Serovar distribution of a DNA sequence involved in the antigenic relationship between Leptospira and equine cornea

Background Horses infected with Leptospira present several clinical disorders, one of them being recurrent uveitis. A common endpoint of equine recurrent uveitis is blindness. Serovar pomona has often been incriminated, although others have also been reported. An antigenic relationship between this bacterium and equine cornea has been described in previous studies. A leptospiral DNA fragment that encodes cross-reacting epitopes was previously cloned and expressed in Escherichia coli. Results A region of that DNA fragment was subcloned and sequenced. Samples of leptospiral DNA from several sources were analysed by PCR with two primer pairs designed to amplify that region. Reference strains from serovars canicola, icterohaemorrhagiae, pomona, pyrogenes, wolffi, bataviae, sentot, hebdomadis and hardjo rendered products of the expected sizes with both pairs of primers. The specific DNA region was also amplified from isolates from Argentina belonging to serogroups Canicola and Pomona. Both L. biflexa serovar patoc and L. borgpetersenii serovar tarassovi rendered a negative result. Conclusions The DNA sequence related to the antigen mimicry with equine cornea was not exclusively found in serovar pomona as it was also detected in several strains of Leptospira belonging to different serovars. The results obtained with L. biflexa serovar patoc strain Patoc I and L. borgpetersenii serovar tarassovi strain Perepelicin suggest that this sequence is not present in these strains, which belong to different genomospecies than those which gave positive results. This is an interesting finding since L. biflexa comprises nonpathogenic strains and serovar tarassovi has not been associated clinically with equine uveitis.


Background
Leptospirosis is a widespread disease that affects wild and domestic animals as well as humans [1]. Animals which have recovered from acute leptospirosis may develop a carrier condition, shedding leptospires in their urine [2].
Horses infected with Leptospira present several clinical disorders, one of them being recurrent uveitis or iridocyclitis [3,4]. In equine recurrent uveitis (ERU), episodes of acute anterior uveitis are separated by quiescent periods of variable duration [5]. Acute signs include blepharospasm, lacrimation, photophobia, myosis, corneal edema and vascularisation, aqueous flare and hypopyon, although no single case shows these signs all together. With repeated attacks, the severity of the lesions becomes more pronounced. The inflammatory process may lead to anterior or posterior synechiae, cataract, iris atrophy, retinal detachment, lens luxation and corneal opacity [3,5].
A common endpoint of ERU is blindness. Horses with uveitis associated with leptospiral seroreactivity are at increased risk of developing blindness with respect to horses which uveitis was attributable to other causes [6]. Serovar pomona has often been incriminated, although others have also been reported.
Parma et al. [7] showed that equine cornea and Leptospira share partial antigenic identity. In fact, corneal opacity was provoked by inoculating horses either with killed Leptospira or equine cornea. Based on these findings, ERU is considered an organ-specific autoimmune disease [8].
The epitopes shared between Leptospira and equine cornea belong to a protein structure located inside this bacterium [9]. A leptospiral DNA fragment that encodes cross-reacting epitopes was cloned and expressed in Escherichia coli as a β-galactosidase-fusion protein [10]. This clone, isolated by expression screening with a polyclonal serum raised against equine cornea proteins, encodes a 90 kDa protein of serovar pomona. Antibodies directed against this leptospiral antigen recognized a 66 kDa equine corneal protein. These findings suggest that an immune response to that leptospiral antigen participates in pathogenesis of equine uveitis. Therefore, it would be interesting to investigate whether this region is spread in the genomes of several serovars of Leptospira.
Before 1989, taxonomy of the leptospires distributed these bacteria between two species, Leptospira interrogans (pathogenic) and L. biflexa (saprophytic) comprising over 200 serovars [11] on the basis of surface agglutinins. For convenience, antigenically related serovars are organized into serogroups.
In the present study, we began with the analysis of serovars of Leptospira by PCR with two primer pairs designed to amplify the region antigenically related to equine cornea.

