|Home | About | Journals | Submit | Contact Us | Français|
T helper 17 (Th17) cells play an important role in mucosal host defense through production of the signature cytokines IL-17 and IL-22. Prostaglandin E2 (PGE2) has been shown to enhance IL-17 production by mature Th17 cells. However, when present during Th17 differentiation, we found that PGE2 inhibited the transcription factor IRF4 and suppressed production of IL-17 but not IL-22. We show that IRF4 was required for IL-17 expression but inhibited IL-22 expression, highlighting the potential for discordant regulation of these two cytokines in Th17 cells. The pathogenic fungus, Cryptococcus neoformans, produces PGE2 and we found that it uses PGE2- and IRF4-dependent mechanisms to specifically inhibit induction of IL-17 during Th17 differentiation. Blockade of host PGE2 during infection led to increased IL-17 production from CD4+T cells and increased survival of mice. These findings suggest that host- or pathogen-derived PGE2 can act directly on Th17 cells during differentiation to inhibit IL-17-dependent anti-microbial responses.
T helper 17 (Th17) cells play an important role in mucosal host defense. Through production of the signature cytokines IL-17 and IL-22, Th17 cells trigger anti-microbial responses such as neutrophil recruitment and anti-microbial peptide secretion (Ouyang et al., 2008). IL-17 plays an important role in responses against pathogenic fungi such as Candida albicans. Mice lacking IL-17 develop more severe oral and skin candidiasis compared to wild type mice (Conti et al., 2009; Kagami et al., 2010) and humans with impaired IL-17 signaling or IL-17 autoantibodies have recurrent mucocutaneous candidiasis (Burbelo et al., 2010; Kisand et al., 2010; Puel et al., 2011; Puel et al., 2010). In addition, IL-17 expression has been correlated with protection against the fungal opportunistic pathogens Cryptococcus neoformans and Pneumocystis jiroveci (Kleinschek et al., 2006; Muller et al., 2007; Rudner et al., 2007; Zhang et al., 2009).
Recent work has shown that prostaglandin E2 (PGE2) can enhance IL-17 production from previously polarized Th17 cells (Boniface et al., 2009; Chizzolini et al., 2008; Napolitani et al., 2009; Yao et al., 2009). PGE2 can exert pro-inflammatory and anti-inflammatory effects depending on the context and target of its action. For example, PGE2 inhibits the proliferation of T cells and suppresses IFNγ production from mature Th1 cells (Betz and Fox, 1991; Harris et al., 2002; Hasler et al., 1983). In contrast, a recent study found that PGE2 enhances IFNγ production during the differentiation of Th1 cells from naive T cells (Yao et al., 2009). These studies suggest that PGE2 differentially affects T cell cytokine production depending on the developmental state of the cell.
In mammals, PGE2 synthesis involves arachidonic acid metabolism by cyclooxygenases (COX). Recognition of fungi by the innate immune system can lead to host PGE2 production (Gagliardi et al., 2010; Smeekens et al., 2010). Interestingly, several pathogenic fungi, including Cryptococcus neoformans and Candida albicans, have been reported to synthesize PGE2 from arachidonic acid by COX-independent enzymatic mechanisms (Erb-Downward and Huffnagle, 2007; Erb-Downward and Noverr, 2007). The virulence factor, laccase, has been identified as the enzyme required for PGE2 synthesis in Cryptococcus neoformans (Erb-Downward et al., 2008; Zhu and Williamson, 2004).
Although recent work has shown that PGE2 enhances IL-17 production from memory Th17 cells in humans and mice, we hypothesized that fungal- or host-derived PGE2 encountered by naïve T cells during early stages of infection may influence the fate of these cells. In contrast to its ability to enhance memory T cell IL-17 production, we found that PGE2 inhibited the production of IL-17 from naïve T cells exposed to Th17 differentiation conditions. Cryptococcus similarly inhibited IL-17 production in a PGE2-dependent manner. We further found that PGE2 inhibited DNA binding and expression of the transcription factor IRF4, resulting in specific blockade of IL-17 but not IL-22, another Th17-associated cytokine. Inhibition of PGE2 synthesis in vivo during cryptococcal infection resulted in increased T cell IL-17 production and improved survival. These results show that host and cryptococcal PGE2 production can contribute to fungal virulence by directly inhibiting the polarization of naïve T cells into IL-17-secreting effector cells.
