ZIF-8 Modified Polypropylene Membrane: A Biomimetic Cell Culture Platform with a View to the Improvement of Guided Bone Regeneration

Abstract

Purpose

Despite the significant advances in modeling of biomechanical aspects of cell microenvironment, it remains a major challenge to precisely mimic the physiological condition of the particular cell niche. Here, the metal–organic frameworks (MOFs) have been introduced as a feasible platform for multifactorial control of cell-substrate interaction, given the wide range of physical and mechanical properties of MOF materials and their structural flexibility.

Results

In situ crystallization of zeolitic imidazolate framework-8 (ZIF-8) on the polydopamine (PDA)-modified membrane significantly raised surface energy, wettability, roughness, and stiffness of the substrate. This modulation led to an almost twofold increment in the primary attachment of dental pulp stem cells (DPSCs) compare to conventional plastic culture dishes. The findings indicate that polypropylene (PP) membrane modified by PDA/ZIF-8 coating effectively supports the growth and proliferation of DPSCs at a substantial rate. Further analysis also displayed the exaggerated multilineage differentiation of DPSCs with amplified level of autocrine cell fate determination signals, like BSP1, BMP2, PPARG, FABP4, ACAN, and COL2A. Notably, osteogenic markers were dramatically overexpressed (more than 100-folds rather than tissue culture plate) in response to biomechanical characteristics of the ZIF-8 layer.

Conclusion

Hence, surface modification of cell culture platforms with MOF nanostructures proposed as a powerful nanomedical approach for selectively guiding stem cells for tissue regeneration. In particular, PP/PDA/ZIF-8 membrane presented ideal characteristics for using as a barrier membrane for guided bone regeneration (GBR) in periodontal tissue engineering.

Introduction

Since the first cell clinical trial on adult mesenchymal stem cells (MSCs) in 1995, these cells have been applied to a wide array of defects, primarily including neurodegenerative diseases,¹ eye disorders,² cardiovascular diseases,³ cartilage and intervertebral disc destruction,⁴ bone loss,^5,6 autoimmune disorders,⁷ and oral and maxillofacial reconstruction.⁸ Dental-related stem cells (like dental pulp stem cells (DPSCs), stem cells from exfoliated deciduous teeth (SHEDs), stem cells from apical papilla (SCAPs) and periodontal ligament stem cells (PDLSCs)) are considered as promising stem cell sources, thanks to their accessibility, noninvasive harvesting, neural crest origin, and exceptional plasticity.⁹ Human DPSCs are reported to have a capability to contribute to the regeneration of a large array of tissues ranging from oral and maxillofacial defects,¹⁰ to corneal disease,¹¹ and pancreatic islet.¹² Despite remarkable progress toward the biomedical application of stem cells,^13,14 isolation, expansion, and differentiation of MSCs under clinical conditions are still challenging.¹⁵

Conventional regenerative procedures are typically associated with the use of biological components that causes serious concerns for translating basic scientific reports from the bench to patient care. Predominantly, batch-to-batch variability, xenogenic contaminants, and costly manufacturing on a large scale currently present challenges to the use of existing regenerative protocols (as reviewed by Adam D. et al¹⁶ and Tan and Barker).¹⁷ Moreover, typical tissue engineering strategies rely on the use of proper scaffolds to support critical behavior of the cells and provide essential mechanical parameters of natural tissues.

Accordingly, there has recently been a tendency toward the regulation of physical and biological behaviors of stem cells by controlling the properties of the substrate materials that interact with the cells.^18–20 In contrast to conventional plastic or glass culture substrates, tailored biomaterials present the ability to mimic different physical and mechanical properties of tissue microenvironments.^21–23 They can directly/indirectly activate different transcriptional programs and trigger specific responses in associated cells.²⁴ In fact, this occurs in a dynamic interaction between cells and the substrate through which cells sense substrate signals and redefine their surrounding environment overtime via secretion of a wide range of matrix components and growth factors.^25–27

