Preservation of osteoblasts and BM-MSCs biological properties after consecutive passages with the thermal-liftoff method

Abstract

Thermo-responsive surfaces based on PNIPAAm provide a novel way to harvest cells with temperature reduction. This cell harvest method makes cell membranes remain intact compared with enzymatic treatment. Nevertheless, it is unclear whether this alternative method tends to preserve cells' biological properties after consecutive passages in the long-term expansion. We constructed poly(N-isopropylacrylamide) (PNIPAAm)-grafted hollow fiber membranes (HFMs) with thermo-responsive properties. The osteoblasts and BM-MSCs of the second passage (P2) were cultured on both the thermo-responsive HFMs and glass coverslips, and then passaged with reducing temperature (thermal-liftoff) and enzymatic treatment. After such consecutive passages, the biological properties of the cells harvested by thermal-liftoff and trypsinization were analyzed and compared. The results showed that a higher viability and cellular proliferation were obtained when cells were detached using the thermal-liftoff method; alkaline phosphatase (ALP) activity and osteocalcin (OCN) protein expression of osteoblasts at P9 and BM-MSCs at P7 with thermal-liftoff were significantly higher than those with trypsinization; and more calcium nodules stained by Alizarin Red S were formed in thermal-liftoff groups. Additionally, for BM-MSCs, more glycosaminoglycan (GAG) was synthesized and secreted into the ECM after chondrogenic differentiation in the thermal-liftoff groups. Therefore, harvesting cells with the thermal-liftoff method is favorable to keep the functions of osteoblasts, and the osteogenic and chondrogenic differentiation of BM-MSCs. Besides, many more extracellular matrix (ECM) proteins remained around or within the cell membranes in the thermal-liftoff groups, which facilitates the preservation of osteoblasts and BM-MSCs' cellular activity, osteoblastic differentiation and mineralization. These results suggest that thermo-responsive HFMs and the matched thermal-liftoff method could be a candidate for harvesting more primitive therapeutic cells (osteoblasts and BM-MSCs) after consecutive passages.

---

1. Introduction

A large number of competent therapeutic cells (in the order of 10⁹ to 10¹¹) are required in regenerative medicine, including tissue engineering and cell transplantation.^1,2 Because of there being an insufficient number of cells derived from humans, animal tissue or body fluids,^3,4 it is important for regenerative medicine to maintain cells' biological characteristics during expansion and harvest. In the process of long-term expansion in vitro, the anchorage-dependent cells undergo serial subculture operations. Therefore, besides a suitable culture substrate and a good culture environment, an appropriate harvest method is also an important condition for cell large-scale expansion. Proteolytic enzyme treatments (e.g. trypsinization) are the most widely employed for harvesting anchorage-dependent cells. Generally, cells are harvested with 0.25% trypsin + 0.02% EDTA at 37 °C for a few minutes followed by trypsin deactivation and cell washing, which is complicated and labor-intensive. Moreover, this treatment results in the degradation of some epitopes of the cell membrane, even though many cells tolerate trypsin digestion for a short time.^5,6 Moreover, trypsin digestion selectively hydrolyzes the protein peptide chain that is the connection between cells and the substrate, breaks the connection between integrin protein and cytoskeleton proteins, and inevitably degrades cell membrane proteins and deposited extracellular matrix (ECM), consequently results in regulating cell adhesion and migration.^5–8 The loss of cells function may be compensated by cells self-repairing. However, the biological characteristic of harvested cells might reduce greatly after frequent passage operations in the long culture term in vitro, thus compromising their therapeutic potency.

Additionally, cell viability decreased significantly when trypsinization time extended from 5 min to 20 min and 60 min.^7,9 And it should be avoided to use of animal-derived enzymes for preventing possible contamination with adventitious agents.¹⁰ Therefore, a non-invasive or less-invasive cell recovery method without proteolytic enzymes would be a candidate for therapeutic cells in the long-term culture in vitro.

Currently, the novel cell harvest strategies have been focused on stimuli-responsive polymers and aptamers mediated detachment.⁹ Stimuli-responsive material undergoes a change in polymer chain conformation when exposed to external stimuli such as changes in temperature, pH, solution ionic strength, solvent composition, light, mechanical force, electric and others.^11,12 The stimuli-responsive substrates have a reversible change in their microstructures from a collapsed to an expanded configuration with external stimuli. Particularly, the change occurs locally at a fast rate and has high selectivity. Poly(N-isopropylacrylamide) (PNIPAAm) is the most popular temperature-responsive polymer for cell culture and detachment strategies. PNIPAAm and its derivatives often exhibit a lower critical solution temperature (LCST) in aqueous solution around 31 °C. Across LCST, these polymers show remarkable hydration-dehydration changes in response to small temperature changes over a narrow range; the thermos-responsive surface based on PNIPAAm presents a completely different geometry and surface wettability. At 37 °C, the thermos-responsive surface exhibits weakly hydrophobic, which is suitable for cells adhesion and growth. When temperature is below LCST, the surface becomes hydrophilic and volume expands, the combined effects force cells to detach from the substrate surface.

In 1990, Okano group¹³ firstly prepared a thermo-responsive surface using PNIPAAm for cells culture and harvest with temperature reduction without proteolytic enzyme treatment. This thermally induced cell harvesting not only simplifies the cell harvest procedure and subsequent manipulations, but also enhances cell viability and therapeutic efficiency. Ma et al.¹⁴ modified microfluidic channel with PNIPAAm and further investigated cell detachment with temperature reduction. The results showed that 89% (COS7) and 97.2% (hMSCs) of viable cells were harvested, which was significantly higher than trypsin digestion with cell viability of 80.6% (COS7) and 75.7% (hMSCs). Liu et al.¹⁵ coated PNIPAAm on silicon nanopillars and further introduced antibody onto PNIPAAm by the anchor biotin-BSA for MCF7 cell isolation. It was reported that 98.8% of captured MCF cells were released from thermo-responsive nanostructured surfaces with minimal decrease in cell viability (<1%). Yang et al.¹⁶ examined cells viability, proliferation, differentiation, immunophenotype and protein residues of released MSCs with the method of trypsinization and thermal-liftoff. The result indicated that MSCs harvested from thermo-responsive films with temperature reduction had higher cell viability, more colonies and stronger differentiation, in comparison with the trypsinization group after three consecutive passages. However, there had no significant difference in cell morphology, immunophenotype. Obviously, thermally induced cell harvesting has an advantage in maintaining cell ability compared with enzyme digestion, whereas it is uncertain that whether cells retain more biological properties when they are harvested from thermo-responsive substrates with temperature reduction after long-time culture and passage processes in vitro.

