Equilibriums of Bowman-Birk inhibitor association with trypsin and α

Equilibriums of Bowman-Birk inhibitor association with trypsin and α-chymotrypsin ... of molecules that disrupt serine protease-proteinaceous inhibito...
0 downloads 0 Views 918KB Size
BOWMAN -BlRK

I N H IBlTOR

ASSOCIATION

Equilibria of Bowman-Birk Inhibitor Association with Trypsin and a-Chymotrypsin? Richard Turner, lrvin E. Liener,* and Rex E. Lovrien

ABSTRACT: Association constants, enthalpies, and stoichiometries of Bowman-Birk soybean inhibitor for trypsin and a-chymotrypsin were measured in the pH range 4-8 a t 2 5 O , 0.01 M Ca2+. The results are quoted in terms of moles of protease active sites, from active site titration. Enthalpies were obtained from calorimetry. The inhibitor was modified by carboxyl group modification, and by tryptic and chymotryptic attack. Association thermodynamics and stoichiometries of the modified inhibitors wtth both proteases were also determined. There is one independent site for each protease on the inhibitor protein. Modification decreases association to some extent, but does not appear to change stoichiometry or protease binding site independency. In the pH 4 region the association enthalpies are endothermic, of the order 6 kcal/mol for both trypsin and

chymotrypsin. With increasing pH, the enthalpies decrease and become exothermic at pH 8 for chymotrypsin. Positive entropies, -50 cal mol-' deg-I, occur at pH 4-5. They decrease as pH increases, but are always positive in sign. The observed enthalpies cannot be assigned to single reactions expected to accompany the overall reaction, such as H + transfer steps. The enthalpies and entropies probably compensate over the p H range 4 -8, with a characteristic temperature of 390 f 30OK. Estimates were made of the rnacromolecular Coulomb charge products in inhibitor-protease interaction. These range from about +5 to -60, over pH range 4--8, depending on the protease. Although interniolecular Coulombic forces cannot be easily delineated at the specific side chain level, they may operate at the macromolecule level.

T h e interaction of proteases with naturally occurring inhibitors has proved to be an attractive system for studying protein-protein interaction (see review by Laskowski and Sealock, I971 ). Among the most thoroughly studied inhibitors are the Kunitz bovine pancreatic inhibitor and the Kunitz soybean inhibitor, both of which combine with trypsin i n a stoichiometric fashion (1:l). The most significant feature of this binding process is the extremely high association constant with reported values of 107-10'4 I./mol, which corresponds to a free energy of association of 10- 15 kcal/mol (Tschesche, 1974). Recent X-ray crystallographic studies of the complexes of trypsin with the pancreatic inhibitor (Ruhlman et al., 1973) and the Kunitz soybean inhibitor (Blow et al., 1974) have provided a closer insight into the nature of the forces which are responsible for the development of the strong specific binding between these protein molecules, the predominant feature of which is evidence for a covalent tetrahedral intermediate. Calorimetric measurements have shown that the binding of the Kunitz soybean inhibitor to trypsin is endothermic (Baugh and Trowbridge, I972), reflecting the relatively high enthalpy of the tetrahedral form, but a large entropy term stabilizes the complex. O f special interest are those protease inhibitors which have the unique capacity to inhibit both trypsin and chymotrypsin a t independent, nonoverlapping binding sites. Examples of these called "double-headed'' inhibitors are turkey ovomucoid (Stevens and Feeney, 1963), the lima bean inhibitor (Krahn and Stevens, 1970), and the BowmanBirk soybean inhibitor (Birk et al., 1967). In the case of the

lima bean and BBI' these two independent binding sites have been identified in relation to their total amino acid sequences (Krahn and Stevens, 1970, 1972; Seidl and Liener, 1971, I972a; Odani et a/.,1972). Unlike the Kunitz soybean inhibitor, however, little information seems to be available concerning the thermodynamic changes which accompany the association of the BBI with T or CT. I n a previous report from this laboratory (Seidl and Liener, 1972b). an attempt was made to evaluate the dissociation constants of the isolated binary complexes of BBI with T or CT as well as the ternary complex of this inhibitor with both of these enzymes. W e report here a further extension of this work in which the interactions of BBI with T and C T have been studied by means of calorimetry. This technique permitted a measure of the true heat of association as a function of pH, which could then be compared to associations obtained by measurements of enzymatic activity, as well as other thermodynamic parameters including free energy, enthalpy, and entropy.

__I_.___.

From the Biochemistry Department. College o f Biological Scicnces. Gortner Laboratory, University of Minnesota. S t . Paul. Minne'rota 55108. Rrcriwd July 30, / 9 7 4 . This work was supported by granta from National Institutes of Health, No. AM-13869. to I.E.1 _, and G M 18807. to R.E.L.

Materials and Methods Twice crystallized bovine trypsin' and three times crystallized bovine cu-chymotrypsin were obtained from Worthington Biochemical Corporation. RBI was prepared according to the method of Birk (Birk and Gertler, 1968) except

'

Abbreviations used are: BBI, Bowman--Birk inhibitor; STI, soybean trypsin inhibitor (Kunitz); T, trypsin: CT, n-chymotrypsin; BBI*T, trypsin modified Bowman-Birk inhibitor; B B I T T , o-chymotrypsin modified Bowman- Birk inhibitor; B T N A , N- benzoyl-l.-tqrosinc-~nitroanilide; BAPA. cu-N-benzoyl-DL-arginine-pnitroanilide HCI: N P A , p - nitrophenyl acetate; N B H , 5-nitro-3H- I .2-bcnn)xathiolc 2.2-dioxide: N P C B, p - nitrophenyl-p- guanidinobcnmatc. 'Since this preparation was not further purificd. i t ma) bc prcaumcd to be a mixture o f CY- and @-trypsin (Schroeder and S h a w , 1968). T h c thermodynamic parameters cited in this paper therefore rcfcr to a mixture containing these two species of tryphin in some unknown proportion. B I O C H E M I S T R Y , VOL..

14,

NO.

2, 1975

275

TLJRNFR. LIENFR. A N D

I O V R I F N

corresponding to the overall reaction were subtracted, and the output is a differential heat. The error was 15% in the p - I mcal range, and 3% in the 0-5 mcal range. for a single measurement. Results

Stoirhionietry of'Association of BBI w i t h Trj.p.vin c i r i d tu-chj,motrypsin Active Sites. The reaction BBI proteolytic enzyme ~2 complex, or more briefly B, PI B . f'. implies a I : I stoichiometry for the reactants. Subscript f denotes free or nonassociiited molecules. A fraction of the total amount of enzyme preparations is sometimes inactive. There arises a question as to whether an inhibitor, BBI i n this case. interacts with all the enzyme molecules, or only those which are truly active. This was investigated by determining stoichiometry o n the basis of the amount of active site titratable enzyme, i n conjunction uith enzymatic activity assays to measure the amount of free enzyme. For the reaction BI. PI i+ B . P, I I