Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M506766200 on September 30, 2005

J. Biol. Chem., Vol. 280, Issue 49, 41077-41089, December 9, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/49/41077    most recent
M506766200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tang, J.
Right arrow Articles by Katz, B. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tang, J.
Right arrow Articles by Katz, B. A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Expression, Crystallization, and Three-dimensional Structure of the Catalytic Domain of Human Plasma Kallikrein*

Jie Tang, Christine Luong Yu, Steven R. Williams, Eric Springman1, Douglas Jeffery, Paul A. Sprengeler, Alberto Estevez, Jun Sampang, William Shrader, Jeff Spencer, Wendy Young, Mary McGrath, and Bradley A. Katz2

From the Department of Structural Chemistry, Celera Genomics, South San Francisco, California 94080

Received for publication, June 21, 2005 , and in revised form, September 29, 2005.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Plasma kallikrein is a serine protease that has many important functions, including modulation of blood pressure, complement activation, and mediation and maintenance of inflammatory responses. Although plasma kallikrein has been purified for 40 years, its structure has not been elucidated. In this report, we described two systems (Pichia pastoris and baculovirus/Sf9 cells) for expression of the protease domain of plasma kallikrein, along with the purification and high resolution crystal structures of the two recombinant forms. In the Pichia pastoris system, the protease domain was expressed as a heterogeneously glycosylated zymogen that was activated by limited trypsin digestion and treated with endoglycosidase H deglycosidase to reduce heterogeneity from the glycosylation. The resulting protein was chromatographically resolved into four components, one of which was crystallized. In the baculovirus/Sf9 system, homogeneous, crystallizable, and nonglycosylated protein was expressed after mutagenizing three asparagines (the glycosylation sites) to glutamates. When assayed against the peptide substrates, pefachrome-PK and oxidized insulin B chain, both forms of the protease domain were found to have catalytic activity similar to that of the full-length protein. Crystallization and x-ray crystal structure determination of both forms have yielded the first three-dimensional views of the catalytic domain of plasma kallikrein. The structures, determined at 1.85 Å for the endoglycosidase H-deglycosylated protease domain produced from P. pastoris and at 1.40 Å for the mutagenically deglycosylated form produced from Sf9 cells, show that the protease domain adopts a typical chymotrypsin-like serine protease conformation. The structural information provides insights into the biochemical and enzymatic properties of plasma kallikrein and paves the way for structure-based design of protease inhibitors that are selective either for or against plasma kallikrein.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The name "kallikrein" was originally introduced in the 1930s to describe the proteolytic activity from the pancreas responsible for cleavage of kininogen (1, 2). The kallikreins are now divided into the following two major categories: plasma kallikrein (EC 3.4.21.34 [EC] ) and tissue kallikreins (3, 4). Tissue kallikreins were initially discovered as an activity from pancreatic or renal tissues effecting the release of kinin from kininogens. Plasma kallikrein was discovered from a hereditary abnormality termed Fletcher trait, characterized by a prolonged partial thromboplastin time that becomes normal if the plasma is exposed to a clot-promoting surface (5). Prekallikrein (PK)3 was identified as the protein whose absence is responsible for the Fletcher trait (6).

Since its discovery in 1965, plasma kallikrein has been biochemically characterized in great detail. The human cDNA of PK encodes a signal peptide of 19 amino acids and a mature prekallikrein of 619 amino acids (7). Activation of PK to kallikrein occurs through a cleavage at the Arg371-Ile372 bond (based on the numbering of the mature protein), producing a two-subunit protein involving a heavy chain and a light chain (6), linked through a disulfide bond between Cys364 and Cys484 (8). The N-terminal heavy chain of 371 amino acids contains four Apple domains whose homologues are also found in factor XI (fXI) (9), whereas the C-terminal light chain forms the protease domain. The two chains have been isolated through reduction and alkylation of the purified enzyme and through affinity capture of the heavy chain on immobilized high molecular weight kinin-Sepharose (10). The purified light chain has amidolytic activity similar to that of the native enzyme, but it is less active in kaolin-dependent coagulation assays. Both the light and heavy chains are glycosylated; PK from human blood contains 15% carbohydrate and has five potential sites for N-glycosylation. Because of heterogeneity in the carbohydrate moiety, the light chain can exist as 36- and 33-kDa isoforms.

Although plasma and tissue kallikreins are related, and share the historical name, kallikrein, for kininogen degrading activity, plasma kallikrein differs from tissue kallikrein members in gene structure, amino acid sequence, protein subunit structure, substrate specificity, and physiological functions (3, 4, 8, 11). Tissue kallikreins, currently comprising 15 identified members (hK1 through hK15), are encoded by genes clustered at 19q13.3-19q13.4 of the human genome, whereas plasma kallikrein is encoded by a single gene on human chromosome 4q35. In contrast to the multidomain structure of plasma kallikrein (homologous to that of fXI), each tissue kallikrein is composed of a protease domain alone. The degree of sequence identity between the protease domain of plasma kallikrein and individual members of the tissue kallikrein family ranges from 30 to 36%, although it is much higher between the protease domains of plasma kallikrein and fXI (66%). Thus, plasma kallikrein bears more resemblance to the coagulation factor, fXI, than to members of the tissue kallikrein family.

Plasma kallikrein plays a central role in the kinin-generating pathways (PK system) and in the surface-mediated "contact system" (8, 12). It is synthesized predominantly in the liver as a proenzyme, prekallikrein, also known as prokallikrein (13). In plasma, ~90% of the circulating PK is complexed with high molecular weight kininogen (HK) (14). On endothelial cells, the PK·HK complex binds to a multiprotein complex that consists of cytokeratin 1, urokinase-type plasminogen activator receptor, and gC1qR, a ubiquitously expressed, multicompartmental cellular protein involved in modulating complement, coagulation, and kinin cascades (15-21). The binding of PK·HK to the multiprotein complex results in the activation of PK by prolylcarboxypeptidase, a constitutively active serine protease located on endothelial membranes (22-23). The activated plasma kallikrein in turn cleaves HK to liberate bradykinin, activates factor XII (fXII) to factor XIIa (fXIIa), and converts prorenin to renin, the central component in the renin-angiotensin system (8, 23). By controlling the release of bradykinin (a potent vasodilator) and the activation of renin (the protease converting angiotensinogen to angiotensin I), plasma kallikrein is deeply involved in blood pressure regulation. Plasma kallikrein also participates in fibrinolysis by activating prourokinase and plasminogen to urokinase and plasmin, respectively (24-27). Finally, plasma kallikrein stimulates neutrophils to aggregate and release their lysosomal contents such as elastase (28, 29). Thus, plasma kallikrein has many important physiological functions, including modulation of blood pressure, complement activation, mediation of inflammatory responses, and maintenance of the balance between fibrin deposition and fibrinolysis (8).

In recent years, tremendous efforts to design specific protease inhibitors for treating diseases have yielded many life-saving drugs (30-33). One such field receiving considerable attention involves the blood coagulation pathways (34). Anticoagulant therapies are in great demand to block undesired thrombosis of surgery patients and to treat many lifethreatening diseases such as stroke, acute coronary syndrome, and deep vein/pulmonary embolism (35). Current therapies with heparin or coumarins have limited efficacy and safety records (36, 37), prompting synthesis of many small molecules intended to inhibit specific target proteases, such as thrombin, factor VIIa (fVIIa), factor IXa (fIXa), and factor Xa (fXa), within the coagulation pathways (34). However, because of the similarity among plasma kallikrein and the coagulation factors, including a preference for arginine at the S1 substrate-binding site, many inhibitors designed for other coagulation factors also inhibit plasma kallikrein.

The therapeutic efficacy of a protease inhibitor relies not only on high affinity for the disease target but also on attenuated affinity for antitargets, whose vital functions need to be maintained. Thus, structural information both for the protease target and for related anti-targets often provides the key for designing potent, selective, and safe inhibitors. High resolution structures can be used to enhance inhibitor selectivity by identifying major structural differences among protease homologues at distal binding sites, such as S4, S3, S2, S1', S2', S3', or S4' (38), or by harnessing more subtle differences at the S1 site. Thus, the determined crystal structures of proteases such as thrombin, fVIIa, and fXa have provided a wealth of information for increasing inhibitor potency and selectivity toward the particular protease target (33, 39-43). However, one important anti-target for these and other serine protease targets is plasma kallikrein, whose structure has been unavailable. As a step toward the design of inhibitors with enhanced selectivity for a target protease like fVIIa and against other anti-targets, including plasma kallikrein, we initiated efforts to establish high level expression systems for plasma kallikrein, to purify and crystallize the protein, and to determine its structure.

