ABSTRACT
Multilocus sequence typing of Borrelia hermsii isolates reveals its divergence into two major genomic groups (GG), but no differences in transmission efficiency or host pathogenicity are associated with these genotypes. To compare GGI and GGII in the tick-host infection cycle, we first determined if spirochetes from the two groups could superinfect the tick vector Ornithodoros hermsi. We infected mice with isolates from each group and fed ticks sequentially on these mice. We then fed the infected ticks on naive mice and measured GGI and GGII spirochete densities in vector and host, using quantitative PCR of genotype-specific chromosomal DNA sequences. Sequential feedings resulted in dual tick infections, showing that GGI or GGII primary acquisition did not block superinfection by a secondary agent. On transmission to naive mice at short intervals after acquisition, ticks with primary GGI and secondary GGII spirochete infections caused mixed GGI and GGII infections in mice. However, ticks with primary GGII and secondary GGI spirochete infections caused only GGII infections with all isolate pairs examined. At longer intervals after acquisition, the exclusion of GGI by GGII spirochetes declined and cotransmission predominated. We then examined GGI and GGII spirochetemia in mice following single inoculation and coinoculation by needle and found that GGI spirochete densities were reduced on multiple days when coinoculated with GGII. These findings indicate that dual GGI-GGII spirochete infections can persist in ticks and that transmission to a vertebrate host is dependent on the order of tick acquisition and the interval between acquisition and transmission events.
INTRODUCTION
Borrelia hermsii is the primary agent of tick-borne relapsing fever in western North America, where it persists in enzootic cycles involving the argasid (soft tick) vector Ornithodoros hermsi and diverse vertebrate hosts (1). Argasid ticks are predominately nest dwelling and feed at each immature stage and multiple times as adults, with the potential for acquisition and transmission of multiple spirochete genotypes (2). As rapid feeders, their occasional infestation of humans is rarely discovered, and spirochete isolation from ticks involved in transmission is virtually unreported. Spirochete DNA analysis from clinical samples, rodent hosts, and ticks indicates divergence into two major B. hermsii genomic groups (GGI and GGII), based on the 16S-23S rRNA gene intergenic spacer (2, 3) and multilocus sequence typing (4). The latter study demonstrated that GGI and GGII isolates are distributed throughout the range of the vector. Clinical isolates from a western Montana site showed sympatry of GGI and GGII spirochetes within a microenvironment where the disease is endemic (4, 5), suggesting that ticks may be exposed to spirochetes of both genotypes during their life cycle. Tick vectors and mammalian hosts can show significant heterogeneity in pathogen populations, as shown by studies on B. burgdorferi sensu lato and sensu stricto isolates (6–8) and Anaplasma marginale strains (9, 10). To date, there are no reports showing infection of field-collected O. hermsi with mixed genotypes of B. hermsii. Detecting mixed spirochete populations in field-collected ticks and hosts requires methods for differential genotyping or serotyping of each sample; in this report, we describe reagents to detect relative GGI-GGII spirochete transmission efficiencies in an experimental enzootic model.
To compare tick infection and transmission, we provided the conditions necessary for superinfection of ticks by sequential feedings on infected mice, fed the ticks on naive mice, and examined both ticks and mice for GGI and GGII spirochete densities. Our three goals were to see if O. hermsi ticks can acquire and maintain GGI and GGII isolates by serial feeding, if the isolate pairs are cotransmitted to naive mice, and if there are reproducible differences in GGI and GGII infections of vector and host. We have previously shown that simultaneous infection by GGI and GGII organisms in O. hermsi does occur by artificial infection and enables cotransmission by the tick vector to mice (11). In contrast, we asked in this study if tick primary acquisition of GGI or GGII spirochetes from an infected mouse affects secondary acquisition of another spirochete population or subsequent transmission to a naive host.
MATERIALS AND METHODS
Mouse infection, tick acquisition, and transmission to mice.B. hermsii GGI isolates DAH, MIL, and BYM; GGII isolates YOR and MTW-2 (4); GGI isolate LAK-4; and GGII isolate LAK-2 (2) were cultured from frozen stocks at 35°C in BSK-II medium (12) supplemented with 12% nonhemolyzed rabbit serum (Pel-Freez, Rogers, AR). Groups of two or three IRW strain inbred mice (13) were injected intraperitoneally or retro-orbitally with isolates at culture passage 10 or lower (relative to initial mouse blood isolation) and monitored by dark-field microscopic evaluation of tail vein blood from days 3 to 7 postinoculation to detect primary spirochetemia (Nikon Eclipse E600; 40× objective; Nikon Instruments, Melville, NY). Mouse experiments were performed at Rocky Mountain Laboratories in compliance with institutional Animal Care and Use Committee guidelines. Uninfected O. hermsi at the second nymphal stage of development, from a laboratory colony maintained at 21°C in a 14–10 light-dark cycle and 89% relative humidity (RH) (above a saturated solution of KCl), were fed on one of the inoculated mice within 2 h of detecting ≥1 × 106 spirochetes/ml blood by dark-field microscopy count. Engorged ticks were held 4 to 9 months at 89% RH, and the molted third-stage nymphs were fed, as described above, on a mouse infected with an isolate from the alternate genomic group to attempt superinfection. At each of the primary and secondary feedings on infected mice, additional colony ticks were also fed as single-acquisition controls. Before transmission assays, the ticks were held for the times specified in Table 1 until feeding on naive IRW female mice. Transmission feeding on mice was followed by daily blood evaluation for spirochetemia, as described above. On the second day on which spirochete presence was ≥1 × 106/ml, or at 7 days postfeed if spirochete density was <106/ml, blood was collected by cardiac puncture or tail bleed. Blood was stored at −20°C until DNA isolation on all samples (Qiaquick DNeasy Kit or QiaAmp DNA Micro Kit; Qiagen, Valencia, CA). Ticks were maintained under colony conditions for 2 to 4 weeks posttransmission before dissection or held for a second transmission feeding on additional mice. Ticks were dissected in phosphate-buffered saline (PBS) on Teflon-printed slides (Electron Microscopy Sciences, Hatfield, PA) for fixation of salivary glands (see below) and recovery of macerated midgut tissue for storage at −20°C until DNA extraction of all samples (Qiaquick DNeasy Kit; Qiagen).
