Proton Bridges in Thrombin-Catalyzed Hydrolysis of Tripeptide 4-Nitroanilides
Linnea Patt
Senior Comprehensive Paper
Catholic University of America
Fall 2003
Abstract:
The purpose of this study was to identify the number of protonic sites and characterize their role in catalysis by thrombin in the hydrolysis of chromogenic substrates that contain some of the P1-P3 specificity sites. For N-Benzoyl-Phe-Val-Arg-pNA HCl, the solvent isotope effects were found to be: at 25.0 0.1 oC, 2.8 0.2 for the Vmax calculations, and 1.2 0.2 for the Vmax/KM calculations; and at 37.0 0.2 oC and 1.6 0.1 for the kobs calculations, and 2.21 0.03 for the v calculations. The fractionation factors were found to be: at 25.0 0.1 oC, 0.37 0.07 for the Vmax calculations, and 0.3 0.3; 2.3 1.8 for the Vmax/KM calculations, and at 37.0 0.2 oC, 0.58 0.09 for the kobs calculations, and 0.39 0.04; 1.2 0.1 for the v calculations. The data are most consistent with having a single proton participate in catalysis.
Introduction:
Blood clotting is an important physiological process in vertebrate animals that works as a defense mechanism to prevent blood loss. Both the formation and breakdown of clots must occur rapidly in a highly regulated environment to ensure proper blood flow and function. This regulation is achieved by a highly complex system called an enzymatic cascade.
Enzymatic cascades are efficient processes in biological systems that ensure a rapid response to an outside trigger. The response begins with an escalation of biological function, such as the increase in concentration and activity of certain enzymes. Trauma and/or rupture to a blood vessel initiates an intrinsic and/or extrinsic pathway that begins the complex cascade of blood clotting. Once the cascade is started, many zymogens, or proenzymes, are activated and the signal is amplified greatly. In each highly regulated step of the process, proteins are cleaved and activated in order to cleave and activate other proteins in the cascade. Thirteen proteins are activated in the cascade culminating in the generation of thrombin,1 as seen in Figure 1.
Thrombin is considered the key enzyme in the clotting cascade because it not only is involved in the last step of the cascade, but it also works to regulate the cascade. To perform its regulatory functions, there are two forms of thrombin: fast and slow.2 At adequately high concentrations of Na+ (0.2M), it performs the last step in the process of conversion of fibrinogen to fibrin monomers. Fibrinogen is made up of three globular units connected by six chains. The tails of four chains are cut, which allows fibrin monomers to assemble into a fibrous array called fibrin, which traps platelets and other blood proteins to form the blood clot.1 Ion fluxes in the blood change throughout the course of blood clotting and eventually the Na+ concentration will drop and the Ca2+ concentration will rise. These conditions enhance thrombin’s affinity for binding the modulator protein, thrombomodulin. In the complex with thrombomodulin, thrombin assumes the slow form and modifies Protein C to APC, activating it. When Protein S, a vitamin K dependant cofactor, binds APC it enhances its ability to bind to phospholipid membranes. Then, APC cleaves factors Va and Xa. The deceleration of the production of thrombin and the clotting cascade is a negative feedback loop.3
Figure 1. Blood Clotting Cascade.4
Structure:
The three dimensional structure of thrombin5, along with other serine protease enzymes, evolved remarkably well to accommodate the chemical transformation that it catalyzes: amide bond cleavage. The active sites and specificity sites are the most important components of the enzyme. The motif that has been conserved in serine proteases is called the catalytic triad, which includes a Ser nucleophile, a His general acid/base catalyst and a carboxylate, Asp or Glu, assisting with a proton bridge. These enzymes also have a specificity site called an oxyanion hole, which helps to stabilize the negative charge on the carbonyl oxygen during the reaction. The catalytic triad is often numbered as Ser195, His 57, and Asp102, as in trypsin and chymotrypsin.6
Specificity is extremely important in thrombin activity. When referring to specificty, there is a common nomenclature referring to different residues on the enzyme and substrate commonly used by enzymologists. The residues on the substrate (P) on either side of the secile bond are identified as P1 - P1`, by increasing numbers going away from the sessile bond, Pi` (i = 1, 2, 3…) toward the C terminus and Pi (i = 1, 2, 3…) toward the N terminus. These correspond to similarly identified residues on the enzyme labeled with an S. 6;7
S2 S2`
P3 P2 P1 P1` P2` P3` Substrate
S3 S1 S1` S3`
For example, in thrombin, Asp189 makes up the S15;8 site and is crucial in the recognition of sustrates because it forms an ion pair with an Arg or Lys residue which makes up the P1 site of the substrate. This forms the salt bridge that makes up the S1-P1 interaction and helps maintain high specificity. The Leu59-Asn62 and Leu144-Gly150 insertion loops are thought to be key in providing specificity.
