Borrelia carolinensis sp. nov., a New (14th) Member of the Borrelia burgdorferi Sensu Lato Complex from the Southeastern Region of the United States

Abstract

Approximately 118 Borrelia isolates were cultured from a variety of rodents, birds, and ticks collected in the southern United States. In addition to a highly diverse group of Borrelia bissettii strains and a homogenous group of Borrelia burgdorferi sensu stricto strains, a group of 16 isolates with unusual characteristics was found. The isolates were cultured from ear biopsy samples of the rodents Peromyscus gossypinus and Neotoma floridana trapped at five localities in South Carolina. A multilocus sequence analysis of the rrf-rrl intergenic spacer, 16S rRNA, fla, ospA, and p66 genes were used to clarify the taxonomic status of the new group of B. burgdorferi sensu lato isolates. Thirteen species of the B. burgdorferi sensu lato complex were used as controls. Unique restriction fragment length polymorphism patterns of the rrf-rrl intergenic spacer region and fla gene were recognized. Unique signature nucleotides were also found in the 16S rRNA gene. A phylogenetic analysis shows that the 16 new isolates cluster together but separately from the other species in the B. burgdorferi sensu lato complex. Our data strongly support the recognition of the 16 isolates as a new B. burgdorferi sensu lato species. We propose to name this genospecies “Borrelia carolinensis” with respect to the place of its currently known geographic location.

Today, there are 13 Borrelia genospecies globally described and recognized. These include Borrelia burgdorferi sensu stricto, B. afzelii, B. garinii, B. bissettii, B. valaisiana, B. lusitaniae, B. andersonii, B. tanukii, B. turdi, B. japonica, B. spielmanii, B. sinica, and, recently recognized, “Borrelia californiensis” (1, 7, 11, 16, 17, 20, 33, 35, 46, 48, 50, 57). These species are differently distributed throughout the world, are differently associated with vectors and hosts, and have different pathogenicity patterns (2, 3, 6, 34, 52, 58).

There have been several dominant tenets in the past concerning Lyme borreliosis in the United States (19, 36). Contrary to dogma, B. burgdorferi sensu lato is present and widely distributed in the southern United States. Our data and those of other investigators confirm that B. burgdorferi sensu lato occurs in many areas where it was not previously thought to occur (4, 10, 18, 21, 26, 27, 29-32, 37-39, 43, 44, 49, 53-55). An analysis of samples from our collection of Borrelia isolates indicates that there is greater genetic variability among southern B. burgdorferi species than that reported for northern strains. A variety of naturally infected mammals and birds were found to serve as reservoirs of B. burgdorferi in the southern United States (see reference 36). Moreover, several species of lizards demonstrated at the molecular level the presence of B. burgdorferi (8). In addition to the large groups of highly diverse strains related to Borrelia bissettii and Borrelia andersonii (22-25) and a group of strains of Borrelia burgdorferi sensu stricto (J. H. Oliver, Jr., N. Rudenko, and M. Golovchenko, unpublished data) (40), a group of 16 isolates with unusual characteristics (41) was recognized among approximately 275 isolates cultured from the southern United States (N. Rudenko and J. H. Oliver, Jr., presented at the 8th Annual Lyme & Other Tick-Borne Diseases Conference, Boston, MA, 26 October 2007).

The main purpose of this report is to clarify the taxonomic status of a novel group of Borrelia isolates from South Carolina (41). Our selection of the genes (rrf-rrl intergenic spacer, 16S rRNA, fla, ospA, and p66) for multilocus sequence analysis of a novel group of spirochetes was based on the availability of control sequences (in GenBank) from 13 species of the Borrelia burgdorferi sensu lato complex that were necessary for comprehensive analysis and comparison.

MATERIALS AND METHODS

Collection of rodents, locations, and Borrelia cultures.

