HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Coombs, G. S. Articles by Madison, E. L. Search for Related Content PubMed PubMed Citation Articles by Coombs, G. S. Articles by Madison, E. L. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 34, 24074-24079, August 20, 1999

Revisiting Catalysis by Chymotrypsin Family Serine Proteases Using Peptide Substrates and Inhibitors with Unnatural Main Chains^*

Gary S. Coombs, Mohan S. Rao^§, Arthur J. Olson^§, Philip E. Dawson^¶**, and Edwin L. Madison^§§¶¶

From the Departments of ^¶ Cell Biology, ^§§ Vascular Biology, ^§ Molecular Biology, and  Chemistry, The Scripps Research Institute, the ^** Skaggs Institute for Chemical Biology, La Jolla, California 92037, the  Department of Molecular Biology, Corvas International, San Diego, California 92121, and ^¶¶ The Torrey Pines Institute for Molecular Studies, San Diego, California 92121

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Chymotrypsin family serine proteases play essential roles in key biological and pathological processes and are frequently targets of drug discovery efforts. This large enzyme family is also among the most advanced model systems for detailed studies of enzyme mechanism and structure/function relationships. Productive interactions between these enzymes and their substrates are widely believed to mimic the "canonical" interactions between serine proteases and "standard" inhibitors observed in numerous protease-inhibitor complexes. To test this central hypothesis we have synthesized and characterized a series of peptide analogs, based on model substrates and inhibitors of trypsin, that contain unnatural main chains. These results call into question a long accepted theory regarding the interaction of chymotrypsin family serine proteases with substrates and suggest that the canonical interactions observed between these enzymes and standard inhibitors may represent nonproductive rather than productive, substrate-like interactions.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

The chymotrypsin family of serine proteases contains many members, such as thrombin (1, 2), factor VII (3, 4), protein C (5), tissue-type plasminogen activator (6-8), urokinase plasminogen activator (9), snake (10-13), and easter (11-15), that play essential roles in important biological and pathological processes including embryonic development, the formation and dissolution of blood clots, angiogenesis, and tumor invasion and metastasis (1-15). This large enzyme family is also among the most extensively investigated and thoroughly characterized model systems for the detailed study of enzyme mechanism and structure/function relationships (16-20).

The three-dimensional structures of numerous complexes between a chymotrypsin family enzyme and a peptide or proteinaceous inhibitor have been solved, and these structures have provided many important insights into the mechanism and structure/function relationships of serine proteases (21, 22). The majority of these protein-protein complexes contain inhibitors known as standard inhibitors because they share a common, standard mechanism of action (22, 23). To date, more than 10 different families of standard mechanism protease inhibitors have been characterized (22, 23). Although they exhibit a similar mechanism of inhibition of serine proteases, neither the primary sequences nor the three-dimensional structures of these families exhibit homology to one another (21, 22).

Despite their unrelated three-dimensional structures, the details of the interaction of the reactive center loops of standard inhibitors from different gene families with the active site cleft of chymotrypsin family enzymes share remarkable similarities (21, 22). For example, the P3- P3' (Schecter and Berger nomenclature (24)) residues of the reactive center loop of standard inhibitors, both as free inhibitors and in complexes with target serine proteases, assume a virtually identical main chain conformation, which has been termed the "canonical" conformation (21, 22). The side chain of the P1 residue of the inhibitor fits into the specificity pocket of the enzyme, and the carbonyl oxygen extends toward the oxyanion hole formed by the main chain amides of Gly-193 and Ser-195. Another invariant feature of these complexes is that the P1-P3 residues of the standard inhibitor form a short, anti-parallel -sheet interaction with residues 214-216 (chymotrypsin numbering system) of the protease (Fig. 1A). This -sheet interaction between protease and inhibitor includes the following three, highly conserved hydrogen bonds between the two main chains: (i) the carbonyl of Ser-214 to the amide-NH of the P1 residue, (ii) the amide-NH of Gly-216 to the carbonyl of the P3 residue, and (iii) the carbonyl of Gly-216 to the amide-NH of the P3 residue (Fig. 1A).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Peptide Synthesis-- Except where noted, peptides were manually synthesized using highly optimized in situ neutralization protocols for t-butoxycarbonyl chemistry solid phase peptide synthesis (25). Peptides were deprotected and cleaved from the resin by treatment with anhydrous HF in the presence of 5% anisole for 1 h at 0 °C. Purification was accomplished by preparative reverse phase-HPLC,¹ and pure fractions were identified by electrospray mass spectrometry and lyophilized. The purified peptides were redissolved in H₂O to yield 20-40 mM stocks. Peptides containing ester linkages were analyzed further by treatment with 1 N NaOH at 37 °C for 30 min. Ester hydrolysis products were then separated by analytical HPLC and analyzed by electrospray mass spectrometry to verify the presence and correct location of the ester bond.

