TPL-2 Regulates Macrophage Lipid Metabolism and M2 Differentiation to Control TH2-Mediated Immunopathology

Persistent TH2 cytokine responses following chronic helminth infections can often lead to the development of tissue pathology and fibrotic scarring. Despite a good understanding of the cellular mechanisms involved in fibrogenesis, there are very few therapeutic options available, highlighting a significant medical need and gap in our understanding of the molecular mechanisms of TH2-mediated immunopathology. In this study, we found that the Map3 kinase, TPL-2 (Map3k8; Cot) regulated TH2-mediated intestinal, hepatic and pulmonary immunopathology following Schistosoma mansoni infection or S. mansoni egg injection. Elevated inflammation, TH2 cell responses and exacerbated fibrosis in Map3k8 –/–mice was observed in mice with myeloid cell-specific (LysM) deletion of Map3k8, but not CD4 cell-specific deletion of Map3k8, indicating that TPL-2 regulated myeloid cell function to limit TH2-mediated immunopathology. Transcriptional and metabolic assays of Map3k8 –/–M2 macrophages identified that TPL-2 was required for lipolysis, M2 macrophage activation and the expression of a variety of genes involved in immuno-regulatory and pro-fibrotic pathways. Taken together this study identified that TPL-2 regulated TH2-mediated inflammation by supporting lipolysis and M2 macrophage activation, preventing TH2 cell expansion and downstream immunopathology and fibrosis.


Introduction
Immune-mediated pathologies and fibrotic scarring are a major cause of global morbidity and mortality.This is due, in part, to a shortage of available drugs and a lack of novel therapeutic targets to limit fibrogenesis, highlighting a major unmet medical need [3,4].Chronic infection resulting in recurring inflammation and wound repair can lead to tissue remodelling, fibrosis and ultimately organ failure.Infection with the parasitic blood fluke, Schistosoma mansoni, can cause severe intestinal and hepatic pathologies caused by fibrotic lesions surrounding trapped parasite material.Parasite eggs become lodged within vascularised tissue invoking a distinctive eosinophil and macrophage (MF)-rich type-2 immune-mediated granuloma [5].T H 2-cell derived IL-4 and IL-13 [6] stimulate IL-4 receptor (IL-4R)-expressing MF's [7,8] to develop an M2 or alternative activation (AA) state characterised by expression of Arginase (Arg1), Resistin-like molecule alpha (Retnla, Fizz-1) and chitinase-like molecules (Chi3l3, Chi3l4) [9].Animal models have indicated that IL-4R-dependent M2-MF's are essential to 1) prevent fatal intestinal damage and sepsis following schistosome infection [7]; 2) orchestrate tissue remodelling and fibrotic responses [10][11][12] and 3) regulate T H 2 cell proliferation and activation [13][14][15].Despite the clear and well-documented importance of M2-MF's during schistosome infection and the resulting immune-mediated protection, pathology and regulation, the critical regulatory proteins that control M2-MF differentiation are poorly understood.
Two distinct inflammatory pathways contribute to fibrogenic responses; classical, proinflammatory type-1/17 and TGF-β-mediated fibrosis [30] and type-2 inflammatory pathways leading to IL-4R-dependent fibrosis [4].It was recently reported that TPL-2-deficient mice, or inhibition of ERK [31], protected mice from type-1/T H 17 and TGFβ-mediated pulmonary fibrosis following bleomycin treatment [31] and from hepatic fibrosis following carbon tetrachloride and methionine choline-deficient diet-induced fibrosis [32].As expected, Map3k8deficient Kupffer cells had reduced TLR-induced IL-1β and pro-fibrotic gene expression, which the authors suggested was responsible for the reduced hepatic fibrosis in vivo.However this was not directly tested.Nevertheless, this study raised the possibility that targeting TPL-2 may forestall the progression of hepatic fibrosis.Indeed many small molecule inhibitors have been developed that block TPL-2 signalling in vitro [2], but none have yet made it in to the clinic.However, it is not known whether TPL-2 contributes to chronic type-2 inflammation and IL-4R-mediated fibrosis.
In this study, we used the well-established Schistosoma mansoni infection model to test whether TPL-2 regulated chronic type-2 associated inflammation, immunopathology and fibrosis.In contrast to the reduced fibrosis observed in Map3k8 -/-mice following chemical and diet-induced fibrosis [32], Map3k8 -/-mice had significantly increased type-2 immune responses with concomitant elevated inflammation and fibrosis surrounding trapped parasite eggs.Using genome-wide transcriptional analysis and metabolic assays we found that TPL-2 was required for lipid oxidative metabolism and M2-MF activation.Specifically, TPL-2 was required for expression of immunoregulatory molecules (Retnla and Arg1) and regulated profibrotic genes (Col genes and Ctgf).Consequently, myeloid cell-specific deletion of Map3k8 resulted in increased type-2 inflammation and significantly increased fibrosis in vivo, phenocopying Map3k8 -/-mice.Collectively, our study identifies a novel and previously unappreciated role for TPL-2 as a molecular regulator of lipolysis in M2-MF's, regulating type-2 inflammation, immunopathology and hepatic fibrosis.