Plasma kallikrein is not only an important anti-target for therapeutic inhibitors designed toward other serine proteases, it is also emerging as an important target itself for some disease states (8, 22, 44, 45). For example, abnormal plasma kallikrein activity is implicated in the symptoms of hereditary angioedema (44, 45), a disease caused by decreased functional levels of the major physiological inhibitor of kallikrein, C1 inhibitor. During acute attacks of this disease, increased levels of bradykinin cause an increase in vascular permeability. Plasma kallikrein inhibitors suppress the defect in vascular permeability in C1 inhibitor knock-out mice (46) and thus may constitute treatment for hereditary angioedema. Other disorders involving plasma kallikrein are septicemia and septic shock (23, 47). The overproduction of bradykinin is implicated in the pathogenesis of septic shock through its ability to lower blood pressure in both humans and animals. Because of the important role of plasma kallikrein in these disease states, a structure of this enzyme would be a critical asset for structure-based design of a potent, selective, and small molecule therapeutic.

High resolution structures of human tissue kallikreins, hK1 (48, 49), hK6 (50), and pro-hK6 (51) have been determined. However, the degree of sequence identity of these tissue kallikrein structures to that of plasma kallikrein is at best 34% (for hK6). More importantly, the loop sequences that form distal binding sites (such as S1'), deemed important for inhibitor design, differ significantly between tissue and plasma kallikrein. Thus the available structures of tissue kallikrein (or of other proteases) can provide only rough guidance in the design of potent, selective inhibitors of plasma kallikrein.

In this report, we describe the expression of the protease domain of plasma kallikrein in the methylotrophic yeast Pichia pastoris and the production of crystallizable protein by removal of Endo H-sensitive glycosylation. We also describe the expression of the protease domain in the baculovirus system as a nonglycosylated form with all three N-linked glycosylation sites removed through site-directed mutagenesis. Both forms were purified and crystallized to yield the first reported crystal structures of the protease domain of human plasma kallikrein (at 1.85 Å for the enzymatically deglycosylated form and at 1.40 Å for the mutagenically deglycosylated form). The structures show that the protease domain of plasma kallikrein adopts a typical chymotrypsin-like serine protease fold and provide insight into substrate selectivity related to the function of this important protease. In addition, well defined structural features, including the unique S1' site, pave the way for the structure-based design of protease inhibitors that are selective either for or against plasma kallikrein.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Expression of Plasma Kallikrein Protease Domain in P. pastoris—A cDNA sequence encoding the plasma kallikrein protease domain, from Asn357 to Ala619 (based on the numbering system of mature prekallikrein, signal sequence not included), in which Cys364 and Cys484 were mutagenized to serines, was inserted into pPIC9. The pPIC9 plasmid containing the insert was linearized with SalI and electroporated into P. pastoris strain KM71. P. pastoris clones were analyzed for expression of the plasma kallikrein protease domain, and an expression clone was inoculated into the complex media BMMY. After overnight incubation at 30 °C with shaking, the cells were then used to inoculate a 15-l bioreactor containing ~8 liters of BMMY media. Fermentation with glycerol feed continued for 24 h to a cell density whose absorbance was 500 at 600 nm. The expression of recombinant plasma kallikrein protease domain was induced by switching the carbon source to methanol for 48-60 h. The fermentation was harvested by centrifuging the culture in a Sorvall RC3B at 5000 rpm for 30 min. The supernatant was recovered and filtered through a 0.22-µm filter to remove particles.

Expression of Mutagenically Deglycosylated Plasma Kallikrein in Sf9 Cells—The nonglycosylated plasma kallikrein protease domain was expressed in the baculovirus system. The glycoprotein 67 signal peptide sequence (MLLVNDSHQGFNKEHTSKMVSAIVLYVLLAAAAHSAFA) was synthesized from overlapping oligonucleotides and cloned into pUC19. The cDNA fragment encoding Asn357-Ala619 of plasma kallikrein (numbering system based on the mature protein sequence of prekallikrein) was fused to the glycoprotein 67 signal peptide sequence by standard techniques. Sequences encoding the glycosylation asparagines, Asn377, Asn434, and Asn475, and the cysteines, Cys364 and Cys484, were mutated to encode Glu377, Glu434, Glu475, Ser364, and Ser484, respectively, through site-directed mutagenesis using a Stratagene QuikChange multisite-directed mutagenesis kit. The mutant cDNA was subcloned into pFastBac1 (Invitrogen). The expression baculovirus stock was obtained by transposition in Escherichia coli strain DH10Bac and by transfection of Sf9 cells with the recovered bacmid. To express mutagenically deglycosylated plasma kallikrein protease domain, Sf9 cells were infected with the recombinant baculovirus at a multiplicity of infection of 5. After infection for 65-72 h, the culture was harvested by centrifugation in a Sorvall RC3B at 5000 rpm for 30 min. The supernatant was recovered and filtered through a 0.22-µm filter.

Purification and Characterization of Recombinant Endo H-deglycosylated Plasma Kallikrein Protease Domain—The recombinant plasma kallikrein protease domain expressed in P. pastoris was secreted into the culture media as a heterogeneously glycosylated zymogen that could be activated through limited trypsin digestion. A typical purification process is summarized in TABLE ONE. The P. pastoris media (6 liters) was concentrated, and buffer was exchanged against 50 mM Na3PO4, pH 6.5, with 10-kDa cut-off ultrafiltration system (Millipore). To remove contaminating proteins, the processed Pichia medium (600 ml) was applied twice to a 300-ml Q-Sepharose Fast Flow (GE Healthcare) column that had been equilibrated with 50 mM Na3PO4, pH 6.5. Flow-through fractions from the Q-Sepharose column were combined and diluted with 2 volumes of 5 mM MES, pH 5.5, and the pH was adjusted to 5.5. This solution was then applied to a Source 15S (15-µm resin) column (1.6 x 10 cm, GE Healthcare) that had been equilibrated with 30 mM MES, pH 5.5. After washing with 10 column volumes of 30 mM MES, pH 5.5, the protein was eluted from the Source S column with 30 mM MES, 350 mM NaCl, pH 5.5. To remove the Endo H-sensitive glycosylation from the recombinant plasma kallikrein, Endo H (Roche Applied Science) was added to the Source S pool at a ratio of 30 milliunits per mg of total protein, and the reaction mixture was incubated at 37 °C for 100 min. The reaction mixture was then diluted with an equal volume of 100 mM Tris-HCl, 300 mM NaCl, pH 7.6, and the final pH adjusted to 7.6. The recombinant plasma kallikrein in the diluted Endo H reaction mixture was activated by adding freshly dissolved sequencing grade trypsin (Roche Applied Science) at a ratio of 2 µg of trypsin per mg of total protein and by incubating at 37 °C for 60 min. The activated, deglycosylated plasma kallikrein protease domain was then loaded onto a 16-ml benzamidine-Sepharose column (high substitute, GE Healthcare) that had been equilibrated with 50 mM Na3PO4, pH 7.0. After washing the column with 5 volumes of 50 mM Na3PO4, 200 mM NaCl, pH 7.0, and then with 3 volumes of 50 mM Na3PO4, pH 7.0, the kallikrein was eluted with 50 mM Na3PO4, 50 mM benzamidine, pH 7.0. Fractions containing protein were pooled and dialyzed against 30 mM MES, pH 5.5. The dialyzed protein solution was further fractionated with a Source 15S column (1 x 10 cm) and eluted with 30 column volumes of 30 mM MES, pH 5.5, containing a linear NaCl gradient from 0 to 400 mM. This Endo H-treated plasma kallikrein protease domain was resolved into four peaks (presumably due to differential, residual Endo H-resistant glycosylation), the first of which eluted at a conductivity of ~ 8 ms/cm. Dynamic light scattering experiments indicated that the protein of each peak was monodisperse. The protein from peak 1 (hereafter referred to as Endo H-deglycosylated plasma kallikrein) was used for structural studies.


