Genomic heterogeneity in Chlorobium limicola: chromosomic and plasmidic differences among strains

ELSEVIER

FEMS Microbiology

Letters I34 (1995) 279-285

Genomic heterogeneity in Chlorobium Zimicola: chromosomic and plasmidic differences among strains Sebastih

Mhdez-Alvarez

a lnstirur de Biologic

a,* , Victkia Pav6n a, Isabel Esteve a, Ricardo Guermo NGria Gaju a

i Microbiologia. Uniwrsifat Auicinoma de Barcelona. Spain ’ Departamenr de Microbiologia. Uniwrsitut de Barcelona, 08028 Barcelona. Spain

Fonummtal

and Departament

Received

18 September

de Genhtica

1995; revised 19 October

1995: accepted

19 October

b ,

08193 BeJlatwra.

1995

Abstract Chromosome analysis by pulsed-field gel electrophoresis shows a high level of genetic heterogeneity between the two subspecies of the green sulfur bacteria Chlorobium limicofu analysed: C. limicola and C. limicola f. s. thiosujfatophilum. Currently, they are differentiated only by the ability to utilize thiosulfate as photosynthetic electron donor, and, by their %G + C content (51% and 58.1%, respectively). However, the capacity to utilize thiosulfate as photosynthetic! electron donor does not appear to be a useful criterion to differentiate between some strains of this species, because this lability is encoded by plasmids that are different depending on the thiosulfatophilum strain analysed. In contrast, this study reveals that the comparison of chromosomal restriction patterns is very useful as an additional aid for the differentiation and identification of C. limicola strains. Keywords:

Chlorobium

striction patterns;

/imicoLa:

Green sulfur bacteria:

Genomic

1. Introduction Chlorobium limicola is an anaerobic and photolithotrophic bacterium, which requires light as energy source and suitable electron donors, such as hydrogen sulfide or elemental sulfur [l]. It includes two formae speciales (f. sp.): C. limicola (Tie-) and C. limicola f. sp. thiosulfatophilum (Tie+) [ 1,2], taxonomically separated by their molar G + C content (5 1% and 5&l%, respectively) and their differ-

’ Corresponding author. Tel.: + 34 (3) 581 1278; Fax: + 34 (3) 58 I 201 I ; E-mail: [email protected]. Federation

heterogeneity;

Pulsed-field

gel electrophoresis;

Chromosomic

macrore-

Plasmid

of European

Microbiological

SSDI 0378.1097(95)00420-3

Societies

ent ability to utilize thiosulfate as electron donor (S,O:-) [3,4]. The analysis of this phenotypic property has provided the basis for the classification of strains within this species. However, molecuhu techniques may provide objectivity and a criterion for taxonomic agreement on the basis of genetic relatedness. Nevertheless, even though in the last few years genetic tools, including conjugation [5], eledtroporation [6], transformation [7], and gene cloning studies [8] have been used, little is known about the igenetics of Chlorobium [9-121. We have previously described that some formae speciales rhiosulfa[ophilum possess a 14-kb plasmid that confers the ability to

280

S. M&de:-All,ure:

et al. / FEMS Microhiolo~~

utilize thiosulfate [7]. However, no other differences in the genetic organization of both tki~~.~ulfat~~phi~um and no thio.su(fatophilum strains have been established to date. Therefore, there appears to be a need for correlation of the phenotypic and genomic differences existing between strains able or unable to use thiosulfate. This paper reports the comparative analysis of chromosome macrorestriction patterns, studied by PFGE, of C. limicola DSM 245 (Tie-), C. limicola f. sp. thiostdfatophilum DSM 249 (Tie+) and three isolated strains. Moreover, it establishes the existence of a new high molecular mass extrachromosoma1 element probably related with thiosulfate metabolism. However, since only five strains were taken into account (in fact, only a few strains of C. limicola are kept in collection), more strains have to be analysed to confirm the taxonomic value of the genetic features described.

