A Bioactive Collagen Membrane that Enhances Bone Regeneration

Abstract

Purpose:

Membranes are an integral component of guided bone regeneration protocols. This pre-clinical study was aimed at enhancing the bioactivity of collagen membranes by incorporating plasmid DNA (pDNA) or chemically modified RNA (cmRNA) encoding bone morphogenetic protein - 9 (BMP-9). In addition, we also endeavored to harness the regenerative potential of the periosteum by creating perforations in the membrane.

Materials and Methods:

Nanoplexes of polyethylenimine (PEI)-nucleic acids (PEI-pDNA or PEI-cmRNA encoding BMP-9) were incorporated into commercially obtained and perforated collagen membranes (PCM) to produce PCM-pDNA(BMP-9) or PCM-cmRNA(BMP-9). After structural characterization, the biodegradation kinetics of PCM, PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9) were assessed in simulated body fluid in vitro. Using a 24-well transwell plate system with bone marrow stromal cells (BMSCs) in the lower chamber and the PCM to be tested in the upper chamber, the in vitro bioactivity of different PCMs was evaluated by measuring various markers for osteogenesis in BMSCs. Alkaline phosphatase activity was assessed in BMSCs, after 7 and 11 days of exposure to PCM, PCM-pDNA(BMP-9) or PCM-cmRNA(BMP-9). Similarly, calcium deposition and Alizarin red staining in BMSCs were assessed after 14 days of exposure to the three different types of PCM. PCMs were then tested in vivo using the calvarial defect model in rats. After 4 weeks, animals were euthanized and bone specimens were harvested for micro-computed tomography and histological assessments.

Results:

Incorporation of pDNA or cmRNA did not alter the biodegradation profile of PCMs. Alkaline phosphatase activity trended towards being higher in BMSCs exposed to PCM-cmRNA(BMP-9) or PCM-pDNA(BMP-9), when compared to BMSCs alone. Similar trends were observed when calcium deposition and alizarin red staining was evaluated. Calvarial bone defects treated with PCM-cmRNA(BMP-9) resulted in significantly higher bone volume/total volume % (BV/TV%), when compared to empty defects and trended towards being higher than defects treated with PCM-pDNA(BMP-9) and PCM alone.

Conclusion:

We demonstrate for the first time that resorbable PCM can be utilized to efficiently deliver pDNA and cmRNA of interest. The released pDNA and cmRNA encoding BMP-9 in this assessment was shown to be functional in vitro as well as in vivo.

Introduction

Resorbable collagen membranes have become an integral part of the standard armamentarium for treating periodontal and bone defects using guided tissue regeneration (GTR) and guided bone regeneration (GBR) approaches [1]. Apart from the primary functions of space maintenance and epithelial exclusion, it is now clear that membranes can effectively be utilized as drug or biomolecule delivery devices. Numerous studies in the last decade had shown that membranes can be utilized to deliver antibiotics ranging from tetracycline to metronidazole [2–6]. Recently, a handful of studies demonstrated that membranes can also be utilized to deliver regenerative factors in a sustained fashion. The researchers evaluated the in vitro efficacy of a collagen membrane containing recombinant platelet derived growth factor – BB (PDGF-BB) in pre-osteoblastic (MC3T3-E1) cells [7]. A sustained release pattern of the growth factor for up to 3 weeks was observed that significantly increased the alkaline phosphatase activity and expression of key osteogenic genes such as RUNX2 in MC3T3-E1 cells [7]. In a subsequent study, the same group tested collagen membranes containing PDGF-BB and growth and differentiation factor-5 (GDF-5) [8]. In addition to in vitro evaluation, they also tested the bone regenerative capacity of the growth factor releasing membrane using surgically created circular (~4 mm diameter) mandibular bone defects in rats. Using micro CT evaluation and histology, they clearly demonstrated significant new bone formation and increased bone density in bone defects covered with GDF-5 releasing membrane, when compared to the other groups tested [8].

Though the earlier studies point to the feasibility of delivering growth factors from membranes, there are several drawbacks in using protein factors including a lack of continued bioavailability of these proteins at the site of implantation [9]. This is due in-part to the inherent instability of proteins in the tissue milieu, conformational changes, degradation and the lack of an optimal delivery system that releases these factors for an adequate period of time [9]. The supraphysiological dosage of these factors to compensate for their instability raises concerns with regard to toxicity [10]. In addition, being recombinant proteins, they are expensive and thus alternative therapeutic approaches are sought.

