Studies of the association and conformational properties of metal-free insulin in alkaline sodium chloride solutions by one- and two-dimensional 1H NMR.

One- and two-dimensional 1H NMR spectroscopy have been employed to probe the association and subsequent conformational changes of metal-free insulin in sodium chloride solution at pH 9 and 9.4. These studies establish that the proton resonances of His(B5) and His(B10) are useful signatures of aggregation and conformation. Changes in chemical shifts and areas of resonances due to the C2 protons of His(B10) and His(B5) and transfer of magnetization experiments served to identify the association as the assembly of tetramer from dimers under our experimental conditions (pH 9.4, [insulin] greater than 1 mM, [NaCl] = 0.1 M). Sodium chloride also alters the equilibrium distribution of species in favor of a tetrameric species. The association equilibrium constant was estimated from area measurements to be approximately 5 x 10(3) M-1 at pH 9.4, 26 +/- 0.1 degrees C, and 0.1 M sodium chloride. Under conditions of 0.1 M sodium chloride concentration, nuclear Overhauser effect experiments in the one- and two-dimensional modes revealed an operative nuclear Overhauser effect between the His(B5) C2 protons and the 2,6 ring protons of a Tyr residue provisionally assigned as Tyr(B16). We conclude that this interaction is a diagnostic signature of a conformational transition whereupon an extended chain from residues B1 to B9 (T-state) is transformed into an alpha-helix (R-state) thus bringing the rings of His(B5) and Tyr(B16) from adjacent subunits across the monomer-monomer interface into van der Waals contact. This conformational flexibility is an added consideration to the discussion of the relevant structure of insulin for receptor binding.     

A novel CoA-independent transacetylase produces the ethanolamine plasmalogen and acyl analogs of platelet-activating factor (PAF) with PAF as the acetate donor in HL-60 cells.

In this study, we demonstrate the presence of a unique membrane-associated transacetylase that transfers the acetate group from platelet-activating factor (PAF) to lysoplasmalogen (in the presence of EDTA and sodium acetate) with the formation of 1-alk-1-enyl-2-acetyl-sn-glycero-3-phosphoethanolamine (alk-1-enylacetyl-GPE). The identity of alk-1-enylacetyl-GPE was confirmed by acid hydrolysis, phospholipases A2 or C treatment and derivatization by fluorodinitrobenzene. The transacetylase has no requirement for Ca2+, Mg2+, or CoA and a broad pH optimum (7.0-8.0) with Km values of 12.0 microM for PAF and 106.4 microM for lysoplasmalogens. The enzyme activity from the isolated membrane fraction is not changed when whole cells are supplemented with 20:4, induced to differentiate into granulocytes, or treated with ionophore A23187. Radyllyso-sn-glycero-3-phosphocholine (GPC), radyllyso-GPE, acyllyso-sn-glycero-3-phosphoserine (GPS), acyllyso-sn-glycero-3-phosphoinositol (GPI), alkyllyso-sn-glycero-3-phosphate (GP), acyllyso-GP, or cis-9-octadecen-1-ol can also serve as acetate acceptors, whereas alkylglycerol, acylglycerol, or cholesterol are inactive. Differences in substrate acceptor specificity, sensitivity toward phenylmethylsulfonyl fluoride, and response to temperature suggest that the CoA-independent transacetylase and the CoA-independent transacylase that transfers long-chain acyl moieties are two separate enzymes. With intact differentiated HL-60 cells, [3H]acetate from [3H]PAF can be incorporated into alk-1-enylacetyl-GPE in the presence of ionophore A23187, but not in its absence. Moreover, phospholipase A2 inhibitors (p-bromophenacyl bromide and mepacrine) block the transacetylation process in whole cell system. These results indicate the production of alk-1-enyllyso-GPE is a rate-limiting factor for the subsequent transacetylation step during cell activation. We conclude that the transacetylase may participate in the biosynthesis of ethanolamine plasmalogen and acyl analogs of PAF, in vivo, fine-tuning of PAF biological responses, and cross-talk between de novo and remodeling pathways of PAF biosynthesis.     

Location of a Bombyx mori receptor binding region on a Bacillus thuringiensis delta-endotoxin.

