Molecular Mechanisms of TGF-beta Signaling Pathway

Division Of Basic Sciences - Nci

Investigators

Linked publications & trials

Abstract

Conventional views on the TGF-beta signaling mechanism center on regulation of Smad transcriptional activities. During the previous site-visit cycle, we made a major discovery of a novel function of Smad3 in reshaping genome-wide splicing patterns in favor of expressing protein isoforms that promote EMT and metastasis 6. Our data indicated that phosphorylation at threonine 179 (T179) in the linker region of Smad3 allows it to form a complex with RNA-binding protein poly(rC) binding protein 1 (PCBP1). Using the heavily spliced CD44 pre-mRNA as a model system, we showed that in response to TGF-beta and EGFR signaling, the Smad3-PCBP1 complex becomes enriched in the SC35 positive nuclear speckles and binds CD44 pre-mRNA directly in the variable exon region. Binding of the Smad3-PCBP1 complex to the inherently weak variable exons denies the access of U2 snRNP auxiliary factor 2 (U2AF2) required for the assembly of spliceosomes, thereby causing the exclusion of variable exons. Phosphorylation at T179 of Smad3 requires inputs from both TGF-beta and mitogenic pathways. In many different types of cancer cells, we found that T179 was already phosphorylated due to elevated EGFR-RAS-MAPK pathway activity sustained through oncogenic mutations. Thus, these findings not only break a conceptual ground in revealing the direct role of Smad3 in regulating splicing patterns, the dual requirement of TGF-beta and oncogenic signaling activation for launching this regulation also provides an insight into how TGF-beta is converted from a potent growth inhibitor of normal epithelial cells into a tumor promoter in advanced cancers. Subsequently, we extended the splicing function of Smad3 to controlling TAK1, which is one of the best characterized non-Smad signal transducers critical for the TGF-beta functions in EMT and apoptosis through activating the c-Jun N-terminal kinase (JNK) and p38 MAPK cascade. TAK1 also plays an essential role in mediating TGF-beta-induced activation of I-kappa B kinase (IKK) and the master transcription factor NF-kB that are required for mounting the EMT response and cell survival. Human and mouse TAK1 pre-mRNAs contain an evolutionarily conserved variable exon 12, whose alternative splicing gives rise to a full length or a shortened isoform 21. We found that the exclusion of the TAK1 variable exon 12 by TGF-beta signaling requires the Smad3-PCBP1 and another independent RNA binding complex containing Fox-1 homolog 2 (Rbfox2) 5. Our data further revealed that both long and short TAK1 isoforms can mediate TGF-beta induction JNK and p38 activation, while only the short TAK1 isoform is essential for mediating TGF-beta-induced NF-kB signaling. These differential activities of TAK1 isoforms enable the TAK1 short isoform to mediate TGF-beta-induced EMT and drug resistance whereas the TAK1 long isoform for promoting apoptosis. These observations offer a harmonious explanation for how a single TAK1 kinase is equipped for carrying out the opposing responses of cell survival and apoptosis to TGF-beta. They also reveal a propensity of the alternatively spliced TAK1 isoform to cause drug resistance due to its activity in supporting EMT and NF-kB survival signaling and suggest a possibility of blocking TGF-beta-induced alternative splicing of TAK1 as a viable strategy to combat resistance to cancer therapeutic drugs. Our previous mechanistic studies of Smad3 in the TGF-beta-induced alternative splicing suggest that in cancer cells, Smad3 disengaged from Smad4 to partner with PCBP1 to regulate alternative splicing 6. Human Smad4 mutations were frequently found in advanced cancers, and Smad4 inactivation have been shown to promote tumor invasion and metastasis, raising the possibility that Smad4 mutation is not only a tumor initiation but also a tumor-promoting event. In supporting our view, Smad4 mutations usually occur after Wnt and EGFR pathway activations in colorectal cancers (CRC), and are associated with cell spreading, liver metastasis, and a poor disease prognosis. According to data from TCGA, Smad4 mutations are present in 13.5% CRC cases, whereas Smad3 accounts for 6%. Interestingly, the majority of Smad3 and Smad4 missense mutations are localized to the SMAD-SMAD interaction interface, suggesting that they may block the interaction required for effecting transcriptional responses to TGF-beta. In order to study the role of Smad4 in the TGF-beta-mediated alternative splicing, we obtained a pair of normal human colon-derived organoid cell lines harboring three (APC, KRAS, and P53) or four (APC, KRAS, P53 and SMAD4) of engineered mutations from Dr. Hans Clevers's lab. When implanted into the caecal wall of nude mice, the quadruple mutant cells were reported to show a significant increase in primary tumor growth and metastases in the liver and lungs compared to the triple mutant, suggesting that Smad4 loss accelerates tumor progression. However, after carefully sequencing these two cell lines, we found that Clevers's group generated these two different CRC organoid cell lines by simultaneous transfection of sgRNA for APC, P53 and SMAD4, so that these two lines harbor different APC, and P53 mutations, and one of the Smad4 allele in the quadruple mutant line encodes a missense mutation. Because these different mutations as well as possible dominant negative effect of Smad4 mutation could complicate the interpretation of the results, we decided to regenerate the quadruple mutant (AKPS) line from the Clever's triple mutant (AKP) line using two sgRNAs that delete exon 3 to exon 5 of Smad4 to make a true Smad4 deletion mutation. We characterized these AKPS cells generated in our group. Comparing to that of parental AKP cells, the AKPS cells displayed hollow-like spheres with larger lumen diameters and thinner epithelial layers when culturing in Matrigel, which is very similar to the morphology reported by Clevers' group for their quadruple mutant cells. These AKPS cells promote primary tumor growth when injected subcutaneously in the nude mice, while promote liver metastasis when injected intrasplenic, thus confirming that Smad4 deletion in the AKP colon organoid cells accelerates colon cancer progression. We next subjected both AKP and AKPS cells to TGF-beta stimulation and performed RNAseq analysis for transcriptional and splicing profiles by comparing these two lines. We found that Smad4 deletion diminished TGF-beta-induced transcriptional responses as expected. Treating AKP cells with TGF-beta induced many alternative splicing events, interestingly, majority of these TGF-beta-induced alternative splicing events, occur in the AKPS cells even in the absence of TGF-beta stimulation. These results indicated that loss of Smad4 is sufficient to drive TGF-beta-dependent alternative splicing events. We are currently investigating further the details of molecular mechanisms of Smad3 and Smad4 in regulating TGF-beta-mediated alternative splicing.