Reactive Oxygen-Scavenging Polydopamine Nanoparticle Coated 3D Nanofibrous Scaffolds for Improved Osteogenesis: Toward an aging in vivo bone regeneration model

Abstract

Tissue engineered scaffolds aimed at the repair of critical-sized bone defects lack adequate consideration for our aging society. Establishing an effective aged in vitro model that translates to animals is a significant unmet challenge. The in vivo aged environment is complex and highly nuanced, making it difficult to model in the context of bone repair. In this work, 3D nanofibrous scaffolds generated by the thermally-induced self-agglomeration (TISA) technique were functionalized with polydopamine nanospheres (PD NPs) as a tool to improve drug binding capacity and scavenge reactive oxygen species (ROS), an excessive build-up that dampens the healing process in aged tissues. PD NPs were reduced by ascorbic acid (rPD) to further improve hydrogen peroxide (H2O2) scavenging capabilities, where we hypothesized that these functionalized scaffolds could rescue ROS-affected osteoblastic differentiation in vitro and improve new bone formation in an aged mouse model. rPDs demonstrated improved H2O2 scavenging activity compared to neat PD NPs, although both NP groups rescued the alkaline phosphatase activity (ALP) of MC3T3-E1 cells in presence of H2O2. Additionally, BMP2-induced osteogenic differentiation, both ALP and mineralization, was significantly improved in the presence of PD or rPD NPs on TISA scaffolds. While in vitro data showed favorable results aimed at improving osteogenic differentiation by PD or rPD NPs, in vivo studies did not note similar improvements in ectopic bone formation an aged model, suggesting that further nuance in material design is required to effectively translate to improved in vivo results in aged animal models.

1. Introduction

The aging bone environment is characterized by several prominent pathologic states that collectively degenerate the body’s natural ability to maintain and repair its tissue^1,2. Chronic inflammation, cellular senescence, accumulation of reactive oxygen species (ROS), among a variety of other factors contribute to the weakened regenerative niche that impedes new bone formation during aging^3–5. Many of these conditions have individually been explored to an extent – however evidence regarding therapeutic approaches targeting these states is underexplored in the context of bone repair, especially in aged animal models. For example, ROS play roles in numerous physiological processes, operating as signaling molecules that help maintain cell differentiation, proliferation, and self-renewal capacity^3,6. However, the extent to which they exert influence on osteoblastic differentiation and overall bone regeneration is not very clear, in vitro and in vivo, and especially rings true in aged models.

Our previously developed TISA technique has demonstrated effectiveness in producing clinically relevant 3D electrospun nanofibrous scaffolds, a capability beyond the reach of traditional electrospinning techniques⁷. Its high elasticity, biomimetic architecture with hierarchically structured macro- and micro-pores, and high degree of porosity grant it many unique strengths that allow it to adapt to numerous in vitro or in vivo bone regeneration scenarios. While its promises for regenerative medicine were clear, TISA scaffolds still faced challenges with bioactivity, drug binding/release and was still unproven in challenged animal models such as aged/inflamed models. Follow-up studies demonstrated a propensity to easily blend multiple polymers, improve bioactivity by compositing with synthetic hydroxyapatite, and readily load protein and small molecules after coating for sustained drug release^8–10. The elasticity enabled fitting of complex-shaped defects in a facile manner, and these numerous improvements yielded significant improvements in osteogenic potential of stem cells in vitro and more new bone formation in vivo. While the strengths of the TISA scaffolds are clear, their potential as the bioactive matrix to improve bone regeneration in an aging model remains to be investigated.

Polydopamine is one very frequently utilized molecule in bioengineering and regenerative medicine^11,12. Derived from naturally-inspired mussels, polydopamine can readily and easily bind to many surfaces via physical adhesion and grant unique properties (e.g., hydrophilicity, drug binding capacity, and etc.) to bio-inert biomaterials^13,14. Polydopamine nanoparticles (PD NPs) in particular are recently seeing much innovation as drug carriers for cancer therapy, anti-inflammatory agents combating synaptic loss, or ROS scavengers in periodontal disease^15–18. PD coatings have been explored in bone regeneration, noting acceleration in new bone formation and improved osteointegration of biomaterials¹⁹. Their anti-inflammatory and anti-oxidant properties could potentially shed light on their mechanisms and lead us toward leveraging this ability to combat inflammation/ROS-challenged bone repair. Chronic inflammation and ROS buildup are two hallmarks of aging and targeting one or both of these via therapeutic interventions are potential tools to combat aging-related conditions^2,20. It is still not well understood what the ROS role is in osteoblastic or osteoclastic activity during repairing and remodeling as we age, and this uncertainty limits our therapeutic strategies using the bioengineering approaches. Efforts to elucidate this mechanism of ROS in basic bone regeneration have arisen through development and study of ROS-scavenging or ROS-responsive smart materials aiming to improve aged bone regeneration^21–23. It is therefore of great interest to determine if the ROS scavenging 3D scaffold can be an effective tool to mitigate ROS-diminished osteoblastic differentiation in vitro and further improve aged bone formation in vivo.