Incorporation of Ester Bonds-- The method for incorporating ester linkages into synthetic peptides has been described (26, 27). -Hydroxy acids were coupled without protection; 2.2 mmol of L-phenyllactic acid, L-lactic acid, or glycolic acid were coupled to the growing peptide resin (0.4 mmol) in N,N-dimethylformamide/dichloromethane (1:1) in the presence of 2.2 mmol of diisopropyl carbodiimide, 2.5 mmol of hydroxybenzotriazole, and 0.8 mmol of N-ethylmorpholine. Coupling proceeded to completion within 5 min for glycolic acid or 20 min for lactic acid or phenyllactic acid. Formation of the ester bond was accomplished by activating 2.2 mmol of t-butoxycarbonyl amino acid with 2.2 mmol of diisopropyl carbodiimide and coupling for 1 h in the presence of 0.8 mmol of N-ethylmorpholine and 0.04 mmol of 4-dimethylaminopyridine.

Kinetics of Peptide Bond Cleavage by Trypsin-- Trypsin was obtained from Sigma and was titrated fluorometrically using the titrant 4-methylumbelliferyl p-guanidinobenzoate. Enzyme concentrations in assays were 2-8 nM. Peptide concentrations ranging from 1 µM to 2 mM were treated with trypsin at 37 °C to achieve 10-20% cleavage. Reactions were stopped by the addition of trifluoroacetic acid to 0.33%, and percent cleavage was measured by reverse phase-HPLC. Kinetic constants were obtained by nonlinear regression analysis of the kinetic data fit to the Michaelis-Menten equation. These constants and their estimated uncertainties agreed very closely with values obtained by Eadie-Hofstee analysis. Kinetic data reported are an average of 3-4 assays.

K_I Determinations-- K_I values for trypsin inhibition were determined by measuring apparent K_m values for hydrolysis of the substrate O-methylsuccinyl-cyclohexyltyrosyl-glycyl-arginyl-p-nitroanilide in the presence of concentrations of inhibitor peptide ranging from 0.25 to 120 µM. K_m apparent values were plotted against inhibitor concentrations, and the resulting X intercept is equal to K_I.

