Allosteric Control of βII-Tryptase by a Redox Active Disulfide Bond*

Background: The S1A serine proteases function in many biological processes and contain a conserved disulfide bond near the active site. Results: This disulfide in the S1A mast cell protease, βII-tryptase, exists in oxidized and reduced states in the enzyme, which influences the specificity and efficiency of the enzyme. Conclusion: βII-Tryptase is regulated by an allosteric disulfide bond. Significance: Other S1A proteases are likely similarly regulated. The S1A serine proteases function in many key biological processes such as development, immunity, and blood coagulation. S1A proteases contain a highly conserved disulfide bond (Cys191–Cys220 in chymotrypsin numbering) that links two β-loop structures that define the rim of the active site pocket. Mast cell βII-tryptase is a S1A protease that is associated with pathological inflammation. In this study, we have found that the conserved disulfide bond (Cys220–Cys248 in βII-tryptase) exists in oxidized and reduced states in the enzyme stored and secreted by mast cells. The disulfide bond has a standard redox potential of −301 mV and is stoichiometrically reduced by the inflammatory mediator, thioredoxin, with a rate constant of 350 m−1 s−1. The oxidized and reduced enzymes have different substrate specificity and catalytic efficiency for hydrolysis of both small and macromolecular substrates. These observations indicate that βII-tryptase activity is post-translationally regulated by an allosteric disulfide bond. It is likely that other S1A serine proteases are similarly regulated.

Over 3% of human genes encode for proteolytic enzymes, and approximately one-quarter of these are serine proteases (1). The name stems from the nucleophilic serine in the active site of the enzyme, which attacks the carbonyl moiety of the substrate peptide bond. The serine proteases function in most biological processes, including development, fertilization, immunity, blood coagulation, digestion, and apoptosis. The majority of serine proteases have a trypsin fold, which is defined by a two ␤-barrel architecture, and trypsin-like substrate specificity, which is a preference for Arg or Lys side chains at the P1 position in the substrate (2). One class of the trypsin-like serine proteases are the mast cell tryptases.
Activated mast cells undergo degranulation, releasing tryptases, proteoglycans, and histamine into the extracellular space, promoting inflammation. Tryptases are one of the most abundant proteins stored in human mast cells (ϳ90% of the total protein in secretory granules), which contribute to inflammatory processes in an incompletely understood manner (3).
Among the identified forms of human tryptase, the ␤ form is by far the most abundant in mast cells (3). The polymorphic TPSAB1 and TPSAB2 gene loci on chromosome 16p13.3 encode four tryptase isoforms: ␣ and three nearly identical forms of ␤ (␤I, ␤II, and ␤III). The ␣-tryptase isoform is inactive and is genetically absent in many human subjects (4), whereas the ␤-tryptases are proteolytically active, with ␤II predominant in the lung and ␤I abundant in the skin (5,6). Little is known about the ␤III form. Because the individual ␤-tryptase isoforms can be difficult to differentiate in vivo, they are often referred to generically as ␤-tryptase. ␤-Tryptase is involved in many acute and chronic inflammatory processes. This includes promotion of tissue remodeling, fibrosis, airway hyperreactivity, and bronchoconstriction (reviewed in Ref. 7). Increased levels of ␤-tryptase are commonly found in the bronchoalveolar lavage fluid of patients with allergic rhinitis (8) and atopic asthma (9). Increased levels of ␤-tryptase are also found in the joints of patients with inflammatory arthritis (10), whereas tryptase knock-out mice have significantly reduced joint disease during induction of inflammatory arthritis models (11).
Despite the growing correlation between ␤-tryptase and inflammatory responses, the actual biological targets of tryptase are relatively unknown. ␤-Tryptase has been proposed to cleave fibronectin, fibrinogen, protease-activated receptor 2, and kininogen and is also thought to activate matrix metalloproteinase-3 and pro-urokinase by discrete proteolysis (reviewed in Ref. 3). Also, no endogenous inhibitors of ␤-tryptase have been identified as yet. This poses puzzling questions as to how ␤-tryptase activity is regulated in vivo.
