Serine Protease and Serine Protease Inhibitors
Michael J. Page and Enrico Di Cera, Department of Biochemistry and Molecular Biophysics, Washington University School of Medicine, St. Louis, Missouri
Serine proteases are among the largest group of proteolytic enzymes in the human genome that play vital roles in health and disease. Regulation of their activity in vivo is mediated by a diverse group of serine protease inhibitors. An overview of the interplay between serine proteases and their inhibitors is provided. In addition, approaches to characterize this relationship are discussed with subsequent emphasis on how such measures apply to pathologies that result from defect of serine proteases or their inhibitors.
Cellular life is brokered on the activity of proteins, and, in turn, control over their concentration and state is vital. Enzymes that hydrolyze peptide bonds, the proteases, are therefore critical ingredients of the genome. Proteases may acts as nonspecific agents of digestion or high-selectivity mediators of posttranslational modification. Proteases are a diverse group of enzymes. Over 180 phylogenetically distinct families of proteases have been identified by the MEROPS protease classification system on the basis of protein fold and additionally separated through phylogenetic relationship (1). Of these families, the largest contingent comprises serine proteases, which are named based on their application of the hydroxyl group of a serine side chain as catalytic nucleophile. Regulation of protease function in vivo is mediated by a minimum of 90 families of protease inhibitors. Together, these two groups control biological functions both inside and outside of the cell. Genetic defects in either protease or inhibitor, therefore, can result in significant pathology with potentially multiple biological pathways impacted.
Serine Proteases and their Inhibitors
More than a third of all known proteolytic enzymes are serine proteases (2). The family name stems from the nucleophilic serine residue within the active site, which attacks the carbonyl moiety of the substrate peptide bond to form an acyl-enzyme intermediate. Nucleophilicity of the catalytic serine is commonly dependent on a catalytic triad of aspartic acid, histidine, and serine—commonly referred to as a charge relay system (3). First observed by Blow over 30 years ago in the structure of chymotrypsin (4), the same combination has been found in four other three-dimensional protein folds that catalyze hydrolysis of peptide bonds. Examples of these folds are observed in trypsin, subtilisin, prolyl oligopeptidase, and ClpP protease. Many other enzyme families use the same catalytic triad, such as asparaginases, esterases, acylases, and β-lactamases (5). Mutagenesis of the aspartic acid to alanine impacts peptide bond hydrolysis to a greater extent than ester hydrolysis, which indicates that a complete catalytic triad is required for the hydrolysis of the stronger peptide bond. It should be noted that several serine protease families use a dyad mechanism in which lysine or histidine is paired with the catalytic serine. Yet, other serine proteases present novel catalytic triads, such as a pair of histidines combined with the nucleophilic serine. In nearly all reported cases, the active site serine can be rendered inactive by generic inhibitors such as diisopropylfluorophosphate and phenylmethanesulfonyl fluoride. We will focus on the most abundant serine protease and serine protease inhibitor families in the human genome. The reader is also referred throughout the text to other recent and more expansive reviews and to the MEROPS database for a more detailed description of the impressive diversity of serine protease and inhibitor structure, function, and activity.
A typical genome contains 2-4% of genes that encode for proteolytic enzymes. The entire complement of peptidases within a genome is referred to as the degradome (6). Of these proteases, a select subset of peptidases underwent considerable gene duplication and divergence. The trypsin-like serine peptidases (Clan PA Family S1 Subfamily A in MEROPS nomenclature) are the largest group of homologous peptidases in the human genome responsible for various critical biological processes. Similar degradome composition is observed in all vertebrates, which indicates that expansion of the S1A peptidase family occurred before emergence of the lineage roughly 450 million years ago. Of 699 peptidases in humans, 178 are serine peptidases, and 138 of them belong to the S1 peptidase family. The chymotrypsin-like fold of the S1 peptidase family presents an ideal catalytic platform that enables high turnover, substrate selectivity, and various modes of regulation in a package readily combined with additional protein domains (7).
Clan PA: SI peptidases—the trypsins
Chymotrypsin-fold proteases are the largest family of peptidases known. Pioneering studies used digestive enzymes such as trypsin and chymotrypsin, which cleave polypeptide chains on the C-terminal side of a positively charged side chain (arginine or lysine) or large hydrophobic residue (phenylalanine, tryptophan, or tyrosine), respectively. In Schechter & Berger nomenclature, the peptide bond hydrolyzed lies between the P1 and P1’ residues of the substrate and is numbered increasingly toward both the N- and C-terminal directions. Interaction occurs in the similarly numbered subsite (S) in the protease. For example, the S2 pocket binds the P2 residue, which is the second amino acid on the N-terminal side of the scissile bond. Determination of other polypeptide sequences of several serine proteases revealed a large family of enzymes. Trypsins and chymotrypsins belong to the S1A family of clan PA, whereas the S1B family comprises various bacterial proteases and the HtrA subgroup of proteases responsible for intracellular protein turnover (8). Both subfamilies share the two β-barrel architecture. Two Greek-key P-barrels comprise the chymotrypsin fold and are homologous in a nontraditional manner (Fig. 1a). β-strand topology of the fold reveals Greek-key architecture in both barrels, yet this topology originates from sequences that run in the opposite direction. Hence, the two halves of the structure are mirror images in protein fold space. Both barrels are functionally partitioned with one end involved in catalysis and a second in regulation. The active site lies in the cleft between them.
