- Volume 136, Issue 4, 1990

Monoclonal antibodies (mAbs) were raised against Penicillium islandicum, a fungus commonly isolated from stored rice grains in South-East Asia. Mice were immunized by a direct, simple method; fresh cell-free surface washings from a solid agar slant culture were injected directly into the peritoneum without prior concentration. Hybridoma supernatants were screened by ELISA. Most of the mAbs raised cross-reacted with other storage fungi and/or uninfected rice grains but three were species-specific. One of these, PI01, was used to develop ELISA and DIP-STICK assays for the detection of P. islandicum in individual grains. All inoculated grains and approximately 90% of grains in natural infected samples from Indonesia tested positively, by ELISA, for P. islandicum. This result and those obtained for discoloured grains from both Indonesia and the Philippines, 32% and 14% respectively, are higher than those obtained by direct plating of surface-sterilized grains. Heat and periodate treatment of the PI01 antigen and binding on Western blots indicate that it is a glycoprotein of M r > 90000. Hyphae of all ages stained uniformly by immunofluorescence using the PI01 antibody but mature conidia stained only weakly.

A numerical taxonomic analysis was performed on 79 phenotypic characters of 147 imperfect yeast species currently assigned to the genus Candida. The characters used were drawn from two monographs on yeast taxonomy. The analysis revealed 10 clusters of three or more species that were similar at the level of 75% or more, and seven clusters containing only one or two species. None of the 10 major clusters contained exclusively species that were traditionally assigned to the genus Torulopsis, while the 12 Candida species of basidiomycetous affinity fell into three clusters with only one species of ascomycetous affinity included. Statistical determination of the five most important differential characters for each cluster failed to show the property of pseudomycelium/mycelium formation as significant for any cluster. The study provides no evidence to support a distinction between taxa that were formerly divided between the genera Candida and Torulopsis and supports previous proposals that these genera should be fused.

A procedure was developed for purifying both the benzyl alcohol dehydrogenase and the benzaldehyde dehydrogenase encoded by the TOL plasmid pWW53 from a single batch of Pseudomonas putida MT53. The procedure involved disruption of the bacteria in the French pressure cell and preparation of a high-speed supernate, followed by chromatography on DEAE-Sephacel, Matrex Gel Red A and Blue Sepharose CL-6B which separated the two enzymes, Phenyl Sepharose CL-4B and Matrex Gel Green A. The final preparations gave single bands on electrophoresis under denaturing and non-denaturing conditions. The subunit Mr values of benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase are 43000 and 56300 respectively. Cross-linking studies with dimethylsuberimidate indicate that both enzymes are probably tetramers, although they run anomalously through gel-filtration columns. The benzyl alcohol dehydrogenase was fairly specific for NAD+ as cofactor but the benzaldehyde dehydrogenase had appreciable activity with NADP+ as well as with NAD+. The optimum pH values are 9·4 and 9·3 for benzyl alcohol dehydrogenase and benzaldehyde dehydrogenase respectively. Benzaldehyde dehydrogenase appears to require a monovalent cation for maximum activity. The apparent K m and maximum velocity values of the two plasmid-encoded dehydrogenases and of the chromosomally encoded benzyl alcohol dehydrogenase and benzaldehyde dehydrogenases I and II from Acinetobacter calcoaceticus were determined for NAD+ and for the unsubstituted substrates and for the monomethyl ring-substituted analogues. The corresponding apparent specificity constants were then calculated. All three benzaldehyde dehydrogenases had very similar substrate profiles, as did the two benzyl alcohol dehydrogenases. The plasmid-encoded benzaldehyde dehydrogenase resembles benzaldehyde dehydrogenase I from A. calcoaceticus (which also requires monovalent cations for activity) in being much more heat-stable than benzaldehyde dehydrogenase II. Overall, the plasmid-encoded benzyl alcohol dehydrogenase from P. putida appears to be remarkably similar to the chromosomally encoded benzyl alcohol dehydrogenase from A. calcoaceticus, and the plasmid-encoded benzaldehyde dehydrogenase is very similar to the two chromosomally encoded benzaldehyde dehydrogenases, particularly benzaldehyde dehydrogenase I.

A restriction endonuclease map was derived for the aromatic amine and m-toluate catabolic plasmid pTDN1 present in Pseudomonas putida UCC22, a derivative of P. putida mt-2. The plasmid is 79±1 kbp in size and can be divided into a restriction-site-deficient region of 51±1 kbp and a restriction-site-profuse region of 28 kbp which begins and ends with directly repeating sequences of at least 2 kbp in length. A mutant plasmid isolated after growth of the host on benzoate had lost the restriction-profuse region by a straightforward recombinational loss retaining one copy of the direct repeat. Analysis of clones, deletion and Tn5 insertion mutants strongly suggested that the meta-cleavage pathway of pTDN1 was situated in the region readily deleted. The catechol 2,3-dioxygenase (C23O) gene of pTDN1 showed no hybridization or restriction homology to previously described C230 genes of TOL plasmids pWW0 and pWW15. In addition, there was little homology between intact pTDN1, pWW0 and pWW15, suggesting the presence of a unique meta-cleavage pathway. We also demonstrated that pTDN1 did not originate from P. putida mt-2 chromosome.

The ability to degrade aromatic amines and m-toluate (Tdn+ phenotype), encoded by plasmid pTDN1, was lost from Pseudomonas putida hosts after subculture in benzoate, succinate, acetate and glucose minimal medium, the fastest rate of loss occurring where benzoate was the substrate. Tdn− cells had either lost the entire pTDN1 plasmid or suffered a recombinational deletion of a specific 26 kbp region. Proportional increase of Tdn− cells resulted from their growth-rate advantage, and additionally, where benzoate was the substrate, from its metabolism via the chromosomal ortho-cleavage pathway incorporating a short lag phase. The ratio of whole plasmid loss to deletion was substrate and pH dependent. Deletion of catabolic genes was not required for loss of pTDN1 but by comparison was a prerequisite for loss of TOL plasmid pWW0. It appeared that m-toluate and benzoate were channelled via chromosomally encoded benzoate oxygenase and dihydroxycyclohexadiene carboxylate dehydrogenase prior to pTDN1 encoded meta-cleavage.