Diversity of biphenyl degraders in a chlorobenzene polluted aquifer

Chemosphere 58 (2005) 529–533 www.elsevier.com/locate/chemosphere

Short Communication

Diversity of biphenyl degraders in a chlorobenzene polluted aquifer Wolf-Rainer Abraham a

a,*

, Dirk F. Wenderoth a, Walter Gla¨ßer

b

Division of Microbiology, GBF—National Research Center for Biotechnology, Mascheroder Weg 1, 38124 Braunschweig, Germany b UFZ—Environment Research Center Leipzig-Halle, Section Hydrogeology, Theodor-Lieser Str. 4, 06120 Halle, Germany Received 27 January 2004; received in revised form 20 August 2004; accepted 31 August 2004

Abstract Biphenyl degrading bacteria (40 strains) have been isolated along a gradient of chlorobenzene pollution from an aquifer which did not contain any PCB to answer the question of how metabolic/catabolic abilities exist in ecosystems that have not been stressed with the relevant substrates is important for intrinsic bioremediations. Only few of the isolates were characterized by 16S rRNA gene sequence analyses as Pseudomonas species while the majority were Grampositive, belonging to the order Actinomycetales and representing the genera Rhodococcus and Arthrobacter. The strains could grow on a variety of chlorobenzoates but no pattern of substrate usage and phylogeny or pollution gradient could be found. Strains which were able to grow on 2,5-dichlorobenzoate were often also able to use 3,4- and 3,5-dichloroand 2,3,5-trichlorobenzoate or those using 2-chlorobenzoate could usually use 2,6-dichlorobenzoate as well. From that results, it is concluded that a highly diverse, basic metabolic activity for PCB degradation existed at this site despite the absence of PCB. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Bitterfeld aquifer; Biphenyl degraders; PCB; Microbial diversity; Chlorobenzoate degradation; Rhodococcus species; Natural attenuation

1. Introduction Although xenobiotica have been introduced into the environment for only the last few centuries, many bacteria are able to degrade a number of them. This is also true for polychlorinated biphenyls (PCB), released into the environment. The origin of the genes enabling many bacteria to use PCB as a substrate is still unknown and the source for many debates (Abraham et al., 2002b). It

* Corresponding author. Tel.: +49 531 618 1419; fax: +49 531 6181 411. E-mail address: [email protected] (W.-R. Abraham).

is interesting whether bacteria which are able to degrade PCB and their intermediate chlorobenzoates could be isolated from a site contaminated with pollutants not containing PCBs. Especially, to elucidate the metabolic potential of the isolates and to connect the pollution profile present at the site with the degradation capabilities of the isolates could be informative. For this study the area of Bitterfeld in Germany was chosen, a site polluted for more than 100 years with a wide variety of pollutants (Weiss et al., 2001). The material of a drilling core to a depth of 30 m at the SAFIRA site (Weiss et al., 1998) near Bitterfeld was used as an inoculum for the isolations. At this site, the water table is 6 m below the surface and a dense layer of lignite is present

0045-6535/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2004.08.074

530

W.-R. Abraham et al. / Chemosphere 58 (2005) 529–533

20 m below the surface which reaches down to 30 m. This lignite layer is loaded with various pollutants and acts as their reservoir (Dermietzel and Christoph, 2001). The water above a depth of 9 m is not polluted although below 9 m, down to the lignite horizon at 20 m, the chlorobenzene concentration increases, reaching 20–30 mg l 1 directly above the lignite interface. Minor pollutants at this site are dichlorobenzenes (Abraham et al., 1997). From the aquifer material, biphenyl degrading strains were isolated along the depth profile and the pollution gradient and their ability to degrade a number of chlorinated benzoates was characterized. The aim was to find a correlation between pollution and the degradation capability of the isolates.

