First report of Borrelia burgdorferi sensu stricto detection in a commune genospecies in Apodemus agrarius in Gwangju, South Korea

Lyme disease is a tick-borne infectious disease caused by the Borrelia burgdorferi sensu lato complex. However, the distribution of Borrelia genospecies and the tissue detection rate of Borrelia in wild rodents have rarely been investigated. Here, we studied 27 wild rodents (Apodemus agrarius) captured in October and November 2016 in Gwangju, South Korea, and performed nested polymerase chain reaction targeting pyrG and ospA to confirm Borrelia infection. Eight rodents (29.6%) tested positive for Borrelia infection. The heart showed the highest infection rate (7/27; 25.9%), followed by the spleen (4/27; 14.8%), kidney (2/27; 7.4%), and lungs (1/27; 3.7%). The B. afzelii infection rate was 25.9%, with the highest rate observed in the heart (7/27; 25.9%), followed by that in the kidney and spleen (both 2/27; 7.4%). B. garinii and B. burgdorferi sensu stricto were detected only in the spleen (1/27; 3.7%). This is the first report of B. burgdorferi sensu stricto infection in wild rodents in South Korea. The rodent hearts showed a high B. afzelii infection rate, whereas the rodent spleens showed high B. garinii and B. burgdorferi sensu stricto infection rates. Besides B. garinii and B. afzelii, B. burgdorferi sensu stricto may cause Lyme disease in South Korea.

www.nature.com/scientificreports/ B. afzelii strains, were identified using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) of the ospC gene and rrf (5S)-rrl (23S) intergenic space 12 . Moreover, in 2020, the prevalence and distribution of five B. burgdorferi sensu lato genospecies (B. afzelii, B. valaisiana, B. yangtzensis, B. garinii, and B. tanukii) were reported based on the results of nested PCR targeting partial flagellin B gene sequences and sequencing in ticks isolated from wild rodents in South Korea 13 . Lyme disease is a Group 3 infectious disease in South Korea, and the number of cases reported to the Korea Disease Control and Prevention Agency has gradually increased from 2011 to 2020, with an average of 15.4 cases reported per year, and 31, 23, 23, and 12 cases reported each year from 2017 to 2020, respectively 14 .
The diagnosis of Lyme disease is confirmed based on positive results in an indirect immunofluorescence assay or enzyme-linked immunosorbent assay via western blot analysis or by isolating and identifying the pathogen from clinical specimens from patients, including blood samples. In addition, the molecular techniques used for classifying and identifying Borrelia spp. and B. burgdorferi include PCR targeting rRNA genes, flaB, recA, p66, and the plasmid-encoded gene ospA; DNA-DNA homology analysis, ribotyping, PCR-RFLP analysis, pulsed-field gel electrophoresis, randomly amplified polymorphic DNA fingerprinting, multilocus sequence typing/multilocus sequence analysis, and whole genome sequencing 2,[15][16][17][18][19][20][21] . The process of culturing clinical specimens to detect B. burgdorferi is labor-intensive, expensive, and applicable only to untreated patients, and therefore, is not used in clinical practice. However, microorganisms can be directly detected in clinical specimens using PCR, and their genotype can be confirmed through sequencing without isolating the pathogens 17 . Nested PCR is known to exhibit a sensitivity 100 times greater than that of conventional PCR; hence, nested PCR can be used to increase the diagnostic sensitivity for Lyme disease 17,22 . Nested PCR targeting the rrf (5S)-rrl (23S) intergenic spacer and ospA (encoding the outer surface protein A) gene or the 16S rRNA and pyrG (encoding CTP synthase) genes was performed, along with sequence analysis, to detect Borrelia DNA in clinical samples 17,20 .
In this study, we investigated the infection rate in 27 wild rodents (A. agrarius) captured in October and November 2016 using nested PCR targeting the Borrelia-specific genes pyrG and ospA and direct DNA sequencing with rodent tissue samples. The distribution and infection rate of Borrelia genospecies in wild rodents, which serve as reservoirs for the pathogens of tick-borne infectious disease, have rarely been investigated. In addition, we reported the rates of Borrelia infection in the different organs of the wild rodents and investigated the differences in the organ-specific detection rate of each Borrelia species.

Results
PCR and tissue detection rates of Borrelia species in captured wild rodents. Twenty-seven wild rodents were captured using Sherman live traps during October and November 2016 in two regions of Gwangju City in South Korea. All captured rodents were identified as A. agrarius. Borrelia-specific pyrG and ospA nested PCR revealed that 8 of the 27 rodents were infected with the pathogens in the spleen, kidney, lungs, and heart (Table 1).
