson-Viitanen et al 1989; Krausslich et al. 1989; Part in et al. 1990; Tritch et al. 1991; Pettit et al. 1994). The newly synthesized proteins are labeled with radioactive 35[S] which allowed visualization by autoradiography. The ba nds were recorded using a Molecular Dynamics PhosphorImager Storm 860, and the intensity of each band was measured with ImageQuant software (GE Healthcare). The calculations we re made based on the known composition of the proteins and the number of labeled methionine (Met ) residues of each protein. The total intensity of all the bands of interest in a given lane was considered 100 and the amount of each product was calculated as percentage of the total amount of labeled substrate in the lane. Expression of the HIV-1 subtype B, A, and C gag polyproteins in an in vitro rabbit reticulocyte lysate transcription-translation system resulted in four major bands with sizes of 55, 41, 25, and 15-17 kD observed on the autoradiogra phs. The 55 kD protein corresponds to the gag precursor. The other 3 bands correspond to the in termediate cleavage products: MA-p2, CA-p2, and NC-p6GAG. In the time frame of our experiments, we were not able to detect the synthesis of entire gag/pol polyprotein (p160) indicating that no active PR, able to carry out the cleavage of gag and gag/pol polyproteins, was synthesized. The transcription-tran slation reaction was followed for 5 hours. The p55 polyproteins cont inue to accumulate but no processing intermediate products were observed (Figure 6-4). The processing of gag polyproteins was accomplished by addition in trans of the active PR that was separa tely expressed in our bacterial system, purified and activated as described previous ly in Material and Methods chapter, page 63.
160 These experiments yielded protei ns with a large range of size s, from 55 kD to less than 2 kD. To be able to capture the en tire picture we used a gradient (4-20%) gel that allows for a good separation. The top layers, with a lower percent of polyacrylamide, trapped and resolved the higher molecular weight proteins while the botto m layers with an increased concentration of polyacrylamide allowed for separa tion of smaller protein. The gels used for this study are denaturant gels so the proteins migrate according to their monomeric molecular weight. In trans processing of HIV-1 subtype B, A, a nd C gag polyproteins by HIV-1 subtype B protease The addition in trans of the active HIV-1 subtype B PR re sulted in ordered processing of the subtype A, B, and C gag (p55) precursors (Fi gure 6-5). As expected, the initial cleavage was observed between the NC and p2 proteins and yi elded two intermediates: MA-p2 (p41) and NCp6GAG (p15). The cleavage of this site proceeded at an increased rate with less than 5% subtype B p55 remaining after 5 min of processing. The cleava ge at the corresponding site within subtype C gag polyprotein occurred at a slightly faster rate, with no p55 remaining after 5 min. The second cleavage occurred at the site be tween the MA and CA (Pettit et al. 2003) and the cleavage product CA-p2 (p25) can be easily visu alized as a intense band of about 25 kD size for all three subtypes. Th e production of p25 increased steadily during the in vitro transcriptiontranslation reaction. However, the amount of p25 was markedly different between the three subtypes analyzed in this study. The p25 quantities for subtypes B and C were roughly similar, with about 10% more subtype C p25 than s ubtype B p25 after one hour of processing. Interestingly, the production of subtype A p25 o ccurred at a slower rate indicating a reduced cleavage rate at the MA/CA cleavag e site. The densitometric analysis showed that, after one hour of processing, the amount of subtype A p25 was 15fold less than that of subtype B or C p25.
161 The subtype A and B MA proteins do not contain Met residues so they cannot be identified on the gels. However, the subtype C MA protein contai ns one Met residue and it can been seen as a faint 17 kD band after 1 min and as a more intense band at the end of the experiment (60 min). The subtype B NC-p6GAG intermediate has a duplicati on of the LQSRPEPTAPP sequence within the p6GAG. Consequently the subtype B NC-p6GAG intermediate migrates on the gel at a slightly different position when compared to the intermediate in subtypes A and C. Worth mentioning is that the NC-p6GAG intermediate products for all th ree subtypes steadily decreases over time, probably due to the HIV PR-mediated cleavage at the p1/p6GAG cleavage site. However, the rates of cleavage at this site s eemed to be different between the three subtypes studied here, with the rate of cleavage for s ubtype C intermediate slightly higher than for subtypes A and B. Also, the amounts of subtype C NC-p6GAG intermediate synthesized at both the beginning and the end of th e experiment are lower when co mpared with subtypes A and B. Due to their low molecular weight and lack of Me t residues, we have not been able to follow the accumulation of the cleava ge products at the p1/p6GAG cleavage site: NC-p1 and p6GAG. In trans processing of HIV-1 subtype B, A, a nd C gag polyproteins by HIV-1 subtype A protease The HIV-1 subtype B, A, and C ga g polyproteins were also cleaved in trans by addition of HIV-1 subtype A PR. The pattern of cleavage is similar to that observed after the addition of subtype B PR. However, the cleavage rates at th e first cleavage site ar e slightly lower when compared to those performed by subtype B PR, shown by the retarded di sappearance of p55, as seen on densitometric analysis (F igure 6-6). Despite this fact, th e final amounts of p25, after one hour of processing, were 3-5-fold higher when compared with those produced by subtype B PRmediated gag cleavage. Interestingly, the pr oduction of subtype A p25 was still lower when
162 compared with the amount of subtype B or C p25. However, the quantitative difference was only 4-fold less subtype A p25 than subtype B or C p25 compared with 15-fold difference observed in the subtype B-PR mediated cleavage. In trans processing of HIV-1 subtype B, A, a nd C gag polyproteins by HIV-1 subtype C protease The cleavage of all three different gag polypr oteins by HIV-1 subtype C PR occurred in a similar pattern to cleavages mediated by the other enzymes, subtype A and B PRs, with a reduced production of subtype A p25 (12-fold lo wer) when compared with subtype B and C p25. Analysis of MA/CA Cleavage Site The MA/CA cleavage site appears to be among th e sites cleaved immedi ately after the first cleavage event. The P5 position in this site harbors a valine (V al) residue for subtype B and C gag polyproteins and a serine (Se r) residue for subtype A gag. Fu rthermore, when an alignment of the cleavage site is performed, a gap of 2 and 3 residues for subtype A and C gag polyproteins, respectively is noted (Figure 6-3). In Vitro Processing Studies of HIV-1 Subtype A and B Gag Polyproteins 124 and QV Variants The mutants for both subtype A and B gag pol yproteins were engi neered using Site Directed Mutagenesis technique. Th e order and the rates of gag pro cessing were studied after the addition of HIV-1 subtype B PR in the reaction mixture, as described above. In trans processing of HIV-1 subtype A gag poly protein S124V and QV variants by HIV-1 subtype B protease The first step was to mutate the MA/CA cleav age site within subtype A gag polyprotein to engineer a site identical to that in subtype B gag. Because only one residue was different between subtype A and B gag pol yproteins, one change was intr oduced: the Ser found at position 124 in subtype A gag was mutated to the corr esponding amino acid in subtype B Val; the
163 mutant was named AgagS124V. We decided on anot her approach as well and engineered a tworesidue insert (glutamine -Val ) that would cover the gap obser ved in the subtype A gag when compared to subtype B gag, at the amino-te rminus position of the MA/CA cleavage site; the mutant was named Agag_QV. The introduction of the S124V mutation at the P5 position of the MA/CA cleavage site increased the rate of processing at this site by a factor of 2, with twice as much p25 produced when compared to V124 subtype A gag (Figur e 5-8). A similar effect was observed for Agag_QV mutant. However, the am ount of p25 generated by these tw o mutants is not equivalent to that of subtype B gag, bot h being about 7-fold less. In trans processing of HIV-1 subtype B gag polyprotein V124L, V124S and QV variants by HIV-1 subtype B protease We also decided to mutate the MA/CA subt ype B cleavage site to mimic the site in subtype A gag polyprotein. We followed the same steps as when we engineered the subtype A gag mutants but backwards. We constructed thre e variants: (1) the amino acid Val at position 124 was substituted by Ser (BgagV124S); (2) two resi dues, glutamine (Gln) and Val, located at the amino-terminus position at the MA /CA cleavage site were removed (Bgag_ QV); (3) the amino acid Val at position 124 was substituted with Leu (BgagV124L). The last mutant harbors a conservative change and it would be useful to further analyze the importance of a hydrophobic amino acid at the P5 position of the cleavage site. The V124L mutations did not appear to affect the rate of cleavage at the MA/CA site, while the V124S decreased the amount of p24 by 6-fold (Figure 6-9). The deletion of QV dipeptide had similar results. However, the produc tion of p24 is at least 2-fold higher than in subtype A gag polyproteins harboring similar ch anges at the N-terminus of the MA/CA site.
164 Discussion To our knowledge there are no other studies analyzing the gag a nd gag/pol processing events in the subtypes A and C. This work represents the first attempt in studying and characterizing the effect of naturally occurri ng polymorphisms located at the cleavage and noncleavage sites in gag and gag/pol polyproteins. In our study we aim to evaluate the phenotypic impact of gag baseline polymorphi sms on HIV-1 gag processing events. The naturally occurring polymor phisms within the gag polyprot ein can be classified as cleavage site and non-cleavage site polymorphisms. The former ones have received a great deal of attention and many studies have demonstrated that some of the amino acid variations within the cleavage site of gag are baseline polymorphi sms while others develop in response to PI treatment and evolution of a catalyt ically impaired PR. It has been previously noted that some of the most frequently mutated positions under PI therapy also happen to be more polymorphic, such as positions P2 to P5 and P3 of the p2/NC cleavage site or position P3 of the p66/INT site (Cote et al. 2001). One study suggest ed that, in subtype B gag pol yprotein, the development of gag cleavage site mutations is associated with heavily mutated PRs (dead end) for which the concomitant evolution of additiona l mutations in the PR and in th e gag substrate may be the only way for the virus to survive in an increasin gly selective environm ent (Doyon et al. 1996; Maguire et al. 2002). More recently, the analysis of resistant vi ral isolates from IDV-treated patients indicated gag adaptation as a common evol utionary pathway, taking place as early as 6 weeks after the start of therapy a nd in the presence of as few as two PR mutations (Zhang et al. 1997). These findings emphasize the relevance of cleavage site mu tations in the evolution of HIV resistance to PIs. Therefore, it was propos ed that the therapy-associated cleavage site mutations should be considered in HIV resistance tests to estimate viral fitness in different clinical settings (V erheyen et al. 2006).