View this table:
[in this window]
[in a new window]
  		TABLE ONE
Purification of recombinant plasma kallikrein from conditioned P. pastoris media

 
Purification and Characterization of Recombinant Mutagenically Deglycosylated Plasma Kallikrein Protease Domain—Mutagenically deglycosylated plasma kallikrein protease domain was purified from 3.0 liters of conditioned media of infected Sf9 cells (TABLE TWO). After slow addition of (NH4)2SO4 to the media (with stirring) to bring the final concentration to 1.8 M, the medium was loaded onto a phenyl-Sepharose high performance column (GE Healthcare, 2.6 x 15 cm) that had been equilibrated with 30 mM Na3PO4, pH 7.0, containing 2 M (NH4)2SO4. The phenyl-Sepharose column was then washed with the equilibration buffer until a base-line absorbance was achieved. Plasma kallikrein protease domain was eluted with 30 mM Na3PO4, pH 7.0, and then dialyzed overnight against 50 mM Na3PO4, pH 6.5. The dialyzed protein was applied to a 150-ml Q-Sepharose Fast Flow column that had been equilibrated with 50 mM Na3PO4, pH 6.5. Flow-through fractions from the Q-Sepharose column (containing the plasma kallikrein) were recovered and diluted with an equal volume of 50 mM Na3PO4, pH 7.6, 300 mM NaCl, and the pH was adjusted to 7.6. Activation of the plasma kallikrein protease domain was achieved by adding sequencing grade trypsin at the ratio of 2 µg of trypsin per mg of total protein and by incubating the reaction mixture at 37 °C for 60 min. The activated protein was further purified with a benzamidine-Sepharose column using conditions similar to those described for purification of Endo H-deglycosylated plasma kallikrein. After concentration by ultrafiltration with a 10-kDa membrane (Millipore), the protein was further purified with a high load Superdex 200 column (1.6 x 60 cm, GE Healthcare) using 30 mM Na3PO4, 140 mM NaCl, pH 7.4, as the eluent. The fractions containing plasma kallikrein were combined and dialyzed against 30 mM MES, pH 5.5. The dialyzed sample was finally purified on a Source 15S column using conditions similar to those described for purification of Endo H-deglycosylated plasma kallikrein. From 3 liters of conditioned media, 0.7 mg of purified plasma kallikrein protease domain was recovered.


View this table:
[in this window]
[in a new window]
  		TABLE TWO
Purification of recombinant kallikrein from conditioned Sf9 media

 
Enzymatic Activity of Full-length Plasma Kallikrein and of the Recombinant Forms of the Protease Domain—Amidolytic activities of fulllength plasma kallikrein and of the recombinant forms of the protease domain were assayed at room temperature in 100-µl reaction mixtures containing 50 mM Tris-HCl, 150 mM NaCl, 2.5% Me2SO, pH 7.6, and 250 µM pefachrome-PK (2AcOH-H-D-butyl-CHA-Arg-pNA, Centerchem, Inc.) in 96-well microtiter plates. The release of p-nitroaniline was monitored at 405 nm with a SpectraMax plate reader controlled by SoftMax Pro software. Enzyme concentration was determined by bicinchoninic acid assays (Pierce). Kinetic parameters were determined by nonlinear least squares regression analysis of the absorbance versus time data for a series of assays involving substrate concentrations ranging from 2 to 500 µM. The values were calculated by fitting the data points to the Michaelis-Menten equation using Kaleidagraph (Synergy Software).

The peptidase specificities of full-length plasma kallikrein and of the recombinant forms of the protease domain were compared with one another by assaying the rate of cleavage of oxidized insulin B chain and the nature of the cleavage products. In each reaction, 75 µg of oxidized insulin B chain (Sigma) in 50 mM Tris-HCl, 150 mM NaCl, pH 7.6, was incubated with 7.5 µg of full-length plasma kallikrein or with 1.5 µg of recombinant protease domain in a 75-µl reaction. In control reactions, no protease was included. After incubation for 1 or 24 h, 25 µl of the reaction mixture was removed, and the reaction was stopped by adding 25 µl of 8 M guanidine HCl, pH 3.0. The digestion mixture was loaded onto a Zorbax C18 column (4.6 mm x 15 cm; Agilent Technologies) with a HP1090 high pressure liquid chromatography system, and the column then washed with 95% mobile phase A (H2O, 0.1% trifluoroacetic acid), 5% mobile phase B (acetonitrile, 0.08% trifluoroacetic acid) for 5 min at a flow rate of 1.0 ml/min. The peptides were eluted from the column with a linear gradient from 5 to 95% mobile phase B over a 45-min period. Peptides in the chromatogram were monitored at 220 nm, and their identities were established by collision-induced dissociation followed by peptide sequencing using a Q-Star mass spectrometer from Applied Biosystems. The oxidized insulin B chain (peak IV) eluted at 22.68 min, the Phe1-Arg22 fragment (peak III) at 20.56 min, Gly23-Ala30 (peak II) at 17.65 min, and Gly23-Lys29 (peak 1) at 17.26 min. The areas of these peaks were integrated to yield the relative amounts after 1 h of incubation (TABLE THREE, top), and after 24 h incubation (TABLE THREE, bottom).


View this table:
[in this window]
[in a new window]
  		TABLE THREE
Relative amounts (percentages) of peptide products produced by incubating oxidized insulin B chain with various forms of plasma kallikrein for 1 or 24 h

The cleavage sites of oxidized insulin B chain are shown by highlighting the P1 residue (R or K) in boldface as follows:

Peak IV, FVNQHLC (SO3H) GSHLVEALYLVC (SO3H) GERGFFYTPKA (oxidized insulin B chain)

Peak III, FVNQHLC (SO3H) GSHLVEALYLVC (SO3H) GER

Peak II, GFFYTPKA

Peak I, GFFYTPK

 
Determination of Inhibition Constants for Full-length Plasma Kallikrein and for Other Serine Proteases—The Ki values involving small molecule inhibitors toward a panel of serine proteases, including plasma kallikrein, were determined at pH 7.4 as described (52). Apparent inhibition constants, K'i values, were calculated from the enzyme hydrolysis rate data collected at various inhibitor concentrations with BatchKi (developed and provided by Dr. Petr Kuzmic, Biokin Ltd., Pullman, MA) using methodology similar to that described for tight-binding inhibitors (53). K'i values were converted to Ki values by the formula, Ki = K'i/(1 + S/Km).

Crystallization and X-ray Diffraction Data Collection of Recombinant Plasma Kallikrein—Human plasma kallikrein (either Endo H-deglycosylated or mutagenically deglycosylated), in 30 mM MES, 75 mM NaCl, pH 5.5, was concentrated to 12-18 mg/ml in the presence of 20 mM benzamidine (Sigma) using the Amicon Centricon-10 system (Millipore). The benzamidine complex of the enzymatically deglycosylated form was crystallized from sitting drops, composed of 0.5 µl of the protein solution and 0.5 µl of the reservoir solution (25% PEG 6K, 0.10 M MES, pH 6.5). The benzamidine complex of the mutagenically deglycosylated form was crystallized from sitting drops composed of 0.5 µl of the protein solution and 0.5 µl of the reservoir solution (0.20 M potassium dihydrogen phosphate, 20% PEG 3350). Crystals of 0.1 mm in each dimension grew within 2 weeks at 17 °C.

X-ray diffraction data sets for plasma kallikrein crystals were collected at -160 °C at the Berkeley ALS Synchrotron ({lambda} = 1.000 Å). For the endo H-deglycosylated plasma kallikrein protease domain (space group = P21212, a = 79.89, b = 63.19, c = 50.31Å), the Quantum 315 CCD detector on beam-line 5.0.2 was used with the following parameters (distance = 235 cm; 24 s per 1.0° frame; 180° total). For the other construct, in which all three asparagines involved in N-glycosylation were changed to glutamates (space group P212121, a = 55.19, b = 57.36, c = 79.85 Å), data were collected with the Quantum 210 CCD detector (distance = 140 cm; 15 s per 1.0° frame; 180° total) at beam-line 5.0.1. The data sets were reduced with Denzo (54).

A Blast search of the Protein Data Bank for the structure most homologous to plasma kallikrein identified the extracellular region of hepsin (39% sequence identity), as an appropriate probe for molecular replacement. (Recently the structure of a much more homologous protease, fXIa (identity = 66%) has been deposited, Protein Data Bank code 1XX9 [PDB] ; see Ref. 55.) The molecular replacement program written by L. Tong as implemented in the X-sight subroutine licensed from Accelrys was used to solve the structure of enzymatically deglycosylated plasma kallikrein, using the hepsin probe structure (Protein Data Bank code 1P57 (56) obtained from the Protein Data Bank (57). The refined structure of enzymatically deglycosylated plasma kallikrein was then used to solve the structure of the mutagenically deglycosylated plasma kallikrein by molecular replacement.