There are several other major elements that make up thrombin. They include: the fibrinogen recognition exosite (anion binding site 1), the heparin binding site (anion binding site 2), and the RGD sequence. The fibrinogen recognition site is relatively far from the active site and is involved in binding fibrinogen, thrombin receptor, fibrin, and thrombomodulin. The heparin binding site is the strongest positively charged patch on -thrombin and is involved in the binding of heparin. It is also suggested that this site is involved in the binding of thrombin to platelets. The RGD sequence is close to the active site and is said to be the platelet binding site.85
Thrombin’s tertiary structure has been studied in great detail and can be defined by simple descriptions. Moving to the qarternary structure, however, thrombin is more difficult to define because of its allostery. It has been found that thrombin has a fast and a slow form, contributing to its role in coagulation and anti-coagulation respectively.9;10 The fast form dominates in high concentrations of Na+ and has a higher specificity for fibrinogen. The slow form dominates in lower concentrations of Na+ and has a higher specificity for protein C. Na+ binds upon fibrinogen binding and is released when protein C binds. The Na+ binding site has been intensely examined and is conserved in many different species, which proves its importance to the function of the enzyme.11 These structural elements all contribute to the effectiveness of thrombin.
Function:
In thrombin-catalyzed reactions, there are three main steps that enable thrombin to cleave the peptide bond, substrate binding, acylation and deacylation.6;12 A schematic of the reaction is shown in Scheme 1 and the elementary steps are shown in Scheme 2.13
Scheme 1.
k1 k2
E + S ES ES` + P`
k-1 k3
E + P``
In the first step, the enzyme (E) binds the substrate (S), an enzyme-substrate complex (ES) is formed. The chemical reactions begin with the nucleophilic attack of Ser195 on the carbonyl carbon of the peptide bond. This attack is promoted by the transfer of a proton from Ser195 to His 57 and by the stabilizing effect of Asp102 on the developing positive charge on His57 through the carboxylate group. A tetrahedral intermediate is produced , which then breaks down when the C-N bond cleaves. This bond cleavage is promoted by the transfer of the proton from His57 to the leaving group. The acylation step, represented by k2, produces an acylenzyme (ES`) along with one of the products (P`), a peptide fragment with a new N terminus. The deacylation step, represented by k3, releases the enzyme along with a second product (P``), a peptide fragment with a new C terminus. Through general-base catalysis by His57, the nucleophilic attack of water on the acyl enzyme causes the hydrolysis of the acyl fragment from thrombin. Both acylation and deacylation involve a possible proton bridge and/or two steps of proton transfer. Experiments have shown that this occurs through proton transfer at the catalytic triad, as in other serine protease enzymes. This process effectively accelerates the reaction by 17-19 orders of magnitude.14;15
Catalysis by thrombin, like all serine proteases, obeys Michaelis-Menten kinetics. If the rates are measured at increasing substrate concentrations, a parabola will be formed that has an asymptote representing Vmax.
v = Vmax[S](2)
KM + [S]
Under steady state conditions, the parameters kcat and KM, the turnover number and the Michaelis constant, respectively, are given by the following equations.
kcat = k2k3;kcat = Vmax(3)
k2 + k3 [Eo]
KM = Ks k3__;Ks = k-1 +k2(4)
k2 + k3 k1
When acylation is the limiting step, k3 > k2, kcat = k2 and KM = Ks. When deacylation is the limiting step, k2 > k3, kcat = k3 and KM = Ks k3/ k2.
The bimolecular encounter of the enzyme and substrate is described by a second order rate constant, kcat / KM. This value is expressed by the following equation.
kcat _ = _____ k1k2k3______(5)
KM k -1k2 + k -1k3 + k2k3
When k3 > k2 or k -1, kcat / KM = k1k2/(k -1+ k2) and when k2 or k3 > k -1, kcat / KM = k1.
The reaction is first order in [S] and second order over all at low substrate concentrations, when [S] < KM. The reaction is zero order in [S] and first order overall at high substrate concentrations, when [S] > KM, showing saturation of the enzyme. Under these conditions, kcat measures mostly the actual transformation of the substrate.
