- Email: [email protected]
gelatin composite fibers in cell culture medium
Materials Science & Engineering C 100 (2019) 862–873
Contents lists available at ScienceDirect
Materials Science & Engineering C journal homepage: www.elsevier.com/locate/msec
Promoting osteogenic differentiation of BMSCs via mineralization of polylactide/gelatin composite fibers in cell culture medium
T ⁎
Man Caoa, Yan Zhoua, Jianping Maob, Pengfei Weia, Dafu Chenc, Renxian Wangc, Qing Caia, , ⁎ Xiaoping Yanga, a
State Key Laboratory of Organic-Inorganic Composites, Beijing Laboratory of Biomedical Materials, Beijing University of Chemical Technology, Beijing 100029, PR China Department of Spine Surgery, Beijing Jishuitan Hospital, Beijing 100035, PR China c Laboratory of Bone Tissue Engineering, Beijing Research Institute of Traumatology and Orthopaedics, Beijing Jishuitan Hospital, Beijing 100035, PR China. b
A R T I C LE I N FO
A B S T R A C T
Keywords: Mineralization Osteocompatibility Culture medium Polylactide Gelatin
Mineralization capability is an important issue in developing bone repairing biomaterials, while it is not quite clear how this feature would act in the presence of cells and influence cell osteogenic differentiation without adding extra osteoinductive factors such as β‑sodium glycerophosphate and dexamethasone. Poly(L‑lactide) (PLLA) and gelatin composite fibers (PG, 1:1 in weight) were electrospun, treated with CaCl2 solution (PG-Ca), and used for mineralization studies by using cell culture media (αMEM, and αMEM + serum). Bone mesenchymal stromal cells (BMSCs) were then seeded and cultured on both PG and PG-Ca fibrous mats for 28 days by only using αMEM + serum. Interestingly, mineral depositions on both PG and PG-Ca fibers were detected in the environment of αMEM or αMEM + serum, in which, PG-Ca fibers demonstrated stronger ability in inducing hydroxyapatite formation than PG fibers, especially in the presence of fetal bovine serum. When BMSCs were cultured on the two kinds of fibrous mats, apatite depositions were still clearly detected, while the depositing amounts decreased in comparison with corresponding cell-free cases. It was ascribed to the consumption of ions by the continuously proliferating BMSCs, whose osteogenic differentiation was significantly promoted even without extra osteoinductive factors, especially on PG-Ca fibrous mats, in comparison with the control group. Therefore, it was confirmed the capability of scaffolding materials in enriching ions like calcium and phosphate around cells was an efficient way to promote bone regeneration.
1. Introduction To meet the challenge in bone regeneration, there is a clear need for osteoinductive materials. These materials should have the capability to promote biomineralization, because new bone-tissue formation in vivo involves an essential process that apatite nucleates and grows on collagen fibrils [1]. Biomineralization studies on various substrates have been intensively carried out in vitro to judge the feasibility of using the materials for in vivo bone defect repairing [2–5]. Common aqueous solutions for these in vitro biomineralization studies are various simulated body fluids (SBF) with different recipes and degrees of supersaturation [6,7]. The SBF proposed by Kokubo et al. in 1991 has ion types and concentrations nearly resembling those in human blood plasma [8], and it was concluded that the examination of apatite formation on a material in SBF was useful and valid for predicting the in vivo bone bioactivity of a material [2–5]. Conventionally, in vitro biomineralization studies using SBFs were
⁎
performed in the absence of cells. For in vivo bone defect implantation, however, cell/material complexes were usually applied to accelerate osteogenesis instead of pure materials [9,10]. In this view, it is meaningful to carry out in vitro biomineralization studies in the presence of bone-related cells such as bone mesenchymal stromal cells (BMSCs). SBFs are not suitable for this purpose because they often contain high concentrations of various ions and are lack of nutrients to support cell growth. Some isotonic solutions, e.g. Ringer's physiological saline, Tyrode's saline, Krebs-Henseleit buffer (KHB), Earle's balanced salt solution (EBSS) and Hanks' balanced salt solution (HBSS), are frequently used in in vitro experiments on organs or tissues, while their amino acid-free feature makes them also improper in conducting apatite deposition experiments with the presence of cells [11]. Reasonably, physiological solutions containing amino acids, vitamins, glucose and inorganic ions simultaneously, such as Eagle's minimum essential medium (Eagle's MEM) and Dulbeccos modified Eagles medium (DMEM), can supply as alternatives to conventional SBFs and
Corresponding authors. E-mail addresses: [email protected] (Q. Cai), [email protected] (X. Yang).
https://doi.org/10.1016/j.msec.2019.02.079 Received 29 August 2017; Received in revised form 16 July 2018; Accepted 20 February 2019 Available online 20 March 2019 0928-4931/ © 2019 Elsevier B.V. All rights reserved.
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
(15 mm × 15 mm) and fixed on plastic rings (ϕ = 10 mm) to avoid mat shrinkage during the crosslinking treatment and the following biomineralization and cell culture studies. The prepared PLLA/gelatin composite fibrous mats were termed as PG for simplification. The PG fibrous mats were immersed in a CaCl2 solution (1 M) at room temperature for 4 h, and then retrieved, washed three times with deionized water, and freeze-dried. These fibrous mats were termed as PG-Ca.
physiological salines [12–15]. Both MEM and DMEM can provide sufficient ionic conditions to induce mineralization, at the same time, are able to maintain cell viability. In practical cell culture, fetal bovine serum (FBS) is usually required to support cell proliferation and differentiation, while it was reported that the presence of proteins or amino acids would lower the apatite formation rate on materials [16–18]. This might bring uncertainties in biomineralization studies along with the proliferation and differentiation of BMSCs in the environment of cell culture medium. Another issue associated with the biomineralization studies in the presence of BMSCs is suggested how the mineralization capability of the materials will influence the osteogenic differentiation of seeded BMSCs. It was revealed that the osteogenic differentiation of BMSCs could be significantly enhanced by properly increasing the concentration of calcium ion in the culture medium [19–21]. Therefore, a hypothesis proposed here is that those materials able to stimulate apatite formation in cell culture medium will favor the osteogenic differentiation of BMSCs by accumulating more calcium and/or phosphate ions around the cells, even if no extra calcium ion or osteoinductive factors like vitamin C, β‑sodium glycerophosphate and dexamethasone are introduced. With these uncertainties in mind, in this study, composite fibrous mats were electrospun from a blend solution containing poly(L‑lactide) (PLLA) and gelatin in the weight ratio of 1:1, and then submitted to biomineralization studies in MEM. Several groups were divided, i.e. PLLA/gelatin composite fibrous mats being pre-treated with CaCl2 solution or not; MEM containing FBS or not; BMSCs being seeded onto the fibrous mats or not. For all the experimental designs, the apatite formation on fibrous mats was systematically evaluated in monitoring changes in weight gaining, morphology, fiber diameter, chemical composition and crystalline structure. In the presence of seeded BMSCs, cell proliferation and osteogenic differentiation were identified both qualitatively and quantitatively for 28 days by using normal αMEM +10% FBS without extra osteoinductive factors. The correlation between biomineralization results and biological behaviors of BMSCs was discussed.
2.3. Sterilization treatment Before both the biomineralization studies and cell culture, sterilization on PG and PG-Ca fibrous mats were required. Briefly, the mats were immersed in 75% ethanol with exposure to ultraviolet light for 2 h, followed by washing three times with phosphate buffer saline (PBS, pH = 7.4) and being soaked in αMEM overnight for further use. 2.4. Cell culture Sprague-Dawley rat BMSCs (purchased from Cell Culture Center, Peking Union Medical College, China) were cultured in αMEM supplemented with 10% FBS, 100 IU/mL penicillin (Sigma) and 100 mg/ mL streptomycin (Sigma) in an incubator (Sanyo, Japan) with 5% CO2 supply at 37 °C and saturated humidity. Until 80% confluence prior to use, the BMSCs were digested by 0.25% trypsin (Sigma) and 0.02% ethylene diamine tetraacetic acid (EDTA) for further use. 2.5. Biomineralization studies To each well of 6-well culture plates, one plastic ring with fixed PG or PG-Ca fibrous mat was placed and physiological solution (10 mL) was added. The biomineralization studies were divided into three groups. In Group I, the physiological solution was αMEM supplemented with 100 IU/mL penicillin and 100 mg/mL streptomycin, and the kinds of samples were termed as PG-αMEM and PG-Ca-αMEM. In Group II, the physiological solution was prepared by adding 10% FBS into the medium used in Group I. In this case, the two kinds of samples were termed as PG-αMEM+FBS, PG-Ca-αMEM+FBS. In Group III, 2 × 103 BMSCs were seeded onto the PG or PG-Ca mats, and cultured with αMEM containing 10% FBS, 100 IU/mL penicillin and 100 mg/mL streptomycin, and the resulting samples were termed as PG-BMSC and PG-Ca-BMSC. All the biomineralization studies were performed in an incubator with 5% CO2 supply at 37 °C and saturated humidity to mimic the situation of cell culture. In all cases, the media were refreshed every 3 days. After being incubated for 3, 7, 14, 21 and 28 days, samples were retrieved from culture plates, rinsed gently with PBS three times. Subsequently, samples including PG-αMEM, PG-Ca-αMEM, PG-αMEM +FBS and PG-Ca-αMEM+FBS were further rinsed gently with deionized water and freeze-dried. BMSC-loaded samples including PGBMSC and PG-Ca-BMSC were fixed by 2.5% glutaraldehyde (Sigma) and stored at 4 °C for further characterizations.
2. Experimental section 2.1. Materials PLLA (Mw = 100,000), 2,2,2‑trifluoroethanol (TFE) (99%) and gelatin (type B, from bovine, pH 4.5–5.5, bloom 240–270) were purchased from Sigma-Aldrich. All of them were used for electrospinning without any further purification. Both 1‑ethyl‑3‑(3‑dimethylaminopropyl) carbodiimide (EDC, 97%) and N‑hydroxysuccinimide (NHS, 97%) were purchased from Sigma. αMEM and FBS were bought from Hyclone and Gibco, respectively, for biomineralization studies and cell culture. Other chemicals involved in this study were of analytically pure grade, which were obtained from Beijing Chemical Plant (China) and used as received. 2.2. Preparation of PLLA/gelatin composite fibers and CaCl2 solution pretreatment
2.6. Characterizations
PLLA/gelatin (1:1 in weight ratio) composite fibrous mats were electrospun as previously reported [22]. Briefly, PLLA (1 g) and gelatin (1 g) were co-dissolved in TFE (10 mL) to get a blend solution after overnight stirring. The solution (10 mL) was loaded into a 20 mL syringe fixed with a stainless needle and electrospun for 8 h under optimized parameters (flow rate: 0.4 mL/h; voltage: 15 kV; receiving distance: 20 cm). Then the as-spun composite fibrous mats were crosslinked by EDC (1.5 wt%) and NHS (nEDC:nNHS = 5:2) at 4 °C in ethanol/deionized water (Vethanol:Vwater = 90:10) for 12 h, followed by being washed three times with deionized water and freeze-dried. Before the crosslinking, the fibrous mats were cut into square pieces
Morphology observations were conducted on scanning electron microscope (SEM, Supera55, Zeiss, Germany) at an accelerating voltage of 15 kV after being sputter-coated with platinum (30 mA, 80 s) using a sputter coater (E5600, Polaron, USA). For each fibrous sample, multiple SEM micrographs were analyzed with Image Tool 2.0 to determine the average fiber diameter. At least, 200 fibers were measured and averaged. To identify chemical compositions and crystalline structures of the minerals deposited onto PG and PG-Ca mats under different situations, Fourier transform infrared (FTIR) and X-ray diffraction (XRD) analysis were applied. FTIR spectra were obtained using infrared spectrometer (Nicolet 6700, USA) with the wavenumber ranging from 863
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 1. Morphology and fiber diameter distribution of electrospun PLLA/gelatin composite fibers before and after being pre-treated with CaCl2 solution: (A–C) PG fibrous mat; (D–F) PG-Ca fibrous mat. The inset in image E indicates the presence of calcium element on PG-Ca fiber.
