No description available.

Diss. (sammanfattning) Uppsala : Uppsala universitet, 2006. / Härtill 5 uppsatser.

Thesis (Ph. D.)--Oregon State University, 2007. / Printout. Includes bibliographical references. Also available on the World Wide Web.

HRV14 3C acts as a protease and has a role in RNA replication in vivo, interacting with a cloverleaf structure in picornaviral genomic RNA. Picornaviral 3C proteases are able to cleave both N- and C-terminally producing 3CDpro, or, 3Cpro and 3DPol, respectively. In order to investigate the mechanisms whereby these alternative processing pathways are adopted an artificial polyprotein system was constructed, composed of viral sequences from the P3 region of the viral polyprotein flanked by reporter genes. Two antibiotic resistance genes (KanR, TetR) were cloned to act as reporter genes flanking the viral region of interest. Analysis of the cleavage products in a coupled TnT system showed whether N- or C-terminal cleavage had occurred. 3Cpro cleaved preferentially at its N-terminus in [KanR3CproTetR] with a lesser degree of cleavage at its C-terminus. When 3ABC was used as the viral component of the system cleavage at the N-terminus of 3Cpro was also observed. The use of 3CDpro as the viral component also had a regulatory effect on the site (N- or C-terminal) of cleavage by 3Cpro. With 3CDpro as the viral component of the artificial reporter polyprotein cleavage occurred at both the N-and C-termini of 3Cpro as well as at the C-terminus of 3Dpol. This surprising result has led to comparisons with the proteolytic action of viral proteases in the caliciviruses and some plant viruses and the proposal of possible evolutionary links between these viruses. The use of the antibiotic resistance genes as reporters allowed investigation into the use of antibiotic resistance phenotypes in E.coli for monitoring cleavage of the artificial polyprotein. Preliminary results indicated that the system may be useful as a genetic screen to quantify large numbers of mutants.

A variety of phosphonamidate-containing peptides were synthesised as potential inhibitors of the HIV-1 protease. These transition state analogues were designed using known sequences from HIV-1 protease substrates and incorporated a unique Phe-Pro scissile bond mimic in an attempt to achieve selectivity over the mammalian aspartic proteases. Such compounds were found to be moderate inhibitors of the HIV-1 protease possessing IC50 values in the 1-100 μM range, both in in vitro and in vivo assays. However, the phosphonamidate methyl ester analogues showed a marked ability to enter cells and this feature was highlighted in the 1:1 ratio of in vivo/ in vitro IC50 values (generally for peptidic inhibitors, this ratio is 10-10000 fold higher, indicating poor cell uptake properties). Optimisation of the methyl ester analogues was attempted by alteration of the binding residues flanking either side of the phosphonamidate moiety. However, such alterations had only a small effect on inhibitor potency and the trends observed for the more potent hydroxyl-containing inhibitors were not seen with our compounds. These results suggest hydrogen-bond donating capacity as a key requirement for potent inhibition of the HIV-1 protease, a feature which the methyl ester analogues lack. Due to the associated problems with peptidic inhibitors, the development towards two novel non-peptidic inhibitors of the HIV-1 protease was undertaken. Both cyclic inhibitors were designed to optimise the key interactions at the core of the active site of the enzyme in an attempt for selective, highly potent, low molecular weight inhibitors. Such inhibitors were designed to replace a structural water molecule in the flap region of the enzyme, whilst maintaining the key interactions with the catalytic aspartates. For this purpose cyclic compounds possessing both alcohol functionality and ring heteroatoms were synthesised. The first sulfur-containing analogue produced a moderate inhibitor of the HIV-1 protease and provided a lead compound for further development. The second seven membered ring analogue was designed on the basis of a recent literature compound and the synthesis incorporated a novel bis-ketone, derived from diethyl L-tartrate. This synthesis has yet to be completed and is currently under investigation in our group by Neil Piggot.

Production of mature infectious adenovirus type 2 depends on the action of the L3 coded 23kd protease which is known to cleave 7 virus proteins and the host cell proteins cytokeratin K7 and K18. Previous studies have shown the enzyme to be a cysteine proteinase with a novel mechanism of activation requiring the presence of an 11 residue peptide derived from the C-terminus of virus structural protein pVI. It has been proposed that the protease is activated within the assembled virion, although other evidence suggests that the protease is active outwith the virus particle, and is partly responsible for the degradation of the intermediate filament system prior to virion release from the infected cell. The subcellular localisation of the 23kd protease during productive adenovirus infection was investigated using a panel of monoclonal antibodies generated during the course of this study. The monoclonal antibodies were partially characterised and those of interest were epitope mapped using a combination of techniques which included limited and chemical proteolysis, and deletion mutagenesis of the recombinant protein, and screening against overlapping peptides corresponding to specific regions within the 23kd protease. The viral 23kd protease, capsid protein pVIII, and the Ll-52kd probable scaffold protein were shown to colocalise within virus-induced intranuclear clear amorphous inclusions late in infection (24 h.p.i onwards). These inclusions were typically associated with crystalline arrays of assembled virions and are believed to be the same sites which contain relocated PML during the earlier stages of infection. The stmctural appearance of these inclusions varied depending on the fixation method used. The distribution of the viral protease, PML and another cellular protein P80-coilin during the late-phase of nuclear transformation was investigated with possible cleavage of both cellular proteins also partially determined. The viral protease and protein pVIII were also shown to colocalise within cytoplasmic structures but were not found to be associated with cytokeratin K18. The degradation of cytokeratin K18 occurred as early as 20 to 22 h.p.i although the intranuclear distribution of pVI at this stage of infection suggested that the viral protease may be regulated by an as yet unknown mechanism. Isoforms of recombinant wild type 23kd protease (but not Ad2tsl) were detected in vitro which suggested that dimerisation may be an in vivo regulatory mechanism, possibly involving intranuclear trafficking.