Delivering DNA or RNA encoding the relevant regenerative factor, instead of its protein form can potentially overcome the above listed limitations associated with protein based approaches [11]. In a proof of concept study, involving the non-viral delivery of pDNA encoding PDGF-B from a collagen scaffold, we have recently demonstrated the feasibility of regenerating bone in calvarial defects in rats [12]. We also subsequently demonstrated that co-delivery of pDNA encoding for fibroblast growth factor-2 (FGF-2) and bone morphogenetic protein-2 (BMP-2) embedded in collagen scaffolds could be an effective way of promoting bone regeneration in diaphyseal long bone radial defects in a diabetic rabbit model [13]. Using a similar experimental approach, our research group was the first to report the efficacy of mRNA therapeutics in bone tissue engineering using chemically-modified mRNA (cmRNA) encoding BMP-2 [cmRNA(BMP-2)] as the active therapeutic agent [14]. Using cmRNA in tissue engineering is a novel approach which has the potential to overcome the limitations of DNA (gene) therapy since cmRNA does not require delivery to the nucleus. In a separate study, we also compared the bone regeneration efficacy of collagen sponges containing cmRNA(BMP-9) with collagen sponges containing cmRNA(BMP-2) and reported superiority of the former, although the difference was not statistically significant [15].

Delivering pDNA or cmRNA encoding a regenerative molecule via a collagen membrane is a positive step towards making these membranes bioactive at the same time as addressing the limitations of protein based approaches explored earlier. The membranes have intricate structural properties that will allow them to be effective pDNA or cmRNA carriers capable of sustained release in order to benefit functional regeneration. It is also clear from recent investigations that by perforating the collagen membranes and by allowing progenitor cells from the periosteum and connective tissue to pass through into the regenerative space, clinical outcomes in GTR and GBR approaches can be improved [16–19] . In this proof of concept study, our objectives were to explore the possibility of incorporating pDNA or cmRNA encoding BMP-9 into PCMs and assess their bone regenerating properties in vitro and in vivo.

Material and Methods

Materials

The GenElute^™ HP endotoxin-free plasmid maxiprep kit, PEI (branched; mol. wt. 25 kDa), and the sodium thiosulfate were purchased from Sigma-Aldrich® (St. Louis, MO). Plasmid DNA encoding human BMP-9 (ORF size 1290 bp) possessing a cytomegalovirus promoter/enhancer was provided by Origene Technologies, Inc. (Rockville, MD). PolyA-120 containing T7 pVAX1 was provided by Life Technologies (Madison, WI). The RNA-easy kit was purchased from Qiagen Inc. The 18S-rRNA and the TaqMan Reverse Transcription Reagents were provided by Applied Biosystems (Foster City, CA). Bovine resorbable bilayer collagen (type I) membranes (Bio-Gide®) were purchased from Geistlich Pharma (Princeton, NJ). Colorimetric alkaline phosphatase (ALP) assay kit and colorimetric calcium assay kit were obtained from Abcam (Cambridge, MA). Quant-iT PicoGreen dsDNA Assay Kit and Quant-iT RNA Assay Kit were obtained from Molecular Probes by Life Technologies (Waltham, MA). Human bone marrow stromal cells (BMSCs) were obtained from American Type Culture Collection (ATCC®, Manassas, VA). Dulbecco’s modified Eagle’s medium (DMEM), trypsin-EDTA (0.25%, 1× solution) and Dulbecco’s phosphate buffered saline (PBS) were purchased from Gibco® (Invitrogen™, Grand Island, NY). Fetal bovine serum (FBS) was purchased from Atlanta Biologicals® (Lawrenceville, GA). Gentamycin sulfate (50 mg/ml) was purchased from Mediatech Inc. (Manassas, VA). Any other solvents or chemicals used that are not mentioned above were purchased from Sigma Aldrich and were reagent grade.

