Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Bioengineering coagulation factor Xa substrate specificity into Streptomyces griseus trypsin Page, Michael J. 2004

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

Item Metadata

Download

Media
831-ubc_2004-902463.pdf [ 12.35MB ]
Metadata
JSON: 831-1.0091823.json
JSON-LD: 831-1.0091823-ld.json
RDF/XML (Pretty): 831-1.0091823-rdf.xml
RDF/JSON: 831-1.0091823-rdf.json
Turtle: 831-1.0091823-turtle.txt
N-Triples: 831-1.0091823-rdf-ntriples.txt
Original Record: 831-1.0091823-source.json
Full Text
831-1.0091823-fulltext.txt
Citation
831-1.0091823.ris

Full Text

Bioengineering Coagulation Factor Xa Substrate Specificity into Streptomyces griseus Trypsin by > Michael J. Page B.Sc, Carleton University, 1998 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Biochemistry and Molecular Biology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 2004 © Michael J. Page, 2004 Abstract Extended substrate specificity is exhibited by a number of highly evolved members of the SI peptidase family, such as the vertebrate blood coagulation proteases. Dissection of this substrate specificity has been hindered by the complexity and physiological requirements of these proteases. In order to understand the mechanisms of extended substrate specificity, a bacterial trypsin-like enzyme, Streptomyces griseus trypsin (SGT), was chosen as a scaffold for the introduction of extended substrate specificity through structure-based genetic engineering. Recombinant and mutant SGT proteases were produced in a B. subtilis expression system, which constitutively secretes active protease into the extracellular medium at greater than 15 mg/L of culture. Comparison of the recombinant wild-type protease to the natively produced enzyme demonstrated near identity in enzymatic and structural properties. To begin construction of a high specificity protease, four mutants in the S1 substrate binding pocket (T190A, T190S, T190V, and T190P) were produced and examined for differences in the Arg:Lys preference. Only the T190P mutant of SGT demonstrated a significant increase in PI arginine to lysine preference - a three-fold improvement to 16:1 - with only a minor reduction in catalytic activity (kcat reduction of 25%). The 1.9 A resolution crystal structure of T190P mutant of SGT in complex with the small molecule inhibitor benzamidine was subsequently determined. The model shows that the increased preference for Arg over Lys side chains in the SI pocket is the result of the second shell residues of the SI pocket, particularly by the N-terminal residue of the protease which does not conflict with the introduced proline ring. ii Using the T190P mutant of SGT as a starting point, coagulation factor Xa (FXa) substrate specificity determinants were then introduced by additional site-directed mutagenesis. To aid in purification of the recombinant proteases a hexa-histidine tag was added to the C-terminus of the protein. Addition of the purification tag reduced the ability of the expression host to produce the enzyme (3 mg/L of culture) but simplified the purification of SGT from the culture medium. Various combinations of a two-residue loop and a number of point mutations at positions 99, 172, 174, 180, and 217 were constructed and characterized in SGT. The mutant bearing mutations at all positions except residue 217 demonstrated a moderate preference for FXa substrates as determined using chromogenic synthetic peptides. However, the kinetic properties of the mutant enzyme suggested that the 172-loop, a member of the S3/S4 substrate binding pocket, is not in a conformation similar to FXa. Addition of the Y217E mutation was designed to stabilize the loop but led to a specific protease similar to coagulation factor XIa and not FXa. These results confirm the evolutionary relationship amongst the vertebrate coagulation proteases and demonstrate the importance and flexibility of the 172-loop. Further, Na + binding, a novel property found in several coagulation proteases, is suggested to play a role in stabilization of the 172-loop and in turn played an important role in the evolution of the vertebrate coagulation cascade. iii Table of Contents Abstract ii Table of Contents iv List of Tables vii List of Figures viii List of Abbreviations x Acknowledgments xi Chapter 1. Introduction 1 1.1 Focus of Research 1 1.2 Overview 1 1.3 Nomenclature of Proteases 4 1.4 Definition of Substrate Specificity 6 1.5 Aspartic, Cysteine, Metallo-, and Threonine Proteases 7 1.6 Catalytic Mechanism of Serine Proteases 9 1.7 Blood Coagulation and Fibrinolysis 14 1.8 Substrate Specificity of Serine Proteases 17 1.9 Genetic Manipulation of Serine Proteases 20 1.10 Streptomyces griseus Trypsin 21 1.11 Statement of Hypothesis 22 1.12 Objectives and Outline 22 Chapter 2. Recombinant Protein Expression of Streptomyces griseus Trypsin in Bacillus subtilis 25 2.1 Introduction 25 2.2 Materials & Methods 28 2.2.1 Plasmids, Bacterial Strains, and Growth Conditions 28 2.2.2 DNA manipulation 28 2.2.3 Protein Purification 29 2.2.4 Kinetic Analysis , 30 2.2.5 Crystallization 30 2.3 Results and Discussion 31 iv 2.3.1 Production and Purification of Recombinant SGT 31 2.3.2 Kinetic Analysis 36 2.3.3 Crystallization and Structure Determination 37 2.3.4 Comparison of the recombinant and Pronase-derived crystal structure 40 2.3.5 Wild-type Native and Recombinant SGT Substrate Binding 41 2.4 Conclusions 42 Chapter 3. Engineering the Primary Substrate Specificity of Streptomyces griseus Trypsin 43 3.1 Introduction 43 3.2 Materials & Methods 44 3.2.1 DNA Manipulation and Protein Purification 44 3.2.2 Kinetic Analysis 45 3.2.3 Crystallization and Structure Determination 45 3.3 Results and Discussion 46 3.3.1 Production of SGT Mutants 46 3.3.2 Kinetic Analysis 46 3.3.3. Crystallization and Structure Refinement 48 3.3.4 Ca*+ binding site of B. subtilis derived SGT 50 3.3.5 T190S and the Loss of y-CH 3 51 3.3.6 T190V and the Effect of a Branched Side Chain 53 3.3.7 T190A and the Loss of y-OH 53 3.3.8 T190P 54 3.3.9 Second Shell Residues 58 3.4 Conclusions 58 Chapter 4. Engineering Coagulation Factor Xa Substrate Specificity into Streptomyces griseus Trypsin 59 4.1 Introduction 59 4.1.1 Overview 59 4.1.2 Choice of Mutations to Mimic FXa-like Specificity 62 4.2 Materials & Methods 65 4.2.1 Plasmids, Bacterial Strains, and Growth Conditions 65 4.2.2 Construction of a Hexahistidine-tagged SGT 65 4.2.3 Sequence analysis of the SI Family peptidases 67 4.2.4 DNA Manipulation 67 4.2.5 Purification of His-tagged SGT and Mutants thereof 68 4.2.6 Characterization of Substrate Specificity 69 4.2.7 Macromolecular Substrate Specificity 69 4.3 Results & Discussion 70 4.3.1 Production of His-tagged SGT 70 4.3.2 Techniques for Characterization of Substrate Specificity of Serine Proteases 71 v 4.3.3 Extended Substrate Specificity of SGT 72 4.3.4 Effect of the 99-loop on the Substrate Specificity of SGT 74 4.3.5 Mechanisms of P3 Selectivity in SI Family Peptidases 76 4.3.6 Role of the 172-loop and Residue 217 in the SI Peptidases 79 4.3.7 Additional Elements Needed for Reconstructing FXa-like Specificity 86 4.3.8 Utility of a FXa-like Protease 89 4.4 Conclusions & Future Directions 92 Chapter 5. General Discussion and Outlook 94 5.1 Substrate specificity determinants of the SI family of Serine Proteases 94 5.2 Molecular Evolution of the SI family of Serine Proteases 98 5.3 Paper, Rock, Scissors Genetic Screening of Trypsin-like Proteases 99 5.5 Future Opportunities 107 5.6 Significance of the Work 108 5.7 Conclusions 109 Appendix A: Structural Alignment of Selected SI Family Peptidases 110 Bibliography I l l vi List of Tables Table 1.1 A short list of biological processes involving proteolysis 3 Table 1.2 A short list of pathologies involving abnormal proteolysis 3 Table 1.3 Select examples of successful protein engineering of serine proteases 21 Table 2.1 Purification table for recombinant SGT (bSGT) from B. subtilis extracellular supernatant 34 Table 2.2 PI arginine to lysine preference of SGT enzymes 37 Table 2.3 Data collection and refinement statistics of wild-type recombinant SGT 39 Table 3.1 Oligonucleotides used to mutate residue 190 in the SGT gene 45 Table 3.2 ES-MS analysis of the four mutants of SGT 46 Table 3.3 PI Arginine to Lysine preference of mutant SGT enzymes 47 Table 3.4 Ki values of benzamidine for recombinant SGT and the four mutant forms 48 Table 3.5 Data collection and refinement statistics for the T190P mutant of SGT 49 Table 4.1 Mutants of SGT constructed to mimic the substrate specificity of FXa 62 Table 4.2 Oligonucleotides used to mutate the SGT gene to mimic residues found in FXa....68 Table 4.3 Steady-state kinetic parameters for the hydrolysis of a series of p-nitroanilide chromogenic substrates by the YSFMP mutant of SGT 83 Table 4.4 Steady-state kinetic parameters for the hydrolysis of a series of p-nitroanilide chromogenic substrates by the YSFMPE mutant of SGT 85 Table 4.5 Cleavage sites of proteases used in processing recombinant proteins 90 Table 5.1 Substrate specificities of SI family peptidases 96 Table 5.2 Substrate specificity determinants of SI family sub-family A peptidases 100 Table 5.3 Inhibition constants of ecotin against a variety of SI family peptidases 104 vii List of Figures Figure 1.1 Distribution of identified proteases based on catalytic type 2 Figure 1.2 Derivation of the Michaelis-Menten equation 6 Figure 1.3. Catalytic mechanism of a typical zinc metalloprotease 9 Figure 1.4 Catalytic triad of serine proteases 10 Figure 1.5 Three dimensional structure of a typical serine protease 12 Figure 1.6 Catalytic mechanism of a serine protease 13 Figure 1.7 Overview of vertebrate blood coagulation 15 Figure 1.8 Protein domains of the vertebrate blood coagulation proteases 16 Figure 1.9 Schecter & Berger nomenclature of protease specificity 18 Figure 2.1 Plasmid map of SGT gene cloned into pWB980 33 Figure 2.2. ES-MS spectrum of purified recombinant SGT 35 Figure 2.3 SDS-PAGE of purified recombinant SGT 36 Figure 2.4 Ramachandran plot of the crystal structure of recombinant wild-type SGT 38 Figure 2.5 Superimposition of the C a traces of native and recombinant wild-type SGT 41 Figure 3.1 Ramachandran plot of the crystal structure of the T190P mutant of SGT 50 Figure 3.2 Comparison of the Ca 2 + binding site in SGT enzymes 52 Figure 3.3 Comparison of the SI binding pocket in SGT enzymes 56 Figure 4.1 Residues involved in the extended substrate specificity of coagulation proteases .63 Figure 4.2 Plasmid construction for the production of recombinant His-tagged SGT 66 Figure 4.3 Purification of a typical His-tagged mutant of SGT 71 Figure 4.4 Substrates used to characterize mutants of SGT with altered substrate specificity 73 Figure 4.5 Normalized kcat /K m values for the T190P, LP and YP mutants of SGT 75 Figure 4.6 S3 & S4 binding pockets of FXa 77 Figure 4.7 S3 and S4 binding pockets of FVIIa 78 Figure 4.8 Conformation of the 172-loop in SI peptidases 80 Figure 4.9 Normalized KJKm values for the YFP and YSFP mutants of SGT 82 Figure 4.10 Na+-binding site in thrombin 88 Figure 4.11 Prothrombin processing by mutants of SGT 91 viii Figure 5.1 Simplified phylogenetic tree of the SI family of peptidases 95 Figure 5.2 Theory behind the Paper, Rocks, Scissors genetic screen 103 Figure 5.3 Dimeric structure of ecotin 104 Figure 5.4 Structure of GFP 106 ix List of Abbreviations A Angstrom unit (1 A = 0.1 run) AMC Aminomethylcoumarin aPC Activated protein C B Crystallographic thermal factor (A2) Bz Benzoyl dNTP Dinucleotidetriphosphate Fo, Fc Observed and calculated structure factors FX Coagulation factor X (similar for other coagulation proteases) FXa Activated coagulation factor X (similar for other coagulation proteases) kcat Catalytic constant Km Michaelis constant LB Luria - Bertani broth NMWL Nominal molecular weight limit PCR Polymerase Chain Reaction Pip Pipecoyl pNA para-nitroanilide r.m.s. Root mean squared Rfree R factors based test set of excluded reflections Rmerge Shk! Si |li(hkl) - <I>(hkl)(| / Shk! Si Ii(hkl) R-cryst £ ||F0bs| - |Fcalc|| / 2 |F0bs| S.D. Standard deviation at 95% confidence interval SDS-PAGE Sodium dodecyl sulfate - polyacrylamide gel electrophoresis Tos Tosyl Tris-HCI Tris(hydroxymethyl)aminomethane hydrochloride Acknowledgments I a m indebted to Ross and Jef f for countless hours o f good times, their use o f " N u m b e r 1", and their abi l i ty to define new verbs. I w o u l d especial ly l i ke to thank S u i - L a m W o n g for p rov id ing the essential components o f the expression system. I appreciate m y parents for their constant support both emot ional ly and f inancia l ly throughout m y l i fe . I thank m y friends, for their help, advice, time and patience. I especia l ly thank Iain, M a r t y , I smai l , A n g u s , M i k e K . , and M a r k and w i s h them the best i n a l l endeavors. A special thanks to T a n y a for a l l o f her helpful comments . T h i s thesis is dedicated to m y be loved D o e x i Chapter 1. Introduction 1.1 Focus of Research How do you create a highly efficient and highly specific enzyme? Over several million years, nature has evolved many enzymes that possess high degrees of specificities. Using methods of genetic manipulation, properties of enzymes can be altered, including their substrate specificity. The present dissertation involves taking a primitive bacterial enzyme and adding substrate specificity where little existed previously. By doing so, we will learn about how molecular evolution has produced substrate specificity in one family of enzymes, the serine proteases. I will use a human protease involved in blood coagulation as our guide and then extend the concepts learned to a system that could produce a variety of other specificities. To begin, I will discuss what is a protease, what do they do and why they are important. From there, I will describe the system and the methods applied to generate a specific protease, what has been done by others and then by the author. Lastly, I will discuss the opportunities that may result from this work. 1.2 Overview Proteolytic enzymes play a diverse number of roles in a variety of essential biological processes, both as non-specific catalysts of protein degradation and as highly specific agents that control physiological events. Hydrolysis of a peptide bond is the key function that proteases fulfill in vivo, and this may result in the activation or destruction of its substrate. Numerous biological processes involving proteolytic activity have been characterized and a wealth of information has been gathered on the five major catalytic classes of these enzymes (Figure 1.1, Table 1.1). Roughly 2 % of all genes in most organisms are proteases, second 1 only to transcription factors. Hence , the importance o f these types o f enzymes i n b io log ica l and commerc ia l settings cannot be understated. Unknown Figure 1.1 D i s t r ibu t ion o f identif ied proteases based on catalytic type. Numerous pa thologica l condit ions are the result o f excessive or insuff icient proteolytic act ivi ty (Table 1.2). These c l i n i ca l situations can arise f rom genetic defects i n the proteases themselves, their natural substrates or their natural inhibi tors . A number o f academic laboratories and pharmaceutical companies are devoted to the product ion o f therapeutic products, such as smal l molecule inhibitors or recombinant proteins, to m i n i m i z e the effects o f protease-related pathologies [1,2]. G i v e n the large number o f c lose ly related proteases found in man, the design o f potent inhibitors w i th m i n i m a l cross react ivi ty is an arduous task. Deta i l ed informat ion on the active site geometry and electronic conf igurat ion o f the protease is a requirement for proper inhibi tor design. These studies have been hampered 2 Biological Process Proteolytic Event Ref. Apoptosis Control of physiological cell death [3] Blood Coagulation Proteolytic cascades of clot formation, fibrinolysis [4] Blood Pressure Renin-angiotensin and kallikrein-kinin systems [5] Digestion Breakdown of protein into tri- and dipeptides; liberation of hormones promoting digestion [6] Fertilization Sperm-Egg interaction, ovulation, ovum implantation and parturition [7] Immunity Complement activation, antigen presentation, chemokine and chemotaxin activation; [8] Intracellular protein level Proteasome-Ubiquitin system [9] Protein Processing Zymogen activation and protein sorting [10] Tissue Remodeling Turnover and repair of the extracellular environment [11] Table 1.1 A short list of biological processes involving proteolysis. Pathology Proteolytic Event Ref. Alzheimer's Disease Processing of amyloid precursor protein [12] Cancer Regulation of apoptosis, tumor growth and invasion [3] Chronic Inflammation Excessive activation of pro-inflammatory cytokines [13] Hemophilia Insufficient levels of coagulation factor activity and slowed clot formation [14] Myocardial Infarction Unregulated coagulation leading to restricted blood flow [15] Parasite Infection Regulation of the parasitic life-cycle [16] Viral Replication Processing of viral coat proteins and other essential replication machinery required for viral infection [17] Table 1.2 A short list of pathologies involving abnormal proteolysis. 3 by the abi l i ty to produce large quantities o f the target enzyme and detailed three d imens iona l structural models o f them. Industrial usage o f proteolyt ic enzymes is widespread and c o m m e r c i a l l y important. F o r example , the product ion o f the protease subt i l i s in for use i n detergents is on the scale o f tons per annum and accounts for 4 0 % o f enzyme sales w o r l d w i d e . Other industrial applicat ions o f proteases include the product ion o f food stuffs, leather, pharmaceuticals , diagnostic reagents, waste management, and si lver recovery [18,19]. Recent advances i n molecular b i o l o g y technology have a l l owed for the development o f proteases wi th improved properties for industr ial use. Some successful examples o f protein engineering include thermostabil i ty, resistance to oxida t ion , alteration o f p H opt ima, and increased catalytic eff ic iency [20]. H o w e v e r , the use o f h igh ly specific proteases for site specif ic proteolysis has for the most part been l imi t ed to academic endeavors due to the diff icult ies associated w i t h p roduc ing h igh puri ty enzymes f rom the complex b i o l o g i c a l systems f rom w h i c h they derive. In the present dissertation, a nove l system has been developed to produce a recombinant protease and the substrate specif ic i ty o f the enzyme altered. T h e results contribute to our understanding o f the substrate specif ici ty determinants o f the t ryps in- l ike f ami ly o f serine proteases. Furthermore, a f ramework for des igning nove l specifici t ies not observed i n nature is then presented based on the work . 1.3 Nomenclature of Proteases Pr io r to a literature rev iew, a few words must be ment ioned on the nomenclature used to describe proteolytic enzymes. T h e International U n i o n o f B i o c h e m i c a l and M o l e c u l a r B i o l o g y Nomencla ture Commi t tee ( I U B M B - N C ) denotes hydrolases acting o n peptide bonds 4 as E . C . 3 . 4 , yet on ly a smal l fraction o f the k n o w n proteases have been g iven fu l l c lass if icat ion. G i v e n the wide dis tr ibut ion o f proteolytic enzymes and their his tor ical s ignif icance, a variety o f c o m m o n names are used. Papa in , c o m m o n l y used to tenderize meat, owes its name to the papaya plant f rom w h i c h it is der ived. Other proteases are named based on the phys io log ica l process i n w h i c h they were first described, such as coagulat ion factor X w h i c h is i nvo l ve d i n vertebrate b l o o d coagulat ion. Unfortunately, a significant number o f proteases owe their name to hav ing s imi la r b iochemica l properties yet are i n v o l v e d i n disparate b i o l o g i c a l processes or hav ing different catalytic mechanisms. F o r example , two o f the many proteases found i n the human l iver , cathepsin B and cathepsin D , are named s imi la r ly but be long to the cysteine and aspartic protease fami ly , respectively. Thus , the nomenclature for each particular protease has become rather mudd led w i t h convent ion . A variety o f terms have been used to refer to enzymes that catalyze the hydrolys is o f a peptide bond . The term peptidase has been suggested by the I U B M B - N C as a preferred alternative to protease [21]. Peptidase refers more easi ly to both exopeptidases (enzymes that cleave peptide bonds at the ends o f a polypept ide chain), endopeptidases (those that cleave peptide bonds i n the midd le o f polypept ide chains) , and oligopeptidases (those that cleave on ly short polypeptides) . Exopept idases are further subdivided into aminopeptidases (c leaving at the N- terminus) and carboxypeptidases (c leaving at the C-terminus) . Other terms found i n the literature inc lude proteinase and proteolyt ic enzyme. In the present thesis, the terms protease and peptidase w i l l be used interchangeably to describe the enzymes studied in the research; a l l o f w h i c h are endopeptidases. 5 1.4 Definition of Substrate Specificity E n z y m e kinet ics are t radi t ional ly defined by M i c h a e l i s - M e n t e n kinet ics w h i c h apply a steady state assumption to the react ion (Figure 1.2). Substrate specif ic i ty is t radi t ional ly defined by the ratio o f the m a x i m u m reaction ve loc i ty cata lyzed by the enzyme per unit t ime (kcat) to the M i c h a e l i s - M e n t e n constant (K m ) . kc a l is equal to the m a x i m a l react ion ve loc i ty (Vmax) d iv ided b y the total amount o f enzyme i n the reaction. In theory, both constants (kcat and K m ) are l i n k e d mathematical ly as K m is dependent on the rates o f each step i n the reaction ( K m = (k.i + k 2 ) / ki). In the present study, the tradit ional def ini t ion o f substrate specif ic i ty w i l l be appl ied (S = k^/Km). E + s ES —> E + p Initial Rate of Reaction: v = d[P] / dt = k [ES] Steady state assumption: d[ES] / dt = 0 Rate of ES formation = Rate of ES breakdown k,[El[Sl = k.^ES] +k 2[ES] B [ES] = k, [E][S] / (k, + k 2 ) (1) Fraction of enzyme in ES: F = (ES] / ([E] + [ES]) (2) (1) + (2) yields: F= [S] / {((k 1 + k 2 ) / k, ) + [S]} (3) Initial rate of reaction: v 0 = v m a x F (4) (3) + (4) yields: v 0 = v m a x [S] / «(k., + k 2 ) / k,) + [S]} (5) MichaelisConstant: K = ( k 1 + k 2 ) / k 1 (6) (valid when k , » k2) (5)+ (6) v0 = v m a x [S] / (K +[S]) Figure 1.2 Der iva t ion o f the M i c h a e l i s - M e n t e n equation. F o r a typ ica l enzyme catalyzed reaction (A) the rate o f change o f the enzyme (E) i n the enzyme-substrate complex (ES) is assumed to be constant throughout the reaction. T h e M i c h a e l i s -M e n t e n equation can then be der ived (B) by in t roducing the M i c h a e l i s constant (K m ) , w h i c h is on ly true i f the rate o f E S formation (ki) is m u c h larger than the rate o f product format ion (k2). 6 1.5 Aspartic, Cysteine, Metallo-, and Threonine Proteases His to r i ca l ly , four mechanist ic classes o f proteolytic enzymes have been recognized based on their catalytic mechanism - aspartic, cysteine, metal lo- , and serine proteases. W i t h the advent o f who le genome sequencing this c lass i f icat ion system has become inadequate as the variety o f catalytic mechanisms ident i f ied i n nature has expanded rapidly . A t present, nearly 18,000 gene sequences for peptidases have been identif ied and over 2000 peptidases have been characterized. Barrett has devised a c lass i f icat ion scheme based o n statistically significant s imilar i t ies i n sequence and structure o f a l l k n o w n proteolytic enzymes, and terms this database M E R O P S (www.merops .ac .uk) [22,23]. Th i s system divides a l l k n o w n proteases into 40 clans and over 169 sub-families, not i n c l u d i n g a group o f putative proteases o f u n k n o w n mechanism. O n l y the four historic classif icat ions o f proteolyt ic enzymes w i l l be ment ioned here. H y d r o l y s i s o f a polypept ide backbone requires three key mechanist ic hurdles to be overcome for efficient catalysis to proceed. Peptide bonds are par t icular ly stable due to the electron resonance between the amide nitrogen and carbonyl group o f the bond . B y the use o f a general ac id , proteases overcome this partial double bond character through the generation o f a negat ively charged tetrahedral intermediate that is s tabi l ized by the active site. Secondly , water is a poor nucleophi le and must be activated, typ ica l ly v i a a general base. Las t ly , amines are poor l eav ing groups and must be expel led f rom the active site pr ior to comple t ion o f the catalytic cyc l e [24]. Proteases accompl i sh these tasks eff iciently and increase the rate o f reaction ~ 1 0 1 0 - f o l d over the uncatalyzed reaction. M o r e o v e r , proteases can catalyze s imi la r reactions i n the hydrolys is o f amides, esters, ani l ides, and thioesters. 7 Aspar t ic proteases and metalloproteases catalyze the hydro lys i s o f the polypept ide backbone through activation o f a water molecule . Metal loproteases comprise the second largest f ami ly o f proteases k n o w n i n nature and typ ica l ly u t i l ize a z inc i o n i n their active site; however , cobalt or manganese are also found (Figure 1.3). In many metalloproteases, a s ingle metal i o n is u t i l i zed i n the catalysis; however two metal ions act ing cocata ly t ica l ly are typ ica l ly found i n proteases conta ining cobalt or manganese. Three amino acid side chains, typ ica l ly H i s , G l u , A s p or L y s residues, are i n v o l v e d i n the co-ordinat ion o f the metal i o n and at least one other residue is required for catalysis [25]. T h e catalytic residue is typ ica l ly G l u i n many metallopeptidases but alternatives exist. A c t i v a t i o n o f the water molecu le i n the aspartic f ami ly o f proteases is the result a pair o f A s p residues that co-ordinate the activated water molecu le [26]. In contrast to nuc leophi l i c attack o f the amide backbone by an activated water molecule , cysteine, serine, and threonine peptidases u t i l ize an amino acid side chain . Cys te ine proteases compromise the third largest f ami ly o f k n o w n peptidases and employ a nuc leophi l i c su l fhydryl f rom a cysteine residue to catalyze the hydro lys i s o f an amide bond [27]. A l t h o u g h less abundant i n nature, threonine peptidases are deeply i n v o l v e d i n the k e y b i o l o g i c a l process o f intracel lular degradation o f polypeptides by the proteasome [28]. T h e overa l l catalytic mechan i sm o f both o f these famil ies o f protease is more s imi lar to that found i n serine proteases, where a nucleophi le as w e l l as a proton donor is required for catalysis. T h e proton donor i n a l l cysteine and threonine peptidases w h i c h have been identif ied is a H i s residue, w h i c h is also true o f a l l k n o w n serine proteases. 8 Figure 1.3. Cata ly t ic mechanism o f a typ ica l z inc metalloprotease. T h e z inc i o n serves to polar ize the carbonyl group o f the substrate as w e l l as facilitate the deprotonation o f the water molecule . Several hydrogen bond ing partners stabil ize the intermediates but are not inc luded i n the diagram for s impl ic i ty . 1.6 Catalytic Mechanism of Serine Proteases N e a r l y a third o f a l l k n o w n proteases are c lass i f ied i n the serine protease f ami ly o f enzymes. T h e f ami ly name stems f rom the nuc leophi l i c serine residue i n the active site o f the enzyme, and the catalytic potency o f this residue is dependent on the A s p - S e r - H i s charge relay system or catalytic triad w h i c h was o r ig ina l ly proposed by B l o w over 30 years ago (Figure 1.4) [29]. These three residues are found i n an ident ical structural pos i t ion i n four different three-dimensional protein folds that catalyze the hydrolys is o f peptide bonds, 9 suggesting four distinct evolut ionary or igins . C o m m o n examples o f these folds are represented by chymot ryps in , subt i l is in , carboxypeptidase Y , and C l p protease. A number o f other enzyme famil ies , i nc lud ing asparaginases, esterases, acylases, and P-lactamases, u t i l ize the A s p - S e r - H i s catalytic triad or variants to generate a strong nucleophi le and promote catalysis [30]. F o r the remainder o f the introduct ion, I w i l l l i m i t the d iscuss ion to the chymot ryps in f ami ly (S I peptidase fami ly) o f serine proteases, w h i c h includes trypsins and elastases. M o r e o v e r , I w i l l u t i l i ze the chymot ryps in number ing system suggested by B l o w to refer to a part icular amino ac id residue. It must be noted that many o f the concepts discussed i n relat ion to chymot ryps in - l ike proteases apply s imi l a r ly to other types o f proteases. His57 His57 / Ser195 / Ser195 Asp 102 CHJ Asp102 — C H , Figure 1.4 Cata lyt ic triad o f serine proteases. T h e A s p , H i s , and Ser combina t ion is found i n serine proteases and other enzyme famil ies that require a nuc leophi l i c serine side chain . Cent ra l to the catalytic triad is the existence o f a hydrogen bond between residue A s p l 0 2 and H i s 5 7 , w h i c h facilitates the abstraction o f the proton f rom S e r l 9 5 and generates a potent nucleophi le . Some controversy exists over whether this hydrogen bond can be described as a l o w barrier hydrogen bond ( L B H B ) , an instance where the p K values between 10 the donor and acceptor are matched. Rejec t ion o f the L B H B theory ma in ly stems f rom the argument that it w o u l d provide no signif icant improvement to catalytic rate enhancement [31,32]. Increasing experimental and theoretical data are supporting this theory and the debate continues [33,34]. S tab i l iza t ion o f the catalytic triad is mediated through a network o f addi t ional hydrogen bonds p rov ided by several h igh ly conserved amino ac id residues surrounding the triad, par t icular ly A l a 5 6 and Ser214 i n the chymot ryps in fami ly o f serine proteases. Signif icant effort has been p laced i n the development o f sma l l molecu le compounds that m i m i c the act ivi ty the catalytic tr iad, but have met w i t h l im i t ed success due to the complex i ty o f the chemistry i n v o l v e d to generate the nuc leophi l i c serine. A c t i v a t i o n o f chymot ryps in - l ike serine proteases requires proteolyt ic processing o f an inact ive zymogen precursor protein. T h i s cleavage occurs at the ident ical pos i t ion i n a l l k n o w n members o f the fami ly : between residues 15 and 16 [24]. The newly created N -terminus produces a conformat ional change i n the enzyme and stabilizes the oxyan ion hole and substrate b ind ing site through formation o f an electrostatic interaction w i t h A s p l 9 4 [29]. T w o P-barrel domains, each formed b y six anti-parallel p-strands, and a C- te rmina l a -he l ix comprise the mature form o f the enzyme (Figure 1.5). B o t h the catalytic residues and substrate b ind ing site l ie i n the cleft between the P-barrel domains , and enzyme-substrate interactions occur w i th both domains. A m i n i m u m o f three disulphide bonds is required to stabil ize the overa l l structure; however f ive or six are c o m m o n l y found i n the f ami ly o f enzymes. 