Cellular memory function from 3D to 2D: Three-dimensional high density collagen microfiber cultures induce their resistance to reactive oxygen species

Abstract

Cell properties generally change when the culture condition is changed. However, mesenchymal stem cells cultured on a hard material surface maintain their differentiation characteristics even after being cultured on a soft material surface. This phenomenon suggests the possibility of a cell culture material to memorize stem cell function even in changing cell culture conditions. However, there are no reports about cell memory function in three-dimensional (3D) culture. In this study, colon cancer cells were cultured with collagen microfibers (CMF) in 3D to evaluate their resistance to reactive oxygen species (ROS) in comparison with a monolayer (2D) culture condition and to understand the effect of 3D-culture on cell memory function. The ratio of ROS-negative cancer cells in 3D culture increased with increasing amounts of CMF and the highest amount of CMF was revealed to be 35-fold higher than that of the 2D condition. The ROS-negative cells ratio was maintained for 7 days after re-seeding in a 2D culture condition, suggesting a 3D-memory function of ROS resistance. The findings of this study will open up new opportunities for 3D culture to induce cell memory function.

Keywords: 3D culture; Cancer tissue; Cellular memory function; Collagen microfibers; Reactive oxygen species.

Graphical abstract

Fig. 1

(A) Fabrication process of CMF, collagen sponge was homogenized at 0.1 mg mL⁻¹ in 85 % ethanol for 6 min and in 70 % ethanol for 1 min subsequently at 4 °C. The obtained CMF was washed with ultra-pure water and lyophilized for 4 days. (B) Schematic illustration of “memory function” of tumor cells isolated from CMF 3D composed of CMF and colorectal cancer cells (HCT116). HCT116 were pre-cultured for 7 days in three different conditions, CMF 3D, spheroid and 2D respectively. The isolated cancer cells were cultured in a monolayer condition for 11 days and the consumed amount of ROS added in culture medium was measured during the culture period to evaluate ROS resistance.

Fig. 2

Structural analyses of CMF. (A) Phase-contrast (Ph) images of collagen sponge fibers and CMF dispersed in water after lyophilization, respectively. (B) Optical microscope (OM) and SEM images of lyophilized collagen sponge and CMF, respectively. (C) Fiber diameter of collagen sponge and CMF calculated from SEM images (n = 6). (D) Fluorescence spectra of 71.4 μM ANS-Mg in CMF and native collagen solution in PBS at 3.5 mg mL⁻¹ at 25 °C. Excitation wavelength was 380 nm. (E) CD spectra for CMF, native collagen and gelatin dissolved in 5 μM acetic acid solutions at 5 μg mL⁻¹ at 25 °C. Statistical comparison between two groups was analyzed by Student's t-test. p value, ***<0.001 denotes statistically significant.

Fig. 3

Developing 3D tissue using CMF. (A) Schematic illustration of the development of CMF 3D tissue. CMF was mixed with a cell suspension and centrifuged at 1150×g for 15 min. (B) HE staining images of 3D tissue using 5 mg of CMF mixed with 1.0 × 10⁶ of NHDF and HCT116. (C) Cell viabilities in three culture methods (2D, spheroid, and CMF 3D) based on the cell count after day 1 of culture, relative to the initial seeding cell number. The cell count was calculated from DNA quantity normalized with the calibration curve against the cell number. Statistical comparisons were performed by one-way ANOVA followed by post hoc Tukey-Kramer multiple comparison tests. “n.s.” denotes no significant differences.

