Strain PM2, a novel methylotrophic fluorescent Pseudomonas sp.

Abstract

A novel bacterial strain, PM2, capable of growing on methanol, was isolated in alkaline conditions from a soil inoculum. This bacterium was characterized at the physiological, biochemical and molecular level. Based on biochemical and molecular data strain PM2 was classified as a novel member of the group of fluorescent pseudomonads. Evidence for the presence of a pyrroloquinoline quinone (PQQ)-linked alcohol dehydrogenase in this organism is presented. Strain PM2 is, to our knowledge, the first example of a methylotrophic Pseudomonas to be characterized in detail. This novel type of metabolism in Pseudomonas broadens even further the metabolic versatility for which this genus is renowned.

Fluorescent, Methylotrophy, Alcohol dehydrogenase, Pseudomonas

1 Introduction

One-carbon (C1) compounds are widespread in the environment. Huge quantities of methane, methanesulfonic acid [1] and methanol [2] are constantly produced by natural biogeochemical reactions and have to be cycled constantly, thus it is no surprise to find that many organisms take advantage of this organic matter for their growth. Oxidative mesophilic methylotrophs are isolated quite easily from natural and man-controlled environments. As a matter of fact, methylotrophic organisms have been found in all major clades of microbial life: the best-studied examples are the Gram-negative methylotrophs, belonging to the α, β and γ subgroups of the Proteobacteria, that contain methanotrophic species [3]. However, methylotrophic species were also found within the Firmicutes (Gram-positives), the Archaea, and the yeasts. Methylotrophic Archaea and some Firmicutes anaerobically disproportionate C1 compounds to CO₂ and methane to generate energy. Oxidative methylotrophs oxidize C1 molecules to CO₂ to gain reducing equivalents. Reports of anaerobic (sulfate-respiring) methylotrophy have been put forward, at least on ecological grounds [4,5], but in most of the cases studied the final electron acceptor is oxygen. Despite the fact that O₂-dependent methylotrophs harbor diverse enzymatic systems for oxidation of C1 compounds, all converge in the production of formaldehyde, which can then be incorporated into cell carbon through several pathways.

Regarding methanol, one oxidation step is sufficient for the conversion to formaldehyde. This oxidation step is carried out by a dehydrogenase in the Gram-negative [6] and the aerobic Gram-positive methylotrophs [7,8], whereas yeasts utilize an alcohol oxidase [9]. Methanol dehydrogenase (MDH) of the Gram-negatives is a very well-characterized quinone-dependent periplasmic enzyme that is very conserved among the different species [10]. The corresponding enzyme in the Gram-positives is a very different complex protein that utilizes NAD(P)⁺ as cofactor [7].

Pseudomonas strains and methylotrophic organisms dwell in the same habitats and are frequently co-isolated. One example was recently provided by Pirttilä[11]: they isolated from Scotch pine buds a Methylobacterium strain and a Pseudomonas synxantha strain. Their assumption is that these microorganisms inhabit pine bud tissues. It is common knowledge that several species of fluorescent Pseudomonas can be found in association with plants. Plant material is also a typical source of methylotrophs and Methylobacteria in particular.

Many methylotrophic bacteria were originally classified within the Pseudomonas genus, but modern classification techniques such as 16S rRNA sequence or fatty acid analysis led to a revision of this situation [12,13]. In an extensive classification effort by Green and Bousfield [14] most of the putative methylotrophic Pseudomonas strains were reclassified under different genera: only three groups of strains had characteristics compatible with the genus Pseudomonas. Two of these strains were fluorescent (produced pyoverdine). To our knowledge (P.N. Green, personal communication) none of these strains was further studied, so very scant information is available on these organisms.

As a consequence, as far as we know, the present work is the first detailed description of a methylotrophic fluorescent Pseudomonas strain.

