ABSTRACT
Human babesiosis is a rapidly emerging and potentially fatal tick-borne disease caused by intraerythrocytic apicomplexan parasites of the Babesia genus. Among the various species of Babesia that infect humans, B. duncani has been found to cause severe and life-threatening infections. Detection of active B. duncani infection is critical for accurate diagnosis and effective management of the disease. While molecular assays for the detection of B. duncani infection in blood are available, a reliable strategy to detect biomarkers of active infection has not yet been developed. Here, we report the development of the first B. duncani antigen capture assays that rely on the detection of two B. duncani-exported immunodominant antigens, BdV234 and BdV38. The assays were validated using blood samples from cultured parasites in human erythrocytes and B. duncani-infected laboratory mice at different parasitemia levels and following therapy. The assays display high specificity with no cross-reactivity with B. microti, B. divergens, Babesia MO1, or P. falciparum. The assay also demonstrates high sensitivity, detecting as low as 115 infected erythrocytes/µl of blood. Screening of 1,731 blood samples from diverse biorepositories, including previously identified Lyme and/or B. microti positive human samples and new specimens from field mice, showed no evidence of B. duncani infection in these samples. The assays could be useful in diverse diagnostic scenarios, including point-of-care testing for early B. duncani infection detection in patients, field tests for screening reservoir hosts, and high-throughput screening such as blood collected for transfusion.
Short summary We developed two ELISA-based assays, BdACA38 and BdACA234, for detecting B. duncani, a potentially fatal tick-borne parasite causing human babesiosis. The assays target two immunodominant antigens, BdV234 and BdV38, demonstrating high specificity (no cross-reactivity with other Babesia species or Plasmodium falciparum) and sensitivity (detecting as low as 115 infected erythrocytes/µl). The assays were validated using in vitro-cultured parasites and infected mice. Screening diverse blood samples showed no evidence of B. duncani active infection among 1,731 human and field mice blood samples collected from the north-eastern, midwestern, and western US. These assays offer potential in diverse diagnostic scenarios, including early patient detection, reservoir animal screening, and transfusion-transmitted babesiosis prevention.
INTRODUCTION
In recent years, vector-borne infectious diseases have emerged as a significant global public health concern (1). The expansion of the geographic range of the tick vectors has led to a notable increase in tick-borne diseases (TBDs), which now account for more than 75% of all reported vector-borne diseases in the United States annually (2–4). Among these tick-borne diseases is human babesiosis, an emerging infectious disease caused by intraerythrocytic apicomplexan parasites of the Babesia genus (5). In 2011, the disease became a US national notifiable condition in 18 states, which led to a significant increase in the number of reported cases. A recent report by the US Centers for Disease Control and Prevention (CDC) documented a rapid surge in human babesiosis cases over the past decade, 16,456 clinical cases were reported between 2011 and 2019 primarily in 37 states, with 98.2% of cases reported concentrated in 10 states (6). While ticks are the main vehicle of transmission of Babesia parasites, blood transfusion remains a major concern despite screening efforts.
To date, eight species of Babesia have been identified as human pathogens and documented to cause mild to severe babesiosis with some cases resulting in fatal outcomes (7). The majority of clinical cases reported to the CDC are attributed to Babesia microti (8, 9), whereas only a few cases of B. duncani babesiosis have been documented in the United States, primarily in Washington, Oregon, and California. However, it is crucial to acknowledge that some of the B. microti cases might be attributed to B. duncani. This observation underscores the significance and justifies the necessity for a new and highly sensitive diagnostic test capable of distinguishing between the two Babesia species. Indeed, as per the CDC Babesiosis Surveillance Report (2011–2015) in the United States, three babesiosis cases were linked to B. duncani in Maryland and Connecticut (10) based on serological criteria. This underscores the concern regarding the presence and potential threat of B. duncani on the East Coast and validates the need for a new, highly sensitive diagnostic test that is specific to B. duncani.
In confirmed cases of B. duncani babesiosis, the parasite was found to induce fulminant and often fatal infections in both immunocompetent and immunocompromised individuals (11–14). Available data suggest that the hard tick Dermacentor albopictus serves as the vector for B. duncani (13).
