Histone demethylase JMJD2D protects against enteric bacterial infection via up-regulating colonic IL-17F to induce β-defensin expression

Histone demethylase JMJD2D (also known as KDM4D) can specifically demethylate H3K9me2/3 to activate its target gene expression. Our previous study has demonstrated that JMJD2D can protect intestine from dextran sulfate sodium (DSS)-induced colitis by activating Hedgehog signaling; however, its involvement in host defense against enteric attaching and effacing bacterial infection remains unclear. The present study was aimed to investigate the role of JMJD2D in host defense against enteric bacteria and its underlying mechanisms. The enteric pathogen Citrobacter rodentium (C. rodentium) model was used to mimic clinical colonic infection. The responses of wild-type and JMJD2D-/- mice to oral infection of C. rodentium were investigated. Bone marrow chimeric mice were infected with C. rodentium. JMJD2D expression was knocked down in CMT93 cells by using small hairpin RNAs, and Western blot and real-time PCR assays were performed in these cells. The relationship between JMJD2D and STAT3 was studied by co-immunoprecipitation and chromatin immunoprecipitation. JMJD2D was significantly up-regulated in colonic epithelial cells of mice in response to Citrobacter rodentium infection. JMJD2D-/- mice displayed an impaired clearance of C. rodentium, more body weight loss, and more severe colonic tissue pathology compared with wild-type mice. JMJD2D-/- mice exhibited an impaired expression of IL-17F in the colonic epithelial cells, which restricts C. rodentium infection by inducing the expression of antimicrobial peptides. Accordingly, JMJD2D-/- mice showed a decreased expression of β-defensin-1, β-defensin-3, and β-defensin-4 in the colonic epithelial cells. Mechanistically, JMJD2D activated STAT3 signaling by inducing STAT3 phosphorylation and cooperated with STAT3 to induce IL-17F expression by interacting with STAT3 and been recruited to the IL-17F promoter to demethylate H3K9me3. Our study demonstrates that JMJD2D contributes to host defense against enteric bacteria through up-regulating IL-17F to induce β-defensin expression.


Introduction
Infection of human-specific gastrointestinal pathogens including enteropathogenic Escherichia coli (EPEC) and enterohaemorrhagic E. coli (EHEC) is a leading cause of diarrhoeal illness and mortality worldwide [1,2].EPEC and EHEC form attaching and effacing (A/E) lesions at the intestinal mucosa surface, followed by colonizing the intestinal epithelial cells [3,4].Citrobacter rodentium is a natural mouse-specific A/E enteric pathogen, which shares virulence factors and an infection strategy with human A/E enteric pathogens EPEC and EHEC [5].Therefore, the model of C. rodentium infection of mice is widely used to study pathogenesis, host innate and adaptive immune responses to infection, and pathogen-host-microbiota interactions and immunometabolism [5,6].
In response to C. rodentium infection, the intestine of mouse produces many cytokines to induce antimicrobial peptides to protect host [7][8][9].Among them, IL-17A and IL-17F play an important role.It has been reported that IL-17A and IL-17F protect mice against C. rodentium infection via the induction of inflammatory mediators and β-defensins [8].Some studies have suggested that IL-17A is important for host defense against C. rodentium infection at the early stage, while IL-17F is more critical for controlling C. rodentium infection at the late stage [10].IL-17A and IL-17F belong to IL-17 family that is composed of six members, including IL-17A,
In this study, we found that JMJD2D was up-regulated in the colonic epithelial cells after C. rodentium infection, suggesting that JMJD2D is involved in C. rodentium-induced colitis.We also found that JMJD2D -/-mice were more sensitive to C. rodentium infection compared to wild-type mice.JMJD2D -/-mice display an impaired expression of IL-17F in the colonic epithelial cells, which contributes to eliminate C. rodentium via increasing antimicrobial peptide expression.Furthermore, JMJD2D -/-mice exhibited a reduced expression of β-defensin 1, βdefensin 3, and β-defensin 4 in the colonic epithelial cells.Mechanistically, JMJD2D cooperated with STAT3 to enhance IL-17F expression to subsequently induce β-defensin expression.

Ethics statement
The animal experimental design was approved by the Ethics Committee of the Ethics Committee of Xiamen University (approval number: XMULAC20190040).All experiments were performed in accordance with approved guidelines and related regulations.

