Detection of Borrelia burgdorferi Sensu Lato and Relapsing Fever Borrelia in Feeding Ixodes Ticks and Rodents in Sarawak, Malaysia: New Geographical Records of Borrelia yangtzensis and Borrelia miyamotoi

Members of the Borrelia burgdorferi sensu lato (Bbsl) complex are etiological agents of Lyme disease (LD), and Borrelia miyamotoi is one of the relapsing fever Borrelia (RFB). Despite the serological evidence of LD in Malaysia, there has been no report from Sarawak, Malaysian Borneo. Thus, this study aimed to detect and characterize Borrelia in rodents and Ixodes ticks from primary forests and an oil palm (OP) plantation in Sarawak. Borrelia yangtzensis (a member of the Bbsl complex) was detected in 43.8% (14/32) of Ixodes granulatus; most of the positive ticks were from the OP plantation (13/14). Out of 56 rodents, B. yangtzensis was detected in four Rattus spp. from the OP plantation and B. miyamotoi was detected in one rodent, Sundamys muelleri, from the primary forest. Further, the positive samples of B. yangtzensis were randomly selected for multilocus sequence analysis (MLSA). The MLSA results of successfully amplified tick samples revealed a clustering with the sequences isolated from Japan and China. This study is the first evidence of B. miyamotoi, a known human pathogen in Malaysia, and B. yangtzensis, which is circulating in ticks and rodents in Sarawak, Malaysian Borneo, and presenting a new geographical record of the Borrelia spp.


Introduction
Members of the Borrelia burgdorferi sensu lato (Bbsl) complex are etiological agents of Lyme disease (LD), and Borrelia miyamotoi is one of the relapsing fever Borrelia (RFB) [1]. There are currently rajah, 2 M. whiteheadi, 45 Rattus spp., and 3 Sundamys muelleri ( Table 1). The rodents that were morphologically assigned to the Rattus spp., R. tiomanicus, and R. tanezumi, were collectively grouped as one Rattus spp., as they were not molecularly identified as a single species.
A total of 32 Ixodes ticks (22 females, 5 nymphs, and 5 larvae) were collected from the rodents and molecularly identified as I. granulatus (Table 1). All the I. granulatus ticks were engorged or semiengorged, except for one larva from the OP plantation. A similar system was also applied for tick sample identity, by using IG as the acronym for I. granulatus.

Detection of the Borrelia spp.
Out of the 56 rodent samples, four Rattus spp. (Sample IDs: OP-007, -014, -018, and -033) from the OP plantation and one S. muelleri (Sample ID: GGNP-04) from the GGNP were positive for the borrelial flagellin gene (flaB) in a polymerase chain reaction (PCR) ( Table 1). The prevalence of Borrelia spp. in rodents from GGNP, KNP, and the OP plantation were 16.7% (1/6), 0% (0/7), and 8.9% (4/45), respectively (Table 1). Subsequent analysis revealed that sequences from these five samples were different from each other. The sequences from OP-007 and OP-014 showed 100% (300/300 bp) and  A total of 32 Ixodes ticks (22 females, 5 nymphs, and 5 larvae) were collected from the rodents and molecularly identified as I. granulatus (Table 1). All the I. granulatus ticks were engorged or semi-engorged, except for one larva from the OP plantation. A similar system was also applied for tick sample identity, by using IG as the acronym for I. granulatus.

Phylogenetic Analysis
Collectively, 19 samples (5 rodents and 14 I. granulatus) were positive for flaB-PCR and were included in the phylogenetic tree construction. Out of 19 samples, 18 were assigned to the clade of Bbsl and clustered together with B. yangtzensis or B. valaisiana-related genospecies ( Figure 2). The remaining rodent sample (GGNP-04) was assigned to the RFB and clustered together with B. miyamotoi ( Figure 2). In addition, the phylogenetic tree based on the 16S rDNA sequences revealed the consistent clustering, as observed in flaB ( Figure 3).  shown in red and blue, respectively. All samples were collected from an oil palm (OP) plantation, except one sample, which was from Gunung Gading National Park (GGNP).

