Conformation and lipid binding properties of four peptides derived from the membrane-binding domain of CTP:phosphocholine cytidylyltransferase

We are probing the mechanism of the lipid selective membrane interactions of CTP:phosphocholine cytidylyltransferase (CT). We have proposed that the membrane binding domain of CT (domain M) consists of a continuous amphipathic alpha-helix between residues approximately 240-295 [Dunne, S. J., et al. (1996) Biochemistry 35, 11975-11984]. This study examined the secondary structure and membrane binding properties of synthetic peptides derived from domain M: a 62mer peptide encompassing the entire domain (Pep62), a 33mer corresponding to the N-terminal portion (PepNH1), and two 33mers corresponding to the three C-terminal 11mer repeats, one with the wild-type sequence (Pep33Ser), and one with the three serines in the nonpolar face substituted with alanine (Pep33Ala). Peptide secondary structure was analyzed by circular dichroism, and lipid interactions were analyzed by a direct vesicle binding assay, by effects of lipid vesicles on peptide tryptophan fluorescence, and by monolayer surface pressure changes. All peptides bound to vesicles as alpha-helices with selectivity for anionic lipids. Binding involved intercalation of the peptide tryptophan into the hydrophobic membrane core. PepNH1, the peptide with the highest positive charge density, showed strong selectivity for anionic lipids. PepNH1 and Pep33Ser did not bind to PC vesicles; however, the more hydrophobic peptides, Pep33Ala and Pep62, did bind to PC vesicles, with apparent partition coefficients for PC that were only approximately 1 order of magnitude lower than those for PC/PG (1/1). Our results suggest that the polar serines interrupting the nonpolar face of the amphipathic helix serve to lower the lipid affinity and thereby enhance selectivity for anionic lipids. Although diacylglycerol is an activator of the enzyme, none of the peptides responded differentially to PC/diacylglycerol vesicles versus pure PC vesicles, suggesting that domain M alone is not sufficient for the enzyme's response to diacylglycerol. Increases in surface pressure at an air-water interface indicated that the domain M peptides had strong surface-seeking tendencies. This supports a binding orientation for domain M parallel to the membrane surface. Binding of CT peptides to spread lipid monolayers caused surface pressure reductions, suggesting condensation of lipids in the formation of lipid-peptide complexes. At low monolayer surface pressures, Pep62 interacted equally with anionic and zwitterionic phospholipids. This suggests that one determinant of the selectivity for anionic lipids is the lipid packing density (area per molecule).     

The amphipathic helix concept: length effects on ideally amphipathic LiKj(i=2j) peptides to acquire optimal hemolytic activity

In a minimalist approach to modeling lytic toxins, amphipathic peptides of LiKj with i=2j composition and whose length varies from 5 to 22 residues were studied for their ability to induce hemolysis and lipid vesicle leakage. Their sequences were designed to generate ideally amphipathic alpha helices with a single K residue per putative turn. All the peptides were lytic, their activities varying by more than a factor of 103 from the shortest 5-residue-long peptide (5-mer) to the longest 22-mer. However, there was no monotonous increase versus length. The 15-mer was as active as the 22-mer and even more than melittin which is used as standard. Partition coefficients from the buffer to the membrane increased in relation to length up to 12 residues, then weakly decreased to reach a plateau, while they were expected to increase monotonously with peptide length and hydrophobicity as revealed from HPLC retention times. Fluorescence labeling by a dansyl group at the N-terminus, or by a W near the CO-terminus, show that up to 12 residues, the peptides were essentially monomeric while longer peptides strongly aggregated in the solution. Lipid affinity was then controlled by peptide length and was found to be limited by folding and self-association in buffer. The lytic activity resulted both from lipid affinity, which varied by a factor of 20-fold, and from efficiency in disturbing the membrane when bound, the latter steeply and monotonously increasing with length. The 15-residue-long peptide, KLLKLLLKLLLKLLK, had the optimal size for highest lytic activity. The shallow location of the fluorescent labels in the lipids is further evidence for a model of peptides remaining flat at the interface.     