We prepared thermo-responsive hollow fiber membranes (HFMs) that achieved cells culture and non-invasive recovery with temperature reduction.¹⁷ In this research, the rat calvaria osteoblasts and rat bone marrow-mesenchymal stem cells were seeded on thermo-responsive HFMs and glass coverslips, and recovered by temperature reduction and trypsinization respectively. After 3, 5, 7 consecutive passages, the cell ability, related proteins, proliferation, osteoblastic differentiation and mineralization of harvested cells were investigated.

2. Materials and methods

2.1 Materials

Cellulose acetate HFMs with an outer diameter and wall thickness of 175 μm and 13 μm, respectively, were obtained from KAWASUMI. N-Isopropylacrylamide (NIPAAm), supplied by J&K Chemical Reagent Co., Ltd, was used as a graft monomer and recrystallized from hexane. CAN was purchased from Sinopharm Chemical Reagent Co., Ltd. as an initiator. Deionized water were used in all procedures.

2.2 Preparation and characterization of thermo-responsive HFMs

Thermo-responsive HFMs were prepared according to the reported protocol.¹⁷ Briefly, after alkaline treatment in 0.05 M NaOH solution at 25 °C for 24 h, the HFMs were washed with deionized water to remove the alkali solution. Then HFMs (total length: 2000 cm) were loaded into a round bottom flask equipped with a three-way stopcock, containing 25 mL nitric acid solution ([HNO₃] = 0.1 M), followed by bubbling with nitrogen for 5 min to deoxygenate the reaction mixture. CAN (0.137 g, 0.25 mmol) was introduced into the reactor to carry out the initiation step. The CAN initiator concentration, initiation time and temperature were 10 mM, 30 min and 60 °C, respectively. After the initiation step, the recrystallized NIPAAm (0.0282 g, 0.25 mmol) were added into the reactor and the graft copolymerization of NIPAAm onto the HFMs was carried out under nitrogen at 25 °C for 6 h. After grafted copolymerization, the sample was rinsed and washed with deionized water to remove the initiator, ungrafted PNIPAAm homopolymer and unreacted monomers. The washed HFMs were dried in oven at 25 °C for at least 24 h. The fourier transform infrared spectroscopy showed the chemical structure of the PNIPAAm-grafted HFMs and elemental analyzer results confirmed the PNIPAAm grafted amount with 2.65 ± 1.73 μg cm⁻² of the thermo-responsive HFMs. The results of dynamic contact angles and proteins adsorption showed that thermo-responsive HFMs displayed temperature sensitivity with temperature changed.

2.3 Isolation and culture of osteoblasts and BM-MSCs

Osteoblasts were isolated from the new-born Sprague Dawley (SD) rat of either sex (Experimental Animal Center of Dalian Medical University, China) according to the reported protocol.¹⁸ Briefly, calvaria from five newborn 24–48 hours rats were excised aseptically, cleansed with PBS and cut into 1 mm × 1 mm × 1 mm size fragments. Then, the bone fragments were treated with 0.25% trypsin solution at 37 °C for 15 min and the supernatant was removed. After cleaning with PBS, the bone fragments were digested with 0.1% collagenase type II (Gibco, USA) at 37 °C water bath for 90 min. The supernatant was collected and centrifuged at 1500 rpm for 5 min. The cell pellet was resuspended in Dulbecco's Modified Eagle's Medium (DMEM, Gibco, USA) containing 10% fetal bovine serum (FBS, Siencell, USA), 100 units per mL penicillin and 100 μg mL⁻¹ streptomycin (Hyclone, USA) and cultured in a humidified 5% CO₂ atmosphere at 37 °C. Medium were refreshed every 3 days. These primary cells were referred as passage 0 (P0). Cells were passaged until reaching approximately 80–90% confluence. The osteoblasts at P2 were used for further studies about the biological properties of cells released from thermo-responsive HFMs by thermal-liftoff and cells detached from the glass coverslips by trypsinization.

Rat bone marrow mesenchymal stem cells (BM-MSCs) were isolated from whole bone marrow aspirates of SD rats (90–120 g, provided by Experimental Animal Center of Dalian Medical University in China) of either sex according to the reported protocol¹⁶ with some modifications. Briefly, the femurs of SD rats were dissected aseptically and cleaned extensively to remove associated soft connective tissues, followed by immersing with 75% (v/v) ethanol for 10 min and rinsing with PBS thoroughly. The distal ends of the bone were then cut open and marrow cavities were flushed with low glucose-Dulbecco's modified Eagle's medium (LG-DMEM, Gibco BRL, USA) supplemented with 10% foetal bovine serum (FBS, Scincell) and 1% penicillin-streptomycin (P/S, Solarbio, China). The obtained cell suspension were centrifuged at 1000 rpm for 5 min to remove the supernatant. The extracted cells were seeded in T25-flasks and cultured in a humidified atmosphere of 5% CO₂ at 37 °C with the culture medium refreshed every 3 days. These primary cells were referred as passage 0 (P0). The adherent cells were passaged when they reached 80–90% confluence. The BM-MSCs of P2 were used for further comparison between the behavior of BM-MSCs released from thermo-responsive HFMs by thermal-liftoff and that of cells detached from the glass coverslips by trypsinization. In addition, the BM-MSCs of P2 were used to detect surface antigens with the flow cytometry (BD Biosciences, USA). In brief, cells were filtrated with cell strainers (70 μm) and centrifuged (1000 rpm, 5 min). Then, cells were resuspended in PBS with a density of 4 × 10⁵ cells per mL. 5 μL of FITC-labeled CD29 (BD Biosciences, USA), 5 μL of FITC-labeled CD34 (BD Biosciences, USA) and 10 μL of FITC-labeled CD90 (BD Biosciences, USA) were respectively added to 500 μL cell suspension and incubated for 30 min on ice-bath without light. After that, the cells were washed twice and resuspended in PBS. The results were analyzed with BDFACSDiva software.