Nine strains were isolated from the ear biopsy samples of Peromyscus gossypinus, 6 from ear biopsy samples of Neotoma floridana, and 1 isolate was cultured from the hard tick Ixodes minor that was feeding on N. floridana (Table 1). Borrelia isolates from the ear tissues were cultured in Barbour-Stoenner-Kelly H medium that contained 0.15% of agarose (Seakem; FMC Bioproducts, Rockland, ME), antibiotics (rifampin, phosphomycin), and fungicide (amphotericin B). The cultures were incubated in 5% CO₂ at 33 to 34°C. When the cultures reached a cell density of 2 × 10⁶ spirochetes/ml, they were stored at −80°C.

TABLE 1.

Borrelia isolates, locations, dates of isolation, and reservoir hosts

Isolate	Location in South Carolina	Date of collection	Host or tick/source
SCCH-6	Mt. Pleasant, Charleston County	February 1995	P. gossypinus/ear clip
SCCH-10	Mt. Pleasant, Charleston County	April 1995	P. gossypinus/ear clip
SCCH-11	Mt. Pleasant, Charleston County	April 1995	N. floridana/ear clip
SCCH-12	Mt. Pleasant, Charleston County	April 1995	N. floridana/ear clip
SCJ-1	Jasper County	December 1996	N. floridana/ear clip
SCJ-5	Jasper County	July 1997	N. floridana/ear clip
SCJ-6	Jasper County	August 1997	N. floridana/ear clip
SCW-13	Wedge Plantation, Charleston County	September 1994	P. gossypinus/ear clip
SCW-14	Wedge Plantation, Charleston County	September 1994	P. gossypinus/ear clip
SCW-19	Wedge Plantation, Charleston County	September 1994	P. gossypinus/ear clip
SCW-21	Wedge Plantation, Charleston County	September 1994	P. gossypinus/ear clip
SCW-22	Wedge Plantation, Charleston County	October 1994	I. minor (male)
SCGT-6	Georgetown County	May 1995	P. gossypinus/ear clip
SCGT-18	Georgetown County	January 1996	N. floridana/ear clip
SCGT-21	Georgetown County	January 1996	P. gossypinus/ear clip
SCSC-1	Sumter County	February 1995	P. gossypinus/ear clip

DNA purification, PCR amplification, cloning, and sequencing.

Total Borrelia DNA was purified using the DNeasy tissue kit (Qiagen) strictly according to the manufacturer's recommendations.

The MasterTaq kit (Eppendorf, Germany) that contained recombinant Taq DNA polymerase from Escherichia coli DH1 and a special 5× TaqMaster PCR enhancer was used for the amplification of the spirochete sequences. The PCR conditions for the amplification of the rrf-rrl, fla, ospA, and p66 genes were as follows: initial denaturation at 96°C for 5 min, followed by 30 cycles at 95°C for 30 s, annealing at 52°C for 30 s, and extension at 72°C for 1 min and the final extension step at 72°C for 10 min. In the case of the amplification of the 16S rRNA gene, the annealing temperature was increased to 58°C. Reactions were set up in a separate area with all precautions (supplies, equipment, and employee's personal safety items, pre- and postamplification activities). A negative control (no template) and positive control (B. burgdorferi B31 DNA) were added to each amplification reaction. The PCR products were separated in 2% (rrf-rrl, ospA, fla, and p66 amplicons) or 1.5% (16S rRNA amplicons) agarose gel, cut off the gel, purified, and submitted for direct sequencing to the University of Washington High-Throughput Genomic Unit (Seattle, WA). All templates were submitted in 96-well, skirted PCR plates. Sequences were determined twice in both directions, using the same specific primers that were used for the amplification of each gene. All sequences were analyzed with EditSeq module of DNASTAR software (DNASTAR, United Kingdom). Database searches used the BLAST programs of the NCBI (Bethesda, MD).

Primers and probes used for analysis of the Borrelia sequences.

All primers and probes used in this study are listed in Table 2.

