Estimation of spin probe clustering in biological membranes.

An iterative spectral subtraction technique has been developed which accurately estimates the proportion of 'dilute' and 'clustered' I(12, 3) (i.e., 5-nitroxide stearate) in human erythrocyte ghosts at 37 degrees C, even if subtractant spectra free from probe-probe interactions cannot be measured due to technical limitations. Gordon et al. ((1985) J. Membrane Biol. 84, 81-95) earlier showed that I(12, 3) occupies a class of high-affinity sites in ghosts at probe/total lipid ratios (P/L) less than 1/2250. Saturation occurs with increasing probe concentration, and, at higher loading, the probe inserts itself at initially dilute sites to form membrane-bound clusters of variable size. Although this model allows determination of the dilute/clustered probe ratio, it requires subtraction of experimental spectra with a 'magnetically dilute' spectrum obtained using P/L less than 1/4600. The new methodology accurately profiles the % probe clustering in human erythrocyte ghosts over the entire P/L range, even if the lowest P/L for the subtractant spectrum contains substantial probe-probe interactions (i.e., P/L of 1/604 or 1/303). Application of either the subtraction technique in Gordon et al. (1985) or the iterative subtraction protocol described here should allow determination of probe clustering in a wide range of I(12, 3)-labeled biological membranes.

Lipid domain formation and ligand-induced lymphocyte membrane changes.

Spectral parameters of spin-labelled phosphatidylcholine, ceramide and cerebroside in the plasma membranes of human blood lymphocytes were measured before and after treatment with various ligands, which included concanavalin-A and phytohemagglutinin. It was found that ligand treatment led to a significant decrease in order of the hydrocarbon chains of the phospholipids. This was accompanied by a clustering of the labelled spingolipids, as estimated by spin-spin interaction, and an increase in the order of their hydrocarbon chains. In the untreated cells the cerebroside fatty acid chain was more ordered than that of the phosphatidylcholine. It was considered that the decrease in phospholipid order was brought about by the sequestration of the more rigid sphingolipids into the patches and caps formed by receptor-ligand complexes. The significance of these changes in lipid distribution and ordering is discussed in relation to the activation of membrane enzyme systems by mitogenic ligands.

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.

The amino-terminal peptide of HIV-1 glycoprotein 41 fuses human erythrocytes.

The ability of synthetic peptides based on the amino-terminus of HIV-1 glycoprotein 41,000 (gp41) to fuse human erythrocytes was investigated. Previous site-directed mutagenesis studies have shown an important role for the N-terminal gp41 domain in HIV-fusion, in which replacement of hydrophobic amino acids with polar residues inhibits viral infection and syncytia formation. Here, a synthetic peptide (FP; 23 amino acid residues 519-541) corresponding to the N-terminus of HIV-1 gp41, and also a FP analog (FP526L/R) with Arg replacing Leu-526, were prepared with solid phase techniques. The lipid mixing and leakage of resealed ghosts triggered by these peptides were examined with fluorescence quenching techniques. Peptide-induced aggregation of human erythrocytes was studied using Coulter counter sizing and scanning electron microscopy (SEM). Using resealed erythrocyte ghosts at physiologic pH, FP induces rapid lipid mixing between red cell membranes at doses previously shown to hemolyze intact cells. FP also causes leakage from resealed ghosts, and promotes the formation of multicelled aggregates with whole erythrocytes. Contrarily, similar FP526L/R concentrations did not induce red cell lysis, lipid mixing, leakage or aggregation. Since the fusogenic potency of FP and FP526L/R parallels earlier gp41 mutagenesis studies showing that substitution of Arg for Leu-526 blocks fusion activity, these data suggest that the N-terminal gp41 domain in intact HIV participates in fusion.

The amino-terminal peptide of HIV-1 glycoprotein 41 lyses human erythrocytes and CD4+ lymphocytes.