Given that PGE2 has been shown to increase IL-17 production from previously polarized Th17 cells, we explored whether PGE2 could act on naïve T cells and influence their differentiation into Th17 cells. We differentiated naïve CD4+ T cells under Th17 polarization conditions (anti-CD3+anti-CD28+IL-6+TGF-β) for 3 days. Intracellular cytokine analysis revealed that the addition of PGE2 at the beginning of polarization strikingly reduced IL-17 production (Figure 1A top panels). In contrast, secondary stimulation of previously polarized Th17 cells for an additional 3 days in combination with PGE2, IL-23, or both, enhanced IL-17 production (Figure 1A bottom panels and (Boniface et al., 2009; Chizzolini et al., 2008; Napolitani et al., 2009; Yao et al., 2009)).
Among the four mammalian PGE2 receptors, T cells express only EP2 and EP4 (Figure S1 and (Boniface et al., 2009; Napolitani et al., 2009; Yao et al., 2009)). Consistent with this, the EP2 agonist butaprost and the EP4 agonist misoprostol inhibited expression of IL-17 transcript and protein, whereas the EP1 and EP3 agonist sulprostone had no effect (Figure 1B). PGE2 was a more potent inhibitor of IL-17 than either receptor agonist alone (Figure 1B), suggesting a combinatorial effect of EP2 and EP4 signaling. Of note, EP2 and EP4 both increase intracellular cAMP, whereas EP1 and EP3 do not (Harris et al., 2002). The ability of forskolin, a cAMP-inducing agent, to mimic the inhibitory effects of PGE2 on IL-17 expression further supported the involvement of EP2 and EP4 (Figure 1C). Taken together, these results indicate that PGE2 can act via T cell EP2 and EP4 receptors to suppress induction of IL-17 during T cell differentiation.
Titration of PGE2 revealed that the inhibitory effect on IL-17 expression is evident at concentrations of 10 nM and greater (Figure 2A). In contrast, these concentrations of PGE2 caused a slight increase in IL-22 production (Figure 2B). The inhibitory effects of PGE2 selectively on IL-17 persisted even when IL-1β and IL-23 were added during Th17 polarization (Figure 2C), a condition that greatly enhanced production of IL-17 and IL-22 in the absence of PGE2 (Chung et al., 2009; Datta et al., 2010). These findings suggest that PGE2 targets a factor that differentially regulates IL-17 and IL-22, two cytokines that are often coordinately regulated in Th17 cells.
TGF-β is known to enhance IL-17 and suppress IL-22 (Rutz et al., 2011). In light of the ability of both PGE2 and TGF-β to differentially regulate these Th17-associated cytokines, we examined whether PGE2 exerted its effects by interacting with the TGF-β pathway. PGE2 still inhibited the low amounts of IL-17 produced by T cells differentiated in the absence of TGF-β, and inhibited IL-17 over a wide range of TGF-β concentrations (Figure 2D). Furthermore, PGE2 did not influence the ability of TGF-β to suppress IL-22 (Figure 2D). These results suggest that PGE2 and TGF-β exert their effects on Th17 differentiation independently of each other.
To identify the mechanism by which PGE2 suppresses IL-17 induction, we next focused on the IL-6 signaling pathway central to Th17 differentiation. Signaling through the IL-6 receptor induces phosphorylation of STAT3, which then translocates to the nucleus and regulates many Th17-associated genes. Patients with Job’s (or hyper IgE) syndrome carry STAT3 mutations that decrease IL-17 production from T cells and predispose to fungal infections (Holland et al., 2007; Milner et al., 2008). In mice, Stat3−/− T cells are defective in production of IL-17, IL-22, Rorγt, and IRF4, consistent with the ability of STAT3 to bind directly to regulatory regions of these genes (Durant et al., 2010). PGE2 did not change the phosphorylation status of STAT3 in response to TGF-β and IL-6, making STAT3 an unlikely target of PGE2 (Figure 3A and Figure S2A and S2B).
IRF4, another transcription factor required for Th17 differentiation, is induced upon TCR activation, and further upregulated by TGF-β and IL-6 (Brustle et al., 2007; Chung et al., 2009). PGE2 suppressed this upregulation of IRF4 during Th17 differentiation (Figure 3B). IRF4 transcripts decreased as early as 24 hours after PGE2 treatment (Figure 3C). PGE2 treatment also led to the downregulation of IRF4 in memory Th17 cells (Figure S2C).
