Janibacter hoylei sp. nov., Bacillus isronensis sp. nov. and Bacillus aryabhattai sp. nov., isolated from cryotubes used for collecting air from the upper atmosphere

Collection of air from 24 km and above

The balloon with the cryosampler payload used for collecting high-altitude air samples (altitude 20–41.4 km) was launched on 20 April 2005 from the National Scientific Balloon Facility of the Tata Institute of Fundamental Research at Hyderabad, India. The cryogenic sampler, comprising a 16-cryoprobe assembly, is similar to one used previously (Wainwright et al., 2003; Shivaji et al., 2006). All 16 cryosampling tubes were autoclaved at 120 °C for 4 h and integrated into the payload in a tissue-culture laboratory which was UV-sterilized overnight. The cryotubes were also checked for sterility after the assembly. For this purpose, three cryotubes were disconnected from the assembly, opened in a laminar flow hood, rinsed with sterile phosphate buffer (0.1 M, pH 7.2) and plated on Luria–Bertani (LB) agar and incubated at 37 °C. No colonies had appeared after 1 month, indicating that the cryotubes were sterile. Therefore, the isolates reported in Table 1⇓ should have originated from the air samples collected at the given altitudes. Cryotubes 1, 3, 5, 7, 9, 12, 13 and 15 were examined at the Centre for Cellular and Molecular Biology, Hyderabad, while cryotubes 2, 4, 6, 8, 10 and 14 were similarly studied at the National Centre for Cell Sciences, Pune. Care was taken that the two laboratories followed similar protocols, and there was frequent interaction and discussion between the two groups to ensure homogeneity of procedures and interpretation.

Detection of viable bacteria from air collected in the cryotubes

All the work related to detection of microbes in air samples from the cryotubes was carried out in a room with very little human traffic and inside a laminar flow hood available exclusively for the handling of the probes. The probes were transferred to the laminar flow hood one at a time and, prior to the transfer, the surface of the cryotubes was cleaned and sterilized with alcohol. All instruments used to unscrew the outlet valve of the cryotube such as spanners and screwdrivers were surface-sterilized with alcohol and flame-heated prior to use. The outlet valve of the cryotube was then connected to a series of two Millipore filtration units using sterile tubing and the air was sequentially filtered through a 0.45 μm filter (Millipore cat. no. HAWP 04700) and a 0.22 μm filter (Millipore cat. no. GSWP 04700) under aseptic conditions in the laminar flow hood. The 0.45 μm and 0.22 μm filters were then used for detection of bacteria. For this purpose, each filter was cut into quarters and one of the quarters was transferred to a nutrient agar plate (0.5 % peptone, 0.3 % beef extract, 0.5 % NaCl and 1.5 % agar, w/v) and incubated at 15 °C. After about 10 days, if no growth was observed, the filter was transferred to an LB agar plate (1 % tryptone, 0.5 % yeast extract, 1.0 % NaCl and 2.0 % agar, w/v; pH 7.2) or an oligotrophic LB agar plate (1/10 LB) and incubated at 15 °C for a further 15 days. Subsequently, if no colonies appeared, the filters were incubated on LB agar at 25 °C for up to a month. It should be noted that, when hydrophobic 0.45 μm (Millipore cat. no. HVHP 04700) and 0.22 μm (cat. no. GVHP 04700) filters were used for filtering the air, a thin layer of medium of the same composition was applied to the surface of the plates before the filters were placed on the surface of the medium.

Two of the remaining quarters of the filter were transferred to minimal salts agar. This medium contained (per 100 ml) 20 ml of a 3.39 % (w/v) sodium dihydrogen phosphate solution, 20 ml 1.5 % (w/v) potassium phosphate, 20 ml 0.25 % (w/v) NaCl, 20 ml 0.5 % (w/v) ammonium chloride and 18 ml Millipore water. The final concentration of agar was 2 %. The medium was also supplemented with 2 ml sterile 20 % glucose and 0.2 ml sterile 0.1 M MgSO₄ solution; the pH was adjusted to 7. This medium was used for the cultivation of organisms which require low nutrients. Blood agar [2.3 % (w/v) peptone, 0.1 % (w/v) corn starch, 0.5 % (w/v) NaCl and 1.5 % (w/v) agar, to which 5 % (v/v) defibrinated blood was added (after autoclaving and cooling the medium)] was used for the cultivation of organisms which require rich media. Plates of minimal salt agar and blood agar were incubated at 30 °C. One quarter of the filter was also incubated at 30 °C on Sabouraud agar (1.0 % enzymic digest of casein, 4 % glucose and 1.5 % agar, w/v) to facilitate fungal growth (data not shown).

