Spatiotemporal assessment of pathogenic Leptospira in subtropical coastal watersheds

Leptospirosis is a growing threat to global public health, manifesting as an emerging and prevalent zoonotic disease in tropical and subtropical areas (Vijayachari et al. 2008; Hartskeerl et al. 2011; Haake & Levett 2015; Dhewantara et al. 2019; Karpagam & Ganesh 2020; Truitt et al. 2020). Leptospirosis contributes to an estimated 60,000 deaths annually globally (Costa et al. 2015). This pathogen-associated disease has been characterized and classified as a neglected tropical and subtropical disease by the World Health Organization (WHO 2003, 2012; Serino-Silva et al. 2018), highlighting the need for increased research efforts to comprehend its epidemiology and control its spread (WHO 2003; Torgerson et al. 2015).

Leptospira, the causative agent of leptospirosis, is considered most prevalent during periods of substantial precipitation (Liang & Messenger 2018) and in agriculture regions, recreational areas, and flood-prone areas (Dobigny et al. 2015). This spirochete gram-negative bacterium, distinguished by its unique helical structure, is currently divided into 35 species distributed across four subclades (Guglielmini et al. 2019). Pathogenic Leptospira species, including Leptospira interrogans, Leptospira borgpetersenii, Leptospira noguchii, and Leptospira weilii, all belonging to the P1 subclade, pose significant health risks (Pappas et al. 2008). Animal vectors, such as canines, livestock, bovines, rodents, and mustelids, play a significant role in the transmission of Leptospira and can affect the epidemiologic patterns of this zoonotic disease (Levett et al. 2001; Pratt & Rajeev 2018). There are many clinical manifestations, such as fever and myalgia, as well as jaundice, nephritis, pulmonary hemorrhages, and cardiac arrhythmias, making it difficult to monitor and control the disease because it is frequently misdiagnosed and underreported (Monteiro et al. 2021). While some human infections are asymptomatic or moderate, some infections can progress toward a serious illness with multiple organ involvement and an increased fatality rate.

Historically, leptospirosis surveillance and research were negligible in arid and subtropical climates of developed nations (Woo et al. 1997; Desai et al. 2009) compared to temperate and humid regions. This oversight obscured the potential risk factors and prevalence patterns in subtropical coastal communities. Coastal regions with mixed-use watersheds have emerged as critical transmission centers for Leptospira, fueled by a complex interaction between environmental, climatic, and human factors (Bharti et al. 2003; Lau et al. 2010). Freshwater streams provide interconnections between upland watersheds and estuaries, providing a pathway for zoonotic microbes and transmitting agents such as rodents and canines. Bacteria can spread through the complex nexus of these connected aquatic ecosystems and have deleterious socioeconomic impacts (Hagan et al. 2016; Casanovas-Massana et al. 2018).

Quantitative polymerase chain reaction (qPCR) analysis is a significant advancement in the field of microbial water quality assessment, as its ability to detect waterborne pathogens and trace the origins of contamination raises the bar notably for environmental science and public health investigations (Harwood et al. 2014; Bridgemohan et al. 2023). qPCR can distinguish fecal contributions from a wide range of host species, including humans, cattle, canines, ruminants, and avians (Zhang et al. 2014). These species contribute to a diverse mosaic of microorganisms in aquatic environments via the alteration of biogeochemical processes, toxin production, biofouling, and oxygen depletion (Falkowski et al. 2008; Paerl & Otten 2013; Dang & Lovell 2016). This fecal source differentiation, informed by qPCR-derived cycle threshold (Ct) values, represents a paradigm shift in water quality and detection strategies. The efficacy of qPCR extends beyond identifying fecal sources; it is also demonstrated in detecting and quantifying waterborne pathogens (Simpson et al. 2002; Borchardt et al. 2021). This feature significantly improves the ability of the scientific and healthcare communities to conduct comprehensive microbial water quality assessments, thereby bolstering strategies to detect and manage the spread of waterborne diseases.

