The progression of Lyme disease in the state of Virginia

Lyme disease has spread in the United States from the northeast to more southern and western parts of the country. It has been shown that the tick which most often carries the Lyme disease pathogen ( ​Borrelia burgdorferi​) in North America, ​Ixodes scapularis ​,​ ​has also spread to these new areas of concern. In Virginia, the disease has progressed from the eastern shore to the western, more mountainous regions. Through obtaining historical museum specimens of mice from 11 locations in Virginia, we seek to determine whether the Lyme disease causing pathogen, B. burgdorferi​, has always been present in the western regions of Virginia, or if the pathogen was brought over by the ​I. scapularis ​ ticks themselves. Thus far, one specimen has tested positive for ​B. burgdorferi​ from Accomack county, located in southeastern Virginia; however no evidence has been found to indicate that the pathogen was present in western sites where cases have increased recently. Introduction: Lyme disease was first diagnosed as a unique condition in 1975 in Lyme, Connecticut, after previously being confused with rheumatoid arthritis (Steere, Et. al, 2004). The disease was discovered when an abnormally large group of children presented with what doctors thought was juvenile rheumatoid arthritis in the fall of 1975 (Elbaum-Garfinkle 2011). In addition to the spatial clustering of cases, many of the children reported tick bites, leading to informing the public of the risk to a novel tick-associated disease. A spirochete was discovered to be present in humans reporting symptoms, and in mice and ticks in the area. By 1982, William Burgdorfer and colleagues isolated the spirochete pathogen, ​Borrelia burgdorferi, ​ from an Ixodes tick and found antibodies for the pathogen in the serum of patients with Lyme disease (Burgdorfer et al. 1982). These discoveries gave insight into the transmission cycle of Lyme disease between small mammals, ticks, and humans. Now, Lyme disease is the most common vector borne disease in the United States (Shapiro, 2014). In the United States, Lyme disease cases have increased from 9,908 cases in 1992 to 19,931 cases in 2006, a 101% annual increase (Bacon et al. 2008). The geographic location of Lyme disease has been expanding since these events unfolded in the 80’s, as highlighted in a surveillance study by Kugeler and colleagues (2015). In the first five-year period that was studied, 1993-1997, 69 counties in the United States were identified as having high Lyme disease incidence, meaning that they were “within a defined, statistically significant high-risk spatial cluster” (Kugeler et al. 2015). Included in these counties, were four outliers in the southeastern United States. In the next five-year period, high incidence increased to 130 counties, and then to 260 counties during the third and last period. Cases were originally concentrated in a small cluster of states in the Northeastern and Midwestern United States such as Connecticut, New York, Massachusetts, New Jersey, Rhode Island, Wisconsin, and Minnesota. However, Lyme disease has since been diagnosed in every state except for Hawaii in between the years 2005-2015 (CDC). Researcher Tammi Johnson examined national parks in the United States for the presence of ticks and the Lyme disease causing pathogen (Johnson et al. 2016). This study shows that the Lyme disease causing pathogen, ​B. burgdorferi, ​ was present in all areas where ​Ixodes scapularis ​ticks were present, and furthermore that there were newly established populations of ​I. scapularis ​ticks in parks in Washington, D.C. and in Greene County, VA. With the expansion of Lyme disease, there is also spatial expansion of the ​Ixodes scapularis ​tick, also known as the black-legged tick or the deer tick. The ​Ixodes scapularis ​tick is the most common vector for the Lyme disease causing pathogen in the United States. This tick is the bridge between the host animals in the wild that harbor the pathogen and humans. Therefore, with expansion of the tick, the disease can emerge where it was not known to be before, due to increasing incidence and reports in humans. Since 1998, the number of counties in the United States where ​Ixodes scapularis ​ticks are established or reported has more than doubled to make for a total of 842 counties across 35 states (Eisen et al. 2016). The directionality of this expansion is northward into upstate New York, New Hampshire, Maine, etc., westward through Pennsylvania, eastern into Ohio and eastern New York, as well as south and southwestward into West Virginia, Virginia, and North Carolina (Eisen et al. 2016). The ​Ixodes