We have observed that the ␤II-tryptase Cys 220 -Cys 248 disulfide bond exists in oxidized and reduced states in the enzyme stored and secreted by mast cells. The disulfide bond is reduced by the inflammatory mediator, thioredoxin. Importantly, the oxidized and reduced isoforms have different specificity and catalytic efficiency for hydrolysis of both small peptide and macromolecular protein substrates. These observations indicate that ␤II-tryptase activity is allosterically regulated by a redox active disulfide bond. The disulfide bond is highly conserved in S1A serine proteases, which suggests that other enzymes in this class will be similarly regulated.

EXPERIMENTAL PROCEDURES
Collection of Secreted Human ␤-Tryptase-The human leukemia mast cell line HMC-1 was kindly provided by Dr. J. H. Butterfield (Mayo Clinic, Rochester, MN). Cells were thawed from liquid nitrogen and grown in flasks containing Iscove's modified Dulbecco's medium supplemented with 10% FBS, 2 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 0.01% ␣-thioglycerol. The cells were split and resuspended at 1 ϫ 10 6 cells/ml into Iscove's modified Dulbecco's medium supplemented with 2 mM L-glutamine, 20 units/ml penicillin, 20 g/ml streptomycin, and no serum. The cells were placed at 37°C for 24 h before collection of the conditioned medium was collected by centrifugation and concentrated using Amicon Ultra 10-kDa cut-off centrifugal filters. For experiments where media supernatant and cell lysates were collected, ␣-thioglycerol and FBS were excluded from the media for 24 h prior to harvesting.
Collection of Murine Bone Marrow Mast Cells-The femurs and tibias of both wild type C57BL/6 and mouse mast cell protease 6 (mMCP-6 Ϫ/Ϫ ) 2 null C57BL/6 mice (12) were harvested and placed in sterile RPMI 1640 medium supplemented with penicillin and streptomycin, after humane sacrifice of the mice. After removing extraneous tissue from the bone, the ends of the bones were removed with a scalpel, and 5 ml of enriched medium (RPMI 1640 medium supplemented with 10% FBS, 2 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, 0.1 mM minimal essential medium nonessential amino acid solution, and 50 M 2-mercaptoethanol) was injected into the core of the bone using a 25-gauge needle, expelling the bone marrow stem cells. The cells were cultured 50% enriched medium and 50% Wehi-3b conditioned medium supplemented with 50 units/ml of recombinant mouse IL-3. Media was changed weekly until Ͼ4 weeks post-harvest, when the presence of mouse bone marrow mast cells was confirmed by preparing cytospins and toluidine blue staining.
Recombinant ␤II-tryptase (Promega) was produced in Pichia pastoris cells and purified from the conditioned medium. ␤II-Tryptase (1 M) was incubated without or with thioredoxin (0.5-1 M) and then labeled with PEG-maleimide for 1 h at room temperature. Human thioredoxin-1 was produced as previously described (13,14). Just before use, the active site disulfide bond of the oxidoreductase was reduced using tris-(2-carboxyethyl)phosphine-agarose (Thermo Scientific) for 30 min at room temperature before desalting the thioredoxin into buffer using a Zeba spin desalting column (Thermo Scientific). The ␤II-tryptase was resolved on SDS-PAGE, transferred to PVDF membrane, and blotted with an anti-tryptase antibody (antitryptase clone AA1, Dako; or anti-tryptase, clone G3, Millipore) and secondary goat anti-mouse peroxidase antibody.