4000 to 500 cm−1 at a resolution of 4 cm−1. XRD patterns were recorded using an X-ray diffractometer (D/Max 2500VB2+ Rigaku, Japan) with a fixed incidence angle of 1° at a scanning rate of 10°/min in the range of 5–60° using CuKα radiation with a monochromator. To determine the Ca/P ratios of deposited minerals, energy-dispersive xray spectrometry (EDX) was performed similarly to SEM observation with exposure time of 180 s. The amounts of deposited minerals were measured by themogravimetric analysis (TGA, TA, Q-50, USA) in atmosphere from room temperature to 700 °C.
CCK-8 solution was added to each well and the system was incubated for 2 h at 37 °C. Then optical density (OD) values were measured by microreader (Muciskan FC) at the wavelength of 450 nm. Four measurements were performed for averaging. 2.8.2. Cell viability To evaluate the potential cytotoxicity of the produced PG and PG-Ca fibrous mats, live/dead staining assay was performed at 3, 7 and 14 days after BMSCs had been seeded onto the mats. PG-BMSC and PGCa-BMSC samples were stained with acridine orange / ethidium bromide (AO/EB), and fluorescent images were captured with a confocal laser scanning microscope (CLSM, TCS SP8, Leica).
2.7. Protein adsorption assay Protein adsorption was performed using 0.33% (w/v) bovine serum albumin (BSA) solution (Sigma). Before being incubated in the BSA solution, PG and PG-Ca fibrous mats (ϕ = 10 mm) were immersed in phosphate buffer saline (PBS) for 1 h. Then, they were transferred into a 24-well culture plate (one sample in each well), and 1.5 mL of BSA solution was added into each well. After 1 h, the concentration of residual protein in the solution was tested by BCA protein assay kit (Thermo, USA). The amount of adsorbed protein (mg/mg mat) was calculated basing on a standard curve determined with BSA solutions of known concentrations.
2.8.3. Cell morphology SEM micrographs were taken after the cells had attached and proliferated on PG or PG-Ca mats for 3, 7, 14, 21 and 28 days. Briefly, cellular constructs were retrieved at pre-determined time points, rinsed by PBS three times and then fixed by 2.5% glutaraldehyde. Afterwards, the fixed samples were dehydrated through a series of graded alcohol solutions, air-dried overnight, and submitted to SEM observation after being sputter-coated with platinum. For fluorescence observation, the retrieved cellular constructs were rinsed three times with PBS, stained by Hochest 33342 and rhodamine phalloidin. Fluorescent images were captured with CLSM under the excitation wavelengths of 350–370 nm and 550–580 nm, respectively.
2.8. Biological evaluations Similar to the situations in performing the biomineralization studies in the presence of BMSCs (Group III), cell viability, proliferation and osteogenic differentiation of seeded BMSCs were evaluated in parallel. In the assessment of osteogenic differentiation, for comparison, two control groups by using osteoinductive medium or not were also conducted by seeding BMSCs directly on the tissue culture polystyrene (TCPS), which was termed as TCPS+OS or TCPS-OS. The osteoinductive medium was prepared by adding 0.05 mmol/L vitamin C (Sigma), 10 mmol/L β‑sodium glycerophosphate (Sigma) and 1 × 10−8 mol/L dexamethasone (Sigma) into the aforementioned culture medium. In all cases, the media were refreshed every 3 days.
2.8.4. Alkaline phosphatase (ALP) activity After the cells or cell/mat complexes were incubated with or without osteoinductive medium for 3, 7, 14, 21 and 28 days, they were retrieved and rinsed by PBS for three times. The cells were then lysated by 400 mL lysate containing 1% Triton X-100, 20 mM Tris and 150 mM NaCl, and cell lysates were then obtained by freezing-thawing three times and centrifuged. The ALP activities in the lysates were then measured by an ALP assay kit (Sigma, USA) using p‑nitrophenyl phosphate (pNPP) method according to the manufacturer's instruction. Briefly, to 30 mL of each lysate, 100 μL of buffered saline solution containing pNPP was added. After being incubated at 37 °C for 15 min, the reaction was stopped with the addition of 150 μL NaOH solution. Absorbance values at 405 nm were then measured on microreader. The data were based on a standard curve provided by the Kit itself to quantify ALP activities. The ALP activities were normalized to the total protein content determined using the BCA assay kit. Four
2.8.1. Cell attachment and proliferation Cell attachment and proliferation was analyzed using Cell Counting Kit-8 (CCK-8, Dojindo, Japan). CCK-8 is a kind of yellow solution that can be reduced to orange by active cells, whose absorbance is directly proportional to cell number. At each predetermined time-point, 10 μL of 864
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 2. SEM observations on the mineralization of PG and PG-Ca fibers by being soaked in αMEM at 37 °C for 3–28 days, together with changes in their average fiber diameters and fiber diameter distributions.
Intensities of the colored solutions were measured at 450 nm on microplate reader. The intensity was inversely proportional to the Col-I concentration, which was calculated basing on the standard curve. The Col-I concentrations were normalized to the total protein content determined using the BCA assay kit. Four measurements were performed for averaging.
measurements were performed for averaging. At 14 days after the cells or cell/mat complexes were incubated with or without osteoinductive medium, they were retrieved for ALP staining. Briefly, the retrieved samples were rinsed by PBS for three times, fixed in 4% paraformaldehyde for 30 min, and then stained with BCIP/NBT ALP color development kit (Solarbio, China) for 30 min according to the manufacturer's instruction. After being washed three times with deionized water, the stained samples were observed and recorded using an optical microscope (CKX41, Olympus).
2.8.6. Calcium deposition Alizarin red staining was performed to qualitatively assess the deposition of calcium on all the substrates at 14 days of cell culture. Briefly, control samples on TCPS and cell/material complexes were washed three times with PBS and fixed with 4% paraformaldehyde in PBS for 30 min. After fixation, all the samples were washed with PBS and immersed in 1 mL of the staining solution (1% alizarin red S in PBS, Cyagen, USA) for 30 min, followed by washing with deionized water to detach non-specific staining from the substrates. Then the substrates were observed with an optical microscope and photos were taken.
2.8.5. Collagen I (Col-I) synthesis Cell lysates were prepared as aforementioned, and the synthesis of Col-I was tested by a Col-I ELISA kit (R&D, USA) following the manufacturer's instruction. Col-I ELISA kit applies the competitive enzyme immunoassay technique utilizing a monoclonal anti-Col-I antibody and a Col-I-HRP conjugate. The assay sample (100 μL) and buffer (10 μL) were incubated together with Col-I-HRP (50 μL) conjugate in pre-coated plates for 1 h. Afterwards, the wells were decanted and washed, followed by adding the substrate for HRP enzyme, in which, the enzymesubstrate reaction formed a blue colored complex. Subsequently, stop solution (50 μL) was added, and the sample solutions turned yellow.
2.9. Statistical analysis The experiments of biological evaluations were performed in 865
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 3. (A, D) TGA, (B, E) FTIR and (C, F) XRD analysis on mineralized PG (A–C) and PG-Ca fibers (D–F) along with soaking time in αMEM.
increased accordingly, from initial ~1 μm to the ~2 μm after 28 days of αMEM immersion (Fig. 2(A4–E4)). This increase in fiber diameter was not contributed by fiber swelling, because PG fibers only showed minor changes in fiber diameters after they were soaked in αMEM for the same periods (Fig. 2(A2–E2)). In comparison with PG-Ca fibers, PG fibers could be seen having quite limited mineral depositions at all the time points. When the soaking time was shorter than 14 days, almost no mineral nucleation was able to be identified on PG fibers (Fig. 2(A1–E1)). The mineralized PG and PG-Ca fibers were collected for characterizations including TGA, FTIR and XRD analysis. TGA curves were obtained under atmosphere, thus, polymeric components would decompose upon heating while inorganic minerals remained. Therefore, those remaining weights after the heating demonstrated the amounts of deposited minerals (Fig. 3(A, D)). For both mineralized PG and PG-Ca fibers, the remaining weights increased as the soaking time being prolonged, and the deposited minerals were more abundant on PG-Ca fibers than on PG fibers at the same time point, which were in accordance with the SEM observations. FTIR spectra of these mineralized fibers are shown in Fig. 3(B, E). In comparison with the spectra of original PG and PG-Ca fibers, the appearance of extra absorption peaks at 1025 cm−1 and 560 cm−1 indicated the deposited minerals mainly being phosphate compounds. With the continuous mineral deposition from 0 d to 28 d, it could be seen those signals belonging to PLLA and gelatin became weaker, while the peaks relating to phosphate group were intensified gradually. Crystal structures of the deposited minerals were then identified with XRD analysis. The initial PG and PG-Ca fibers demonstrated amorphous structure with no obvious diffraction signal (Fig. 3(C, F)). By being soaked in αMEM for several days, a diffraction peak was detected at 2θ = 16.6o, which was ascribed to the crystal structure of PLLA. Another two diffraction peaks at 2θ = 26o and 2θ = 32o were emerged for both PG and PG-Ca fibers after they were immersed 14 days in αMEM. Referring to the standard pattern of HA (PDF#01-084-1998), the two peaks at 2θ = 26o and 2θ = 32o were assigned to the (0 0 2) and (2 1 1) crystal plane of HA, respectively. Their intensities turned stronger and sharper in line with longer soaking time, especially in the PG-Ca case, which indicated that the ability of HA formation was obviously stronger on PG-Ca fibers than on PG fibers in αMEM. Reflecting by the shapes of the diffraction peaks, in both
quadruplicate (n = 4), and the results were presented as mean ± standard deviation (SD). Statistical difference was determined using Student's t-test for independent samples. Differences between groups of *p < 0.05 were considered statistically significant. 3. Results 3.1. Pre-treating PG fibers with CaCl2 solution Electrospinning from PLLA/gelatin blend solution could fabricate continuous and bead-free fibers with smooth surface as shown in Fig. 1(A, B). And the average diameter of PG fibers was estimated 0.96 μm (Fig. 1C). By soaking the PG fibrous mat in CaCl2 aqueous solution (1 M) at room temperature for 4 h, the morphology of retrieved PG-Ca fibers displayed some distortion after being freeze-dried, but remaining continuous and showing no hint of agglomeration between fibers (Fig. 1(D, E)). The PG-Ca fibers were measured having the average fiber diameter of 0.98 μm, which was close to that of PG fibers. From the inset in Fig. 1E, calcium element was clearly detected on the PG-Ca fiber, which revealed the pre-adsorption of calcium ions onto PG fibers. 3.2. Mineralization on PG and PG-Ca fibers in αMEM Mineralization studies were firstly carried out in αMEM to determine its feasibility in inducing apatite nucleation and growth. Both PG and PG-Ca fibrous mats were soaked in αMEM with the media being refreshed every 3 days, and the systems were placed in CO2 incubator to mimic the cell culture conditions. From Figs. 2 and S1, clearly, mineral depositions onto both PG and PG-Ca fibers were observed along with longer soaking time, indicating the occurrence of mineralization. The difference between the PG case and the PG-Ca case was their different mineral nucleating and depositing rates. Within 3 days, minerals were already detected on PG-Ca fibers, more and more minerals deposited gradually when the soaking time proceeded to 28 days (Fig. 2(A3–E3)). Those mineral depositions did not damage the fibrous network, on the contrary, the depositions arranged from loosely packed morphology into a kind of shell coating on PG-Ca fibers, showing flaky like structure (Fig. S2). And thus, the average diameters of mineralized PG-Ca fibers 866
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 4. SEM observations on the mineralization of PG and PG-Ca fibers by being soaked in αMEM containing 10% of FBS at 37 °C for 3–28 days, together with their changes in average fiber diameters and fiber diameter distributions.