2 Materials and methods

2.1 Chemicals and media

Except where otherwise stated, all chemicals were of analytical grade and obtained from Aldrich, Sigma or Merck. Methane was used as a 50:48:2 air:CH₄:CO₂ mixture.

2.2 Enrichment, isolation and maintenance

Pre-enrichment was performed by maintaining a 50 g sample of orchard soil at 5°C for 6 months and spiking it every 2 weeks with 5 ml of a sterile 0.1% (v/v) solution of methanol. Enrichment of isolates was carried out at 26°C using 2 g of this pre-enriched soil and the adapted minimal medium MinE, as described in Kelly et al. [15], with the addition of 0.1% (v/v) methanol. For routine growth, methanol concentrations of 0.2 or 0.4% (v/v) were used. In some cases higher concentrations were employed as described. Formaldehyde was used at 2 mM. Sugars at 10 mM. All other carbon sources at a 20 mM concentration, except for sodium dodecyl sulfate (SDS), which was incorporated at 1 mM. When yeast extract (YE) was used for growth, a concentration of 0.5% (w/v) was employed. Buffering of the pH was obtained using TABS 10 mM pH 9.6, or phosphate 11 mM pH 6.8, as stated.

2.3 Physiological characterization

Growth rate and yield determinations were performed in at least three replicate cultures grown with shaking. The possibility of anaerobic growth was tested in sloppy agar columns (as described in [16]). Levels of resistance to antibiotics and metals were determined in agar. For antibiotics, susceptibility disks (Oxoid) were used according to the manufacturer's instructions. pH changes in the medium were detected by the addition of bromocresol purple to the agarized medium, or by direct pH measurement of the supernatants. The flagella Ryu stain was used according to Heinbrook et al. [17]. Growth temperatures were tested using MinE+YE 0.5% (w/v) agarized medium. King's media A and B were prepared using formulations from Pronadisa (Barcelona, Spain) according to the manufacturer's instructions. The arginine dihydrolase and gelatinase tests were performed according to Smibert and Krieg [18]. In all cases appropriate positive and negative controls were used for comparison.

2.4 Biochemical characterization

Quinone and fatty acid analyses were performed at DSMZ, Braunschweig, Germany (http://www.dsmz.de) by thin layer chromatography (TLC), diode array detector/high performance liquid chromatography (HPLC) and by gas chromatography, respectively. The guanosine+cytosine content (mol% G+C) of the genomic DNA was determined by the HPLC technique described in Mesbah et al. [19] at the BCCM/LMG Culture Collection Laboratories, University of Gent, Belgium. The oxidase test was performed with Difco oxidase stain dropper. The Gram stain and catalase detection were performed as previously described [18]. Hydroxypyruvate reductase (HPR) was assayed according to standard protocols [20]. MDH was assayed essentially as described in Day and Anthony [21], but the reaction-starting reagent was the substrate (and not polyethylene sulfonate (PES)).

2.5 Molecular techniques

Genomic DNA was extracted using the UltraClean™ microbial genomic DNA isolation kit, MOBIO Laboratories, CA, USA. Denaturing of DNA was performed by treatment with alkali: 0.25 vol of denaturing solution (NaOH 0.1 M, ethylenediamine tetraacetic acid (EDTA) 1 mM) were added to DNA suspensions; the mix was incubated at room temperature for 5 min and neutralized with the addition of 0.3 vol of Na acetate 3 M pH 4.5. Slot blotting was carried out using a Hybri-Slot™ apparatus (Life Technologies). The positive control probe was the polymerase chain reaction (PCR)-amplified 16S rRNA gene obtained from strain PM2. The mxaF probe was obtained from Methylobacterium extorquens AM1 genomic DNA using primers mxaF1003 and mxaF1561, as described in [10]. Probe labelling and hybridization were performed using the AlkPhos Direct kit by Amersham Pharmacia Biotech according to the manufacturer's instructions.

