Introduction

The intestinal epithelium is a monolayer of columnar epithelial cells lining the luminal surface and functions as a physical and immunological barrier between the host and its luminal contents1. Two major differentiated cell types are defined within the intestinal epithelium, (a) absorptive enterocytes and (b) secretory lineages, which include Paneth cells that release antimicrobial factors, goblet cells that produce mucins, enteroendocrine cells that secrete hormones and tuft cells which play a key role in helminth expulsion2,3.

Parasitic helminth infections are a major public health problem and affect more than 1.5 billion people worldwide4,5. The host defense against helminth infection is orchestrated by activation of type 2 immunity6,7. One of the characteristic type 2 immune responses is intestinal epithelial remodeling8. Although tuft and goblet cells constitute only a minor fraction of intestinal epithelial cells (IECs), tight control of their differentiation is crucial for protection against infections with gastrointestinal parasitic worms9,10. Chemosensory tuft cells sense the intestinal parasitic infection and produce IL25, which causes accumulation of type 2 innate lymphoid cells (ILC2s) and other IL4/IL13 producing cell types in the lamina propria (LP). In turn IL4/IL13 signals to the stem cells within the intestinal crypt and promotes tuft and goblet cell hyperplasia11,12,13. STAT6 is a well-known transcription factor to mediate type 2 immune responses following IL4/IL13 activation, playing important roles in the clearance of intestinal parasites and allergic disorders14,15. In response to IL4/IL13, STAT6 proteins undergo tyrosine phosphorylation and are transported to the nucleus, where along with co-factors, STAT6 regulates the expression of its target genes16. Confirming the requirement for STAT6 in anti-helminth responses, mice lacking STAT6 show abrogated expansion of tuft and goblet cells and compromised worm expulsion after helminth infections12,17,18,19. Although STAT6 has been mainly shown to function in immune cells, there is a growing appreciation of its role in non-immune cells, including IECs. Selective expression of constitutively activated STAT6 in IECs promotes proliferation and differentiation of tuft and goblet cells and protects against gastrointestinal helminth infections, indicating activation of STAT6 in epithelial cells is critical for secretory cell differentiation20.

The Sirtuin family members (SIRT1-7) are evolutionarily conserved proteins with enzymatic activity of NAD+-dependent deacetylase or mono-ADP-ribosyltransferase21. Sirtuin 6 (SIRT6) is a nuclear and chromatin-bound protein that emerges as an important epigenetic regulator controlling longevity, genome stability, metabolism, and inflammation22,23. Sirt6 systemic knockout causes severe hypoglycemia and premature death, while transgenic mice overexpressing SIRT6 have an extended lifespan in males24,25,26. Notably, IEC-specific Sirt6 knockout mice are more susceptible to dextran sulfate sodium (DSS)-induced colitis. Importantly, the protective role of SIRT6 in colitis maybe mediated by R-spondin-1 because SIRT6 fails to prevent proinflammatory cytokines induced YAMC (a normal colonic epithelial cell line) cell death when Rspo1 is silenced by siRNA27. Conversely, transgenic mice overexpressing SIRT6 exhibit improvement of DSS-induced colitis progression probably due to attenuated activation of NF-kB and c-Jun pathways in the colon28. However, the roles of SIRT6 in regulating intestinal secretory cell lineage development and type 2 mucosal immunity in response to helminth infections have not been studied.

In the present study, we demonstrate that epithelial SIRT6 is essential for tuft cell development and maintaining intestinal type 2 immunity in response to helminth infection. In IECs, SIRT6 inhibits SOCS3 transcription through histone H3K56 deacetylation, thereby enhancing STAT6 activity to promote intestinal epithelial remodeling for helminth expulsion. This study unveils a novel role of SIRT6 in type 2 mucosal immunity in the context of parasitic infections.

Results

Sirt6 deletion in IECs leads to impaired tuft cell development and intestinal type 2 immunity

Sirt6 floxed mice were crossed with Villin-Cre mice to generate IEC-specific Sirt6 knockout mice (IEC-KO, genotype: Sirt6flox/flox;VillinCre) and control mice (LoxP, genotype: Sirt6flox/flox). Knockout of Sirt6 in the jejunal or colonic epithelium of IEC-KO mice was confirmed at both mRNA and protein levels (Fig. 1a, b). IEC-KO mice were born at the expected Mendelian frequencies and exhibited no overt abnormalities. In addition, H&E staining of the jejunum and colon tissues revealed no noticeable mucosal abnormalities in IEC-KO mice compared with LoxP mice (Supplementary Fig. 1a–d). To understand the functional outcome of Sirt6 ablation in IEC differentiation, we analyzed the mRNA expression of cell lineage markers in the jejunal and colonic IECs. Interestingly, we found Dclk1 (tuft cell marker) expression was significantly reduced in IEC-KO mice compared with LoxP littermates (Fig. 1c, d). In contrast, epithelial expression of markers for goblet cells (Muc2), enterocytes (Slc5a1), enteroendocrine cells (Chga), Paneth cells (Lyz1), and stem cells (Lgr5) was comparable in LoxP and IEC-KO mice (Fig. 1c, d). To test whether Sirt6 ablation altered the epithelial cell life cycle, we further analyzed cell proliferation (Ki67) and apoptosis (TUNEL) in the jejunum but found no significant differences between LoxP and IEC-KO mice (Supplementary Fig. 1e–h).

