Search

The cytotoxic peptide from honeybee venom, melittin, and a synthetic peptide analogue of it lyse human erythrocytes in a biphasic process. The kinetics of the lysis in 0.30 M sucrose, 0.01 M sodium phosphate, pH 7.30 at 4 degrees C were investigated. Our results show that melittin rapidly binds to the outer surface of the erythrocyte membrane, and the surface-bound monomers produce transient openings through which approximately 40 hemoglobin molecules can escape. Concomitantly, the melittin loses its ability to effect the process, presumably by translocation through the bilayer. The half-life for this process is 1.2 min. In a much slower process, dimers of this internalized melittin again produce transient membrane openings in a steady state. On a molar basis, the synthetic peptide analogue produces a fast process comparable to that caused by melittin, but is more efficient in the slow phase. Escape of hemoglobin and of carbonic anhydrase through the openings is diffusion controlled. These results suggest that the functional units necessary for the activity of melittin-like cytotoxic peptides are a 20 amino acid amphiphilic alpha-helix with a hydrophobic:hydrophilic ratio greater than 1 and a short segment with a high concentration of positive charges.

A basic, amphiphilic alpha helix is a structural feature common to a variety of inhibitors of calmodulin and to the calmodulin-binding domains of myosin light chain kinases. To aid in recognizing this structural feature in sequences of peptides and proteins we have developed a computer algorithm which searches for sequences of appropriate length, hydrophobicity, helical hydrophobic moment, and charge to be considered as potential calmodulin-binding sequences. Such sequences occurred infrequently in proteins of known crystal structure. This algorithm was used to find the most likely site in the catalytic (gamma) subunit of phosphorylase b kinase for interaction with calmodulin (the delta subunit). A peptide corresponding to this site (residues 341-361 of the gamma subunit) was synthesized and found to bind calmodulin with approximately an 11 nM dissociation constant. A variant of this peptide in which an aspartic acid at position 7 in its sequence (347 of the gamma subunit) was replaced with an asparagine was found to bind calmodulin with approximately a 3 nM dissociation constant.

The question of how the amino acid sequence of a protein specifies its three-dimensional structure remains to be answered. Proteins are so large and complex that it is difficult to discern the features in their sequences that contribute to their structural stability and function. One approach to this problem is de novo design of model proteins, much simpler than their natural counterparts, yet containing sufficient information in their sequences to specify a given function (for example, folding in aqueous solution, folding in membranes, or formation of ion channels). Designed proteins provide simple model systems for understanding protein structure and function.

Wild-type and pseudorevertant signal peptides of the lamB gene product of Escherichia coli interact with lipid systems whereas a nonfunctional deletion mutant signal peptide does not. This conclusion is based on interaction of synthetic signal peptides with a lipid monolayer-water surface, conformational changes induced by presence of lipid vesicles in an aqueous solution of signal peptide, and capacities of the peptides to promote vesicle aggregation. Analysis of the signal sequences and previous conformational studies suggest that these lipid interaction properties may be attributable to the tendency of the functional signal peptides to adopt alpha-helical conformations. Although the possibility of direct interaction between the signal peptide and membrane lipids during protein secretion is controversial, the results suggest that conformationally related amphiphilicity and consequent membrane affinity of signal sequences are important for function in vivo.

Calmodulin is known to bind target enzymes and basic, amphiphilic peptides in a Ca2(+)-dependent manner. Recently, we introduced a photoaffinity label, p-benzoylphenylalanine (Bpa), into the sequence of a model, alpha-helical, calmodulin-binding peptide. When the Bpa residue was introduced at the third position of the peptide, Met-144 on the C-terminal domain of calmodulin was labeled, whereas when the photolabel was placed at the thirteenth position, Met-71 on the N-terminal domain was labeled. Assuming that both peptides bind in similar orientations, these results are not consistent with the crystal structure of calmodulin, in which the domains are held at a significant distance from one another by a long alpha-helical segment. To test the assumption that both peptides bind in similar orientations, we have synthesized a calmodulin-binding peptide with the photolabel in both the third and the thirteenth positions. Upon photolysis, this peptide forms a cross-link between Met-71 and Met-124 on the N- and C-terminal domains, respectively. Furthermore, a peptide with a Bpa in the thirteenth position and a Trp residue in the third position was also synthesized. After photocross-linking the Bpa residue of this peptide to Met-71 of calmodulin, it could be shown that the fluorescence properties of the Trp residue were consistent with its side chain being buried in a hydrophobic pocket on the C-terminal domain of calmodulin. These data indicate that, when complexed with basic, amphiphilic peptides, calmodulin can adopt a conformation in which its two domains are significantly closer than in the crystal structure of the uncomplexed protein.

