Emended descriptions of the genera Myxococcus and Corallococcus , typification of the species Myxococcus stipitatus and Myxococcus macrosporus and a

The genus Corallococcus was separated from the genus Myxococcus mainly on the basis of differences in morphology and consistency of swarms and fruiting bodies of the respective members. Phylogenetic, chemotaxonomic and physiological evidence is presented here that underpins the separate status of these phylogenetically neighbouring genera. Emended descriptions of the two genera are presented. The data also suggest that the species Corallococcus macrosporus belongs to the genus Myxococcus. To the best of our knowledge, the type strains of the species Myxococcus macrosporus and Myxococcus stipitatus are not available from any established culture collection or any other source. A Request for an Opinion is made regarding the proposal that strain Cc m8 (5DSM 14697 5CIP 109128) be formally recognized as the neotype strain for the species Myxococcus macrosporus, replacing the designated type strain Windsor M271, and that strain Mx s8 (5DSM 14675 5JCM 12634) be formally recognized as the neotype strain for the species Myxococcus stipitatus, replacing the designated type strain Windsor M78.


Introduction
Within the family Myxococcaceae, delineation of genera, descriptions of species and affiliations of strains to species are based primarily on the morphology of vegetative cells, swarming colonies, fruiting bodies and myxospores.The genus Corallococcus was established for strains that build fruiting bodies with finger-like projections or of even more bizarre, coralloid shape with cartilaginous consistency.In contrast, fruiting bodies of Myxococcus species are larger and globular, form knobs, either constricted at the base or on a stalk, and are of a soft consistency (Reichenbach & Dworkin, 1981).Other properties cited by Reichenbach (2005a) as being useful for the differentiation of the two genera are that Myxococcus strains are generally sensitive to kanamycin, whereas about 60 % of Myxococcus isolates grow in the presence of streptomycin, that Corallococcus strains show opposite reactions towards these antibiotics, that Corallococcus strains produce squalene while Myxococcus strains do not and that Myxococcus strains are more tolerant of NaCl (up to 2 %).
The genus Myxococcus was described by Thaxter (1892), who first recognized that the myxobacteria, formerly assigned to the fungi, were in fact bacteria.The species names Myxococcus disciformis, M. flavescens, M. fulvus, M. stipitatus, M. xanthus, M. virescens, M. macrosporus and M. coralloides have since been validly published.M. disciformis was later reclassified as Angiococcus disciformis (Hook et al., 1980).In 1981, Reichenbach and Dworkin proposed the genus name Corallococcus for the former genus 'Chondrococcus', but the genus name had no standing in nomenclature until recently.In his description of the genus Corallococcus, Reichenbach (2005c) included the taxa M. coralloides, 'Myxococcus exiguus' and 'Chondrococcus macrosporus' as Corallococcus coralloides comb.nov., Corallococcus exiguus nom.rev., comb.nov.and Corallococcus macrosporus comb.nov., respectively.The genus name was accepted as being validly published, but the species name Corallococcus macrosporus was not accepted as being synonymous with Myxococcus macrosporus on formal grounds (Reichenbach, 2007;

see details below).
A small number of taxonomic studies have recently been published that elucidate the phylogenetic status of the genera Myxococcus and Corallococcus.In a broad 16S rRNA gene sequence-based analysis of almost all myxobacterial genera described at that time, Spro ¨er et al. (1999) found members of the genus Corallococcus (genus name not validly published at the time) branching close but adjacent to members of the genus Myxococcus.M. macrosporus strains Cc m7 and Cc m8 (5DSM 14697) did not cluster with other members of the genus Corallococcus but within the radiation of Myxococcus strains.As the 16S rRNA gene sequence similarities among myxococcal strains were higher than 98.2 % and only slightly lower between myxococci and members of the genus Corallococcus (97-98 %), no decision on the taxonomic status of Myxococcus macrosporus was made.