2.6 16S rRNA sequence analysis

The 16S rRNA gene of strain PM2 was amplified by PCR using the primers f27 and r1492 [22] under standard PCR conditions. The amplified fragments were cloned into the pGEM T-Easy vector (Promega) and sequenced using 16S-specific primers f27, f357 and r1492 [22]. BLAST analyses were performed through the National Center for Biotechnological Information online service (http://www.ncbi.nih.gov/BLAST). The 16S rRNA gene sequences were aligned using the BioEdit program (version 4.8.8) [23] and analyzed using the programs SEQBOOT (1000 iterations), DNADIST (Kimura 2-parameter), NEIGHBOR, and CONSENSE of the PHYLIP package [24]. An alignment of 19 sequences of 1343 nucleotides was used. The sequence of the near-complete 16S rRNA gene from strain PM2 was deposited in the GenBank databases under the Accession Number AY151820.

2.7 Protein analysis

Cell pellets from 50–500 ml cultures were resuspended in 0.5 ml sonication buffer [20] after which cell lysis was obtained using a Branson Sonifier 250 sonicator with four cycles of 30 s at 50% duty cycle intercalated with four cycles of 30 s off duty. The lysates were centrifuged at 16 000×g to eliminate particulate matter and the extracts were stored at −80°C until used. The protein content of the cell-free extracts was measured by standard methods [25]. Cell-free extracts were analyzed by SDS–polyacrylamide gel electrophoresis (PAGE) in 12.5% acrylamide gels according to standard procedures [26].

After electrophoresis, proteins were electroblotted onto a nitrocellulose membrane (Hybond-C, Amersham). The membrane was incubated with the primary antibody, specific for the large subunit of MDH (MxaF) (rabbit IgG anti-MxaF) at a 1:5000 dilution, and subsequently with the secondary antibody, goat anti-rabbit IgG linked to horseradish peroxidase (Sigma), at a 1:5000 dilution. Immunodetection was performed by chemiluminescence, using a kit from Amersham (RPN 2109). The membranes were exposed to a Hybond-electrochemoluminescent (ECL) film (Amersham) for 1–3 s.

3 Results and discussion

3.1 Enrichment and isolation

A pre-enrichment was performed exposing an orchard soil sample to methanol as described in Section 2. From this pre-enriched soil sample two bacterial strains, PM1 and PM2, were obtained by standard enrichment and purification techniques on minimal medium MinE with TABS 10 mM pH 9.6 as buffering agent and methanol 0.1% (v/v) as carbon source. Strain PM1 is a typical Methylobacterium, a pink-pigmented facultative methylotroph (PPFM). Strain PM2 is described below.

3.2 Characterization of strain PM2

Strain PM2 formed white, regular, smooth, mucous, 1 mm wide colonies on minimal medium agar plates. Cells of this bacterium appeared at microscopic inspection as motile rods, ca. 1×3–4 µm, showing one polar flagellum. When cultures grown in liquid minimal medium were exposed to ultraviolet (UV) light (254 nm), a feeble greenish fluorescence was observed in the culture, although this pigment was not detectable on solid medium. A strong yellow pigmentation was observed on King's media A and B. No pyocyanin (blue pigment) was observed on King's media.

Cells of this strain were Gram-negative. Strain PM2 had positive, although weak, catalase and oxidase reactions and tested positive in the arginine dihydrolase and gelatinase tests. Table 1 shows the carbon sources tested for this bacterium. It is worth noting the metabolic versatility of strain PM2, typical of the genus Pseudomonas: it was able to grow on a number of sugars and organic acids, glycerol, ethanol, benzoate, even SDS. Of the C1 compounds tested, growth could be observed only with methanol and formate, while none of the substituted amines or formaldehyde supported growth. By contrast, the addition of low concentration (2 mM) of formaldehyde to a culture growing on methanol provoked an inhibitory effect. Strain PM2 tolerated methanol concentrations up to 5% (v/v, corresponding to 1.23 M). In sloppy agar columns this bacterium was not capable of growth far from the agar surface, even in the presence of nitrate. It could grow rapidly between 8 and 34°C, but was inhibited at 37°C. Growth, although slow, occurred at temperatures as low as 0°C. In liquid medium pH 5 and 12 were inhibitory for growth of this organism, whereas pH 6.8 (phosphate) or 9.6 (TABS) supported growth.

