Bioactive glass ions induce efficient osteogenic differentiation of human adipose stem cells encapsulated in gellan gum and collagen type I hydrogels

1 Bioactive glass ions induce efficient osteogenic differentiation of human adipose stem cells encapsulated in gellan gum and collagen type I hydrogels Kaisa Vuornosa,b,*, Miina Ojansivua,b, Janne T Koivistoc,d, Heikki Häkkänene, Birhanu Belayf, Toni Montonenf, Heini Huhtalag, Minna Kääriäinenh, Leena Hupai, Minna Kellomäkic, Jari Hyttinenf, Janne A Ihalainene, Susanna Miettinena,b aAdult Stem Cell Group, BioMediTech, Faculty of Medicine and Health Technology , Tampere University, Tampere, Finland. P.O. BOX 100, FI-33014 Tampere University, Finland, [email protected], [email protected], [email protected]. bResearch, Development and Innovation Centre, Tampere University Hospital, Tampere, Finland. P.O. BOX 2000, FI-33521 Tampere, Finland. cBiomaterials and Tissue Engineering Group, BioMediTech, Faculty of Medicine and Health Technology , Tampere University, Tampere, Finland. P.O. BOX 527, FI-33101 Tampere, Finland,[email protected], [email protected]. dHeart Group, BioMediTech, Faculty of Medicine and Health Technology , Tampere University, Tampere, Finland. P.O. BOX 100, FI-33014 Tampere University, Finland. eNanoscience Center, University of Jyväskylä, Jyväskylä, Finland. P.O. BOX 35, FI-40014 University of Jyväskylä, Finland, [email protected], [email protected]. fComputational Biophysics and Imaging Group, BioMediTech, Faculty of Medicine and Health Technology , Tampere University, Tampere, Finland. P.O. BOX 527, FI-33101 Tampere, Finland, [email protected], [email protected], jari.hy[email protected]. gFaculty of Social Sciences, Tampere University, Tampere, Finland. P.O. BOX 100, FI-33014 Tampere University, Finland, [email protected]. This is the accepted manuscript of the article, which has been published in Materials Science and Engineering: C. 2019, 99, 905-918. https://doi.org/10.1016/j.msec.2019.02.035 2 hDepartment of Plastic and Reconstructive Surgery, Tampere University Hospital, P.O. BOX 2000, FI-33521 Tampere, Finland, [email protected] iJohan Gadolin Process Chemistry Centre, Åbo Akademi University, Åbo, Finland. Biskopsgatan 8, FI-20500 Åbo, Finland, [email protected]. *Corresponding author. Address: Tampere University, BioMediTech, Faculty of Medicine and Health Technology, Adult Stem Cell Group, P.O. BOX 100, FI-33014 Tampere University, Finland; Tel.: +358 40 190 1789; Fax: +358 3 3551 8498; E-mail: [email protected] (K. Vuornos). 3 Abstract Background: Due to unmet need for bone augmentation, our aim was to promote osteogenic differentiation of human adipose stem cells (hASCs) encapsulated in gellan gum (GG) or collagen type I (COL) hydrogels with bioactive glass (experimental glass 2-06 of composition [wt-%]: Na2O 12.1, K2O 14.0, CaO 19.8, P2O5 2.5, B2O3 1.6, SiO2 50.0) extract based osteogenic medium (BaG OM) for bone construct development. GG hydrogels were crosslinked with spermidine (GG-SPD) or BaG extract (GG-BaG). Methods: Mechanical properties of cell-free GG-SPD, GG-BaG, and COL hydrogels were tested in osteogenic medium (OM) or BaG OM at 0, 14, and 21d. Hydrogel embedded hASCs were cultured in OM or BaG OM for 3, 14, and 21d, and analyzed for viability, cell number, osteogenic gene expression, osteocalcin production, and mineralization. Hydroxyapatite-stained GG-SPD samples were imaged with Optical Projection Tomography (OPT) and Selective Plane Illumination Microscopy (SPIM) in OM and BaG OM at 21d. Furthermore, Raman spectroscopy was used to study the calcium phosphate (CaP) content of hASC-secreted ECM in GG-SPD, GG-BaG, and COL at 21d in BaG OM. Results: The results showed viable rounded cells in GG whereas hASCs were elongated in COL. Importantly, BaG OM induced significantly higher cell number and higher osteogenic gene expression in COL. In both hydrogels, BaG OM induced strong mineralization confirmed as CaP by Raman spectroscopy and significantly improved mechanical properties. GG-BaG hydrogels rescued hASC mineralization in OM. OPT and SPIM showed homogeneous 3D cell distribution with strong mineralization in BaG OM. Also, strong osteocalcin production was visible in COL. Conclusions: Overall, we showed efficacious osteogenesis of hASCs in 3D hydrogels with BaG OM with potential for bone-like grafts. 4 Keywords: adipose stem cell, bioactive glass, osteogenic differentiation, gellan gum hydrogel, collagen type I hydrogel 1. Introduction Increasing number of musculoskeletal defects and the growth of the ageing population augment the demand for functional engineered bone grafts. Autologous bone is limited while, besides shortage of allograft bone, it might also pose a risk of adverse effects and graft rejection [1, 2]. Human adipose stem cells (hASCs) are abundant and accessible adult stem cells and suitable for the development of bone constructs [3]. For bone applications, 3D hydrogels offer an adaptable approach of a free form construct with high elasticity and malleable mechanical properties, instead of limited conventional scaffold structure. 