Plasmacytoid dendritic cells facilitate T helper cell cytokine responses throughout Schistosoma mansoni infection

Abstract

Plasmacytoid dendritic cells (pDCs) are potent producers of Type I IFN (IFN-I) during viral infection and respond to IFN-I in a positive feedback loop that promotes their function. IFN-I shapes DC responses during helminth infection, impacting the ability of DCs to support Th2 responses. However, the role of pDCs in type 2 inflammation is unclear. Previous studies have shown that pDCs are dispensable for hepatic or splenic Th2 responses during the early stages of murine infection with the trematode Schistosoma mansoni, at the onset of parasite egg laying. However, during S. mansoni infection, an ongoing Th2 response against mature parasite eggs is required to protect the liver and intestine from acute damage, and how pDCs participate in immune responses to eggs and adult worms in various tissues beyond acute infection remains unclear. We now show that pDCs are required for optimal Th2 cytokine production in response to S. mansoni eggs in the intestinal-draining mesenteric lymph nodes (MLNs) throughout infection, and for optimal egg-specific IFNγ at later timepoints of infection. Further, pDC depletion at chronic stages of infection led to increased hepatic and splenic pathology, as well as abrogated Th2 cell cytokine production and activation in the liver. In vitro, MLN pDCs supported Th2 cell responses from infection-experienced CD4⁺ T cells, a process dependent on pDC IFN-I responsiveness, yet independent of antigen. Together, these data highlight a previously unappreciated role for pDCs and IFN-I in maintaining and reinforcing type 2 immunity in the LNs and inflamed tissue during helminth infection.

Introduction

Plasmacytoid dendritic cells (pDCs) are known for potent production of Type I IFN (IFN-I), which is especially important for innate activation during viral infection (1, 2). Recent studies demonstrated that pDCs can develop from a lymphoid, as well as myeloid precursor population, and thus the majority of pDCs belong to a separate lineage from conventional DC (cDC) subsets (3, 4). Although pDCs can express MHC Class II and costimulatory molecules (5, 6), they do not stably express MHC II and exogenous antigen on their surface, they lack essential antigen-processing machinery (7, 8), and they express a very restricted suite of pattern-recognition receptors (PRRs) (9), making their relevance as an antigen-presenting cell (APC) population uncertain. Thus, more broadly, the role of pDCs and their function beyond provision of IFN-I in settings outside of viral infection, such as during type 2 inflammation, remains unclear.

pDCs can be found in the T cell zones of peripheral LNs in the steady state, and their migration into this site is enhanced under inflammatory conditions, such as during bacterial or viral infection (10, 11). Additionally, pDCs home to the intestines in specific settings (10, 12) and exert influence on intestinal immune status, for example by orchestrating tolerance against orally delivered antigens in the mesenteric lymph nodes (MLNs) (12, 13). The gastrointestinal tract, along with the liver, is one of the primary sites affected by infection with the medically important helminth Schistosoma mansoni (14), where egg transit through tissues causes significant damage and leads to the development of a Th2 response (14, 15). These observations suggest a potential role for pDCs in modulating immune responses in the tissue during helminth infection. As we have recently shown that cDC populations induce egg-specific Th2 responses in the MLN (16), pDCs are likely not important APCs in this context, and the role of pDCs in supporting Th2 response to eggs and their antigens remains unknown.

pDCs are the primary producers of IFN-I in antiviral responses (7) and pDC activation and function, including IFN-I production, is dependent on an IFN-I autocrine feedback loop via pDC expression of IFNAR1, the IFN-I receptor (17–19), while pDC-derived IFN-I can also drive optimal cDC function during viral infection (20). IFN-I activation in mice also occurs in response to several helminths and their antigens (Ag), including S. mansoni, Heligmosomoides polygyrus and Nippostrongylus brasiliensis (21–24). Further, cDCs require IFN-I signaling via the IFN-I receptor (IFNAR) to initiate type 2 responses against type 2 Ag (21, 22). In response to S. mansoni egg Ag, cDCs, not pDCs, were the primary source of IFN-I (21). However, whether IFN-I influences pDC function during type 2 inflammation, or pDCs are required for Th2 responses during active S. mansoni infection at the egg laying or chronic stages, remains unclear.

In murine models of allergic airway inflammation, rather than acting as APCs in priming type 2 immunity, pDCs can negatively regulate Th2 activation (25, 26). TLR7/8 activation can initiate pDC production of IFN-I that inhibits the function of lung group 2 innate lymphoid cells (ILC2s), reducing airway hypersensitivity (27). In contrast, a recent report suggested that pDC accumulation in the lung during viral infection facilitates type 2 allergic airway exacerbations in an ovalbumin- (OVA) driven model, with antibody-mediated depletion of pDCs reducing type 2 cytokines in the draining LN (28). In humans, along with cDCs, pDC numbers in the blood were increased during infection with the soil-transmitted nematode Strongyloides stercoralis, compared to numbers in endemic healthy and anthelmintic-treated individuals (29). We previously demonstrated that, unlike reports from allergic airway models, negative regulation of type 2 inflammation is not a hallmark pDC function in the early stages of S. mansoni infection (6). Further, pDCs were not essential for promoting Th2 responses in the liver or spleen during the early stages of infection when egg-laying has just recently commenced (6). Despite this, hepatic pDCs isolated ex vivo were able to support OVA-specific TCR transgenic CD4⁺ T cell production of IL-13 in vitro. These data implicated a role for pDCs in promoting development or maintenance of the adaptive immune response against S. mansoni.

