Genetic diversity of culturable bacteria in oil-contaminated rhizosphere of Galega orientalis

Environmental Pollution 139 (2006) 244e257 www.elsevier.com/locate/envpol

Genetic diversity of culturable bacteria in oil-contaminated rhizosphere of Galega orientalis Minna M. Jussila*, German Jurgens, Kristina Lindstro¨m, Leena Suominen Department of Applied Chemistry and Microbiology, Viikki Biocenter, PO Box 56, FI-00014, University of Helsinki, Finland Received 18 October 2004; accepted 13 May 2005

Bacterial diversity during rhizoremediation in oil-contaminated soil is characterized by a combination of molecular methods. Abstract A collection of 50 indigenous meta-toluate tolerating bacteria isolated from oil-contaminated rhizosphere of Galega orientalis on selective medium was characterized and identified by classical and molecular methods. 16S rDNA partial sequencing showed the presence of five major lineages of the Bacteria domain. Gram-positive Rhodococcus, Bacillus and Arthrobacter and gram-negative Pseudomonas were the most abundant genera. Only one-fifth of the strains that tolerated m-toluate also degraded m-toluate. The inoculum Pseudomonas putida PaW85 was not found in the rhizosphere samples. The ability to degrade m-toluate by the TOL plasmid was detected only in species of the genus Pseudomonas. However, a few Rhodococcus erythropolis strains were found which were able to degrade m-toluate. A new finding was that Pseudomonas migulae strains and a few P. oryzihabitans strains were able to grow on m-toluate and most likely contained the TOL plasmid. Because strain specific differences in degradation abilities were found for P. oryzihabitans, separation at the strain level was important. For strain specific separation (GTG)5 fingerprinting was the best method. A combination of the single locus ribotyping and the whole genomic fingerprinting techniques with the selective partial sequencing formed a practical molecular toolbox for studying genetic diversity of culturable bacteria in oil-contaminated rhizosphere. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Oil contamination; Rhizosphere; Bacterial diversity; Molecular identification; TOL plasmid

1. Introduction The perennial forage legume of Galega orientalis (goat’s rue), with its nitrogen-fixing symbiont Rhizobium galegae H1174, and various other rhizosphere bacteria have good potential for rhizoremediation of oil-contaminated soil (Suominen et al., 2000), since plants enhance the bioremediation of oil-contaminated soils by

* Corresponding author. Tel. C358 9 191 59279; fax: C358 9 191 59322. E-mail address: minna.m.jussila@helsinki.fi (M.M. Jussila). 0269-7491/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2005.05.013

stimulating microbial degradation activity in the rhizosphere (Radwan et al., 1995). Not only do the plant roots supply nutrients such as amino acids, carbohydrates and organic acids for rhizosphere bacteria (Anderson et al., 1993), but they may also help bacteria degrade toxic organic chemicals by releasing phospholipid surfactants that modify the physical and chemical properties of the rhizosphere and thus the bioavailability of organic pollutants (Read et al., 2003). Root exudates may create favourable conditions also for co-metabolism. In addition to plant surfactants, Nielsen and Sørensen (2003) have shown that many Pseudomonas fluorescens strains produce different cyclic lipopeptides

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

with antifungal and biosurfactant properties in specific habitats like the rhizosphere. Pseudomonas putida PaW85 contains an archetypal and conjugative TOL plasmid called pWW0 (Bayley et al., 1977; Williams and Murray, 1974). Self-transmissible TOL plasmids have the potential to spread to other bacteria in the soil. The xyl genes located on TOL plasmids code for enzymes responsible for the degradation of monoaromatic oil compounds, BTEX (benzene, toluene, ethyl benzene and xylenes) (Assinder and Williams, 1990). Petrol stations and garages are typical sources of these highly soluble oil compounds. The metacleavage pathway on TOL plasmids differs significantly from the chromosomal b-ketoadipate or ortho-cleavage pathway in being able to tolerate alkyl substituents on the catechol. This enables TOL-harbouring bacteria to utilize also e.g. meta-toluate (3-methylbenzoate), our model compound. One of the key enzymes in the meta-cleavage pathway is the xylE gene product called catechol 2,3dioxygenase, the substrate of which is 3-methylcatechol. The breakdown of catechol can be detected as a yellow product called 2-hydroxymuconic semialdehyde. PCR-based DNA-typing methods of culturable bacteria are nowadays universally applicable, simple and rapid. However, oil-degrading bacteria have not yet been studied systematically with these methods. Besides the knowledge of total bacterial communities, we still need information on individual culturable bacteria in order to combine different metabolic functions to relevant bacterial species. Restriction fragment length polymorphism analysis (RFLP) of amplified 16S rRNA genes provides an estimate of the phylogenetic relationships between bacteria at species and higher taxonomic levels (Gurtler et al., 1991; Laguerre et al., 1994), while repetitive sequence-based polymerase chain reaction (rep-PCR) genomic fingerprinting is used to produce species and strain specific fingerprints of different bacterial genomes (Laguerre et al., 1996; Nick et al., 1999; Versalovic et al., 1991, 1994). DNA primers corresponding to repetitive extragenic palindromic (REP), enterobacterial repetitive intergenic consensus (ERIC) and a subunit of the BOX element (BOXA) sequences as well as to a repetitive trinucleotide ((GTG)5) are available for this purpose. REP and ERIC have been the most popular primers thus far (de Bruijn, 1992; Judd et al., 1993; Louws et al., 1994). Most of the studies using these DNA-fingerprinting methods have focused on one species or genus at a time. We, however, wanted to know, how molecular identification methods could be used for parallel identification of several bacterial genera from the complex environment of oilcontaminated rhizosphere. In the present study we applied molecular techniques to settle a compact set of methods suitable for monitoring the diversity of culturable bacteria during in situ bioremediation, and to gain genetic information about

245

potential oil degraders in the contaminated rhizosphere. The aims of the present study were (1) to establish a collection of bacteria from oil-contaminated rhizosphere with the capacity to grow on a medium containing m-toluate, (2) to phenotypically characterize the culture collection by determining the tolerance of the isolates to m-toluate and their capacity to degrade m-toluate, (3) to taxonomically characterize the collection of indigenous m-toluate-tolerating isolates by using 16S rDNA PCRRFLP, rep-PCR genomic fingerprinting and partial 16S rRNA gene sequencing.