3D hydrogel culture simulates more effectively the natural elastic cell microenvironment allowing higher degrees of freedom to form cellular interactions compared to traditional stiffer biomaterials. For instance, natural polymers have been studied as native microenvironments for stem cells, such as protein based collagen type I (COL), while inexpensive polysaccharide gellan gum (GG) offers more tailorable mechanical properties to support stem cell differentiation. Moreover, GG and COL hydrogels have been already reported as hydrophilic, biocompatible, bioresorbable, and also, adaptable hydrogel scaffolds suitable for bone tissue engineering applications [4-9]. On the other hand, bioactive glasses (BaGs) have been widely applied as medical implant materials and have been shown well applicable for bone grafts [10-12] and, in addition, as strong osteogenic inducers of the hASCs without any added chemical supplements such as growth factors [13, 14]. Despite their suitability for bone applications, hydrogel biomaterials in themselves lack bone mineralization enhancing components. To date, the majority of the GG and COL hydrogel studies for bone applications have sought to increase hydrogel mineralization and mechanical properties by the addition of nanosized or larger BaG particles to the hydrogel matrix [4, 7, 15, 16]. Hydrogel mineralization requires robust support from added components in in vitro conditions mimicking the 5 physiological conditions. To achieve this, in an earlier study, we demonstrated that BaG dissolution ions were strong inducers of hASC osteogenic differentiation, and showed that the ions dissolved from the specific composition of experimental silica-based BaG 2-06 combined with osteogenic medium (OM) components induced calcium phosphate (CaP) mineral accumulation already after 14 days of culture in 2D culture conditions [13]. The ionic dissolution products of experimental glass 2-06 have been previously analyzed high in Ca2+, K+, and B+, and the detailed ionic composition of the bioactive glass extract (BaG ext) of experimental glass 2-06 has been reported in a published study [13]. Thus, we hypothesized that the BaG ext ionic dissolution products would promote equally strong osteogenic differentiation of hASCs in 3D hydrogel culture. In addition, since small molecules like cationic spermidine (SPD) have been demonstrated to interact with anionic polymers such as GG [17, 18], we hypothesized the divalent Ca2+ cations in the BaG ext to function as potential ionic crosslinkers for GG hydrogel. To the best of our knowledge, GG and COL have not been previously combined with BaG ionic species alone for bone tissue engineering applications with embedded hASCs. We also used efficient imaging techniques to assess the cell distribution, form and mineralization with in-house-built Optical Projection Tomography (OPT) and Selective Plane Illumination Microscopy (SPIM) systems that have been already applied to cell imaging in 3D hydrogel culture, mass transport studies, and characterization of 3D hydrogels [19-21]. Therefore, the aim of this study was to develop an effective hASC-laden hydrogel mineralization method for the development of engineered bone constructs. The mechanical properties of cell-free hydrogels were tested for bone applications with or without incubation in serum-containing OM and BaG ext based osteogenic medium (BaG OM) media. GG and COL hydrogels combined with BaG OM induction were compared for osteogenic differentiation of hydrogel-encapsulated hASCs, and to that end hASC viability, adhesion, cell number, osteocalcin production by immunofluorescence staining, mineralization, and the gene expression of osteogenic marker genes were analyzed. Additionally, the potential of GG ionic crosslinkers SPD and BaG ext to support hASC 6 mineralization was tested in the control OM. The transparent GG-SPD cell-hydrogel sample mineralized residues were imaged with OPT and SPIM in OM and BaG OM at 3 weeks. Also, the Raman spectra of hASC-secreted ECM and mineralized residues in GG-SPD, GG-BaG, and COL was measured at 3 weeks of culture in the BaG OM condition. 2. Materials and methods 2.1 Hydrogel scaffolds 2.1.1 Gellan gum hydrogel scaffolds. GG (low acyl, Mw 1.0 kg/mol; Gelzan CM; Sigma-Aldrich, St. Louis, MO, USA) sterile filtered (0.2 µm) solution of 0.5 % (w/v) concentration was crosslinked with either 16 % (v/v) SPD (BioXtra; Sigma-Aldrich) of 1 mg/mL concentration in 10 % (w/w) sucrose (Sigma-Aldrich) in deionized water or with BaG ext (experimental glass 2-06) [13] containing Dulbecco’s Modified Eagle Medium/Ham’s Nutrient Mixture F-12 (DMEM/F-12 1:1; Thermo Fisher Scientific, Waltham, MA, USA), 1 % L-glutamine (GlutaMAX; Thermo Fisher Scientific), and 1 % antibiotics/antimycotic containing 100 U/mL penicillin/100 U/mL streptomycin (P/S; Lonza, Basel, Switzerland) to yield GG-BaG hydrogel samples. The cell pellet was resuspended into GG solution at +37°C for hASC encapsulation into 3D hydrogel. The cellhydrogel solution was manually mixed to an ionic crosslinker for immediate gelation. The cell culture medium was added on top of the gelated hASC-laden hydrogel samples. 