Phagocytes produce prostaglandin E2 in response to cytosolic Listeria monocytogenes

Listeria monocytogenes is an intracellular bacterium that elicits robust CD8+ T-cell responses. Despite the ongoing development of L. monocytogenes-based platforms as cancer vaccines, our understanding of how L. monocytogenes drives robust CD8+ T-cell responses remains incomplete. One overarching hypothesis is that activation of cytosolic innate pathways is critical for immunity, as strains of L. monocytogenes that are unable to access the cytosol fail to elicit robust CD8+ T-cell responses and in fact inhibit optimal T-cell priming. Counterintuitively, however, activation of known cytosolic pathways, such as the inflammasome and type I IFN, lead to impaired immunity. Conversely, production of prostaglandin E2 (PGE2) downstream of cyclooxygenase-2 (COX-2) is essential for optimal L. monocytogenes T-cell priming. Here, we demonstrate that vacuole-constrained L. monocytogenes elicit reduced PGE2 production compared to wild-type strains in macrophages and dendritic cells ex vivo. In vivo, infection with wild-type L. monocytogenes leads to 10-fold increases in PGE2 production early during infection whereas vacuole-constrained strains fail to induce PGE2 over mock-immunized controls. Mice deficient in COX-2 specifically in Lyz2+ or CD11c+ cells produce less PGE2, suggesting these cell subsets contribute to PGE2 levels in vivo, while depletion of phagocytes with clodronate abolishes PGE2 production completely. Taken together, this work demonstrates that optimal PGE2 production by phagocytes depends on L. monocytogenes access to the cytosol, suggesting that one reason cytosolic access is required to prime CD8+ T-cell responses may be to facilitate production of PGE2.


Introduction
Listeria monocytogenes is a Gram-positive, intracellular pathogen that elicits robust T-cell responses. Though both CD4 + and CD8 + T-cells contribute to development of Listeriainduced immunity, experiments involving adoptive transfers or in vivo depletion highlight the importance of CD8 + T-cells specifically in mediating protective responses [1,2]. As such, L. monocytogenes has been used for decades as a model to understand how CD8 + T-cell responses are primed [3]. Understanding these responses has become more pressing recently as Listeriabased platforms aiming to drive CD8 + T-cell responses are in clinical trials as cancer immunotherapies [4]. Initial work showed that critical signals promoting Listeria-stimulated T-cell responses are provided acutely, as bacterial clearance with antibiotics as early as 24 hours-post infection has minimal impact on the kinetics of CD8 + T-cell responses [5]. This work highlights the role of early signals in informing Listeria-stimulated cell mediated adaptive responses.
One early signal impacting T-cell responses is the inflammatory environment induced during infection. The importance of the inflammatory milieu on priming T-cell responses has been solidified by multiple groups using antigen-pulsed, matured dendritic cells in combination with non-antigen expressing L. monocytogenes as an inflammatory boost [6,7]. These studies enable discrimination between antigen presentation and inflammation and demonstrate that wild-type L. monocytogenes provides an optimal inflammatory milieu to drive T-cell priming [6,7]. The inflammatory boost provided through wild-type L. monocytogenes infection led to increased T-cell responses, whereas use of strains that specifically alter the inflammatory milieu led to suboptimal responses [6,7]. L. monocytogenes activates a number of innate pathways that contribute to the inflammatory milieu. In particular, multiple groups have focused on the role of various cytosolic innate immune pathways, as previous research demonstrated the necessity of cytosolic access in priming cell-mediated immunity [8][9][10]. L. monocytogenes utilizes a cytolysin, listeriolysin O (LLO), to escape from phagosomes directly into the cytosol and LLO-deficient mutants that are unable to access the cytosol inhibit T-cell priming and generate tolerizing immune responses [11,12]. Despite the importance of cytosolic access for priming T-cell responses, multiple cytosol-dependent innate pathways are counterintuitively detrimental to immunity, including STING-dependent type I interferon [13,14] as well as

Unprimed macrophages and dendritic cells upregulate PGE 2 -synthesizing enzymes in response to cytosolic L. monocytogenes
Infection of BMDMs and BMDCs with wild-type L. monocytogenes led to expression of the genes necessary for PGE 2 production (Fig 1A-1C). To assess whether these cells could utilize these enzymes to produce PGE 2 , we assessed PGE 2 production in culture supernatant by mass spectrometry. Surprisingly, supernatant from both BMDMs and BMDCs had no detectable PGE 2 compared to PBS-treated controls, both during infection with wild-type or Δhly L. monocytogenes (Fig 1D). This suggests that either enzyme expression was not high enough to induce detectable PGE 2 , or there may be additional post transcriptional modifications required for enzyme activity. Analysis of PGE 2 from BMDMs or BMDCs deficient in COX-2 had no detectable PGE 2 , as expected ( Fig 1D).

