Vaccine-linked chemotherapy induces IL-17 production and reduces cardiac pathology during acute Trypanosoma cruzi infection

Chagas disease resulting from Trypanosoma cruzi infection leads to a silent, long-lasting chronic neglected tropical disease affecting the poorest and underserved populations around the world. Antiparasitic treatment with benznidazole does not prevent disease progression or death in patients with established cardiac disease. Our consortium is developing a therapeutic vaccine based on the T. cruzi flagellar—derived antigen Tc24-C4 formulated with a Toll-like receptor 4 agonist adjuvant, to complement existing chemotherapy and improve treatment efficacy. Here we demonstrate that therapeutic treatment of acutely infected mice with a reduced dose of benznidazole concurrently with vaccine treatment – also known as “vaccine-linked chemotherapy”—induced a TH17 like immune response, with significantly increased production of antigen specific IL-17A, IL-23 and IL-22, and CD8 + T lymphocytes, as well as significantly increased T. cruzi specific IFNγ-producing CD4 + T lymphocytes. Significantly reduced cardiac inflammation, fibrosis, and parasite burdens and improved survival were achieved by vaccine-linked chemotherapy and individual treatments. Importantly, low dose treatments were comparably efficacious to high dose treatments, demonstrating potential dose sparing effects. We conclude that through induction of TH17 immune responses vaccine-linked chemotherapeutic strategies could bridge the tolerability and efficacy gaps of current drug treatment in Chagasic patients.


Results
Concurrent Tc24-C4 vaccine with BZN increases survival and reduces parasite burdens. To evaluate protection given by the concurrent Tc24-C4 vaccine in combination with BZN, we monitored survival and collected blood to measure parasitemia until 53 days of infection (53dpi). As expected, infected untreated control mice exhibited an acute lethal infection with only 40% survival at 53 dpi. Each treated-experimental group had significantly increased survival compared with the untreated control group (Table 1). Importantly, both groups that received the concurrent vaccine with BZN, using either a high or low-dose vaccine, had similar survival (100%) as the control groups treated with the curative or low-dose of BZN. Reduced but still high survival rates (90%) were observed in mice treated with both high and low doses of E6020-SE alone. The group treated with the low-dose vaccine alone had 80% survival rate, which did not reach statistical significance compared with the infected untreated control group (Table 1). Parasitemia, as measured by the area under the curve Scientific Reports | (2021) 11:3222 | https://doi.org/10.1038/s41598-021-82930-w www.nature.com/scientificreports/ (AUC) was also significantly reduced by both low and high dose vaccine + BZN treatment as well as low dose BZN and high dose BZN alone (Fig. 1A,B). Importantly, both low and high dose vaccine + BZN groups had significantly lower AUC when compared to low dose BZN alone, suggesting a synergistic effect of the combined treatment (Fig. 1A,B). Further, both low and high dose vaccine + BZN groups showed the same levels of protection as the group receiving high dose BZN. While mice vaccinated with either a high or low-dose of the vaccine alone or E6020-SE alone showed a trend to a lower parasitemia compared to infected untreated controls, these reductions did not reach statistical significance (Fig. 1A,B). Supporting the reduced parasitemia, mice treated with both low and high dose vaccine + BZN had significantly lower cardiac parasite burdens compared to the infected untreated mice, and compared to mice treated with low dose BZN alone (Fig. 1C). In contrast, the groups treated with low-dose of the vaccine, low-dose BZN, high and low-dose E6020 failed to reduce the car- Table 1. Survival to T. cruzi acute infection. Mice were infected with 500 trypomastigotes of T. cruzi and followed up daily for up to 53 days post-infection. Percentages of survival were analyzed using Log-rank (Mantel-Cox) test comparing treated groups vs infected untreated group and statistical differences were considered when P values * P ≤ 0.05, ** P ≤ 0.005. Concurrent Tc24-C4 vaccine with BZN protects mice against acute cardiac inflammation and fibrosis. We also evaluated the potential of vaccine-linked chemotherapy to protect against tissue damage caused by T. cruzi acute infection by measuring cardiac inflammatory cell inflammation and fibrosis. The percentage of the inflammatory cell infiltrate in the heart of mice treated with either a high-or low-dose of vaccine + BZN, as well as high dose vaccine alone, was significantly lower compared with the infected untreated control group (Fig. 1D). Concurrently, fibrosis was significantly reduced by all experimental treatments except low-dose E6020 , when compared to infected untreated control mice (Fig. 1E). Together, these results suggest that vaccine-linked chemotherapy also protects against cardiac damage caused by acute T. cruzi acute infection.