TABLE 2.

Primers and probes used in this study

Target	Primer or probe	Species	Nucleotide sequence	Position (nt)	Reference
rrf-rrl intergenic spacer region	5S(rrf)F	B. burgdorferi sensu lato	CTGCGAGTTCGCGGGAGA	229-246	47
	23S(rrl)R	B. burgdorferi sensu lato	TCCTAGGCATTCACCATA	482-465	47
	Sl	B. burgdorferi sensu lato	CTTTGACCATATTTTTATCTTCCA	453-430	51
	Ss	B. burgdorferi sensu stricto	AACACCAATATTTAAAAAACATAA	322-299	51
	Ga	B. garinii	AACATGAACATCTAAAAACATAAA	322-298	51
	Af	B. afzelii	AACATTTAAAAAATAAATTCAAGG	305-278	51
	VS116	B. valaisiana	CATTAAAAAAATATAAAAAATAAATTTAAGG	305-278	51
rrs	16SMF	B. burgdorferi sensu lato	AGAGTTTGATCCTGGCTTAG	1529-1510	13
	16SMR	B. burgdorferi sensu lato	CCTCCCTTACGGGTTAGAA	84-103	13
	fD3	B. burgdorferi sensu lato	AGAGTTTGATCCTGGCTTAG	8-27	20
	T50	B. burgdorferi sensu lato	GTTACGACTTCACCCTCCT	1497-1479	20
	S5	B. burgdorferi sensu lato	GAGGAATAAGCTTTGTAGGA	442-462	20
	RS11	B. burgdorferi sensu lato	CTTAACCCAACACCTCACAGC	1086-1066	20
fla	Fla outF	B. burgdorferi sensu lato	AARGAATTGGCAGTTCAATC	271-290	8
	Fla outR	B. burgdorferi sensu lato	GCATTTTCWATTTTAGCAAGTGATG	767-743	8
	PROBE	B. burgdorferi sensu lato	CACATATTCAGATGCAGACA	300-319	59
	BbslSD	B. burgdorferi sensu lato	GCTGTAAATATTTATGCAGCTAATGTTGC	556-584	15
	FL 8	B. burgdorferi sensu stricto	CTCTGGTGAGGGAGCTCAAACTGCTCAGGATGCACCGGTTCAAGAGGGT	594-642	45
	FL 15	B. garinii	CTCTGGTGAAGGAGCTCAGGCTGCTCAGACTGCACCTGTTCAAGAAGG	594-641	45
	FL 16	B. afzelii	TGCTGGTGAGGGAGCTCAAGCTGCTCAGGCTGCACCTGTTCAAGAGGGTGCT	594-644	45
	BbssSD	B. burgdorferi sensu stricto	GGTTCAAGAGGGTGTTCAACAGG	630-652	15
	BgSD	B. garinii	AGGAGCTCAGGCTGCTCAGA	603-622	15
	BafSD	B. afzelii	CTCAAGAAGAAGGAGCTCAGCAA	644-666	15
	BjSD	B. japonica	GCCTGTTCAAGAAGGCATTCAACAG	627-651	15
	BandSD	B. andersonii	TGTTCAAGAGGGTATTCAACAG	630-651	15
	BvSD	B. valaisiana	ACACCTGTTGAAGAAGGTGCTCAACAG	625-652	15
	Bgp127SD	B. bissettii	GAAGGTGTTCAGCAAGAAGG	637-656	15
ospA	N1	B. burgdorferi sensu lato	GAGCTTAAAGGAACTTCTGATAA	184-206	14
	C1	B. burgdorferi sensu lato	GTATTGTTGTACTGTAATTGT	745-722	14
	Detect. probe	B. burgdorferi sensu lato	CTTGAAGGCGTAAAAGCTGACAAA	226-249	12
	SSO probe	B. burgdorferi sensu lato	ACTAATGTTTTGCCATCTTCTTTGAA	329-304	28
p66	outer1	B. burgdorferi sensu lato	CGAAGATACTAAATCTGT	147-158	8
	outer2	B. burgdorferi sensu lato	GCTGCTTTTGAGATGTGTCC	442-461	8

Analysis of Borrelia sequences.