Primed BMDMs and BMDMs produce PGE 2 during cytosolic L. monocytogenes infection
The lack of PGE 2 produced by BMDMs and BMDCs in response to L. monocytogenes infection was surprising given the upregulation of Ptgs2 and Ptges transcripts. Other innate pathways, such as the inflammasome, require a priming step in order to induce optimal activation. We hypothesized that naïve bone marrow-derived phagocytes may similarly require additional stimulation in order to produce PGE 2 . To test this hypothesis, we treated BMDMs and BMDCs overnight with the TLR2 agonist PAM3CSK4 (PAM) before infection with wild-type and Δhly L. monocytogenes and again analyzed transcript expression. PAM alone induced a small amount of expression of Ptgs2 expression in both BMDMs and BMDCs (Fig 1A). Infection with wild-type L. monocytogenes led to a significant increase in expression that was less robust in Δhly L. monocytogenes-infected cells, similar to the effect seen in unprimed cells ( Fig 1A). Ptges expression, alternatively, had a larger increase in transcript expression during PAM-priming, both during wild-type and Δhly L. monocytogenes infection of BMDMs and BMDCs ( Fig 1B). Furthermore, PAM treatment alone induced expression of Ptges similar to that induced during infection in BMDMs (Fig 1B). Taken together, these data suggest that cytosolic access accentuates expression of Ptgs2, where TLR signaling alone is sufficient to induce Ptges expression. We also assessed expression of Pla2g4a in TLR-primed BMDMs and BMDCs and, similar to unprimed cells, saw no changes in expression (S1A Fig). Additionally, Ifnb1 transcript was again induced during cytosolic infection, where Il1b expression was induced during PAM-treatment alone, as well as during infection with wild-type or Δhly L. monocytogenes (S1B and S1C Fig).
As Ptgs2 expression was also dependent on cytosolic access in primed BMDMs and BMDCs, we next assessed expression of COX-2 protein in primed cells infected with wild-type or Δhly L. monocytogenes by western blot. Similar to unprimed cells, BMDMs had similar levels of COX-2 protein during infection with wild-type or Δhly L. monocytogenes (Fig 1C). In BMDCs, alternatively, COX-2 protein expression was reduced during infection with Δhly L. monocytogenes compared to wild-type infection (Fig 1C), again suggesting that COX-2 protein expression in BMDCs is potentiated by cytosolic access.
We hypothesized that priming BMDMs and BMDCs with PAM would stimulate the cells to produce PGE 2 during infection with wild-type L. monocytogenes. To test this hypothesis, we Wild-type or COX-2 -/-BMDMs or BMDCs were infected with indicated strains of L. monocytogenes at an MOI of 10 +/-the TLR2 agonist PAM3CSK4 and assessed 6hpi for expression of Ptgs2 (encoding COX-2) and Ptges (encoding mPGES-1) by qRT PCR (A-B) or COX-2 protein by western blot (C). Supernatant was harvested and assessed for PGE 2 by mass spectrometry (D). Mass spectrometry data was normalized to d-PGE 2 and fold change is relative to PBS treated controls. Data are a combination of at least two independent experiments (A,B, D, and western blot quantification), or a representative of at least two independent experiments (western blot image). Significance was determined by a one-way ANOVA with Bonferroni's correction. � p < 0.05, �� p < 0.01, ��� p < 0.001, ���� p < 0.0001. https://doi.org/10.1371/journal.ppat.1009493.g001

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes assessed production of PGE 2 in the supernatant of primed BMDMs and BMDCs by mass spectrometry. BMDMs and BMDCs were treated overnight with PAM before infection with wildtype and Δhly L. monocytogenes. Six hours post-infection, cell supernatant was assessed for PGE 2 . In contrast to unprimed BMDM and BMDCs, wild-type infection of primed cells led to a significant increase in PGE 2 production compared to PBS-treated controls ( Fig 1D). Previous data showed L. monocytogenes-stimulated PGE 2 production in peritoneal macrophages [23,25]. Our data suggest that priming BMDMs prior to infection induces the cells to behave more like tissue resident macrophages in respect to PGE 2 production. Furthermore, the ability of BMDMs to produce PGE 2 provides a tool to efficiently study PGE 2 synthesis in macrophages during infection. PAM-primed COX-2 deficient BMDMs and BMDCs again led to no PGE 2 production, solidifying the necessity of COX-2 activity in PGE 2 production ( Fig 1D). To characterize the kinetics of PGE 2 production, we analyzed PGE 2 production via ELISA from two to eight hours post-infection in BMDMs. During wild-type infection, PGE 2 was produced as early as two hours post-infection (S1D Importantly, maximal PGE 2 production in primed BMDM and BMDCs was dependent on cytosolic access, as infection with Δhly L. monocytogenes led to significantly reduced PGE 2 levels ( Fig 1D). No PGE 2 was detected during Δhly L. monocytogenes infection at any earlier or later timepoints assessed, suggesting that vacuole-constrained infection does not simply alter PGE 2 production kinetics (S1D Fig). Furthermore, when BMDMs were infected with substantially more Δhly L. monocytogenes (MOI 50 compared to MOI 10 wild-type L. monocytogenes) there was still significantly less PGE 2 production at six hours-post infection (S1E Fig). Taken together, these results suggest that cytosolic L. monocytogenes induces robust PGE 2 production in TLR-primed BMDMs and BMDCs.
Given our data that cytosolic access potentiated PGE 2 production in BMDMs and BMDCs, we next assessed whether known cytosolic pathways influenced production of PGE 2 . Two innate cytosolic pathways activated by L. monocytogenes, inflammasomes and type I IFN, can influence PGE 2 production in other infection models [25][26][27]. Accordingly, we assessed PGE 2 production ex vivo during infection with strains of L. monocytogenes that alter activation of these pathways. Infection of BMDMs with ΔtetR L. monocytogenes (a strain that hyperactivates type I IFN) [28] or ΔmdrMTAC (a strain that induces less type I IFN) [29] did not alter PGE 2 production (S1F Fig). In contrast, infection with a strain that hyperactivates the inflammasome, Lm-pyro, led to a small but reproducible decrease in PGE 2 production (S1F Fig), suggesting that inflammasome activation may influence PGE 2 production in response to L. monocytogenes infection.
Lastly, we also sought to understand whether PGE 2 specifically was being induced, or if there was a more broad increase eicosanoid production. To test the hypothesis that L. monocytogenes induces production of other eicosanoids, we analyzed production of prostaglandin D 2 (PGD 2 ), thromboxane B 2 (TXB 2 ), and leukotriene B 4 (LTB 4 ). However, we saw no changes production of these eicosanoids by wild-type L. monocytogenes (S2A and S2B Fig). This interesting observation suggests that macrophages and dendritic cells preferentially induce PGE 2 in response to infection with cytosolic L. monocytogenes.

