Graham Kemp

Identification of the enzyme required for activation of the small ubiquitin-like protein SUMO-1.

The ubiquitin-like protein SUMO-1 is conjugated to a variety of proteins including Ran GTPase-activating protein 1 (RanGAP1), IκBα, and PML. SUMO-1-modified proteins display altered subcellular targeting and/or stability. We have purified the SUMO-1-activating enzyme from human cells and shown that it contains two subunits of 38 and 72 kDa. Isolation of cDNAs for each subunit indicates that they are homologous to ubiquitin-activating enzymes and to the Saccharomyces cerevisiae enzymes responsible for conjugation of Smt3p and Rub-1p. In vitro, recombinant SAE1/SAE2 (SUMO-1-activating enzyme) was capable of catalyzing the ATP-dependent formation of a thioester linkage between SUMO-1 and SAE2. The addition of the SUMO-1-conjugating enzyme Ubch9 resulted in efficient transfer of the thioester-linked SUMO-1 from SAE2 to Ubch9. In the presence of SAE1/SAE2, Ubch9, and ATP, SUMO-1 was efficiently conjugated to the protein substrate IκBα. As SAE1/SAE2, Ubch9, SUMO-1, and IκBα are all homogeneous, recombinant proteins, it appears that SUMO-1 conjugation of IκBα in vitrodoes not require the equivalent of an E3 ubiquitin protein ligase activity.

Crustins: enigmatic WAP domain-containing antibacterial proteins from crustaceans

Crustins are antibacterial proteins of ca. 7-14 kDa with a characteristic four-disulphide core-containing whey acidic protein (WAP) domain, expressed by the circulating haemocytes of crustaceans. Over 50 crustin sequences have been now reported from a variety of decapods, including crabs, lobsters, shrimp and crayfish. Three main types seem to occur but all possess a signal sequence at the amino terminus and a WAP domain at the carboxyl end. Differences between types lie in the structure of the central region. Those crustins purified as the native protein or expressed recombinantly all kill Gram-positive bacteria, and gene studies have shown that they are constitutively expressed, often at high levels, but show no consistent patterns of change in expression following injection of bacteria. This variable response to infection is enigmatic but indicates that these proteins could perform additional functions, perhaps as immune regulators in recovery from wounding, trauma or physiological stress.

The adenovirus protease is activated by a virus-coded disulphide-linked peptide

In common with many other viruses, adenoviruses code for a protease essential for the development of infectivity. Recombinant adenovirus protease was active in crude in vitro complementation assays but was inactive with peptide or purified protein substrates. Activity was reconstituted by a component of adenovirus virions, which was identified as GVQSLKRRRCF, a peptide derived from the virus protein pVI. Synthetic peptides were used to demonstrate that the cysteine is essential and that the disulphide-linked dimer is required for activity. It is proposed that the adenovirus protease is a cysteine protease and that its activation by the peptide involves thiol-disulphide interchange, which serves to expose the active site cysteine. This represents a novel strategy for controlling the activity of a protease that is required for virus maturation.

Isolation and characterisation of oncorhyncin II, a histone H1-derived antimicrobial peptide from skin secretions of rainbow trout, Oncorhynchus mykiss.

A potent antimicrobial peptide, tentatively named oncorhyncin II, was isolated from an acid extract of rainbow trout skin secretions. Amino acid sequencing showed that the first 17 residues of oncorhyncin II are identical to residues 138-154 of histone H1 from rainbow trout. Matrix-assisted laser desorption ionization mass spectrometry revealed that the purified peptide has a molecular mass of 7195.3Da. Taken together, these data indicate that oncorhyncin II is a 69-residue C-terminal fragment of histone H1, probably phosphorylated at two residues. Oncorhyncin II has minimal inhibitory concentrations in the submicromolar range against Gram-(+) as well as Gram-(-) bacteria and it does not display significant haemolytic activity towards trout erythrocytes. The purified peptide was found to induce a marked destabilisation of planar lipid bilayers without the formation of stable ion channels. Oncorhyncin II is possibly a cleavage product of histone H1 with a potentially important role in mucosal defence of rainbow trout.

Anti-microbial properties of histone H2A from skin secretions of rainbow trout, Oncorhynchus mykiss

Skin exudates of rainbow trout contain a potent 13.6 kDa anti-microbial protein which, from partial internal amino acid sequencing, peptide mass fingerprinting, matrix-associated laser desorption/ionization MS and amino acid analysis, seems to be histone H2A, acetylated at the N-terminus. The protein, purified to homogeneity by ion-exchange and reversed-phase chromatography, exhibits powerful anti-bacterial activity against Gram-positive bacteria, with minimal inhibitory concentrations in the submicromolar range. Kinetic analysis revealed that at a concentration of 0.3 microM all test bacteria lose viability after 30 min incubation. Weaker activity is also displayed against the yeast Saccharomyces cerevisiae. The protein is salt-sensitive and has no haemolytic activity towards trout erythrocytes at concentrations below 0.3 microM. Reconstitution of the protein in a planar lipid bilayer strongly disturbs the membrane but does not form stable ion channels, indicating that its anti-bacterial activity is probably not due to pore-forming properties. This is the first report to show that, in addition to its classical function in the cell, histone H2A has extremely strong anti-microbial properties and could therefore help contribute to protection against bacterial invasion.

