aDipartimento di Biologia Animale e Ecologia Marina, Salita Sperone 31, 98166 Messina, ItalybDipartimento di Biologia Animale e Genetica, Via Romana 17-19, 50125 Firenze, Italy
(Submitted 17 October 1988; accepted 1 December 1998)
The marine luminous bacteria are microor-ganisms of great interest because of their abilityto emit visible light, uncommon within theprokaryotes, and their ubiquity in the marineenvironment as planktonic or free-living,saprophytic, light organ- and gut-symbiotic(commensal), and even pathogenic [18].
The marine species known thus far are repre-sentative of the genera Vibrio, Photobacteriumand Shewanella, all belonging to the family Vibri-onaceae within the γ subdivision of the Proteo-bacteria. In particular, four clusters of speciescan be distinguished according to theirDNA/DNA similarity: i) the Vibrio harveyigroup (V. harveyi, V. splendidus, V. orientalis); ii)the Vibrio fischeri group (V. fischeri, V. logei); iii)the Photobacterium group (P. leiognathi, P. phos-phoreum); iv) Shewanella hanedai; and v) Vibriovulnificus [18].
In the past, identification of marine luminousbacteria based only on a phenotypic approachwas often equivocal because of their highlyvariable phenotypes. Recent molecular tech-niques, including hybridization with luxAprobes [12, 28] and analysis of protein-coding
* Correspondence and reprintsTel.: 0039 90 676 5523; fax: 0039 90 393 409;[email protected]: ARDRA, amplified ribosomal DNA restric-tion analysis; MA, bacto-marine agar (Difco); MB, bacto-marine broth (Difco); NCIMB, National Collection of In-dustrial and Marine Bacteria; PCR, polymerase chainreaction; RAPD, random amplified polymorphic DNA;RDP, Ribosomal Data Project; Ta, annealing temperature,TAE; Tris-acetate-EDTA (buffer); TE, Tris-EDTA (buffer).
sequences [24], improved identification but re-vealed certain limitations. Hybridization withluxA probes, for example, was highly species-specific for the majority of marine luminousisolates except for V. harveyi. Indeed, the V.harveyi probe showed cross-reactivity with luxAgenes of two closely related species: V. vulnificusand V. orientalis [28]. Two V. vulnificus biotypesare known thus far [23]: biotype 1, a clinicalstrain, is an opportunistic human pathogen;biotype 2, an environmental strain, is primarilyan eel pathogen, but also an opportunisticpathogen for humans [2]. Some V. harveyi and V.splendidus strains were recognised pathogensfor penaeid larvae and snooks in hatcheries [3,13], and thus careful recognition of these speciesin sea water samples is necessary.
A molecular approach was also used fordetermining phylogenetic relationships amongtype strains of the family Vibrionaceae on thebasis of small subunit rRNA gene sequencing.Comparison of the 16S rDNA sequences led toconstruction of a phylogenetic tree for the genusVibrio and related genera [8, 11, 20].
In the present paper, we report the identifica-tion and characterisation of 15 luminous bacte-rial strains isolated from near-shore TyrrhenianSea waters off the northeastern coast of Sicily,using a polyphasic approach based on a combi-nation of physiological tests and moleculartechniques including restriction analysis of am-plified 16S rDNA (ARDRA) [25], determinationof the 16S rDNA sequence [22], random ampli-fied polymorphic DNA (RAPD) [26, 27] andanalysis of plasmid content.
2. Materials and methods
2.1. Sampling and cultivation
The samples were collected from differentsites in coastal waters of the Tyrrenhian Sea offnortheastern Sicily. Sampling and bacterial cul-tivation and isolation were carried out accord-ing to Maugeri [17].
2.2. Bacterial strains
Twenty bacterial strains listed in table I wereused in this work: 15 of them were luminous
strains isolated from Tyrrhenian Sea waters andfive were luminous type strains purchased fromNCIMB.
2.3. Phenotypic characterisation
The isolates were examined for 58 morpho-logical, biochemical and physiological charac-ters according to Alsina and Blanch [1] andCaruso et al. [4]. Enzymatic activities and car-bohydrate fermentation were tested by the min-iaturized kits API 20E and API 20NE(bioMérieux). Glucose oxidation/fermentationwas tested in Leifson OF/F medium modifiedfor marine bacteria.
2.4. Amplification of 16S rDNA
The strains were grown overnight on MA at20 °C. A single colony of each strain was pickedup with a sterile toothpick, suspended in 20 µLof sterile distilled water and lysed by heating at95 °C for 10–20 min. After lysis, the cell suspen-sion was cooled in ice, briefly spun in a micro-centrifuge and used for the PCR.