Fig. 1: IEC-specific deletion of Sirt6 results in defective tuft cell expansion and intestinal type 2 immune responses.
figure 1

2–3-month-old male naive and/or H. poly-infected LoxP and IEC-KO mice were subjected to the following assays on day 14 post-infection. a, b Western blot (a) and qPCR (b) analyses of SIRT6 in the jejunal and colonic IECs of naive mice. n = 5 mice/group in (b); p = 0.0028 (jejunum) and 0.0014 (colon). c, d qPCR analysis of the expression of IEC markers in the jejunal (c) and colonic (d) IECs of naive mice. Dclk1, Muc2, Slc5a1, Chga, Lyz1, n = 5 mice/group; Lgr5, n = 4 mice/group; p = 0.0096 (Dclk1) in (c). e Jejunal tuft cells were examined by immunostaining with anti-DCLK1 antibody (200X). f Jejunal goblet cells were examined by Alcian blue staining (200X). g Quantification of tuft (DCLK1 positive) cells shown in (e). n = 4, 4, 5, 4 mice/group, respectively; 15 crypt-villus units counted for each mouse; from left to right, *p = 0.0188, <0.0001, 0.0093; #p < 0.0001 for both. h Quantification of goblet (Alcian blue positive) cells shown in (f). n = 4, 4, 5, 4 mice/group, respectively; 15 crypt-villus units counted for each mouse; from left to right, *p = 0.0005; #p = 0.0001, 0.0021. i, j Analysis of parasite burden by quantification of adult worms in intestinal lumen (i) and eggs in feces (j). n = 5 mice/group; p = 0.0243 in (i) and 0.0029 in (j). k qPCR analysis of tuft and goblet cell markers expression in the jejunal IECs. n = 4, 4, 5, 5 mice/group, respectively; from left to right, Dclk1 (*p = 0.0497, 0.0004; #p = 0.0005, 0.0002); Il25 (*p = 0.0474, <0.0001; #p < 0.0001, 0.0006); Pou2f3 (*p = 0.032, 0.0007; #p = 0.0006, 0.0004); Muc2 (*p = 0.0296, 0.0466; #p = 0.0214); Retnlb (*p = 0.0016; #p = 0.0019, 0.0018). l Representative FACS plots of H. poly-infected mice analyzed for IL17RB+CD90+ ILC2s gated from Lin−CD45+CD3−CD4− cells in the intestinal lamina propria. n = 4 mice/group. m qPCR analysis of type 2 cytokines expression in the jejunum of mice infected with H. poly. n = 10, 6, 6, 11 mice/group for Il4, Il5, Il9, Il13, respectively; p = 0.0484 (Il4) and 0.0269 (Il13). n Jejunal IL13 levels were measured by ELISA. n = 7 mice/group; *p = 0.0114; #p = 0.0072. Data are presented as mean ± SEM. All p values were generated by two-tailed unpaired t test. *p < 0.05. In (g, h, k, n), *p < 0.05 vs LoxP naive, #p < 0.05 vs LoxP H. poly. Scale bars, 100 µm. Source data are provided as a Source Data file.

Full size image

In the small intestine (SI), tuft cells sense helminth infections and initiate a type 2 immune response characterized by epithelial remodeling. Heligmosomoides polygyrus (H. poly), which induces type 2 immune responses, is a natural intestinal dwelling parasitic helminth of mice29,30. Indeed, we observed noticeable tuft and goblet cell hyperplasia during infection of mice with H. poly, as shown by qPCR and histological analyses (Supplementary Fig. 2a–d). To further investigate the role of SIRT6 in helminth infection-induced intestinal epithelial remodeling, we infected LoxP and IEC-KO mice with H. poly and analyzed on day 14 p.i., a time point at which adult worms were detected in the SI and eggs were found in the feces31. Both the tuft cell numbers and the expression of tuft cell marker genes including Dclk1, Il25, and Pou2f3 were reduced in the jejunum of naive IEC-KO mice (Fig. 1e, g, k). As expected, H. poly infection led to dramatic tuft cell hyperplasia in the jejunum of LoxP mice but blunted tuft cell expansion in IEC-KO mice (Fig. 1e, g, k). Although the goblet cell abundance was not significantly altered in naive IEC-KO mice, we found the numbers of goblet cell and the cell markers (Muc2, Retnlb) expression were significantly reduced in the jejunum of H. poly-infected IEC-KO mice compared with LoxP littermates, indicating that H. poly infection-induced goblet cell hyperplasia is compromised when epithelial Sirt6 is ablated (Fig. 1f, h, k). As a result of the defective epithelial remodeling, IEC-KO mice had more adult worms in the SI and passed more eggs into the feces (Fig. 1i, j).

Since intestinal tuft cell-ILC2 circuit mediates epithelial responses to helminth infections, we then performed flow cytometric analysis to determine the frequency of intestinal LP ILC2s in H. poly-infected LoxP and IEC-KO mice. Surprisingly, IEC-KO mice had comparable frequency of ILC2s compared with LoxP mice (Fig. 1l). Notably, H. poly-infected IEC-KO mice had lower mRNA expression of type 2 cytokine genes in the jejunum than LoxP mice, suggesting that LP ILC2s from IEC-KO mice maybe functionally impaired (Fig. 1m). Since IL13 released by ILC2 is critical for intestinal epithelial remodeling in response to type 2 immunity, we further explored the IL13 levels in the jejunum by ELISA. Consistently, compared with LoxP controls, IEC KO mice had reduced jejunal IL13 levels after H. poly infection (Fig. 1n). Collectively, these data demonstrate that SIRT6 is required for tuft cell development and intestinal anti-helminth responses.

