Cell counting kit-8 (CCK-8) and polyvinylidene fluoride (PVDF) membrane were purchased from Beyotime Biotechnology (Shanghai, China). BCA protein assay kit, protease inhibitor cocktail and RIPA lysis buffer were purchased from CWBiotech Co. Ltd. (Beijing, China). Rat-tail type I collagen were purchased from Solarbio Science & Technology Co., Ltd. (Beijing, China). Ara-C, DNR and discoidin domain receptor 1 (DDR1)-IN-1 were purchased from MedChemExpress USA (New Jersey, USA), RPMI-1640 media was purchased from Hyclone (GE Healthcare Life Sciences, Logan, UT, USA). Fetal bovine serum (FBS) was purchased from Biological Industries (Beit Haemek, Israel). CD2-PC5 (cat. no. A07745), CD3-ECD (cat. no. A07748), CD4-FITC (cat. no. A07751), CD8-PE (cat. no. A07756), CD34-ECD (cat. no. B49202), (cat. no. B36294), all purchased from Beckman coulter, Inc. (Brea, CA, USA). Stat3 (cat. no. 9139), DDR1 (no. 5583), p-STAT3 (no. 9145), GAPDH (cat. no. 5174s), primary antibodies, goat anti-mouse (cat. no. 7076s) and rabbit HRP-conjugated secondary antibody (cat. no. 14708s) were purchased from Cell Signaling Technology, Inc., (Danvers, MA, USA). PCL scaffolds were purchased from 3D Biotek LLC (New Jersey, USA).

The study was approved by the Ethics Committee of Qingdao University (Qingdao, China).

Collagen, a major extracellular matrix (ECM) component, is coated to scaffold to make cell adhere better and as a ligand of DDR1 (11). DDR1 functions as the unique receptor tyrosine kinase (RTK) which can be activated by collagen (12,13). PCL scaffolds were transferred into the 24-well plates and coated with collagen type I in different concentrations (10, 20, 40, 80, 160 and 320 µg/ml). Then the PCL scaffold was dried in biosafety hood overnight.

The acute T leukemia cell line Jurkat, Clone E6-1 (ATCC^® TIB-152™) was cultured in RPMI-1640 supplemented with 10% FBS and 1% penicillin/streptomycin. The cells were maintained at 37°C with 5% CO₂ and 95% air (13). For experiments on scaffolds, the seeding density is different from 2.5×10⁵ to 5×10⁵ cells/ml according to experiments and the seeding method was in strict accordance to the 3D Insert™ cell seeding protocol.

Cell seeding efficiency was determined 4 h after seeding into scaffold by counting the number of cells remaining in the media (14). Cell proliferation/viability was assessed for 6, 24, 48, 96, 144 and 168 h using CCK-8 reagent. Jurkat cells were seeded at a density of 2.5×10⁵ cells/ml and incubated with CCK-8 reagent for 2 h at 37°C. A total of 100 µl solution was taken for recording the optical density after incubation. Absorbance at 450 nm was measured using the Tecan Safire 2 multifunctional microplate reader.

In order to visualize the growth of the cells in the scaffold, the scaffold was examined by SEM (JSM-840; JEOL, Ltd., Tokyo, Japan) following cultivation with the Jurkat cells for 72 and 144 h. Briefly, the scaffold was treated with 3% glutaral for 4 h. Then, the scaffold was washed twice with PBS, following 30, 50, 70, 80 and 100% ethanol for dehydration. Then the carbon dioxide critical point dryer (XD-1; Eiko, Japan) was used to dry the scaffold. Prior to SEM analysis, samples were coated with gold for 120 sec using auto fine coater (IB-3; EiKo, Japan).

Flow cytometry analysis was performed to analyze whether the cell phenotype changes after cultivation in scaffold for one week. The expression of CD2, CD3, CD4, CD8, CD34 and CD45 were analyzed. Briefly, nonadherent cells were removed by PBS washing, then the cells were digested from the scaffold by collagenase treatment. After digestion, cells were harvested by gently pipetting the cell suspension up and down five times within the scaffold. Then the supernatant was centrifuged at 100 × g for 5 min to pellet the cells. The cells pellet was resuspended in PBS and incubated with antibodies for 15 min at 37°C. Flow cytometry analysis was performed using FC500 (Beckman Coulter, Inc., Brea, CA, USA). Isotypes were used as a control.

