Implicación del receptor de glucocorticoides NR3C1 del epitelio intestinal en la respuesta a esteroides, la esteroidogénesis y la neoplasia

  1. Arredondo Amador, María
Supervised by:
  1. Fermín Sánchez de Medina López-Huertas Director
  2. María Olga Martínez Augustin Co-director

Defence university: Universidad de Granada

Fecha de defensa: 08 November 2019

Committee:
  1. José Juan García Marín Chair
  2. Rosario Jiménez Moleón Secretary
  3. María Victoria García-Mediavilla Committee member
  4. Enrique José Cobos del Moral Committee member
  5. Emma Muiños López Committee member

Type: Thesis

Abstract

1 Introduction. Glucocorticoids (GC) are molecules with pleiotropic effects, which are attained by way of a fairly complex mechanism of action that basically depends on the interaction with its primary biological target, the cytoplasmic glucocorticoid receptor (GR). The mechanism of action of GC can divided into three different aspects: the GR activity as a transcription factor, activating (transactivation) or reducing (repression) the expression of different genes, and the interaction of the GR with other transcription factors, resulting in modulation of their transcriptional activity by tethering (transrepression). The vast immunomodulatory effects of GC are the basis of their effectiveness in the treatment of IBD. These are thought to arise mainly from transrepression and classical repression, while the majority of the side effects are related with the transactivation. Inflammatory bowel disease (IBD) is a chronic relapsing inflammatory disorder of the gastrointestinal tract that encompasses two idiopathic and major inflammatory diseases: Crohn's disease and ulcerative colitis. Both forms of IBD significantly impair quality of life, and require prolonged medical and/or surgical interventions. What makes it particularly challenging is its still unknown cause, its unpredictable presentations and symptoms, the less than optimal treatments, and a rise in its incidence and prevalence in many areas of the world. One of the main complications of IBD is the development of colitis-associated colorectal cancer (CAC). The successive periods of inflammation and intestinal regeneration to which the intestine is exposed promotes an hyperproliferative state of the epithelium, generate free radicals that are able to damage DNA and favor the appearance of dysplasias, first step of carcinoma. Nowadays, there is a wide variety of synthetic GC available for clinical use by virtue of their anti-inflammatory and immunosuppressive effects. Among those receiving more attention by clinicians involved in the treatment of IBD are prednisolone and metilprednisolone, administered systemically (intravenous for instance), and oral budesonide, which pursues a local effect taking advantage of its significant first pass metabolism. The use of GC in the treatment of IBD has some significant limitations. At present, only around 40% of the patients respond normally to these drugs, while the remainder do not respond at all or respond at the beginning of the treatment but become dependent when they reach the remission state. Besides, now it is fully accepted that GC administration does not offer any benefit when administered trying to prolong the remission state. The experimental evidence currently available on the application of synthetic GC in different models of colitis has been gathered after their empirical use in humans. Our laboratory has recently described that budesonide exerts both deleterious and protective effects in experimental colitis induced by DSS in mice. On one side, budesonide increased body weight loss, bacterial translocation, sepsis, and bleeding, with increased mortality. On the other side, inflammatory parameters such as MPO or IL6 where lower compared to the non-treated group. Furthermore, the intestinal epithelium is able to synthetize GC independently of adrenal glands. These GC act in an autocrine or paracrine manner, being essential in the intestinal homeostasis and regulating multiple processes, principally inflammatory responses and proliferation. GR knockout mice are not fully viable, mainly because of a failure in pulmonary maturation which leads to premature death of the mice. Several tissue specific KO mice have been generated in the last few years. Skin (a tissue relatively similar to the intestinal epithelia) GR KO mice showed altered development and a basal inflammatory status that increased the susceptibility to detergent agents. Intestinal GR KO mice have been previously generated but only studied for its impact on glucose absorption and gastroparesis. Mice carrying a conditional deletion of the intestinal epithelial