PHD3 Stabilizes the Tight Junction Protein Occludin and Protects Intestinal Epithelial Barrier Function*

Background: Prolyl hydroxylases (PHDs) are linked to inflammatory bowel diseases (IBD); however, the exact role of PHD3, a member of PHD family, in IBD is unknown. Results: Deletion of Phd3 in mice intestinal epithelial cells causes spontaneous colitis. Conclusion: PHD3 protects the intestinal epithelial barrier function. Significance: The results are helpful for the development of therapeutic strategies for IBD. Prolyl hydroxylase domain proteins (PHDs) control cellular adaptation to hypoxia. PHDs are found involved in inflammatory bowel disease (IBD); however, the exact role of PHD3, a member of the PHD family, in IBD remains unknown. We show here that PHD3 plays a critical role in maintaining intestinal epithelial barrier function. We found that genetic ablation of Phd3 in intestinal epithelial cells led to spontaneous colitis in mice. Deletion of PHD3 decreases the level of tight junction protein occludin, leading to a failure of intestinal epithelial barrier function. Further studies indicate that PHD3 stabilizes occludin by preventing the interaction between the E3 ligase Itch and occludin, in a hydroxylase-independent manner. Examination of biopsy of human ulcerative colitis patients indicates that PHD3 is decreased with disease severity, indicating that PHD3 down-regulation is associated with progression of this disease. We show that PHD3 protects intestinal epithelial barrier function and reveal a hydroxylase-independent function of PHD3 in stabilizing occludin. These findings may help open avenues for developing a therapeutic strategy for IBD.

Inflammatory bowel disease (IBD) is a chronic inflammatory condition with severe pathology and limited therapeutic options (14,15). The major types of IBD are ulcerative colitis (UC) and Crohn disease. The mechanism that underlies the pathogenesis of IBD is poorly understood. A common feature of IBD is the loss of intestinal epithelial barrier function, leading to inflammatory response and barrier disruption (16). Intestinal epithelial cells establish and maintain the barrier that functions to separate the internal tissues from luminal stuffs. To form an intact epithelial cell layer, the paracellular pathway between cells must be sealed. This is mediated by the junction complex, which is composed of the tight junction (TJ) and subjacent adherens junction (17). TJ is the rate-limiting step in transepithelial transport and the principal determinant of intestinal permeability. Increased intestinal permeability and disrupted TJ may lead to bowel inflammation (17,18). Occludin is the first identified TJ protein and contributes to TJ stabilization and optimal barrier function (17,19).
Recent studies indicate that prolyl hydroxylase activity can affect the development of inflammation. It was demonstrated that the PHD inhibitors could ameliorate experimental colitis in mice (20,21), and PHD1 was demonstrated to regulate intes-tinal epithelial cell apoptosis in the inflamed colon (22). Administration of PHD inhibitors represents a potential therapeutic approach for chronic inflammatory disease (23)(24)(25). Although PHDs are found to be linked to inflammation, the exact role of each member of PHD family in development of IBD remains elusive. We demonstrate in this manuscript that genetic ablation of Phd3 in IECs leads to spontaneous colitis in mice, and PHD3 deficiency in IECs decreases occludin level with defect of barrier function, indicating that PHD3 is a molecule defending against colitis. Further, we find that low expression level of PHD3 is correlated with high disease severity of human UC, implying that down-regulation of PHD3 is associated with progression of UC. Our results suggest that PHD3 plays an important role in maintaining the intestinal epithelial barrier function.

Materials and Methods
Cell Culture and Reagents-293T cells and human colon cancer RKO cells were cultured in DMEM with 10% FBS. The human colon cancer Caco-2 cells were grown in DMEM with 20% FBS. Mouse colon cancer CT26. WT cells were grown in RPMI 1640 medium containing 10% FBS. All media were supplemented with 100 units/ml penicillin and 100 mg/ml streptomycin. The cells were incubated at 37°C, 5% CO 2 . Dimethyloxaloylglycine (DMOG), an inhibitor of PHDs, was purchased from Frontier Scientific. Cycloheximide, and MG132 were from Sigma. Dextran sulfate sodium (DSS) was from MP Biomedicals.
