Molecular recognition of macrocyclic peptidomimetic inhibitors by HIV-1 protease

High-resolution crystal structures are described for seven macrocycles complexed with HIV-1 protease (HIVPR). The macrocycles possess two amides and an aromatic group within 15-17 membered rings designed to replace N- or C-terminal tripeptides from peptidic inhibitors of HIVPR. Appended to each macrocycle is a transition state isostere and either an acyclic peptide, nonpeptide, or another macrocycle. These cyclic analogues are potent inhibitors of HIVPR, and the crystal structures show them to be structural mimics of acyclic peptides, binding in the active site of HIVPR via the same interactions. Each macrocycle is restrained to adopt a P-strand conformation which is preorganized for protease binding. An unusual feature of the binding of C-terminal macrocyclic inhibitors is the interaction between a positively charged secondary amine and a catalytic aspartate of HIVPR. A bicyclic inhibitor binds similarly through its secondary amine that lies between its component N-terminal and C-terminal macrocycles. In contrast, the corresponding tertiary amine of the N-terminal macrocycles does not interact with the catalytic aspartates. The amine-aspartate interaction induces a 1.5 Angstrom N-terminal translation of the inhibitors in the active site and is accompanied by weakened interactions with a water molecule that bridges the ligand to the enzyme, as well as static disorder in enzyme flap residues. This flexibility may facilitate peptide cleavage and product dissociation during catalysis. Proteases [Aba(67,95)]HIVPR and [Lys(7),Ile(33),Aba(67,95)]- HIVPR used in this work were shown to have very similar crystal structures.

Structure-activity relationships of omega-conotoxins MVIIA, MVIIC and 14 loop splice hybrids at N and P/Q-type calcium channels

The omega-conotoxins are a set of structurally related, four-loop, six cysteine containing peptides, that have a range of selectivities for different subtypes of the voltage-sensitive calcium channel (VSCC). To investigate the basis of the selectivity displayed by these peptides, we have studied the binding affinities of two naturally occurring omega-conotoxins, MVIIA and MVIIC and a series of 14 MVIIA/MVIIC loop hybrids using radioligand binding assays for N and P/Q-type Ca2+ channels in rat brain tissue. A selectivity profile was developed from the ratio of relative potencies at N-type VSCCs (using [I-125]GVIA radioligand binding assays) and P/Q-type VSCCs (using [I-125]MVIIC radioligand binding assays). in these peptides, loops 2 and 4 make the greatest contribution to VSCC subtype selectivity, while the effects of loops 1 and 3 are negligible. Peptides with homogenous combinations of loop 2 and 4 display clear selectivity preferences, while those with heterogeneous combinations of loops 2 and 4 are less discriminatory. H-1 NMR spectroscopy revealed that the global folds of MVIIA, MVIIC and the 14 loop hybrid peptides were similar; however, several differences in local structure were identified. Based on the binding data and the 3D structures of MVIIA, GVIA and MVIIC, we have developed a preliminary pharmacophore based on the omega-conotoxin residues most Likely to interact with the N-type VSCC. (C) 1999 Academic Press.

Solution structure of alpha-conotoxin ImI by H-1 nuclear magnetic resonance

alpha-Conotoxin ImI derives from the venom of Conus imperialis and is the first and only small-peptide ligand that selectively binds to the neuronal alpha(7) homopentameric subtype of the nicotinic acetylcholine receptor (nAChR). This receptor subtype is a possible drug target for several neurological disorders. The cysteines are connected in the pairs Cys2-Cys8 and Cys3-Cys12, To date it is the only alpha-conotoxin with a 4/3 residue spacing between the cysteines, The structure of ImI has been determined by H-1 NMR spectroscopy in aqueous solution, The NMR structure is of high quality, with a backbone pairwise rmsd of 0.34 Angstrom for a family of 19 structures, and comprises primarily a series of nested beta turns. Addition of organic solvent does not perturb the solution structure. The first eight residues of ImI are identical to the larger, but related, conotoxin EpI and adopt a similar structure, despite a truncated second loop. Residues important for binding of ImI to the alpha 7 nAChR are all clustered on one face of the molecule. Once further binding data for EPI and ImI are available, the ImI structure will allow for design of novel alpha(7) nAChR-specific agonists and antagonists with a wide range of potential pharmaceutical applications.