In this study, we have addressed the impact of pDC depletion on T cell cytokine production in the intestine-draining MLNs during active S. mansoni infection. We found that pDCs supported optimal Th2 cytokine production at this site at the onset of parasite egg-laying, as well as during the shift to chronic infection. However, pDCs were not fundamentally required for Th2 priming in the skin-draining popliteal LNs (PLNs) following subcutaneous injection of S. mansoni eggs, an in vivo model of acute egg challenge. Nevertheless, a role for pDCs during active S. mansoni infection was also evident in other tissue sites at later stages of infection, with hepatic and splenic pathology exacerbated by increased parasite egg burdens, and the hepatic Th2 response severely abrogated, by pDC depletion later in infection. Together, these data suggest that pDCs have a dynamic role to play at both priming and effector sites affected by egg transit during S. mansoni infection. Importantly, MLN pDCs capably supported type 2 cytokine production by infection-derived MLN CD4⁺ T cells, dependent on pDC expression of IFNAR1. The role of pDCs was not limited to enhancing Th2 responses, as optimal IFNγ production in response to egg challenge and patent S. mansoni infection also depended on the presence of pDCs in the dLNs. Thus, we have identified a previously unappreciated role for pDCs in supporting established CD4⁺ T cell effector responses during chronic helminth infection and demonstrate a novel requirement for pDC IFN-I responsiveness to enable optimal Th2 responses.

Materials and Methods

Mice, infections and immunizations

BDCA-2-DTR, Ifnar1^−/− and KN2 female mice, on the C57BL/6 background, were bred and maintained at the University of Manchester under specific pathogen-free conditions (30–32). Control C57BL/6 (WT) mice were either bred and maintained in-house or obtained commercially from Harlan. Experimental mice were infected percutaneously with ~40–80 cercariae. S. mansoni eggs were isolated from C57BL/6 mouse livers and stored at −80°C before being used to immunize mice s.c. in each rear footpad, with 2,500 eggs in 50 μl PBS. For pDC depletion, mice were injected i.p. every 48h from d32–42 or d42–56 with 4 ng /g diphtheria toxin (Sigma-Aldrich) in PBS. Endotoxin-free soluble egg Ag (SEA) from S. mansoni was prepared in-house as previously described (33). All experiments were conducted under a license granted by the Home Office (UK), in accordance with local guidelines.

Provision of parasite material

The life cycle of the NMRI (Puerto Rican) strain of S. mansoni was maintained by routine infections (~180 cercariae) of female TO mice obtained from Harlan and Biomphalaria glabrata (NMRI albino and pigmented hybrid) snails (12 miracidia/snail) at Aberystwyth University. All procedures performed on mice adhered to the United Kingdom Home Office Animals (Scientific Procedures) Act of 1986 (project license PPL 40/3700 and P3B8C46FD) as well as the European Union Animals Directive 2010/63/EU and were approved by Aberystwyth University’s Animal Welfare and Ethical Review Body (AWERB).

Flow cytometry

Following cell isolation, cell populations were assessed by flow cytometry. Cells were first stained with LIVE/DEAD Fixable Aqua or UV (ThermoFisher) and, following FcR-Block, were stained with combinations of the following mAb: CD4, CD11b, CD11c, CD40, CD44, CD45R, CD64, CD80, CD86, CD205, CD317, Lineage (CD3, CD19, CD49b, Ly6G, NK1.1), F4/80, Foxp3, huCD2, Ly6C, MHC II, TCRβ and XCR1. All antibodies for flow cytometry were purchased from BD, ThermoFisher or Biolegend. Samples were acquired on a FACS LSR or LSR Fortessa flow cytometer using BD FACS Diva Software and analyzed with FlowJo 9 software (Tree Star, Inc.) to ascertain proportional expression of markers. When gating on DC populations, cells were gated on live, singlet, lineage⁻ cells, and monocytes (Ly6C^hiCD11b^hi) and macrophages (F4/80⁺CD64⁺) were first excluded from analysis. pDCs were gated as Ly6C^midCD11b⁻CD11c^midCD317⁺CD45R⁺, cDC1s as CD11c^hiMHC II^hiCD11b^loCD205^hi/XCR1^hi, and cDC2s as CD11c^hiMHCII^hiCD11b^hiXCR1^lo/CD205^lo. Tregs were gated as TCRβ⁺NK1.1⁻CD4⁺Foxp3⁺, and Teff as TCRβ⁺NK1.1⁻CD4⁺Foxp3⁻CD44^hiCD62L^lo.

S. mansoni egg counts

Livers and intestines from infected mice were digested in 4% potassium hydroxide (15 ml/gram of liver tissue; 7.5 ml/gram of intestine tissue) at 37°C overnight. 100 μl aliquots of the digests were evaluated on gridded petri dishes and the eggs counted at 10x magnification. Each digest was examined in triplicate and the mean results were used to extrapolate the total number of eggs /gram of tissue.

