- Email: [email protected]
Matrix stiffness determines the fate of nucleus pulposus–derived stem cells
Biomaterials 49 (2015) 68e76
Contents lists available at ScienceDirect
Biomaterials journal homepage: www.elsevier.com/locate/biomaterials
Matrix stiffness determines the fate of nucleus pulposusederived stem cells Yosi Navaro a, Nadav Bleich-Kimelman a, Lena Hazanov b, Iris Mironi-Harpaz b, Yonatan Shachaf b, Shai Garty c, g, Yoav Smith d, Gadi Pelled a, e, f, Dan Gazit a, e, f, Dror Seliktar b, Zulma Gazit a, e, f, * a
Skeletal Biotech Laboratory, The Hebrew UniversityeHadassah Faculty of Dental Medicine, Ein Kerem, Jerusalem 91120, Israel Department of Biomedical Engineering, TechnioneIsrael Institute of Technology, Haifa 32000, Israel Department of Molecular Cell Biology, Weizmann Institute of Science, Rehovot 76100, Israel d Genomic Data Analysis Unit, The Hebrew UniversityeHadassah Medical School, The Hebrew University of Jerusalem, Jerusalem 91120, Israel e Department of Surgery, Cedars-Sinai Medical Center, Los Angeles, CA 90048, USA f Board of Governors Regenerative Medicine Institute, Cedars-Sinai Medical Center, Los Angeles, CA 90048, USA g Max Planck Institute for Intelligent Systems, Stuttgart, Germany b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 September 2014 Accepted 20 January 2015 Available online
Intervertebral disc (IVD) degeneration and consequent low-back pain present a major medical challenge. Nucleus pulposusederived stem cells (NPeSCs) may lead to a novel therapy for this severe disease. It was recently shown that survival and function of mature NP cells are regulated in part by tissue stiffness. We hypothesized that modification of matrix stiffness will influence the ability of cultured NP-SCs to proliferate, survive, and differentiate into mature NP cells. NP-SCs were subcultured in threedimensional matrices of varying degrees of stiffness as measured by the material's shear storage modulus. Cell survival, activity, and rate of differentiation toward the chondrogenic or osteogenic lineage were analyzed. NP-SCs were found to proliferate and differentiate in all matrices, irrespective of matrix stiffness. However, matrices with a low shear storage modulus (G0 ¼ 1 kPa) promoted significantly more proliferation and chondrogenic differentiation, whereas matrices with a high modulus (G0 ¼ 2 kPa) promoted osteogenic differentiation. Imaging performed via confocal and scanning electron microscopes validated cell survival and highlighted stiffness-dependent cell-matrix interactions. These results underscore the effect of the matrix modulus on the fate of NP-SCs. This research may facilitate elucidation of the complex cross-talk between NP-SCs and their surrounding matrix in healthy as well as pathological conditions. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Elasticity Matrix stiffness Fibrinogen Hydrogel Stem cells Intervertebral disc
1. Introduction Intervertebral disc (IVD) degeneration and consequent low-back pain present a major medical challenge with no optimal solution in sight. In Western society, this pathological condition is prevalent among people younger than 45 years of age and percentage of the work force affected varied from 2% to 8% with days of absence from work per patient [1].
* Corresponding author. Board of Governors Regenerative Medicine Institute, Cedars-Sinai Medical Center, Los Angeles, CA 90048, USA. Fax: þ1 (310) 248 8066. E-mail address: [email protected] (Z. Gazit). http://dx.doi.org/10.1016/j.biomaterials.2015.01.021 0142-9612/© 2015 Elsevier Ltd. All rights reserved.
The IVD consists of three major anatomic zones: the nucleus pulposus (NP), the annulus fibrosus (AF), and cartilage endplates. These three anatomic zones are distinct but uniquely attached, contributing to the mechanical function of the IVD [2,3]. The NP is a gel-like substance that contributes to the load-bearing capacity of the IVD and sustains the flexion/extension and lateral bending are spine movements required for many daily activities. This gelatinous substance plays an important role in the IVD's mechanical function by redistributing spinal compressive loads [4]. During IVD degeneration, the disc's biophysical and biochemical properties are altered. The pathogenesis of IVD degeneration is complex and principally relies on resident NP cells to revitalize tissue [3,4]. The NP is originally derived from the notochord [5e7]; the NP cells in the immature nucleus are smaller and contain more
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
condensed and smaller nuclei than the notochordal cells [8]. By early adulthood, the NP becomes populated by chondrocyte-like cells [2,5,6]. It is well established that the NP contains NP-derived stem cells (NPeSCs), which maintain homeostasis in this tissue [2,7,8] and have been shown to commit to osteogenic and chondrogenic lineages in vitro [9]. With respect to the biochemical composition of NP tissue, it is highly hydrated and contains negatively charged sulfated glycosaminoglycans (sGAGs), collagens, and other noncollagenous proteins [3,6]. On initiation of IVD degeneration, the biosynthesis of proteoglycans and collagen is inhibited [1]. Specifically, the NP cell population alters its biochemical and secretory function, resulting in decreases in proteoglycan content and in aggrecan and type II collagen gene expression, as well as biophysical changes such as decreased osmotic pressure, cell volume, and fixed charge density [4]. These alterations initiate a catabolic modification in the structure of the extracellular matrix (ECM), which normally maintains IVD functionality [5]. Evidently, these alterations also affect the cell's proliferation rate and differentiation potential [9]. Recent studies have shown that interactions between NP cells and the ECM depend not only on matrix protein composition, but also on matrix stiffness; thus, changes in cell fate and phenotype are likely to be modulated by tissue stiffness and the mechanical microenvironment of the cells [10]. These observations are in complete agreement with those of others who demonstrated earlier that the mechanical stiffness of the ECM has a profound effect on the fate of mesenchymal stem cells (MSCs) [11e14]. For example, Choi et al. [15] documented that the fate of MSCs is directed down an osteogenic lineage when these cells are exposed to certain ECM mechanical properties. Moreover, the secretion of calcified ECM by osteoblast-like cells was shown to be a mechanosensitive response to substrate stiffness [16]. Similarly, others have demonstrated that ECM stiffness modulates the proliferation and differentiation of mature NP cells in vitro [5,6,11,12,17]. Recently, we demonstrated that cultures of NP-SCs from degenerate discs differ in their proliferation and differentiation capacities when compared to cells from healthy discs [9]. We proposed that ECM stiffness and changes in the elasticity of the degenerated disc matrix are associated with the impaired function of resident NP-SCs, which probably contributes to the onset of the IVD degeneration process. However, to the best of our knowledge, there have been no studies to date that document the effect of matrix stiffness on the fate of NP-SCs. There is a growing consent that 3D models recreate significant characteristics of the microenvironment providing more relevant biological information than 2D models. In the conventional 2D cultures, cells propagate in monolayers on a solid substrate; they grow flat and reach a short height, relatively fixed. Cells that are grown in a 3D model keep a 3D structure, more versatile, with measurable dimensions all round; furthermore, the interactions between close cells cultured in 3D are all around, not restricted to the boundaries of the cells in a particular plane, as in 2D cultures [18]. In the current investigation, we set out to explore the role of matrix stiffness on the fate of NP-SCs in culture. To do this, we developed a hydrogel biomaterial system that can encapsulate NP-SCs in a threedimensional (3D) culture, sustain the cells' survival, and provide mechanical cues to the cells based on variations in the storage shear modulus of the encapsulating milieu. This tunable hydrogel is a semi-synthetic material made from adducts of fibrinogen and poloxameric block copolymers called Tetronic 1307. We took advantage of the distinct physical properties of the fibrinogeneTetronic adducts to control the physical properties of the resulting hydrogel in a manner independent from other critical material properties such as ligand density (i.e. fibrinogen concentration). These materials enabled us to investigate the in vitro fate of
69
NP-SCs in a highly controlled mechanical environment and provided us with a potential scaffold material for NP regeneration. Accordingly, we addressed the hypothesis that modification of matrix stiffness can be used to enhance the ability of NP-SCs to proliferate, survive, and differentiate into mature NP cell populations. 2. Materials and methods 2.1. Hydrogel manufacturing 2.1.1. Tetronic 1307etetraacrylate synthesis The acrylation of Tetronic 1307 (T1307, also known as Pluracare1307, O-BASF MW ¼ 18 kDa) was conducted in the same way as the deacrylation of polyethylene glycol [19]. Briefly, the process was carried out under an Argon atmosphere by reacting T1307etetraol in a solution of dichloromethane (Aldrich, Sleeze, Germany) and toluene (Bio-Lab, Jerusalem, Israel) with acryloylechloride (Merck KGaA, Darmstadt, Germany) and triethylamine (TEA) (Fluka, Buchs, Switzerland) at a molar ratio of 150% relative to the hydroxyl groups. The final product was precipitated out in petroleum ether (40 e60 C) (Bio-Lab). The solid polymer was dried under vacuum conditions for 48 h. The product was characterized by proton nuclear magnetic resonance (NMR) to determine the average number of acryl groups on the T1307 molecule. 2.1.2. Conjugation of T1307 to fibrinogen Fibrinogen was conjugated to Tetronic-tetraacrylate (T1307-TA) by a Michaeltype addition reaction. To conjugate fibrinogen to the synthetic polymers, an 8.3-mg/ml solution of fibrinogen in PBS (150 mM) with 8 M urea was supplemented with tris (2-carboxyethyl) phosphine hydrochloride (TCEP) (Sigma, St. Louis, MO, USA) at a molar ratio of 1.5:1 TCEP to fibrinogen cysteines. After dissolution of the fibrinogen, the functionalized polymer (T1307-TA) in a solution of PBS and 8 M urea (260 mg/ml) was added at a molar ratio of 4:1 synthetic polymer to fibrinogen cysteines (T1307-TA). The reaction was incubated for 3 h at room temperature, after which the volume of reaction was doubled by adding PBS and 8 M urea. The conjugated protein was precipitated out by adding 6 volumes of acetone (Bio-Lab) to the final solution. The precipitate was dissolved in PBS containing 8 M urea at a protein concentration of 7e9 mg/ml and then dialyzed against PBS at 4 C for 2 days with two changes of PBS per day (Spectrum, MW cutoff 12e14 kDa). The net fibrinogen concentration in the Tetronic1307efibrinogen (TF) precursor was determined using a Nano-drop ND-2000 spectrophotometer. The final product was characterized according to previously published protocols [20]. 2.2. Biomechanical analysis of the hydrogel 2.2.1. In situ rheological characterization The in situ hydrogel formation, mechanical properties, and cross-linking kinetics were characterized using an AR-G2 shear rheometer (TA Instruments, New Castle, DE, USA) equipped with a Peltier plate temperature-controlled base. Time-sweep oscillatory tests were performed in 50-mm parallel-plate quartz geometry using 600 ml of acellular TF precursor solution containing 0.1% w/v Irgacure 2959 initiator (Ciba, Basel, Switzerland). To monitor the in situ liquid-to-solid transition (gelation), the acellular precursor solution was first held at 4 C for 1 min to allow the liquid to equilibrate, followed by a 7-min exposure to UV light (365 nm, 2 mW/cm2). Then the temperature was raised to 37 C where it was held for an additional 5 min without UV light. To find the linear viscoelastic region of the time-sweep tests, oscillatory strain (0.1e10%) and frequency sweeps (0.1e10 Hz) were conducted in two separate samples at 4 C (following exposure to UV light) and again at 37 C (the temperature was raised to 37 C after exposure to UV light at 4 C, as before). The linear viscoelastic region was found to be in the range of 2% strain and 1 Hz frequency (data not shown). Using time-sweep experiments, the viscoelastic material properties, including the storage and loss modulus values (G0 and G00 ), as well as the phase angle, were continuously recorded [21]. The plateau storage modulus, G0 was reported as the characteristic measure of the elastic properties of the hydrogels. The reported G0 was taken as real part of complex shear modulus, G* ¼ G0 þ iG'0 at the conclusion of the time-sweep test. Consequently, the G00 values were one to two orders of magnitude lower then the G0 values. Two acellular hydrogel compositions were characterized for the experimental design in the proceeding sections: a low modulus (G0 ¼ 1 kPa) and a high modulus (G0 ¼ 2 kPa) material. The two test group materials were created by adding increasing amounts of Tetronic 1307-TA to the TF, and they were identified by their initial in situ measurements of storage shear modulus. The inclusion of cells into the precursor solutions did not alter their in-situ rheological properties (data not shown). 2.2.2. Ex situ rheological characterization Ex situ measurements of hydrogel modulus were designed to measure how the hydrogel modulus was affected by cultured cells after certain durations in vitro. For this, precast hydrogels cultured at various time points were placed in an AR-G2 shear rheometer (TA Instruments, New Castle, DE, USA) equipped with 20-mm parallel-plate geometry. To ensure that the ex situ rheological characterization results were valid, the sample diameter was always matched to the diameter of the