2. Experimental Sampling: At the site in Bitterfeld, Germany, contaminated with chlorobenzene, the saturated zone began at 6 m below the surface and a layer of lignite was present between 20–30 m below the surface with concentrations of chlorobenzene of 27 mg l 1 and 1,4-dichlorobenzene of 0.3 mg l 1 peaking directly above the lignite layer (Popp and Mo¨der, 1997). In March 1997, a core of 10 cm diameter (Safbit 13/97) was obtained from a drilling 30 m deep into the contaminated aquifer. For those parts of the drill core which were above the lignite horizon, the material was pooled for each meter, while the core containing the lignite layer was cut and samples of 5 g were taken at the surface of the cuttings, as detailed in Abraham et al. (1997). The samples were stored at 4 °C and used for enrichment within 24 h. Isolation of bacteria and determination of the metabolic potential: Strains were enriched, under aerobic conditions, on medium M9 with biphenyl as the sole source of carbon (Abraham et al., 1999). Bacteria were isolated from soil on a minimal agar medium (8 g NH4H2PO4, 0.2 g yeast extract, 2 g K2HPO4, 0.5 g MgSO4 Æ 7H2O, 0.5 g Na2SO4, 0.5 g NaCl, 10 mg ZnCl2 Æ 2H2O, 8 mg MnSO4 Æ 7H2O, 10 mg FeSO4 Æ 7H2O, and 50 mg CaCl2 in 1 l of distilled water) (M9 medium) and biphenyl, as carbon source. The biphenyl was applied via the gas phase to the culture on the agar surface. A bacterial suspension was streaked out on the agar in a dilution series. Pure colonies were identified by nearly full-length 16S rRNA gene sequences and comparison with public and in-house databases, as described by Abraham et al. (2002a). Screening on substrate usage was performed in liquid culture with 1 g l 1 carbon source and OD600 was determined after 1, 2, 3, 5 and 7 days. Growth was determined as ‘‘fair’’ for an 25–49% increase of OD600, as ‘‘moderate’’ for an 50–99% increase and as ‘‘good’’ for an increase of more than 100% (Table 2). Statistical analysis: For the correlation of metabolic activities and the generation of dendrograms the software

Table 1 Bacterial isolates grown on biphenyl and used in this study. The closest relative and its 16S rRNA gene sequence similarity is given Strain

Depth

Nearest affiliation

WAB338 WAB344 WAB325 WAB327 WAB328 WAB331 WAB332 WAB333 WAB315 WAB317 WAB321 WAB322 WAB305 WAB306 WAB310 WAB311 WAB312 WAB314 WAB389 WAB390 WAB391 WAB392 WAB393 WAB394 WAB364 WAB365 WAB367 WAB369 WAB377 WAB378 WAB379 WAB381 WAB382 WAB383 WAB384 WAB396 WAB354 WAB355 WAB360 WAB361

1m 1m 10 m 10 m 10 m 10 m 10 m 10 m 19 m 19 m 19 m 19 m 20 m 20 m 20 m 20 m 20 m 20 m 23.7 m 23.7 m 23.7 m 23.7 m 23.7 m 23.7 m 25.7 m 25.7 m 25.7 m 25.7 m 25.7 m 25.7 m 25.7 m 25.7 m 25.7 m 25.7 m 25.7 m 26.7 m 27.7 m 27.7 m 27.7 m 27.7 m

Rhodococcus koreensis 99% Rhodococcus erythropolis 98% Pseudomonas sp. NBO1-H 99% Promicromonospora enterophila 98% Arthrobacter polychromogenes 99% Rhodococcus erythropolis 99.8% Arthrobacter polychromogenes 99% Arthrobacter sp. CAB 1 95% Micrococcus lylae strain d10 97% Arthrobacter sulfureus 98% Pseudomonas sp. ST 5 98% Pseudomonas brennerii 99% Rhodococcus fascians 99% Arthrobacter sulfureus 98% Arthrobacter polychromogenes 99% Pseudomonas rhodesiae 99% Acinetobacter sp. V4.ME-25 99% Pseudomonas veronii 99.8% Rhodococcus sp. 98% Corynebacterium xerosis 98% Dietzia sp. ES 18 97% Rhodococcus sp. 99.1% Corynebacterium sp. 99% Dietzia maris 99% Dietzia natronolimnaea 99% Dietzia sp. ES 18 98% Dietzia sp. ES 18 97% Arthrobacter sulfureus 98% Arthrobacter sp. CAB 1 99% Arthrobacter polychromogenes 98% Rhodococcus sp. UFZ-B520 99% Rhodococcus fascians 99% Rhodococcus luteus (DSM 43673) 99% Rhodococcus sp. 98.5% Rhodococcus sp. 99% Rhodococcus sp. LG3 99% Rhodococcus erythropolis 100% Arthrobacter sp. 99.7% Dietzia sp. 99.8% Rhodococcus fascians 99%

StatisticaÒ 6.0 (StatSoft) has been used. The substrates were compared by their support of bacterial growth and the strains by their metabolic activity applying multivariate exploratory techniques. The cluster analysis used unweighted pair-group average and the standard Pearson product–moment correlation coefficients basing the reliability analysis on those correlations.