In a more extensive study on the intrageneric structure of the genus Corallococcus (Stackebrandt & Pa ¨uker, 2005;Stackebrandt et al., 2007), 37 strains of Corallococcus, Corallococcus macrosporus DSM 14697 T and M. xanthus DSM 16526 T were analysed with respect to 16S rRNA, gyrB, rpoB, fusA and lepA gene sequence similarities.In each of the dendrograms calculated from these genes, Corallococcus macrosporus DSM 14697 T clustered adjacent to M. xanthus DSM 16526 T but significantly outside the cluster defined by Corallococcus coralloides DSM 2259 T and Corallococcus exiguus DSM 14696 T .This finding was corroborated by significant differences in fatty acid profiles, Riboprint patterns and metabolic reactions (Stackebrandt et al., 2007).It thus appeared that the morphology-based formal decision to classify M. macrosporus as a member of the genus Corallococcus is not supported by the chemotaxonomic, genetic and metabolic properties analysed.
Results of these studies raised afresh the question of whether the separate status of the genera Corallococcus and Myxococcus is justified and how and where to delineate the genus boundaries.The aims of this study were threefold.First, to amend the physiological and chemotaxonomic description of Myxococcus species in order to accumulate evidence as to whether or not the separate status of the two genera Myxococcus and Corallococcus is justified, second, to support the taxonomic position of Corallococcus macrosporus as a member of the genus Myxococcus, and third, to propose neotype strains for the species M. macrosporus and M. stipitatus, for which culturable type strains are not available from any established culture collection or any other source.

Differentiation of the genus Corallococcus from the genus Myxococcus
Genomic properties and the composition of whole-cell fatty acids and physiological properties of several strains of the genus Corallococcus were described previously (Spro ¨er et al., 1999;Stackebrandt & Pa ¨uker, 2005;Stackebrandt et al., 2007).In this study, these investigations were expanded to representatives of different Myxococcus species.Published 16S rRNA gene sequences were aligned by the ae2 editor (Maidak et al., 1997) and by ARB (Ludwig et al., 2004) and dendrograms were generated using neighbour-joining (Saitou & Nei, 1987) and least-squares algorithms (De Soete, 1983) as well as maximumlikelihood analysis (Olsen et al., 1994).Bootstrap analysis was used to evaluate the significance of the tree topology statistically by performing 1000 resamplings (Felsenstein, 1993).The topologies of all dendrograms thus generated were very similar (Fig. 1; least-squares analysis) in that Corallococcus strains and Myxococcus strains formed two well-separated clades supported by bootstrap values of 99 %.The exception was Corallococcus macrosporus DSM 14697 T , which clustered within the Myxococcus clade in all the dendrograms generated, adjacent to the three highly related species M. virescens, M. flavescens and M. xanthus.
Strains from the DSMZ were used as reference strains for physiological tests and fatty acid analysis (Table 1).The cultures were maintained on bakers' yeast agar (VY/2; DSM medium 9).Test media were inoculated with cells suspended in medium MD1 (l 21 : 3 g casitone, 0.7 g CaCl 2 .2H 2 O, 2 g MgSO 4 .7H 2 O) after growth on CY agar (l 21 : 3 g casitone, 1 g CaCl 2 .2H 2 O, 1 g yeast extract, 15 g agar; pH 7.2) for 2 days.Incubations were carried out at 28 u C. Starch and xylan hydrolysis were tested on overlay agar.The first layer consisted of water agar (l 21 : 1 g CaCl 2 .2H 2 O and 15 g agar agar), the second of starch or xylan agar (l 21 : 2 g native potato starch or 5 g oat spelt xylan, 0.5 g casitone, 1 g CaCl 2 .2H 2 O, 1 g MgSO 4 .7H 2 O, 10 g agar).The plates were evaluated after covering with Lugol's solution after 5 days (starch) and 4 days (xylan) of incubation.Clear haloes with a diameter larger than that of the swarms and staying for longer than 5 min were scored as positive.Salt tolerance was measured as growth on CY agar plates containing 15 g NaCl l 21 , incubated for up to 3 weeks.Aesculin hydrolysis was tested on slants with the composition (l 21 ) 3 g casitone, 1 g yeast extract, 1 g aesculin, 1 g sodium citrate and 0.05 g ground ferric citrate.Slants were read after 8 days.Tolerance against antibiotics was determined on VY/2 agar plates containing 250 mg kanamycin sulphate or 50 mg streptomycin sulphate l 21 .Growth was observed for up to 5 days.