11 Figure 1.5 Three d imens iona l structure o f a typ ica l serine protease o f the S I f a m i l y o f peptidases. Componen ts o f the catalytic triad are shown i n stick f o r m and are located i n between the two (3-barrel domains. F igure 1.6 depicts the generally accepted mechanism o f serine protease catalyzed hydrolys is o f a peptide bond [24]. Ini t ia l ly , the h y d r o x y l o x y g e n o f Ser 195 attacks the carbonyl o f the peptide substrate as a result o f Hi s57 i n the catalytic t r iad act ing as a general base (Steps I and II). T h e oxyan ion tetrahedral intermediate is s tabi l ized by the backbone atoms o f G l y l 9 3 (not depicted) and S e r l 9 5 that generate a pos i t ive ly charged pocket w i t h i n the active site (Step III). Co l l apse o f the tetrahedral intermediate generates the acy l -enyzme intermediate and s tabi l izat ion o f the newly created N- te rminus is mediated b y H i s 5 7 (Steps I V and V ) . Ev idence for the existence o f the acyl-enzyme intermediate was p rov ided i n 1954 by Hart ley and K i l b e y [35]. In these in i t i a l experiments a pre-steady state burst o f product 12 Asp102 C H . II His57 / Ser195 \ / CH, \ / " N — C / ^ H O His57 / H,C Ser195 \ w7 °. a-/ Asp102 CM; N — C ' l / V H O Formation of the Michaelis Complex His 57 / H,C Ser195 \ / CH, o. a---" Asp102—CH, H O VI Acyl-enzyme Intermediate His57 „ J Ser19S V\ \ Asp102 C H , O C, 111 His57 / H,C Ser195 \ / CH, o. a-Asp102 C H , rr. Tetrahcdral Intermediate V I I / Asp102 C H ; His57 / H 2C a--' Sen 95 \ . © _ CHj V _ e Tetrahcdral Intermediate IV His57 / Asp102—CH, Serf 95 \ CH, , / H O H B O VIII Asp102 C H , His57 / H , C Serf 95 \ ' * - H - O ' C H J a--' o—c H O Free C-terminus generated Figure 1.6 Cata lyt ic mechan i sm o f a serine protease. Format ion o f a tetrahedral intermediate is the key conformat ional step i n the acyla t ion and deacylat ion reactions. 13 correct ly identif ied that a bond to a h y d r o x y l moie ty w i t h i n chymot ryps in was i n v o l v e d i n the reaction mechanism. In the second ha l f o f the mechanism, a water molecu le displaces the free polypept ide fragment and attacks the acyl -enzyme intermediate (Step V I ) . A g a i n , the oxyan ion hole stabilizes the second tetrahedral intermediate o f the pathway and col lapse o f this intermediate liberates a new C-terminus . 1.7 Blood Coagulation and Fibrinolysis Vertebrate b l o o d coagulat ion and f ibr inolys is can serve as a useful paradigm for the study o f proteolysis i n a b i o l o g i c a l setting. T h e process serves as a mode l for pathologies associated w i t h improper proteolysis , molecula r evo lu t ion through gene dupl ica t ion and divergence, as w e l l as understanding molecular recogni t ion and substrate specif ici ty. In i t ia l ly recognized as a cascade o f events that leads to ampl i f ica t ion and rate enhancement, the feedback pathways o f the c lo t t ing cascade have on ly recently become elucidated. A t the site o f an injury that leads to disrupt ion o f the integrity o f a b l o o d vessel , a rapid and specific response must be employed to prevent excessive b l o o d loss and to restrict bacterial infect ion [36]. In vivo, the format ion o f a f ibr in c lo t requires a m i n i m u m o f f ive proteases: coagulat ion factor X I ( F X I ) , coagulat ion factor I X ( F I X ) , coagulat ion factor V U ( F V I I ) , coagulat ion factor X ( F X ) , and prothrombin . These enzymes circulate i n the b l o o d stream at l o w concentrations i n inact ive, zymogen forms. A c t i v a t i o n o f these zymogens requires the site-specific proteolysis o f one or more peptide bonds, l iberat ing a free N -terminus and promot ing proteolytic act ivi ty (active forms o f the enzymes are denoted wi th a lower case "a" , such as F X a ) [37]. L o c a l i z a t i o n o f the proteases to membrane surfaces at the site o f injury is p rov ided by three co-factors: activated coagulat ion factor V ( F V a ) , activated 14 coagula t ion factor V I I I (FVII Ia ) and tissue factor (TF) [38]. In combina t ion w i t h a phospho l ip id bi layer these co-factors promote the rate o f clot format ion ~ 1 0 6 - f o l d (Figure 1.7). Down-regu la t ion o f the pathway is mediated i n part by another protease, activated protein C (aPC) , w h i c h cleaves two o f the co-factors, F V a and F V I I I a , at specif ic posit ions i n the prote in and inactivates them [39]. T h r o m b i n plays a p ivo ta l role i n the process as it activates protein C , F V , and FVTII as w e l l as a number o f other s igna l ing proteins that recrui t cells and proteins to the site o f damage [40]. F i g u r e 1.7 O v e r v i e w o f vertebrate b l o o d coagulat ion. F i v e proteases are i n v o l v e d i n the formation o f a c ross- l inked f ibr in b lood clot (factors X I a , I X a , V i l a , X a , and thrombin). Three accessory proteins (factors V i l l a and V a , and tissue factor (TF)) are i n v o l v e d i n co- loca l iza t ion o n a phosphol ip id surface ( P L ) and catalytic rate enhancement o f the entire process. Prote in C ( P C ) is activated ( aPC) by thrombin and leads to inh ib i t ion o f the process by c leav ing the co-factors. 15 B i o c h e m i c a l characterization o f the pur i f ied components o f the b l o o d coagulat ion pathway in vitro has shown that each protease i n the pathway prefers to recognize and cleave a particular sequence o f amino acids. Substrate specif ici ty o f these proteases, however , is not extremely strict o w i n g to the requirement for the process to occur rapid ly [41,42]. Substrate specif ici ty is a combina t ion o f addi t ional regions o f the enzyme that contribute to molecula r recogni t ion as w e l l as to the l oca l architecture o f the substrate b i n d i n g site i n the active site o f the protease. Interactions between proteins are p rov ided by addi t ional protein domains i n the polypept ide sequence (Figure 1.8) [43]. Conf igura t ion o f these domains provides some clues to the evolut ionary history o f vertebrate b l o o d coagulat ion. Prothrombin FVII ©000©^ ©©©©©^ Protein C 00000T£RoI] Key: Apple Domain EGF Domain Propeptide Cf^j) Fibronection Type I Domain ^ ) Gia Domain ^ j n ^ Fibronectin Type II Domain Kringle Domain I PROT i Protease Domain Signal Peptide Figure 1.8 Prote in domains o f the vertebrate b l o o d coagulat ion proteases. 16 Through gene sequence analysis, gene dupl ica t ion and divergence o f the coagulat ion proteases probably occurred pr ior to the emergence o f the vertebrate l ineage. C o m p a r i s o n o f the publ i shed gene sequences f rom a number o f organisms ranging f rom jawless vertebrates to humans shows that the domain organizat ion o f the coagulat ion machinery is h igh ly conserved i n a l l vertebrates [44]. S l igh t variations are k n o w n to exist, however , i nc lud ing the absence o f the contact system ( F X I , F X I I , and ka l l ik re in ) i n f ish. Dool i t t l e has proposed that the format ion o f the core o f the pathway occurred roughly 450 m i l l i o n years ago [43]. In the intervening t ime, a significant amount o f molecula r evo lu t ion has taken place resul t ing i n numerous changes to the gene sequence thereby result ing i n paral le l op t imiza t ion o f the relevant proteins for their phys io log i ca l role. 1.8 Substrate Specificity of Serine Proteases H y d r o l y s i s o f a polypept ide cha in requires proper recogni t ion, orientation and b ind ing o f the polypept ide backbone. Thus , the residues adjacent to the scissi le bond have a s ignif icant impact on the rate o f hydro lys is . N e a r l y 30 years ago, Schecter and Berger described the substrate-protease interaction and their system has been adopted i n the literature (Figure 1.9)[45]. In this mode l the scissi le peptide bond is surrounded by subsites o n the protease. Substrate amino acids are termed P (for peptide) and the subsites o f the protease that interact w i th them are ca l led S (for subsite). Substrate residues extending towards the N -terminus o f the substrate are numbered P 2 , P 3 , P 4 and so forth. Converse ly , substrate residues extending towards the C-terminus are labeled P 2 ' , P 3 \ P 4 ' and onwards. Reg ions i n the protease are numbered according to the substrate. F o r example , the P I residue is bound i n the S I pocket. Theore t ica l ly , a large amount o f var ia t ion i n substrate specif ici ty can result f rom 17 the 20 possibi l i t ies o f amino ac id side chains, and a w ide divers i ty o f specif ic i ty is observed i n the nature. Protease Scissile Bond Figure 1.9 Schecter & Berger nomenclature o f protease specif ici ty. B r o a d l y specific proteases recognize and act at a site dictated by a single amino ac id i n a polypept ide chain whereas h igh ly specific proteases recognize a short m o t i f consis t ing o f three to eight amino ac id residues. The b i o l o g i c a l process i n w h i c h the protease is i n v o l v e d dictates the leve l o f specif ic i ty . Diges t ive enzymes found i n the gut, such as t rypsin and chymot ryps in , recognize and cleave polypeptides based on the presence o f a s ingle type o f amino ac id ( A r g / L y s and Phe /Trp /Tyr , respect ively) . Hence , a polypept ide substrate w o u l d typ ica l ly be degraded into mul t ip le fragments for further processing and absorption. In contrast, a number o f b i o l o g i c a l processes require more specific proteolysis . A s ment ioned previous ly , the vertebrate b l o o d clot t ing cascade relies on the specific cleavage o f each member o f the pathway to funct ion properly. A number o f other b i o l o g i c a l processes require s imi la r levels o f specif ici ty, par t icular ly when used for s ignal ing purposes such as hormone and chemokine act ivation [46]. H o w e v e r , a trade of f exists between the l eve l o f substrate specif ici ty and the catalytic eff ic iency o f the enzyme. 18 By demonstrating a h igh degree o f select ivi ty, the preferred substrate has a s low rate of association wi th the enzyme. In turn, this generates a decreased catalytic rate relat ive to non-specif ic enzymes. In vivo such a scenario is not often preferred and alternate mechanisms of substrate specif ici ty are employed . A s ment ioned i n the discuss ion o f the b l o o d coagulat ion system, addi t ional protein domains are associated wi th a protease doma in to a id p ro te in - . protein interactions. Thus , phys io logy has p laced a barrier o n the l eve l o f specif ic i ty that a protease might possess. Proteases o f the b l o o d coagulat ion system display a marked preference for certain amino ac id side chains i n the P I to P 4 posit ions and hydro lyze them rapid ly . Al l coagulat ion proteases have t rypsin- l ike specific at the p r imary ( P I ) posi t ion and prefer to hydro lyze peptide bonds on the C- te rmina l side o f A r g or L y s residues [47]. Ex tended substrate specif ici ty exists i n a l l coagulat ion proteases i n the S 2 to S4 b i n d i n g pockets. O n the basis o f their specif ici ty, both F X a and thrombin are w i d e l y used for the site-specific cleavage o f recombinant proteins after I l e - G l u - G l y - A r g and L e u - V a l - P r o - A r g sequences i n a polypeptide chain , respect ively [48]. S i m i l a r sequences to those preferred are k n o w n to be h y d ro l y zed i n vivo by the two proteases. K i n e t i c analysis o f these proteases has revealed that both enzymes can effectively hydro lyze other sequences o f amino acids [41,49-51]. F o r example , F X a was in i t i a l ly thought to have a strict preference for a smal l amino ac id side chains at P 2 ( G l y ) . H o w e v e r , large residues (Trp, Phe) at this pos i t ion are preferred in vitro. T h r o m b i n can be genet ical ly manipulated to prefer one o f its two cleavage sites, w h i c h are different i n sequence and structure [52,53]. B a s e d on these discrepancies, it seems possible to engineer at least this l eve l o f substrate specif ic i ty into a broadly specific t rypsin- l ike protease. 19 1.9 Genetic Manipulation of Serine Proteases Great strides i n b iotechnology have been made i n the past decade that a l low for the design o f enzymes w i t h desirable properties. Examples o f successful modif icat ions made through protein engineering include increased stability or act ivi ty at the extremes o f temperatures or p H , resistance to oxida t ion , and stabili ty i n non-aqueous environments [20]. M e t h o d s to introduce these properties i nvo lve mutagenesis o f a target enzyme either through structure based design or by random mutagenesis combined w i t h some form o f genetic selection [54]. A rat ional design strategy requires significant amounts o f informat ion, par t icular ly three d imens iona l structures o f the in i t i a l enzyme as w e l l as the knowledge o f the regions that w o u l d be i nvo lved . Converse ly , a randomized mutagenesis procedure combined w i t h selection or screening requires no informat ion o f the sequence, structure or mechan i sm and has been w i d e l y adopted for the alteration o f b iochemica l properties o f enzymes [55]. Centra l to a l l forms o f protein engineering is the creation o f a nove l protein by adding a nove l funct ion that was not possessed by the target protein. Crea t ion o f a h igh ly specific protease suffers f rom the pract ical diff icult ies associated wi th p roduc ing a k ine t i ca l ly worse enzyme. A s ment ioned previous ly , for a protease to exhib i t a h igh degree o f substrate specif ic i ty , some compensat ion i n catalytic ef f ic iency must be made. A number o f studies have shown that it is possible to swi tch substrate specifici t ies amongst disparate members o f the protease fami ly . F o r example , Heds t rom demonstrated the convers ion o f a t rypsin- l ike enzyme, w h i c h prefers P I A r g / L y s residues, into a chymot ryps in -l i ke enzyme that prefers Phe /T rp /Ty r at P I [56-59]. The change i n specif ici ty required mutagenesis o f three surface loops i n the enzyme as w e l l as a number o f other point mutations. Importantly, the regions changed do not contact the substrate direct ly and the 20 result ing enzyme is inefficient at ca ta lyz ing the hydro lys i s o f peptide bonds. These observations demonstrate the inherent complex i ty o f des igning improved substrate specif ic i ty in the chymot ryps in fami ly . Other modif icat ions o f the substrate specif ic i ty i n the S I f ami ly of proteases have also been successful (Table 1.3). A number o f specif ici ty determinants have been uncovered by mutagenesis targeted to probe addi t ional features o f the enzyme, such as z y m o g e n processing and protease-inhibitor interactions [60-64]. T h e weal th o f b iochemica l and structural informat ion avai lable suggests the abi l i ty to design extended substrate specif ici ty o f the S 2 to S 4 pockets. Protease Engineered Property Ref. T r y p s i n Conve r s ion to elastase-like pr imary specif ici ty [65] Increased P I specif ic i ty towards A r g side chains [66] Increased P I specif ici ty towards L y s side chains [67] SI ' Eng inee r ing to favor basic residues [68] K a l l i k r e i n M o d i f i c a t i o n o f S 2 b ind ing pocket [69] F I X a Conve r s ion to F X a - l i k e extended substrate specif ici ty [70] T h r o m b i n Al te ra t ion o f P 2 - P 4 preference creating an anticoagulant protease [52,53] Table 1.3 Select examples o f successful protein engineering o f serine proteases. 1.10 Streptomyces griseus Trypsin Streptomyces griseus t rypsin ( S G T ) was in i t i a l ly pur i f ied f rom Pronase - a commerc i a l preparation o f secreted proteases - and characterized on the basis o f its hydro lys is of N a - b e n z o y l - L - a r g i n i n e e thyl ester (B AEE) and casein and its inh ib i t ion by soybean trypsin inhibi tor [71]. Further characterization o f its specif ici ty and inh ib i t ion identif ied S G T as a 21 typica l broad specif ici ty t ryps in- l ike serine protease o f the S I f ami ly that hydrolyzes polypept ide chains on the C- te rmina l side o f basic residues ( A r g and L y s ) [72]. B a s e d on sequence alignments, S G T is more s imi la r to bovine t rypsin than to other bacterial serine proteases [73,74]. The structure o f S G T was subsequently determined b y x-ray crystal lography and refined to 1.7 A , and revealed a three-dimensional fo ld that is also more s imi lar to mammal i an serine proteases than bacterial proteases [75,76]. A l t h o u g h similar i t ies exist at the sequence and structural l eve l , S G T differs f rom its mammal i an homologues i n its reduced number o f amino ac id insertions i n the polypept ide cha in when a l igned by either sequence or structure (Append ix A ) . M o r e o v e r , S G T contains on ly three disulf ide bonds rather than the f ive or s ix typ ica l ly observed i n the S I f ami ly o f proteases. These differences suggest that S G T c o u l d be used as a mode l scaffold to study the substrate specif ic i ty and other properties demonstrated by mammal i an proteases. 1.11 Statement of Hypothesis I f a l l substrate specif ici ty determinants o f coagulat ion factor X a are k n o w n , then their introduct ion into Streptomyces griseus t rypsin w i l l result i n a protease w i t h s imi lar substrate selectivity. 1.12 Objectives and Outline A l t h o u g h a weal th o f data is avai lable, several questions remain about the overa l l mechan i sm o f specif ic i ty o f serine proteases. A m o n g s t the poo r ly characterized details o f serine proteases are the f l ex ib i l i ty o f the active site and its influence on substrate specif ici ty, the role o f water molecules i n the active site and the poss ib i l i ty o f des igning ul t ra-high 22 specif ici ty serine proteases that can be tai lored to desired reactions. In this study, S G T is developed as a mode l for m a m m a l i a n serine protease specif ici ty as it has s imi la r i ty i n both sequence and structure, and is der ived f rom a bacterial source that should a l low for product ion i n other bacteria and hence a l low genetic modi f ica t ion o f the protein. Based on the structural s imilar i t ies observed near and around the active site o f S . griseus t rypsin compared to m a m m a l i a n serine proteases, site-directed mutagenesis should produce a ca ta lyt ica l ly active protease wi th the specif ici ty o f factor X . F o u r questions are to be addressed i n this study: (1) Does the recombinant S G T protein produced f rom Bacillus subtilis have s imi la r enzymatic properties as the w i l d type protein? (2) W h a t mutations are required to increase the pr imary specif ic i ty o f the enzyme towards A r g side chains? (3) W h a t point mutations or surface loops near the active site confer a greater degree o f extended specif ici ty in the enzyme? (4) W h a t are the complete requirements to convert S G T to coagulat ion factor X a - l i k e specif ic i ty? By engineering substrate specif ici ty into a protease where l i t t le exists, a number o f benefits w i l l result. A s ment ioned previous ly , proteases are used i n the site specific cleavage o f recombinant proteins and the enzymes used are cost ly due to their product ion f rom b lood . Proteases resul t ing f rom a bacterial expression system w o u l d cost far less and facilitate increased usage. Through the design o f specif ici ty s imi la r to a coagulat ion factor, one can examine the mechanisms b y w h i c h the proteases generate specif ici ty and the roles o f other regions o f the polypept ide that might be i n v o l v e d i n specif ic i ty . F o r example , nearly a l l 23 coagulat ion factor proteases have a specific sod ium b i n d i n g site and the b i n d i n g o f sod ium results i n a 3- to 5-fold rate increase i n catalysis [77]. S o d i u m b i n d i n g i n thrombin also alters the substrate specif ici ty o f the enzyme, yet dissect ion o f this process has been l imi ted o w i n g to the interconnected relat ionship amongst b iochemica l events w i th in the enzyme. Deve lopment o f a recombinant bacterial expression system for S. griseus t rypsin (Chapter 2) was a c ruc ia l obstacle to overcome i n this research. U s i n g this expression system, the p r imary specif ic i ty o f the protease was genet ical ly engineered to favor A r g over L y s side chains at the P I posi t ion (Chapter 3). Subsequent mutagenesis o f the S 2 to S4 pockets o f S G T was carr ied out to engineer coagulat ion factor X a substrate specif ici ty to the enzyme (Chapter 4). O n the basis o f these results, a f ramework for the product ion o f nove l proteases w i t h substrate specificit ies not observed i n nature is out l ined and other future directions are discussed (Chapter 5). 24 Chapter 2. Recombinant Protein Expression of Streptomyces griseus Trypsin in Bacillus subtilis 2.1 Introduction Deve lopment o f an efficient, cost effective and scaleable recombinant protein expression system is the first step i n protein engineering. So lub le , active and h igh puri ty protein must result f rom an efficient expression system. A number o f organisms have been used for this purpose i nc lud ing those f rom bacteria, fungi , yeast, and eukaryot ic c e l l l ines . Pro te in expression i n l ower organisms, such as gram-posi t ive and gram-negative bacteria, costs s ignif icant ly less but the drawback o f these systems is their inab i l i ty to produce complex proteins. L o n g e r product ion times and higher costs are associated w i t h eukaryotic based systems; however , their use is usual ly required when the protein to be produced is large (>60 k D a ) , has a complex fo ld and contains disulphide bonds or requires post-translational modi f ica t ion (such as g lycosyla t ion) . F o r protein engineering, bacterial , fungal , or yeast expression hosts are typ ica l ly used due to their amenabi l i ty to genetic manipula t ion and eff ic iency o f protein product ion. Successful expression o f a recombinant protein i n l ower organisms requires a number o f favorable b iochemica l features. L o w toxic i ty , s imple structural fo ld , l ack o f d isulphide bonds, l ack o f post-translational modi f ica t ion , and smal l size tend to help product ion. Recombinan t expression is also s ignif icant ly inf luenced by the properties o f the gene that encodes the polypept ide. O p t i m a l codon usage and lack o f secondary structure have been shown to be problematic i n the expression o f several proteins [78]. In many instances, 25 however , hav ing favorable characteristics at the genetic and protein leve l may s t i l l not result in successful expression and a great deal o f tr ial and error i n different systems is needed. Bac te r ia l protein expression is w i d e l y performed i n Escherichia coli. A s a host for genetic manipula t ion , E. coli is unr iva l led i n the divers i ty and s impl ic i ty o f methods established for genetic manipula t ion . Pur i f i ca t ion and alteration o f DNA i n these gram-negative bacteria is straightforward and h igh ly reproducible . Indeed, m u c h o f the history o f molecula r b i o l o g y is the result o f the study o f this organism. Unfortunately, E. coli is not as adept at the product ion o f foreign proteins. A number o f attempts have been made to engineer the genome o f this organism to increase its capacity for recombinant protein product ion, yet no universal solut ion has been found [79]. F o r this reason a number o f other bacteria have been investigated as alternatives i nc lud ing Bacilli [80,81], Lactobacilli [82], Streptomyctes [83], Pseuodomonads [84] and Caulobacter [85]. Importantly, these alternatives have the abi l i ty to secrete heterologous proteins outside o f the ce l l . Secret ion eases subsequent pur i f ica t ion, reduces the toxic i ty associated wi th the protein, and improves the rate o f format ion o f d isulphide bonds. On the basis o f its bacterial source, Streptomyces griseus t rypsin is suggested as a good target for recombinant protein expression and subsequent protein engineering. S G T has many features that are required for successful protein engineering. The structure o f the enzyme has been determined at h igh resolut ion and a weal th o f structure-function informat ion has been described for h igh ly s imi la r enzymes [75,76,86]. A l t h o u g h i t has a number o f desirable properties for recombinant expression such as its smal l size and s imple fo ld , several features o f the gene and protein cou ld compl ica te the product ion o f the recombinant protein. Transla t ion o f the m R N A encoding the protein may be hampered by the h igh guanine and 26 cytosine content o f the S G T gene (70%). Several authors have shown that sub-opt imal codon usage i n the first several codons can dramat ical ly decrease protein expression [87,88]. In the reducing environment o f a bacterial c e l l , the nascent polypept ide cha in may have d i f f icul ty fo rming the three disulphide bonds required to stabil ize the structure o f S G T [89]. Las t ly , i f the protease is produced i n an active form, i t m a y degrade components o f the c e l l as w e l l other S G T polypeptides. These properties indicate that product ion o f the recombinant protein may be diff icul t . O n c e produced by an expression host, a recombinant protein must be pur i f ied to homogenei ty. In this process, some form o f protein capture to concentrate and crudely pur i fy the protein is typ ica l ly l i n k e d w i t h one or more chromatographic separations. Throughout this procedure the loss o f protein, whether due to instabi l i ty or the act ivi ty o f contaminants, must be m i n i m i z e d . G i v e n the large body o f literature on serine proteases, par t icular ly t ryps in- l ike proteases o f the chymot ryps in fami ly , a number o f reagents are available c o m m e r c i a l l y and can be used i n a variety o f pur i f ica t ion methods. O n e part icular advantage o f us ing a bacterial expression host is the lack o f post-translational modi f ica t ion i n the target protein w h i c h min imize s sample heterogeneity i n the protein pur i f ied . Thus , recombinant t ryps in- l ike proteases der ived f rom a bacterial source should be easi ly pur i f ied i n h igh y i e ld . To begin in t roducing substrate specif ici ty into S G T , an efficient expression system was required. I chose a bacterial expression system that w o u l d facilitate d o w n stream processing and future high-throughput studies. B. subtilis is an excel lent expression system for S G T due to its abi l i ty to secrete active recombinant protein into the extracellular environment. C o m p a r i s o n o f the enzymatic properties and three d imens iona l structure o f the pur i f ied protein to the nat ively der ived protein f rom Streptomyces griseus show that both 27 proteins are ident ical . These studies p rov ide the basis for the subsequent engineering o f substrate specif ici ty. 2.2 Materials & Methods 2.2.1 Plasmids, Bacterial Strains, and Growth Conditions Escherichia coli was g r o w n us ing standard methods [90]. B. subtilis strain W B 7 0 0 was g r o w n i n super-rich m e d i u m [91] or on tryptose b l o o d agar base (Di fco) at 3 7 ° C . F o r the B. subtilis ca r ry ing p l a smid p W B 9 8 0 [92], kanamyc in was added to a f inal concentration o f 10 p,g ml" 1 i n both l i q u i d and so l id media . 2.2.2 DNA manipulation Procedures for genomic (5. griseus ( A T C C 10137)) and p l a smid D N A manipula t ion were carr ied out us ing established protocols [90]. P l a s m i d D N A was pur i f ied us ing a Q I A p r e p spin min iprep k i t (Qiagen). E n z y m e s were obtained f rom N e w E n g l a n d B i o l a b s and R o c h e M o l e c u l a r B i o c h e m i c a l s . F o r P C R ampl i f ica t ion o f the SprT gene encoding S G T , the f o l l o w i n g ol igonucleot ides were designed to mainta in reading frame o f the sacB signal peptide present i n p l a smid p W B 9 8 0 : 5 ' - g g a a g c t t t t g c a G T C G T C G G C G G A A C C C G C G C G G - 3 ' 5 ' - g g t o t a g a t t a G A G C G T G C G G G C G G C C G A G G - 3 ' (restriction sites are underl ined, the SprT gene specific sequences are g iven i n upper case). T h e P C R fragment was first c loned into pBluesc r ip t K S + (Stratagene) and then sub-cloned into p W B 9 8 0 us ing the H i n d l l l and X b a l restrict ion enzyme sites contained i n the o l igonucleot ide primers. Transformat ion o f B. subtilis strain W B 7 0 0 was performed by the 28 method o f S p i z i z e n [93]. D N A sequence analysis o f the c loned gene was performed us ing the B i g D y e Terminator k i t and analyzed on an A B I 3700 D N A Sequencer ( A p p l i e d Biosys tems) . 