Nitriliruptor alkaliphilus gen. nov., sp. nov., a deep-lineage haloalkaliphilic actinobacterium from soda lakes capable of growth on aliphatic nitriles, and proposal of Nitriliruptoraceae fam. nov. and Nitriliruptorales ord. nov.

Organic compounds containing a C≡N bond (nitriles) are important intermediates and building blocks in the synthesis of organic compounds. There are also a few examples of naturally occurring nitriles, which are either formed by cyanogenic plants from cyanide (Vetter, 2000) or produced during anaerobic degradation of amino acids (Harper & Gibbs, 1979). Most nitriles are toxic and difficult to degrade. Nevertheless, many bacterial species and some fungi can hydrolyse the nitrile bond via two different enzymic pathways (nitrile hydratase and nitrilase) with the corresponding carboxylic acid and ammonia as final products. Microorganisms possessing these enzymes are valuable biocatalysts and can be used either in organic synthesis or in environmental biotechnology (Kobayashi & Shimizu, 2000; Banerjee et al., 2002). Previously, all known nitrile-degrading microorganisms were neutrophilic, i.e. growing optimally at neutral pH values and in non-saline media.

We recently described the first example of a bacterium from soda lake sediments, Natronocella acetinitrilica, capable of growth with acetonitrile and propionitrile as carbon, energy and nitrogen source at haloalkaline conditions (Sorokin et al., 2007a). Furthermore, we also demonstrated the possibility of biodegradation of more complex nitriles, such as isobutyronitrile (iBN) [(CH₃)₂CHCN], by two haloalkaliphilic consortia, enriched from soda lakes and soda soils (Sorokin et al., 2007b). In the soda lake consortium, the key reaction in iBN hydrolysis was performed by the novel haloalkaliphilic actinobacterium described in the present paper. This bacterium is considered to represent a novel, deeply rooted cluster within the class Actinobacteria.

The isolation of an iBN-degrading consortium from soda lake sediments at pH 10 and the growth and metabolic properties of its members have been described previously (Sorokin et al., 2007b). A stable co-culture enriched from soda lake sediments capable of complete utilization of iBN at pH 10 and 0.6 M total Na⁺ consisted of a motile spirillum as the dominant morphotype and short, non-motile rods. The spirillum was isolated from the mixed culture by using isobutyroamide, which is the first product of iBN hydrolysis in the nitrile hydratase/amidase pathway. The organism was identified as a member of the genus Marinospirillum in the class Gammaproteobacteria (Sorokin et al., 2007b). Further experiments showed that this bacterium played the role of scavenging of iBN hydrolysis products released by the rod-shaped phenotype. The latter was obtained in pure culture by using high concentrations of iBN (up to 20 mM), which proved toxic for the Marinospirillum satellite organism. The iBN-hydrolysing strain was designated ANL-iso2^T.

Phase-contrast photomicrographs were obtained by using a Zeiss Axioplan Imaging 2 microscope. For electron microscopy, cells were fixed with glutaraldehyde (final concentration 3 %, v/v) and positively contrasted with 1 % (w/v) uranyl acetate. For thin sectioning, cells were fixed in 1 % (w/v) OsO₄+0.5 M NaCl for 3 h at room temperature, washed and stained overnight with 1 % (w/v) uranyl acetate, dehydrated in an ethanol series and embedded in Epon resin. Thin sections were stained with 1 % (w/v) lead citrate. Fatty acids of the cellular lipids were extracted with a mixture of methanol/chloroform and analysed by GC-MS according to Zhilina et al. (1997). The cell-wall fraction was obtained and analysed according to Streshinskaia et al. (1979). A cell suspension in 1 % (w/v) SDS was disrupted by sonication and heated briefly (100 °C, 5 min). Cell walls were separated by fractional centrifugation and were purified by using trichloroacetic acid and trypsin. Quantitative determination of the cell-wall amino acids was performed with an LC 600 E amino acid analyser (Biotronic) after acid hydrolysis (6 M HCl, 100 °C, 18 h). Isomers of diaminopimelic acid (DAP) were determined by TLC on cellulose (modified method of Hasegawa et al., 1983).

Strain ANL-iso2^T formed flat, spreading microcolonies on the iBN alkaline agar after 1 month of incubation. Cells were small, non-motile rods, occurring singly or in pairs under most conditions (Fig. 1a⇓). When grown on glucose and yeast extract, chain formation was observed (Fig. 1b⇓). The cell wall was of the Gram-positive type (Fig. 1c⇓). The cell wall contained alanine, glutamic acid and meso-DAP in a molar ratio of 2 : 1 : 1, which indicated the presence of the A1γ-type peptidoglycan (Schleifer & Kandler, 1972). Small amounts (up to a molar ratio of 0.4) of other amino acids were also detected, which suggested the presence of additional peptides. Furthermore, glucose, galactose and glycerol were found in the cell-wall hydrolysates (hydrolysis at 2 M HCl, 3 h, 100 °C). These data, along with the presence of glycerol phosphates, phosphodiesters and an unidentified amino sugar, as revealed by electrophoresis of the cell-wall hydrolysates, suggested the presence of an anionic carbohydrate-containing polymer(s) in the cell wall of strain ANL-iso2^T (Naumova et al., 2001). Analysis of whole-cell fatty acids of strain ANL-iso2^T grown with iBN at pH 10, 0.6 M Na⁺ and 30 °C demonstrated a dominance of saturated components, including C_16 : 0 (19 %), iso-C₁₄ (17 %), C_14 : 0 (13 %) and iso-C₁₆ (8 %). Unsaturated components were represented by C_16 : 1ω7 (16 %), C_17 : 1ω8 (5 %) and C_18 : 1ω9 (4 %).

