Plasmid encoding miRNA-200c delivered by CaCO₃-based nanoparticles enhances rat alveolar bone formation

Abstract

Aim

miRNAs have been shown to improve the restoration of craniofacial bone defects. This work aimed to enhance transfection efficiency and miR-200c-induced bone formation in alveolar bone defects via plasmid DNA encoding miR-200c delivery from CaCO₃ nanoparticles.

Materials & methods

The CaCO₃/miR-200c delivery system was evaluated in vitro (microscopy, transfection efficiency, biocompatibility) and miR-200c-induced in vivo alveolar bone formation was assessed via micro-computed tomography and histology.

Results

CaCO₃ nanoparticles significantly enhanced the transfection of plasmid DNA encoding miR-200c without inflammatory effects and sustained miR-200c expression. CaCO₃/miR-200c treatment in vivo significantly increased bone formation in rat alveolar bone defects.

Conclusion

CaCO₃ nanoparticles enhance miR-200c delivery to accelerate alveolar bone formation, thereby demonstrating the application of CaCO₃/miR-200c to craniofacial bone defects.

Plain language summary

The restoration of craniofacial bone defects is surgically complex and requires the combined use of bone grafts and regenerative biomaterials. miRNAs are small biomolecules that have been shown to improve bone regeneration in large bone defects. The aim of this work was to develop a nanoparticle-based delivery system to sustain the release of miRNAs to improve the restoration of craniofacial bone defects. The results of this study demonstrated that CaCO₃ nanoparticles extend the delivery of miRNAs to enhance bone formation in a craniofacial bone defect animal model in a therapeutically safe manner that improves upon conventional nanoparticle materials for bone regeneration. The findings attest to the regenerative properties of miRNAs and further indicate the potential application of CaCO₃-based nanoparticles in restoring large bone defects.

Oral and craniofacial bone reconstruction after oral cavity cancer excision represents a significant challenge for surgeons. More than 20% of oral cancer patients present with bone invasion that requires maxillectomy or mandibulectomy and subsequent osseous reconstruction [1–3]. Vascularized bone grafts and osteocutaneous free flaps are the current gold-standard treatment for surgical reconstruction [4–6]. However, the surgical challenges associated with vascularized bony reconstructions are technically demanding, with craniofacial reconstructive surgeries demonstrating a nearly 40% failure and complication rate due to infections, hardware malfunction, osseous malunion and autologous bone resorption [7]. When failure occurs, patients requiring a second surgery are at further risk of complications, including increased donor site morbidity and risk of infection [8]. In addition, patients suffering from oral cancer with oral-craniofacial reconstruction often require adjuvant radiotherapy. Radiotherapy may impair bone formation and graft osseointegration and ultimately lead to graft loss [9]. Although exogenous osteogenic factors, such as US FDA-approved bone morphogenetic proteins (BMPs; RHBMP-2/7), are incorporated into grafts to promote bone regeneration at the defect site, BMPs are unstable and their short half-life requires supraphysiological dose administration, which has been linked to adverse side effects, including ectopic bone formation, bone resorption and other pro-oncogenic concerns [10–14]. Furthermore, in patients treated for oral and neck cancers, approximately two-thirds of all patients experience tumor relapse, with most cases occurring less than 1 year after primary tumor resection [15]. Thus, a safer and more effective therapeutic with strong bone regeneration and anticancer properties is needed to improve the clinical outcomes of craniofacial reconstruction for oral cancer patients.

In considering agents to combat these regeneration and restoration challenges, miRNAs (miRs) have recently emerged as tools to promote osteogenic differentiation and bone formation [16,17]. miRs are small, noncoding RNAs that play crucial roles in craniofacial bone metabolism and remodeling by epigenetically regulating biological processes, including cell cycle regulation, differentiation, migration and apoptosis [18,19]. Moreover, by targeting inhibitory regulators of osteogenic differentiation, some specific miRs have been found to upregulate the WNT/β-catenin and BMP/SMAD signaling pathways to promote bone formation [20]. In particular, miR-200c, a member of the miR-200 family, has been found to have anticancer and enhanced bone regeneration capabilities that inhibit bone loss, improve osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs), and promote in vivo bone regeneration in critical-sized calvarial defect animal models [21,22]. Previous studies have found that miR-200c effectively improves osteogenic differentiation and bone formation by targeting multiple components of the osteogenesis pathway, including NOGGIN (a BMP antagonist), SOX2 (a WNT signaling antagonist), AKT/β-catenin and SMAD7 [20,23]. miR-200c additionally possesses potent anticancer properties that inhibit the epithelial-to-mesenchymal transition in oral cancer initiation and metastases [24–27]. Furthermore, miR-200c has been found to suppress multiple proinflammatory cytokines and mediators, including IL-6, IL-8, CCL-5 and IFRγ-1, that activate bone resorption by stimulating osteoclastogenesis [28–31]. The authors' previous work has effectively demonstrated the osteogenic and anti-inflammatory properties of using naked plasmid DNA (pDNA) encoding miR-200c to improve in vivo bone formation and regeneration in calvarial animal models; however, the capacity of miR-200c to regenerate bone in defect models that more appropriately mimic the conditions of oral-craniofacial reconstructive surgery, as in an alveolar bone defect model, has yet to be evaluated.