Restimulation assays

Livers were perfused, diced, and digested at 37°C for 45 min using 0.4 U/ml Liberase (Roche) and 80 U/ml DNase I Type IV (Sigma-Aldrich). The digested liver was then passed through a 100 μm cell strainer with the aid of a syringe plunger. Leukocytes were separated from other liver cells by resuspension in 33% isotonic Percoll (GE Healthcare) and centrifugation at 700 x g. Pelleted cells were resuspended and passed through a 40 μm cell strainer to obtain a single-cell suspension and remove S. mansoni eggs. RBCs were ACK lysed, and cells were counted and resuspended for use. Single cell suspensions of PLN, MLN or liver leukocytes (1 × 10⁶ cells/ml) were cultured in X-vivo 15 medium (Lonza) containing 2 mM L-Glutamine and 50 μm 2-ME (ThermoFisher) in 96 well plates at 37°C 5% CO₂ with, or without, 15 μg/ml SEA. After 72h, supernatants were harvested and analyzed for IL-4, IL-5, IL-10, IL-13 and IFNγ using paired mAb and recombinant cytokine standards (Biolegend and Peprotech). Medium alone values were subtracted from Ag-restimulated cytokine levels for each sample.

Co-culture assays

pDCs and T cells were FACS-sorted from the MLNs of S. mansoni-infected animals. This assay was set up as previously described (21, 34). 50,000 CD4⁺ T cells were cultured in 96 well plates for 3d with 2,500 pDCs, with, or without, 1 μg/ml anti-CD3 (produced in-house). On d3, either supernatants were taken for cytokine analysis by ELISA or cells were analyzed for huCD2 expression (IL-4 production) by flow cytometry.

Statistical analysis

Statistical analysis was carried out using JMP software (SAS), analyzed using linear mixed effects models with a fixed effect of experimental group and a random effect of experiment day, to account for any significant differences between experiments. Model assumptions of normality and homogeneous variance were assessed by a visual analysis of the raw data and the model residuals. Right-skewed data was log or square root transformed. Experimental groups were considered statistically significant if the fixed effect F test p-value was less than or equal to 0.05. Post-hoc pairwise comparisons between experimental groups were made using Tukey’s HSD multiple-comparison test. Statistical outliers were identified using the extreme studentized deviate method and omitted prior to the mixed effect model analysis. Graphs of results were shown as mean±SEM of untransformed data using Prism version 6 (GraphPad).

Results

S. mansoni infection alters the proportion, number and activation status of MLN DC subsets

During S. mansoni infection, a high proportion of eggs released by adult female worms are forced into the liver by blood flow in the portal vein. However, in order to perpetuate the helminth lifecycle, some eggs must rupture through the intestinal wall to exit the host (14, 15). Egg transit causes significant pathology in the intestine, as well as the liver, over the course of infection. We have previously shown that egg injection into the intestinal serosa elicits Ag-bearing cDCs to migrate to the MLNs where they promote Th2 polarization (16). However, little is known about which DC subsets are evident in the MLNs over the course of active S. mansoni infection.

To address this question, we harvested MLN cells at different timepoints throughout S. mansoni infection (Fig. 1a). Assessment began at day (d) 28 of infection, around the time that adult female worms start to deposit eggs that transit through the tissues (14). The window between d28–42 of infection also induces ‘acute’ Th2 priming, predominantly in response to egg Ag. By later, ‘chronic’, timepoints, from around d56 of infection, mice develop extensive Th2-driven immunopathology and display a more regulatory immunophenotype (14). Comparing d28, d42 and d56 post-infection, we found that the proportion of pDCs (CD317⁺Ly6C⁺CD11c⁺, Fig. 1b) within the MLNs was significantly increased in mice at both d28 and d56 of infection compared to naïve mice, and that the total numbers of pDCs were increased compared to naïve at d42–56 of infection (Fig. 1c). Frequencies of cDC subsets, defined as cDC1s or cDC2s according to their differential expression of surface markers CD11b and CD205/XCR1 (Fig. 1b) (35) did not change throughout infection as a proportion of all live cells. Similarly, the total number of cDC1s was significantly increased at d42–56 of infection compared to naïve animals, while cDC2s transiently increased in number at d42 of S. mansoni infection (Fig. 1c). At every timepoint, pDCs remained a smaller percentage of total MLN cells, and a smaller number of cells, than cDC1s or cDC2s, as previously shown in the liver during S. mansoni infection (6).

Figure 1. The kinetics and activation of DC populations over the course of S. mansoni infection.