The conventional catalytic triad in S1 family peptidases mediates peptide bond hydrolysis (Fig. 1b). Hydrogen bonding between residue Asp-102 and His-57 (chymotrypsin numbering is used) facilitates abstraction of the proton from Ser-195 and creates a potent nucleophile. Stabilization of the catalytic triad is mediated through a network of additional hydrogen bonds provided by several highly conserved amino acid residues that surround the triad, particularly Thr-54, Ala-56, and Ser-214, and buttressed more by a disulphide bond between residues 42 and 58. The reaction proceeds via pair tetrahedral intermediates. In the first step, nucleophilic attack by the serine yields an oxyanion intermediate stabilized by the backbone amides of Gly-193 and Ser-195. Collapse of the tetrahedral intermediate generates an acyl-enzyme intermediate, and stabilization of the newly created N-terminus is mediated by His-57. Hartley and Kilbey provided evidence for the acyl-enzyme intermediate in 1954 (9). In these initial experiments, a pre-steady state burst of product correctly identified that a bond to a hydroxyl moiety within chymotrypsin was involved in the reaction mechanism. The second half of the mechanism is a reversal of the first step, in which a water molecule displaces the free polypeptide fragment and attacks the acyle-enzyme intermediate. Again, the oxyanion hole stabilizes the second tetrahedral intermediate, and collapse of this intermediate liberates a new C-terminus in the substrate. Central to the regulation of peptidase activity is zymogen activation.
Figure 1. (a) Two β-barrel architecture of the S1A peptidase family (PDB 1OS8). Following activation, the zymogen arm stabilizes the active site cleft that lies in between the two barrels. An additional α-helix is at the C-terminus. (b) The canonical catalytic triad is generated by the spatial arrangement of Asp-102, His-57, and Ser-195 positioned to facilitate hydrogen bond formation and abstraction of the proton of hydroxyl moiety.
Activation of most chymotrypsin-like serine peptidases requires proteolytic processing of an inactive zymogen precursor protein. Cleavage of the proprotein precursor occurs at an identical position in all known members of the family: between residues 15 and 16. The nascent N-terminus induces conformational change in the enzyme through formation of an intramolecular electrostatic interaction with Asp-194 that stabilizes both oxyanion hole and substrate-binding site (10). Zymogen activation provides a powerful mechanism of regulation that endows temporal control over protease activity, an ability to escape premature enzyme inhibition, and places these enzymes within the context of chains of proteolytic events. Many proteases of the coagulation and immune pathways are regulated more through allosteric mechanisms that involve monovalent cations (Na+), divalent cations (Ca2+), glycosaminoglycans, and protein cofactors (11). These properties derive from the structure of the chymotrypsin fold and combine to produce proteolytic networks responsible for key biological processes responsible for human health.
Several vital processes rely on clan PA peptidases. Chief among them are blood coagulation and the immune response, which involve cascades of sequential zymogen activation. In both systems, the chymotrypsin-fold peptidase domain is combined with one more associated protein domains, including apple, CUB, EGF, fibronectin, kringle, sushi, and von Willebrand factor domains. These protein domains are on the N-terminus as an extension of the propeptide segment of the peptidase. Such a trend of N-terminal-associated domains in the S1A peptidase family is common across all forms of life. The domain architecture pairs well with the zymogen activation mechanism, which liberates the proper N-terminus to enable catalytic activity. Often, the associated protein domains remain attached to the peptidase domain through a covalent disulphide bond on the opposing surface of the protease active site. Many associated domains are entirely encoded by a single exon in their peptidase gene and suggest an important role for exon shuffling during molecular evolution of clan PA.
