Influence of polymyxins on the structural dynamics of Escherichia coli lipid membranes.

Polymyxins are a family of nonribosomic cationic peptide antibiotics highly effective against Gram-negative bacteria. Two members of this family, Polymyxins B and E (PxB, PxE), form molecular vesicle-vesicle contacts and promote a selective exchange of phospholipids at very low concentrations in the membrane, a biophysical phenomenon that can be the basis of their antibiotic mode of action. To get more insight into the interaction of these antibiotics with the lipid membrane, their effect on the structural dynamics of bilayers prepared with lipids extracted from the membrane of Escherichia coli was determined using fluorescently labeled phopholipids. Steady-state anisotropy measurements with probes that localize at different positions in the membrane give information on the effects of polymyxins on the mobility of the phospholipids. Results with PxB, PxE, colymycin M and polymyxin B nonapeptide (PxB-NP), a deacylated derivative with no antibiotic properties, are compared. At low peptide concentrations (<2 mol%) PxB and PxE bind to the membranes superficially, affecting very slightly the ordering of the lipids at the outermost part of the bilayer. Above this concentration, PxB and PxE insert more deeply in the bilayer, increasing lipid order both in the gel and liquid-crystal states and modifying phase transitions. Fluorescence experiments with pyrene labeled phospholipids indicate that the increase in lipid packing is accompanied by an enrichment of phospholipids in the bilayers. In contrast, colymycin M and PxB-NP did not modify lipid packing or phase transition, nor did they induce microdomain formation. The possible significance of these results in the antibiotic mode of action of PxB and PxE is discussed. The combination of spectroscopic techniques described here can be useful as part of a general method of screening for new antibiotics that act on the membrane by the same mechanism as polymyxins.

NFAT1 enhances HIV-1 gene expression in primary human CD4 T cells.

Cyclosporin A (CsA) is a potent inhibitor of the NFAT family of transcription factors that enhance T cell activation. The observation that human immunodeficiency virus type 1 (HIV-1)-positive transplant recipients have a reduced HIV-1 viral burden during treatment with CsA suggested that NFAT may play a direct role in enhancing transcription of the HIV-1 viral genome. Two sets of NFAT binding sites were identified in the HIV-1 long terminal repeat (LTR) promoter by in vitro footprinting with full-length recombinant NFAT protein, and gel shift analysis of nuclear protein from polyclonally activated primary CD4 T cells revealed specific binding of NFAT1 to the NFkappaB binding sites of the HIV-1 LTR. Activation of primary CD4 T cells transiently transfected with a HIV-1 LTR luciferase reporter plasmid, lacking the NFAT binding sites in the upstream putative negative regulatory element but maintaining the NFkappaB/NFAT sites, demonstrated increased HIV-1 gene expression when cotransfected with a NFAT1 expression vector. Moreover, CsA, FK506, and a dominant-negative NFAT1 protein independently inhibited HIV-1 LTR promoter activity in CD4 T cells stimulated with phorbol ester and calcium ionophore. In primary human CD4 T cells, CsA also inhibited promoter activity directed by multimers of binding sites for NFAT, while having no effect on NFkappaB multimer-driven promoter activity. Increasing NFAT1 levels in CD4 T cells transiently transfected with a HIV-1 provirus also increased p24 protein expression. Thus, NFAT may be a target for prevention of HIV-1 LTR-directed gene expression in human CD4 T cells.

Analysis and comparison of the mouse and human immunoglobulin heavy chain JH-Cmu-Cdelta locus.

We report here 23,686 bases of contiguous DNA sequences from the mouse germline immunoglobulin heavy chain (H) constant (C) mu delta region. The sequence spans the joining (JH) regions, the mu constant region (C mu), the delta constant region (C delta) coding regions, a domain relic, the mu switch region (S mu), seven blocks of simple sequence repeats, a large unique sequence inverted repeat, a large unique sequence forward repeat, and all of the intervening material. A comparison of this 23.7-kb region with the corresponding human C mu/C delta region reveals clear homology in the coding and introns of C mu but not in the 5' flanking J gene segments nor in the intergenic and C delta regions. This mixed pattern of similarity between the human and the mouse sequences contrasts with high levels of similarity found in the T-cell receptor C alpha/C delta region and alpha and beta myosin genes and the very low levels found in the gamma-crystallin, XRCC1, and beta-globin gene clusters. The human and mouse comparison further suggests the incorporation of novel sequences into expressed genes of IgD.

Functional analysis of the V gamma 3 promoter of the murine gamma delta T-cell receptor.