The restoration of large bone defects from tumor excisions would greatly benefit from an efficient gene delivery system that can enhance transfection and effectiveness. As an alternative to growth factor-mediated bone regeneration, viral and nonviral gene delivery methods have been used to deliver vectors carrying plasmids that encode osteogenic molecules, such as miR-200c [32]. Viral vectors have a high transfection efficiency but are costly and difficult to produce, and they are further limited by issues with carcinogenesis and immunogenesis [33]. Recently, nonviral vectors, such as high-molecular weight, branched polyethylenimine (PEI), have become increasingly popular for pDNA delivery, as they circumvent the safety concerns and limitations of viral methods [34,35]. PEI transfection efficiencies are high among the nonviral methods; however, PEI has been found to be cytotoxic and shows limited biodegradability [36]. Alternatively, calcium phosphates (CaPs) have been used for decades as effective gene carriers to introduce pDNA into many cell types through CaP/pDNA coprecipitation [37–39]. More recently, CaCO₃, a well-studied mineral biomaterial, has demonstrated excellent osteoconductivity, biocompatibility and biodegradability and possesses significant applicability for gene delivery in bone regeneration [40,41]. As a nanoparticle, CaCO₃ can easily be coprecipitated with pDNA to form a nanocomplex system for gene delivery. Furthermore, coprecipitation of CaCO₃ with protamine sulfate (PS), a clinically used antidote for heparin-induced anticoagulation, has been found to substantially improve CaCO₃-mediated pDNA transfer into cells [42–44]; however, its efficacy for gene delivery and tissue regeneration in vivo has not been studied. This evidence strongly suggests that CaCO₃/PS nanoparticle composites have the potential to deliver pDNA encoding miR-200c to promote osteogenic differentiation and bone formation to restore bone defects caused by craniofacial reconstructive surgeries.

In this study, the authors investigated bone formation induced by CaCO₃/miR-200c coprecipitation under both in vitro and in vivo conditions. They found that the CaCO₃/miR-200c nanocomplexes significantly increased miR-200c expression in human embryonic palatal mesenchymal (HEPM) cells and sustained the overexpression with no inflammatory or cytotoxic effects. They also found that their CaCO₃/miR-200c nanocomplexes significantly increased in vivo bone formation in their alveolar bone defect model. These findings attest to the regenerative capacity of miR-200c in restoring large bone defects and further indicate its potential application to the treatment of craniofacial bone defects from reconstructive surgeries.

Materials & methods

Preparation & characterization of the CaCO₃/pDNA nanocomplex

CaCO₃/pDNA nanocomplexes were created by incorporating pDNA into CaCO₃/PS coprecipitated nanoparticles via a self-assembly method based on previous literature [38,42–46]. Two separately prepared solutions were combined to form the nanocomplexes. Specifically, pDNA encoding miR-200c at different doses was mixed with 16 μl of CaCl₂ (0.5 M) and then diluted to a final volume of 60 μl (solution A). Solution B was prepared by mixing 16 μl of NaCO₃ (0.01 M) with PS at different doses and then diluting it to a total volume of 40 μl. Solutions A and B were mixed to produce the CaCO₃/PS/miR-200c nanocomplexes. The CaCO₃/pDNA nanocomplexes were visualized using field-emission scanning electron microscopy (FE-SEM; Hitachi S-4800, Tokyo, Japan).

The hydrodynamic particle diameter and zeta potential of the CaCO₃/pDNA nanocomplexes were measured using dynamic light scattering (DLS; Zetasizer Nano ZS; Malvern Instruments, Worcestershire, UK). To assess hydrodynamic diameter, a 0.5 mg/ml solution of the CaCO₃/pDNA nanocomplexes in a 5 mM HEPES buffer ([4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid]; Thermo Fisher Scientific, MA, USA) at pH 7 was placed in a disposable cuvette for DLS measurements in backscattering mode (173°). Calcite was assumed to be the primary crystal form of the nanoparticles, with an ordinary refractive index of 1.486 and an absorption of 0.1. Water at 20°C was the main component of the dispersant and resulted in a viscosity of 1.0031 cP and refractive index of 1.330. DLS data revealed both primary and aggregated nanoparticle distributions. The primary nanoparticle distribution was analyzed using a modified Gaussian function in Origin 2018. To assess the zeta potential of the nanocomplexes, surface potential was calculated as the average of two measurements at pH 7 and pH 9 at 20°C using the same DLS device and operating parameters, only substituting the cuvette to a DTS1060 cell.

Transfection efficiency of pDNA encoding miR-200c influenced by ratios of CaCO₃:PS

The authors used preosteoblast, HEPM cells (ATCC, VA, USA), to determine the transfection efficiency of pDNA encoding miR-200c using the CaCO₃/pDNA nanocomplex delivery system. HEPM cells were cultured and expanded in DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin (Life Technologies, NY, USA) at 37°C, 5% CO₂. To determine the influence of CaCO₃:PS ratio on miR-200c transfection efficiency, the authors transfected HEPM cells (2 × 10⁴ cells/well; 12-well plate) with CaCO₃/PS nanoparticles incorporating pDNA encoding miR-200c at different CaCO₃:PS ratios (1:0.03125, 1:0.125, 1:0.25 and 1:0.5). For CaCO₃/miR-200c transfection, HEPM cells were transfected with CaCO₃/miR-200c nanocomplex (1 μg) in 1 ml Gibco™ Minimal Essential Medium (Opti-MEM^®; Thermo Fisher Scientific) for 16 h, after which the transfecting medium was replaced with fresh DMEM Complete Medium for continued cell culture. After 3 days, the transfection efficiency of the CaCO₃/miR-200c nanocomplexes with differing PS ratios was assessed via quantitative reverse transcription PCR (qRT-PCR; performed using technical triplicates).

Analysis of transfection efficiency of pDNA encoding miR-200c delivered by CaCO₃-based nanoparticle systems

To determine the transfection efficiency of the authors' CaCO₃/miR-200c nanocomplexes and to compare them with other delivery systems, the authors transfected HEPM cells (2 × 10⁴ cells/well; 12-well plate) with pDNA encoding empty vector (EV) or miR-200c using naked pDNA alone, branched PEI (MW: 25 kDa; Sigma-Aldrich, MO, USA) or CaCO₃/PS coprecipitated nanoparticles. For this study, the seven treatment groups included: 1) untreated control; 2, 3) naked pDNA encoding EV or miR-200c (1 μg); 4, 5) PEI/pDNA nanocomplex encoding EV or miR-200c (1 μg); and 6, 7) CaCO₃/pDNA nanocomplex encoding EV or miR-200c (1 μg).

