Development of serine protease inhibitors displaying a multicentered short (<2.3 A) hydrogen bond binding mode: inhibitors of urokinase-type plasminogen activator and factor Xa.

Novel scaffolds that bind to serine proteases through a unique network of short hydrogen bonds to the catalytic Ser195 have been developed. The resulting potent serine protease inhibitors were designed from lead molecule 2-(2-hydroxyphenyl)1H-benzoimidazole-5-carboxamidine, 6b, which is known to display several modes of binding. For instance, 6b can recruit zinc and bind in a manner similar to that reported by bis(5-amidino-2-benzimidazolyl)methane (BABIM) (Nature 1998, 391, 608-612).(1) Alternatively, 6b can bind in the absence of zinc through a multicentered network of short (<2.3 A) hydrogen bonds. The lead structure was optimized in the zinc-independent binding mode toward a panel of six human serine proteases to yield optimized inhibitors such as 2-(3-bromo-2-hydroxy-5-methylphenyl)-1H-indole-5-carboxamidine, 22a, and 2-(2-hydroxybiphenyl-3-yl)-1H-indole-5-carboxamidine, 22f. Structure-activity relationships determined that, apart from the amidine function, an indole or benzimidazole and an ortho substituted phenol group were also essential components for optimal potency. The affinities (K(i)) of 22a and 22f, for example, bearing these groups ranged from 8 to 600 nM toward a panel of six human serine proteases. High-resolution crystal structures revealed that the binding mode of these molecules in several of the enzymes was identical to that of 6b and involved short (<2.3 A) hydrogen bonds among the inhibitor hydroxyl oxygen, Ser195, and a water molecule trapped in the oxyanion hole. In summation, novel and potent trypsin-like serine protease inhibitors possessing a unique mode of binding have been discovered.

Expression, purification, and characterization of deglycosylated human pro-prostate-specific antigen.

Wild-type and deglycosylated forms of human prostate-specific antigen were expressed in Chinese hamster ovary (CHO) cells as zymogens. ProPSA was collected from conditioned medium and purified using a single cation-exchange chromatographic step for the deglycosylated form and cation-exchange followed by gel filtration chromatography for the wild-type form. Recombinant wild-type proPSA produced in CHO cells has an average MW of 34.5 kDa, whereas the deglycosylated proPSA has a MW of 32.4 kDa. Both forms of proPSA were activated in vitro and the kinetic properties measured for the deglycosylated PSA are very similar to those of the wild-type recombinant PSA and the native PSA isolated from seminal fluid. These results suggest that deglycosylated PSA is likely to be very similar to native PSA with respect to its three-dimensional structure and will provide a homogeneous protein preparation necessary for X-ray crystallographic analysis.

Structural basis for selectivity of a small molecule, S1-binding, submicromolar inhibitor of urokinase-type plasminogen activator.

BACKGROUND: Urokinase-type plasminogen activator (uPA) is a protease associated with tumor metastasis and invasion. Inhibitors of uPA may have potential as drugs for prostate, breast and other cancers. Therapeutically useful inhibitors must be selective for uPA and not appreciably inhibit the related, and structurally and functionally similar enzyme, tissue-type plasminogen activator (tPA), involved in the vital blood-clotting cascade. RESULTS: We produced mutagenically deglycosylated low molecular weight uPA and determined the crystal structure of its complex with 4-iodobenzo[b]thiophene 2-carboxamidine (K(i) = 0.21 +/- 0.02 microM). To probe the structural determinants of the affinity and selectivity of this inhibitor for uPA we also determined the structures of its trypsin and thrombin complexes, of apo-trypsin, apo-thrombin and apo-factor Xa, and of uPA, trypsin and thrombin bound by compounds that are less effective uPA inhibitors, benzo[b]thiophene-2-carboxamidine, thieno[2,3-b]-pyridine-2-carboxamidine and benzamidine. The K(i) values of each inhibitor toward uPA, tPA, trypsin, tryptase, thrombin and factor Xa were determined and compared. One selectivity determinant of the benzo[b]thiophene-2-carboxamidines for uPA involves a hydrogen bond at the S1 site to Ogamma(Ser190) that is absent in the Ala190 proteases, tPA, thrombin and factor Xa. Other subtle differences in the architecture of the S1 site also influence inhibitor affinity and enzyme-bound structure. CONCLUSIONS: Subtle structural differences in the S1 site of uPA compared with that of related proteases, which result in part from the presence of a serine residue at position 190, account for the selectivity of small thiophene-2-carboxamidines for uPA, and afford a framework for structure-based design of small, potent, selective uPA inhibitors.

UDP-N-acetylglucosamine acyltransferase of Escherichia coli. The first step of endotoxin biosynthesis is thermodynamically unfavorable.