2.2.3 Protein Purification In order to purify the native and recombinant protease to homogenei ty , a pur i f icat ion strategy was developed at l o w p H to m i n i m i z e autolysis. Af te r centrifugation to remove cel lu lar debris (5,000 x g, 1 hr), recombinant S G T was purif ied f rom the supernatant o f 1 L o f B. subtilis W B 7 0 0 culture. Sequential a m m o n i u m sulphate fractionation was carr ied out at 3 0 % and 8 5 % saturation. The 8 5 % ( N H ^ S C M fraction pellet was resuspended i n 20 m M sod ium acetate buffer p H 4.5, d i a lyzed against the same buffer and appl ied to a c o l u m n (15 c m x 1.5 cm) o f S P Sepharose Fast F l o w ( A m e r s h a m Pharmacia) . Af te r extensive wash ing wi th 20 m M sod ium acetate buffer conta in ing 50 m M N a C I , p H 4.5, the bound proteins were eluted wi th 20 m M sod ium acetate buffer, p H 4.5, conta ining 150 m M N a C I . T h e active fractions were poo led and appl ied to a B e n z a m i d i n e Sepharose 4 Fast F l o w c o l u m n (8 c m x 0.75 cm) ( A m e r s h a m Pharmacia) . The c o l u m n was washed wi th 20 m M sod ium acetate buffer conta ining 500 m M N a C I , p H 4.5, and the enzyme was eluted i n the same buffer conta ining i n addi t ion 40 m M benzamidine H C 1 (Sigma) . Fract ions conta in ing active protease were poo led , concentrated and d ia lyzed against 10 m M T r i s - H C I buffer conta in ing 150 m M N a C I and 20 m M C a C l 2 , p H 7.6 us ing a 10,000 N M W L Ultrafree-4 centrifugal filter unit ( M i l l i p o r e ) . G e l f i l t rat ion through a c o l u m n (45 c m x 0.75 cm) o f Sephadex G - 7 5 ( A m er s ham Pharmacia) was performed using 10 m M T r i s - H C I conta ining 150 m M N a C I and 20 m M C a C l 2 , p H 7.6. S i m i l a r l y , native S G T was isolated f rom 1 g o f extracellular filtrate f rom S. griseus (Sigma) . The f ina l protein concentration was determined by U V absorbance at 280 nm, us ing the 29 ext inct ion coefficient 37,100 M " 1 cm" 1 [94] or by a B C A protein assay k i t (Pierce). A c t i v e site titration was performed us ing 4-ni t rophenyl p ' -guanidinobenzoate and a standard curve o f p -ni t rophenol (Sigma) . S o d i u m dodecylsulphate po lyac ry lamide ge l electrophoresis and Coomass ie B l u e staining were performed according to standard procedures [90]. N- t e rmina l protein microsequence analysis was performed by the Un ive r s i t y o f V i c t o r i a - G e n o m e B C Proteomics Centre (V ic to r i a , Canada) . Electrospray-mass spectrometry was carr ied out on a P E - S c i e x A P I 300 triple quadrupole mass spectrometer (Sciex) equipped wi th an Ionspray ion source. The mass spectrometry was performed by D r . S. H e i n the Withers laboratory (Dept. o f Chemis t ry , U B C ) . 2.2.4 Kinetic Analysis K i n e t i c analysis was performed i n 10 m M T r i s - H C I buffer conta ining 150 m M N a C I , 20 m M C a C l 2 , and 0.1 % P E G 8000, p H 7.6. A standard o f 7-amino 4-methy lcoumar in ( A M C ) was used to quantify the rates o f hydro lys is o f the f luorogenic substrates T o s - G l y - P r o -A r g - A M C and T o s - G l y - P r o - L y s - A M C (Bachem). A m i n i m u m o f six substrate concentrations ranging f rom 1 to 50 \iM was used. T h e f ina l concentration o f the enzyme i n each assay was 0.5 n M . Non- l inea r regression o f the in i t i a l react ion rates and ca lcula t ion o f the k inet ic parameters were performed us ing the Graphpad P r i s m 3.0 software (Graphpad). 2.2.5 Crystallization In previous studies, crystals o f the native S G T were obtained through batch crys ta l l iza t ion us ing ( N H 4 ) 2 S 0 4 [75,76]. In the current study, proteins were c rys ta l l ized us ing s imi la r condi t ions (10-15 m g / m L protein, 1.5 M ( N H ^ S O ^ 10 m M c a l c i u m acetate, p H 6.2) 30 except hanging drop vapor diffusion was u t i l i zed where the reservoir contained 1.55 M (NH 4 ) 2 S04. Crysta ls appeared i n two to three weeks to dimensions o f approximately 0.3 x 0.3 x 0.3 m m . D a t a were col lected at 100 K (Oxford Cryostream) wi th a M a r 3 4 5 detector mounted on a R i g a k u R U - 2 0 0 X - r a y generator (50 k V , 100 m A ) wi th O s m i c focus ing mirrors . Crysta ls were soaked br ief ly in 2 0 % g lyce ro l , 2.2 M (NH4)2S04for cryoprotect ion pr ior to data co l lec t ion . D a t a were processed us ing the HKL package and refined us ing C N S vers ion 1.1 i n combina t ion w i t h X t a l v i e w [95-97]. The prev ious ly reported native S G T structure ( P D B entry 1 S G T ) was used as a mode l for r i g i d body refinement o f the structure [76]. 2.3 Results and Discussion 2.3.1 Production and Purification of Recombinant SGT Previous studies demonstrated the abi l i ty to produce soluble t ryps in- l ike enzymes i n the per ip lasmic space o f E. coli [98-100]. In our experiments, however , E. coli was incapable of generating soluble S G T despite us ing a variety o f p l a smid constructs i n a number o f bacterial host strains. T h e presence o f three disulf ide bonds, the h igh G + C % content o f the SprT gene (70%) and the toxic i ty o f the recombinant protein are possible reasons for the lack of product ion o f recombinant proteins i n E. coli [78,88]. T o overcome these l imita t ions B. subtilis W B 7 0 0 , a strain that is deficient i n seven proteases, was used to produce sufficient y ie lds o f recombinant S G T for kinet ic and structural analysis [101]. U n l i k e E. coli, B. subtilis is capable o f secreting proteins into the extracellular environment, w h i c h facilitates rap id detection, pur i f ica t ion and analysis o f recombinant proteins. 31 Secret ion o f proteins into the extracellular m e d i u m is facili tated by the presence o f a single p lasma membrane i n the gram-posi t ive bacter ium B. subtilis. F o l l o w i n g translation by the r ibosome, a nascent polypept ide chain is targeted for secretion by the presence o f a c leavable amino- terminal s ignal peptide. O n the basis o f genome sequence analysis, over 300 proteins (-7.3 % o f all genes) have been postulated for secretion i n B. subtilis [102-104]. T h e abi l i ty to secrete such a w ide divers i ty o f proteins by this organism has been used for the product ion o f a number o f recombinant proteins [105-107]. Pro te in secretion i n a l l bacteria, i nc lud ing B. subtilis, is p r imar i l y due to the ATP-dependent Sec pathway. R e c o g n i t i o n o f the s ignal peptide is mediated by the s ignal recogni t ion particle protein complex , w h i c h shuttles the unfolded protein to the c e l l membrane [108]. R e m o v a l o f the s ignal peptide occurs as the denatured protein translocates across the c e l l membrane and is typ ica l ly carr ied out by the type I s ignal peptidase S i p S i n B . subtilis [109]. In our expression system, secretion o f S G T into the extracellular environment was mediated by fusing the gene to the s ignal peptide sequence o f levansucrase (SacB) (Figure 2.1). Thus , the N- terminus o f the protein is not accessible to the active site and the protease is inact ive unt i l it is secreted f rom the ce l l . C leavage after the A l a - P h e - A l a sequence at the junc t ion o f the fusion protein by S i p S generates the correct N- te rminus . S G T can then fo ld i n the compara t ive ly non-reducing environment outside o f the c e l l . B y m i m i c k i n g the native organism for the product ion o f the recombinant S G T , we have developed an efficient and nove l expression system for the product ion o f t ryps in- l ike enzymes. 32 Xbal (4376) Replicase Kanamycin nucleotidyltransferase F i g u r e 2.1 P l a s m i d map o f S G T gene c loned into p W B 9 8 0 for recombinant protein expression i n B. subtilis. The gene was c loned to mainta in the reading frame o f the S a c B signal peptide w h i c h facilitates secretion o f the recombinant protein into the extracel lular environment. Recombinan t protein yields o f >15 m g / L o f culture m e d i u m were obtained wi th in 24 hours o f growth at 37°C. T h e four step purif icat ion typ ica l ly produced 10-15 m g / L o f B. subtilis culture wi th an overa l l y i e l d o f 8 0 % (Table 2.1). Four separation techniques were appl ied to y i e l d the highest puri ty enzyme possible. The methods were chosen based on their compat ib i l i ty and m i l d condi t ions . Fo r example , an affinity chromatography step us ing soybean trypsin inhib i tor was found to b ind S G T effectively. H o w e v e r , r emova l o f the protein 33 f rom this type o f c o l u m n required p H 2.0 and it was feared that such condit ions may destroy unstable mutants o f S G T . V o l . To ta l To ta l Speci f ic Pur i f i ca t ion Y i e l d ( m L ) Pro te in A c t i v i t y A c t i v i t y (fold) (%) (mg) (mmol/s) (pmol/s /mg) M e d i a 1000 16000 59 4 1 100 ( N H 4 ) 2 S 0 4 50 400 53 134 34 90 precipitate S P Sepharose 15 18.3 51 2774 694 87 B e n z a m i d i n e 5 12.4 49 3856 964 81 Sepharose G-75 Superdex 4 12.2 48 3 9 5 5 989 81 Table 2.1 Pur i f ica t ion table for recombinant S G T ( b S G T ) f rom B . subtilis extracellular supernatant. A c t i v i t y was measured by the hydrolys is o f the chromogenic substrate B z - U e - G l u - A r g - p N A at 4 0 u M i n 10 m M T r i s - H C I , 20 m M C a C l 2 , p H 7.6. Na t ive and recombinant S G T were pur i f ied to homogenei ty pr ior to analysis. Pur i ty was assessed by several different cri teria. B y electrospray ion iza t ion mass spectrometry, the expected and observed masses for the native and recombinant protein were w i t h i n experimental error (23106.9 and 23107.0 amu, respectively) . F o r both proteins the integrated data f rom the m/z+ fragments y ie lded a single unambiguous peak wi th m i n i m a l background (Figure 2.2). In addit ion, when the protease was treated w i t h phenylmethanesulfonyl f luoride, S D S - P A G E showed a single band o f the expected molecula r weight (Figure 2.3). O n l y two autolytic fragments o f S G T were observed when the protease was bo i l ed pr ior to S D S - P A G E i n the absence o f a strong inhibi tor . 34 A m i n o - t e r m i n a l sequence analysis revealed that the wi ld- type enzyme had a single unambiguous sequence N H 2 - V a l - V a l - G l y - G l y - T h r - A r g corresponding to the publ i shed S G T sequence (12). Together w i th the protein assays and active site titration data, these results suggest that the f inal protein preparation was greater than 9 9 % pure. BioSpec Reconstruct for +Q1: 5.92 min (6 scans) from BSGT 1.4e5H 231 07.0 21047.0 21637.0 LA L< A ^ ^ J w a k l a ^ i i .. 21000 23000 Mass, amu 1.80e5 cps 24355.0 25000 Figure 2.2. E S - M S spectrum o f pur i f ied recombinant S G T . T h e single m/z peak at 23,107.0 amu wi th m i n i m a l background indicates the h igh puri ty o f the recombinant protein. 35 kDa A 97.4 66.2 45.0 31.0 21.5 14.4 F i g u r e 2.3 S D S - P A G E o f pur i f ied recombinant S G T . L a n e A : b S G T (0.5 //g) inactivated wi th P M S F prior to addit ion o f S D S - P A G E load ing buffer and b o i l i n g . Lane B : b S G T (1.0 jug) wi thout P M S F inh ib i t ion . The central lane o f the ge l contains low-range protein molecular mass standards ( B i o - R a d ) whose masses are g iven on the left-hand side o f the gel . The ge l was stained wi th Coomass ie B r i l l i a n t B l u e . 2.3.2 K i n e t i c A n a l y s i s T w o f luorogenic peptide substrates, T o s - G l y - P r o - A r g - A M C and T o s - G l y - P r o - L y s -A M C , were used to moni tor the P I A r g : L y s preference o f the native and recombinant proteases (Table 2.2). T h e kinet ic parameters o f the native and recombinant wi ld - type S G T are s imi la r to previously reported values using the same pair o f substrates [66]. O n the basis o f the s imi lar rates o f hydro lys i s o f these peptide substrates, we can conc lude that the recombinant protein behaves ident ical ly to the native protease. 36 T o s - G l y - P r o - A r g - A M C T o s - G l y - P r o - L y s - A M C S R / S K * kcat kcat / Km kcat K m kcat / Km (min") ( u M ) (min" u M " ) ( m i n ) (u.M) (min" u.M") S G T 4880 2.3 2122. 1670 3.6 464 4.6 ± 4 1 0 ± 0 . 2 ± 120 ± 0 . 4 b S G T 4570 2.0 2285 1520 3.2 475 4.8 ± 1210 ± 0 . 2 ± 6 0 ± 0 . 2 * S R / S K = ( T o s - G l y - P r o - A r g - A M C k ^ / K m ) / ( T o s - G l y - P r o - L y s - A M C kc, / Km) Table 2.2 P I arginine to lys ine preference o f S G T enzymes. A r g : L y s preference was measured b y amido ly t i c act ivi ty o f the native ( S G T ) , recombinant S G T ( b S G T ) us ing two f luorogenic peptides, T o s - G l y - P r o - A r g - A M C and T o s - G l y -P r o - L y s - A M C . V a l u e s obtained i n triplicate ± S . D . 2.3.3 Crystallization and Structure Determination X - r a y diffraction data were obtained for the recombinant wi ld- type S G T at 1.5 A resolut ion. Da ta co l lec t ion and refinement statistics are g iven in Tab le 2.3. T h e recombinant protease crys ta l l ized i n the C222] space group and contained one molecu le per asymmetr ic unit and a Mat thews coefficient o f 2.2 A 3 / D a (Table 2.3). T h e structure was deposited i n the P D B database as l O S S . D u r i n g refinement, l o w R c r y st and R f r e e values were obtained (0.19 and 0.22). These were accompanied by excel lent stereochemistry ind ica t ing a h igh qual i ty mode l . Inspection o f the Ramachandran plot revealed that a l l non-g lyc ine backbone atoms are i n a l l owed regions, w i t h on ly A s n l 7 8 adopting a conformat ion i n the generously a l l owed region (Figure 2.4). The overa l l B-factors for the polypept ide atoms were l o w ( -13 A 2 ) , and regions wi th h igh B-factors were l imi ted to solvent exposed regions that are not i n v o l v e d i n crystal pack ing . The structure o f S G T is the highest qual i ty reported to date l i k e l y due to the 37 decreased radiation damage as the present crystals were ana lyzed under cryogenic condi t ions wi th a shorter co l l ec t ion t ime. Phi (degrees) F i g u r e 2.4 Ramachandran plot o f the crystal structure o f recombinant wi ld - type S G T . A l l non-glycine residues are i n the a l lowed conformat ion. The plot was calculated us ing PROCHECK[110] 38 Data co l lec t ion Reso lu t ion (A) Tota l Observat ions Completeness (%) Average redundancy I/oT Rmerge (%) Refinement statistics Space group C e l l d imensions (a,b,c) * M o l e c u l e s per asymmetr ic unit Rcryst Rfree Prote in atoms* Solvent atoms per asymmetric unit Average B-factor for protein (A2) Average B-factor for water (A2) Occupancy o f C a 2 + (B A 2 ) B o n d length deviations (A) B o n d angle deviations (°) 1.5 ( 1 . 5 5 - 1.65) 25923 84.1 (74.1) 2.8 20.4 (5.1) 4.1 (18.2) C 2 2 2 i 50.04, 69.82, 119.65 1 0.196 0.222 1623 211 12.9 21.0 0.53 (14.98) 0.007 1.4 a = P = y = 90° ; i nc lud ing alternate side cha in conformations Table 2.3 Da ta co l lec t ion and refinement statistics o f wi ld- type recombinant S G T . Statistics for the highest resolut ion shel l are g iven i n parentheses. 39 2.3.4 Comparison of the recombinant and Pronase-derived crystal structure A l l amino ac id residues i n the mode l o f the recombinant S G T were c lear ly identif ied and posi t ioned. In 1 S G T [76], several residues had weak or absent side-chain density. In the present structure, most o f these densities were c lear ly resolved, al though several residues lacked density i n the terminal atoms o f their side-chains (Thr20, G l n 7 5 , L y s 8 2 , Thr98 , Ser236, and A r g 2 4 3 ) indica t ing disorder o f these solvent exposed atoms. Residues 77 and 79 were mode led as G l y and A l a i n 1 S G T but are two Ser residues by D N A sequence analysis [74]. These Ser residues were c lear ly resolved i n the current electron density map. T h e electron density o f one sulphate i on was observed i n the oxyan ion hole o f the substrate b ind ing site and was inc luded i n the structure. The pos i t ion o f this sulphate is conserved i n anionic sa lmon trypsin ( P D B entry 1 B I T ) , bovine trypsin ( 1 T L D ) , and porc ine pancreatic elastase ( 3 E S T ) [111-113]. Alternate conformations were observed for G l n l 9 2 , w h i c h either points into the solvent or forms a pair o f hydrogen bonds wi th the backbone N H o f G l y l 4 8 o f an adjacent S G T molecule . In 1 S G T the same residue was noted as hav ing h igh m o b i l i t y [76]. Differences i n the C a + 2 b ind ing site i n the recombinant crystal structure are discussed i n Chapter 3. The overa l l differences between the native ( 1 S G T ) and wi ld- type recombinant enzyme are minor (Figure 2.5), w i th a root mean square deviat ion o f a l l 892 atoms i n the C a backbone o f 0.27 A. The largest deviat ion i n backbone occurs at the 174-loop where peptide bond o f A l a l 7 7 a adopts a 180° rotation compared to the native structure. 40 F i g u r e 2.5 Super impos i t ion o f the C a traces o f native and recombinant wi ld - type S G T . T h e two structures ( P D B I D 1 S G T and 1 0 S 8 ) are indis t inguishable as indicated by the smal l r.m.s devia t ion between the C u backbone atoms o f the two structures (0.27 A ) . 2.3.5 W i l d - t y p e N a t i v e a n d R e c o m b i n a n t S G T Subs t ra t e B i n d i n g A negatively charged residue (D189) is present in the S I pocket and confers the pr imary specif ici ty o f S G T and other t rypsin- l ike enzymes towards pos i t ive ly charged A r g or L y s side chains [114]. Based on the structure o f 1 S G T , D 1 8 9 is located at the base o f a narrow cy l ind r i ca l cleft that can accommodate these side chains [76]. T h e sl ight preference for A r g over L y s in this pocket is due to the requirement for a b r idg ing water molecu le between the shorter lysy l - s ide chain and D 1 8 9 [66]. The y - O H o f residue T 1 9 0 interacts directly w i t h the substrate v i a hydrogen bonding . Bo th L y s and A r g side chains adopt 41 favorable conformations for interaction wi th D 1 8 9 and the h y d r o x y l group o f T 1 9 0 . K i n e t i c analysis o f the native and recombinant S G T proteases demonstrated an A r g : L y s preference o f 4 :1 . Ex tended substrate specif ici ty is largely absent in a l l b roadly specific t ryps in- l ike enzymes. A c c e s s i b i l i t y o f the catalytic triad i n S G T is not hindered by the structure o f the active site. Crys t a l structures o f other t rypsin- l ike enzymes have demonstrated that the peptide backbone o f substrate residues P I to P3 forms an anti-parallel (3-sheet w i t h residues 214 to 216 i n the enzyme [24]. A n a l y s i s o f the substrate specif ic i ty o f the S 2 to S4 pockets i n bovine trypsin has shown an absence for preference o f any side cha in at these posit ions [115,116]. T h e structure o f S G T suggests an ident ical mode o f substrate b ind ing and an absence o f substrate specif ic i ty i n the S 2 to S4 pockets. 2.4 Conclusions A nove l expression system for the product ion o f recombinant S G T has been developed us ing B. subtilis. H i g h puri ty protease resulted f rom a four-step pur i f ica t ion protocol . T h e recombinant protein demonstrates ident ical b iochemica l and structural properties to the Streptomyces der ived protease. O n the basis o f the h igh l eve l o f product ion , puri ty, and abi l i ty to crysta l l ize the protein, the recombinant protein is h igh ly amenable for engineering extended substrate specif ici ty into the enzyme. 42 Chapter 3. Engineering the Primary Substrate Specificity of Streptomyces griseus Trypsin 3.1 Introduction Substrate specif ici ty is a k e y concept i n the analysis o f serine proteases. Sequence analysis studies show that the S I f ami ly o f t ryps in- l ike enzymes l i k e l y evo lved f rom a c o m m o n ancestral gene [117]. T h e cu lmina t ion o f incremental evolut ionary steps led to the appearance o f a number o f h igh ly specific proteases such as those found i n the vertebrate b l o o d coagulat ion cascades. These proteases fu l f i l l regulatory roles i n ce l lu lar processes that are dist inct f rom their more p r imi t ive roles as degradative and protective enzymes [118]. M u c h data have been col lected on the specif ici ty determinants o f serine proteases, result ing i n a c lassif icat ion system based on their p r imary specif ic i ty (S I pocket) . T h e specif ic i ty o f t ryps in- l ike enzymes at the S I pocket is largely defined by the presence o f a negatively charged side cha in at posi t ion 189 [114]. O p t i m a l b i n d i n g o f the pos i t ive ly charged substrate ( A r g or L y s side chains) to residue 189 is mediated by residue 190 [66,119]. In t ryps in- l ike serine proteases, pos i t ion 190 is occup ied by a l imi ted number o f amino acids. Degradat ive proteases wi th l o w pr imary specif ic i ty ( A r g : L y s preference o f 4:1) d isplay G i n , T h r or Ser at pos i t ion 190, whereas proteases wi th h igh pr imary specif ic i ty ( A r g : L y s preferences greater than 7:1) contain A l a or Ser at this pos i t ion [120]. Prev ious studies have shown that mutagenesis o f pos i t ion 190 can be used to manipulate the substrate specif ici ty o f t rypsin to favor cleavage after either A r g or L y s side chains [66,67,119]. S i m i l a r to degradative vertebrate t ryps in- l ike enzymes, S G T demonstrates a pr imary substrate preference o f A r g L y s o f 4 :1 . In the previous chapter I described the product ion o f 43 fu l ly active recombinant S G T f rom B. subtilis. U s i n g this system, mutants o f S G T were constructed wi th altered preference for arginine to lys ine ( A r g : L y s ) . Muta t ions were designed to m i m i c those found i n other t ryps in- l ike enzymes. SGT mutant T 1 9 0 P is considerably more active and less Arg - spec i f i c when compared wi th the previous ly publ i shed S 1 9 0 P mutat ion created i n rat anionic t rypsin [67]. K i n e t i c and structural analysis o f the mutant protease shows that both the act ivi ty and specif ic i ty o f the enzyme is affected by residues surrounding residue 190. These results further our understanding o f the pr imary substrate specif ici ty o f t ryps in- l ike enzymes. B a s e d on the ease of product ion and pur i f icat ion o f the recombinant protein i n our B. subtilis expression system, SGT is an ideal scaffold for the introduct ion o f addit ional mutations to enhance the substrate specif ici ty o f the S 2 to S 4 b ind ing pockets. 3.2 Materials & Methods 3.2.1 DNA Manipulation and Protein Purification U s i n g the prev ious ly described S G T gene c loned into pBluescr ip t K S + p lasmid , mutagenesis was performed on the gene us ing a Q u i k C h a n g e site-directed mutagenesis k i t (Stratagene) as described by the manufacturer. Ol igonucleot ides used for mutagenesis are p rov ided i n Tab le 3.1. D N A sequence analysis o f the c loned gene and mutants was performed us ing the B i g D y e Terminator k i t and analyzed on an A B I 3700 D N A Sequencer ( A p p l i e d Biosys tems) . M u t a n t S G T genes were sub-cloned into p lasmid p W B 9 8 0 and transformed into B. subtilis W B 7 0 0 as described previous ly . M u t a n t S G T proteins were expressed and pur i f ied in an ident ica l manner as the wi ld- type . 44 Mutation Oligon ucleotide T 1 9 0 A 5 ' - G G C G T C G A C G C C T G C C A G G G T - 3 ' T 1 9 0 P 5 ' - G G C G T C G A C C C C T G C C A G G G T - 3 ' T 1 9 0 V 5 ' - G G C G T C G A C T T C T G C C A G G G T - 3 ' T 1 9 0 V 5 ' - G G C G T C G A C G T C T G C C A G G G T - 3 ' Table 3.1 Ol igonucleot ides used to mutate residue 190 i n the S G T gene. The reverse complement sequences o f these ol igonucleot ides were also used i n the mutagenesis. 3.2.2 Kinetic Analysis K i n e t i c analysis was performed in 10 m M Tr is H C 1 buffer conta ining 150 m M N a C I , 20 m M C a C l 2 , and 0.1 % P E G 8000, p H 7.6. A standard o f 7-amino 4-methy lcoumar in ( A M C ) was used to quantify the rates o f hydrolys is o f the f luorogenic substrates T o s - G l y - P r o -A r g - A M C and T o s - G l y - P r o - L y s - A M C (Bachem). A m i n i m u m o f s ix substrate concentrations ranging f rom 1 to 500 u . M was used. E n z y m e concentrations ranged f rom 0.5 to 10 n M . B e n z a m i d i n e concentrations ranged f rom 5 to 200 n M . Non- l inea r regression o f the in i t i a l reaction rates and ca lcula t ion o f the kinet ic parameters were performed us ing the Graphpad P r i s m 3.0 software (Graphpad). 3.2.3 Crystallization and Structure Determination Crys ta l l i za t ion condi t ions for the T 1 9 0 P mutant o f S G T were s imi la r to the wi ld - type protein; however 25 m M benzamidine was inc luded i n the sample buffer. Crysta ls appeared i n two to three weeks to dimensions o f 0.3 x 0.3 x 0.3 m m . Da ta co l lec t ion and structure refinement were ident ical to those used for the wi ld- type recombinant structure. E lec t ron 45 density o f the inhib i tor i n the S I pocket o f the mutant enzyme was evident, and was inc luded i n the f inal mode l . 3.3 Results and Discussion 3.3.1 Production of SGT Mutants Expres s ion levels o f the four mutant proteases were comparable to the wi ld- type S G T protein suggesting that the mutations were not detrimental to the overa l l stabili ty and fo ld ing o f the proteases. T o ensure accurate kinet ics , the complete method us ing the four-step pur i f icat ion pro tocol was appl ied to each protein. E S - M S analysis o f the recombinant proteases demonstrated the presence o f the mutation, as w e l l as h igh puri ty o f the sample (Table 3.2). Theoretical Mass Observed Mass Difference (amu) (amu) (amu) T 1 9 0 A 23077.0 23079.0 2~0 T 1 9 0 P 23103.0 23098.8 4.2 T 1 9 0 S 23093.0 23090.0 3.0 T 1 9 0 V 23107.1 23102.0 5.1 Table 3.2 E S - M S analysis o f the four mutants o f S G T . 3.3.2 Kinetic Analysis T w o f luorogenic peptide substrates, T o s - G l y - P r o - A r g - A M C and T o s - G l y - P r o - L y s -A M C , were used to moni tor the P I A r g : L y s preference o f recombinant and mutant proteases (Table 3.3). V a l u e s for the native and wi ld - type recombinant S G T are p rov ided i n the table 46 for reference. Mutants T 1 9 0 P and T 1 9 0 A demonstrated a significant increase i n P I A r g : L y s preference over the wi ld- type enzyme o f 18:1 and 8:1, respectively. A l l four mutants showed increased K m values for the substrates tested suggesting that the S I pocket o f S G T is op t imized for substrate b ind ing , a feature that has been observed i n other t rypsin- l ike enzymes [66,67,119]. A s imi lar trend o f fo ld differences for the K i value o f the smal l molecu le inhib i tor benzamidine was observed, w i th the except ion o f the T 1 9 0 P mutant (Table 3.4). Tos-Gly-Pro-Arg-AMC Tos-Gly-Pro-Lys-AMC S R / S K * kcat kcat / K m kcat kcat / K m (min") (MM) (min" uM") (min-) (MM) (min" uM") S G T 4880 ±410 2.3 ±0.2 2122 1670±120 3.6 ± 0.4 464 4.6 b S G T 4570±1210 2.0 ±0 .2 2285 1520 ± 6 0 3.2 ±0 .2 475 4.8 T 1 9 0 A 4950 ± 470 12.9 ± 1.0 384 192 ± 32 4.4 ± 0.6 44 8.7 T 1 9 0 P 3 6 1 0 ± 1 3 0 67 ± 5 54 527 ± 2 166 ± 2 3 18 T 1 9 0 S 6036 ± 561 6.1 ±0 .6 990 2584 ±141 10.1 ± 1.2 256 3.9 T 1 9 0 V 2300 ±177 224 ± 20 11 791 ± 13 231 ± 2 3.4 3.2 S R / S K = ( T o s - G l y - P r o - A r g - A M C kcat / K r a ) / ( T o s - G l y - P r o - L y s - A M C kcat / K r a ) Table 3.3 P I A r g i n i n e to L y s i n e preference o f mutant S G T enzymes. A r g : L y s preference was measured by amidoly t ic act ivi ty o f the native ( S G T ) , recombinant ( b S G T ) and mutants o f S G T us ing two f luorogenic peptides, Tos -G l y - P r o - A r g - A M C and T o s - G l y - P r o - L y s - A M C . V a l u e s obtained i n triplicate ± S . D . 47 K; (uM) b S G T 2.7 ± 0 . 8 T 1 9 0 A 9.9 ± 0.7 T 1 9 0 P 16.4 ± 2 . 2 T 1 9 0 S 4.8 ± 0 . 8 T 1 9 0 V 197 ± 37 Table 3.4 K j values o f benzamidine for recombinant S G T and the four mutant forms. 3.3.3. Crystallization and Structure Refinement X - r a y diffraction data were obtained for the T 1 9 0 P mutant o f S G T at 1.9 A resolut ion. Da ta co l lec t ion and refinement statistics are g iven i n Tab le 3.5. A s wi th the recombinant w i l d -type, the mutant c rys ta l l ized in the C222i space group, contained one molecule per asymmetric unit and a Mat thews coefficient o f 2.3 A 3 / D a . T h e mutant structure was deposited i n the P D B database as 1 0 S 8 . D u r i n g refinement, l o w Rcryst and Rf r e e values were obtained (0.17 and 0.21). These were accompanied by excel lent stereochemistry (Figure 3.1). 48 T190P SGT Data co l lec t ion Reso lu t ion ( A ) To ta l Observat ions Completeness (%) Average redundancy Vol Rmerge (%) Refinement statistics Space group C e l l d imensions (a,b,c) * M o l e c u l e s per asymmetr ic unit Rcryst Rfree Prote in atoiW Solvent atoms per asymmetr ic unit Ave rage B-fac tor for protein ( A 2 ) Ave rage B-fac tor for water ( A 2 ) Occupancy o f C a 2 + (B A 2 ) B o n d length deviations (A) B o n d angle deviations (°) 1.9 (1.93 - 2 . 0 5 ) 14471 89.7 (79.1) 7.1 44.4 (25.2) 3.1 (5.6) C 2 2 2 , 50.08, 69.60, 119.83 1 0.167 0.212 1632 193 12.8 21.4 0.43 (17.50) 0.007 1.4 * a = P = y = 9 0 ° ; T i nc lud ing alternate side chain conformations Table 3.5 Da ta co l lec t ion and refinement statistics for the T 1 9 0 P mutant o f S G T i n complex wi th benzamidine. Statistics for the highest resolut ion shel l are g iven i n parentheses. 49 Phi (degrees) F i g u r e 3.1 Ramachandran plot o f the crystal structure o f the T 1 9 0 P mutant o f S G T . A l l non-g lyc ine residues are in the a l lowed conformat ion. The plot was calculated us ing P R O C H E C K [110] 3.3.4 C a 2 + b i n d i n g site o f B . sub t i l i s d e r i v e d S G T In the previous ly reported structure ( 1 S G T ) the c a l c i u m b ind ing site consisted o f the A s p 165 and G l u 2 3 0 carboxylate groups, two w e l l ordered water molecules and the carbonyl oxygen atoms o f residues A l a l 7 7 a and G l u l 8 0 (Figure 3.2) [76]. In the present structures, A s p l 6 5 was shown to adopt a different conformation than observed prev ious ly . In the w i l d -type recombinant S G T structure, the carboxylate o f A s p 165 forms a bidentate electrostatic interaction wi th A r g l 6 9 rather than the structural C a 2 + ion . In the T 1 9 0 P m o d e l , A s p l 6 5 adopts a pair o f alternative conformations, either facing the C a 2 + ion or towards A r g l 6 9 . 50 M o d e l i n g o f the two posit ions at ha l f occupancy generated a l ower B -factor for the conformat ion i n v o l v e d i n the electrostatic interaction wi th A r g l 6 9 . In both structures, the carbonyl oxygen o f A l a l 7 7 a was oriented away f rom the C a 2 + i o n and does not appear to be i n v o l v e d in the interaction. Three ordered water molecules are associated wi th the i on , rather than the two prev ious ly observed. In contrast, a single disordered water molecu le wi th a h igh B- fac to r (41 .4 A 2 ) was found near the C a 2 + i o n i n the T 1 9 0 P crystal structure. In other t rypsin- l ike proteases, C a 2 + ions have been found prev ious ly w i t h h igh B-factors relative to the overa l l structure suggesting less than fu l l occupancy (25). The reduced occupancy o f the ion is not surpris ing as a related enzyme, Streptomyces erythraeus t rypsin, lacks G l u 2 3 0 and no c a l c i u m is evident i n the structure [121]. M o r e o v e r , the suggested C a 2 + b ind ing site i n S G T is comple te ly different than that observed i n mammal i an t ryps in- l ike enzymes [76]. W h e n taken wi th previous observations that the c a l c i u m ion plays no role i n catalysis but p lays a role on ly i n structural stabil i ty i n l o w / h i g h p H solutions, these data suggest that the c a l c i u m b ind ing site has weak affinity for the i o n [72]. 