2.1.2 Collagen type I hydrogel scaffolds. Commercially available COL (rat tail collagen type I, 3.0 mg/mL, Gibco; Thermo Fisher Scientific) was gelated with 10x phosphate buffered saline (10x PBS; Lonza) and 1 N NaOH (Sigma-Aldrich) according to the manufacturer’s protocol. Briefly, the ice cold 10x PBS and 1 N NaOH were mixed and COL stock was added while kept on ice. The cell pellet was resuspended and mixed into the non-gelated COL mixture followed by immediate gelation in RT and the cell culture medium was added on top of the gelated hydrogels. 7 2.2 Mechanical testing The acellular GG-SPD, GG-BaG, and COL hydrogel samples were mechanically tested by compression testing. The 0.875 cm3 samples (n=3–6) of approximately 4.0–6.5 mm of height and 12.2 mm diameter were cast in custom-made molds and incubated overnight at +37°C under parafilm without media to ensure complete hydrogel gelation before compression testing or media incubation initiation. Unconfined compression was performed with a constant 10 mm/min strain rate in air environment in RT to 65 % strain from their original height at 0, 14, and 21 days of incubation in OM or BaG OM media. The compressive load was measured by the Bose 5100 BioDynamic ElectroForce (TA Instruments, New Castle, DE, USA) instrument equipped with a 225 N load sensor and the data was recorded with the WinTest 4.1 software (WinTest, Yokohama, Japan). The compressive modulus under tension was calculated by MS Excel (Microsoft, Redmond, WA, USA) based on the slope of the linear region of the stress versus strain curve where the Hooke’s law holds [17]. 2.3 Adipose stem cell isolation and cell expansion The hASCs were obtained from subcutaneous adipose tissue of six healthy female donors of 52±5 years in surgeries at the Tampere University Hospital Department of Plastic Surgery between 2014– 2015 with the patients’ written informed consent, in accordance with the Ethics Committee of the Pirkanmaa Hospital District’s, Tampere, Finland, ethical approval (R15161). The hASCs were isolated as reported previously [22]. Briefly, the adipose tissue was cut up and tissue was digested by collagenase type I (1.5 mg/mL; Thermo Fisher Scientific) in maintenance medium containing DMEM/F-12 1:1 (Thermo Fisher Scientific), 5 % human serum (Biowest, Nuaillé, France), 1 % Lglutamine (GlutaMAX; Thermo Fisher Scientific), and 1 % P/S. The cells were expanded in maintenance medium . Flow cytometry analysis was performed, and the overall results of the cell surface marker flow cytometry analysis (see Supplementary file 1) indicated the mesenchymal 8 origin of the cells in accordance with literature [3, 23, 24]. The isolated hASCs were tested and reported negative for mycoplasma contamination. 2.4 Osteogenic induction and cell culture Osteogenic induction was initiated immediately after plating by adding 0.3 cm3 either OM optimized for hASC osteogenic differentiation [22] or BaG OM to the hydrogel encapsulated cell constructs. The cells were plated at a density of 950,000 cells/cm3, encapsulated in respective hydrogels which were cast in 48-well plate wells (Nunclon; Sigma-Aldrich) in a volume of 0.2 cm3. The BaG ext was prepared as reported previously from bioactive glass 2-06 (wt-%: Na2O 12.1; K2O 14.0; CaO 19.8; P2O5 2.5; B2O3 1.6; SiO2 50.0) [13]. Briefly, 87.5 mg/mL of BaG granules (500– 1000 µm) of bioactive glass 2-06 were disinfected with 70 % ethanol washes for 10 min repeated twice and air dried in RT for 2 h followed by incubation for 24 h at +37°C to dissolve ions into the maintenance medium without human serum. After incubation, 5 % human serum (Biowest) was added to the sterile filtered (0.2 µm) BaG ext. The BaG ext was prepared fresh each 14 days to avoid any risk of precipitates. For the Raman spectroscopy analyses, the cells were cultured in phenol red free media to avoid fluorescence interference with the Raman spectra [13]. The control cell cultures were maintained in OM. Different media compositions are listed in Table 1. During the experiments, medium was changed every other day. The experiments were carried out at hASC passage 3–6. 9 Table 1. Composition of media. Medium Composition Maintenance medium (MM) DMEM/F-12 1:1, 5 % HS, 1 % L-glutamine, 1 % P/S Osteogenic medium (OM) 5 nM Dex, 250 µM AsA2P, 10 mM β-GP in MM BaG osteogenic medium (BaG OM) OM in BaG ext base Raman spectroscopy cell culture medium Phenol red free DMEM/F-12 1:1 including Lglutamine, 5 % HS, 1 % P/S Raman spectroscopy phenol red free BaG OM OM in phenol red free BaG ext base MM, maintenance medium; OM, osteogenic medium; BaG OM, bioactive glass extract osteogenic medium; HS, human serum (Biowest); P/S, 100 U/mL penicillin/streptomycin (Lonza); Dex, dexamethasone (Sigma-Aldrich); AsA2P, L-ascorbic acid 2-phosphate (SigmaAldrich); β-GP, beta-glycerophosphate (Sigma-Aldrich); L-glutamine (Thermo Fisher Scientific). 