165 The non-cleavage site polymorphisms have also been indicated as possible mechanisms of adaptation for compensating for the reduced cataly tic activity of mutant PRs. Moreover, certain amino acid substitutions in non-cleavage sites have been shown to contribute to the development of high levels of viral resistance to multiple PI s. These data strongly suggest that non-cleavage site amino acid substitutions in the gag protein recover the reduced replicative fitness of HIV-1 caused by mutations in the viral PR and may open a new avenue for designing PIs that resist the emergence of PI-resistant HIV-1 (Gatanaga et al. 2002). Such non-cleavage site gag mutations should render the polyprotein cleavage si tes more accessible to the PR, make polymerization of viral proteins more efficacious, and/or make assembly and disassembly more efficient. Alteration(s) of other unknown functions of gag prot eins may also contribute to the HIV-1 acquisition of resist ance to PIs, but it appears that HIV-1 resistance to PIs is acquired with multiple mechanisms (Gatanaga et al. 2002). Previous studies observed for the first time K416N and E399Vsubstitutions in the NC protein, a nd E455A and E455V substitutions in the p6GAG protein in resistant viruses suggesting that th ere are additional possibilities for gag adaptation besides the previously described substitutions in the cleavage sites surrounding the p2, NC, p1 and p6GAG peptides (Mammano et al. 1998). Analysis of consensus and individual seque nces of the non-B subt ype gag polyproteins showed an increased number of polymorphic residues when compar ed with subtype B gag (Los Alamos 2007). The subtype A and C clinical isolates analyzed in this study showed cleavage site sequences that sometimes differed considerably from the consensus sequences in subtype B gag after which HIV PR peptide substrates are of ten designed (Figure 6-3) None of the cleavage site polymorphisms present at the cleavage site that have been previ ously described as drug resistance-associated mutations were obser ved within our HIV-1 subtype A and C gag
166 sequences. Also these clones do not harbor any of the previously described non-cleavage site substitution, except E481K, a baseline polymorphi sm in subtype C gag polyprotein, that appears to enhance the replicative capac ity of the virus (ref), and L496P, a polymorphism in subtype A gag, that arises as a post-PI treatment mutation in patients harboring HIV1 subtype B (Ho et al. manuscript in preparation). The subtype A and C gag polyproteins analyzed in this study also harbored several small gaps and inserts (Figure 6-2). Two small gaps ar e located at the amino-terminus of the MA/CA cleavage site, in the carboxyl-termi nus of the MA. Several groups de termined the structure of the MA protein through NMR analysis (Massiah et al. 1994; Matthews et al 1994). They observed that the C-terminus 20 amino acid residues do not adopt a rigid conformati on in the solution. It appears that the C-termini of the monomers form helical arms that fit into a cleft between the helices A and C of the adjacent monomer, a common feature in the viral architecture (Stuart et al. 1993). Consequently, the small gaps within this region do not interfere with a secondary structure but might affect the MA/MA interactio ns, that are important for both gag and gag/pol dimerization and for the formation of the viral MA shell. Another gap is located within the p2/NC cleavage site of subtype C gag, at the N-terminus of the p2 protein. However, as discussed later in this chapter, there appear to be no consequences of this gap on the rate of the order of gag processing. The FLQSRPEPTAPP duplication wa s observed within subtype B p6GAG protein. The influence of this duplicat ion on the results of our in vitro experiments is not clear. Previous studies showed that the complete or partial dupli cation of PTAP appears to provide an advantage for survival of the virus in the presence of NRTIs and PIs (Ibe et al 2003). Insertions such as the PTAP repeat or SRPE in p6 reduce the enzyma tic activity of the wild-type HIV-1 PR, while
167 restoring gag processing and s ubsequent replication of protease inhibito r-resistant variant (Tamiya et al. 2004). A truncated p6 did not inhi bit processing of p160 ga g/pol in COS-7 cells, while Demirov et al. showed that in certain T cell lines, mutation in the PTAP motif did cause processing defects (Yu et al. 1998; Demirov et al. 2002). All of these studies were performed in a cellular system where the budding a nd releasing of whole virions was analyzed. In the cell-free context of our experiment set-up, we can fo llow neither the assembly nor the budding. We believe the influence of the duplic ation over the rate and order of the processing is minimal, if any, in our system. We analyzed the processing events in gag polyprotein cleavage by the PR from the same subtype, but also cleaved by the other two subtyp e PRs. This kind of subtype interplay would occur in the case of a recombinant form, wher e two different subtypes recombine and form a virus with a new genetic makeup (T akeb et al. 2004). This approach enabled us to evaluate the cleavage that occurs in the recombinant fo rms and would also bring information about compatibility between combinations of naturally occurring polymorphisms within the PR and within the gag polyproteins. We observed that the order of gag cleavage is similar between all three HIV-1 subtypes analyzed in this study. As expect ed, upon addition of active HIV PR in trans the first cleavage occurred at the p2/NC cleavage site closely fo llowed by MA/CA cleavage. When the same gag polyprotein was cleaved in trans by PRs from different subtypes the pattern of cleavage was highly similar. However, the same PR rendered di fferent rates of cleavage for different cleavage sites. It appears that subtype B PR cleaves the p2/NC site well and has a slightly lower MA/CA cleavage rate. Subtype A and C PR s exhibited a relatively slower cleavage rate for p2/NC when compared to subtype B PR, but they showed an increased MA/CA cleavage rate.
168 Independent of the PR added in trans the p2/NC cleavage appears to occur at a reduced rate for subtype A gag when compared to B a nd C gag polyproteins. The p2/NC cleavage site is very polymorphic among these three gag polyprotei ns, each having a unique combination at the P3-P5 positions. In the same time, we observe d a marked reduction in the amount of p25 produced by in trans processing of subtype A gag. Even when processed by the subtype A PR, the amount of subtype A p25 is 4-fold less than the amount of p25 that resulted from the cleavage of subtype B gag polyprotein by the su btype A PR. The slightly decreased p2/NC cleavage rate in subtype A gag polyprotein might be due to seve ral factors: (1) the baseline polymorphisms at the p2/NC site, (2) the baselin e non-cleavage site pol ymorphisms within p2, NC or the other structural proteins within subt ype A gag, and/or (3) negative feed-back from the MA/CA site. With so many baseline variations it might be difficult to pi npoint the exact source of this difference in the p2/NC cleavage rate among subtypes, and probably would not be feasible to pursue this avenue due to the f act that such small difference might not have significant influence on the virus re plicative capacity or PI treatment susceptibility. However, we believe that it is important to mention that even if small, these differences in cleavage rates might increase upon acquisition of other ARV treatment-induced mutations, and the overall effect on the virus could be significant. As mentioned above, the cleavage of the MA /CA within subtype A gag proceeded at a much slower rate when compared with thos e for subtype B and C gag polyproteins. Upon a closer analysis of the amino acid sequence, we noticed that there is one residue difference between subtype A and the other two subtypes: the Val residue f ound at the P5 position within subtype B and C is mutated to a Ser in subtype A gag.
169 In order to understand the role of this cleavage site polymorphism on the cleavage rate at MA/CA site, we created a subtype B MA/CA cleavage site within subtype A gag. Our rationale was that if the cleavage site sequence has the domi nant influence on the cleavage rate, the rate of cleavage of the mutated subtype A MA/CA site w ould be comparable with those the subtype B. If there are other determinants within gag polyprotein, most like ly the rates of cleavage at the MA/CA site would be different between thes e two subtypes. We engineered two mutants designated AgagS124V and Agag_QV, where the Ser was mutated to a Val, and the Gln-Val dipeptide was inserted at the N-terminus of the cleavage site (Figure 6-3). The in vitro transcription-transla tion analysis was conducted as pr eviously described, and the in trans cleavage was performed by the subtype B PR. Th e amount of p25 increased by 2-fold for the subtype A gag mutants but did not equal the p25 production upon cleavage of subtype B gag. We believe that this might be due to non-cle avage site polymorphisms present within gag polyprotein. We also engineered a subtype A MA/CA cleavag e site within subtype B gag. We obtained three mutants BgagV124L, BgagV124S and Bgag_ QV, as explained in the Results section of this chapter. The production of p25 was not affected by the V124L mutation (Figure 6-9), but the cleavage rate at the MA/CA cl eavage site decreased significa ntly upon introducing V124S or deleting the Glu-Val dipeptide. After 90 min, the amount of p25 for BgagV124S was twice as much as for subtype A gag and 3-times less wh en compared to subtype B gag. The p25 amount in Bgag_ QV mutant was slightly great er than for BgagV124S. These results point to the fact that the base line polymorphisms with in gag polyprotein play a dominant role in modulating the ra te of cleavage. In this case, it appears that rate of cleavage is modulated by both the cleavage and the non-cleava ge site polymorphisms. Our results indicate
170 that the presence of hydrophobic re sidue at position P5 of the MA /CA site appears to be a requirement for an enhanced cleavage rate. These results add to previous studies that suggested that the MA/CA site might be more sensitive to mutations than other pr ocessing sites (Margolin et al. 1990; Partin et al. 1990; Pettit et al. 1994). The importan ce of the non-cleavage site polymorphisms is underlined by the fact that ev en after we introduced a MA/CA cleavage site identical with that in subtype B, the amount of p25 produced by processing of subtype A gag is still much lower than for subtype B. Also, when a subtype A MA/CA cleavage site was engineered in subtype B gag the drop in the cl eavage rate was not to the level encountered in subtype A gag. These results also point not only to the effect of cleavag e and non-cleavage site polymorphisms but also to a probable cooperative effect among different regions in gag. A better understand of the role of non-cleavage site polym orphisms could possible come from crystallographic or NMR st udies of gag polyprotein which c ould provide us with a detailed map of the interactions and structural arrang ements among different components of gag. Such information would be useful in designing nove l therapeutic approach es to disrupt these interactions. An interesting finding was that the sequence analysis through the Los Alamos Database showed that the consensus sequence of subtype A sub-subtype A2 contains a serine at the same position. Sequence analysis of the rest of the ga g and gag/pol genes and proteins of the subtype A clone analyzed here revealed that it belongs to subtype A sub-subtype A2 group. This subsubtype is confined to small region in the wester n part of Africa, and is not frequently found in any other part of the world. There might be a connection between a slower processing rate and the reduced number of HIV-posit ive persons infected with th is sub-subtype but many other factors, such as ethnicity, geographically conf ined population groups, and access to care have to
171 be taken in consideration. Also, this in vitro assay of gag processing cannot replace phenotypic studies on the replicative capacity of the virus harboring this poly morphism. It is possible that, due to other compensatory factors not assayed here, the replicative capacity of the vi rus is not to be affected at the same extent as seen with the in vitro assay. PR processing activity depends on (1) the pr imary amino acids sequence of PR, (2) the composition of the cleavage sites, (3) determinants in gag, and (4) the accessi bility to the site of the PR (Pettit et al. 1994; Goode now et al. 2002). We showed th at determinants in gag could play a dominant role in modulating the rate of cleavage. This research and previous studies support the sequential, ordered substrate cl eavage by the HIV-1 PR. The idea of strict requirements for gag processing is also supporte d by the fact that, despite many differences between the three subtypes studied here, the or der and, largely, the rate of cleavage are comparable. Further mutational and structural studies would bring more insight and would pinpoint the key residues involve d in coordinating the processing events. Also future studies where PI will be added in the reaction mixture should answer the question whether this naturally occurring polymorphism represen ts a mechanism through which the virus can acquire faster drug resistance. The validity of these experiments is limited by the extent to which this in vitro processing system mimics the precursor cleavage that accompanies virus assemble in vivo (Pettit et al. 2004). Several studies support the theory that the spontaneous processing of in vitro -translated precursors is an appropriate model for processing in vivo (Carter and Zybarth 1994; Platt and Haffar 1994; Zybarth and Carter 1995). Several groups (Moody et al. 1995; Spearman and Ratner 1996; Lee et al. 1999) have observed that gag or pol precursors produced in rabbit reticulocite lysate (RRL) multimerize into higher -order components similar to those observed in
172 virions. These studies indicate that the concentrati on of gag/pol precursors in this system is high enough to support multimerization. Other lines of evidence suggest that the intermediates and final products that arise during se quential processing of gag polyprotein with this assay correlate well with the processing events observed in infected cells (Mer vis et al. 1988; Gowda et al. 1989). Others (Platt and Haffar 1994; Hermida-Matsumoto and Resh 1999) have noted efficient membrane association of translated gag in RRL Campbell et al. found that components within the reticulocyte lysate are required for proper multimerization of purified HIV gag into 110-nm particles in vitro and later linked that requirement to specific phospholipids (Campbell and Rein 1999; Campbell et al. 2001). The correct interpretation of the bands seen on the autoradiographs was confirmed by Western blot analysis (data not shown). Despite the high similarity in the amino sequence of the CA protein among all HIV-1 subtypes, the monoc lonal antibodies against subtype B that we initially used to identify the cleavage products contai ning p24 (NIH AIDS reagent program) did not bind to subtype C p24. However, a second polyclonal antibody did re cognize the subtype C p24. One of the questions we asked was why we we re not able to observe processing in the absence of a forced frameshift mutation. It has been well documented that the frameshift occurs in 5% of translational events and is the result of the ribosomal slippage along UUUUUUA sequence (Jacks et al. 1988). Our initial expectations were that we would be able to document processing without addition of PR in trans or engineering a forced frameshift mutation. However, even after allowing the reaction to pr oceed for as long as 3 hours, no processing was observed either by visually inspecting the au toradiographs or by performing densitometric analysis (Figure 6-4). Several previ ous studies showed that processing in vivo occurs in a limited
173 space (the interior of the virion) where the concen tration of gag and gag/pol, and so of the HIV-1 PR embedded within gag/pol, is high. This allows for efficient dimerization and thus activation of the PR and for the initiation of gag/pol cleavag e. This spatial limitation and the high number of gag and gag/pol polyproteins directed at the cell membrane are the key factors in promoting processing and maturation of the virions. Even if we assume that the frameshift occurs with the same frequency in our system, the concentratio n of gag/pol might be too low to promote the efficient dimerization of the protease monomers This tentatively explains how the processing events happen upon addition of an increased concentration of PR. Despite its limitations, this in vitro system has several advantag es: (1) it is significantly faster than any other expression -purification systems using prokar yotic or eukaryotic cells; (2) it does not involve the generation of live HIV; therefore it does not requ ire special biosafety facilities; (3) it is limited to the mechanistic relationship of PR and gag or gag/pol, thus eliminating variable factors that would complicat e even more the interpretation of the results. The present data, taken together, suggest that HIV-1 resistance to PIs is associated with primary and secondary mutations in the viral PR and is also associated with the cleavage site amino acid substitutions in gag together with substitutions at non-cleavage sites. We believe it is a necessity to examine regions within gag pol yprotein for predictive sequences that, in combination with naturally occurring polymor phisms or ARV-induced mutations in PR, will enhance genotype and drug susceptibility predic tions and allow customization of therapeutic regimens, especially needed in patients failing multiple anti-HIV drug regimens.
