Neil Parkinson

Phylogenetic structure of Xanthomonas determined by comparison of gyrB sequences

Previously, we have produced a phylogeny of species type strains from the plant-pathogenic genus Xanthomonas based on gyrB sequences. To evaluate this locus further for species and infraspecies identification, we sequenced an additional 203 strains comprising all the pathovar reference strains (which have defined plant hosts), 67 poorly characterized pathovars, currently classified as Xanthomonas campestris, and 59 unidentified xanthomonads. The well-characterized pathovars grouped either in clades containing their respective species type strain or in clades containing species related to Xanthomonas axonopodis. The Xanthomonas euvesicatoria, Xanthomonas perforans and Xanthomonas alfalfae species complex, Xanthomonas fuscans and Xanthomonas citri were discriminated as X. axonopodis-related clades and comprised a large proportion of unidentified strains as well as 80 pathovars representing all the X. axonopodis pathovars and many poorly characterized pathovars, greatly increasing the plant host ranges of the constituent species. Most xanthomonads from these three large clades were isolated from a taxonomically diverse range of plant hosts, including many weed species, from field systems in India, suggesting that these lineages became established and diversified in agricultural areas in this region. The majority of these xanthomonads had minimal sequence diversity, consistent with rapid and highly extensive pathovar diversification that has occurred in relatively recent times. Low-intensity farming practices may have provided conditions conducive to pathovar development, and evidence for pathovar diversification within other regional angiosperm floras is discussed. The gyrB locus was sufficiently discriminating to identify diversity within many species. Seven branches or clades were sufficiently distinct to be considered as potential novel species. This study has provided a comprehensive xanthomonad classification framework and has firmly established gyrB sequencing as a rapid and efficient identification tool.

A new form of arthropod phenoloxidase is abundant in venom of the parasitoid wasp Pimpla hypochondriaca.

We have recently identified phenoloxidase (PO) activity among several biologically active factors in venom from the parasitoid wasp Pimpla hypochondriaca. We have now isolated three genes, designated POI, POII and POIII, from a cDNA library made from venom-producing glands and found that their products are related to pro-phenoloxidases (PPOs), which are expressed as proenzymes in haemocytes and which mediate immune processes in arthropods. This is the first report of PO as a venom constituent. Amino acid sequence comparisons between the three Pimpla POs and PPOs revealed several notable differences, including the absence of sequences which specify the site of proteolytic activation in insect PPOs and the unprecedented occurrence of signal peptide sequences. NH(2)-terminal amino acid analysis of PO purified from venom yielded a peptide sequence matching the predicted mature NH(2) termini of POI and POII, confirming the authenticity of the signal peptide and indicating that proteolytic processing, other than to remove the signal peptide, does not occur in the wasp. Expression of POI, analysed by Northern hybridization, was approximately uniform from the time of adult emergence to day 6 post-emergence, after which it declined. A novel means of host immune suppression, mediated by the unregulated activity of venom PO in the haemocoel, is proposed.

cDNAs encoding large venom proteins from the parasitoid wasp Pimpla hypochondriaca identified by random sequence analysis.

Venom from the parasitoid wasp Pimpla hypochondriaca contains numerous proteins, has potent in vitro anti-haemocytic properties, and disrupts host encapsulation responses. By sequencing 500 cDNAs randomly isolated from a venom gland library, we have identified 60 clones that encode proteins containing potential secretory signal sequences. To identify cDNAs encoding particular venom proteins, N-terminal amino acid sequences were determined for large (>30 kDa) venom proteins that had been separated using a combination of gel filtration and SDS-PAGE. We describe five of these cDNAs, which encoded residues that matched with the N-terminal sequences of previously undescribed venom proteins. cDNAs vpr1 and vpr3 encoded related proteins of approximately 32 kDa that were found in widely different fractions of gel filtration-separated venom. Neither vpr1 nor vpr3 were closely related to any other protein in the GenBank database, suggesting that they are highly specialised venom components. vpr2 encoded a 57-kDa polypeptide that was similar to a Drosophila protein, of unknown function, which lacks a signal sequence. A fourth clone, tre1, encoded a 61-kDa protein with extensive sequence similarity to trehalases. The 76-kDa sequence encoded by lac1 contained three regions which were very similar to histidine-rich copper-binding motifs, and could be aligned with the laccase from the fungus Coprinus cinereus. This study represents a significant step towards a holistic view of the molecular composition of a parasitoid wasp venom.