In pyrG nested PCR, the overall rate of positive response to Borrelia species was 29.6% (8/27). Among the studied organs, the detection rate was the highest in the heart (25.9%, 7/27). The kidney and spleen showed positive detection rates of 7.4% (2/27) and 14.8% (4/27), respectively. The B. afzelii infection rate was 25.9% (7/27) in the wild rodents. The heart showed the highest positive detection rate (25.9%, 7/27), and the kidney and spleen had a positive detection rate of 7.4% (2/27), respectively. The infection rate for both B. garinii and B. burgdorferi sensu stricto was 3.7% (1/27), and both bacteria were detected only in the spleens.
In ospA nested PCR, the heart tissues from 6 of 27 wild rodents exhibited a positive response to B. afzelii, with an infection rate of 22.2%. The infection rate was 3.7% (1/27) in the spleen, kidney, and lung tissues. In a wild rodent that showed a Borrelia-positive result in pyrG nested PCR (Chosun M10-8 Sp, which indicates the spleen of wild rodent no. 8 captured in October 2016), B. burgdorferi sensu stricto was also detected by ospA nested PCR in the spleen.
Using pyrG nested PCR, the B. afzelii infection rate was found to be 25% (3/12) in October and 26.7% (4/15) in November, and the hearts of the animals showed the highest infection rate in both months. With respect to B. garinii, a positive response was detected in the spleen of only one animal captured in October (Chosun M10-11 Sp). This animal exhibited co-infection with B. garinii (detected in the spleen) and B. afzelii (detected in the heart, kidney, and lungs). Additionally, infection by B. burgdorferi sensu stricto in the spleen was confirmed in only one wild rodent captured in October (Chosun M10-8 Sp).

Discussion
Lyme disease is a zoonotic disease transmitted by ticks and caused by the B. burgdorferi sensu lato complex, a group comprising approximately 20 species. B. burgdorferi sensu stricto, B. afzelii, B. garinii, B. valaisiana, and B. lusitaniae have been reported to cause the disease 23 .
In this study, we performed nested PCR targeting the pyrG and ospA genes of Borrelia species and detected infection by Borrelia genospecies, including B. afzelii, B. garinii, and B. burgdorferi sensu stricto, in A. agrarius, with a 29.6% (8/27) positive detection rate for Borrelia species.
In a study conducted in 2008, conventional PCR targeting ospC, a gene specific to Borrelia, was performed using genomic DNA extracted from 1618 ticks (420 pools) and 369 rodents (A. agrarius) captured close to the demilitarized zone of Gyeonggi Province, South Korea. Contrary to our results, the positive rate for B. burgdorferi sensu lato infection was found to be 1% (16/420) in ticks; however, the Borrelia infection rate in rodents was not reported 24 . We attempted to retrieve additional published reports on the detection rate of Borrelia genospecies in wild rodents, including A. agrarius, in South Korea; however, we did not find such reports. Recently, the B. afzelii detection rate was reported in ticks parasitizing domestic and wild animals in South Korea, including     23 .
In the present study, a high positive detection rate for B. afzelii (25.9%) was found in specimens obtained from captured wild rodents, and the highest rate was observed in the heart tissues. While B. garinii and B. burgdorferi sensu stricto infection was also detected (3.7%), these bacteria only infected the spleen. In 2020, Cadavid et al. reported that when immunosuppressed adult Macaca mulatta were inoculated with B. burgdorferi, B. burgdorferi exhibited tropism for the meninges in the central nervous system and for connective tissues. Additionally, significant inflammation was noted only in the heart, and immunosuppressed animals inoculated with B. burgdorferi exhibited cardiac fiber degeneration and necrosis 27 . In 2017, Grillon et al., reported that B. burgdorferi sensu stricto and B. afzelii target the skin of mice regardless of the route of inoculation and cause persistent skin infection 28 . These results differ from ours, probably because we detected Borrelia spirochetes from each organ of the captured wild rodents, whereas Grillon et al. detected Borrelia from each organ after inoculating a susceptible animal.
Lyme disease spirochetes exhibit strain-and species-specific differences in tissue tropism. For example, infection by B. burgdorferi sensu stricto, B. garinii, and B. afzelii, the three major spirochetes causing Lyme disease, is characterized by distinct but overlapping clinical signs. Infection by B. burgdorferi sensu stricto, the most common causative agent of Lyme disease in the USA, is closely associated with arthritis, whereas that by B. garinii is related to neuroborreliosis, and that by B. afzelii is related to acrodermatitis (a type of chronic skin lesion). B. burgdorferi and B. garinii isolates were shown to cause severe arthritis in immunocompromised mice in animal studies 29,30 . The correlation between pancarditis and the marked tropism of B. burgdorferi in cardiac tissues was also reported in studies involving the autopsy of patients with sudden cardiac deaths associated with Lyme carditis 31 .