The above animal tests were performed in accordance with the guidelines of ‘Instructive notions with respect to caring for laboratory animals’ for animal experimentation by the Ministry of Science and Technology of the People's Republic of China in 2006. Bioethics and Medicine Committee of Dalian University of Technology have approved the experiments.

2.4 Osteoblasts and BM-MSCs cultivation and harvest on thermo-responsive HFMs

Thermo-responsive HFMs were sterilized with high-pressure steam and rinsed with PBS for 3 times, and soaked in DMEM medium containing 10% FBS overnight in 6-well tissue culture plate for subsequent use. 50 μL osteoblasts or BM-MSCs suspension of P2 was gently dripped on the HFMs surface at a density of 2 × 10⁴ cells per cm². After incubation for 4 h at 37 °C in order to allow the cells to attach on the HFMs surface, cell-HFMs were transferred to another 6-well tissue culture plate and final volume of the medium was adjusted to 2 mL per well. The cell-HFMs constructs were incubated at 37 °C with medium changed every 2 or 3 days. When reaching 80–90% confluence, the osteoblasts or BM-MSCs were harvested by temperature reduction. Briefly, the cold fresh serum-free culture medium (less than 20 °C) were introduced after removing medium. Then cell-HFMs constructs were incubated at 20 °C for 30 min. When cell morphology changed from spread to round, the cells were released from the surface of thermo-responsive HFMs after being gently pipetted. Then, the released cells were reseeded on other thermo-responsive HFMs for subculture. These handlings were repeated 7 times for osteoblasts and 5 times for BM-MSCs and these cultures were referred to as the thermal-liftoff groups.

Some studies have shown that there is no obvious difference in cells adhesion, differentiation and morphology cultured on filter membranes, glass and plastic;¹⁹ and there is also no obvious differences in cellular morphology, viability and ECM production for cells on thermo-responsive substrates, glass coverslips and non-glass substrates.^16,20,21 We believe the diversities in the performance of released cells in the thermal-liftoff groups and trypsinization groups are independent on the diverse culture substrates. Thus, glass coverslips were employed as positive controls in this study. The osteoblasts or BM-MSCs at P2 were seeded on the sterilized glass coverslips in 24-well culture plates at a density of 2 × 10⁴ cells per cm². Cells were cultured in an incubator at 37 °C with medium changed every 2 or 3 days. When reaching 80–90% confluence, cells were passaged with proteolytic enzyme treatment and reseeded on other glass coverslips. In brief, the culture medium were removed and cells were rinsed twice with PBS. Then, 50 μL of 0.25% trypsin + 0.02% EDTA solution (Gibco-Invitrogen) was added and incubated at 37 °C for 5 min. These handlings were repeated 7 times for osteoblasts and 5 times for BM-MSCs, and these cultures were referred as trypsinization groups.

2.5 Protein analysis of the released osteoblasts and BM-MSCs

The extracellular matrix components (fibronecti (FN) and laminin (LN)) of cells released from thermo-responsive HFMs with thermal-liftoff and glass coverslips with trypsinization were evaluated through immunofluorescent staining. Briefly, the released cells in thermal-liftoff groups and trypsinization groups were dropped on glass slides and incubated at 4 °C refrigerator overnight. Then, the cell smears were fixed with 4 °C 4% paraformaldehyde for 20 min and washed 3 × 5 min with PBS. 10% goat serum was added and incubated at room temperature for 60 min to block non-specific interaction. After that, primary antibody solutions (rabbit anti-fibronectin, anti-laminin (Abcam)) were added for incubation at 4 °C overnight. Then, the primary antibody solutions were removed and the cell smears were washed with PBS for 5 min, FITC-conjugated goat anti-rabbit IgG (Santa Cruz Biotechnology, Inc.) and Hoechst 33342 were added for incubation at room temperature for 60 min. The anti-fluorescein quencher was dropped after the cell smears were washed with PBS and air-dried. The cells were observed under an inverted fluorescence microscope (OLYMPUS BX71, Japan).

In addition, the total protein, fibronectin (FN) and laminin (LN) carried with the released cells were evaluated through quantitative analysis with BCA Protein Assay Kit (Beyotime, China), ELISA Kits of Fibronectin and Laminin (Boster, China). The released cells in thermal-liftoff groups and trypsinization groups were washed twice with cold PBS, centrifuged for 5 min at 1000 rpm and then incubated with 500 μL radioimmunoprecipitation (RIPA) + 5 μL phenylmethylsulfonyl fluoride (PMSF) for 30 min on ice. The cell lysates were centrifuged at 13 500 rpm at 4 °C for 5 min to remove cell debris. Then, the supernatants were collected and stored frozen for BCA assay and ELISA assay. The quantitative analysis of total protein, FN and LN were performed according to the manufacturer's protocol. The assayed protein mass was normalized to the cell number.

2.6 Cell viability and proliferation of the released osteoblasts and BM-MSCs

For evaluating the viability of the released osteoblasts (P5, P7, P9) and BM-MSCs (P5, P7), the cells were stained with Calcein-Am/PI. The viable and dead cells were imaged using a fluorescent microscope (OLYMPUS BX71, Japan) with appropriate filter sets.

The Alamar Blue (AB) assay (YEASEN, USA) was used to determine cells proliferation of the released osteoblasts (P5, P7, P9) and BM-MSCs (P5, P7) following the manufacturers' instructions. AB is a non-toxic aqueous fluorescent dye that does not affect phenotype, viability or proliferation of the cells. Briefly, the released cells were collected and seeded on 96-well plate at a density of 1 × 10³ cells per well. At 1, 3, 5, 7 day, the cells were incubated in serum-free medium supplemented with 10% (v/v) AB fluorescent dye for 4 h at 37 °C. Then the assay solution in each well was transferred to a black 96-well plate in triple and fluorescence was measured by an enzyme-linked immune detector (Thermo scientific, USA) at excitation and emission wavelengths of 550 nm and 590 nm, respectively.

