In vitro biotinylation provides quantitative recovery of highly purified active lactose permease in a single step.

Consler et al. [Consler, T. G., Persson, B. L., et al. (1993) Proc. Natl. Acad. Sci. U.S.A. 90, 6934-6938] described a one-step purification of lactose permease, a hydrophobic membrane transport protein, from Escherichia coli. Permease constructs containing a biotin acceptor domain are biotinylated in vivo, followed by solubilization and avidin affinity purification. Although a high degree of purity is obtained, only about 15-20% of the permease is recovered due to incomplete biotinylation. In this communication, a simple modification is described that allows quantitative recovery of highly purified permease. Membranes containing permease with the biotin acceptor domain from the Klebsiella pneumoniae oxaloacetate decarboxylase are extracted with 5 M urea or treated with dicyclohexylcarbodiimide to inactivate F1/Fo ATPase and biotinylated in vitro with biotin ligase, ATP and d-biotin. Subsequently, the membranes are harvested, washed to remove free biotin and solubilized with 2% n-dodecyl-beta-D-maltopyranoside. Biotinylated permease is then purified in one step by affinity chromatography on monomeric avidin-Sepharose. The purified material is homogeneous and exhibits full activity with respect to ligand binding and counterflow.

Synthetic peptides corresponding to the four P regions of Electrophorus electricus Na+ channel: interaction with and organization in model phospholipid membranes.

The hydropathy plot of the alpha subunit of the voltage-gated Na+ channel reveals four homologous repeats, each of which is homologous to Shaker type K+ channel monomer and contains six putative transmembrane segments and a hydrophobic segment within the loop connecting transmembrane segments S5 and S6. Current models predict that the four homologous segments [designated H5 or P regions (PR)] from the S5-S6 loop of each repeat lie in the aqueous pore. Peptides corresponding to the P regions of the four domains of the Electrophorus electricus (eel) Na+ channel (25-36 aa long, designated as PR-I, PR-II, PR-III, and PR-IV) and a 23-mer preceding PR-II (designated pre-PR-II) were synthesized and fluorescently labeled. The segments were then structurally and functionally characterized for their interaction with phospholipid membranes. Although the sequences of the four P regions are significantly different, they all bind to zwitterionic phospholipid membranes with similar partition coefficients (approximately 10(4) M-1). The pre-PR-II does not bind membranes at all. Resonance energy transfer measurements, between donor/acceptor-labeled pairs of peptides, revealed that besides the PR-I/PR-III pair, all other pairs form heteroaggregates but do not coassemble with unrelated membrane-bound peptide. Circular dichroism (CD) spectroscopy revealed that PR-I, PR-II, and PR-III adopt similar partial alpha-helical structures (approximately 30%) in 40% trifluoroethanol and in solutions of 1% sodium dodecylsulfate (SDS). The PR-IV (36 aa) adopts approximately 18% alpha-helical structure, and pre-PR-II gives a low CD signal. These findings are in line with proposed models in which the P regions are packed in close proximity in the lumen of the hydrophobic core of the channel. Furthermore, the finding that the PRs adopt similar partial alpha-helical structures in two different hydrophobic environments might suggest that partial alpha-helical structures also exist in the native channel as proposed by recent models. The results are discussed in terms of proposals that various regions of membrane proteins participate in driving folding or oligomerization of the parent molecules.

pH- and ionic strength-dependent fusion of phospholipid vesicles induced by pardaxin analogues or by mixtures of charge-reversed peptides.