2.5 Cell viability and cell number Cell viability was analyzed with Live/Dead fluorescence staining (Thermo Fisher Scientific) at 3 and 14 days, as described previously [25]. The living cells were stained with 0.5 mM calcein acetoxymethyl ester (green stain) and necrotic cells were stained with 0.25 mM ethidium homodimer-1 (red stain) for 45 min in RT. The samples were imaged using an epifluorescence Olympus IX51 microscope and Olympus DP30BW digital camera (Olympus, Tokyo, Japan). Cell number was measured based on the total amount of DNA with CyQUANT Cell Proliferation Assay Kit (Thermo Fisher Scientific) according to the manufacturer’s protocol at 14 and 21 days. Briefly, the cells were lysed with 0.1 % Triton X-100 buffer (Sigma-Aldrich) and the hydrogel samples were homogenized mechanically by the Ultra-Turrax tissue homogenizer (IKA Labortechnik, Staufen, Germany). The lysed samples were stored at –80°C until analysis after a freeze-thaw cycle. A working solution was prepared with the kit provided CyQUANT GR dye and 16 OM incubation alone increased COL compressive modulus significantly at 21 days. Additionally, at 21 days, the GG-SPD (BaG OM) samples had the highest compressive moduli measured at approximately 40.0 kPa and significantly higher compared to GG-SPD (OM) and COL (BaG OM). For both GG-SPD and COL hydrogels, the BaG OM condition had significantly higher compressive modulus compared to the OM condition at the 21-day time point. Overall, the COL as well as GGBaG samples showed ductile behavior and plastic deformation properties without a fracture point, whereas by default, the GG-SPD samples had higher resistance to deformation and more brittle behavior which indicated a stiffer structure, and also had a clear fracture point in the stress-strain curve (See Supplementary file 3). 3.2 Adipose stem cell viability maintained and cell number increased in 3D hydrogels The viability of hASCs was analyzed at 3 and 14 days of 3D hydrogel culture (Fig. 2, see Supplementary file 4). The cells remained well viable during culture. The COL (OM) samples contracted after 4 days and were therefore unavailable for a later time point. 17 Fig. 2 Cell viability. Viability of hASCs encapsulated in GG-SPD (OM), GG-SPD (BaG OM), COL (OM) and COL (BaG OM) 3D hydrogels at 3 and 14 days. The COL (OM) samples 18 contracted after 3 days. Representative images with hASCs from 1 donor (n=1). Scale bar 200 µm. Figure in color in print. 2-column fitting figure. The cell number based on total DNA content was analyzed at 14and 21-day time points (Fig. 3). Fig. 3 Cell number. Cell number based on total DNA content of hASCs encapsulated in 3D hydrogels at 14 and 21 days. The analyzed hASCs were from 3 donors (n=9). Data are presented as mean + SD. Significant difference with p < 0.05. a The combined COL (OM) and COL (BaG OM) cell numbers are significantly higher compared to the combined GG-SPD (OM) and GG-SPD (BaG OM) cell numbers at 14 days; b The combined 14and 21-day COL (BaG OM) cell numbers are significantly higher compared to the combined 14and 21-day GG-SPD (BaG OM) cell numbers. 1-column fitting figure. The cell number increased the most for the BaG OM cultured samples with the significantly highest cell number for the COL (BaG OM) at 14 days. Also, at 14 days, the COL hydrogels in OM and BaG OM combined together had a significantly higher cell number compared to the combined GGSPD hydrogels in OM and BaG OM. Further, the combined cell number results at 14 and 21 days of 19 the COL (BaG OM) condition showed a significantly higher cell number compared to the combined 14and 21-day GG-SPD (BaG OM) cell numbers. 3.3 Higher expression of osteogenic marker genes of adipose stem cells in collagen type I hydrogel scaffolds The gene expression of osteogenic marker genes was analyzed by qRT-PCR at 14 and 21 days (Fig. 4). 20 Fig. 4 Gene expression. Gene expression of osteogenic marker genes of hASCs in GG-SPD (OM), GG-SPD (BaG OM), COL (OM), and COL (BaG OM) 3D hydrogels at 14 and 21 days. Significant difference with p < 0.05. (A) DLX5 gene expression significantly higher in a COL (BaG OM) compared to GG-SPD (OM) at 21 days; (B) OSX gene expression significantly higher for b the combined COL (OM) and COL (BaG OM) sample expression at 14 and 21 days compared to the combined GG-SPD (OM) and GG-SPD (BaG OM) sample expression at 14 and 21 days; (C) RUNX2 gene expression significantly higher for c the combined COL (OM) and COL (BaG OM) sample expression at 21 days compared to the combined GG-SPD (OM) and GG-SPD (BaG OM) sample expression at 21 days; (D) ALPL gene expression. The results were relativised to the control condition of GG-SPD (OM) at 