Of the 15 cryotubes from which air was sampled, only a single novel bacterial colony was detected, in cryotube 11, corresponding to air collected at an altitude of 40–41.4 km (Table 1⇑). When placed on minimal salts agar, the filter paper from cryotube 11 showed a single cream-coloured colony after 24 days of incubation at 30 °C, and it was designated strain PVAS-1^T.

During the filtration process, plates containing fungal and bacterial media were left open in the laminar flow hood throughout the operation to check the sterility of the hood. These plates were incubated at 25 °C (for fungi) or 37 °C (bacteria) for 7 days prior to use, to confirm the sterility of the plates. Colonies were not detected on any of these plates, indicating that the sampling was done under sterile conditions.

Detection of viable bacteria from washes of empty cryotubes

It seemed possible that some bacteria might have remained at the bottom of the tubes or attached to their polished walls and thus escaped detection. To check this possibility, all 15 empty cryotubes were injected with 200 ml sterile phosphate buffer (0.1 M, pH 7.2) and agitated at 22 °C for 6 h in a shaker. The liquid was then removed using sterile tubing and a syringe and filtered through a 0.22 μm Millipore filter (47 mm diameter). These filters were each cut aseptically into four sectors and each sector was placed on a plate containing one of four different media, LB agar, minimal salts agar, blood agar and Sabouraud agar, and incubated under the conditions described above. The plates were observed for the appearance of colonies over a period of 1 month.

From the 15 cryotube wash samples, 12 bacterial colonies were detected (Table 1⇑). Nine (PVAS-2, -3, -4, -5, -6, -8 and -10 and B5W22-1 and B5W22-2) showed more than 98 % 16S rRNA gene sequence similarity to reported species (Table 1⇑). Therefore, further attempts were not made to characterize them to the species level.

Morphological, biochemical and chemotaxonomic characteristics

Morphological, growth and biochemical studies of viable colonies were performed using standard methods (Holding & Collee, 1971; Smibert & Krieg, 1994). LB agar was used for growth and maintenance and for the determination of phenotypic and chemotaxonomic characteristics of all strains except PVAS-2 and -3. Strains PVAS-2 and -3 were grown on a medium containing 0.25 % (w/v) K₂HPO₄, 0.225 % (w/v) NaH₂PO₄, 0.05 % (w/v) (NH₄)₂SO₄, 0.02 % (w/v) MgSO₄ . 7H₂O and 0.1 % (v/v) methanol. The shape, size and motility of the strains were ascertained using a Leitz Diaplan phase-contrast microscope with an oil-immersion objective (×100). The sensitivity of the cultures to antibiotics was determined by using antibiotic discs (HiMedia). Utilization of various carbon compounds as sole carbon sources was tested in mineral liquid medium supplemented with 0.2 % filter-sterilized carbon source as described previously (Shivaji et al., 2006). Fatty acid and lipid composition (Sato & Murata, 1988; Kiran et al., 2004), DNA G+C content (Shivaji et al., 1992, 2005) and DNA–DNA hybridization (Tourova & Antonov, 1987; Shivaji et al., 1992) were determined according to standard procedures. Analyses of polar lipids, respiratory quinones and meso-diaminopimelic for PVAS-1^T were carried out by the Identification Service of the DSMZ (Braunschweig, Germany). Isoprenoid quinones were also extracted according to the method of Collins et al. (1977), separated by HPLC and identified as described previously (Reddy et al., 2003). Peptidoglycan was prepared and analysed according to the method described by Komagata & Suzuki (1987).

Detailed phenotypic and chemotaxonomic characteristics (listed in Tables 2⇓, 3⇓ and 4⇓ and in the species descriptions) were compared with those of Bacillus silvestris DSM 12223^T, Bacillus megaterium MTCC 428^T, Janibacter anophelis CCUG 49715^T, Janibacter terrae DSM 13876^T, Janibacter melonis DSM 16063^T, Janibacter limosus DSM 11140^T and Janibacter corallicola DSM 18906^T.

