Myxobacteria in high moor and fen: An astonishing diversity in a neglected extreme habitat

Abstract Increasing antibiotic resistances of numerous pathogens mean that myxobacteria, well known producers of new antibiotics, are becoming more and more interesting. More than 100 secondary metabolites, most of them with bioactivity, were described from the order Myxococcales. Especially new myxobacterial genera and species turned out to be reliable sources for novel antibiotics and can be isolated from uncommon neglected habitats like, for example, acidic soils. Almost nothing is known about the diversity of myxobacteria in moors, except some information from cultivation studies of the 1970s. Therefore, we evaluated the myxobacterial community composition of acidic high moor and fen both with cultivation‐independent 16S rRNA clone bank analysis and with cultivation. Phylogenetic analyses of clone sequences revealed a great potential of undescribed myxobacteria in high moor and fen, whereby all sequences represent unknown taxa and were detected exclusively by cultivation‐independent analyses. However, many clones were assigned to sequences from other cultivation‐independent studies of eubacterial diversity in acidic habitats. Cultivation revealed different strains exclusively from the genus Corallococcus. Our study shows that the neglected habitat moor is a promising source and of high interest with regard to the cultivation of prospective new bioactive secondary metabolite producing myxobacteria.

: Origin of clones achieved in this study.

Isolation of predators with bait bacteria isolated from the moor samples
As an approach to the moor-environment, non-myxobacterial bacteria were also isolated from the samples from which the clone banks have been established and offered as bait. On 1/10 Luria Bertani-agar [per liter: 1.0 g tryptone, 0.5g yeast extract, 1.0 g NaCl; pH 5.5] different dilutions of samples A1, B4, B7, B9, B18, B21, B24, C1 and C2 (in water) were plated and as many morphological different bacteria as possible were isolated. The DNA of the pure cultures was extracted and 16S rRNA gene PCR with primer F27/R1525 was carried out. 16S rRNA genes of bait organisms were partial sequenced with primer F27, F518 and R1100. Consensus sequences were compared with sequences of the NCBI database. Eight different bacteria could be isolated from samples B18, B21, C1, and C2 (Table S2) and were streaked alive cross like on water agar. The moor samples were placed at the end of the cross and incubated at room temperature.

Sample treatment
Water samples were treated in four different ways: From 25 ml centrifuged moor water the pellets were treated in the same way as the solid samples. In addition moor water was diluted 1:1000. A total of 20 µl of this dilution was dropped on water agar/E. coli and filter paper (Stan 21). The third approach was plating 100 µl of this dilution on water agar and Stan 21. After drying, E. coli bait and filter pieces were added, respectively. The last approach was filtering 10 ml moor water with a water jet pump through a cellulose filter. The filter was placed on VY/2-agar (per liter: Baker's yeast: 5.00 g, CaCl2 x 2 H2O: 1.36 g, agar (Difco): 20 g, Vitamin B12: 0.50 mg. Sterilized vitamin B12 solution was added after autoclaving separately by filtration).

Sequence analyses
For a first assignment, 16S rRNA genes of the cultures were sequenced using primer F27 and F518.
For six cultivated representatives of the OTUs (C8, B1, C6, C4, B17, C17), full length 16S rRNA genes were sequenced using additionally primer F357 (Muyzer et al., 1993), F945, R1078 and R1525 to assure that both nucleotide directions were covered. The sequences of these six cultures have been deposited at GenBank under accession numbers KP18974-KP18979. The specificity of the two primer combinations (FW2/FW5 and R1525) was checked by PCR using genomic DNA of 21 representatives derived from the phyla Actinobacteria, Cyanobacteria, Firmicutes, α-, β-, and γ-Proteobacteria (Table   S1) and additional 22 representatives of the order Myxococcales (Table S2). PCR applicability of extracted DNA from the representatives was initially checked with eubacterial primers F27/R1525 as described for the isolated strains. Afterwards the PCR conditions for the semi-specific myxobacterial primer sets (FW2/FW5 -R1525) were tested. The annealing temperature was optimized using sequential gradient PCR reactions, starting with a temperature range between 50°C and 70°C. The PCR products were checked via agarose gel and the temperature range was narrowed down until an optimal annealing temperature for both primer sets could be defined, such as 61.1°C for combination FW2/R1525 and 65.5°C for FW5/R1525. These conditions were used for the amplification of myxobacterial 16S rRNA genes in the further process.

Data analysis
The 16S rRNA gene sequences acquired via clone bank and from the isolates were checked for quality using the program BioEdit (free available). The sequences from the cultures were assembled into consensus sequences. The single sequences from clones as well as the consensus sequences from cultures were compared with the NCBI database entries. Closely related myxobacterial sequences, representing different species, were imported into the ARB database (version 14.02.2005 database; http://www.arb-home.de) and aligned together with the sequences acquired in this study. A distance matrix tree was constructed with 16S rRNA sequences of 56 myxobacterial type strains using the Neighbour-Joining method (Saitou and Nei, 1987) and Jukes Cantor correction (Jukes and Cantor, 1969). The topology of the phylogenetic tree was built by bootstrap analysis of 1000 operations. As recommended in ARB, no filter was used in bootstrapping. Sequences sharing more than 99% similarity, calculated with the similarity matrix tool in ARB, were grouped into operational taxonomic units (OTUs). For phylogenetic analyses, the 16S rRNA gene of a cultivated representative of each OTU, if there was one, was fully sequenced. Only the type and representative myxobacterial strains as well as clones of uncultured bacteria with highest similarity to the novel isolates were shown in the phylogenetic tree. Table S3 shows the type strains and their corresponding 16S rRNA gene accession numbers used for the construction of the phylogenetic tree. Non Myxococcales-sequences were excluded from our analyses (data not shown). Single sequences or sequences of clones from OTUs which did not include type strain sequences were blasted in GenBank to figure out the next cultivated relative. In all cases, the next cultivated relative belongs to the Myxococcales. So we are sure that all our clone sequences included in this study are of myxobacterial origin.