Comparative analysis of the LPS biosynthetic loci of the genetic subtypes of serovar Hardjo: Leptospira interrogans subtype Hardjoprajitno and Leptospira borgpetersenii subtype Hardjobovis

FEMS Microbiology Letters 177 (1999) 319^326

Comparative analysis of the LPS biosynthetic loci of the genetic subtypes of serovar Hardjo: Leptospira interrogans subtype Hardjoprajitno and Leptospira borgpetersenii subtype Hardjobovis Alejandro de la Pen¬a-Moctezuma, Dieter M. Bulach, Thareerat Kalambaheti, Ben Adler * Bacterial Pathogenesis Research Group, Department of Microbiology, Monash University, Clayton, Vic. 3168, Australia Received 25 June 1999 ; accepted 28 June 1999

Abstract Although Leptospira borgpetersenii subtype Hardjobovis and L. interrogans subtype Hardjoprajitno belong to different species, they are serologically indistinguishable and are therefore classified as serovar Hardjo. Since LPS is the major antigen involved in serological classification, this implies that the LPS of these subtypes is identical. Comparison of the LPS biosynthetic loci (rfb) of the subtypes revealed remarkable similarity, with 32 and 31 origins of replication (orfs) in the Hardjoprajitno and Hardjobovis rfb loci, respectively. The order and orientation of these orfs were identical with the exception of an additional orf in Hardjoprajitno between orfs 4 and 5 and intergenic sequences differing between the subtypes. The Hardjoprajitno rfb locus has been divided into four intercalated regions based on sequence similarity to other leptospiral rfb loci. orfJ1^orfJ14 as well as orfJ21^orfJ22 are more similar to regions of the rfb locus of L. borgpetersenii subtype Hardjobovis. orfJ15^orfJ20 as well as orfJ23^orfJ31 are almost identical to the corresponding orfs in L. interrogans serovar Copenhageni. We propose that the progenitor Hardjoprajitno strain, containing an rfb locus which closely resembled the Copenhageni locus, acquired orfs 1^14 and orfs 21^22 from subtype Hardjobovis resulting in two serologically indistinguishable subtypes of serovar Hardjo which in turn constituted the main bovine-adapted leptospiral serovar. ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : rfb Locus; Phylogeny ; Hardjo; Lipopolysaccharide biosynthesis; Leptospira

1. Introduction Leptospirosis is a worldwide zoonosis caused by spirochaetes of the genus Leptospira [1,2]. DNA hybridisation analysis has led to the division of the pathogenic leptospires into species, namely Lepto-

* Corresponding author. Tel.: +61 (3) 9905 4815; Fax: +61 (3) 9905 4811; E-mail: [email protected]

spira borgpetersenii, Leptospira inadai, Leptospira interrogans, Leptospira kirschneri, Leptospira noguchi, Leptospira santarosai and Leptospira weilii [3,4]. Serologically, leptospires are classi¢ed into more than 200 serovars and 24 serogroups based on antigenic relatedness [2]. The serovar-based classi¢cation of Leptospira depends on the diversity of structure of the surface-exposed components of lipopolysaccharide (LPS). Notably, LPS is the only protective antigen identi¢ed thus far [5]. Furthermore, leptospiral

0378-1097 / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 9 9 ) 0 0 3 3 3 - X

