There is an acute need for alternatives to modern bone regeneration techniques, which have in vivo morbidity and high cost. Dental pulp stem cells (DPSCs) constitute an immunocompatible and easily accessible cell source that is capable of osteogenic differentiation. In this study, we engineered economical hard-soft intercalated substrates using various thicknesses of graphene/polybutadiene composites and polystyrene/polybutadiene blends. We investigated the ability of these scaffolds to increase proliferation and induce osteogenic differentiation in DPSCs without chemical inducers such as dexamethasone, which may accelerate cancer metastasis.

For each concentration, samples were prepared with dexamethasone as a positive control. Proliferation studies demonstrated the scaffolds’ effects on DPSC clonogenic potential: doubling times were shown to be statistically lower than controls for all substrates. Confocal microscopy and scanning electron microscopy/energy dispersive X-ray spectroscopy indicated widespread osteogenic differentiation of DPSCs cultured on graphene/polybutadiene substrates without dexamethasone. Further investigation of the interaction between hard-soft intercalated substrates and cells can yield promising results for regenerative therapy.


Current mainstream bone regeneration techniques, such as autologous bone grafts, have many limitations, including donor site morbidity, graft resorption, and high cost.1,2 An estimated 1.5 million individuals suffer from bone-disease related fractures each year, and about 54 million individuals in the United States have osteoporosis and low bone mass, placing them at increased risk for fracture.2,3,4,5 Bone tissue scaffold implants have been explored in the past decade as an alternative option for bone regeneration treatments. In order to successfully regenerate bone tissue, scaffolds typically require the use of biochemical growth factors that are associated with side effects, such as the acceleration of cancer metastasis.6,7 In addition, administering these factors in vivo is a challenge.6 The purpose of this project was to engineer and characterize a scaffold that would overcome these obstacles and induce osteogenesis by controlling the mechanical environment of the implanted cells.

First isolated in 2001 from the dental pulp chamber, dental pulp stem cells (DPSCs) are multipotent ecto-mesenchymal stem cells.8,9 Previous studies have shown that these cells are capable of osteogenic, odontogenic, chondrogenic, and adipogenic differentiation.10,11,12 Due to their highly proliferative nature and various osteogenic markers, DPSCs provide a promising source of stem cells for bone regeneration.11

An ideal scaffold should be able to assist cellular attachment, proliferation and differentiation.13 While several types of substrates suitable for these purposes have been identified, such as polydimethylsiloxane14 and polymethyl methacrylate15, almost all of them require multiple administrations of growth factors to promote osteogenic differentiation.6 In recent years, the mechanical cues of the extracellular matrix (ECM) have been shown to play a key role in cell differentiation, and are a promising alternative to chemical inducers.16,17

Recent studies demonstrate that hydrophobic materials show higher protein adsorption and cellular activity when compared to hydrophilic surfaces; therefore, we employed hydrophobic materials in our experimental scaffold.18,19,20,21 Polybutadiene (PB) is a hydrophobic, biocompatible elastomer with low rigidity. Altering the thickness of PB films can vary the mechanical cues to cells, inducing the desired differentiation. DPSCs placed onto spin-casted PB films of different thicknesses have been observed to biomineralize calcium phosphate, supporting the idea that mechanical stimuli can initiate differentiation.6,16,17 Atactic polystyrene (PS) is a rigid, inexpensive hydrophobic polymer.22 As PB is flexible and PS is hard, a polymer blend of PS-PB creates a rigid yet elastic surface that could mimic the mechanical properties of the ECM.

Recently, certain carbon compounds have been recognized as biomimetic.23 The remarkable rigidity and elasticity of graphene, a one-atom thick nanomaterial, make it a compelling biocompatible scaffold material candidate.24 Studies have also shown that using a thin sheet of graphene as a substrate enhances the growth and osteogenic differentiation of cells.23

We hypothesized that DPSCs plated on hard-soft intercalated substrates—specifically, graphene-polybutadiene (G-PB) substrates and polystyrene-polybutadiene (PS-PB) substrates of varying thicknesses—would mimic the elasticity and rigidity of the bone ECM and thus induce osteogenesis without the use of chemical inducers, such as dexamethasone (DEX).

Materials and Methods

G-PB and PS-PB solutions were prepared through dissolution of varying amounts of graphene and PS in PB-toluene solutions of varying concentrations. Graphene was added to a thin PB solution (3 mg PB/mL toluene) to create a 1:1 G-PB ratio by mass. Graphene was added to a thick PB solution (20 mg PB/mL toluene) to create 1:1 and 1:5 G-PB ratios by mass. PS was added to a thick PB solution to create 1:1, 1:2, and 1:4 PS-PB blend ratios by mass. Spincasting was used to apply G-PB and PS-PB onto silicon wafers as layers of varying thicknesses (thin PB: 20.5nm, thick PB: 202.0nm).25 Subsequently, DPSCs were plated onto the coated wafers either with or without dexamethasone (DEX). Following a culture period of eight days, the cells were counted with a hemacytometer to determine proliferation, and then stained with xylenol orange for qualitative analysis of calcification. Cell morphology and calcification of stained cells were determined through confocal microscopy and phase contrast fluorescent microscopy. Cell modulus and scaffold surface character were determined using atomic force microscopy. Finally, cell biomineralization was analyzed using scanning electron microscopy.