Here, we evaluate the effects of the PD NPs on osteoblastic differentiation, reduce them to effectively scavenge ROS, and then use them to functionalize TISA scaffolds with the aim of improving their drug binding/release, osteogenic and new bone forming capabilities both in vitro and in vivo. We hypothesize that PD NP functionalized TISA could be an effective tool to rescue ROS-induced reduction in osteogenesis, and mitigate ROS buildup during repair in an aged in vivo model to more effectively repair a bone defect.

2. Experimental

2.1. Materials

Polycaprolactone (PCL, MW=80k), gelatin (G1890), ethanol, Dichloromethane (DCM), N,N-dimethylformamide (DMF), Dopamine-Hydrochloride (Dopa-HCl), and Ammonium Hydroxide were purchased from Sigma-Aldrich (St. Louis, MO). All chemicals and materials used in this research were used without further purification.

2.2. Electrospinning and TISA Scaffold synthesis

TISA scaffolds were generated as described in our previous publications^7,8. Briefly, PCL nanofibrous (NF) mats with fiber diameter between 50–200nm were synthesized via electrospinning with a high-voltage power supply (Gamma High Voltage Research Inc. Ormond Beach, FL) and a laboratory-produced roller (25cm diameter). NFs were broken into short fibers by mechanical grinding in liquid nitrogen, and then the shortened fibers were separated individually and well dispersed in a mixed solvent of water/gelatin/ethanol with weight ratio of 4/2/1. The well-dispersed individual short NFs were spontaneously agglomerated into 3D structure via submersion of the glass bottle containing the suspension into a water bath at 55°C for 3 minutes. The suspension was then quenched in an ice bath for 30min to prevent further shrinkage of NF agglomeration. The resulting 3D scaffold was retrieved from the bottle and rinsed several times with water to remove residual gelatin and ethanol. Finally, the scaffold was lyophilized to retain the 3D porous morphology for future experiments. For cell culture experiments, TISA scaffolds were frozen in water and cut into discs by 5mm biopsy punch, and then cut to roughly 2mm thickness by a razor blade.

2.3. Polydopamine nanoparticle coating and subsequent reduction

Polydopamine nanoparticles (PD NPs) were generated by the Stöber process. 0.25g of Dopamine-HCl was dissolved in 5ml DI water. The dopamine-HCl solution was slowly added into a solution of DI water/Ethanol/NH₄OH (45ml/20ml/1ml) while stirring at 300rpm, then the reaction was continued overnight. For coating of PD NPs on the TISA scaffolds, 5mm TISA discs were submerged in the resulting PD NP solution for 24 hours in a shaker, and then rinsed 3x by DI water to remove residual NPs, ethanol, and NH₄OH. For reduction of PD NPs, NP suspension was centrifuged and residual ethanol and NH₄OH was aspirated, and then NPs were re-suspended in DI water. The re-suspended PD NPs were soaked in 300mM ascorbic acid for 30 minutes, and then centrifuged and rinsed again with DI water to remove residual ascorbic acid. NP reduction was confirmed by a Nicolet iS50 FTIR Spectrometer (Thermo Fisher). For TISA-PD scaffolds, the PD coated discs were submerged in 300mM ascorbic acid for 30 minutes, and then rinsed 3x by DI water to remove residual ascorbic acid. Scaffolds were then lyophilized and stored in a desiccator for future use in cell culture experiments.