Spirochete densities in mouse blood prior to tick feeds, intervals between acquisition tick feeds, and intervals to transmission feeds on mice
Immunofluorescence microscopy on salivary glands.For the direct fluorescent antibody (DFA) test, dissected tick salivary glands on slides were rinsed briefly in distilled water, allowed to dry, fixed in acetone for 10 min, and stored at −20°C. The water rinse did not appear to damage tissue integrity but increased tissue adherence to slides upon rehydration. Thawed slides were rehydrated in PBS, blocked for 1 h in Image-iT FX Signal Enhancer (Invitrogen, Grand Island, NY), washed in PBS, blocked for 30 min in PBS-0.75% bovine serum albumin (BSA), and incubated for 30 min with fluorescent conjugates of monoclonal antibodies to Vtp-5 (H3548) and Vtp-6 (H4825) (4) diluted in PBS-0.75% BSA. The slides were washed in PBS, air dried, and mounted in Vectashield mounting medium (Vector Laboratories, Burlingame, CA). Total numbers of fluorescent spirochetes in each pair of glands were determined, using a Nikon Eclipse E800 fluorescence microscope (Nikon); images were captured with a DXM1200C digital camera and ACT-1C software, version 1.0.1.5 (Nikon).
Mouse coinfection by needle inoculation.To obtain mouse-passaged spirochetes, isolates MTW-2, DAH, MIL, and YOR were adjusted to 1 × 108 (GGI isolates) or 1 × 107 (GGII isolates) spirochetes/ml BSK-II (Petroff-Hausser chamber; EMS, Washington, PA), and 0.1 ml was injected retro-orbitally into IRW adult female mice. The 10-fold difference in GGI-GGII inoculation densities resulted in similar spirochetemia densities in blood at 3 to 5 days postinfection (data not shown). Citrated blood (10 to 20 mM sodium citrate, final concentration) was collected during the first spirochetemia by cardiac puncture; 0.1 ml was cultured in 6.5 ml medium, and the remaining blood was centrifuged at 200 × g for 20 min to recover plasma. To recover spirochetes, 0.2 to 0.3 ml of plasma was centrifuged at 13,000 × g for 10 min, and pelleted spirochetes were resuspended for culture in 6.5 ml of medium in an adaptation of previously published methods (14). At 3 to 11 days postinoculation, first-passage cultures from whole blood or plasma were frozen at a spirochete density of ≥107/ml at −80°C in aliquots containing 20% rabbit serum and 20% glycerol. To compare spirochetemic densities in mice, thawed aliquots of the first-passage cultures were centrifuged, resuspended in BSK medium, and incubated for 2 to 3 days to produce second-passage cultures. Second-passage cultures were adjusted to 2 × 104 spirochetes/ml (Petroff-Hausser counting chamber) and injected retro-orbitally into IRW female mice as 0.1-ml suspensions containing 1 × 103 GGI, 1 × 103 GGII, or 1 × 103 (each) GGI and GGII isolates (five mice per group). The density of spirochetes was monitored daily by dark-field microscopy, and on days 3 through 6 postinoculation, 5 μl of tail vein blood was collected and stored at −20°C until DNA extraction, with 1 μg carrier RNA, on all samples (QiaAmp DNA Micro Kit; Qiagen).
Quantitative PCR (qPCR) for GGI and GGII B. hermsii detection in ticks and mice.Quantification of the B. hermsii 16S rRNA gene (rrs) and alleles at the vtp locus (vtp-5 and vtp-6) was performed with primer-probe sets previously described (11). Targets for differentiating GGI and GGII were identified by alignment of the GGI DAH chromosome (GenBank accession number CP000048.1) and a preliminary GGII MTW-2 chromosome assembly (T. G. Schwan, S. J. Raffel, and S. F. Porcella, unpublished data) with BLASTN version 2.2.25 (15), accessed at the National Center for Biotechnology Information website (http://blast.ncbi.nlm.nih.gov). Nonaligned gaps of ≥4 nucleotides (nt) were screened for the presence of distinct DAH and MTW-2 qPCR primer-probe targets, as determined with ABI Primer Express (version 2.0.0; Applied Biosystems, Foster City, CA). Forward and reverse primers to a region of the bh0260 open reading frame (ORF) in DAH (bh0260-1 allele) and to a region of bh0260 in MTW-2 (bh0260-2 allele; GenBank accession number KC997179) were tested for specificity as PCR primer pairs in MacVector (version 11.1.2: MacVector, Inc., Cary, NC). Primer and probe oligonucleotide sequences for detecting bh0260-1 in DAH were as follows: forward primer, AGTTTGGAGGATGCTAAGTCGAAT; probe, 6-carboxyfluorescein (FAM)-ATGCTTTGCCTCACTAAACCTAATGGGCAAAT-6-carboxytetramethyl-rhodamine (TAMRA); and reverse primer, TCTTCGGCGAATCCCTCTT. For detecting bh0260-2 in MTW-2, the sequences were as follows: forward primer, TGTAGAGCATGAAGAGTCGGTTTT; probe, FAM-ACCATCTTGTCCAACAACACTGCCAAATTCTT-TAMRA; and reverse primer, CAACATTATCCAAACCAACATCTAGAG. All oligonucleotides were purchased from Sigma-Aldrich (St. Louis, MO). Each qPCR target assay was performed in triplicate on DNA samples. Typically, 25 μl of the 60-μl total eluate from DNA purification was diluted with 25 μl 1 mM Tris-0.1 mM EDTA, pH 8.0, and used in 5-μl aliquots for each rrs, bh0260-1, and bh0260-2 target. Target sequences were detected with DyNAmo Flash Probe qPCR Master Mix, as recommended by the supplier (Thermo Fisher Scientific, Waltham, MA), in 20-μl reaction volumes and 300 nM primer and 50 nM probe concentrations, using an ABI Prism 7900HT Sequence Detection System (Applied Biosystems). Thermal-cycling parameters were as follows: one cycle each of 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. The reported values are from assays with ≥90% amplification efficiency (slope of the threshold cycle [CT] versus log spirochete equivalents, less than or equal to −3.6). The bh0260-1 qPCR was specific for amplification of the GGI target to 3.9 × 105 GGII spirochetes per reaction, whereas the bh0260-2 reaction was specific for amplification of the GGII target to ≥2.7 × 105 GGI spirochetes per reaction (data not shown). Standards representing 3 × 105 to 3 × 10−1 spirochetes per assay consisted of 5 μl serially diluted DAH (rrs and bh0260-1 assays) and MTW-2 (bh0260-2 assay) DNA. Detection sensitivity was routinely less than one spirochete per reaction, due to the multiple chromosome copies per spirochete (16). In qPCR determinations on blood DNA, the lower limit of detection varied from 2.5 × 102 spirochetes/ml (100 μl extracted blood) to 5 × 103 spirochetes/ml (5 μl extracted blood). Calculation of spirochete numbers was based on the observed average yield of RNase A-treated DNA (DNeasy Kit; Qiagen) from passage 8 DAH (rrs and bh0260-1 assays) and passage 9 MTW-2 (bh0260-2 assay) of 106.5 ± 5.0 ng/107 spirochetes; the previously reported DNA content of 12.1 Mb/cell for the B. hermsii HS1 isolate (16), using a value of 660 g/mol nucleotide pair, provides a theoretical value of 133 ng DNA/107 spirochetes. No correction was made for the extraction efficiencies of experimental DNA tissue sources or for the chromosomal copy numbers per cell under different growth conditions.