2.7 Osteoblastic differentiation and mineralization

The effect of thermal-liftoff and enzymatic treatment on osteoblastic differentiation and mineralization of osteoblasts and BM-MSCs was evaluated through ALP assay, osteocalcin (OCN) protein assay and Alizarin Red S staining. In brief, the released osteoblasts (P5, P7, P9) and the released BM-MSCs (P5, P7) were seeded on 96-wells/24-wells plate at the density 2 × 10⁴ cells per cm² and cultured for 3 days with DMEM containing 15% FBS and 1% P/S. Then, the culture medium was changed to be osteogenic medium (DMEM, 15% FBS, 1% P/S, 10 mM β-glycerophosphate, 50 μM ascorbic acid and 100 nM dexamethasone). After 10 days, ALP activity of osteoblasts was assayed using a BCIP/NBT alkaline phosphatase color development kit (Beyotime Institute of Biotechnology) and alkaline phosphatase assay kit (Nanjing Jiancheng Bioengineering Institute). After 4 weeks, OCN protein of osteoblasts was quantitatively assayed with rat bone gla protein/osteocalcin (BGP/OCN) ELISA kit (Bio-Swamp Life Science) and Alizarin Red S staining was conducted. In BM-MSCs osteogenic differentiation experiment, ALP activity was assayed after induction with osteogenic medium for 2 weeks, cells were stained with Alizarin Red S and OCN protein was quantitatively assayed at 4 weeks in accordance with the manufacturer's protocol. The results were normalized to the cell number.

2.8 Chondrogenic differentiation of BM-MSCs

The chondrogenic differentiation potential of BM-MSCs harvested with thermal-liftoff and trypsinization methods were evaluated with toluidine blue staining. In brief, 50 μL of the harvested BM-MSCs with a density of 1 × 10⁶ cells per mL were seeded on 24-wells plate. After incubation for 4 h to allow the cells to completely adhesion, chondrogenic medium (HG-DMEM, 1% FBS, 50 μg mL⁻¹ ascorbate-2-phosphate, 100 nmol L⁻¹ dexamethasone, 10 ng mL⁻¹ TGF-β1, 1% ITS⁺ premix) were added. After incubation for 2 weeks, the cells were fixed with 4% paraformaldehyde and stained with toluidine blue.

2.9 Statistical analysis

Statistical analysis was performed using a two-tailed Student's t-test. A p-value of <0.05 was considered to indicate statistical significance. All the values were expressed as the means ± SD of triplicate experiments.

3. Results

3.1 Expression of surface marker protein for BM-MSCs

Previous studies had shown that BM-MSCs usually exhibited around 95% positive staining for CD29, CD44, CD73, CD90 and CD105 and expressed low levels of CD34 and CD45 (≤2%).^22,23 Therefore, we marked the cells with CD29, CD90 and CD34. The results were shown in Fig. 1. 94.0% and 92.6% of the cell population positively expressed CD29 and CD90, respectively. However, they expressed the hematopoietic marker CD34 with only 2.1%. 0.5% of the cell population negatively expressed CD45. The results indicated that these cells were BM-MSCs.

Fig. 1 Expression of the surface marker proteins for BM-MSCs. The expressing rate of the no-treatment control group, CD34, CD45, CD29 and CD90 were 0.6%, 2.1%, 0.5%, 94.0% and 92.6% respectively.

3.2 Cell detachment

The osteoblasts and BM-MSCs' detachment from thermo-responsive HFMs with temperature reduction were shown in Fig. 2. At 37 °C, cells spread on the surface of thermo-responsive HFMs. When exposed to cold non-FBS DMEM (approximately 20 °C), the surface of thermo-responsive HFMs changed to be more hydrophilic.¹⁷ After incubation with cold non-FBS DMEM for 30 min, cells turned to be round. And, cell clusters were completely released after being gently pipetted.

Fig. 2 Cells morphology on thermo-responsive HFMs after reducing temperature from 37 °C to 20 °C. The spreaded cells changed to be rounded after temperature reduction and cell clusters completely detached after being gently pipetted. Scale bar: 100 μm.

3.3 Protein analysis

The distribution of fibronectin and laminin carried by the cells on thermo-responsive HFMs and on glass coverslips, as well as by the released cells in thermal-liftoff groups and trypsinization groups were shown in Fig. 3(A–D) and 4(A–D). The fibronectin and laminin around adherent cells on thermo-responsive HFMs and glass coverslips showed strong green fluorescence. The fibronectin localized between cells of the inner cell mass and assembled into a fibrillar matrix; the laminin of attached cells distributed mainly surrounding the cell membrane. As shown in Fig. 3(E, I, M, G, K and O) and 4(E, I, G and K), the fibronectin and laminin of cell clusters in thermal-liftoff groups showed strong green fluorescence, which indicated that cells in thermal-liftoff groups detached with a large amount of fibronectin and laminin. For the dispersed cells in trypsinization groups, some cells expressed little or weak green fluorescence, which revealed that only a small amount of fibronectin and laminin detached together with the cells after enzymatic treatment.

Fig. 3 Representative immunofluorescence staining of membrane proteins of fibronectin (A, B, E, F, I, J, M, N) and laminin (C, D, G, H, K, L, O, P) of adherent osteoblasts on thermo-responsive HFMs (A, C) and glass coverslips (B, D), cell cluster harvested by thermo-liftoff (E, I, M, G, K, O) and dispersed cells harvested by trypsinization (F, J, N, H, L, P). The cell passage number is 2 (A–D), 5 (E–H), 7 (I–L), 9 (M–P). Proteins were stained green with FITC-labeled secondary antibody, and cell nuclei stained blue with Hoechst 33342 dye. The pictures of different colors in the same field were overlaid with Image-Pro plus 6.0 software. Scale bar: 200 μm.

Fig. 4 Representative immunofluorescence staining of membrane proteins of fibronectin (A, B, E, F, I, J) and laminin (C, D, G, H, K, L) of adherent BM-MSCs on thermo-responsive hollow fiber membranes (A, C) and glass coverslips (B, D), disperse d cells harvested by trypsinization (F, J, H, L) and cell cluster harvested by thermal-liftoff (E, I, G, K). The cell passage number is 2 (A–D), 5 (E–H), 7 (I–L). Proteins stained green with FITC-labeled secondary antibody, and cell nuclei stained blue with Hoechst 33342 dye. The pictures of different colors in the same field were overlaid with Image-Pro plus 6.0 software. Scale bar: 200 μm.