UDP-N-acetylglucosamine acyltransferase of Escherichia coli catalyzes the reaction, UDP-GlcNAc + R-3-hydroxymyristoyl-ACP--> UDP-3-O-(R-3-hydroxymyristoyl)-GlcNAc + ACP. Using Matrex Gel Green A and heparin-agarose, we have purified the enzyme to near homogeneity from a strain that overproduces it 474-fold. The subunit molecular mass determined by SDS-gel electrophoresis is approximately 30 kDa, consistent with results of previous radiolabeling experiments in mini-cells. The amino-terminal sequence (Met-Ile-Asp-Lys-Ser-Ala-Phe-Val-His-Pro) and the amino acid composition of the purified protein are consistent with DNA sequencing (Coleman, J., and Raetz, C. R. H. (1988) J. Bacteriol. 170, 1268-1274). At saturating concentrations of the second substrate, the apparent Km values for UDP-GlcNAc and R-3-hydroxymyristoyl-ACP are 99 and 1.6 microM, respectively. There is an absolute requirement for the R-3-hydroxy moiety of the fatty acyl-ACP substrate; myristoyl-ACP binds effectively (IC50 = 2 microM) but is inactive (< 0.01%) as an alternate substrate. The most remarkable feature of the reaction is its unfavorable equilibrium constant, Keq approximately equal to 0.01, which is not predicted by model S-->O acyl transfer reactions. Thus, although UDP-GlcNAc acyltransferase catalyzes the first unique step of lipid A biosynthesis, it is the second enzyme (the deacetylase) that commits the substrates to this pathway. The specific activity of the deacetylase is elevated approximately 5-fold when lipid A synthesis is inhibited.

Anti-immune complex antibodies enhance affinity and specificity of primary antibodies.

Antibodies have previously been described that enhance the binding of a second antibody to its antigen. The origin of this effect has been variously ascribed to binding to a neodeterminant on the Fc region, to a combined determinant representing portions of the second antibody and the immunogen, and to a ligand-induced conformation of the Fab fragment. This paper describes an antibody that recognizes an immune complex of an antibody to tetrahydrocannabinol (THC). The antibody binds the anti-THC antibody at an epitope recognized by an anti-idiotype antibody that is capable of blocking THC binding. The ability of various THC derivatives to enhance or inhibit binding taken together with equilibria and kinetic data support a model in which the anti-immune complex antibody interacts through adventitious binding to pendant groups on the THC derivatives. This type of interaction offers the opportunity to increase the sensitivity and specificity of immunoassays beyond the limits imposed by normal antibody binding. The implications of these findings with regard to earlier observations of anti-immune complex antibodies are discussed.

Engineering of genetically detoxified pertussis toxin analogs for development of a recombinant whooping cough vaccine.

Pertussis toxin (PT) is an important protective antigen in vaccines against whooping cough, and a genetically detoxified PT analog is the preferred form of the immunogen. Several amino acids of the S1 subunit were identified as functionally critical residues by site-directed mutagenesis, specifically, those at positions 9, 13, 26, 35, 41, 58, and 129. Eighty-three mutated PT operons were introduced into Bordetella parapertussis, and the resultant toxin analogs were screened for expression levels, enzymatic activity, residual toxicity, and antigenicity. While more than half of the mutants were found to be poorly secreted or assembled, the rest were fully assembled and most were highly detoxified. Single mutations resulted in up to a 1,000-fold reduction in both toxic and enzymatic activities, while PT analogs with multiple mutations (Lys-9 Gly-129, Glu-58 Gly-129, and Lys-9 Glu-58 Gly-129) were 10(6)-fold detoxified. Operons coding for stable and nontoxic mutants shown to express a critical immunodominant protective epitope were returned to the chromosome of Bordetella pertussis by allelic exchange. In vivo analysis of the toxin analogs showed a dramatic reduction in histamine sensitization and lymphocytosis-promoting activities, paralleling the reduction in toxic activities. All mutants were protective in an intracerebral challenge test, and the Lys-9 Gly-129 analog was found to be significantly more immunogenic than the toxoid. PT analogs such as those described represent suitable components for the design of a recombinant whooping cough vaccine.

Purification and properties of lipid A disaccharide synthase of Escherichia coli.

