Transplacental transmission of tick-borne Babesia microti in its natural host Peromyscus leucopus

Background Babesia microti is an emerging tick-borne pathogen and the causative agent of human babesiosis. Mathematical modeling of the reproductive rate of B. microti indicates that it cannot persist in nature by horizontal tick-host transmission alone. We hypothesized that transplacental transmission in the reservoir population contributes to B. microti persistence and emergence in North American rodent populations. Methods Peromyscus leucopus were collected from Connecticut and Block Island, Rhode Island and analyzed using a highly specific quantitative PCR (qPCR) assay for infection with B. microti. Results In April, 100% (n = 103) of mice were infected with B. microti. Females exhibited significantly higher parasitemia than their offspring (P < 0.0001) and transplacental transmission was observed in 74.2% of embryos (n = 89). Transplacental transmission of B. microti is thus a viable and potentially important infectious pathway in naturally infected rodent species and should be considered in future theoretical and empirical studies. Conclusions To our knowledge, this study is the first to report transplacental transmission of B. microti occurring in its natural reservoir host, P. leucopus, in the United States and the only study that provides a quantitative estimate of parasitemia. This vector-independent pathway could contribute to the increased geographic range of B. microti or increase its abundance in endemic areas. Electronic supplementary material The online version of this article (10.1186/s13071-018-2875-8) contains supplementary material, which is available to authorized users.


Background
Babesia microti (Apicoplexa: Sporozoea) is a zoonotic intraerythrocytic apicomplexan parasite and is responsible for almost all cases of human babesiosis in the United States [1]. Human babesiosis is an emerging tickborne disease that shares the same vector, the blacklegged tick (Ixodes scapularis), and dominant reservoir host, the white-footed mouse (Peromyscus leucopus), as the causative agent of Lyme disease, the spirochete Borrelia burgdorferi [2]. During the last 20 years, human babesiosis has spread in the United States [3,4] following a similar trajectory to that of Lyme disease, although with a temporal lag [5][6][7]. Factors accounting for the delayed spread of babesiosis compared to Lyme disease include lower fitness in the enzootic cycle because of lower transmission from infected host to tick and lower trans-stadial transmission, greater asymptomatic infection in humans, insufficient physician awareness, and underdiagnosis in non-or newly-endemic areas [1,5,6].
An integrated measure of B. microti fitness (the basic reproductive number, R 0 ) was estimated to be lower than the threshold for pathogen persistence (R 0 < 1) under ecological conditions identified in long-term field studies [6,8] implying that emergence of this pathogen should be unlikely in nature. However, despite the low predicted R 0 , B. microti has not only persisted in multiple locations with high infection prevalence in ticks regionally [5,[9][10][11], but it is also geographically expanding in the Northeast and upper Midwest regions of the USA [6]. This paradoxical finding suggests additional mechanism(s) enhancing B. microti transmission and persistence in the enzootic cycle may be occurring. Previous field studies reported higher average infection prevalences of B. burgdorferi (22.37%) compared to B. microti (9.53%) in nymphal I. scapularis ticks throughout the New England area [6]. However, very little is known about the B. microti infection status of P. leucopus in nature.
The aims of this research were to (i) determine if transplacental transmission of B. microti occurs in naturally infected, wild reservoir P. leucopus mice and (ii) quantify the level of transplacental transmission of host populations in B. microti-endemic coastal New England. As a comparison for early season infection prevalence we also screened for the presence of B. burgdorferi in the same host population; B. burgdorferi is not known to be transmitted transplacentally [12,13]. Animals were removed from traps, morphological characteristics (age, sex, weight, body measurements, etc.) were collected, and attached larval and nymphal ticks were counted and removed from the ears and body. All animals were euthanized with an overdose of isoflurane and necropsied immediately in the field. All organs and tissues were preserved in liquid nitrogen, blood samples were dried on Whatman FTA cards (Fisher Scientific, Pittsburg, PA, USA), and the carcasses were wrapped in Whirl-Pak bags (Whirl-Pak®, Nasco, Fort Atkinson, WI, USA) and frozen in liquid nitrogen; embryos of pregnant females remained within the carcass. All samples were transferred to -80°C for long term storage upon return to the laboratory. All animal procedures were in accordance with guidelines approved by the Columbia University Institutional Animal Care and Use Committee (IACUC no. AC-AAAL3656).