Mice
JMJD2D -/-mice on a C57BL/6 background were generated as described previously [16].Wildtype mice were used as controls.Six-week-old male mice were used in all experiments.All experimental procedures were performed in accordance with animal protocols approved by the Laboratory Animal Center of Xiamen University.

C. rodentium infection of mice and quantification of the bacterial burden
C. rodentium strain ATCC51459 was cultured into Luria-Bertani (LB) medium for 12 h at 37˚C and 180 rpm shaking.Mice were administrated with 1.0 × 10 9 CFU C. rodentium in 0.2 ml PBS by oral gavage after fasting for 8 h.The number of live bacteria was determined by serially diluting and spreading on a MacConkey agar plate for 18-24 h at 37˚C.For determination of C. rodentium in feces and tissue samples, feces, colons, feces + cecums, livers, spleens, and lungs were collected, weighed, immersed with 1ml PBS and homogenized in a Tissuelyser-24 for a total of 6 min at 30 Hz at room temperature.Feces and tissue homogenates were serially diluted in PBS and plated in triplicate onto MacConkey agar plates.C. rodentium colonies were identified as pink colonies and counted after an incubation time of 20 h at 37˚C.

PLOS PATHOGENS
Protective role of JMJD2D in enteric bacterial infection

Colon culture
The 0.5 cm distal colon was dissected and washed three times with cold PBS containing streptomycin and penicillin to removed feces.The colon segments were cultured in 1 mL RPMI 1640 medium without FBS but with streptomycin and penicillin for 24 h at 37˚C with 5% CO 2 .Supernatants were harvested at 13,000×g for 10 min at 4˚C.And the supernatants were frozen in -80˚C for further analysis.

Bone marrow chimeric mice
A single lethally irradiated (9 Gry) eight-week-old recipient wild-type mice and JMJD2D -/- mice were reconstituted with 1× 10 7 femoral bone marrow cells of either wild-type mice or JMJD2D -/-mice via intravenous injection 4 h later after irradiation.The chimeric mice were divided into four groups: WT to WT, WT to JMJD2D -/-, JMJD2D -/-to WT, JMJD2D -/-to JMJD2D -/-mice.Reconstituted mice were administered with antibiotics in the drinking water for 3 weeks during the recovery phase.The efficiency of bone marrow transplantation was determined after 4 weeks via digital PCR using peripheral blood genomic DNA and the averaged efficiency of bone marrow transplantation was calculated to exceed 90% (S1 Fig) .Eight weeks after bone marrow transplantation, the chimeric mice were infected with C. rodentium and sacrificed on day 7 after C. rodentium infection.

Cytokines measure
The concentrations of IL-6, IL-1β, and TNF-α in the colon culture supernatants were quantified by Ready-SET-Go ELISA sets (eBioscience) according to manufacturer's instructions.

Histopathology
The distal colon tissues were removed, fixed in 10% neutral formalin and embedded in paraffin.Five-μm sections were prepared and stained with hematoxylin and eosin (H&E).Histological scoring for inflammation and tissue damage was determined as described previously [19].

Immunohistochemistry
Five-micrometer paraffin sections were cut, deparaffinized and rehydrated.Antigens were retrieved by soaking in preheated citrate buffer (pH 6) under microwave heating for 20 min.Endogenous peroxidases were quenched by incubating with 3% H 2 O 2 for 15 min.Nonspecific binding sites were blocked with 5% BSA for 1 h at room temperature.And then the sections were incubated overnight with rabbit anti-C.rodentium antibody at 4˚C.The next day, sections were washed and incubated with EliVision plus kits (Maixin) at room temperature.Peroxidase activity was assessed by DAB reagents.

Immunofluorescence staining
Tissue section preparation and antigen repair were consistent with oboved immunohistochemistry assay.Nonspecific binding sites were blocked with 5% BSA for 1 h at room temperature.And then the sections were incubated overnight with rabbit anti-Ki67 (ab15580, Abcam) antibody at 4˚C overnight.Then sections were washed and incubated with Goat anti-Rabbit IgG (H+L) Alexa Fluor Plus 555 (A32732, Invirogen) at room temperature for 1h.Sections were washed with PBS for three times, and mounted with Solarbio Mounting medium with DAPI (S2110).Images were acquired using Zeiss AxioScan7.

PLOS PATHOGENS
Protective role of JMJD2D in enteric bacterial infection

Isolation of colonic epithelial cells
The colonic epithelial cells were isolated from the colons using a modification of a previously described protocol [19].