The initial day 14 wave of fetal thymocytes express a gamma delta T-cell receptor (TCR). This surface TCR is generated by preferential rearrangement of V gamma 3 and V delta 1 recombination segments. To delineate the role of regulatory sequences in this expression, we have analyzed the V gamma 3 promoter control region under the regulation of its cognate C gamma 1 enhancer. Transcription initiates 25 bases downstream from a TATTAA sequence at a consensus initiator motif. The minimal 5' promoter sequences supporting expression by transient analysis extend -243 nucleotides from the +1 start site. Three regulatory sequences in this region have been defined by deletion and mutagenesis: a consensus CTF/NF-1 site at -55, an Ets homology sequence at -65, and a degenerate, but crucial, SP-1 site at -100. The presence of additional sequences downstream of the start site which extend through the leader intron were necessary for expression. In contrast to other TCR or immunoglobulin variable regions, one or more strong upstream suppressor sequences resembling silencer elements have been observed. A 311-bp fragment, positions -586 to -897, exhibited strong repressing activity regardless of orientation when placed upstream of heterologous promoters.

Primary structure of the porin protein of Haemophilus influenzae type b determined by nucleotide sequence analysis.

Sequencing techniques for single- and double-stranded DNA were used to determine the nucleotide sequence of the gene encoding P2, the major outer membrane (porin) protein of Haemophilus influenzae type b (Hib). The open reading frame encoding the P2 protein comprised 361 amino acid codons. Comparison of the inferred amino acid sequence with data obtained by amino acid sequencing of the N terminus of the mature or fully processed P2 protein revealed that this protein has a signal peptide composed of 20 amino acids. N-terminal amino acid sequencing of tryptic peptides derived from purified P2 allowed direct identification of 158 of the 341 amino acids in the fully processed P2 protein; there was 100% correlation between these amino acid sequences and that inferred from the nucleotide sequence. The amino acid sequence of Hib P2 protein had 23 to 25% homology with the sequence of the OmpF porin of Escherichia coli and with that of the Neisseria gonorrhoeae porin P.IA. Codon usage in the Hib P2 gene was significantly different from that observed for a gene encoding a porin of E. coli. DNA hybridization studies indicated that there is a single copy of the P2 gene in the Hib chromosome. The availability of the nucleotide and amino acid sequences for the Hib P2 protein will facilitate investigation of the antigenic characteristics and structure-function relationship of this porin.

Cloning of the gene encoding the major outer membrane protein of Haemophilus influenzae type b.

The major outer membrane protein (P2) of Haemophilus influenzae type b (Hib) with an apparent molecular weight of 37,000 to 40,000 has been previously shown to function as a porin and also as a target for antibodies protective against experimental Hib disease. The gene encoding the Hib P2 protein was cloned by using a shuttle vector capable of replication in both Escherichia coli and H. influenzae. The amino acid sequence of the amino terminus of the Hib P2 protein was determined and used to design an oligonucleotide probe corresponding to the first 20 amino acids of this protein. This oligonucleotide probe was used to identify Hib chromosomal DNA fragments containing the Hib P2 gene. These DNA fragments were ligated into the plasmid vector pGJB103 and then used to transform a rec-1 mutant of H. influenzae Rd. Recombinant clones expressing the Hib P2 protein were identified in a colony blot-radioimmunoassay by using a monoclonal antibody specific for a surface epitope of the Hib P2 protein. The gene encoding this Hib protein was present on a 10-kilobase Hib DNA insert in the recombinant plasmid. Transformation experiments involving the recombinant plasmid suggested that unregulated synthesis of Hib P2 is a lethal event in E. coli. The recombinant Hib P2 protein was exposed on the surface of the recombinant H. influenzae strain. This recombinant strain was used to develop a system for detecting polyclonal serum antibodies directed against surface determinants of the Hib P2 protein. The availability of the gene encoding the Hib P2 protein should facilitate investigation of both the immunogenicity and the structure-function relationship(s) of this major outer membrane protein.

Identification and localization of integral membrane proteins of virulent Treponema pallidum subsp. pallidum by phase partitioning with the nonionic detergent triton X-114.

Integral membrane proteins of Treponema pallidum subsp. pallidum (T. pallidum) were identified by phase partitioning with the nonionic detergent Triton X-114; antigens with apparent molecular masses of 47, 38, 36, 34, 32, 17, and 15 kilodaltons (kDa) were identified in the detergent phase. Immunoblotting with murine monoclonal antibodies directed against pathogen-specific 47- and 34-kDa T. pallidum antigens confirmed their presence in the detergent phase. Endoflagellar proteins of T. pallidum were not detected in immunoblots of detergent-phase proteins when monospecific antisera directed against endoflagella of the nonpathogenic T. phagedenis biotype Reiter were used. At detergent concentrations (0.02 and 0.1%) which appeared to solubilize selectively the outer membranes of treponemes radiolabeled with 35S in vitro, limited amounts of detergent-phase proteins were immunoprecipitated. Greater amounts of detergent-phase proteins were extracted at higher detergent concentrations (0.5 and 2.0%) which resulted in both outer membrane solubilization and ultrastructural derangements of the residual cytoplasmic bodies. Furthermore, Triton X-114 extraction of both intact treponemes and organisms without outer membranes yielded detergent phases with similar protein profiles. The results of these experiments indicate that the hydrophobic proteins identified by Triton X-114 are not located exclusively in the T. pallidum outer membrane. The results are also consistent with the hypothesis that the T. pallidum outer membrane is a protein-deficient lipid bilayer.

