Enzymatic Characterization, Subsite Specificity Exploration, and Inhibition Analysis of Malarial Aspartic Proteinases Using Enzyme Kinetics, Combinatorial Chemistry, and Molecular Modeling

Material Information

Enzymatic Characterization, Subsite Specificity Exploration, and Inhibition Analysis of Malarial Aspartic Proteinases Using Enzyme Kinetics, Combinatorial Chemistry, and Molecular Modeling
Copyright Date:


Subjects / Keywords:
Active sites ( jstor )
Amino acids ( jstor )
Antimalarials ( jstor )
Catalysis ( jstor )
Enzymes ( jstor )
Libraries ( jstor )
Malaria ( jstor )
Parasites ( jstor )
pH ( jstor )
Vacuoles ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright the author. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:


This item is only available as the following downloads:

Full Text




Copyright 2006 by Peng Liu


This document is dedicated to my parents for their endless love.


iv ACKNOWLEDGMENTS I am indebted to Professor Ben M. Dunn fo r his invaluable scientific guidance. His brilliant personality and prude nt attitude toward science set up an excellent example throughout my graduate career. I would like to thank the members of my supervisory committee, Dr. John B. Dame, Dr. Susan C. Frost, Dr. Robert McKenna and Dr. Julie Maupin-Furlow, for their guidance and expert ise. Their helpful suggestions greatly support the research work presented. I would like to thank my colleagues in the Dunn Lab for their kindness and scientific and spiritual suppor t, especially Dr. Jos C. Clemente, Dr. Rebecca E. MooseClemente and Dr. Patrica O O’Brien, who “d ecoyed” me into this special family. I would like to thank th e people who provided tec hnical support during my graduate research. These folks are Mr. Char les A. Yowell, Dr. Jorge F. Bonilla, Dr. Stanley M. Stevens, Jr., Mr. Scott H. McClung, Mr. Alfred Y. Chung and Ms. Ran Zheng.


v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES...........................................................................................................xi ABBREVIATIONS...........................................................................................................xv ABSTRACT.......................................................................................................................xx CHAPTER 1 INTRODUCTION........................................................................................................1 Malaria........................................................................................................................ ..1 History of Malaria.................................................................................................1 Current Status, Prophylaxis and Treatment...........................................................2 Malaria Parasite..........................................................................................................12 Causative Agent and Life Cycle..........................................................................12 Drug Resistance...................................................................................................13 Plasmepsins.................................................................................................................16 Genetic Family....................................................................................................16 Physiological Functions.......................................................................................19 Potential Targets of Novel Antimalarials............................................................22 Characterization of Naturally-occu rring and Recombinant Plasmepsin.............25 Combinatorial Chemistry............................................................................................28 2 MATERIALS AND METHODS...............................................................................33 Introduction.................................................................................................................33 Materials.....................................................................................................................34 Methods......................................................................................................................36 Production of Zymogen and the Ma ture Form of Recombinant Pf PM1 and Pb PM4.............................................................................................................36 Site-directed mutations of pro Pb PM4..........................................................36 Expression of recombinant pro Pf PM1 and pro Pb PM4................................39 Inclusion body preparation...........................................................................39 Pro Pf PM1 inclusion body denaturation and in vitro refolding....................40


vi Pro Pb PM4 inclusion body denaturation and in vitro refolding...................41 Purification of pro Pf PM1 and pro Pb PM4....................................................42 Mature Pf PM1 and Pb PM4 preparation and purification............................42 Characterization of Zymogen and th e Mature Form of Recombinant Pf PM1 and Pb PM4.......................................................................................................43 Self-processing and catalysis optimization of pro Pf PM1 and pro Pb PM4...43 Catalysis optimization of mature Pf PM1 and Pb PM4.................................44 N-terminal sequencing analysis...................................................................45 Kinetic parameter determination..................................................................46 Combinatorial Chemistry Based Subsite Preference Studies on Pf PM1 and Pb PM4.............................................................................................................47 Combinatorial libraries.................................................................................47 Primary subsite preferen ces—Spectroscopic assays....................................48 Secondary subsite preferences—LC-MS.....................................................48 Data processing............................................................................................51 Inhibition analyses........................................................................................53 Inhibition Studies of Potential Antimalarials......................................................57 Enzyme and inhibitor preparation................................................................57 Dissociation constant ( Ki) measurement......................................................57 Molecular modeling.....................................................................................57 In vitro antiparasitic activity assays.............................................................59 3 PRODUCTION AND ENZYMATIC CHARACTERIZATION OF RECOMBINANT PLASMEPSIN 1 ( Pf PM1) FROM Plasmodium falciparum ........61 Introduction.................................................................................................................61 Results........................................................................................................................ .64 Generation of the Recombinant Pro Pf PM1 K110pN Mutant..............................64 Expression, in vitro Refolding and Purification of the Recombinant Pro Pf PM1 K110pN..........................................................................................65 Optimal Conditions for the Catalysis of the Pf PM1 K110pN Mutant................70 Auto-activation of the Recombinant Pro Pf PM1 K110pN Mutant......................70 Isolation of Mature Pf PM1 and Screening its Optimal pH Condition for Catalysis...........................................................................................................73 Catalytic Efficiency of Mature Pf PM1 Hydrolysis on Chromogenic Peptide Substrates.........................................................................................................76 Inhibition Effects of Compounds on Pf PM1.......................................................76 Discussion...................................................................................................................79 Conclusion..................................................................................................................83 4 PRODUCTION AND ENZYMATIC CHARACTERIZATION OF RECOMBINANT PLASMEPSIN 4 ( Pb PM4) FROM Plasmodium berghei ............85 Introduction.................................................................................................................85 Results........................................................................................................................ .86 Cloning of the Recombinant Wild-type Pro Pb PM4............................................86 Expression, in vitro Refolding, and Purification of Recombinant Pro Pb PM4....89


vii Self-processing of Recombinant Pro Pb PM4.......................................................93 Optimal Conditions for Pb PM4 Catalysis...........................................................96 Isolation and Characterization of Mature Pb PM4...............................................99 Comparison of Kinetic Parameters of Pb PM4 Species.......................................99 Inhibition Analyses of Compounds Bound to Pb PM4......................................103 Discussion.................................................................................................................104 Conclusion................................................................................................................107 5 EXPLORATION OF SUBSITE PREFERENCES OF Pf PM1 AND Pb PM4 USING COMBINATORIAL CHEMISTR Y BASED PEPTIDE LIBRARIES.......109 Introduction...............................................................................................................109 Results.......................................................................................................................113 Subsite Preferences............................................................................................113 Primary specificity.....................................................................................113 Secondary specificity.................................................................................119 Inhibitor Design.................................................................................................131 Characterization of Single Peptidomimetic Inhibitors......................................131 Kinetic analyses..........................................................................................131 Molecular modeling...................................................................................133 Discussion.................................................................................................................135 Conclusion................................................................................................................140 6 HIV-1 CLINICAL PROTEASE INHI BITOR–INHIBITION ANALYSES ON PLASMEPSINS AND ANTIPARASITIC ACTIVITIES ON Plasmodium falciparum CULTURE.............................................................................................143 Introduction...............................................................................................................143 Results.......................................................................................................................144 Kinetic Analyses................................................................................................146 Molecular Modeling..........................................................................................148 In vitro Antiparasitic Activity...........................................................................150 Discussion.................................................................................................................159 Conclusion................................................................................................................164 7 INHIBITION ANALYSES OF PRIM AQUINE-STATINE “DOUBLE DRUG” COMPOUNDS AGAINST PLASMEPSINS...........................................................166 Introduction...............................................................................................................166 Results.......................................................................................................................170 Enzyme Inhibition.............................................................................................170 Molecular Modeling..........................................................................................173 Discussion.................................................................................................................177 Conclusion................................................................................................................179 8 FUTURE DIRECTIONS..........................................................................................181


viii LIST OF REFERENCES.................................................................................................199 BIOGRAPHICAL SKETCH...........................................................................................220


ix LIST OF TABLES Table page 1-1 Potential factors for th e resurgence of malaria...........................................................5 1-2 X-ray crystal struct ures of plasmepsins...................................................................29 2-1 Primers for the site-directed mutations of pro Pb PM4.............................................37 2-2 Cycling parameters for site d-directed mutagenesis of pro Pb PM4...........................37 2-3 Normalization factors to calibrate the quantities of di gestion products of octapeptide library pools..........................................................................................55 3-1 Average representative yields duri ng the production and pur ification of the recombinant pro Pf PM1 from 1 liter culture of expression......................................67 3-2 Kinetic parameters of na turally-occurring and recombinant Pf PM1 digestion on chromogenic peptide substrates...............................................................................77 3-3 Dissociation constants of inhibitory compounds bound to Pf PM1..........................78 4-1 Average representative yields duri ng the production and pur ification of the recombinant wild type pro Pb PM4 from 1 liter culture of expression......................91 4-2 Comparison of kinetic parameters of Pb PM4 species digestion on chromogenic peptide substrates...................................................................................................102 4-3 Dissociation constants of inhibitory compounds bound to Pb PM4.......................103 5-1 Initial cleavage velociti es (AU/sec) of the P1 comb inatorial peptide library pools by Pf PM1 and Pb PM4............................................................................................115 5-2 Initial cleavage velociti es (AU/sec) of the P1’ combinatorial peptide library pools by Pf PM1 and Pb PM4..................................................................................116 5-3 Optimal peptide sequence for Pf PM1 and Pb PM4 determined from analyses of the P1 and P1’ combinatorial libraries...................................................................131 5-4 The inhibitory activities of peptidomimetic inhibitors on Pf PM1 and Pb PM4.....134


x 5-5 Amino acid residues that comprise th e S3-S3’ subsite pockets of human and malaria aspartic proteinase series...........................................................................141 6-1 Inhibition constants of clin ical HIV-1 PIs on plasmepsins....................................147 6-2 Amino acid residues of Pf PM4 involving in hydrophobic interactions with APV and RTV.................................................................................................................158 6-3 The anti-parasitic activities (IC50) of clinical HIV-1 prot ease inhibitors on the P. falciparum 3D7 strain.............................................................................................160 6-4 Experimental and calculated free ener gies for the binding of APV and RTV to Pf PM4.....................................................................................................................162 7-1 The dissociation constants of PQ-statine “double-drug” compounds on plasmepsin inhibition.............................................................................................174


xi LIST OF FIGURES Figure page 1-1 Barks of Quinquina calisaya (from Bolivia)..............................................................3 1-2 Stamps highlighting the WHO ma laria eradication campaign (1955-1978)..............4 1-3 Geographic distribution of malaria............................................................................6 1-4 The Biological larvicides—the Guppy and Gambusia fish........................................8 1-5 Examples of antimalarial drugs................................................................................11 1-6 The life cycle of human malaria parasites................................................................14 1-7 The amino acid sequence features of plasmepsins of the human malaria parasite P. falciparum ............................................................................................................18 1-8 The hemoglobin degradation pathway in Plasmodium falciparum .........................21 1-9 Structures of plasmepsin inhibitors..........................................................................24 1-10 Combinatorial inhibitor libraries for sc reening highly selective plasmepsin 2........31 2-1 Overview of the QuikChange site-directed mutagenesis method..........................38 2-2 The HPLC gradient elution program for separation of cleavage peptides...............50 2-3 The elution order of enzymatic dige stion products during the separation by reverse phase HPLC.................................................................................................52 2-4 LC-MS analyses of Pf PM1 digested P1F library.....................................................55 3-1 Nucleotide and protein sequence of the full length pro Pf PM1................................62 3-2 Molecular modeling structure of mature Pf PM1......................................................63 3-3 Sequence alignment of the Pf PM1 and Pf PM2 prosegment portion that potentially binds in the active site cleft....................................................................66 3-4 Gel filtration purification chromatogram of the pro Pf PM1 K110pN mutant..........68


xii 3-5 SDS-PAGE analysis of overexpre ssion and purification of the pro Pf PM1 K110pN mutant........................................................................................................69 3-6 Determination of the optimal conditions for the catalysis of Pf PM1.......................71 3-7 SDS-PAGE analysis of time-reso lved auto-activation of the pro Pf PM1 K110pN mutant.......................................................................................................................72 3-8 Determination of the optimal pH condition for the catalysis of mature Pf PM1......74 3-9 N-terminal protein sequenci ng analyses on the auto-converted Pf PM1..................75 3-10 SDS-PAGE analysis of purified recombinant mature Pf PM1 materials for crystal tray setup..................................................................................................................82 4-1 Nucleotide and protein sequence of pro Pb PM4.......................................................87 4-2 Molecular modeling structure of mature Pb PM4.....................................................88 4-3 Site-directed mutagenesis of pro Pb PM4..................................................................89 4-4 SDS-PAGE analysis of overexpression and purification of recombinant wild type pro Pb PM4........................................................................................................90 4-5 Size exclusion purification chro matogram of the wild type pro Pb PM4..................92 4-6 SDS-PAGE analysis of time-resolved self-processing of wild type pro Pb PM4.....94 4-7 SDS-PAGE analysis of time-reso lved self-processing of the pro Pb PM4 L117pE mutant.......................................................................................................................95 4-8 Determination of the optimal conditi ons for the catalysis of wild type Pb PM4......97 4-9 Determination of the optimal c onditions for the catalysis of the Pb PM4 L117pE mutant.......................................................................................................................98 4-10 pH-dependent catalytic activity profile of mature Pb PM4....................................100 4-11 N-terminal protein sequencing analyses on self-processing of pro Pb PM4...........101 5-1 Schematic diagram of a peptide substrat e fitting to the active site cleft of an endopeptidase.........................................................................................................110 5-2 Schematic diagram of the enzymatic dige stion of individual P1 and P1’ library pools.......................................................................................................................112 5-3 The P1 amino acid preferences of the malarial aspartic proteinases, Pf PM1 (A) and Pb PM4 (B).......................................................................................................117


xiii 5-4 The P1’ amino acid preferences of the malarial aspartic proteinases, Pf PM1 (A) and Pb PM4 (B).......................................................................................................118 5-5 The P3 amino acid preferences of the malarial aspartic proteinase Pf PM1...........121 5-6 The P3 amino acid preferences of the malarial aspartic proteinase Pb PM4..........122 5-7 The P2 amino acid preferences of the malarial aspartic proteinase Pf PM1...........124 5-8 The P2 amino acid preferences of the malarial aspartic proteinase Pb PM4..........125 5-9 The P2’ amino acid preferences of the malarial aspartic proteinase Pf PM1.........127 5-10 The P2’ amino acid preferences of the malarial aspartic proteinase Pb PM4.........128 5-11 The P3’ amino acid preferences of the malarial aspartic proteinase Pf PM1.........129 5-12 The P3’ amino acid preferences of the malarial aspartic proteinase Pb PM4.........130 5-13 Structures of the inhibitors desi gned from the combinatorial approach................132 5-14 Molecular modeling of the Pf PM1-compound 1 complex (A) and the Pb PM4compound 3 complex (B).......................................................................................136 6-1 Structures of FDA-appr oved clinical HIV-1 PIs....................................................145 6-2 Superimposition of crystallographi c structures of HIV-1 protease and Pf PM4.....151 6-3 The Pf PM4-APV molecular model........................................................................152 6-4 The Pf PM4-RTV molecular model........................................................................154 7-1 Schematic diagram showing the catalyti c mechanism of aspartic proteinases......169 7-2 General structural formula of the primaquine-statine “double-drug” compounds.169 7-3 Structure of the modeled Pf PM4-compound 2 complex........................................178 8-1 Amino acid sequence alignment of Pf PM1 and Pf PM4 and surface residue mutations on Pf PM1...............................................................................................183 8-2 DNA shuffling........................................................................................................184 8-3 Complementation of the split GFP fragment.........................................................186 8-4 GFP fragment 1-10 OPT expression system..........................................................187 8-5 GFP 11 M3 fusion system......................................................................................188


xiv 8-6 The induced expression of r ecombinant GFP 1-10 (A) and pro Pf PM2-GFP 11 fusion protein (B) in E. coli ....................................................................................190 8-7 SDS-PAGE analysis of expression of the full length (A) and semipro Pf HAP (B)192 8-8 Gel filtration purification chromatogram of the recombinant semipro Pf PM2 and semipro Pf HAP.......................................................................................................194 8-9 Far UV circular dichoism (CD) spectrum of the semipro Pf HAP..........................195 8-10 Diagram of two combinatorial librarie s that explore the influences between subsites of different domains..................................................................................196


xv ABBREVIATIONS Angstrom A. Anopheles ABC ATP-binding-cassette ACT artemisinin-based combination therapies AIDS acquired immunodeficiency syndrome AnTet anhydrotetracycline AU absorption unit AUC the area under the curve BCE before Common Era CaCl2 calcium chloride CCI combinatorial chemistry inhibitor CD circular dichroism CDC centers for disease and control prevention CE Common Era cm centimeter cpm counts per minute DDT dichloro-diphenyl-trichloroethane DMSO dimethyl sulfoxide E-64 L-3-carboxy-2,3-trans-epoxypropion yl-leucylamido(4guanidino)butane E. coli Escherichia coli


xvi EDTA ethylene diaminotetraacetic acid ER endoplasmic reticulum methyleneamino ([-CH2-NH-]) FPLC fast protein liquid chromatography FV food vacuole g gravitational force GFP green fluorescence protein gm gram HAP H isto-A spartic P roteinase HCl hydrochloric acid HIV-1 human immunodefici ency virus type 1 HP high performance h hour IB inclusion body IC50 50% Inhibition Concentration i.d. inner diameter IPTG isopropylthiogalactopyranoside LSB Laemmli sample buffer kcat turnover number kcat/ Km specificity constant kDa kilo Daltons Ki dissociation/inhibition constant Ki ap apparent dissociati on/inhibition constant


xvii Km Michaelis-Menten constant kV kilovolts LB Luria broth LC liquid chromatography M molar Ci microCurrie MCS multiple clone site MD molecular dynamics MES 2-(4-morpholino)-ethane sulfonic acid g microgram mg milligram MgCl2 magnesium chloride mL milliliter m micrometer mm millimeter mM millimolar MS mass spectrometry MWCO molecular weight cut off NaCl sodium chloride NIH the National Institutes of Health nL nanoliter nm nanometer nM nanomolar


xviii Nph pnitrophenylalanine OD280 optical density at 280 nanometers OD600 optical density at 600 nanometers P. Plasmodium PAGE polyacrylamide gel electrophoresis Pb Plasmodium berghei Pb PM4 Plasmodium berghei plasmepsin 4 PCR polymerase chain reaction Pf Plasmodim falciparum pfcrt Plasmodim falciparum chloroquine-resistance transporter pfmdr 1 Plasmodium falciparum multidrug-resistance 1 Pf PM1 Plasmodium falciparum plasmepsin 1 pH negative log of the hydrogen ion concentration PI protease inhibitor pM picomolar Pm Plasmodim malariae PM plasmepsin PMSF phenylmethylsulfonyl fluoride Po Plasmodim ovalae PQ primaquine pro Pb PM4 semi-zymogen form of Plasmodium berghei plasmepsin 4 pro Pf PM1 semi-zymogen form of Plasmodium falciparum plasmepsin 1 psi pound-force per square inch


xix Pv Plasmodim vivax PVDF polyvinylidene difluoride rpHPLC reverse phase high performance liquid chromatography rpm revolutions per minute sec second SDS sodium dodecyl sulfate spp. species plural sqrt the square root of SSI single substrate inhibitor TFA trifluoroacetic acid Tris tris (hydroxymethyl) aminomethane UV ultra-violet V volt(s) Vmax maximum velocity w/v weight per unit volume WHO the World Health Organization


xx 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 ENZYMATIC CHARACTERIZATION, SU BSITE SPECIFICITY EXPLORATION AND INHIBITION ANALYSIS OF MALARI AL ASPARTIC PROTEINASES USING ENZYME KINETICS, COMBINATOR IAL CHEMISTRY AND MOLECULAR MODELING By Peng Liu August 2006 Chair: Ben M. Dunn Major Department: Biochemi stry and Molecular Biology Malaria, one of the greates t scourges of human societ y, claims nearly 3 million lives annually. The causative agent, the protozoa Plasmodium spp., has developed drug resistance increasingly in the last two and a half decades. Food vacuole plasmepsins, a group of aspartic poteinase homologues, have become potential targets for novel antimalarial drug design. Plasmepsin 1 of the human malaria parasite Plasmodium falciparum initiates the essential hemoglobin degradation process to pr ovide nutrient and create space for parasite development. Plasmepsin 4 of the rodent malaria parasite Plasmodium berghei is a potential candidate for assessment of dr ug metabolism in the mouse model. The recombinant zymogen forms of these tw o plasmepsins were overexpressed in E. coli . In vitro refolding and auto-activation of th e proenzyme molecules were optimized. In addition, the best conditions for the catalysi s by the mature enzymes were determined.


xxi The S3-S3’ subsite specificities of the tw o investigated plasmepsins were studied using two chromogenic octapeptide combinator ial libraries. The primary specificities at the S1 and S1’ subsites and the secondary spec ificities at the S3, S2, S2’ and S3’ subsites were determined. Specific inhi bitors of the two plasmepsin s were identified and single peptide inhibitors were s ynthesized. The specificities of these compounds on the two investigated plasmepsins were confirmed by enzyme inhibition and were exhibited by molecular modeling. Evidence indicates the anti-parasitic activit ies of some of the FDA approved HIV-1 protease inhibitors. The inhibiti on constants of all the seven inhibitors against seven food vacuole plasmepsins, the potential targets insi de malarial parasites, were determined, the results were interpreted by mo lecular modeling, and the anti-p arasitic activities of these protease inhibitors were tested against the cultured Plasmodium falciparum . In addition, the inhibition effects of a group of systematically synthesized primaquine-statine double drug compounds on seven food vacuole plasmepsins were studied. Inhibition constants were determin ed and compounds that could block multiple plasmepsin activities with high binding affi nities were identified. Putative binding modes of these tight binding inhibitors were built upon X-ray crystal structures using molecular modeling.


1 CHAPTER 1 INTRODUCTION Malaria History of Malaria Malaria, one of the oldest infectious diseases, has been known to plague mankind for millennia. This disease derives its name fr om the Italian Mal’aria (“bad air”) since it used to be common in the fetid mars hy areas around Rome (Boyd 1949). Malaria supposedly originated from the jungles of Af rica and was spread to the Mediterranean shores, India and South East As ia along with human migration. The history for man to study malaria was documented more than 4000 years ago. In 2700 BCE, several characteristic symptoms of malaria were described in the Nei Ching, the Canon of Medicine, edited by the earliest Chinese Emperor, Huang Ti (Bruce-Chwatt 1988). It was not until the late 17th century that the first detaile d description of the clinical picture of malaria was presented by Mort on (Bruce-Chwatt 1988). As more knowledge on malaria has been mastered, man began to ex plore approaches to treat this disease. During the second century BCE, the Qinghao plan t, also known as the annual or sweet wormwood, was described in the medical treatise, 52 Remedies, found in the Mawangdui Tomb of the Han Dynasty. In 340 CE, the an tifever properties of Qinghao were first described by the alchemist Ge Hong of th e East Jin Dynasty (Bruce-Chwatt 1988). The active ingredient of Qinghao, known as artemisi nin, was isolated by Chinese scientists in 1971. In 1640, Huan del Vego first employed the tincture of the cinchona bark (Figure 11) for treating malaria. The active ingredient of the cinchona bark, quinine, was extracted


2 by Pelletier and Caventou in 1820. Since then, mo re antimalarials, such as chloroquine, mefloquine, pyrimethamine, and dichlo ro-diphenyl-trichloroethane (DDT), have been identified and employed for mala ria prophylaxis and treatment. With the success of these antimalarials as part of the reason, the World Health Organization (WHO) launched an ambitious pl an in 1955 to eradicate malaria worldwide (Figure 1-2). Eradication was successful in na tions with temperate climates and seasonal malaria transmission; however, some nati ons of sub-Saharan Africa were excluded completely from the eradication campai gn. In late 1970s, the emergence of drug resistance, widespread resist ance to available insecticid es, wars and massive population movements, and difficulties in obtaini ng funding from donor countries resulted in abandonment of the malaria eradication project. Current Status, Prophylaxis and Treatment Malaria, successfully eradicated or contro lled in most areas of the world in early 1960s, has resurrected in the past two and a half decades. Part of the reasons for the resurgence of malaria are liste d in Table 1-1. Currently, malaria affects more than 2.4 billion people (more than 40% of the world popul ation) mainly living in the tropical and subtropical region (Figure 1-3). Each year, malaria infects an estimated 300-500 million individuals and causes approximately 100 million clinical cases with a 16% growth rate. Annually, malaria claims 1.5 million lives worl dwide with 85% of d eaths occurring in sub-Sahara Africa (Breman 2001; Greenwood and Mutabingwa 2002). Malaria has been recognized as the No. 1 priority tropical disease of the World Health Organization. Prophylaxis and treatment of malaria can be achieved from all three levels: host, vector and causative agent.


3 Figure 1-1. Barks of Quinquina calisaya (from Bolivia), the active ingredient of which is quinine. Figure is adapted from CDC 2006a.


4 Figure 1-2. Stamps highli ghting the WHO malaria erad ication campaign (1955-1978). Figure is adapted from CDC 2006b.


5 Table 1-1. Potential factors fo r the resurgence of malaria Man made Complacency and laxity in anti mala rial campaigns; conflicts and wars; travel and migrations; deteriorating health systems; poverty; population increase Parasite Drug resistance Vector Insecticide resistance Environmental and climatic change Global warming—increased breeding a nd life span of the insect vector Table is adapted from Greenwood and Mutabingwa 2002.


6 Figure 1-3. Geographic di stribution of malaria. Malaria us ually occurs in regions where environmental conditions allow parasite reproduction in the vector. Thus, malaria is generally restricted to tr opical and subtropical areas (see map). However, this distribution might be aff ected by climatic changes, especially global warming, and population travels a nd immigrations. Figure is adapted from CDC 2006c.


7 Man is the most important link in the ma larial control chain. People living in endemic areas as well as travelers to such ar eas should be encouraged to take protective measures to avoid mosquito bites. These include closing the doors and windows in the evenings; using mosquito repellant lotions, creams, mats or coils; and using bednets, especially insecticide treated bednets. Individuals at high ri sk should be asked to take antimalarials regularly. Early diagnosis and treatment are necessary to reduce mortality and prevent further transmission. Additionall y, complete treatment should be ensured. However, the execution of these measures has no t yet been prospective thus far, partially due to the lack of experienced staffs and shortage of funding and medicines. Malaria is transmitted by female anophelin e mosquitoes. Malaria control in the vector level can start by prev enting development of mosquito larvae. Measures include eliminating water collection so as to prev ent mosquitoes from laying eggs, and using chemicals, such as themiphos and fenthion, and biological larvicid es, such as Guppy and Gambusia fish (Figure 1-4). Space, residual, or combined spraying of insecticide such as organophosphate insecticides and pyrethroids can kill adult mosquitoes. However, the main mosquito vectors, Anopheles gambiae and A. funestus , have already been developing resistance to the pyrethroid type insecticides ; therefore, new effective substitutes are under development. With the aid of the ongoing A. gambiae genome project (Holt et al. 2002; Kaufman et al. 2002; Mongin et al. 2004), substantial progress is being made in the creation of genetical ly modified malaria-resistant mosquitoes (Coleman and Alphey 2004), which, however, are still far from being wildly used. The battle against malaria can certainly be carried out by affecting the life cycle of the parasite so as to block its development and growth and even kill the parasite. The


8 Figure 1-4. The Biological la rvicides—the Guppy and Gambusia fish. Figure is adapted from Kakkilaya 2006.


9 history of recognition of malaria and understand ing the biology of mala ria parasites is the history of discovery of antimalarials. To da te, various types of naturally-occurring and synthesized drugs (Figure 1-5) of high poten cy and specificity have been or are being used to cure malaria or attenuate the severity of malaria symptoms. Quinoline antimalarials originate from quinine (Meshni ck and Dobson 2001). Later its derivatives, such as chloroquine and amodiaquine (O’N eill et al. 1998; Ridley and Hudson 1998; Tilley et al. 2001), were devel oped. For a long time, these compounds have been used as potent drugs to treat malaria, but resistance has rapi dly developed in the past two decades (Dorsey et al. 2001; Wellems and Plowe 2001) . Another group of antimalarials are artemisinin and its derivatives. These drugs al so act against gametocytes, a sexual stage form of the malaria parasite. Additionally, drugs of the arte misinin class supposedly exert activity through interaction w ith heme in the food vac uole (Meshnick and Dobson 2001; Meshnick et al. 1996). It appears that iron in heme, the by-product toxic to malarial parasites from hemoglobin metabolism, can induce degradation of artemisinin (Posner and Oh 1992). Intermediates, resulting from a cascade of rearrangements and fragmentations of artemisini n and its analogs (Gu et al . 1999; Posner et al. 1998, 1996), alkylate some specific parasite proteins, such as the heme-binding malarial translationally controlled tumor protein (TCTP) (Bhisuttibhan et al. 1999), wh ich are critical for parasite survival (Rodriguez et al. 2002; Wang and Wu 2000). Artemisinin and its derivatives have been used increasingly since the 1980s (WHO 1998, 2001; Hayne s 2001; Li and Wu 1998; Meshnick 2001; Price 2000). A third group of antimalarials are the antifolate class. These drugs are inhibitors of one-carbon transf er reactions in nucle otide biosynthesis and amino acid metabolism pathways (Sherman 1998). Presently, the most widely used


10 regimen is the combination of pyrimethami ne and sulphadoxine; however, resistance seems to develop rapidly (Takechi et al . 2001; White 1998). Atovaquone/proguanil fixeddose combination is another treatment fo r malaria. Atovaquone interferes with mitochondrial electron transport. Resist ance has developed against atovaquone (Looateesuwan et al. 1999); hence this combination. In addition, commonly used antibiotics, such as tetracycline, doxycyclin e and clindamycin, inhibit parasite growth and are being used increasingly with othe r antimalarials (WHO 2001). Extensive use of “single-drug” malaria medicines has spurred ra pid development of dr ug-resistant parasite species. So far, artemisinins are the only t ype of traditional drugs that have not been documented for treatment failure. Novel drug targets in the life cycle of the malaria parasite are being identified and potentia l drugs against these targets are under investigation. Despite these efforts, artemi sinin-based combination therapies (ACT) are currently the most effective medici ne available to treat malaria. Development of malaria vaccines is a nother strategy for prophylaxis. Vaccine studies on malaria parasites have been carried out for decades, yet there are still no commercially available ones thus far due to the complexity the parasites present. First, compared with viruses and bacteria, mala ria parasites bear much larger genomes encoding more proteins (G ardner et al. 2002a). Secondl y, malaria parasites have multistage life cycles and different antigens are expressed in varied stages (Bozdech et al. 2003). Thirdly, malaria parasites, P. falciparum in particular, have enormous variability in their proteins, which easily enable them to evade host immune defences (Richie and Saul 2002). On the other hand, evidence shows th at malaria vaccines are feasible. First of all, immunization with irradiated sporozoite s at least partially protects animals and


11 humans from being infected by sporozo ites (Clyde 1990; Collins and Contacos 1972; Egan et al. 1993; Nussenzweig et al. 1967; Rieckmann et al. 1979). Secondly, people repeatedly infected by malaria can develop imm unity to protect against clinical diseases (Baird 1995). Thirdly, vaccines already in ha nd can protect animals and humans against malaria infection (Kester et al. 2001; Stow er et al. 2001). Fourthly, vaccines show success in protecting mosqu itoes from infection by P. falciparum and P. vivax (Carter et al. 2000; Hisaeda et al. 2000). Vaccine studies are still in progress along with hope and challenges. Figure 1-5. Examples of antimalarial dr ugs. Figure is adapted from Ridley 2002.


12 Malaria Parasite Causative Agent and Life Cycle Malaria is caused by the protozoa Plasmodium species. The genus Plasmodium is estimated to include at least 172 spp., of wh ich 89 infect reptiles (Telford 1994), 32 infect birds (van Riper et al. 1994) and 51 infect mammals. Hitherto, four Plasmodium spp., Plasmodium falciparum , Plasmodium vivax , Plasmodium ovalae and Plasmodium malariae , have been identified to inf ect human beings, among which P. falciparum is responsible for over 90% of human deaths from malaria (Mi ller et al. 2002). Plasmodium spp. are insect vector-borne protozoa n parasites. Each parasite needs a female mosquito and a vertebrate host to accomplish its life cycle. The sexual and asexual stages of life cycle are carried out in the mosquito and the vertebrate host, respectively. As an exam ple, the life cycle of P. falciparum is reviewed in brevity (Figure 1-6). The asexual stage of the life cycle in man starts when a female anopheline mosquito takes a blood meal (Despommier et al. 1987). As a result, sporozoites are released into the bl oodstream of the human host and de livered to the liver. There, sporozoites invade parenchymal cells to in itiate the exoerythrocytic cycle. In the hepatocytes, sporozoites undergo asexual reproduction known as exoerythrocytic schizogony, which renders rupture of liver cells to release thousands of merozoites into the blood stream. The erythrocytic cycle is initiated when these released merozoites invade red blood cells. Inside the erythroc ytes, merozoites go through a trophic period when the parasites experience enlargement. Th e early immature trophozo ite is referred to as “ring form” due to its morphology. Trophozoite enlargement is accompanied by active metabolisms such as hemoglobin proteolysis and host cytoplasm ingestion. At the end of


13 the trophic period, multiple rounds of erythroc ytic schizogony results in rupture of erythrocytes and release of da ughter merozoites. It is parasi tes in the erythrocytic stage that cause the symptoms of ma laria. Most of the merozoite s invade fresh red blood cells to reinitiate the cycle. A minor portion, on th e other hand, differentiate into sexual forms known as macrogametocytes and microgametocytes. The sexual stage is carried out inside the transmission vector. As a female mosquito takes a blood meal, gametocytes are i ngested and further differentiated as macrogametocytes and microgametocytes. Fl agellated microgametes, released from microgametocytes, fertilize macrogametes to form zygotes in the midgut of mosquito. Zygotes are further developed as motile ookinet es which penetrate th e gut epithelial cells and become oocysts. As the oocysts under go asexual reproduction, sporozoites are developed and released into the hemocoel of the mosquito host. The sporozoites then migrate to and invade the salivary glands a nd are ready to be deliv ered to human being again. Drug Resistance The emergence and worldwide spread of drug-resistant parasites have presented malaria as one of the priority issues to the international health community. While chloroquine resistance of the most lethal species, P. falciparum , emerged 35 years ago (Campbell et al. 1979), the fi rst chloroquine-resistant P. vivax infection was confirmed in the late 1980s (Rieckmann et al. 1993; Whit by et al. 1989). In the past two and a half decades, drug-resistant Plasmodium has been rapidly developed and globally spread. To date, P. falciparum has been reported to show resi stance to most regimens except artemisinins.


14 Figure 1-6. The life cycle of human malaria parasites. When a blood meal is taken, a malaria-infected female Anopheles mosquito inoculates sporozoites into the human host. Sporozoites infect hepatocy tes and mature into schizonts, which rupture to release merozoites. Following this initial replication in the liver (exoerythrocytic schizogony), the parasites undergo asexual multiple rounds of replication in the eryt hrocytes (erythrocytic schi zogony). Merozoites infect erythrocytes. The ring stage trophozoites ma ture into schizonts, which rupture to release merozoites. Some parasites differentiate into sexual erythrocytic stages (gametocytes). The gametocytes, microgametocytes and macrogametocytes, are ingested by an Anopheles mosquito during a blood meal. The multiplication of parasites in the mosquito is known as the sporogonic cycle. While in the mosquito's stomach, the microgametes penetrate the macrogametes to generate zygotes. The zygotes subsequently become motile and elongated, known as ookinetes, which invade the midgut wall of the mosquito and develop into oocysts. The oocysts grow, rupture, and release sporozoites, which are delivered to the mosquito's salivary glands. Inoculation of the sporozoites into a new human host perpetuates the malaria life cycle. Figure is adapted from Oaks 1991.


15 Chloroquine and related 4-substituted qui noline drugs kill malaria parasites by blocking the detoxification of heme. As previously di scussed, trophozoites digest hemoglobin to provide nutrients and create sp ace for their developmen t, proliferation and osmotic protection (Lew et al. 2003). A t oxic byproduct of hemoglobin degradation is free heme. Malaria parasites polymerize heme to form an inert crystalline pigment known as hemozoin (Francis et al. 1997b; Slater 1993). Chloroquine accumulates in the acidic food vacuole and effectively blocks the seque stration of lethal heme into hemozoin (Sullivan et al. 1996). Chloroquine-resistant P. falciparum reduces the concentration of chloroquine in the food vacuol e potentially due to mutations of two independent genetic sources, pfcrt ( P. falciparum chloroquine-resistance trans porter) (Fidock et al. 2000) and pfmdr 1 ( P. falciparum multidrug-resistance 1) (F oote et al. 1989). The 13-exon pfcrt encodes a 45 kDa transmembrane protein loca lized to the membrane of the food vacuole and has no significant homology to the ABC transporter family. Pfcrt genes from resistant cells contain 5 mu tations, all of which are restricted within or near transmembrane domains. However, the mechanism by which pfcrt affects chloroquine levels is still poorly understood. Pfmdr encodes the P-glycoprotein homologue 1 (Pgh1) protein and is homologous to the ABC transpor ter family. Point mutations responsible for resistance have been identified (Reed et al. 2000). Another major therapeutic regimen for malaria treatment is the synergistic combination of pyrimethamine and sulfadoxi ne, which are representative drugs of antifolate and sulfonamide class, respec tively. Suldadoxine and pyrimethamine are inhibitors of dihydropteroate synthase and dihydrofolate reductase, respectively, both of which are key enzymes in the protein and nucleic acid synthesis pathway of Plasmodium


16 spp. Application of the combined therapy le ads to an orderly accumulation of point mutations in the two enzymes and developmen t of drug resistance (Cortese et al. 2002; Kublin JG et al. 2002; Plowe CV et al. 1997; Roper C et al. 2003). Additionally, drug resistance to atovaquone, a mitochondrial el ectron transport inhibitor, has been confirmed, but fortunately, the clinical us e of atovaquone has been salvaged by the simultaneous application of proguanil, an an tifolate class drug (Canfield et al. 1995; Looareesuwan et al. 1996). A P. falciparum strain resistant to multiple drugs (resistant to chloroquine, sulfadoxine/pyrimethamine and mefloquine, and partially resistant to quinine and quinidine) have been reported in Indonesia (Syafruddin et al. 2003). Experiments indicate that in terms of acquiring drug resistance, the frequency of this strain (W2) was 1000 times higher than that of the strain (D6) fully sensitive to conventional drugs (Rathod et al. 1997). The ability for parasites to rapidly acquire drug resistance fo rces us to find new molecular entities to counter the challenge. A potential group of targets for novel drug design are the plasmepsins. Plasmepsins Genetic Family Plasmepsins belong to Clan AA, Family A1 of the aspartic proteinase superfamily. The name plasmepsin comes from Plasm odium and pepsin , a common aspartic proteinase with similar molecular structur e. At least 10 plasmepsins have been identified from the complete P. falciparum genome (Coombs et al. 2001; Gardner et al. 2002a, b). Plasmepsin 1, 2, 4 and the closely related histo-aspartic prot einase (HAP) have been immunolocalized in a special acidic or ganelle of the malaria parasite, known as the food vacuole during the erythroc ytic stage (Banerjee et al. 2002). The gene loci encoding


17 plasmepsin 4, 1, 2 and HAP lie sequentially in a cluster near the left telomere on the “+”strand of chromosome 14 and span a 20 kb region, with approximate 4 kb noncoding sequence separating each gene (Banerjee et al . 2002; Dame et al. 2003). Each of the P. falciparum food vacuole plasmepsins, encoded as a single polypeptide, is composed of a prosegment of approximately 124 residues followed by the mature enzyme of 326-329 residues in length. In additi on, their mature enzymes share high amino acid sequence identity (59%) to each other (Dame et al. 2003). P. falciparum specific oligonucleotide micr oarray studies on the asexual intraerythrocytic developmental cycle confirm the expression profiles of the food vacuole plasmepsins in the erythrocytic stage (Bozdech et al. 2003). In addition, the expression of plasmepsin 5, 9 and 10 of P. falciparum has been identified over the erythrocytic stage but they do not reside in the food vacuole (B anerjee et al. 2002; Bozdech et al. 2003). However, plasmepsin 6-8 fail to be detected in that particular cycle, indicating their possible roles in the liver or sexual stage. These six non-food vacuole plasmepsins are located in 5 different chromosomes and shar e poor amino acid sequence identity to each other as well as to the food vacuole plasme psins (Coombs et al. 2001; Dame et al. 2003; Gardner et al. 2002a; Hall et al. 2002). Additionally, plasmepsin 5-10 of P. falciparum have rather more unique sequence featur es than their food vacuole plasmepsin homologues (Figure 1-7). First of all, their prosegment is of a different size, from 47 residues ( Pf PM8) up to 259 residues ( Pf PM10); secondly, their mature enzyme segments are of variable leng ths due to the presence of insertions ( Pf PM5, 7 and 9), deletion ( Pf PM8) or C-terminal extensions ( Pf PM5 and 9).