Cell lines
The murine colorectal cancer cell line CMT93, and human embryonic kidney cell line HEK293T were obtained from the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China).All cell lines were grown in Dulbecco's modified Eagle's medium (DMEM) supplementing with 10% FBS (ThermoFisher Scientific, Waltham, US) and penicillin/streptomycin in an incubator at 37˚C and 5% CO 2 .

Establishment of stable JMJD2D-knockdown cell lines
Two different JMJD2D shRNA plasmids were purchased from Sigma-Aldrich (St.Louis.US).Control and JMJD2D shRNA plasmids were transfected into CMT93 cells using Lipofectamine 2000 (ThermoFisher Scientific, Waltham, US) according to the manufacturer's instructions.Puromycin (1 μg/mL) was used to select stable JMJD2D-knockdown cell lines.

Luciferase reporter assay
The IL-17F promoter reporter was transfected into CMT93 cells using Lipofectamine 2000 (ThermoFisher Scientific, Waltham, US) according to the manufacturer's instructions.The activities of IL-17F promoter reporter were analyzed by the luciferase reporter assay system (Promega, Madison, US) according to the manufacturer's instructions.The renilla reniformis luciferase reporter acted as an internal control.

Co-IP assays
The interaction between JMJD2D and STAT3 in CMT93 cells was analyzed by the Co-IP assay as described previously [20].The supernatants were immunoprecipitated with 2 μg of the following antibodies or control immunoglobulin G (IgG) with rotation for overnight at 4˚C: anti-JMJD2D (Abcam, Cat#ab93694) and anti-STAT3 (#9139S; CST), and the immunoprecipitants were then used to western blot.

Statistical analysis
All bar charts and statistical analyses were performed using GraphPad Prism 5.All data were shown as the Mean + SEM.The statistically significant effects between mean values (P <0.05) were assessed with the two-tailed Student's t-test.

PLOS PATHOGENS
Protective role of JMJD2D in enteric bacterial infection

JMJD2D -/-mice are more sensitive to C. rodentium infection compared to wild-type mice
To define the role of JMJD2D in C. rodentium-induced colitis, we first detected the expression of JMJD2D in the colonic epithelial cells of wild-type mice orally infected with C. rodentium.As shown in Fig 1A , JMJD2D expression was markedly increased in the colonic epithelial cells isolated from mice on days 7 and 14 after C. rodentium infection, implying that JMJD2D in the colonic epithelial cells may be participated in the clearance of C. rodentium.We then orally infected wild-type and JMJD2D -/-mice with C. rodentium and measured body weight and bacterial burdens in the feces, colons, spleens, lungs, and livers for up to 28 days.JMJD2D -/-mice exhibited more body weight loss on days 10-14 after C. rodentium infection compared to wildtype mice (Fig 1B).Furthermore, JMJD2D -/-mice displayed 10-100 folds higher C. rodentium burdens in the feces, colons, and cecums + feces compared to wild-type mice on days 7, 14, and 21 after infection (Fig 1C).The C. rodentium burdens in the spleens, lungs, and livers of JMJD2D -/-mice were also increased compared to wild-type mice on days 7, 14, and 21 after infection, suggesting that JMJD2D -/-mice exhibit more severe systemic dissemination of C. rodentium compared to wild-type mice.Consistently, immunohistochemical staining of C. rodentium showed that more C. rodentium bacteria, which penetrated deeply into crypts, were observed in the colons of JMJD2D -/-mice compared to wild-type mice on days 7 and 14 after infection (Fig 1D).These results indicated that JMJD2D plays an important protective role in host defense against C. rodentium infection.
C. rodentium infection leads to colonic inflammation and injury [21].A shorter colon length, which is a feature of colonic inflammation and injury, was observed in JMJD2D -/-mice compared to wild-type mice on days 7 and 14 after infection (Fig 1E).H&E staining was performed to score colonic pathology in the colons of JMJD2D -/-mice and wild-type mice without or with C. rodentium infection.Histological analysis of colons from JMJD2D -/-mice infected with C. rodentium showed more serious submucosal inflammation and edema compared to wild-type mice, resulting in significantly higher histological scores in the colons (Fig 1F).Furthermore, the levels of TNF-α and IL-6 in the colonic culture supernatant of JMJD2D -/-mice on day 14 after infection were significantly more than those in the colonic culture supernatants of wild-type mice (Fig 1G).Taken together, JMJD2D -/-mice exhibited more serious colonic pathology and produced more inflammatory cytokines compared to wild-type mice after C. rodentium infection.

