Kinase Chemogenomics: Targeting the Human Kinome for Target Validation and Drug Discovery

Chemogenomics is a gene family-based approach to drug discovery and target validation. This review will summarize the application of this interdisciplinary approach to the protein kinases of the human genome with emphasis upon the synergies and efficiencies to be gained. Specific examples from the SAPK-family will be discussed.

Mechanistic role of an NS4A peptide cofactor with the truncated NS3 protease of hepatitis C virus: elucidation of the NS4A stimulatory effect via kinetic analysis and inhibitor mapping

Infection by hepatitis C viruses (HCVs) is a serious medical problem with no broadly effective treatment available for the progression of chronic hepatitis. The catalytic activity of a viral serine protease located in the N-terminal one-third of nonstructural protein 3 (NS3) is required for polyprotein processing at four site-specific junctions. The three-dimensional crystal structure of the NS3-NS4A co-complex [Kim, J. L., Morgenstern, K. A., Lin, C., Fox, T., Dwyer, M. D., Landro, J. A., Chambers, S. P., Markland, W., Lepre, C. A., O'Malley, E. T., Harbeson, S. L., Rice, C. M., Murcko, M. A., Caron, P. R., & Thomson, J. A. (1996) Cell 87, 343-355] delineates a small hydrophobic region within the 54-residue NS4A protein that intercalates with and makes extensive contacts to the core of the protease. The current investigation addresses the mechanism of NS3 protease catalytic activation by NS4A utilizing a small synthetic NS4A peptide (residues 1678-1691 of the virus polyprotein sequence) and the recombinantly expressed protease domain of NS3. The addition of NS4A dramatically increased NS3 kcat and kcat/Km catalytic parameters when measured against small peptide substrates representing the different site-specific junctions of the polyprotein. The catalytic effect of natural and non-natural amino acid substitutions at the P1 position in a 5A/5B peptide substrate was investigated. NS3-NS4A demonstrated a marked catalytic preference for the cysteine residue commonly found in authentic substrates. The pH dependence of the NS3 hydrolysis reaction is not affected by the presence of NS4A. This result suggests that NS4A does not change the pKa values of the active site residues of NS3 protease. A steady state kinetic analysis was performed and indicated that the binding of NS4A and the peptide substrate occurs in an ordered fashion during the catalytic cycle, with NS4A binding first. Two distinct kinetic classes of peptidyl inhibitors based upon the 5A/5B cleavage site were identified. An NS4A-independent class is devoid of prime residues. A second class of inhibitors is NS4A-dependent and contains a natural or non-natural cyclic amino acid substituted for the commonly found P1' residue serine. These inhibitors display an up to 80-fold increase in affinity for NS3 protease in the presence of NS4A. Sequential truncation of prime and P residues from this inhibitor class demonstrated the fact that the P4' and P1' residues are crucial for potent inhibition. The selectivity of this NS4A effect is interpreted using a model of the 5A/5B decapeptide substrate bound to the active site of the NS3-NS4A structure.

Crystal structure of the hepatitis C virus NS3 protease domain complexed with a synthetic NS4A cofactor peptide

An estimated 1% of the global human population is infected by hepatitis C viruses (HCVs), and there are no broadly effective treatments for the debilitating progression of chronic hepatitis C. A serine protease located within the HCV NS3 protein processes the viral polyprotein at four specific sites and is considered essential for replication. Thus, it emerges as an attractive target for drug design. We report here the 2.5 angstrom resolution X-ray crystal structure of the NS3 protease domain complexed with a synthetic NS4A activator peptide. The protease has a chymotrypsin-like fold and features a tetrahedrally coordinated metal ion distal to the active site. The NS4A peptide intercalates within a beta sheet of the enzyme core.

Calpain inhibitor AK295 protects neurons from focal brain ischemia. Effects of postocclusion intra-arterial administration

BACKGROUND AND PURPOSE

This research was performed to determine whether a selective inhibitor of the calcium-dependent protease, calpain, could reduce ischemia-associated brain damage when peripherally administered after a vascular occlusion.