18 Figure 1-7. The amino acid sequence featur es of plasmepsins of the human malaria parasite P. falciparum . The prosegments, mature enzyme segments and deleted segment of mature enzyme are represented by white rectangles, black rectangles and dot rectangl e, respectively. The insert ed loops and C-terminal extension regions are repr esented by circles and st raight lines, respectively. Pf PM1, Pf PM2, Pf HAP and Pf PM4 all share a similar enzyme size (prosegment of 124 amino acid long plus mature enzymes of 326-329 amino acid long). However, the sizes of non-food vacuole plasmepsins are distinctive. The prosegment lengths are: Pf PM5 = 86 amino acids, Pf PM6 = 85 amino acids, Pf PM7 = 76 amino acids, Pf PM8 = 47 amino acids, Pf PM9 = 215 amino acids and Pf PM10 = 259 amino acids. In addition, the mature enzyme segments are of variable lengths. Pf PM5 has three insertions: the first is 14 amino acids, the second is 46 amino acids and the third is 28 amino acids and Pf PM5 also has a C-terminal extension of 71 amino acids. The Pf PM6 mature enzyme segment is similar in length to the food vacuole enzymes represented at the top, while Pf PM7 has a single inser tion of 18 amino acids. The mature enzyme segment of Pf PM8 has an N-terminus that is 13 amino acids shorter than the other plasmepsins. Pf PM9 has an insertion of 61 amino acids and a C-terminal extension of 16 amino acids. The mature enzyme part of Pf PM10 is the same size as the food vacuole enzymes.


19 For the other three nonfalciparum parasites that infect man, only one plasmepsin has been identified from each species that is highly homologous to the food vacuole plasmepsins of P. falciparum . These three plasmepsins are more closely related to each other (756% identity) than to the food vacuole plasmepsins of P. falciparum , and are denoted as Pv PM4, Po PM4 and Pm PM4 as they share the hi ghest amino acid sequence identity (65%) with Pf PM4 (Dame et al. 2003). Besides, the amino acid sequence features of these three plasmepsins are qui te similar to those of their food vacuole plasmepsin orthologs of P. falciparum . In parallel, the presence of orthologs of all six non-food vacuole plasmepsins of P. falciparum in the three nonfalciparum parasites infecting man has been verified by compar ative genomics methods (Dame et al. 2003). The studies on genomes of malaria parasi tes that infect rodents, such as P. berghei and P. chabaudi , are also in progress. As an example, P. berghei , quite like the three nonfalciparum parasites infecting man, c ontains one unique orthlog ( Pb PM4) closely related the food vacuole plasmepsin and six non-food v acuole plasmepsin orthlogs (Carlton et al. 1999; Dame et al. 2003). Physiological Functions The successes of in vitro culture of bloodstage P. falciparum (Trager and Jensen 1976) and isolation of the food vacuole (Goldbe rg et al. 1990) created feasibility to study the physiological functions of plasmepsins of this acidic organelle. The main function of plasmepsins in the food vacuole is degrad ation of hemoglobin, the major cytosolic protein of erythrocytes, to provide nutr ients and make development space for the increasing schizonts. Hemoglobin molecules are engulfed by the inf ected parasites and delivered to the food vacuole (Rudzinska et al. 1965; Goldberg 1993), where they are digested in an ordered and speci fic pathway (Figure 1-8). As for P. falciparum ,


20 plasmepsin 1 and 2, originally known as aspart ic hemoglobinase 1 and 2 (Francis et al. 1994), initiate the hemoglobin dissection by hydrolysis of the 33-34 peptide bonds in the hinge region of hemoglobin (Gluzman et al. 1994). The initial cleavage unravels the tetrameric protein, which is then subjected to proteolysis by other aspartic proteinases such as HAP and plasmepsin 4. The resulting large peptide fragments are further digested by cysteine proteases falcipain2 and -3 (Francis et al. 1996; Salas et al. 1995; Sijwali et al. 2001) to break down to small peptides of 15-20 amino acids long. These peptides are further degraded by a metalloprotease falc ilysin (Eggleson et al. 1999). The resulting oligopeptides may be exported out of the food vacuole for terminal degradation to single residues in the parasite cytoso l (Kolakovich et al. 1997). This digestion process is quite efficient as up to 75% of hemoglobin of th e infected erythrocytes can be broken down within a few hours (Ball et al. 1948; Roth et al. 1986). The physiological roles for food vacuol e plasmepsins other than hemoglobin digestion are not clearly unders tood; however, evidence indicate that plasmepsin 2 is able to digest native spectrin, ac tin and protein 4.1 from erythroc yte ghosts in the neutral pH milieu of the cytosol compartment (Le Bonniec et al. 1999). Later, recombinant plasmepsin 4 was proven to have the ability to digest spectrin in acidic and near neutral pH conditions (Wyatt and Berry 2002). Further evidence suggests that there exist local regions of acidification with in parasitized red blood cells (Hayashi et al. 2000), which may also be consistent with in vivo activities of plasmepsins against cytoskeleton proteins, although the presence of plasmepsin 4 in the cytosolic region, unlike plasmepsin 2, still needs to be verified.


21 Figure 1-8. The hemoglobin degradation pathway in Plasmodium falciparum . Hemoglobin degradation inside the mala ria parasite is performed in an ordered stepwise way. Aspartic, cystei ne and metalloproteases as well as cytosolic exopeptidases are involved in this process. Figure is adapted from Coombs et al. 2001.


22 Little is known about the behaviors of the non-food vacuole plasmepsins of P. falciparum . The naturally-occurring plasmepsin 5 has recently been described as an integral membrane protein that is located in the endoplasmic reticulum (ER). This evidence suggests a potential pr otein processing role played by this enzyme (Klemba and Goldberg 2005). Potential Targets of Novel Antimalarials A large body of evidence demonstrates th at a variety of aspartic proteinase inhibitors can block the development and grow th of malaria parasi tes and consequently kill the parasites. Development of highl y selective plasmepsin inhibitors and investigation of their related antimalarial activities is an essential step for creating the new generation of antimalarial drugs. Pepstatin A (Figure 1-9), the general inhibitor of aspart ic proteinases, kills the cultured P. falciparum before trophozoite development with an IC50 value of 4 M and has a significant effect on sc hizonts maturation. When used in combination with E-64, a cysteine proteinase inhibitor, pepstatin A can inhibit the parasite growth with a strong synergistic effect (Bailly et al. 1992). SC-50083 (Figure 1-9) is a peptidomimetic inhibitor highly specific to plasmepsin 1. When fed to the cultures of chloroquine-sensitive P. falciparum HB-3 strain, SC50083 kills the parasite by blocking the bulk of hemoglobin degrada tion. Like pepstatin A, SC-50083, when used in combination with E-64, shows strong synergistic antiparasitic activity (Gluzman et al. 1994; Francis et al. 1994). The antiparasitic activities of a group of pe ptidomimetic plasmepsin 1 inhibitors (Ro 40-4388, Ro 40-5576, Ro 42-1118 and Ro 17-7109) (Figure 1-9), designed by Hoffmann-La Roche Inc, have been teste d. These inhibitors exhibit comparable IC50


23 values of inhibition of parasite growth between the chloroquin e-sensitive NF54 and chloroquine-resistant K1 strain. Ro 40-4388 and Ro 40-5576 are the two most effective inhibitors with IC50 values in the nanomolar to low mi cromolar range. Each of these two compounds, however, when used in combinati on with chloroquine in the NF54 strain, shows an antagonistic e ffect (Moon et al. 1997). A nonpeptidyl diphenylurea derivative WR268961 (Figure 1-9) from the Walter Reed chemical database shows inhibition of P. falciparum growth with IC50 values in the nanomolar range for both the ch loroquine-sensitive (D6) and chloroquin e-resistant strains (W2) (Jiang et al. 2001). Parasites fed w ith WR268961 present enlarged food vacuoles relative to the control. The high level of degraded hemoglobin and the weak in vitro inhibition of plasmepsins 2 ( Ki = 6.1 M) may indicate that WR268961 does not inhibit hemoglobin processing initiators, plasmepsin 1 and 2, but halts further degradation. Given the prevalence of HIV/AIDS-malaria coinfection in subSahara Africa, the effects of HIV-1 clinical drugs on malari a treatment are intriguing. Recently, several groups have reported their studies on the anti parasitic activities of the clinical HIV-1 protease inhibitors. Skinner-Adams et al. revealed that saquina vir, ritonavir and indinavir directly inhibited the growth of P. falciparum multidrug-resistant strain Dd2 and chloroquine-sensitive strain 3D7 in vitro at clinically relevant concentrations (SkinnerAdams et al. 2004). Parikh et al. examined the antiparasitic effects of all seven protease inhibitors on cultured drug -sensitive and drug-resistant P. falciparum strains. Results indicated all compounds bloc ked parasite development at clinically relevant concentration and lopinavir was the most potent compound (Parikh et al. 2005). Andrews et al. examined morphologies and hemoglobin digestion patterns of protease inhibitor


24 Figure 1-9. Structures of plas mepsin inhibitors. Figure is adapted from Francis et al. 1994; Jiang et al. 2001, and Moon et al. 1997.


25 treated parasites (Andr ews et al. 2006). Although antipar asitic activities were shown for most of the compounds, distin ct parasite morphologies were observed among different inhibitors; besides, unlike ch loroquine, hemoglobin diges tion has been significantly halted in the presence of protease inhibitors, and yet the digestion pr ofiles were distinct from that of pepstatin A. In addition, in vivo antimalarial activities of protease inhibitors were studied on a P. chabaudi infected murine model. The most effective regimens against infection were combin ations of ritonavir-saquina vir and ritonavir-lopinavir. Targeted genetic disruption of the P. falciparum food vacuole plasmepsins by homologous recombination reveal ed that the parasite can su rvive with deletions in each of the individual gene and appear morphologically normal. Pf PM1, Pf PM2, and Pf PM4/ Pf PM1 disruptions have elongated reproduc tion time compared with the parental cell line due to reduced growth rate. The Pf PM2 knockout construct exhibits abnormal mitochondrial morphology, and deletion of Pf PM4 results in accumulation of electrondense vesicles in the food v acuole (Liu et al. 2005; Omar a-Opyene et al. 2004). These findings indicate that there may be fu nctional redundancy between the four P. falciparum food vacuole plasmepsins. For this reason, it is necessary to develop novel antimalarial drugs that can inhibit multiple targets of this proteinase family. Characterization of Naturally-occurring and Recombinant Plasmepsin Remarkable progress has been made on ch aracterization of plasmepsins in the kinetic and structural level in the past d ecade. Naturally-occurring plasmepsins can be isolated directly from para site extracts. The relatively low yields of the naturallyoccurring forms led to expression of recombinant plasmepsins in the heterologous system E. coli .


26 Plasmepsin 1 and 2 of P. falciparum are the first two enzymes identified as hemoglobinases. Naturally-occurring plasmeps in 1 and 2 have been purified from the food vacuole of the parasite (Gluzman et al. 1994; Go ldberg et al. 1991). The semiproenzyme forms of plasmepsin 1 and 2 lacking the potential transmembrane region have been successfully expressed in E. coli . Approximately 20 mg of purified activatable proplasmepsin 2 can be obtained from 1 liter of cell culture. The recombinant plasmepsin 2 can perform auto-activation at the optimal catalysis pH of 4.4 leaving a prosegment of 12 residues attached to the ma ture enzyme (Hill et al. 1994). The naturally-occurring and recombinant forms of plasmepsin 2 share comparable kinetic parameters on varied chromogenic and fluorogenic peptide substr ates despite the extra prosegment portion (Luker et al. 1996; Tyas et al. 1999). However, plasmepsin 1 is not a similar case. First, the wild type recombinant proplasmepsin 1 can not perform au to-activation. Selfprocessing was achieved by mutating K110p into V110p; secondly, only 65 g of fully activatable zymogen were obtained from 1 lite r of cell culture (Moon et al. 1997); thirdly, the recombinant plasmepsin 1 is up to 10-fold less efficient than its naturally-occurring counterpart in cleavage of ch romogenic and fluorogenic pep tide substrates (Luker et al. 1996; Tyas et al. 1999). Later, the recombinant plasmepsin 4 ort hologs of human mala ria parasites were produced in E. coli . Like plasmepsin 2, these proplasmepsin 4 enzymes are easily activated by self-cleavag e, the optimal pH for zymogen ac tivation and catalysis is around 4.5 (Li et al. 2004; Westling et al. 1997; Wyatt and Berry 2002). The kinetic parameters of plasmepsin 2 a nd plasmepsin 4 orthologs hydrolysis of a series of single chromogenic peptide substrat es with single residue substitutions at P3,


27 P2, P2’ and P3’ positions were determined (Li et al. 2004; Westli ng et al. 1999; Westling et al. 1997). From the results, the best subs titute for each position was selected, and the optimal substrate for each enzyme was obtai ned. The peptide bonds between P1 and P1’ were modified to a reduced noncleavable methyleneamino bond to develop the single substrate inhibitors (SSIs). Furthermore, th e subsite preferences of plasmepsin 2 and plasmepsin 4 orthlogs were studied using a co mbinatorial chemistry approach. A series of inhibitors with maximal deviation of bindi ng affinities between plasmepsins and human cathepsin D were developed and named the co mbinatorial chemistry inhibitors (CCIs). The inhibition effects of these SSIs and CCIs on plasmepsins were investigated and some can serve as lead compounds for designing novel antimalarials (Beyer et al. 2005). One of the food vacuole plasmepsins of P. falciparum is the histo-aspartic proteinase (HAP). It has one of the catalytic aspartic acids replaced by a histidine, and the catalytic activity of HAP has been confirme d from its purified na turally-occurring form. In addition, the optimal pH for the natura lly-occurring HAP is around 6.0, around one pH unit higher than the acidic catalytic milieu fo r typical aspartic protei nases (Banerjee et al. 2002). Also, the two active site residues, His34 and Asp214 and a spatially approximated Thr220 form a potential threonine proteinase catalytic triad. The catalysis mechanism for HAP is still elusive. Some evidence shows that the catalytic activity of naturallyoccurring HAP can be totally blocked by the serine proteinase i nhibitor PMSF at 1 mM, but the reaction can also be completely st opped by the aspartic proteinase inhibitor pepstatin A at 1 M (Banerjee et al. 2002). Studies of the mechanism of HAP using a combination of homology modeling, automated docking searches and molecular dynamics simulation indicate this enzyme func tions like an aspartic proteinase (Bjelic


28 and qvist 2004). The recombinant HAP has been overexpressed in E. coli with the yield of 10 mg/L cell culture. Howe ver, no catalytic activity ha s been detected from the refolded material (Berry et al. 1999). In parallel, studies on non-food vacuol e plasmepsin characterization are in progress. As an example, recombinant plasmepsin 5 of P. falciparum has been successfully expressed as inclusion body form in E. coli . Refolding is being attempting to produce active materials (Melissa R. Marzahn, Prof. Ben M. Dunn group, personal communication). Characterization of plasmepsins of the murine malaria parasites, such as plasmepsin 4 enzymes of P. berghei and P. chabaudi , has also been reported (Humphreys et al. 1999; Martins et al. 2003; Martins et al. 2006). X-ray crystallography has become an impor tant tool for understanding active site features and dissecting the in teractions between plasmepsin s and inhibitor molecules. More than 15 plasmepsin crystal structures have been reported to the Protein Data Bank thus far (Table 1-2), most of wh ich characterize the structures of Pf PM2 in complex with varied inhibitory compounds. These findings, together with the continuously augmented structural knowledge, certainly benefit th e development of novel drugs with higher selectivity and more potency. Combinatorial Chemistry In the late 1980s requirement for synt hesis of numerous chemical compounds rapidly and inexpensively spawned a new branch of chemistry known as combinatorial chemistry. Application of comb inatorial chemistry approaches to characterize proteinligand interactions was initiat ed in early 1990s, when the “t ea bag” and “split and mix” technologies were used on library synthe sis (Houghten et al. 1 991; Lam et al. 1991). Originally, proteins/enzymes we re screened by compounds attach ed to resins (Nery et al.


29 Table 1-2. X-ray crystal st ructures of plasmepsins. Enzyme PDB ID Complexed Resolution () Reference 1SME pepstatin A 2.70Silva et al. 1996 1PFZ pro Pf PM2 1.85Bernstein et al. 1999 1LEE rs367 1.90Asojo et al. 2002a 1LF2 rs370 1.80Asojo et al. 2002a 1LF3 EH58 2.70Asojo et al. 2003 1M43 pepstatin A 2.40Asojo et al. 2003 1LF4 mature enzyme 1.90Asojo et al. 2003 1ME6 statine-based inhibi tor 2.70Freire et al. 2004 1XDH pepstatin A 1.70Prade 2005a 1XE5 pepstatin analogue 2.40Prade 2005b 1XE6 pepstatin analogue 2.80Prade 2005c 2BJU achiral inhibitor 1.56Prade et al. 2005 1W6H a novel inhibitor with bulky P1 side chain 2.24Lindberg et al. 2006a Pf PM2 1W6I pepstatin A 2.70Lindberg et al. 2006b 1QS8 pepstatin A 2.50Bernstein et al. 2003 Pv PM4 1MIQ pro Pv PM4 2.50Bernstein et al. 2003 Pf PM4 1LS5 pepstatin A 2.80Asojo et al. 2002b Pm PM4 2ANL KNI-764 3.30Clemente et al. 2006 Structural data are availabl e in Protein Data Bank ( ).


30 1997). The presence of the resins not only affects the accuracy of screening due to the potential alteration of proteinligand binding mode (St Hilarie et al. 1999), but also limits the types of substances can be made usi ng combinatorial chemistry. To overcome the limitations of the “on-bead” bi ological screening analysis, methods have been developed to perform synthesis in solution and even co mbine the advantages of both the “on-bead” and the solution method (Borman 1996; Borman 1998; Borman 1999; Han et al. 1995; Service 1996; Service 1997). Methods have also been devised to allow screening analysis of library compounds in soluti on (Backes et al. 2000; Meldal et al. 1994; Salmon et al. 1993). Using combinatorial chemistry approaches , a variety of low-molecular-weight inhibitors of plasmepsin 2 with nanomolar ra nge of inhibition have been identified. By comparison of the crystal structures of plas mepsin 2 and human cathepsin D in complex with pepstatin A, the high sequence homol ogy of the active site region between the two enzymes was confirmed. Subsequently, Ha que et al. employed a human cathepsin D inhibitor library containing 1039 mechanism-ba sed compounds (Figure 1-10 a) to screen against plasmepsin 2 to identify lead co mpounds. These lead molecules were further optimized by individually examining the thr ee sites (R1, R2 and R3) of variations. The resulting inhibitors showed low nanomolar inhibition with molecular weight around 600 Da and selectivity up to 15-fold over hu man cathepsin D (Haque et al. 1999). Furthermore, Carroll et al. designed a 13020-member combinatorial library based on a statine template containing four functi onal group variations (Figure 1-10 b). Out of 13020 compounds, 64 molecules were identified to be highly selective to plasmepsin 2 over human cathepsin D (Carroll et al. 1998b) . Carroll et al. al so synthesized and


31 Figure 1-10. Combinatorial i nhibitor libraries for screeni ng highly selective plasmepsin 2. (a) 1039-member combinatorial library with hydroxylethylamine as core structure; (b) 13020-member combinator ial library with statine as core structure; (c) 18900-member combinator ial library with statine and three cyclic diamino acids as core structure. Figure is adapted from Batra et al. 2002.


32 evaluated a 18900-member statine-based combin atorial library (Figure 1-10 c). Specific plasmepsin 2 inhibitors with nanomolar range of inhibition were obtained (Carroll et al. 1998a). Beyer et al. designed two sets of positional-scanning chromogenic peptide substrate libraries for screening active site pr eferences of plasmepsins and human aspartic proteinases. A total of 19 amino acid residues were screened at each subsite from S3-S3’. The best substitute for each position was determined, from which the best substrate for each enzyme was obtained. By comparison of the subsite preferences between plasmepsins and human aspartic proteinases, highly specific plasmepsin peptidomimetic inhibitors were designed (Beyer et al. 2005). This thesis is composed of 8 chapters: th is chapter briefly covers the previous and current research on malaria, the causative agent Plasmodium spp. and the potential drug design targets—plasmepsins; chapter 2 desc ribes in detail the materials and methods employed in research experiments; chapte r 3 and 4 focus on production and enzymatic characterization of plasmepsin 1 ( Pf PM1) of the human malaria parasite P. falciparum and plasmepsin 4 ( Pb PM4) of the rodent malaria parasite P. berghei ; chapter 5 depicts the subsite preference screen and hi ghly selective inhibitor design for Pf PM1 and Pb PM4 based on the combinatorial library approaches; chapter 6 and 7 focus on the inhibition analysis of the clinical HIV-1 protease i nhibitors and the prim aquine-statine “double drug” compounds on plasmepsins; chapter 8 is the future directions.


33 CHAPTER 2 MATERIALS AND METHODS Introduction This chapter expatiates upon chemical a nd biological materials, instruments and methods employed throughout this study. The experimental methods are described as follows: 1) Production of r ecombinant plasmepsins, which includes site-directed mutations, overexpression of encoded en zyme constructs, inclusion body extraction, purification and denaturation, in vitro refolding, purification of refolded materials, selfprocessing of zymogens, and purification of mature enzymes; 2) Characterization of purified recombinant plasmepsins, including de termination of optimal pH and incubation time conditions for proenzyme conversion a nd enzymatic catalysis, time-resolved SDSPAGE analyses of self-processing events , N-terminal sequencing analyses and determination and comparison of kinetic parameters ( kcat, Km, kcat/ Km and Ki); 3) Investigation of subsite pr eferences of plasmepsins—De termination of the primary substrate specificities by comparison of initial cleavage velocities, determination of the secondary substrate specificities with LC -MS, synthesis of single substrates and inhibitors, and determination and co mparison of kinetic parameters ( kcat, Km, kcat/ Km and Ki) are involved in this section; 4) Analys es of the inhibition e ffects of synthesized compounds and HIV-1 protease inhibitors (P I) on plasmepsins and their antiparasitic activities. For this se ction, the methods include determin ation of dissociation constants ( Ki), molecular modeling, and in vitro antiparasitic assays.


34 Materials All chemical reagents were purchased from either Fisher Scientific or Sigma unless otherwise noted. The pET-3a expression constructs enc oding the sequences of the wild type pro Pf PM1, the pro Pf PM1 K110pN mutation and the wild type pro Pb PM4 were generous gifts from Professor John B. Dame, Univer sity of Florida, Gainesville, Florida. The primers for PCR and site-directed mutagenesis were designed using Oligo6.0 (Molecular Biology Insight s Inc.) and synthesized by Invitr ogen. Plasmid constructs were extracted using QIAprep Spin Miniprep Kits (Qiagen) . Products of conventional PCR and restriction endonuclease digested DNA fragments were purified using a QIAEX II Gel Extraction Kit (Qiagen). The Taq DNA polymerase was from Qiagen. The T4 DNA ligase, the Vent DNA polymerase, restriction endonuc leases and dNTPs were from New England Biolab. The shrimp alkaline phospha tase was from Promega. Site-directed mutagenesis reactions were set up using QuikChange Site-Directed Mutagenesis Kits (Stratagene). PCRs were performed using an iCycler Thermal Cycler (BioRad). The DNA electrophoresis equipment was purchased from Bio-Rad. The One Shot TOP10 chemically competent E. coli cells and the BL21 Star (DE3) pLysS One Shot chemically competent E. coli cells were from Invitrogen. The Spectra/Por Molecularporous Memb rane Tubings (MWCO = 6-8 kDa, 8.0 mL/cm) were from Fisher Scientific. The Hi Trap Q HP 5 mL columns packed with the Q Sepharose high performance resin were empl oyed for anion exchange chromatography. A HiLoad 16/60 column packed with Superdex 75 prep grade resin and a Superdex HR10/30 column packed with Superdex 75 prep grade resin were used for size exclusion chromatography. These columns were purchas ed from Amersham Biosciences. Anion


35 exchange and size exclusion chromatogra phy were performed on a Pharmacia Biotech FPLC system. The equipments and reagents fo r SDS-PAGE analysis were from Bio-Rad. The PVDF membranes were from Fisher Sc ientific. N-terminal protein sequencing analysis was performed by the Proteomics Core Facility at the University of Florida, Gainesville, Florida. All centrifugation work was performed on a model J2-21 centrifuge (Beckman), a GS-15R benchtop centrifuge (Beckman) and a microcentrifuge 5417 C (Brinkmann Instruments Inc.) Colony-growing plates were incubated in an Isotemp 500 series incubator (Fisher Scientific). Waterbath incubations were perf ormed in an Isotemp waterbath incubator (Fisher Scientific). E. coli cell inoculations and protein expressions were performed in either a Series 25 Incubator Shaker (New Brunswick Scientific) or an Innova 4000 Incubator Shaker (New Brunswick Scientific). Protein concentration work wa s performed using either Vivaspin 15R or Vivaspin 6 concentrators (VIVASCIENCE). All kinetic assays were performed using a Cary 50 Bio UV-Visible spectrophotometer (Varian). Data were processed using SigmaPlot 2000 and EnzFitter1.05 (EGA). LC-MS analysis, peptide synthesis and purification, and amino acid analysis were performed by the Proteomi cs Core Facility at the University of Florida, Gainesville, Florida. The group of synthesized “double-drug” co mpounds was a gift from Professor Enrica Bosisio at University of Milano, M ilano, Italy. The clinical HIV-1 protease inhibitor series were furnished by NIH AIDS Reagent and Reference program.


36 Culture of the Plasmodium falciparum 3D7 strain and the antiparasitic assays were performed by the laboratory of Professor J ohn B. Dame at the University of Florida, Gainesville, Florida. Methods In this “Methods” section, pro Pf PM1 and pro Pb PM4 represent both the wild type and the mutant forms of the recombinant enzymes unless specially noted. Production of Zymogen and th e Mature Form of Recombinant Pf PM1 and Pb PM4 Site-directed mutations of pro Pb PM4 Mutations were designed specifically wher e self-cleavage occurred to block the cleavage sites. Site-directed mutagenesis also allowed us to discover the sequential events of proenzyme conversion and helped us to id entify relatively stable mature enzyme forms for structural studies. The target residues to be mutated we re determined by N-terminal protein sequencing analysis of self-cleavage events. In order to block the processing sites, the site-directed mutations were designed accord ing to the primary subsite preferences of Pb PM4 at the S1 pocket. The frequency of E. coli codon usage and the relative “GC” content of the primers were also considered. Pairs of fully complemented primers (Table 2-1) were designed. The PCR r eactions were driven by the Turbo high-fidelity DNA polymerase (Stratagene). The methylated template was then digested by Dpn I (New England Biolab) at 37 C for 1 h. The proce sses of site-directed mutagenesis are shown in Figure 2-1. The program employed for genera tion of these mutations is shown in Table 2-2. About 1-1.5 L of the resulted materials were transformed into 25 L of the chemically competent BL21 StarTM (DE3) pLysS E. coli cells (Invitrogen).


37 Table 2-1. Primers for the site-directed mutations of pro Pb PM4 Primera Sequenceb pro Pb PM4 L117pE (+) 5’–CATAAAAGAATCATTCAAATTATTAAAATCAGGTga ATTAAAAAAAGAGC’ pro Pb PM4 L117pE (-) 5’–GCTCTTTTTTTAATtcACCTGATTTTAATAATTTGAAT GATTCTTTTATG’ pro Pb PM4 L112pE L117pE (+) 5’–CATAAAAGAATCATTCAAAgaATTAAAATCAGGTga ATTAAAAAAAGAGC’ pro Pb PM4 L112pE L117pE (-) 5’–GCTCTTTTTTTAATtcACCTGATTTTAATtcTTTGAATG ATTCTTTTATG’ a L117pE indicated a single Leu to Glu muta tion at the 117p residue of the prosegment; L112pE L117pE indicated a double Leu to Gl u mutation at the 112p and 117p residue of the prosegment. b Nucleotides in lower case indicated the locations of the mutations. Table 2-2. Cycling parameters for sited-directed mutagenesis of pro Pb PM4 Segment Cycles Temperature (C) Time (min) 1 195 0.5 95 0.5 55 1 2 18 68 12 3 14


38 Figure 2-1. Overview of the QuikChange site-directed mutagenesis method. Figure is adapted from the instruc tion manual, Stratagene.


39 Expression of recombinant pro Pf PM1 and pro Pb PM4 The recombinant wild type and mutated forms of pro Pf PM1 and pro Pb PM4 were overexpressed in E. coli . Each of these recombinant pro Pf PM1 and pro Pb PM4 forms encoded the C-terminal prosegment of 48 ami no acid residues in length plus the mature enzyme. One liter of LB medium containing 50 g/mL ampicillin and 34 g/mL chloramphenicol was inoculated with 20 mL of overnight culture of the BL21 StarTM (DE3) pLysS E. coli cells bearing the constructs afor ementioned. The cells were grown at 37 C with 250 rpm shaking speed until OD600 reached 0.6. A final concentration of 1 mM IPTG was utilized to induce gene expression. After 3 h induction, cell pellets were ha rvested by centrifugation at 13,000 g, 4 C, for 15 min and stored at -20 C until ready for inclusion body purification. Inclusion body preparation The cell pellets were resuspended in i ce cold buffer 1 (0.01 M Tris-HCl, pH 8.0, 0.02 M MgCl2, 0.005 M CaCl2, 4.2 mL/gm of cell mass) and lysed twice via a French Pressure cell at 1000 psi. A final concentra tion of 80 Kunitz units/mL of DNase I was introduced to the lysate and incubated at r oom temperature for 15 min. About 5-10 mL of the lysate was laid upon 10 mL of 27% (w/v) sucrose and centrifuged at 12,000 g, 4 C, for 45 min. The pellets were resuspended in buffer 2 (10 mM Tris-HCl, pH 8.0, 1 mM EDTA, 2 mM -mercaptoethanol, 100 mM NaCl). About 5-10 mL of the resuspension was laid upon 10 mL of 27% (w/v) sucros e and centrifuged at 12,000 g, 4 C, for 45 min. The pellets were then resuspended in buffer 3 (50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 2.5 mM -mercaptoethanol, 0.5% Triton-X100) and centrifuged at 12,000 g, 4


40 C, for 15 min. The IB pellets were finally washed with buffer 4 (50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 2.5 mM -mercaptoethanol) and harves ted by centrifugation at 12,000 g, 4 C, for 15 min. The purified wet in clusion body materials were weighed and resuspended in buffer 5 (10 mM Tris-H Cl, pH 8.0, 1 mM EDTA) to the final concentration of 100 mg/mL and stored at -80 C until ready to use. Pro Pf PM1 inclusion body denaturation and in vitro refolding The inclusion body materials of the wild type pro Pf PM1 and the pro Pf PM1 K110pN mutant were thawed on ice and resu spended in buffer 5 (10 mM Tris-HCl, pH 8.0, 1 mM EDTA) to the final concentration of 40 mg/mL. An equal volume of glacial acetic acid was added to the resuspension to fully dissolve the inclusion body materials by gently mixing and incubating at room temp erature for 10 min. The samples were then centrifuged at 3,000 g, 4 C, for 10 min. The supernatant was added drop by drop to the freshly prepared deionized denaturation bu ffer (6 M urea, 50 mM sodium phosphate, pH 8.5, 500 mM sodium chloride). The pH value of the solution was adjusted immediately back to pH 8.5 by slow addition of 10 M sodi um hydroxide to the protein solutions while stirring. The protein samples we re allowed to denature whil e stirring slowly at room temperature for about 2 h. Any undissolved material was removed by centrifugation at 13,500 g, room temperature for 30 min, and the supernatant was passed through a 0.22 m filter. The resulted supernatant was dialyzed agai nst the dialysis buffer (20 mM Tris-HCl, pH 8.0) at 4 C. The dialysis buffer was changed every 6 h 3 more times. The final dialysate was centrifuged at 13,000 g, 4 C for 30 min and filtered through a 0.22 m filter to remove any precipitates.


41 The soluble dialysate was subjected to catalytic activity test. The pro Pf PM1 solutions were preincubated in 0.1 M sodium acetate, pH 4.5, at 37 C for 20 min and mixed with 40 M of the chromogenic peptide subs trate A: Lys-Pro-Ile-Leu-Phe*NphArg-Leu (Nph = paranitrophenylalanine, and * repr esents the bond where cleavage occurs). The initial cleavage velocities we re measured using a Cary 50 Bio UV-Visible spectrophotometer. The dialys ate was stored at 4 C be fore further purification. Pro Pb PM4 inclusion body denaturation and in vitro refolding The inclusion body materials of th e wild type and mutated pro Pb PM4 were thawed on ice and directly dissolved in the denatura tion buffer (freshly pr epared deionized 6 M urea, 50 mM sodium phosphate, pH 8.5, 500 mM sodium chloride). Denaturation was carried out at room temperat ure for 2 h while stirring slow ly. Any undissolved material was removed by centrifugation at 13,000 g, 4 C for 30 min, and the supernatant was passed through a 0.22 m filter. The supernatant was dialyzed against 20 mM Tris-HCl, pH 8.0 at 4 C. The dialysis buffer was changed every 6 h 3 more times. The resulting dialysate was centrifuged at 13,000 g, 4 C for 30 min and filtered through a 0.22 m membrane to remove any precipitates. The soluble dialysate was subjected to catalytic activity test. The pro Pb PM4 solutions were incubated in 0.1 M sodium c itrate, pH 5.0, at 37 C for 5 min and mixed with 40 M of the chromogenic peptide substrat e A: Lys-Pro-Ile-Le u-Phe*Nph-Arg-Leu (Nph = paranitrophenylalanine, and * represents the bond where cleavage occurs). The initial cleavage velocities were m easured using a Cary 50 Bio UV-Visible spectrophotometer (Varian). The dialysate was st ored at 4 C before further purification.