JMJD2D expression in non-bone marrow-derived cells controls C. rodentium infection
To determine which JMJD2D-expressing cells are important for the control of C. rodentium infection, we performed reciprocal bone marrow transfer experiments between JMJD2D -/- mice and wild-type mice.Mice were orally infected with C. rodentium 8 weeks after bone marrow transfer and monitored the body weight and C. rodentium burdens in the feces, colons, and spleens for 7 days.The C. rodentium burdens in the feces from WT BM !WT mice were lower than those from WT BM !JMJD2D -/-mice, but were comparable with those from JMJD2D -/-BM!WT mice on days 7 after infection (Fig 2A).Additionally, the C. rodentium burdens in the colons from JMJD2D -/-BM !JMJD2D -/-mice were higher than those from JMJD2D -/-BM !WT mice, but were comparable with those from WT BM !JMJD2D -/- mice on days 7 after infection (Fig 2B).Similar results were also observed in the spleens (Fig 2C).Consistently, anti-C.rodentium staining showed that more C. rodentium bacteria were

PLOS PATHOGENS
Protective role of JMJD2D in enteric bacterial infection observed in the colons from JMJD2D -/-BM !JMJD2D -/-mice compared to those from JMJD2D -/-BM !WT mice, but were comparable with those from WT BM !JMJD2D -/- mice on days 7 after infection (Fig 2D).Furthermore, histological analysis of colons from WT BM !JMJD2D -/-mice and JMJD2D -/-BM !JMJD2D -/-mice infected with C. rodentium showed higher histology score compared to WT BM !WT mice and JMJD2D -/-BM!WT mice (Fig 2E).Taken together, these results indicated that JMJD2D expression in non-bone marrow-derived cells plays a critical role in controlling C. rodentium infection.

JMJD2D deficiency impairs the expression of β-defensins and IL-17F in the colonic epithelial cells during C. rodentium infection
Due to JMJD2D expression in non-bone marrow-derived cells controls C. rodentium infection, we hypothesized that JMJD2D expression in the colonic epithelial cells plays a critical role in controlling C. rodentium infection.To test it, we examined the expression of antimicrobial genes such as Reg family and β-defensins, which play critical roles in controlling C. rodentium infection [7,8], in the colonic epithelial cells of wild-type and JMJD2D -/-mice infected with C. rodentium.As shown in S2 Fig, there was no significant difference in the expression of Reg 3β, Reg 3γ, CRAMP, RELM-β, S100A8, S100A9, and Muc5Ac in the colonic epithelial cells of wild-type and JMJD2D -/-mice infected with C. rodentium; however, JMJD2D deficiency impaired the expression of β-defensin-1, β-defensin-3, and β-defensin-4, but not β-defensin-2, in the colonic epithelial cells of mice without or with C. rodentium infection (Fig 3A -3D).These results indicated that JMJD2D expression in the colonic epithelial cells promotes the expression of β-defensin-1, β-defensin-3, and β-defensin-4 to control C. rodentium infection.
It has been reported that IL-17A and IL-17F can induce the expression of β-defensins [8].Therefore we detected the expression of IL-17A and IL-17F in the colonic epithelial cells of wild-type and JMJD2D -/-mice infected with C. rodentium.The expression of IL-17F in the colonic epithelial cells of JMJD2D -/-mice was significantly reduced compared to wild-type mice on days 0, 7 and 14 after C. rodentium infection (Fig 3E), whereas the expression of IL-17A in the epithelial cells was comparable between wild-type and JMJD2D -/-mice (Fig 3F).Consistent with the mRNA levels of IL-17F, the protein levels of IL-17F in the colonic culture supernatants of JMJD2D -/-mice were markedly decreased compared to those of wild-type mice on days 0, 7, and 14 after C. rodentium infection (Fig 3G).JMJD2D deficiency did not impair the expression of IL-17 receptors IL-17RA and IL-17RC in the colonic epithelial cells (Fig 3H and 3I).Taken together, these results implicated that JMJD2D expression in the colonic epithelial cells may promote the expression of β-defensin-1, β-defensin-3, and β-defensin-4 during C. rodentium infection by enhancing IL-17F expression.

