The Foxp3+ regulatory T-cell population requires IL-4Rα signaling to control inflammation during helminth infections

Forkhead box P3 (Foxp3+) regulatory T (Treg)-cell function is controlled by environmental cues of which cytokine-mediated signaling is a dominant component. In vivo, interleukin-4 (IL-4)-mediated signaling via IL-4 receptor alpha (IL-4Rα) mediates Treg cell transdifferentiation into ex-Foxp3 T helper 2 (Th2) or T helper 17 (Th17) cells. However, IL-4-mediated signaling also reinforces the Foxp3 Treg compartment in vitro. We generated Foxp3-specific IL-4Rα-deficient mice and demonstrated differential efficiency of IL-4Rα deletion in male (approximately 90%) and female (approximately 40%) animals, because of cyclic recombinase (Cre)-mediated X-linked foxp3 inactivation. Irrespective of the degree of IL-4Rα deletion within the Foxp3+ Treg cell population, mice showed exacerbation of immune effector responses with aggravated tissue pathology in tissue-dwelling helminth infections (Schistosoma mansoni or Nippostrongylus brasiliensis). Mechanistically, IL-4Rα deletion in males and females led to a reduced expression of Foxp3 and subsequently an impaired accumulation of Foxp3+ Treg cells to inflamed tissues. In-depth cellular typing by flow cytometry revealed that the impairment of IL-4Rα-mediated signaling during helminth infections decreased the ability of central Treg cells to convert into effector Treg (eTreg) cells and caused a significant down-regulation of markers associated with Treg cell migration (C-X-C motif chemokine receptor 3 [CXCR3]) and accumulation in inflamed tissues (GATA binding protein 3 [GATA3]) as well as survival (B cell lymphoma 2 [Bcl-2]). These findings unprecedentedly, to our knowledge, uncover a role for IL-4Rα signaling in the positive regulation of Foxp3+ Treg cell function in vivo. Complementing our past knowledge on a widely reported role for IL-4Rα signaling in the negative regulation and transdifferentiation of Foxp3+ Treg cells in vivo, our present findings reveal the host requirement for an intact, but not reduced or potentiated, IL-4Rα-mediated signaling on Foxp3+ Treg cells to optimally control inflammation during helminth infections.

Forkhead box P3 (Foxp3 + ) regulatory T (Treg)-cell function is controlled by environmental cues of which cytokine-mediated signaling is a dominant component. In vivo, interleukin-4 (IL-4)-mediated signaling via IL-4 receptor alpha (IL-4Rα) mediates Treg cell transdifferentiation into ex-Foxp3 T helper 2 (Th2) or T helper 17 (Th17) cells. However, IL-4-mediated signaling also reinforces the Foxp3 Treg compartment in vitro. We generated Foxp3-specific IL-4Rα-deficient mice and demonstrated differential efficiency of IL-4Rα deletion in male (approximately 90%) and female (approximately 40%) animals, because of cyclic recombinase (Cre)-mediated X-linked foxp3 inactivation. Irrespective of the degree of IL-4Rα deletion within the Foxp3 + Treg cell population, mice showed exacerbation of immune effector responses with aggravated tissue pathology in tissue-dwelling helminth infections (Schistosoma mansoni or Nippostrongylus brasiliensis). Mechanistically, IL-4Rα deletion in males and females led to a reduced expression of Foxp3 and subsequently an impaired accumulation of Foxp3 + Treg cells to inflamed tissues. In-depth cellular typing by flow cytometry revealed that the impairment of IL-4Rα-mediated signaling during helminth infections decreased the ability of central Treg cells to convert into effector Treg (eTreg) cells and caused a significant down-regulation of markers associated with Treg cell migration (C-X-C motif chemokine receptor 3 [CXCR3]) and accumulation in inflamed tissues (GATA binding protein 3 [GATA3]) as well as survival (B cell lymphoma 2 ). These findings unprecedentedly, to our knowledge, uncover a role for IL-4Rα signaling in the positive regulation of Foxp3 + Treg cell function in vivo. Complementing our past knowledge on a widely reported role for IL-4Rα signaling in the negative regulation and transdifferentiation of Foxp3 + Treg cells in vivo, our present findings reveal the host requirement for an intact, but not reduced a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