The restriction fragment length polymorphism (RFLP) analysis of Borrelia sequences was done in silico using the free software available at http://insilico.ehu.es (5). The rrf-rrl intergenic spacer region was digested with the MseI and DraI restriction endonucleases; the fla gene was digested with HapII, HhaI, HincII, CelII, and DdeI. All RFLP patterns obtained were compared with patterns already published (11, 47, 58). The alignment of the rrf-rrl region, fla, and ospA genes (“virtual hybridization”) with the corresponding probes (Table 2) was conducted using the MegAlign module of DNASTAR software (DNASTAR, United Kingdom).

Phylogenetic analysis.

Available sequences from different loci (rrf-rrl, 16S rRNA, fla, ospA, p66) of all 13 control species of the Borrelia burgdorferi sensu lato complex were used as controls in the phylogenetic analysis. The alignments were done with ClustalX (version 1.81) (56). Identical sequences were excluded from analysis. Obviously misaligned characters were shifted manually. The phylogenetic reconstruction for the molecular data was inferred by using a maximum parsimony heuristic search, performed in PAUP* 4.0, beta version 10, by implementing the tree-bisection-reconnection algorithm. Gaps were treated as missing characters. Branch supports were calculated by bootstrap analyses (1,000 replicates for a molecular data set). The results were confirmed using the maximum likelihood method.

Nucleotide sequence accession numbers.

Sequences determined in this study have been deposited in GenBank and given the indicated accession numbers: EU072425 to EU072440 for the rrf-rrl intergenic spacer sequences, EU085403 to EU085418 for the 16S rRNA gene sequences, EU076485 to EU076500 for the fla gene sequences, EU076501 to EU076516 for the p66 gene sequences, and EU085387 to EU085402 for the ospA gene sequences.

RESULTS

Borrelia cultures.

Sixteen Borrelia isolates, discussed in this paper, were cultured from the ear biopsy samples (clips) of two main rodent reservoirs of Borrelia burgdorferi in the southern United States. All cultures that were proved to be contamination free are stored in the Borrelia culture collection of the James H. Oliver, Jr., Institute of Arthropodology and Parasitology, Georgia Southern University at Statesboro, and are available upon request. Repeated double-directional sequencing of the PCR products from five selected loci gave identical results for every discussed South Carolina isolate, confirming in this way the presence of one Borrelia species in each culture.

Analysis of the rrf-rrl intergenic spacer region.

The exact sizes of the rrf-rrl spacer regions (253 nucleotides [nt]) of the tested isolates were determined by direct sequencing of the purified amplicons. The rrf-rrl intergenic spacers of the B. carolinensis strains exhibited 100% sequence identity in this locus. The results of the in silico digestion with MseI and DraI revealed RFLP patterns different from any of those known for B. burgdorferi sensu lato species or genomic groups, confirming the novelty of the South Carolina species (Table 3). The “virtual hybridization” results of the B. carolinensis rrf-rrl region showed 100% identity to the probes designed for the detection of spirochetes from the B. burgdorferi sensu lato group (59), while available species-specific oligonucleotides were not identical to any part of the analyzed intergenic spacer region of B. carolinensis.

TABLE 3.