FEMSLE 8923 29-7-99

320

A. de la Pen¬a-Moctezuma et al. / FEMS Microbiology Letters 177 (1999) 319^326

LPS elicits the production of agglutinating, opsonic, protective antibodies in mice and hamsters [6,7]. Mitchison et al. [8] described the organisation and con¢rmed the functionality of the rhamnose biosynthetic region, part of the LPS biosynthetic (rfb) locus, in L. interrogans serovar Copenhageni. More recently, Kalambaheti et al. [9] mapped and sequenced the rfb locus of L. borgpetersenii serovar Hardjo subtype Hardjobovis, identifying a cluster of 31 open reading frames (orfs) all transcribed in the same direction. Serovar Hardjo is the main cause of leptospirosis in Australia and New Zealand [10,11] as well as in other parts of the world [1,2]. The genetic heterogeneity of strains antigenically grouped as serovar Hardjo was ¢rst demonstrated by restriction endonuclease analysis [12], leading to delineation of two subtypes, Hardjoprajitno and Hardjobovis [13]. The prevalent infecting subtype in cattle and humans is Hardjobovis [12,14]. By contrast, a study of the isolates obtained from bovine clinical and pathological material showed that more than 75% of Hardjo isolates were subtype Hardjoprajitno [15]. Intriguingly, subtypes Hardjobovis and Hardjoprajitno are classi¢ed into di¡erent species, L. borgpetersenii and L. interrogans, respectively, based on whole genomic DNA hybridisation [13]. We have proposed that insight into the origins of these subtypes will be provided by comparison of the Hardjo rfb loci. Accordingly, in this paper, we report a comparison of the rfb loci of subtypes Hardjobovis and Hardjoprajitno, which reveals a remarkable conservation of gene content and order. An additional comparison with the rfb locus of serovar Copenhageni has enabled the proposal of a hypothesis relating to the origin of the Hardjo subtypes.

2. Materials and methods 2.1. Bacterial strains and growth media L. interrogans serovar Copenhageni (strain L45) was described previously [16] while L. borgpetersenii serovar Hardjo subtype Hardjobovis (strain L171) was obtained from Dr. R.B. Marshall, Massey University, New Zealand. L. interrogans serovar Hardjo subtype Hardjoprajitno (strain L375) was obtained

from the Royal Tropical Institute, Amsterdam, The Netherlands. Strains were cultured in EMJH medium with added pyruvate [17]. 2.2. PCR and sequencing Chromosomal DNA was prepared by the CTAB lysis method [18]. Paired primers, based on the nucleotide sequences of the Hardjobovis or Copenhageni rfb loci, were used for PCR to generate sequencing templates from Hardjoprajitno chromosomal DNA. PCR was performed by using the Expand High Fidelity system (Roche Molecular Biochemicals) with an annealing temperature of 56³C. The paired primers used to generate a set of overlapping amplicons by PCR, which span the entire Hardjoprajitno rfb locus, are listed in Table 1. Sequencing of the amplicons was performed by primer walking, using the BigDye Ready Reaction Dye Deoxy terminator cycle sequencing kit in a Perkin-Elmer Applied Biosystems 373A automated sequencer (www.perkinelmer.com). Sequences were assembled with Sequencher 3.0 (www.genecodes.com). Analysis and comparison of sequences was performed using the Australian National Genomic Information Service (www.angis.org.au), through which the GCG suite of programs is available [19].

3. Results 3.1. Common gene layout Earlier comparisons of the rfb locus of serovar Copenhageni and subtype Hardjobovis revealed a common gene layout in the region spanning orf24^ orf29, nucleotide sequence similarity is approximately 90% across this region of the loci [9]. An additional sequence determined from the Copenhageni rfb locus reveals that the common gene order extends from orf11^orf31 (Bulach et al., manuscript in preparation). On the basis of the common gene layout in the loci from Hardjobovis (L. borgpetersenii) and Copenhageni (L. interrogans), it was assumed that a common gene layout might be found in Hardjoprajitno. Using combinations of primers, previously designed for the determination of the nucleotide sequence of Hardjobovis and Copenhageni

FEMSLE 8923 29-7-99

A. de la Pen¬a-Moctezuma et al. / FEMS Microbiology Letters 177 (1999) 319^326

321

Table 1 Overlapping fragments obtained by PCR used as templates for DNA sequencing Fragment

