Seroprevalence of Leptospira spp. in Blanding's Turtles (Emydoidea blandingii) from DuPage County, Illinois USA

Leptospira spp. are spirochetes that are important pathogens in both domestic animals and wildlife. Because of the ubiquitous nature of these organisms, dogs and cattle are commonly vaccinated against these bacteria to minimize morbidity, mortality, and economic losses. Leptospira interrogans is also considered an important re-emerging zoonotic disease that is thought to be increasing in prevalence as a result of urban sprawl (Ward et al., 2004a,b). A primary reason for the increased reporting of cases is attributed to increased contact between humans, domestic animals, and wildlife at the urban–wildlife interface. Several mammals are known to serve as important reservoirs for certain serovars of Leptospira spp. For example, raccoons (Procyon lotor) are commonly infected with Leptospira kirschneri serovar Grippotyphosa, dogs with L. interrogans serovar Canicola, and cattle with Leptospira borgpetersenii serovar Hardjo (Mitchell et al., 1999; Bolin, 2000). Other important reservoirs for the many pathologic serotypes are not well defined, leaving much of the epidemiology of this pathogen in many of these urban–wildlife interfaces unknown.

Since the 1960s there has been an interest in characterizing the role aquatic chelonians play in the epidemiology of Leptospira spp. (Abdulla and Karstad, 1962; Andrews et al., 1965; Charon et al., 1974; Glosser et al., 1974), although much of the work to date has been experimental in nature. Blanding's turtles (Emydoidea blandingii) experimentally infected with L. interrogans serovar Pomona were found to become leptospiremic and shed leptospires (Abdulla and Karstad, 1962), suggesting that this species could play a role in the transmission of leptospirosis in an aquatic ecosystem. In addition to the findings in Blanding's turtles, Andrews et al. (1965) found a two-thirds higher rate of seroconversion in red eared sliders (Trachemys scripta elegans) following infection compared with a terrestrial species of turtle (Eastern Box Turtle, Terrapene carolina carolina). However, this should not be unexpected as Leptospira spp. have a strong evolutionary association with water.

The purpose of this study was to estimate the seroprevalence of Leptospira spp. in wild Blanding's turtles from DuPage County, Illinois, United States to assess the potential for their exposure to this pathogen and to enrich our understanding of the role these animals might play in the epidemiology of Leptospira spp. The biological hypotheses tested in this study were that: 1) the seroprevalence of Leptospira spp. would be less than ≤10%, 2) adult animals would have a higher seroprevalence than would subadult animals, and 3) there would be no difference in the seroprevalence between males and females or between trapping locations of the turtles.

This research study was performed in accordance to the regulations set forth by the University of Illinois Institutional Animal Care and Use Committee (IACUC protocol no. 11-096).

Study area and site selection: Trapping for this study was conducted in the Forest Preserve District of DuPage County, Illinois, USA. The preserve is comprised of 25,000 acres of protected prairies, woodlands, and wetlands. Animals were trapped at five different locations within the preserve.

Subjects and serum collection: Turtles sampled for this study were part of an ongoing head-start program through the Forest Preserve District of DuPage County. Adult female turtles were fitted with a very high-frequency transmitter that allowed them to be located in the field. Other animals were opportunistically sampled when they were encountered. Turtles were captured by hand and manually restrained for sampling. Each turtle was identified by a passive integrated transponder tag. A physical examination was performed on each animal. Morphometrics collected from each animal included body weight, carapace length and width, and plastron length and width. Sexing was done using external sexual characteristics including tail size. Aging (adult, juvenile) was done based on the size of the animals. Blood was collected from the subcarapacial sinus using a 3-ml syringe and a 22-ga needle. The blood samples were placed in red-top serum separator tubes and centrifuged at 349 × g for 20 min. Serum was removed, placed in a cryovial tube (Becton-Dickinson, Franklin Lakes, NJ), and transported on frozen gel packs to the University of Illinois (Urbana, Illinois) for testing.