To help determine the structures of some of the loops, automated fitting was carried out with the APR/warp package (warpNtrace mode) (58). Structures were refined manually by analysis of (|Fo| - |Fc|), {alpha}c and of (2|Fo| - |Fc|), {alpha}cmaps, followed by refinement with X-PLOR (59, 60). Map analysis and X-PLOR refinement were iterated until no further improvement in Rcryst or in the models could be achieved. X-ray diffraction and refinement statistics and Protein Data Bank accession codes are provided in TABLE FOUR.


View this table:
[in this window]
[in a new window]
  		TABLE FOUR
Crystallography of the two recombinant forms of the protease domain of human plasma kallikrein

 
Synthesis of Inhibitors—Synthesis of inhibitors is described in detail elsewhere (61-63).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Purification and Characterization of the Endo H- and Mutagenically Deglycosylated Plasma Kallikrein Protease Domain—The protease domain of plasma kallikrein (Asn357-Ala619) expressed in P. pastoris was heterogeneously glycosylated and secreted as a zymogen. The zymogen was activated by trypsin treatment, which removes the N-terminal peptide. Heterogeneity in the recombinant protease domain results from incomplete carbohydrate processing during biosynthesis. Endo H was used to remove most of the carbohydrate and to convert the protease domain to a single band on SDS-PAGE. This Endo H-deglycosylated protein could be further fractionated into three to four species with cation-exchange chromatography using Mono S or Source S columns (data not shown). These species differed in molecular mass by 1000-2000 Da, most likely due to differential glycosylation. Each was monodisperse as analyzed by dynamic light scattering. The Km values of the various species were similar to one another, as were the kcat/Km values, using pefachrome-PK as the substrate (data not shown). We used the first Source S peak for crystallization experiments because the extent of post-translational modification was the least, and the purification yield was the highest.



View larger version (30K):
[in this window]
[in a new window]
  		FIGURE 1.
Characterization of nonglycosylated recombinant plasma kallikrein protease domain from conditioned Sf9 cell media. The purified protein was analyzed with a reverse phase Zorbax CN column using water, 0.1% trifluoroacetic acid (mobile phase A) and acetonitrile, 0.08% trifluoroacetic acid (mobile phase B), and with electrospray mass spectrometry using a Q-Star mass spectrometer. The amino acid sequence of the mutant plasma kallikrein construct is shown. The N377E, N434E, and N475E mutations at the N-glycosylation sites are colored green, and the C364S and C484S mutations are shown in purple. The N-terminal residues (Asn357-Arg371) and C-terminal residues (Ala613-Ala619) residues, absent in the final purified protein, are colored red. (These C-terminal residues are removed during trypsin treatment.) Each of the cysteine pairs involved in the four disulfides is color-coded and given the appropriate superscript (S1, S2, S3, or S4).

 
To obtain a nonglycosylated plasma kallikrein protease domain and to confirm that the differences in molecular weight among Source S peaks were from N-glycosylation, we mutagenized all three N-glycosylation sites (Asn377, Asn434, and Asn475) to glutamates, and we expressed the mutant in a baculovirus system. The mutant kallikrein protease domain was secreted into culture media as a zymogen and could be activated by trypsin digestion. The purified mutant protease was homogeneous by cation-exchange chromatography and was resolved as a single peak on Mono S (data not shown). Liquid chromatography-mass spectrometry analysis of the purified nonglycosylated mutant plasma kallikrein protease domain indicated that it included residues from Ile372 to Lys612 (Ile16 to Lys248 basing on chymotrypsin numbering scheme) without any post-translational modification (Fig. 1). This protein, hereafter referred to as mutagenically deglycosylated plasma kallikrein, was found to be monodisperse by dynamic light scattering analysis and was used for crystallization and for structural studies.

Comparison of Activity of Recombinant Forms of Plasma Kallikrein Protease Domain and of the Full-length Protein from Blood—The activities of purified full-length plasma kallikrein and of the two recombinant deglycosylated protease domains were compared in substrate specificity analysis using oxidized insulin B chain and in kinetic analysis using pefachrome-PK. In the insulin B chain cleavage experiment (TABLE THREE), the native full-length plasma kallikrein from blood and the two recombinant forms of the protease domain each cleaved peptide bonds with arginine or lysine at the S1 site. Each of the three purified kallikreins first cleaved the oxidized insulin B chain between Arg22 and Gly23, thus converting the 30-residue insulin B chain into two fragments, an N-terminal 22-mer and a C-terminal 8-mer. In longer digestions, each protease then removed the last alanine residue from the insulin B chain C-terminal 8-residue peptide by cleaving the Lys29-Ala30 peptide bond (TABLE THREE). To compare the substrate affinity and enzymatic activity of the purified full-length protein and of the recombinant forms of the protease domain, pefachrome-PK was used to measure Km and kcat/Km values (TABLE FIVE). All three purified proteases had a similar Km value toward pefachrome-PK (~100 µM); the kcat/Km values were also similar to one another (~1 x 106). Thus, the recombinant forms of the plasma kallikrein protease domain from P. pastoris or from Sf9 cells have catalytic activities similar to that of the full-length native plasma kallikrein.


View this table:
[in this window]
[in a new window]
  		TABLE FIVE
Enzyme kinetic parameters of full-length plasma kallikrein and of the two recombinant forms of the protease domain

 
Crystal Structures of Endo H-deglycosylated and Mutagenically Deglycosylated Human Plasma Kallikrein Protease Domain—There is extensive sequence conservation between the protease domain of plasma kallikrein and of other trypsin-like serine proteases. The sequence encoding the protease domain of plasma kallikrein is 39% identical with that of the protease domain of hepsin and 38% with that of bovine trypsin. Thus it is not surprising that the structures of both mutagenically and Endo H-deglycosylated plasma kallikrein (Fig. 2) are typical of the chymotrypsin serine protease family. Core main residues of plasma kallikrein superimpose well on the corresponding residues in bovine trypsin-benzamidine (PDB code 1C5P), although major differences in main chain occur in loop regions (Fig. 3, a and b).

The plasma kallikrein protease shares lower sequence identity with the tissue kallikreins (33% on average) than it does with trypsin (38%) or with hepsin (39%), but all of these proteases adopt similar three-dimensional structures. The superposition of the main chain of plasma kallikrein protease onto that of the tissue kallikrein, hK6 (PDB code 1LO6)) (50) is shown in Fig. 3c. Of the available crystal structures of tissue kallikrein (hK1, Protein Data Bank code 1SPJ (48); hK6, Protein Data Bank code 1LO6 (50); and pro-hk6, Protein Data Bank code 1GVL (51)), hK6 exhibits an S1 site that is more similar to that of plasma kallikrein than does hK1. The S1 site of hK1 contains a serine at position 226 whose side chain is hydrogen-bonded to one of the carboxylate oxygens of Asp189. The presence of serine at 226 enlarges the S1 site of hK1 compared with that of plasma kallikrein, as the latter has a glycine at this position.

The structures of the Endo H- and mutagenically deglycosylated forms of the protease domain of human plasma kallikrein are very similar to one another (Fig. 4a). The protease domain has two six-stranded {beta}-barrels forming the substrate-binding sites, (His57 and the catalytic triad, Asp102, and Ser195) is in a typical arrangement (Fig. 2). Benzamidine is bound at the S1 site, making hydrogen-bonded salt bridges with Asp189(Fig. 5), as in other trypsin-like serine proteases (52). The three N-linked glycosylation sites, Asn21, Asn72, and Asn113, occur within loops on the surface of the molecule. In agreement with the observation that the Endo H-deglycosylated and mutagenically degly-cosylated forms of the plasma kallikrein protease domain have similar proteolytic activities (TABLE FIVE), their core structures superimpose well on one another, with an r.m.s. deviation of 0.17 Å for the 61 C-{alpha} atoms used for superposition (Fig. 4a). However, there are seven loops whose structures differ significantly between Endo H- and mutagenically deglycosylated kallikrein (Fig. 4, a and b). These differences occur in regions of crystal packing constraints. Despite the looser packing of mutagenically deglycosylated kallikrein (42% solvent) compared with that of the Endo H-deglycosylated counterpart (28% solvent), there are nevertheless about as many regions (six) influenced by packing in the latter as in the former (five) (Fig. 4b). None of the major differences between the structures of Endo H-deglycosylated and mutagenically deglycosylated plasma kallikrein occur at the glycosylation sites (Fig. 4, a and b), suggesting that these replacements cause minor structural changes.