Cytosolic access is required for PGE 2 production in vivo
Production of PGE 2 by TLR-primed BMDMs and BMDCs ex vivo is potentiated by cytosolic access. To assess whether L. monocytogenes induces PGE 2 in a cytosol-dependent manner in vivo, we infected mice intravenously with wild-type and Δhly L. monocytogenes and assessed PGE 2 levels in the spleen twelve hours post-infection, previously defined as the peak PGE 2 response to infection [16]. Wild-type L. monocytogenes led to an eight-fold increase in PGE 2 (Fig 2A). Infection with Δhly L. monocytogenes strikingly showed no increase in PGE 2 over mock-immunized controls (Fig 2A). To ensure that the reduced PGE 2 production was not due to differences in bacterial burdens, mice were infected at a dose of wild-type (10 5 bacteria) and Δhly L. monocytogenes (10 7 bacteria) that led to comparable burdens ( Fig 2B). This shows that the absence of PGE 2 in Δhly L. monocytogenes-infected mice is not just due to reduced bacterial burdens. Taken together, these data highlight that cytosolic access is necessary for in vivo induction of PGE 2 .
Previous data demonstrated that infection with Δhly L. monocytogenes fails to induce robust cell-mediated immunity [11,12]. Furthermore, coinfection of Δhly L. monocytogenes with wild-type L. monocytogenes suppresses T-cell priming associated with wild-type infection, demonstrating that infection with vacuole-constrained L. monocytogenes induces an immunosuppressive environment [11]. We hypothesized that infection with Δhly L. monocytogenes may actively inhibit PGE 2 production, contributing to impaired T-cell responses. To address this hypothesis, we infected mice with wild-type L. monocytogenes, Δhly L. monocytogenes, or a coinfection of both strains and assessed PGE 2 levels twelve hours post-infection. Wild-type L. monocytogenes infection induced PGE 2 whereas Δhly L. monocytogenes failed to induce production, as expected ( Fig 2C). Strikingly, coinfection of the two strains led to little to no PGE 2 production ( Fig 2C) suggesting that vacuole-constrained strains of L. monocytogenes actively impair production of PGE 2 . Active inhibition of PGE 2 production, combined with the previously described impacts of TLR-dependent IL-10 [11], could be one of multiple reasons why Δhly L. monocytogenes fails to induce protective immunity.

CD11c + and Lyz2 + cells produce PGE 2 during L. monocytogenes infection in vivo
Our data identified PGE 2 production by macrophages and dendritic cells ex vivo ( Fig 1D). Furthermore, previous groups have reported that macrophage and dendritic cell subsets are heavily infected early during in vivo infection, a timepoint where we have previously detected increases in splenic PGE 2 [16,22]. From these data, we next hypothesized that macrophages and/or dendritic cells were responsible for producing PGE 2 in vivo that is necessary for optimal T-cell priming. Accordingly, we hypothesized that deletion of PGE 2 production in these cells specifically would 1) reduce the splenic levels of PGE 2 and 2) impair the robust CD8 + Tcell responses normally induced by L. monocytogenes. To test these hypotheses, we generated mice deficient in COX-2 selectively in CD11c + cells or Lyz2 + cells using the cre/lox system. Mice containing loxP sites flanking the COX-2-encoding gene (COX-2 fl/fl ) were crossed with mice expressing the cre recombinase under the CD11c (CD11c-cre + ) or Lyz2 (Lyz2-cre + ) promoters. COX-2 fl/fl CD11c-cre + and COX-2 fl/fl Lyz2-cre + mice were immunized with 10 7 CFU of a live-attenuated, vaccine strain of L. monocytogenes (LADD L. monocytogenes) currently used in clinical trials as a cancer therapy platform [30]. The LADD strain is deficient in two major virulence genes, actA and inlB, but retains immunogenicity, making it safe for clinical use [30,31]. The LADD vaccine strain was used here to enable analysis of T-cell responses in floxed mice as discussed below and induces similar levels of PGE 2 [16].
We first hypothesized that immunization of COX-2 fl/fl CD11c-cre + and/or COX-2 fl/fl Lyz2-cre + mice with LADD L. monocytogenes would lead to lower levels of splenic PGE 2 compared to control mice. Immunization of COX-2 fl/fl CD11c-cre + and COX-2 fl/fl Lyz2-cre + mice each showed reduced levels of PGE 2 production, leading to only 60% of the PGE 2 induced during immunization of control mice ( Fig 3A). However, deletion of COX-2 in either CD11c + or

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes Cytosolic access is necessary for L. monocytogenes-stimulated PGE 2 production in vivo. C57BL/6 mice were infected with 10 5 wild-type or 10 7 Δhly L. monocytogenes. 12hpi spleens were harvested for eicosanoid extraction and mass spectrometry (A) and livers were harvested for bacterial burdens (B). C57BL/6 mice were infected with 10 5 wild-type, 10 7 Δhly L. monocytogenes, or a combination of both strains. 12hpi spleens were harvested for eicosanoid extraction and mass spectrometry (C). Data are representative of two independent experiments. Mass spectrometry data was normalized to d-PGE 2 levels and fold change is compared to PBS controls. Significance was determined by a one-way ANOVA with Bonferroni's correction. �� p < 0.01, ���� p < 0.0001. https://doi.org/10.1371/journal.ppat.1009493.g002

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes Lyz2 + cells did not fully abrogate production of PGE 2 as observed in mice globally deficient in mPGES-1 (mPGES-1 -/-) ( Fig 3A). This suggests that CD11c + and Lyz2 + cells each contribute to PGE 2 production and that deletion of COX-2 in either is not sufficient to completely prevent PGE 2 production. We also assessed PGE 2 levels in mice deficient in COX-2 selectively in T-cells and PGE 2 production was still observed (S3 Fig). As T-cells are not known to be

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes infected by L. monocytogenes, this is consistent with our hypothesis suggesting PGE 2 production specifically from infected cell subsets. PGE 2 is critical for generating optimal T-cell responses in response to L. monocytogenes, as immunization of mPGES-1-deficient mice or treatment of mice with a COX-2-specific pharmacological inhibitor leads to impaired CD8 + T-cell responses [16]. We next hypothesized that the decreased PGE 2 production in the COX-2 fl/fl CD11c-cre + or COX-2 fl/fl Lyz2-cre + mice would be sufficient to similarly impair CD8 + T-cell responses. We specifically assessed the impact on CD8 + T-cells as these are the predominate cells contributing to protective immunity to L. monocytogenes [1,2] and are cells of interest in L. monocytogenes-based cancer immunotherapeutics [4]. To test this hypothesis, we immunized mice with 10 7 LADD L. monocytogenes expressing the model antigens B8R and OVA. Seven days after immunization, splenocytes were isolated, stimulated with B8R or OVA, and production of IFNγ was assessed by flow cytometry. Despite decreased PGE 2 production in these mice, T-cell responses were not affected both in percent IFNγ + as well as number of IFNγ + T-cells per spleen (Figs 3B, 3C, S4A and S4B). Similarly, the number of antigen-specific T-cells measured by B8R tetramer was unchanged in these mice compared to control mice (S4C Fig). This suggests that the PGE 2 remaining in these mice was sufficient to prime productive T-cell responses. Due to its short in vivo half-life, PGE 2 asserts its effects locally [32]. It is possible that while global splenic PGE 2 levels are decreased, the local concentrations of PGE 2 are sufficient to prime T-cell responses. Taken together, these data suggest that although Lyz2 + and CD11c + cells contribute to production of PGE 2 during L. monocytogenes infection, PGE 2 production by either cell subset alone these cells is not necessary for T-cell priming, as CD8 + T-cell responses are not impacted by loss of PGE 2 production in either subset.

