cGAS–STING and MyD88 Pathways Synergize in Ly6Chi Monocyte to Promote Streptococcus pneumoniae-Induced Late-Stage Lung IFNγ Production

The cyclic GMP–AMP synthase–stimulator of interferon genes (cGAS–STING) pathway senses DNA and induces type I interferon (IFN) production. Whether and how the STING pathway crosstalk to other innate immune pathways during pathogen infection, however, remains unclear. Here, we showed that STING was needed for Streptococcus pneumoniae-induced late, not early, stage of lung IFNγ production. Using knockout mice, IFNγ reporter mice, intracellular cytokine staining, and adoptive cell transfer, we showed that cGAS–STING-dependent lung IFNγ production was independent of type I IFNs. Furthermore, STING expression in monocyte/monocyte-derived cells governed IFNγ production in the lung via the production of IL-12p70. Surprisingly, DNA stimulation alone could not induce IL-12p70 or IFNγ in Ly6Chi monocyte. The production of IFNγ required the activation by both DNA and heat-killed S. pneumococcus. Accordingly, MyD88−/− monocyte did not generate IL-12p70 or IFNγ. In summary, the cGAS–STING pathway synergizes with the MyD88 pathway in monocyte to promote late-stage lung IFNγ production during pulmonary pneumococcal infection.


INTRODUCTION
During pathogen infections, multiple innate immune signaling pathways are activated. Stimulator of interferon genes (STING) is essential for cytosolic DNA-induced type I interferon (IFN) production but largely dispensable for Toll-like receptors (TLRs) activations (1)(2)(3). Currently, it is not clear if and how the STING pathway crosstalks with another innate immune pathway during infections.
Streptococcus pneumoniae is an extracellular bacterial pathogen that causes pneumonia, sinusitis, otitis media, septicemia, and meningitis (4,5). A recent study found that the STING-mediated cytosolic DNA sensing pathway is activated during pulmonary S. pneumoniae infection (6). However, pneumococcal infection-induced proinflammatory cytokines, including tumor necrosis factor a (TNFa), interleukin (IL)-6, and IL-1b, are largely intact in the STING −/− mice, and the bacterial burden in the lung, spleen, and blood were comparable between STING −/− and wild-type (WT) mice (6). Thus, STING seems to be dispensable for the initial innate immunity to S. pneumoniae including the control of bacterial burden.
IFNg promotes M1-macrophage development that not only phagocyte and kill the bacteria but also contribute to tissue injury. Streptococcus pneumoniae infection induces lung IFNg. In patients with S. pneumoniae sepsis, plasma IFNg was elevated and correlated with increased mortality (7). IFNg −/− mice are more resistant than the WT mice in developing pneumococcal meningitis (8). For pneumococcal pneumonia, IFNg −/− mice or pretreatment of mice with anti-IFNg neutralizing Ab had either no effect on mortality (9,10) or lead to decreased survival (11,12). Thus, the role of lung IFNg production during pulmonary pneumococcal infection remains controversial.
In this report, we found that there were two waves of lung IFNg productions by different immune cells during pulmonary pneumococcal infection. STING is required for the late, not early, stage of lung IFNg production. Notably, the production of IFNg required the activation of both STING and MyD88 pathways indicating previous unknown crosstalk between STING and TLRs pathways during infection.
All mice are on a C57BL/6 background. Mice were housed and bred in the Animal Research Facility at the University of Florida. All experiments with mice were performed by the regulations and approval of the Institutional Animal Care and Use Committee from the University of Florida (protocol number 201909362).

Detection of the Lung Cytokine Production
Mice were intranasally infected with S. pneumoniae D39. Mice were sacrificed by CO 2 asphyxiation at the indicated time points. The lungs were subsequently perfused with cold PBS, washed in PBS once, and stored in a 1.0-ml tissue protein extraction reagent (T-PER) containing protease inhibitors (Roche, Indianapolis, IN). The lungs were homogenized using a Bertin Technology Minilys tissue homogenizer. Lung homogenates were spun at 14,000×g for 15 min at 4°C. The supernatant was collected and analyzed for cytokine production using ELISA.