Solution structure of a defensin-like peptide from platypus venom

Three defensin-like peptides (DLPs) were isolated from platypus venom and sequenced. One of these peptides, DLP-1, was synthesized chemically and its three-dimensional structure was determined using NMR spectroscopy. The main structural elements of this 42-residue peptide were an anti-parallel beta-sheet comprising residues 15-18 and 37-40 and a small 3(10) helix spanning residues 10-12. The overall three-dimensional fold is similar to that of beta-defensin-12, and similar to the sodium-channel neurotoxin ShI (Stichodactyla helianthus neurotoxin I). However, the side chains known to be functionally important in beta-defensin-12 and ShI are not conserved in DLP-1, suggesting that it has a different biological function. Consistent with this contention, we showed that DLP-1 possesses no anti-microbial properties and has no observable activity on rat dorsal-root-ganglion sodium-channel currents.

Chemical synthesis and folding pathways of large cyclic polypeptide: Studies of the cystine knot polypeptide kalata B1

Kalata B1 is a member of a new family of polypeptides, isolated from. plants, which have a cystine knot structure embedded within an amide-cyclized backbone. This family of molecules are the largest known cyclic peptides, and thus, the mechanism of synthesis and folding is of great interest. To provide information about both these phenomena, we have synthesized kalata B1 using two distinct strategies. In the first, oxidation of the cysteine residues of a linear precursor peptide to form the correct disulfide bonds results in folding of the three-dimensional structure and preorganization of the termini in close proximity for subsequent cyclization. The second approach involved cyclization prior to oxidation. In the first method, the correctly folded peptide was produced only in the presence of partially hydrophobic solvent conditions. These conditions are presumably required to stabilize the surface-exposed hydrophobic residues. However,; in the synthesis,involving cyclization prior to oxidation, the cyclic reduced peptide folded to a significant degree in the absence of hydrophobic solvents and even more efficiently in the presence of hydrophobic solvents. Cyclization clearly has a major effect on the folding pathway and facilitates formation of the correctly disulfide-bonded form in aqueous solution; In addition to facilitating folding to a compact stable structure cyclization has an important effect on biological activity as assessed by hemolytic activity.

NMR studies of the low-density lipoprotein receptor-binding peptide of apolipoprotein E bound to dodecylphosphocholine micelles

Circular dichroism and NMR spectroscopy have been used to determine the structure of the low-density lipoprotein (LDL) receptor-binding peptide, comprising residues 130-152, of the human apolipoprotein E. This peptide has little persistent three-dimensional structure in solution, but when bound to micelles of dodecylphosphocholine (DPC) it adopts a predominantly alpha-helical structure. The three-dimensional structure of the DPC-bound peptide has been determined by using H-1-NMR spectroscopy: the structure derived from NOE-based distance constraints and restrained molecular dynamics is largely helical. The derived phi and psi angle order parameters show that the helical structure is well defined but with some flexibility that causes the structures not to be superimposable over the full peptide length. Deuterium exchange experiments suggest that many peptide amide groups are readily accessible to the solvent, but those associated with hydrophobic residues exchange more slowly, and this helix is thus likely to be positioned on the surface of the DPC micelles. In this conformation the peptide has one hydrophobic face and two that are rich in basic amino acid side chains. The solvent-exposed face of the peptide contains residues previously shown to be involved in binding to the LDL receptor.

Identification of a new ligand binding domain in the al subunit of the inhibitory glycine receptor

Four discontinuous extracellular sequence domains have been proposed to form the ligand binding sites of the ligand-gated ion channel receptor superfamily. In this study, we investigated the role of 12 contiguous residues of the inhibitory glycine receptor that define the proposed loop A ligand binding domain; Using the techniques of site-directed mutagenesis and patch-clamp electrophysiology, four of the 12 residues were shown to have impaired ligand binding. Three mutants, I93A, A101H, and N102A, resulted in significant (17-44-fold) increases in the agonist EC50 values as compared with the wild-type glycine receptor, whereas Hill coefficients, I-max values, and antagonist affinity remained largely unaffected. Consideration of receptor efficacy values indicates that these residues are involved in ligand binding rather than channel activation. A fourth mutant, W94A, failed to give rise to any glycine-activated currents, although cell-surface expression was observed, suggesting that this residue may also be involved in agonist binding. These data provide the most extensive characterization of the loop A ligand binding domain available to date and define two new residue locations, Ile(93) and Asn(102), as contributing to the four-loop model of ligand binding.