Therefore, a broad range of biomaterials has been exploited, which commonly include natural or synthetic hydrogels,²⁸ electrospun polymeric fibers,²⁹ and patterned plastic or glass-based substrates. In particular, paper-based cell culture platforms have become increasingly popular for in vitro expansion of cells as well as biomedical applications, like designing anti-biofouling membranes,³⁰ biosensors,^31,32 tissue regeneration, drug screening, and disease modeling.³³ The biocompatibility, porous and flexible structure, low-cost, and easy high-throughput manufacturing make paper scaffolds the ideal subjects for construction of complex tissues. For instance, resorbable/non-resorbable bioactive barrier membranes are extensively utilized in guided bone regeneration (GBR), for preventing invasion of non-osteogenic cells to defect site.^34,35 Some recent clinical and preclinical studies reported reasonable improvement in cell adhesion, proliferation and osteopromotion by use of polypropylene (PP) membrane for alveolar defects.^36–39 The PP membrane is an impermeable porous platform with high rigidity and elastic memory.⁴⁰ Owing to its hydrophobic nature, PP membrane could provide isolated spaces for regeneration of damaged tissues through minimizing infiltration of cells and biomolecules. However, the pristine PP membrane is not favorable for adhesion, organization, and proper activity of proteins/cells, underline the importance of surface modification approaches for expanding its biomedical application.^41,42

Besides the inherent chemical properties of biomaterials, various physical and mechanical parameters, such as stiffness,^43–45 topography,^46–48 viscosity,⁴⁹ density,^33,50 cellular internalization,⁵¹ degradation rate,⁵² and biomaterial groups^53–55 have been applied to control stem cell behaviors including viability, proliferation, motility, spreading and differentiation capacity (extensively reviewed in Ref).^20,56–58 For example, it is now clear that the gradient of substrate stiffness in physiological condition effectively regulates the essential activity of cells and has a profound influence on their fate.^28,59–61 Moreover, micro- and nano-scale roughness can directly affect the cellular behavior of MSCs, such that various structural features of the materials regulate the proliferation or differentiation potential of the cells.^62–64 In addition, the level of surface hydrophilicity selectively controls the tendency of cells to attach and spread on the substrate, which alters cellular activities by interfering with integrin-related signaling pathways.^65–67

It is of paramount importance to note that the existing approaches usually target cell microenvironment via altering a single key factor, rather than reconstitution of a whole-cell niche (reviewed by Tewary et al).⁶⁸ Although some more recent studies gained higher efficiency and more specificity via a combination of two distinct approaches,^69–72 there is still a big gap to comprehensive mimicry of the natural physical and mechanical condition of cells.

Metal-organic frameworks (MOFs) have been used in wide-ranging applications, such as gas storage and separation,^73–76 molecular sieving,^77,78 sensing,^79–82 and catalysis.^83–85 Because of their impressive capacities, MOFs have emerged as a promising material type for various biomedical applications, especially in drug delivery and biosensing platforms (reviewed in Ref).35,^86,87 Moreover, recent progress in designing and manufacturing of MOF nanostructures leading to their novel biomedical applications, such as osteogenesis promotion,^88–90 photosensitizer,⁹¹ biomolecule vehicle,⁹² Intracellular sensing,^93,94 bioimaging,⁹⁵ and biocatalysis.⁹⁶ They are highly ordered porous materials composed of metal coordination centers connected by organic linkers.⁹⁷ Among them, zeolitic imidazolate frameworks (ZIFs) have attracted intense interest due to their impressive stability along with their high porosity and surface area. They are generally composed of tetrahedrally coordinated metal ions (M= Zn²⁺, Co²⁺, Cd²⁺, Mg²⁺, etc.) and imidazolate derivatives (Im).

MOFs are porous structures, which typically have exhibited considerable stability in physiological conditions and possess a good potential for in situ- and post-synthetic functionalization with certain biomolecules, either on metals or organic ligands.^98–100 Namely, Immobilization of enzyme-MOF composites on porous and flexible membranes introduces an effective approach for designing bioactive substrates.^101,102 Moreover, MOFs serve as an ideal host for adsorption and conjunction of short peptides, antibodies, and nucleic acids (reviewed by Kempahanumakkagari et al).¹⁰³ These merits make the MOF layers well suitable for supporting the cellular behaviors through both in vitro and natural conditions.

Additionally, using various metal groups and numerous organic molecules, MOFs form a wide range of structures with a highly tunable pore size (usually 0.4–6 nm) and surface area (500–4500 m²/g).^104,105 Different configuration, porosity, and functionality of MOFs can be achieved by controlling framework interpenetration.106 Also, size of particles could be precisely adjusted by controlling effective parameters in synthesis procedures or applying different mechanisms.¹⁰⁷,108 They are composed of a permanent pore space which can adsorb guest molecules via changing in unit cell volume or subnetwork displacement. For example, ZIF-8 can absorb larger guest molecules by linker rotation109 or present modified porous nature under reasonable pressure.⁵¹ These intriguing chemical and structural properties of MOFs make these materials an ideal candidate for mimicking the physical and mechanical characteristics of natural tissue.