JMJD2D promotes the expression of β-defensins by enhancing IL-17F expression in colorectal cancer cells
To verify that IL-17F can promote β-defensin expression, we transfected IL-17F specific small interfering RNA to knock down IL-17F expression in murine colorectal cancer cell line CMT93 and then examined the effects of IL-17F knockdown on β-defensin expression.As shown in Fig 4A, the expression of β-defensin-1, β-defensin-3, and β-defensin-4 was markedly decreased in IL-17F-knockdown CMT93 cells, demonstrating that IL-17F can promote β-days after C. rodentium infection.(G) The levels of TNF-α and IL-6 in the colonic culture supernatant of JMJD2D -/-mice on day 14 after infection were significantly more than those in the colonic culture supernatants of wild-type mice.The protein levels of TNFα, IL-6 and IL-1β in the colons from JMJD2D -/- mice and wild-type mice (n = 5) on days 7, 14

PLOS PATHOGENS
Protective role of JMJD2D in enteric bacterial infection

Fig 2. JMJD2D expression in non-bone marrow-derived cells controls C. rodentium infection. (A-C)
The C. rodentium burdens in the feces (A), colons (B), and spleens (C) from WT BM !WT mice were lower than those from WT BM !JMJD2D -/-mice, but were comparable with those from JMJD2D -/-BM!WT mice on days 7 after infection.(D) More C. rodentium bacteria were observed in the colons from JMJD2D -/-BM !JMJD2D -/-mice compared to those from JMJD2D -/-BM !WT mice, but were comparable with those from WT BM !JMJD2D -/-mice on days 7 after infection.

JMJD2D promotes IL-17F expression via enhancing STAT3 signaling
To reveal the mechanisms by which JMJD2D promotes IL-17F expression to further induce βdefensin expression in response to C. rodentium infection, we treated wild-type and JMJD2Dknockdown CMT93 cells with heat-killed C. rodentium and then examined the expression of JMJD2D, IL-17F, and β-defensins.We first examined the effect of heat-killed C. rodentium treatment on the expression of JMJD2D.As shown in Fig 5A , heat-killed C. rodentium treatment induced JMJD2D expression.We previously showed that JMJD2D expression could be induced by activating NF-κB signaling [17], we therefore hypothesized that heat-killed C. rodentium treatment may activate NF-κB signaling to induce JMJD2D expression.Indeed, heat-killed C. rodentium treatment could activate NF-κB signaling as demonstrated by increased phosphorylation of p65 (S3A Fig) , suggesting that JMJD2D expression is induced by activated NF-κB signaling in response to C. rodentium infection.We next examined the effect of heat-killed C. rodentium treatment on the expression of IL-17F and β-defensins.As shown in Fig 5B , heat-killed C. rodentium treatment markedly induced the mRNA expression of IL-17F, β-defensin-1, β-defensin-3, and β-defensin-4, but the induction was significantly reduced in JMJD2D-knockdown CMT93 cells.Furthermore, heat-killed C. rodentium induced IL-17F promoter reporter activity about 7-fold in wild-type CMT93 cells, whereas the induction of IL-17F promoter reporter activity was dramatically decreased to 2-fold in JMJD2D-knockdown CMT93 cells (Fig 5C).These results suggested that JMJD2D promotes IL-17F expression at the transcriptional levels.
To find out which transcription factor JMJD2D can cooperate with to promote IL-17F expression, we screened the sequence of mouse IL-17F promoter in Jaspar databases and PROMO databases to search for the potential transcription factor binding sites.We found a transcription factor STAT3 binding site (GTTCTAAGAAA, -1825~-1815) on the IL-17F promoter, implicating that STAT3 may bind to this site to induce IL-17F expression.Additionally, western blotting results showed that heat-killed C. rodentium treatment activated STAT3 signaling as demonstrated by increased phosphorylation of STAT3 in control CMT93 cells, whereas the induction of STAT3 phosphorylation was reduced in JMJD2D-knockdown CMT93 cells (Fig 5D).Similarly, STAT3 phosphorylation was impaired in the colonic epithelial cells of JMJD2D -/-mice on days 7 and 14 after C. rodentium infection (S3B Fig) To determine the role of STAT3 binding site in STAT3-induced IL-17F expression, we performed promoter reporter mutation assays.Co-transfection of JMJD2D and STAT3 could synergistically induce the activity of wild-type IL-17F promoter reporter (with STAT3 binding site), whereas it failed to induce the activity of STAT3 binding site-absent or STAT3 binding site-mutated IL-17F promoter reporter (Fig 5G and 5H).To define whether JMJD2D and STAT3 could be recruited to the STAT3 binding site on the IL-17F promoter without or with heat-killed C. rodentium treatment, we performed ChIP assays.As shown in Fig 5I , STAT3 and JMJD2D could be recruited to the STAT3 binding site on the IL-17F promoter without heat-killed C. rodentium treatment, and the recruitment of JMJD2D and STAT3 on the IL-17F promoter was increased after heat-killed C. rodentium treatment.JMJD2D knockdown reduced the recruitment of JMJD2D and STAT3 on the IL-17F promoter without or with heat-killed C. rodentium treatment (Fig 5I).Additionally, JMJD2D knockdown dramatically elevated H3K9me3 levels on the IL-17F promoter (Fig 5I), suggesting that JMJD2D is responsible for demethylating H3K9me3 on the IL-17F promoter.Taken together, these results suggested that JMJD2D promotes IL-17F expression by enhancing STAT3 signaling via increasing STAT3 phosphorylation and recruitment on the STAT3 binding site on the IL-17F promoter to demethylate H3K9me3.