70
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
rheometer geometry (i.e. 20 mm). A stainless-steel punch and Teflon die were used to create 20-mm disks from the cultured hydrogel constructs before each test in the rheometer [22]. Time-sweep oscillatory tests were performed at 37 C for 4 min using 2% strain and 1 Hz on the precast 20-mm disc-shaped constructs. To avoid slippage of the hydrogels from the geometry, an abrasive patch of the same size (20 mm) was glued to the rheometer plate. The gap size was controlled using the normal force sensor in the rheometer and kept within 0.5 N compressive force. All measurements exhibited a plateau throughout the whole test; thus the first 60 s of the measurement were averaged, and the storage modulus was recorded as the real part of the complex shear modulus (i.e. G0 ). 2.3. Isolation and culture of NP-SCs NP-SCs were isolated and cultured in a manner previously described [9]. Briefly, some NP substance was harvested from healthy porcine IVDs using sterile surgical tools. After isolation, the NP substance was subjected to overnight enzymatic digestion using Dulbecco's modified Eagle medium (DMEM) with 4.5 g/L D-glucose (Biological Industries, Beit-Haemek, Israel) supplemented with 10% fetal calf serum (FCS) (Biological Industries), 2 mM L-glutamine (Biological Industries), 1 mM sodium pyruvate (Sigma, St. Louis, MO, USA), 100 U/ml penicillinestreptomycin (Biological Industries), 25 mM HEPES (Biological Industries), 2 mg/ml collagenase type IV (Sigma), and 0.15 mg/ml hyaluronidase (Sigma) at 37 C with stirring. Debris of the NP substance was filtered using 100-mm nylon mesh (BD Falcon, Bedford, MA, USA). Trypan blue exclusion dye and an automated cell counter (Countess™, Invitrogen, Paisley, UK) were used to count the viable cells. The NP-derived cells were plated at a density of 1.8 105 cells/cm2 in complete medium containing DMEM with 4.5 g/L D-glucose containing 10% FBS (Biological Industries), 100 U/ml penicillinestreptomycin (Biological Industries), and 2 mM L-glutamine (Biological Industries). The medium was changed after 72 h and then every 3e4 days. Upon confluence, the cells were trypsinized using 0.25% trypsin-EDTA (Invitrogen, Paisley, UK) and replated at a density of 5 103 cells/cm for expansion. The cells were cultured to the 2nd or 3rd passage (P2 and P3, respectively) and used for the assays described in the proceeding sections. Bone marrowederived MSCs (BM-MSCs), which were used as controls in some experiments, were isolated and cultured as previously reported [23,24]. 2.3.1. Three-dimensional culture of NP-SCs in tetronic 1307efibrinogen (TF) NP-SCs were dispersed in TF hydrogel precursor solution, encapsulated after gelation, and subcultured within 3D TF hydrogels for up to 21 days, similar to previously described protocols [22]. Briefly, NP-SCs (at P2 or P3) were seeded in 1-kPa or 2-kPa hydrogels containing 0.1% w/v Irgacure 2959 initiator (Ciba) at a seeding density of 3$106 cells/ml. The hydrogel precursors were prepared fresh for each experiment. Each TF hydrogel (total volume 50 ml) was dispensed in a 5-mmdiameter silicone mold. The hydrogels were incubated for 5 min at 4 C and then cross-linked under long-wave UV light (365 nm, 4e5 mW/cm2, Vilber Lourmat, Marne la Vallee, France) for 5 min. The cross-linked cell-seeded hydrogels were incubated in a humidified incubator (37 C, 5% CO2) in culture medium as described above. Cell culture medium was changed every 3e4 days throughout the experiment. In some experiments, a 2D culture plate was used as a control group. 2.3.2. Histological analysis For the histological analysis, the hydrogels were fixed in 4% formalin, passed through a graded series of ethanol solutions, and embedded in paraffin. Sections (5-mm thick) were cut from each paraffin block by using a motorized microtome (Leica Microsystems, Wetzlar, Germany). For detection purposes the cells were labeled with CM-DiI before inclusion [25]. Slides containing cells labeled with CM-DiI were mounted using GVA Mounting Solution (Invitrogen). CM-DiI stain was captured using fluorescent microscopy. Additional slides were stained with hematoxylin and eosin (H&E). 2.4. Cell proliferation Cell survival and metabolic activity were assessed 7 and 14 days after cell seeding by quantifying the dsDNA content with the aid of the PicoGreen assay (Molecular Probes, Eugene, OR, USA), as previously described [26]. Briefly, 50 ml cell-containing hydrogel samples were homogenized and enzymatically digested. The samples were then diluted in 100 ml buffer (10 mM TriseHCl, 1 mM EDTA, pH 7.5) and the fluorescent PicoGreen dye was added. The samples were analyzed following a 5-min incubation. Protein content was measured using the BCA assay (Pierce, Rockford, IL, USA) according to the manufacturer's protocol. 2.5. Differentiation assays Differentiation of NP-SCs toward the chondrogenic or osteogenic lineage was analyzed using reverse-transcription polymerase chain reaction (RT-PCR) followed by quantitative real-time PCR against pivotal chondrogenic (aggrecan, sox9) or osteogenic (osteocalcin, osteopontin) genes. In addition, biochemical assaysda dimethylmethylene blue (DMMB) assay for sGAGs content and an alkaline phosphatase (ALP) colorimetric assaydwere performed to validate the chondrogenic and osteogenic phenotypic expression of cultured cells.
2.5.1. Chondrogenic differentiation analysis Total RNA was extracted from cells that were seeded in the TF hydrogels (1 kPa, 2 kPa) or in pellet culture as a control. Pellet cultures were prepared as described elsewhere [25]. RNA extraction was performed on Days 1, 7, and 14. To dissolve the harvested hydrogels, we incubated them for 1.5 h in 1 mg/ml Collagenase C7657 (Sigma) and 3 mg/ml Collagenase C2674 solution (Sigma) (for 1-kPa and 2-kPa hydrogels, respectively). RNA was extracted using the RNeasy Mini kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's protocol. RNA was retrotranscribed using random primers and reverse transcriptase (Promega Corp., Madison, WI, USA); Aggrecan and b-actin gene expression were analyzed using quantitative real-time PCR. Quantitative real-time PCR was performed with the aid of the Step One system and program (ABI, Foster City, CA, USA) and analyzed using the relative quantification method. The primers that we used included the TaqMan gene expression assay ACAN-Ss03374825_m1 (ABI) for aggrecan, the TaqMan assay on demand ACTB-Ss03376160_u1 (ABI) for b-actin, and the Ss03392406_m1 (ABI) for sox9. Aggrecan and sox9 gene expression was normalized to the housekeeping gene b-actin and calibrated to the gene expression of cells in the pellet culture on Day 1. Production of sGAGs by NP-SCs was assessed using the dimethylmethylene blue (1,9-DMMB, Sigma, St. Louis, MO, USA) assay, which was modified to suit assessment of sGAGs accumulation in hydrogels. Each sample was suspended in a digestion solution containing 125 mg/ml papain, 100 mM phosphate buffer, 10 mM cysteine, and 10 mM EDTA, pH 6.3 (all from Sigma), and incubated overnight in 65 C. Finally, each 50 ml of digested cells containing hydrogels were homogenized and enzymatically digested. 10 ml of each sample were diluted with 250 ml of DMMB. The optical density was measured with a spectrophotometer at a wavelength of 525 nm. 2.5.2. Osteogenic differentiation analysis Total RNA was extracted, reverse transcribed, and analyzed quantitatively using real-time PCR as described above. The expression of osteocalcin (Ss03373655_s1, BGLAP, ABI) and osteopontin, also known as secreted phosphoprotein 1 (Ss03391322_g1, SPP1, ABI) genes, was normalized to the housekeeping gene b-actin and calibrated to the gene expression of NP-SCs cultured on 2D plates on Day 1. In addition, The ALP activity was determined for the hydrogel treatments. On Days 7 and 14, the hydrogels were digested using collagenase, as described above, and the ALP activity was assessed as previously described [25]. Briefly, 50-ml cell-containing hydrogel samples were homogenized and enzymatically digested. The samples were then diluted into a 400-ml phosphate substrate working reagent, whirled in a vortex device, and incubated for 10 min in 37 C. The optical density was measured at a wavelength of 404 nm. Values were normalized to the sDNA amount, which was measured using the PicoGreen assay. 2.6. Cell morphology analysis The NP-SC morphology was analyzed using immunofluorescence staining against the protein paxillin, a marker for focal adhesion points, and counterstained with fluorescent dyes to detect the cytoskeleton protein actin. To further visualize cellmatrix interactions, samples were imaged using scanning electron microscopy (SEM). 2.6.1. Paxillin and actin immunofluorescence Hydrogels seeded with NP-SCs were harvested on Days 7 and 14 and fixed using 10% formalin solution; afterward they were washed with 0.3% Triton and PBS. Following an overnight incubation with blocking solution (1% BSA), the hydrogels were incubated with purified mouse antiepaxillin primary antibody (dilution 1:250; BD Biosciences Pharmingen, San Diego, CA, USA) for 2 h at room temperature and for an additional 2 h at 4 C. Subsequently, the samples were incubated with secondary antibody for 4 h at room temperature (Jackson ImmunoResearch, West Grove, PA, USA). For counterstains, actin filaments were stained using PhalloidinTRITC (Sigma), and the nuclei were stained with SYTOX blue (dilution 1:2500; Molecular Probes, Eugene, OR, USA) for 30 min. Finally, the hydrogels were washed 3 times in PBS at room temperature, incubated overnight in PBS, and visualized using the Zeiss LSM 700 Laser scanning microscope (Thornwood, NY, USA). Paxillin staining was quantified using MatLab (Mathworks, Natick, MA). The software was used for processing images of cells labeled for actin and paxillin. Paxillin staining was calculated as the ratio of paxillin staining to the cell surface area (as identified by staining for actin). 2.6.2. High-resolution scanning electron microscopy (HR-SEM) HR-SEM was used to ascertain the cellular organization within the hydrogels. Sample preparation for the SEM included cell fixation and dehydration, as previously described [26]. Electron micrographs were obtained using a Sirion highresolution scanning electron microscope (HR-SEM, FEI, Eindhoven, The Netherlands) with the voltage set at 3e5 kV. The samples were fixed using 2.5% w/v glutaraldehyde (Sigma) and 2% w/v paraformaldehyde (PFA, Sigma) in 200 mM cacodylic acid (Merck) and 10 nM CaCl2. After washing, the samples were dehydrated using increasing concentrations of ethanol, followed by drying with a Balzers CPD 030 Critical Point Dryer (CPD). Samples were spatter-coated for 60 s until an approximate 10-nm palladium/gold (Pd/Au) coating was formed and subsequently imaged.
Y. Navaro et al. / Biomaterials 49 (2015) 68e76 2.7. Statistical analysis Data are expressed as means ± standard errors. Significant differences between data sets were found using the Student's t-test (two way); the minimal criterion for significance was set at p < 0.05.
71
on Day 1 (Fig. 1b). Although the stiffness of the 2-kPa matrix was slightly higher than that of the 1-kPa matrix on Day 14, this difference was not statistically significant.
3. Results
3.3. Cell survival enhanced in 1-kPa hydrogels
3.1. Mechanical properties of the hydrogels
Cell survival and proliferation were analyzed using both the PicoGreen assay (which measures double-stranded DNA) and the BCA assay (which measures total protein content) on Days 7 and 14. NP-SCs were found to proliferate in both types of matrices. Cells seeded in the 1-kPa matrix displayed higher proliferation rates than cells seeded in the 2-kPa matrix over time (Fig. 2a). Moreover, the amount of protein in the hydrogels on Day 14 was significantly higher in the 1-kPa matrix than in the 2-kPa matrix (Fig. 2b). Cells were observed on Day 7, expanding throughout the gel space, with low cell-to-cell contact, and a light preponderance of elongated shape in the cells encapsulated inside the 1-kPa matrix (Fig. 2c, DiI staining).