Deletion of COX-2 in both Lyz2 + and CD11c + cells further reduces splenic PGE 2 levels
Our data showed that single deletions of COX-2 in CD11c + or Lyz2 + cells reduced PGE 2 , but not to baseline values. We next hypothesized that PGE 2 production by either of these subsets individually was sufficient for T-cell priming and that to observe impaired T-cell responses we would have to eliminate PGE 2 production in both CD11c + and Lyz2 + cells. To do this, we crossed the COX-2 fl/fl CD11c-cre + and COX-2 fl/fl Lyz2-cre + mice, leading to mice with a COX-2 deletion in both cell subsets (COX-2 fl/fl CD11c-cre + Lyz2-cre + ). We assessed the ability of these mice to produce PGE 2 by mass spectrometry and found that PGE 2 was further reduced, with about 40% the amount PGE 2 produced compared to immunized control mice ( Fig 4A). This suggests that CD11c + and Lyz2 + cells combined to produce the majority of PGE 2 during immunization with L. monocytogenes.
Due to further reduced PGE 2 production in our mice deficient in COX-2 in both CD11c + and Lyz2 + cells, we next assessed CD8 + T-cell responses in these mice. Mice again were immunized with 10 7 vaccine strain of L. monocytogenes expressing the model antigens B8R and OVA and assessed for IFNγ production seven days later. Despite diminished PGE 2 production, CD8 + T-cell responses remained intact in the double COX-2 deficient mice, both in percent and number (Figs 4B, 4C, S5A and S5B). Similarly, antigen-specific T-cells measured by B8R tetramer were also unchanged compared to wild-type controls (S5C Fig). This suggest that even the small amount of PGE 2 produced locally is sufficient to drive CD8 + T-cell responses.

Depletion of phagocytes eliminates PGE 2 production in vivo
Our ex vivo data highlighted the ability of BMDMs and BMDCs to produce PGE 2 in response to cytosolic L. monocytogenes. However, deletion of COX-2 in Lyz2 + and CD11c + cells did not

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes completely abrogate PGE 2 production in vivo. These data led us to hypothesize that other phagocytic cell subsets not effectively targeted by these cre-drivers may be producing the residual PGE 2 , such as marginal zone macrophages (MZMs), metallophilic macrophages, or other CD11b + cells more broadly [33,34]. To test this hypothesis, we utilized short-term clodronate liposomes to rapidly deplete phagocyte populations in the spleen. Mice were depleted with clodronate liposomes 24 hours prior to immunization with L. monocytogenes [35]. Twelve hours post-immunization, spleens were harvested and assessed for PGE 2 by mass spectrometry. Additionally, splenocytes were assessed for CD11b + and CD11c + populations by flow cytometry to confirm clodronate efficacy. Clodronate treatment led to significantly fewer CD11b + cells and a trend for decreased CD11c + cells (S6A and S6B Fig). Treatment of mice with clodronate prior to infection with L. monocytogenes completely eliminated PGE 2 production compared to infected control mice ( Fig 5A). Importantly, bacterial burdens were equivalent between clodronate and mock-treated mice (Fig 5B). Pretreatment with a control empty liposome, encapsome, actually increased PGE 2 production compared to infected control mice, potentially due to increased bacterial burdens (Fig 5A and 5B). To ensure that the abrogated PGE 2 was due to phagocyte depletion and not due to overall increased levels of cell death associated with clodronate-mediated depletion, additional mice were treated with a B cell-depleting anti-CD20 antibody 24 hours prior to infection with L. monocytogenes. Treatment with anti-CD20 resulted in depletion of~50% of splenic B-cells, yet did not alter PGE 2 production compared to mice treated with an isotype control (S6C and S6D Fig). These data suggest that the loss of PGE 2 production in the context of clodronate treatment was specific to depletion of phagocytes and that generic immune cell depletion and associated cell death does not prevent PGE 2 production. Taken together, these data demonstrate that phagocytic cell populations are critical for PGE 2 production in vivo following L. monocytogenes immunization.
Loss of antigen presenting cells through clodronate treatment leads to impaired CD8 + Tcell activation due to loss of antigen presenting cells, making analysis of CD8 + T-cell responses in this model not informative [36,37]. Given this, we alternatively assessed the possibility that other phagocytic cells targeted by clodronate, but not the COX-2 fl/fl Lyz2-cre + , could contribute to PGE 2 production. Complete elimination of PGE 2 production with clodronate treatment suggested that the residual PGE 2 in the COX-2 fl/fl CD11c-cre + Lyz2-cre + mice was due to a phagocytic cell that was not effectively targeted in these mice. Previous data showed that although the Lyz2-cre used in this study is highly efficient at deletion of loxP flanked genes in some macrophage subsets, it is only minimally successful at deleting genes of interest in other subsets, such as MZMs [33]. MZMs, characterized by expression of MARCO, are heavily infected early in L. monocytogenes infection [22]. We hypothesized that the residual PGE 2 we detected in the COX-2 fl/fl CD11c-cre + Lyz2-cre + mice may be due to inefficient deletion in macrophage subsets such as these. To assess the role of MZMs in PGE 2 production, we assessed expression of COX-2 by immunohistochemistry. Mice were immunized with 10 7 vaccine strain of L. monocytogenes and spleens were harvested three and ten hours later. Spleen cryosections were then stained for L. monocytogenes, COX-2, and MARCO. Uninfected mice had COX-2 staining in the periarteriolar lymphoid sheath (PALS) with little expression in the marginal zone (MZ) (Fig 5C and 5D). As early as three hours post-immunization COX-2 staining was observed in the MZ, with approximately 50% of COX-2 colocalizing with MARCO + cells (Fig 5C and 5D). Expression of COX-2 in the MZ was maintained at 10hpi, again showing approximately 50% colocalization with MARCO (Fig 5C and 5D). Furthermore, L. monocytogenes colocalized with COX-2 and MARCO expressing cells, suggesting that infected MZMs may be producing PGE 2 (Fig 5D). Expression of COX-2 suggests that MZMs, or other non-CD11c/Lyz2 expressing phagocytes within the marginal zone, could be capable of producing PGE 2 in vivo and may be contributing to the PGE 2 remaining in the COX-2 fl/fl CD11c-cre + Lyz2-cre + mice. To test the hypothesis that MZMs could contribute to the residual PGE 2 observed in the double transgenic mice, we assessed expression of COX-2 by immunohistochemistry in the COX-2 fl/fl CD11c-cre + Lyz2-cre + mice. The mice were infected with 10 7 LADD L. monocytogenes or mock-infected with PBS and spleens were harvested 10 hours later. COX-2 expression in the PALS was completely eliminated, confirming the efficacy of the Lyz-cre and CD11-cre in deletion of COX-2 (S6E Fig). Importantly, COX-2 expression was still detected in the MZ and colocalized with MARCO, supporting the hypothesis that MZMs may contribute to the residual PGE 2 (S6E Fig). Taken together, our data suggest that multiple myeloid derived subsets can contribute to PGE 2 production, including Lyz2 + cells, CD11c + cells, and possibly MZMs. Complete reductions in PGE 2 by depletion of phagocytic cells such as these with clodronate treatment is consistent with our data showing that PGE 2 is produced from phagocytic cells infected with cytosolic L. monocytogenes.