42 Purification of pro Pf PM1 and pro Pb PM4 The soluble dialysate materials of pro Pf PM1 and pro Pb PM4 were primarily purified using HiTrap Q HP 5 mL anion ex change columns. The columns were first equilibrated alternatively with elution bu ffer A (20 mM Tris-HCl , pH 8.0) and elution buffer B (20 mM Tris-HCl, pH 8.0, 500 mM NaCl ). The refolded materials were loaded onto the columns, washed with elution buffer A for 5 min, and the protein was subsequently eluted with a gradie nt of 0-0.5 M sodium chloride. The OD280 and catalytic activity of each fracti on were tested. The catalytic activity assays were carried out si milarly as those of testing the dialysates. The OD280 values were read from a Cary 50 Bio UV-Vi sible spectrophotometer. Pro Pf PM1 was eluted at a NaCl concentration of 0.23 M, while the elution peak of pro Pb PM4 corresponded to 0.3 M NaCl. The peak fractions were pooled and c oncentrated using the Vivaspin 15R concentrator (MWCO = 5 kDa, VIVASCIENCE) until OD280 reached 2.5 for pro Pf PM1 and 1.5 for pro Pb PM4. The concentrated samples were centrifuged at 24,000 g, 4 C for 10 min to remove a ny precipitates. About 250 L of concentrated pro Pf PM1 sample (~ 3 mg/mL) was further purified by size ex clusion chromatography using a Superdex 75 HR10/30 column. About 3 mL of concentrated pro Pb PM4 sample (~ 3 mg/mL) was subjected to gel filtration chromat ography using a HiLoad 16/60 column. The OD280 and catalytic activity of each fract ion were tested. Fractions bearing catalytic activities were pooled and stored at 4 C. Mature Pf PM1 and Pb PM4 preparation and purification The mature Pf PM1 could be isolated from the pro Pf PM1 K110pN mutant. The zymogen of the mutated form was incubated in 0.1 M sodium formate, pH 4.0, at 37 C


43 for 1 h. The maturation process was quenched by diluting the resulted material into 20 mM Tris-HCl, pH 8.0, 8-fold in excess of vol ume while stirring slow ly, thus raising pH near 8.0. The protein solution was loaded ont o a HiTrap Q HP 5 mL anion exchange column and eluted with a gradient of 0-0.5 M NaCl. The mature Pf PM1 was eluted at 0.28 M NaCl. The OD280 and catalytic activity of each fraction were tested. Fractions bearing catalytic activity were pooled and stored at 4 C. The mature Pb PM4 could be isolated from the pro Pb PM4 L117pE mutant. The pro Pb PM4 L117pE was incubated in 0.1 M sodium citrate, pH 5.0, at 37 C for 10 min. The proenzyme conversion was stopped by addi ng the protein sample into 20 mM TrisHCl, pH 8.0, 8-fold in excess of volume while stirring slowly to bring pH close to 8.0. The diluted material was loaded onto a HiTr ap Q HP 5 mL anion exchange column. The mature Pb PM4 was eluted at 0.35 M NaCl in a 00.5 M linear gradient. Each fraction was tested and fractions bearing catalytic ac tivity were treated as discussed above. Characterization of Zymogen and the Mature Form of Recombinant Pf PM1 and Pb PM4 Self-processing and catalysis optimization of pro Pf PM1 and pro Pb PM4 Most of the pepsin-like aspa rtic proteinases, including plasmepsins, acquire their catalytic activities via a self-processing event under ac idic conditions (Dunn 2002). In proper acidic pH milieus, the prosegme nt of zymogen undergoes conformational alteration (Khan et al. 1999). As a result, a majority of the prosegment is cleaved off by the mature enzyme allowing the latter to gain its function. The optimal conditions for self-cleavage were studied as follows: As for pro Pf PM1, the purified zymogen was incubated at 37 C with one fifth volume of 0.5 M acidic buffer series: 0.5 M sodium formate, pH 3.5; 0.5 M sodium


44 formate, pH 4.0; 0.5 M sodium acetate, pH 4.5; 0.5 M sodium citrate, pH 5.0; 0.5 M sodium citrate, pH 5.5 and 0.5 M sodium phosphate, pH 6.0. An equal amount of sample was withdrawn after 0, 5, 10, 15, 20, 25, 30, 60, and 120 min incubation. The reactions were stopped by addition of the 5 Laemmli sample buffer (LSB) and boiling for 10 min. The conversion from pro Pf PM1 to mature enzyme was monitored by SDS-PAGE. In addition, the optimal conditi ons for the catalysis of pro Pf PM1 were determined. As zymogen required auto-activation to show catalytic activity, the best conditions for the combined events of self-processing and ma ture enzyme catalysis were screened. The purified pro Pf PM1 was treated in similar c onditions aforementioned. At each preincubation time point, samples we re withdrawn and mixed with 100 M of peptide substrate A. The initial cleav age velocities were measured on a Cary 50 Bio UV-Visible spectrophotometer. The resulti ng initial rates were normaliz ed with overall the highest velocity set to 100 percent. For each specifi c condition, three individual assays were performed, from which the average normalized velocities and standard errors were determined. The condition allowing pro Pf PM1 to show the highest catalytic activity was defined as its optimal catalysis condition. Using the same approaches, the self-processing and optimal catalysis of pro Pb PM4 were studied, except that th e incubation time before samples withdrawn were 0, 5, 10, 30, 60, 120, 240, 480 min and overnight. Catalysis optimization of mature Pf PM1 and Pb PM4 The purified mature Pf PM1 and Pb PM4 were incubated at 37 C with one fifth volume of acidic buffer series of differen t pH: 0.5 M sodium formate, pH 3.5; 0.5 M sodium formate, pH 4.0; 0.5 M sodium acetat e, pH 4.5; 0.5 M sodium citrate, pH 5.0; 0.5 M sodium citrate, pH 5.5 and 0.5 M sodium phosphate, pH 6.0. Without considering the


45 conversion issue, a 3 min preincubation tim e was given for temperature equilibration. The resulting materials were mixed with 100 M of the chromogenic peptide substrate A. The initial cleavage velocity at each speci fic condition was measured and normalized with the highest initial rate set to 100 pe rcent. The average normalized velocities and standard error were determined from th ree individual assays. The optimal pH for enzymatic catalysis was determined by comparison of the relative hydrolyzing rates. N-terminal sequencing analysis The purified samples required to be analyzed were loaded onto a 10% Tris-Tricine denatured polyacrylamide gel. Protein samp les ran through the stacking gel at 50 V for 30 min and through the separating gel at 120 V for 2 h. At the end of SDS-PAGE, the gel was immediately soaked into the transf erring buffer (10 mM MES, pH 6.0, 20% methanol) while slowly shaking for about 20 min. Meanwhile, a PVDF membrane was soaked in 100% methanol for about 20 min. The protein was transferred onto the PVDF membrane at room temperature for 2 h with the PowerPac 200 (Bio-Rad) set to 90 V. The PVDF membrane was immediately soaked in distilled deionized water while slowly shaking. Water was changed 5 times to wash any salt off. The membrane was stained in 0.02% (w/v) Coomassie Brilliant Blue R-250, 40% methanol, 5% glacial acetic acid, 55% distilled deionized water for 30 sec and dest ained in 40% methanol, 5% glacial acetic acid, 55% distilled deionized water for up to 1 min. The membrane was washed thoroughly with distilled deioni zed water and dried at room temperature between two 3 mm Whatman filter paper (Bio-Rad), wrapped with aluminum foil and stored at 4 C. The N-terminal amino acid sequences were determined by the Edman degradation method (Edman 1949) on an Applied Bios ystems 470A protein sequencer at the Proteomics Core Facility, University of Florida, Gainesville, Florida.


46 Kinetic parameter determination Substrate hydrolysis and Km. Chromogenic peptide substrates were dissolved in 20% DMSO, 10% formic acid, 70% distilled deionized water unless specially noted. Substrate stock solutions were quantified by amino acid analysis (Atherton 1989; Moore et al. 1958; Moore and Stein 1948). Substrate hydrolysis as says were set up in the optimal catalysis conditions of the enzymes. The substrate hydrolysis was defined on the spectrophotometer as the decrease of the av erage absorbance from 284-324 nm (Dunn et al. 1994; Scarborough et al. 1993). The initial clea vage velocities (AU/sec) for at least six substrate concentrations ( M) were measured on a Cary 50 Bio UV-Visible spectrophotometer. The observed rates in AU/se c were converted to M/sec by dividing by the total absorbance change for complete digest ion of each individual substrate of definite concentration. From the convert ed initial rates an d related substrate concentrations, the Vmax and Km values were determined by the equation v = Vmax [ S ] / ( Km + [ S ]) and Marquardt analysis (Marquardt 1963) using th e single substrate program of the enzyme kinetic module 1.0 of SigmaPlot 2000 (Version 6.10) (Spec Science). Active site titration and kcat. Inhibitor stock solutions were prepared in 100% DMSO unless specially noted and quantified by amino acid analysis. The tested enzyme was inhibited by different concentrations of the highly specific aspartic proteinase competitive inhibitor pepstatin A. The ini tial hydrolyzing rates (AU/sec) of the bound enzyme on a substrate of known concentration ( M) were measured via a Cary 50 Bio UV-Visible spectrophotometer. The total c oncentration of the active enzyme was determined from the initial cleavage veloci ties and related pepstatin A concentrations (nm) using the Henderson equation (Henders on 1972; Leatherbarrow et al. 1985) and the tight-binding program in the enzyme kinetic module 1.0 of SigmaPlot 2000 (Version


47 6.10) (Spec Science). The kcat value was calculated from the equation kcat = Vmax (AU/sec)/ ( A /[ S ]*[ E ]), where A is the total absorbance decrease under a specific substrate concentration, and [ E ] is the total active enzyme concentration. Dissociation constant ( Ki) measurement. For a tight binding ( Ki = 50 pM-10 nM) competitive inhibitor, the initial rates (AU/s ec) of enzymatic cleavage of a chromogenic peptide substrate of known concentration ( M) in the presence of di fferent concentrations (nM) of inhibitors were measured as described above. The Ki value was determined by fitting the initial cleavage velocities and re lated inhibitor concentrations into the competitive tight binding inhibitor equation (Morrison 1969): v = {(0.5 Vmax /[ E ])/( Km/[ S ] + 1)}*{([ E ]-[ I ]Ki ap) + sqrt(([ E ]-[ I ]Ki ap)2 + (4[ E ] Ki ap))}, where Ki ap = ( Ki*([ S ]/ Km + 1)), of the Enzfitter1.05 program (EGA). For a non-tight binding ( Ki = 50 nM-10 M) competitive inhibitor, the initial rates (AU/sec) of enzymatic cleavage of at least si x different concentrati ons of a chromogenic peptide substrate in the presence of at leas t two different concentrations of inhibitors were measured and the Ki value was determined by fitti ng the initial hydr olyzing rates and related substrate and inhibitor concentrations into the equation: v = ([ S ]* Vmax)/([ S ] + ( Km * (1+ [ I ]/ Ki)) of the single substrate-single inhi bitor (competitive) program in the Enzyme Kinetic Module 1.0 of SigmaPlo t 2000 (Version 6.10) (Spec Science). Combinatorial Chemistry Based Subsite Preference Studies on Pf PM1 and Pb PM4 Combinatorial libraries The P1 and P1’ combinatorial libraries, which have previously been employed for addressing substrate specificitie s of human and malarial aspart ic proteinases (Beyer et al. 2005), were used here for studyi ng the subsite preferences of Pf PM1 and Pb PM4. The synthesis and purificatio n of these two sets of libraries have been discussed (Beyer 2003).


48 Primary subsite preferences—Spectroscopic assays The primary substrate specificities of Pf PM1 and Pb PM4 at the P1 and P1’ position were determined from the initial hydr olyzing rate of each peptide pool. For the enzyme preparation, mature Pf PM1 was purified from the K110pN mutant. About 800 nM of the mature Pf PM1 was preincubated in 0.1 M sodium citrate, pH 5.5, at 37 C for 3 min. For Pb PM4, about 1 M of the wild type pro Pb PM4 was preincubated in 0.1 M sodium citrate, pH 5.0, at 37 C fo r 5 min. For the peptid e library preparation, each lyophilized substrate pool was dissolved in filtered distilled deionized water making the stock concentration approximately 1.25 mM . The solutions were filtered through a 0.45 m Costar cellulose acetate tube filter by centrifugation at 20,000 g for 5 min to remove any undissolved material. The initial rates of cleavage on 100 M of peptide pools were measured at 37 C using a Cary 50 Bio UV-Visible spectrophotometer . Due to the different locations of the chromophore (Nph), a decrease or an increase of the average absorbance from 284-324 nm was observed during enzymatic digestion of the P1 or P1’ library pools, respectively (Dunn et al. 1994). These initia l cleavage velocities were then normalized with the maximal rate set to 100 percent. Secondary subsite preferences—LC-MS The best three peptide pools of each lib rary set were subjected to secondary substrate specificity studies on th e P3, P2, P2’ and P3’ position. The enzyme and peptide pool preparati on was described above. The complete process of enzyme-catalyzed hydrolysis wa s monitored on the Cary 50 Bio UV-Visible spectrophotometer. Regardless of the enzyme species, the complete digestion of P1 library pools required approximately 2 h; while hydrolysis of the P1’ library pools was


49 complete in about 1 h. The total alterations of the average absorbance from 284-324 nm were calculated from these observations. The enzymatic digestion time allowing only a 510% of substrate hydrolysis, i.e. the linear pha se of the kinetic reaction, was determined. Such time periods were allowed for enzymatic digestions on 100 M of peptide pools. The reactions were stopped by addition of 1% (v/v) of 14 M ammonium hydroxide to raise the pH above 8.0. The resulting materials were immediately frozen in -80 C until LC-MS analyses were performed. The cleaved peptide products were subject ed to separation via capillary reverse phase high performance liquid chromatography (rpHPLC). Peptide poo ls were thawed on ice and diluted 20-fold with solvent A (5% acetonitrile, 94.9 % distilled deionized water and 0.1% glacial acetic acid), and 10 L of the diluted sample was subsequently loaded onto a C18 capillary trap (300 m i.d. 5 mm—packed with C18 PepMap 100, 5 m, 100 , LC Packings, San Francisco, CA). Th e sample was desalted with the loading buffer (3% acetonitrile, 96.9% distilled deinoi zed water, 0.1% glaci al acetic acid and 0.01% TFA) for 5 min and backflushed onto th e analytical column. Capillary rpHPLC separation of cleavage products was pe rformed on a self-packed 20 cm 75 m i.d. Alltima C18 reverse phase column (particle size: 5 m) (Alltech Associates, Deerfield, IL) in combination with an Ultimate Ca pillary HPLC system (LC Packings, San Francisco, CA) operated at a flow rate of 200 nL/min. The program for HPLC isolation is described in Figure 2-2. Inline mass spectrometry analyses of the column eluate were accomplished using a Thermo-Finnigan LCQ Deca quadrupole ion trap mass spectrometer (Thermo Electron Corp, San Jose, CA). The analyses were impl emented in the electrospray ionization mode


50 Figure 2-2. The HPLC gradient elution program for separa tion of cleavage peptides. Solvent A: 5% acetonitrile, 94.9% dis tilled deionized water and 0.1% glacial acetic acid; Solvent B: 3% acetonitril e, 96.9% distilled deionized water and 0.1% glacial acetic acid.


51 (ESI) with the following tec hnical parameters: sheath gas (N2) = 0, aux gas (N2) = 0, spray voltage = 2 kV, capillary temperature = 175 C, capillary voltage = 33 V and tube lens offset = 20 V. Data processing The elution orders of the pentapeptide and tripeptide product series have been determined from the LC-MS analyses (Beyer 2003) (Figure 2-3). For those isomeric peptides with identical molecular masses (X aa = Ile, Leu, or norLe u) or with similar molecular mass (Xaa = Lys or Gln), single pentapeptides and tripeptides have been synthesized to determined their elution orders. Peptide quantity was determined by integrating the area under the curve (AUC) for [M+H]+ and [M+2H]2+ ions of the pentapeptides and [M+H]+ ions of the tripeptides via the Qual Browser program of the X-Calibur 1.3 software package (Applied Biosystems, Foster City, CA). An example of gross mass ch romatogram of isolated cleavage peptides was shown in Figure 2-4, where the chromatograms of single peptides could be segregated based on their unique mass and rela tive retention time. The AUCs of peaks representing each identified product were then calculated. An example of the primarily quantitated results was al so listed in Figure 2-4. During the synthesis of peptide library pools, the abilities of va ried amino acids to be incorporated in the studied P3, P2, P2’ a nd P3’ position were diverse. As a result, the relative quantities of peptide pools were not quite co mparable. For this reason, three groups of combinatorial octapeptides with mi xtures at one position, Lys-Pro-Ile-Glu-PheNph-Xaa-Leu, Lys-Pro-Ile-Xa a-Nph-Phe-Gln-Leu and LysPro-Ile-Glu-Nph-Phe-GlnXaa, have been synthesized and subjected to LC-MS analyses. A semiquantitated profile


52 Figure 2-3. The elution order of enzymatic digestion produ cts during the separation by reverse phase HPLC. The retention time of peptide was extended in the direction of the arrows pointed. Within each group, the elution order might be slightly varied for different peptide pools.


53 for the relative abundances of these octape ptides has been obtained, from which the normalization factors for the pentapeptide and tripeptide products have been determined (Beyer 2003) (Table 2-3). The original relative abunda nces of peptide products were subsequently calibrated by the normalization factors and further normaliz ed with the greatest quantities set to 100 percent. For each pool, the LC-MS analysis was repeated 3-4 times, from which the average relative abundances a nd standard errors were calculated. The optimal residues preferred by each subsite were determined. Inhibition analyses The inhibition effects of plasmepsin singl e substrate inhibito r (SSI) series and combinatorial chemistry inhibitor ( CCI) series (Beyer et al. 2005) on Pf PM1 and Pb PM4 were studied. The enzymes were prepared as follows: the purified mature Pf PM1 was incubated in 0.1 M sodium citrate, pH 5.5, at 37 C for 3 min and the purified pro Pb PM4 was incubated in 0.1 M sodium citrate, pH 5.0, at 37 C for 5 min. These incubations were a prerequisite for inhibition study. All pe ptidomimetic inhibitors were dissolved in 100% DMSO and filtered through a 0.45 m Costar cellulose acetate tube filter by centrifugation at 20,000 g for 5 min to remove any particulate. The concentrations of resulted solutions were quantified by am ino acid analysis. The approaches for determination of dissociation constants ( Ki) were discussed in the “Dissociation constant ( Ki) measurement” part of the section “Cha racterization of Zymogen and the Mature Form of Recombinant Pf PM1 and Pb PM4”.




55 Figure 2-4. LC-MS analyses of Pf PM1 digested P1F library. The mass chromatogram of the cleaved products was dissected into chromatograms of pentaand tripeptides with their retention times and AUCs hi ghlighted. The retention time, AUC and calculated relative abundance for each peptide product are listed.


56 Table 2-3. Normalization fact ors to calibrate the quantitie s of digestion products of octapeptide library pools Xaa Normalization Factora Xaa Normalization Factorb Xaa Normalization Factorc K 1.99847661 K 2.115677599 K 12.108807890 R 2.22301096 R 2.200301571 R 9.518931657 H 2.73283484 H 2.527029589 H 15.534099470 N 2.54025050 N 2.332184449 N 14.168701140 S 4.41105215 S 3.623665245 S 9.596173945 Q 2.76947174 Q 2.430593722 Q 12.310344600 A 4.38288726 A 1.791020944 A 3.995346816 G 6.49171419 G 2.188329152 G 5.029890057 T 2.44177973 T 2.298321438 T 12.080757200 E 3.85956928 E 5.012095013 E 8.576156976 D 3.65042599 D 1.985050990 D 12.988210390 P 2.52292095 P 2.004296055 P 10.684259900 V 1.46398107 V 1.541689753 V 2.838692789 Y 3.24389833 Y 1.715635729 Y 6.618056019 I 1.00000000 I 1.000000000 I 1.000000000 L 2.62200999 L 1.355005387 L 2.181609183 nL 2.62200999 nL 1.532933968nL 1.899688097 F 4.61529792 F 1.880778303 F 2.656791987 W 3.80859080 W 2.266828896 W 5.502620617 Data are adapted from Beyer 2003. a Determined from LC-MS analysis of octape ptide library Lys-Pro-Ile-Glu-Phe-Nph-XaaLeu and used for calibration of pentapeptide and tripeptide products of P1 library pools. b Determined from LC-MS analysis of octape ptide library Lys-Pro-Ile-Xaa-Nph-Phe-GlnLeu and used for calibration of pentap eptide products of P1’ library pools. c Determined from LC-MS analysis of octape ptide library Lys-Pro-Ile-Glu-Nph-Phe-GlnXaa and used for calibration of trip eptide products of P1’ library pools.


57 Inhibition Studies of Potential Antimalarials Enzyme and inhibi tor preparation The expression and purification of Pf PM1 and Pb PM4 were discussed in the section “Production of Zymogen and the Mature Form of Recombinant Pf PM1 and Pb PM4”. The preparation of these two enzymes for kinetic studies was discussed in the “Inhibition analyses” part of the section “Combinatorial Chemistry Based Subsite Preference Studies on Pf PM1 and Pb PM4”. All the other plasmepsin enzymes were expressed in E. coli and purified in inclusion body form (Hill et al. 1994). The incl usion bodies were purified, denatured, and refolded in vitro , and the folded materials were pu rified (Westling et al. 1997). The zymogen forms of these plasmepsins were incubated in 0.1 M sodium acetate, pH 4.5, at 37 C for 5 min prior to inhibition studies. All inhibitors were prepared as described in the “Active site titration and kcat” part of the section “Characteri zation of Zymogen and the Ma ture Form of Recombinant Pf PM1 and Pb PM4”. The concentrations of nonpeptidomimetic inhibitors were determined by titration of a tightly bound enzyme with known concentration. Dissociation constant ( Ki) measurement The approaches for determining inhib ition constants were discussed in the corresponding part of the sect ion “Characterization of Zymo gen and the Mature Form of Recombinant Pf PM1 and Pb PM4”. Molecular modeling Pf PM4–PQ-statine “double-drug” compounds modeling. The model of compound 2 (see Table 7-1) was built under the command “BUILD/EDIT>GET


58 FRAGMENT” of Sybyl7.1 (Tripos Inc.). The energy of the resulting molecule was minimized to 0.05 kcal/(mol 2) using the Powell met hod in the “COMPUTE” menu. Pepstatin A was employed as the templa te ligand for modeling. Pepstatin A was extracted from the crystal structure of Pf PM4 in complex with pepstatin A (1LS5, 2.80 ) (Asojo et al. 2003). The backbone and transition state mimic hydroxyl group of the statine portion of compound 2 was manually ov erlapped with those of pepstatin A, and the resulting conformation was docke d into the active site cleft of Pf PM4. Connolly surfaces of the enzyme active clefts we re generated using the “MOLCAD Surface” command in the “View” menu. The conforma tions of the bridge group and the PQpeptide portion were subsequently rea ssigned following stereochemistry rules, considering potential hydrogen-bonding and hydrophobic interactions, and avoiding steric hindrance against subs ite comprising residues of Pf PM4. Such a binding conformation of compound 2 was further en ergy-minimized while keeping the enzyme molecule rigid. Energy minimization was performed using the Powell method for 1000 cycles under the command “COMPUTER>M INIMIZE SUBSET” of Sybyl7.1 (Tripos Inc.). Pf PM4–HIV-1 protease inhibitors modeling. Superimposing. The crystal structure of Pf PM4 in complex with pepstatin A (1LS5, 2.80 ) (Asojo et al. 2002b) and HIV-1 protease in complex with ritonavir (1RL8, 2.00 ) (Rezacova et al. 2005) or HIV-1 proteas e in complex with amprenavir (1T7J, 2.20 ) (King et al. 2005; Surleraux et al. 2005) were superimposed using the Sybyl7.1 molecular graphics program (Tripos Inc.). The catalytic motifs (~Asp32-Thr33-Gly34~ and ~Asp215-Ser/Thr216-Gly217~, pepsin numbering) were overlapped with the


59 plasmepsin structure as the reference and a llowing the HIV-1 protease molecules to fit. This was performed under the command “Biopolymer>Compare Structures>Fit Monomers”. Docking and energy minimization. Using the Sybyl7.1 program, the ligands, pepstatin A and amprenavir or ritonavir, were extracted from their respective structures. The transition state relevant hydroxyl group and the backbone of HIV-1 protease inhibitors were overlapped with those of pepstatin A. The coordinates of protease inhibitors were further modified by fitting their side chains into the active site pockets of Pf PM4 following stereochemistry rules. The re sulting molecules were docked into the active site cleft of Pf PM4. The free energies of the enzyme-inhibitor complexes were subsequently minimized using the program CNS (Brnger et al. 1998) . The whole enzyme molecule was fixed, and the inhibitors were allowed rotation and translation in the enzyme active site. The cycle of minimization was set to 200. The conformation of the Pf PM4–HIV-1 protease inhibitor complex with the lowest free en ergy was adopted as the model structure. The hydrogen bonding interactions between the enzyme and inhibitors and the accommodation of inhibitors into the active site cleft of Pf PM4 were illustrated using the program PyMOL (DeLano Scientific LLC). In vitro antiparasitic activity assays Antimalarial activities of the synthetic compounds were measured in vitro using the 3H-hypoxanthine uptake assays (Desjardins et al. 1979; Fidock et al. 1998). The tested compounds were dissolved in 100% DMSO making the final concentration 50 mM. A series of 2-fold dilutions we re performed with the concentr ations of the working solution 0.1-50 M. The dilution series were added to P. falciparum 3D7 cultures (0.5-1.0%


60 parasitemia, 1.5-2.0% hematocrit) in low hypoxanthine containing medium in 96-well plates at a final volume of 200 L per we ll. The antiparasitic assays at each drug concentration were performed in triplicate. The parasites were cultured at 37C for 48 h in gas (95% N2, 4% CO2, 1% O2). After 48 h, 100 L of the medium was removed and replaced by 100 L of the lowhypoxanthine medium containing 3H-hypoxanthine at a concentration of 5Ci/mL. After an additional 24 h, cells were harvested onto glass fiber filters and washed thoroughly with distilled wa ter. Dried filters were placed in sample bags and immersed in scintillation fluid (Per kinElmer), and radioactive emissions were counted in a 1450 MicroBeta reader (PerkinElmer). Percentage reduction in hypoxanthine uptake (a marker of growth inhibition) was calculated as follows: reduction = 100 [(geometric mean cpm of no-drug samples)-(mean cpm of test samples)]/(geometric mean cpm of no-drug samples). The percentage reductions were plotted as a function of drug c oncentrations. Fifty percent in hibitory concentrations (IC50) were determined using the variable-slope sigmoidal dose-response nonlinear regression equation (Systat Software Inc.). All the para site cultures and antiparasitic activity test assays were performed by Jorge A. Bon illa Ph.D. of Prof. John B. Dame group, University of Florida, College of Vete rinary Medicine, Gainesville, Florida.


61 CHAPTER 3 PRODUCTION AND ENZYMATIC CHARACTERIZATION OF RECOMBINANT PLASMEPSIN 1 ( Pf PM1) FROM Plasmodium falciparum Introduction In the food vacuole of th e human malaria parasite Plasmodium falciparum , four plasmepsins have been identified, known as plasmepsin 1, 2, 4 and HAP (histo-aspartic proteinase) (Coombs et al. 2001). The functi ons of these food vacuole plasmepsins are mainly involved in hemoglobin degradation, which was originally detected from the initiation of hydrolysis on hemoglobin by a purified naturally-occurring aspartic proteinase (Goldberg et al. 1991). This enzyme is currently recognized as plasmepsin 1, previously designated as aspartic hemoglobina se I (Francis et al . 1994) or PFAPG (Dame et al. 1994). Plasmepsin 1 shares a high amino acid se quence identity and a similar expression pattern with its food vacuole plasmepsin pa ralogs (Banerjee et al. 2002; Dame et al. 2003). The plasmepsin 1 gene encodes a polypeptide composed of a 123 amino acid long prosegment plus a mature portion of 329 am ino acids with a potential transmembrane motif residing in the N-terminus of the pros egment (Figure 3-1). The mature portion of the plasmepsin 1 polypeptide is supposedly folded to a pepsin-like bilobal tertiary structure (Figure 3-2) similar to those of Pf PM2 and Pf PM4 (Asojo et al. 2002a,b; Asojo et al. 2003) based on their highly similar primary structures. Immunoelectron microscopy of intraery throcytic trophozoites reveals that plasmepsin 1 is expressed as a transmembr ane protein and is transported from the


62 Figure 3-1. Nucleotide and protein sequence of the full length pro Pf PM1 (GenBank Accession No. NC_004317). The potential transmembrane motif is underlined; the truncated prosegment (hi ghlighted in blue) plus the mature enzyme sequence (highlighted in red) form the construct for expression. The pair of active site motifs is highlighted in green.


63 Figure 3-2. Molecular mode ling structure of mature Pf PM1. The modeled ribbon structure was built by Amrita Madabushi Ph.D. of Prof. Robert McKenna group based upon the X-ray crys tallographic structure of Pf PM2 in complex with rs370 (1LF2, 1.80 ) (Asojo et al. 2002a) using the program SWISSMODEL (Peitsch 1996; Schwede T et al . 2003). The two cat alytic residues, Asp 34 and Asp 214, were rendered in space-filled presentation.


64 peripheral membranes of the parasite to th e food vacuole (Francis et al. 1994). The conversion of zymogen to mature enzyme ha s been started during the transportation. SC50083, a strong inhibitor of plasmepsin 1 (IC50 = 0.5-0.6 M), kills cultured P. falciparum parasites by presumably blocking hem oglobin degradation (Francis et al. 1994). Other inhibitors of plasmepsin 1, such as Ro40-4388 and Ro40-5576, bear a similar antiparasitic activity (Moon et al. 1997). These pieces of evidence indicate that plasmepsin 1 may serve as one of the ke y targets for novel antimalarial drug design. This chapter describes expression, in vitro refolding, purification and enzymatic characterization of a recombinant propla smepsin 1 mutant, K110pN. The successful production and characterization of the active enzyme provi des knowledge on the design of selective inhibitors of plasmepsin 1. Results Generation of the Recombinant Pro Pf PM1 K110pN Mutant A semipro Pf PM1 fragment lacking the first 75 re sidues of the prosegment due to the potential toxic effect of the transm embrane motif was amplified from the intraerythrocytic stag e cDNA library of the P. falciparum 3D7 strain. Th e PCR fragment was subcloned into the Bam H I site of the pET-3a expression vector. Site-directed mutagenesis at 110p was performed to replace Lys with Asn. The resu lting construct was confirmed by DNA sequencing analysis and transformed into the BL21 DE3 pLysS E. coli cell line. All the work was performed by Ch arles A. Yowell of Prof. John B. Dame’s laboratory, University of Florida, College of Veterinary Medicine, Ga inesville, Florida.


65 Expression, in vitro Refolding and Purification of the Recombinant Pro Pf PM1 K110pN Previous studies indicated that the wild type recombinant pro Pf PM1 could not perform auto-activation (Moon et al. 1997). The amino acid sequences of the prosegments of Pf PM1 and its paralog Pf PM2, whose auto-activation can be easily conducted, were compared. A residue deviation (Lys for Pf PM1 and Asn for Pf PM2) at 110p, the neighbor of a self-cleavage “hot -spot” site Phe111p-Phe/Leu112p was identified (Figure 3-3). If self-cleavage occurs between 111p and 112p, the residue at 110p will potentially occupy the S2 pocket of th e active site. The properties of residues at this position may have a critical role in determining the efficiency of self-cleavage. Therefore, the Lys110p was mutated to Asn to potentially confer the auto-activation ability. A recombinant semi-pro Pf PM1 K110pN form with the last 48 residues of the prosegment plus the mature enzyme was expressed at 37 C for 3 h by 1mM IPTG induction. A total of 350 mg of purified incl usion body material was obtained from 1 liter of cell culture, accounting for approximately 14% of the total cell mass (Table 3-1). Subsequently, the inclusion body material was dissolved in glacial acetic acid and denatured in 6 M urea solution. The inclusion body preparation of pro Pf PM1 was not completely solubilized without the aid of glacial acetic acid. Remarkable protein aggregates were observed dur ing refolding. The rest of the soluble materials were subjected to gel filtration ch romatography to isolate the folded proenzyme of proper size from the misfolded material. Results from the size exclusion chromatography revealed that the refolded pro Pf PM1 had the molecular mass of 43 kDa. Additionally, enzyme material from peak fractions showed catal ytic activity on the chromogenic peptide


66 Pf PM2 THKLKNYIKE S V N F*L N SGLT KTNYLG SSND Pf PM1 QHRLKNYIKE S L K F*F K TGLT QKPHLG NAGD K110pNQHRLKNYIKE S L N F*F K TGLT QKPHLG NAGD S4S2S1’ S3S1S2’ 110P1* = Potential cleavage sitePf= P. falciparumPf PM2 THKLKNYIKE S V N F*L N SGLT KTNYLG SSND Pf PM1 QHRLKNYIKE S L K F*F K TGLT QKPHLG NAGD K110pNQHRLKNYIKE S L N F*F K TGLT QKPHLG NAGD S4S2S1’ S3S1S2’ 110P1* = Potential cleavage sitePf= P. falciparum Figure 3-3. Sequence alignment of the Pf PM1 and Pf PM2 prosegment portion that potentially binds in the active site clef t. The prosegment portion is in black, the sequence of the mature enzyme is highlighted in green. The potential cleavage site is highlighted in blue. The active site pockets each residue proposed to project into ar e illustrated. The mutated re sidue is highlighted in red.


67 Table 3-1. Average representa tive yields during the produc tion and purification of the recombinant pro Pf PM1 from 1 liter expression. Production and purification st eps Average yields (mg)a Cell pellet (wet) 2520b Inclusion body (wet) 350b 6 M urea denaturation 45c Refolded dialysate 3.2c Anion exchange chromatography 2.1c Gel filtration chromatography 1.2c a The product yield of each step was the average result from three independent expressions. b Weights of products were direct ly measured after centrifugation. c The concentration of soluble protein was determined using OD280, with the extinction coefficiency 280 = 41,510 M-1 cm-1 (a theoretical value calculated from the sequence of semipro Pf PM1 using ProtParam (Gasteiger et al. 2005)).


68 Figure 3-4. Gel filtration purification chromatogram of the pro Pf PM1 K110pN mutant. The protein concentration of each fraction was represented by OD280. A 50 L of sample from each fraction was prei ncubated in 0.1 M sodium acetate, pH 4.5, at 37 C for 20 min, the enzymatic activity of each fraction was determined by measuring the initial cleavage rates on the chromogenic peptide substrate, Lys-Pro-Ile-Leu-Phe*NphArg-Leu (100 M). The recombinant zymogen material was collected as the 43 kDa monomer in a single peak.


69 Figure 3-5. SDS-PAGE analysis of overe xpression and purification of the pro Pf PM1 K110pN mutant. Lane 1: high molecular weight marker (RPN 756, Amersham); Lane 2: total cell lysate be fore IPTG induction (T = 0); Lane 3: total cell lysate after 3 h IPTG induction (T = 3); La ne 4: purified inclusion body form; Lane 5: soluble dialysate; Lane 6: purified zymogen by anion exchange chromatography; Lane 7: purified zymogen by gel filtration chromatography.


70 substrate A: Lys-Pro-IleLeu-Phe*Nph-Arg-Leu (* represents the bond where the cleavage occurs, and Nph = para -nitrophenylalanine) (Figure 3-4). Products from each preparation step were analy zed by SDS-PAGE (Figure 3-5). Optimal Conditions for the Catalysis of the Pf PM1 K110pN Mutant The catalytic activity profiles of the Pf PM1 K110pN mutant at different pH conditions were depicted from the initial cl eavage velocities on th e chromogenic peptide substrate A following a variety of preincubation times (Figure 3-6). Pf PM1 showed significant catalytic activity w ithin a narrow pH range of 4.0-4.5. Catalytic activities of this enzyme at other acidic conditions, such as pH 3.5 (Figure 3-6) and pH 5.0-6.0 (data not shown) were also tested but were hardly detected. At pH 4.0-4.5, the enzyme required 20 min to maximize its initial cleavage velo city. Longer preincubation times diminished the rates of substrate hydrolysis. Overall, th e optimal condition for this enzyme variant to perform catalysis is at pH 4.5 with 20 min preincubation. Such hydrolysis assays reflect the combined effects of zymogen maturati on and active enzyme catalysis. These two events can be individually a ddressed by SDS-PAGE analysis of self-cleavage of the pro Pf PM1 K110pN mutant and by studying th e optimal condition of mature Pf PM1 catalysis. Auto-activation of the Recombinant Pro Pf PM1 K110pN Mutant The molecular conversion from pro Pf PM1 to mature Pf PM1 was detected using SDS-PAGE based upon a time-dependent 5kDa molecular weight reduction under acidic conditions (Figure 3-7). The auto-activation events of the pro Pf PM1 K110pN mutant could be effectively tracked at pH 4.0 and 4.5. The conversion at pH 4.0 represented a quick self-cleavage


71 Figure 3-6. Determination of the optim al conditions for the catalysis of Pf PM1. The pro Pf PM1 K110pN mutant was incubated at 37 C in 0.1 M sodium formate, pH 3.5, 0.1 M sodium formate, pH 4.0, and 0.1 M sodium acetate, pH 4.5. The hydrolyses of the activated Pf PM1 on the chromogenic peptide substrate, LysPro-Ile-Leu-Phe*Nph-Arg-Leu (100 M) we re monitored at a variety of time points from 0 to 4 h, and the corresp onding initial cleavage velocities were measured. The results were normalized w ith the maximal initial rates set to 100 percent. The assays were repeated three times, from which the average and standard errors were calculated and plotted ag ainst the incubation time. The zymogen can be efficiently activated and performs catalysis at pH 4.0 and 4.5. The optimal condition for the catalysis of Pf PM1 is at pH 4.5 with 20 min preincubation.


72 Figure 3-7. SDS-PAGE analysis of time -resolved auto-activation of the pro Pf PM1 K110pN mutant. All the auto -activation assays were performed at 37 C by incubating 3 g of pro Pf PM1 in 0.1 M sodium formate, pH 3.5 (I); 0.1 M sodium formate, pH 4.0 (II); 0.1 M s odium acetate, pH 4.5 (III); and 0.1 M sodium citrate, pH 5.0 (IV). The convers ion from zymogen to mature enzyme was monitored at incubation time indicated (unit: min). The self-cleavage of pro Pf PM1 was effectively carried out at pH 4.0 and 4.5, as shown in (II) and (III).


73 process, where a majority of enzyme mate rials were activated within 20 min and the conversion was completed after 1 h incuba tion. At pH 4.5, self-cleavage was slowed down so that less than half of proenzyme materials were activ ated after 20 min incubation and the whole conversion required more than 2 h to complete. In addition, at both pH conditions, the total amount of mature Pf PM1 obtained was less than that of zymogen, indicating that the enzyme committed self-deg radation along with specific conversion. At pH 3.5, the protein was not able to mainta in stability as severe self-degradation dominated conversion followi ng acidification. As a resu lt, only a minute amount of mature enzyme was retained, which partiall y explained the weak catalytic activities observed in the substrate hydr olysis assays. At pH 5.0, th e enzyme conversion started after 15 min incubation along with subtle self-degradation. However, a major portion remained as proenzyme after 4 h incuba tion resulting in undet ectable enzymatic activities. Meanwhile, intermediate enzyme species during self-cleavage were observed at pH 3.5-4.5. These findings suggest that the auto-activation of pro Pf PM1 is a pH-dependent, time-resolved stepwise event, and that the optimal pH condition for pro Pf PM1 to conduct auto-activation is pH 4.0-4.5, where the enzyme exhibit efficient catalysis. And at such pH milieus, the converted enzyme is able to maintain stability for at least 2 h to allow isolation of the mature Pf PM1. Isolation of Mature Pf PM1 and Screening its Optimal pH Condition for Catalysis The mature Pf PM1 was prepared by converting the pro Pf PM1 K110pN mutant at pH 4.0, 37 C for 1 h and further purified via anion exchange chroma tography. The initial hydrolysis rates of the resulting enzyme at varied acidic pH conditions on the chromogenic peptide substrate A were measured (Figure 3-8). The mature Pf PM1


74 Figure 3-8. Determination of the optimal pH condition for the catalysis of mature Pf PM1. The mature Pf PM1 was incubated at 37 C for 3 min in a series of acidic buffers from pH 3.5.0. The initia l cleavage velocities of mature Pf PM1 on the chromogenic peptide substr ate, Lys-Pro-Ile-Leu-Phe*Nph-ArgLeu (100 M), were measured. The results were normalized with the maximal hydrolysis rate set to 100 percent. The average and standard error from duplicate assays were plotted against pH values. The optimal pH for mature Pf PM1 to perform catalysis was pH 5.5.