Tetronic®1307 is a synthetic tetrafunctional ethylene oxide/ propylene oxide block copolymer that, much like other members of its poloxameric family (e.g. Pluronic®F127), displays a reversible temperature-induced physical solegel transition and an irreversible UV light–activated chemical cross-linking (photo-polymerization). Conjugating fibrinogen, a biological protein, to Tetronic®1307 produces a new biosynthetic precursor, which retains its reversible temperature-induced transition and photo-polymerization (Fig. 1a) and is also susceptible to protease degradation and consequent cell-mediated remodeling (data not shown). Similar to other protein-polymer conjugates [27], this protein-based material also conveys inductive signals to cells through bioactive sites on the fibrinogen backbone, as well as through other biomechanical properties such as the matrix modulus. A hydrogel was formed from the TF precursor in the presence of NP-SCs by non-toxic free-radical polymerization using light activation. Two different TF materials, each having a shear storage modulus (G0 ) of 1 kPa or 2 kPa, were fabricated (Fig. 1a) to study the effect of matrix stiffness on the proliferation, differentiation, and scaffold remodeling of NP-SCs. 3.2. Ex situ modulus enhanced after 14 days in culture The ex situ modulus of the NP-SCsecontaining scaffold with an initial modulus of 1 kPa (measured by in situ rheometry) increased gradually between Days 1 and 14, resulting in a significant 60% increase in the shear storage modulus of the construct (Fig. 1b, p < 0.05). No difference was detected in the ex situ mechanical properties of the construct containing NP-SCs with an initial modulus of 2 kPa between Day 1 and Day 7; however, by Day 14, the matrix modulus increased significantly (by 40%) compared to that
3.4. Lower modulus induces differentiation of NP-SCs toward a chondrogenic lineage The expression of aggrecan, a pivotal chondrogenic gene in NP-SCs, was higher in cells seeded in hydrogels with the lower shear storage modulus (1 kPa) than in cells cultured in the 2-kPa hydrogels. A comparison of gene expression of cells seeded in the 1-kPa constructs after 7 days in culture revealed a 4-fold increase in aggrecan expression (n ¼ 8) over that measured on Day 1 (Fig. 3a) and significantly higher expression than cells seeded in the 2-kPa hydrogel at the same time point. All results were normalized with aggrecan expression at Day 1 of culture. All groups were cultured under complete medium conditions with no chondrogenic supplements. The evaluation of gene expression of the chondrogenic gene sox9 likewise indicated higher expression in cells seeded in the 1-kPa construct than in cells seeded in the 2-kPa construct (Fig. 3b). The expression level of type-II collagen was also tested, but no significant differences were found between cells seeded in 1-kPa and 2-kPa hydrogels on Days 7 and 14 (data not shown). Nevertheless, the DMMB assay revealed significantly higher levels of
Fig. 1. In situ and ex situ rheological testing of TF materials: Characterization of the shear storage modulus (G0 , Pa) to examine mechanical differences between groups. a) In situ rheological time-sweep measurements show lower and higher shear storage moduli (G0 , Pa) as a function of cross-linking time and temperature. The first stage shows chemical cross-linking at 4 C, and the second stage shows physical cross-linking at 37 C. b) Ex situ shear storage moduli (G0 , Pa) of 1-kPa and 2-kPa TF materials containing NP-SCs at several time points (n ¼ 3; *, **p < 0.05). The classification of treatment groups is based on the value of the modulus obtained for acellular hydrogels using in situ rheology measurements (1 and 2 kPa).
72
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
Fig. 2. Effect of 1- and 2-kPa TF matrices on NP-SC survival and proliferation. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices. Cell survival and proliferation were assessed using two different types of assays. a) DNA quantification: Samples were processed 7 and 14 days after cell seeding, and the concentrations of double-stranded DNA were determined using the PicoGreen assay. Bars indicate SE (n ¼ 7, *p < 0.05). b) Protein content: Samples were processed 7 and 14 days after cell seeding, and protein concentrations were determined using the BCA assay. Protein concentration secreted by the cells was higher in the low-elasticity modulus, as revealed by the BCA assay. Bars indicate SE (n ¼ 3, *p < 0.05). c) Spatial distribution of the cells in the matrices 7 days after cell seeding: Fluorescence images of DiI stained cells and H&E staining of processed gels. Scale bars vary according to the different captions. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
sGAGs accumulation in the 1-kPa group than in the 2-kPa group on Days 7 and 14 (Fig. 3c). Significant differences were also noted when the 1-kPa hydrogel group was compared with the control group (NPeSCs in pellet culture) on Day 14. 3.5. Higher modulus induces differentiation of NP-SCs toward the osteogenic lineage NP-SCs that were cultured in 2-kPa hydrogels without osteogenic supplements for 14 days displayed induction of osteogenic differentiation. At Day 14, osteocalcin expression was 15 times greater in the 2-kPa group than in the 1-kPa group (Fig. 4a). Osteopontin expression was significantly higher in the 2-kPa group than in the 1-kPa group on Day 7 (Fig. 4b). Consistently, ALP activity in NP-SCs cultured in the 2-kPa hydrogels was significantly higher
than that in cells cultured in the 1-kPa hydrogels on Day 7; this ALP activity decreased in both groups by Day 14 (Fig. 4c). 3.6. Variations in the matrix modulus affect NP-SCs morphology and cell-matrix interactions Confocal microscopy and SEM (Fig. 5a) were used to analyze cell survival and cell-matrix interactions. Both types of hydrogels provided the cells with a favorable environment for appropriate cell attachment and growth. The 1-kPa hydrogels were characterized by a rough topography, containing more surface area and thus promoting increased adhesion and anchoring. These cell-matrix contacts promoted a stretched and long, spindle-shaped fibroblast-like morphology, as evidenced by actin-stained cells (Fig. 5b). In contrast, NP-SCs seeded in 2-kPa hydrogels exhibited a smoother
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
73
Fig. 3. Effect of 1-kPa and 2-kPa TF matrices on induction of chondrogenesis with NP-SCs. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices. Chondrogenesis was assessed using three independent assays on Days 1, 7, and 14; and the results were always normalized to data from Day 1. a) Aggrecan gene expression (by quantitative PCR): Samples were processed 1, 7, and 14 days after cell seeding. Chondrogenic gene expressiondAggrecandwas measured by quantitative real-time PCR. Cells subcultured in 1-kPa hydrogels revealed a higher rate of Aggrecan gene expression. Bars indicate SE (n ¼ 8, *p < 0.05). b) Chondrogenic gene expressiondSox9dwas measured by quantitative real-time PCR. Cells subcultured in 1-kPa hydrogels revealed a higher rate of Sox9 gene expression. Bars indicate SE (n ¼ 8, *p < 0.05). c) Sulfated glycosaminoglycans (sGAGs) content was quantified by performing a dimethylmethylene blue (DMMB) assay. Samples were processed on Days 1 and 14 after cell seeding, and sGAGs accumulation in the hydrogels was measured. Bars indicate SE (n ¼ 3, *,**p < 0.05).
topography, with cells tightly surrounded by the gel (Fig. 5a). Those cells displayed a spherical morphology with less stretched out actin fibers (Fig. 5b). The SEM cell morphology observations were in complete agreement with the confocal microscopy images. Quantitative analysis of focal adhesions using paxillin staining at Day 14 (Fig. 5b) revealed more paxillin stain per cell in cells cultured in 1-kPa hydrogels than in cells in the 2-kPa group; this could indicate that NP-SCs attach better to the softer matrix. 4. Discussion Previous studies have shown that matrix stiffness affects the fate of BM-MSCs [11e13] as well as mature NP cell organization, function, and morphology [5,11,12]. Survival, genotype, and phenotypic expression of MSCs were all shown to be affected by the modulus of the cellular microenvironment [5,10e12,28]. In this study, we tested the effect of matrix stiffness on NP-SCs by using a Tetronic-fibrinogen (TF) hydrogel that allowed us to control the mechanical properties of the cell-encapsulating biomaterial while keeping other material properties, such as ligand density (i.e. fibrinogen concentration), unchanged. Using various assays we analyzed the effect of the matrix modulus on the survival and differentiation of NP-SCs. Our results reveal that lower-modulus hydrogelsd1 kPa, similar to the physiological modulus found in
native IVDdpromoted cell survival and chondrogenic differentiation, as well as enhanced cell attachment to the matrix. Unlike previous studies in which BM-MSCs were used [11,12,28], here we investigated MSCs isolated from lumbar IVDs. The porcine IVD used in this study is known to share common characteristics with the human IVD [29e31]. Therefore, primary cells isolated from a porcine IVD closely resemble human NP-derived SCs. Consequently, the elastic shear modulus of the human NP substance was widely investigated and reported to be in the range of 0.1 kPae1 kPa [4e6,17]. This variability in the elastic modulus of IVD is mainly due to the different test types that have been used. Here, we used the low modulus matrix (1-kPa group) to mimic the mechanical properties of the healthy NP and a slightly stiffer matrix (2-kPa group), which represents a more cross-linked version of the same material. We used these two groups (compliant and stiff, respectively) to investigate the cell response to matrix elasticity, while keeping other properties, such as fibrinogen concentration, unchanged. In this context, the TF biomaterial provided us with the ability to control cross-linking density without changing the fibrinogen composition by adding small amounts of additional cross-linker to the hydrogel precursor solution (Fig. 1a). The NP-SCs were cultured within these two types of hydrogels. Assessments were made based on cell differentiation, proliferation, and viability; and the results were correlated to changes in the initial material
74
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
Fig. 4. Effect of 1- and 2-kPa TF matrices on induction of osteogenesis with NP-SCs. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices, and gene expression was measured by quantitative real-time PCR for a) Osteocalcin and b) Osteopontin 7 and 14 days after cell seeding. The gene expression rate was higher in cells seeded in 2-kPa hydrogels. Bars indicate SE (n ¼ 5, *p < 0.05). c) Alkaline phosphatase (ALP) (colorimetric assay) activity was determined 7 and 14 days after cell seeding.