S1 peptidases in the human genome are, for the most part, phylogenetically grouped into six functional categories: digestion, coagulation and immunity, tryptase, matriptase, kallikrein, and granzymes. Various enzymes are involved in the breakdown of proteins in the digestive system. The trypsins, chymotrypsins, and elastases are endopeptidases that breakdown polypeptides into shorter chains. More digestion is mediated by various exopeptidases (12). In particular, carboxypeptidases A and B from the M14 family of zinc-dependent metalloproteases shorten the nascent peptides through complementary selectivity toward basic or aromatic residues. Inappropriate release of trypsin from the digestive system signals proinflammatory responses typically mediated by tryptase-like S1 peptidases. Tryptases are major components in secretory granules of mast cells that are unique among clan PA peptidases because of their homotetrameric quaternary structure (13). Like trypsin, tryptases mediate proinflammatory signaling through protease-activated receptors 2, yet definition of other substrates in health and disease states remain elusive. Matriptases are membrane-bound S1 peptidases that bear primary substrate selectivity similar to trypsin (14). Again, physiological substrates of this subfamily of peptidases are largely unknown, yet high gene expression levels for matriptases are associated with various cancer types. Similar association with cancer has led to great interest in the large family of kallikreins (15), a family commonly known for its role in regulation of blood pressure through the kinin system (16). Granzymes are mediators of directed apoptosis by natural killer cells and cytotoxic T cells that play key roles in the defense against viral infection (17). Notably, unique among clan PA is the primary selectivity of granzymes toward acidic residues in the P1 position of the substrates. Of the wide diversity of proteases in clan PA family S1, the mediators of immunity and blood coagulation have been particularly well studied.
Clan SC: peptidase diversity in the α/β fold
Clan SC peptidases are a/p hydrolase-fold enzymes that consist of parallel β-strands surrounded by α-helices. The a/p hydrolase-fold provides a versatile catalytic platform that, in addition to achieving proteolytic activity, can either act as an esterase, lipase, dehalogenase, haloperoxidase, lyase, or epoxide hydrolase (18). Six phylogenetically distinct families of clan SC are known, and only four of them have known structure. Catalytic amenability of the α/β hydrolase-fold may underlie why clan SC peptidases are the second largest family of serine peptidases in the human genome. Other mechanistic classes need not use the catalytic serine and instead use cysteine or glutamic acid (19). Clan SC peptidases present an identical geometry to the catalytic triad observed in clans PA and SB, yet this constellation is ordered differently in the polypeptide sequence. Substrate selectivity develops from the a-helices that surround the central β-sheet core. Within clan SC, carboxypeptidases from family S10 are unique for their ability to maintain catalytic activity in acidic environments. Nearly all serine peptidases have activity restricted within the range of neutral to alkaline pH. Many clan SC peptidases hydrolyze substrates on the C-terminal side of a proline residue with several exceptions. Both endoproteolytic and exoproteolytic activities occur in clan SC, which contrasts the trend in other serine peptidase families in which members are predominantly one or the other. For examples of differing selectivity in clan SC, prolyl oligopeptidase from family S9 cleaves peptide bonds within peptides, and prolyl aminopeptidase from family S33 removes N-terminal proline and hydroxyproline residues from peptides (20). Substrate selectivity for peptides shorter than 30 amino acids in length is derived from the two-domain architecture. An N-terminal seven-bladed propeller restricts access to the C-terminal α/β hydrolase domain and, in turn, the site of peptide bond hydrolysis (21). On the basis of their selectivity toward smaller peptides and not full-length proteins, clan SC peptidases are thought to be particularly important in cell signaling mechanisms.
In humans, clan SC peptidases are often associated with proline-specific N-terminal processing of peptides and proteins, yet many present a nonproteolytic function. S9 is the largest family of clan SC peptidases with 41 homologs in the human genome. Of these homologs, prolyl oligopeptidase (POP) and dipeptidyl peptidase IV (DPP-IV) are the best characterized. The crystal structure of POP revealed the two-domain architecture and basis for substrate selectivity. Notably, no naturally occurring inhibitor of this family of proteases has been found. A putative role for POP has been suggested in the metabolism of various neuropeptides (22). DPP-IV presents a similar two-domain architecture (23). DPP-IV is a transmembrane protein responsible for processing hormones and chemokines. Only three S10 family peptidases have been identified in the human genome, and their biological roles remain to be uncovered. Of three S28 family peptidases in humans, only dipeptidyl-peptidase II (DPP-II) is characterized. DPP-II catalyzes release of two N-terminal amino acids when proline or alanine is in the P1 position. Eighteen S33 family peptidases are in the human genome; however, many of them do not display peptidase activity. For example, several of these enzymes catalyze hydrolysis of epoxide bonds into diols and play a role in detoxification or cellular signaling (24).
Clan SB: family S8—subtilisins
Clan SB peptidases are prevalent in plant and bacterial genomes with few representatives in a given animal genome. However, these proprotein convertases are vital for protein processing in all metazoa. The archetype of clan SB is subtilisin. Subtilisin was originally discovered in the gram-positive bacterium Bacillus subtilis and like chymotrypsin was one of the earliest protein crystal structures determined (25). The catalytic triad of aspartic acid, histidine, and serine is found in the exact geometric organization observed in the peptidases of clans PA and SC, yet the surrounding protein fold bears no similarity (Fig. 2). Clan SB also contains a second family of peptidases S53, the sedolisins. In these peptidases, the histidine general base is substituted by a glutamic acid, and the tetrahedral intermediate is stabilized by a negatively charged carboxyl group from an aspartic acid rather than through partial positive charges. Subtilisins have proven extremely useful for protein engineering studies. Successful examples of engineering the subtilisin scaffold include substrate selectivity, thermal stability, cold adaptation, stability in nonaqueous solvents, fluoride activation, and ability to act as a peptide ligase (26). Many engineering studies on subtilisin have led to greatly improved cleaning agents for use in laundry detergent. Physiological function of clan SB peptidases tends to be nutrition-oriented with select roles in protein processing. Most clan SB peptidases prefer to hydrolyze substrates on the C-terminal side of large hydrophobic residues. However, proprotein processing peptidases such as kexin and furin cleave following a pair of dibasic residues (27). Most clan SB peptidases are secreted outside the cell or localized to the cell membrane. A notable exception is the tripeptidyl-peptidases responsible for intracellular protein turnover.