To exclude the possibility that the epithelial abnormalities observed in IEC-KO mice were due to the side effects of Cre expression driven by Vil1 promoter, we evaluated the jejunal anti-helminth responses in Sirt6flox/+ mice and heterozygous Sirt6flox/+:VillinCre mice. As expected, Sirt6flox/+:VilCre mice had no apparent changes in the numbers of tuft and goblet cells compared with Sirt6flox/+ mice (Supplementary Fig. 3a–c). Accordingly, worm burden on day 14 after H. poly infection was also comparable in Sirt6flox/+ and Sirt6flox/+:VilCre mice (Supplementary Fig. 3d, e). These results confirm the specificity of SIRT6 in regulating anti-helminth responses in the intestinal epithelium.

Succinate triggers a type 2 immune responses in the SI by activating tuft cells, thereby leading to the epithelial remodeling32,33. To test whether SIRT6 is also required for succinate-induced tuft cell expansion, we supplemented mouse drinking water with 150 mM succinate for 7 days and measured tuft cell hyperplasia in the jejunum of LoxP and IEC-KO mice. Succinate treatment induced a significant increase in the numbers of both tuft and goblet cells in WT mice, however, Sirt6 deletion in IECs compromised the succinate-induced tuft and goblet cell hyperplasia (Supplementary Fig. 4a–c).

IL4&IL13 treatment is not sufficient to rescue the defective H. poly infection-induced tuft and goblet cell hyperplasia in IEC-KO mice

As we shown above, tuft cells, ILC2s and epithelial progenitors comprise a response circuit that mediates intestinal epithelial remodeling associated with type 2 immunity. To explore to what extent the defective type 2 immunity caused by diminished tuft cell abundance contributes to the impaired anti-helminth responses observed in IEC-KO mice, we treated H. poly-infected LoxP and IEC-KO mice with recombinant murine IL4/IL13 mixture for 8 days (started on 6 d.p.i.) and assessed the phenotypes of epithelial remodeling and worm clearance. Importantly, rIL4/rIL13 treatment greatly increased the resistance of mice to H. poly infection and largely rescued the impaired ability to expel H. poly worms in IEC-KO mice, as illustrated by comparable numbers of adult worms in the lumen and eggs in the feces between the two genotypes (Fig. 2a, b). In contrast, rIL4/rIL13 treatment did not rescue the blunted tuft and goblet cell expansion following H. poly infection in IEC-KO mice (Fig. 2c–g). These data suggest that epithelial Sirt6 ablation impairs the activity of IL4/IL13 signaling pathway in IECs and thereby inhibits the type 2 immunity-induced intestinal epithelial remodeling. STAT6, a transcription factor mediating type 2 immune responses, is found in the cell in its latent form. Upon cytokine stimulation it is phosphorylated by the Jak kinases at Tyr641, this leads to dimerization and translocation of STAT6 to the nucleus to induce transcription14. Indeed, we observed significantly reduced Y641 phosphorylation of STAT6 in the jejunal epithelium of H. poly-infected IEC-KO mice in the absence or presence of rIL4/rIL13 treatment, implying a potential role of SIRT6 in IL4/IL13-STAT6 pathway regulation in IECs (Fig. 2h, i).

Fig. 2: Recombinant IL4&IL13 treatment fails to rescue the impaired tuft and goblet cell hyperplasia caused by epithelial Sirt6 ablation.
figure 2

3-4-month-old male LoxP and IEC-KO mice were injected (i.p.) daily with vehicle or rIL4&rIL13 (1ug/mouse) from day 6 after H. poly infection and subjected to the following analyses on day 14 post-infection. a, b Analysis of parasite burden by quantification of adult worms in intestinal lumen (a) and eggs in feces (b). n = 5 mice/group; form left to right, *p = 0.019, 0.0038, 0.0038; #p = 0.004 in (a); *p = 0.03, 0.0466; #p = 0.0092 in (b). c qPCR analysis of tuft and goblet cell markers expression in the jejunal IECs. n = 5 mice/group; form left to right, Dclk1 (*p = 0.0125, 0.021; #p = 0.0092, 0.0348), Trpm5 (*p = 0.0058, 0.0288, 0.0422; #p = 0.0144), Retnlb (*p = 0.0368; #p = 0.0235). d Jejunal tuft cells were examined by IHC staining with anti-DCLK1 antibody (200X). e Quantification of tuft (DCLK1 positive) cells shown in (d). n = 5 mice/group; 20 crypt-villus units counted for each mouse; from left to right, *p = 0.0104, 0.0021; #p = 0.0002, 0.0008. f Jejunal goblet cells were examined by Alcian blue staining (200X). g Quantification of goblet (Alcian blue positive) cells shown in (f). n = 5 mice/group; 20 crypt-villus units counted for each mouse; from left to right, *p = 0.0149, <0.0001; #p < 0.0001, 0.0003. h, i Western blot (h) and quantification (i) analyses of P-STAT6 (Y641) and SIRT6 in the jejunal IECs. n = 3, 3, 4, 4 mice/group, respectively; from left to right, P-STAT6 (*p = 0.0159; #p = 0.0136, 0.0161), SIRT6 (*p = 0.0118, 0.0113; #p = 0.0099, 0.0102). Data are presented as mean ± SEM. All p values were generated by two-tailed unpaired t test. *p < 0.05 vs LoxP-Veh, #p < 0.05 vs LoxP-rIL4&13. Scale bars, 100 µm. Source data are provided as a Source Data file.