Amplification of 16S rDNA was performedusing 2 µL of each lysed cell suspension accord-ing to the procedure described previously [6].The amplification products were analysed byagarose gel (1.2% w/v) electrophoresis in TAEbuffer (0.04 M Tris-acetate, 0.001 M EDTA),containing 1 µg/mL of ethidium bromide.
Primers P0 and P6 (table II) were designed onthe basis of the conserved eubacterial sequences
Table I. Bacterial strains used in this study.
Strains Source orreference Strains Source or
reference
Lu 1 This work Lu 12 This workLu 2 This work Lu 15 This workLu 3 This work Lu 17 This workLu 4 This work Lu 18 This workLu 5 This workLu 6 This workLu 7 This work V. splendidus NCIMB 1Lu 8 This work V. harveyi NCIMB 1280Lu 9 This work V. fischeri NCIMB 1281Lu 10 This work P. leiognathi NCIMB 2193Lu 11 This work V. orientalis NCIMB 2195
222 Caccamo et al.
and are located at the 5’ and 3’ ends of 16SrDNA, enabling amplification of nearly all ofthe gene. The primers were synthesized bystandard phosphoramidite chemistry, depro-tected, dried, dissolved in TE buffer (10 mMTris-HCl, pH 8.0, 1 mM EDTA), and used with-out further purification.
2.5. Restriction analysis of amplified 16S rDNA(ARDRA)
Approximately 1.5 µg of amplified 16S rDNAwas cleaved with 3 units of the restrictionenzyme AluI (Bohringer Mannheim) in a totalvolume of 20 µL at 37 °C for 3 h. The enzymewas then inactivated by heating the reactionmixtures at 65 °C for 10 min. The reaction prod-ucts were analysed by agarose gel (2.5% w/v)electrophoresis in TAE buffer (0.04 M Tris-acetate, 0.001 M EDTA) 1X with 1 µg/mLethidium bromide.
2.6. Sequencing of 16S rDNA
The determination of the 16S rDNA nucle-otide sequence was performed by the method ofSanger et al. [21] using the Promega fmoly(thermal cycle) DNA Sequencing Kit andα-dATP 35S to label the DNA synthesized. Theamplified 16S rDNA was purified from thereaction mix by agarose gel (0.8% w/v) electro-phoresis in TAE 1X with 1 µg/mL of ethidiumbromide. A small agarose slice containing theband of interest (observed under long UV,312 nm) was excised from the gel and purifiedusing a QIAquick Gel Extraction kit (Quiagen)according to the manufacturer’s instructions.
The reactions were performed using theprimers listed in table II, which were designedon the basis of conserved eubacterial sequences.Each primer that did not show palindromicsequences longer than four bases or comple-mentarity at the 3’ extreme was denatured at90 °C for two min before use. The annealingtemperature (Ta) used in cycle sequencing was50 °C. Reactions were performed for 25 thermalcycles according to the instruction manual.
2.7. Analysis of sequence data
The 16S rDNA nucleotide sequences obtainedwere aligned with the most similar ones of theRDP [14] using RDP utilities. The alignmentwas manually checked and analysed.
2.8. Nucleotide sequence accession numbers
The 16S rDNA nucleotide sequences obtainedwere deposited in GenBank with accessionnumbers from AF094701 to AF094710. Numbersrelative to the sequences of luminous referencestrains are listed in table III.
2.9. RAPD fingerprinting
Random amplification was performed on2 µL of each cell lysate in a 25 µL final volume of1 × amplification buffer Mg-free (Promega)supplemented with 3 mM MgCl2, 500 ng ofprimer, 200 µM of each dNTP and 0.625 U of TaqDNA polymerase (Promega). The reaction mix-tures were incubated in a thermal cycler (GeneAmp PCR System 9600 Perkin-Elmer) at 90 °Cfor 1 min and at 95 °C for 1 min 30 s, then cycled
Table II. Oligonucleotides used as primers for 16S rDNA amplification and sequencing.
The position indicates the annealing of the primer 3’end to the E. coli 16S rDNA in forward (f) or reverse (r) orientation.
Characterisation of marine luminous bacteria 223
45 times through the following temperatureprofile: 95 °C for 30 s, 36 °C for 60 s, 75 °C for2 min; finally, the reactions were incubated at75 °C for 10 min, then at 60 °C for 10 min andstored at 4 °C.
