Findings Suggest Cholesterol Plays a Surprising Role in Pancreatic Cancer

Although a majority of pancreatic cancer cells are highly dependent on endogenous biosynthesis of cholesterol, new research from investigators at Fox Chase Cancer Center has demonstrated that some more aggressive pancreatic cancers are completely independent of this process.

Disruption of cholesterol biosynthesis by Nsdhl knockout or treatment with statins transforms glandular pancreatic carcinomas in mice to the more aggressive basal subtype via the activation of SREBP1. When SREBP1 is cleaved and goes to the nucleus, this induces Tgfb1 expression, autocrine TGF-ß–SMAD2/3 signaling, and epithelial-mesenchymal transition (EMT).

A recent multifaceted investigation by a group of scientists at Fox Chase and collaborators across the United States proved this finding and published the data in “Cholesterol Pathway Inhibition Induces TGF-β Signaling to Promote Basal Differentiation in Pancreatic Cancer,” which appeared in the October 2020 issue of the journal Cancer Cell.¹

Navigating the Cholesterol Pathway

Pancreatic ductal adenocarcinoma (PDAC) is poised to become the second-leading cause of cancer death in the United States by 2030. The median survival for patients with basal PDAC is much lower than that of the classical subtype (6.3 months vs 10.4 months, respectively), according to data from the COMPASS study (NCT02750657).²

There are 2 major types of pancreatic cancer cells: classic, or glandular, and the more aggressive and treatment-resistant basal. Defined on the basis of transcriptional profiling, classic pancreatic cancer cells tend to grow in clusters and maintain features of epithelial differentiation. The basal variant features cancer cells that look like fibroblasts and are notoriously resistant to chemotherapy, resulting in far shorter life expectancies for patients (Figure).

To understand the metabolic dependencies of pancreatic cancer, we initially hypothesized that, by blocking biosynthesis of cholesterol in pancreatic cancer cells, we could suppress tumor development. We took advantage of a genetic mouse model in which we could turn off one of the cholesterol biosynthetic genes, Nsdhl. This protection happened in mice, in which cancer cells retained a single copy of the tumor-suppressor Trp53 gene. However, when the Trp53 gene was completely removed, the protection was lost and, surprisingly, the pancreatic cancer switched from classic to the basal subtype.

To replicate the genetic experiment with a pharmacological cholesterol pathway inhibitor, we next treated mice with a commonly prescribed atorvastatin (Lipitor). In this experiment, we found that tumors pharmacologically deprived of cholesterol exhibit more aggressive, sarcomatoid histology. Analyses by single-cell RNA sequencing were illuminating and allowed us to decipher a potential culprit mechanism that was responsible for the cholesterol pathway–regulated cancer cell differentiation switch.

In Nsdhl-deficient murine pancreatic cancers, we observed remarkable increases in the population of mesenchymal carcinoma cells (we dubbed as EMT cells), which constituted more than 80% of the entire cancer cell population. We noted that in Nsdhl-deficient cancer cells, which maintained epithelial differentiation features, there was a significant increase in the expression of Tgfb1—a growth factor responsible for EMT and a hallmark feature of the EMT subset of pancreatic cancer cells.

Our next task was to explain how the cholesterol pathway blockade could regulate TGF-ß signaling. In vitro depletion of cholesterol in cancer cells activated a transcriptional factor SREBP1 and unexpectedly induced Tgfb1 expression. This increased amount of secreted TGFB1 in the culture media triggered the canonical TGF-ß signaling cascade in PDAC cells.

We then turned our attention to 55 pancreatic tumor tissue samples of previously untreated patients who had undergone surgery for localized pancreatic adenocarcinoma. We compared a sample of patients who had not taken statins with 14 patients who were routinely taking statins until the day of surgery. The abundance of EMT cells and levels of TGF-ß signaling in patients taking statins was inversely proportionate to the level of lipids in the blood.

If patients exhibited high lipids, their percentage of EMT cells was low; if their cholesterol was low, there was an increase in EMT cells. These results suggest that blood lipids and blood cholesterol may be involved in the regulation of the emergence of more aggressive pancreatic cancer cells in tumors.

Clinical Implications

Because basal tumors are inherently treatment-resistant and aggressive, identifying this new metabolic regulator that can promote the EMT and basal conversion—which is ultimately responsible for chemotherapy resistance and metastatic spread—should help investigators overcome a major obstacle for anticancer therapy. Based on these data, we also obtained a greater appreciation of the intratumoral clonal heterogeneity, in that a given human or mouse cancer consists of multiple clones of cells. As in mice, human tumors contain both mesenchymal (or EMT) and epithelial carcinoma cells, as well as some subset of cells in between—a whole spectrum of clones.

When a tumor is subjected to chemotherapy and the patient has taken statins, that combination may favor the growth of potentially more aggressive cancer cell clones. We are just beginning to understand this “clonal competition”; the more we supply selective pressure to pancreatic cancer cells, the more one type of cell predominates. Uncovering their respective sensitivities to treatments and drugs will deepen our understanding of the clonal shifts.

Our results may be relevant for patients who should not continue taking statins when combating advanced metastatic cancers. In fact, I have started critically reviewing my patients’ medication lists and stopping statins if the potential benefits become irrelevant in lieu of their advanced cancer diagnosis, especially if their cancer carries Trp53 mutations.

Another unanswered question: Could statins act differently in the setting of wildtype versus mutant Trp53? The answer could potentially explain why no clear relationship has been established between statin use, PDAC incidence, and survival. Even more speculatively, a precipitous drop in blood lipids, which we sometimes observe in cases of poor nutrition and cachexia, could contribute to cancer aggressiveness.

To obtain greater clarity regarding the relationship between diet, medication, and pancreatic cancer aggressiveness, we are beginning to look at a larger collection of human samples to determine how blood lipids correlate with a patient’s nutrition, medicines, and percentage of EMT cells in their tumors.

1. Gabitova-Cornell L, Surumbayeva A, Peri S, et al. Cholesterol pathway inhibition induces TGF-ß signaling to promote basal differentiation in pancreatic cancer. Cancer Cell. 2020;38(4):567583.e11. doi:10.1016/j.ccell.2020.08.015
2. Aung KL, Fischer SE, Denroche RE, et al. Genomics-driven precision medicine for advanced pancreatic cancer: early results from the COMPASS trial. Clin Cancer Res. 2018;24(6):13441354. doi:10.1158/1078-0432.CCR-17-2994