Animals-Phd3 flox/flox (Phd3 ϩ/ϩ ) mice were generated as described (26). The Villin-Cre mice obtained from the Model Animal Research Center of Nanjing University were mated with Phd3 flox/flox mice to generate Phd3 flox/flox :Villin-Cre (Phd3 IEC-KO ) mice. The mice were housed in laboratory cages at 23 Ϯ 3°C with a humidity of 35 Ϯ 5% in a 12-h dark/light cycle with a free access to a regular chow diet (Shanghai Laboratory Animal Co. Ltd., Shanghai, China), under a specified pathogen-free condition. All animals were maintained and used in accordance with the guidelines of the Institutional Animal Care and Use Committee of the Institute of Nutritional Sciences.
Tissue Staining-The tissues of colon were formalin-fixed, paraffin-embedded, and hematoxylin/eosin (H&E)-stained in a regular way. Immunohistochemical staining was as described (10). The immunohistochemistry score is calculated by combining an estimate of percentage of immunoreactive cells (quantity score) with an estimate of staining intensity (intensity score) (27). For the quantity score: no staining, 0; 0 -25% of cells stained, 1; 26 -50%, 2; 51-75%, 3; 76 -100%, 4. For the intensity score: 0, negative; 1, weak; 2, moderate; 3, strong. If there is multifocal immunoreactivity and there are significant differences in staining intensities between foci, the average of the least intense and most intense staining was recorded. immunohistochemistry score ϭ (quantity score ϫ intensity score)/2. The evaluation was done by two double-blinded pathologists, and the average of their results is designated as the final score. The histopathologic grade of inflammation was also scored by clinical pathologists. The biopsies of UC patients were stained after informed consent, and the study protocol was approved by the ethical committee of Zhongshan Hospital.
Preparation and staining of frozen tissue sections were as follows. Tissues were fixed in 4% paraformaldehyde for 2 h, followed by a 10-min wash in PBS at 4°C three times. The tissues were immersed in 30% sucrose in PBS until subsidence. Tissues were embedded in OCT (optimal cutting temperature) and stored at Ϫ80°C. Sections (ϳ7 m) obtained in a cryostat and collected on poly-L-lysine-coated slides were dried in air for 1-2 h and were then rehydrated in PBS (5 min) twice. The slides were incubated in blocking buffer (5% BSA) for 60 min, followed by incubation with the first antibody at 4°C overnight. The slides were washed in PBS (5 min, three times) and then incubated with fluorochrome-conjugated second antibody for 60 min in dark. After washing as described above, the slides were incubated with DAPI for 2 min. The slides were washed, mounted, and covered with coverslips.
DSS Colitis-Sex-and age-matched littermates (9 -11 weeks) received DSS (2.5%) in drinking water for 7 days. Mice body weight was recorded daily. The disease activity index was determined as described (22).
In Vivo Epithelial Permeability Assay-The in vivo intestinal epithelial permeability was determined as described (22). Briefly, the age-matched female littermates were orally administered (0.6 mg/g of body weight) a FITC-dextran solution (70 kDa, 80 mg/ml). After 4 h, the mice were sacrificed, blood was obtained by cardiac puncture, and plasma was separated for determination of FITC by fluorometry at 488 nm. The distribution of FITC-dextran in sectioned colonic tissue was determined by fluorescence microscopy.
In Vitro Permeability Assay-Caco-2 cells were cultured on Transwells with polyester membrane insert (Corning) allowing proper cellular polarization with formation of an apical (upper compartment) and basolateral face (lower compartment). The insert was pretreated with DMEM overnight before cell plating. Caco-2 cells were seeded at a density of 0.5 ϫ 10 5 cells/insert. The medium was replaced with fresh medium every 2 days. After 18 days, the cells were transfected with PHD3 siRNA oligonucleotides for 6 days (28). Briefly, the medium in both upper and lower compartments was replaced with OPTI medium containing 20% FBS. The siRNA oligonucleotides in reagent (Lipofectamine 3000; Invitrogen) were added to the upper compartment. The cells were incubated for 2 days. This was performed three times. Then the medium in both compartments was replaced with OPTI medium containing 20% FBS. Occludin construct mixed in transfection reagent was added to the upper compartment. After 24 h, the medium in both compartments was replaced with fresh DMEM. FITC-dextran (10 kDa, 10 g/ml) was added to the upper compartment, and the cells were incubated at 37°C for 2 h. The concentration of FITCdextran in the bottom compartment was measured in a spectrophotometer (excitation at 485 nm and emission at 530 nm).