Introduction
Regulatory T (Treg) cells, the central component in the regulation of the immune system, play a pivotal role in the maintenance of self-tolerance and immune homeostasis [1]. Research on the molecular bases of Treg cells' function has revealed the X-linked transcription factor forkhead box P3 (Foxp3) as uniquely expressed in Treg cells. How Foxp3 + Treg cells are regulated is a critical question yet to be fully explored. Growing evidence has established that Foxp3 + Treg cell function is greatly affected by the integrity of their receptors and the cytokines available in the milieu. Our knowledge of the intricacies of such regulation has now considerably expanded.
Cytokines such as interleukin-2 (IL-2), interleukin-15 (IL-15) [2][3][4], interferon gamma (IFN-γ) [5,6], interleukin-12 (IL-12) [7], interleukin-6 (IL-6) [8,9], and interleukin-4 (IL-4) [10][11][12][13][14][15] have been reported to provide critical signals in the modulation of Treg cells' development and function. IL-4, the canonical cytokine defining Type 2 immune responses, signals through the IL-4 receptor alpha (IL-4Rα) to mitigate Treg cell function during Type 2 diseases [16,17]. Recent reports have shown that augmentation of IL-4Rα signaling through gain-offunction mutation [14,15] or chronic Type 2 inflammation [18] leads to a drastic reduction in Foxp3 + Treg cell population and impairment of Treg cell suppressive function, which in turn drive their reprogramming toward T helper 2 (Th2)-like or T helper 17 (Th17)-like cells [14,15,18], favoring the notion of an inhibitory role for this receptor in Treg cell function. However, treatment of cluster of differentiation 4 (CD4) + cluster of differentiation 25 (CD25) + Treg cells in vitro with IL-4 has been shown to have an antiapoptotic role, an augmentation of the rate of Foxp3 expression, and potentiation of Foxp3 + Treg cell suppressive function [10], suggesting an unappreciated supporting role for IL-4Rα-mediated signaling in Foxp3 + Treg cell function in vitro yet to be validated in vivo. In the present study, we show that Foxp3 + Treg cells in secondary lymphoid organs constitutively express IL-4Rα and up-regulate the receptor expression upon S. mansoni (Sm) infection. We provide evidence that either 40% or 90% of deletion of the IL-4 receptor specifically within the Foxp3 + Treg cell population leads to aggravated tissue inflammation during helminth infections (Sm and N. brasiliensis [Nb] infections) and that this occurs as a result of a weakened Foxp3 + Treg cell compartment paralleled by an uncontrolled effector T-cell compartment. Adding to previous reports that have defined a restraining role for IL-4Rα on Foxp3 + Treg cells in vivo, we now present complementary evidence indicating that this receptor is equally needed, in full, in vivo for Foxp3 + Treg cell ability to control inflammation during helminth infections.
To assess the functional impairment of IL-4Rα-mediated signaling on CD4 + Foxp3 + Treg cells in our newly generated model, pooled spleen and MLN cells of naïve male and female Foxp3 cre IL-4Rα −/lox mice and their littermate controls were cultured with or without recombinant IL-4 (rIL-4) for 40 hr and/or 1 hr, and then, levels of IL-4Rα surface expression and signal transducer and activator of transcription 6 (STAT6) phosphorylation were measured by flow cytometry, respectively. CD19 + B cells and CD4 + Foxp3 − T cells from Foxp3 cre IL-4Rα −/lox mice and their littermate control had a comparable level of IL-4Rα expression, as determined by IL-4Rα geometric mean fluorescence intensity (GMFI), after the addition of rIL-4 ( Fig 2J  and 2K). In agreement, no major differences in the level of phosphorylated STAT6 (p-STAT6) either at baseline (S2A and S2B Fig) or after rIL-4 stimulation (S2A Fig and S2C-1E) in both populations (CD19 + B cells and CD4 + Foxp3 − T cells) were noted between Foxp3 cre IL-4Rα −/ lox mice and their littermate controls. In contrast, a proportion of rIL-4-stimulated CD4 + Foxp3 + Treg cells derived from Foxp3 cre IL-4Rα −/lox mice were significantly impaired in their ability to up-regulate IL-4Rα expression compared to CD4 + Foxp3 + Treg cells derived from their littermate control (Fig 2J and 2K). In support, STAT6 phosphorylation upon rIL-4 stimulation in CD4 + Foxp3 + Treg cells in IL-4Rα −/lox mice was significantly higher compared to Foxp3 cre IL-4Rα −/lox mice (S2A and S2C Fig), indicating that the IL-4Rα signaling pathway on CD4 + Foxp3 + Treg cells is impaired in our Foxp3 cre IL-4Rα −/lox mice. Of interest, we noted that at baseline, male mice do have a higher level of p-STAT6 when compared to female mice (S2B Fig). In our Foxp3 cre IL-4Rα −/lox mice, even though the increase in STAT6 phosphorylation upon rIL-4 stimulation in Foxp3 + Treg cell compartment was higher in female mice than males (S2D and S2E Fig) because of the differential deletion of il4ra gene on Foxp3 + Treg cells, the initial higher level of p-STAT6 noticed in male mice brought the absolute final level of p-STAT6 in male and female mice to a comparable level (S2C Fig).

Impairment of IL-4Rα within Foxp3 + Treg cells neither alters Foxp3 + Treg cell compartment nor breaks tolerance under a steady state
Equipped with the aforementioned murine model, we first interrogated the need for Foxp3 + Treg cells to express IL-4Rα expression under a steady state in vivo. Naïve Foxp3 cre IL-4Rα −/ lox mice and littermate control, IL-4Rα −/lox mice, were examined for alteration of Foxp3 + Treg cell compartments, overall tissue pathologies, and tissue cellularities. Foxp3 cre IL-4Rα −/lox mice had similar tissue frequencies of CD4 + Foxp3 + Treg cells among cluster of differentiation 3 (CD3) + T cells when compared to their littermate control (Figs 3A and 3B and S3A). No aberrant changes in body weight (Figs 3C and S3B), vital organs' weight (Figs 3D and S3C), or organs' cellularities (Figs 3E and S3D) were noted. Foxp3 cre IL-4Rα −/lox mice also displayed normal lymphocyte compartments in their primary lymphoid organ (thymus), secondary lymphoid organs (spleen and MLN), and peripheral tissues (lung and liver) (Fig 3F-3K and S3E-2I). Together, our findings suggest that no major physical alterations are consequent to the impairment of IL-4Rα-mediated signaling within the Foxp3 + T-cell compartment in our Foxp3 cre IL-4Rα −/lox mouse model under a steady state.
Mice with a Treg cell-specific deletion of GATA3, a transcription factor closely associated with IL-4Rα-mediated signaling [24], have been shown to develop a spontaneous inflammatory disorder with an increased release of inflammatory soluble mediators [25]. To address whether the deletion of IL-4Rα within Foxp3 + Treg cells would also instruct a spontaneous and perhaps more subtle inflammatory response in Foxp3 cre IL-4Rα −/lox mice, serum levels of soluble inflammatory mediators and liver enzymes and cytokine production by CD4 + T cells were determined. Serum levels of IL-4, tumor necrosis factor alpha (TNF-α), immunoglobulin E (IgE; Figs 3L and S3J), and aspartate aminotransferase (AST) (Figs 3M and S3K) in Foxp3 cre IL-4Rα −/lox mice were similar to those of littermate controls. All CD4 + T cell-derived cytokines tested were also similar between Foxp3 cre IL-4Rα −/lox mice and littermate controls (Figs 3N and 3O and S3L). Collectively, these results show that in vivo under a steady state, impairment of IL-4Rα-mediated signaling within the Foxp3 + T-cell compartment neither alters Foxp3 + Treg cell compartments nor results in spontaneous inflammatory disorder in Foxp3 cre IL-4Rα −/lox mice.