2. Materials and methods 2.1. Microcosms and the harvest of rhizosphere soil samples Microcosms with oil- (OS) or m-toluate-contaminated soil (MS) were set up. Soil types were classified according to Elonen (1971) and basic soil characteristics were analysed by Viljavuuspalvelu Ltd (Mikkeli, Finland). Oil-contaminated, dry and humous sandy soil (4% organic C; pH 7.3; conductance 10!mS/cm 10.0; levels of exchangeable nutrients mg/l: Ca, 9350; P, 17; K, 122; Mg, 574; S, 1820; total N, 0.17) was from a 20-year-old land-farming field for oil refinery wastes in southern Finland. This oil soil contained 10% total hydrocarbons (THC) as measured gravimetrically by carbon tetrachloride extraction (SFS 3009, 1980). The original oil soil was diluted with fine sand to a 2% final THC concentration. The concentrations of calcium and sulphur were very high in the OS. Metal concentrations (mg/kg DW; As, 53; Hg, 0.67; Cd, 0.45; Cr, 94; Cu, 123; Pb, 40; Ni, 145; Zn, 312; V, 233) were determined from the oil soil by Neste, Corporate Technology, Analytical Research (Porvoo, Finland). Due to the high metal concentration, the OS had high conductance. The concentrations of arsenic, copper, mercury, nickel, zinc and vanadium exceeded maximum values for agricultural use. MS was prepared from agricultural soil by adding 10% m-toluate (pH 13) to get the final m-toluate concentration of 3000 mg/l of soil. Agricultural soil was from the Partala Research Station for Ecological Agriculture (MTT, Juva, Finland) (4% organic C; pH 6.3; conductance 10!mS/cm 2.5; levels of exchangeable nutrients mg/l: Ca, 1920; P, 2.6; K, 133; Mg, 195; S, 42.3; total N, 0.18). This soil was moist and humous fine sand. It contained no indigenous Rhizobium galegae bacteria. Plants in OS or MS were treated in two different ways and grown in an open greenhouse at an average summer temperature of 16  C in Helsinki: Galega orientalis seeds inoculated with R. galegae HI174 were grown with or without the presence of a peat layer inoculated with Pseudomonas putida PaW85. Peat-based P. putida PaW85 inoculants were prepared as described by Suominen et al.

246

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

(2000). A thin P. putida PaW85 peat layer (1.5 g; 109 CFU/g) was added between the contaminated soil layer and a clean soil top layer. G. orientalis seeds were surface sterilized, peat-inoculated (105 CFU R. galegae/seed) according to Elomestari Ltd (Juva, Finland), and planted (20 seeds/pot) into the 10 cm clean soil top layer of 9-l microcosm pots. Fertilizer was applied and the plants were thinned out to the final number of four per pot as described by Suominen et al. (2000). After a 4-month summer season the plant roots had spread well to the contaminated soils. Rhizosphere samples were harvested from soil adhering to roots. Samples from six replicates per treatment were combined. The remaining samples were frozen (ÿ20  C).

2.2. Phenotypic screening of the culture collection 2.2.1. Isolation, purification and cultivation of rhizosphere bacteria The bacterial strains isolated and used in this work are listed in Table 1. Bacteria were extracted from the rhizosphere of G. orientalis by blending 1 g wet wt of the rhizosphere soil with 9 ml of Tween buffer (100 ml of 1% Tween80 in 0.9% NaCl) for 1 min before serial dilution. Bacterial strains were grown and isolated on DEF8based agar (Lindstro¨m and Lehtoma¨ki, 1988), supplemented with cycloheximide, CHX (250 mg/ml) to inhibit fungal growth, for 2e3 days at C28  C. Glucose (1 g/l), mannitol (1 g/l) or rhizosphere extract (100 ml/l) were used alone or in different combinations as a carbon source. Rhizosphere extract was prepared by mixing (1:5) rhizosphere soil (wet wt) with 0.1 M Na-phosphate buffer (pH 6.8). The mixture was stirred for 30 min, centrifuged for 10 min and filter sterilized (diameter 0.2 mm). Rhizosphere extract was stored at ÿ20  C until used. For some isolations soil extract agar (SEA) containing (per litre; pH 6.8) soil extract 100 ml, yeast extract 5 g, K2HPO4 0.2 g, glucose 1 g, agar 15 g, and CHX 250 mg was used. Soil extract was prepared like rhizosphere extract, but the mixing relation of the soil and the buffer was 1:1. The selective agent in all the media was m-toluate (from 1 to 7 g/l). Because some of the isolates grew as tight complexes consisting of two to three different strains, they had to be diluted in Tweenphosphate buffer (per litre, pH 6.5; K2HPO4$3H2O 3.67 g, KH2PO4 4.62 g, Tween80 1 ml) and subcultured at least three times on the same media. Gram staining was performed to ensure the purity of the isolates. Isolated strains were maintained at C4  C on tryptone yeast extract agar (TY) (Beringer, 1974) supplemented with 250 mg/l m-toluate and 250 mg/ml CHX. R. galegae strains were maintained on yeast extract mannitol agar (YEM) (Lindstro¨m et al., 1983) supplemented with Congo red (Merck; 25 mg/ml). Cells for DNA isolation were grown in TY broth with 250 mg/l m-toluate for

1 to 2 days at C28  C. R. galegae strains were grown to saturation in YEM broth for 2 days at C28  C. The strains were stored in TY or YEM broth with 15% (v/v) glycerol at ÿ70  C. For long time storage the strains were freeze-dried and deposited in the HAMBI culture collection. 2.2.2. meta-Toluate utilization/tolerance test The ability of the isolates to utilize different concentrations of m-toluate as the sole carbon and energy source was tested on DEF8 agar (degradation test). The ability of the colonies to grow in the presence of varying m-toluate concentrations was tested on TY agar (tolerance test). Overnight cultures (10 ml) were dropped upon the test agars containing up to 15 g/l m-toluate at intervals of 1 g/l. After 3 days the maximal growth was visually assessed. 2.2.3. Indirect assay for TOL plasmids: the catechol spray test The existence of the catabolic key enzyme catechol 2,3-dioxygenase, encoded by TOL plasmids, was proven indirectly by dropping 500 mM catechol (Sigma) upon bacterial colonies grown on TY agar. Positive colonies turned yellow within a few seconds due to the conversion of catechol to 2-hydroxymuconic semialdehyde (Williams et al., 1988). 2.3. DNA isolation of rhizosphere bacteria Cells grown in the presence of m-toluate were washed with the corresponding, unselective growth broth. Total genomic DNA was isolated from 1.5 ml pure cultures by the hexadecyltrimethyl ammonium bromide (CTAB) method (Wilson, 1989). RNA was removed by DNasefree RNase (Boehringer Mannheim; 5 mg/ml) according to the manufacturer’s instructions. The DNA pellet was resuspended in 40 ml 0.1! TE buffer and stored at ÿ20  C. For some gram-positive strains the method using diatomaceous earth (Celite, analytical filter aid, BDH Laboratory Supplies, Poole, England) as a DNAbinding solid support was used. In this method, nucleic acids were purified directly from the lysate according to Boom et al. (1990) and Heyd and Diehl (1996) with the following modifications: 100 ml of cells in deionised distilled water (DDW) were suspended in 700 ml GTC buffer (7 M guanidiumeHCl, 100 mM Tris pH 7.4, 10 mM CDTA). Two hundred microlitres of pumic stone slurry (Celite in DDW 1:1) was added and the mixture was strongly vortexed for 15 min. After brief spinning the pellet was washed once with GTC buffer and twice with washing buffer (40 mM Tris pH 7.4, 5 mM EDTA, 400 mM NaCl, 55% ethanol). The pellet was then washed with 70% ethanol and dried. Nucleic acids were

Table 1 Characteristics of the bacterial strains selected with m-toluate from the oil-contaminated rhizosphere of Galega orientalis HAMBI, Acc. noa (isolate)

