Chorismate mutase:prephenate dehydratase from Acinetobacter calcoaceticus. Purification, properties and immunological cross-reactivity

The bifunctional P protein (chorismate mutase: prephenate dehydratase) from Acinetobacter calcoaceticus has been purified. It was homogeneous in polyacrylamide gels and was more than 95% pure on the basis of the immunostaining of purified P protein with the antibodies raised against the P protein. The native enzyme is a homodimer (Mr = 91,000) composed of 45-kDa subunits. A twofold increase in the native molecular mass of the P protein occurred in the presence of L-phenylalanine (inhibitor of both activities) or L-tyrosine (activator of the dehydratase activity) during gel filtration. Chorismate mutase activity followed Michaelis-Menten kinetics with a Km of 0.55 mM for chorismate. L-Phenylalanine was a relatively poor non-competitive inhibitor of the mutase activity. The chorismate mutase activity was also competitively inhibited by prephenate (reaction product). Substrate-saturation curves for the dehydratase activity were sigmoidal showing positive cooperativity among the prephenate-binding sites. L-Tyrosine activated prephenate dehydratase strongly but did not abolish positive cooperativity with respect to prephenate. L-Phenylalanine inhibited the dehydratase activity, and the substrate-saturation curves became increasingly sigmoidal as phenylalanine concentrations were increased with happ values changing from 2.0 (no phenylalanine) to 4.0 (0.08 mM L-phenylalanine). A sigmoidal inhibition curve of the dehydratase activity by L-phenylalanine gave Hill plots having a slope of -2.9. Higher ionic strength increased the dehydratase activity by reducing the positive cooperative binding of prephenate, and the sigmoidal substrate-saturation curves were changed to near-hyperbolic form. The happ values decreased with increase in ionic strength. Antibodies raised against the purified P protein showed cross-reactivity with the P proteins from near phylogenetic relatives of A. calcoaceticus. At a greater phylogenetic distance, cross-reaction was superior with P protein from Neisseria gonorrhoeae than with that from the more closely related Escherichia coli.     

Modulation of dolichyl-phosphomannose synthase activity by changes in the lipid environment of the enzyme

Rat liver dolichyl-phosphomannose synthase (GDP mannose-dolicholphosphate mannosyltransferase; EC 2.4.1.83) was previously shown to catalyze optimal rates of mannosyl transfer to dolichyl-P when the polyprenol acceptor was incorporated into a phosphatidylethanolamine (PE) matrix that has a tendency to adopt a nonbilayer (hexagonal HII) phase [Jensen, J. W., & Schutzbach, J. S. (1985) Eur. J. Biochem. 153, 41-48]. The present investigations now further define the properties of the lipid environment that are essential for mannosyltransferase activity. Monogalactosyl diglyceride (MGDG), a glycoglycerolipid that prefers a nonbilayer-phase organization in isolation, was shown to provide a suitable lipid matrix for synthase activity. By comparison, the enzyme was not activated by digalactosyl diglyceride (DGDG), which forms stable bilayer structures upon hydration. Enzyme activity in MGDG/DGDG mixtures decreased as the proportion of DGDG in the dispersion was increased. Although bilayer-forming phospholipids supported low rates of mannosyl transfer, enzyme activity was stimulated by the addition of MGDG to either phosphatidylcholine (PC) or PE/PC (1:1) membranes. The incorporation of agents known to destabilize bilayer structures including dolichols, ubiquinone, dodecane, and cholesterol into PE/PC (1:1) membranes also increased the rate of mannosyl transfer. Enzyme activity in PC membranes was stimulated by the presence of gramicidin and also by greatly increased concentrations of the substrate, dolichyl-P. The results demonstrate that the enzyme does not have a requirement for PE and suggest that the physical state of the lipid matrix is an important determinant for reconstitution of the synthase and polyprenol phosphate substrate in a productive complex. The formation of an enzyme/lipid complex was demonstrated by sucrose density gradient centrifugation and could be correlated with the lipid requirements for enzyme activity.     