Figure 2. Subtilisin (PDB 1SCN) presents an identical catalytic triad to that observed in other serine proteases and enzymes yet within an entirely different protein fold.
Within the human genome, 10 clan SB peptidases have been identified; nine belong to the S8 family, and only one is from the S53 family. Although well known for their role in processing proteins along the secretion pathway, new roles for proprotein convertases are emerging, including regulation of plasma protein levels. Tripeptidyl-peptidase I (TPP-I) is the sole representative from family S53 in the human genome and one of many lysosomal peptidases responsible for protein turnover (28). TPP-I removes three amino acids from the N-terminus of small peptides. Mutations in TPP-I are associated with infantile neuronal ceroid lipofuscinosis (Batten disease), the most common neurodegenerative disorder in children, which is characterized by intracellular accumulation of autofluorescent lipopigments.
Over 90 phylogenetically distinct families of protease inhibitors have been classified by the MEROPS database. We will focus the current discussion on the most abundant and well-characterized groups.
Canonical mechanism: kunitz- and kazal-type inhibitors
In general, protease inhibitors interact with the peptidase in a canonical substrate-like manner with the protease. In this situation, three to four residues interact in an antiparallel β-sheet fashion within the enzyme active site. Many of these protease inhibitors are small proteins typically 30 to 120 amino acids in length (29). Often, these smaller proteases inhibitors contain many disulphide bonds. Many proteases have extremely high melting temperatures (>80 ° C) and retain their native conformation in the presence of strong chaotropes. Despite heterogeneity of sequence, each of these inhibitors presents a nearly identical conformation in reactive site loop that restricts proteolytic activity. In this process, the reactive site loop does not undergo conformational change when in complex with the protease and the P1 residue is positioned to place the carbonyl carbon at a very short distance to Ser-195. The carbonyl oxygen, in turn, points toward the oxyanion hole that forms H-bonds with the amide groups of Gly-193 and Ser-195. The amide nitrogen of the P1 residue is directed toward the Oy of Ser-195 not facing Ser-214 as typically observed in natural substrates. The latter change is thought to shorten during the catalytic cleavage process (30). The loop may be combined with other structural elements to mediate inhibition, including a P-hairpin following the loop or a disulphide bond in close proximity to the scissile bond. During interaction with the protease, the scissile bond remains intact and may undergo a slight deformation from planarity or a more distorted Michaelis complex depending on the enzyme and inhibitor pairing. Few specific contacts define the interaction between protease and inhibitor outside of the active site, where most are typically hydrophobic. In turn, composition of the reactive site loop contributes the most significant energetics in the binding process. Altering the P1 residue by mutagenesis, therefore, can be used to shift the inhibition profile dramatically. Lack of additional contacts ensures that most serine protease inhibitors regulate activity of multiple proteases in vivo.
MEROPS identifies Kunitz-type inhibitors as families I2 and I3, yet they seem to have developed separately in evolutionary history. Families I2 and I3 are referred to as “Kunitz-A” and “Kunitz-P” for their origin from animals and plants, respectively. Aprotinin, also known as bovine pancreatic trypsin inhibitor, was one of the first protease inhibitors identified and isolated by Kraut and coworkers in 1930. The I2 family is considerably more homogenous and thought to inhibit only S1 peptidases. In contrast, the I3 family is split into two phylogenetic groups, I3A and I3B, both of which typically inhibit S1 peptidases, yet members of the I3A family can also potentially inhibit the A1 family aspartyl proteases and the C1 family cysteine proteases. The first structure of an I3 inhibitor was the complex of soybean trypsin inhibitor with trypsin by Sweet and colleagues (31). The structure presents a β-barrel architecture capped by a pair of β-strands stabilized by two disulphide bonds (Fig. 3). The physiologic function of the Kunitz-type proteases remains unknown for many family members other than those in man. Prevention against digestion or invasion from pathogens has been suggested based on a common abundance in seeds. In man, tissue factor pathway inhibitor is a key Kunitz-type inhibitor responsible for regulating blood clot formation. On the basis of their potency, Kunitz-type inhibitors were among the first examined for therapeutic application. Aprotinin was approved for clinical application in coronary-artery bypass graft surgery in 1993. Fifteen years later, considerable controversy has developed over its use given an associated risk of mortality and the availability of less expensive lysine analogs that achieve the same goals (32).