Two peptides from the amino terminus of human interferon-beta were synthesized corresponding to amino acids 1-21 and 18-45. The peptides were conjugated to bovine serum albumin, and rabbits were immunized with either the (1-21)- or the (18-45)-peptide conjugate. Antibodies to the synthetic peptides were detected in the sera using a radioimmunoassay with 125I-labeled peptide. Two of the antisera, one against peptide 1-21 and one against peptide 18-45, immunoprecipitated [35S]interferon-beta. The former was used to study the biosynthesis of interferon-beta in human diploid fibroblasts. In cells induced with double-stranded RNA (poly(I:C] to synthesize interferon-beta, two intracellular proteins with estimated molecular weights of 23,000 and 18,000 were precipitated with the antiserum. Three exocellular proteins from the same induced cells were precipitated with molecular weights of 23,000, 18,000, and 10,000. Reduced amounts of the intra- and exocellular Mr = 23,000 component and enhanced amounts of the Mr = 18,000 component were observed when induced cells were treated with the glycosylation inhibitor tunicamycin. Neither the antibody to peptide 1-21, the antibody to peptide 18-45, nor a combination of both antibodies neutralized the interferon-beta antiviral activity. We conclude that the amino terminus of interferon-beta may not be involved in the binding of interferon-beta to its receptor.

The specifically mutated enzyme thiol beta-lactamase has been expressed in Escherichia coli by means of the trp promoter and purified to homogeneity. The gene for this enzyme results from a single base change N410 A----T in the gene of pBR322 RTEM beta-lactamase (EC 3.5.2.6, penicillinase, penicillin amido-beta-lactamhydrolase) which alters the codon for the active site Ser 70 to that for Cys. Precursor thiol beta-lactamase is processed to give the same NH2-terminal sequence as that for wild type enzyme. In contrast to the wild type enzyme, thiol beta-lactamase contains one free titratable thiol group/molecule. Thiol beta-lactamase catalyzes the hydrolysis of beta-lactams with a substrate specificity that is distinct from that of wild type enzyme. For benzyl-penicillin and ampicillin, the Km values are similar to wild type values although the kcat values are 1-2% that of wild type enzyme. For the cephalosporin nitrocefin, the Km is greater than 10-fold that of the wild type and the kcat is at least as large as the kcat for the wild type enzyme. Thiol beta-lactamase is different from wild type beta-lactamase in that it is not competitively inhibited by boric acid although a small degree of noncompetitive inhibition does occur. Whereas the circular dichroism spectra of both enzymes are nearly identical, thiol beta-lactamase at 40 degrees C is 3-fold more resistant to trypsin than is the wild type enzyme.

Based on a consideration of the probable structure of calmodulin and some natural peptides known to interact with it, two calmodulin-binding peptides were designed. These peptides bind to calmodulin in helical conformations and are capable of forming electrostatic and hydrophobic interactions with calmodulin. Their dissociation constants for binding (less than or equal to 210 and 400 pM) place them as the tightest-binding inhibitors of calmodulin thus far reported. The study of the interactions of these and similar peptides with calmodulin will provide valuable insights into the mechanisms whereby calmodulin binds to target enzymes, and it also serves as an excellent model system for exploring the physical chemistry of protein-protein interaction.

Upon the addition of interferon (IFN) to cultured human cells, the expression of genes encoding the 56K and the guanylate binding protein (GBP) is specifically induced. We have analyzed their expression at the protein and the mRNA levels and studied how their regulation differs in cells treated with different IFNs. In the type I IFN (alpha and beta)-treated cells, we detected the accumulation of the 56K protein primarily in the cytoplasm. The 56K protein was undetectable in untreated cells or in cells treated with type II IFN (IFN-gamma). In contrast, a greater amount of GBP was synthesized in cells treated with type II IFN than in cells treated with type I IFN. The differential induction of these two proteins correlates well with the relative amounts of their mRNAs in type I and type II IFN-treated cells. In addition, the IFN-induced synthesis of the 56K protein was found in certain cell lines in which the GBP synthesis was not detected. These results suggest that the regulation of these two genes requires dissimilar factors which are activated or induced to different extents by type I and type II IFNs.

Amino-terminal sequence analysis of human T-cell growth factor indicated that the amino acid in position 3 of the polypeptide chain was modified. Examination of the N-terminal octapeptide using the amino acid analyzer and mass spectrometry demonstrated that position 3 was a threonine which was linked to N-acetyl-D-galactosamine. This site of glycosylation is of practical significance since it appears to play a role in the selectivity of a monoclonal antibody for the factor.