Flow Cytometry Analysis
Mice were intranasally infected with S. pneumoniae D39. Mice were sacrificed by CO 2 asphyxiation at the indicated time points. The lungs were subsequently perfused with cold PBS. Excised lungs were cut into small pieces and digested in Roswell Park Memorial Institute (RPMI) containing 200 mg/ml DNase I (Roche) and 25 mg/ml Liberase TM (Roche) at 37°C for 2 h. Red blood cells were then lysed using ACK lysis buffer (Gibco), and a single-cell suspension was prepared and analyzed by BD LSR Fortessa flow cytometry.
The following Abs from Biolegend were used in the flow cytometry: Ly6C (HK1.

Intracellular Staining
The intracellular cytokine staining was performed using the Cytofix/Cytoperm ™ kit from BD Biosciences. Briefly, mice were intranasally administered with either S. pneumoniae D39 ∼8-10 × 10 6 CFU in 50 µl of 1× Ultrapure PBS or PBS alone. The lungs were perfused and harvested at 24 and 48 h postinfection and washed in PBS followed by 2 h of digestion in RPMI containing 200 µg/ml DNAse I (Roche), 25 µg/ml Librase TM (Roche), and Golgi-plug 1 µg/µl (BD Bioscience). Digested lungs were processed to prepare the single lung cell suspension in RPMI containing Golgi-plug 1 µg/µl. The cells were fixed in Cytofix/perm buffer (BD Biosciences) in the dark for 20 min at room temperature (RT). Fixed cells were washed and kept in Perm/Wash buffer at 4°C. The Golgi-plug was present during every step before fixation. Cells were stained with cytokinespecific staining antibodies in perm buffer in the dark for 30 min at RT. Cells were washed, and the single-cell suspension was prepared to be analyzed by BD LSRFortessa flow cytometry.

Ex Vivo Monocyte Culture and Activation
Ly6C hi monocytes were purified from bone marrow cells using EasySep Mouse Monocyte Isolation kit (STEMCELL Technologies). Purified monocytes were cultured in RPMI (Invitrogen) with 10% FBS, 2 mM L-glutamine, 1 mM sodium pyruvate, 10 mM HEPES buffer, 1% non-essential amino acids, 50 mM 2-mercaptoethanol, and1% Pen/Strep. Cells were stimulated with 5 × 10 6 CFU/ml heat-killed S. pneumoniae (HKSP) (InvivoGen), 2 µg/ml apoptotic DNA transfected with lipofectamine 3000 (15), or both for 17 h at 37°C. To generate apoptotic DNA, we isolated splenocytes from WT mice and cultured those in complete RPMI for 3 days at 37°C without changing the media; cells were used to isolate genomic DNA using a Qiagen kit. The supernatant of stimulated Ly6C hi monocytes was analyzed for cytokine production.

Statistical Analysis
All data were expressed as means ± SEM. Statistical significance was evaluated using Prism 9.1 software to perform one-way ANOVA Tukey's multiple comparison test.

Streptococcus pneumoniae-Induced
Innate Immune Responses Are Largely Intact in STING −/− Mice We investigated the role of STING in host defense against pulmonary S. pneumoniae infection. Streptococcus pneumoniaeinduced lung inflammatory cytokines IL-6, IL-1b, TNFa, and chemokines KC, MCP-1 were not altered in the STING −/− mice (Figures S1A-E). Interestingly, there was no detectable S. pneumoniae-induced lung IFNb protein ( Figure S1F). Lung bacterial burden and total proteins in the BAL fluid, an indication of lung damage, were also not significantly different between STING −/ − and WT mice ( Figures S1G, H). Pneumococcal infection recruits neutrophils into the lung that are critical for the host defense (16). There was no difference in the total numbers of recruited neutrophils (CD11b hi Ly6G + ) between WT and STING −/− mice (Figures S1I, J). Thus, STING is largely dispensable for the innate immune responses to pulmonary pneumococcal infection, which is consistent with a recent report (6).