Figure 3. Complex formation between trypsin and soybean trypsin inhibitor exemplifies the mechanism of Kunitz-type inhibitors (PDB 1AVW). Greatly reduced peptide bond hydrolysis rates lead to inhibition.
Kazal-type inhibitors are classified as family I1 by the MEROPS database. The name is derived from pancreatic secretory inhibitor, which is now termed SPINK1, originally isolated by Kazal and coworkers in 1948. The SPINK family (serine protease inhibitor, Kazal) plays important roles in the digestive system, lungs, skin, and likely many other tissues in the body. Six SPINKs can be identified in the human genome, and each contains multiple repeats of the Kazal-type fold. Mutations in SPINK1 are associated with hereditary pancreatitis (33). Netherton syndrome is a rare disorder that affects the skin of patients and results in ichthyosiform dermatosis and hair shaft abnormalities. Patients with Netherton syndrome are found to have a mutation on chromosome 5 in the SPINK5 gene, which encodes an array of 15 Kazal-type inhibitor domains (34). Considerable biochemical characterization has been carried out on the ovomucoid inhibitors of the Kazal family (35). Notably, as multiple Kazal-type domains are often found within a single polypeptide chain, they need not inhibit the same type of protease or protease specificity. For example, dog bikazin contains two Kazal-type domains in which one domain prefers trypsin, and the other prefers chymotrypsin.
Bait-and-trap mechanism: serpins
Serpins are found in all kingdoms of life, yet their presence in a given organism is not, which suggests the family has undergone significant gene transfer and loss. The family name was coined by Carrell and Travis as an acronym of serine protease inhibitor (36). Serpins are the most abundant form of serine protease inhibitor in the human genome. With the exception of fungi, all multicellular eukaryotes seem to possess one or more serpin genes. However, despite their ubiquity, few physiological functions are ascribed to serpins outside those known in man, and, in particular, they are associated with pathologies. Most serpins are irreversible inhibitors of serine proteases of the S1 family of peptidases with select family members inhibiting S9 of subtilisins and cysteine proteases. Several members of the serpin family have lost their ability to act as inhibitors and acquired new functions such as ovalbumin and pigment epithelium-derived factor. The unique mechanism of protease inhibition by serpins has received considerable attention.
Serpins consist of a conserved core of three β-sheets and eight or nine a-helices that act collectively in the inhibitory mechanism. As with the Kazal- and Kunitz-type inhibitors, the mechanism involves a surface exposed loop that is termed the reactive center loop (RCL). The RCL presents a short stretch of polypeptide sequence bearing the P1-P1’ scissile bond. Like other serine protease inhibitor families, the P1 residue dominates the thermodynamics that govern the interaction between protease and inhibitor. Exposure of the P1 residue to solvent is typically brokered by 15 amino acids N-terminal to the P1 residue and 5 residues on the C-terminal “prime” side of the scissile bond. Evidence for dramatic conformational change in the inhibitory mechanism was first provided by the crystal structure of the cleaved form of α1-antitrypsin (37). In this structure and unlike the native form, the reactive center loop was not solvent exposed but occurred as an additional β-strand within the core of the structure.
Dramatic conformational change of both inhibitor and protease is the most recognizable feature of the serpin mechanism (Fig. 4). After formation of the Michaelis-Menten encounter complex, the reactive center loop is cleaved, and an acyl enzyme intermediate is formed as in the normal serine protease catalytic mechanism. However, after bond hydrolysis, the RCL rapidly inserts into the central P-sheet of the serpin, which yields an overall stability enhancement to the inhibitor and traps the acyl-enzyme intermediate. It largely unknown how this change in conformation occurs. Several studies suggest the complex undergoes transient exchange between expelled and partially inserted states (38). Integration of the β-strand into the structure flips the protease from the “top” of the structure to the complete opposite side of the serpin. Because of this, the local environment of the catalytic triad is distorted and therefore cannot complete the catalytic cycle (39). When the process is finished, the previously adjacent P1 and P1’ residues are separated by over 70 A. During this process, the protease has been converted from a metastable free state into a more energetically favored relaxed bound state. Serpins have a considerably lower melting temperature (TM ~60 °C) in isolation than when cleaved (TM > 100 °C). Unrelated in sequence or structure, the macroglobulin family of protease inhibitors similarly applies scissile bond cleavage, yet the subsequent step involves entrapment of the protease in a cage-like architecture (40). Although most serpins apply this mechanism to inhibit serine proteases irreversibly, a select group has been shown to act reversibly. For example, protein C inhibitor, also known as PAI-1, acts as a reversible inhibitor to the single-chain urokinase-type plasminogen activator. Moreover, several serpins are known to integrate their cleaved RCL into another serpin molecule in trans (41). In turn, serpins can undergo polymerization, which becomes relevant in several pathological conditions.