Functional studies assessed the cytolytic activity of the amino terminal peptide (FP-I; 23 residues 519-541) of the glycoprotein 41,000 (gp41) of the Human Immunodeficiency Virus Type-1 (HIV-1). Synthetically prepared FP-I efficiently hemolyzed human red blood cells at 37 degrees C, with 40% lysis at 32 microM. Kinetic studies indicated that FP-I induced maximal hemolysis in 30 min, probably through tight binding of the peptide with the red cell membrane. The Phe-Leu-Gly-Phe-Leu-Gly (residues 526-531) motif in FP-I apparently plays a critical role in lysis of red cells, since no hemolytic activity was observed for an amino-acid-substituted FP-I in which the unique Phe-Leu-Gly-Phe-Leu-Gly was converted to Ala-Leu-Gly-Ala-Leu-Gly. As neither smaller constituent peptides (e.g., residues 519-524 and residues 526-536) nor a N-terminal flanking peptide (e.g., residues 512-523) induced red cell hemolysis, the entire 23-residue (519-541) sequence of FP-I may be required for hemolytic activity. FP-I was also cytolytic with CD4(+)-bearing Hut-78 cells, with 40% lysis at approx. 150 microM. These results are consistent with an earlier hypothesis that the N-terminal peptide of gp41 may partially contribute to the in vivo cytopathic actions of HIV-1 infection (Gallaher, W.R. (1987) Cell 50, 327-328).

The amino-terminal peptide of HIV-1 glycoprotein 41 interacts with human erythrocyte membranes: peptide conformation, orientation and aggregation.

Structural studies assessed interactions between the amino-terminal peptide (FP-I; 23 residues 519-541) of the glycoprotein 41,000 (gp41) of Human Immunodeficiency Virus Type-1 (HIV-1) and human erythrocyte membranes and simulated membrane environments. Peptide binding was examined at sub-hemolytic (approx. less than 5 microM) and hemolytic (greater than or equal to 5 microM) doses (Mobley et al. (1992) Biochem. Biophys. Acta 1139, 251-256), using circular dichroism (CD) and Fourier-transform infrared (FTIR) measurements with FP-I, and electron spin resonance (ESR) studies employing FP-I spin-labeled at either the amino-terminal alanine (FP-II; residue 519) or methionine (FP-III; position 537). In the sub-lytic regime, FP-I binds to both erythrocyte lipids and dispersions of SDS with high alpha-helicity. Further, ESR spectra of FP-II labeled erythrocyte ghosts indicated peptide binding to both lipid and protein. In ghost lipids, FP-II was monomeric and exhibited low polarity and rapid, anisotropic motion about its long molecular axis (i.e., alpha-helical axis), with restricted motion away from this axis. The spin-label at the amino-terminal residue (Ala-519) is insensitive to the aqueous broadening agent chromium oxalate and buried within the hydrophobic core of the membrane; the angle that the alpha-helix (residues 519-536) makes to the normal of the bilayer plane is either 0 degree or 40 degrees. Contrarily, ESR spectra of ghost lipids labeled with sub-lytic doses of FP-III indicated high mobility and polarity for the reporter group (Met-537) at the aqueous-membrane interface, as well as extreme sensitivity to chromium oxalate. At lytic FP-I doses, CD and FTIR showed both alpha-helix and beta-structure for peptide in ghost lipids or detergent, while ESR spectra of high-loaded FP-II in ghost membranes indicated peptide aggregates. Membrane aggregates of FP-I may be involved in hemolysis, and models are suggested for N-terminal gp41 peptide participation in HIV-induced fusion and cytolysis.

A spin label study of the ionic strength dependent conformational change in the human Ia molecule.

Purified human Ia molecules were labelled with maleimide or isothiocyanate spin labels or by reacting "TEMPAMINE" spin label with the neuraminic acid of their carbohydrate residues. It was found that increasing the ionic strength from 0.05 to 0.75 markedly increased the dipolar interaction between the maleimide-attached labels, but no effect was found of ionic strength or motion on dipolar interaction with the other two labels. The effect of increasing ionic strength could be blocked by the prior addition of Ia-specific antibody, but could not be reversed by the addition of antibody after ionic strength was increased. These findings complement an earlier finding that increasing ionic strength over the range 0.05-0.75 has an inhibitory effect on the combination of Ia with its antibody. Because the maleimide spin labels attach predominantly to SH groups it is suggested that increasing ionic strength causes conformational changes in the immunoglobulin loop region which alter the accessibility of the Ia antibody-binding site.

Electron spin resonance spectroscopy in the study of lymphoid cell receptors.