The Pensacola Bay Estuary System and its surrounding watershed comprise a diverse aquatic ecosystem and are a possible zoonotic exposure area for Leptospira as a result of their extensive exposure to a diverse array of environmental variables and diverse fecal host populations (Smith et al. 2009; Tao et al. 2012; Schrandt et al. 2016; Moss & Snyder 2017; Kapoor et al. 2018; Bridgemohan et al. 2023). Leptospirosis epidemiology related to other environmental variables in the Pensacola Bay System (PBS) can be used as a model to illustrate how this microorganism appears in subtropical regions: factors such as frequent rainfall, storm events, warm temperatures, and the potential for inundation and varying levels of residential, industrial, and recreational use (Croteau et al. 2023) can influence the potential for pathogens such as Leptospira to be present in freshwater, estuarine, and marine environments (Hacker et al. 2020).

The goal of this study was to characterize the spatiotemporal distribution, prevalence, and co-occurrence of pathogenic Leptospira, fecal indicator bacteria (FIB) Escherichia coli, and enterococci in diverse aquatic ecosystems in the PBS. Our focus on the PBS is intended to contribute to the global comprehension of leptospirosis by yielding insights applicable to this region and similar regions worldwide. This task has been made achievable with the application of qPCR techniques, the rapid detection of Leptospira in environmental matrices, and the identification of likely sources (Bachoon et al. 2010; Rawlins et al. 2014; Riediger et al. 2016). This was confirmed by successfully implementing a hydrolysis probe qPCR assay targeting the LipL32 gene (Stoddard et al. 2009; Rawlins et al. 2014; Truitt et al. 2020). The technique permits the robust detection and characterization of pathogenic Leptospira spp. in various aquatic ecosystems, providing novel insights into the molecular microbial ecology of vulnerable coastal subtropical regions.

The area has a subtropical climate and is affected by its proximity to the Gulf of Mexico. The summer months have high annual precipitation (30–50% of the 1,600 mm, on average) and exhibit less temperature variation along the coast than in the inland regions. Hurricanes and tropical storms frequently impact the summer and autumn seasons (NFWMD 2017).

The study involved a systematic analysis of the spatial and temporal variations in water physical and chemical properties based on sonde measurements and water samples collected monthly at 44 locations across all four seasons. The water samples and measurements were collected at a depth of 0.3 m below the surface. Basic water quality measurements were measured during each sampling event using an Aqua TROLL 500 portable multiparameter sonde (In situ Inc.), calibrated monthly following the manufacturer-recommended procedures outlined in the instrument Vu-Situ software. For sampling, three samples of 100 mL each were taken from each site: one for quantifying E. coli, one for quantifying enterococci, and one for DNA extraction to identify Leptospira. Each sampling container was rinsed three times with the surface water from its respective collection point to reduce the likelihood of contamination or data distortion. After rinsing and sample collection, containers were marked for traceability and securely sealed to prevent contamination from the outside. The containers were placed on ice to preserve sample integrity. Six hours after the initial collection, samples were promptly transported to the Watershed Management Laboratory at the West Florida Research Education Center in Milton, Florida, USA, where they were processed for further analysis.

Several environmental and weather variables were gathered using remotely sensed sources. Precipitation data were obtained for each site from nearby Weather Underground stations for 24- and 72-h intervals before each sampling. Land cover was characterized using the 2019 National Land Cover Dataset from the US Geological Survey. The spatial analysis for land cover focused on a 500-m radius region upstream of each inland (stream) site and a 500-m radius centered on the sampling location for coastal sites. The land cover within each circle was stratified into five categories: forested areas (wooded wetlands, evergreen forests, or mixed deciduous forests), open water (parts of the Pensacola or Perdido Bay Estuary System), low-intensity development (<25% impervious), moderate-intensity development (25–50% impervious), and high-intensity development (>50% impervious). Land cover for each site was classified based on the predominant type; some sites were evenly divided between two categories (open water/high-intensity developed, open water/medium-intensity developed, and medium-intensity developed/forested), resulting in seven distinct groupings.