N-terminal hydrophobic residues of the G-protein ADP-ribosylation factor-1 insert into membrane phospholipids upon GDP to GTP exchange

GDP/GTP exchange modulates the interaction of the small G-protein ADP-ribosylation factor-1 with membrane lipids: if ARF(GDP) is mostly soluble, ARF(GTP) binds tightly to lipid vesicles. Previous studies have shown that this GTP-dependent binding persists upon removal of the N-terminal myristate but is abolished following further deletion of the 17 N-terminal residues. This suggests a role for this amphipathic peptide in lipid membrane binding. In the ARF(GDP) crystal structure, the 2-13 peptide is helical, with its hydrophobic residues buried in the protein core. When ARF switches to the GTP state, these residues may insert into membrane lipids. We have studied the binding of ARF to model unilamellar vesicles of defined composition. ARF(GDP) binds weakly to vesicles through hydrophobic interaction of the myristate and electrostatic interaction of cationic residues with anionic lipids. Phosphatidylinositol 4,5-bis(phosphate) shows no specific effects other than strictly electrostatic. By using fluorescence energy transfer, the strength of the ARF(GTP)-lipid interaction is assessed via the dissociation rate of ARF(GTPgammaS) from labeled lipid vesicles. ARF(GTPgammaS) dissociates slowly (tau(off) approximately 75 s) from neutral PC vesicles. Including 30% anionic phospholipids increases tau(off) by only 3-fold. Reducing the N-terminal peptide hydrophobicity by point mutations had larger effects: F9A and L8A-F9A substitutions accelerate the dissociation of ARF(GTPgammaS) from vesicles by factors of 7 and 100, respectively. This strongly suggests that, upon GDP/GTP exchange, the N-terminal helix is released from the protein core so its hydrophobic residues can interact with membrane phospholipids.     

Effect of the N-terminal glycine on the secondary structure, orientation, and interaction of the influenza hemagglutinin fusion peptide with lipid bilayers

The amino-terminal segment of the membrane-anchored subunit of influenza hemagglutinin (HA) plays a crucial role in membrane fusion and, hence, has been termed the fusion peptide. We have studied the secondary structure, orientation, and effects on the bilayer structure of synthetic peptides corresponding to the wild-type and several fusogenic and nonfusogenic mutants with altered N-termini of the influenza HA fusion peptide by fluorescence, circular dichroism, and Fourier transform infrared spectroscopy. All peptides contained segments of alpha-helical and beta-strand conformation. In the wild-type fusion peptide, 40% of all residues were in alpha-secondary and 30% in beta-secondary structures. By comparison, the nonfusogenic peptides exhibited larger beta/alpha secondary structure ratios. The order parameters of the helices and the amide carbonyl groups of the beta-strands of the wild-type fusion peptide were measured separately, based on the infrared dichroism of the respective absorption bands. Order parameters in the range 0.1-0.7 were found for both segments of the wild-type peptide, which indicates that they are most likely aligned at oblique angles to the membrane normal. The nonfusogenic but not the fusogenic peptides induced splitting of the infrared absorption band at 1735 cm(-1), which is assigned to stretching vibrations of the lipid ester carbonyl bond. This splitting, which reports on an alteration of the hydrogen bonds formed between the lipid ester carbonyls and water and/or hydrogen-donating groups of the fusion peptides, correlated with the beta/alpha ratio of the peptides, suggesting that unpaired beta-strands may replace water molecules and hydrogen-bond to the lipid ester carbonyl groups. The profound structural changes induced by single amino acid replacements at the extreme N-terminus of the fusion peptide further suggest that tertiary or quaternary structural interactions may be important when fusion peptides bind to lipid bilayers.     