Furthermore, the fibronectin, laminin and total proteins carried by the released cells were quantitatively analyzed, as summarized in Tables 1 and 2. The total proteins, fibronectin and laminin of thermal-liftoff groups were significantly higher than those of trypsinization groups. The result demonstrates that the thermal-liftoff method makes released cells retain much more extracellular matrix proteins compared with those of trypsin digestion. Additionally, the proteins carried by harvested cells in thermal-liftoff groups remained at a high level after 7 and 5 passage cycles. And the annexed proteins released cells of trypsinization groups witnessed a downward trend, it is predicted that enzymatic treatment might damage the function of cells synthesizing and secreting extracellular matrix proteins to a certain extent.

Table 1 Average protein mass of total proteins, fiberonectin and laminin carried by the harvested osteoblasts normalized to cell number^a

Protein	Thermal-liftoff group			Trypsinization group
	P5	P7	P9	P5	P7	P9
a *P < 0.05, **P < 0.01. Compared with trypsinization group.
Total protein (ng per cell)	0.0761 ± 0.0007**	0.0550 ± 0.0026**	0.0662 ± 0.0021*	0.0492 ± 0.0008	0.0270 ± 0.0023	0.0445 ± 0.0026
Fibronectin (pg per cell)	2.6779 ± 0.2025**	2.5904 ± 0.1313**	2.4412 ± 0.2404**	0.6057 ± 0.0966	0.2717 ± 0.0440	0.5583 ± 0.0656
Laminin (pg per cell)	0.0797 ± 0.0032**	0.0552 ± 0.0383	0.0922 ± 0.0252**	0.0434 ± 0.0064	0.0340 ± 0.0062	0.0118 ± 0.0047

---

Table 2 Average protein mass of total proteins, fibronectin and laminin carried by the harvested BM-MSCs normalized to cell number^a

Protein	Control group	Thermal-liftoff group		Trypsinization group
	P2	P5	P7	P5	P7
a *P < 0.05, **P < 0.01. Compared with trypsinization group.
Total protein (ng per cell)	0.0293 ± 0.0002	0.0668 ± 0.0005**	0.0987 ± 0.0078**	0.0108 ± 0.0004	0.0142 ± 0.0004
Fibronectin (pg per cell)	0.8816 ± 0.5133	0.7592 ± 0.0284	1.0394 ± 0.0282**	0.5802 ± 0.1395	0.4049 ± 0.0684
Laminin (pg per cell)	0.0132 ± 0.0066	0.3616 ± 0.2285**	0.3350 ± 0.0767**	0.0072 ± 0.0017	0.0054 ± 0.0015

---

3.4 Cell viability and proliferation

To investigate the effect of harvest methods on cell viability, the released cells in thermal-liftoff groups and trypsinization groups were stained with Calcein-AM/PI. As shown in Fig. 5(A–C, A′ and B′), the detached cell clusters in thermal-liftoff groups showed a high cell viability (strong green fluorescence), dead cells (red fluorescence) were rarely observed. In contrast, dead cells were clearly visible (Fig. 5(D–F, C′ and D′)) in trypsinization groups. Furthermore, cell viability was assessed quantitatively with IPP software (Table 3). The osteoblasts and BM-MSCs viabilities of thermal-liftoff groups were more than 90% over consecutive passage cycles, whereas enzymatic treatment caused cell death approximately 15–33%.

Fig. 5 (1) Cell live/dead staining of osteoblasts harvested by thermal-liftoff method (A–C) and trypsinization method (D–F). A, D: P5; B, E: P7; C, F: P9. (2) Cell live/dead staining of BM-MSCs harvested by thermal-liftoff method (A′, B′) and trypsinization method (C′, D′). A′, C′: P5; B′, D′: P7. Cells were stained with Calcein-Am/PI. The scale bars: A, B, C, D, E, A′, C′, D′: 200 μm; F and B′: 100 μm.

Table 3 Cell viability of thermal-liftoff group and trypsinization group

Cell viability	Thermal-liftoff group			Trypsinization group
	P5	P7	P9	P5	P7	P9
Osteoblast	91.22 ± 7.04%	92.59 ± 1.92%	90%	74.05 ± 6.04%	67.59 ± 17.13%	71.29 ± 5.68%
BM-MSCs	96.42 ± 1.14%	95.59 ± 2.37%	—	84.96 ± 9.32%	77.67 ± 8.44%	—

---

The proliferation of detached BM-MSCs was also assayed with Alamar Blue (Fig. 6). In the earlier culture time, there is no statistical difference in cells proliferation between thermal-liftoff groups and trypsinization groups. After 5 days and 7 days culture, the BM-MSCs in thermal-liftoff groups demonstrated statistically higher cellular proliferation than those in trypsinization groups (P < 0.05 and P < 0.01). In the earlier cultivation period, the advantage of the thermal-liftoff method in cellular proliferation was not fully displayed, which may be attributed to the cell population status when they detached from substrates. The released cells in thermal-liftoff groups appeared in clusters and they took at least 20 h to spread completely when they were seeded on tissue culture plates. And we found that the larger the cell cluster, the longer the time needed for cells to spread. On the other hand, the harvested cells with enzymatic treatment were dispersed single cells and they were completely spread after 16 h (ESI Fig. 1†). It was reported that the cells with a highly spreading shape had an increase in DNA synthesis and cellular proliferation.²⁴ Therefore, thermal-liftoff groups were similar to trypsinization groups in cellular proliferations during the earlier culture time. Finally, because a large amount of ECM proteins (FN, LN etc.), which can promote cellular proliferation,^25,26 were simultaneously detached together with cells from thermo-responsive HFMs with temperature drop, the cellular proliferation of thermal-liftoff group was statistically higher than that of trypsinization group during the latter culture time.

Fig. 6 The proliferation of BM-MSCs harvested by thermal-liftoff from thermo-responsive HFMs and trypsinization from glass coverslips. The data are expressed as the mean ± SD of triplicate experiments (*, P < 0.05; **, P < 0.01).

3.5 Osteoblastic differentiation and mineralization in vitro

The effect of harvesting methods on osteoblastic differentiation of osteoblasts and BM-MSCs was further investigated. Bone formation stages involve cell proliferation, followed by extracellular matrix (ECM) maturation and mineralization. During the differentiation stage of bone cells, a high-level expression of ALP firstly occurred, and followed by ECM maturation. The matured ECM will mineralize at the end and form mineralized nodules, which are osteoblastic phenotypic markers and represent the final stage of osteoblastic differentiation. Therefore, ALP activity, OCN protein content and alizarin red S staining of the harvested cells were evaluated.