Although various strategies have been reported for fabrication of MOF thin layers, in situ crystallization and growth on chemically modified supports is of great interest.110 Polydopamine (PDA) is a mussel-inspired adhesive polymer, which gives numerous unshared electron pairs for secondary reactions.111 Recently, PDA-based sticky platforms have been used to attach MOF nutrients on a wide range of organic and inorganic supporting materials.⁸²,112

Although MOFs are increasingly set to become useful tools in the biomedical area, their potential to form cell culture substrates is not fully explored yet. In consequence, this paper takes a new look at the capability of the ZIF-8 thin layer to support essential activities of dental pulp stem cells (DPSCs) under in vitro condition. Following the characterization of specific mechanical properties of the PP/PDA/ZIF-8 membrane, we evaluated primary cellular attachment, proliferation rate, and multilineage differentiation of cells on this platform. Fundamental behaviors of DPSCs were compared to tissue culture plates (TCP), as the gold standard for cell culture substrates. The ZIF-8 nano structure point to new opportunities for surface functionalization of PP membranes as a nanomedicine with significant advantages for using in GBR therapies.

Materials and Methods

Preparation and Characterization of Modified Membranes

A thin film layer of ZIF-8 was fabricated on a PP substrate, as we documented earlier.⁸² In brief, the membranes were cleaned carefully with absolute ethanol and immersed in a fresh solution of polydopamine/poly(ethylene imine) (PDA/PEI, 2mg/mL of each) for 4 hours. Following rinsing with deionized water, the samples were incubated in an aqueous solution of 2.74 mg/mL of Zn(NO3)₂ and 56.6 mg/mL of 2-methylimidazole for 2 hours (Scheme 1).

Scheme 1.

Schematic presentation of substrate preparation workflow.

The physical and mechanical properties of the ZIF-8 modified substrates then were assessed in comparison to PDA/PEI coated and intact PP membranes, to characterize the substrates. For this, the surface chemistry of the ZIF-8 layer was evaluated using a Fourier transform infrared spectrometer (Alpha II FT-IR, Bruker, Massachusetts, USA) at a resolution of 4 cm⁻¹ in the range of 4000–400 cm⁻¹. X-ray diffraction (XRD) was done using Empyrean Thin-Film XRD Xpert Materials Research Diffractometer (MRD) with the 2-theta method (range 5^ͦ −40^ͦ, step size 0.025^ͦ, 100 s/step) was used to study the crystalline structure of ZIF8 thin layer. High Score Plus software was applied for XRD analysis, and peaks were compared with ZIF-8 Database peaks from Powder Diffraction File 4 (PDF-4) organic database (reference ID: 00–062-1030).113 The morphological and topological characteristics of membranes were also analyzed by scanning electron microscopy (SEM, Zeiss EVO 15LS, Germany). Analysis of surface topography of coated and uncoated specimens was performed in contact mode by using Atomic force microscopy (AFM, JPK NanoWizard 2, Germany).

Furthermore, the static water contact angle was assessed using a Theta optical tensiometer (Biolin Scientific, Sweden) paired with OneAttention software. The surface free energy (SFE) of the membranes was determined by measuring the contact angles of three types of liquids (ie, water, glycerol, di-iodomethane) according to the acid-base Van Oss method. The elastic modulus (stiffness) of the substrates was determined using Bruker Dimension ICON SPM (Bruker Corporation, USA) equipped with an OTESPA-R3 cantilever operated in the PeakForce Tapping mode. The scan size was set to 150 nm with a scan rate of 0.4 Hz, and the resolution of the images was set to 256 samples/line.

Culture and Expansion of DPSCs

Human dental pulp stem cells were obtained and characterized from extracted third molar teeth, as previously described.114 The experiments were performed under approval of the Ethics Committee of the Royan Institute (NO. IR.ACECR.ROYAN.REC.1397.290), in accordance with the Declaration of Helsinki. Written informed consent was obtained from all volunteers to participate in the study. Isolated cells were cultured and expanded under standard culture condition, using DMEM medium supplemented with 10% fetal bovine serum (FBS), 1% glutamax, and 1% penicillin/streptomycin (all from Gibco, Paisley, UK). For further experiments, DPSCs (passages 4–6) were seeded with 2.5x10⁴/cm² density on substrates, following UV sterilization. Membranes were pre-incubated in complete culture media for overnight, before cell seeding.