RFLP analysis of the 5S-23S rRNA intergenic spacer of species from the Borrelia burgdorferi sensu lato complex

Species	Amplicon size (bp)	In silico RFLP pattern for indicated restriction enzyme
		DraI	MseI
B. burgdorferi B31	254	144, 52, 29, 28	107, 52, 38, 29, 28
B. afzelii VS461	246	174, 52, 20	107, 68, 51, 20
B. andersonii 21133	266	No site	119, 67, 52, 28
B. bissettii DN127	257	144, 53, 33, 27	107, 52, 38, 33, 37
B. bissettii 25015	253	173, 53, 27	107, 52, 34, 27, 17, 12, 4
B. californiensis 443	253	173, 80	107, 51, 37, 30, 28
B. carolinensis	253	173, 53, 27	107, 67, 52, 27
B. garinii 20047	253	201, 52	107, 95, 51
B. garinii NT29	253	144, 57, 52	107, 57, 51, 38
B. japonica HO14	236	No site	107, 78, 51
B. lusitaniae PotiB2	257	145, 83, 29	107, 82, 39, 29
B. sinica CMN3	235	157, 49, 29	107, 48, 38, 29, 13
B. spielmanii A14S	225	158, 67	108, 66, 51
B. tanukii Hk501	245	173, 52, 20	174, 51, 20
B. turdi Ya501	248	144, 81, 23	107, 51, 38, 21, 16, 8, 7
B. valaisiana VS116	255	203, 52	174, 51, 23, 7

Analysis of the partial fla gene.

Amplicons of the fla gene from B. carolinensis (497 nt) isolates were highly homogeneous among themselves and did not reveal identity to the known spirochete species. Four nucleotide substitutions that cause two amino acid substitutions (isolates SCCH6 and SCGT18) were the only differences among the B. carolinensis isolates. The “virtual hybridization” results of the B. carolinensis fla gene showed 100% identity to two probes designed for the detection of B. burgdorferi sensu lato but showed no identity to an additional seven (plus three) species-specific probes for the identification of B. burgdorferi sensu stricto B. garinii, B. afzelii, B. japonica, B. andersonii, B. valaisiana, and B. bissettii (Table 2). The RFLP pattern for B. carolinensis was unique (Table 4) and undescribed previously (11).

TABLE 4.

RFLP analysis of the partial sequence of the fla gene of species from the Borrelia burgdorferi sensu lato complex

Species	fla size (bp)	In silico RFLP pattern for indicated restriction enzyme
		CelII	DdeI	HapII	HhaI	HincII
B. burgdorferi B31	488		338, 150	349, 139
B. afzelii VS461	488	380, 108	305, 108, 42, 33
B. andersonii 21133	488		238, 150, 100
B. bissettii DN127	487		220, 117, 78, 72		404, 83
B. californiensis 443a	456a		225, 117, 111, 3a		281, 175a
B. carolinensis	488	365, 123	221, 117, 78, 45, 27
B. carolinensis SCGT18	488		221, 117, 78, 72
B. garinii 20047	488		329, 78, 72, 9		405, 83	453, 35
B. japonica HO14	488		305, 150, 33		348, 140
B. lusitaniae PotiB2	488		338, 150		393, 83, 12	453, 35
B. sinica CMN3	488		338, 150		300, 105, 83	453, 35
B. spielmanii A14S	488		338, 150		348, 140
B. tanukii Hk501	488		338, 150		256, 232
B. turdi Ya501	488		221, 150, 117		405, 83	453, 35
B. valaisiana VS116	488		188, 135, 117, 33, 15

^aThe partial flagellin sequence of B. californiensis available in GeneBank (accession number DQ393346) differs significantly in size and represents the region shifted 218 nt down to the 3′ end of the gene in comparison to the rest of the control species: i.e. its RFLP pattern is irrelevant.

Analysis of the 16S rRNA gene.

Amplification of the partial 16S rRNA (rrs) gene involved three sets of primers (Table 2). Adjusted to the sizes of the control sequences available in GenBank (1,362 nt), B. carolinensis rrs sequences were aligned with the rrs sequences from 13 spirochete species of the B. burgdorferi sensu lato complex and representative spirochete species from the relapsing fever group. The uniqueness of the signature nucleotides in the B. carolinensis isolates, which are phylogenetically significant in species identification, was revealed (Table 5). B. burgdorferi B31 was used as a baseline for the alignment, and the signature nucleotide positions were numbered according to the full rrs sequence of this species.