(a) C57BL/6 WT mice were infected with ~40–80 S. mansoni cercariae on day (d)0, and mesenteric LNs (MLNs) harvested from naïve (N) mice, and at d28, d42, d56 of infection. (b) Representative plots of MLN DC gating. (c) Percentage of total live cells and total numbers of pDCs, cDC1s and cDC2s in naïve (N) or S. mansoni infected (Sm) MLNs. (d) Fold change in geometric mean fluorescence intensity (gMFI) of MHC II, CD40, CD80 and CD86 on the surface of DC subsets over the course of infection, normalized to expression levels on cells from control mice. Results are mean±SEM. Data from 2 experiments pooled (n=4–13 mice/ group) and analyzed using a linear fixed effect model with pairwise comparison, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Although cDC expression of MHC II and co-stimulatory molecules is essential for CD4⁺ T cell priming (36), minimal upregulation of these surface markers has been reported against helminths, including in splenic cDCs and pDCs during S. mansoni infection (37). In agreement with this, we observed no significant increase in MHC II levels (geometric mean fluorescence intensity, gMFI) on all DC subsets in the MLN of infected compared to naïve mice (Fig. 1d), with MHC II levels in fact decreasing in mice between d42 and d56 of infection compared to naïve mice. MLN pDCs significantly upregulated CD40 expression on their surface at d28 of infection compared to cells from naïve mice, whilst cDC subsets only upregulated CD40 compared to controls around the development of chronic infection at d56. Thus, the MLN pDC population was responsive to S. mansoni infection during the early stages of the immune response, before egg production, highlighting the potential for a previously unappreciated role for pDCs at this site following schistosome infection.

pDCs are required for optimal T helper cytokine responses in the gut-draining LN during S. mansoni infection

pDC depletion early in S. mansoni infection (d32–42 post-infection) had no significant effect on Th2 development in the liver or spleen (6), but the impact of pDCs on the Th2 response at other sites affected by egg transit or at other timepoints post-S. mansoni infection has remained unclear. To assess the influence of pDCs on T helper cell development in the gut-draining MLNs during S. mansoni infection, we used BDCA-2-DTR pDC-depletable mice, which express the human diphtheria toxin receptor (DTR) gene under control of the promoter for BDCA-2, a marker of human pDCs, for effective and specific pDC depletion (30). We first depleted pDCs between d32-d42 of S. mansoni infection (Fig. 2a), a timeframe that correlates with the onset of egg deposition and the commencement of Th2 priming, and our previous assessment of the impact of pDCs and CD11c⁺ cell depletion on hepatic and splenic Th2 responses during infection (6, 38). The proportion of pDCs from total live cells was variable in the MLNs of WT and DTx-treated animals (Fig. 2b), thus calculated pDC depletion at this site was roughly 50% in naïve mice, and 35–40% depletion in S. mansoni-infected animals (Fig. 2c). Despite incomplete depletion of pDCs in the MLNs, levels of parasite-specific IL-4 and IL-13 produced by MLN cells, as measured by ELISA of cell-free supernatants, were significantly impaired in infected animals following pDC depletion on d32–42 post-infection (Fig. 2d), suggesting a potential site-specific role for pDCs in promotion of the Th2 cell response. To investigate the role of pDCs later in infection, when the Th2 response is in a more chronic phase (14), we next depleted pDC populations from d42 to d56 of infection (Fig. 2e), 2–4 weeks after initial onset of egg deposition. This led to more robust pDC depletion in the MLNs at this timepoint (Fig. 2f–g) and defective S. mansoni egg Ag-specific recall responses in the MLNs, including a significant reduction in IL-5, IL-13 and IFNγ production (Fig. 2h). While Ag-specific IL-4 from MLN cells was reduced following pDC depletion earlier in infection, at this timepoint egg-specific IL-4 was not significantly impacted.

Figure 2. pDCs are required for optimal T cell cytokine production in the gut-draining lymph node throughout S. mansoni infection.

WT or BDCA-2-DTR (DTR) mice were infected with ~40–80 S. mansoni (Sm) cercariae on d0 (naïve (N)). For assessment of pDCs in acute infection from d32 until d42 (a), or for assessment at the onset of chronic infection, from d42 until d56 (e), mice were treated i.p. with DTx before MLN harvest. (b, f) Proportion and total numbers of pDCs from MLN cells on d42 (b) or d56 (f) of infection. (c, g) pDC depletion in DTR mice calculated compared to pDC percentages in WT mice on d42 (c) or d56 (g) of infection. (d, h) After SEA restimulation, or culture with medium alone, MLN cell-free supernatants were analyzed by ELISA for a schistosome egg-specific recall cytokine response. Medium alone values were subtracted from Ag-restimulated samples. Results are mean±SEM. Data from 3 experiments pooled (n=4–12 mice/ group) analyzed using a linear fixed effect model with pairwise comparison. *P<0.05, **P<0.01, ***P<0.001, P****<0.0001.

Parasitemia was unaltered after pDC depletion by DTx treatment at the onset of egg-laying, with WT and BDCA-2-DTR mice displaying similar egg burdens in the liver and gut (Fig. 3a), and similar or diminished immunopathology at d42 of S. mansoni infection (Fig. 3b). In contrast, there was a significant increase in liver egg counts and a trend for increased intestinal egg deposition following depletion at the later timepoint (Fig. 3c), accompanied by significantly heightened splenomegaly and hepatomegaly (Fig. 3d). Thus, defective Th2 cytokine levels in the absence of pDCs could impact the ability of the host to tolerate infection as it reaches chronicity.

Figure 3. pDCs are required to restrain parasite egg-laying and host pathology during chronic S. mansoni infection.