Samples for FIB analysis were processed to quantify the most likely most probable number (MPN) of E. coli and enterococci using the Colilert^®-18 and Enterolert^®-18/Quanti-Tray^® 2000 system from IDEXX Laboratories following NEMI Standard Method 9223B (USEPA 2012). IDEXX Quanti-trays were incubated at 35.5 and 42.5 °C for E. coli and enterococci, respectively. The most likely number (MPN/100 mL) was derived based on the number of fluorescent cells using statistical quantification tables.

A total of 132 water samples were analyzed for Leptospira, part of a larger sampling procedure across the Pensacola Bay region; samples were selected for Leptospira analysis if they had high levels of E. coli (>410 MPN/100 mL), with an additional 20% of samples exhibiting low E. coli concentrations to ensure a comprehensive inclusion of various sites, mainly urban and peri-urban areas. For DNA extraction, samples were filtered with a 0.45-mm-pore nitrocellulose membrane filter (Type GS, Millipore, Billerica, MA, USA) within 6 h of collection. To process the filters, we used a modified methodology described by Bachoon et al. (2010) utilizing the MoBio UltracleanTM Soil DNA Kit (Carlsbad, CA, USA). The extracted DNA was identified via a Nanodrop ND-1000 Spectrophotometer (Wilmington, DE).

Extracted DNA from samples was subjected to a microbial assay to determine the presence of L. interrogans, known to cause leptospirosis, using qPCR. This technique accurately identified Leptospira bacteria in environmental samples (Rawlins et al. 2014; Truitt et al. 2020). The bacteria were detected by targeting a 242-bp segment of the LipL32 gene, and each sample was analyzed in duplicate using the Qiagen QuantiTect Probe PCR kit and the Bio-Rad CFX96 system (Hercules, California 94547, USA) based on the methodology specified by Truitt et al. (2020).

For each qPCR reaction, we used 10 ng of extracted DNA per 20 μL of reaction mixture. The reaction mixture also included 500 nM of each primer (as detailed in Table 1) and 200 nM of the probe. The initial temperatures (95 °C for 15 min) were followed by 40 cycles of denaturation (95 °C for 10 s) and annealing/extension (64 °C for 30 s). L. interrogans serovar Pomona and E. coli strain K-12 provided positive and negative controls. In addition, nontemplate controls (NTCs) were included to confirm the assay's specificity (Truitt et al. 2020). In each qPCR assay conducted, we rigorously included NTCs to ascertain the specificity of our assays and to meticulously screen for any potential contamination. These NTCs, comprising reaction mixtures devoid of DNA template, served as critical controls to confirm the absence of nonspecific amplification or contamination. This practice ensured the integrity and reliability of our assay results. In addition, we meticulously organized the layout of standards within each assay, spanning a comprehensive range of known concentrations. This strategic arrangement facilitated the precise quantification of target DNA in our samples by establishing a robust calibration curve for accurate analysis.

After initial detection, affirmative samples were reanalyzed in triplicate. The number of genome copies was quantified using standard curves (Truitt et al. 2020). Using DNA extracted from L. interrogans serovar Pomona and E. coli strain K-12, both positive and negative controls were constructed. Also included were NTCs using DNA-free containers. Using the genome size of L. interrogans (4 659.9 Mb), the method outlined by Levett et al. (2005) was utilized to calculate genome equivalents. The assay's lower detection limit was determined using positive control DNA serial dilutions. Variations in quantification cycle (Cq) values between a sample and its 10-fold dilution were analyzed to identify potential PCR inhibitors (Ahmed et al. 2020; Truitt et al. 2020). Only PCRs with an efficiency of >90% were considered acceptable.

We utilized JMP Pro (JMP^®, Version 17, SAS Institute Inc., Cary, NC, 1989–2023) to analyze the nature of the relationships between Leptospira, FIB, water quality parameters, climatic variables, seasons, and land cover for the 132 samples with the highest levels of bacteria that were subjected to DNA extraction and Leptospira analysis. These statistical techniques included bivariate correlations, summarized with a Pearson bivariate correlation analysis, and generalized linear modeling (GLM), with the goal of providing insight into whether differences in these parameters have a substantial effect on the presence of Leptospira.

