We describe a new approach to identifying and quantifying pathogen-related leptospires in environmental water sources. A quantitative molecular and taxonomical method based on combining real-time reverse transcriptase PCR and DNA sequencing provided an important measure of human risk both for acquiring leptospirosis and for developing severe disease. Urban slums in Iquitos, Peru had high concentrations of pathogenic Leptospira and L. interrogans-type leptospires in surface waters such as those found in the market area and along the riverfront. In contrast, a rural area near Iquitos had lower concentrations of pathogenic Leptospira in surface waters such as wells. These Leptospira were less likely than those in the urban environment to be L. interrogans (the more highly pathogenic species). The concentration and species of leptospires in the water sources correlated with risk of severe human leptospirosis.
The outcome of human infection by Leptospira
can be divided into host and microbiological factors. Different genetic backgrounds, for example, containing mutations altering the major histocompatibility complex [25
] or innate immunity [26
] might modulate outcome of human infection. Virulence factors of Leptospira
may differ among infecting strains, and which ones may be variably expressed in different mammalian hosts [27
] are important factors likely to modulate outcome of infection. Here we show that the type and quantity of leptospires in surface waters that are potential sources of human infection are also related to the outcome of infection. This approach to the identification of uncultivated pathogens in environmental surface waters may be applicable to assessing risk for other water-borne infectious diseases such as salmonellosis, shigellosis, tularemia, and cryptosporidiosis.
We used molecular taxonomical approaches to directly link both the concentration and the species of pathogen-related Leptospira in environmental surface waters to the types of Leptospira that infect humans. We found that L. interrogans species, more commonly associated with severe human leptospirosis, were found significantly more often in urban than in rural environmental water sources where species other than L. interrogans, such as L. santarosai (associated with pigs and cattle) and intermediately pathogenic Leptospira spp. (of uncertain mammalian reservoirs) predominated. This differential distribution of leptospiral sequences in environmental water sources mirrored that of human isolates from urban and rural settings. We interpret these data to indicate that one important factor determining the risk of severe human leptospirosis, together with pathogen-host immune interactions, is both leptospiral species and concentration of pathogen-related Leptospira present in environmental surface waters. These findings are important because these factors are targets amenable to public health intervention, more so than bacterial or host factors. After validation in other settings and on a larger scale, the molecular identification and quantification of Leptospira in environmental water samples may be useful for guiding environmental remediation to prevent leptospirosis in endemic regions.
Epidemiological approaches to identifying risk factors for acquiring leptospirosis have traditionally focused on assessing potential zoonotic reservoirs, behavioral risk factors, and various sociodemographic features of human populations. While other investigators in the past have sought to identify pathogen-related leptospires in environmental sources using culture techniques, this approach has proven to be difficult and unreliable, being hindered in two ways: (1) saprophytic leptospires predominate in the environment that are morphologically similar to pathogen-related Leptospira but grow faster; and (2) the culture methods for identifying pathogen-related Leptospira are laborious, time-consuming, and insensitive. However, while DNA amplification techniques have been published and used for diagnosis of leptospirosis, they cannot identify all the leptospiral serovars in one reaction nor distinguish leptospiral species.
In both urban and rural settings, anthropogenic influences on ecology and environment seem to drive leptospiral transmission to humans. The quantification of leptospiral DNA copies in the water samples tested shows that in the living area of Belen, gutter and stagnant river water samples had higher concentrations of leptospires than did rainwater collection and underground water sources. This finding can be explained by the presence of chronically infected rodents, dogs, and pigs contaminating these water sources, as opposed to the underground sources, which are often protected from animal contact. In rural Padrecocha, the freshwater stream samples were positive most of the time compared to the freshwater well samples. This may be explained by the fact that many inhabitants use chlorine for water decontamination and detergent and soap for clothes washing and bathing near the wells; leptospires are inhibited at low detergent concentrations [11
]. Alternatively, it is possible that the Leptospira-
negative wells were relatively protected from environmental runoff, thus avoiding higher level contamination from animal urine. The high and persistent positivity of the freshwater streams (especially in the stream that runs through the village) may be due to the presence of cattle and pig farms and also of wild mammals along the streams. These findings are also important because they can explain the persistent sources of transmission in these populations; activities of daily living are strongly linked to water exposure (fishing, bathing, clothes washing, etc.).