Concurrent Tc24-C4 vaccine with BZN induces a robust T cell-mediated immune response.
To evaluate the participation T cell subsets induced by vaccine-linked chemotherapy in the control of acute T. cruzi infection spleen cells harvested from survivor mice were stimulated in vitro with Tc24-C4 protein or T. cruzi parasite lysate, stained with fluorescent antibodies and analyzed by flow cytometry, as described. Results indicated that high-dose vaccine + BZN, high-dose vaccine alone, and low-dose vaccine alone induced a significantly higher percentage of Tc24-C4 specific CD8 + T cells compared to infected untreated control mice ( Fig. 2A). Moreover, mice treated with low-dose vaccine + BZN had a significantly higher percentage of parasite-specific CD4 + IFNγ producing-T cells compared to infected untreated control mice (Fig. 2B). There were no significant differences in IFNγ producing-CD8 + T cells between the infected untreated mice and any of the treatment groups (data not shown). Thus, our data demonstrate that our vaccine, alone or in combination with low dose BZN treatment, induces a strong antigen-specific T cell response.
Concurrent Tc24-C4 vaccine with BZN elicited a robust Tc24-C4-specific T H 17 like immune response. We previously showed that vaccine-linked chemotherapy induces a balanced antigen specific T H 1/T H 2 cytokine profile in acutely infected mice 36 . Here we further investigated the antigen specific T H 17 cytokine profile induced by vaccine-linked chemotherapy. We found that mice treated with a low-dose of vaccine + BZN had significantly higher levels of Tc24-C4 specific IL-17, IL-22, IL-23, IL-6, TNFα and IL-5 production when compared to infected untreated controls, and significantly increased IL-17 and TNFα when compared to low dose vaccine alone, suggesting increased T H 17 like responses in addition to T H 1 like and T H 2 like responses (Fig. 3A). Likewise, mice receiving a low-dose of the vaccine alone had significantly higher levels of IL-23, TNF-α, IL-5 and a trend toward high levels of IL-17A and IL-22. High-dose vaccine + BZN treatment induced significantly increased Tc24-C4 specific IL-23, IL-21, IL-6, IL-4, IL-5, IL-10 and TNFα when compared to infected untreated control mice, and increased IL-17 and TNFα when compared to high dose vaccine alone (Fig. 3B). In contrast, high dose vaccine + BZN induced lower levels of IL-17 and IL-22 compered to vaccine alone (Fig. 3B). Described previously as a regulatory cytokine, IL-10 was significantly increased in infected untreated mice compared to mice treated with high-dose vaccine + BZN (Fig. 3B). Additionally, mice treated with low dose vaccine + BZN showed significantly higher levels of T H 17 (IL-17A, IL-22) and T H 2 (IL-4, IL-5, IL-10) when compared to mice treated with low-dose BZN alone (Fig. 3C), while mice treated with high dose www.nature.com/scientificreports/ vaccine + BZN showed increased IL-4, IL-5, and IL-6 when compared to low dose BZN alone (Fig. 3D). Conversely, the cytokine IL-23 was significantly higher in the group treated with low-dose BZN alone compared to vaccine + BZN ( Fig. 3C-D). Levels of IFNγ and TNF-α were found similar between mice treated with either low or high-dose vaccine + BZN and low-dose BZN alone ( Fig. 3C-D) As expected, treatment with low-dose BZN alone did not induce Tc24-C4 specific T H 2 cytokine release ( Fig. 3C-D).