Bone marrow stromal cells (BMSCs)

BMSCs were maintained in DMEM (supplemented with 10% fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA), 1mM Glutamax^™ (Gibco), 1mM sodium pyruvate (Gibco), and 1% gentamycin (50 μg/ml)) in a humidified incubator at 37°C and 5% CO₂ (Sanyo Scientific Auto flow, IR direct heat CO₂ incubator). BMSCs were passaged using 0.25% trypsin-EDTA (Life Technologies). In this study, BMSCs were used at passages 3 to 4. Cells were cultured on 75 cm² polystyrene cell culture flasks (Corning, NY, USA).

Isolation and purification of pDNA encoding BMP-9 (pDNA(BMP-9))

DH5α^™ bacteria (E. coli; chemically competent) were transformed with pDNA(BMP-9) and then grown in Lennox L Broth (LB Broth) overnight at 37°C in a shaking incubator at 300 rpm. pDNA(BMP-9) was then purified using a GenElute™ HP maxiprep kit. The quality of the extracted pDNA was assessed via a NanoDrop 2000 UV-Vis Spectrophotometer (Thermoscientific, Wilmington, DE) by determining the absorbance ratio (A₂₆₀/A₂₈₀ nm). A₂₆₀ was used to determine pDNA concentrations and yields.

Preparation of cmRNA encoding BMP-9 (cmRNA(BMP-9))

cmRNA encoding BMP-9 was prepared as described previously [15]. In brief, BMP-9 complementary DNA (cDNA) was subcloned into a PolyA-120 containing T7 pVAX1. Subsequent to linearization with XbaI, pDNA purity was determined. Next using MEGAscript T7 Transcription Kits (Life Technologies) mRNA encoding BMP-9 was generated. The mRNA was capped with the anti-reverse cap analog (ARCA; 7-methyl (3’-O-methyl) GpppGm7G (5’)ppp(5’)G). Next, capped modified ribonucleic acid triphosphates were added to the reaction in the form of pseudouridine-5’-triphosphate and 5-methylcytidine-5’-triphosphate (Ѱ(1.0)m5C(1.0)) at a ratio of 100%. The manufactured mRNA underwent purification and was then assessed for purity and size.

Fabrication and characterization of pDNA(BMP-9) and cmRNA(BMP-9) nanoplexes

Nanoplexes of PEI-pDNA(BMP-9) and PEI-cmRNA(BMP-9) were generated at an N/P ratio of 10, in order to achieve optimal transfection efficacies as elucidated earlier [12]. The nanoplexes were fabricated following our previously reported protocol [14]. Briefly, nanoplexes were prepared by adding 500 μL of PEI (2 mg/mL) solution to 500 μL pDNA or cmRNA solution (encoding BMP-9) containing 50 μg pDNA or cmRNA and vortexed for 30 s, followed by incubation at room temperature for 30 min. The final volume of the complexes utilized for in vitro experiments was 20 μl, containing 1 μg pDNA or cmRNA.

Nanoplexes were characterized for their size, size distribution and zeta-potential using dynamic light scattering measurements (Zetasizer Nano-ZS, Malvern Instruments, Westborough, MA). Using 10 mm diameter polystyrene cuvettes, the particle size and particle size distribution were measured by photon correlation spectroscopy (PCS), using dynamic laser light scattering (4mW He–Ne laser with a fixed wavelength of 633nm, 173° backscatter at 25°C). Using the same instrument, zeta-potentials of the complexes in water were assessed via polystyrene folded capillary cells. Zeta-potential measurements were performed by the laser scattering method (Laser Doppler Micro-electrophoresis, He–Ne laser, 633 nm, and 17° light scatter at 25°C). All measurements were performed in triplicate.

Development of resorbable membranes incorporating PEI-pDNA(BMP-9) and PEI-cmRNA(BMP-9) nanoplexes [PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9)]

Commercially available non-cross linked collagen membranes (Bio-Gide®, Geistlich Pharma, NJ) were uniformly perforated using a microneedle roller (Dermaroller®, 1.0 mm). Briefly, the membranes (40 × 50 mm) were placed on a flat surface and the microneedle roller was rolled in 4 perpendicular lines over the membrane surface, 10 times each for a total of 40 times, with the application of constant pressure. Subsequent to nanoplex preparation (N/P ratio of 10), they were injected into the hand cut circular (6 mm diameter) perforated circular collagen membrane using a sterile 28 gage needle. The nanoplex-embedded collagen membranes were then freeze-dried for 5 h. The thickness of dry bilayer membrane was about 2.5 mm and the wet membrane thickness was approximately 1 mm. The membrane was then cut to required dimensions, prior to experiments.