For naked pDNA transfection, HEPM cells were transfected with naked pDNA encoding EV or miR-200c (1 μg) in 1 ml Opti-MEM^® for 16 h, after which the transfecting medium was replaced with fresh DMEM Complete Medium for continued cell culture. For PEI transfection, the PEI/pDNA (3:1 ratio) nanocomplex was formed by adding 4 μg of PEI solution to 1 μg of pDNA encoding miR-200c solution, and the combined solution was mixed via pipetting for 30 s. The PEI/pDNA mixture was then incubated at room temperature for 30 min to allow for nanocomplex formation. The HEPM cells were transfected using PEI/pDNA nanocomplex encoding EV or miR-200c (1 μg) in 1 ml Opti-MEM^® for 6 h, after which the transfecting media was replaced with fresh DMEM Complete Medium for continued cell culture. For the CaCO₃/pDNA nanocomplex transfection, CaCO₃/pDNA nanocomplexes were formed as previously described using a 1:0.25 CaCO₃:PS ratio, and HEPM cells were transfected with CaCO₃/pDNA nanocomplexes encoding EV or miR-200c (1 μg) in 1 ml Opti-MEM^® for 16 h, after which the transfecting medium was replaced with fresh DMEM Complete Medium for continued cell culture. After 3 days, miR-200c transfection efficiency induced by the different delivery systems was assessed via qRT-PCR, using the methods previously described (performed using technical triplicates).

Analysis of biocompatibility of pDNA encoding miR-200c delivered by CaCO₃-based nanoparticle systems

The influence of the nonviral delivery systems on the viability of transfected HEPM cells was assessed via a MTT cell viability assay kit according to the manufacturer's protocol (Biotium, CA, USA). HEPM cells were seeded into 96-well plates (5 × 10³ cells/well) and transfected, as previously described, with either pDNA encoding EV or miR-200c (0.1 μg) using the nonviral delivery system (either PEI or CaCO₃/PS nanoparticles), with untreated HEPM cells as the control group. The transfected cells were then cultured for 3 days, after which MTT solution was added and absorbance was measured and normalized to the untreated control HEPM cells following manufacturer protocol on a spectrophotometer (SpectraMax iD3; Molecular Devices, CA, USA) at 570 nm, with background absorbance set to 630 nm.

The authors also quantified proinflammatory cytokines induced by nonviral delivery systems. HEPM cells transfected with naked pDNA, PEI or CaCO₃ nanocomplex were cultured in DMEM Complete Medium for 3 days and the mRNA of inflammatory biomarkers for IL-6 and IL-8 was evaluated via qRT-PCR (each treatment performed using technical triplicates).

qRT-PCR analysis for miR-200c & inflammatory biomarker expression

Total cellular RNA from cultured HEPM cells was extracted using a miRNeasy Mini Kit (Qiagen, CA, USA). The concentration and purity of total RNA were quantified using a NanoDrop™ One Microvolume UV-Vis Spectrophotometer (Thermo Fisher Scientific). To measure miR-200c expression, the mirScript II reverse transcription kit and the mirScript SYBR Green PCR Kit (Qiagen) were used and normalized to GAPDH, an internal control for human cells, via a comparative Ct (ΔΔCt) method. Expression of IL-6 and IL-8 was performed on a CFX Connect™ (Bio-Rad, CA, USA) using the SYBER^® Premix Ex Tag™ II Kit (Takara Bio, Inc, Kusatsu, Japan). Gene expression was calculated and compared with GAPDH via a comparative Ct (ΔΔCt) method. The primer sequences are listed in Table 1.

Table 1. . Primer sequences used for in vitro quantitative reverse transcription PCR analyses.

Gene	Forward primer	Reverse primer
GAPDH	5′-CATCACTGCCACCCAGAAGACTG-3′	5′-ATGCCAGTGAGCTTCCCGTTCAG-3′
IL-6	5′-ACTCACCTCTTCAGAACGAATTG-3′	5′-CCATCTTTGGAAGGTTCAGGTTG-3′
IL-8	5′-ACTGAGGATTGAGAGTGGAC-3′	5′-AACCCTCTGCACCCAGTTTTC-3′

Evaluating the release of CaCO₃/miR-200c from collagen sponges

To investigate the differences in release of pDNA from collagen sponges with different delivery systems, the authors quantified pDNA release profiles and the cellular uptake of pDNA encoding miR-200c released from collagen sponges incorporating naked pDNA or CaCO₃/pDNA nanocomplexes across different time points. To study pDNA release, collagen sponges were incorporated with either naked pDNA encoding miR-200c (10 μg) or CaCO₃/pDNA nanocomplexes encoding miR-200c (10 μg). Similarly, to evaluate cellular uptake of released pDNA encoding miR-200c, collagen sponges were incorporated with either naked pDNA (5 μg), or CaCO₃/pDNA nanocomplexes encoding miR-200c (5 μg; n = 3/condition). For both studies, collagen sponges were loaded with either treatment (20 μl/sponge), then lyophilized overnight. Each sponge was then placed into a 1.5 ml tube filled with 500 μl phosphate-buffered saline (PBS) and placed on a shaker to continuously shake at 150 r.p.m. and 4°C for the duration of each study. For the release study, at each time point, half of the PBS solution was removed for dsDNA analysis and fresh PBS solution was added back to the tube. The concentration of pDNA released from the sponges was quantified using a Qubit™ dsDNA HS Assay Kit (Invitrogen, MA, USA) at distinct time points following the manufacturer's protocols. For the cellular uptake study, at each time point, the collagen sponges were removed from the PBS solution and subsequently seeded with HEPM cells (5 × 10⁵ cells/sponge). The cellular uptake of pDNA encoding miR-200c by the HEPM cells in relation to the release of naked pDNA or CaCO₃/pDNA nanocomplexes from the collagen sponges was measured using qRT-PCR after 48 h.