To date, the true incidence of B. duncani babesiosis remains unknown. Accurate determination of the true prevalence of B. duncani babesiosis is challenging due to the absence of specific and reliable diagnostic tests that can determine active infection. Over the years, various diagnostic techniques including microscopy, serological tests like immunofluorescence assay (IFA), immunoblot and enzyme-linked immunosorbent assay (ELISA), and nucleic acid-based tests such as nested PCR, reverse transcription-polymerase chain reaction (RT-PCR), and transcription-mediated amplification (TMA), have been employed to detect Babesia infections. Morphologically, B. duncani-infected red blood cells (RBCs) are indistinguishable from those infected with B. microti or other Babesia species and bear resemblance to the ring-stage forms of Plasmodium falciparum, the causative agent of severe human malaria. Therefore, microscopy cannot be reliably used to diagnose B. duncani active infection in a blood sample. With the exception of cases confirmed by both microscopy and PCR (14–19), most unverified cases attributed to this parasite have relied on serological assays, which cannot reliably distinguish between past and active infections. A 2018 publication by Scott and Scott summarized reports by physicians and naturopathic physicians of 1119 cases of B. duncani between 2011 and 2017 in Canada, with most cases occurring in the Pacific Coast region (20). However, no positive blood smears or details about the sensitivity or specificity of the serological and molecular assays used were described in that report. Furthermore, no appropriate controls were included in these studies.
Therefore, these claims warrant further validation, which could be facilitated by sharing blood samples with established biorepositories and large-scale blood screening in different geographic areas in the US. Nucleic acid-based detection techniques are indeed more specific and sensitive and have been effective in identifying positive B. microti blood samples. Nevertheless, in some instances, these techniques have been shown to detect residual parasite DNA or RNA even after the parasite has been eliminated (21, 22). Recent studies using microscopic examination of blood samples from B. microti-infected animals treated with drugs that cleared the infection demonstrated the presence of residual DNA after treatment with antibabesial drugs (22). Studies using B. microti-infected blood revealed that antigen detection assays, which rely on the detection of highly abundant and often immunodominant secreted antigens of the parasite, exhibit a stronger correlation with active B. microti infection compared to Nuclic Acid-detection assays (23). Therefore, efforts to develop similar assays for the detection of active infections by other Babesia species could be an invaluable resource to estimate the true incidence of babesiosis caused by each of the Babesia species that cause infection in humans. This is particularly important in light of the rapid environmental changes impacting the transmission of these diseases and other TBDs (24, 25).
A significant breakthrough in the field of B. duncani biology emerged with the establishment of the first continuous in vitro culture system of B. duncani in human erythrocytes (26–28). Furthermore, the recent sequencing and annotation of the parasite’s genome (29) made it possible to apply a multi-faceted approach that incorporates genomic, transcriptomic, and proteomic analyses to identify unique and species-specific proteins that are exported by the parasite from infected RBCs into the host plasma.
In this study, we report the discovery of two highly reliable biomarkers for active B. duncani infection. Babesia duncani-specific antigen capture assays (BdACAs) that detect these antigens were developed and exhibit high sensitivity, are capable of detecting as low as 115 parasites/µl of blood and can distinguish active infections from past exposures, and can distinguish between B. microti and B. duncani to provide accurate species infections, important for public health.
MATERIALS AND METHODS
Ethics statement
All animal experiments described in this study followed the Yale University institution guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committee at Yale University (protocol number 2023-07689).
Babesia duncani infection in mice
C3H/HeJ or SCID mice (C.B-17/IcrHsd-Prkdcscid) 5-6 week, female mice from Jackson laboratories were infected by injecting 1×106 or 1x 10 7 B. duncani WA-1 strain or B. microti LabS1 strain, respectively by infected erythrocytes intravenously (IV) from the stock mouse. Blood samples were collected at different time points by a retro-orbital bleed or cardiac puncture, and parasitemia was monitored by light microscopy of Giemsa-stained blood smears. Where indicated, plasma or serum was separated from whole blood by centrifugation at 6000g for 10 min, and packed mouse RBC (mRBC) was lysed with 1% saponin. Hemolysate and parasite pellets were collected after centrifugation at 9300g for 10 min at 4°C.
Babesia duncani in vitro culture
Babesia duncani WA-1 parasites were cultured continuously in human erythrocytes in DMEM-F12 media (28), as described previously (26). (Included in Supplementary methods).
Isolation of extracellular vesicles from plasma, culture supernatant, and hemolysate
Extracellular membrane vesicles containing protein cargo released from the parasite were isolated from supernatant/plasma and hemolysate of B. duncani in vitro and in vivo culture by EXO Quick ULTRA EV isolation kit (EQULTRA-20A-1) following the user manual. Where indicated, vesicles were also isolated by the ultracentrifugation method, as mentioned previously (30). Briefly, 1.5 ml of human RBC (hRBC) and mRBC samples (uninfected and B. duncani infected plasma/serum or supernatant and hemolysate) were sequentially centrifuged at 500g for 30 min, followed by 16,000g for 45 min. These samples were then centrifuged at 120,000g for 14h at 4°C using a Sorvall MTX 150 micro-ultracentrifuge with an S52-ST swinging bucket rotor (Thermo Fisher) and ultracentrifuge pellet (Up/UHp) and ultracentrifugation supernatant (Us/UHs) fractions were collected (31).