Map3k8-deficient mice develop increased T H 2-mediated immunopathology and fibrosis following infection with Schistosoma mansoni
Following maturation and worm pairing, gravid worms release hundreds of eggs, many of which traverse the wall of the intestine and are released into the environment via the fecal route.However, many eggs do not successfully reach the intestinal lumen but instead become trapped in the intestinal wall or within vascularised organs, particularly the liver.An eosinophil and MF-rich fibrotic granuloma forms around trapped eggs causing significant tissue damage, orchestrated by CD4 + T H 2 cells and a highly polarised type-2 immune response [5].
To test whether TPL-2 contributed to S. mansoni-associated intestinal and hepatic pathology and fibrosis, we infected Map3k8 -/-mice with 50 S. mansoni cercariae.
CD4 + T H 2 cell-derived IL-4 and IL-13 are essential for granuloma formation [6], mobilising and activating a suite of innate immune cells, including MF's and eosinophils, and promoting local collagen deposition.T H 2 cell-mediated inflammatory responses are controlled by Foxp3 + regulatory T (T REG ) cells [34], which restrain T H 2 cell expansion.It was previously suggested that T cell intrinsic TPL-2 regulates T H 2 [35] and Foxp3 + T REG cell differentiation [36].However, these conclusions were based on in vitro experiments and were not tested in vivo.To determine whether Map3k8 -/-mice had dysregulated T H 2 and Foxp3 + T REG responses following S. mansoni infection, we crossed Map3k8 -/-mice with Il4 gfp and Foxp3 rfp reporter mice, generating dual-reporter Map3k8 -/-mice (Map3k8 -/-Foxp3 rfp Il4 gfp ).These reporter mice allowed us to accurately and simultaneously monitor T H 2 (Il4 gfp+ ) and Foxp3 + T REG (Foxp3 rfp+ ) cells in Map3k8 -/-mice without the requirement for re-stimulation or intra-nuclear staining.Map3k8deficiency did not alter CD4 + CD25 + Foxp3 rfp+ T REG cell frequencies in the spleen, mesenteric lymph node (MLN) or in the local liver tissue, indicating that TPL-2 was not required for T REG cell development or recruitment following S. mansoni infection (Fig 1G, top row).However, CD4 + CD44 + Il4 gfp+ T H 2 cells in both lymphoid tissues and the liver were significantly increased in Map3k8 -/-mice compared to WT mice (Fig 1G, middle row).Map3k8-deficiency also increased the frequency of Il4 gfp+ Foxp3 rfp+ cells in the MLN.
Pharmacological inhibition of MEK1/2, a downstream target of TPL-2, protected mice from bleomycin induced fibrosis [31].We have previously reported that bleomycin-induced fibrosis is mediated by a pro-inflammatory type-1/type-17 and TGFβ driven response, distinct from type-2 mediated pulmonary fibrosis [30].It therefore remained unclear whether TPL-2 contributed to type-2 driven pulmonary fibrosis.To test this we treated mice intravenously with S. mansoni eggs to invoke type-2 inflammation in the lungs leading to the development of pulmonary fibrosis, as previously described [30].Similar to responses in the liver, Map3k8 -/-mice had increased collagen staining in the lung and increased hydroxyproline levels, compared to WT mice given S. mansoni eggs (S2 Fig) T cell-intrinsic TPL-2 does not regulate T H 2-mediated immunopathology following S. mansoni infection It has previously been reported that T cell-intrinsic TPL-2 regulates T H 2 cell differentiation in vitro and acute type-2 inflammation in the airways [35], however it has remained unclear whether T cell-intrinsic TPL-2 regulates T H 2 cell differentiation and function in vivo.To formally test whether T cell-intrinsic TPL-2 contributed to the enhanced inflammation and fibrosis observed in Map3k8 -/-mice (Fig 1) we restricted Map3k8 deficiency to T cells using Cd4 Cre Map3k8 fl/fl mice.Deletion of Map3k8 in T cells (Cd4 Cre Map3k8 fl/fl ) had no impact on methods.E) Expression of Col3 and Col6 was determined from RNA extracted from liver or small intestinal tissue.Data is expressed relative to HPRT.F) Hydroxyproline was quantified in liver tissue from naïve and infected animals.G) Frequency of T REG (CD4 + CD25 + Foxp3 RFP+ ) and T H 2 (CD4 + CD44 + Il4 GFP+ ) cells in the spleen, mesenteric lymph nodes (MLN) and liver were determined by FACS.All experiments are representative of 2-3 independent experiments with 5-10 mice/genotype.* p< 0.05 as assessed by two-tailed Mann-Whitney test. .IL-5 and IL-10 production was significantly increased in re-stimulated MLN cells from Map3k8 -/-mice, compared to WT cells; however production of these cytokines was not affected when Map3k8 was deleted in T cells only (Fig 2E).IL-17 production was low and unchanged between all groups, however IFNγ secretion from lymph node cells was reduced in Map3k8 -/ mice and Cd4 Cre Map3k8 fl/fl mice, in line with a previous report [18].To further test whether T cell intrinsic TPL-2 was required for T H 2 cell differentiation, we isolated naïve T cells (TCRβ +- CD4 + CD44 _ ) from WT and Map3k8 -/-mice and polarised them under T H 1 or T H 2 conditions in vitro.Similar frequencies of IFNγ + or IL-4 + cells were observed between WT and Map3k8 -/ -T cells, respectively (Fig 2F ), suggesting that T cell-intrinsic TPL-2 does not contribute to T H 1 or T H 2 differentiation in vitro.Taken together, these data indicate that T cell-intrinsic TPL-2 is required for optimal IFNγ secretion in vivo, but does not contribute to T H 2 cell differentiation in vitro or in vivo, and that T cell-intrinsic TPL-2 does not contribute to T H 2 cell-mediated immunopathology following S. mansoni infection.

