Establishment and Optimisation of Anti-Babesia microti Drug Ecacy Evaluation Method

Background Babesiosis, an infectious zoonotic parasitic disease that occurs globally, is most commonly caused by Babesia microti. For a severe infection, a combination of exchange transfusion and support-therapy is used and, for a mild infection, single antibiotics, or a combination of antibiotics and antiprotozoal drugs, are used for 7–10 days for the treatment of babesiosis; however, as these treatments have problems, a new therapy is needed. Although aetiological and molecular biology methods are often used in drug screening, these methods have disadvantages and there is no standard anti-B. microti drug screening method. Method This study was consisted of basal test, drug-suppressive test and drug-therapeutic test. BALB/c mice were used as the animal model and robenidine hydrochloride (ROBH) was used as a positive drug. By modifying Peter’s 4-day suppression method, immunosuppressive and subpassage tests were conducted, in combination with microscopy and real-time uorescent quantitative PCR (qPCR), the current anti-B. microti drug screening method was optimised. SPSS was used for data analysis.


1)
In the basal test, there were signi cant differences in the erythrocyte infection rate (EIR) and relative value of the target gene (P = 0.011 and 0.012, respectively) among days. The blood of infected mice 28,35, and 37 days post-infection (dpi) could infect healthy second-generation mice, even if no parasites could be observed under a light microscope. Resurgence occurred when the immunity of an infected mouse decreased to a certain extent after using an immunosuppressant.

2)
In the drug-suppressive test, no B. microti was observed under the light microscope in any ROBH group after drug administration. Signi cant differences were observed in the EIR and relative value of the target gene (both P < 0.001) between the control group and ROBH groups. No B. microti was observed after the administration of an immunosuppressant or in second-generation mice in ROBH-treated groups; however, B. microti was observed in control and proguanil hydrochloride-treated groups.