Comparison of acetylcholinesterase genes from cattle ticks

We describe the isolation and characterisation of two putatively new acetylcholinesterase genes from the African cattle ticks Boophilus decoloratus and Rhipicephalus appendiculatus. The nucleotide sequences of these genes had 93% homology to each other and 95% and 91% identity, respectively, to the acetylcholinesterase gene from an Australian strain of another cattle tick, Boophilus microplus. Translation of the nucleotide sequences revealed putative amino acids that are essential for acetylcholinesterase activity: the active site serine, and the histidine and glutamate residues that associate with this serine to form the catalytic triad. All known acetylcholinesterases have three sets of cysteines that form disulfide bonds; however, the acetylcholinesterase genes of these three species of ticks encode only two sets of cysteines. Acetylcholinesterases of B. microplus from South Africa, Zimbabwe, Kenya and Mexico had 98-99% identity with acetylcholinesterase from B. microplus from Australia, whereas acetylcholinesterase from B. microplus from Indonesia was identical to that from Australia. Preliminary phylogenetic analyses surprisingly indicate that the acetylcholinesterases of ticks are closer phylogenetically to acetylcholinesterases of vertebrates than they are to those of other arthropods. (C) 1999 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved.

Plant cyclotides: A unique family of cyclic and knotted proteins that defines the cyclic cystine knot structural motif

Several macrocyclic peptides (similar to 30 amino acids), with diverse biological activities, have been isolated from the Rubiaceae and Violaceae plant families over recent years. We have significantly expanded the range of known macrocyclic peptides with the discovery of 16 novel peptides from extracts of Viola hederaceae, Viola odorata and Oldenlandia affinis. The Viola plants had not previously been examined for these peptides and thus represent novel species in which these unusual macrocyclic peptides are produced. Further, we have determined the three-dimensional struc ture of one of these novel peptides, cycloviolacin O1, using H-1 NMR spectroscopy. The structure consists of a distorted triple-stranded beta-sheet and a cystine-knot arrangement of the disulfide bonds. This structure is similar to kalata B1 and circulin A, the only two macrocyclic peptides for which a structure was available, suggesting that despite the sequence variation throughout the peptides they form a family in which the overall fold is conserved. We refer to these peptides as the cyclotide family and their embedded topology as the cyclic cystine knot (CCK) motif. The unique cyclic and knotted nature of these molecules makes them a fascinating example of topologically complex proteins. Examination of the sequences reveals they can be separated into two subfamilies, one of which tends to contain a larger number of positively charged residues and has a bracelet-like circularization of the backbone. The second subfamily contains a backbone twist due to a cis-Pro peptide bond and may conceptually be regarded as a molecular Moebius strip. Here we define the structural features of the two apparent subfamilies of the CCK peptides which may be significant for the likely defense related role of these peptides within plants. (C) 1999 Academic Press.

The role of disulfide-linked dimerization in interleukin-3 receptor signaling and biological activity

Cysteine residues 86 and 91 of the beta subunit of the human interleukin (hIL)-3 receptor (h beta c) participate in disulfide-linked receptor subunit heterodimerization. This linkage is essential for receptor tyrosine phosphorylation, since the Cys-86 --> Ala (Mc4) and Cys-91 --> Ala (Mc5) mutations abolished both events. Here, we used these mutants to examine whether disulfide-linked receptor dimerization affects the biological and biochemical activities of the IL-3 receptor. Murine T cells expressing hIL-3R alpha and Mc4 or Mc5 did not proliferate in hIL-3, whereas cells expressing wild-type h beta c exhibited rapid proliferation. However, a small subpopulation of cells expressing each mutant could be selected for growth in IL-3, and these proliferated similarly to cells expressing wild-type h beta c, despite failing to undergo IL-3-stimulated h beta e tyrosine phosphorylation. The Mc4 and Mc5 mutations substantially reduced, but did not abrogate, IL-3-mediated anti-apoptotic activity in the unselected populations. Moreover, the mutations abolished IL-3-induced JAK2, STAT, and AKT activation in the unselected cells, whereas activation of these molecules in IL-3-selected cells was normal. In contrast, Mc4 and Mc5 showed a limited effect on activation of Erk1 and -2 in unselected cells. These data suggest that whereas disulfide-mediated cross-linking and h beta c tyrosine phosphorylation are normally important for receptor activation, alternative mechanisms can bypass these requirements.

