INTRODUCTION

TOP ABSTRACT INTRODUCTION HUMAN TRYPTASE FAMILY STRUCTURE OF -TRYPTASE... REGULATION OF TRYPTASE BY... INTERACTION WITH SUBSTRATES AND... CONSEQUENCES OF TRYPTASE... CONCLUSIONS REFERENCES

Keywords: airways; mast cells; tissue remodeling; tryptase

By using aniline dyes to stain histological slides in the 1870s, Paul Ehrlich discovered heavily granulated cells he believed to be the products of the "stuffing" of connective tissue cells. Accordingly, he chose the name mast cells ("stuffed cells"). The accumulation of mast cells in chronically inflamed and "overfed," that is, remodeled tissues already noted by Ehrlich has subsequently been verified by numerous studies and extended to a variety of disorders including, for example, fibrotic lung diseases and asthma. However, contrary to Ehrlich's assumption that mast cells are only by-products, they are now envisaged to contribute directly to the pathogenesis of these disorders. Among the plethora of proinflammatory, vasoactive, chemotactic, and growth-promoting mediators synthesized by mast cells, interest has focused on tryptases. These proteinases are highly and selectively expressed in mast cells and are stored in exceptionally large amounts, in a fully enzymatically active state, in their secretory granules, a fact that is increasingly utilized for the detection of mast cells and their degranulation. More importantly, tryptases possess unique features that distinguish them from other trypsin-like proteinases, and they seem to be capable of inducing and driving various processes that contribute to (chronic) inflammation and tissue remodeling.

	HUMAN TRYPTASE FAMILY

TOP ABSTRACT INTRODUCTION HUMAN TRYPTASE FAMILY STRUCTURE OF -TRYPTASE... REGULATION OF TRYPTASE BY... INTERACTION WITH SUBSTRATES AND... CONSEQUENCES OF TRYPTASE... CONCLUSIONS REFERENCES

Human tryptases comprise a family of trypsin-like serine proteinases that are encoded by at least three genes located together with four or more adjacent tryptase-like pseudogenes on chromosome 16p13.3 (1, 2). The tryptases, identified at the cDNA and protein level, are divided into three groups, that is, -, -, and - or transmembrane tryptases (TMTs). Within each group, members show a high degree of sequence identity ( 98%), suggesting that they may be allelic variants of each other. So far nine tryptases (1, 2; 1a, 1b, 2, 3; and 1, 2, TMT) have been sequenced; however, the number of isoenzymes may be considerably larger as the human tryptase loci are highly polymorphic (3). The deduced amino acids sequences of the mature - and -tryptases are ~ 90% identical whereas the /TM tryptases are less closely related (~ 50% identity with - and -tryptases) and contain a C-terminal hydrophobic domain, a feature not found in other tryptases.

Mast cells appear to express -, -, and -tryptases. The main types stored in their secretory granules, however, are -isoenzymes, which accumulate in much larger amounts than any of the other granule-associated serine proteinases of leukocytes and lymphocytes, comprising as much as 25% of the mast cell protein. Therefore, -tryptases also are the major isoenzymes that are released during mast cell degranulation and that are isolated from normal human lung and skin tissues. Because the exact isoenzyme composition of isolated tryptases is usually not known, the enzymatically active tetramers of such preparations, as well as of -tryptases produced recombinantly, will subsequently be referred to as "tryptase."

	STRUCTURE OF -TRYPTASE TETRAMER

TOP ABSTRACT INTRODUCTION HUMAN TRYPTASE FAMILY STRUCTURE OF -TRYPTASE... REGULATION OF TRYPTASE BY... INTERACTION WITH SUBSTRATES AND... CONSEQUENCES OF TRYPTASE... CONCLUSIONS REFERENCES

Among the serine proteinases, tryptases are unique insofar as they are enzymatically active in the form of a noncovalently linked heparin-stabilized tetramer, resistant to most proteinaceous inhibitors, and display an unusual substrate specificity with a preference for peptidergic over macromolecular substrates (see below for details). These distinguishing features are well explained by the 3 Å crystal structure of the human lung 2-tryptase tetramer in complex with the inhibitor 4-amidinophenylpyruvic acid (APPA) (4).