ESR spectroscopy is a sensitive method which can be used to monitor a wide range of changes in and near the plasma membranes of lymphoid cells during such events as ligand-induced receptor patching and capping, activation and endocytosis and phagocytosis. The advantages of the technique are rapidity, sensitivity, small sample size (10(6)-10(7) cells), and nondestructive nature. Spin labels that are attached to a range of intrinsic and extrinsic molecules give information about the fluidity and polarity of the environment in which they are located. Because of the occurrence of interaction when probes are sequestered in restricted regions of the membrane, ESR spectroscopy is also a valuable technique for measuring the formation of domains in the cell membrane.

The amino-terminal peptide of HIV-1 gp41 interacts with human serum albumin.

Structural and functional studies were made to assess interactions between human serum albumin (HSA) and the amino-terminal peptide (FP-I; 23-residue peptide 519-541) of glycoprotein 41,000 (gp41) of human immunodeficiency virus type-1 (HIV-1). Circular dichroism (CD) spectroscopy indicated that the peptide binds to albumin with dominant alpha-helical character. Peptide binding to albumin was also examined using FP-I spin labeled at either the amino-terminal alanine (FP-II; residue 519) or methionine (FP-III; position 537). Electron spin resonance (ESR) spectra of FP-II bound to HSA at 38 degrees C indicated that the spin label at the amino-terminal residue (Ala-519) was motionally restricted. The ESR spectrum of 12-nitroxide stearate (12-NS)-labeled HSA was identical to that obtained with FP-II, indicating that the reporter groups for the 12-NS and FP-II probes are similarly bound to albumin. Contrarily, ESR spectra of HSA labeled with FP-III indicated high mobility for the reporter group (Met-537) at the aqueous-protein interface. This suggests that the N-terminal gp41 peptide binds as an alpha helix (residues 519-536) to fatty acid sites on HSA, such that Ala-519 of the peptide residues in the interior of the protein while Met-537 lies outside the protein in aqueous solution. It is also of interest that addition of HSA to human red blood cells dramatically reduced the ability of FP-I to induce hemolysis, presumably through peptide-albumin binding that inhibited FP-I interactions with red cell membranes. The significance of these results focuses on the following three points. The first is that high serum levels of albumin may limit the efficacy of anti-HIV therapies using peptides based on the N-terminal gp41 domain. The second is that the elucidation of FP-I and HSA interactions with physical techniques may provide clues on the molecular features underlying viral FP-I combination with receptors on the target cell surface. Last, the affinity of albumin for the N-terminal gp41 peptide may play a subordinate role in the blocking of HIV infectivity in vitro that has been reported for chemically modified albumins.

Antivirals that target the amino-terminal domain of HIV type 1 glycoprotein 41.

Functional and structural studies were made to assess whether a class of antiviral agents targets the N-terminal domain of the glycoprotein 41,000 (gp41) of human immunodeficiency virus type 1 (HIV-1). Previous experiments have shown that the amino-terminal peptide (FP-I; 23 amino acids, residues 519-541) of HIV-1 gp41 is cytolytic to both human erythrocytes (non-CD4+ cells) and Hut-78 cells (CD4+ lymphocytes). Accordingly, FP-I-induced hemolysis may be used as a surrogate assay for evaluating the role of the N-terminal gp41 domain in HIV-cell interactions. Here, we studied the blocking of FP-I-induced lysis of erythrocytes by the following anti-HIV agents: (1) IgG [i.e., anti-(518-541) IgG] raised to an immunoconjugate of Arg-FP-I, (2) apolipoprotein A-1 (apo A-1) and a peptide based on apo A-1, (3) dextran sulfate, (4) gp41 peptide (residues 637-666), and (5) anionic human serum albumins. Dose-response curves indicated that their relative potency in inhibiting FP-I-induced hemolysis was approximately correlated with their previously reported anti-HIV activity. Electron spin resonance (ESR) studies showed that FP-I spin labeled at the N-terminal alanine binds to anti-(518-541) IgG, dextran sulfate, and anionic albumins. The high in vitro antiviral activity and low cytotoxicity of these agents suggest that blocking membrane-FP-I interactions offers a novel approach for AIDS therapy or prophylaxis.

Cytotoxic activity of the amino-terminal region of HIV type 1 Nef protein.