A novel photoreactive amino acid has been incorporated synthetically into two model peptides and the calmodulin-binding domain from myosin light chain kinase. Cross-linked photoadducts of each peptide with calmodulin have been prepared and digested by chemical and/or enzymatic methods to determine the site of label attachment. Depending on the position of the photoprobe in the peptide sequence, either Met-144 or Met-71 is photolabeled. These results are discussed in relation to the three-dimensional structure of calmodulin obtained crystallographically and the known solution properties of calmodulin.

By using interactive computer graphics, two models for calmodulin have been constructed based on the structures of two functionally and structurally related proteins, intestinal calcium-binding protein and carp parvalbumin. The two models have been compared and contrasted to the parent proteins with respect to proportion of solvent-exposed hydrophobic residues, solvent-accessible surface area, and side-chain packing. Electrostatic potential surfaces generated for the models suggest a probable binding site for basic amphiphilic alpha-helical peptides located between the last E and F helices in the second domain of calmodulin. Both electrostatic and hydrophobic complementarity can contribute to stabilization of a peptide-protein complex in this region.

The question of how the primary amino acid sequence of a protein determines its three-dimensional structure is still unanswered. One approach to this problem involves the de novo design of model peptides and proteins that should adopt desired three-dimensional structures. A systematic approach was aimed at the design of a four-helix bundle protein. The gene encoding the designed protein was synthesized and the protein was expressed in Escherichia coli and purified to homogeneity. The protein was shown to be monomeric, highly helical, and very stable to denaturation by guanidine hydrochloride (GuHCl). Thus a globular protein has been designed that is capable of adopting a stable, folded structure in aqueous solution.

Ion channel proteins are important for the conduction of ions across biological membranes. Recent analyses of their sequences have suggested that they are composed of bundles of alpha-helices that associate to form ion-conducting channels. To gain insight into the mechanisms by which alpha-helices can aggregate and conduct ions, three model peptides containing only leucine and serine residues were synthesized and characterized. A 21-residue peptide, H2N-(Leu-Ser-Ser-Leu-Leu-Ser-Leu)3-CONH2, which was designed to be a membrane-spanning amphiphilic alpha-helix, formed well-defined ion channels with ion permeability and lifetime characteristics resembling the acetylcholine receptor. In contrast, a 14-residue version of this peptide, which was too short to span the phospolipid bilayer as an alpha-helix, failed to form discrete, stable channels. A third peptide, H2N-(Leu-Ser-Leu-Leu-Leu-Ser-Leu)3-CONH2, in which one serine per heptad repeat was replaced by leucine, produced proton-selective channels. Computer graphics and energy minimization were used to create molecular models that were consistent with the observed properties of the channels.

Calmodulin has recently been shown to form exceptionally tight, calcium-dependent complexes with several natural peptides (Kdiss greater than 10(-7) M). These peptides were demonstrated to be capable of forming basic, amphiphilic alpha-helices. To further illustrate the importance of this structural feature for calmodulin binding, several other amphiphilic alpha-helical peptides were tested for their ability to bind calmodulin. To monitor complexes of high affinity (greater than 10(8) M-1), a new competition assay was devised with Sepharose 4B-conjugated melittin. Stoichiometries were assessed by electrophoresis and equilibrium size exclusion chromatography. Three peptides, which were designed to form idealized amphiphilic alpha-helices were tested. The basic peptides, N alpha-9-fluorenylmethoxycarboxyl-(FMOC)-(Leu-Lys-Lys-Leu-Leu-Lys-L eu)1 and FMOC-(Leu-Lys-Lys-Leu-Leu-Lys-Leu)2 bind calmodulin in a 1:1 complex with dissociation constants of 150 and 3 nM, respectively. The acidic peptide, FMOC-(Leu-Glu-Glu-Leu-Leu-Glu-Leu)2 failed to bind calmodulin, even at micromolar concentrations. Complex formation between calmodulin and the 14-residue basic peptide leads to an increase in the helicity of the complex which is attributed to an increase of about 50% in the helicity of the peptide. Calmodulin also interacts with the neutral alpha-helical peptide toxin delta-hemolysin. Concomitant with binding, the fluorescence maximum of the unique Trp residue increases 2-fold and is blue-shifted. A dissociation constant could not be unambiguously estimated though, since delta-hemolysin has a strong tendency to self-aggregate. The above data support our hypothesis that a basic, amphiphilic alpha-helix is a structural feature which underlies the calmodulin-binding properties common to a variety of peptides.