Design of dimerization inhibitors of HIV-1 aspartic proteinase: a computer-based combinatorial approach.

Inhibition of dimerization to the active form of the HIV-1 aspartic proteinase (HIV-1 PR) may be a way to decrease the probability of escape mutations for this viral protein. The Multiple Copy Simultaneous Search (MCSS) methodology was used to generate functionality maps for the dimerization interface of HIV-1 PR. The positions of the MCSS minima of 19 organic fragments, once postprocessed to take into account solvation effects, are in good agreement with experimental data on peptides that bind to the interface. The MCSS minima combined with an approach for computational combinatorial ligand design yielded a set of modified HIV-1 PR C-terminal peptides that are similar to known nanomolar inhibitors of HIV-1 PR dimerization. A number of N-substituted 2,5-diketopiperazines are predicted to be potential dimerization inhibitors of HIV-1 PR.

Lipopeptides as dimerization inhibitors of HIV-1 protease.

In AIDS therapy, attempts have been made to inhibit the virus-encoded enzymes, e.g. HIV-1 protease, using active site-directed inhibitors. This approach is questionable, however, due to virus mutations and the high toxicity of the drugs. An alternative method to inhibit the dimeric HIV protease is the targeting of the interface region of the protease subunits in order to prevent subunit dimerization and enzyme activity. This approach should be less prone to inactivation by mutation. A list of improved 'dimerization inhibitors' of HIV-1 protease is presented. The main structural features are a short 'interface' peptide segment, including non-natural amino acids, and an aliphatic N-terminal blocking group. The high inhibitory power of some of the lipopeptides [e.g. palmitoyl-Tyr-Glu-Leu-OH, palmitoyl-Tyr-Glu-(L-thyronine)-OH, palmitoyl-Tyr-Glu-(L-biphenyl-alanine)-OH] with low nanomolar Ki values in the enzyme test suggests that mimetics with good bio-availability can be derived for AIDS therapy.

A rapid and sensitive bacterial assay to determine the inhibitory effect of 'interface' peptides on HIV-1 protease co-expressed in Escherichia coli.

The HIV-1 protease is essential for maturation of virus particles and is, therefore, an attractive target for antiviral drugs. The function of this protease depends on the dimerization of two identical subunits. Commonly used protease inhibitors are directed mainly against the active site of the enzyme which often leads to viral resistance. To determine the inhibitory effect of peptides interfering with the dimerization site of the HIV-1 protease, a recombinant bacterial screening assay was established. Escherichia coli was co-transformed with two different plasmids, expressing the 'interface' peptide and an active HIV-1 protease toxic for the bacteria. Co-expression of inhibitory peptides overcomes the incomplete membrane transmission of supplemented inhibitors and leads to a direct interaction of the inhibitory peptide and the HIV-1 protease. The inhibitory effect of co-expressed peptides was measured by an increased growth of co-transformed bacteria, compared with a slowly growing E. coli control culture only expressing the HIV-1 protease. Using this assay several penta- and hexa-peptides were screened for their ability to inhibit HIV-1 protease activity. One of these peptides showed a significant inhibitory effect on co-expressed recombinant HIV-1 protease.

Screening of inhibitors of HIV-1 protease using an Escherichia coli cell assay.

To evaluate the available peptidic and pseudopeptidic inhibitors of HIV protease for their possible in vivo activity, a screening test using Escherichia coli was established. E. coli cells carrying the plasmid pET9c-PR containing the gene for HIV-1 protease under the control of a T7-promotor are grown in the absence and in the presence of inhibitors. The action of the toxic protease produced by the cells is counteracted by the inhibitors. Provided sufficient membrane permeability of the inhibitors exists, this results in accelerated cell growth. From the peptides known to be active in an in-vitro enzyme test, most compounds inhibit HIV protease only to a limited degree in this test. However, two short peptides (Ac-Ser-Tyr-Glu-Leu and Lys-Ile-Ser-Tyr-Asp-Tyr) protect cell growth to an considerabe extent, thus indicating that they reach the E. coli cytosol and there block HIV protease. Two pseudopeptides known to be very potent in the enzyme test (SDZ PRI 053 and CIBA 61755) also inhibit HIV-1 protease strongly in this cell growth test.