Discussion
Cytosolic access is required to effectively generate cell-mediated immunity to L. monocytogenes [10][11][12]. Decades of work has focused on understanding the cytosol-dependent processes necessary for T-cell priming, a topic that has gained interest recently due to use of L. monocytogenes as a cancer immunotherapy platform. Our data suggest that one reason cytosolic access is important may be to facilitate phagocyte production of PGE 2 , an eicosanoid required to generate optimal CD8 + T-cell responses [16]. We showed that PGE 2 is produced by BMDMs and BMDCs ex vivo. Importantly, this pathway is potentiated by cytosolic access, as vacuole-constrained L. monocytogenes induce lower production of PGE 2 . Furthermore, infection of mice with a vacuole-constrained L. monocytogenes strain led to no increase of PGE 2 over mock immunized controls. Lastly, we showed that Lyz2 + and CD11c + cells contribute to PGE 2 production in vivo as deletion of COX-2 in these subsets led to decreased PGE 2 levels, however other clodronate-sensitive phagocyte populations also contribute to PGE 2 production following L. monocytogenes immunization. This work leads to many new questions including how cytosolic L. monocytogenes activates this pathway, how immune cells discriminate which eicosanoid to produce in response to infection, how even small concentrations of PGE 2 still lead to productive PGE 2 responses, and how PGE 2 facilitates optimal T-cell priming.
One intriguing hypothesis is that PGE 2 synthesis during L. monocytogenes infection is driven by an innate cytosolic sensor. L. monocytogenes elicits a number of innate pathways that could contribute to differential activation of the PGE 2 -synthesis pathway. One possibility is that induction of type I IFN influences PGE 2 production. Type I IFN can be induced cytosolically by L. monocytogenes through recognition of cyclic diadenosine monophosphate (c-di-AMP). Upon entry into the cytosol, L. monocytogenes secretes c-di-AMP through multidrug resistance transporters [28,38] where it is recognized by either the reductase controlling NF-κB (RECON) [39] or stimulator of IFN genes (STING) [40,41]. STING activation leads to type I interferon induction [40,41], and was originally hypothesized to be critical for T-cell

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes responses. Paradoxically, however, type I IFN inhibits cell-mediated immunity to L. monocytogenes [14]. Interestingly, there has been well documented crosstalk between the PGE 2 and type I IFN pathways during infections with other pathogens such as influenza and M. tuberculosis [26,27]. In the context of influenza, Coulombe et al. showed that infection led to upregulation of PGE 2 and a subsequent decrease in type I IFN [26]. In contrast to L. monocytogenes, type I IFN is important in generating cell-mediated immune responses to influenza. Accordingly, diminished type I IFN due to increased PGE 2 reduces both acute and protective immunity during influenza infection. On the other hand, Mayer-Barber et al. recently showed that inhibition of type I IFN during M. tuberculosis infection led to an increased level of PGE 2 in an IL-1-dependent manner [27]. This correlated with better bacterial control. Due to crosstalk between these two pathways, it seems possible that recognition of c-di-AMP and subsequently upregulation of type I IFN may also be playing a role in PGE 2 production during L. monocytogenes immunization. Here, we show a preliminary analysis of the role of type I IFN on PGE 2 production during L. monocytogenes infection. In an ex vivo model, strains of L. monocytogenes that alter type I IFN levels did not change PGE 2 production, suggesting that perhaps type I IFN does not regulate PGE 2 . Despite these ex vivo findings, future in vivo analysis of PGE 2 levels in mice deficient in STING or the type I IFN receptor (IFNAR) is necessary to conclusively define links between these two pathways. Should type I IFN negatively influence PGE 2 production in vivo, use of L. monocytogenes strains that have reduced secretion of c-di-AMP and subsequently less type I IFN could be an avenue of further research for immunotherapeutic platforms.
Another cytosolic pathway that may influence PGE 2 levels during L. monocytogenes infection is the inflammasome. Inflammasomes are multiprotein complexes that recognize a wide range of pathogen associated molecular patterns [42][43][44]. Wild-type L. monocytogenes infection leads to a small amount of inflammasome activation, largely through the absent in melanoma 2 (AIM2) inflammasome [45]. The AIM2 inflammasome recognizes cytosolic DNA that is released during bacteriolysis within the cytosol [45][46][47]. Originally, inflammasomes such as AIM2 were known to have two major downstream effects, the release of proinflammatory cytokines IL1β/IL-18 and the induction of a lytic form of cell death, pyroptosis, characterized by formation of membrane pores by the protein Gasdermin D [48][49][50]. Seminal work by von Moltke et al. introduced a new downstream effect, the activation of an eicosanoid storm, including PGE 2 [25]. This work, as well as supporting recent work, showed elevated levels of PGE 2 after inflammasome activation [25,51]. One possible hypothesis stemming from this work is that induction of membrane pores during pyroptosis leads to calcium influx, activating cPLA2 and releasing arachidonic acid from the membrane. This model would suggest that use of mice deficient in caspase-1 or Gasdermin D would lead to lower levels of PGE 2 production. Intriguingly, our ex vivo data suggest the opposite during L. monocytogenes infection. During infection with a strain of L. monocytogenes that hyperactivates the inflammasome we found a slight, but reproducibly significant decrease in PGE 2 production. We hypothesize that this is likely due to increased pyroptosis in host cells, leading to a lower number of viable cells able to produce PGE 2 . Alternatively, the diminished PGE 2 could be due to crosstalk between these two pathways. This finding is consistent with other data showing that hyperactivation of the inflammasome leads to impaired cell-mediated immune responses [6,52], leading to the possibility that one reason for impaired responses in the context of robust inflammasome activation may be due to lower PGE 2 . Further analysis of PGE 2 levels in vivo during hyper-inflammasome activation and/or in mice lacking inflammasome components will be important to elucidate these details. The role of inflammasomes as well as type I IFN are intriguing avenues to understand signaling pathways driving PGE 2 production during L. monocytogenes infection.