3)
In the drug-therapeutic test, there were signi cant differences in the EIR and relative value of the target gene between the 100 mg/kg ROBH groups and the control group (P = 0.049 and < 0.001, respectively) but not between ROBH groups and atovaquone-azithromycin group (P = 0.587 and 0.418, respectively). None of the second-generation mice in any drug-treated group were infected, whereas all the second-generation mice in the control group were infected; parasites could be observed in the control group from the 2 nd day after the administration of an immunosuppressant, several parasites were observed in the atovaquone-azithromycin group from the 10 th day after the administration of an immunosuppressant, and no parasites were observed in ROBH groups.
Conclusion ROBH can be used as a positive drug in anti-B. microti drug screening trials. qPCR results can be used as a judgment standard in preliminary screening tests and the combination of immunosuppressive and subpassage experiments can be used to validate drug e cacy. Background Babesiosis, a zoonosis usually transmitted by tick bites, occurs globally. Babesia species live in the red blood cells of mammals and cause similar symptoms to malaria. Currently, more than 100 species and genotypes of Babesia that can infect mammals have been identi ed worldwide, including seven species can cause babesiosis in humans-Babesia microti, Babesia divergens, Babesia bovis, Babesia canis, Babesia duncani, Babesia venatorum (once named Babesia sp. EU1), and a new Babesia sp., KO1 (similar to Babesia ovine) [1,2]. Since the rst case of babesiosis in humans was reported in Yugoslavia in 1957 [3], more cases have been reported all over the world, including in the USA, France, Germany, Australia, China, Vietnam, and Myanmar; most cases have been reported in the USA [4], most of which were infected by B. microti [5].
Most people infected by B. microti show mild symptoms, such as fever and anaemia, and return to normal after 2-4 weeks [6]. If these people do not reach hospital in time, or if the parasite goes undetected, they are likely to become B. microti-carriers. In carriers, recurrence occurs in low immune states and the parasite can be transmitted mother-to-child, blood donor to immunode cient blood transfusion recipient, or organ transplant donor to recipient [7,8].
Exchange transfusion and support-therapy are usually used to treat patients with a severe B. microti infection, which is characterised by parasitaemia (≥ 10%), severe haemolysis, and organ failure [9]. In patients who experience more than 3 months of parasitaemia but no symptoms, or mild to moderate infection, combination treatment with antibiotics and antiprotozoal drugs, such as quinine-clindamycin, quinine-azithromycin, and atovaquone-azithromycin combinations, is used for 7-10 days [1,10]. However, these treatments have side effects, relapse can occur, and 'resistance' has occurred in patients with immune dysfunction [11]. Therefore, new and more effective treatments need to be explored.
Several approaches are used to determine anti-B. microti drug e cacy in drug-screening trials. An aetiological method, observation by microscopy after blood smear staining, is the most commonly used. Mahmoud, using BALB/c mice, conducted in vivo anti-B. microti drug screening experiments for two drugs [12] and showed that the detection rate of this method is 1000 B. microti/mL; however, more smears are needed to improve the detection rate when parasitaemia is not high. Aetiological methods require experts to distinguish parasites; otherwise, parasites are easily omitted or mistakenly identi ed [13].
Marley, using golden Syrian hamsters (Mesocricetus auratus), conducted microscopy combined with subpassage tests and observed that an 8-aminoquinoline, WR238605, has good e cacy [14]. This method has high accuracy but a long experimental period, large number of test animals, and high requirements for experimenters. Yao conducted preliminary screening experiments on 19 drugs (antibiotics and antiprotozoal drugs) based on this method in BALB/c mice and demonstrated that robenidine hydrochloride (ROBH), a drug used for the prevention of coccidiosis in chickens, has the best e cacy on B. microti [15].
Molecular biology methods, such as the original polymerase chain reaction (PCR), nested PCR, and realtime uorescent quantitative PCR (qPCR), are often used in the diagnosis of protozoan infections. There are advantages and disadvantages to each PCR method. Original PCR has a detection limit of 30 parasites and low sensitivity, accuracy, and speci city [16,17]. The sensitivity of nested PCR is higher, 2.4 parasites/µL, but is easily contaminated because of the two rounds of ampli cation and false-positive results occur frequently [16]. Because the target gene is in the DNA of the parasite in nested PCR, time is needed to clear all DNA after the parasites are killed; thus, in drug e cacy screening tests, a positive result in nested PCR cannot re ect whether a drug is effective promptly. The sensitivity of real-time qPCR (with SYBR Green) is the same as that of nested PCR but the detection limit of the TaqMan probe is up to 12 parasites/µL. Furthermore, real-time PCR can not only be qualitative but also quantify the target gene by Ct value (threshold cycle) [18,19].
Mohamed established the Babesia uorescence assay (BFA) using SYBR Green I to stain DNA and nested PCR to observe nucleic acid residues in animal organs and determine drug e cacy. However, the speci city of BFA is not good, the dye stains not only parasite DNA but also BALB/c mice and microbial DNA, and the test time is long. Additionally, removing blood from mice tail veins for in vitro cultivation allows environmental factors to in uence the results. Therefore, the use of BFA is inadequate to determine drug e cacy [20].
Some immunological methods, such as the indirect uorescent antibody test (IFAT), enzyme-linked immunosorbent assay (ELISA), and immune chromatography test (ICT), are also used in the diagnosis of protozoan infections. IFAT is regularly used to detect Babesia and diagnose babesiosis in the USA; its sensitivity and speci city are 88%-96% and 90%-100%, respectively [21]. ELISA is suitable for a large number of serum samples but cross-reactions with other sera occur [22][23][24]. ICT can detect a speci c antibody in the serum and show the result using colour. Additionally, as it is simple, accurate, and intuitive, it is widely used as an on-site test [22]. However, antibodies are produced more than 1 week from infection and a positive serum antibody result does not necessarily represent a current infection. Therefore, immunological methods are not applicable for B. microti drug screening in the laboratory [13].
Based on this previous research, in this study, current anti-B. microti drug screening methods were optimised using BALB/c mice as the animal model and ROBH as the positive drug. Peter's 4-day suppression method was modi ed and immunosuppressor resurgence and subpassage tests, with microscopy and real-time uorescent qPCR, were used to establish an e cient, rapid, and accurate method to evaluate anti-B. microti drug e cacy in vivo. The results of this study may lay the foundation for the screening of reliable anti-B. microti drugs.