Surface and morphological characterization of membranes

The gene activated and control membranes were characterized for surface topography using scanning electron microscopy (SEM, Hitachi Model S-4800, Japan). Briefly, the membranes were mounted on a SEM aluminum stub and sputter-coated with conductive gold-palladium using an argon beam K550 sputter coater (Emitech Ltd., Kent, England). Images were captured using the Hitachi S-4800 SEM operated at 3 kV accelerating voltage and a current (I) of 10 A.

Evaluation of the biodegradation properties of the membranes

The biodegradation rate of the gene activated and control membranes was examined in vitro. The membranes (6 mm in diameter) (n=3) were incubated in 15 mL of simulated body fluid at 37 °C, with continuous shaking at 300 rpm. After incubating for 10, 21, and 30 days, the membranes were removed, washed using distilled water and air dried and weighed. The degradation rate of the membranes were expressed as a percentage, calculating the relative weight loss of the specimens using the following formula:

$$MassLoss(\%)=\frac{Dryweight(t0)-Dryweight(tx)}{Dryweight(t0)}\times 100$$

Where t0 = initial dry weight, tx = dry weight after different incubation time (10, 21, 30 days).

Quantification of nanoplex release from the membranes

The 6 mm diameter PCMs containing pDNA(BMP-9) and cmRNA(BMP-9) nanoplexes (10 μg pDNA or cmRNA) were completely submerged in 1× PBS (500 μL) and kept in an incubator shaker at 37 °C and at 300 rpm. At 0.167, 1, 1.167, 3, 4, and 5 h, supernatants were removed and replaced with fresh PBS. The amounts of pDNA(BMP-9) and cmRNA(BMP-9) nanoplexes released from the membrane over time were determined using the PicoGreen and Quant-iT RNA assay kits. Each assay condition was performed in triplicate.

Determination of ALP activity

BMSCs were seeded at 5 × 10⁴ cells/well 24 h prior to treatments into the wells of 24-well plates and then treated with the PCM-pDNA(BMP-9) or PCM-cmRNA(BMP-9). The induction of ALP activity was assessed at 7 and 11 days after transfection. ALP activity was assessed using a colorimetric ALP assay kit. Each assay condition was performed in triplicate and normalized with the concentrations of total cellular proteins. Enzyme activity was expressed as nanomoles of p-nitrophenol produced per minute per milligram of total cellular protein.

Quantitative and qualitative detection of in vitro mineralization

Ca²⁺ detection assay:

Degree of mineralization and released Ca²⁺ ions from BMSCs was quantitatively evaluated by using a calcium detection kit. The BMSCs were seeded at a density of 2 × 10⁶ cells/well in 24-well plates and then assayed using the kit, 14 days post transfection with the PCM-pDNA(BMP-9) or PCM-cmRNA(BMP-9).

Calcium salt sedimentation staining:

BMSCs were seeded into 24-well plates at 5 × 10⁴ cells/well. The calcium salt sedimentation was evaluated using an alizarin red stain as reported previously [14]. Briefly, 14 days post transfection with the PCM-pDNA(BMP-9) or PCM-cmRNA(BMP-9), the cells were rinsed with PBS (1×) for 5 min and then fixed with 10% formalin (Alfa Aesar, Ward Hill, MA) for 10 min. The cells were then rinsed with water, and then incubated with 2% alizarin red S (pH = 4.1 - 4.3) (Sigma-Aldrich) for 10 min. The plate was then rinsed several times with distilled water until no alizarin red was observed seeping into solution. The cells were then dried for imaging. The stained calcium mineral deposits were imaged with a Nikon TE-300 inverted microscope.