75 Figure 3-9. N-terminal protein sequencing analyses on the auto-converted Pf PM1. The naturally-occurring mature enzyme porti on is underlined. The mature enzyme sequence starts with Asn (Gluzman et al. 1994), which is indicated by the arrowhead. Two major cleavage events , occurring between Phe111p-Phe112p and between L116p-Thr117p, contributed even ly to the final mature species of this study. The N-terminal residues of the two resulting products are indicated by the arrows.


76 performed its catalysis most efficiently at pH 5.5. While at pH 5.0, 80% of the full catalytic activity was shown; only approximate ly 30% of the catalytic activity could be obtained from the mature Pf PM1 at pH 4.0 or pH 4.5, the optimal condition for zymogen auto-conversion. N-terminal seque ncing analysis on the mature Pf PM1 species revealed that self-cleavage occurred exclusivel y and equally between Phe111p-Phe112p and between Leu116p-Thr117p (Figure 3-9). Catalytic Efficiency of Mature PfPM1 Hydrolysis on Chromogenic Peptide Substrates The kinetic parameters ( kcat, Km and kcat/ Km) of the converted mature Pf PM1 cleavage of varied chromogenic peptide s ubstrates were determined (Table 3-2). Compared with the naturally-occurring fo rm (Tyas et al. 1999), the recombinant Pf PM1 showed comparable binding affinities ( Km) on substrate B and C; however, the catalytic constant ( kcat) of the recombinant Pf PM1 was approximately 6-fold and 2-fold lower than that of the naturally-occurri ng counterpart for substrate B and C, respectively, which resulted in up to a 7-fold less efficient catal ysis on these two substrates. In addition, the kinetic parameters of mature Pf PM1 on the peptide substrate A, utilized for testing catalytic activity, were also determined. Inhibition Effects of Compounds on Pf PM1 The dissociation constants of three competitive inhibitors against Pb PM4 were determined (Table 3-3). Pepstatin A is a ge neral tight binding inhi bitor of pepsin-like aspartic proteinases. Ro40-4388 is a peptidom imetic inhibitor with high binding affinity specific to Pf PM1 (Moon et al. 1997; Tyas et al . 1999). Both compounds showed strong antiparasitic activity on P. falciparum culture (Bailly et al. 1992; Moon et al. 1997). SQ3000 is a naturally-occurring compound that ca n be isolated from several species of


77Table 3-2. Kinetic parameters of naturally-occurring and recombinant Pf PM1 digestion on chromogenic peptide substrates. Plasmepsin 1 Recombinant K110pN Naturally-occurringa Substrate kcat (s-1) Km ( M) kcat/ Km (mM-1 s-1) kcat (s-1) Km ( M) kcat/ Km (mM-1 s-1) (A) Lys-Pro-Ile-Leu-Phe*Nph-Arg-Leu 1.1 0.113.2 1.785 15n.d.n.d.n.d. (B) Leu-Glu-Arg-Ile-Phe*Nph-Ser-Phe 2.6 0.18.6 0.2302 141682000 (C) Lys-Glu-Leu-Val-Phe*Nph-Ala-Leu-Lys 2.9 0.29.0 1.5322 5968750 a Data of the naturally-occurring form of Pf PM1 are cited from Tyas et al. 1999. n.d. = not determined.


78 Table 3-3. Dissociation constant s of inhibitory compounds bound to Pf PM1. Ki (nM) Inhibitor Naturally-occurringa Recombinant K110pN pepstatin A 0.71.4 0.2 Ro40-4388 9 25 4 SQ 3000 n.d.5581 669 a Data of the naturally-occurring form of Pf PM1 are cited from Tyas et al. 1999. n.d. = not determined N H H N N H H N N H OH OH OH O O O O O O pepstatin A O NH H N N H O O O O Ro40-4388 O OH H SQ 3000O N H H N N H H N N H OH OH OH O O O O O O pepstatin A O NH H N N H O O O O Ro40-4388 O OH H SQ 3000O

PAGE 100

79 marine sponges (Quideau et al. 2002). SQ 3000, also known as puupehenone, showed significant binding affinity to several plasme psins and remarkable antiparasitic activity (unpublished data). While pepstatin A show ed a low nanomolar binding affinity, the inhibition constant of Ro40-4388 was also in nanomolar magnitude. The Ki values of the recombinant Pf PM1 were comparable with those of th e naturally-occurring form of these two compounds. SQ 3000, on the other hand, exhibited the weakest binding to Pf PM1 with the Ki value in the micromolar range. Discussion A total of four plasmepsins have so far been identified to reside in the food vacuole and appear to be involved in initiation of hemoglobin degr adation. As their functional redundancy has been demonstrated by genetic disruption studies (Liu et al. 2005; OmaraOpyene et al. 2004), novel potent drugs agai nst plasmepsins are expected to block catalytic activities of more than one member. Understanding the similarities and deviations of the active site features of different plas mepsins relies on production and characterization of each of th e enzyme species. The naturally-occurring plasmepsin 1, as the initiator of the whole hemoglobin degrad ation process, has shown unique enzymatic properties on inhibitor binding, and needs to be further characterized at both the kinetic and the structural level. Preparation of sufficient amounts of the naturally-occurring plasmepsin 1 has proven to be difficult to accomplish, thus our efforts depend on successful production of the recombinant form using, for instance, the heterologous E. coli expression system. Previous studies indicated that the wild t ype recombinant proplasmepsin 1 was not able to conduct auto-activation (Moon et al. 1997). Therefore, we attempted to engineer the prosegment of wild type zymogen to crea te a self-cleavage site. As a successful

PAGE 101

80 precedent, Beyer et al. introduced an oligope ptide substrate sequence to the propart of human cathepsin D zymogen to replace the orig inal one. The resulting mutant was able to conduct auto-activation and the activated species showed comp arable kinetic parameters with the naturally-occurring enzyme (Bey er and Dunn 1996). Cons idering the ease of pro Pf PM2 auto-activation, we comp ared residues between pro Pf PM1 and pro Pf PM2 in the vicinity of a potential activation site, Phe111p-Phe/Leu112p, and substituted one of the most divergent residues hypothetically occupying the S2 pocket of plasmepsin 1 during activation for that found in pro Pf PM2 to generate the K110pN mutant. In addition, expression of the full length pro Pf PM1 in E. coli led to production of a truncated zymogen with the propart cut to only 52 resi dues (Luker et al. 1996) probably due to the toxicity of the membrane-spanning domain (F igure 3-1). Therefore, a truncated version of pro Pf PM1 mutant lacking the transmembrane domain was produced in E. coli . An average of 1.2 mg of purifie d activatable plasmepsin 1 zymogen was obtained from 1 liter of cell culture. However, consider ing the perfect overl apping between the OD280 and catalytic activity profile of the size exclusi on chromatography (Figur e 3-4), this further purification step might not be necess ary. Thus, 2.1 mg of activatable pro Pf PM1 could be produced on average out of 1 liter of cell cu lture representing nearly 5% of the total protein material. In terms of the yield, this has been an improvement compared with production of the previously reported recombinant Pf PM1 K110pV mutant where 0.9 mg of fully activatable zymogen was purified from 14 liters of cell cultu re representing only 0.6% of the original material (Moon et al . 1997). Different cons tructs (K110pN vs. K110pV) and denaturation and refolding met hods applied on the aggregate protein may be responsible for this yield deviation. Meanwhile, the conve rted species from these two

PAGE 102

81 mutants shared similar catalytic efficiency and binding affinity to substrates listed in Table 3-2 and inhibitor pe pstatin A (Moon et al. 1997). In addition to studies of the inhibition e ffects on the kinetic level, attempts have been made to grow crystals of Pf PM1 in complex with compounds of strong binding affinity. Highly purified and c oncentrated protein samples ar e required for this purpose. From 4 liters of cell culture, approximately 40 L of 6 mg/mL of purified mature Pf PM1 was extracted. Shown in Figure 3-10 is the SDS-PAGE analysis of mature Pf PM1 for crystal tray setup. Despite this, efforts still ne ed to be made to improve the yield of active Pf PM1 as crystals have not been successfu lly obtained. One strategy involves employing directed evolution (Arnold 1996), in this case, the DNA shuffling technique (Stemmer 1994) to build up a library of Pf PM1 variants, from which the ones that can increase the solubility will be selected using fluorescen ce emitted from green fluorescence protein (GFP) as readout (Cabantous et al. 2005). A series of point mutations will provide the source for the library construction. More deta ils of this method will be described in Chapter 8 “FUTURE DIRECTIONS”. Interestingly, the optimal pH for the auto-activation of pro Pf PM1 and that for the catalysis of mature Pf PM1 differ by approximately one pH unit. In one prospect, at the optimal condition (pH 5.5) for the catalysis of mature Pf PM1, the zymogen does not conduct efficient self-processing, hence no act ivity can be observed. In the other prospect, at the conditions (p H 4.0-4.5) for the most effici ent self-activation, only about 30% of full catalytic activities of mature enzy me is exhibited. As the pH of the transport vesicles and the food vacuole inside the P. falciparum parasite is around 5.0-5.4 (Krogstad et al. 1985; Yayon et al. 1984), it se ems impossible for the naturally-occurring

PAGE 103

82 Figure 3-10. SDS-PAGE analysis of purified recombinant mature Pf PM1 materials for crystal tray setup (Coomassie blue stain) . Lane 1-6: peak fractions of purified mature Pf PM1 (~ 38 kDa) before concentratio n; Lane 7-8: concentrated (6.6 mg/mL) mature Pf PM1–inhibitor complex, 40-fold dilution, loading volume: 10 L for lane 7, and 4 L for lane 8.

PAGE 104

83 pro Pf PM1 to conduct auto-activation in such compartments as well. Additionally, the junction between the prosegment and mature enzyme is not a favorite cleavage site for plasmepsins, indicating that there may exist another unrelated protease that acts as the convertase. Although the identity of this enzyme has not been defined yet, evidence confirming its presence is accumulating (Ban erjee et al. 2003; Francis et al. 1997a). Compared with the naturally-occurring form, the self-activated recombinant Pf PM1 is less catalytically efficient (Table 3-2) . This may be due to the presence of the prosegment tail up to 12 residues long, whic h could interfere with the efficiency of substrate turnover during catalysis. Conversely, Pf PM2 shares similar kcat values between its naturally-occurring and recombinant form , as a majority (62%) of the activated enzyme species retain only 2 residues of th e prosegment (Tyas et al. 1999). From this point of view, improvement of the catalytic efficiency of Pf PM1 relies on identification of the potential convertase. Neverthe less, the two distinct sources of Pf PM1 exhibit comparable binding affinities both on pe ptide substrates and on peptidomimetic inhibitors. The recombinant Pf PM1 may thus serve as a repr esentative of its naturallyoccurring counterpart for compound binding and inhibition analyses. Conclusion The recombinant pro Pf PM1 K110pN mutant has been overexpressed in an E. coli expression system. The inactive inclusion bodies can be extracted and the protein refolded. Approximately 5% of the inclusion body aggregates is conve rted to activatable Pf PM1 zymogen from 1 liter of cell culture. The pro Pf PM1 K110pN mutant conducts auto-activation most efficiently at pH 4.0 and pH 4.5; while, th e resulting mature Pf PM1 performs optimal catalysis at pH 5.5. N-termin al protein sequencing an alysis reveals that activation occurs at Phe111p-Phe112p and Leu116p-Thr117p. The recombinant Pf PM1

PAGE 105

84 shows less catalytic efficiency on substrat e hydrolysis than it s naturally-occurring counterpart, and yet they share comparable binding affinities on varied compounds. The successful production and characteri zation of active recombinant Pf PM1 provide further knowledge on the group of plasmepsin homol ogues and may benefit novel antimalarial drug design.

PAGE 106

85 CHAPTER 4 PRODUCTION AND ENZYMATIC CHARACTERIZATION OF RECOMBINANT PLASMEPSIN 4 ( Pb PM4) FROM Plasmodium berghei Introduction Plasmodium berghei is one of the four malaria para site species that infect rodents of western Africa (Kendrick 1978). In spite of their phylogenetic distance (Waters et al. 1991; Kendrick 1978), the basic biol ogy of the murine parasites closely resembles that of the human malaria parasites (Landau and B oulard 1978; Sinden 1978; Aikawa and Seed 1980). P. berghei is one of the few mammalian pa rasites whose bloodstages can be cultured in vitro , and massively collected and purifie d. In addition, cultu re conditions of P. berghei are highly similar to those for culture of P. falciparum (Janse et al. 1984; Janse et al. 1989). This valuable technique allows comparison of drug susceptibility of P. falciparum and P. berghei . Further more, P. berghei infected murine species can serve as a reliable animal model for assessment of in vivo drug metabolism at the pharmacodynamics and pharmacokinetics level. A total of seven genes have been iden tified to encode pepsin-like aspartic proteinases from the available data of the P. berghei Genome Project (Humphreys et al. 1999; Dame et al. 2003). Comparativ e genomics has been employed on the P. berghei sequence database to identify or thlogs of plasmepsins from P. falciparum . So far, only one enzyme species that shares high se quence similarity with the food vacuole plasmepsins of P. falciparum has been recognized. This plasmepsin of P. berghei shares the highest amino acid sequence id entity with plasmepsin 4 ( Pf PM4) of P. falciparum

PAGE 107

86 (54.5% for pro Pf PM4, and 58.2% for mature Pf PM4) and is nominated as plasmepsin 4 ( Pb PM4) of P. berghei (Dame et al. 2003). Located within chromosome 10 (Carlton et al. 1999), the Pb PM4 gene encodes a single polypeptide of 451 amino acids in length and is expressed as a zymogen containing a 124 amino acid long prosegment, of which a potential type II transmembrane domain re sides in the N-terminus (Figure 4-1). Molecular modeling of the mature enzyme indicat es its bilobal tertiary structure is highly composed of -strands (Figure 4-2) as most of the pepsin-like aspartic proteinases of the A1 family. Food vacuole plasmepsins, as a group, have been considered as prime targets for novel antimalarial drug design. Th e potency and toxicity of le ad compounds need to be studied in a mammalian system, such as pa rasite-infected mice. The high identity between Pb PM4 and food vacuole plasmepsins from parasites infecting man allows this species to mimic the target of antimalaria l compounds in a parasite-infected murine model. This chapter describes the cloning, expression, in vitro refolding and enzymatic characterization of the re combinant wild type pro Pb PM4. In addition, a pro Pb PM4 mutant (L117pE) allows production of a st able form of self-processed mature Pb PM4. This stable mature Pb PM4 was isolated and characterized as well and may serve as a subject for future structural studies. Results Cloning of the Recombinant Wild-type Pro Pb PM4 A hydrophobic segment of 20 amino acids in length is located approximately 35 residues from N-terminus of proplasmepsins. Expression of full length sequence results in a low yield of inclusion bodies likely due to the toxicity of this potential

PAGE 108

87 Figure 4-1. Nucleotide and protein sequence of pro Pb PM4 (GenBank Accession No. AJ223308). Shown is the full length seque nce. The potential transmembrane segment is underlined; the truncated prosegment (blue) plus the mature enzyme sequence (red) is the construct for expression. The pair of active site motifs are highlighted in green.

PAGE 109

88 Figure 4-2. Molecular mode ling structure of mature Pb PM4. The modeled ribbon structure was built by Amrita Madabushi Ph.D. of Prof. Robert McKenna group based upon the X-ray crys tallographic structure of Pf PM2 in complex with rs370 (1LF2, 1.80 ) (Asojo et al. 2002a) using the program SWISSMODEL (Peitsch 1996; Schwede T et al . 2003). The two cat alytic residues, Asp 34 and Asp 214, are rendered in space-filled presentation.

PAGE 110

89 transmembrane segment (Dame et al. 1994; L uker et al. 1996; Moon et al. 1997). Thus, a truncated form of the Pb PM4 zymogen lacking the N-terminal 75 residues was cloned from the genomic DNA of the P. berghei ANKA strain and inserted into the Bam H I site of pET-3a vector by Charles A. Yowell of Prof . John B. Dame’s laboratory, University of Florida, College of Veterinary Medicine, Gain esville, Florida. Muta tions were generated at the 112p and 117p residue of the prosegme nt, where Leu residues were replaced by Glu residues (Figure 4-3). The clones were confirmed by DNA sequencing analyses. Such a mutation was made based on the resu lts of the S1 subsite preference of Pb PM4 discussed in Chapter 5. Wild Type ttcaaattattaaaatcaggtttattaaaa F K L L K S G L L K Mutant ttcaaa gaa ttaaaatcaggt gaa ttaaaa F K E L K S G E L K 117p 112p Figure 4-3. Site-directe d mutagenesis of pro Pb PM4. The L112pE L117pE double mutant was generated based up on the L117pE mutant. Expression, in vitro Refolding, and Purification of Recombinant Pro Pb PM4 The IPTG-induced expression of the semipro Pb PM4 was carried out at 37 C for 3 h. E. coli cells were harvested and inclusion bodies were isolated. The protein from the inclusion body materials was denatured a nd refolded. The resulting materials were purified by anion exchange and size exclus ion chromatography. Shown in Figure 4-4 is the SDS-PAGE analysis of a protei n sample of each step. Semipro Pb PM4 was successfully purified from the total cell lysate. Similar results were obtained for the L117pE and L112pE L117pE mutants (data not shown). Meanwhile, the average yield of

PAGE 111

90 Figure 4-4. SDS-PAGE analysis of overexpression and purific ation of recombinant wild type pro Pb PM4. M: high molecular weight mark er; 1: total cell lysate before IPTG induction; 2: total cell lysate after 3 h IPTG induction; 3: purified inclusion body form; 4: soluble dial ysate; 5: purified zymogen by anion exchange chromatography; 6: pur ified zymogen by gel filtration chromatography.

PAGE 112

91 Table 4-1. Average representa tive yields during the produc tion and purification of the recombinant wild type pro Pb PM4 from 1 liter culture of expression. Production and purification st eps Average yields (mg)a Cell pellet (wet) 2780b Inclusion body (wet) 420b 6 M urea denaturation 60c Refolded dialysate 13c Anion exchange chromatography 8.3c Gel filtration chromatography 1.0c a The product yield of each step was the average result from three individual expressions. b Weights were directly measured after centrifugation. c The concentration of soluble protein was determined using OD280 with the extinction coefficiency 280 = 41,510 M-1 cm-1 (a theoretical value calculated from the sequence of semipro Pb PM4 using ProtParam (Gasteiger et al. 2005)).

PAGE 113

92 Figure 4-5. Size exclusion purification chromatogram of the wild type pro Pb PM4. Catalytically active material s (the second peak of OD280) were isolated from misfolded protein. An aliquot of 250 L of sample was used for testing OD280 of each fraction, and 20 L of sample from each fraction was used for catalytic activity test using the ch romogenic substrate: Lys-Pro-Ile-LeuPhe*Nph-Arg-Leu (100 M).

PAGE 114

93 product of each step from one liter cell cultu re was determined from three individual experiments (Table 4-1). The inclusion bodies accounted for about 15% of the total cell mass indicating that overexpression has b een achieved. A necessary size exclusion chromatography step allowed isolation of catalytically active materials from the misfolded protein (Figure 4-5). As a result , approximately 1.7% (w/w) of the denatured proenzyme materials was pur ified as highly active pro Pb PM4 of 43kDa. Similar yield profiles were obtained for the L117pE a nd L112pE L117pE mutants (data not shown). Self-processing of Recombinant Pro Pb PM4 The auto-cleavage event of the wild type pro Pb PM4 was studied at every half pH unit from pH 3.5-6.0. Shown in Figure 46 are SDS-PAGE analyses of pro Pb PM4 autocleavage time courses from pH 4.5-6.0. Self-p rocessing was efficiently conducted at pH 4.5 and pH 5.0 with an incubation time of le ss than 5 min. On the other hand, cleavage of the prosegment was significantly slowed down at pH 5.5 so that an incubation of more than 2 h was required to start the conver sion; while self-pro cessing at pH 6.0 was completely halted during a 12 h acidificat ion. These observations suggested autocleavage of pro Pb PM4 was a pH-sensitive and time-dependent process. Processing intermediates were observed during the convers ion (Figure 4-6 III), which indicated selfprocessing of pro Pb PM4 was a stepwise event. N-terminal protein sequencing analyses of the final processing at pH 4.5-6.0 exclusivel y identified the self-d igestion site between Leu117p and Leu118p (Figure 4-11), suggesting that self-processing of wild type pro Pb PM4 was site-specific under acidic c onditions. In addition, the wild type pro Pb PM4 maintained poor stability during se lf-processing as si gnificant amounts of enzyme committed nonspecific self-degradation.

PAGE 115

94 Figure 4-6. SDS-PAGE analysis of time-re solved self-processing of wild type pro Pb PM4. All the self-processing events were performed at 37 C, by incubating 2.4 g of wild type pro Pb PM4 in 0.1 M sodium acetate, pH 4.5 (I), 0.1 M sodium citrate, pH 5.0 (II), 0.1 M sodium citrate, pH 5.5 (III), and 0.1 M sodium phosphate, pH 6.0 (IV). The c onversion processes from zymogen to mature enzyme were monitored at time indicated (unit: min, O/N = overnight).

PAGE 116

95 Figure 4-7. SDS-PAGE analysis of time -resolved self-processing of the pro Pb PM4 L117pE mutant. All the se lf-processing events were performed at 37 C, by incubating 2.4 g of wild type pro Pb PM4 in 0.1 M sodium acetate, pH 4.5 (I), 0.1 M sodium citrate, pH 5.0 (II), 0.1 M sodium citrate, pH 5.5 (III), and 0.1 M sodium phosphate, pH 6.0 (IV). The c onversion processes from zymogen to mature enzyme were monitored at time indicated (unit: min, O/N = overnight).

PAGE 117

96 A L117pE mutant was then generated pr imarily to block the resulting autocleavage site and self-processing was subsequent ly investigated on this species. Shown in Figure 4-7 are SDS-PAGE analyses of time re solved auto-cleavage from pH 4.5-6.0. This particular mutation was not able to totally block self-diges tion which was carried out at other potential sites. This conclusion was confirmed by N-terminal protein sequencing analysis, as three different diges tion sites, F110p-K111p, L112p-L113p, and L113pK114p, were identified from the converted enzyme at pH 5.0 (Figure 4-11). Selfprocessing of this L117pE mutant was pH-s ensitive and was carried out in a timedependent manner as well. The auto-digestion pr ocess could be efficiently finished within 1 min and 5 min at pH 4.5 and pH 5.0, resp ectively. Although time required for complete conversion of zymogen at pH 5.5 was el ongated up to about 30 min, it was a more dynamic process than that of the wild type . Specific self-digestion was not observed at pH 6.0 until after 2 h incubation, and self-process ing of this mutated zymogen at such pH was accompanied with self-degradation. Additio nally, the stability of this converted L117pE mutant was well maintained under acidic conditions for hours and no intermediate converting species emerged. Th ese valuable properties of the mutant allowed isolation of the mature form of Pb PM4. Optimal Conditions for Pb PM4 Catalysis The optimal conditions for the wild type Pb PM4 to perform enzymatic catalysis were studied. The catalytic activity for the wild type Pb PM4 on hydrolysis of the chromogenic octapeptide subs trate A (Lys-Pro-Ile-Leu-Phe*Nph-Arg-Leu) was measured for each self-processing condition. Shown in Fi gure 4-8 are the time-resolved catalysis profiles of the wild type Pb PM4 from pH 3.5-6.0. Overall, the optimal pH for the wild type Pb PM4 to perform catalysis was around pH 5.0-5.5. An approximately 5 min

PAGE 118

97 Figure 4-8. Determination of the optimal c onditions for the catalysis of wild type Pb PM4. Experiments were performed at 37 C, in acidic buffer with pH from 3.5-6.0. The wild type enzyme was aci dified for the time indicated. Subsequently, enzyme-catalyzed initia l hydrolyzing rates on the chromogenic substrate, K-P-I-L-F*Nph-R-L (100 M), were measured and normalized with the maximal initial velocity set to 100 percent. Results were averaged from three individual assays.

PAGE 119

98 Figure 4-9. Determination of the optimal conditions for the catalysis of the Pb PM4 L117pE mutant. Experiments were perfor med at 37 C, in acidic buffer with pH from 4.5-6.0. The mutated enzyme was acidified for the time indicated. Subsequently, enzyme-catalyzed initia l hydrolyzing rates on the chromogenic substrate, K-P-I-L-F*Nph-R-L (100 M), were measured and normalized with the maximal initial velocity set to 100 percent. Results were averaged from three individual assays.

PAGE 120

99 preincubation was required for the enzyme to reach full activity and a majority of catalytic activity could be maintained give n a less than 30 min preincubation under such pH conditions. As the wild type Pb PM4 remained as the proenzyme form during 2 h preincubation at pH 5.5, this particular enzyme was capab le of enzymatic catalysis despite the presence of the prosegment. In addition, the optimal catalysis cond itions were determined for the L117pE mutant. Shown in Figure 4-9 are the enzy matic catalysis profiles from pH 4.5-6.0. Similar to the wild type, the mutated enzyme favored to perform catalysis at pH 5.0-5.5 where full enzymatic activity was shown afte r one minute of preincubation. Catalytic activity started to significan tly decrease with more than half an hour preincubation. Meanwhile, the mutated pro Pb PM4 zymogen was catalytically active in conditions such as pH 5.0 with 1 min preincubation and pH 5.5 with less than 10 min preincubation, which further confirmed that Pb PM4 was catalytically active ev en in the presence of the prosegment if only proper acid ic pH conditions were given. Isolation and Characterization of Mature Pb PM4 The mature form of Pb PM4, as has been discussed, could be processed from the L117pE mutant and further purified. The optimal pH conditions for enzymatic catalysis by mature Pb PM4 were addressed and the results ar e shown in Figure 4-10. This mature Pb PM4 showed maximal cataly tic activity at pH 5.5, and approximately 95% of full activity at pH 5.0. These pH conditions were co nsidered the optimal catalysis milieus for the mature enzyme. Comparison of Kinetic Parameters of Pb PM4 Species Varied forms of Pb PM4 have been created and show n to be catalytically active. The kinetic parameters of these enzyme ve rsions were determined using chromogenic

PAGE 121

100 Figure 4-10. pH-dependent cataly tic activity profile of mature Pb PM4. Experiments were performed at 37 C in acidic buffers from pH 3.5-6.0. Purified mature Pb PM4 was preincubated for 3 min. In itial hydrolyzing rates of enzyme catalysis on the chromogenic substrate, K-P-I-L-F*Nph-R-L (100 M), were then measured and normalized with the maximal initial velocity set to 100 percent. Results were averaged from three individual assays.

PAGE 122

101 Figure 4-11. N-terminal protein sequen cing analyses on self-processing of pro Pb PM4. Self-cleavage of pro Pb PM4 was carried out in a stepwise way. Shown is the semipro Pb PM4 and the converted enzyme species from the wild type Pb PM4, the L117pE and L112pE L117pE mutant . The two residues at 112p and 117p are highlighted with arrows. The N-te rminal residues of converted enzymes are labeled and highlighted with arro wheads. The N-terminal part of the expression construct belonging to pET-3a and the mature Pb PM4, whose sequence is determined by amino aci d sequence alignment with human plasmepsin homologues, are repres ented as black and blue boxes, respectively. The sequence of the 48 amino acid long prosegment (77p-124p) of pro Pb PM4 is listed in red.

PAGE 123

102Table 4-2. Comparison of kinetic parameters of Pb PM4 species digestion on chromo genic peptide substrates Processed from wild type pro Pb PM4 Processed from pro Pb PM4 L117pE Purified from pro Pb PM4 L117pE Wild type pro Pb PM4 pro Pb PM4 L117pE Substrate kcat (s-1) Km ( M) kcat/ Km ( M-1 s-1) kcat (s-1) Km ( M) kcat/ Km ( M-1 s-1) kcat (s-1) Km ( M) kcat/ Km ( M-1 s-1) kcat (s-1) Km ( M) kcat/ Km ( M-1 s-1) kcat (s-1) Km ( M) kcat/ Km ( M-1 s-1) K-P-I-LF*Nph-R-L 34.9 5.7 2.1 0.2 16.5 2.9 40.5 4.7 1.4 0.1 28.9 4.2 28.5 1.8 1.2 0.1 23.8 2.5 n.d.>100n.d.n.d.>100n.d. K-P-I-QF*Nph-R-L 24.5 1.8 3.8 0.3 6.4 0.7 22.8 1.5 3.5 0.3 6.5 0.8 26.9 3.1 3.4 0.2 8.0 1.1 n.d.>100n.d.n.d.>100n.d. Kinetic assays were performed in 0.1 M sodi um citrate, pH 5.0, at 37 C, with 5 min preincubation for the converted species; an d performed in 0.1 M sodium citrate, pH 5.5, at 37 C, with 5 min preincubation for the pro Pb PM4 species. n.d. = not determined

PAGE 124

103 peptide substrates: Lys-Pro-Ile-Leu-Phe*N ph-Arg-Leu and Lys-Pr o-Ile-Gln-Phe*NphArg-Leu (* represents the pe ptide bond for cleavag e) (Table 4-2). In order to understand the disparities of catalytic efficiency of different enzyme preparations, consistent conditions were adopted. For the converted en zyme forms, kinetic studies were carried out at pH 5.0 with 5 min pr eincubation and for the pro Pb PM4 forms, at pH 5.5 with 3 min preincubation. The three converted Pb PM4 species, Pb PM4 directly processed from the wild type, Pb PM4 directly processed from the L117pE mutant, and Pb PM4 purified from the converted L117pE mutant, exhibited comparable kcat and kcat/ Km values for both substrates, suggesting there be no prosegme nt mediated inhibition on the converted enzyme. However, the binding affini ties of the two substrates on pro Pb PM4 were remarkably lower than the converted forms (> 100 M vs. < 5 M), causing catalytic constants ( kcat) inaccessible to be accurately determined. Inhibition Analyses of Compounds Bound to Pb PM4 The binding affinities of three competitiv e inhibitors, pepstatin A, Ro40-4388 and SQ 3000 to Pb PM4 were determined (Table 4-3). Pb PM4 was converted from the wild type pro Pb PM4 at 37 C, pH 5.0 with 5 min incu bation. While pepstatin A showed a subnanomolar binding affinity, the inhibiti on constant of Ro40-4388 was in nanomolar magnitude. SQ 3000, on the other hand, exhibited the weakest binding to Pb PM4 with the Ki value in the micromolar range. Table 4-3. Dissociation constant s of inhibitory compounds bound to Pb PM4. Inhibitor Ki (nM) pepstatin A 0.11 0.02 Ro40-4388 135 21 SQ 3000 4123 50

PAGE 125

104 Discussion Plasmepsins have been recognized as pr omising targets for new antimalarial drug design. As part of the progressive developm ent of potential plasme psin inhibitors, the antiparasitic and antimalarial pr operties of these compounds on an animal model, such as P. berghei infected murine species, are urgently required to be investigated. For this purpose, it is of great significance to characterize the plasmepsins of P. berghei that are homologous to those of hu man malaria parasites. Plasmepsin 4 ( Pb PM4) of P. berghei has been identified as, thus far, the only orthlog closely related to food vacuole plasmepsins of P. falciparum (Dame et al. 2003). In addition, the amino acid sequence of Pb PM4, as shown in Figure 4-1, is considered, by comparative genomics studies, highly similar to that of its intronless counterparts of the P. falciparum species (Banerjee et al. 2002, Dame et al. 2003). In spite of these findings, little is known about the tim ing of expression and physio logical functions of the naturally-occurring Pb PM4. Recently, by the immunofluorescent assay microscopy technique, Pb PM4 has been located both in the f ood vacuole and in the cytosol of P. berghei culture of the bloodstage (Prof. John B. Dame, personal communication). Since Pf PM2 and Pf PM4 have the ability to hydrolyze spectri n at neutral pH of the cytosol (Le Bonniec et al. 1999; Wyatt and Be rry 2002), the roles played by Pb PM4 in the life cycle of the parasite also may not be limited to hemoglobin digestion in the food vacuole. The recombinant pro Pb PM4 shows full catalytic activity in the presence of the prosegment. This special behavior has not b een known in any other plasmepsins studied so far. However, a similar feature has been identified in another aspartic proteinase, secretase (BACE), where the prosegment doe s not suppress enzyme activity but appears to facilitate proper folding of the active proteinase domain (Shi et al. 2001). As the

PAGE 126

105 recombinant Pf PM2 can be expressed and correctly fold ed in the absence of its truncated prosegment (Istvan and Goldberg 2005), th e sequence encoding the last 326 amino acid residues of Pb PM4 was cloned and expressed in E. coli . However, the expression failed possibly due to self-degradation, indicating the N-terminal prosegment may play a critical role of stabilizing the mature enzyme. In the presence of the truncated prosegment, Pb PM4 has Km values two orders greater than thos e for the converted enzyme. This may be due to the competitive binding of the N-term inal flexible segment to the active site cleft or because the active site cleft resulting from incomplete conformational alteration does not properly fit the substrates. Several cleavage sites in the proregion of Pb PM4 have been identified by Nterminal protein sequencing analysis and th e catalytic activities of varied converted species have been confirmed by kinetic anal ysis. Strictly, these processed products may all be pseudomature Pb PM4s, since the naturally-occurring mature enzyme has not yet been isolated and studied. Solutions to quest ions, such as whether the enzyme functioning in the parasite is the same species as that determined by sequence alignment, and whether the naturally-occurring enzyme shares comparable kinetic features with the recombinant forms, will rely upon isolation and puri fication of the naturally-occurring Pb PM4. In addition, self-processing of recombinant pro Pb PM4 is carried out in multiple steps, which was clearly observed from cleavage of the L112pE L117pE mutant from pH 5.5 (results not shown) and confirmed by N-terminal sequenci ng analyses of the wild type and mutated pro Pb PM4. Multiple cleavage sites generated by auto-activation has been monitored in other plasmepsins, such as Pf PM1 (Chapter 3), Pf PM2 (Tyas et al. 1999) and food vacuole plasmepsins of nonfalciparum parasites infecting man, such as Po PM4

PAGE 127

106 and Pm PM4 (Li et al. 2004). The pro Pb PM4 K111pI mutant, a catalytically active form, was previously reported (Humphreys et al. 1999 ). N-terminal protei n sequencing analysis of the converted mature enzyme resulted in two major cleavage si tes: one digestion occurred between Phe110p and Ile111p and the other was between Leu112p and Leu113p, which is consistent with the results we have obtained from the L117pE mutant. Additionally, we have iden tified a minor cleavage site between Leu113p and Lys114p accounting for no more than 10% of the pr ocessing event. Meanwhile, a remarkable prosegment digestion can even occur in th e presence of Lys111p. These findings indicate that the hot spots for Pb PM4 self-processing are where hydrophobic residues reside in the potential P1 position regardless of the properties of residue s at the P1’ position. The catalytic efficiency (represented by kcat and kcat/ Km) and binding affinity (represented by Km) of converted Pb PM4 species are quite comparable on chromogenic substrates. The consistent performance of enzymatic catalysis among wild type and the mutated enzyme and among the purified and nonpur ified sample indicates that the extra 4-7 residues at the N-terminus of Pb PM4 do not affect the catalytic properties of Pb PM4 and a further purification step seems not necessary for kinetic studies of this enzyme. Meanwhile, the values of kinetic parameters on the substrates as well as the inhibitor pepstatin A are in the same magnitude as other food vacuole plasmepsins of human malaria parasites (Li et al. 2004; Tyas et al. 1999; Westling et al . 1999; Westling et al. 1997), indicating the active site cleft of the folded mature Pb PM4 and those of its orthlogs are quite alike. Less than 2% of total protein is converte d to catalytically active material. This relatively low yield compared with the recombinant Pf PM2 preparation is mainly due to

PAGE 128

107 the major portion of misfolded protein. Di fferent refolding approaches could be employed to improve the yield of catalyti cally active material. On the other hand, mutations that can enhance the solubility of Pb PM4 are required for yield improvement. Mutations can be designed, according to the molecular modeling structure, on the surface of the enzyme by substitution of hydrophobic residues for polar and charged ones. A directed evolution methods (Arnold 1996; Stemmer 1994) can be employed to reshuffle these mutations. The resulting mutation recombinants can be systematically screened using a GFP split system (Caban tous et al. 2005). This st rategy of improvement of recombinant enzyme yields will be discussed in detail in chapter 8. Mutations that enhance production of catalytic ally active and stable enzyme materials are expected to benefit X-ray crystallogr aphy based enzyme–inhibito r interaction studies. Conclusion The wild type and two mutated recombinant plasmepsin 4s ( Pb PM4s) of the rodent malaria parasite P. berghei have been successfully expressed in E. coli . Catalytically active enzyme materi als were obtained by in vitro refolding of protein from inclusion bodies. All three versions of Pb PM4 performed efficient se lf-processing at pH 4.5-5.5. The optimal conditions for the three versions of Pb PM4 to perform catalysis are pH 5.05.5 with about 5 min of preinc ubation at 37 C. A converted Pb PM4 of high stability was isolated from self-processing of the L117pE mutant. The optimal pH for catalysis of this mature species at 37 C was also around 5.0-5.5. The converted species from the wild type and from the L117pE mutant showed co mparable catalytic e fficiency and binding affinity on chromogenic substrates. The converted mature Pb PM4 showed similar values of kinetic parameters on chromogenic substr ates and the typical pepsin-like aspartic proteinase inhibitor, pepstatin A. These studies may help understand the enzymatic

PAGE 129

108 features of Pb PM4, a potential drug target in a useful malaria parasite infected animal model.

PAGE 130

109 CHAPTER 5 EXPLORATION OF SUBSITE PREFERENCES OF Pf PM1 AND Pb PM4 USING COMBINATORIAL CHEMISTRY BASED PEPTIDE LIBRARIES Introduction Extensive structural studie s have demonstrated that most endopeptidases bear elongated active site clefts that can interact with substrate residues. Such binding clefts favor accommodation of substrates with an extended -strand conformation. The binding mode of a peptide substrate to an endopeptidase is shown in Figure 5-1. The subsite pockets of the endopeptidase are labeled as S4 -S3’; while the side chains of the ligand residues, projecting to the co rresponding subsites, are named P4-P3’ from the amino to carboxyl terminus (Schechter and Berger 1967). In such an arra ngement, the side chain of each residue interplays with residues that comprise a defined pocket. Subsite preferences of endopeptidases originate from distinctive interactions between residues of the ligand and receptor in each binding pocket of different enzymes as well as in different pockets of a defined enzyme. Recognition of s ubsite preferences between homologous endopeptidases can facilitate th e design of specific substrates and inhibitors for targeted enzymes. For this purpose, two sets of combinatorial chemistry based peptide libraries were employed to investigate the S3-S3’ subsite preferences of Pf PM1 (Chapter 3) and Pb PM4 (Chapter 4). The P1 combinatorial library wa s composed of 19 indi vidual peptide pools. The sequence backbone, Lys-ProXaa-Glu-P1*Nph-Xaa-Leu (Nph = paranitrophenylanaline, and * represents the pept ide bond to be cleaved) was designed (Beyer

PAGE 131

110 Figure 5-1. Schematic diagram of a peptide subs trate fitting to the active site cleft of an endopeptidase. The subsite pockets are nom inated S4-S3’ and the side chains that are accommodated in these subsites are nominated P4-P3’. The enzymecatalyzed hydrolysis occurs at the peptide bond between P1 and P1’ (Schetcher and Berger 1967). Subsites S3, S1 and S2’ are located in the amino-terminal domain of the enzyme, such as the plasmepsins; whereas subsites S4, S2, S1’ and S3’ reside in the carboxy-terminal domain of the enzyme.