Cell Metabolic Activity Assay

To evaluate the viability and proliferation of DPSCs on substrates, the MTS assay was carried out applying cell titer 96 aqueous one solution (Promega, WI, USA) after 8 hours, 1, 3, 5 and 7 days, according to the manufacturer’s instruction. Briefly, cultured cells at each time point were incubated with MTS/PMS solution for 3.5 hours. Therefore, the absorbance of produced formazan crystals was measured at 450 nm using a microplate reader (Fluostar Optima, BMG Lab Technologies, Germany) and normalized to the cell-free solution.

Cell Viability Assay

Carboxyfluorescein diacetate succinimidyl ester (CFSE; Sigma, Munich, Germany) was used to evaluate the viability of cultured cells on the ZIF-8 substrate. After one day, the attached cells were incubated with serum-reduced (1% FBS) medium contained five µM CFSE for 30 minutes. Next, the esterase reaction was quenched by adding serum-supplemented (10%) medium, and dead cells were stained by five µg/mL propidium iodide (PI) for one minute. Imaging by fluorescence microscopy (Olympus, BX51, Japan) showed live DPSCs as green CFSE labeled cells, while dead cells were counterstained with PI.

Morphological and Cytoskeletal Analysis of DPSCs

Cultured substrates were examined by SEM on day 1, to monitor the morphological features of DPSCs. Before imaging, cells were fixed in 2.5% glutaraldehyde and dehydrated with increasing concentration of ethanol.

To visualize the cells’ cytoskeleton, F-actin microfilaments were stained at day one and seven post-seeding, by using phalloidin-TRITC; following fixation with 4% paraformaldehyde and permeabilization by 0.2% Triton X-100 (all from Sigma, Munich, Germany). Cell nuclei were counterstained with DAPI (Sigma, Munich, Germany), and monitored under fluorescence microscopy (Olympus, BX51, Japan). The number of cells was quantified by counting the DAPI-stained nuclei from five random low-magnification fields. Cell spread area was also measured in images by applying Image J software (Version 1.42q, National Institutes of Health, USA).

Evaluation of Differentiation Potential

To gain more insights on the ability of DPSCs to differentiate into different lineages on the substrate of ZIF-8, a confluent monolayer of cells was maintained with specific induction media. Briefly, osteogenesis was promoted by facing cells with normal culture medium containing 50 mg/mL ascorbic acid, 10 mM β-glycerophosphate, and 10 nM dexamethasone. Adipogenic induction was also carried out using growth medium supplemented with 100 nM dexamethasone, 50 µg/mL β-glycerophosphate, and 50 µg/mL indomethacin. Further, chondrogenesis was induced via DMEM medium enriched with 10 ng/mL TGFB1, 100 nM dexamethasone, 50 µg/mL ascorbic acid, 100 µg/mL sodium pyruvate, 40 µg/mL L-proline, 1% ITS and 1% FBS.

Following three weeks induction, the level of differentiation was quantified using qRT-PCR for lineage-specific markers (osteogenesis: osteopontin (OPN/BSP1) and bone morphogenetic protein 2 (BMP2), adipogenesis: peroxisome proliferator-activated receptor gamma (PPARG) and fatty acid binding protein 4 (FABP4), and chondrogenesis: collagen 2 (COL2A1) and aggrecan (ACAN)). All specific gene expressions were normalized to GAPDH housekeeping gene (See Supplementary Table 2 for details). Total RNA was extracted using TRIzol reagent (Invitrogen, CA, USA), and cDNA synthesis was carried out by applying the Amplisense cDNA Synthesis kit (AmpliSens, Moscow, Russia). qRT-PCR reactions were done by Applied Biosystems Step One Plus (ABI, CA, USA) and data analysis performed through the ddCt method to compare to pre-treatment cells (day 0).

In addition, an immunocytochemistry assay was carried out for further evaluation of differentiation at the protein level. Cells were stained with integrin-binding sialoprotein (IBSP/BSP2), PPARG, and COL2A1, which are used for labeling osteoblasts, adipocytes, and chondrocytes, respectively. To prepare samples for immunostaining, cells were fixed, permeabilized (as we mentioned in section 5.5), and stained with specific primary and secondary antibodies (see Supplementary Table 3 for details). Imaging was performed using fluorescence microscopy (Olympus, BX51, Japan).

