Viable leptospires could be recovered from four dogs with confirmed acute leptospiral infection, and further molecular and serological characterization enabled the identification of L. interrogans serogroups Icterohaemorrhagiae (strains 4, 5 and 9) and Canicola (strain 6) as causative agents for canine leptospirosis in the present study.
Despite canine acute infection caused by pathogenic Leptospira being largely reported in Latin America (29), the characterization of the infecting strains at a serovar or serogroup level is restricted to a few studies that have successfully isolated the pathogen in culture media (5,12,18,19,30–34). In recent years, isolation and proper characterization of the infecting strains in dogs with acute leptospirosis has shown that dogs can be exposed to a wide variety of serogroups (6), depending on the geographical region (e. g. urban or rural settings, tropical or temperate climate), local environmental and behavioral conditions (contact with wildlife fauna, sanitation, restriction of roaming behavior), and implementation of preventive measures against infection (use of vaccines, rodent control policies). Such variety may impose serious challenges to produce effective vaccines containing leptospiral strains circulating locally.
Furthermore, several countries have reported an antigenic shift due the extensive use of vaccines (10), giving rise to serovars not included in traditional vaccine compositions, highlighting the importance of using dogs as sentinels to access strains circulating locally.
Our results corroborate previous serological findings showing that Icterohaemorrhagiae is likely the main causative pathogen for acute canine infection in Brazil (5,19,35,36).These findings are also in agreement with the few studies that have consistently characterized Icterohaemorrhagiae as the main infecting serogroup in symptomatic dogs (5,19), reinforcing the hypothesis that dogs are highly exposed to environmental contamination promoted by rodents.
Serogroup Icterohaemorrhagiae is also considered the main causative agent for human leptospirosis in Brazil, notably in São Paulo city (37,38). The UNISA veterinary hospital is located at the south area of São Paulo, a region with the highest incidence rates for human leptospirosis, with an annual average of 57 reported human cases of leptospirosis between 2017 and 2019 (39). The area is characterized by having the highest precipitation rates in the city (40), with many humans living in low-income communities with poor sanitation, which can predispose not only humans, but also dogs to environmental exposure to pathogenic Leptospira spp. Under such conditions, dogs can act as important sentinels (41), and the diagnosis of acute canine infection and further characterization of the infecting Leptospira spp. may assist local health authorities to prevent leptospiral transmission by mitigating risk factors shared both by humans and dogs. Moreover, most districts in the area have a rural-urban interface configuration, with intense wildlife fauna interaction, which can lead to the transmission of leptospires maintained by wild reservoirs, reinforcing the importance of implementing local epidemiological surveillance using dogs as sentinels. In fact, a recent serological survey conducted by Diodato et al. (2016)(42) have found that Autumnalis is the most frequently serogroup found in asymptomatic semi-domiciled and stray dog populations from the south area of São Paulo. Interestingly, even in face of the successful characterization of Icterohaemorrhagiae and Canicola isolates in the present study, several dogs with acute infection confirmed by MAT presented the highest titers against Autumnalis serogroup.
Acute infection caused by Autumnalis was recently reported in humans and dogs from Japan(43), thus indicating a possible zoonotic transmission of this particular serogroup. In Latin America, serological surveys conducted in the islands of Barbados (44), Saint Kitts (45) and Trinidad (46) have also found Autumnalis as the predominant serogroup among dogs, and the isolation of serovar Bim (serogroup Autumnalis) from an asymptomatic carrier dog have also been described (31).
In Brazil, Autumnalis was identified as the main infecting serogroup found in stray and asymptomatic dog populations from different parts of the country (47–55), however confirmation of infection by isolation and molecular/serological characterization hasn’t been described yet. Unfortunately, isolation was not achieved in samples taken from dogs presenting the highest titers against Autumnalis in the present study, and the actual infecting serogroup attributed to those cases could not be determined. The MLVA analysis have also failed to identify serogroup identity in these samples, and despite L. interrogans infection was confirmed by 16SrRNA sequencing analysis, molecular typing methods using single-gene sequencing have poor discriminatory power to identify leptospires at a serovar/serogroup level. Therefore, serological evidence of Autumnalis infection should be interpreted with caution, as MAT results might have a low correlation with the actual infecting serogroup due to cross-reaction between serovars (56,57), especially in acute cases, where paradoxical reaction is common. Moreover, cross-reactions between serogroup Icterohaemorrhagiae and Autumnalis have been described after natural infection by Icterohaemorrhagiae serogroup (58), and the use of commercial vaccines containing Icterohaemorrhagiae representatives may also potentially induce cross-reactions between both serogroups (50,59).
In the present study, the association of PCR and MAT tests allowed the diagnosis of leptospirosis in 17 of the 31 dogs included in the study (54.8%), demonstrating that part of these animals would have had a false-negative result if only one of the tests had been performed. Despite the benefits of associating both tests for the diagnosis of acute canine leptospirosis, the PCR was able to diagnose more infected dogs when used as a single strategy, thus indicating the benefits of this method for the definitive diagnosis at the early stages of the disease, as previously reported (5,60).
The PCR results from 12 positive samples were confirmed by sequencing and phylogenetic analysis, and all sequences clustered with sequences belonging to L. interrogans representatives. This approach has allowed the identification of unexpected leptospires infecting dogs in Brazil, such as L. santarosai (5), L. kirchnneri (61) and L. noguchii (21), evidencing that this approach could be useful to circumvent the role of dogs in the transmission chain of leptospirosis.
The MLVA analysis have confirmed species identity of the isolates, with 100% agreement with MAT using polyclonal antibodies. This molecular technique has been successfully used to identify particular serogroups in clinical samples in which isolation could not be achieved (62). In the present study, most samples with presenting leptospiral DNA confirmed by the 16SrRNA PCR protocol have not yielded MLVA Positive results, probably due to the insufficient DNA concentration. Nevertheless, MLVA was able to identify infection by serogroup Icterohaemorrhagiae in two dogs with no MAT titers, and the application of this technique should be incorporated in future studies tackling canine acute leptospiral infection.
Despite the usefulness of molecular methods for the diagnosis of acute leptospirosis, the use of PCR may lead to false positive results in cases where non-leptospiral diseases mimicking acute leptospirosis are concomitantly present in dogs presenting chronic urinary shedding of leptospires. Moreover, no proper comparison between MAT and PCR results could be performed in consequence of the high lethality of the disease, which hampered the evaluation of MAT convalescent titers in most dogs.
The lethality was strikingly high among the suspected cases studied (67.7%), reinforcing previous studies from our group (5). These findings highlight the urgent need for implementing new therapeutic strategies in canine leptospirosis, as observed in clinical trials with human subjects (63). Additionally, the high proportion of dogs with no confirmation of infection reveals that investigation of other causes leading to clinical signs similar to acute leptospirosis should be investigated, and that the current diagnostic strategies should be improved for the diagnosis of the disease.