All procedures were performed in accordance with the protocols approved by the Institutional Animal Care and Use Protocol Committee of the University of Idaho (protocol #2018-19). All procedures adhered to the U.S. National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals.

Six Welsh cross horses (HO-397, HO- 403, HO-394, HO-402, HO-396 and HO-398) were used in this study. All the horses were more than 6 months of age.

T. equi Florida strain was used in this study. This is the reference parasite strain and its genome sequence has been published in GenBank (GenBank ID number: 11168).

Since equine TGF-β1 is not available as an analyte in Millipore’s MILLIPLEX® MAP Equine Cytokine Magnetic Bead Panel kit, the horse transforming growth factor beta1(TGFB1) ELISA Kit (Cusabio, Houston, Tx) was used to assess expression of the cytokine in the serum samples. The assay was performed in accordance with the manufacturer’s instructions and like the Milliplex assay, 4PL was used to determine TGF-β1 concentrations in the samples and the Assayfit pro software (AssayCloud, Nijmegen, Netherlands) was used to perform the analysis.

The Theileria equi antibody test kit (VMRD, Pullman) was used to perform competitive ELISA for the detection of EMA-1 specific- antibodies in the serum samples. The assay was performed in accordance with the manufacturer’s instructions, and the results were reported as % inhibition. Samples that yielded ≥ 40% inhibition were considered positive while samples that exhibited inhibition of < 40% were considered negative.

Two-way ANOVA available in GraphPad Prism 10.2.1 (GraphPad Software, San Diego, CA) was used to perform statistical analyses at a significance level (α) of 0.05.

The schizont and pre-schizont stages of T. equi development were not evaluated as part of this study. However, the horses manifested typical equine piroplasmosis symptoms and disease dynamics ( Supplementary Table 1 ). As expected, a drastic decline in PCV was observed in the primarily infected horses compared to secondarily infected horses. The decline was observed as soon as 2 DPI and by 7 DPI, HO-402’s PCV had decreased by more than 50% ( Supplementary Table 1 ). By 16 DPI, all the horses in the primary infected group had a decline in PCV of greater than 50% with a significant decline observed at 14 DPI (P=0.0098) ( Figure 2A and Supplementary Table 1 ). On the other hand, PCV remained mostly within the normal range of 30% to 47% in the secondary infected horses ( Supplementary Table 1 ) and no significant difference was observed in PCV levels pre- and post- secondary infection ( Figure 2B ). Peak rectal temperature coincided with significant decline in PCV in the primary group (P=0.0098) ( Figure 2A ). Similarly, an inverse relationship was observed between PCV and parasite DNA load ( Figure 3 ). The highest parasite DNA load assessed by quantification of DNA was observed at 14 DPI in both groups and this coincided with pyrexia in HO-402 ( Supplementary Table 1 ).

While the primarily infected animals manifested clinical signs such as anorexia, lethargy, malaise, and pale mucous membranes, none of the secondary infected animals showed these clinical signs. This is possibly because the primary exposure to the T. equi merozoite stabilate induced protective immune responses against the parasite (13). In this regard, the parasite DNA load in the secondarily infected group was much lower than in the primarily infected group. This resulted in less severe hemolysis and thus minimal reduction in PCV compared to the primary infected animals ( Supplementary Table 1 ). Collectively, symptoms observed in the primary infected group are consistent with the undesirable effects of pro-inflammatory responses. This prompted us to investigate the patterns of expression of cytokines that could have influenced the development of the observed symptoms.

Although the primary infected horses manifested symptoms that might be linked to inflammation, they did not express many of the pro-inflammatory cytokines whose expression was assessed i.e. IL-1β, IFN-γ, IL-6 and TNFα ( Table 1 ). Minimal levels of IL-17A, IL-12 and RANTES were expressed by one animal each ( Table 1 ), and there was no obvious correlation between their expression and the symptoms observed ( Supplementary Table 1 ). Notably, expression of IL-10 was observed in two of the three primary infected horses i.e. HO-402 and HO-396 with significant concentrations observed in the latter ( Table 1 ).

Similarly, most of the pro-inflammatory cytokines i.e., IL-1β, IL-6, IL-17A, IFN-γ, IL-4 and IL-12 were not expressed in the secondarily infected group, except for RANTES and TNF-α. ( Table 2 ) The former was expressed at minimal levels by two of the three animals while the latter was expressed by HO-394 only ( Table 2 ). There was also no correlation between expression of the pro-inflammatory cytokines and symptoms observed except for elevated temperature and expression of RANTES by HO-394 at 28 dpi ( Supplementary Table 1 ). Like the primarily infected group, expression of IL-10 was observed in all the persistently infected horses ( Table 2 ). The cytokine was expressed at higher concentrations in the secondarily infected horses than in the primarily infected horses at 14 and 21 DPI ( Figure 4 ). Additionally, the highest concentrations of the cytokine were observed at 14 DPI, and this coincided with the maximum parasite load in both groups ( Figure 3 ).