Identification of the Redox Active Disulfide Bond in ␤II-Tryptase Using Mass Spectrometry-␤II-Tryptase (1 M) was incubated without or with reduced thioredoxin (0.5 M) for 1 h at 30°C in phosphate-buffered saline. We employed mass spectrometry with differential cysteine labeling to identify the redox active disulfide bond in the enzyme (15). Briefly, any free thiols in the samples were labeled with iodoacetamide before SDS-PAGE. Gels were stained with SYPRO Ruby (Invitrogen), and bands corresponding to tryptase were excised, destained, dried, and incubated with 100 mM DTT and then washed and incubated with 50 mM methyl methanethiolsulfonate (Sigma-Aldrich). The slices were washed and dried before digestion of proteins with 20 ng/l of trypsin (Promega) or 12 ng/l of chymotrypsin (Roche Applied Science) in 25 mM NH 4 CO 2 overnight at 25°C. Peptides were eluted from the slices with 5% formic acid, 50% acetonitrile and separated by liquid chromatography on an Ultimate 3000 HPLC (Dionex, Amsterdam, The Netherlands). Samples (2.5 l) were concentrated and desalted onto a micro C18 precolumn (500 m ϫ 2 mm; Michrom Bioresources, Auburn, CA) with H 2 O:CH 3 CN (98:2, 0.05% trifluoroacetic acid) at 15 l/min. After a 4-min wash, the precolumn was switched (Valco 10 port valve; Dionex) into line with a fritless nano column (75 m ϫ ϳ10 cm) containing C18 medium (5 , 200 Å Magic; Michrom) manufactured as described (16). Peptides were eluted using a linear gradient of H 2 O:CH 3 CN (98:2, 0.1% formic acid) to H 2 O:CH 3 CN (64:36, 0.1% formic acid) at 250 nl/min over 30 min. High voltage (2000 V) was applied to low volume tee (Upchurch Scientific), and the column tip was positioned ϳ0.5 cm from the heated capillary (280°C) of an Orbitrap Velos (Thermo Electron, Bremen, Germany) mass spectrometer. Positive ions were generated by electrospray, and the Orbitrap operated in data-dependent acquisition mode.
A survey scan m/z 350 -1750 was acquired in the Orbitrap (resolution ϭ 30,000 at m/z 400, with an accumulation target value of 1,000,000 ions) with lockmass enabled. Up to the 10 most abundant ions (Ͼ5,000 counts) with charge states of Ͼϩ2 were sequentially isolated and fragmented within the linear ion trap using collisionally induced dissociation with an activation q of 0.25 and an activation time of 30 ms at a target value of 30,000 ions. The m/z ratios selected for MS/MS were dynamically excluded for 30 s. MS data were searched using Mascot (V2.2; Matrix Science) against the nonredundant database from the National Center for Biotechnology Information. Search parameters were: precursor tolerance, 6 ppm; and product ion tolerances, Ϯ0.6 Da. Methionine oxidation, Cys-methyldisulfide, and Cys-carboxyamidomethyl were selected as variable modifications with full tryptic and chymotryptic cleavage of up to five missed cleavages. To determine the extent of alkylated cysteine residues in tryptase, the relative ion abundance of peptides containing Cys-carboxyamidomethyl and Cys-methyldisulfide was used. To calculate ion abundance of peptides, extracted ion chromatograms were generated using the XCalibur Qual Browser software (v2.1.0; Thermo Scientific). The area was calculated using the automated peak detection function built into the software. The ratio of carboxyamidomethyl to methyldisulfide (the methyl methanethiolsulfonate adduct) labeling represents the fraction of the cysteine in the population that is in the reduced state.
Chemical Properties of the ␤II-Tryptase Cys 220 -Cys 248 Disulfide Bond-Recombinant ␤II-tryptase (1 M) was incubated in argon-flushed phosphate-buffered saline containing 0.1 mM EDTA, 0.5 mM DTT ox (oxidized DTT or 4,5-dihydroxy-1,2dithiane; Sigma) and various concentrations of DTT red (reduced DTT; dithiothreitol; Sigma) for 18 h at room temperature to allow equilibrium to be reached. Microcentrifuge tubes were flushed with argon prior to sealing to prevent oxidation during the incubation period. Reactions were quenched with 2 mM 3-(N-maleimido-propionyl) biocytin (MPB; Invitrogen) for 60 min at 22°C and unreacted MPB quenched with glutathione (4 mM) for 60 min at 22°C. Samples were resolved on SDS-PAGE as described above, protein-transferred to PVDF membrane, and blotted with 1:10,000 dilution of streptavidin-peroxidase to detect the MPB label. Band intensity was quantified using ImageJ version 1.43u software. The fraction of reduced Cys 220 -Cys 248 disulfide bond was measured from the intensity of labeling with MPB with reference to labeling of fully reduced disulfide bond. The results were expressed as the ratio of reduced to oxidized protein and fitted to Equation 1, where R is the fraction of reduced protein at equilibrium, A is the fraction of reduced ␤II-tryptase in the native recombinant protein, and K eq is the equilibrium constant. The standard redox potential (E 0 Ј) of the Cys 220 -Cys 248 bond was calculated using the Nernst equation (Equation 2), using a value of Ϫ307 mV for the redox potential of the DTT disulfide bond (17). The rate of reduction by thioredoxin and rate of air oxidation of the Cys 220 -Cys 248 disulfide were measured by incubating 1 M recombinant ␤II-tryptase and 10 or 1 M thioredoxin at 37°C in argon-flushed phosphate-buffered saline containing 0.1 mM EDTA. Aliquots were removed at specified time points, and free thiols were immediately labeled with 200 M MPB for 1 h before quenching unreacted MPB with 400 M glutathione.
