Search

Attempts to increase protein stability by insertion of novel disulfide bonds have not always been successful. According to the two current models, cross-links enhance stability mainly through denatured state effects. We have investigated the effects of removal and addition of disulfide cross-links, protein flexibility in the vicinity of a cross-link, and disulfide loop size on the stability of Cucurbita maxima trypsin inhibitor-V (CMTI-V; 7 kD) by differential scanning calorimetry. CMTI-V offers the advantage of a large, flexible, and solvent-exposed loop not involved in extensive intra-molecular interactions. We have uncovered a negative correlation between retention time in hydrophobic column chromatography, a measure of protein hydrophobicity, and melting temperature (T(m)), an indicator of native state stabilization, for CMTI-V and its variants. In conjunction with the complete set of thermodynamic parameters of denaturation, this has led to the following deductions: (1) In the less stable, disulfide-removed C3S/C48S (Delta Delta G(d)(50 degrees C) = -4 kcal/mole; Delta T(m) = -22 degrees C), the native state is destabilized more than the denatured state; this also applies to the less-stable CMTI-V* (Delta Delta G(d)(50 degrees C) = -3 kcal/mole; Delta T(m) = -11 degrees C), in which the disulfide-containing loop is opened by specific hydrolysis of the Lys(44)-Asp(45) peptide bond; (2) In the less stable, disulfide-inserted E38C/W54C (Delta Delta G(d)(50 degrees C) = -1 kcal/mole; Delta T(m) = +2 degrees C), the denatured state is more stabilized than the native state; and (3) In the more stable, disulfide-engineered V42C/R52C (Delta Delta G(d)(50 degrees C) = +1 kcal/mole; Delta T(m) = +17 degrees C), the native state is more stabilized than the denatured state. These results show that a cross-link stabilizes both native and denatured states, and differential stabilization of the two states causes either loss or gain in protein stability. Removal of hydrogen bonds in the same flexible region of CMTI-V resulted in less destabilization despite larger changes in the enthalpy and entropy of denaturation. The effect of a cross-link on the denatured state of CMTI-V was estimated directly by means of a four-state thermodynamic cycle consisting of native and denatured states of CMTI-V and CMTI-V*. Overall, the results show that an enthalpy-entropy compensation accompanies disulfide bond effects and protein stabilization is profoundly modulated by altered hydrophobicity of both native and denatured states, altered flexibility near the cross-link, and residual structure in the denatured state.

Serine proteinase protein inhibitors follow the standard mechanism of inhibition (Laskowski M Jr, Kato I, 1980, Annu Rev Biochem 49:593-626), whereby an enzyme-catalyzed equilibrium between intact (I) and reactive-site hydrolyzed inhibitor (I*) is reached. The hydrolysis constant, Khyd, is defined as [I*]/[I]. Here, we explore the role of internal dynamics in the resynthesis of the scissile bond by comparing the internal mobility data of intact and cleaved inhibitors belonging to two different families. The inhibitors studied are recombinant Cucurbita maxima trypsin inhibitor III (rCMTI-III; Mr 3 kDa) of the squash family and rCMTI-V (Mr approximately 7 kDa) of the potato I family. These two inhibitors have different binding loop-scaffold interactions and different Khyd values--2.4 (CMTI-III) and 9 (CMTI-V)--at 25 degrees C. The reactive-site peptide bond (P1-P1') is that between Arg5 and Ile6 in CMTI-III, and that between Lys44 and Asp45 in CMTI-V. The order parameters (S2) of backbone NHs of uniformly 15N-labeled rCMTI-III and rCMTI-III* were determined from measurements of 15N spin-lattice and spin-spin relaxation rates, and [1H]-15N steady-state heteronuclear Overhauser effects, using the model-free formalism, and compared with the data reported previously for rCMTI-V and rCMTI-V*. The backbones of rCMTI-III [(S2) = 0.71] and rCMTI-III* [(S2) = 0.63] are more flexible than those of rCMTI-V [(S2) = 0.83] and rCMTI-V* [(S2) = 0.85]. The binding loop residues, P4-P1, in the two proteins show the following average order parameters: 0.57 (rCMTI-III) and 0.44 (rCMTI-III*); 0.70 (rCMTI-V) and 0.40 (rCMTI-V*). The P1'-P4' residues, on the other hand, are associated with (S2) values of 0.56 (rCMTI-III) and 0.47 (rCMTI-III*); and 0.73 (rCMTI-V) and 0.83 (rCMTI-V*). The newly formed C-terminal (Pn residues) gains a smaller magnitude of flexibility in rCMTI-III* due to the Cys3-Cys20 crosslink. In contrast, the newly formed N-terminal (Pn' residues) becomes more flexible only in rCMTI-III*, most likely due to lack of an interaction between the P1' residue and the scaffold in rCMTI-III. Thus, diminished flexibility gain of the Pn residues and, surprisingly, increased flexibility of the Pn' residues seem to facilitate the resynthesis of the P1-P1' bond, leading to a lower Khyd value.