Receptor binding studies were performed with 125I-labeled trypsin-activated insecticidal toxins, CryIA(a) and CryIA(c), from Bacillus thuringiensis on brush-border membrane vesicles (BBMV) prepared from Bombyx mori larval midgut. Bioassays were performed by gently force feeding B. mori with diluted toxins. CryIA(a) toxin (LD50; 0.002 micrograms) was 200 times more active against B. mori larvae than CryIA(c) toxin (LD50; 0.421 micrograms) and showed high-affinity saturable binding. The Kd and the binding site concentration for CryIA(a) toxin were 3.5 nM and 7.95 pmol/mg, respectively. CryIA(c) toxin (Kd, 50.35 nM; Bmax, 2.85 pmol/mg) did not demonstrate high-affinity binding to B. mori BBMV. Control experiments with CryIA(a) and CryIA(c) toxins revealed no binding to mouse small intestine BBMV and nonspecific binding to pig kidney BBMV. These data provide evidence that binding to a specific receptor on the membrane of midgut epithelial cells is an important determinant with respect to differences in insecticidal spectrum of insecticidal crystal proteins. To locate a B. mori receptor binding region on the CryIA(a) toxin, homologous and heterologous competition binding studies were performed with a set of mutant proteins which had previously been used to define the B. mori "specificity domain" on this toxin (Ge, A. Z., Shivarova, N. I., and Dean, D. H. (1989) Proc. Natl. Acad. Sci. U.S.A. 86, 4037-4041). These mutant proteins have had regions of their genes reciprocally exchanged with the cryIA(c) gene. A B. mori receptor binding region on CryIA(a) toxin includes the amino-terminal portion of the hypervariable region, amino acids 332-450, which is identical to the previously described B. mori specificity determining region. These data provide direct evidence that delta-endotoxins contain a tract of amino acids that comprise a binding region and as a results determines the specificity of a toxin.     

Purification and identification of the heat-stable factor required for pregnenolone-binding protein activity. Evidence that the factor is adenosine 3',5'-diphosphate.

This paper presents data identifying adenosine 3',5'-diphosphate (3',5'-ADP) as the small heat-stable factor essential for the active steroid binding complex of the adrenocortical pregnenolone-binding protein (PBP). Factor activity obtained from the boiled supernatant of partially purified PBP was isolated by high performance liquid chromatography using weak anion-exchange and hydrophobic (C18) chromatography sequentially. The purified material retained characteristic factor activity and presented a UV spectrum identical to that for authentic 3',5'-ADP. Mass spectroscopic analysis of the isolated factor revealed an M-H ion of appropriate mass (m/z = 426) and a decomposition pattern for the M-H ion that was consistent with the structure of 3',5'-ADP. The studies presented here demonstrate that authentic 3',5'-ADP can categorically substitute for factor prepared from the soluble fraction of the guinea pig adrenal. Specifically, 3',5'-ADP potentiated ligand binding of partially purified native PBP and restored binding capacity to alkaline phosphatase-inactivated PBP in a dose-dependent manner. As is the case for adrenocortical factor activity, these effects were negated by pretreating the 3',5'-ADP with calf intestinal alkaline phosphatase. Other nucleotides similarly tested, including ADP isomers, were ineffective as factor substitutes. The sulfated form of 3',5'-ADP (i.e. 3'-phosphoadenosine 5'-phosphosulfate) demonstrated some potential for restoring binding capacity to phosphatase-inactivated PBP; however, this compound was clearly inhibitory rather than stimulatory for native PBP activity. Taken collectively, the data overwhelmingly demonstrate that 3',5'-ADP is in fact the molecule required by the PBP for high affinity steroid binding complex formation. It is not yet known whether 3',5'-ADP acts allosterically or contributes directly to the structure of the steroid binding site.     

On the location and function of tyrosine beta 331 in the catalytic site of Escherichia coli F1-ATPase.