Tooth socket model: surgical preparation & animal care

All in vivo animal experiments were performed under the approval from the Office of Animal Resources at the University of Iowa. The surgical protocols were followed by the policies and guidelines provided by the Institutional Animal Care and Use Committee and all animal surgeries were performed under sterile conditions. All biological agents (pDNAs, CaCO₃ etc.) and collagen sponges were prepared and implanted into 12-week-old male Sprague Dawley rats (Charles River Laboratories, MA, USA) under sterile conditions in a surgical environment. Under general anesthesia using ketamine/xylazine, the maxillary teeth and surrounding tissues were disinfected with 2% iodine tincture. After the first molar (M1) was removed on the left and right sides of the maxilla, a 2-mm-wide by 1 mm-deep hole was created in the alveolar bone of the molar cavity using a 2 mm implant twist drill. The cavity was then filled with an experimental treatment group (n = 3–5/treatment): no treatment (defect control); collagen sponge incorporating CaCO₃/pDNA nanocomplexes encoding EV (0.1 μg); or collagen sponge incorporating CaCO₃/pDNA nanocomplexes encoding miR-200c (0.1 μg). Each animal received one treatment per maxillary side. After treatment designation and implantation, each cavity was sutured closed. The rats were euthanized after 3 weeks, and full maxillary palates containing the M1 alveolar bone implants were harvested. Harvested tissues were first rinsed in PBS, then fixed in 4% paraformaldehyde.

miR-200c tooth socket micro-computed tomography imaging & quantification

Micro-computed tomography (μCT) imaging was performed to evaluate new bone formation within the tooth socket defect space. Images of the tooth socket defects were scanned and reconstructed via a μCT scanner with Bruker software (Skyscan model 1272; Bruker, Kontich, Belgium) using the following acquisition parameters: 70 kV, 142 μA, 0.6 rotation step, 1 mm Al filter and 14 μm pixel size. Dragonfly 2021.3 software (Object Research Systems, Montreal, Canada) was then used to analyze bone formation in the socket defects. Briefly, reconstructed palatal images were loaded into Dragonfly and, using grayscale intensity thresholding, low-density soft tissues were removed and a dataset including all dense bone materials was created. After thresholding, a spherical, 0.75-mm-diameter region of interest (ROI) was placed in the defect space to emulate the defect created by the dental burr. The spherical ROI was cut in half across the native alveolar bone line running across the top of the defect, thus creating a 1-mm-wide by 0.75-mm-deep defect ROI, which was used to evaluate bone growth in the defect region. Bone formation in the defects was collected as a volume percentage of dense bone dataset values within the defect ROI. Bone volume data were collected in this manner, utilizing the same thresholds, across all samples, and representative 2D and 3D images were acquired to visualize bone formation in the defects.

Histological evaluation of in vivo bone formation in tooth socket defects

The explanted maxillas were decalcified for 8 h using a decalcification solution (Decalcifying Solution–Lite; Sigma-Aldrich). The decalcified samples were then cleared with xylene and embedded in paraffin. The entire embedded sample, which included the defect with collagen sponge implant and surrounding maxillary tissue, was cut into 7 μm coronal sections and stained with hematoxylin and eosin and Masson's trichrome stain using standard protocols. Representative sections were selected for staining at distinct intervals throughout the sample, starting from the middle of the sample and working outward at an interval sampling distance of 0.1 mm (n = 5). Corresponding images of the stained tissues were taken using an inverted microscope (Eclipse Ts2; Nikon Instruments, Inc, NY, USA) to examine the bone formation occurring from the treated collagen sponge within the tooth socket defect.

Statistical analysis

Descriptive statistics were conducted for both in vitro and in vivo investigations. A one-way analysis of variance with post hoc Tukey's honestly significant difference test was used to determine whether there was a significant difference between treatment groups for both the in vitro miR-200c transfection and biocompatibility studies. For the in vivo study, a one-way analysis of variance with post hoc Tukey's honestly significant difference test was utilized to evaluate whether there were significant differences in bone formation across treatment groups. The Shapiro–Wilks test was also applied to verify the assumption of normality. All statistical tests completed for the in vitro and in vivo quantifications used a significance level of 0.05, and each graphic depicts mean values and associated standard deviations. Statistical analyses and associated figures were created via GraphPad Prism (version 8.1.2.; GraphPad Software, Inc., CA, USA).

Results

CaCO₃ as a nonviral delivery system efficiently delivers pDNA encoding miR-200c

CaCO₃/miR-200c nanocomplexes were created by incorporating pDNA encoding miR-200c into CaCO₃/PS coprecipitated nanoparticles via a self-assembly method. The CaCO₃/miR-200c nanocomplexes were observed via SEM imaging to have a well-defined, reproducible, rhombohedral geometry characteristic of the calcite crystalline phase of CaCO₃, with the nanocomplexes displaying an average diameter of 32 nm (Figure 1A). Hydrated CaCO₃/miR-200c nanocomplexes exhibited a hydrodynamic diameter of 67 nm with a zeta potential of -5.17 mV (Figure 1B). After characterizing the CaCO₃/miR-200c nanocomplexes, the authors evaluated how the CaCO₃:PS ratio influenced miR-200c transfection of HEPM cells in vitro. They found that the CaCO₃:PS ratio of 1:0.25 statistically significantly increased expression of miR-200c in comparison with untreated control HEPM cells (**p < 0.01) as well as all other tested CaCO₃:PS ratios (*p < 0.05; Figure 1C). No significant differences were observed between the control HEPM cells and all other tested CaCO₃:PS ratios. From these results, the authors investigated the transfection efficiency of the CaCO₃/miR-200c nanocomplexes with the optimized CaCO₃:PS ratio (1:0.25) in comparison with other transfection methods using nonviral delivery systems such as naked pDNA and PEI. The authors found that transfecting HEPM cells with CaCO₃/miR-200c statistically significantly increased miR-200c expression in comparison with untreated control HEPM cells and all other nonviral transfection systems, including naked pDNA and PEI (***p < 0.001; Figure 1D). Furthermore, PEI/miR-200c transfection statistically significantly increased miR-200c expression in HEPM cells compared with untreated control HEPM cells and naked pDNA transfection (*p < 0.05); however, miR-200c expression by PEI/miR-200c transfection was statistically significantly less than that of CaCO₃/miR-200c transfection (*p < 0.05). Transfection with naked pDNA encoding miR-200c increased miR-200c expression in comparison with untreated control HEPM cells and HEPM cells transfected with naked pDNA encoding EV, PEI/EV and CaCO₃/EV; however, this increase in miR-200c expression was not significant. There were additionally no observable differences in miR-200c expression found when comparing untreated HEPM control cells and delivery systems using pDNA encoding EV.

