Analysis of the Trichuris suis excretory/secretory proteins as a function of life cycle stage and their immunomodulatory properties

Parasitic worms have a remarkable ability to modulate host immune responses through several mechanisms including excreted/secreted proteins (ESP), yet the exact nature of these proteins and their targets often remains elusive. Here, we performed mass spectrometry analyses of ESP (TsESP) from larval and adult stages of the pig whipworm Trichuris suis (Ts) and identified ~350 proteins. Transcriptomic analyses revealed large subsets of differentially expressed genes in the various life cycle stages of the parasite. Exposure of bone marrow-derived macrophages and dendritic cells to TsESP markedly diminished secretion of the pro-inflammatory cytokines TNFα and IL-12p70. Conversely, TsESP exposure strongly induced release of the anti-inflammatory cytokine IL-10, and also induced high levels of nitric oxide (NO) and upregulated arginase activity in macrophages. Interestingly, TsESP failed to directly induce CD4+ CD25+ FoxP3+ regulatory T cells (Treg cells), while OVA-pulsed TsESP-treated dendritic cells suppressed antigen-specific OT-II CD4+ T cell proliferation. Fractionation of TsESP identified a subset of proteins that promoted anti-inflammatory functions, an activity that was recapitulated using recombinant T. suis triosephosphate isomerase (TPI) and nucleoside diphosphate kinase (NDK). Our study helps illuminate the intricate balance that is characteristic of parasite-host interactions at the immunological interface, and further establishes the principle that specific parasite-derived proteins can modulate immune cell functions.


T. suis genes and excretory/secretory proteins (TsESP) display stage-specific expres-
sion. Genomic sequencing of T. suis predicted the presence of 9,832 genes, comparable to T. muris and T. trichura (9,403 and 9,856, respectively), with an average open reading frame length of 2,384 bp (Supplementary  Table S1). To assess changes in gene expression as a function of life cycle stage, transcriptomic analysis was performed on 10-, 16-, 21-, 28-day larvae (L 2 , L 3 , and L 4 , respectively) and 42-day adult worm (L 5 ) stages and male or female adult worms and these results were compared to ES products identified by proteomics.
To examine ESP released by T. suis, parasites isolated from infected pigs at 10, 16, 21, 28, and 42 days post-inoculation and cultivated in vitro for up to 72 h. Conditioned medium was collected for protein analysis. SDS-PAGE analysis showed that proteins with MW ranging from ~10-180 kDa were released by 21-and 28-day larvae and adult worms (Fig. 1A). Protein output was notably lower for 10-and 16-day larvae. Proteomic analysis of TsESP isolated from the various life stages collectively identified ~350 proteins using single preparations for 10-, 16-, and 21-day larvae, and three biological replicates for 28-day larvae and adult worms (Supplementary  Table S2). Based on Gene Ontology (GO) analysis and biological function, these proteins were classified into 37 groups with structural proteins, glycolytic enzymes, chaperones, proteases, protease inhibitors, and uncharacterized proteins representing the most abundant classes (Fig. 1B). Interestingly, ~20 of the uncharacterized proteins detected in this analysis appeared to be specific to Trichuris and shared very little or no sequence homology with other helminth proteins (Supplementary Table S3). Bioinformatic analyses showed that ~26% of the proteins in the TsESP proteomes contained a signal sequence and ~15% contained 1-4 predicted transmembrane domains.