1) Using a combination of site-directed mutagenesis and fluorescence spectroscopy we have studied the location and function of residue beta Y331 in the catalytic site of Escherichia coli F1-ATPase. The fluorescent analog lin-benzo-ADP was used as a catalytic-site probe, and was found to bind to three sites in normal F1, with Kd1 = 0.20 microM and Kd2,3 = 5.5 microM. lin-Benzo-ATP was a good substrate for hydrolysis. 2) The mutants investigated were beta Y331F, L, A and E. kcat/KM for ATP hydrolysis in purified F1 was reduced according to the series Y greater than or equal to F greater than L greater than A greater than E, with E being severely impaired; concomitant decreases in binding affinity for lin-benzo-ADP were seen. 3) Fluorescence properties of lin-benzo-ADP bound to F1 differed widely, depending on the residue present at position beta 331. Red shifts of excitation and emission spectra occurred with F and L residues, but not with Y, A, or E. There was strong quenching of fluorescence with wild-type (Y), partial quenching with A, and no quenching with F, L, or E. 4) We conclude that (a) the environment around the bound adenine moiety in the catalytic site is nonpolar, (b) residue beta 331 is part of the adenine-binding subdomain and when tyrosine is the residue, the phenolic hydroxyl makes direct interaction with the fluorophore, (c) an aromatic residue is not absolutely required at position beta 331 for catalytic function, but an increase in polarity leads to functional impairment, and (d) in terms of fluorescence response of bound lin-benzo-ADP all three catalytic sites behaved the same. 5) F1 from mutant beta Y297F bound lin-benzo-ADP with the same fluorescence and binding characteristics as normal F1, and catalytic properties were similar to normal. Therefore, there was no reason to conclude that residue beta Y297 is involved in binding the adenine moiety of ATP.     

Distinct domains of an oligotopic membrane protein are Sec-dependent and Sec-independent for membrane insertion.

Leader peptidase of Escherichia coli spans the plasma membrane twice with its amino terminus on the periplasmic surface of the membrane and its large carboxyl-terminal domain protruding into the periplasm. To monitor the transfer of the amino terminus of leader peptidase to the periplasm, we have constructed a fusion protein between the 18-residue amino-terminal periplasmic domain of Pf3 bacteriophage coat protein and the beginning of leader peptidase. We find that neither the SecA or SecY proteins nor a transmembrane electrochemical potential is required for insertion of the amino terminus, while the transfer of the carboxyl-terminal domain of leader peptidase has these requirements. The first 35 residues of leader peptidase, which include the first hydrophobic domain and the carboxyl-terminal positively charged cluster, are sufficient to insert the amino terminus. When positively charged residues are introduced before the first transmembrane segment, translocation of the amino terminus is abolished. These studies in protein membrane topogenesis, showing that there are different requirements for amino and carboxyl termini insertion, indicate that multiple mechanisms exist even within the same protein.     

Regulation of vascular cell adhesion molecule 1 on human dermal microvascular endothelial cells.

Vascular endothelial cell adhesion molecule 1 (VCAM-1) is an adherence molecule that is induced on endothelial cells by cytokine stimulation and can mediate binding of lymphocytes or tumor cells to endothelium. Because these interactions often occur at the level of the microvasculature, we have examined the regulation of expression of VCAM-1 in human dermal microvascular endothelial cells (HDMEC) and compared it to the regulation of VCAM-1 in large vessel human umbilical vein endothelial cells (HUVEC). Both cell populations were judged pure as assessed by expression of von Willebrand factor and uptake of acetylated low density lipoprotein. Expression of VCAM-1 was not detectable on either unstimulated HDMEC or HUVEC when assessed by ELISA or flow cytometry. Stimulation of either HDMEC or HUVEC with TNF-alpha resulted in a time- and dose-dependent induction of VCAM-1. However, although TNF-alpha-induced cell surface and mRNA expression of VCAM-1 in HDMEC was transient, peaking after 16 h of stimulation, TNF stimulation led to persistently elevated cell surface expression of VCAM-1 on HUVEC. IL-1 alpha also induced cell surface expression of VCAM-1 on HUVEC in a time- and dose-dependent manner, but stimulation of HDMEC with IL-1 alpha at doses up to 1000 U/ml failed to induce significant cell surface expression. However, IL-1 alpha induced time- and dose-dependent increases in ICAM-1 on HDMEC. Similarly, IL-4 induced VCAM-1 expression and augmented TNF-alpha-induced expression on HUVEC but did not affect VCAM-1 expression on HDMEC. Binding of Ramos cells to cytokine-stimulated endothelial cell monolayers correlated with VCAM-1 induction. Increased binding was seen after stimulation of HDMEC with TNF-alpha, which was blocked by anti-VCAM-1 mAb, but no increases in binding were noted after stimulation of HDMEC monolayers with IL-1 alpha. These data provide additional evidence for the existence of endothelial cell heterogeneity and differences in cell adhesion molecule regulation on endothelial cells derived from different vascular beds.