PAGE 132

111 et al. 2005) from the previous substrate specifi city studies of aspart ic proteinases (Dunn et al. 1986; Dunn and Hung 2000; Westling et al. 1999). Each pool is nominated after the residue at the P1 position. Th ese residues include the 20 natu ral amino acids in addition to norLeu, except for Met and Cys. A mixtur e of the aforementioned 19 amino acids is incorporated in the P3 and P2’ position, respectively, which offers a total of 361 peptide species (19 19) for an indi vidual pool, and 6859 for the whole library (19 19 ). The P1’ combinatorial library, on the other hand, contains the sequence backbone LysPro-Ile-Xaa-Nph*P1’-Gln-Xaa. For each peptide pool, one of the 20 natural amino acids plus norLeu, excluding Met and Cys, is inco rporated in the P1’ position; while the mixture of these residues is accommodate d in the P2 and P3’ position. A specific cleavage of each of the P1 or P1’ libra ry pools generates 19 pentapeptides and 19 tripeptides as shown in Figure 5-2. In this study, the primary substrate specificities of Pf PM1 and Pb PM4 at the P1 and P1’ position were determined by comparison of the initial hydrolysis rate of peptide library pools; while the secondary specificiti es at the P3, P2, P2’ and P3’ position were identified by measuring the relative abundan ces of cleavage products using the in-line liquid chromatography mass spectrometry (LC-MS) techniques. Peptidomimetic inhibitors of Pf PM1 and Pb PM4 with high selectivity against human cathepsin D were therefore developed. The second part of this chapter encompasses in vitro inhibition analyses of a series of peptidomimetic inhibitors on Pf PM1 and Pb PM4. These compounds, developed from single substrate kinetic analyses (Li et al . 2004; Westling et al. 1999; Westling et al. 1997) or screened out from combinatorial chemis try based peptide libraries (Beyer et al.

PAGE 133

112 Figure 5-2. Schematic diagram of the enzyma tic digestion of individual P1 and P1’ library pools.

PAGE 134

113 2005), are designed to target the food vacuol e plasmepsins of human malaria parasites ( Pf PM2, Pf PM4, Pv PM4, Po PM4 and Pm PM4). The dissociation constants ( Ki) of nine compounds against Pf PM1 and Pb PM4 were determined and potential binding features of the best compounds were explored using molecular modeling. All these studies enrich the knowledge for designing peptidomimetic inhibitors of plasmepsins with high specificity and selectivity. Results Subsite Preferences The hydrolyses of the P1 and P1’ combinatorial libraries by Pf PM1 and Pb PM4 were performed at their optimal catalysis conditions, which have been discussed in Chapter 3 and 4, respectively. The amino aci d acceptances at each subsite were ranked according to the normalized cleavage velocity for primary specificities or to the percent relative abundances of the hydr olyzed pentaor tripeptides for secondary specificities. Primary specificity The sole difference between peptide pools of a defined library dwells in the amino acids of the P1 or P1’ position. Therefore, the initial rates, re presenting the primary specificities, are directly affected by the residue identities at these two cleavage sites. The initial cleavage velocities of the two newly characterized proteinases, Pf PM1 and Pb PM4, on the P1 and P1’ library pools ar e listed in Table 5-1 and Table 5-2, respectively. These values were further pr esented as normalized percentage with the maximum rate set to 100 and the normalized values for both enzymes were plotted against each peptide pool to define the P1 (F igure 5-3) and P1’ specificity (Figure 5-4). Specificity of the S1 subsite. As shown in Figure 5-3 (A), Pf PM1 exclusively preferred accommodation of bul ky hydrophobic residues in the S1 subsite. The best two

PAGE 135

114 substitutions were phenylalani ne and leucine. In what fo llows, tyrosine and norleucine were moderately acceptable. Other P1 residue s, including tryptophan and isoleucine, led to poor recognition by Pf PM1. Similar to the outcome for Pf PM1, phenylalanine was the most favorite P1 amino acid substitution for Pb PM4. However, the branched hydrophobic residues, leucine and norleucine, the two second best substitutions, only resulted in 40 percent of the maximal cleavage efficiency. Four other P1 amino aci ds, asparagines, glutamine, tyrosine and tryptophan, also exhibited minor acceptances; whereas digesti on of the rest of the P1 peptide libraries by Pb PM4 were hardly allowed. These findings suggested that both Pf PM1 and Pb PM4 preferred fitting large hydrophobic residues in the S1 pocket and Pb PM4 seemed more tolerant to residues of different properties than Pf PM1. Specificity of the S1’ subsite. The optimal P1’ amino acid substitutions for Pf PM1 and Pb PM4 were consistently hydrophobic re sidues (Figure 5-4). Instead of phenylalanine, leucine and norle ucine were the most favorab le amino acids at P1’ for Pf PM1 and Pb PM4, respectively. For Pf PM1, the other aromatic or branched hydrophobic substitutions except for tryptophan spanned 40-75 percent of the maximal cleavage efficiency; whereas, these residues were also well fitted in the S1’ subsite of Pb PM4 as such peptide pools were digested mo re than 70 percent as efficiently as the P1’-norLeu library was. Notably, for both enzyme s, the initial rates decreased as the size of side chains decreased from leucine to glycine or expanded to tryptophan, which was potentially due to lacking or too close surface contacts with residues of the S1’ subsite. In addition, similar to the results from the P1 library, both Pf PM1 and Pb PM4 did not

PAGE 136

115 Table 5-1. Initial cleavage velocities (AU/s ec) of the P1 combinatorial peptide library pools by Pf PM1 and Pb PM4. P1 Amino Acid Pf PM1 P1 Amino Acid Pb PM4 F 24.2 F 43.6 L 22.2 nL 18.2 Y 9.82 L 15.0 nL 9.50 N 10.7 W 1.04 Q 8.92 A 0.87 Y 7.77 P 0.68 W 7.37 N 0.60 I 2.80 G 0.57 V 2.60 D 0.55 S 1.92 R 0.53 P 1.80 K 0.40 D 1.78 V 0.35 A 1.69 E 0.28 G 0.87 S 0.28 E 0.84 I 0.21 K 0.75 T 0.20 R 0.34 Q 0.16 T 0.25 H 0.07 H 0.06 The initial hydrolysis rates of peptide pools were listed in descending order for Pf PM1 and Pb PM4.

PAGE 137

116 Table 5-2. Initial cleavage velocities (AU/s ec) of the P1’ combinatorial peptide library pools by Pf PM1 and Pb PM4. P1’ Amino Acid Pf PM1 P1’ Amino Acid Pb PM4 L 159.1 nL 75.0 nL 116.6 Y 72.1 F 101.9 F 67.4 V 101.0 L 67.0 I 72.0 I 60.5 Y 63.7 V 55.0 A 50.4 T 15.4 T 40.3 A 15.0 S 23.7 W 11.0 W 13.6 R 8.43 Q 5.50 P 3.55 K 3.10 N 2.84 N 2.10 H 2.51 P 1.90 D 1.99 R 1.79 E 1.88 G 1.53 K 1.62 E 0.96 Q 1.59 H 0.24 G 1.46 D 0.12 S 1.21 The initial hydrolysis rates of peptide pools were listed in descending order for Pf PM1 and Pb PM4.

PAGE 138

117 Figure 5-3. The P1 amino acid preferences of the malarial aspartic proteinases, Pf PM1 (A) and Pb PM4 (B).

PAGE 139

118 Figure 5-4. The P1’ amino acid preferences of the malarial aspartic proteinases, Pf PM1 (A) and Pb PM4 (B).

PAGE 140

119 exhibit remarkable hydrolysis on peptide pools bearing most of the polar or charged P1’ residues. Secondary specificity The best peptide pools among the P1 a nd P1’ combinatorial libraries were determined from spectroscopic assays. These libraries were subse quently utilized to explore the extended secondary subsite preferences of Pf PM1 and Pb PM4 by LC-MS analysis. A specific cleavage between the P1 and P1’ residue produced a 19 non-prime side pentapeptide subgroup and a 19 prime si de tripeptide one. Peptides within each subgroup only differed in one amino acid subs titution, which was identified based on the mass and retention time of its related peptid e. The assigned products were quantified by integrating their relevant ion peaks. The three optimal P1 libraries hydrolyzed by Pf PM1 were P1-Phe, P1-Leu and P1Tyr; and P1-Phe, P1-norLeu and P1 -Leu were the best triad for Pb PM4. Therefore these pools were digested and analyzed to discover the S3 and S2’ subsite preferences. On the other hand, exploring the S2 and S3’ specificiti es relies on studies of the P1’ libraries. P1’-Leu, P1’-norLeu and P1’-Phe, the three pools with the highest initial velocities were selected for Pf PM1 and P1’-norLeu, P1’-Tyr a nd P1’-Phe were chosen for Pb PM4. Notably, the extended subsite preferences sole ly represented the results of the enzymecatalyzed zero-order reactions. In all figures for the secondary specificities, the residues are listed in order of the retention time of their related peptides. Specificity of the S3 subsite. The results of the S3 subs ite preferences from LCMS analysis of catalytic hydrolys is of the best P1 libraries by Pf PM1 and Pb PM4 are shown in Figure 5-5 and Figure 5-6, respectively.

PAGE 141

120 Large hydrophobic amino acids including bot h the branched and aromatic residues were strongly favored in the S3 subsite of Pf PM1. Phenylalanine was overall the optimal substitution of this particul ar position. The S3 subsite preferences were somehow deviated in the context of different P1 re sidues. As an example, P3-Leu was equally favored with P3-Phe in the P1-Leu library, but was merely detected in the P1-Tyr pool. Further, only the peptides containing aroma tic phenylalanine and tryptophan at P3 could be recognized by Pf PM1 when Tyr resided at P1. Such evidence indicated that the subsite preferences significantly depended on the cont ext of the peptide sequence fitted to the proteinases. Similarly, hydrophobic residues of bulky size were exclusively favored in the S3 subsite of Pb PM4 for all the three libraries st udied. Peptide products containing phenylalanine, leucine, norleuc ine and isoleucine at P3 we re the most abundant in the tested library pools. P3-Trp were preferre d when phenylalanine or leucine, but not norleucine was in the P1 position; whereas P3 -Tyr containing peptide products were only enriched in the P1-Phe library. Surprisingl y, the P3-Val containi ng pentapeptides could only be detected when phenylalanine dwe lled in P1, which was possibly because the relatively short side chain of valine insuffici ently interacted with residues composing the S3 subsite pockets. In contra st, polar and charged amino acids were hardly recognized in the P3 position except for glutamic acid, wh ich maintained a 15-20 percent of maximum abundance in all three tested libraries. Specificity of the S2 subsite. From LC-MS analysis of the best P1’ libraries, specificities of the S2 subsites for Pf PM1 and Pb PM4 are depicted in Figure 5-7 and Figure 5-8, respectively.

PAGE 142

121 Figure 5-5. The P3 amino acid preferences of the malarial aspartic proteinase Pf PM1.

PAGE 143

122 Figure 5-6. The P3 amino acid preferences of the malarial aspartic proteinase Pb PM4.

PAGE 144

123 The S2 subsite resides in the carboxy-ter minal domain of plasmepsins. At this position, Pf PM1 could accept varied types of residue s. For all three libraries, the best amino acid substitution was consistently seri ne, which was followed by isoleucine. Most of the hydrophobic residues were moderately a ccepted regardless of the different sizes, except for the largest tryptophan, which was disfavored by Pf PM1. Another interesting finding came from glycine and glutamic acid. Bo th residues were inconsistently accepted in the context of different P1’ amino acids , as the disfavored P2-Gly and P2-Glu containing peptides in the P1’-Phe library we re decently accepted in the other two pools. These findings, in addition to features exhibi ted in the S3 subsites of both enzymes, again indicated the mutual influences among resi dues of varied positions upon enzyme binding. Similarly, the S2 subsite of Pb PM4 was highly tolerant to P2 amino acids of different properties as most of the substitutions rendered at least 20 percent of the maximum abundance except for the three basic re sidues, lysine, arginine and histidine. For the three P1’ libraries investigated, glutamic acid was consistently the most favorable P2 residue and the runners-up were isoleucine and serine. Despite these residues, the rest of the hydrophobic amino acids were, on average, more favorable than hydrophilic ones. Nevertheless, unlike the hydrophobicity-overwhe lmed S3 pocket, the S2 subsite of Pb PM4 as well as Pf PM1 allowed acceptance of resi dues with varied features. Specificity of the S2’ subsite. Studies on the S2’ subsite preferences relied on analysis of the relative abundances of digested tripeptides from the best P1 libraries. The P2’ residue specificities of Pf PM1 and Pb PM4 were exhibited in Figure 5-9 and Figure 510, respectively.

PAGE 145

124 Figure 5-7. The P2 amino acid preferences of the malarial aspartic proteinase Pf PM1.

PAGE 146

125 Figure 5-8. The P2 amino acid preferences of the malarial aspartic proteinase Pb PM4.

PAGE 147

126 For Pf PM1, the results indicated that glutamine was dominantly favored in the S2’ subsite for the three tested libraries. Peptides containing the other s ubstitutions were all less than 30% abundant, though the acceptances of residues of varied features were comparable and detectable. As the results from the P1-Phe, P1-Leu a nd P1-norLeu libraries showed, serine and glutamine were generally pref erred in the S2’ subsite of Pb PM4, except that for the P1Leu library, P2’-Trp was accepted as the be st. Hydrophobic residues except for proline and hydrophilic residues such as threonine and glutamic acid can be accommodated equally well in the S2’ pocket. However, char ged residues such as aspartic acid and the three basic amino acids we re not favored. Overall, Pb PM4 adopted residues of varied types of the P2’ position similar as it behaved at the S2 subsite. The two most favored P2’ substitutions of Pb PM4 were serine and glutamine. Specificity of the S3’ subsite. Figure 5-11 and Figure 5-12 revealed the S3’ subsite preferences for Pf PM1 and Pb PM4, respectively. The relative abundances were obtained from LC-MS analysis of the tripeptide isolated from the optimal P1’ libraries. Pf PM1 exclusively preferred bulky hydrophobic re sidues at the S3’ subsite, such as leucine, isoleucine, norleucine, phenylalani ne and tryptophan. Over all the best amino acid substitutions were phenylalanine and isol eucine. In the P1’-Phe library, valine and tyrosine were also well accepted. Further, Pb PM4 conducted selection from an even narrower range for the S3’ subsite compared with the results from Pf PM1. As shown in Figure 5-12, only the aromatic amino acids, tryptophan and phenylal anine were well accepted for the P1’-Leu and P1’-Tyr libraries and the isomeric leucin e, norleucine and isoleucine were also

PAGE 148

127 Figure 5-9. The P2’ amino acid preferences of the malarial aspartic proteinase Pf PM1.

PAGE 149

128 Figure 5-10. The P2’ amino acid preferences of the malarial as partic proteinase Pb PM4.

PAGE 150

129 Figure 5-11. The P3’ amino acid preferences of the malarial as partic proteinase Pf PM1.

PAGE 151

130 Figure 5-12. The P3’ amino acid preferences of the malarial as partic proteinase Pb PM4.

PAGE 152

131 counted in the P1’-Phe library. Peptides c ontaining other P3’ amino acid substitutions were hardly detected in these library pools. Inhibitor Design The best P3-P3’ amino acid substitutions for Pf PM1 and Pb PM4 were extracted from the combinatorial library studies and listed in Table 5-3. Further, the subsite preferences of Pf PM1 and Pb PM4 were compared with those of human cathepsin D (hCatD) (Beyer 2003). Amino acid substitutions drawing the largest difference on initial velocities or relative abundan ces from hCatD were incorpor ated and the P1-P1’ peptide bonds were modified as methyleneamino [-CH2-NH-]. Such a strategy may facilitate identification of selective pe ptidomimetic inhibitors of Pf PM1 and Pb PM4 against their most homologous human aspartic proteinase . The sequences and structures of these inhibitors are shown in Figure 5-13. Table 5-3. Optimal peptide sequence for Pf PM1 and Pb PM4 determined from analyses of the P1 and P1’ combinatorial libraries P5 P4 P3 P2 P1 P1’ P2’ P3’ Pf PM1 Lys Pro Phe Leu Ser Phe Leu Gln Phe Ile Pb PM4 Lys Pro Phe Leu Ile Glu Phe Nle Ser Gln Trp Phe Ile The optimal amino acid substitutions of each lib rary and those with more than 95% of normalized initial velocity or relative a bundance within each pool are considered. The best residues are listed on top. The best substitution is put on top. Nle = norleucine. Characterization of Single Peptidomimetic Inhibitors Kinetic analyses The dissociation constants ( Ki) of nine peptidomimetic inhibitors were determined on the two newly characterized plasmepsins, Pf PM1 and Pb PM4, and the results are shown in Table 5-4. Developed using comb inatorial chemistry methods, compounds 1-5

PAGE 153

132 Figure 5-13. Structures of the inhibitors designed from the combinatorial approach. The reduced scissile bonds between the P1 and P1’ amino acid substitutions are represented by ([-CH2-NH-]).

PAGE 154

133 are selective inhibitors of food vacuole pl asmepsins vesus hCatD (Beyer et al. 2005). Compounds 6-9, on the other hand, are designe d based on kinetic analyses of single peptide substrates. Combining the amino acid su bstitutions that fit best at each subsite, these compounds are generally strong plasmepsin-binding inhibitors. Overall, these peptidomimetic compounds were not tight-bindi ng inhibitors of Pf PM1 as seven of nine inhibited the enzyme in the micromolar range. These compounds also showed poor selectivity for Pf PM1 against hCatD. Although two compounds (compounds 5 and 9) had comparable inhibito ry activities between these two enzymes, the other seven inhibited Pf PM1 1000-fold more weakly than hCatD. This further implied the necessity of designing peptidom imetic inhibitors specific to Pf PM1. In addition, compounds 1-6 exhibited a sim ilar inhibition profile to Pf PM1 as that to Pf PM2, which indicated that the subsite features of Pf PM1 were somehow more similar to those of Pf PM2. On the other hand, five of nine compounds bound to Pb PM4 in the picomolar to nanomolar range and only one in micromol ar magnitude. Overall, most of the Ki values of Pb PM4 were comparable with their plasme psin 4 orthologs from human malaria parasites, indicating that they shared simila r active site structural features. Additionally, compound 3 selectively inhibited Pb PM4 by a factor of more than 70-fold when compared to inhibition of hCatD. This co mpound may serve as a reference for designing tight-binding inhibitors selective to Pb PM4 based on the combinatorial library analyses. Molecular modeling Coincidentally, the tightes t binding inhibitors of Pf PM1 (compound 1) and Pb PM4 (compound 3) are also the most selective ag ainst hCatD. Hence, the binding modes of

PAGE 155

134Table 5-4. The inhibitory activitie s of peptidomimetic inhibitors on Pf PM1 and Pb PM4. Dissociation Constant ( Ki)c Inhibitora Sequenceb Pf PM1 Pb PM4 Pf PM2d Pf PM4d Pv PM4d Po PM4d Pm PM4d hCatDd 1 KPnLSnL LQI 72.8 8.8 nM 375 54 nM 13.9 1.8 nM 21.7 2.7 nM 97 14 nM 187 29 nM 160 26 pM 219 21 nM 2 KPVEF RQT 31 5 M 502 56 nM >20 M 2.4 0.3 nM 14.4 2.1 nM 39.0 4.5 nM 10.3 1.3 nM 30.4 2.0 nM 3 KPLEF YRV 38 6 M 120 31 pM 19.5 4.0 M 476 87 pM 684 87 pM 3.2 0.5 nM 342 47 pM 8.5 0.6 nM 4 KPLEF FRV 5.5 0.7 M 1.4 0.1 nM 4.3 0.8 M 85 14 pM 582 84 pM 3.2 0.5 nM 3.7 0.6 nM 4.7 0.4 nM 5 KPFEL AWT 42 7 M 8.1 0.8 M 16.6 3.1 M 12.7 1.6 M 9.8 1.8 M > 20 M 9.0 1.2 M 12.7 1.2 M 6 KPIEF FAL 8.3 1.2 M 2.7 0.2 nM 2.3 0.3 M 209 29 pM 114 21 pM 251 48 pM 138 26 pM 1.9 0.4 nM 7 KPFnLF FSR 2.8 0.3 M 164 35 pM 75 10 pM 244 41 pM 208 34 pM 600 110 pM 2.2 0.2 nM 1.4 0.2 nM 8 KPIIF FRL 1.7 0.2 M 1.2 0.1 nM 117 14 nM 530 70 pM 1.2 0.2 nM 5.5 0.5 nM 5.8 0.9 nM 2.1 0.7 nM 9 KPInLF FRL 318 42 nM 104 19 pM 68 7 pM 193 39 pM 83 16 pM 410 70 pM 63 10 pM 500 99 pM a Inhibitors 1-5 were develope d from combinatorial chemistry based peptide lib rary screening studies (Beyer et al. 2005) and inhibitors 6-9 were derived from kinetic analyses of single peptide substrate series (Li et al. 2004; Wes tling et al. 1999; Wes tling et al. 1997). b = -CH2-NH-. c Dissociation constants in the picomolar, nanomolar and microm olar ranges are highlighted in red, blue and black colors, respectively. d Dissociation constants are from Beyer 2003

PAGE 156

135 compounds 1 and 3 to their co rresponding enzymes were explored (Figure 5-14). Both compounds were fitted into the active site clefts. The -strand conformation was adopted with each residue occupying a single subsite except for P5-Lys, which was exposed to the solvent. As for compound 1, the unbranched li pophilic side chains of the two norleucine residues were projected into the S3 and S1 pockets of Pf PM1. Their relatively small sizes might cause insufficient interactions with residues deep in the subsites. This could partially explain their not-qui te-strong binding affinity. The Pf PM1 CCI has larger hydrophobic residues at P3 (Phe) and P1 (Leu) as well as at P3’ (Phe). Such alterations may further contribute to enzyme-inhibitor in teraction and improve the binding affinity. On the other hand, compound 3 exhibited better accommodation in the Pb PM4 model structure as P1-Phe, P1’-Tyr and P2’Arg shared proper hydrophobic or hydrogenbonding interactions with enzyme residues. As a result, Pb PM4 was inhibited in the subnanomolar range. The Pb PM4 CCI incorporates a bulkier P3-Tyr and P3’-Phe and substitutes P1’-Tyr with Arg. Such adjustme nts may further enhan ce the solubility and Pb PM4 binding affinity. Discussion The substrate specificities of Pf PM1 and Pb PM4 were investigated using the combinatorial approach. These results were subsequently compared with the subsite preferences of ot her plasmepsins ( Pf PM2, Pf PM4, Pv PM4, Po PM4 and Pm PM4) and human aspartic proteinase homologues (human pepsin A, human cathepsin D and E) from previous studies using the same libraries (Beyer et al. 2005). The primary specificities of the two ne wly characterized enzymes reveal high consistency with the other enzyme homologue s. In the S1 subsites, nine of the ten enzymes accept phenylalanine as the best a nd the S1’ subsites also favor bulky

PAGE 157

136 Figure 5-14. Molecu lar modeling of the Pf PM1-compound 1 complex (A) and the Pb PM4-compound 3 complex (B). Compounds 1 and 3 were fitted into the active site cleft of Pf PM1 and Pb PM4 correspondingly. The surfaces of the active site clefts were generated base d on the electrostatic potential. Hydrogen bonding interactions were highlighted in cyan dashed lines. The active site pockets and the enzyme residues that maintain hydrogen bonding interactions with the inhibitors were labeled.

PAGE 158

137 hydrophobic side chains, though the optimal subs titutions are shared by five distinct residues (Phe, Leu, Tyr, Ile and Nle). This is not surprising since residues comprising these two sites are highly conserved amongst these prot einases (Table 5-5). For Pf PM1 and Pb PM4, the strong hydrophobicity of the P1 position is of particular importance, which has been implied by the N-terminal sequencing analyses of the proenzyme selfprocessing events in Chapter 3 and 4. For the convenience of comparison of the s econdary specificities, the results from digestion of the P1-Phe and P1’-Phe libr aries were employed for these homologous proteinases. The hydrophobicity preferences are well maintained for the S3 and S3’ subsites among the ten aspartic proteinases investig ated. In the S3 subsite, the best three substitutions for all tested enzymes are limited within the three aliphatic residues, Ile, Leu and Nle and the three aromatic ones, Tyr, Ph e and Trp. Similar outcome is revealed for the S3’ pocket. However, residues that are co mposed of these S3 subsites are not quite conserved as expected (Table 5-5). Although hydrophobic residues account for the majority, their identities are fairly divergent except for the highly consistent Phe117, which indicates that this residue may play a critical role in interacting with the P3 amino acids . Pb PM4 has the most significant differences on the S3 residues from other enzymes in that three serines are involved in S3 pocket construction. In this case, the hydroxyl groups of these serines may project off the pocket to maintain hydrophobicity. As for the S3’ subsite, plasmepsin 4 orthologs share id entical composing residues, which are quite distinct from the three human as partic proteinases as well as Pf PM1 and Pf PM2 (Table 55). Despite the different composition, all th e ten enzymes share a similar specificity

PAGE 159

138 profile with the best substitution being Ile. Notably, P3-Tyr and P3-Phe are comparably favored in all tested enzymes except for Pb PM4, where P3-Phe (99%) is overwhelmingly preferred versus P3-Tyr (3%). This may be because the spatially approached amino acids of Pb PM4 force the S3’ composing residues to adopt conformations that favor accommodation of a non-hydroxyl benzyl ring. Unlike the stringent specificities exhibited in the S3 and S3’ subsites, the substrate specificities at S2 and S2’ were extended so that polar residues such as serine and glutamine are well accepted in both human en zymes and plasmepsins. Residues that are composed of the S2 subsite are quite cons erved among these enzymes, especially among plasmepsin 4 orthologs (Table 5-5). As a result, the best three amino acid substitutions for plasmepsin 4 enzymes are consistently Gl u, Ile and Ser, and P2-Ile and P2-Ser are the most favorites for every tested enzyme. On the other hand, S2’ composing residues share relatively poor similarities between varied en zymes, especially for residue 74 (Table 5-5). Results reveal that human aspartic pr oteinases prefer accepting hydrophobic residues, whereas glutamine is also greatly accepted by most of plasmepsins as well. These findings indicate the direct infl uences of diverse subsite resi dues on substrate preferences and a possibly major contribution of the S2’ subsite on overall enzymatic specificity. Due to the difficulties of production of active recombinant enzyme, subsite specificities of Pf PM1 and Pb PM4 have not been well studie d previously. Gluzman et al. investigated Pf PM1 digestion on its natural substr ate hemoglobin and identified six cleavage sites, three on and chain each ( 3334, RMF-LSF; 4647, PHF-DLS; 9899, VNF-KLL; 3132, GRL-LVV; 41-42, QRF-FES; and 129-130, QAAYQK; represents the cleavage site) (Gluzman et al. 1994). A ll six sites are specific for

PAGE 160

139 Pf PM1 digestion except for 3334. Other than the unique hydrophobic features of P1 residues, amino acids with varied properti es emerge across the S3-S3’ subsite, which does not agree well with the results from the combinatorial chemistry approach. This fact not only confirms the essential role th e hydrophobic P1 residue plays on substrate recognition and processing by Pf PM1 but also indicates the binding specificity of a peptide is determined by the en tire sequence that spans the active site cleft rather than single isolated residues. Tyas et al . studied the specificity constants ( kcat/ Km) of a series of chromogenic peptide substrates for Pf PM1, with focus on the P3, P2 and P2’ positions (Tyas et al. 1999). The peptide s ubstrate K-E-F-V-F*Z-A-L-K (Z = para nitrophenylalanine, and * repres ents the cleavage bond) was used as a reference. Both the P3F to P3L and P2V to P2N alterations decrea se the specificity constants more than 2fold and the P2’A to P2’R alteration maintains a similar kcat/ Km as the reference. These findings are generally consiste nt with the results from combinatorial library studies. Siripurkpong et al. employed a random peptide to analyze the subsite preferences at the prime side of Pf PM1 (Siripurkpong et al. 2002).The resu lts reveal that the S1’ subsite prefers hydrophobic residues and the S2’ subsite switches its favorites to polar residues, which generally agree with the results from the combinatorial libraries. Controversially, serine was also considered as a favorite P1 substitution in that study, which may be due to different peptide constructi ons and analysis approaches. Peptidomimetic compounds with high bindi ng affinity and/or selectivity to some other food vacuole plasmepsins have been prev iously developed (Beyer et al. 2005, Li et al. 2004; Westling et al. 1997) Being homologues, Pf PM1 and Pb PM4 should have comparable performance on interacting with th ese inhibitors as othe r plasmepsins, which

PAGE 161

140 has been proved by kinetic analysis. The most specific tight-binding inhibitors of Pf PM1 and Pb PM4 chosen from the experimental data are subsequently compared with the ones developed from combinatorial ch emistry. The two compounds for Pf PM1 are closely related (KPnLSnL LQI vs. KPFSL LQF), whereas the two for Pb PM4 are quite different (KPLEF YRV vs. KPFEF RQF). Hence, it will be very interesting to know if inhibitors developed from the combinatorial libraries could further improve the binding affinity and selectivity for Pf PM1 and Pb PM4, which will rely on inhibition analysis of such compounds on the plasmepsins and homologous human proteinases. Using the combinatorial chemistry based pe ptide libraries, the subsite specificities of two essential plasmepsins, Pf PM1 and Pb PM4, were systematically investigated for the first time. Results from this study in conjunction with the discussions on enzyme preparation and primary characterization in Chapter 3 and 4 broaden our knowledge on kinetic and structural features of these two aspartic proteinases. In addition, selective peptidomimetic inhibitors of these two enzymes have been proposed, as could be considered potential lead compounds fo r designing drugs of high potency and bioavailability and low toxicity. Conclusion The S3-S3’ subsite preferences of two malaria aspartic proteinases Pf PM1 and Pb PM4 were discovered using two co mbinatorial peptide libraries. Pf PM1 and Pb PM4 showed similar and yet unique substrate spec ificities. Overall, the S3, S1, S1’ and S3’ subsites favored accommodation of bulky hydrophobic amino acids; whereas the preferences for the S2 and S2’ pockets were extended as polar residues such as serine, glutamine and glutamic acid could also be well accepted. The optimal peptide sequences

PAGE 162

141 Table 5-5. Amino acid residues that comprise the S3-S3’ subsite pockets of human and malaria aspartic proteinase series. S3 hPepA hCatD hCatE Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 9 Y Y Y V F V V I V S 12 M A M V I L I L L L 13 E Q E M M M M V M S 111 F T T A T I I V L I 115 A A A G S S V I A S 117 F F F F F F F F F F S2 hPepA hCatD hCatE Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 76 G G G V V G G G G G 219 S S S S S S T S S S 222 T V T T T T T T T T 287 Q M Q I I L L I L I 289 M M L V L V V V V V S1 hPepA hCatD hCatE Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 30 V V I I I I I L I I 32 D D D D D D D D D D 34 G G G G G G G G G G 75 Y Y Y Y Y Y Y Y Y Y 76 G G G V G G G G G G 77 T S T S S S S S S S S1’ hPepA hCatD hCatE Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 189 Y Y Y Y Y Y Y Y F Y 213 I I I I I V I I I I 215 D D D D D D D D D D 217 G G G G G G G G G G 218 T T T T T T T T T T 300 I I I I I I I I I I S2’ hPepA hCatD hCatE Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 73 I I I M M I I L I I 74 T H Q N N S T L T L 128 I I L L L L L L L L 130 S V V I I I I I I V S3’ hPepA hCatD hCatE Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 76 G G G V V G G G G G 289 M M L V L V V V V V 291 L I I L F I I I I I 292 P P H N P D D D D D Amino Acid residues are numbered based on the sequence of human pepsin A. Residues that comprise subsites of Pf PM1 and Pb PM4 are highlighted in red.

PAGE 163

142 for Pf PM1 and Pb PM4 were K-P-F-S-F-L-Q-F and K-P-F-E-F-nL-S-W, respectively. The substrate specificities of these two en zymes were compared with that of human cathepsin D, the most homologous human aspa rtic proteinase to plasmepsins, and reduced peptidomimetic inhibitors that were selective to Pf PM1 and Pb PM4 were proposed. These compounds were K-P-F-S-L L-Q-F and K-P-Y-E-F R-Q-F for Pf PM1 and Pb PM4, respectively. In addition, the inhibition effects of these two plasmepsins on a series of peptidomimetic compounds specific to human food vacuole plasmepsins were studied. Similar to its plasmepsin 4 orthologs, Pb PM4 was mostly inhibited within nanomolar magnitude; whereas Pf PM1 shared the inhibition profile with Pf PM2 maintaining most of the Ki values in the micromolar range. From the inhibition analysis, the best inhibitors of Pf PM1 and Pb PM4 were selected and their potential binding modes were illustrated. This study furt her complements active site characteristics of plasmepsins and benefits designi ng of effective antimalarial drugs.

PAGE 164

143 CHAPTER 6 HIV-1 CLINICAL PROTEASE INHI BITOR–INHIBITION ANALYSES ON PLASMEPSINS AND ANTIPARASITIC ACTIVITIES ON Plasmodium falciparum CULTURE Introduction Malaria and HIV-1/AIDS are two major endemic infectious diseases haunting public health worldwide. Currently, malaria afflicts approximately 40% of the world population and causes up to 2.7 million deaths annually with over 90% of the victims dwelling in sub-Saharan Africa (Breman 2001; Greenwood and Mutabingwa 2002). HIV1/AIDS affects nearly 50 million individuals gl obally and half of the infected populations are residents of the sub-Saharan Africa re gion (UNAIDS 2006). Therefore, coinfections of malaria and HIV-1/AIDS in this area are significantly prevalent. Highly specific and efficient drug compounds against potential targets need to be developed for battling against such infectious diseases. As a successful precedent in the antiinfective campaign, HIV-1 protease inhibitors (PI) have been widely used in part of the clinical treatment. So far, seven HIV-1 PIs have been approved by FDA as clinical antiretroviral agents and recently anothe r two novel compounds, tipranavir (Best and Haubrich 2006) and darunavir (FDA 2006, de Meye r et al. 2005), has joined this group. These compounds are specific to the HIV-1 prot ease and are tight bi nding inhibitors in the subnanomolar range (Clemente et al. 2004; Muzammil et al. 2003). On the other hand, actions must be taken to overcome the increasingly developed resistance of malaria parasite to traditional drugs over the past two and a half d ecades. Plasmepsins, a group of homologous aspartic proteinases of malaria pa rasites, have been considered potential

PAGE 165

144 targets for novel antimalarial drug design (B ailly et al. 1992; Moon et al. 1997; Silva et al. 1996). Despite belonging to different aspartic proteinase subfamilies, plasmepsins (A1 family) and the HIV-1 protease (A2 family) sh are some sequence identity and tertiary structure similarity (Pechik et al. 1989). Several groups recen tly have reported that HIV-1 PIs, such as ritonavir, saquinavi r and lopinavir, showed antimala rial activities at clinically relevant concentrations against in vitro cultured P. falciparum in the intraerythrocytic stage (Andrews et al. 2006; Pa rikh et al. 2005; Savarino et al . 2005; Skinner-Adams et al. 2004). In addition, previous studies indicated that inhibitors of food vacuole (FV) plasmepsins, such as pepstatin A, SC-50083 and Ro40-4388 coul d kill the cultured P. falciparum by blocking hemoglobin digestion (Baill y et al. 1992; Francis et al. 1994; Moon et al. 1997), a critical ev ent for the parasite initiate d by plasmepsins. Being the unique group of aspartic protei nases identified so far from the malarial parasite genome (Coombs et al. 2001; Gardner et al. 2002a, b), plasmepsins, especially FV plasmepsins, are highly suspected as the inhibitory target s of HIV-1 PIs. To te st this hypothesis, the inhibition constants ( Ki) of seven clinical HIV-1 PIs ag ainst seven FV plasmepsins from human and rodent malaria parasites were de termined, the distinct binding modes of amprenavir and ritonavir on Pf PM4 were illustrated and th e effects of PIs on blocking growth of a chloroquine-sensitive P. falciparum strain at the bloodstage were studied. Results Seven HIV-1 PIs, atazanavir (ATV), ampr enavir (APV), indinavir (IDV), lopinavir (LPV), nelfinavir (NFV), r itonavir (RTV) and saquinavir (SQV), have been approved by FDA as clinical antiretrovira ls (Figure 6-1). The inhibiti on effects of these compounds

PAGE 166

145 Figure 6-1. Structures of FDA-approved clinical HIV-1 PI s. Structures were drawn usi ng ChemDraw Std 8.0 (Cambridgesoft).

PAGE 167

146 were tested on seven recombinant plasmepsin s. Six are from human malaria parasites: plasmepsin 1, 2 and 4 ( Pf PM1, Pf PM2 and Pf PM4) from P. falciparum , and plasmepsin 4 enzymes ( Pv PM4, Po PM4 and Pm PM4) from P. vivax , P. ovalae and P. malariae , respectively; and an additional plasmepsin 4 ( Pb PM4) is from the rodent malarial parasite P. berghei . Kinetic Analyses Shown in Table 6-1 are the dissociation c onstants of protease inhibitors binding to plasmepsins. Not surprisingly, these inhibitors show much weaker binding to plasmepsin with the average dissociation constant in th e high nanomolar to low micromolar range compared with their low nanomolar to subnanomoar inhibition against the HIV-1 protease (Clemente et al. 2004; Roberts et al. 1990). Among the seven inhibitors, RTV showed overall the strongest bind ing affinities with the lowest Ki value (18 nM) for Po PM4 and the highest (245 nM) for Pf PM2. LPV and SQV were also reasonably strong inhibitors of the tested plasmepsins, with mo st of the binding affi nities in the nanomolar range. On the other hand, APV exhibited ove rall the weakest inhibition against the plasmepsins studied with all dissociation cons tants in micromolar magnitude. In addition, most of the Ki values for ATV and IDV were also in the micromolar range. Comparison of binding affinities of the PI se ries within each plasmepsin indicated that the Ki values varied from 60-fold for Pv PM4 up to more than 1300-fold for Po PM4. Such dramatic deviations reflect distinct stru ctural features of the active site clefts among plasmepsins despite their high sequence homology. In addition, Pf PM2, with a 72-fold dissociation constant range, was overall th e most weakly inhibited enzyme, as the Ki values were almost exclusiv ely in the micromolar range.

