The ESRP1-GPR137 axis contributes to intestinal pathogenesis

Essential revisions:

1) A central point of the manuscript is the effect on intestinal barrier integrity upon decreased ESRP1 expression. Here the authors should extend their attempts to explain the underlying mechanism and should perform a targeted and more comprehensive analysis of tight junction proteins. Moreover, increased bacterial translocation and anti-commensal antibody response, but normal fecal albumin, lipocalin and FITC-dextran permeability may indicate specific defects in mucus.

To better assess barrier integrity in Triaka mice, we consulted the literature to compile an extensive list of (intestinal) genes belonging to the occludine-, claudine-, tight junction protein (TJP)- and junctional adhesion molecule (JAM) family (1, 2). We then analyzed our RNA-sequencing data to assess the transcript levels of these 21 selected genes.

We found that E-cadherin (Cdh1) transcripts were most abundant, showing in wild-type (WT) colonic intestinal epithelial cells (cIECs) a ~2 fold higher expression level than messages from the next coming gene, Cldn7. Yet, Cdh1 transcript levels were decreased in Triaka compared to WT cIECs (Author response image 1A). These findings on different Cdh1 expression could be independently validated by real time quantitative PCR analysis from additional Triaka and WT cIECs (Author response image 1B). They also further corroborate our data indicating reduced E-cadherin protein expression on Triaka cIECs (Figure 3A and B).

E-cadherin is an epithelial adherens junction protein with central function for the formation of adherens junctions and desmosomes. Consequently, constitutive Cdh1-deficiency in IECs is neonatal lethal (3), while E-cadherin loss in IECs of adult mice leads to intestinal disease due to loss of intestinal epithelial integrity (4). Furthermore, Cdh1-deficient IECs showed compromised barrier function (3). Taken together, it is thus conceivable that the reduction of E-cadherin expression on Triaka IECs accounts for the reduced intestinal tightness in Esrp1-mutant mice. As cIECs have been described to be more dependent on E-cadherin expression for their integrity than IECs from the small intestine (4), this may possibly explain why in Triaka mice the large intestine – but not the small intestine – shows reduced intestinal electrical resistance ex vivo (see main manuscript).

Relatively to E-cadherin, the other (tight) junction proteins were less expressed in mouse IECs. Surprisingly, Cldn8 and Tjp2 were upregulated in Triaka versus WT cIECs, with 1.23 fold and 1.96 fold increase, respectively (Author response image 1A). This may represent compensatory changes subsequent to E-cadherin down-modulation, as previously reported for other cell-cell adhesion genes (5).

In summary, these results suggest that downregulation of E-cadherin is the most relevant alteration in junction protein expression in Triaka versus WT cIECs, which likely contributes to impaired barrier integrity in Esrp1^Triaka mutant mice.

Author response image 1

Download asset Open asset

Does Esrp-1 regulate splicing in Muc2 or any enzymes known for modifications/glycosylation of Muc2?

Muc2 and MUC2 modifying enzymes – ESRP1-dependent mRNA splicing:

To address this point, we have specifically compared the splicing pattern of Mucin 2 (Muc2) isoforms in Triaka versus WT cIECs. In agreement with the data represented in Figure 2B, there was no difference in the relative expression of Muc2 isoforms. This suggests that ESRP1 does not regulate the splicing of Muc2 transcripts.

We next investigated whether ESRP1 may possibly affect the splicing of transcripts of genes involved in the post-translational modification of MUC2. MUC2 protein becomes heavily O-glycosylated (6, 7), a process that is controlled by different glycosyltransferases (8). The initiation of O-glycosylation is performed by peptidyl-GalNAc transferases which add GalNAc to cell surface and secreted proteins being processed in the Golgi apparatus (7).