Animals and surgical procedure

Fisher (CDF®) male white rats (F344/ DuCrl, 14 weeks old, 250 g) were purchased from Harlan Laboratories (Indianapolis, IN) and housed and cared for in the animal facilities. All animal protocols performed here were approved by the University of Iowa Institutional Animal Care and Use Committee, Iowa. Under anesthesia (intra-peritoneal injection of ketamine (80 mg/kg)-xylazine (8 mg/kg) mixture), a sagittal incision in the scalp was made and the soft tissues were reflected using blunt dissection to expose the calvarium. Two critical-sized (5 mm diameter) calvarial defects were created on the parietal bone, on both sides of the sagittal suture, using a round carbide bur. The created defects were randomly assigned to one of the following groups: 1) Empty defect (n=5), 2) defects covered with PCM-pDNA(BMP-9) (n=5), and 3) defects covered with PCM-cmRNA(BMP-9) (n=6). The membranes were 6 mm in diameter and covered the defects adequately. The rough surface of the membrane containing the bioactive molecules faced the periosteum and the smooth surface was placed facing the defect. Where applicable, the membranes were hand packed into the defects. Next, using sterile silk sutures the incision was closed and meloxicam (2 mg/kg) was administered subcutaneously for post-operative pain management. Rats were euthanized after 4 weeks and the calvarial bone samples were harvested and fixed in neutral buffered formalin (10%) for analyses (fig 1).

Fig. 1.

Schematic illustration of the rat surgical procedure. Rats were euthanized after 4 weeks, and newly formed bone was analyzed using μCT scanning and histological staining.

Micro-CT analysis and histological observation of the regenerated bone

Bone volume and bone connectivity density were quantified using three-dimensional x-ray micro-computed tomography (μCT) imaging as previously described [14]. For histological observations, the fixed specimens (in 10% neutral buffered formalin), were decalcified using a Surgipath Decalcifier II procedure. Then samples were dehydrated using increasing concentrations of ethanol, and then treated with xylene (Merck, Germany), and embedded in paraffin. Sagittal histological sections (5 μm thick) were made in the central part of the wound and samples were collected on Superfrost Plus Slides (Fisher Scientific, Pittsburgh, PA). Sections were deparaffinized and stained with Hematoxylin-Eosin (H&E). To qualitatively evaluate in vivo bone regeneration after 4 weeks; six sections, representing the central area of each defect, including intact native bone margins surrounding the reconstructed defects, were used to assess new bone formation and bridging of the created defect. The Olympus Stereoscope SZX12 and an Olympus BX61 microscope, both equipped with a digital camera, were utilized for bright field examination of the slides.

Data presentation and statistical analysis

Numerical data are represented as mean ±standard error of the mean (SEM). All statistical analyses were performed using statistical and graphing software, GraphPad Prism version 6 for windows (GraphPad Software Inc., San Diego, CA). Unless stated otherwise, one-way ANOVA analysis of variance followed by Tukey’s multiple comparison post-test was used to compare all groups pair-wise. Differences were considered significant at p-values that were less than or equal to 0.05.

Results and Discussion

In this study for the first time we provided critical information on the feasibility and effectiveness of a bioactive membrane-based non-viral gene delivery system for bone regeneration with far reaching clinical applications in dentistry and medicine.

Nanoplex formation and characterization

The isolated cmRNA and pDNA were complexed with the cationic PEI into nanoplexes in order to facilitate the cellular uptake and consequently achieve higher transfection efficacy. The PEI-cmRNA(BMP-9) or PEI-pDNA(BMP-9) nanoplexes were prepared at the amine to phosphate (N/P) ratio of 10 which has been shown to have highest transfection efficacy with the lowest cell toxicity [12]. The size of the nanoplexes ranged from 90 to 120 nm with a narrow size distribution, with the average polydispersity index (PDI) less than 0.2. The zeta-potential of the complexes was in the range of +33 to +45 mV. Nanosized complexes smaller than 150 nm with a positive zeta-potential are both considered favorable factors for efficient in vitro cellular uptake [20] and in vivo distribution and diffusion [21].

SEM analysis of perforated collagen membranes

Using SEM, the PCMs and gene activated PCMs were characterized for their surface topography before and after lyophilization. The SEM images showed highly interconnecting structures and the lyophilization did not appear to have any significant effect on the morphology and porosity of the membranes (fig. 2). In addition, there was no notable difference between the empty PCMs and gene activated PCMs. The PCMs provide a framework for growth and maintain space, excluding soft tissue from the defect site.

Fig. 2.

Representative digital (a) and SEM (b and c) images of the perforated collagen membranes before and after lyophilization (Scale bar, 1.00 mm; Inset: SEM images with scale bar = 20.0 μm).