Map3k8 regulates lipolysis for proficient alternative activation of Macrophages
Oxidative lipid metabolism is a metabolic programme recently reported to be essential for M2-MF activation [44].It also has previously been reported that TPL-2, MEK 1/2 and ERK 1/2 from RNA extracted from liver.Data is expressed relative to HPRT and presented as a fold-change relative in infected WT mice.[ [45][46][47][48] can regulate lipid metabolism in a variety of different cells.We therefore hypothesised that the compromised M2-MF activation of Map3k8-deficient MF's was due to reduced lipid metabolism.To investigate this possibility, we analysed the expression of 220 genes involved in lipid metabolism from the transcriptional data obtained from WT and Map3k8 -/-M2-MF 's (S1 Table) and identified 16 TPL-2-dependent genes involved in lipid metabolism that were up regulated in WT M2-MF's but not in Map3k8 -/-M2-MF's (Fig 7A).These genes included the LDL receptor, Olr1, which is required for lipid uptake [49] and Adipoq encoding adiponectin, which promotes lipid oxidation [50] and the alternative activation of human MF's [51].In addition, Aldh1a2 (also referred to as Raldh2) which catalyses the synthesis of the lipid metabolismpromoting metabolite, retinoic acid, from retinaldehyde [52] and the NFκB-regulated sialyltransferase, St8sia1 [53], which catalyzes the transfer of sialic acid from CMP-sialic acid to GM3 to produce gangliosides, were reduced in Map3k8 -/-M2-MF's, compared to WT M2-MF's.Several of these genes are downstream of TPL-2 (S5 Fig) , supporting our hypothesis that TPL-2 regulates lipolysis in M2-MF's.Together these changes in gene expression in Map3k8-deficient M2-MF's were consistent with TPL-2 signalling regulating lipolysis.
To formally test whether lipid metabolism was compromised in Map3k8 -/-M2-MF's, we used extracellular flux analysis and measured the oxygen consumption rate (OCR) and spare respiratory capacity (SRC, the quantitative difference between maximal uncontrolled OCR, and the initial basal OCR, indicative of commitment to oxidative phosphorylation) in un-stimulated and IL-4/IL-13 mediated M2-MF's.At baseline, un-stimulated WT and Map3k8 -/-MF's had a similar OCR and SRC (S6 Fig) .However, Map3k8 -/-M2-MF's had significantly reduced OCR and SRC (Fig 7B and 7C), indicating that TPL-2 is required for lipid metabolism in M2-MF's, and providing a mechanistic explanation for reduced M2-MF's in Map3k8 -/-mice.
Map3k8 -/-mice and LysM Cre Map3k8 fl/fl mice, which had compromised M2-MF activation, had elevated Th2 cell responses.It has previously been reported that M2-MF's can directly regulate T cell responses [13][14][15].We therefore hypothesised that Map3k8 -/-M2-MF's would not regulate Th2 cell differentiation and proliferation as well as WT M2-MF's.To test this hypothesis, we co-cultured WT or Map3k8 -/-BMMF's with naïve cell trace violet (CTV)-labelled CD4 + CD44 − Il4 gfp-OTII + T cells in the presence of IL-4, IL-13 and OVA and determined the proliferation (CTV dilution) and differentiation (Il4 gfp expression) of T cells.After 3 days, 24% of T cells had proliferated and differentiated when co-cultured with WT MF's (Fig 7D, top row middle panel).However, co-culture of T cells with Map3k8 -/-MF's led to significantly more Th2 cell differentiation and proliferation (~38%, Fig 7D, bottom row middle panel).Finally, to determine whether lipid metabolism contributed to MF-mediated regulation of Th2 cell proliferation and differentiation we pre-treated MF's for 6 hours with Orlistat, an irreversible lipase inhibitor, prior to co-culture with T cells.Orlistat treated WT MF's led to more Th2 cell proliferation and differentiation (~32%, Fig 7D, top row right panel), phenocopying Map3k8 -/-MF's and indicating that lipid metabolism in IL-4/IL-13 activated MF's was required for optimal MF-mediated control of Th2 cell proliferation, as previously reported [44].Of note, Orlistat treated Map3k8 -/-MF's only led to a small increase in Th2 cell proliferation, suggesting that lipid metabolism was already at a minimum in Map3k8 -/-MF's.
Taken together this study has demonstrated that TPL-2 is a critical regulator of immunemediated pathology and fibrosis following S. mansoni infection, functioning as an important metabolic regulator in M2-MF activation.