Leu(10) of alpha-conotoxin PnIB confers potency for neuronal nicotinic responses in bovine chromaffin cells

Two alpha-conotoxins PnIA and PnIB (previously reported as being mollusc specific) which differ in only two amino acid residues (AN versus LS at residues 10 and 11, respectively), show markedly different inhibition of the neuronal nicotinic acetylcholine receptor response in bovine chromaffin cells, a mammalian preparation. Whereas alpha-conotoxin PnIB completely inhibits the nicotine-evoked catecholamine release at 10 mu M, with IC50 = 0.7 mu M, alpha-conotoxin PnIA is some 30-40 times less potent. Two peptide analogues, [A10L]PnIA and [N11S]PnIA were synthesized to investigate the extent to which each residue contributes to activity. [A10L]PnIA (IC50 = 2.0 mu M) completely inhibits catecholamine release at 10 mu M whereas [N11S]PnIA shows Little inhibition. In contrast, none of the peptides inhibit muscle-type nicotinic responses in the rat hemi-diaphragm preparation. We conclude that the enhanced potency of alpha-conotoxin PnIB over alpha-conotoxin PnIA in the neuronal-type nicotinic response is principally determined by the larger, more hydrophobic leucine residue at position 10 in alpha-conotoxin PnIB. (C) 2000 Elsevier Science B.V. All rights reserved.

FHA domain boundaries of the Dun1p and Rad53p cell cycle checkpoint kinases

Dun1p and Rad53p of the budding yeast Saccharomyces cerevisiae are members of a conserved family of cell cycle checkpoint protein kinases that contain forkhead-associated (FHA) domains. Here, we demonstrate that these FHA domains contain 130-140 residues, and are thus considerably larger than previously predicted by sequence comparisons (55-75 residues), In vivo, expression of the proteolytically defined Dun1p FHA domain, but not a fragment containing only the predicted domain boundaries, inhibited the transcriptional induction of repair genes following replication blocks, This indicates that the non-catalytic FI-IA domain plays an important role in the transcriptional function of the Dun1p protein kinase. (C) 2000 Federation of European Biochemical Societies.

Discovery and characterization of a family of insecticidal neurotoxins with a rare vicinal disulfide bridge

We have isolated a family of insect-selective neurotoxins from the venom of the Australian funnel-web spider that appear to be good candidates for biopesticide engineering. These peptides, which we have named the Janus-faced atracotoxins (J-ACTXs), each contain 36 or 37 residues, with four disulfide bridges, and they show no homology to any sequences in the protein/DNA databases. The three-dimensional structure of one of these toxins reveals an extremely rare vicinal disulfide bridge that we demonstrate to be critical for insecticidal activity. We propose that J-ACTX comprises an ancestral protein fold that we refer to as the disulfide-directed beta-hairpin.

Identification of regions within the third FnIII-like domain of the IL-5R alpha involved in IL-5 interaction

Previously, two binding sites for interleukin 5 (IL-5) were identified on the IL-5 receptor alpha chain (IL-5R alpha). They are located within the CD loop of the first fibronectin type III (FnIII)-like domain and the EF loop of the second FnIII-like domain. The first binding site was identified by exploiting the different abilities of human IL-5R alpha (hIL-5R alpha) and mouse IL-5R alpha (mIL-5R alpha) to bind hIL-5. Here we show that ovine IL-5 (oIL-5) has the ability to activate the hIL-5R alpha but not the mIL-5R alpha. By using chimeras of the mIL-5R alpha and hIL-5R alpha we demonstrate that residues within the first and third FnIII-like domains of mIL-5R alpha are responsible for this lack of activity. Furthermore, mutation of residues on hIL-5R alpha to mIL-5R alpha within the predicted DE and FG loop regions of the third FnIII domain reduces oIL-5 activity, These results show that regions of the third FnIII domain of IL-5R alpha are involved in binding, in addition to the regions in domains one and two of the IL-5R alpha that were identified in an earlier study. (C) 2000 Academic Press.

Synthesis and structure elucidation of kappa-casein (45-87) - A two dimensional nuclear magnetic resonance study

We have shown that 44 amino acid residues N-terminal segment of kappa-casein exhibits considerable a-helical structure. This prompted us to investigate the structures of the remaining segments of kappa-casein. Thus, in this study the chemical synthesis and structure elucidation of the peptide 45-87 amino acid residues of kappa-casein is reported. The peptide was assembled using solid phase peptide synthesis methodology on pam resin, cleaved via HF, freeze dried and, after purification, characterised by mass spectrometry (observed m/z 4929; calculated mit 4929.83). The amino acid sequence of the peptide is: CKPVALINNQFLPYPYYAKPAAVRSPAQILQWQVLSNTVPAKA Its structure elucidation has been carried out using circular dichroism (CD) and nuclear magnetic resonance (NMR) techniques. CD spectrum of the peptide shows it to be a random structure in water but in 30% trifluoroethanol the peptide exhibits considerable structure. The 1D and 2D NMR spectra corroborated the results of CD. The structure elucidation of the peptide using TOCSY and NOESY NMR techniques will be discussed.