Thus, we compiled a list of all peptidyl-GalNAc transferases (a total of 19 in mice) as well as other glycosyltransferases previously identified in mouse IECs (9, 10); this list included Galnt1, Galnt2, Galnt3, Galnt4, Galnt5, Galnt6, Galnt7, Galnt9, Galnt10, Galnt11, Galnt12, Galnt13, Galnt14, Galnt15, Galnt16, Galnt17, Galnt18, Galnt19, Galnt20, St6galnac6, St6galnac2, St6gal1, St3gal6, St3gal4, Gal3st2, Fut2, Chst4, B4galnt2, B4galnt1, B3gnt7, B3gnt3, B3galt5, B4galt4, B4galt1, Gcnt3, C1galt1c1 and C1galt1. Analysis of our RNA sequencing data for splicing events in transcripts of these genes – which was performed for those transcripts expressed in sufficient levels in murine cIECs for this kind of analysis – did not uncover potential targets of ESRP1-dependent splicing in our model. This is also indicated in Figure 2B.

MUC2 function also depends on proteolytic cleavage i.e. for the expansion of the mucus layer at the border of the inner to outer mucus layer of the colon (11). MUC2 is cleaved either through endogenous (12, 13) or bacteria-derived enzymes (14). Since the precise identities of these endogenous enzymes are currently unknown (12), we assessed whether ESRP1-dependent splicing may target transcripts from genes encoding proteolytic enzymes present in the inner and other mucus layer (12). Of these analyzed genes (Apeh, Blmh, Casp1, Cndp2, Dnpep, Ggh, Klk, Lap3, Mep1b, Npepps, Pa2g4, Park7, Pgcp, Prep, Psma2, Psma4, Psma5, Psma6, Psma7, Psmb1, Psmb2, Psmb3, Psmb4, Psmb5, Psmb6, Rnpep, Thop1, Usp5), none were found to overlap with identified ESRP1 targets in our dataset (see also Figure 2B).

Lastly, we also assessed datasets from previous studies analyzing ESRP1-dependent splicing events in skin, urethral or pituitary cells (15-17). Systematic search of all the above-mentioned genes did not reveal isoforms of Muc2 or of genes encoding enzymes involved in MUC2 post-translational modifications that are spliced in an ESRP1-dependent manner. However, it has to be mentioned that some of the above-listed genes are preferentially expressed in the intestine and may therefore not have been detected in these different RNA sequencing studies.

In summary, these data suggest that Muc2 or MUC2-modifying genes are not directly affected by ESRP1-mediated mRNA splicing.

Muc2 and MUC2 modifying enzymes – transcript levels:

We next surmised that Esrp1^Triaka might have indirectly altered the relative expression of these different genes involved in post-translational modification of MUC2 (all listed above). We thus analyzed our RNA sequencing data to assess whether transcript levels of these genes were differently expressed in Triaka versus WT cIECs. Of the 65 analyzed genes, only few displayed significant changes, with minor changes in expression levels, i.e. ~30% up- or downregulation in Triaka versus WT cIECs (upregulated genes: Galnt3, 1.27 fold, Lap3, 1.3 fold change; downregulated genes: St6galnac6, 0.73 fold; St3gal6, 0.79 fold; St3gal4, 0.67 fold and B3galt5, 0.73 fold change; Author response image 2).

Finally, while gene ontology analysis disclosed that the expression of genes involved in cell cycle and proliferation was affected in Triaka compared to WT cIECs (Figure 3—figure supplement 1 and Figure 3—source data 1, main manuscript), no alterations were detected in pathways or gene sets involved in barrier function or mucus production/modification (Author response table 1 below).

As these changes in expression are rather marginal and since gene ontology analysis did not identify alterations in barrier function or mucus production/modification in Triaka cIECs, we conclude that the differences in transcript levels of Galnt3, Lap3, St6galnac6, St3gal6, St3gal4, B3galt5 are likely not biologically relevant.

Author response image 2

Download asset Open asset

Muc2 and MUC2 modifying enzymes – protein levels in situ:

We also assessed the expression of MUC2 protein as a surrogate of the mucus layer in cIECs, at steady-state. The thickness of the inner MUC-2 layer was similar between Triaka and WT (Author response image 3A and B). Taken together, these different analyses do not provide evidence for overtly altered or impaired MUC2 expression, modification or function in the Triaka versus WT intestine.