Intact IL-4Rα-mediated signaling on Foxp3 + Treg cell is required for the regulation of exacerbated immune responses
The strategy of probing subtle immune impairments, nonapparent during a steady state, under more inflammatory settings has previously proven to be efficient in unveiling hidden immune defects [5,27]. Moreover, the aforementioned enhancement of survival and Foxp3 expression of CD4 + CD25 + Treg cells upon induction of IL-4/IL-4Rα-mediated signaling in vitro and the striking higher expression of IL-4Rα by Foxp3 + Treg cells observed during infection warranted us to further address the possible higher need for this receptor by Foxp3 + Treg cells, in vivo, during inflammatory disease conditions. To do so, IL-4Rα −/lox , Foxp3 cre IL-4Rα −/lox , and IL-4Rα −/− mice were infected with Sm cercariae and euthanized 8 wk thereafter. Deletion of il-4rα within the Foxp3 + Treg cell population was confirmed at the genomic level by qPCR ( Fig 4A) and at the protein level by flow cytometry (Figs 4B and S5A and S5B), at which specific deletion of IL-4Rα on Foxp3 + Treg cells (more efficient in males, when compared to females), but not on CD19 + B cells, was observed. Flow cytometry analysis of Foxp3 + Treg cells within the Sm-diseased livers showed significant reduction of Foxp3 + Treg cell infiltration into the liver of Sm-infected Foxp3 cre IL-4Rα −/lox when compared to their littermate controls (Figs 4C and 4D and S5C). Furthermore, we noted that Foxp3 + Treg cells in the liver of Sm-infected Foxp3 cre IL-4Rα −/lox mice had a significant reduction in the expression levels of cells. (O) Frequency of IFN-γ, IL-4, IL-10, and IL-13-expressing CD4 + T cells. MLN cells from naïve mice were restimulated with PMA/Ionomycin in the presence of monensin, after which CD4 + T cells stained intracellularly for indicated cytokines. Results are representative of two independent experiments with 7-9 mice/group. Data are expressed as mean ± S.E.M. NS, P > 0.05; � P < 0.05, �� P < 0.001, ��� P < 0.0001 by two-tailed unpaired Student t test. Underlying data can be found in S1 Data. AST, aspartate aminotransferase; CD3, cluster of differentiation 3; CD4, cluster of differentiation 4; CD8, cluster of differentiation 8; CD19, cluster of differentiation 19; FMO, fluorescence minus one; Foxp3, forkhead box P3; FSC, forward scatter; IFN-γ, interferon gamma; IgE, immunoglobulin E; IL-4, interleukin-4; IL-10, interleukin-10; IL-13, interleukin-13; IL-4Rα, interleukin-4 receptor alpha; MLN, mesenteric lymph node; ND, not detectable; NS, not significant; PMA, phorbol myristate acetate; SSC-A, side scatter A; TNF-α, tumor necrosis factor alpha; Treg, regulatory T.  Together, these findings strongly suggest that the impairment of IL-4/IL-4Rα-mediated signaling within the Foxp3 + T-cell compartment diminishes Foxp3 + Treg cell recruitment to affected tissues during inflammation and impairs Treg cell ability to up-regulate suppressive factor, culminating into a heightened immune activation.
We next questioned the nature of the heightened immune response observed during inflammation in Foxp3 cre IL-4Rα −/lox mice. To do so, MLN cells from Sm-infected Foxp3 cre IL-4Rα −/lox mice and littermate controls were probed for CD4 + T cell-expressing GATA3 as a marker of Th2 responses [24], T-box transcription factor (T-bet) as a marker of T helper 1 (Th1) responses [28], RAR-related orphan receptor gamma (RORγt) as a marker of Th17 responses [29], and B cell lymphoma 6 (Bcl-6) as a marker of T follicular responses [30,31]. Consistent with the Th2 hegemony that is characteristic of Sm infection [32], MLN CD4 + T cells from Sm-infected Foxp3 cre IL-4Rα −/lox mice overexpressed GATA3 (Figs 4N and S5L and S6A); however, all other T-cell polarization markers remained unaltered compared to littermate control (Figs 4N and S5L and S6A). Of importance, CD4 + GATA binding protein 3 (GATA3) + cells from Sm-infected Foxp3 cre IL-4Rα −/lox mice produced more IL-4, IL-10, and IL-13 compared to littermate controls (Figs 4O and S5M and S6B). Together, these suggest that CD4 + GATA3 + responses likely drive the heightened immune response and the elevated cytokine production observed in Foxp3 cre IL-4Rα −/lox mice during inflammation. Nevertheless, with the recent description of ex-Foxp3 T cells [14,15,18], this observation could be confounded by the likelihood of cytokine production by newly formed ex-Foxp3 + T cells dually expressing GATA3 and Foxp3 and releasing cytokines following deletion of IL-4Rα. To the presence of monensin. (N) Frequency of indicated transcription factor-expressing CD4 + T cells in the MLN 8 wk post infection.