2. Materials

Lrtters

134 ( I9051 279-285

mEinstein m -’ s- ’ continuous illumination. Cultures were grown in Pfennig minimal medium at pH 6.6-6.9 [13] to about IO’ cell mll ’ 2.2. DNA preparation tion

and restriction

Genomic DNA was embedded in low-meltingpoint agarose to prepare DNA inserts as described by Smith et al. [ 141. Prior the restriction digestion of the chromosomal DNA, the cell debris and proteinase K were removed by subjecting agarose blocks to PFGE (150 s pulse time, 330 V cm-‘). Blocks were then stored with TE buffer at 4°C. Restriction assays were carried out as described [ 15,161. Restriction enzymes, selected for their specific cleavage sites, were purchased from New England Biolabs (Beverly, MA) and Boehringer Mannheim (Mannheim, Germany). 2.3. PulsedTfield gel electrophoresis

and methods

2.1. Strains and growth conditions The bacterial strains studied are listed in Table I. New isolates were identified on the basis of their biochemical, morphological and ultrastructural properties [2]. Strains UdG6038 and UdG6041 were analysed at the Department of Microbiology of the University of Girona. Cultures were grown photolithoautotrophically in rubber cap bottles at 25°C and 50

Table I Bacterial strains ’ Strain

Source

Thiosulfate utilization

% G+C

C. limicolr C. limicoh f. thiosul~ltophilurn BF 8000 h UdG6038 ’ UdG6041’

DSM DSM UAB’ UdG’ UdG

+ + + -

5 I .o 58.1 ND ND ND

’ Symbols: -, negative: +, positive; ND, not determined. h Strain isolated by M. Alguer6. UAB. UAB, Universitat Autbnoma de Barcelona. ’ Strain isolated by Imma Pibernat. UdG. ’ UdG, Universitat de Girona. ’ Strain isolated by Imma Pibernat, UdG.

enzyme diges-

(PFGEI

PFGE [ 171 was performed in a Pharmacia-LKB apparatus. Gels were made of 1% agarose (SeaKern LE Agarose, FMC, Rockland, USA), and run at 15°C in modified TBE buffer (100 mM Tris, 100 mM boric acid, 0.2 mM EDTA, final pH 8-8.4). Different resolution windows were obtained by varying the pulse time for the different restriction endonucleases assays. Saccharomyces cereLisiae chromosomes, phage A concatemers and phage A DNA digested with the restriction endonucleases EcoRI and Hind111 were used as size standards. For topological studies, chromosomes of S. pombe were used as markers, and DNA inserts were run at 4500 s pulse time for 168 h. at a field strength of 3 V cm- ’ [IS]. 2.4. Plasmid analysis Plasmids smaller than 50 kb were detected by conventional electrophoresis when the DNA isolated by the alkaline method [ 151 was electrophoresed through a 0.7% agarose gel for 6 h under a constant electric field of 5 V cm- ’ . The presence of plasmids larger than 50 kb was studied in samples subjected to PFGE, at a field

strength of 10 V cm-’ using different pulse times (loo- 150 s) without restriction cleavage. 2.5. Southern

blot

at 68°C. The last washing step at 68°C ensures a high level of homology when a hybridization signal is detected.

analysis

Fractionated DNA was transferred onto Hybond-N Nylon membranes (Amersham) as described [ 191. The DNA probes were labelled with the DIG system (Boehringer Mannheim, Mannheim, Germany) according to the instructions of the manufacturer. Hybridization experiments with the DNA probes were performed in 5 X SSC, 2.0% (w/v) Blocking Reagent for nucleic acid hybridization, 0.1% Nlaurylsarcosine, 0.02% SDS at 68°C. Filters were washed twice, first in 0.1% SDS, 2 X SSC (0.3 M NaCl, 30 mM sodium citrate, pH 7.0) at room temperature and afterwards in 0. I % SDS, 0. I X SSC

3. Results 3. I. Restriction

endonuclease

and C. limicola

thiosu~fatophilum

23.1kb m

of C. limicola

chromosotnk

DNA

In order to compare the restriction pattern of the chromosomes from different strains of C. limicola, a battery of different restriction endonucleases was assayed: AseI, BclI, DraI, EcoRV, Pact, F$iI and SwaI. Among all of them we decided to utilize the restriction endonucleases PacI and SwaI, because they gave the best resolution pattern in PFGE. Chro-