Author response image 3

Download asset Open asset

Current explanations for the reduced intestinal barrier in Triaka mice:

Our additional data presented above suggest that the diminished tightness of the Triaka intestine is caused by mechanisms different than a direct effect of Esrp1^Triaka on the splicing pattern of tight junction or adherence junction genes, or of Muc2 or MUC2-modifiers. In addition, the expression of the majority of these genes in IECs is not or only marginally affected by the Triaka mutation.

Therefore, the reduced intestinal barrier in Triaka mice is likely caused by a combination of the following three parameters:

i) Epithelial-to-mesenchymal transition (EMT) is a process by which epithelial cell-cell junctions are destabilized and the apical-basal polarity of epithelia is lost (18, 19). ESRP1 is a well-established regulator of EMT (20, 21), and our results provide several lines of evidence for EMT-associated phenotypes in Triaka IECs, including decreased proliferative capacity and decreased expression of E-cadherin (an adherens junction protein). Indeed, downregulation of E-cadherin expression helps the destabilization of cell adhesions and the loss of apical-basal polarity to increase cell motility during EMT (18, 22). E-cadherin downregulation is therefore considered a hallmark of EMT (18). Transcription factors regulating EMT that include SNAIL, SLUG, TWIST, and the ZEB family all repress E-cadherin expression during EMT (18). Of note, inhibition of ZEB1/2 promotes tight junction assembly of breast epithelial cells (23). In contrast to these EMT-promoting factors, ESRP1 expression correlates with that of E-cadherin (24). ESRP1 overexpression has been reported to prevent EMT, among others by maintaining high levels of E-cadherin and the localization of this protein at cell junctions, while ESRP1 knockdown induced the opposite effects (25). Given the above-discussed role of E-cadherin for intestinal epithelial integrity, it is thus likely that EMT-driven downregulation of E-cadherin contributes to the diminished intestinal barrier of Triaka mice.

ii) Moreover, we reveal in our study that distinct ESRP1-specific GPR137 isoforms induced different trans-epithelial electrical resistance in monolayers of transduced intestinal epithelial CMT-93 cells. Indeed, CMT-93 cells transduced with Gpr137_Long – the isoform preferentially expressed in Esrp1^Triaka IECs – displayed lower barrier integrity compared to cells transduced with the Gpr137_Short isoform. Thus, we conclude that alteration in the relative frequency of GPR137 isoform partly underlie the defective barrier in Triaka mice (Figure 5).

iii) In addition to this involvement of reduced E-cadherin expression and altered generation of GPR137 isoforms, there may be other mechanisms contributing to the diminished barrier integrity in Triaka mice. For instance, CD44 is also an important regulator of barrier function (26) and a promoter of EMT (25, 27). Moreover, distinct CD44 isoforms were shown to affect the barrier function in endothelial cells (28) and to induce E-cadherin loss at epithelial cell-cell junctions (25). This is already discussed in the manuscript:

“…Besides Gpr137, we also revealed other genes with altered splicing patterns in Triaka compared with WT cIECs, which likely also contribute to the intestinal phenotype in Triaka mice. Among those ESRP1 target genes, Cd44 is probably the most studied for its intestinal function. CD44 and its isoforms negatively regulate IEC apoptosis (Zeilstra et al., 2008; Lakshman et al., 2004) and participate in various cellular functions, such as proliferation, adhesion, and migration (Ponta et al., 2003).”

2) The authors suggest that the ESRP1-mutated tumors are more aggressive and show signs of EMT. This appears to only be supported by an expression analysis of E-cadherin. The existence of EMT features should be supported by the inclusion of additional markers.

To address this point, we compared the expression of additional epithelial and mesenchymal markers commonly used to assess EMT. Analysis of our RNA sequencing data indicated a consistent trend for downregulation or upregulation of several epithelial- or mesenchymal-regulated gene transcripts in Triaka compared to WT cIECs, respectively (Author response image 4A).