Cell Proliferation and Morphology

To ensure the biocompatibility of graphene, cell proliferation studies were conducted on all G-PB substrates. Results showed that G-PB did not inhibit DPSC proliferation. The doubling time was lowest for 1:1 G-thick PB, while doubling time was observed to be greatest for 1:1 G-thin PB. Multiple two-sample t-tests showed that the graphene substrates had significantly lower doubling times than standard plastic monolayer (p < 0.001).

After days 3, 5, and 8, cell morphology of the DPSCs cultured on the G-PB and PS-PB films was analyzed using phase contrast fluorescent microscopy. Images showed normal cell morphology and growth, based on comparison with the control thin PB and thick PB samples. After day 16 and 21 of cell incubation, morphology of the DPSCs cultured on all substrates was analyzed using confocal microscopy. There was no distinctive difference among the morphology of the DPSC colonies. DPSCs appeared to be fibroblast-like and were confluent in culture by day 15 of incubation.

Modulus Studies

In order to establish a relation between rigidity of the cells and rigidity of the substrate the cells were growing on, modulus measurements were taken using atomic force microscopy. Modulus results are included in Figure 3.

Differentiation Studies

After day 16 of incubation, calcification of DPSCs on all substrates was analyzed using confocal microscopy. Imaging indicated preliminary calcification on all substrates with and without DEX. Qualitatively, the DEX samples exhibited much higher levels of calcification than their non-DEX counterparts, as evident in thin PB, 1:1 G-thin PB, and 1:4 PS-PB samples.

After days 16 and 21 of incubation, biomineralization by the DPSCs cultured on the substrates was analyzed by scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDX). The presence of white, granular deposits in SEM images indicates the formation of hydroxyapatite, which signifies differentiation. This differentiation was confirmed by the presence of calcium and phosphate peaks in EDX analysis. Other crystals (not biomineralized) were determined to be calcium carbonates by EDX analysis and were not indicative of DPSC differentiation.

On day 16, only thin PB induced with DEX was shown to have biomineralized with sporadic crystal deposits. By day 21, all samples were shown to have biomineralized to some degree, except for 1:1 PS-PB (non-DEX) and 1:4 PS-PB (non-DEX). Heavy biomineralization in crystal and dotted structures was apparent in DPSCs cultured on 1:1 G-thin PB (both with and without DEX). Furthermore, samples containing graphene appeared to have greater amounts of hydroxyapatite than the control groups. All PS-PB substrates biomineralized in the presence of DEX, while only 1:2 PS-PB was shown to biomineralize without DEX. While results indicate that while PS-PB copolymers generally require DEX for biomineralization and differentiation, this is not the case for G-PB substrates. Biomineralization occurred on DPSCs cultured on G-PB substrates without DEX, demonstrating the differentiating ability of the G-PB mechanical environment and its interactions with DPSCs.


This study investigated the effect of hard-soft intercalated scaffolds on the proliferation and differentiation of DPSCs in vitro. As cells have been shown to respond to substrate mechanical cues, we monitored the effect of ECM-mimicking hard-soft intercalated substrates on the behavior of DPSCs. We chose graphene and polystyrene as the hard components, and used polybutadiene as a soft matrix.

By using AFM for characterization of the G-PB composite and PS-PB blend substrates, we demonstrated that all surfaces had proper phase separation and uniform dispersion. This ensured that DPSCs would be exposed to both the hard peaks and soft surfaces during culture, allowing us to draw valid conclusions regarding the effect of substrate mechanics.

Modulus studies on substrates indicated that 1:1 G-thin PB was the most rigid substrate and control thin PB was the second most rigid. In general, G-PB substrates were 8-20 times stiffer than PS-PB substrates. The high relative modulus of graphene-based substrates can be attributed to the stiffness of graphene itself. The cell modulus appeared highly correlated to the substrate modulus, both indicating that the greater the stiffness of the substrate, the greater the stiffness of DPSCs cultured on it and supporting the finding that substrate stiffness affects the cell ECM.27,28 For example, 1:1 G-thin PB had the highest surface modulus, and DPSCs grown on 1:1 G-thin PB had the highest cell modulus. Conversely, DPSCs grown on 1:4 PS-PB had the lowest cell modulus, and 1:4 PS-PB had one of the lowest relative surface moduli. Another notable trend involves DEX; cells cultured with DEX had greater moduli than cells cultured without, suggesting a possible mechanism used by DEX to enhance stiffness and thereby osteogenesis of DPSCs.