Vector Construction and Cell Transfection-PHD3 plasmid was constructed as described (10). PHD3(H196A), PHD3 (H135A), and PHD3(H135A/H196A) vectors were generated by site-directed mutagenesis (10). The vector encoding occludin was constructed to pCMV-flag2B plasmid. The plasmid encoding GST-Itch fusion protein was constructed by inserting PCR-generated DNA fragment encoding Itch into pGEX4T1. The human gene encoding Itch was cloned, and the Itch protein consists of 903 amino acids. Enzymes BamHI and SalI were chosen for linking the human Itch CDS into the pGEX-4T-1 vector. The GST tag is in the NH 2 terminus of the Itch protein.
The construct pET28a-PHD3 for producing His-PHD3 was generated by PCR. Escherichia coli BL21-Gold(DE3)pLysS cells were transformed with pGEX4T1 or pET28a expression vectors and treated with isopropyl-D-thiogalactoside (0.1 mM) for 4 h. The Itch plasmid was from Dr. Lin Li (Institute for Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences).
Cytokine Measurements-The mice colon mucosa was scraped and homogenized in liquid nitrogen. The homogenized scrapings were lysed in radioimmune precipitation assay buffer. The lysates were centrifuged, and the supernatant was collected for determination of cytokines using ELISA kits (R&D) as per the manufacturer's instruction.
Statistical Analysis-The data represent means Ϯ S.E. from three independent experiments except where indicated. Statistic analysis was performed using the unpaired two-tailed Student's t test or two-way analysis of variance at a significance level of p Ͻ 0.05.

Results
IECs-specific Deletion of PHD3 Leads to Spontaneous Colitis in Mice-The Phd3 flox/flox (Phd3 ϩ/ϩ ) and Phd3 flox/flox Villin-Cre (Phd3 IEC-KO ) mice were generated as described (26). The offspring were born at a Mendelian ratio and developed normally. The results of Western blotting indicate that PHD3 was efficiently ablated from the intestinal epithelial cells, but not from cells of other tissues such as small intestinal Peyer's patches and livers of Phd3 IEC-KO mice (Fig. 1A), indicating the tissue specificity of knock-out of PHD3. Compared with Phd3 ϩ/ϩ mice, the body weights of Phd3 IEC-KO littermates were reduced (Fig. 1B). Deletion of PHD3 in IECs decreased colon length of mice (Fig. 1C). H&E staining of colon tissue sections revealed lymphocyte infiltration in colon epithelium of Phd3 IEC-KO mice (Fig. 1D). The ICAM-1 is a marker of immune activation and response, and its expression is increased in the inflamed intestinal mucosa of UC and Crohn disease patients (29). We stained the colon sections with ICAM-1 antibody and found that Phd3 IEC-KO mice had a stronger ICAM-1 staining than Phd3 ϩ/ϩ littermates did (Fig. 1E). We determined by real time PCR the expression of Tnf␣, Il-1␤, and Il-6 in colon mucosa. The transcript levels of these genes in colon mucosa of Phd3 IEC-KO mice are increased (Fig. 1F). The protein levels of these cytokines are also elevated (Fig. 1G). These results suggest that genetic ablation of Phd3 in IECs leads to spontaneous colitis in mice.
PHD3 Deficiency Confers on Mice Higher Susceptibility to DSS-induced Colitis -We determined whether deletion of PHD3 in IECs conferred upon mice more sensitivity to DSS, a toxin that disrupts the intestinal barrier function and induces colitis. When subjected to DSS treatment, the Phd3 IEC-KO mice exhibited more severe wasting ( Fig. 2A), more colon shortening (Fig. 2, B and C), and higher disease activity (Fig. 2D). Histological analyses show that the colon epithelium of Phd3 IEC-KO mice had increased crypt loss, lamina propria collapse, areas of mucosal erosion, and lymphocyte infiltration, as compared with the Phd3 ϩ/ϩ littermates (Fig. 2E). The expression levels of Tnf␣, Il-1␤, and Il-6 in colon mucosa of Phd3 IEC-KO mice are higher than those from control littermates (Fig. 2F). These results suggest that Phd3 IEC-KO mice are more sensitive to DSS challenge.

