60, 5-g


Genomic Variations in Mosquitocidal Strains of Bacillus sphericus Detected by Ml3 DNA Fingerprinting AGNES ABADJIEVA, Department

of Genetics,



of Microbiology,







4, 1991;


of Sciences,


1113, &&aria

June 14, 1991

Ml3 DNA are the result of the sequences which exhibit length variation due to variable number of tandem reand local strains belonging to different serotypes was exam- peats, homologous to both &base pair clusters ined by DNA fingerprinting. A phage Ml3 DNA probe de- (GAGGGTGGXGGXTCT) within the protein III gene tected a number of variable fragments in the restriction di- of the Ml3 molecule, first used as a probe by Vassart et gestsof total strain DNAs. The patterns of band distribution al. (1987). The differentiation of the toxic strains of this showed a certain homology among mosquitocidal strains, species on the basis of morphological and biochemical expressedby similarity index D and might be a reliable cricharacteristics is almost impossible (Baumann et al., terion for assessingthe level of genomic similarity between 1984; de Barjac et al., 1980). The serological classificaclosely related strains. An important advantage of DNA fingerprinting is the differentiation of one bacterial strain from tion led to grouping of the pathogenic strains into five (de Barjac et al., 1985). Genomic differanother, both expressing common phenotype and possessing H-serogroups ences or similarities among serotypes have not been highly similar genomic portions. The strain variation reexplored. vealed by the Ml3 probe will be useful for characterization Therefore, the use of a modern molecular analysis of individual strains within a serotype. It could help as well for investigation of differto solve someuncertain casesbased on the results obtained such as DNA fingerprinting ences among strains and serotypes of B. sphaericus at by other methods of identification. o 1~ Academic PMS, the DNA level seemed appropriate. To evaluate KEY WORDS: Bacillus sphaericus; fingerprinting; Ml3 whether this technique is useful to give information probe; serotypes; mosquitocidal strains. about genetic relatedness at the strain level B. sphaericus serotypes and strains within a serotype were anaINTRODUCTION lyzed. Minisatellites comprise arrays of unique nucleotide MATERIALSANDMETHODS units which are tandemly repeated. They exist as dispersed families in the genomes of humans, animals, Bacterial strains. The B. sphaericus strains studied plants, and microorganisms. A subset of the repeated are listed in Table 1. Local strains were isolated from DNA sequences contains a short “core” sequence. Jef- vectors (Culex, Culiseta and Anopheles larvae) and freys and his colleagues have developed a DNA fingermosquito breeding sites. All strains were grown in nuprinting system based on the hypervariability of these trient broth (Difco) at 30°C. regions due to extensive variation in the number of Preparation of DNA and restriction endonucleuse repeats (Jeffreys et al., 1985a). Cloned DNA segments cleavage. Procedures used in preparing DNA fingercontaining minisatellites have been used as hybridizaprints have been described previously (Abadjieva et al., tion probes which detect alleles from multiple loci un- 1990; Miteva et al., 1990). Total DNA was isolated acder the conditions of relaxed stringency. The resulting cording to Hintermann et al. (1981). DNA (10 Fg> from complex patterns or “DNA fingerprints” are individueach sample was cleaved by restriction endonucleases ally specific. and electrophoresed in 25cm-long 1% agarose gels at 2 Recently we have reported the presence of repeated V/cm for 18 hr. XbaI, BgZI, EcoRI, PvuI, PvuII, BcZI, sequences in a group of Gram-positive microorganisms HindII, HindIII, PstI, KpnI, H&I, Hue111 single and detected by the use of Ml3 bacteriophage DNA as a double digestions were performed. Gels were Southern probe (Miteva et al., 1990). Bacillus sphaericus, a miblotted to nitrocellulose membranes Hybond C (Amercroorganism of great antimosquito potency was also sham). studied (Abadjieva et al., 1990). The hybridization proHybridization. M13mp8 single-stranded DNA was files in the genome of this microorganism generated by 5 The genomic variation of Bacillus sphaericus reference




Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.