TABLE 5.

Analysis of signature nucleotides of the 16S rRNA of species from the Borrelia burgdorferi sensu lato complex and representatives from the relapsing fever group

Group and species	Nucleotide at indicated positiona
	116	126	143	170	171	175	203	243	252	273	323	470	472	588	626	675	707	990	1028	1111	1132	1249	1255
B. burgdorferi sensu lato complex
B. burgdorferi B31	G	T	T	G	G	G	C	C	A	A	T	A	T	A	C	T	A	A	T	A	G	A	A
B. afzelii	G	T	C	G	G	A	C	T	G	G	T	A	C	A	G	C	G	A	T	A	A	A	A
B. andersonii	G	T	C	G	G	A	C	C	A	G	T	A	T
B. bissettii DN127	G	T	T	A	G	A	C	C	A	G	T	A	T	A	C	T	A	A	T	A	G	G	A
B. bissettii 25015	G	T	T	A	G	G	C	C	A	G	T	A	T	A	C	T	G	A	T	A	G	A	A
B. californiensis	G	T	C	A	G	G	C	C	A	G	T	G	T	A	A	T	G	A	T	A	G	A	A
B. carolinensis	G	T	T	G	A	A	T	C	A	G	C	A	C	G	C	T	A	A	C	G	G	A	A
B. carolinensis SCJ-1	G	T	T	G	A	A	T	C	A	G	C	A	C	G	C	T	A	G	C	G	G	A	A
B. garinii	G	C	T	A	G	A	C	C	A	G	T	A	T	A	A	T	A	A	T	A	G	A	A
B. japonica	G	C	C	G	G	A	C	T	A	A	T	G	T	G	G	T	A	A	T	A	G	A	G
B. lusitaniae	G	T	C	A	G	A	C	C	A	G	T	A	T	A	A	T	A	A	T	A	G	A	G
B. sinica	G	T	T	G	G	A	C	T	A	A	T	A	C	G	A	T	A	A	T	A	G	A	G
B. spielmanii	G	T	C	G	G	A	C	T	G	G	T	G	T	A	A	C	G	A	T	A	A	A	A
B. tanukii	G	C	C	A	G	A	C	T	G	G	T	G	T	A	C	C	G	A	T	A	G	G	A
B. turdi	G	T	T	G	G	A	C	T	G	G	T	G	T	A	G	C	G	A	C	A	G	A	A
B. valaisiana	G	T	C	G	G	A	C	T	G	G	T	A	T	A	A	C	G	A	T	A	A	A	A
Relapsing fever group
B. hermsii	A	T	T	G	G	A	C	C	A	T	T	A	T	G	A	C	G	A	T	G	A	T	A
B. lonestari	A	C	T	A	G	A	C	C	A	T	T	C	T	G	A	C	G	A	T	G	A	T	A
B. miyamotoi	A	C	T	A	G	A	C	C	G	T	T	A	T	G	A	C	G	A	T	G	A	T	A
B. parkeri	A	T	T	G	G	A	C	C	A	T	T	C	T	G	A	C	G	A	T	G	A	T	A
B. recurrentis	A	T	T	G	G	G	C	C	A	T	T	C	T	G	G	C	G	A	T	G	A	T	A
B. turicatae	A	T	T	G	G	A	C	C	A	T	T	C	T	G	A	C	G	A	T	G	A	T	A

^aThe bold type indicates unique signature nucleotides in the Borrelia carolinensis 16S rRNA gene.

Analysis of the ospA gene.