The crystal structure reveals that the tetramer is not a compact tetrahedral body as predicted by initial models based on monomeric proteinases. Rather, the monomers (arbitrarily assigned as A, B, C, and D in Figure 1) are positioned at the corners of a flat rectangular frame, leaving a continuous central pore. All monomers of the tetramer are nearly equivalent in structure; however, only monomer A is identical to C, and only monomer B is identical to D, as the symmetry is disturbed because of the conformations of the Tyr-75 side chains located in the A-B and C-D interfaces and the slightly different environment of each monomer in the crystal.

View larger version (124K): [in this window] [in a new window]   Figure 1.   Front, side, and top views onto the tryptase tetramer. The four monomers A, B, C, and D (clockwise) shown in ribbons are arranged at the corners of a flat, framelike structure. The four active sites are directed toward the central pore; the catalytic triad residues Ser-195, His-57, and Asp-102 are shown in orange. The six unique loops that differ most from the archetypal serine proteinase trypsin and realize all contacts between neighboring monomers via two distinct interfaces are colored. Thus, for example, monomer A interacts with monomer D via the 60-loop (blue), the 97-loop (magenta), and 173-flap (turquoise), and with monomer B via the 37-loop (red ), the 70-80-loop (yellow), and the 147-loop including the 152-"spur" (green); this loop protrudes from each of the monomers toward the central pore and partially obstructs its entrances.

The tryptase monomer exhibits the typical -strand-dominated fold seen in other trypsin-like proteinases and is remarkably similar to the archetypal serine proteinases trypsin and chymotrypsin: more than 160 of the 245 amino acid residues of the monomer are topologically equivalent to those of trypsin and chymotrypsin. Six surface loops (colored in Figure 1), however, differ greatly in conformation and partially in length from those of other trypsin-like proteinases. These six loops are located around the active site, border and shape the outer rim of the active site cleft, and at the same time realize all contacts with the neighboring monomers.

The two distinct interfaces that connect each monomer to its two neighbors differ considerably in size and quality. The interface between monomers A and D (and the equivalent between monomers B and C), which is realized by the 173-flap, the 97-loop, and the 60-loop of both monomers, is relatively large (contact surface area 1,075 Å²). Besides a number of hydrophobic contacts, it comprises a salt bridge (Asp-60B to Arg-224) and four hydrogen bonds that stabilize this contact. In comparison, the A-B (and the equivalent C-D) interface, which involves the 147-loop, the 70-80-loop, and the 37-loop of both monomers, is considerably smaller (540 Å²), exclusively hydrophobic in nature, and lacks any stabilizing hydrogen bonds and salt bridges. Furthermore, this interface presumably is destabilized by the charges of the two Arg-150 residues that oppose each other and the asymmetric conformation of the two Tyr-75 residues that would clash if arranged in a symmetrical manner. Thus these small and exclusively hydrophobic A-B and C-D interfaces appear to be the built-in shear points of the tryptase tetramer.

	REGULATION OF TRYPTASE BY PROTEOGLYCANS AND INHIBITORS

TOP ABSTRACT INTRODUCTION HUMAN TRYPTASE FAMILY STRUCTURE OF -TRYPTASE... REGULATION OF TRYPTASE BY... INTERACTION WITH SUBSTRATES AND... CONSEQUENCES OF TRYPTASE... CONCLUSIONS REFERENCES

Tryptase is colocalized in the secretory granules of mast cells and secreted bound to heparin proteoglycans (7). These acidic aminoglycans are likely to bind to two clusters of positively charged residues that are located on the peripheral surface adjacent to the A-B and C-D interfaces, and thus bridge and stabilize these shear points of the tetramer. In vitro, tryptase can be stabilized by high salt concentrations, which strengthen the hydrophobic interactions and weaken the electrostatic arginine-arginine repulsion within this interface.