GR (NR3C1ΔIEC) were generated using the Cre/loxp technology to evaluate if the harmful effects of BD was due to its action on the epithelial receptor. The colon of these mice exhibited an inflammatory status, as evidenced by an increase of S100A9 and p-STAT3 expression in colonic tissue, with systemic consequences (high plasmatic levels of TNF and IL6). This inflammation was transitory, as it was present 7 days postdeletion but largely reduced after 14 days. Regarding the response of NR3C1ΔIEC mice to experimental colitis by dextran sodium sulfate (DSS), they were protected, although in early stages of the illness they showed more rectal bleeding and increased expression of inflammatory parameters (S100A9 and p-STAT3). This Doctoral Thesis will try to elucidate the causes of the colonic inflammation in NR3C1ΔIEC in basal conditions and the involvement of intestinal epithelial GR on high-dose BD treatment, on intestinal steroidogenesis and on the development of colitis-associated colorectal cancer (CAC). 2 Material and methods. To carry out the objectives of the present Thesis, we used a wide variety of techniques, including tissue culture techniques, intestinal organoid obtention and culture, real time PCR, histology techniques, flow cytometry, ion transport in Ussing chambers, Western blot, ELISA, RNAseq as well as DSS induced colitis, adrenalectomy and AOM/DSS induced CAC. Inducible tissue-specific knockout mice of the GR in the intestinal epithelia (NR3C1ΔIEC) were generated by using Cre-loxp technology in a C57 background by the administration of tamoxifen for 5 days. Mice GRloxp/loxp lacking the Cre gene were used as wild-type controls (WT). In vitro studies were done in organoids coming from jejunum of NR3C1ΔIEC and WT. 3 Aims. On the basis of the scientific background presented in the Introduction and our own previous results, our aims in the present Doctoral Thesis are: 1. To elucidate the causes and consequences of the transitory colonic inflammation due to the lack of intestinal epithelial NR3C1 in basal conditions. 2. To characterize in depth the dual role of intestinal epithelial NR3C1 in the experimental colitis induced by DSS. 3. To characterize the influence of the intestinal epithelial receptor NR3C1 in the regulation of systemic and local intestinal steroidogenesis, as well as the secondary modulation of the inflammatory response. 4. To study the role of the intestinal epithelial NR3C1 receptor in the development of colitis-associated colorectal cancer. 4 Results and discussion. Our results can be split into four sections. Causes and consequences of intestinal epithelial NR3C1 deletion in mice To elucidate the origin of the colonic inflammation due to the lack of intestinal epithelial GR, WT and NR3C1ΔIEC mice were sacrified 7 days after the first tamoxifen injection. The RNAseq analysis of colonic epithelial cells revealed that the intestinal epitelium GR is implicated in several biologic processes, mainly related with immune response, inflammation and defense. In addition, intraepithelial lymphocytes (IELs) and lamina propria mononuclear cells (LPMCs) were analyzed by flow cytometry. A slight increase in TCRγδ+ lymphoctytes was noted in the absence of intestinal GR. Moreover, the intestinal barrier function was evaluated and an increased transmural FITC-dextran flux was detected in the colon of NR3C1ΔIEC mice in vitro in Ussing chambers, as well as a reduction of the presence of ZO-1 at surface intestinal epithelium by immunofluorescence. Together, these results show that GR is basally activated and essential to maintain intestinal homeostasis. Colonic inflammation seems to be caused by a leaky gut and an increase in the proinflammatory signaling due to intestinal epithelial GR loss. Finally, the consequences of GR deletion were evaluated at long-term (8 weeks) and were found to be almost completely nonexistent, confirming the transitory nature of the colonic inflammation. Implications of intestinal epithelial NR3C1 expression in experimental colitis and high-dose of BD treatment. Intestinal epithelial GR has a deleterious role in DSS experimental colitis since his absence provokes an improvement in the general state of mice (at 7 days), as well as a great reduction in mortality in prolonged expositions to DSS 2.5% (w:v) (21 days). Since the epithelial GR is involved in the deleterious response to therapeutic doses of budesonide, an experiment was carried out with a supratherapeutic dose in order to assess whether the benefit of the GC can be increased in the absence ot the epithelial receptor. It was found that mice were protected against death although not against weight loss; this effect was restricted to