modulus. Our data show strong correlations between the initial modulus of the 3D culture matrix and the biological response of the NP-SCs after several days in culture. Regarding the ex situ shear storage modulus, it was evident that the mechanical properties of both the compliant and stiff hydrogels increased in response to the NP-SC culture. We speculate that a higher increase in the modulus of the compliant material was the result of enhanced modulus-dependent cell remodeling (and proliferation), leading to heightened ECM production in these constructs. Dikovsky et al. [32] showed that MSC spreading in PEGfibrinogen 3D matrices occurred more rapidly when the storage modulus was low. Moreover, Kesselman et al. [22] showed that fibroblasts cultured within soft 3D PEG-fibrinogen hydrogels produced ECM and significantly increased their matrix modulus after as little as 7 days in culture. In the current study, we observed that the gradual increase in the modulus of the compliant (1-kPa) hydrogel between Days 1 and 14 and the increase in the modulus of the 2-kPa scaffold observed between Days 7 and 14 were consistent with higher proliferation rates and improved spindled cell morphologies (Fig. 2a and Fig. 5b for the 1-kPa matrix) as well as possibly associated with protein secretion (Fig. 2b). NP-SCs seeded in 1-kPa matrices also exhibited a higher rate of chondrogenic differentiation (Fig. 3aec). Aggrecan deposition is considered a hallmark of chondrogenesis and comprises a major part of the IVD's ECM [33]. Aggrecan and sox9 gene expression and sGAGs accumulation were significantly higher in cells cultured in
the soft 1-kPa hydrogels compared with cells cultured in the stiffer 2-kPa hydrogels (Fig. 3aec). The increase in accumulation of sGAGs in cells cultured in 1-kPa hydrogels (Fig. 3c) is consistent with reports from previous studies in which soft matrices were used [12,13]. However, in this study we cultured the cells in 3D matrices, which better simulate in vivo conditions than 2D cell cultures [5]. The underlying mechanism that regulates induction of chondrogenesis in NP-SCs in 1-kPa hydrogels is unknown and needs to be further investigated. However, the mechano-responsiveness of the cells to the ECM plays a major role in the induction of chondrogenesis [34]. Moreover, both the high availability of cellecell interactions and the low oxygen tension inside the hydrogel could also promote chondrogenesis. NP-SCs seeded in 2-kPa matrices showed increased expression of osteogenic differentiation markers (Fig. 4aec). Expression of osteocalcin (a late osteogenic gene marker) was significantly higher in cells cultured in 2-kPa hydrogels than in cells cultured in 1-kPa hydrogels (Fig. 4a). Additionally, expression of osteopontin (a midto-late osteogenic marker) was significantly higher in cells cultured in 2-kPa hydrogels after 1 week of culture compared with cells cultured in 1-kPa matrices (Fig. 4b). These results correlate with those of previous studies using human BM-MSCs and human osteoblasts, which also indicates that a cell fate toward the osteogenic lineage is favored when more rigid substrates are used [14]. Moreover, ALP activity (Fig. 4c) showed a marked elevation in cells cultured in 2-kPa hydrogels compared with cells cultured in 1-kPa
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
75
Fig. 5. Effect of 1- and 2-kPa TF matrices on NP-SC morphology and adhesion to the matrix. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices. a) Scanning electron microscopy (SEM) reveals the cell morphology of the samples 7 and 14 days after cell seeding. b) Confocal microscopy of NP-SCs 14 days after seeding, using fluorescent staining of actin filaments, corroborated the results of the SEM images by showing that cells grown in 1-kPa hydrogels are much more spread out than cells grown in 2-kPa hydrogels. Paxillin protein can be seen labeled in red, actin in green, and cell nuclei in blue. c) Paxillin quantification: Samples were processed 7 and 14 days after cell seeding, and the percentage of paxillin was quantified relative to the cell surface area. Bars indicate SE (n ¼ 6, *p < 0.05). (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
matrices after 7 days. The mechanism by which NP-SCs are genetically and phenotypically induced toward osteogenesis is still uncertain, but some possibilities include ECM stiffness, cell morphology, and other interactions with the ECM, all of which can explain cell phenotype in the unique environment of stiffer IVD tissues such as the annulus fibrosus [4,6,17]. In similar fashion, we speculate that changes in the NP matrix stiffness caused by a degenerative process may inhibit the differentiation of NP-SCs and prevent tissue regeneration within the disc. Variations in matrix stiffness also revealed a profound effect on NP-SC morphology. NP-SCs’ adherence to the fibrinogen domains [35] of the hydrogels showed different degrees of attachment and very different cellular morphologies when cultured in 1-kPa and 2-kPa matrices (Fig. 5a and b). Although it is commonly accepted that osteoblastic cells present spindle morphology while chondrocytic cells tend to be more rounded, here we show that cells seeded in 1-kPa matrices and differentiated to chondrogenic cells, exhibited a spindledeshape morphology, unlike cells seeded in 2-kPa matrices, which were mostly spherical and expressed osteogenic markers of differentiation. Moreover, focal adhesion points, visualized by paxillin staining, were more abundant in cells that were cultured in 1-kPa hydrogels with significant differences observed on Day 14. It is important to indicate that the hydrogel used in this study is not composed of the ECM molecules that are found in the NP, bone or cartilage and therefore we could not expect similar cell morphology as seen in vivo. It is also possible that the lower elastic modulus matrix facilitated better cellecell and
cell-matrix interactions, while cells seeded in stiffer matrices were less able to move around and create cellecell interactions due to the higher stiffness of the encapsulating microenvironment. Another possibility is that the rough topography of the 1-kPa matrix, as demonstrated by SEM images, was due to the exposure of fibrinogen domains that allowed cells to adhere and spread, and thus to acquire their more spindled shape. Interestingly, these finding are in the same line with recently published results from Kim et al. [36]. Authors of previous studies reported that variation in the cellular response to matrix stiffness in different cell types is receptor dependent [37]. Moreover, the use of 2D or 3D matrices has a profound effect on cell morphology, adhesion, and spreading [37e39]. It has also been suggested that cells exert contractile forces and “interpret” substrate deformation to determine a preferred direction for their movements [39e41]. Our findings of NP-SC behavior in the TF hydrogels are in agreement with the results of other studies, revealing the effect of the matrix modulus on the fate of MSCs [5,7,11,12,14,28,42]. However, additional studies are still required to better elucidate the exact cellular mechanism that is responsible for the observed cell behavior. We envision that future treatments for degenerated IVD pathology could potentially utilize TF hydrogels in combination with NP-SCs. The material properties of the TF hydrogels can be fine-tuned to achieve modulus values that mimic the native IVD microenvironment. Accordingly, we have shown that when NP-SCs are cultured in hydrogels having such a modulus, it can help facilitate the proliferation of these cells and direct them to commit
76
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
to a specific and desirable lineage. Hence, our results may pave the way to new tissue-engineering strategies for NP regeneration. 5. Conclusions Our data indicate that NP-SCs proliferate in TF hydrogels. Moreover, we conclude that NP-SCs cultured within hydrogels having a lower (1-kPa) elastic modulus were promoted toward a chondrogenic differentiation pathway, as evidenced by quantitative real-time PCR and the DMMB assay, whereas cells cultured in higher modulus (2-kPa) hydrogels were promoted toward an osteogenic differentiation pathway. These results highlight the effect of matrix stiffness on the fate of NP-SCs. This research may pave the way for a better understanding of the complex cross-talk of cells and their matrix during disc degeneration, as well as present novel therapeutic strategies for IVD regeneration. Acknowledgments We gratefully acknowledge funding support from the National Institutes of Health (RO3 AR057143, Z.G.), Israel Science Foundation grant (no. 614/09, D.G.) and the Sir Zelmen Cowen Universities Fund Blue Sky Research Grant (D.G.). References [1] Manchikanti L, Singh V, Falco FJ, Benyamin RM, Hirsch JA. Epidemiology of low back pain in adults. Neuromodulation e J Int Neuromodulation Soc 2014;17(Suppl. 2):3e10. [2] Hsieh AH, Twomey JD. Cellular mechanobiology of the intervertebral disc: new directions and approaches. J Biomech 2010;43:137e45. [3] Niosi CA, Oxland TR. Degenerative mechanics of the lumbar spine. Spine J : Off J North Am Spine Soc 2004;4:202Se8S. [4] Neidlinger-Wilke C, Galbusera F, Pratsinis H, Mavrogonatou E, Mietsch A, Kletsas D, et al. Mechanical loading of the intervertebral disc: from the macroscopic to the cellular level. Eur Spine J e Off Publ Eur Spine Soc Eur Spinal Deformity Soc Eur Sect Cerv Spine Res Soc 2014;23(Suppl. 3):S333e43. [5] Gilchrist CL, Darling EM, Chen J, Setton LA. Extracellular matrix ligand and stiffness modulate immature nucleus pulposus cell-cell interactions. PLoS One 2011;6:e27170. [6] Setton LA, Chen J. Cell mechanics and mechanobiology in the intervertebral disc. Spine (Phila Pa 1976) 2004;29:2710e23. [7] Blanco JF, Graciani IF, Sanchez-Guijo FM, Muntion S, Hernandez-Campo P, Santamaria C, et al. Isolation and characterization of mesenchymal stromal cells from human degenerated nucleus pulposus: comparison with bone marrow mesenchymal stromal cells from the same subjects. Spine (Phila Pa 1976) 2010;35:2259e65. [8] Henriksson H, Thornemo M, Karlsson C, Hagg O, Junevik K, Lindahl A, et al. Identification of cell proliferation zones, progenitor cells and a potential stem cell niche in the intervertebral disc region: a study in four species. Spine (Phila Pa 1976) 2009;34:2278e87. [9] Mizrahi O, Sheyn D, Tawackoli W, Ben-David S, Su S, Li N, et al. Nucleus pulposus degeneration alters properties of resident progenitor cells. Spine J e Off J North Am Spine Soc 2013. [10] Hwang PY, Setton LA, Chen J, Jing L, Hoffman BD. The role of extracellular matrix elasticity and composition in regulating the nucleus pulposus cell phenotype in the intervertebral disc. J Biomech Eng 2014. [11] Huebsch N, Arany PR, Mao AS, Shvartsman D, Ali OA, Bencherif SA, et al. Harnessing traction-mediated manipulation of the cell/matrix interface to control stem-cell fate. Nat Mater 2010;9:518e26. [12] Guilak F, Cohen DM, Estes BT, Gimble JM, Liedtke W, Chen CS. Control of stem cell fate by physical interactions with the extracellular matrix. Cell Stem Cell 2009;5:17e26. [13] Seliktar D. Designing cell-compatible hydrogels for biomedical applications. Science 2012;336:1124e8. [14] Engler AJ, Sweeney HL, Discher DE, Schwarzbauer JE. Extracellular matrix elasticity directs stem cell differentiation. J Musculoskelet Neuronal Interact 2007;7:335.
[15] Choi JS, Harley BA. The combined influence of substrate elasticity and ligand density on the viability and biophysical properties of hematopoietic stem and progenitor cells. Biomaterials 2012;33:4460e8. [16] Ishaug-Riley SL, Crane-Kruger GM, Yaszemski MJ, Mikos AG. Three-dimensional culture of rat calvarial osteoblasts in porous biodegradable polymers. Biomaterials 1998;19:1405e12. [17] Setton LA, Chen J. Mechanobiology of the intervertebral disc and relevance to disc degeneration. J Bone Jt Surg Am 2006;88(Suppl. 2):52e7. [18] Knight E, Przyborski S. Advances in 3D cell culture technologies enabling tissue-like structures to be created in vitro. J Anat 2014. [19] Halstenberg S, Panitch A, Rizzi S, Hall H, Hubbell JA. Biologically engineered protein-graft-poly(ethylene glycol) hydrogels: a cell adhesive and plasmindegradable biosynthetic material for tissue repair. Biomacromolecules 2002;3:710e23. [20] Dikovsky D, Bianco-Peled H, Seliktar D. The effect of structural alterations of PEG-fibrinogen hydrogel scaffolds on 3-D cellular morphology and cellular migration. Biomaterials 2006;27:1496e506. [21] Plotkin M, Vaibavi SR, Rufaihah AJ, Nithya V, Wang J, Shachaf Y, et al. The effect of matrix stiffness of injectable hydrogels on the preservation of cardiac function after a heart attack. Biomaterials 2014;35:1429e38. [22] Kesselman D, Kossover O, Mironi-Harpaz I, Seliktar D. Time-dependent cellular morphogenesis and matrix stiffening in proteolytically responsive hydrogels. Acta Biomater 2013. [23] Aslan H, Zilberman Y, Arbeli V, Sheyn D, Matan Y, Liebergall M, et al. Nucleofection-based ex vivo nonviral gene delivery to human stem cells as a platform for tissue regeneration. Tissue Eng 2006;12:877e89. [24] Bosch P, Pratt SL, Stice SL. Isolation, characterization, gene modification, and nuclear reprogramming of porcine mesenchymal stem cells. Biol Reprod 2006;74:46e57. [25] Aslan H, Zilberman Y, Kandel L, Liebergall M, Oskouian RJ, Gazit D, et al. Osteogenic differentiation of noncultured immunoisolated bone marrowderived CD105þ cells. Stem Cells 2006;24:1728e37. [26] Garty S, Kimelman-Bleich N, Hayouka Z, Cohn D, Friedler A, Pelled G, et al. Peptide-modified “smart” hydrogels and genetically engineered stem cells for skeletal tissue engineering. Biomacromolecules 2010;11:1516e26. [27] Shachaf Y, Gonen-Wadmany M, Seliktar D. The biocompatibility of PluronicF127 fibrinogen-based hydrogels. Biomaterials 2010;31:2836e47. [28] Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix elasticity directs stem cell lineage specification. Cell 2006;126:677e89. [29] Mosekilde L, Kragstrup J, Richards A. Compressive strength, ash weight, and volume of vertebral trabecular bone in experimental fluorosis in pigs. Calcif Tissue Int 1987;40:318e22. [30] Heino TJ, Alm JJ, Moritz N, Aro HT. Comparison of the osteogenic capacity of minipig and human bone marrow-derived mesenchymal stem cells. J Orthop Res 2012;30:1019e25. [31] Lundin O, Ekstrom L, Hellstrom M, Holm S, Sward L. Exposure of the porcine spine to mechanical compression: differences in injury pattern between adolescents and adults. Eur Spine J 2000;9:466e71. [32] Dikovsky D, Bianco-Peled H, Seliktar D. Defining the role of matrix compliance and proteolysis in three-dimensional cell spreading and remodeling. Biophys J 2008;94:2914e25. [33] Hidaka C, Quitoriano M, Warren RF, Crystal RG. Enhanced matrix synthesis and in vitro formation of cartilage-like tissue by genetically modified chondrocytes expressing BMP-7. J Orthop Res 2001;19:751e8. [34] Cassinelli EH, Hall RA, Kang JD. Biochemistry of intervertebral disc degeneration and the potential for gene therapy applications. Spine J 2001;1:205e14. [35] Almany L, Seliktar D. Biosynthetic hydrogel scaffolds made from fibrinogen and polyethylene glycol for 3D cell cultures. Biomaterials 2005;26:2467e77. [36] Kim TH, An DB, Oh SH, Kang MK, Song HH, Lee JH. Creating stiffness gradient polyvinyl alcohol hydrogel using a simple gradual freezingethawing method to investigate stem cell differentiation behaviors. Biomaterials 2015;40: 51e60. [37] Yeung T, Georges PC, Flanagan LA, Marg B, Ortiz M, Funaki M, et al. Effects of substrate stiffness on cell morphology, cytoskeletal structure, and adhesion. Cell Motil Cytoskelet 2005;60:24e34. [38] Bard JB, Hay ED. The behavior of fibroblasts from the developing avian cornea. Morphology and movement in situ and in vitro. J Cell Biol 1975;67:400e18. [39] Lo C-M, Wang H-B, Dembo M, Wang Y-l. Cell movement is guided by the rigidity of the substrate. Biophys J 2000;79:144e52. [40] Pelham Jr RJ, Wang Y. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A 1997;94:13661e5. [41] Sheetz MP, Felsenfeld DP, Galbraith CG. Cell migration: regulation of force on extracellular-matrix-integrin complexes. Trends Cell Biol 1998;8:51e4. [42] Kong HJ, Polte TR, Alsberg E, Mooney DJ. FRET measurements of cell-traction forces and nano-scale clustering of adhesion ligands varied by substrate stiffness. Proc Natl Acad Sci U S A 2005;102:4300e5.