Method 1. Establishing B. microti-BALB/c mice model
The B. microti Peabody mjr strain, ATCC®PRA-99TM, obtained from the Institute of Laboratory Animal Sciences, Chinese Academy of Medical Sciences, was conserved by our institute. Female BALB/c mice, aged 6-8 weeks, were inoculated by an intraperitoneal injection of 200 μL of mice blood that contained 10 7 B. microti-infected erythrocytes, as previously described [25]. The number of mice used in basal test, drug-suppressive test and drug-therapeutic test was 153, 252 (36 mice in one group, seven groups in total) and 135 (45 mice in one group, three groups in total) respectively.

Staining and microscopic examination
Blood was removed from mice tail tips to prepare thin smears on microscope slides. All smears were airdried, xed in absolute methanol, stained with Giemsa stain (Solarbio, Beijing, China), and the parasites were counted under ×1000 magni cation using a bright-eld microscope (Nikon, Tokyo, Japan). Five thousand erythrocytes were scanned for each sample, the number of the erythrocytes infected with B. microti was recorded, and the EIR of each sample was calculated.

Nucleic acid testing
Genomic DNA was extracted from erythrocytes in ethylenediaminetetraacetic acid-anticoagulated blood using the DNeasy Blood & Tissue Kit (Qiagen, Shanghai, China). A speci c fragment of B. microti 18S rRNA was set as the target fragment and the primers were Bab 2 (5 -GTTATAGTTTATTTGATGTTCGTTT-3 ) and Bab 3 (5 -AAGCCATGCGATTCGCTAAT-3 ), as used in a previous study [26]. The mice actin gene was used as a reference gene and the primers were actin-qF (5 -GTGCTATGTTGCTCTAGACTTCG-3 ) and actin-qR (5 -ATGCCACAGGATTCCATACC-3 ). Real-time uorescence qPCR was performed in a 10 μL reaction mixture containing 5 μL of 2×AceQ qPCR SYBR Green Master Mix (Low ROX Premixed), 1 μL of DNA template, 0.2 μL of each primer (10 mM), and 3.2 μL of ddH 2 O. The reactions were performed under the following conditions: predegeneration at 95 °C for 5 min, followed by 40 cycles of 10 s at 95 °C and 30 s at 60 °C. The conditions for the dissociation curve analysis were 95 °C for 5 min, 60 °C for 60 s, and 95 °C for 15 s. qPCR was repeated three times for each sample. The qPCR reactions were performed using ABI QuantStudio7.

Subpassage testing
Six healthy female BALB/c mice, aged 6-8 weeks, were inoculated with anticoagulated blood (200 μL for each) from six infected mice in each group 28 days post-infection (dpi), as described in a previous study [14,15]. Thin blood smears were prepared on microscope slides from 7 dpi to 30 dpi, stained with Giemsa stain, and observed under the microscope, as described in Method 2, for parasites. A negative result was recorded if no parasite was detected.

Immunosuppressor testing
A dexamethasone sodium phosphate injection was used as the immunosuppressor, as described previously [27]. Thin blood smears from six mice were prepared on microscope slides 2 days after immunosuppressor injection (200 μL for each). The next steps followed Method 2. A positive result was recorded if parasites were observed.

Drug-suppressive and drug-therapeutic tests
Inoculated BALB/c mice (using Method 1) were treated with drug 4 h after infection (0 dpi) in the drugsuppressive test [28] and 10 dpi in the drug-therapeutic test, based on the weight of the mice (0.2 mL/10 g), by intragastric gavage. The drug dosage was prepared according to clinical guidelines and previous study ( Table 1). Three mice in each group were randomly selected as samples at each time point in both the drugsuppressive and drug-therapeutic tests. Thin blood smears were prepared on microscope slides and nucleic acid testing was conducted for each sample.
Six mice were used in subpassage testing to infect six second-generation mice 28 dpi and the immunosuppressor was used for 5 days from 28 dpi.