FBS, as revealed by TGA data, were the rates in inducing apatite nucleation and accumulation (Fig. 5(A, B)). For the same kind of fibers (PG or PG-Ca), the amounts of deposited minerals was higher at the same soaking time point if the αMEM contained FBS. By using BSA solution, protein adsorption abilities of PG and PG-Ca fibers were tested. The results turned out to be that PG-Ca fibrous mat had higher protein adsorption amount (0.108 ± 0.003 mg/mg mat) than PG fibrous mat (0.028 ± 0.002 mg/mg mat). Then, it was suggested that the protein adsorption abilities of fibrous mats, especially the PG-Ca fibrous mat, led to their stronger potentials in inducing mineral depositions when the systems contained FBS. In these two cases, from Fig. S3(C, D), the Ca/P ratios of deposited minerals were still found above 2 for samples of 28 days.
cases, the crystallinity of the formed HA was not high after 28 days of mineralization in αMEM. Ca/P ratios of deposited minerals on both PG and PG-Ca fibers were further assessed with EDX, as shown in Fig. S3(A, B), they were estimated over 2 for samples of 28 days, slightly deviating from the 1.67 of stoichiometric HA. 3.3. Mineralization on PG and PG-ca fibers in αMEM+FBS To mimic the liquid condition used in cell culture, 10% of FBS was added into αMEM, followed by immersing PG and PG-Ca fibers into the media similarly for 28 days. As shown in Figs. 4 and S4, mineral nucleation and continuous deposition were clearly observed on both kinds of fibers, the changes in fiber morphology and diameter were closely in accordance with those images shown in Figs. 2 and S1. The PG-Ca fibers still displayed stronger ability in inducing mineralization than PG fibers, and the formed minerals displayed flaky like morphology (Fig. S5). From both FTIR (Fig. 5(B, E)) and XRD (Fig. 5(C, F)) analysis, the deposited minerals on PG and PG-Ca fibers were confirmed crystalline HA, showing minor difference from the former mineralization results in αMEM. The things different in the presence of FBS from the cases absent of
3.4. Mineralization on PG and PG-Ca fibers along with cell proliferation Attachment and proliferation of BMSCs on PG and PG-Ca fibrous mats were evaluated and the results are shown in Fig. 6. For ease of comparison, the attachment was normalized basing on the 4 h data in control, and the proliferation was normalized to the first day data in each case, respectively. It could be seen BMSCs intending to attach onto 867
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 5. (A, D) TGA, (B, E) FTIR and (C, F) XRD analysis on mineralized PG and PG-Ca fibers along with soaking time in αMEM containing 10% of FBS.
Fig. 6. (A) Attachment and (B) proliferation of BMSCs being cultured on PG and PG-Ca fibers.
morphology was able to be detected. Thus, PG and PG-Ca fibrous mats were of high cell affinity and insignificant cytotoxicity. Along with the cell proliferation, SEM observations were conducted on both cell/fiber complexes at low and high magnifications. From the images of lower magnifications (Fig. S6), cells could be seen adhering and spreading well on all fibrous mats, showing continuous proliferation along with longer culture time. At the later stage of culture (14–28 days), cells had excreted abundant extracellular matrix and grew confluent. The images in higher magnification revealed the details of PG and PG-Ca fibers without cell covering in order to study apatite deposition (Fig. 8). It was interesting to find that mineralization on both kinds of fibers could still occur along even with the cell proliferation, accordingly, fiber diameters increased gradually. Nucleation and mineral deposition were determined faster and more abundant on PG-Ca fibers than on PG fibers as the images shown. The deposited minerals were confirmed phosphate compounds by FTIR spectra (Fig. 9(B, E)) and weakly crystalline apatite by XRD patterns (Fig. 9(C, F)). The intensities of corresponding signals in these figures were weaker than those in Fig. 5, i.e. the mineralization in αMEM containing 10% FBS, indicating the presence of BMSCs having influence on mineral depositions. Other proofs to support this point came from the comparisons of TGA measurements (Figs. 5(A, D) and 9A(A, D)) and fiber diameters (Figs. 4 and 8) between the same kind of fibers (PG or PG-Ca) in the
fibrous mats in comparison with the flat TCPS surface, besides, PG fibers were able to enhance cell attachment more efficiently than PG-Ca fibers (Fig. 6A). Continuous proliferation was detected for BMSCs cultured on TCPS and PG fibrous mat during the 28 days, indicating the strong growth ability of BMSCs (Fig. 6B). On PG-Ca fibrous mats, BMSCs proliferated normally from 1 day to 14 days, while cell growth rate slowed down when the culture time was longer than 14 days. Moreover, BMSCs proliferated significantly faster on both fibrous mats than on TCPS within 7 days, while the growth rate of BMSCs on TCPS surpassed those cells on both fibrous mats as the culture time beyond 14 days. These variations suggested that BMSCs seeded on different substrates might go through different biological behaviors, mainly in their osteogenic differentiation, because the substrates had different capacities in inducing mineral deposition. Live/dead staining assay was applied to confirm the viability of BMSCs cultured on both fibrous mats and TCPS along with proliferation. As shown in Fig. 7A, cell numbers could be seen increasing gradually from 3 days to 14 days on all the substrates, showing consistent results with the cell proliferation data. The cells were dominantly stained green with few red spots, which indicated high cell viability in all the cases. The fluorescent images of cytoskeleton and nucleus staining vividly illustrated the spreading cell morphology, and the cell growth along with culture time (Fig. 7B). No deformation in cell 868
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 7. (A) Live/dead staining with AO/EB and (B) cell morphology staining with Hochest 33342/rhodamine phalloidin after BMSCs being cultured on PG and PG-Ca fibers for 3, 7 and 14 days. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
that osteogenic differentiation of BMSCs had progressed to some extent by judging the calcium depositions. Thus, the difference in the mineral capability between PG and PG-Ca fibers might bring different influence on osteogenic differentiation of BMSCs. To clarify this issue, osteogenic differentiation studies were conducted by culturing BMSCs on both fibrous mats without adding osteoinductive factors (i.e. PG-OS and PG-Ca-OS groups), while control groups were conducted on TCPS with (TCPS+OS) or without (TCPSOS) osteoinductive factors. Results of quantitative analysis on ALP activity and Col-I synthesis are presented in Fig. 11. For the two control groups, clearly, the osteogenic differentiation of BMSCs could be significantly enhanced by using osteoinductive medium, while the ALP activity and Col-I content were quite low if the osteoinductive factors were absent. But in the case of PG-OS group, the osteogenic differentiation of seeded BMSCs was found slightly enhanced in comparison with the TCPS-OS group, especially when the co-culture time was longer than 14 days. Interestingly, BMSCs in the PG-Ca-OS group demonstrated much stronger potential in osteogenic differentiation than those in the PG-OS group, which was confirmed by the significantly higher levels in both ALP activity and Col-I content in the former case.
cases of BMSCs present or absent under the same soaking time point. Along with cell proliferation, apparently, the amounts of deposited minerals were less and the fiber diameters were thinner in comparison with those cell-free groups. Combined with the cell proliferation results (Fig. 6B), these facts strongly suggested the significant correlations between cell biological behaviors and mineralization capacities of substrates used for cell culture. Interestingly, as revealed by Fig. S3(E, F), the Ca/P ratios of deposited minerals were identified around 1.6–1.7 for samples of 28 days in the presence of BMSCs. 3.5. Osteogenic differentiation of BMSCs along with mineralization Close looks at cells after being cultured on PG-Ca fibrous mats for 3, 14 and 28 days are presented in Fig. 10. Mineral depositions were not only detected on polymeric fibers, but also abundant on spreading cells. Moreover, the amounts of deposited minerals on cells increased along with longer culture time. However, no such phenomenon was found for cells cultured on PG fibers (Fig. S7). Even no extra osteoinductive factors (e.g. vitamin C, β‑sodium glycerophosphate, dexamethasone) were applied in the cell proliferation studies, thereby, it could be inferred 869
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 8. SEM observations on the mineralization of PG and PG-Ca fibers along with cell proliferation by being cultured for 3–28 days, together with their changes in average fiber diameters and fiber diameter distributions.
nucleation induced by scaffolding materials is able to be guaranteed [2,3,26]. Even so, it can not represent the real situation when the systems contain live cells and culture medium containing numerous amino acids and proteins is used to support cell growth. Culture medium, which is rich of various ions, thus can be a better choice than SBF to carry out biomineralization studies for the same purpose, especially in the case of adding FBS into culture medium to mimic cell culture conditions [11,12]. In this study, hydrophilic PG composite fibers were electrospun and used for the biomineralization studies of being soaked in different culture media (αMEM, or αMEM+FBS). The gelatin component in the composite fibers was able to provide nucleation sites for mineral deposition due to presence of functional group like carboxyl on fiber surface [17,27]. To promote the nucleation process, one group of PG fibers was pre-treated with CaCl2 solution, and calcium ions were preadsorped onto fiber surface via interaction with functional groups from the gelatin component [28]. When PG and PG-Ca fibers were soaked in αMEM, mineral depositions were readily induced, increasing in
Similarly, qualitative evaluations on ALP and calcium deposition of BMSCs being cultured in different groups for 14 days revealed the same trend. Undoubtedly, as shown in Fig. 12, the stained area was the largest and deepest in the TCPS+OS group, followed by PG-Ca-OS > PGOS > TCPS-OS group. All these results suggested that fibrous substrates favored the osteogenic differentiation of BMSCs in comparison with flat TCPS, and fibrous substrate having higher mineralization capability could further promote osteogenic differentiation via accelerating apatite formation. 4. Discussion The success of bone repairing via the strategy of tissue engineering depends significantly on the capability of scaffolding material in inducing osteogenic differentiation [23–25]. To evaluate the osteocompatibility of various bone tissue engineering scaffolds, in vitro biomineralization using SBF is regarded proper and effective as if the homogeneous nucleation in SBF can be inhibited and heterogeneous 870
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 9. (A, D) TGA, (B, E) FTIR and (C, F) XRD analysis on mineralized PG and PG-Ca fibers along with cell proliferation.
Fig. 10. SEM images showing mineral depositions on both PG-Ca fibers and spreading BMSCs along with cell proliferation by being cultured for 3, 14 and 28 days.
Fig. 11. (A) ALP activity and (B) Col-I synthesis of BMSCs being cultured on PG and PG-Ca fibrous mats without using osteoinductive factors (PG-OS and PG-Ca-OS groups) in comparison with the cases of TCPS-OS and TCPS+OS controls.