The cells were lysed with RIPA buffer containing phosphatase and protease inhibitor cocktail. Protein was extracted from cells using an ultrasonicator for two cycles, 5 sec each time with a 10-second interval. After incubation for 15 min on ice, the lysate was centrifuged at 15,493 × g for 20 min. After centrifugation, the supernatant was collected and quantified with BCA protein assay. Total protein (20 µg) was separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF membrane. After blocking with 5% non-fat milk for 1 h at room temperature, the membranes were probed with primary antibody against DDR1, STAT3, p-STAT3 and GAPDH overnight at 4°C, followed by incubation with HRP-conjugated secondary antibody. The immunoreactive bands were visualized using Fusion FX7 detection reagents (Vilber Lourmat, Marne-la-Vallée, France). Autoradiograms were scanned and the labeled bands were quantified using Image J analysis software (National Institutes of Health, Bethesda, MD, USA).

3D culture platforms are capable of mimicking microenvironments in vivo, which provide greater physiological relevance in comparison to conventional 2D cultures (15–18). Ara-C (1 µM) and DNR (500 nM) were used in our experiments according to the clinical application 5×10⁵/ml Jurkat cells were seeded into the tissue culture plate systems (TCPS), uncoated PCL scaffold (UPS) and collagen type I coated polycaprolactone scaffold (CTCPS) for 24 h. Nonadherent cells were removed along with the medium and new medium with or without drugs was added. For TCPS, cells were taken out with medium and centrifuged. After removing the medium, the cells were seeded into wells and medium with or without drugs was added. The medium was replaced every 48 h. Das et al found that inhibition of DDR1 expression significantly enhanced chemosensitivity of breast cancer cells to genotoxic drugs (19). Thus, DDR1 signaling provides a new target for therapeutic intervention to leukemia cells. DDR1-IN-1, an orally bioavailable DDR1 inhibitor (20,21), was used in our experiments.

All experimental data were analyzed by Student's t-test. All experiments were done three times. Data are presented as mean ± standard error. P<0.05 was considered to indicate a statistically significant difference. All tests were completed using PRISM 6 (GraphPad Software, Inc., La Jolla, CA, USA).

Collagen type I is one of the main bone marrow ECM proteins and also the ligand of DDR1 (22). In order to mimic the bone marrow microenvironment and further study the function of DDR1, PCL scaffold coating with collagen type I was used in our experiments (Fig. 1). The effect of different concentrations of collagen type I on the cell seeding and cell proliferation was improved compared to UPS. Indeed, collagen type I coating enhanced the seeding efficiency. The seeding efficiency increased gradually at low concentration of collagen type I, but decreased at high concentration (160 or 320 µg/ml) when compared with 80 µg/ml collagen type I coated PCL scaffolds (Fig. 2).

To investigate the effect of PCL scaffold on cell proliferation, we compared proliferation of Jurkat cells on TCPS, UPS and 80 µg/ml CTCPS for 168 h. CCK-8 reagent was used to analyze the extent of cell proliferation/viability. Different growth kinetics were observed for Jurkat cells cultured on different conditions. Coating the PCL scaffold with collagen type I enhanced cell proliferation compared with the TCPS and UPS, especially for a longer time culture. For the initial 96 h, the TCPS revealed the maximum cell proliferation/viability compared with the UPS and CTCPS. However, with the prolongation of the cell culturing, the CTCPS showed better cell proliferation than the UPS. Moreover, CTCPS can maintain cell proliferation for 168 h. After 166 h, the TCPS revealed a significant deterioration in cell proliferation (Fig. 3).

