The study of microRNAs (miRNAs) became intertwined with cancer biology at an early stage. In one of the first publications reporting altered miRNA expression in cancer, Michael et al. found that loss of miR-143/145 is associated with colorectal cancer.1 This and other seminal studies opened the door to a new field of research, which over the past decade has revealed that miRNA dysregulation is involved in all aspects of malignancy including tumor initiation, progression and metastasis. As one of the most reproducible observations in this vast literature, loss of miR-143/145 has been reported in subsequent studies of colorectal cancer and other epithelial tumor types. Through repression of KRAS, MYC, ERK5 and other oncogenes, overexpression of miR-143/145 inhibits proliferation and survival of cancer cells in cell culture and xenografts.2,3 Therefore, miR-143/145 have been widely accepted as prototypical miRNA tumor suppressors. In a recent study,4 we attempted to complement these reports by examining the endogenous functions of miR-143/145 in mice. We found that loss of miR-143/145 does not lead to any discernable defects in intestinal development or homeostasis. This was not entirely unexpected, as a growing number of studies have reported that miRNAs are dispensable for development but perform important functions under stress conditions.5 Indeed, we found that miR-143/145 were essential for regeneration of epithelial cells following intestinal injury elicited by dextran sulfate sodium (DSS). Adjacent to DSS-induced ulcers, proliferation of wild-type epithelial cells switches from a hierarchical mode restricted to stem cells and transit-amplifying cells to a regenerative mode wherein all cells in the crypts multiply. However, in miR-143/145 null mice, epithelial cells failed to make this transition, resulting in inefficient epithelial repair, sustained inflammation and eventual death of the animals. In the course of investigating the cellular context of this pro-regenerative function, we were surprised to find that miR-143/145 expression is abundant in smooth muscle cells of muscularis mucosae and muscularis externa, myofibroblasts, and other cells of mesenchymal lineages, but completely absent in human and mouse intestinal epithelial cells (Fig. 1). This finding was further corroborated by genetic studies, as mice with mesenchymal deletion of miR-143/145 recapitulated the regenerative defect, while epithelial ablation of miR-143/145 had no discernable effect. Figure 1. miR-143/145 in intestinal regeneration and tumorigenesis. miR-143/145 are abundant in the mesenchymal cells of the intestine, including myofibroblasts and smooth muscle, but completely absent in the epithelium. Although dispensable for development and ... Smooth muscle cells in the GI tract form 2 distinct layers: muscularis mucosae and muscularis externa, which are classic histological structures that, by mechanical contraction, facilitate movement of food along the digestive tract. Previous studies have mainly focused on their mechanical and electrophysiology properties. Given the abundant expression of miR-143/145 in smooth muscle, we deleted the miRNAs specifically within this lineage to determine the contribution of this cell-type to the regenerative defect. Remarkably, smooth muscle-specific miR-143/145 deletion significantly impaired intestinal wound healing, uncovering a previously unrecognized signaling function for this lineage that provides instructive cues during epithelial regeneration. Paracrine signals may be elaborated by smooth muscle cells themselves (Fig. 1). In addition, we showed that within zones of injury, smooth muscle cells adopt a myofibroblast phenotype and migrate into the lamina propria, where they may functionally coordinate epithelial repair. This behavior is reminiscent of smooth muscle cells of the vasculature, which are known to switch between contractile and secretory phenotypes in injury settings.6 We demonstrated that miR-143/145-deficient myofibroblasts secrete an excess of the novel miR-143 target insulin-like growth factor binding protein 5 (IGFBP5), which sequesters IGF ligands and may thereby impair intestinal epithelial proliferation. Further studies are necessary to elucidate the broader signaling program that is dysregulated in miR-143/145-deficient smooth muscle and myofibroblasts. This approach is likely to yield important insights into the mechanisms that govern intestinal wound healing and may reveal new therapeutic strategies for ulcerative colitis, Crohn's disease, and other intestinal diseases. Our findings also have important implications for colorectal cancer. The absence of miR-143/145 in normal and cancerous epithelium clearly rules out a cell autonomous tumor suppressor function for these miRNAs. We suspect that the previously reported loss of miR-143/145 in colorectal cancer simply reflects the distinct cellular composition of normal and neoplastic intestinal tissue. Similar misinterpretations may have occurred in other cancer settings, underscoring the need for careful analyses to resolve the settings wherein these miRNAs may function in a cell autonomous manner to influence cancer cell behavior. Importantly, our findings do not invalidate the reported anti-tumorigenic activity of miR-143/145 in colorectal cancer cells or potential therapeutic strategies based upon delivery of these miRNAs. Nevertheless, the effects of such treatments on mesenchymal cells in target tissues have to be carefully evaluated. Finally, it is tempting to speculate that miR-143/145 could influence colorectal tumorigenesis through their ability to regulate mesenchymal-to-epithelial paracrine signaling (Fig. 1). It has been demonstrated that colorectal tumors exhibit degraded epithelial barrier function, which allows microbial penetration and tumor-promoting inflammation.7 Perhaps loss of miR-143/145 in the mesenchyme, which could impair a compensatory local epithelial regenerative response, exacerbates tumor-associated inflammation and thereby enhances malignant progression. Regardless, these new insights into the natural functions of miR-143/145 serve as yet another reminder that cancer cells do not live in isolation. Stromal components, such as myofibroblasts, vasculature, immune cells, and extracellular matrix, have to be taken into consideration to gain a full understanding of miRNA functions in tumorigenesis.