Statistical analysis.Statistical analyses were performed with tests described in individual experiments, using Prism version 6.0b (Prism for Macintosh; GraphPad Software, La Jolla, CA).
RESULTS
Identification of chromosomal qPCR targets for GGI-GGII differentiation.Our experimental rationale in this study was to sequentially infect ticks with two isolates of B. hermsii, one from each of the genomic groups, and then observe the transmission of these isolates to naive mice. As we possess multiple isolates from GGI and GGII, the identification of phenotypic differences between GGI and GGII must involve several pairs of isolates to verify that phenotypes are reproducible among members of each genomic group. In each of the tick superinfections, we identified characteristics of the GGI-GGII spirochete isolates that explain their inclusion in the design. To quantitate GGI and GGII spirochetes in mixed infections, we previously targeted two alleles at the plasmid-borne vtp locus to differentiate spirochete isolates relative to the 16S rRNA-encoding rrs locus on the chromosome (11). B. hermsii Vtp is produced during the tick infection portion of the infectious cycle (17). The GGI isolates that have been typed possess vtp-1, -2, -3, -4, or -6, while GGII isolates possess vtp-1, -5, or -7. This distribution is noncongruent with the phylogeny calculated with the multiple loci used to characterize the GGI-GGII divergence (4). Consequently, using the vtp locus as a qPCR target limits the selection of GGI-GGII spirochetes for mixed-infection analysis to those with divergent vtp alleles. We surveyed GGI DAH and GGII MTW-2 chromosomal DNA alignments to identify sequences that were divergent between these isolates and identified the bh0260 ORF as a possible GGI-GGII qPCR target. This 1,197-nt ORF showed 95% identity between DAH and MTW-2, but a 36-nt segment retained only 61% identity, allowing oligonucleotide design for qPCR differentiation of GGI DAH and GGII MTW-2 alleles. In qPCR amplification with bh0260 primer-probe oligonucleotides on DNA from DAH, MTW-2, and other GGI and GGII isolates, we concluded that there was conservation of the divergent DAH and MTW-2 target sequences within the respective GGI-GGII clades. Specificity and sensitivity limits of qPCR for GGI and GGII bh0260 alleles are provided in Materials and Methods. The correlation of the spirochete number determined by qPCR for rrs and the bh0260 targets of either GGI or GGII isolates was highly significant (r = 0.965; P = 0.00003) compared to the correlation between rrs and vtp-6 or vtp-5 alleles of the same isolates (r = 0.232; P = 0.616). We therefore used the results of the qPCR assay for the bh0260 alleles to identify relative numbers of GGI and GGII spirochetes in tick and mouse DNA samples.
Tick acquisition of GGI and GGII isolates from infected mice.As indicated above, we examined spirochete transmission by feeding superinfected ticks on naive mice and then quantified the density of GGI and GGII B. hermsii in the ticks and in the recipient mice. Table 1 provides values for the spirochete densities in mouse blood prior to primary- and secondary-acquisition feedings for the eight superinfected tick groups generated in the course of this study. In six of these groups, densities were not significantly different between the two feedings. In two of the groups, MTW-2–DAH and LAK-4–LAK-2, spirochete densities were significantly different. In both cases, the GGII isolate was at higher density than the GGI isolate. Table 1 also shows the time elapsed between tick primary and secondary acquisitions and between secondary acquisition and transmission feedings. Ticks from the last four groups were used either for transmission feeding at 6 to 10 months after the secondary acquisition or for serial-transmission feeding at 9 to 15 months and at 12 to 18 months after the secondary acquisition (see below). The time intervals between feedings were adequate for developmental progression of ticks, including molt (18), which increases the probability of full engorgement at each feeding.
Spirochete transmission to mice at three ticks per mouse.Five months after the secondary-acquisition feeding, the GGI DAH- and GGII MTW-2-exposed and the GGI MIL- and GGII YOR-exposed tick cohorts were fed in groups of three ticks per naive mouse, with six mice per group, to measure the efficiency of transmission by qPCR analysis of GGI and GGII densities in blood. Two weeks after feeding, tick midguts were dissected for qPCR analysis of GGI and GGII spirochete densities. Tick salivary glands were recovered to enumerate spirochetes by the DFA test with monoclonal antibodies to Vtp-6 (GGI DAH and MIL) and Vtp-5 (GGII MTW-2 and YOR). Some salivary glands were not recovered for DFA, and these tick-mouse sets were eliminated from acquisition-transmission analyses. To estimate the potential GGI and GGII spirochete exposure per recipient mouse at the time of tick feeding, we calculated the sum of the three midgut GGI qPCR values, the three GGII qPCR values, the three salivary gland pair Vtp-6 values, and the three salivary gland pair Vtp-5 DFA values for each set of three ticks that had fed on a mouse. These exposure estimates for each mouse per group were then averaged as four values to calculate the mean GGI and GGII spirochete densities per tick group (Fig. 1). The qPCR spirochete densities per milliliter of mouse blood were the simple means of the data for the recipient mice in each group.