Triterpenes as potential dimerization inhibitors of HIV-1 protease.

HIV-1 protease is a homodimeric enzyme. A beta-sheet consisting of the four terminal segments provides the main driving force for dimerization of the per se inactive protomers. Several short peptides with sequences related to the terminal sequences of the protease are able to inhibit dimerization by blocking the 'interface' part of the monomers. From the structures of such inhibitory peptides a 'pharmacophore' could be derived. By using a prominent distance from this pharmacophore scaffold for a library search (Cambridge Structural Database), non-peptide inhibitors of HIV-1 protease with polycyclic triterpene structure could be found. The IC50 constants of these compounds are near 1 microM. One of the triterpenes, the ursolic acid (Ki = 3.4 microM), was further kinetically analysed (according to Zhang). The shape of the graph confirms the expected mechanism of dimerization inhibition.

The inhibition of human immunodeficiency virus proteases by 'interface peptides'.

The active human immunodeficiency virus type 1 (HIV-1) protease has a homodimeric structure, the subunits are connected by an 'interface' beta-sheet formed by the NH2- and COOH-terminal amino acid segments. Short peptides derived from these segments are able to inhibit the protease activity in the range of micromolar IC50 values. We have further improved the inhibitory power of such peptides by computer modelling. The best inhibitor, the palmitoyl-blocked peptide Pam-Thr-Val-Ser-Tyr-Glu-Leu, has an IC50 value of less than 1 microM. Some of the peptides also showed very good inhibition of the HIV-2 protease. The C-terminal segment of the HIV-1 matrix protein, Acetyl-Gln-Val-Ser-Gln-Asn-Tyr, also inhibits HIV-1 protease. Kinetic studies confirmed the 'dissociative' mechanism of inhibition by the peptides. Depending on the peptide structure and ionic strength, both dimerization inhibition and competitive inhibition were observed, as well as synergistic effects between competitive inhibitors and interface peptides.

Facile syntheses of C2-symmetrical HIV-1 protease inhibitors.

With the goal of obtaining inexpensive yet potent anti-AIDS drugs, simple inhibitors of HIV-1 protease were synthesised. The C2-symmetrical pseudopeptidic substrate analogues can be prepared as inhibitors for HIV-1 protease starting from symmetrical ketones 3a-d by a facile four-step synthesis. After bromination of 3a-d to alpha,alpha'-dibromoketones 4a-d, we synthesised the diamino compounds 6a-c by Gabriel synthesis, which were then coupled with Z-valine to yield inhibitors including a central hydroxy group 8a-d a-i by azidation, reduction with LiAlH4 and coupling of the beta,beta'-diaminohydroxy compounds with appropriate peptides. The first set of compounds showed only weak inhibition whereas the latter reach Ki values of up to 3.0 microM.

Analysis of substrate cleavage by recombinant protease of human T cell leukaemia virus type 1 reveals preferences and specificity of binding.