presence of protein ingredients in comparing corresponding samples (Figs. 4 and 5). This phenomenon seemed inconsistent with reports that addition of proteins or amino acids into SBF would stabilize calcium and phosphate ions from nucleation and deposition because of their high water solubility [29–31]. But we found that both PG and PG-Ca fibrous mat were able to adsorb proteins, especially the PG-Ca fibrous mat, which was thought the key point to strength the mineralization
amounts as the soaking time being prolonged. As expected, the PG-Ca fibers were indeed able to enhance the accumulation of minerals on fibers and the formation of crystalline HA in comparison with PG fibers (Figs. 2 and 3). With the addition of FBS, the mineralization behaviors on both PG and PG-Ca fibers were quite similar to the former αMEM cases, while the amounts of the deposited minerals were increased apparently in the 871
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 12. ALP staining and alizarin red staining for calcium depositions of BMSCs being cultured on PG and PG-Ca fibrous mats without using osteoinductive factors (PG-OS and PG-Ca-OS groups) for 14 days in comparison with the cases of TCPS-OS and TCPS+OS controls. Scale bar = 250 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
capacities of the two fibrous mats in αMEM containing FBS. Then it was curious to know how the fibrous mats to behave in inducing mineralization in the presence of cells and along with cell proliferation. Every thing has two faces. Another issue concerned in performing this cell culture experiment was how the mineralization capability of fibrous mat to influence osteogenic differentiation of seeded BMSCs without adding extra osteoinductive factors. There were two hypotheses here. One hypothesis was the mineralization on fibers being inhibited due to the consumption of ions by cells. The other hypothesis was the fibers able to accumulate ions around adjacent cells and to stimulate osteogenic differentiation of cells thereof. From our experimental results, these two hypotheses were solidly proven. Along with cell proliferation as long as 28 days, mineral depositions on both PG and PG-Ca fibers were definitely detected and PG-Ca fibers still demonstrated stronger ability in inducing apatite formation. In comparison with those cell-free cases, however, the amounts of deposited minerals significantly decreased in the presence of cells, which indicated the usage of ions by cells indeed interfering mineralization (Figs. 4 and 8). On the other hand, the osteogenic differentiation of BMSCs was significantly enhanced on PG-Ca fibers than on both PG fibers and TCPS in the absence of extra osteoinductive factors (Figs. 11 and 12). It was reported that properly higher concentration of calcium ions in culture medium could favor osteogenic differentiation via stimulating calcium ion sensitive receptors on cell membranes [32–34]. And it was observed that BMSCs cultured on PG-Ca fibers themselves were covered with abundant mineral depositions in addition to those mineralized fibers (Fig. 10). Reasonably, it was inferred that the strong capability of PG-Ca fibers in attracting ions and in inducing mineral deposition created a favorable micro-environment for adjacent BMSCs taking the lead in proceeding into the differentiation. For the mineralized fibers in different cases, it was found the deposited minerals were flaky like, which were normal morphology of HA [17,27]. However, their Ca/P ratios were different (Fig. S3). It could be seen the Ca/P ratios were above 2 in the cases of cell being absent (with or without FBS), while it was more closer to the 1.67 of HA in the presence of BMSCs. The reason was suggested that PG and PG-Ca fibers were affinity to calcium ions due to the carboxyl groups originating from gelatin component, which likely led to higher Ca/P ratios if the deposited minerals had not transformed into well crystallized HA (Figs. 3 and 5). In the presence of BMSCs, the assumption of calcium ion by cells and the calcium deposition alongside osteogenic differentiation of BMSCs were suggested the possible reasons to lead the closer Ca/P ratios to that of stoichiometric HA. Summarized from all aforementioned discussions, it was reliable to say that the osteocompatibility of a bone tissue engineering scaffolding material was able to be reflected by its mineralization capability in SBF
or in cell culture medium. When the material was co-cultured with cells, its strong capability in inducing mineralization played vital functions in enriching osteoinductive ions around those cells attaching and spreading on the material. Therefore, the osteogenic differentiation of cells was thus significantly promoted even in the absence of commonly used osteoinductive factors like β‑sodium glycerophosphate and dexamethasone. 5. Conclusion Electrospun PG composite fibrous mats could mimic the microstructure of collagen fibrous network in native ECM, therefore, were good substrates used in biomineralization studies. Its capability in inducing apatite formation could be strengthened by being simply pretreated with CaCl2 solution, from which, those pre-adsorbed calcium ions could act as pre-nucleation sites to accelerate mineral depositions. Cell culture medium and cell culture medium containing FBS were both tested effective in providing sufficient ions to form apatite depositions on PG and PG-Ca fibers, and the PG-Ca fibers displayed significantly higher potential in protein adsorption and faster rate in mineralization. These features endowed the PG-Ca fibrous mats strong ability in promoting mineral depositions on seeded cells and thus accelerating osteogenic differentiation of the cells via the enrichment of ions like calcium and phosphate ions. The opposite side of this phenomenon was the amounts of deposited minerals on the fibrous mats being reduced due to the consumption of ions by cells. To draw a conclusion, it was feasible and efficient to prepare biomaterials with strengthened mineralization capability in cell culture systems, which would play vital function in facilitating bone regeneration. Acknowledgements The authors acknowledged the financial support from National Key R&D Program of China (2017YFC1104302/4300), National Natural Science Foundation of China (51873013), and Beijing Municipal Natural Science Foundation (7182068, 7161001). Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.msec.2019.02.079. References [1] J.D. Pasteris, B. Wopenka, E. Valsami-Jones, Bone and tooth mineralization: why apatite? Elements 4 (2008) 97–104.
872
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
[20] L. Yang, S. Perez-Amodio, F.Y. Groot Barrère-de, V. Everts, C.A. van Blitterswijk, P. Habibovic, The effects of inorganic additives to calcium phosphate on in vitro behavior of osteoblasts and osteoclasts, Biomaterials 31 (2010) 2976–2989. [21] Y.K. Liu, Q.Z. Lu, R. Pei, H.J. Ji, G.S. Zhou, X.L. Zhao, R.K. Tang, M. Zhang, The effect of extracellular calcium and inorganic phosphate on the growth and osteogenic differentiation of mesenchymal stem cells in vitro: implication for bone tissue engineering, Biomed. Mater. 4 (2009) 025004. [22] X. Yang, Q. Xu, N. Yan, G. Sui, Q. Cai, X. Deng, Structure and wettability relationship of coelectrospun poly(L-lactic acid)/gelatin composite fibrous mats, Polym. Adv. Technol. 22 (2011) 2222–2230. [23] S.H. Lee, H. Shin, Matrices and scaffolds for delivery of bioactive molecules in bone and cartilage tissue engineering, Adv. Drug Deliv. Rev. 59 (2007) 339–359. [24] S. Wang, F. Hu, J. Li, S. Zhang, M. Shen, M. Huang, X. Shi, Design of electrospun nanofibrous mats for osteogenic differentiation of mesenchymal stem cells, Nanomedicine (2017) 30083–30087 S1549-9634. [25] S. Stratton, N.B. Shelke, K. Hoshino, S. Rudraiah, S.G. Kumbar, Bioactive polymeric scaffolds for tissue engineering, Bioact. Mater. 1 (2016) 93–108. [26] K. Bleek, A. Taubert, New developments in polymer-controlled, bioinspired calcium phosphate mineralization from aqueous solution, Acta Biomater. 9 (2013) 6283–6321. [27] Q. Cai, Q. Xu, Q. Feng, X. Cao, X. Yang, X. Deng, Biomineralization of electrospun poly(L-lactic acid)/gelatin composite fibrous scaffold by using a supersaturated simulated body fluid with continuous CO2 bubbling, Appl. Surf. Sci. 257 (2011) 10109–10118. [28] K. Iijima, A. Sakai, A. Komori, Y. Sakamoto, H. Matsuno, T. Serizawa, M. Hashizume, Control of biomimetic hydroxyapatite deposition on polymer substrates using different protein adsorption abilities, Colloids Surf. B: Biointerfaces 130 (2015) 77–83. [29] X. Yang, B. Xie, L. Wang, Y. Qin, Z.J. Henneman, G.H. Nancollas, Influence of magnesium ions and amino acids on the nucleation and growth of hydroxyapatite, Crys. Eng. Comm. 13 (2011) 1153–1158. [30] B. Palazzo, D. Walsh, M. Iafisco, E. Foresti, L. Bertinetti, G. Marta, C.L. Bianchi, G. Cappelletti, N. Roveri, Amino acid synergetic effect on structure, morphology and surface properties of biomimetic apatite nanocrystals, Acta Biomater. 5 (2009) 1241–1252. [31] K. Wang, Y. Leng, X. Lu, F. Ren, Calcium phosphate bioceramics induce mineralization modulated by proteins, Mater. Sci. Eng. C 33 (2013) 3245–3255. [32] S. Maeno, Y. Niki, H. Matsumoto, H. Morioka, T. Yatabe, A. Funayama, Y. Toyama, T. Taguchi, J. Tanaka, The effect of calcium ion concentration on osteoblast viability, proliferation and differentiation in monolayer and 3D culture, Biomaterials 26 (2005) 4847–4855. [33] X. Zhang, S. Meng, Y. Huang, M. Xu, Y. He, H. Lin, J. Han, Y. Chai, Y. Wei, X. Deng, Electrospun gelatin/-TCP composite nanofibers enhance osteogenic differentiation of BMSCs and in vivo bone formation by activating Ca2+-sensing receptor signaling, Stem Cells Int. 2015 (2015) 507154. [34] J. Ye, W. Ai, F. Zhang, X. Zhu, G. Shu, L. Wang, P. Gao, Q. Xi, Y. Zhang, Q. Jiang, S. Wang, Enhanced proliferation of porcine bone marrow mesenchymal stem cells induced by extracellular calcium is associated with the activation of the calciumsensing receptor and ERK signaling pathway, Stem Cells Int. 2016 (2016) 6570671.
[2] M. Bohner, J. Lemaitre, Can bioactivity be tested in vitro with SBF solution? Biomaterials 30 (2009) 2175–2179. [3] T. Kokubo, H. Takadama, How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 27 (2006) 2907–2915. [4] A.A. Zadpoor, Relationship between in vitro apatite-forming ability measured using simulated body fluid and in vivo bioactivity of biomaterials, Mater. Sci. Eng. C 35 (2014) 134–143. [5] T. Kokubo, S. Yamaguchi, Novel bioactive materials developed by simulated body fluid evaluation: surface-modified Ti metal and its alloys, Acta Biomater. 44 (2016) 16–30. [6] X. Lu, Y. Leng, Theoretical analysis of calcium phosphate precipitation in simulated body fluid, Biomaterials 26 (2005) 1097–1108. [7] X. Chen, Y. Li, P.D. Hodgson, C. Wen, Microstructures and bond strengths of the calcium phosphate coatings formed on titanium from different simulated body fluids, Mater. Sci. Eng. C 29 (2009) 165–171. [8] T. Kokubo, Bioactive glass ceramics: properties and applications, Biomaterials 12 (1991) 155–163. [9] H.J. Park, S.J. Yu, K. Yang, Y. Jin, A.N. Cho, J. Kim, B. Lee, H.S. Yang, S.G. Im, S.W. Cho, Paper-based bioactive scaffolds for stem cell-mediated bone tissue engineering, Biomaterials 35 (2014) 9811–9823. [10] Q. Yao, J.G. Cosme, T. Xu, J.M. Miszuk, P.H. Picciani, H. Fong, H. Sun, Three dimensional electrospun PCL/PLA blend nanofibrous scaffolds with significantly improved stem cells osteogenic differentiation and cranial bone formation, Biomaterials 115 (2017) 115–127. [11] A.C. Tas, The use of physiological solutions or media in calcium phosphate synthesis and processing, Acta Biomater. 10 (2014) 1771–1792. [12] J.T. Lee, Y. Leng, K.L. Chow, F. Ren, X. Ge, K. Wang, X. Lu, Cell culture medium as an alternative to conventional simulated body fluid, Acta Biomater. 7 (2011) 2615–2622. [13] R.R. Rao, A. Jiao, D.H. Kohn, J.P. Stegemann, Exogenous mineralization of cellseeded and unseeded collagen-chitosan hydrogels using modified culture medium, Acta Biomater. 8 (2012) 1560–1565. [14] Ł. John, M. Bałtrukiewicz, P. Sobota, R. Brykner, Ł. Cwynar-Zając, P. Dzięgiel, Noncytotoxic organic-inorganic hybrid bioscaffolds: an efficient bedding for rapid growth of bone-like apatite and cell proliferation, Mater. Sci. Eng. C 32 (2012) 1849–1858. [15] V. Wagener, S. Virtanen, Protective layer formation on magnesium in cell culture medium, Mater. Sci. Eng. C 63 (2016) 341–351. [16] P. Liu, J. Tao, Y. Cai, H. Pan, X. Xu, R. Tang, Role of fetal bovine serum in theprevention of calcification in biological fluids, J. Cryst. Growth 310 (2008) 4672–4675. [17] Y. Guo, J. Lan, C. Zhang, M. Cao, Q. Cai, X. Yang, Mineralization on polylactide/ gelatin composite nanofibers using simulated body fluid containing amino acid, Appl. Surf. Sci. 349 (2015) 538–548. [18] D.C. Bassett, L.M. Grover, F.A. Müller, M.D. McKee, J.E. Barralet, Serum protein controlled nanoparticle synthesis, Adv. Funct. Mater. 21 (2011) 2968–2977. [19] S. An, J. Ling, Y. Gao, Y. Xiao, Effects of varied ionic calcium and phosphate on the proliferation, osteogenic differentiation and mineralization of human periodontal ligament cells in vitro, J. Periodontal Res. 47 (2012) 374–382.