1

Summary of the characteristics of strain PM2

Cell morphology	rods
Gram reaction	negative
DNA G+C content	59.2%
Catalase	±
Oxidase	±
Gelatinase	+
Arginine dihydrolase	+
C1 carbon sources	methanol, formate
Pigment	water-soluble greenish fluorescence
Molar yield on methanol	6.1 g mol−1
Max concentration of methanol tolerated	5% (v/v)
Sensitivity to antibiotic	Str, Kan, Gen, Tet
Resistance to antibiotic	Trp, Pen, Rif, Nal, Chl, Fus, Ery
Growth T range	0°C≤T≤34°C
Growth on	mannose, glucose, galactose, fructose, sucrose, sorbitol, mannitol, benzoate, acetate, citrate, succinate, glycerol, ethanol, 1-propanol, 2-propanol, acetone, SDS
No detectable growth on	methane, mono-, di- or tri-methylamine, methanesulfonate, methylsulfate, maltose, lactose, lactate

Str, streptomycin; Kan, kanamycin; Gen, gentamicin; Tet, tetracycline; Trp, trimethoprim; Pen, penicillin; Rif, rifampin; Nal, nalidixic acid; Chl, chloramphenicol; Fus, fusidic acid; Ery, erythromycin.

Strain PM2 grew very poorly on 1-propanol (0.15%, w/v) alone, but in the presence of YE 0.2% (w/v) and 1-propanol final OD was twice that with YE only.

This bacterium acidified the medium when utilizing methanol, whereas, in contrast, when formate was used as carbon source, the pH rose. Maximum observed growth rates were 0.149 h⁻¹ (T_d=6.7 h) on methanol, and 0.395 h⁻¹ (T_d=2.5 h) on YE. Average molar yield on methanol was 6.1±0.72 g mol⁻¹, a low value when compared to other aerobic methylotrophs (9–14 g mol⁻¹) [27]. Molar yield on formate was 3.32±0.47 g mol⁻¹.

Strain PM2 was resistant to trimethoprim, penicillin, rifampin, nalidixic acid, chloramphenicol, fusidic acid and erythromycin and sensitive to streptomycin, kanamycin, gentamicin and tetracycline. This strain was tolerant only to low concentrations of Cd (40 ppm) (in the form of CdCl₂), 1 ppm of Hg (as HgCl₂) and 1 ppm of Cr (K₂Cr₂O₇). On the other hand it showed resistance to the presence of high levels of As (3000 ppm) (KH₂AsO₄).

3.3 Classification of strain PM2

A near-complete 16S rRNA gene was amplified from the genomic DNA of strain PM2, cloned and sequenced. By BLAST searches and subsequent phylogenetic analysis (Fig. 1) it became evident that this strain's rRNA sequence clusters firmly within the fluorescent group of the Pseudomonas genus. The organisms most closely related to strain PM2 at the rRNA level were P. synxantha strain G (with 99.7% identity to strain PM2) [11], Pseudomonas sp. G2, a glyphosate-degrading strain (Chen, M. et al., unpublished) (99.7%), and Pseudomonas azotoformans (99.4%) (Anzai, Y. et al., unpublished). Unfortunately, the methylotrophic Pseudomonas strains described in Green and Bousfield [14] were never characterized at the level of 16S rRNA sequence and we are not able to assess their phylogenetic relationship with strain PM2. Genomic DNA of strain PM2 was analyzed and the percentage of G+C was determined as 59.2%.