Figure 1. . Characterization of CaCO₃/miR-200c nanocomplex particles and evaluation of miR-200c transfection efficiency using CaCO₃ nanoparticles in comparison with naked plasmid DNA and polyethylenimine transfection.

(A) Scanning electron microscopy image of CaCO₃/miR-200c nanocomplex particles. (B) Analysis of the hydrodynamic particle diameter for the CaCO₃/miR-200c nanocomplex particles measured via dynamic light scattering. (C) Relative expression levels of miR-200c from human embryonic palatal mesenchymal cells cultured for 3 days after transfection with different CaCO₃:protamine sulfate ratios. (D) Relative expression levels of miR-200c from human embryonic palatal mesenchymal cells cultured for 3 days after transfection using different transfection systems. Column means that do not share a letter are statistically significantly different using the post hoc Tukey's honestly significant difference test (p < 0.05; performed in triplicate). Scale bar: 100 nm.

EV: Empty vector; PEI: Polyethylenimine; PS: Protamine sulfate.

The CaCO₃-based nanoparticle delivery system is highly biocompatible

Biocompatibility of the different nonviral delivery systems was assessed via an MTT assay and proinflammatory marker quantification via qRT-PCR using HEPM cells. Cytotoxicity of the PEI and CaCO₃ nanocomplex delivery systems was evaluated by an MTT assay 3 days after initial HEPM cell transfection with each delivery system (Figure 2A). Untreated HEPM cells were used as a control. After 3 days, the viability of HEPM cells treated with the CaCO₃/EV nanocomplex was not significantly reduced in comparison with untreated HEPM control cells. However, HEPM cells transfected with the PEI/EV system exhibited a statistically significant reduction in cell viability after 3 days compared with untreated HEPM cells (**p < 0.01).

Figure 2. . Assessment of inflammatory and cytotoxic effects of nonviral CaCO₃ and polyethylenimine nanoparticle delivery systems.

(A) Evaluation of the biocompatibility of different nonviral delivery systems with human embryonic palatal mesenchymal cells assessed via MTT assay (**p < 0.01; performed in triplicate). (B & C) Normalized fold change of IL-6 (B) and IL-8 (C) transcripts from human embryonic palatal mesenchymal cells 3 days after transfection with different delivery systems.

**p < 0.01; performed in triplicate.

EV: Empty vector; PEI: Polyethylenimine.

Proinflammatory marker expression for IL-6 (Figure 2B) and IL-8 (Figure 2C) was quantified via qRT-PCR 3 days after HEPM cell transfection with naked pDNA encoding EV, PEI/EV or CaCO₃/EV nanocomplex, with untreated HEPM cells as the control. After 3 days of culture, the HEPM cells transfected with the PEI/EV system exhibited a statistically significant increase in IL-6 expression in comparison with untreated HEPM cells and CaCO₃/EV nanocomplex transfected cells (**p < 0.01) and HEPM cells transfected with naked pDNA encoding EV (***p < 0.001). HEPM cells transfected with the CaCO₃/EV nanocomplex did not significantly increase IL-6 expression in comparison with untreated HEPM control cells. Furthermore, after 3 days of culture, the HEPM cells transfected with the PEI/EV system showed an increase in IL-8 expression in comparison with untreated HEPM control cells and HEPM cells transfected with either naked pDNA encoding EV or CaCO₃/EV nanocomplex; however, this increase in IL-8 expression was not significant. HEPM cells transfected with the CaCO₃/EV nanocomplex additionally did not significantly increase IL-8 expression in comparison with untreated HEPM control cells.

CaCO₃-based nanoparticles extend the overexpression of miR-200c in collagen sponges in vitro

pDNA release from collagen sponges incorporating naked pDNA encoding miR-200c or CaCO₃/miR-200c nanocomplex was quantified using a high-sensitivity dsDNA kit (Figure 3A). From the authors' release study, they found that naked pDNA was rapidly released from the collagen sponges, with an approximately 83.78% cumulative release of naked pDNA after 6 h and 88.59% cumulative release after 24 h. However, limited release of miR-200c from the collagen sponges incorporating CaCO₃/miR-200c nanocomplexes was measured after 24 h. For the cellular uptake study, HEPM cells were seeded onto collagen sponges incorporating either naked pDNA encoding miR-200c or CaCO₃/miR-200c nanocomplexes that had been placed in PBS solution to release incorporated components for 0, 6 or 24 h (Figure 3B). The authors found that HEPM cells seeded on the CaCO₃/miR-200c nanocomplex-incorporated collagen sponges significantly increased miR-200c overexpression about ten-times higher than collagen sponges incorporating naked pDNA encoding miR-200c (****p < 0.0001) and approximately 4500-times higher than collagen sponge control samples (****p < 0.0001). Furthermore, the extent of miR-200c overexpression in HEPM cells was not decreased significantly during the extended time period when the CaCO₃/miR-200c nanocomplex-incorporated collagen sponges were immersed in PBS solution. The collagen sponges incorporating naked pDNA encoding miR-200c also significantly increased miR-200c overexpression in the seeded HEPM cells compared with the HEPM cell-seeded collagen control sponges (****p < 0.0001); however, the collagen sponges incorporating naked pDNA encoding miR-200c displayed a significant decrease in miR-200c overexpression in HEPM cells after 24 h in PBS solution compared with 0-h and 6-h time points (**p < 0.01).

Figure 3. . Quantification of naked miR-200c and CaCO₃/miR-200c nanocomplex release from collagen sponges and their effect on miR-200c overexpression.

(A) Release profile of naked plasmid DNA encoding miR-200c and CaCO₃/miR-200c nanocomplexes from collagen sponges over 24 h. (B) miR-200c overexpression in human embryonic palatal mesenchymal cells seeded on collagen sponges incorporating naked plasmid DNA encoding miR-200c or CaCO₃/miR-200c nanocomplexes across different release time points.

**p < 0.01; ****p < 0.0001; n = 2/treatment, performed in duplicate.

ns: Not significant.