Statisticalanalysis
EIR was evaluated by the percentage of infected erythrocytes. The relative value of the target gene was calculated using the 2 -∆∆Ct method and 1 dpi mice from each group were used as a reference. In the basal test, a paired samples t-test was used to analyse the mean values of EIR and relative value of the target gene of all samples and a linear correlation analysis was used. Repeated measures analysis of variance (two factors and multi-level) was used to analyse the mean values of EIR and relative value of the target gene of all samples in both drug-suppressive and drug-therapeutic tests. All data analyses were performed using IBM SPSS software, version 19.0. P < 0.05 was considered signi cant.

Basal test
Staining, microscopic examination, and nucleic acid testing were performed daily for 46 days from 0 dpi. Subpassage testing and immunosuppressor testing were conducted 28, 35, and 37 dpi. The immunosuppressor was administered daily for 7 days.

EIR
All inoculated mice were successfully infected with B. microti.
Dynamic EIR changes were observed in the examination of thin blood smears (Fig. 1). The acute period was 7-10 dpi and the highest infection rate was detected 9 dpi, with an average EIR of 37%. Few parasites were detected in the chronic infection stage after 20 dpi, which is in accordance with the research of Xu [25] and Cai [29]. The P value was 0.011, which indicated that there were signi cant differences among daily EIRs.
B. microti was small in the early stage of infection and showed diverse morphology, such as dot, ring, pear, amoeba, rod, and Maltese cross shapes. With the increase in EIR, the infected erythrocyte gradually swelled. Several parasites could be observed in one erythrocyte and dark chromatin was observed in parasites. As B. microti escaped from a ruptured erythrocyte, the number of parasites in that erythrocyte decreased and the infected erythrocyte returned to normal. As same as described in Cai's study [29].

Nucleic acid testing
The change tendency is shown in Fig. 1. The highest relative value of the target gene was observed 9 dpi and the peak stage was 7-10 dpi, which was followed by dynamic changes. Statistical analysis showed signi cant differences among daily relative values of the target gene (P = 0.012).

Subpassage testing
B. microti was observed in the thin blood smear of all second-generation mice 4 to 15 days after infection by the anticoagulated blood of rst-generation mice on 28, 35, and 37 dpi. It indicates although no parasites were observed in the thin smear 28, 35, and 37 dpi, the blood was infectious.

Immunosuppressor testing
B. microti was observed in the thin blood smear 2 days after the intraperitoneal injection of immunosuppressor 28, 35, and 37 dpi; however, the infection rate was low. The EIR reached 36.2% 7 days after the daily intraperitoneal injection of immunosuppressor and continued to increase, up to 61.1%. This indicated that B. microti would appear in the blood again when the immunity of the infected mice in the chronic stage decreases.

Drug-suppressive test
All inoculated mice were successfully infected with B. microti. Both the examination time point and number of mice in one group were reduced based on the results from the basal experiment, and 4, 7, 11, 15, 20, and 28 dpi were set as examination time points.

EIR
Blood was extracted from mice tail tips to prepare thin smears on microscope slides at each time point.
No parasite was observed in the ROBH group (EIR = 0.00%). The morphology of red blood cells and B. microti in ROBH and control groups after infection under a transmission electron microscope is shown in Fig. 2. There was no B. microti in mice red blood cells in the ROBH group and the shape of red blood cells tended to be more normal than that in the control group.
There were signi cant differences between the ROBH and control groups, the ROBH and chloroguanide hydrochloride groups (both P < 0.001). The EIR in the control and chloroguanide hydrochloride groups showed dynamic changes and the peak stage was 7-11 dpi (Fig. 3). There were no signi cant differences between the control group and the 4-and 10-day chloroguanide hydrochloride group (P = 0.334 and 0.950, respectively).