Amplifying primers

Primer sequences

A

BAP329 BAP316 BAP663 BAP700 BAP278 BAP859 BAP228 BAP867 BAP268 BAP86 BAP775 BAP774 BAP622 BAP677 BAP534 BAP535

5P-CAG 5P-CAG 5P-GTA 5P-GAA 5P-GAT 5P-AAC 5P-CTA 5P-ACG 5P-ATG 5P-TGT 5P-CAR 5P-GCN 5P-TTT 5P-AGC 5P-GGA 5P-AGT

B C D E F G H

GGG ACT TAC TAC ACG GTC ATA TGG CAG TCC ATG CCR TAT AAA ATT GTA

CTG TGA GAG GGA TAT CAT CTT ATC TGT GCA ATG TGN CAG TCC CGT TAA

CAA ATA AAC GAC GTC TCC GCC GTT AAC TAA GGN SNY GCT AGG TGG ATC

TTG ATG CCT TGT GAC TTC TTG TTC CAA TCG GAY TGN GGC AGG GAT GCT

rfb loci, it was possible to amplify the fragments listed in Table 1 from the Hardjoprajitno rfb locus by PCR. The nucleotide sequence of the Hardjoprajitno rfb locus spanning orfJ1^orfJ31 comprises 37 918 bp (GenBank: AF144879). The sequence has a G+C content of 33.8% as compared to 36.2% found in the Hardjobovis rfb locus, 31.3% in the Copenhageni rfb locus and 33.5^40.5% found in Leptospira spp. genomes [2,4]. 3.2. Comparison with the Hardjobovis locus Analysis of the Hardjoprajitno rfb locus sequence allowed for the identi¢cation of 32 orfs encoded on the same strand. Homologues of the 31 orfs previously described in Hardjobovis [9] were observed in the Hardjoprajitno rfb locus. Comparison of the encoded proteins revealed that homologues ranged from 63 to 99% identity and that the orfs were arranged in the same order as in the Hardjobovis rfb locus (Table 2). orfJ36 was an additional reading frame in Hardjoprajitno, spanning 615 bp and is located between orfJ4 and orfJ5. The deduced protein sequence has no similarity to any protein in the database. Comparison of the nucleotide sequences among the rfb loci of subtypes Hardjobovis and Hardjoprajitno revealed that identity in the coding regions was

ACG-3P GGC-3P TGG AAT CCC-3P ACG-3P CGG-3P CGG-3P CCC-3P GAG-3P TTG-3P GAR GCN GCR CAR TGG-3P AAC-3P GAT CCT CCG AGA

GGC-3P

GTY GG-3P TCY TC-3P

GCG-3P TGC-3P

Size of fragment (bp)

Region of rfb locus

3 978

Upstream orf1^orf2

4 555

orf2^orf6

5 133

orf6^orf11

3 089

orf11^orf13

6 325

orf12^orf15

5 877

orf15^orf20

5 976

orf20^orf24

9 246

orf24^orf31

above 80% except for orfs J4, J7, J17 and J23 (Fig. 1 and Table 2). Intergenic regions were signi¢cantly less-conserved. In the intergenic regions between orfs 4^5, orfs 12^13, orfs 18^19, orfs 21^22, orfs 22^23, orfs 27^28 and orfs 29^30. Additional sequences were found in either Hardjobovis or Hardjoprajitno (Fig. 1). In contrast, substantial intergenic regions exist between orfs 14^15, orfs 16^17 and orfs 17^18 in both subtypes. The sequences in these regions di¡er signi¢cantly. Kalambaheti et al. [9] detected the presence of a member of the IS5 family of IS elements in the orfs H14^H15 intergenic region of the Hardjobovis rfb locus, while in Hardjoprajitno, there is an intergenic region of a similar size but no insertion element. 3.3. Comparison with the Copenhageni locus In contrast to the comparison of the HardjobovisHardjoprajitno rfb loci, where there was a 86% average nucleotide sequence identity across all 31 conserved reading frames, the Copenhageni-Hardjoprajitno comparison identi¢ed two regions with almost 100% nucleotide sequence identity (98.6%). These regions span from orfJ15 to orfJ20 (region B) and orfJ23^orfJ31 (region D). There is an additional sequence between orfs J29^J30 in Hardjoprajitno that includes 14 copies of a 46-bp direct repeat (Fig. 1). Comparison of the Copenhageni and Hardjopra-

FEMSLE 8923 29-7-99

322

A. de la Pen¬a-Moctezuma et al. / FEMS Microbiology Letters 177 (1999) 319^326

jitno loci in the region spanning from orfJ1 to orfJ14 (region A, Fig. 1) identi¢ed a signi¢cant divergence in organisation. No homologues to orfJ1^orfJ9 are present in the Copenhageni locus and a reading frame between orfC10 and orfC11, identi¢ed as orfC32, is not present in either Hardjobovis or Hardjoprajitno. orfJ21^orfJ22 has been designated region C (Fig. 1). Identity between the nucleotide sequence of Copenhageni and Hardjoprajitno in this region is less than for the corresponding comparison between