Human T cell leukaemia virus type 1 (HTLV-1) protease (PR14) was expressed in bacteria and purified by gel filtration. A continuous spectrophotometric assay was used to measure the kinetic parameters of substrate hydrolysis by PR14. Several peptide substrates containing HTLV-1 sequences known to be cleaved by PR14 were used. Cleavage analysis showed that the affinity with which PR14 binds these substrates is higher than that previously reported for HTLV-1 Gag peptides. Also, the affinities of peptides containing the sites involved in autocleavage of protease from its precursor are higher than for the peptides containing sites required for structural protein maturation. This suggests that the autocatalysis of protease from its own precursor has priority over other cleavage reactions and supports similar observations of an ordered hierarchy of processing events by retroviral proteases. As the N- and C-terminal regions of retroviral aspartic proteases are known to contribute to stability of the dimer by forming antiparallel beta-strands, short peptides corresponding to these terminal sequences of HTLV-1 protease were tested for their ability to inhibit cleavage of substrates by PR14. Inhibition was seen with a C-terminal peptide corresponding exactly to the C-terminal 11 amino acids of the processed PR14, whereas a peptide containing a sequence situated further from the C terminus was less effective. An inhibitor of the protease of human immunodeficiency virus type 1, Ro 31-8959, was found to be a poor inhibitor of PR14.

The inhibition of HIV-1 protease by interface peptides.

Previous studies have shown that some peptides derived from one of the terminal amino acid segments of the homodimeric HIV-1 protease show moderate inhibition of this enzyme probably by interfering with the "interface" structure formed by the four terminal segments of the dimer. Different peptides, with improved inhibitory potency, were devised by computer modelling, synthesized, and tested. Ac-TVSFNF, the short peptide with the best inhibition so far (IC50 = 80 microM) is identical with the C-terminal part of the gag-pol frame shift protein p6*. This suggests a regulatory function of p6* as a dimerization inhibitor of HIV protease in the virion. Peptides derived from the active site sequence of PR are inactive. The two terminal hexapeptides of reverse transcriptase are also inactive in the HIV-1 PR activity assay.

Inhibition of HIV-1 protease by short peptides derived from the terminal segments of the protease.

The active HIV-1 protease is a homodimeric enzyme. A beta-sheet consisting of N- and C-terminal segments provides the main driving force for dimerization of the inactive protomers. Several short peptides with sequences derived from the N- and C-termini of the protease were tested for inhibition of protease activity and for inhibition of HIV-1 replication in lymphocytes. Medium inhibitory activity was found with each of the peptides in the enzyme test and no inhibition of the lymphocytes was found up to 200 micrograms/ml. The enzyme tests indicate that HIV-1 protease is the target of the inhibitory action. Synergistic action could not be found with pairs of the peptides derived from the two different termini. Prolonged incubation with one of the peptides increased inhibition indicating a slow dissociation of the protease dimers. No cytotoxic effect of the inhibitors could be found below 200 micrograms/ml.

alpha 2-Macroglobulin is cleaved by HIV-1 protease in the bait region but not in the C-terminal inter-domain region.

alpha 2-Macroglobulin is cleaved by human immunodeficiency virus-1 protease. The cleavage site is the Phe684-Tyr685 bond in the "bait region", an exposed part of alpha 2-macroglobulin, creating the "F-form". The methylamine derivative of alpha 2-macroglobulin is also cleaved at the same bond. The homologous chicken ovomacroglobulin does not form an F-form structure with the protease, although, F-form generation by other enzymes is known. This is possibly due to the lack of a suitable cleavage sequence in the corresponding region of ovomacroglobulin. In human alpha 2-macroglobulin, the interdomain segment between the main part of the molecule and the receptor-binding C-terminal domain is not cleaved by the HIV protease although typical cleavage sequences occur. In AIDS, therefore, HIV protease from infected cells in unlikely to interfere with receptor-binding of alpha 2-macroglobulin.

HIV-1 reproduction is inhibited by peptides derived frm the N- and C-termini of HIV-1 protease.

Two octapeptides derived from the sequence of the N- and C-termini of HIV-1 protease were tested for their ability to inhibit HIV-1 reproduction. Weak inhibitory activity was found with each of the two peptides. It is assumed that HIV-1 protease is the target of the inhibitory action. In spite of the moderate inhibitory activity the results are encouraging since they may be improved by various means.

Cleavage of phosphorylase kinase and calcium-free calmodulin by HIV-1 protease.