Downstream of STAT3 and IRF4, Rorγt is a transcriptional activator that controls expression of both IL-17 and IL-22 (Ivanov et al., 2006; Sanos et al., 2009; Trifari et al., 2009; Yang et al., 2007). PGE2 suppressed the induction of Rorγt by TGF-β and IL-6 (Figure 3D). Rorα, which is also important for Th17 cell development and is regulated similarly to Rorγt (Yang et al., 2008), was also suppressed by PGE2 (Figure 3E), suggesting the target of PGE2 is upstream of both Rorα and Rorγt. PGE2 also downregulated AHR (Figure 3F), a transcription factor important for IL-17 and IL-22 expression (Quintana et al., 2008; Veldhoen et al., 2008). However, downregulation of AHR is associated with decreased IL-22. Therefore, the increased IL-22 seen with PGE2 was inconsistent with an AHR-mediated effect.
Since PGE2 did not affect STAT3 phosphorylation but affected Th17-associated transcription factors downstream of STAT3, we investigated which of these effects could account for the decreased IL-17 and increased IL-22 expression seen with PGE2. A dramatic reduction in IL-17, IL-22, IRF4, and Rorγt in Stat3−/− cells confirmed the central role of STAT3 in Th17 differentiation (Figure S3A). STAT3 upregulates IL-17 expression by directly binding to the IL-17 promoter (Chen et al., 2006). Binding of STAT3 to the IL-22 promoter has not been demonstrated, but the decreased IL-22 expression in Stat3−/− cells may be explained by the requirement of STAT3 for Rorγt production, important for both IL-17 and IL-22 expression (Ivanov et al., 2006; Nurieva et al., 2007; Sanos et al., 2009; Trifari et al., 2009; Yang et al., 2007). Indeed, loss of Rorγt (Rorc−/−) resulted in decreased relative expression of both IL-17 and IL-22 (Figure 4A). IL-17 transcript and protein expression were completely Rorγt-dependent, but low amounts of Rorγt-independent IL-22 were expressed under low TGF-β conditions (Figure S3B). Taken together, these results indicate that PGE2 modulates canonical Th17-inducing transcriptional pathways. However, the coordinate regulation of IL-17 and IL-22 by AHR and Rorγt led us to hypothesize that IRF4, whose role in IL-22 regulation has been incompletely defined, may be a candidate mediator of the differential effect of PGE2 on these two cytokines.
IRF4 expression was unaffected by Rorγt deficiency (Figure 4A ), suggesting that loss of IL-17 and IL-22 production in the absence of Rorγt was not mediated by IRF4. Surprisingly, in the absence of IRF4, IL-17 expression was suppressed and IL-22 was increased (Figure 4B). siRNA gene targeting of IRF4 during Th17 differentiation confirmed the decreased IL-17 and increased IL-22 production (Figure 4C). When IL-23 and IL-1β were included during Th17 differentiation, the absence of IRF4 still led to decreased IL-17 and increased IL-22 expression (Figure 4D). The effects of IRF4 deficiency were similar in the absence of TGF-β, and increasing doses of TGF-β still inhibited IL-22 in IRF4-deficient cells (Figure S3C). These data identify that IRF4 is required for IL-17 expression but acts as an inhibitor of IL-22 expression.
Unlike STAT3 and Rorγt deficiency, IRF4 deficiency mimicked the ability of PGE2 to inhibit IL-17 expression and promote IL-22 expression. Loss of IRF4 also reproduced the effect of PGE2 on expression of other genes (Figure 5A), further suggesting that the effect of PGE2 on IL-17 and IL-22 may be mediated by its suppression of IRF4. Polarization of IRF4-deficient naïve T cells confirmed that IRF4 was absolutely required for IL-17 induction (Figure 5B). The failure of PGE2 to augment IL-22 in IRF4-deficient cells indicated that IRF4 mediated this effect. Unexpectedly, the increased IL-22 expression seen with IRF4 deficiency was suppressed by PGE2 (Figure 5B), suggesting that IRF4 was not only required for PGE2-mediated enhancement of IL-22 expression but that an additional mechanism of PGE2-mediated inhibition of IL-22 was unmasked in the absence of IRF4. Consistent with the induction of multiple pathways by PGE2, retroviral overexpression of IRF4 failed to reverse the effects of PGE2 on Th17 differentiation, indicating that IRF4 is necessary but not sufficient to mediate these effects (Figure 5C). One possible implication of these results is that beyond transcriptional repression of IRF4 itself, PGE2 also induces other factors that prevent IRF4 DNA binding or transcriptional activity. Indeed, PGE2 abolished IRF4 binding at IL-17 and IL-22 promoter regions (Figure 5D ). The inhibition of binding was detectable 18 hours after PGE2 exposure, before IRF4 protein started to decrease (Figure S4A). Since IRF4 promotes its own expression (Lehtonen et al., 2005), these data suggest that PGE2 first prevents DNA binding by IRF4, which then leads to decreased expression of IRF4 transcript and protein.