Strain ANL-iso2^T completely degraded iBN, utilizing it as a sole source of energy, carbon and nitrogen. However, growth of the pure culture was very slow, not only with nitriles but also with other growth substrates in general. It converted iBN rapidly into isobutyroamide, isobutyrate and ammonia during the initial phase with little biomass growth. Subsequent actual biomass growth was very slow, with μ_max=0.035 h⁻¹. Strain ANL-iso2^T was also capable (after adaptation) of growth on a range of linear aliphatic nitriles with different chain lengths, such as propionitrile (C₃), butyronitrile (C₄), valeronitrile (C₅) and capronitrile (C₆), as the carbon, energy and nitrogen source, with μ_max values similar to that with iBN. However, it was not able to grow on acetonitrile (C₂). Other compounds utilized as single carbon and energy source included the respective organic acids produced from the aliphatic nitriles listed above, acetate, malate, fumarate, succinate, pyruvate, citrate, d-glucose, d-fructose, maltose, d-mannose, melezitose, α,α-trehalose, sucrose, d-arabinose, inositol, cellobiose and glycerol. Growth was also possible on complex media, including yeast extract or peptone broth, but was much slower than in the presence of organic acids or sugars. Anaerobic fermentative growth with glucose, fructose or iBN plus nitrate was not observed. Strain ANL-iso2^T was able to hydrolyse and utilize gelatin, but not starch, casein, cellulose, chitin, pectin or xylane.

The pH range for growth of strain ANL-iso2^T at 0.6 M Na⁺ with iBN was 8.4–10.4 (optimum growth at pH 9.0–9.5) and the salt range for growth at pH 10 was 0.1–2.0 M total Na⁺ (optimum growth at 0.2–0.3 M), qualifying the organism as a moderately salt-tolerant, obligate alkaliphile (Fig. 2⇓). In contrast to the conditions for growth, hydrolysis of iBN by washed cells and cell-free extract was possible at acidic to neutral pH with an optimum at pH 8, indicating intracellular localization of the nitrile hydratase/amidase system (Fig. 2⇓). The metabolic potential of strain ANL-iso2^T was not limited by the ability to hydrolyse C₃–C₆ aliphatic nitriles, which are utilized as growth substrates. Experiments with washed cells grown with iBN demonstrated that the actual range of nitriles being converted was much broader than utilized for growth, including such industrially important nitriles as acrylonitrile, metacrylonitrile, acetonitrile, benzonitrile, nicotinonitrile and adiponitrile, but also more industrially challenging nitriles such as tertiary nitriles and α-azidonitriles. In all cases, nitrile hydratase activity was two to three orders of magnitude higher than the corresponding amidase activity.

Isolation of DNA and subsequent determination of G+C content were performed according to Marmur (1961) and Marmur & Doty (1962), respectively. Isolation of template DNA, amplification of the 16S rRNA gene of strain ANL-iso2^T and its sequencing have been described previously (Sorokin et al., 2007a). The 16S rRNA gene sequence of strain ANL-iso2^T was aligned with those from GenBank by using clustal w. Phylogenetic trees were reconstructed via four different algorithms by using the treeconw software package (Van de Peer & De Wachter, 1994). Pairwise evolutionary distances (expressed as estimated changes per 100 nt) were calculated by using the method of Jukes & Cantor (1969). Phylogenetic trees were then reconstructed by using three different algorithms. As the tree topology was very similar in all three cases, final results are presented for the neighbour-joining analysis only. Bootstrap analysis (based on 1000 replications) was used to validate the reproducibility of the branching patterns of the trees. Phylogenetic analysis based on 16S rRNA gene sequencing placed the nitrile-utilizing alkaliphile within the class Actinobacteria as a novel deep lineage (Fig. 3⇓). It formed an independent cluster together with several uncultured actinobacteria from various habitats, including soda lakes in the USA, Egypt and Russia (Humayoun et al., 2003; Mesbah et al., 2007; Foti et al., 2008). However, strain ANL-iso2^T showed no more than 85 % 16S rRNA gene sequence similarity to any recognized taxon. Therefore, the only safe conclusion that could be drawn from 16S rRNA gene sequence comparison was that the novel bacterium is a member of the class Actinobacteria. At the time of writing, the class Actinobacteria consists of six orders, among which the order Actinomycetales is the most numerous and diverse (see Fig. 3⇓). Strain ANL-iso2^T, despite being a typical actinobacterium in its morphotype, phylogenetically forms an independent lineage. The lack of such actinobacteria in culture might originate from their very slow growth and specialization on exotic toxic substrates such as nitriles. It is interesting to note that actinobacterial genera such as Rhodococcus, Arthrobacter, Nocardia and Gordonia contain many highly active and versatile nitrile-hydrolysing strains, including some that are industrially important (Bunch, 1998; Kobayashi & Shimizu, 2000; Brady et al., 2004).

Strain ANL-iso2^T can be regarded as the first haloalkaliphilic versatile nitrile degrader and as a representative of a novel deep phylogenetic lineage within the class Actinobacteria. Based on unique phenotypic properties and distinct phylogeny, we suggest that strain ANL-iso2^T represents a novel species of a new genus, for which the name Nitriliruptor alkaliphilus gen. nov., sp. nov. is proposed, and that this forms a novel family Nitriliruptoraceae fam. nov. and a new order Nitriliruptorales ord. nov. within the class Actinobacteria.