In the absence of stabilizing aminoglycans or high salt concentrations, the tetramer spontaneously and rapidly disintegrates into monomers with little known enzymatic and biologic activity (8). This inactivation, a complex multistep process whose reversibility appears be a function of the pH (9), is accompanied by conformational changes consistent with the conversion of the active site to a zymogen-like structure (10). Structural features support the hypothesis that on tetramer dissociation the monomers undergo transformation from an active conformation stabilized by intermonomer contacts into a zymogen-like state that is favored because of intramonomer interactions. Thus, in the tetramer, the AD-type intermonomer contacts are likely to support insertion of the N-terminal Ile-16 residue of each monomer into the Ile-16 pocket, where its free -amino group forms a solvent-inaccessible salt bridge with the carboxylate group of Asp-194, a requirement for a functional substrate recognition site and thus the enzymatic activity (11). In the isolated monomers, however, a second acidic anchoring point (residues Asp-143, Asp-145, and Asp-147) may successfully compete with Asp-194 for the binding of the N-terminal Ile-16 -amino group, and the monomer thus converts into a zymogen-like conformation that is stabilized by the zymogen triad residues His-40 and Ser-32.

The stabilization of the enzymatically active tetramer by aminoglycans and its dissociation into inactive monomers appear to be the predominant mechanisms regulating the enzymatic activity of tryptase in vivo. In contrast to virtually all other serine proteases, human tryptase is resistant to endogenous proteinase inhibitors, resulting in a prolonged catalytic activity in the extracellular space and even in plasma. Tryptase is not inhibited by large proteinase inhibitors such as ₂-macroglobulin and the serpins (12) as they are far too bulky to fit into the central pore and to interact with one of the active centers. Also, smaller Kunitz-type inhibitors such as bovine pancreatic trypsin inhibitor (aprotinin) and classic Kazal-type inhibitors (e.g., the Kunitz domaine of the Alzheimer -amyloid precursor protein) cannot inhibit tryptase as their distal poles would collide with the adjacent monomers. Similarly, based on docking experiments SLPI (secretory leukocyte protease inhibitor, or mucus proteinase inhibitor [MPI]), a major serine proteinase inhibitor in human airways, and even elafin (skin-derived antileukoprotease [SKALP]), a smaller protein corresponding to the inhibitory active second domain of SLPI, should not be able to bind to an active center to inhibit tryptase (Figure 2). Although both Alter and coworkers (12) and our group could not detect any significant effect of SLPI on tryptase, Wright and coworkers reported that SLPI inhibits tryptase with a subnanomolar equilibrium dissociation constant for the complex (13). Potentially, SLPI "inhibits" tryptase in a fashion similar to antithrombin III, lactoferrin, and myeloperoxidase, which destabilize the enzymatically active tetramer by competing for the binding of heparin (12, 14, 15). The relevance of this mechanism for the regulation of tryptase in vivo remains to be determined as the results obtained in vitro depend largely on the nature and the concentration of the aminoglycan used to stabilize tryptase.

View larger version (123K): [in this window] [in a new window]   View larger version (133K): [in this window] [in a new window]   Figure 2.   Models of the interaction of the inhibitors (A) SLPI (secretory leukocyte protease inhibitor) and (B) elafin with the tryptase tetramer. SLPI and elafin (red) were docked into the active site of monomer A by superposition of the proteinase moiety of their known complexes with chymotrypsin (12a) and porcine pancreatic elastase (12b), respectively. If bound to tryptase, both inhibitors would collide with monomer D and the A-D interface.

Both the lack of an endogenous inhibitor and the resistance to Kunitz-type inhibitors distinguish human tryptase from those of most other species. For example, bovine tryptase is colocalized in mast cells with its endogenous inhibitor aprotinin and rat tryptase with trypstatin, a fragment of the inter--trypsin inhibitor light chain. Furthermore, Kunitz-type inhibitors such as aprotinin and trypstatin inhibit bovine, canine, and rat tryptases but have little affinity for human tryptase; on the other hand, LDTI (leech-derived tryptase inhibitor), an atypical Kazal-type inhibitor, is a tight-binding inhibitor of the human but not the bovine enzyme (16). These differences are likely to reflect variations between species in the structures of the active centers, their positioning within the tetramer, and the size of the tetramer's central pore and thus may also indicate differences in substrate specificity.