The transcriptional repressor Bcl6 antagonizes IRF4 transcriptional activity (Gupta et al., 1999) and is negatively regulated by IRF4 in B cells (Saito et al., 2007), making it a candidate mediator of IRF4 inhibition by PGE2. Overexpression of Bcl6 in T cells suppresses Th17 cell development (Nurieva et al., 2009). Consistent with negative regulation by IRF4 and an inhibitory role on Th17 differentiation, Bcl6 protein and mRNA expression increased with PGE2 treatment (Figure S4B). Loss of Bcl6 had little effect on IL-17, IL-22, IRF4, or Rorγt expression, indicating Bcl6 was not required for Th17 polarization (Figure S4C). Bcl6 deficiency did not alter the inhibitory effect of PGE2 on IL-17 but enhanced the ability of PGE2 to augment IL-22 expression (Figure S4D), suggesting that Bcl6 inhibited IL-22 under these conditions and that induction of Bcl6 may contribute to the paradoxical inhibition of IL-22 by PGE2 when IRF4 is not present to counterbalance these effects. Thus, PGE2 appears to exert its effects on multiple transcription factors whose interplay during Th17 differentiation remains to be fully understood, but these data taken together suggest that inhibition of IRF4 is a central mechanism by which PGE2 alters Th17 differentiation.
Cryptococcus neoformans, an opportunistic fungal pathogen that is a leading cause of mortality in HIV-infected individuals worldwide (Park et al., 2009), produces PGE2 using the enzyme laccase (Erb-Downward et al., 2008). Laccase-deficient Cryptococcus neoformans (Δlac1) strains are less virulent than the wild type H99 strain (Pukkila-Worley et al., 2005; Salas et al., 1996). We speculated that inability of Δlac1 to produce PGE2 (Zhu and Williamson, 2004) contributes to its decreased virulence. To test our hypothesis that fungal PGE2 production could directly alter T cell polarization, we differentiated naïve CD4+ T cells under Th17 polarization conditions (anti-CD3+anti-CD28+IL-6+TGF-β) in the presence of either wild type H99 or Δlac1 mutant Cryptococcus. We found that H99 inhibited T cell production of both IRF4 and IL-17, whereas Δlac1 did not (Figure 6A and 6B). Neither H99 nor Δlac1 suppressed IL-22 production, consistent with a specific inhibitory effect on IL-17 production (Figure 6C). IRF4 inhibition by H99 was reversed by the addition of antagonists to the PGE2 receptors EP2 or EP4 (Figure 6D). IL-17 inhibition by H99 was reversed by the EP2 antagonist, but not by the EP4 antagonist (Figure 6E). The inability of EP4 antagonism to recover IL-17 expression despite the recovery of IRF4 expression may be due to inhibition of IL-17 by enhanced FoxP3 expression under these conditions (data not shown) and suggests differential roles for EP2 and EP4 receptors in regulating IL-17 expression. The ability of PGE2 receptor antagonists to reverse IRF4 and IL-17 inhibition is consistent with cryptococcal inhibition of IL-17 being driven by PGE2 and not by other functions of laccase. These data show that cryptococcal PGE2 can directly suppress IL-6-mediated production of IL-17 by blocking the upregulation of IRF4 required for IL-17 expression (Figure S5).