METHODS

A variation of the rat middle cerebral artery occlusion model was used. A range of doses of AK295 (a novel calpain inhibitor synthesized for this purpose) was continuously infused through the internal carotid artery, beginning 1.25 hours from the initiation of the occlusion. Rats were killed at 21 hours, and the infarct volume was quantified.

RESULTS

Postocclusion (1.25-hour) infusion of the calpain inhibitor AK295 elicited a dose-dependent neuroprotective effect after focal ischemia. The highest dose tested (3 mg/kg per hour) afforded the maximum effect, illustrated by a 32% reduction in infarct volume 21 hours after the ischemia (vehicle, 81.7 +/- 4.7 mm3; AK295, 54.9 +/- 6.9 mm3; P < .007).

CONCLUSIONS

These data provide the first evidence that a peripherally administered calpain inhibitor can protect against ischemic brain damage. They offer further support for an important role of calpain proteolysis in the brain degeneration associated with cerebral ischemic events and suggest that selective calpain inhibitors provide a rational, novel, and viable means of treating such neurodegenerative problems.

Stereospecific synthesis of peptidyl alpha-keto amides as inhibitors of calpain

Peptidyl alpha-keto amides have been synthesized and tested as inhibitors of the cysteine protease calpain. A stereospecific synthesis was devised in which Cbz-dipeptidyl-alpha-hydroxy amides were oxidized with TEMPO/hypochlorite to the corresponding alpha-keto amides. This oxidation was accomplished in good yields and without epimerization of the chiral center adjacent to the ketone. The potent inhibition of porcine calpain I by the L,L diastereomers, combined with the poor inhibition by the L,D diastereomers, established the requirement for the all-L stereochemistry of the active inhibitor. The early lead inhibitors were very hydrophobic and, therefore, poorly soluble in aqueous solutions. Using the stereospecific route, new compounds were prepared with polar groups at the C- and N-termini. These modifications resulted in more soluble inhibitors that were still potent inhibitors of calpain. Studies of the stability of these alpha-keto amides showed that absolute stereochemistry can be maintained in acidic and unbuffered environments but general base-catalyzed epimerization of the chiral center adjacent to the ketone occurred rapidly. The alpha-hydroxy precursors were inactive as inhibitors of calpain, which supports the hypothesis that the alpha-keto compounds reversibly form an enzyme-bound tetrahedral species that results from the nucleophilic addition of the catalytic thiol of calpain to the electrophilic ketone of the inhibitor.

An echistatin C-terminal peptide activates GPIIbIIIa binding to fibrinogen, fibronectin, vitronectin and collagen type I and type IV

Integrin binding to proteins often involves recognition of domains containing the arginine-glycine-aspartate (RGD) motif. Different binding affinities and specificities of the integrin-ligand protein interactions involve additional protein domains. The n.m.r. structure of the snake-venom protein echistatin suggested that the C-terminal portion of the molecule might be important, in addition to the RGD domain, in binding to the integrin glycoprotein IIbIIIa (GPIIbIIIa) [Saudek, Atkinson and Pelton (1991) Biochem. 30, 7369-7372]. The synthetic C-terminal peptide, echistatin-(40-49), PRNPHKGPAT, (1) inhibited binding of GPIIbIIIa to immobilized echistatin (IC50 3-6 mM), but did not inhibit binding of GPIIbIIIa to immobilized fibrinogen (up to 5 mM peptide), (2) activated GPIIbIIIa binding to fibronectin and vitronectin, usual ligands for the activated integrin, (3) activated binding of GPIIbIIIa to collagen type I and type IV, proteins not usually regarded as ligands for the integrin, and (4) stimulated 125I-fibrinogen binding by human platelets. These findings argue for an interaction of this non-RGD domain in echistatin with GPIIbIIIa, leading to activation of the integrin and extension of the ligand specificity to include immobilized collagen.