PHD3 Deficiency Disrupts the Intestinal Epithelial TJ and
Barrier Function-HIF-1␣ and HIF-2␣ are implicated in intestinal inflammation (30 -32). We found that deletion of PHD3 in IECs had little effect on expression of HIF-1␣ and HIF-2␣ in colon epithelial cells (Fig. 2G), which is consistent with recent finding that a single knock-out of PHD family member does not lead to accumulation of HIF␣ in intestinal epithelium (33). These results imply that the PHD3 deficiency-induced colitis might not be through the HIF pathway.
Intact intestinal epithelial TJ is critical in protecting against inflammation (17,18). Disrupted TJ is an important cause of barrier dysfunction and intestinal inflammation (17). Because the Phd3 IEC-KO mice had intestinal inflammation (Fig. 1), we presumed that PHD3 deficiency in IECs influenced intestinal epithelial TJ. To know this, we stained the sections using antibody against occludin, an important TJ protein (19,34). Much less occludin staining was observed in colon epithelium of Phd3 IEC-KO mice (Fig. 3A), indicating a flaw of TJ. Defect of TJ may lead to failure of the epithelial barrier function. We tested this by oral administration of FITC-dextran in mice. Microphotography shows that more FITC-dextran passed through the colon epithelial barrier of Phd3 IEC-KO mice (Fig. 3B), whereas most of the dye was retained at the surface of the barrier of Phd3 ϩ/ϩ littermates, and Phd3 IEC-KO mice had more FITCdextran in serum than Phd3 ϩ/ϩ littermates did (Fig. 3B). We also stained with occludin antibody the colon epithelium of young mice (4 -5 weeks). Similarly, disruption of occludin was observed in colon epithelium of these young Phd3 IEC-KO mice (Fig. 3C). There is little difference in body weights between Phd3 ϩ/ϩ mice and Phd3 IEC-KO littermates (Fig. 3D). The colonic morphology is normal, and there is little lymphocyte infiltration in colon epithelium of the young Phd3 IEC-KO mice (Fig. 3E). The results indicate that a loss of occludin precedes intestinal inflammation. These data suggest that PHD3 deficiency in IECs leads to defects of intestinal epithelial TJ and failure of barrier function.
To confirm that PHD3 regulates intestinal TJ and barrier function, we did in vitro experiments using Caco-2 cells. Knockdown of PHD3 led to reduction of occludin of the cells (Fig. 3F), which indicates a defect of TJ. Occludin plays a critical role in the formation of the TJ seal, and its defect disturbs the TJ permeability barrier (35), facilitating macromolecule flux across the barrier (28). Therefore, we determined in our subsequent experiment the paracellular permeability of monolayer of Caco-2 cells using FITC-dextran. Knockdown of PHD3 enhanced the FITC-dextran level in lower compartment, indicating an increased paracellular permeability of the monolayer (Fig. 3G). Overexpression of occludin suppressed the PHD3 knockdown-enhanced the dye permeation flux (Fig. 3G). Altogether, the results suggest that the PHD3 deficiency-induced decrease of occludin is possibly the cause of defect of intestinal TJ barrier.
Death of IECs is another reason for barrier dysfunction. Knockdown of PHD3 did not lead to cell death of Caco-2 and  AUGUST 14, 2015 • VOLUME 290 • NUMBER 33 RKO cells (Fig. 3H), implying that the PHD3 deficiency-induced barrier function defect is not due to cell death. All of these results suggest that PHD3 functions to maintain the intestinal epithelial homeostasis, probably through ensuring the TJ barrier function.

PHD3 Stabilizes Occludin and Protects Intestinal Barrier
PHD3 Enhances the Stability of Occludin-The aforementioned data show that deletion of PHD3 decreased the protein level of occludin. We tested this in 293T and colon epithelial cells and found that knockdown of PHD3 decreased protein levels of occludin in these cells (Fig. 4A). In this experiment, PHD3 expression was moderately decreased in siRNA-treated cells. Accordingly, the decrease of occludin was modest. In accord with this, overexpression of PHD3 increased the protein level of occludin (Fig. 4B) (the upper bands are possible posttranslational modification of occludin according to data sheet of this antibody). The data suggest that PHD3 up-regulates occludin.