Contents lists available at ScienceDirect
Biomaterials journal homepage: www.elsevier.com/locate/biomaterials
Matrix stiffness determines the fate of nucleus pulposusederived stem cells Yosi Navaro a, Nadav Bleich-Kimelman a, Lena Hazanov b, Iris Mironi-Harpaz b, Yonatan Shachaf b, Shai Garty c, g, Yoav Smith d, Gadi Pelled a, e, f, Dan Gazit a, e, f, Dror Seliktar b, Zulma Gazit a, e, f, * a
Skeletal Biotech Laboratory, The Hebrew UniversityeHadassah Faculty of Dental Medicine, Ein Kerem, Jerusalem 91120, Israel Department of Biomedical Engineering, TechnioneIsrael Institute of Technology, Haifa 32000, Israel Department of Molecular Cell Biology, Weizmann Institute of Science, Rehovot 76100, Israel d Genomic Data Analysis Unit, The Hebrew UniversityeHadassah Medical School, The Hebrew University of Jerusalem, Jerusalem 91120, Israel e Department of Surgery, Cedars-Sinai Medical Center, Los Angeles, CA 90048, USA f Board of Governors Regenerative Medicine Institute, Cedars-Sinai Medical Center, Los Angeles, CA 90048, USA g Max Planck Institute for Intelligent Systems, Stuttgart, Germany b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 17 September 2014 Accepted 20 January 2015 Available online
Intervertebral disc (IVD) degeneration and consequent low-back pain present a major medical challenge. Nucleus pulposusederived stem cells (NPeSCs) may lead to a novel therapy for this severe disease. It was recently shown that survival and function of mature NP cells are regulated in part by tissue stiffness. We hypothesized that modification of matrix stiffness will influence the ability of cultured NP-SCs to proliferate, survive, and differentiate into mature NP cells. NP-SCs were subcultured in threedimensional matrices of varying degrees of stiffness as measured by the material's shear storage modulus. Cell survival, activity, and rate of differentiation toward the chondrogenic or osteogenic lineage were analyzed. NP-SCs were found to proliferate and differentiate in all matrices, irrespective of matrix stiffness. However, matrices with a low shear storage modulus (G0 ¼ 1 kPa) promoted significantly more proliferation and chondrogenic differentiation, whereas matrices with a high modulus (G0 ¼ 2 kPa) promoted osteogenic differentiation. Imaging performed via confocal and scanning electron microscopes validated cell survival and highlighted stiffness-dependent cell-matrix interactions. These results underscore the effect of the matrix modulus on the fate of NP-SCs. This research may facilitate elucidation of the complex cross-talk between NP-SCs and their surrounding matrix in healthy as well as pathological conditions. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Elasticity Matrix stiffness Fibrinogen Hydrogel Stem cells Intervertebral disc
1. Introduction Intervertebral disc (IVD) degeneration and consequent low-back pain present a major medical challenge with no optimal solution in sight. In Western society, this pathological condition is prevalent among people younger than 45 years of age and percentage of the work force affected varied from 2% to 8% with days of absence from work per patient [1].
* Corresponding author. Board of Governors Regenerative Medicine Institute, Cedars-Sinai Medical Center, Los Angeles, CA 90048, USA. Fax: þ1 (310) 248 8066. E-mail address: [email protected] (Z. Gazit). http://dx.doi.org/10.1016/j.biomaterials.2015.01.021 0142-9612/© 2015 Elsevier Ltd. All rights reserved.
The IVD consists of three major anatomic zones: the nucleus pulposus (NP), the annulus fibrosus (AF), and cartilage endplates. These three anatomic zones are distinct but uniquely attached, contributing to the mechanical function of the IVD [2,3]. The NP is a gel-like substance that contributes to the load-bearing capacity of the IVD and sustains the flexion/extension and lateral bending are spine movements required for many daily activities. This gelatinous substance plays an important role in the IVD's mechanical function by redistributing spinal compressive loads [4]. During IVD degeneration, the disc's biophysical and biochemical properties are altered. The pathogenesis of IVD degeneration is complex and principally relies on resident NP cells to revitalize tissue [3,4]. The NP is originally derived from the notochord [5e7]; the NP cells in the immature nucleus are smaller and contain more
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
condensed and smaller nuclei than the notochordal cells [8]. By early adulthood, the NP becomes populated by chondrocyte-like cells [2,5,6]. It is well established that the NP contains NP-derived stem cells (NPeSCs), which maintain homeostasis in this tissue [2,7,8] and have been shown to commit to osteogenic and chondrogenic lineages in vitro [9]. With respect to the biochemical composition of NP tissue, it is highly hydrated and contains negatively charged sulfated glycosaminoglycans (sGAGs), collagens, and other noncollagenous proteins [3,6]. On initiation of IVD degeneration, the biosynthesis of proteoglycans and collagen is inhibited [1]. Specifically, the NP cell population alters its biochemical and secretory function, resulting in decreases in proteoglycan content and in aggrecan and type II collagen gene expression, as well as biophysical changes such as decreased osmotic pressure, cell volume, and fixed charge density [4]. These alterations initiate a catabolic modification in the structure of the extracellular matrix (ECM), which normally maintains IVD functionality [5]. Evidently, these alterations also affect the cell's proliferation rate and differentiation potential [9]. Recent studies have shown that interactions between NP cells and the ECM depend not only on matrix protein composition, but also on matrix stiffness; thus, changes in cell fate and phenotype are likely to be modulated by tissue stiffness and the mechanical microenvironment of the cells [10]. These observations are in complete agreement with those of others who demonstrated earlier that the mechanical stiffness of the ECM has a profound effect on the fate of mesenchymal stem cells (MSCs) [11e14]. For example, Choi et al. [15] documented that the fate of MSCs is directed down an osteogenic lineage when these cells are exposed to certain ECM mechanical properties. Moreover, the secretion of calcified ECM by osteoblast-like cells was shown to be a mechanosensitive response to substrate stiffness [16]. Similarly, others have demonstrated that ECM stiffness modulates the proliferation and differentiation of mature NP cells in vitro [5,6,11,12,17]. Recently, we demonstrated that cultures of NP-SCs from degenerate discs differ in their proliferation and differentiation capacities when compared to cells from healthy discs [9]. We proposed that ECM stiffness and changes in the elasticity of the degenerated disc matrix are associated with the impaired function of resident NP-SCs, which probably contributes to the onset of the IVD degeneration process. However, to the best of our knowledge, there have been no studies to date that document the effect of matrix stiffness on the fate of NP-SCs. There is a growing consent that 3D models recreate significant characteristics of the microenvironment providing more relevant biological information than 2D models. In the conventional 2D cultures, cells propagate in monolayers on a solid substrate; they grow flat and reach a short height, relatively fixed. Cells that are grown in a 3D model keep a 3D structure, more versatile, with measurable dimensions all round; furthermore, the interactions between close cells cultured in 3D are all around, not restricted to the boundaries of the cells in a particular plane, as in 2D cultures [18]. In the current investigation, we set out to explore the role of matrix stiffness on the fate of NP-SCs in culture. To do this, we developed a hydrogel biomaterial system that can encapsulate NP-SCs in a threedimensional (3D) culture, sustain the cells' survival, and provide mechanical cues to the cells based on variations in the storage shear modulus of the encapsulating milieu. This tunable hydrogel is a semi-synthetic material made from adducts of fibrinogen and poloxameric block copolymers called Tetronic 1307. We took advantage of the distinct physical properties of the fibrinogeneTetronic adducts to control the physical properties of the resulting hydrogel in a manner independent from other critical material properties such as ligand density (i.e. fibrinogen concentration). These materials enabled us to investigate the in vitro fate of
69
NP-SCs in a highly controlled mechanical environment and provided us with a potential scaffold material for NP regeneration. Accordingly, we addressed the hypothesis that modification of matrix stiffness can be used to enhance the ability of NP-SCs to proliferate, survive, and differentiate into mature NP cell populations. 2. Materials and methods 2.1. Hydrogel manufacturing 2.1.1. Tetronic 1307etetraacrylate synthesis The acrylation of Tetronic 1307 (T1307, also known as Pluracare1307, O-BASF MW ¼ 18 kDa) was conducted in the same way as the deacrylation of polyethylene glycol [19]. Briefly, the process was carried out under an Argon atmosphere by reacting T1307etetraol in a solution of dichloromethane (Aldrich, Sleeze, Germany) and toluene (Bio-Lab, Jerusalem, Israel) with acryloylechloride (Merck KGaA, Darmstadt, Germany) and triethylamine (TEA) (Fluka, Buchs, Switzerland) at a molar ratio of 150% relative to the hydroxyl groups. The final product was precipitated out in petroleum ether (40 e60 C) (Bio-Lab). The solid polymer was dried under vacuum conditions for 48 h. The product was characterized by proton nuclear magnetic resonance (NMR) to determine the average number of acryl groups on the T1307 molecule. 2.1.2. Conjugation of T1307 to fibrinogen Fibrinogen was conjugated to Tetronic-tetraacrylate (T1307-TA) by a Michaeltype addition reaction. To conjugate fibrinogen to the synthetic polymers, an 8.3-mg/ml solution of fibrinogen in PBS (150 mM) with 8 M urea was supplemented with tris (2-carboxyethyl) phosphine hydrochloride (TCEP) (Sigma, St. Louis, MO, USA) at a molar ratio of 1.5:1 TCEP to fibrinogen cysteines. After dissolution of the fibrinogen, the functionalized polymer (T1307-TA) in a solution of PBS and 8 M urea (260 mg/ml) was added at a molar ratio of 4:1 synthetic polymer to fibrinogen cysteines (T1307-TA). The reaction was incubated for 3 h at room temperature, after which the volume of reaction was doubled by adding PBS and 8 M urea. The conjugated protein was precipitated out by adding 6 volumes of acetone (Bio-Lab) to the final solution. The precipitate was dissolved in PBS containing 8 M urea at a protein concentration of 7e9 mg/ml and then dialyzed against PBS at 4 C for 2 days with two changes of PBS per day (Spectrum, MW cutoff 12e14 kDa). The net fibrinogen concentration in the Tetronic1307efibrinogen (TF) precursor was determined using a Nano-drop ND-2000 spectrophotometer. The final product was characterized according to previously published protocols [20]. 2.2. Biomechanical analysis of the hydrogel 2.2.1. In situ rheological characterization The in situ hydrogel formation, mechanical properties, and cross-linking kinetics were characterized using an AR-G2 shear rheometer (TA Instruments, New Castle, DE, USA) equipped with a Peltier plate temperature-controlled base. Time-sweep oscillatory tests were performed in 50-mm parallel-plate quartz geometry using 600 ml of acellular TF precursor solution containing 0.1% w/v Irgacure 2959 initiator (Ciba, Basel, Switzerland). To monitor the in situ liquid-to-solid transition (gelation), the acellular precursor solution was first held at 4 C for 1 min to allow the liquid to equilibrate, followed by a 7-min exposure to UV light (365 nm, 2 mW/cm2). Then the temperature was raised to 37 C where it was held for an additional 5 min without UV light. To find the linear viscoelastic region of the time-sweep tests, oscillatory strain (0.1e10%) and frequency sweeps (0.1e10 Hz) were conducted in two separate samples at 4 C (following exposure to UV light) and again at 37 C (the temperature was raised to 37 C after exposure to UV light at 4 C, as before). The linear viscoelastic region was found to be in the range of 2% strain and 1 Hz frequency (data not shown). Using time-sweep experiments, the viscoelastic material properties, including the storage and loss modulus values (G0 and G00 ), as well as the phase angle, were continuously recorded [21]. The plateau storage modulus, G0 was reported as the characteristic measure of the elastic properties of the hydrogels. The reported G0 was taken as real part of complex shear modulus, G* ¼ G0 þ iG'0 at the conclusion of the time-sweep test. Consequently, the G00 values were one to two orders of magnitude lower then the G0 values. Two acellular hydrogel compositions were characterized for the experimental design in the proceeding sections: a low modulus (G0 ¼ 1 kPa) and a high modulus (G0 ¼ 2 kPa) material. The two test group materials were created by adding increasing amounts of Tetronic 1307-TA to the TF, and they were identified by their initial in situ measurements of storage shear modulus. The inclusion of cells into the precursor solutions did not alter their in-situ rheological properties (data not shown). 2.2.2. Ex situ rheological characterization Ex situ measurements of hydrogel modulus were designed to measure how the hydrogel modulus was affected by cultured cells after certain durations in vitro. For this, precast hydrogels cultured at various time points were placed in an AR-G2 shear rheometer (TA Instruments, New Castle, DE, USA) equipped with 20-mm parallel-plate geometry. To ensure that the ex situ rheological characterization results were valid, the sample diameter was always matched to the diameter of the