3.3.5 T190S a n d the Loss of 7-CH3 Degradat ive proteases i n v o l v e d i n digestive and protective functions typ ica l ly possess Ser or T h r residues at pos i t ion 190. H o w e v e r , proteases exh ib i t ing higher substrate specif ici ty m a y also possess these residues [120]. The T 1 9 0 S mutant o f S G T demonstrated no significant increase i n overa l l substrate specif ic i ty , yet a mino r increase i n catalytic act ivi ty (kc at increase o f 25%) and a 3-fold increase i n K m (Table 3.3) were observed for both the A r g and L y s conta in ing substrates. Los s o f the y - C H 3 is un l ike ly to s ignif icant ly increase the solvent accessibi l i ty o f D 1 8 9 , and it is more l i k e l y that the increased m o b i l i t y o f the y-OH results i n 51 the increased K m values for both substrates. Th i s is va l id i f koai is used as a measure o f the stabil ization o f the transit ion state. In eukaryotic t rypsin- l ike enzymes that conta in Ser at posi t ion 190, the space that w o u l d be occupied by a y - C H 3 is f i l l ed by me thy l groups f rom either residue 16 or 138 usual ly i n the fo rm o f isoleucine side chains. In S G T , both o f these residues are valine and possess one fewer methyl group. F i g u r e 3.2 C o m p a r i s o n o f the C a 2 + b ind ing site i n S G T enzymes: ( A ) native, ( B ) wi ld- type recombinant and (C) T 1 9 0 P mutant o f S G T . The number o f water molecules ( • ) that co-ordinate the structural c a l c i u m ion (O), as w e l l as the conformat ion o f the amino acid l igands differ in a l l three structures. 52 3.3.6 T190V and the Effect of a Branched Side Chain In contrast to the kinet ics o f T 1 9 0 S , the T 1 9 0 V mutant demonstrated a 2-fold reduct ion i n k c a t i n combinat ion w i t h nearly a 100-fold increase i n K m for both A r g and L y s conta ining substrates (Table 3.3). Replacement o f the y - O H wi th a methyl group removes the hydrogen bond ing capacity o f the residue and reduces the solvent accessibi l i ty o f D 1 8 9 . T h i s results i n a destabi l ized transition state complex relative to the wi ld- type protease and a weaker electrostatic interaction between the substrate and enzyme. T h e 100-fold increase i n K m for A r g conta ining substrates compared to the 70- fo ld increase i n K m for L y s conta in ing substrates indicates the smal l increase in vo lume has a more significant effect on the longer and bu lk ie r a rg inyl side chain . D 1 8 9 is not typ ica l ly observed wi th V I 9 0 or 116 i n naturally occur r ing t rypsin- l ike enzymes l i k e l y due to the poor catalytic eff ic iency o f these combinat ions o f side chains. 3.3.7 T190A and the Loss of y-OH A majori ty o f vertebrate enzymes i n v o l v e d i n phys io log i ca l regulat ion, such as the coagulat ion factor serine proteases, possess A l a at pos i t ion 190 and exhibi t A r g : L y s substrate specificit ies ranging f rom 7:1 for coagulat ion factor X a to greater than 14:1 for bovine thrombin [119]. T h e T 1 9 0 A mutat ion i n S G T demonstrates the molecula r basis for the predominance o f A l a at this pos i t ion i n h igh ly specific proteases favor ing a P I A r g residue. O p t i m a l rates o f catalysis at l o w concentrations o f substrate tend to be requisite characteristics o f these vertebrate enzymes. K i n e t i c analysis o f the T 1 9 0 A mutant revealed no change i n k c a t for A r g conta ining substrates but rather a 10-fold reduct ion i n k c a t for L y s conta in ing substrates (Table 3.3). A s noted previous ly , the increase i n solvent accessibi l i ty o f D 1 8 9 53 should stabil ize the transition state complex . H o w e v e r , the loss o f hydrogen bond ing (previously p rov ided by the y - O H ) exhibits a more pronounced effect on the L y s conta in ing substrate due to its requirement o f an ordered b r idg ing water molecu le to D 1 8 9 . 3.3.8 T190P U n l i k e the T 1 9 0 A mutat ion, whose effects were predominant ly on the kc a t o f the L y s -conta in ing substrate, the T 1 9 0 P mutat ion affected the K m s igni f icandy. The K m for the A r g containing substrate was 35-fo ld higher than the wi ld- type , compared to the 46- fo ld increase for the L y s containing substrate (Table 3.3). S i m i l a r l y , the reduct ion i n kc a t for the A r g substrate (25%) was s ignif icant ly less than the L y s substrate (66%). Together, these changes generate an overa l l A r g to L y s preference o f 18 to 1. A previous report [66] analyzed the S 1 9 0 P mutant o f rat anionic t rypsin, w h i c h is analogous to the T 1 9 0 P constructed i n S G T . W h e n us ing the same pair o f substrates used i n this study, wi ld- type rat anionic t rypsin exhibi ts a s imi la r p r imary substrate specif ic i ty to S G T . H o w e v e r , the S 1 9 0 P mutat ion i n rat anionic t rypsin results i n a h igh ly specific protease that favors the A r g substrate 135-fold over the L y s conta in ing substrate. T h i s increase i n substrate specif ici ty was combined wi th a greater than 10-fold reduction i n kc a t for both A r g and L y s substrates. These authors suggested that Tyr228 may be i n v o l v e d i n steric c lash ing w i t h the prol ine r ing at pos i t ion 190, leading to reduced activi ty. In S G T , residue 228 is also T y r suggesting an alternate b ind ing mode o f this mutant. T o address these discrepancies, the crystal structure o f the T 1 9 0 P mutant i n complex wi th the smal l molecule inhibi tor benzamidine was investigated. B i n d i n g o f the benzamidine inhib i tor to the T 1 9 0 P mutant is nearly ident ica l to that observed i n other t rypsin- l ike proteases [112]. The prol ine r i ng o f residue 190 does not adopt 54 a conformat ion that occludes the negat ively charged carboxylate group o f A s p 189, nor does it conf l ic t w i th T y r 2 2 8 suggesting that the p rev ious ly characterized specif ici ty o f the S 1 9 0 P mutant i n rat anionic t rypsin was the result o f second shell residues at posi t ions 16 or 138 (Figure 3.3). T h e mutat ion does not s ignif icant ly affect the conformat ion o f any o f the residues surrounding T 1 9 0 P , i nc lud ing the c r i t i ca l A s p l 8 9 . The loca l r.m.s. deviat ion is l o w (0.70 A) for a l l 96 atoms wi th in a 5 A radius o f residue 190. H o w e v e r , the backbone carbonyl group o f A s p l 8 9 is rotated 4 5 ° relative to the wi ld - type structure due steric constraints o f the prol ine residue. Rota t ion o f this carbonyl group does not disrupt the hydrogen bond wi th the backbone nitrogen o f residue 17. Hence , the moderate increase i n A r g to L y s substrate specif ic i ty o f this mutant is the l i k e l y the result o f the strengthened interaction between the substrate and A s p 189 i n a more hydrophobic environment. T h e effect on lysy l - s ide chains is more predominant due to the l ack o f hydrogen bond ing o f the prol ine r i ng to the substrate or a b r idg ing water molecule and w o u l d reduce the rate o f association o f the side chain w i t h A s p l 8 9 . 55 F i g u r e 3.3 C o m p a r i s o n o f the SI b ind ing pocket i n S G T enzymes: ( A ) the recombinant wi ld - type structure and (B) the T 1 9 0 P mutant o f S G T complexed wi th the benzamidine inhib i tor (Benz) . In the wi ld - type structure, the y - O H points towards the S1 pocket and provides a H - b o n d i n g group for the substrate. M u t a t i o n o f residue 190 to prol ine removes this H - b o n d i n g capaci ty wi thout disrupt ing the c r i t i ca l D 1 8 9 . A l t h o u g h A s p 189 adopts a s imi la r conformat ion to that found i n the wi ld- type protease, it is possible that the b ind ing o f the benzamidine inhib i tor stabil izes the conformat ion o f this side cha in through formation o f the electrostatic interaction. A n a l y s i s o f 56 the inh ib i t ion constants o f benzamidine wi th the wi ld - type recombinant protease and four mutants reveals the basis for the increased specif ici ty wi thout loss o f catalytic act ivi ty. S i m i l a r fo ld differences for the K m values for the peptide substrates and values relative to the wi ld- type are observed for a l l mutants except T 1 9 0 P . The Kj value is s ix - fo ld higher than the wi ld- type , whereas the K m values for the A r g and L y s conta in ing substrates increase 35-fo ld and 46- fo ld , respect ively (Tables 3.3 & 3.4). A s the inhib i tor forms a direct electrostatic interaction wi th A s p 189, the minor difference o f the inh ib i tory constants suggests that i n the T 1 9 0 P mutant the interaction occurs i n a more hydrophobic environment and is not accompanied by structural rearrangement o f the S1 b ind ing pocket . Whereas T 1 9 0 A demonstrates the interactions present at the S I b ind ing pocket i n the majori ty o f proteases wi th a P I A r g preference, the kinet ic analysis o f T 1 9 0 P suggests a potential intermediate i n the evolu t ion o f the vertebrate coagulat ion cascade. O n l y two natural proteases have been identif ied that possess A s p 189 and P r o 190 - hagfish pro thrombin and human k a l l i k r e i n 10, yet neither protein has been characterized wi th respect to substrate specif ici ty. B o t h genes have been characterized by D N A sequence analysis o f a number o f over lapping c D N A l ibrary clones [122-124]. T h e presence o f P r o at this pos i t ion suggests a s imi lar specif ici ty as that observed for the T 1 9 0 P o f S G T . M o r e o v e r , the residues surrounding residue 190 are ident ical to those found i n S G T . D N A sequence analysis o f the hagfish pro thrombin gene reveals that many o f the features attributed to substrate specif ici ty, such as the 60- and 99-loops, are ident ical . Hence , the reduced catalytic act ivi ty and higher K m values caused by the prol ine at this pos i t ion may be compensated by an increased concentrat ion o f the enzyme or substrate i n the b l o o d stream o f this p r imi t ive vertebrate. 57 3.3.9 Second Shell Residues A number o f catalytic or structural studies i n v o l v i n g the structure based design o f enzyme properties have demonstrated an important role for second shell residues surrounding the mutat ion o f interest [125]. In serine proteases, residue 190 interacts direct ly w i t h the side chains o f residues 16 (the N- te rminus o f the protein), 138 and 228. Res idue 228 is a h igh ly conserved tyrosine i n k n o w n t rypsin- l ike enzymes. Residues 16 and 138 are restricted to hydrophobic side chains ( V a l , He, and L e u ) . T h e differences i n kinet ic parameters observed between mutations made i n this study, rat anionic t rypsin and human trypsin (type I) are l i k e l y due to the presence or absence o f methyl groups w i th in this pai r o f residues [66,67,119]. M o r e o v e r , mutagenesis o f residue 16 i n rat anionic t rypsinogen II has been demonstrated to affect p r imary substrate specif ici ty o f the enzyme [126]. A more detailed invest igat ion o f these residues is required to understand the basis o f substrate specif ic i ty w i t h i n this f a m i l y o f important enzymes. 3.4 Conclusions O n the basis o f the ease o f product ion o f S G T i n the B. subtilis expression system, it is possible to enhance the specif ici ty at the S 2 to S 4 b ind ing pockets by either structure-based design or a directed evolu t ion strategy. Introduction o f an affinity tag, such as a hexahist idine tag, w o u l d speed the pur i f icat ion process o f the protein and facilitate characterization o f recombinant mutant proteases. T h e abi l i ty to crys ta l l ize this molecu le readi ly supports S G T as a mode l scaffold for understanding the mechanisms o f substrate specif ic i ty i n the h igh ly evo lved serine protease fami ly . 58 Chapter 4. Engineering Coagulation Factor X a Substrate Specificity into Streptomyces griseus Trypsin 4.1 Introduction 4.1.1 Overview Archi tec ture o f the active site plays a key role i n the phys io log i ca l functions o f serine proteases. A l t h o u g h the catalytic machinery is s imi lar , i f not ident ical , w i th in the f ami ly o f serine proteases, the residues compr i s ing the active site dictate function. Differences i n the active site lead to substrate specif ic i ty and the l eve l o f regulat ion by protease inhibi tors . De ta i l ed understanding o f the molecular basis o f substrate specif ici ty is needed to design inhibi tors for therapeutic applications. M o r e o v e r , the abi l i ty to tai lor protease specif ici ty to meet specia l ized needs is an attainable goa l . A t present, i t has not been established what m a x i m u m levels o f substrate specif ici ty c o u l d exist on the serine protease scaffold. F e w studies have i n v o l v e d i m p r o v i n g the extended substrate specif ici ty o f serine proteases. Proteases have been extensively characterized wi th respect to enzymat ic properties. Rates o f catalysis against numerous l ibraries o f peptide and polypept ide substrates that are both natural and synthetic i n o r ig in are readi ly avai lable [41,127-131]. Inhibi t ion constants are s imi la r ly abundant due to the importance o f phys io log ica l regulat ion and for imped ing the progression o f a pathology or prevent ing one f rom deve lop ing [132-137]. In addi t ion, a vast amount o f sequence and structural data can be found i n the databases. Indeed, the quantity o f informat ion c o m p i l e d on proteases is daunting. U l t ima te ly these data reflect properties that 59 result f rom approximately fifty amino acids that create and surround the active site o f serine proteases. Less than twenty amino ac id residues are i n v o l v e d i n enzyme-substrate interactions i n the S I peptidases [24]. A number o f residues i n v o l v e d in substrate b i n d i n g exert their influence v i a backbone contacts or by s tabi l izat ion o f the entire structure o f the protease domain . Some examples inc lude residues 214 to 216 w h i c h are i n v o l v e d in format ion o f the anti-parallel P-strand between enzyme and substrate and three disulf ide bonds between residues 42 and 58, 168 and 182, and 191 and 220 that stabil ize the entire domain . In both instances, these residues are h igh ly conserved throughout the serine protease f ami ly and do not s ignif icant ly modulate specif ici ty, but rather assist i n the enzyme-substrate interactions and stabil ize the transition state o f the catalytic process [138,139]. Thus , variat ion in substrate specif ici ty can be reduced i n complex i ty to roughly ten amino acid posi t ions. These residues are located on a l l sides o f the active site and their effects can be altered b y their l oca l environments and p r o x i m a l residues. Ex tended substrate specif ici ty is a ha l lmark o f the serine proteases o f vertebrate b l o o d coagulat ion. These proteases serve as useful models to understand substrate specif ic i ty throughout the entire f ami ly o f S I peptidases. Coagu la t ion factor X a ( F X a ) is important due to its central role i n coagulat ion and wide spread use i n biotechnology-related applications requi r ing site specific proteolysis . The predominant feature o f F X a substrate specif ici ty is a two residue loop at posi t ion 99 [140]. Insertion at this site i n the sequence posi t ions a large aromatic side cha in , T y r 9 9 , i n the S 2 pocket and restricts P 2 substrate residues to smal l G l y side chains. Less is k n o w n about the other determinants o f substrate specif ici ty i n F X a . 60 K i n e t i c data for the hydro lys is o f a number o f different peptide substrates suggests that the S3 and S 4 pockets o f F X a d isplay m i n i m a l select ivi ty [41]. The S3 specif ici ty pocket is typ ica l ly the least stringent o f the pockets throughout the S1 peptidase fami ly due to the solvent exposure o f the P3 side chain . Modera t e ly s ized hydrophobic amino acids are preferred by the S4 b ind ing pocket w h i c h is dominated by hydrophobic side chains [116]. G i v e n the selectivi ty o f these pockets, it seems possible to engineer F X a - l i k e specif ici ty into a broadly specific protease. M o r e o v e r , it is l i k e l y that a protease wi th a stricter preference for the I l e - G l u - G l y - A r g F X a cleavage sequence c o u l d be engineered. Measures o f substrate specif ici ty o f a protease can differ substantially based on the type o f substrate employed . Proteolyt ic cleavage o f a short peptide can misrepresent the rate o f hydrolys is o f a single bond in a fo lded protein. De ta i l ed characterization o f F X a us ing comprehensive peptide substrate l ibraries revealed that large, planar hydrophobic residues were preferred i n the S2 pocket over smal l residues [41,42]. Such an observation contrasts w i t h sequence analysis o f in vivo F X a substrates. A m i n o ac id sequences o f prothrombin , the predominant in vivo substrate o f F X a , f rom organisms spanning 450 m i l l i o n years o f vertebrate evolu t ion show conservation o f the I l e - G l u / A s p - G l y - A r g recogni t ion sequence [122,123]. M o r e o v e r , numerous studies have employed F X a for the site specific proteolysis o f recombinant proteins [127]. Reports o f non-specif ic proteolysis b y F X a are obv ious ly not easy to f ind i n the literature. H o w e v e r , the large number o f successful results confirms selectivi ty for the cleavage sequence is h igh . 61 4.1.2 Choice of Mutations to Mimic FXa-like Specificity Prev ious ly , I described the design o f a protease wi th a h igh p r imary specif ic i ty for A r g i n the P I pos i t ion us ing S G T . U s i n g the op t imal mutant o f S G T f rom this w o r k as a starting point, the design o f F X a - l i k e specif ici ty was pursued. T o convert the specif ici ty o f S G T into F X a , s ix mutations were int roduced into S G T to enhance select ivi ty o f the S2 , S3 and S4 pockets (Table 4.1). Muta t ions were chosen based on sequence conservation i n k n o w n F X a proteins f rom various species and inspect ion o f various x-ray crystal structures o f members o f the S I peptidase fami ly . These mutations i n S G T are i nvo lved i n enzyme-substrate interactions as w e l l as op t imiza t ion o f the overa l l architecture o f the active site (Figure 4.1). M a n y o f the mutations created also m i m i c what is found i n activated protein C ( aPC) , factor I X a (FLXa) , factor X I a , and factor V i l a (FVIIa ) . Name Mutations T 1 9 0 P T 1 9 0 P L P 99- loop, T 1 9 0 P Y P 99- loop, T 9 9 Y , T 1 9 0 P Y F P 99- loop, T 9 9 Y , N 1 7 4 F , T 1 9 0 P Y S F P 99- loop, T 9 9 Y , Y 1 7 2 S , N 1 7 4 F , T 1 9 0 P Y S F M P 99- loop, T 9 9 Y , Y 1 7 2 S , N 1 7 4 F , E 1 8 0 M , T 1 9 0 P Y S F M P E 99- loop, T 9 9 Y , Y 1 7 2 S , N 1 7 4 F , E 1 8 0 M , T 1 9 0 P , Y 2 1 7 E Table 4.1 Mutants o f S G T constructed to m i m i c the substrate specif ic i ty o f F X a . F o r the 99- loop mutation, two residues ( L y s and G l u at posi t ions 96 and 97) were inserted s imi la r to that found i n the F X a polypept ide sequence. 62 Select iv i ty o f the S2 pocket in F X a is generated by the insert ion o f two amino ac id residues at pos i t ion 99. Ex tens ion o f the polypeptide chain at this loca t ion posit ions the T y r 9 9 side cha in into the S2 pocket and restricts access to smal l aliphatic side chains [140]. T w o mutations i n S G T were required to engineer selectivity for P 2 side chains. Firs t , a two residue insert ion was created. Second, T h r 9 9 T y r was added to the loop construct to constrain the S 2 pocket i n a s imi lar fashion to F X a . Figure 4.1 Residues invo lved i n the extended substrate specif ici ty o f coagulat ion proteases. The N a + b ind ing site is k n o w n to play a role i n the substrate specif ici ty o f thrombin and l ies adjacent to the S I pocket. E 1 8 0 M was introduced into S G T to remove the electrostatic effects o f the negat ively charged G l u residue and m i m i c a residue conserved i n the entire S I fami ly peptidases. Sequence analysis shows that M e t 180 is h igh ly conserved and exists in approximately 6 0 % o f the fami ly . Howeve r , a functional role for residue 180 has not been described i n the literature. Res idue G l u 180 i n S G T is one o f several negatively charged side chains near the S4 pocket 63 that generates an overa l l negative electrostatic potential i n the region. A s hydrophobic P 4 residues are preferred i n F X a , i t was conjectured that r emova l o f the charged moie ty i n the S4 pocket w o u l d improve substrate b ind ing . Y 1 7 2 S and N 1 7 4 F mutations i n S G T were constructed to facilitate proper pos i t ion ing o f the 172-loop and increase the hydrophobic i ty o f the S4 pocket. In S G T , T y r l 7 2 is bur ied i n the core o f the structure i n a s imi lar fashion to other t ryps in- l ike enzymes [76]. H o w e v e r , Ser 172 in F X a has a conformat ion that exposes the side chain to solvent and is i nvo lved i n the pos i t ion ing o f residue 174 [141]. M u t a t i o n at the equivalent posi t ions i n bovine trypsin l ed to a f lex ib le 172-loop and suggests replacement o f residues 172 to 174 is not sufficient to create the S4 pocket o f F X a [142]. Thus , s tabi l izat ion o f the loop should be important for op t imiza t ion o f the extended substrate specif ici ty i n F X a . Y 2 1 7 E was created i n S G T to secure the 172-loop conformat ion observed i n F X a . Res idue 217 is poor ly conserved i n the S I fami ly , yet conserved i n a l l coagulat ion proteins identif ied f rom a variety o f vertebrate species [122,123]. In the crystal structure o f human F X a , residue 217 forms two charge assisted hydrogen bonds w i t h S e r l 7 3 v i a the h y d r o x y l group o f the side cha in and amide group o f the backbone [143]. Unfortunately, the G l y l 7 3 S e r mutant o f S G T has not been characterized yet. U s i n g B. subtilis as an expression host, a number o f mutants at the residues discussed above were constructed i n S G T and their enzymatic properties determined. Substrate specif ici ty was investigated by compar i son o f the rates o f hydro lys i s o f a sma l l l ibrary o f commerc i a l l y avai lable chromogenic peptides. The mutant bearing a l l seven mutations d i d not produce a protease wi th F X a - l i k e specif ici ty, but rather specif ic i ty towards coagulat ion factor X I a ( F X I a ) substrates was observed. A n intermediate mutant w i t h s ix mutations led to a 64 protease wi th moderate F X a - l i k e specif ici ty. These results conf i rm previous studies that demonstrated a role for the 99- loop and 172-loop i n F X a and related enzymes. Res idue 217 is conf i rmed as a determinant o f substrate specif ici ty. Howeve r , questions about the detailed role o f residue 217 are raised by the data. Further, residue 180 has been identif ied as a determinant o f the substrate specif ici ty i n the S I fami ly o f peptidases. 4.2 Materials & Methods 4.2.1 Plasmids, Bacterial Strains, and Growth Conditions E. coli was g r o w n us ing standard methods [90]. P l a s m i d D N A was pur i f ied us ing a Q I A p r e p spin min iprep k i t (Qiagen) and manipulated using standard protocols [90]. E n z y m e s were obtained f rom N e w E n g l a n d B i o l a b s and R o c h e M o l e c u l a r B i o c h e m i c a l s . B. subtilis strain W B 7 0 0 was g r o w n i n super-rich m e d i u m [91] or on tryptose b l o o d agar base (Di fco) at 3 7 ° C . F o r the B. subtilis car ry ing p l a smid p W B 9 8 0 [92], kanamyc in was added to a f ina l concentration o f 10 pig raL"1 i n both l i q u i d and so l id media . 4.2.2 Construction of a Hexahistidine-tagged SGT P l a s m i d S G T pET-28(a)+ was prev ious ly constructed i n an unsuccessful attempt to express recombinant S G T by c l o n i n g the gene into the X b a l and X h o l restriction endonuclease cleavage sites o f the p lasmid . Inspection o f electrospray mass spectrometry data o f aged samples o f recombinant S G T f rom B. subtilis suggested that A r g 2 4 3 was weak ly susceptible to autolysis. T h e mutat ion A r g 2 4 3 S e r was constructed us ing the Q u i k C h a n g e site-directed mutagenesis k i t (Stratagene) as suggested by the manufacturer us ing the ol igonucleot ide 5 ' - G C C T C G G C C G C C A G C A C G C T C G A G C A C - 3 ' and reverse complement 65 ol igonucleot ide . The C- te rmina l por t ion o f the S G T gene was sub-cloned f rom pET28(a )+ into b S G T p W B 9 8 0 v i a Pvu I I and A v r l l cleavage sites (Figure 4.2). Thus , S G T was c loned in frame w i t h a hexa-hist idine tag ( H i s R p W B 9 8 0 ) . A s this p l a smid construct was to be used for sub-c loning mutants o f S G T f rom the pBluescr ip t K S + E. coli p l a smid , a smal l por t ion o f the gene was deleted by digest ing wi th N a r l and self- l igation to y i e l d A N a r l H i s R p W B 9 8 0 . A s a result, successful sub-c loning o f mutants w o u l d restore protease act ivi ty. Xbal Xhol Avrll + c Xbal Xhol ^ 1 KHHOHH^  pET-28(a)+ Xbal Xhol Avrll SGTTTS^ S + Arg243Ser Hindll I Xbal A v r U H bSGT pWB980 Hindlll Xhol HisR pWB980 ^ ) Hindlll Xhol Narl ANarl HisR pWB980 F i g u r e 4.2 P l a s m i d construction for the product ion o f recombinant His- tagged S G T ( H i s R p W B 9 8 0 ) and a deletion mutant construct for easi ly ident i fy ing successful sub-c l o n i n g o f mutant S G T genes ( A N a r l H i s R p W B 9 8 0 ) . 66 4.2.3 Sequence analysis of the SI Family peptidases A l l sequences for the S I f ami ly sub-family A peptidases were downloaded f rom the M E R O P S database (http:/ /www.merops.ac.uk, release date 04-03-2002) [22]. Sequences were pre-al igned by the curators o f the database. T h e complete al ignment o f 740 proteases was pasted into M i c r o s o f t E x c e l , and then converted into single co lumns corresponding to ind iv idua l residues i n the polypept ide chain . Dis t r ibu t ion o f amino acids at each pos i t ion o f interest was obtained us ing the C O U N T I F function w i t h i n the program. T h e f i le served as a useful database that l i nked amino acids at a defined posi t ion i n the polypept ide sequence to characterized enzymes. Reg ions where insertions or deletions occurred i n the fami ly were not handled w e l l by this method and inspect ion o f k n o w n crystal structures and publ i shed literature were required. 4.2.4 DNA Manipulation U s i n g the previous ly described S G T gene c loned into pBluesc r ip t K S + p lasmid (Chpt. 2.2.2), mutagenesis was performed on the gene us ing a Q u i k C h a n g e site-directed mutagenesis k i t (Stratagene) as described by the manufacturer. Ol igonucleot ides used for mutagenesis are p rov ided i n Tab le 4.2. D N A sequence analysis o f the c loned gene and mutants was performed us ing the B i g D y e Terminator k i t and analyzed on an A B I 3700 D N A Sequencer ( A p p l i e d Biosys tems) . Mutan t S G T genes were sub-cloned into p l a smid A N a r l H i s R p W B 9 8 0 v i a H i n d l l l and Pvu I I restrict ion sites and transformed into B. subtilis W B 7 0 0 by the method o f S p i z i z e n [93]. 67 Mutation Oligonucleotide 9 9 - L o o p 5 ' - C A G G C C C C C C G G C T A C A A C A A G G A G G G C A C C G G C A A G G A C T G G - 3 ' T 9 9 Y 5 ' - C A A G G A G G G C T A C G G C A A G G A C - 3 ' E 1 8 0 M 5 ' - C T C G T G G C C A A C G A G A T G A T C T G C G C C G G A T A C - 3 ' N 1 7 4 F 5 ' - T C C G C G T A C G G C T T C D G A G C T C G T G G C C - 3 ' Y 1 7 2 S 5 ' - G C C G C T C C G C G T C C G G C T T C G A G C T - 3 ' N 1 7 4 F Y 2 1 7 E 5 ' - A G C T G G G G C G A G G G C T G C G C C - 3 ' Table 4.2 Ol igonucleot ides used to mutate the S G T gene to m i m i c residues found i n F X a . T h e reverse complement sequences o f these ol igonucleot ides were also used i n the mutagenesis. 4.2.5 Purification of His-tagged SGT and Mutants thereof Overn igh t B. subtilis 20 m L cultures were used to inoculate 250 m L o f super-rich broth conta ining 10 [ i g / m L k a n a m y c i n and g r o w n for 16 hrs at 3 7 ° C . The supernatant was harvested by centrifugation (30 min . , 5000 rpm) and then passed over a T a l o n affinity resin c o l u m n ( B D Biosc iences ) (10 c m x 0.75 cm) equil ibrated i n wash buffer (50 m M sod ium phosphate, 500 m M N a C I , p H 8.2). The c o l u m n was washed wi th 10 c o l u m n volumes o f wash buffer and the recombinant protein eluted wi th wash buffer conta ining 100 m M imidazo le . A c t i v e fractions conta ining recombinant protein were concentrated us ing a 10,000 N M W L Ultrafree-4 centrifugal filter unit ( M i l l i p o r e ) and d ia lyzed wi th 100 m M T r i s - H C I , 150 m M N a C I , 20 m M CaCl2, p H 7.6 i n the same unit. Recombinan t proteins were stable at 4 ° C for months. Pro te in quantif ication was ident ical to that described prev ious ly (Chpt. 2.2.3). 68 4.2.6 Characterization of Substrate Specificity K i n e t i c analysis was performed i n 10 m M T r i s - H C I buffer conta ining 150 m M N a C I , 20 m M C a C l 2 , and 0.1 % P E G 8000, p H 7.6 at 2 5 ° C . React ions (300 uT) were prepared i n 96-w e l l microplates us ing either a R o b o S e q 4204 or R o b o G o laboratory automation system ( M W G B i o t e c h A G , Ebersburg , Germany) and measured us ing a Labsystems M u l t i s k a n Ascen t plate reader. Peptide substrates were handled as suggested by their manufacturers (Diapharma, A m e r i c a n Diagnost ica) . Assessment o f specif ici ty us ing chromogenic peptide substrates was estimated by direct analysis o f the hydro lys is o f each substrate at 40 \iM at two stages o f the puri f icat ion. In this method, the specif ic i ty constant (kc a t /Km) was determined f rom the slope o f the natural logar i thm o f substrate remain ing as funct ion o f t ime. Deta i l ed kinet ic analyses were performed on a m i n i m u m o f s ix substrate concentrations ranging f rom 20 to 600 p M and enzyme concentrations o f 10 to 70 n M . H i g h e r substrates concentration were not examined due to solubi l i ty diff icul t ies w i t h the peptides. 