Differential gene expression analysis was performed using DESeq2 across 3 sample sets consisting of a total of 8 RNA-Seq samples (early larvae: 10-, 16-, 17-, and 21-day; larvae: 28-day; adult, 3 replicates; Supplementary  Table S3). This analysis identified, (i) 1,226 genes that were more highly expressed in early larvae and enriched for the GO terms related to ribosome, kinase, and peptidase activity, and less likely than other genes to be T. suis-specific (relative to T. trichiura, T. muris, and 3 host species; 261 genes, P = 1 × 10 −7 ; Supplementary  Table S4), (ii) 208 genes were more highly expressed in 28-day larvae and enriched for GO terms related to cell adhesion and signaling, and motor activity, and (iii) 662 genes more highly expressed in adult worms, enriched for GO terms related to kinase activity, toxin responses, and carbohydrate metabolism, and genes likely to be T. suis-specific (226 genes, P = 0.0001; Fig. 2A, Supplementary Tables S3, S4). Gene expression levels corresponding to protein sets detected by proteomics (Fig. 2B) are shown for the early larvae (Fig. 2C), the 28-day larvae (Fig. 2D), and the adult (Fig. 2E) RNA-Seq samples, indicating that highly-expressed genes in each stage are more likely to be detected by proteomic analysis; since these proteins appear to be targeted for release into the extracellular environment. Genes corresponding to proteins detected in early larvae TsESP proteomic studies are more likely to be overexpressed in the early larvae in comparison to other life cycle stages (P = 0.05), as are those corresponding to proteins detected in the 28-day larvae (P = 0.04) and the TsESP (P = 0.02; Fig. 2F, Supplementary  Table S3). In sum, the proteomic and transcriptomic analyses revealed life stage-specific expression patterns of ES proteins which might reflect the different developmental requirements of the larval stages and suggest that the host immune response is exposed to a concoction of parasite-derived molecules that varies in composition as the parasite develops. TsESP inhibit pro-inflammatory cytokine production and induce IL-10 secretion. TsESP from adult worms was shown to suppress inflammatory responses in human monocyte-derived dendritic cells 29 . Similarly, TsESP collected from L 1 hatched in vitro reduced clinical signs of allergic airway disease in a murine model 30 . One aim of this study was to directly compare the immunomodulatory effects of TsESP released from larvae and adult worms. To examine these effects, a concentration-dependent experiment was first performed by pre-incubating BMDM with 1-50 µg/mL of TsESP from 28-day larvae and adult worms. The addition of TsESP alone failed to induce TNFα production (Fig. 3A); however, inhibition of TNFα release (>70%; P < 0.001) was observed in BMDM cultures pretreated with adult TsESP as low as 1 µg/mL prior to stimulation with interferon-gamma (IFNγ) and LPS, potent inducer of TNFα. Although a similar concentration-dependent response was observed with 28-day larvae TsESP, higher concentrations (≥25 µg/mL) were required to reduce TNFα secretion (>69%; P < 0.001) (Fig. 3A), which could reflect differences in the levels of specific immunoactive T. suis proteins in these TsESP preparations. In contrast, BMDM pretreated with TsESP alone was sufficient to produce robust levels of the anti-inflammatory cytokine IL-10 ( Fig. 3B).
Indeed, the addition of ≥25 µg/mL of 28-day larvae or ≥1 µg/mL of adult worm TsESP triggered a marked increase in IL-10 secretion (P < 0.001) in unstimulated or IFNγ/LPS-stimulated BMDM cultures. Interestingly, TsESP from adult worms elicited significantly greater IL-10 levels compared to 28-day larvae TsESP when tested at equal concentrations. In addition, pretreatment of BMDC with 28-day larvae and adult worm TsESP reduced IL-12p70 secretion by ~80% (P < 0.001) following stimulation of cells with CpG-ODN (Fig. 3C), a TLR9 ligand and known inducer of IL-12p70. Control experiments showed that TsESP treatment alone failed to stimulate production of IL-12p70 by BMDC. To verify that the immunomodulatory effects were linked to specific protein component(s), TsESP preparations were subjected to extensive proteolysis at an elevated temperature to degrade proteinaceous material. Addition of TsESP to BMDC prior to CpG-ODN stimulation decreased IL-12p70 secretion ( Supplementary Fig. S1); however, treatment of 28-day larva or adult worm TsESP with proteinase K 38 abolished the capacity of TsESP to inhibit IL-12p70 production following CpG-ODN stimulation. In fact, protease treatment dramatically augmented the stimulatory effect on IL-12p70 production, resulting in significantly higher levels of cytokine secretion as compared to CpG-ODN control cultures. Collectively, these data revealed that TsESP from larvae and adult worms altered cytokine profiles in BMDM and BMDC despite concomitant treatment with exogenous pro-inflammatory stimuli. Moreover, these observations not only confirm the immunosuppressive properties of TsESP on innate immune cells, but also highlight the greater potency of adult worm TsESP as compared to day-28 larvae.
TsESP induce nitric oxide (NO) production and arginase-1 (ARG-1) expression. Parasitic nematodes skew the polarization of macrophages from classically-activated (M1) to alternatively-activated macrophages or suppressive M2 through different mechanisms 29,39,40 . The general paradigm for macrophage polarization states that M1 produce high levels of NO, while M2 display increased levels of ARG-1 41,42 . Surprisingly, Western blot analyses revealed that treatment of BMDM with 28-day and adult TsESP alone induced expression of nitric oxide synthase 2 (NOS2 or iNOS) and ARG-1 (Fig. 4A). Concomitant treatment of BMDM with a combination of IFNγ and LPS with or without TsESP resulted in robust upregulation of NOS2; however, the levels were notably higher following treatment with adult TsESP. TsESP from 28-day larvae had a less pronounced effect, regardless of the addition of exogenous cytokines or LPS. In contrast, TsESP treatment alone caused an increase in ARG-1 expression, while concomitant addition of IL-4 and IL-13 with or without IL-10 drastically enhanced its expression in TsESP-treated and control cultures.