CaCO₃/pDNA miR-200c nanocomplexes enhance alveolar bone formation in vivo

The capacity of the CaCO₃/pDNA nanocomplex system to promote in vivo bone formation was evaluated using an M1 tooth socket defect rat model. Collagen sponges incorporated with CaCO₃/EV (1 μg) or CaCO₃/miR-200c (1 μg) were implanted into M1 tooth socket defects in rats for 3 weeks, after which alveolar bone formation was assessed via μCT imaging (Figure 4A) and bone volume percentage quantification (Figure 4B). Through μCT imaging, the authors were able to visualize bone tissue growth within the defect space of the M1 tooth socket. They found that the untreated defect samples visually presented the least amount of bone formation in comparison with the samples treated with collagen sponges incorporating either CaCO₃/EV or CaCO₃/miR-200c. Furthermore, based on μCT imaging alone, the defects treated with CaCO₃/miR-200c displayed the greatest amount of bone formation within the regenerating defect space in comparison with the untreated and CaCO₃/EV treated defects. This visually observable increase in bone formation for the CaCO₃/miR-200c treated defects was quantitatively confirmed using μCT quantitative analysis, in which new bone formation occurring in the defect space was measured as a bone volume percentage. Based on the quantitative analysis, the authors found that defects treated with collagen sponges incorporating CaCO₃/miR-200c had a statistically significantly increased production of new bone formation (50.88%) in comparison with untreated defects (22.29%; *p < 0.05) and defects treated with collagen sponges incorporating CaCO₃/EV (16.53%; **p < 0.01). No significant differences were found in bone volume percentage between the untreated defects and the defects treated with collagen sponges incorporating CaCO₃/EV.

Figure 4. . Micro-computed tomography analysis of bone formation induced by collagen sponges incorporating CaCO₃/miR-200c nanocomplexes in maxillary tooth socket bone defects.

(A) Representative micro-computed tomography images of the top and cross-sectional views displaying new bone formation (green) observed in maxillary tooth socket defects with different treatments 3 weeks post-treatment. (B) Quantitative analysis of bone volume percentage in defects with different treatments.

*p < 0.05; **p < 0.01; n = 3–5.

BV: Bone volume; EV: Empty vector; TV: Tissue volume.

After μCT imaging and quantitative analysis, the explanted palates were sectioned and stained for histological examination. In the hematoxylin and eosin- and Masson's trichrome-stained sections (Figure 5), the authors observed limited areas of scattered bone formation in the untreated defects and the defects treated with collagen sponges incorporating CaCO₃/EV. Instead, for these two groups, they observed more fibrous tissue formation than new bone formation. However, in the defects treated with collagen sponges incorporating CaCO₃/miR-200c, they observed increased amounts of new bone formation within the defect space with minimal fibrous tissue formation.

Figure 5. . Histological analysis of new bone formation induced by collagen sponges incorporating CaCO₃/miR-200c nanocomplexes in maxillary tooth socket bone defects.

Microphotographs of the cross sections of tooth socket defects with different treatments stained by hematoxylin and eosin and Masson's trichrome staining at different magnifications. Scale bars: 1 mm (1×), 200 μm (4×, 10×).

EV: Empty vector; F: Fibrous tissues; NB: New bone.

Discussion

The development of advanced exogenous therapeutics with strong osteogenic properties that enhance bone formation is necessary to effectively treat craniofacial bone defects caused by diseases that require reconstructive surgery. This study, for the first time, revealed that CaCO₃/miR-200c nanocomplexes formed through the coprecipitation of osteogenic miR-200c with CaCO₃ nanoparticles significantly improve cell transfection, increase cellular miR-200c overexpression and enhance bone formation in a rat alveolar bone tooth socket defect model. These findings demonstrate the safe, potent and restorative capabilities of miR-200c when used with CaCO₃ nanoparticles to enhance bone formation in large defects. These results significantly advance the potential of miR-based therapeutics for clinical bone formation.

In overcoming the challenges of growth factor-based regenerative strategies, miRs have emerged as critical osteogenic biomolecules to enhance bone formation. In this study, the authors coprecipitated CaCO₃ with pDNA encoding miR-200c to create CaCO₃/miR-200c nanocomplexes. To characterize the morphology and particle diameter of both dried and hydrated CaCO₃/miR-200c nanocomplexes, the authors utilized SEM imaging and DLS. Using SEM, they found that their CaCO₃/miR-200c nanocomplexes had an average diameter of 32 nm, which is well within the reported 1–100 nm nanoparticle diameter range (Figure 1A). Furthermore, they observed the CaCO₃/miR-200c nanocomplexes to have a characteristic nanoscale size and rhombohedral morphology consistent with the calcite crystalline phase of CaCO₃ using SEM imaging. Several prior studies have determined that the combination of CaCl₂ and Na₂CO₃ precursors in the presence of water without cell culture medium or polyphosphate creates the calcite crystalline phase of CaCO₃ [47–51]. Calcite is thermodynamically stable and is reported to not dissolve in cell culture medium, which the authors observed in this study. Combined with these published findings, the present study's CaCO₃ preparation protocol and SEM images indicate that the CaCO₃ nanocomplexes formed in this study are characteristic of the calcite crystalline phase of CaCO₃; however, further analyses using x-ray diffraction and Fourier transform infrared spectroscopy should be performed to confirm the crystalline phase. Using DLS, the authors determined that the CaCO₃/miR-200c nanocomplexes had a surface charge zeta potential of -5.17 mV (Figure 1B). It is reported that particles with zeta potentials beyond ±30 mV are typically considered stable; however, for particles with zeta potentials between ±30 mV, electrostatic repulsive forces alone are insufficient to promote long-term stability of particles, leading to an increased likelihood for particle aggregation with time [52–54]. The low zeta potential of the present study's CaCO₃ nanocomplexes could cause particles to aggregate over time and at higher particle concentrations, and DLS is reported to have limited accuracy in measuring aggregated samples with high polydispersity indexes [55–57]. Therefore, the authors of the present study used freshly prepared, low-concentration (0.5 mg/ml) CaCO₃ samples in HEPES buffer for the DLS measurements and found that the CaCO₃ nanocomplexes had a hydrodynamic particle diameter size of 67 nm (Figure 1B). They further found that with increased measurement time (i.e., greater than 30 min) and CaCO₃ concentration, the measured polydispersity index increased, indicating the aggregation of CaCO₃ particles. The authors also observed that the solvent composition influenced the tendency of the particles to aggregate; HEPES-buffered CaCO₃ nanocomplexes displayed a low polydispersity index with limited aggregation in comparison with CaCO₃ particles measured in PBS solution that displayed high polydispersity and rapid, uncontrolled aggregation. Evidence from prior studies indicates that pH, the concentration of H⁺ ions and the presence or absence of proteins in the suspending medium greatly influence particle surface potential, stability and ultimately quantification capabilities via DLS [58]. One study found that the presence of bovine serum albumin in cell culture medium increased colloidal stability and hindered particle aggregation [59]. The present study's results further indicate that solvent choice influences particle aggregation and ultimately affects DLS measurement capabilities.