873
Contents lists available at ScienceDirect
Materials Science & Engineering C journal homepage: www.elsevier.com/locate/msec
Promoting osteogenic differentiation of BMSCs via mineralization of polylactide/gelatin composite fibers in cell culture medium
T ⁎
Man Caoa, Yan Zhoua, Jianping Maob, Pengfei Weia, Dafu Chenc, Renxian Wangc, Qing Caia, , ⁎ Xiaoping Yanga, a
State Key Laboratory of Organic-Inorganic Composites, Beijing Laboratory of Biomedical Materials, Beijing University of Chemical Technology, Beijing 100029, PR China Department of Spine Surgery, Beijing Jishuitan Hospital, Beijing 100035, PR China c Laboratory of Bone Tissue Engineering, Beijing Research Institute of Traumatology and Orthopaedics, Beijing Jishuitan Hospital, Beijing 100035, PR China. b
A R T I C LE I N FO
A B S T R A C T
Keywords: Mineralization Osteocompatibility Culture medium Polylactide Gelatin
Mineralization capability is an important issue in developing bone repairing biomaterials, while it is not quite clear how this feature would act in the presence of cells and influence cell osteogenic differentiation without adding extra osteoinductive factors such as β‑sodium glycerophosphate and dexamethasone. Poly(L‑lactide) (PLLA) and gelatin composite fibers (PG, 1:1 in weight) were electrospun, treated with CaCl2 solution (PG-Ca), and used for mineralization studies by using cell culture media (αMEM, and αMEM + serum). Bone mesenchymal stromal cells (BMSCs) were then seeded and cultured on both PG and PG-Ca fibrous mats for 28 days by only using αMEM + serum. Interestingly, mineral depositions on both PG and PG-Ca fibers were detected in the environment of αMEM or αMEM + serum, in which, PG-Ca fibers demonstrated stronger ability in inducing hydroxyapatite formation than PG fibers, especially in the presence of fetal bovine serum. When BMSCs were cultured on the two kinds of fibrous mats, apatite depositions were still clearly detected, while the depositing amounts decreased in comparison with corresponding cell-free cases. It was ascribed to the consumption of ions by the continuously proliferating BMSCs, whose osteogenic differentiation was significantly promoted even without extra osteoinductive factors, especially on PG-Ca fibrous mats, in comparison with the control group. Therefore, it was confirmed the capability of scaffolding materials in enriching ions like calcium and phosphate around cells was an efficient way to promote bone regeneration.
1. Introduction To meet the challenge in bone regeneration, there is a clear need for osteoinductive materials. These materials should have the capability to promote biomineralization, because new bone-tissue formation in vivo involves an essential process that apatite nucleates and grows on collagen fibrils [1]. Biomineralization studies on various substrates have been intensively carried out in vitro to judge the feasibility of using the materials for in vivo bone defect repairing [2–5]. Common aqueous solutions for these in vitro biomineralization studies are various simulated body fluids (SBF) with different recipes and degrees of supersaturation [6,7]. The SBF proposed by Kokubo et al. in 1991 has ion types and concentrations nearly resembling those in human blood plasma [8], and it was concluded that the examination of apatite formation on a material in SBF was useful and valid for predicting the in vivo bone bioactivity of a material [2–5]. Conventionally, in vitro biomineralization studies using SBFs were
⁎
performed in the absence of cells. For in vivo bone defect implantation, however, cell/material complexes were usually applied to accelerate osteogenesis instead of pure materials [9,10]. In this view, it is meaningful to carry out in vitro biomineralization studies in the presence of bone-related cells such as bone mesenchymal stromal cells (BMSCs). SBFs are not suitable for this purpose because they often contain high concentrations of various ions and are lack of nutrients to support cell growth. Some isotonic solutions, e.g. Ringer's physiological saline, Tyrode's saline, Krebs-Henseleit buffer (KHB), Earle's balanced salt solution (EBSS) and Hanks' balanced salt solution (HBSS), are frequently used in in vitro experiments on organs or tissues, while their amino acid-free feature makes them also improper in conducting apatite deposition experiments with the presence of cells [11]. Reasonably, physiological solutions containing amino acids, vitamins, glucose and inorganic ions simultaneously, such as Eagle's minimum essential medium (Eagle's MEM) and Dulbeccos modified Eagles medium (DMEM), can supply as alternatives to conventional SBFs and
Corresponding authors. E-mail addresses: [email protected] (Q. Cai), [email protected] (X. Yang).
https://doi.org/10.1016/j.msec.2019.02.079 Received 29 August 2017; Received in revised form 16 July 2018; Accepted 20 February 2019 Available online 20 March 2019 0928-4931/ © 2019 Elsevier B.V. All rights reserved.
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
(15 mm × 15 mm) and fixed on plastic rings (ϕ = 10 mm) to avoid mat shrinkage during the crosslinking treatment and the following biomineralization and cell culture studies. The prepared PLLA/gelatin composite fibrous mats were termed as PG for simplification. The PG fibrous mats were immersed in a CaCl2 solution (1 M) at room temperature for 4 h, and then retrieved, washed three times with deionized water, and freeze-dried. These fibrous mats were termed as PG-Ca.
physiological salines [12–15]. Both MEM and DMEM can provide sufficient ionic conditions to induce mineralization, at the same time, are able to maintain cell viability. In practical cell culture, fetal bovine serum (FBS) is usually required to support cell proliferation and differentiation, while it was reported that the presence of proteins or amino acids would lower the apatite formation rate on materials [16–18]. This might bring uncertainties in biomineralization studies along with the proliferation and differentiation of BMSCs in the environment of cell culture medium. Another issue associated with the biomineralization studies in the presence of BMSCs is suggested how the mineralization capability of the materials will influence the osteogenic differentiation of seeded BMSCs. It was revealed that the osteogenic differentiation of BMSCs could be significantly enhanced by properly increasing the concentration of calcium ion in the culture medium [19–21]. Therefore, a hypothesis proposed here is that those materials able to stimulate apatite formation in cell culture medium will favor the osteogenic differentiation of BMSCs by accumulating more calcium and/or phosphate ions around the cells, even if no extra calcium ion or osteoinductive factors like vitamin C, β‑sodium glycerophosphate and dexamethasone are introduced. With these uncertainties in mind, in this study, composite fibrous mats were electrospun from a blend solution containing poly(L‑lactide) (PLLA) and gelatin in the weight ratio of 1:1, and then submitted to biomineralization studies in MEM. Several groups were divided, i.e. PLLA/gelatin composite fibrous mats being pre-treated with CaCl2 solution or not; MEM containing FBS or not; BMSCs being seeded onto the fibrous mats or not. For all the experimental designs, the apatite formation on fibrous mats was systematically evaluated in monitoring changes in weight gaining, morphology, fiber diameter, chemical composition and crystalline structure. In the presence of seeded BMSCs, cell proliferation and osteogenic differentiation were identified both qualitatively and quantitatively for 28 days by using normal αMEM +10% FBS without extra osteoinductive factors. The correlation between biomineralization results and biological behaviors of BMSCs was discussed.
2.3. Sterilization treatment Before both the biomineralization studies and cell culture, sterilization on PG and PG-Ca fibrous mats were required. Briefly, the mats were immersed in 75% ethanol with exposure to ultraviolet light for 2 h, followed by washing three times with phosphate buffer saline (PBS, pH = 7.4) and being soaked in αMEM overnight for further use. 2.4. Cell culture Sprague-Dawley rat BMSCs (purchased from Cell Culture Center, Peking Union Medical College, China) were cultured in αMEM supplemented with 10% FBS, 100 IU/mL penicillin (Sigma) and 100 mg/ mL streptomycin (Sigma) in an incubator (Sanyo, Japan) with 5% CO2 supply at 37 °C and saturated humidity. Until 80% confluence prior to use, the BMSCs were digested by 0.25% trypsin (Sigma) and 0.02% ethylene diamine tetraacetic acid (EDTA) for further use. 2.5. Biomineralization studies To each well of 6-well culture plates, one plastic ring with fixed PG or PG-Ca fibrous mat was placed and physiological solution (10 mL) was added. The biomineralization studies were divided into three groups. In Group I, the physiological solution was αMEM supplemented with 100 IU/mL penicillin and 100 mg/mL streptomycin, and the kinds of samples were termed as PG-αMEM and PG-Ca-αMEM. In Group II, the physiological solution was prepared by adding 10% FBS into the medium used in Group I. In this case, the two kinds of samples were termed as PG-αMEM+FBS, PG-Ca-αMEM+FBS. In Group III, 2 × 103 BMSCs were seeded onto the PG or PG-Ca mats, and cultured with αMEM containing 10% FBS, 100 IU/mL penicillin and 100 mg/mL streptomycin, and the resulting samples were termed as PG-BMSC and PG-Ca-BMSC. All the biomineralization studies were performed in an incubator with 5% CO2 supply at 37 °C and saturated humidity to mimic the situation of cell culture. In all cases, the media were refreshed every 3 days. After being incubated for 3, 7, 14, 21 and 28 days, samples were retrieved from culture plates, rinsed gently with PBS three times. Subsequently, samples including PG-αMEM, PG-Ca-αMEM, PG-αMEM +FBS and PG-Ca-αMEM+FBS were further rinsed gently with deionized water and freeze-dried. BMSC-loaded samples including PGBMSC and PG-Ca-BMSC were fixed by 2.5% glutaraldehyde (Sigma) and stored at 4 °C for further characterizations.
2. Experimental section 2.1. Materials PLLA (Mw = 100,000), 2,2,2‑trifluoroethanol (TFE) (99%) and gelatin (type B, from bovine, pH 4.5–5.5, bloom 240–270) were purchased from Sigma-Aldrich. All of them were used for electrospinning without any further purification. Both 1‑ethyl‑3‑(3‑dimethylaminopropyl) carbodiimide (EDC, 97%) and N‑hydroxysuccinimide (NHS, 97%) were purchased from Sigma. αMEM and FBS were bought from Hyclone and Gibco, respectively, for biomineralization studies and cell culture. Other chemicals involved in this study were of analytically pure grade, which were obtained from Beijing Chemical Plant (China) and used as received. 2.2. Preparation of PLLA/gelatin composite fibers and CaCl2 solution pretreatment
2.6. Characterizations
PLLA/gelatin (1:1 in weight ratio) composite fibrous mats were electrospun as previously reported [22]. Briefly, PLLA (1 g) and gelatin (1 g) were co-dissolved in TFE (10 mL) to get a blend solution after overnight stirring. The solution (10 mL) was loaded into a 20 mL syringe fixed with a stainless needle and electrospun for 8 h under optimized parameters (flow rate: 0.4 mL/h; voltage: 15 kV; receiving distance: 20 cm). Then the as-spun composite fibrous mats were crosslinked by EDC (1.5 wt%) and NHS (nEDC:nNHS = 5:2) at 4 °C in ethanol/deionized water (Vethanol:Vwater = 90:10) for 12 h, followed by being washed three times with deionized water and freeze-dried. Before the crosslinking, the fibrous mats were cut into square pieces
Morphology observations were conducted on scanning electron microscope (SEM, Supera55, Zeiss, Germany) at an accelerating voltage of 15 kV after being sputter-coated with platinum (30 mA, 80 s) using a sputter coater (E5600, Polaron, USA). For each fibrous sample, multiple SEM micrographs were analyzed with Image Tool 2.0 to determine the average fiber diameter. At least, 200 fibers were measured and averaged. To identify chemical compositions and crystalline structures of the minerals deposited onto PG and PG-Ca mats under different situations, Fourier transform infrared (FTIR) and X-ray diffraction (XRD) analysis were applied. FTIR spectra were obtained using infrared spectrometer (Nicolet 6700, USA) with the wavenumber ranging from 863
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 1. Morphology and fiber diameter distribution of electrospun PLLA/gelatin composite fibers before and after being pre-treated with CaCl2 solution: (A–C) PG fibrous mat; (D–F) PG-Ca fibrous mat. The inset in image E indicates the presence of calcium element on PG-Ca fiber.