	INTERACTION WITH SUBSTRATES AND RECEPTORS

TOP ABSTRACT INTRODUCTION HUMAN TRYPTASE FAMILY STRUCTURE OF -TRYPTASE... REGULATION OF TRYPTASE BY... INTERACTION WITH SUBSTRATES AND... CONSEQUENCES OF TRYPTASE... CONCLUSIONS REFERENCES

Tryptase has a puzzling substrate specificity: like trypsin, tryptase rapidly hydrolyzes peptide substrates carboxyterminal of arginine and lysine residues; unlike trypsin, however, it cleaves only a few proteins. This unusual specificity results from an active site that is not suitable to confer specificity but is located within the central pore of the tetramer. Thus, the active site of each monomer is similar in structure to that of trypsin with a nearly identical S1 specificity pocket, an S2 subsite that is open and larger than that of trypsin, and a fully blocked S3/S4 subsite. The access to the four active sites of the tetramer, however, is severely limited as they are directed toward the central pore. Tryptase may thus be considered as a member of the "self-compartmentalizing proteases," a small group of proteolytic enzymes (e.g., Gal6, the tricorn protease and the proteasome) that form compartments through self-association and segregate their proteolytic active sites to the interior of these compartments.

With a cross-section of ~ 40 × 15 Å at the entrances, the central pore is just large enough for elongated peptides of the diameter of an -helix to thread through and to interact with the active sites. Thus, neuropeptides such as calcitonin gene-related peptide (CGRP), vasoactive intestinal peptide (VIP), and peptide histidine methionine (PHM) as well as peptide hormones such as vasopressin and kinetensin (reviewed in Tam [17]) can easily enter the central pore to be cleaved and destroyed. Similarly, kininogen fragments resulting from the cleavage by neutrophil elastase at inflammatory sites are small enough to enter the central pore of the tryptase tetramer, where they may even dock to adjacent active sites so that the octapeptide bradykinin can be excised by (quasi)simultaneous cleavages (18). However, a small size alone apparently is not a sufficient prerequisite for a polypeptide to be recognized as a tryptase substrate, as the tachykinins substance P and neurokinin A as well as neurotensin are not hydrolyzed despite potential tryptic cleavage sites.

Few macromolecular substrates are cleaved by tryptase, leading, for example, to the activation of the zymogens of stromelysin 1 and urokinase-type plasminogen activator (19, 20), and the inactivation of fibronectin and of the procoagulant functions of high molecular mass kininogen and fibrinogen (21). These large proteins that do not fit into the central pore of the tetramer must extend cleavable surface loops of > 20 Å to interact with an active site. Such loops may be extracted from the substrate structure by secondary interactions with negatively charged amino acid residues that cluster on the inner pore-facing surface; docking studies with models of prostromelysin (MMP-3) and the prourokinase suggest that the propeptides of these zymogens could otherwise not interact with the active centers within the tetramer. Alternatively, it is possible that the observed cleavages of some macromolecular proteins are due to tryptase monomers, which may have some residual activity (8) withbased on the structure of the active sitebroad trypsin-like specificity.

In addition to the few soluble substrates known and discussed above, tryptase cleaves and activates the "protease-activated receptor 2" (PAR-2), which may be one of its main purposes. PAR-2 is a member of a subgroup of the G protein-coupled, seven-transmembrane domain cell surface receptors that are activated by a proteolytic mechanism: The cleavage of the N-terminal extracellular domain unmasks a new amino terminus that subsequently functions as a tethered ligand to activate the receptor (for reviews see (Cocks and Moffatt [24] and Coughlin [25]). The activation cleavage site (. . .KGR SLIGKV. . .) in the N-terminal domain of PAR-2 is located ~ 40 amino acid residues upstream from the predicted first transmembrane segment and thus presented at sufficient distance from the cell membrane to interact with an active site inside the tryptase tetramer. Because of the required proteolytic cleavage, activation of PARs is irreversible; for a prolonged response to occur, receptors on the cell surface need to be replenished by translocation from intracellular pools and subsequently by resynthesis that may require stimulation of their expression by inflammatory mediators (26). Evidence is accumulating that many of the cellular effects of tryptase (see below) are mediated by PAR-2; on the basis of indirect evidence, however, tryptase-like trypsin activates cells by both PAR-2-dependent and -independent mechanisms (27, 28). It also remains to be determined whether tryptase can elicit all the effects that are induced by other PAR-2 agonists, that is, trypsin and peptides imitating the tethered ligand; contrary to the results obtained with tryptase (see below), these effects include the release from airway epithelial cells of bronchodilatatory prostanoids, which have protective effects in the airways (29, 30).