The released cells in thermal-liftoff groups and trypsinization groups were reseeded on 24-wells plate. After cultivation for 10 days in osteogenic medium, ALP activity was assessed. As shown in Fig. 7, ALP positive cells were stained and visualized with dark blue-violet color. The ALP activity of osteoblasts (P9) harvested with thermal-liftoff method was significantly higher than those of trypsinization groups (P < 0.05). For BM-MSCs, the ALP activity of thermal-liftoff group was also considerably higher than that of trypsinization group (P < 0.05 and P < 0.01). Moreover, the ALP activity of BM-MSCs at P5 and P7 harvested with the trypsinization method displayed a significant decrease compared with BM-MSCs at P2 (P < 0.05 and P < 0.01). The BM-MSCs of thermal-liftoff group at P5 didn't show a significant difference compared with BM-MSCs at P2 in ALP activity, whereas the difference in ALP activity between the P7 of thermal-liftoff group and P2 was noticeable (P < 0.05). After incubation for 4 weeks in osteogenic medium, the cells of the two groups were stained with alizarin red S solution. After 3, 5 and 7 continuous passages, the osteoblasts harvested with thermal-liftoff formed much more calcium nodules than cells harvested with trypsinization (Fig. 8). The same situation also occurred in BM-MSCs. Meanwhile, OCN protein content in the cell culture supernatant was evaluated. As shown in Fig. 9. The difference in osteoblasts' OCN protein expression level was not statistically significant between thermal-liftoff groups and trypsinization groups for both P5 and P7, whereas the significant difference was seen at P9 (P < 0.01). However, for the BM-MSCs at P7, the OCN protein expression level of thermal-liftoff group was significantly higher than that of trypsinization group (P < 0.01). Besides, the OCN protein expression levels of BM-MSCs at P5 and P7 showed a downward trend compared with BM-MSCs at P2 regardless of harvesting methods. Noticeably, there was a substantial decline in the OCN protein expression level of trypsinization groups (P < 0.01). After 5 consecutive passages, the OCN protein expression level of thermal-liftoff groups was much lower than that of P2. Taken together, the results revealed that harvesting cells from thermo-responsive surface with temperature drop could better maintain the osteoblastic differentiation and mineralization ability during the long-term passage cultivation.

Fig. 7 ALP staining of osteoblasts (A) and BM-MSCs (B) harvested by thermal-liftoff and trypsinization; ALP activity assays of osteoblasts (C) and BM-MSCs (D). The data are expressed as the mean ± SD for n = 6 (*, P < 0.05; **, P < 0.01).

Fig. 8 Alizarin Red S staining of osteoblasts (A) and BM-MSCs (B) harvested by thermal-liftoff and trypsinization. Scale bar: 1000 μm.

Fig. 9 BGP/OCN protein expression of osteoblasts and BM-MSCs harvested with thermal-liftoff and trypsinization. Error bars represent means ± SD for n = 6. (*, P < 0.05; **, P < 0.01).

3.6 Chondrogenic differentiation of BM-MSCs

The harvested BM-MSCs of thermal-liftoff groups and trypsinization groups were reseeded on 24-wells plate. After cultivation for 2 weeks in chondrogenic medium, cells were stained with toluidine blue. As shown in Fig. 10, BM-MSCs in both thermal-liftoff groups and trypsinization groups were positive stained for toluidine blue, which indicated that the potential of chondrogenic differentiation of BM-MSCs in all groups. However, toluidine blue staining revealed more glycosaminoglycan (GAG) were synthesized and secreted into the ECM in thermal-liftoff groups than that of trypsinization groups. GAG is a typical and major component of the cartilage extracellular matrix. Therefore, the BM-MSCs harvested and passaged with thermal-liftoff owned more potential of chondrogenic differentiation.

Fig. 10 Toluidine blue staining of BM-MSCs after induction 2 weeks. A: P2; B: thermal-liftoff group, P5; C: thermal-liftoff group, P7; D: trypsinization group, P5; E: trypsinization group, P7.

4. Discussion

The inconvenience of calvaria, tibia and femur obtained from human tissues and the low frequency of MSC in bone marrow (0.01–0.001%) results in the low numbers of BM-MSCs and osteoblasts in the primary culture.^27–29 In fact, only 10⁵ to 10⁶ of BM-MSCs and 10⁶ to 10⁷ of osteoblasts are usually obtained in the primary culture when they were reached approximately 80–90% of confluence from 25–30 mL of human bone marrow aspirates and bone specimens, respectively.^30–32 And the population doubling times of BM-MSCs and osteoblasts are 30–70 h in the early passages (<7 trypsin passages).^30,33–35 In order to obtain 10⁹ to 10¹¹ cells for regenerative medicines, the cells need to be cultured for 4–7 weeks with 5–8 times passaging. Besides, osteoblasts within passage 10 were used in the researches of regenerative medicines;^28,36,37 and, BM-MSCs at passage 3–9 were investigated mostly because they displayed a similar colony-forming unit-fibroblast (CFU-F) potential and a constant growth rate.^27,29 Thus osteoblasts of P2 after 3, 5, 7 consecutive passages and BM-MSCs of P2 after 3, 5 consecutive passages were evaluated in this study.

Traditionally, anchorage-dependent cells are passaged and harvested with trypsin solution, which can digest extracellular matrices (ECM) and chelate Ca²⁺ ions, resulting in disrupt cell–cell junctions. As reported previously, enzyme digestion degrades plasma membrane proteins and deposited ECMs, leading to a reduction in cells viability and even multi-differentiation potential.^38,39 These findings are consistent with our present results. In the experiment of Calcein/AM staining, we found that there were many dead cells among released osteoblasts and BM-MSCs after enzymatic treatment; namely, cell viability reduced. Additionally, the proliferation of BM-MSCs fell consistently from P2 to P7.