4000 to 500 cm−1 at a resolution of 4 cm−1. XRD patterns were recorded using an X-ray diffractometer (D/Max 2500VB2+ Rigaku, Japan) with a fixed incidence angle of 1° at a scanning rate of 10°/min in the range of 5–60° using CuKα radiation with a monochromator. To determine the Ca/P ratios of deposited minerals, energy-dispersive xray spectrometry (EDX) was performed similarly to SEM observation with exposure time of 180 s. The amounts of deposited minerals were measured by themogravimetric analysis (TGA, TA, Q-50, USA) in atmosphere from room temperature to 700 °C.
CCK-8 solution was added to each well and the system was incubated for 2 h at 37 °C. Then optical density (OD) values were measured by microreader (Muciskan FC) at the wavelength of 450 nm. Four measurements were performed for averaging. 2.8.2. Cell viability To evaluate the potential cytotoxicity of the produced PG and PG-Ca fibrous mats, live/dead staining assay was performed at 3, 7 and 14 days after BMSCs had been seeded onto the mats. PG-BMSC and PGCa-BMSC samples were stained with acridine orange / ethidium bromide (AO/EB), and fluorescent images were captured with a confocal laser scanning microscope (CLSM, TCS SP8, Leica).
2.7. Protein adsorption assay Protein adsorption was performed using 0.33% (w/v) bovine serum albumin (BSA) solution (Sigma). Before being incubated in the BSA solution, PG and PG-Ca fibrous mats (ϕ = 10 mm) were immersed in phosphate buffer saline (PBS) for 1 h. Then, they were transferred into a 24-well culture plate (one sample in each well), and 1.5 mL of BSA solution was added into each well. After 1 h, the concentration of residual protein in the solution was tested by BCA protein assay kit (Thermo, USA). The amount of adsorbed protein (mg/mg mat) was calculated basing on a standard curve determined with BSA solutions of known concentrations.
2.8.3. Cell morphology SEM micrographs were taken after the cells had attached and proliferated on PG or PG-Ca mats for 3, 7, 14, 21 and 28 days. Briefly, cellular constructs were retrieved at pre-determined time points, rinsed by PBS three times and then fixed by 2.5% glutaraldehyde. Afterwards, the fixed samples were dehydrated through a series of graded alcohol solutions, air-dried overnight, and submitted to SEM observation after being sputter-coated with platinum. For fluorescence observation, the retrieved cellular constructs were rinsed three times with PBS, stained by Hochest 33342 and rhodamine phalloidin. Fluorescent images were captured with CLSM under the excitation wavelengths of 350–370 nm and 550–580 nm, respectively.
2.8. Biological evaluations Similar to the situations in performing the biomineralization studies in the presence of BMSCs (Group III), cell viability, proliferation and osteogenic differentiation of seeded BMSCs were evaluated in parallel. In the assessment of osteogenic differentiation, for comparison, two control groups by using osteoinductive medium or not were also conducted by seeding BMSCs directly on the tissue culture polystyrene (TCPS), which was termed as TCPS+OS or TCPS-OS. The osteoinductive medium was prepared by adding 0.05 mmol/L vitamin C (Sigma), 10 mmol/L β‑sodium glycerophosphate (Sigma) and 1 × 10−8 mol/L dexamethasone (Sigma) into the aforementioned culture medium. In all cases, the media were refreshed every 3 days.
2.8.4. Alkaline phosphatase (ALP) activity After the cells or cell/mat complexes were incubated with or without osteoinductive medium for 3, 7, 14, 21 and 28 days, they were retrieved and rinsed by PBS for three times. The cells were then lysated by 400 mL lysate containing 1% Triton X-100, 20 mM Tris and 150 mM NaCl, and cell lysates were then obtained by freezing-thawing three times and centrifuged. The ALP activities in the lysates were then measured by an ALP assay kit (Sigma, USA) using p‑nitrophenyl phosphate (pNPP) method according to the manufacturer's instruction. Briefly, to 30 mL of each lysate, 100 μL of buffered saline solution containing pNPP was added. After being incubated at 37 °C for 15 min, the reaction was stopped with the addition of 150 μL NaOH solution. Absorbance values at 405 nm were then measured on microreader. The data were based on a standard curve provided by the Kit itself to quantify ALP activities. The ALP activities were normalized to the total protein content determined using the BCA assay kit. Four
2.8.1. Cell attachment and proliferation Cell attachment and proliferation was analyzed using Cell Counting Kit-8 (CCK-8, Dojindo, Japan). CCK-8 is a kind of yellow solution that can be reduced to orange by active cells, whose absorbance is directly proportional to cell number. At each predetermined time-point, 10 μL of 864
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 2. SEM observations on the mineralization of PG and PG-Ca fibers by being soaked in αMEM at 37 °C for 3–28 days, together with changes in their average fiber diameters and fiber diameter distributions.
Intensities of the colored solutions were measured at 450 nm on microplate reader. The intensity was inversely proportional to the Col-I concentration, which was calculated basing on the standard curve. The Col-I concentrations were normalized to the total protein content determined using the BCA assay kit. Four measurements were performed for averaging.
measurements were performed for averaging. At 14 days after the cells or cell/mat complexes were incubated with or without osteoinductive medium, they were retrieved for ALP staining. Briefly, the retrieved samples were rinsed by PBS for three times, fixed in 4% paraformaldehyde for 30 min, and then stained with BCIP/NBT ALP color development kit (Solarbio, China) for 30 min according to the manufacturer's instruction. After being washed three times with deionized water, the stained samples were observed and recorded using an optical microscope (CKX41, Olympus).
2.8.6. Calcium deposition Alizarin red staining was performed to qualitatively assess the deposition of calcium on all the substrates at 14 days of cell culture. Briefly, control samples on TCPS and cell/material complexes were washed three times with PBS and fixed with 4% paraformaldehyde in PBS for 30 min. After fixation, all the samples were washed with PBS and immersed in 1 mL of the staining solution (1% alizarin red S in PBS, Cyagen, USA) for 30 min, followed by washing with deionized water to detach non-specific staining from the substrates. Then the substrates were observed with an optical microscope and photos were taken.
2.8.5. Collagen I (Col-I) synthesis Cell lysates were prepared as aforementioned, and the synthesis of Col-I was tested by a Col-I ELISA kit (R&D, USA) following the manufacturer's instruction. Col-I ELISA kit applies the competitive enzyme immunoassay technique utilizing a monoclonal anti-Col-I antibody and a Col-I-HRP conjugate. The assay sample (100 μL) and buffer (10 μL) were incubated together with Col-I-HRP (50 μL) conjugate in pre-coated plates for 1 h. Afterwards, the wells were decanted and washed, followed by adding the substrate for HRP enzyme, in which, the enzymesubstrate reaction formed a blue colored complex. Subsequently, stop solution (50 μL) was added, and the sample solutions turned yellow.
2.9. Statistical analysis The experiments of biological evaluations were performed in 865
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 3. (A, D) TGA, (B, E) FTIR and (C, F) XRD analysis on mineralized PG (A–C) and PG-Ca fibers (D–F) along with soaking time in αMEM.
increased accordingly, from initial ~1 μm to the ~2 μm after 28 days of αMEM immersion (Fig. 2(A4–E4)). This increase in fiber diameter was not contributed by fiber swelling, because PG fibers only showed minor changes in fiber diameters after they were soaked in αMEM for the same periods (Fig. 2(A2–E2)). In comparison with PG-Ca fibers, PG fibers could be seen having quite limited mineral depositions at all the time points. When the soaking time was shorter than 14 days, almost no mineral nucleation was able to be identified on PG fibers (Fig. 2(A1–E1)). The mineralized PG and PG-Ca fibers were collected for characterizations including TGA, FTIR and XRD analysis. TGA curves were obtained under atmosphere, thus, polymeric components would decompose upon heating while inorganic minerals remained. Therefore, those remaining weights after the heating demonstrated the amounts of deposited minerals (Fig. 3(A, D)). For both mineralized PG and PG-Ca fibers, the remaining weights increased as the soaking time being prolonged, and the deposited minerals were more abundant on PG-Ca fibers than on PG fibers at the same time point, which were in accordance with the SEM observations. FTIR spectra of these mineralized fibers are shown in Fig. 3(B, E). In comparison with the spectra of original PG and PG-Ca fibers, the appearance of extra absorption peaks at 1025 cm−1 and 560 cm−1 indicated the deposited minerals mainly being phosphate compounds. With the continuous mineral deposition from 0 d to 28 d, it could be seen those signals belonging to PLLA and gelatin became weaker, while the peaks relating to phosphate group were intensified gradually. Crystal structures of the deposited minerals were then identified with XRD analysis. The initial PG and PG-Ca fibers demonstrated amorphous structure with no obvious diffraction signal (Fig. 3(C, F)). By being soaked in αMEM for several days, a diffraction peak was detected at 2θ = 16.6o, which was ascribed to the crystal structure of PLLA. Another two diffraction peaks at 2θ = 26o and 2θ = 32o were emerged for both PG and PG-Ca fibers after they were immersed 14 days in αMEM. Referring to the standard pattern of HA (PDF#01-084-1998), the two peaks at 2θ = 26o and 2θ = 32o were assigned to the (0 0 2) and (2 1 1) crystal plane of HA, respectively. Their intensities turned stronger and sharper in line with longer soaking time, especially in the PG-Ca case, which indicated that the ability of HA formation was obviously stronger on PG-Ca fibers than on PG fibers in αMEM. Reflecting by the shapes of the diffraction peaks, in both
quadruplicate (n = 4), and the results were presented as mean ± standard deviation (SD). Statistical difference was determined using Student's t-test for independent samples. Differences between groups of *p < 0.05 were considered statistically significant. 3. Results 3.1. Pre-treating PG fibers with CaCl2 solution Electrospinning from PLLA/gelatin blend solution could fabricate continuous and bead-free fibers with smooth surface as shown in Fig. 1(A, B). And the average diameter of PG fibers was estimated 0.96 μm (Fig. 1C). By soaking the PG fibrous mat in CaCl2 aqueous solution (1 M) at room temperature for 4 h, the morphology of retrieved PG-Ca fibers displayed some distortion after being freeze-dried, but remaining continuous and showing no hint of agglomeration between fibers (Fig. 1(D, E)). The PG-Ca fibers were measured having the average fiber diameter of 0.98 μm, which was close to that of PG fibers. From the inset in Fig. 1E, calcium element was clearly detected on the PG-Ca fiber, which revealed the pre-adsorption of calcium ions onto PG fibers. 3.2. Mineralization on PG and PG-Ca fibers in αMEM Mineralization studies were firstly carried out in αMEM to determine its feasibility in inducing apatite nucleation and growth. Both PG and PG-Ca fibrous mats were soaked in αMEM with the media being refreshed every 3 days, and the systems were placed in CO2 incubator to mimic the cell culture conditions. From Figs. 2 and S1, clearly, mineral depositions onto both PG and PG-Ca fibers were observed along with longer soaking time, indicating the occurrence of mineralization. The difference between the PG case and the PG-Ca case was their different mineral nucleating and depositing rates. Within 3 days, minerals were already detected on PG-Ca fibers, more and more minerals deposited gradually when the soaking time proceeded to 28 days (Fig. 2(A3–E3)). Those mineral depositions did not damage the fibrous network, on the contrary, the depositions arranged from loosely packed morphology into a kind of shell coating on PG-Ca fibers, showing flaky like structure (Fig. S2). And thus, the average diameters of mineralized PG-Ca fibers 866
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 4. SEM observations on the mineralization of PG and PG-Ca fibers by being soaked in αMEM containing 10% of FBS at 37 °C for 3–28 days, together with their changes in average fiber diameters and fiber diameter distributions.