Hardjobovis and Hardjoprajitno. Comparison of the encoded proteins from Hardjoprajitno with Hardjobovis revealed that these are much more closely related (OrfH21 97% and OrfH22 96% similarity) than Copenhageni (OrfC21 58% and OrfC22 84% similarity). 3.4. Comparative overview The genetic organisation and nucleotide sequence of the locus in subtype Hardjoprajitno showed that it

Table 2 Relationship of rfb genes and Rfb proteins of Hardjoprajitno with homologues in Hardjobovis and Copenhageni Hardjoprajitno orfs

J1 J2 J3 J4 J36 J5 J6 J7 J8 J9 J10 J11 J12 J13 J14 J15 J16 J17 J18 J19 J20 J21 J22 J23 J24 J25 J26 J27 J28 J29 J30 J31

Hardjobovisa

Copenhagenia

Putative functionsb

Nucleotide

Peptide

Nucleotide

Peptide

90 92 80 72 NP 88 85 77 91 94 96 95 88 91 93 83 86 76 80 86 82 91 92 75 85 85 87 91 86 86 85 80

91 (96) 97 (99) 83 (93) 78 (89) NP 91 (93) 93 (97) 87 (93) 92 (96) 95 (98) 99 (100) 99 (99) 86 (92) 92 (96) 91 (95) 90 (97) 83 (90) 72 (85) 80 (90) 88 (98) 87 (95) 94 (97) 94 (96) 63 (78) 90 (94) 86 (93) 90 (94) 97 (98) 86 (93) 89 (93) 85 (93) 85 (93)

NPc NP NP NP NP NP NP NP NP NP 67 64 67 69 ^ 99 98 99 100 99 99 59 70 95 98 99 99 99 99 99 99 99

NPc NP NP NP NP NP NP NP NP NP 65 (82) 58 (77) 65 (77) 72 (85) 49 (21) 99 (100) 99 (99) 99 (99) 100 (100) 99 (100) 100 (100) 35 (58) 68 (84) 95 (97) 100 (100) 99(100) 98 (98) 100 (100) 99 (100) 99 (99) 99 (99) 99 (99)

Glycosyltransferase Unknown Aminotransferase (histidine biosynthesis) Cyclase (histidine biosynthesis) Unknown Glycosyltransferase Dehydratase Mannose biosynthesis UDP-Glu-4-epimerase Oxidoreductase UDP-Gal-NAc-epimerase UDP-Glc-NAc-epimerase Fucosamine transferase Und-hexose-P-transferase O-Antigen polymerase wzy Flippase, wzx Unknown Unknown Dideoxy-sugar biosyntesis Isomerase Aminotransferase Glycosyltransferase Glycosyltransferase Glycosyltransferase Rhamnose biosynthesis Rhamnose biosynthesis Rhamnose biosynthesis Rhamnose biosynthesis Rhamnosyltransferase Rhamnosyltransferase Rhamnosyltransferase Transporter protein

ö = No detectable similarity. The shaded blocks indicate the regions of greater similarity to L. borgpetersenii (Hardjobovis) or L. interrogans (Copenhageni). a Percentage nucleotide identity or peptide identity (similarity) across the entire gene. b [9]. c NP = not present in this serovar.

FEMSLE 8923 29-7-99

Fig. 1. Graphical comparison of the organisation and nucleotide sequence of the rfb locus of (a) Hardjoprajitno and Hardjobovis and (b) Hardjoprajitno and Copenhageni. For each comparison, a pairwise alignment of sequences was constructed using GAP and PLOTSIMILARITY (window size 200 bases) [19] was used to draw the graphical relationship. Where gaps of greater than 100 bases have been added to optimise the alignment of nucleotide sequences, shaded boxes have been added to indicate the location of the gap. Flanking the line diagram of the Copenhageni and Hardjobovis rfb loci are sequence similarity and identity values comparing the protein sequence encoded by each orf in the Copenhageni and Hardjobovis rfb loci with the corresponding encoded protein in the Hardjoprajitno rfb locus. These values were determined using BESTFIT [19]. Region A (orfs 1^14), region B (orfs 15^20), region C (orfs 21^22) and region D (orfs 23^31) are indicated by hatched lines.