Embryo necropsy
To ensure no cross contamination of maternal blood occurred during the necropsy of embryos from the female's body cavity, a new pair of autoclaved dissecting utensils were used for each adult female. Utensils were soaked in a 10% bleach solution and then in 70% ethanol for at least 2 min between sampling embryos from the same adult female. All embryos were removed from the body cavity of the female, washed with a saline solution, photographed, immediately cut into sections around each embryo, and then transferred to a separate sterile Petri dish. Embryos were found in various stages of development. Because the gestation time of P. leucopus is 22-28 days, embryos were separated into three developmental age categories (Week 1, Week 2, Week 3; Fig. 1) based on the size and the presence or absence of various developmental characteristics for each embryo (i.e. eye spot, placenta, limbs and tail, etc.). Week 1 embryos were very small with little to no distinguishing characteristics. Week 2 embryos were larger with a small placenta, could be removed from the uterus, and a distinctive eyespot was observed. Week 3 embryos were large with almost fully developed characteristics. A 25 mg sample of the following tissue types was collected: the uterine arteries and veins, the uterus surrounding each individual embryo, the placenta if present, and the embryonic sac surrounding each embryo if available. Utensils were cleaned with bleach and ethanol between Week 1 embryos were very small with little to no distinguishing characteristics. Week 2 embryos were larger and an embryo could be removed from its location in the uterus, a distinctive eye-spot was also observed. Week 3 embryos were large with almost fully developed characteristics each type of tissue collection. If the embryo was early in development the whole embryo was used for DNA extraction or the embryo was cut in half to equal 25 mg. If the embryo was in a later developmental stage the heart of the embryo was removed and used in the extraction process. All embryo samples were screened for the presence of B. microti using the quantitative PCR protocols described in the following section.

DNA extraction and PCRs
For all embryos and embryonic tissues, a 25 mg sample was collected as described in the previous section, DNA was extracted by hand using the Qiagen DNeasy Blood and Tissue kit following the manufacturer's protocol (Qiagen, Valencia, CA, USA), and DNA concentration was analyzed using a spectrophotometer (Denovix Inc, New Castle County, Delaware, USA).
For adult samples, a 3 mm ear punch biopsy was collected from each mouse and DNA was extracted using the QIAcube HT DNA extraction system (Qiagen). DNA concentration was measured for each sample using a spectrophotometer (Denovix Inc). Each ear punch biopsy sample collected from adults was then tested in duplicate for the presence of B. burgdorferi using a quantitative PCR (qPCR) specific for a unique 69 bp segment of the 16S rRNA gene: forward primer (5'-GGC GGC ACA CTT AAC ACG TTA G-3'), reverse primer (5'-GCT GTA AAC GAT GCA CAC TTG GT-3'), probe (6FAM-TTC GGT ACT AAC TTT TAG TTA A-MGBNFQ) [14]. Samples were run on a 7500 real-time PCR system (Applied Biosystems®, ThermoFisher Scientific, Waltham, WA, USA) using TaqMan Fast Advanced chemistry (ThermoFisher Scientific, Waltham, WA) and cycling conditions consisted of: 95°C for 20 s, followed by 40 cycles of 95°C for 3 s and 60°C for 30 s. Because B. burgdorferi is an extracellular bacterium and primarily observed in tissues, we restricted our screening of B. burgdorferi to only adult ear punch biopsy samples.
For adult individuals, DNA from dried blood samples was extracted from FTA cards using the QIAcube HT DNA extraction system or by hand using the Qiagen DNeasy Blood and Tissue kit following the manufacturer's protocol (Qiagen). DNA concentration was analyzed using a spectrophotometer (Denovix Inc). Embryos, embryonic tissues, and adult blood samples, were screened in duplicate for the presence of B. microti using a qPCR designed specifically for detecting a 104 bp section of the 18S rRNA gene of B. microti: forward primer (5'-AAC AGG CAT TCG CCT TGA AT-3'), reverse primer (5'-CCA ACT GCT CCT ATT AAC CAT TAC TCT-3'), probe (6FAM-CTA CAG CAT GGA ATA ATG A-MGBNFQ) [15]. Babesia microti is primarily an intracellular erythrocytic protozoan and therefore we analyzed embryos, embryonic tissues, and adult dried blood samples for the presence of B. microti. Because blood samples could not be obtained from embryos, a 25 mg sample of heart tissue was also collected and analyzed from a subset of adult mice for direct comparison to embryos and embryonic tissues.
Average cycle threshold (CT) and quantity values were collected from each run and mean infection prevalence (number of infected individuals/total number of individuals) was calculated for each location. qPCR standards were constructed by separately cloning the aforementioned targeted regions of B. burgdorferi and B. microti into pUC57-Kan plasmids (GENEWIZ, Inc., South Plainfield, NJ, USA). A dilution series (10 6 -1 copy number dilutions) for each pathogen was developed by combining a single uninfected I. scapularis nymph (courtesy of the CDC) and a known amount of plasmid DNA followed by DNA extraction [14,15]. Quantification and normalization of each sample were completed as previously described [14,16,17]; results are expressed as gene copies per picogram (pg) of total DNA.