Concurrent Tc24-C4 vaccine with BZN induced a robust total parasite T H 17 cytokine response.
We also evaluated and compared the cytokine release response against T. cruzi total parasite antigens. As expected, infected untreated mice showed a T H 2 biased cytokine profile with higher levels of IL-10, IL-5, IL-4 in contrast with IL-17A, IL-23, IL-22 and IL-12p70 levels that were undetectable ( Fig. 3E-F). Also, higher levels of IFNγ were observed in the infected untreated group compared to all other treatment groups ( Fig. 3E-F). Mice treated with a low-dose vaccine + BZN showed significantly lower levels of T H 2 cytokines (IL-4, IL-5, and IL-10) as well as higher levels of IL-17A, IL-23 and IL-22 compared to the infected untreated group (Fig. 3E). Interestingly, mice treated with high-dose vaccine + BZN showed a T H 17 like immune response with significantly higher levels of IL-17A, IL-23 and IL-22 and significantly lower levels of IL-10 and IL-4 ( Fig. 3E). Radars graphs showing the total parasite-specific cytokine patterns from the groups treated with low or highdose E6020-SE alone showed an overall T H 2 biased profile with increased IL-4, IL-5 and IL-10 cytokines, but with significantly lower levels compared to infected untreated mice ( Supplementary Fig. 1A). Tc24-C4 specific IL-22 was significantly reduced and IL-23 was significantly increased in mice treated with high-dose E6020-SE alone compared to infected untreated mice ( Supplementary Fig. 1B). As expected, the groups treated with E6020 alone had higher levels of soluble IFNγ compared to mice infected untreated. Nevertheless, the T H 17 profile in those mice was slightly increased (IL-21 and IL-22 compared to infected untreated mice ( Supplementary Fig. 1).

The T H 17 like cytokine response strongly correlated with reduced parasite burdens. To inves-
tigate if parasite burdens were associated with specific cytokines, we performed a correlation analysis comparing parasitemia and cardiac parasite burden with the T H 17 cytokines IL-17A, IL-21, IL-22, and IL-23. Importantly, a significant negative correlation (r = -0.75, P-value = 0.030) was found between cardiac parasite burdens and IL-22 in mice treated with the high-dose vaccine + BZN (Fig. 4A). Further, significant negative correlations were found in mice treated with the curative dose of BZN between IL-17A and both parasitemia and cardiac parasite burden (r = -0.826, P-value Spearman 0.013; r = -0.86, P < 0.0001 respectively) ( Fig. 4B and C, respectively), as well as between the cytokine IL-22 and both parasitemia and cardiac parasite burden (r = -0.812, P-value Spearman 0.010; r = -0.764, P < 0.0001 respectively) ( Fig. 4D and E, respectively). To begin to explore the correlation between CD4 + and CD8 + cytokine producing cells and secreted cytokines from cultured splenocytes, we performed a correlation analysis between the percent of CD4 + IFNγ + cells or CD8 + IFNγ + cells measured by flow cytometry and individual secreted cytokines measured by Luminex. We found that mice treated with low dose vaccine + BZN had a significant negative correlation with secreted TNFα (Supplementary Fig. 2A). In contrast, mice treated with high dose vaccine + BZN had a significant positive correlation with secreted TNFα ( Supplementary Fig. 2B). No other significant correlations were identified between specific secreted cytokines or cytokine producing T cells and cardiac parasites or parasitemia.