Full size image

Deficiency of SIRT6 in intestinal organoids leads to compromised IL13-induced tuft and goblet cell expansion

To further explore whether SIRT6 regulates helminth infection-induced epithelial remodeling in an epithelial cell autonomous manner, we used an ex vivo organoid culture system that allows physiological responses of isolated intestinal epithelium to be analyzed in the absence of stromal cues. We set up intestinal organoid cultures from LoxP and IEC-KO mice and treated the organoids with vehicle or rIL13. Sirt6 deletion in organoids resulted in significantly reduced mRNA expression of tuft cell markers (Dclk1, Trpm5, Pou2f3) in both vehicle- and rIL13-treated conditions (Fig. 3a). Consistent with the in vivo data, we also found noticeably downregulated expression of goblet cell markers (Muc2, Retnlb) in rIL13-treated IEC-KO organoids (Fig. 3a). To further validate our qPCR results, we stained vehicle- or rIL13-treated organoids with DCLK1 or MUC2 antibodies. Consistently, vehicle-treated IEC-KO organoids had fewer tuft cells than LoxP organoids. As expected, treatment of IEC-KO organoids with rIL13 triggered significantly less tuft cell hyperplasia than that detected in LoxP organoids (Fig. 3b, c). Although the goblet cell numbers were comparable in vehicle-treated LoxP and IEC-KO organoids, we observed a significant reduction of IL13-induced goblet cell expansion in IEC-KO organoids compared with LoxP organoids (Fig. 3d, e).

Fig. 3: Loss of SIRT6 in intestinal organoids causes defective IL13-induced tuft and goblet cell hyperplasia.
figure 3

LoxP and IEC-KO intestinal organoids were treated with vehicle or rIL13 (25 ng/ml) for 48  h and then subjected to the following analyses. a qPCR analysis of IEC markers expression. n = 4 biological replicates/group; from left to right, Dclk1 (*p = 0.0001, 0.0014, 0.0001; #p = 0.0012, 0.01); Trpm5 (*p < 0.0001, <0.0001, 0.0018; #p < 0.0001, 0.0015); Pou2f3 (*p = 0.0303, 0.0018, <0.0001; #p = 0.0018, 0.017); Muc2 (*p < 0.0001, 0.002; #p < 0.0001, 0.0154); Retnlb (*p = 0.0087, 0.0013; #p = 0.0087); Slc5a1 (*p = 0.0019, 0.0116; #p = 0.0032); Lyz1 (*p = 0.049, 0.0339; #p = 0.0008); Lgr5 (*p = 0.0265, 0.0326; #p = 0.0082). b, d Tuft and goblet cells were labeled by anti-DCLK1 (b) and anti-MUC2 (d), respectively, in frozen sections of organoids (200X). c Quantification of tuft cells shown in (b). n = 15 organoids/group; from left to right, *p = 0.0301, <0.0001, 0.001; #p < 0.0001 for both. e Quantification of goblet cells shown in (d). n = 15 organoids/group; from left to right, *p < 0.0001, 0.0176; #p < 0.0001, 0.0032. Data are presented as mean ± SEM. All p values were generated by two-tailed unpaired t test. *p < 0.05 vs LoxP-Veh, #p < 0.05 vs LoxP-rIL13. Scale bars, 50 µm. Source data are provided as a Source Data file.

Full size image

Epithelial deletion of Sirt6 causes defective STAT6 activity

To uncover the functional link between SIRT6 and IL4/IL13-STAT6 pathway, we first analyzed the expression of SIRT6 and phosph-STAT6 (Y641) in the jejunal epithelium of mice infected with H. poly for different time intervals. Immunostaining analysis demonstrated that SIRT6 and P-STAT6 (Y641) were both primarily localized in the nucleus of crypt epithelial cells (Fig. 4a, b). Importantly, along with the rapid induction of STAT6 (Y641) phosphorylation, H. poly infection noticeably increased and sustained SIRT6 expression (Fig. 4a, b). It is worthy to note that the specificity of SIRT6 and P-STAT6 (Y641) antibodies was validated by immunostaining of the jejunum tissues from IEC-KO and Stat6-/- mice, respectively (Supplementary Fig. 5a, b). Consistently, western blot results revealed that the protein levels of SIRT6 and P-STAT6 (Y641) were markedly increased on day 6 p.i. and kept high levels to day 14 p.i., indicative of the potential involvement of SIRT6 in regulating anti-helminthic STAT6 activity (Fig. 4c, d). We went on to examine whether Sirt6 deletion leads to impaired STAT6 phosphorylation. As expected, STAT6 (Y641) phosphorylation was significantly reduced in the jejunal IECs of IEC-KO mice in both naive and H. poly-infected states (Fig. 4e–g). In addition, despite the absence of detectable STAT6 (Y641) phosphorylation in vehicle-treated intestinal organoids (probably due to the lack of type 2 cytokines in the organoid culture system), IEC-KO organoids had significantly reduced IL13-induced STAT6 (Y641) phosphorylation compared with LoxP organoids (Fig. 4h, i).