Gramreaction/ phylogen. groupd

Tolerance to m-tot (g/l)e

m-tot as the sole C source (g/l)e

m-totgroupf

16Sribotypeg

(GTG)5geno typeh

The most similar species based on 16S rDNA sequences, Acc. no.i

OS/GR OS/GR OS/GR OS/GR C P MS/GR MS/GRC P OS/GR C P OS/GR OS/GR OS/GR C P OS/GR C P

DEF/RE/1 DEF/glu/5 DEF/glu/5 DEF/RE/4 DEF/RECman/2 DEF/RE/7 DEF/RECman/2 DEF/glu/5 DEF/glu/5 DEF/glu/5 DEF/RECglu/2

e/g e/g e/g e/g e/g e/g e/g e/g e/g e/g e/g

13B 11B 13B 11B 13 13 13 11B 13B 11B 11

7B 0 0 0 3; 13B 13 4; 13B 0 7B 0 8

2a 3a 3a 3a 1c, CC 1a, CC 1c, CC 3a 2a 3a 1b, CC

1A 1A 1A 1A 1A 1A 1A 1C 1C 2A 2B

7 7 7 7 7 7 2 7 7 10 5

Pseudomonas NS NS NS Pseudomonas Pseudomonas Pseudomonas NS Pseudomonas Pseudomonas Pseudomonas

H2401, AF501346 (BN4c) H2402, AF501340 (AN5a) H2405, AF501343 (AN6bl) H2615 (AN6b2)

OS/GR C P OS/GR C P OS/GR C P OS/GR C P

SEA/SECglu/5 SEA/SECglu/5 SEA/SECglu/5 SEA/SECglu/5

e/a e/a e/a e/a

6 8 5 5

0 0 0 0

3b 3b 3c 3c

3 4 5 5

3 1 11 11

H2380, AF501365 (5a) H2384, AF501357 (8a)

OS/GR C P OS/GR C P

DEF/RE/2 DEF/RE/3

e/b e/b

4 3

0 0

3c 3c

7A 7B

9 8

H2378, AF501342 H2616 (2) H2617 (3a) H2382, AF501364 H2618 (7a1) H2387 (13a) H2619 (22a) H2620 (22b) H2392, AF501355 H2621 (22c) H2393, AF501338 H2399, AF501353 H2404, AF501344 H2403, AF501345 H2390, AF501339 H2385, AF501363 H2622 (4a) H2379, AF501359 H2381, AF501356 H2623 (5d1) H2372, AF501358 H2383, AF501360 H2624 (7b4) H2373, AF501335

OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR OS/GR C P OS/GR C P OS/GR C P OS/GR OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR C P OS/GR C P

DEF/glu/5 DEF/glu/5 DEF/glu/5 DEF/RE/5 DEF/RE/1 DEF/RE/7 DEF/RE/4 DEF/RE/4 DEF/RE/4 DEF/RE/4 DEF/glu/5 SEA/SECglu/5 SEA/SECglu/5 SEA/SECglu/5 DEF/RE/1 DEF/RE/7 DEF/glu/5 DEF/glu/5 DEF/RE/2 DEF/RE/2 DEF/RE/3 DEF/RE/1 DEF/RE/1 DEF/RE/3

C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC C/hGC

11 11 11 11 9 13 ND ND 11 ND 13 11 11 11 11B 13 13 13 13 13 13 13 13 11

0 0 0 1 0 0 ND ND 0 ND 0 0 5 0 0 0 0 0 0 0 0 0 0 0

3a 3a 3a 2b 3a 3a ND ND 3a ND 3a 3a 2b 3a 3a 3a 3a 3a 3a 3a 3a 3a 3a 3a

8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 11A 11A 11A 11A 11A 11B 11B 11C

14 14 14 14 14 14 14 14 14 14 14 14 14 15 16 17 21 21 21 21 21 21 21 19

H2386, AF501362 H2611 (21a) H2612 (21b) H2613 (23a) H2396, AF501350 H2397, AF501349 H2374, AF501336 H2614 (11) H2388, AF501341 H2391, AF501361 H2394, AF501352

(12)

(27b) (29) (19a) (14) (20) (26a)

(1)

(6a)

(22B1) (25a) (N1b) (AN6a) (BN5a) (16a) (10) (4b) (5b) (8b13a) (7b3) (9a)

ID%j

oryzihabitans, D84004

99.3%

oryzihabitans, D84004 oryzihabitans, D84004 oryzihabitans, D84004

99.2% 99.3% 99.4%

oryzihabitans, D84004 jessenii, AF068259 migulae, AF074383

99.2% 99.0% 99.1%

Devosia riboflavina, D49423 Ochrobactrum anthropi, AJ242580 Agrobacterium tumefaciens, M11223 NS

99.0% 98.4% 99.8%

Ralstonia eutropha, AB015605 Ralstonia eutropha, AB015605

97.6% 97.6%

Rhodococcus erythropolis, AJ131637 NS NS Rhodococcus erythropolis, AJ131637 NS NS NS NS Rhodococcus erythropolis, AJ011328 NS Nocardia calcarea, AB037105 Rhodococcus erythropolis, AJ131637 Rhodococcus erythropolis, AJ131637 Rhodococcus erythreus, X79289 Nocardia calcarea, AB037105 Rhodococcus erythropolis, AJ131637 NS Arthrobacter histidinolovorans, X83406 Arthrobacter histidinolovorans, X83406 NS Arthrobacter histidinolovorans, X83406 Arthrobacter histidinolovorans, X83406 NS Arthrobacter aurescens, X83405

100%

100%

100% 99.7% 99.9% 99.8% 99.8% 99.9% 99.5% 99.4% 99.4% 99.4% 99.4% 99.4%

(continued on next page)

247

Isolation medium/ C source/ m-tot (g/l)c

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

Soil/ treatmentb

248

Table 1 (continued ) Soil/ treatmentb

Isolation medium/ C source/ m-tot (g/l)c

Gramreaction/ phylogen. groupd

Tolerance to m-tot (g/l)e

m-tot as the sole C source (g/l)e

m-totgroupf

16Sribotypeg

(GTG)5geno typeh

The most similar species based on 16S rDNA sequences, Acc. no.i

ID%j

H2407, AF501368 (N1a1d) H2408, AF501366 (AN4d)

OS/GR C P OS/GR C P

SEA/SECglu/5 SEA/SECglu/5

C/hGC C/hGC

ND ND

ND ND

ND ND

12 13

23 18

Kocuria kristinae, X80749 Micrococcus luteus, AJ276811

99.5% 99.6%

H2625 (N1a1b) H2406, AF501367 H2395, AF501351 H2398, AF501354 H2375, AF501337 H2400, AF501347 H2389, AF501348

OS/GR C P OS/GR C P MS/GR OS/GR C P OS/GR C P OS/GR C P OS/GR

SEA/SECglu/5 SEA/SECglu/5 DEF/RECman/2 SEA/SECglu/5 SEA/SECglu/5 SEA/SECglu/5 DEF/RECman/2

C/1GC C/1GC C/1GC C/1GC C/1GC C/1GC C/1GC

ND ND 11 11 9 9 6

ND ND 0 0 0 0 0

ND ND 3a 3a 3a 3a 3b

9A 9A 9B 9C 9C 9C 10

20 20 22 12 12 12 13

NS Bacillus Bacillus Bacillus Bacillus Bacillus Bacillus

99.1% 98.8% 99.9% 98.8% 99.9% 99.9%

OS/GR C P OS/GR

DEF/5 YEM/man/0

e/g e/a

13 4

13 0

1a, CC 3c

1B 6

4 6

(N1a1c) (27a1) (N1a1a) (AN4a) (BN4a) (15)

Reference strains H1828, L28676 (I2) H1174, Y12355 (I1) a

macroides, X70312 macroides, X70312 macroides, AF157696 macroides, X70312 macroides, AF157696 thuringiensis, AF155954