2.4. Morphological and Chemical Characterization

Scanning Electron Microscopy (SEM) was utilized to characterize surface and nanofibrous morphology of TISA-PD scaffolds. Scaffolds were sputter coated with either gold or palladium prior to imaging. Samples were imaged with a Hitachi S-4800 SEM, with varying accelerating voltages between 2–10kV. Water contact angle was measured by goniometer (Ramé-Hart) and imaged by X software. Briefly, ES scaffolds were immobilized on glass slides by double-sided tape, and a 2.0 μL water droplet was rested on the surface for 5 seconds and images were subsequently captured and contact angle determined. For surface chemistry analysis, a Fourier Transform-Infrared Spectroscope (FT-IR, Nicolet is50) was employed to detect surface chemistry changes to nanoparticles. PD NPs were centrifuged to remove residual solvent, and rinsed 3x with DI water, and then freeze-dried to obtain powder for FTIR analysis.

AFM height images and nanoindentation measurements were collected using a MFP-3D AFM (Asylum Research, Oxford Instruments, Goleta, CA) under ambient temperature and pressure. All AFM imaging and nanoindentation measurements were performed using Silicon Nitride tips (MikroMasch, Watsonville, CA, CSC37) with a nominal spring constant of 0.7N/m and a typical tip radius of curvature of 8 nm. AFM height images were collected in intermittent mode (AC mode) at a typical scan rate of 1 Hz. AFM nanoindentation experiments were performed by measuring force versus vertical piezo displacement curves at multiple positions over several representative nanofibers that were then fit using the Johnson–Kendall–Roberts (JKR) contact model to determine the Young’s modulus from each force measurement at a particular nanofiber position^24–26. Repeated force-indentation plots were collected on 18 PCL and 16 PD-coated PCL nanofibers at four-six distinct positions across each fiber. The elastic modulus and Poisson’s ratio of the AFM probe were assumed to be 200 GPa and 0.20, respectively, and the Poisson’s ratio of nanofibers was assumed to be 0.33. The reported Young’s modulus results for PCL and PD-coated PCL fibers correspond to average and one standard deviation.

2.5. In vitro Hydrogen Peroxide Scavenging and BMP2 Release

Amplex Red Hydrogen Peroxide/Peroxidase assay kit (Thermo Fisher) was used to determine Hydrogen peroxide scavenging potential of PD NPs and PD NP coated TISA scaffolds according to manufacturer’s instructions. Briefly, 100 μM hydrogen peroxide solution was generated, and PD or rPD NPs were added to each solution at a concentration of 10mg/ml for 30 minutes to react. After 30 minutes, NPs were removed via centrifuge, and the resulting supernatant was analyzed for remaining hydrogen peroxide to determine scavenging potential of the NPs.

For BMP2 release from scaffolds, TISA, TISA-PD, and TISA-rPD scaffolds were sterilized with 70% ethanol solution for 30 minutes then rinsed 3x with dPBs and blotted dry with surgical gauze, and 100ng BMP2 was dropped onto scaffolds suspended from 10uL of dPBS. Scaffolds were incubated for 30 minutes on ice then submerged in 200 uL of dPBs at 37°C for elution study. At various pre-determined time points, dPBS was withdrawn and replaced with fresh dPBS and frozen for later analysis, which was carried out with a BMP2 ELISA kit (Peprotech).

2.6. In Vitro Cell Study

2.6.1. Scaffold preparation, Sterilization, and Cell Seeding

TISA scaffolds were cut into 5mm diameter × 2mm thick discs by biopsy punch and razor blade after being frozen in DI water. For sterilization, scaffolds were immersed in 70% ethanol for 30 minutes followed by rinse with sterile dPBS 3x for 30 minutes each, followed by incubation in minimum essential medium α (α-MEM, Gibco, Waltham, MA). Excess medium was removed by resting scaffolds on sterile gauze prior to cell seeding. For seeding, various concentrations of cells were added to scaffolds by dropping on to scaffolds in a volume of 10 μl of medium. A total cell count of 3.0×10⁵ cells were seeded onto scaffolds for the experiments for ALP Activity and total calcium content. Scaffolds were cultured on non-treated 35mm culture dishes at 37°C with 5% CO2.