Small Peptide Anilide Substrates:
Natural substrates of serine protease enzymes are not easily detected analytically and are usually quite expensive. However, substrate mimics can be used to study the mode of proton transfer or effect of individual S-P or S’-P’ interactions on the efficiency of catalysis. Substrate mimics are usually short petide chains with a good leaving group releasing an easily identifiable signal molecule. The most widely used leaving group for serine protease studies is a para-nitro-aniline (pNA) for two very practical reasons. First, the cleavage of the nitroanalide-peptide bond catalyzed by the enzyme strongly resembles the natural process and pNA is released in proportion to the enzymatic activity. Second, the absorption maximum of free pNA after cleavage is much different than when it is still attached to the substrate. Before cleavage, the peptide is colorless with an absoption maximum in the UV. As pNA is released, it can be easily monitored at its absorption maximum of 400nm.
The substrates used for these reactions have very small chains that resemble the natural substrates of the enzyme. These substrates, N-Benzoyl-Phe-Val-Arg-pNA and N-D-Phe-Pip-Arg-pNA, are not as selective as the natural substrates, since natural substrates are much larger and interact with different binding pockets and selectivity regions of the enzyme. These substrates do, however, provide the S1/P1, S2/P2, and S3/P3 interactions that occur in natural substrates.
Solvent Isotope Effects and Proton Inventories:16-18
Serine proteases, through the activity of the catalytic triad, function through general acid-base catalysis that involves proton transfer and proton bridges. One question that can be posed regarding this mechanism is how active a role do the proton transfers and bridges play in catalysis. In the rate-determining step of the reaction, the mode of proton interaction can be most accurately measured by solvent isotope effect and partial solvent isotope effect. The theory behind this method explains that rapidly exchanging protonic sites will equilibrate with the solvent and in the presence of heavy water, these protonic sites will be replaced by deuterium. Rate measurements are taken in buffered water, heavy water, and mixtures of the two. Since the catalytic triad contains these rapidly exchanging protonic sites, proton transfer to and from these sites will be replaced by deuterium in the presence of heavy water. Because of its smaller mass-dependent vibrational frequencies, deuterium reacts slower than protium and, therefore, catalytic reactions will proceed slower in deuterium oxide than in protium oxide. Reactions in a mixture of the two will proceed at a rate somewhere between that in water and heavy water. The rate ratios produced in buffered water and heavy water are highly dependent on the atom fraction of D, n, in the solution. The equation, derived by Gross and Bulter, shows the relationship as:
TS RS
Vn = Voi (1n+nTi) /j (1n+nRj) (6)
Where Vn is velocity in a mixed solvent, Vo is velocity in water, n is atom fraction of deuterium, RS is reactant state, R is RS fractionation factor, TS is the transition state, and T is TS fractionation factor. The TS product is over the TS fractional factor and the RS product is over the RS fractionation factor. In essence, the fractionation factors are inverse equilibrium isotope effects, KD/KH, for the exchange between a structural site on RS or TS and a bulk solvent site. Through the least squares fitting procedure, the contributing isotope effects can be determined. Many complex models can be derived from this equation, however, the most common simplifications involve the assumption of a unit fractionation factor of most RSs and that, in most hydrolytic enzymes, there are one or two active-site units that contribute, producing: 1) Vn = Vo(1n+nT) or 2) Vn = Vo(1n+nT1)(1n+nT2), respectively. In order to acquire a correct proton inventory, the pH/pD dependence must be known for the reaction. Then, the reactions need to be performed at a pH plateau in identical H2O and D2O solutions. The models can be applied to the data and the best fit is based on chemical reasoning and statistical analysis.16-18
Goal:
The purpose of this study was to identify the number of protonic sites and characterize their role in catalysis by thrombin in the hydrolysis of chromogenic substrates that contain some of the P1-P3 specificity sites.
Experimental
Solutions: Tris buffer stock solutions were made with 0.0067M Tris(hydroxymethyl)-aminomethane hydrochloride, 0.0133M Tris(hydroxymethyl)-aminomethane, 0.1% PEG, and 0.03M NaCl for both H2O and D2O. Buffer combinations were made using the two stock solutions in the following proportions: 15% D2O/85% H2O, 33%D2O/67% H2O, 50% D2O/50% H2O, 67% D2O/33% H2O, 85% D2O/15% H2O.