70
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
rheometer geometry (i.e. 20 mm). A stainless-steel punch and Teflon die were used to create 20-mm disks from the cultured hydrogel constructs before each test in the rheometer [22]. Time-sweep oscillatory tests were performed at 37 C for 4 min using 2% strain and 1 Hz on the precast 20-mm disc-shaped constructs. To avoid slippage of the hydrogels from the geometry, an abrasive patch of the same size (20 mm) was glued to the rheometer plate. The gap size was controlled using the normal force sensor in the rheometer and kept within 0.5 N compressive force. All measurements exhibited a plateau throughout the whole test; thus the first 60 s of the measurement were averaged, and the storage modulus was recorded as the real part of the complex shear modulus (i.e. G0 ). 2.3. Isolation and culture of NP-SCs NP-SCs were isolated and cultured in a manner previously described [9]. Briefly, some NP substance was harvested from healthy porcine IVDs using sterile surgical tools. After isolation, the NP substance was subjected to overnight enzymatic digestion using Dulbecco's modified Eagle medium (DMEM) with 4.5 g/L D-glucose (Biological Industries, Beit-Haemek, Israel) supplemented with 10% fetal calf serum (FCS) (Biological Industries), 2 mM L-glutamine (Biological Industries), 1 mM sodium pyruvate (Sigma, St. Louis, MO, USA), 100 U/ml penicillinestreptomycin (Biological Industries), 25 mM HEPES (Biological Industries), 2 mg/ml collagenase type IV (Sigma), and 0.15 mg/ml hyaluronidase (Sigma) at 37 C with stirring. Debris of the NP substance was filtered using 100-mm nylon mesh (BD Falcon, Bedford, MA, USA). Trypan blue exclusion dye and an automated cell counter (Countess™, Invitrogen, Paisley, UK) were used to count the viable cells. The NP-derived cells were plated at a density of 1.8 105 cells/cm2 in complete medium containing DMEM with 4.5 g/L D-glucose containing 10% FBS (Biological Industries), 100 U/ml penicillinestreptomycin (Biological Industries), and 2 mM L-glutamine (Biological Industries). The medium was changed after 72 h and then every 3e4 days. Upon confluence, the cells were trypsinized using 0.25% trypsin-EDTA (Invitrogen, Paisley, UK) and replated at a density of 5 103 cells/cm for expansion. The cells were cultured to the 2nd or 3rd passage (P2 and P3, respectively) and used for the assays described in the proceeding sections. Bone marrowederived MSCs (BM-MSCs), which were used as controls in some experiments, were isolated and cultured as previously reported [23,24]. 2.3.1. Three-dimensional culture of NP-SCs in tetronic 1307efibrinogen (TF) NP-SCs were dispersed in TF hydrogel precursor solution, encapsulated after gelation, and subcultured within 3D TF hydrogels for up to 21 days, similar to previously described protocols [22]. Briefly, NP-SCs (at P2 or P3) were seeded in 1-kPa or 2-kPa hydrogels containing 0.1% w/v Irgacure 2959 initiator (Ciba) at a seeding density of 3$106 cells/ml. The hydrogel precursors were prepared fresh for each experiment. Each TF hydrogel (total volume 50 ml) was dispensed in a 5-mmdiameter silicone mold. The hydrogels were incubated for 5 min at 4 C and then cross-linked under long-wave UV light (365 nm, 4e5 mW/cm2, Vilber Lourmat, Marne la Vallee, France) for 5 min. The cross-linked cell-seeded hydrogels were incubated in a humidified incubator (37 C, 5% CO2) in culture medium as described above. Cell culture medium was changed every 3e4 days throughout the experiment. In some experiments, a 2D culture plate was used as a control group. 2.3.2. Histological analysis For the histological analysis, the hydrogels were fixed in 4% formalin, passed through a graded series of ethanol solutions, and embedded in paraffin. Sections (5-mm thick) were cut from each paraffin block by using a motorized microtome (Leica Microsystems, Wetzlar, Germany). For detection purposes the cells were labeled with CM-DiI before inclusion [25]. Slides containing cells labeled with CM-DiI were mounted using GVA Mounting Solution (Invitrogen). CM-DiI stain was captured using fluorescent microscopy. Additional slides were stained with hematoxylin and eosin (H&E). 2.4. Cell proliferation Cell survival and metabolic activity were assessed 7 and 14 days after cell seeding by quantifying the dsDNA content with the aid of the PicoGreen assay (Molecular Probes, Eugene, OR, USA), as previously described [26]. Briefly, 50 ml cell-containing hydrogel samples were homogenized and enzymatically digested. The samples were then diluted in 100 ml buffer (10 mM TriseHCl, 1 mM EDTA, pH 7.5) and the fluorescent PicoGreen dye was added. The samples were analyzed following a 5-min incubation. Protein content was measured using the BCA assay (Pierce, Rockford, IL, USA) according to the manufacturer's protocol. 2.5. Differentiation assays Differentiation of NP-SCs toward the chondrogenic or osteogenic lineage was analyzed using reverse-transcription polymerase chain reaction (RT-PCR) followed by quantitative real-time PCR against pivotal chondrogenic (aggrecan, sox9) or osteogenic (osteocalcin, osteopontin) genes. In addition, biochemical assaysda dimethylmethylene blue (DMMB) assay for sGAGs content and an alkaline phosphatase (ALP) colorimetric assaydwere performed to validate the chondrogenic and osteogenic phenotypic expression of cultured cells.
2.5.1. Chondrogenic differentiation analysis Total RNA was extracted from cells that were seeded in the TF hydrogels (1 kPa, 2 kPa) or in pellet culture as a control. Pellet cultures were prepared as described elsewhere [25]. RNA extraction was performed on Days 1, 7, and 14. To dissolve the harvested hydrogels, we incubated them for 1.5 h in 1 mg/ml Collagenase C7657 (Sigma) and 3 mg/ml Collagenase C2674 solution (Sigma) (for 1-kPa and 2-kPa hydrogels, respectively). RNA was extracted using the RNeasy Mini kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's protocol. RNA was retrotranscribed using random primers and reverse transcriptase (Promega Corp., Madison, WI, USA); Aggrecan and b-actin gene expression were analyzed using quantitative real-time PCR. Quantitative real-time PCR was performed with the aid of the Step One system and program (ABI, Foster City, CA, USA) and analyzed using the relative quantification method. The primers that we used included the TaqMan gene expression assay ACAN-Ss03374825_m1 (ABI) for aggrecan, the TaqMan assay on demand ACTB-Ss03376160_u1 (ABI) for b-actin, and the Ss03392406_m1 (ABI) for sox9. Aggrecan and sox9 gene expression was normalized to the housekeeping gene b-actin and calibrated to the gene expression of cells in the pellet culture on Day 1. Production of sGAGs by NP-SCs was assessed using the dimethylmethylene blue (1,9-DMMB, Sigma, St. Louis, MO, USA) assay, which was modified to suit assessment of sGAGs accumulation in hydrogels. Each sample was suspended in a digestion solution containing 125 mg/ml papain, 100 mM phosphate buffer, 10 mM cysteine, and 10 mM EDTA, pH 6.3 (all from Sigma), and incubated overnight in 65 C. Finally, each 50 ml of digested cells containing hydrogels were homogenized and enzymatically digested. 10 ml of each sample were diluted with 250 ml of DMMB. The optical density was measured with a spectrophotometer at a wavelength of 525 nm. 2.5.2. Osteogenic differentiation analysis Total RNA was extracted, reverse transcribed, and analyzed quantitatively using real-time PCR as described above. The expression of osteocalcin (Ss03373655_s1, BGLAP, ABI) and osteopontin, also known as secreted phosphoprotein 1 (Ss03391322_g1, SPP1, ABI) genes, was normalized to the housekeeping gene b-actin and calibrated to the gene expression of NP-SCs cultured on 2D plates on Day 1. In addition, The ALP activity was determined for the hydrogel treatments. On Days 7 and 14, the hydrogels were digested using collagenase, as described above, and the ALP activity was assessed as previously described [25]. Briefly, 50-ml cell-containing hydrogel samples were homogenized and enzymatically digested. The samples were then diluted into a 400-ml phosphate substrate working reagent, whirled in a vortex device, and incubated for 10 min in 37 C. The optical density was measured at a wavelength of 404 nm. Values were normalized to the sDNA amount, which was measured using the PicoGreen assay. 2.6. Cell morphology analysis The NP-SC morphology was analyzed using immunofluorescence staining against the protein paxillin, a marker for focal adhesion points, and counterstained with fluorescent dyes to detect the cytoskeleton protein actin. To further visualize cellmatrix interactions, samples were imaged using scanning electron microscopy (SEM). 2.6.1. Paxillin and actin immunofluorescence Hydrogels seeded with NP-SCs were harvested on Days 7 and 14 and fixed using 10% formalin solution; afterward they were washed with 0.3% Triton and PBS. Following an overnight incubation with blocking solution (1% BSA), the hydrogels were incubated with purified mouse antiepaxillin primary antibody (dilution 1:250; BD Biosciences Pharmingen, San Diego, CA, USA) for 2 h at room temperature and for an additional 2 h at 4 C. Subsequently, the samples were incubated with secondary antibody for 4 h at room temperature (Jackson ImmunoResearch, West Grove, PA, USA). For counterstains, actin filaments were stained using PhalloidinTRITC (Sigma), and the nuclei were stained with SYTOX blue (dilution 1:2500; Molecular Probes, Eugene, OR, USA) for 30 min. Finally, the hydrogels were washed 3 times in PBS at room temperature, incubated overnight in PBS, and visualized using the Zeiss LSM 700 Laser scanning microscope (Thornwood, NY, USA). Paxillin staining was quantified using MatLab (Mathworks, Natick, MA). The software was used for processing images of cells labeled for actin and paxillin. Paxillin staining was calculated as the ratio of paxillin staining to the cell surface area (as identified by staining for actin). 2.6.2. High-resolution scanning electron microscopy (HR-SEM) HR-SEM was used to ascertain the cellular organization within the hydrogels. Sample preparation for the SEM included cell fixation and dehydration, as previously described [26]. Electron micrographs were obtained using a Sirion highresolution scanning electron microscope (HR-SEM, FEI, Eindhoven, The Netherlands) with the voltage set at 3e5 kV. The samples were fixed using 2.5% w/v glutaraldehyde (Sigma) and 2% w/v paraformaldehyde (PFA, Sigma) in 200 mM cacodylic acid (Merck) and 10 nM CaCl2. After washing, the samples were dehydrated using increasing concentrations of ethanol, followed by drying with a Balzers CPD 030 Critical Point Dryer (CPD). Samples were spatter-coated for 60 s until an approximate 10-nm palladium/gold (Pd/Au) coating was formed and subsequently imaged.
Y. Navaro et al. / Biomaterials 49 (2015) 68e76 2.7. Statistical analysis Data are expressed as means ± standard errors. Significant differences between data sets were found using the Student's t-test (two way); the minimal criterion for significance was set at p < 0.05.
71
on Day 1 (Fig. 1b). Although the stiffness of the 2-kPa matrix was slightly higher than that of the 1-kPa matrix on Day 14, this difference was not statistically significant.
3. Results
3.3. Cell survival enhanced in 1-kPa hydrogels
3.1. Mechanical properties of the hydrogels
Cell survival and proliferation were analyzed using both the PicoGreen assay (which measures double-stranded DNA) and the BCA assay (which measures total protein content) on Days 7 and 14. NP-SCs were found to proliferate in both types of matrices. Cells seeded in the 1-kPa matrix displayed higher proliferation rates than cells seeded in the 2-kPa matrix over time (Fig. 2a). Moreover, the amount of protein in the hydrogels on Day 14 was significantly higher in the 1-kPa matrix than in the 2-kPa matrix (Fig. 2b). Cells were observed on Day 7, expanding throughout the gel space, with low cell-to-cell contact, and a light preponderance of elongated shape in the cells encapsulated inside the 1-kPa matrix (Fig. 2c, DiI staining).