Pseudomonas putida PaW85, L28676 Rhizobium galegae, Y12355

100% 100%

HAMBI, the Microbial Culture Collection at the Division of Microbiology, Department of Applied Chemistry and Microbiology, University of Helsinki, Finland; Acc. no, the number of the 16S rDNA sequence in DDBJ/EMBL/GenBank nucleotide sequence databases. b OS, oil soil, MS, m-toluate soil; GR, Galega orientalis inoculated with Rhizobium galegae; CP, Pseudomonas putida PaW85/pWWO inoculant in the peat layer. c DEF, defined medium DEF No 8, YEM, yeast extract-mannitol medium with streptomycin (500 mg/ml); SEA, soil extract agar; glu, glucose; man, mannitol; RE, rhizosphere extract, SE, soil extract. d Phylogenetic groups: a, b, g, a-, b- and g-subdivision of Proteobacteria, respectively, hGC and lGC, gram-positive bacteria with high and low G C C content, respectively. e Observation limit for m-toluate tolerance and utilization was 1 g/l, B, brown pigment, ND, no data. f Grouping of rhizosphere isolates according to their ability to degrade/tolerate m-toluate: 1, m-toluate degraders which were catechol-positive (CC), and used m-toluate excellently (a, 13 g/l), well (b, 8 g/l) or moderately (c, up to 4 g/l), 2, m-toluate degraders which were catechol-negative, and used m-toluate well (a, 7 g/l) or moderately (b, up to 5 g/l), 3, m-toluate tolerants which were catechol-negative, and tolerated m-toluate excellently (a, up to 13 g/l), well (b, up to 8 g/l) or moderately (c, up to 5 g/l). g Rhizosphere isolates were grouped according to the similarity (0.7) of their restriction patterns into 13 16S rDNA PCR-RFLP ribotypes in computer-assisted analysis combining restriction enzymes AluI, MspI and RsaI, AeC, subdivision was based on the visual detection of profiles of 16S rDNA PCR-RFLP analysis with ten restriction enzymes (AluI, MseI, MspI, RsaI, CfoI, DdeI, HaeIII, HinfI, MboI and TaqI). h Rhizosphere isolates were grouped according to the similarity (0.7) of their genomic fingerprinting patterns into 23 (GTG)5-genotypes in computer-assisted analysis. i 16S rDNA nucleotide sequences were identified by WWW/BLASTN 2.1.1 using DDBJ/EMBL/GenBank nucleotide sequence databases, NS, not sequenced. j Identification was based on partial sequencing of 16S rRNA genes.

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

HAMBI, Acc. noa (isolate)

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

eluted with 100 ml of 10 mM TriseHCl (pH 9). After heating the suspension at C55  C for 2 min, vortexing and brief spinning, the supernatant containing DNA was ready for use. Standard techniques were used for estimating the DNA concentration in agarose gels (Sambrook et al., 1989).

2.4. Genomic fingerprinting by rep-PCR The oligonucleotide primers used for genomic fingerprinting by rep-PCR were REP1R-I and REP2-I, ERIC1R and ERIC2, BOXA1R, and (GTG)5 (Versalovic et al., 1991, 1994) synthesized by the Institute of Biotechnology, University of Helsinki. PCR amplifications were carried out in 25 ml reaction volumes containing 1! Dynazyme reaction buffer (10 mM TriseHCl pH 8.8, 6 mM MgCl2, 50 mM KCl, 0.1% Triton X-100, 10% (v/v) DMSO, 160 mg/ml BSA) modified from Versalovic et al. (1994) and Nick et al. (1999), 1 mM deoxynucleoside triphosphates (Pharmacia Biotech), 35 pmol REP/ 30 pmol ERIC/10 pmol BOX or 20 pmol (GTG)5 primers, 2 U DynazymeÔ II DNA polymerase (Finnzymes) and 50 ng template DNA. The amplifications were performed with a PTC-200 Peltier Thermal Cycler (MJ Research) with the temperature profile as slightly modified from Versalovic et al. (1991) by de Bruijn (1992). For the REP primers the cycles used were: an initial denaturation at 95  C for 6 min followed by 30 cycles of a three-step PCR program (94  C for 1 min, 40  C for 1 min and 65  C for 8 min) and a final extension at 65  C for 16 min. For the ERIC, BOX and GTG primers, the first denaturation step at 95  C lasted for 7 min and the annealing temperature was kept at 52  C. The reproducibility of the results was confirmed by repeating the amplifications twice in individual PCR runs. Ten microlitres of each PCR reaction was electrophoresed on 1.5% agarose gels for 5 h 45 min at 65 V. The fingerprints were recorded on a UV transilluminator provided with a video camera (Panasonic CCTV) connected to a computer including the SMViewII software. The patterns were printed on thermal paper with Video Copy Processor P68E (Mitsubishi) or the gels were photographed to Polaroid Type 55 positive/ negative film. The REP-, ERIC-, BOX- and GTG-PCR profiles were analysed separately and in different combinations as combined gels with the GelCompar 4.1 software (Applied Maths, Kortrijk, Belgium). The similarities between stored pairs of patterns were calculated by both the Pearson product-moment correlation coefficient and the Dice band-matching coefficient and the groupings were visualized as a dendrogram by the UPGMA clustering algorithm. In construction of the final (GTG)5-PCR dendrogram, GelCompar 4.1 was used with Pearson product-moment correlation

249

coefficient (Opt fine, 1.5%) and UPGMA clustering method. The reproducibility of the GelCompar analysis was evaluated and the threshold values for different bacterial groups were estimated by making cluster analyses of the size standards and the control strains from the same and different agarose gels. As an indication of the quality of a cluster analysis, the value of cophenetic correlation (c.c.) was calculated for each constructed tree. The c.c. value describes the productmoment correlation between all original matrix similarities and all corresponding similarity values derived from the dendrograms. If the UPGMA dendrogram faithfully represents the similarity matrix, the c.c. value will be high (more than 90%). 2.5. RFLP analysis of PCR-amplified 16S rRNA genes Primers fD1 and rD1 described by Weisburg et al. (1991) were used for PCR amplification of 1500-bp sequences of genes coding for 16S rRNA. The 50-ml PCR mixture contained 1! reaction buffer Dynazyme (Finnzymes), 200 mM dNTPs (Pharmacia Biotech), 0.1 mM (each) primers, 1 U of DynazymeÔ II DNA polymerase (Finnzymes) and 25 ng of bacterial DNA as template. Alternatively, 1.5 U of Red Hot DNA polymerase (Advanced Biotechnologies) with 1! reaction buffer IV and 2 mM MgCl2 was used. Amplification was performed in a PTC-200 Peltier Thermal Cycler (MJ Research) with the temperature profile modified from Laguerre et al. (1994): An initial 3-min denaturation step at 95  C was followed by 30 cycles of a three-step PCR program (94  C for 1 min, 55  C for 1 min and 72  C for 1 min) and a final 3-min extension step at 72  C. Amplified DNA was examined by electrophoresis in 1% agarose gel with 5-ml aliquots of PCR product. Aliquots (9 ml) of PCR-amplified 16S rDNA were digested with restriction endonucleases (5 U/reaction) (Laguerre et al., 1994): AluI, MseI, MspI, RsaI (MBI Fermentas), CfoI, DdeI, HaeIII, HinfI, MboI and TaqI (Promega). Restricted DNA was analysed in 4% agarose gels. After electrophoresis at 200 V for 2.5 h, the gels were recorded, printed and photographed as previously described. For the construction of the dendrogram GelCompar 4.1 was used with Pearson product-moment correlation coefficient (Opt fine, 2%) and UPGMA clustering method. 2.6. Molecular identification by partial 16S ribosomal DNA sequencing The 16S rRNA genes were partially (850e1000 bp) sequenced. The amplification with primers pA and pF# (25 pmol each) (Edwards et al., 1989) was performed as