Results

Substrate Characterization

A thin layer of ZIF-8 crystals was observed in coated samples, through SEM imaging of modified membranes, and verified by XRD analysis and FTIR absorbance spectra (Figure 1A–C). Even though PP and PP/PDA specimens do not present crystalline XRD pattern, exhibited diffraction peaks of PP/PDA/ZIF8 sample are highly matched with the ZIF8 pattern obtained from Cambridge Structural Database (CSD) as well as the simulated PP/PDA/ZIF8 XRD pattern. PDA/PEI surface modification is accompanied by two distinct IR peaks at 3200–3600 cm⁻¹, due to the stretching vibration of alcohol, catechol and amine bonds, and 1680 cm⁻¹, mainly related to C=N bonds between PEI and PDA. Among new peaks revealed as the result of ZIF8 coating, a characteristic peak at 420 cm⁻¹ refers to Zn-N stretching, and the other notable peak at 1600 cm⁻¹ corresponds to N-H bending. In addition, substrate treatment with PDA and PDA/ZIF8 coating dramatically decrease the water contact angle from 130 ͦ ± 1 for PP to 54 ͦ ± 8 and 27.8 ͦ ± 7, respectively. In contrast, SFE values rise significantly after PDA treatment and increase further after ZIF8 coating (Figure 1D). Furthermore, we found that PDA/PEI modification leads to a notably smoother substrate (Ra = 57 nm ± 1.6, Rq = 71.4 nm ± 5.8) compared to pristine PP membrane (Ra = 213.9 nm ± 18.5, Rq = 279.7 nm ± 27.3). On the other hand, as expected, ZIF8 in situ crystallization significantly increases the roughness value of the modified membrane (Ra = 244.2 nm ± 27, Rq = 376.3 nm ± 97.1), because of the crystalline structure of the thin layer (Figure 1E). Figure 1F shows 4.25 GPa Young’s modulus (from a surface roughness analysis via AFM force mapping) for PP/PDA/ZIF8 samples, which is significantly higher than both PP/PDA (1.84 GPa) and PP (0.99 GPa) specimens.

Figure 1.

Characterization of modified substrates. (A) SEM imaging for exploring the morphological feature of PP/PDA and PP/PDA/ZIF-8 substrates. Images with higher magnification are shown as inserts. (B) XRD crystalline patterns of the PP, PP/PDA, and PP/PDA/ZIF-8 membranes compared to the simulated ZIF-8 coated sample and ZIF-8 database. (C) FTIR spectra obtained from PP, PP/PDA, and PP/PDA/ZIF-8 samples. Blue and green dash lines show characteristic peaks for ZIF-8 and PDA-PEI, respectively. (D) Water contact angle measurement and calculated surface free energy for PP, PP/PDA, and PP/PDA/ZIF-8. Each sample was assessed in three replicates for 10 seconds. Data is represented as mean ± SEM, * p<0.05. (E) 3D AFM topographical images and measured surface roughness values of PP, PP/PDA, and PP/PDA/ZIF-8 substrates. (F) Surface elastic modulus measurement for PP, PP/PDA, and PP/PDA/ZIF-8 substrates by PeakForce AFM. Individual values are presented for 256 samples/line in each specimen with the corresponding mean value.

Abbreviations: PP, polypropylene; PDA, polydopamine; PEI, polyethyleneimine.

Cell Attachment on ZIF-8 Thin Film

The first set of cell culture experiments on MOF thin films revealed a proper attachment of DPSCs on PP/PDA/ZIF-8 substrate with a level comparable to the TCP control group (Figure 2A and B). Quantitative analysis showed that nearly twice the amount of cells adhered to ZIF-8, while the spread area of the cells did not change much (Figure 2C). We also assessed the primary metabolic activity of cells attached to samples by MTS assay. Surprisingly, cells cultured on ZIF-8 exhibited markedly reduced activity, in contrast to the control group of TCP (Figure 2D). Morphological studies by SEM imaging provided additional evidence for satisfactory adherence of cells on ZIF-8, likewise TCP surface (Figure 2E and F). Furthermore, strong evidence of the viability of the attached cells is presented in Figure 2G and H, which shows the high number of CFSE⁺/PI⁻ cells on TCP and ZIF-8 layer.

Figure 2.

