Is the presence of Leptospira in the environment merely random? An epidemiological explanation based on serial analysis of water samples

: Human leptospirosis cannot be investigated without studying zoonotic and environmental aspects of the disease. The objectives of this study are to explore the abundance of Leptospira in different climate zones of Sri Lanka and to describe the presence of Leptospira in the same water source at different time points. First, water and soil samples were collected from different parts of the country; second, water sampling continued only in the dry zone; and finally, serial sampling of water from ten open wells was performed at five different time points. Quantitative PCR of water and metagenomic sequencing of soil were performed to detect Leptospira. In the first component, 2 out of 12 water sites were positive, and both were situated in the wet zone. Very small quantities of the genus Leptospira were detected by metagenomic analysis of soil. Only 5 out of 26 samples were positive in the second component. Six, five, four, five, and six wells were positive in serial measurements of the third component. All wells were positive for at least one measurement, while only one well was positive for all measurements. Proximity to the tank and greater distances from the main road were significant risk factors associated with well positivity for Leptospira . The presence of Leptospira was not consistent, indicating the random abundance of Leptospira in the natural environment.


Introduction
Integrating human, animal and environmental health is essential in the control and prediction of zoonotic diseases. While investigations on animal and human interfaces are increasing, greater incorporation of environmental and ecosystem components is highlighted as a missing link in the One Health approach(1). Leptospirosis, a globally widespread and neglected tropical disease, also lacks adequate investigations linking animal and environmental factors to human infection. Various definitive and intermediate hosts, such as livestock, domestic pets, and wild or feral animals, harbour Leptospira in their proximal convoluted tubules of renal nephrons and excrete Leptospira via urine(2). These excreted Leptospira enter the human body through abrasions of the skin, mucus membranes or conjunctiva and cause leptospirosis (3). Different mechanisms have been acquired by Leptospira for adaptation to different environments(4).
As leptospirosis is a zoonotic disease transmitted by mammals, birds, rodents and marsupials, people who have direct contact with animals or animal products or reside or work close to animal habitats are considered at risk for infection(2). Hunters(5), sewer workers(6), butchers(7), veterinarians(8) and dairy farmers(9) are reported as major risk groups for the disease through direct exposure to animals, whereas farmers (10) and mine workers(11,12) have exposure via contaminated water sources. Studies have shown that contaminated water is a major source of disease transmission, as the disease is associated with floods, rainfall and recreational activities in water (13,14). Unlike in direct exposure, Leptospira has to enter the host within a short period after being shed into the environment or has to survive in water for a considerable period of time to cause disease by water contamination. Evidence suggests that Leptospira can survive in water for several days to more than one year(15). Additionally, it has been revealed that Leptospira can cause infection in susceptible individuals even after prolonged starvation of the pathogen(15). However, not all people who are exposed to contaminated water develop infection. This phenomenon warrants further exploration of the mechanism of Leptospira transmission.
Sri Lanka is a leptospirosis hotspot(16,17), and the disease causes significant morbidity and mortality despite its underestimation in Sri Lanka(18,19). The major modes of exposure to leptospirosis in Sri Lanka are paddy farming and working in gem mines(20). This finding indicates that indirect exposure through water sources is more common in Sri Lanka than direct exposure to animals. Evidence suggests that the infecting species and clinical patterns of leptospirosis vary among geographical locations in the country (21). This indicates that the natural survival of Leptospira could vary among those areas. There are three major climate zones in Sri Lanka, namely, the wet zone, the dry zone and the intermediate zone (22). The wet zone receives high rainfall on average and frequently reports more leptospirosis cases than other zones, while the dry zone reports leptospirosis cases predominantly during the rainy season(23). Livestock, farming practices and wildlife are also different among these zones. All these factors may lead to varying degrees of Leptospira survival in natural water sources. The objectives of this study are to explore the presence of Leptospira in the environment around human habitats where leptospirosis cases are reported in different climate zones and to perform a time series evaluation of the abundance of Leptospira in natural water sources, the main human-animal interface of disease transmission.

Results
Water and soil sample collection was performed at 12 sites in nine districts representing five out of the nine provinces of Sri Lanka (Table 1). Of the water samples tested from 12 sites, only the samples from Mawanella (an abundant paddy field) and Mathara (a paddy field) tested positive for Leptospira ( Figure 1). The 16S rRNA amplicon sequencing data were analysed from eleven sites. Taxonomy assigned based on the RefSeq database via MG-RAST showed that in the soil microbiome, the relative abundance of the genus Leptospira was minute compared to that of other organisms. The highest relative abundance (105.6) was reported from the sample of Rathnapura 1 (Figure 1).
The 26 sites included in the second component (dry zone) included water samples from large human-made irrigation tanks/lakes (n=6), paddy fields (n=6), rainwater collections (n=4), rivers/natural water streams (n=4), natural water pools (n=2), water canals (n=2) and wells (n=2). Of these, a single site was strongly positive for Leptospira, while sites 2, 3, 22, 24 and 25 were positive ( Figure 2). The strongly positive site 9 was a well from which water was being used for both agriculture and household activities but not for drinking.    Table 2 summarizes the local and environmental factors associated with the number of times that each well was positive. All the wells shared similar characteristics, while the positivity was higher in the wells situated close to the water tank (lake) and away from the main road.  Figure 4 shows the association between well positivity and distance from the nearest water tank in kilometres. It clearly shows that when the distance from the water tank is reduced, the number of times that the well is positive increases.