4.2.7 Macromolecular Substrate Specificity H u m a n prothrombin and F X a were purchased f rom Haemato log ica l Technologies . P ro th rombin (3.5 pg) was digested w i t h mutants o f S G T (60 ng) and F X a (60 ng) overnight at r o o m temperature i n 10 m M T r i s - H C I buffer conta ining 150 m M N a C I , 20 m M C a C b , and 0.1 % P E G 8000, p H 7.6. Proteolyt ic fragments were resolved by S D S - P A G E f o l l o w i n g a standard pro toco l [90]. 69 4.3 Results & Discussion 4.3.1 Production of His-tagged S G T In order to facilitate pur i f ica t ion and analysis o f a large number o f mutants, s impl i f ica t ion o f the four step pur i f icat ion scheme described prev ious ly was required (Chpt. 2.2.3). Init ial attempts to produce recombinant S G T i n E. coli were unsuccessful , yet produced a construct bearing S G T i n frame wi th a hexa-hist idine tag i n the pET-28a(+) p lasmid . Several his t idine residues i n succession promote b i n d i n g to metal ions such as C u + 2 , +2 +2 Ni , or C o . A s the metal ions can be i m m o b i l i z e d onto an appropriate chromatography m e d i u m , capture o f recombinant protein f rom complex samples is greatly s impl i f i ed . The C -terminal region o f S G T bearing the tag was amenable for subc lon ing f rom pET-28a(+) into the B. subtilis p l a smid p W B 9 8 0 through a fortuitous A v r l l restriction site. T o ensure the tag remained on the protease, the A r g 2 4 3 S e r mutat ion was introduced into S G T to protect against potential autolytic cleavage. Y i e l d s o f S G T bearing a hexa-hist idine tag at the C-terminus o f the protein were three to f ive times lower than the non-tagged construct. M a x i m a l yie lds o f recombinant protein bear ing the tag d i d not exceed 5 m g / L o f culture compared to 15 m g / L for the wi ld - type construct. B i n d i n g o f the recombinant protease to various c o m m e r c i a l l y avai lable N i + 2 or +2 Co -chelated resins was poor suggesting that the tag was part ia l ly bur ied i n the protein. Inspection o f the crystal structure o f S G T suggests that the first two H i s residues m a y l ie i n a sha l low cleft on the enzyme surface. H o w e v e r , h igh puri ty protein resulted f rom the pur i f icat ion (Figure 4.3). The reduced y i e l d o f S G T bearing the tag m a y hinder fo ld ing o f the enzyme. Al te rna t ive ly , the tag may promote interactions wi th the negatively charged pept idoglycan found on the c e l l w a l l o f the bacteria. H o w e v e r , the ease o f protein pur i f icat ion 70 based on the tag outweighed the demand for greater amounts o f protein. Y i e l d s o f each o f the mutants were s imi lar suggesting the mutations were not detrimental to protein fo ld ing . These results further support B. subtilis as an excel lent expression host for the p roduc t ion o f proteases and other proteins that are diff icul t to produce in E. coli. A B C D m -23 kDa F i g u r e 4.3 Pur i f ica t ion o f a typ ica l His- tagged mutant o f S G T f rom B. subtilis culture us ing T a l o n metal affinity resin. Lane A , 1 m L supernatant; L a n e B , 1 m L c o l u m n f low through; L a n e C , 1 m L wash buffer. Lane D : concentrated recombinant protein (2 Hg). Samples were concentrated us ing tr ichloroacetic ac id in lanes A-C. 4.3.2 T e c h n i q u e s f o r C h a r a c t e r i z a t i o n o f Subs t r a t e S p e c i f i c i t y o f S e r i n e Pro teases Character izat ion o f the substrate specif ici ty o f wi ld - type F X a us ing comprehensive libraries o f smal l peptide substrates revealed the select ivi ty o f the protease is not strict [41,42]. Interestingly, the P 2 preference for large, planar hydrophobic substrates over G l y side chains has been noted i n these studies. Moreover , P3 and P 4 side chains are w e a k l y selected for . O n the basis o f this specif ic i ty , F X a is thought to resemble a l o w ef f ic iency trypsin rather than a h igh ly selective th rombin [41]. One drawback o f the l ibraries used to analyze the substrate specif ici ty is the potential error caused by differences in conformat ion o f each o f the peptides in w h i c h each peptide may adopt differing conformations and lead to bias. 71 In the present study, we have used commerc ia l substrates that differ f rom the natural polypept ide substrates or short synthetic peptides composed o f L - a m i n o acids. These substrates show enhanced specif ic i ty for members o f the coagula t ion cascade and have been w i d e l y used i n c l i n i c a l applicat ions. Non-s tandard amino acids are present i nc lud ing op t imal b l o c k i n g groups at their N - t e r m i n i to generate h igh ly specific substrates (Figure 4.4). C o m p a r i s o n o f the rates o f hydrolys is o f these substrates shows that the mutations created i n S G T have altered the active site geometry and substrate specif ici ty o f the enzyme. 4.3.3 Extended Substrate Specificity of SGT W i t h the T 1 9 0 P mutant o f S G T as the starting point, extended substrate specif ic i ty at the S 2 to S4 posit ions was introduced by substitution o f residues found i n F X a . T h e T 1 9 0 P mutant was chosen over the T 1 9 0 A mutation for the higher K m values. A s the K m value for most peptides bear ing the T 1 9 0 P mutant was approximately 50 to 100 p M , chromogenic substrates w o u l d be useful i n the kinet ic analysis o f mutants. In contrast, the T 1 9 0 A mutant o f S G T displayed K m values be low 10 u M for each substrate. K i n e t i c dissect ion o f subsequent mutants based on the T 1 9 0 A mutat ion w o u l d require the use o f fluorescent substrates that are less c o m m o n l y used for coagulat ion proteases. M o r e o v e r , the T 1 9 0 P mutant permitted estimation o f the specif ici ty constant ( k c a t / K m ) us ing a direct interpretation o f the rate o f hydro lys i s o f peptide substrate. In this method, a single substrate concentration that is w e l l be low the K m value is hydro lyzed and the rate measured by spectrophotometry. T h e natural logar i thm o f substrate remain ing as funct ion o f t ime is plotted as a straight l ine whose slope approximates k c a t / K m . A number o f mutants o f S G T were characterized by this method and it was invaluable for determining successful alterations o f substrate specif ici ty. 72 S-2222 S-2238 Bz-lle-Glu(y-OR)-Gly-Arg-pNA H-D-Phe-Pip-Arg-pNA R=H (50%) and R=CH3(50%) S-2288 S-2302 H-D-lle-Pro-Arg-pNA H-D-Pro-Phe-Arg-pNA S-2366 pyroGlu-Pro-Arg-pNA F i g u r e 4.4 Substrates used to characterize mutants o f S G T wi th altered substrate specif ici ty (Diapharma, W e s t Chester, O h i o ) . Non-s tandard amino acids are incorporated into the substrate for improved d iscr iminat ion among proteases. 73 4.3.4 Effect of the 99-loop on the Substrate Specificity of SGT Substrate specif ici ty o f the S 2 b ind ing pocket i n coagulat ion proteases is affected by the presence o f a two or three amino ac id insert ion termed the 99- loop. In F X a , the insertion o f two residues facilitates pos i t ion ing o f T y r 9 9 and restricts access to the S 2 pocket. Character izat ion o f substrate selectivi ty o f S G T revealed preferences inherent w i th in the protease (Figure 4.5). Spectrozyme P C a and S-2366 were preferred two-fo ld over a l l other substrates u t i l i zed . B o t h substrates have P r o at P 2 and this l i k e l y leads to presentation o f the A r g side cha in i n a conformation more favorable for hydro lys is . Importantly, the F X a preferred substrates were not preferred by the T 1 9 0 P mutant o f S G T . T w o mutants were designed to show the importance o f this loop i n generating select ivi ty i n the S 2 pocket (Figure 4.5). Firs t , two residues, L y s 9 6 and G l u 9 7 , were introduced to lengthen the 99- loop i n S G T (denoted " L P b S G T " ) . A l l substrates examined were hydro lyzed w i t h h i g h l y s imi la r react ion rates to the T 1 9 0 P mutant us ing L P b S G T . A s this mutant presented the smaller, hydroph i l i c side cha in T h r at pos i t ion 99, no selectivity i n the S 2 b ind ing pocket was anticipated. B o t h o f the inserted residues should orient their carbonyl oxygens into the S3 /S4 pocket i f the loop has the same conformat ion observed i n other proteases. Introduction o f T 9 9 Y led to a non-specific protease hav ing s imi lar preference for a l l substrates. These results suggested that the 99- loop was i n a s imi la r conformat ion to that observed i n F X a and that addi t ional determinants o f substrate specif ici ty were needed to reconstitute the desired specif ic i ty . No tab ly , substrates bearing P 2 G l y were more effect ively hyd ro lyzed suggesting a s imi lar conformat ion o f the loop to F X a . Structural s imi lar i ty o f the F X a 99- loop is observed i n a P C . T h i s l i k e l y explains the maintained preference for a P C preferred substrates (Spectrozyme P C a and S-2366) i n each o f 74 the mutants characterized, i nc lud ing the mutant L P o f S G T . In a P C , residue 99 is also T h r . M u t a t i o n o f this pos i t ion i n a P C wi th substitutions found i n F X a led to a protease wi th a s imi lar a substrate specif ic i ty and inhibi tory profile to F X a . In the same study, the opposite mutations i n F X a ( Y 9 9 T ) led to a s imi la r swi tch ing o f specif ic i ty [140]. Importantly, as determined through peptide substrates, substrate specifici ty d i d not correlate w i t h the hydrolys is o f macromolecular substrates and demonstrates the c ruc ia l role for addit ional protein-protein interactions i n the coagulat ion proteases. M S-2266 • S-2366 • S-2222 • SpectrozymeFXa B SpectrozymePCa • S-2302 T190P Figure 4.5 N o r m a l i z e d k c a t / K m values for the T 1 9 0 P , L P and Y P mutants o f S G T . V a l u e s derived f rom independent measurements done in tr iplicate ( ± 10% S .D. ) . W e a k preference for a P C substrates is exhibi ted by the enzyme in i t i a l ly . Introduction o f the two residue loop and T 9 9 Y generates a broadly specific protease. A number o f other serine proteases have insertions at pos i t ion 99. La rge insertions (greater than 5 residues) are found in several ka l l ik re ins , the C l s protease, as w e l l as complement factor B [144,145]. Compared to S G T , other proteases have a two or three residue insert ion at this region inc lud ing thrombin and pancreatic elastase II. In each o f these enzymes, the loop plays a role i n determining the substrate specif ic i ty o f the enzyme o f both the S2 and S3 substrate b i n d i n g pockets. Increasing the length o f the inser t ion correlates wi th 75 decreasing catalytic eff ic iency [145]. Futures studies for engineering substrate specif ic i ty should apply variable lengths o f amino ac id insertions and composi t ions o f the 99- loop. 4.3.5 Mechanisms of P3 Selectivity in SI Family Peptidases Select iv i ty for P 3 residues is poor i n nearly a l l S I peptidases. The enzyme-substrate interaction is l imi t ed due to solvent exposure o f the P 3 side cha in . In F X a , P 3 b ind ing is generated by the side chains o f residues 192 and 215. Throughout the entire f ami ly o f S I peptidases, residue 215 is h igh ly conserved as a large planar side cha in ( W , F , Y ) . A l l crystal structures determined to date have shown the side cha in i n a conformat ion that borders the S4 specif ici ty pocket (Figure 4.6). H o w e v e r , the backbone carbonyl group o f residue 215 is i n v o l v e d i n a hydrogen bond wi th the P 3 residue o f the substrate. In the crystal structures o f wi ld- type recombinant b S G T and T 1 9 0 P mutant, residue 192 displayed a h igh degree o f f l ex ib i l i ty and was mode led as two conformations. F l e x i b i l i t y o f this residue has been demonstrated i n several coagulat ion proteases and the alternate conformations facilitate interactions wi th the P 3 and P 2 ' specif ic i ty pockets (Figure 4.7). A s both S G T and F X a both possess G i n residues at pos i t ion 192, mutat ion was not required. H o w e v e r , future w o r k should i n v o l v e mutagenesis o f this residue due to its invo lvement i n substrate specif ici ty and inh ib i t ion by protease inhibi tors . Res idue 192 has evo lved to not disrupt substrate b ind ing and plays a role i n enzyme inh ib i t ion . Studies i n v o l v i n g protein C , F X a , and thrombin have shown that residue 192 is important i n protease-inhibitor contacts and is a key basis for differential inh ib i t ion [146-148]. M e t / G l u / G l n substitutions at pos i t ion 192 in coagulat ion factor F X a d i d not s ignif icant ly alter substrate specif ici ty us ing peptide substrates s imi lar to those used i n this study [143]. 76 Presumably, mutation o f the residue to a L y s side chain i n S G T w o u l d generate an increased preference for ac idic side chains at P 3 . Unfortunately, L y s 192 has not been int roduced into any S I peptidases by site-directed mutagenesis yet the residue exists natural ly i n some members o f the fami ly . 170-1 Figure 4.6 S3 & S4 b ind ing pockets o f F X a . G i n 192 plays a l im i t ed role i n determining the select ivi ty o f the S3 pocket s temming f rom inherent f l ex ib i l i t y o f the side chain . In the F X a structure without a substrate, the side cha in points away f rom S3 pocket. The mode l depicted is the superimposi t ion o f F X a ( P D B I D 1 H C G ) and F V I I a i n complex w i t h 1 ,5 -dansy l -Glu -Gly-Arg-ch lo romethy l ketone ( P D B I D 1 C V W ) , wi th the F V I I a structure hidden. 77 F i g u r e 4.7 S3 and S4 b ind ing pockets o f F V I I a . In the crystal structure o f human coagulat ion factor V i l a in complex wi th 1 ,5 -dansy l -G lu -Gly -Arg -ch lo rome thy l ketone ( P D B I D 1 C V W ) , L y s 192 interacts w i th the negat ively charged side cha in i n P 3 o f the inhibi tor . Sequence analysis o f the S1 peptidases reveals a subset o f proteases that conta in pos i t ive ly charged side chains at posi t ion 192. These proteases show S3 select ivi ty for ac id ic P3 residues. No tab ly , v e n o m b i n A and b i l ineob in from the moccas in snake (Agkistrodon bilineatus) have L y s 192 [149,150]. These proteases are found i n the v e n o m o f the snake and 78 m i m i c F X a by act ivat ing thrombin at the same posi t ion i n the polypept ide. A l t h o u g h crystal structures o f these proteases have not been reported, molecula r mode l ing o f b i l i neob in suggests that residue 192 occupies an ident ica l pos i t ion to that observed i n F X a , thrombin and S G T [151]. Crea t ion o f an electrostatic interaction between the substrate and enzyme may decrease the rate o f deacylat ion dur ing catalysis. Rates o f substrate hydro lys i s by L y s 192 bear ing proteases are 10-fold lower than observed for the coagulat ion proteases. Thus , the Q 1 9 2 K mutant o f S G T should be characterized and m a y y i e l d a stronger preference for ac idic side chains at P 3 wi th a concomitant reduction i n catalytic eff iciency. H o w e v e r , it is l i k e l y that the h igh f l ex ib i l i ty o f residue 192 w i l l l i m i t the m a x i m a l str ingency for P 3 side chains. T w o mechanisms have evo lved to increase the selectivity o f the S3 pocket i n the S I f ami ly o f peptidases. In rat mast c e l l protease II, the absence o f a disulphide bond between residues 191 and 220 generates addi t ional enzyme-substrate contacts and enlarges the S3 pocket [152]. P 3 selectivity can also be generated by a secondary protein as evident i n the structure o f staphylokinase i n complex wi th the protease domain o f p lasmin . Staphylokinase acts as a co-factor and inserts several side chains into the S3 /S4 pocket. A s a result the substrate specif ici ty and inhib i tory prof i le o f p l a smin are drast ical ly altered [153]. Thus , the serine protease scaffold is h igh ly amenable for further engineering o f selectivi ty i n the S3 b i n d i n g pocket. 4.3.6 Role of the 172-loop and Residue 217 in the SI Peptidases In order to fo rm the S4 pocket i n F X a , a stretch o f residues f rom 172 to 174 adopt a conformat ion dist inct f rom that observed i n broadly specific t rypsin l i k e enzymes (Figure 4.8). In S G T and other non-specif ic serine proteases, a large hydrophobic side chain at residue 79 172 buries i tself into core o f the enzyme. In turn, the backbone o f residues 173 and 174 adopts a conformat ion that exposes the side chains o f these residues away f r o m the S 4 pocket. In F X a , F I X a , thrombin and a P C residue 172 is Thr , M e t , or Ser. In these proteases the 172-loop adopts an "up" conformat ion and bounds the S4 pocket. In order to generate the proper conformat ion o f the 172-loop, two mutations were made i n S G T : Y 1 7 2 S and N 1 7 4 F . F i g u r e 4.8 Confo rma t ion o f the 172-loop i n S I peptidases: S G T ( A ) , F X a ( B ) , a P C (C) and F V I I a (D) . In F X a and a P C , the 172-loop adopts a conformat ion such that residue 172 is not bur ied in the core o f the enzyme w h i c h is evident in a l l structures o f broad specif ic i ty t ryps in- l ike enzymes. F V I I a has a large insertion at residue 170, and generates substrate specif ici ty i n a different manner than other coagula t ion proteases. 80 Introduction o f N 1 7 4 F alone into S G T y ie lded a protease wi th select ivi ty for S-2366, a substrate designed for quantif icat ion o f F X I a and a P C . B a s e d on the crystal structure o f T 1 9 0 P S G T , N 1 7 4 F should not affect the specif ic i ty o f the S4 pocket unless the conformat ion o f the 172-loop is distorted. T h e alteration o f substrate specif ic i ty may be due to the Phe side cha in at pos i t ion 174 replacing residue T y r l 7 2 its bur ied conformation. Subsequent mutagenesis o f Y 1 7 2 S suggested that Phe 174 was i n a bur ied conformat ion as it resulted i n a very litt le change i n substrate specif ici ty compared to the Y F P mutant o f S G T (Figure 4.9). A recent study reported the crystal structure o f rat anionic t rypsin w i t h s imi la r mutations i n the 172-loop, and suggests the basis for a P C and F X I a - l i k e specif ici ty [142]. In the crystal structure o f a mutant o f rat anionic t rypsin w i t h the Ser-Ser-Phe sequence substituted at the 172 to 174 posi t ions, the 172 loop adopts a nove l conformat ion i n w h i c h Phe 174 buries i nward i n the structure o f the enzyme s imi lar to that anticipated i n the mutants o f S G T [142]. A s a result, the conformat ion o f the side chains at posit ions 172 and 173 do not m i m i c those observed i n F X a even though the polypept ide sequence is the same. The S4 pocket is enlarged as the 172-loop extends farther away f rom the structure. I f a s imi la r conformat ion occurred i n the Y F P and Y S F P mutants o f S G T then the preference for substrates w i t h larger side chains at P 4 might be more favored. B o t h S-2366 and S-2266 have large hydrophobic groups at P 4 (pyroglutamic ac id and D-va l ine , respect ively) . 81 1.00 090 • S-2266 • S-2366 • S-2222 080 0 70 0 60 050 0 40 0 so 0 ? 0 0 10 000 Y F P Y S F P F i g u r e 4.9 N o r m a l i z e d kcJKm values for the Y F P and Y S F P mutants o f S G T . Introduction o f Y 1 7 2 S y ie lded no substantial change i n substrate specif ic i ty . V a l u e s der ived f rom independent measurements done i n triplicate (± 10% S .D. ) . A d d i t i o n o f the E 1 8 0 M mutat ion to the Y S F P S G T mutant ( Y S F M P ) l ed to s ignif icant improvement i n F X a - l i k e specif ici ty (Table 4.3). In any k n o w n crystal structure o f an S I fami ly peptidase, the M e t side chain at residue 180 makes no direct contacts w i th the substrate or any component o f the protein. A l imi ted diversi ty o f amino ac id variat ion is observed at posi t ion 180 i n the S I peptidases wi th a M e t residue present i n - 6 0 % o f the a l l proteins i n the fami ly . O n l y a few o f S I peptidases bear a posi t ively charged side cha in , such as L y s or A r g , at this posi t ion. Pos i t ion ing o f residue 180 is achieved through a P-hairpin formed by residues 177 to 180. T h e N H group o f residue 180 forms a hydrogen bond w i t h the C = 0 backbone o f residue 177. In S G T , the carboxylate moei ty o f G l u 180 also forms a hydrogen bond wi th the N H group o f the amide backbone o f V a i 177. Muta t ion o f residue 180 should alter the electrostatic environment o f the S3 pocket as w e l l as permit an alternate conformat ion o f the 172-loop. Dissec t ion o f the kinet ic constants o f the Y S F M P mutant bear ing the E 1 8 0 M mutat ion indicates moderate reconsti tution o f F X a - l i k e properties. B o c - L e u - G l y - A r g was the most 82 specific substrate for the mutant protease indica t ing the S3 pocket is hydrophobic i n the mutant protease. Substrates designed for the quantif icat ion o f F X a had the lowest K m values o f a l l substrates (S-2222 and Spect rozyme F X a ) . H o w e v e r , substrates w i t h the lowest K m values were not associated w i t h the highest turnover numbers. Ef f ic ien t turnover o f n o n - F X a preferred substrates s t i l l occurred suggesting the extended b ind ing pocket o f F X a was not present. B a s e d on the observed properties o f the Y F P , Y S F P and Y S F M P mutants o f S G T , s tabi l izat ion o f the 172-loop was impl ica ted as be ing necessary for generating F X a - l i k e extended substrate specif ici ty. Further, crystal lographic analysis is required to observe the conformat ion o f the 172-loop caused by these mutations. Substrate K m (uM) kcat (min) kc a t / K m (uM" min") FXa W K m (uM min") L G R 460 4664 10.1 -B oc-Leu-Gly-Arg-pN A ± 1 2 ± 8 1 1 S-2222 241 1765 7.3 4.9 Bz-Ue-Glu(Y-OR)-Gly-Arg-pNA (R=-HorCH3at50%) ± 3 6 ± 9 5 9 S-2366 511 3683 7.2 -pyroGlu-Pro-Arg-pNAHCl ± 6 7 ± 1209 Spect rozyme P C a H-D-Lys(g-Cbo)-Pro-Arg-pNA 623 ± 139 3788 ± 6 2 0 6.1 4.1 Spect rozyme F X a M-D-CHG-Gly-Arg-pNA 356 ± 2 0 1800 ± 2 4 0 5.1 18.2 S-2266 2271 4017 1.8 -H-D-Val-Leu-Arg-pNA ± 4 6 8 ± 1687 S-2238 1396 1056 0.8 1.2 H-D-Phe-Pip-Arg-pNA ± 8 0 ± 3 9 8 S-2302 1467 512 0.3 -H-D-Pro-Phe-Arg-pNA ± 8 0 ± 2 2 0 S-2251 1857 527 0.3 -H-D-Val-Leu-Lys-pNA ± 103 ± 2 6 0 Table 4.3 Steady-state k inet ic parameters for the hydrolys is o f a series o f p-ni t roani l ide chromogenic substrates by the Y S F M P mutant o f S G T compared to that observed wi th F X a under s imi lar react ion condi t ions. ( F X a data f rom ref. [140]). V a l u e s obtained i n triplicate ( ± S .D . ) . 83 W i t h i n the coagulat ion proteases, residue 217 plays a role i n substrate selectivity. Muta t ions at this pos i t ion have been characterized i n thrombin, coagulat ion factor V i l a (FVI Ia ) , and coagulat ion factor I X a ( F I X a ) [143,154,155]. In these studies, residue 217 was described as a determinant o f P 2 / P 3 selectivi ty v i a formation o f direct contact w i t h the substrate. Conc lus ions drawn were based on analysis o f substrate specif ic i ty us ing on ly a few peptide substrates and p r imar i ly through enzyme- inhib i tor interactions evident i n crystal structures. The latter can lead to misrepresentation o f the true funct ion o f peripheral residues o f the active site, such as residue 217, as the interaction between enzyme and inhib i tor is typ ica l ly far tighter than for a natural substrate. M o r e o v e r , later studies established the s ignif icance o f a sod ium b ind ing site i n several o f the coagulat ion factor proteases. B i n d i n g o f a sod ium i o n improves the catalytic eff ic iency o f the enzyme b y structural changes i n the protease domain . Res idue 217 has been impl ica ted i n s tabi l izat ion o f the i o n b i n d i n g site [156]. Inspection o f crystal structures o f F X a , V i l a , and thrombin i n complex w i t h peptide based inhibi tors supports neither o f these theories comple te ly [141,157,158]. Res idue 217 is typ ica l ly a negatively charged side-chain throughout the S I f ami ly o f peptidases, yet this residue is a T y r i n S G T . In several crystal structures o f coagulat ion proteases, the ca rboxy l group o f the G l u side chain is i n v o l v e d i n the formation o f two charge assisted hydrogen bonds w i t h residue 173 i n the 172-loop. Hence , it was postulated that introduct ion o f this side cha in w o u l d lead to stabil izat ion o f the loop i n an upwards conformat ion and enhance F X a -l i k e extended substrate specif ici ty. K i n e t i c analysis o f the Y S F M P E mutant o f S G T bearing the G l u 2 1 7 y ie lded a protease more s imi la r to F X I a and not F X a (Table 4.4). L i t t l e is k n o w n about the substrate specif ici ty o f F X I a in vitro and no crystal structure has been reported for the protein. Inspection o f the 84 polypept ide sequence o f human F X I a shows the presence o f M e t l 8 0 and G l u 2 1 7 , and a three residue insert ion loop at pos i t ion 99 that w o u l d present either Ser or G l y into the S 2 pocket [159]. Crys ta l lographic analysis is required to determine whether the introduced G l u 2 1 7 side cha in adopts a conformat ion that restricts the P 4 pocket or whether it forms an electrostatic interaction w i t h A r g 2 2 2 . A s i n the Y S F P and Y S F M P mutations, the T y r l 7 2 i n the polypept ide sequence o f F X I a may possess a 172-loop i n the down conformation. Hence , a l l o f the determinants thought to generate F X a - l i k e specif ic i ty are present i n F X I a w i t h the except ion o f T y r 9 9 and may exp la in substrate specif ici ty o f the Y S F M P E mutant o f S G T . Substrate K m (uM) kcat (min) kcat/ K m (uM" min") FXa kcat/ K m (uM" min") S-2366 pyroGlu-Pro-Arg-pNAHCl 526 ± 2 3 6691 ± 1409 12.7 -S-2302 H-D-Pro-Phe-Arg-pNA 568 ± 3 9 5487 ± 1569 9.7 -Spect rozyme F X a M-D-CHG-Gly-Arg-pNA 167 ± 3 7 819 ± 180 4.9 18.2 S-2266 H-D-Val-Leu-Arg-pNA 488 ± 3 3 1624 ± 3 4 3 3.3 -Spect rozyme P C a H-D-Lys(g-Cbo)-Pro-Arg-pNA 1952 ± 3 4 5870 ± 1 5 5 3 3.0 4.1 L G R Boc-Leu-Gly-Arg-pNA 177 ± 2 2 196 ± 8 6 1.1 -S-2238 H-D-Phe-Pip-Arg-pNA 316 ± 3 9 198 ± 8 6 0.6 1.2 S-2222 Bz-ne-Glu(y-OR)-Gly-Arg-pNA (R = H or C H 3 at 50%) 2341 ± 126 1066 ± 7 0 1 0.5 4.9 Table 4.4 Steady-state k inet ic parameters for the hydrolys is o f a series o f p-ni t roani l ide chromogenic substrates by the Y S F M P E mutant o f S G T compared to that observed w i t h F X a under s imi lar react ion condi t ions . ( F X a data f rom ref. [140]). V a l u e s obtained i n triplicate ± S . D . 85 4.3.7 Additional Elements Needed for Reconstructing FXa-like Specificity A d d i t i o n a l mutations must be made to improve further F X a - l i k e selectivi ty o f the S G T mutants constructed i n the present study. Non-add i t ive effects o f mutations are c o m m o n for mutations that are residues i n c lose p r o x i m i t y [153,160,161]. Hence , the absence o f one specif ici ty determinant m a y hinder the abi l i ty o f other residues to function properly. Studies attempting reconsti tution o f F X a specif ic i ty i n other proteases may y i e l d clues as to what element is absent f rom the Y S F M P E mutant o f S G T . Unfortunately, these studies for the most part have focused exc lus ive ly on the 99- loop [70]. Ra t anionic trypsin ( R A T ) was mutated i n several o f the regions characterized i n this study but d i d not lead to an enzyme w i t h F X a - l i k e properties [142]. Muta t ions o f residue 190, the 99- loop, and the 172-loop were combined i n R A T but were characterized on the basis o f enzyme inh ib i t ion rather than substrate specif ici ty. Inh ib i t ion constants (K;) o f inhibi tors o f F X a were typ ica l ly 10-fold higher w i t h the combined mutant o f R A T . O n the basis o f structure data, certain inhibi tors c o u l d assist the s tabi l izat ion the 172-loop i n the proper upwards conformation. Based on the data f rom S G T mutants i n the present study, i t appears that the conformat ion o f the 172-loop is the key ingredient mi s s ing for reconsti tution o f the extended substrate specif ici ty o f F X a . Several possibi l i t ies exist for further s tabi l izat ion o f the 172-loop. Introduction o f Y 2 1 7 E was unsuccessful at i m p r o v i n g the selectivi ty o f the Y S F M P mutant o f S G T and m a y be due to the requirement o f a h y d r o x y l group at the side cha in o f residue 173, w h i c h is a G l y i n S G T . Residues w i th in a 6 A radius o f residues 172 to 174 i n F X a inc lude residues 167 to 176, 182, 215 to 217, 224 and 227. C o m p a r i s o n o f the crystal structures o f S G T and F X a show that most o f these residues are ident ical and exist i n h igh ly s imi la r conformations i n 86 both proteases. Substitutions o f the residues adjacent to the 172-loop i n the polypept ide sequence may further stabil ize the loop. H o w e v e r , the largest differences i n structure are evident at residues 224 and 227. Importantly, these residues are i n v o l v e d i n the format ion o f the N a + b i n d i n g site i n F X a . It is possible that introduct ion o f a N a + b ind ing site into Y S F M P E S G T w i l l reconstruct the enzymatic properties o f F X a . If the site were present, bur ia l o f residue 172 or 174 into the core o f the enzyme c o u l d be disfavored. Importantly, the Y S F M P E mutant has kinet ic properties s imi la r to F X I a w h i c h also does not possess a N a + b i n d i n g site. Sequence analysis o f the coagulat ion proteases indicates that i n their evolut ionary past, a l l other coagula t ion proteases had a N a + b ind ing site. Site specific sod ium b i n d i n g has been demonstrated to p lay a key role i n the catalytic eff ic iency o f the coagulat ion factor proteases i nc lud ing F X a , a P C , and thrombin. D i C e r a demonstrated the c ruc ia l role o f residue 225 i n the serine proteases for b ind ing a single sod ium ion near the active site o f the enzyme [162]. B i n d i n g o f sod ium and no other a lka l i metal to these enzymes generates a 3- to 5-fold increase i n catalytic eff iciency. In nearly a l l serine proteases, residue 225 is occupied by a P ro and these proteases do not demonstrate increased catalytic act ivi ty i n the presence o f sod ium ions. H o w e v e r , i n the coagulat ion factor proteases, this residue is Ty r . Absence o f P ro at pos i t ion 225 a l lows proper pos i t ion ing o f the carbonyl group o f the preceding residue i n the polypept ide sequence to b i n d the N a + i on [77]. In addi t ion to residue 225, an extensive network o f hydrogen bonds and water molecules facilitates s tabi l izat ion o f a sod ium i o n i n a pos i t ion that is very near the S1 b ind ing pocket (Figure 4.10). M u c h research has been devoted to the characterization o f the i o n b ind ing site i n thrombin, activated protein C and coagulat ion factor X a [156,163,164]. D i s rup t ion o f any component o f the ion b ind ing site readi ly abolishes b ind ing and catalytic act ivi ty . T h e 87 complex i ty o f these enzymes has l imi ted our abi l i ty to understand the structural and thermodynamic effects o f sod ium b ind ing . However , the crystal structure o f thrombin in the absence o f N a + suggests that removal o f the ion destabilizes the entire S I pocket and changes the conformat ion o f the C y s l 6 8 - C y s l 8 2 disulf ide bond [165]. The S I pocket l ies adjacent to residue 172 i n S G T or residue 174 in Y 1 7 2 S N 1 7 4 F mutants o f S G T i n the bur ied conformation. Therefore, the conformat ion o f the loop c o u l d be altered b y the presence o f an occupied N a + - b i n d i n g site F i g u r e 4.10 N a + - b i n d i n g site in thrombin. Octahedral co-ordinat ion stabil izes the i o n . B a c k b o n e carbonyl groups and water molecules are i n v o l v e d i n the interaction. Three ion-pairs surround the site and provide stability. Introduction o f a sod ium b ind ing site into a mutant S G T to improve substrate specifici ty w o u l d invo lve signif icant mutagenesis o f the gene as nearly a l l o f the residues that 88 comprise the site are absent i n S G T . In thrombin , the N a + b ind ing site is located between two surface loops beginning at residues 180 and 220 [77]. T h e ion b i n d i n g site is 15-20 A distal f rom the catalytic triad, yet l ies wi th in 5 A f rom D 1 8 9 . A cy l i nd r i ca l cavi ty occupied by up to sixteen water molecules helps to stabil ize the site. B o u n d N a + is coordinated octahedrally by two carbonyl oxygen atoms prov ided by R 2 2 1 a a n d K 2 2 4 , and four bur ied water molecules . One o f these water molecules hydrogen bonds to the side cha in o f D 1 8 9 establishing a direct l i n k between the sod ium i o n and the S I site [52]. O v e r a l l stabili ty o f t h e N a + site is p rov ided by three i o n pairs, R 2 2 1 a - E 1 4 6 , K 2 2 4 - E 2 1 7 , and D 2 2 2 - R 1 8 7 . S o d i u m b ind ing involves residues adjacent to the substrate b ind ing pocket and may p lay a role i n generating stringent substrate specif ici ty. In particular, residues 192 and 217 were both targets for mutagenesis i n the present study and have been impl ica ted i n sod ium b ind ing . 