Discussion
Liver fibrosis and cirrhosis, which is responsible for over 1.5 million fatalities per year [54], can develop following a variety of infectious insults, including chronic infection [4].Diseases TPL-2 regulates pro-fibrotic and immuno-regulatory pathways in M2 macrophages.Bone marrow-derived macrophages were stimulated with IL-4 and IL-13 for 24 hours.Cells were harvested, RNA extracted and genome-wide transcriptional expression was determined by microarray analysis using 3 biological replicates.A) Heat map of differentially regulated genes in un-stimulated and IL-4+IL-13 stimulated cells.B) Ingenuity pathways analysis of transcriptional profiles of differentially regulated genes.C-F) Venn diagram and bar graphs of TPL-2 dependent (1), common (2) and TPL-2-regulated genes characterized by persistent T H 2 cytokine responses, such as chronic helminth infections, are associated with the development of significant tissue pathology and fibrotic scarring.Although the molecular pathogenesis of fibrotic diseases are slowly emerging [4], there are very few novel therapeutic candidates progressing through clinical trials [55], highlighting a significant unmet medical need.
Two distinct inflammatory axes contribute to inflammation-driven fibrosis; type-1/ T H 17 mediated inflammation [30] and type-2 driven fibrosis [56].In this study we established that the Map3 kinase, TPL-2, is an important negative regulator of chronic type-2 inflammationdriven fibrosis following schistosome infection.These data are in contrast to a previous study testing the role of TPL-2 in three models of pro-inflammatory type-1/17-associated fibrosis (carbon tetrachloride-, methionine-choline-deficient diet-and bile duct ligation-induced fibrosis) [57,58].In two of these three models Map3k8 -/-mice had significantly reduced fibrosis [32].These seemingly contrasting results most likely reflect the different inflammatory events contributing to the fibrogenic response.For example, it has been widely reported that TPL-2 is required for pro-inflammatory type-1/17-associated inflammation and immunity [17][18][19][20][21][22][23][24][25][26][27][28][29].It therefore stands to reason that TPL-2 would be required for pro-inflammatory type-1/17-associated fibrosis, as reported by Perugorria and colleagues [32].In contrast, TPL-2 appears to function as a negative regulator of type-2 inflammation in the lung and liver (Fig with increased acute [35] and chronic type-2 inflammation in Map3k8 -/-mice, as presented here.In this context, the exacerbated type-2 inflammatory response in Map3k8 -/-mice resulted in increased fibrosis.If these animal models reflect human disease, focused strategies targeting TPL-2 would benefit from identifying a prognostic biomarker and treating patients with type-1/17-associated fibrosis, rather than patients with type-2-associated fibrosis.
Inflammation-driven fibrosis involves a co-ordinated and often dysregulated wound healing response involving a variety of migratory leukocytes activating local stroma.TPL-2 is expressed in both leukocytes and local stroma and therefore identifying where TPL-2 was regulating the fibrogenic process was essential for us to identify how TPL-2 regulated fibrosis.It has been previously suggested that increased acute T H 2 responses in Map3k8 -/-mice was due to a T cellintrinsic role for TPL-2 [35], however this was not tested in vivo.Similarly, increased intestinal inflammation and tumorigenesis in Map3k8 -/-mice was attributed to a reduced frequency of Foxp3 + T REG cells, [36], however again this was not specifically tested in vivo.Using Map3k8 -/ -Il4 gfp Foxp3 rfp mice and re-stimulated local lymph nodes we identified that TPL-2 negatively regulated the differentiation, expansion and/or recruitment of T H 2 cells, however this was not due to a T cell-intrinsic role for TPL-2, as mice with a T cell-intrinsic deletion of Map3k8 mounted similar T H 2 responses as WT mice.Furthermore, exacerbated hepatic fibrosis observed in Map3k8 -/-mice was not observed in mice with a T cell-specific deletion of Map3k8, indicating that T cell-intrinsic TPL-2 had no impact on type-2-mediated inflammation or fibrosis in these systems.