250

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

described for 16S rDNA PCR-RFLP. The thermal cycler was programmed as described previously (KoukilaKa¨hko¨la¨ et al., 1995). The amplification products were purified for sequencing with MicroSpin S-400 HR columns (Pharmacia) and sequenced in both directions with primers pD# and pF# (Edwards et al., 1989). Sequencing was done by using a Dye Terminator Cycle Sequencing kit and an ABI 377 DNA Sequencer (Applied Biosystems) by the Institute of Biotechnology, University of Helsinki. 16S rDNA nucleotide sequences were identified by WWW/BLASTN 2.1.1 (8/2000) (http:// www.ncbi.nlm.nih.gov/blast) (Altschul et al., 1990, 1997) using DDBJ/EMBL/GenBank nucleotide sequence databases. Thirty-four 16S rRNA gene nucleotide sequences obtained in this study were deposited in the DDBJ/ EMBL/GenBank sequence databases under the accession numbers AF501335eAF501368 (Table 1). 2.7. Phylogenetic analysis The phylogenetic analysis of partial 16S rDNA sequences was performed using the ARB software package (http://www.arb-home.de) (Ludwig et al., 2004). The ARB FastAligner utility was used for automatic sequence alignment and the resulting alignments were manually verified against known secondary structure regions. A phylogenetic tree was inferred by performing neighbour-joining analysis with Jukese Cantor corrected distances.

3. Results and discussion 3.1. Identification and phylogenetic analysis of the isolates from oily rhizosphere Fifty randomly selected strains, 17 gram-negative and 33 gram-positive, were included in a collection of indigenous m-toluate-tolerating bacteria from the oilcontaminated rhizosphere of G. orientalis (Table 1). Thirty-four isolates, representing different branches in 16S rDNA PCR-RFLP and (GTG)5-PCR fingerprinting analyses, were chosen for partial sequencing of PCRamplified 16S rRNA genes. The nucleotide identity percentages of individual isolates to the closest validity identified phylogenetic neighbour in the DDBJ/EMBL/ GenBank databases as compared by partial 16S rRNA gene sequences are shown in Table 1. Most of the strains matched the nucleotide sequences found in the databases very well (99.0e100%). However, two strains obtained lower matches at 98.4 and 97.6% homology to Ochrobactrum anthropi and Ralstonia eutropha, respectively. Some of our 16S rDNA sequences were similar to database sequences of strains found in the rhizosphere of other plants than G. orientalis: Pseudomonas sp. with rape, Devosia sp. with potato and Ochrobactrum sp. with

wheat (Table 1; http://www.ncbi.nlm.nih.gov/blast). Many of our isolates were similar to strains detected in connection with degradation of various compounds: limonene which is a monoterpene from citrus and can be used to replace petroleum-based solvents like toluene and xylene (Rhodococcus; H2378, H2382, H2385, H2399, H2404), chlorobenzene (Rhodococcus; H2392, H2403), trichloroethylene (Ralstonia; H2380, H2384), 2,4-dinitrophenol (Nocardia; H2390, H2393), and resin-acid degrading psychrophiles (Pseudomonas; H2394) obtained from arctic soils contaminated with hydrocarbons. Our results showed high phylogenetic diversity in the form of five major lineages of the Bacteria domain (Fig. 1.): 26 isolates were placed in gram-positive bacteria with high DNA G C C content and seven in grampositive bacteria with low DNA G C C content, 12 in the g-subdivision of Proteobacteria, five in the a-subdivision of Proteobacteria, and two in the b-subdivision of Proteobacteria. Most of our isolates were gram-positive bacteria representing the genera Rhodococcus (high G C C%), Bacillus (low G C C%) and Arthrobacter (high G C C%). In the study of naturally degrading microbial communities by Wagner-Do¨bler et al. (1998) the microcosm enrichments by polychlorinated biphenyl (PCB) crystals were strongly dominated by gram-positive bacteria, while Pseudomonas strains were not isolated. Despite Pseudomonas (g-Proteobacteria) being the most abundant gram-negative genus among our strains, we did not find the inoculant strain P. putida PaW85 in our samples. Thus, P. putida PaW85 harbouring a TOL plasmid did not have any ecological supremacy over intrinsic bacteria. Interestingly, the TOL plasmid was found in Pseudomonas migulae strains and in a few P. oryzihabitans strains by catechol test (Table 1). These strains were also able to grow on m-toluate. These findings are new and have not been reported before. Previous studies have shown that classical enrichment tends to yield Pseudomonas strains and related gramnegative strains because of their high growth rates (Wagner-Do¨bler et al., 1998). Our gram-negative and gram-positive isolates grew well on both minimal DEF8 medium with or without rhizosphere-extract and on soil extract agar (SEA). SEA medium yielded a greater variety of both gram-negative and gram-positive isolates. DEF8, in turn, seemed to favour Pseudomonas and Arthrobacter strains. The oil soil in the compost field had been contaminated for a long period. Thus, the selection of different species able to tolerate m-toluate was higher in OS than in MS. The dominance of either gram-positive or gramnegative bacteria in soil might depend on the toxic compound and on the level of its toxicity. Sandaa et al. (1999) found that in soils of low heavy-metal concentration more sequences clustered to the gram-positive bacteria with high DNA G C C content compared to soils of high heavy-metal concentration where most of

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

251

Fig. 1. Phylogenetic tree inferred using the ARB software package by neighbour-joining analysis with JukeseCantor distance correction based on alignment of partial 16S rDNA sequences (900 bp), showing the phylogenetic affiliation of m-toluate tolerating isolates from oil-contaminated Galega orientalis rhizosphere. The scale bar represents 0.05 changes per nucleotide position. The number of isolates is indicated inside the brackets. Bacterial species either in bold or underlined in the figure includes catechol positive or catechol negative isolates able to degrade m-toluate, respectively.

the bacteria represented a-Proteobacteria. In our oilcontaminated rhizosphere, gram-positive bacteria dominated, while the best m-toluate degraders were found among the genus Pseudomonas. The most effective m-toluate degraders were isolated when m-toluate or rhizosphere-extract with m-toluate, but without additional sugars, was used as the sole carbon source. An environment rich in simple carbon and energy sources may well decrease the rate of degradation of more complex competing compounds. On the other hand, compounds secreted from plant roots might work as a trigger of the degradation of some harmful compounds (Anderson et al., 1993). 3.2. Responses of the bacterial isolates towards meta-toluate Only one-fifth of the strains tested were able to degrade m-toluate at variable levels (Table 1). All the isolates

which could degrade m-toluate also tolerated it very well. Only one strain, Pseudomonas oryzihabitans (H2397), was catechol positive (m-toluate group 1a in Table 1) and degraded m-toluate as efficiently (able to grow with 13 g/l m-toluate in the medium) as the inoculant strain P. putida PaW85 (HI828). Catechol positive P. migulae (H2394) also had a high m-toluate utilization level (8 g/l) but the catechol reaction was weaker (1b in Table 1). P. oryzihabitans strains H2396 and H2374 used m-toluate up to 3e4 g/l and above that concentration they produced a brown pigmentation up to 13 g/l m-toluate in the medium (1c in Table 1). Their catechol reaction was also weaker than that of P. putida PaW85. The reduction of the catechol 2,3-dioxygenase activity might indicate that the cell did not completely support TOL functions. In this study, however, it was not verified whether the TOL plasmid had conjugated from P. putida PaW85 to the other Pseudomonas species or whether these species harboured an indigenous TOL plasmid.