Naïve (N) and S. mansoni (Sm)-infected WT or DTR mice were treated with DTx, as in Fig. 1. On d42 (a) or d56 (c) of Sm infection, livers and small intestines were isolated from infected animals, and parasite egg counts performed. (b, d) Splenomegaly and hepatomegaly determined by spleen and liver weights as a percentage of total body weight on d42 (b) or d56 (d) of Sm infection. Results are mean±SEM. Data from 2 (c-d) or 4 (a-b) experiments pooled (n=7–20 mice/ group) analyzed using a linear fixed effect model with pairwise comparison. *P<0.05, **P<0.01, P****<0.0001.

We addressed the possibility that other DC populations could be altered in response to pDC depletion and found that DTx treatment of WT or BDCA-2-DTR animals had no impact on monocyte, macrophage (6), or cDC subset frequencies in d42 S. mansoni-infected mice (Supplementary Fig. 1a), although there was a minor decrease in cDC1s in the MLNs of some naïve DTx-treated BDCA-2-DTR mice. At d56 of infection there was a significant increase in cDC2s in the MLNs and significant decrease in liver cDC1s in both WT and BDCA-2-DTR mice compared to naive controls (Supplementary Fig. 1b). Thus, non-pDC DC subset alteration does not likely influence the deficit in Th2 cytokine responses in the MLN following pDC depletion at day 32–42 or 42–56 post-infection. Together, these data demonstrated that pDCs were required for optimal CD4⁺ T cell Th2 development in the gut-draining MLNs over the course of active S. mansoni infection, from early after commencement of egg laying. Further, pDCs were required for IFNγ production at later stages of infection.

pDCs are not required for Th2 priming against S. mansoni eggs

CD11c⁺ myeloid cells are required for optimal Th2 induction following murine S. mansoni infection or egg injection (38), with pDCs specifically promoting CD4⁺ T cell cytokine responses to eggs antigens in the MLNs during infection (Fig. 2). To directly address whether pDCs are required for Th2 priming against schistosome eggs, we injected PBS- or DTx-treated WT or BDCA-2-DTR mice subcutaneously with PBS or S. mansoni eggs (Supplementary Fig. 2a), a system widely used to investigate acute type 2 response induction to eggs in the dLNs without the additional complexities involved in ongoing infection (34, 38). DTx-treatment had no significant impact on Th2 development or egg-specific IFNγ production following subcutaneous S. mansoni egg injection (Supplementary Fig. 2b), even though PLN pDCs were efficiently depleted in egg-injected mice (Supplementary Fig. 2c–e), with no impact on cDC populations (Supplementary Fig. 2f–g). Thus, pDCs were not required for T helper cell polarization in a model of acute egg challenge, but rather were a key contributing cell type for optimal Th2 and Th1 cytokine responses during patent S. mansoni infection (Fig. 2).

pDCs sustain MLN CD4⁺ T cell responses during chronic S. mansoni infection

To better understand the function of pDCs in T helper responses during S. mansoni infection, we began by analyzing the influence of pDC depletion on CD4⁺ T cell subpopulations in this context. There was no significant impact of pDC depletion on proportions or total numbers of CD44⁺CD62L^loCD4⁺ T effector (Teff) cells or Foxp3⁺ T regulatory (Treg) in the MLNs of infected mice at d42 after pDC depletion (Fig. 4a–b). At d56 of infection, there was a comparable increase in the proportion of Teff but not Tregs in both WT and BDCA-2-DTR mice compared to naïve controls (Fig. 4c). However, when total numbers of CD4⁺ T cell populations were assessed at this timepoint, there was a significant impairment in expansion of both Teff and Tregs in the MLNs of S. mansoni-infected BDCA-2-DTR mice compared to WT controls (Fig. 4d). These data demonstrate that pDCs have a key role to play in the effective development or maintenance of both effector and regulatory CD4⁺ T cell populations as S. mansoni infection progresses from d42 to d56.

Figure 4. pDCs from S. mansoni infected mice support optimal CD4⁺ T cell responses in vivo and Th2 polarization in vitro.

MLNs were harvested from naïve (N) and S. mansoni (Sm)-infected WT and DTR mice on d42 or d56 of infection, after pDC depletion from d32–42 or d42–d56, respectively, and analyzed for CD4⁺ T cell populations. (a, c) Percentage of CD4⁺ T cells on d42 (a) or d56 (c) of S. mansoni infection that are CD44⁺ Teff and Foxp3⁺ Tregs in the MLN following DTx treatment. (b, d) Total numbers of T cell populations in the MLN at d42 (b) or d56 (d) of infection following pDC depletion. (e) Schematic to show co-culture regimen. T cells from d42 S. mansoni (Sm)-infected mice were cultured for 3d with, or without, MLN pDCs from naïve (N) or d42 Sm-infected mice, in the presence, or absence, of anti-CD3 (aCD3). (f) On d3 of culture, supernatants were collected from cultures and assessed for cytokine production by ELISA. (g) KN2 MLN T cells from Sm mice were cultured in the presence or absence of WT or Ifnar1^−/− MLN Sm pDCs with aCD3. On d3, CD4⁺ T cells were stained for huCD2 to identify the proportion of huCD2⁺ (IL-4⁺) T cells. Results are mean±SEM. Data from 2–3 experiments pooled (n=6–7 mice/ group (a–d), n=8–14 wells/ group (f–g)) analyzed using a linear fixed effect model with pairwise comparison. *P<0.05, **P<0.05, ***P<0.001, P****<0.0001.

