Although new cases of colorectal cancer (CRC) are on the decline, CRC is amongst the top four most commonly diagnosed cancers in the United States according to the latest NCI SEER report. Underlying risk factors such as smoking, obesity, and family genetic history can lead to malignant transformation of intestinal epithelial cells. At the molecular level, mutations in the genes APC, LKB1, MLH2, and MYH have been associated with colorectal tumor initiation or progression. However, regulation at the transcriptional-translational interface by RNA-binding proteins (RBPs) has not been well characterized within CRC. To clarify the role of the RBPs LIN28B and IMP1 in intestinal cell differentiation, regeneration, and malignant transformation, CAMB alumna Dr. Priya Chatterji and colleagues from the laboratory of Anil Rustgi recently defined a novel signaling axis between LIN28B, IMP1, and the downstream WNT signaling pathway.
LIN28B is a master transcriptional regulator that suppresses let-7 miRNA. Subsequently, this suppression increases the transcription of let-7 miRNA targets including IMP1. Previous members from the Rustgi laboratory established that IMP1 acts as an oncogene in IMP1-overexpression CRC xenograft models. Priya encountered an early obstacle when determining the LIN28B regulated translatome and deciphering downstream targets of LIN28B unique to intestinal epithelium repair and colorectal tumor initiation. Priya stated that “since both LIN28B and IMP1 are RNA binding proteins that bind thousands of targets, it was important to figure out how to focus my study. My collaboration with Dr. Premal Shah at Rutgers to do the ribosome profiling really guided my study.” Priya performed ribosomal profiling in SW480 CRC cells and observed that IMP1 was the most notably enriched post translational regulatory element when LIN28B was overexpressed. Subsequent CRISPR/Cas9 deletion of IMP1 in two CRC cell lines followed by RNA-seq and gene set enrichment analysis (GSEA) led to the discovery that IMP1 was a significant translational regulator of pathways associated with colorectal cancer, in particular the WNT signaling pathway.
Priya first examined clinical datasets to search for correlations between LIN28B and IMP1 in CRC patients. She found that LIN28B and IMP1 transcript levels positively correlated when interrogating colorectal cancer patient RNA-seq dataset from The Cancer Genome Atlas database. Additionally, previous studies implicate IMP1 as an oncogene and driver of tumor progression in CRC, independent of LIN28B expression. Using intestine-specific LIN28B overexpression genetic murine models or LIN28B overexpression xenograft murine models, Priya manipulated IMP1 expression to characterize IMP1 regulation on cell differentiation and CRC tumor growth properties. She found that regardless of IMP1 protein expression level, tumor volume decreased in mice injected with LIN28B overexpressing colorectal cancer cells. On the contrary, mice with intestinal-specific LIN28B overexpression acquired multiple tumors and increased counts of proliferating cells in normal adjacent intestinal tissue with IMP1 loss. Notably, there was no tumor formation in genetic models of IMP1 loss alone. However, the combination of LIN28B overexpression with IMP1 loss led to undifferentiated tumors. These data imply that IMP1 acts as an inhibitor in colorectal tumor initiation when LIN28B is overexpressed within the intestine.
Building upon the inhibitory role of IMP1 in CRC tumor initiation, Priya hypothesized that IMP1 may regulate cell stemness and epithelial regeneration within normal intestinal epithelia. Intestinal epithelial-specific deletion of IMP1 led to the regeneration of crypt cells in mice following whole body irradiation - a stressor used to ablate most intestinal cell types. To further examine the stem cell compartment effected by IMP1 loss, Priya interrogated the mRNA expression of stem cell-related genes including WNT target genes and found an increase in mRNA expression of these genes. Furthermore, beta-catenin, a protein important for nuclear transcription of WNT target genes, was more abundant in the nucleus of LIN28B-overexpressing cells, and enriched for genes that encompass the WNT signaling pathway. Additionally, GSEA performed on RNA-seq of IMP1 wildtype or IMP1 deleted intestinal stem cell tissue disclosed a WNT target gene enrichment signature in the absence of IMP1. These observations led to the hypothesis that IMP1 impacts downstream canonical WNT targets in the intestine.
To complement her hypothesis, Priya utilized an IMP1 overexpression model to see if IMP1 overexpression had an opposite effect on tumor growth or intestinal regeneration in the presence of LIN28B overexpression. Priya recounts that, “it was necessary to test my hypothesis out in multiple systems both in vitro, in vivo and ex vivo.” IMP1 overexpression resulted in no tumor formation irrespective of LIN28B overexpression status, suggesting IMP1 expression by itself was not enough to generate tumors. When comparing LIN28B overexpression/IMP1 deletion to a LIN28B/IMP1 double overexpression model, the double overexpression model showed a decrease in WNT target mRNA expression. This finding suggested that IMP1 was a switch for WNT signaling and IMP1 expression may be a result of another upstream protein involved in tumor initiation.
Taken together, Priya’s findings illustrate a more intricate signaling network between LIN28B, IMP1, and WNT downstream targets. The presence, absence, or overexpression of LIN28B defines the effects of IMP1 expression on tumor burden, differentiation potential, and epithelial cell repair within the colorectal tumor environment or normal intestinal ecosystem. IMP1 loss promotes expansion of the intestinal stem cell compartment and augments LIN28B-mediated colorectal tumor progression. Ongoing efforts from Priya and other members of the Rustgi lab to further investigate IMP1 expression in colonic epithelial repair resulted in the additional European Molecular Biology Organization Reports publication, “Posttranscriptional regulation of colonic epithelial repair by RNA binding protein IMP1/IGF2BP1" published in late March. Using conventional therapeutic approaches to target signaling pathways in CRC is difficult due to toxicity in normal intestinal epithelium. This study suggests that inducing IMP1 may be a beneficial therapeutic strategy for a subset of CRC cases in which LIN28B is overexpressed.
The LIN28B-IMP1 signaling axis in the intestinal epithelium. LIN28B overexpression with IMP1 loss leads to tumor progression in colorectal tumor cells and rapid expansion of intestinal stem cells. (Adapted from Chatterji, P et al. 2018)
Chatterji, P, Hamilton, K E et al. (2018). The LIN28B-IMP1 post-transcriptional regulon has opposing effects on oncogenic signaling in the intestine. Genes & Development. 32: 1020-1034.