The solution structure and internal dynamics of the reactive-site (Lys44-Asp45 peptide bond) hydrolyzed form of recombinant Cucurbita maxima trypsin inhibitor-V (rCMTI-V*) were characterized by the application of two-dimensional 1H-15N NMR methods to the uniformly 15N-labeled protein. The 1H-15N chemical shift correlation spectra of rCMTI-V* were assigned, and the chemical shift data were compared with those available for rCMTI-V [Liu, J., Prakash, O., Cai, M., Gong, Y., Huang, Y., Wen, L., Wen, J. J., Huang, J.-K., & Krishnamoorthi, R. (1996) Biochemistry 35, 1516-1524] and CMTI-V* [Cai, M., Gong, Y., Prakash, O., & Krishnamoorthi, R. (1995) Biochemistry 34, 12087-12094] for which three-dimensional solution structures have been determined. It was deduced that the solution structure of rCMTI-V* was almost the same as that of CMTI-V*. 15N spin-lattice and spin-spin relaxation rate constants (R1 and R2, respectively) and ¿1H¿-15N steady-state heteronuclear Overhauser effects were measured for the peptide NH units and arginine and tryptophan N epsilon H groups in rCMTI-V*, and the model-free parameters [Lipari, G., & Szabo, A. (1982) J. Am. Chem. Soc. 104, 4546-4559, 4559-4570] were computed. Most of the backbone of rCMTI-V* is found to be highly constrained (S2 = 0.85), including the N-terminal residues 3-6 (S2 = 0.77). Residues 39-44, forming the C-terminal fragment of the binding loop, exhibit increased mobility (S2 = 0.51); however, the N-terminal segment (residues 46-48) retains rigidity as in the intact form (S2 = 0.83). The S2 values, 0.78 and 0.59, respectively, of Arg50 and Arg52 side chain NHs provide evidence not only for the conservation of the Arg hydrogen-bonds with the binding loop segments but also for the difference in strength between them. This is consistent with the earlier observation made from a study of rCMTI-V at two different pHs and its R50 and R52 mutants [Cai, M., Huang, Y., Prakash, O., Wen, L., Dunkelbarger, S. P., Huang, J.-K., Liu, J., & Krishnamoorthi, R. (1996) Biochemistry 35, 4784-4794]. The dynamical results suggest the mainchain oxygen atom of Asp45 as the hydrogen bond acceptor of Arg50. Residues Trp9 and Trp54, which interact with many others in the protein scaffold and the binding loop region, respectively, remain rigid in the cleaved inhibitor with the S2 values of 0.84 and 0.71 determined for their respective N epsilon Hs. The internal dynamics of rCMTI-V* was compared with that of the noncovalent complex formed between the two fragments of reactive-site-hydrolyzed chymotrypsin inhibitor-2 from barley seeds [CI-2; Shaw, G. L., Davis, B., Keeler, J., & Fersht, A. R. (1995) Biochemistry 34, 2225-2233], another potato I family inhibitor that lacks the Cys3-Cys48 disulfide present in rCMTI-V*.