The qPCR assay for L. interrogans determined by DNA isolated from water samples using a Ct value below 40 was positive, and the efficiency was significant (p ≤ 0.9870). The assay's detection threshold was set at five genome copies of L. interrogans. The LipL32 gene qPCR assay indicated the presence of pathogenic Leptospira in the water samples. Ten distinct locations exhibited positive results for the pathogen during various months and seasons, highlighting the spatiotemporal potential of this pathogen.

Leptospirosis is a zoonotic pathogen that is spread by rodents, domesticated pets (canine), wildlife, and livestock (McDonough 2001) and is exacerbated by poor sanitation and environmental events, such as flooding (Grassmann et al. 2017). It can be transmitted indirectly through the eyes, mouth, and nose through contact with contaminated soil, water, or animal tissues (Zhao et al. 2019; Govindan et al. 2021). Leptospirosis remains a threat, particularly in tropical and subtropical systems (Goarant 2016), such as the PBS. Leptospirosis imposes a significant burden on public health worldwide (WHO 2011): estimates suggest a prevalence of over a million cases per year, as well as a mortality rate of approximately 59,000, highlighting the potential severity of this disease (WHO 2018; Warnasekara et al. 2019). Given the prevalence of underdiagnosis and underreporting, particularly in resource-limited settings, it is likely that these numbers understate the actual impact of leptospirosis. The disease's clinical overlap with other illnesses, such as meningitis, hepatitis, and a spectrum of fevers, contributes to the likely underestimation of its prevalence (Terpstra 2003; Musso & La Scola 2013; Warnasekara et al. 2019).

Our results build on existing research demonstrating the utility of using the LipL32 protein as a marker for identifying Leptospira (Xue et al. 2010; Schreier et al. 2013). The spatial and temporal distribution of Leptospira in PBS from January to December 2022 highlighted the prevalence of this species in response to varying land cover types. Bacteria are found in waterbodies that flow through or are adjacent to a range of land cover types from rural to developed areas, high to low levels of development, and mixed-use areas adjacent to open water systems. It was more common in lower-order streams than in larger streams, with fourth- to seventh-order streams lacking detections. Similarly, Leptospira was detected in smaller bayous and bay samples but not in marine samples. This trend could suggest a possible dilution effect: As the body of water size increases, the detection capacity decreases. Previous research has found more definitive relationships between the spatial and temporal characteristics of Leptospira and environmental factors and climate (Mann 2015; Lopez et al. 2019). Although other studies have found the disease to occur predominantly in warm, humid seasons (Vanasco et al. 2008; Desvars et al. 2011), our findings reveal the presence of Leptospira in samples collected in March, April, August, September, November, and December. This suggests a broader temporal distribution of the disease than previously documented. Our findings reveal a broader temporal distribution of Leptospira, with detections across multiple seasons, including both warmer and cooler months. This may indicate unique environmental or ecological conditions in our study area that facilitate the survival and proliferation of Leptospira beyond the traditionally recognized peak seasons.

In addition, in contrast to the previous research, we found no greater prevalence of Leptospira following rainfall compared to during dry periods; in other regions, the incidence of leptospirosis is amplified during extreme climatic events such as heavy rainfall, floods, and shifts in the Oceanic Niño Index (Schneider et al. 2012; Weinberger et al. 2014). In addition, its prevalence has been found to correlate with natural disasters such as hurricanes and typhoons (Su et al. 2011), as well as exposure to nature during work or recreational activities (Gaynor et al. 2007; Burton 2015; Bierque et al. 2020). In other subtropical countries, such as Argentina, where leptospirosis is a notifiable disease, there is a dearth of precise epidemiological data (Vanasco et al. 2008). Typically, outbreaks occur in an episodic fashion, primarily in reaction to alterations in the environment, such flooding (Izazola et al. 2018) or excessive precipitation (Guimarães et al. 2014). Notably, floods significantly affect rodent reservoir behavior, reducing their natural habitats and prompting their migration toward human settlements (Lopez et al. 2019). This subsequently increases the likelihood of disease transmission. However, the role of various animal reservoirs in the spread of disease within comparable regions remains obscure.