A monoclonal antibody that blocks adhesion of Chlamydomonas mt+ gametes.

During the mating reaction (fertilization) in the biflagellated alga, Chlamydomonas reinhardtii, mt+ and mt- gametes adhere to each other via their flagella and subsequently fuse to form quadriflagellated zygotes. In the studies reported here, we describe a monoclonal antibody directed against an mt+ flagellar surface molecule. The antibody blocks the adhesiveness of mt+ gametes, isolated mt+ flagella, and detergent extracts thereof. It has no effect on mt- gametes. Cyanogen bromide-activated Sepharose beads derivatized with the antibody bind only mt+ gametes; mt- gametes and mt+ and mt- vegetative cells are unreactive with the derivatized beads. The interaction of mt+ gametes with the beads is dynamic and cells continuously bind, detach, and rebind to the beads. Surprisingly, antibody-derivatized beads that have been incubated with mt+ gametes acquire the ability to bind mt- gametes. Moreover, extraction of the preincubated beads with detergents releases active mt+ adhesion molecules. The evidence suggests that binding of the antibody to the flagellar surface adhesion molecules causes their release from the flagellar surface, possibly mimicking the normal mechanism of flagellar de-adhesion.

Flagellar adhesion in Chlamydomonas induces synthesis of two high molecular weight cell surface proteins.

Because our previous studies (Snell, W.J., and W.S. Moore, 1980, J. Cell Biol. 84:203- 210) on the mating reaction of chlamydomonas reinhardtii showed that there was an adhesion-induced turnover of proteins whose synthesis is induced during aggregation. Analysis by SDS PAGE and autoradiography showed that proteins of 220,000 M(r) and 165, 000 M(r) (designated A(1) and A(2) respectively) consistently showed a high rate of synthesis only in flagella or flagellar membrane-enriched fractions prepared from aggregating gametes. Since the two proteins were soluble in the non-ionic detergent NP-40 and were removed from intact cells by a brief pronase treatment, it is likely that A(1) and A(2) are membrane proteins expose on the cell surface. A(1) and A(2) were each synthesized by gametes of both mating types (mt(-) and mt(+)) and synthesis of these two proteins could be detected in the normal mating reaction (wild type mt(-) and mt(+)), in mixtures of mt(-) and impotent mt(+) gametes (which could aggregate but not fuse), and in mixtures of gametes of a single mating type with isolated flagella of the opposite mating type. Cells aggregating in tunicamycin, an inhibitor of protein glycosylation, lost their adhesiveness during aggregation and did not synthesize the 220,000 M(r) protein but instead produced a protein (possibly an underglycosylated form of A(1)) of slightly lower mol wt. The 220,000 and 165,000 M(R) proteins appeared to be flagellar proteins and not cell wall proteins because A(1) and A(2) did not co-migrate with previously identified cell wall proteins, and synthesis of the two proteins could not be detected in flagella-less (bald-2) mutant cells. Analysis of the adhesive activity of sucrose gradient fraction of detergent (octyl glucoside)-solubilized flagellar membranes revealed that fractions containing A(1) and A(2) did not have detectable adhesive activity. The possibility remains that A(1) and A(2) are adhesion molecules whose activity could not be measured in the assay we used. Alternatively, the 220,000 and 165,000 M(r) proteins may be inactivated adhesion molecules or else they may be flagellar surface proteins involved only indirectly in the adhesion process.

Lidocaine reversibly inhibits fertilization in Chlamydomonas: a possible role for calcium in sexual signalling.

A flagellar adhesion-induced signal sent during the mating reaction of the biflagellate alga, Chlamydomonas reinhardtii, initiates release of cell-wall-degrading enzymes, activation of mating structures, and cell fusion. The nature of this signal is unknown, but it may be mediated by an adhesion-induced change (activation) of flagellar tips. The studies reported here show that lidocaine, a local anesthetic that is reported to interfere with the movement of divalent cations across cell membranes, reversibly blocks cell wall loss and gametic fusion without blocking adhesion or flagellar tip activation. In these experiments lidocaine inhibited both the initial rates and the extent of wall loss and zygote formation. Studies with gametes of a paralyzed flagellar mutant, pf 17, revealed that lidocaine also blocked flagellar surface motility (visualized as movement of polystyrene beads) at concentrations of the inhibitor which also prevented gametic fusion. The concentration of lidocaine required to block cell fusion was dependent on the concentration of calcium or magnesium in the medium. In the absence of added calcium, 0.5 mM lidocaine inhibited fusion by 70%. In 0.5 mM calcium, 0.5 mM lidocaine had no effect on fusion and 2 mM lidocaine was required for 90% inhibition. The results suggest that divalent cations may play a critical role in sexual signalling in Chlamydomonas.