To determine if increased NOS2 expression resulted in enhanced NO production, levels of NO were estimated in the culture medium by measuring nitrite concentrations. Treatment of BMDM with either 28-day larvae or adult worm TsESP alone was sufficient to triggered a sharp increase in NO levels (P < 0.01) (Fig. 4B). Furthermore, NO concentrations were higher in cultures pretreated with TsESP followed by exposure to a combination IFNγ and LPS (P < 0.05), factors known to induce NO production 43 . Surprisingly, TsESP-induced NO production by BMDM remained elevated despite treatment with IL-4 and IL-13 with or without IL-10 (P < 0.001), both previously shown to inhibit NOS2 transcription [44][45][46] . Elevated arginase activity was also observed when cells were exposed to TsESP with or without additional IFNγ and LPS treatment (Fig. 4C). Consistent with the increased expression of ARG-1 (Fig. 4A), BMDM treated with IL-4 and IL-13 showed a robust upregulation in arginase activity which was further elevated in the presence of IL-10. Addition of TsESP had a pronounced additive effect with the latter cytokines leading to higher arginase activity as compared to control cultures (P < 0.01) (Fig. 4C). The simultaneous upregulation of NOS2 and ARG-1 suggests that TsESP did not polarize BMDM to an M1 or M2 phenotype, as typically defined, but rather towards a phenotype reminiscent of myeloid-derived suppressor cells (MDSC) 47 with potential suppressive abilities.

TsESP-polarized myeloid cells suppress antigen-specific CD4 + T cell proliferation.
Given the phenotype observed with TsESP-treated myeloid cells, we sought to determine if these cells could alter antigen-specific expansion of CD4 + T cells. Co-cultures of BMDC pulsed with OVA 323-339 peptide and OT-II CD4 + T cells induced robust T cell division as shown by the dilution of the CFSE signal (Fig. 5A), which corresponded to a ~60% increase in T cell proliferation (Fig. 5B). However, treatment of BMDC with adult worm or larval TsESP prior to co-culturing caused a marked and significant inhibition of OT-II CD4 + T cell division (~40% reduction in cell proliferation as compared to medium control) after a 72 h incubation; Fig. 5A,B). Determination of cytokine levels in the supernatants revealed high levels of IL-10 in cultures containing BMDC pretreated with adult worm TsESP, while supernatants from the medium control and BMDC pretreated with 28-day larvae TsESP contained only minimal levels of IL-10 ( Fig. 5C). A similar analysis showed that cultures containing BMDC pretreated with TsESP either from adult worms or day-28 larvae had increased levels of IFNγ and TNFα (Fig. 5D,E).
To assess which cells in BMDC:OT-II CD4 + T cell co-cultures released cytokines (IL-10, IFNγ, TNFα), intracellular cytokine staining was performed after a 48 h incubation. Cells were stained for CD11c or CD4 cell surface markers to differentiate BMDC and T cells, respectively, and doubly stained to detect intracellular levels of IFNγ and IL-10 or IFNγ and TNFα. Dual intracellular staining of T cell populations following stimulation with OVA peptide presented by BMDC showed that 2.4-3.5% of CD4 + T cells expressed IFNγ (Fig. 5F). Co-staining of the latter population for TNFα confirmed that ~2.0% of the CD4 + T cells in the medium control and TsESP-treated cultures expressed IFNγ and TNFα. The frequency of IFNγ + and TNFα + CD4 + T cells increased to ~3.0% in cultures treated with TsESP. CD4 + T cells expressing both IFNγ and IL-10 were detected at a frequency of ~1.7% in the medium control and 28-day larvae TsESP while the frequency of IFNγ + IL-10 + CD4 + T cells increased to ~2.8% in the adult worm TsESP-treated cultures, which may account for the notably higher level of IL-10 detected in the culture supernatant following pretreatment of BMDC with adult TsESP (Fig. 5E).