Cloning, expression, and purification of recombinant proteins
Plasmid DNA constructs for BdV234 and BdV38 were synthesized by GenScript, Inc., NJ. Details description of protein expression and purification provided in the supplementary methods.
Immunodetection of BdV234 and BdV38
Polyclonal antibodies raised in rabbits (Cocalico Biologicals, Inc., PA) against the indicated B. duncani proteins were purified and used for the immunodetection analysis as described in detail (32). Briefly, supernatant (S), hemolysate (H), and parasite pellet (P) fractions collected either from B. duncani in vitro culture in hRBC (18% parasitemia) or B. duncani infected (24% parasitemia) or uninfected mRBCs were mixed with 4X Laemmli sample buffer (3:1 ratio), boiled at 80°C for 5 min and separated on 4-20% Mini-protean gels (Bio-Rad, P4568093). The gels were then transferred onto nitrocellulose membranes (Bio-Rad, 1620115), probed with anti-BdV234 and BdV38 primary antibodies (1:250 dilution) and goat-anti-rabbit horseradish peroxidase (HRP)-conjugated IgG Ab (1: 5000 dilution, (Thermo Fisher, 31466). Blots were developed by Super signalTM West Pico PLUS chemiluminescent substrate (Thermo Fisher, 34577) and imaged using the LI-COR Odyssey-Fc imaging system.
Immunofluorescence assay (IFA)
Immunofluorescence assay was performed as reported previously (30, 32) and described briefly in the supplementary methods.
BdACA using B. duncani in vitro culture supernatant and hemolysate fractions
Babesia duncani antigen capture sandwich ELISAs were performed as described earlier (22, 23), with few modifications. Briefly, 96-well plates were coated with 5µg/ml of capture Abs (BdV234 and BdV38) and incubated for 2h at room temperature, followed by blocking with 5% BSA in PBST (0.05% Tween-20) at 37°C for 1h. Plates were then incubated overnight at 4°C with either (1) BdV38 and BdV234 recombinant protein (100 ng/well), (2) a 1 in 10 dilution of B. duncani in vitro culture supernatant, hemolysate, and whole blood, or (3) plasma/serum (heat-inactivated for 30 min at 56°C), hemolysate and whole blood from uninfected or B. duncani infected C3H/HeJ mice. Whole blood samples from American Red Cross (ARC), Lyme Disease Biobank (LDB), and field-derived mouse samples were incubated similarly. The plates were then washed four times with PBST, and 5 μg/ml biotinylated detection Ab (BdV234 and BdV38) were added and incubated at room temperature for 2h. Following this, plates were washed four times with PBST and incubated with HRP conjugated streptavidin antibody in 1:5000 dilution at room temperature for 1h, followed by washing and incubation with 50 μl 3,3’,5,5’ tetramethylbenzidine (TMB) substrate for 7 min in the dark. The reaction was stopped by adding 50 μl of 0.1N HCl, and the Optical Density at 450 nm was measured using a BioTek Synergy Mx plate reader.
Detection of BdV234 and BdV38 antigens in hemolysate from drug-treated B. duncani in vitro culture or infected mouse blood
In vitro B. duncani culture with an initial parasitemia of 1% (cultured in 5% hematocrit with DMEMF-12 containing hRBCs) was exposed to tafenoquine (SML0396, Sigma Aldrich) at 1x IC50 (3 µM) and 2x IC50 (6 µM) concentrations for three generations. Both untreated and tafenoquine treated parasite cultures were collected at specified time points (0, 3, 6, 12, 24, 36, 48, 60, and 72 hours post treatment), lysed with 1% saponin, and the resulting hemolysates were used for ELISA. Untreated infected and uninfected hRBCs were included as controls. Similarly, hemolysates were prepared from blood from 12 C3H/HeJ (3 male, 3 female) mice infected with a high dose of B. duncani (106 parasites) and used in ELISA to detect BdV234 and BdV38. Six of these mice were treated with tafenoquine for 5 days (DPI 3-7), and the remaining six were treated with vehicles (PEG400). Vehicle-treated and uninfected mice hemolysates were used as controls.
Genomic DNA extraction and Real-Time PCR
To further confirm the presence or absence of B. duncani DNA in infected blood, qPCR analyses were conducted as previously reported (33). Briefly, blood from B. duncani in-vitro culture and B. duncani infected mouse at various parasitemia levels were collected at specific time points post-infection, followed by gDNA isolation by the DNeasy Blood and Tissue kit (Qiagen, 69504). The isolated genomic DNA were used to amplify the B. duncani internal transcribed spacer (ITS) sequence in its nuclear rRNA (18s-rRNA) by using BdITS1-F 5′-GCTTCCTAACCCGAGACCAA-3′ and BdITS1-R 5′-CACTGGCGGGGTGAAAAGTA-3′ primers (34). The reaction mixture contained 1× advanced universal SYBR green super mix (1725270; Bio-Rad), 1 μl of each primer (10 μM), and 2 μl of template DNA.