For the rate of reduction of the ␤-tryptase disulfide bond, the data were fit to one-phase exponential association using Prism (GraphPad). For the rate of oxidation of ␤-tryptase, the data were fit to one-phase exponential decay using Prism (GraphPad).
Hydrolysis of Fibronectin and Fibrinogen by Oxidized and Reduced ␤II-Tryptase-␤II-Tryptase was prepared as described for the chromogenic substrate assay. Any unpaired cysteines in the thioredoxin were alkylated with 25 mM iodoacetamide for 1 h to prevent thioredoxin from reducing disulfide bonds in fibronectin or fibrinogen. Alkylation of reduced ␤II-tryptase with iodoacetamide had no effect on activity in the chromogenic substrate assay (data not shown). Oxidized or reduced ␤II-tryptase (20 nM) was incubated with fibronectin (Sigma-Aldrich; 10 mg/ml) or fibrinogen (Sigma-Aldrich; 5 mg/ml) at room temperature for 21 h in either 50 mM HEPES, pH 7.4 buffer containing 30 mM NaCl, 10% glycerol, and 1 mg/ml PEG 6000 or 50 mM MES, pH 6.2 buffer containing 30 mM NaCl, 10% glycerol, and 1 mg/ml PEG 6000. Aliquots were removed at various time points, and cleavage of fibronectin was measured by resolving the reactions on SDS-PAGE and staining with silver (Thermo Scientific). ␤II-Tryptase is not visible in the stained gels because of the low concentration, and thioredoxin (17.2 kDa) runs significantly below the fibronectin (30 -50 kDa) and fibrinogen (200 -400 kDa) peptides and does not interfere with interpretation of protein cleavage.

Native Human ␤-Tryptase and the Equivalent Murine
Enzyme Contain a Reduced Disulfide Bond-The human mast cell line, HMC-1, produces, stores, and secretes ␤-tryptase (18). We confirmed secretion of the active enzyme by chromogenic substrate assays of conditioned medium from serum-deprived cells (not shown). Conditioned medium and lysate of HMC-1 cells was incubated with a branched polyethylene glycolmaleimide (three 12-PEG units). Maleimides specifically react with thiolates at neutral pH (19). This alkylator adds a mass of ϳ2.5 kDa per labeled thiol, and incorporation is measured by increase in molecular mass by SDS-PAGE and immunoblotting of the protein of interest. Both the secreted (Fig. 1A) and lysate (Fig. 1B) ␤-tryptase incorporated up to two molecules of PEGmaleimide, indicating the presence of two unpaired cysteine thiols in the enzyme. The crystal structure of ␤II-tryptase shows that all eight cysteines in each monomer participate in a disulfide bond (20). This result indicates that ␤-tryptase stored and secreted by HMC-1 cells contains a reduced disulfide bond. The majority of the ␤-tryptase was labeled with PEG-maleimide, indicating that most of the secreted enzyme was in the reduced state. This finding suggested that ␤-tryptase is post-translationally regulated by a redox active disulfide bond.
The murine equivalent of ␤-tryptase is mMCP-6, which shares 77% sequence identity to human ␤II-tryptase and has similar substrate specificity (3,21). mMCP-6 contains the same conserved disulfide bonds as human ␤II-tryptase. Bone marrow mast cells were harvested from wild type and mMCP-6 null C57BL/6 mice, and the redox state of mMCP-6 in bone marrow mast cell lysate was measured by labeling with PEG-maleimide. The majority of the mMCP-6 is reduced (Fig. 1C), which is consistent with the nature of the enzyme stored and secreted by human mast cells (Fig.  1, A and B). No bands were observed in the mMCP-6 null mice, indicating that the antibody is specific for mMCP-6.