Myristoylated 21- and 25-residue N-terminal peptides of the Nef protein of HIV-1 lysed human erythrocytes and were cytotoxic toward a human CD4+ T cell line, CEM, and primary human peripheral blood mononuclear cells (PBMCs). The corresponding nonmyristoylated N-terminal peptides were only very weakly hemolytic and cytotoxic. A myristoylated peptide consisting of residues 31-50 of Nef was neither hemolytic nor cytotoxic. Alteration of the tryptophan residue at position 13 to a serine did not change the hemolytic and cytotoxic activity. Studies of the ultraviolet fluorescence of the tryptophan at position 5 in the peptide, using an artificial membrane system and fluorescence-quenching agents that inserted into the bilayer at different levels, suggested that myristoylation results in this residue being brought into contact with the upper hydrocarbon region of the lipid bilayer of the cell membrane. This tryptophan is flanked by a number of polar residues that would maintain it in this position, resulting in a considerable increase in disorder in the upper regions of the lipid bilayer, leading to its destabilization and to lysis. The cytotoxic activity of the myristoylated Nef fragments may, in part, explain the killing and deletion of cells, especially in lymphoid tissues, during HIV infection.

Structural requirements for the cytotoxicity of the N-terminal region of HIV type 1 Nef.

We have found that the hemolytic and cytotoxic activities of myristoylated Nef N-terminal peptides require a net positive charge in the first seven amino residues of the sequence. The activities are considerably less dependent on the secondary structure of the peptides. Film balance studies showed that both active and inactive peptides interacted with neutral phospholipid monolayers, suggesting that binding to neutral lipids was not a sufficient condition for lytic activity. It was also found that nonmyristoylated N-terminal peptide did not interact to the same extent with the monolayer, indicating that myristoylation was essential for lipid interaction. It is considered that the positively charged residues of the proximate N terminus of Nef interact with acidic lipids of biological membranes, reinforcing the weak membrane-targeting properties of the myristyl chain. Parallels are drawn between this mode of interaction with membranes and that of members of the Src family of proteins, which are also myristoylated and have positively charged residues in their proximate N termini. In particular, these proteins and Nef also have serine residues in their proximal N-terminal regions, which when phosphorylated could neutralize the positive charge and thus provide a mechanism for modulating membrane interaction.

Fusogenic activity of amino-terminal region of HIV type 1 Nef protein.

We have studied two isoforms of Nef, Nef-27 and Nef-25, which were produced in E. coli. Nef-25 lacked the first 18 N-terminal residues of Nef-27 and both were nonmyristylated. Nef-27 fuses small unilamellar dipalmitoyl phosphatidylcholine vesicles (SUVs), as indicated by enhanced light scattering of SUVs and lipid mixing using concentration-dependent fluorescence dequenching. Nef-27 also causes the appearance of a shifted isotropic peak in the 31P NMR spectra of these vesicles, suggesting that protein interactions induce nonlamellar lipid structures. Recombinant Nef-25, which lacks only the 18 N-terminal residues of Nef-27, does not fuse vesicles and has little effect on the 31P NMR spectra. On the other hand, synthetic peptides consisting of 18 or 21 of the N-terminal residues of Nef-27 are strongly membrane perturbing, causing vesicle fusion and inducing isotropic peaks in the 31P NMR spectrum. Endogenous fluorescence spectra of the N-terminal peptide (21 residues) with SUVs show that the N-terminal sequence of Nef may achieve these perturbing effects by inserting its hydrophobic side into the lipid bilayer. Theoretical calculations using hydrophobic moment plot analysis indicate that short-length stretches (i.e., six amino acid residues) of the N-terminal sequence may insert into the lipid bilayer as multimeric alpha helices or beta sheets. The above-described membrane activities of Nef-27, which principally reside in its N-terminal domain, may play critical role(s) in certain functional properties of the full-length protein. For example, the fusogenic activity of the N-terminal sequence may be involved in the extracellular release of Nef-27, much of which appears to be associated with small membrane vesicles. The fusion activity may also be relevant to the ability of Nef-27 to downregulate CD4 and IL-2 receptors when this protein is electroporated into cultured lymphocytes, an activity not possessed by Nef-25.

Efficacy of fusion peptide homologs in blocking cell lysis and HIV-induced fusion.