Discussion
Vaccine linked-chemotherapy is a promising strategy combining the beneficial effects of both vaccine immunotherapy and an antiparasitic drug to control acute T. cruzi infection 36 . In this study, we demonstrated that either high or low-dose vaccine given concurrently with low-dose BZN significantly reduced cardiac parasite burdens, parasitemia, cardiac inflammatory infiltrate and cardiac fibrosis in acutely infected mice. Further, concurrent vaccine + BZN treatment induced antigen-specific cytokines indicative of T H 17 like immune responses, in addition to T H 1 and T H 2 like immune responses. These data confirm and expand on our prior results, where by using a sequential vaccine-linked chemotherapy strategy with treatment initiated on 7 days post infection, parasitemia was reduced by 24% and the secreted cytokine profile from splenocytes suggested a balanced T H 1/ T H 2 immune response 36 . Here we improved on that result, achieving a 98.73% reduction in parasitemia by concurrently administering vaccines and low-dose of BZN starting 7 days post-infection. We recently reported on the partially protective effect of TLR4 agonist adjuvant in acutely infected mice mediated in part by reduction of tissue parasite burdens 39 . Here we confirmed our prior results showing partial efficacy of individual treatments with benznidazole, vaccine, or the TLR4 agonist adjuvant E6020 alone as evidenced by significantly increased survival, reduced parasite burdens and reduced fibrosis. Importantly in this study only vaccine + BZN at both high and low doses and high dose vaccine alone significantly reduced cardiac inflammatory infiltrate, suggesting that both vaccine and low dose benznidazole are needed to achieve maximal synergistic effects. However, due to the partial protection afforded by each individual treatment additional studies combining specific components, such as E6020 SE + BZN, would be required to further define the relative contribution and synergism of the Tc24 C4 antigen, the TRL4 agonist adjuvant, and benznidazole on vaccine-linked chemotherapy efficacy. We also showed that the effects of high and low dose vaccine + BZN on parasite burdens and tissue pathology were roughly equivalent, indicating that the low dose vaccine + BZN could be advanced in further testing as a dose sparing strategy. Parasite persistence has been demonstrated to drive the persistent low grade inflammation and post-inflammatory fibrosis of symptomatic chronic Chagas disease, which results in pathologic cardiac remodeling, and dysfunctional electrical conduction 13,40,41 . Indeed, myocardial fibrosis in humans determined by non-invasive techniques like magnetic resonance imaging was correlated to severe heart disease 14,42 . Recently, we also showed in a mouse model of chronic infection that cardiac fibrosis correlated with increased cardiac strain on MRI, and www.nature.com/scientificreports/ increased serum levels of the pro-fibrotic biomarkers TGFβ, platelet derived growth factor-D (PDGF-D), and connective tissue growth factor (CTGF) 43 . Our group has previously shown that therapeutic vaccination with a high dose prototype vaccine (100 µg Tc24 + 25 µg E6020-SE) significantly reduced cardiac fibrosis in chronically infected mice 35 . Building on these findings, ongoing studies are evaluating the effect of low dose vaccine-linked chemotherapy in a chronic infection model as a dose sparing strategy to improve cardiac health. Evaluating cardiac function using non-invasive cardiac imaging, such as echocardiography, as well as evaluating general animal health in the ongoing chronic studies will be key to determine potential clinical improvement, and may suggest specific parameter for monitoring clinical outcomes in future human clinical trials. We measured the major cytokines representing T H 17, T H 1, and T H 2 immune responses to define their participation in the protection against T. cruzi infection conferred by vaccine + BZN. Vaccine-linked chemotherapy induced a T H 17 like profile, with significantly higher levels of IL-17, IL-22, and IL-23 compared to infected untreated controls. Further, low dose vaccine + BZN induced increased IL-17A secretion when compared to low dose vaccine alone. However, high dose vaccine + BZN induced IL-17A levels that were lower than high dose vaccine alone. Recent findings have shown that the cytokine IL-17A in particular is critical for eliciting a T H 17 immune response, and is necessary for T. cruzi host protection 16 . IL-17A has been shown to regulate cardiac inflammation in experimentally infected animals mediated by neutrophil derived IL-10, and IL-17RA signaling is necessary for survival of parasite specific CD8 + T cells 37,38,44 . Further, it has been confirmed in Chagasic patients that increased IL-17A expression and increased CD4 + IL-17A producing cells correlate with less severe cardiac disease 45,46 . Thus, we conclude that the decreased cardiac inflammatory infiltrate and parasite burdens and increased antigen specific CD8 + cells is mediated in part by increased IL-17 induced by vaccine-linked chemotherapy. The interesting finding of reduced IL-17 secretion in the high dose vaccine + BZN treatment compared to high dose vaccine alone may be due to further reductions in parasite burdens from the higher doses which may in turn reduce the parasite stimulus for increased IL-17. Additional studies exploring different doses of vaccine and BZN would be necessary to elucidate this.