These findings were further validated for several of these markers by qPCR.Expression of the EMT-inducing transcription factors Zeb1 and Zeb2 was increased in Triaka cIECs (Author response image 4B-C). Correspondingly, Cdh1 and Esrp1 expression was downregulated in these cells, in agreement with previous studies (24, 25) (Author response image 1 and 4D, respectively). In addition, we show in the manuscript that, compared to controls, Triaka intestinal tumors display increased expression of TGF-β1 protein, a well-known EMT inducer (32, 33) (Figure 4—figure supplement 2).

Taken together, these results indicate a more mesenchymal transcription pattern in Triaka cIECs, therefore suggesting that these cells underwent partial EMT. We added this information to the revised main manuscript (see also new figure: Figure 4—figure supplement 3):

“…These features of Esrp1^Triaka CRC lesions likely resulted from a partial EMT signature expressed by Triaka cIECs, prior to transformation (Figure 4—figure supplement 3).”

Author response image 4

Download asset Open asset

3) The human data, which enriches the vast amount of granular experiments performed in order to study the Esrp1^Triaka model, highlight that Esrp1 deserves attention from the community and certainly warrants further studies. For the human data on survival tables, please highlight the patient characteristics and the exclusion of differences that might explain survival benefits.

To specifically address these comments, we first performed a univariate analysis assessing ESRP1 expression or clinicopathological features relatively to patient survival. Parameters found to be statistically significant in this first univariate analysis were subsequently analyzed in a multivariate analysis. In this multivariate analysis, ESRP1 expression was found to be a significant predictor of CRC patient survival when controlling for other confounding factors. We accordingly updated this information in the revised manuscript as follows:

“Furthermore, low nuclear ESRP1 expression was associated with reduced patient survival (P = 0.0456), larger tumors (P = 0.0034), lymphatic invasion (P = 0.0466), advanced pT-stage (P = 0.02) and presence of nodal metastasis (P = 0.016) (Figure 6C and Figure 6—source data 1). ESRP1 expression was determined to be an independent prognostic factor, after adjusting for the confounding effects of pT, pN, pM, tumor budding, and lymphatic invasion (Figure 6—source data 2). This indicates a possible tumor-suppressive role of ESRP1 in CRC.”

In addition, we added a new table showing the results of the univariate and multivariate survival analysis (Figure 6—source data 2) and we updated the section on “Evaluation of ESRP1 in CRC and IBD cases” in the Materials and methods of the revised manuscript.

Minor points:

1) While the authors clearly show altered expression of two known targets of Esrp1 and decide to study hypofunctional Esrp1 in intestinal epithelial cells, the exact consequence of the Esrp1-M161V substitution and its effects on Esrp1 function remains difficult to determine. Specifically, while both Cd44v5 and Fgfr2-IIIb expression serve as proof-of-principle and are diminished in Esrp1Triaka mice, the degree to which they are diminished differs vastly. Is it possible that reduced function of Esrp1 (albeit to different degrees for each individual target) or potentially additional 'off'-target effects (modifying some of the 35 identified Ersp1-targets) explain parts of the phenotype. Please expand discussion on these topics.

We presently do not have evidence that off-target effects may be caused by Esrp1^Triaka. Nevertheless, it has to be mentioned here that as many as 14 of the 35 identified genes with different relative frequency of splicing isoforms in Esrp1^Triakacompared to Esrp1^WT cIECs have been previously reported to be targets of Esrp1 in tissues other than intestine (15-17, 34). This further supports the notion of a hypomorphic instead of a neomorphic activity (or gain-of-function) of ESRP1^Triaka.