Babesia microti positive samples, determined via qPCR, were then subjected to a standard PCR using a specific primer set to amplify 437 bp of the B. microti 18S rRNA gene (forward PIRO-A: 5'-AAT ACC CAA TCC TGA CAC AGG G-3' and reverse PIRO-B: 5'-TTA AAT ACG AAT GCC CCC AAC-3') [18]. Amplification was performed using Platinum Superfi 2× PCR Master Mix (Invitrogen, ThermoFisher Scientific, Waltham, WA, USA) under the following conditions: 98°C for 30 s, followed by 40 cycles of 98°C for 10 s, 60°C for 10 s, and 72°C for 1 min, with a final elongation step of 72°C for 5 min. PCR products were subjected to gel electrophoresis on a 1% agarose gel stained with ethidium bromide. Amplicons that produced bands of the correct size (~450 bp) were submitted for Sanger sequencing (Eurofins, Louisville, KY, USA) in both the forward and reverse directions. Consensus sequences were constructed and aligned with other orthologous B. microti sequences deposited in the GenBank database and previously described as human-infecting Clade 1 and nonhuman-associated Clade 2 and Clade 3 pathogens (AY144696, AY144701, AY144690; respectively) [19]. Alignments were completed by hand using BioEdit Sequencing Alignment Editor [20] and using the online Multiple Sequence Comparison by Log-Expectation Alignment (MUSCLE) program. Sequence diversity and a pairwise distance matrix were constructed using the maximum composite likelihood model in MEGA version 7.0 software [21].

Statistical analyses
Infection prevalence was calculated for sex of adults, embryonic tissue type, and embryonic stage of development. Significant differences between the proportion of individuals infected and not-infected with B. burgdorferi were evaluated using a Fisher's exact test. A Student's ttest was used to compare the number of nymphs collected from April mice, individuals infected with B. burgdorferi and/or B. microti in April mice between males and females, between CT and BI, and to compare female infection to embryo infection overall and between trapping sessions. A nonparametric Kruskal-Wallis test was performed to assess differences in B. microti infection in the stages of embryonic development and the average number of embryos in each developmental stage. A pairwise Wilcoxon test was used to determine the variation between each stage of development. Finally, an analysis of variance (ANOVA) was completed to determine differences in parasitemia between the three stages of embryonic development. Geometric means of duplicate samples were used in all calculations.

Results
All mice collected in April and only pregnant females (n = 6) and their embryos (n = 26) collected in July were necropsied and analyzed for the presence of B. burgdorferi and B. microti using qPCR. In the April sampling session, a total of 63 P. leucopus mice were collected from the Connecticut (CT) locations and 40 from the Block Island (BI) locations (Table 1). Of these, the combined proportion of mice with attached I. scapularis nymphs was 57.89% (n = 32 CT mice, n = 26 BI mice). Overall, 12.62% (CT = 9.53%; BI = 17.5%) of mice were infected with B. burgdorferi and 100% with B. microti. The average B. burgdorferi log mean copy number per pg total DNA (MCN) did not vary significantly between CT and BI (Table 1) or between infected males (n CT = 4; n BI = 5) and females (n CT = 2; n BI = 2). No significant differences in B. microti MCN were observed between CT and BI (Table 1) or between infected males (n = 55; MCN ± standard error = 5.25 ± 0.09) and females (n = 48; MCN = 5.20 ± 0.10).
Of the total small mammals sampled in 2016, 20 P. leucopus females were pregnant (n CT = 12; n BI = 8), including three and five mice collected on BI in April and July, respectively. All 12 pregnant females from CT were collected in April. One pregnant meadow vole (Microtus pennsylvanicus) was collected on BI in July and was included in some of the analyses. Four pregnant mice and the vole tested positive for B. burgdorferi (25.0%). All blood samples collected from adult pregnant females tested positive for the presence of B. microti via qPCR (100%) and no significant differences were observed between CT and BI females. The number of embryos per female (mouse only) ranged from 2-6 with an average of 4.5 embryos per female ( Table 2; Additional file 1: Table S1). A significant difference among the average number of embryos was observed (Kruskal-Wallis χ 2 = 8.13, df = 2, P = 0.0172). The average number of embryos in Week 1 (n = 3.4) was not significantly different from the number of embryos in Week 2 (n = 4.8; Wilcoxon test P = 0.0660), however there were significantly more Week 3 embryos compared to Week 1 (n = 5.1; Wilcoxon test P = 0.0360), but not compared to Week 2 (Wilcoxon test P = 0.4340; Fig. 2a). There were no significant differences observed in the average number of embryos between trapping sessions.