Collection of biological samples used for screening B. duncani antigen/infection
Two hundred human whole blood samples were obtained from the American Red Cross (ARC), of which 100 samples were positive for B. microti infection as detected by transcriptionally mediated amplification (TMA) positive and the remaining 100 samples were TMA negative for B. microti, the samples were collected in areas endemic for B. microti. (Connecticut, Delaware, Massachusetts, Maryland, Maine, Minnesota, New Hampshire, New Jersey, New York, Pennsylvania, Virginia, Vermont). Additionally, 770 whole blood samples were obtained from the Lyme Disease Biobank (LDB) with 317 samples enrolled with signs and symptoms of early Lyme disease and 282 controls as previously described (35) and were collected from East Hampton, Martha’s Vineyard, and Wisconsin (2014–2020). Similarly 171 samples were collected from the west coast (California), including 11 enrolled with early Lyme disease, 10 controls, and 150 with peresistent Lyme symptoms. Whole blood samples were also obtained from 761 white-footed mice (Peromyscus leucopus), a prominent reservoir host of B. microti, (unscreened) that were collected from western Pennsylvania, USA, and used in this study.
RESULTS
Identification of B. duncani secreted antigens
The genome of B. duncani has been extensively studied and characterized. Its annotation identified 4222 proteins, of which 479 are predicted to be members of the parasite’s secretome (29). To identify proteins exported by the parasite into the host red blood cell and/or host environment, we used a Nanotrap-based proteomics approach (36, 37) on hemolysate (H) and secreted (S) fractions collected from a culture of B. duncani-infected human erythrocytes as shown in Fig. 1A. Mass spectrometry analysis identified 27 proteins in the S fraction (Table 1) and 39 proteins in the H fraction. Of these exported proteins, BdV234 and BdV38 were selected based on their peptide abundance, high expression levels of their encoding genes as determined by RNAseq, and predicted antigenicity profiles (Table 1). The genes encoding these two proteins were cloned into the pGEX 6p-1 and pET21a (+) expression vectors, respectively, and the encoded recombinant proteins were expressed in E. coli, purified on glutathione Sepharose high-performance (BdV234) and Ni-NTA agarose (BdV38) affinity columns and used to generate specific polyclonal antibodies in rabbits. The antibodies were then affinity-purified and evaluated for their specificity using recombinant proteins as well as total extracts from a culture of B. duncani-infected erythrocytes. The resulting affinity-purified anti-BdV234 and anti-BdV38 antibodies were used to detect the expression and cellular distribution of the native proteins in the supernatant (S), hemolysate (H), or pellet (P) fractions from cultures of B. duncani-infected human erythrocytes and control uninfected human erythrocytes (Fig. 1B). Consistent with the export of these proteins, immunoblot analyses identified the native BdV234 and BdV38 in all three fractions while no signal could be detected in the fractions isolated from the uninfected erythrocytes (Fig. 1B). We then assessed the cellular distribution of the same antigens in mice infected with B. duncani. Interestingly, whereas the antigens could be readily detected in the H and P fractions prepared from mouse blood collected from B. duncani-infected C3H/HeJ mice, the signal from the plasma (Pl) fraction was very weak, likely the result of dilution in the large volume of plasma collected from infected animals (Fig. 1B). Additionally no signal could be detected from the Pl, H, or P fractions prepared from blood collected from uninfected mice, further confirming the specificity of the antibodies raised against these proteins (Fig. 1B).
Evidence of vesicular-mediated secretion of BdV234 and BdV38 from B. duncani-infected erythrocytes
Recent studies in B. microti and B. divergens have demonstrated that several proteins of these parasites’ are actively exported by using vesicular-mediated export mechanisms (30, 38). To assess the mode of export of BdV234 and BdV38, the culture S and H fractions of B. duncani-infected human erythrocytes, as well as the Pl and H fractions of B. duncani-infected mice were subjected to ultracentrifugation at 120,000 x g for 14h at 4°C to separate the soluble (Us/UHs) and vesicle-associated membrane (Up/UHp) fractions. These resulting fractions were then analyzed by immunoblotting using anti-BdV234 and anti-BdV38 antibodies to assess the presence or absence of the antigens. As shown in Fig. 1C, the proteins were primarily found associated with the Up/UHp fractions in both human and mouse infected samples (Fig. S1-A), indicating that once exported from the infected erythrocytes, these antigens are associated with vesicles. This association was further validated using the ExoQuick® vesicle isolation system, which was used to separate the supernatant fraction (Es/EHs) from the vesicle-rich pellet fraction (Ev/EHv) on plasma, supernatant, and hemolysate fractions collected from uninfected or B. duncani-infected blood (Fig. S1-B). As shown in Fig. 1D, both antigens were detected in the vesicle fractions (EV or HEV) from the S, H, and Pl fractions obtained from B. duncani-infected human and mouse RBCs, respectively. As a control, no signal could be detected in similar fractions prepared from blood collected from uninfected human erythrocytes in vitro or uninfected C3H/HeJ mice (Fig. 1D). Together these data demonstrate that BdV234 and BdV38 are exported by B. duncani-infected erythrocytes through a vesicular-mediated mechanism.