Overexpression of PHD3 had little influence on occludin mRNA levels in Caco-2 or RKO cells (Fig. 4E), indicating that PHD3 induces occludin at a post-transcriptional level. We found that expression of PHD3 prevented degradation of occludin and enhanced the half-life of this protein (Fig. 4F). These results suggest that PHD3 functions to stabilize occludin. In addition to occludin, we also determined the expression of other TJ proteins including ZO-1, ZO-2, ZO-3, claudin-1, and claudin-2. Neither overexpression nor knockdown of PHD3 had effects on expression of these proteins (Fig. 4, G and H).
PHD3 Stabilizes Occludin through Itch-Itch is an E3 ligase of occludin (37). We therefore asked whether PHD3 stabilized occludin through Itch. PHD3 had no effect on the expression of Itch (Fig. 5, A and B). We next determined whether PHD3 interfered with the interaction between Itch and occludin. Knockdown of PHD3 increased (Fig. 5C) and overexpression of PHD3 decreased (Fig. 5D) the association of Itch and occludin, suggesting that PHD3 prevents Itch from binding occludin. Immunoprecipitation experiments show that the endogenous PHD3 has association with Itch (Fig. 5E). Further, His-PHD3 interacted with GST-Itch (Fig. 5F), implying that PHD3 binds Itch directly. We found that PHD3 inhibited the interaction between Itch and occludin in a dose-dependent manner (Fig.  5G). Together, these results suggest that PHD3 binds Itch and prevents Itch from targeting occludin, leading to accumulation of occludin. A working model is proposed (Fig. 5H).
PHD3 Expression Is Inversely Correlated with Severity of UC-Finally, we determined the expression level of PHD3 in colon tissues taken from UC patients. The colonic biopsies from 15 UC patients with different grades of inflammation were examined. The protein levels of PHD3 in samples displaying high inflammatory activity are lower, as compared with those with low inflammatory activity (Fig. 6A). Statistic analysis indicates that there is an inverse correlation between protein level of PHD3 and severity of UC (Fig. 6B), suggesting that the decrease of PHD3 is associated with progression of UC. The protein levels of occludin in biopsies of patients with high disease severity are lower than those with low disease severity (Fig.  6C), and there is an association between protein level of PHD3 and that of occludin (Fig. 6D). The data provide evidence that PHD3 regulates expression of occludin positively.

Discussion
In this manuscript, we have demonstrated that genetic ablation of Phd3 in mice IECs leads to spontaneous colitis and exacerbates experimental colitis. We find that PHD3 stabilizes occludin and ensures the intestinal epithelial barrier function.
These results suggest that PHD3 plays an important role in protection of intestinal epithelial barrier function.
PHDs inhibitors have been shown to ameliorate experimental colitis in murine models (20,21). This is probably due to the suppression of prolyl hydroxylase activities. One therefore might expect that deletion of PHD3 would more or less protect intestinal epithelial barrier function. However, our results show that PHD3 loss leads to colitis and exacerbates experimental colitis in mice, implying a protective role for this protein. Thus, PHD3 may have an important role in maintaining intestinal barrier function. We here delineate a novel function of PHD3 in stabilizing occludin independent of its hydroxylase activity. Our data suggest that the stabilization of occludin by PHD3 may contribute to the protection of intestinal epithelial barrier.
Although PHDs inhibitors were found to suppress IBD in murine models, the underlying mechanisms are not fully understood, and to what extent each PHD member is involved in reducing the inflammatory burden remains to be deter- . Cell death was determined by means of propidium iodide staining. Briefly, the cells transfected with control or siPHD3 oligonucleotides were grown on 24-well plates at a confluence of ϳ50%. After 48 h, the cells were stained with propidium iodide (1 g/ml) for 20 min, and cell death was determined using the ArrayScan VTI HCS (Thermo Scientific). Cell survival was determined by calculating the ratio of propidium iodide-negative cells to total cells. The right panel shows the efficiency of PHD3 knockdown. Scale bar, 100 m. AUGUST 14, 2015 • VOLUME 290 • NUMBER 33

JOURNAL OF BIOLOGICAL CHEMISTRY 20585
mined. Of the three PHDs (PHD1-3), PHD2 appears to be the primary one that contributes the majority of prolyl hydroxylase activity (26,38), and PHD3 is more effective in suppression of HIF-2␣ (39). A few studies showed that HIF-1␣ prevented mice from intestinal inflammation (31,40,41). In contrast, HIF-2␣ was found to activate the inflammatory response in the intestinal epithelium and promote mice colitis (32). In one study, epithelial HIF-1 signaling was also shown to promote DSS-induced colitis in mice (42). These results implicate the complexity of the PHD/HIF system in IBD. Moreover, the PHDs inhibitors might also target other molecules and other types of cells, implying that the mechanisms that PHDs inhibitors defend against colitis are complex. Here, we demonstrate that PHD3 stabilizes occludin independent of its hydroxylase, suggesting a critical role of PHD3 in TJ barrier function. It remains unclear that how and to what extent the prolyl hydroxylase activity of PHD3 contributes to the intestinal barrier function. This needs further investigation.