Bacterial Strains Studied Strain

Pathogenic activity


Phage group*

Kellen Q




SSII - 1 WHO 1593 WHO 2362 WHO 2297 CCEB 640

+ + + + -


5a5b 5a5b 25 4

2 3 3 4


Local Local Local Local Local Local Local Local

+ + + + + + + +

5a5b’ 5a5b’ 5a5b” 5a5b” ND ND ND ND


03 066 068 089 41 98 106 116

Origin Obtained from de Barjac (Inst. Pasteur, IP) IP l-l IP 2-l IP 5-1 IP 5-10 IP 25-1 Obtained from 0. Lysenco This laboratory This laboratory This laboratory This laboratory This laboratory This laboratory This laboratory This laboratory

Note. ND, not determined. a From de Bajac et al. (1985). * From Yousten (1984). ’ From de Barjac (personal communication).

labeled with [a-32PldCTP by Klenow polymerase extension from the Ml3 sequencing primer to specific activity of 1.0 x 10’ cpm/pg using a commercial kit (Amersham). Blots were prehybridized in 5 x SSC, 5 x Denhardt’s solution, 0.1% sodium dodecyl sulfate, 5 mM EDTA at 57 and 54°C for 1 or 2 hr and hybridized overnight under the same conditions. Filters were washed in 2~ SSC, 0.1% sodium dodecyl sulfate at the hybridization temperatures and exposed at - 70°C with an intensifying screen. Several autoradiographs were used for evaluation of the fragment patterns. Fragments from phage X DNA cut with Hind111 were used as size markers. RESULTS


Comparison of DNA Fingerprints Different Serotypes


DNA Ml3 probe was hybridized to blots of doubledigested genomic DNA from reference strains B. sphaericus, belonging to five serotypes: 2363 (serotype 5), 2297 (serotype 251, Kellen Q (serotype 11, SSII-1 (serotype 2), and CCEB 640 (serotype 4, Fig. 1). The number and distribution of the observed hybridization bands revealed substantial differences in the DNA fingerprinting patterns and generated a unique spectrum for any of these samples. At the same time several fragments similar in size were detected in all of the serotypes of mosquitocidal B. sphaericus studied (5, 25, 1, 2). Fingerprint bandsharing level between two individuals was proposed to evaluate the degree of similarity between their ge-


.i 123451


FIG. 1. DNA fingerprinting patterns from strains B. sphuericus belonging to different serotypes. Panel (a) XbaI-EcoRI digests; (b) PuuI-EcoRI digests. Strains: lane 1, 2362 (serotype 5a5b); lane 2, 2297 (serotype 25); lane 3, Kellen Q (serotype la); lane 4, SSZZ-1 (serotype 2); lane 5, 640 (serotype 4).

nomes. The position and number of bands were used to calculate the probability that a band present in one serotype is matched by one of approximately similar mobility and intensity in another, expressed by index D according to Jeffreys et al. (198513.) Differentiation between serotypes was conducted using four different double restrictions. Table 2 shows the results from all pairwise comparisons. Similarity index D was calculated as D = 2 no.AB/ (no.A + no.B) where no.AB was the number of bands shared by two serotypes A and B, possessing no.A and no.B bands each (Lansmann et al., 1981; Jeffreys et al., 1985b). The probability that A and B share all fragments (P) was calculated as the average DAB was raised to the power of the average number of fragments per genome (Nybom et al., 1990). The considerable DNA similarity among all mosquitocidal serotypes could be explained by common genome segments, containing minisatellites, which are conserved in the evolution of antimosquito species. In the analyses of serotype variation only one apathogenic strain (640) was used. It was completely different from all the others and the individual similarity index values, resulting from its comparison with the mosquito pathogens were approximately five times lower (Table 2). This result confirmed that the B. sphaericus group encompassed a large number of strains, pathogenic or not for mosquitoes, and often only slightly related.