Cellular localization of selected B. duncani secreted proteins
In order to investigate the localization and cellular distribution of BdV234 and BdV38 secretory proteins in B. duncani, immunofluorescence assays (IFA) were conducted on uninfected or B. duncani-infected human erythrocytes by using affinity-purified BdV234 and BdV38 antibodies from rabbit sera. Anti-rabbit Alexafluor-488 was used for the detection of BdV234 and BdV38, and α-Band III antibody was used as a membrane marker for human RBCs. Interestingly, BdV234 was not only found to be localized inside the parasites but was also found in the host erythrocytes displaying punctate vesicular staining (Fig. 2), which indicating that this GPI-anchored protein is secreted by the parasite into the host cytosol. On the other hand, BdV38 was found to be localized primarily within the parasites. The rabbit pre-immune sera served as a control in our IFA, and no localization signal with BdV234 and BdV38 was observed other than the nucleus and RBC membrane (Fig. 2).
Development of B. duncani antigen capture Assays BdACAs
The discovery that BdV234 and BdV38 could be readily detected in the hemolysate and environment of B. duncani-infected erythrocytes led us to investigate the use of these two proteins as biomarkers of an active B. duncani infection. We used the anti-BdV234 and anti-BdV38 antibodies to design an antigen capture sandwich ELISA assay (BdACA) to detect the exported proteins in B. duncani-infected blood (Fig. 3A). The BdACA assays, one for each antigen, were initially optimized by employing two-fold serial dilutions of recombinant BdV234 and BdV38 proteins to generate a standard curve. The ELISA data was graphed with optical density against known antigen concentrations to obtaining sigmoidal curves (Fig. 3B). To further validate the assays for native proteins at the cellular level, an in vitro culture of B. duncani-infected human erythrocytes at 18% parasitemia with 5% hematocrit was prepared, and the supernatant and hemolysate fractions were collected, subjected to two-fold serial dilutions, and used to detect the exported antigens. From our findings, both BdV234 and BdV38 effectively identified the presence of total parasite proteins that were secreted into the culture supernatant and established a positive correlation with the percentage of parasitemia counted by microscopy (Fig. 3C). In addition, when the assays were conducted on the culture hemolysate samples, we observed a comparable correlation accompanied by a marginal rise in the absorbance values (Fig. 3D). In both assay conditions, we observed nearly undetectable levels of the proteins when using the supernatant and hemolysates from cultures of uninfected human RBCs. We extended the BdACA validation into the established in vivo mouse model, utilizing serum and plasma from B. duncani-infected mice (with ∼24% parasitemia) (Fig. 3E), along with the H fractions (Fig. 3F). We observed robust detection of the parasite antigens when probing the mouse hemolysate fractions with both antibodies, while the levels were merely undetectable in the serum or plasma fractions (Fig. 3E). As expected, there was no detection in the serum, plasma, or hemolysate fractions from uninfected mice.