Computational Docking and Molecular Modeling-- The AutoDock 3.0 suite of programs (28, 29) was used to dock the tripeptides (FFR, FGR, and GAK) and a synthetic molecule N-(4-toluenesulfonyl)-L-m-amidino-phenylalanyl-piperidine (3-TAPAP) to trypsin. The structure of the 3-TAPAP-trypsin complex is known (30), but the structures of trypsin with the tripeptides (FFR, FGR, and GAK) are not known. To model trypsin-tripeptide complexes, the structure of trypsin at 1.9 Å was taken from the Protein Data Bank (1pph) (30). No water molecules were considered during the docking of the two peptides, FFR and FGR, that contained a P1 arginine residue. During the docking of GAK, however, we included a single water molecule within the P1 pocket that is very highly conserved in trypsin complexes with molecules that contain a P1 lysine. Polar hydrogen atoms were added to the protein using the SYBYL modeling package (Tripos Associates, Inc.). The CH, CH₂, and CH₃ groups of the amino acid residues were treated as united atoms. No protein atoms were moved during the docking calculations. During the docking simulations, the amide torsions in each of the peptides were fixed. The initial structures of the tripeptides were generated using the Insight II molecular modeling program from Molecular Simulations Inc. (31). In the present work a genetic algorithm was used as the search method. In each docking simulation the population size was set to 50, and 27,000 generations were run. We performed 100 docking runs for each of the tripeptides and 3-TAPAP. The structures generated were then clustered using a root mean square tolerance value of 2.0 Å. In the case of 3-TAPAP our procedure produced 8 clusters. For FFR, FGR, and GAK we generated 3, 4, and 7 clusters, respectively, with a maximum of 9 individual structures in each cluster. The lowest energy complex of 3-TAPAP with trypsin closely resembles the x-ray structure. In the lowest energy complex of FFR, the Arg side chain was not placed in the specificity pocket. However, in the second lowest energy conformer, approximately 1.1 kcal/mol higher than conformer-1, the Arg side chain is placed in the specificity pocket. The predicted free energy of binding for FFR, FGR, and GAK complexes was 14.45, 12.32, and 15.5 kcal/mol, respectively, and these AutoDock-derived low energy complexes (conformer-2 in the case of FFR and conformer-1 in the cases of FGR and GAK) were subsequently optimized using the DISCOVER suite of programs (31). In this energy optimization procedure both the protein and the substrates were allowed to move. We used a cff91 force field for the energy refinement. The optimization was continued until the gradient fell below 0.1 kcal/Å.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Analysis of Hydrogen Bond Contributions to Peptide Hydrolysis-- It is generally accepted that the canonical conformation of the reactive center loop of standard inhibitors represents a substrate-like binding mode for chymotrypsin family enzymes and that the three highly conserved hydrogen bonds described above make important contributions to the catalysis of substrate cleavage by these proteases. To test this assertion, we synthesized and characterized model peptide substrates of trypsin and also variants of these substrates in which either the P3 or P2 amide-NH was replaced by an oxygen atom, converting the corresponding peptide bond into an ester.

Replacement of the amide bond with an ester bond in the context of a peptide is an established strategy for investigating the role of peptide main chain hydrogen bonding in biochemical interactions (26, 27, 32-35). Amide and ester bonds are very similar in terms of structure and conformational preferences. Both occur predominantly in the trans planar configuration (36), and in the context of peptide chains, ester and amide bonds have similar Ramachandran plots (37). In contrast to their structural similarity, however, the hydrogen bonding properties of ester linkages within a peptide main chain are altered in two major ways compared with natural peptide bonds. First, the ester carbonyl has a pK_a that is 4 units lower than that of an amide carbonyl; consequently, the ester carbonyl is a much weaker hydrogen bond acceptor than an amide carbonyl (35, 38). Second, the hydrogen bond donating amide-NH is replaced with the electronegative oxygen atom of the ester.

Fig. 1B presents an alignment of the ester containing substrates with a normal peptide substrate in the canonical conformation. As indicated, replacement of the P3 residue with an -hydroxy acid would eliminate the hydrogen bond between the P3 amide-NH in the normal peptide substrate and the carbonyl of Gly-216 of trypsin. In addition, replacement of the P2 residue with an -hydroxy acid would substantially weaken the hydrogen bond between the neighboring P3 carbonyl and the amide-NH of Gly-216 (Fig. 1B).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   A, representation of the canonical binding mode of a peptide to trypsin that indicates highly conserved hydrogen bonds with dashed lines (46). B, alignment, in the canonical conformation, of the P3 residue of a native peptide substrate and two peptide analogs with trypsin. Dotted lines represent favorable hydrogen bonding interactions, whereas dotted lines covered by an X indicate canonical hydrogen bonding interactions that are eliminated or substantially compromised by chemical alteration of the main chain in the peptide analogs.

Trypsin cleaved model substrates I-III, which contained natural peptide main chains, efficiently, with k_cat/K_m values of approximately 1.3 × 10⁶ M¹ s¹ (peptide I), 3.6 × 10⁶ M¹ s¹ (peptide II), or 3.2 × 10⁶ M¹ s¹ (peptide III) (Table I). Modified peptides IV-VI, which contained an ester linkage between the P4 and P3 residues, were also cleaved efficiently by trypsin with k_cat/K_m values of 2.8 × 10⁶ M¹ s¹, 3.9 × 10⁶ M¹ s¹, or 3.4 × 10⁶ M¹ s¹, respectively (Table I). Loss of the ability to form the hydrogen bond between the P3 amide-NH of a substrate and the carbonyl of Gly-216 of trypsin, as observed in complexes of chymotrypsin-like enzymes and standard inhibitors, had little or no effect on productive interactions between trypsin and the three model peptide substrates. These "atomic mutations" in the model substrates produced either a 2.2-fold (peptide IV) or a 1.1-fold (peptides V and VI) increase in catalytic efficiency. These modest alterations in k_cat/K_m correspond to a G for stabilization of the transition state during catalysis of only 0.47 (peptide IV), 0.05 (peptide V), or 0.04 (peptide VI) kcal/mol.