14 days. The relative expression of DLX5 and OSX had high variance and are presented partly in Log(10) scale. The hASCs were isolated from 3 donors (n=6). Group medians are indicated with a horizontal line. 1.5-column fitting figure. All the osteogenic marker genes were statistically significantly higher in gene expression in COL in OM or in BaG OM compared to GG-SPD at 3 weeks, excluding ALPL expression. Due to the nonGaussian distribution, the statistical analyses of the gene expression medians was conducted with the Mann-Whitney test. The DLX5 expression was significantly higher for the COL (BaG OM) compared to GG-SPD (OM) at 21 days. The OSX expression was significantly higher for the combined COL (BaG OM) results at 14 and 21 days compared to the combined results for GG-SPD (OM) at 14 and 21 days. The gene expression of RUNX2 of the combined COL in OM and BaG OM samples at 21 days was statistically significantly higher than that of the combined GG-SPD samples in OM and BaG OM at the 21-day time point. 21 3.4 Strong immunocytochemical osteocalcin staining of adipose stem cells in collagen type I hydrogel scaffolds Immunocytochemical staining results of OC and DAPI stained hASCs encapsulated in GG and COL hydrogels in different media conditions at 21 days are shown in Fig. 5 (see Supplementary file 5). Fig. 5 Osteocalcin immunofluorescence staining. Representative images of osteocalcin and DAPI stained hASCs in GG-SPD (OM), GG-SPD (BaG OM), COL (OM), and COL (BaG OM) 3D hydrogels at 21 days. The blank controls are shown in lower left corner. The hASCs were isolated from 2 donors (n=2). Scale bar 200 µm. Figure in color in print. 2-column fitting figure. Strong OC immunofluorescence staining was detected in the COL samples in both media conditions, although some OC staining in GG-SPD was visible. 22 3.5 Efficient mineralization of adipose stem cells in 3D hydrogel scaffolds The hASC secreted mineralized matrix of hydroxyapatite residues were stained with the OsteoImage assay and hASC nuclei with DAPI and imaged in 3D hydrogels at 21 days (Fig. 6A). The hydroxyapatite content was also quantified with the OsteoImage assay analysis (Fig. 6B). 23 Fig. 6 Mineralization. (A) Representative images of hydroxyapatite residues and cell nuclei of hASCs stained with the OsteoImage assay and DAPI in 3D hydrogels at 21 days in GG-SPD (OM), GG-SPD (BaG OM), and COL (BaG OM). The COL (OM) samples contracted before 21 days. The blank controls are presented in the lower left corner. The hASCs were isolated from 1 donor (n=1). Scale bar 200 µm; (B) The measured OsteoImage hydroxyapatite mineralization fluorescence count of hASCs in 3D hydrogel at 21 days. Data are presented as mean + SD. a Significant difference from GG-SPD (OM) with p < 0.05. The hASCs were isolated from 3 donors and 2 parallel samples of each condition were tested, out of which for GG-SPD (OM) and (BaG OM) groups a total of 5 samples (n=5), and for COL (BaG OM) 4 samples (n=4) were available for analysis at 21-day time point. Figure in color in print. 2-column fitting figure. The OsteoImage stained hydroxyapatite residue imaging supported the quantified mineralization results with the strongest hydroxyapatite staining for the COL (BaG OM), while also strong staining was seen in the GG-SPD (BaG OM) condition, and where moderate staining was visible in the GGSPD (OM) at the 21-day time point. 3.6 Optical Projection Tomography and Selective Plane Illumination Microscopy for high resolution 3D hydrogel scaffold imaging OPT and SPIM images were acquired for brightfield transmission and fluorescence emission modes, respectively (Fig. 7). Optical opacity of the COL hydrogel prevented its OPT and SPIM imaging. 24 Fig. 7 OPT and SPIM 3D imaging of hASC-hydrogel constructs. Representative label-free brightfield OPT 3D reconstructed images of hASCs encapsulated in (A) GG-SPD (OM); (C) GGSPD (BaG OM); and OsteoImage hydroxyapatite stained fluorescence SPIM 3D images of hASCs in 3D hydrogels in (B) GG-SPD (OM); (D) GG-SPD (BaG OM) at 21 days. The hASCs were isolated from 1 donor (n=1). Scale bar 100 µm. Figure in color in print. 2-column fitting figure. 25 For the brightfield OPT 3D reconstructed images, the variation in light attenuation between the cells and the hydrogel structure allowed to visualize the distribution of the cells in each projection image (Fig. 7A; 7C; see Supplementary file 9; 10; 3D reconstruction videos). The cells and the surrounding shaded mineralized ECM were visible in transmission mode brightfield images, and thus offered proof of concept that label-free OPT can be applied to 3D hydrogel cell culture mineralization studies. The 3D stack of multi-focal fluorescence SPIM imaging showed the fluorescent-labeled mineralization clearly with strong hydroxyapatite stain in the GG-SPD (BaG OM) sample (Fig. 7B; 7D; see Supplementary file 11; 12; 3D reconstruction videos) whereas the fluorescent-stained hydroxyapatite was scarce in the OM condition (see Supplementary file 13; 14; 3D reconstruction video). Any closer inspection of cell morphology was restricted by the optically dense mineralization. 