It was previously reported that exposure to ESP from other intestinal nematodes H. polygyrus and T. circumcincta induced differentiation of naïve splenocytes into CD4 + CD25 + FoxP3 + T reg cells 26 , an effect we did not observe with TsESP treatment (Supplementary Fig. S2). Flow cytometric analyses showed that the expression of the CD25 and FoxP3 markers in concanavalin (Con) A-stimulated splenocytes did not differ when control cultures (4.1 ± 1.4%) were compared to TsESP-treated cultures (28-day larvae 5.3 ± 1.3%; adult worm 5.0 ± 1.5%). In contrast, ESP from adult H. polygyrus worms induced a significant increase (14.1 ± 2.7%, P < 0.05) in de novo CD4 + CD25 + FoxP3 + T cells, as previously reported 26 . Overall, these data suggest that the suppressive effects of TsESP are mediated by TsESP-polarized myeloid cells that limit anti-specific CD4 + T cell proliferation without directly inducing T reg cells.
To identify proteins linked to the immunomodulatory activity, the three pools of GPC fractions with highest bioactivity (16-18, 19-21, and 28-30) were subjected to LC-MS/MS analysis. Collectively, 33 proteins were  Table S5). With the exception of ferritin, Sm ribonucleoprotein, aspartyl aminopeptidase, and α-mannosidase 2C1, all other proteins were previously observed in the crude 28-day larvae or adult worm proteomes. In fractions 16-18, angiotensin-converting enzyme was the only protein also detected in the crude 28-day larvae and adult worm proteomes. The bulk of proteins exhibiting TNFα inhibitory and IL-10 stimulating activity were in fractions 19-21; although several of these proteins were also found in fractions 28-30. Among these, there were a number of proteases, glycolytic enzymes, kinases, dehydrogenases, and the iron and cholesterol binding proteins, ferritin and the Niemann-Pick C1 protein, respectively. A number of uncharacterized proteins were also detected, with proteins D918-01198 and D918-10108 being specific to Trichuris spp. (Supplementary Table S5).
Since the amount of protein found in each of the GPC bioactive fractions was limiting, it precluded the possibility of performing additional chromatographic separation steps to obtain single highly purified protein components. Therefore, we generated recombinant versions of several T. suis proteins, including triosephosphate isomerase (TPI), nucleoside diphosphate kinase (NDK), and a small nuclear ribonucleoprotein (SNRP), in an E. coli expression system to assess and validate their immunomodulatory activity. These three candidates were selected for initial studies since they are found in ESPs from multiple helminths. Bioinformatic analysis showed that T. suis SNRP, TPI, and NDK have predicted MWs of 8.0, 27, and 26 kDa and share ~83, 61, and 75% sequence identity with their mammalian homologues, respectively. An interesting feature of the T. suis NDK is a 76 amino acid C-terminal extension that is absent from the mammalian homologue, but shares ~40% sequence identity with autophagy related protein 2 from Trichinella. Treatment of BMDM with recombinant TPI and NDK inhibited IFNγ/LPS-induced TNFα secretion in a concentration-dependent manner (25 to ~6 µg/mL for TPI, and 25 to ~3 µg/mL for NDK) (Fig. 7A). In contrast, SNRP had no inhibitory activity at the tested concentrations, but instead appeared to potentiate TNFα production.
To examine the cellular mechanisms underlying the anti-inflammatory effects of TPI and NDK, we assessed by Western blots the phosphorylation (activation) and expression of different mediators. Specifically, we monitored the phosphorylation of STAT3 and the expression of CCAAT/enhancer-binding protein β (C/EBPβ), two transcription factors critical in MDSC functions 47 . Treatment of BMDM with adult TsESP, TPI, or NDK without additional stimulation led to the phosphorylation of STAT3 (Y705) and induced expression of C/EBPβ) (Fig. 7B). Also, expression of nuclear factor interleukin-3 regulated (NFIL-3), a target gene of STAT3-mediated transcription and a suppressor of Il12b transcription 48 , was induced upon exposure to TsESP. Of note, STAT3, C/EBPβ, and NFIL3 are all important transcription mediators driving the expression of numerous anti-inflammatory genes [48][49][50][51] . In addition, a marked increase was observed in the expression of TTP (tristetraprolin), a protein which binds specific mRNAs, such as TNFα and IFNγ 52 , and targets them for degradation. Of note, native TsESP elicited a stronger response for these targets, suggesting that several proteins in this mixture contribute to the effect. In sum, the identification of potential protein candidates supports the concept that specific parasite-derived molecules can be linked to the immunomodulatory properties of TsESP and may promote the immunosuppressive functions of myeloid cells.  from these parasites [18][19][20][21][22][23][24][25][26][28][29][30] . In this study, we have focused on proteins released by T. suis; however we recognized that these helminths release other small metabolic products such as PGE2 which are capable of inhibiting pro-inflammatory responses 59 .