174 Figure 6-1. HIV-1 gag processing. In cis cleavage occurs when the HIV-1 PR cuts within the same gag/pol polyprotein in which it is embedded, and in trans cleavage occurs when the PR cleaves the gag polyprotein.
175 Figure 6-2. Alignment of the gag polyproteins of HIV-1 subtype B, C and A. The near fulllength clones of gag/pol gene from subtypes A, B and C were obtained from NIH AIDS Reference and Reagent Program. The alignment was performed with ClustalW (Expasy Proteomics Server).
176 Figure 6-2. Continued
177 Figure 6-3. Alignment of the cleavage sites within gag polyproteins of HIV1 subtypes, B, C, and A.
178 Figure 6-4. The in vitro processing without the frames hift and without the addition in trans of the HIV protease. Protein marker is shown in the left and the gag polyproteins in the right of the gel. The time points are i ndicated in the top of the gel.
179 A B (min) Figure 6-5. In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype B protease. A) Autoradiographs. Th e time points are indicated in the top of the gel. The protein marker is in the left and the processing intermediates are in the right of the gel. B) Densitometric Anal ysis. The time indicated on the x-axis is shown in minutes. The legend is: p55, p41, p24, p15.
180 A B (min) Figure 6-6. In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype A protease. A) Autora diographs. The time points ar e indicated in the top of the gel. The protein marker is in the left and the processing intermediates are in the right of the gel. B) Densitometric Anal ysis. The time indicated on the x-axis is shown in minutes. The legend is: p55, p41, p24, p15.
181 A B (min) Figure 6-7. In trans processing of HIV-1 subtype B, A, and C gag polyproteins by HIV-1 subtype C protease. A) Autoradiographs. The time points are indicated in the top of the gel. The protein marker is in the left and the processing intermediates are in the right of the gel. B) Densitometric Analysis The time indicated on the x-axis is shown in minutes. The legend is: p55, p41, p24, p15.
182 A B (min) Figure 6-8. In trans processing of HIV-1 subtype A gag pol yprotein S124V and QV variants by HIV-1 subtype B protease. A) Autoradiogr aphs. The black triangles in the top of the gel indicate 7 time points from 0 to 90 min after subtype B PR addition. The protein marker is in the left of the gel. B) Densitometric Analysis. The time indicated on the x-axis is shown in minutes. The legend is: p55, p41, p24, p15.
183 A B (min) Figure 6-9. In trans processing of HIV-1 subtype B gag polyprotein V124S and QV variants by HIV-1 subtype B protease. A) Autoradiogr aphs. The black triangles in the top of the gel indicate 5 time points from 0 to 90 mi n. The protein marker is in the left of the gel. B) Densitometric Analysis. The time indicated on the x-axis is shown in minutes. The legend is: p55, p41, p24, p15.
184 Figure 6-10. HIV-1 matrix protei n (PDB code 1TAM). It has 5 -helices, A to E, (shown in red) connected through unstructured coils (shown in green). The last 20 amino acids at the C-terminus adopt a flexible coil-lik e conformation, important for MA-MA dimerization. Figure rendered with Pymol (DeLano Scientific). NH2 HOOCA D B C E
185 LIST OF REFERENCES Alexander CS, Montessori V, Wynhoven B et al. (2002) Prevalence and response to antiretroviral therapy of non-B subtypes of HIV in antiretroviral-naive individuals in British Columbia. Antivir Ther. 7:31-35 Alkhatib G, Liao F, Berger EA, Farber JM, Peden KW (1997) A new SIV co-receptor, STRL33. Nature. 388:238 Atlas A, Granath F, Lindstrom A, Lidman K, Li ndback S, Alaeus A (2005) Impact of HIV type 1 genetic subtype on the outcome of antiretrovi ral therapy. AIDS Res Hum Retroviruses. 21:221-227 Baeten D (2007) HIV-1 subtype D infection is associated with faster disease progression than subtype A, in spite of similar HIV-1 pl asma viral loads. In: 14th Conference on Retroviruses and Opportunistic Infections, Los Angeles Baxter JD, Schapiro JM, Boucher CA et al. (2006) Genotypic changes in human immunodeficiency virus type 1 protease associated with reduced susceptibility and virologic response to the protease inhib itor tipranavir. J Vi rol. 80:10794-10801. Epub 12006 Aug 10723 Bhoopat L, Eiangleng L, Rugpao S et al. (2001) In vivo identifica tion of Langerhans and related dendritic cells infected with HIV-1 subtype E in vaginal mucosa of asymptomatic patients. Mod Pathol. 14:1263-1269 Blackard JT, Renjifo B, Fawzi W et al. (2001) HIV-1 LTR subtype and pe rinatal transmission. Virology. 287:261-265 Bocket L, Cheret A, Deuffic-Burban S et al. (2005) Impact of human immunodeficiency virus type 1 subtype on first-line antiretroviral therapy effectiveness. Antivir Ther 10:247-254 Borsetti A, Ohagen A, Gottlinger HG (1998) The C-terminal half of the human immunodeficiency virus type 1 Ga g precursor is sufficient for efficient particle assembly. J Virol. 72:9313-9317 Bray M, Prasad S, Dubay JW et al. (1994) A sm all element from the Mason-Pfizer monkey virus genome makes human immunodeficiency virus type 1 expression and replication Revindependent. Proc Natl A cad Sci U S A. 91:1256-1260 Brenner B, Turner D, Oliveira M et al. ( 2003) A V106M mutation in HIV-1 clade C viruses exposed to efavirenz confers cross-resistan ce to non-nucleoside re verse transcriptase inhibitors. Aids 17:F1-5 Brodine SK, Mascola JR, Weiss PJ et al. (1995) Detection of di verse HIV-1 genetic subtypes in the USA. Lancet 346:1198-1199
186 Brunger AT, Adams PD, Clore GM et al. ( 1998) Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr. 54:905-921 Bryant M, Ratner L (1990) Myristoylation-dependent replic ation and assembly of human immunodeficiency virus 1. Proc Natl Acad Sci U S A. 87:523-527 Buckman JS, Bosche WJ, Gorelick RJ (2003) Human immunodeficiency virus type 1 nucleocapsid zn(2+) fingers are required fo r efficient reverse transcription, initial integration processes, and protection of ne wly synthesized viral DNA. J Virol. 77:14691480 Burger DM, de Graaff M, Wu is EW, Koopmans PP, Hekster YA (1997) Determination of indinavir, an HIV-protease inhibitor, in human plas ma by reversed-phase highperformance liquid chromatography. J Ch romatogr B Biomed Sci Appl. 703:235-241 Burniston MT, Cimarelli A, Colgan J, Curtis SP, Luban J (1999) Human immunodeficiency virus type 1 Gag polyprotein multimerization requires the nucleocapsid domain and RNA and is promoted by the capsid-dimer interface and the basic region of matrix protein. J Virol. 73:8527-8540 Buve A, Carael M, Hayes RJ et al. (2001) Mu lticentre study on factors determining differences in rate of spread of HIV in sub-Saharan Af rica: methods and prevalence of HIV infection. Aids 15 Suppl 4:S5-14 Calazans A, Brindeiro R, Brinde iro P et al. (2005) Low accumulati on of L90M in protease from subtype F HIV-1 with resistance to prot ease inhibitors is caused by the L89M polymorphism. J Infect Dis. 191:1961-1970. Epub 2005 Apr 1928 Campbell S, Fisher RJ, Towler EM et al. (2001) Modulation of HIV-like particle assembly in vitro by inositol phosphates. Proc Natl Acad Sci U S A. 98:10875-10879. Epub 12001 Aug 10828 Campbell S, Rein A (1999) In vitro assembly pr operties of human immunode ficiency virus type 1 Gag protein lacking the p6 domain. J Virol. 73:2270-2279 Cane PA, de Ruiter A, Rice P, Wiselka M, Fox R, Pillay D (2001) Resistance-associated mutations in the human immunodeficiency vi rus type 1 subtype c protease gene from treated and untreated patients in the Unite d Kingdom. J Clin Microbiol 39:2652-2654 Caride E, Hertogs K, Larder B et al. (2001) Genot ypic and phenotypic evidence of different drug-resistance mutation patterns between B and non-B subtype isolates of human immunodeficiency virus type 1 found in Br azilian patients failing HAART. Virus Genes 23:193-202
187 Carpenter CC, Fischl MA, Hammer SM et al. (1997) Antiretroviral therapy for HIV infection in 1997. Updated recommendations of the Intern ational AIDS Society-USA panel. Jama 277:1962-1969 Carter C, Zybarth G (1994) Pro cessing of retroviral Gag polypro teins: an in vitro approach. Methods Enzymol 241:227-253 Centlivre M, Sommer P, Michel M et al. (2006) The HIV-1 clade C promoter is particularly well adapted to replication in the gut in primary infection. Aids. 20:657-666 Clapham PR, Weiss RA (1997) I mmunodeficiency viruses. Spoilt for choice of co-receptors. Nature. 388:230-231 Clemente JC, Coman RM, Thiaville MM et al. (2006) Analysis of HIV-1 CRF_01 A/E protease inhibitor resistance: structur al determinants for maintaining sensitivity and developing resistance to atazanavir. Biochemistry. 45:5468-5477 Clemente JC, Hemrajani R, Blum LE, Goode now MM, Dunn BM (2003) Secondary mutations M36I and A71V in the human immunodeficien cy virus type 1 prot ease can provide an advantage for the emergence of the prim ary mutation D30N. Biochemistry 42:1502915035 Clemente JC, Moose RE, Hemrajani R et al. (2 004) Comparing the accumulation of activeand nonactive-site mutations in the HI V-1 protease. Biochemistry. 43:12141-12151 Collaborative Computational Project n ( 1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr. 50:760-763 Collier AC (1996) Efficacy of combination anti retroviral therapy. A dv Exp Med Biol 394:355372 Coman RM, Robbins A, Goodenow MM, McKe nna R, Dunn BM (2007) Expression, purification and preliminary X-ray cr ystallographic studies of the human immunodeficiency virus 1 subtype C protease. Acta Crystallo graph Sect F Struct Biol Cryst Commun. 63:320-323. Epub 2007 Mar 2030 Condra JH, Holder DJ, Schleif WA et al. (1996) Genetic correlates of in vivo viral resistance to indinavir, a human immunodefici ency virus type 1 protease inhibitor. J Virol 70:82708276 Condra JH, Schleif WA, Blahy OM et al. (1995) In vivo emergence of HIV-1 variants resistant to multiple protease inhibitors. Nature. 374:569-571 Cote HC, Brumme ZL, Harrigan PR (2001) Huma n immunodeficiency vi rus type 1 protease cleavage site mutations associated with prot ease inhibitor cross-resistance selected by indinavir, ritonavir, and/or saquinavir. J Virol. 75:589-594