Multilocus Sequence Analysis of the Borrelia spp.
All I. granulatus samples included for the multilocus sequence analysis (MLSA) were successfully amplified for the eight housekeeping genes. In the phylogenetic tree based on the concatenated MLSA genes, the Borrelia spp. from I. granulatus were located in the clade of B. yangtzensis ( Figure 4). This trend was also confirmed in the other phylogenetic trees based on flaB and 16S rDNA (Figures 2 and 3). For the four rodent samples included in MLSA, a minimum of two (1/4), four (2/4), and six (1/4) housekeeping genes were successfully amplified. Therefore, the phylogenetic inferences of the Borrelia species in both rodent and tick samples were done on a per-gene basis. The phylogenetic analysis based upon each housekeeping gene showed that the rodent samples (OP-007, -014, -018, and -033) were located in the clade of B. yangtzensis with I. granulatus from this study

Multilocus Sequence Analysis of the Borrelia spp.
All I. granulatus samples included for the multilocus sequence analysis (MLSA) were successfully amplified for the eight housekeeping genes. In the phylogenetic tree based on the concatenated MLSA genes, the Borrelia spp. from I. granulatus were located in the clade of B. yangtzensis (Figure 4). This trend was also confirmed in the other phylogenetic trees based on flaB and 16S rDNA (Figures 2 and 3). housekeeping genes were successfully amplified. Therefore, the phylogenetic inferences of the Borrelia species in both rodent and tick samples were done on a per-gene basis. The phylogenetic analysis based upon each housekeeping gene showed that the rodent samples (OP-007, -014, -018, and -033) were located in the clade of B. yangtzensis with I. granulatus from this study ( Figure S1).

Discussion
We investigated Borrelia spp. in I. granulatus and rodents, L. sabanus, M. rajah, M. whiteheadi, Rattus spp., and S. muelleri from primary forests (GGNP and KNP) and an OP plantation in Sarawak, Malaysia. We identified B. yangtzensis from I. granulatus and Rattus spp. and B. miyamotoi from S. muelleri. This study is the first evidence of B. miyamotoi in Malaysia and B. yangtzensis in Sarawak, Malaysian Borneo.
Borrelia yangtzensis detected from I. granulatus in this study was formerly known as B. valaisianarelated genospecies since phylogenetic inferences showed a close relation but a clear distinction to B. valaisiana (a member of the Bbsl complex in Europe) [33]. However, unlike B. valaisiana that utilize birds as the reservoir host and Ixodes ticks as the vector [33,34], B. yangtzensis is maintained and transmitted through a natural infection cycle between rodents and Ixodes ticks [35]. Isolations of B. yangtzensis from different rodent species were recorded from Rattus spp., S. murinus, Mus spp., and A. agrarius in Japan, China, and Taiwan [35][36][37], as well as from different Ixodes tick species, such as I. nipponensis in South Korea and I. granulatus in Japan and China [35,38,39]. Recently, a Borrelia sp. closely related to B. yangtzensis was detected in Peninsular Malaysia from I. granulatus collected from different rodent species [30]. Takhampunya et al. [40] also identified B. yangtzensis from one rodent and two tick pools of Ixodes spp. collected from rodents in northern Thailand. This is similar to our study, as we detected B. yangtzensis from Rattus spp. and I. granulatus for the first time in Sarawak. These results suggest that B. yangtzensis is circulated and Rattus spp. and I. granulatus play the roles of the natural reservoir and vector, respectively, in Sarawak. However, there is a limitation in our study; as all of the positive Ixodes ticks were engorged, we could not rule out whether borrelial DNA detected from the tick samples was from the blood meal host or not. Thus, further investigations of B. yangtzensis in unfed I. granulatus are required to confirm the vector species of this bacterium in Sarawak. In addition, I. granulatus has been documented from migratory birds in Taiwan by Kuo et