Evaluation of the biodegradation properties of the membranes

The results of in vitro degradation rate of the empty PCMs, gene and RNA or transcript-activated PCMs are shown in figure 3. Biodegradation is likely to occur by dissolution of the membrane struts exposed to simulated body fluid. Initially (approximately up to 8 days), the degradation rate of the membranes was relatively slow, however, subsequent to day 8 the rate of degradation dramatically increased. . The mass loss was observed to be independent of the loaded component, as there was no notable difference between the tested PCMs.

Fig. 3.

Mass loss from the empty PCM and gene activated membranes over time (n=3).

Release profile of nanoplexes from PCMs

The ability of the PCMs to release and deliver nanoplexes was assessed by measuring cumulative release of pDNA(BMP-9) and cmRNA(BMP-9) nanoplexes from PCMs in PBS. As shown in Figure 4, pDNA(BMP-9) and cmRNA(BMP-9) nanoplexes embedded in PCMs had similar release profiles. Approximately 65–75% of the incorporated pDNA(BMP-9) and cmRNA(BMP-9) nanoplexes were released within the 24 h followed by sustained release of plasmids over the next 4 days.

Fig. 4. Cumulative release of pDNA(BMP-9) and cmRNA(BMP-9) nucleic acids from PCMs.

Data were extracted from three independent experiments (n=3) and is expressed as mean ± SEM.

Determination of ALP activity

The ability of pDNA(BMP-9) and cmRNA(BMP-9) nanoplexes embedded in PCMs to promote osteoblastic differentiation in BMSCs was assessed by measuring the earlier osteogenic marker, ALP, on days 7 and 11. Results demonstrated that ALP activity was significantly increased in both treatment groups by day 11 compared to the control group (untreated BMSCs), and this enhanced ALP activity was also noted on day 7 (albeit not significantly so for pDNA(BMP-9)) (fig. 5).

Fig. 5. Induction ALP activity by BMSCs treated with PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9).

Data were pooled from three independent experiments in quadruplicate (n=4) and is expressed as mean ± SEM. Statistical differences between treated and untreated BMSCs groups were evaluated by one way ANOVA followed by Tukey’s post-test (**p < 0.01; *p < 0.05).

Quantitative and qualitative detection of in vitro mineralization

Extracellular matrix calcification is a later stage marker of the osteoblastic differentiation. Calcium deposition by BMSCs in the presence of the PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9) was evaluated after 14 days using a calcium colorimetric detection assay kit (quantitative analysis) and alizarin red staining (qualitative analysis).

Quantitative analysis demonstrated that calcium concentrations were increased significantly in both treatment groups after 14 days, compared to the control (untreated BMSCs) group (p value <0.01, Tukey’s multiple comparison post-test) (fig. 6.). In addition, calcium concentrations for BMSCs exposed to PCM-cmRNA(BMP-9) were significantly higher compared to BMSCs exposed to PCM-pDNA(BMP-9) (p value <0.05).

Fig. 6. Quantitative calcium mineralization assay.

Calcium colorimetric detection assay to evaluate calcium mineralization levels was performed on BMSC suspensions 14 days post-treatment with PCMs loaded with indicated nanoplex treatment (n = 3). a = statistically significant difference between control and pDNA(BMP-9), b = statistically significant difference between control (BMSC) and cmRNA(BMP-9), and c = statistically significant difference between cmRNA-BMP-9 and pDNA-BMP-9. Statistical differences were evaluated by one way ANOVA followed by Tukey’s post-test (**p < 0.01; *p < 0.05) and values are expressed as mean ± SEM.

The qualitative study showed the BMSCs treated with PCM-cmRNA(BMP-9) appeared to stain red more intensely compared to BMSCs treated with PCM-pDNA(BMP-9) and untreated BMSCs. Image J analysis was used to quantify the optical density of the alizarin red staining (fig. 7).

Fig. 7. Qualitative calcium mineralization assay.

Alizarin red staining was performed on BMSCs, 14 days post treatment with PCMs loaded with indicated treatment. (a) BMSCs exposed to PCM-cmRNA(BMP-9) (b) BMSCs exposed to PCM-pDNA(BMP-9) (c) BMSCs untreated control (empty PCM, and (d) Integrated optical density of the alizarin red staining as determined by Image J analysis. Scale bar 1.00 mm.