The Cys 220 -Cys 248 Disulfide Bond Is Reduced in ␤II-Tryptase-The identity of the redox active disulfide bond in ␤II-tryptase was determined using recombinant enzyme produced in yeast cells and mass spectrometry. Differential cysteine labeling allows identification of the individual cysteines that are unpaired and those that participate in a disulfide bond. The unpaired cysteines were alkylated with iodoacetamide and the disulfide-bonded cysteines with methyl methanethiolsulfonate following reduction with DTT. The ratio of carboxyamidomethyl (the iodoacetamide adduct) to methyldisulfide (the methyl methanethiolsulfonate adduct) labeling represents the fraction of the cysteine in the population that is in the reduced state. We were able to detect seven of the eight cysteines in ␤II-tryptase and so could measure the redox state of all four disulfide bonds in the enzyme.
Recombinant ␤II-tryptase appears as two bands of ϳ34 and 30 kDa on SDS-PAGE because of different glycosylation patterns (3) (Fig. 2A). The Cys 220 -Cys 248 disulfide bond was reduced in ϳ30% of ␤II-tryptase molecules in the preparation, which increased to ϳ80% following incubation of 1 M enzyme with 0.5 M thioredoxin ( Fig. 2B and Table 1). Thioredoxin is a relevant protein reductant in the cell and in inflammatory environments where ␤-tryptase is found. Thioredoxin gene expression is up-regulated, and the protein is secreted by immune cells during inflammation, leading to a high local concentration and also elevated blood levels (22). High serum levels of thioredoxin have been observed in patients with asthma, rheumatoid arthritis, and heart failure (23,24), and serum levels correlate with disease activity in rheumatoid arthritis (25,26). The ϳ50% increase in reduction of the Cys 220 -Cys 248 disulfide bond in an incubation of  0.5 mol of thioredoxin/mol of ␤II-tryptase indicates that the reaction is stoichiometric. Thioredoxin did not influence the redox state of the other three disulfide bonds in ␤II-tryptase.
To test the veracity of the mass spectrometry analysis, the reduced ␤II-tryptase was labeled with PEG-maleimide. A fraction of the ␤II-tryptase incorporated two molecules of PEGmaleimide, in accordance with partial reduction of the Cys 220 -Cys 248 disulfide bond, and incubation with a molar equivalent of thioredoxin fully reduced the enzyme, which is consistent with the mass spectrometry analysis (Fig. 2C).
Chemical Properties of the ␤II-Tryptase Cys 220 -Cys 248 Disulfide Bond-The redox potential and rate of reduction of the ␤II-tryptase Cys 220 -Cys 248 disulfide bond by thioredoxin were measured. The standard redox potential of the Cys 220 -Cys 248 disulfide bond was determined using oxidized and reduced DTT. The MPB labeling of fully reduced ␤II-tryptase was used to calculate the ratio of reduced to oxidized ␤II-tryptase as a function of the ratio of reduced to oxidized DTT (Fig. 3A). A standard redox potential of Ϫ301 mV was calculated (Fig. 3B). Analysis of the DTT titration results indicated that 24% of the ␤II-tryptase

-Cys 248 disulfide bond in ␤II-tryptase is reduced in a fraction of the molecules, and the oxidized fraction can be reduced by thioredoxin
The fraction of peptides containing Cys 220 or Cys 248 that labeled with IAM increased upon addition of thioredoxin (bold face type), indicating reduction of the Cys 220 -Cys 248 disulfide bond. Thioredoxin selectively reduced this bond because there was no appreciable change in the IAM labeling of the other six cysteines in the protease (Cys 59 Fig. 3B). This result is consistent with the fraction of reduced ␤II-tryptase molecules in the preparation measured by mass spectrometry (Table 1) and PEG-maleimide labeling (Fig. 2C).
The rate of reduction of the ␤II-tryptase Cys 220 -Cys 248 disulfide bond by thioredoxin was measured under pseudo-first order conditions (Fig. 3C). A k obs of 0.21 Ϯ 0.05 min Ϫ1 was calculated, which equates to a second order rate constant of 350 M Ϫ1 s Ϫ1 (Fig. 3D). The rate of oxidation of the Cys 220 and Cys 248 dithiol was calculated from reactions of equimolar ␤IItryptase and thioredoxin (Fig. 3E). The bond was reduced after ϳ30 min, and the reduced thioredoxin was consumed in the reaction. Oxidation of the Cys 220 and Cys 248 thiols in the enzyme exposed to ambient air was monitored with time. The k obs for the oxidation was 0.006 Ϯ 0.002 min Ϫ1 (Fig. 3F,  inset), which is 35-fold slower than the rate of reduction of the bond by a 10-fold molar excess of thioredoxin.