Discussion
We investigated Borrelia spp. in I. granulatus and rodents, L. sabanus, M. rajah, M. whiteheadi, Rattus spp. and S. muelleri from primary forests (GGNP and KNP) and an OP plantation in Sarawak, Malaysia. We identified B. yangtzensis from I. granulatus and Rattus spp. and B. miyamotoi from S. muelleri. This study is the first evidence of B. miyamotoi in Malaysia and B. yangtzensis in Sarawak, Malaysian Borneo.
Borrelia yangtzensis detected from I. granulatus in this study was formerly known as B. valaisiana-related genospecies since phylogenetic inferences showed a close relation but a clear distinction to B. valaisiana (a member of the Bbsl complex in Europe) [33]. However, unlike B. valaisiana that utilize birds as the reservoir host and Ixodes ticks as the vector [33,34], B. yangtzensis is maintained and transmitted through a natural infection cycle between rodents and Ixodes ticks [35]. Isolations of B. yangtzensis from different rodent species were recorded from Rattus spp., S. murinus, Mus spp. and A. agrarius in Japan, China, and Taiwan [35][36][37], as well as from different Ixodes tick species, such as I. nipponensis in South Korea and I. granulatus in Japan and China [35,38,39]. Recently, a Borrelia sp. closely related to B. yangtzensis was detected in Peninsular Malaysia from I. granulatus collected from different rodent species [30]. Takhampunya et al. [40] also identified B. yangtzensis from one rodent and two tick pools of Ixodes spp. collected from rodents in northern Thailand. This is similar to our study, as we detected B. yangtzensis from Rattus spp. and I. granulatus for the first time in Sarawak. These results suggest that B. yangtzensis is circulated and Rattus spp. and I. granulatus play the roles of the natural reservoir and vector, respectively, in Sarawak. However, there is a limitation in our study; as all of the positive Ixodes ticks were engorged, we could not rule out whether borrelial DNA detected from the tick samples was from the blood meal host or not. Thus, further investigations of B. yangtzensis in unfed I. granulatus are required to confirm the vector species of this bacterium in Sarawak. In addition, I. granulatus has been documented from migratory birds in Taiwan by Kuo et al. [41]. Migratory birds are known to play an important role in the dispersal of Bbsl, with previous reports involving Ixodes ticks from Japan, South Korea, and Russia [42][43][44]. Moreover, studies in Canada by Scott et al. showed that migratory birds disperse Bbsl-infected ticks over a long distance and across geographical barriers [45,46]. Thus, investigation of ticks collected from migratory birds in Malaysia might help expand our knowledge of Bbsl, including B. yangtzensis.
Multilocus sequence analysis was first introduced by Margos et al. [47] for depicting the evolutionary processes of B. burgdorferi. By targeting multi loci, eight housekeeping genes were developed for the Bbsl complex [47] and have been subsequently used in other studies to characterize the complexity of the Bbsl genospecies [33,35]. In recent years, this method has also proven to be useful in comparing the intraspecific diversity, elucidating population genetic structure, and other ecological aspects that may contribute to the transmission dynamics of the Bbsl genospecies [48,49]. Further, MLSA has been used to confirm B. yangtzensis from the isolates of ticks and rodents from China and Japan [33]. Based on MLSA, the phylogeny inference revealed that the isolates formed two sister clusters, with each cluster consisting of isolates from both China and Japan. Concordantly, in our study, the concatenated sequences of B. yangtzensis from I. granulatus were also located in two sister clades.
On the human pathogenic aspect, B. valaisiana had been regarded as the causative agent of LD in humans but was recently proven otherwise [50]. Two human LD cases caused by B. valaisiana-related genospecies were reported from Japan and China [51,52], but Margos et al. [33,50] later on ratified them be B. yangtzensis and suggested B. valaisiana as a non-human pathogenic. Although B. yangtzensis may potentially be a human pathogenic, there is currently no study providing further evidence. Even though the significance of B. yangtzensis in humans and animals is not yet fully understood, the findings of B. yangtzensis in our study imply the likelihood that the bacterium circulates within the Ixodes ticks and rodents in primary forests and OP plantation in Sarawak. Borrelia spp. closely related to B. yangtzensis were also detected from I. granulatus in Peninsular Malaysia [26]. Of note, I. granulatus is a rare parasite of humans with few related reports. Yun et al. [53] identified only one female of I. granulatus from 261 ticks they collected from humans in South Korea. A checklist of ticks from Thailand, dated back to 1983, documented that humans could be hosts for this tick species [54]. However, it was not clear in these studies whether the I. granulatus collected were biting humans. Thus, more investigations are required to evaluate the pathogenicity to humans and understand the transmission cycle of B. yangtzensis in Malaysia.