188 Dauber DS, Ziermann R, Parkin N et al. (2002) A ltered substrate specific ity of drug-resistant human immunodeficiency virus type 1 protease. J Virol. 76:1359-1368 de Mendoza C, Morello J, Garcia-Gasco P, Rodr iguez-Novoa S, Soriano V (2007) Tipranavir: a new protease inhibitor for the treatment of antiretroviral-experienced HIV-infected patients. Expert Opin Pharmacother. 8:839-850 de Oliveira T, Engelbrecht S, Janse van Re nsburg E et al. (2003) Variability at human immunodeficiency virus type 1 subtype C protease cleavage site s: an indication of viral fitness? J Virol 77:9422-9430 De Wit S, Boulme R, Poll B, Schmit JC, Clumeck N (2004) Viral load and CD4 cell response to protease inhibitor-containing regimens in s ubtype B versus non-B treatment-naive HIV-1 patients. Aids. 18:2330-2331 Demirov DG, Orenstein JM, Fr eed EO (2002) The late domai n of human immunodeficiency virus type 1 p6 promotes virus release in a cell type-dependent ma nner. J Virol. 76:105117 Deng HK, Unutmaz D, KewalRamani VN, Litt man DR (1997) Expr ession cloning of new receptors used by simian and human imm unodeficiency viruses. Nature. 388:296-300 DeVita VT, Jr., Broder S, Fauci AS, Kovacs JA, Chabner BA (1987) Developmental therapeutics and the acquire d immunodeficiency syndrome. Ann Intern Med. 106:568581 Doranz BJ, Rucker J, Yi Y et al. (1996) A dualtropic primary HIV-1 isolate that uses fusin and the beta-chemokine receptors CKR-5, CKR3, and CKR-2b as fusion cofactors. Cell. 85:1149-1158 Dorfman T, Bukovsky A, Ohagen A, Hoglund S, Gottlinger HG (1994a) Functional domains of the capsid protein of human immunodeficie ncy virus type 1. J Virol. 68:8180-8187 Dorfman T, Mammano F, Haseltine WA, Gottlinger HG (1994b) Role of the matrix protein in the virion association of the human i mmunodeficiency virus type 1 envelope glycoprotein. J Virol. 68:1689-1696 Doualla-Bell F, Avalos A, Gaol athe T et al. (2006) Impact of human immunodeficiency virus type 1 subtype C on drug resi stance mutations in patients from Botswana failing a nelfinavir-containing regimen. Antim icrob Agents Chemother. 50:2210-2213 Doyon L, Croteau G, Thibeault D, Poulin F, P ilote L, Lamarre D (1996) Second locus involved in human immunodeficiency virus type 1 re sistance to protease i nhibitors. J Virol. 70:3763-3769
189 Dragic T, Litwin V, Allaway GP et al. (1996) HIV-1 entry into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature. 381:667-673 Dumans AT, Soares MA, Machado ES et al. (2004) Synonymous genetic polymorphisms within Brazilian human immunodeficiency virus Type 1 subtypes may influence mutational routes to drug resistance. J In fect Dis 189:1232-1238. Epub 2004 Mar 1212 Dunn BM, Gustchina A, Wlodawer A, Kay J (1994) Subsite prefer ences of retroviral proteinases. Methods Enzymol 241:254-278 Ehrlich LS, Liu T, Scarlata S, Chu B, Carter CA (2001) HIV-1 capsid protein forms spherical (immature-like) and tubular (mature-like) part icles in vitro: stru cture switching by pHinduced conformational changes. Biophys J. 81:586-594 Erickson JW, Burt SK (1996) Structural mechanisms of HIV drug resistance. Annu Rev Pharmacol Toxicol 36:545-571 Erickson JW, Gulnik SV, Markowitz M (1999) Protease inhibitors: resistance, cross-resistance, fitness and the choice of initial and salvage therapies. Aids 13:S189-204 Erickson-Viitanen S, Manfredi J, Viitanen P et al. (1989) Cleavage of HIV-1 gag polyprotein synthesized in vitro: sequential cleavag e by the viral protease. AIDS Res Hum Retroviruses. 5:577-591 Esposito L, Vitagliano L, Sica F, Sorrentino G, Zagari A, Mazzarella L (2000) The ultrahigh resolution crystal structure of ribonuclease A containing an isoaspartyl residue: hydration and sterochemical analysis. J Mol Biol. 297:713-732 Fleury H, Recordon-Pinson P, Caumont A et al (2003) HIV type 1 diversity in France, 19992001: molecular characterizati on of non-B HIV type 1 subtyp es and potential impact on susceptibility to antiretroviral drugs AIDS Res Hum Re troviruses 19:41-47 Foulkes-Murzycki JE, Scott WR, Schiffer CA (2007) Hydrophobic sliding: a possible mechanism for drug resistance in human i mmunodeficiency viru s type 1 protease. Structure. 15:225-233 Frater AJ, Dunn DT, Beardall AJ et al. (2002) Comparative res ponse of African HIV-1-infected individuals to highly ac tive antiretroviral th erapy. Aids. 16:1139-1146 Freedberg DI, Ishima R, Jacob J et al. (2002) Rapid structural fluctu ations of the free HIV protease flaps in solution: relationship to crystal stru ctures and comparison with predictions of dynamics calcula tions. Protein Sci. 11:221-232 Gamble TR, Vajdos FF, Yoo S et al. (1996) Cr ystal structure of hu man cyclophilin A bound to the amino-terminal domain of HIV-1 capsid. Cell. 87:1285-1294
190 Gamble TR, Yoo S, Vajdos FF et al. (1997) Structure of the carboxyl -terminal dimerization domain of the HIV-1 capsid protein. Science. 278:849-853 Garrus JE, von Schwedler UK, Pornillos OW et al. (2001) Tsg101 and the vacuolar protein sorting pathway are essential for HIV-1 budding. Cell. 107:55-65 Gaschen B, Taylor J, Yusim K et al. (2002) Diversity considerati ons in HIV-1 vaccine selection. Science 296:2354-2360 Gatanaga H, Suzuki Y, Tsang H et al. (2002) Amino acid substitutions in Gag protein at noncleavage sites are indispensable for the development of a high multitude of HIV-1 resistance against protease inhibito rs. J Biol Chem. 277:5952-5961. Epub 2001 Dec 5910 Gitti RK, Lee BM, Walker J, Summers MF, Yoo S, Sundquist WI (1996) Structure of the aminoterminal core domain of the HIV-1 capsid protein. Science. 273:231-235 Gong YF, Robinson BS, Rose RE et al. (2000) In vitro resistance profile of the human immunodeficiency virus type 1 protease inhibitor BMS-232632. Antimicrob Agents Chemother. 44:2319-2326 Gonzalez LM, Aguiar RS, Afonso A et al. ( 2006) Biological charac terization of human immunodeficiency virus type 1 subtype C protease carrying indi navir drug-resistance mutations. J Gen Virol. 87:1303-1309 Gonzalez LM, Brindeiro RM, Tarin M et al. (2 003) In vitro hypersus ceptibility of human immunodeficiency virus type 1 subtype C pr otease to lopinavir. Antimicrob Agents Chemother. 47:2817-2822 Goodenow MM, Bloom G, Rose SL et al. (2002) Naturally occu rring amino acid polymorphisms in human immunodeficiency virus type 1 (H IV-1) Gag p7(NC) and the C-cleavage site impact Gag-Pol processing by HI V-1 protease. Virology 292:137-149 Gordon M, De Oliveira T, Bishop K et al. (2003) Molecular charac teristics of human immunodeficiency virus type 1 subtype C vi ruses from KwaZulu-Natal, South Africa: implications for vaccine and antiretrovir al control strategies. J Virol 77:2587-2599 Gottlieb MS, Groopman JE, Weinstein WM, Fahey JL, Detels R (1983) The acquired immunodeficiency syndrome. Ann Intern Med 99:208-220 Gottlinger HG, Dorfman T, Sodroski JG, Haseltine WA (1991) Effect of mutations affecting the p6 gag protein on human immunodefi ciency virus particle rel ease. Proc Natl Acad Sci U S A. 88:3195-3199 Gottlinger HG, Sodroski JG, Haseltine WA (1989) Role of capsid precursor processing and myristoylation in morphogenesis and infectiv ity of human immunodeficiency virus type 1. Proc Natl Acad Sci U S A. 86:5781-5785
191 Gowda SD, Stein BS, Engleman EG (1989) Iden tification of protein intermediates in the processing of the p55 HIV-1 gag precursor in cells infected with recombinant vaccinia virus. J Biol Chem. 264:8459-8462 Groopman JE, Gottlieb MS (1983) Ac quired immune deficiency syndrome. AIDS: the widening gyre. Nature 303:575-576 Gross I, Hohenberg H, Wilk T et al. (2000) A conformational switc h controlling HIV-1 morphogenesis. Embo J. 19:103-113 Grossman Z et al. (2004) Mutation D30N is not preferentially selected by human immunodeficiency virus type 1 subtype C in th e development of resist ance to nelfinavir. Antimicrob Agents Chemother 48:2159-2165 Gustchina A, Weber IT (1990) Comparison of in hibitor binding in HIV-1 protease and in nonviral aspartic proteases: the role of the flap. FEBS Lett. 269:269-272 Hamelberg D, McCammon JA (2005) Fast peptidyl cis-trans isomer ization within the flexible Gly-rich flaps of HIV-1 proteas e. J Am Chem Soc. 127:13778-13779 Hammond J, Larder BA, Schinazi RF, Mellors JW (1997) Mutation in retroviral genes associated with drug resistance. In: Korber B (ed) Human Retroviruses and AIDS, Los Alamos, NM, pp 207-249 Harrigan PR, Hertogs K, Verbiest W et al. (1999) Baseline HIV drug resistance profile predicts response to ritonavir-saquinavir protease inhi bitor therapy in a comm unity setting. Aids. 13:1863-1871 Heaslet H, Rosenfeld R, Giffin M et al. (to be published) Confor mational flexibility in the flap domains of ligand free HIV protease. To be published Henderson LE, Benveniste RE, Sowder R, Cope land TD, Schultz AM, Oroszlan S (1988a) Molecular characterization of gag proteins from simian immunodeficiency virus (SIVMne). J Virol. 62:2587-2595 Henderson LE, Copeland TD, Sowder RC, Schultz AM, Oroszlan S (1988b) Human Retrovirus, Cancer and AIDS: Aproaches to Prevention and Therapy. In: Bol ognesi D (ed) Human Retrovirus, Cancer and AI DS. Liss, New York, pp 135-147 Henderson LE, Sowder RC, Copeland TD, Oroszlan S, Benveniste RE (1990) Gag precursors of HIV and SIV are cleaved into six proteins found in the ma ture virions. J Med Primatol 19:411-419