Cellular attachment and viability analysis of DPSCs on the ZIF-8 modified substrate. F-actin arrangement of DPSCs cultured on (A) TCP and (B) PP/PDA/ZIF-8 substrates after one day. F-actin filaments were visualized via labeling with phalloidin-TRITC (red), and nuclei were stained with DAPI (blue). Scale bar = 200 µm. Magnified inserts showed the typical morphological features of cells on each substrate. Scale bar = 20 µm. (C) Bar graph presenting the quantification of cell count (black bar) and cell spread area (brown bar). (D) Analysis of cell metabolic activity was carried out after 8 hours, using MTS assay. SEM imaging of cultured cells on (E) TCP and (F) PP/PDA/ZIF-8 substrates, following one-day incubation in normal culture condition. The viability of cultured DPSCs was evaluated on (G) TCP and (H) PP/PDA/ZIF-8 substrates on day 1. Viable cells are stained in green as the result of CFSE cleavage, while dead cells are showed by PI-positive cells (red points). Quantified data are represented as mean ± SEM from three independent experiments, * p<0.05.

On the other hand, notable populations of cells were found on the intact polypropylene support, which formed an aggregation feature with a very low level of cell spreading. This phenomenon was confirmed by SEM imaging, and the basal activity of cells in this configuration was defined regarding MTS results (Supplementary Figure 2). Not surprisingly, DPSCs finely attached, extended, and retained their fibroblastic morphology on PDA modified PP membrane (Supplementary Figure 3).

Cell Proliferation on Modified Membranes

As for TCP, cytoskeletal imaging of cells on PP/PDA/ZIF-8 substrate after seven days clearly showed the significant propagation of cells in comparison to the first day (Figure 3A and B). Further quantitative analysis of figures highlighted that a greater number of cells with a higher spread area were observed on the ZIF-8 layer than for a standard tissue culture plate (Figure 3C). Aligned with these findings, MTS analysis revealed a reasonable absorbance on day one through both tested groups, which gradually rose to a near-constant (Figure 3D). While the TCP group showed higher initial absorbance and peaked sooner (day 5), cells on PP/PDA/ZIF-8 coated membrane grew rapidly, as overtaking the control after seven days.

Figure 3.

Proliferation potential of DPSCs on the ZIF-8 modified substrate. Cytoskeletal F-actin staining of DPSCs cultured on (A) TCP and (B) PP/PDA/ZIF-8 substrates after seven days. F-actin filaments were visualized via labeling with phalloidin-TRITC (red), and nuclei were stained with DAPI (blue). Scale bar = 200 µm. The insets are showing a higher-magnification view of the larger image. Scale bar = 20 µm. (C) Quantification analysis of cell number (black bar) and cell spread area (brown bar) of phalloidin-stained cells after a week. (D) Investigation of the proliferation rate of cultured DPSCs via analysis of cellular metabolic activity with MTS assay over seven days.

Even though cellular activity and the total spreading of DPSCs attached to the PP membrane displayed slight elevation during the period, the overall number of cells did not increase (Supplementary Figure 2). On the other hand, Supplementary Figure 3 provides evidence on the proliferation and expansion of DPSCs on the PP/PDA modified membrane.

Multilineage Differentiation of DPSCs on ZIF-8-Coated Membranes

The multilineage differentiation potential of DPSCs on the ZIF-8 layer was assessed following 3-weeks induction under specific culture conditions. Gene expression analysis exhibited significant elevation of expression of all critical lineage-specific markers in cells cultured on PP/PDA/ZIF8 in both RNA and protein levels. We found a substantially higher expression level of BSP1 and BMP2 not only concerning day 0 (external control sample) but also to the treated cells on the TCP substrate (Figure 4A). On the other hand, the expression of PPARG and FABP4 enhanced with an almost similar level in both experimental groups (Figure 4B). Interestingly, two-dimensional chondrogenic induction resulted to form an aggregation of cells exceptionally expressed COL2A1 and ACAN; however, it was much fewer than the cells cultured on TCP (Figure 4C). Immunofluorescent imaging also revealed considerable expression of BSP2, PPARG, and COL2A1 in DPSCs, which were treated with osteo-, adipo-, and chondrogenic induction medium on TCP and PP/PDA/ZIF-8 substrate (Figure 4D–J).

Figure 4.