The reaction products were analysed by aga-rose gel (2% w/v) electrophoresis in TAE buffer1X with 1 µg/mL ethidium bromide.
Two different 10-nt primers were used: AP5(5’-TCCCGCTGCG-3’) and AP12 (5’-CGGCCCCTGC-3’). Primers were synthesizedby standard phosphoramidite chemistry, depro-tected, dried, dissolved in TE buffer (10 mMTris-HCl pH 8.0, 1 mM EDTA) and used with-out further purification.
2.10. Analysis of plasmid content
Plasmids were extracted from cell pellets of3 mL cultures in MB using the Miniprep extrac-tion kit (BioRad) and analysed by agarose gel(0.8% w/v) electrophoresis.
3. Results
3.1. Phenotypic characterisation
All strains were motile rod-shaped Gram-negative cells, oxidase-positive and O/129
(10 µg and 150 µg)-sensitive. They gave identi-cal responses to the following cultural andbiochemical tests: growth at NaCl 0% (-),growth up to NaCl 8% (+), growth at 4 °C (-),growth at 45 °C (-), acid from: glucose (+),arabinose (-), inositol (-), rhamnose (-), sorbitol(-), assimilation of: arabinose (-), N-acetyl-glucosamine (-), caprate (-), adipate (-), growthon lactate (+), nitrate reduction (+), sulphideproduction (-), Voges-Proskauer reaction (-),ONPG (+), tryptophan-deaminase (-), gas fromglucose (-).
The results from other tests are reported intable IV.
The cluster analysis was performed on 32selected characters (table IV) by using the Jac-card similarity coefficient (SJ) and the un-weighted average linkage clustering method.
The resulting dendrogram (figure 1) showedthe correlation (SJ > 0.80) of all fifteen isolateswith V. harveyi NCIMB 1280 and their subdivi-sion into four clusters or phena.
3.2. ARDRA
When the 16S rDNA of each of the 20 bacte-rial strains was amplified by PCR, an amplifi-cation fragment of about 1450 bp (not shown)was obtained. ARDRA was performed with theenzyme AluI which often generates species-specific restriction patterns, enabling bacterialstrains to be divided into groups correspondingto a given species [5-7, 9, 15].
Table III. 16S rDNA nucleotide sequences of luminous speciescollected from RDP.
Genbankentry
Genbankentry
P. phosphoreum D11186 V. splendidus Z21731X74687 X74724Z19107 V. orientalis Z21731
P. leiognathi D11184 D11215X74686 X74719
V. cholerae D11199 V. logei D11207D11198 X74708D11197D11196 V. fischeri X70640X74695 D11202X74697 X74702D11195
V. vulnificus X74727 V. harveyi D11205X74726 M58172X56582 X56578
Figure 1. Dendrogram resulting from cluster analysis on 32phenotypic characters by using the SJ coefficient and the un-weighted average linkage method.
224 Caccamo et al.
Three different ARDRA patterns were recog-nised (figure 2 and table V) within the 15 iso-lates. Ten environmental strains showed a re-striction pattern identical to that of the referencestrain V. harveyi, NCIMB 1280 (group A), onestrain showed identity with the reference strainV. splendidus NCIMB 1 (group B), the remaining4 strains (group C) exhibited a pattern differentfrom those of the reference strains.
Table IV. Results of phenotypic characterisation tests.
Strains Lu Lu Lu Lu Lu Lu Lu Lu Lu Lu Lu Lu Lu Lu Lu V.s. V.h. V.f. P.l. V.o.
Table V. Subdivision of the 15 luminous isolates into threeARDRA groups.
ARDRA group Bacterial strains Type strain pattern
A Lu 1, Lu 2, Lu 4, Lu 6,Lu 7, Lu 9, Lu 10,Lu 15, Lu 17, Lu 18
V. harveyi 1280
B Lu 3 V. splendidus 1C Lu 5, Lu 8, Lu 11,
Lu 12---
Characterisation of marine luminous bacteria 225
3.3. 16S rDNA sequencing and analysis
Data obtained from 16S rDNA restrictionanalysis were confirmed by determining the 16SrDNA sequence of seven randomly chosenstrains showing the same ARDRA pattern ofV. harveyi NCIMB 1280 (group A), from position350 to 450. This region, according to the RDPand Kita-Tsukamoto [11], contains a sequencewhich may be considered as a signature forluminous Vibrio species. As shown in figure 3,all seven strains showed, in this region,
a 16S rDNA sequence typical of V. harveyi.Moreover, the almost complete 16S rDNA se-quence of the isolate Lu1 (belonging to the sameARDRA group) was determined. The analysis ofthe sequence confirmed previous results allow-ing the assignment of Lu1 to the species V. har-veyi.