2.6.2. Alkaline Phosphatase (ALP) Activity

ALP was measured using an EnzoLyte nNPP Alkaline Phosphatase Assay Kit (AnaSpec, San Jose, CA) as previously described^7,10. After 7 days of culture, scaffolds were rinsed with dPBS prior to lysing. After rinsing, scaffolds were immersed into lysis buffer for 10 minutes in an ice bath and mechanically ground using forceps. Ensuing lysate was then transferred to a microcentrifuge tube and then centrifuged at 2500xG at 4°C. Resulting supernatant was mixed with p-nitrophenyl phosphate and incubated for 30 minutes at 37°C, and then absorbance at 405nm was quantified using a microplate reader. ALP activity was normalized against total protein content, which was measured by BCA protein kit (Thermo Scientific). Lysate generated from the ALP kit was incubated with working solution from the BCA kit and absorbance at 562nm was quantified with a microplate reader.

2.6.3. Total Calcium Content

Cell seeded scaffolds were evaluated for total calcium content by using a total calcium LiquiColor kit (Stanbio Laboratory, TX). After 21 days of culture, scaffolds were rinsed in dPBS and dissected by razor blade into small pieces. Calcium was extracted using 1ml of 6M hydrochloric acid. 10 μl of resulting supernatant was added to 1ml of working solution from the LiquiColor kit and incubated according to the manufacturer’s instructions, where the absorbance was measured at 550nm by microplate reader.

2.7. In Vivo Bone Regeneration

In vivo bone regeneration was carried out by a mouse ectopic model. Animal surgeries were performed according to protocols approved by the office of the Institutional Care and Use Committee (IACUC) of the University of Iowa. Inbred B57BL/6J mice (19-month-old, female) were provided by National Institute on Aging (NIA) for use in this study. As per our previous protocol, mice were shaved on the dorsal side and an antiseptic (70% ethanol) was applied to the skin on the surgical area. Each mouse received a total of 4 scaffolds, implanted under an incision on the dorsal side. TISA, TISA/PD, and TISA/rPD scaffolds were implanted with an addition of 2 μg of rhBMP2 (PeproTech, Inc.), plus a fourth group of TISA without rhBMP2. After 4 weeks of initial surgery, mice were euthanized and ossicles were retrieved and fixed in 10% neutral buffered formalin followed by transfer to cold 70% ethanol for follow-up analysis.

2.8. Radiographic and Histological Analysis

Four weeks post-implantation, radiographic analysis was carried out on the fixed ossicles using an In-Vivo Extreme small imaging micro-CT system (Bruker, Billerica, MA, USA). Histological staining was utilized after decalcification in 15% EDTA for two days, followed by storage in 70% ethanol, and sectioning and cutting in paraffin. Samples were subsequently cut and stained with Hematoxylin and Eosin (H&E) and then microscopically observed.

3. Results

3.1. Morphology of PD-NP TISA Scaffolds

Fig. 1, A shows the SEM image of the TISA scaffolds containing agglomerated PCL nanofibers with large pores. PD NPs generated by the Stӧber method had size about 20 nm with a narrow size distribution as shown in the SEM image (Fig. 1, B). The PD NPs could be reliably coated on TISA scaffolds via simply soaking them in a water/ethanol solution. After coating, the scaffold nanofibers were evenly coated by NPs (Fig. 1, C) and the scaffold was darkened as shown in the inset of Fig. 1, C. After refreshing the PD NP coating solution multiple times, the NP density could be increased on the TISA scaffold (Fig. 1, D). For continued experiments, a single coating layer was utilized for simplicity and no notable difference in ALP activity change of MC3T3-E1 cells compared to multiple coating layers (data not shown).

Fig. 1 –

SEM Morphology of TISA scaffolds (A), PD nanoparticles (B), PD Nanoparticles on TISA scaffolds (C) and reduced PD nanoparticles on TISA scaffolds (D). Macroscopic images of TISA (A, inset) and TISA/PD (C, inset) are included as well.

3.2. Physical and Chemical characteristics of Scaffolds

After morphological observation by SEM, physical characteristics were investigated via contact angle goniometry and AFM. Several influencing factors for osteogenic cellular differentiation include wettability (as measured by contact angle) and matrix elasticity or stiffness such as Young’s modulus. After coating with PD NPs, water contact angle was sharply reduced from 120.02° ± 1.54 to 12.70° ± 2.0, indicating a sharp and significant increase in wettability from pure PCL fibers (Fig. 3, A,B). AFM height imaging was employed to visualize individual NPs on the surface (Fig. 2, C,D) and AFM nanoindentation measurements were utilized to calculate young’s modulus of the individual nanofibers, where PD coated PCL fibers showed a significant increase in the modulus from 1.5 ± 0.4 MPa to 13.9 ± 1.1 MPa (Fig. 2, E). FTIR spectra of nanoparticles before and after reduction via ascorbic acid are visible in Fig S1. After ascorbic acid treatment, a broad -OH band around 3400 cm⁻¹ was observed, characteristic of O-H stretching of phenolic hydroxyl, suggesting at least partial reduction of quinone subunits of the polydopamine molecules.