We currently do not have a definitive explanation as to why hypofunctional Esrp1^Triaka has a distinct impact on Cd44v5 and Fgfr2-IIIb splicing. These differences may be reflective of i) non-physiological in vitro conditions, e.g. due to the use of transfected cell lines, ii) target gene-specific effects of Esrp1^Triaka (i.e. exons of transcripts from distinct genes may be spliced by ESRP1 at different rates. Indeed, Esrp-regulated splicing events have been reported to show variable sensitivity to the loss of Esrp1 (and/or Esrp2) (15, 24)), iii) or they may also depend on the relative abundance of target pre-mRNA. The latter was observed, e.g., for the splicing of Fgfr2-IIIb that was not affected by the Esrp1^Triaka mutation in vivo, at steady-state. Inflammatory conditions upregulate Fgfr2 expression (35-37), and we accordingly observed an increase in Fgfr2-IIIb transcripts in cIECs from mice with mild colitis. Interestingly, these inflammatory conditions also revealed the splicing defects of ESRP1^Triaka, as evidenced by reduced Fgfr2-IIIb levels in Triaka compared to WT cIECs (Author response image 5). Yet these results are preliminary, as it is not known how inflammation may further modulate Esrp1 expression or function.

Author response image 5

Download asset Open asset

We also discuss now these points in the revised manuscript, respectively:

“…ESRP1^Triaka however showed reduced levels of Cd44v5 and Fgfr2-IIIb inclusion, with 1.8 and 3.3 fold induction, respectively (Figure 1A and B). This variation in the extent of in vitro splicing of Cd44 versus Fgfr2 by ESRP1^Triaka likely relates to the fact that Esrp1-regulated splicing events show distinct sensitivity to Esrp1 loss (15, 24).”

“…Although our data indicate reduced splicing activity from the Triaka mutation, a possible neomorphic effect of ESRP1^Triaka cannot be fully excluded, and further investigation using other genetic models is required to examine this aspect.”

Reviewer #1:

[…] The following points should be addressed to improve this study:

1) Is there any difference in the expression of components of the tight junctions? This would support the authors claim about a barrier defect.

Please refer to Essential Revision #1 as an answer to this comment.

2) Is there any clear evidence for the existence of EMT apart from changes in E-cadherin?

Please refer to Essential Revision #2 as an answer to this comment.

Reviewer #2:

[…] Minor criticism is listed below:

1) Increased bacterial translocation and anti-commensal antibody response, but normal fecal albumin, lipocalin and FITC-dextrane permeability may indicate specific defects in mucus. Does Esrp-1 regulate splicing in Muc2 or any enzymes known for modifications/glycosylation of Muc2?

Please refer to Essential Revision #1 as an answer to this comment.

Reviewer #3:

[…] The major limitation of this study lies in the utilization of the Esrp1^Triaka model itself. While clearly the authors show altered expression of two known targets of Esrp1 and conclude therefore to study hypofunctional Esrp1 in intestinal epithelial cells, the exact consequence of the Esrp1-M161V substitution and therefore on Esrp1 function remains difficult to determine. Therefore, the relevance of studying this specific point mutation in the absence of an association with human disease is at least somewhat limited. Specifically, while both Cd44v5 and Fgfr2-IIIb expression that serve as proof of principle are diminished in Esrp1^Triaka mice the degree to which they are diminished differs vastly. Therefore it cannot be fully addressed if indeed reduced function of Esrp1 (albeit to different degrees for each individual target) or additional off-target effects (artificially modifying some of the 35 identified Ersp1-targets) explain parts of the phenotype. Have the authors considered at all to study Esrp1 heterozygous KO mice (Esrp1+/-) or mice exhibiting a heterozygous conditional deletion within the intestinal epithelium? Alternatively, a conditional deletion of both alleles at a later time point e.g. 6 weeks of age utilizing e.g. Vcre-ERT2 mice might provide additional insight.

Please refer to Minor Point #1 as a partial answer to this comment.

Heterozygous ESRP1 knockout (Esrp1^+/-) mice have been reported not to show any phenotype or splicing defects. Specifically, epidermal splicing of Arhgef11, Enah, Fgfr1-IIIb, or Fgfr3-IIIb was altered in homozygous Esrp1^-/- compared to WT mice, yet it was not altered in heterozygous Esrp1^+/- animals (15). Accordingly, it is unlikely that Esrp1^+/- mice show an intestinal phenotype.