GGI and GGII B. hermsii in tick midgut, salivary gland, and mouse blood following transmission by tick bite with three ticks per mouse. Ticks fed at primary and secondary acquisitions on infected mice were fed in groups of three per mouse on naive recipient mice at 5 months after the secondary-acquisition feed. The values represent qPCR-detected GGI and GGII B. hermsii in DNA from tick midguts dissected 2 weeks posttransmission and blood collected during initial spirochetemia after tick feeding (qPCR; means and standard deviations [SD]) and DFA-detected Vtp-6-positive GGI and Vtp-5-positive GGII B. hermsii in pairs of tick salivary glands that were dissected 2 weeks posttransmission (DFA; means and SD). Tick mean values were calculated from the total for each set of three midguts or salivary gland pairs per mouse. The numbers of tick-mouse samples per transmission assay were as follows: (A) DAH–MTW-2, n = 5, and MTW-2–DAH, n = 4; (B) MIL-YOR, n = 6, and YOR-MIL, n = 5. The brackets indicate significant differences between GGI and GGII values by unpaired t test: *, P = 0.05 to 0.01; **, P = 0.01 to 0.001; ***, P < 0.001.
The DAH–MTW-2 acquisition and transmission results (Fig. 1A) show that primary GGI DAH colonization of ticks did not preclude secondary GGII MTW-2 colonization. The individual midgut qPCR results indicated that 14 of the 18 ticks were infected with GGI DAH and 17 of the 18 were infected with GGII MTW-2; 13 of the 18 ticks were superinfected with both isolates. Tick midguts had significantly higher GGII than GGI spirochete densities. Salivary glands were equally infected with both GGI and GGII populations, as determined by DFA. Mice infected by these ticks showed significantly higher GGII MTW-2 than GGI DAH spirochete densities in blood. The reciprocal GGII MTW-2–GGI DAH results (Fig. 1A) show that tick primary GGII infection did not prevent secondary GGI acquisition. Of the 18 ticks fed, 17 midguts were infected with GGII MTW-2 and 13 were infected with GGI DAH; 13 of 18 ticks were superinfected with both isolates. The MTW-2–DAH tick midguts showed similar GGI-GGII spirochete densities, but in the salivary glands, GGII MTW-2 was more abundant than GGI DAH. In mouse blood, GGII MTW-2 was detected at >106 spirochetes/ml, but no GGI DAH was observed. However, these MTW-2–DAH results are qualified by the higher density of MTW-2 in blood at primary acquisition than of DAH at secondary acquisition (Table 1). In assays of parallel transmission to the dual-acquisition ticks, the ticks infected with only one isolate (see Materials and Methods) were also fed in groups of three per mouse with two mice per tick group. All mice in these groups showed spirochetemias, with the expected GGI or GGII qPCR product in tick midgut and mouse blood DNA. The spirochetemia densities in mouse blood were equivalent to or greater than those observed after feeding doubly infected ticks (data not shown).
A second superinfection experiment compared GGI MIL and GGII YOR spirochete acquisition and transmission. MIL-YOR tick midguts and salivary glands showed equivalent GGI-GGII spirochete densities (Fig. 1B). The 18 ticks that fed on recipient mice included 17 infected with GGI MIL and 18 infected with GGII YOR; 17 of 18 ticks were superinfected. Upon transmission to mice, the GGI MIL spirochete density was significantly higher in mouse blood than the GGII YOR density. In GGII YOR-GGI MIL tick midguts, YOR density appeared higher than that of MIL, but it was not statistically significant (unpaired t test, P = 0.1474), likely due to the large standard deviation in the YOR data. In the salivary gland DFA test, GGII YOR showed significantly greater density than GGI MIL. Seventeen of the 18 YOR-MIL ticks fed on mice were infected by GGII YOR, and 18 ticks were infected by GGI MIL; 17 of the 18 ticks were superinfected. Transmission by YOR-MIL ticks to mice resulted in GGII YOR at approximately 106 spirochetes/ml, while the GGI MIL was not detected. This result was similar to the MTW-2–DAH results in Fig. 1A; in the YOR-MIL experiment, YOR and MIL spirochete densities were equivalent at primary and secondary spirochete acquisition (Table 1). The control singly infected ticks were tested as described above for transmission to mice. Each mouse became spirochetemic, and the expected GGI or GGII target was detected in a qPCR assay of the tick midgut and mouse blood DNA. As in the first experiment, the spirochetemias after transmission from singly infected ticks were greater than or equivalent to those seen after feeding doubly infected tick groups (data not shown).
Detection of spirochetes by immunofluorescence microscopy.The two salivary gland preparations in Fig. 2 show that sequential GGI and GGII acquisition resulted in a pattern of nonhomogeneous infection of individual type II granular acini, the lobular organelles largely comprised of secretory cells surrounding a central lumen (19, 20). Spirochetes formed aggregates of varying density, but their localization relative to tick cells was not identified. A portion of a GGI MIL-GGII YOR-infected gland in Fig. 2A shows Vtp-6-positive MIL infection of multiple acini and Vtp-5-positive YOR infection of separate acini, with coinfection of several acini by both types of spirochetes. Figure 2B shows a portion of a salivary gland from a GGII YOR-GGI MIL-infected tick, with Vtp-5-positive YOR localized to one acinus and Vtp-6-positive MIL localized to separate acini. Similar distributions of GGI and GGII spirochetes to granular acini was observed in the ticks infected with MTW-2 and DAH. In all the posttransmission salivary glands examined, spirochetes infected only granular acini and were not observed in agranular acini.
DFA of B. hermsii in the salivary glands of infected O. hermsi. Fixed tissue was incubated with monoclonal antibodies H4825 to Vtp-6-Alexa 568 conjugate (red channel) and H3548 to Vtp-5-Alexa 488 conjugate (green channel) and viewed at ×200 magnification. Salivary gland duct and tracheal branches (blue channel) were detected by autofluorescence. (A) Gland from tick with primary GGI MIL (Vtp-6; red) and secondary GGII YOR (Vtp-5; green) infections. (B) Gland from tick with primary GGII YOR (Vtp-5; green) and secondary GGI MIL (Vtp-6; red) infections. Scale bar in panel B, 20 μm.