The cell proliferation/viability result showed that 80 µg/ml CTCPS was able to support cell survival and proliferation better. It is very important for cells to maintain the initial phenotype and not differentiate with the prolongation of the cell culture. The expression of CD2, CD3, CD4, CD8, CD34, CD45 in Jurkat cells cultured in different conditions such as UPS and 80 µg/ml CTCPS were analyzed at 168 h. The results showed that the percentage of cells with the phenotype of CD2⁺CD3⁺CD4^+dimCD8⁻CD34⁻CD45^+dim was maintained for 168 h of culture on the scaffolds compared to that cultured on TCPS (Fig. 4). Similarly, the phenotype of Jurkat cell on the CTCPS and UPS showed no difference in comparison to the TCPS for 168 h.

The 80 µg/ml CTCPS was more suitable for cell proliferation and growth for a longer period without changing cell phenotype. So we used 80 µg/ml CTCPS for drug testing applications. Jurkat cells were treated with Ara-C and DNR or their combination (Fig. 5). CCK-8 reagent was used to test cell cytotoxicity of Jurkat cells cultured on TCPS, UPS and 80 µg/ml CTCPS. Jurkat cells revealed high sensitivity to Ara-C cultured on TCPS. Cell proliferation reduced to 89.58±3.26, 62.86±6.84 and 53.36±3.37% for 48, 144 and 168 h respectively on TCPS, according to the CCK-8 results (Fig. 5A). Cells cultured on 80 µg/ml CTCPS showed resistance to drug treatment, as shown by the result that cell proliferation was 103.98±5.3, 88.69±2.19, 77.11±3.47% at 48, 144 and 168 h. Moreover, the drug sensitivity of cells cultured on UPS was between TCPS and CTCPS, with the relative percentage of 87.98±3.74, 73.28±5.5 and 63.65±0.81% at 48, 144 and 168 h. Cells cultured under the three conditions showed more sensitivity to DNR or Ara-C + DNR combination than Ara-C alone. A significant death of Jurkat cells was observed in TCPS (57.67±8.02, 42±4.58%) when compared to CTCPS (78±5.57, 57.33±5.03%) in the presence of DNR for 96 and 168 h (Fig. 5B; *P<0.05 vs. TCPS). A similar trend of combination use of Ara-C and DNR was obtained in TCPS (24.67±4.73%) and UPS cultures (39±7.55%) in comparison to 80 µg/ml CTCPS (46.33±4.04%) for 168 h (Fig. 5C; **P<0.01 vs. TCPS). In general, the cells cultured on 80 µg/ml CTCPS showed drug resistance for a longer culture period compared with TCPS and UPS.

We have shown that cell cultured on 80 µg/ml CTCPS revealed chemoresistance compared to TCPS. We further explored the possible mechanisms that lead to cell chemoresistance. Evidence suggests that high expression of DDR1 and p-STAT3 is related to cell proliferation, invasion, migration and chemoresistance of malignant tumors (23,24). STAT3 is the downstream effector of DDR1 pathway and STAT3 can be phosphorylated by DDR1 (25,26). Therefore, we wanted to verify whether the drug resistance to Ara-C and DNR of cells cultured in 3D scaffold is related to the high expression of DDR1 and p-STAT3. The levels of DDR1 and p-STAT3 were highly expressed on the UPS and 80 µg/ml CTCPS compared to the TCPS (Fig. 6). Next, we chose DDR1-IN-1, a potent and selective DDR1 receptor tyrosine kinase inhibitor (21), to further analyze whether DDR1 inhibition would reverse the chemoresistance. The concentration of DDR1-IN-1 used was determined in Jurkat cells cultured in TCPS (Fig. 7A). IC₅₀ of 335 nM with DNR was chosen for further experiments. DDR1-IN-1 (1 µM) which is sufficient to inhibit DDR1 autophosphorylation alone did not inhibit cell proliferation after incubation for 48 h, whereas Ara-C (Fig. 7B) and DNR (Fig. 7C) singly or in combination (Fig. 7D), along with DDR1-IN-1 induced significant cell death, wherein almost all the cells were wiped out in contrast with DDR1-IN-1 alone at 144 h.