Reduction of the Cys 220 -Cys 248 ␤II-Tryptase Disulfide Bond Does Not Influence Tetramer Formation-␤-Tryptase forms a ϳ134-kDa homotetramer that is stored in mast cells as the fully active form rather than as a pro-enzyme, both of which are unique traits among the S1A proteases. The crystal structure of ␤-tryptase is a tetramer composed of four quasiequivalent monomers arranged in a square flat ring (20). In the tetramer, all of the active sites face into the central pore (20), limiting both substrate and inhibitor access (27). Whether or not the ␤-tryptase monomer possesses enzyme activity is debated.
The effect of thioredoxin reduction of the Cys 220 -Cys 248 disulfide bond on tetramer formation was assessed using gelatin zymography. The zymogram separates the monomers from tetramers in mild denaturing conditions, and enzyme activity is detected from degradation of the gelatin embedded in the gel. Only the tetrameric ␤II-tryptase is active in this assay, and thioredoxin treatment had no obvious effect on tetramer formation or gelatinase activity (Fig. 4).
Oxidized and Reduced ␤II-Tryptase Cleave a Small Peptide Substrate with Different Efficiency-The Cys 220 -Cys 248 disulfide bond links two ␤-loop structures that define the rim of the active site pocket of ␤II-tryptase (20). Structural changes to this region have previously been shown to alter enzymatic activity of the tryptases (28 -30). Thus, it was anticipated that cleavage of this disulfide bond would alter the S1 pocket region and influence the catalytic properties of the enzyme. This was tested using the chromogenic substrate, pyro-Glu-Pro-Arg-p-nitroanilide.
We tested the activity of ␤II-tryptase at neutral pH (7.4) and at pH 6.2 because the intracellular pH of mast cells and granules, where ␤II-tryptase is stored, has been found to range from 5.2 to 6.5 (31)(32)(33). The enzyme also functions in acidic inflammatory environments such as the synovial fluid of rheumatoid arthritic joints (34,35) and the airways of asthmatics (36,37). Synovial fluid acidosis positively correlates with radiological knee joint destruction in patients with rheumatoid arthritis (35), and the average pH of airway vapor in patients with acute asthma is 5.23 (36). Acidic pH has also been shown to alter ␤-tryptase activity toward various substrates (38 -41), making it an important variable to test. Because ionic strength has been found to influence ␤-tryptase activity and the equilibrium between monomer and tetramer (38 -41), we also selected two different ionic strengths for our assays. The catalytic efficiency of oxidized versus reduced ␤II-tryptase was therefore tested at both neutral (7.4) and acidic (pH 6.2) pH and also at physiological (I 150) and low (I 50) ionic strength.
There was a small difference in the maximal velocity (V max ) of the oxidized versus reduced enzyme at neutral pH and physiological ionic strength (Fig. 5A) but no difference at low ionic strength (Fig. 5B). The difference in V max was more pronounced at pH 6.2 at either ionic strength (Fig. 5, C and D). There was no significant change in K m at either pH or ionic strength (Table 2).
Oxidized and Reduced ␤II-Tryptase Cleave Macromolecular Substrates with Different Efficiency and Specificity-The redox state of the ␤-tryptase Cys 220 -Cys 248 disulfide bond also influences cleavage of the macromolecular substrates fibronectin and fibrinogen. ␤II-Tryptase cleaves the large substrate fibronectin into multiple, smaller peptides ranging from ϳ30 -50 kDa in size, which can be visualized using SDS-PAGE and silver staining. ␤II-Tryptase has also been found to cleave fibrinogen, which can be visualized as a shift from a molecular mass of ϳ340 kDa to a slightly smaller protein ϳ238 -268 kDa on SDS-PAGE (42).