A new class of high affinity ligands for the neurokinin A NK2 receptor: psi (CH2NR) reduced peptide bond analogues of neurokinin A4-10

Analogues of [Leu10]NKA4-10 were synthesized in which each of the amide bonds was sequentially replaced with the reduced amide psi (CH2NH) bond to determine the effect of this structural modification on the antagonism of NKA binding to the HUB NK2 receptor. [psi (CH2-NH)9,Leu10]NKA4-10 (6) retained significant affinity for the NK2 receptor (IC50 = 115 nM) and showed weak partial stimulation of PI turnover (approximately 10-15% of NKA maximum). 6 behaves as a competitive antagonist of NKA-stimulated PI turnover with a pA2 = 6.7. The secondary amine of the psi (CH2NH) moiety of 6 was converted to a tertiary amine by alkylation. This modification was found to have a small effect upon receptor affinity but did result in attenuation of partial agonist activity. A combination of amino acid substitutions and psi (CH2NH) alkylation yielded [beta Ala8,psi (CH2N(CH2)2CH3)9,Phe10]NKA4-10 (21) which has very high affinity for the HUB NK2 receptor. This compound inhibited [125I]NKA binding with an IC50 = 1 nM which is equal to the receptor affinity of NKA. Compound 21 also shows very weak partial agonism of PI turnover (< or = 5% of NKA maximum) which makes this the most potent member of a new class of NKA ligands: psi(CH2NR)9-NKA4-10 analogues which potently antagonize NKA binding and possess minimal partial agonist activity.

Conformation of a neurokinin antagonist in solution. 2D NMR and restrained molecular dynamics study

The hexapeptide [cyclo(Leu1 psi(CH2NH2)Leu2-Gln3-Trp4-Phe5-Gly6)]+1 is a potent antagonist of neurokinin A activity in tissues of hamster urinary bladder. The solution conformation of this cyclic hexapeptide has been characterized by the combined use of two dimensional nuclear magnetic resonance spectroscopy and restrained molecular dynamics. The proton spectrum of the peptide was fully assigned by the sequential assignment procedure. Interproton distances were derived from crosspeak volumes in two dimensional Nuclear Overhauser Effect spectra, and dihedral angles were calculated from appropriate coupling constants. Temperature coefficients of the amide protons were determined. Restrained molecular dynamics simulations were carried out using the backbone interproton distances as constraints. During 210 ps of restrained molecular dynamics the peptide interconverted among three closely related families of conformations. These interconversions occurred at picosecond timescales under the simulation conditions.

Airway effects of inhaled bradykinin, substance P, and neurokinin A in sheep

The effects of inhaled bradykinin (BK), substance P (SP), and neurokinin A (NKA) on pulmonary resistance and airway responsiveness to carbachol were studied in conscious allergic sheep. Inhaled BK (20 breaths, 0.1 to 5.0 mg.ml-1) caused dose-dependent increases in pulmonary resistance. Neither inhaled SP nor NKA (20 breaths, 0.1 to 1.0 mg.ml-1) produced significant bronchoconstriction in allergic sheep. However, the response to SP could be enhanced (p less than 0.05) by pretreatment with the neutral endopeptidase inhibitor, thiorphan (40 breaths, 1 mg.ml-1). Sheep that were allergic to Ascaris suum antigen were 5.9 times (p less than 0.05) more sensitive to the constrictor effects of BK than nonallergic sheep. BK-induced bronchoconstriction was blocked in a dose-dependent fashion by the BK beta 2-receptor antagonist, NPC 567 (D-arginine[hydroxyproline3,D-phenylalanine7]BK). Atropine (0.2 mg.kg-1, intravenously) and nedocromil sodium (1 mg.kg-1 in 3 ml of saline, aerosolized) significantly inhibited the BK-induced bronchoconstriction by 97% and 43%, respectively. Chlorpheniramine (2 mg.kg-1, intravenously) had no effect. NKA caused a transient increase in airway responsiveness in allergic sheep, producing a mean 1.9-fold leftward shift in dose-response curves to aerosolized carbachol (p less than 0.05). This hyperresponsiveness was not evident 24 hours after NKA challenge. Neither SP nor BK changed airway responsiveness. Thus, in allergic sheep, inhaled BK caused a more pronounced bronchoconstriction than that observed in nonallergic sheep. The bronchoconstriction was blocked by a BK-receptor antagonist and appeared to be partially mediated via cholinergic reflexes.(ABSTRACT TRUNCATED AT 250 WORDS)