M2-MF's also supply proline for collagen synthesis and produce a variety of pro-fibrotic factors to promote wound healing, collagen deposition and if dysregulated, fibrotic scarring [59].We observed increased expression of pro-fibrotic genes (Col1a1, Col3a1, Col5a2 and Ctgf) in Map3k8 -/-M2-MF's, which may explain the elevated fibrosis observed in Map3k8 -/-mice.The mechanism of TPL-2-mediated regulation of pro-fibrotic mediators is not completely understood.It was recently reported that TPL-2 regulates hepatocyte growth factor (HGF) production in fibroblasts [60], in part, by reducing sensitivity to TGFβ.In IL-13-dependent fibrosis, TGFβ can inhibit some pro-fibrotic pathways [33].If Map3k8-deficient M2-MF's also have decreased sensitivity to TGFβ, this may explain the increased expression of several pro-fibrotic mediators in Map3k8-deficient M2-MF's, however this requires further study.Together, these two observations provide some explanations for the increased T H 2 cell responses and exacerbated hepatic and pulmonary fibrosis observed in Map3k8 -/-mice.Indeed, specific deletion of Map3k8 in LysM-expressing cells phenocopied Map3k8 -/-mice, with increased T H 2 inflammation and fibrosis following S. mansoni infection or S. mansoni egg injection, compared to WT mice.The LysM Cre system appears to efficiently delete floxed alleles in tissue-resident MF's, but not so well in newly recruited myeloid cells [61,62].The exacerbated immunopathology observed in LysM-Cre Map3k8 fl/fl mice suggests that Map3k8 may have an important function in tissue-resident MF's, rather than newly recruited myeloid cells that may not have efficiently deleted Map3k8.
A paralleled increase in pro-fibrotic factors in Map3k8 -/-M2-MF's identified a novel TPL-2-regulated pro-fibrotic axis in M2-MF's.Whether this was also due to compromised metabolic programme in Map3k8 -/-M2-MF's is currently unclear.In vivo, the increased T H 2-mediated inflammation, as a result of compromised immune regulation by Map3k8 -/-M2-MF's, may further exacerbate pro-fibrotic pathways, providing a severely dysregulated respiratory capacity (SRC), the quantitative difference between maximal uncontrolled OCR, and the initial basal OCR, is depicted in the plot.C) Basal oxygen consumption rates (OCR) and spare respiratory capacity (SRC) in WT and Map3k8 -/-M2 macrophages.D) WT or Map3k8-deficent bone marrow-derived macrophages (BMDM) were generated from 3 individual mice and co-cultured with a pool of cell trace Violet (CTV)-labelled naïve OTII CD4 + CD44 − Il4 gfp-T cells and stimulated with IL-4 and IL-13 for 3 days.Th2 cell differentiation (Il4 gfp expression) and proliferation (CTV dilution) was determined by FACS after 3 days.In some wells BMDM were pre-treated with Orlistat for 6 hours and washed, prior to co-culture.Data is representative of 2-3 independent experiments with a minimum of 3 biological replicates per experiment.* p< 0.05 as assessed by two-tailed Mann-Whitney test.doi:10.1371/journal.ppat.1005783.g007microenvironment.In summary, this study has identified that TPL-2 is an important metabolic regulator M2-MF's in vitro and that myeloid cell intrinsic TPL-2 critically controlled chronic type-2-mediated inflammation and fibrosis in vivo.