Interaction of the lantibiotic nisin with membranes revealed by fluorescence quenching of an introduced tryptophan

Nisin is a lantibiotic produced by strains of Lactococcus lactis subsp. lactis. The target for nisin action is the cytoplasmic membrane of gram-positive bacteria. To aid understanding of its mode of action, the interaction of nisin with vesicles of differing phospholipid composition were investigated by fluorescence techniques, using a variant of nisin in which the isoleucine at position 30 was replaced by a tryptophan residue. Activity of the site-directed variant containing tryptophan was established to be similar to that of the wild-type peptide. Fluorescence experiments showed a blue shift of the emission wavelength maximum in the presence of lipid vesicles, indicating that the tryptophan residue enters a more hydrophobic environment. Quenching experiments with aqueous and membrane-restricted quenchers (iodide and spin-labelled lipids, respectively) both confirmed a non-aqueous environment for the Trp30 residue, and implied that the residue resides between 0.36 nm and 0.52 nm from the centre of the membrane, depending on the lipid identity. The results clearly demonstrate that nisin interacts strongly with the hydrophobic phase of lipid vesicles. This interaction is stronger in the presence of negatively charged lipids suggesting their importance in the functional interaction of nisin with membranes.     

A single-residue deletion alters the lipid selectivity of a K+ channel-associated peptide in the beta-conformation: spin label electron spin resonance studies

Lipid-peptide interactions with the 27-residue peptide of sequence KLEALYILMVLGFFGFFTLGIMLSYIR reconstituted as beta-sheet assemblies in dimyristoylphosphatidylcholine bilayers have been studied by electron spin resonance (ESR) spectroscopy with spin-labeled lipids. The peptide corresponds to residues 42-68 of the IsK voltage-gated K+ channel protein and contains the single putative transmembrane span of this protein. Lipid-peptide interactions give rise to a second component in the ESR spectra of lipids spin-labeled on the 14C atom of the chain that corresponds to restriction of the lipid mobility by direct interaction with the peptide assemblies. From the dependence on the lipid/peptide ratio, the stoichiometry of lipid interaction is found to be about two phospholipids/peptide monomer. The sequence of selectivity for lipid association with the peptide assemblies is in the order phosphatidic acid > stearic acid = phosphatidylserine > phosphatidylglycerol = phosphatidylcholine. Comparison with previous data for a corresponding 26-residue mutant peptide with a single deletion of the apolar residue Leu2 (Horvath et al., 1995. Biochemistry 34:3893-3898), indicates a very similar mode of membrane incorporation for native and mutant peptides, but a strongly modified pattern and degree of specificity for the interaction with negatively charged lipids. The latter is interpreted in terms of the relative orientations of the charged amino acid side chains in the beta-sheet assemblies of the native and deletion-mutant peptides.     

Pore-forming peptides induce rapid phospholipid flip-flop in membranes

A kinetic model for pore-mediated and perturbation-mediated flip-flop is presented and used to characterize the mechanism of peptide-induced phospholipid flip-flop in bilayers. The model assumes that certain peptides can bind to and aggregate within the membrane. When the aggregate attains a critical size (M peptides), a channel is created that results in a fast flip-flop of phospholipids. In addition, certain peptides induce flip-flop through perturbation of the membrane without forming a pore. Donor phospholipid vesicles with an asymmetrical distribution of the fluorescent phospholipid 1-oleoyl-2-[12-[(7-nitro-1,2,3-benzoxadiazol-4- yl)amino]dodecanoyl]phosphatidylcholine (NBD-PC) were used to measure the extent of flip-flop by quantitating the decrease in fluorescence as the NBD-PC exchanged from the donor vesicles to acceptor vesicles that contained a quencher of the NBD fluorescence. Flip-flop curves generated at lipid/peptide ratios ranging from 30/1 to 300000/1 could be well-simulated by the model. Pore-forming peptides, such as melittin or the synthetic peptide GALA (WEAALAEALAEALAEHLAEALAEALEALAA), induce rapid phospholipid flip-flop with half-times for flip-flop of seconds at low peptide/vesicle ratios. The deduced pore sizes are M = 10 +/- 2 for GALA and M = 2 - 4 for melittin. The synthetic peptide LAGA (WEAALAEAEALALAEHEALALAEAELALAA) can catalyze flip-flop via bilayer perturbation. In contrast, hydrophobic peptides such as gramicidin A and valinomycin intercalate into the membrane, but induce little flip-flop. Modeling of the kinetics of phospholipid translocation supports pore formation as the key factor in accelerating phospholipid flip-flop. Thus, amphipathic segments from membrane proteins may account for non-energy-dependent phospholipid flip-flop in biological membranes.     