Fig. 4: IEC ablation of Sirt6 attenuates the Tyr641 phosphorylation of STAT6.
figure 4

a–d 2-month-old C57BL/6J male mice were infected with H. poly and analyzed on indicated days post-infection. a, b Immunostaining for SIRT6 (a) and P-STAT6 (Y641) (b) in the jejunum. c, d Western blot (c) and quantification (d) analyses of SIRT6 and P-STAT6 (Y641) in the jejunal IECs. n = 3 mice/group; p = 0.0102 (6 dpi), 0.006 (9 dpi), 0.0203 (14 dpi) for SIRT6; p = 0.0259 (6 dpi), 0.0025 (9 dpi), 0.0089 (14 dpi) for P-STAT6. e–g 2–3-month-old male naive and H. poly-infected LoxP and IEC-KO mice were subjected to the following assays on day 14 post-infection. e Immunostaining for P-STAT6 (Y641) in the jejunum. f, g Western blot (f) and quantification (g) analyses of P-STAT6 (Y641) in the jejunal IECs. n = 3 mice/group; from left to right, *p = 0.0166, 0.0423; #p = 0.0177, 0.0136. h, i Western blot (h) and quantification (i) analyses of P-STAT6 (Y641) in vehicle- or rIL13-treated (25 ng/ml, 48 h) organoids. n = 3 biological replicates/group; p = 0.0451. Data are presented as mean ± SEM. All p values were generated by two-tailed unpaired t test. In d, *p < 0.05 vs naive; In g, *p < 0.05 vs LoxP-naive, #p < 0.05 vs LoxP-H. poly; Scale bars, 200X, 100 µm; 400X, 25 µm. Source data are provided as a Source Data file.

Full size image

Furthermore, we investigated the effects of SIRT6 on the transcriptional activity of STAT6 using luciferase reporter assay in HEK293T cells. It should be pointed out that HEK293T cells lack endogenous STAT6, but they express all other components of the IL4/IL13 signaling and are able to activate an IL4/IL13 responsive promoter upon ectopic expression of STAT634. Using a luciferase reporter containing two tandem copies of STAT6 binding sites, we found that SIRT6 overexpression promoted, while SIRT6 knockdown suppressed the transcriptional activity of STAT6 in the absence or presence of rIL13 stimulation (Fig. 8t, u). Together, these results suggest that SIRT6 is able to augment STAT6 activity by increasing IL13-induced Tyr641 phosphorylation.

Transgenic mice with IEC-specific overexpression of constitutively activated STAT6 exhibits enhanced tuft and goblet cell differentiation

To directly explore the role of STAT6 in intestinal epithelium, we generated transgenic mice that express a constitutively active form of mouse STAT6 (STAT6vt), driven by mouse Vil1 gene promoter (TgSTAT6vt) (Fig. 5a). STAT6vt contains two amino acid mutations (V547A/T548A) in the SH2 domain resulting in Jak kinase independent Tyr641 phosphorylation35. We obtained 3 transgenic mouse lines (line-10, line-13 and line-18) expressed transgene STAT6vt at different levels. qPCR analysis of the jejunal IECs identified line-18 as the highest expresser of STAT6vt (~22-fold increase), followed by line-13 (~7-fold increase) and line-10 (~2-fold increase) (Fig. 5c). The protein levels of STAT6vt in these 3 transgenic lines were also confirmed by western blot analysis (Fig. 5b). Importantly, STAT6vt was only expressed in the intestine but not in other tissues including stomach, liver, lung, heart, and kidney of line-18 mice, indicating an intestinal epithelial-specific expression pattern of STAT6vt (Supplementary Fig. 6a). As we showed above, line-13 transgenic mice exhibited moderate epithelial expression of STAT6vt and no outwardly apparent phenotype, we therefore selected transgenic line-13 for further study (hereafter referred to as TgSTAT6vt).

Fig. 5: IEC-specific overexpression of constitutively activated STAT6 (STAT6vt) promotes tuft and goblet cell differentiation.
figure 5

2-month-old male WT and TgSTAT6vt mice (including line-10, line-13 and line-18) were subjected to the following assays. a The strategy for creating TgSTAT6vt mice. b, c Western blot (b) and qPCR analyses (c) of TgSTAT6vt expression in the jejunal IECs. n = 7 mice/group in (c); p = 0.0013 (line-10), <0.0001 (line-13), <0.0001 (line-18). d Representative H&E images of jejunum tissues. e Analyses of jejunal villus length and crypt depth. n = 30 villi or crypts/group; p < 0.0001 (villus length, crypt depth). f qPCR analysis of IEC markers expression in the jejunal IECs. Dclk1, Lyz1, n = 4 mice/group; Trpm5, Retnlb, n = 5 mice/group; Slc5a1, n = 5 (WT) and 4 (Tg) mice; Chga, Lgr5, n = 4 (WT) and 3 (Tg) mice; p = 0.0259 (Dclk1), 0.0044 (Trpm5), 0.0491 (Retnlb), 0.0322 (Lyz1). g Tuft and goblet cells were examined by DCLK1 immunostaining and Alcian blue staining, respectively in the jejunum (200X). (h) Quantification of tuft and goblet cells shown in (g). n = 4 mice/group; 20 crypt-villus units counted for each mouse; p = 0.0018 (tuft cell) and 0.0126 (goblet cell). i Tuft and goblet cells were labeled by anti-DCLK1 and anti-MUC2, respectively, in frozen sections of organoids (200X). j Quantification of tuft and goblet cells shown in (i). n = 10 organoids/group; p < 0.0001 (tuft cell, goblet cell). Data are presented as mean ± SEM. All p values were generated by two-tailed unpaired t test. *p < 0.05. Scale bars, 100 µm in (d, g); 50 µm in (i). Source data are provided as a Source Data file.