Cells of strain PVAS-1^T are Gram-positive, coccoid, non-endospore-forming and non-motile and occur singly or in clumps. Some biochemical characteristics are listed in Table 2⇑. meso-Diaminopimelic acid is the diagnostic diamino acid in the cell-wall peptidoglycan and iso-C_16 : 0 (50.4 %), iso-C_18 : 0 (5.1 %), 10-methyl C_17 : 0 (10.9 %), C_17 : 1Δ9c (12.1 %) and C_18 : 1Δ9c (6.0 %) are the predominant fatty acids. Mycolic acids are absent. The predominant isoprenoid quinone present is MK-8(H₄). The DNA G+C content is 72.8 mol%. These characteristics and phylogenetic analysis of the 16S rRNA gene sequence indicated that PVAS-1^T is a member of the genus Janibacter.

Cells of strains B3W22^T and B8W22^T are Gram-positive, rod-shaped, endospore-forming and catalase-positive, with C_15 : 0, C_16 : 0, iso-C_15 : 0, anteiso-C_14 : 0, anteiso-C_16 : 0 and C_16 : 1Δ9c as the predominant fatty acids (Tables 3⇑ and 4⇑). The polar lipid profile of strain B3W22^T contains phosphatidylglycerol (PG), phosphatidylethanolamine (PE), phosphatidylserine (PS) and one unidentified lipid, whereas that of strain B8W22^T has only PG and PE. MK-6, MK-7 and MK-8 are the menaquinones in strain B3W22^T whereas, in B8W22^T, only MK-7 was present (Table 3⇑). The DNA G+C contents of the two strains are 38–40 mol%. These characteristics indicated that strains B3W22^T and B8W22^T are members of the genus Bacillus (Tables 3⇑ and 4⇑).

Phylogenetic analysis

The 16S rRNA gene was amplified from genomic DNA, purified and sequenced as described previously (Shivaji et al., 2000; Pidiyar et al., 2004). To ascertain the phylogenetic affiliation of the novel strains, the almost-complete 16S rRNA gene sequences of the bacterial isolates were manually corrected and aligned using clustal_x (Thompson et al., 1994). Phylogenetic trees were constructed based on neighbour-joining (Saitou & Nei, 1987) and maximum-parsimony methods using mega 3.1 (Kumar et al., 2004). Evolutionary distances were determined with Kimura's two-parameter model (Kimura, 1980). Bootstrap analysis (Felsenstein, 1993) was performed for 1000 replications. Reference sequences were retrieved from GenBank under the accession numbers indicated on the trees.

Phylogenetic analysis of the 16S rRNA gene sequence of strain PVAS-1^T (1196 bp) indicated that its closest relatives were Janibacter anophelis CCUG 49715^T (98 % similarity) and Janibacter terrae DSM 13953^T (98 %) (Lang et al., 2003). Both the neighbour-joining tree (Fig. 1⇓) and the maximum-parsimony tree (not shown) revealed that strain PVAS-1^T clustered most closely with these strains. DNA–DNA relatedness of PVAS-1^T with Janibacter terrae DSM 13953^T was 29.8 % and with Janibacter anophelis DSM 18333^T was 13.7 %, and the reciprocal reactions in both cases gave values of about 9–10 % . Thus, strain PVAS-1^T represents a novel species of the genus Janibacter, for which the name Janibacter hoylei sp. nov. is proposed.

Strains B3W22^T and B8W22^T are related at the 16S rRNA gene sequence level (91 % similarity), but the sequences differ by more than 2.5 %, indicating that they probably represent different species. blastn analysis indicated that the nearest phylogenetic neighbour of B3W22^T is Bacillus silvestris H3-23^T (99.5 % similarity), with which it forms a clade in the phylogenetic tree, and it is separated from Caryophanon latum DSM 14151^T and Caryophanon tenue DSM 14152^T, with which it exhibits about 97 and 96 % similarity, respectively (Fig. 2⇓). The members of the genus Caryophanon are characterized by the presence of slightly curved to straight trichomes, and are thus clearly different from B3W22^T.