The use of thermo-responsive substrates based on PNIPAAm has created a novel approach to harvest cells. In contrast with trypsinization method, the detachment of cells from thermo-responsive substrates can be achieved simply by lowering the temperature of the culturing medium down to below the LCST. In addition, higher cell viability of the harvested cell was demonstrated.^20,21 However, little studies were focused on whether this thermal-liftoff method favors to preserve the detached cells biological properties after consecutive passaging in the long-term culture. This research showed that thermal-liftoff method can better retain the detached cells' biological properties compared with trypsinization method.

The superiority in cell viability and proliferation of the harvested cells in the thermal-liftoff group could be ascribed to the concurrent detachment of an extracellular matrix (ECM) protein, fibronectin (FN). Laminiar (LN).⁴⁰ ECM is composed of glycosaminoglycans and proteoglycans, collagens and non-collagenous glycoproteins (such as laminin, tenascin and fibronectin), synthesized by the cells and secreted to the extracellular, or cell surface.²⁵ ECM forms a complex network structure by the connection between cells surface receptor integrin proteins and ECM proteins. This structure can anchor cell culture substrates and provide a suitable place for cells survival and migration. Moreover, cell proliferation and differentiation are regulated through the signal transduction based on ECM network.^25,26

In our research, the immunofluorescence staining images of FN and LN indicated that more FN and LN carried with the released cells of thermal-liftoff groups compared with those of trypsinization groups, and quantitative analysis showed that the amounts of total proteins, FN and LN carried with the harvested cells in thermal-liftoff groups are approximately 1.5–62 times compared with that of trypsinization groups. Additionally, the difference of two group cells carried total proteins, FN and LN was growing with the passage number. Taken together, these results indicated that more deposited ECM proteins were carried with the released cells with thermal-liftoff method, which may facilitate the maintenance of cell viability and cellular proliferation.⁴¹ In contrast, enzymatic treatment damaged ECM proteins and even influenced ECM synthesis and secretion, resulting in the loss of cell viability and cellular proliferation.

It is well known that ECM proteins have a profound influence on cell behaviors by affecting not only cell adhesion, viability and proliferation but also cell differentiation.²⁵ Osteoblastic differentiation and mineralization of osteoblasts and BM-MSCs was essential for regenerative medicines. ALP and OCN protein expression are important phenotypic markers for osteoblastic early-stage differentiation and latter-stage differentiation respectively. The formation of calcium phosphate salts or mineral deposition is a primary function of osteoblastic cells. Therefore, ALP activity, osteocalcin (OCN) protein expression and calcium deposits of the released cells after consecutive passages were evaluated. In this research, ALP of osteobalsts and BM-MSCs released with thermal-liftoff and trypsinization was positive expressed after serially passaged. However, ALP activity of thermal-liftoff group was higher than that of trypsinization group, especially when the number of passage increased. The difference of OCN protein expression and calcium deposition between the two groups was similar to that of ALP. In addition, we found that there had no obvious difference in the ALP activity and OCN protein expression of BM-MSCs at P5 detaching with temperature reduction from thermo-responsive in comparison with BM-MSCs at P2. Whereas, a substantial decline happened to BM-MSCs at P5 in trypsinization groups. Therefore, thermal-liftoff method not only better preserve osteoblastic functions of osteoblasts and BM-MSCs but also keep stronger osteogenic differentiation and chondrogenic differentiation potential of BM-MSCs, which should be attributed to thermal-liftoff retaining a large amount of ECM proteins. Previously reports indicated that ECM proteins played a vital role in regulating bone formation and osteoblastic differentiation mainly through the integrin mediated signal transduction by ERK/MAPK pathway. FN interacts with integrin α2β1, α3β1, α4β1 and α5β1, LN interacts with α1β1, α2β1, α3β1 and αvβ5. And further studies confirmed that FN was essential for early osteoblastic matrix mineralization.^42–44 In this study, Alizarin Red S staining was utilized as a good indicator for detecting early onset mineralization. The results showed more calcium nodules formation in thermal-liftoff groups compared with trypsinization groups, which indicated that detaching osteoblasts and BM-MSCs from thermo-responsive HFMs with temperature reduction was benefitted for the bone formation.

5. Conclusions

In this study, we investigated the effect of two different subculture methods, thermal-liftoff and trypsinization, on osteoblasts and BM-MSCs biological properties after 3, 5 and 7 consecutive passages. The results confirmed that harvesting cells from thermo-responsive surface with temperature reduction could better maintain cellular viability, proliferation, osteoblastic differentiation and mineralization, chondrogenic differentiation during the long-term cultivation in vitro. The preservation of osteoblasts and BM-MSCs biological properties were related to the large amount of ECM proteins carried with the detached cells in thermal-liftoff groups. Therefore, thermo-responsive HFMs and the matched thermal-liftoff method can be a candidate for harvesting more primitive therapeutic cells for regenerative medicines.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (31170945/31370991), Fok Ying Tung Education Foundation (132027), the State Key Laboratory of Fine Chemicals (KF1111) and the Fundamental Research Funds for the Central Universities (DUT12JB09) and SRF for ROCS, SEM.