male mice, which were the most affected overall on the other hand. Thus female WT mice show an upregulated expression of proinflammatory markers such as S100a8, S100a9, Tnf or Il6, and a higher phosphorylation of STAT3. The reason for this sex-dependent effect is unclear at present, but it may be related to the fact that male mice appear to possess a more efficient intestinal steroidogenesis and a higher glucocorticoid-dependent antiinflammatory signaling (male WT overexpress Tsc22d3 and Nr5a2 compared to female WT mice). The loss of intestinal GR in females may be protective via increased epithelial steroidogenesis as a result of the loss of the negative feedback influence of the receptor, or simply because of the underlying inflammation is greater. In vitro stimulation of intestinal organoids from WT and NR3C1ΔIEC mice with TNF+FBS and LPS unveiled the potent inhibition of chemokine expression (Cxcl1 and Ccl2) performed by GR, consistent with increased proinflammatory signaling in vivo. Implication of epithelial NR3C1 in intestinal steroidogenesis. The intestinal epithelial GR seems to exert a negative feedback regulation on intestinal steroidogenesis, as mice lacking the receptor have increased production of corticosterone and augmented intestinal epithelial Cyp11a1 expression (rate limiting enzyme of the process) under an inflammatory challenge such as experimental colitis. On the other hand, in vitro studies using intestinal organoids revealed higher levels of corticosterone on the NR3C1ΔIEC supernatant. Experiments in adrenalectomized mice evidenced that intestinal corticosterone levels are not significantly affected (while they are reduced by 40% in the skin). Conversely, plasma corticosterone levels are unchanged in knockout mice, indicating that intestinal steroidogenesis is not able to partially supply the production from adrenal glands, even in the absence of GR. Role of intestinal epithelial NR3C1 in the development of colitis associated colorectal cancer. Previous studies of our group, as well as the RNAseq analysis and the experiments using organoids, revealed an increased proliferative state of the intestinal epithelium of NR3C1ΔIEC compared to WT mice. CAC was induced to both genotypes using the AOM/DSS model. NR3C1ΔIEC mice showed a larger number of tumors than WT mice. However, they were less invasive than those carried by controls. In addition, NR3C1ΔIEC mice expressed increased levels of promoters and mediators of the epithelial to mesenchymal transition (EMT) or growth factors like Snai1, Snai2, Lgr5, Egfr, Tgfb1 or Myc. These results demonstrate a moderate antitumoral action of GR. 5 Conclusions. 1. The intestinal epithelial glucocorticoid receptor is basally activated in the mouse colon in vivo. Its loss provokes transitory colonic inflammation, presumably due to alterations in the intestinal permeability and/or a higher proinflammatory signaling by the epithelium and infiltrated TCRγδ+ lymphocytes, where the intestinal microbiota seems to play an essential role, as this response is not present in the small intestine. 2. The intestinal epithelial glucocorticoid receptor has a dual role in DSS-induced experimental colitis. Initially, it exerts a deleterious effect, as its absence worsens animal status in the early phase of exposure to DSS. However, as the inflammatory response is established, the deletion of the receptor is clearly protective, to the point of preventing 48% of animal deaths in a chronic colitis protocol. 3. Budesonide treatment of experimental colitis induced by DSS at supratherapeutic doses is strengthened by the absence of the intestinal epithelial glucocorticoid receptor. This effect is observed only in female mice. The mechanism may be related to a specific interaction of the receptor, or simply to the fact that inflammation is more pronounced in females. These results suggest that the clinical response of intestinal inflammation to glucocorticoids could be improved in the absence of epithelial actions. 4. The intestinal epithelial NR3C1 collaborates in the regulation of the novo corticosterone synthesis, possibly through a negative feedback. The resulting increased intestinal corticosterone synthesis is correlated to the protection against colitis, and therefore could be mechanistically involved. 5. In mice, intestinal steroidogenesis is sufficient to the maintenance of the local levels of corticosterone, while in the skin a partial dependence is observed. 6. The deletion of intestinal epithelial NR3C1 increases cellular proliferation and causes a higher tumor load in the AOM/DSS induced experimental colitis-associated colorectal cancer model, although these tumors are less invasive.