Previous studies utilizing CaCO₃ for gene delivery have coprecipitated CaCO₃ with PS to significantly improve CaCO₃-mediated pDNA transfer into cells. The present study found that the optimal CaCO₃:PS ratio to significantly enhance miR-200c transfection into HEPM cells was 1:0.25 (Figure 1C). Utilizing this optimal ratio, the authors successfully found that the CaCO₃/miR-200c nanocomplexes significantly increased miR-200c transfection into HEPM cells in comparison with PEI and naked pDNA transfection (Figure 1D). PEI transfection also increased miR-200c expression in HEPM cells but to a significantly lesser degree than that of the CaCO₃/miR-200c nanocomplexes. These results demonstrate that through using an optimized CaCO₃:PS ratio, CaCO₃/miR-200c nanocomplexes can be effectively developed to significantly enhance miR-200c transfection. Based on the authors' previous work with miR-200c, they have found that transfection via naked pDNA encoding miR-200c significantly increases miR-200c in both human and rat BMSCs and increases the expression of osteogenic markers, including RUNX2, OCN and OPG. The authors chose to use HEPM cells for this study because they have previously confirmed miR-200c expression outcomes in human and rat BMSCs and wanted to test human preosteoblasts. Based on their previously published findings, they believe the significantly amplified miR-200c expression in HEPM cells transfected via their CaCO₃/miR-200c nanocomplexes would also increase osteogenic marker expression in HEPM cells in a similar manner to their prior results using BMSCs.

This study evaluated the biocompatibility of CaCO₃/miR-200c nanocomplexes in relation to other nonviral transfection methods. The authors again utilized HEPM cells, as the results from this human-derived cell line would be translatable to clinical situations with human patients. From the MTT assay, the authors determined that CaCO₃ nanocomplex transfection did not affect HEPM cell viability in comparison with untreated cells (Figure 2A). HEPM cells transfected via CaCO₃/miR-200c nanocomplexes also did not significantly increase the expression of IL-6 or IL-8 in comparison with untreated cells (Figure 2B & C). PEI transfection, however, significantly increased IL-6 expression in comparison with untreated cells and CaCO₃/miR-200c-transfected cells. PEI transfection also increased IL-8 expression, but not significantly. These data indicate that these CaCO₃/miR-200c nanocomplexes are biocompatible, noncytotoxic and safe for human cell transfection. The results also affirm evidence published in the literature and further substantiate the cytotoxic and proinflammatory properties of PEI as a nonviral transfection mechanism.

Based on the authors' previous study demonstrating the slow-release potential of collagen coatings, they evaluated the influence of collagen on the release of their nanocomplexes [21]. The CaCO₃/miR-200c nanocomplexes were developed from published protocols, yet few studies have evaluated the release profile of CaCO₃/pDNA nanoparticles. For the authors' release study, they utilized a high-sensitivity dsDNA assay kit to measure pDNA release, which uses fluorescent probes to tag dsDNA for quantification. The authors found that naked pDNA encoding miR-200c incorporated into collagen sponges was rapidly released over 24 h but observed limited release of CaCO₃/miR-200c (Figure 3A). The small amount of pDNA released from the CaCO₃/miR-200c-incorporated collagen samples could be explained by an inability of the fluorescent probe of the dsDNA kit to effectively tag pDNA bound to CaCO₃ nanoparticles. Coprecipitation of pDNA with CaCO₃ nanoparticles creates a CaCO₃/pDNA polymer network in which the pDNA polymer is surrounded by CaCO₃ nanoparticles and the electrostatic interactions between the CaCO₃ and pDNA polymer chain create a condensed polymer network [60]. The binding of CaCO₃ nanoparticles to the pDNA polymer chain may have limited the ability of the fluorescent probe to effectively tag pDNA bound to CaCO₃ nanoparticles released from the collagen sponges. This may explain why few studies have published results for pDNA release from CaCO₃ nanoparticles, as current methods require the dissociation of pDNA from CaCO₃ for effective quantification. One previous study evaluated pDNA release from CaP-based nanoparticles by using two different pH-buffered solutions to dissolve nanoparticles prior to dsDNA quantification [41]. However, the authors of the present study could not effectively use this method to measure released dsDNA, possibly due to differences between the previous studies' CaP and the present study's CaCO₃ nanoparticles. In order to test whether there is an interaction between the collagen and CaCO₃/miR-200c nanocomplexes that extends the release of the nanocomplexes, the authors seeded HEPM cells onto naked pDNA and CaCO₃/miR-200c-incorporated collagen sponges that had been immersed into PBS for 0, 6 or 24 h. Using this approach, they quantified the cellular uptake of pDNA remaining in collagen after PBS exposure at different time points as an indicator of naked pDNA or CaCO₃/miR-200c release from the collagen sponges (Figure 3B). They further found a significant decrease in miR-200c expression for the naked miR-200c-collagen sponges after 24 h in PBS, indicating that pDNA was released from these samples and this depletion decreased the ability of the HEPM cells to amplify miR-200c overexpression to the same extent as the 0-h and 6-h samples. However, there were no significant differences in miR-200c overexpression for CaCO₃/miR-200c-collagen after 0, 6 and 24 h in PBS, suggesting that the release of nanocomplexes from collagen is sustained for 24 h. The dissimilarities in the release of naked pDNA and CaCO₃/pDNA from collagen sponges are likely due to differences in electrostatic interactions and pDNA encapsulation by the CaCO₃ nanoparticles. The quick release of pDNA from the collagen sponge is likely due to electrostatic repulsive forces between the negatively charged pDNA molecules and the negatively charged collagen protein. For the CaCO₃/pDNA samples, however, coprecipitation of pDNA with CaCO₃ nanoparticles created a nanocomplex system that encapsulated pDNA within the CaCO₃ nanoparticles and likely slowed and sustained pDNA release over time as the CaCO₃ particles dissolved [61,62].