Intranasal infection with Cryptococcus neoformans in BALB/c mice leads to pulmonary infection and mortality. We postulated that PGE2 produced in the lungs by either the host or fungus during infection might suppress IL-17 production from developing T cells and thus decrease immunity to the fungus. Deletion of the laccase gene renders Cryptococcus less virulent in mice; this may be due to loss of fungal PGE2 production or due to loss of other laccase functions. PGE2 can also be produced in infected lung by host cells. Since it is not feasible to specifically inhibit PGE2 synthesis by cryptococcal laccase, we targeted host PGE2 induction with the cyxclooxygenase inhibitor indomethacin. Treatment with indomethacin beginning one day after Cryptococcus infection reduced PGE2 metabolite concentrations in the lung (Figure 7A). Indomethacin modestly increased survival of infected mice by 3–4 days (Figure 7B). This increased survival required CD4+ T cells, as CD4+ T cell depletion before infection caused mice to succumb at the same rate as the control group (Figure 7C). Splenic CD4+ T cells isolated from infected mice treated with indomethacin produced more IL-17, but not IL-22, IFN-γ, or IL-4 (Figure 7D). In addition to CD4+ T cells, other cell types including γδ T cells, NK cells, and neutrophils have been reported to produce IL-17. One week after infection, indomethacin-treated mice had an increased number of IL-17+ leukocytes (CD45+) in bronchoalveolar lavage fluid (BALF) compared to control mice (Figure 7E). Most of these were CD4+ T cells and a smaller fraction were γδ T cells (Figure 7E). Less than 0.1% of IL-17-producing cells in BALF were neutrophils with no significant differences between the groups (data not shown). Significant numbers of NK cells were not found. There was an overall increase in IL-17+ cells in BALF two weeks post-infection, but the effect of indomethacin was not sustained at this time point (Figure 7F). Concordant with its effects on IL-17, indomethacin decreased lung cryptococcal burden one week after infection but cryptococcal burden was equivalent to untreated mice by two weeks (Figure 7G). This early effect of indomethacin on IL-17 expression and cryptococcal burden contributed to the survival advantage seen in treated mice since this was reversed with IL-17 blockade (Figure 7H). IL-17 blockade or CD4 depletion had little effect on infected PBS-treated control mice (Figure 7C and 7H), consistent with IL-17 already being suppressed during infection, a likely contributing factor to the recognized but incompletely understood susceptibility of this mouse model to Cryptococcus. The observed effects of indomethacin treatment indicate that PGE2 blockade can augment immunity to fungal infections by increasing early T cell IL-17 production.
Our data show that PGE2 acts during Th17 cell differentiation to suppress IL-17 while enhancing IL-22 expression. Although IL-17 and IL-22 often appear to be coordinately regulated in Th17 cells, the potential for discordant expression is evident with the identification of “Th22” cells that express IL-22 but not IL-17 (Duhen et al., 2009; Trifari et al., 2009). Compared to the regulation of IL-17, the regulation of IL-22 is poorly characterized. We show that IRF4, a transcription factor critical for IL-17 expression, suppresses IL-22 expression. By inhibiting IRF4, PGE2 simultaneously enhances IL-22 and suppresses IL-17 expression. The influence of PGE2 on other transcription factors, such as Bcl6, may contribute to this effect. This differential activity of IRF4 is consistent with its known ability to act as either a transcriptional activator or repressor at different loci depending on cofactor binding (Marecki and Fenton, 2002). A previous study reported loss of IL-22 production in IRF4-deficient T cells in vitro, but stable IL-22 production in vivo (Brustle et al., 2007). The sustained IL-22 production from T cells in vivo is consistent with our results, and the in vitro loss of IL-22 production may reflect different experimental conditions that included use of TNF-α during polarization and resting cells for two days before cytokine analysis.
IRF4 not only regulates Th17 cell differentiation but also plays a crucial role in the differentiation of other T cell lineages (Lohoff et al., 2002; Staudt et al., 2010; Zheng et al., 2009). Of note, PGE2 has little effect on Th2 cells in vitro (Betz and Fox, 1991), suggesting that the effects on IRF4 are specific to IL-6-dependent Th17 differentiation. The mechanisms responsible for this specificity remain to be determined. Because IRF4 can interact with many different factors, such as PU.1, FoxP3, and Bcl6, and since IRF4 can regulate itself, it is possible that PGE2 targets an IRF4-associated factor that regulates IRF4 specifically in the context of IL-6 effects.