Structure effects of double D-amino acid replacements: a nuclear magnetic resonance and circular dichroism study using amphipathic model helices

D-Amino acid replacements and the determination of resulting structural changes are a useful tool to recognize amphipathic helices in biologically active peptides such as neuropeptide Y and corticotropin-releasing factor. In this paper the secondary structures of one amphipathic alpha-helical peptide and its double D-amino acid analog have been determined by means of 1H NMR and CD spectroscopies under equivalent conditions. The chemical shifts (NH and C alpha H) and the analysis of nuclear Overhauser effects show a split of the continuous helix for the all-L peptide into two helices at the position of double D-amino acid replacement. Hydrogen exchange rates correlate with water accessibilities in the hydrophobic/hydrophilic face and confirm the amphipathic helical structure in the all-L peptide as well as in its double D-amino acid analog. A significantly accelerated hydrogen isotope exchange rate is observed for the D-Ala9 backbone proton, implying an increased flexibility at that position. These results show that the incorporation of an adjacent pair of D-amino acids only causes a local change in structure and flexibility, which makes the double D replacement interesting as a tool for specific helix-disturbing modifications to search for helical conformations in biologically active peptides.     

Identification of two distinct properties of class II major histocompatibility complex-associated peptides

We have examined the interactions of various peptides with the mouse class II major histocompatibility complex molecule I-Ak. The peptides were derived from the model protein hen egg white lysozyme (HEL). The immunodominant peptide of HEL is a 10-mer, residues 52-61. Our previous work established that this sequence contains the key residues for binding and presentation to T cells. Now we show that the binding of this 10-mer sequence resulted in complexes of I-Ak and peptide that, in SDS/PAGE (without boiling the protein), rapidly dissociated from the component alpha and beta chains. The binding interactions were studied in vitro, by incubating purified I-Ak and radiolabeled peptide, or ex vivo, by using antigen-presenting cells incubated with peptides. Peptides with additional residues at either the amino or carboxyl terminus behaved dramatically differently. Complexes of I-Ak with the longer peptides were stable to SDS/PAGE. Very few amino acid additions result in the change from unstable to stable complexes. The important issue here is that when cultured with HEL, antigen-presenting cells selected the HEL peptides containing the 52-61 sequences that favored stability [Nelson, C. A., Roof, R. W., McCourt, D. W. & Unanue, E. R. (1992) Proc. Natl., Acad. Sci. USA 89, 7380-7383]. Also, from other studies, such sequences correlate with a high immunogenicity of the peptide. We conclude that there are structural features of peptides that change the stability of the class II molecule and that are independent of the "core" peptide seen by the T cells.     

The helix-hinge-helix structural motif in human apolipoprotein A-I determined by NMR spectroscopy

The conformation of a synthetic peptide of 46 residues from apoA-I was investigated by fluorescence, CD, and 2D NMR spectroscopies in lipid-mimetic environments. ApoA-I(142-187) is mainly unstructured in water but helical in SDS or dodecylphosphocholine (DPC), although the peptide only associates with DPC at approximately the critical micellar concentration. Solution structures of apoA-I(142-187) were determined by distance geometry calculations based on 450 (in DPC-d38) or 397 (in SDS-d25) NOE-derived distance restraints, respectively. Backbone RMSDs for superimposing the two helical regions 146-162 and 168-182 are 0.98 +/- 0.22 (2.38 +/- 0.20) and 1.99 +/- 0.42 (2.02 +/- 0.21) A in DPC (SDS), respectively. No interhelical NOE was found, suggesting that helix-helix interactions between the two helical domains in apoA-I(142-187) are unlikely. Similar average, curved helix-hinge-helix structures were found in both SDS and DPC micelles with the hydrophobic residues occupying the concave face, indicating that hydrophobic interactions dominate. Intermolecular NOESY experiments, performed in the presence of 50% protonated SDS, confirm that the two amphipathic helices and Y166 in the hinge all interact with the micelle. The involvement of Y166 in lipid binding is supported by fluorescence spectroscopy as well. On the basis of all the data above, we propose a model for the peptide-lipid complexes wherein the curved amphipathic helix-hinge-helix structural motif straddles the micelle. The peptide-aided signal assignment achieved for apoA-I(122-187) (66mer) and apoA-I suggests that such a structural motif is retained in the longer peptide and most likely in the intact protein.     