FBS, as revealed by TGA data, were the rates in inducing apatite nucleation and accumulation (Fig. 5(A, B)). For the same kind of fibers (PG or PG-Ca), the amounts of deposited minerals was higher at the same soaking time point if the αMEM contained FBS. By using BSA solution, protein adsorption abilities of PG and PG-Ca fibers were tested. The results turned out to be that PG-Ca fibrous mat had higher protein adsorption amount (0.108 ± 0.003 mg/mg mat) than PG fibrous mat (0.028 ± 0.002 mg/mg mat). Then, it was suggested that the protein adsorption abilities of fibrous mats, especially the PG-Ca fibrous mat, led to their stronger potentials in inducing mineral depositions when the systems contained FBS. In these two cases, from Fig. S3(C, D), the Ca/P ratios of deposited minerals were still found above 2 for samples of 28 days.
cases, the crystallinity of the formed HA was not high after 28 days of mineralization in αMEM. Ca/P ratios of deposited minerals on both PG and PG-Ca fibers were further assessed with EDX, as shown in Fig. S3(A, B), they were estimated over 2 for samples of 28 days, slightly deviating from the 1.67 of stoichiometric HA. 3.3. Mineralization on PG and PG-ca fibers in αMEM+FBS To mimic the liquid condition used in cell culture, 10% of FBS was added into αMEM, followed by immersing PG and PG-Ca fibers into the media similarly for 28 days. As shown in Figs. 4 and S4, mineral nucleation and continuous deposition were clearly observed on both kinds of fibers, the changes in fiber morphology and diameter were closely in accordance with those images shown in Figs. 2 and S1. The PG-Ca fibers still displayed stronger ability in inducing mineralization than PG fibers, and the formed minerals displayed flaky like morphology (Fig. S5). From both FTIR (Fig. 5(B, E)) and XRD (Fig. 5(C, F)) analysis, the deposited minerals on PG and PG-Ca fibers were confirmed crystalline HA, showing minor difference from the former mineralization results in αMEM. The things different in the presence of FBS from the cases absent of
3.4. Mineralization on PG and PG-Ca fibers along with cell proliferation Attachment and proliferation of BMSCs on PG and PG-Ca fibrous mats were evaluated and the results are shown in Fig. 6. For ease of comparison, the attachment was normalized basing on the 4 h data in control, and the proliferation was normalized to the first day data in each case, respectively. It could be seen BMSCs intending to attach onto 867
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 5. (A, D) TGA, (B, E) FTIR and (C, F) XRD analysis on mineralized PG and PG-Ca fibers along with soaking time in αMEM containing 10% of FBS.
Fig. 6. (A) Attachment and (B) proliferation of BMSCs being cultured on PG and PG-Ca fibers.
morphology was able to be detected. Thus, PG and PG-Ca fibrous mats were of high cell affinity and insignificant cytotoxicity. Along with the cell proliferation, SEM observations were conducted on both cell/fiber complexes at low and high magnifications. From the images of lower magnifications (Fig. S6), cells could be seen adhering and spreading well on all fibrous mats, showing continuous proliferation along with longer culture time. At the later stage of culture (14–28 days), cells had excreted abundant extracellular matrix and grew confluent. The images in higher magnification revealed the details of PG and PG-Ca fibers without cell covering in order to study apatite deposition (Fig. 8). It was interesting to find that mineralization on both kinds of fibers could still occur along even with the cell proliferation, accordingly, fiber diameters increased gradually. Nucleation and mineral deposition were determined faster and more abundant on PG-Ca fibers than on PG fibers as the images shown. The deposited minerals were confirmed phosphate compounds by FTIR spectra (Fig. 9(B, E)) and weakly crystalline apatite by XRD patterns (Fig. 9(C, F)). The intensities of corresponding signals in these figures were weaker than those in Fig. 5, i.e. the mineralization in αMEM containing 10% FBS, indicating the presence of BMSCs having influence on mineral depositions. Other proofs to support this point came from the comparisons of TGA measurements (Figs. 5(A, D) and 9A(A, D)) and fiber diameters (Figs. 4 and 8) between the same kind of fibers (PG or PG-Ca) in the
fibrous mats in comparison with the flat TCPS surface, besides, PG fibers were able to enhance cell attachment more efficiently than PG-Ca fibers (Fig. 6A). Continuous proliferation was detected for BMSCs cultured on TCPS and PG fibrous mat during the 28 days, indicating the strong growth ability of BMSCs (Fig. 6B). On PG-Ca fibrous mats, BMSCs proliferated normally from 1 day to 14 days, while cell growth rate slowed down when the culture time was longer than 14 days. Moreover, BMSCs proliferated significantly faster on both fibrous mats than on TCPS within 7 days, while the growth rate of BMSCs on TCPS surpassed those cells on both fibrous mats as the culture time beyond 14 days. These variations suggested that BMSCs seeded on different substrates might go through different biological behaviors, mainly in their osteogenic differentiation, because the substrates had different capacities in inducing mineral deposition. Live/dead staining assay was applied to confirm the viability of BMSCs cultured on both fibrous mats and TCPS along with proliferation. As shown in Fig. 7A, cell numbers could be seen increasing gradually from 3 days to 14 days on all the substrates, showing consistent results with the cell proliferation data. The cells were dominantly stained green with few red spots, which indicated high cell viability in all the cases. The fluorescent images of cytoskeleton and nucleus staining vividly illustrated the spreading cell morphology, and the cell growth along with culture time (Fig. 7B). No deformation in cell 868
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 7. (A) Live/dead staining with AO/EB and (B) cell morphology staining with Hochest 33342/rhodamine phalloidin after BMSCs being cultured on PG and PG-Ca fibers for 3, 7 and 14 days. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
that osteogenic differentiation of BMSCs had progressed to some extent by judging the calcium depositions. Thus, the difference in the mineral capability between PG and PG-Ca fibers might bring different influence on osteogenic differentiation of BMSCs. To clarify this issue, osteogenic differentiation studies were conducted by culturing BMSCs on both fibrous mats without adding osteoinductive factors (i.e. PG-OS and PG-Ca-OS groups), while control groups were conducted on TCPS with (TCPS+OS) or without (TCPSOS) osteoinductive factors. Results of quantitative analysis on ALP activity and Col-I synthesis are presented in Fig. 11. For the two control groups, clearly, the osteogenic differentiation of BMSCs could be significantly enhanced by using osteoinductive medium, while the ALP activity and Col-I content were quite low if the osteoinductive factors were absent. But in the case of PG-OS group, the osteogenic differentiation of seeded BMSCs was found slightly enhanced in comparison with the TCPS-OS group, especially when the co-culture time was longer than 14 days. Interestingly, BMSCs in the PG-Ca-OS group demonstrated much stronger potential in osteogenic differentiation than those in the PG-OS group, which was confirmed by the significantly higher levels in both ALP activity and Col-I content in the former case.
cases of BMSCs present or absent under the same soaking time point. Along with cell proliferation, apparently, the amounts of deposited minerals were less and the fiber diameters were thinner in comparison with those cell-free groups. Combined with the cell proliferation results (Fig. 6B), these facts strongly suggested the significant correlations between cell biological behaviors and mineralization capacities of substrates used for cell culture. Interestingly, as revealed by Fig. S3(E, F), the Ca/P ratios of deposited minerals were identified around 1.6–1.7 for samples of 28 days in the presence of BMSCs. 3.5. Osteogenic differentiation of BMSCs along with mineralization Close looks at cells after being cultured on PG-Ca fibrous mats for 3, 14 and 28 days are presented in Fig. 10. Mineral depositions were not only detected on polymeric fibers, but also abundant on spreading cells. Moreover, the amounts of deposited minerals on cells increased along with longer culture time. However, no such phenomenon was found for cells cultured on PG fibers (Fig. S7). Even no extra osteoinductive factors (e.g. vitamin C, β‑sodium glycerophosphate, dexamethasone) were applied in the cell proliferation studies, thereby, it could be inferred 869
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 8. SEM observations on the mineralization of PG and PG-Ca fibers along with cell proliferation by being cultured for 3–28 days, together with their changes in average fiber diameters and fiber diameter distributions.
nucleation induced by scaffolding materials is able to be guaranteed [2,3,26]. Even so, it can not represent the real situation when the systems contain live cells and culture medium containing numerous amino acids and proteins is used to support cell growth. Culture medium, which is rich of various ions, thus can be a better choice than SBF to carry out biomineralization studies for the same purpose, especially in the case of adding FBS into culture medium to mimic cell culture conditions [11,12]. In this study, hydrophilic PG composite fibers were electrospun and used for the biomineralization studies of being soaked in different culture media (αMEM, or αMEM+FBS). The gelatin component in the composite fibers was able to provide nucleation sites for mineral deposition due to presence of functional group like carboxyl on fiber surface [17,27]. To promote the nucleation process, one group of PG fibers was pre-treated with CaCl2 solution, and calcium ions were preadsorped onto fiber surface via interaction with functional groups from the gelatin component [28]. When PG and PG-Ca fibers were soaked in αMEM, mineral depositions were readily induced, increasing in
Similarly, qualitative evaluations on ALP and calcium deposition of BMSCs being cultured in different groups for 14 days revealed the same trend. Undoubtedly, as shown in Fig. 12, the stained area was the largest and deepest in the TCPS+OS group, followed by PG-Ca-OS > PGOS > TCPS-OS group. All these results suggested that fibrous substrates favored the osteogenic differentiation of BMSCs in comparison with flat TCPS, and fibrous substrate having higher mineralization capability could further promote osteogenic differentiation via accelerating apatite formation. 4. Discussion The success of bone repairing via the strategy of tissue engineering depends significantly on the capability of scaffolding material in inducing osteogenic differentiation [23–25]. To evaluate the osteocompatibility of various bone tissue engineering scaffolds, in vitro biomineralization using SBF is regarded proper and effective as if the homogeneous nucleation in SBF can be inhibited and heterogeneous 870
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 9. (A, D) TGA, (B, E) FTIR and (C, F) XRD analysis on mineralized PG and PG-Ca fibers along with cell proliferation.
Fig. 10. SEM images showing mineral depositions on both PG-Ca fibers and spreading BMSCs along with cell proliferation by being cultured for 3, 14 and 28 days.
Fig. 11. (A) ALP activity and (B) Col-I synthesis of BMSCs being cultured on PG and PG-Ca fibrous mats without using osteoinductive factors (PG-OS and PG-Ca-OS groups) in comparison with the cases of TCPS-OS and TCPS+OS controls.