In a previous cross-sectional study, the prevalence of anti-leptospiral antibodies was found to be significantly higher in residents of Belen (28%) than in rural areas near Iquitos (17%) [10
]. This observation suggests the possibility that protective immunity against severe disease from repeated infection may develop in areas with high leptospirosis transmission, especially if high frequency of infection leads to cross-serovar protection. These findings are also consistent with the finding that most of the cases of severe leptospirosis seen in Iquitos come from urban districts of the city rather than from periurban slums such as Belen or rural areas such as Padrecocha [9
]. Such possibilities need to be tested in prospective, population-based studies of acquired immunity to symptomatic leptospirosis.
This study had several limitations. First, because of technical limitations in leptospiral cultivation, we were not able to identify leptospiral serovars in the water samples. Second, the real-time PCR assay used in this study, previously reported to be specific for the detection of pathogen-related Leptospira
], is not as specific as previously reported—we were able to identify noncultivated leptospires of unknown pathogenicity. In fact, we were able to identify new Leptospira
of unknown pathogenicity, since none has yet been identified in mammals. We have provisionally termed this group of Leptospira
“clade C.” Because these clade C sequences were mostly found in rural settings, this quantitative real-time PCR cross-reactivity problem does not invalidate the basic finding of higher densities of pathogens in ground water from urban settings. Our data do confirm published reports, in that all positive samples we identified were confirmed by sequencing, even those samples that had high Ct values beyond the limit of detection in the standard curve (low leptospiral counts considered as suspicious [unpublished data]). Other problems with this methodology are expense and availability of the equipment. However, the advantages of this method make possible the study of large areas in a short period of time, such as we studied here, allowing for the rapid deployment of environmental risk assessment in endemic regions, particularly in the setting of epidemics. Finally, we did not use high-throughput sequencing of 16S rRNA gene fragments to provide a large-scale representation of leptospiral sequences in the water samples. Such a study is ongoing.
Identifying the strains present in environmental water samples may provide insight into the relevance of local mammal species to the dissemination of leptospires and occurrence of human leptospirosis in Iquitos. For example, most strains sequenced from Padrecocha belonged to possibly new intermediate leptospiral species (MAM, JN Ricaldi, and JMV, unpublished data). 16S rRNA gene sequences found in water sources have also been detected in cattle, pigs, and bats, suggesting that in this rural setting, these animals may contribute to human leptospirosis (CBC, MAM, and JMV, unpublished data). In Belen, where rat populations are dense, we detected 16S rDNA gene sequences similar to several different L. interrogans
serovars. Rats common in Belen (R. norvegicus
and R. rattus)
carry L. interrogans
with identical 16S rDNA gene sequences (unpublished data), suggesting the potential importance of these animals to the transmission of leptospirosis in the area. In addition, of 21 cases of leptospirosis recorded among patients from urban and periurban Iquitos and outlying rural areas, we noted that L. interrogans
clones were detected almost exclusively in urban/periurban habitats, while clones from rural areas significantly comprised other leptospiral genomospecies, including L. santarosai
and L. noguchii,
which have been shown to be carried by cattle and wildlife reservoirs (CBC, MAM, and JMV, unpublished data). Furthermore, in this study, L. interrogans
was found to be the primary cause of severe leptospirosis in Iquitos [9
]. Taken together, these observations suggest that real-time detection and sequencing are potentially useful for determining the significance of local reservoirs of leptospires, and that an association with rats is a significant risk factor for acquiring severe leptospirosis in Iquitos.
Leptospirosis continues to be overlooked globally as a major public health threat. Not only does leptospirosis have highly variable, nonspecific clinical presentations, but diagnosis is difficult and requires a high index of suspicion particularly with regard to potential environmental exposure. In the Amazon region of Peru, leptospirosis is an important cause of morbidity and mortality. More effective, timely, and efficient methods need to be deployed for rapid clinical diagnosis and for environmental risk assessment. PCR-based detection and identification of leptospiral DNA, which avoid the problems inherent to isolation of fastidious organisms from contaminated sources, can be useful for assessing and monitoring risk of human populations to Leptospira-contaminated water. The methodologies presented here need to be validated further in prospective studies in different regions of the world where leptospirosis is common. The researchers conducting these studies should provide their data to Ministries of Health and policy-making agencies critical for justifying public health and occupational medicine campaigns to control leptospirosis transmission. General health promotion should also continue to use common-sense strategies such as wearing shoes and eliminating trash from the home and local environments.