The results of the physiological tests are summarized in Table 2; a compilation of results of every strain studied is supplied as Supplementary Table S1 in IJSEM Online.The strains of Corallococcus coralloides and Corallococcus exiguus hydrolysed starch and xylan so strongly that colourless haloes that exceeded the size of the swarms appeared after covering of the agar plates with Lugol's solution.In contrast, the strains of the genus Myxococcus, including Corallococcus macrosporus, only showed very weak, lightblue haloes that never exceeded the size of the swarms.
Aesculin hydrolysis, detected as darkening of the medium, was positive (though weak for some strains) for strains of Corallococcus coralloides and Corallococcus exiguus and negative for strains of the genus Myxococcus and of Corallococcus macrosporus.Three features mentioned by Reichenbach (2005) as being useful to distinguish members of the genera Corallococcus and Myxococcus were re-evaluated.Most of the Myxococcus strains and the strains of Corallococcus macrosporus showed growth on agar plates containing 1.5 % NaCl, while the authentic Corallococcus strains did not.Four of nine selected Corallococcus strains were resistant to kanamycin.Resistance to streptomycin was displayed by strains of the species M. virescens, M. flavus, M. fulvus and M. stipitatus, while three out of four strains of M. xanthus were susceptible to this antibiotic.The Corallococcus strains did not grow in the presence of streptomycin, except Corallococcus macrosporus, which again showed the same reactions as the Myxococcus strains (Table 2).
In order to substantiate the discrimination between the two genera, commercially available test strips for bacterial identification were tested.Neither API 20NE (bioMe ´rieux) nor Biolog GN microplates (testing for the metabolism of 95 substrates) could be adapted in such a way that meaningful reproducible results were obtained with myxobacteria.In contrast, test strips that predominantly determine enzyme reactions, namely API Campy and API ZYM (bioMe ´rieux), resulted in patterns for Corallococcus strains that could be evaluated (Stackebrandt et al., 2007).However, the features offered by these strips were not helpful in distinguishing members of the genera Myxococcus and Corallococcus, as Myxococcus strains displayed the same reactions as described for Corallococcus (results not shown).Only the reduction of triphenyltetrazolium chloride tended to be positive for Corallococcus and negative for Myxococcus strains.
Biomass for whole-cell fatty acid analysis was cultivated on CY agar for 2 days at 28 u C, after inoculation with a suspension of cells grown on VY/2 agar.Fatty acid methyl esters were obtained as described previously (Ka ¨mpfer & Kroppenstedt, 1996) and separated by a gas chromatograph (model 5898A; Hewlett Packard).Peaks were automatically computed using the Microbial Identification standard software package (Sasser, 1990).
The whole-cell fatty acid patterns of the Corallococcus strains were published previously (Stackebrandt et al., 2007).The patterns of the Myxococcus strains determined after propagation of the cells under the same conditions are listed in Supplementary Table S2.Branched fatty acids were the predominant components in strains belonging to the genus Myxococcus, headed by iso-C 15 : 0 and iso-C 17 : 0 .The profiles included the hydroxylated acids iso-C 14 : 0 3-OH (1.4-10.4%), iso-C 15 : 0 3-OH (1.1-2.6 %) and iso-C 17 : 1 3-OH (1.2-3.7 %).The following differences were found in comparison with the Corallococcus strains (excluding Corallococcus macrosporus): C 16 : 1 v5c amounted for at least 9.9 % in Myxococcus strains, while the proportion was 3.2 % or less in Corallococcus strains; iso-C 13 : 0 and anteiso-C 17 : 1 were present in proportions of at least 2.7 and 12.0 %, respectively, in Corallococcus strains, but significantly less in Myxococcus strains; C 16 : 1 v11c was present in Myxococcus strains but not in the Corallococcus strains (Table 2).The strains of Corallococcus macrosporus displayed profiles similar to the profiles of members of the genus Myxococcus.