B

242Skbw

analysis

23

5M

4242.5kb

-23.1 kb

Fig. I. Chromosomic DNA of C. limiroku DSM 249 (2), DSM 245 (3). BF8000 (I), UdG6041 (lane 4) and UdG6038 (lane 5) digested with SwI. Molecular mass marker was A concatemers plus A DNA digested with Hind111 (M). Gel (A) was run at 5-35 s pulse time for 24 h a~ a field strength of 2 V cm-’ (CHEF system). Gel (B) was run at S-SO s pulse time for 24 h at a field strength of 2.5 V cm” (OFAGE system).

282

S. Mhdez-Alvarez et al. / FEMS Microbiology L.erters134 ( 1995) 279-285

mosomic DNA of C. limicola DSM 245 (Tie-) and C. limicola f. sp. thiosulfatophilum DSM 249 (Tie+) generated a few DNA fragments, when digested with these two enzymes. When inserts of C. limicola f. sp. thiosulfatophilum DSM 249 (Tie’) DNA were cleaved with restriction endonucleases PucI and SwaI and then subjected to PFGE at different pulse times (from 5 to 80 s), 6 (data not shown) and 1l (lanes 2 in Fig. 1) bands were identified, respectively. In the case of C. limicola DSM 245 (Tie-), the restriction endonuclease PucI generated 7 bands (data not shown), and SwaI yielded 16 (lanes 3 in Fig. I). Fig. 2 is a semilogarithm representation comparing the restriction patterns of the two strains.

Table 2 Size of the chromosomic restriction fragments from C. limicoh DSM 245 and C. limicofu DSM 249 for the enzymes PacI and SW1 DSM 249

DSM 249

DSM 245

DSM 245

Pacl

SWUI

SWUI

Pm1

860 560 485 441 271 60

804 556 271 224 218 194 176 88 40 20 15

3.2. Determination cola DSM DSM 249

of chromosomic sizes of C. limi245 and C. limicolu thiosulfatophilum

The sizes of the chromosomes were estimated by summing the individual fragment lengths in each restriction endonuclease digest. For each restriction analysis of a strain, the whole range of fragment size was subdivided into several regions, and band positions were determined from the gel with optimum

4

Fig. 2. Semilogarithmic representation of the chromosomic restriction patterns of C. limicola DSM 249 (I. 3) and DSM 245 (2, 4) for the endonucleases PncI (3, 4) and SwaI (I, 2).

2677

2616

400 400 291 291 242 200 170 115 13 65 60 50 45 40 17 15 2414

915 650 300 225 200 125 100

2515

resolution in the respective molecular mass range. The size of each DNA fragment was determined by calibration using linear DNA molecular markers as reference. Chromosomic DNA of C. limicola DSM 245 (Tie-) digested with either PucI or SwaI generated fragments ranging from 100 to 9 15 kb or from 15 to 400 kb, respectively (Table 2). In the case of C. limicola thiosulfatophilum DSM 249 (Tie+), PucI generated fragments ranging from 60 to 860 kb, and SwaI yielded fragments ranging from 15 to 804 (Table 2). The total chromosome size of both strains was calculated from the sum of the fragment sizes generated by the digested chromosomic DNA. The size of the C. limicola f. sp. thiosulfatophilum DSM 249 (Tie+) chromosome is 2646.5 + 30.5 kb, and that of C. limicola DSM 245 (Tie-) is 2464.5 f 50.5 kb. It must be noted that sometimes some faint bands can be observed in the SwaI digestions of chromosomic DNA of C. limicolu DSM 245 (Fig. 1: lanes A3 and I33). They probably correspond to any not removed extrachromosomal element that has never been detected or to partially digested chromosomic DNA.