Nucleic acid testing
The relative value of the target gene was < 1.00 in the ROBH group. Signi cant differences were observed between the ROBH and control groups and the ROBH and chloroguanide hydrochloride groups (both P < 0.001). Additionally, there were no signi cant differences between 4-and 10-day ROBH treated groups for both 100 mg/kg and 60 mg/kg doses. The relative value of the target gene in the chloroguanide hydrochloride and control groups was highest on 7 dpi and 11 dpi, respectively, and declined gradually to < 0.10 from 28 dpi. There were signi cant differences between the chloroguanide hydrochloride and control groups (P < 0.001) and the 4-and 10-day treated chloroguanide hydrochloride groups (P < 0.001).

Subpassage testing
B. microti was observed in the thin blood smear of second-generation mice in both chloroguanide hydrochloride and control groups 7 to 15 days after infection by the anticoagulated blood of rstgeneration mice 28 dpi. No parasites were observed in the thin smear of second-generation mice in all ROBH groups.

Immunosuppressor testing
Thin blood smears of mice in each group were prepared 3-30 dpi and B. microti was observed in both chloroguanide hydrochloride and control groups but not in the ROBH group. The results of repeated experiments were consistent. This indicated that, after infected mice in the chronic stage are treated by ROBH, B. microti would not appear in blood again.

3.Drug-therapeutic test
The combination of atovaquone and azithromycin is most commonly used for the treatment of patients with B. microti infection and was used as a reference to study the e cacy of ROBH. Mice were treated with the drug daily for 10 days from 10 dpi and blood was inspected 10, 13, 15, 18, 21, 24, 28, 31, and 56 dpi. All inoculated mice were successfully infected with B. microti.

EIR
Dynamic changes in EIR are shown in Fig. 4. The EIR reduced to 0.00%, 0.33%, and 0.70% in the ROBH, atovaquone-azithromycin combination, and control groups, respectively, 13 dpi. The P value was 0.049, which indicated that there was a signi cant difference between 100 mg/kg ROBH and control groups. However, there were no signi cant differences between the 100 mg/kg ROBH and atovaquoneazithromycin combination groups and the control and atovaquone-azithromycin combination groups (P = 0.587 and 0.108, respectively).

Nucleic acid testing
The relative value of the target gene was < 0.10 in the ROBH group from 21 dpi, > 0.10 at any examination time point in the atovaquone-azithromycin combination group, and > 0.30 until 31 dpi and > 0.10 until 56 dpi in the control group. There were signi cant differences between the 100 mg/kg ROBH and control groups and the atovaquone-azithromycin combination and control groups (both P < 0.001). However, there was no signi cant difference between the 100 mg/kg ROBH and atovaquone-azithromycin combination groups.

Subpassage testing
B. microti was only observed in the thin blood smear of second-generation mice in the control group.

Immunosuppressor testing
Thin blood smears of mice in each group were prepared 3-30 dpi and B. microti was observed in both atovaquone-azithromycin combination and control groups but not in the ROBH group. Only a small number of B. microti was observed in the atovaquone-azithromycin combination group after 10 days of immunosuppressor administration.

Discussion
The aim of this study was to evaluate anti-Babesia drug e cacy evaluation methods. Three tests were carried out: basal test, drug-suppressive test, and drug-therapeutic test. The difference between drugsuppressive and drug-therapeutic tests was the timing of drug administration. In the drug-suppressive test, the drug was administered daily for 4 or 10 days from 4 h after inoculation, whereas in the drugtherapeutic test, the drug was administered daily for 10 days from the 10 th day after inoculation, when parasitaemia was more acute. To examine the results of all three tests, thin blood smear staining and microscopic examination, nucleic acid testing, subpassage testing, and immunosuppressor testing were conducted.