View larger version (45K):
[in this window]
[in a new window]
  		FIGURE 2.
Secondary structure of plasma kallikrein. {beta}-Sheet, {alpha}-helix, and the remaining structural elements are yellow, red, and cyan, respectively. Bound benzamidine, the active site residues (His57, Asp102, Ser195), and the glycosylation sites (Asn21, Asn72, and Asn113) are shown.

 
Inhibition of Plasma Kallikrein by Short Hydrogen Bond-mediated Serine Protease Inhibitors—Previously, we developed a large suite of small molecule inhibitors (2-(2-phenol)-benzimidazoles and 2-(2-phenol)-indoles) as serine protease inhibitors (61-67). Inhibition by such scaffolds is typically mediated by short hydrogen bonds at the active site, often via a multicentered short hydrogen bond array involving the active site Ser195, a water in the oxyanion hole, the phenolate of the inhibitor, and the benzimidazole (or indole) nitrogen of the inhibitor (see diagram in TABLE SIX). The Ki values for a set of such inhibitors toward plasma kallikrein and other proteases are provided in TABLE SIX, part a.


View this table:
[in this window]
[in a new window]
  		TABLE SIX
Inhibition constants of inhibitors for full-length plasma kallikrein and for other proteases

 
Plasma kallikrein stands out from the other trypsin-like serine proteases in the degree of inhibition by these simple short hydrogen bond-mediated inhibitors; the inhibitors on average are more potent for plasma kallikrein than for the other proteases. For example, the simplest scaffold, CG-02, is more potent for plasma kallikrein (Ki = 0.020 µM) than for any of the remaining proteases (Ki = 0.041-1.7 µM). Similarly, the benzimidazole isostere, CG-01, is more potent for plasma kallikrein (Ki = 0.22 µM) than for the other proteases (Ki = 0.55-24 µM). The average increase in potency toward plasma kallikrein over the other proteases is significant for both CG-01 (29-fold) and CG-02 (24-fold). Although these inhibitors were synthesized in previous programs without the help of the plasma kallikrein structure, the newly solved crystal structures presented here indicate many of the structural features in the substrate-binding sites that are potentially useful in further improving this class of inhibitors or in the design of inhibitors with different scaffolds.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Expression, Purification, and Activity of Recombinant Forms of Plasma Kallikrein Protease Domain—The first step in determining the structure of plasma kallikrein was to establish an abundant source of crystallizable protein. The catalytic subunit of plasma kallikrein (the light chain), isolated through reduction and alkylation of the inter chain disulfide bond, retains full enzymatic activity (10). However, because of the reduction and alkylation of disulfides and because of heterogeneity in glycosylation (involving the 33- and 36-kDa forms), the purified light chain is not an ideal source for crystallization studies. Thus, we set out to establish expression and purification systems to obtain sufficient amounts of recombinant plasma kallikrein protease domain for structural studies. In the first system (TABLE ONE), the protease domain was expressed as a zymogen in P. pastoris and converted to an active protease by trypsin treatment. In the second system (TABLE TWO), all three N-glycosylation sites, Asn377, Asn434, and Asn475, were mutagenized to glutamates, and the mutant protease domain was expressed in the baculovirus/Sf9 system as an unglycosylated zymogen, subsequently activated by trypsin treatment. In both systems, Cys364 and Cys484 were also changed to serines to prevent disulfide formation between the protease and nonprotease domains. Each form of plasma kallikrein has enzymatic activity similar to that of native plasma kallikrein when assayed with high molecular weight kininogen,4 with insulin B chain (TABLE THREE) or with synthetic peptide substrates (TABLE FIVE), in agreement with the previous observation that the light chain has activity similar to that of the full-length protein (10). The purified proteins are readily crystallizable, providing good systems for plasma kallikrein structural studies.

N-Linked oligosaccharides affect the activity of some serine proteases (68) but not others (50, 56). In the protease subunit of plasma kallikrein, all three N-glycosylation sites are on the surface of the protein, removed from substrate-binding pockets (Fig. 2), and thus glycosylation should not affect the enzymatic activity. The similarity in enzymatic activities among full-length plasma kallikrein and the two recombinant forms of the protease domain support this expectation. The N-linked glycosylation of plasma kallikrein may facilitate its biosynthesis and/or stability; the detailed function(s) of glycosylation remain to be elucidated.



View larger version (47K):
[in this window]
[in a new window]
  		FIGURE 3.
a, superposition of main chains of trypsin (orange) and of enzymatically deglycosylated plasma kallikrein protease domain (blue) based on 61 core residues (16-21, 28-33, 40-46, 51-57, 101-108, 139-143, 155-158, 189-197, 212-217, and 227-229) (r.m.s. = 0.33 Å). The active site triad (His57, Asp102, and Ser195) is shown along with Asp189 at the base of the S1 site. b, sequence alignment of trypsin and plasma kallikrein based on structural superposition involving the enzymatically deglycosylated form (pkal 1). For not italicized sequence segments, the differences between corresponding C-{alpha} coordinates of the superimposed trypsin and plasma kallikrein structures are less than 1.0 Å. Compared sequences

 



View larger version (44K):
[in this window]
[in a new window]
  		FIGURE 4.
a, superposition of main chains of mutagenically deglycosylated plasma kallikrein protease domain (cyan) onto the enzymatically deglycosylated form (orange) based on 61 core residues (16-21, 28-33, 40-46, 51-57, 101-108, 139-143, 155-158, 189-197, 212-217, and 227-229) (r.m.s. = 0. 17 Å). The active site triad (His57, Asp102, and Ser195) is shown along with Asp189 at the base of the S1 site. The Asn -> Glu glycosylation site mutations are labeled in blue. Positions where the structures of the two forms of the plasma kallikrein protease domain differ significantly from one another are labeled in red. b, summary of the loop differences between the two recombinant plasma kallikrein forms.

 
Structural Features of Substrate-binding Site S1 and Implications in Structure-based Drug Design—At the bottom of the S1 site of plasma kallikrein lies negatively charged Asp189, which forms hydrogen-bonded salt bridges with positively charged Arg or Lys residues that are the P1 residues of protein or peptide substrates for this protease. This interaction with Asp189 constitutes the structural basis for the selectivity of plasma kallikrein for Arg or Lys at the S1 site (Fig. 5) (52). The role of Asp189 in recognition at the S1 site of plasma kallikrein is apparent in the structure of the benzamidine complex, in which Asp189 makes hydrogen-bonded salt bridges with the amidino group (Fig. 5), as in other trypsin-like serine proteases (52). Other hydrogen bonds to benzamidine involve the carbonyl group of Gly219 and a water molecule. Although the structures of the S1 sites of trypsin-like serine proteases resemble one another, one important difference often occurs at residue 190 (52, 62, 66). Trypsin-like serine proteases can be classified into three subgroups: Ala190, Ser190, or Thr190 proteases. The Ser190 and Ala190 group proteases each constitute about 32% of the more than 100 trypsin-like serine proteases in the human genome (67). Plasma kallikrein is an Ala190 protease, with no hydrogen-bonding capacity for the Ala190 side chain. In Ala190 protease-amidine complexes (such as plasma kallikrein-benzamidine; see Fig. 5), the sole hydrogen-bonding partner to one of the amidine protons is a water molecule, and in the corresponding Ser190 protease complexes (such as trypsin-benzamidine), there are two hydrogen-bonding acceptors for the amidine proton, O-{gamma}Ser190 and the water. Thus, inhibitors designed to displace the water are often highly selective for Ser190 proteases over Ala190 counterparts (such as plasma kallikrein) because of the resulting hydrogen-bonding deficit at the S1 site (62, 66, 67). Thus, viewed as an anti-target, plasma kallikrein offers the S1 site as a potential locus for engineering selectivity against this protease, if the target of interest is a Ser190 protease, such as urokinase. For example, CG-06 (Y = Cl, X = N, and R = H; see TABLE SIX, part c), whose chloro ortho to the amidine displaces the water from the S1 site, is 90-fold selective for urokinase (Ki = 0.009 µM) over plasma kallikrein (Ki = 0.80 µM) (62).



View larger version (28K):
[in this window]
[in a new window]
  		FIGURE 5.
S1 site of plasma kallikrein-benzamidine superimposed on a |Fo| - |Fc| omit map contoured at 2.75 {sigma}.

 



View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 6.
Model of CG-05 bound to plasma kallikrein, showing expected interaction of the diacid group of the inhibitor with Lys192 of the enzyme, along with the multicentered short hydrogen bond array observed at the active site of the trypsin complex. Short hydrogen bonds are shown in cyan.

