Strains of “atypical Legionella-like organisms” (ALLO) were studied by deoxyribonucleic acid homology. Strains WIGA (ALLO1) and MI-15 (ALLO2) were previously shown by Garrity et al. (G. M. Garrity, A. Brown, and R. M. Vickers, Int. J. Syst. Bacteriol. 30:609-614, 1980) to be related, and the name Fluoribacter bozemanae was proposed for them. We now show that strains NY-23 (ALLO4) and TEX-KL are also closely related genetically. Studies with strain LS-13 demonstrate a lower, but significant, degree of homology of LS-13 with strains WIGA, MI-15, and NY-23, justifying inclusion of all of these strains in a single genus, distinct from Legionella. It is proposed, therefore, that because of phenotypic similarities and genetic relatedness, strains NY-23, TEX-KL, and LS-13 be classified as members of the genus Fluoribacter, which also includes strains WIGA and MI-15. Because NY-23 and LS-13 are the type strains of Legionella dumoffii and L. gormanii, respectively, we propose the transfer of these species to the genus Fluoribacter as F. dumoffii (Brenner et al.) comb. nov. and F. gormanii (Morris et al.) comb. nov.
Deoxyribonucleic acid (DNA)-DNA hybridization and numerical analysis of electrophoretic protein patterns were carried out to determine the taxonomic position of 37 “Klebsiella pneumoniae-like” strains from nonclinical origin (soil and water) and previously studied by numerical taxonomy (group L of Gavini et al., Ann. Microbiol. [Inst. Pasteur] 128B:45-59, 1977). The DNA interrelatedness for 20 selected strains of group L was at least 87%, except for two strains (63 and 69%). Protein electrophoretograms indicated likewise that at least 85% of the L strains constitute a genetically homogeneous group. The latter technique enabled the distinction of strains of group L from those of K. pneumoniae, K. oxytoca, and an unnamed Klebsiella cluster (group K) related to K. pneumoniae and K. oxytoca. On the basis of their phenotypic characters, protein electrophoretic patterns, and DNA-DNA relatedness, we propose for these strains the name Klebsiella terrigena sp. nov., with strain CIP 80-07 (= CUETM 77-176) as the type strain. The DNA of this strain was found to be 49 to 64% homologous to K. oxytoca DNA, 48 to 63% to K. pneumoniae DNA, and 51% to K. ozaenae DNA.
Eleven Ureaplasma isolates from cattle, chosen to represent the serological diversity of these strains, were compared with the eight serovars (serotypes) of Ureaplasma urealyticum (ureaplasmas from humans) by sodium dodecyl sulfatepolyacrylamide gel electrophoresis of polypeptides labeled by growing organisms in the presence of [35S]methionine. The bovine isolates had many common polypeptides, as did the human isolates. Furthermore, the bovine isolates could be distinguished from the human ones. Within the bovine strains, there appeared to be a relationship between the polypeptides present and the serological structure. The results further support the view that ureaplasmas isolated from cattle should be regarded as a species different from U. urealyticum.
The deoxyribonucleic acid homologies among 33 strains of coryneform bacteria with various cell wall types reflected the previous grouping of these organisms by Yamada and Komagata on the bases of the mode of cell division, principal amino acids in the cell wall, and the base composition of deoxyribonucleic acid. In addition, the following four homology clusters were found: (i) six strains of corynebacteria with a glycolyl type of cell wall, (ii) four strains of glutamic acid-producing bacteria, (iii) two strains of Brevibacterium linens, and (iv) three strains of coryneform bacteria containing LL-diaminopimelic acid in the cell wall.
I assessed the genetic relatedness of 45 strains of heterocyst-forming cyanobacteria assigned to eight genera by Rippka et al. and of 19 undescribed strains of the same group by in vitro reassociation of radioiodinated deoxyribonucleic acids. The members of the genera Nodularia, Cylindrospermum, Chlorogloeopsis, and Fischerella formed discrete clusters (intrageneric values of relative binding, more than 55%) and showed intergeneric relatedness of less than 40%, results consistent with the classification proposed by Rippka et al. The genus Nostoc was heterogeneous; four strains previously assigned to Anabaena appeared to belong to Nostoc. The genus Calothrix comprised four clusters with various degrees of internal homogeneity and two strains which showed low relatedness to any others. The general relatedness (i.e., relative binding) of heterocyst formers to various non-heterocystous cyanobacteria belonging to groups II and III (both unicellular and filamentous) was on the order of 10 to 20%.