Our results with TsESP are largely in agreement with two previous reports, the first of which demonstrated the anti-inflammatory properties of adult TsESP on intestinal epithelial cells (IECs) and monocyte-derived dendritic cells 29 , and the second, which showed the immunosuppressive properties of L 1 larvae TsESP in vitro and in a murine model of allergic airway disease 30 . Our study extends these findings by revealing that TsESP from adult worms is more active than TsESP derived from L 4 larvae (28-day), which may be linked to stage-specific expression profiles and/or the relative abundance of bioactive proteins, as suggested by the results of our transcriptomic and proteomic analyses. These observations offer insights into changes that occur between the parasite and the host immune system as the larvae develop into adult worms. The subtle yet significant differences in immunomodulatory abilities of different life stages could represent important considerations in the design of a novel therapeutic tool for addressing pathologies associated with immune-mediated disorders.
Since the time that this work was performed, an additional two T. suis genome assemblies (female and male) have been published 60 (Supplementary Table S1). These two genome assemblies were slightly larger than our assembled genome (71 mbp and 74 mbp for female and male respectively, vs 64 mbp for our assembly) and were annotated with more protein coding genes (14,261 and 14,436, vs 9,832 genes). However, our protein coding gene count was comparable to other published Trichuris species (9,403 for T. muris and 9,856 for T. trichiura 61 ), and our genome completion estimation using BUSCO 62 indicated that our genome was 95% complete similar to the 94.7% completion rate for the other two published versions (Supplementary Table S1), suggesting that we are capturing the complete functional gene set with fewer genes. Furthermore, deduced protein sequence similarity matching using NCBI BLASTp (v2.6.0+) indicated that 90.7% and 90.5% of our genes had matches (E ≤ 10 −5 ) in the previously published female and male T. suis genomes, respectively 60 , while the two genomes had 94.5% of genes matching each other. The higher gene count in the published genomes may be the result of gene fragmentation resulting from having more than 10-fold as many contigs in the genome assembly (Supplementary Table S1). Here, we identify biological functions enriched among early larvae (kinases, peptidases, T. suis-specific gene functions), 28-day larvae (cell adhesion, motor activity), and adult stage T. suis (toxin responses, carbohydrate metabolism, T. suis-specific function) using DESeq2 to identify differentially-expressed genes (Supplementary Table S4). The previous genome paper 60 also had RNA-Seq datasets from L1/L2, L3, L4 and adult stages (in addition to tissue-specific datasets from the male and female posterior body, and the stichosome) but did not perform a similar statistical differential analysis for enrichment for comparison to our results. Hence, our study supports but also complements previous reports by assessing stage-specific gene expression profiles and enrichment of specific biological functions, thus giving greater insight into the developing worms.
Although CpG and LPS (TLR9 and TLR4 ligands, respectively), and IFNγ can trigger distinct signaling events, they also activate common signaling cascades suck as MAPK (e.g. p38, ERK) and increase immune-related gene expression by enhancing the activity of different transcription factors including NFκB and AP-1 63,64 . In our study, we have observed that TsESP treatment was able to inhibit both CpG-mediated IL-12p70 and IFNγ/LPS-mediated TNFα production by innate immune cells. While it remains to be established if the parasite-derived molecules target multiple independent pathways or common downstream effectors of TLR4, TLR9, and IFNγR, our results demonstrate the potency of these molecules to inhibit inflammation without being limited to a single stimulus. This observation is an important consideration especially in the context of treatments of immune-mediated disorders, which are often caused by various molecular cues 4 . In addition, our observations offer valuable clues as to potential signaling events affected which could direct future studies. Furthermore, our study revealed an interesting effect of TsESP on the polarization of primary macrophages and dendritic cells. Macrophages treated with the parasite products adopted several features akin to myeloid-derived suppressor cells (MDSC), such as the simultaneous expression of NOS2 and ARG1, production of IL-10, higher levels of phosphorylated STAT3, and expression of C/EBPβ (reviewed in 47 ). Moreover, TTP, a potent repressor of translation of inflammatory mediators, can be constitutively expressed by tumor-associated macrophages (TAM) 65 similar to our findings in BMDM treated with TsESP. Another transcription factor induced by TsESP treatment was NFIL3; this protein is important in colonic macrophages to limit inflammation upon exposure to enteric microbiota 48 . Interestingly, Nfil3 −/− mice develop microbiota-dependent colitis 66 . The expansion of MDSC during infections with Taenia crassiceps 67 , H. polygyrus 68,69 , and Schistosoma japonicum 70 has been reported; however, evidence for the involvement of ES products in the generation of immunosuppressive myeloid cell subsets in this context would be an important addition to our understanding of host-pathogen interactions. In addition, TsESP-treated dendritic cells display the ability to inhibit antigen-specific CD4 + T cell expansion 47 . Curiously, co-culture of TsESP-treated BMDC and OT-II CD4 + T cells stimulated with OVA peptide generated IFNγ + IL10 + double-positive CD4 + T cells. This phenotype has been observed in leishmaniasis 71 and toxoplasmosis 72 . The underlying mechanism involves an early phase of IFNγ to eradicate the pathogen followed by a regulatory phase or "cytokine switch" during which IL-10 is produced by the same IFNγ + cells, essentially leading to a negative feedback loop to limit inflammation 73 . Hence, TsESP could modulate CD4 + T cell responses indirectly by altering the phenotype of antigen presenting cells.