The bifurcation in the phylogenetic tree matches the differences in fatty acid profiles and grouped species of the genus Myxococcus and Corallococcus macrosporus in one cluster and the remaining Corallococcus strains in another.Since this demarcation is in good agreement with the physiological properties (based upon different reactions towards starch, xylan and aesculin, tolerance of NaCl, streptomycin and kanamycin), we found evidence to support the separate status of the two genera Myxococcus and Corallococcus, as indicated in the emended descriptions below.

Emended description of the genus Myxococcus
Thaxter 1892 (Approved Lists 1980) Morphological features of cells, swarms, myxospores and fruiting bodies are as described by Reichenbach (2005b).Members of the genus do not decompose starch or xylan on agar plates and do not hydrolyse aesculin.Variable for growth in the presence of 1.5 % NaCl.Sensitive to kanamycin.Most strains are tolerant of 50 mg streptomycin sulphate l 21 .Whole-cell fatty acid patterns have the following characteristics when grown on agar medium CY: main components iso-C 15 : 0 and C 16 : 1 v5c; contain no iso-C 14 : 0 but contain C 16 : 1 v11c; content of iso-C 13 : 0 1.6 % or  1980).The Approved Lists drew heavily on the eighth edition of Bergey's Manual of Determinative Bacteriology (Buchanan & Gibbons, 1974) as a source for locating suitable type material, as well as the source of many of the species and genus descriptions.In this edition of Bergey's Manual, strain Windsor M78 T is referred to as a reference strain of Myxococcus stipitatus Thaxter 1897, while Myxococcus macrosporus is typified by the work of Krzemieniewska & Krzemieniewski (1926), and strain M271 T is referred to as a reference strain.Despite the fact that these two strains became the type strains of these species on the Approved Lists, the term 'reference strain' rather than 'type strain' was used for these organisms in the first edition of Bergey's Manual of Systematic Bacteriology (Brockman, 1989).
While the genus affiliation of M. stipitatus has not been changed over decades, the taxonomic history of M. macrosporus is chequered.The species was originally placed in the genus 'Chondrococcus' (Jahn, 1924) as 'Chondrococcus macrosporus' by Krzemieniewska & Krzemieniewski (1926) and retained, though illegitimate (the name has previously been used for an alga; Jeffers & Holt, 1961), in this genus until the seventh edition of Bergey's Manual (Breed et al., 1957).When the taxonomic importance of the consistency of slime of fruiting bodies (firm in strains of the genus 'Chondrococcus', while deliquescent in strains of the genus Myxococcus) was disregarded (Zahler & McCurdy, 1974), the eighth edition of Bergey's Manual (Buchanan & Gibbons, 1974)  be placed properly in the genus Myxococcus, the type of Myxococcus macrosporus would be strain M271 T , a rather undesirable state of affairs.
The situation is best solved by a Request for an Opinion regarding the proposal that the type of Myxococcus macrosporus be represented by a neotype, strain Cc m8 (deposited as DSM 14697 5CIP 109128).Since this strain is already the type of Corallococcus macrosporus, a consequence of this action would be that the names Corallococcus macrosporus and Myxococcus macrosporus should be considered to be homotypic synonyms, as originally intended by Reichenbach (2005c).We consider that Corallococcus macrosporus is a member of the genus Myxococcus and, as already pointed out above, the description of Corallococcus macrosporus provided by Reichenbach (2005c) demonstrates that the proposed neotype strain Cc m8 agrees closely with the descriptions given by Krzemieniewska & Krzemieniewski (1926) and Zahler & McCurdy (1974).