An immunological study of nine strains of pure-cultured members of the order Actinomycetales isolated from symbiotic root nodules was undertaken to establish the taxonomic relationships of these organisms. Based on the data from a simple double-diffusion (Ouchterlony) precipitation procedure, these isolates, which tentatively have been classified in the genus Frankia (family Frankiaceae), could be divided into two major groups. The first group, designated serotype I, included all of the Actinomycetales strains isolated from Alnus, Comptonia, and Myrica host plants. The second group, serotype II, included Actinomycetales strains isolated from Elaeagnus host plants. No ambiguous cross-reactions with unrelated members of the Actinomycetales were observed. Crushed nodule suspensions cross-reacted in the same manner as pure-cultured frankiae, suggesting that the technique could be important in screening for novel, naturally occurring strains. A comparative study in which an immunofluorescence procedure was used failed to show reliable correlations among the Actinomycetales strains tested. We concluded that the immunodiffusion technique was the better of the two methods for Frankia strain characterization.
Plant specificities and deoxyribonucleic acid homologies were studied among 122 strains of Rhizobium. Some strains were assigned to species on the basis of their source of isolation and present nodulation capabilities, but many did not fit into one of the six currently recognized species of the genus Rhizobium. Among those strains assigned to species were many which also nodulated plants outside their species-specific, cross-inoculation group. Conversely, isolates from a wide variety of plants could be designated Rhizobium phaseoli since they were capable of nodulating Phaseolus vulgaris. Acid production and growth rate on yeastmannitol agar were tested for all strains. Some strains grew rapidly but did not produce an acid reaction; these were grouped with the fast growing acid producers. Deoxyribonucleic acid homology was used to identify four genetic groups of fastgrowing, acid-producing rhizobia. Group 1 included strains of Rhizobium trifolii (except strains obtained from Trifolium lupinaster, Rhizobium leguminosarum, Rhizobium phaseoli (obtained from Phaseolus vulgaris), and two strains obtained from Neptunia gracilis. Group 2 comprised six American strains obtained from crown vetch (Coronilla varia), sainfoin (Onobrychis vicifolia), and Sophora spp. Species status for this group should remain tentative until further strains have been studied. Group 3 corresponded with Rhizobium meliloti as presently defined. Group 4 included fast-growing Lotus rhizobia, two strains obtained from T. lupinaster, and a wide variety of previously unclassified strains. Nine fastgrowing strains could not be included in any of these groups. The nine slowgrowing, non-acid producing strains included in this study showed < 10% homology with DNAs from seven fast-growing reference strains. The relationships between subgroups in group 1 are discussed, and the genetic diversity of strains obtained from Phaseolus vulgaris is examined. It is proposed that fast-growing rhizobia comprise at least four species corresponding with the four genetic groups described.
Deoxyribonucleotide sequence relationships among currently recognized Bordetella species (Bordetella pertussis, Bordetella parapertussis, and Bordetella bronchiseptica) were examined by deoxyribonucleic acid (DNA) hybridization involving the hydroxyapatite batch procedures of Brenner and co-workers. The results indicated that strains from all species tested were highly related. At the stringent criterion (80°C), the relative binding of B. pertussis DNA to B. parapertussis DNA was 75 ± 9%, and to B. bronchiseptica DNA it was 73 ± 8%. Intraspecies binding was 93 ± 8%. Under similar conditions, the relative binding of B. parapertussis DNA to B. bronchiseptica DNA was 85 ± 9%. The various so-called Bordetella species may be reconsidered as representing different subspecies belonging to a single species.
Three strains of a gram-negative coccus which exhibits twitching motility and which grows in a monolayer of cells on non-lipid-containing media were isolated from freshwater. A new genus, Agitococcus, with a single species, A. lubricus sp. nov., is proposed for these strains. The type strain of this species is UQM 1981.
Strains of anaerobic, sporeforming bacteria were isolated by using adenine as a carbon and energy source. Strain WA-1 could utilize all naturally occurring purines, as well as such products of purine degradation as glycine and some of its derivatives, as substrates. The products formed were acetate, formate, carbon dioxide, and ammonia. The organism depended strictly on the availability of selenium compounds for growth. Selenite and molybdate supplementation of the medium promoted the formation of active formate dehydrogenase and xanthine dehydrogenase. The molar growth yield on adenine or hypoxanthine was 10.0 g of dry weight per mol of substrate. The organism's doubling time was 80 min when the strain was grown at its optimum temperature, 36°C. The organism had a guanine plus cytosine content of 29 mol% (by the thermal denaturation method). It shared similarities with Clostridium acidiurici and “C. cylindrosporum” (not on the Approved Lists of Bacterial Names), but it could be differentiated from these by deoxyribonucleic acid-deoxyribonucleic acid homology. Therefore, it is described as a new species, Clostridium purinolyticum. The ability of the type strain, WA-1 (DSM 1384), to grow on adenine and glycine was the most significant difference between it and the two above-mentioned species.
The name Pseudomonas maltophilia Hugh and Ryschenkow 1961 is not on the Approved Lists of Bacterial Names. The taxon to which this name was applied is a distinct entity, and it can be differentiated from the other named species of Pseudomonas. Consequently, the name Pseudomonas maltophilia is revived for the same organism to which the name originally referred. The type strain of P. maltophilia is ATCC 13637 (= 810-2 = RH 1168).