The ospA gene of B. burgdorferi is genetically heterologous, which is useful in the species-specific separation of B. burgdorferi sensu lato strains. The sequences of all 16 B. carolinensis ospA amplicons (561 nt) were identical. Alignment of the partial sequences of the B. carolinensis ospA gene with two oligonucleotide probes, designed for the detection of B. burgdorferi sensu lato species (Table 2), revealed 100% identity within two different regions of the gene. The clustering of Borrelia strains in the phylogenetic tree based on ospA DNA sequences is in agreement with the classification based on the sequence analysis of conserved chromosomal genes, as well as with data obtained by pulsed-field gel electrophoresis and randomly amplified polymorphic DNA fingerprinting (2, 12).

Analysis of the p66 gene.

All sequences of the partial p66 genes (315 nt) from 16 B. carolinensis isolates were identical. In contrast to the amount of data available in GenBank for the analysis of other spirochete loci, such as the rrf-rrs intergenic spacer, 16S rRNA gene, fla gene, or ospA gene, the data about the B. burgdorferi p66 gene are rather limited. It is known that the p66 gene exhibits considerable sequence conservation among Lyme disease Borrelia species but is highly variable in size and sequence in the relapsing fever Borrelia group. Due to these differences, the p66 gene is successfully employed for the separation of Borrelia spp. among the two groups and the phylogenetic placement of new strains.

Phylogenetic analysis.

The distance matrix for phylogenetic analysis was generated from the adjusted and aligned 2,977 nucleotides of the partial rrs gene (1,360 nt), the partial ospA gene (561 nt), the partial fla gene (488 nt), the partial p66 gene (315 nt), and the complete rrf-rrl spacer (253 nt) of each isolate. Identical sequences were excluded from the analysis, resulting in four representatives of the B. carolinensis species—SCGT6, SCGT18, SCJ1, and SCW22—that represent the rest of the 13 identical isolates. A complete analysis of the concatenated sequences of all five loci for phylogenetic purposes failed because of the unavailability of the sequences from the corresponding genes of the 13 control Borrelia species in GenBank. The phylogenetic analysis was completed in three steps. An analysis of the concatenated sequences of the fla, rrs, and rrf-rrl loci of B. carolinensis with the 13 control species of Lyme disease complex clearly exhibited the separation of spirochete species into two groups. One group includes the isolates from Eurasia, whereas the other group includes the typical U.S. strains (B. burgdorferi ss, B. andersonii, B. bissettii, and B. californiensis) and isolates from B. carolinensis species that formed a separate cluster. The allocation of B. carolinensis isolates on the tree was supported by the distinctively different PCR-RFLP pattern of the fla gene and rrf-rrl spacer and the unique signature nucleotides on the 16S rRNA sequence (Fig. 1A). The addition of the fourth locus (the partial ospA gene) to the analysis confirmed the separation of B. carolinensis isolates into an independent branch. Analysis of the four loci lacks two species from the Lyme disease complex (B. sinica and B. tanukii) because of sequence unavailability (Fig. 1B). The concatenated matrix of the five loci (p66, ospA, fla, rrs, and rrf-rrl) includes only five control species from the Lyme disease complex. The results of the phylogenetic analysis confirm once again that isolates of B. carolinensis species are closely associated and constitute a new taxon in the B. burgdorferi sensu lato complex (Fig. 1C).

FIG. 1.

Phylogenetic analysis based on concatenated sequences of control Borrelia species available from databases and obtained in this study. (A) Three loci—fla, rrs, and the rrf-rrl intergenic spacer region; (B) four loci—ospA, fla, rrs, and the rrf-rrl intergenic spacer region; (C) five loci—p66, ospA, fla, rrs, and the rrf-rrl intergenic spacer region. Trees were constructed with the maximum parsimony method. Support values were obtained in a bootstrap procedure based on 1,000 replications. Trees are unrooted. The short names for the Borrelia species used in Fig. 1 are: B.b.B31, Borrelia burgdorferi B31; B.afz., Borrelia afzelii VS461; B.and., Borrelia andersonii 21133; B.biss., Borrelia bissettii DN127; B.calif., Borrelia californiensis CA443; B.gar., Borrelia garinii 20047; B.jap., Borrelia japonica HO14; B.lus., Borrelia lusitaniae PotiB2; B.sin., Borrelia sinica CMN3; B.spielm., Borrelia spielmanii A14S; B.tan., Borrelia tanukii HK501; B.tur., Borrelia turdi Ya501; B.val., Borrelia valaisiana VS116. The new member of the Borrelia burgdorferi sensu lato complex, Borrelia carolinensis, is represented by isolates SCGT18, SCGT6, SCJ1, and SCW22.