	CONSEQUENCES OF TRYPTASE RELEASE IN THE AIRWAYS

TOP ABSTRACT INTRODUCTION HUMAN TRYPTASE FAMILY STRUCTURE OF -TRYPTASE... REGULATION OF TRYPTASE BY... INTERACTION WITH SUBSTRATES AND... CONSEQUENCES OF TRYPTASE... CONCLUSIONS REFERENCES

Although the number of known substrates is limited compared with other trypsin-like proteases, tryptase initiates and drives a variety of processes relevant to airway inflammation and remodeling. These include airway fibrosis and extracellular matrix turnover, angiogenesis, airway smooth muscle and epithelial cell hyperplasia, inflammation, and alterations of bronchial tone (Figure 3).

View larger version (32K): [in this window] [in a new window]   Figure 3.   Road map of some of the processes initiated by mast cell tryptase (thick arrows) relevant for airway inflammation and tissue remodeling. Chymase, a serine protease that is stored and secreted together with tryptase by a subset of airway mast cells, may contribute, for example, by stimulating vascular permeability and airway gland serous cell secretion, and by cleavage of constituents of the extracellular matrix and the epithelial cell glycocalyx. Modified and updated with permission from Reference 30a.

The ability of tryptase to stimulate fibroblast proliferation and chemotaxis has been well documented (27, 31). In addition, tryptase increases the synthesis and secretion of type I collagen (32, 33), suggesting a role particularly in chronic fibrosis because this type is preferentially found in established fibrotic lesions. Tryptase also increases the amounts of collagenolytic activity secreted by cultured fibroblasts (33) and activates the zymogens of stromelysin 1 and urokinase-type plasminogen activator (19, 20), which, in turn, can activate plasminogen and various matrix metalloproteases; the capacity of this array of proteases to alter the collagen, elastin, and other protein/proteoglycan composition of the extracellular matrix of the airways is enormous. Whether tryptase itself can degrade macromolecular proteins as it has been reported for fibronectin (21) and even intact type VI collagen microfibrils needs to be reconsidered in light of the structure; the cleavages seen in these preparations may, for example, be secondary to the activation of small amounts of prourokinase or pro-MMPs present.

Tryptase may further contribute to airway remodeling by stimulating proliferation of airway epithelial cells (34), bronchial smooth muscle cells (35), and endothelial cells (36, 37). Besides being a mitogen, tryptase may modulate the migration of cells by cleaving fibrinogen and potentially fibronectin that are enriched in the airways in association with injury and inflammation; the cleavage of fibrinogen destroys an Arg-Gly-Asp (RGD) sequence motif at amino acids 572-573 of its -chain (23) that is recognized by cell surface _v3-integrins present, for example, on smooth muscle cells, endothelial cells, and mast cells.

The many proinflammatory actions of tryptase that have been demonstrated in vitro and in vivo include the ability to cause edema formation and the accumulation of eosinophils and neutrophils (38, 39). These actions are at least in part secondary to the degranulation of mast cells (38, 40, 41), thereby providing a potentially important positive feedback mechanism by which the release of tryptase may amplify and perpetuate the response of mast cells to the initial stimulus. Likely, the ability of tryptase to stimulate the expression of ICAM-1 and IL-8, a potent neutrophil chemoattractant, in endothelial and epithelial cells (34, 36), also contributes to these effects; together with neutrophil elastase, tryptase may subsequently process high molecular mass kininogen to the proinflammatory bradykinin even if it is oxidized and therefore resistant to hydrolysis by kallikreins (18).

Particular interest has been focused on the ability of tryptase to modulate airway smooth muscle tone and responsiveness. Tryptase potentiates histamine-induced contractions both in canine and human bronchi via calcium-related mechanisms (42, 43); interestingly, in human bronchi this effect is confined to bronchi sensitized to antigens. In addition, rather complex interactions between tryptase and sensory nerves may affect bronchomotor tone. Thus, tryptase has been shown to stimulate sensory nerves to release neuropeptides via a PAR-2- mediated mechanism (44). Some of the neuropeptides released, that is, the bronchodilating peptides VIP and PHM and the vasodilator/bronchoconstrictor CGRP, may subsequently be rapidly cleaved and destroyed by tryptase, whereas bronchoconstrictory tachykinins, for example, substance P, neurokinin A, and neurokinin B, resist inactivation by tryptase (17). Thus, tryptase may induce neurogenic inflammation and at the same time modulate the consequences, shifting the overall response toward a bronchoconstrictory phenotype. These tryptase-mediated and -modulated interactions may be particularly relevant due to the close anatomic association of mast cells with nerve fibers and further complicated by the ability of neuropeptides to induce degranulation of some mast cell subtypes.