252

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

Catechol negative P. oryzihabitans H2386 and H2388 were also able to use m-toluate as a sole carbon source at high concentrations (7 g/l) (2a in Table 1). In addition, the latter isolates constantly produced some brown pigment on different media containing m-toluate (both in the degradation and tolerance tests). An interesting finding was that various strains of P. oryzihabitans had differences in their degradation abilities. Interestingly, despite being catechol negative, the grampositive Rhodococcus erythropolis strains H2382 and H2404 were, to some extent, able to use m-toluate (1e5 g/l) (mtoluate group 2b in Table 1). Candidus et al. (1994) had earlier found from Rhodococcus rhodochrous CTM a catechol 2,3-dioxygenase (C23O) which could accept both 3methylcatechol and 2,3-dihydroxybiphenyl as substrates. The gene for this C23O was situated on a plasmid pTC1. Kulakov et al. (1998) found in turn a C23O gene, edoC, from their Rhodococcus strains. It was identical to an ipbC gene from the isopropylbenzene operon of R. erythropolis BD2. Three other extradiol dioxygenases from their Rhodococcus strains had various degrees of homology to different extradiol dioxygenases. The edoB gene, which was also found in a R. erythropolis NCIMB 13065 that did not utilize aromatic compounds, showed the existence of a silent pathway(s) for degradation of aromatic compounds. R. rhodochrous K37, which was able to metabolize PCBs in alkaline conditions, was shown to contain eight genes encoding extradiol dioxygenase with 2,3-dihydroxybiphenyl 1,2-dioxygenase activity (Taguchi et al., 2004). Two of these genes were situated on a linear plasmid. All eight BphC products exhibited much higher substrate activity for 2,3-dihydroxy-biphenyl than for catechol, 3-methylcatechol or 4-methylcatechol. Similar substrate activity was observed in R. erythropolis TA421 isolated from a termite ecosystem (Kosono et al., 1997). According to the nucleotide sequence databases, C23O genes of Rhodococcus and Pseudomonas are not closely related. Over 50% of the strains were neither catechol positive nor were they able to break down m-toluate, although they tolerated it at various concentrations. Species from several gram-positive genera (Rhodococcus, Arthrobacter, Nocardia, Bacillus) tolerated very high amounts (9e13 g/l) of m-toluate (m-toluate group 3a in Table 1). Being catechol negative, a few P. oryzihabitans strains (H2611, H2612, H2613, H2614), P. jessenii (H2391) and gram-positive Nocardia calcarea (H2390) were excellent tolerants (11e13 g/l) but they produced some brown pigment constantly (3a in Table 1). Gram-negative Devosia riboflavina (H2401) and Ochrobactrum anthropi (H2402), and gram-positive Bacillus thuringiensis (H2389), B. macroides (H2375 and 2400) and R. erythropolis (H2618) also tolerated surprisingly high amounts (6e8 g/l) of m-toluate (3b in Table 1). Gramnegative R. galegae (H1174), Agrobacterium tumefaciens (H2405) and Ralstonia eutropha (H2380 and H2384) tolerated m-toluate well (3e5 g/l) (3c in Table 1).

The best m-toluate degraders isolated in this study belonged to the genus Pseudomonas. Most solventtolerant strains isolated have been found to belong to the genus Pseudomonas (Isken and de Bont, 1998). Other genera, such as Bacillus (Abe et al., 1995) and Rhodococcus (Andreoni et al., 2000), have also been shown to include solvent-tolerant strains. We showed that several gram-positive strains representing the genera Rhodococcus, Arthrobacter, Nocardia and Bacillus tolerated very high amounts of m-toluate. In addition, some Rhodococcus strains could even degrade m-toluate but not with the meta-pathway in the TOL plasmid. Most certainly, intrinsic potential exists for various degradation processes in soil. 3.3. Molecular typing methods for grouping of heterogeneous rhizosphere bacteria 3.3.1. 16S rDNA ribotyping: a minimal set of three restriction endonucleases needed In the single locus 16S rDNA RFLP analysis, ten restriction enzymes were tested alone and in different combinations for their discriminatory power to detect simultaneously a wide variety of bacterial genera (data not shown). The subdivision of ribotypes was based on the visual detection of the profiles made with all ten enzymes (Table 1). Computer-assisted analysis of the 16S rDNA PCR-RFLP patterns created with AluI, MspI and RsaI and using a similarity value of 0.7 revealed seven gram-negative and six gram-positive ribotypes among 52 m-toluate tolerating bacteria from oil-contaminated Galega rhizosphere (Fig. 2.). Different Arthrobacter species could be separated but not the subtypes of Bacillus except of B. thuringiensis. In addition, the subdivision of ribotypes in the genus Pseudomonas was only partly recognized. None of the ten restriction endonucleases was useful to distinguish species within the Rhodococcus/Nocardia branch. Different sets of enzymes may be needed to resolve strains within certain species. Interestingly, AluI alone could be used to separate different genera from each other. It could also, to some extent, differentiate strains even at species level inside the genera Ralstonia and Pseudomonas. On the other hand, gram-negative and gram-positive bacteria were slightly intermingled in computer-assisted analysis based on only AluI. 3.3.2. Genomic fingerprinting by rep-PCR: the most useful method was (GTG)5-PCR The grouping results derived from rep-PCR genomic fingerprinting with all four primer pairs were generally in good agreement with each other (data not shown). When primers were used in pairwise combinations or all four primers together, no important additional information was achieved. (GTG)5-PCR produced the

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

253

Fig. 2. A UPGMA dendrogram of the isolates from oil-contaminated Galega orientalis rhizosphere based on 16S rDNA PCR-RFLP fingerprinting analysis with combined patterns of three restriction endonucleases: AluI, MspI and RsaI. The cophenetic correlation value of the dendrogram was 93.2% representing trustworthy the similarity matrix. E or e, catechol positive or negative, respectively; B, able to break down m-toluate as the sole C source.

clearest, best-separated fingerprints consisting of both small and large bands and they were reproducible both with gram-negatives and gram-positives (Fig. 3.). At the similarity level of 70%, gram-negative and grampositive strains were clustered into 11 and 12 (GTG)5groups, respectively (Fig. 4a,b). The groups represented different species. Species in the same genera were grouped together but the location of the genera did not necessarily reflect their phylogenetic relationships. Only Pseudomonas and Bacillus species were mostly interspersed through the dendrograms demonstrating their heterogeneous genomic nature. In addition, Ralstonia eutropha was divided into two (GTG)5genotypes, though grouped with 0.62 similarity, and

the Rhodococcus/Nocardia brand into four (GTG)5genotypes, which formed a group at the 0.39 similarity level. Despite the careful standardization at any level, the fingerprints even from the same strain may vary so much that the similarity percents in dendrograms cannot be taken as absolute identity measures. Instead, in this study it was shown that (GTG)5-PCR genomic fingerprinting could also be used to infer the degree of shifts in the community structure at species level. Our results are supported by the observation that the genomic structure of a bacterium, as deduced from its genomic fingerprint, represents an accurate reflection of its taxonomic and phylogenetic position based on total genomic

254

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

Fig. 3. (GTG)5-PCR fingerprinting of some isolates from oil-contaminated Galega orientalis rhizosphere. M, lSinI molecular marker. 1, Rhizobium galegae H1174; 2, Pseudomonas putida PaW85 H1828; 3, Rhodococcus erythropolis H2378; 4, H2616; 5, H2617; 6, Arthrobacter histidinolovorans H2379; 7, A. histidinolovorans H2381; 8, A. histidinolovorans H2383; 9, A. histidinolovorans H2372; 10, H1614; 11, P. oryzihabitans H2386; 12, H2387; 13, P. oryzihabitans H2388; 14, Bacillus thuringiensis H2389; 15, Nocardia calcarea H2390. Identification of the isolates is indicated in Table 1.