Although there is an unequivocal propensity of TsESP to inhibit inflammatory responses, the identity of the proteins with immune-altering activity has remained largely undetermined for T. suis and other nematodes. Using a combination of chromatographic techniques and proteomic analyses, we demonstrated that subsets of TsESP display immunomodulatory capabilities that recapitulate to a large extent effects observed with native TsESP. Three candidates were chosen based on their presence within biologically active GPC fractions and previous literature reports of orthologs in ESP from many other parasitic helminths, namely TPI and NDK, which displayed immunosuppressive activity, and SNRP, which did not have the effects seen with TPI and NDK mimicking the bioactivity observed in native TsESP. TPI is a ubiquitously expressed enzyme that catalyzes the aldose-ketone isomerization of dihydroxyacetone phosphate to glyceraldehyde-3-phosphate, a key reaction in the glycolysis pathway. TPI has been previously identified as the most abundantly secreted protein by Brugia malayi, the causative agent of human lympathic filariasis, and is required for optimal fecundity 74 . TPI is also highly secreted by other parasites, such as the nematodes Meloidogyne incognita 75 and Haemonchus contortus 76 , and the trematode Schistosoma mansoni 77 . It remains unclear, however, how this enzyme alters the phenotype of host immune cells.
Evidence for a moonlighting function of TPI has been reported for Trichomonas vaginalis 78 and the fungal pathogen Paracoccidioides brasiliensis 79 , where surface-associated TPI interacts with extracellular matrix components such as laminin and fibronectin to facilitate adhesion and invasion. Although this alternate function for TPI does not address molecular mechanisms implicated in immunomodulation, it does suggest that uncharacterized and perhaps unexpected roles for such enzymes in the infection process.
The cytoplasmic enzyme NDK catalyzes the exchange of a terminal phosphate from a donor NTP to an acceptor NDP to produce a triphosphate, a necessary step for the synthesis of DNA and RNA. Interestingly, NDK also is released into the extracellular milieu by parasites and other microbes and contributes to the inhibition of innate immune responses 80 . It is possible that the T. suis NDK exerts its effect by degrading extracellular ATP (eATP) to reduce purinergic signaling through P2X 7 receptors and inflammasome-mediated responses, preventing release of pro-inflammatory cytokines, as observed with other pathogens 80,81 . Moreover, eATP levels affect macrophage polarization; higher levels of this eATP favor M1 polarization 82,83 .
Despite their striking effects on innate immune cells, how these recombinant proteins or other proteins in TsESP exert immunomodulatory effects remains unclear. It is tempting to postulate that a mechanism involving the cross-linking of surface molecules/receptors that triggers a signaling cascade may lead to the internalization of T. suis proteins. For instance, C-type lectin receptors (CLR), pattern-recognition receptors (PRRs) expressed on DCs and macrophages, are involved in the recognition and internalization of carbohydrates and polypeptides 84 . The Mycobacterium-derived mannose-capped lipoarabinomannans (ManLAMs) bind to the mannose receptor DC-SIGN to trigger an inhibitory signal that blocks LPS-induced activation and suppress dendritic cell functions and IL-12 secretion 85,86 . Helminths express glycosylated molecules, several of which bind to innate receptors and modulate host immune responses 87 . Schistosome soluble egg antigens (SEA), a complex mixture of glycosylated proteins and lipids, are internalized by dendritic cells through different CLR, namely DC-SIGN, MGL, and the mannose receptor, and to co-localize with MHC-II-positive lysosomal compartments 88 . It will also be interesting to determine if similar mechanisms are implicated in the response to TsESP exposure, but also how these molecules block TLR ligand-mediated and cytokine-induced inflammatory responses.