DISCUSSION

From the five sites of sample collection of B. carolinensis, only one, Sumter County, South Carolina, is inland. The others are of coastal location. Georgetown and Charleston are neighboring counties and are separated from Jasper County by Beaufort County. The numbers of samples collected in each locality are different (Table 6). Nine strains of B. carolinensis were isolated from P. gossypinus, six from N. floridana, and one from the hard tick I. minor (male) feeding on an N. floridana rodent. Ixodes minor usually does not bite humans; however, it appears to be more important as a maintenance vector in the enzootic cycle of B. burgdorferi sensu lato than the “bridge” vector Ixodes scapularis, which feeds on nonhuman species and humans (36). The possible presence of undescribed Borrelia species together with common species for the United States was previously discussed (36, 42). Northern coastal California and the northeastern and upper Midwest states traditionally are recognized as the three principal regions in the United States where Lyme disease is endemic. But during the last decade, at least three Borrelia species (B. burgdorferi sensu stricto B. bissettii, and B. andersonii) were identified and already described in southeastern states (Georgia, Florida, and South Carolina) (22-25, 36-43). As many as 69% of Peromyscus gossypinus rodents, trapped in coastal localities, were proved to be naturally infected (38, 39). The presence of B. burgdorferi sensu lato in the region suggests that coastal sites in the southeast might represent a risk for contracting Lyme borreliosis. The description of a new species, Borrelia carolinensis, further adds support to the conclusion that spirochete diversity is greater in the southeastern United States than in the northern states.

TABLE 6.

Geographical distribution and prevalence of B. carolinensis in South Carolina

Location	Total no. of samples collected	No. (%) of samples identified as B. carolinensis	No. of samples from the indicated source
			P. gossypinus	N. floridana	I. minor
Charleston County, Mt. Pleasant	23	4 (17.3)	2	2	0
Charleston County, Wedge Plantation	47	5 (10.6)	4	0	1
Jasper County	4	3 (75)	0	3	0
Georgetown County	14	3 (21.4)	2	1	0
Sumter County	6	1 (16.6)	1	0	0
Total	94	16 (17)	9	6	1

The multilocus sequence analysis of 16 South Carolina isolates revealed unknown RFLP patterns of their rrf-rrl intergenic spacer regions and fla genes, unique signature nucleotides in the 16S rRNA gene, and separate clustering from the other species in phylogenetic analysis based on concatenated sequences of five selected loci of new species. The results of the multilocus sequence analysis of novel Borrelia species, presented in this paper, clearly define the new taxonomic status of the South Carolina group of spirochete isolates in the B. burgdorferi sensu lato complex. The cotton mouse P. gossypinus and the eastern wood rat N. floridana are shown to be the primary reservoir hosts of B. carolinensis. The geographical distribution of both rodent species and the knowledge that they are parasitized by I. scapularis, Ixodes affinis, I. minor, Dermacentor variabilis, and Amblyomma maculatum (9) may be used as indirect evidence of the possible distribution of B. carolinensis isolates in the United States. However, as previously stated, the maintenance of Borrelia species is dependent on the relative abundance of reservoir hosts and vector-competent ticks and the intensity of host-vector interactions. Perhaps the identification of well-established populations of B. carolinensis was predetermined by natural factors that exist in the southeastern part of the United States.