A similar situation is seen with Myxococcus stipitatus.
Because of the unavailability of the type strain Windsor M78 T , Reichenbach (2005b) designated a new type strain Mx s8 rather than proposing that a neotype be designated.Here, we request an Opinion to the effect that strain Mx s8 (5DSM 14675 5JCM 12634) serve as the neotype strain for Myxococcus stipitatus.The description of Myxococcus stipitatus provided by Reichenbach (2005b) demonstrates that the proposed neotype strain agrees closely with the descriptions given by Thaxter (1897) and Zahler & McCurdy (1974).We are aware that in both cases the designation of a neotype strain can only become effective 2 years after publication of this proposal.
Profiles of phenotypic characters and fatty acids of individual strains are available with the online version of this paper.IP: 54.70.40.11On: Sat, 03 Aug 2019 19:11:29

Table 2 .
Features that differentiate the genera Myxococcus and Corallococcus Myxococcus (Corallococcus) macrosporus includes strain Cc m8.Results for individual strains are given in Supplementary TablesS1 and S2.V, Variable results (numbers of positive strains given in parentheses); for reactions to antibiotics, positive indicates resistance, growth, and negative indicates sensitivity, no growth; ND, not detected., of summed feature 4 (including anteiso-C 17 : 1 ) 7.4 % or less and of C 16 : 1 v5c at least 9.9 %.The G+C content of the DNA is 65-71 mol%.The type species is Myxococcus fulvus (Cohn 1875)Jahn 1911 (Approved Lists 1980).NaCl.Sensitive to streptomycin.Some strains are tolerant of 250 mg kanamycin sulphate l 21 .Whole-cell fatty acid patterns have the following characteristics: main components are iso-C 15 : 0 , anteiso-C 17 : 1 and iso-C 17 : 0 ; contain iso-C 14 : 0 but no C 16 : 1 v11c, content of iso-C 13 : 0 at least 2.7 % and of C 16 : 1 v5c 3.2 % or less.The G+C content of the DNA is 66-68 mol%.The type species is Corallococcus coralloides (Thaxter 1892) Reichenbach 2007.Myxococcus stipitatusThaxter 1897, with the type strain Windsor M78 T , and Myxococcus macrosporus (Krzemieniewska and Krzemieniewski 1926) Zahler and McCurdy 1974, with the type strain Windsor M271 T , were included on the Approved Lists of Bacterial Names (Skerman et al., ND 1.2-1.3Downloaded from www.microbiologyresearch.orgby IP: 54.70.40.11On: Sat, 03 Aug 2019 19:11:29 less (Reichenbach, 2005c)the genera Myxococcus, 'Chondrococcus' (including 'Chondrococcus macrosporus') and Angiococcus in the genus Myxococcus, with several being listed as species incertae sedis.In drawing up the Approved Lists, Skerman et al.Corallococcus macrosporus is listed as a member of the new genus Corallococcus ('Chondrococcus'  Jahn 1924, 85)in the second edition of Bergey's Manual of Systematic Bacteriology(Reichenbach, 2005c).Besides Corallococcus macrosporus, the genus contained Corallococcus coralloides and the novel species Corallococcus exiguus.The criteria for establishing a new genus distinct from Myxococcus were morphological, such as the cartilaginous consistency of fruiting bodies, their bizarre shape and the large number of small fruiting bodies.
Reichenbach (2005c)to be available and could not be included in these studies.However, it is clear thatReichenbach (2005c)considers Corallococcus macrosporus and Myxococcus macrosporus to be synonyms, despite the fact that two different types have been designated.If the species Corallococcus macrosporus was to be considered to be a heterotypic synonym of Myxococcus macrosporus and International Journal of Systematic and Evolutionary Microbiology 59 Downloaded from www.microbiologyresearch.orgby IP: 54.70.40.11On: Sat, 03 Aug 2019 19:11:29