pDCs can support Th2 cytokine production by MLN T cells isolated from S. mansoni infection

Having demonstrated that pDCs were essential both for optimal Th2 cytokine and IFNγ output by MLN cells (Fig. 2), we next addressed whether MLN pDCs from active S. mansoni infection could directly enhance cytokine production by infection-derived MLN CD4⁺ T cells. To do so, pDCs sorted from the MLNs of naïve mice or mice at d42 of S. mansoni infection were cultured with MLN CD4⁺ T cells also isolated from d42 infected mice (Fig. 4e). MLN pDCs from either naïve or infected mice significantly enhanced levels of Th2 cytokine production by CD4⁺ T cells isolated from infected animals, compared to that observed from T cells cultured alone (Fig. 4f). These data demonstrate that pDCs, while incapable of priming Th2 polarization of naïve T cells (6), could competently support Th2 cytokine production from infection-derived, antigen-experienced, CD4⁺ T cells. Of note, in this context, pDCs failed to significantly enhance CD4⁺ T cell IFNγ production.

IFN-I signaling via IFNAR can promote optimal DC activation and function during inflammation (17–20), including during type 2 immune activation driven by helminth Ag, where cDC induction of Th2 responses is dependent on their ability to respond to IFN-I (21, 22). Further, IFN-I is produced in vivo in response to helminths, including following immunization with S. mansoni egg Ag and during active S. mansoni infection (21, 24). To investigate whether pDC promotion of type 2 cytokine production from CD4⁺ T cells was also facilitated by IFN-I responsiveness, we isolated pDCs from the MLNs of WT or Ifnar1^−/− S. mansoni-infected mice and co-cultured them with CD4⁺ T cells isolated from infected WT KN2 cytokine reporter mice, which express huCD2 on their surface when IL-4 is produced (21, 21, 32). Optimal pDC-mediated support of type 2 cytokine production from infection-derived CD4⁺ T cells was dependent on the ability of pDCs to respond to IFN-I, as the percentage of huCD2⁺ CD4⁺ T cells was significantly reduced when pDCs were Ifnar1-deficient, in comparison to WT cells (Fig. 4g). Thus, the ability of pDCs to facilitate type 2 cytokine production by MLN T cells from infection was promoted by pDC responsiveness to IFN-I.

pDCs support hepatic Th2 cytokine responses as S. mansoni infection progresses

We hypothesized that the impaired T cell cytokine response evident in the MLNs of pDC-depleted mice during S. mansoni infection might have a subsequent impact on the immune function in the effector sites, such as the liver, at later stages of infection. This hypothesis was supported by increased egg burden and pathology after depletion later in infection (Fig. 3c–d). DTx treatment between d42–56 of infection (Fig. 5a) led to effective depletion of hepatic pDCs in BDCA-2-DTR mice (Fig. 5b–c) but did not significantly reduce hepatic cDC1s or cDC2s during infection (Supplementary Fig. 1b) (39). When we depleted pDCs at this stage of infection (from d42), egg Ag-induced IL-4 levels in the supernatants of cultured liver cells were significantly reduced at d56, with a trend for decreased IL-13 levels compared to WT controls (Fig. 5d). IL-5 and IL-10 were not significantly affected at this site or timepoint, and pDCs were not required for maintaining Ag-specific IFNγ in the liver (Fig. 5d), in contrast to the MLN phenotype at this timepoint (Fig. 2). The decrease in parasite-specific IL-4 and IL-13 from liver cells following pDC depletion was consistent with increased splenomegaly and hepatomegaly (Fig. 3c–d), the defects in MLN Th2 cytokines (Fig. 2h) and CD4⁺ T cell responses (Fig. 4c–d) evident at this timepoint. Despite a minor increase in the percentage of Tregs in pDC-depleted naïve mice compared to controls, the proportion of hepatic Teff and Tregs at d56 of S. mansoni infection was unaffected by pDC depletion (Fig. 5e). However, the total numbers of Tregs and Teff were significantly diminished in infected pDC-depleted mice compared to WT animals during this phase of infection (Fig. 5f).

Figure. 5. pDCs are essential for effective Th2 cytokine production in the liver in chronic S. mansoni infection.

(a) Livers were harvested from DTx-treated naïve (N) and S. mansoni (Sm)-infected WT or DTR mice on d56 of infection, as in Fig. 1c. (b) Liver cells were purified and analyzed by flow cytometry for the presence of pDCs, shown as a percentage of all live leukocytes in the liver and as total numbers. (c) pDC depletion in N and Sm DTR mice calculated by comparing pDC percentages in WT and DTR mice. (d) Supernatants from liver cells of N and Sm WT/DTR mice were cultured with SEA or medium (M) alone and analyzed by ELISA for schistosome egg-specific recall cytokine responses. M alone values were subtracted from Ag-restimulated samples. (e-f) Percentage of liver CD4⁺ T cells (e) or total cell numbers (f) that are CD44^hi Teff and Foxp3⁺ Tregs from N/ Sm mice on d56 of infection. Results are mean±SEM. Data from 2 experiments pooled (n=6–10 mice/ group), analyzed using a linear fixed effect model with pairwise comparison. *P<0.05, **P<0.01, ***P<0.001, P****<0.0001.