DNA-DNA hybridisation values (Rademaker et al., 2000). There are only two previous reports on the use of the (GTG)5 primer for genomic fingerprinting: one on gram-negative rhizobia (Nick et al., 1999) and another on gram-positive Lactobacillus (Gevers et al., 2001). We can recommend (GTG)5 fingerprinting for both gram-negative and gram-positive bacteria to be successful with rep-PCR analysis.

4. Conclusions Only one-fifth of the strains that tolerated m-toluate in oil-contaminated Galega rhizosphere also degraded m-toluate. Some of the tolerant strains might be involved in the degradation, probably also the co-metabolising, of

other hydrocarbons present in the oil-contaminated soil. The ability to degrade m-toluate by a TOL plasmid was detected only in species of the genus Pseudomonas. In addition, a few Rhodococcus erythropolis strains were found to be able to degrade m-toluate most likely by a different kind of catechol 2,3-dioxygenase compared to one coded by the TOL plasmid. The inoculum Pseudomonas putida PaW85 was not found in our rhizosphere samples. Thus, some supreme intrinsic degradation potential exist in oil-contaminated rhizosphere. Various Pseudomonas oryzihabitans strains differed in their ability to degrade m-toluate. P. migulae strains and a few P. oryzihabitans strains were able to grow on m-toluate and most likely contained the TOL plasmid. These findings have not been reported before. Separation at the strain level was important because strain specific differences in degradation abilities were found for P. oryzihabitans. (GTG)5-PCR was the best method for strain specific fingerprinting. 16S ribotyping, (GTG)5-genotyping and selective partial sequencing of 16S rDNA genes formed our molecular toolbox for studying different perspectives of genetic diversity of culturable bacteria in oil-contaminated rhizosphere. (GTG)5-genotyping proved to be a powerful method differentiating bacterial species. At the same time, it separated even different strains among Pseudomonas oryzihabitans and Ralstonia eutropha. It also revealed some intricacies in the naming of Rhodococcus/Nocardia and Bacillus species. Generally, the (GTG)5-genotype and the 16S-ribotype (or subtype) corresponded to each other very well representing a species. In some cases, 16S-ribosubgrouping revealed more differences inside species than (GTG)5-genotyping, like in the case of Arthrobacter histidinolovorans. On the other hand, the restriction endonucleases used might not be the best ones for all the genera (Rhodococcus/ Nocardia). Thus, the information obtained from 16Sribotyping and (GTG)5-genotyping complemented each other. When combined with selective partial sequencing, they offered rapidly gainable and relevant phylogenetic information to be used in microbial ecology studies of rhizosphere populations.

Acknowledgments This work was supported by grants from the Maj and Tor Nessling Foundation, the Academy of Finland, and Ekokem Oy. We thank Alexander Kraft for help with the isolation of rhizosphere bacteria. Sirpa Tiikkainen and Minna Sinkkonen are acknowledged for their help in the greenhouse, and Seppo Kaijalainen during DNA isolation. Eeva-Liisa Ko¨ssi is thanked for technical assistance in repeating parts of the rep-PCR and 16S rDNA PCR-RFLP analysis.

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

255

Fig. 4. UPGMA dendrograms of the gram-positive (a) and gram-negative (b) isolates from oil-contaminated Galega orientalis rhizosphere based on genomic fingerprinting by rep-PCR with (GTG)5 primers. The reference strains Rhizobium galegae and Pseudomonas putida PaW85 (not isolated from the soils) were included in the dendrogram (b) for reference. The value of cophenetic correlation with gram-negative bacteria was 92.7% and with gram-positive bacteria 95.2% representing trustworthy the similarity matrix. E or e, catechol positive or negative, respectively; B, able to break down m-toluate as the sole C source.

256

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257

References Abe, A., Inoue, A., Usami, R., Moriya, K., Horikoshi, K., 1995. Degradation of polyaromatic hydrocarbons by organic solventtolerant bacteria from deep sea. Biosci. Biotech. Biochem. 59, 1154e1156. Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local alignment search tool. J. Mol. Biol. 215, 403e410. Altschul, S.F., Madden, T.L., Scha¨ffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389e3402. Anderson, T.A., Guthrie, E.A., Walton, B.T., 1993. Bioremediation in the rhizosphere: plant roots and associated microbes clean contaminated soil. Environ. Sci. Technol. 27, 2630e2636. Andreoni, V., Bernasconi, S., Colombo, M., van Beilen, J.B., Cavalca, L., 2000. Detection of genes for alkane and naphthalene catabolism in Rhodococcus sp. strain 1BN. Environ. Microbiol. 2, 572e577. Assinder, S.J., Williams, P.A., 1990. The TOL plasmids: determinants of the catabolism of toluene and the xylenes. Adv. Microb. Physiol. 31, 1e69. Bayley, S.A., Duggleby, C.J., Worsey, M.J., Williams, P.A., Hardy, K.G., Broda, P., 1977. Two modes of loss of the tol function from Pseudomonas putida mt-2. Mol. Gen. Genet. 154, 203e204. Beringer, J.E., 1974. R factor transfer in Rhizobium leguminosarum. J. Gen. Microbiol. 84, 188e198. Boom, R., Sol, C.J.A., Salimans, M.M.M., Jansen, C.L., Wertheim van Dillen, P.M.E., van der Noordaa, J., 1990. Rapid and simple method for purification of nucleic acids. J. Clin. Microbiol. 28, 495e503. Candidus, S., van Pee, K.H., Lingens, F., 1994. The catechol 2,3dioxygenase gene of Rhodococcus rhodochrous CTM: nucleotide sequence, comparison with isofunctional dioxygenases and evidence for an active-site histidine. Microbiology 140, 321e330. de Bruijn, F.J., 1992. Use of repetitive (repetitive extragenic palindromic and enterobacterial repetitive intergeneric consensus) sequences and the polymerase chain reaction to fingerprint the genomes of Rhizobium meliloti isolates and other soil bacteria. Appl. Environ. Microbiol. 58, 2180e2187. Edwards, U., Rogall, T., Blo¨cker, H., Emde, M., Bo¨ttger, E.C., 1989. Isolation and direct complete nucleotide determination of entire genes. Characterization of a gene coding for 16S ribosomal RNA. Nucleic Acids Res. 17, 7843e7853. Elonen, P., 1971. Particle-size analysis of soil. Acta Agral. Fenn. 122, 1e122. Gevers, D., Huys, G., Swings, J., 2001. Applicability of rep-PCR fingerprinting for identification of Lactobacillus species. FEMS Microbiol. Lett. 205, 31e36. Gurtler, V., Wilson, V.A., Mayall, B.C., 1991. Classification of medically important clostridia using restriction endonuclease site differences of PCR-amplified 16S rDNA. J. Gen. Microbiol. 137, 2673e2679. Heyd, M.L., Diehl Jr., J.D., 1996. Recovering DNA from agarose gels with pumice. BioTechniques 20, 394e398. Isken, J., de Bont, J.A.M., 1998. Bacteria tolerant to organic solvents. Extremophiles 2, 229e238. Judd, A.K., Schneider, M., Sadowsky, M.J., de Bruijn, F.J., 1993. Use of repetitive sequences and the polymerase chain reaction technique to classify genetically related Bradyrhizobium japonicum serocluster 123 strains. Appl. Environ. Microbiol. 59, 1702e1708. Kosono, S., Maeda, M., Fuji, F., Arai, H., Kudo, T., 1997. Three of the seven bphC genes of Rhodococcus erythropolis TA421, isolated from a termite ecosystem, are located on an indigenous plasmid associated with biphenyl degradation. Appl. Environ. Microbiol. 63, 3282e3285.