Characterization of the Adenovirus Proteinase: Substrate Specificity

Peptides were synthesized based on the cleavage sites in the adenovirus type 2 proteins pVI and pVII. The synthetic peptides were incubated with disrupted, purified adenovirus as a source of proteinase and specific cleavages were monitored by fast protein liquid chromatography and amino acid analysis. Using this approach it was established that all the peptides cleaved were of the form M(L)XGX decreases G or M(L)XGG decreases X. Thus we have shown that the adenoviral proteinase recognizes a specific secondary structure formed by a sequence of at least five amino acids, the main determinants of specificity being two and four residues to the N-terminal side of the bond cleaved. We were able to examine the relevant structural features of the peptide substrates by utilizing the CHEM-X molecular modelling package. Using our consensus sequence we were able to predict the cleavage sites in the viral proteins pVIII, pre-terminal protein (pTP), 11K and IIIa. Octapeptides containing the predicted sites in pVIII and the pTP were synthesized and shown to be cleaved by the proteinase.

Chemical mechanisms underlying the vasodilator and platelet anti-aggregating properties of S-nitroso-N-acetyl-DL-penicillamine and S-nitrosoglutathione

The chemistries of S-nitroso-DL-penicillamine (SNAP) and S-nitrosoglutathione (GSNO) in relation to their ability to relax vascular smooth muscle and prevent platelet aggregation have been investigated. Metal ion catalysis greatly accelerates the decomposition of SNAP, but has little effect on GSNO. Instead, NO release from GSNO is effected either by NO transfer to a free thiol (e.g. cysteine), or by enzymatic cleavage of the glutamyl-cystyl peptide bond. In both cases the resulting nitrosothiol (i.e. S-nitrosocysteine and S-nitrosocystylglycine, respectively) is susceptible to metal ion catalysed NO release. We conclude that transnitrosation or enzymatic cleavage are obligatory steps in the mechanism of NO release from GSNO, whereas SNAP needs only the presence of metal ions to effect this process. The different modes of NO production may go some way towards explaining the different physiological effectiveness of these S-nitrosothiols as vasodilators and inhibitors of platelet aggregation.

Oncorhyncin III: a potent antimicrobial peptide derived from the non-histone chromosomal protein H6 of rainbow trout, Oncorhynchus mykiss

The partial N-terminal amino acid sequence of the antimicrobial peptide reported in the present paper has been submitted to the TrEMBL database under the accession number P83338. A 6.7 kDa antimicrobial peptide was isolated from trout skin secretions using acid extraction followed by cation-exchange chromatography, (t)C(18) solid-phase extraction, and C(18) reversed-phase HPLC. The molecular mass of this peptide, which is tentatively named oncorhyncin III, is 6671 Da, as determined by matrix-assisted laser-desorption ionization MS. N-terminal amino acid sequencing revealed that the first 13 residues of oncorhyncin III are identical with those of the non-histone chromosomal protein H6 from rainbow trout. Hence these data combined with the MS results indicate that oncorhyncin III is likely to be a cleavage product of the non-histone chromosomal protein H6 (residues 1-66) and that it probably contains two methylated residues or one double methylation. The purified peptide exhibits potent antibacterial activity against both Gram-positive and Gram-negative bacteria, with minimal inhibitory concentrations in the submicromolar range. The peptide is sensitive to NaCl, and displays no haemolytic activity towards trout erythrocytes at concentrations below 1 microM. Scanning electron microscopy revealed that oncorhyncin III does not cause direct disruption of bacterial cells. Reconstitution of the peptide in planar lipid bilayers strongly disturbs the membranes, but does not induce the formation of stable ion channels. Taken together, these results support the hypothesis that oncorhyncin III plays a role in mucosal innate host defence.

Interaction between hnRNPA1 and IκBα Is Required for Maximal Activation of NF-κB-Dependent Transcription

ABSTRACT Transcriptional activation of NF-κB is mediated by signal-induced phosphorylation and degradation of its inhibitor, IκBα. NF-κB activation induces a rapid resynthesis of IκBα which is responsible for postinduction repression of transcription. Following resynthesis, IκBα translocates to the nucleus, removes template bound NF-κB, and exports NF-κB to the cytoplasm in a transcriptionally inactive form. Here we demonstrate that IκBα interacts directly with another nucleocytoplasmic shuttling protein, hnRNPA1, both in vivo and in vitro. This interaction requires one of the N-terminal RNA binding domains of hnRNPA1 and the C-terminal region of IκBα. Cells lacking hnRNPA1 are defective in NF-κB-dependent transcriptional activation, but the defect in these cells is complemented by ectopic expression of hnRNPA1. hnRNPA1 expression in these cells increased the amount of IκBα degradation, compared to that of the control cells, in response to activation by Epstein-Barr virus latent membrane protein 1. Thus in addition to regulating mRNA processing and transport, hnRNPA1 also contributes to the control of NF-κB-dependent transcription.

Characterization of the Adenovirus Proteinase: Development and Use of a Specific Peptide Assay

An assay system has been developed for the adenovirus endoproteinase which utilizes the synthetic peptide MSGGAFSW, derived from the cleavage site of the adenovirus type 2 (Ad2) protein pVI. MSGGAFSW was shown to be cleaved at the G-A bond when incubated with a source of Ad2 proteinase. Digestions were readily monitored by either fast protein liquid chromatography or thin layer electrophoresis, enabling the rapid production of quantitative data. Comparison of the peptide assay with a previously described [35S]methionine assay system showed it to be faster, cleaner and less prone to extreme conditions of pH and ionic strength. The effect on adenovirus proteinase activity of a number of inhibitors was assessed using both the [35S]methionine and peptide assays. Identical inhibitor profiles were obtained and these suggested that the adenovirus enzyme is a cysteine proteinase. The 23K gene product, thought to be the proteinase, contains cysteine and histidine residues at positions 122 and 54, respectively, that could constitute part of the active site of a cysteine proteinase. These amino acids and their surrounding residues are conserved in all serotypes examined and appear to bear some resemblance to those in the putative active sites of the 3C proteinases in picornaviruses.