S. mansoni infection and egg-induced pulmonary inflammation
Mice were infected percutaneously via the tail with 50 cercariae of a Puerto Rican strain of S. mansoni (NMRI) obtained from Biomphalaria glabrata snails, kindly provided by Dr. Quentin Bickle, LSHTM.Infection intensity was determined following perfusion and granuloma size was determined from 10-20 individual granulomas per tissue sample, measured Image J. Tissue pathology was analysed following Masson's trichrome (Collagen, Blue; Nuclei, black/dark blue; Muscle, cytoplasm, Red) staining of 5μm sections from paraffin embedded samples.Intestinal pathology was determined using a comprehensive scoring system taking into account the level of infiltration, disruption and severity of the intestinal architecture [68].Hydroxyproline content was quantified in liver tissue using a hydroxyproline assay kit according to the manufacturers recommendations (Cambridge Biosciences, UK).Tissue eggs were quantified by digesting a known weight of liver tissue with collagenase and liberase and isolating eggs on a discontinuous percoll gradient, as previously described [69].For intravenous delivery of S. mansoni eggs, eggs were washed extensively in PBS, with 5000k delivered in 200μl of sterile PBS.

Bone marrow derived macrophage culture
Bone marrow cells were plated to a density of 5 x 10 6 cells per 90-mm bacterial Petri dish (Sterilin) in 10ml of DMEM/F-12 Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 with GlutaMAX supplement (Gibco) supplemented with 10% FBS, antibiotics, 20% L-cell conditioned medium, L-Glutamine (1%), HEPES (1%), Sodium Pyruvate (1%) and β-mercaptoethanol (1%).After 4 days of culture, 10ml of additional medium was added and cells were cultured for a further 3 days.Non-adherent cells were washed away and the adherent cells were collected in 5ml PBS with 5% FBS and 2.5mM EDTA.For experiments the cells were re-plated in medium with 1% FBS without L-cell supplement and were incubated overnight before stimulation.Cells were stimulated with IL-4 and IL-13 (20ng/ml) (R&D systems) or LPS (100ng/ml) (Alexis Biochemicals).In some experiments, BMDM were generated from three individual mice and co-cultured with a pool of naïve OTII CD4 + CD44 − Il4 gfp-T cells during stimulation with OVA peptide (323-339) (Invivogen).In some of these co-culture experiments, BMDM were pre-treated for 6 hours with Orlistat (100μM; Cayman), prior to co-culture.

Isolation of hepatic macrophages
Liver tissue was perfused and the organ was collected in gentleMACS columns (Miltenyi Biotec) in incomplete RPMI 1640 (Gibco).The tissues were dissociated in incomplete RPMI 1640 with Liberase TL (0.5mg/ml) (Roche), Collagenase (4μg/ml) (Roche) and DNAse (7.5μg/ml) using the gentleMACS dissociator (Miltenyi Biotec).The partly digested tissue fragments were incubated at 37°C for 45 min, following which the tissues were completely dissociated.The cellular fraction was run through a 100μm filter and the cells were centrifuged at 50g for 3 min, to pellet the cells and the supernatant fraction was centrifuged at 320g for 5min.The cellular fractions from both steps were pooled and collected in 1X HBSS Hanks balanced salt solution and mixed with OptiPrep (Sigma) to get a 17% w/v solution and overlayed with 1X GBSS Gey's balanced salt solution.The samples were centrifuged at 400g for 15min at room temperature with no brakes and the enriched layer of cells were collected from the interface of the GBSS and the 17% solution.The cells were stained and FACS sorted as Live/ CD45 + /CD3 − /CD19 − /NK1.1 − /Ly6G − /SiglecF − /CD11b + /F4/80 + .