1.Basal test
Each method veri ed the results of the others. The trends of EIR and relative value of the target gene in qPCR were consistent, the result of the correlation analysis was signi cant (Fig. 5; Pearson r = 0.8313, R 2 = 0.6910), and the peak stage based on both measurements was 7-10 dpi. When the EIR was 0.00% after 20 dpi, although it was di cult to con rm infection by microscopy, nucleic acid testing showed a positive result. Simultaneously, second-generation mice were infected, which indicated that infected mice in the chronic stage were infectious and that B. microti would recur when infected mice in the chronic stage entered a weakened immune state.

2.Drug-suppressive test
In this test, Peter's 4-day suppression method was modi ed. Considering the period of clinical therapy, the treatment time was prolonged, and regardless of ROBH dose and treatment time, the EIR was 0.00% from 4 dpi. The relative value of the 18S rRNA encoding gene was < 1.00. Both subpassage and immunosuppressor showed negative results. The results of the control group were similar to those of the basal test.
We also studied the e cacy of chloroguanide hydrochloride, which has a similar structure to ROBH, in the drug-suppressive test. The recommended dosage for chloroguanide hydrochloride is 40 mg/kg but this was toxic and lead to mice death. Therefore, the dosage was gradually lowered to 25 mg/kg. The result showed that, despite its guanidino structure, chloroguanide hydrochloride could not inhibit B. microti, which is concordant with the results of our previous study [15].

3.Drug-therapeutic test
Generally, patients enter a hospital when symptoms appear. However, symptoms show after the incubation period of B. microti, which can be 1-4 weeks after tick bites and 1-9 weeks after blood transfusion [6]. We treated mice for 10 days from the day the EIR reaches its peak stage, and used the common combination therapy of atovaquone and azithromycin as a reference, to investigate the e cacy of ROBH and verify the accuracy of the method we developed.
The combination of atovaquone and azithromycin relieved parasitaemia quickly but recrudescence appeared in immunosuppressor testing, whereas the ROBH-treated group showed negative results in all four methods. This indicates that the e cacy of the atovaquone and azithromycin combination may be inferior to that of ROBH.
We evaluated these four B. microti-detecting methods in their ability to evaluate drug e cacy and observed that thin blood smear staining and microscopy was most intuitive, although it missed parasites easily and required specialists. Additionally, it was di cult to determine whether all parasites died or were simply not detected when we failed to identify B. microti by microscopy. Subpassage testing and immunosuppressor testing had high accuracy but a long time and many animals were required. qPCR had high accuracy and could process many samples at one time. The relative value of the target gene was calculated using the 2 -∆∆Ct method; thus, both qualitative and quantitative results could be offered by qPCR.
Additionally, nested PCR was used in drug e cacy evaluation in this study (data not shown) but always showed a positive outcome in three tests from 1 dpi to the end of the test. This result is in accordance with those by Xu [26]. Because the elimination of parasite DNA takes time, nested PCR, which provides qualitative results only, could not re ect drug e cacy promptly.
Consequently, we suggest that ROBH is used as a positive drug in anti-Babesia drug screening experiments. Methodologically, the modi ed Peter's 4-day suppression method could be used in preliminary screening and qPCR could be conducted to detect mice blood 4-11 dpi and, if the relative value of the target gene is < 0.10, the tested drug can be considered effective in the inhibition of parasite growth. Then, a drug-therapeutic test could be carried out, in which subpassage testing and immunosuppressor testing could be conducted 28 dpi. If both show negative results, the tested drug can be assumed to be effective for the treatment of babesiosis.

Conclusion
In our study, an anti-B. microti drug e cacy evaluation method was established and optimised. ROBH can be used as a positive drug in anti-B. microti drug screening trials in the laboratory. The time and number of animals required can be reduced using qPCR as a judgment standard in anti-B. microti drug preliminary screening tests. Finally, the combination of immunosuppressive and subpassage experiments can be used to validate drug e cacy.

Consent for publication
Not applicable.

Availability of data and material
The datasets supporting the conclusions of this article are included within the article.

Competing interests
The authors declare that they have no competing interests.