References

1) Ulluwishewa D, Anderson RC, McNabb WC, Moughan PJ, Wells JM, Roy NC. Regulation of tight junction permeability by intestinal bacteria and dietary components. J Nutr. 2011;141(5):769-76. doi: 10.3945/jn.110.135657.

2) Luissint AC, Nusrat A, Parkos CA. JAM-related proteins in mucosal homeostasis and inflammation. Semin Immunopathol. 2014;36(2):211-26. doi: 10.1007/s00281-014-0421-0.

3) Bondow BJ, Faber ML, Wojta KJ, Walker EM, Battle MA. E-cadherin is required for intestinal morphogenesis in the mouse. Developmental biology. 2012;371(1):1-12. doi: 10.1016/j.ydbio.2012.06.005.

4) Schneider MR, Dahlhoff M, Horst D, Hirschi B, Trulzsch K, Muller-Hocker J, Vogelmann R, Allgauer M, Gerhard M, Steininger S, Wolf E, Kolligs FT. A key role for E-cadherin in intestinal homeostasis and Paneth cell maturation. PloS one. 2010;5(12):e14325. doi: 10.1371/journal.pone.0014325.

5) Chen A, Beetham H, Black MA, Priya R, Telford BJ, Guest J, Wiggins GA, Godwin TD, Yap AS, Guilford PJ. E-cadherin loss alters cytoskeletal organization and adhesion in non-malignant breast cells but is insufficient to induce an epithelial-mesenchymal transition. BMC cancer. 2014;14:552. doi: 10.1186/1471-2407-14-552.

6) Johansson ME, Ambort D, Pelaseyed T, Schutte A, Gustafsson JK, Ermund A, Subramani DB, Holmen-Larsson JM, Thomsson KA, Bergstrom JH, van der Post S, Rodriguez-Pineiro AM, Sjovall H, Backstrom M, Hansson GC. Composition and functional role of the mucus layers in the intestine. Cellular and molecular life sciences: CMLS. 2011;68(22):3635-41. doi: 10.1007/s00018-011-0822-3.

7) Arike L, Hansson GC. The Densely O-Glycosylated MUC2 Mucin Protects the Intestine and Provides Food for the Commensal Bacteria. Journal of molecular biology. 2016;428(16):3221-9. doi: 10.1016/j.jmb.2016.02.010.

8) Bennett EP, Mandel U, Clausen H, Gerken TA, Fritz TA, Tabak LA. Control of mucin-type O-glycosylation: a classification of the polypeptide GalNAc-transferase gene family. Glycobiology. 2012;22(6):736-56. doi: 10.1093/glycob/cwr182.

9) Johansson ME, Jakobsson HE, Holmen-Larsson J, Schutte A, Ermund A, Rodriguez-Pineiro AM, Arike L, Wising C, Svensson F, Backhed F, Hansson GC. Normalization of Host Intestinal Mucus Layers Requires Long-Term Microbial Colonization. Cell host & microbe. 2015;18(5):582-92. doi: 10.1016/j.chom.2015.10.007.

10) Zhang L, Tian E, Ten Hagen KG. UDP-N-Acetyl-Α-D-Galactosamine: Polypeptide N-Acetylgalactosaminyltransferases (ppGalNAc-Ts). In: Taniguchi N, Honke K, Fukuda M, Narimatsu H, Yamaguchi Y, Angata T, editors. Handbook of Glycosyltransferases and Related Genes. Tokyo: Springer Japan; 2014. p. 495-511.

11) Johansson ME, Larsson JM, Hansson GC. The two mucus layers of colon are organized by the MUC2 mucin, whereas the outer layer is a legislator of host-microbial interactions. Proc Natl Acad Sci U S A. 2011;108 Suppl 1:4659-65. doi: 10.1073/pnas.1006451107.