Similarly, variants of the three model peptides that contained an ester linkage between the P3 and P2 residues (peptides VII-IX) were also cleaved efficiently by trypsin with k_cat/K_m values of 2.3 × 10⁶ M¹ s¹ (peptide VII), 3.8 × 10⁶ M¹ s¹ (peptide VIII), or 2.3 × 10⁶ M¹ s¹ (peptide IX) (Table I). Replacement of the P2 amide-NH of the model substrates with an oxygen atom, therefore, produced a 1.1-fold increase in catalysis by trypsin (peptide VIII), reduced catalysis by 28% (peptide IX), or increased catalysis by a factor of approximately 1.8 (peptide VII). These small effects on catalysis correspond to a G for stabilization of the transition state during catalysis of 0.03 (peptide VIII), +0.20 (peptide IX), or 0.35 (peptide VII) kcal/mol (Table I).

Efficient catalysis by trypsin of peptide substrates IV-IX is inconsistent with the current belief that the three, highly conserved hydrogen bonds observed in complexes between chymotrypsin family enzymes and the P3-P1 residues of peptidic or "standard" proteinaceous inhibitors represent productive, substrate-like interactions between enzyme and substrate. Alteration of the main chain of the model substrates to eliminate the possibility of forming one of these hydrogen bonds (peptides IV-VI) or to greatly weaken another of these hydrogen bonds, if formed (peptides VII-IX), had little effect on the hydrolysis of the substrates by trypsin. One hypothesis that is consistent with these data is that the widely observed, canonical conformation of the reactive center loop of standard inhibitors actually represents a nonproductive binding mode to chymotrypsin family enzymes. This new hypothesis is particularly attractive because the canonical conformation has been defined by observation of complexes between proteases and standard inhibitors, which bind tightly but, unlike good substrates, are cleaved very slowly. An alternative, new hypothesis, which is also consistent with our data, is that substrates do bind chymotrypsin family enzymes in a canonical conformation, but that at least two of the three hydrogen bonds between the P3-P1 residues of the substrate and residues 214-216 of the enzyme contribute equally to stabilization of the ground state (the Michaelis complex) and the transition state.

Analysis of Hydrogen Bond Contributions to Trypsin Inhibition-- To test further the hypothesis that the canonical conformation represents a nonproductive, rather than a productive, substrate-like binding mode to chymotrypsin family enzymes, we used a model inhibitory peptide of trypsin (peptide X) to perform experiments similar to those described above for model trypsin substrates. This inhibitory peptide was initially isolated from a combinatorial peptide library by Eichler and Houghten (39) and was shown by Coombs et al. (40) to be a tight binding, slowly hydrolyzed substrate of trypsin. Alteration of the main chain of the inhibitory peptide to introduce an ester linkage between the P4 and P3 (peptide XI) or P3 and P2 (peptide XII) residues increased catalysis of the inhibitory peptide by trypsin by a factor of 1.8 or 2.0, respectively. These small increases in catalysis correspond to a G for stabilization of the transition state during catalysis of approximately 0.4 kcal/mol.

