Additivity-based design of the strongest possible turkey ovomucoid third domain inhibitors for porcine pancreatic elastase (PPE) and Streptomyces griseus protease B (SGPB)

We describe here successful designs of strong inhibitors for porcine pancreatic elastase (PPE) and Streptomyces griseus protease B (SGPB). For each enzyme two inhibitor variants were designed. In one, the reactive site residue (position 18) was retained and the best residues were substituted at contact positions 13, 14, and 15. In the other variant the best residues were substituted at all contact positions except the reactive site where a Gly was substituted. The four designed variants were: for PPE, T(13)E(14)Y(15)-OMTKY3 and T(13)E(14)Y(15)G(18)M(21)P(32)V(36)-OMTKY3, and for SGPB, S(13)D(14)Y(15)-OMTKY3 and S(13)D(14)Y(15)G(18)I(19)K(21)-OMTKY3. The free energies of association (ΔG(0)) of expressed variants have been measured with the proteases for which they were designed as well as with five other serine proteases and the results are discussed.

Analysis of sequence-reactivity space for protein-protein interactions

Sequence-reactivity space is defined by the relationships between amino acid type choices at some residue positions in a protein and the reactivities of the resulting variants. We are studying Kazal superfamily serine proteinase inhibitors, under substitution of any combination of residue types at 10 binding-region positions. Reactivities are defined by the standard free energy of association for an inhibitor against an enzyme, and we are interested in both the strength (the free energy value) and specificity (relative free energy values for one inhibitor against different enzymes). Characterizing the structure of such a space poses several interesting questions: (1) How many sequences achieve particular strength and specificity characteristics? (2) What is the best such sequence? (3) What are some nearly-as-good alternatives? (4) What are their common residue type characteristics (e.g., conservation and correlation)? Although these problems are all highly combinatorial in nature, this article develops an efficient, integrated mechanism to address them under a data-driven model that predicts reactivity for given sequences. We employ sampling and a novel deterministic distribution propagation algorithm, in order to determine both the reactivity distribution and sequence composition statistics; integer programming and a novel branch-and-bound search algorithm, in order to optimize sequences and enumerate near-optimal sequences; and correlation-based sequence decomposition, in order to identify sequence motifs. We demonstrate the value of our mechanism in analyzing the Kazal superfamily sequence-reactivity space, providing insights into the underlying biochemistry and suggesting hypotheses for further experimental consideration. In general, our mechanism offers a valuable tool for investigating the available degrees of freedom in protein design within a combined computational-experimental framework.

Identification of insulin receptor substrate 1 serine/threonine phosphorylation sites using mass spectrometry analysis: regulatory role of serine 1223

Insulin receptor substrate 1 (IRS-1), an intracellular substrate of the insulin receptor tyrosine kinase, also is heavily phosphorylated on serine and threonine residues, and several serine phosphorylation sites alter the function of IRS-1. Because of the large number of serine/threonine residues, position-by-position analysis of these potential phosphorylation sites by mutagenesis is difficult. To circumvent this, we have employed matrix-assisted laser desorption/ionization time-of-flight and HPLC-electrospray ionization tandem mass spectrometry techniques to scan for serine and threonine residues that are phosphorylated in full-length human IRS-1 ectopically expressed in cells using an adenoviral vector. This approach revealed 12 phosphorylation sites on serine or threonine residues, 10 of which were novel sites. Seven of these sites were in proline-directed motifs, whereas five were in arginine-directed sites. Sequence inspection suggested that phosphorylation of Ser1223 might alter the interaction of IRS-1 with the protein tyrosine phosphatase Src homology domain 2 (SH2)-containing phosphatase-2 (SHP-2). Mutation of Ser1223 to alanine to prevent phosphorylation resulted in increased association of SHP-2 with IRS-1, decreased insulin-stimulated tyrosine phosphorylation of IRS-1 in CHO/IR cells, and decreased insulin-stimulated association of the p85 regulatory subunit of phosphatidylinositol-3-kinase with IRS-1. This mutation had no effect on association of IRS-1 with the insulin receptor. Sequence analysis showed the Ser1223 region to be widely conserved evolutionarily. These data suggest that phosphorylation of Ser1223 dampens association of IRS-1 with SHP-2, thereby increasing net insulin-stimulated tyrosine phosphorylation.

Testing of the additivity-based protein sequence to reactivity algorithm

The standard free energies of association (or equilibrium constants) are predicted for 11 multiple variants of the turkey ovomucoid third domain, a member of the Kazal family of protein inhibitors, each interacting with six selected enzymes. The equilibrium constants for 38 of 66 possible interactions are strong enough to measure, and for these, the predicted and measured free energies are compared, thus providing an additional test of the additivity-based sequence to reactivity algorithm. The test appears to be unbiased as the 11 variants were designed a decade ago to study furin inhibition and the specificity of furin differs greatly from the specificities of our six target enzymes. As the contact regions of these inhibitors are highly positive, nonadditivity was expected. Of the 11 variants, one does not satisfy the restriction that either P(2) Thr or P(1)' Glu should be present and all three measurable results on it, as expected, are nonadditive. For the remaining 35 measurements, 22 are additive, 12 are partially additive, and only one is (slightly) nonadditive. These results are comparable to those obtained for a set of 398 equilibrium constants for natural variants of ovomucoid third domains. The expectation that clustering of charges would be nonadditive is modified to the expectation that major nonadditivity will be observed only if the combining sites in both associating proteins involve large charge clusters of the opposite sign. It is also shown here that an analysis of a small variant set can be accomplished with a smaller subset, in this case 13 variants, rather than the whole set of 191 members used for the complete algorithm.

Additivity-based prediction of equilibrium constants for some protein-protein associations

For many protein families, such as serine proteinases or serine proteinase inhibitors, the family assignment predicts reactivity only in general terms. Both detailed specificity and quantitative reactivity are lacking. We believe that, for many such protein families, algorithms can be devised by defining the subset of n functionally important sequence positions, making the 19n possible single mutants and measuring their reactivity. Given the assumption that the contributions of the n positions are additive, the reactivities of the 20(n) variants can be predicted. This is illustrated by an almost complete algorithm for the Kazal family of protein inhibitors of serine proteinases.