 



View larger version (96K):
[in this window]
[in a new window]
  		FIGURE 7.
a, model of tetrahedral intermediate complex of plasma kallikrein bound by P1, P2, and P3 portion of substrate with Arg at S1 site and Met at S2 site. b, model of tetrahedral intermediate complex of trypsin bound by P1, P2, and P3 portion of a substrate with Arg at S1 site and Pro at S2 site.

 
Another important structural feature bordering the S1 site of plasma kallikrein is Lys192. Sequence analysis of the trypsin-like proteases of the human genome shows that the majority of proteases contain glutamine at residue 192 (60%), followed by lysine (14%) and glutamate (5.4%). As proteases with Lys192 are a minority, a potential strategy for the development of inhibitors selective for plasma kallikrein could rely on specific interactions with Lys192. For example, in the serine protease inhibitor series that make short hydrogen bonds at the active site (64-67) (TABLE SIX, part a), substituents para to the phenol interact with residue 192. In a model of the plasma kallikrein-bound structure of CG-05, one of the oxygens of the succinate makes a hydrogen-bonded salt bridge with Lys192 (Fig. 6). The role in potency and selectivity of residue 192 is revealed in the inhibition data in TABLE SIX, parts a and b, in which both residue 192 of the protease as well as the interacting group of the inhibitor are varied.

The data in TABLE SIX, part a, show that addition of a succinate para to the phenol of the inhibitor results in a significant increase in potency toward serine proteases with a lysine at position 192 (plasma kallikrein and fVIIa). In the case of plasma kallikrein, addition of a succinate para to the phenol incurs a 27-fold increase in potency for the benzimidazole (CG-04, Ki = 0.0081 µM) over that of the parent (CG-01, Ki = 0.215 µM) or a 29-fold increase for the indole (CG-05, Ki = 0.0007 µM) over that of the parent (CG-02, Ki = 0.020 µM). Similarly, for fVIIa the succinate results in increases in potency of 12- and 5.6-fold. The increase in potency toward plasma kallikrein and fVIIa imparted by succinate reflects a favorable interaction of this group with Lys192. Sizeable increases are also observed with acetate para to the phenol (data not shown).

The effect of incorporating a succinate into the inhibitor scaffold is generally minor for the Gln192 proteases (TABLE SIX, part a). Changes in potency resulting from the succinate range from 0.26- to 4.3-fold, with an average change of 1.4 ± 0.8 for eight Gln192 proteases (not all of which are shown in TABLE SIX). For the Glu192 protease, thrombin, the succinate group decreases potency by as much as 8.6-fold (for CG-04, Ki = 210 µM compared with CG-01, Ki = 24 µM), probably reflecting an unfavorable electrostatic repulsion between this group and the Glu192 side chain. Thus, the presence of Lys192 of plasma kallikrein provides a means for introducing selectivity toward plasma kallikrein and against other non-Lys192 serine proteases (TABLE SIX, part b). Conversely, it might be expected that incorporation of an appropriate basic group ortho to the phenol would allow development of selectivity against plasma kallikrein because of the resulting unfavorable electrostatic repulsion. Note that features other than the Lys192 side chain must also play important roles in selectivity of the compounds in TABLE SIX. For example, the succinate-containing CG-05 shows substantial selectivity, 84-fold, toward plasma kallikrein against fVIIa, even though both are Lys192 proteases.



View larger version (42K):
[in this window]
[in a new window]
  		FIGURE 8.
Model of CG-01 bound to plasma kallikrein, showing the expected location of the terminal phenyl group of the inhibitor in the S1' site.

 
Implications for Enzymatic Specificity of Structural Features of Substrate-binding Site S2—The S2 site of plasma kallikrein is bordered by the His57 and Trp215 side chains and by the main chain of residues Ser97 to Gly99 (Fig. 7a). The physiological substrate of plasma kallikrein (kininogen) is cleaved in two positions, in which either the Phe or Met side chain occupies the S2 site (69). This preference for Phe or Met at S2 is in contrast to that of thrombin or of trypsin, which prefer a Pro at S2 (70). Substrates with Pro at the S2 site are suboptimal substrates for plasma kallikrein (71).

In order to investigate the structural basis for the S2 site selectivity of plasma kallikrein versus that of trypsin, tri-peptide portions (P1, P2, and P3) of substrates were docked into these proteases, and the resulting models of the catalytic complexes were compared. The key differences in the S2 site of trypsin (or of thrombin) versus that of plasma kallikrein are the positions of loop residues 94-99 and the identity of residue 99, Leu in trypsin (Fig. 7b) versus Gly in plasma kallikrein (Fig. 7a). The lack of a side chain at position 99 creates space at the S2 site of plasma kallikrein (Fig. 7a) that is absent in trypsin (Fig. 7b) or in thrombin. In addition, loop residues 95-99, defining a wall of the S2 site of plasma kallikrein, are shifted toward the substrate in plasma kallikrein compared with their position in trypsin, thus creating a cavity for accommodation of a Phe or Met side chain at the S2 site of plasma kallikrein (Fig. 7a). This cavity is absent in trypsin (Fig. 7b). Thus, the structure of plasma kallikrein affords a convincing explanation for the S2 preference of this protease.

Structural Features of Substrate-binding Site S1' to Harness for Drug Design—The S1' site of trypsin, formed from residues 35-42 and 57-60, is relatively shallow. The location of the S1' site on the superimposed C-{alpha} structures of trypsin and plasma kallikrein is shown in Fig. 3a. In many important protease targets like tryptase, thrombin, hepsin, fVIIa, urokinase-type plasminogen activator, tissue-type plasminogen activator, and plasmin, additional loop residues (the 60s loop) inserted after residue 59 in trypsin, provide another wall to the S1' site. This additional boundary creates a well defined S1'-binding pocket with distinct structural features that can be exploited in the development of selective inhibitors. Plasma kallikrein is one such protease that contains a "60's loop," and an important difference between plasma kallikrein and trypsin is the insertion of residues Gly60A, Leu60B, and Pro60C at the S1' site. This difference is also apparent in the comparison of plasma kallikrein versus tissue kallikrein, hK6 (Fig. 3c). The location of this loop is shown in Fig. 3, a and c, in which the first residue, Gly60A, is labeled. The S1' site region of plasma kallikrein is further elaborated from that of trypsin or of hK6 because of insertion into plasma kallikrein of additional loop residues after residue 35.

In our other serine protease inhibitor development programs, the S1' site has been used extensively for engineering selectivity toward particular protease targets into inhibitors that make short hydrogen bonds at the active site (61-66). In plasma kallikrein, the side chains of residues Leu41, Leu60B, and Trp60D form a binding pocket that is readily accessible from the para position of the distal phenyl group of a typical, simple short hydrogen bond-mediated scaffold, like CG-01 (Fig. 8). In addition, both the main chain and side chain of Asp60 are accessible for interactions from substituents on the meta position of the distal phenyl of bound CG-01. Fortunately, this region is free of crystal packing interactions in the Endo H-deglycosylated kallikrein, and should therefore reflect the true solution structure. The unique S1' site of plasma kallikrein is thus an attractive locus for engineering selectivity into plasma kallikrein inhibitors by using previously developed, active site-directed, and short hydrogen bond-mediated inhibitor scaffolds.


   FOOTNOTES
 
The atomic coordinates and structure factors (codes 2ANW and 2ANY) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 Present address: Locus Pharmaceuticals, Inc., 4 Valley Square, 512 E. Township Line Rd., Blue Bell, PA 19422. Back

2 To whom correspondence should be addressed: Dept. of Structural Chemistry, Celera Genomics, 180 Kimball Way, South San Francisco, CA 94080. Tel.: 650-866-6270; Fax: 650-866-6654; E-mail: brad.katz{at}celera.com.