OF B. sphericus

TABLE2 Analysis Comparison


of Ml3

probe hybridization



of mosquitocidal strains from different serotypes (2362, 2297, Kellen Q and SSZZ-1) No. fragments per genome

XbaI-EcoRI PuuI-EcoRI PstI-PUUII KpnI-Hue111

11.5 11.75 8.5 8.5

Comparison of mosquitocidal serotypes and one apathogenic serotype (strain 640)


(1.3) (0.5) (1.2) (1.3)

No. fragments per genome


0.59 (0.03) 0.76 (0.01) 0.51(0.07) 0.60 (0.07)

2.3 3.9 33 1.3

from B. sphuericus


x x x x

11.4 11.8 7.8 8.8

10m3 10-2 10-3 10-2

(1.1) (0.44) (1.9) (1.3)

All results in the table are obtained under relaxed stringency (54°C); D is calculated as an average no.ABl(n0.A + no.B) where no.A and no.B are the number of fragments in strains A and B, and no.AB only bands that have similar intensity and roughly equal mobility in two strains were taken as shared DA, raised to the power of the average number of fragments per genome; standard deviation within

The significantly similar clustering of minisatellites in the toxic strains demonstrates a good correlation with the data of Krych et al. (1980). On the basis of

within a Serotype

We wanted to check whether any differences between mosquitocidal strains belonging to one and the same serotype could be detected. In our investigation



A c


0.19 (0.06) 0.16 (0.005) O.ll(O.07) 0.16 (0.06)

P 5 4 1.6 9.9

x x x x

10-g lo-lo lo-* 10-s

of all pairwise comparisons: DAB = 2 is the number shared by both strains; bands; P is calculated as the average parentheses.

two reference strains (1593 and 2362) were used. Our decision to choose these two was related to the previous studies. Both strains belonged to phage group 3 (Yousten, 1984) and 5a5b serogroups (de Barjac et al., 1985). In DNA-DNA hybridization experiments these strains exhibited fragments identical in size when hybridized to clones, containing different parts of the B. sphaericus toxin gene (Baumann et al., 1987). The sequence analyses of the coding and flanking regions of the corresponding toxin genes in both microorganisms were shown to be identical (Berry and Hindley, 1987; Hindley and Berry, 1987). So it, was interesting to compare DNA fingerprinting patterns of strains containing highly similar genomic portions. Our results showed that both strains revealed great similarity (Fig. 2A, lanes 1 and 2). The profiles differed only in one or two

DNA homologies these authors readily distinguished all strains of B. sphaericus toxic to mosquitoes (forming the IIA homology subgroup) from the other homology groups and subgroups. This parallelism demonstrates that the level of band-sharing between the DNA fingerprinting patterns of two related strains is a reliable criterion for assessing the genomic similarity between them. Comparison of DNA Fingerprints




B c


,23 e.4

(I,5 4.3 2,3 2p









FIG. 2. Intra- and interserotype comparison of DNA fingerprinting patterns from strains B. sphericus revealed at (A) high (57°C) and (B) relaxed (54°C) stringency. (a) PstI-HindIII digests; (b) &&Hind111 digests; (c) BglI-Hind111 digests; (d) EcoRI digests. Strains: lane 1, 2362 (serotype 5a5b); lane 2, 1593 (serotype 5a5b); lane 3, Kellen Q (serotype la).