Both the rate of cleavage of fibronectin and the sites of cleavage differ for the oxidized and reduced enzyme, which is more pronounced at acidic pH. ␤II-Tryptase cleaves fibronectin more rapidly and at more sites at pH 6.2 (Fig. 6B) than pH 7.4 (Fig. 6A), and there is significantly less fibronectin cleavage by reduced ␤II-tryptase at both pH. Incubation of ␤II-tryptase with redox-inactive thioredoxin mimicked the results in the absence of thioredoxin. This experiment controlled for the possibility that thioredoxin was acting as a competitive substrate for ␤II-tryptase in the assay and so biasing the results. Reduced ␤II-tryptase was also less efficient at cleaving fibrinogen at neutral pH (Fig. 6C), but both forms of the enzymes were equally  NOVEMBER 29, 2013 • VOLUME 288 • NUMBER 48 efficient at pH 6.2 (Fig. 6D). ␤II-Tryptase has been observed to cleave fibrinogen at a faster rate in acidic pH (43).

DISCUSSION
The tryptases belong to the serine protease subfamily S1A, a group of trypsin-like, neutral serine proteases categorized by their function: coagulation and immunity, digestion, tryp-tase, kallikrein, granzymes, and matriptase (44). A global alignment of 79 human S1A proteases found that the Cys 191 -Cys 220 disulfide bond (chymotrypsin numbering) is conserved in all forms of human tryptase and in all S1A serine proteases, except for the cathepsins and granzymes, which lack the disulfide bond (44). There is a small difference in the maximal velocity (V max ) of the oxidized versus reduced enzyme at neutral pH and physiological ionic strength (A), but no difference at low ionic strength (D). The difference in V max was more pronounced at pH 6.2 at either ionic strength (B and C). There was no significant change in the Michaelis constant (K m ) at either pH or ionic strength ( Table 2). The data points and errors are the means Ϯ S.E. of three determinations, and the solid lines represent the nonlinear least squares fit of the data to the Michaelis-Menten competitive substrate inhibition model.

TABLE 2 Kinetic parameters for hydrolysis of pyro-Glu-Pro-Arg-p-nitroanilide by oxidized and reduced ␤II-tryptase
The values were determined in experiments of the type shown in Fig. 4. Errors are Ϯ S.E. The structural/functional role of the Cys 191 -Cys 220 disulfide bond has been investigated in some S1A proteases. In coagulation factor XII, the disulfide bond (Cys 540 -Cys 571 in factor XII) is essential for activity. In the rare blood coagulation disorder known as Hageman trait, Cys 571 is mutated to a serine in factor XII, which ablates the disulfide bond and inactivates the enzyme (45). Elimination of the disulfide bond in the blood coagulation protease thrombin (Cys 540 -Cys 571 in thrombin) results in a ϳ100-fold loss of activity toward peptide substrates (46). The disulfide bond in trypsin (Cys 173 -Cys 197 in trypsin) is also important for activity, although there are conflicting reports on the nature of the change in activity upon elimination of the bond. One study found that ablation of the disulfide bond has no effect on the function, structure, K m , or stability of the enzyme but does affect the k cat for hydrolysis of peptide substrates (47). Another study reported that eliminating the disulfide bond increases the K m for peptide substrates and shifts substrate specificity from trypsin-like to chymotrypsin-like (48).

Conditions
Some disulfide bonds, like some peptide bonds, are cleaved in the mature protein. These disulfide bonds are known as allosteric disulfides (49,50). The bonds are reductively cleaved by oxidoreductases or by thiol-disulfide exchange to control protein function (51,52). Reduction of an allosteric disulfide bond results in a functional change at another site in the protein, and changes in ligand binding, substrate hydrolysis, proteolysis, or oligomer formation have been documented (reviewed in Refs. 49 and 50). Approximately 30 mammalian, plant, bacterial, or viral proteins have been found to contain allosteric disulfides so far, and a few have been linked to unwanted thrombosis (␤2glycoprotein I and factor XI) and impaired blood pressure regulation (angiotensinogen) in humans. The findings reported herein indicate that the ␤-tryptase Cys 220 -Cys 248 disulfide is an allosteric bond.