1

Consensus tree of neighbor-joining analysis of 16S rRNA sequences. Numbers at branch nodes indicate the bootstrap values over 1000 replicates. Only values >700 are shown. Abbreviations: M.=Methylobacterium; R.=Rhizobium; P.=Pseudomonas.

The quinones present in strain PM2 could be identified as ubiquinone mainly of the Q-9 type. Small amounts (3.5%) of Q-8 could also be detected. The fatty acid pattern found is composed of unbranched saturated and unsaturated fatty acids, mainly 16:1 ω7c (palmitoleic acid), 16:0 (palmitic acid) and 18:1 ω7c (cis vaccenic acid). The diagnostic 2- and 3-hydroxy fatty acids with chain length of 10 and 12 carbons were also found. Based on these chemo-taxonomical data, strain PM2 was found most similar to Pseudomonas putida.

All this evidence together with the observation of the greenish fluorescence clearly endorses the classification of strain PM2 within the group of fluorescent Pseudomonas.

3.4 Catabolism of methanol by strain PM2

In order to understand how strain PM2 utilizes methanol, soluble protein extracts were obtained from this organism after growth on methanol, ethanol, 2-propanol or succinate and analyzed in SDS–PAGE. A polypeptide of slightly higher size than the landmark large subunit of MDH typical of all known Gram-negative methanol oxidizers (around 67 kDa in size) was greatly induced by growth on methanol, to a lesser extent by ethanol and to low levels by 2-propanol (Fig. 2). A Western blot of the gel shown in Fig. 2 was challenged with an anti-MxaF polyclonal antibody preparation. The result (Fig. 3) showed cross-hybridization of the ca. 70 kDa polypeptide of strain PM2 with the antibody. It also confirmed that the expression of the same polypeptide was strongly induced during growth on methanol and ethanol. This polypeptide was present at lower levels during growth on 2-propanol and some residual expression was detectable also on succinate.

2

SDS–PAGE: Lane M, molecular mass protein markers (Pharmacia); lane 1, strain PM2 after growth on methanol; lane 2, strain PM2 after growth on ethanol; lane 3, strain PM2 after growth on 2-propanol; lane 4, strain PM2 after growth on succinate; lane 5, M. extorquens strain AM1 after growth on methanol, for comparison.

3

Western blot of SDS–PAGE shown in Fig. 2 hybridized with anti-MxaF antibody. Lane M, molecular mass protein markers (Pharmacia); lane 1, strain PM2 after growth on methanol; lane 2, strain PM2 after growth on ethanol; lane 3, strain PM2 after growth on 2-propanol; lane 4, strain PM2 after growth on succinate; lane 5, M. extorquens strain AM1 after growth on methanol.

Soluble extracts of cells grown on methanol, formate, ethanol, 2-propanol or succinate were used to assay for the presence of MDH-like activity. We found this type of enzyme in alcohol-grown cell-free extract, although with much higher specific activity when the organism was grown on methanol or ethanol (see Table 2). This alcohol dehydrogenase was similar to the MDH of Methylobacterium[28] in that its substrate range was limited to primary alcohols.

2

Levels of alcohol dehydrogenase-specific activity in strain PM2 (nmol of methanol oxidized min⁻¹ mg protein⁻¹)

Growth substrate	Assay substrate
	Methanol	Ethanol	1-Propanol	1-Butanol	1-Pentanol
Methanol	330.6±66.4	291.1±53.8	198.5±63.3	110.1±22.5	136.8±42.5
Formate	nd	nd	NT	NT	NT
Ethanol	248.5±75.9	256.0±22.8	113.5±45.7	70.8±35.0	104.0±42.5
2-Propanol	116.6±64.2	107.9±14.1	24.2±5.6	NT	NT
Succinate	nd	nd	nd	NT	NT

nd, =none detected; NT, not tested.