A. de la Pen¬a-Moctezuma et al. / FEMS Microbiology Letters 177 (1999) 319^326

FEMSLE 8923 29-7-99

323

324

A. de la Pen¬a-Moctezuma et al. / FEMS Microbiology Letters 177 (1999) 319^326

consists of four discrete regions in which region A (orfs J1^J14) and region C (orfs J21^J22) are more closely related to subtype Hardjobovis with an average similarity of 95.5%, excluding intergenic sequences. Region B (orfs J15^J20) and region D (orfs J23^ J31) resemble closely sequences from Copenhageni with an average similarity of 99.3% (Fig. 1).

4. Discussion 4.1. Evolution of the Hardjoprajitno subtype Analysis of the sequences of the rfb loci of the two subtypes of serovar Hardjo revealed a high level of peptide sequence similarity (94%) and the conserved order of 31 orfs, consistent with the encoded proteins producing a surface-expressed polysaccharide that is indistinguishable by serological methods. Additional comparison of the rfb loci of subtype Hardjoprajitno and serovar Copenhageni, both L. interrogans, revealed a very high (98.6%) nucleotide identity in regions B and D (Fig. 1 and Table 2). We proposed that the rfb locus of the Hardjoprajitno progenitor strain closely resembles that of serovar Copenhageni at least across the B and D regions. This progenitor strain, as a result of host or environmental adaptation, has subsequently acquired the Hardjobovis-speci¢c cluster of genes identi¢ed as region A (orfs J1^ J14) and substituted existing reading frames in region C (orfs J21 and J22), resulting in a `Hardjo phenotype' (Fig. 1). The mechanisms by which these genetic changes occurred are unknown, but it is reasonable to assume that they were acquired by lateral transfer from Hardjobovis, resulting in the present two serologically indistinguishable subtypes of serovar Hardjo. In order to examine in Hardjoprajitno the relative rate of nucleotide sequence change in the Copenhageni-like (regions B and D) or Hardjobovis-like regions (A and C), it must be assumed that the sequences from which they were derived were almost identical to the sequences in Copenhageni and Hardjobovis loci, respectively. On this basis, we observed that nucleotide sequences of the Copenhageni-like regions are more highly conserved (98.5%) than the Hardjobovis-like regions (88.4%). A general overview of the nucleotide sequence di¡erences in the

Hardjobovis-like region reveals that changes are restricted to the third base of the codon in a coding region (data not shown). This is consistent with the adaptation of nucleotide sequences acquired laterally to the codon usage in a new host and is additional support for the proposed derivation of the Hardjoprajitno genetic subtype. On the other hand, this is in contrast to the Copenhageni-like regions where it is proposed that no such adaptive force is present. Convergent evolution resulting in the `Hardjo phenotype' is likely to be an advantage for Hardjoprajitno due to host-serovar associations. Most human cases of Hardjo infection appear to be caused by Hardjobovis [10,11], which is also the most commonly isolated subtype from the urinary tract of asymptomatic cattle [2]. On the other hand, at least one report has suggested that reproductive tract isolates are more frequently the Hardjoprajitno subtype [15] and more likely to cause severe leptospirosis than the Hardjobovis subtype. Indeed, one study has found no association of serovar Hardjo with bovine abortion in a geographical situation where Hardjobovis was the predominant endemic subtype [20]. These ¢ndings are consistent with the hypothesis we have proposed based on sequence comparisons, that Hardjobovis is a bovine-adapted serovar of lower pathogenicity whereas Hardjoprajitno is a serovar of higher pathogenicity for cattle that has undergone host adaptation via gene acquisition from Hardjobovis. 4.2. Common functions between serovars Logically, the orfs in Hardjoprajitno that remain unchanged from the proposed progenitor strains, i.e. regions B and D (Fig. 1), must encode proteins with the same function in the context of LPS biosynthesis between Hardjoprajitno, Hardjobovis and Copenhageni. Thus, it is possible to predict that common sugars should be present in the LPS of such leptospiral serovars including for example rhamnose (orfs 24^27) [8] and the activated sugar(s) (orfs 17^20) [9]. In the same context, the functions of the putative glycosyltransferases encoded by orfs 23, 28, 29 and 30, as well as the putative transporter proteins, encoded by orf31 and orf15, are probably the same functions in each of Hardjoprajitno, Hardjobovis and Copenhageni.

