Rhesus monkey kidney epithelial cells (LLC-MK2 cells) (ATCC) were routinely cultured in high-glucose Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% fetal bovine serum (FBS) (Thermo Fisher, Waltham, MA, USA) (DMEM10) at 37°C and 5% CO₂. These cells were infected with the Trypanosoma cruzi Y stain in high-glucose DMEM, supplemented with 2% FBS (DMEM2) (Thermo Fisher) to maintain the parasite.

Supernatant of T. cruzi-infected LLC-MK2 cells was collected, centrifuged in a 15-ml tube (Corning, Corning, NY, USA) at 1,350×g for 10 min and washed twice with DMEM2. After the last centrifugation, the pellet was left in DMEM2 for 2 h at 37°C and 5% CO₂ to allow live trypomastigotes to swim. The supernatant containing live trypomastigotes was collected and filtered in a bottle top 0.22-µm filter (Corning). The filter was washed with DMEM2 to resuspend the live trypomastigotes, which were transferred to a new 15-ml tube and incubated at 56°C for 10 min. Parasite DNA was obtained by incubating approximately 200 million heat-killed trypomastigotes in 500 µl of lysis buffer (Tris.HCl 0.1 M, pH 8.5; EDTA 5 mM, pH 8.0; NaCl 0.2 M, SDS 0.2%, and 100 µg of proteinase K in water) at 37°C and 600 rpm for 18 h, followed by precipitation with isopropanol at 8,600×g for 5 min, washing with ethanol at 70%, centrifugation at 8,600×g for 5 min, and resuspension in 25 µl of DNAse/RNAse-free water. The ratios of absorbance at 260/280 and 260/230 nm were used to assess the purity of DNA with a Nanodrop 2000 (Thermo Fisher). The parasite DNA used in transfection experiments had 260/280 and 260/230 ratios of 1.92 and 2.21, respectively.

RAW264.7-Lucia™ ISG and RAW264.7-Lucia™ ISG-STING-KO macrophages (InvivoGen, Toulouse, France) were plated in 24-well plates (Corning) at a density of 10⁵ cells per well in 500 µl of DMEM2 24 h before infection, exposure to heat-killed trypomastigotes, or transfections. The cells were incubated for 16 h with 3 × 10⁶ live or heat-killed trypomastigotes per well in 300 µl of DMEM2, washed with PBS (Thermo Fisher), and incubated for additional 24 h in 300 µl of DMEM2. Supernatant was collected and total RNA extracted. Alternatively, the cells were transfected with 80 ng of parasite DNA complexed with lipofectamine 2000 (Thermo Fisher) in 300 µl of OPTIMEM (Thermo Fisher) per well, accordingly to manufacturer’s instructions. As experimental controls, the cells were transfected either with 1.0 μg/ml of c-di-GMP (InvivoGen) or 0.5 μg/ml of Poly I:C (InvivoGen) complexed with lipofectamine 2000 in 300 µl of OPTIMEM. Lipofectamine 2000 and OPTMEM were used in negative control wells. Supernatant was collected and total RNA extracted 24 h after transfection.

Twenty microliters of supernatant from infected, heat-killed parasite-exposed or transfected RAW264.7-Lucia™ ISG and RAW264.7-Lucia™ ISG-STING-KO macrophages were mixed with 50 μl of QUANTI-Luc™ (InvivoGen) and immediately read in a Smart Line TL luminometer (Titertek Berthold, Pforzheim, Germany), with an acquisition time of 1 s for determination of luciferase activity. The detection of IFN-β, IL-6, and IL-12 cytokines in the supernatant of infected RAW264.7-Lucia™ ISG and RAW264.7-Lucia™ ISG-STING-KO macrophages was performed using the Mouse Custom ProcartaPlex kit (Thermo Fisher), accordingly to manufacturer’s instructions. The samples were read with a MagPix Luminex system (Merck Millipore, Burlington, MA, USA) and analyzed using the Milliplex Analyst software (Merck Millipore).

The study was approved by the Ethics Committee on the Use of Animals (CEUA) of the Faculty of Medicine, University of Sao Paulo (FMUSP), under protocol number 1567/2020, and carried out in accordance with Brazilian Federal Law number 11,794 on scientific use of animals and the National Institutes of Health guide for the care and use of laboratory animals.