4.3.8 Utility of a FXa-like Protease Factor X a is rout inely used for the cleavage o f recombinant fusion proteins, yet the preference for the I E G R cleavage sequence is not strict. F o r example , i n honey bee prepromeli t t in , a related sequence, V L G R , was readi ly c leaved by F X a [166]. A l t h o u g h this situation is not c o m m o n and can be prevented by inspect ion o f the polypeptide sequence o f the recombinant protein, a more specific protease is desirable. A number o f vendors supply alternative proteases w i t h differ ing recogni t ion sequences, i nc lud ing T E V protease [167], thrombin [127], enterokinase [127], and P r e S c i s s i o n ™ protease [127] (Table 4.5). Howeve r , many plasmids currently emp loyed contain the F X a cleavage m o t i f in-frame wi th c o m m o n restriction sites for c lon ing and recombinant expression o f proteins. T o assess the abi l i ty o f mutant S G T proteases to hydro lyze the bond after the I l e - G l u - G l y - A r g sequence i n 89 macromolecular substrates, pro thrombin was digested w i t h l im i t ed amounts o f each mutant constructed i n S G T (Figure 4.11). P ro th rombin was chosen as it is the substrate o f F X a in vivo and has a number o f potential cleavages sites. Protease Cleavage Sequence Notes Enterokinase Factor X a A s p - A s p - A s p - A s p - L y s I l e - G l u / A s p - G l y - A r g ^ P I ' can not be a P ro . Secondary cleavage sites (due to l o w P 3 and P 4 preference) Secondary cleavage sites. Cons ide rab ly more expensive than F X a Less useful for r emova l o f C- te rmina l tags R e m a i n i n g sequence after hydro lys is is problematic T h r o m b i n L e u - V a l - P r o - A r g ^ G l y - S e r T E V protease G l u - A s n - L e u - T y r - P h e -G l n ^ G l y L e u - G l u - V a l - L e u - P h e -G l n i G l y - P r o P reSc i s s ion protease Table 4.5 Cleavage sites o f proteases used i n processing recombinant proteins. Accura te processing o f prothrombin was not achieved by any o f the mutants constructed i n this study. Overn igh t digestions are typ ica l ly used for site specific proteolysis o f recombinant fusion proteins and hence a s imi lar strategy was employed . C l o s e inspect ion o f the digest ion pattern shown i n F igure 4.11 shows that on ly Y S F M P S G T yields a proteolytic fragment o f s imi la r size to the B - c h a i n o f thrombin. H o w e v e r , addi t ional cleavages occur and leave on ly trace amounts o f the desired fragment. T h e Y S F M P E mutant o f S G T d id not process prothrombin to y i e l d any fragment suggestive o f F X a - l i k e specif ici ty and confirms the data p rov ided through hydrolys is o f smal l peptide substrates. P ro th rombin act ivation is a poor representative o f what w o u l d be anticipated for the cleavage o f typ ica l recombinant proteins. In particular, several regions o f the protein are op t imized for hydro lys i s 90 in vivo. Therefore, i t is l i k e l y that the Y S F M P mutant o f S G T c o u l d be used as an alternative for the site specific proteolysis o f recombinant proteins as an alternative to F X a . B —Prothrombin —Prethrombin 1 —Thrombin B-chain —Fragment 1.2 —Pro tease —Fragment 1 j 1 ~"1 Prothrombin Fragment 1 Fragment 2 FXa Activation Fragment 1.2 A-chain B-chain Fragments F i g u r e 4.11 P ro th rombin processing by mutants o f S G T : T 1 9 0 P ( B ) , Y P ( C ) , Y S F P ( D ) , Y S F M P (E) and Y S F M P E (F) mutants o f S G T compared to F X a ( G ) and undigested pro thrombin ( A ) . O n l y the Y S F M P mutant o f S G T processed trace amounts o f a product the same size as thrombin. The Y S F M P E mutant o f S G T contains a l l k n o w n specif ici ty determinants o f F X a , but d i d not generate any o f the fragments associated w i t h the activation o f prothrombin. 91 4.4 Conclusions & Future Directions Extended substrate specif ici ty i n F X a is the result o f four amino acids at posit ions 99, 174 ,180 , and 192. These residues are posi t ioned accurately by residues 172 and 217, and a two amino ac id insert ion at pos i t ion 99. Other residues that surround these posit ions are l i k e l y i n v o l v e d i n mino r op t imiza t ion o f the electrostatic environment and stabili ty o f the region. Substi tut ion o f the key residues o f F X a into S G T created a protease w i t h s imi lar substrate specif ic i ty that was more s imi la r to F X I a rather than F X a . B a s e d on these f indings, the development o f a protease wi th more stringent specif ici ty for the preferred F X a cleavage sequence may be possible through addi t ional mutation o f S G T par t icular ly by addi t ion o f the N a + b ind ing site as found i n F X a or thrombin. Cent ra l to the success o f this research was the use o f B. subtilis for product ion o f the recombinant protein and mutants thereof. The ease o f protein pur i f icat ion and l o w cost o f product ion is a significant advantage over p rev ious ly reported systems. Structural and sequence s imi lar i ty o f S G T to F X a suggests that substrate specif ic i ty o f any coagulat ion protease cou ld be re-created i f not bettered. Increased stringency for the I l e - G l u - G l y - A r g F X a c o u l d be generated by mutagenesis o f the 99- loop and pos i t ion 192 i n the polypept ide sequence o f the Y S F M P E mutant o f b S G T . Character izat ion o f the specif ici ty o f F X a shows a weak preference at the P 2 , P3 and P 4 posit ions o f the substrate. Op t imiza t ion o f the 99- loop created i n S G T w i l l be needed to decrease f l ex ib i l i ty and restrict access to the S2 pocket. H o w e v e r , it is u n k n o w n what mutations w i l l create r ig id i ty i n the 99- loop. Mutagenesis o f residue 192 to a L y s amino acid should increase the specif ici ty o f the S3 pocket for negatively charged side chains. Increased stringency i n the S4 pocket may be created by strengthening the interaction o f residues 173 and 217. In particular, mutants bearing Ser, Thr , or L y s at pos i t ion 192 may stabil ize the loop 92 further. E a c h o f these mutations w i l l . l i k e l y cause a decrease i n the catalytic eff ic iency o f the enzyme. The suggested amino acids are l i k e l y not observed i n the wi ld- type F X a protein due to the phys io log ica l requirement o f eff iciency. E v o l u t i o n has decreased the potential harmful effect o f less than perfect substrate specif ici ty through l inkage o f addi t ional protein domains that facilitate protein-protein and pro te in- l ip id interactions. 93 Chapter 5. General Discussion and Outlook 5.1 Substrate Specificity Determinants of the SI family of Serine Proteases Increasing the substrate specif ici ty through mutagenesis o f the S I to S4 pockets i n S G T was successful. S i x mutations were combined , a two residue insert ion and five point mutations, to m i m i c the active site architecture o f coagulat ion factor X a . O n l y the combined mutant demonstrated a strong preference for the desired I l e - G l u - G l y - A r g recogni t ion sequence. Introduction o f a seventh mutat ion i n S G T , Y 2 1 7 E , d i d not further improve the specif ici ty towards F X a preferred substrates. A s anticipated, specif ici ty results largely f rom the amino ac id residues w h i c h constitute the enzyme-substrate interface but addi t ional regions o f the protease are important. A further increase i n specif ici ty w i l l require s tabi l izat ion o f the active site architecture and opt imiza t ion o f the electrostatic environment o f the entire active site. These mutations w i l l i n v o l v e second shel l or more distal residues throughout the protein. M o l e c u l a r evolu t ion over mi l l i ons o f years has accompl i shed these tasks and has generated diverse proteases and substrate specifici t ies. A number o f proteases share s imi la r architecture and substrate specif ici ty determinants as S G T . In the M E R O P S database, S G T is c lass i f ied as a member o f C l a n S A , F a m i l y S I , Subfami ly A peptidases [168]. N e a r l y 1000 proteases in this f ami ly have been identif ied, i nc lud ing the vertebrate and invertebrate coagulat ion factor proteases, ka l l ik re ins , granzymes, and, complement proteases (Figure 5.1). No tab ly , these proteases demonstrate t ryps in- l ike , chymot ryps in - l ike , and elastase-like pr imary substrate specifici t ies (Table 5.1). A l t h o u g h the overa l l architecture o f these proteases is s imi lar , variat ions i n the active site o f the enzyme facilitate differing specifici t ies. 94 r - S 1 A -S1-j— Insect Trypsins i~L SGT Invertebrate Trypsins r Plasmin I L Vertebrate Trypsins , _ | '— Vertebrate Chymotrypsins — Kallikreins I I — Coagulation Factor Proteases I— C1r C1s Proteases Elastases Granzymes & Cathepsin G Insect Chymotrypsins Complement Factor B — S1B—— Glutamyl endopeptidase \— S 1 C — P r o t e a s e Do S 1 D — L y s y l endopeptidase — S1E — Streptogrisin A I— S1F —Astrovirus serine protease F i g u r e 5.1 S i m p l i f i e d phylogenet ic tree o f the S I f ami ly o f peptidases. In each o f the S I sub-families o f peptidases, a diversi ty o f substrate specificit ies are found. Sub-fami ly A is considerably larger than the other sub-families w h i c h are l imi t ed i n dis tr ibut ion to gram posit ive/negative bacteria and viruses. 95 Preferred Peptidases PI Substrate Notes Ref. Granzyme A R,K Monomelic, dimeric and oligomeric structure [41,169] alters specificity Granzyme B D P4:1 > V, P3: E > G; Preference for D at PI, and [41,170,171] results from residue 226 Granzyme H W,Y,F [172] Granzyme M M,L [173] Human glandular R S1' preference for S [128] kallikrein 2 Trypsin R,K Slight S1 preference for R > K [41,174] Plasmin R,K Strong S1 preference for K > R [41] Cruzain Not P/I Very broad specificity [41] Duodenase L,K,Y,F [175] Trocarin R>K Many snake venom proteases have specificities similar to coagulation factor proteases [176,177] Table 5.1 Substrate specificit ies o f S I f ami ly peptidases. Chymot ryps in - l i ke proteases require proper pos i t ion ing o f large hydrophobic side chains i n the pr imary b ind ing pocket for efficient catalysis. Heds t rom demonstrated that convers ion o f a t rypsin- l ike enzyme to a chymot ryps in - l ike enzyme required extensive mutagenesis o f the S I pocket [56,57,59]. In addi t ion to point mutations at posit ions 189, 216, and 226 several loops adjacent to the pocket were required to generate the change i n the pr imary substrate specif ici ty but resulted i n a poor ly active enzyme. Importantly, the altered loops do not contact the substrate d i recdy. Improvement o f the catalytic eff ic iency against amide substrates was achieved by mutagenesis o f Y 1 7 2 W [57]. In the coagulat ion proteases, the side cha in o f residue 172 is exposed to the solvent and i n v o l v e d i n the S 4 substrate b i n d i n g pocket [178]. In the present study, mutation o f residue 172 i n S G T was required to m i m i c F X a - l i k e specif ici ty o f the S4 pocket. H o w e v e r , bur ia l o f the aromatic side cha in at pos i t ion led to an alternate conformat ion i n the 172-loop. These studies demonstrate that mul t ip le amino acids act i n concert and that s imi lar posi t ions i n the polypept ide sequence can affect differ ing specif ici ty pockets i n the fo lded protein. 96 A m i n o ac id insertions i n the protein sequence are a major component i n the substrate specif ici ty o f serine proteases. Hydrogen bonds, d ipole moments, steric constraints, and altered electrostatic environments can be generated by the introduct ion o f one or more residues into the protein. In the present work , addi t ion o f a two residue loop at pos i t ion 99 was important i n altering the properties o f the S 2 pocket. L o o p insertions are found i n many members o f the S I fami ly o f proteases. In addi t ion to pos i t ion 99, loop insertions are found at other posi t ions i n the protease sequence in the S I f ami ly peptidases. T h r o m b i n possesses a 10 amino ac id insert ion at posi t ion 60 as w e l l as a shorter loop at posi t ion 148 that are i n v o l v e d i n substrate recogni t ion [179,180]. These loops have l imi ted f l ex ib i l i t y that has on ly been demonstrated by mutagenesis o f the protease [181]. Other posit ions amenable for loop insertions include residue 170 found i n several coagulat ion proteases [182], the " k a l l i k r e i n l o o p " at posi t ion 90 [183], and also at residue 70 i n the neuropsins [184]. The loops can also be i n v o l v e d i n b iochemica l properties other than substrate specif ici ty such as enzyme regulat ion, protein-protein interactions and zymogen act ivat ion. Ex tended substrate specif ici ty w i th in the S I f ami ly o f peptidases can invo lve residues on the N - and C- te rmin i o f the scissile bond o f the substrate. Vertebrate coagulat ion proteases possess substrate specif ic i ty i n the S I to S 4 pockets. S m a l l side chains at P I ' are also favored by many proteases due to their decreased steric hindrance w i t h the enzyme. M e m b e r s o f the k a l l i k r e i n f ami ly d isp lay select ivi ty at P 2 ' for A r g side chains. Ra t and human mast c e l l proteases have a marked preference for ac idic residues at P 2 ' due to the presence o f L y s 4 0 , A r g l 4 3 , and L y s l 9 2 [185]. Hence , it seems possible that a s imi la r strategy employed to 97 generate coagulat ion factor X a specif ici ty i n S G T cou ld be appl ied to generate proteases h igh ly selective for amino acids on both sides o f the scissi le bond. 5.2 Molecular Evolution of the SI family of Serine Proteases O v e r the course o f evolut ion , proteases have been incorporated into a w ide variety o f ce l lu lar processes. A l t h o u g h the protease domain provides the catalytic machinery , addi t ional protein domains are l i nked to add functionali ty. Substrate recogni t ion and cel lu lar (or extracellular) loca l iza t ion are two c o m m o n roles o f these domains . F o r example , the C U B domains facilitate protein-protein interactions [186], G i a domains promote pro te in- l ip id interactions [187] and fibronectin domains loca l i ze proteins to extracellular f ib r in depositions [188,189]. Based on the diversi ty o f associated domains , construct ion o f an accurate phylogenet ic tree is a diff icul t task. A number o f approaches have been taken to dissect the evolu t ion o f the serine proteases. These studies can be d iv ided into those that examine the entire protease sequence inc lud ing the associated protein domains , and those w h i c h examine on ly the protease doma in or parts thereof [190-194]. A l l studies have supported the existence o f a s ingle ancestor for the entire S I f ami ly o f peptidases [120]. A p p r o x i m a t e l y 4 0 amino ac id posi t ions are h igh ly conserved throughout the f ami ly to produce a consistent three d imens iona l structure. Var ia t ions i n these residues are l imi ted to conservative mutations that preserve the polar i ty and size o f the side cha in [117]. T h e catalytic triad H i s - A s p - S e r is present i n a l l members o f the fami ly [195]. A carboxyl-group conta ining amino ac id at pos i t ion 194 is also absolutely conserved amongst the fami ly . A s p / G l u l 9 4 provides the electrostatic interaction wi th the N -terminus o f the protease domain (residue 16) [196]. Format ion o f this interaction stabilizes the 98 oxyan ion hole and the active site catalytic triad. L i m i t e d divers i ty at a part icular residue and close p rox imi ty to the active site suggests involvement i n enzyme-substrate interactions and specif ici ty. A n a l y s i s o f the amino ac id distr ibution at each residue i n the substrate b ind ing pockets shows a l imi ted number o f possibi l i t ies exist i n nature (Table 5.2). Determinants characterized i n the present w o r k (residues 180, 190, and 217) are moderately conserved, and can be mutated further to y i e l d nove l proteases. Other active site residues wi th less sequence conservat ion (residues 215 and 228) have not been characterized wi th respect to substrate specif ici ty but l i k e l y can be manipulated to influence specif ici ty. Together w i t h the var iabi l i ty i n size and compos i t ion o f the loops that surround the active site, the abi l i ty to design proteases w i t h specificit ies that do not exist i n nature seems possible. A t present, it has not been established h o w h igh ly specific proteases cou ld be developed i n the laboratory aside f rom structure based design. Elements o f the present dissertation c o u l d be used i n such a system. 5.3 Paper, Rock, Scissors Genetic Screening of Trypsin-like Proteases D e s i g n o f substrate specif ici ty by structure based techniques is l imi t ed by our inab i l i ty to understand the complex interactions that occur i n a protein structure. Muta t ions thought to exhib i t an effect often.result i n absent or opposing results to that expected [125,197]. In the present study, both T 1 9 0 P and Y 2 1 7 E mutations i n S G T yie lded s ignif icant ly different kinet ic properties f rom what were anticipated. Thus , alternate approaches are required. R a n d o m i z e d mutagenesis o f the who le gene or parts thereof combined w i t h genetic select ion has been w i d e l y adopted for the addi t ion or manipula t ion o f enzymat ic properties. 99 Substrate Binding Pocket S1 S2 S3/P4 16 17 189 190 194 221 228 214 180 192 215 216 217 I 72 V 60 D 69 S 36 D 99 A 50 Y 72 S 93 M 61 Q 46 W 63 G 90 E 14 A.A. % V 19 I 22 S 13 A 31 E 1 G 29 F 19 T 2 Q 7 K 11 F 19 V 5 Y 13 A.A. % L 1 A 2 G 9 T 20 N 5 N 2 A 1 H 3 N 10 Y 8 A 1 V 7 A.A. % A 1 L 2 N 2 Q 2 S 1 E 3 F 2 H 1 I 1 L 7 A.A. % V V D T D A Y S E Q W G Y S G T I V D A D A Y S M Q W G E F X a . V V D P D A Y s M Q W G E Y S F M P E b S G T T a b l e 5.2 Substrate specif ici ty determinants o f S I f ami ly sub-family A peptidases (not i nc lud ing loop regions). Sequence al ignment o f 740 proteases i n the S I f ami ly peptidases shows a l imi ted divers i ty o f amino acids exists at the substrate b i n d i n g region i n a l l k n o w n members o f the f ami ly (Sequence analysis described i n Chapter 4.2.3). T h e final mutant construct ( Y S F M P E ) d i d not y i e l d the desired specif ici ty, even though a l l k n o w n specif ici ty determinants were incorporated Conve r t i ng trypsin, or other p r imi t ive proteases, into h igh ly specific enzymes suffers f rom the disadvantage that the wi ld- type enzyme w i l l a lmost a lways be more active towards any substrate than the target enzyme. T y p i c a l directed evolu t ion strategies emp loy creating nove l activities or substrate specifici t ies that are comple te ly disparate f rom the wi ld - type enzyme [198,199]. F o r example , the wi ld- type enzyme does not catalyze a part icular react ion 100 or does so poor ly and the screen selects for increased act ivi ty. S u c h a screen can not be appl ied to increase proteolyt ic substrate specif ici ty f rom a p r imi t ive enzyme. A genetic screen for nove l substrate specif ici ty must also compensate for the large divers i ty o f possible substrates ar is ing f rom 20 amino acids at each posi t ion. A three residue stretch o f amino acids can have 8000 different possible permutations. M o r e o v e r , the side chains have a degree o f s imi la r i ty that can not easi ly be accounted for. Thus , ident if icat ion o f the desired mutant protein must re ly on strategies employed i n nature. In particular, protease inhibi tors may be useful for in f luenc ing the molecular evolu t ion o f proteases. Recent molecular evolu t ion o f H I V proteases serves as an excel lent mode l for directed evolu t ion to escape inh ib i t ion . F o r the virus to reproduce, a number o f proteins are produced as precursors that must be c leaved for act ivat ion [200,201]. O n the basis o f this requirement, a number o f protease inhibi tors have been designed and appl ied c l i n i c a l l y i n the treatment o f this disease [202]. Unfortunately, the virus is k n o w n to mutate rapid ly and a number o f mutations have been described that direct ly lead to resistance against inh ib i t ion [203]. These mutations are located at the active site cleft, as w e l l as at distal regions o f the protease. No tab ly , resistance to inh ib i t ion produced altered substrate specif ici ty o f the enzyme [204]. Muta t ions i n the protein substrates at the site o f cleavage have been demonstrated to accompany the mutations that generate inhib i tor resistance [205]. These observations suggest that evolu t ion directed by inh ib i t ion is a v a l i d concept for the design o f h igh ly specif ic proteases. Genet ic screening for nove l substrate specif ici ty c o u l d be generated by c o m b i n i n g three components: a protease, an inhibi tor that affects the wi ld- type but not the target protease, and a means by w h i c h to v isua l ize act ivi ty. Paper, R o c k , Scissors genetic screening 101 is put forth to accompl i sh these tasks (Figure 5.2). Co-express ion o f an inhib i tor w i t h the protease w i l l provide the direct ion for the evolut ion by d is t inguishing proteases that are s imi lar to the wi ld- type i n the active site. V a r i a b l e s ized loops at different posi t ions i n the polypept ide sequence o f S G T c o u l d be added and the who le gene randomly mutated b y various means [206,207]. O n l y active proteases wi th altered active site geometry or properties w o u l d overcome the inh ib i t ion and potential ly y i e l d mutants w i t h improved substrate specif ici ty. Detec t ion o f protease act ivi ty cou ld apply a sensitive fluorescent substrate that is added direct ly to the so l id growth med ium. Format ion o f a halo w o u l d show proteolyt ic act ivi ty and s ignal a potential ly important mutant that c o u l d be characterized further. S igni f icant technical hurdles are evident for this system to function. In particular, the generation o f large mutant l ibraries i n B. subtilis is s ignif icant ly more di f f icul t than E. coli and no k n o w n protease substrate w o u l d be amenable for in vivo selection. T o overcome these l imita t ions new techniques and nove l proteins must be developed. E. coli t rypsin inhibi tor , ecot in , is a potent inhib i tor o f t rypsin- l ike serine proteases w i t h several properties amenable for use i n directed evolut ion . T h e b i n d i n g mode o f ecot in is ident ical to that observed w i t h a natural substrate [208]. E c o t i n displays a number o f useful properties, i nc lud ing thermostabil i ty and stabili ty i n extreme p H . The inhib i tor exists as a d imer i n solut ion and has two regions i n v o l v e d i n inhibitor-protease interactions (Figure 5.3) [209]. M a r k e d differences i n inhib i tory strengths are observed us ing the wi ld- type protein and h igh specif ici ty enzymes (Table 5.3). Coagula t ion Factor X a and thrombin have a 100- to 1000-fold lower inh ib i t ion constant compared to broadly specific proteases [210]. A more potent inhib i tor against t rypsin- l ike enzymes can be constructed through mutagenesis o f the PI M e t to A r g [211]. In contrast, the inhib i tor can be converted to a monomer ic fo rm w i t h 102 1000-fold less inh ib i tory strength towards t rypsin- l ike enzymes [132]. Thus , ecot in can be genetical ly manipulated to exhib i t a broad range o f inhib i tory strengths spanning at least six orders o f magnitude. Di rec ted evolut ion to generate substrate specif ic i ty c o u l d use ecotin mutants iteratively, such that the inhibi tory strength was increased after each round o f selection. c Active Protease + + + -Altered Specificity + Inhibited + - + -Production of Halo + F i g u r e 5.2 Theory behind the Paper, R o c k s , Scissors genetic screen. A . Ini t ia l ly , the wi ld- type protease is b l o c k e d from hyd ro lyz ing a substrate that is inc luded i n the so l id growth m e d i u m and no halo is evident. B . R a n d o m mutagenesis o f the protease gene w i l l lead to mutant proteases that escape inh ib i t ion and a halo surrounding g r o w i n g bacterial colonies w i l l s ignal potential clones. C . Severa l poss ibi l i t ies are accounted for by this screen include the removal o f non-active mutants and d i sc r imina t ion f rom enzymes w i t h wi ld - type characteristics. 103 F i g u r e 5.3 D i m e r i c structure o f ecotin. The pr imary b i n d i n g site interacts wi th the target protease i n a conformat ion ident ical to that observed wi th substrates. A secondary protease b ind ing site is provided by the other cha in . P ro tease K j ( n M ) B o v i n e t rypsin <1 Factor X a 54 H u m a n leukocyte elastase 55 H u m a n F X I I a 89 H u m a n K a l l i k r e i n 163 T a b l e 5.3 Inhib i t ion constants o f ecotin against a variety o f S I f ami ly peptidases. T h r o m b i n , activated protein C , tissue-type p lasminogen activator and p la smin are poor ly inhibi ted by wi ld - type ecotin. 104 E c o t i n binds a protease in a s imi la r fashion as a substrate and this property c o u l d be used to further control the directed evolu t ion o f protease specif ici ty. Mutagenes is o f the three residues preceding the P I residue c o u l d be used to alter the Ki o f the inhibi tor . The mutated sequence o f the protein w o u l d be the opposite o f the recogni t ion site o f the protease. F o r example , monomer ic ecotin presenting G l n - L y s - T r p - M e t i n P 4 to P I should poor ly inhibi t coagulat ion factor X a w h i c h prefers the sequence U e - G l u - G l y - A r g . H o w e v e r , this mutant inhibi tor should s t i l l restrict the act ivi ty o f S G T . E c o t i n derives f rom a bacterial source and product ion o f the recombinant inhibi tor i n B. subtilis should be straightforward. G i v e n the h igh leve l o f recombinant protein expression i n E. coli, y ie lds i n B. subtilis should provide a sufficient molar excess o f inhibi tor . Detect ion o f a h igh ly specific protease i n h igh throughput screening requires a substrate that is w e l l defined, readi ly quantif ied, and preferably inexpensive. A c o m m o n l y used method for detection o f protease act ivi ty i n bacterial colonies is the addi t ion o f s k i m m i l k powder to the so l id m e d i u m . Protease act ivi ty leads to a zone o f clearance, or halo , around the co lony . Unfortunately, h igh specif ic i ty proteases do not hydro lyze s k i m m i l k effectively. Al te rna t ive ly , peptide substrates s imi la r to that employed i n the present w o r k can be synthesized w i t h fluorescent l eav ing groups. H o w e v e r , these peptides are cost ly to produce. Green fluorescent protein ( G F P ) is c o m m o n l y used as reporter protein for gene expression and cel lu lar loca l iza t ion and cou ld be engineered for detection o f protease act ivi ty [212-214]. T h e fluorophore is generated f rom three adjacent residues, S e r - T y r - G l y , that are sequestered on the inside an 11-stranded p-bar re l (Figure 5.4) [215]. T h e protein displays h igh thermostabil i ty and extreme resistance to proteolysis result ing f rom the t ight ly packed structure [216]. Recent ly , W i l l i a m s described three regions in G F P that cou ld be rendered 105 sensitive to site specific proteolysis [217]. However , cleavage o f the protein at any one o f these sites d i d not lead to a decrease i n fluorescence. These results were l i k e l y due to the stability o f the P -ba r re l structure w h i c h d id not unfold after hydro lys i s o f a s ingle bond. C o m b i n i n g two insertion mutations lead to a protein that d i d not fo ld proper ly , fo rming inc lus ion bodies in their E. coli expression system, and further research was not continued ( M . W i l l i a m s , personal communica t ion) . R a n d o m mutagenesis and screening o f a double insertion mutant o f G F P cou ld be readily performed to select for soluble protein that fluoresces under UV l ight [218]. G F P w o u l d then be a converted to a useful reagent for the detection o f proteolytic act ivi ty in vivo. A c t i v e protease w o u l d be evident by the format ion o f a halo surrounding a B. subtilis co lony i f the fluorescent protein was added di rect ly to the so l id growth m e d i u m in the proposed screen. F i g u r e 5.4 Structure o f G F P and potential regions for insertion o f a protease recogni t ion sequence ( P D B code 1 E M A ) . Fo r example , I l e - G l u - G l y - A r g - S e r inserted at posit ions 172 and 189 w o u l d a l l ow F X a to cleave G F P twice , re leasing a strand o f the P -ba r re l , and r emov ing the fluorescent properties o f the molecu le . 