Native peptide inhibition. Specific inhibition of type II phospholipases A2 by synthetic peptides derived from the primary sequence

The binding of low molecular weight type II phospholipase A2 (EC) to membrane surfaces and hydrolysis of phospholipid are thought to involve the formation of a hydrophobic channel into which a single substrate molecule diffuses before cleavage. The floor and right side of the channel are provided by hydrophobic residues 2, 5, and 9 of an amphipathic amino-terminal helix. The channel is postulated to form via a conformational change in this helix and inward movement of a hydrophobic flap (residue 69 side chain). We show that the amino-terminal tryptic peptide of human type II phospholipase A2 forms a noncovalent complex with the tryptic peptide from residues 70-74 of the enzyme. Further, the 70-74-peptide sequence (FLSYK) dose-dependently inhibits phospholipid hydrolysis in a mixed micelle assay. This native peptide inhibition also occurred with type II enzymes from Crotalus durissus and Crotalus atrox, which have different amino acid sequences at the amino terminus as well as different 70-74 regions of the molecules. Despite significant conservation of tertiary structure among the enzymes, inhibition by each peptide is specific to the enzyme from which the peptide sequence is derived. We propose that these native peptides inhibit enzyme activity via a sequence-specific, noncovalent interaction with the amino-terminal residues of the enzyme, thereby preventing the conformational change on binding to the micelle interface. These experiments demonstrate a new method for specific inhibition of phospholipase A2 which, in principle, would be applicable to other biologically active polypeptides and proteins.     

Structure, synthesis, and activity of dermaseptin b, a novel vertebrate defensive peptide from frog skin: relationship with adenoregulin

A novel antimicrobial peptide, designated dermaseptin b, was isolated from the skin of the arboreal frog Phyllomedusa bicolor. This 27-residue peptide amide is basic, containing 3 lysine residues that punctuate an alternating hydrophobic and hydrophilic sequence. In helix-inducing solvent, dermaseptin b adopts an amphipathic alpha-helical conformation that most closely resembles class L amphipathic helixes, with all lysine residues on the polar face of the helix. The peptide exhibits growth inhibition activity in vitro against a broad spectrum of pathogenic microorganisms including yeast and bacteria as well as various filamentous fungi that are responsible for severe opportunistic infections accompanying acquired immunodeficiency syndrome and the use of immunosuppressive agents. Maximized pairwise sequence alignment of dermaseptin b and dermaseptin s, a 34-residue antimicrobial peptide previously isolated from Phyllomedusa sauvagii, reveals 81% amino acid identity. No other significant similarity was found between dermaseptin b and any prokaryotic or eukaryotic protein, but similarity was found with adenoregulin (38% amino acid postional identity), a 33-residue peptide that enhances binding of agonists to the A1 adenosine receptor. The synthetic replicates of dermaseptin b and adenoregulin displayed similar but nonidentical spectra of antimicrobial activity, and both peptides were devoid of lytic effect on mammalian cells. Accordingly, the observation that adenoregulin enhances binding of agonists to the adenosine receptor may in fact be a consequence of its ability to alter the structure of biological membranes and to produce signal transduction via interactions with the lipid bilayer, bypassing cell surface receptor interactions.     