192 Hermida-Matsumoto L, Resh MD (1999) Huma n immunodeficiency virus type 1 protease triggers a myristoyl switch that modulates membrane binding of Pr55(gag) and p17MA. J Virol. 73:1902-1908 Hertogs K, Bloor S, De Vroey V et al. (2000) A novel human immunodeficiency virus type 1 reverse transcriptase mutational pattern conf ers phenotypic lamivudine resistance in the absence of mutation 184V. Anti microb Agents Chemother 44:568-573 Hill M, Tachedjian G, Mak J (2005) The packagi ng and maturation of the HIV-1 Pol proteins. Curr HIV Res. 3:73-85 Hill MK, Shehu-Xhilaga M, Crowe SM, Mak J (2 002) Proline residues within spacer peptide p1 are important for human immunodeficiency viru s type 1 infectivity, protein processing, and genomic RNA dimer stab ility. J Virol. 76:11245-11253 Hirsch MS, Brun-Vezinet F, D'Aqui la RT et al. (2000) Antiretrovira l drug resistance testing in adult HIV-1 infection: recommendations of an International AIDS Society-USA Panel. Jama 283:2417-2426 Hirsch VM, Olmsted RA, Murphey-Corb M, Pur cell RH, Johnson PR (1989) An African primate lentivirus (SIVsm) closely rela ted to HIV-2. Nature 339:389-392 Holguin A, Alvarez A, Soria no V (2002) High prevalence of HIV-1 subtype G and natural polymorphisms at the protease gene among HIV-infected immi grants in Madrid. Aids 16:1163-1170 Holguin A, Paxinos E, Hertogs K, Womac C, So riano V (2004) Impact of frequent natural polymorphisms at the protease gene on the in vitr o susceptibility to protease inhibitors in HIV-1 non-B subtypes. J Clin Virol. 31:215-220 Hornak V, Okur A, Rizzo RC, Simmerling C ( 2006a) HIV-1 protease flaps spontaneously close to the correct structure in simulations following manual placement of an inhibitor into the open state. J Am Chem Soc. 128:2812-2813 Hornak V, Okur A, Rizzo RC, Simmerling C (2006b) HIV-1 protease flaps spontaneously open and reclose in molecular dynamics simulati ons. Proc Natl Acad Sci U S A. 103:915-920. Epub 2006 Jan 2017 Huang M, Orenstein JM, Martin MA, Freed EO ( 1995) p6Gag is required for particle production from full-length human immunodeficiency vi rus type 1 molecular clones expressing protease. J Virol. 69:6810-6818. Hugen PW, Verweij-van Wissen CP, Burger DM Wuis EW, Koopmans PP, Hekster YA (1999) Simultaneous determination of the HIV-prot ease inhibitors indinavir, nelfinavir, saquinavir and ritonavir in human plasma by reversed-phase high-performance liquid chromatography. J Chromatogr B Biomed Sci Appl. 727:139-149
193 Hung M, Patel P, Davis S, Green SR (1998) Im portance of ribosomal frameshifting for human immunodeficiency virus type 1 particle asse mbly and replication. J Virol. 72:4819-4824 Ibe S, Shibata N, Utsumi M, Kaneda T (2003) Se lection of human immun odeficiency virus type 1 variants with an insertion mutation in the p6(gag) and p6(pol) gene s under highly active antiretroviral therapy. Microbiol Immunol 47:71-79 Iga M, Matsuda Z, Okayama A et al. (2002) Rapid phenotypic assay for human immunodeficiency virus type 1 protease using in vitro tr anslation. J Virol Methods 106:25-37 Ishima R, Freedberg DI, Wang YX, Louis JM Torchia DA (1999) Flap opening and dimerinterface flexibility in the fr ee and inhibitor-bound HIV protease, and their implications for function. Structure. 7:1047-1055 Jacks T, Power MD, Masiarz FR, Luciw PA, Ba rr PJ, Varmus HE (1988) Characterization of ribosomal frameshifting in HIV-1 ga g-pol expression. Nature 331:280-283 Jacobo-Molina A, Ding J, Nanni RG et al. (1993) Crys tal structure of human immunodeficiency virus type 1 reverse transcriptase comp lexed with double-stranded DNA at 3.0 A resolution shows bent DNA. Proc Natl Acad Sci U S A. 90:6320-6324 Jones TA (1978) A graphics model building and re finement system for macromolecules. Journal of Applied Crystallography:268-272 Jones TA, Zou JY, Cowan SW, Kjeldgaard M (1 991) Improved methods for building protein models in electron density maps and the lo cation of errors in these models. Acta Crystallogr A. 47:110-119 Kanki PJ, Hamel DJ, Sankale JL et al. (1999) Human immunodeficiency virus type 1 subtypes differ in disease progressi on. J Infect Dis. 179:68-73 Kantor R, Katzenstein D (2003) Polymorphism in HIV-1 non-s ubtype B protease and reverse transcriptase and its potential impact on drug susceptibility and drug resistance evolution. AIDS Rev 5:25-35 Kantor R, Katzenstein D (2004) Drug resistance in non-subtype B HIV-1. J Clin Virol 29:152159 Kantor R, Machekano R, Gonzales MJ, Dupni k K, Schapiro JM, Shafer RW (2001) Human Immunodeficiency Virus Reverse Transcript ase and Protease Sequence Database: an expanded data model integrating natural langua ge text and sequence analysis programs. Nucleic Acids Res 29:296-299
194 Kaplan AH, Manchester M, Sw anstrom R (1994) The activity of the protease of human immunodeficiency virus type 1 is initiated at the membrane of infected cells before the release of viral proteins and is required for release to occur with maximum efficiency. J Virol 68:6782-6786 Karacostas V, Wolffe EJ, Nagashima K, Gonda MA, Moss B (1993) Overexpression of the HIV1 gag-pol polyprotein results in intracellular activation of HIV-1 protease and inhibition of assembly and budding of virus-li ke particles. Virology. 193:661-671 Karpas A (2004) Human retrovir uses in leukaemia and AIDS: reflections on their discovery, biology and epidemiology. Biol Rev Camb Philos Soc 79:911-933 Kempf DJ, Isaacson JD, King MS et al. (2001) Identification of genotypic changes in human immunodeficiency virus protease that correlate wi th reduced susceptibility to the protease inhibitor lopinavir among vira l isolates from protease inhi bitor-experience d patients. J Virol 75:7462-7469 Kilby JM, Sfakianos G, Gizzi N et al. (2000) Safety and pharmacokinetics of once-daily regimens of soft-gel capsule saquinav ir plus minidose ritonavir in human immunodeficiency virus-negative adults. Antimicrob Agents Chemother. 44:2672-2678 Kohl NE, Emini EA, Schleif WA et al. (1988) Ac tive human immunodeficiency virus protease is required for viral infectivity. Proc Natl Acad Sci U S A. 85:4686-4690 Kolli M, Lastere S, Schiffer CA (2006) Co-evolution of nelfinavir-resistant HIV-1 protease and the p1-p6 substrate. Virology. 347:405-409. Epub 2006 Jan 2020 Kondo E, Gottlinger HG (1996) A conserved L XXLF sequence is the major determinant in p6gag required for the incorpor ation of human immunodeficien cy virus type 1 Vpr. J Virol. 70:159-164 Kozal MJ, Shah N, Shen N et al. (1996) Ex tensive polymorphisms observed in HIV-1 clade B protease gene using high-density ol igonucleotide arrays Nat Med. 2:753-759 Krausslich HG, Ingraham RH, S koog MT, Wimmer E, Pallai PV, Carter CA (1989) Activity of purified biosynthetic proteinase of human i mmunodeficiency virus on natural substrates and synthetic peptides. Proc Natl Acad Sci U S A. 86:807-811 Kurowski M, Kaeser B, Sawyer A, Popescu M, Mrozikiewicz A ( 2002) Low-dose ritonavir moderately enhances nelfinavir expo sure. Clin Pharmacol Ther. 72:123-132 Laeyendecker O, Li X, Arroyo M (2006) Th e Effect of HIV Subt ype on Rapid Disease Progression in Rakai, Uganda. In: 13th C onference on Retroviruses and Opportunistic Infections, Los Angeles, CA
195 Laskowski RA, MacArthur MW, Moss DS, Thor nton JM (1993) PROCHECK: a program to check the stereochemical quality of pr otein structures. Journal of Applied Crystallography:283-291 Le Tiec C, Barrail A, Goujard C, Taburet AM (2005) Clinical pharmacoki netics and summary of efficacy and tolerability of atazanav ir. Clin Pharmacokinet 44:1035-1050 Lee PP, Linial ML (1994) Efficient particle form ation can occur if the matrix domain of human immunodeficiency virus type 1 Gag is substi tuted by a myristylation signal. J Virol. 68:6644-6654 Lee YM, Liu B, Yu XF (1999) Formation of vi rus assembly intermediate complexes in the cytoplasm by wild-type and assembly-defec tive mutant human im munodeficiency virus type 1 and their association with membranes. J Virol. 73:5654-5662 Liao WH, Chiu HC, Wang CT (2004) Effects of mutations in an HIV-1 gag gene containing a 107-codon tandem repeat in the matrix region on assembly and processing of the protein product. J Med Virol 74:528-535 Liao WH, Wang CT (2004) Char acterization of human immunode ficiency virus type 1 Pr160 gag-pol mutants with truncations downstr eam of the protease domain. Virology 329:180188 Liao Z, Cimakasky LM, Hampton R, Nguyen DH Hildreth JE (2001) Lipid rafts and HIV pathogenesis: host membrane cholesterol is required for infection by HIV type 1. AIDS Res Hum Retroviruses. 17:1009-1019 Lindberg J, Pyring D, Lowgren S et al. (2004) Symmetric fluor o-substituted diol-based HIV protease inhibitors. Ortho-fluorinated and me ta-fluorinated P1/P1'-b enzyloxy side groups significantly improve the antiviral activity a nd preserve binding efficacy. Eur J Biochem. 271:4594-4602 Loemba H, Brenner B, Parniak MA et al. (2002) Genetic divergence of human immunodeficiency virus type 1 Ethiopian clade C reverse tr anscriptase (RT) and rapid development of resistance against nonnucleos ide inhibitors of RT. Antimicrob Agents Chemother 46:2087-2094 Logsdon BC, Vickrey JF, Martin P et al. (2004) Crystal structures of a multidrug-resistant human immunodeficiency virus ty pe 1 protease reveal an expanded active-site cavity. J Virol. 78:3123-3132 Los Alamos HD (2007) Los Alamos HIV Database. Los Alamos National Laboratory. Accessed from http://www.hiv.lanl.gov/content/index, on June 2007