Figure 4. A large conformational change defines the mechanism of serpin inhibition. Conversion of the Michaelis complex (PDB 1OPH) into cleaved trapped conformation (PDB 1 EZX) traps the RCL of serpin into the p-strand core of inhibitor. A significant gain in stability, therefore, is imparted to the entire serpin structure.
Serpins play key regulatory functions in man. α1-antitrypsin serves a major role in protecting the connective tissue of the lungs from leukocyte-released elastase. The C1 inhibitor restricts proteases of the immune system from unwanted proteolysis and inflammation. Two plasminogen activator inhibitors control fibrinolysis. Viral serpins have also been described that broker their survival and propagation through restricting these same proteolytic pathways. Control of blood clot formation is through antithrombin. However, unlike other serpin family members, the interaction between clotting factor protease and antithrombin is greatly facilitated by heparin or heparin sulfate glycosaminoglycans, which bind to the inhibitor to mediate this effect (42). In turn, antithrombin is directed toward regulation of protease activity at cell surfaces such as the vascular endothelium, which display heparin in various forms.
Exosite recognition: hirudin and other anticoagulant molecules
Numerous strategies have evolved in different pathogenic and parasitic organisms to alter the coagulation cascade for their own benefit. Snake and leech venoms have yielded a plethora of serine proteases and serine protease inhibitors that bear this trait. Many inhibitors function through hijacking the macromolecular recognition regions in blood clotting factors, the earliest known example of which is hirudin, a 66-amino acid residue protein that is an extremely tight binding and selective inhibitor of the blood coagulation protease thrombin. Indeed, the use of leeches in medicine dates back to the Greeks in 200 B.C. However, it was not until 1884 that Haycraft isolated the anticoagulant agent from medicinal leech saliva and termed this agent hirudin. Selective and tight binding results from the cooperative binding at both the anion binding exosite responsible for fibrinogen recognition and active site of thrombin. Notably, unlike nearly all other protease inhibitors, active site recognition involves formation of a parallel β-sheet. Hirudin has been derivatized and modified in various ways to develop direct thrombin inhibitors. Two recombinant forms, lepirudin and desirudin, are available for clinical use to prevent deep-vein thrombosis after surgery and treat heparin-induced thrombocytopenia. However, various problems have limited their use, including bleeding problems, short half-life, and development of antihirudin antibodies. Additional modification of the hirudin platform and other medicinal chemistry strategies aimed at thrombin inhibition is well described in Reference 43.
Characterization of Protease-Inhibitor Interactions
Defining the selectivity and potency of an inhibitor relies on accurate characterization of the protease. In particular, values of kcat and kcat/KM are readily obtained using basic enzymology and various commercially available chromogenic and fluorogenic substrates. Proteases act on a single substrate during the catalytic cycle. Therefore, models to interpret data follow classical descriptions of competitive, noncompetitive, and mixed inhibition. As with traditional enzymology, curve-fitting measures combined with graphical validation of data is suggested as a more accurate approach than the use of initial rates analysis. Use of IC50 is to be avoided as the values of this measure are only a crude comparison between reversible inhibitors. For serine proteases, the use of irreversible inhibitors, such as chloromethyl ketones, are extremely useful for determining the amount of active protease in a given protein preparation. Many serine protease inhibitors form stable complexes with their target proteases that can be resolved via gel electrophoresis as a simple and effective means of visualization. However, many protease-inhibitor interactions require more advanced data treatment. Some examples include slow tight-binding mechanisms and higher-order stoichiometries of inhibition (44). Lastly, conformational change can be measured through various biophysical techniques including UV and fluorescence spectroscopy, circular dichroism, and isothermal titration calorimetry.
Serine Proteases and their Inhibitors in Disease
As illustrated above, the human genome encodes a large and diverse population of serine proteases and serine protease inhibitors. Design of small-molecule inhibitors to restrict proteolytic activity continues to garner attention in the pharmaceutical industry. Given the many related proteases in the genome, this task is particularly challenging. Early work focused on active site-directed therapies. However, as evidenced by the naturally occurring serine protease inhibitors, active site recognition enables the regulation of multiple protease targets. Minimization of unwanted side effects is then a significant hurdle. More recent effort has sought the development of therapeutics that focus on other regions of the protease. The crucial role for such regions in biological systems is demonstrated by the blood coagulation and immune system proteases, in which macromolecular recognition is dependent on exosites and the allosteric communication of these regions with the active site.