Two lines of evidence obtained in vivo support the hypothesis that the mechanisms outlined above contribute to the regulation of airway tone, responsiveness, and inflammation. Thus, in allergic sheep the inhalation of aerosolized human tryptase causes bronchoconstriction and airway hyperresponsiveness (40), and pretreatment with inhibitors directed against human tryptase before allergen challenge reduces the early- and late-phase responses and the development of airway hyperresponsiveness and eosinophilic inflammation (45). Moreover, a first clinical trial has suggested that tryptase inhibitors may lessen the response of subjects with mild asthma to inhaled allergen (46). However, similar effects were also obtained after administration of SLPI in sheep (13), which in our opinion does not inhibit tryptase (see above). The second line of evidence stems from genetic linkage analyses performed in mice, revealing that a low airway responsiveness to acetylcholine and serotonin is associated with a point mutation that leads to the activation of a cryptic splice site and thus a deficiency in MMCP-7, one of the two tryptases of this species (47). Potentially, the polymorphisms of the human tryptase genes described (3) provide similar mechanisms regulating airway tone and responsiveness in humans.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION HUMAN TRYPTASE FAMILY STRUCTURE OF -TRYPTASE... REGULATION OF TRYPTASE BY... INTERACTION WITH SUBSTRATES AND... CONSEQUENCES OF TRYPTASE... CONCLUSIONS REFERENCES

In summary, the unique tetrameric architecture explains many of the unusual biochemical and biological properties of -tryptases and supports their proposed role in chronic inflammation and tissue remodeling. Once released by mast cell degranulation, tryptases remain in tissues for a prolonged period of time because of the macromolecular size of the tetramer, which limits its diffusion, and because of its binding to heparin and presumably other extracellular aminoglycans; in addition, once released tryptase may amplify and perpetuate the initial response by stimulating mast cell degranulation. Tryptase remains enzymatically active because the active sites are located within the tetramer's central pore, which hinders the access and binding of endogenous proteinase inhibitors that rapidly inactivate other trypsin-like proteases. At the same time, however, the size of the central pore imposes a uniquely narrow substrate specificity; only flexible peptidergic substrates that can thread through this pore such as neuropeptides and elastase-induced kininogen fragments, and proteinaceous substrates and cell surface receptors with a cleavable surface loop of > 20 Å, can be processed or destroyed.

Tryptase should therefore not be considered as a classic digestive enzyme such as trypsin or neutrophil elastase but rather as an enzyme with two jobs: to act as a (neuro)peptidase and as an initiator that prevails enzymatically active in tissues and starts various processes by cleaving a few key proteins, by activating the zymogens of other proteases, and by stimulating cells via PAR-2 or other mechanisms. Obviously, the consequences and their permanence largely depend on the initial presence and on the resynthesis of these key substrates as they can be activated only once. Both presence and resynthesis are likely to be tissue and species dependent and at least in part require the stimulation of their expression by inflammatory mediators, which may explain differences in the responses to tryptases seen, for example, in the proliferation of fibroblasts from skin and lung and the contraction of normal and sensitized human bronchi.

To untangle these complex and often indirect actions of tryptases and to validate their relevance for human disease, studies are currently underway in diverse in vitro systems and in animal models. However, tryptases show marked species differences; in particular, human tryptases appear to be uniquely resistant to Kunitz-type inhibitors, which are endogenous inhibitors of the tryptases of most other species. This resistance is likely to reflect species variations in the architecture of the active tetramer or the structure of the active centers and thus to be indicative for differences in substrate specificity. Therefore, care should be taken in extrapolating results obtained in animals to human (patho)physiology and in interpreting data obtained with inhibitors directed against human tryptases in other species as they may be considerably less active and less specific than expected.