106 In summary, a directed evolu t ion strategy for the development o f nove l proteolyt ic substrate specificit ies is possible . T w o proteins must be created and characterized pr ior to va l ida t ing such an approach. T h e recombinant expression o f S G T i n B . subtilis provides a useful starting point towards this goa l and is a substantial improvement over prev ious ly reported systems [59,98]. Mutagenes is o f the cleavage sequence presented by G F P and the inhib i tor w i l l permit screening for any specif ic i ty desired. N o v e l substrate b i n d i n g pockets m a y result that differ f rom that observed elsewhere i n the fami ly , yet produce s imi lar and more stringent substrate specifici t ies 5.5 F u t u r e O p p o r t u n i t i e s A number o f nove l features exist in the t rypsin scaffold that are poor ly understood and c o u l d be generated on a s impl i f i ed scaffold such as S G T . T h e role o f sod ium b i n d i n g i n generating catalytic eff ic iency was prev ious ly discussed as a potential avenue for research. Induced fit mechanisms o f substrate specif ici ty have been described, par t icular ly i n the complement system [219]. A u t o l y t i c act ivat ion induced by receptor b i n d i n g is a w e l l k n o w n phenomenon i n the vertebrate b l o o d coagulat ion cascade [220]. Z y m o g e n activation mechanisms are k n o w n to differ i n the S I f ami ly peptidases [221-224]. L i n k a g e o f the protease domain to other protein domains cou ld be studied to generate nove l funct ion and proteolyt ic activit ies. Las t ly , l i t t le is understood i n the mechanism o f catalytic rate enhancement caused by phospho l ip id and co-factor b ind ing i n the coagulat ion cascade [225]. These molecula r properties are l i k e l y not dist inct mechanisms acting alone. Structural p r o x i m i t y and direct interactions w i th adjacent amino ac id side chains suggest complex relationships w i th substrate b ind ing and the catalytic process have yet to be fu l ly understood. 107 5.6 Significance of the Work In the present dissertation, a nove l expression system for t ryps in- l ike enzymes has been produced and op t imized . Prev ious studies have for the most part used eukaryotic proteases w h i c h are inherently more di f f icul t to w o r k w i t h based on their evolu t ion to meet phys io log ica l function. I have demonstrated that p r imi t ive t rypsin- l ike enzymes der ived f rom a bacterial source can be used as a scaffold for engineering substrate specif ic i ty and functional properties s imi la r to eukaryot ic proteases. A number o f mutations were created i n S G T that increased the pr imary specif ic i ty for A r g conta ining substrates and the extended substrate specif ici ty was improved to part ia l ly m i m i c F X a . These results show that the substrate specif ici ty o f any protease i n the S1 fami ly o f peptidases c o u l d be dupl icated us ing a s imi lar approach. Perfect m i m i c r y o f F X a substrate specif ic i ty was not achieved l i k e l y due to the requirement o f addi t ional second shell residues i nvo lved i n op t imiza t ion o f the b i n d i n g pocket . Future w o r k c o u l d focus on further reproduct ion o f F X a specif ici ty through s imi lar amino ac id substitutions. A n alternative approach that involves r andom mutagenesis and a nove l genetic screen has also been described that m a y y i e l d h igh ly specific proteases that bear litt le sequence s imi lar i ty w i t h k n o w n proteases. Di rec ted evolu t ion o f extended substrate specif ici ty is technical ly cha l leng ing and, i f successful, can s ignif icant ly expand the repertoire o f protease technology. H i g h l y selective serine proteases w o u l d be useful i n a number o f applicat ions. T h e Y S F M P mutant o f S G T bearing six mutations c o u l d be used for the site specific proteolysis o f recombinant proteins as an alternative to F X a . In the future, it is anticipated that proteases c o u l d be designed wi th levels o f substrate specif ic i ty approaching that found i n restriction 108 endonucleases [226]. These enzymes c o u l d be used for hydro lys is o f proteins wi thout the need for manipula t ion o f the D N A sequence for addi t ion o f a protease recogni t ion site. M o r e o v e r , peptide l iga t ion through reverse proteolysis has been described and cou ld be combined wi th h igh ly selective mutants o f S G T [227]. Thus , nove l proteins c o u l d be produced more rap id ly i n an approach comparable to combinator ia l chemistry. 5.7 C o n c l u s i o n s Substrate specif ici ty o f the S I f ami ly peptidases is der ived f rom a few amino acids in the protein sequence. D u e to the requirement o f second shell residues for opt imiza t ion o f the substrate b i n d i n g site, structure based design o f selectivi ty is a diff icul t yet achievable goa l . Eng inee r ing specif ic i ty is then l imi ted by what exists i n nature. M o l e c u l a r evolut ion has not explored many o f the possibi l i t ies o f substrate specif ici ty due to the phys io log ica l requirement for efficient catalysis. Di rec ted evolut ion , as proposed i n the present dissertation or by other means, is the next l i k e l y step i n the progression o f protease technology. 109 Appendix A: Structural Alignment of Selected SI Family Peptidases Structures o f S I f ami ly peptidases were a l igned us ing the combina tor ia l extension method avai lable on-l ine at http://cl.sdsc.edu/ce.html [228]. Conserved residues i n a l l 8 polypeptides are denoted wi th * and mutations constructed i n S G T are l isted be low the alignment. 60-loop 1SGT 1TLD 3 TGI 1PFX 1AUT 1HCG 1CVW 1PPB 1 WGGTRAAQGEFPFMVRLSM 1 IVGGYTCGANTVPYQVSLNSGYH 1 IVGGYTCQENSVPYQVSLNSGYH 1 IVGGENAKPGQFPWQVLLNGKIDA- -1 LIDGKMTRRGDS PWQVVLLDSKKKL -1 IVGGQECKDGECPWQALLINEENEG-1 IVGGKVCPKGECPWQVLLLVNGAQ- -GCGGALYAQDIVLTAAHCV FCGGSL INSQWWSAAHCY FCGGSLINDQWWSAAHCY FCGGSIINEKWWTAAHCIEP- -ACGAVLIHPSWVLTAAHCMDES-FCGGTILSEFYILTAAHCLYQA-LCGGTLINTIWWSAAHCFDKI -- --SGSGNNT--SITATGG K S- -GIQVRLG K S- -RIQVRLG G -V- -KITWAG K KLLVRLG K RFKVRVG K NWRNLIAVLG 1 IVEGSDAEIGMS PWQVMLFRKS PQELLCGASLISDRWVLTAAHCLLYP PWDKNF TENDLLVRIG 99-loop I I 1 SGT : 54 WDLQ- -S-G-AAVKVRSTKVLQAPGYN G-TGKDWALIKLAQPIN QPTLKIAT-T T 1TLD: 52 EDNINWE-G-NEQFISASKSIVHPSYN-SNT-LNNDIMLIKLKSAASLNSRVASISLPT-S C 3TGI: 52 EHNINVLE-G-NEQFVNAAKIIKHPNFD-RKT-LNNDIMLIKLSSPVKLNARVATVALPS-S C 1PFX: 55 EYNTEETEP- -TEQRRNVIRAIPHHSYNATVNKYSHDIALLELDEPLTLNSYVTPICIAD-KEYTNI - F 1AUT: 56 EYDLRRWE-K-WELDLDIKEVFVHPNYS-KST-TDNDIALLHLAQPATLSQTIVPICLPD-SGLAEREL 1HCG: 56 DRNTEQEE-G-GEAVHEVEWIKHNRFT-KET-YDFDIAVLRLKTPITFRMNVAPACLPE-RDWAESTL 1CVW : 58 EHDLSE-H-DGDEQSRRVAQVIIPSTYV- PGT -TNHDIALLRLHQPWLTDHWPLCLPERTFSERT-L 1PPB: 65 KHSRTRYE-RNIEKISMLEKIYIHPRYN-WRENLDRDIALMKLKKPVAFSDYIHPVCLPD-RETAAS-L KE Y * * 96 99 -172-loop-1SGT: 104 AYNQGTFTVAGWGA-1TLD: 110 ASAGTQCLISGWGN-3TGI: 110 APAGTQCLISGWGN-1PFX: 120 LK-FGSGYVSGWGR-1AUT: 120 NQAGQETLVTGWGY• 1HCG: 120 MT-QKTGIVSGFGR-1CVW: 122 AF-VRFSLVSGWGQ-1PPB: 130 LQAGYKGRVTGWGN-NRE - GG SQQRYLLKANVPFVSDAACRSAY GNELVANEEICAGY -TKSSGT SYPDVLKCLKAPILSDSSCKSAY PGQIT-SNMFCAGY-TLSSGV NEPDLLQCLDAPLLPQADCEAS Y PGKIT - DNMVCVGF -VFNRG RSATILQYLKVPLVDRATCLRST KFTIY-SNMFCAGF -HSSREKEAKRN-RTFVLNFIKIPWPHNECSEVM SNMVS - ENMLCAGI -THEKGRQS TRLKMLEVPYVDRNSCKLS S SFIIT - QNMFCAGY -LLDRG ATALELMVLNVPRLMTQDCLQQSRKVGDS PNIT - E YMFCAGY -LKETWTANVGKGQPSVLQWNLPIVERPVCKDST RIRIT-DNMFCAGY-* * S F M * ** 172 174 180 1SGT 160 - PDT---GGVDTC- -QGDSGGPMFRK DNADEWIQVGIVSWGYGCARPGYPGVYTEVSTFASAI 1TLD 166 -L- E- --GGKDSC--QGDSGGPWCS GKLQGIVSWGSGCAQKNKPGVYTKVCNYVSWI 3TGI 166 -L- E- --GGKDSC--QGDSGGPWCN GELQGIVSWGYGCALPDNPGVYTKVCNYVDWI 1PFX 174 -H-E- --GGKDSC--QGDSGGPHVTEVEGT-- SFLTGIISWGEECAVKGKYGIYTKVSRYVNWI 1AUT 180 -L- G---DRQDAC--EGDSGGPMVASFHGT-- WFLVGLVSWGEGCGLLHNYGVYTKVSRYLDWI 1HCG 174 -D-T- - -KQEDAC--QGDSGGPHVTRFKDT-- YFVTGIVSWGEGCARKGKYGIYTKVTAFLKWI 1CVW 181 -S- D---GSKDSC- -KGDSGGPHATHYRGT-- WYLTGIVSWGQGCATVGHFGVYTRVSQYIEWL 1PPB 191 -K-PDEGKRGDAC--EGDSGGPFVMKSPFNNR WYQMGIVSWGEGCDRDGKYGFYTHVFRLKKWI *p* * * * * * * ****E*** * ** * * 190 217 %ldentity to SGT Ref. 1SGT 217 ASAARTL Streptomyces griseus trypsin 100 .0 [76] 1TLD 217 KQTIASN Bovine beta-trypsin 35. 1 [112] 3 TGI 217 QDTIAAN Rat anionic trypsin 32 . 7 [229] 1PFX 229 KEKTK-- Human coagulation factor IXa 34 . 0 [230] 1AUT 235 HGHIRDK Human activated protein C 32. 2 [231] 1HCG 229 DRSMKTR Human coagulation factor Xa 32 . 5 [141] 1CVW 236 QKLMRSE Human coagulation factor V i l a 35. 9 [158] 1PPB 251 QKVIDQF Human alpha-thrombin 33 . 2 [157] 110 B i b l i o g r a p h y 1. Katakura , S., Nagahara , T . , Hara , T . & Iwamoto, M . (1993). A nove l factor X a inhibi tor: structure-activity relationships and selectivi ty between factor X a and thrombin. Biochem. Biophys. Res. Commun. 197, 965-72. 2 . Sandberg, H . , Almsted t , A . , Brandt , J . , G r a y , E . , Ho lmqu i s t , L . , Oswa ldsson , U . , Sebr ing, S. & M i k a e l s s o n , M . (2001). Structural and functional characteristics o f the B-domain-dele ted recombinant factor V I I I protein, r - V I I I S Q . Thromb. Haemost. 85, 93-100. 3 . Denaul t , J . B . & Salvesen, G . S. (2002). Caspases: keys i n the ign i t ion o f c e l l death. Chem. Rev. 102, 4489-500. 4. D a v i e , E . W . , Fu j ikawa , K . & K i s i e l , W . (1991). The coagulat ion cascade: ini t ia t ion, maintenance, and regulat ion. Biochemistry 30, 10363-70. 5. Natesh, R . , Schwager , S. L . , Sturrock, E . D . & A c h a r y a , K . R . (2003). Crys t a l structure o f the human angiotensin-convert ing enzyme- l i s inopr i l complex . Nature All, 551-4. 6. Neurath , H . (1984). E v o l u t i o n o f proteolytic enzymes. Science 224, 350-7. 7. K l e m m , U . , M u l l e r - E s t e r l , W . & E n g e l , W . (1991). A c r o s i n , the pecul iar sperm-specific serine protease. Hum. Genet. 87, 635-41 . 8. Sattar, R . , A l i , S. A . & A b b a s i , A . (2003). B io in fo rmat ics o f granzymes: sequence compar i son and structural studies on granzyme f a m i l y by h o m o l o g y mode l ing . Biochem. Biophys. Res. Commun. 308, 726-35. 9. K o s t o v a , Z . & W o l f , D . H . (2003). F o r w h o m the b e l l tol ls: protein qual i ty cont ro l o f the endoplasmic re t icu lum and the ubiquit in-proteasome connect ion. EMBO J. 22, 2309-17. 1 0 . Tay lo r , N . A . , V a n D e V e n , W . J . & Creemers , J . W . (2003). C u r b i n g activation: proprotein convertases i n homeostasis and pathology. FASEB J. 17, 1215-27. 11. S tamenkovic , I. (2003). Ext race l lu la r matr ix remodel l ing : the role o f matr ix metalloproteinases. J. Pathol. 200, 448-64. 12 . Se lkoe , D . J . (1998). T h e c e l l b i o l o g y o f be ta-amyloid precursor protein and preseni l in in A lzhe imer ' s disease. Trends Cell Biol. 8, 447-53. 13 . Yager , D . R . & N w o m e h , B . C . (1999). The proteolytic environment o f chronic wounds . Wound Repair Re gen. 7 , 4 3 3 - 4 1 . I l l 14. B o l t o n - M a g g s , P . H . & Pas i , K . J . (2003). Haemophi l i a s A and B . Lancet 361 , 1801-9. 15. G u z m a n , L . A . & L i n c o f f , A . M . (1997). Th rombo ly t i c Therapy: T h e Treatment o f C h o i c e i n A c u t e M y o c a r d i a l Infarction. J. Thromb. Thrombolysis 4 , 337-343. 16. M c K e r r o w , J . H . , Sun , E . , Rosenthal , P . J . & B o u v i e r , J . (1993). T h e proteases and pathogenici ty o f parasitic protozoa. Annu. Rev. Microbiol. Al, 821-53. 17. M a l i a r , T . , B a l a z , S., Tand l i ch , R . & Sturdik, E . (2002). V i r a l pro te inases-poss ib le targets o f ant iviral drugs. Acta Virol. 46 , 131-40. 18. Gupta , R . , B e g , Q . K . & L o r e n z , P . (2002). Bac te r i a l a lkal ine proteases: molecular approaches and industr ial applicat ions. Appl. Microbiol. Biotechnol. 59, 15-32. 19. R a o , M . B . , Tanksale , A . M . , Ghatge, M . S. & Deshpande, V . V . (1998). M o l e c u l a r and b io technologica l aspects o f m i c r o b i a l proteases. Microbiol. Mol. Biol. Rev. 62 , 597-635. 20. B r y a n , P . N . (2000). Pro te in engineering o f subt i l is in . Biochim. Biophys. Acta 1543, 203-222. 21 . Barrett, A . J . (1997). Nomencla ture Commi t tee o f the International U n i o n o f B iochemis t ry and M o l e c u l a r B i o l o g y ( N C - I U B M B ) . E n z y m e Nomencla ture . Recommendat ions 1992. Supplement 4: corrections and additions (1997). E u r . J . Biochem. 250, 1-6. 22. R a w l i n g s , N . D . & Barrett , A . J . (1999). M E R O P S : the peptidase database. Nucleic Acids Res. 21, 325-31. 23. Barrett, A . J . , T o l l e , D . P . & R a w l i n g s , N . D . (2003). M a n a g i n g peptidases i n the genomic era. Biol. Chem. 384, 873-82. 24. Heds t rom, L . (2002). Serine protease mechan i sm and specif ic i ty . Chem. Rev. 102, 4501-24. 25. Lowthe r , W . T . & Mat thews , B . W . (2002). Meta l loaminopept idases : c o m m o n functional themes i n disparate structural surroundings. Chem. Rev. 102 ,4581-608 . 26. D u n n , B . M . (2002). Structure and mechanism o f the pepsin- l ike f ami ly o f aspartic peptidases. Chem. Rev. 102, 4431-58. 27. L e c a i l l e , F . , Ka le ta , J . & B r o m m e , D . (2002). H u m a n and parasitic papain- l ike cysteine proteases: their role i n phys io logy and pathology and recent developments i n inhibi tor design. Chem. Rev. 102, 4459-88. 28. O r l o w s k i , M . & W i l k , S. (2000). Cata lyt ic activit ies o f the 20 S proteasome, a mult icatalyt ic proteinase complex . Arch. Biochem. Biophys. 383, 1-16. 112 29. Mat thews , B . W . , S ig ler , P . B . , Henderson, R . & B l o w , D . M . (1967). Three-d imens iona l structure o f tosyl-a lpha-chymotrypsin . Nature 214, 652-6. 30. D o d s o n , G . & W l o d a w e r , A . (1998). Cata ly t ic triads and their relatives. Trends Biochem. Sci. 23 , 347-52. 31 . War she l , A . , Papazyan, A . & K o l l m a n , P . A . (1995). O n low-barr ier hydrogen bonds and enzyme catalysis. Science 269, 102-6. 32. War she l , A . & Papazyan, A . (1996). Ene rgy considerations show that low-barr ier hydrogen bonds do not offer a catalytic advantage over ordinary hydrogen bonds. Proc. Natl. Acad. Sci. USA 93 , 13665-70. 33. Topf , M . , V a r n a i , P . & Richards , W . G . (2002). A b in i t io Q M / M M dynamics s imulat ion o f the tetrahedral intermediate o f serine proteases: insights into the active site hydrogen-bonding network. J. Am. Chem. Soc. 124, 14780-8. 34. Frey , P . A . , Whi t t , S. A . & T o b i n , J . B . (1994). A low-barr ier hydrogen bond i n the catalytic triad o f serine proteases. Science 264, 1927-30. 35. Har t ley , B . S. & K i l b e y , B . A . (1954). Biochem. J. 56, 289-97. 36. V e l d m a n , A . , Hof fman , M . & Ehrenfor th , S. (2003). N e w insights into the coagulat ion system and impl ica t ions for new therapeutic options wi th recombinant factor V i l a . Curr. Med. Chem. 10, 797-811. 37. Ruf , W . , Dorfleutner, A . & R i e w a l d , M . (2003). Spec i f ic i ty o f coagulat ion factor s ignal ing. J. Thromb. Haemost. 1, 1495-503. 38. A h m a d , S. S., L o n d o n , F . S. & W a l s h , P . N . (2003). The assembly o f the factor X -activating complex on activated human platelets. J. Thromb. Haemost. 1, 48-59. 39. Dah lback , B . & V i l l o u t r e i x , B . O . (2003). M o l e c u l a r recogni t ion i n the protein C anticoagulant pathway. J. Thromb. Haemost. 1, 1525-34. 40. D a v i e , E . W . (1995). B i o c h e m i c a l and molecula r aspects o f the coagulat ion cascade. Thromb. Haemost. 74, 1-6. 4 1 . Har r i s , J . L . , B a c k e s , B . J . , Leonet t i , F . , M a h r u s , S., E l l m a n , J . A . & C r a i k , C . S. (2000). R a p i d and general p ro f i l ing o f protease specif ici ty by us ing combinator ia l f luorogenic substrate l ibraries. Proc. Natl. Acad. Sci. USA 97 , 7754-9. 42 . B a c k e s , B . J . , Har r i s , J . L . , Leonet t i , F . , C r a i k , C . S. & E l l m a n , J . A . (2000). Synthesis o f posi t ional-scanning l ibraries o f f luorogenic peptide substrates to define the extended substrate specif ici ty o f p l a smin and thrombin. Nat. Biotechnol. 18, 187-93. 113 4 3 . Dav idson , C . J . , Tuddenham, E . G . & M c V e y , J . H . (2003). 450 m i l l i o n years o f hemostasis. J. Thromb. Haemost. 1, 1487-94. 44. J iang, Y . & Doo l i t t l e , R . F . (2003). The evolut ion o f vertebrate b l o o d coagulat ion as v i ewed f rom a compar i son o f puffer fish and sea squirt genomes. Proc. Natl. Acad. S c i . U S A 100, 7527-32. 45. Schechter, I. & Berger , A . (1967). O n the size o f the active site i n proteases. I. Papain . Biochem. Biophys. Res. Commun. 27, 157-62. 4 6 . O v e r a l l , C . M . (2002). M o l e c u l a r determinants o f metalloproteinase substrate specif ici ty: matrix metalloproteinase substrate b ind ing domains , modules , and exosites. Mol. Biotechnol. 22, 51-86. 4 7 . Mat t ler , L . E . & B a n g , N . U . (1977). Serine protease specif ic i ty for peptide chromogenic substrates. Thromb. Haemost. 38, 776-92. 48 . N a g a i , K . & Thogersen, H . C . (1984). Generat ion o f beta-globin by sequence-specific proteolysis o f a h y b r i d protein produced i n Esche r i ch i a c o l i . Nature 309, 810-2. 49 . R ichardson , P . L . (2002). The determination and use o f op t imized protease substrates in drug d iscovery and development. Curr. Pharm. Des. 8, 2559-81 . 5 0 . Turk , B . E . , H u a n g , L . L . , P i r o , E . T . & Cant ley , L . C . (2001). Dete rmina t ion o f protease cleavage site motifs us ing mixture-based oriented peptide l ibraries. N a t . Biotechnol. 19, 661-7. 5 1 . M a l y , D . J . , Leonet t i , F . , Backes , B . J . , Dauber , D . S., Har r i s , J . L . , C r a i k , C . S. & E l l m a n , J . A . (2002). Expedient solid-phase synthesis o f f luorogenic protease substrates us ing the 7-amino-4-carbamoylmethylcoumar in ( A C C ) f luorophore. J . O r g . Chem. 67, 910-5. 5 2 . Gruber , A . , C a n t w e l l , A . M . , D i Cera , E . & Hanson , S. R . (2002). T h e thrombin mutant W 2 1 5 A / E 2 1 7 A shows safe and potent anticoagulant and anti thrombotic effects in v i v o . J. Biol. Chem. 277, 27581-4. 53 . D a n g , Q . D . , G u i n t o , E . R . & d i Cera , E . (1997). Ra t iona l engineer ing o f act ivi ty and specif ic i ty i n a serine protease. Nat. Biotechnol. 15 ,146-9 . 54. Fersht, A . & Win te r , G . (1992). Protein engineering. Trends Biochem. Sci. 17, 292-5. 55 . Y o u , L . & A r n o l d , F . H . (1996). Direc ted evolut ion o f subt i l i s in E i n B a c i l l u s subti l is to enhance total act ivi ty i n aqueous d imethyl formamide . Protein Eng. 9, 77-83. 56 . K u r t h , T . , U l l m a n n , D . , Jakubke, H . D . & Hedst rom, L . (1997). C o n v e r t i n g trypsin to chymot ryps in : structural determinants o f S I ' specif ic i ty . Biochemistry 36, 10098-104. 114 57 . Heds t rom, L . , Farr-Jones, S., Kettner, C . A . & Rutter, W . J . (1994). Conve r t i ng trypsin to chymotryps in : ground-state b i n d i n g does not determine substrate specif ici ty. Biochemistry 33 , 8764-9. 58 . Heds t rom, L . , Perona , J . J . & Rutter, W . J . (1994). Conver t ing t rypsin to chymotryps in : residue 172 is a substrate specif ici ty determinant. Biochemistry 33 , 8757-63. 59 . Heds t rom, L . , S z i l a g y i , L . & Rutter, W . J . (1992). Conver t ing trypsin to chymotryps in : the role o f surface loops. Science 255, 1249-53. 60. Heds t rom, L . (1996). T r y p s i n : a case study i n the structural determinants o f enzyme specif ici ty. Biol. Chem. 3 7 7 , 4 6 5 - 7 0 . 61 . Reza ie , A . R . & O l s o n , S. T . (1997). Cont r ibu t ion o f lys ine 6 0 f to S I ' specif ic i ty o f thrombin. Biochemistry 36, 1026-33. 6 2 . Grutter, M . G . , Priest le, J . P . , Rahue l , J . , Grossenbacher, H . , B o d e , W . , Hofsteenge, J . & Stone, S. R . (1990). Crys t a l structure o f the thrombin-hi rudin complex : a nove l mode o f serine protease inh ib i t ion . EMBO J. 9, 2361-5. 63 . W a n g , S. X . , E s m o n , C . T . & Flet ter ick, R . J . (2001). Crys ta l structure o f th rombin-ecotin reveals conformational changes and extended interactions. Biochemistry 40 , 10038-46. 6 4 . V i n d i g n i , A . , W i n f i e l d , M . , A y a l a , Y . M . & D i Cera , E . (2000). R o l e o f residue Y 9 9 in tissue p lasminogen activator. Protein Sci. 9, 619-22. 65 . H u n g , S. H . & Heds t rom, L . (1998). Conve r t i ng trypsin to elastase: substitution o f the SI site and adjacent loops reconstitutes esterase specif ici ty but not amidase act ivi ty . Protein Eng. 11 ,669-73 . 66 . E v n i n , L . B . , Vasquez , J . R . & C r a i k , C . S. (1990). Substrate specif ici ty o f t rypsin investigated by us ing a genetic selection. Proc. Natl. Acad. Sci. USA 87, 6659-63. 67 . Perona, J . J . , E v n i n , L . B . & C r a i k , C . S. (1993). A genetic selection elucidates structural determinants o f arginine versus lys ine specif ic i ty i n t rypsin. Gene 137, 121-6. 6 8 . K u r t h , T . , Grahn , S., Thormann , M . , U l l m a n n , D . , Hofmann , H . J . , Jakubke, H . D . & Heds t rom, L . (1998). Eng inee r ing the S I ' subsite o f t rypsin: design o f a protease w h i c h cleaves between dibasic residues. Biochemistry 3 7 , 1 1 4 3 4 - 4 0 . 69 . W a n g , J . , C h a o , J . & C h a o , L . (1992). Spec i f ic i ty determinants o f rat tissue k a l l i k r e i n probed by site-directed mutagenesis. Protein Eng. 5, 569-75. 115 7 0 . Hopfner , K . P . , Brandstetter, H . , Karcher , A . , K o p e t z k i , E . , Huber , R . , E n g h , R . A . & B o d e , W . (1997). Conve r t i ng b l o o d coagulat ion factor TXa into factor Xa : dramatic increase i n amidoly t ic act ivi ty identifies important active site determinants. EMBO J. 16, 6626-35. 7 1 . T rop , M . & B i r k , Y . (1968). The t rypsin- l ike enzyme f rom Streptomyces griseus (pronase). Biochem. J. 109, 475-6. 7 2 . O la f son , R . W . & S m i l l i e , L . B . (1975). E n z y m i c and phys icochemica l properties o f Streptomyces griseus t rypsin. Biochemistry 14, 1161-7. 7 3 . Ola f son , R . W . , Jurasek, L . , Carpenter, M . R . & S m i l l i e , L . B . (1975). A m i n o ac id sequence o f Streptomyces griseus t rypsin. C y a n o g e n bromide fragments and complete sequence. Biochemistry 14, 1168-77. 7 4 . K i m , J . C , C h a , S. H . , Jeong, S. T . , O h , S. K . & B y u n , S. M . (1991). M o l e c u l a r c l o n i n g and nucleotide sequence o f Streptomyces griseus t rypsin gene. Biochem. Biophys. Res. Commun. 181, 707-13. 7 5 . R e a d , R . J . , Brayer , G . D . , Jurasek, L . & James, M . N . (1984). C r i t i c a l evaluat ion o f comparat ive mode l b u i l d i n g o f Streptomyces griseus t rypsin. Biochemistry 23 , 6570-5. 7 6 . Read , R . J . & James, M . N . (1988). Ref ined crystal structure o f Streptomyces griseus trypsin at 1.7 A resolut ion. J. Mol. Biol. 200, 523-51. 7 7 . D i Cera , E . , G u i n t o , E . R . , V i n d i g n i , A . , D a n g , Q . D . , A y a l a , Y . M . , W u y i , M . & T u l i n s k y , A . (1995). T h e N a + b i n d i n g site o f thrombin. / . Biol. Chem. 270, 22089-92. 78. H a n n i g , G . & M a k r i d e s , S. C . (1998). Strategies for op t im iz ing heterologous protein expression i n Esche r i ch i a c o l i . Trends Biotechnol. 16, 54-60. 7 9 . P r i n z , W . A . , A s l u n d , F . , H o l m g r e n , A . & B e c k w i t h , J . (1997). The role o f the th ioredoxin and glutaredoxin pathways i n reducing protein disulf ide bonds i n the Esche r i ch i a c o l i cy toplasm. J. Biol. Chem. 272, 15661-7. 80 . U d a k a , S., Tsukagosh i , N . & Yamagata , H . (1989). B a c i l l u s brevis , a host bacter ium for efficient extracellular product ion o f useful proteins. Biotechnol. Genet. Eng. Rev. 7, 113-46. 81 . W o n g , S. L . (1995). Advances i n the use o f B a c i l l u s subtilis for the expression and secretion o f heterologous proteins. Curr. Opin. Biotechnol. 6, 517-22. 82 . Ve r r i p s , C . T . & van den B e r g , D . J . (1996). Barr iers to appl ica t ion o f genetical ly modi f i ed lactic ac id bacteria. Antonie Leeuwenhoek 70, 299-316. 8 3 . B i n n i e , C , Cossar , J . D . & Stewart, D . I. (1997). Heterologous biopharmaceut ical protein expression i n Streptomyces. Trends Biotechnol. 15, 315-20. 116 84. Schweizer , H . P . , H o a n g , T . T . , Propst, K . L . , Ornelas , H . R . & Karkhof f -Schweize r , R. R . (2001). V e c t o r design and development o f host systems for Pseudomonas. Genet. Eng. (TV. Y). 23 , 69-81 . 85 . B i n g l e , W . H . , N o m e l l i n i , J . F . & Smi t , J . (2000). Secret ion o f the Caulobacter crescentus S-layer protein: further loca l iza t ion o f the C- te rmina l secretion s ignal and its use for secretion o f recombinant proteins. J. Bacteriol. 182, 3298-301. 86 . Perona, J . J . , Heds t rom, L . , Rutter, W . J . & Flet ter ick, R . J . (1995). Structural or igins of substrate d iscr iminat ion i n t rypsin and chymot ryps in . Biochemistry 34, 1489-99. 87. M a k r i d e s , S. C . (1996). Strategies for ach iev ing h igh- leve l expression o f genes i n Esche r i ch ia c o l i . Microbiol. Rev. 60, 512-38. 88 . K a n e , J . F . (1995). Effects o f rare codon clusters on h igh- leve l expression o f heterologous proteins i n Esche r i ch ia c o l i . Curr. Opin. Biotechnol. 6, 494-500. 8 9 . Jurado, P . , R i t z , D . , B e c k w i t h , J . , de L o r e n z o , V . & Fernandez, L . A . (2002). P roduc t ion o f functional s ingle-chain F v antibodies i n the cy top lasm o f Esche r i ch i a c o l i . J. Mol. Biol. 320, 1-10. 90. Sambrook, J . , Fr i t sch , E . F . & Man ia t i s , T . (1989). Molecular Cloning: A Laboratory Manual, 2nd Ed., C o l d Spr ing Harbor Laboratory , C o l d Spr ing Harbor , N.Y. 9 1 . H a i l i n g , S. M . , Sanchez -Anza ldo , F . J . , Fukuda , R . , D o i , R . H . & Meares , C . F . (1977). Z i n c is associated wi th the beta subunit o f D N A - d e p e n d e n t R N A polymerase of B a c i l l u s subtil is . Biochemistry 16, 2880-4. 9 2 . W u , S. C . & W o n g , S. L . (1999). Deve lopment o f i m p r o v e d p U B l 10-based vectors for expression and secretion studies i n B a c i l l u s subti l is . J. Biotechnol. 72, 185-95. 93 . S p i z i z e n , J . (1958). Transformat ion o f b iochemica l ly deficient strains o f B a c i l l u s subtilis by deoxyribonucleate . Proc. Natl. Acad. Sci. USA 4 4 , 1 0 7 2 - 1 0 7 8 . 94. Hatanaka, Y . , Tsunematsu, H . , M i z u s a k i , K . & M a k i s u m i , S. (1985). Interactions o f derivatives o f guanidinophenylalanine and guanid inophenylg lyc ine wi th Streptomyces griseus t rypsin. Biochim. Biophys. Acta 832, 274-9. 9 5 . M c R e e , D . E . (1999). X t a l V i e w / X f i t — A versatile program for manipula t ing atomic coordinates and electron density. J. Struct. Biol. 125, 156-65. 9 6 . Brunger , A . T . , A d a m s , P . D . , C l o r e , G . M . , D e L a n o , W . L . , G r o s , P . , Grosse-Kuns t leve , R . W . , J iang , J . S., K u s z e w s k i , J . , N i l g e s , M . , Pannu , N . S., Read , R . J . , R i c e , L . M . , S imonson , T . & War ren , G . L . (1998). Crys ta l lography & N M R system: A new software suite for macromolecular structure determination. Acta Crystallogr. D. Biol. Crystallogr. 54 ( P t 5) , 905-21. 117 97. O t w i n o w s k i , Z . & M i n o r , D . (1997). Processing of X-ray Diffraction Data Collected in Oscillation Mode. Me thods E n z y m o l . , M a c r o m o l e c u l a r Crys ta l lography, Part A (Carter, C . W . & Sweet, R . M . , Eds . ) , 276, A c a d e m i c Press. 98. Vasquez , J . R . , E v n i n , L . B . , H i g a k i , J . N . & C r a i k , C . S. (1989). A n expression system for trypsin. J. Cell. Biochem. 39, 265-76. 99. F u k u o k a , S. & Nyaruhucha , C . M . (2002). Express ion and funct ional analysis o f rat P 2 3 , a gut hormone- inducible i soform o f t rypsin, reveals its resistance to proteinaceous trypsin inhibi tors . Biochim. Biophys. Acta. 1588, 106-12. 100. M a n o s r o i , J . , Tayapiwatana, C , G o t z , F . , Werner , R . G . & M a n o s r o i , A . (2001). Secret ion o f active recombinant human tissue p lasminogen activator derivatives i n Esche r i ch i a c o l i . Appl. Environ. Microbiol. 67, 2657-64. 101. Y e , R . , K i m , J . H . , K i m , B . G . , Szarka , S., Sihota , E . & W o n g , S. L . (1999). H i g h -leve l secretory product ion o f intact, b i o l o g i c a l l y active staphylokinase f rom B a c i l l u s subtil is . Biotechnol. Bioeng. 62, 87-96. 102. T ja l sma, H . , B o l h u i s , A . , Jongbloed , J . D . , B r o n , S. & van D i j l , J . M . (2000). S igna l peptide-dependent protein transport i n B a c i l l u s subtil is: a genome-based survey o f the secretome. Microbiol. Mol. Biol. Rev. 64, 515-47. 103. An t e lmann , H . , T ja l sma, H . , V o i g t , B . , Ohlmeie r , S., B r o n , S., van D i j l , J . M . & Hecker , M . (2001). A proteomic v i e w on genome-based s ignal peptide predict ions. Genome Res. 11, 1484-502. 104. Kuns t , F . , Ogasawara , N . , M o s z e r , I., A l b e r t i n i , A . M . , A l l o n i , G . , A z e v e d o , V . , Ber tero, M . G . , Bessieres, P . , B o l o t i n , A . , Borcher t , S., Bor r i s s , R . , Bours ie r , L . , Brans , A . , B r a u n , M . , B r i g n e l l , S. C , B r o n , S., B rou i l l e t , S., B r u s c h i , C . V . , C a l d w e l l , B . , Capuano , V . , Carter, N . M . , C h o i , S. K , C o d a n i , J . J . , Conner ton , I. F . , D a n c h i n , A. & et a l . (1997). T h e complete genome sequence o f the gram-posi t ive bacter ium B a c i l l u s subtil is . Nature 390, 249-56. 