A mechanically-induced colon cancer cell population shows increased metastatic potential

Abstract

Background: Metastasis accounts for the majority of deaths from cancer. Although tumor microenvironment has been shown to have a significant impact on the initiation and/or promotion of metastasis, the mechanism remains elusive. We previously reported that HCT-8 colon cancer cells underwent a phenotypic transition from an adhesive epithelial type (E-cell) to a rounded dissociated type (R-cell) via soft substrate culture, which resembled the initiation of metastasis. The objective of current study was to investigate the molecular and metabolic mechanisms of the E-R transition.

Methods: Global gene expressions of HCT-8 E and R cells were measured by RNA Sequencing (RNA-seq); and the results were further confirmed by real-time PCR. Reactive oxygen species (ROS), anoikis resistance, enzyme activity of aldehyde dehydrogenase 3 family, member A1 (ALDH3A1), and in vitro invasion assay were tested on both E and R cells. The deformability of HCT-8 E and R cells was measured by atomic force microscopy (AFM). To study the in vivo invasiveness of two cell types, athymic nude mice were intra-splenically injected with HCT-8 E or R cells and sacrificed after 9 weeks. Incidences of tumor development and metastasis were histologically evaluated and analyzed with Fisher's exact test.

Results: Besides HCT-8, E-R transition on soft substrates was also seen in three other cancer cell lines (HCT116, SW480 colon and DU145 prostate cancer). The expression of some genes, such as ALDH3A1, TNS4, CLDN2, and AKR1B10, which are known to play important roles in cancer cell migration, invasion, proliferation and apoptosis, were increased in HCT-8 R cells. R cells also showed higher ALDH3A1 enzyme activity, higher ROS, higher anoikis resistance, and higher softness than E cells. More importantly, in vitro assay and in vivo animal models revealed that HCT-8 R cells were more invasive than E cells.

Conclusions: Our comprehensive comparison of HCT-8 E and R cells revealed differences of molecular, phenotypical, and mechanical signatures between the two cell types. To our knowledge, this is the first study that explores the molecular mechanism of E-R transition, which may greatly increase our understanding of the mechanisms of cancer mechanical microenvironment and initiation of cancer metastasis.

Figure 1

Multiple cancer cells lines show E-R transition on appropriate soft substrates. (a) HCT-116 cells cultured on 10 kPa PA gel substrates (coated with fibronectin) form cell colonies in 2–5 culture days. (b) HCT-116 cells begin to dissociate from colonies on the 10th day. On PS substrate or other stiffness PA gels under same culture condition, they do not show dissociation. (c) DU-145 cells cultured on 10 kPa PA gel substrates (coated with fibronectin) form cell colonies in 3–7 culture days. (d) DU-145 cells begin to dissociate from colonies on the 19th day. On PS substrate or other stiffness PA gels under same culture condition, they do not show dissociation. Scale bar: 100 μm.

Figure 2

Summary of differentially expressed genes in HCT-8R cells as compared to E cells. RNA-Seq analyses were performed as described in Methods. The data are expressed as fold expression. The bar graph represents the average fold expression changes of R cells compared to E cells cultured under different conditions (PS3, Gel3, and PS17) that do not allow an E to R transition. The graphical and tabulated data summarize the up-regulated (green) as compared to down-regulated (red) genes. The composite results suggest that, most of the differentially up-regulated genes in R cells are associated with the functionalities of cell proliferation, motility, metabolism, invasive phenotype, colorectal adenocarcinoma and tumor metastasis. The differentially down-regulated genes in R cells are associated with tumor suppression and inhibition of apoptosis.

Figure 3

Verification of selected gene and protein expression. Verification of selected gene and protein expression was performed using RT-qPCR and direct ALDH enzyme-staining assays as described in Methods. (a) RT-qPCR was used to verify overexpression of selected genes identified by RNA-Seq analyses. The blue, pink, and green bars represent 3 separate experiments each consisting of 3 replicate samples for each primer set. One-way ANOVA analyses indicate the mean value increases in gene expression between E and R cells in all experiments for TENS4 (p < 0.03), ALDH3A1 (p < 0.01), CLDN-2 (p < 0.01) and AKR1B10 (P < 0.005) represent statistically significant differences. (b-c) fluorometric ALDH enzyme assays were used to directly measure the relative expression of ALDH in R and E cells in cell monolayers undergoing E-R transition on soft substrates. 90% of R cells, as well as some E cells in the cell islands undergoing E-R transition express relatively high amounts of ALDH activity. Scale bar: 100 μm. (d) The integrated cellular fluorescence intensity of ALDH for R cells was 4–5 fold higher as compared to E cells.

Figure 4

Biophysical properties of E and R cells. (a-b) Reactive oxygen species of HCT-8 E and R cells detected by staining with 5-(and-6)-carboxy-2′, 7′-dichlorodihydrofluorescein diacetate. Left: Fluorescent pictures. Right: DIC pictures. Scale bar: 20 μm. (c) Comparison of ROS production in E and R cells. (d) Comparison of basement membrane invasiveness of E and R cells. R is HCT-8 R cells harvested from 20 kPa gels on after 17 days culture. E is HCT-8 cells cultured on PS for 10 days. (e) Phase-contrast photomicrographs of HCT-8 R (top) and E cells (bottom) cultured on PS prior to harvesting and invasion assay. Scale bar: 100 μm. (f) Assembled 4x4 picture set of invaded HCT-8 E cells after 48 hr incubation. Assembled 4x4 picture set of invaded HCT-8 R cells after 48 hr incubation (purple foci indicates invaded cells). (g-h) Comparison of the anchorage independent growth and viability of HCT-8 E (g) and R cells after culture on non-functionalized gels for 8 days. The percentages of viable E and R cells were determined by Trypan-blue staining (i). Red arrows in (g) indicate the dead cells. (j) Atomic force microscopy measurement of stiffness of HCT-8 cells before and after E-R transition. The square-dot and circle-dot curves represent the force vs. indentation displacement of HCT-8 E and R cells, respectively. The data were fitted using an improved Hertz model to extract the Elastic modulus of cells. (j) Comparison of the cell elasticity of E and R cells. (k) Comparison of actin organization in E and R cells on PS. Rhodamine phalloidin (520/650, red) was used specifically stain F-actin filaments. The spatial fluorescent intensity distributions of actin expression across E and R cells body are indicated by dashed lines a and b). Scale bar: 10 μm.

Figure 5

Macroscopic and microscopic aspects of tumors. Macroscopic illustration of tumors in the spleen (a & d) and liver (b & e) of HCT8 cell-injected nude mice. c & f illustrate the similar histological growth pattern and cellular morphology of HCT-8E (c) and HCT-8R (f) cells in the liver (H&E staining, dots show the division of the liver tissue on the top and tumor on the bottom).