Deletion of IL-4Rα within Foxp3 + Treg cell population impairs their ability to control tissue inflammation in diseases
To further appraise the consequences of the heightened immune responses observed in diseased Sm-infected Foxp3 cre IL-4Rα −/lox mice, egg-driven fibrogranulomatous inflammation, as well as Foxp3 + Treg cells' infiltration in the liver of Sm-infected Foxp3 cre IL-4Rα −/lox mice, was microscopically assessed in comparison with that of Sm-infected littermate controls. We noted that Sm infection resulted in enlarged egg-driven granulomas in the liver of Foxp3 cre IL-4Rα −/ lox mice when compared to granulomas in the liver of their littermate control (Fig 5A and 5B).
In agreement with this observation, we found that the average number of Foxp3 + Treg cells within egg-driven granulomas in Foxp3 cre IL-4Rα −/lox mice was significantly reduced when compared to the amount of Foxp3 + Treg cells recruited to the liver of littermate controls ( Fig  5C and 5D). Furthermore, collagen levels, attesting the degree of tissue fibrosis, were considerably higher in the livers of Sm-infected Foxp3 cre IL-4Rα −/lox mice when compared to the levels reported in Sm-infected littermate controls ( Fig 5E). Indeed, colorimetric measurement of 4-hydroxyproline, a direct product of acid hydrolysis of collagen, confirmed our observed increased level of collagen in the liver of Sm-infected Foxp3 cre IL-4Rα −/lox mice (Figs 4F and S5F). In fact, with similar egg burdens in the guts and livers of Foxp3 cre IL-4Rα −/lox mice and their littermate control (Fig 5G), Sm-infected Foxp3 cre IL-4Rα −/lox mice still displayed much larger coalescing granulomas (Fig 5H), as indicated by hematoxylin-eosin (HE) (Fig 5A) and chromotrope aniline blue (CAB) staining ( Fig 5E). Similar to the liver, gut tissues of Sminfected Foxp3 cre IL-4Rα −/lox mice revealed elevated levels of cellular recruitment around trapped parasite eggs (Fig 5I), which was associated with a more pronounced deposition of collagen (Fig 5J), as confirmed by gut hydroxyproline content (Figs 5K and S5O), indicating a heightened fibrogranulomatous inflammation around Sm-trapped eggs in Foxp3 cre IL-4Rα −/ lox mice when compared to littermate controls.
Since Cre expression can generate a phenotype of its own [33][34][35][36][37][38][39], we sought to address whether the uncontrolled immune responses and the exaggerated fibrogranulomatous inflammation noted in the Sm-infected Foxp3 cre IL-4Rα −/lox mice were due to the specific deletion of IL-4Rα on the Foxp3 + Treg cells or were an artefact effect from the Cre transgene alone. To test that, IL-4Rα +/+ and Foxp3 cre IL-4Rα +/+ mice were infected with 100 Sm cercariae, and then, Foxp3 + Treg cell compartments and fibrogranulomatous inflammation in liver and gut were investigated 8 wk post infection. We noted that IL-4Rα expression on either CD19 + B cells or  Taken together, these observations indicate that the impairment of IL-4Rα-mediated signaling, but not the Cre transgene expression, specifically within the Foxp3 + Treg cell compartment during experimental schistosomiasis drives a poor accumulation of Foxp3 + T cells in the inflamed tissues and consequently results in elevated host fibrogranulomatous responses around the trapped parasite eggs.
We next aimed to test whether this would also hold true in another inflammatory helminth disease model. We found that following subcutaneous infection of Foxp3 cre IL-4Rα −/lox and IL-4Rα −/lox littermate control mice with Nb, the airways of Nb-infected Foxp3 cre IL-4Rα −/lox mice showed much heavier mucus production 9 d post infection (S9A and S9B Fig). Moreover, we noted that Foxp3 + Treg cells' recruitment to the inner layer of the alveoli of the lungs of Nb-infected Foxp3 cre IL-4Rα −/lox mice (S9C and S9D Fig) was diminished, indicating that in this model as well the impairment of IL-4Rα-mediated signaling within the Foxp3 + T-cell compartment impairs their ability to accumulate at the inflamed tissue and to control local tissue inflammation. Collectively, these results clearly suggest, in two different and major inflammatory helminth models, that impairment of IL-4Rα-mediated signaling within the Foxp3 + Tcell compartment leads to uncontrolled immune responses and exacerbated tissue(s) inflammation, underscoring a hitherto-unappreciated role of IL-4Rα-mediated signaling on Foxp3 + Treg cells to control inflammatory responses in vivo.

IL-4Rα-mediated signaling on Foxp3 + Treg cell is required for the generation of potent effector Foxp3 + Treg (eTreg) cell population
Next, we investigated the mechanisms underpinning the poor ability of Foxp3 + Treg cells to accumulate in the inflamed tissues in our diseased Foxp3 cre IL-4Rα −/lox mice. First, we assessed the effect of IL-4Rα deletion specifically on Foxp3 + T-cell compartment for its ability to build up the eTreg cells needed to contain the immune responses at the site of inflammation [40]. To address that, central Foxp3 + Treg (cTreg) cell and eTreg cell populations in hepatic lymph nodes (hLNs) and MLNs of Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk were assessed in comparison to those of Sm-infected littermate controls. In both lymph nodes, deletion of IL-4Rα specifically on Foxp3 + Treg cells resulted in the accumulation of cTreg cells (Figs 6A and 6B and S10A), which was associated with a significant reduction in the pool of eTreg cell Regulation of Foxp3+ Treg cells via IL-4 receptor population (Figs 6A and 6C and S10B), indicating that IL-4Rα-mediated signaling on Foxp3 + Treg cells could be playing a role in inducing the conversion of cTreg cell to eTreg cell. Furthermore, the ability of these eTreg cells to maintain or up-regulate the expression of C-X-C motif chemokine receptor 3 (CXCR3), a chemoattractant receptor required for the migration of eTreg cells to nonlymphoid tissues (in particular, liver [40,41]), was significantly impaired in both hLNs (Fig 6D and 6E) and MLNs (S10C and S10D Fig) of Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk. In support, the expression level of GATA3, a transcription factor required for Treg cell accumulation at the site of inflammation [42], within the Foxp3 + Treg cell in the liver of Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk was diminished when compared to littermate controls (Fig 6F and 6G). Together, these results suggest that IL-4Rαmediated signaling on Foxp3 + Treg cell is required for the maintenance or up-regulation of the markers required for migration and accumulation of Foxp3 + Treg cell in the inflamed tissue.
Then, we traced back whether the reduction in the Foxp3 + Treg cell population in the liver of Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk was due to a global defect in Foxp3 + Treg cells or in the Treg cell de novo conversion in vivo, since eTreg cell can be thymic-derived and/or peripherally induced. To address that, we analyzed the thymic-derived Foxp3 + Treg cells (Foxp3 + Helios + Treg cells) and the peripherally iTreg cells (Foxp3 + Helios − Treg cells) in the liver of Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk and their littermate controls. We noted that in the liver of Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk, both Foxp3 + Helios + and Foxp3 + Helios − Treg cells were significantly reduced (Fig 6H and 6I), suggesting a global defect in the Foxp3 + Treg cell population.
Finally, inasmuch as the reduction of Foxp3 + Treg cell population was independent of its proliferation as Foxp3 + Treg cells in the liver of Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk and their littermate control had a comparable expression level of Ki-67 (Fig 6J and 6K), we therefore asked whether the reduction of Foxp3 + Treg cell population in the liver could be possibly driven, in addition to their poor migration and accumulation, by the lessened ability of these Foxp3 + Treg cells to survive in the absence of IL-4Rα-mediated signaling. Indeed, we noted a diminished expression level of the pro-survival factor B cell lymphoma 2 (Bcl-2; Fig 6L  and 6M) within the Foxp3 + Treg cell population in the liver of the Foxp3 cre IL-4Rα −/lox mice infected with Sm for 8 wk when compared to their littermate control. In support, those Foxp3 + Treg cell population had a higher expression level of the apoptotic marker caspase-3 (Fig 6N  and 6O) as well as the fixable viability stain (Fig 6P and 6Q), indicating that the Foxp3 + Treg cell population in the liver has a higher propensity to undergo apoptosis in the absence of IL-4Rα-mediated signaling, which recapitulates our in vitro data mentioned earlier. Collectively, our results suggest that IL-4Rα-mediated signaling is required for the conversion of cTreg cells into eTreg cells. These eTreg cells further need the IL-4Rα-mediated signaling to maintain or up-regulate markers required for their migration, accumulation, and survival at the site of inflammation.