The adenovirus codes for a protease which is essential for virion infectivity. This protease requires the presence of a peptide cofactor in order to develop optimal activity. This peptide, GVQSLBCRRRCF, originates from the C-terminal of a viral protein, pVI, and some evidence regarding its specificity came from observations showing that neither of the peptides GVQSLKRRRAF or KRRRCF was able to activate the protease, indicating that both the cysteine and the N-terminal were important in the activation process. However, the mechanism by which the peptide activates the protease has never been elucidated. In this project, several factors contributing to the activation mechanism of the human adenovirus type 2 protease were studied, such as the peptide N-terminal length and composition, the environment close to the cysteine and the distance between the N-terminal and the cysteine, in view of assessing the relevance of each of these parameters in the activation process and proposing a mechanism of activation. Based on the above studies, attempts of protease inhibition were also performed based on the activation process rather than on the blocking of the active site, and the relevance of these results was related with the proposed activation mechanism. An attempt to clone an avian adenovirus protease was also performed, in order to try and compare the activation processes between the two proteases.

It was confirmed that U. urealyticum produces an IgAl protease, which cleaves human IgAl only into intact Fab and Fcalpha fragments. By N-terminal amino acid sequencing of Fcalpha fragments, the site of digestion was identified as a Pro235-Thr236 peptide bond within the a chain hinge-region of IgAl. A number of assay systems were examined for their ability to detect and estimate IgAl protease activity. A reliable and reproducible immunoblotting method was developed, in conjunction with a quantifiable assay utilising [125I] IgAl. By these methods, IgAl protease activity was identified in fourteen serotypes of U. urealyticum, all of which appeared to digest IgAl at the same Pro235-Thr236 peptide bond. The enzyme was active over a broad range of pH (pH 3-10) and was inhibited by the serine-protease inhibitors 3,4-DCI and DFP. The IgAl protease was not located in 'spent' ureaplasma cultivation medium but appeared to be cell-associated. The activity was solubilised by a number of non-ionic detergents which were required in purification buffers to maintain enzyme stability, further suggesting a membrane-bound location. Although the enzyme was not purified to homogeneity, a number of protocols were established which provide a basis for future work. A genomic library of U. urealyticum DNA was produced and a variety of strategies adopted for identification of the iga gene. Radiolabelled DNA probes were generated from a plasmid containing the iga gene for N. gonorrhoeae (pIP503). By Southern blot hybridisation, no significant homology was identified between the heterologous probes and ureaplasma genomic DNA. Based on regions of high nucleotide conservation between the iga genes from N. gonorrhoeae and H. influenzae, degenerate PCR primers were designed. While amplification products did not appear to contain regions of the iga, such an approach may be adapted and extended for use in future studies.

A series of experiments, designed to investigate two mutually incompatible theories of the catalysis carried out by pepsin, the archetypal aspartic protease, were undertaken. Mechanism-activated active-site probes, based on acyl hydrazides, were synthesised, but could not be shown to inactivate pepsin. Experiments designed to trap a covalent intermediate in pepsin catalysis were also carried out, but did not provide any evidence for such an intermediate. A number of methyl hydrogen 1-aminoalkyl phosphonates have been synthesised. They were used, by co-workers in the group, in the synthesis of phosphonamidate-containing penta- and hexapeptide-based inhibitors for the aspartic protease from HIV-1. These compounds were found to be inhibitors in both in vitro and in vivo assays. The best inhibitors had IC50 values in the low micromolar range. Molecular modelling was used to develop a model for the interaction of these compounds with the active site of the protease. This model was used to rationalise some of the results obtained from the inhibitors. Attempts were made to clone, overexpress and purify the protease from E. coli. The protease was purified to homogeneity but no activity could be observed. Various attempts to obtain activity were unsuccessful.

The substrate specificity, classification, expression and control of the adenovirus encoded protease, the activity of which is required for the production of virions, is described. The use of synthetic peptides has shown that the protease cleaves sites of the form (M,L,I)XGG-X or (M,L,I)XGX-G and these consensus sequences have been used to identify potential cleavage sites in all the known substrates of the protease. Putative cleavage sites have also been found in a number of other adenovirus proteins including the major coat proteins, the hexon and the penton, that had not previously been considered as substrates of the protease. The octapeptide, MSGGAFSW, has been used to develop a specific assay for the protease where digestion is monitored by HPLC reverse phase chromatography. Purification of the protease from adenovirus particles and the preparation of antipeptide sera against the L3 23-kDa protein have been used to confirm that the latter is the protease and that it is not proteolytically activated. Inhibitor studies reveal that the enzyme is inhibited by the thiol directed reagents iodoacetate, PCMB and DTDP, and activated by DTT or cysteine suggesting that it is a member of the cysteine class of proteases. Analysis of the sequence of the enzyme, however, shows that it is not a papain-like enzyme and it is suggested that it might be a member of the recently identified subclass of cysteine proteases that are related to trypsin. The 23-kDa protease has been expressed using E.coli and baculovirus expression systems and a purification schedule for the protein was developed using anion exchange and hydrophobic interaction chromatography. The purified enzyme, however, was not able to cleave the peptide substrate MSGGAFSW and the conclusion was drawn that the protein is probably synthesised as an apoenzyme. One of the substrates of the protease, the pre-terminal (pTP) protein was expressed in insect cells using a recombinant baculovirus. Coinfections of cells with recombinant 23-kDa and pTP baculoviruses resulted in efficient processing of the pTP to the intermediate terminal protein (iTP) in situ. The partially purified baculovirus expressed 23-kDa protein was not, however, found to digest the pTP in vitro, supplying further evidence that the adenovirus protease requires a cofactor.