Diversity of proteases and inhibitors results in a wide range of pathologies that result from disruption in either serine protease or serine protease inhibitor. The earliest descriptions of such imbalances are within the blood coagulation cascade. For example, hemophilia B results from deficiency in coagulation factor IX. In contrast, excessive activation of immune system serine proteases produces inflammatory states. Errors in serine protease inhibitors can have consequences on multiple biological systems. However, overlapping inhibition by multiple families of inhibitor can, in certain instances, lessen the severity of the pathology. Genetic abnormalities in the serpins have also been associated with polymerization and therefore belong to the category of conformational disease. Emphysema, cirrhosis, and thrombosis may result from such aberrant conformational transitions. Neuroserpin may also play a key role in familial encephalopathy because of the formation of inclusion body-like material (45). Understanding the molecular mechanisms of limited proteolysis and their regulation in vivo remains a challenging and insightful venue to improve human health.
1. Rawlings ND, Morton FR, Kok CY, Kong J, Barrett AJ. MEROPS: the peptidase database. Nucleic Acids Res. 2008; 36:D320-325.
2. Page MJ, Di Cera E. Serine peptidases: classification, structure and function. Cell. Mol. Life Sci. In press.
3. Hedstrom L. Serine protease mechanism and specificity. Chem. Rev. 2002; 102:4501-4524.
4. Blow DM, Birktoft JJ, Hartley BS. Role of a buried acid group in the mechanism of action of chymotrypsin. Nature 1969; 221:337-340.
5. Dodson G, Wlodawer A. Catalytic triads and their relatives. Trends Biochem. Sci. 1998; 23:347-352.
6. Lopez-Otin C, Overall CM. Protease degradomics: a new challenge for proteomics. Nat. Rev. Mol. Cell Biol. 2002; 3:509-519.
7. Page MJ, Macgillivray RT, Di Cera E. Determinants of specificity in coagulation proteases. J. Thromb. Haemost. 2005: 3;2401-2408.
8. Pallen MJ, Wren BW. The HtrA family of serine proteases. Mol. Microbiol. 1997; 26:209-221.
9. Hartley BS, Kilby BA. The reaction of p-nitrophenyl esters with chymotrypsin and insulin. Biochem. J. 1954; 56:288-297.
10. Bode W, Fehlhammer H, Huber R. Crystal structure of bovine trypsinogen at 1-8 A resolution. I. Data collection, application of Patterson search techniques and preliminary structural interpretation. J. Mol. Biol. 1976: 106:325-335.
11. Page MJ, Di Cera E. Role of Na+ and K+ in enzyme function. Physiol. Rev. 2006; 86:1049-1092.
12. Whitcomb DC, Lowe ME. Human pancreatic digestive enzymes. Dig Dis. Sci. 2007; 52:1-17.
13. Pereira PJ, Bergner A, Macedo-Ribeiro S, Huber R, Matschiner G, Fritz H, Sommerhoff CP, Bode W. Human beta-tryptase is a ring-like tetramer with active sites facing a central pore. Nature 1998; 392:306-311.
14. Lin CY, Anders J, Johnson M, Sang QA, Dickson RB. Molecular cloning of cDNA for matriptase, a matrix-degrading serine protease with trypsin-like activity. J. Biol. Chem. 1999; 274: 18231-18236.
15. Diamandis EP, Yousef GM, Luo LY, Magklara A, Obiezu CV The new human kallikrein gene family: implications in carcinogenesis. Trends Endocrinol. Metab. 2000; 11:54-60.
16. Proud D, Kaplan AP Kinin formation: mechanisms and role in inflammatory disorders. Annu. Rev. Immunol. 1988; 6:49-83.
17. Barry M, Bleackley RC. Cytotoxic T lymphocytes: all roads lead to death. Nat. Rev. Immunol. 2002; 2:401-409.
18. Nardini M, Dijkstra BW. Alpha/beta hydrolase fold enzymes: the family keeps growing. Curr. Opin. Struct. Biol. 1999; 9:732-737.
19. Hotelier T, Renault L, Cousin X, Negre V, Marchot P, Chatonnet A. ESTHER, the database of the alpha/beta-hydrolase fold superfamily of proteins. Nucleic Acids Res. 2004; 32:D145-147.
20. Rea D, Fulop V. Structure-function properties of prolyl oligopeptidase family enzymes. Cell Biochem. Biophys. 2006; 44:349-365.
21. Kuhn B, Hennig M, Mattei P. Molecular recognition of ligands in dipeptidyl peptidase IV. Curr. Top Med. Chem. 2007; 7:609-619.
22. Garcia-Horsman JA, Mannisto PT, Venalainen JI. On the role of prolyl oligopeptidase in health and disease. Neuropeptides 2007; 41:1-24.
23. Rasmussen HB, Branner S, Wiberg FC, Wagtmann N. Crystal structure of human dipeptidyl peptidase IV/CD26 in complex with a substrate analog. Nat. Struct. Biol. 2003; 10:19-25.
24. Arand M, Cronin A, Adamska M, Oesch F. Epoxide hydrolases: structure, function, mechanism, and assay. Methods Enzymol. 2005; 400:569-588.