Discussion
Studies on Foxp3 + Treg cell function have uncovered a critical regulatory role for the cytokines present in the milieu (reviewed in [9]). IL-4, the canonical cytokine of Type 2 immune responses that signals via IL-4Rα, signaling on Treg cells has been shown to promote Treg cell reprogramming toward Th17 [15] or Th2 [14,18] cells. As a consequence, enhanced signaling via IL-4Rα in Treg cells has resulted in impaired suppressive function and heightened inflammation, establishing the concept of a negative regulation of Treg cell activity by IL-4Rα-mediated signaling. The present study, however, shows that IL-4Rα-mediated signaling is also Results pooled from two independent experiments with 3-4 mice/group. Data are expressed as mean ± S.E.M. NS, P > 0.05; � P < 0.05, �� P < 0.001, ��� P < 0.0001 by two-tailed unpaired Student t test. Underlying data can be found in S1 Data. Bcl-2, B cell lymphoma 2; CD44, cluster of differentiation 44; CD62L, L-selectin; cTreg, central regulatory T; CXCR3, C-X-C motif chemokine receptor 3; eTreg, effector regulatory T; Foxp3, forkhead box P3; GATA3, GATA binding protein 3; GMFI, geometric mean fluorescence intensity; hLN, hepatic lymph node; IL-4Rα, interleukin-4 receptor alpha; NS, not significant; Sm, S. mansoni; Treg, regulatory T. required by Treg cells to control inflammation in diseases (Fig 7). First, in an experimental murine model of schistosomiasis, we demonstrated that Foxp3 + Treg cells up-regulated IL-4Rα expression. Further analyses on the need for the host to up-regulate this receptor during schistosomiasis were then conducted using a murine model of specific deletion of IL-4Rα within the Foxp3 + Treg cell population, termed Foxp3 cre IL-4Rα −/lox mice. Specific removal of IL-4Rα in a part or almost totally on the Foxp3 + Treg cell population exacerbated tissue inflammation during experimental Sm-and Nb-mediated diseases. Aggravated immunopathology was associated with a decline in Treg cell accumulation in the diseased tissues and the reduced expression of suppressive markers, paralleled by an augmentation of immune effector responses as translated in the Sm model in Foxp3 cre IL-4Rα −/lox mice. Collectively, these findings unprecedentedly, to our knowledge, uncovered a critical need for IL-4Rα-mediated signaling on Foxp3 + Treg cells to control tissue immunopathology in helminthiases. Such a need