Micro-CT analysis of regenerated bone

The effect of PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9) on osteogenesis was studied in critical sized calvarial bone defects model in rats. Four weeks after in vivo implantation of PCMs, rats were sacrificed and regenerated bone tissue was assessed for its volume and connectivity density using micro-computed tomography (μCT) scans.

Qualitative analysis demonstrated an increase in mineralized bone matrix in groups treated with PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9), compared to control groups (empty defect without PCM) (Fig.8.). Quantitative assessment of μCT images of the newly formed bone was used to report the average bone volume of the regenerated bone as a fraction of the total tissue volume of interest (BV/TV) and connectivity density. The distribution of BV/TV of defects treated with PCM-cmRNA(BMP-9) was 1.12-fold higher compared to defects treated with PCM-pDNA(BMP-9), albeit not significantly. However, there was significantly increased BV/TV when both treatment groups were compared to the empty defect (control) group (Fig. 9.). The connectivity density of the regenerated bone in the defects treated with PCM-cmRNA(BMP-9) or PCM-pDNA(BMP-9) was 10-fold higher and 19-fold higher, respectively, when compared to the control group. There was no significant difference between two treatment groups with respect to connectivity density (Fig. 9.).

Fig.8. Qualitative assessment of in vivo bone repair.

Representative μCT images showing the level of bone tissue regenerated after 4 weeks in calvarial defects treated with: (a) defect left empty without PCM, (b) PCM-pDNA(BMP-9), and (c) PCM-cmRNA(BMP-9). Scale bars, Intact images: 1.00 mm and 3.8 mm diameter region of interest: 100 μm.

Fig.9. Quantitative analysis of in vivo bone repair.

(a) Evaluation of bone volume fraction and (b) connectivity density of regenerated bone after 4 weeks in defects treated with: empty defects (n=5), PCM-pDNA(BMP-9) (n=5), and PCM-cmRNA(BMP-9) (n=6). Significant differences between PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9) treatments and empty defect were evaluated by one way ANOVA followed by Tukey’s post-test (**p < 0.01; *p < 0.05). Values are expressed as mean ± SEM.

Histological observation

Images of bone sections stained with H&E further validated the μCT results. There was complete bridging of the calvarial defects as evidenced by the mineralized regenerated bone tissue in both the PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9) treated groups. The control group failed to form new mineralized bone and mostly promoted soft tissue formation. (Fig. 10). In the process of bone healing, osteoid formation and bone matrix mineralization is heavily influenced by BMP-9 signaling [22]. We postulate that the PCMs provided structural support and a platform for gradual release of nanoplexes to transfect nearby cells. Cells transfected with BMP-9 produce greater levels of BMP-9 which promote osteogenesis and osteoprogenitor recruitment into the membrane that subsequently differentiate [23].

Fig.10.

Illustrative histology sections indicating the extent of nascent bone formation at the defect sites at 4 weeks in response to (a) empty defect, (b) PCM-pDNA(BMP-9), and (c) PCM-cmRNA(BMP-9). NB-native bone and RB-regenerated bone. The complete bridging of new bone is indicated by the arrows. Scale bar, 50 μm.

In this proof of concept study, we demonstrated the feasibility of using collagen membranes as an effective biomolecules (pDNA or cmRNA) delivery device, in addition to their normal functions of space maintenance and soft tissue exclusion. Our results confirmed that the pDNA and cmRNA released from the membrane was able to successfully transfect native cells at the implanted site. This strategy, which utilizes local and sustained delivery of biomolecules at the bone-periosteal interface, has significant translational potential and clinical relevance. Our ultimate goal is to use such bioactive membranes to regenerate bony defects in the craniofacial complex, such as alveolar ridge defects to gain bone volume, prior to dental implant placement (Figure 11).

Fig.11.

Schematic of the bioactive membrane based non-viral gene or transcript delivery system functioning at bone-periosteal interface.

Conclusions

In summary, we demonstrated that perforated bioactive membranes can act as non-viral gene delivery devices for tissue regeneration purposes. The results from the present study indicated that PCM-pDNA(BMP-9) and PCM-cmRNA(BMP-9), nanoplexes enhanced osteogenic differentiation compared to controls; as evidenced from the higher ALP activity and calcium mineralization in vitro, and significantly higher distribution of bone volume fraction in defects in vivo.