We then performed PCR amplification of the genomic DNA from strain PM2 using primers designed previously [10] to amplify the widest known range of mxaF genes (coding for the large subunit of MDH). Although all positive controls (mxaF primers with M. extorquens AM1 genomic DNA and 16S rRNA-specific primers, with strain PM2 genomic DNA) amplified as expected, we obtained no amplification of the expected fragment (ca. 515 bp) of gene mxaF with strain PM2. In a further experiment the mxaF amplicon obtained with the same set of primers and M. extorquens AM1 genomic DNA was used as a probe in slot-blot hybridization against genomic DNA of strain PM2. The results were negative at a washing temperature of 65°C and a very weak signal was detectable at 55°C.

All these data taken together suggest that this bacterial strain utilizes a pyrroloquinoline quinone (PQQ)-linked alcohol dehydrogenase to oxidize methanol to formaldehyde, but also indicate that this enzyme is not so closely related to the well-known Gram-negative MDH. The 70 kDa polypeptide was expressed at very low levels during growth on 2-propanol. This result and the failure of the dehydrogenase to oxidize secondary alcohols in vitro support that the enzyme is not responsible for the catabolism of this substrate. However, the presence of the 70 kDa polypeptide also during growth on 2-propanol and succinate and of MDH-like activity in 2-propanol-grown cells, although at lower levels, suggest that, in strain PM2, this enzyme may be also used as a house-keeping, non-specific alcohol dehydrogenase.

It has been known for some time that P. aeruginosa ATCC 17933 employs a PQQ-linked periplasmic alcohol dehydrogenase to catabolize ethanol [29]. The sequence of the respective gene (exaA) and the structure of the dehydrogenase are known [30,31] and structural and biochemical similarities between ExaA and MDH have been recognized. Searching the publicly available genome of P. putida KT2440 (http://www.tigr.org) shows that also this species possesses a putative quinoprotein alcohol dehydrogenase gene. The identity between the exaA genes and Methylobacterium mxaF is very low (54–56%), while the identity/similarity values are definitely significant at the protein level (33–34% identity, 48–50% similarity). These figures could easily explain the apparent absence of mxaF gene in strain PM2 and the simultaneous presence of cross-hybridization in immunodetection. Our hypothesis is that strain PM2 utilizes a homolog of ExaA to catabolize methanol. A similar case of a non-methanol-specific alcohol dehydrogenase recruited to allow growth on methanol was previously reported in Thiosphaera panthotropha strain Tp9002 [32].

In order to determine whether strain PM2 utilizes the serine pathway to fix C1 carbon, we looked for the presence of the diagnostic enzyme HPR in methanol- and formate-grown cell extracts and found specific activities of 98.3 and 57.1 nmol min⁻¹ mg protein⁻¹, respectively. The presence of HPR activity indicates that C1 carbon is fixed to cell carbon via the serine cycle. However, both the levels of HPR and the molar yields on methanol are low and may suggest the utilization at the same time of a less efficient carbon fixation pathway. At present no direct evidence is available to determine whether the serine cycle is the only carbon-fixing pathway at work in this organism. It is worth noting that in the genomes of both P. aeruginosa PAO1 and P. putida KT2440 putative HPR genes (hyp) are present.

Methylotrophy within the genus Pseudomonas seems to be confined to a very limited number of strains. This work is the first report relating on the mechanism of methanol utilization in a true Pseudomonas. The recognition of this new type of metabolism broadens our knowledge and expectations on this genus already renowned for its wide metabolic versatility.

Acknowledgments

We wish to thank Ana Rita Figueredo for her valuable technical assistance. Our gratitude goes to Carlos Miguez for his kind gift of M. extorquens strain AM1 and to J. Colin Murrell, Don P. Kelly, Chris Anthony and Célia Manaia for their helpful suggestions. We are indebted to Hirohide Toyama, Yamaguchi University, Japan for the gift of the anti-MxaF antibody. We acknowledge the 5^th Framework European Union project FPQLK5-CT-200-01528 for financial support.