Tetronic®1307 is a synthetic tetrafunctional ethylene oxide/ propylene oxide block copolymer that, much like other members of its poloxameric family (e.g. Pluronic®F127), displays a reversible temperature-induced physical solegel transition and an irreversible UV light–activated chemical cross-linking (photo-polymerization). Conjugating fibrinogen, a biological protein, to Tetronic®1307 produces a new biosynthetic precursor, which retains its reversible temperature-induced transition and photo-polymerization (Fig. 1a) and is also susceptible to protease degradation and consequent cell-mediated remodeling (data not shown). Similar to other protein-polymer conjugates [27], this protein-based material also conveys inductive signals to cells through bioactive sites on the fibrinogen backbone, as well as through other biomechanical properties such as the matrix modulus. A hydrogel was formed from the TF precursor in the presence of NP-SCs by non-toxic free-radical polymerization using light activation. Two different TF materials, each having a shear storage modulus (G0 ) of 1 kPa or 2 kPa, were fabricated (Fig. 1a) to study the effect of matrix stiffness on the proliferation, differentiation, and scaffold remodeling of NP-SCs. 3.2. Ex situ modulus enhanced after 14 days in culture The ex situ modulus of the NP-SCsecontaining scaffold with an initial modulus of 1 kPa (measured by in situ rheometry) increased gradually between Days 1 and 14, resulting in a significant 60% increase in the shear storage modulus of the construct (Fig. 1b, p < 0.05). No difference was detected in the ex situ mechanical properties of the construct containing NP-SCs with an initial modulus of 2 kPa between Day 1 and Day 7; however, by Day 14, the matrix modulus increased significantly (by 40%) compared to that
3.4. Lower modulus induces differentiation of NP-SCs toward a chondrogenic lineage The expression of aggrecan, a pivotal chondrogenic gene in NP-SCs, was higher in cells seeded in hydrogels with the lower shear storage modulus (1 kPa) than in cells cultured in the 2-kPa hydrogels. A comparison of gene expression of cells seeded in the 1-kPa constructs after 7 days in culture revealed a 4-fold increase in aggrecan expression (n ¼ 8) over that measured on Day 1 (Fig. 3a) and significantly higher expression than cells seeded in the 2-kPa hydrogel at the same time point. All results were normalized with aggrecan expression at Day 1 of culture. All groups were cultured under complete medium conditions with no chondrogenic supplements. The evaluation of gene expression of the chondrogenic gene sox9 likewise indicated higher expression in cells seeded in the 1-kPa construct than in cells seeded in the 2-kPa construct (Fig. 3b). The expression level of type-II collagen was also tested, but no significant differences were found between cells seeded in 1-kPa and 2-kPa hydrogels on Days 7 and 14 (data not shown). Nevertheless, the DMMB assay revealed significantly higher levels of
Fig. 1. In situ and ex situ rheological testing of TF materials: Characterization of the shear storage modulus (G0 , Pa) to examine mechanical differences between groups. a) In situ rheological time-sweep measurements show lower and higher shear storage moduli (G0 , Pa) as a function of cross-linking time and temperature. The first stage shows chemical cross-linking at 4 C, and the second stage shows physical cross-linking at 37 C. b) Ex situ shear storage moduli (G0 , Pa) of 1-kPa and 2-kPa TF materials containing NP-SCs at several time points (n ¼ 3; *, **p < 0.05). The classification of treatment groups is based on the value of the modulus obtained for acellular hydrogels using in situ rheology measurements (1 and 2 kPa).
72
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
Fig. 2. Effect of 1- and 2-kPa TF matrices on NP-SC survival and proliferation. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices. Cell survival and proliferation were assessed using two different types of assays. a) DNA quantification: Samples were processed 7 and 14 days after cell seeding, and the concentrations of double-stranded DNA were determined using the PicoGreen assay. Bars indicate SE (n ¼ 7, *p < 0.05). b) Protein content: Samples were processed 7 and 14 days after cell seeding, and protein concentrations were determined using the BCA assay. Protein concentration secreted by the cells was higher in the low-elasticity modulus, as revealed by the BCA assay. Bars indicate SE (n ¼ 3, *p < 0.05). c) Spatial distribution of the cells in the matrices 7 days after cell seeding: Fluorescence images of DiI stained cells and H&E staining of processed gels. Scale bars vary according to the different captions. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
sGAGs accumulation in the 1-kPa group than in the 2-kPa group on Days 7 and 14 (Fig. 3c). Significant differences were also noted when the 1-kPa hydrogel group was compared with the control group (NPeSCs in pellet culture) on Day 14. 3.5. Higher modulus induces differentiation of NP-SCs toward the osteogenic lineage NP-SCs that were cultured in 2-kPa hydrogels without osteogenic supplements for 14 days displayed induction of osteogenic differentiation. At Day 14, osteocalcin expression was 15 times greater in the 2-kPa group than in the 1-kPa group (Fig. 4a). Osteopontin expression was significantly higher in the 2-kPa group than in the 1-kPa group on Day 7 (Fig. 4b). Consistently, ALP activity in NP-SCs cultured in the 2-kPa hydrogels was significantly higher
than that in cells cultured in the 1-kPa hydrogels on Day 7; this ALP activity decreased in both groups by Day 14 (Fig. 4c). 3.6. Variations in the matrix modulus affect NP-SCs morphology and cell-matrix interactions Confocal microscopy and SEM (Fig. 5a) were used to analyze cell survival and cell-matrix interactions. Both types of hydrogels provided the cells with a favorable environment for appropriate cell attachment and growth. The 1-kPa hydrogels were characterized by a rough topography, containing more surface area and thus promoting increased adhesion and anchoring. These cell-matrix contacts promoted a stretched and long, spindle-shaped fibroblast-like morphology, as evidenced by actin-stained cells (Fig. 5b). In contrast, NP-SCs seeded in 2-kPa hydrogels exhibited a smoother
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
73
Fig. 3. Effect of 1-kPa and 2-kPa TF matrices on induction of chondrogenesis with NP-SCs. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices. Chondrogenesis was assessed using three independent assays on Days 1, 7, and 14; and the results were always normalized to data from Day 1. a) Aggrecan gene expression (by quantitative PCR): Samples were processed 1, 7, and 14 days after cell seeding. Chondrogenic gene expressiondAggrecandwas measured by quantitative real-time PCR. Cells subcultured in 1-kPa hydrogels revealed a higher rate of Aggrecan gene expression. Bars indicate SE (n ¼ 8, *p < 0.05). b) Chondrogenic gene expressiondSox9dwas measured by quantitative real-time PCR. Cells subcultured in 1-kPa hydrogels revealed a higher rate of Sox9 gene expression. Bars indicate SE (n ¼ 8, *p < 0.05). c) Sulfated glycosaminoglycans (sGAGs) content was quantified by performing a dimethylmethylene blue (DMMB) assay. Samples were processed on Days 1 and 14 after cell seeding, and sGAGs accumulation in the hydrogels was measured. Bars indicate SE (n ¼ 3, *,**p < 0.05).
topography, with cells tightly surrounded by the gel (Fig. 5a). Those cells displayed a spherical morphology with less stretched out actin fibers (Fig. 5b). The SEM cell morphology observations were in complete agreement with the confocal microscopy images. Quantitative analysis of focal adhesions using paxillin staining at Day 14 (Fig. 5b) revealed more paxillin stain per cell in cells cultured in 1-kPa hydrogels than in cells in the 2-kPa group; this could indicate that NP-SCs attach better to the softer matrix. 4. Discussion Previous studies have shown that matrix stiffness affects the fate of BM-MSCs [11e13] as well as mature NP cell organization, function, and morphology [5,11,12]. Survival, genotype, and phenotypic expression of MSCs were all shown to be affected by the modulus of the cellular microenvironment [5,10e12,28]. In this study, we tested the effect of matrix stiffness on NP-SCs by using a Tetronic-fibrinogen (TF) hydrogel that allowed us to control the mechanical properties of the cell-encapsulating biomaterial while keeping other material properties, such as ligand density (i.e. fibrinogen concentration), unchanged. Using various assays we analyzed the effect of the matrix modulus on the survival and differentiation of NP-SCs. Our results reveal that lower-modulus hydrogelsd1 kPa, similar to the physiological modulus found in
native IVDdpromoted cell survival and chondrogenic differentiation, as well as enhanced cell attachment to the matrix. Unlike previous studies in which BM-MSCs were used [11,12,28], here we investigated MSCs isolated from lumbar IVDs. The porcine IVD used in this study is known to share common characteristics with the human IVD [29e31]. Therefore, primary cells isolated from a porcine IVD closely resemble human NP-derived SCs. Consequently, the elastic shear modulus of the human NP substance was widely investigated and reported to be in the range of 0.1 kPae1 kPa [4e6,17]. This variability in the elastic modulus of IVD is mainly due to the different test types that have been used. Here, we used the low modulus matrix (1-kPa group) to mimic the mechanical properties of the healthy NP and a slightly stiffer matrix (2-kPa group), which represents a more cross-linked version of the same material. We used these two groups (compliant and stiff, respectively) to investigate the cell response to matrix elasticity, while keeping other properties, such as fibrinogen concentration, unchanged. In this context, the TF biomaterial provided us with the ability to control cross-linking density without changing the fibrinogen composition by adding small amounts of additional cross-linker to the hydrogel precursor solution (Fig. 1a). The NP-SCs were cultured within these two types of hydrogels. Assessments were made based on cell differentiation, proliferation, and viability; and the results were correlated to changes in the initial material
74
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
Fig. 4. Effect of 1- and 2-kPa TF matrices on induction of osteogenesis with NP-SCs. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices, and gene expression was measured by quantitative real-time PCR for a) Osteocalcin and b) Osteopontin 7 and 14 days after cell seeding. The gene expression rate was higher in cells seeded in 2-kPa hydrogels. Bars indicate SE (n ¼ 5, *p < 0.05). c) Alkaline phosphatase (ALP) (colorimetric assay) activity was determined 7 and 14 days after cell seeding.
modulus. Our data show strong correlations between the initial modulus of the 3D culture matrix and the biological response of the NP-SCs after several days in culture. Regarding the ex situ shear storage modulus, it was evident that the mechanical properties of both the compliant and stiff hydrogels increased in response to the NP-SC culture. We speculate that a higher increase in the modulus of the compliant material was the result of enhanced modulus-dependent cell remodeling (and proliferation), leading to heightened ECM production in these constructs. Dikovsky et al. [32] showed that MSC spreading in PEGfibrinogen 3D matrices occurred more rapidly when the storage modulus was low. Moreover, Kesselman et al. [22] showed that fibroblasts cultured within soft 3D PEG-fibrinogen hydrogels produced ECM and significantly increased their matrix modulus after as little as 7 days in culture. In the current study, we observed that the gradual increase in the modulus of the compliant (1-kPa) hydrogel between Days 1 and 14 and the increase in the modulus of the 2-kPa scaffold observed between Days 7 and 14 were consistent with higher proliferation rates and improved spindled cell morphologies (Fig. 2a and Fig. 5b for the 1-kPa matrix) as well as possibly associated with protein secretion (Fig. 2b). NP-SCs seeded in 1-kPa matrices also exhibited a higher rate of chondrogenic differentiation (Fig. 3aec). Aggrecan deposition is considered a hallmark of chondrogenesis and comprises a major part of the IVD's ECM [33]. Aggrecan and sox9 gene expression and sGAGs accumulation were significantly higher in cells cultured in
the soft 1-kPa hydrogels compared with cells cultured in the stiffer 2-kPa hydrogels (Fig. 3aec). The increase in accumulation of sGAGs in cells cultured in 1-kPa hydrogels (Fig. 3c) is consistent with reports from previous studies in which soft matrices were used [12,13]. However, in this study we cultured the cells in 3D matrices, which better simulate in vivo conditions than 2D cell cultures [5]. The underlying mechanism that regulates induction of chondrogenesis in NP-SCs in 1-kPa hydrogels is unknown and needs to be further investigated. However, the mechano-responsiveness of the cells to the ECM plays a major role in the induction of chondrogenesis [34]. Moreover, both the high availability of cellecell interactions and the low oxygen tension inside the hydrogel could also promote chondrogenesis. NP-SCs seeded in 2-kPa matrices showed increased expression of osteogenic differentiation markers (Fig. 4aec). Expression of osteocalcin (a late osteogenic gene marker) was significantly higher in cells cultured in 2-kPa hydrogels than in cells cultured in 1-kPa hydrogels (Fig. 4a). Additionally, expression of osteopontin (a midto-late osteogenic marker) was significantly higher in cells cultured in 2-kPa hydrogels after 1 week of culture compared with cells cultured in 1-kPa matrices (Fig. 4b). These results correlate with those of previous studies using human BM-MSCs and human osteoblasts, which also indicates that a cell fate toward the osteogenic lineage is favored when more rigid substrates are used [14]. Moreover, ALP activity (Fig. 4c) showed a marked elevation in cells cultured in 2-kPa hydrogels compared with cells cultured in 1-kPa