The side chains of Arg50 and Arg52 iin Cucurbita maxima trypsin inhibitor-V (CMTI-V) anchor the binding loop to the scaffold region [Cai, M., Gong, Y., Kao, J.L-F., & Krishnamoorthi, R. (1995) Biochemistry 34, 5201-5211]. The consequences of these hydrogen-bonding and electrostatic interactions on the conformational flexibility and stability of the binding loop were evaluated by trypsin-catalyzed hydrolysis of CMTI-V mutants, in which each of the arginines was individually replaced with Ala, Lys, or Gln by genetic engineering methods. All mutants exhibited significantly increased vulnerability to the protease attack at many sites, including the reactive-site (Lys44-Asp45 peptide bond), with the R50 mutants showing much more pronounced effects than the R52 counterparts. For CmTI-V and the mutants studied, a qualitative correlation was inferred between binding loop flexibility and retention time on a reverse-phase high-pressure liquid chromatography C-18 column. The R50 mutants were found to be more flexible than the corresponding R52 versions. These results demonstrate that Arg50 contributes more to the stability and function of CMTI-V. The differing strengths of the hydrogen bonds made by Arg50 and Arg52 were characterized by determining the internal dynamics of their side chains at pH 5.0 and 2.5: 15N NMR longitudinal and transverse relaxation rates and 15N-1H nuclear Overhauser effect (NOE) enhancements were measured for the main-chain and side-chain NH groups in 15N-labeled recombinant CMTI-V (rCMTI-V) and the model-free parameters [Lipari, G., & Szabo, A.(1982) J. Am. Chem. Soc. 104, 4546-59; 4559-4570] were calculated. At both pH 5.0 and 2.5, the arginines at positions 26, 47, 58 and 66 are found to be highly mobile, as the caluculated general order parameters, S2 values, of their NepsilonH groups fall in the range 0.03-0.18. The corresponding values for Arg50 amd Arg52 are 0.73 and 0.63, respectively, at pH 5.0, thus confirming that the two arginines are rigid and hydrogen bonded. At pH 2.5, these hydrogen bonds are still retained with Arg50 appearing to be more restrained (S2 = 0.71) than Arg52 (S2 = 0.56). This is consistent a greater contribution by Arg50 to the conformational stability of the reactive-site loop in CMTI-V. The results also indicate that the Arg50 and Arg52 side chains are not hydrogen-bonded to carboxylate groups, which would be protonated at pH 2.5 and, hence, unavailable for hydrogen-bonding interactions. The overall folding of rCMTI-V appears not to be significantly affected by the pH change, as indicated by comparisons of 1H and 15N chemical shifts, sequential NOE cross-peaks, and S2 values of the backbone atoms, and the conserved side-chain dynamics of Trp9 and Trp54--residues that are involved in hydrophobic and hydrogen-bonding interactions with others in the protein core and the binding loop, respectively.