Dickeya species relatedness and clade structure determined by comparison of recA sequences

Using sequences from the recA locus, we have produced a phylogeny of 188 Dickeya strains from culture collections and identified species relatedness and subspecies clade structure within the genus. Of the six recognized species, Dickeya paradisiaca, D. chrysanthemi and D. zeae were discriminated with long branch lengths. The clade containing the D. paradisiaca type strain included just one additional strain, isolated from banana in Colombia. Strains isolated from Chrysanthemum and Parthenium species made up most of the clade containing the D. chrysanthemi type strain, and the host range of this species was extended to include potato. The D. zeae clade had the largest number of sequevars and branched into two major sister clades that contained all of the Zea mays isolates, and were identified as phylotypes PI and PII. The host range was increased from six to 13 species, including potato. The recA sequence of an Australian sugar-cane strain was sufficiently distinct to rank as a new species-level branch. In contrast to these species, Dickeya dadantii, D. dianthicola and D. dieffenbachiae were distinguished with shorter branch lengths, indicating relatively closer relatedness. The recA sequence for the type strain of D. dadantii clustered separately from other strains of the species. However, sequence comparison of three additional loci revealed that the D. dadantii type strain grouped together with the six other D. dadantii strains that were sequenced. Analysis of all four loci indicated that the D. dadantii strains were most closely related to D. dieffenbachiae. Three further branches (DUC-1, -2 and -3) were associated with these three species, which all diverged from a common origin and can be considered as a species complex. The large clade containing the D. dianthicola type strain comprised 58 strains and had little sequence diversity. One sequevar accounted for the majority of these strains, which were isolated nearly exclusively from eight hosts from Europe. Isolation of this sequevar on multiple occasions from Dianthus and (more recently) potato demonstrates that this lineage has become established in these species. The D. dadantii clade comprised 11 sequevars, and the known host range of the species was extended from eight to 19 species. New hosts included several ornamental species and potato. The clade DUC-1 was made up exclusively of potato strains originating from Europe, which had identical sequences, whilst DUC-2 strains were isolated mostly from a variety of monocotyledonous species. A single strain from Aglaonema sp. made up DUC-3. A single sequevar constituted the D. dieffenbachiae clade. The phylogenetic method described will provide a simple means for identification to the species and intraspecies level, which will support efforts to control these pathogens based on monitoring and surveillance.

Purification of pimplin, a paralytic heterodimeric polypeptide from venom of the parasitoid wasp Pimpla hypochondriaca, and cloning of the cDNA encoding one of the subunits.

We have previously detected a paralytic factor in gel filtration-separated venom from the endoparasitoid wasp Pimpla hypochondriaca which is active against the fly Musca domestica. Now we have further purified this factor, which we have called pimplin, by reverse phase chromatography, and established using SDS-PAGE that it has a molecular mass of approximately 22 kDa. A 40 ng dose of pimplin administered to adult M. domestica by intrahaemocoelic injection was sufficient to kill all flies tested. Treatment of pimplin with beta-mercaptoethanol prior to SDS-PAGE analysis resulted in the appearance of two polypeptides of approximately 15 and 6 kDa, indicating that pimplin is a heterodimer whose polypeptides are linked through a disulphide bond. Subunit masses of 10.544 and 6.318 kDa were determined using MALDI-TOF analysis indicating that the larger subunit migrates anomalously in SDS-PAGE. Using an oligonucleotide probe designed from N-terminal sequence obtained for the 15 kDa polypeptide, we have isolated a cDNA (pim1) encoding this larger pimplin subunit. The N-terminal amino acid sequence of pim1 occurred 28 residues beyond a predicted signal peptide cleavage site, indicating that pim1 is synthesised as a pre-propolypeptide which is secreted and proteolytically cleaved to yield the mature polypeptide stored in the venom sac. Beginning at the fourth residue of the mature pim1 venom polypeptide is a stretch of 46 residues consisting of alternating prolines, the significance of which is discussed in terms of possible host processing.

Proposal of Xanthomonas translucens pv. pistaciae pv. nov., pathogenic to pistachio (Pistacia vera)

Strains of Xanthomonas translucens have caused dieback in the Australian pistachio industry for the last 15 years. Such pathogenicity to a dicotyledonous woody host contrasts with that of other pathovars of X. translucens, which are characterized by their pathogenicity to monocotyledonous plant families. Further investigations, using DNA-DNA hybridization, gyrB gene sequencing and integron screening, were conducted to confirm the taxonomic status of the X. translucens pathogenic to pistachio. DNA-DNA hybridization provided a clear classification, at the species level, of the pistachio pathogen as a X. translucens. In the gyrB-based phylogeny, strains of the pistachio pathogen clustered among the X. translucens pathovars as two distinct lineages. Integron screening revealed that the cassette arrays of strains of the pistachio pathogen were different from those of other Xanthomonas species, and again distinguished two groups. Together with previously reported pathogenicity data, these results confirm that the pistachio pathogen is a new pathovar of X. translucens and allow hypotheses about its origin. The proposed name is Xanthomonas translucens pv. pistaciae pv. nov.