3. Results and discussion On various media, a number of strains were isolated from samples taken from the drilling core (Abraham

W.-R. Abraham et al. / Chemosphere 58 (2005) 529–533

531

Table 2 Growth of the isolates on chlorinated benzoates and PCB (Aroclor 1221) Strain

Bzt

2-Cl

3-Cl

4-Cl

2,3-Cl

2,4-Cl

2,5-Cl

2,6-Cl

3,4-Cl

3,5-Cl

2,3,5-Cl

2,4,6-Cl

A1221

WAB338 WAB344

+++ +++

+ ++

++ ++

+ –

+ –

+++ ++

++ ++

++ ++

++ ++

+ ++

+ ++

++ ++

++ ++

WAB325 WAB327 WAB328 WAB331 WAB332 WAB333

– ++ +++ + +++ +++

– + ++ + + +

– – ++ ++ + +

– – ++ ++ ++ ++

– – + ++ + ++

– + +++ ++ +++ ++

– + ++ ++ + ++

– + ++ ++ + ++

– + + ++ – +

+ + + + – +

+ + + ++ + +

+++ + ++ +++ ++ ++

+ ++ +++ – +++ –

WAB315 WAB317 WAB321 WAB322

– ++ – –

– – – –

– – – –

– – – –

++ ++ ++ ++

+ + + ++

++ ++ ++ +

+ + + –

+ + + +

+ ++ + +

++ ++ ++ +

+ ++ ++ +

+ ++ ++ ++

WAB305 WAB306 WAB310 WAB311 WAB312 WAB314

+ ++ +++ – – ++

++ + + – – +

++ + + – + +

++ ++ ++ – – +

++ + ++ + ++ +

+++ ++ +++ ++ + +

++ ++ ++ + ++ +

++ + + + + +

+ + + + ++ +

+ + + + + –

++ ++ ++ + ++ –

++ ++ ++ + + ++

– ++ +++ + + –

WAB389 WAB390 WAB391 WAB392 WAB393 WAB394

+ +++ +++ +++ ++ +++

++ ++ + ++ ++ ++

++ – – + ++ –

++ – – ++ + –

++ – – ++ ++ –

++ ++ + ++ ++ ++

++ +++ + ++ +++ ++

++ ++ + ++ ++ ++

++ +++ ++ + ++ +++

+ ++ + – ++ ++

++ ++ + + +++ ++

+++ ++ + ++ +++ ++

– +++ ++ +++ ++ +++

WAB364 WAB365 WAB367 WAB369 WAB377 WAB378 WAB379 WAB381 WAB382 WAB383 WAB384

+++ ++ ++ +++ +++ + + + +++ + +

++ ++ ++ + + + + + – + +

+ ++ ++ + + + + + – + +

+ + ++ + + ++ ++ + + + +

+ ++ ++ + + + + + – ++ +

++ ++ +++ + +++ – +++ ++ ++ ++ +

++ ++ ++ + + – ++ + ++ ++ +

++ + ++ + + + ++ ++ + ++ ++

++ + ++ + + – + ++ – +++ +

+ + + + – – + + – ++ +

++ ++ ++ + + + + ++ – +++ ++

++ +++ +++ ++ + +++ + +++ – +++ +++

++ – – – – – – – – ++ –

WAB396 WAB354 WAB355 WAB360 WAB361

– +++ +++ +++ ++

– ++ ++ + ++

– +++ ++ + ++

– ++ ++ + ++

– ++ ++ + ++

++ +++ ++ + +++

++ ++ ++ – ++

++ ++ ++ ++ ++

+++ ++ + – +

++ + + – +

++ ++ ++ + ++

++ ++ +++ +++ ++

+++ ++ ++ +++ –

Isolates from different depths are separated by broader lines (+++ = good growth, ++ = moderate growth, + = fair growth, – = no growth; Bzt = benzoate, x-Cl = x-chlorobenzoate, A1221 = Aroclor 1221).