Fig. 3.

BMP2-induced ALP Activity of MC3T3-E1 cells cultured for 7 days on TISA scaffolds, normalized by total protein content. Inset shows image of TISA (A), and TISA-PD (B) scaffolds prior to cell seeding to demonstrate PD stability to sterilization. Data are expressed as mean±SD (n= 3).

Fig. 2.

Physical properties of PD coated PCL. Contact angle of neat PCL (A) and PD-coated PCL fibers (B). Morphology of PD coated fibers (C) and individual nanospheres (D) as shown by AFM. Young’s modulus of individual nanofibers as measured by AFM, where error bars correspond to one standard deviation (E, n = 5).

3.3. PD NP-stem cell interactions

PD NP interaction with MC3T3-E1 cells was evaluated as an early marker for osteogenic activity. Initial experiments showed that PD coating on TISA scaffold was capable of improving the ALP activity of MC3T3-E1 cells after 7days in culture, and this was subsequently investigated with BMP2-induced ALP activity, where the trend continued with a further increase in ALP activity (Fig. 3). PD NP coatings remained stable after sterilization in 70% ethanol (Fig. 3, inset), and even remained after 7days of culture in growth medium (data not shown).

3.4. Hydrogen Peroxide Scavenging effects of PD NPs

Polydopamines ability to scavenge reactive oxygen species has shown inconsistent conclusions in prior literature, with limited effectiveness or high amounts required, often leading to cytotoxicity^27,28. In our initial experimentation, PD NPs alone did little to reduce presence of hydrogen peroxide in solution, so we reduced the NPs with ascorbic acid with the goal of improving the capability of peroxide scavenging. After reduction, rPD NPs showed a significant decrease of hydrogen peroxide in solution as compared to equal mass of non-reduced PD NPs (Fig. 4). When this effect was studied on cell activity in vitro, interestingly both PD and rPD NPs rescued the ALP activity of MC3T3-E1 cells cultured in presence of hydrogen peroxide. This could potentially suggest NPs have some interaction with the cells that improve cells ability to differentiate independent of the direct scavenging activity of rPD NPs.

Fig. 4.

Hydrogen Peroxide scavenging capacity of reduced PD NPs after 30 minutes in 100uM H2O2 (left). ALP activity of MC3T3-E1 cells treated with H2O2 in presence or absence of PD NPs (right). Data are expressed as mean±SD (n= 3).

3.5. BMP2-induced in vitro osteogenic differentiation of rPD-TISA

To study whether or not this trend carried over to BMP2-induced in vitro ALP activity, TISA PD and rPD scaffolds were investigated under presence of BMP2 and similar trends were observed. BMP2 release from scaffolds was observed up to 2 weeks, where TISA scaffolds released the highest amount after 14 days, and both TISA/PD and TISA/rPD had lower burst release and slower release curves up to 14 days (Fig S2). BMP2 addition into scaffolds noted a significant increase in ALP levels for all TISA groups (Fig. 5). Interestingly, rPD group showed a significant increase within the BMP2 group in terms of raw ALP activity compared to TISA, while PD coated scaffolds did not. After normalization by BCA kit for total protein, the significant difference dissipated – potentially due to an improved ability of rPD scaffolds to protect cell growth and differentiation. Moving to long-term marker total calcium content as shown in figure 6, we see the trend of significant improvement is most notable in the non-BMP2 induced groups for rPD over neat TISA.

Fig. 5.

ALP Activity as normalized by total protein content (left, blue) and total ALP activity (right, green) after 7days of culture on scaffolds on TISA and PD or rPD-TISA scaffolds. Data are expressed as mean±SD (n= 3).

Fig. 6.

Total calcium content of MC3T3-E1 cells on scaffolds after culturing for 28days. Data are expressed as mean±SD (n= 3).