12) Johansson ME, Phillipson M, Petersson J, Velcich A, Holm L, Hansson GC. The inner of the two Muc2 mucin-dependent mucus layers in colon is devoid of bacteria. Proc Natl Acad Sci U S A. 2008;105(39):15064-9. doi: 10.1073/pnas.0803124105.

13) Schutte A, Ermund A, Becker-Pauly C, Johansson ME, Rodriguez-Pineiro AM, Backhed F, Muller S, Lottaz D, Bond JS, Hansson GC. Microbial-induced meprin β cleavage in MUC2 mucin and a functional CFTR channel are required to release anchored small intestinal mucus. Proc Natl Acad Sci U S A. 2014;111(34):12396-401. doi: 10.1073/pnas.1407597111.

14) Lidell ME, Moncada DM, Chadee K, Hansson GC. Entamoeba histolytica cysteine proteases cleave the MUC2 mucin in its C-terminal domain and dissolve the protective colonic mucus gel. Proc Natl Acad Sci U S A. 2006;103(24):9298-303. doi: 10.1073/pnas.0600623103.

15) Bebee TW, Park JW, Sheridan KI, Warzecha CC, Cieply BW, Rohacek AM, Xing Y, Carstens RP. The splicing regulators Esrp1 and Esrp2 direct an epithelial splicing program essential for mammalian development. eLife. 2015;4. doi: 10.7554/eLife.08954.

16) Bebee TW, Sims-Lucas S, Park JW, Bushnell D, Cieply B, Xing Y, Bates CM, Carstens RP. Ablation of the epithelial-specific splicing factor Esrp1 results in ureteric branching defects and reduced nephron number. Developmental dynamics: an official publication of the American Association of Anatomists. 2016;245(10):991-1000. doi: 10.1002/dvdy.24431.

17) Lekva T, Berg JP, Lyle R, Heck A, Ringstad G, Olstad OK, Michelsen AE, Casar-Borota O, Bollerslev J, Ueland T. Epithelial splicing regulator protein 1 and alternative splicing in somatotroph adenomas. Endocrinology. 2013;154(9):3331-43. doi: 10.1210/en.2013-1051

en.2013-1051 [pii].

18) Lamouille S, Xu J, Derynck R. Molecular mechanisms of epithelial-mesenchymal transition. Nat Rev Mol Cell Biol. 2014;15(3):178-96. doi: 10.1038/nrm3758.

19) Huang RY, Guilford P, Thiery JP. Early events in cell adhesion and polarity during epithelial-mesenchymal transition. Journal of cell science. 2012;125(Pt 19):4417-22. doi: 10.1242/jcs.099697.

20) Warzecha CC, Carstens RP. Complex changes in alternative pre-mRNA splicing play a central role in the epithelial-to-mesenchymal transition (EMT). Semin Cancer Biol. 2012;22(5-6):417-27. doi: 10.1016/j.semcancer.2012.04.003

S1044-579X(12)00059-4 [pii].

21) Yang Y, Park JW, Bebee TW, Warzecha CC, Guo Y, Shang X, Xing Y, Carstens RP. Determination of a Comprehensive Alternative Splicing Regulatory Network and Combinatorial Regulation by Key Factors during the Epithelial-to-Mesenchymal Transition. Molecular and cellular biology. 2016;36(11):1704-19. doi: 10.1128/MCB.00019-16.

22) Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest. 2009;119(6):1420-8. doi: 10.1172/JCI39104.

23) Elsum IA, Martin C, Humbyert PO. Scribble regulates an EMT polarity pathway through modulation of MAPK-ERK signaling to mediate junction formation. Journal of cell science. 2013;126(Pt 17):3990-9. doi: 10.1242/jcs.129387.

24) Warzecha CC, Sato TK, Nabet B, Hogenesch JB, Carstens RP. ESRP1 and ESRP2 are epithelial cell-type-specific regulators of FGFR2 splicing. Molecular cell. 2009;33(5):591-601. doi: S1097-2765(09)00069-0 [pii]

10.1016/j.molcel.2009.01.025.

