ANTI-LEPTOSPIRE AGGLUTININS IN EQUINE SERA, FROM SÃO PAULO, GOIAS, AND MATO GROSSO DO SUL, BRAZIL, 1996-2001

Equine leptospirosis can present a non-symptomatic form, an acute clinical form, or even develop chronically, causing reproductive alterations, such as abortion and recurrent uveitis. Since the prevalence of leptospirosis in several countries and regions is widely reported, the objective of this study was to verify the prevailing equine leptospirosis in different regions of Brazil. Sera from 1402 blood samples from horses of different age, sex, breed, and purpose were examined. These samples came from southeastern and central west states of Brazil. The method utilized was the Microscopic Agglutination Test (MAT), with 12 different Leptospira serovars. From the sera tested, 754 (54%) were positive for one (385) or more (372) serovars. These results were higher when compared to national and international levels. The most commonly found serovars were icterohaemorrhagiae (37.01%), suggesting exposure to rodents, castellonis (16.97%), and djasiman (15.19%). There were significant differences of reagents between sexes, and a tendency toward higher positivity with age. Distribution of sera-reagents related to aptitude showed a markedly higher value for work animals than for sporting ones. Higher rates were found for animals with undefined breed. There were no significant differences related to regional origin. As an indication of multiple exposure, significant associations were observed between the following serovars: castellonis and djasiman; castellonis and grippotyphosa; castellonis and copenhageni; castellonis and icterohaemorrhagiae; castellonis and pomona; canicola and pomona; canicola and djasiman; djasiman and copenhageni; icterohaemorrhagiae and djasiman; icterohaemorrhagiae and pyrogenes; copenhageni and pomona. These results showed the necessity of further studies on the epidemiology of this disease in equines and its relationship to human illness.


INTRODUCTION
Leptospiroses in equines can present a clinically non-symptomatic form, or develop to acute or chronic clinical form (13).The acute form, characterised by septicemia, encephalitis, and hepatonephritis is rare.Chronic manifestations are more frequent, with loss of weight, various degrees of conjunctivitis, ceratite, uveitis, and reproductive alterations, such as abortion (24).
The most commonly used method in the diagnostic laboratories is the microscopic seroagglutination test applying representative serovars strains as antigens.It is also used in investigating epidemics, since it indicates the serogroups present in a population (17).
Serologic research in equines has demonstrated that exposure to and consequent infection by this agent is distributed worldwide.The predominating serogroups varied with the leptospiric antigens utilized in the studies, and with geographic location (5), reflecting the animals exposure to serovars carried by other domestic or wild animals (10,14).
Considering the variety prevalent in different countries, this study aimed to verify the presence of anti-Leptospira antibodies in equines from different states of Brazil, its relation with age, sex, origin, and aptitude, as well as to examine the seroconversion associations between pairs of serovars, as an indication of multiple exposure.From the 1402 samples of equine sera examined, 71 (5.06%) were from males, 617 (44.01%) from females, and 714 (50.93%) had no information about the sex.The ages varied from 1 month to 19 years old.With regard to the aptitude, 445 (31.74%) presented data, 138 were sporting animals, and 307 work animals.In relation to the breeds, 408 (29.10%) horses had defined breed, mainly Quarter-horse, pure bred Arabian, and "Pantaneira", a local breed of western Mato Grosso state.In regard to the geographical origin, 659 samples were from São Paulo, 145 from Mato Grosso do Sul, 31 from Goiás, and the remainder, of unknown origin.
Agglutinations up to 1:100 sera dilution were considered positive.

Statistical Analysis
The association between the epidemiologic variables and the seroconvertion was evaluated using the Chi-square test, with α=0.05%, utilizing the EpiInfo software, version 6.04.The association between seroconversion in a horse at higher titers among those serovars, as an indication of multiple exposure, was evaluated using the Chi-square or the Fisher tests, depending on the number of observations.The titers higher than the top limit of the confidence interval of titers distribution for each serovar were considered high.