et al., 1999). From the strains isolated on biphenyl as the sole source of carbon and energy, 40 have been selected for further study. Only few of them belonged to Gramnegative phyla, Pseudomonas veroni being one of the phylogenetically closest relatives. Strains of Pseudomonas species could be isolated only down to a depth of 19 m and were not found in the enrichments from the lignite layer. However, most of the strains were Gram-positive, which is surprising because Gram-posi-

tive strains presented only a minor population at this site (Vogt et al., 2002). It is interesting to note that one of the strains, WAB 379, was very closely related to Rhodococcus sp. UFZ-B520, isolated from the same site as a chlorobenzene degrader (Vogt et al., 2000). Sequencing of the 16S rRNA genes of the strains and comparison with international and in-house data bases revealed that the strains belonged mainly to the order Actinomycetales and here to the genera Rhodococcus and Arthrobacter,

W.-R. Abraham et al. / Chemosphere 58 (2005) 529–533 110 100 90 80 70 60 50 40

Bzt

2-Cl

2,6-Cl

3-Cl

4-Cl

2,4-Cl

2,3-Cl

2,5-Cl

2,4,6-Cl

3,4-Cl

3,5-Cl

20

2,3,5-Cl

30 A1221

while some strains were members of the genera Corynebacterium, Dietzia and Promicromonospora. No phylogenetic trend could be observed along the depth and pollution gradients (Table 1). The growth of the isolates have been tested on various substrates. Besides the PCB mixture Aroclor 1221 (Shiu and Mackay, 1986), benzoates and various mono-, di- and trichlorobenzoates were tested as growth substrates. The chlorobenzoates have been chosen as substrates because of their formation as intermediates during the degradation of PCB. Most microorganisms possess either the ability to degrade the first aromatic ring in the biphenyl skeleton (upper degradation pathway), leaving the second one untouched, or they degrade the resulting chlorobenzoates (lower degradation pathway) (Mokross et al., 1990). The results are listed in Table 2. A wide variety of metabolic activities have been found in these strains. To elucidate whether some of these activities are coupled, i.e., whether there was a correlation between one activity and another one, statistical analyses have been performed (Fig. 1). It was observed that the ability of the strains to degrade 2-chloro-benzo-

Similarity

532

Fig. 1. Dendrogram comparing the substrates based on their usage by the 40 isolates. Abbreviation of the substrates are the same as in Table 2.

ate correlates, highly significant (p < 0.001) and positively, with the use of 3-chlorobenzoate. Correlations have been observed between: (i) the degradation of

Fig. 2. Similarity of the strains based on their metabolic diversity. Below the dendrogram the different depths are given from where the strains have been isolated.

W.-R. Abraham et al. / Chemosphere 58 (2005) 529–533

2,5-dichlorobenzoate and 3,4-, 3,5-dichloro- and 2,3,5trichlorobenzoate, and (ii) 3,4- and 3,5-dichloro- and 2,3,5-trichlorobenzoate. From Fig. 1 it is also obvious that the use of Aroclor 1221, 2,3-dichloro- or 2,4,6-trichloro-benzoate did only correlate weakly with any of the other substrates. It is possible that these correlations can be interpreted as the occurrence of isoenzymes or similarities in the active centers of the dioxygenases involved in degradation of these chlorobenzoates. Such an interpretation is likely because the strains tested were phylogenetically closely related. A dendrogram based on the metabolic activities listed in Table 2 for all strains revealed no correlation of the metabolic activities with the pollution gradient, i. e., with depth (Fig. 2). However, the degradation of some substrates seemed to be depth-dependent. While almost no isolate from 25.7 m was able to grow on Aroclor 1221, this substrate was degraded very well by most of the isolates from 23.7 and 27.7 m. All isolates from 19 m lacked the ability to grow on the mono-chlorinated benzoates and only one of four could grow on benzoate. Benzoate was also not degraded by all isolates from 20 m but was a very good substrate for most isolates from 23.7 m. These observations point to an adaptation of the microbial communities at the different depths to some compounds similar to these substrates. In summary, it has been shown that a rich diversity of biphenyl degrading strains have been found at a polluted site, although PCB has never been produced here. Most of the cultivable strains were Gram-positive and the genus Rhodococcus comprised most of the strains. Neither a pattern in the phylogeny of the strains with depth could be found nor the metabolic diversity of the isolates could be correlated with depth. The concentrations of the pollutant present at this site, mainly chlorobenzene, did not correlate with the biphenyl and chlorobenzoate degrading activities of the strains. It is concluded that a background activity of PCB degradation existed, at least, at this site, which displayed a large metabolic diversity with high redundancies. It is tempting to speculate that such a diversity could be the optimal starting point for strain selection and optimisation for PCB degradation, both in natural attenuation and in biotechnological applications. Acknowledgments The assistance of Dagmar Wenderoth and Tanja Jeschke in the isolation of the strains and Annette Kru¨ger in 16S rRNA gene sequencing is greatly acknowledged. This study has been partly supported by HGF strategy funds
System integrated environmental biotechnology of BMBF and HGF, Germany.