Relationship of sidechain hydrophobicity and alpha-helical propensity on the stability of the single-stranded amphipathic alpha-helix

The aim of the present investigation is to determine the effect of alpha-helical propensity and sidechain hydrophobicity on the stability of amphipathic alpha-helices. Accordingly, a series of 18-residue amphipathic alpha-helical peptides has been synthesized as a model system where all 20 amino acid residues were substituted on the hydrophobic face of the amphipathic alpha-helix. In these experiments, all three parameters (sidechain hydrophobicity, alpha-helical propensity and helix stability) were measured on the same set of peptide analogues. For these peptide analogues that differ by only one amino acid residue, there was a 0.96 kcal/mole difference in alpha-helical propensity between the most (Ala) and the least (Gly) alpha-helical analogue, a 12.1-minute difference between the most (Phe) and the least (Asp) retentive analogue on the reversed-phase column, and a 32.3 degrees C difference in melting temperatures between the most (Leu) and the least (Asp) stable analogue. The results show that the hydrophobicity and alpha-helical propensity of an amino acid sidechain are not correlated with each other, but each contributes to the stability of the amphipathic alpha-helix. More importantly, the combined effects of alpha-helical propensity and sidechain hydrophobicity at a ratio of about 2:1 had optimal correlation with alpha-helix stability. These results suggest that both alpha-helical propensity and sidechain hydrophobicity should be taken into consideration in the design of alpha-helical proteins with the desired stability.     

A physicochemical approach for predicting the effectiveness of peptide-based gene delivery systems for use in plasmid-based gene therapy

Novel synthetic peptides, based on carrier peptide analogs (YKAKnWK) and an amphipathic peptide (GLFEALLELLESLWELLLEA), have been formulated with DNA plasmids to create peptide-based gene delivery systems. The carrier peptides are used to condense plasmids into nanoparticles with a hydrodynamic diameter (DH) ranging from 40 to 200 nm, which are sterically stable for over 100 h. Size and morphology of the carrier peptide/plasmid complex have been determined by photon correlation spectroscopy (PCS) and transmission electron microscopy (TEM), respectively. The amphipathic peptide is used as a pH-sensitive lytic agent to facilitate release of the plasmid from endosomes after endocytosis of the peptide/plasmid complex. Hemolysis assays have shown that the amphipathic peptide destabilizes lipid bilayers at low pH, mimicking the properties of viral fusogenic peptides. However, circular dichroism studies show that unlike the viral fusion peptides, this amphipathic peptide loses some of its alpha-helical structure at low pH in the presence of liposomes. The peptide-based gene delivery systems were tested for transfection efficiency in a variety of cell lines, including 14-day C2C12 mouse myotubes, using gene expression systems containing the beta-galactosidase reporter gene. Transfection data demonstrate a correlation between in vitro transfection efficiency and the combination of several physical properties of the peptide/plasmid complexes, including 1) DNA dose, 2) the zeta potential of the particle, 3) the requirement of both lytic and carrier peptides, and 4) the number of lysine residues associated with the carrier peptide. Transfection data on 14-day C2C12 myotubes utilizing the therapeutic human growth hormone gene formulated in an optimal peptide gene delivery system show an increase in gene expression over time, with a maximum in protein levels at 96 h (approximately 18 ng/ml).     

Structure, orientation and affinity for interfaces and lipids of ideally amphipathic lytic LiKj(i=2j) peptides

The behavior of lytic ideally amphipathic peptides of generic composition LiKj(i=2j) and named LKn, n=i+j, is investigated in situ by the monolayer technique combined with the recently developed polarization modulation IR spectroscopy (PMIRRAS). A change in the secondary structure occurs versus peptide length. Peptides longer than 12 residues fold into alpha-helices at interfaces as expected from their design, while enough shorter peptides, from 9 down to 5 residues, form intermolecular antiparallel beta-sheets. Analysis of experimental and calculated PMIRRAS spectra in the amide I and II regions show that peptides are flat oriented at the interfaces. Structures and orientation are preserved whatever the nature of the interface, air/water or DMPC monolayer, and the lateral pressure. Peptide partition constants, KaffPi, are estimated from isobar surface increases of DMPC monolayers. They strongly increase when Pi decreases from 30 mN/m to 8 mN/m and they vary with peptide length with an optimum for 12 residues. This non-monotonous dependence fits with data obtained in bilayers and follows the hemolytic activity of the peptides. Lipid perturbations due to peptide insertion essentially detected on the PO4- and CO bands indicate disorder of the lipid head groups. Lysis induced on membranes by such peptides is proposed to first result from their flat asymmetric insertion.     