High sensitivity of BdACAs for detection of B. duncani infection
The sensitivity of the BdACAs was assessed using a culture of B. duncani at ∼9 x 106 B. duncani-iRBCs per ml, which was subsequently subjected to two-fold dilutions and allowed to proliferate for 24h in culture. For the comparison of BdACA with qPCR, we isolated the genomic DNA from the same samples and used it for PCR detection of the B. duncani 18s rRNA (Fig. 4A). The H fractions from the same cultures were further used to evaluate the sensitivity of the BdACAs. The Bd38ACA could detect as low as 1 x 103 iRBCs, while Bd234ACA could detect as low as 2 x 103 iRBCs (Fig. 4B-C). No reactivity was observed in uninfected human RBCs, as expected. The limit of blank (LOB) marked the threshold above which all values were considered positive, while the limit of detection (LOD) for Bd38ACA and Bd234ACA was determined to be 3 x of LOB (OD450=0.3). The sensitivity of BdACA was also assessed to establish the minimum detectable level of parasitemia. A 1 ml of B. duncani in vitro culture with an initial parasitemia of 0.5% (5% hematocrit) was allowed to grow over time, and parasitemia was monitored. After isolating the hemolysate (1 ml), genomic DNA extraction was performed to facilitate quantitative polymerase chain reaction (qPCR) analysis (Fig. 4D). The BdACA assay employed 5 microliters per well alongside the previously mentioned hemolysate volume. Our findings demonstrated that the BdACA method effectively identified parasitemia levels as low as 0.5%, consistent with the expected sensitivity. Additionally, as the secretion of antigens increased over time, our assays could detect more secreted antigens (Fig. 4E-F). Similarly, to validate the sensitivity of the assay in vivo samples, blood was collected from day post-infection (dpi) 0-7 from C3H/HeJ mice. Genomic DNA isolated from parasites at different time points was used in qPCR using B. duncani 18s rRNA forward and reverse primers that detected parasites as early as 0-dpi (Fig. 4G), while Bd38ACA and Bd234ACA could detect the antigens as early as 2 and 3 dpi, respectively (Fig. 4H-I).
BdV38 and BdV234 as biomarkers for evaluating antibabesial treatment efficacy
An ideal biomarker of active Babesia infection is one whose levels show an excellent positive correlation with parasitemia levels and does not remain in the blood following clearance of infection due to drug treatment. To assess whether BdV38 and BdV234 meet these criteria, BdACAs were conducted on in vitro B. duncani cultures under both untreated and drug-treated conditions. Parasite cultures with 1 % parasitemia were initially inoculated and collected at 12h intervals over a 72h period (three parasite erythrocytic lifecycles) following treatment with either vehicle control (DMSO) or tafenoquine (TQ) at 3 µM (1x IC50) and 6 µM (2x IC50). The samples were assessed for parasitemia detection in both conditions through microscopic analysis of Giemsa-stained blood smears. Observations from Giemsa-stained blood smears at stipulated time intervals of post-inoculation displayed continuous parasite growth until 72h (Fig. 5A). The decline in % parasitemia over time and the inhibitory impact of 1x and 2x IC50 of tafenoquine were evident, detecting parasitemia as low as 0.05% and 0.1% at 48h compared to the untreated (UT) parasites, which reached approximately 10% after 72h post-inoculation (Fig. 5A). The detection of BdV38 from the total parasite lysate positively correlated with the viability of the parasite over time in untreated (UT) samples (Fig. 5B). In contrast, the curve declined over time and reached a plateau after 48h with tafenoquine treatment (1x or 2x) (Fig. 5B). For post-drug treatment, as tafenoquine cleared the parasites from the culture, the total absorbance showed nearly no difference compared to the uninfected RBC lysate. Upon qPCR analysis of the same samples to assess parasite detection, the presence of B. duncani DNA was indicated in both the untreated and treated conditions at stipulated time intervals. However, a clear difference was observed between the uninfected and infected samples (Fig. 5C). Similarly, we used whole blood samples from uninfected and B. duncani-infected (106) mice that were treated with tafenoquine (TQ) or vehicle. Blood samples were tested to assess parasite growth by microscopy (Fig. 5D), and BdV38ACA was used to detect BdV38 secretion that declined after 7 dpi once the parasites were cleared by the drug treatment (Fig. 5E). qPCR results suggested DNA remained in circulation even after the parasites were cleared (Fig. 5F). Thus, the Bd38ACA assay reliably detected only active infections. Similarly we tested these in vito and in vivo samples with anti-BdV234 antibody and found the similar results as with BdV38 (Fig. S3).
Bd38ACA and Bd234ACA are highly specific assays for B. duncani detection
To assess the specificity of the BdACA assays, we tested blood samples infected with other Babesia species and similar apicomplexan parasites, such as P. falciparum from in vitro culture as well as from animal hosts. We also compared the sensitivity of BdACA from uninfected and B. microti (106) infected mouse blood. Uninfected and B. duncani-infected mouse blood was used as a control. Our in vivo results indicated that Bd38ACA (Fig. 6A) and Bd234ACA (Fig. 6C) are highly specific for B. duncani-secreted proteins compared to B. microti. Similarly, the assays were conducted on in vitro blood samples from different cultures of B. divergens (Rouen87 strain), Babesia MO1, or P. falciparum (3D7 strain), each with 10% parasitemia grown in human erythrocytes. Our result indicated that Bd38ACA (Fig. 6B) and Bd234ACA (Fig. 6D) only detect the respective proteins secreted by B. duncani, and no signal above the lower limit detection of OD 0.3 could be detected in these samples.