Six- to 8-week-old male wild-type BALB/c, wild-type C57BL6, and STING-KO mice with a C57BL6 background were maintained at the Tropical Medicine Institute II, Faculty of Medicine, University of Sao Paulo. BALB/c and C57BL6 mice were purchased from the Faculty of Medicine, University of Sao Paulo. STING-KO mice were kindly provided by Dr. Baber and are derived from his previously described laboratory colony (46). The animals were housed in groups of up to 5 per cage in a room with controlled light and temperature (12 h light/dark cycles, 21°C ± 2°C) and free access to food and water.

The Trypanosoma cruzi Y strain was maintained in BALB/c mice and used to infect wild-type C57BL6 and STING-KO mice. Blood was collected from euthanized BALB/c mice at the peak of infection and centrifuged at 200×g for 10 min. The supernatant was collected, centrifuged at 3,800×g, and the pellet of parasites resuspended in RPMI1640 (Thermo Fisher). Fifty thousand trypomastigotes in 200 µl of RPMI1640 were intraperitoneally injected in each C57BL6 and STING-KO mouse. Parasitemia was monitored by counting the number of trypomastigotes in 5 µl of fresh blood collected from the tail vein as previously described (47).

RAW264.7-Lucia™ ISG and RAW264.7-Lucia™ ISG-STING-KO macrophages were plated and incubated with live trypomastigotes for 16 h, as previously described. The cells were washed with PBS and incubated for an additional 48 h in 300 µl of high-glucose DMEM without phenol red (Nova Biotecnologia, Ribeirao Preto, Brazil). Alternatively, splenocytes from 4-, 7-, and 13-day-infected C57BL6 and STING-KO mice were incubated for 48 h at a density of 5 × 10⁵ cells in 200 µl of high-glucose DMEM without phenol red per well. The supernatant was collected and centrifuged at 15,000×g for 5 min. The Nitric Oxide Assay kit (Thermo Fisher) was used accordingly to manufacturer’s instructions for total nitrate and nitrite detection with an Epoch spectrophotometer (BioTek, Winooski, VT, USA).

Total RNA extraction from RAW264.7 macrophages, hearts, and spleens was performed using Trizol reagent (Thermo Fisher), RNeasy Fibrous Tissue kit (Qiagen, Hilden, Germany), and RNeasy mini kit (Qiagen), respectively. Synthesis of cDNA was performed using the Superscript II Reverse Transcriptase (Thermo Fisher), accordingly to manufacturer’s instructions. Real-time PCR was performed using Power SyBr green master mix (Thermo Fisher) and a QuantStudio 12k thermocycler (Thermo Fisher) with the following parameters: 95°C for 15 min, 40 cycles of 95°C for 15 s, and 60°C for 1 min. The primer sequences were as follows: HPRT1 forward 5′-GTTGGGCTTACCTCACTGCT-3′; HPRT1 reverse 5′-GCAAAAAGCGGTCTGAGGAG-3′; IFN-β forward 5′-TGGGAGATGTCCTCAACTGC-3′; IFN-β reverse 5′-CCAGGCGTAGCTGTTGTACT-3′; IL-6 forward 5′-CCCCAATTTCCAATGCTCTCC-3′; IL-6 reverse 5′-GGATGGTCTTGGTCCTTAGCC-3′; IL-12 forward 5′-GAAGTCCAATGCAAAGGCGG-3′; IL-12 reverse 5′-GAACACATGCCCACTTGCTG-3′; TNF-α forward 5′-ATGGCCTCCCTCTCATCAGT-3′; TNF-α reverse 5′-TTTGCTACGACGTGGGCTAC-3′; CXCL9 forward 5′-CCAAGCCCCAATTGCAACAA-3′; CXCL9 reverse 5′-AGTCCGGATCTAGGCAGGTT-3′; IFN-γ forward 5′-AGCAAGGCGAAAAAGGATGC-3′; IFN-γ reverse 5′-TCATTGAATGCTTGGCGCTG-3′; PRF1 forward 5′-TGGTGGGACTTCAGCTTTCC-3′; PRF1 reverse 5′-GAAAAGGCCCAGGAGGAACA-3′.