The solution structure of recombinant Cucurbita maxima trypsin inhibitor-V (rCMTI-V), whose N-terminal is unacetylated and carries an extra glycine residue, was determined by means of two-dimensional (2D) homo and 3D hetero NMR experiments in combination with a distance geometry and simulated annealing algorithm. A total of 927 interproton distances and 123 torsion angle constraints were utilized to generate 18 structures. The root mean squared deviation (RMSD) of the mean structure is 0.53 A for main-chain atoms and 0.95 A for all the non-hydrogen atoms of residues 3-40 and 49-67. The average structure of rCMTI-V is found to be almost the same as that of the native protein [Cai, M., Gong, Y., Kao, J.-L., & Krishnamoorthi, R. (1995) Biochemistry 34, 5201-5211]. The backbone dynamics of uniformly 15N-labeled rCMTI-V were characterized by 2D 1H-15N NMR methods. 15N spin-lattice and spin-spin relaxation rate constants (R1 and R2, respectively) and [1H]-15N steady-state heteronuclear Overhauser effect enhancements were measured for the peptide NH units and, using the model-free formalism [Lipari, G., & Szabo, A. (1982) J. Am. Chem. Soc. 104, 4546-4559, 4559-4570], the following parameters were determined: overall tumbling correlation time for the protein molecule (tau m), generalized order parameters for the individual N-H vectors (S2), effective correlation times for their internal motions (tau e), and terms to account for motions on a slower time scale (second) due to chemical exchange and/or conformational averaging (R(ex)). Most of the backbone NH groups of rCMTI-V are found to be highly constrained ((S2) = 0.83) with the exception of those in the binding loop (residues 41-48, (S2) = 0.71) and the N-terminal region ((S2) = 0.73). Main-chain atoms in these regions show large RMSD values in the average NMR structure. Residues involved in turns also appear to have more mobility ((S2) = 0.80). Dynamical properties of rCMTI-V were compared with those of two other inhibitors of the potato I family--eglin c [Peng, J. W., & Wagner, G. (1992) Biochemistry 31, 8571-8586] and barley chymotrypsin inhibitor 2 [CI-2; Shaw, G. L., Davis, B., Keeler, J., & Fersht, A. R. (1995) Biochemistry 34, 2225-2233]. The Cys3-Cys48 linkage found only in rCMTI-V appears to somewhat reduce the N-terminal flexibility; likewise, the C-terminal of rCMTI-V, being part of a beta-sheet, appears to be more rigid.

Reactive-site (Lys44-Asp45 peptide bond) hydrolyzed Cucurbita maxima trypsin inhibitor-V (CMTI-V*) was prepared and characterized: In comparison to the intact form, CMTI-V* exhibited markedly reduced inhibitory properties and binding affinities toward trypsin and human blood coagulation factor XIIa. The equilibrium constant of trypsin-catalyzed hydrolysis, Khyd, defined as [CMTI-V*]/[CMTI-V], was measured to be approximately 9.4 at 25 degrees C (delta G degrees = -1.3 kcal.mol-1). From the temperature dependence of delta G degrees, the following thermodynamic parameters were estimated: delta H degrees = 1.6 kcal.mol-1 and delta S degrees = 9.8 eu. In order to understand the functional and thermodynamic differences between the two forms, the three-dimensional solution structure of CMTI-V* was determined by a combined approach of NMR, distance geometry, and simulated annealing methods. Thus, following sequence-specific and stereospecific resonance assignments, including those of beta-, gamma-, delta-, and epsilon-hydrogens and valine methyl hydrogens, 809 interhydrogen distances and 123 dihedral angle constraints were determined, resulting in the computation and energy-minimization of 20 structures for CMTI-V*. The average root mean squared deviation in position for equivalent atoms between the 20 individual structures and the mean structure obtained by averaging their coordinates is 0.67 +/- 0.15 A for the main chain atoms and 1.19 +/- 0.23 A for all the non-hydrogen atoms of residues 5-40 and residues 48-67.(ABSTRACT TRUNCATED AT 250 WORDS)