105. Nagarajan, V . , Ramaley , R . , A lbe r t son , H . & C h e n , M . (1993). Secret ion o f streptavidin f rom B a c i l l u s subtil is . Appl. Environ. Microbiol. 59 , 3894-8. 106. K a k u d o , S., Y o s h i k a w a , K , T a m a k i , M . , Nakamura , E . & Teraoka , H . (1992). Secretory expression o f a glutamic-acid-specif ic endopeptidase (SPase) f rom Staphylococcus aureus A T C C 1 2 6 0 0 i n B a c i l l u s subtil is . Appl. Microbiol. Biotechnol. 38, 226-33. 107. Wi t t e , V . , W o l f , N . , Diefenthal , T . , R e i p e n , G . & Dargatz , H . (1994). Heterologous expression o f the c los t r ipain gene f rom C l o s t r i d i u m h i s to ly t i cum i n Esche r i ch i a c o l i and B a c i l l u s subtil is: maturation o f the c los t r ipain precursor is coup led wi th self-act ivat ion. Microbiology 140 ( P t 5), 1175-82. 118 108 . van Wely, K . H . , S w a v i n g , J . , F reud l , R . & Driessen , A . J . (2001). Trans loca t ion o f proteins across the c e l l envelope o f Gram-pos i t ive bacteria. FEMS Microbiol. Rev. 25 , 437-54. 1 0 9 . Da lbey , R . E . , L i v e l y , M . O . , B r o n , S. & van D i j l , J . M . (1997). The chemistry and e n z y m o l o g y o f the type I s ignal peptidases. Protein Sci. 6, 1129-38. 110 . V a g u i n e , A . A . , R i c h e l l e , J . & W o d a k , S. J . (1999). S F C H E C K : a uni f ied set o f procedures for evaluat ing the qual i ty o f macromolecular structure-factor data and their agreement wi th the atomic m o d e l . Acta Crystallogr. D. Biol. Crystallogr. 55 ( P t 1), 191-205. 111 . M e y e r , E . , C o l e , G . , Radhakr i shnan , R . & E p p , O . (1988). Structure o f native porc ine pancreatic elastase at 1.65 A resolutions. Acta Crystallogr. B. 44 ( P t 1), 26-38. 1 1 2 . Bar tun ik , H . D . , Summers , L . J . & Bar t sch , H . H . (1989). Crys t a l structure o f bovine beta-trypsin at 1.5 A resolut ion i n a crystal f o r m wi th l o w molecula r p a c k i n g density. A c t i v e site geometry, i on pairs and solvent structure. J. Mol. Biol. 210, 813-28. 1 1 3 . B e r g l u n d , G . I., Smalas, A . O . & H o r d v i k , A . (1995). Structure o f A n i o n i c S a l m o n T r y p i n i n a 2nd Crys ta l F o r m . Acta Crystallogr. D. Biol. Crystallogr. 51 , 725-730. 114. Graf , L . , Jancso, A . , S z i l a g y i , L . , H e g y i , G . , Pinter , K . , Naray -Szabo , G . , H e p p , J . , M e d z i h r a d s z k y , K . & Rutter, W . J . (1988). Electrostat ic complementar i ty w i t h i n the substrate-binding pocket o f t rypsin. Proc. Natl. Acad. Sci. USA 8 5 , 4 9 6 1 - 5 . 1 1 5 . Edwards , P . D . , Mauge r , R . C , Cot t re l l , K . M . , M o r r i s , F . X . , P ine , K . K . , Sylvester , M. A . , Scott, C . W . & F u r l o n g , S. T . (2000). Synthesis and enzymat ic evaluat ion o f a PI arginine aminocoumar in substrate l ibrary for t ryps in- l ike serine proteases. Bioorg. Med. Chem. Lett. 10, 2291-4. 1 1 6 . F u r l o n g , S. T. , M a u g e r , R . C , Str impler , A . M . , L i u , Y . P . , M o r r i s , F . X . & Edwards , P . D . (2002). Synthesis and phys ica l characterization o f a P I arginine combinator ia l l ibrary , and its appl icat ion to the determination o f the substrate specif ici ty o f serine peptidases. Bioorg. Med. Chem. 10, 3637-47. 1 1 7 . R y p n i e w s k i , W . R . , Perrakis , A . , V o r g i a s , C . E . & W i l s o n , K . S. (1994). Evo lu t iona ry divergence and conservat ion o f t rypsin. Protein Eng. 7, 57-64. 118. Perona, J . J . & C r a i k , C . S. (1997). Evo lu t iona ry divergence o f substrate specif ici ty w i th in the chymot ryps in - l ike serine protease fo ld . J. Biol. Chem. 272 , 29987-90. 1 1 9 . S ichler , K . , Hopfner , K . P . , K o p e t z k i , E . , Huber , R . , B o d e , W . & Brandstetter, H . (2002). T h e influence o f residue 190 i n the S I site o f t rypsin- l ike serine proteases on substrate selectivity is universa l ly conserved. FEBS Lett. 530, 220-4. 119 120. Rose , T . & D i Cera , E . (2002). Substrate recogni t ion drives the evolu t ion o f serine proteases. J Biol Chem 277, 19243-6. 1 2 1 . Y a m a n e , T . , K o b u k e , M . , Tsutsui , H . , T o i d a , T . , S u z u k i , A . , A s h i d a , T. , Kawa ta , Y . & Sak iyama , F . (1991). Crys ta l structure o f Streptomyces erythraeus t rypsin at 2.7 A resolut ion. J. Biochem. (Tokyo). 110, 945-50. 122. B a n f i e l d , D . K . & M a c G i l l i v r a y , R . T . (1992). Par t ia l characterization o f vertebrate prothrombin c D N A s : ampl i f ica t ion and sequence analysis o f the B cha in o f thrombin f rom nine different species. Proc. Natl. Acad. Sci. USA 89, 2779-83. 1 2 3 . B a n f i e l d , D . K . , I rw in , D . M . , W a l z , D . A . & M a c G i l l i v r a y , R . T . (1994). E v o l u t i o n o f prothrombin: i sola t ion and characterization o f the c D N A s encoding ch icken and hagfish prothrombin . J. Mol. Evol. 38, 111-SI. 124. L i u , X . L . , Waze r , D . E . , Watanabe, K . & B a n d , V . (1996). Identif ication o f a nove l serine protease-like gene, the expression o f w h i c h is down-regulated dur ing breast cancer progression. Cancer Res. 56, 3371-9. 125. T o b i n , M . B . , Gustafsson, C . & H u i s m a n , G . W . (2000). Di rec ted evolut ion: the 'rational ' basis for ' irrational ' design. Curr. Opin. Struct. Biol. 10, 421-7. 126. Heds t rom, L . , L i n , T . Y . & Fast, W . (1996). H y d r o p h o b i c interactions control zymogen activation i n the t rypsin f ami ly o f serine proteases. Biochemistry 3 5 , 4 5 1 5 -23. 127. Jenny, R . J . , M a n n , K . G . & L u n d b l a d , R . L . (2003). A cr i t ica l r ev iew o f the methods for cleavage o f fusion proteins w i th thrombin and factor Xa. Protein Expr. Purif. 31 , 1-11. 128. Clout ie r , S. M . , Chagas , J . R . , M a c h , J . P . , G y g i , C . M . , Le i s inger , H . J . & Deperthes, D. (2002). Substrate specif ici ty o f human k a l l i k r e i n 2 ( h K 2 ) as determined by phage display technology. Eur. J. Biochem. 269, 2747-54. 129. Sheppeck, J . E . , 2nd, K a r , H . , G o s i n k , L . , Wheat ley , J . B . , Gjerstad, E . , Lof tus , S. M . , Z u b i r i a , A . R . & Jane, J . W . (2000). Synthesis o f a statistically exhaustive fluorescent peptide substrate l ibrary for p ro f i l i ng protease specif ic i ty . Bioorg. Med. Chem. Lett. 10, 2639-42. 130. P a c i n i , L . , V i t e l l i , A . , F i l o c a m o , G . , Ba r tho lomew, L . , Brunet t i , M . , Tramontano, A . , Steinkuhler , C . & M i g l i a c c i o , G . (2000). In v i v o selection o f protease cleavage sites by us ing ch imer ic S indbis vi rus l ibraries. J. Virol. 74, 10563-70. 1 3 1 . Rosse , G . , K u e n g , E . , Page, M . G . , Schauer -Vukas inov ic , V . , G i l l e r , T . , L a h m , H . W . , H u n z i k e r , P . & Schlatter, D . (2000). R a p i d identif icat ion o f substrates for nove l proteases us ing a combinator ia l peptide l ibrary. J. Comb. Chem. 2, 461-6. 120 132 . P a l , G . , S z i l a g y i , L . & Graf , L . (1996). Stable monomer ic fo rm o f an o r ig ina l ly d imer ic serine proteinase inhibi tor , ecot in, was constructed v i a site directed mutagenesis. FEBS Lett. 385, 165-70. 1 3 3 . He r ron , D . K . , G o o d s o n , T . , Jr., W i l e y , M . R . , W e i r , L . C . , K y l e , J . A . , Y e e , Y . K , Tebbe, A . L . , T ins l ey , J . M . , M e n d e l , D . , Masters , J . J . , F r anc i skov i ch , J . B . , Sawyer , J . S., Be igh t , D . W . , Ra tz , A . M . , M i l o t , G . , H a l l , S. E . , K l i m k o w s k i , V . J . , W i k e l , J . H . , Eas twood , B . J . , Towner , R . D . , G i f f o r d - M o o r e , D . S., Craft , T . J . & Smi th , G . F . (2000). 1,2-Dibenzamidobenzene inhibi tors o f human factor Xa. J. Med. Chem. 43 , 859-72. 134. A b a t o , P . , C o n r o y , J . L . & Seto, C . T . (1999). Combina to r i a l l ibrary o f serine and cysteine protease inhibi tors that interact w i t h both the S and S' b ind ing sites. J . M e d . Chem. 4 2 , 4 0 0 1 - 9 . 1 3 5 . K i c z a k , L . , K o s c i e l s k a , K , O t l e w s k i , J . , C z e r w i n s k i , M . & D a d l e z , M . (1999). Phage display selection o f P I mutants o f B P T I directed against f ive different serine proteinases. Biol. Chem. 380, 101-5; 136. M e l d a l , M. (1998). The solid-phase enzyme inhibi tor l ibrary assay. Methods Mol. Biol. 87, 75-82. 1 3 7 . Pannekoek, H . , van M e i j e r , M . , Schleef, R . R . , Loskutoff , D . J . & Barbas , C . F . , 3rd. (1993). Func t iona l d isplay o f human plasminogen-act ivator inhib i tor 1 ( P A I - 1 ) on phages: nove l perspectives for structure-function analysis b y error-prone D N A synthesis. Gene 128, 135-40. 1 3 8 . Ross , J . , J iang , H . , Kanost , M . R . & W a n g , Y . (2003). Serine proteases and their homologs i n the D r o s o p h i l a melanogaster genome: an in i t ia l analysis o f sequence conservat ion and phylogenet ic relationships. Gene 304, 117-31. 1 3 9 . Terado, T . , N o n a k a , M . I., N o n a k a , M . & K i m u r a , H . (2002). Conserva t ion o f the modular structure o f complement factor I through vertebrate evolu t ion . Dev. Comp. Immunol. 26, 403-13. 140 . Reza ie , A . R . (1996). R o l e o f residue 99 at the S2 subsite o f factor X a and activated protein C i n enzyme specif ici ty. J. Biol. Chem. 271 , 23807-14. 1 4 1 . Padmanabhan, K , Padmanabhan, K . P . , T u l i n s k y , A . , Park, C . H . , B o d e , W . , Huber , R . , B lankensh ip , D . T . , C a r d i n , A . D . & K i s i e l , W . (1993). Structure o f human d e s ( l -45) factor X a at 2.2 A resolut ion. J. Mol. Biol. 232, 947-66. 1 4 2 . R e y d a , S., Sohn , C , K l e b e , G . , R a i l , K . , U l l m a n n , D . , Jakubke, H . D . & Stubbs, M . T . (2003). Reconst ruct ing the b ind ing site o f factor X a i n t rypsin reveals l igand- induced structural plast ici ty. J. Mol. Biol. 325, 963-77. 121 1 4 3 . H s u , Y . C , H a m a g u c h i , N . , C h a n g , Y . J . & L i n , S. W . (2001). The dist inct roles that G i n - 1 9 2 and G l u - 2 1 7 o f factor I X p lay i n select ivi ty for macromolecu la r substrates and inhibi tors . Biochemistry 40 , 11261-9. 144. Gabor iaud , C , R o s s i , V . , B a l l y , I., A r l a u d , G . J . & Fon tec i l l a -Camps , J . C . (2000). Crys ta l structure o f the catalytic domain o f human complement e l s : a serine protease wi th a handle. EMBO J. 19, 1755-65. 145. Ca rva lho , A . L . , Sanz, L . , Baret t ino, D . , R o m e r o , A . , Calvete , J . J . & R o m a o , M . J . (2002). Crys t a l structure o f a prostate k a l l i k r e i n isolated f rom sta l l ion seminal p lasma: a homologue o f human PSA. J. Mol. Biol. 322, 325-37. 146. Reza ie , A . R . & E s m o n , C . T . (1996). M o l e c u l a r basis o f residue 192 part icipat ion i n determination o f coagulat ion protease specif ici ty. Eur. J. Biochem. 242, 477-84. 147. Reza ie , A . R . & E s m o n , C . T . (1995). Cont r ibu t ion o f residue 192 i n factor X a to enzyme specif ici ty and function. 7. Biol. Chem. 2 7 0 , 1 6 1 7 6 - 8 1 . 148. Reza ie , A . R . & E s m o n , C . T . (1993). C o n v e r s i o n o f glutamic ac id 192 to glutamine i n activated protein C changes the substrate specif ic i ty and increases react ivi ty toward macromolecular inhibi tors . J. Biol. Chem. 268, 19943-8. 149. N i k a i , T . , Ohara , A . , K o m o r i , Y . , F o x , J . W . & Sugihara, H . (1995). P r i m a r y structure of a coagulant enzyme, b i l i neob in , f rom A g k i s t r o d o n bil ineatus venom. Arch. Biochem. Biophys. 318, 89-96. 150. K o m o r i , Y . , N i k a i , T . , Ohara , A . , Y a g i h a s h i , S. & Sugihara, H . (1993). Effect o f b i l ineob in , a thrombin- l ike proteinase f rom the venom o f c o m m o n cant i l (Agk i s t rodon bil ineatus). Toxicon 31 , 257-70. 1 5 1 . Castro, H . C , S i l v a , D . M . , C r a i k , C . & Z i n g a l i , R . B . (2001). Structural features o f a snake v e n o m thrombin- l ike enzyme: thrombin and t rypsin on a single catalytic platform? Biochim. Biophys. Acta 1547, 183-95. 152. Reming ton , S. J . , W o o d b u r y , R . G . , Reyno lds , R . A . , Mat thews , B . W . & Neurath , H . (1988). T h e structure o f rat mast c e l l protease II at 1.9-A resolut ion. Biochemistry 27, 8097-105. 153. E s m o n , C . T . & Mather , T . (1998). S w i t c h i n g serine protease specif ici ty. Nat. Struct. Biol. 5, 933-7. 154. Ts i ang , M . , Ja in , A . K . , D u n n , K . E . , Rojas , M . E . , L e u n g , L . L . & G i b b s , C . S. (1995). Func t iona l mapp ing o f the surface residues o f human thrombin. J. Biol. Chem. 210, 16854-63. 122 1 5 5 . C h a n g , Y . J . , H a m a g u c h i , N . , C h a n g , S. C , Ruf , W . , Shen, M . C . & L i n , S. W . (1999). Engineered recombinant factor V I I Q 2 1 7 variants w i th altered inhib i tor specifici t ies. Biochemistry 38 ,10940 -8 . 156 . Zhang , E . & T u l i n s k y , A . (1997). The molecular environment o f the N a + b i n d i n g site of thrombin. Biophys. Chem. 63 , 185-200. 157 . B o d e , W . , M a y r , I., B a u m a n n , U. , Huber , R . , Stone, S. R . & Hofsteenge, J . (1989). The refined 1.9 A crystal structure o f human alpha-thrombin: interaction w i t h D - P h e -P r o - A r g chloromethylketone and signif icance o f the T y r - P r o - P r o - T r p insert ion segment. EMBO J. 8, 3467-75. 1 5 8 . K e m b a l l - C o o k , G . , Johnson, D . J . , Tuddenham, E . G . & Har los , K . (1999). Crys t a l structure o f active si te-inhibited human coagulat ion factor V i l a (des-Gla) . J. Struct. Biol. 127, 213-23. 1 5 9 . Fu j ikawa , K . , C h u n g , D . W . , Hendr i ckson , L . E . & D a v i e , E . W . (1986). A m i n o ac id sequence o f human factor XI, a b l o o d coagulat ion factor w i t h four tandem repeats that are h igh ly homologous wi th p lasma preka l l ik re in . Biochemistry 25 , 2417-24. 160 . W e l l s , J . A . (1990). A d d i t i v i t y o f mutat ional effects i n proteins. Biochemistry 29 , 8509-17. 1 6 1 . L i C a t a , V . J . & A c k e r s , G . K . (1995). Long-range , smal l magnitude nonaddi t iv i ty o f mutat ional effects i n proteins. Biochemistry 34, 3133-9. 1 6 2 . G u i n t o , E . R . , C a c c i a , S., Rose , T . , Futterer, K . , W a k s m a n , G . & D i Cera , E . (1999). Unexpec ted c ruc ia l role o f residue 225 i n serine proteases. Proc. Natl. Acad. Sci. USA 96, 1852-7. 1 6 3 . Schmidt , A . E . , Padmanabhan, K . , U n d e r w o o d , M . C , B o d e , W . , Mather , T . & Bajaj , S. P . (2002). The rmodynamic l inkage between the S I site, the N a + site, and the Ca2+ site i n the protease domain o f human activated protein C ( A P C ) . S o d i u m i o n i n the APC crystal structure is coordinated to four carbonyl groups f rom two separate loops. J. Biol. Chem. 277, 28987-95. 164. U n d e r w o o d , M . C , Z h o n g , D . , Mathur , A . , H e y d u k , T . & Bajaj , S. P . (2000). The rmodynamic l inkage between the S I site, the N a + site, and the Ca2+ site i n the protease domain o f human coagulat ion factor xa . Studies on catalytic eff ic iency and inhibi tor b ind ing . J. Biol. Chem. 275, 36876-84. 165 . P ineda , A .O . , Savvides , S . N . , W a k s m a n , G . , & d i Cera , E . (2003). Crys ta l structure o f the anticoagulant s low form o f thrombin. J. Biol. Chem. 277, 40177-80. 1 6 6 . H e , M . , J i n , L . & Aus ten , B . (1993). Spec i f ic i ty o f factor X a i n the cleavage o f fusion proteins. J. Protein Chem. 12, 1-5. 123 1 6 7 . Phan , J . , Zdanov , A . , E v d o k i m o v , A . G . , Tropea , J . E . , Peters, H . K . , 3rd, Kapust , R . B . , L i , M . , W l o d a w e r , A . & W a u g h , D . S. (2002). Structural basis for the substrate specif ic i ty o f tobacco etch virus protease. J. Biol. Chem. 277, 50564-72. 1 6 8 . R a w l i n g s , N . D . , T o l l e , D . P . & Barrett, A . J . (2004). M E R O P S : the peptidase database. Nucleic Acids Res. 32, D 1 6 0 - 4 . 1 6 9 . B e l l , J . K . , Goe tz , D . H . , M a h r u s , S., Har r i s , J . L . , Flet ter ick, R . J . & C r a i k , C . S. (2003). T h e o l igomer ic structure o f human granzyme A is a determinant o f its extended substrate specif ici ty. Nat. Struct. Biol. 10, 527-34. 1 7 0 . Odake , S., K a m , C . M . , Naras imhan , L . , Poe , M . , B l a k e , J . T . , Krahenbuh l , O . , Tschopp , J . & Powers , J . C . (1991). H u m a n and murine cytotoxic T lymphocyte serine proteases: subsite mapp ing w i t h peptide thioester substrates and inh ib i t ion o f enzyme act ivi ty and cy to lys is b y isocoumarins . Biochemistry 30, 2217-27. 1 7 1 . P h a m , C . T . , Thomas , D . A . , M e r c e r , J . D . & L e y , T . J . (1998). P roduc t ion o f fu l ly active recombinant murine granzyme B i n yeast. J. Biol. Chem. 273, 1629-33. 1 7 2 . Edwards , K . M . , K a m , C . M . , Powers , J . C . & Trapani , J . A . (1999). T h e human cytotoxic T ce l l granule serine protease granzyme H has chymot ryps in - l ike (chymase) act ivi ty and is taken up into cy toplasmic vesicles reminiscent o f granzyme B -conta ining endosomes. J. Biol. Chem. 214, 30468-73. 1 7 3 . Smyth , M . J . , Sayers, T . J . , Wi l t rou t , T . , Powers , J . C . & Trapani , J . A . (1993). M e t -ase: c l o n i n g and distinct ch romosomal locat ion o f a serine protease preferentially expressed i n human natural k i l l e r cel ls . J. Immunol. 151, 6195-205. 174. Page, M . J . , W o n g , S. L . , Hewi t t , J . , Strynadka, N . C . & M a c G i l l i v r a y , R . T . (2003). Eng inee r ing the pr imary substrate specif ici ty o f Streptomyces griseus trypsin. Biochemistry 42 , 9060-6. 175. Z a m o l o d c h i k o v a , T . S., Vorotyntseva , T . I. & A n t o n o v , V . K . (1995). Duodenase, a new serine protease o f unusual specif ici ty f rom bov ine duodenal mucosa . Pur i f i ca t ion and properties. Eur. J. Biochem. 227, 866-72. 1 7 6 . Joseph, J . S., C h u n g , M . C , Jeyaseelan, K . & K i n i , R . M . (1999). A m i n o ac id sequence o f t rocarin, a pro thrombin activator f rom Tropidech is carinatus venom: its structural s imi la r i ty to coagulat ion factor Xa. Blood 94, 621-31 . 177 . M a t s u i , T . , Fu j imura , Y . & T i t an i , K . (2000). Snake v e n o m proteases affecting hemostasis and thrombosis. Biochim. Biophys. Acta 1411, 146-56. 1 7 8 . Kamata , K . , K a w a m o t o , H . , H o n m a , T . , Iwama, T . & K i m , S. H . (1998). Structural basis for chemica l inh ib i t ion o f human b l o o d coagulat ion factor Xa. Proc. Natl. Acad. S c i . U S A 95 , 6630-5. 124 1 7 9 . D i B e l l a , E . E . & Scheraga, H . A . (1996). T h e role o f the insert ion loop around tryptophan 148 i n tthe act ivi ty o f thrombin. Biochemistry 35, 4427-33. 1 8 0 . T o s o , R . , P ino t t i , M . , H i g h , K . A . , P o l l a k , E . S. & Berna rd i , F . (2002). A frequent human coagulat ion Factor V I I mutation ( A 2 9 4 V , c l 5 2 ) i n loop 140s affects the interaction wi th activators, tissue factor and substrates. Biochem. J. 363, 411-6. 1 8 1 . van de Loch t , A . , B o d e , W . , Huber , R . , L e B o n n i e c , B . F . , Stone, S. R . , E s m o n , C . T . & Stubbs, M . T . (1997). T h e thrombin E 1 9 2 Q - B P T I complex reveals gross structural rearrangements: impl ica t ions for the interaction wi th ant i thrombin and th rombomodul in . EMBO J. 16, 2977-84. 1 8 2 . Soej ima, K . , Y u g u c h i , M . , M i z u g u c h i , J . , T o m o k i y o , K . , Nakash ima , T . , N a k a g a k i , T . & Iwanaga, S. (2002). T h e 99 and 170 loop-modi f i ed factor V i l a mutants show enhanced catalytic act ivi ty wi thout tissue factor. J. Biol. Chem. 277, 49027-35. 183 . T i m m , D . E . (1997). The crystal structure o f the mouse glandular ka l l ik re in -13 (prorenin conver t ing enzyme). Protein Sci. 6, 1418-25. 184. O k a , T. , H a k o s h i m a , T . , Itakura, M . , Y a m a m o r i , S., Takahashi , M . , Hash imoto , Y . , Shiosaka , S. & K a t o , K . (2002). R o l e o f loop structures o f neuropsin i n the act ivi ty o f serine protease and regulated secretion. J. Biol. Chem. 277, 14724-30. 1 8 5 . K a r l s o n , U . , Pejler, G . , Tomasin i -Johansson, B . & H e l l m a n , L . (2003). Ex tended substrate specif ici ty o f rat mast ce l l protease 5, a rodent alpha-chymase w i t h elastase-l i ke p r imary specif ic i ty . J. Biol. Chem. 278, 39625-31. 186. G a l , P . & A m b r u s , G . (2001). Structure and function o f complement act ivat ing enzyme complexes : C l and M B L - M A S P s . Curr. Protein Pept. Sci. 2, 43-59. 1 8 7 . H u a n g , M . , R i g b y , A . C , M o r e l l i , X . , Grant , M . A . , H u a n g , G . , Fur ie , B . , Seaton, B . & Fur ie , B . C . (2003). Structural basis o f membrane b i n d i n g by G i a domains o f v i t amin K-dependent proteins. Nat. Struct. Biol. 10, 751-6. 1 8 8 . Horrevoets , A . J . , Smi lde , A . , de V r i e s , C . & Pannekoek, H . (1994). T h e specific roles of finger and kr ing le 2 domains o f tissue-type p lasminogen activator dur ing i n vi tro f ibr inolys is . J. Biol. Chem. 269, 12639-44. 1 8 9 . Smi th , B . O . , D o w n i n g , A . K . , D r i s c o l l , P . C , Dudgeon , T . J . & C a m p b e l l , I. D . (1995). T h e solut ion structure and backbone dynamics o f the f ibronect in type I and epidermal growth factor-l ike pai r o f modules o f tissue-type p lasminogen activator. Structure 3, 823-33. 1 9 0 . K r e m , M . M . & D i Cera , E . (2001). M o l e c u l a r markers o f serine protease evolu t ion . EMBOJ. 20, 3036-45. 125 191. Patthy, L . (1990). Evo lu t iona ry assembly o f b l o o d coagulat ion proteins. Semin Thromb. Hemost. 16, 245-59. 192. Patthy, L . (1990). E v o l u t i o n o f b l o o d coagulat ion and f ibr inolys is . Blood Coagul. Fibrinolysis 1, 153-66. 193 . Patthy, L . (1985). E v o l u t i o n o f the proteases o f b l o o d coagulat ion and f ibr inolys is by assembly f rom modules . Cell. 4 1 , 657-63. 194. K r e m , M . M . , Rose , T . & D i Cera , E . (2000). Sequence determinants o f funct ion and evolu t ion i n serine proteases. Trends Cardiovasc. Med. 10, 171-6. 195 . Takeda-Shi taka , M . & U m e y a m a , H . (1998). E l u c i d a t i o n o f the cause for reduced act ivi ty o f abnormal human p lasmin conta in ing an A l a 5 5 - T h r mutat ion: importance o f h igh ly conserved A l a 5 5 i n serine proteases. FEBS Lett. 425, 448-52. 196. Heds t rom, J . , Sa in io , V . , Kemppa inen , E . , Haapia inen , R . , K i v i l a a k s o , E . , Schroder, T . , Le inonen , J . & Stenman, U . H . (1996). Se rum complex o f t rypsin 2 and alpha 1 anti trypsin as diagnostic and prognostic marker o f acute pancreatitis: c l i n i c a l study i n consecutive patients. B M J 313, 333-7. 197. N i x o n , A . E . & Firest ine, S. M . (2000). Ra t iona l and "irrat ional" design o f proteins and their use in biotechnology. IUBMB Life 49 , 181-7. 198. V o i g t , C . A . , Kauf fman , S. & W a n g , Z . G . (2000). Ra t iona l evolut ionary design: the theory o f i n vi tro protein evolut ion . Adv. Protein Chem. 55, 79-160. 199 . A r n o l d , F . H . & M o o r e , J . C . (1997). O p t i m i z i n g industr ial enzymes b y directed evolu t ion . Adv. Biochem. Eng. Biotechnol. 5 8 , 1 - 1 4 . 200. Kramer , R . A . , Schaber, M . D . , S k a l k a , A . M . , Gangu ly , K , Wong-S taa l , F . & Reddy , E . P . (1986). H T L V - I I I gag protein is processed i n yeast cel ls b y the virus p o l -protease. Science 231 , 1580-4. 2 0 1 . Lightfoote , M . M . , C o l i g a n , J . E . , F o l k s , T . M . , F a u c i , A . S., M a r t i n , M . A . & Venkatesan, S. (1986). Structural characterization o f reverse transcriptase and endonuclease polypeptides o f the acquired immunodef ic iency syndrome retrovirus. J . Virol. 60 , 771-5. 2 0 2 . B r i k , A . & W o n g , C . H . (2003). H I V - 1 protease: mechan i sm and drug discovery. O r g . Biomol. Chem. 1, 5-14. 2 0 3 . Condra , J . H . , Schleif , W . A . , B l a h y , O . M . , G a b r y e l s k i , L . J . , Graham, D . J . , Quintero , J . C , Rhodes , A . , R o b b i n s , H . L . , R o t h , E . , Shivaprakash, M . & et a l . (1995). In v i v o emergence o f HIV-1 variants resistant to mul t ip le protease inhibi tors . Nature 314, 569-71 . 126 204. W i l s o n , S. I., P h y l i p , L . H . , M i l l s , J . S., G u l n i k , S. V . , E r i c k s o n , J . W . , D u n n , B . M . & Kay , J . (1997). Escape mutants o f HIV-1 proteinase: enzymic eff ic iency and susceptibi l i ty to inh ib i t ion . Biochim. Biophys. Acta 1339 ,113-25 . 205 . R o b i n s o n , L . H . , M y e r s , R . E . , Snowden , B . W . , T isda le , M . & B l a i r , E . D . (2000). HIV type 1 protease cleavage site mutations and v i ra l fitness: impl ica t ions for drug susceptibi l i ty phenotyping assays. AIDS Res. Hum. Retroviruses 16, 1149-56. 206. C h i r u m a m i l l a , R . R . , Mura l idha r , R . , Marchant , R . & N i g a m , P . (2001). Improv ing the qual i ty o f industr ia l ly important enzymes by directed evolu t ion . Mol. Cell. Biochem. 224, 159-68. 207. D a l b y , P . A . (2003). O p t i m i s i n g enzyme funct ion b y directed evolu t ion . Curr. Opin. Struct. Biol. 13, 500-5. 208. M c G r a t h , M . E . , E r p e l , T . , Bystroff , C . & Het te r ick , R . J . (1994). M a c r o m o l e c u l a r chelat ion as an improved mechan i sm o f protease inh ib i t ion : structure o f the ecot in-t rypsin complex . EMBO J. 13, 1502-7. 209. M c G r a t h , M . E . , E r p e l , T . , B r o w n e r , M . F . & Het te r ick , R . J . (1991). Expres s ion o f the protease inhibi tor ecot in and its co-crys ta l l iza t ion w i t h t rypsin. J. Mol. Biol. 222 , 139-42. 210. Seymour , J . L . , L indqu i s t , R . N . , Denn i s , M . S., Mof fa t , B . , Yansura , D . , R e i l l y , D . , Wess inger , M . E . & Lazarus , R . A . (1994). E c o t i n is a potent anticoagulant and reversible tight-binding inhib i tor o f factor Xa. Biochemistry 33 , 3949-58. 211 . P a l , G . , Sprengel , G . , Patthy, A . & Graf , L . (1994). Al te ra t ion o f the specif ici ty o f ecot in, an E . c o l i serine proteinase inhibi tor , by site directed mutagenesis. FEBS Lett. 342, 57-60. 212. L e f f e l , S. M . , M a b o n , S. A . & Stewart, C . N . , Jr. (1997). App l i ca t ions o f green fluorescent protein i n plants. BioTechniques 23, 912-8. 213 . C o r m a c k , B . (1998). Green fluorescent protein as a reporter o f transcription and protein loca l iza t ion i n fungi . Curr. Opin. Microbiol. 1, 406-10. 214. P h i l l i p s , G . J . (2001). Green fluorescent protein—a bright idea for the study o f bacterial protein loca l iza t ion . FEMS Microbiol. Lett. 204, 9-18. 215 . O r m o , M . , Cubi t t , A . B . , K a l l i o , K , Gross , L . A . , Ts i en , R . Y . & R e m i n g t o n , S. J . (1996). Crys ta l structure o f the A e q u o r e a v ic tor ia green fluorescent protein. Science 273, 1392-5. 216. Ts i en , R . Y . (1998). The green fluorescent protein. Annu. Rev. Biochem. 67, 509-44. 127 2 1 7 . C h i a n g , C . F . , O k o u , D . T . , G r i f f i n , T . B . , Verret , C . R . & W i l l i a m s , M . N . (2001). Green fluorescent protein rendered susceptible to proteolysis: posi t ions for protease-sensitive insertions. Arch. Biochem. Biophys. 394, 229-35. 218. W a l d o , G . S. (2003). Genet ic screens and directed evolu t ion for protein so lubi l i ty . Curr. Opin. Chem. Biol. 7, 33-8. 2 1 9 . Tay lo r , F . R . , B i x l e r , S. A . , B u d m a n , J . I., W e n , D . , Karpusas , M . , R y a n , S. T . , J aworsk i , G . J . , Safar i-Fard, A . , Po l l a rd , S. & W h i t t y , A . (1999). Induced fit act ivat ion mechan i sm o f the except ional ly specific serine protease, complement factor D . Biochemistry 38, 2849-59. 220. Eigenbrot , C . (2002). Structure, funct ion, and act ivat ion o f coagulat ion factor V I I . Curr. Protein. Pept. Sci. 3, 287-99. 221 . Venka teswar lu , D . , Perera, L . , Darden , T . & Pedersen, L . G . (2002). Structure and dynamics o f zymogen human b l o o d coagulat ion factor X. Biophys. J. 82, 1190-206. 2 2 2 . Eigenbrot , C , Ki rchhofe r , D . , Denn i s , M . S., Santel l , L . , Lazarus , R . A . , Stamos, J . & U l t s c h , M . H . (2001). T h e factor V E zymogen structure reveals reregistration o f beta strands dur ing act ivat ion. Structure {Comb.) 9, 627-36. 2 2 3 . H ink-Schaue r , C , Estebanez-Perpina, E . , W i l h a r m , E . , Fuentes-Prior , P . , K l i n k e r t , W . , B o d e , W . & Jenne, D . E . (2002). The 2 . 2 - A crystal structure o f human pro-granzyme K reveals a r i g i d zymogen w i t h unusual features. J. Biol. Chem. 277, 50923-33. 224. Budayova-Spano , M . , L a c r o i x , M . , Thie lens , N . M . , A r l a u d , G . J . , Fon tec i l l a -Camps , J . C . & Gabor iaud , C . (2002). The crystal structure o f the zymogen catalytic domain o f complement protease C l r reveals that a disruptive mechan ica l stress is required to trigger act ivat ion o f the C l complex . EMBO J. 21 , 231-9. 2 2 5 . K r i s h n a s w a m y , S., Jones, K . C . & M a n n , K . G . (1988). Prothrombinase complex assembly. K i n e t i c mechan i sm o f enzyme assembly on phospho l ip id vesicles . J. Biol. Chem. 263, 3823-34. 2 2 6 . P i n g o u d , A . & Jeltsch, A . (1997). Recogn i t i on and cleavage o f D N A by type-II restrict ion endonucleases. Eur. J. Biochem. 246, 1-22. 227. Bordusa , F . (2002). Proteases i n organic synthesis. Chem. Rev. 102, 4817-68. 228 . Sh indya lov , I. N . & Bourne , P . E . (1998). Prote in structure al ignment by incremental combinator ia l extension ( C E ) o f the op t imal path. Protein Eng. 11, 739-47. 229 . Pasternak, A . , R i n g e , D . & Heds t rom, L . (1999). C o m p a r i s o n o f anionic and cat ionic trypsinogens: the anionic act ivat ion domain is more f lex ib le i n solut ion and differs i n its mode o f B P T I b i n d i n g i n the crystal structure. Protein Sci. 8, 253-8. 128 230. Brandstetter, H . , Bauer , M . , Huber , R . , L o l l a r , P . & B o d e , W . (1995). X - r a y structure of c lo t t ing factor IXa: active site and module structure related to X a s e act ivi ty and hemoph i l i a B . Proc. Nat.lAcad. Sci USA 92, 9796-800. 2 3 1 . Mather , T . , Oganessyan, V . , H o f , P . , Huber , R . , F o u n d l i n g , S., E s m o n , C . & B o d e , W . (1996). T h e 2.8 A crystal structure o f Gla-domain less activated protein C . EMBO J. 15, 6822-31. 129 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0091823/manifest

Comment

Related Items