scapularis ​tick has four life cycle stages: the egg, larvae, nymph, and adult. Throughout each of its mobile stages (larvae, nymph, and adult) it feeds on different hosts, making it a three-host tick (Paskewitz n.d.). The eggs are usually deposited in late spring and then hatch into larvae in the summer. At this stage, they can feed on a variety of mammals during which they can pick up the pathogen that causes Lyme disease. During the next spring, the larvae transition to the nymphal stage when they are most likely to pick up pathogens and transmit to humans. Temperature and day-length are predictors of host-seeking behavior in ticks, thus allowing for predictability of infection during certain times of the year. However, studies in Europe have shown the ability for ticks to adapt to climate change, increasing the overall risk of disease (Gilbert et al. 2014). In the fall, the nymphs transition to adults, making them larger and therefore more easily detectable on humans or domestic animals. The expression of genes of the Lyme disease causing pathogen, ​Borrelia burgdorferi​, have been shown to change during the different life cycle stages of the tick (Gilmore et al. 2001). Researcher Joe Piesman and colleagues examined ​B. burgdorferi ​transmission and its relation to the amount of time a tick is attached to its host (Piesman et al. 1986). They found that pathogen transmission increased from 7.1% for a 24 hour exposure period to 35.7% for a 48 hour exposure period, and finally to 92.9% for a 72 hour exposure period. Although most cases of Lyme disease occur between northern Virginia and New England, recent evidence has shown that Lyme disease has intensified in the state of Virginia, and has spread southward (Lantos et al. 2015). In relation to the state of Virginia specifically, the relative abundance of ticks was said to be “greater near the Atlantic Ocean than further inland” (Casteel and Sonenshine 1996). This was discovered by tick sampling using denim cloth tick flags attached to a wooden dowel along randomly selected transects at the three different locations. In 2013, the incidence of Lyme disease in Virginia was 11.2 cases per 100,000 people, compared to the incidence in North Carolina at 0.4 cases per 100,000 people (Lantos et al. 2015). Virginia was first reported to have Lyme disease in 1982, when three cases were reported (Stroube and Miller 1992). Publicity of the disease increased in 1987, allowing for more awareness and yielding a total of 27 cases in VA (Stroube and Miller 1992). Each year the number of cases has increased. According to the Center for Disease Control, Lyme disease cases in Virginia have increased from 274 cases in 2005 to 1102 confirmed cases in 2015, with an additional 437 possible cases (Adams et al. 2014). Lyme disease has spread to the more western and mountainous regions of Virginia since 2007. A previous study showed that tick populations were rarer in highly elevated areas and were also slower to develop in these areas (Leighton et al. 2012). The geographic expansion of Lyme disease is correlated with the expansion of the ​Ixodes scapularis ​tick to more western parts of Virginia. The number of counties in Virginia with ​Ixodes scapularis ​ticks has increased from 4 to 29 reported counties and from 8 to 43 established counties from 1995 to 2015 (Eisen et al. 2016). The progression of ​I. scapularis ​ ticks to more southern and western regions of Virginia is exhibited in this expansion, and it has been found that, in contrast to previous findings, the highest densities of ticks are in the higher elevated mountainous regions (Brinkerhoff et al. 2014). Virginia is not the only state where expansion has been observed, revealed in the documentation by researcher Sarah Hamer in Michigan (Hamer et al. 2011). Hamer and colleagues explored Lyme disease expansion in Michigan, and found that there are cryptic, or hidden, species that can harbor the ​Borrelia burgdorferi ​pathogen, and undetectably so, because the bridge vector, ​Ixodes scapularis ​was not present. The transfer of the pathogen from mammal to mammal was accomplished by a different tick species, one that does not commonly bite humans or domestic animals. Therefore, Hamer found that ​Borrelia burgdorferi ​was able to persist in the population through a cryptic bird-rabbit-tick cycle, remaining undetected until the ​Ixodes scapularis ​ticks arrived and served as the vector to humans and domestic animals. Since ​Ixodes scapularis ​is a general feeder, when it invades an area it may feed on any small mammal, including rabbits or wild birds, and pick up the pathogen. It then acts as the bridge vector between these small mammals and