Defective intestinal epithelial barrier, characterized by increased intestinal permeability, is an important pathogenic factor that contributes to the development of intestinal inflammation (17,34). Occludin is a main component of the TJ (19,34), and its down-regulation has been observed in UC (43) and Crohn disease (44) patients. Occludin has been found to play a critical role in formation of TJ seal (35), and its defect facilitates macromolecule flux across the intestinal epithelial barrier (28), which may trigger activation of the immune system. Whereas the occludin knock-out mice displayed a complex phenotype, the morphology of intestinal TJ seems intact (45). This has led some to conclude that occludin is not essential to TJ barrier function. Many other studies both in vivo and in vitro, however, indicate that occludin is a critical regulator of TJ barrier function (28, 46 -48). Our results show that deletion of PHD3 leads to disruption of occludin and increases TJ permeability and defect of intestinal barrier function. Expression of exogenous occludin could prevent the PHD3 knockdown-induced permeability of Caco-2 monolayer. These results suggest that PHD3 ensures intestinal barrier function, at least in part, through stabilizing occludin. Examination of UC biopsies shows that the PHD3 level is associated with occludin level, providing evidence that PHD3 regulates occludin positively. Further, we find that a reduction of PHD3 parallels the severity of UC. We cannot draw the conclusion that the increased severity of UC is totally due to decreased PHD3; however, considering the fact that PHD3 stabilizes occludin and occludin enhances TJ barrier, our data suggest that loss of PHD3 contributes to the progression of UC.
HIF-1␣ and HIF-2␣ are implicated in intestinal inflammation (30 -32). In our work, we did not observe an obvious increase of HIF-1␣ or HIF-2␣ in intestinal epithelial cells from Phd3 IEC-KO mice (Fig. 2G), which is consistent with recent finding that single knock-out of PHD family member does not lead to accumulation of HIF␣ (33). These results suggest that PHD3 protects against colitis not through HIF pathway. Because PHD3 has targets other than HIF␣ (7-13), one cannot exclude the possibility that other targets, in addition to occludin, are also involved. It is possible that the PHD3 deficiency-induced   antibody. B, the protein level of PHD3 in colon epithelium is inversely correlated with severity of UC. The Spearman's rank correlation coefficient is calculated. r ϭ Ϫ0.658, p ϭ 0.008, n ϭ 15. C, representative of biopsies stained with occludin antibody. D, the protein level of PHD3 is associated with that of occludin in biopsies. The Spearman's rank correlation coefficient is calculated. r ϭ 0.552, p ϭ 0.033, n ϭ 15. Scale bar, 100 m. AUGUST 14, 2015 • VOLUME 290 • NUMBER 33 background). Additionally, given that colonic flora have an important effect on colitis phenotypes (49), different colonic flora in mice bred in-house are also likely to affect the phenotypic manifestations in these two models.

PHD3 Stabilizes Occludin and Protects Intestinal Barrier
There is lack of an effective treatment for IBD, which has led to investigators seeking a new therapeutic approach. Disruption of occludin is an aspect of a number of diseases, and prevention of its down-regulation is protective against these diseases, such as TNF-induced barrier loss (47). Our results show that PHD3 stabilizes occludin, which may open avenues for developing a strategy for IBD therapy. PHDs are involved in many physiological processes (1,2), and accordingly administration of generic PHD inhibitors may potentially disturb various side effects and has potential limitations (25). The finding that PHD3 stabilizes occludin independent of its hydroxylase may offer an opportunity to target this pathway without disturbing the oxygen-sensing system that is critical for many processes. For example, it might be feasible to deliver hydroxylasedeficient PHD3 to promote occludin TJ formation and fight against colitis.