For detection of parasite DNA in the hearts of infected animals, we extracted DNA using the FlexiGene Kit (Qiagen), accordingly to manufacturer’s instructions and used previously described primer sequences (48). Real-time quantitative PCR was performed using Power Sybr green Master mix and the Quanti Studio 3 thermocycler (Thermo Fisher). The β-actin gene was used as an endogenous control and the calculation of parasitism in the heart was based on a T. cruzi DNA dilution curve.

TSBK20 peptide (ANYKFTVL-NH2) was synthesized by manual solid phase peptide synthesis on NovaSyn TGR R resin (Merck) using the Fmoc/tBu strategy. 2-(1H-Benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU) (Merck) and N,N-diisopropylethylamine (DIPEA) were used in the coupling reactions and N,N-dimethylformamide (DMF) was used as solvent. Purity (>97%) was determined by RP-HPLC (Shimadzu, Kyoto, Japan). The peptide was resuspended in dimethyl sulfoxide (DMSO) at a stock concentration of 10 mg/ml and used in immunological assays at a concentration of 10 µg/ml.

Spleens from uninfected and 13-day-infected C57BL6 and STING-KO mice were aseptically removed and disrupted using 70 µm Cell Strainer (Corning). Red blood cells were lysed using ACK lysis buffer (Thermo Fisher); the samples were centrifuged at 300×g for 5 min and splenocytes resuspended in R10 medium (RPMI-1640 supplemented with 10% FBS, 2 mM l-glutamine, 1 mM sodium pyruvate, 1% vol/vol nonessential amino acid solution, 1% vol/vol vitamin solution, 40 µg/ml of gentamicin, and 5 × 10⁻⁵ M 2β-mercaptoethanol, all from Thermo Fisher). Splenocytes were plated in 96-well round-bottom plates (Corning) at a density of 0.5 × 10⁶ cells in 200 µl of R10 medium and stimulated with 10 µg/ml of the T. cruzi H2-K^b-restricted peptide TSKB20 in the presence of 5 µg/ml of brefeldin A (BioLegend, San Diego, CA, USA) for 14 h at 37°C and 5% CO₂. DMSO and PMA (50 ng/ml) plus ionomycin (500 ng/ml) (Sigma, St. Louis, MO, USA) were used as negative and positive control stimuli, respectively.

After stimulation, the cells were transferred to 96-well V-bottom plates (Corning), centrifuged at 300×g for 5 min and stained with the monoclonal antibodies anti-CD3 APC-Cy7 (BD Biosciences, Franklin Lakes, NJ, USA), anti-CD4 PerCP (BD Biosciences), and anti-CD8 PE-Cy7 (BD Biosciences) diluted in PBS for 30 min at 4°C. The cells were washed twice with PBS and fixed with BD Cytofix/Cytoperm™, accordingly to manufacturer’s instructions. Thereafter, the cells were washed twice with BD Perm/Wash™ buffer and stained with the monoclonal antibodies anti-IFN-γ APC (BD Biosciences) and anti-Perforin PE (BioLegend) diluted in BD Perm/Wash™ buffer for 30 min at 4°C. The cells were washed twice with BD Perm/Wash™ buffer and resuspended in PBS. The samples were acquired with a FACS Canto II (BD Biosciences) cytometer and analyzed with FlowJo 10 software (BD Biosciences).

Heart samples were fixed in a 10% buffered formalin solution, dehydrated in an increasing concentration of ethanol (Merck), and embedded in paraffin. The blocks were sectioned with a thickness of 5 μm and stained with hematoxylin-eosin (H&E). The pathologist performed blinded histological analysis and provided a score for the intensity of myocarditis, as follows: (0) absence of myocarditis: absence or minimal focal inflammatory infiltrate; (1) mild myocarditis: mild, focal, or multifocal inflammatory infiltrate, with little cardiomyocyte aggression; (2) moderate myocarditis: clear inflammatory infiltrate, predominantly multifocal with occasional diffuse areas (coalescence), with multiple foci of cardiomyocyte aggression; and (3) intense myocarditis: exuberant inflammatory infiltrate, predominantly diffuse, with multiple foci of cardiomyocyte aggression.

The results were analyzed using the Graph Pad Prism 8 software. We used Mann-Whitney U test for comparisons between 2 parameters and two-way ANOVA, Tukey’s, and Bonferroni’s tests for multiple comparisons. Pearson’s correlation coefficient was used for correlation analysis.

