Biomechanical regulation of breast cancer metastasis and progression

Abstract

Physical activity has been consistently linked to decreased incidence of breast cancer and a substantial increase in the length of survival of patients with breast cancer. However, the understanding of how applied physical forces directly regulate breast cancer remains limited. We investigated the role of mechanical forces in altering the chemoresistance, proliferation and metastasis of breast cancer cells. We found that applied mechanical tension can dramatically alter gene expression in breast cancer cells, leading to decreased proliferation, increased resistance to chemotherapeutic treatment and enhanced adhesion to inflamed endothelial cells and collagen I under fluidic shear stress. A mechanistic analysis of the pathways involved in these effects supported a complex signaling network that included Abl1, Lck, Jak2 and PI3K to regulate pro-survival signaling and enhancement of adhesion under flow. Studies using mouse xenograft models demonstrated reduced proliferation of breast cancer cells with orthotopic implantation and increased metastasis to the skull when the cancer cells were treated with mechanical load. Using high throughput mechanobiological screens we identified pathways that could be targeted to reduce the effects of load on metastasis and found that the effects of mechanical load on bone colonization could be reduced through treatment with a PI3Kγ inhibitor.

Figure 1

Mechanical strain regulates gene transcription of cell adhesion, drug metabolism and proliferation genes. MDA-MB-231 breast cancer cells were treated with mechanical strain for 24 h at 0, 7.5, and 15% maximal strain at a frequency of 1 Hz. Total RNA was isolated and RNA sequencing was performed (n = 4). (A) Hierarchical clustering of the most significantly regulated genes. FPKM = fragments per kilobase of transcript per million mapped reads. (B) Venn diagrams for significantly regulated genes in the 7.5% and 15% mechanical strain groups. (C) The top five most upregulated and downregulated gene ontology groups for cells treated with 7.5% strain. (D) The top five most upregulated and downregulated gene ontology groups for cells treated with 15% strain. R software was used to create parts of this figure.

Figure 2

Mechanical strain decreases proliferation and increases drug resistance in breast cancer cells. (A) Images of Bax and Bcl-2 immunostaining for cyclic mechanical strains of 0–17.5% strain. Bar = 100 μm. (B) Relative expression of pro-apoptotic Bax to anti-apoptotic Bcl-2 protein expression. *p < 0.05 versus 0% strain (n = 10). (C) Relative gene expression of Bax to Bcl-2 after 24 h of mechanical strain (n = 10). (D) Response of MDA-MB-231 breast cancer cells to drug treatment with paclitaxel, doxorubicin, or 5-fluorouracil under 0%, 7.5%, or 15% mechanical strain (n = 8). *p < 0.05 versus static conditions. ^†p < 0.05 versus static conditions under no treatment and under static conditions with the pharmacological treatment with same concentration as the indicated group. (E) Western blotting for cells treated with 7.5% strain in combination with DMSO, asciminib (Asc; Abl1 inhibitor), radotinib (Rad; c-Abl1 inhibitor), AZD1480 (Azd; Jak2 inhibitor), Ruxolitinib (Rux; Jak1/2 inhibitor), or a PI3K inhibitor (PI3K). (F) Western blotting for cells treated with 7.5% strain and the indicated inhibitors. Full-length blots/gels are presented in Supplemental Figs. S6–S7.