References

1. P. Singh and D. J. Williams, J. Tissue Eng. Regener. Med., 2008, 2, 307–319.
2. C. J. Koh, J. Am. Soc. Nephrol., 2004, 15, 1113–1125.
3. A. P. Donghui Jing, John M. Canty and Emmanuel S. Tzanakakis, Tissue Eng., Part B, 2008, 14, 393–406.
4. M. E. Bernardo, A. M. Cometa, D. Pagliara, L. Vinti, F. Rossi, R. Cristantielli, G. Palumbo and F. Locatelli, Best Pract. Res., Clin. Haematol., 2011, 24, 73–81.
5. D. Melican, R. Butler, N. Hawkins, L. H. Chen, E. Hayden, M. Destrempes, J. Williams, T. Lewis, E. Behboodi, C. Ziomek, H. Meade, Y. Echelard and W. Gavin, Theriogenology, 2005, 63, 1549–1563.
6. A. Tamura, J. Kobayashi, M. Yamato and T. Okano, Biomaterials, 2012, 33, 3803–3812.
7. B. Sutradhar, J. Park, G. Hong, S. Choi and G. Kim, Pak. Vet. J., 2010, 30, 232–238.
8. J. V. Olsen, S. E. Ong and M. Mann, Mol. Cell. Proteomics, 2004, 3, 608–614.
9. Q. Zheng, S. M. Iqbal and Y. Wan, Biotechnol. Adv., 2013, 31, 1664–1675.
10. J. Collett, A. Crawford, P. V. Hatton, M. Geoghegan and S. Rimmer, J. R. Soc., Interface, 2007, 4, 117–126.
11. Y. C. Chen, R. Xie and L. Y. Chu, J. Membr. Sci., 2013, 442, 206–215.
12. N. G. Patel and G. Zhang, Organogenesis, 2013, 9, 93–100.
13. N. Yamada, T. Okano and H. Sakai, Macromol. Rapid Commun., 1990, 11, 571–576.
14. D. Ma, H. Chen, Z. Li and Q. He, Biomicrofluidics, 2010, 4, 44107.
15. H. Liu, X. Liu, J. Meng, P. Zhang, G. Yang, B. Su, K. Sun, L. Chen, D. Han, S. Wang and L. Jiang, Adv. Mater., 2013, 25, 922–927.
16. L. Yang, F. Cheng, T. Liu, J. R. Lu, K. Song, L. Jiang, S. Wu and W. Guo, Biomed. Mater., 2012, 7, 035003.
17. M. Zhuang, T. Liu, K. Song, D. Ge and X. Li, Mater. Sci. Eng., C, 2015, 55, 410–419.
18. K. D. Song, T. Q. Liu, X. Q. Li, Z. F. Cui, X. Y. Sun and X. H. Ma, Biomed. Environ. Sci., 2007, 20, 91–98.
19. T. F. M. Hans-Jürgen Boxberger, Biol. Cell, 1994, 82, 109–119.
20. H. S. Yang, O. Jeon, S. H. Bhang, S. H. Lee and B. S. Kim, Cell. Transplant., 2010, 19, 1123–1132.
21. D. Y. Shi, D. Ma, F. Q. Dong, C. Zong, L. Y. Liu, D. Shen, W. J. Yuan, X. M. Tong, H. W. Chen and J. F. Wang, Biosci. Rep., 2010, 30, 455.
22. M. Dominici, K. Le Blanc, I. Mueller, I. Slaper-Cortenbach, F. Marini, D. Krause, R. Deans, A. Keating, D. Prockop and E. Horwitz, Cytotherapy, 2006, 8, 315–317.
23. T. K. Georg Siegel, U. Hermanutz-Klein, K. Bieback, H. Northoff and R. Schäfer, BMC Med., 2013, 11, 1–20.
24. R. G. Thakar, Q. Cheng, S. Patel, J. Chu, M. Nasir, D. Liepmann, K. Komvopoulos and S. Li, Biophys. J., 2009, 96, 3423–3432.
25. T. Rozario and D. W. DeSimone, Dev. Biol., 2010, 341, 126–140.
26. F. M. Watt and W. T. Huck, Nat. Rev. Mol. Cell Biol., 2013, 14, 467–473.
27. L. G. Chase, U. Lakshmipathy, L. A. Solchaga, M. S. Rao and M. C. Vemuri, Stem Cell Res. Ther., 2010, 1, 8.
28. Q. Wan, T. Schoenmaker, I. D. Jansen, Z. Bian, T. J. de Vries and V. Everts, Bone, 2016, 86, 10–21.
29. J. W. Zhao, Z. L. Gao, H. Mei, Y. L. Li and Y. Wang, Am. J. Med. Sci., 2011, 341, 460–468.
30. P. J. Hanley, Z. Mei, A. G. Durett, G. Cabreira-Hansen Mda, M. Klis, W. Li, Y. Zhao, B. Yang, K. Parsha, O. Mir, F. Vahidy, D. Bloom, R. B. Rice, P. Hematti, S. I. Savitz and A. P. Gee, Cytotherapy, 2014, 16, 1048–1058.
31. M. E. Bernardo, M. A. Avanzini, C. Perotti, A. M. Cometa, A. Moretta, E. Lenta, C. D. Fante, F. Novara, A. D. Silvestri, G. Amendola, O. Zuffardi, R. Maccario and F. Locatelli, J. Cell. Physiol., 2007, 211, 121–130.
32. H. Declercq, Biomaterials, 2004, 25, 757–768.
33. C. E. LeBlon, M. E. Casey, C. R. Fodor, T. Zhang, X. Zhang and S. S. Jedlicka, Differentiation, 2015, 90, 1–15.
34. Y. Koshihara, M. Kawamura, S. Endo, C. Tsutsumi, H. Kodama, H. Oda and S. Higaki, In Vitro Cell. Dev. Biol., 1989, 25, 37–43.
35. S. B. Usuludin, X. Cao and M. Lim, Biotechnol. Bioeng., 2012, 109, 1248–1258.
36. F. P. Koch, C. Merkel, T. Ziebart, R. Smeets, C. Walter and B. Al-Nawas, Clin. Oral Investig., 2012, 16, 79–86.
37. S. Ehnert, K. Falldorf, A.-K. Fentz, P. Ziegler, S. Schröter, T. Freude, B. G. Ochs, C. Stacke, M. Ronniger, J. Sachtleben and A. K. Nussler, Bone Rep., 2015, 3, 48–56.
38. F. B. Zhang, L. Li, B. Fang, D. L. Zhu, H. T. Yang and P. J. Gao, Biochem. Biophys. Res. Commun., 2005, 336, 784–792.
39. F. Sugiura, H. Kitoh and N. Ishiguro, Biochem. Biophys. Res. Commun., 2004, 316, 233–239.
40. H. E. Canavan, X. H. Cheng, D. J. Graham, B. D. Ratner and D. G. Castner, Langmuir, 2005, 21, 1949–1955.
41. B. K. Sabrina Dey, M. R. Alexander, C. Alexander and F. R. A. J. Rose, J. Mater. Chem., 2011, 21, 6883–6890.
42. R. Globus, S. Doty, J. Lull, E. Holmuhamedov, M. Humphries and C. Damsky, J. Cell Sci., 1998, 111, 1385–1393.
43. A. M. M. R. K. G. C. H. Damsky, J. Cell Sci., 1997, 110, 2187–2196.
44. S. Mathews, R. Bhonde, P. K. Gupta and S. Totey, Differentiation, 2012, 84, 185–192.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra21946b

---