The four members of ARDRA group C, on thebasis of the Lu8 full sequence and the sequenceof the Lu5 signature region as well, were shownto be closely related to V. harveyi and to group A.
Figure 2. Agarose gel electrophoresis of amplified 16S rDNA digested with restriction endonuclease AluI. L, DNA ladder 100 bp.
Figure 3. 16S rDNA alignment of Vibrio luminous species, from position 350 to 450 (according to E. coli numeration). The sequenceof this region, according to the RDP [11, 14], may be considered as a signature for luminous Vibrio species.
226 Caccamo et al.
3.4. RAPD fingerprinting
The degree of intraspecific genetic variabilityamong strains of the main ARDRA group wasinvestigated by RAPD fingerprinting analysis oftotal DNA of each strain in two experimentswith primers AP5 and AP12. In particular, am-plification with primer AP5 showed a mono-morphic band (300 bp) in nine of the 10 envi-ronmental isolates in the group A (figure 4).
3.5. Analysis of plasmid content
Analysis of plasmid content showed the pres-ence of plasmid molecules of different sizes in
some strains belonging to ARDRA group A.Hybridization carried out using as probe theplasmid DNA of isolate Lu4 showed a correla-tion with plasmid molecules of four isolateshaving different RAPD patterns (figure 5).
4. Discussion
Preliminary results from phenotypic charac-terisation tests (table IV) enabled us to classifythe fifteen luminous isolates analysed in thiswork into the family Vibrionaceae.
The absence of PHB accumulation and thepositivity of the following diagnostic tests [18]:
Figure 4. RAPD patterns of genomic DNA of the tyrrhenian strains. On the left, amplification with the primer AP12; on the right,amplification with the primer AP5.
Figure 5. Southern blot hybridization of plasmid extracted from tyrrhenian luminous strains. Plasmid DNA from isolate Lu4 was usedas probe.
Characterisation of marine luminous bacteria 227
growth up to NaCl 10%, amylase, lipase, gelati-nase, growth on cellobiose, mannitol, propi-onate and L-tyrosine, excluded their belongingto the genus Photobacterium. In particular, thepresence of P. phosphoreum, unusual in Tyrrhe-nian Sea coastal waters due to temperaturesbeing higher than 4 °C, had been previouslyexcluded by negative growth at 4 °C and nega-tive gas production from glucose.
Negative growth at 4 °C, growth up to NaCl10%, indole production and positive gelatinasealso excluded the presence of V. logei.
The correlation between the isolates groupedinto four phena and the V. harveyi NCIMB 1280type strain, shown in the dendrogram obtainedby cluster analysis on 32 selected phenotypiccharacters (figure 1), was further investigated bymolecular methods.
Molecular analysis (ARDRA and 16S rDNAsequencing) enabled the 15 isolates to be subdi-vided into three groups (table V).
ARDRA group A included ten isolates whichwere assigned on the basis of full or partialsequencing of the 16S rDNA to the speciesV. harveyi, in agreement with both ARDRA andphenotypic analyses.
The four isolates belonging to ARDRA groupC appeared to be closely related to the V. harveyitype strain. Comparison of the full 16S rDNAsequence obtained with those available in data-banks precluded Lu8 as belonging to the speciesV. harveyi. Analysis of the signature regions,with special reference to that of Lu5, neverthe-less suggested that group C could be considereda subspecies of V. harveyi or a tyrrhenian biovardifferent from other reference strains, the 16SrDNAs of which have already been sequenced.The similarity between the restriction patternsof ARDRA group A and group C (figure 2) lendsupport to this hypothesis.
Analysis of the signature region also sug-gested that strains belonging to ARDRA groupC could not be assigned to the species V. vulnifi-cus or V. logei, the reference strains of whichwere not included in this work.
These data are in agreement with phenotypicdata which led to the grouping of species be-longing to ARDRA group C into two phena (2
and 3), separated from phena 1 and 4, includingthe species belonging to ARDRA group A (fig-ure 1).