3.7 Raman spectroscopic analysis verified mineralized hydroxyapatite residues and gellan gum mineralization improved with bioactive glass extract ionic crosslinking The Raman spectra of hASCs in GG-SPD (BaG OM), GG-BaG (BaG OM) and COL (BaG OM) were measured at 21 days (Fig. 8). 32 mineralization processes [5, 7], the mineral deposition was clearly accelerated in interactions with the cell-secreted matrix in our present results. Moreover, these results verified that the GG-SPD hydrogels allowed the steady unobstructed diffusion [20] of the ionic species supporting hASC osteogenic differentiation and homogeneous mineral deposition. What is more, the cell-containing samples were also positively stained for the late osteogenic marker OC, which was shown to increase together with the quantitatively measured and imaged mineralized content thus confirming hASC osteogenic differentiation towards bone-like cells in 3D hydrogel culture. All in all, the strong OC and hydroxyapatite staining in the COL hydrogel, in addition to the highest Raman spectroscopic measurement result indicated that COL hydrogel combined with the BaG OM as the most efficient osteogenic inducer of the hydrogel encapsulated hASCs. In the future and especially for in vivo studies for larger critical sized bone defects, however, cell survival, diffusion of nutrients, and removal of waste products also in the central parts of the construct should be secured with an adequate vessel structure for vascularized bone-like graft development. Also, the effect of mineralization on collagen hydrogel in vivo degradation would require further investigation in the future. While COL hydrogel has been extensively tested for potential injectable bone tissue engineering applications [5, 49-51], we also tested an abundant and economic GG hydrogel for hASC osteogenic differentiation. On the whole, our results indicated that the BaG OM induction was required for efficient hASC osteogenic differentiation encapsulated in 3D GG-SPD and COL hydrogels. The BaG OM was chosen based on previous studies in our group that showed in 2D culture the efficiency of the bioactive glass 2-06 BaG ext for hASC osteogenic induction [13]. Importantly, the current results confirmed that the dissolved BaG ions with the OM supplements together with evolving hydrogel mechanical properties were efficient to induce hASC osteogenic differentiation. Further, we hypothesized that crosslinking GG hydrogel with the BaG ext would enhance hASC mineralization encapsulated in 3D GG, based on the rich Ca2+ content of the BaG 33 ext and since GG crosslinking requires small cations [52]. Also, Ca2+ ions are implicated in the ECM mineralization process [46]. Indeed, the poor performance of cell-loaded GG hydrogel samples in control OM was rescued by Ca2+ crosslinking with the BaG ext. What is more, the mineralization results suggested that a separate cationic crosslinker might be omitted when using the novel BaG ext for GG gelation, where a statistically significantly higher hydroxyapatite content of the hASC-laden GG-BaG hydrogels was compared to GG-SPD samples in control OM at 3 weeks. Further studies are still required to determine the potential of the BaG ext hydrogel crosslinking for the hASC mineralization process in more detail. Overall, BaG OM induced efficient osteogenic differentiation and strong mineralization of hASCs in 3D hydrogels thus supporting further the osteoblast-like cell maturation combined with enhanced mechanical properties of the mineralized matrix in both the GG and COL hydrogels. Therefore, these novel results combining osteoinducing BaG ions and 3D hydrogel stem cell culture have considerable potential for the development of a wide variety of applications for bone tissue engineered constructs. 5. Conclusions In this study, we studied enhanced hASC osteogenic induction with BaG ionic dissolution products in GG-SPD and COL 3D hydrogel culture in BaG OM compared to regular OM. Incubation in BaG OM significantly reinforced GG and COL hydrogel mechanical properties and showed a stiffening behavior similar to an evolving bone matrix. In both media conditions, the hASCs were well viable embedded within 3D hydrogel, where the GG-SPD encapsulated hASCs had a tight and round cell morphology, whereas in COL hydrogel elongated and spread morphologies were observed. With the BaG OM induction, the hASCs in COL hydrogel showed significantly higher osteogenic marker gene expression. On the whole, the BaG OM culture significantly enhanced hASC potency to mineralize in both the GG-SPD and COL 3D hydrogels, however, hASC-laden COL (BaG OM) hydrogels showed highest mineralization and hydroxyapatite content confirmed by Raman 34 spectroscopy analysis together with the strongest OC staining results. Importantly, the