Leptospira MAT: Serial, 2-fold dilutions of serum from 1:25–1:800 were evaluated against seven serovars commonly found in Illinois (Leptospira interrogans serovars: Autumnalis, Bratislava, Canicola, Icterohaemorrhagiae, Pomona; Leptospira kirschneri serovar Grippotyphosa; Leptospira borgpetersenii serovar Hardjo). The antigen was prepared from cultures grown in Probumin® media (Millipore, Billeria, MA) and centrifuged at 349 × g for 10 min at room temperature to remove dead bacteria. The supernatant was diluted 1:6 with sterile phosphate-buffered saline (PBS).

Serum samples were centrifuged at 349 × g for 1 min at room temperature to remove any red blood cells and lipids, pipetted into a 96-well flat bottom plate, and diluted 1:12.5 to 1:800 with PBS. Fifty microliters of antigen were added to the 50 μl of diluted sera.

Plates were incubated at room temperature for at least 2 h and examined using dark field microscopy. The end point was determined by the last positive (>10% agglutination) dilution. A titer ≥1:25 was considered positive (Veterinary Diagnostics Laboratory Standard Operating Procedure, University of Illinois, Urbana, IL). The serovar that had the highest titer was considered causative. An animal was labeled as “multiple titer” if it had reactions to multiple serovars at the same titer levels.

Statistical methods: The sample size used for this study was based on the following assumptions: an expected prevalence of <10%, a population size of 40 animals, an alpha = 0.05, and a power of 0.8. The 95% binomial confidence intervals were calculated for appropriate proportions. Fisher exact tests were used to look for associations between serostatus and age, sex, and location. A Mann–Whitney test was used to analyze differences between juvenile and adult titers. A Kruskal–Wallis test was used to determine differences in titers for sex (male, female, unknown) and the five different locations within the forest preserve. Significance was determined when a P-value was ≤0.05. Statistical analyses were performed using SPSS 19.0 (SPSS Inc., Chicago, IL).

A total of 31 turtles were sampled in this study. The majority of the turtles were females (87.0%, 27/31), which was not unexpected because they are being monitored by one of the authors (DT) for the head-start program. The remaining animals were either males (6.5%, 2/31) or too young to sex (6.5%, 2/31). The age cohorts for this population were divided into adults (80.6%, 25/31) or subadults (19.4%, 6/31). Most of the turtles were captured at site 1 (74.1%, 23/31); the remaining turtles were captured at sites 2 (9.7%, 3/31), 3 (6.5%, 2/31), 4 (6.5%, 2/31), and 5 (3.2%, 1/31). The majority of Blanding's turtles in this study were found to be seropositive for Leptospira spp. (93.5%, 29/31; 95% CI: 84.8, 100). The two negative animals were both adults and from two different locations (site 1 and site 2). All of the subadult turtles had serum titers ≤1:50, except one animal with a titer of 1:100. A titer of 1:100 was the most common (41.4%, 12/29), followed by 1:200 (24.1%, 7/29), 1:50 (17.2%, 5/29), 1:25 (13.8%, 4/29), and 1:400 (3.5%, 1/29) (Figure 1). Serovar Icterohemorrhagiae was most-commonly detected (58.6%, 17/29), followed by serovar Grippotyphosa (17.2%, 5/29) and serovar Brataslava (3.5%, 1/29) (Figure 2). Six turtles (20.7%) reacted to two or more serovars at equal titers (Icterohemorrhagiae and Canicola, 1:200, n = 1; Icterohemorrhagiae and Canicola, 1:100, n = 3; Icterohemorrhagiae and Canicola, 1:25, n = 1; Icterohemorrhagiae, Grippotyphosa, and Canicola, 1:25, n = 1). No significant differences in seropositivity were noted by sex (P = 0.97), age (P = 0.67), or location (P = 0.84). No significant differences were noted in titer between age groups (P = 0.43), age (P = 0.23), or location (P = 0.27).

The results of this study confirm that wild Blanding's turtles in the Forest Preserve District of DuPage County, Illinois are exposed to and produce an immunological response to Leptospira spp. This finding, in combination with the experimental work by Abdulla and Karstad (1962), suggests that Blanding's turtles could play an important role in the epidemiology of Leptospira spp. in habitats where the turtle and the pathogen are both found.