Some discrepancies neverthless remain, i.e.,grouping of the strains Lu17 (ARDRA group A)and Lu3 (ARDRA group B), the latter identifiedas a V. splendidus species, into phena 3 and 2respectively (figure 1). These anomalies wereexpected because of the great variety of factorsaffecting phenotypic expression of environmen-tal bacteria, the main reason why a classificationbased only on phenotypic data usually providesan inadequate representation of prokaryotictaxa.
As suggested by RAPD analysis (figure 4)carried out on members of ARDRA group A, thedegree of intraspecific biodiversity in luminousstrains is so high that we may be in the presenceof ‘regional’ biodiversity, and new and differentbiovars of type strains might be defined inrelation to the different environmental sources.This was previously hypothesized by Martin-Kearley and Gow [16], who, in the Newfound-land coastal waters, found Vibrio species differ-ent from the previously described type strains.
The presence of plasmid molecules in differ-ent strains of the same species, as demonstratedfor some members of ARDRA group A (figure 5),and the correlation between these plasmidsdetected by Southern hybridization, suggestedthe possibility of horizontal gene transfer thatsupports this biodiversity.
Data obtained suggest that ARDRA with en-zyme AluI could represent a powerful tool forthe identification of the luminous bacteria, sinceit gives different restriction patterns for eachreference strain, enabling discrimination be-tween closely related species, i.e., V. harveyi,V. orientalis and V. splendidus.
Further information on identification can alsobe provided by sequencing in the 16S rDNAsignature region (350, 450) that shows basesubstitutions particular to each luminous spe-cies (figure 3).
A combination of these two techniques couldbe a valuable strategy for rapid and reliableidentification of luminous bacterial strains iso-lated from the marine environment, in place of
228 Caccamo et al.
lengthy phenotypic characterisation and certainmolecular methods employed thus far givingequivocal results. Moreover, the use of thisstrategy could be highly relevant in other taxo-nomic studies in order to achieve better resolu-tion of discrimination between closely relatedbacterial species.
Finally, biodiversity observed in luminousstrains isolated from the Tyrrhenian Sea, thoughpoorly explored up to now, may be useful formany biotechnological applications, as reportedby Hill et al. [10]. Roda et al. [19] used a V.harveyi strain isolated from marine sea waternear the Aeolian Islands (Sicily) as a sensitivemodel biotest in cytotoxicity assays based onluminescence inhibition by toxic compounds.
[1] Alsina M., Blanch A.R., Improvement and update of a set of keys forbiochemical identification of Vibrio species, J. Appl. Bacteriol. 77(1994) 719–721.
[2] Amaro C., Biosca E.G., Vibrio vulnificus biotype 2 pathogenic for eelsis also an opportunistic pathogen for humans, Appl. Environ. Mi-crobiol. 62 (1996) 1454–1457.
[3] Baticados M.C.L., Lavilla-Pitogo C.R., Cruz-Lacierda E.R., De LaP-ena L.D., Sunaz N.A., Studies on the chemical control of luminousbacteria V. harveyi and V. splendidus isolated from diseased P. mon-odon larvae and rearing water, Dis. Aquat. Org. 9 (1991) 133–139.
[4] Caruso G., Zaccone R., Crisafi E., Distribution and numericaltaxonomy of Vibrionaceae in the waters of the Straits of Messina,Mi-crobiologica 19 (1996) 155–166.
[5] DiCello F., Pepi M., Baldi F., Fani R., Molecular characterisation of an-alkane degrading bacterial community and identification of a newspecies Acinetobacter venetianus, Res. Microbiol. 148 (1997)237–249.
[6] DiCello F., Bevivino A., Chiarini L., Fani R., Paffetti D., TabacchioniS., Dalmastri C., Biodiversity of a Burkholderia cepacia populationisolated from maize rhizosphere at different plant growth stages,Appl. Environ. Microbiol. 63 (1997) 4485–4493.
[7] DiCello F., Fani R., A molecular strategy for the study of naturalbacterial communities by PCR-based techniques, Minerva Biotech-nol. 8 (1996) 126–134.
[8] Dorsch M., Lane D., Stackebrandt E., Towards a phylogeny of thegenus Vibrio based on 16S rRNA sequences, Int. J. Syst. Bacteriol. 42(1992) 58–63.
[9] Grifoni A., Bazzicalupo M., DiSerio C., Fancelli S., Fani R., Identifi-cation of Azospirillum strains by restriction fragment length poly-morphism of the 16S rDNA and of the histidine operon, FEMSMicrobiol. Lett. 127 (1995) 85–91.
[10] Hill P.J., Denyer S.P., Stewart G.S.A.B., Rapid assays based on in vivobacterial luminescence, Microbiol. Eur.1 (1993) 16–21.