Mechanism of synergism between antimicrobial peptides magainin 2 and PGLa

The antimicrobial peptides magainin 2 and PGLa, discovered in the skin of the African clawed frog, Xenopus laevis, exhibit marked synergism [Westerhoff, H. V., Zasloff, M., Rosner, J. L., Hendler, R. W., de Waal, A., Vaz Gomes, A., Jongsma, A. P. M., Riethorst, A., and Juretic, D., Eur. J. Biochem. 228, 257-264 (1995)], although the mechanism is not yet clear. They are believed to kill bacteria by permeabilizing membranes. In this study, we examined the interactions of these peptides in lipid bilayers. PGLa, like magainin 2, preferentially interacts with acidic lipids, forming an amphipathic helix. The peptide induces the release of a water-soluble dye, calcein, entrapped within liposomes. The coexistence of magainin 2 enhances membrane permeabilization, which is maximal at a 1:1 molar ratio. Fluorescence experiments using L18W-PGLa revealed that both peptides form a stoichiometric 1:1 complex in the membrane phase with an association free energy of -15 kJ/mol. Single amino acid mutations in magainin 2 significantly altered the synergistic activity, suggesting that precise molecular recognition is involved in complex formation. The complex as well as each component peptide form peptide-lipid supramolecular complex pores, which mediate the mutually coupled transbilayer transport of dye, lipid, and the peptide per se. The rate of pore formation rate is in the order complex >/= PGLa > magainin 2, whereas the pore lifetime is in the order magainin 2 > complex > PGLa. Therefore, the synergism is a consequence of the formation of a potent heterosupramolecular complex, which is characterized by fast pore formation and moderate pore stability.     

Identification of an inhibitory domain of CTP:phosphocholine cytidylyltransferase

The function of the putative amphipathic helices between residues 236 and 314 of CTP:phosphocholine cytidylyltransferase was examined by constructing two truncation mutants; CT314 was missing the entire phosphorylation segment, whereas CT236 was missing both the region with the putative amphipathic helices and the phosphorylation segment. Stable cells lines expressing these truncation mutants in Chinese hamster ovary 58 cells were isolated and characterized. CT314 was predominantly soluble in control cells but became membrane-associated in cells treated with oleate, which also causes translocation of wild-type cytidylyltransferase. CT236 was found to be soluble both in control cells and in cells treated to cause translocation. These results strongly suggest that the membrane-binding site is located within residues 237-314. When assayed for activity in vitro, the mutant forms were catalytically active in the presence of exogenous lipids. CT236, moreover, was as active in the absence of lipids as in their presence, whereas CT314 required lipids for activity. The rate of phosphatidylcholine synthesis in cells expressing CT236 was considerably higher than in wild-type cells, consistent with the enzyme being constitutively active in the cells. These results indicate that residues 237-314 constitute an inhibitory segment; when this segment is removed from the catalytic domain by truncation or by binding to membranes, an inhibitory constraint is removed and cytidylyltransferase is activated.     