The demethylase activity of JMJD2D is required for the induction of IL-17F expression
To determine whether the demethylase activity of JMJD2D is required for the induction of IL-17F transcription, we examined the effects of wild-type JMJD2D and demethylase-defective mutant JMJD2D-S200M on STAT3 phosphorylation and the expression of IL-17F and βdefensins in CMT93 cells.Overexpression of wild-type JMJD2D could enhance STAT3 phosphorylation, but overexpression of JMJD2D-S200M couldn't enhance STAT3 phosphorylation (Fig 6A).Overexpression of wild-type JMJD2D could increase the mRNA levels of IL-17F, βdefensin-1, β-defensin-3, and β-defensin-4, but overexpression of JMJD2D-S200M couldn't increase the expression of these genes (Fig 6B).These results suggested that the demethylase activity of JMJD2D is required for promoting the expression of IL-17F, β-defensin-1, β-defensin-3, and β-defensin-4 in CMT93 cells.We next examined the effects of wild-type JMJD2D and demethylase-defective mutant JMJD2D-S200M on IL-17F promoter reporter activity in the absence or presence of STAT3.As shown in Fig 6C , overexpression of JMJD2D increased IL-17 promoter reporter activity in the absence or presence of STAT3, but overexpression of JMJD2D-S200M couldn't increase IL-17 promoter reporter activity.Furthermore, overexpression of JMJD2D decreased H3K9me3 levels on the IL-17F promoter, but overexpression of (MOI = 200) for 2 h activated STAT3 signaling as demonstrated by increased phosphorylation of STAT3 in control CMT93 cells, whereas the induction of STAT3 phosphorylation was reduced in JMJD2D-knockdown CMT93 cells.(E) JMJD2D could interact with STAT3 and the interaction was enhanced by heat-killed C. rodentium treatment.Co-IP analysis of the interaction of JMJD2D and STAT3 in CMT93 cells.(F) Co-transfection of JMJD2D and STAT3 synergistically induced IL-17F promoter reporter activity.(G) Schematic maps of different IL-17F promoter constructs.(H) Co-transfection of JMJD2D and STAT3 synergistically induced the activity of wild-type IL-17F promoter reporter (with STAT3 binding site), whereas it failed to induce the activity of STAT3 binding site-absented or STAT3 binding site-mutated IL-17F promoter reporter.