Lipid A disaccharide synthase of Escherichia coli catalyzes the reaction 2,3-diacyl-GlcN-1-P + UDP-2,3-diacyl-GlcN----2',3'-diacyl-GlcN (beta,1'----6)2,3-diacyl-GlcN-1-P + UDP (Ray, B. L., Painter, G., and Raetz, C. R. H. (1984) J. Biol. Chem. 259, 4852-4859). Using a strain that overproduces the enzyme about 200-fold we have devised a simple purification to near homogeneity, utilizing two types of dye-ligand resins and heparin-agarose. The overall purification starting with membrane-free extracts was 54-fold (16,000-fold relative to wild-type extracts) with a 31% yield. The subunit molecular mass determined by sodium dodecyl sulfate gel electrophoresis is approximately 42,000 daltons, and the native enzyme appears to be a dimer. The amino-terminal sequence is (X)-(Thr)-Glu-Gln-(X)-Pro-Leu-Thr-Ie-Ala..., consistent with the results predicted from the DNA sequence, Met-Thr-Glu-Gln-Arg-Pro-Leu-Thr-Ile-Ala.... The purified enzyme displays a strong kinetic preference for sugar substrates bearing two fatty acyl moieties, but it is, nevertheless, very useful for the semisynthetic preparation of many lipid A analogs. Gel filtration studies demonstrate that the natural substrates (2,3-diacyl-GlcN-1-P and UDP-2,3-diacyl-GlcN) form micelles (n approximately equal to 300), rather than bilayers, under conditions used to assay the enzyme. Unlike most enzymes of glycerophospholipid synthesis, the lipid A disaccharide synthase does not require the presence of a detergent for catalytic activity. At 1 mM UDP-2,3-diacyl-GlcN the Vmax and Km values for 2,3-diacyl-GlcN-1-P are 14,028 +/- 513 nmol/min/mg and 0.27 +/- 0.02 mM. When 2,3-diacyl-GlcN-1-P is maintained at 1 mM, they are 12,368 +/- 472 nmol/min/mg and 0.11 +/- 0.01 mM for UDP-2,3-diacyl-GlcN.

Inhibition of foetal pulmonary choline-phosphate cytidylyltransferase under conditions favouring protein phosphorylation.

Choline-phosphate cytidylyltransferase (EC 2.7.7.15) activity from 25- and 29-day-foetal rabbit lungs was inhibited in both the cytosolic and the microsomal fractions by preincubation with MgATP. The inhibition of the cytosolic enzyme was greater when measured with added phosphatidylglycerol (PG) than without (78-89% versus 50-55%), whereas the inhibition of the microsomal enzyme did not exhibit this distinction (66-72% versus 60-70%). When preincubated with the buffer alone, the cytosolic enzyme was activated to a greater extent by added PG than was the microsomal enzyme (13-14-fold versus 2-3-fold). However, after preincubation with MgATP, the cytosolic enzyme was activated to a smaller extent by added PG (3-6-fold). The inhibition of the enzyme by MgATP required a preincubation and was absent when ADP or AMP was substituted for ATP. Moreover, ATP analogues such as adenosine 5'-[beta, gamma-methylene]triphosphate and adenosine 5'-[gamma-thio]triphosphate also failed to inhibit the enzyme when substituted for ATP in the preincubation. The inhibition by MgATP was not affected by including cyclic AMP in the preincubation, but Ca2+ ions alone or plus diacylglycerol in the preincubation increased the inhibition slightly. The inhibition was abolished by including an inhibitor of cyclic-AMP-dependent protein kinase in the preincubation. These observations, taken collectively, point to the inhibition of foetal pulmonary cytidylyltransferase through the phosphorylation of a protein and suggest that this key enzyme in lung surfactant production may be regulated through this mechanism.

Substrate specificities and structure-activity relationships for acylation of antibiotics catalyzed by kanamycin acetyltransferase.

Antibiotic resistance caused by the presence of the plasmid pMH67 is mediated by the aminoglycoside acetyltransferase AAC(6')-4, also known as kanamycin acetyltransferase. Bacteria harboring the plasmid are resistant to the kanomycins plus a broad range of other deoxystreptamine-containing aminoglycosides but not to the gentamicins XK62-2 and C1 which are substituted at the 6'-position. Substrate specificity studies on the purified enzyme, however, now show that the enzyme acetylates an even broader range of aminoglycosides, including the gentamicins XK62-2 and C1. The enzyme also accepts several acyl-CoA esters, which differ in nucleotide as well as in acyl chain length. Application of the method of analysis of structure-activity data developed earlier for gentamicin acetyltransferase [Williams, J. W., & Northrop, D. B. (1978) J. Biol. Chem. 253, 5908-5914] to the kinetic data obtained for AAC(6')-4 shows that the turnover of the acylation reaction is limited by catalysis and not by the rate of release of either the acetylated antibiotic or CoA. Most structural changes in aminoglycosides cause changes in rates of release, and only drastic changes, near the 6'-amino group, affect catalysis. The structural requirements on aminoglycosides for enzymatic activity run parallel to the structural requirements for antibacterial activity.