Based on previous research with wild aquatic chelonians (Andrews et al., 1965; Silva et al., 2009), and an absence of any historical (necropsy) leptospiral findings in this population of Blanding's turtles, it was expected that the seroprevalence of Leptospira spp. in these turtles would be low (<10%). Instead, this study found that the majority of animals were exposed to the pathogens, including the subadult animals. The 95% CI for this population suggests that the overall prevalence in this population could range between 85% to nearly 100%. Therefore, populations without a history of leptospirosis should not be excluded from consideration of exposure to the pathogen unless specifically tested.

The high seroprevalence in these turtles suggests that the normal activities of this species can lead to a high rate of exposure to Leptospira spp. Blanding's turtle habitat includes marshes, prairie wetlands, wet sedge meadows, and shallow, vegetated areas of lakes. These areas tend to have little to no water flow and present ideal locations for the development and persistence of Leptospira spp. Blanding's turtles are also primarily carnivorous (e.g., snails, insects, crayfish, and small vertebrates) and may be ingesting prey exposed to the bacteria. This species can also travel long distances on land (primarily during breeding season or in search of food) and exposed animals could spread leptospires to other areas. The findings in the current study appear to be echoed in other studies screening aquatic turtles for Leptospira spp. In Andrews et al. (1965), the aquatic red-eared slider was found to have a higher seroprevalence to Leptospira spp. than did the terrestrial Eastern box turtle. The red-eared sliders live in aquatic areas similar to the Blanding's turtles, while box turtles would not be expected to be exposed to the same aquatic environments on a regular basis. Another study found a similar result, with all of the sampled European pond turtles (Emys orbicularis) and a red-eared slider being seropositive (4/4, 100%) while the tortoises tested (Indian star tortoise, Geochelone elegans; spurthighed tortoise, Testudo graeca; Hermann's tortoise, Testudo hermanni; radiated tortoise, Geochelone radiata; and Russian tortoise, Testudo horsfieldii) had a lower seroprevalence (5/52, 9.6%) (Lindtner-Knific et al., 2013). Finally, Silva et al. (2009) tested 40 fresh-water turtles (29 D'Orbigny's slider, Trachemys dorbigni and 11 Hilaire's side-necked turtle, Phrynops hilarii) caught in urban lakes and observed serum titers >1:100 in 11 (27.5%) animals. When combining the results of the current study with the three previous published articles evaluating aquatic turtles, it is apparent that these animals are being exposed to Leptospira spp. in their environment, and likely play a role in the epidemiology of the disease, but future studies to quantify infection or shedding of the organism are needed to further clarify their role.

No significant differences in the seropositivity among the different age groups, sexes, or locations trapped were identified. However, there was a natural sampling bias in this study because the majority of the animals were adult females (87%) and they were from a single location (74.1%). Follow-up studies evaluating a larger sampling of males and subadults from different habitats would be necessary to elucidate whether differences in sex, age, or location are relevant to the seroprevalence of Leptospira spp. in Blanding's turtles. However, even considering these limitations, it is important to note that all of the subadult animals were seropositive, both males were positive, and positive turtles were found at each trapping location. This further supports the likelihood that the general behaviors of this chelonian in this habitat result in an increased risk of exposure to Leptospira spp. Characterizing these specific behaviors could be useful in identifying risk factors that may be used to reduce the likelihood of exposure to the pathogen for not only Blanding's turtles but other species too, including humans.