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
75
Fig. 5. Effect of 1- and 2-kPa TF matrices on NP-SC morphology and adhesion to the matrix. NP-SCs were seeded at a density of 3$106 cells/ml in 3D TF matrices. a) Scanning electron microscopy (SEM) reveals the cell morphology of the samples 7 and 14 days after cell seeding. b) Confocal microscopy of NP-SCs 14 days after seeding, using fluorescent staining of actin filaments, corroborated the results of the SEM images by showing that cells grown in 1-kPa hydrogels are much more spread out than cells grown in 2-kPa hydrogels. Paxillin protein can be seen labeled in red, actin in green, and cell nuclei in blue. c) Paxillin quantification: Samples were processed 7 and 14 days after cell seeding, and the percentage of paxillin was quantified relative to the cell surface area. Bars indicate SE (n ¼ 6, *p < 0.05). (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
matrices after 7 days. The mechanism by which NP-SCs are genetically and phenotypically induced toward osteogenesis is still uncertain, but some possibilities include ECM stiffness, cell morphology, and other interactions with the ECM, all of which can explain cell phenotype in the unique environment of stiffer IVD tissues such as the annulus fibrosus [4,6,17]. In similar fashion, we speculate that changes in the NP matrix stiffness caused by a degenerative process may inhibit the differentiation of NP-SCs and prevent tissue regeneration within the disc. Variations in matrix stiffness also revealed a profound effect on NP-SC morphology. NP-SCs’ adherence to the fibrinogen domains [35] of the hydrogels showed different degrees of attachment and very different cellular morphologies when cultured in 1-kPa and 2-kPa matrices (Fig. 5a and b). Although it is commonly accepted that osteoblastic cells present spindle morphology while chondrocytic cells tend to be more rounded, here we show that cells seeded in 1-kPa matrices and differentiated to chondrogenic cells, exhibited a spindledeshape morphology, unlike cells seeded in 2-kPa matrices, which were mostly spherical and expressed osteogenic markers of differentiation. Moreover, focal adhesion points, visualized by paxillin staining, were more abundant in cells that were cultured in 1-kPa hydrogels with significant differences observed on Day 14. It is important to indicate that the hydrogel used in this study is not composed of the ECM molecules that are found in the NP, bone or cartilage and therefore we could not expect similar cell morphology as seen in vivo. It is also possible that the lower elastic modulus matrix facilitated better cellecell and
cell-matrix interactions, while cells seeded in stiffer matrices were less able to move around and create cellecell interactions due to the higher stiffness of the encapsulating microenvironment. Another possibility is that the rough topography of the 1-kPa matrix, as demonstrated by SEM images, was due to the exposure of fibrinogen domains that allowed cells to adhere and spread, and thus to acquire their more spindled shape. Interestingly, these finding are in the same line with recently published results from Kim et al. [36]. Authors of previous studies reported that variation in the cellular response to matrix stiffness in different cell types is receptor dependent [37]. Moreover, the use of 2D or 3D matrices has a profound effect on cell morphology, adhesion, and spreading [37e39]. It has also been suggested that cells exert contractile forces and “interpret” substrate deformation to determine a preferred direction for their movements [39e41]. Our findings of NP-SC behavior in the TF hydrogels are in agreement with the results of other studies, revealing the effect of the matrix modulus on the fate of MSCs [5,7,11,12,14,28,42]. However, additional studies are still required to better elucidate the exact cellular mechanism that is responsible for the observed cell behavior. We envision that future treatments for degenerated IVD pathology could potentially utilize TF hydrogels in combination with NP-SCs. The material properties of the TF hydrogels can be fine-tuned to achieve modulus values that mimic the native IVD microenvironment. Accordingly, we have shown that when NP-SCs are cultured in hydrogels having such a modulus, it can help facilitate the proliferation of these cells and direct them to commit
76
Y. Navaro et al. / Biomaterials 49 (2015) 68e76
to a specific and desirable lineage. Hence, our results may pave the way to new tissue-engineering strategies for NP regeneration. 5. Conclusions Our data indicate that NP-SCs proliferate in TF hydrogels. Moreover, we conclude that NP-SCs cultured within hydrogels having a lower (1-kPa) elastic modulus were promoted toward a chondrogenic differentiation pathway, as evidenced by quantitative real-time PCR and the DMMB assay, whereas cells cultured in higher modulus (2-kPa) hydrogels were promoted toward an osteogenic differentiation pathway. These results highlight the effect of matrix stiffness on the fate of NP-SCs. This research may pave the way for a better understanding of the complex cross-talk of cells and their matrix during disc degeneration, as well as present novel therapeutic strategies for IVD regeneration. Acknowledgments We gratefully acknowledge funding support from the National Institutes of Health (RO3 AR057143, Z.G.), Israel Science Foundation grant (no. 614/09, D.G.) and the Sir Zelmen Cowen Universities Fund Blue Sky Research Grant (D.G.). References [1] Manchikanti L, Singh V, Falco FJ, Benyamin RM, Hirsch JA. Epidemiology of low back pain in adults. Neuromodulation e J Int Neuromodulation Soc 2014;17(Suppl. 2):3e10. [2] Hsieh AH, Twomey JD. Cellular mechanobiology of the intervertebral disc: new directions and approaches. J Biomech 2010;43:137e45. [3] Niosi CA, Oxland TR. Degenerative mechanics of the lumbar spine. Spine J : Off J North Am Spine Soc 2004;4:202Se8S. [4] Neidlinger-Wilke C, Galbusera F, Pratsinis H, Mavrogonatou E, Mietsch A, Kletsas D, et al. Mechanical loading of the intervertebral disc: from the macroscopic to the cellular level. Eur Spine J e Off Publ Eur Spine Soc Eur Spinal Deformity Soc Eur Sect Cerv Spine Res Soc 2014;23(Suppl. 3):S333e43. [5] Gilchrist CL, Darling EM, Chen J, Setton LA. Extracellular matrix ligand and stiffness modulate immature nucleus pulposus cell-cell interactions. PLoS One 2011;6:e27170. [6] Setton LA, Chen J. Cell mechanics and mechanobiology in the intervertebral disc. Spine (Phila Pa 1976) 2004;29:2710e23. [7] Blanco JF, Graciani IF, Sanchez-Guijo FM, Muntion S, Hernandez-Campo P, Santamaria C, et al. Isolation and characterization of mesenchymal stromal cells from human degenerated nucleus pulposus: comparison with bone marrow mesenchymal stromal cells from the same subjects. Spine (Phila Pa 1976) 2010;35:2259e65. [8] Henriksson H, Thornemo M, Karlsson C, Hagg O, Junevik K, Lindahl A, et al. Identification of cell proliferation zones, progenitor cells and a potential stem cell niche in the intervertebral disc region: a study in four species. Spine (Phila Pa 1976) 2009;34:2278e87. [9] Mizrahi O, Sheyn D, Tawackoli W, Ben-David S, Su S, Li N, et al. Nucleus pulposus degeneration alters properties of resident progenitor cells. Spine J e Off J North Am Spine Soc 2013. [10] Hwang PY, Setton LA, Chen J, Jing L, Hoffman BD. The role of extracellular matrix elasticity and composition in regulating the nucleus pulposus cell phenotype in the intervertebral disc. J Biomech Eng 2014. [11] Huebsch N, Arany PR, Mao AS, Shvartsman D, Ali OA, Bencherif SA, et al. Harnessing traction-mediated manipulation of the cell/matrix interface to control stem-cell fate. Nat Mater 2010;9:518e26. [12] Guilak F, Cohen DM, Estes BT, Gimble JM, Liedtke W, Chen CS. Control of stem cell fate by physical interactions with the extracellular matrix. Cell Stem Cell 2009;5:17e26. [13] Seliktar D. Designing cell-compatible hydrogels for biomedical applications. Science 2012;336:1124e8. [14] Engler AJ, Sweeney HL, Discher DE, Schwarzbauer JE. Extracellular matrix elasticity directs stem cell differentiation. J Musculoskelet Neuronal Interact 2007;7:335.
[15] Choi JS, Harley BA. The combined influence of substrate elasticity and ligand density on the viability and biophysical properties of hematopoietic stem and progenitor cells. Biomaterials 2012;33:4460e8. [16] Ishaug-Riley SL, Crane-Kruger GM, Yaszemski MJ, Mikos AG. Three-dimensional culture of rat calvarial osteoblasts in porous biodegradable polymers. Biomaterials 1998;19:1405e12. [17] Setton LA, Chen J. Mechanobiology of the intervertebral disc and relevance to disc degeneration. J Bone Jt Surg Am 2006;88(Suppl. 2):52e7. [18] Knight E, Przyborski S. Advances in 3D cell culture technologies enabling tissue-like structures to be created in vitro. J Anat 2014. [19] Halstenberg S, Panitch A, Rizzi S, Hall H, Hubbell JA. Biologically engineered protein-graft-poly(ethylene glycol) hydrogels: a cell adhesive and plasmindegradable biosynthetic material for tissue repair. Biomacromolecules 2002;3:710e23. [20] Dikovsky D, Bianco-Peled H, Seliktar D. The effect of structural alterations of PEG-fibrinogen hydrogel scaffolds on 3-D cellular morphology and cellular migration. Biomaterials 2006;27:1496e506. [21] Plotkin M, Vaibavi SR, Rufaihah AJ, Nithya V, Wang J, Shachaf Y, et al. The effect of matrix stiffness of injectable hydrogels on the preservation of cardiac function after a heart attack. Biomaterials 2014;35:1429e38. [22] Kesselman D, Kossover O, Mironi-Harpaz I, Seliktar D. Time-dependent cellular morphogenesis and matrix stiffening in proteolytically responsive hydrogels. Acta Biomater 2013. [23] Aslan H, Zilberman Y, Arbeli V, Sheyn D, Matan Y, Liebergall M, et al. Nucleofection-based ex vivo nonviral gene delivery to human stem cells as a platform for tissue regeneration. Tissue Eng 2006;12:877e89. [24] Bosch P, Pratt SL, Stice SL. Isolation, characterization, gene modification, and nuclear reprogramming of porcine mesenchymal stem cells. Biol Reprod 2006;74:46e57. [25] Aslan H, Zilberman Y, Kandel L, Liebergall M, Oskouian RJ, Gazit D, et al. Osteogenic differentiation of noncultured immunoisolated bone marrowderived CD105þ cells. Stem Cells 2006;24:1728e37. [26] Garty S, Kimelman-Bleich N, Hayouka Z, Cohn D, Friedler A, Pelled G, et al. Peptide-modified “smart” hydrogels and genetically engineered stem cells for skeletal tissue engineering. Biomacromolecules 2010;11:1516e26. [27] Shachaf Y, Gonen-Wadmany M, Seliktar D. The biocompatibility of PluronicF127 fibrinogen-based hydrogels. Biomaterials 2010;31:2836e47. [28] Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix elasticity directs stem cell lineage specification. Cell 2006;126:677e89. [29] Mosekilde L, Kragstrup J, Richards A. Compressive strength, ash weight, and volume of vertebral trabecular bone in experimental fluorosis in pigs. Calcif Tissue Int 1987;40:318e22. [30] Heino TJ, Alm JJ, Moritz N, Aro HT. Comparison of the osteogenic capacity of minipig and human bone marrow-derived mesenchymal stem cells. J Orthop Res 2012;30:1019e25. [31] Lundin O, Ekstrom L, Hellstrom M, Holm S, Sward L. Exposure of the porcine spine to mechanical compression: differences in injury pattern between adolescents and adults. Eur Spine J 2000;9:466e71. [32] Dikovsky D, Bianco-Peled H, Seliktar D. Defining the role of matrix compliance and proteolysis in three-dimensional cell spreading and remodeling. Biophys J 2008;94:2914e25. [33] Hidaka C, Quitoriano M, Warren RF, Crystal RG. Enhanced matrix synthesis and in vitro formation of cartilage-like tissue by genetically modified chondrocytes expressing BMP-7. J Orthop Res 2001;19:751e8. [34] Cassinelli EH, Hall RA, Kang JD. Biochemistry of intervertebral disc degeneration and the potential for gene therapy applications. Spine J 2001;1:205e14. [35] Almany L, Seliktar D. Biosynthetic hydrogel scaffolds made from fibrinogen and polyethylene glycol for 3D cell cultures. Biomaterials 2005;26:2467e77. [36] Kim TH, An DB, Oh SH, Kang MK, Song HH, Lee JH. Creating stiffness gradient polyvinyl alcohol hydrogel using a simple gradual freezingethawing method to investigate stem cell differentiation behaviors. Biomaterials 2015;40: 51e60. [37] Yeung T, Georges PC, Flanagan LA, Marg B, Ortiz M, Funaki M, et al. Effects of substrate stiffness on cell morphology, cytoskeletal structure, and adhesion. Cell Motil Cytoskelet 2005;60:24e34. [38] Bard JB, Hay ED. The behavior of fibroblasts from the developing avian cornea. Morphology and movement in situ and in vitro. J Cell Biol 1975;67:400e18. [39] Lo C-M, Wang H-B, Dembo M, Wang Y-l. Cell movement is guided by the rigidity of the substrate. Biophys J 2000;79:144e52. [40] Pelham Jr RJ, Wang Y. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc Natl Acad Sci U S A 1997;94:13661e5. [41] Sheetz MP, Felsenfeld DP, Galbraith CG. Cell migration: regulation of force on extracellular-matrix-integrin complexes. Trends Cell Biol 1998;8:51e4. [42] Kong HJ, Polte TR, Alsberg E, Mooney DJ. FRET measurements of cell-traction forces and nano-scale clustering of adhesion ligands varied by substrate stiffness. Proc Natl Acad Sci U S A 2005;102:4300e5.