Metabolism assays
Day 7 bone marrow-derived monocytes (BMDM) were cultured at a density of 5x10 5 cells in an XF24 plate (Seahorse Bioscience) over night.On day 8 cells were stimulated with 20ng/ml of recombinant IL-4 and IL-13 (R&D systems).On day 9 media was replaced with XF base medium (Seahorse Bioscience) supplemented with 100x Glutamax (Gibco), 100X Sodium Pyruvate (SIGMA) and 25mM glucose and the plate incubated for 10-30 minutes in a non-CO 2 incubator at 37°C.For analysis of basal oxygen consumption rate (OCR) and extracellular acidification rate (ECAR), cells were analysed with XF-24 Extracellular Flux Analyzer (Seahorse Bioscience) under standard Seahorse running protocol.The Seahorse cartridge was loaded at a final concentration of 10μM Oligomycin (SIGMA O4876-5mg), 15μM FCCP (triflourocarbonylcyanide phenylhydrazone, SIGMA C2920-10mg), 1μM Rotenone (SIGMA R8875) plus 10μM Antimycin A (A8674-25mg) in ports A, B and C respectively.The bioenergetics profile consisted of basal OCR measurements in the absence of drugs and OCR/ECAR following the injection of drugs.All OCR/ECAR/SRC analyses were obtained from 5 replicates in 3 independent repeats.
LC-MS for arginase assay.At day 7, non-adherent cells were washed away and the adherent cells were collected in 5ml PBS with 5% FBS and 2.5mM EDTA.Cells were re-plated in medium with 1% FBS without L-cell supplement in a 6-well plate.Cells were kept for 12-18 hours, then stimulated with IL-4 and IL-13 (20ng/ml) (R&D systems) for 24hrs.Cells were washed twice with PBS, metabolites were extracted with ice cold mixture of acetonitrile/methanol/water (2:2:1, v/v/v).Plates were kept on dry ice for 5-10 min.Cells were collected, vortexed vigorously then spun for 20 min at 16,000g at 4°C.The supernatant was stored at -80°C until analysis.Residual protein content in the cell lysates was determined with Bicinchoninic acid method (BCA assay, Pierce).Prior to the LC-MS analysis, 100μL of the cell lysates were acidified with an equal volume of acetonitrile containing 0.2% acetic acid.The mixture was spun for 10 minutes at 16,000g, at 4°C. 2μL of the acidified mixture were used for the LC-MS analysis.Chromatography was performed on Agilent 1200 LC system, including a solvent degasser, binary pump and temperature-controlled auto-sampler.Samples were inject to a Cogent Diamond Hydride Type C silica column (150mm×2.1mmi.d, 4μm particle size and 100A pore size) and eluted with flow rate of 0.4 mL/min.The gradient employed is based on the number 3, according to [70].Metabolites were detected using an Agilent Accurate Mass 6230 TOF apparatus, as previously described [71].Detected m/z 175.11895 at 16.8 min and 133.0972 at 16.6 min were identified as arginine and ornithine, respectively, on the basis of unique accurate mass-retention time identifiers and spectral data.The reported abundances were normalized to the residual protein content in each corresponding sample.
qRT-PCR, ELISA, western blotting and serum LP qRT-PCR.Tissue samples were frozen in RNAlater (Sigma) and homogenized in QIAzol (Qiagen).Cell pellets were lysed in Buffer RLT (Qiagen).Total RNA was isolated as per manufacturer's protocol.100ng-1μg of total RNA was reverse transcribed using the QuantiTect Reverse Transcription Kit (Qiagen).cDNA produced was used for real time quantitative PCR with Power SyBrGreen (Applied Biosystems).The expression levels of different genes were normalised to hypoxanthine-guanine phosphoribosyl transferase (HPRT) expression and expressed as fold change relative to naïve or PBS-treated WT samples.
Western blotting.For immunoblotting, cell lysates were normalized to equal total protein content and resolved on 10% Criterion TGX Gels (Biorad).Separated proteins were transferred onto Trans-Blot Turbo PVDF transfer (Biorad) membranes (Immobilon).Specific bound antibodies were visualized by chemiluminescence (Merck Millipore).Endotoxin levels in serum samples were measured using an LAL chromogenic assay, according to manufacturer's recommendations (Pierce, Thermo fisher).

RNA, qRT-PCR and microarray
RNA was isolated from tissues and cells using RNAeasy mini spin columns according to manufacturers' instructions (Qiagen).cDNA was generated from 5ng of total RNA using WT-Ovation Pico system (version 1) RNA Amplification System followed by double stranded cDNA and 1B), despite a similar egg burden (S1 Fig) and serum LPS level as S. mansoni-infected WT mice (S1 Fig).Similarly, intestinal inflammation was also significantly increased in Map3k8 -/-mice (Fig 1C and 1D).Consistent with the increased collagen staining observed in Map3k8 -/-mice, collagen-synthesising genes, Col3 and Col6, were both significantly elevated in the liver and small intestine of Map3k8 -/-mice, compared to WT controls (Fig 1E), with significantly more hydroxyproline in the liver of Map3k8 -/-mice (Fig 1F).

Fig 1 .
Fig 1. Map3k8 -/-mice develop increased hepatic and intestinal inflammation and fibrosis following S. mansoni infection.WT and Map3k8 -/-mice were infected percutenously with 50 S. mansoni cercariae and analysed at 8 weeks post-infection.A & C) Perfused tissue was fixed and embedded in paraffin before sectioning and staining with Masson's trichrome.B) Granuloma size was determined from 10-20 individual granulomas per sample measured using Image J. Scale bars are 1000μm (top), 200μm (middle) and 100μm (bottom).D) Intestinal pathology score, as described in In the lung tissue and local draining thoracic lymph nodes (TLN), Map3k8 -/-mice had increased Th2 cell frequency (S2 Fig) promoting increased Il13, Col6 and Mmp12 expression in the lung (S2 Fig).Collectively, these data indicate that TPL-2 is an important negative regulator of type-2 inflammation, immunopathology and fibrosis following S. mansoni infection or S. mansoni egg induced pulmonary fibrosis in vivo.