A total of 89 mouse embryos were collected and tested for the presence of B. microti. Week 1 embryos showed the highest infection prevalence (91.67%), whereas Week 3 embryo infection was lowest (56.10%); the average infection prevalence among all embryos was 74.16% ( Table 2). The average B. microti MCN was significantly different among the three stages of embryonic development (Kruskal-Wallis χ 2 = 16.04, df = 2, P = 0.0003). Embryos from Week 1 (MCN = 2.55 ± 0.17) and Week 2 (MCN = 2.49 ± 0.18) displayed significantly higher parasitemia than embryos from the Week 3 stage of development (MCN = 1.64 ± 0.11; Wilcoxon test P = 0.0003 and P = 0.0022, respectively; Fig. 2b), but not between each other (Wilcoxon test P = 0.8948). Other reproductive tissue types (uterus, placenta, embryonic sac) also tested positive for B. microti (Additional file 2: Table S2), but due to the differences in embryo developmental stage not all tissue types could be collected at each of the three stages.

Discussion
Our study demonstrates for the first time that transplacental transmission of B. microti occurs in wild, naturally-infected populations of P. leucopus and M. pennsylvanicus in Connecticut and on Block Island, RI, USA, with 74% transmission efficiency from mother to fetuses. Although studies have previously reported transplacental transmission of Babesia canis canis in canines [22,23], B. microti in BALB/c laboratory mice [24], and humans [25,26] only one other study has reported vertical transmission of B. microti occurring in naturally infected rodent populations [27]. Tolkacz and colleagues [27] reported vertical transmission of B. microti in two See Additional file 1: Table S1 for detailed information of each individual female a b All individuals were infected with B. microti in April, in contrast to only 12.6% of individuals infected with B. burgdorferi. These mice could have been infected by nymphal I. scapularis in the spring or could have survived and maintained the infection through the winter months (chronic infection). Because nymphal I. scapularis were found on 50.8% of mice collected in CT and 65.0% of mice collected on BI, it is possible these mice all became infected by questing nymphs once they emerged from their winter burrows. However, because a higher proportion of nymphal I. scapularis are typically infected with B. burgdorferi (22.37% average infection prevalence) than B. microti (9.53% average infection prevalence) in the New England area (comparing 10 studies and 9335 nymphal I. scapularis samples) [6], the higher B. microti infection prevalence in mice requires an alternative explanation. A combination of vertical transmission and chronic infection with B. microti through the winter months would amplify B. microti even before most ticks emerged from diapause, typically in mid to late May [6], providing a transmission advantage compared to B. burgdorferi. The duration of B. microti infection in naturally infected (via tick vector) P. leucopus is unclear; however, chronic infection has been shown to persist in laboratory P. leucopus for seven months and possibly longer [9,10,28,29]. Chronic infection of B. microti was also observed for approximately five months in naturally infected voles from Russia [30]. In these ways mice infected over the transmission season (April-July) could remain infected until the following spring, when transplacental transmission could further increase early-season transmission. Resistance to reinfection is unknown in Peromyscus; consequently, mice could either become re-infected or B. microti could become reactivated every transmission season resulting in chronic infection and an increase of the pathogen in the population via vertical transmission. Reactivation of certain pathogens has been shown to occur during stressful life events (i.e. migration in birds) [31,32]. Seasonal physiological stress (i.e. cold temperatures during winter months) may allow for reactivation of B. microti and provide an additional explanation for prolonged infection in wild P. leucopus.
The relative timing of insemination and infection in the mother may be important determinants of the survivorship and infection outcome of fetuses [24,33]. The breeding season of P. leucopus in our study region starts around March, while nymphal tick activity typically starts in April or May, although both events are highly dependent on seasonal variation in temperature [34]. Furthermore, peak B. microti parasitemia in P. leucopus occurs 14 days after infection via tick bite [35]. Given the differences in mouse and tick life cycles, females captured in April were more likely inseminated before B. microti infection via ticks. The lower average number of Week 1 compared to Week 3 embryos may be explained by the higher likelihood of reabsorption of some or all embryos when the female is inseminated simultaneously or before infection. Reabsorption of embryos was suggested in a study using BALB/c laboratory mice which found that females inseminated in the acute phase of B. microti infection (0-12 days) did not produce any offspring [24]. Reabsorption of fetuses when mice are infected early in the gestation period has also been observed in other pathogens, such as Neospora caninum in BALB/c laboratory mice [33]. A significantly higher number of Week 3 embryos with significantly lower parasitemia compared to embryos in earlier stages of development (Week 1 and 2) was observed. This may also indicate that females became infected with B. microti after insemination or during later stages of gestation and the pathogen had not fully disseminated to all embryos when the female was euthanized. Experimental manipulation of the timing of infection and insemination is required to fully characterize the affect B. microti may have on developing fetuses and pathogen persistence.