533

References Abraham, W.-R., Tesar, M., Nogales, B., Lu¨nsdorf, H., 1997. Phylogenetische und funktionelle Vielfalt der Aromaten abbauenden Bakterien im kontaminierten Aquifer von Bitterfeld. In: Weiß, H., Teutsch, G., Daus, B. (Eds.), Sanierungsforschung in regional kontaminierten Aquiferen (SAFIRA), UFZ-Bericht 27/1997, pp. 131–136. Abraham, W.-R., Wenderoth, D., Jeschke, T., 1999. Physiologische Diversita¨t Aromaten-abbauender Bakterien im kontaminierten Aquifer von Bitterfeld. In: Weiß, H., Daus, B., Teutsch, G. (Eds.), SAFIRA. 2. Statusbericht: Modellstandort, Mobile Testeinheit, Pilotanlage. UFZ-Bericht 17/ 1999, pp. 40–45. Abraham, W.-R., Stroempl, C., Bennasar, A., Vancanneyt, M., Snauwaert, C., Swings, J., Smit, J., Moore, E.R.B., 2002a. Phylogeny of Maricaulis Abraham et al. 1999 and proposal of Maricaulis virginensis sp. nov., M. parjimensis sp. nov., M. washingtonensis sp. nov. and M. salignorans sp. nov. Int. J. Syst. Evol. Microbiol. 52, 2191–2201. Abraham, W.R., Nogales, B., Golyshin, P., Pieper, D.H., Timmis, K.N., 2002b. Polychlorinated biphenyl-degrading microbial communities in soils and sediments. Curr. Opin. Microbiol. 5, 246–253. Dermietzel, J., Christoph, G., 2001. The impact of a lignite seam on contaminated groundwater in the aquifer system of the Bitterfeld region. Water, Air Soil Pollut. 125, 157–170. Mokross, H., Schmidt, E., Reineke, W., 1990. Degradation of 3-chlorobiphenyl by in vivo constructed hybrid pseudomonads. FEMS Microbiol. Lett. 71, 179–186. Popp, P., Mo¨der, M., 1997. Organische Analytik. In: Weiß, H., Teutsch, G., Daus, B. (Eds.). Sanierungsforschung in regional kontaminierten Aquiferen (SAFIRA), UFZ-Bericht 27/1997, pp. 73–75. Shiu, W.Y., Mackay, D., 1986. A critical review of aqueous solubilities, vapor pressures, Henrys law constants, and octanol-water partition coefficients of the polychlorinated biphenyls. J. Phys. Chem. Ref. Data 15, 911–929. Vogt, C., Lorbeer, H., Alfreider, A., Wuensche, L., Babel, W., 2000. Chlorobenzene degrading bacteria isolated from a chlorobenzene-contaminated aquifer system. Genbank submission AF235011. Vogt, C., Alfreider, A., Lorbeer, H., Ahlheim, J., Feist, B., Boehme, O., Weiss, H., Babel, W., Wu¨nsche, L., 2002. Two pilot plant reactors designed for the in situ bioremediation of chlorobenzene-contaminated ground water: hydrogeological and chemical characteristics and bacterial consortia. Water, Air Soil Pollut.: Focus 2, 161–170. Weiss, H., Kopinke, F.-D., Popp, P., Wu¨nsche, L., 1998. In situ remediation research in a complexly contaminated aquifer: the SAFIRA test site at Bitterfeld, Germany. NATO/CCMS Pilot study: Evaluation of Demonstrated and Emerging Technologies for the Treatment of Contaminated Land and Groundwater (Phase III), Special Session, no. 229, EPA 542-R-98-003, pp. 84–91. Weiss, H., Teutsch, P., Fritz, P., Daus, B., Grathwohl, P., Trabitzsch, R., Feist, B., Ruske, R., Bo¨hme, O., Schirmer, M., 2001. Sanierungsforschung in regional kontaminierten Aquiferen (SAFIRA)—1. Information zum Forschungsschwerpunkt am Standort Bitterfeld. Grundwasser 6 (3), 113–122.