FEMSLE 8923 29-7-99

A. de la Pen¬a-Moctezuma et al. / FEMS Microbiology Letters 177 (1999) 319^326

4.3. Divergent functions between serovars Conversely, where there has been a change from the predicted Hardjoprajitno progenitor strain, it is possible to predict a change in function. That is the case for the glycosyltransferases located in region C (Fig. 1), identi¢ed as Orfs 21 and 22, which are predicted to have a function that di¡ers between Copenhageni (58 and 84% similarity, respectively) and the two Hardjo subtypes (97 and 96% similarity, respectively), although the resultant di¡erences may merely re£ect changes in the linkage between the donor and acceptor sugars. The product of orfH14 is predicted to be the Oantigen polymerase (Wzy) [9]. In other Gram-negative bacteria, the function Wzy is dependent on the subunit structure [21] and as such, the function would necessarily vary between Hardjoprajitno and its progenitor. In addition, upstream of orfJ11 are several orfs that are not present in Copenhageni nor in any one of 10 L. interrogans serovars that have been examined thus far (manuscript in preparation). In this region of the Hardjoprajitno rfb locus (orfs J1^J9), the corresponding orfs in Hardjobovis are genes predicted to encode enzymes involved in the biosynthesis of activated sugars including galactose and N-acetyl-galactosamine as well as two glycosyltransferases [9], thus suggesting that these sugars may not be present in the Copenhageni LPS. Their highly conserved order in the serologically indistinguishable subtypes Hardjobovis and Hardjoprajitno and their absence in an unrelated serovar clearly suggest that this cluster of genes encodes antigenic/serologic speci¢city. The comparisons of rfb loci presented here have enabled the proposal of a theory that predicts that the Hardjoprajitno subtype rfb locus is derived from a strain closely related to serovar Copenhageni which acquired sections of the Hardjobovis rfb locus by horizontal genetic transfer. The fact that there are no signi¢cant di¡erences identi¢ed between the rfb loci of the Hardjo genetic subtypes is consistent with the comparison of loci that encode enzymes for the biosynthesis of LPS which are serologically indistinguishable. Similarity analysis, to identify putative functions encoded by the orfs found in the rfb loci, has allowed for predictions about the sugar content of the LPS of these serovars to be made.

325

These ¢ndings suggest that the genetic basis for serological di¡erences among leptospiral serovars is related to speci¢c orfs in their respective rfb loci. Detailed analysis of these rfb loci will therefore provide a molecular basis for the serological classi¢cation of Leptospira and for immunity in leptospirosis.

Acknowledgements This work was supported by a grant from the National Health and Medical Research Council, Canberra. A. de la Pen¬a-Moctezuma was the recipient of an Overseas Postgraduate Research Scholarship, a Monash Graduate Scholarship and a scholarship from the DGAPA, Universidad Nacional Auto¨noma de Me¨xico. The authors acknowledge the excellent technical assistance of Vicki Vallance and Ian McPherson.