These data demonstrate that, in addition to playing a key role in supporting Teff responses in the MLN early during S. mansoni infection, pDCs enhance type 2 cytokine production and CD4⁺ T cell responses at the liver effector site at later stages of infection, including effective Th2 function and Treg generation. They highlight for the first time a key role for pDCs in supporting optimal effector CD4⁺ T cell responses to parasite egg Ag during S. mansoni infection in both priming and effector sites in the MLN and liver, respectively. In the absence of pDCs, abrogated CD4⁺ T cell-mediated adaptive immunity leads to increased immunopathology, with the potential to impact the long-term fitness and survival of the host.

Discussion

We have identified that pDCs are essential for optimal CD4⁺ T cell responses during chronic S. mansoni infection. An LN-specific impairment in Th2 cytokines during early infection was associated with defective accumulation of CD44⁺ activated T effector T cells and Tregs (Fig.s 2 and 4) – the CD4⁺ T cell subsets involved in orchestrating, and regulating, the Th2 response in the tissue as infection progresses (40). This then also led to a defect in Th2 cytokine output in the effector site, the liver, at the onset of chronic infection (Fig. 5). This defect in CD4⁺ T cell-driven type 2 immunity in the absence of pDCs was associated with and likely caused increased parasite egg deposition and exacerbated infection-driven pathology, resulting in worsened splenomegaly and hepatomegaly (Fig. 3). All together, these findings highlight a role for pDCs in supporting optimal CD4⁺ T cell function during ongoing S. mansoni infection.

Our data suggest that there are site-specific differences in the timing of a requirement for pDCs to promote optimal immune activation against S. mansoni, comparing the liver and MLNs. This dichotomy may relate to differences in pDC localization, as well as tissue signals and/or architecture in the MLNs and liver during S. mansoni infection. During viral infection, pDC recruitment by CCL3/4 chemokines to T cell-cDC1 clusters support T cell priming in the LN paracortex, and formation of these clusters is essential for optimal T cell function (20). Our in vitro coculture assays suggest that, when pDCs and CD4⁺ T cells are in close proximity, pDCs from naïve as well as S. mansoni-infected mice can enhance type 2 cytokine production (Fig. 4). Thus, as infection progresses, increased immune cell infiltration, changes in liver architecture (Fig. 3) and chemokine production (14) may lead to an increased role for pDCs in promoting CD4⁺ T cell function at this site. Intriguingly, S. mansoni eggs produce a chemokine-binding protein that blocks chemokines, including CCL3, from binding their receptors (41). Additionally, high CCL3 levels are associated with increased disease severity in S. mansoni-infected mice and chronic infection in humans (42), supporting a role for these signals in regulating immune function during S. mansoni infection that has yet to be fully explored. Further, pDCs infiltrate into the skin following tissue injury, where pDC IFN-I production promotes wound healing, a process promoted by the release of host-derived nucleic acids from damaged tissue (43). Thus, increased tissue damage in the liver and release of host damage signals and immune signals at later stages of infection may promote pDC function at this site, as chronic S. mansoni infection develops.

Importantly, pDC depletion was not always complete in the MLNs, with naïve and d42 infected animals showing depletion levels below 50% in some instances, compared to the average ~80% depletion we observed in other contexts (Fig. 2 and Supplementary Fig. 2). This may reflect the fact that pDCs are a very small population in this tissue under these conditions, so are difficult to quantify accurately. Despite this, incomplete pDC depletion still had a significant impact on Th2 cytokine production from MLN cells at early timepoints of infection (Fig. 2d), as well as later in infection, when pDC depletion was more consistent (Fig. 2h). pDC depletion earlier in infection led to decreased egg-specific IL-4 and IL-13 in the MLN, while depletion at d42–56 caused a significant defect in MLN IL-5, IL-13 and IFNγ, but not IL-4 (Fig. 2). This discrepancy may reflect dynamic changes in the pDC and/or CD4⁺ T cell populations in response to chronic levels of parasite egg deposition, evidenced by differences in absolute cytokine levels at different timepoints of infection. In this work, we have primarily focused on the role of pDCs in activating T cell production of type 2 cytokines, but our data also strongly support a role for pDCs in the generation of parasite-specific IFNγ. pDCs act in a supporting role to cDC1s in IFNγ induction in other settings (20), and thus they may co-operate with cDC1s to promote Th1 cytokine production during S. mansoni infection (44).

We have not assessed here the impact of defective Th2 polarization in the MLNs on the effector response in the intestinal lamina propria, due to the technical difficulty of isolating immune cells from S. mansoni-infected intestinal tissue. However, we hypothesize a similar impact on Th2 cytokine responses at this site, given that pDCs are found abundantly in the small intestine in naïve mice (12) and can acquire gut tropism under inflammatory conditions (10, 12, 45). Thus, future studies should address whether pDCs also play a role in intestinal Th2 function during S. mansoni infection.