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes It is also possible that production of PGE 2 by L. monocytogenes is independent of known cytosolic pathways. Identification of other unknown censors could be accomplished by assessing PGE 2 levels in response to different L. monocytogenes mutants. Mutant strains of L. monocytogenes that differentially induce PGE 2 could provide insight as to the cytosolic censors involved. One additional hypothesis is that PGE 2 -production is independent of a cytosolic sensor completely and instead is driven by LLO-mediated pore formation. Though LLO is tightly regulated transcriptionally, translationally, and posttranslationally to be most active in the vacuole, a small amount of LLO may remain active in the cytosol of cells [53][54][55]. It is possible that this small amount of LLO induces pore formation in the cell membrane and allows calcium influx, subsequently activating cPLA2. Use of strains that further restrict LLO production in the cytosol, such as new strains that excise hly once L. monocytogenes has entered the cytosol [56], could help assess the role of LLO-mediated pores on PGE 2 production.
In addition, our data show an interesting phenotype where BMDCs and PAM-primed BMDMs selectively produce PGE 2 in response to L. monocytogenes infection rather than a global increase in eicosanoid production. Analysis of the eicosanoid milieu in cell culture supernatant show an increase in PGE 2 production during wild-type L. monocytogenes infection, but little to no changes in other eicosanoids such as PGD 2 , TXB 2 , or LTB 4 . This raises the question of how a cell determines which eicosanoid to produce in response to different stimuli. The eicosanoid produced in different conditions is dependent on terminal synthases [21]. Therefore, the expression and activity of these synthases determine the resulting eicosanoid milieu. Multiple factors impact expression of different synthases including cytokines, hormones, and microbial products [57]. For example, expression of mPGES-1 can be induced by LPS and prostaglandin D 2 synthases, though less well understood, can be upregulated by glucocorticoids [58][59][60]. Conversely, anti-inflammatory cytokines such as IL-10 can inhibit expression of mPGES-1 [61]. Given that vacuole constrained L. monocytogenes have previously been demonstrated to induce high levels of IL-10 leading to inhibition of T-cell priming, it is interesting to speculate that the inhibition of PGE 2 production observed in our co-infection experiments could be tied to this previously observed induction of IL-10. Use of other L. monocytogenes strains that differentially activate cytokines or are deficient in different microbial PAMPs could be informative as to which signal specifically leads to enhanced levels of mPGES-1 transcript. In addition, activity of each synthase also may dictate which eicosanoids are produced [57]. Terminal synthase activity can be modulated by posttranslational modification (such as phosphorylation) as well as presence of cofactors (such as ATP and glutathione) [57]. Depletion of essential cofactors during metabolic or oxidative stress could influence the induced inflammatory milieu [57]. The post transcriptional regulation highlights the necessity of assessing endpoint eicosanoid production rather than simply transcript or protein levels, as these other factors influencing activity can alter which eicosanoids ultimately are produced. This is particularly true in our data, as despite seeing upregulation of mPGES-1 transcript during infection of unprimed BMDMs and BMDCs, we failed to see PGE 2 production. This suggests that perhaps some additional modification is necessary to induce mPGES-1 activity during L. monocytogenes infection. Similarly, our data suggest that cytosolic access is not required for COX-2 translation in primed BMDMs, but is required to fully induce PGE 2 production. This again suggests that further modification of COX-2 may be important for optimal activity. In contrast, cytosolic access is required for both COX-2 protein translation and PGE 2 production in BMDCs. Perhaps PGE 2 production in BMDCs is driven by COX-2 protein expression, whereas in BMDMs it is driven by post translational modifications. Further analysis on the role of phosphorylation and cofactor availability will help elucidate these details of regulation.