Spirochete transmission to mice with a single tick per mouse.In the above-described experiments, three ticks were fed on each recipient mouse. The use of multiple vectors per host may obscure the relationship between spirochete densities in the tick and the infectious dose observed in the mouse blood. We therefore examined mouse infection after single ticks had fed, using additional GGI-GGII pairs to determine if the apparent dominance of GGII in transmission to mice was observed in additional spirochete isolates. To address the possibility that the vtp-5 allele possessed by both GGII MTW-2 and YOR could enhance fitness in the tick relative to GGI isolates with vtp-6, we first compared GGI BYM and GGII LAK-2 isolates, which both possess the vtp-1 allele. A single vtp allele in both primary and secondary acquisitions would normalize unspecified tick vector-spirochete interactions prior to mouse transmission.
Ticks were infected as in the first two experiments, but transmission involved feeding one tick per mouse at 8 to 10 months after the secondary acquisition. DFA testing of salivary glands was not performed, as reagents for GGI-GGII serological differentiation were not available. The primary- and secondary-acquisition spirochetemias in mouse blood used to infect the GGI BYM–GGII LAK-2 ticks were not significantly different (Table 1). The transmission results shown in Fig. 3A included only tick-mouse pairs in which the tick midgut contained both GGI and GGII spirochetes by qPCR, indicating superinfection (five of the six ticks assayed in each group). The BYM–LAK-2 midgut spirochete densities were similar for GGI and GGII in the transmitting ticks, as were GGI and GGII spirochetemia densities in the recipient mice. For the LAK-2–BYM ticks, spirochete densities in mouse blood for the primary- and secondary-acquisition feedings were equivalent (Table 1). The transmission results showed that midgut GGI-GGII spirochete densities were similar (Fig. 3A). However, GGII LAK-2–GGI BYM mouse blood contained only GGII spirochetes, consistent with the GGII-dominant pattern seen above with MTW-2–DAH and YOR-MIL transmission in which the GGII spirochete was acquired first by the ticks. All tick controls infected with only one spirochete type caused spirochetemias in mice, and qPCR on these tick midgut and recipient mouse blood DNA samples detected the GGI or GGII target of the spirochete type with which ticks had been infected. Spirochete densities in mouse blood after single-infection tick feeds were greater than or equivalent to the densities measured after double-infection tick feeds (data not shown).
GGI and GGII B. hermsii in tick midgut and mouse blood following transmission by tick bite with one tick per mouse. Ticks fed at primary and secondary acquisitions on infected mice were fed individually on naive recipient mice at 6 to 10 months after the secondary-acquisition feed. The values represent means and SD for qPCR-detected GGI and GGII spirochetes in DNA from tick midguts dissected 2 weeks posttransmission and blood collected during initial spirochetemia after tick feeding. The numbers of tick-mouse samples per transmission assay were as follows: (A) BYM–LAK-2 and LAK-2–BYM, n = 5 for both groups (both isolates possess the vtp-1 allele). (B) LAK-4–LAK-2 and LAK-2–LAK-4, n = 5 for both groups (both isolates were from a single focus of endemicity). The brackets indicate significant differences between GGI and GGII values by unpaired t test: *, P = 0.01 to 0.05.
We performed a second experiment with dually infected GGI LAK-4 and GGII LAK-2 ticks, which were isolated from the same focus of endemicity in western Montana (4, 5). LAK-4 has the vtp-6 allele (Schwan et al., unpublished), while LAK-2 contains the vtp-1 allele; there are no vtp allele-matched GGI-GGII isolates in our collection from a single focus. We again sequentially infected ticks with the GGI-GGII pair and tested single ticks at 6 to 10 months after the secondary-acquisition feed for transmission to mice. The LAK-4–LAK-2 acquisition data (Table 1) indicated that the primary GGI LAK-4 density was significantly lower than the GGII LAK-2 density at secondary acquisition, but the LAK-2 primary-acquisition–LAK-4 secondary-acquisition densities showed no significant difference. The posttransmission tick midgut qPCR evaluations showed that one of the six transmitting ticks in each group was not superinfected, reducing the tick-mouse sample size to five per group. Tick midgut GGI and GGII spirochetes achieved similar densities in both the LAK-4–LAK-2 and the LAK-2–LAK-4 groups (Fig. 3B). Spirochetemic blood from LAK-4–LAK-2 mice also showed similar GGI-GGII spirochete densities. In the GGII LAK-2–GGI LAK-4 mouse blood, however, only GGII spirochetes were detected. Singly infected control ticks were successfully transmitting to mice, as evidenced by similar or higher spirochetemia densities in mouse blood than with doubly infected ticks and by evidence of expected GGI or GGII spirochetes in qPCR of tick midgut and mouse blood DNA (data not shown).
Spirochete transmission to mice upon repeated feeding by single ticks.As the primary acquisition of the GGII isolates by ticks appeared to inhibit transmission of secondary-acquisition GGI isolates at 5 to 10 months postacquisition, we investigated if this pattern occurred at longer intervals by examining two rounds of transmission feedings by single ticks on naive mice. For these experiments, we used ticks remaining from the cohorts mentioned above that were superinfected with BYM and LAK-2 and with LAK-4 and LAK-2. These ticks were assessed for first transmission at 9 to 15 months after acquisition feeding was complete (Table 1); the first-transmission feedings occurred on two separate occasions due to inadequate recovery of engorged ticks on first attempted feedings. All second-transmission feedings were performed at uniform intervals of 3 months after the first-transmission dates, at 12 to 18 months after acquisition feeding. One month after the second transmission, ticks were dissected for isolation of DNA from the midgut. All eight of the GGI BYM–GGII LAK-2 ticks and four of the eight GGII LAK-2–GGI BYM ticks were superinfected when examined after second transmission. In the LAK-4 and LAK-2 tick cohorts, seven of the eight GGI LAK-4–GGII LAK-2 ticks and all six of the GGII LAK-2–GGI LAK-4 ticks were superinfected after the second transmission. As in the above-described experiments, mouse-tick data were excluded for those ticks infected with only one isolate.
The GGI BYM–GGII LAK-2 ticks showed similar midgut GGI and GGII spirochete densities after second transmission to mice, and mouse blood showed that these ticks transmitted both spirochete types at the two feedings (Fig. 4A), a result similar to transmission by this tick cohort at 10 months postacquisition (Fig. 3A). In the GGII LAK-2–GGI BYM ticks, midgut spirochete densities were similar for the GGI and GGII isolates. The first and second transmissions resulted in GGI and GGII spirochetes in mouse blood at similar densities (Fig. 4A), in contrast to results at 8 months postacquisition, when GGI BYM spirochetes were not detected in mouse blood (Fig. 3A).