S. M&de:-Akarez

3.3. Chromosome strains

identification

et al. / FEM.5 Microbiology

of the newly isolated

For comparison of chromosomic restriction patterns with newly isolated strains, the enzyme PacI was rejected, because it has a high exonuclease activity, which makes it difficult to obtain restriction patterns for some strains. Thus, the chromosome restriction patterns of C. limicola f. sp. thiosulfutophilum DSM 249 (Tie+) and C. limicola DSM 245 (Tie’) for the enzyme SwaI were compared with that of the three newly isolated strains in order to contrast their genetic assignation with that obtained by phenotypic characters. As expected, the DNA of strain BFSOOO (Tie+), which is able to utilize thiosulfate, has the same restriction pattern (Fig. 1A, lane 1) as C. limicolu f. sp. thiosulfutophilum DSM 249 (Tie+) (Fig. IA, lane 2). However, the restriction patterns of the strains UdG6038 (Tie’) (Fig. lB, lane 5) and UdG6041 (Tie-) (Fig. 1A, lane 4) are different from those of both C. limicola f. sp. thiosulfatophilum DSM 249 (Tie+) and C. limicola DSM 245 (Tie-).

Letters 134 (1995)

279-285

283

PFGE, and it was studied to detect any possible role in the thiosulfate metabolism. With that purpose, it was eluted from an agarose gel and labelled as described in Materials and methods. The resulting probe (probe 6038) was then hybridized with the 14-kb plasmid isolated from the thiosulfbtophilum strains DSM 249 (Tie+) and BF8000 (Tie+), and from C. limicolu DSM 245 transformants (Tie’). The same hybridization bands were observed as those given by the probe derived from the 14-kb plasmid (data not shown). The large size of the probe 6038 ensures that a hybridization signal is only detected if most of the target DNA is represented at the probe. This result suggests that this high molecular mass extrachromosomal element determines some functions related to the thiosulfate metabolism and that it confers the ability to utilize thiosulfate to strain UdG6038 (Tio’). As a negative control, strain UdG6041 (Tie-), which is unable to utilize thiosulfate, was analysed. No extrachromosomal elements were detected in this strain.

4. Discussion 3.4. Extrachromosomal sulfate metabolism

material

related to the thio-

We have previously described the isolation of a 14-kb plasmid from the forma specialis thiosuffutophilum strain C. limicola DSM 249 (Tie+). This plasmid confers the ability to utilize thiosulfate, as it was demonstrated by transformation of C. limicola DSM 245 (Tie-) [7]. Plasmid DNA isolated from transformants (Tie+) matched the C. limicola DSM 249 (Tie+) plasmid DNA in size, in restriction patterns for EcoRV and Hind111 and also hybridized with a labelled probe generated from C. limicola DSM 249 (Tie+) plasmid DNA (unpublished results). The presence of any extrachromosomal element related to thiosulfate utilization was examined in the two newly isolated thiosulfatophilum strains. As previously reported [7], the 14-kb plasmid was detected in the strain BF8000. The presence of such a plasmid has been examined also in strain UdG6038 (Tie+), which is also able to metabolize thiosulfate. Although the 14-kb plasmid was not detected in this strain, an extrachromosomal element of high molecular mass (approximately 650 kb) was isolated by

Several of the traditional phenotypic traits used for species characterization in anoxygenic photosynthetic bacteria have an uncertain taxonotmic value. Thus, a genotypically based classification ~of anoxygenie phototrophic strains, both above and’ below the species level, is urgently needed. One useful approach is the comparison of macrorestriction patterns, as we have done to contrast the phenotypic identification of three strains of the purple sulfur bacteria Chromatium ninosum [20]. In this report, the genomic organization of the two formae speciales of Chlorobium limicola were analysed by PFGE, including chromosomic macrorestriction patterns, chromosome sizes, and presence of extrachromosomal material related with sthiosulfate metabolism. The chromosome comparison presented for the strains C. limicola f. sp. thiosulfatophilum DSM 249 (Tie+) and C. limicolu DSM 245 (Tie-) was performed by analysing chromosomic macrorestriction patterns by PFGE. This is the first comparative study of the two formae at the genomic level. The results indicate a similar chromosome size for both (2646.5