Although previous data have suggested that a TBK-1/IRF3-dependent signaling pathway is essential for IFN-β induction in response to T. cruzi infection, the role of STING remained to be determined (45). To address this question, we used RAW264.7-Lucia™ ISG and RAW264.7-Lucia™ ISG-STING-KO macrophages, which are sufficient or deficient for STING expression, respectively, and designed to secrete luciferase into the culture medium in response to activation of IRF-dependent signaling pathways. We incubated these cells with live or heat-killed T. cruzi Y strain trypomastigotes for 16 h, removed residual parasites, and incubated for additional 24 h to collect supernatant and total RNA for evaluation of luciferase activity and gene expression ( Figure 1A ). We observed that although infection of STING-KO macrophages promoted activation of IRF-dependent pathways, this activation was significantly lower than that observed for RAW264.7 ISG macrophages upon T. cruzi infection ( Figure 1B ). On the other hand, no differences in activation of IRF-dependent pathways were found in heat-killed T. cruzi-exposed STING-KO or RAW264.7 ISG macrophages when compared with their respective controls ( Figure 1B ). In line with our luciferase results, we observed significantly lower IFN-β, IL-6, and IL-12 gene expression in STING-KO-infected macrophages when compared with RAW264.7 ISG-infected macrophages, while no induction of these genes was observed in either STING-KO or RAW264.7 ISG macrophages exposed to heat-killed T. cruzi ( Figures 1C–E ). Infection with T. cruzi resulted in similar induction of TNF-α gene expression in STING-KO and RAW264.7 ISG macrophages, but no significant response was observed upon exposure to heat-killed trypomastigotes ( Supplementary Figure S1A ). We also evaluated NO production in response to infection and found no difference when comparing STING-KO and RAW264.7 ISG macrophages ( Supplementary Figure S2A ). Overall, our results indicate that STING deficiency negatively impacts activation of IRF-dependent pathways and cytokine expression in response to Trypanosoma cruzi infection, while heat-killed trypomastigotes failed to promote activation of RAW264.7 macrophages.

DNA from adenovirus 5, herpes simplex virus, Listeria monocytogenes Plasmodium sp., and Leishmania donovani activates STING-dependent signaling in a variety of cells, indicating that STING signaling may have a role in immune responses to multiple pathogens (46, 49–53). Therefore, we hypothesized that STING would be required for cytokine induction in response to T. cruzi DNA. To test our hypothesis, we transfected RAW264.7-Lucia™ ISG and RAW264.7-Lucia™ ISG-STING-KO cells with T. cruzi Y strain DNA, collected supernatant and total RNA 24 h after transfection, and evaluated luciferase activity and gene expression ( Figure 2A ). We found that STING was essential for DNA-mediated activation of IRF-dependent pathways, as STING-KO cells showed significantly lower luciferase activity upon transfection ( Figure 2B ). As observed for infection, STING-KO cells had significantly lower IFN-β, IL-6, and IL-12 gene expression in response to T. cruzi DNA ( Figures 2C–E ). Although STING-KO and RAW264.7 ISG cells had similar TNF-α gene expression upon infection, we found that parasite DNA transfection resulted in increased TNF-α gene expression in RAW264.7 ISG cells when compared with nontransfected control, which was not observed in STING-KO cells ( Supplementary Figure S1B ). To ensure our system was working properly, we transfected RAW264.7 ISG and STING-KO cells with poly IC (TLR3 ligand) and c-di-GMP (STING ligand). We observed that while STING-KO cells were responsive to poly IC, c-di-GMP elicited no luciferase activity or gene expression. In contrast to infection and parasite DNA, c-di-GMP failed to induce IL-6 and IL-12 gene expression while poly IC failed to induce IL-12 gene expression in RAW264.7 ISG cells ( Supplementary Figures S3A–D ). Taken together, these results indicate that T. cruzi DNA activates STING-dependent signaling.