PGE2 also appears to have disparate activities on T cells depending on their differentiation status. In contrast to the inhibition of IL-17 in naïve T cells undergoing Th17 polarization, the addition of PGE2 to previously polarized Th17 cells enhances IL-17 production. This difference may be due to PGE2 targeting multiple pathways that play distinct roles at these two stages. PGE2 inhibits IL-17 by suppressing IRF4 downstream of IL-6 receptor signaling, which is critical for Th17 polarization. After polarization, signals other than IL-6, such as IL-23, are important to promote IL-17 expression. PGE2 inhibited induction of IRF4 in response to IL-23 in polarized Th17 cells, but this did not inhibit IL-17 and IL-22 production. PGE2 can upregulate expression of IL-23R and IL-1R in mature T cells (Boniface et al., 2009). This augmentation of IL-23 and IL-1 pathways by PGE2 might explain the response of memory Th17 cells to PGE2 that appears to be IRF4-independent. Consistent with this divergent effect on IL-6 versus IL-23 pathways, cAMP-dependent enhancement of IL-17 production is seen during IL-6-independent Th17 differentiation with IL-1β and IL-23 (Datta et al., 2010). Understanding this differential regulation of T cell polarization under varying cytokine contexts will be important to predict the effects of therapeutic interventions such as PGE2 blockade with commonly used non-steroidal anti-inflammatory drugs (NSAIDs).
As mentioned previously, PGE2 can have different effects depending on the context and cellular target of action. Its pleiotropic effects on innate immunity, adaptive immunity, and the neuroendocrine axis make the net immune effect of systemic alterations in PGE2 difficult to predict. The ability of several fungal pathogens to secrete their own PGE2 raises the intriguing possibility that these pathogens may be able to use this pathway to manipulate specific aspects of host immunity in the local milieu during infection. We find that cryptococcal PGE2 can directly affect T cell differentiation by suppressing IRF4 and thus IL-17 production. By inhibiting Th17 differentiation locally during the early stages of infection, this may prevent the development of optimal host defense.
Several studies have shown that protection from cryptococcal infection correlates with IL-17 production. For instance, treatment of mice with IL-23 enhanced IL-17 expression and increased survival compared to IL-12 treatment (Kleinschek et al., 2010). In addition, IL-17 expression is high in Il4−/−Il13−/− mice after infection and this correlates with survival (Zhang et al., 2009). In these models, IL-17 seems to play a role in controlling early infection locally in the lung but the effects of IL-17 neutralization during natural infection are modest (Hardison et al., 2010), consistent with preexistent suppression of IL-17 by cryptococcal and host-derived PGE2. This suppressed IL-17 response likely contributes to the susceptibility of the mouse model to Cryptococcus, since our studies show that blockade of host PGE2 production during infection increases early IL-17 production by CD4+ T cells that results in improved survival. This ability of CD4+ T cell functional enhancement to increase mouse resistance suggests that the susceptibility of the mouse model is at least partially dependent on inadequate CD4+ T cell function, mimicking the inadequate CD4+ T cell functionality and numbers usually seen in humans who are susceptible to cryptococcal disease. Of note, a similar benefit of PGE2 blockade in human disease would require sufficient residual naïve CD4+ T cell numbers to mount an IL-17 response. While our studies show a clear role for CD4+ T cells, IL-17-producing γδ T cells detected in the lung may also contribute to the immune response against Cryptococcus. The role of IL-22, and the potential effect of its enhancement by PGE2, in fungal infections is less clear, with evidence that it may contribute to immunity in some contexts (Burbelo et al., 2010; Kisand et al., 2010; Puel et al., 2010; Zelante et al., 2011) but not in others (Kagami et al., 2010).
Although inhibition of host PGE2 synthesis with indomethacin lowered broncheoalveolar lavage PGE2 metabolite concentrations, it is still possible that cryptococcal PGE2 accumulates in microenvironments of the lung. Since Cryptococcus uses COX-independent mechanisms to generate PGE2, currently available pharmacologic COX inhibitors are ineffective in inhibiting cryptococcal PGE2 synthesis. Developing specific inhibitors of cryptococcal PGE2 synthesis to use alongside COX inhibitors that block host PGE2 may be a strategy to enhance the host anti-fungal response and optimize therapy against this important fungal pathogen. Our work further suggests that specific PGE2-mediated mechanisms (e.g. IRF4 inhibition) may be targeted to selectively modulate the pleiotropic effects of PGE2 on host immunity.
See Supplemental Experimental Procedures for additional details.
Single cell suspensions of spleens from 6–8 week old C57BL/6 mice were obtained using a GentleMACs tissue dissociator. After ACK lysis of red blood cells, CD4+ T cells were enriched by negative selection (Miltenyi Biotec). Naïve T cells (CD4+CD62L+CD44low) were sorted on a FACSAria. Purity ranged from 98–99%.
Anti-CD3ε (10 μg/ml) was coated overnight onto a 24-well plate. Naïve T cells were then cultured at 2×106 cell/ml/well in High Glucose DMEM with 10% FCS (Hyclone), 2 mM Glutamax, 0.1 mM non-essential amino acids, 1 mM sodium pyruvate, 55 μM beta-mercaptoethanol, 100 U/ml penicillin, and 100 μg/ml streptomycin (all from Invitrogen). Anti-CD28, anti-IL-4, and anti-IFNγ were added to the wells with or without TGF-β, IL-6, IL-1β, IL-23, or PGE2. When indicated, titrated numbers of Cryptococcus neoformans grown overnight on starvation plates were added to the wells. After three days, cells were harvested for RNA isolation or flow cytometry and supernatants were harvested for ELISA. For restimulation, cells were replated at day 3 on plates coated with anti-CD3 and soluble IL-23 or PGE2 for an additional three days. Reagent concentrations are detailed in Supplemental Experimental Procedures.
For intracellular staining, cells were activated for four hours with 50 ng/ml PMA and 1 μM ionomycin (Sigma-Aldrich). GolgiPlug (BD Biosciences) was added for the final three hours. Cells were then stained with Aqua Live/Dead stain (Invitrogen), fixed in 2% paraformaldehyde and permeabilized in Cytoperm buffer (BD Biosciences). FoxP3 staining buffers (eBioscience) were used for IRF4 and Bcl6 staining. Cells were stained with PE-conjugated anti-IL-22 (clone 1H8PWSR, eBioscience), APC-conjugated anti-IL-17A (clone eBio17B7, eBioscience), or FITC-conjugated anti-IFN-γ (clone XMG1.2, eBioscience). Data were acquired on a LSR II or LSR Fortessa flow cytometer (BD Biosciences) and analyzed with Flowjo (Treestar). For sorting, cells were enriched as described above and then stained with FITC-labeled anti-CD4 (clone RM4-5, eBioscience), PE-labeled anti-CD62L (clone MEL-14, eBiosciences), and APC-labeled anti-CD44 (clone IM7, eBioscience). IRF4 was visualized with a PE-conjugated rat anti-mouse antibody (clone 3E4, Santa Cruz Biotechnology). Anti-phospho-STAT3 (pY705) conjugated to Alexa Fluor 488 was used per the manufacturer’s instructions (BD Biosciences). PE-conjugated Bcl6 antibody was from eBioscience. For BALF analysis, antibodies against the following antigens were used: CD45 (30-F11), CD4 (GK1.5), CD3 (17A2), γδ (eBioGL3), CD49b (DX5), Gr-1 (RB6-8C5), CD11b (M1/70), and IL-17 (eBio17B7) (all from eBioscience).
Phospho-STAT3 (pY705) was detected in T cell supernatants using the Pathscan Inflammation Multi-Target Sandwich ELISA kit (Cell Signaling Technology). The IL-17 ELISA set was from eBioscience and the IL-22 ELISA set was from Antigenix. Multiplex cytokine detection in BALF was performed using the Bio-Plex suspension array system (Bio-Rad).
Real Time data was obtained using Taqman One-Step RT-PCR Master Mix reagents (Applied Biosystems). Samples were run in duplicate and normalized to the housekeeping gene, beta-actin. Data was reported according to the ΔΔCt method: ΔΔCt = ΔCtsample − ΔCtreference. Primers are listed in Supplemental Experimental Procedures.
Naïve mouse CD4+ T cells (2×106) were cultured under IL-17-inducing conditions with or without PGE2 (1.0 μM) for 0, 10, or 30 minutes. Cells were stimulated with anti-CD3+anti-CD28 T activator Dynabeads (Invitrogen), anti-IL-4 (10 μg/ml), anti-IFN-γ (10 μg/ml), IL-6 (10 ng/ml) and human TGF-β (1 ng/ml). Whole cell lysates were made in RIPA buffer with Phenylmethanesulfonyl fluoride, Phosphatase Inhibitor Cocktail 2, and Protease Inhibitor Cocktail (Sigma). For Western blotting, protein was separated by SDS-PAGE, transferred to nitrocellulose membranes using the iBlot Gel System (Invitrogen), and probed using anti-pSTAT3 (D3A7 (Tyr705); Cell Signaling Technology), anti-GAPDH (D15H11; Cell Signaling Technology), or anti-STAT3 (9132; Cell Signaling Technology). IRDye 680LT donkey anti-rabbit antibody (LI-COR) was used for detection. Membranes were imaged with an Odyssey Infrared Imaging System and Odyssey Classic Software (LI-COR).
siRNA gene targeting of IRF4 expression was achieved using ON-TARGETplus SMARTpool siRNA from Dharmacon. A non-targeting pool was used as a negative control. Naive T cells were stimulated (anti-CD3+anti-CD28+IL-6+TGF-β) for 18 hours before nucleofection of siRNAs (Amaxa). Cytokines were then replaced and cells were harvested 72 hours after the initiation of polarization. Decreased IRF4 expression was confirmed by flow cytometry.
Naive CD4+ T cells were activated overnight (18 hours) as described above with or without 1 μM PGE2. Cells were fixed, nuclei lysed, and chromatin sheared using a Misonex sonicator and cup horn. Immunoprecipitation was according to the Millipore Magna-ChIP protocol. To precipitate IRF4 complexes, 2 μg of IRF4 polyclonal goat antibody was used (M-17, Santa Cruz Biotech). Goat IgG was used as an isotype control. Percent input was determined by removing an aliquot of sheared chromatin prior to immunoprecipitation and comparing amplification of this DNA with amplification of the precipitated chromatin using Sybr Green and real-time PCR (7500Fast, Applied Biosystems). IRF4-binding sites in the IL-17 promoter regions were previously reported (Kwon et al., 2009). Primers are listed in Supplemental Experimental Procedures.
Phoenix E cells were transfected with empty or IRF4-expressing retroviral plasmids (pMSCV-IRES-GFP) using the calcium-phosphate method. After 48 hours viral supernatants were collected, filtered through a 0.45-μm filter, and added to T cells (2×106) that had been cultured for 24 hours under Th17 conditions with or without PGE2(1 μM). The culture medium was replaced with 1 ml of viral supernatant and Polybrene (10 μg/ml, Millipore), and cells were centrifuged for 1hr at 1,000 g and 30° C. Cells were then incubated for 1 hr at 37° C before replacement of viral supernatant with pre-warmed culture medium supplemented with respective cytokines at half the initial concentration. Cells were analyzed by flow cytometry 48 hours later.
Cryptococcus neoformans strain H99 was grown for two days on YPD agar plates. On the day of infection, cells were scraped off the plates, resuspended in PBS, and counted. Mice were anesthetized with isoflurane and 1×106 Cryptococcus cells in 20 μl PBS was administered via the nostril. Beginning one day after infection, mice were injected intraperitoneally with 100 μg of indomethacin or vehicle in PBS every other day. For in vitro T cell studies, spleens were isolated after 13 days. Colony forming units (CFU) were obtained by homogenizing whole lungs at one and two weeks post-infection. For depletion of CD4+ cells, 1 mg of CD4 antibody (GK1.5, BioXCell) or isotype control (LTF-2, BioXCell) was injected intraperitoneally one day prior to infection. Flow cytometry one week later confirmed 99.9% depletion of CD4+ T cells in the blood. For blockade of IL-17, mice were injected with 100 μg of IL-17 antibody (17F3, BioXCell) or isotype control (MOPC-21, BioXCell) every other day starting on the day of infection. BALF was obtained by lavage of the lungs with 1 ml cold PBS.
PGE2 metabolite concentration in BALF was determined in duplicate samples using Prostaglandin E Metabolite EIA Kit (Cayman).
Statistics were calculated using the unpaired Student’s t test (two-tailed) with Prism software (GraphPad). *P < 0.05, **P < .001, and ***P < 0.001.
We thank R. Munford (NIAID) for discussion and comments on the manuscript, and members of the J. O’Shea (NIAMS) and W. Leonard (NHLBI) laboratories for assistance in identifying ChIP binding regions. This work was supported by the Intramural Research Program of the NIH, NIAID.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.