PAGE 168

147Table 6-1. Inhibition constants of clinical HIV-1 PIs on plasmepsins Inhibition constant ( Ki) Protease Inhibitor Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 amprenavir 20 4 M 6.2 0.7 M6.6 1.1 M1.7 0.2 M 12 1 M1.5 0.2 M12 1 M atazanavir 447 44 nM 5.1 0.8 M16 3 M3.7 0.6 M 3.5 0.5 M4.9 0.6 M7.1 0.7 M indinavir 899 109 nM 18 3 M 486 87 nM 1.1 0.1 M 24 3 M1.7 0.2 M7.0 1.8 M lopinavir 291 29 nM 1.6 0.2 M1.3 0.3 M 508 63 nM 378 47 nM678 61 nM806 76 nM nelfinavir 1.3 0.1 M 2.9 0.3 M 481 65 nM457 55 nM 2.9 0.3 M 762 117 nM527 51 nM ritonavir 100 8 nM 245 25 nM56 8 nM62 10 nM 18 3 nM23 2 nM88 8 nM saquinavir 534 68 nM 2.3 0.3 M 283 46 nM715 97 nM 791 98 nM343 32 nM87 8 nM Micromolar inhibition was highlight ed in blue and nanomolar inhi bition was highlighted in red.

PAGE 169

148 Molecular Modeling Kinetic analyses indicated that APV wa s the weakest PI bound to plasmepsins, while RTV was the strongest. Molecular mode ling was subsequently utilized to propose the binding modes of the two inhibitors. Pf PM4 was utilized as th e enzyme model due to its high sequence homology with the other tested plasmepsins and the availability of the crystallographic structure (1LS5, 2.80 ). Figur e 6-2 shows the superposition of crystal structures of HIV-1 proteas e-PI complexes with the Pf PM4-pepstatin A complex based on their conserved catalytic motifs (~As p32-Thr33-Gly34~ and ~Asp215-Ser/Thr216Gly217~, pepsin numbering). APV and RTV were then docked into the active site of Pf PM4 and energy-minimized. As for the Pf PM4-APV complex, APV was extracted from the energy minimized model structure and overlapped with pepstatin A (Figure 6-3 A). Notably, the coordinates of the hydroxyl group of APV were different from those of pepstatin A in that the hydrogen-bonding interactions between this tr ansition state mimic and OD1 and OD2 of the catalytic dyad were lost (Figure 6-3 C) . The hydroxyl group shifted farther off the carboxyl oxygens of Asp32 and Asp215 (pepsi n numbering), beyond the distances of hydrogen-bonding interactions. A total of nine residues from Pf PM4 were identified to have hydrophobic interactions with moieties of APV (Figur e 6-3 C, D and Table 6-2). These residues stabilize the four hydrophobic side chains of APV that were accommodated in the S2-S2’ subsites. It was worth mentioning that the 4aminobenzenyl group in the S2 subsite only maintained weak hydrophobic interactions w ith Thr217. This increased binding energy and potentially attenuated its binding affinity to Pf PM4. In addition, four hydrogenbonding interactions were assigned between APV and Pf PM4 (Figure 6-3 C). Two

PAGE 170

149 residues of Pf PM4, Gly36 and Gly216, are located near the catalytic sites and the other two, Ser76 and Gly78, reside in the flap region. The hydrogen-bonding interactions occurred between those residues and the amino and carbonyl group of the carbamate moiety as well as the sulfonamide group. Thes e interactions further stabilize binding of the main chain of APV. RTV adopts two distinct orientat ions to fit in the active si te cleft of HIV-1 protease due to the symmetrical struct ure of both the ligand and the protein. For this reason, both orientations of RT V were docked into Pf PM4 (Figure 6-4 B and F). As for the first orientation, in order to ove rlap with pepstatin A, RTV was initially translated along the backbone of pepstatin A to find the starting conformation to dock into PfPM4. The final minimized result reveal ed that the hydroxyl groups from pepstatin A and RTV were 2.8 apart (Figure 6-4 A). The energy minimized Pf PM4-RTV model indicated that the hydroxyl group formed one hydrogen-bondi ng interaction (3.1 ) with OD1 of Asp214; while the distances to the carboxyl oxygens of Asp34 were more than 4.0 , beyond the range of hydrogen-bonding interactions. Such a binding orientation of RTV al lowed development of eight hydrogenbonding connections with seven Pf PM4 residues (Figure 6-4 C), six of which were distributed either in the flap region (Ser76, Gly78 and Ser79) or in the catalytic site (Gly36, Asp214 and Gly216). Additionally, six of the eight hydrogen-bonds benefited locking the main chain conformation of RTV. Another tw elve residues of Pf PM4 were involved in forming hydrophobic inte ractions with RTV, including all nine residues that interacted with APV (Figure 6-4 C, D and Table 6-2). These residues contributed to stabilizing hydrophobic side chai ns of RTV in the S3-S2’ subsites. The three additional

PAGE 171

150 residues, Met15, Ile32 and Phe 120, created interactions with the benzyl and 2-isopropyl4-thiazolylmethylen e group in the S1 and S3 pocket. The binding mode of RTV in the reverse or ientation was also st udied. As shown in Figure 6-4 G, the hydroxyl group did form a hydrogen bond of 3.0 with the side chain of Asp214 but no interactions with Asp34, whic h was similar as what was shown in the first orientation. Seven residues were relate d to hydrogen-bonding interactions with the reversed inhibitor and five of them were conserved in both RTV models. In addition, five of the seven hydrogen-bonding interactions i nvolved the RTV main chain, a correlative feature with the APV and the other RTV model. Despite th ese similarities, only seven residues were identified to interact with th e hydrophobic side chains of RTV. In the S2 subsite, the 5-thiazolylmethyl ene group only shared weakly hydrophobic interaction with Thr217, similar to the case of the 4aminophenyl group of APV. Another hydrophobic group, 2-isopropyl-4-thiazolylmethylene, inte racted with Ile133 on one side but was solvent-exposed on the other side. In vitro Antiparasitic Activity The seven HIV-1 protease inhibito rs were subsequently fed to P. falciparum cultures to investigate their activities on blocking growth and development of the parasite. These experiments were done by Dr . Jorge A. Bonilla of Prof. John B. Dame group, University of Florida, Gainesville, Florida. The antipar asitic activity was represented as the concentration (IC50) of inhibiting fifty percent of uptaking 3Hhypoxanthine by the chloroquine-sensitive 3D7 strain in the bloodstage. The concentration of DMSO was maintained at 0.1% in all tests and did not inhibit the growth of parasite culture. The results (Table 6-3) indicated that lopinavi r bore the strongest antiPlasmodium activity among these protease i nhibitors with the average IC50 value of

PAGE 172

151 Figure 6-2. Superimposition of crystallogra phic structures of HIV-1 protease and Pf PM4. Overlapping is based on the two catalytic motifs (~Asp32-Thr33Gly34~ and ~Asp215-Ser/Thr216-Gly217~, pepsin numbering) (red and blue for the HIV-1 protease and Pf PM4, respectively) usi ng the program Sybyl7.1 (Tripos Inc.). (A) Overlapping of crysta l structure of HIV-1 protease (red) in complex with APV (red) (1T7J, 2.20 , King et al. 2005, Surleraux et al. 2005) with that of Pf PM4 (green) in complex with pepstatin A (green) (1LS5, 2.80 , Asojo et al. 2002b); (B) Overla pping of crystal structure of HIV-1 protease (blue) in complex with RTV (blue) (1RL8, 2.00 , Rezacova et al. 2005) with that of Pf PM4 (green) in complex with pepstatin A (green). The two catalytic aspartates are also shown.

PAGE 173

152 Figure 6-3. The Pf PM4-APV molecular model. (A) Energy-minimized APV (cyan for carbon atoms) is superimposed on pepstatin A (orange for carbon atoms). The hydroxyl groups of the ligands are identif ied with arrows. (B) The tertiary structure of Pf PM4 in complex with APV. Pf PM4 are colored by secondary structure (red for -helices, yellow for -strands and green for random coils). The two catalytic aspartates and APV are highlighted in stick mode. (C) Hydrophobic and hydrogen-bonding in teractions of APV with Pf PM4. APV and its hydrogen-bonding related residue s are shown in stick mode. Hydrogen bonds are presented by dashes. Residues involved in hydrophobic interactions with APV are highlighted in red. The tw o catalytic aspartates are highlighted in blue. All residues are labeled. (D) AP V fitted to the active site pockets of Pf PM4. APV is modeled in stick mode. Connolly surface is generated based on electrostatic potent ial. The flap and 290’s loop region as well as the active site pockets are labeled. Picture A is prepared using Sybyl7.1 (Tripos Inc.) and picture B, C and D are prepared using PyMOL (DeLano Scientific LLC).

PAGE 174

153 Figure 6-3. Continued

PAGE 175

154 Figure 6-4. The Pf PM4-RTV molecular model. (A) Superposition of energy-minimized RTV on pepstatin A; (B) The tertiary structure of Pf PM4 in complex with RTV; (C) Hydrophobic and hydrogen-bondi ng interactions of RTV with Pf PM4; (D) RTV fitted to the active site pockets of Pf PM4. (E)-(H) repeat (A)-(D) with RTV reversely oriented. All the descriptions to the models are same as Figure 6-3.

PAGE 176

155 Figure 6-4. Continued

PAGE 177

156 Figure 6-4. Continued

PAGE 178

157 Figure 6-4. Continued

PAGE 179

158 Table 6-2. Amino acid residues of Pf PM4 involving in hydrophobi c interactions with APV and RTV. PI Subsite Functional Group Residues APV S2 S1/S3 S1’ S2’ 4-aminophenyl N-isobutyl benzyl tetrahydro-3-furyl Thr217 Tyr77, Ile123 Tyr192, Ile294, Ile300 Ser37, Ile75, Tyr77, Leu131 RTV S3 S2 S1 S1’ S2’ 2-isopropyl-4thiazolylmethylene isopropyl benzyl benzyl 5-thiazolylmethylene Met15, Ile32, Phe120 Thr217, Ile300 Ile32, Tyr77, Ile123, Phe120 Tyr192, Ile294, Ile300 Ser37, Ile75, Tyr77, Ile123, Leu131 RTV reverse orientation S2 S1/S3 S1’ S2’ 5-thiazolylmethylene benzyl benzyl isopropyl Thr217 Tyr77, Ile114, Ile123 Ile294, Ile300 Tyr77, Ile133

PAGE 180

159 3.2 M, much lower than the concentration it can accumulate in serum; while amprenavir barely exhibited inhibition on parasite growth within the range of drug concentration tested (up to 50 M). The other five protease inhibitors showed comparable antiparasitic activities. The IC50 values of ritonavir and indinavir were close to th eir clinically relevant drug concentrations in the bl ood stream (Table 6-3). These results suggested that LPV, RTV and IDV may play an antimalarial role in P. falciparum -infected human body. Discussion In the last two years, several FDA-a pproved clinical HIV-1 PIs have been discovered to possess antimalarial activi ties against parasite culture within pharmacologically relevant concentrations. Ho wever, the molecular targets of these PIs have not been clearly known yet. One or more of the plasmepsins are likely to be the targets for the PIs’ antiparas itic effects. Such a hypothe sis requires experimental verification. Meanwhile, four Plasmodium spp. have been known to infect human beings, but only P. falciparum can be cultured in vitro . In this case, the antiPlasmodium effects of PIs on the other three species are hard to be assessed directly; while studies on inhibition of plasmepsins on the nonfalciparum species by PIs may provide evidence for their potential blocking effects on these parasi tes. Therefore, the binding affinities of seven PIs on a series of plasmepsins from hu man and rodent malarial parasites have been determined. RTV was the tightest binding inhibitor against the tested plasmepsins, while APV was overall the weakest. Over all plasmeps ins tested, the binding affinity difference between RTV and APV is approximately two or ders of magnitude. The model structures of the Pf PM4-RTV and Pf PM4-APV complexes were crea ted using the strategy of

PAGE 181

160 Table 6-3. The antiparasitic activities (IC50) of clinical HIV-1 prot ease inhibitors on the P. falciparum 3D7 strain. Inhibitors IC50 ( M) Maximum concentration in human serum ( M) amprenavir > 50 15.2a atazanavir 12.4 5.68.7b indinavir 18.8 3.617.2c lopinavir 3.4 0.115.6d nelfinavir 15.6 1.16.0e ritonavir 17.1 1.815.5f saquinavir 15.9 1.45.5g a Goujard et al. 2003 b Bristol-Myers Squibb prescribing information c Boffito et al. 2002 d Abbott prescribing information e Agouron prescribing information f Abbott prescribing information g Roche prescribing information; Veldkamp et al. 2001

PAGE 182

161 manual docking followed by energy relaxation. The binding modes of the two PIs to Pf PM4 were compared and the unique features were addressed. A common feature shared by these model structures was that a majority of hydrogen-bonding interactions were related to the main chai n atoms of the inhibitors while hydrophobic interactions ma inly occur in the active site pocket with the bulky side chain group. As APV was only half th e size of RTV, the hydrogen-bond donors and acceptors of APV were less than those of RTV, which allowed APV to more readily dissociate. In addition, APV ha s four hydrophobic side chains projecting to active site pockets, and yet the benzene ring, the larg est hydrophobic group at the S2 subsite, was barely rigidified in the modeled structur e. On the other hand, RTV has four more hydrogen-bonds to stabilize th e backbone and three more hydrophobic residues to lock five hydrophobic side chain moieties in the activ e sites. Each side chain group interacted with at least three hyd rophobic residues of Pf PM4. This allowed RTV to gain more interactions with the enzyme and become more robustly associated. Because of the asymmetrical structure of the Pf PM4 active site cleft, the binding mode of RTV in the reverse orientation was not expected to be similar as the other orientation. The model structur e did not support RTV adopting the reverse orientation for the decreased hydrophobic interactions and poten tially flexible thia zolylmethylene group at both termini of RTV. Mode ling the reversely oriented AP V has also been attempted but abandoned in that spa tial hindrances occurred betw een the 4-aminophenyl group and protein residues comprising the S1’ and S3’ pockets. Alternatively, an automatic docking strategy was implemented by our collaborator Hugo Gutirrez-de-Tern Ph.D. of Uppsal a University, Uppsala, Sweden. The

PAGE 183

162Table 6-4. Experimental and calculated free energies for the binding of APV and RTV to Pf PM4. ligand-surrounding interactions (kcal/mol) c Compounda Gbind, x-score (kcal/mol) Gbind, exp (kcal/mol) b Gbind, LIE (kcal/mol) vdw l-s p V el l-s p V vdw l-s wV el l-s wV APV -9.8-5.9 0.4-65.6 0.3-65.9 0.4-39.6 0.1-62.7 0.5 APV (reverse orientation) -9.9 -7.1 -3.3 0.6-67.8 0.3-58.0 1.0-39.6 0.1-62.7 0.5 RTV -12.7-7.8 0.8-97.4 0.3-80.0 0.9-56.2 0.5-79.0 1.2 RTV (reverse orientation) -11.7 -10.0 -4.8 0.8-90.6 0.1-75.2 0.9-56.2 0.5-79.0 1.2 The homology model of Pf PM4 (Gutirrez-de-Tern et al. 2006, accepted by Bioc hemistry) was used for the MD simulations a Binding modes of APV and RTV we re discussed in the text b The experimental free energy values were calc ulated from the experimentally determined Ki values using Gbind = RT ln Ki c The calculated average electrostatic el l-s p V and nonpolar vdw l-s p Venergies for ligand-surrounding (l-s) interactions. Th e subscripts p and w denote simulations of the ligand in complex with the protein and free in water, respectively.

PAGE 184

163 conformations of APV and RTV were extracted from the same crystal structures (1T7J and 1RL8) and automatically docked into the active site cleft of a Pf PM4 model structure (Gutirrez-de-Tern et al. 2006, accepted by Biochemistry) using the program GOLD v.3.0.1 (Jones et al. 1997). For APV, two bi nding modes were identified from the docking algorithm—the one described above and the reversely oriented conformation. Molecular dynamics (MD) stimulations we re then applied to each binding mode. Snapshots of the dynamic conformations of ligands were collect ed during the MD simulations. The binding energy for each a dopted conformation were calculated and averaged. The resulting binding energies for di fferent orientations of APV are shown in Table 6-4. Although comparable results were shown by the X-score function, results from the LIE scoring function indicated that the APV binding mode discovered from our manual docking strategy was more stable th an its reverse orientation. Similarly, the binding mode of RTV preferred by manual doc king had a lower bindi ng energy than its reversely oriented conformation, which has been confirmed by the calculation from both score functions. Notably, the binding energies derived from the model structures did not quite accurately match the experimental resu lts, and this may result from the highly flexible ligands as well as the absence of a high resolution X-ray crystal structure of Pf PM4. Nevertheless, consistent results were achieved from different modeling approaches, which strongly supported th e proposed binding modes of both PIs. Data from kinetic analyses indicated strong binding of RT V, LPV and SQV on food vacuole plasmepsins. The P. falciparum culture assays confirmed the potential inhibition activities of LPV, RTV and I DV in the human body. However, among these PIs, the dissociation constants were not generally proportional to the IC50 values. When

PAGE 185

164 fed to cultured parasites, the drugs need to penetrate four membrane layers in order to reach plasmepsins inside the food vacuole (Fra ncis et al. 1997a; Klemba et al. 2004); how well they can execute blocking activities is quite dependent on how efficient they can approach to the targets. The relatively weak antiparasitic activities SQV and RTV exhibited may be due to their poor permea tion abilities, which may be enhanced by covalently linking PIs with cell-penetrati ng peptides such as Tat48-60 and SynB3 (Russelle et al. 2001; Vives et al. 1997). Previously, a related study focused on th e hemoglobin digestion pattern of PI treated parasites. The kinetic data shown here provide direct evidence to support the notion that the HIV-1 PIs block the growth of malaria parasites by inhibition of plasmepsins. Further proof may rely on th e gene knockout and/or RNAi techniques to disrupt the expression of plasmepsin familie s and study antiparasitic activities of PIs under such constructs. And yet due to the f unctional redundancy of plasmepsin family members (Dame et al. 2004), such a goal may only be achieved when the actions of all plasmepsins were blocked, which has been recently indicated (Parikh et al. 2006). Conclusion The direct inhibition effects of seven FDA-approved clinical HIV-1 protease inhibitors on seven food vacuole plasmepsins from four human and one rodent malarial parasites were studied. Ritonavir, lopinavi r and saquinavir showed overall the tightest binding affinities against plasmepsins, wh ereas amprenavir was the weakest binding inhibitor. The distinctive binding mo des of ritonavir and amprenavir on Pf PM4 were illustrated in modeled structures. In vitro parasite culture expe riments indicated the antimalarial activities of lopina vir, ritonavir and indinavir. These studies provided direct

PAGE 186

165 evidence on plasmepsin-mediated malarial para site growth inhibiti on by HIV-1 protease inhibitors.

PAGE 187

166 CHAPTER 7 INHIBITION ANALYSES OF PRIM AQUINE-STATINE “DOUBLE DRUG” COMPOUNDS AGAINST PLASMEPSINS Introduction The rapid development of drug resistance from malaria parasites, especially Plasmodium falciparum , has remarkably prompted the resurgence of this ancient infectious disease in the last two and a ha lf decades. Potential drug targets, such as plasmepsins, have been recognized and studi ed for a new round of antimalarial campaign. As aspartic proteinase ho mologues, plasmepsins are expressed and believed to function in multiple stages of the parasite life cycle (Banerjee et al. 2002; Bozdech et al. 2003). Those that reside in the food vacuol e (FV) initiate hemoglobin degradation, a critical process for normal parasite deve lopment and growth (Banerjee et al. 2002). Inhibitors of FV plasmepsins, such as pepstatin A, Ro40-4388 and SC-50083, block hemoglobin degradation and kill cultured parasites (Bailly et al. 1992; Gluzman et al. 1994; Francis et al. 1994; Moon et al. 1997) , indicating these FV plasmepsins are attractive targets for novel antimalarial drug design. A total of four FV plasmepsins (plasmep sin 1, 2, 4 and HAP) are expressed in the bloodstage of P. falciparum (Banerjee et al. 2002), and onl y one ortholog of plasmepsin 4 for each of the other three human malaria parasites ( P. vivax , P. ovalae and P. malariae ) play a similar role as a hemoglobinase (Dam e et al. 2003; Li et al . 2004; Westling et al. 1997). Although these homologous enzymes share high amino acid sequence identity (Dame et al. 2004), they carry distinctive bi nding specificities, whic h has been revealed

PAGE 188

167 by related X-ray crystal structures and kinetic analyses (Asojo et al. 2002a, b; Asojo et al. 2003; Bernstein et al. 1999; Bernstein et al . 2003; Clemente et al. 2006; Freire et al. 2004; Li et al. 2004; Madabushi et al. 2005; Prade 2005a, b, c; Prade et al. 2005; Silva et al. 1996; Westling et al. 1999; We stling et al. 1997). On the other hand, targeted genetic disruption studies show that the cultured P. falciparum can survive in the case of knocking out one or two of the FV plasmepsin genes (Liu et al. 2005; Omara-Opyene et al. 2004), which indicates that functional re dundancy may exist between the four FV plasmepsins. These findings remind us that ideal antimalarial drugs must be able to effectively block multiple plasmepsins a nd that these compounds must bear high selectivity against human aspartic protei nases. Following this thought, statine based inhibitors have been designed with sta tine, a mechanism-based pharmacophore as the featured binding unit (Ca rroll et al. 1998a, b). Statine, a nonnatural amino acid, mimics the tetrahedral intermediate during peptide bond hydrolysis (Figure 7-1). Most of the statine based inhibitors studied show strong binding affinities to plasmepsins (Banerje e et al .2002; Li et al . 2004; Luker et al. 1996; Siripurkpong et al. 2002; Ty as et al. 1999; Westing J et al. 1997); however, as peptide-derived compounds, they are also ch aracterized with metabolic instability and poor cell permeation ability. Compensation of the drawbacks of such peptidomimetic compounds has been observed in the HIV/AI DS therapy studies where a “double-drug” strategy was introduced to incorporate a nucle oside reverse transcriptase inhibitor (NRTI) with an allophenylnorstatine c ontaining protease inhibitor (P I) (Kimura et al. 1999; Kiso et al. 1999; Matsumoto et al. 2001a, b). The NRTI can facilitate membrane penetration for PIs (Chan et al. 1993; Domin et al. 1993; Hu ang et al. 1994) and parent drugs can be

PAGE 189

168 regenerated from intracellular hydrolysis to ta rget on different entitie s and synergistically contribute to antiretroviral activities. The success of th e “double-drug” design on HIV control inspired us to apply a similar strategy to malaria. In this case, the antimalarial primaquine (PQ) was incorporated as the s econd drug conjugated to the statine-based compounds. Primaquine is the most effective 8-aminoquinoline drug to date in malaria treatment. It acts against P. vivax and P. ovalae of the exoerythrocytic stage and against the gametocytes of P. falciparum ; however, the targets and action mechanism of PQ are still unclear. In addition, PQ has not been wi dely administered because of its toxicity (Carson 1984). Further studies s how that the side effects of PQ are attenuated and the activity is enhanced when it is linked with tripeptides, such as Lys-Leu-D-Val-NH2 and Lys-Leu-D-Ala-NH2, which potentially protect PQ agai nst rapid metabolism (Philip et al. 1988). Therefore, the peptide derivative s of PQ were employed in this study. The scaffold of “double-drugs” was constr ucted by covalently conjugating the PQ peptide prodrugs and the “~Ile-Leucinylstatin e” portion via chemical bridge groups of different size. These compounds were designed and synthesized by Prof. Enrica Bosisio’s group at University of Milan, Milan, Italy and their general structure is shown in Figure 7-2. Kinetic assays were performed to determine the dissociation constants ( Ki) values of these inhibitors against seven FV plas mepsins, i.e., plasmepsin 1, 2 and 4 ( Pf PM1, Pf PM2 and Pf PM4) from P. falciparum ; plasmepsin 4 enzymes ( Pv PM4, Po PM4 and Pm PM4) from P. vivax , P. ovalae and P. malariae ; and a plasmepsin 4 ortholog ( Pb PM4) from the rodent malaria parasite P. berghei . Following that, a “double-drug” inhibitor (compound 2) was docked into the active site cleft of Pf PM4. A possible binding mode of

PAGE 190

169 Figure 7-1. Schematic diagram showing the cata lytic mechanism of aspartic proteinases. Figure is adapted from James et al. 1992, and Veerapandian et al. 1992. Figure 7-2. General structural formula of the primaquine-statine “double-drug” compounds. Figure is adapte d from Romeo et al. 2004.

PAGE 191

170 the PQ-statine compounds was proposed. Data from these studies provide more information on specific subsite features of plasmepsins and on improvement of the binding affinity of newly synthesized compounds. Results Enzyme Inhibition The inhibitory activities of compounds 1-18 on the seven FV aspartic proteinases, Pf PM1, Pf PM2, Pf PM4, Pv PM4, Po PM4, Pm PM4 and Pb PM4 were determined and the resulting Ki values are listed in Table 7-1. Th e PQ-statine “double-drug” compounds were overall strong inhibitors against the tested plasmepsin homologues with most of the dissociation constants in the subnanomola r to nanomolar range. Compounds 1-8 exhibited the tightest binding of subnanomolar to low nanomolar magnitude to all seven enzymes except for Pf PM1 and Pb PM4. The inhibitory activiti es of such compounds on these two enzymes were widely divergent with more than 80-fold and 20-fold deviation, respectively. While their i nhibitory activities against Pf PM1 were fairly strong (except for compound 3), their binding affinities to Pb PM4 were 1-2 orders of magnitude weaker than those to Pf PM2 and the other plasmepsin 4 homologs. Compounds 1-8 share a similar structure scaffold, which is com posed of the PQ-dip eptide portion, the leucinylstatine moiety and a linker group. Diffe rent chemical linkers separate compound 1 from compounds 2-8, which are further di scriminated by the divergent dipeptide sequences conjugated with PQ. On the ot her hand, compound 18 showed the weakest inhibition of high nanomolar to low micromol ar magnitude against the studied enzymes. Surprisingly, it blocked the activity of Pb PM4 fairly well. Lacking the PQ-dipeptide moiety, this compound is half the size of compounds 1-8.

PAGE 192

171 In order to address the cont ributions of different moietie s to the enzyme inhibition effects, these statine derivatives were categorized by structural composition. First, the compounds containing both PQ-peptid e and statine can be subdivided into five unique groups based on the chemical li nkers of different size. Compounds 1, 2, 14, 16 and 17 bear different linkers and ye t share the “PQ-Lys-Leu~” and “~IleLeucinylstatine” moieties. Hen ce, inhibitory acti vities of these five compounds were compared to explore the favored binding feat ures of the bridge groups. As the physical size of chemical linkers sequentially d ecreased for compounds 1, 2, 14, 16 and 17, the Ki values increased for Pf PM4 (up to 240-fold), Pv PM4 (up to 70-fold), Po PM4 (up to 400fold) and Pb PM4 (up to 16-fold). Such a princi ple was also established in the Pf PM1 (up to 9-fold) and Pf PM2 (up to 70-fold) change in inhibi tion, except that the worst inhibitor was compound 16. However, inhibition of Pm PM4 was insensitive to the size of linkers as all five compounds bound in the subnanomol ar to low nanomolar range. Based on these data, it seemed that a relatively longe r and bulkier bridge group could potentially improve the overall binding affinities of the “double-drug” compounds. Secondly, the inhibition effects of PQ-prodrug containing compounds were investigated. The Ki values of compound 13 were co mpared with those of compound 2. The presence of the “PQ-Lys~” moiety in compound 2 consistently enhanced the binding affinities to all seven plasmepsins from 4-fold ( Pm PM4) up to 78-fold ( Pf PM4). A similar outcome was revealed from the ot her two inhibitors, compounds 14 and 15, which bore a different bridge group. In this cas e, compound 15, lacking the “PQ-Lys~” portion, was bound to the enzymes 2( Pm PM4) to 96-fold ( Po PM4) weaker than compound 14.

PAGE 193

172 These findings indicated that th e “PQ-Lys~” moiety improved the in vitro inhibition of the statine-based compounds against FV plasmepsins. Thirdly, using the Ki data from compounds 2 and 13 as references, the contributions of the PQ and attached amino acids to plasme psin inhibition were further explored. First of all, when PQ was directly linked to Le u, the binding affinities of compound 9 to the three plasmepsins from P. falciparum were 5-fold decreased relative to those of compound 13. While compound 9 inhibited Pm PM4 even better than compound 2, the inhibitory activities of compounds 9 and 13 were comparable as to Pv PM4, Po PM4 and Pb PM4. Secondly, in compound 11, a Lys was at tached to Leu, which, compared to compound 13, led to a slight in crease of binding affinities to plasmepsin 4 enzymes, a 15fold improvement to Pf PM1 and a less favored binding to Pf PM2. Appending PQ to compound 11 (i.e., compound 2) rendered a comparable Ki values to Pf PM1 and Pm PM4 but enhanced the binding affinities to all th e other plasmepsins. Thus, inhibition data from compounds 9 and 11 further implied the positive impact of the “PQ-Lys-” moiety on plasmepsin inhibition. Analogous results came from inhibitory studies of compound 12, where the N6t -tert-Butyloxycarbonyl-L-lysine ( t -Boc-Lys) replaces Lys. This suggested that the presence of a t -Boc group at the side chai n terminus of Lys did not remarkably affect compound inhibition. In addition, when PQ was directly linked to the bridge group, compared to compound 13, compound 10 showed a remarkable improvement on binding to Pv PM4, Po PM4 and Pb PM4, a slight 2-fold binding affinity enhancement against Pf PM1, Pf PM2 and Pm PM4 and a similar Ki value for Pf PM4. These findings indicated large hydrophobic gr oups were favored in the P4 position. Also notably, Pm PM4 was well inhibited de spite these variations.

PAGE 194

173 Fourthly, the potential roles of the “~Ly s-Leu~” dipeptide on enzyme inhibition were studied. First of all, th e Lys residue was modified as t -Boc-Lys in compound 3. Compared to compound 2, this decoration main ly resulted in a 12-fold binding affinity decrease to Pf PM1 and Po PM4 while maintaining similar inhibition against the other enzymes, which supported the notion that addition of t -Boc to the side chain of Lys had minor effects on the compounds’ binding abilities. The Lys residue was then substituted by Leu and inhibition of Pf PM2 and Pf PM4 were attenuated by 15and 60-fold, respectively, by the resulting compound 7. Add itionally, an alanine scan was performed in these two positions. A Lys to Ala alterati on in compound 6 did not affect the overall enzyme binding significantly. A Leu to Ala substitution in compound 4 enhanced binding to Pm PM4 13-fold but reduced the affinity to Pf PM4 34-fold. When both residues were replaced by Ala, compound 5 improved binding to Pf PM1 19-fold but increased the Ki values to Pf PM2 and Pb PM4 56and 34-fold, respectivel y. Furthermore, the dipeptide was replaced by the (4-amino) benzoyl group that had a similar backbone size as dipeptide. As a result, compound 8 remarkably enhanced the binding affinity exclusively to Pm PM4 by two orders of magnitude. These find ings indicated that the presence of a dipeptide long linker between PQ and the br idge group generally enhanced the binding abilities to the tested FV plasmepsins. The active site features of different plasmepsins were somehow reflected from their sensitivity to the divergent compositions of peptide linkers. Molecular Modeling The PQ-statine “double drug” compounds are significantly larger than most of investigated inhibitors of aspa rtic proteinases. No ligands of similar structure have been crystallized in complex with aspartic prot einases. The absence of a homologous template

PAGE 195

174Table 7-1. The dissociation constants of PQ-stati ne “double-drug” compounds on plasmepsin inhibition Dissociation Constant ( Ki) (nM) Compound Structure Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 1 9.9 1.8 3 0.2 6 0.1 0.2 0.1 0.4 0.1 0.14 0.03 18 2 2 19.0 3.9 0.4 0.09 0.5 0.1 1.2 0.2 1.6 0.3 1.2 0.1 44 4 3 230 30 3.0 0.1 2.6 0.4 1.3 0.1 19.2 1.4 1.5 0.3 199 29 4 33.1 5.2 1.8 0.4 17.0 3.4 0.28 0.02 1.3 0.4 0.09 0.01 123 20 5 1.01 0.45 22.4 1.3 0.7 0.1 2.3 0.4 7.5 0.4 1.7 0.3 1481 216 6 74.0 10.5 0.5 0.2 0.51 0.06 0.24 0.03 2.5 0.3 0.51 0.10 279 42 O N H N N H H N N H H N N H O O O O O N H 2 OH O N H O O O N H N N H H N N H H N N H O O O O O N H 2 OH O N H O O N H N N H H N N H H N N H O O O O O H N O H O N H O O O O N H N N H H N N H H N N H O O O O O N H 2 O H O N H O O N H N N H H N N H H N N H O O O O O O H O N H O O N H N N H H N N H H N N H O O O O O OH O N H O

PAGE 196

175Table 7-1. Continued Dissociation Constant ( Ki) (nM) Compound Structure Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 7 57.8 4.3 6.2 0.2 30 1 2.5 0.3 5.3 1.3 5.0 0.7 210 29 8 7.6 1.2 13.8 1.2 0.3 0.04 2.0 0.2 3.3 0.5 0.0078 0.0021 179 31 9 765 56 199 29 245 28 15 2 21 2 0.3 0.07 644 33 10 81.2 7.0 8 2 58 5 0.7 0.1 3 0.4 2.2 0.5 159 19 11 10.7 1.3 61 7 8.7 0.6 11 2 17 2 1.2 0.3 243 30 12 16.7 3.2 57 7 19 2 4 1 2 0.4 1.4 0.5 761 45 O N H N N H H N N H H N N H O O O O O OH O N H O O N H N N H H N N H O O O OH O N H O N H O N O H N H N N H O N H H N N H O O O O O OH H N N H O O OH O N H O O N H N N H O O H N N H O N H H N N H O O O O O N H 2 O OH O H N N H O N H H N N H O O O O O H N O OH O O

PAGE 197

176Table 7-1. Continued Dissociation Constant ( Ki) (nM) Compound Structure Pf PM1 Pf PM2 Pf PM4 Pv PM4 Po PM4 Pm PM4 Pb PM4 13 158 20 17 2 39 7 10 2.5 44 4 4.5 0.3 1461 119 14 26.7 2.6 4.1 0.5 21 2 1.5 0.2 1.7 0.3 0.6 0.1 75 10 15 330 46 118 12 257 35 47 8 164 26 1.4 0.1 1737 78 16 91.1 9.3 28 4 29 4 16 1.3 38 5 0.5 0.04 124 13 17 35.7 3.2 6 1 120 16 14 3 161 18 2.1 0.3 285 30 18 727 57 334 38 981 136 17 3 1265 90 19 1.4 144 13 H N N H O O OH O N H O N H O H O O O N H N N H H N N H O O N H 2 O H N O N H H N O OH O H N N H O O OH O N H O O N H N N H O O N H N N H H N N H O O O N H 2 H N N H H N O OH O O O N H N N H H N N H O O O N H 2 O H N N H H N O OH O O N H H N N H O O O H O

PAGE 198

177 complicates proposition of binding modes of su ch compounds to the plasmepsins. On the other hand, the statine portion is usually cons idered as a transition state analogue for the catalysis of aspartic protei nases. Therefore, leucinylstatine of the “double-drug” compound is predicted to be located at S1 and S1’. Based on this notion, compound 2, one of the tightest binding inhibi tors of this series, was doc ked into the active site of Pf PM4 using the crysta l structure of the Pf PM4–pepstatin A complex (1LS5, 2.80 , Asojo et al. 2002b) as the template. A potential binding mode of compound 2 to Pf PM4 is shown in Figure 7-3. The “~Ile-Leucinylstatin e” moiety is accommodated in the S2-S1’ pockets. The bridge group, therea fter, extends toward the nonprime side of the active site cleft and leads the “PQ-Lys~” portion to a pot ential binding site comprising residues in the C-terminal domain. Such extra binding c ontributions from the “PQ-Lys~” moiety may partially explain the st ronger binding affinities for corresponding inhibitors. And a bridge group of an appropriate size, lik e that of compound 2, may enhance compound binding not only by itself but also by accurate ly transferring the “PQ-Lys~” moiety to that binding pocket. Another modeling involve d docking compound 2 into the active site of Pf PM2 using the crystal structure of the Pf PM2–pepstatin A (1M43, 2.40 , Asojo et al. 2003) as the template. A similar re sult was revealed (data not shown). Discussion Kinetic analyses indicated that most of the PQ-statine “double-drug” compounds were high affinity inhibitors against plasme psins. These compounds obviously need to be further tested against human aspartic protei nase homologues, especially human cathepsin D (hCatD), to assess their inhibition selectiv ity. Primary studies on this aspect are being carried out by our collaborators of University of Milan, Milan, Italy. Results showed that compound 6, 9 and 12 inhibited hCatD two orde rs of magnitude weaker than they

PAGE 199

178 Figure 7-3. Structure of the modeled Pf PM4-compound 2 complex. The cyan ribbon structure represents the Pf PM4 molecule. Compound 8 is highlighted as balland-stick. The Connolly surface is colore d based on the electrostatic potential of the enzyme active site. The binding poc kets in the active site (S3-S1’) are labeled. While the leucinylstatine moiety resides in the S1-S1’ subsite, the “PQ-Lys~” moiety extends outside the act ive site to potentially interact with residues of a C-terminal loop region.

