We identified a high prevalence of infection with a clonal L. borgpetersenii genotype in a diverse urban R. norvegicus community in Johannesburg, South Africa’s largest metropole. The spatial structuring exhibited by the two most common R. norvegicus haplotypes and evidence for the presence of at least four haplotypes not previously detected in South Africa suggest that the R. norvegicus community in Johannesburg is a result of multiple invasion events. The most common haplotype (RN01), along with two new haplotypes (RN16 , RN17), was genetically similar to haplotypes previously identified in Johannesburg and Durban [6], a coastal city on South Africa’s east coast. However, all the haplotypes identified in Johannesburg were distinct from the single haplotype identified in Cape Town. Further Rattus spp. genotyping, both within Johannesburg and nationally, would be valuable to better understand the invasion history of these key hosts.
Despite the heterogenous R. norvegicus community identified in Johannesburg we identified a single clonal Leptospira genotype consistent with L. borgpetersenii serogroup Javanica in R. norvegicus across the metropole and in the only R. rattus identified in this study. Although clonality is inferred based on sequencing a limited number of loci, for each of these phylogenetically informative loci [12] multiple samples were sequenced and no genetic variation was noted. A similar lack of genetic diversity in L. borgpetersenii serogroup Javanica strains in Rattus spp. hosts has been noted across Malaysia using pulsed-field gel electrophoresis (PFGE), the gold standard for genotyping of Leptospira [20]. Moreover, the lfb1 sequence from a single L. borgpetersenii infection identified as part of a mixed infection in Cape Town was also consistent with L. borgpetersenii lfb1 sequences identified in Johannesburg suggesting that, as in Malaysia [20], this strain may be widely distributed in South Africa. However, as this infection was identified as part of a mixed infection, further genotyping at additional loci was not possible.
The lack of L. interrogans in Rattus spp. in Johannesburg is notable, given the widely recognised host-pathogen association between Rattus spp. and serogroups in this species [21] and recent evidence that L. interrogans is found in R. norvegicus in Cape Town [7]. It is possible that, despite the complex invasion history suggested by R. norvegicus haplotyping, none of the R. norvegicus introduced to Johannesburg were infected with L. interrogans. However, even where studies have demonstrated significant levels of L. borgpetersenii infection in Rattus spp., L. interrogans was the more commonly detected species [20]. Leptospira interrogans and L. borgpetersenii differ in their ability to survive in the environment [22] and field studies in rodent hosts suggest that warm, moist environments favour the transmission of L. interrogans while environmental conditions are less important for the transmission of L. borgpetersenii [23,24]. Therefore, it is more likely that the environmental conditions in Johannesburg (cold, dry winters) explains the absence of L. interrogans, whereas environmental conditions in Cape Town (coastal, temperate conditions with wet winters and warm summers) allow the transmission of both Leptospira species.
Within Johannesburg, there were significant differences in the prevalence of infection between administrative regions, consistent with previous findings of high variance in infection prevalence in urban R. norvegicus in Copenhagen, Denmark [25]. Previous studies have suggested a link between increased L. borgpetersenii prevalence in rodent hosts and urban areas characterised by mixed residential and commercial use in Malaysian Borneo [23]. In this study, the region with the lowest prevalence (region B) includes upmarket residential areas and business districts with less habitat heterogeneity than other regions, which include a mix of residential, business and industrial areas, large informal settlements and semi-rural areas. In this study, spatial resolution below the level of regions was not possible. Therefore, further work is required to identify the drivers of spatial variation in Leptospira prevalence in Rattus spp. in urban areas in South Africa.
Although the prevalence of Leptospira spp. infection in Rattus spp. varies widely in Africa [4], the prevalence (44%, 75/171) identified here, and in a previous study undertaken in Cape Town [7], are amongst the highest identified in similar contexts in the region. For example, the prevalence noted in this surveillance study is similar to that identified in surveillance-based sampling undertaken in urban Antananarivo in Madagascar (49%, 47/96) [26], while the prevalence (67%, 8/12) identified in R. norvegicus during an outbreak in Cape Town [7] is similar to that identified during an outbreak investigation (68%, 17/25) undertaken in Reunion Island [27].