The strain of B. microti was supplied by American Type Culture Collection (ATCC) under the number ATCCR PRA-99^TM and was maintained in our laboratory by serial passage of parasitized erythrocytes in mice (Zhao et al., 2016). The colony of R. haemaphysaloides ticks was derived from water buffalo in Hubei Province in southern China (Zhou et al., 2006) and the colony of Hyalomma asiaticum ticks was derived from sheep in the Xinjiang Autonomous Region in northern China (Li et al., 2014). The colony of parthenogenetic Haemaphysalis longicornis ticks was derived from deer in the Shanghai Wildlife Park in China (Zhou J. et al., 2013). A single engorged female was used to establish each tick colony. Tick rearing was performed in the laboratory in a dark incubator at 25°C with 92% relative humidity and feeding on a New Zealand white rabbit. After three generations had been maintained under laboratory conditions, the colonies of ticks were established in Shanghai Veterinary Research Institute, Chinese Academy of Agricultural Sciences, Shanghai, China. According to general principles, sexual maturity animals were chosen for this study. Six-week-old KM mice (Shanghai Silaike Experimental Animal Co. Ltd.). Two-month-old SD rats (Shanghai Silaike Experimental Animal Co. Ltd.), Three-week-old Yellow hair chicks (Shanghai Songlian Experimental Animal Co. Ltd.), 2-month-old Changshang pigs (Shanghai Nongxi Pig Farm), 3-month-old Berger’s dogs (Shanghai Songlian Experimental Animal Co. Ltd.), and 1-month-old goats (Shanghai Songlian Experimental Animal Co. Ltd.) were used in this study. The experimental animals were treated following the approved guidelines from the Animal Care and Use Committee of the Shanghai Veterinary Research Institute (Approval number SHVRI-MO-0023), all surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering.

Blood containing B. microti-infected erythrocytes were collected by cardiac puncture from mice infected with B. microti. After measuring the total erythrocyte count and infection rate, parasitized erythrocytes were adjusted to 1 × 10⁸/mL by phosphate buffered saline (0.01M PBS, pH7.4). The B. microti-infected erythrocytes were then injected either intraperitoneally for rat or intravenously for chick, pig, dog, and goat with 1 × 10⁷ parasitized erythrocytes, three individuals were infected for each kinds of domestic animals. Blood samples were then collected from the animals once every 5 days until 55 days after inoculation.

Thin blood smears with Giemsa staining were used to examine Babesia parasites in infected red blood cells (RBCs) under a light microscope, a method that is currently used in most diagnostic laboratories (Mosqueda et al., 2012). The peripheral bloods from infected animals were taken, and then made thin blood smears with Giemsa staining for microscopic observation.

DNA was extracted from 300-μl of EDTA preserved whole blood using the QIAamp DNA minikit blood and body fluids protocol (Qiagen) according to the manufacturer’s instructions and eluted in an equal volume of elution buffer. Extracted DNA (5 μl) was used as a template for quantitative PCR following the methodology (Rollend et al., 2013), Forward and reverse primers and probe sequences are: Bm18Sf-AACAGGCATTCGCCTTGAAT, Bm18Sr-CCAACT GCTCCTATTAACCATTACTCT, and Bm18Sp-6FAM-CTA CAGCATGGAATAATGA-MGBNFQ, respectively. The quantitative PCR was performed using ABI 7500 instruments. Nested-PCR was used to detect parasites in blood samples by a previously published protocol (Persing et al., 1992).

Whole blood samples were collected regularly from infected animals and the sera were collected and kept at -30°C. Enzyme-linked immunosorbent assay (ELISA) was carried out to detect the antibody responses against B. microti infection, which used the recombinant BmSA1 as the coated antigen (Luo et al., 2011). Briefly, 96-well microplates were coated with the antigen, glutathione S-transferase (GST)- BmSA1 and GST (negative control) at a concentration of 250 ng/well. Animal sera diluted at 1:100 with PBS containing 0.05% Tween 20 were used as the source of primary antibody and HRPO-conjugated anti-chicken IgG (ID, A30-107P) or anti-pig IgG (ID, A100-105P), or anti-goat IgG (ID, A50-100P), or anti-dog IgG (ID, A40-123P), or anti-rat IgG (ID, A110-105P) was used as the second antibody respectively (Bethyl Laboratories, United States). After addition of substrate [2,20-azino-bis (3-ethylbenzthiazoline-6-sulfonic acid)], the optical density (OD) value was recorded at 415 nm by SpectraMax M5 (Molecular Devices, United States).

EDTA-preserved blood was collected from infected animals as described above, blood examination was carried out using the Mindray BC-2800 Auto Hematology Analyzer (Mindray, China). Number of white blood cells (WBC), number of RBCs, hematocrit (HCT), and mean corpuscular volume (MCV) were checked.

Approximately 600 larvae were applied equally to feed on 3 mice with B. microti parasitemia of 20%. Each mouse was collared with a plastic cap with central drilling to prevent grooming and then ticks were allowed to feed to repletion on infected mice. Parts of engorged larvae ticks were pooled (10 larvae ticks) for B. microti qPCR detection at once; the other engorged larvae were allowed to molt in the dark at 25°C, 92% relative humidity for 2–4 weeks. Newly molted tick nymphs were collected (individual) and used for qPCR detection. In the same way, about 100 nymphs were applied equally to feed on 3 mice with B. microti parasitemia of 20%. Some of the engorged nymphs and newly molted adult ticks from the engorged nymphal were collected (individual) and used for B. microti qPCR detection respectively. Also, about 60 adult female ticks (add 60 adult male ticks for R. haemaphysaloides and H. asiaticum) were applied equally to feed on 10–30 mice with B. microti parasitemia of 20% depending on different species of ticks. Some of engorged female ticks (individual) and the pool (10 larvae ticks) of newly molted larval ticks from the single engorged female were collected and used for B. microti qPCR detection.