Thermotropic lipid phase separation in the human immunodeficiency virus

The presence of thermodependent lipid domains in the envelope of the human immunodeficiency virus (HIV) was studied. HIV was propagated in Hut-78 cells and purified by differential-gradient centrifugation. Since the virus was highly infectious in cell culture and Western blots of detergent-inactivated HIV showed envelope proteins when exposed to sera containing anti-HIV antibodies, this viral preparation was not deficient in 'spike' or 'knob' particles. Electron spin resonance (ESR) studies of intact HIV labeled with 5-nitroxide stearate (5-NS) indicated that a temperature-dependent lipid phase separation occurs with a high onset at approx. 42 degrees C and a low onset at approx. 15 degrees C. Cooling below 42 degrees C induces 5-NS clustering. Similar phase separations with high onsets at approx. 37-38 degrees C were previously identified in 5-NS labeled human erythrocytes (cholesterol/phospholipid (C/P) molar ratio = 0.90) and cholesterol-loaded (C/P = 0.85-0.98) rat liver plasma membranes. These were attributed to a temperature-sensitive redistribution of endogenous lipid components such that 5-NS is excluded from cholesterol-rich domains and tends to reside in cholesterol-poor domains at low temperatures. Since HIV has a lipid envelope with a similarly high C/P of 0.88 (Aloia et al. (1988) Proc. Natl. Acad. Sci. USA 85, 900-904), cholesterol-rich and cholesterol-poor domains also probably exist in HIV at physiologic temperatures. The reduced stability and infectivity of HIV noted on heating above 42 degrees C may be due, in part, to the abolition of these thermodependent domains.     

Physiochemical study of rock crab lipoproteins

Physicochemical studies have been carried out on the hemolymph and egg lipoproteins of the rock crab (Cancer antennarius). Analytical ultracentrifugal analyses of vitellogenic female HDL3 revealed the presence of two types of lipoproteins. The first with a sedimentation rate of 5.35 S was comparable to lipoproteins in male and non-vitellogenic female hemolymph. The second with a sedimentation rate of 10.74 S was comparable to the major lipoprotein of egg yolk. A similar comparison could be made following electrophoretic analyses in native polyacrylamide gels. Electrophoresis in SDS-polyacrylamide gels revealed three major apolipoproteins common to egg and vitellogenic HDL3. A fourth apolipoprotein was found in both male and female HDL3. In contrast to mammalian HDL, none of these crustacean apolipoproteins had a molecular weight less than 82 000. One of these apolipoproteins appears to be comparable physicochemically to the enteric form of apolipoprotein B in mammals.     

Preliminary X-ray studies of the tetra-heme cytochrome c3 and the octa-heme cytochrome c3 from Desulfovibrio gigas

A tetra-heme and an octa-heme cytochrome c3 from the sulfate bacterium Desulfovibrio gigas have been crystallized. Diffraction quality crystals of the tetra-heme cytochrome are obtained from solution by the addition of polyethylene glycol at pH 6.5. The crystals are orthorhombic, space group P2(1)2(1)2 with unit cell parameters a = 42.27 A, b = 52.54 A and c = 52.83 A. The octa-heme cytochrome crystals develop from low ionic strength solutions of phosphate or Tris-Cl in the pH range 6.2-7.6. The crystals belong to the trigonal system, space group P3(1) or the enantiomorph P3(2), with unit cell parameters a = b = 57.4 A, c = 97.3 A, gamma = 120 degrees. Single crystal diffraction studies of the structures of these two low-potential cytochromes are in progress.     

Insulin-specific suppressor T cell factors

Murine antibody responses to heterologous insulins are controlled by MHC-linked immune response genes. Although nonresponder mice fail to make antibody when injected with nonimmunogenic variants of insulin, we have recently shown that nonimmunogenic variants stimulate radioresistant, Lyt- 1+2- helper T cells that support secondary antibody responses. However, the helper activity can not be detected unless dominant, radiosensitive Lyt-1-2+, I-J+ suppressor T cells are removed. In this paper we report that extracts of primed Lyt-2+ suppressor T cells contain insulin-specific suppressor factors (TsF) that are capable of replacing the activity of suppressor T cells in vitro. The activity of these factors is restricted by MHC-linked genes that map to the I-J region, and immunoadsorption studies indicated that they bind antigen and bear I-J-encoded determinants. Insulin-specific TsF consists of at least two chains, one-bearing I-J and the other the antigen-binding site. Furthermore, mixing of isolated chains from different strains of mice indicates that the antigenic specificity is determined by the antigen-binding chain and the MHC restriction by the H-2 haplotype of the source of the non-antigen-binding, I-J+ chain. Moreover, mixtures containing antigen-binding chain from allogeneic cell donors and I-J+ chain from responder cell donors have activity in cultures containing responder lymphocytes. This suggests that preferential activation of suppressor T cells, rather than differential sensitivity to suppression, results in the nonresponder phenotype to insulin.     