Discussion
Human leptospirosis is mainly investigated in a disjointed manner, ignoring the zoonotic nature of the disease and the importance of the animal-human interface in disease transmission. Focusing on the transmission process at the animal-human interface is required to explain the disease transmission patterns in zoonotic diseases. Cross-sectional studies with a single time point description of environmental contamination only partially explain the actual risks and transmission pattern of the disease. In this study, we aimed to describe Leptospira in soil and water together with serial sampling of water sources to describe the existence of Leptospira in the natural environment.
With regard to water samples, the finding that two sites in the wet zone were positive while none of the sites in the dry and intermediate zones were positive for Leptospira is compatible with the reported incidence of leptospirosis, as the wet zone reports nearly two times the number of cases compared to that in the dry zone(24). As the environment of the wet zone is favourable for the growth and survival of Leptospira, the probability of detecting the organism in samples is expected to be higher. The observed difference could also be due to the diversity of Leptospira in different geographical areas, as described previously(25). It has been shown that Leptospira can survive in vitro as well as in the natural environment through biofilm formation with the environmental microbiota. Therefore, Leptospira can survive even in nutrient-free environments(26,27). On the other hand, the nutrients required for Leptospira survival could be different in the two climatic zones, and further studies are needed to explain the differences we observed. Nutrient availability could be a main reason for the observed diversities of water samples between the climate zones. The diversity of the soil microbiome may be a contributory factor to the differences we observed, as shown in the 16S amplicon sequencing data of the soil samples tested at the same sites(28). Species-or strain-specific differences in the natural survival of Leptospira, with a specific focus on geographical, environmental and climatic factors, need further exploration(15).
An emerging hypothesis is that virulent Leptospira survive in soil for a long period and come to the surface when the soil is washed away during the rainy season(15,29,30).
Therefore, the probability of detecting Leptospira could be higher in wet zone due to its surface wetness throughout the year. In the dry zone, relatively low rainfall is received for a short period of time (22)

Limitations
The sensitivity of PCR is considered low when the concentration of Leptospira is low(33).
As Leptospira is diluted in water, there is a high probability of failing to detect the existing Leptospira from the source of water collection. We used a 2-step centrifugation protocol to concentrate Leptospira. Although it was an optimized procedure, there was a probability of losing a considerable number of Leptospira in the pellet of the 1 st centrifugation step.
Metagenomic analysis is highly dependent on the database used. Therefore, we accept that there is a probability of missing species of the genus Leptospira that are not present in the MG-RAST database. Direct comparison of the abundances of Leptospira in water and soil could not be performed, as two different analysis techniques were used. Although the PCR was negative, there was a theoretical possibility of Leptospira being present in place with the same water collection other than the site of sample collection.

Study design and setting
This study included three major components of environmental sample collections, as illustrated in Figure 6: island-wide (including most parts of the island) water and soil  Table 1, soil samples were collected from the same eleven sites where the water samples had been collected. Site Katugasthota (Figure 1) was a deep canal where we collected only water due to the practical inconvenience of soil sample collection. All sites were selected on the basis of probable exposure history of con-

Sample collection and transport
Four water samples were collected from each site, and a 1-metre gap was maintained between the sample collection locations within the sites. Ten millilitres of water was collected in a sterile 15 ml Falcon tube using a clean plastic container, and the lid was closed were sent for metagenomic analysis.

Sample Processing, DNA extraction and PCR testing
There is no optimized best method for concentrating Leptospira from water samples (34,35).
We found that a two-step protocol suggested by Paula et al. to concentrate Leptospira from urine produced better results than the available protocols for water(36). Therefore, centrifugation was conducted in two steps. First, samples were centrifuged at 3000 rpm for 5 minutes. Second, each supernatant was transferred to two microcentrifuge tubes (1.5 mL) and centrifuged at 15000 rpm for 10 minutes, and the supernatants were discarded. Samples collected from the same sites were pooled for the extraction of DNA. DNA was ex-

Definition of PCR positivity
A positive curve with melting temperature was considered a positive replicate.
If only one replicate was positive, the sample was considered positive. If two or more replicates were positive, the site was considered strongly positive.

Data analysis
The total number of times that a well was positive out of five measurements was considered the dependent variable. Wells were categorized into two groups based on the presence or absence of the risk factors shown in Table 2. Two-sample t tests were used to compare the number of positive risk factors present and the number of risk factors absent. A P value less than 0.05 was considered significant.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1.