Ten KM mice were infested with 60 nymphs each, which were previously infected with B. microti as larvae. Approximately 50 μl tail bloods of mice were collected for parasite detection every 3 days starting from day 7 to day 30 after tick infestation. When no B. microti infection was detected after blood samples were microscopically examined, each mouse was sacrificed, bled and a fresh mice was inoculated with the blood. The monitoring process was conducted again as mentioned above for another 30 days until positive mice were found.

For each blood sample, a thin blood smear was stained with Giemsa-stain in order to examine parasites by microscope. Forty-five microliters of blood were used for DNA extraction, and qPCR was performed to detect B. microti infection. Blood collection was stopped when the mice were proved to be infected upon microscopic examination.

Differences in the positive rate of ticks were tested by Chi-square, which was performed using SPSS 20.0 software. A probability P-value < 0.05 was considered statistically significant.

Following infection with B. microti, parasitemia increased rapidly in rats, parasites were observed in RBCs 1 day after infection, with parasitemia peaking at 29.1% at day 6, then declining rapidly to <1% at day 11 post-infection (Supplementary Figure S1). Although it was difficult to detect parasites in RBCs of infected rats after day 11 post-infection, a few parasites could still be observed over the following few days. By contrast, no parasites were observed in the RBCs of dogs, goat, pigs and chicks at any time in the experiment periods (data not shown).

As shown in Table 1, the parasite DNA could not be detected by quantitative PCR and nested-PCR in the blood samples of all host animals from 5 to 55 days after infection except the rat. The copy numbers of target gene of B. microti were never less than 10³ at any point during the experiment in the rat in real-time PCR (Supplementary Figure S2). However, B. microti did not to be detected in the blood samples of dogs, goats, pigs and chicks after 5 days following initial infection by both qPCR and nested-PCR (data not shown).

As shown in Figure 1, compared with rat serum before the infection, the specific antibodies against B. microti were detectable in rats as early as 10 days post-infection and a significant antibodies levels were found on 20 days post-infection, peaking at 30 days post-infection (OD > 0.8) and then kept high levels until the end of experiments. In dogs and goats, unexpectedly, antibody response was quick. Significant antibody levels were found on 10 days post-infection, peaking at 15 days post-infection and then antibody responses decreased quickly and disappeared on 20 and 30 days post-infection for goats and dogs respectively. By contrast, all serum samples from the experimental pigs and chicks showed no any specific antibody responses against B. microti.

As shown in Figure 2, during early infection in rats, the values of RBC and HCT were below the value of the normal range at day 5 post-infection. The numbers of RBCs of infected rats increased at 10 days post-infection and returned to the normal range at 15 days; the values of HCT returned to the normal range at 10 days post-infection. The numbers of leukocytes increased rapidly at day 5 post-infection, peaking at day 10 post-infection which was eight times the maximum normal value (20.9 × 10⁹ /L), returning to within the normal range at day 15 post-infection. The mean value of MCV of infected rats was also higher than the normal range at day 10 post-infection and returned normal at day 15 post-infection. By contrast, no significant changes from normal values were recorded in the levels of RBC, HCT, WBC and MCV in experimental infected dogs, goats, pigs and chicks (Supplementary Figures S3–S6).

Tick larvae were fed on the mice infected with B. microti for 3–5 days, 600–650 larvae of H. longicornis, R. haemaphysaloides and H. asiaticum were engorged and collected respectively. As shown in Table 2, engorged larvae pools of three species of ticks were all 100% positive (20/20) by PCR, which indicated that the larvae ticks had ingested B. microti successfully. However, positive rates of molted ticks (nymphs) were 93.3, 86.7, and 83.3% for H. longicornis, R. haemaphysaloides and H. asiaticum respectively, which showed no significant difference between different species of ticks.

About 150 engorged nymphs were collected from infected mice after feeding of H. longicornis, R. haemaphysaloides, and H. asiaticum. All 20 of the engorged nymphs tested by PCR were positive, however, positive rates of molted ticks (adult) were 70, 55, and 50% for H. longicornis, R. haemaphysaloides, and H. asiaticum respectively (Table 3), which showed significant difference between H. longicornis and the other two species of ticks. This result may suggest the different ability for transmission in this developmental stage.

About 50 engorged females of each tick species were collected from infected mice after feeding. All 20 of the engorged female ticks tested by PCR were positive, but no positive larvae tick pools from single female ticks were found in all tested samples in H. longicornis, R. haemaphysaloides, and H. asiaticum (Table 4).

As shown in Table 5, most of infected nymphs had successfully engorged on mice between 3 and 5 days. However, B. microti was not detected in mice in the first 30 days after tick infestation and the second 30 days after collected blood was blindly passed by inoculation. Parasites were detected by qPCR in 2 out of 10 mice in the third 30 days after blood passage. The parasite was not found in the blood smears of mice until the fourth 30 days after blind passage of blood (Supplementary Figure S7). It took about 4 month to firstly detect B. microti infection after tick infestation test.