Blood screening using Bd38ACA
To evaluate the effectiveness of BdACA as screening tools for detecting active B. duncani infections, we employed the Bd38ACA assay to screen 1,731 whole blood specimens from diverse locations across the US. These encompassed 200 human whole blood samples (100 B. microti-positive and 100 B. microti-negative) from the American Red Cross (ARC); 770 human whole blood samples from the Lyme Disease Biobank (LDB), previously screened for Lyme disease and B. microti; and 761 whole blood samples from white-footed mice collected from Western Pennsylvania (Fig. 7A). In this study, the lower limit of detection (LOD; OD450 = 0.3) was defined as thrice of the LOB (lower limit of blank), which corresponds to a B. duncani parasitemia of 0.3% (Fig. 7B). All 1,731 samples, analyzed in this study, showed OD450 values below 0.6, which when compared to the standard curve for B. duncani parasitemia levels corresponds to parasitemia levels below 0.55% (Fig. 7A). Of these 1731 samples, 8 had OD450 values between 0.31 and 0.59. Among the 761 white-footed mouse samples, 756 showed signals below the LOD and 5 samples showed signals above the LOD but below an OD450 of 0.35 (< than 2x LOD). Similarly, out of the 770 Lyme Disease Biobank samples, 3 showed signals above the LOD with OD450 of 0.31, 0.49 and 0.59. The 8 samples with OD450 signals above the LOD using the BdV38-based detection assay, were further confirmed to be within this same range when tested using the BdV234ACA assay (not shown), suggesting that these samples either carry B. duncani at parasitemia levels below 0.5% or the BdV38 and BdV234 assays detect antigens from other closely related pathogens. To gain further insights into the source of the signals in these samples (N=9, one ARC30 sample also included), PCR-based amplification analyses were conducted using the primer pairs listed in Table 2. Using primers designed to be specific for B. duncani (specific to HSP-70, CelTos, and BdV38 genes), B. microti (specific to BmITS and Bm GPI12 genes), B. divergens (specific to AMA-1 gene) and B. MO1 (specific to the helicase gene), our PCR analyses showed that these samples do not contain DNA from these species (Fig. S4-S5). To assess whether these samples contain DNA from other Babesia species, we selected samples with the highest OD450 among these for PCR analysis using a previously reported primer pair (39) that amplifies a 200 bp of the 18 S rRNA gene from multiple Babesia species (Fig. 6A). The nine samples included 1 from ARC, 3 from Lyme Biobank and 5 from field mice. As shown in Fig. 6 B, six of the 9 samples tested positive by PCR. Sequencing of the 200 bp PCR product revealed sequences with 98 to 100% identity to 18S rRNA genes from Babesia species of Clade VI (B. divergens, capreoli, B. odocoilei, and B. venatorum) (Table 3). To further evaluate these data, a new primer set was designed to bind to conserved regions of the 18S rRNA genes of Babesia species and amplify a 736 bp. Using this primer set against all 9 samples, we found 5 (field mice) samples to be positive by PCR. However, sequencing of the fragments failed to identify a specific species at more than 90% sequence identify (Fig. S6 C-D). Together these data suggest that the 8 samples may contain new species of Babesia or closely related organisms that the BdACAs can also detect. However, the exact identity of these new species could not be determined due to limited amount of material available for large scale genomic analyses.
DISCUSSION
Herein, we report the development of the first antigen capture assays, BdACAs, for the detection of active B. duncani infection. The assays detect the presence of B. duncani-secreted antigens BdV38 (31 kDa) or BdV234 (27 kDa) in blood samples from infected cultures or animals with high sensitivity and specificity. The BdACAs exhibit a remarkable sensitivity in detecting these secreted B. duncani antigens from as little as a single drop of blood, equivalent to 1.6 x 107 red blood cells, in both human and mouse samples. Furthermore, the assay specifically distinguishes B. duncani infections when tested against other human babesiosis-causing parasites (B. microti, B. divergens (Rouen87), Babesia MO1) or related protozoan species (P. falciparum) that causes human malaria. These assays, therefore, open up new possibilities for efficient and convenient diagnosis of B. duncani infection, ensuring timely intervention and improved healthcare outcomes. The versatility of this assay makes it useful in various setups, including point-of-care (POC) tests in clinics or at home. Its potential applications include confirming acute B. duncani infection, screening clinical samples, and ensuring the safety of blood supplies at blood donation centers, thereby preventing transfusion-transmitted babesiosis. Additionally, it offers a valuable tool for conducting epidemiological surveys to detect B. duncani infection in humans and reservoir hosts. Given the urgent need for improved diagnostic tests to combat vector borne or transfusion-transmitted babesiosis, as seen with other Babesia sp. in the USA, the development of this assay represents a significant stride toward achieving that goal (40, 41) and the first report of an ELISA based antigen capture assay for B. duncani.