Contrary to earlier reports, we have found that tri- and hexapeptides analogous or homologous with segments of the 23-residue N-terminal fusion sequence (FS) of the viral transmembrane glycoprotein gp41 (residues 517-539) did not significantly inhibit HIV-1-induced syncytium formation, using an uninfected cell-infected cell fusion assay. In contrast, we found that the high molecular weight apolipoprotein A-1 and a 23-residue analog of the FS, with the phenylalanine residues at positions 524 and 527 replaced with alanine residues, were effective inhibitors. Although the tripeptides were ineffective as inhibitors of syncytium formation, we found a number of them inhibited red cell lysis induced by the synthetic peptide AVGIGALFLGFLGAAGSTMGARS (based on the HIV-1 gp41 FS). This effect was also seen with apolipoprotein A-1. The Ala524,527 analog of the fusion sequence could not be tested in this system because it was hemolytic. We concluded that the smaller peptides were effective inhibitors of hemolysis because they interfered with pore formation by the fusion sequence peptide, either by disrupting the pores or by preventing the peptide from adopting the alpha-helical conformation found in the pores. On the other hand, membrane fusion, which is a prelude to syncytium formation, has been shown to require the fusion sequence in the beta-strand conformation. We argue that small peptides would be unable to block interaction between such strands, although larger molecules, such as apolipoprotein A-1 and the Ala524,527 analog, would be able to do so and thus inhibit fusion. It seems, therefore, that a successful drug directed against the FS-cell membrane interaction stage of syncytium formation would need to be of relatively high molecular weight and complexity.

Decreased synthesis of hepatic satellite DNA in pyrrolizidine alkaloidosis of the sheep.

the proportion of heavy satellite in the DNA isolated from the livers of sheep suffering from experimental pyrrolizidine alkaloidosis is significantly lower (3.5%) than that found in the DNA from lovers of normal sheep (12%). Dehydroheliotridine, the major unbound, relatively stable metabolite of lasiocarpine and heliotrine, the alkaloids used in the study, was found to inhibit selectively the semiconservative replication of the satellite DNA in cultures of ovine kidney cells. It is suggested that the inhibition of the synthesis of satellite DNA may be related to an attack by the metabolite on the pericentromeric region where the majority of the satellite sequences are located.

The binding of dehydroheliotridine to DNA and the effect of it and other compounds on repair synthesis in main and satellite band DNA.

This study was aimed at elucidating the mechanisms of the preferential depression of satellite DNA synthesis by dehydroheliotridine (DHH). DHH was found to induce repair synthesis to the same extent in both main and satellite band DNA in cultured sheep lymphocytes. This was also the case with acridine orange, nitrogen mustard (HN2) and ethyl methane-sulphonate (EMS). Using analytical equilibrium ultracentrifugation no difference was found between the extents of in vitro binding of DHH by main and satellite band DNA. From these results it was concluded that the depression of the synthesis of satellite DNA could not be explained by either its preferential binding of DHH or by less effective repair mechanisms. Radiolabelled DHH when added to synchronized cultures of ovine kidney cells was found to be preferentially bound to the satellite DNA (one DHH molecule to 6000 nucleotides) compared with the main band DNA (one to 10 000). When 5-bromodeoxyuridine (BUdR) was added to the cultures no DHH label was found in the heavy, semiconservatively replicated DNA band. From these findings it is suggested that attack may occur during mitosis where all of the satellite DNA may be undergoing synthesis at the same time, thus explaining the increased amount of DHH bound to the satellite.

ESR spectroscopy in the study of antigen processing--uptake of spin-labelled antigens by macrophages.

Macrophages were briefly pulsed with a spin-labelled synthetic polypeptide, poly(L-tyrosine:L-glutamic acid) poly DL-alanine:poly L-lysine (n-TGAL) in the presence and absence of anti-TGAL-antibody, and the electron spin resonance (ESR) spectra of the cell suspension compared with the spectrum of free n-TGAL in solution. Spectral analysis indicated two cell-associated n-TGAL pools, one composed of freely rotating label held in an aqueous environment, susceptible to protease digestion and ascorbate reduction, and a second highly concentrated pool, sequestered intracellularly, and held within a highly ordered, polar microenvironment. The ESR analyses were completed within minutes of antigen pulsing, employed very small numbers of live cells, and did not damage the cells being tested. The utility of the technique in screening fatty acid-antigen conjugates for macrophage uptake was demonstrated.