GGI and GGII B. hermsii in tick midgut and mouse blood following serial transmission to naive mice by tick bite with one tick per mouse. Ticks fed at primary and secondary acquisitions on infected mice were fed individually on naive recipient mice at 9 to 15 months after the secondary-acquisition feed, held for 3 months, and refed on naive mice. The values represent means and SD for qPCR-detected GGI and GGII spirochetes in DNA from tick midguts dissected 1 month after the second transmission (trans.) and from blood collected during initial spirochetemia posttransmission. The numbers of tick-mouse samples per transmission assay were as follows: (A) BYM–LAK-2, n = 7, and LAK-2–BYM, n = 4; (B) LAK-4–LAK-2, n = 7, and LAK-2–LAK-4, n = 6. The brackets indicate significant differences between GGI and GGII values by unpaired t test: *, P = 0.05 to 0.01; **, P = 0.01 to 0.001; ***, P < 0.001.
The GGI LAK-4–GGII LAK-2 ticks showed equivalent midgut spirochete densities (Fig. 4B), but both the first- and second-transmission events resulted in significantly higher GGII LAK-2 than GGI LAK-4 values in mouse blood, contrasting with results obtained at the 10-month interval (Fig. 3B), in which LAK-4 and LAK-2 spirochetes were present at similar densities in blood. The LAK-2–LAK-4 ticks, with similar midgut GGI and GGII spirochete densities (Fig. 4B), produced a GGII LAK-2 spirochetemia in mice at first-transmission feeding but no detectable GGI LAK-4 spirochetemia, as seen in the 6-month-transmission results with this tick cohort (Fig. 3B). After the second feeding, LAK-4 spirochetes were detected in mouse blood, but at significantly lower density than LAK-2 (Fig. 4B).
GGI and GGII initial spirochetemia by needle inoculation.Transmission from ticks superinfected with GGI and GGII spirochetes in the above-described experiments often resulted in significantly greater GGII spirochete densities in mouse blood. The GGII isolates may have had a higher density at tick transmission, or they may have replicated at a higher rate in the host, resulting in higher density during initial spirochetemia. In the absence of a method to quantify the spirochete inoculum transmitted by superinfected ticks, we artificially infected mice with GGI, GGII, or combined GGI-GGII inocula and then sampled blood during the initial spirochetemias to compare densities. To obtain spirochete populations with similar passage histories (21), two isolates each from GGI and GGII were used at second passage from infected mouse blood, as described in Materials and Methods. In preliminary GGI DAH and GGII MTW-2 needle inoculations, we found that 1 × 103 spirochetes of either isolate caused onset of spirochetemia (detected by dark-field microscopy) at 4 days postinjection. This observation is within the range of spirochetemia onset when mice are infected by tick bite. We therefore inoculated five mice per group with 1 × 103 GGI DAH spirochetes, 1 × 103 GGII MTW-2 spirochetes, or 1 × 103 spirochetes of each. GGI MIL and GGII YOR were similarly inoculated in a second series of mice. Mouse blood was collected for DNA extraction at days 3 through 6 postinoculation and assayed by qPCR to obtain relative GGI and GGII spirochete densities (Fig. 5 and 6). The GGI and GGII results are displayed in panels A and B, respectively, of each figure to clarify the relative spirochete densities under single-inoculation and coinoculation conditions. Infection with single isolates caused increased spirochete density from day 3 to a maximum on day 4 or 5. The GGII MTW-2 isolate reached a higher density in mouse blood on days 3 and 4 than GGI DAH, and GGII YOR reached a slightly higher density than GGI MIL. However, we did not measure the growth rates of these isolates in blood.
GGI and GGII B. hermsii in mouse blood during initial spirochetemia following single- and dual-needle inoculation with DAH and MTW-2. Cultured DAH or MTW-2 spirochetes (1 × 103) were injected intravenously on day 0 postinoculation (p.i.) in a final volume of 0.1 ml for individual inoculations; coinoculations (DAH + MTW-2) consisted of 1 × 103 DAH and 1 × 103 MTW-2 spirochetes in 0.1 ml. The values represent means ± SD of qPCR-detected GGI (A) and GGII (B) spirochetes in DNA from blood collected during spirochetemia on days 3 to 6 postinoculation (n = 5 for each group). No GGII spirochetes were detected in mice inoculated with DAH, and no GGI spirochetes were detected in mice inoculated with MTW-2; thus, these data are not plotted. Three DAH–MTW-2 and four MTW-2 mice were euthanized after blood collection on day 5. Significant differences between single and coinoculated values by unpaired t test are indicated: *, P = 0.01 to 0.05. n.t., no t test on day 6 for a single MTW-2 mouse.
GGI and GGII B. hermsii in mouse blood during initial spirochetemia following single- and dual-needle inoculation with MIL and YOR. Cultured MIL or YOR spirochetes (1 × 103) were injected intravenously on day 0 p.i. in a final volume of 0.1 ml for individual inoculations; coinoculations (MIL + YOR) consisted of 1 × 103 MIL and 1 × 103 YOR spirochetes in 0.1 ml. No GGII spirochetes were detected in mice inoculated with MIL, and no GGI spirochetes were detected in mice inoculated with YOR; thus, these data are not plotted. The values represent means ± SD of qPCR-detected GGI (A) and GGII (B) spirochetes in DNA from blood collected during spirochetemia on days 3 to 6 postinoculation (n = 5 for each group). Note that the density range differs from that in Fig. 5. Significant differences between single and coinoculated values by unpaired t test are indicated: *, P = 0.05 to 0.01; **, P = 0.01 to 0.001.
When coinoculated with MTW-2, DAH densities on days 3 through 5 were significantly less than with DAH inoculated alone (Fig. 5A). MTW-2 density reached a higher spirochetemia in coinoculated mice on day 4 (Fig. 5B). At day 4 postinoculation, mice in the MTW-2 and DAH–MTW-2 groups showed signs of morbidity, with four of the five in the MTW-2 group and three of the five in the DAH–MTW-2 group requiring euthanasia on day 5. The Mantel-Cox log-rank survival test indicated significant mortality in the DAH–MTW-2 (P = 0.0495) and MTW-2 (P = 0.0143) groups compared to the group of mice inoculated with DAH alone. Due to this morbidity, no statistical test was performed between single and coinoculation GGII spirochete densities on day 6 (Fig. 5B).