Females exhibited significantly higher parasitemia compared to their offspring. Although the different tissues investigated could partially account for this difference, we found no significant differences in MCN in a subset of samples of adult heart tissue and blood. Decreased parasitemia in embryos may thus be due to limited placenta permeability. Although the placenta is known to function as a barrier to potentially harmful microbes and pathogens [36], multiple parasites and pathogens have developed ways of evading the placental barrier to infect a fetus, including another apicomplexan, Plasmodium spp. [37], as well as Toxoplasma gondii [38], Trypanosoma spp. [39], and several nematodes [40,41]. Similar to Plasmodium spp., B. microti sporozoites invade host red blood cells and are able to breach the placental barrier to infect a female's offspring [42]; exactly how these pathogens are able to diffuse through the placenta is unknown.
Tolkacz and colleagues [27] found different genotypes of B. microti in female voles and their offspring, although the mechanisms for these differences were not explained. This was not the case in our study, with sequences from P. leucopus and M. pennsylvanicus revealing high similarity between females and fetuses and overall. Low genome-wide sequence diversity was also observed in other studies [43,44]. Some studies suggest that B. microti represents a rich species complex consisting of three distinct clades in the United States using the 18S rRNA gene; however genetic diversity within each clade is relatively invariant [19]. Strains in Clade 1 are found associated with human-biting Ixodes species, locations known to be endemic for human babesiosis, and exhibit the least amount of genetic diversity. Whereas strains in Clade 2 were isolated from carnivores and those in Clade 3 were found to infect rodents; neither Clade 2 nor 3 have been found to infect humans. Sequences from this study most closely aligned with the reference sequence from Clade 1 and these sequences showed a low sequence diversity of 0.001. One of our sequences (2033-Arteries) exhibited higher diversity than predicted by previous studies; this might have occurred due to multiple infections or interference from other pathogens during Sanger sequencing (resulting in a noisy chromatograph) and most likely does not represent a unique B. microti species.
Our field-based results indicate that transplacental transmission of B. microti is a potentially important pathway for infection in wild rodent species, P. leucopus and M. pennsylvanicus, which may partially explain B. microti emergence and geographic expansion. In endemic areas, transplacental transmission may enhance early season amplification of B. microti, contributing to higher infection prevalences in both hosts and vectors. Additional enhancement mechanisms have been described, including increased transmission of B. microti to ticks feeding on hosts coinfected with B. burgdorferi based in field and laboratory studies [6,11,35] and amplification driven by tick aggregation, as shown in an empirically-informed model [45]. Transplacental transmission in natural populations may also facilitate B. microti maintenance and spread in small rodent populations in areas with limited I. scapularis occurrence. For instance, in areas of Alaska, California, Colorado, Florida, Maine, and Montana that lack I. scapularis, small rodent populations (voles, shrews, cotton rats, other Peromyscus spp.) were infected with both human-infecting and nonhuman-associated strains of B. microti [46][47][48][49][50]. Some evidence suggests that nidicolous ticks (e.g. I. angustus, I. muris, I. spinipalpis) or other exophilic ticks (I. pacificus) may contribute moderate enzootic persistence in certain areas [48] in a 'cryptic cycle' , but vertical transmission may also play a role. Irrespective of the maintenance mechanism in the enzootic cycle, transmission to humans and emergence would only occur when I. scapularis (a 'bridge' vector to humans) invades a region where B. microti had been enzootically maintained, as has been shown for B. burgdorferi [49,51].
To confirm the relevance of transplacental and vertical transmission in B. microti, studies in a laboratory setting are required to specifically quantify the relative efficiencies of B. microti transmission routes, including tick-to-host transmission, vertical transmission in relation to insemination and infection timing, chronic infection, and xenodiagnoses and transmission efficiency of infected offspring. More extensive field studies quantifying the role of vertical transmission and chronic infection of B. microti should then be conducted to assess how these transmission pathways influence this pathogen's persistence and transmission, in comparison to B. burgdorferi. As winter temperatures in these areas continue to increase each year [52], overwinter survival of mice may increase [53,54], further enhancing the early season advantage of B. microti through chronic infection and vertical transmission.