The kinetic mechanism of kanamycin acetyltransferase derived from the use of alternative antibiotics and coenzymes.

Kinetic data for the 6'-aminoglycoside-modifying enzyme, AAC(6')-4, also named kanamycin acetyltransferase, have been collected for six aminoglycoside antibiotics (amikacin, gentamicin C1a, kanamycin A, neomycin B, sisomicin, and tobramycin) and three coenzymes (acetyl-CoA, n-propionyl-CoA, and n-butyryl-CoA). The initial velocity pattern using acetyl-CoA favors a ping-pong kinetic mechanism (Kia = -0.34 +/- 0.34 microM), but the pattern using n-propionyl-CoA supports a sequential one (Kia = 2.7 +/- 0.8 microM). Kinetic analyses using alternative substrates confirm the sequential mechanism and, moreover, clearly identify a random order of addition of antibiotic and coenzyme because V/K values of antibiotics varied 40-fold as the identity of the coenzyme was changed and V/K values of coenzyme varied 13-fold as the identity of the antibiotic was changed. One or both sets of values would have remained unchanged if the mechanism were either ordered sequential or ping-pong. Combining these results with structure-activity data which argue for a rapid rate of release of substrates and products relative to the rate of enzymatic turnover (Radika, K., and Northrop, D. B. (1984) Biochemistry 25, in press) establishes the kinetic mechanism as rapid equilibrium random sequential.

A new kinetic diagnostic for enzymatic mechanisms using alternative substrates.

When the concentration of substrate A is varied against fixed and saturating concentrations of alternative substrates B, B', B", etc., and B is similarly varied against alternatives of A, unique pairs of kinetic patterns are generated for each of the major classes of kinetic mechanism for enzymatic catalysis involving two substrates. Thus, a determination of mechanism may be obtained from just two kinetic patterns. The diagnostic has the following advantages over previous usages of alternative substrates and the traditional approach using initial velocity patterns: efficiency (fewer assays are needed), accuracy (large changes in slopes and intercepts are possible), and precision of assays (one substrate is always saturating).

Purification of two forms of kanamycin acetyltransferase from Escherichia coli.

Kanamycin acetyltransferase acylates aminoglycoside antibiotics using acetyl-CoA, and thereby conveys bacterial resistance to several clinically important antibiotics, notably amikacin. The enzyme was quantitatively and reproducibly released from Escherichia coli W677 harboring plasmid pMH67 by a modified osmotic shock procedure (bacterial cells are incubated overnight in sucrose and again without sucrose before onset of osmotic shock). The enzyme was purified by dye-ligand chromatography on Affi-Gel Blue in addition to antibiotic affinity chromatography on neomycin-Sepharose-4B. The activity did not increase with subsequent chromatography on ion-exchange, hydrophobic, or molecular-exclusion gels. However, both dye-ligand and molecular-exclusion chromatography, as well as disc-gel electrophoresis, separated the purified enzyme equally into two active protein fractions. Based on the more active of the two forms, the purification was 112-fold with a specific activity of 1.9 IU/mg. The less-active form has an unusual absorbance spectrum, with a maximum near 255 nm, which cannot be explained by the amino acid composition. Chromatography of this form alone regenerated both forms, suggesting that the enzyme is noncovalently conjugated to an uncharged chromophore, such as a lipid. The purified enzyme has a very sharp pH optimum at 5.5 with a plateau on the alkaline side, but is most stable between pH 8.5 and 9.5. Data from electrophoresis in the presence of sodium dodecyl sulfate and gel-filtration on Ultrogel AcA 44 are consistent with a tetrameric protein of 60-70,000 Da.

Correlation of antibiotic resistance with Vmax/Km ratio of enzymatic modification of aminoglycosides by kanamycin acetyltransferase.

Kinetic data for the antibiotic-modifying enzyme kanamycin acetyltransferase AAC(6')-IV have been determined for five aminoglycoside antibiotics (amikacin, gentamicin C1a, kanamycin A, sisomicin, and tobramycin) and compared with close-interval determinations of the minimal inhibitory concentrations of the same antibiotics against Escherichia coli W677 harboring the resistance plasmid pMH67. These minimal inhibitory concentrations for the resistant bacteria varied from 80 to 800 micrograms/ml. Of the kinetic parameters Vmax, Km, and Vmax/Km ratio only Vmax/Km ratio had a linear correlation with minimal inhibitory concentrations (r = +0.818) at pH 7.8, where all antibiotics produced substrate inhibition, but not at pH 6.0, where they did not. The correlation with only Vmax/Km ratio has implications regarding the expression of resistance within the dynamics of the bacterial cell (i.e., antibiotic uptake versus modification), whereas substrate inhibition presents an opportunity to search for new chemotherapeutic agents which will combat resistance directly.