STING Is Required for S. pneumoniae Induced Lung IL-12p70 by Monocyte/ Monocyte-Derived Cells
Interestingly, S. pneumoniae-induced lung IL-12p70 was lost in STING −/− mice ( Figure 1A). Streptococcus pneumoniae secretes cyclic di-AMP that is a STING ligand. STING can also be activated via cGAS that senses cytosolic DNA (17)(18)(19). We found that cGAS −/− mice failed to make lung IL-12p70, suggesting that S. pneumoniae-induced lung IL-12p70 was likely induced by DNA, not cyclic di-AMP ( Figure 1B). As a control, S. pneumoniae induced lung TNFa was unaltered in the cGAS −/− mice ( Figure S1K).
Two Waves of Lung IFNg Production During S. pneumoniae Infection IL-12p70 drives IFNg production (21,22). Streptococcus pneumoniae infection induces strong IFNg production in the lungs (9)(10)(11)(12). We found that IL-12p70 −/− mice failed to make IFNg upon S. pneumoniae infection at 48 hpi, but not at 24 hpi ( Figure 2A). We reasoned that there were at least two waves of lung IFNg production during S. pneumoniae infection, and only the second wave of lung IFNg was dependent on IL-12p70. CCR2 −/− mice lack lung IL-12p70 production. Similar to the IL-12p70 −/− mice, CCR2 −/− mice were defective in the late, but not early S. pneumoniae-induced lung IFNg production ( Figure 2B).
LysM cre STING fl/fl Mice Lack S. pneumoniae Induced Late-Stage Lung IFNg Production Monocyte/monocyte-derived cells are critical for lung IL-12p70 and the late-stage IFNg production (Figures 1 and 2). We hypothesized that STING expression in monocyte drove lung IFNg production. We examined S. pneumonia-induced lung IFNg production in CD11c cre STING fl/fl and LysM cre STING fl/fl mice. CD11c cre STING fl/fl mice delete STING gene in alveolar macrophage and dendritic cells (DCs), while LysM cre STING fl/fl mice delete STING gene in alveolar macrophage, interstitial macrophage, monocyte/monocyte-derived cells, and neutrophils (24). Notably, neutrophils do not express STING, while NK cells and Ly6C hi monocytes have strong STING expression (13,14).
We found that LysM cre STING fl/fl , not the CD11c cre -STING fl/fl mice, were defective in the late-stage lung IFNg production by S. pneumoniae infection (Figures 3D, E), suggesting that STING expressing in interstitial macrophage, monocyte/monocytederived cells, not DCs or alveolar macrophage, was likely To further establish that STING expression in monocyte/ monocyte-derived cells is critical for lung IFNg production, we adoptively transferred (i.n) WT bone marrow Ly6C hi monocyte into STING −/− mice at 16 hpi and determined lung IFNg production at 48 hpi. We found that STING −/− mice receiving WT Ly6C hi monocyte produced lung IFNg at 48 hpi ( Figure 3F). We concluded that STING expression in monocyte/monocytederived cells promotes the late-stage lung IFNg production during S. pneumoniae infection.
Besides monocyte, neutrophils and NK cells produce lung IFNg at 48 hpi ( Figure 2F). We proposed that STING expression in monocyte/monocyte-derived cells produces IL-12p70 that drove late-stage lung IFNg production by NK cells and neutrophils during pneumococcal infection ( Figure 3G). Indeed, intranasal administration of recombinant IL-12p70 at 16 hpi restored IFNg production in STING −/− mice ( Figure 3H).