196 Louis JM, Nashed NT, Parris KD, Kimmel AR, Je rina DM (1994) Kinetics and mechanism of autoprocessing of human immunode ficiency virus type 1 prot ease from an analog of the Gag-Pol polyprotein. Proc Na tl Acad Sci USA. 91:7970-7974 Louis JM, Wondrak EM, Kimmel AR Wingfield PT, Nashed NT ( 1999) Proteolytic processing of HIV-1 protease precursor kinetics and mechanism. J Biol Chem 274:23437-23442 Loveday C, Burke A, Van Hooff F, Johnson MC ( 2001) Equivalence in viro logical responses to HAART in patients with HIV-1 subtype non-B infection: a ca se-controlled study (n=100). In: The 8th Conferen ce on Retroviruses and Opport unistic Infections, Chicago Maguire MF, Guinea R, Griffin P et al. (2002) Changes in human immunodeficiency virus type 1 Gag at positions L449 and P453 are linked to I50V protease mutants in vivo and cause reduction of sensitivity to amprenavir and improved viral fitness in vitro. J Virol. 76:7398-7406 Mahalingam B, Boross P, Wang YF et al. (2002) Combining mutations in HIV-1 protease to understand mechanisms of resi stance. Proteins 48:107-116 Mahalingam B, Louis JM, Hung J, Harrison RW, We ber IT (2001) Structural implications of drug-resistant mutants of HIV-1 protease: high-resolution crysta l structures of the mutant protease/substrate analogue co mplexes. Proteins 43:455-464 Mahalingam B, Wang YF, Boross PI et al. (2004) Crystal structures of HIV protease V82A and L90M mutants reveal cha nges in the indinavir-bindi ng site. Eur J Biochem 271:15161524 Mammano F, Ohagen A, Hoglund S, Gottlinger HG (1994) Role of the major homology region of human immunodeficiency virus type 1 in virion morphogenesis. J Virol. 68:4927-4936 Mammano F, Petit C, Clavel F (1998) Resistance-associated loss of viral fitness in human immunodeficiency virus type 1: phenotypic analysis of prot ease and gag coevolution in protease inhibitortreated patients. J Virol. 72:7632-7637 Mansky LM, Temin HM (1995) Lower in vivo muta tion rate of human i mmunodeficiency virus type 1 than that predicted from the fidelity of purified reverse transcriptase. J Virol. 69:5087-5094 Margolin N, Heath W, Osborne E, Lai M, Vlahos C (1990) Substitutions at the P2' site of gag p17-p24 affect cleavage efficiency by HIV1 protease. Biochem Biophys Res Commun. 167:554-560 Marquardt DW (1963) An Algorith m for Least-Squares Estimation of Nonlinear Parameters. J. Soc. Ind. Appl. Math. 11:431-441
197 Martin P, Vickrey JF, Protease G et al. (2005) "Wide-open" 1.3 A structure of a multidrugresistant HIV-1 protease as a dr ug target. Structure. 13:1887-1895 Martinez-Picado J, Savara AV, Sutton L, D'Aquila RT (1999) Re plicative fitne ss of protease inhibitor-resistant mutants of human imm unodeficiency virus type 1. J Virol. 73:37443752 Martin-Serrano J, Zang T, Bieniasz PD (2001) HIV-1 and Ebola virus encode small peptide motifs that recruit Tsg101 to sites of particle assembly to facilitate egress. Nat Med. 7:1313-1319 Massiah MA, Starich MR, Pasc hall C, Summers MF, Christ ensen AM, Sundquist WI (1994) Three-dimensional structure of the human immunodeficiency vi rus type 1 matrix protein. J Mol Biol. 244:198-223 Matthews BW (1968) Solvent content of protein crystals. J Mol Biol. 33:491-497 Matthews S, Barlow P, Boyd J et al. (1994) Struct ural similarity between the p17 matrix protein of HIV-1 and interferon-gamma. Nature. 370:666-668. McCormack GP, Glynn JR, Crampin AC et al (2002) Early evolution of the human immunodeficiency virus type 1 subtype C ep idemic in rural Malawi. J Virol 76:1289012899 McDermott J, Farrell L, Ross R, Barklis E (1996) Structural analysis of human immunodeficiency virus type 1 Gag protein interactions, usin g cysteine-specific reagents. J Virol. 70:5106-5114 McPherson A (1982) Preparation and Analysis of Protein Crystals. Wiley & Son, New York McQuade TJ, Tomasselli AG, Liu L et al. (1990) A s ynthetic HIV-1 protea se inhibitor with antiviral activity arrest s HIV-like particle matura tion. Science. 247:454-456 Merritt EA (1999a) Comparing anis otropic displacement parameters in protein structures. Acta Crystallogr D Biol Crystallogr. 55:1997-2004 Merritt EA (1999b) Expanding th e model: anisotropic displacem ent parameters in protein structure refinement. Acta Crysta llogr D Biol Crystallogr. 55:1109-1117 Merritt EA, Bacon DJ (1997) Raster3D Version 2: photorealistic molecular graphics. Methods of Enzymology 227:505-524 Mervis RJ, Ahmad N, Lillehoj EP et al (1988) The gag gene products of human immunodeficiency virus type 1: alignmen t within the gag open reading frame, identification of posttranslat ional modifications, and evidence for alternative gag precursors. J Virol 62:3993-4002
198 Michaelis L, Menten ML (1913) Die Kinetik de r Invertinwerkung. Biochemische Zeitschrift 49:333 Miller M, Schneider J, Sathyana rayana BK et al. (1989) Structur e of complex of synthetic HIV-1 protease with a substrate-based inhibito r at 2.3 A resoluti on. Science 246:1149-1152 Mitsuya Y, Winters MA, Fessel WJ et al. (2006) N88D facilitates the co-occurrence of D30N and L90M and the development of multidrug resistance in HIV type 1 protease following nelfinavir treatment failure. AI DS Res Hum Retroviruses. 22:1300-1305 Molla A, Korneyeva M, Gao Q et al. (1996) Or dered accumulation of mutations in HIV protease confers resistance to rito navir. Nat Med. 2:760-766 Moody MD, Pettit SC, Shao W et al. (1995) A side chain at position 48 of the human immunodeficiency virus type-1 protease fl ap provides an additional specificity determinant. Virology. 207:475-485 Moore JP (1997) Coreceptors: im plications for HIV pathogenesi s and therapy. Science. 276:5152 Murakami T, Freed EO (2000) Genetic ev idence for an inter action between human immunodeficiency virus type 1 matrix and al pha-helix 2 of the gp41 cytoplasmic tail. J Virol. 74:3548-3554 Murray MC, Embree JE, Ramdahin SG, Anzala AO, Njenga S, Plummer FA (2000) Effect of human immunodeficiency viru s (HIV) type 1 viral geno type on mother-to-child transmission of HIV-1. J Infect Dis. 181:746-749 Muzammil S, Ross P, Freire E (2003) A major role for a set of non-active site mutations in the development of HIV-1 protease drug resistance. Biochemistry 42:631-638 Navaza J (2001) Implementation of molecular repl acement in AMoRe. Acta Crystallogr D Biol Crystallogr. 57:1367-1372. Epub 2001 Sep 1321 Navia MA, Fitzgerald PM, McKeever BM et al. (1989) Three-dimensional structure of aspartyl protease from human immunodeficiency virus HIV-1. Nature 337:615-620 Nguyen DH, Hildreth JE (2000) Evidence for bud ding of human immunodef iciency virus type 1 selectively from glycolipid-enriched me mbrane lipid rafts. J Virol. 74:3264-3272 Nicholson LK, Yamazaki T, Torchi a DA et al. (1995) Flexibility a nd function in HIV-1 protease. Nat Struct Biol. 2:274-280
199 Olsen DB, Stahlhut MW, Rutkowski CA, Schock HB, vanOlden AL, Kuo LC (1999) Non-active site changes elicit broad-based cross-resistance of the HIV-1 protease to inhibitors. J Biol Chem 274:23699-23701 Oroszlan S, Luftig RB (1990) Retroviral proteinases. Curr Top Microbiol Immunol 157:153-185 Osmanov S, Pattou C, Walker N, Schwardla nder B, Esparza J (2002) Estimated global distribution and regional spr ead of HIV-1 genetic subtypes in the year 2000. J Acquir Immune Defic Syndr 29:184-190 Otwinowski Z, Minor W (1997) Pr ocessing of x-ray diffraction da ta collected in oscillation mode. In: Methods in Enzymology, pp 307-326 Paillart JC, Gottlinger HG (1999) Opposing effect s of human immunodeficiency virus type 1 matrix mutations support a myristyl switch m odel of gag membrane targeting. J Virol. 73:2604-2612 Partin K, Krausslich HG, Ehrlich L, Wimmer E, Carter C (1990) Mutational analysis of a native substrate of the human immunodeficiency vi rus type 1 proteinase. J Virol. 64:3938-3947 Patick AK, Duran M, Cao Y et al. (1998) Genot ypic and phenotypic charact erization of human immunodeficiency virus type 1 variants isolated from patie nts treated with the protease inhibitor nelfinavir. Antimicrob Agents Chemother 42:2637-2644 Peeters M (2001) The genetic vari ability of HIV-1 and its implic ations. Transfus Clin Biol 8:222-225 Peeters M, Sharp PM (2000) Ge netic diversity of HIV-1: th e moving target. Aids 14:S129-140 Perez EE, Rose SL, Peyser B et al. (2001) Human immunodeficiency virus type 1 protease genotype predicts immune and viral respons es to combination therapy with protease inhibitors (PIs) in PI-naive patien ts. J Infect Dis 183:579-588. Epub 2001 Jan 2019 Perno CF, Cozzi-Lepri A, Balotta C et al. (2002) Low prevalence of primary mutations associated with drug resistance in antiviralnaive patients at therapy initiation. Aids 16:619-624 Perrin V, Mammano F (2003) Parameters driving th e selection of nelfinavir-resistant human immunodeficiency virus type 1 variants. J Virol 77:10172-10175 Perryman AL, Lin JH, McCammon JA (2006) Restra ined molecular dynamics simulations of HIV-1 protease: the first step in validati ng a new target for drug design. Biopolymers. 82:272-284 Petropoulos CJ, Parkin NT, Limoli KL et al. (2 000) A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1. Antimicrob Agents Chemother 44:920-928