Fig. 4

Evaluation of ROS resistance in CMF 3D. (A) Elastic moduli of CMF tissue constructed under 1, 5, and 10 mg CMF and 1.0 × 10⁶ of HCT116 cells at 25 °C (n = 6). p value, ***<0.001 denotes statistically significant. (B) Comparison of histograms of FACS analysis using CellROX™ Deep Red Reagent between 2D, spheroid and CMF 3D culture in HCT116. Cells were cultured using three methods for 7 days and then isolated from the scaffolds. Cells were treated with 100 μM of tert-Butyl hydroperoxide (TBHP) and double stained with 1 μM SYTOX™ Blue Nucleic Acid Stain and CellROX™ Deep Red Reagent. Cells below a threshold in the 448/45 channel were defined as live cells and cells below a threshold in the 660/10 channel among live cells were defined as ROS-negative cells. To determine the threshold, 2D cultured cells without TBHP were used as a negative control. (C) The population of ROS-negative cells in total cells via gating of doublet removal of HCT116 and MDA-MB-231 cultured in 2D, spheroid and 1, 5, and 10 mg of CMF 3D for 7 days (n = 3). *** denotes that “CMF 10 mg” is statistically significant at p < 0.01 compared to all other samples. ** denotes that “CMF 5 mg” is statistically significant at p < 0.05 compared to “CMF 1 mg” and p < 0.01 compared to “2D” and “Spheroid”. * denotes that “CMF 1 mg” is statistically significant at p < 0.05 compared to “2D”. ††† denotes that “CMF 10 mg” is statistically significant at p < 0.01 compared to “2D” and “Spheroid”. †† denotes that “CMF 5 mg” is statistically significant at p < 0.01 compared to “2D” and “Spheroid”. † denotes that “CMF 1 mg” is statistically significant at p < 0.01 compared to “2D” and “Spheroid”. (D) Quantitative-PCR analysis of GSH related genes, GSR and GCL analyzed by the ΔΔCt method from HCT116 and MDA-MB-231 cultured in 2D, spheroid and CMF 3D (5 mg) for 7 days (n ≥ 3). Each target was normalized using PPIA as an internal control. Target expressions were compared as a percentage against 2D culture values. p value, * <0.05 denote statistically significant. (E) Fluorescence images of WB signals of GSR, GCL and β-tubulin as an internal control from HCT116 cultured in 2D, spheroid and CMF 3D culture for 7 days (n = 3). Full blotting image was displayed in Figure S5. of the Supporting Information. (F) Comparison of GSR and GCL protein expression of HCT116 cultured in 2D, spheroid and CMF 3D (5 mg) for 7 days (n = 3). Each protein expression was calculated from fluorescence intensity from WB images using ImageJ and normalized using β-tubulin as an internal control. Target expressions were compared as a percentage against 2D culture values. p value, * <0.05 denote statistically significant. Statistical comparisons were performed by one-way analysis of variance (ANOVA) followed by post hoc Tukey-Kramer multiple comparison tests.

Fig. 5

Evaluation of hypoxic condition in CMF 3D. (A) Comparison of Partial O₂ pressure in tissue using an Oxoplate® (n = 3). HCT116 were seeded on each well of an Oxoplate in 2D, spheroid and CMF 3D. The bottom fluorescence intensity of the plate was measured. Fluorescence intensity at Ex. 540 nm/Em 650 nm was used as an indicator, while Ex. 540 nm/Em 590 nm was used as a reference. Both were used for calculation of O₂ saturation (%) after calibration using O₂ saturated water and O₂ free water at 25 °C. Oxygen partial pressure (mmHg) was calculated from oxygen saturation at 20 °C and atmospheric pressure of 1013 hPa. (B) Quantitative-PCR analysis of hypoxia related genes, HIF1α, SLC2A1, BNIP3 and PDK1 analyzed by the ΔΔCt method from HCT116 cultured in 2D, spheroid and CMF 3D (5 mg) for 7 days (n = 3). Each target was normalized using PPIA as an internal control. Target expressions were compared as a percentage against 2D culture values. (C) Immunohistochemical images of paraffin cross sections of 2D, spheroid and CMF 3D tissues. Statistical comparisons were performed by one-way ANOVA followed by post hoc Tukey-Kramer multiple comparison tests. p value, * <0.05 and ** <0.01 denote statistically significant.

Fig. 6

Evaluation of “memory function” of cells in CMF 3D. (A) Schematic illustration of culture scheme to evaluate “memory function” of cells. HCT116 were precultured in each culture method (2D, spheroid, CMF 3D) and isolated using enzymes from their scaffold. All isolated cells (Ex 2D, Ex spheroid and Ex CMF) were seeded to 2D culture for 11 days and ROS resistance was evaluated at Day 0, 1, 4, 7 and 11. (B) The change in the percentage of ROS-negative cells in total cells via gating of doublet removal from Day 0 to Day 11 was analyzed in Ex 2D, Ex spheroid, and Ex CMF. (C) Quantitative-PCR analysis of GSH related genes, GSR and GCL analyzed by the ΔΔCt method from HCT116 cultured in Ex 2D, Ex spheroid and Ex CMF (n = 3) from Day 0 to Day 11. Each target was normalized using PPIA as an internal control. Target expressions were compared as a percentage against 2D culture values. Statistical comparisons were performed by one-way ANOVA followed by post hoc Tukey-Kramer multiple comparison tests. p value, * <0.05 and ** <0.01 denote statistically significant.