Occurrence of Biased mTOR Signaling in Hepatocellular Carcinoma

Biased mTOR Signaling Identified as a Key Driver in Hepatocellular Carcinoma Progression

In oncology, the mTOR (mammalian target of rapamycin) pathway is often viewed as a monolithic switch. When it is "on," the cell grows and builds mass. When it is "off," those processes stall. Because hyperactivation of this pathway is a hallmark of many cancers, it has long been a primary target for drug development. However, in hepatocellular carcinoma (HCC)—a highly lethal liver cancer—simply turning the switch off hasn't always worked.

A new study suggests we have been looking at the switch incorrectly. Instead of a simple binary state, the researchers found that HCC employs "biased mTOR signaling." In this state, the pathway doesn't just turn on. It skews its activity toward specific downstream branches while suppressing others. This selective redirection allows cancer cells to optimize their metabolism for survival. This makes traditional single-target therapies less effective.

The Failure of Monolithic Inhibition

The challenge in treating HCC lies in its molecular heterogeneity (diversity in genetic drivers). Because the disease arises from different causes, the underlying drivers are not uniform. While 40–50% of HCC cases show elevated mTOR activity, the complexity of the pathway makes it difficult to predict which specific arm fuels tumor growth.

Current strategies often rely on "rapalogs"—drugs like Rapamycin that specifically inhibit mTORC1 (mechanistic target of rapamycin complex 1). mTORC1 is the branch responsible for ribosome biogenesis (building protein machinery) and translation. However, the efficacy of these drugs is often limited by resistance. If a cancer cell can bypass the blocked mTORC1 branch by using other components, the drug fails to halt progression. The authors note that the success of targeted therapies is often undermined by the pathways they attempt to suppress.

Decoding the Skewed Signaling Architecture

To investigate this, the researchers established a rapid preclinical model using C57BL/6J mice. By administering a single dose of the carcinogen DEN followed by CCl4 (carbon tetrachloride), they created a "DCI" model. This model develops HCC within just four months .

This allowed them to observe disease progression over time.

The study identifies a sophisticated redistribution of signaling activity termed "biased mTOR signaling." Rather than a global surge in activity, the tumor shows a paradoxical split:

1. Downregulation of the Canonical Axis: The researchers observed a suppression of the mTORC1-S6K-S6 signaling axis. Specifically, they reported a 4-fold reduction in S6K activity .

They also found a 2-fold reduction in the phosphorylation (the chemical activation) of S6 and AKT. 2. Upregulation of Alternative Branches: Simultaneously, other downstream targets were heavily recruited. The study found significant increases in the phosphorylation of mTORC1-ULK1, mTORC1-4EBP1, and the mTORC2-PKCα axis . 3. Transcriptional Driving Forces: This wasn't just a change in protein activity. The total levels of these substrates—ULK1, 4EBP1, and PKCα—were increased by 3- to 5-fold . This was driven by increased mRNA expression (the instructions for making proteins).

This mechanism suggests the cancer cell is re-engineering the pathway. By increasing the abundance of specific substrates like ULK1 (involved in autophagy, or cellular recycling) and 4EBP1 (involved in translation regulation), the cell biases mTOR activity toward specific metabolic tasks.

Evidence from Mice and Human Cohorts

The researchers validated these findings by bridging mouse data with human clinical data. Using the TCGA-LIHC dataset, which includes 365 liver cancer patients, they performed survival analyses. They wanted to see if these "biased" markers actually mattered for patient outcomes.

The results were striking. The authors report that elevated levels of EIF4EBP1 and ULK1 were strongly associated with a rapid decline in 5-year progression-free survival (PFS) in HCC patients [Figure S4]. Furthermore, the overexpression of PKCα showed a robust association with poor survival in both HCC and wider pan-cancer cohorts [Figure S5]. This confirms that the skewed signaling in the mouse model is a clinically relevant feature of human malignancy.

Testing different inhibition strategies provided the final evidence. In the 6w-DCI mouse model, Rapamycin failed to inhibit tumor growth .

In contrast, Torin 1—a dual inhibitor that targets both mTORC1 and mTORC2—significantly reduced tumor burden . It behaved similarly to the FDA-approved multi-kinase inhibitor Sorafenib .

Limitations in the Mechanistic Narrative

The study leaves several questions unanswered. First, the longitudinal analysis could not definitively prove that fibrosis (scarring of liver tissue) causes HCC progression. Because the researchers could not capture a stage with significant fibrosis but no tumors, they could only conclude that fibrosis and tumor growth are correlated .

Second, the utility of standard serum markers (blood tests) for liver function remains limited. Markers like albumin, cholesterol, and alanine transaminase (ALT) increased during intermediate disease phases. However, they failed to reliably detect the earliest stages of HCC development . This suggests clinicians cannot yet rely on routine blood work to catch these tumors early. Finally, while ULK1 showed a strong connection to HCC progression, its impact on overall survival across broader pan-cancer cohorts was relatively modest.

The Verdict: Dual Targeting is Essential

The evidence suggests the "monolithic switch" model of mTOR signaling is insufficient for treating liver cancer. Because HCC cells use biased signaling to bypass mTORC1-specific inhibitors, the therapeutic focus must shift.

The study demonstrates that targeting both the mTORC1 and mTORC2 complexes is necessary to blunt tumor progression. For practitioners and researchers, this implies that clinical trials focusing only on rapalogs may be mismatched against the signaling reality of HCC. Moving forward, characterizing a patient's specific signaling bias will likely be essential for successful targeted therapy.

Figures from the paper