The role of the cytokine IL-23 in T. cruzi infection is poorly described, however, its participation in bacterial infections such as Listeria monocytogenes is critical for neutrophil recruitment mediated also by IL-17 receptor A (IL-17RA) in an IFNγ independent manner 47 . Moreover, both cytokines (IL-17A/IL-23) have been shown to have a protective role during other vacuole-bound bacteria and Toxoplasma gondii infections 48,49 . Thus, our findings indicate that resistance to T. cruzi infection might be associated to IL-17A/IL-23 axis. Another T H 17 cytokine that could be associated with protection against T. cruzi is IL-22. Besides its protective role in T. cruzi immunity being not well elucidated, efforts have been made to define its association with parasite genetic diversity 50 or pathogenicity 51 . Results from the latter study reported that lower total serum levels of IL-22 in mice were associated with intestinal damage during chronic T. cruzi-infection. Potentially, the consistently increased levels of either Tc24-C4 specific or total T. cruzi parasite-specific IL-22 from mice treated with vaccine + BZN, in sharp contrast with undetectable levels from infected untreated mice, would be strongly related to protection. Additionally, increased IL-22 levels from mice therapeutically treated with vaccine + BZN significantly correlated with decreased T. cruzi parasite burdens in cardiac tissue. Even though associating a single cytokine with clinical outcomes might be difficult due to their pleiotropic expression, these data suggest IL-22 may be a useful biomarker for treatment efficacy 50 . Based on previous studies in vitro, a T H 17 immune profile seems to be paramount for control of acute T. cruzi infection, likely through NADPH oxidase mechanisms with the collaboration of IL-21 and IL-22 16 .  www.nature.com/scientificreports/ In addition to demonstrating the T H 17 like responses induced by vaccine-linked chemotherapy, we also confirmed that this strategy induces T H 1 and T H 2 like immune responses. The low dose vaccine + BZN treatment induced significantly increased antigen specific CD8 + cells, total parasite CD4 + IFNγ + cells, and the cytokines TNFα, and IL-5 suggesting a mixed T H 1/T H 2 response. The T H 1 cytokines TNFα, IL-12, and IFNγ, are wellknown to be necessary for activation of antigen-presenting cells (APCs), and stimulation of inducible Nitric Oxide Synthase (iNOS), a critical killing mechanism for intracellular T. cruzi parasites [52][53][54][55] . In addition, CD8 + T lymphocytes are essential for control of T. cruzi infection, and both vaccines and benznidazole treatment have been shown to induce protective CD8 + T lymphocyte populations 56,57 . Together, these data suggest that the mechanism of protection afforded by our vaccine-linked chemotherapy strategy can also be attributed to the induction of T H 1 and T H 2 driven effector mechanisms.
