Adhesion to the Brain Endothelium Selects Breast Cancer Cells with Brain Metastasis Potential

Abstract

Tumor cells metastasize from a primary lesion to distant organs mainly through hematogenous dissemination, in which tumor cell re-adhesion to the endothelium is essential before extravasating into the target site. We thus hypothesize that tumor cells with the ability to adhere to the endothelium of a specific organ exhibit enhanced metastatic tropism to this target organ. This study tested this hypothesis and developed an in vitro model to mimic the adhesion between tumor cells and brain endothelium under fluid shear stress, which selected a subpopulation of tumor cells with enhanced adhesion strength. The selected cells up-regulated the genes related to brain metastasis and exhibited an enhanced ability to transmigrate through the blood-brain barrier. In the soft microenvironments that mimicked brain tissue, these cells had elevated adhesion and survival ability. Further, tumor cells selected by brain endothelium adhesion expressed higher levels of MUC1, VCAM1, and VLA-4, which were relevant to breast cancer brain metastasis. In summary, this study provides the first piece of evidence to support that the adhesion of circulating tumor cells to the brain endothelium selects the cells with enhanced brain metastasis potential.

Keywords: biomechanics; brain metastasis; endothelial adhesion; fluid shear stress; mechanobiology.

Figure 1

Brain-endothelium-adhesion-based selection of breast cancer cells. (a) The illustration of in vitro system for brain-endothelium-adhesion-based selection of breast cancer cells (created with biorender.com). (b,c) The adhesion of tumor cells to the brain endothelium under various levels of fluid shear stress. WT cells were labeled with red cell tracker and perfused into the flow chamber slides to adhere to the brain endothelium under 0.25, 0.5, 1, 2.5, and 5 dyn/cm² shear stress for 15 min. The number of adhered cells was counted under a fluorescence microscope. Scale bar = 100 μm. n = 3. (d,e) The FAS cells exhibited enhanced adhesion ability under shear stress. WT cells (labeled with red cell tracker) and FAS cells or SAS cells (labeled with green cell tracker) were equally mixed and perfused into the flow chamber at the rate of 1 dyn/cm² to adhere to the brain endothelium for 15 min. The number of adhered cells was counted under a microscope and normalized to the WT group. Scale bar = 100 μm. n = 3. All data are represented by mean ± SEM. The statistics among groups were analyzed using one-way ANOVA with the post hoc Bonferroni test. (ns: no significance, *** p < 0.001, **** p < 0.0001).

Figure 2

FAS cells up-regulate brain metastasis genes. (a) FAS cells showed enhanced expressions of brain metastasis-related genes. The gene expression was evaluated using qPCR. The statistics were calculated using two-way ANOVA with the post hoc Tukey test. n = 3. (b,c) The expressions of COX2 and SerpinB2 in all groups were tested using immunofluorescence staining. Scale bar = 10 μm. (d,e) Quantification of the fluorescence intensity of COX2 and SerpinB2 in (b,c). The statistics among three groups were calculated based on one-way ANOVA with the post hoc Bonferroni test. n = 50. (ns: no significance, * p < 0.05, *** p < 0.001, **** p < 0.0001).

Figure 3

FAS cells exhibit enhanced adhesion to the brain endothelium and BBB transmigration ability. (a,b) FAS cells exhibited enhanced adhesion strength on the endothelium. All groups were labeled with green cell tracker and added into the flow chamber slides to adhere to the brain endothelium for 30 min. Then, 0.25, 0.5, 1, and 2 dyn/cm² wall shear stress were applied to each slide for 15 min, respectively. The cancer cells remaining on the brain endothelium after each treatment of fluid shear stress were counted using the fluorescence microscope. Two-way ANOVA along with post hoc Tukey test were used to calculate the statistics. Scale bar = 100 μm. n = 3. (c) The illustration of trans-endothelial migration assay (created by biorender.com). (d,e) FAS cells exhibited enhanced BBB transmigration ability. An hCMEC/D6 monolayer was cultured on the top of the insert membrane and cancer cells were added to transmigrate through the monolayer to the lower chamber. The transmigrated cancer cells (marked by the green cell tracker) were imaged and counted using fluorescence microscopy. One-way ANOVA along with post hoc Bonferroni test was performed to analyze the statistics. Scale bar = 50 μm. n = 3. All data are represented by mean ± SEM. (ns: no significance, ** p < 0.01, **** p < 0.0001).

Figure 4

FAS cells exhibit advantages in cell adhesion and survival within the soft brain environment. (a–c) FAS cells had an enhanced adhesion ability on 0.6 kPa soft matrices. The same number of cancer cells from the WT group (labeled with red cell tracker) was mixed with cancer cells from the FAS group or the SAS group (labeled with green cell tracker) and seeded on the same 0.6 kPa polyacrylamide gels coated with collagen I. After 15 min, these cells were gently washed with PBS. This illustration was created using Biorender.com (a). The remaining cells were imaged and counted under fluorescence microscope. Scale bar = 100 μm. n = 3. (d,e) FAS cells exhibited lower cell apoptosis on soft matrices. All groups were seeded on 0.6 kPa polyacrylamide gels coated with collagen I in low-FBS medium overnight. The dead cells were then marked with PI and tested through flow cytometry. n = 3. All data are represented by mean ± SEM. The statistics among three groups were calculated based on one-way ANOVA with the post hoc Bonferroni test. (** p < 0.01, **** p < 0.0001).

Figure 5

The FAS cells up-regulate multiple adhesion molecules that are clinically relevant to breast cancer brain metastasis. (a) FAS cells had enhanced expressions of brain metastasis-related adhesion genes. qPCR was conducted to test the gene expression. The statistics were analyzed based on two-way ANOVA along with the post hoc Tukey test. n = 3. (b) The highly expressed adhesion molecules in the FAS cells were clinically related to brain metastasis. The gene expression data were collected from the GEO public database (GSE173661, n = 25 samples). The paired comparison was analyzed using Graphpad. For each gene, a paired Student’s t-test was used to analyze the difference between the statistics in two groups. (c,d) The expressions of MUC1 and VCAM1 were tested through immunofluorescence staining. Scale bar = 10 μm. (e,f) Quantification of the expressions of MUC1 and VCAM1 in (c,d). The statistics among three groups were calculated based on one-way ANOVA with the post hoc Bonferroni test. n = 50. All data are represented by mean ± SEM. (ns: no significance, **** p < 0.0001).