Understanding the nature of the molecular dialogue between the parasite and its host that drives immune modulation could lead to the development of new strategies for combating autoimmune disorders. A focused and targeted approach in the treatment of these debilitating diseases using parasite-derived molecules will circumvent the disadvantages associated with the inoculation of patients with live organisms.

Materials and Methods
Isolation of T. suis life cycle stages. Pig (Sus scrofa) management and handling procedures were approved by the Beltsville Area Animal Care and Use Committee (Protocol #10-011) following Institutional Animal Care and Use Committees (IACUC) guidelines. Mixed-sex pigs (crossbred: Landrace × Yorkshire × Poland China) of approximately 3 months of age were inoculated with a single dose of infective T. suis eggs (2 × 10 4 eggs/pig), as previously described 9 . Swine were housed in stalls with a nonabsorptive concrete floor surface, two pigs per pen, and had access to water and feed ad libitum. The diet was a corn-soybean formulation containing 16% crude protein and vitamins and minerals that exceeded National Research Council guidelines 9 . Following euthanasia, T. suis larval stages and adult worms were collected at various time points, between day 10 to 42 post-inoculation from the cecum and proximal colon. Larvae and adult worms were collected and pooled from 2 to 9 infected pigs (average of 4-5 pigs for each parasite stage, for each infection trials). Larval stages were isolated after pig tissue was washed free of adherent fecal material with warm tap water and then incubated for 3 h in PBS. Adult worms were manually removed from the tissue using fine forceps. 10-and 16-day larvae were freed of debris by floatation over a 40% Percoll; 21-and 28-day larvae were collected by swirling around a curved Pasteur pipette; 42-day adult worms (L5) were picked individually free of debris. Worms were washed five times, for 10 min each, by sedimentation in wash buffer (Hank's Balanced Salt Solution (HBSS) pH 7.2 supplemented with 500 U/mL penicillin, 500 µg/mL streptomycin, 1.25 µg/mL amphotericin B, and 350 µg/mL chloramphenicol) (Gibco, Carlsbad, CA) at a ratio of 10 parts wash buffer to 1 part worms. Pigs were free of other helminth infections as measured by routine screening of the herd and the absence of worms in the intestinal contents and liver lesions at necropsy.

Production of T. suis excretory-secretory protein (TsESP). Parasites were incubated in wash buffer
O/N at 37 °C, 10% CO 2 in air then resuspended in culture medium (DMEM with 4.5 g/mL glucose, L-glutamine, and sodium bicarbonate at pH 7.2, 250 U/mL penicillin, 250 µg/mL streptomycin, 0.625 µg/mL amphotericin B, and 400 µg/mL chloramphenicol) (Sigma-Aldrich, St. Louis, MO) at a concentration of ~20 adult worms/mL or ~500 larvae/mL and incubated at 37 °C, 5% CO 2 . Culture medium was replaced every 24 h for a period up to 72 h, then worms were collected by filtration using a Whatman filter paper. Conditioned culture media were sterilized by filtration through a 0.22 µm Amicon bottle top filter (Millipore, Billerica, MA). TsESP were concentrated under nitrogen pressure (55 psi) using a stirred ultrafiltration chamber (Amicon model 8400, Millipore) with a Millipore Ultracel 3 kDa MWCO ultrafiltration disc. Protein concentrations were measured using the RCDC assay (Bio-Rad, Hercules, CA), according to the manufacturer's specifications, and the samples were aliquoted and stored at −80 °C. Norcross, GA), and the yield and purity assessed by Bioanalyzer (Agilent Technologies, Santa Clara, CA). Whole-genome shotgun fragment and paired-end sequencing libraries (3 kb and 8 kb) were constructed from the gDNA, as described 89 , and sequenced on the Illumina HiSeq. 2000 platform (392x). Linker and adapter sequences were trimmed, and cleaned reads were assembled using ALLPATHS-LG 90 . A repeat library was generated using Repeatmodeler (http://www.repeatmasker.org/RepeatModeler.html), rRNA genes were identified using RNAmmer (http://www.cbs.dtu.dk/cgi-bin/nph-sw_request?rnammer), and tRNAs were identified with tRNAscan-SE 91 . Non-coding RNAs were identified by sequence homology search of the Rfam database (http://selab.janelia.org/software.html). Repeats and predicted RNAs were then masked using RepeatMasker 92 .