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes Another pressing question generated from this work is how productive T-cell responses were induced in mice deficient in COX-2 in both CD11c + and Lyz2 + cells despite reduced PGE 2 levels. Here, we show that these mice produce substantially reduced PGE 2 , yet still induce wild-type CD8 + T-cell responses. One hypothesis that we explored in this work is that other cell subsets not efficiently targeted by our cre/lox model were still producing PGE 2 . Certain cell subsets such as MZMs do not have effective gene deletion using the Lyz2 + promoter to drive cre recombinase expression [33]. Our immunohistochemistry data suggest that MZMs still may be capable of producing PGE 2 in the COX-2 fl/fl CD11c-cre + Lyz2-cre + mice (S6D Fig). For this reason, we hypothesized that subsets such as these may still be producing sufficient levels of PGE 2 to drive T-cell responses. One way to assess the role of MARCO + MZMs is use of a new cre recombinase-driving promoter, SIGN-R1, developed by Pirgova et al [34]. SIGN-R1 is a lectin binding receptor expressed on MZMs and drives more efficient deletion of genes by the cre/lox system [34]. Generation of triple COX-2 knockout mice that express the SIGN-R1-cre in combination with our reported COX-2 fl/fl Lyz2-cre + CD11c-cre + model could be informative about the role of MZMs in production of PGE 2 . Though we show that Lyz2 + and CD11c + cells contribute to PGE 2 , analysis of MZMs and other myeloid cells will further understanding of PGE 2 production.
The lack of diminished cell-mediated immunity could also be due to local acting effects of PGE 2 . It is possible that even if PGE 2 levels are below detection at a whole spleen level, certain cells are able to produce PGE 2 locally in sufficient concentration to drive T-cell responses. More sensitive measures of PGE 2 , such as quantitative mass spectrometry imaging recently developed, would be required to analyze local responses such as these [62]. These novel techniques enable analysis of location of PGE 2 and other eicosanoids within a spleen and could detect lower concentrations [62]. Similarly, the sensitivity of the receptor PGE 2 is acting upon during L. monocytogenes infection could influence how much PGE 2 is necessary for inducing a response. PGE 2 binds primarily to four receptors, EP1-4 [63]. EP3 and EP4 are higher affinity receptors (kD~1nM compared to 10-15nM for EP1/2) [63,64]. Should the higher affinity receptors be identified as the important receptors for influencing immunity during L. monocytogenes infection, even lower concentrations of PGE 2 still induced in our COX-2 fl/fl Lyz2-cre + CD11c-cre + model may be sufficient for cell-mediated responses. Further analysis as to relevant receptors and which cells they are expressed on could help elucidate these details.
Lastly, how PGE 2 facilitates T-cell responses in the context of L. monocytogenes immunization remains unknown. In innate immune cells, PGE 2 influences expression of co-stimulatory and activation markers. PGE 2 signaling in dendritic cells upregulates the co-stimulatory molecules OX40L and 4-1BBL [65], thereby promoting T-cell proliferation. Similarly, PGE 2 signaling in macrophages leads to polarization towards a more inflammatory M1 phenotype [66] and aids in activation [67]. Furthermore, PGE 2 promotes migration of innate cell subsets, leading to enhanced migration towards CCL21 [68,69] and MCP-1 [70,71]. These proinflammatory functions suggest that PGE 2 may be acting to enhance immunity through its local effects on innate immune cells. PGE 2 may also be influencing immunity more directly on T-cell subsets, such as through polarization of T-cells towards a Th1 phenotype [72]. Additionally, PGE 2 leads to higher expression of OX-40L, OX-40, and CD70 directly on T-cells, promoting T-cell interactions and sustaining immune responses [65]. In order to more fully understand how PGE 2 facilitates T-cell responses to L. monocytogenes, a comprehensive analysis of these effects on both T-cells and innate immune cells is required.
We and others have shown that innate immune responses substantially influence cell-mediated immune responses, particularly the inflammatory milieu induced during infection. Here, we present evidence that one pathway critical for immunity, induction of PGE 2 , is dependent on access to the cytosol. Furthermore, we show that PGE 2 is produced by macrophages and

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes dendritic cells. These data suggest analysis and modulation of eicosanoid levels, particularly PGE 2 levels, may be informative to improve the use of L. monocytogenes-based immunotherapeutic platforms.

Ethics statement
This work was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All protocols were reviewed and approved by the University of Wisconsin-Madison Institutional Animal Care and Use Committee.

Bacterial strains
The Listeria monocytogenes strains used in this study were all in the 10403s background. The attenuated (LADD) strain used in the analysis of T-cell responses was in the ΔactAΔinlB background as previously described and engineered to express full length OVA and the B8R 20-27 epitope [30]. OVA and B8R 20-27 were constructed as a fusion protein under the control of the actA promoter with the secretion signal of the amino terminal 300bp of the ActA gene [15]. This fusion protein was integrated into the site-specific pPL2e vector as previously described [15].

Mouse strains
Six-to eight-week-old C57BL/6 male and female mice were obtained from the NCI and Charles River NCI facility. Ptgs2 -/-(COX-2 -/-) mice were obtained from Jackson Laboratory and maintained as heterozygote breeding pairs. Ptges -/-(mPGES1 -/-) mice lacking microsomal PGE synthase have been previously described [73][74][75]. In order to generate cell-type specific COX-2 knockout mice, COX-2 fl/fl mice (stock number 030785) were obtained from Jackson Laboratory and crossed with Lyz2-cre (stock number 004781), CD11c-cre (stock number 008068), or CD4-cre expressing mice (stock number 022071), all also obtained from Jackson Laboratory. Double Lyz2-cre and CD11c-cre expressing mice were generated by crossing COX-2 fl/fl Lyz2-cre + mice with COX-2 fl/fl CD11c-cre + mice. Genotypes were confirmed by PCR using the primer pairs in Table 1. In experiments assessing PGE 2 production and T-cell responses in these mice, COX-2 flf/l cremice were used as controls for initial experiments, followed by duplicate experiments using either COX-2 flf/l creor WT C57BL/6 mice. Experiments utilizing COX-2 fl/fl cremice or C57BL/6 mice induced consistent PGE 2 levels and T-cell responses, suggesting that insertion of the loxP site independently did not influence COX-2 expression.

Mouse genotype
Forward (

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes

BMDM and BMDC generation and infection
Bone marrow-derived macrophages and dendritic cells were made using six-to eight-weekold Ptgs2 -/-(COX-2 -/-) or C57BL/6 mice as previously described [15,76]. Briefly, bone marrow was harvested and macrophages were cultured in the presence of M-CSF from transfected 3T3 cell supernatant for six days with a supplement of M-CSF at day three and frozen down for storage. Dendritic cells were cultured in the presence of 20ng/ml recombinant GM-CSF (BD Biosciences, San Jose, CA) for 7 days with a supplement of 20ng/mL GM-CSF every third day. For infection, BMDMs or BMDCs were plated at 1x10 6 cells/well in a 12 well dish overnight +/-100ng/mL PAM3CSK4. The following morning, cells were infected with indicated strains of L. monocytogenes or PBS control at an MOI of 10 unless otherwise indicated. Thirty minutes later, supernatant was removed and replaced with medium containing 50μg/mL gentamycin to remove extracellular bacteria. At the indicated times, cells were harvested for western blot or qRT PCR and supernatant was harvested for eicosanoid analysis as described below.

qRT PCR
RNA was isolated from BMDMs or BMDCs using the RNAqueous-Micro Total RNA Isolation Kit (Invitrogen), and DNAse treated with Turbo DNAse (Invitrogen) according to manufacturer's instructions. 500ng total RNA was reverse transcribed in 10μL reactions using the iScript cDNA Synthesis Kit (BioRad) according to manufacturer's instructions and cDNA was diluted 10-fold using molecular grade water (Invitrogen). 2.5μL diluted cDNA was used as template in a 10μL qRT-PCR reaction performed in duplicate using gene-specific primers and Kapa SYBR Green Universal qPCR mix (KAPA Biosystems) according to manufacturer's instructions using a BioRad CFX Connect Real-Time PCR System. The sequences of gene-specific primers are shown in Table 2. Data was analyzed using Excel and all RNA abundances were calculated by using a standard curve of synthesized template (Integrated DNA Technologies, G-Blocks) and are normalized to ActB (β-actin).