Other research in Florida USA's PBS has indicated that elevated FIB levels are common (Genthner et al. 2005; Albrecht et al. 2022), which may be caused by fecal contamination from human, sewage, ruminant, canine, and avian origins (Bridgemohan et al. 2023). It also revealed the presence of other pathogens, such as Campylobacter jejuni, and parasitic organisms (Hellein et al. 2011; Gordon et al. 2013). However, none of the studies reported the presence of zoonotic organisms such as Leptospira. In our study, we identified that the prevalence of Leptospira had no clear association with any specific land cover. In addition, a notable seasonal pattern emerged, with milder, rainier months exhibiting an increase in Leptospira incidence (Table 3). These results suggest the need for a broader analysis of the prevalence of pathogens such as Leptospira to develop meaningful interventions and management actions to improve safety in freshwater and coastal systems.

The relevance and significance of our findings pertain to the management of water resources in relation to public health. Most notably, our results suggest that the reliance on a single FIB as a metric for evaluating bacteriological water quality may oversimplify microbial dynamics. Such an approach may inadvertently obscure more nuanced microbial risks, a point that becomes apparent when assessing potential threats such as leptospirosis. Our investigation revealed seven instances in which Leptospira was detected, but enterococci levels – the standard FIB for marine and estuarine environments – mostly remained below standard threshold values for impairment (USEPA 2012). However, these Leptospira detections were concurrent with elevated E. coli concentrations, suggesting that actual health hazards may be underestimated if only enterococci are used as a water quality indicator. These results illuminate a potential flaw in the use of a single FIB as an indicator of water quality found elsewhere with Leptospira (Truitt et al. 2020) and highlight the potential for qPCR to provide a more comprehensive approach to accurately reflect the microbial risk in recreational aquatic environments.

Our findings also support the concept of the ‘One Health’ approach, an all-encompassing health strategy that emphasizes the interdependence of human health, environmental health, and animal health. This approach, which is endorsed by prominent global organizations including the United Nations and the World Health Organization, places significant emphasis on the need for multidisciplinary and multisectoral collaboration to effectively tackle intricate health issues that arise from the interplay between humans, faunae, and the environment (Laing et al. 2020; Hillier et al. 2021).

The complexity of public health issues often necessitates a coordinated response from local, state, and nongovernmental organizations (Maibach et al. 2008; Khan et al. 2018), in which the active participation of regulatory bodies in conceiving and implementing mitigation strategies is vital (Kohl et al. 2013). Local health departments and the CDC can improve early detection and patient management through enhanced monitoring programs and health education campaigns (Rose et al. 2001; Ventola 2015). Efforts in which local advocacy groups collaborate can enhance the resilience of policies governing resource management and provide a more robust defense against threats to community wellbeing (Deitch et al. 2021). Utilizing these coordinated efforts can substantially advance strategies for preserving recreational water quality and public health.

In the current landscape of molecular microbial ecology, using qPCR in water quality evaluation is a significant strategy (Bonadonna et al. 2019; Tiwari et al. 2022). Its capability to detect a wide-ranging spectrum of waterborne pathogens, especially those that pose significant threats to both human and animal communities in coastal ecosystems, represents a significant advance in the fields of environmental surveillance and public health protection (King et al. 2007; Morrison et al. 2008; Lama & Bachoon 2018; Bridgemohan et al. 2023). Numerous studies investigating pathogen detection have demonstrated that qPCR is a precise and effective instrument (Holman et al. 2014; Bridgemohan et al. 2022): It provides a high level of sensitivity and specificity for the identification and quantification of target pathogens, offering a depth to water quality assessments that is beneficial but not widely performed (Bustin & Nolan 2004; Siqueira & Rôças 2017).