The solution structure of Cucurbita maxima trypsin inhibitor-V (CMTI-V), which is also a specific inhibitor of the blood coagulation protein, factor XIIa, was determined by 1H NMR spectroscopy in combination with a distance-geometry and simulated annealing algorithm. Sequence-specific resonance assignments were made for all the main-chain and most of the side-chain hydrogens. Stereospecific assignments were also made for some of the beta-, gamma-, delta-, and epsilon-hydrogens and valine methyl hydrogens. The ring conformations of all six prolines in the inhibitor were determined on the basis of 1H-1H vicinal coupling constant patterns; most of the proline ring hydrogens were stereospecifically assigned on the basis of vicinal coupling constant and intraresidue nuclear Overhauser effect (NOE) patterns. Distance constraints were determined on the basis of NOEs between pairs of hydrogens. Dihedral angle constraints were determined from estimates of scalar coupling constants and intraresidue NOEs. On the basis of 727 interproton distance and 111 torsion angle constraints, which included backbone phi angles and side-chain chi 1, chi 2, chi 3, and chi 4 angles, 22 structures were calculated by a distance geometry algorithm and refined by energy minimization and simulated annealing methods. Both main-chain and side-chain atoms are well-defined, except for a loop region, two terminal residues, and some side-chain atoms located on the molecular surface. The average root mean squared deviation in the position for equivalent atoms between the 22 individual structures and the mean structure obtained by averaging their coordinates is 0.58 +/- 0.06 A for the main-chain atoms and 1.01 +/- 0.07 A for all the non-hydrogen atoms of residues 3-40 and 49-67. These structures were compared to the X-ray crystallographic structure of another protein of the same inhibitor family-chymotrypsin inhibitor-2 from barley seeds [CI-2; McPhalen, C. A., & James, M. N. G. (1987) Biochemistry 26, 261-269]. The main-chain folding patterns are highly similar for the two proteins, which possess 62% sequence differences. However, major differences are noted in the N- and C-terminal segments, which may be due to the presence of a disulfide bridge in CMTI-V, but not in CI-2.

Cucurbita maxima trypsin inhibitor-V (CMTI-V) is also a specific inhibitor of human blood coagulation factor beta-factor XIIa. A recombinant version of CMTI-V has allowed probing of roles of individual amino acid residues including the reactive site residue, lysine (P1), by site-directed mutagenesis. The K44R showed at least a 5-fold increase in inhibitory activity toward human beta-factor XIIa, while there was no change toward bovine trypsin. This result demonstrates that beta-factor-XIIa prefers an arginine residue over lysine residue, while trypsin is non-specific to lysine or arginine in its binding pocket. On the other hand, the specificity of CMTI-V could be changed from trypsin to chymotrypsin inhibition by mutation of the P1 residue to either leucine or methionine (K44L or K44M).

The gene encoding for a pumpkin (Curcubita maxima) trypsin inhibitor CMTI-V was synthesized chemically. The synthetic gene was prepared from four overlapping oligonucleotides by overlapping extension. The synthetic gene was amplified by polymerase chain reaction, cloned into a T7 expression vector and expressed in Escherichia coli as a fusion protein. The clone, namely 70-1, encoded a fusion protein containing 7 amino acid residues of the N-terminus of the bacterial protein rho 10 and the entire 68 residues of CMTI-V. The wild-type fusion protein constituted approximately 15% of the total bacterial protein mass and was purified to homogeneity in a single step by antibody affinity chromatography. The wild-type fusion protein possesses inhibitory activity toward trypsin and beta-Factor XIIa, but to a lesser extent when compared to the natural CMTI-V. A mutant, T43A, in which threonine at position 43 (P2 position) was replaced by alanine, was constructed. This mutant showed considerably lower specific inhibitory activity toward both trypsin and beta-Factor XIIa.

15N NMR assignments were made to the backbone amide nitrogen atoms at natural isotopic abundance of intact and reactive-site (Arg5-Ile6) hydrolyzed Cucurbita maxima trypsin inhibitor III (CMTI-III and CMTI-III*, respectively) by means of 2D proton-detected heteronuclear single bond chemical shift correlation (HSBC) spectroscopy, utilizing the previously made sequence-specific 1H NMR assignments (Krishnamoorthi et al. (1992) Biochemistry 31, 898-904). Comparison of the 15N chemical shifts of the two forms of the inhibitor molecule revealed significant changes not only for residues located near the reactive-site region, but also for those distantly located. Residues Cys3, Arg5, Leu7, Met8, Cys10, Cys16, Glu19, His25, Tyr27, Cys28 and Gly29 showed significant chemical shift changes ranging from 0.3 to 6.1 ppm, thus indicating structural perturbations that were transmitted throughout the molecule. These findings confirm the earlier conclusions based on 1H NMR investigations.