We also examined lung monocyte infiltration in STING −/− mice during S. pneumoniae infection. We observed a mild decrease in lung Ly6C hi monocytes in STING −/− mice at 48 hpi ( Figures 3I, J). Furthermore, STING −/− had similar S. pneumoniae-induced MCP-1 production as the WT mice (S1D). We, thus, preferred the hypothesis that STING expression in monocyte senses DNA and drives IL-12p70 production to promote late-stage lung IFNg production.   Figure S3A). Ly6C hi monocyte, NK cells, or moMAC, however, contained few labeled bacteria ( Figures  S3B-D), indicating that the S. pneumoniae may not directly release pathogen DNA into the cytosol of these cells. During live infection, besides live bacteria and bacteria components, there were dead host cells in the lung that could release their DNA and activate the cGAS-STING pathway. We activated monocyte with mouse genomic DNA isolated from apoptotic mouse splenocytes ( Figure 4A). We isolated Ly6C hi monocytes from the bone marrow and stimulated them with mouse genomic apoptotic DNA. Surprisingly, we found that the apoptotic DNA alone did not generate IL-12p70 or IFNg in the isolated Ly6C hi monocyte ( Figures 4B, C). We also used cGAMP to activate the Ly6C hi monocyte. Again, cGAMP did not induce IL-12p70 ( Figure 4B). As control, both DNA and cGAMP activated Ly6C hi monocytes to produce TNFa and IFNb ( Figures S4A, B). Thus, DNA sensing alone cannot activate Ly6C hi monocyte to produce IL-12p70 or IFNg.
Heat-Killed Streptococcus pneumoniae and Apoptotic DNA Together Induce IL-12p70 and IFNg in Ly6C hi Monocyte During live S. pneumoniae infection, monocyte and monocytederived cells likely encounter both PAMP, e.g., TLR agonists, and DAMP, e.g., host DNA released from dead cells. We hypothesized that the cGAS-STING pathway may synergize with the TLR pathway to induce IL-12p70 and IFNg in Ly6C hi monocyte. We activated monocyte with HKSP plus apoptotic DNA. Indeed, HKSP/DNA stimulation induced IL-12p70 and IFNg in monocyte ( Figures 4B, C). Similarly, HKSP/cGAMP induced IL-12p70 and IFNg (Figures 4B, C). HKSP alone did not induce IL-12p70 or IFNg in the monocyte (Figures 4B, C). We concluded that the induction of IL-12p70-IFNg in monocyte required the synergistic activation of DNA and TLRs signaling.
To ask if TLR2 is required for lung IFNg production, we infected TLR2 −/− mice with S. pneumoniae. Unlike the STING −/− or cGAS −/− mice, the lung IFNg production was similar in WT and TLR2 −/− mice ( Figure 4G). We suspected that additional TLRs pathways, such as TLR9 (26), might synergize with the cGAS-STING pathway for lung IFNg production during pathogen infection, compensating for the loss of TLR2.
To establish that MyD88 are required for S. pneumoniae induced lung IFNg, we adoptively transferred (i.n.) WT, STING −/− , or MyD88 −/− monocytes to CCR2 −/− mice and examined the IFNg production in the lung by S. pneumonia. Unlike the WT monocyte, neither STING −/− nor MyD88 −/− monocyte restored lung IFNg production in the CCR2 −/− mice ( Figure 4H). Thus, both MyD88 and STING expression in the monocyte are required for lung IFNg production in vivo.