Cell morphology studies indicated normal growth and normal cell shape on all substrates. Cell proliferation studies indicated that all samples had significantly lower cell doubling times than standard plastic monolayer (p < 0.001). Results confirm that graphene is not cytotoxic to DPSCs, which supports previous research.27

SEM/EDX indicated that DPSCs grown on thick PB soft substrates appeared to have increased proliferation but limited biomineralization. In contrast, cells on the hardest substrates, 1:1 G-thin PB and thin PB, exhibited slower proliferation, but formed more calcium phosphate crystals, indicating greater biomineralization and osteogenic differentiation. The success of G-PB substrates in inducing osteogenic differentiation may be explained by the behavior of graphene itself. Graphene can influence cytoskeletal proteins, thus altering the differentiation of DPSCs through chemical and electrochemical means, such as hydrogen bonding with RGD peptides.29,30 In addition, G-PB substrate stiffness may upregulate levels of alkaline phosphatase and osteocalcin, creating isometric tension in the DPSC actin network and resulting in greater crystal formation.30 Overall, the proliferation results indicate that cells that undergo higher proliferation will undergo less crystal formation and osteogenic differentiation (and vice-versa).

The data presented here indicate that hard-soft intercalated substrates have the potential to enhance both proliferation and differentiation of DPSCs. G-PB substrates possess greater differentiation capabilities, whereas PS-PB substrates possess greater proliferative capabilities. Within graphene-based substrates, 1:1 G-thin PB induced the greatest biomineralization, performing better than various other substrates induced with DEX. This indicates that substrate stiffness is a potent stimulus that can serve as a promising alternative to biochemical factors like DEX.


The development of an ideal scaffold has been the focus of significant research in regenerative medicine. Altering the mechanical environment of the cell offers several advantages over current strategies, which are largely reliant on growth factors that can lead to acceleration of cancer metastasis. Within this study, the optimal scaffold for growth and differentiation of DPSCs was determined to be the 1:1 G-thin PB sample, which exhibited the greatest cell modulus, crystal deposition, and biomineralization. In addition, our study indicates two key relationships: one, the correlation between substrate and cell rigidity, and two, the tradeoff between scaffold-induced proliferation and scaffold-induced differentiation of cells, which depends on substrate characteristics. Further investigation of hard-soft intercalated substrates holds potential for developing safer and more cost-effective bone regeneration scaffolds.


  1. Spin-Neto, R. et al. J Digit Imaging. 2011, 24(6), 959–966.
  2. Rogers, G. F. et al. J Craniofac Surg. 2012, 23(1), 323–327.
  3. Bone health and osteoporosis: a report of the Surgeon General; Office of The Surgeon General: Rockville, 2004.
  4. Christodoulou, C. et al. Postgrad Med J. 2003, 79(929), 133–138.
  5. Hisbergues, M. et al. J Biomed Mater Res B. 2009, 88(2), 519–529.
  6. Chang, C. et al. Ann J Mater Sci Eng. 2014, 1(3), 7.
  7. Jang, JY. et al. BioMed Res Int. 2011, 2011.
  8. d’Aquino, R. et al. Stem Cell Rev. 2008, 4(1), 21–26.
  9. Liu, H. et al. Methods Enzymol. 2006, 419, 99–113.
  10. Jimi, E. et al. Int J Dent. 2012, 2012.
  11. Gronthos, S. et al. Proc Natl Acad Sci. 2000, 97(25), 13625–13630.
  12. Chen, S. et al. Arch Oral Biol. 2005, 50(2), 227–236.
  13. Daley, W. P. et al. J Cell Sci. 2008, 121(3), 255–264.
  14. Kim, S.J. et al. J Mater Sci Mater Med. 2008, 19(8), 2953–2962.
  15. Dalby, M. J. et al. Nature Mater. 2007, 6(12), 997–1003.
  16. Engler, A. J. et al. Cell. 2006, 126(4), 677–689.
  17. Reilly, G. C. et al. J Biomech. 2010, 43(1), 55–62.
  18. Schakenraad, J. M. et al. J Biomed Mater Res. 1986, 20(6), 773–784.
  19. Lee, J. H. et al. J Biomed Mater Res. 1997, 34(1), 105–114.
  20. Ruardy, T. G. et al. J Colloid Interface Sci. 1997, 188(1), 209–217.
  21. Elliott, J. T. et al. Biomaterials. 2007, 28(4), 576–585.
  22. Danusso, F. et al. J Polym Sci. 1957, 24(106), 161–172.
  23. Nayak, T. R. et al. ACS Nano. 2011, 5(6), 4670–4678.
  24. Goenka, S. et al. J Control Release. 2014, 173, 75–88.
  25. Extrand, C. W. Polym Eng Sci. 1994, 34(5), 390–394.
  26. Oh, S. et al. Proc Natl Acad Sci. 2009, 106(7), 2130–2135
  27. Jana, B. et al. Chem Commun. 2014, 50(78), 11595–11598.
  28. Nayak, T. R. et al. ACS Nano. 2010, 4(12), 7717–7725.
  29. Banks, J. M. et al. Biomaterials. 2014, 35(32), 8951–8959.
  30. Arnsdorf, E. J. et al. J Cell Sci. 2009, 122(4), 546–553.