[Leu9 psi(CH2NH)Leu10]-neurokinin A (4-10) (MDL 28,564) distinguishes tissue tachykinin peptide NK2 receptors

The neurokinin A analogue, MDL 28,564 (Asp-Ser-Phe-Val-Gly-Leu-CH2NH-Leu-NH2), inhibited 125I-NKA binding to hamster urinary bladder NK2 receptors with a KI of 130 nM. For rat submaxillary gland NK1 receptors and cerebral cortical NK3 receptors, the KI's for MDL 28,564 were greater than 250 microM and greater than 500 microM, respectively. MDL 28,564 did not relax dog carotid artery (NK1 tissue) or contract rat portal vein (NK3 tissue). In guinea-pig trachea tissues, MDL 28,564 stimulated phosphatidylinositol turnover and induced contraction with maximum effects similar to those of neurokinin A. In hamster urinary bladder tissue, MDL 28,564 stimulated phosphatidylinositol turnover with maximum effect only 10% of that of neurokinin A, did not produce sustained contraction itself and antagonized NKA-induced contraction. MDL 28,564 also produced full contraction in rabbit pulmonary artery (NK2 tissue) but was inactive in rat vas deferens (NK2 tissue). These data with MDL 28,564 are consistent with the NK2 receptors in guinea-pig trachea and rabbit pulmonary artery being different from those in hamster urinary bladder and rat vas deferens.

Inhibition of aminopeptidases by peptides containing ketomethylene and hydroxyethylene amide bond replacements

Inhibitors of aminopeptidase enzymes have been prepared by the synthesis of peptide substrate analogues in which the scissile amide bond has been replaced with the hydrolytically stable ketomethylene (-COCH2-) and hydroxyethylene [-CH(OH)CH2-] functionalities. Two synthetic strategies were used to prepare the inhibitors, and the advantages and disadvantages of each are discussed. The synthesis of peptides that contain the hydroxyethylene isostere was complicated by competing lactone and lactam formation, and attempts to prepare free N-terminal dipeptide hydroxyethylene isostere derivatives were unsuccessful. All ketomethylene isosteres examined were weak inhibitors of both leucine aminopeptidase and aminopeptidase M. However, the ketomethylene inhibitor LysK(RS)Phe (58) (Ki = 4 nM) is a potent inhibitor comparable to the natural product, arphamenine A (ArgKPhe; Ki = 2.5 nM). Normal Michaelis-Menten kinetics for inhibition of membrane leucine aminopeptidase are observed in the absence of magnesium ion, but nonlinear kinetics were obtained in the presence of Mg2+.

Inhibition of arginine aminopeptidase by bestatin and arphamenine analogues. Evidence for a new mode of binding to aminopeptidases

The synthesis and inhibition kinetics of a new, potent inhibitor of arginine aminopeptidase (aminopeptidase B; EC 3.4.11.6) are reported. The inhibitor is a reduced isostere of bestatin in which the amide carbonyl is replaced by the methylene (-CH2-) moiety. Analysis of the inhibition of arginine aminopeptidase by this inhibitor according to the method of Lineweaver and Burk yields an unusual noncompetitive double-reciprocal plot. The replot of the slopes versus [inhibitor] is linear (Kis = 66 nM), but the replot of the y intercepts (1/V) versus [inhibitor] is hyperbolic (Kii = 10 nM, Kid = 17 nM). These results provide evidence for a kinetic mechanism in which the inhibitor binds to the S1' and S2' subsites on the enzyme, not the S1 and S1' subsites occupied by dipeptide substrates. Furthermore, structure-activity data for a series of ketomethylene dipeptide isosteres in which the amide (-CONH-) of a dipeptide is replaced with the ketomethylene (-COCH2-) moiety show that the S1 and S1' subsites preferentially bind basic and aromatic side chains, respectively. These results are in agreement with the known substrate specificity of arginine aminopeptidase. The structure-activity data for several bestatin analogues, however, show that these compounds do not bind to the S1 and S1' sites of arginine aminopeptidase. A comparison of the data provides evidence that bestatin inhibits arginine aminopeptidase and possibly other aminopeptidases by binding to the S1' and S2' sites of the enzyme.