Figure 3

Mechanical load enhances the adhesion of cancer cells to endothelial cells and collagen I under shear stress. (A) Diagram of the experimental design. The cells were first treated with mechanical load in a high throughput system and then adhesion measured in a high throughput flow device. (B) MDA-MB-231 breast cancer cells were mechanically strained at maximal strain from 0 to 17.5% at a frequency of 0.1 and 1 Hz for 24 h. Initial adhesion of strained cells under 0.5 dynes/cm² shear stress to a TNF-α treated endothelial monolayer was measured relative to the static group. *p < 0.05 compared to the static group (n = 8). (C) Relative adhesion of the cells treated with mechanical load after detachment shear stress up to 20 dynes/cm². (D) Adhesion of cells to endothelial cells and isolated ECM molecules including laminin (LM), vitronectin (VN), collagen I (COL I), collagen II (COL 2) and fibronectin (FN). *p < 0.05 versus control group with the same adhesion substrate. (E) Adhesion of cells to endothelial cells and ECM after detachment with shear stresses up 20 dynes/cm² (n = 8). *p < 0.05 versus control group with the same adhesion substrate. (F) Initial adhesion of cells to endothelial cells in the presence of integrin inhibitors. Adhesion is to endothelial cells treated with TNF-α unless otherwise noted. (G) Relative adhesion of cells after detachment up to 20 dynes/cm² (n = 8). Adhesion is to endothelial cells treated with TNF-α unless otherwise noted. *p < 0.05 versus the HUVEC group. ^†p < 0.05 versus the TNF-α treated HUVEC group. ^‡p < 0.05 versus the TNF-a treated HUVEC with 5% strain group.

Figure 4

High throughput mechanobiological screens for blocking load-induced enhancement of cancer cells to endothelial cells under shear stress. (A) Diagram of experimental protocol. The cells are loaded in the presence of compounds from a kinase inhibitor library and then the adhesion to inflamed endothelial cells under flow is measured. (B) Heat map of the adhesion and detachment of cancer cells with treatment with kinase inhibitors. (C) Example graphs from the kinase screen for a PI3K inhibitor and Jak3 inhibitor. (D) Results of the kinase inhibitor screen. Compounds in the lower left portion of the graph have reduced initial adhesion or τ₅₀ to endothelial cells. The τ₅₀ is an index of the strength of adhesion. It is the shear stress needed to cause detachment of 50% of the cells calculated from a curve fit to the detachment of the cells under increases shear stress.

Figure 5

Strain differentially activates Smad2/3 and Yap/Taz in breast cancer and epithelial cells. (A) MDA-MB-231 cells were treated with mechanical strain for 24 h at 1 Hz with varying magnitudes of maximal strain. Immunostaining for phospho-Smad2/3, Smad2/3 and Yap/Taz was performed. Bar = 100 μm. (B) Quantification of total Smad2/3 in the cells (n = 20). *p < 0.05 versus static group (n = 20). (C) Quantification of phosphorylated Smad2/3 in cells treated mechanical load (n = 20). *p < 0.05 versus the static group for the same subcellular location. (D) Quantification of nuclear to cytoplasmic Yap/Taz staining (n = 20). *p < 0.05 versus static group. (E-J) Western blotting for lysates from cells treated with mechanical load for 24 h and inhibitors Lck, PI3K, Syk and Yap (Vert; verteporfin). Full-length blots/gels are presented in Supplemental Figs. S12–S17.

Figure 6

Summary diagram of the mechantransduction mechanisms supported by the studies for the enhancement of survival and adhesion by mechanical load.

Figure 7

Mechanical loading decreases tumor growth and increases metastasis to the skull in immune compromised mice. (A) Luciferase expressing MDA-MB-231 cells were grown under static or mechanically loaded conditions for 24 h and then implanted into the mammary fat pad of nu/nu mice. Radiance of mice with orthotopic tumor implantation at 22 days. (B) Quantification of the radiance in the mammary fat pad. *p < 0.05 versus static group (n = 10). (C) Laser speckle imaging of mice with orthotopic implantation of MDA-MB-231 cells after mechanical conditioning. (D) Ratio of perfusion of the mammary fat pad with tumor implantation to contralateral control fat pad. *p < 0.05 between indicated groups (n = 10). (E) Images of luminescence as measured by IVIS for nu/nu given an intravenous injection of MDA-MB-231 cells. The cells were cultured under static or 7.5% strain for 24 h prior to injection. (F) Quantification of luminescence in the head of the mice given an intravenous injection of MDA-MB-231 cells at day 42 following injection. *p < 0.05 versus static group (n = 8–9). (G) Luminescence in the heads of NOD Scid mice 42 days after tail vein injection of MDA-MB-231 cells. (H) Quantification of the luminescence in the head of the mice after 42 days. *p < 0.05 versus indicated group (n = 7–8).