Sequence-specific hydrogen-1 NMR assignments were made to all of the 29 amino acid residues of reactive-site-hydrolyzed Cucurbita maxima trypsin inhibitor I (CMTI-I*) by the application of two-dimensional NMR (2D NMR) techniques, and its secondary structural elements (two tight turns, a 3(10)-helix, and a triple-stranded beta-sheet) were identified on the basis of short-range NOESY cross peaks and deuterium-exchange kinetics. These secondary structural elements are present in the intact inhibitor [Holak, T. A., Gondol, D., Otlewski, J., & Wilusz, T. (1989) J. Mol. Biol. 210, 635-648] and are unaffected by the hydrolysis of the reactive-site peptide bond between Arg5 and Ile6, in accordance with the earlier conclusion reached for CMTI-III* [Krishnamoorthi, R., Gong, Y.-X., Lin, C. S., & VanderVelde, D. (1992) Biochemistry 31, 898-904]. Chemical shifts of backbone hydrogen atoms, peptide NH's, and C alpha H's, of CMTI-I* were compared with those of the intact inhibitor, CMTI-I, and of the reactive-site-hydrolyzed, natural, E9K variant, CMTI-III*. Cleavage of the Arg5-Ile6 peptide bond resulted in changes of chemical shifts of most of the backbone atoms of CMTI-I, in agreement with the earlier results obtained for CMTI-III. Comparison of chemical shifts of backbone hydrogen atoms of CMTI-I* and CMTI-III* revealed no changes, except for residues Glu9 and His25. However, the intact forms of the same two proteins, CMTI-I and CMTI-III, showed small but significant perturbations of chemical shifts of residues that made up the secondary structural elements of the inhibitors.(ABSTRACT TRUNCATED AT 250 WORDS)

The solution structure of reactive-site hydrolyzed Cucurbita maxima trypsin inhibitor III (CMTI-III*) was investigated by two-dimensional proton nuclear magnetic resonance (2D NMR) spectroscopy. CMTI-III*, prepared by reacting CMTI-III with trypsin which cleaved the Arg5-Ile6 peptide bond, had the two fragments held together by a disulfide linkage. Sequence-specific 1H NMR resonance assignments were made for all the 29 amino acid residues of the protein. The secondary structure of CMTI-III*, as deduced from NOESY cross peaks and identification of slowly exchanging hydrogens, contains two turns (residues 8-12 and 24-27), a 3(10)-helix (residues 13-16), and a triple-stranded beta-sheet (residues 8-10, 29-27, and 21-25). This secondary structure is similar to that of CMTI-I [Holak, T. A., Gondol, D., Otlewski, J., & Wilusz, T. (1989) J. Mol. Biol. 210, 635-648], which has a Glu instead of a Lys at position 9. Sequential proton assignments were also made for the virgin inhibitor, CMTI-III, at pH 4.71, 30 degrees C. Comparison of backbone hydrogen chemical shifts of CMTI-III and CMTI-III* revealed significant changes for residues located far away from the reactive-site region as well as for those located near it, indicating tertiary structural changes that are transmitted through most of the 29 residues of the inhibitor protein. Many of these residues are functionally important in that they make contact with atoms of the enzyme in the trypsin-inhibitor complex, as revealed by X-ray crystallography [Bode, W., Greyling, H. J., Huber, R., Otlewski, J., & Wilusz, T. (1989) FEBS Lett. 242, 285-292].(ABSTRACT TRUNCATED AT 250 WORDS)