bands, but still enough to distinguish between them. a b C As the number of comparable fragments was small we tried to obtain more information by hybridizing washed filters with radiolabeled Ml3 under less stringent conditions (54°C instead of 57°C). In such experiments (Fig. 2B) additional fainter bands were revealed and the constant ones were also observed. The strains could again be distinguished on the basis of more extensive variation, so relaxed stringency could be more useful in the comparing of closely related strains. Their genomic DNA shared more M13-related bands and showed fewer differences in a pairwise comparison. That is why it is recommendable for better differentiation to use cleavage with several restriction endonucleases. The comparison of the reference strains from 5a5b serogroup (1593 and 2362) revealed similarity index values higher than those demonstrated among serotypes (Table 3). Another strain (Kellen Q), belonging to serotype la and reported to share 79% homology with 1593 (Krych et al., 1980) was also examined. The data from their fingerprint analysis are shown in Table 3 and Fig. 2 1234 (lanes 2 and 3). The probe Ml3 picked up genomic dif123 12345 ferences demonstrated by segments of unique position. FIG. 3. Comparison of DNA fingerprinting patterns from local The calculated indexes of the two strains were close to isolates of B. sphoericus with reference strains. (al HindIII-PuuII those obtained in the investigation of different sero- digests. Strains: lane 1,2362; lane 2,068; lane 3, 106; lane 4,03. (b) types (Table 1, Fig. 1). The GC-rich core sequence of PstI-PuuII digests. Strains: lane 1, 2297; lane 2, 2362; lane 3, 089. Differences in local strains are indicated by the arrows on the autoMl3 contrasted with the low GC content of the DNA radiographs. (c) BclI-HindIII digests. Strains: lane 1, 2362; lane 2, from both strains-35 mol% (Krych et al., 1980). This 41; lane 3, 98; lane 4, 106; lane 5, 116. suggests that using an AT-rich probe could permit a wider range of designated probe sequences in DNA finduced stringency conditions. The fragment profiles gerprinting assays. That is why a more detailed com- were obtained by digestions with four- and six-base parison of these strains with a probe possessing high recognizing restriction enzymes. In two cases (strains AT content, for example, DXYS15 (Simmler et al., 03 and 089) divergences from the reference strain 2362 19871, would be of interest. were revealed (Figs. 3a and 3b). As both strains are Figure 3 represents the results of our hybridization highly toxic to Culex mosquito larvae (LCso 8.2 x lo2 study of local pathogenic isolates performed under reand 7.5 X lo2 spores/ml, respectively) and their fingerprint patterns differed in segregation of new-length bands, they could become new representatives of the TABLE 3 collection of mosquitocidal strains of the species. Intra- and Interserotype Similarities of B. sphaericus Another group of four local strains of B. sphaericus Probability Probability relatively toxic to Aedes larvae showed BcZI-Hind111 No. that a that A hybridization patterns identical to the reference strain No. fragments fragment and B 2362 (Fig. 3c), although they came from different geoStrains and informative in A is share all per graphical areas. enzymes fragments genome also in B fragments 236211593 P&I-HirKlIII BclI-HindIII BgZI-Hind111 EcoRI 1593lKellen Q P&I-Hind111 BcZI-HindI BglI-Hind111 EcoRI

24 25 27 11

12 12.5 13.5 5.5

0.91 0.96 0.96 0.72

3.2 6.0 6.6 1.6

x x x x

10-i 10-l 10-l 10-i

22 19 27 9

11 8.5 13.5 4.5

0.63 0.63 0.66 0.44

6.2 1.9 3.6 2.4

x x x x

1O-3 10-2 1O-3 10m2

Note. All results in the table are obtained under relaxed stringency (54°C).


of the Method

Both reproducibility and genetic stability of serotypes were studied by preparing fingerprints from strains after multiple passages on Difco agar. Patterns identical to the original HindIII digestions were obtained with all serotypes studied. Confirmation of these results was obtained by restriction enzyme cleavage with EcoRI (data not shown). The results obtained in this study prove that the method of DNA fingerprinting reflects the high infor-


mative value of the hypervariable loci in the genomes of related strains of one species. The investigation of reference and local strains of B. sphaericus showed the applicability of this technique to display genetic similarities and differences within the B. sphaericus species. An important advantage of DNA fingerprinting is the differentiation of one bacterial strain from another, both expressing common phenotype. As the antimosquito potency of B. sphaericus is far from being well known and new strains are still being discovered this method promises to become an efficient tool in strain identification. It could be used as well as an additional character to the numerical classification of bacteria especially in doubtful cases. Genetic dissection of the mechanisms involved in the production of antigens will be considerably more difficult than studies of proteins, which can be related to few genes. Nevertheless it seems worthwhile to use the molecular genetic techniques to explore serotypic differences among strains of B. sphaericus. Recent studies show that a tandem repeat probe can detect another probe, which hybridizes with a new discriminative set of DNA fragments (Simmler et al., 1987; Washio et al., 1989). The effective strategy for the future experiments could be the using of conserved regions of similarity for the development of more specific probes for analysis of B. sphaericus strains. ACKNOWLEDGMENTS This investigation received financial support from the UNDPI WORLD BANK/WI-IO Special Programme Research and Training in Tropical Diseases (TDR). REFERENCES Abadjieva, A. N., Grigorova, R. T., and Miteva, V. I. 1990. DNA fingerprinting of the mosquito pathogen Bacillus sphaericus with Ml3 DNA as a probe. Lett. Appl. Bact. 10, 141-143. Berry, C., and Hindley, J. 1987. Bacillus sphaericus strain 2362: identification and nucleotide sequence of the 41,9 kDa toxin gene. Nucleic Acids Res. 15, 5891. Baumann, L., Okamoto, K., Unterman, B. M., Lynch, M. J., and Baumann, P. 1984. Phenotipic characterization of Bacillus thuringiensis and Bacillus cereus. J. Invertebr. Pathol. 44, 329-341. Baumann, P., Baumann, L., Bowditch, R. D., and Broadwell, A. H.