284

S. Mlndez-Akarez

et al. / FEMS Microbiology

k 30.5 kb for C. limicola f. sp. thiosulfatophilum DSM 249 and 2464.5 + 50.5 kb for C. limicola DSM 245), but very different restriction patterns (Fig. 21, as expected due to their different %G + C (58.1% and 51%, respectively). In fact, strains belonging to the same bacterial species can not differ in more than 5% in their G + C content. Otherwise, the differences observed between the two formae with respect to the number of bands generated by PucI and SwaI amalso coherent with the %G + C. Thus, the lower G + C content of C. limicolu (Tie-) explains that these two enzymes, which contain only AT base pairs in their recognition sequences, have a slightly higher cleaving frequency. On the other hand, the other character used to differentiate these two formae speciales, which is the ability to utilize thiosulfate as electron donor, has been demonstrated to be encoded by a 14-kb plasmid [7] in strain C. limicolu f. sp. thiosulfatophilum DSM 249. Moreover, such a plasmid was easily transferred by transformation to the strain C. limicola DSM 245 [7], which suggests that the capacity to utilize thiosulfate has an uncertain taxonomic value. Comparison of chromosomic restriction patterns analysed by PFGE has been reported to be a useful tool for the identification of strains at the species [21] and subspecies [22] level, and for determining genetic relatedness between species of the same genus [23]. For that reason, SwaI chromosomic restriction patterns from five different strains of C. limicola were analysed by PFGE, obtaining four different profiles. These results indicate a high genetic heterogeneity between different strains phenotypically assigned to the species C. limicola. Due to the importance that the character ‘ability to use thiosulfate’ has had until now, the presence of the 14-kb plasmid or any related plasmid was analysed in all thiosulfatophilum strains. This plasmid has been previously detected in the strain BF8000 [7], whereas it was not observed in the thiosulfutophilum strain UdG6038. However, a high molecular mass extrachromosomal element, which hybridized with the 14-kb plasmid from C. limicolu DSM 249, was isolated from strain UdG6038, which suggests that this element is also related to thiosulfate metabolism. Nevertheless, further analyses must be performed to demonstrate this suggestion. If it

Lztters 134 llYY51270-285

would be confirmed, we could conclude that the ability of C. limicola to utilize thiosulfate is encoded by the extrachromosomal material present in each strain. In conclusion, the comparison of restriction patterns by using PFGE indicates that various strains phenotypically assigned to the species C. limicolu contain high differences in their chromosomal primary structure. In fact, phenotypic characterization of Chlorobium strains is based on a few characters of uncertain taxonomic value. Therefore, it may be useful to analyse the chromosomes from the different Chlorobium species in order to clarify the taxonomic assignment of strains belonging to this genus.

Acknowledgements We are grateful to the Department of Biology of the University of Girona for providing strains UdG6038 and UdG6041 cultures. We thank J. Checa for technical assistance and M. Piqueras for critically reading the manuscript. This work was supported by CICYT grant PB94-0730 to I.E. and MAR 91-0874 to R.G. Research by S.M. was supported by a scholarship of the Autonomous Government of Catalonia.

References [I] Pfennig,

[2]

[3] [4]

[5]

[6]

N. and Ttiiper, H.G. (1992) The family Chlorobiaceae. In: The Prokaryotes (Ballows, A., Triiper, H.G., Dworkin, M., Harder, W. and Schleifer, K., Eds.), pp. 35833592. Springer-Verlag. Berlin. Pfennig, N. (1989) Green sulfur bacteria. In: Bergey’s Manual of Systematic Bacteriology (Murray, R.G.E., Brenner, D.J., Bryant, M.P., HOI, J.G., Krieg, N.R., Moulder, J.W., Pfennig, N., Sneath, P.H.A., Staley, J.T. and Williams, ST., Eds.), pp. I682- 1709. Williams & Wilkins, Baltimore, MD. Brune. D.C. (1989) Sulfur oxidation by phototrophic bacteria. Biochim. Biophys. Acta 975, 189-221. TrUper. H.G., Lorenz, C., Schedel, M. and Steinmetz, M. (1988) Metabolism of thiosulfate in Chlorobium. In: Green Photosynthetic Bacteria (Olson, J.M., Ormerod, J.G., Amesz, J., Stackebrandt, E. and Triiper, H.G., Eds.), pp. 189-200. Plenum Press, New York, NY. Wahlund, T.M. and Madigan, M.T. (1995) Genetic transfer by conjugation in the thetmophilic green sulfur bacterium Chlorobium tepidum. J. Bacterial. 177. 2583-2588. Kjaerulff, S., Diep, D.B.. Okkels, J.S., Scheller, H.V. and Ormerod, J.G. (1994) Highly efficient integration of foreign