A discordant finding in this study was that neither antigen-specific nor total parasite IFNγ release was significantly increased in our vaccine + BZN concurrent treatments. IFNγ plays an essential role in the recruitment T cells in the initial stage of T. cruzi infection and is critical for protective immunity in both mice and humans 34,35,58,59 . However, overexpression leads to undesirable inflammatory effects 60 . IL-10 has been described as a T H 2 regulatory cytokine that inhibits MHC class II expression and excessive production of IFNγ and TNFα 61 . Here we did not observe significantly increased IL-10 secretion induced by vaccine alone or vaccine + BZN when compared to infected untreated mice. In fact, IL-10 secretion appeared to be slightly reduced, although this decrease was not statistically significant. One explanation for this result is that vaccine-linked chemotherapy controlled both parasite burdens and cardiac inflammation such that by 53 days post infection, the immune response is reflective of a balanced, healing response. Another explanation is that in this model, the antigen specific T H 17 immune response may contribute more to controlling cardiac pathology compared to IL-10. Still another possibility is that under some as yet undefined circumstance in mice or related to dose and formulation, the vaccine might protect through either T H 17 or balanced T H 1/T H 2 pathways. Further studies are necessary to identify and characterize the specific cells involved in both parasite control and cardiac pathology, including cardiac specific lymphocytes. In contrast to IFNγ, both low and high dose vaccines, either alone or in combination with BZN treatment, did induce significantly increased levels of IL-6. IL-6 is a pleiotrophic cytokine that is essential for parasite specific immune responses and resistance to T. cruzi infection 62 . However, IL-6 is also known to promote fibrosis in the setting of extensive inflammation 63 . Since we observed reduced fibrosis and inflammation in our model, it is likely that the anti-inflammatory effects of the vaccine reduced the expected www.nature.com/scientificreports/ biologic impact of this cytokine. Future studies will further evaluate the relationship between treatment induced cytokines and pro-fibrotic biomarkers to determine their relative contribution to cardiac pathology. In summary, we demonstrated that a concurrent vaccine-linked chemotherapy treatment strategy can achieve protection against acute T. cruzi infection by immune mediated control. The antigen specific cytokine profile suggests a balanced immune response, with T H 1 like, T H 2 like and T H 17 like responses. The protective immune response induced by our concurrent vaccine-linked chemotherapy strategy is similar to other protective vaccines [64][65][66] . Further, the comparable efficacy of the low dose vaccine + BZN to the high dose vaccine + BZN indicates that there is no practical benefit to using higher doses. This work provides key proof of concept for employing multi-modal treatment options to bridge the efficacy and tolerability gaps of benznidazole treatment alone, but additional knowledge of the mechanisms of disease will assist in further optimizing treatment strategies. Ongoing studies in our lab using vaccine-linked chemotherapy in chronically infected mice are exploring the effect of this strategy on cardiac pathology, including inflammation and fibrosis, as well as cardiac function and liver health as measures of improved clinical outcomes. Additionally, defining functional differences in immune cells within the periphery, i.e., antigen specific CD8 + IFNγ + , CD4 + IFNγ, and CD4 + IL-17A + cells, will better define the immune response induced by vaccine-linked chemotherapy. Further, defining the cardiac specific inflammatory response, for example by staining for inflammatory markers by immunohistochemistry, will provide critical information about the mechanisms of disease that correlate with cardiac health. Together, these data will ultimately assist in the identification of additional targets for therapeutic intervention.

Material and methods
Ethics statement. All studies were approved by the Institutional Animal Care and Use Committee of Baylor College of Medicine described under assurance numbers D16-00,475 (current) and 3823-01 (previous) and were performed in strict compliance with the 8 th Edition of The Guide for Care and Use of Laboratory Animals 67 .
Parasite and mice. T. cruzi H1 parasites, originally isolated from a human case in Yucatan, Mexico were maintained by serial passage in mice every 25 to 28 days until infection 68 . Ninety female BALB/c mice were obtained at 6-week-old from Taconic vendor (Taconic Biosciences, Inc) and housed in groups of 5 animals, with ad libitum food and water and a 12 h light/dark cycle. Mice were allowed to acclimate for one week prior to studies.