3.6. Ectopic bone formation on PCL TISA/PD and rPD scaffolds

To investigate the effects of PD and rPD influence on BMP2-induced bone formation in vivo, TISA, TISA/PD, and TISA/rPD scaffolds were ectopically implanted into the dorsal side of mice for 4 weeks. Each scaffold group received an equal dose of BMP2 (2 ug/scaffold), and a fourth TISA group with no added BMP2 was used as a negative control. After retrieving the ossicles after 4 weeks of implantation, histological staining revealed new bone formation from each group as depicted in figure 7. TISA +BMP2 (Fig. 7A) and TISA/rPD+BMP2 group (Fig. 7C) showed similarly significant new bone formation, with a higher amount evident by histological examination and the quantitative μCT data (Fig. 7E) compared to TISA/PD+BMP2 group (Fig. 7B). Whereas the in vitro data supported stronger osteogenic differentiation in PD and rPD coated TISA scaffolds, this trend was not replicated in vivo results. Unexpectedly, the TISA/PD+BMP2 group significantly inhibited new bone formation compared to TISA+BMP2 and TISA/rPD+BMP2 groups, contrary to the in vitro trend where both PD and rPD TISA groups improved BMP2-induced osteogenic activity.

Fig. 7.

Histological staining by H&E of ossicles after 4 weeks implantation in vivo. Groups are as follows: TISA+BMP2 (A), TISA/PD+BMP2 (B), TISA/rPD+BMP2 (C), and TISA/rPD (D). Scale bars = 200 μM. New bone area as quantified by μCT (E). Data are expressed as mean±SD (n= 3).

4. Discussion

Three-dimensional scaffolds must address numerous challenges when considering their use as engineered therapeutics for bone defects. These challenges become more numerous as compounding factors such as age and/or other comorbidities are added to the picture. PCL-TISA scaffolds as previously reported are robustly capable as biomimetic scaffolds for bone regeneration in vivo yet require varied functionalization to overcome the bioinert nature and limited drug binding capacity of PCL^7,8,10,29. Polymer blends and compositing with various hydroxyapatite (HA) coatings demonstrate capacity to tackle the challenges of bioinertness and drug binding/release, and show marked improvement in new bone formation compared to pure PCL-TISA^8–10,29. Yet, considerations for aging-related conditions such as high inflammation and high ROS presence are still to be undertaken. This work utilizes a natural-inspired, polydopamine nanoparticle ubiquitous coating to realize multi-faceted improvements to TISA scaffolds in hydrophilicity, osteoblastic differentiation, and ROS scavenging capacity.

Cell-material interactions drive many important aspects of cell culture in engineered biomaterial scaffolds including but not limited to seeding effectiveness, cell penetration, and eventually proliferation and differentiation^30,31. Not only does surface chemistry play a role in guiding osteoblastic differentiation, but also do the physical properties such as elasticity, stiffness, or hydrophilicity^32–34. PD NPs ubiquitously coated PCL-TISA scaffolds in this work, where coating intensity could be controlled by varying treatment time as seen via SEM. Coating of NPs was also able to be carried out without compromising the hierarchical macro- and micro-porous structure of scaffolds. Follow-up physical analysis noted an improvement in hydrophilicity as seen via goniometry and nanofiber modulus as examined by AFM. These physical contributions alone are strong potential reasons for the noted improvement in ALP activity of cells cultured on scaffolds after one week. While cell-material interactions can be modified to improve qualities for in vitro purposes, we can also note improvements in drug-material interactions as seen by the significant improvement in the BMP2 group during ALP activity experiments. PD NP coating dramatically enhanced the BMP2-induced ALP activity of cells cultured on TISA scaffolds in this work. While BMP2 is a well-known potent bone-forming agent, its short half-life in growth medium significantly drives up its cost and dosage requirements. PD coating has been shown to capture BMP2 in previous research for bone tissue regeneration^35,36; the increased surface area of the PD nanoparticles in this work are likely to provide a similar effect in protecting BMP2 as seen in the improved ALP activity.