Repeat count data are available in Supplementary Table S6. A total of 9,832 protein coding genes as predicted using a combination of the ab initio programs Snap 93 , Fgenesh 94 , and Augustus 95 , and the MAKER annotation pipeline 96 . A consensus gene set based on these predictions was generated using a hierarchical approach developed at The Genome Institute 97 and gene product naming was determined by BER (http://ber.sourceforge.net). Transmembrane domains and classical secretion peptides were predicted using Phobius 98,99 , and non-classical secretion signals were predicted using SecretomeP 100 . Proteins were assigned to KEGG orthologous groups, pathways and pathway modules using KEGGscan 101 with KEGG release 68 102 . Associations with InterPro protein domains and Gene Ontology (GO) classifications were inferred using InterProScan [103][104][105] . All available functional annotations are available in Supplementary Table S3. Functional enrichment of GO terms related to particular subsets of proteins was calculated using FUNC 106 with an adjusted P-value cutoff of 0.01 (Supplementary  Table S4). Orthologous protein families were defined using the OrthoMCL package 107 with an inflation factor 1.5, based on the complete deduced proteomes from the T. suis genome annotation, and of T. muris, T. trichiura 61 , and host species Homo sapiens, Mus musculus, and Sus scrofa (retrieved from Genbank, January 2014) 108 .
Nitric oxide (NO) production and arginase enzymatic activity. Culture supernatants were collected and NO production was determined using the Griess reagent according to manufacturer's specifications (Biotium, Hayward, CA), while cells were kept to test for arginase activity (see below). Absorbance was measured at 548 nm with a Synergy H4 (Biotek) plate reader, and nitrite concentrations were calculated using linear regression obtained from serial dilution of nitrite standards. Arginase activity was measured using the colorimetric reaction between α-ISPP (Sigma-Aldrich) and urea, a technique adapted from previously published work 115 .
Color development was measured at 540 nm using a Synergy H4 (Biotek) plate reader, and arginase enzymatic activity was calculated using values obtained from urea standards and expressed as milli-enzyme units per million cells (mU/10 6 cells).

Intracellular cytokine staining (ICS).
Co-cultures of BMDC and OT-II CD4 + T cells containing CD11c + BMDC pretreated with larvae or adult worm T. suis ESP, or left untreated for 2 h and then 5 × 10 6 purified OT-II CD4 + T cells were added in the presence of 1 nM OVA 323-339 peptide (Anaspec). Cells were cultured at 37 °C, 5% CO 2 for 48 h, and Brefeldin A (eBioscience) was added during the last 4 h of culture to inhibit cytokine secretion. Non-adherent T cells were harvested and adherent BMDC were collected using 1X TrypLE (Invitrogen); cells were processed separately for flow cytometry analysis. Recombinant protein expression. Genes encoding the recombinant T. suis triosephosphate isomerase (TPI, D918-00560), nucleoside diphosphate kinase (NDK, D918-00383), and small nuclear ribonucleoprotein (SNRP, D918-00505) 505 were obtained as a synthetic gene from GenScript cloned into expression vector pET15b vector (Millipore). E. coli ER2566 (New England Biolabs) transformed with the pET15b-TPI, pET15b-NDK, or pET15b-505 vectors were grown at 37 °C and protein expression was induced with 15 mM benzyl alcohol and 0.2 mM IPTG. Bacterial cell pellets were resuspended in PBS (pH 7.2) and cells were lysed by sonication, NaCl was added (0.5 M final) and clarified supernatants were applied to Ni-NTA columns. Columns were washed with PBS, followed by 50 mL of PBS with 2% Triton X-114 to remove endotoxins, then with 25 mL of PBS. Recombinant proteins were eluted with a 80, 160, and 320 mM imidazole step gradient in PBS. Fractions containing a homogeneous preparation were concentrated using an Amicon ultracentrifugal filter unit, and the buffer was exchanged for PBS.

Statistical analyses.
Where applicable, statistical significance was determined using one-way ANOVA followed by Bonferroni post-hoc test; calculations were performed using Prism software (GraphPad Software, La Jolla CA). Differences were considered significant when *P < 0.05, **P < 0.01, ***P < 0.001, while not significant differences are indicated by "ns".

Data Availability
Raw reads from transcriptomic analyses were deposited in the GenBank Sequence Read Archive under BioProject ID PRJNA179528 (SRX1838974 -SRX1838986). Genomic sequencing and gene annotation data are available on the NCBI sequence read archive (Biosample ID SAMN02721381, BioProject ID PRJNA179528). All other datasets generated and analysed during the current study are available upon request.