Th2 polarization against acute egg challenge did not fundamentally require pDCs, at least within the first week following exposure to S. mansoni eggs (Supplementary Fig. 2). Thus, it may take several weeks of continuous egg deposition during infection for pDCs to become important during Th2 maintenance in the dLNs, with this only becoming evident in effector sites at later timepoints. In S. mansoni egg-driven pulmonary fibrosis, macrophages are important for maintaining T cell responses in the tissue, but not the dLNs (46), while ILC2s, basophils and cDCs have been implicated to play a role in promoting the maintenance of Th2 cells and Tregs at sites of inflammation and at mucosal barriers (47–50). This highlights a potential role for other innate cell populations in promoting optimal Th2 cytokine production in the liver at early timepoints. Many pDC studies have failed to identify a role for these cells in direct CD4⁺ T cell priming (7). Our co-culture system demonstrated that pDCs are effective at facilitating Th2 cytokine responses in a population of Ag-experienced, or preactivated, CD4⁺ T cells without exogenous antigen (Fig. 4) (6), suggesting that Ag presentation by pDCs is less important. Of note, recent studies highlight that the majority of pDC precursors are lymphoid, rather than myeloid (3, 4), with the suggestion that pDCs may be redesignated as an innate lymphoid cell population (4). If this is indeed the case, pDCs could be compared to group 2 innate lymphoid cells (ILC2s) in their capacity to support Th2 function, rather than myeloid DCs. How pDCs promote optimal CD4⁺ T cell function during S. mansoni infection remains to be determined. Our coculture experiments suggest that any direct pDC-T cell interaction may not depend on Ag (Fig. 4), indeed pDCs from naïve and S. mansoni-infected mice failed to support Ag-specific T cell proliferation in vitro (6). Thus, pDCs may act in other ways to potentiate Th cytokine production and CD4⁺ T cell subset expansion, for example, by driving T cell survival or cytokine production.

Prior work addressing pDC function in type 2 immunity has focused on the lung, particularly in the context of allergic airway models. Early studies using antibody depletion of pDCs in OVA-driven airway allergic inflammation suggested an anti-inflammatory role for these cells (25, 26). In agreement with this, we show that pDCs are required for Treg recruitment and/or expansion in the MLNs and liver during chronic S. mansoni infection (Fig. 4 and 5). However, a coincident reduction in effector T cell populations (Fig. 4 and 5) and Th2 cytokine production (Fig. 2 and 5) in the MLN and liver is inconsistent with regulation being the only role for pDCs in type 2 immunity. In agreement with our findings, a more recent study highlighted an influx of pDCs into the lung following OVA-allergen challenge, with antibody-depletion of pDCs leading to reduced Th2 cytokines and type 2-mediated inflammation (27). This study highlighted a role for OX40L and IL-25 in regulating pDC function in this context. However, IL-25-deficient mice do not display any significant exacerbation in liver granuloma responses during chronic S. mansoni infection (51). Thus, further work is needed to address the importance of different type 2-associated signals for mediating pDC function in S. mansoni infection.

We have highlighted a previously unappreciated role for IFN-I signaling in directing pDC function in type 2 inflammation (Fig. 4). Our work emphasizes that DC subsets depend on IFN-I, not only for type 2 priming by cDCs (21, 22), but also for optimal Th2 cytokine responses supported by pDCs. We and others have previously shown that several facets of cDC function are altered in the absence of IFN-I responsiveness, including surface activation and migratory capacity (21, 22). In contrast, studies investigating the role of pDCs in ILC2-mediated allergic lung inflammation in response to Alternaria fungus with IFN-I-inducing TLR ligands suggested that pDC IFN-I inhibits ILC2-mediated type 2 immunity (28). This study is perhaps more applicable to situations where tissues are responding to type 1 pathogens that drive high level IFN-I production, alongside a type 2-promoting pathogen or allergen, for example during virus and helminth coinfection (52). Thus, further work is needed to address the signals that promote pDC potentiation of Th2 cytokine output, including the importance of parasite or host-derived Ag and the broader role of IFN-I in type 2 immunity. In particular, it remains to be determined whether pDCs (17–20) or other cell types are the primary source of IFN-I during helminth infection, and whether pDC-derived IFN-I plays a role in promoting CD4⁺ T cell function. Other potential cellular sources include cDCs – shown to produce IFNα6 in response to S. mansoni egg Ag (21), or damaged stromal and epithelial cells that release innate IFNs in response to damage (53, 54).

In summary, we have revealed that pDCs play an integral role in optimal Th2 cytokine responses and Treg accumulation in both LNs and effector sites during S. mansoni infection (Fig.s 2, 4 and 5). Impairment in immune activation following pDC depletion was accompanied by a significant increase in splenic and hepatic pathology during chronic infection, likely due to a dramatic increase in egg deposition in the liver and gut (Fig. 3). Further work will be needed to identify the mechanisms employed by pDCs to influence Th2 and Treg responses, such as whether pDCs are required for the proliferation, recruitment, maintenance, or survival of T cells during S. mansoni infection. Our work highlights a novel role for pDCs and IFN-I during S. mansoni infection, which may have implications for the future development of targeted immunotherapies and vaccines designed to combat the debilitating disease caused by this important human pathogen.