Protease-catalyzed splicing of peptide bond

Ph D Students
SR Balajee
Lavanya Anantharaman
Srijita Banerjee

The two major themes of the project are: A) understanding the structural and mechanistic imperatives of peptide ligation reactions catalyzed by proteases and, B) engineering of new and novel hemoglobin tetramers to delineate intermolecular interactions in the sickle hemoglobin fiber.

A.    Protease-catalyzed splicing of native peptide fragments

Model studies on protease-mediated peptide ligation and protein semisynthesis are the objectives of the project.

During the current reporting period, we continued to address the mechanistic issues related to crowding-mediated reversal of proteolysis of the aforementioned coiled-coil and fragment complementing protein systems.

As reported previously, V8 protease-mediated synthesis of LIAA and VLAA (but not VVAA) from their respective complementary fragments (residues 1-11 and 12-20) proceeded in the presence of dextran. The Far-UV CD spectrum of the reactant peptide pairs (1:1 mixture, discontiguous) in all cases was featureless in both the presence and absence of dextran. In contrast, CD spectra of contiguous peptides were distinct. While VVAA displayed randomized conformation, LIAA and VLAA exhibited helical structures that remained largely unperturbed in the presence of dextran indicating the absence of direct interactions between dextran and the peptides. However, the ratio of q₂₂₂/q₂₀₈ for both peptide was found to be higher in the presence of dextran as compared to the dilute solution. The ratio for LIAA was found to be 1.01 and 1.04 in the absence and presence of dextran respectively. A value of q₂₂₂/q₂₀₈ > 1 is diagnostic of the presence of coiled coil helix in aqueous solution. Although the q₂₂₂/q₂₀₈ ratio of VLAA was significantly increased in the presence of dextran, values were much lower as compared with LIAA in both the absence (0.91 Vs 1.01) and presence (.97 Vs 1.04) of dextran. Thus dextran did not shift the equilibrium to synthesis by direct interactions with either product or reactant, rather the synthetic reaction was facilitated in the presence of dextran because crowding promoted association of helices to form coiled-coils that lead to the reduction in excluded volume. Further corroborative evidence of crowding effect also emanated from the fact that synthesis of LIAA proceeded as efficiently in polyethyleneglycol (PEG) as in dextran solutions.

Previously we reported subtilisin-mediated reformation of native triosephosphate isomerase (TIM) from mutiple fragments in the presence of dextran or PEG. We also reported that a single nick in ribonuclease S (RNase S) could not be ligated under similar conditions. The effect of crowding agent on the enzymatic activity of RNase S as well as RNase A was evaluated to see if dextran or PEG caused any dissociation of the S-peptide- S-protein complex because a weakening or dissociation of the complex would shift the equilibrium to a system that is no longer fragment complementing, and might lack the stereo-chemical proximity of the ligating ends. The enzymatic assays of RNase S or RNase A incubated in dextran under the conditions of reverse proteolysis yielded similar specific activity as that in the presence of dilute solution (absence of dextran). In contrast, both RNase A and RNase S, under similar conditions but in the presence of PEG retained only about 70% of the enzymatic activity. The fact that dextran did not exert any deleterious effect on the enzymatic activity of RNase S suggest that the interactions of S-peptide and S-protein in the RNase S complex were maintained in the presence of dextran. The diminution of activity in the presence of PEG in RNase S need not necessarily emanate from destabilization of RNase S complex per se considering that PEG also reduces the activity of covalent RNase A by the same amount. These results univocally establish that the lack of the desired excluded volume effect is the cause of the failure of RNase A reformation in the crowded milieu.

B.    Chemo-enzymic engineering of proteins

Site-specific incorporation of natural and/or nonstandard amino acids in hemoglobin to probe the intermolecular interactions in sickle hemoglobin (HbS) fiber are the objectives of the project.

During the current reporting period, we studied the polymerization behavior of the above two hybrid tetramers. These HbS tetramers are designated a₂(GGH)b₂^S and a₂(GGQ)b₂^S based on the fact that a-chain of langur also contains amino acid sequence changes at two other positions (a19Ala®Gly, a21Ala®Gly) in addition to that at a78 (Asn®His). The gelation concentration (Csat) of the hybrid tetramers was determined to ascertain the influence of sequence changes on the polymerization reaction. The Csat value (polymer solubility) of a₂ (GGH)b₂^S was similar to native HbS (28 Vs 30 mg/ml) while that of a₂(GGQ)b₂^S (50 mg/ml) was found to be about 1.7 fold higher. At a first glance, these results suggest that Gln at 78 positions inhibits polymerization and that His at this site or Ala to Gly substations at 19 or 21 positions have no effect on HbS polymerization. This interpretation appear consistent with the notion that residues 19 and 21 are not known to be contact points and that imidazole side chain at the 78 site presumably provides similar polymer contacts as that available with Asn residue present at this site in native HbS.

That the polymerization mechanism of the above hybrids was not as simple as expressed above became evident from polymerization behavior of a semisynthetic mutant HbS, a₂(GGN)b₂^S, that contained only the two Gly substitutions of the hybrids. Interestingly, the Csat of this mutant was very similar to that of a₂(GGQ)b₂^S hybrid implying that the inhibitory effect of the aGGQb^S hybrid actually emanated from the presence of Gly residues at the AB corner and that Asn to Gln change at the EF corner had no effect on the polymerization process. Further, considering that Gly residues at the AB corner are inhibitory, His 78 must enhance the polymerization and compensate for the inhibitory effect of the Gly residues. In this scenario, the interactions of AB and EF regions must be additive.

In order to test the above surmise and to establish the interaction-linkage between AB and EF regions in the polymerization process, we made simultaneous perturbations at a16 and a78 sites. The idea to couple a16 with His78 was based on the fact that this is an established contact point of the fiber and the perturbation of this AB region site is known to generate an inhibitory effect. Therefore, if His78 is a polymerization-enhancing residue and its effect is additive with AB region contacts, His78 should lower the intrinsic inhibitory effect of a16 in the double mutant in much the same way as the a₂(GGH)b₂^S hybrid. This was indeed the case as demonstrated by the polymerization experiments of mutant HbS containing a double mutant a-chain in which Lys16®Gln mutation at the AB corner was introduced along with Asn78®His. The single mutant HbS (Lys16®Gln) produced a Csat of 39 mg/ml that was consistent with the known inhibitory activity of a16. Interestingly, Csat of the double mutant was 17 mg/ml as against 30 mg/ml for native HbS. Thus, the presence of His78 reduced the Csat of the single mutant by about 22 units. Further, a simple arithmetic would show that His78 also pulled down the inhibitory effect due to the Gly substitutions at the AB corner by the same amount (22 units) as the Csat of the a₂(GGH)b₂^S and a₂(GGQ)b₂^S constructs were 28 mg/ml and 50 mg/ml respectively. These considerations lead to the unequivocal conclusion that the interactions of the AB and EF corner residues are additive. From the data, it can be deduced that the Csat of the HbS with a point mutation of His78 would be 8 mg/ml (since the Csat of HbS is 30 mg/ml, 30-22=8). Incidentally, a point mutation with such high polymerization-enhancing strength is not known thus far. This finding has implications for transgenic mouse model for sickle cell anemia.

Publications

Original peer-reviewed articles

1.     Balajee SR and Roy RP (2002) Volume exclusion effect as a driving force for reverse proteolysis: Implications for polypeptide assemblage in a macromolecular crowded milieu. J Biol Chem 277: 43253-43261.