Discussion
In this study, we disclosed that JMJD2D plays an important protective role in C. rodentiuminduced colitis.Primarily, JMJD2D -/-mice displayed an impaired clearance of C. rodentium and more severe tissue pathology compared with wild-type mice.And JMJD2D -/-mice exhibited an impaired induction of IL-17F, β-defensin-1, β-defensin-3, and β-defensin-4 in the colonic epithelial cells during C. rodentium infection.At the molecular level, JMJD2D cooperates with STAT3 to enhance IL-17F expression to further induce the expression of β-defensin-1, β-defensin-3, and β-defensin-4.It is known that hemopoietic cells including neutrophils, monocytes/macrophages, CD4 + T cells, and dendritic cells play a critical role in clearing C. rodentium infection [22][23][24][25].In addition, nonhemopoietic cells, predominantly the colonic epithelial cells, are also important in control of C. rodentium infection [26].It has been reported that the mRNA levels of JMJD2D were highly expressed in hematopoietic system (bone marrow, lymph nodes, and spleen), hematopoietic cells (dendritic cells and macrophages) [27], and colonic epithelial cells [17].In this study, to determine in which cells JMJD2D contributes to host defense against C. rodentium infection, we performed bone marrow transplantation experiments in wild-type and JMJD2D -/-mice.We found that JMJD2D -/-recipients that received bone marrow from wildtype or JMJD2D -/-mice exhibited more C. rodentium burdens in the colons and spleens compared to wild-type recipient mice that received bone marrow from wild-type or JMJD2D -/- mice on day 7 after C. rodentium infection, suggesting that JMJD2D in nonhemopoietic cells, most likely in colonic epithelial cells, is required for clearance of C. rodentium.
Innate immunity, including the intestinal barrier comprised of mucus layer, epithelial cells, and the nonepithelial cell (most notable leukocytes) [28], is the first line of defense against pathogens.Antimicrobial peptides secreted by intestinal epithelial cells, which can directly kill or inactivate bacteria, are key effectors of innate immunity [29,30].Some antimicrobial peptides, including defensins, cathelicidins, and regenerating islet-derived protein (REG) family, have been found to contribute to host defense against C. rodentium infection [7,30,31].Notably, β-defensins play a key role in host defense against C. rodentium infection [30,32].In this study, we found that the C. rodentium burdens in JMJD2D -/-mice were higher than those in wild-type mice on days 7 and 14 after C. rodentium infection, and meanwhile the expression of β-defensin-1, β-defensin-3, and β-defensin-4 in the colonic epithelial cells of JMJD2D -/-mice was significantly decreased compared with wild-type mice.Thus, we propose that higher C. rodentium burdens in JMJD2D -/-mice are caused by a reduced innate immunity response, such as expression of β-defensin-1, β-defensin-3, and β-defensin-4 after C. rodentium infection, which lead to more severe colonic injury.
IL-17F-deficient mice are susceptibility to C. rodentium infection due to the defect in βdefensin production [8], indicating that IL-17F are required for the control of C. rodentium infection.In addition to innate immune cells, IL-17F is highly expressed in colonic epithelial cells [8].Because our data showed that JMJD2D in nonhemopoietic cells was important for host defense against C. rodentium infection, so we focus on studying the expression of IL-17F in the colonic epithelial cells.When JMJD2D -/-mice displayed more C. rodentium burdens compared with wild-type mice on days 7 and 14 after C.rodentium infection, we observed a decreased expression of IL-17F and β-defensins in the colonic epithelial cells of JMJD2D -/- mice.Thus, impaired IL-17F and β-defensin production in the colonic epithelial cells contributes to the sensitivity of JMJD2D -/-mice to C. rodentium.
It has been reported that the transcription factor STAT3 in the colonic epithelial cells can be activated by C. rodentium infection, and intestinal epithelial STAT3 is essential for the clearance of C. rodentium infection [33].Our recent study has demonstrated that JMJD2D can serve as a coactivator for STAT3 to enhance its transcriptional activity [34].Implied by these information, when we found a STAT3 binding site on the IL-17F promoter, we hypothesized that JMJD2D may cooperate with STAT3 to induce IL-17F expression.Indeed, our current data showed that JMJD2D could be recruited to the STAT3 binding site on the IL-17F promoter and JMJD2D could cooperate with STAT3 to enhance IL-17F transcription, indicating that JMJD2D regulates IL-17F expression at the transcriptional level through activation of STAT3 signaling.In the current study, we also observed JMJD2D deficiency reduced STAT3 phosphorylation without or with C. rodentium infection.The possible mechanism could involve JMJD2D-mediated regulation of upstream signaling pathways that converge on STAT3 phosphorylation.Our recent study has demonstrated that JMJD2D can elevate IFNGR1 expression, which modulated STAT3 phosphorylation in CMT93 cells [34].
Our previous study has shown that JMJD2D -/-mice are more sensitive to dextran sulfate sodium-induced colitis (DSS) due to the defect in the colonic regeneration, which is characterized by an increased in colonic epithelial cells apoptosis and a decrease in colonic epithelial cell proliferation via impaired Hedgehog signaling activation [17], indicating that JMJD2D plays an important protective role in DSS-induced colitis.In the current study, we also observed lower body weight and shorter colon length in JMJD2D -/-mice compared to wild-type mice on days 7 and 14 after infection (Fig 1B and 1E).Considering the important role of IL-23/IL-22 axis in the initiation of immune response, especially crypt proliferation and regenerative repair, we detected the expression levels of IL-23 and IL-22 in colon and isolated intestinal epithelial cells.No significant differences were observed in the expression of IL-22 and IL-23 in the colons of wild-type and JMJD2D -/-mice on days 7 and 14 after C. rodentium infection.On day 14 after C. rodentium infection, JMJD2D deficiency led to impaired IL-22 expression in colonic epithelial cells.No significant difference was observed in IL-23 expression in colonic epithelial cells of wild-type and JMJD2D -/-mice infected with C. rodentium (S4A Fig) .Additionally, we employed Ki67 staining to define the regenerative crypt hyperplasia response.The results revealed comparable numbers of Ki67 + cells per crypt between wild-type mice and JMJD2D -/-mice without or with C. rodentium infection (S4B Fig) .This might be attributed to the peak expression of IL-22 occurring on the third day after infection [7], while differences in regenerative response occurred in the early stages following infection.Furthermore, we treated wild-type and JMJD2D-knockdown CMT93 cells with IL-22 and then examined the expression of JMJD2D, p-STAT3, and Reg3β, as well as the recruitment of STAT3 on Reg3β promoter in wild-type and JMJD2D-knockdown CMT93 cells without or with IL-22 treatment.JMJD2D knockdown reduced IL-22-induced STAT3 phosphorylation and Reg3β expression, as well as reduced the recruitment of STAT3 on the Reg3β promoter (S4C-S4E Fig) , suggesting that JMJD2D may play a crucial regulatory role in the IL-22-STAT3-Reg3β-mediated damage repair, and the specific role and molecular mechanisms remain to be elucidated, and we propose this as a new topic for future investigation.Our current study mainly focused on the protective role of JMJD2D in C. rodentium infection via increasing β-defensin production by upregulation of IL-17F expression.Therefore, our findings suggest that JMJD2D plays essential protective roles in intestine in response to different stimulations via different mechanisms.
Based on the current and previous studies, we propose a working model for the role of JMJD2D in protecting colonic epithelial cells from C. rodentium infection: C. rodentium infection activates NF-κB signaling to induce JMJD2D expression; on one hand, JMJD2D activates STAT3 signaling by enhance STAT3 phosphorylation; on the other hand, JMJD2D cooperates with STAT3 to promote IL-17F expression to further induce the expression of antimicrobial