The fusogenic properties of the neurotoxin paradaxin and eight of its analogues with small unilamellar vesicles (SUV), composed of egg phosphatidylcholine and phosphatidylserine (PC/PS), were investigated. Fusion was demonstrated by a lipid-mixing assay and by an increase in vesicle size as revealed by electron microscopy. The lipid-mixing assay was performed at either neutral (pH 6.8) or acidic (pH 4.5) conditions, in solutions containing either high or low salt concentrations. A low level of fusion could be induced at neutral pH only by pardaxin derivatives with amino groups at both the peptide's backbone and N-terminus. However, a marked enhancement in the fusogenic activity occurred when amino groups were present also in the C-terminus. Pardaxin analogues in which amino groups were substituted by carboxylic groups induced elevated levels of fusion only at high salt concentrations where enhancement of aggregation occurs, and acidic pH, which increased alpha-helicity. The influence of mutual interactions between pardaxin's analogues possessing complementary charges on the lipid-mixing process was also studied. At neutral pH and high salt, an inactive acidic analogue increased the fusogenic activity of a complementary-charged basic peptide. However, such mutual interactions at low salt concentrations reduced the fusogenic activity of the pardaxin analogues. Analogues containing D-amino acids were not fusogenic, thus demonstrating the structural specificity of these observations. The results indicate that the charge, alpha-helical structure, and aggregation of peptide monomers play an important role in the fusogenic ability of polypeptides.

Interaction of antimicrobial dermaseptin and its fluorescently labeled analogues with phospholipid membranes.

Dermaseptin, a 34 amino-acid residue antimicrobial polypeptide [Mor, A., Nguyen, V. H., Delfour, A., Migliore-Samour, D., & Nicolas, P. (1991) Biochemistry 30, 8824-8830] was synthesized and selectively labeled at its N-terminal amino acid with either 7-nitrobenz-2-oxa-1,3-diazole-4-yl (NBD), rhodamine, or fluorescein. The fluorescent emission spectra of the NBD-labeled dermaseptin displayed a blue-shift upon binding to small unilamellar vesicles (SUV), reflecting the relocation of the fluorescent probe to an environment of increased apolarity. Titrations of solutions containing NBD-labeled dermaseptin with SUV composed of zwitterionic or acidic phospholipids were used to generate binding isotherms, from which were derived surface partition constants of (0.66 +/- 0.06) x 10(4) M-1 and (2.8 +/- 0.3) x 10(4) M-1, respectively. The shape of the binding isotherms, as well as fluorescence energy transfer measurements, suggests that some aggregation of membrane-bound peptide monomers occurs in acidic but not in zwitterionic vesicles. The preferential susceptibility of the peptide to proteolysis when bound to zwitterionic but not to acidic SUV suggests that these aggregates might then penetrate a relatively short distance into the hydrophobic region of the acidic membrane. Furthermore, the results provide good correlation between the peptide's strong binding and its ability to permeate membranes composed of acidic phospholipids, as revealed by a dissipation of diffusion potential and a release of entrapped calcein from SUV.

Interaction of D-amino acid incorporated analogues of pardaxin with membranes.

The influence of specific L- to D-amino acid substitutions on the interaction of pardaxin, a shark repellent neurotoxin polypeptide, with phospholipid vesicles and human erythrocytes is described. Twelve modified, truncated, or fluorescently labeled [with the fluorophore 7-nitrobenz-2-oxa-1,3-diazole-4-yl (NBD) at their N-terminal amino acid] analogues of pardaxin were synthesized by a solid-phase method. Fluorescence measurements were used to monitor the interaction of the analogues with membranes [Rapaport, D., & Shai, Y. (1991) J. Biol. Chem. 266, 23769-23775]. Upon titration of solutions containing the NBD-labeled peptides with small unilamellar vesicles, the fluorescent emission spectra of all NBD-labeled peptides displayed similar blue-shifts, in addition to enhanced intensities, upon relocation of the probe to the more apolar environment. Binding isotherms were constructed from which surface partition constants, in the range of 10(4) M-1, were derived. The existence of an aggregation process, suggested by the shape of the binding isotherms, could be associated only with those analogues in which the N-helix (residues 1-9) was not perturbed. The alpha-helical content of the analogues was estimated by circular dichroism (CD) spectroscopy, both before and after binding to vesicles at neutral pH. The ability of the peptides to dissipate a diffusion potential and to cause calcein release, as well as to lyse human erythrocytes, served to functionally characterize the peptides. The results support a two alpha-helix model, with a bend at position 13, as best describing pardaxin in its membrane-bound state.(ABSTRACT TRUNCATED AT 250 WORDS)