Koukila-Ka¨hko¨la¨, P., Springer, B., Bo¨ttger, E.C., Paulin, L., Jantzen, E., Katila, M.-L., 1995. Mycobacterium branderi sp. nov., a new potential human pathogen. Int. J. Syst. Bacteriol. 45, 549e553. Kulakov, L.A., Delcroix, V.A., Larkin, M.J., Ksenzenko, V.N., Kulakova, A.N., 1998. Cloning of new Rhodococcus extradiol dioxygenase genes and study of their distribution in different Rhodococcus strains. Microbiology 144, 955e963. Laguerre, G., Allard, M.-R., Revoy, F., Amarger, N., 1994. Rapid identification of rhizobia by restriction fragment length polymorphism analysis of PCR-amplified 16S rRNA genes. Appl. Environ. Microbiol. 60, 56e63. Laguerre, G., Mavingui, P., Allard, M.-R., Charnay, M.-P., Louvrier, P., Mazurier, S.-I., Rigottier-Gois, L., Amarger, N., 1996. Typing of rhizobia by PCR DNA fingerprinting and PCRrestriction fragment length polymorphism analysis of chromosomal and symbiotic gene regions: application to Rhizobium leguminosarum and its different biovars. Appl. Environ. Microbiol. 62, 2029e2036. Lindstro¨m, K., Jarvis, D.B.W., Lindstro¨m, P.E., Patel, J.J., 1983. DNA homology, phage-typing, and cross-nodulation studies of rhizobia infecting Galega species. Can. J. Microbiol. 29, 781e789. Lindstro¨m, K., Lehtoma¨ki, S., 1988. Metabolic properties, maximum growth temperature and phage sensitivity of Rhizobium sp. (Galega) compared with other fast-growing rhizobia. FEMS Microbiol. Lett. 50, 277e287. Louws, F.J., Fulbright, D.W., Stephens, C.T., de Bruijn, F.J., 1994. Specific genomic fingerprints of phytopathogenic Xanthomonas and Pseudomonas pathovars and strains generated with repetitive sequences and PCR. Appl. Environ. Microbiol. 60, 2286e2295. Ludwig, W., Strunk, O., Westram, R. et al., 2004. ARB: a software environment for sequence data. Nucleic Acids Res. 32, 1363e1371 (http://www.arb-home.de). Nick, G., Jussila, M.M., Hoste, B., Niemi, R.M., Kaijalainen, S., de Lajudie, P., Gillis, M., de Bruijn, F.J., Lindstro¨m, K., 1999. Rhizobia isolated from root nodules of tropical leguminous trees characterized using DNA-DNA dot-blot hybridisation and repPCR genomic fingerprinting. Syst. Appl. Microbiol. 22, 287e299. Nielsen, T.H., Sørensen, J., 2003. Production of cyclic lipopeptides by Pseudomonas fluorescens strains in bulk soil and in the sugar beet rhizosphere. Appl. Environ. Microbiol. 69, 861e868. Rademaker, J.L.W., Hoste, B., Louws, F.J., Kersters, K., Swings, J., Vauterin, L., Vauterin, P., de Bruijn, F.J., 2000. Comparison of AFLP and rep-PCR genomic fingerprinting with DNA-DNA homology studies: Xanthomonas as a model system. Int. J. Syst. Evol. Microbiol. 50, 665e677. Radwan, S., Sorkhoh, N., El-Nemr, I., 1995. Oil biodegradation around roots. Nature 376, 302. Read, D.B., Bengough, A.G., Gregory, P.J., Crawford, J.W., Robinson, D., Scrimgeour, C.M., Young, I.M., Zhang, K., Zhang, X., 2003. Plant roots release phospholipid surfactants that modify the physical and chemical properties of soil. New Phytol. 157, 315e326. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual, 4 vols., second ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sandaa, R.-A., Torsvik, V., Enger, Ø., Daae, F.L., Castberg, T., Hahn, D., 1999. Analysis of bacterial communities in heavy metalcontaminated soils at different levels of resolution. FEMS Microbiol. Ecol. 30, 237e251. SFS 3009, 1980. Determination of oil and grease in water. Gravimetric method. Finnish Standards Association (SFS), Helsinki. Suominen, L., Jussila, M.M., Ma¨kela¨inen, K., Romantschuk, M., Lindstro¨m, K., 2000. Evaluation of the Galega-Rhizobium galegae system for the bioremediation of oil-contaminated soil. Environ. Pollut. 107, 239e244. Taguchi, K., Motoyama, M., Kudo, T., 2004. Multiplicity of 2,3-dihydroxybiphenyl dioxygenase genes in the gram-positive

M.M. Jussila et al. / Environmental Pollution 139 (2006) 244e257 polychlorinated biphenyl degrading bacterium Rhodococcus rhodochrous K37. Biosci. Biotechnol. Biochem. 68, 787e795. Versalovic, J., Koeuth, T., Lupski, J.R., 1991. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res. 19, 6823e6831. Versalovic, J., Schneider, M., de Bruijn, F.J., Lupski, J.R., 1994. Genomic fingerprinting of bacteria using repetitive sequencebased polymerase chain reaction. Methods Mol. Cell. Biol. 5, 25e40. Wagner-Do¨bler, I., Bennasar, A., Vancanneyt, M., Stro¨mpl, C., Bru¨mmer, I., Eichner, C., Grammel, I., Moore, E.R.B., 1998. Microcosm enrichment of biphenyl-degrading microbial communities from soils and sediments. Appl. Environ. Microbiol. 64, 3014e3022.

257

Weisburg, W.G., Barns, S.M., Pelletier, D.A., Lane, D.J., 1991. 16S ribosomal DNA amplification for phylogenetic study. J. Bacteriol. 173, 697e703. Williams, P.A., Murray, K., 1974. Metabolism of benzoate and the methylbenzoates by Pseudomonas putida (arvilla) mt-2: evidence for the existence of a TOL plasmid. J. Bacteriol. 120, 416e423. Williams, P.A., Taylor, S.D., Gibb, L.E., 1988. Loss of the toluenexylene catabolic genes of TOL plasmid pWW0 during growth of Pseudomonas putida on benzoate is due to a selective growth advantage of ‘cured’ segregants. J. Gen. Microbiol. 134, 2039e2048. Wilson, K., 1989. Preparation of genetic DNA from bacteria. In: Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K. (Eds.), Current Protocols in Molecular Biology. John Wiley, New York, pp. 2.4.1e2.4.5.