OF B. sphaericus

1987. Cloning of the gene for the larvicidal toxin of Bacillus 2362: Evidence for the family of related sequences. J. 169, 4061-4067. de Barjac, H., Veron, M., and Cosmao Dumanoir, V. 1980. Characterisation biochimique et serologique de souches de Bacillus sphaericus pathogenes ou non pour les moustiques. Ann. Microbiol. (In& Pasteur). 131B, 191-201. de Barjac, H., Larget-Thiery, I., Cosmao Dumanoir, V., and Ripouteau, H. 1985. Serological classification of Bacillus sphaericus strains on the basis of toxicity to mosquito larvae. Appl. Microbial. Biotech. 21, 85-90. Hindley, J., and Berry, C. 1987. Identification, cloning and sequence analysis of the Bacillus sphaericus 1593 41,9 kDa larvicidal toxin gene. Mol. Microbial. 1, 187-194. Hintermann, G., Grameri, R., Keiser, T., and Hutter, R. 1981. Restriction analysis of the Streptomyces glaucescens genome by agarose gel electrophoresis. Arch. Microbial. 130, 218-222. Jeffreys, A. J., Wilson, V., and Thein, S. L. 1985a. Hypervariable “minisatellite” regions in human DNA. Nature (London) 314, 6773. Jeffreys, A. J., Wilson, V., and Thein, S. L. 1985b. Individual specific “fingerprints” of human DNA. Nature (London) 316, 7679. Krych, V. K., Johnson, J. J., and You&en, A. A. 1980. Deoxyribonucleic acid homologies among strains of Bacillus sphaericus. Znt. J. sphaericus Bacterial.

Syst. Bacterial.

30, 476484,

Lansmann, R., Shade, R., Shapira, J., and Avise, J. 1981. The use of restriction endonucleases to measure mitochondrial DNA sequence relatedness in natural populations. J. Mol. Evol. 17, 214226. Miteva, V. I., Abadjieva, A. N., and Grigorova, R. T. 1990. Ml3 bacteriophage DNA as a probe for DNA fingerprinting in Grampositive microorganisms. Syst. Appl. Microbial. 13, 350-353. Nybom, H., Rogstad, S. H., and Schaal, B. A. 1990. Genetic variation detected by use of Ml3 DNA fingerprint probe in Malus, Prunus and Rubus (Rosaceae). Theor. Appl. Genet. 79, 153-156. Simmler, M. C., Johnsson, C., Petit, C., Rouyer, F., Vergnaud, G., and Weissenbach, J. 1987. Two highly polymorphic minisatellites from the pseudoautohosomal region of the human sex chromosomes. EMBO J. 6, 963-969. Vassar-t, G., Georges, M., Monsier, R., Brocas, H., Lequarre, A. A., and Christophe, D. 1987. A sequence in Ml3 phage detects hypervariable minisatellites in human and animal DNA. Science 235, 683-684. Washio, K., Misava, S., and Ueda, S. 1989. Probe walking: Development of novel probes for DNA fingerprinting. Hum. Genet. 83, 223-226. Yousten, A. 1984. Bacteriophage typing of the mosquito pathogenic strains of Bacillus sphaericus. J. Invertebr. Pathol. 43, 124-125.

Genomic variations in mosquitocidal strains of Bacillus sphaericus detected by M13 DNA fingerprinting.

The genomic variation of Bacillus sphaericus reference and local strains belonging to different serotypes was examined by DNA fingerprinting. A phage ...
2MB Sizes 0 Downloads 0 Views