Parasite lysate. Parasite lysates was prepared using 10 6 epimastigotes T. cruzi H1 obtained from logarithm grown phase of epimastigotes cultured in LIT media at 28 °C. The parasites were disrupted in PBS containing protease inhibitor cocktail (cOmplete ultra-tablets, Roche) by three freeze-thaw cycles. Further, parasites were sonicated for 15 s at the minimum setting three times and then centrifuged at 15,000 g for 30 min. The soluble fraction was aliquoted and stored at -80 °C until use. Protein concentration was determined by BCA protein assay kit (ThermoFisher) according to manufacturer's instructions.
Immunotherapeutic treatment. All mice were infected with 500 blood form T. cruzi H1 trypomastigotes by intraperitoneal injection. Infected mice were randomly divided into nine experimental groups of 10 mice each ( Table 2). The recombinant Tc24-C4 antigen was expressed and purified according to previously published work 26 . The adjuvant E6020 was dissolved in a 2% stable squalene emulsion (SE), acquired through Eisai Inc. Vaccine formulations were freshly prepared and mixed just before injection. Prepared vaccines were injected subcutaneously on day 7 and boosted at day 14 post-infection. BZN powder (ELEA Laboratory) was reconstituted in solution with 95% HPMC (0.5% hydroxypropylmethylcellulose/ 0.4% Tween 80/ 0.5% benzyl alcohol in deionized water) and 5% DMSO to a final concentration of 10 mg/mL and prepared daily, immediately before oral administration to mice. Mice receiving low-dose (25 mg/kg) daily oral BZN (BZN) were treated for 1 week from 7 days until 14 days post-infection. Mice receiving high-dose (100 mg/kg) daily oral BZN were treated for 20 days from 7 days until 26 days post-infection. Survival was recorded daily throughout the experiment and Table 2. Experimental groups. All mice were infected with 500 T. cruzi H1 by intraperitoneal injection. Treated mice were prime-immunized subcutaneously on day 7 and boosted at day 14 post-infection. BZN was administered by oral administration once daily beginning 7 days post-infection.  36 . Briefly, total DNA was isolated from blood and cardiac tissue using a DNEasy blood and tissue kit (Qiagen) and 4 ng of DNA from blood, or 50 ng of DNA from cardiac tissue was used per reaction. The oligonucleotides primers used to amplify the satellite region of T. cruzi nuclear DNA were forward 5′ AST CGG CTG ATCDTTT TCG A 3′ and reverse 5′ AAT TCC TCC AAG CAG CGG ATA 3′. The PCR reaction included a sequence-specific, fluorescently oligonucleotide probe 5′ 6-FAMCAC ACA CTG GAC ACC AAMGB 3′. Parasite equivalents were calculated based on a standard curve and data was normalized to GAPDH (primers 5′ CAA TGT GTC CGT CGT GGA TCT 3′ and 5′ GTC CTC AGT GTA GCC CAA GATG 3′, probe 5′ 6-FAM CGT GCC GCC TGG AGA AAC CTGCC MGB 3′, Life Technologies).

Quantification of inflammatory infiltrate cells and cardiac fibrosis.
Heart tissue samples were fixed in 10% neutral buffered formalin and routinely processed for paraffin embedding, sectioning and staining with Hematoxylin and Eosin, or Masson's Trichrome for inflammation infiltrate or fibrosis measurement, respectively. For inflammation analysis, five representative pictures from each slide stained were captured at 10X magnification using and Amscope ME580 bright field microscope. For inflammatory infiltrate cells the images were analyzed using FIJI software (National Institute of Health) based on the number of pixels corresponding to cells nuclei to estimate the number of inflammatory infiltrate cells per area 69 . To quantify fibrosis, five representative pictures from each slide stained were captured at 20X magnification with Micron software in Micromaster microscope and the percentage of the blue-colored area corresponding to collagen also was determined using FIJI software (National Institute of Health).