Multilineage differentiation potential of DPSCs on the ZIF-8 modified substrate. Evaluation of the multilineage differentiation potential of DPSCs cultured on PP/PDA and PP/PDA/ZIF-8, and TCP substrates at mRNA level, after three weeks induction. Osteo-, adipo-, and chondrogenesis were assessed respectively via measurement of the relative expression of (A) BSP1 and BMP2, (B) FABP4 and PPARG, and (C) COL2A1 and ACAN compared to pre-treated cells (Day 0). Data represented as mean ± SEM from three independent experiments, * p<0.05. (D) The quantitative analysis of fluorescence intensity of immunostaining assay for lineage-specific markers (BSP2: Osteoblasts; PPARG: Adipocytes; COL2A1: Chondrocytes) in the level of protein at the end of 21 days induction time for DPSCs cultured on TCP and PP/PDA/ZIF-8. The representative images of stained DPSC for (E, F) BSP2, (G, H) PPARG, and (I, J) COL2A1. In all images, nuclei were stained in blue with DAPI. Scale bar is 200 µm.

Discussion

Despite rapid progress toward the biomedical application of stem cells, there remains a high demand for clinical-grade stem cell products as well as functional supports to promote tissue regeneration. Over the past two decades, bioengineering strategies investigate for suitable biomaterials to develop safe patient-specific regeneration therapies. In line with our previous studies on MOF-modified substrates,^{77,82,101,102,110,115} we proposed these nanostructures as tunable cell culture platforms with considerable potential to simultaneously regulating different microenvironmental cues. This idea is supported by the diverse physical and mechanical characteristics of more than 20,000 components belong to MOF family, such as elastic properties, inherent porosity, stiffness, and particle size.^116,117 From the literature, it is mainly governed by the structural flexibility, framework breathing, and the chemistry of subunits, and influenced by the fabrication conditions.^117–119 Besides, there are very limited reports exploited particular characteristics of MOFs to specifically control cellular behavior. For instance, controllable gas-release from photoactive MOF crystals has been applied for local regulation of cellular signaling pathways through in vitro culture conditions.^120,121 A number of recent approaches have utilized MOF nanoparticles for improving osteoconductivity/osteoinductivity of standard bone substituents, such as titanium (Ti), poly‐l‐lactic acid (PLLA), and calcium phosphate (CaP).^88,89,122 This emerging paradigm is quite governed by the precise concentration of metal ions and the stability of MOF nanostructure. Accordingly, it is deduced that applying MOF layers could improve their stability at the cell-substrate interaction surface and decline their probable toxic effect.

According to our hypothesis, we examined the cellular response of human DPSCs to the particular physio-mechanical cues of the ZIF-8 thin layer, as a well-known example of MOFs. All experiments were also performed for the PP membrane as well as PP/PDA to ascertain whether the findings were related to the ZIF-8 layer. Razmjou et al previously investigated the physical properties of the ZIF-8 dense layer with an approximate thickness of 800 nm and sensible adsorption capacity for silver particles.⁸² Generally, this study showed that forming the ZIF-8 layer significantly improved the hydrophilicity, SFE, nano-roughness, and elastic modulus of the PP membrane. Indeed, the chemical composition analysis identified plenty of N-H and Zn-N bonds in the ZIF-8 crystalline layer.

Based on the critical objectives of tissue engineering, cell culture platforms are required to hold suitable surface properties to promote stem cell adhesion, proliferation, and differentiation.^123–125 The higher number of adhered DPSCs on the ZIF-8 layer, as compared to TCP as the “gold standard” substrate for cell culture, highlights its superiority not only for in vitro but also for in vivo and clinical experiments. In fact, the relative increased roughness and greater surface area provided by ZIF-8 film (37.46 m²/g, graphs are shown in Supplementary Figure 1 and data are summarized in Supplementary Table 1) along with the active basic nitrogen atoms in the structure of ZIF-8 crystals seems to play critical roles in the primary attachment of cells.¹²⁶

Although DPSCs were successfully attached to a ZIF-8 layer, they were not able to expand very well and find an appropriate fibroblastic morphology within one day. The reduced level of metabolic activity at this stage correlates with substantial evidence in the literature that showed the reduced cell spreading area, indicating a significant drop in cellular metabolism.^127–129 Interestingly, after seven days, they formed a compact monolayer of cells that overtook the PP/PDA group and TCP. Furthermore, the significant higher surface free energy is also an essential factor for enhanced cell attachment following PDA modification. This result is in accordance with reported findings that considered surface free energy as an effective parameter for cell-substrate interaction, in addition to wettability.¹³⁰

Regarding the importance of efficient differentiation promotion by cell culture platforms through homing conditions,^131–133 we evaluated the multilineage potential of DPSCs during long-term culturing on a ZIF-8 layer. It is suspected that considerable upregulation in all three lineage-related genes is at least partially related to the higher surface area provided by the ZIF-8 layer. Even more important factor may be harnessing of cell-secreted stimulatory signals by internal active sites in the ZIF-8 crystalline layer.⁹⁹ The trapping of soluble growth factors and ECM components secreted by differentiating DPSCs on the substrate resulting in subsequent amplification of the differentiation process.