BaG extcrosslinked GG-BaG hydrogels promoted significantly higher hASC mineralization even in the control OM. The OPT and SPIM techniques were evaluated as efficient methods for emerging 3D hydrogel cell culture imaging and analysis applications. These results demonstrated the significant potential of BaG OM induction and novel 3D hydrogel culture methods for the osteogenic differentiation of hASCs towards bone-like cells and for bone regeneration applications. Competing interests The authors declare that they have no competing interests. Funding This study was financially supported by the Finnish Funding Agency for Innovation (TEKES Business Finland), the Human Spare Parts Project, the Academy of Finland, The City of Tampere Science Fund, The Finnish Cultural Foundation's Pirkanmaa Regional Fund, The Finnish Foundation for Technology Promotion, and the Competitive State Research Financing of the Expert Responsibility area of Tampere University Hospital. Authors' contributions KV contributed to planning of the work, carried out most of the data collection and analysis, drafted and edited the manuscript. MO contributed to planning of the work and participated in manuscript preparation and editing. JK, HHä, BB, and TM participated in data collection and analysis and manuscript preparation and editing. HHu was consulted for statistical analyses, contributed to the data analysis and participated in manuscript preparation and editing. LH, MK, JH, and JI participated in the planning of the work and manuscript preparation and editing. SM contributed to planning and coordination of the work, interpretation of data, and participated in manuscript preparation and editing. 35 Acknowledgements The authors wish to thank Mari Lehti-Polojärvi, MSc, for the OPT and SPIM imaging cell culture platforms, Computational Biophysics and Imaging Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland, , the Tampere Imaging Facility (TIF), Tampere University, for their support in imaging services, and Jenny Parraga Meneses, PhD, Biomaterials and Tissue Engineering Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Finland , and Nick Walters, PhD, Adult Stem Cell Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland , and for technical support Adult Stem Cell Group laboratory technicians Miia Juntunen, MSc, Anna-Maija Honkala, laboratory engineer, and Sari Kalliokoski. Supplementary material Supplementary file 1: Flow cytometric analysis and cell characterization.docx. Flow cytometric surface marker expression analysis and cell characterization results. Supplementary file 2: Raman spectra of acellular hydrogels.docx. Raman spectra of acellular blank and acellular hydroxyapatite doped hydrogel constructs. Supplementary file 3: Representative stress-strain curves.docx. Compressive modulus mechanical testing of cell-free GG-SPD, GG-BaG, and COL hydrogels after 0-, 14-, and 21-day incubation in OM and BaG OM media. Supplementary file 4: Cell viability.docx. Viability of hASCs encapsulated in GG-SPD (OM), GG-SPD (BaG OM), COL (OM) and COL (BaG OM) 3D hydrogels at 3 and 14 days. Supplementary file 5: Immunocytochemical staining of osteocalcin.docx. Osteocalcin production and nuclei of hASCs encapsulated in GG-SPD (OM), GG-SPD (BaG OM), COL (OM) and COL (BaG OM) 3D hydrogels at 21 days. 36 Supplementary file 6: Cell viability crosslinker comparison.docx. Viability of hASCs encapsulated in GG-SPD (OM) and GG-BaG (OM) 3D hydrogels at 3, 14, and 21 days. Supplementary file 7: Cell number crosslinker comparison.docx. Cell number based on total DNA content of hASCs encapsulated in GG-SPD (OM) and GG-BaG (OM) 3D hydrogels at 14 and 21 days. Supplementary file 8: Mineralization crosslinker comparison.docx. OsteoImage hydroxyapatite and DAPI staining of hASCs encapsulated in GG-SPD (OM) and GG-BaG (OM) 3D hydrogels at 21 days. Supplementary file 9: Brightfield OPT imaging of GG-SPD in OM.avi. Brightfield OPT imaging 3D reconstruction of GG-SPD cell-hydrogel construct in OM condition at 21 days. Supplementary file 10: Brightfield OPT imaging of GG-SPD in BaG OM.avi. Brightfield OPT imaging 3D reconstruction of GG-SPD cell-hydrogel construct in BaG OM condition at 21 days. Supplementary file 11: Fluorescence SPIM imaging of GG-SPD in BaG OM.avi. Fluorescence SPIM imaging 3D reconstruction of GG-SPD cell-hydrogel construct with fluorescence stained hydroxyapatite in BaG OM condition at 21 days. Supplementary file 12: Combined OPT and SPIM imaging of GG-SPD in BaG OM.avi. Combined OPT and SPIM imaging 3D reconstruction of GG-SPD cell-hydrogel construct in BaG OM condition at 21 days. Supplementary file 13: Fluorescence SPIM imaging of GG-SPD in OM.avi. Fluorescence SPIM imaging 3D reconstruction of GG-SPD cell-hydrogel construct with fluorescence stained hydroxyapatite in OM condition at 21 days. 37 Supplementary file 14: Combined OPT and SPIM imaging of GG-SPD in OM.avi. Combined OPT and SPIM imaging 3D reconstruction