humans in the area, and is responsible for the spread of Lyme disease. Sampling from vertebrates in locations where ​I. scapularis ​is absent can indicate whether ​B. burgdorferi​ is maintained in a cryptic cycle of transmission using alternate and as-yet unidentified vector species. The white footed mouse, or ​Peromyscus leucopus ​, which was the species from which samples were taken in this study, has been called the “most competent disease reservoir” (LoGiudice et al 2002). Studies have shown that the dilution effect, or the presence of high species diversity in the host’s community, reduces possible vector infection because it dilutes the community (LoGiudice et al.). LoGiudice and colleagues found that squirrels had the highest dilution potential, “reducing infection prevalence by ~58%.” A study in Connecticut found that out of blood serum samples from 514 mice, 75% tested positive for ​Borrelia burgdorferi​ (Bunikis et al. 2004). Compared to other small rodents, the white footed mouse was found to be the most competent vector for B. burgdorferi with 90% infection rate, as compared to 75% for chipmunks and 5.5% for meadow voles (Mather et al. 1989). Borrelia burgdorferi has a variety of outer surface proteins that are expressed at different stages of both vertebrate-to-tick and tick-to-vertebrate transmission, including the lipoprotein outer surface protein C (ospC). This protein is upregulated when the ​Borrelia ​ leave the gut of the tick and migrate to the salivary glands, so that they can escape and possibly infect a host (Pal et al. 2004). Bunikis et al. (2004) found that ospC was the most recognized antigen to ​B. burgdorferi present in the ​Peromyscus leucopus ​mouse. My goal is to assess the historical distribution of ​B. burgdorferi ​in Virginia ​P. leucopus ​by using preserved museum specimens. I will test the hypothesis that the ​B. burgdorferi ​pathogen has always been present in the western parts of Virginia through a cryptic cycle rather than being introduced by the spatial expansion of the ​Ixodes scapularis ​ticks using historical mice specimens. The use of preserved museum mice specimens presents difficulties for both DNA extraction and PCR, as the preserved specimens could be degraded or also contain PCR inhibitors in the preservation methods. The specimens were preserved in a dry manner, however the specific chemicals used were unknown. Previous research on museum specimens has pointed to the use of ethanol to clean the samples before extraction (Paabo 1989). Wandeler et al. described that preserved museum specimens can be expected to be degraded and therefore diluted, implying that PCR for preserved specimens is usually restricted to products smaller than 200 base pairs (2007). However, museum samples have been shown to be of use in genetic analyses to help determine phylogenies or taxonomies (Freeland et al. 2007, Su et al. 1999). By testing mouse tissues collected from different areas within Virginia prior to the increase in Lyme disease incidence, I can document the historical presence or absence of enzootic transmission. I obtained ​P. leucopus ​samples from 11 different counties in Virginia: Accomack, Augusta, Botetourt, Giles, Montgomery, Northampton, Patrick, Princess Anne, Rockbridge, Rockingham, and Smyth. The dates in which the mice were acquired range from 1915 to 1991. The range of dates and locations of these mice, all before the peak of Lyme disease in Virginia, will allow us to determine whether ​Borrelia burgdorferi ​was always present in the western regions of Virginia cryptically and whether the ​Ixodes scapularis ​ ticks are simply a zoonotic bridge between the host animals and humans, or whether the pathogen was brought to the west by the spreading ​Ixodes scapularis ​ticks themselves. Methods: Preserved white footed mice samples were obtained from the Virginia Museum of Natural History in Martinsville, VA. The mice were preserved with unknown chemicals, in a dry state, therefore potentially rendering the DNA degraded or non amplifiable due to the presence of a PCR inhibitor. A total of 126 samples were acquired. The tissue samples were taken from ears of the white footed mouse, ​Peromyscus leucopus ​and range in age from years 1915-1991. A map showing the locations of obtained samples can be found in Figure 1. Historical Mice Specimen DNA Amplification: The museum specimens have been preserved using chemicals that could have potentially denatured or fragmented the DNA. If this were the case, amplification via PCR would be difficult and maybe impossible. To confirm presence of DNA, a nanodrop was performed for samples 001-010 and 107-116, as well as for three positive controls (201H, 119-01, and 199-02). We attempted to amplify the Cytochrome B gene which codes for a protein that is found in the mitochondria of eukaryotic cells (Howell). First, a PCR was done with a DNA gradient, to determine the optimal amount of DNA (uL) for the PCR reaction. Extracted DNA amounts of 2 uL, 5 uL, and 10 uL were used with known positive control DNA in a mix, to confirm that preserved specimen DNA was not inhibiting PCR reaction. This PCR was successful. To amplify just the museum specimen DNA, primers were designed for Cytochrome Oxidase and for a smaller PCR product (~117 bp) due to possible degradation in the tissues. These and subsequent primers were designed using NCBI primer blast function. The primer blast provides melt temperatures and so annealing temperatures were designed to be 3-5 degrees C less. To determine an optimal annealing temperature for the reaction, a PCR temperature gradient was performed from 46.9-53.2 degrees Celsius with a known positive control. Another temperature gradient PCR was performed from 40.2-46.7 degrees Celsius. Using the annealing temperature that produced the strongest bands from the previous PCR (46 degrees Celsius), a PCR was performed with five of the museum specimens and a positive and negative control. Borrelia Confirmation: Primers for the ospC (outer surface protein C) gene were designed to assess the presence of Borrelia burgdorferi​ in the samples. Primers were designed with preference for highly conserved sequences for this gene in ​B. burgdorferi ​that was not longer than 200 base pairs. A gradient of positive control dilutions was used to test all three of the new primer sets. Another PCR was performed on the previous PCR products to yield higher amplification. Primer set 2 was then used on positive controls for a nested PCR and this was repeated a second time, however this reaction was unsuccessful both times. Therefore, new primers were designed for a nested PCR, which included both an outer set and an inner set, to increase specificity. An MgCl ​ 2​ gradient was performed on the nested PCR to determine optimal amount of the compound. qPCR was then used with these primers to target smaller sequences and to visually compare to positive controls. qPCR mixture included the same mixture for regular PCR, but with qPCR specific taq: 0.4 uL MgCl ​ 2​, 10 uL taq, 2 uL Fin primer (Bunikis et al 2004), 2 uL R8 primer, 4 uL H ​ 2​O, and 2 uL DNA for each museum specimen. qPCR results were compared to positive and negative controls in terms of Cq value and visually, in order to asses the relative amount of amplification, in the qPCR program. After qPCR, a regular PCR was performed to visualize the bands and then the possible positives from that were extracted from the gel and sent out for sequencing. Primer table is provided in Table 4. Nested primers were designed to target the Intergenic Spacer (IGS) between the 16s and 23s ribosomal sequences, which is a commonly used genetic target to detect and characterize ​B. burgdorferi​. Any potential positives that were identified via the IGS nested PCR and gel electrophoresis were extracted using PCR clean-up protocol and sent to Operon Eurofins lab Louisville, KY for sequencing. To date, DNA from samples 001-078 and 107-126 have been extracted and samples 001-019 and 107-126 have been run via qPCR targeting the ospC gene. Follow up IGS nested PCRs have been run on samples 001-026. To continue, the rest of the samples’ DNA will be extracted, run with the qPCR protocol, with the IGS nested PCR, and possibly sequenced. Additional PCRs may be necessary if qPCR indicates a potential positive, but IGS PCR is unsuccessful due to possible degradation. RESULTS: Historical Mice Specimen DNA Amplification: To confirm that DNA was present in adequate amounts in the preserved museum specimens, a nanodrop was performed for samples 001-010 and 107-116, as well as for three positive controls: 201H, 119-01, and 199-02 (Table 1). The results show that the samples have detectable DNA, with an average concentration of 14.19 ng/uL, whereas the 3 positive controls have an average concentration of 18.73 ng/uL. To determine an optimal annealing temperature for the Cytochrome oxidase primers, temperature gradient PCRs were performed. The first temperature gradient that was performed from 46.9-53.2 degrees celsius revealed that the most optimal annealing temperatures were around the lower limit of our temperature gradient (Figure 2). Therefore, another temperature gradient was performed from 40.2-50.0 degrees celsius (Figure 3). The bands that were produced for annealing temperatures 45.4 and 46.7 degrees celsius appeared equally strong. 