Our in vitro results suggested that STING is required for induction of cytokines involved in T. cruzi immunity. Therefore, we hypothesized that STING-KO mice would have lower immune activation and would be less effective in controlling the parasite during acute infection. To test our hypothesis, we intraperitoneally infected C57BL6 and STING-KO mice ( Figure 3A ) and found that STING-KO mice had significantly higher parasitemia from days 5 to 8 after infection ( Figure 3B ). Real-time PCR analysis showed that both groups of animals had similar amounts of T. cruzi DNA in the heart on days 4 and 7 after infection ( Figure 3C ). However, we observed significantly higher amounts of T. cruzi DNA in the hearts of STING-KO-infected animals at day 13 after infection ( Figure 3C ), indicating that STING-dependent signaling plays a role in parasite control.

We performed heart histological analysis of STING-KO and C57BL6-infected animals and observed no difference in the magnitude of inflammatory infiltration 13 days after infection ( Figures 3D, E ). Real-time PCR analysis of the heart tissue at days 4, 7, and 13 after infection revealed a kinetic increase in the expression of genes related to immune control of the parasite in both groups of animals. Notably, IFN-β gene expression was significantly lower in the hearts of STING-KO mice at days 7 and 13 after infection when compared with C57BL6 mice ( Figure 4A ). IL-12, CXCL9, IFN-γ, and perforin gene expression was significantly lower in the hearts of STING-KO mice 13 days after infection ( Figures 4C–F ). No significant difference in IL-6 gene expression was found in the hearts of STING-KO-infected mice when compared with C57BL6 mice ( Figure 4B ), while significantly lower TNF-α gene expression was observed in the hearts of STING-KO mice 13 days after infection ( Supplementary Figure S1C ), indicating that STING deficiency negatively impacts the expression of key genes related to protection against acute T. cruzi infection.

Given that all immunological genes evaluated in our study were previously shown to contribute to protection against acute T. cruzi infection, we performed correlation analysis to understand whether the magnitude of gene expression would be associated with parasite control. We found that IFN-β, IL-6, and IL-12 gene expression had no correlation with T. cruzi DNA in the heart at day 13 after infection ( Figures 4G–I ). However, CXCL9, IFN-γ, and perforin gene expression was inversely correlated with T. cruzi DNA ( Figures 4J–L ). Moreover, we observed a positive correlation among CXCL9, IFN-γ, and perforin gene expression in the heart of infected animals ( Supplementary Figures S4A–C ), suggesting that CXCL9-mediated recruitment of IFN-γ and perforin-expressing cells may have had a positive impact on parasite control at day 13 after infection.

To have a more systemic view of the immune response to the parasite in the context of STING signaling, we evaluated the spleens of STING-KO- and C57BL6-infected mice at days 4, 7, and 13 after infection. We observed that IFN-β, IL-6 and IL-12 gene expression was higher at day 4 after infection and decreased overtime in both groups of animals ( Figures 5A–C ). Notably, STING-KO-infected mice had significantly lower IFN-β and IL-6 gene expression in the spleen at day 4 after infection when compared with C57BL6-infected mice ( Figures 5A, B ), indicating that STING-dependent signaling may play a role in early induction of key cytokines against T. cruzi. On the other hand, IL-12 gene expression ( Figure 5C ) and TNF-α gene expression were similar in the spleens of STING-KO- and C57BL6-infected mice in all time points ( Supplementary Figure S1D ). We also evaluated NO production by splenocytes and found no significant differences between groups ( Supplementary Figure S2B ), suggesting the involvement of other signaling pathways in immunity to T. cruzi.

To further exploit the role of STING signaling in T. cruzi-driven immunity, we performed flow cytometry using splenocytes to investigate IFN-γ and perforin production by CD8⁺ T cells against a T. cruzi H-2K^b-restricted peptide named TSKB20 ( Figure 5D ). As expected, we found very low numbers of splenic IFN-γ, perforin, and IFN-γ/perforin-producing CD8⁺ T cells in uninfected animals ( Figures 5E–G ). On the other hand, we observed significantly lower numbers of splenic TSKB20-specific IFN-γ and IFN-γ/perforin-producing CD8⁺ T cells in STING-KO-infected mice when compared with C57BL6-infected mice ( Figures 5H, J ), while the numbers of splenic TSKB20-specific CD8⁺ T cells producing only perforin were similar in both groups ( Figure 5I ). Collectively, our results indicate that STING signaling promotes expression of innate cytokines and generation of CD8⁺ T cells against T. cruzi.