The current strategies employed for the detection of subclinical infections of intraerythrocytic parasites suffer from notable limitations. Microscopy, although widely used, exhibits low sensitivity and specificity, making it challenging to differentiate between various species of Babesia and other apicomplexan parasites, such as Plasmodium. Serological tests such as immunofluorescence assays (IFA) lack reliability in distinguishing between acute and chronic infections. On the other hand, PCR based assays that show high sensitivity to detect parasite DNA can detect the DNA even after the infection has been cleared, as observed in the case of B. microti (22). An ineffective diagnostic tool also hampers the proper understanding of the epidemiological distribution of the parasite in its reservoir hosts.
One key advantage leveraged in our approach is the use of an in culture-in mouse (ICIM) model system of B. duncani (42) infection, which allowed us to systematically examine different aspects of our assay, including optimizing the assay using various sample types such as hemolysate, supernatant, whole blood, plasma, and serum. Additionally, the recent comprehensive genome analysis of B. duncani by Singh et. al. (29) shed light on the abundance and nature of proteins, epigenome, and transcriptome analysis identified classes of candidate virulence factors, antigens for diagnosis of active infection, and several attractive drug targets. Studies in B. microti have shown that secreted proteins are excellent biomarkers for the detection of active parasite infection (43, 44). These findings led to the development of antigen capture assays (ACAs) for the detection of the secreted antigen BmGPI12 (22, 23). Here, we have used an integrated approach consisting of genomic, transcriptomic, and proteomic analyses to identify secreted proteins of B. duncani, which could be suitable as biomarkers of active infection from different fractions (S, H, and P) (Table 1). Among the pool of exported proteins, BdV234 and BdV38 were specifically chosen due to their substantial peptide abundance and notably elevated expression levels (Table 1). We reasoned that those proteins with higher abundance and containing signal sequences were predominantly destined for the extracellular space, where they could play a crucial role in eliciting cell signaling and facilitating communication between host and parasite to promote infection propagation. To our surprise, our initial attempts to detect BdV234 and BdV38 directly from the plasma or serum of B. duncani-infected mice were unsuccessful. However, by employing ultracentrifugation and the Exo-Quick approach, we were able to purify extracellular vesicles (EVs) that contained these parasite proteins. The EVs containing these B. duncani proteins may play a crucial role in the trafficking and delivery of these proteins, allowing them to be expressed on the outer membrane of infected RBCs similar to that reported for B. microti (30, 38). Our IFA results, from in vitro B. duncani-infected parasites, further support the cellular distribution of these proteins. Further investigation is warranted to explore the precise mechanism by which these proteins are sorted and packaged into EVs, as well as their specific role in mediating their transport to the surface of infected RBCs. Understanding the molecular processes involved in this vesicular transport machinery will contribute to our overall comprehension of B. duncani pathogenesis and host-parasite interactions.
This study assessed the field suitability of Bd38ACA by screening a diverse array of human and mouse whole blood samples. The assay exhibited minimal cross-reactivity with both human and field-derived white-footed mouse samples. However, a subset (0.4%) of samples showed low reactivity with Bd38ACA, at levels significantly lower in comparison to the positive control samples. Subsequent PCR analysis confirmed the absence of B. duncani parasites in these samples but the exact organisms, likely a species closely related to B. duncani, that cause the cross-reactivity detected in the BdACA assays remain to be further characterized.
In conclusion, our findings demonstrate that BdACA is a reliable diagnostic assay for the detection of active B. duncani infection. It is suitable for large-scale screening of the blood supply and further optimization of the assay is warranted to test uninfected and infected human blood samples and compare with the current method (NAT assays) for blood screening. The development of such a diagnostic tool is a significant step forward in the fight against B. duncani and can significantly improve the accuracy and efficiency of diagnosis, treatment, and surveillance of this emerging tick-borne disease.
Data Availability
All data produced in the present work are contained in the manuscript
Author contributions
MC, PV, DD, LZ, JC, and AP: Investigation, methodology, formal analysis, visualization, writing original draft, review, and editing. SW: Sample collection and analysis. CB: Conceptualization, supervision, funding acquisition, project administration, writing original draft, review, and editing. All authors have read and agreed to the published version of the manuscript. All authors contributed to the article and approved the submitted version.
Funding
The research described herein was supported by the Global Lyme Alliance Foundation. CBM research is also supported by NIH grants AI138139, AI123321, AI152220 and AI153100, and AI136118, the Steven and Alexandra Cohen Foundation (Lyme 62 2020), and The Blavatnik Family Foundation.
Acknowledgments
We thank the American Red Cross and the Lyme Disease Biobank for providing the blood samples used in this study.