In this study, serologic testing was used to assess the leptospiral status of the Blanding's turtles. While the results represent an important first step in elucidating the epidemiology of this pathogen in these turtles, and strongly suggest that Blanding's turtles in this study site are being exposed to Leptospira spp., they provide no information about how the pathogen affects these turtles or whether the turtles can serve as reservoirs for the spirochete in their habitat. It is also important to note that, in this study, animals with a titer ≥1:25 were considered seropositive. Previous studies have used higher titer cut-offs to determine seropositivity including ≥1:250 (Charon et al., 1974), ≥1:200 (Glosser et al., 1974), and >1:100 (Silva et al., 2009). The lower cut-off of 1:25 used in this study was determined based on previous work done with wildlife serology in mammals and reptiles at the University of Illinois Diagnostics Laboratory. The MAT is considered to have a high specificity because antibodies against Leptospira spp. rarely cross-react with other bacteria (World Organization for Animal Health, 2008); thus, false positives are less likely and positive results tend to be true positives. However, cross-reaction can occur among serovars, especially at lower titers. To limit the likelihood of misinterpreting results in this study, the highest serovar titer was considered positive. This study found serovar Icterohemorrhagiae was most common, followed by serovar Grippotyphosa and serovar Brataslava. All three serovars have been previously detected in wildlife (Mitchell, 1999; Richardson and Gauthier, 2003; Davis et al., 2008). Stray or owned dogs are another potential source of infection, as canids have also been found to be infected with all three of these serovars (Bolin, 2000; Davis et al., 2008). Serovars are often location-dependent, and a survey of the local wildlife would help further characterize the results seen in this study. Drainage from surrounding areas into the DuPage County Forest Preserve increases the range and type of wildlife that could be affecting the common serovars in this watershed.

As noted previously, one limitation of the MAT test is that it only evaluates exposure. Additional testing using culture, real-time polymerase chain reaction (PCR), or both from antemortem samples collected from the turtles (e.g., cloacal swabs or urine) could be used to further characterize the role of the turtles in disseminating the bacteria into the environment. Also, specific postmortem sampling (culture, PCR, and histopathology) for Leptospira spp. from turtles found dead in the forest preserve (e.g., hit by car, killed by predators) or while in captivity could be useful. Sample collection from the environment, including water samples, substrate, and other vertebrate species, could also be collected to better ascertain the point sources of Leptospira spp. and further refine the epidemiology of this pathogen in this ecosystem.

The population of Blanding's turtles that were sampled in the Forest Preserve District of DuPage County was found to have a high exposure to Leptospira spp. At this time, it is not known how this pathogen may affect this population of endangered animals, as none of the animals were found to be sick nor have any of the dead animals submitted for necropsy from this population been diagnosed with leptospirosis. However, the development of a virulent form of the pathogen in this environment could be catastrophic because the majority of the turtles are being exposed to the bacteria in this system. The high seroprevalence of Leptospira spp. in the turtles is also cause for concern for humans and other animals utilizing this forest preserve. Because the trapping sites within the forest preserve are commonly used for recreation by humans and dogs, it is possible that exposure to Leptospira spp. may occur. Precautions can be taken by individuals using these areas for recreation by not using water from the ponds for cleaning human products intended for cooking or eating and not drinking the water. Also, minimizing the potential contact for pets following these same practices is warranted, along with not allowing them to drink standing water from puddles.

Similarly, it is unknown whether the Blanding's turtles from this population could actively serve as a source of infection to the personnel working with them, as no testing has been done that has found the organism on culture or PCR. However, this work is being planned to further elucidate the potential risks to individuals handling the turtles and to better define the role of the Blanding's turtles in the epidemiology of Leptospira spp. Individuals working with aquatic chelonians should consider the potential for being exposed to this pathogen if the area has been previously identified as positive for Leptospira spp.

Implementing biosecurity protocols, such as wearing gloves and washing hands thoroughly after handling animals, is highly recommended. It is also important to consider the potential risk for exposure to pathogens when bringing these animals into a captive environment. At the facility described in this article, female turtles are captured and held in captivity for the head-start program every spring. These animals are housed near birds and mammals, so there is the potential for cross-contamination between different species. Again, confirming whether the animals are merely exposed to the pathogen or actually infected is required to determine true risk. In cases where disease status is unknown, risk can be minimized by changing gloves and washing hands in-between animals and ensuring that drainage of the housing areas is separate.

The authors wish to thank the DuPage County Forest Preserve staff for their help in collecting the samples and Fluker Farms (Port Allen, Louisiana) for providing funding for this project.