PAGE 200

179 inhibited PfPM2 in terms of 50% inhibiti on concentration (IC50). These promising results must be further confirmed with other plasmepsin enzymes. Another intriguing question is how these co mpounds affect the growth of cultured parasites. A series of compounds have been tested, among which compounds 1 and 2 showed low micromolar IC50 values on both chloroquine-sensitive (D10) and chloroquine-resistan t (W2) strains of P. falciparum (Romeo et al. 2004). Meanwhile, Mr. Carlos R. Sulsona of Prof. John B. Dame group, University of Florida, Gainesville, Florida, tested the antiparas itic activities of compounds 2, 4 and 6 on the chloroquinesensitive (3D7) st rain. All three co mpounds showed antiPlasmodium effects at the concentration of 5 M. These pieces of evidence sugge sted the PQ-statine “double-drug” compounds bore antiparasitic activities. Furt her, Romeo et al. f ound that, without the “PQ-Lys~” moiety, the IC50 values of Pf PM2 and P. falciparum inhibition decreased by 200and 40-fold, respectively; whereas th e primaquine group itself did not have significant effects to plasmepsin and parasite inhibition (Romeo et al. 2004). Therefore, it seems that the “PQ-Lys~” moiety not only faci litates enzyme inhibition directly but also helps drug delivery to kill the parasite, t hough detailed mechanisms for both aspects are not clear yet. Conclusion The inhibition effects of a group of pr imaquine-statine “double-drug” compounds were kinetically analyzed agai nst seven plasmepsin homologues ( Pf PM1, Pf PM2, Pf PM4, Pv PM4, Po PM4, Pm PM4 and Pb PM4). Most of these compounds were nanomolar inhibitors of the tested plasmeps ins. The presence of the primaquine-lysine moiety consistently enhances binding affin ity of the compounds. The chemical group that covalently linked primaquine and the statin e portion also remarkably affected enzyme

PAGE 201

180 inhibition. A relatively long and bulky linker was preferred. A potential binding mode of a “double-drug” compound to plasmepsin was pr oposed: a long bridge group allowed the associated primaquine-lysine portion to reac h a binding pocket outside the active site; and the extra binding energies from the primaquine -lysine moiety as well as the contributions from the interactions with a larger chemical linker significantly improved inhibition against plasmepsins.

PAGE 202

181 CHAPTER 8 FUTURE DIRECTIONS Despite that the plasmepsin homologues share high amino acid sequence identity and tertiary structure similarity, the refoldi ng features they exhibi ted are quite distinct. Under the same refolding conditions, the recombinant pro Pf PM2 and plasmepsin 4 zymogens of human malaria parasites mainta in solubility during renaturation with the initial refolding protei n concentration of 0.15 mg/mL (determined by OD280) and more than 90% of the zymogens fold uniformly to the expected size of approximately 43 kDa. Although the recombinant pro Pf PM1 follows the latter trend, the severe precipitation during its refolding is hard to attenuate ev en in the presence of different chemical additives. On the other hand, pro Pb PM4 maintains stability in the folding milieu, but the resulting materials differ in size with only about 10% highly active and matching the corresponding size. As a result, the am ounts of recombinant activatable pro Pf PM1 and pro Pb PM4 obtained from heterologous E. coli cell expression are adequate for enzymatic characterization on the kinetic analysis level, but barely meet the requirement for X-ray crystallography-based structural studies. Impr ovement of the refolding properties can be carried out by engineering pr otein molecules and/or by alteration or modification of the folding methods. As for the former, one of th e strategies is to create and recombine plasmepsin mutants by directed evolu tion (Arnold 1996; Stemmer 1994) followed by genetic screening of mutant li braries using a GFP split system (Cabantous et al. 2005) to select variants with higher solubility.

PAGE 203

182 Directed evolution is a labor atory process used on isol ated molecules or microbes to cause mutations and identify subsequent ad aptations to novel environments. One of the powerful forms of direct e volution is DNA shuffling (Ste mmer 1994). As a method of in vitro homologous recombination, DNA shuffling is carried out by PCR-mediated random reassembly of DNA fragments of muta nt genes via the homology-based priming property. Due to its poor solubility during the refolding process, the recombinant pro Pf PM1 will be chosen as the experimental subject. A modeled structure of pro Pf PM1 has been created using SWISS-MODEL by Amrita Mada bushi Ph.D. of Prof. Robert McKenna’s laboratory, University of Florida, College of Medicine, Gainesvill e, Florida. Residues located at the surface region were easily assi gned from the modeled structure. The amino acid sequences of pro Pf PM1 and the efficiently refolded pro Pf PM4 were aligned. Without altering the overall catalytic ac tivity, hydrophobic and nonpolar residues of pro Pf PM1 were substituted by hydrophili c and polar ones from the pro Pf PM4 paralog (Figure 8-1). These point mutations will be ge nerated individually as the source of the library of pro Pf PM1 variants. In order to exclude a ny biased recombination, every single mutation product will be included. On the ot her hand, spatially close residues may have cooperative effects on protein folding, which al so will be considered. These mutants will be amplified by PCR and the PCR products will be subjected to DNa se I digestion. The condition for DNase I digestion will be contro lled such that DNA fragments of 10-50 bps will be obtained. The purified fragments will be reassembled in a PCR-like reaction by a high fidelity DNA polymerase. The resulting mixed pro Pf PM1 mutants will be further

PAGE 204

183 Figure 8-1. Amino acid sequence alignment of Pf PM1 and Pf PM4 and surface residue mutations on Pf PM1. The mature enzyme sequences of Pf PM1 and Pf PM4 are aligned. The active site residues are high lighted in red, residues residing in surface loop regions are highlighted in blue. Residues of Pf PM1 that are attempted to be substituted by those of Pf PM4 are underlined. A total of 23 mutations are designed, which will e nhance surface hydrophilicity of the target enzyme.

PAGE 205

184 Figure 8-2. DNA shuffling (Stemmer 1994). (A) A pool of mutated genes as well as the wild-type form is collected as substrates of the shuffling reaction. (B) The mutant library is fragmented with DNase I. (C) The fragments are reassembled into full-length genes by in vitro recombination via a minus primer PCR reaction. (D) The recomb inant fragment-shuffled genes are obtained by a plus primer PCR reaction after the reassembly, some examples of the recombinant fragment-shuffled ge nes are shown. (E) Se lected pool of improved recombinants gives the starting point for the next round of shuffling reaction.

PAGE 206

185 amplified by a conventional PCR to integrate the fragments. The whole process for the mutant library preparation is illustrated in Figure 8-2. The GFP split system is composed of a major GFP 1-10 portion and a minor GFP 11 fragment. The former represents the N-terminal 10 -strands (residue 1-214) of the intact molecule and the latter is the pep tide of the C-terminal strand (residue 214-230). The pro Pf PM1 coding sequence will be cloned at the N-terminal of the GFP 11 sequence and expressed as a fusion protein, and the large fragment GFP 1-10 will be expressed separately. Correctly folded pro Pf PM1 potentially allows the GFP 11 tag to be solventaccessible such that the two GFP fragments can be noncovalently associated and produce fluorescence under excitation conditions. On the other hand, misfolding of pro Pf PM1 will most likely wrap the GFP 11 fragment in side resulting in no fluorescence emission (Figure 8-3) (Cabant ous et al. 2005). The optimized GFP 1-10 portion (GFP 1-10 OP T) has been cloned into a modified pET-28a vector and the expres sion can be induced by IPTG (Figure 8-4). The optimized GFP 11 fragment (GFP 11 M3) has been cloned in to a modified pTet vector bearing an anhydrotetracycline (AnTet)-inducible tet promoter (Figure 8-5) (Lutz and Bujard 1997). The sequences of both peptides have been op timized to improve folding and solubility (Cabantous et al. 2005). The update ve rsion, GFP 1-10 OPT and GFP 11 M3, were obtained as a kind gift fro m Dr. Geoffrey S Waldo, Un iversity of New Mexico, Albuquerque, New Mexico. In order to confirm the expression of bot h fragments, the tr uncated version of pro Pf PM2 (the last 48 amino acid residues of th e prosegment plus the mature enzyme sequence) was cloned into the pTET-SpecR vector. The GFP 1-10 fragment and the

PAGE 207

186 Figure 8-3. Complementation of the split GFP fragment. The targeted protein and GFP fragment 11 is expressed as a fusion pr otein. The large fragment GFP 1-10 is expressed separately. The targeted pr otein, when properly folded, will most likely leave the GFP 11 tag exposed. Th e small and large GFP fragment will associate and form the fluorophore to develop fluorescence. When the targeted protein commits misfolding, wh ich may cause the tag buried inside, the two GFP fragments can not intera ct and fluorescence will not emit. X = Targeted protein, L = flexible peptide linker. Figure is adapted from Cabantous et al. 2005.

PAGE 208

187 Figure 8-4. GFP fragment 1-10 OPT expression system. (A) The pET-28a vector ( ). Expression is under the control of the T7 promoter and can be induced by IPTG. (B) Cloning Cassette for GFP fragment 1-10 OPT (Cabantous et al . 2005). The cassette is inserted in between Nde I and Xho I of the pET-28a vector. The GFP fragment 1-10 OPT is cloned in between Spe I and Kpn I.

PAGE 209

188 Figure 8-5. GFP 11 M3 fusion system (C abantous et al. 2005). (A) pTet-SpecR expression vector. The vector conf ers spectinomycin resistance. The expression is under the control of tet promoter (Lutz and Bujard1997), and is able to be induced by anhydrotetracyclin e (AnTet). The gene of fusion protein is accommodated in the multiple clone sites (MCS). (B) Cloning cassette for the fusion protein. The targeted gene is cloned in between Nde I and Bam H I. The GFP 11 M3 tag dwells in between Spe I and Kpn I. A (GGGS)2 linker covalently collects the two parts. 6 Histag resides upstream of the Nde I site. (C) Amino acid sequence of th e GFP 11 WT and M3 tag. The three mutations of GFP 11 M3 peptide is highlighted in red.

PAGE 210

189 pro Pf PM2-GFP 11 fusion protein were expressed in BL21 (DE3) pLysS E. coli by the induction of 1mM IPTG and 1mg/mL of AnTe t, respectively. The expression of both proteins was shown in Figure 8-6. The identi ties of the GFP 1-10 and the fusion protein were confirmed by MALDI-TOF mass spectrom etry following in-gel trypsin digestion treatment. The fragments of the pro Pf PM1 variant library will be su bcloned in fusion with the GFP 11 fragment. The resulting construct wi ll be cotransformed with the GFP 1-10–pET28a construct into the BL21 (DE3) pLysS E. coli expression strain. The expression of the pro Pf PM1-GFP 11 fusion protein a nd the GFP 1-10 fragment will be sequentially induced. The transformants will first be blotted onto a nitrocellulose membrane and placed onto AnTet containing LB plates. After induced expression, the inducer molecules will be allowed to diffuse away on plain LB plates. Following that, the transformants will be further induced by IPTG. Sequentially induced expression is necessary to avoid the false positive case wh ere the two portions of the GFP molecule associate before protein folding. The resulti ng colonies will be ex cited at wavelength 488 nm and emitted fluorescence at wavelength 520 nm will be observed. Those colonies producing brighter fluorescence will be coll ected. The mutated encoding sequences will be reshuffled again and transformants with higher intensity of fluorescence will be screened. The whole process will be repeated several times and the sequence of the final clones will be determined by DNA sequencing analyses, based on which site-directed mutagenesis will subsequently be performed to generate variants that can improve pro Pf PM1 solubility.

PAGE 211

190 Figure 8-6. The induced expression of recombinant GFP 1-10 (A) and pro Pf PM2-GFP 11 fusion protein (B) in E. coli . (A) M: low molecular weight marker (RPN 755, Amersham); 1: total cell lysate befo re IPTG induction (T = 0); 2: total cell lysate after 3 h IPTG induction (T = 3); 3: 1st wash of inclusion body (IB) (total); 4: 1st wash of IB (supernatant); 5: 1st wash of IB (pellet); 6: last wash of IB (total); 7: purif ied IB; M’: high molecular weight marker (RPN 756, Amersham). (B) M’: high molecular weight marker (RPN 756, Amersham); 14: total cell lysate after 0-3 h IP TG induction (T = 0, 1, 2, 3); 5-7: 1st wash of IB (total, supernatant, pe llet); 8: purified IB. The bands for in-gel trypsin digestion-mass spectrometry analys es are highlighted with arrows.

PAGE 212

191 We believe the DNA shuffling mediated mutant recombination and the GFP split system aided screening can allow us to iden tify plasmepsin variants with solubility improvement. A potential risk of employing this strategy on plasmepsins lies on their proteinase property. The sol uble, properly folded prot einase may degrade endogenous protein resources, although at neutral pH conditions of E. coli this event is hardly to occur due to the inability for plasmeps in zymogens to activate themselves. Development of highly potent plasmeps in-targeted antimalarial drugs relies on fully understanding characteristics of enzyme individuals. HAP, one of the plasmepsin members residing in the food vacuole, is far from being well studied. Among Plasmodium spp. infecting primates, only two species, P. falciparum infecting man and P. reichenowi infecting chimpanzee, has so far b een identified to possess HAP (Prof. John B. Dame, personal communication). With one of the active site residues Asp34 substituted by His, and approached Ser/ Thr residues, HAP is arguably rather a serine/threonine protease. The catalytic m echanism HAP may take and the contribution of the potentially unique features of HAP-i nhibitor interaction to novel antimalarials inspire us to study this particular enzyme from both the kinetic and the structural level. The semipro Pf HAP with the last 48 residues of th e propart plus the mature enzyme portion was cloned from the bloodstage cDNA library of the P. falciparum 3D7 strain. The expression of this recombinant HAP was first performed in the eukaryotic system Pichia pastoris . Production of the enzyme failed as the transcripts we re internally terminated. The expression of both the full length and the semipro Pf HAP was then carried out in E. coli (Figure 8-7). The semipro Pf HAP but not the full length HAP was overexpressed. The inclusion body materials of the semipro Pf HAP were prepared as

PAGE 213

192 Figure 8-7. SDS-PAGE analys is of expression of the full length (A) and semipro Pf HAP (B). M: high molecular weight marker (RPN 756, Amersham); Lane 1-4: total cell lysate before (T = 0) and after 1, 2 and 3 h IPTG induction; Lane 5-8: supernatant of different inclusion body purification steps; Lane 9: purified inclusion body form.

PAGE 214

193 described in Chapter 2. Because a His6-tag was linked in the N-terminus of the proenzyme, inclusion body materials were denatured in 6 M urea solution as for denaturation of pro Pb PM4 and primarily purified usi ng the Ni-chelating affinity chromatography. The resulting materials were dialyzed against 20 mM Tris-HCl, pH 8.0 at a final concentration of 0.3 mg/mL. Prot ein folding was verified by gel filtration chromatography (Figure 8-8) and circular dichroism (CD) spectrometry (Figure 8-9). However, the catalytic activity of recombinant HAP is hard to capture possibly due to the low efficiency of auto-activation or the nonspecific substrate used. In order to confer the auto -activation ability to pro Pf HAP, we have attempted to mutate the potential S2 pocket binding re sidue Lys110p during self -processing to Glu, Asn and Ile. The resulting species were st ill unable to perform self-conversion. Since prosegment residues of plasmepsins are also homologous, pro Pf HAP maturation may be accomplished with the help of its paralogs, such as Pf PM2 and Pf PM4, and this hypothesis needs to be tested. Another way to activate pro Pf HAP is to fuse it with another protein, such as thioredoxin, and when enterokinase digests the fusion protein, it will release the prosegment of HAP (Xiao et al. 2005). Combinatorial chemistry based approaches have previously been employed (Boss et al. 2003; Deperthes 2002; Hernandez and Roush 2002; Lien et al. 1999; Richardson 2002; Salemme et al. 1997) to study the s ubsite preferences of a variety of endopeptidases. The P1 and P1’ peptide librar ies have specifically been applied to determine active site features of human and ma laria aspartic proteina ses of the A1 family. As new members of plasmepsins are incr easingly being identif ied and successfully produced, their unique characte ristics on binding and hydrolyz ing peptide substrates are

PAGE 215

194 Figure 8-8. Gel filtration purification chromatogram of the recombinant semipro Pf PM2 and semipro Pf HAP. The elution of a majority of the pro Pf HAP materials was synchronized with that of the self-activatable pro Pf PM2 paralog, indicating that most of the recombinant HAP zymogen materials were folded to a similar size as their pro Pf PM2 homologue.

PAGE 216

195 Figure 8-9. Far UV circul ar dichoism (CD) spectrum of the semipro Pf HAP.

PAGE 217

196 Figure 8-10. Diagram of two combinatorial lib raries that explore the influences between subsites of different domains. Library A probes the linked P2-P1’ specificity and library B examines the preferen ces of S1-S2’. Residues from other position adopt the best ones speci fic for individual enzymes.

PAGE 218

197 intriguing for investigation. Hence, the tw o libraries were app lied to the subsite preferences of Pf PM1 and Pb PM4. Despite this, a further study on subsite preferences can be carried out from the following two aspects. First, the mutual influences of amino aci d substitutions from adjacent positions on fitting to an enzyme need to be reconsider ed. The interdependency of residues in the context peptide sequence have been well observe d in this and previous studies. The P1 and P1’ combinatorial libraries are design to investigate the interplays between residues projecting to the same domain. Besides, ne w combinatorial libraries are required to explore the mutual influences of neighboring re sidues that dwelled in pockets of different domains. Two examples of such constructi ons are depicted in Figure 8-10. The first library (Figure 8-10 A) is developed to expl ore simultaneously the preferences of the S2, S1 and S1’ subsite; while the second (Figur e 8-10 B), the S1, S1’ and S2’ subsite. The best amino acid substitutions from previous studies will be adopted for the rest of the residues in these libraries. For such two libraries, since no chromophores are incorporated, equivalent amounts of peptide pools enzymes as well as the enzyme must be employed. Identification and quantification of the cleavage products will fully rely on LC-MS. Secondly, the building blocks of combinatoria l libraries should be more extensive. Besides the natural amino acids, the nonnatura l amino acids as well as other functional groups could also be incorpor ated. Numerous studies have been carried out on subsite specificities of plasmepsins from either combinatorial library screening or single substrate/inhibitor analyses. Functional gr oups from substrates with high cleavage efficiency and inhibitors with high binding affinity will be documented. A collection of

PAGE 219

198 strongly favored functional groups for a specific subsite will be equally incorporated to the corresponding position. Digestion of the resulting combinatorial libraries can be analyzed and compounds with hi gher specificity can be obtaine d from a broader range of screening. Plasmepsins, the aspartic proteinase fa mily of malaria para sites, have been considered potential targets for novel antim alarial drugs. A successful journey of drug discovery requires a comprehe nsive understanding the feat ures of targets from the structural, kinetic and chemical level. Studies addressed in this thesis complement our knowledge to two new members of plas mepsin enzymes. The production and achievement of active enzymes were described; in what follows, the subsite specificities of the two enzymes were systematically discussed using combinatorial chemistry approaches. Inhibition analyses of single mechanism-based compounds (clinical HIV-1 PIs and statine-based “doubl e-drug” compounds) on the plasmepsin families further enriched information on unique binding featur es of these homologues. All these findings may contribute to design ideal inhi bitors against the plasmepsins.

PAGE 220

199 LIST OF REFERENCES Aikawa M, Seed TM. 1980. Morphology of Plasmodia . In Malaria: Epidemology, Chemotherapy, Morphology, and Metabolism (Kreier JP ed.). Vol. 1, pp. 285-344. Academic Press, New York, London, Toronto, Sydney, San Francisco. Andrews KT, Fairlie DP, Madala PK, Ray J, Wyatt DM, Hilton PM, Melville LA, Beattie L, Gardiner DL, Reid RC, Stoermer MJ, Skinner-Adams T, Berry C, McCarthy JS. 2006. Potencies of human immunodeficien cy virus protea se inhibitors in vitro against Plasmodium falciparum and in vivo against murine malaria. Antimicrob Agents Chemother 50(2): 639-648. Arnold FH. 1996. Directed evolutio n: creating biocatalysts for the future. Chem Eng Sci 51(23): 5091-5102. Asojo OA, Afonina E, Gulnik SV, Yu B, Er ickson JW, Randad R, Mehadjed D, Silva AM . 2002a. Structures of Ser205 mu tant plasmepsin II from Plasmodium falciparum at 1.8 in complex with the inhibitors rs367 and rs370. Acta Crystallogr Sect D Biol Crystallogr 58(Pt 12): 2001-2008. Asojo OA, Gulnik SV, Afonina E, Yu B, Ellman JA, Haque TS, Silva AM. 2002b (structure released). Crystal st ructure of plasmepsin IV from P. falciparum in complex with pepstatin A. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Asojo OA, Gulnik SV, Afonina E, Yu B, Ellman JA, Haque TS, Silva AM. 2003. Novel uncomplexed and complexed structures of pl asmepsin II, an aspartic protease from Plasmodium falciparum . J Mol Biol 327(1): 173-181. Atherton D. 1989. Successful PTC amino acid analysis at the picomolar level. In Techniques in Protein Chemistry (Hugli TE ed.). pp. 273-283. Academic Press, San Diego, New York, Berkeley, Bo ston, London, Sydney, Tokyo, Toronto. Backes BJ, Harris JL, Leonetti F, Craik CS , Ellman JA. 2000. Synthesis of positionalscanning libraries of fluorogeni c peptide substrates to define the extended substrate specificity of plasmin and thrombin. Nat Biotechnol 18(2): 187-193. Erratum in: Nat Biotechnol 18(5): 559. Bailly E, Jambou R, Savel J, Jaureguiberry G. 1992. Plasmodium falciparum : differential sensitivity in vitro to E-64 (cysteine protease inhibitor) and pepstatin A (aspartyl protease inhibitor). J Protozool 39(5): 593-599.

PAGE 221

200 Baird JK. 1995. Host age as a determinant of naturally acquired immunity to Plasmodium falciparum . Parasitol Today 11(3): 105-111. Ball EG, McKee RW, Anfinsen CB, Cruz WO, Geiman QM. 1948. Studies on malarial parasites: ix. chemical and metabolic changes during growth and multiplication in vivo and in vitro . J Biol Chem 175(2): 547-571. Banerjee R, Francis SE, Goldberg DE. 2003. F ood vacuole plasmepsins are processed at a conserved site by an acid ic convertase activity in Plasmodium falciparum . Mol Biochem Parasitol 129(2): 157-165. Banerjee R, Liu J, Beatty W, Pelosof L, Klemba M, Goldberg DE. 2002. Four plasmepsins are active in the Plasmodium falciparum food vacuole, including a protease with an active-site histidine. Proc Natl Acad Sci USA 99(2): 990-995. Batra S, Srinivasan T, Rastogi SK, Kundu B. 2002. Identification of enzyme inhibitors using combinatorial libraries. Curr Med Chem 9(3): 307-319. Bernstein NK, Cherney MM, Loetscher H, Ridley RG, James MNG. 1999. Crystal structure of the novel aspartic protei nase zymogen proplasmepsin II from Plasmodium falciparum . Nat Struct Biol 6(1): 32-37. Bernstein NK, Cherney MM, Yowell CA, Da me JB, James MNG. 2003. Structural insights into the activation of P. vivax plasmepsin. J Mol Biol 329(3): 505-524. Berry C, Humphreys MJ, Matharu P, Granger R, Horrocks P, Moon RP, Certa U, Ridley RG, Bur D, Kay J. 1999. A distinct member of the aspartic proteinase gene family from the human malaria parasite Plasmodium falciparum . FEBS Lett 447(2-3): 149-154. Best B, Haubrich R. 2006. Tipranavir: a prot ease inhibitor for multi-drug resistant HIV-1. Exp Opin Investig Drugs 15(1): 59-70. Beyer BB. 2003.Targeted chromogenic octapeptid e combinatorial libraries: exploration of the primary and extended subsite specifi cities of human and malarial aspartic endopeptidases. Doctoral dissertation in Biochemistry and Molecular Biology, University and Florida, College of Medi cine. Gainesville, Florida. Ref Type: Thesis/Dissertation. Beyer BB, Johnson JV, Chung AY, Li T, Madabushi A, Agbandje-McKenna M, McKenna R, Dame JB, Dunn BM. 2005. Ac tive-site specificity of digestive aspartic peptidases from the four species of Plasmodium that infect humans using chromogenic combinatorial peptide li braries. Biochemistry 44(6): 1768-1779. Beyer BM, Dunn BM. 1996. Self-activati on of recombinant human lysosomal procathepsin D at a newly engineered cleavage junction, "sho rt" pseudocathepsin D. J Biol Chem 271(26): 15590-15596.

PAGE 222

201 Bhisuttibhan J, Philbert MA, Fujioka H, Aikawa M, Meshnick SR. 1999. The Plasmodium falciparum translationally controlled tumor protein. Subcellular localization and calcium binding. Eur J Cell Biol 78(9): 665-670. Bjelic S, qvist J. 2004. Com putational prediction of stru cture, substrate binding mode, mechanism, and rate for a malaria prot ease with a novel type of active site. Biochemistry 43(46): 14521-14528. Boffito M, Bonora MS, Raiteri R, Reynolds HE, Hoggard PG, Sinicco A, Back DJ, Di Perri G. 2002. Pharmacokinetic evaluation of indinavir and indinavir/ritonavircontaining antiretroviral regimens in a c linical setting. Ther Drug Monit 24(4): 574-576. Borman S. 1996. Combinatorial chemists focus on small molecules, molecular recognition, and automation. Chem Eng News 74(7): 29-54. Borman S. 1998. Biosynthesis combinat orially. Chem Eng News 76(37): 29-30. Borman S. 1999. Reducing time to drug discovery. Chem Eng News 77(10): 33-48. Boss C, Richard-Bildstein S, Weller T, Fisc hli W, Meyer S, Binkert C. 2003. Inhibitors of the Plasmodium falciparum parasite aspartic protease plasmepsin II as potential antimalarial agents. Curr Med Chem 10(11): 883-907. Boyd MF. 1949. Historical review. In Malariology (Boyd MF ed.). Vol. 1, pp. 3-25. WB Saunders Company, Philadelphia and London. Bozdech Z, Llins M, Pulliam BL, W ong ED, Zhu J-C, DeRisi JL. 2003. The transcriptome of the intraeryth rocytic developmental cycle of Plasmodium falciparum . PLoS Biol 1(1): 85-100. Breman JG. 2001. The ears of the hippopotamus: manifestations, determinants, and estimates of the malaria burden. Am J Trop Med Hyp 64(1-2 Suppl): 1-11. Bruce-Chwatt LJ. 1988. History of malaria from prehistory to eradication. In Malaria: Principles and Practice of Malariology (W ernsdorfer WH, Sir McGregor I eds.). Vol. 1, pp. 1-60. Churchill Livingstone, Edinburgh. Brnger AT, Adams PD, Clore GM, DeLano WL , Gros P, Grosse-Kunstleve RW, Jiang J-S, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL. 1998. Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr Sect D Biol Crystallogr 54(Pt 5): 905-921. Cabantous S, Terwilliger TC, Waldo GS. 2005. Protein tagging and detection with engineered self-assembling fragments of green fluorescent protein. Nat Biotechnol 23(1): 102-107.

PAGE 223

202 Campbell CC, Chin W, Collins WE, Teutsch SM, Moss DM. 1979. Chloroquine-resistant Plasmodium falciparum from East Africa: cultivati on and drug sensitivity of the Tanzanian I/CDC strain from an Ameri can tourist. Lancet 2(8153): 1151-1154. Canfield CJ, Canfield CJ, Pudney M, Gu tteridge WE. 1995. Interactions of atovaquone with other antimalarial drugs against Plasmodium falciparum in vitro . Exp Parasitol 80(3): 373-381. Carlton JMR, Galinski MR, Barnwell JW , Dame JB. 1999. Karyotype and synteny among the chromosomes of all four speci es of human malaria parasite. Mol Biochem Parasitol 101(1-2): 23-32. Carroll CD, Johnson TO, Tao S, Lauri G, Orlo wski M, Gluzman IY, Goldberg DE, Dolle RE. 1998a. Evaluation of a structure-based statine cyclic diamino amide encoded combinatorial library against plasmepsin II and cathepsin D. Bioorg Med Chem Lett 8(22): 3203-3206. Carroll CD, Patel H, Johnson TO, Guo T, Or lowski M, He ZM, Cavallaro CL, Guo J, Oksman A, Gluzman IY, Connelly J, Chelsky D, Goldberg DE, Dolle RE. 1998b. Identification of pot ent inhibitors of Plasmodium falciparum plasmepsin II from an encoded statine combinatorial librar y. Bioorg Med Chem Lett 8(17): 2315-2320. Carson PE. 1984. 8-aminoquinolines. In Antimalarial Drugs II. Current Antimalarials and New Drug Developments. Handbook of Expe rimental Pharmacology (Peters W, Richards WHG eds.). Vol. 68, pp. 83-122. Springer-Verlag, Berlin, Heidelberg, New York, Tokyo. Carter R, Mendis KN, Miller LH, Molinea ux L, Saul A. 2000. Malaria transmissionblocking vaccines—how can their developm ent be supported? Nat Med 6(3): 241244. CDC 2006a. The history of malaria, an ancient disease. Quinine (early 17th century). In . Last Accessed: July, 2006. Ref Type: Electronic Citation. CDC 2006b. The history of malaria, an ancien t disease. Eradicati on efforts worldwide: success and failure (1955-1978). In y/index.htm#eradicationworldwide . Last Accessed: July, 2006. Ref Type: Electronic Citation. CDC 2006c. Geographic distri bution and epidemiology. Geographic distribution. In ribution_epi/distribution.htm . Last Accessed: July, 2006. Ref Type: Electronic Citation. Chan TCK, Shaffer L, Redmond R, Penni ngton KL. 1993. Permeation and metabolism of anti-HIV and endogenous nucleosides in hum an immune effector cells. Biochem Pharmacol 46(2): 273-278.

PAGE 224

203 Clemente JC, Govindasamy L, Madabushi A, Fisher SZ, Moose RE, Yowell CA, Hidaka K, Kimura T, Hayashi Y, Kiso Y, Agbandje-McKenna M, Dame JB, Dunn BM, McKenna R. 2006. Structure of the aspart ic protease plasmepsin 4 from the malarial parasite Plasmodium malariae bound to an allophen ylnorstatine-based inhibitor. Acta Crystallogr D Biol Crystallogr 62(3): 246-252. Clemente JC, Moose, RE, Hemrajani R, Wh itford LRS, Govindasamy L, Reutzel R, McKenna R, Agbandje-McKenna M, Goodenow MM, Dunn BM. 2004. Comparing the accumulation of activeand nonactive-site mutations in the HIV-1 protease. Biochemistry 43(38): 12141-12151. Clyde DF. 1990. Immunity to falciparum and vivax malaria induced by irradiated sporozoites: a review of the University of Maryland studies , 1971-75. Bull World Health Organ 68(Suppl): 9-12. Coleman PG, Alphey L. 2004. Genetic control of vector populations: an imminent prospect. Trop Med Int Health 9(4): 433-437. Collins WE, Contacos PG. 1972. Imm unization of monkeys against Plasmodium cynomolgi by X-irradiated sporozo ites. Nature 236(67): 176-177. Coombs GH, Goldberg DE, Klemba M, Berry C, Kay J, Mottram JC. 2001. Aspartic proteases of Plasmodium falciparum and other parasitic prot ozoa as drug targets. Trends in Parasitol 17(11): 532-537. Cortese JF, Caraballo A, Contreras CE, Pl owe CV. 2002. Origin and dissemination of Plasmodium falciparum drug-resistance mutations in South America. J Infect Dis 186(7): 999-1006. Dame JB, Reddy GR, Yowell CA, Dunn BM, Kay J, Berry C. 1994. Sequence, expression and modeled structure of an aspa rtic proteinase from the human malaria parasite Plasmodium falciparum . Mol Biochem Parasitol 64(2): 177-190. Dame JB, Yowell CA, Opyene LO, Carlton JMR, Cooper RA, Li T. 2003. Plasmepsin 4, the food vacuole aspartic proteinase found in all Plasmodium spp. infecting man. Mol Biochem Parasitol 130(1): 1-12. Deperthes D. 2002. Phage display substrate: a blind method for determining protease specificity. Biol Ch em 383(7-8): 1107-1112. Desjardins RE, Canfield CJ, Haynes JD, Chulay JD. 1979. Quantitative assessment of antimalarial activity in vitro by a semiautomated microdilution technique. Antimicrob Agents Chemother 16(6): 710-718. Despommier DD, Karapelon JW. 1987. Part 1 Protozoa. In Parasite Life Cycles (Despommier DD, Karapelon JW eds.). pp. 28-37. Springer-Verlag, New York, Berlin, Heidelberg, London, Paris, Tokyo.

PAGE 225

204 Domin BA, Mahony WB, Zimmerman TP. 1993. Membrane permeation mechanisms of 2',3'-dideoxynucleosides. Bioche m Pharmacol 46(4): 725-729. Dorsey G, Fidock DA, Wellems TE, Rose nthal PJ. 2001. Mechanisms of quinoline resistance. In Antimalarial Chemotherapy. Mechan isms of Action, Resistance, and New Directions in Drug Discovery (Rosen thal PJ ed.). pp. 153-172. Humana Press, Totowa, New Jersey. Dunn BM. 2002. Structure and mechanism of the pepsin-like family of aspartic peptidases. Chem Rev 102(12): 4431-4458. Dunn BM, Hung SH. 2000. The two sides of en zyme-substrate specificity: lessons from the aspartic proteinases. Bioc him Biophys Acta 1477(1-2): 231-240. Dunn BM, Jimenez M, Parten BF, Valler MJ, Rolph CE, Kay J. 1986. A systematic series of synthetic chromophoric substrates for aspartic proteinases. Biochem J 237(3): 899-906. Dunn BM, Scarborough PE, Davenport R, Swietnic ki W. 1994. Analysis of proteinase specificity by studies of peptide subs trates: the use of UV and fluorescence spectroscopy to quantitate ra tes of enzymatic cleavage. In Peptide Analysis Protocols (Dunn BM, Pennington MW eds.). Methods in Molecular Biology. Vol. 36, pp. 225-243. Humana Press, Totowa, New Jersey. Edman P. 1949. A method for the determinati on of the amino acid se quence in peptides. Arch Biochem Biophys 22 (3): 475-480. Egan JE, Hoffman SL, Haynes JD, Sadoff JC , Schneider I, Grau GE, Hollingdale MR, Ballou WR, Gordon DM. 1993. Humoral i mmune responses in volunteers immunized with irradiated Plasmodium falciparum sporozoites. Am J Trop Med Hyg 49(2):166-173. Eggleson KK, Duffin KL, Goldberg DE. 1999. Identification and characterization of falcilysin, a metallopeptidase involved in hemoglobin catabolism within the malaria parasite Plasmodium falciparum . J Biol Chem 274(45): 32411-32417. FDA. June, 2006. FDA approves new HIV treatme nt for patients who do not respond to existing drugs. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Fidock DA, Nomura T, Talley AK, Cooper RA, Dzekunov SM, Ferdig MT, Ursos LM, Sidhu AB, Naude B, Deitsch KW, Su XZ , Wootton JC, Roepe PD, Wellems TE. 2000. Mutations in the P. falciparum digestive vacuole transmembrane protein Pf CRT and evidence for their role in ch loroquine resistance. Mol Cell 6(4): 861871.

PAGE 226

205 Fidock DA, Nomura T, Wellems, TE. 1998. Cycloguanil and its parent compound proguanil demonstrate dist rict activities against Plasmodium falciparum malaria parasites transformed with human dihydr ofolate reductase. Mol Pharmacol 54(6): 1140-1147. Foote SJ, Thompson JK, Cowman AF, Kemp DJ. 1989. Amplification of a multidrug resistant gene in some chlor oquine-resistant isolates of P. falciparum . Cell 57(6): 921-930. Francis SE, Banerjee R, Goldberg DE. 1997a. Bi osynthesis and maturation of the malaria aspartic hemoglobinases plasmepsins I and II. J Biol Chem 272(23): 14961-14968. Francis SE, Gluzman IY, Oksman A, Bane rjee D, Goldberg DE. 1996. Characterization of native falcipain, an enzyme involved in Plasmodium falciparum hemoglobin degradation. Mol Bioche m Parasitol 83(2): 189-200. Francis SE, Gluzman IY, Oksman A, Knicker bocker A, Mueller R, Bryant ML, Sherman DR, Russell DG, Goldberg DE. 1994. Molecula r characterization and inhibition of a Plasmodium falciparum aspartic hemoglobinase. EMBO J 13(2): 306-317. Francis SE, Sullivan DJ Jr, Goldberg, DE. 1997b. Hemoglobin metabolism in the malaria parasite Plasmodium falciparum . Annu Rev Microbiol 51(1): 97-123. Freire E, Nezami AG, Amzel LM. 2004 (struc ture released). Cr ystal structure of plasmepsin II, an aspartic protease from Plasmodium falciparum , in complex with a statine-based inhibitor. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Gardner MJ, Hall N, Fung E, White O, Berri man M, Hyman RW, Carlton JM, Pain A, Nelson KE, Bowman S, Paulsen IT, James K, Eisen JA, Rutherford K, Salzberg SL, Craig A, Kyes S, Chan M-S, Nene V, Shallom SJ, Suh B, Peterson J, Angiuoli S, Pertea M, Allen J, Selengut J, Haft D, Mather MW, Vaidya AB, Martin DM, Fairlamb AH, Fraunholz MJ, Roos DS, Ralph SA, McFadden GI, Cummings LM, Subramanian GM, Mungall C, Venter JC, Carucci DJ, Hoffman SL Newbold C, Davis RW, Fraser CM, Barrell B. 2002a. Genome sequence of the human malaria parasite Plasmodium falciparum . Nature 419(6906): 498-511. Gardner MJ, Shallom SJ, Carlton JM, Salzberg SL, Nene V, Shoaibi A, Ciecko A, Lynn J, Rizzo M, Weaver B, Jarrahi B, Bre nner M, Parvizi B, Tallon L, Moazzez A, Granger D, Fujii C, Hansen C, Pederson J, Feldblyum T, Peterson J, Suh B, Angiuoli S, Pertea M, Allen J, Selengut J, White O, Cummings LM, Smith HO, Adams MD, Venter JC, Carucci DJ, Ho ffman SL, Fraser CM. 2002b. Sequence of Plasmodium falciparum chromosomes 2, 10, 11 and 14. Nature 419(6906): 531534.

PAGE 227

206 Gasteiger E, Hoogland C, Gattiker A, Duva ud S, Wilkins MR, Appel RD, Bairoch A. 2005. Protein identification and analys is tools on the ExPASy server. In The Proteomics Protocols Handbook (Walker JM ed.). pp. 571-607. Humana Press, Totowa, New Jersey. Gluzman IY, Francis SE, Oksman A, Smith CE, Duffin KL, Goldberg DE. 1994. Order and specificity of the Plasmodium falciparum hemoglobin degradation pathway. J Clin Invest 93(4): 1602-1608. Goldberg DE. 1993. Hemoglobin degradation in Plasmodium -infected red blood cells. Semin Cell Biol 4(5): 355-361. Goldberg DE, Slater AF, Beavis R, Chait B, Cerami A, Henderson GB. 1991. Hemoglobin degradation in the human malaria pathogen Plasmodium falciparum : a catabolic pathway initiated by a specific aspartic protease. J Exp Med 173(4): 961969. Goldberg DE, Slater AFG, Cerami A, He nderson GB. 1990. Hemoglobin degradation in the malaria parasite Plasmodium falciparum : an ordered process in a unique organelle. Proc Natl Acad Sci USA 87(8): 2931-2935. Goujard C, Vincent I, Meynard JL, Choudet N, Bollens D, Rousseau C, Demarles D, Gillotin C, Bidault R, Taburet AM . 2003. Steady-state pharmacokinetics of amprenavir coadministered with ritonavi r in human immunodeficiency virus type 1-infected patients. Antimicrob Agents Chemother 47(1): 118-123. Greenwood B, Mutabingwa T. 2002. Ma laria in 2002. Nature 415(6872): 670-672. Gu J, Chen K, Jiang H, Leszczynski J. 1999. A model molecule study of the O-centered and the C-centered free radical interm ediates of artemisinin. J Mol Struct (Theochem) 491(1-3): 57-66. Gutirrez-de-Tern H, Nervall M, Ersmark K, Liu P, Janka LK, Dunn BM, Hallberg A, qvist J. 2006. Inhibitor binding to the plasmepsin IV aspartic protease from Plasmodium falciparum . Accepted by Biochemistry.