DNA–DNA hybridization studies indicated that the relatedness between B3W22^T and Bacillus silvestris DSM 12223^T is only 29 % (reciprocal reaction 37 %). Furthermore, these two strains exhibit phenotypic and chemotaxonomic differences (Tables 3⇑ and 4⇑) that imply that B3W22^T represents a novel species; the name proposed is Bacillus isronensis sp. nov.

In the phylogenetic tree, strain B8W22^T forms a robust clade with its nearest phylogenetic neighbour Bacillus megaterium IAM 13418^T (Fig. 2⇑), with which it shows 99.7 % similarity at the 16S rRNA gene sequence level. Despite this high similarity, it is observed that, at the whole genome level, the DNA–DNA relatedness between B8W22 and Bacillus megaterium MTCC 428^T is only 35 % (reciprocal reaction 43 %), indicating that B8W22^T could be assigned to a novel species. Furthermore, B8W22^T and Bacillus megaterium MTCC 428^T differ in their phenotypic and chemotaxonomic characteristics (Tables 3⇑ and 4⇑), implying that B8W22^T does indeed represent a novel species, for which the name proposed is Bacillus aryabhattai sp. nov. Similarity of B8W22^T to other Bacillus species at the 16S rRNA gene sequence level was less than 95 %.

Sensitivity to UV radiation

Cultures were grown to an OD₆₆₀ of 1.0 at 30 °C and the sensitivity to UV radiation was determined by exposure to a UV lamp (UV-B, 15 W64; Sankyo Denki) as described previously (Shivaji et al., 2006). Cultures of Bacillus silvestris DSM 12223^T_, Bacillus megaterium MTCC 428^T, Micrococcus luteus MTCC 106^T, Janibacter limosus DSM 11140^T and Janibacter terrae DSM 13876^T were used as controls in these experiments. The results indicate that strains B3W22^T, B8W22^T and PVAS-1^T are more resistant to UV irradiation than their nearest phylogenetic neighbours (Supplementary Table S1, available in IJSEM Online).

Conclusions

The bacterial strains identified are unlikely to be laboratory contaminants, because no such cultures were handled in the laboratory. The control plates that were exposed to the air flow of the laminar flow hood during the entire air filtration procedure did not show any colonies. Furthermore, all possible precautions were taken to rule out contamination of the cryotubes throughout the assembly and deployment stages of the experiment; it is unlikely that they came from the cryotube assembly facility and survived the transit through the stratosphere, because the control cryotubes were completely sterile. It has been argued that micro-organisms can travel across interplanetary space during routine meteoritic exchanges between the Earth and Mars (Gladman et al., 1996; Mileikowsky et al. 2000; Nicholson et al., 2000). Furthermore, it is possible that micro-organisms reach the stratosphere from the Earth as a result of volcanic eruptions, the updraft caused by blue lightning strikes (Pasko et al., 2002), thunderstorms and forest fires (Fromm et al., 2004) and gravitophotophoresis (Rohatschek, 1996). Finally, in 1982, Hoyle and Wickramasinghe proposed the theory of ‘Panspermia’ (Hoyle & Wickramasinghe, 1986, 1993, 1999).

Description of Janibacter hoylei sp. nov.

Janibacter hoylei (hoy′le.i. N.L. gen. n. hoylei of Hoyle, named after Sir Fred Hoyle, the famous English astronomer).

Colonies on LB agar are creamish, entire, round and 1–2 mm in diameter. Cells stain Gram-positive and show oxidative metabolism; they are non-motile, non-endospore-forming cocci, 0.4–0.7 μm in diameter, that occur singly or in clumps. Good growth (visible colonies with a diameter of 1 mm) occurs after 2 days of incubation on nutrient agar at 25–30 °C. Grows in LB broth at 20–40 °C and at pH 5–10, with optimum growth at 30 °C and pH 9. Grows in the presence of 5 % (w/v) NaCl in LB broth and exhibits weak growth in LB broth containing 10 % NaCl. Resistant to UV radiation. Catalase- and oxidase-positive; H₂S is not produced. Voges–Proskauer negative, indole-negative and reduces nitrate to nitrite. Results of carbon source utilization tests are shown in Table 2⇑. Does not produce acid from alcohols or the following sugars: d-glucose, d-fructose, d-galactose, lactose, maltose, cellobiose, d-mannitol, d-mannose, raffinose, l-rhamnose, d-ribose, d-sorbitol, trehalose and d-xylose. meso-Diaminopimelic acid is the diagnostic diamino acid in the cell-wall peptidoglycan. The main fatty acids present are iso-C_16 : 0, 10-methyl C_17 : 0, C_18 : 1Δ9c and C_17 : 1Δ9c. The predominant isoprenoid quinone present is MK-8(H₄). In whole-cell hydrolysates, ribose and glucose are present, but arabinose, galactose and mannose are absent. Mycolic acids are absent. The DNA base composition of the type strain is 72.8 mol% G+C.