46 degrees C was chosen as the most optimal annealing temperature. To ensure efficiency of these primers, a PCR was performed using 46 degrees C as the annealing temperature, with five of the museum specimens and a positive control (201H) (Figure 4). The bands appeared at the expected length of 117 base pairs. This confirmed that there was amplifiable DNA in the museum specimens. Confirmation of ​Borrelia burgdorferi ​in historical specimens: In order to confirm the presence of ​Borrelia burgdorferi ​in the museum specimens, 3 different primer sets were designed to target ospC. All were tested via PCR with positive control dilutions (10 ​-1​, 10​-2​, 10​-3​, 10​-4​) (Figure 5). Primer set 2 appeared to be the most promising, with bands around 300 base pairs, as expected, for three dilutions. Primer set 3 also showed a band around 300 base pairs, but only for the most concentrated DNA dilution. A pseudo-nested PCR (repeated PCR of the same primers, using template from first round) was performed on the previous PCR products to increase amplification for the less concentrated dilutions (Figure 6). There was possible primer dimer for primer set 2, with the lower bands still appearing around 300 base pairs as expected, but other bands appearing higher at a smaller base pair location on the gel (less than 100 bp). The bands produced from primer set 3 appeared slightly smaller than the expected product length, and bands also appeared in two of the negative control lanes (Figure 6). A PCR was then ran for primer set 2 with known positive mouse DNA. This PCR failed two times, as no bands appeared on either gel. New nested PCR primers were then designed with an expected product length of 150 base pairs. Since regular PCR protocol includes MgCl ​ 2​ in the reaction, a MgCl ​2​ gradient was performed for the nested PCR (Table 2). All 16 combinations were run on a gel (Figure 7). qPCR was also used with these primers to quickly visualize the museum specimens in comparison to positive control dilutions, and because qPCR has the ability to target smaller chunks of DNA (Figure 8). This qPCR indicated that a positive ospC mouse specimen would appear between 10 ​-3​ and 10 ​-4​ DNA concentrations, as indicated by our known positive mouse specimen. This first qPCR indicated that the five preserved museum specimens tested were negative for ospC, as their amplification curves appeared around the negative controls. A qPCR, done on mice specimen 001-010, positive controls (2), and negative controls both with (2) and without (2) primers showed that a positive control Cq value can be expected to be around 25.04 (Figure 8, Table 3). The negative controls with primers had an average Cq value of 31.97. A lower Cq value is interpreted as a higher concentration of the targeted sequence of DNA. The mice specimens with Cq values that appeared relatively lower and whose curves appeared visually closer to the positive controls versus the negative controls were noted to be potential positives. qPCR Cq values can be found in Table 3. For samples that appeared relatively positive in that they were closer to positive control Cq values versus negative were tested with the ospC nested PCR (Bunikis et al. 2004) and then a gel was run (Figure 9). The bands on the gel appeared around the expected base pair length, however positive control dilution 10 ​-4​ appeared smeared with its brightest point near 1000 base pairs and the negative control lane also produced a band in similar length to the samples and positive control. Therefore, a PCR purification (PCR clean-up protocol) was done on each of the lanes and sent for sequencing. The sequencing was inconclusive and showed possible contamination. Intergenic Spacer (IGS) primers for the sequence between 16s and 23s ribosomal sequences were used for an expected product length of about 1000 base pairs (Bunikis et al. 2004). First, mice specimen 011-026 were tested along with 3 positive control dilutions (10 ​-2​, 10​-3​ , 10​-4​) and one known positive control “46” (Figure 10). Sample 024 appeared similar to all positive controls. Also to note, one of the negative controls appeared similar to the known positive designated “46.” In order to determine whether 024 was a true positive, a nested IGS PCR was run with samples 023, 024, and 025, as well as with the two positive control dilutions (10 ​-3​, 10​-4​), the known positive “46” and two negative controls. The gel from this PCR clearly only showed bands in the positive control lanes and no bands in the sample lanes (Figure 11). From the original IGS gel (Figure 9), sample 014 also appeared around 1000 base pairs. This sample was sent in for sequencing and it was confirmed positive.