Results and Discussion
A region of the leptospiral DNA fragment that encodes epitopes involved in the antigenic cross-reactivity between this bacterium and equine cornea was subcloned and sequenced. The obtained sequence was deposited in the GenBank database under accession no. AY046585. Two primer pairs were designed from the sequence to amplify that region by PCR: S3a/S3b and S4a/S4b, which produce 253 and 152 bp PCR amplicons, respectively.
Several samples of leptospiral DNA, from reference strains and isolates from Argentina, were analysed by PCR with these primers (Table 1). Reference strains from serovars canicola, icterohaemorrhagiae, pomona, pyrogenes, wolffi, bataviae, sentot, hebdomadis and hardjo rendered products of the expected sizes with both pairs of primers (Fig. 1). The specific DNA region was also amplified from isolates from Argentina belonging to serogroups Canicola and Pomona. A single product was obtained in all cases of positive amplification.
Another set of primers (G1/G2 and B64-I/ B64-II) previously described [17,18] was used as a positive control of leptospiral DNA (results not shown). With these primers, L. borgpetersenii and all L. interrogans strains here mentioned were amplified, suggesting that the failure to amplify the sequence related antigenically to equine cornea from L. borgpetersenii serovar tarassovi was not due to template degradation or to the presence of PCR inhibitors Therefore, this strain is not likely to carry this sequence or presents variations in DNA fine structure that prevent PCR primer annealing.
As primers G1/G2 and B64-I/ B64-II are specific for pathogenic leptospires, the presence of leptospires in the aliquot taken from the culture of L. biflexa serovar patoc was determined by microscopic observation.
These results suggest that in the nonpathogenic strain analysed, the sequence related to the antigen mimicry with equine cornea is not present. In addition, this sequence could not be detected either in one of the pathogenic strains. Interestingly, it belongs to a different genomospecies (L. borgpetersenii) than the other pathogenic strains According to Faine [2], there has been a tendency to extrapolate and generalize a conclusion to all leptospires from observations on one or few strains studied. But nowadays, newer advances on genetic groupings dictates that much of the conventionally accepted knowledge has to be studied critically again with modern techniques to ascertain the extent of genotypic and phenotypic variation. This study contributes to the knowledge of the distribution of a DNA sequence, which is present in L. biflexa serovar pomona, among different serovars of Leptospira.
Additional studies with different strains of L. borgpetersenii will be necessary to know if this sequence is not present in this genomospecies. It would be also interesting to investigate whether corneal epitopes crossreactive with this bacterium are present in different breeds of horses and in other species of the family Equidae (as ERU is also the leading cause of blindness in mules).

Conclusions
A DNA sequence of sevovar pomona related to the antigen mimicry with equine cornea was detected in several strains of Leptospira belonging to different serovars, including reference strains and isolates from Argentina. Therefore, this sequence is not exclusively present in serovar pomona.
The results obtained with L. biflexa serovar patoc strain Patoc I and L. borgpetersenii serovar tarassovi strain Perepelicin suggest that this sequence is not present in these strains, which belong to different genomospecies than those which gave positive results. This is an interesting finding since L. biflexa comprises nonpathogenic strains and serovar tarassovi has not been associated clinically with ERU.

Bacteria and culture conditions
Reference strains of Leptospira used in this publication are listed in Table 1. Six clinical isolates from Argentina were also studied (3 belonging to serogroup Canicola and 3 to Pomona). Leptospires were maintained in Fletcher medium [19].

Sample preparation
Aliquots were taken from cultures, diluted 1/10 in water and boiled for 10 min. Five microliters were used as the template for PCR amplification.

Primers
Primer sequences for the indicated region were as follows. PCR amplifications were performed as follows: initial denaturation at 94°C for 2 min, followed by 30 cycles of denaturation at 94°C for 1 1 / 2 min, annealing at 66°C (primer pair S3a/S3b) or 57°C (primer pair S4a/S4b) for 1 1 / 2 min, extension at 72°C for 2 min and a final extension at 72°C for 5 min. Reaction products were analysed in a 1.5% agarose gel stained with ethidium bromide.