Western blots
BMDMs or BMDCs were harvested and protein was extracted using the Pierce SDS-PAGE Sample Prep Kit (Thermo) according to the manufacturer's instructions. Total protein content was measured by the Pierce BCA Protein Assay Kit (Thermo) and equivalent protein levels were loaded into a polyacrylamide gel (BioRad). Samples were transferred onto a nitrocellulose membrane using a semi-dry transfer apparatus before blocking with a 5% skim milk solution for thirty minutes at room temperature. After washing 3x with PBS-T, the membrane was incubated overnight at 4˚C with the primary antibodies anti-COX-2 (1:200, Cayman Chemical) and anti-β-actin loading control (1:1000, ThermoFisher) in a 5% bovine serum albumin solution. The following day samples were washed with PBS-T before being incubated with secondary antibodies (anti-rabbit 800 at 1:10,000, anti-mouse 680 at 1:5,000). Samples were

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes imaged on a LiCor imager and analyzed via ImageStudio. Sample signal was normalized to βactin and relative abundance was compared to wild-type L. monocytogenes.

In vivo immunizations and pharmacological treatments
L. monocytogenes of the wild-type, attenuated (LADD), or Δhly background were grown overnight in brain heart infusion media at 30C. The bacteria were back diluted 1:5 and allowed to grow to log phase (OD0.4-0.6,~1-1.5 hours) at 37C, with aeration, prior to infection. Bacteria were diluted in PBS and mice were infected with 200μL at the indicated doses intravenously. For bacterial burden analysis, mice were sacrificed at 12hpi and livers were homogenized in 0.1% Nonidet P-40 in PBS and plated on Luria-Bertani plates. For splenic macrophage depletion, 200μL clodronate, PBS control, or endosome lipid control (Encapsula Nano Sciences) were given intravenously 24 hours prior to bacterial infection according to the manufacturer's instructions. For B-cell depletion, mice were dosed intraperitoneally with 12.5μg of Ultra-LEAF Purified anti-mouse CD20 (clone SA271G2) or isotype control (clone RTK4530) in 100μL PBS 24 hours prior to bacterial infection. Depletion efficacy of relevant cell subsets was confirmed by assessing abundance of splenic CD11b + cells (clone M1/70), CD11c + cells (clone N418), or B220 + cells (clone RA3-6B2) by flow cytometry. Celecoxib (Cayman Chemical) was milled into standard mouse chow (Envigo) at 100mg/kg and fed ad lib for 48 hours before and after immunization [16,77].

Eicosanoid measurement
In vivo eicosanoid levels were assessed by mass spectrometry. Ex vivo analysis was assessed by mass spectrometry or ELISA, as indicated. For ex vivo PGE 2 analysis by ELISA, BMDMs were plated and infected as described above. Supernatant was harvested at indicated timepoints and diluted 1:100 in ELISA buffer (Cayman Chemical). PGE 2 was detected using a PGE 2 ELISA kit (Cayman Chemical) following the manufacturer's instructions. For mass spectrometry eicosanoid extractions, spleens from mice were harvested at twelve hours post immunization and flash frozen in tubes containing 50ng deuterated PGE 2 standard (Cayman Chemical) in 5μL methanol and stored overnight at -80C. For ex vivo extractions, 1mL of supernatant was flash frozen in tubes similarly containing 50ng deuterated PGE 2 standard in 5μL methanol. The following day, two mL of ice cold methanol were added to the tissue culture supernatant or spleens. Spleens were homogenized in glass homogenizers. Samples then were incubated at 4C for 30 minutes. Next, cellular debris was removed by centrifugation and samples were concentrated to 1mL volume before being acidified with pH 3.5 water and loaded onto conditioned solid phase C18 cartridges. Samples were washed with hexanes before eluting using methyl formate followed by methanol. Samples were concentrated using a steady stream of nitrogen gas and suspended into 55:45:0.  [78,79]. Samples were normalized to deuterated PGE 2 levels to ensure that differences between samples was not due to extraction efficiency. Average PGE 2 ion count for each sample was then compared to PBS or WT infected samples, as indicated.

T-cell analysis
Mice were sacrificed seven days after immunization and splenocytes were isolated as previously described [6]. In brief, red blood cells were lysed using ACK buffer and then splenocytes were counted using a Z1 Coulter counter. For tetramer analysis, splenocytes were immediately

Cryosection preparation and immunofluorescence microscopy of infected spleens
C57BL/6 or COX-2 fl/fl CD11c-cre + Lyz2-cre + mice were infected intravenously by tail vein injection of 10 7 LADD L. monocytogenes in 150 μl of PBS. Mice were sacrificed at 3 or 10 hpi and spleens harvested and snap frozen in OCT for immunofluorescence microscopy as described previously [22]. Uninfected mice were used as negative controls. Briefly, 5μm spleen cryosections were cut using a Leica CM1850 cryostat, mounted on Superfrost Plus microscope slides (Thermo Fisher) and stored at -80˚C until use.  Anti-CD20 treatment depletes splenic B220 + cells, but does not influence PGE 2 production. C57BL/6 mice were dosed with 200μL clodronate, liposome control (encapsome), or PBS 24 hours prior to immunization with 10 7 LADD L. monocytogenes. 12hpi spleens were harvested and assessed for CD11b + (A) and CD11c + (B) populations by flow cytometry. C57BL/6 mice were dosed with 50μg anti-CD20 or isotype control 24 hours prior to immunization with 10 7 LADD L. monocytogenes. 12hpi spleens were harvested and assessed

PLOS PATHOGENS
Phagocytes produce prostaglandin E 2 in response to cytosolic Listeria monocytogenes