[11] Kita-Tsukamoto K., Oyaizu H., Nanba K., Simidu U., Phylogeneticrelationships of marine bacteria mainly members of the familyVibrionaceae determined on the basis of 16S rRNA sequences, Int. J.Syst. Bacteriol. 1 (1993) 8–19.
[12] Lee K., Ruby E.G., Detection of the light organ symbiont Vibriofischeri in hawaiian seawater by using lux gene probes, Appl. Environ.Microbiol. 58 (1992) 942–947.
[13] Liu P.C., Lee K.K., Chen S.N., Pathogenicity of different isolates ofVibrio harveyi in tiger prawn Penaeus monodon, Lett. Appl. Microbiol.22 (1996) 413–416.
[15] Martinez-Murcia A.J., Acinas S.G., Rodriguez-Valera F., Evaluation ofprokaryotic diversity by restrictase digestion of 16S rDNA directlyamplified from hypersaline environments, FEMS Microbiol. Ecol.17(1995) 247–256.
[16] Martin-Kearley J., Gow J.A., Numerical taxonomy of Vibrionaceaefrom Newfoundland coastal waters, Can. J. Microbiol. 40 (1994)355–361.
[17] Maugeri T.L., Planktonic luminous bacteria in the waters of theStraits of Messina, in: Guglielmo L., Manganaro A., De Domenico E.(Eds.), The Straits of Messina Ecosystem, Dipartimento di BiologiaAnimale ed Ecologia Marina Messina, 1995, pp.135–142.
Characterisation of marine luminous bacteria 229
[18] Nealson K.H., Hastings J.W., The Luminous Bacteria, in: Balows A.,Trüper H.G., Dworkin M., Harder W., Schleifer K.H. (Eds.), TheProkaryotes, Springer-Verlag New York, NY, 1991, pp. 625–639.
[19] Roda A., Maugeri T.L., Polimeni C., Bolognani I., Matteuzzi D.,Girotti S., Ferri E., Use of thermostable luminescent bacteria as arapid cytotoxicity test, J. Biolum. Chemilum. 9 (1994) 315.
[20] Ruimy R., Breittmayer V., Elbaze P., Lafay B., Boussemart O.,Gauthier M., Christen R., Phylogenetic analysis and assessment ofthe genera Vibrio Photobacterium Aeromonas and Plesiomonas de-duced from small-subunit rRNA sequences, Int. J. Syst. Bacteriol. 44(1994) 416–426.
[21] Sanger F., Nicklen S., Coulson A.R., DNA sequencing with chainterminating inhibitors, Proc. Natl. Acad. Sci. USA 74 (1977)5463–5466.
[22] Stackebrandt E., Goodfellow M,(Eds.), Nucleic acids techniques inbacterial systematics, John Wiley & Sons, New York, NY, 1991.
[23] Stelma G.N., Reyes A.L., Peeler J.T., Johnson C.H., Spaulding P.L.,Virulence characteristics of clinical and environmental isolates ofVibrio vulnificus, Appl. Environ. Microbiol. 58 (1992) 2776–2782.
[24] VanErt L.M., Wimpee C.F., Makemson J.C., Identification of lumi-nous marine bacteria based on sequence analysis of the lux A gene,in: Guerrero R., Pedrós-Alió C. (Eds.), Trends in Microbial Ecology,Spanish Society for Microbiology, Barcelona, 1993, pp. 511–514.
[25] Vaneechoutte M., Rossau R., De Vos P., Gillis M., Janssens D., PaepeN., DeRouck A., Fiers T., Claeys G., Kersters K., Rapid identifica-tion of bacteria of the Comamonadaceae with amplified ribosomalDNA-restriction analysis (ARDRA), FEMS Microbiol. Lett. 93(1992) 227–234.
[26] Welsh J., McClelland M., Fingerprinting genome using PCR witharbitrary primers,Nucleic Acids Res.18 (1990) 7213–7218.
[27] Williams J.G.K., Kubelick A.R., Livak K.J., Rafalski J.A., Tingey S.V.,Polymorphism generated by arbitrarily primed PCR in the mouse:application to strain identification and genetic mapping, NucleicAcids Res. 18 (1990) 6531–6535.
[28] Wimpee C.F., Nadeau T.L., Nealson K.H., Development of species-specific hybridization probes for marine luminous bacteria by usingin vitro DNA amplification, Appl. Environ. Microbiol. 57 (1991)1319–1324.