Borrelia miyamotoi, the causative agent of RF, was first isolated from I. persulcatus ticks in Japan. To date, several Ixodes tick species are considered as a vector of B. miyamotoi. For instance, I. scapularis and I. pacificus are the vectors reported in the United States and Canada, I. ricinus in Europe, and I. persulcatus in Europe and Asia [16,23,[55][56][57]. So far, the reservoir hosts, based on the geographical distribution, for B. miyamotoi are still not well understood, but rodents and birds have been considered as the reservoir hosts in some regions [58,59]. In this study, we detected B. miyamotoi from S. muelleri in GGNP; this is the first report of B. miyamotoi in Malaysia. However, none of the I. granulatus examined in this study were positive for B. miyamotoi. Furthermore, the infection rate of B. miyamotoi was lower than that of B. yangtzensis in this study. Generally, the infection rate of B. miyamotoi in rodents and ticks appears to be lower than that of Bbsl species as per previous studies in Japan and Russia [56,58]. Further, another study on the prevalence of B. miyamotoi infection in I. scapularis conducted in Canada was low (<1%) [57]. Moreover, the reported prevalence of B. miyamotoi in questing Ixodes ticks ranged from 1.3% in I. Ricinus to 3.6% in I. persulcatus [60]. For future studies, the sample size of the rodents and ticks should be increased to find the vector tick species and to describe the diversity and distribution of B. miyamotoi in Sarawak. In addition, B. miyamotoi has been recently reported from H. concinna in Northeastern China [23]. In Europe, migratory birds have been reported as the reservoir host of B. miyamotoi or play a role in the dispersal of tick vectors [59]. Thus, the investigations of other tick species and birds may provide more in-depth insights of B. miyamotoi in Sarawak.
The sampling in this study was conducted only once for the primary forests (GGNP and KNP) and the OP plantation in different seasons, which yielded a small sample size, especially in the sampling during the wet season. A small sample size and lack of sampling repetition may have contributed to the low number of positive samples in this study; as B. miyamotoi was only positive in one rodent, and B. yangtzensis was not detected in the rodents from GGNP and KNP. In addition, only rodent spleens were used in this study for Bbsl and RFB screening. Future work to estimate the prevalence should include ear biopsies and other internal organs, as Bbsl and B. miyamotoi may not have the same strategies for maintenance and dissemination in the same reservoir host [61]; therefore, different tissue may yield different detection rate [62]. Despite the incomparable rodent numbers, the number of Ixodes ticks from the primary forests and OP plantation were fairly similar (14 and 18, respectively). In the OP plantation, 13 ticks were positive for B. yangtzensis; in contrast, in the primary forests, only one tick from KNP was positive. Land conversion with a human-dominated ecosystem could have a potent effect on reservoirs and the zoonotic risk, because of the alterations of host diversity and composition [63,64]. The difference observed in this study might be reflected by the variation between the primary forest and OP plantation. A follow-up study to evaluate this hypothesis should encompass a larger sampling size with repetition.
In conclusion, we examined Borrelia spp. in rodents and ticks from primary forests and an OP plantation. We reported the presence of B. miyamotoi for the first time in Malaysia. We also reported the first detection of B. yangtzensis, which was characterized using MLSA, in both rodents and I. granulatus in Sarawak. These findings of Borrelia spp. in Sarawak provide evidence of a new geographical record. Our study warrants the need for further investigations as it is important to determine how the Borrelia spp. may impact public health in Malaysia.