Structural basis for specificity of retroviral proteases

The Rous sarcoma virus (RSV) protease S9 variant has been engineered to exhibit high affinity for HIV-1 protease substrates and inhibitors in order to verify the residues deduced to be critical for the specificity differences. The variant has 9 substitutions (S38T, I42D, I44V, M73V, A100L, V104T, R105P, G106V, and S107N) of structurally equivalent residues from HIV-1 protease. Unlike the wild-type enzyme, RSV S9 protease hydrolyzes peptides representing the HIV-1 protease polyprotein cleavage sites. The crystal structure of RSV S9 protease with the inhibitor, Arg-Val-Leu-r-Phe-Glu-Ala-Nle-NH2, a reduced peptide analogue of the HIV-1 CA-p2 cleavage site, has been refined to an R factor of 0.175 at 2.4-A resolution. The structure shows flap residues that were not visible in the previous crystal structure of unliganded wild-type enzyme. Flap residues 64-76 are structurally similar to residues 47-59 of HIV-1 protease. However, residues 61-63 form unique loops at the base of the flaps. Mutational analysis indicates that these loop residues are essential for catalytic activity. Side chains of flap residues His 65 and Gln 63' make hydrogen bond interactions with the inhibitor P3 amide and P4' carbonyl oxygen, respectively. Other interactions of RSV S9 protease with the CA-p2 analogue are very similar to those observed in the crystal structure of HIV-1 protease with the same inhibitor. This is the first crystal structure of an avian retroviral protease in complex with an inhibitor, and it verifies our knowledge of the molecular basis for specificity differences between RSV and HIV-1 proteases.     

Molecular basis for prokaryotic specificity of magainin-induced lysis

Magainins and mastoparans are examples of peptide antibiotics and peptide venoms, respectively. They have been grouped together as class L amphipathic helixes [Segrest, J.P., et al. (1990) Proteins 8, 103-117] because of similarities in the distribution of Lys residues along the polar face of the helix. Class L venoms lyse both eukaryotic and prokaryotic cells whereas class L antibiotics specifically lyse bacteria. The structural basis for the specificity of class L antibiotics is not well understood. Sequence analysis showed that class L antibiotics have a Glu residue on the nonpolar face of the amphipathic helix; this is absent from class L venoms. We synthesized three model class L peptides with or without Glu on the nonpolar face: 18LMG (LGSIWKFIKAFVGGIKKF), [E14]18LMG and [G5,E14]18LMG. Hemolysis, bacteriolysis, and bacteriostasis studies using these peptides showed that the specificity of lysis is due to both the presence of a Glu residue on the nonpolar face of the helix and the bulk of the nonpolar face. Studies using large unilamellar phospholipid vesicles showed that the inclusion of cholesterol greatly inhibited leakage by the two Glu-containing peptides. These results cannot be attributed to changes in the phase behavior of the lipids caused by the inclusion of cholesterol or to differences in the secondary structure of the peptides. These results suggest that eukaryotic cells are resistant to lysis by magainins because of peptide-cholesterol interactions in their membranes that inhibit the formation of peptide structures capable of lysis, perhaps by hydrogen bonding between Glu and cholesterol. Bacterial membranes, lacking cholesterol, are susceptible to lysis by magainins.     

Interaction of class A amphipathic helical peptides with phospholipid unilamellar vesicles

The exchangeable apolipoproteins are important in determining the structure/function properties of lipoproteins. These proteins typically contain varying amounts of amphipathic helices. Five model peptides, 18A, Ac-18A-NH2, Ac-18R-NH2, 37pA, and 37aA, have been designed to investigate variations of the amphipathic alpha-helix structural motif on their lipid-binding properties. These include the 18-residue peptides, 18A and Ac-18A-NH2, examples of class A helices, and Ac-18R-NH2, which has the positions of acidic and basic residues interchanged relative to 18A. Three larger peptides were also studied: 36A, a dimer of 18A, 37pA and 37aA, dimers of 18A coupled by Pro (18A-Pro-18A) and Ala (18A-Ala-18A), respectively. We report here the results of a thermodynamic characterization of the binding properties of these peptides to small unilamellar vesicles of POPC. Partition coefficients, Kp, were determined by fluorescence spectroscopy and binding enthalpies, deltaH, by titration calorimetry. These parameters were used to obtain the free energies, deltaG0, and entropies, deltaS0, of binding. The results of this study indicate Kp values on the order of 10(5), with interactions being enthalpically but not entropically favored in all cases. The presence of positively charged residues at the interface (18A and Ac-18A-NH2) enhances binding but has little effect on the extent of bilayer penetration. The presence of tandem repeats decreases lipid affinities for these small, highly curved bilayers. Our results are consistent with the idea that interaction appears to be confined largely to the surface, with some degree of penetration of the hydrophobic face of the helix into the interior of the bilayer.