3 The abbreviations used are: PK, prekallikrein; Endo H, endoglycosidase H; HK, high molecular weight kininogen; PDB, protein data bank; r.m.s., root mean square; MES, 2-(N-morpholino)ethanesulfonic acid; fXI, factor XI; fVIIa, factor VIIa; fIXa, factor IXa; fXa, factor Xa; hK, human tissue kallikreins. Back

4 J. Tang, and E. Springman, unpublished results. Back


   ACKNOWLEDGMENTS
 
We are grateful to the staff at the Advanced Light Source, Berkeley, CA, for help with crystallographic data collection. The Advanced Light Source is supported by the Director, Office of Science, Office of Basic Energy Sciences, Materials Sciences Division, of the United States Department of Energy under Contract. DE-AC03-76SF00098 at Lawrence Berkeley National Laboratory.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Kraut, H., Frey, E. K., and Werle, E. (1930) Hoppe-Seyler's Z. Physiol. Chem. 189, 97-106
  2. Werle, E. (1934) Biochem. Z. 269, 415-434
  3. Clements, J. A., Hooper, J. D., and Odorico, D. M. (2004) in Handbook of Proteolytic Enzyme (Barret, A. J., Rawlings, N. D., and Woessner, J. I., eds) 2nd Ed., pp. 1569-1577, Elsevier, London
  4. Yousef, G. M., and Diamandis, E. P. (2001) Endocr. Rev. 22, 184-204[Abstract/Free Full Text]
  5. Hathaway, W. E., Belhasen, L. P., and Hathaway, H. S. (1965) Blood 26, 521-532[Abstract/Free Full Text]
  6. Wuepper, K. D., and Cochrane, C. G. (1972) J. Exp. Med. 135, 1-20[Abstract]
  7. Chung, D. W., Fujikawa, K., McMullen, B. A., and Davie, E. W. (1986) Biochemistry 25, 2410-2417[CrossRef][Medline] [Order article via Infotrieve]
  8. Colman, R. W. (2004) in Handbook of Proteolytic Enzymes (Barrett, A. J., Rawlings, N. D., and Woessner, J. F., eds) 2nd Ed., pp. 1644-1651, Elsevier, London
  9. McMullen, B. A., Fujikawa, K., and Davie, E. W. (1981) Biochemistry 26, 2050-2056
  10. van der Graaf, F., Tans, G., Bouma, B. N., and Griffin, J. H. (1982) J. Biol. Chem. 257, 14300-14305[Abstract/Free Full Text]
  11. Borgono, C. A., and Diamandis, E. P. (2004) Nat. Rev. Cancer 4, 876-890[CrossRef][Medline] [Order article via Infotrieve]
  12. Kitchens, C. S. (2002) Arch. Pathol. Lab. Med. 126, 1382-1386[Medline] [Order article via Infotrieve]
  13. Ciechanowicz, A., Bader, M., Wagner, J., and Ganten, D. (1993) Biochem. Biophys. Res. Commun. 197, 1370-1376[CrossRef][Medline] [Order article via Infotrieve]
  14. Mandle, R. J., Colman, R. W., and Kaplan, A. P. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 4179-4183[Abstract/Free Full Text]
  15. Hasan, A. A., Zisman, T., and Schmaier, A. H. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 3615-3620[Abstract/Free Full Text]
  16. Shariat-Madar, Z., Mahdi, F., and Schmaier, A. H. (1999) J. Biol. Chem. 274, 7137-7145[Abstract/Free Full Text]
  17. Colman, R. W., Pixley, R. A., Najamunnisa, S., Yan, W., Wang, J., Mazar, A., and McCrae, K. R. (1997) J. Clin. Investig. 100, 1481-1487[Medline] [Order article via Infotrieve]
  18. Herwald, H., Dedio, J., Kellner, R., Loos, M., and Muller-Esterl, W. (1996) J. Biol. Chem. 271, 13040-13047[Abstract/Free Full Text]
  19. Joseph. K., Ghebrehiwet, B., Peerschke, E. I., Reid, K. B., and Kaplan, A. P. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 8552-8557[Abstract/Free Full Text]
  20. Mahdi, F., Shariat-Madar, Z., Todd, R. F., III, Figueroa, C. D., and Schmaier, A. H. (2001) Blood 97, 2342-2350[Abstract/Free Full Text]
  21. Schmaier, A. H. (2002) J. Clin. Investig. 109, 1007-1009[CrossRef][Medline] [Order article via Infotrieve]
  22. Shariat-Madar, Z., Mahdi, F., and Schmaier, A. H. (2002) J. Biol. Chem. 277, 17962-17969[Abstract/Free Full Text]
  23. Shariat-Madar, Z., and Schmaier, A. H. (2004) J. Endotoxin. Res. 10, 3-13[CrossRef]
  24. Colman, R. W. (1969) Biochem. Biophys. Res. Commun. 35, 273-279[CrossRef][Medline] [Order article via Infotrieve]
  25. Ichinose, A., Fujikawa, K., and Suyama, T. (1986) J. Biol. Chem. 261, 3486-3489[Abstract/Free Full Text]
  26. Hauert, J., Nicoloso, G., Schleuning, W. D., Bachmann, F., and Schapira, M. (1989) Blood 73, 994-999[Abstract/Free Full Text]
  27. Schmaier, A. H. (2000) Curr. Opin. Hematol. 7, 261-265[CrossRef][Medline] [Order article via Infotrieve]
  28. Kaplan, A. P., Kay, A. B., and Austen, K. F. (1972) J. Exp. Med. 135, 81-97[Abstract]
  29. Wachtfogel, Y. T., Kucich, U., James, H. L., Scott, C. F., Schapira, M., Zimmerman, M., Cohen, A. B., and Colman, R. W. (1983) J. Clin. Investig. 72, 1672-1677[Medline] [Order article via Infotrieve]
  30. Ni, Z. J., and Wagman, A. S. (2004) Curr. Opin. Drug Discovery Dev. 7, 446-459[Medline] [Order article via Infotrieve]
  31. Adams, J., and Kauffman, M. (2004) Cancer Investig. 22, 304-311[CrossRef][Medline] [Order article via Infotrieve]
  32. Docherty, A. J., Crabbe, T., O'Connell, J. P., and Groom, C. R. (2003) Biochem. Soc. Symp. 70, 147-161
  33. Maryanoff, B. E. (2004) J. Med. Chem. 47, 769-787[CrossRef][Medline] [Order article via Infotrieve]
  34. Wieland, H. A., Laux, V., Kozian, D., Lorenz, M. (2003) Curr. Opin. Investig. Drugs 4, 264-271[Medline] [Order article via Infotrieve]
  35. Kyrle, P. A., and Eichinger, S. (2002) Vasa 31, 163-166[CrossRef][Medline] [Order article via Infotrieve]
  36. Yu, A., Jacobson, S. H., Bygden, A., and Egberg, N. (2002) Clin. Chem. Lab. Med. 40, 21-26[CrossRef][Medline] [Order article via Infotrieve]
  37. Levine, M. N., Raskob, G., Landefeld, S., and Kearon, C. (2001) Chest 119, Suppl. 1, 108-121
  38. Schechter, I., and Berger, A. (1967) Biochem. Biophys. Res. Commun. 27, 157-162[CrossRef][Medline] [Order article via Infotrieve]
  39. Banner, D. W., D'Arcy, A., Chène, C., Winkler, F. K., Guha, A., Konigsberg, W. H., Nemerson, Y., and Kirchhofer, D. (1996) Nature 380, 41-46[CrossRef][Medline] [Order article via Infotrieve]
  40. Parlow, J. J., Case, B. L., Dice, T. A., Fenton, R. L., Hayes, M. J., Jones, D. E., Neumann, W. L., Wood, R. S., Lachance, R. M., Girard, T. J., Nicholson, N. S., Clare, M., Stegeman, R. A., Stevens, A. M., Stallings, W. C., Kurumbail, R. G., and South, M. S. (2003) J. Med. Chem. 46, 4050-4062[CrossRef][Medline] [Order article via Infotrieve]
  41. Brandstetter, H., Kühne, A., Bode, W., Hubert, R., von der Saal, W., Wirthensohn, K., and Engh, R. A. (1996) J. Biol. Chem. 271, 29988-29992[Abstract/Free Full Text]
  42. Wiley, M. R., Weir, L. C., Briggs, S., Bryan, N. A., Buben, J., Campbell, C., Chirgadze, N. Y., Conrad, R. C., Craft, T. J., Ficorilli, J. V., Franciskovich, J. B., Froelich, L. L., Gifford-Moore, D. S., Goodson, T., Jr., Herron, D. K., Klimkowski, V. J., Kurz, K. D., Kyle, J. A., Masters, J. J., Ratz, A. M., Milot, G., Shuman, R. T., Smith, T., Smith, G. F., Tebbe, A. L., Tinsley, J. M., Towner, R. D., Wilson, A., and Yee, Y. K. (2000) J. Med. Chem. 43, 883-899[CrossRef][Medline] [Order article via Infotrieve]