Monoclonal antibodies recognizing single amino acid substitutions in hemoglobin

Four monoclonal antibodies (mAb) to non-human primate hemoglobin referred to as Cap-4, Cap-5, Rh-2, and Rh-4, and two mAb to human hemoglobin, referred to as H-1 and H-3 were isolated and were partially characterized. Binding studies with these mAb on a panel of hemoglobins and isolated alpha and beta globin chains revealed a unique reactivity pattern for each mAb. Amino acid sequence analysis of the antigens used to generate the binding data suggests that the specific recognition of certain hemoglobin antigens by each mAb is controlled by the presence of a particular amino acid at a specific position within the epitope. The use of synthetic peptides as antigens confirmed this observation for five of the mAb. No synthetic peptides were tested with the sixth mAb, Rh-2. The amino acids required for binding of mAb Cap-4, Cap-5, Rh-4, and Rh-2 to hemoglobin are alanine at beta 5, threonine at beta 13, glutamine at beta 125, and leucine at alpha 68. The non-human primate hemoglobin antibodies require a specific amino acid that is not present in human hemoglobin. The amino acid required for binding of Cap-4, Cap-5, and Rh-4 could arise by a single base change in the beta globin gene, whereas the amino acid required for Rh-2 binding would only occur if two base changes occurred. Thus these mAb are candidate probes for a somatic cell mutation assay on the basis of the detection of peripheral blood red cells that possess single amino acid substituted hemoglobin as a result of single base substitutions in the globin genes of precursor cells.     

Comparative action of glyphosate as a trigger of energy drain in eubacteria.

Escherichia coli, Bacillus subtilis, and Pseudomonas aeruginosa, each possessing a 5-enolpyruvylshikimate 3-phosphate synthase that is sensitive to inhibition by glyphosate [N-(phosphonomethyl)glycine], provide a good cross-section of organisms exemplifying the biochemical diversity of the aromatic pathway targeted by this potent antimicrobial compound. The pattern of growth inhibition, the alteration in levels of aromatic-pathway enzymes, and the accumulation of early-pathway metabolites after the addition of glyphosate were distinctive for each organism. Substantial intracellular shikimate-3-phosphate accumulated in response to glyphosate treatment in all three organisms. Both E. coli and P. aeruginosa, but not B. subtilis, accumulated near-millimolar levels of shikimate-3-phosphate in the culture medium. Intracellular backup of common-pathway precursors of shikimate-3-phosphate was substantial in B. subtilis, moderate in P. aeruginosa, and not detectable in E. coli. The full complement of aromatic amino acids prevented growth inhibition and metabolite accumulation in E. coli and P. aeruginosa where amino acid end products directly control early-pathway enzyme activity. In contrast, the initial prevention of growth inhibition in the presence of aromatic amino acids in B. subtilis was succeeded by progressively greater growth inhibition that correlated with rapid metabolite accumulation. In B. subtilis glyphosate can decrease prephenate concentrations sufficiently to uncouple the sequentially acting loops of feedback inhibition that ordinarily link end product excess to feedback inhibition of 3-deoxy-D-arabino-heptulosonate 7-phosphate synthase by prephenate. The consequential unrestrained entry is an energy-rich substrates into the aromatic pathway, even in the presence of aromatic amino acid end products, is an energy drain that potentially accounts for the inability of end products to fully reverse glyphosate inhibition in B. subtilis. Even in E. coli after glyphosate inhibition and metabolite accumulation were allowed to become fully established, a transient period where end products were capable of only partial reversal of growth inhibition occurred. The distinctive metabolism produced by dissimilation of different carbon sources also profound effects upon glyphosate sensitivity.     

Analysis of the spatial organization of microtubule-associated proteins

We have developed microdensitometer-computer correlation techniques to analyze the arrangement of microtubule arms and bridges (i.e., microtubule-associated proteins [MAPs]). A microdensitometer was used to scan immediately adjacent to the wall of longitudinally sectioned microtubules in positive transparency electron micrographs. Signal enhancement procedures were applied to the digitized densitometer output to produce a binary sequence representing the apparent axial spacing of MAP projections. These enhanced records were analyzed in two ways. (a) Autocorrelograms were formed for each record and correlogram peaks from a group of scans were pooled to construct a peak frequency histogram. (b) Cross-correlation was used to optimize the match between each enhanced record and templates predicted by different models of MAP organization. Seven symmetrical superlattices were considered as well as single axial repeats. The analyses were repeated with randomly generated records to establish confidence levels. Using the above methods, we analyzed the intrarow bridges of the Saccinobaculus axostyle and the MAP2 projections associated with brain microtubules synthesized in vitro. We confirmed a strict 16-nm axial repeat for axostyle bridges. For 26 MAP2 records, the only significant match was to a 12-dimer superlattice model (P less than 0.002). However, we also found some axial distances between MAP2 projections which were compatible with the additional spacings predicted by a 6-dimer superlattice. Therefore, we propose that MAP2 projections are arranged in a "saturated 12-dimer, unsaturated 6-dimer" superlattice, which may be characteristic of a wide variety of MAPs.