When GGI MIL was coinoculated with GGII YOR, MIL spirochete density was significantly less on days 4 through 6 than with single inoculation of MIL (Fig. 6A). In the MIL-YOR-coinoculated mice, YOR density was significantly reduced on day 5 compared to spirochete densities with YOR single inoculation (Fig. 6B). After MIL and YOR single inoculation or coinoculation, there were no signs of morbidity during spirochetemia.
DISCUSSION
In examining B. hermsii competency for tick superinfection, we found that spirochetes in an infected blood meal can successfully superinfect the midgut of a previously infected tick. The first series of experiments involved transmission by three ticks to each recipient mouse. Of the 72 ticks tested, 60 were superinfected with the GGI and GGII isolates to which they had been exposed. In subsequent experiments with one tick per mouse, we reported only the results of transmission by ticks that were superinfected. Nevertheless, these ticks were the majority of the total in their cohorts, with 20 out of 24 superinfected in the single-transmission series and 25 out of 30 total in the double-transmission series. O. hermsi acquisition of relapsing fever spirochetes is not ensured upon exposure to infected hosts (22), and the present data show that cofeeding ticks on an infected host does not ensure uniform acquisition. Spirochete failure to colonize ticks occurred at both primary and secondary feedings on infected mice and may have been due to the lack of full engorgement. These rates of superinfection appear to be much higher than those achieved with artificial infection of O. hermsi, where we observed infection of 5 first-stage nymphs out of 15 that were immersed as larvae in a suspension of GGI DAH and GGII MTW-2 (11). In that report, we verified spirochete infection of ticks by immunofluorescence after the transmission feed but did not directly assay for superinfection in the ticks. Superinfection of three or more of the five infected nymphs was inferred by detection of vtp-6 and vtp-5 sequences in blood DNA from three out of six mice, after feeding two or three ticks per mouse. Thus, acquisition by immersion in spirochete suspensions and acquisition by feeding on infected mice result in dually infected ticks that can transmit the dual spirochete infection to mice. Tick acquisition of dual infection by feeding on coinfected mice, which we have not attempted, would also provide useful information on GGI-GGII competition in the enzootic cycle.
As an obligate parasite, B. hermsii persists by proliferating in susceptible hosts, increasing the probability of acquisition by tick vectors and of transmission to new vertebrate hosts. We measured spirochete efficiency of transmission by evaluating the density of the primary and secondary spirochete populations in the tick vector and in mouse blood after transmission. It was apparent from the 5- to 10-month short-interval transmissions that a GGII primary tick infection interfered with transmission to mice of a GGI secondary tick infection, but not the converse. This interference was first seen with GGII MTW-2 and YOR spirochetes as primary tick acquisitions, when they blocked the transmission of GGI secondary acquisitions (Fig. 1). This blocking was observed with ticks that had significantly higher GGII densities in the salivary glands than the secondary GGI isolates. The GGII MTW-2 isolate, in contrast to the GGII YOR, achieved a dominant population during mouse infection whether it was the primary or secondary spirochete acquisition by ticks. These results suggested greater transmission efficiency by GGII isolates but were qualified by the use of multiple ticks for each transmission. This heterogeneity of GGI-GGII tick transmission sources at the start of mouse infection could result in a greater range of infection outcomes. The single-tick, short-interval transmission data (Fig. 3), however, reiterated the finding that GGII spirochetes could interfere with tick transmission of a secondary GGI spirochete population. The densities of GGI and GGII spirochetes in tick midguts did not reveal primary- versus secondary-population differences, but spirochetemias in mouse blood showed the dominance of the primary GGII populations upon transmission. The choice of GGI BYM and GGII LAK-2 as isolates with the same vtp-1 allele gave a transmission outcome similar to those with isolates GGI DAH and GGII MTW-2 or GGI MIL and GGII YOR bearing vtp-6 and vtp-5 alleles, suggesting vtp alleles are not a significant factor in transmission efficiency. The GGI LAK-4 (vtp-6) and GGII LAK-2 (vtp-1) isolates from a single focus of endemicity (2, 4, 5) repeated the pattern of a GGII (LAK-2) primary acquisition interfering with transmission with a GGI (LAK-4) secondary acquisition. To provide further evidence of this interference, we verified that ticks that had acquired GGI spirochetes only at the secondary-acquisition feedings were competent to transmit these isolates. Thus, when GGII spirochetes were the primary acquisition, they blocked transmission of GGI secondary acquisitions at 5 to 10 months.
The GGII interference with productive infection of recipient mice by GGI spirochetes did not persist, because longer, 11- to 14-month intervals between acquisition and transmission allowed secondary GGI spirochetes to infect mice. Using the same GGI BYM–GGII LAK-2 and GGI LAK-4–GGII LAK-2 cohorts of superinfected ticks, we allowed individual ticks to sequentially feed on two naive mice and found that the primary GGII LAK-2 tick infections did not prevent transmission of either secondary GGI BYM or GGI LAK-4 spirochetes (Fig. 4). The extended time interval between the secondary acquisition and the first transmission feed, and the refeeding of individual ticks, resulted in detectable cotransmission of both spirochete populations. GGI isolates, as a secondary tick acquisition, may require an extended period to migrate to the salivary glands and successfully compete with GGII spirochetes for transmission at tick feeding. In nest-dwelling soft ticks, such as O. hermsi, intervals between feeding depend on host behavior. B. hermsii transmission may depend on the availability of naive hosts, as seroconversion provides broad immunity to reinfection (23). Cumulatively, the short-interval and long-interval serial tick-feeding tests show that the transmission efficiency of the second acquisition, either GGI or GGII, is positively correlated with incubation time in the vector and possibly with the number of transmission feedings.