Full size image

Intriguingly, TgSTAT6vt mice had reduced jejunal villus height and crypt depth compared with WT controls (Fig. 5d, e). Furthermore, qPCR analysis revealed a significantly augmented expression of genes encoding tuft (Dclk1, Trpm5) and goblet (Retnlb) markers in TgSTAT6vt mice compared with WT mice, indicating STAT6 in IEC promotes tuft and goblet cell differentiation (Fig. 5f). In contrast, we observed a significant reduction of Lyz1 expression in the jejunal IECs of TgSTAT6vt mice (Fig. 5f). No changes in the expression of genes encoding Slc5a1 (enterocyte), Chga (enteroendocrine cell), and Lgr5 (stem cell) were detected (Fig. 5f). To exclude the possibility that intestinal epithelial phenotype observed in TgSTAT6vt mice was caused by the side effects of transgene random insertion, we also examined the mRNA expression of IEC lineage markers in the jejunal IECs of line-18 mice. Consistently, mRNA levels of Dclk1 and Retnlb in line-18 mice were significantly higher than those in WT mice, while Lyz1 mRNA expression was dramatically lower (Supplementary Fig. 6d). Interestingly, while the epithelial expression of Slc5a1, Chga, and Lgr5 were comparable in WT and TgSTAT6vt (line-13) mice, their levels were significantly lower in line-18 mice than in control mice (Fig. 5f, Supplementary Fig. 6d). This discrepancy is likely attributed to the different expression levels of transgene STAT6vt between these 2 mouse lines. Notably, we also observed a more dramatic reduction of jejunal villus height in line-18 mice, indicating epithelial STAT6 directly regulates villus development in the SI (Supplementary Fig. 6b, c).

To further analyze whether augmented expression of cell markers correlated with the increased tuft and goblet cell numbers in TgSTAT6vt mice, we stained jejunum tissues for DCLK1 and Alcian blue. As expected, TgSTAT6vt mice had more tuft and goblet cells in the jejunal epithelium than WT mice (Fig. 5g, h). To test if tuft and goblet cell expansion is an IEC-intrinsic regulation by STAT6, we examined the tuft and goblet cell abundance in the intestinal organoids. Indeed, TgSTAT6vt organoids had significantly more tuft and goblet cells than WT organoids, revealing an epithelium-autonomous effect of STAT6 in tuft and goblet cell differentiation (Fig. 5i, j).

IEC STAT6vt overexpression rescues impaired intestinal anti-helminth responses in IEC-KO mice

To better understand the transcriptional program regulated by SIRT6 in the intestinal epithelium, we conducted RNA-seq analysis on the jejunal IECs from LoxP and IEC-KO mice. Among the 188 significantly differentially expressed genes (log2(fold-change)>1.25, and p < 0.05), 58 genes were downregulated and 130 genes were upregulated (Supplementary Fig. 7a). Gene Set Enrichment Analysis (GSEA) revealed that both tuft-1 and tuft-2 identity gene sets were enriched in downregulated genes by Sirt6 deficiency, again supporting that epithelial SIRT6 is critical for intestinal tuft cell development (Supplementary Fig. 7b)36. Notably, unbiased gene ontology analysis revealed that pathways like positive regulation of inflammatory response, anatomical structure homeostasis and biosynthesis of specialized pro-resolving mediators (SPMs) were downregulated while pathways involved in leukocyte activation, regulation of defense response, and regulation of leukocyte cell-cell adhesion were upregulated (Supplementary Fig. 7c). Furthermore, to gain insight into the underlying molecular mechanisms of STAT6 regulated IEC development, we also performed RNA-seq analysis of the jejunal IECs from TgSTAT6vt mice and WT littermates (Supplementary Fig. 7d). When compared the gene expression changes (log2(fold-change) from DESeq2) of IEC-KO (normalized to LoxP) and TgSTAT6vt (normalized to WT), we found that many TgSTAT6vt downregulated genes were upregulated in Sirt6 IEC-KO mice (Supplementary Fig. 7e), indicating that STAT6 and SIRT6 may be involved in the same pathway to regulate intestinal epithelial homeostasis.

Since Sirt6 deletion leads to the compromised activation of STAT6 in response to type 2 cytokines in IECs, we hypothesized that overexpression of constitutively activated STAT6vt may rescue the epithelial abnormalities in IEC-KO mice. To test this hypothesis, we crossed IEC-KO mice with TgSTAT6vt mice to generate IEC-KO-TgSTAT6vt mice. Indeed, IEC-KO-TgSTAT6vt mice had significantly more tuft and goblet cells in the jejunal epithelium than IEC-KO mice (Supplementary Fig. 8a, b). This rescue effect was also reflected by the increased expression of Dclk1, Trpm5, and Retnlb in the IECs of IEC-KO-TgSTAT6vt mice (Supplementary Fig. 8c). Furthermore, a significant increase in the numbers of tuft and goblet cells were observed in IEC-KO-TgSTAT6vt organoids (Supplementary Fig. 8d, e).