Regulation of Foxp3+ Treg cells via IL-4 receptor
for Foxp3 + Treg cells for this signaling axis is yet to be examined in allergic Th2 inflammation and Th1-or Th17-dominated diseases.
Recently published reports have shown that augmentation of IL-4Rα-mediated signaling on Foxp3 + Treg cells leads to an impaired suppressive activity of those cells with subsequent reprogramming into ex-Foxp3 Th2 or Th17 cells [14,15,18]. Although these observations would suggest that IL-4Rα signaling might not favor Treg cell activity, our present analyses revealed that Foxp3 + Treg cells have a basal expression level of IL-4Rα under a steady state, pointing at a possible requirement for this receptor in Treg cell biology. Our assumption is in accordance with a previous attempt by other investigators that have reported the potency of IL-4 to prevent spontaneous apoptosis and reduction of Foxp3 expression in the cultures of isolated Treg cells [10]. In fact, IL-4 stimulation of isolated cultures of CD4 + CD25 + Treg cells in our study recapitulated such a need for IL-4Rα-mediated signaling to prevent cell death and to maintain or enhance Foxp3 expression, indicating a hitherto-unappreciated need for IL-4Rα-mediated signaling in Treg cell biology and activity. IL-4Rα expression by Foxp3 + Treg cells was further enhanced following Sm infection, suggesting a strong requirement for this receptor either to preserve Treg cell activity or foster their transdifferentiation into ex-Foxp3 Th2 cells. In fact, the reprogramming of Foxp3 + Treg cells into functional effector ex-Foxp3 Th2 cells has now been demonstrated during Heligmosomoides polygyrus infection [18], during which ex-Foxp3 Th2 cells generated in an IL-4Rα-dependent manner were shown to contribute to the overall antiparasitic Th2 response [18]. Hence, it is possible that IL-4Rα up-regulation on Foxp3 + Treg cells during acute experimental schistosomiasis observed in our study could occur as an attempt for the host to rapidly mount a protective Th2 immune response during acute schistosomiasis [32]. More experiments are clearly needed at this point to address the likelihood of such a contribution of ex-Foxp3 in the host protective Th2 response during experimental schistosomiasis.
IL-4Rα removal specifically from the Foxp3 + Treg cell compartment was achieved by making use of the Cre-lox system. A murine model, Foxp3 cre IL-4Rα −/lox , was obtained in which IL-4Rα was deleted specifically within the Foxp3 + Treg cell compartment in a dosage-dependent manner whereby IL-4Rα is partially deleted in females and quasi-completely deleted in males. The more efficient Cre-mediated IL-4Rα deletion achieved in male when compared to female Foxp3 cre IL-4Rα −/lox mice strongly fits the phenomenon of random inactivation of the X chromosome that can take place in females but not males, as previously reported [43,44]. Our approach herein described could therefore set a precedent for the generation of transgenic mouse models with various levels of knockdown of a target gene.
We found, consistent with the literature [10], that isolated cultures of CD25 + Treg cells tightly depended on IL-4Rα for survival and maintenance of Foxp3 expression. However, partial or quasi-complete removal of IL-4Rα from the Foxp3 + Treg cell population did not affect the Treg cell compartment during steady state. A possible explanation could be that the remnant IL-4Rα on Foxp3 + Treg cells in our Foxp3 cre IL-4Rα −/lox model is sufficient to preserve the Foxp3 + Treg cell compartment and thus tolerance during a steady state in which randomly occurring inflammatory responses are not common. This is consistent with previous studies that have attempted to define the physiological importance of factors expressed by Foxp3 + Treg cells. Specifically, in these studies, Foxp3-specific GATA3-or T-bet-deficient mice were clinically indistinguishable from their littermate controls and showed no defect on the function of Foxp3 + Treg cells under a steady state in young mice [5,27,42]. In older mice (6 mo upwards), however, the targeting of GATA3 within the Foxp3 + Treg cell population led to inflammatory disease [25]. It is therefore feasible that specific deletion of IL-4Rα from the Foxp3 + Treg cell compartment might also lead to apparent physiological defects only in old mice, but this is still to be addressed experimentally. Notably, inflammation in these models of Foxp3-specific GATA3 or T-bet deletion uncovered more robustly a defect in the Foxp3 + Treg cell compartment that translated into physiological impairments [27]. Similarly, infection of our Foxp3 cre IL-4Rα −/lox mice resulted in elevated inflammatory responses in helminth-mediated disease models. This was defined by increased levels of Foxp3 − T-cell proliferation and activation as well as a clear augmentation of cytokine production, indicating a more potent effector response following either partial or quasi-complete deletion of IL-4Rα from the Foxp3 + Treg cell compartment during inflammation. It is also important to point out that we observed similarly heightened cytokine release and inflammation in diseased female Foxp3 cre IL-4Rα −/lox mice, in which the less efficient deletion of IL-4Rα was observed when compared to Sm-infected male Foxp3 cre IL-4Rα −/lox mice. The possible explanations could be that (i) disruption of the IL-4Rα signaling pathway on Foxp3 + Treg cells as low as 40% is enough to completely disequilibrate the inflammation-regulation balance during infection, and/or (ii) the functional characterization of males and females mice, by using STAT6 phosphorylation as a readout, demonstrated that the absolute final levels of STAT6 phosphorylation in cells from male and female Foxp3 cre IL-4Rα −/lox mice were comparable (S1C Fig), indicating that male and female Foxp3 cre IL-4Rα −/lox mice do have a similar defect at the functional level of the IL-4Rα signaling pathway on Foxp3 + Treg cell population. Such observations could provide an explanation for the similar phenotype in Foxp3 cre IL-4Rα −/lox female and male mice during inflammation, when elevated IL-4 is produced.
As of yet, the general observation of heightened inflammatory responses in diseased Foxp3-specific IL-4Rα-impaired mice could be rooted under two opposing but not mutually exclusive explanations-i.e., (i) IL-4Rα as a critical receptor in conferring to Foxp3 + Treg cells the necessary potency to control inflammatory responses in disease and/or (ii) the need for this IL-4Rα receptor to foster Foxp3 + Treg cells transdifferentiation into effector T cells. The former explanation does find support in our observation of a significant reduction in Foxp3 expression levels within the Foxp3 + Treg cell population following removal of IL-4Rα from the Foxp3 compartment in Sm-infected Foxp3 cre IL-4Rα −/lox mice. In fact, Foxp3 expression is the defining factor that endows Treg cell with a suppressive ability, and its continuous expression is required to maintain the transcriptional and functional programs of Treg cells [45][46][47][48]. Consequently, our observation of a reduced Foxp3 expression during inflammation following IL-4Rα removal on Foxp3 + Treg cells defines a need for this receptor in stabilizing and promoting Foxp3 expression by Foxp3 + Treg cells in vivo. Although the latter explanation of a negative regulation of Foxp3 + Treg cells by IL-4Rα signaling in favor of ex-Foxp3 transdifferentiation can be appropriately dismissed by using fate-reporter mice, our present study, as it stands, convincingly indicates that the T cells responsible for the increased cytokine production reported in Foxp3 cre IL-4Rα −/lox diseased mice expressed not Foxp3 but rather GATA3, a profile inconsistent with recently reported ex-Foxp3 T cells during helminth infections [18]. This argued against Foxp3 Treg cells' transdifferentiation into cytokine-producing effector T cells as mediating the heightened inflammation observed in diseased Foxp3 cre IL-4Rα −/lox mice.
Partial or quasi-complete removal of IL-4Rα within the Foxp3 + Treg cell compartment of Foxp3 cre IL-4Rα −/lox diseased mice led to an impaired accumulation of Treg cells to the site of inflammation. Mechanistically, various processes can mediate such an impaired accumulation of Foxp3 cre IL-4Rα −/lox Treg cells observed in our infectious settings, including defects in (i) the extrathymic Treg cell conversion, (ii) the proliferation capacity, (iii) the generation of eTreg cells, (iv) the migration and the accumulation of eTreg cells at the site of inflammation, and/or (v) the survival of Treg cells. We found no obvious defect of CD25 − CD4 + cell precursors from Foxp3 cre IL-4Rα −/lox mice in converting into Foxp3 + Treg cells in vitro, arguing against a defect in extrathymic Treg cell conversion. Ex vivo analyses of Foxp3 + Treg cells from diseased Foxp3 cre IL-4Rα −/lox mice showed no signs of impaired proliferation, dismissing Foxp3 + Treg cell-defective proliferation as a possible cause of the poor accumulation reported. However, we found that cTreg cells in the lymph nodes of diseased Foxp3 cre IL-4Rα −/lox mice had a poor ability to convert into eTreg cells, resulting in a significant reduction in the eTreg cell population required to populate the nonlymphoid tissues that further contain the inflammatory immune responses [40]. Furthermore, migration and accumulation of these eTreg cells were impaired, as indicated by the reduction in the expression level of CXCR3 and GATA3, respectively. Our results are in full accordance with previous reports on the need of Foxp3 + Treg cells to express GATA3 and CXCR3 to migrate and accumulate in the inflamed tissues and in the liver, respectively [9,41,42]. In addition to the reduction in the pool of eTreg cell population and their poor ability to migrate and accumulate at the site of inflammation, Foxp3 + Treg cells in the liver of diseased Foxp3 cre IL-4Rα −/lox mice had a higher propensity to undergo apoptosis, recapitulating our in vitro data, which showed that CD25 + Treg cells' survival and stability (as judged by Foxp3 expression) are compromised in the absence of IL-4.
Overall, our results support the idea that an intact, rather than a potentiated or diminished, IL-4Rα-mediated signaling is optimal to endow Foxp3 + Treg cells with their most efficient ability to control inflammation in disease as illustrated here in helminthiases. As per our current findings, such an optimal response is most likely achieved by preventing Foxp3 + Treg cell death and stabilizing the expression of canonical Treg cell-suppressive and associated markers such as Foxp3 and GATA3. In light of past reports that have defined a negative role for IL-4Rα-mediated signaling on Treg cell activity in Th2-dominated settings, our present finding of a negative role of IL-4Rα deficiency on Treg cells during inflammation further highlights the fact that the same cytokine can have both positive and negative effects on Treg cell activity and reinstates the notion that balance is key. Promoting or impairing IL-4Rα-mediated signaling yielded a similar result of Foxp3 + Treg cell impairment, therefore strongly arguing against the benefits of uninformedly modulating this cytokine receptor on Foxp3 + Treg cells to ameliorate tolerance, particularly in the current global context where helminthiasis are highly common. This is therefore particularly important in light of recent therapeutic advances for the control of highly pathogenic cases of asthma for which IL-4Rα targeting has recently been introduced and is attracting further interest for other inflammatory conditions [14,15].
Conclusively, in this study, a case is made for the preservation of intact levels of IL-4Rα signaling on Foxp3 + Treg cells to ensure optimal regulatory responses in vivo, and caution is therefore raised for the careful manipulation of this signaling axis in disease.