of GG-SPD cell-hydrogel construct in OM condition at 21 days. Vitae Kaisa Vuornos, MSc, works as a doctoral student in the Adult Stem Cell Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland. Miina Ojansivu, PhD, works as a post doctoral researcher at the Karolinska Institutet, Solna, Sweden, in the Department of Medical Biochemistry and Biophysics. Janne T Koivisto, MSc, works as a doctoral student in the Biomaterials and Tissue Engineering Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland, and also works in the Heart Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland. 38 Heikki Häkkänen works as a Laboratory Engineer in the Spectroscopy for Detecting Dynamics of Biomolecules Group, Nanoscience Center, in the Department of Biological and Environmental Science in the University of Jyväskylä, Jyväskylä, Finland. Birhanu Belay, MSc, works as a doctoral student in the Computational Biophysics and Imaging Group, BioMediTech, at the Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland. 39 Toni Montonen, MSc, works as a doctoral student in the Computational Biophysics and Imaging Group, BioMediTech, at the Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland. Heini Huhtala, MSc, works as a University Instructor in the Faculty of Social Sciences, Tampere University, Tampere, Finland. Minna Kääriäinen, MD, PhD, works as a plastic surgeon and chief physician in the Department of Plastic and Reconstructive Surgery at the Tampere University Hospital, Tampere, Finland. Leena Hupa, DSc, is the Professor of Inorganic Chemistry at the Johan Gadolin Process Chemistry Centre, Åbo Akademi University, Åbo, Finland. 40 Professor Minna Kellomäki, DSc, is the Principal Investigator of the Biomaterials and Tissue Engineering Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Finland. Professor Jari Hyttinen, DSc, is the Principal Investigator in the Computational Biophysics and Imaging Group (www.tut.fi/cbig), BioMediTech, at the Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland. Professor Janne A Ihalainen, PhD, is the Principal Investigator in the Spectroscopy for Detecting Dynamics of Biomolecules Group, Nanoscience Center, and the Head of Department in the Department of Biological and Environmental Science in the University of Jyväskylä, Jyväskylä, Finland. 41 Susanna Miettinen, PhD, is an Associate Professor in the Faculty of Medicine and Health Technology in the Tampere University and the Principal Investigator in the Adult Stem Cell Group, BioMediTech, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland. References [1] A. Chatterjea, G. Meijer, C. van Blitterswijk, J. de Boer, Clinical application of human mesenchymal stromal cells for bone tissue engineering, Stem Cells Int., (2010) 215625. 10.4061/2010/215625 [doi]. [2] M. Jakob, F. Saxer, C. Scotti, S. Schreiner, P. Studer, A. Scherberich, M. Heberer, I. Martin, Perspective on the evolution of cell-based bone tissue engineering strategies, Eur.Surg.Res., 49 No. 1 (2012) 1-7. 10.1159/000338362 [doi]. 48 [45] J.A. Killion, S. Kehoe, L.M. Geever, D.M. Devine, E. Sheehan, D. Boyd, C.L. Higginbotham, Hydrogel/bioactive glass composites for bone regeneration applications: synthesis and characterisation, Mater.Sci.Eng.C.Mater.Biol.Appl., 33 No. 7 (2013) 4203-4212. 10.1016/j.msec.2013.06.013 [doi]. [46] H. Orimo, The mechanism of mineralization and the role of alkaline phosphatase in health and disease, J.Nippon Med.Sch., 77 No. 1 (2010) 4-12. JST.JSTAGE/jnms/77.4 [pii]. [47] A.J. Leite, B. Sarker, T. Zehnder, R. Silva, J.F. Mano, A.R. Boccaccini, Bioplotting of a bioactive alginate dialdehyde-gelatin composite hydrogel containing bioactive glass nanoparticles, Biofabrication, 8 No. 3 (2016) 035005-5090/8/3/035005. 10.1088/1758-5090/8/3/035005 [doi]. [48] J. Lewandowska-Lancucka, S. Fiejdasz, L. Rodzik, M. Koziel, M. Nowakowska, Bioactive hydrogel-nanosilica hybrid materials: a potential injectable scaffold for bone tissue engineering, Biomed.Mater., 10 No. 1 (2015) 015020-6041/10/1/015020. 10.1088/1748-6041/10/1/015020 [doi]. [49] B. Sarker, J. Hum, S.N. Nazhat, A.R. Boccaccini, Combining collagen and bioactive glasses for bone tissue engineering: a review, Adv.Healthc.Mater., 4 No. 2 (2015) 176-194. 10.1002/adhm.201400302 [doi]. [50] M.M. Villa, L. Wang, J. Huang, D.W. Rowe, M. Wei, Bone tissue engineering with a collagenhydroxyapatite scaffold and culture expanded bone marrow stromal cells, J.Biomed.Mater.Res.B.Appl.Biomater., 103 No. 2 (2015) 243-253. 10.1002/jbm.b.33225 [doi]. [51] T. Long, J. Yang, S.S. Shi, Y.P. Guo, Q.F. Ke, Z.A. Zhu, Fabrication of three-dimensional porous scaffold based on collagen fiber and bioglass for bone tissue engineering, J.Biomed.Mater.Res.B.Appl.Biomater., 103 No. 7 (2015) 1455-1464. 10.1002/jbm.b.33328 [doi]. [52] E. Miyoshi, T. Takaya, K. Nishinari, Rheological and thermal studies of gel-sol transition in gellan gum aqueous solutions, Carbohydrate Polymers, 30 No. 2-3 (1996) 109-119.