S. Me’ndez-Akurez

[7]

[8]

[9]

[IO]

[I l]

[12] [ 131

(141

[15]

et al. / FEMS Microbiology

DNA into the genome of the green sulfur bacterium Chlorobium l?briufonne by homologous recombination. Photosynth. Res. 4 I, 211-283. MCndez-Alvarez, S., Pavbn, V., Esteve, I., Guerrero, R. and Gaju, N. (I 994) Transformation of Chlorobium limicola by a plasmid that confers the ability to utilize thiosulfate. J. Bacterial. 176, 7395-7397. Avissar, Y.J. and Beale, S.I. (1990) Cloning and expression of a structural gene from Chlorobium ~ibrioforme that complements the hemA mutation in Escherichiu coli. J. Bacteriol. 172, 1656- 1659. Donohue, T.J. and Kaplan, S. (1991) Genetic techniques in Rhodospirillaceae. In: Bacterial Genetic Systems. Methods Enzymol. 204, 459-485. Marrs, B.L. (1983) Genetics and molecular biology. In: The Phototrophic Bacteria: Anaerobic Life in the Light (Ormerod, J.G., Ed.), pp. 186-214. Oxford University Press, Oxford. Saunders, U.A. (1992) Genetics of the photosynthetic prokaryotes. In: Photosynthetic Prokaryotes (Mann, N.H. and Carr, N.G., Eds.). pp. 121-145. Plenum Press, New York, NY. Scolnik, P.A. and Marrs, B.L. (1987) Genetic research with photosynthetic bacteria. Annu. Rev. Microbial. 41, 703-726. Van Gemerden, H. and Beeftink, H.H. (1983) Ecology of phototrophic bacteria. In: The Phototrophic Bacteria (Ormerod, J.G., Ed.), pp. 146- 185. Blackwell, Oxford. Smith, CL. and Cantor, C.R. (1987) Purification, specific fragmentation and separation of large DNA molecules. In: Recombinant DNA. Methods Enzymol. 155, 449-467. McClelland, M., Jones, M.. Patel, Y. and Nelson, M. (1987)

[ 161

[17]

[18]

[19]

[20]

[2 I ]

[22]

]23]

Letters

134

C1995)279-285

285

Restriction endonucleases for pulsed field mapping of bacterial genomes. Nucleic Acids Res. 15, 5985-6005. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. 2nd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Schwartz, D.C. and Cantor, C.R. (1984) Separation of yeast chromosome sized DNAs by pulsed field gel gradient electrophoresis. Cell 37, 67-75. Smith, C.L. and Condemine, G. (1990) New approaches for physical mapping of small genomes. J. Bacterial. 172. I l671172. Smith, C.L., Klco, CR. and Cantor, C.R. (1988) Pulsed field gel electrophoresis and the technology of large DNA molecules. In: Genome Analysis: A Practical Approach (Davies, K.E., Ed.), pp. 21 l-272. IRL, Oxford. Gaju, N., Pa&n, V., Marin, I., Esteve, I., Guerrero, R. and Amils, R. (1995) Chromosome map of the phototrophic anoxygenic bacterium Chromutium rinosum. FEMS Microbiol. Lett. 126, 241-248. Frey, J., Haldimann, A. and Nicolet, J. ( 1992) Chromosomal heterogeneity of various Mvxplasma h~poneumoniat~ field strains. Int. I. Syst. Bacterial. 42, 275-280. Sameeh, S.M., Garcia, M.M. and Taylor, D.E. (1992) Differentiation of the subspecies of Camp~lobactrr fetus by genomic sizing. Int. J. Syst. Bacterial. 42, 446-450. Khattak, M.N. and Matthews, R.C. (1993) Genetic relatedness of Bordetellu species as determined by macrarestriction digests resolved by pulsed-field gel electrophoresis. Int. J. Syst. Bacterial. 43, 659-664.