200 Pettit SC, Clemente JC, Jeung JA, Dunn BM, Ka plan AH (2005) Ordered processing of the human immunodeficiency virus ty pe 1 GagPol precursor is influenced by the context of the embedded viral protea se. J Virol. 79:10601-10607 Pettit SC, Everitt LE, Choudhury S, Dunn BM, Kapl an AH (2004) Initial cleavage of the human immunodeficiency virus type 1 GagPol precursor by its activat ed protease occurs by an intramolecular mechanism. J Virol 78:8477-8485 Pettit SC, Gulnik S, Everitt L, Kaplan AH ( 2003) The dimer interfaces of protease and extraprotease domains influence the activation of protease and the specificity of GagPol cleavage. J Virol. 77:366-374 Pettit SC, Moody MD, Wehbie RS et al. (1994) The p2 domain of human immunodeficiency virus type 1 Gag regulates sequential proteo lytic processing and is required to produce fully infectious virion s. J Virol. 68:8017-8027 Piacenti FJ (2006) An update and review of an tiretroviral therapy. Pharmacotherapy. 26:11111133 Pillay D, Sinka K, Rice P, Peters B, Clarke J, Workman J (2000) Impact of HIV-1 subtype on NNRTI resistance mutations. Antiviral Therapy 5:128 Pillay D, Walker AS, Gibb DM et al. (2002) Im pact of human immunodeficiency virus type 1 subtypes on virologic response and emergen ce of drug resistance among children in the Paediatric European Network for Treatment of AIDS (PENTA) 5 trial. J Infect Dis. 186:617-625. Epub 2002 Aug 2009 Pires IL, Soares MA, Speranza FA et al. (2004) Prevalence of human immunodeficiency virus drug resistance mutations and subtypes in drug -naive, infected individuals in the army health service of Rio de Janeiro, Brazil. J Clin Microbiol 42:426-430 Platt EJ, Haffar OK (1994) Characterization of human immunodeficiency virus type 1 Pr55gag membrane association in a cell-free system: requirement for a C-terminal domain. Proc Natl Acad Sci U S A 91:4594-4598 Pollard VW, Malim MH (1998) The HIV-1 Re v protein. Annu Rev Microbiol 52:491-532 Poon DT, Li G, Aldovini A (1998) Nucleocapsid and matrix protei n contributions to selective human immunodeficiency virus type 1 genom ic RNA packaging. J Virol. 72:1983-1993 Pope M, Frankel SS, Mascola JR et al. (1997) Hu man immunodeficiency vi rus type 1 strains of subtypes B and E replicate in cutaneous dendr itic cell-T-cell mixtur es without displaying subtype-specific tropism. J Virol. 71:8001-8007
201 Prabu-Jeyabalan M, Nalivaika E, Schiffer CA (2000) How does a symmetric dimer recognize an asymmetric substrate? A substrate comp lex of HIV-1 protease. J Mol Biol. 301:12071220 Prabu-Jeyabalan M, Nalivaika EA, King NM, Schi ffer CA (2003) Viabilit y of a drug-resistant human immunodeficiency virus ty pe 1 protease variant: struct ural insights for better antiviral therapy. J Virol 77:1306-1315 Quinn TC, Wawer MJ, Sewankambo N et al. (2002) Viral load and heterosexual transmission of human immunodeficiency virus type 1. N Engl J Med 2000; 342: 921-929 Reicin AS, Ohagen A, Yin L, Hoglund S, Goff SP (1996) The role of Gag in human immunodeficiency virus type 1 virion morphogenesis and earl y steps of the viral life cycle. J Virol. 70:8645-8652 Renjifo B, Gilbert P, Chaplin B et al. (2004) Pr eferential in-utero transm ission of HIV-1 subtype C as compared to HIV-1 subt ype A or D. Aids. 18:1629-1636 Richards AD, Phylip LH, Farmerie WG et al. (1 990) Sensitive, soluble chromogenic substrates for HIV-1 proteinase. J Biol Chem. 265:7733-7736 Rick SW, Erickson JW, Burt SK (1998) Reactio n path and free energy calculations of the transition between alternate conformations of HIV-1 protease. Proteins 32:7-16 Ringhofer S, Kallen J, Dutzler R et al. (1999) X-ray structure and conf ormational dynamics of the HIV-1 protease in complex with the inhibitor SDZ283-910: agreement of timeresolved spectroscopy and molecular dynami cs simulations. J Mol Biol 286:1147-1159 Robertson DL, Anderson JP, Bradac JA et al. (2000) HIV-1 nomenclature proposal. Science 288:55-56 Rodenburg CM, Li Y, Trask SA et al. (2001) Near full-length clones and reference sequences for subtype C isolates of HIV type 1 from three different continents. AIDS Res Hum Retroviruses 17:161-168 Roge BT, Katzenstein TL, Nielse n HL, Gerstoft J (2003) Drug resi stance mutations and outcome of second-line treatment in patients with firs t-line protease inhibito r failure on nelfinavircontaining HAART. HIV Med. 4:38-47 Romano L, Venturi G, Ferruzzi R et al. (2000) Detection of ge notypically drug-resistant HIV-1 variants and non-B subtypes in recently infected antiretroviral-naive ad ults in Italy. Aids 14:2204-2206 Rose RE, Gong YF, Greytok JA et al. (1996) Human immunodeficiency virus type 1 viral background plays a major role in development of resistance to prot ease inhibitors. Proc Natl Acad Sci U S A 93:1648-1653
202 Rossmann MG (1990) The molecular replacemen t method. Acta Crystallogr A. 46:73-82 Rusconi S, La Seta Catamancio S, Citterio P et al. (2000) Suscep tibility to PNU-140690 (Tipranavir) of human immunodeficiency virus t ype 1 isolates derived from patients with multidrug resistance to other protease inhibitors. Antimicrob Agents Chemother. 44:1328-1332 Rutenber E, Fauman EB, Keenan RJ et al. (1993) Structure of a non-pept ide inhibitor complexed with HIV-1 protease. Developi ng a cycle of structure-base d drug design. J Biol Chem 268:15343-15346 Saksena N (1998) HUMAN IMMUNODEFICIENCY VIRUSES. Biology, Immunology and Molecular Biology. Medical System S.p.A., Westmead Samson M, Libert F, Doranz BJ et al. (1996) Resistance to HIV-1 in fection in caucasian individuals bearing mutant alleles of th e CCR-5 chemokine receptor gene. Nature. 382:722-725 Sanches M, Krauchenco S, Martins NH, Gu stchina A, Wlodawer A, Polikarpov I (2007) Structural characterization of B and non-B subtypes of HIVprotease: insights into the natural susceptibility to drug resistance developmen t. J Mol Biol. 369:1029-1040. Epub 2007 Mar 1024 Sanches M, Martins NH, Calazans A et al. (2004) Crystallization of a non-B and a B mutant HIV protease. Acta Crystallogr D Biol Crystallogr 60:1625-1627 Scarlata S, Ehrlich LS, Cart er CA (1998) Membrane-induced alterations in HIV-1 Gag and matrix protein-protein intera ctions. J Mol Biol. 277:161-169 Schieferstein C, Buhk T (2005) HAART Side E ffects. In: Hoffmann C, Kamps BS (eds) HIV Medicine. Flying Publisher Schroder AR, Shinn P, Chen H, Berry C, Ecker JR, Bushman F (2002) HIV1 integration in the human genome favors active genes a nd local hotspots. Cell. 110:521-529 Scott WR, Schiffer CA (2000) Curling of flap tips in HIV-1 protease as a mechanism for substrate entry and tolerance of dr ug resistance. Structure. 8:1259-1265 Servais J, Lambert C, Fontaine E et al. (2001) Variant human immunodeficiency virus type 1 proteases and response to combination thera py including a protease inhibitor. Antimicrob Agents Chemother 45:893-900 Shafer RW, Chuang TK, Hsu P, White CB, Katzenstein DA (1999) Sequence and drug susceptibility of subtype C protease from human immunodeficiency virus type 1 seroconverters in Zimbabwe. AIDS Res Hum Retroviruses 15:65-69
203 Shafer RW, Jung DR, Betts BJ, Xi Y, Gonzal es MJ (2000) Human i mmunodeficiency virus reverse transcriptase and protease sequence database. Nucleic Acids Res 28:346-348 Shafer RW, Rhee SY, Pillay D et al. (2007) HIV-1 protease and re verse transcriptase mutations for drug resistance surve illance. Aids. 21:215-223 Shehu-Xhilaga M, Crowe SM, Mak J (2001) Mainte nance of the Gag/Gag-Pol ratio is important for human immunodeficiency virus type 1 R NA dimerization and viral infectivity. J Virol. 75:1834-1841 Sheldrick GM (1997) The SHELX97 Manual. Univ ersity of Gottingen, Gottingen, Germany. Accessed from http://shelx.uni-ac.gw dg.de/SHELX/index.html, on June 2007 Sheldrick GM, Schneider TR ( 1997) Methods in Enzymology. Methods in Enzymology:319-343 Simon F, Mauclere P, Roques P et al. (1998) Identification of a new human immunodeficiency virus type 1 distinct from gr oup M and group O. Nat Med 4:1032-1037 Soares MA, De Oliveira T, Brindeiro RM et al. (2003) A specific subtype C of human immunodeficiency virus type 1 circ ulates in Brazil. Aids 17:11-21 Somogyi P, Jenner AJ, Brierley I, Inglis SC (1993) Ribosomal pa using during translation of an RNA pseudoknot. Mol Ce ll Biol. 13:6931-6940 Spearman P, Horton R, Ratner L, Kuli-Za de I (1997) Membrane binding of human immunodeficiency virus type 1 matrix protein in vivo supports a conformational myristyl switch mechanism. J Virol. 71:6582-6592 Spearman P, Ratner L (1996) Human immunodefi ciency virus type 1 capsid formation in reticulocyte lysates. J Virol. 70:8187-8194 Spinelli S, Liu QZ, Alzari PM, Hirel PH, Poljak RJ (1991) The three-dime nsional structure of the aspartyl protease from the HIV1 isolate BRU. Biochimie. 73:1391-1396 Spira S, Wainberg MA, Loemba H, Turner D, Brenner BG (2003) Impact of clade diversity on HIV-1 virulence, antiretroviral drug sens itivity and drug resistance. J Antimicrob Chemother 51:229-240 Stanford U (2007) HIV Drug Resistance Database Stanford University, California, USA. Accessed from http://hivdb.stanford.edu, on June 2007 Stephen AG, Worthy KM, Towler E et al. (2002) Identification of HIV-1 nucleocapsid protein: nucleic acid antagonists with cellular anti-HIV activity. Biochem Biophys Res Commun. 296:1228-1237