presence of protein ingredients in comparing corresponding samples (Figs. 4 and 5). This phenomenon seemed inconsistent with reports that addition of proteins or amino acids into SBF would stabilize calcium and phosphate ions from nucleation and deposition because of their high water solubility [29–31]. But we found that both PG and PG-Ca fibrous mat were able to adsorb proteins, especially the PG-Ca fibrous mat, which was thought the key point to strength the mineralization
amounts as the soaking time being prolonged. As expected, the PG-Ca fibers were indeed able to enhance the accumulation of minerals on fibers and the formation of crystalline HA in comparison with PG fibers (Figs. 2 and 3). With the addition of FBS, the mineralization behaviors on both PG and PG-Ca fibers were quite similar to the former αMEM cases, while the amounts of the deposited minerals were increased apparently in the 871
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
Fig. 12. ALP staining and alizarin red staining for calcium depositions of BMSCs being cultured on PG and PG-Ca fibrous mats without using osteoinductive factors (PG-OS and PG-Ca-OS groups) for 14 days in comparison with the cases of TCPS-OS and TCPS+OS controls. Scale bar = 250 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
capacities of the two fibrous mats in αMEM containing FBS. Then it was curious to know how the fibrous mats to behave in inducing mineralization in the presence of cells and along with cell proliferation. Every thing has two faces. Another issue concerned in performing this cell culture experiment was how the mineralization capability of fibrous mat to influence osteogenic differentiation of seeded BMSCs without adding extra osteoinductive factors. There were two hypotheses here. One hypothesis was the mineralization on fibers being inhibited due to the consumption of ions by cells. The other hypothesis was the fibers able to accumulate ions around adjacent cells and to stimulate osteogenic differentiation of cells thereof. From our experimental results, these two hypotheses were solidly proven. Along with cell proliferation as long as 28 days, mineral depositions on both PG and PG-Ca fibers were definitely detected and PG-Ca fibers still demonstrated stronger ability in inducing apatite formation. In comparison with those cell-free cases, however, the amounts of deposited minerals significantly decreased in the presence of cells, which indicated the usage of ions by cells indeed interfering mineralization (Figs. 4 and 8). On the other hand, the osteogenic differentiation of BMSCs was significantly enhanced on PG-Ca fibers than on both PG fibers and TCPS in the absence of extra osteoinductive factors (Figs. 11 and 12). It was reported that properly higher concentration of calcium ions in culture medium could favor osteogenic differentiation via stimulating calcium ion sensitive receptors on cell membranes [32–34]. And it was observed that BMSCs cultured on PG-Ca fibers themselves were covered with abundant mineral depositions in addition to those mineralized fibers (Fig. 10). Reasonably, it was inferred that the strong capability of PG-Ca fibers in attracting ions and in inducing mineral deposition created a favorable micro-environment for adjacent BMSCs taking the lead in proceeding into the differentiation. For the mineralized fibers in different cases, it was found the deposited minerals were flaky like, which were normal morphology of HA [17,27]. However, their Ca/P ratios were different (Fig. S3). It could be seen the Ca/P ratios were above 2 in the cases of cell being absent (with or without FBS), while it was more closer to the 1.67 of HA in the presence of BMSCs. The reason was suggested that PG and PG-Ca fibers were affinity to calcium ions due to the carboxyl groups originating from gelatin component, which likely led to higher Ca/P ratios if the deposited minerals had not transformed into well crystallized HA (Figs. 3 and 5). In the presence of BMSCs, the assumption of calcium ion by cells and the calcium deposition alongside osteogenic differentiation of BMSCs were suggested the possible reasons to lead the closer Ca/P ratios to that of stoichiometric HA. Summarized from all aforementioned discussions, it was reliable to say that the osteocompatibility of a bone tissue engineering scaffolding material was able to be reflected by its mineralization capability in SBF
or in cell culture medium. When the material was co-cultured with cells, its strong capability in inducing mineralization played vital functions in enriching osteoinductive ions around those cells attaching and spreading on the material. Therefore, the osteogenic differentiation of cells was thus significantly promoted even in the absence of commonly used osteoinductive factors like β‑sodium glycerophosphate and dexamethasone. 5. Conclusion Electrospun PG composite fibrous mats could mimic the microstructure of collagen fibrous network in native ECM, therefore, were good substrates used in biomineralization studies. Its capability in inducing apatite formation could be strengthened by being simply pretreated with CaCl2 solution, from which, those pre-adsorbed calcium ions could act as pre-nucleation sites to accelerate mineral depositions. Cell culture medium and cell culture medium containing FBS were both tested effective in providing sufficient ions to form apatite depositions on PG and PG-Ca fibers, and the PG-Ca fibers displayed significantly higher potential in protein adsorption and faster rate in mineralization. These features endowed the PG-Ca fibrous mats strong ability in promoting mineral depositions on seeded cells and thus accelerating osteogenic differentiation of the cells via the enrichment of ions like calcium and phosphate ions. The opposite side of this phenomenon was the amounts of deposited minerals on the fibrous mats being reduced due to the consumption of ions by cells. To draw a conclusion, it was feasible and efficient to prepare biomaterials with strengthened mineralization capability in cell culture systems, which would play vital function in facilitating bone regeneration. Acknowledgements The authors acknowledged the financial support from National Key R&D Program of China (2017YFC1104302/4300), National Natural Science Foundation of China (51873013), and Beijing Municipal Natural Science Foundation (7182068, 7161001). Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.msec.2019.02.079. References [1] J.D. Pasteris, B. Wopenka, E. Valsami-Jones, Bone and tooth mineralization: why apatite? Elements 4 (2008) 97–104.
872
Materials Science & Engineering C 100 (2019) 862–873
M. Cao, et al.
[20] L. Yang, S. Perez-Amodio, F.Y. Groot Barrère-de, V. Everts, C.A. van Blitterswijk, P. Habibovic, The effects of inorganic additives to calcium phosphate on in vitro behavior of osteoblasts and osteoclasts, Biomaterials 31 (2010) 2976–2989. [21] Y.K. Liu, Q.Z. Lu, R. Pei, H.J. Ji, G.S. Zhou, X.L. Zhao, R.K. Tang, M. Zhang, The effect of extracellular calcium and inorganic phosphate on the growth and osteogenic differentiation of mesenchymal stem cells in vitro: implication for bone tissue engineering, Biomed. Mater. 4 (2009) 025004. [22] X. Yang, Q. Xu, N. Yan, G. Sui, Q. Cai, X. Deng, Structure and wettability relationship of coelectrospun poly(L-lactic acid)/gelatin composite fibrous mats, Polym. Adv. Technol. 22 (2011) 2222–2230. [23] S.H. Lee, H. Shin, Matrices and scaffolds for delivery of bioactive molecules in bone and cartilage tissue engineering, Adv. Drug Deliv. Rev. 59 (2007) 339–359. [24] S. Wang, F. Hu, J. Li, S. Zhang, M. Shen, M. Huang, X. Shi, Design of electrospun nanofibrous mats for osteogenic differentiation of mesenchymal stem cells, Nanomedicine (2017) 30083–30087 S1549-9634. [25] S. Stratton, N.B. Shelke, K. Hoshino, S. Rudraiah, S.G. Kumbar, Bioactive polymeric scaffolds for tissue engineering, Bioact. Mater. 1 (2016) 93–108. [26] K. Bleek, A. Taubert, New developments in polymer-controlled, bioinspired calcium phosphate mineralization from aqueous solution, Acta Biomater. 9 (2013) 6283–6321. [27] Q. Cai, Q. Xu, Q. Feng, X. Cao, X. Yang, X. Deng, Biomineralization of electrospun poly(L-lactic acid)/gelatin composite fibrous scaffold by using a supersaturated simulated body fluid with continuous CO2 bubbling, Appl. Surf. Sci. 257 (2011) 10109–10118. [28] K. Iijima, A. Sakai, A. Komori, Y. Sakamoto, H. Matsuno, T. Serizawa, M. Hashizume, Control of biomimetic hydroxyapatite deposition on polymer substrates using different protein adsorption abilities, Colloids Surf. B: Biointerfaces 130 (2015) 77–83. [29] X. Yang, B. Xie, L. Wang, Y. Qin, Z.J. Henneman, G.H. Nancollas, Influence of magnesium ions and amino acids on the nucleation and growth of hydroxyapatite, Crys. Eng. Comm. 13 (2011) 1153–1158. [30] B. Palazzo, D. Walsh, M. Iafisco, E. Foresti, L. Bertinetti, G. Marta, C.L. Bianchi, G. Cappelletti, N. Roveri, Amino acid synergetic effect on structure, morphology and surface properties of biomimetic apatite nanocrystals, Acta Biomater. 5 (2009) 1241–1252. [31] K. Wang, Y. Leng, X. Lu, F. Ren, Calcium phosphate bioceramics induce mineralization modulated by proteins, Mater. Sci. Eng. C 33 (2013) 3245–3255. [32] S. Maeno, Y. Niki, H. Matsumoto, H. Morioka, T. Yatabe, A. Funayama, Y. Toyama, T. Taguchi, J. Tanaka, The effect of calcium ion concentration on osteoblast viability, proliferation and differentiation in monolayer and 3D culture, Biomaterials 26 (2005) 4847–4855. [33] X. Zhang, S. Meng, Y. Huang, M. Xu, Y. He, H. Lin, J. Han, Y. Chai, Y. Wei, X. Deng, Electrospun gelatin/-TCP composite nanofibers enhance osteogenic differentiation of BMSCs and in vivo bone formation by activating Ca2+-sensing receptor signaling, Stem Cells Int. 2015 (2015) 507154. [34] J. Ye, W. Ai, F. Zhang, X. Zhu, G. Shu, L. Wang, P. Gao, Q. Xi, Y. Zhang, Q. Jiang, S. Wang, Enhanced proliferation of porcine bone marrow mesenchymal stem cells induced by extracellular calcium is associated with the activation of the calciumsensing receptor and ERK signaling pathway, Stem Cells Int. 2016 (2016) 6570671.
[2] M. Bohner, J. Lemaitre, Can bioactivity be tested in vitro with SBF solution? Biomaterials 30 (2009) 2175–2179. [3] T. Kokubo, H. Takadama, How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 27 (2006) 2907–2915. [4] A.A. Zadpoor, Relationship between in vitro apatite-forming ability measured using simulated body fluid and in vivo bioactivity of biomaterials, Mater. Sci. Eng. C 35 (2014) 134–143. [5] T. Kokubo, S. Yamaguchi, Novel bioactive materials developed by simulated body fluid evaluation: surface-modified Ti metal and its alloys, Acta Biomater. 44 (2016) 16–30. [6] X. Lu, Y. Leng, Theoretical analysis of calcium phosphate precipitation in simulated body fluid, Biomaterials 26 (2005) 1097–1108. [7] X. Chen, Y. Li, P.D. Hodgson, C. Wen, Microstructures and bond strengths of the calcium phosphate coatings formed on titanium from different simulated body fluids, Mater. Sci. Eng. C 29 (2009) 165–171. [8] T. Kokubo, Bioactive glass ceramics: properties and applications, Biomaterials 12 (1991) 155–163. [9] H.J. Park, S.J. Yu, K. Yang, Y. Jin, A.N. Cho, J. Kim, B. Lee, H.S. Yang, S.G. Im, S.W. Cho, Paper-based bioactive scaffolds for stem cell-mediated bone tissue engineering, Biomaterials 35 (2014) 9811–9823. [10] Q. Yao, J.G. Cosme, T. Xu, J.M. Miszuk, P.H. Picciani, H. Fong, H. Sun, Three dimensional electrospun PCL/PLA blend nanofibrous scaffolds with significantly improved stem cells osteogenic differentiation and cranial bone formation, Biomaterials 115 (2017) 115–127. [11] A.C. Tas, The use of physiological solutions or media in calcium phosphate synthesis and processing, Acta Biomater. 10 (2014) 1771–1792. [12] J.T. Lee, Y. Leng, K.L. Chow, F. Ren, X. Ge, K. Wang, X. Lu, Cell culture medium as an alternative to conventional simulated body fluid, Acta Biomater. 7 (2011) 2615–2622. [13] R.R. Rao, A. Jiao, D.H. Kohn, J.P. Stegemann, Exogenous mineralization of cellseeded and unseeded collagen-chitosan hydrogels using modified culture medium, Acta Biomater. 8 (2012) 1560–1565. [14] Ł. John, M. Bałtrukiewicz, P. Sobota, R. Brykner, Ł. Cwynar-Zając, P. Dzięgiel, Noncytotoxic organic-inorganic hybrid bioscaffolds: an efficient bedding for rapid growth of bone-like apatite and cell proliferation, Mater. Sci. Eng. C 32 (2012) 1849–1858. [15] V. Wagener, S. Virtanen, Protective layer formation on magnesium in cell culture medium, Mater. Sci. Eng. C 63 (2016) 341–351. [16] P. Liu, J. Tao, Y. Cai, H. Pan, X. Xu, R. Tang, Role of fetal bovine serum in theprevention of calcification in biological fluids, J. Cryst. Growth 310 (2008) 4672–4675. [17] Y. Guo, J. Lan, C. Zhang, M. Cao, Q. Cai, X. Yang, Mineralization on polylactide/ gelatin composite nanofibers using simulated body fluid containing amino acid, Appl. Surf. Sci. 349 (2015) 538–548. [18] D.C. Bassett, L.M. Grover, F.A. Müller, M.D. McKee, J.E. Barralet, Serum protein controlled nanoparticle synthesis, Adv. Funct. Mater. 21 (2011) 2968–2977. [19] S. An, J. Ling, Y. Gao, Y. Xiao, Effects of varied ionic calcium and phosphate on the proliferation, osteogenic differentiation and mineralization of human periodontal ligament cells in vitro, J. Periodontal Res. 47 (2012) 374–382.
873