DISCUSSION
In this report, we showed that STING synergizes with MyD88 to induce IL-12p70 and IFNg in the lung. STING is particularly required for the late-stage (48 hpi) lung IFNg production during S. pneumoniae infection. This unique requirement likely reflects the need for IL-12p70 production since the initial lung IFNg production by S. pneumoniae is IL-12p70 independent (23).
Previously, Temizoz et al. showed that cGAMP, in combination with CpG ODN, stimulated IFNg production in PBMCs (26). Furthermore, cGAMP and CpG ODN together, acting as an antigen-free anticancer agent, reduced tumor size significantly compared to cGAMP alone in the EG-7 and B16-F10 mouse tumor models (26). They further showed that IL-12p70 was required for the synergistic induction of IFNg in PBMCs (26). Different from ours, Temizoz et al. showed that type I IFN was needed for the IFNg production (26). Type I IFN is required for IL-18 production in moMACs (27). IL-18, also known as IFNg-inducing factor, can induce IFNg production (21,22,28). In our experimental setting, lung production of IFNg does not require type I IFN. Nevertheless, it is likely that that type I IFN, via the production of IL-18, together with IL-12p70, could further augment IFNg production.
It has long been known that DCs production of IL-12p70 requires at least two stimuli (29)(30)(31)(32)(33). This dual requirement is likely a safeguard to avoid the possible detrimental effects of uncontrolled IL-12p70-medicated Th1 responses. Napolitani et al. showed that in both human and mouse DCs, TLR3 and TLR4 potently synergized with TLR7, TLR8, and TLR9 to induce IL-12p70 and IL-23, leading to enhanced and sustained Th1 responses (29). Here, we found that STING-mediated cytosolic DNA sensing pathway synergize with TLR2 pathway in monocyte for IL-12p70 and IFNg production. Nevertheless, it is likely that STING pathway can synergize with other PRRs for IL-12p70 and IFNg production because TLR2 −/− mice did not have defect in lung IFNg production during S. pneumoniae infection. How STING and TLRs synergistically induce IL-12p70 is unclear. TLRs activation takes place on the plasma membrane or endosome, while STING activation happens at the ER-Golgi interface. It is tempting to speculate that a spatiotemporal activation of STING and TLRs may aid in IL-12p70 production.
The discovery of two waves of lung IFNg production during S. pneumoniae infection may clarify the role of IFNg in pneumococcal infection. Using IFNg −/− mice or anti-IFNg neutralizing Ab, previous studies were inconclusive (9)(10)(11)(12). We speculated that the two waves of lung IFNg may play opposite roles in host defense against pneumococcal infection. The neutrophils-mediated, IL-12p70-independent early lung IFNg may be beneficial by generating M1 macrophages to neutralize bacteria. The late-stage lung IFNg in pneumococcal infection, however, may be detrimental because in patients with S. pneumoniae sepsis, IFNg was elevated and correlated with increased mortality (7). We speculated that persistent IFNg production may promote sustained inflammation that may enhance tissue damage and mortality. Ly6C hi monocyte has emerged as a key player in pathogeninduced IFNg production in the mucosal surface (34,35). Ly6C hi monocytes are rapidly recruited to sites of infection and differentiate into macrophages and dendritic cells. Two recent studies found that CCR2 −/− mice, which lack infiltrating Ly6C hi monocyte, produced significantly less IFNg in the lung during pulmonary Legionella pneumophila infection (34,35). Similar to our finding, they found that IL-12p70 is required for IFNg production and identified infiltrating monocyte as the major source of IL-12p70 (34,35). Another study found that during vaginal HSV-2 infection, Ly6C hi monocytes produce IL-18, which activates NK cells to produce IFNg (36). Thus, a new paradigm emerges that during mucosal pathogen infection, infiltrating Ly6C hi monocyte produces IL-12p70 or IL-18 that instructs NK cells or T cells to produce IFNg.
In summary, the activation of the STING pathway in monocyte/monocyte-derived cells can synergize with the MyD88 pathway to drive IFNg production during pneumococcal infection that may influence the development of adaptive immunity.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
All experiments with mice were performed by the regulations and approval of the Institutional Animal Care and Use Committee from the University of Florida (protocol number 201909362).

AUTHOR CONTRIBUTIONS
HRT, SP, and LJ conceived the research. LJ designed the experiments, wrote the manuscript, and supervised the research. SP, HT, HG, SM, and LJ performed experiments and analyzed the data. SP drafted the manuscript. All authors contributed to the article and approved the submitted version.

FUNDING
This work was supported by NIH grants AI110606, AI125999, AI132865, and HL152163 (to LJ). SM was supported through The American Association of Immunologists Careers in Immunology Fellowship Program.