References [1] Heath, S.E. and Johnson, R. (1994) Leptospirosis. J. Am. Vet. Med. Assoc. 205, 1518^1523. [2] Faine, S. (1994) Leptospira and Leptospirosis. CRC Press, Boca Raton, FL. [3] Ramadass, P., Jarvis, B.D., Corner, R.J., Penny, D. and Marshall, R.B. (1992) Genetic characterisation of pathogenic Leptospira species by DNA hybridisation. Int. J. Syst. Bacteriol. 42, 215^219. [4] Yasuda, P.H., Steigerwalt, A.G., Sulzer, K.R., Kaufmann, A.F., Rogers, F. and Brenner, D.J. (1987) Deoxyribonucleic acid relatedness between serogroups and serovars in the family Leptospiraceae with proposals for seven new Leptospira species. Int. J. Syst. Bacteriol. 37, 407^415. [5] Cinco, M., Delneri, D. and Ban¢, E. (1992) Immunodominant antigens recognised by the human immune response to infection by organisms of the species Leptospira interrogans serogroup Australis. FEMS Immunol. Med. Microbiol. 4, 287^297. [6] Jost, B.H., Adler, B., Vinh, T. and Faine, S. (1986) A monoclonal antibody reacting with a determinant on leptospiral lipopolysaccharide protects guinea pigs against leptospirosis. J. Med. Microbiol. 22, 269^275. [7] Farrelly, H.E., Adler, B. and Faine, S. (1987) Opsonic monoclonal antibodies against lipopolysaccharide antigens of Leptospira interrogans serovar hardjo. J. Med. Microbiol. 23, 1^7. [8] Mitchison, M., Bulach, D.M., Vinh, T., Rajakumar, K., Faine, S. and Adler, B. (1997) Identi¢cation and characterisation of the dTDP-rhamnose biosynthesis and transfer genes of the lipopolysaccharide-related rfb locus in Leptospira interrogans serovar Copenhageni. J. Bacteriol. 179, 1262^1267.

FEMSLE 8923 29-7-99

326

A. de la Pen¬a-Moctezuma et al. / FEMS Microbiology Letters 177 (1999) 319^326

[9] Kalambaheti, T., Bulach, D.M., Rajakumar, K. and Adler, B. (1999) Genetic organisation of the lipopolysaccharide O-antigen biosynthetic locus of Leptospira borgpetersenii serovar Hardjobovis. Microb. Pathog. (in press). [10] Adler, B. and Faine, S. (1980) The epidemiology of human leptospirosis in Australia. Commun. Dis. Intell. 80-21, 2^4. [11] Blackmore, D.K. (1982) Risks of contracting leptospirosis on the dairy farm. N.Z. Med. J. 95, 649^652. [12] Robinson, A.J., Rammadas, P., Lee, A. and Marshall, R.B. (1982) Di¡erentiation of subtypes within Leptospira interrogans serovars hardjo, balcanica and tarassovi by bacterial restriction-endonuclease DNA analysis (BRENDA). J. Med. Microbiol. 15, 331^338. [13] Ramadass, P. and Marshall, R.B. (1990) Species di¡erentiation of Leptospira interrogans serovar hardjo strain Hardjobovis from strain Hardjoprajitno by DNA slot blot hybridisation. Res. Vet. Sci. 49, 194^197. [14] Thiermann, A.B., Handsaker, A.L., Foley, J.W., White, F.H. and Kingscote, B.F. (1986) Reclassi¢cation of North American leptospiral isolates belonging to serogroups Mini and Sejroe by restriction endonuclease analysis. Am. J. Vet. Res. 47, 61^66. [15] Ellis, W.A., Thiermann, A.B., Montgomery, J., Handsaker,

[16]

[17]

[18]

[19] [20]

[21]

A., Winter, P.J. and Marshall, R.B. (1988) Restriction endonuclease analysis of Leptospira interrogans serovar hardjo isolates from cattle. Res. Vet. Sci. 44, 375^379. Faine, S. and van der Hoeden, J. (1964) Virulence-linked colonial and morphological variation in Leptospira. J. Bacteriol. 88, 1493^1496. Johnson, R.C., Walby, J.K., Henry, R.A. and Auran, N.E. (1973) Cultivation of parasitic leptospires : E¡ect of pyruvate. Appl. Microbiol. 26, 118^119. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.A., Smith, J.A. and Struhl, K. (1997) Current protocols in Molecular Biology. Willey Interscience and Sons, NY. (1994) Program Manual for the Wisconsin Package version 8. Madison, WI. Chappel, R.J., Millar, B.D., Adler, B., Hill, J., Je¡ers, M.J., Jones, R.T., McCaughan, C.J., Mead, L.J. and Skilbeck, N.W. (1989) Leptospira interrogans serovar hardjo is not a major cause of bovine abortion in Victoria. Aust. Vet. J. 66, 330^333. Burrows, L.L. and Lam, J.S. (1999) E¡ect of wzx (rfbX) mutations on A-band and B-band lipopolysaccharide biosynthesis in Pseudomonas aeruginosa O5. J. Bacteriol. 181, 973^ 980.

FEMSLE 8923 29-7-99