  43. Maignan, S., and Mikol, V. (2001) Curr. Top. Med. Chem. 1, 161-174[CrossRef][Medline] [Order article via Infotrieve]
  44. Walport, M. J. (2001) N. Engl. J. Med. 344, 1058-1066[Free Full Text]
  45. Colman, R. W. (1999) Immunopharmacology 43, 103-108[CrossRef][Medline] [Order article via Infotrieve]
  46. Han, E. D., MacFarlane, R. C., Mulligan, A. N., Scafidi, J., and Davis, A. E., III (2002) J. Clin. Investig. 109, 1057-1063[CrossRef][Medline] [Order article via Infotrieve]
  47. De La Cadena, R. A., Suffredini, A. F., Page, J. D., Pixley, R. A., Kaufman, N., Parrillo, J. E., and Colman, R. W. (1993) Blood 81, 3313-3317[Abstract/Free Full Text]
  48. Laxmikanthan, G., Blaber, S. I., Bernett, M. J., Scarisbrick, I. A., Juliano, M. A., and Blaber, M. (2005) Proteins 58, 802-814[CrossRef][Medline] [Order article via Infotrieve]
  49. Katz, B. A., Liu, B., Barnes, M., and Springman, E. (1998) Protein Sci. 7, 875-885[Medline] [Order article via Infotrieve]
  50. Bernett, M. J., Blaber, S. I., Scarisbrick, I. A., Dhanarajan, P., Thompson, S. M., and Blaber, M. (2002) J. Biol. Chem. 277, 24562-24570[Abstract/Free Full Text]
  51. Gomis-Ruth, F. X., Bayes, A., Sotiropoulou, G., Pampalakis, G., Tsetsenis, T., Villegas, V., Aviles, F. X., and Coll, M. (2002) J. Biol. Chem. 277, 27273-27281[Abstract/Free Full Text]
  52. Katz, B. A., Mackman, R., Luong, C., Radika, K., Martelli, A., Sprengeler, P. A., Wang, J., Chan, H., and Wong, L. (2000) Chem. Biol. 7, 299-312[CrossRef][Medline] [Order article via Infotrieve]
  53. Kuzmic, P., Sideris, S., Cregar, L. M., Elrod, K. C., Rice, K. D., and Janc, J. W. (2000) Anal. Biochem. 281, 62-67[CrossRef][Medline] [Order article via Infotrieve]
  54. Otwinowski, Z., and Minor, W. (1993) in Proceedings of the CCP4 Study Weekend, January 29-30, 1993, SERC Daresbury Laboratory, Warrington, UK, pp. 56-62
  55. Pandey, P., Babine, R. E., Gorga, J. C., Seidl, K. J., Gelfand, E., Weaver, D. T., Abdel-Meguid, S. S., and Strickler, J. E. (2005) J. Biol. Chem. 280, 4704-4712[Abstract/Free Full Text]
  56. Somoza, J. R., Ho, J. D., Luong, C., Ghate, M., Sprengeler, P. A., Mortara, K., Shrader, W. D., Sperandio, D., Chan, H., McGrath, M. E., and Katz, B. A. (2003) Structure 11, 1-20[Medline] [Order article via Infotrieve]
  57. Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N., and Bourne P. E. (2000) Nucleic Acids Res. 28, 235-242[Abstract/Free Full Text]
  58. Perrakis, A., Harkiolaki, M., Wilson, K. S., and Lamzin, V. S. (2001) Acta Crystallogr. Sect. D Biol. Crystallogr. 57, 1445-1450[CrossRef][Medline] [Order article via Infotrieve]
  59. Brünger, A. T. (1990) X-PLOR Manual, A System for X-ray Crystallography and NMR, Version 3.1, pp. 187-206, Yale University, New Haven, CT
  60. Brünger, A. T. (1992) Nature 355, 472-474[CrossRef][Medline] [Order article via Infotrieve]
  61. Verner, E., Katz, B. A., Spencer, J., Allen, D., Hataye, J., Hruzewicz, W., Hui, H., Kolesnikov, A., Li, Y., Luong, C., Martelli, A., Radika, K., Rai, R., She, M., Shrader, W., Sprengeler, P., Trapp, S., Wang, J., Young, W. B., and Mackman, R. L. (2001) J. Med. Chem. 44, 2753-2771[CrossRef][Medline] [Order article via Infotrieve]
  62. Mackman, R. L., Katz, B. A., Breitenbucher, J. G., Hui, H. C., Verner, E., Luong, Liu, L., and Sprengeler, P. A. (2001) J. Med. Chem. 44, 3856-3871[CrossRef][Medline] [Order article via Infotrieve]
  63. Mackman, R. L., Hui, H. C., Breitenbucher, J. G., Katz, B. A., Luong, C., Martelli, A., McGee, D., Radika, K., Sendzik, M., Spencer, J. R., Sprengeler, P. A., Tario, J., Verner, E., and Wang, J. (2002) Bioorg. Med. Chem. Lett. 12, 2019-2022[CrossRef][Medline] [Order article via Infotrieve]
  64. Katz, B. A., Spencer, J. R., Elrod, K., Luong, C., Mackman, R. L., Rice, M., Sprengeler, P. A., Allen, D., and Janc, J. (2002) J. Am. Chem. Soc. 124, 11657-11668[CrossRef][Medline] [Order article via Infotrieve]
  65. Katz, B. A., Elrod, K., Verner, E., Mackman, R. L., Luong, C., Shrader, W. D., Sendzik, M., Spencer, J. R., Sprengeler, P. A., Kolesnikov, A., Tai, V. W., Hui, H. C., Breitenbucher, J. G., Allen, D., and Janc, J. W. (2003) J. Mol. Biol. 329, 93-120[CrossRef][Medline] [Order article via Infotrieve]
  66. Katz, B. A., Sprengeler, P. A., Luong, C., Verner, E., Elrod, K., Kirtley, M., Janc, J., Spencer, J., Breitenbucher, J. G., Hui, H., McGee, D., Allen, D., Martelli, A., and Mackman, R. L. (2001) Chem. Biol. 8, 1107-1121[CrossRef][Medline] [Order article via Infotrieve]
  67. Katz, B. A., Luong, C., Ho, J. D., Somoza, J. R., Gjerstad, E., Tang, J., Williams, S. R., Verner, E., Mackman, R. L., Young, W., Sprengeler, P. A., Chan, H., Mortara, K., Janc, J. W., and McGrath, M. E. (2004) J. Mol. Biol. 344, 527-547[CrossRef][Medline] [Order article via Infotrieve]
  68. Oka, T., Hakoshima, T., Itakura, M., Yamamori, S., Takahashi, M., Hashimoto, Y., Shiosaka, S., and Kato, K. (2002) J. Biol. Chem. 277, 14724-14730[Abstract/Free Full Text]
  69. Kitamura, N., Takagaki, Y., Furuto, S., Tanaka, T., Nawa, H., and Nakanishi, S. (1983) Nature 305, 545-549[CrossRef][Medline] [Order article via Infotrieve]
  70. Harris, J. L., Backes, B. J., Leonetti, F., Mahrus, S., Ellman, J. A., and Craik, C. S. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 7754-7759[Abstract/Free Full Text]
  71. McRae, B. J., Kurachi, K., Heimark, R. L., Fujikawa, K., Davie, E. W., and Powers, J. C. (1981) Biochemistry 20, 7196-7206[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
K. W. Rickert, P. Kelley, N. J. Byrne, R. E. Diehl, D. L. Hall, A. M. Montalvo, J. C. Reid, J. M. Shipman, B. W. Thomas, S. K. Munshi, et al.
Structure of Human Prostasin, a Target for the Regulation of Hypertension
J. Biol. Chem., December 12, 2008; 283(50): 34864 - 34872.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
L. Beinrohr, V. Harmat, J. Dobo, Z. Lorincz, P. Gal, and P. Zavodszky
C1 Inhibitor Serpin Domain Structure Reveals the Likely Mechanism of Heparin Potentiation and Conformational Disease
J. Biol. Chem., July 20, 2007; 282(29): 21100 - 21109.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
S. P. Bajaj, A. E. Schmidt, S. Agah, M. S. Bajaj, and K. Padmanabhan
High Resolution Structures of p-Aminobenzamidine- and Benzamidine-VIIa/Soluble Tissue Factor: UNPREDICTED CONFORMATION OF THE 192-193 PEPTIDE BOND AND MAPPING OF Ca2+, Mg2+, Na+, AND Zn2+ SITES IN FACTOR VIIa
J. Biol. Chem., August 25, 2006; 281(34): 24873 - 24888.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
280/49/41077    most recent
M506766200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tang, J.
Right arrow Articles by Katz, B. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tang, J.
Right arrow Articles by Katz, B. A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2005 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement