Syk‐dependent glycolytic reprogramming in dendritic cells regulates IL‐1β production to β‐glucan ligands in a TLR‐independent manner

Abstract Dendritic cells (DCs) activated via TLR ligation experience metabolic reprogramming, in which the cells are heavily dependent on glucose and glycolysis for the synthesis of molecular building blocks essential for maturation, cytokine production, and the ability to stimulate T cells. Although the TLR‐driven metabolic reprogramming events are well documented, fungal‐mediated metabolic regulation via C‐type lectin receptors such as Dectin‐1 and Dectin‐2 is not clearly understood. Here, we show that activation of DCs with fungal‐associated β‐glucan ligands induces acute glycolytic reprogramming that supports the production of IL‐1β and its secretion subsequent to NOD‐, LRR‐ and pyrin domain‐containing protein 3 (NLRP3) inflammasome activation. This acute glycolytic induction in response to β‐glucan ligands requires spleen tyrosine kinase signaling in a TLR‐independent manner, suggesting now that different classes of innate immune receptors functionally induce conserved metabolic responses to support immune cell activation. These studies provide new insight into the complexities of metabolic regulation of DCs immune effector function regarding cellular activation associated with protection against fungal microbes.

myeloid cells, such as macrophages and DCs, and on a small subset of lymphocytes. 5 in macrophages and DCs. 5,9,10 Although innate immune responses to whole yeast cell-wall preparations are known to be mediated by both Dectin molecules and TLR-2 receptors, 11,12 studies have shown that modified yeast-derived molecules such as whole glucan particles (WGP) or depleted zymosan (ZD), which is deprived of TLR stimulatory capacity by treatment of zymosan (Zy) with hot alkali, produce Dectinspecific inflammatory responses, such as reactive oxygen species (ROS) and cytokine production in a TLR-independent manner. 13,14 Furthermore, Dectin-specific responses are nonredundant with TLRmediated innate immune responses, as Dectin-1-deficient mice display impaired immune responses against C. albicans, resulting in compromised resistance to fungal infection. 10 Stimulation of PRRs by PAMPs, such as microbial pathogens, and/or danger-associated molecular patterns (DAMPs), such as ATP and uric acid crystals, results in release of proinflammatory cytokines by DCs and macrophages. Among these cytokines, IL-1 plays a crucial role in both local and systemic inflammation as one of the earliest inflammatory mediators released by activated innate immune cells.
The secretion of biologically active IL-1 is mediated by the cleavage of pro-IL-1 by active caspase-1, and is tightly controlled by 2 distinct cellular signals. PRR activation, which induces the transcription and translation of immunologically inactive pro-IL-1 serves as a first signal, which is followed by the second signal that causes the cleavage of pro-IL-1 by active caspase-1. Different forms of a heterogeneous group of multiprotein complexes, termed inflammasomes, are responsible for activating caspase-1. These inflammasome complexes consist of sensor proteins, such as nucleotide-binding oligomerization domain like receptor family proteins (NLRs) that can detect PAMPs and/or DAMPs, an adaptor protein apoptosis-related speck-like protein (ASC), a pyrin or caspase recruitment domain, and effector caspase-1. Inflammasome activation gives rise to autocatalysis and activation of caspase-1, which then exerts a catalytic activity on pro IL-1 . [15][16][17] Among different types of NLR inflammasomes, activation of the NLRP3 inflammasome is most abundantly studied in the field of myeloid immune cells in response to various microbial pathogens, including fungi, and plays a documented downstream role in Dectin-1initiated innate immune responses. [18][19][20] Recent studies have reported that NLRP3 inflammasomes are required for immune responses against C. albicans mediated by both Dectin-1 and TLR2; however, IL-1 secretion is completely ablated in Dectin-1 deficient mice compared to that in TLR2 knockout mice. 19,21,22 Although the immune effector responses in these studies have been reported to require both TLR2 and Dectin receptors to fight fungal infections, Dectin-1-mediated effects on downstream NLRP3 inflammasome activation are much more pronounced. 21 Activation through TLRs causes DCs to exhibit rapid changes in cellular metabolism characterized most prominently by increased rates of aerobic glycolysis. TLR-driven glycolytic reprogramming critically supports the survival and immune function of DCs by satisfying both the energetic and nutrient substrate requirements associated with the rapid anabolic synthesis of proteins and effector molecules during DC activation. [23][24][25][26] Pharmacologic or genetic approaches to inhibit the induction of aerobic glycolysis during DC activation result in attenuation of DC maturation and lead to subsequent adverse effects on DC effector functions, including inflammatory cytokine secretion, antigen presentation, and T cell stimulation. [25][26][27] A prominent recent study has demonstrated that monocytes challenged with a -glucan cell wall component of C. albicans up-regulate aerobic glycolysis, albeit the underlying mechanism for this process is epigenetically regulated after a long-term fungal challenge. 28 Although activation of the NLRP3 inflammasome in macrophages is supported by aerobic glycolysis, 18 and inhibition of glycolysis attenuates IL-1 release and inflammasome activation in macrophages, 18,29 little is known about the how Dectinspecific acute glycolytic induction contributes to inflammasome activation in DCs. Here, we report that DC stimulation with Dectin-specific ligands drives aerobic glycolysis to support inflammasome-dependent IL-1 release, in a TLR-independent manner.
Upon activation of Dectin-1, the downstream adaptor motif of its cytoplasmic tail, which is similar to immunoreceptor tyrosine-based activation motifs (ITAMs), serves as a docking site for spleen tyrosine kinase (Syk). 30 Syk signaling plays an important role in lymphocyte proliferation and survival. In macrophages and DCs, phosphorylation of Syk induces inflammatory responses such as cytokine secretion, 31 ROS production, and NLRP3 inflammasome activation. 30,32 The PI3K/Akt axis regulates the proliferation, survival, and metabolic activities in many cells. [33][34][35] We have shown the activation of PI3K and Akt signal transduction during the TLR-driven glycolytic burst in DCs. 24 However, whether the PI3K/Akt axis is associated with Syk activity in Dectin-stimulated DCs is not understood. Additionally, TLR signaling has been shown to depend on MyD88, 36,37 which is an essential signaling adaptor molecule in most TLR signaling pathways.
Although some reports indicate the involvement of MyD88 in the immune responses to fungal pathogens, 31,38,39 the role of MyD88 in Dectin receptor activation is controversial. 38,40 In this report, we demonstrate that NLRP3-mediated IL-1 secretion from cells primed by a Dectin-specific ligand is facilitated by a Syk-dependent and PI3K/Akt-dependent glycolytic reprogramming that is independent of TLR-MyD88.
Although the importance of glycolytic reprogramming for DC effector function has been extensively studied in TLR-driven activation, DC metabolic regulation in response to fungal pathogens is understudied. In this study, we show that the Dectin-mediated glycolytic burst is regulated by Syk activation in a TLR-independent manner. Syk inhibition abolished the Dectin-driven glycolytic burst and downstream IL-1 activation, while sparing MyD88-dependent TLR stimulation. Collectively, our study suggests that DCs engage in a Syk-dependent signaling mechanism responsible for driving the Dectin-specific glycolytic burst in support of IL-1 production and release. These findings show that CLRs, in addition to TLRs, are sufficient for driving early glycolytic reprogramming in DCs, and that this activity is important for sustaining the early inflammatory responses to fungal-associated ligands.

Mouse DC culture and activation
Bone marrow-derived DCs (BMDCs) were generated as described in

Western blot analysis
DCs were lysed using lysis buffer with Pierce protease and phos- Cleaved caspase-1 and cleaved IL-1 blots were performed on these concentrated supernatant preparations.

Flow cytometry and cytokine measurements
Above mentioned antibodies were used for flow cytometry.

Quantitative real-time PCR of Il-1 expression
RNA was isolated with an RNAeasy Kit (Qiagen, Germantown, MD) and cDNA was synthesized with an iScript cDNA Synthesis Kit (Biorad, Hercules, CA). Il-1 Taqman primer probes (Applied Bioscience System, Foster City, CA) and AB7500 sequence detection system or QuantStudio 3.0 were used for relative mRNA expression. mRNA relative quantitative values were calculated based on 2(-ΔΔCT) and were normalized to untreated DCs with -actin used as a housekeeping gene control.

Statistical analysis
Data were analyzed with GraphPad Prism software (version 6.0). Samples were analyzed using paired t-test, 1-way and 2-way ANOVA.
ANOVA tests were post-calculated by Tukey's multiple comparison test. Results are means +SD or +SEM, and statistical values are represented with an asterisk as significant when P values were below 0.05.

Dectin-mediated activation drives glycolytic reprogramming
In order to identify Dectin-specific acute metabolic reprogramming in DCs and to determine the contribution of these metabolic changes to early immune responses by these cells, we utilized an array of agonistic ligands specific to TLRs alone (LPS), Dectin-1/2 alone (ZD, Curdlan, WGP), or ligands that interact with both simultaneously (Zy).
We first characterized the ability of these different agonists to induce acute glycolytic reprogramming, termed "glycolytic burst," in GM-CSFdifferentiated BMDCs by metabolic extracellular flux analysis (Agilent,  (Fig. 1C). Furthermore, we observed the involvement of glycolytic metabolism to support activation-associated TNF-and IL-12 production in response to each ligand (Fig. 1D). Consistent with our metabolic data ( Fig. 1A and B), proper DC activation by all ligands, including Dectin-specific ligand, ZD, requires glycolytic reprogramming (Fig. 1D).

Dectin-dependent glycolytic burst and maturation requires Syk signaling
Dectin-driven immune responses are documented to be mediated by both Syk-dependent and -independent mechanisms. 5,42 Based on this, we were interested in testing the requirement for Syk signaling in only a partial inhibition was observed for the TLR-specific agonist LPS and dual agonist Zy (Fig. 2A). The partial effect of Syk inhibition on LPS-mediated glycolytic reprogramming is consistent with reports that Syk constitutively binds to the cytoplasmic tail of TLR4 and is known to play a role in TLR4-mediated signaling. 43 Up-regulated surface expression of the co-stimulatory molecules CD40 and CD86 in response to Dectin-specific ligands (ZD, Curdlan, WGP) was strongly attenuated by Syk inhibition, whereas TLR-driven maturation (LPS) was unimpaired, and dual-agonist maturation (Zy) was only modestly impaired, by Syk inhibition (Fig. 2B). These data indicate that Sykdependent signals are sufficient to regulate Dectin-mediated glycolytic reprogramming and maturation in DCs.

Dectin-mediated metabolic reprogramming is independent of TLR/MyD88 signals
In order to confirm that Dectin-specific metabolic reprogramming and activation was independent of TLR-signaling, we stimulated  3C). These data confirm that Dectin/Syk mediated signals are sufficient to drive DC metabolic reprogramming even in the absence of the TLR2/MyD88 signaling axis that is also well documented to regulate innate immune responses to fungal-associated ligands.

Dectin-mediated glycolytic reprogramming requires a PI3K/TBK-1/Akt signaling axis
Syk signaling is known to play a crucial role in Dectin-mediated innate immune responses. 22,30,31,44 In previously published work, we have identified that phosphorylation of Akt T308 is required for hexokinase association with the mitochondria and glycolytic reprogramming in TLR-stimulated DCs. 24 We thus hypothesized that similar signaling pathways may regulate acute metabolic responses to Dectin-specific activation of DCs. To test this, we first examined expression of total and T308-phosphorylated Akt in BMDCs stimulated with Zy or ZD (Fig. 4A). We observed that DC activation by both ligands induces the phosphorylation of Akt T308, suggesting that TLR and CLR signaling mechanisms likely converge to mediate glycolytic reprogramming (Fig. 4A). PI3K inhibition resulted in reduced phosphorylation of Akt T308 in response to ZD, suggesting that PI3K is involved in regulating Dectin-dependent signaling that leads to Akt T308 phosphorylation (Fig. 4B). This is in contrast to previously published data showing that PI3K is not required for acute glycolytic reprogramming in response to TLR stimulation in DCs. 24 In previous studies, we have shown that early TLR-mediated glycolytic reprogramming in DCs is mediated by signaling through TBK1. 24 Consistent with these findings, inhibition of TBK1 completely ablated the phosphorylation of Akt T308 in ZD activation (Fig. 4C), suggesting that TLR and Dectin-dependent glycolysis induction pathways converge at this point. Consistent with these data, Akt inhibition attenuated acute glycolytic reprogramming in both Zy-and ZD-stimulated DCs (Fig. 4D, left), whereas PI3K inhibition significantly impacted glycolysis induction only in ZD-stimulated cells (Fig. 4D, right). In TLR-stimulated DCs, TBK1 phosphorylation of Akt has been demonstrated to regulate acute glycolytic burst independent of PI3K signaling, whereas the PI3K/Akt axis plays a major role in maintaining long-term glycolytic activity in these cells. 24,26 In contrast to these findings for acute glycolytic induction to TLR-mediated stimulation, our data suggest that Dectin-dependent Akt T308 activation is mediated by both PI3K and TBK-1. with -glucan components of C. albicans up-regulate aerobic glycolysis, these studies emphasized the long-term effect of -glucan-treated monocytes and the metabolic changes described in these cells have been documented to be regulated by epigenetic changes. 28,45 However, the requirement for Dectin-specific acute metabolic regulation for NLRP3 mediated IL-1 secretion has not been explicitly characterized in DCs. We hypothesized that the aerobic glycolysis observed in Dectin-activated BMDCs (Figs. 1 and 2) is required to support IL-1 production and secretion triggered by the NLRP3 inflammasome in a Syk-dependent manner. To test this, we activated BMDCs with LPS, Zy, and ZD in the presence or absence of 2DG for 6 h and examined the transcriptional expression of IL-1 (Fig. 5A). mRNA expression of IL-1 did not change regardless of glycolysis inhibition, indicating that IL-1 production is not regulated by glycolysis at the transcriptional level, a finding consistent with the lack of a glycolytic requirement for transcription of inflammatory genes in response to TLR stimulation of DCs. 24 However, using an ELISA for secreted IL-1 , we found that IL-1 production and secretion was significantly reduced in 2DG-inhibited DCs at 5 h post-activation (Fig. 5B). This ELISA assay did not discriminate between pro-and cleaved IL-1 , and so does not directly assess NLRP3 inflammasome activation; rather, it depicts an impact on IL-1 protein translation that is consistent with the attenuation of cytokine protein production seen in TLR-activated 2DG-inhibited DCs. 24 To more directly address the impact of glycolysis on NLRP3-dependent activity, the levels of both cleaved caspase-1 and cleaved IL-1 were detected in both cellular lysates and culture supernatants (Fig. 5C). Cleaved caspase-1 and cleaved IL-1 were both readily detected in lysates and supernatants of cells stimulated with LPS, Zy, or ZD in the presence of the inflammasome complex activator nigericin, with ZD-stimulated cells showing the lowest levels of these cleaved molecules (Fig. 5C). Cleaved caspase-1 and cleaved IL-1 levels in BMDCs stimulated with the TLR/CLR agonists in the presence of nigericin were significantly attenuated by 2DG-mediated glycolysis inhibition (Fig. 5D). It is worth noting that in our hands, cleaved IL-1 levels were consistently more sensitive to 2DG inhibition than cleaved caspase-1, and that the effect of 2DG on cleaved caspase-1 was fairly variable ranging from near-complete inhibition  (Fig. 5E). In addition, cleaved IL-1 production and secretion was attenuated for all ligands by 2DG inhibition at both "0" and "4" h time points, with more dramatic impact in the 2DG "0" group and a more modest impact for the 2DG "4" group (Fig. 5E). These data suggest that Dectin-mediated and Syk-dependent IL-1 activity is regulated at both the translational level and inflammasome-dependent secretion level by glucose metabolism (Fig. 5D and E).

DISCUSSION
Stimulation of myeloid immune cells with fungal ligands induces metabolic reprogramming. 28,46 Several studies have also shown that activation (of immune cells) by fungal pathogens causes inflammasome formation 17,32,44,47 and induces the Syk-kinase signaling pathway. 22,30,31,44 Although these studies add significant value to the understanding of host immune responses against fungal pathogens, an integrated understanding of fungal-mediated acute metabolic changes in the regulation of specific immune outcomes in DCs, such as inflammasome-dependent cytokine responses, has not been previously described. In this study, we demonstrate that the Dectinspecific activation of DCs induces glycolysis-dependent IL-1 production via the Syk-mediated signal transduction pathway, which drives activation of the PI3K/TBK1/Akt axis. showed that Dectin-mediated Syk-dependent IL-1 production in BMDCs is heavily reliant on glucose and glycolytic reprogramming.
Glucose metabolism in TLR-mediated activation has been widely studied in DCs and innate myeloid immune cells alike, 49 whereas LPS induces aerobic glycolysis but not OXPHOS. 51 Differential functional outcomes could also arise from cellular participation in divergent glucose metabolism. In the example of monocytes stimulated with TLR2 ligands, phagocytic activity of these cells is preferentially supported by OXPHOS whereas cytokine production is facilitated by glycolysis and OXPHOS. 51 In parallel to TLR-mediated metabolic changes, recent work has highlighted the differential metabolic outcomes from dimorphic stages of C. albicans, leading to diverse functional responses. 46 Of note, since we have recently shown that TLRmediated DC activation relies on alternative metabolic pathways such as glycogen metabolism, 52 it is tempting to hypothesize that Dectin-1 mediated responses may participate in similar aspects of activationassociated glucose metabolism.
In summary, our findings identify Dectin-specific acute glycolytic reprogramming events that support DC effector responses to fungalassociated -glucan ligands. However, the complex and redundant nature of the metabolic networks demands additional work at an organism level to better understand the regulatory mechanisms