N-Benzoyl-Phe-Val-Arg-pNA HCl (Sigma-Aldrich) and N-D-Phe-Pip-Arg-pNA (Diapharma) were dissolved in DMSO for the stock solutions. For Michaelis-Menten analysis, N-Benzoyl-Phe-Val-Arg-pNA stock solution was used at a concentration of 0.01952M and was diluted progressively from 4KM to 0.04KM. For initial rates experiments using N-Benzoyl-Phe-Val-Arg-pNA, a 0.01952M stock solution was used and diluted to 3.2 x 10-4 M in the cuvette and for initial rates studies using N-D-Phe-Pip-Arg-pNA, a 3.2 x 10-3 M stock solution was used and diluted to 6.4 x 10-5M in the cuvette. For progress curve analysis, a 4.88x 10-4 M N-Benzoyl-Phe-Val-Arg-pNA stock solution was used and diluted to 9.76 x 10-6 M in the cuvette.
Enzyme solutions were made with human alpha thrombin purchased from Enzyme Research Laboratories, Inc. with activity of 3181 NIH units/mg. These were diluted with buffer to (5-2) x 10-6 M and, thus were at (5-2) x 10-8 M in the cuvette.
Kinetics: Rate measurements were taken by a Perkin-Elmer Lambda 6 UV/Vis Spectrophotometer taking 1000 data points at 400nm using a program called PECSS. The temperature for Michaelis-Menten experiments was regulated by a Techne Tempette TE-8A circulating water bath and for all of the other experiments, by a Brinkmann MGW Lauda RM-20 circulating water bath. In all kinetic experiments, 970L buffer, 20L substrate stock solution, and 10L enzyme stock solution was used.
Michaelis-Menten experiments: The buffer was inserted into the cuvette and incubated in the cell compartment for 10 minutes. At that point, substrate and enzyme were added and 1000 absorption data points were acquired. This was done with all of the substrate concentrations and the PECSS software was used to calculate the slope (OD/sec) of each run. The data was entered into the computer program Grafit and fitted to a Michaelis-Menten equation, where Vmax and Vmax/KM were calculated. An example is shown in Figure 2.
Initial rates: The buffer was incubated in the cell compartment for 5 minutes; then, enzyme was added and the mixture was incubated for another 10 minutes. Substrate was added and absorption data points were acquired to get the initial rate of the reaction. PECSS was used to calculate the OD/sec.
Figure 2. Michaelis-Menten profile for the thrombin-catalyzed hydrolysis of N-Benzoyl-Phe-Val-Arg-pNA in 0.02M Tris buffer at pH 8.3 containing 0.03M NaCl, 0.1%PEG-4000 and at 25 0.2 oC.
Progress curve analysis: These experiments were set up exactly like the initial rate studies except the reactions were allowed to run to almost completion so the entire curve could be seen. The data was taken and fit to the first-order rate equation using the program Grafit. Rate constants were calculated. An example is shown in Figure 3.
Figure 3. Progress curve of thrombin-catalyzed hydrolysis of N-Benzoyl-Phe-Val-Arg-pNA in 0.02M Tris buffer at pH 8.3 containing 0.03M NaCl, 0.1%PEG-4000 and at 37 0.2 oC.
Data Analysis: For each data set, each calculated value was compared to the value found in the D2O buffer and the rate constant ratios were plotted against the corresponding n value. For example, for the initial rate data, a vn:vD2O ratio was the average of three repeats under a set of conditions, calculated and divided by the average value of the rate of rate constant obtained in D2O, and plotted against n. These curves were fitted to different proton inventory models where solvent isotope effect and TS/S were calculated. The model with the lowest reduced chi2 was plotted. The models are shown in Table 1.
Table 1: Models for fitting proton inventory data.
Information obtainedEquation
TS1 Vn = VH (1 - n + n 1)
TS1, solv. Vn = VH (1 - n + n 1) Sn
2TS1 Vn = VH (1 - n + n )2
2TS1, solv. Vn = VH (1 - n + n )2 Sn
TS1, TS2 Vn = VH (1 - n + n 1)(1 - n + n 2)
TS1, TS2, solv. Vn = VH (1 - n + n 1)(1 - n + n 2) Sn
TS, RSVn = VH (1 - n + n 1)/(1 - n + n R)
Results:
Table 2 shows examples of fitting data to various models to determine the fractionation factors for the thrombin-catalyzed hydrolysis of N-Benzoyl-Phe-Val-Arg-pNA. The model of choice is give in bold face.