Spleen processing and cultures. At day 53, post-infection mice were euthanized and spleens were disrupted mechanically using a 100 µm cell strainer in presence of RPMI supplemented with 10% fetal bovine serum, antibiotics (1X penicillin-streptomycin) and L-Glutamine. Red blood cells were lysed using ammoniumchloride potassium (ACK) lysis buffer. Spleen cells were adjusted in 5 mL of RPMI supplemented and viability determined by acridine orange and propidium iodide (AOPI) live/dead dye using a Cellometer Auto 2000 cell counter (Nexcelom Biosciences, Lawrence, MA). Spleen cells were incubated at 1 × 10 6 live-cells per well in a round-bottom 96-well plate and stimulated either with 100 µg/mL recombinant Tc24-C4 protein, 5 µg/mL T. cruzi parasite lysate, 20 ng/mL PMA-1 mg/mL Ionomycin as a positive control, or media only for 96 h at 37 °C with 5% CO 2 36 . After incubation, supernatants were obtained from cell cultures by centrifugation at 300 g for 5 min and then aliquoted and frozen until use.

T cell cytokines analysis by flow cytometry.
To determine CD4 + and CD8 + lymphocytes sub-populations in antigen re-stimulated splenocytes were washed twice with RPMI and surface staining was performed using antibodies anti-CD3 fluorescein isothiocyanate (FITC) clone 145-2C11 (eBioscience), anti-CD8 peridinin chlorophyll protein (PerCP-Cy5.5) clone 53-6.7 (BD Bioscience), and anti-CD4 Alexa Fluor® 700 clone RM4-5 (eBioscience). For intracellular cytokine staining, cells were incubated with 4.1 µg/mL brefeldin A for the last four hours of incubation. After surface staining cells were fixed and permeabilized with BD Cytofix/Cytoperm (BD, Bio Sciences) and stained with anti-IFNγ allophycocyanin (APC) clone XMG1.2 (eBioscience) 59 . Samples were acquired on the analyzer LSR Fortessa and at least 100,000 total events were analyzed using FlowJo 8.7 software. Cells were gated on forward and side scatter for lymphocytes, exclusion of viability dye, and singlet populations ( Supplementary Fig. 3A-C, respectively). CD3 + ( Supplementary Fig. 3D) , CD8 + and CD4 + gates (Supplementary Fig. 3E) were determined based on surface stained cells. Quadrant gating was used to determine the percentage of CD8 + IL-4 + ( Supplementary Fig. 3F), CD4 + IL-4 + ( Supplementary Fig. 3G), CD4 + IFNγ + (Supplementary Fig. 3H) and CD8 + IFNγ + cells (( Supplementary Fig. 3I). The percentage of antigen specific or total parasite specific cells for each target was calculated by subtracting the percent of media stimulated cells from the percent of antigen-stimulated cells for each mouse. In cases where the antigen stimulated cells was lower than media alone, the percentage is expressed as a negative number.

Measurement of cytokines.
Antigen-specific cytokines release from re-stimulated spleen cells were measured using a multiplexed bead-based assay Milliplex MAP Mouse T H 17 (Millipore) and Luminex technology. Supernatant IL-10, IL-5, IL-4, IL-6, IL-21, IL-22, IL-23, IL-17A, IFNγ, IL-12p70 and TNF-α cytokines were measured following manufacturer protocol. Assay was adapted in 96-well wall-less plate format 70 . Briefly, 25 µL of the sample was placed in drops over the magnetic plate and incubated overnight at 4 °C with antibodyimmobilized beads. The plates were washed 3 times and incubated with detection antibodies by 1 h at room temperature (RT). To reveal the plates were incubated 30 min RT with streptavidin-phycoerythrin as substrate and placed in Luminex Magpix Reader® and data was analyzed with Bio-plex Manager 6.0 and graphed into radar plot using R software. Statistical analysis. Parasitemia throughout the experiment was depicted with the area under curve (AUC). Radar plots were created using package radarchart in R interface 71 . For this, cytokines concentrations data was normalized by applying log base 10 transformation. Transformed values were summarized using the