In contrast to the modest effects on catalysis, removal of specific hydrogen bonding capacity by replacement of a main chain amide-NH group with an oxygen atom significantly compromised the inhibitory properties of peptide X (Table II). Replacement of the P3 amide-NH of this peptide with an oxygen atom (peptide XI) increased the K_I for trypsin by a factor of approximately 8.5. Similarly, replacement of the P2 amide-NH of peptide VII with an oxygen atom (peptide XII) increased the K_I for trypsin by a factor of approximately 7. These increases in K_I correspond to G values of approximately 1.2 kcal/mol, which is consistent with the loss of a single important hydrogen bond and similar to previous studies of amide to ester replacements within an -helix that destabilized T4 lysozyme by 0.7-0.9 kcal/mol (34) and within an antiparallel -sheet that reduced the stability of staphylococcal nuclease by 1.5-2.5 kcal/mol (35). These data strongly suggest that both the main chain amide-NH and carbonyl groups of the P3 glycine of peptide X form hydrogen bonds, presumably with the main chain amide-NH and carbonyl groups of Gly-216 of trypsin, that significantly enhance the inhibitory potency of the peptide. Consequently, these data support the hypothesis that the widely observed canonical conformation of the reactive center loop of standard inhibitors represents a nonproductive binding mode for chymotrypsin family enzymes.

Prediction of Alternate Substrate Binding Modes by Computational Docking-- The kinetic data described above appear inconsistent with assertions that the "canonical conformation" represents the productive binding mode of substrates to chymotrypsin family serine proteases. Consequently, because direct examination of catalysis by serine proteases is not possible using current structural methods, we examined whether computational docking methods and molecular modeling could be used to propose an alternative binding mode for substrates that was consistent with our biochemical data. The program AutoDock 3.0 (28, 29) was used to dock the tripeptides FGR (the P3-P1 residues of model substrate I) and FFR (P3-P1 for model substrates II and III) to trypsin (Protein Data Bank code 1pph (30)). Prior to docking, we acylated the amino terminus and amidated the carboxyl terminus of each peptide, because we have previously shown that these modifications significantly enhanced the hydrolysis of short peptide substrates by trypsin (41). AutoDock 3.0 allowed free rotation around all sp³ type bonds in the substrate, only constraining the peptide bond (28, 29). The trypsin structure was fixed during docking. Low energy complexes for both substrates were further minimized using DISCOVER (31), which allowed rotational freedom for both the substrate and the side chains of the enzyme.

Both final, minimized complexes predicted a similar, "noncanonical" binding mode for the small substrates (Fig. 2, A and B). In agreement with the biochemical data described above, neither the amide-NH nor the carbonyl group of the P3 residue of the substrate forms a hydrogen bond with trypsin in either minimized complex. In the model of the FFR-trypsin complex, the amide-NH and carbonyl oxygen of the P3 residue actually point away from the trypsin 214-216 main chain and extend into solvent. The P3 amide-NH is located approximately 5 Å from the carbonyl group of Gly-216 of trypsin, whereas the P3 carbonyl is located approximately 7 Å from the amide group of Gly-216. Similarly, these distances are, respectively, approximately 6 Å and 5 Å in the trypsin-FGR model (Fig. 2, A and B).

View larger version (40K): [in this window] [in a new window]   Fig. 2.   A, active site region of the modeled, minimized complex between trypsin (Protein Data Bank code 1pph) (30) and the tripeptide FFR. B, active site region of the modeled, minimized complex between trypsin and the tripeptide FGR. C, active site region of the modeled, minimized complex between trypsin and the tripeptide GAK.

In addition to the absence of canonical, main chain hydrogen bonding, the side chain interactions of the modeled substrates also differ from those in the canonical conformation (Figs. 2, A and B, and 3). In both minimized complexes, the side chain of the P2 residue extends away from the protease surface and into solvent, and the side chain of the P3 residue contacts the protease surface where it occupies the "aryl binding site" (2, 42) of the enzyme. In contrast, in the canonical binding mode the P2 residue extends toward the protease surface and fills the S2 pocket, whereas the P3 residue extends away from the active site cleft and into the solvent (Fig. 3). The placement of substrate side chains in the modeled complex, rather than in the canonical complex, appears consistent with our previous demonstration that the ability of a peptide substrate to discriminate between the closely related chymotrypsin family enzymes tissue-type plasminogen activator and urokinase plasminogen activator is mediated primarily by the P3 residue (43-45).

View larger version (101K): [in this window] [in a new window]   Fig. 3.   Superposition of the P3-P1 residues (green) of the trypsin-bovine pancreatic trypsin inhibitor complex (Protein Data Bank code 2ptc) (47) and the modeled, minimized substrate FGR (magenta). Canonical, main chain hydrogen bonding interactions between bovine pancreatic trypsin inhibitor and trypsin are indicated by yellow spheres.

