Collagen Fiber Alignment and Internal Architecture Cooperatively Regulate Breast Cancer Cell Migration in a Subtype-Specific Manner

Introduction: : Breast cancer accounts for approximately 30% of all newly diagnosed cancers annually, making it the most prevalent cancer among women. Metastatic breast cancer involves the migration of tumor cells through the extracellular matrix (ECM) and into the circulatory system, where they can form secondary tumors at distant sites. Collagen, a primary structural protein of the ECM, plays a critical role in this process. In breast cancer, the presence of highly aligned collagen fibers oriented perpendicularly to the tumor-stromal boundary has been identified as a prognostic indicator of metastasis (TACS3). This alignment can be quantitatively assessed using the Coefficient of Alignment (CoA), defined as the percentage of collagen fibers oriented within 15 degrees of the dominant (mode) fiber angle. Additionally, the forward-to-backward scattered second harmonic generation (F/B) ratio, which represents the fiber internal structure (FIS), has been correlated with patient survival in invasive ductal carcinoma. Elevated CoA values have independently been associated with increased metastatic potential, while the FIS (via the F/B ratio) has also been linked to clinical outcomes in a subtype-specific manner: higher F/B ratios predict poorer survival in basal-like breast cancers, whereas lower F/B ratios are associated with worse prognosis in luminal subtypes. Although alignment and FIS independently predict metastatic risk, the combined impact of fiber alignment and internal structure on tumor cell motility remains poorly understood. To address this gap, this study uses an engineered in vitro ECM model to study the hierarchical or synergistic relationships between these collagen-related properties.

Materials and

Methods: : Using the previously described microfluidic device with stepwise contracting geometry [1], we fabricated 3D collagen gels with CoA and F/B ratios representing the values identified from banked patient tumor samples. Confocal reflectance microscopy (CRM) and second harmonic generation (SHG) microscopy were used to quantify collagen fiber alignment and fiber internal structure, respectively. To examine the combined effects of collagen alignment and internal structure on cell migration, six experimental conditions were generated by combining two levels of CoA (unaligned: < 0.5; aligned: > 0.5) with three levels of F/B (low: < 6; medium: 6–10; high: > 10). Cell migration experiments were performed using MDA-MB-231(luminal subpopulation) and MCF7 (basal subpopulation) breast cancer cells. Time-lapse microscopy was used to capture cell migration trajectories. A custom MATLAB script reoriented all tracks to a common origin (0,0) and fitted an ellipse encompassing 99% of the trajectories to represent the overall migration pattern; the aspect ratio of this ellipse was used as a metric of “directional migration”. Additionally, trajectories were analyzed using the Persistent Random Walk (PRW) model to quantify migration speed and persistence. To ensure that collagen stiffness and pore size were not confounding variables, mechanical stiffness and pore characteristics were quantified across all conditions.

[1] A. Ahmed et al., “Microengineered 3D Collagen .,” Adv. Mater. Technol 2021, Doi: 10.1002/admt.202001186.

Results, Conclusions, and Discussions:: To investigate the role of extracellular matrix architecture in regulating cell migration, six experimental conditions were created by systematically varying collagen fiber alignment (CoA) and internal structure (via F/B), while controlling for stiffness and pore size. Mechanical characterization confirmed no significant differences in stiffness or porosity across conditions (Fig. 1), confirming that observed migration behaviors were due to CoA and F/B. Directional migration, quantified via the aspect ratio of ellipses fit to reoriented cell trajectories, revealed cell line–specific responses. In MDA-MB-231 cells, both CoA and F/B independently influenced migration directionality, with no interaction effect—suggesting a hierarchical relationship between these matrix features (Fig. 2, A-B). In contrast, MCF7 cells exhibited no significant changes in directional migration in response to either CoA or F/B, indicating reduced sensitivity to matrix-guided directionality (Fig. 2, C-D). Migration speed analyses further indicated that FIS was the primary regulator of motility in MDA-MB-231 (luminal) cells, while CoA had no measurable effect (Fig. 3, A-B). Again, no interaction was observed, indicating a dominant role for FIS. Conversely, in MCF7 cells, both CoA and F/B contributed independently to speed without synergy (Fig. 3, C-D), suggesting these matrix features can modulate luminal cell motility even when directional guidance is minimal.
Collectively, these findings reveal subtype-specific differences in sensitivity to ECM structural features between basal-like and luminal breast cancer cells. MDA-MB-231 cells exhibited a greater reliance on collagen organization for directed migration, indicating that ECM architecture plays a more prominent role in their motility. In contrast, MCF7 cells showed limited directional response to matrix cues, suggesting reduced guidance by physical matrix features, though their migration speed was still influenced by the collagen environment. These results emphasize the importance of ECM structural features in shaping migration behavior in a subtype-specific manner. By dissecting how cancer cells interpret CoA and F/B cues, this study contributes to a deeper understanding of tumor–microenvironment interactions. It also motivates future studies to determine the extent to which ECM structural features contribute to predicting metastatic propensity in distinct cancer subtypes.

Acknowledgements and/or References (Optional)::