Ethics statement
All mice were maintained in specific-pathogen-free barrier conditions in individually ventilated cages at the University of Cape Town biosafety level 2 animal facility. Experimental mice were sex-and age-matched and used between 6 and 8 wk of age. All the experimental work was in strict accordance with the recommendations of the South African national guidelines and of the University of Cape Town practice for laboratory animal procedures, as in ethics protocols 014/003 and 016/027 approved by the Animal Research Ethics Committee of the Faculty of Health Science, University of Cape Town. All efforts were made to minimize animals' suffering. Prior to percutaneous infection with Sm cercariae, animals were anesthetized by intraperitoneal injection of a cocktail of ketamine (100 mg/kg) and xylazine (10 mg/kg) and monitored for 5 min to confirm deep anesthesia. Anesthesia was confirmed by the absence of pedal reflex (toe pinch) and eyeblink reflex amid a regular respiratory rate. The anesthesia duration was of a maximum of 30 min. During the anesthesia phase, animals were exposed to an infrared lamp to help them maintain their core body temperature. This procedure was performed and duly cared for by trained and authorized researchers. Post infection, animals were monitored until regaining of consciousness, and moistened food was added to the cage bedding. Upon reaching the study's experimental end point and/or the protocol-defined humane end point, animals were euthanized under this study by exposure to an excess of halothane (4% in air) for 5 minutes. Death was confirmed either by neck dislocation or exsanguination by cardiac puncture. Death was not a predetermined end point in any of the arms of this study.

Mice infection and challenge
Sm infection. Mice were percutaneously infected via the abdomen, using stainless steel rings, with 100 viable cercariae of a Puerto Rican strain of Sm obtained from infected Biomphalaria glabrata snails (NMRI strain, NR-21962, provided by Biomedical Research Institute, Rockville, MD, United States). Mice were euthanized 8 wk post infection, and eggs isolated from liver and ileum were counted as previously described [49].
Nb infection. Mice were injected subcutaneously with 500 Nb L3 larvae suspended in 0.9% NaCl using a 21-G needle (Braun, Melsungen, Germany). Mice were euthanized 9 d post infection, and tissue samples were collected for analyses.

Cell isolation
Thymus and MLN single-cell suspensions were prepared by mechanical squeezing through a 40 μM cell strainer (Falcon, Corning, MA, US). Single-cell spleen suspensions were prepared by mechanical squeezing through a 70 μM cell strainer (Falcon). Liver lymphomyeloid cells were isolated following a modified version of the method of Gossen and colleagues and Hardy and colleagues [50,51]. Briefly, single-cell suspensions from the liver tissues were prepared by chopping them into 10 mm 3 small pieces and incubating them in Iscove's Modified Dulbecco's Medium (IMDM) containing 220 U/mg Collagenase I (Gibco, Waltham, MA, US), 13 U/mg DNase I (Sigma, St. Louis, MO, US), and 5% inactivated fetal calf serum (iFCS; Gibco) (digestion buffer) for 30 min at 37˚C under constant rotation. The resulting suspension was mechanically squeezed through a 100 μM sterile cell strainer (Falcon), followed by centrifugation at 1,200 rpm for 10 min at 4˚C. Supernatant was discarded and the cells resuspended in 36% of isotonic Percoll (Sigma). This suspension was mixed thoroughly, and separation was performed at 500 g, without breaks, for 10 min at 4˚C. For lungs, single-cell suspensions were prepared by chopping the lungs into 10 mm 3 pieces, incubating them in digestion buffer for 30 min at 37˚C, and mechanically squeezing them through a 70 μM cell strainer. Erythrocytes were lysed using RBC lysis buffer. Cells were washed; resuspended in IMDM containing 10% iFBS, 100 U/ml penicillin (Gibco), and 100 μg/ml streptomycin (Gibco) (IMDM culture medium); and checked for viability and cell number by trypan blue staining.