As a key index of osteogenic differentiation during bone development and regeneration,^134,135 the highly amplified expression of BSP1 and BMP2 on PP/PDA/ZIF-8 platform is in line with our primary hypothesis on promising potential of the ZIF-8 layer for exaggerating the innate osteogenic potential of DPSCs. Indeed, our findings showed that the ZIF-8 coating drastically increased the rigidity of matrix, which is extensively reported to have a major impact on the commitment of MSCs to osteogenic differentiation.^37,136,137 The higher elastic modulus of PP/PDA/ZIF-8 substance compared to the reported value for standard tissue culture plates (≈ 1 GPa) is consistent with the higher level of osteogenesis on the ZIF-8 layer.^138,139 Stiff materials predominantly regulate cell fate via modulating integrin interactions, reorganizing adhesion ligands, increasing cytoskeletal tension,¹⁴⁰ and inducing epigenetic modification.¹⁴¹

As a result of the chondrogenic induction of DPSCs through two-dimensional culture conditions, we were surprised to detect 3D cartilage-like structures after three weeks. This finding is in complete agreement with the proper physiological environment of chondrocytes, which function in loose contact with the substrate.^142–144 It seems that the stiff surfaces provided by both test and control groups prompted the cells to disrupt cell-substrate adhesions, migrate, and form features more resembling the natural cartilage environment. The elevated expression of COLL2A confirmed the notable level of chondrogenesis on both substrates. The higher stiffness value of PP/PDA/ZIF-8 is accompanied by lower chondrogenic efficiency, compared with TCP. In contrast to remarkable osteo- and chondrogenesis, the relative weaker adipogenic differentiation observed maybe because of the stiff structure of the substrates.

Hence, these attractive features render PP/PDA/ZIF-8 platform suitable for GBR therapies. GBR is a surgical technique that occludes the in-growth of adjacent soft tissue into periodontal bone defects via utilizing a flexible bioactive barrier membrane. The ideal GBR membrane is required to improve bone regeneration and appropriately integrate into the host tissue.³⁴ The recent clinical studies provided evidences for recruitment of cells into pristine PP membrane during GBR processes.^38,39,145 Similarly, we found that DPSCs can be attached on PP membrane and enhanced their metabolic activity over time; however, they did not spread properly even after seven days. This phenomenon was also observed for adipose-derived stem cells (ADSCs) grown on poly(l-lactide) acid (PLLA) film.⁶⁵ It seems that the inherent hydrophobic characteristic of the support materials strongly inhibited the normal expansion of MSCs in both cases. Whereas, significant improvement in cellular attachment on the PP/PDA/ZIF-8 platform may be reflected in the marked increase in surface hydrophilicity, obtained through PDA surface modification, as previously described by Schendzielorz et al.¹⁴⁶

Conclusion

In summary, we have obtained satisfactory results proving the promising capacity of ZIF-8 thin film as a stimulating agent on the cell culture substrate. This report opens up a new field of study of the applications of MOFs for the manufacturing of substrates able to mimic various mechanochemical properties of natural tissues. As an example, the PP/PDA/ZIF-8 showed remarkable capacity for the primary adhesion of DPSCs in comparison to other surface modification strategies (Supplementary Table 4). Exceptional expression of lineage-specific markers on PP/PDA/ZIF-8 substrate, especially in the case of bone-specific markers, presents higher efficiency rather than immobilization of some biological components (Supplementary Table 5). Favorable characteristics of the MOF nanostructures could minimize the requirement for biologically active induction molecules, and provide a chemical-based alternative to cell microenvironments. Further studies should concentrate on designing particular substrates with the desired stiffness, roughness, and geometrical features. On a broader level, we can apply a wide range of post functionalization and pore engineering approaches to mimic physiological conditions. In the light of these findings, PP/PDA/ZIF-8 platform exhibits profound capacity to use as a promising nanomedical tool in GBR procedures, apart from its in vitro application for mimicking natural microstructure of hard tissues.

Disclosure

Ms Fatemeh Ejeian reports grants from the Iran National Science Foundation (INSF) and Biotechnology Development Council of the Islamic Republic of Iran during the conduct of the study. Professor Vicki Chen reports grants from Australian Research Council during the conduct of the study. The authors report no other potential conflicts of interest for this work.