204 Strack B, Calistri A, Craig S, Popova E, Go ttlinger HG (2003) AIP1/ALIX is a binding partner for HIV-1 p6 and EIAV p9 functioning in virus budding. Cell. 114:689-699 Stuart DT, Upton C, Higman MA, Niles EG McFadden G (1993) A poxvirus-encoded uracil DNA glycosylase is essential for vi rus viability. J Virol. 67:2503-2512 Sugiura W, Matsuda Z, Yokoma ku Y et al. (2002) Interferen ce between D30N and L90M in selection and development of protease inhibitorresistant human immu nodeficiency virus type 1. Antimicrob Agents Chemother. 46:708-715 Takeb EY, Kusagawa S, Motomura K (2004) Mo lecular epidemiology of HIV: tracking AIDS pandemic. Pediatr Int. 46:236-244 Tamiya S, Mardy S, Kavlick MF, Yoshimura K, Mistuya H (2004) Amino acid insertions near Gag cleavage sites restore the otherwis e compromised replication of human immunodeficiency virus type 1 variants re sistant to protease i nhibitors. J Virol. 78:12030-12040 Tanuri A, Vicente AC, Otsuki K et al. (1999) Genetic variation and susceptibilities to protease inhibitors among subtype B a nd F isolates in Brazil. Antimicrob Agents Chemother 43:253-258 Tapia N, Franco S, Puig-Basagoiti F et al. (2003) Influence of huma n immunodeficiency virus type 1 subtype on mother-to-child transmission. J Gen Virol. 84:607-613 Tomasselli AG, Heinrikson RL (1994) Specificity of retroviral proteas es: an analysis of viral and nonviral protein substrates. Methods Enzymol 241:279-301 Tomasselli AG, Sarcich JL, Barrett LJ et al. (1993) Human immunodeficiency virus type-1 reverse transcriptase and ribonucl ease H as substrates of the viral protease. Protein Sci. 2:2167-2176 Toth G, Borics A (2006) Flap opening mechan ism of HIV-1 protease. J Mol Graph Model. 24:465-474. Epub 2005 Sep 2026 Tozser J, Gustchina A, Weber IT, Blaha I, Wondr ak EM, Oroszlan S (1991) Studies on the role of the S4 substrate binding site of HIV proteinases. FEBS Lett. 279:356-360 Tritch RJ, Cheng YE, Yin FH, Erickson-Viitanen S (1991) Mutagenesis of protease cleavage sites in the human immunodefi ciency virus type 1 gag pol yprotein. J Virol. 65:922-930 Turner BG, Summers MF (1999) Structural biology of HIV. J Mol Biol. 285:1-32 U.S. Department of Health and Human Services PoCPfToHI (2006) Guide lines for the use of antiretroviral agents in HIV-1 infected adults and adolesce nts. Accessed from www.aidsinfo.nih.gov on June 2007
205 UNAIDS/WHO (2006) AIDS epidemic update December 2006. UNAIDS/WHO. Accessed from http://www.unaids.org, on June 2007 Velazquez-Campoy A, Todd MJ, Vega S, Freire E (2001) Catalytic efficiency and vitality of HIV-1 proteases from African viral subtype s. Proc Natl Acad Sci U S A 98:6062-6067 Velazquez-Campoy A, Vega S, Fl eming E et al. (2003) Protease inhibition in African subtypes of HIV-1. AIDS Rev 5:165-171 Velazquez-Campoy A, Vega S, Fr eire E (2002) Amplification of the effects of drug resistance mutations by background polymorphisms in HIV-1 protease from African subtypes. Biochemistry 41:8613-8619 Veldkamp AI, van Heeswijk RP, Mulder JW et al. (2001) Steady-state pharmacokinetics of twice-daily dosing of saquinavir plus ritonavir in HIV-1-infected individuals. J Acquir Immune Defic Syndr. 27:344-349 Vergne L, Peeters M, Mpoudi-Ngol e E et al. (2000) Genetic dive rsity of protease and reverse transcriptase sequences in nonsubtype-B human immunodeficien cy virus type 1 strains: evidence of many minor drug resistance mutati ons in treatment-naive patients. J Clin Microbiol 38:3919-3925 Vergne L, Stuyver L, Van Houtte M, Butel C, Delaporte E, Peeters M (2006) Natural polymorphism in protease and reverse transcri ptase genes and in vitro antiretroviral drug susceptibilities of non-B HIV-1 strains from tr eatment-naive patients. J Clin Virol. 36:4349. Epub 2006 Mar 2024 Verheyen J, Litau E, Sing T et al. (2006) Co mpensatory mutations at the HIV cleavage sites p7/p1 and p1/p6-gag in therapy-naive and th erapy-experienced patients. Antivir Ther 11:879-887 von Schwedler UK, Stuchell M, Muller B et al (2003) The protein network of HIV budding. Cell. 114:701-713 Vondrasek J, Wlodawer A (2002) HIVdb: a database of the structures of human immunodeficiency virus prot ease. Proteins 49:429-431 Wainberg MA, Drosopoulos WC, Sa lomon H et al. (1996) Enhanced fidelity of 3TC-selected mutant HIV-1 reverse transc riptase. Science 271:1282-1285 Walker PR, Pybus OG, Rambaut A, Holmes EC (2005) Comparative pop ulation dynamics of HIV-1 subtypes B and C: subt ype-specific differences in patterns of epidemic growth. Infect Genet Evol 5:199-208. Epub 2004 Dec 2007
206 Wallace AC, Laskowski RA, Thornton JM (1995) LI GPLOT: a program to generate schematic diagrams of protein-ligand inte ractions. Protein Eng. 8:127-134 Weber IT, Miller M, Jaskolski M, Leis J, Skalka AM, Wlodawer A ( 1989) Molecular modeling of the HIV-1 protease and its substr ate binding site. Science. 243:928-931 Weidle PJ, Malamba S, Mwebaze R et al. (2002) Assessment of a pilo t antiretroviral drug therapy programme in Uganda: patients' res ponse, survival, and drug resistance. Lancet. 360:34-40 Wester CW, Kim S, Bussmann H et al. (2005) Initial response to highly active antiretroviral therapy in HIV-1C-infected a dults in a public sector treatment program in Botswana. J Acquir Immune Defic Syndr. 40:336-343 Wiegers K, Rutter G, Kottler H, Tessmer U, Hohenberg H, Kraussl ich HG (1998) Sequential steps in human immunodeficiency virus partic le maturation revealed by alterations of individual Gag polypr otein cleavage site s. J Virol 72:2846-2854 Williams JW, Morrison JF (1979) The kinetics of reversible tight-binding inhibition. Methods Enzymol 63:437-467 Wlodawer A, Erickson JW (1993) Structure-based inhibitors of HIV-1 protease. Annu Rev Biochem 62:543-585 Wlodawer A, Miller M, Jaskolski M et al. (1989) Conserved folding in retroviral proteases: crystal structure of a synthetic HIV-1 protease. Science 245:616-621 Wondrak EM, Louis JM (1996) Influence of fla nking sequences on the dimer stability of human immunodeficiency virus type 1 protease. Biochemistry. 35:12957-12962 Wu TD, Schiffer CA, Gonzales MJ et al. (2003) Mutation patterns and structural correlates in human immunodeficiency virus type 1 protease following di fferent protease inhibitor treatments. J Virol. 77:4836-4847 Wyma DJ, Kotov A, Aiken C (2000) Evidence for a stable interaction of gp41 with Pr55(Gag) in immature human immunodeficiency virus type 1 particles. J Virol. 74:9381-9387 Xie D, Gulnik S, Gustchina E et al. (1999) Drug resistance mutations can affect dimer stability of HIV-1 protease at neutra l pH. Protein Sci. 8:1702-1707 Yang C, Li M, Newman RD et al. (2003) Genetic diversity of HIV-1 in western Kenya: subtypespecific differences in mother-tochild transmission. Aids. 17:1667-1674 Yu XF, Chen J, Shao Y, Beyrer C, Lai S (1998a) Two subtypes of HI V-1 among injection-drug users in southern China. Lancet 351:1250
207 Yu XF, Dawson L, Tian CJ, Flexner C, Dette nhofer M (1998b) Muta tions of the human immunodeficiency virus type 1 p6Gag domain re sult in reduced reten tion of Pol proteins during virus assembly. J Virol. 72:3412-3417 Yuan X, Yu X, Lee TH, Essex M (1993) Mu tations in the N-terminal region of human immunodeficiency virus type 1 matrix protein block intrace llular transport of the Gag precursor. J Virol. 67:6387-6394 Zennou V, Mammano F, Paulous S, Mathez D, Clavel F (1998) Loss of vira l fitness associated with multiple Gag and Gag-Pol processing defects in human immunodeficiency virus type 1 variants selected for resistance to protease inhibitors in vivo. J Virol 72:3300-3306 Zhang H, Orti G, Du Q et al. (2002) Phylogenetic and phenotypic analysis of HIV type 1 env gp120 in cases of subtype C mother-to-child transmission. AIDS Res Hum Retroviruses 18:1415-1423 Zhang YM et al. (1997) Drug resi stance during indinavir therapy is caused by mutations in the protease gene and in its Gag substrat e cleavage sites. J Virol. 71:6662-6670 Zhou J, Imamichi H, Imamichi T et al. (2004) Small-molecule inhibition of human immunodeficiency virus type 1 re plication by specific targeting of the fi nal step of virion maturation. J Virol. 78:922-929 Zhou W, Parent LJ, Wills JW, Resh MD (1994) Identification of a membrane-binding domain within the amino-terminal region of human immunodeficiency virus type 1 Gag protein which interacts with acidic phos pholipids. J Virol. 68:2556-2569 Ziermann R, Limoli K, Das K, Arnold E, Petrop oulos CJ, Parkin NT (2000) A mutation in human immunodeficiency virus type 1 pr otease, N88S, that causes in vitro hypersensitivity to amprenavir. J Virol 74:4414-4419 Zybarth G, Carter C (1995) Domains upstream of the protease (PR) in human immunodeficiency virus type 1 Gag-Pol influence PR autoprocessing. J Virol 69:3878-3884 Zybarth G, Krausslich HG, Partin K, Carter C (1994) Proteolytic activity of novel human immunodeficiency virus type 1 proteinase proteins from a precursor with a blocking mutation at the N terminus of the PR domain. J Virol 68:240-250
208 BIOGRAPHICAL SKETCH Roxana Maria Coman was born in August 1976 to Victoria and Florin Coman, and grew up in Barlad, Vaslui, Romania, with her sister Anca Cristina. She received her MD from Carol Davila University of Medicine and Pharmacy, Bucharest, Romania, in September 2000. Roxana entered the Interdisciplinary Program in Biomed ical Sciences at the College of Medicine, University of Florida in the summer of 2003. One year later she joined th e laboratory of Dr. Ben M. Dunn and began training towards her doctorate After her graduation, she plans to pursue a dual career as a clinic ian and researcher.