Cell culture
Pooled cells from spleen and MLN from naïve Foxp3 cre IL-4Rα −/lox mice and their littermate control were cultured at 5 × 10 6 cells/ml in RPMI medium (Lonza, Walkersville, MD, US) supplemented with 10% iFBS and penicillin and streptomycin (100 U/ml and 100 μg/ml) (RPMI culture medium) for 40 hr at 37˚C in a humidified atmosphere containing 5% CO 2 , in medium supplemented with 0 or 1 ng/ml rIL-4 (BD Biosciences). For measuring the level of STAT6 phosphorylation, spleen and MLN cells from naïve Foxp3 cre IL-4Rα −/lox mice, their littermate control, and global knock-out mice were cultured at 1 × 10 6 cells/ml RPMI culture medium for 1 hr at 37˚C in a humidified atmosphere containing 5% CO 2 , in medium supplemented with 0 or 10 ng/ml rIL-4. After the incubation period, cells were harvested, washed in FACS buffer, and then stained for IL-4Rα and/or p-STAT6. IL-4Rα expression and level of STAT phosphorylation were detected by flow cytometry. In other settings, sorted CD4 + CD25 + T cells from naïve Foxp3 cre IL-4Rα −/lox mice and their littermate control were cultured at 1 × 10 6 cells/ml in RPMI medium for 18 or 36 hr, at 37˚C in a humidified atmosphere containing 5% CO 2 , with 0 or 10 ng/ml rIL-4; then, cells were checked for survival and Foxp3 expression by flow cytometry.

In vitro Treg cell conversion assay
CD4 + T cells were enriched from a pool of naïve spleens and MLN cells by using EasySep Mouse T Cell Isolation Kit (Stemcell Technologies, Vancouver, Canada, Catalogue no. 19852A) as per the manufacturer's instruction, and then, the CD4 + CD25 − population was sorted by using BD FACSAria III cell sorter (BD Biosciences). Sorted CD4 + CD25 − cells (2.5 × 10 6 cells/ml) were cultured with plate-bound anti-CD3 (10 μg/ml) in the presence of soluble anti-CD28 (2 μg/ml) and different concentrations of TGFβ (0, 1, or 2 ng/ml) (BD Pharmingen) in 96-well, flat-bottomed plates in triplicates. After 3 d, the iTreg cells were analyzed by flow cytometry for CD25 and Foxp3 expression.

Serum analyses
Cytokines and IgE. Serum cytokines (IL-4 and TNF-α) and total IgE titer were measured as follows. Blood was collected in serum separator tubes (BD Biosciences) and centrifuged at 8,000 g for 10 min at 4˚C to separate serum. The Nunc MicroWell 96-Well Microplates (Thermo Fisher Scientific) were coated with 0.5 ng/ml capture antibody in PBS and incubated overnight at 4˚C. Plates were washed and blocked with 2% (w/v) milk powder for 2 hr at 37˚C, and samples were loaded and incubated overnight at 4˚C. Detecting, biotin-labeled antibodies were added and incubated for 2 hr at 37˚C, and then, avidin-horseradish peroxidase or avidinalkaline phosphatase was added and incubated for 1 hr at 37˚C. Plates were developed by adding TMB Microwell Peroxidase Substrate (KPL, Gaithersburg, MD, US) or p-nitrophenyl phosphate disodium salt hexahydrate (Sigma), respectively. The absorbance was read at 450 or 405 nm using a VersaMax microplate spectrophotometer (Molecular Devices, Sunnyvale, CA, US).
Liver enzymes. Hepatocellular damage was assessed by measuring the serum level of AST at the National Health Laboratory Service of South Africa (Cape Town).

Histology
Sm infection. Liver and gut tissues were fixed in 4% (v/v) formaldehyde in PBS, embedded in wax, and processed, and 5-7 μm sections were stained with HE. Micrographs of liver and gut granuloma were captured using Nikon Eclipse 90i (Nikon, Minato, Tokyo). A granuloma diameter of 20-50 granulomas per animal was determined using Nikon NIS-Elements imaging software (Nikon Corporation). For fibrosis assessment, tissue sections were stained with chromotrope 2R and CAB and counterstained with Wegert's hematoxylin for collagen staining [23].
Nb infection. Lung tissues were fixed in 4% (v/v) formaldehyde in PBS, embedded in wax, and processed, and 5 μm sections were stained with periodic acid-Schiff reagent (PAS) to visualize mucus-producing goblet cells [52]. Micrographs of lung sections were captured using Nikon Eclipse 90i. Cytokines (IL-4, IL-5, IL-10, IL-13, IFN-γ, and IL-17, all from BD Pharmingen) were measured in the protein extracts by sandwich ELISA as described previously [23]. Cytokine values were normalized according to the protein content measured by Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, catalogue no. 23225).

Hydroxyproline quantification
Hydroxyproline content as a measure of collagen production was quantified using a modified protocol [53]. In brief, weighed liver samples were hydrolyzed overnight at 110˚C in 6 M HCl and then filtered through Whatman filter papers. Filtrate was neutralized with 1% phenolphthalein and titrated against 10 M NaOH. An aliquot was mixed with isopropanol and added to a chloramine-T/citrate buffer solution (pH 6.0) (Sigma). Ehrlich's reagent solution (25 g pdimethyl-amino-benzaldehyde, 37.5 ml 60% perchloric acid) was added and measured at 570 nm using a VersaMax microplate spectrophotometer (Molecular Devices). Hydroxyproline levels were calculated by using 4-hydroxy-L-proline (Calbiochem, San Diego, CA, US) as standard, and results were expressed as μg hydroxyproline per weight of liver tissue that contained 10 4 eggs.