All measurements of GGI-GGII spirochete densities in the ticks were by necessity taken after the ticks had transmitted spirochetes to mice. These data allowed us to identify whether one or both isolates were in the vector but could not indicate the density or distribution of spirochetes at the time of transmission. The relative GGI-GGII spirochete densities in tick midguts did not show a strong correlation with densities in recipient mouse blood (Fig. 1, 3, and 4), but the density (Fig. 1) and distribution (Fig. 2) in salivary glands suggest some indications of transmission outcomes that need further study. In our approach to superinfections with GGI and GGII spirochetes, the tick group size allocated to feed on a single spirochetemic mouse was limited. The tick nymphs that engorged at the primary-acquisition feeding had to molt, feed on a second spirochetemic mouse, and molt again before they were tested for transmission. The limited size of infected tick cohorts did not allow us to obtain detailed records on changes in spirochete density and distribution within ticks prior to transmission. A report on antigenic switching by the B. hermsii GGI DAH isolate indicated that spirochetes are detectable in salivary glands as early as 33 days after acquisition (17), and a later report on the same isolate found that if the density of spirochetes at the time of acquisition by second-stage nymphs was ≥106 spirochetes/ml, the transmission efficiency was 100% at the time of the next feeding, but the interval between acquisition and transmission feeds was not reported (22). To assess spirochete populations in the midgut and salivary glands at time points prior to transmission feeding would require a large cohort of nymphs at the outset to fully understand transmission dynamics. Earlier observations of dissemination of Borrelia duttonii in Ornithodorus moubata (24) indicate that the salivary glands are only one of several organs that are invaded by this species of relapsing fever spirochete, as the ovaries, central ganglion, malpighian tubules, and coxal organs are also infected. The pattern of O. hermsi salivary gland infection observed in Fig. 2, in which GGI and GGII spirochetes primarily colonized separate granular acini and were never observed in agranular acini, suggests the latter are inaccessible to spirochetes or do not support their growth. A more precise model of B. hermsii infection of salivary glands has to correlate patterns of spirochete localization and release with the secretory functions of acini, as initiated with studies in Ornithodoros savignyi (25).
After passage from the tick salivary glands to the feeding lesion, relapsing fever spirochetes must replicate in host blood to be acquired by new tick vectors. To model GGI-GGII competition during the first spirochetemia, we assayed spirochete densities in mouse blood after needle inoculation. Densities of the GGI DAH and MIL isolates declined significantly in the blood when coinoculated with GGII MTW-2 and YOR (Fig. 5 and 6). These results are consistent with attenuation of GGI spirochetes in the blood in the presence of GGII isolates, with implications for successful transmission from infected ticks. However, GGII YOR density also decreased on day 5 when coinoculated with MIL, but GGII MTW-2 showed a significant increase on day 4 when coinjected with GGI DAH. The GGII isolates therefore tend to depress GGI density in mixed-blood spirochetemias, while GGII MTW-2 density appears to be enhanced by the presence of a GGI isolate. As a possible factor affecting spirochete growth in blood, all GGI isolates examined to date possess a pseudogene at the adenine deaminase-coding ade locus, whereas GGII isolates carry an intact ade ORF (26). If the GGI mutation eliminates adenine conversion to hypoxanthine, in the context of the essential purine salvage mechanisms utilized by Borrelia species (27, 28), it may have a significant effect on GGI-GGII spirochete competition in both vertebrate host blood and tick vector tissues. Targeted mutagenesis of the ade locus in GGII isolates or allelic replacement in GGI isolates by available methods (29) may reveal whether the gene function contributes to the apparent GGII suppression of GGI spirochetes in mice coinfected with GGII spirochetes.
The MTW-2 isolate may have one or more additional variant loci not shared with other GGII isolates. In the tick acquisition and transmission data (Fig. 1 and 3), this isolate differs from YOR and LAK-2 isolates in that it is the dominant isolate after transmission, even if it is secondarily acquired by ticks. After needle inoculation of mice, MTW-2 caused morbidity at an inoculum at which YOR did not. Further coinoculation experiments using a single GGI isolate as a competing population are needed to verify differential MTW-2–YOR or MTW-2–LAK-2 virulence. With this evidence, comparison of genomic DNA sequences between MTW-2 and the other GGII isolates may lead to identification of loci involved in relative virulence.
Several decades of isolating GGI and GGII B. hermsii from sites and clinical specimens in western North America indicate that both genotypes are maintained in enzootic cycles (2). The phenotypic differences we observed here suggest that GGII organisms have a replication rate advantage over GGI spirochetes that would increase their prevalence, but this would only occur if short-interval (<1 year between nymphal feedings) transmission were frequent. If long-interval equilibration of coinfecting GGI and GGII populations in ticks were prevalent, it seems possible that GGI and GGII spirochetes could maintain similar frequencies of infection in the tick vector. Modeling of parasite infection rates by May and Nowak (30, 31) indicates that when multiple strains are capable of coinfecting a given host, the strains are clustered at maximum virulence phenotypes. In the present study, such a strain may be represented by the GGII isolate MTW-2, which caused morbidity in mice. In contrast, a pattern of single-strain host infection suggested to them a broader range of virulence phenotypes, including strains that cause lower host mortality and thereby increase acquisition by the vector. Such a pattern may be represented in this work by the GGI isolates and possibly most GGII isolates. We have shown here that the pathogen vector can be superinfected with different genotypes of B. hermsii. We have also shown that superinfected ticks can cotransmit spirochetes with mixed genotypes to a laboratory mouse host. In enzootic cycles, each wild rodent host species may respond to the vector and its spirochete burden in distinctive patterns. If GGI isolates are found at prevalences approximately equal to those of faster-replicating GGII isolates, applying the May-Nowak model to the vector would predict that most ticks contain a single spirochete strain. Thus, seroprevalence surveys that differentiate between GGI and GGII exposure in rodent hosts, possibly based on recognition of factor H-binding protein isoforms A1 and A2 (32, 33), would lead to a better understanding of GGI-GGII spirochete distribution in foci of endemicity. If foci exist where GGII exposure is dominant, the model suggests this may be correlated with an increased availability of naive hosts in which higher virulence is necessary for pathogen survival.
ACKNOWLEDGMENTS
We gratefully acknowledge Merry Schrumpf and Dave Mead for Alexa Fluor conjugates of the monoclonal antibodies; Anita Mora for assistance with the figures; Tammi Johnson for discussions on infectious disease modeling; and Sarah Ward, Dan Dulebohn, and Clayton Jarrett for critical reading of the manuscript. We also thank the reviewers for their contributions.
This project was supported by the Division of Intramural Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD.
FOOTNOTES
- Received 30 April 2013.
- Returned for modification 13 May 2013.
- Accepted 22 May 2013.
- Accepted manuscript posted online 28 May 2013.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
The authors have paid a fee to allow immediate free access to this article.