To further determine whether the impaired anti-helminth responses in IEC-KO mice were due to defective IEC STAT6 activity, we infected WT, TgSTAT6vt, IEC-KO, IEC-KO-TgSTAT6vt mice with H. poly and analyzed epithelial responses and worm burden in these mice on day 14 after infection. The overexpression of STAT6vt and the ablation of Sirt6 in jejunal epithelium were confirmed by western blot analysis (Fig. 6a, b). Following infection, TgSTAT6vt mice had obviously more jejunal tuft and goblet cells than WT mice (Fig. 6d-f). Intriguingly, although the IEC-KO mice had the lowest abundance of tuft and goblet cells among the 4 groups of mice we studied, we observed comparable numbers of jejunal tuft and goblet cells in TgSTAT6vt and IEC-KO-TgSTAT6vt mice (Fig. 6d-f). These findings were also verified by qPCR analysis of tuft (Dclk1, Trpm5) and goblet (Retnlb) cell markers (Fig. 6c). In accord with a critical role of tuft and goblet cell expansion in H. poly worm expulsion, we found that IEC-KO mice had markedly higher worm burden than WT mice, while IEC overexpression of STAT6vt significantly relieved the worm burden in both WT and IEC-KO mice, as shown by fewer adult worms in the SI and eggs in the feces (Fig. 6g, h). In addition, organoids derived from these 4 groups of mice were cultured and stimulated with rIL13. Consistently, we observed comparable numbers of tuft and goblet cells in TgSTAT6vt and IEC-KO-TgSTAT6vt organoids (Fig. 6i–k). Together, these results reveal that overexpression of STAT6vt in IECs remarkably rescued the impaired intestinal type 2 immunity caused by Sirt6 deletion.

Fig. 6: IEC STAT6vt overexpression rescues the defective intestinal type 2 immunity resulting from Sirt6 deletion.
figure 6

2–3-month-old male mice with 4 genotypes (WT, TgSTAT6vt, IEC-KO, IEC-KO-TgSTAT6vt) were infected with H. poly and subjected to the following analyses on day 14 post-infection. a, b Western blot (a) and quantification (b) analyses of P-STAT6 (Y641) and SIRT6 in the jejunal IECs. n = 3 mice/group; from left to right, P-STAT6 (*p = 0.0083; #p = 0.0074, 0.0026); SIRT6 (*p = 0.0358, 0.0344; #p = 0.0002). c qPCR analysis of the expression of Dclk1, Trpm5 and Retnlb in the jejunal IECs. Dclk1, Trpm5, n = 5 mice/group; Retnlb, n = 4 mice/group; from left to right, Dclk1 (*p = 0.0054, 0.0019; #p = 0.0027, 0.0494); Trpm5 (*p = 0.0389, 0.0085; #p = 0.0184); Retnlb (*p = 0.0466; #p = 0.0232, 0.0154). d Tuft and goblet cells were examined by DCLK1 immunostaining and Alcian blue staining, respectively in the jejunum (200X). e, f Quantification of tuft (e) and goblet (f) cells shown in (d). n = 3, 4, 3, 3 mice/group, respectively in (e) and n = 4 mice/group in (f); 30 crypt-villus units counted for each mouse; form left to right, *p = 0.0253, 0.0119; #p = 0.0007, 0.0017 in (e); *p = 0.0059, 0.0264, 0.0104; #p < 0.0001 for both in (f). g, h Analysis of parasite burden by quantification of adult worms in intestinal lumen (g) and eggs in feces (h). n = 5, 4, 5, 4 mice/group; form left to right, *p = 0.0458, 0.0016; #p = 0.0001, 0.0005 in (g); *p = 0.0487, 0.0268; #p = 0.0051, 0.0052 in (h). i Tuft and goblet cells were labeled by anti-DCLK1 and anti-MUC2, respectively, in frozen sections of rIL13-treated organoids (200X). j, k Quantification of tuft (j) and goblet (k) cells shown in (i). n = 10 (j) and 9 (k) organoids/group; form left to right, *p = 0.0307, 0.0003, 0.0281; #p = 0.0003, <0.0001 in (j); *p = 0.0181, 0.0002; #p = 0.0002, 0.0009 in (k). Data are presented as mean ± SEM. All p values were generated by two-tailed unpaired t test. *p < 0.05 vs WT, #p < 0.05 vs IEC-KO. Scale bars, 100 µm in (d); 50 µm in (i). Source data are provided as a Source Data file.

Full size image

IEC overexpression of SIRT6 in mice promotes intestinal type 2 immune responses

To further confirm that SIRT6 indeed modulates type 2 immunity-induced intestinal epithelial remodeling, SIRT6 IEC-specific transgenic mice (IEC-Tg) were generated by crossing transgene SIRT6 floxed mice which encompass a floxed STOP cassette in front of the SIRT6 transgene, with Villin-Cre mice37,1.