Fig 6 .
Fig 6.The demethylase activity of JMJD2D is required for the up-regulation of IL-17F.(A) Overexpression of wild-type JMJD2D could enhance STAT3 phosphorylation, but overexpression of JMJD2D-S200M could not.(B) Overexpression of wild-type JMJD2D could increase the mRNA levels of IL-17F, βdefensin-1, β-defensin-3, and β-defensin-4, but overexpression of JMJD2D-S200M could not.(C) Overexpression of JMJD2D increased IL-17 promoter reporter activity in the absence or presence of STAT3, but overexpression of JMJD2D-S200M could not.(D) Overexpression of JMJD2D decreased H3K9me3 levels on the IL-17F promoter, but overexpression of JMJD2D-S200M could not.Data are shown as mean±SEM.Results are representative of three independent experiments.***P<0.001.https://doi.org/10.1371/journal.ppat.1012316.g006 peptide β-defensins to eliminate C. rodentium by interacting with STAT3 and being recruited to the IL-17F promoter to demethylate H3K9me3 (Fig 7).S4 Fig. (A) No significant differences were observed in the expression of IL-22 and IL-23 in the colons of wild-type and JMJD2D -/-mice on days 7 and 14 after C. rodentium infection.On day 14 after C. rodentium infection, JMJD2D deficiency led to impaired IL-22 expression in colonic epithelial cells.No significant difference was observed in IL-23 expression in colonic epithelial cells of wild-type and JMJD2D -/-mice infected with C. rodentium.(B) Ki67 staining revealed comparable numbers of Ki67 + cells per crypt between wild-type mice and JMJD2D -/- mice without or with C. rodentium infection.(C, D) JMJD2D knockdown reduced IL-22-induced STAT3 phosphorylation (C) and Reg3β expression and (D).(E) JMJD2D knockdown reduced the recruitment of STAT3 on the Reg3β promoter without or with IL-22 treatment.Results are representative of three independent experiments.**P<0.01;***P<0.001.(TIF) S1 Data.Excel spreadsheet containing, in separate sheets, the underlying numerical data and statistical analysis for Figure panels.(XLSX)