25. Wright CS, Alden RA, Kraut J. Structure of subtilisin BPN’ at 2.5 angstrom resolution. Nature 1969; 221:235-242.
26. Bryan PN. Protein engineering of subtilisin. Biochim. Biophys. Acta 2000; 1543:203-222.
27. Hosaka M, Nagahama M, Kim WS, Watanabe T, Hatsuzawa K, Ikemizu, J, Murakami K, Nakayama K. Arg-X-Lys/Arg-Arg motif as a signal for precursor cleavage catalyzed by furin within the constitutive secretory pathway. J. Biol. Chem. 1991; 266:12127-12130.
28. Turk B, Turk D, Turk V. Lysosomal cysteine proteases: more than scavengers. Biochim. Biophys. Acta. 2000; 1477:98-111.
29. Bode W, Huber R. Natural protein proteinase inhibitors and their interaction with proteinases. Eur. J. Biochem. 1992; 204:433-451.
30. James MN, Sielecki AR, Brayer GD, Delbaere LT, Bauer CA. Structures of product and inhibitor complexes of Streptomyces griseus protease A at 1.8 A resolution. A model for serine protease catalysis. J. Mol. Biol. 1980; 144:43-88.
31. Sweet RM, Wright HT, Janin J, Chothia CH, Blow DM. Crystal structure of the complex of porcine trypsin with soybean trypsin inhibitor (Kunitz) at 2.6-A resolution. Biochemistry 1974; 13:4212-4228.
32. Sedrakyan A, Treasure T, Elefteriades JA. Effect of aprotinin on clinical outcomes in coronary artery bypass graft surgery: a systematic review and meta-analysis of randomized clinical trials. J. Thorac. Cardiovasc. Surg. 2004; 128:442-448.
33. Witt H, Luck W, Hennies HC, Classen M, Kage A, Lass U, Landt O, Becker M. Mutations in the gene encoding the serine protease inhibitor, Kazal type 1 are associated with chronic pancreatitis. Nat. Genet. 2000; 25:213-216.
34. Chavanas S, Bodemer C, Rochat A, Hamel-Teillac D, Ali M, Irvine AD, Bonafe JL, Wilkinson J, Taieb A, Barrandon Y, Harper JI, de Prost Y, Hovnanian A. Mutations in SPINK5, encoding a serine protease inhibitor, cause Netherton syndrome. Nat. Genet. 2000; 25:141-142.
35. Laskowski M Jr, Kato I. Protein inhibitors of proteinases. Annu. Rev. Biochem. 1980; 49:593-626.
36. Travis J, Owen M, George P, Carrell R, Rosenberg S, Hallewell RA, Barr PJ. Isolation and properties of recombinant DNA produced variants of human alpha 1-proteinase inhibitor. J. Biol. Chem. 1985; 260:4384-4389.
37. Loebermann H, Tokuoka R, Deisenhofer J, Huber R. Human alpha 1-proteinase inhibitor. Crystal structure analysis of two crystal modifications, molecular model and preliminary analysis of the implications for function. J. Mol. Biol. 1984; 177:531-557.
38. Whisstock JC, Bottomley SP. Molecular gymnastics: serpin structure, folding and misfolding. Curr. Opin. Struct. Biol. 2006; 16:761-768.
39. Huntington JA, Read RJ, Carrell RW. Structure of a serpin protease complex shows inhibition by deformation. Nature 2000; 407:923-926.
40. Sottrup-Jensen L. Alpha-macroglobulins: structure, shape, and mechanism of proteinase complex formation. J. Biol. Chem. 1989; 264:11539-11542.
41. Zhou A, Carrell RW. Dimers initiate and propagate serine protease inhibitor polymerisation. J. Mol. Biol. 2008; 375;36-42.
42. Bourin MC, Lindahl U. Glycosaminoglycans and the regulation of blood coagulation. Biochem. J. 1993; 289(Pt 2):313-330.
43. Schwienhorst A. Direct thrombin inhibitors - a survey of recent developments. Cell. Mol. Life. Sci. 2006; 63:2773-2791.
44. Schechter NM, Plotnick MI. Measurement of the kinetic parameters mediating protease-serpin inhibition. Methods 2004; 32:159-168.
45. Davis RL, Shrimpton AE, Holohan PD, Bradshaw C, Feiglin D, Collins GH, Sonderegger P, Kinter J, Becker LM, Lacbawan F, Krasnewich D, Muenke M, Lawrence DA, Yerby MS, Shaw CM, Gooptu B, Elliott PR, Finch JT, Carrell RW, Lomas DA. Familial dementia caused by polymerization of mutant neuroserpin. Nature 1999; 401:376-379.
The MEROPS database of peptidases and inhibitors is an invaluable resource that can be found at http://merops.sanger.ac.uk/
Protease Inhibitors, Mechanisms of
Approaches to Enzyme Inhibition