A pattern is appearing in the configurations of allosteric disulfides and the secondary structures that they link (49,50). The disulfide bond, or cystine residue, is defined by five angles and different combinations of these angles define 20 possible cystine configurations (51,52). Three of the 20 disulfide bond configurations are emerging as allosteric configurations: the ϪRHStaple, ϪLHHook and Ϫ/ϩRHHook bonds (49). The most common secondary structures linked by allosteric disulfides are ␤-strands and/or ␤-loops (49). The ␤-tryptase Cys 220 -Cys 248 disulfide exists in two different configurations in the crystal structure of the tetramer (20), ϪRHHook and ϪRHSpiral, and the bond links two extended ␤-loops (53). The disulfide is also exposed to solvent (Fig. 7), in particular the Cys 220 residue, FIGURE 6. Oxidized and reduced ␤II-tryptase cleave macromolecular substrates with different efficiency and specificity. A and B, oxidized or reduced ␤-tryptase was incubated with fibronectin, and the cleavage products were resolved on SDS-PAGE and silver-stained. As a control for the possible effects of thioredoxin on the substrate, ␤II-tryptase was also incubated with a redox-inactive form of thioredoxin (inactive trx). Oxidized and reduced ␤II-tryptase produce different fibronectin peptides at both pH 7.4 (A) and 6.2 (B). Intact fibronectin can be seen near the top of the gel, whereas cleaved portions of fibronectin are ϳ30 -40 kDa in size. C and D, oxidized or reduced ␤-tryptase was incubated with fibrinogen, and the cleavage products were resolved on SDS-PAGE and silver-stained. Oxidized but not reduced ␤II-tryptase cleaves fibrinogen at pH 7.4. There is little to no difference in the fibrinogen cleavage products produced by the oxidized and reduced forms of ␤II-tryptase at pH 6.2. which is likely an important factor in the accessibility of this bond to thioredoxin.
Reduction of the Cys 220 -Cys 248 disulfide bond is predicted to alter the shape of the active site pocket, which would explain the altered substrate specificity and catalytic efficiency of the reduced enzyme. The two ␤-loops linked by the Cys 220 -Cys 248 disulfide bond shape the S1 substrate pocket that binds the residue immediately N-terminal of the target peptide bond (the P1 position). The importance of the S1 substrate pocket is demonstrated in the tryptase isoform, ␣-tryptase. Despite ␣-tryptase being 92% identical to ␤II-tryptase, it is completely inactive because of a single amino acid change that distorts the S1 pocket and blocks substrate binding (28 -30). Distortion of the ␤II-tryptase S1 pocket is likely to be occurring when the Cys 220 -Cys 248 disulfide bond is reduced, leading to the changes we observed.
Thioredoxin is a potentially relevant ␤II-tryptase reductant in vivo but may not be the only one. Both ␤II-tryptase and thioredoxin are released into the extracellular space during physiological inflammation and are found in synovial fluid in rheumatoid arthritis and bronchoalveolar fluid in asthma (10,11,(23)(24)(25)(26)54). Thioredoxin concentrations up to 100 nM have been reported in synovial fluid of rheumatoid arthritic joints (25,54). Other oxidoreductases are secreted during inflammation and other conditions, including protein-disulfide isomerase (55) and ␥ interferon-inducible lysosomal thiolreductase (GILT) (56,57). It is possible that protein-disulfide isomerase and GILT play a role in regulating ␤II-tryptase through redox control of the Cys 220 -Cys 248 allosteric disulfide bond. Interestingly, GILT is optimally active in acidic pH (58), which is an environment in which ␤-tryptase functions (34 -37).
The finding that ␤II-tryptase is controlled by an allosteric disulfide bond is informative, since very little is known about the regulation of ␤II-tryptase activity because no endogenous inhibitors have been identified. ␤-Tryptase forms a homotetramer held together by polar and hydrophobic interactions that is further stabilized by heparin proteoglycans (20,38,39,41,59,60). Our results indicate that the disulfide bond is reduced in the tetramer in the absence of heparin. How heparin influences the equilibrium between oxidized and reduced ␤IItryptase and the activities of these two forms will be the subject of future studies.
There is precedence for allosteric regulation of a S1A protease. The blood coagulation protease, thrombin, exists in three forms: a low activity sodium-free form, a high activity sodiumbound form, and a form that is inactive and unable to bind sodium (2,61). Conversion of the inactive to highly active form occurs very rapidly, and modulation of this equilibrium is thought to regulate thrombin activity in vivo. Redox control of the conserved Cys 191 -Cys 220 disulfide bond in S1A proteases is possibly a general mechanism of allosteric regulation of this family of enzymes, which is predicted to influence their function in a number of important biological processes.