In contrast to the substrate-derived peptides FGR and FFR (Fig. 2, A and B), AutoDock predicted that the inhibitor-derived peptide GAK bound trypsin in a canonical conformation (Fig. 2C). The predicted, lowest energy enzyme-GAK complex contained the characteristic short, antiparallel -sheet between enzyme and peptide and included all three conserved, canonical hydrogen bonds between the two main chains. For example, the distance between the P3 carbonyl oxygen of the peptide and the amide-NH of glycine 216 was approximately 2.8 A, and the distance between the P3 amide-NH of the peptide and the carbonyl oxygen of glycine 216 was approximately 2.7 A. This predicted, canonical interaction between trypsin and GAK, an inhibitor-derived peptide, is consistent with the hypothesis that the canonical conformation represents a nonproductive interaction. Moreover, this result is also consistent with the demonstration in this study that loss of the ability to form either of two canonical hydrogen bonds significantly reduced the inhibitory potency of the GAK-containing peptide.

Conclusions-- The assumption that peptide substrates bind to chymotrypsin-like enzymes in a canonical conformation is difficult to reconcile with two experimental results: (i) the observation of this study that the ability of the P3 amide-NH and carbonyl groups to form strong hydrogen bonds with the trypsin main chain does not contribute to the catalysis of model substrates and (ii) our previous demonstration that the P3 residue, which extends away from the protease surface and into solvent in the canonical conformation, can mediate the tissue-type plasminogen activator/urokinase plasminogen activator selectivity of peptide substrates. By contrast, an alternative binding mode, suggested by computational methods in which the P3 residue occupies the aryl binding site of the enzyme, immediately suggests a clear molecular rationale for both of these experimental observations. These data also raise the possibility that more than one productive binding mode may exist for substrates of chymotrypsin-like enzymes that exhibit broad specificity. Loss of this ability to interact productively with multiple substrate binding modes may be one mechanism by which these enzymes can evolve stringent specificity.

These studies raise fundamental questions regarding the interactions of peptide substrates with chymotrypsin family serine proteases, one of the most important model systems of enzymology, and provide provocative new hypotheses to be tested in future experiments. We suggest that the precise, productive binding mode(s) of substrates to chymotrypsin family serine proteases must be considered an open question.

	FOOTNOTES

^* This work was supported in part by National Institutes of Health Grant R01 HL52475 (to E. L. M.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Corvas International, 3030 Science Park Rd., San Diego, CA 92121.

	ABBREVIATIONS

The abbreviations used are: HPLC, high pressure liquid chromatography; 3-TAPAP, N-(4-toluenesulfonyl)-L-m-amidino-phenylalanylpiperidine.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				C.-T. Chen, E.-W. Yang, H.-J. Hsu, Y.-K. Sun, W.-L. Hsu, and A.-S. Yang Protease substrate site predictors derived from machine learning on multilevel substrate phage display data Bioinformatics, December 1, 2008; 24(23): 2691 - 2697. [Abstract] [Full Text] [PDF]

				H.-J. Hsu, K.-C. Tsai, Y.-K. Sun, H.-J. Chang, Y.-J. Huang, H.-M. Yu, C.-H. Lin, S.-S. Mao, and A.-S. Yang Factor Xa Active Site Substrate Specificity with Substrate Phage Display and Computational Molecular Modeling J. Biol. Chem., May 2, 2008; 283(18): 12343 - 12353. [Abstract] [Full Text] [PDF]

				E. S. Radisky and D. E. Koshland Jr. From the Cover: A clogged gutter mechanism for protease inhibitors PNAS, August 6, 2002; 99(16): 10316 - 10321. [Abstract] [Full Text] [PDF]

				J. R. Hesselberth, D. Miller, J. Robertus, and A. D. Ellington In Vitro Selection of RNA Molecules That Inhibit the Activity of Ricin A-chain J. Biol. Chem., February 18, 2000; 275(7): 4937 - 4942. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Coombs, G. S. Articles by Madison, E. L. Search for Related Content PubMed PubMed Citation Articles by Coombs, G. S. Articles by Madison, E. L. Social Bookmarking                      What's this?