Prevalence of Borrelia burgdorferi in adult female ticks (Ixodes scapularis), Wisconsin 2010-2013

The incidence of Lyme disease in some areas of Wisconsin, U.S.A. has more than tripled since 1991(Hoang-Johnson 2010). Tick abundance and the prevalence of Borrelia burgdorferi in questing ticks are essential components of risk assessment for Lyme disease borreliosis. Published surveys completed overseas in various countries (Germany, Japan, China, Norway) have shown prevalence rates of Borrelia spp. infection ranging from about 30-40% in adult Ixodes spp. (Fu et al. 2013, Murase et al. 2013, Rauter et al. 2002, Soleng and Kjelland 2013). In Wisconsin, unpublished surveys also reveal that as many as 40-50% of Ixodes scapularis adults in some areas may be infected. In recently published studies from the eastern United States, prevalence rates of B. burgdorferi in adult I. scapularis ranged from 27% to 45.2% (Han et al. 2014, Schulze et al. 2013, Hamer et al. 2010). In Wisconsin, the prevalence of B. burgdorferi in I. scapularis nymphs collected from managed red pine forests from 2009 to 2013 was approximately 30% (Lee et al. 2014). To estimate the prevalence of adult questing ticks carrying the Lyme disease bacterium in this pilot study, we randomly collected 341 questing adult female Ixodes scapularis from known and suspected tick habitats throughout regions of Wisconsin during the springs of 2010 through 2013. Nymphal and adult tick activity has been reported to be the highest in May and June (Reye et al. 2010). These locations generally were wooded with a medium-dense tree canopy and contained leaf litter. Sites with tall grasses on the wooded edges with agricultural plains were also chosen. The extent of urbanization was not recorded but in general was minimal. Ticks were collected by walking the sampling area and removing ticks from the collector's clothing or using a flannel drag cloth. The same sites were not necessarily sampled each year as different individuals did the sampling each year in order to maximize the number of sampling locations (average of 24 sites each year). Ticks were collected from 21 counties in primarily west-central Wisconsin (Figure 1). Only adult female ticks (no nymphs or adult males) were assayed in this study. Individual ticks were stored in 70% ethanol before the DNA extraction procedure. Later they were removed from ethanol and allowed to dry before being bisected using a sterile 18-gauge needle. The tick halves were transferred to a sterile microfuge tube, and 200 μl of homogenization solution (0.1M NaCl, 0.2M Tris-HCl pH 8.0, and 0.05M EDTA) was added. Ticks were then homogenized for 90 s with a Kontes Pestle Homogenizer, and SDS (final concentration of 0.5%) and Proteinase K (final concentration 150 μg/ml) were added. Tubes were incubated overnight at 56° C, followed by heating for 10 min at 95° C, and centrifuging at 10,000 g for 10 s. The DNA was extracted with 100 μl of phenol/STE and 100 μl of chloroform, then spun for 90 s at 10,000 g. Two volumes of 95% ethanol were added to the top aqueous layer and placed on ice for 5 min. Tubes were then spun at 12,500 g in a refrigerated microfuge for 15 min. Ethanol was removed with a fine pipette tip, the pellets were dried in a speed-vac, and 50 μl of sterile water was added to each pellet. Samples were stored at -20° C until use. The real-time PCR assay (Failsafe PCR System, Epicentre, Madison, WI) was based on Pietila et al. (2000), utilizing published primer sequences targeting the recA gene of B. burgdorferi. Five ml samples were assayed in a SyBr Green PCR platform (Bio-Rad Mini-Opticon, Hercules, CA) using 2 mmol/liter of primers nTM17.F (5′GTGGATCTATTGTATTAGATGAGGCTCTCG 3′) and nTM17.R (5′ GCCAAAGTTCTGCAACATTAACACCTAAAG 3′) to amplify a 222 bp region of the recA gene. The PCR conditions were as follows: 95° C for 3 min, 95° C for 10 s, 59° C for 5 s, and 72° C for 10 s 45 times. A melting curve analysis from 55° C to 95° C, read every 0.2° C, and held 1 s showed a Tm of 82° C for the 222 bp product. Each run included a 1 ng positive control (purified B. burgdorferi sensu stricto strain B31 DNA, ATCC 35210) and a negative control (minus DNA). All positive controls were mixed after all reagents were added to test samples in a PCR safety hood. The size of the fragments was confirmed on a 3% agarose gel. Samples considered positive after the run had a Tm of ∼82° C and a band at 222 bp that matched the positive control. Bands from several random samples and positive controls meeting the above criteria were Sanger-sequenced (University of Wisconsin Biotechnology Center, Madison, WI). All showed 99% homology to B. burgdorferi recA gene (GenBank Acc. no. U23457.1). No attempt was made to further separate B. burgdorferi into genospecies based on Tm or further sequence analysis. One hundred twenty-two (35.7%) of 341 adult female Ixodes scapularis collected from all regions during the springs of 2010 through 2013 were positive for Borrelia burgdorferi (Table 1, Figure 1). The central region (Chippewa, Clark, Marathon, Eau Claire Counties) had the highest average prevalence: 79/203 (38.9%, 95% confidence limits (C.L) +/-7.9) based on data pooled for the four years. The southwest region and west-central counties had similar average prevalence rates of 13/43 (30.2%, C.L.+/-14.9), and 26/73 (35.6%, C.L.+/-11.7), respectively, during the collecting period. Among individual counties with at least 18 sampled ticks, those counties with the highest average prevalence rates were Chippewa (central) with 12/18 (66.7%, C.L. +/-24.5), Dunn (west-central) with 24/54 (44.4%, C.L.+/-14.2), and Eau Claire (central) with 62/170 (36.5%, C.L.+/-7.5), while the lowest was Vernon county (southwest) with 2/23 (8.7%, C.L.+/-13.7). The yearly prevalence of infected ticks from all regions was determined and graphed as a likelihood distribution function (Figure 2) (Albert 2009, Albert 2014, Hoff 2009). Over the four-year period, the estimated prevalence increased each year from about 21% to about 50% (Figure 2, Table 1). Given the different sample sizes and the associated distribution of the underlying actual prevalences, we cannot conclude that the actual prevalences are increasing monotonically with time. Even so, the estimated prevalence in 2010 and the distributions of prevalence in 2012 and 2013 had very little overlap (Figure 2). This apparent difference in prevalence may be due to an actual increase or to biased sampling in 2010 or there is a lot of variability across years. No effort was made to identify specifics (only the general location) of the habitat where the Borrelia-carrying ticks were collected. Thus, we are not able to correlate specific biotic factors including types of vegetation, acorns, leaf litter, or host animal densities with the frequency of infected ticks. Which factors are most critical in maintenance and transmission of the bacterium remains to be determined. This pilot study adds to the identification of areas in Wisconsin with a high abundance of infected ticks. Continued surveillance and research on identification of biotic and abiotic factors contributing to infected tick abundance will help explain the recent increases in human Lyme disease incidence in Wisconsin. We thank Benjamin Bonis, Kristina Weimer, and Georgeann Vargas for helping develop the PCR assay, Dr. Evan Weiher for statistical analysis, and Infectious Disease Ecology classes for collecting ticks. This study was funded by the Office of Research and Sponsored Programs (ORSP) at the University of Wisconsin-Eau Claire.