The forces behind EMT and tumor metastasis

Metastasis is the primary cause of mortality for cancer patients. Understanding the molecular and cellular mechanisms that underlie metastatic spread is critical for the development of effective therapeutic approaches. Acquisition of migratory and invasive properties by tumors cells is mediated by induction of Epithelial-Mesenchymal Transition (EMT). This program is orchestrated by a network of transcription factors including Twist1, Snai1/2, and Zeb1/2. Recent studies have demonstrated the requirement of a dynamic regulation of EMT during tumor progression to first promote tumor cell dissemination and then allow metastatic outgrowth. However, we currently lack a complete understanding of how this process is induced and then repressed during the metastasis cascade.¹ Past studies largely have focused on biochemical signals that induce EMT including TGF-β, hypoxia, and inflammatory cytokines. Landmark studies identified a functional role for the mechanical forces generated by stiffening matrix during tumor progression in driving tumor malignancy.^2,3 How such biomechanical cues are recognized and subsequently transduced into biochemical and transcriptional signals to modulate tumor cell behavior is unclear.

Recently, we identified Twist1 as a critical mechano-mediator that drives matrix stiffness-induced EMT.⁴ Mechanistically, at low stiffness correlating to the mechanical properties of normal breast tissue, Twist1 is sequestered in the cytoplasm and thus unable to induce an EMT transcriptional program. Increasing matrix stiffness releases Twist1 from its cytoplasmic binding partner, G3BP2, resulting in nuclear translocation of Twist1 to drive EMT. Consistent with this model, loss of G3BP2 led to constitutive nuclear localization of Twist1 and synergized with increasing matrix stiffness to induce EMT and tumor cell invasion. These findings were validated in tumor models in vivo, as loss of G3BP2 led to increased local tumor invasion and distant lung metastasis. Understanding the upstream mechanotransduction machinery that regulates the interaction between TWIST1-G3BP2 is of great interest for future studies.

An interesting finding from this study is that the mechanoregulation of Twist1 and YAP, a previously identified mechanosensitive transcription co-activator,⁵ are likely through distinct molecular mechanisms. Both Twist1 and YAP translocate from the cytoplasm to the nucleus in response to increasing matrix stiffness in an integrin-dependent manner, highlighting the critical role of integrin signaling in cellular mechanotransduction. Interestingly, in contrast to YAP mechanoregulation, Twist1 nuclear translocation in response to increasing matrix stiffness is not sensitive to changes in cell shape, epithelial polarity and adherens junctions. These results suggest the existence of multiple cellular mechanotransduction pathways. However, it remains unclear how integrins stimulation acts in conjunction with other signaling pathways to trigger an appropriate mechano-sensitive response depending on the cellular context. Delineating these and other undiscovered mechanotransduction pathways will continue to expand our understanding of how biomechanical signals from the extracellular matrix (ECM) integrate with other inputs to regulate cellular behavior.

Another interesting observation is that TGF-β signaling is not sufficient to induce EMT at low stiffness and synergizes with high matrix stiffness to induce EMT. These results are consistent with previous findings that increasing matrix stiffness regulates the response to TGF-β signaling, shifting from inducing apoptosis to EMT.⁶ This suggests that biochemical and biomechanical inputs from the microenvironment converge together to regulate EMT. Furthermore, this highlights our observation that the mechanical properties, matrix stiffness specifically, of normal breast tissue are sufficient to suppress the induction of EMT in response to a single biochemical cue. These results suggest that ECM remodeling represents a potential EMT checkpoint in breast tumor progression. Significant ECM remodeling by the stromal and/or tumor compartment may be a critical event during metastatic progression. Consistent with this hypothesis, we and others have found that the presence of disorganized collagen fibers, which is a marker of decreased matrix stiffness, correlates with improved patient survival.⁷

Finally, of particular interest is whether mechanoregulation is generalizable to multiple tumor types beyond breast cancer. Given that different normal tissue types have different mechanical properties, how the increasing matrix stiffness of neoplastic tissue affects cell behavior in each of these contexts remains to be determined. Breast cancer may be particularly susceptible to biomechanical changes as the matrix stiffness of normal breast tissue is very low. Alternatively, the relative increase in matrix stiffness in tumor tissue may be what determines cellular response, and thus ECM remodeling leading to increased matrix stiffness will be a critical input in multiple tumor types. In particular, pancreatic tissue is known to become highly fibrotic as the tumor progresses. Recent evidence show that, even if driven by a specific genotype, pancreatic lesions are likely to evolve, through ECM stiffening, to oncogene-independent progression to PDAC. Whether cells sense absolute or relative changes in matrix stiffness, and how cellular identity defines the response to these changes are critical open questions with significant implications for the role of mechanical properties of the tumor microenvironment.

In conclusion, the emerging field of mechanotransduction has identified previously unrecognized critical inputs for tumor progression and metastasis. The identification of Twist1 as a critical mechano-mediator expands our understanding of the functional role that mechanical signals from the tumor microenvironment play. Continued investigation of mechanotransduction mechanisms will be critical for the rational development of effective therapeutic approaches and a more complete understanding of the molecular and cellular processes underlying tumor progression.