qPCR has emerged as a powerful instrument for public health protection (Hameed et al. 2018), with its proficiency in the accurate detection and quantification of waterborne pathogens beneficial to the formulation of effective strategies to reduce disease transmission and mortality (Rainbow et al. 2020). This application of qPCR in PBS exemplifies the utility of molecular microbiological techniques in modern health prevention and risk detection strategies, providing a compelling illustration of the efficacy of qPCR in analyzing water quality. Specifically, we used qPCR to detect Leptospira, adding to the growing body of research supporting its application. However, as with any scientific technique, qPCR has its limitations. One such disadvantage is the associated cost, which may prevent its widespread adoption in contexts with limited resources. Nonetheless, the extraordinary sensitivity and specificity of qPCR demonstrate its value, making a compelling case for its increased use in water quality assessment. As progress is made in the field of molecular microbial ecology, it becomes imperative to consider strategies that can counterbalance the cost factor, ensuring a broader and more equitable application of this potent public health protection tool.

Our year-long study in the PBS and its adjacent watershed revealed the spatiotemporal distribution and prevalence of Leptospira and its relationship with FIB E. coli and Enterococci in various aquatic ecosystems, upstream estuarine areas, and coastal recreational areas. Our study highlights the health risks associated with these bacteria and calls for extensive public health interventions and sustainable strategies for the aquatic ecosystem management. The research has global applicability, highlighting the value of novel techniques such as qPCR in informing health policies and interventions in resource-constrained contexts. Our results revealed the presence (n = 12) of Leptospira and co-occurrence with elevated levels of FIB such as E. coli, even in developed nations such as the United States, highlighting the need for more rigorous monitoring and containment, particularly in coastal recreational areas. The prevalence of these organisms has far-reaching consequences on human and animal health and socioeconomic dynamics. We encourage global interdisciplinary collaboration among scientists, health practitioners, and policymakers to address the global initiative against leptospirosis and acknowledge the situation's extent beyond traditional geographic boundaries.

We thank the funding agencies at the US Environmental Protection Agency and US Department of Agriculture (Grants MX-00D86419 and FLA-WFC-005577, respectively) and the facilities of the UF IFAS West Florida Research and Education Center and the microbiology laboratory at Georgia College, finally, to the watershed management laboratory team including Emily Harmon, Patrick Garrison, and Caitlin Turnbull, at UF IFAS WFREC for their field and laboratory assistance.

Support of this research was funded by the United States Environmental Protection Agency Grant number MX-00D86419 and the United States Department of Agriculture Hatch Grant number FLA-WFC-005577.

Both writers and scientists contributed to the development of the research concept, the process of writing it up, and the design. The individuals involved in this study, namely, Ronell Bridgemohan, Matthew Deitch, Emily Harmon, Matt Whiles, P. Christopher Wilson, Eban Bean, Dave Bachoon, Aaden Redhead, and Jodell Nicholas, were responsible for conducting material preparation, sampling, data collection, and analysis. Ronell Shamir is a prominent figure in the field of computer science. Hemsley Bridgemohan, Matthew Jaeger Deitch, and Emily Harmon conducted a series of scientific procedures including sampling, water quality testing, enumeration of FIB, membrane filtering, and nutrient testing. DNA extraction and polymerase chain reaction were conducted by Dave Bachoon, Aaden Redhead, and Jodell Nicholas.

The individuals who contributed to this study by assisting in data analysis, manuscript write-up, and statistical analysis were Matt Whiles, Eban Bean, and Patrick Christopher Wilson. The research manuscript was initially authored by Ronell Bridgemohan, and subsequent iterations of the document were subject to scrutiny by each researcher/author involved. All individuals have thoroughly reviewed and provided their consent for the ultimate iteration. The data gathered and examined in the course of this study, as well as the research conclusions, may be acquired by contacting the respective author upon inquiry.

All authors have read, understood, and have complied as applicable with the statement on ‘Ethical responsibilities of Authors’ as found in the Instructions for Authors and are aware that with minor exceptions, no changes can be made to authorship once the paper is submitted.’ This is an observational study. The Research Ethics Committee has confirmed that no ethical approval is needed.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.