ROS mitigation is one potentially strong strategy to combat aging-complicated injury or repair. PD is utilized in this work due to its significant potential for scavenging of ROS. Some previous work reports PD alone is capable of significantly reducing ROS without any further modification^15,16, whereas other groups note a more limited role of pure PD that can be improved after some modifications^37–39. In our work, PD NPs alone show a limited ability for peroxide scavenging, so the nanoparticles underwent reduction via ascorbic acid, and a significant improvement in peroxide scavenging potential was subsequently noted. Antioxidant activity arising from PD reduction can likely be attributed to neutralization via electron donation from the reduced polydopamine, reducing the overall amount in solution and in growth medium. In cells cultured in presence of neat and reduced NPs, interestingly both groups were capable of rescuing ALP activity of cells cultured in hydrogen-peroxide containing osteoconductive medium. The osteoconductive medium contains ascorbic acid in addition to beta-glycerophosphate, further suggesting that the ascorbic acid reduction component of NPs is a primary reason for their peroxide scavenging capabilities.

While the in vitro data followed a clear story in improving osteogenic potential and ROS scavenging capabilities, the in vivo story was not quite as clear. An ectopic bone formation model was chosen due to the ability for it to provide decisive results with the TISA scaffold and various functionalizations from previous studies⁷. While not reductive to in vitro markers, TISA/PD scaffolds significantly reduced BMP2-induced new bone formation in the in vivo ectopic model. The TISA/rPD group however saw a complete rescuing of this drop, interestingly enough. As noted from our previous research with TISA²⁹, BMP2 is essential for ectopic bone formation, so the final group of TISA/PD without BMP2 has little to no new bone growth as expected. This discrepancy from in vitro to in vivo data could arise from several potential sources, starting with drug binding activity – firstly, as BMP2 binding/retention could be affected by reduction, this may allow for a high variance in binding and actual release from scaffolds, potentially sequestering much of the BMP2 on TISA/PD scaffolds and blocking pharmacologic effect. This could also reduce the initial burst release, which is another well-noted beneficial factor to new bone formation in vivo^40,41. As seen from the in vitro release curves, TISA scaffolds released the most significant overall amount of BMP2, likely helping explain the new bone volume from in vivo. However, the TISA/PD and TISA/rPD BMP2 release curves are very similar, suggesting some other factor such as reduction chemistry may be at play. Secondly, ROS buildup has spatiotemporal concerns in vivo – ROS levels in tissue may fluctuate heavily at various time points during regeneration, making a single point in vitro model a poor predictor of scavenging potential in vivo⁴². Not only does the foreign body response produce significant ROS, but biomaterial degradation byproduct can release particles that induce ROS generation in cells, or impede antioxidant production by tissue surrounding implants⁴³. rPD NPs may also have decreased potency in ROS scavenging over time, as the in vitro experiments are short-term compared to the 4+ week in vivo experiments – rPD NPs could potentially be oxidized by the ROS themselves as one possible mechanism for this loss in potency. Attacking ROS alone is also potentially over-simplifying the aged regenerative niche, as chronic systemic and local inflammation certainly contribute to the challenges facing aging-induced bone repair.

In vitro trends of TISA/rPD scaffolds looked very promising to reduce ROS levels and improve osteogenic differentiation, however, some limitations should be addressed in future lines of this research. Young/healthy mice as a control versus aged mice will help draw a better understanding of the differences in regeneration in an aging-complicated model. Additionally, a bone defect model would be a more clinically relevant model for new bone formation and would more likely be accompanied by the increased ROS environment/inflammatory factors that should be considered in the material design for use in aged animal models. To better understand the impact of ROS scavenging in vivo, tracking changes of local ROS in aging mice during the repair process would be an important gauge to determine sustained effect of scavenging ability of PD or rPD NPs. And finally, it is very strongly suggested that the aged in vivo regenerative niche is more complex and nuanced than a simple increase in ROS; additional consideration to chronic inflammatory environment, senescent cell populations, and other factors should be examined in a deeper level through future studies via improvements of drug delivery and biomaterial approaches.

5. Conclusion

To develop a more effective biomaterial-based therapeutic intervention for aging-compromised bone repair, polydopamine-nanoparticle coated TISA scaffolds were synthesized to tackle the challenges of a high ROS environment during bone repair. After nanoparticle reduction, TISA/rPD scaffolds showed significant improvements in the areas of osteoblastic differentiation and hydrogen peroxide scavenging capacity, indicating the scaffolds potential for further in vivo experiments. TISA/rPD scaffolds maintained strong new BMP2-induced bone formation capacity, although not significantly improved compared to TISA, where further refinement of the model and considerations for other parameters in the aging-compromised in vivo model would likely yield stronger improvements in future lines of experiments.

Data Availability Statement

Data that support the findings of this study are available upon request from the authors upon reasonable request.