25) Brown RL, Reinke LM, Damerow MS, Perez D, Chodosh LA, Yang J, Cheng C. CD44 splice isoform switching in human and mouse epithelium is essential for epithelial-mesenchymal transition and breast cancer progression. J Clin Invest. 2011;121(3):1064-74. doi: 10.1172/JCI44540.

26) Kirschner N, Haftek M, Niessen CM, Behne MJ, Furuse M, Moll I, Brandner JM. CD44 regulates tight-junction assembly and barrier function. The Journal of investigative dermatology. 2011;131(4):932-43. doi: 10.1038/jid.2010.390.

27) Xu H, Tian Y, Yuan X, Wu H, Liu Q, Pestell RG, Wu K. The role of CD44 in epithelial-mesenchymal transition and cancer development. OncoTargets and therapy. 2015;8:3783-92. doi: 10.2147/OTT.S95470.

28) Zhang P, Fu C, Bai H, Song E, Dong C, Song Y. CD44 variant, but not standard CD44 isoforms, mediate disassembly of endothelial VE-cadherin junction on metastatic melanoma cells. FEBS Lett. 2014;588(24):4573-82. doi: 10.1016/j.febslet.2014.10.027.

29) Zeilstra J, Joosten SP, Dokter M, Verwiel E, Spaargaren M, Pals ST. Deletion of the WNT target and cancer stem cell marker CD44 in Apc(Min/+) mice attenuates intestinal tumorigenesis. Cancer research. 2008;68(10):3655-61. doi: 10.1158/0008-5472.CAN-07-2940.

30) Lakshman M, Subramaniam V, Jothy S. CD44 negatively regulates apoptosis in murine colonic epithelium via the mitochondrial pathway. Experimental and molecular pathology. 2004;76(3):196-204. doi: 10.1016/j.yexmp.2003.12.009.

31) Ponta H, Sherman L, Herrlich PA. CD44: from adhesion molecules to signalling regulators. Nat Rev Mol Cell Biol. 2003;4(1):33-45. doi: 10.1038/nrm1004.

32) Miettinen PJ, Ebner R, Lopez AR, Derynck R. TGF-β induced transdifferentiation of mammary epithelial cells to mesenchymal cells: involvement of type I receptors. The Journal of cell biology. 1994;127(6 Pt 2):2021-36.

33) Xu J, Lamouille S, Derynck R. TGF-β-induced epithelial to mesenchymal transition. Cell research. 2009;19(2):156-72. doi: 10.1038/cr.2009.5.

34) Dittmar KA, Jiang P, Park JW, Amirikian K, Wan J, Shen S, Xing Y, Carstens RP. Genome-wide determination of a broad ESRP-regulated posttranscriptional network by high-throughput sequencing. Molecular and cellular biology. 2012;32(8):1468-82. doi: 10.1128/MCB.06536-11.

35) Huang JJ, Joh JW, Fuentebella J, Patel A, Nguyen T, Seki S, Hoyte L, Reshamwala N, Nguyen C, Quiros A, Bass D, Sibley E, Berquist W, Cox K, Kerner J, Nadeau KC. Eotaxin and FGF enhance signaling through an extracellular signal-related kinase (ERK)-dependent pathway in the pathogenesis of Eosinophilic esophagitis. Allergy Asthma Clin Immunol. 2010;6(1):25. doi: 10.1186/1710-1492-6-25.

36) Liu CJ, Jin JD, Lv TD, Wu ZZ, Ha XQ. Keratinocyte growth factor gene therapy ameliorates ulcerative colitis in rats. World journal of gastroenterology: WJG. 2011;17(21):2632-40. doi: 10.3748/wjg.v17.i21.2632.

37) D'Amici S, Ceccarelli S, Vescarelli E, Romano F, Frati L, Marchese C, Angeloni A. TNFalpha modulates Fibroblast Growth Factor Receptor 2 gene expression through the pRB/E2F1 pathway: identification of a non-canonical E2F binding motif. PloS one. 2013;8(4):e61491. doi: 10.1371/journal.pone.0061491.