Lyme borreliosis spirochetes exhibit a high detection rate in a specific organ depending on the species, which could suggest a certain preference for the organ. These findings suggest that Lyme borreliosis spirochetes infecting rodents can be detected in the heart and spleen tissues, as B. afzelii exhibit a high detection rate in rodent hearts, whereas B. garinii and B. burgdorferi exhibit high detection rates in rodent spleens. According to Matuschka et al., infected Norway rats that served as reservoirs for Lyme disease spirochetes increased the infection risk for visitors to a city park in central Europe 32 . Based on the relatively high rate of Borrelia infection (29.6%) in the captured rodents in this study, the risk of Lyme disease in the Gwangju city, South Korea is predicted to be high. Therefore, additional research should be conducted to study the causative agents of Lyme disease in South Korea and further elucidate their prevalence and tissue tropism.
In conclusion, this is the first study to show the presence of B. burgdorferi sensu stricto in rodents captured in South Korea. B. afzelii, one of the causative agents of Lyme disease, exhibited a high positive detection rate (25.9%) in wild rodents, specifically in the heart tissues, captured in the areas around a metropolitan city in the southwestern region of South Korea, whereas B. garinii and B. burgdorferi exhibited high detection rates in the spleen.
Our findings suggest that along with B. garinii and B. afzelii, B. burgdorferi sensu stricto may also act as a causative agent of Lyme disease in South Korea, and different Borrelia species exhibit different tissue detection rates.  33 . The two regions in which the mouse traps were placed are located on a boundary that divides the urban and rural areas, and there were five types of locations (fallow ground, a ridge between rice fields, a boundary between a forest and field, area surrounding tombs, and area surrounding water) selected in both regions. For capturing wild rodents, 10 Sherman live traps were placed on a deserted area at each location once a month in October and November. Peanut butter-coated biscuits were used as the bait for the wild rodents, and the traps were set at approximately 10 a.m. and removed at approximately 8 p.m. Twenty-seven rodents were captured (twelve in October and fifteen in November). After capture, the animals were euthanized via inhalation of 5% isoflurane in accordance with an approved animal use protocol, and the spleen, kidneys, lungs, and heart were harvested and stored at − 20 °C 33 . PCR amplification. To detect Borrelia DNA, nested PCR targeting the pyrG and ospA genes of Borrelia species was performed using genomic DNA extracted from the tissue specimens. For pyrG nested PCR, pyrG-1F/pyrG-1R primers (for the initial PCR step) and pyrG-2F/pyrG-2R primers (for nested PCR) were used 20 . For ospA nested PCR, Borrel-ospAF1/Borrel-ospAR1 primers (for the initial PCR step) and Borrel-ospAF2/Borrel-ospAR2 primers (for nested PCR) were used 17 . The primer sequences are listed in Table 4. PCR was performed using the AmpliTaq Gold 360 Master Mix (Applied Biosystems, Foster City, CA, USA) and an Applied Biosystems Veriti 96-Well Thermal Cycler. An enzyme reaction solution of 20 µL was used in the primary PCR; this solution was composed of 1 μL each of the forward and reverse primers (5 μM), 10 μL of Master Mix, 2 μL of a GC enhancer, and 4 μL of distilled water. Nested PCR was performed with the same reaction solution used in the initial PCR step, using the initial PCR product as the template. PCR was performed using gene-specific PCR primers at specific annealing temperatures under the following cycling conditions: 10 min at 94 °C for the predenaturation step, 30 cycles of 20 s at 94 °C, 30 s at the different annealing temperatures, 30 s-1 min at 72 °C, and a final extension step of 7 min at 72 °C. In each PCR run, a negative control (reaction mixture without the template DNA) was included. The genomic DNA of B. burgdorferi B31 Clone 5A1 was used as the positive control. The annealing temperatures are listed in Table 4. Upon the completion of PCR, the products were separated by electrophoresis on 1.2% agarose gels containing ethidium bromide.
Nucleotide sequencing. A QIAquick Gel Extraction Kit (Qiagen) was used to purify the PCR products, which were directly sequenced using the PCR primers and an automated sequencer (ABI Prism 3730XL DNA analyzer; Applied Biosystems) at Solgent (Deajeon, South Korea). To identify the bacteria, the sequences were analyzed using the BLAST network service (Ver 2.33; http:// www. techn elysi um. com. au/ chrom as. html) available from the National Center for Biotechnology Information (National Institutes of Health, Rockville, MD, USA).
Sequence similarity and phylogenetic analyses. The DNA sequence identity, contig generation, and homology comparison were confirmed using Lasergene v6 (DNASTAR, Madison, WI, USA) and the NCBI