The type strain is PVAS-1^T (=MTCC 8307^T =DSM 21601^T =CCUG 56714^T), isolated from an air sample collected at an altitude of 40–41.4 km using a cryosampler.

Description of Bacillus isronensis sp. nov.

Bacillus isronensis (is.ro.nen′sis. N.L. masc. adj. isronensis arbitrary name pertaining to ISRO, the acronym of the Indian Space Research Organization, which largely funded the studies in which the type strain was isolated).

Colonies on nutrient agar are white, entire, round and 3–4 mm in diameter. Cells produce round terminal endospores and are motile. Grows at 5–37 °C and pH 6–10. Does not grow at 42 °C or at pH 4 or 11. Tolerates up to 5.8 % NaCl. Resistant to UV radiation. Grows on peptone. Positive for oxidase, lipase, gelatinase, starch hydrolysis, caseinase, tryptophan deamination, nitrate reduction and indole production. Produces acid from cellobiose and utilizes a number of sugars, amino acids and other carbon compounds as sole carbon sources (Table 3⇑). The type strain is sensitive to a number of antibiotics (Table 3⇑). The fatty acids present are C_15 : 0, C_16 : 0, C_17 : 0, iso-C_13 : 0, iso-C_15 : 0, anteiso-C_14 : 0, anteiso-C_16 : 0, C_16 : 1Δ9c, C_16 : 1Δ11 and C_18 : 1Δ9 (Table 4⇑). The interpeptide bridge of the cell-wall peptidoglycan consists of Lys–Glu–Ala. The lipids present are PG, PE, PS and an unidentified lipid. MK-6, MK-7 and MK-8 are the menaquinones present. The DNA G+C content of the type strain is 40 mol%.

The type strain is B3W22^T (=MTCC 7902^T =JCM 13838^T), isolated from a cryogenic tube used for collecting an air sample from an altitude of 27–30 km.

Description of Bacillus aryabhattai sp. nov.

Bacillus aryabhattai (a.ry.a.bhat′ta.i. N.L. gen. n. aryabhattai of Aryabhatta, named after the renowned Indian astronomer of the 5th century AD).

Colonies on nutrient agar are peach-coloured, entire, round, flat and 5–8 mm in diameter. Cells produce ellipsoidal and central endospores and are motile. Growth occurs at 10–37 °C and pH 6–10. Does not grow at 42 °C or at pH 4 or 11. Tolerates up to 11.6 % NaCl. Resistant to UV radiation. Grows on peptone. Positive for phosphatase, DNase, tryptophan deamination, ornithine decarboxylase and malonate utilization (Table 3⇑). Produces acid from various carbon sources and utilizes a number of sugars, amino acids and other carbon compounds as sole carbon sources (Table 3⇑). The type strain is sensitive to a number of antibiotics (Table 3⇑). The fatty acids present are C_14 : 0, C_15 : 0, C_18 : 0, iso-C_15 : 0, anteiso-C_14 : 0, anteiso-C_16 : 0 and C_16 : 1Δ9c (Table 4⇑). The interpeptide bridge of the cell-wall peptidoglycan consists of meso-diaminopimelic acid–Ala–Glu. The lipids present are PG and PE. MK-7 is the menaquinone present. The DNA G+C content of the type strain is 38 mol%.

The type strain is B8W22^T (=MTCC 7755^T=JCM 13839^T), isolated from a cryogenic tube used for collecting an air sample from an altitude of 40–41.4 km.