RNA sequencing

Total RNAs of isolated jejunal IECs were extracted by using the RNeasy Plus Mini Kit (Qiagen) following the manufacturer’s instructions. RNA concentration and integrity of each sample were measured on an Agilent Bioanalyzer. A cDNA library was prepared and sequenced according to the Illumina standard protocol by Vazyme Biotech. Sequencing reads were mapped to the mouse reference genome (mm10) using Hisat2 2.1.0 with default settings54. Differentially expressed genes were analyzed using counts from HTSeq-count (version 0.13.5)55. Only genes that have at least 10 reads in total were then fed into DESeq2 with masking structural RNAs and non-coding RNAs for calculating fold change56. Genes that are significantly changed were selected by using log2(fold change) > 1.25 for upregulated genes and log2(fold change) < −1.25 for downregulated genes with p value < 0.05 as cut off. The gene ontology analysis was performed using Metascape for up- and downregulated genes57. To generate a ranked gene list for pre-ranked gene set enrichment analysis (GSEA), genes were ranked by their log2(fold-change) in IEC-KO versus LoxP (n = 3)58. The RNA-seq data were deposited at Gene Expression Omnibus (GEO) under accession ID GSE202470 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE202470).

Western blot

Tissues and cells were lysed in RIPA buffer supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF) and Roche cOmplete protease inhibitor cocktail. Protein extracts were separated on an SDS-PAGE gel, transferred to nitrocellulose membranes, and blotted with the indicated primary antibodies at 4 °C for overnight. Horseradish peroxidase (HRP)-conjugated secondary antibodies were given for 1 h. The immune complexes were detected using the ECL detection reagents (Beyotime). Band densitometries were obtained with ImageJ software. Detailed information of the primary antibodies used in this study is listed in Supplementary Table 2.

Histology, immunohistochemistry, and immunofluorescence

Intestinal tissues were coiled into “Swiss rolls” and fixed in 4% paraformaldehyde (PFA), embedded in paraffin and sectioned at a thickness of 5 µm. Organoids were fixed in 4% PFA, embedded in optimal cutting temperature (OCT) compound and sectioned at a thickness of 10 µm. Antigen retrieval was performed in citrate buffer using a pressure cooker. Paraffin sections of intestinal tissues were stained with H&E staining according to standard procedures. Immunohistochemistry (IHC) analysis was performed using IHC detection kit (ZSGB Bio) following manufacture’s guidelines. Alcian blue staining was carried out using Alcian blue stain kit (Vector Labs) following manufacturer’s instructions. For immunofluorescence, tissue slides were blocked with 3% BSA, 0.2% Tween-20 in PBS, incubated with primary antibodies (1:100 to 1:200 dilution) overnight, and secondary antibodies (1:400 dilution) for 1 h. TUNEL assay was carried out using DeadEnd Fluorometric TUNEL System (Promega) following manufacturer’s instructions. Images were taken with a Nikon Eclipse Ni-U microscope and DS-Fi2 color CCD (Nikon) through ImageView software. For cell number quantification, 3–5 random fields per sample were captured. In the intestine, cell number was normalized to the number of crypt/villus units. In the organoid, cell number was normalized to nuclei number as indicated by DAPI staining. Detailed information of the primary antibodies used in this study is listed in Supplementary Table 2.

ELISA

Jejunum (~2 cm) was removed, flushed with saline, cut into several pieces and then lysed in RIPA buffer using Dounce homogenizer. IL13 levels in jejunal tissue lysate were analyzed using Mouse IL-13 DuoSet ELISA (R&D Systems) according to manufacture’s protocol.

Flow cytometry

Intestinal LP lymphocytes were isolated according to an established protocol59. Briefly, freshly harvested jejunum (~10 cm) was flushed, opened longitudinally, and cut into ~5–6 pieces. DTT and EDTA solutions were used sequentially to remove intraepithelial lymphocytes and epithelial cells. The remaining tissue was washed and digested with collagenase for 45 min at 37 °C with shaking. LP lymphocytes were pelleted after filtering the digested tissue through 100 µm filters. Cells were fixed and then stained with relevant antibodies at 4 °C for 30 min. Flow cytometry data were acquired on CytoFLEX (Beckman Coulter) and analyzed with FlowJo (v10.6.2). When gating cells, single cells were selected for viable, lineage negative (CD19-CD11b-CD49b-CD11c-Gr-1-B220-), CD45+ cells. Within the CD3-CD4- subset, ILC2 population is positive for CD90 and IL-17RB. Detailed information of the antibodies used in this study is listed in Supplementary Table 2.

Chromatin immunoprecipitation (ChIP)

NCM460 cells were fixed with 1% formaldehyde at RT for 15 min. Chromatins were lysed and sonicated to an average size of ~250 bp. Immunoprecipitation was performed using anti-FLAG and anti-Acetyl-H3K56 antibodies at 4 °C overnight. Reversion of cross-linking was performed by heating samples at 65 °C for overnight and DNA was purified using phenol-chloroform extraction. ChIP DNA amount for gene promoters of interest was analyzed by real-time PCR and normalized to that of a housekee** gene, ACTB. Sequence information of the primers used in this study is listed in Supplementary Table 1.

Luciferase reporter assay

HEK293T cells were plated onto 24-well plates and co-transfected with p2xSTAT6-Luc2P (luciferase reporter), PGL4.74 (Renilla luciferase), and expression constructs. Transfection was conducted using Lipofectamine 2000 (Invitrogen). 48 h after transfection, cells were treated with vehicle or rIL13 for 2 h. Then, cells were lysed and luciferase activity was measured using the Dual-Luciferase Reporter Assay System and GloMax™ Systems (Promega). Data were presented as relative Firefly luciferase activities normalized to Renilla luciferase activities.

Statistics and reproducibility

All data are presented as mean ± SEM. Data were analyzed using two-tailed unpaired Student’s t-test. p < 0.05 was considered as significant. If not otherwise mentioned, each experiment was repeated independently with similar results at least three times.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.