doi: 10 .
1371/journal.ppat.1005783.g001granuloma development in the liver (Fig 2A and 2B) or small intestine (Fig 2C) following S. mansoni infection.Similarly, fibrosis (Fig 2A and 2C) and expression of collagen synthesising genes, Col3 and Col6, were not affected following the deletion of Map3k8 in CD4 + cells (Fig 2D) restricted Map3k8 deletion to Lysozyme M-expressing cells using LysM Cre Map3k8 fl/fl mice (S3 Fig).Mice with myeloid cell-specific deletion of Map3k8 had significantly more inflammation with larger hepatic (Fig 3A and 3C) and intestinal (Fig 3B) granulomas and more severe intestinal pathology (Fig 3D), without any appreciable change in serum LPS (S3 Fig).Of note, a distinct collagen-rich fibrotic ring surrounded hepatic granulomas in LysM Cre- Map3k8 fl/fl mice, which was absent in mice with WT myeloid cells.Increased collagen staining in the liver was supported by increased expression of collagen-synthesising genes, Col3 and Col6 (Fig 3E) and increased hydroxyproline (Fig 3F).Similar to Map3k8 -/-mice, mice with myeloid cell-specific deletion of Map3k8 had elevated type-2 cytokine secretions (IL-13, IL-5 and IL-10) following lymph node re-stimulation without any appreciable change in IFNγ or IL-17A secretion (Fig 3E).Similarly, elevated expression of Il13 but not Il1b, Tgfb, Il17a, Ifng, Tnfa or Il6 was observed in LysM Cre Map3k8 fl/fl mice, compared to control mice (S3 Fig).These data clearly indicated that macrophage/myeloid cell intrinsic-TPL-2 contributed significantly to the regulation of T H 2-mediated inflammation and fibrosis following S. mansoni infection.

Fig 2 .
Fig 2. T cell-intrinsic Map3k8 does not contribute to exacerbated inflammation and pathology following S. mansoni infection.Cd4 Cre Map3k8 +/+ and Cd4 Cre Map3k8 fl/fl mice were infected percutenously with 50 S. mansoni cercariae and analysed at 8 weeks postinfection.A-C) Perfused tissue was fixed and embedded in paraffin before sectioning and staining with Masson's trichrome.B) Granuloma size was determined from 10-20 individual granulomas per sample measured using Image J. D) Expression of Col3 and Col6 was determined from RNA extracted from liver or small intestinal tissue.Data is expressed relative to HPRT and shown as a fold-change relative to uninfected mice.E) Mesenteric lymph node cells were re-stimulated with anti-CD3 for 3 days.Cytokines were measured in supernatants, by ELISA.F) Naive T cells (CD4 + CD44 − CD25 − CD62L + ) were FACS purified from WT and Map3k8 -/-mice and cultured under T H 1 and T H 2 conditions.Frequencies of CD44 + IFNγ + and CD44 + IL-4 + cells were determined by intracellular FACS analysis on day 7.All experiments are representative of 2-3 independent experiments with 5-10 mice/genotype.* p< 0.05 as assessed by two-tailed Mann-Whitney test.doi:10.1371/journal.ppat.1005783.g002

Fig 5 .
Fig 5. TPL-2 is required for M2 activation of Macrophages, in vitro.A-D) Bone marrow-derived macrophages (BMDM) were stimulated with IL-4 and IL-13 for 6 or 24 hours, as indicated.Cells were harvested, RNA extracted and gene expression was determined by qRT-PCR and expressed relative to un-stimulated genotype control cells.E) BMDM's were stimulated for 1.5, 3 and 6 hours, as indicated.Total Protein was extracted with phosphorylated and total protein levels of STAT6, ERK, p38a, JNK and α-tubulin determined by western blot.All experiments are representative of 2-3 independent experiments with 3-5 biological replicates and 3 technical replicates in each experiment.* p< 0.05 as assessed by two-tailed Mann-Whitney test.doi:10.1371/journal.ppat.1005783.g005

Fig 6 .
Fig 6.TPL-2 regulates pro-fibrotic and immuno-regulatory pathways in M2 macrophages.Bone marrow-derived macrophages were stimulated with IL-4 and IL-13 for 24 hours.Cells were harvested, RNA extracted and genome-wide transcriptional expression was determined by microarray analysis using 3 biological replicates.A) Heat map of differentially regulated genes in un-stimulated and IL-4+IL-13 stimulated cells.B) Ingenuity pathways analysis of transcriptional profiles of differentially regulated genes.C-F) Venn diagram and bar graphs of TPL-2 dependent (1), common (2) and TPL-2-regulated genes

Fig 7 .
Fig 7. TPL-2 regulates lipolysis in M2 macrophages and regulation of T H cell differentiation and proliferation.Bone marrow-derived macrophages (BMDM) were stimulated with IL-4 and IL-13 for 24 hours.A) Analysis of genes involved in lipid metabolism was performed by Ingenuity pathways analysis (S1 Table)with Map3k8-dependent genes depicted in a heat map.B) After 24hrs of stimulation with IL-4 and IL-13 oxygen consumption rates (OCR) were determined in M2 macrophages using an XF-96 Extracellular Flux Analyzer (EFA) during sequential treatments with oligomycin, FCCP, and rotenone/antimycin.Spare