Ethics Approvals
The collecting of rodents and ticks was approved by the Sarawak Forests Department, Malaysia   Table 1). The sampling period for each site ranged from 5-10 days. The rodents were captured using collapsible cage traps; their tentative species, sex, breeding status, and body measurements were recorded. The captured rodents were individually anesthetized using isoflurane. Next, the ticks attached to each rodent were collected and kept separately in 70% ethanol. In addition, the selected rodents were euthanized following the method described by Taylor et al. [58] for internal organs collection. Finally, the harvested organs were kept in 70% ethanol and subsequently stored at −20 • C after being transferred back to the facility; the spleen samples of the rodents were used in this study.

DNA Preparation and Species Identification of Rodents and Ticks
DNA was extracted from the rodent spleens at the Faculty of Resource Science and Technology, Universiti Malaysia Sarawak, using a Wizard ® Genomic DNA Purification Kit (Promega, Madison, WI, USA), following the manufacturer's instructions. The DNA samples and ticks were sent to Hokkaido University, Japan, where the subsequent screenings and analyses of the samples were conducted.
For the molecular identification of the rodent species, we amplified a fragment of the cytochrome c oxidase subunit 1 (CO1) by polymerase chain reaction (PCR) using the primer pairs BatL5310 and R6036R (Table 3) [65]. The PCRs were conducted in a 20 µL reaction mixture using the Ex Taq Hot Start version (Takara Bio, Shiga, Japan) with the following conditions: 30 cycles of denaturation at 94 • C for 30 s, annealing at 48 • C for 30 s, and extension at 72 • C for 60 s.
The tick genera or species were morphologically identified based on the taxonomic keys [66,67], and a fragment of their mitochondrial 16S ribosomal DNA (16S rDNA; Table 3) was confirmed by sequencing [68]. One leg of the ticks was removed, and DNA was extracted from it by using the hot alkaline extraction method previously described by Mtambo et al. [69], with some modifications. Briefly, we added 10 µL of 100 nM of sodium hydroxide and incubated it at 95 • C for 10 min, followed by adding 2 µL of tris-hydrochloride buffer (pH 7.0). The PCR was conducted using Tks Gflex DNA Polymerase (Takara Bio, Shiga, Japan), with the following conditions: initial denaturation at 94 • C for 1 min, followed by 40 cycles of 98 • C for 10 s, 55 • C for 15 s, and 68 • C for 24 s, and a final extension at 68 • C for 5 min.
The amplification products from rodents and ticks were electrophoresed on a 1.2% agarose gel with Midori Green Direct DNA stain (Nippon Genetics, Tokyo, Japan) and visualized with a BLooK LED transilluminator (GeneDireX, Las Vegas, NV, USA). The Sanger sequencing was performed using the BigDye Terminator version 3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA). The sequencing products were analyzed on an ABI Prism 3130 x genetic analyzer (Applied Biosystems, Foster City, CA, USA), according to the manufacturer's instructions. The sequences were compared with public databases using the Nucleotide Basic Local Alignment Search Tool (BLASTn) (https://blast.ncbi.nlm.nih.gov/Blast.cgi). In addition, the Barcode of Life Data System (BOLD; http://www.barcodinglife.org) was also used for rodent identification [70].
After identifying the tick species, the remaining body of the Ixodes ticks was washed with sterile phosphate-buffered saline and individually crushed with Micro Smash MS-100R (TOMY, Tokyo, Japan) for 30 s at 2500 rpm. The DNA was extracted using the Wizard ® Genomic DNA Purification Kit (Promega, Madison, WI, USA), as described above.

Screening of the Borrelia spp.
DNA from the rodent spleens and Ixodes ticks were subjected to the screening of Bbsl and RFB using a nested PCR targeting the flaB, which is a 345 bp amplicon (Table 3) [71]. The PCR conditions were as follows: 25 and 30 cycles of denaturation at 94 • C for 30 s, 55 • C and 50 • C of annealing for 30 sec, and extension at 72 • C for 1 min in the first and second PCR, respectively. The positive samples of the flaB-PCR were further characterized by additional PCRs, targeting 16S rDNA, which is approximately a 1370 bp amplicon. For the Ixodes ticks, a single PCR with BF1 and BR1 primers was performed [73] (Table 3). While, for the rodent samples, universal primers targeting bacterial 16S rDNA were added for the first PCR [72], BF1 and BR1 primers were added for the second PCR. The PCR conditions for the single and nested PCRs were identical, i.e., 35 cycles of denaturation at 94 • C for 30 s, annealing at 55 • C for 30 s, followed by extension at 72 • C for 90 s. The PCR was conducted using the Ex Taq Hot Start version (Takara Bio, Shiga, Japan) with a reaction mixture of 20 µL. DNA from "Candidatus Borrelia fainii" strain Qtaro [74] and molecular-grade water were used for the positive and negative controls, respectively. Finally, the electrophoresis, PCR product purification, and Sanger sequencing were performed as mentioned in the subsection "DNA preparation and species identification of rodents and ticks".

Multilocus Sequence Analysis of the Borrelia spp.
Four rodent and six tick samples were randomly selected from the flaB and 16S rDNA PCR positive samples and were used for MLSA based on the sequences of eight housekeeping genes (clpA, clpX, nifS, pepX, pyrG, recG, rplB, and uvrA). In order to obtain the sequences from these genes, we employed previously described methods with slight modification [47]. Briefly, nested PCRs using Ex Taq Hot Start version (Takara Bio, Shiga, Japan) were performed without the touchdown step, initially. Then, for the samples that failed to amplify, we repeated the PCR with the touchdown step. Finally, the PCR products were observed with gel electrophoresis and purified using a FastGene Gel/PCR Extraction Kit (Nippon Genetics, Tokyo, Japan), followed by Sanger sequencing.