Two main subtypes of colorectal cancer (CRC) exist; one subtype displays active wingless (Wnt) signaling through inactivating mutations targeting the Wnt negative regulator adenomatous polyposis coli (APC). The other subtype originates from sessile serrated adenomas and is enriched in activating mutations of the V-raf murine sarcoma viral oncogene homolog B (BRAF) oncogene, whereby a valine to glutamate substitution at codon 600 causes constitutive activation of the BRAF kinase and the downstream mitogen-activated protein kinase (MAPK) pathway [1,2,3].

Cholesterol biosynthesis through the mevalonate pathway is associated with an increased risk of developing CRC [4]. Tumorigenesis in mice carrying mutations in the Apc gene (Apcmin) displays augmented cholesterol biosynthesis and is reversed by genetic or pharmacologic inhibition of the pathway [5]. Evidence suggests that mutant BRAF regulates the transcription factor sterol regulatory element binding protein-1 (SREBP1), a master regulator of cholesterol and lipid metabolism, in several malignancies, including CRC [6]. Nevertheless, whether cholesterol metabolism contributes to BRAF-mutant serrated neoplasia has not been thoroughly investigated. Recently, we identified increased expression of the mevalonate and cholesterol metabolism signature in datasets of human BRAF-mutant and/or serrated neoplasias as compared to normal tissues or CRCs harboring a wild-type BRAF oncogene. Using a mouse model carrying inducible expression of BrafV600E in the intestinal epithelium, we confirmed transcriptional activation of the same signature, which was reversed by pharmacologic inhibition of the MAPK pathway [7]. Moreover, inhibition of the mevalonate pathway with statins, a class of drugs widely prescribed in the treatment of cardiovascular conditions, which inhibit the rate-limiting enzyme 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase, prevented the establishment of hyperplastic crypts in the intestine of BrafV600E-mutant mice [7]. Thus, even in the context of serrated CRC, cholesterol biosynthesis is increased and pro-tumorigenic. However, since a subset of serrated lesions does not harbor mutations in the BRAF oncogene, the dependence of the signature’s expression on the MAPK pathway should be further investigated. The mevalonate pathway also synthesizes isoprenoids, fueling the biosynthesis of Coenzyme Q and protein farnesylation. Currently, it is unclear whether the anti-cancer efficacy of statins depends on their impact on the biosynthesis of sterols and/or isoprenoids [4, 5, 8]. Moreover, when comparing the expression of the cholesterol gene signature between BRAF wild-type adenomas and normal tissues, we did not observe enrichment in the neoplastic specimens, suggesting that transcriptional regulation might play a lesser role in the activation of cholesterol metabolism in wnt-driven CRC.

An interesting observation distinguishes serrated and non-serrated neoplasia. During Wnt-driven tumorigenesis in Apcmin mice, sterols act as mitogens and stimulate the proliferation of intestinal stem cells (ISCs) without affecting survival [5]. In contrast, in BRAF-mutant intestinal epithelium, we observed that cholesterol biosynthesis protects crypt cells from apoptosis without affecting proliferation [7]. This result, which agrees with known anti-apoptotic properties of statins [9], might highlight biological differences between crypt cells with constitutively activated Wnt or MAPK pathways, which could shape the role of ISCs in tumor progression. Indeed, whereas Wnt-driven tumors originate from LGR5+ canonical ISCs, serrated lesions likely develop through the dedifferentiation of intestinal cells [2]. In addition, serrated CRCs encompass a population of LGR5-negative cancer cells characterized by a fetal-like gene signature driven by the Hippo-pathway effectors YAP/TAZ [2, 10, 11]. We have shown that the expression of BrafV600E suffices to enrich the fetal gene signature in an MAPK-dependent fashion [7]. However, whether cholesterol metabolism contributes to the establishment of this subtype of fetal-like cells remains to be investigated.

Finally, what is the clinical relevance of these observations? Can statin treatment reduce the incidence of BRAF-mutant CRC? Although statin administration reverses crypt hyperplasia in BRAF-mutant mice, we could not formally demonstrate an anti-tumor impact of statins in this subtype of CRC. This may be related to the inherent limitations of the BrafV600E mouse model we employed, which displays restrained tumor burden and long latency, hindering efforts aimed at assessing a direct impact on tumor development [7, 12,13,14]. Assessment of the preventive efficacy of statins in mouse models with higher disease penetrance [11,12,13,14] would resolve this conundrum. However, recent data indicate that the sensitivity of CRC cell lines to statin-induced cell death relies on an intact Bone Morphogenetic Protein (BMP) pathway [15]. CRC cell lines sensitive to statins display activation of the BMP pathway. Inhibiting the BMP pathway abolishes statin-induced apoptosis in otherwise sensitive cells. More specifically, statin-sensitive cell lines (e.g., RKO, HCT116, DLD1) express the BMP-related protein suppressor of mothers against decapentaplegic (SMAD4), which is lost/mutated in resistant cells (e.g., HT29, SW480). In a follow-up epidemiological study, no association between statin use and the risk of developing CRC with a mutation in BRAF was reported, but statins were associated with an overall reduced risk of CRC and with a larger reduction of SMAD4-positive CRC [16]. We confirmed these findings, showing that BRAFV600E CRC cells with mutations in SMAD4 are resistant to statin treatment in vitro [7]. Therefore, despite the fact that mutant BRAF drives cholesterol biosynthesis, the mutational status of SMAD4 is likely to be a more accurate predictor of the protection conferred by the use of statins than mutated BRAF. Further insights into the biochemical scenarios linking cholesterol metabolism with CRC may help refine the targeted treatment of this disease.