Phosphorylase kinase and calcium-free calmodulin are digested by human immunodeficiency virus-1 protease. In phosphorylase kinase, the alpha subunit is preferentially hydrolyzed at arg748-val749. The beta subunit is cleaved only slowly at leu678-pro679, and calmodulin, the integral delta subunit of phosphorylase kinase, is not cleaved at all. However, free calmodulin in the calcium-depleted form showed to be a good substrate for the protease. Here the cleavage occurs at phe65-pro66 and met71-met72. This fast hydrolysis of free calmodulin can be blocked by micromolar concentrations of Ca2+ or millimolar concentrations of Mg2+.

Electron microscopical examination of different aggregation and decomposition states of phosphorylase kinase. Identification and computer averaging of the alpha gamma delta fragment.

A sample of phosphorylase kinase aged by storage was subjected to chromatography on a size exclusion column. Samples from the four major peaks were analysed by electron microscopy. The first major peak consisted of oligomers of the enzyme monomers which showed no structural order. The second peak mainly consisted of dimers of phosphorylase kinase connected to each other by the outer tipes of the "wings", thus identifying a binding locus through cross-labelling. The third peak showed the already known typical forms of phosphorylase kinase ("butterflies", "chalice form"). The last peak yielded structures identical to the alpha gamma delta fragments obtained by lithium bromide cleavage of the intact enzyme. These structures correspond to the curved outer parts of the cup substructures of the chalice and butterfly forms. The result could be verified by computer averaging of the alpha gamma delta fragments. This finding confirms reports (Chan, K.-F. J. & Graves, D.J. (1984) Calcium Cell Funct. 5, 1-31) that the beta-subunits (missing in the distally situated fragment) are located in the central parts of the molecules. The good quality of the averages of the low molecular mass fragment (about 200 kDa) recommends computer averaging of electron micrographs of partial complexes as suitable method for the study of protein complexes.

Electron microscopical localization of the alpha 2-macroglobulin thiol ester sites.

The active thiol ester groups of alpha 2-macroglobulin (alpha 2M) were reacted with a biotin derivative and the sites labelled with avidin-ferritin complexes. Electron micrographs show a strong preference of attachment of the ferritins to the ends of the rods of the H-shaped molecules. A mutual "cross-labelling" was observed in an alpha 2M preparation which yielded dimers of the molecules which must have been formed during purification. The molecules were mostly attached to each other at the ends of the rods of the H-shaped molecules. It is concluded that the thiol esters responsible for the covalent attachment of the proteinases (and other molecules) may be located more in the distal parts of the alpha 2M molecules, while the proteinase molecules are finally trapped near to the centre of the alpha 2M molecules.

Electron microscopical investigation of citrate lyase single molecules.

Electron micrographs of citrate lyase from Rhodopseudomonas gelatinosa and Klebsiella aerogenes reveal two characteristic molecular forms. The "basket" form and the "star" form were subjected to two-dimensional image reconstruction using a technique involving averaging of superposed single molecular images after rotational correlation. A three-dimensional image reconstruction shows that the images of these forms can be interconverted by rotation and that they therefore represent different views of the same structure.

Electron microscopical structure analysis of yeast fatty-acid synthase at low resolution.

From an electron microscopical tilt series of the multi-enzyme complex yeast fatty-acid synthase eight three-dimensional molecular structures were obtained by three-dimensional reconstruction of single molecules. The structures confirm present concepts showing a well defined central wall and a sequential arrangement of protein domains in the form of "arches". Additional structural details as ring-shaped parts in the central walls are recognizable. Because of the flattening and the irregular structural deterioration of the single molecules three-dimensional averaging was only partially successful; however, a satisfactory average from five molecules could be obtained. Attempts to find the symmetry of the subunit arrangement by applying correlation methods and by establishing a novel type of correlation analysis ("correlation tables") did not yield a clear proof. However, several strong indications of D3 symmetry were found.