PAGE 228

207 Hall N, Pain A, Berriman M, Churcher C, Harris B, Harris D, Mungall K, Bowman S, Atkin R, Baker S, Barron A, Brooks K, Buckee CO, Burrows C, Cherevach I, Chillingworth C, Chillingworth T, Christodoulou Z, Clark L, Clark R, Corton C, Cronin A, Davies R, Davis P, Dear P, D earden F, Doggett J, Feltwell T, Goble A, Goodhead I, Gwilliam R, Hamlin N, Hance Z, Harper D, Hauser H, Hornsby T, Holroyd S, Horrocks P, Humphray S, Ja gels K, James KD, Johnson D, Kerhornou A, Knights A, Konfortov B, Kyes S, La rke N, Lawson D, Lennard N, Line A, Maddison M, McLean J, Mooney P, Moul e S, Murphy L, Oliver K, Ormond D, Price C, Quail MA, Rabbinowitsch E, Ra jandream MA, Rutter S, Rutherford KM, Sanders M, Simmonds M, Seeger K, Sharp S, Smith R, Squares R, Squares S, Stevens K, Taylor K, Tivey A, Unwin L, Whitehead S, Woodward J, Sulston JE, Craig A, Newbold C, Barrell BG. 2002. Sequence of Plasmodium falciparum chromosomes 1, 3-9 and 13. Nature 419(6906): 527-531. Han H, Wolfe MM, Brenner S, Janda KD. 1995. Liquid-phase combinatorial synthesis. Proc Natl Acad Sci USA 92(14): 6419-6423. Haque TS, Skillman AG, Lee CE, Habashita H, Gluzman IY, Ewing TJ, Goldberg DE, Kuntz ID, Ellman JA. 1999. Potent, low-mo lecular-weight non-pe ptide inhibitors of malarial aspartyl protease plas mepsin II. J Med Chem 42(8): 1428-1440. Hayashi M, Yamada H, Mitamura T, Horii T, Yamamoto A, Moriyama Y. 2000. Vacuolar H(+)-ATPase localized in plasma membranes of malaria parasite cells, Plasmodium falciparum , is involved in regional ac idification of parasitized erythrocytes. J Biol Chem 275(44): 34353-34358. Haynes RK. 2001. Artemisinin and derivatives: th e future for malaria treatment? Curr Opin Infect Dis 14(6): 719-726. Henderson PJF. 1972. A linear equation that de scribes the steady-state kinetics of enzymes and subcellular particles inte racting with tightly bound inhibitors. Biochem J 127(2): 321-333. Hernandez AA, Roush WR. 2002. Recent advances in the synthesis, design and selection of cysteine protease inhibitors. Cu rr Opin Chem Biol 6(4): 459-465. Hill J, Tyas L, Phylip LH, Kay J, Dunn BM, Berry C. 1994. High level expression and characterization of Plasmepsin II , an aspartic proteinase from Plasmodium falciparum . FEBS Lett 352(2): 155-158. Hisaeda H, Stowers AW, Tsuboi T, Collins WE, Sattabongkot JS, Suwanabun N, Torii M, Kaslow DC. 2000. Antibodies to malari a vaccine candidates Pvs25 and Pvs28 completely block the ability of Plasmodium vivax to infect mosquitoes. Infect Immun 68(12): 6618-6623.

PAGE 229

208 Holt RA, Subramanian GM, Halpern A, Sutt on GG, Charlab R, Nusskern DR, Wincker P, Clark AG, Ribeiro JM, Wides R, Salz berg SL, Loftus B, Yandell M, Majoros WH, Rusch DB, Lai Z, Kraft CL, Abril JF , Anthouard V, Arensburger P, Atkinson PW, Baden H, de Berardinis V, Baldwin D, Benes V, Biedler J, Blass C, Bolanos R, Boscus D, Barnstead M, Cai S, Cent er A, Chaturverdi K, Christophides GK, Chrystal MA, Clamp M, Cravchik A, Curwen V, Dana A, Delcher A, Dew I, Evans CA, Flanigan M, Grundschober-Freimoser A, Friedli L, Gu Z, Guan P, Guigo R, Hillenmeyer ME, Hladun SL, Hogan JR, H ong YS, Hoover J, Jaillon O, Ke Z, Kodira C, Kokoza E, Koutsos A, Letunic I, Levitsky A, Liang Y, Lin JJ, Lobo NF, Lopez JR, Malek JA, McIntosh TC, Meiste r S, Miller J, Mobarry C, Mongin E, Murphy SD, O'Brochta DA, Pfannkoch C, Qi R, Regier MA, Remington K, Shao H, Sharakhova MV, Sitter CD, Shetty J, Smith TJ, Strong R, Sun J, Thomasova D, Ton LQ, Topalis P, Tu Z, Unger MF, Wa lenz B, Wang A, Wang J, Wang M, Wang X, Woodford KJ, Wortman JR, Wu M, Yao A, Zdobnov EM, Zhang H, Zhao Q, Zhao S, Zhu SC, Zhimulev I, Coluzzi M, della Torre A, Roth CW, Louis C, Kalush F, Mural RJ, Myers EW, Adams MD, Smith HO, Broder S, Gardner MJ, Fraser CM, Birney E, Bork P, Brey PT, Venter JC, Weissenbach J, Kafatos FC, Collins FH, Hoffman SL. 2002. The genome sequence of the malaria mosquito Anopheles gambiae . Science 298(5591): 129-149. Houghten RA, Pinilla C, Blondelle SE, Appel JR, Dooley CT, Cuervo JH. 1991. Generation and use of synthetic peptide co mbinatorial libraries for basic research and drug discovery. Nature 354(6348): 84-86. Huang QQ, Yao SY, Ritzel MW, Paters on AR, Cass CE, Young JD. 1994. Cloning and functional expression of a complement ary DNA encoding a mammalian nucleoside transport protein. J Biol Chem 269(27): 17757-17760. Humphreys MJ, Moon RP, Klinder A, Fowler SD, Rupp K, Bur D, Ridley RG, Berry C. 1999. The aspartic proteinase from the rodent parasite Plasmodium berghei as a potential model for plasmepsins fr om the human malaria parasite, Plasmodium falciparum . FEBS Lett 463(1-2): 43-48. Istvan ES, Goldberg DE. 2005. Distal substrat e interactions enhan ce plasmepsin activity. J Biol Chem 280(8): 6890-6896. James MNG, Sielecki AR, Hayakawa K, Ge lb MH. 1992. Crystallographic analysis of transition state mimics bound to peni cillopepsin: difluorostatineand difluorostatone-containing peptid es. Biochemistry 31(15): 3872-3886. Janse CJ, Boorsma EG, Ramesar J, Grobbee MJ , Mons B. 1989. Host cell specificity and schizogony of Plasmodium berghei under different in vitro conditions. Int J Parasitol 19(5): 509-514. Janse CJ, Mons B, Croon JJ, van der Kaay HJ. 1984. Long-term in vitro cultures of Plasmodium berghei and preliminary observations on gametocytogenesis. Int J Parasitol 14(3): 317-320.

PAGE 230

209 Jiang S, Prigge ST, Wei L, Gao Ye, Huds on TH, Gerena L, Dame JB, Kyle DE. 2001. New class of small nonpeptidyl compounds blocks Plasmodium falciparum development in vitro by inhibiting plasmepsins. Antimicrob Agents Chemother 45(9): 2577-2584. Jones G, Willett P, Glen RC, Leach AR, Ta ylor R. 1997. Development and validation of a genetic algorithm for flexible docking. J Mol Biol 267(3): 727-748. Kakkilaya BS. April, 2006 (last updated ). Mosquito control: source reduction. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Kaufman TC, Severson DW, Robinson GE. 2002. The Anopheles genome and comparative insect genomics. Science 298(5591): 97-98. Kendrick RK. 1978. Taxonomu, z oogeography and evolution. In Rodent Malaria (Kendrick RK, Peters W eds.). pp. 1-52. Academic Press, London, New York, San Francisco. Kester KE, McKinney DA, Tornieporth N, Ockenhouse CF, Heppner DG, Hall T, Krzych U, Delchambre M, Voss G, Dowler MG, Palensky J, Wittes J, Cohen J, Ballou WR; RTS,S Malaria Vaccine Evaluati on Group. 2001. Efficacy of recombinant circumsporozoite protein vaccine regimens against experimental Plasmodium falciparum malaria. J Infect Dis 183(4):640-647. Khan AR, Bernstein NK, Bergmann EM, Ja mes MNG. 1999. Struct ural aspects of activation pathways of aspartic protea se zymogens and viral 3C protease precursors. Proc Natl Acad Sci USA 96(20): 10968-10975. Kimura T, Matsumoto H, Matsuda T, Hamawa ki T, Akaji K, Kiso Y. 1999. A new class of anti-HIV agents: synthesis and activity of conjugates of HIV protease inhibitors with a reverse transcriptase inhibito r. Bioorg Med Chem Lett 9(6): 803-806. King NM, Prabu-Jeyabalan M, Nalivaika EA, Wigerinck PBTP, de Bthune MP, Schiffer CA. 2005 (structure released). Crystal structure of inhib itor amprenavir in complex with a multi-drug resistant variant of HIV-1 protease (L63P/V82T/I84V). In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Kiso Y, Matsumoto H, Yamaguchi S, Kimu ra T. 1999. Design of small peptidomimetic HIV-1 protease inhibitors and prodrug forms. Lett in Peptide Sci 6(5-6): 275-281. Klemba M, Goldberg DE. 2005. Characteriza tion of plasmepsin V, a membrane-bound aspartic protease homolog in the endoplasmic reticulum of Plasmodium falciparum . Mol Biochem Para sitol 143(2): 183-191.

PAGE 231

210 Kolakovich KA, Gluzman IY, Duffin KL, Go ldberg DE. 1997. Generation of hemoglobin peptides in the acidic digestive vacuole of Plasmodium falciparum implicates peptide transport in amino acid produc tion. Mol Biochem Parasitol 87(2): 123-135. Krogstad DJ, Schlesinger PH, Gluzman IY. 1985. Antimalarials increase vesicle pH in Plasmodium falciparum . J Cell Biol 101(6): 2302-2309. Kublin JG, Dzinjalamala FK, Kamwendo DD, Malkin EM, Cortese JF, Martino LM, Mukadam RA, Rogerson SJ, Lescano AG, Molyneux ME, Winstanley PA, Chimpeni P, Taylor TE, Plowe CV. 2002. Molecular markers for failure of sulfadoxine-pyrimethamine and chlo rproguanil-dapsone treatment of Plasmodium falciparum malaria. J Infect Dis 185(3): 380-388. Lam KS, Salmon SE, Hersh EM, Hruby VJ, Kazmierski WM, Knapp RJ. 1991. A new type of synthetic peptide library for identifying liga nd-binding activity. Nature 354(6348): 82-84. Erratum in: 1992. Na ture 360(6406): 768; 1992. Nature 358(6385): 434. Landau I, Boulard Y. 1978. Life cycles and morphology. In Rodent Malaria (Kendrick RK, Peters W eds.). pp. 53-84. Academic Press, London, New York, San Francisco. Leatherbarrow RJ, Fersht AR, Winter G. 1985. Transition-state st abilization in the mechanism of tyrosyl-tRNA synthetase reve aled by protein engi neering. Proc Natl Acad Sci USA 82(23): 7840-7844. Le Bonniec S, Deregnaucourt C, Redeker V, Banerjee R, Grellier P, Goldberg DE, Schrevel J. 1999. Plasmepsin II, an acidic hemoglobinase from the Plasmodium falciparum food vacuole, is active at neutral pH on the hos t erythrocyte membrane skeleton . J Biol Chem 274(20): 14218-14223. Lew VL, Tiffert T, Ginsberg H. 2003. Exce ss hemoglobin digestion and the osmotic stability of Plasmodium falciparum -infected red blood ce lls. Blood 101(10): 41894194. Li T, Yowell CA, Beyer BB, Hung SH, Wes tling J, Lam MT, Dunn BM, Dame JB. 2004. Recombinant expression and enzymatic subs ite characterization of plasmepsin 4 from the four Plasmodium species infecting man. Mo l Biochem Parasitol 135(1): 101-109. Li Y, Wu Y-L. 1998. How Chinese scientists discovered qinghaosu (artemisinin) and developed its derivatives. Wh at are the future perspectives? Med Trop 58(3 Suppl): 9-12. Lien S, Francis GL, Graham LD. 1999. Combin atorial strategies for the discovery of novel protease specificities. Comb Ch em High Throughput Screen 2(2): 73-90.

PAGE 232

211 Lindberg J, Johansson P-O, Rosenquist A, Kv arnstroem I, Vrang L, Samuelsson B, Unge T. 2006a (structure released). Structural study of a novel inhibitor with bulky P1 side chain in complex with plasmepsin II--implication for drug design. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Lindberg J, Johansson P-O, Rosenquist A, Kv arnstroem I, Vrang L, Samuelsson B, Unge T. 2006b (structure released). Structural study of a novel inhibitor with bulky P1 side chain in complex with plasmepsin II--implication for drug design. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Liu J, Gluzman IY, Drew ME, Goldberg DE. 2005. The role of Plasmodium falciparum food vacuole plasmepsins. J Biol Chem 280(2): 1432-1437. Looareesuwan S, Chulay J, Canfield C, Hutchinson D. 1999. Malarone (atovaquone and proguanil hydrochloride): a review of its clinical development for treatment of malaria. Malarone Clinical Trials Study Group. Am J Trop Med Hyg 60(4): 533541. Looareesuwan S, Viravan C, Webster HK, Kyle DE, Hutchinson DB, Canfield CJ. 1996. Clinical studies of atovaquone, alone or in combination with other antimalarial drugs, for treatment of acute uncomplicated malaria in Thailand. Am J Trop Med Hyg 54(1): 62-66. Luker KE, Francis SE, Gluzman IY, Goldberg DE. 1996. Kinetic analysis of plasmepsin I and II, aspartic proteases of the Plasmodium falciparum digestive vacuole. Mol Biochem Parasitol 79(1): 71-78. Lutz R, Bujard H. 1997. Independent and tight regulation of transcriptional units in Escherichia coli via the LacR/O, the TetR/O and AraC/I1-I2 regulatory elements. Nucleic Acids Res 25(6): 1203-1210. Marquardt DW. 1963. An algorithm for least-s quares estimation of nonlinear parameters. J Soc Indust Appl Math 11(2): 431-441. Martins TM, Domingos A, Berry C, Wyatt DM. 2006. The activity and inhibition of the food vacuole plasmepsin from the rodent malaria parasite Plasmodium chabaudi . Acta Trop 97(2): 212-218. Martins TM, Novo C, do Rosario VE, Do mingos A. 2003. Aspartic proteases from Plasmodium chabaudi : a rodent model for human ma laria. Acta Trop 89(1): 1-12. Matsumoto H, Kimura T, Hamawaki T, Kumagai A, Goto T, Sano K, Hayashi Y, Kiso Y. 2001a. Design, synthesis, and biological evaluation of anti-HIV double-drugs: conjugates of HIV protease inhibitors with a reverse transcriptase inhibitor through spontaneously cleavable linkers . Bioorg Med Chem 9(6): 1589-1600.

PAGE 233

212 Matsumoto H, Matsuda T, Nakata S, Mitoguc hi T, Kimura T, Hayashi Y, Kiso Y. 2001b. Synthesis and biological ev aluation of prodrug-type anti-HIV agents: ester conjugates of carboxylic acidcontaining dipeptide HIV pr otease inhibitors and a reverse transcriptase inhibitor. Bioorg Med Chem 9(2): 417-430. Meldal M, Svendsen I, Breddam K, Auzann eau FI. 1994. Portion-mixing peptide libraries of quenched fluorogenic substrates for complete subsite mapping of endoprotease specificity. Proc Natl Acad Sci USA 91(8): 3314-3318. Meshnick SR. 2001. Artemisinin and its derivatives. In Antimalarial Chemotherapy. Mechanisms of Action, Resistance, a nd New Directions in Drug Discovery (Rosenthal PJ ed.). pp. 191-202. Huma na Press, Totowa, New Jersey. Meshnick SR, Dobson MJ. 2001. The history of antimalarial drugs. In Antimalarial Chemotherapy. Mechanisms of Action, Resi stance, and New Directions in Drug Discovery (Rosenthal PJ ed.). pp. 1526. Humana Press, Totowa, New Jersey. Meshnick S, Taylor T, Kamchonwongpaisan S. 1996. Artemisinin and the antimalarial endoperoxides: From herbal remedy to targeted chemotherapy. Microbiol Rev 60(2): 301-315. de Meyer S, Azijn H, Surleraux D, Jochmans D, Tahri A, Pauwels R, Wigerinck P, de Bthune M-P. 2005. TMC114, a novel human immunodeficiency virus type 1 protease inhibitor-resistant viruses, incl uding a broad range of clinical isolates. Antimicrob Agents Chemother 49(6): 2314-2321. Miller LH, Baruch DI, Marsh K, Doumbo, OK. 2002. The pathogenic basis of malaria. Nature 415(6872): 673-679. Mongin E, Louis C, Holt RA, Birney E, Collins FH. 2004. The Anopheles gambiae genome: an update. Trends Parasitol 20(2): 49-52. Moon RP, Tyas L, Certa U, Rupp K, Bur D, Jacquet C, Matile H, Loetscher H, Leitch FG, Kay J, Dunn BM, Berry C, Ridley RG. 1997. Expression and characterization of plasmepsin I from Plasmodium falciparum . Eur J Biochem 244(2): 552-560. Moore S, Spackman DH and Stein WH. 1958. Chromatography of amino acids on sulfonated polystyrene resins: an impr oved system. Anal Chem 30(7): 1185-1190. Moore S and Stein WH. 1948. Photometric ninhydrin method for use in the chromatography of amino acids. J Biol Chem 176(1): 367-388. Morrison JF. 1969. Kinetics of the reversible in hibition of enzyme-catalyzed reactions by tight-binding inhibitors. Bioc him Biophys Acta 185(2): 269-286. Muzammil S, Ross P, Freire E. 2003. A major role for a set of non-active site mutations in the development of HIV-1 prot ease drug resistance. 42(3): 631-638.

PAGE 234

213 Nery ED, Juliano MA, Meldal M, Svendsen I, Scharfstein J, Walmsley A, Juliano L. 1997. Characterization of the substrate spec ificity of the major cysteine protease (cruzipain) from Trypanosoma cruzi using a portion-mixing combinatorial library and fluorogenic peptides. Biochem J 323(2): 427-433. Nussenzweig RS, Vanderberg J, Most H, Orton C. 1967. Protective mmunity produced by the injection of X-irradiated irradiated sporozoites of Plamodium berghei . Nature 216(111): 160-162. Oaks SC Jr, Mitchell VS, Pearson GW, Carpenter CCJ. 1991. Bachground. The parasite. Life cycle. In Malaria: Obstacles and Opportunities (Oaks SC Jr, Mitchell VS, Pearson GW, Carpenter CCJ eds.). pp. 22-36. National Academy Press, Washington DC. Omara-Opyene AL, Moura PA, Sulsona CR, Bo nilla JA, Yowell CA, Fujioka H, Fidock DA, Dame JB. 2004. Genetic disruption of the Plasmodium falciparum digestive vacuole plasmepsins demonstrates thei r functional redundancy. J Biol Chem 279(52): 54088-54096. O’Neill PM, Bray PG, Hawley SR, Ward SA, Park BK. 1998. 4-Aminoquinolines: past, present, and future: a chemical perspe ctive. Pharmacol Ther 77(1): 29-58. Parikh S, Gut J, Istvan E, Goldberg DE, Havlir DV, Rosenthal PJ. 2005. Antimalarial activity of human immunodeficiency virus type 1 protease inhibitors. Antimicrob Agents Chemother 49(7): 2983-2985. Parikh S, Liu J, Sijwali P, Gut J, Goldbe rg DE, Rosenthal PJ. 2006. Antimalarial effects of human immunodeficiency virus type 1 pr otease inhibitors differ from those of the aspartic protease inhi bitor pepstatin. Antimicrob Agents Chemother 50(6): 2207. Pechik IV, Gustchina AE, Andreeva NS, Fedorov AA. 1989. Possible role of some groups in the structure and function of HIV-1 protease as revealed by molecular modeling studies. FEBS Lett 247(1): 118-122. Peitsch MC. 1996. ProMod and Swiss-Model: internet-based tools for automated comparative protein modelling. Biochem Soc Trans 24(1): 274-279. Philip A, Kepler JA, Johnson BH, Carroll FI. 1988. Peptide derivatives of primaquine as potential antimalarial agen ts. J Med Chem 31(4): 870-874. Plowe CV, Cortese JF, Djimde A, Nwanya nwu OC, Watkins WM, Winstanley PA, Estrada-Franco JG, Mollinedo RE, Avila JC, Cespedes JL, Carter D, Doumbo OK. 1997. Mutations in Plasmodium falciparum dihydrofolate reductase and dihydropteroate synthase and epidemiologi c patterns of pyrimethamine-sulfadoxine use and resistance. J Infect Dis 176(6): 1590-1596.

PAGE 235

214 Posner GH, Cumming JN, Woo SH, Ployprad ith P, Xie S, Shapiro TA. 1998. Orally active antimalarial 3-substituted trioxanes: new synthetic methodology and biological evaluation. J Med Chem 41(6): 940-951. Posner GH, Oh CH. 1992. Regiospecifically oxygen-18 labeled 1,2,4-trioxane: a simple chemical model system to pr obe the mechanism(s) for th e antimalarial activity of artemisinin (qinghaosu). J Am Chem Soc 114(21): 8328-8329. Posner GH, Park SB, Gonzalez L, Wang D, Cumming JN, Klinedinst D, Shapiro TA, Bachi MD. 1996. Evidence for the importance of high-valent Fe=O and of a diketone in the molecular mechanism of ac tion of antimalarial trioxane analogs of artemisinin. J Am Chem Soc 118(14): 3537-3538. Prade L. 2005a (structure released). Structure of plasmepsin II in complex with pepstatin A. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Prade L. 2005b (structure released). Structure of plasmepsin II in complex of a pepstatin analogue. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Prade L. 2005c (structure released). Structure of plasmepsin II in complex of a pepstatin analogue. In . Last Accessed: July, 2006. Ref Type: Electronic Citation. Prade L, Jones AF, Boss C, Richard-Bildstei n S, Meyer S, Binkert C, Bur D. 2005. X-ray structure of plasmepsin II complexed with a potent achiral inhibitor. J Biol Chem 280(25): 23837-23843. Price RN. 2000. Artemisinin drugs: novel anti malrial agents. Exp Opin Invest Drugs 9(8): 1815-1827. Quideau S, Lebon M, Lamidey AM. 2002. Enantiospecific synthesis of the antituberculosis marine sponge metabolit e (+)-puupehenone. The arenol oxidative activation route. Or g Lett 4(22): 3975-3978. Rathod PK, McErlean T, Lee PC. 1997. Variations in frequencies of drug resistance in Plasmodium falciparum . Proc Natl Acad Sci USA 94(17): 9389-9393. Reed MB, Sallba KJ, Caruana SR, Kirk K, Cowman AF. 2000. Pgh1 modulates sensitivity and resistance to multiple antimalarials in Plasmodium falciparum . Nature 403(6772): 906-909. Rezacova P, Brynda J, Sedlacek J, Konvalinka J, Fabry M, Horejsi M. 2005 (structure released). Crystal structure of the comple x of resistant strain of HIV-1 protease (V82A mutant) with ritonavir. In . Last Accessed: July, 2006. Ref Type: Electronic Citation.

PAGE 236

215 Richardson PL. 2002. The determination and use of optimized protease substrates in drug discovery and development. Curr Pharm Des 8(28): 2559-2581. Richie TL, Saul A. 2002. Progress and chal lenges for malaria vaccines. Nature 415(6872): 694-701. Ridley RG. 2002. Medical need, scientific opportunity and the dr ive for antimalarial drugs. Nature 415(6872): 686-693. Ridley RG, Hudson AT. 1998. Quinoline antimal arials. Exp Opin Ther Patents 8(2): 121136. Rieckmann KH, Beaudoin RL, Cassells JS, Se ll KW. 1979. Use of attenuated sporozoites in the immunization of human volunteers against falciparum malaria. Bull World Health Organ 57(Suppl 1): 261-265. Rieckmann KH, Yeo AE, Davis DR, Hutton DC, Wheatley PF, Simpson R. 1993. Recent military experience with malaria chemoprophylaxis. Med J Aust 158(7): 446-449. van Riper C III, Atkinson CT, Seed TM. 1994. Plasmodia of birds. In Parasitic protozoa, 2nd edn (Kreier JP ed.). Vol. 7, pp. 73-140. Academic Press, San Diego, New York, Boston, London, Sydney, Tokyo, Toronto. Roberts NA, Martin JA, Kinchington D, Br oadhurst AV, Craig JC, Duncan IB, Galpin SA, Handa BK, Kay J, Krohn A, Lambert RW, Merrett JH, Mills JS, Parkes KEB, Redshaw S, Ritchie AJ, Taylor DL, Th omas GJ, Machin PJ. 1990. Rational design of peptide-based HIV proteinase inhibitors. Science 248(4953): 358-361. Rodriguez M, Claparols C, Robert A, Me unier B. 2002. Alkylation of microperoxidase11 by the antimalarial drug artemisini n. Chem Biol Chem 3(11): 1147-1149. Romeo S, Dell’Agli M, Parapini S, Luca R, Galli G, Mondani M, Sparatore A, Taramelli D, Bosisio E. 2004. Plasmepsin II inhibition and antiPlasmodial activity of primaquine-statine ‘double-drugs’ . Bioorg Med Chem Lett 14(11): 2931-2934. Roper C, Pearce R, Bredenkamp B, Gumede J, Drakeley C, Mosha F, Chandramohan D, Sharp B. 2003. Antifolate antimalarial resist ance in southeast Africa: a populationbased analysis. Lancet 361(9364): 1174-1181. Roth EF Jr, Brotman DS, Vanderberg JP, Sc hulman S. 1986. Malarial pigment-dependent error in the estimation of hemoglobin content in Plasmodium falciparum -infected red cells: implications for metabolic and biochemical studies of the erythrocytic phases of malaria. Am J Trop Med Hyg 35(5): 906-911. Russelle C, Smirnova M, Clair P, Lefauconnier J-M, Chavanieu A, Calas B, Scherrmann J-M, Temsamani J. 2001. Enhanced deliv ery of doxorubicin into the brain via peptide vector-mediated strategy: saturati on kinetics and specificity. J Pharmacol Exp Ther 296(1): 124-131.

PAGE 237

216 Rudzinska MA, Trager W, Bray RS. 1965. Pi nocytotic uptake and the digestion of hemoglobin in malaria parasites. J Protozool 12(4): 563-576. St Hilarie PM, Willert M, Juliano MA, Juliano L, Meldal M. 1999. Fluorescencequenched solid phase combinatorial libraries in the characterization of cysteine protease substrate specificit y. J Comb Chem 1(6): 509-523. Salas F, Fichmann J, Lee GK, Scott MD, Ro senthal PJ. 1995. Functional expression of falcipain, a Plasmodium falciparum cysteine proteinase, supports its role as a malarial hemoglobinase. In fect Immun 63(6): 2120-2125. Salemme FR, Spurlino J, Bone R. 1997. Serendi pity meets precision: the integration of structure-based drug design and combinat orial chemistry for efficient drug discovery. Structure 5(3): 319-324. Salmon SE, Lam KS, Lebl M, Kandola A, Kh attri PS, Wade S, Patek M, Kocis P, Krchnak V, Thorpe D, Felder S. 1993. Disc overy of biologically active peptides in random libraries: solution-phase testing after staged ort hogonal release from resin beads. Proc Natl Acad Sci USA 90(24): 11708-11712. Savarino A, Cauda R, Cassone A. 2005. Aspartic proteases of Plasmodium falciparum as the target of HIV-1 protease inhi bitors. J Infect Dis 191(8): 1381-1382. Scaborough PE, Guruprasad K, Topham C, Richo GR, Conner GE, Blundell TL, Dunn BM. 1993. Exploration of subsite binding sp ecificity of human cathepsin D through kinetics and rule-based molecula r modeling. Protein Sci 2(2): 64-276. Schechter I, Berger A. 1967. On the size of the active site in protease, I. papain. Biochem Biophys Res Commun 27(2): 157-162. Schwede T, Kopp J, Guex N, Peitsch MC. 2003. SWISS-MODEL: An automated protein homology-modeling server. Nucl eic Acids Res 31(13): 3381-3385. Service RF. 1996. Combinatorial chemistry hits the drug market. Science 272(5266): 1266-1268. Service RF. 1997. High-speed materi als design. Science 277(5325): 474-475. Sherman IW. 1998. Purine and pyrimidine metabolism of asexual stages. In Malaria— Parasite Biology, Pathogenesis and Pr otection (Sherman IW ed.). pp. 177-184. ASM Press, Washington DC. Shi XP, Chen E, Yin KC, Na S, Garsky VM, Lai MT, Li YM, Platch ek M, Register RB, Sardana MK, Tang MJ, Thiebeau J, Wood T, Shafer JA, Gardell SJ. 2001. The pro domain of -secretase does not confer strict zy mogen-like properties but does assist proper folding of the protease domain. J Biol Chem 276(13): 10366-10373.

PAGE 238

217 Sijwali PS, Shenai BR, Gut J, Singh A, Rosenthal PJ. 2001. Expression and characterization of the Plasmodium falciparum haemoglobinase falcipain-3. Biochem J 360(Pt 2): 481-489. Silva AM, Lee AY, Gulnik SV, Maier P, Co llins J, Bhat TN, Collins PJ, Cachau RE, Luker KE, Gluzman IY, Francis SE, Oksm an A, Goldberg DE, Erickson JW. 1996. Structure and inhibition of plasmepsin II, a hemoglobin-degrading enzyme from Plasmodium falciparum . Proc Natl Acad Sci USA 93(19): 10034-10039. Sinden RE. 1978. Cell biology. In Rodent Malaria (Kendrick RK, Peters W eds.). pp. 85169. Academic Press, London, New York, San Francisco. Siripurkpong P, Yuvaniyama J, Wilairat P, Goldberg DE. 2002. Active site contribution to specificity of the aspart ic proteases plasmepsin I a nd II. J Biol Chem 277(43): 41009-41013. Skinner-Adams TS, McCarthy JS, Gard iner DL, Hilton PM, Andrews KT. 2004. Antiretrovirals as antimalarial ag ents. J Infect Dis 190(11): 1998-2000. Slater AFG. 1993. Choloquine: mechanism of drug action and resistance in Plasmodium falciparum . Pharmacol Ther 57(2-3): 203-235. Stemmer WPC. 1994. DNA shuffling by ra ndom fragmentation and reassembly: in vitro recombination for molecular evolutio n. Proc Natl Acad Sci USA 91(22): 1074710751. Stowers AW, Cioce V, Shimp RL, Lawson M, Hui G, Muratova O, Kaslow DC, Robinson R, Long CA, Miller LH. 2001. E fficacy of two alternate vaccines based on Plasmodium falciparum merozoite surface protein 1 in an Aotus challenge trial. Infect Immun 69(3): 1536-1546. Sullivan DJ Jr, Gluzman IY, Russell DG, Goldberg DE. 1996. On the molecular mechanism of chloroquine’s antimalarial action. Proc Natl Acad Sci USA 93(21): 11865-11870. Surleraux DLNG, Tahri A, Verschueren WG, Pille GM, de Kock HA, Jonckers TH, Peeters A, de Meyer S, Azijn H, Pauw els R, de Bthune MP, King NM, PrabuJeyabalan M, Schiffer CA, Wigerinck PB TP. 2005. Discovery and selection of TMC114, a next generation of HIV-1 proteas e inhibitor. J Med Chem 48(6): 18131822. Syafruddin D, Asih PB, Aggarwal SL, Sh ankar AH. 2003. Frequency distribution of antimalarial drug-resistant alleles among isolates of Plasmodium falciparum in Purworejo district, Central Java Provin ce, Indonesia. Am J Trop Med Hyg 69(6): 614-620.

PAGE 239

218 Takechi M, Matsuo M, Ziba C, MacHeso A, Butao D, Zungu IL, Chakanika I, Bustos MD. 2001. Therapeutic efficacy of sulphadoxi ne/pyrimethamine and susceptibility in vitro of P. falciparum isolates to sulphadoxine -pyremethamine and other antimalarial drugs in Malawian childr en. Trop Med Int Health 6(6): 429-34. Telford SR Jr. 1994. Plasmodia of reptiles. In Parasitic protozoa, 2nd edn (Kreier JP ed.). Vol. 7, pp. 1-71. Academic Press, San Diego, New York, Boston, London, Sydney, Tokyo, Toronto. Tilley L, Loria P, Foley M. 2001. Chloroquine and other quinoline antimalarials. In Antimalarial Chemotherapy. Mechanisms of Action, Resistance, and New Directions in Drug Discovery (Rosenth al PJ ed.). pp. 87-122. Humana Press, Totowa, New Jersey. Trager W, Jensen JB. 1976. Human mala ria parasites in continuous culture. Science 193(4254): 673-675. Tyas L, Gluzman I, Moon RP, Rupp K, Wes tling J, Ridley RG, Kay J, Goldberg DE, Berry C. 1999. Naturally-occurring and recombinant forms of the aspartic proteinases plasmepsins I and II from the human malaria parasite Plasmodium falciparum. FEBS Lett 454(3): 210-214. UNAIDS 2006. 2006 Report on the global AIDS epidemic. In ta/2006GlobalReport/default.asp . Last Accessed: July, 2006. Ref Type: Electronic Citation. Veerapandian B, Cooper JB, Sali A, Bl undell TL, Rosati RL, Dominy BW, Damon DB, Hoover DJ. 1992. Direct observation by X -ray analysis of the tetrahedral "intermediate" of aspartic prot einases. Protein Sci 1(3): 322-328. Veldkamp AI, van Heeswijk RP, Mulder JW, Meenhorst PL, Schreij G, van der Geest S, Lange JM, Beijnen JH, Hoetelmans RM . 2001. Steady-state pharmacokinetics of twice-daily dosing of saquinavir plus rito navir in HIV-1-infected individuals. J Acquir Immune Defic Syndr 27(4): 344-349. Vives E, Brodin P, Lebleu B. 1997. A truncated HIV-1 Tat protein ba sic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. J Biol Chem 272(25): 16010-16017. Wang D-Y, Wu Y-L. 2000. A possible an timalarial action mode of qinghaosu (artemisinin) series compounds. Alkylati on of reduced glutathione by C-centered primary radicals produced from antim alarial compound qinghaosu and 12-(2,4dimethoxyphenyl)-12-deoxoqinghaosu. Chem Commun (22):2193-2194. Waters AP, Higgins DG, McCutchan TF. 1991. Plasmodium falciparum appears to have arisen as a result of lateral transfer be tween avian and human hosts. Proc Natl Acad Sci USA 88(8): 3140-3144.

PAGE 240

219 Wellems TE, Plowe CV. 2001. Chloroquine-resis tant malaria. J In fect Dis 184(6): 770776. Westling J, Cipullo P, Hung SH, Saft H, Dame JB, Dunn BM. 1999. Active site specificity of plasmepsin II. Protein Sci 8(10): 2001-2009. Westling J, Yowell CA, Majer P, Erickson JW, Dame JB, Dunn BM. 1997. Plasmodium falciparum , P. vivax , and P. malariae : a comparison of the active site properties of plasmepsins cloned and expressed from th ree different species of the malaria parasite. Exp Parasi tol 87(3): 185-193. Whitby M, Wood G, Veenedaal JR, Ri eckmann K. 1989. Chloroquine resistant Plasmodium vivax . Lancet 2(8676): 1395. White NJ. 1998. Drug resistance in malari a. Br Med Bull 54(3): 703-715. WHO 2001. The use of antimalarial drugs: re port of an informal consultation, 13-17 November 2000. Report No. WHO/CDS/RBM/2001.33. WHO 1998. The use of artemisinin and its deri vatives as anti-malar ial drugs. Report of a joint CTD/DMP/TDR informal consultation, Geneva, 10-12 June 1998. Report No. WHO/MAL/98.1086. Wyatt DM, Berry C. 2002. Activity and inhib ition of plasmepsin IV, a new aspartic proteinase from the malaria parasite, Plasmodium falciparum . FEBS Lett 513(2-3): 159-162. Xiao H-G, Ogawa M, Yada RY. 2005. Recomb inant expression of active histo-aspartic proteinase. The 4th general meeting of th e international proteolysis society. Quebc City, Quebc, Canada. Yayon A, Cabantchik ZI, Ginsburg H. 1984. Iden tification of the acidic compartment of Plasmodium falciparum -infected human erythrocytes as the target of the antimalarial drug chloroquine. EMBO J 3(11): 2695-2700.

PAGE 241

220 BIOGRAPHICAL SKETCH Peng Liu was born on a snowing New Year’s Eve, January 28, 1979, in Dingzhou, Hebei Province, the People’s Republic of China. In September 1997, Peng started his undergraduate career at Peki ng University, Beijing, China. His major was biotechnology. Peng spent his senior year at the National Laboratory of Protein and Gene Engineering focusing on the Arabidopsis thaliana genome and proteome research. He finished his dissertation, “C onstruction of Arabidopsis Mutant Collection and Functional Gene Isolation”, and graduated with a B.S. in July 2001. In August 2001, Peng joined the IDP program of University of Florida, College of Medicine, to pursue his Ph.D. degree. In June 2002, he joined the laboratory of Dis tinguished Professor Ben M. Dunn. Peng spent four years on enzymatic characterization, subsite specificity and enzyme-inhibitor interaction studies of malarial aspartic pr oteinases under the guida nce of Professor Dunn.