DISCUSSION
There are numerous reports of serologic studies on equine leptospirosis in several countries, demonstrating the importance of this infection in equines.Large occurrence of titers in apparently healthy animals suggests that the majority of the infections are non-symptomatic, with rapid elimination of the infecting serotype (14).Most of these studies denote the availability of laboratories, as well as private veterinary clinics that make these diagnoses (17).
Difficulty in interpreting serologic tests based on a single sample is reported, since the MAT results can be negative or low, during the acute phase of the illness, or in animals that aborted, due to the interval between the infection and abortion (22).Thus, it is suggested that seropositive result is an evidence of exposure to the agent, but it does not confirm the disease (3).On experimental infection of five horses with L. interrogans serovar copenhageni, there was an increase of serologic titers from the 7 th day after the inoculation.The highest level was reached on the 18 th day, when the geometric titer averaged 4851.3.After the 24 th day of inoculation, the titers began to decrease, varying from 50 to 400.On day 100, the titers geometric level was 152.8.The infected horses presented clinical alterations, such as hyperthermia, anorexia, jaundice, petechial hemorrhages in the mouth, and recurrent uveitis (13).
In Canada, titers for the icterohaemorrhagiae, bratislava, copenhageni, and autumnalis serovars were commonly found, 94.6, 56.6, 46.5, and 43.5%, respectively (16).In the United States of America (USA), it was observed that 56% of the 2967 positive animals reacted for one or more serovars, with prevalence of bratislava (4).In France, among 1645 horses, a prevalence of 67.0% was reported, with titers of 200 or over, and predominance of the Australis and Gryppotyphosa serogroups (21).
However, considering the lack of definition for significant serologic titers of infection in the equine species (9), as well as the utilization of various Leptospira antigens, and the different geographic origins (14), it is difficult to compare results.
In the present work, greater response to icterohaemorrhagiae was observed (37.01%), which was also found in other studies in Brazil (1,7,18).Infection by the icterohaemorrhagiae serogroup is maintained by rodents that share the environment with equines, what can explain this response (16).The risk factors associated with equines exposure to the pomona, autumnalis, and bratislava serovars were studied in New York, and it was concluded that seroconversion of these serovars is also related to exposure to rodents (3).
Other animals may be a source of infection by Leptospira for the equine species, as evidenced in this study.
Great number of reactions to the castellonis serovar (16.97%), which belongs to the Ballum group, emphasize the importance of the constant association between horses and bovines, since this serogroup is very common in that species (21).The same reason can explain the presence of the hardjo serovar (4.49%), once bovines are responsible for maintaining it (10).
Reaction to the canicola serovar (5.58%) can be explained by the fact that horses are usually next to dogs, carriers of this serovar (17).The pomona serovar is maintained by swine and bovines, and the equine species is considered an accidental host (23).It is recognized as the most important agent of recurrent uveitis in horses (5), so that its occurrence in healthy animals is smaller than in clinically affected ones (23).In other studies in Brazil, a higher prevalence of this serovar was observed (2,6,20), whereas in this work, anti-Leptospira agglutinins were observed in only 5.42% of the samples.
Horses are constantly exposed to the pomona serovar, probably due to the presence of rodents.The animals infected by pomona serovar in this study showed titers over 1600, probably because these decreased more slowly and the animals remained reagents for a longer period (3).
It was suggested that the bratislava serovar is adapted to the equine species (10), however, in this study, the seroreactivity against this serovar was 2.57%, whereas in Rio de Janeiro it was 11.70% (18).It has been proposed that exposure to this serovar is related to wild animals (3).
Other researches on seroepidemics in equines, carried out in Brazil, did not utilize this serovar (18).
Incidence of infection was significantly higher among females, probably due to herd management.This result disagrees with the ones obtained in Canada, where similar results for both sexes were found (16).
In this study, higher exposure to the agent was observed among animals of undetermined breed and those used for work, in agreement with the data found in Canada ( 16) and USA (3); confirming the importance of management practices on the risk of being seropositive to

Leptospira interrogans.
With regard to the age, it was verified that a tendency toward positive response increased with age, confirming the results obtained by other authors (3,16,19), and presuming that with the elapse of time, the animals have higher chances of contracting the infection.
Serologic reactions to various serovars are common, which can be explained by multiple exposure to more than one serovar.It is also possible that some cross-reactivity occurred between the serovars used in MAT (4).Since it is believed that cross-reactivity occurs predominately in lower titers, this evaluation was based on the association between serovars on samples with titers higher than the top limit of confidence interval of titer distribution, according to the Barwick et al. study (4).Therefore, the incidence of reactions to more than one serovar must be considered not only as a result of diagnostic test of the serum but also a reflection of the inaccuracies of the MAT (19).
As an indication of multiple exposure, significant associations were observed for all pairs of antigen tested (pomona, hardjo, icterhaemorrhagiae, gryppotyphosa, canicola, autumnalis, and bratislava), except for gryppotyphosa / pomona and gryppotyphosa / autumnalis, which presented significant associations.In order to evaluate these associations, titers were considered high for values equal to or higher than 1600.Cross-reactivity between some of the serovars could explain the large number of low titers (Table 1), also noted in another study (4).Despite the significant association of high titers between serovar pairs indicating multiple exposure (4), associations between the copenhageni and icterohaemorrhagiae serovars, pertaining to the same serogroup (Icterohaemorrhagiae), and between hardjo and wolffi, both of the Serjoe serogroup, suggest the occurrence of crossreactivity among these serovars.In fact, cross-reactivity between the serovars hardjo and wolffi are fairly common, since both belong to the same serogroup and present similar antigens (8).
Additional studies on the epidemiology of this disease in equines are required, including isolation of the bacteria, in order to maximize the measures of control effectiveness and to prevent this disease in the equine and other mammal species (4).In Brazil, Leptospira interrogans, serovar icterohaemorrhagia, was isolated from an aborted equine fetus (12).
Further references of isolation and identification of Leptospira in equines from Brazil are not found in literature, with a large blank space on the true participation of some serovars in the genesis of this clinical disease in the equine species.

CONCLUSION
High prevalence of anti-Leptospira agglutinins in horses suggests that animals are possible carriers, eliminating the agent with urine into the environment, consequently infecting other animals, including man.These results show the necessity of more studies regarding the epidemiology of this disease in equines and its relationship to the human disease.
Samples of equine sera from various states of the southeast and central-west of Brazil were examined and sent to the Laboratory of Zoonosis Diagnostic, Department of Hygiene and Public Health, School of Veterinary Medicine and Animal Husbandry, São Paulo State University (UNESP), Botucatu, São Paulo State, Brazil, for the Microscopic Agglutination Test (MAT), from 1996 to 2001.

Figure 1 .
Figure 1.Results of microscopic sero-agglutination for Leptospira spp in 158 samples of equine serum, according to the age, Botucatu, 2003.

Table 2 .
P Values for associations between serovar 1 pairs of Leptospira in samples with titers higher than the upper limit of confidence interval of titer distribution, calculated by the Chisquare or Fisher test, according to the number of observations, Botucatu, 2003.

Table 3 .
Occurrence of anti-Leptospira agglutinins in equines, according to origin, number of animals, percentage of reagents, and most frequent serovar.Botucatu 2003.