Analysis of expression of TGF-β1 using a competitive inhibition ELISA showed that all the horses in this study expressed the cytokine. Primarily infected animals expressed higher levels of the cytokine compared to secondarily infected horses especially at 7, 21, 28 and 42 DPI ( Figure 5 ). Notably, significant expression of TGF-β1 was observed at 7 and 14 DPI in the primarily and secondarily infected groups respectively (* p<0.0213). Like IL-10, the highest levels of TGF-β1 expression in both groups were observed at 14 DPI which coincided with the highest parasite load ( Figure 3 ).

IL-10 is a potent anti-inflammatory cytokine that limits the detrimental effects of host pro-inflammatory responses to pathogens thus averting tissue damage (32). The cytokine was originally considered to be primarily produced by CD4 T_H2 and CD4 T regulatory (T_reg) cell subset but recent studies have shown that it can also be produced by many categories of leukocytes, with the major sources being T helper cells, monocytes, macrophages and dendritic cells (32). Studies have also indicated that IL-10 is a key molecule in underpinning B cell differentiation, proliferation and antibody production thus promoting humoral immune responses (33).

TGF-β regulates several T cell functions including inhibition of the differentiation of T helper 1 (Th1) and T helper 2 (Th2) cells and promotion of development of T_reg and T follicular helper (T_fh) cells (34). These are essential in promoting anti-inflammatory responses (35), and antibody responses (36), respectively.

Since pro-inflammatory cytokines were not consistently expressed in this study, and very low levels were observed in cases where expression occurred, it is likely that the high levels of IL10 and TGF-β1 suppressed the pro-inflammatory cytokines to below detection levels. There is also the probability that expression of the anti-inflammatory cytokines was not associated with modulation of pro-inflammatory responses. In this regard, we investigated the possibility of an association between expression of the anti-inflammatory cytokines and development of antibodies by evaluating the production of antibodies against T. equi EMA-1 protein. Our findings showed significant production of antibodies against T. equi during primary infection at 21 to 49 DPI ( Figure 6A ). As expected, antibodies against EMA-1 were present in the secondary infected group at 0 DPI and no significant increase in antibody production was observed in this group except for a significant decline at 21 DPI ( Figure 6B ). Existence of a correlation between IL-10 and TGF-β expression, and EMA-1 antibody production was observed, with increasing antibody levels coinciding with high expression of the cytokines ( Figures 6A, B ) and parasite DNA levels at 14 DPI. This is consistent with our previous findings which showed that significant antibody responses against the rhoptry-associated proteins were observed in these animals, and high levels were also observed at 14 DPI (13). The direct relationship between high parasite load and expression of IL-10 and TGF-β further suggests that the cytokines were produced in response to the high level of parasites thus leading to antibody production and subsequent reduction in the parasite load.

Collectively, unlike the findings observed in studies involving transforming Theileria species (23, 37), significant pro-inflammatory responses were not observed in this study. Additionally, our findings were contrary to the observations made on serological levels of cytokines during natural T. equi infection (27). Notably, this could be due to the fact that this study focused solely on the merozoite stages of T. equi development thus suggesting that the cytokine responses observed by Mostafavi et al. could be associated with schizogony and thus further studies are required to ascertain if this is the case.

The findings obtained from this study provide insight into the relationship between T. equi merozoites and its hosts by highlighting the cell mediated immune responses that occur in response to the parasite. The study demonstrates that merozoites trigger substantial anti-inflammatory responses and minimal pro-inflammatory responses, strongly suggesting that the symptoms observed during primary exposure to the parasite can be associated with hemolysis and subsequent anemia, but not pro-inflammatory responses. It is worth noting that even though hematocrit was stable in secondarily infected horses, the slight drop observed in one of the horses suggests that some carrier animals that are clinically inapparent might not exhibit a very strong immune response thus posing challenges to the equine industry. The host cellular immune response to parasites that transiently infect nucleated equid host cells such as T. equi and related Theileria-like apicomplexan parasites is technically difficult to investigate. However, from the current study it seems clear that expression of IL-10 and TGF-β in the merozoite stages of infection are very likely associated with stimulation of antibody production, but not dampening of pro-inflammatory immune responses even though both pro- and anti- inflammatory cytokines were co-expressed in some cases. Notably, it is likely that the changes in cytokine levels could have occurred due to cell activation. Further studies are therefore required to confirm if this could be a factor in this study. Additional studies are also required to evaluate recall of cytokine responses by re-stimulating immune cells from T. equi-infected horses in vitro.