The authors' prior published studies used naked pDNA encoding miR-200c delivered via bulbus injections, miR-200c-incorporated collagen sponges and, more recently, 3D-printed miR-200c-incorporated collagen-coated scaffolds to regenerate large and critical-sized bone defects. From these studies, they have demonstrated the osteogenic and anti-inflammatory properties of miR-200c and have found that miR-200c can effectively induce bone formation in rat calvarial defect models using these different miR-200c delivery approaches. However, like the authors' prior works, most published studies only utilize calvarial models to study bone formation. To regenerate defects caused by diseases such as oral cancer, there is a need to use in vivo models that more appropriately mimic the conditions of craniofacial reconstructive surgery, such as in an alveolar bone defect model. Few published studies have used tooth socket defect models to study alveolar bone regeneration and, to the authors' knowledge, none have used miRs to enhance alveolar bone formation, until now. The present study utilized an M1 tooth socket defect model to investigate the capacity of CaCO₃/miR-200c nanocomplexes to enhance alveolar bone formation. Based on the μCT imaging and quantification results (Figure 4), they found that the CaCO₃/miR-200c-incorporated collagen sponges significantly increased the percentage of new bone formation in the alveolar bone defects in comparison with untreated defects and defects treated with collagen sponges incorporating CaCO₃/EV. They further confirmed this increase in new bone formation in the CaCO₃/miR-200c-treated samples via hematoxylin and eosin and Masson's trichrome staining (Figure 5). In the histology sections, they observed more new bone formation in the alveolar bone defects treated with CaCO₃/miR-200c-incorporated collagen sponges, whereas the untreated defects and defects treated with CaCO₃/EV-incorporated collagen sponges displayed more fibrous tissue assembly than new bone formation. In prior studies, the authors assessed bone formation at longer time intervals, such as after 6 weeks. However, in this study, they were interested in determining if the CaCO₃/miR-200c nanocomplexes could induce bone regeneration at earlier stages and, thus, decided to assess regeneration after 3 weeks. The authors assume that with longer implantation times, they would find increased bone formation across all samples, with the defects treated with CaCO₃/miR-200c-incorporated collage sponges displaying significantly more bone formation; however, additional, longer-term studies would need to be completed to confirm this. The data from this study indicate that the CaCO₃/miR-200c nanocomplexes significantly enhance alveolar bone formation in vivo. These results further attest to the osteogenic and regenerative properties of miR-200c and build upon the conclusions of the authors' previous in vivo studies using miR-200c for calvarial bone regeneration. In congruence with the previous findings, data from this study support miR-200c as a potent and effective osteogenic agent that has significant application in treating different types of craniofacial bone defects.

The findings of this study demonstrate that CaCO₃/miR-200c nanocomplexes significantly enhance the transfection of miR-200c into HEPM cells in a biocompatible and noncytotoxic manner. These data establish the CaCO₃/pDNA nanocomplex system as an effective tool for nonviral gene delivery and a safer alternative to PEI transfection. Furthermore, CaCO₃ nanoparticles can effectively sustain the release of miR-200c from collagen sponges and extend enhanced expression of miR-200c in vitro. The results of the in vivo study also demonstrate that the CaCO₃/miR-200c nanocomplexes can effectively enhance bone formation in an alveolar bone defect model, further authenticating miR-200c as a potent osteogenic agent applicable to craniofacial bone restoration. This evidence strongly indicates that CaCO₃/miR-200c nanocomplexes significantly enhance transfection of pDNA encoding miR-200c into cells to promote osteogenesis and bone formation and effectively illustrates the therapeutic potential of miR-200c to restore bone defects caused by craniofacial reconstructive surgeries and diseases.

Conclusion

There is a critical need to develop advanced osteogenic therapeutics to effectively regenerate bone defects caused by craniofacial diseases such as oral cancer. miRNAs, specifically miR-200c, are biomolecules with potent osteogenic and anticancer properties that have significant application as novel regenerative agents to treat craniofacial bone defects. In this study, we successfully demonstrated that CaCO₃/miR-200c nanocomplexes significantly enhance plasmid transfection in a noncytotoxic, biocompatible manner in comparison with naked pDNA- and PEI-based transfection strategies. We further demonstrated, for the first time, that CaCO₃/miR-200c nanocomplexes significantly enhance in vivo bone formation in an alveolar bone defect model. These results strongly support CaCO₃-based nanoparticles as a safe and efficient nonviral system for cellular plasmid delivery and further advance the clinical potential of miR-based regenerative strategies to restore craniofacial bone defects.

Summary points.

• Craniofacial bone defects are surgically and therapeutically challenging to treat, and current regenerative methods are limited in their abilities to safely and effectively restore large bone defects.

• miRNAs play crucial roles in craniofacial bone development and regeneration, and miR-200c has been found to effectively enhance osteogenesis and in vivo bone regeneration to heal bone defects.

• CaCO₃-based nanoparticles have recently been utilized as nonviral vectors for plasmid DNA-based cell transfection; however, their efficiency in vivo and biocompatibility in comparison with other nonviral strategies have yet to be determined.

• Coprecipitation of CaCO₃ with protamine sulfate at an optimal ratio increases human embryonic palatal mesenchymal cell transfection.

• CaCO₃/miR-200c nanocomplexes created via the coprecipitation of CaCO₃/protamine sulfate with plasmid DNA encoding miR-200c enhances miR-200c transfection into human embryonic palatal mesenchymal cells.

• CaCO₃ nanocomplexes are biocompatible and noncytotoxic, unlike polyethylenimine transfection.

• CaCO₃/miR-200c nanocomplexes enhance in vivo bone formation in an alveolar bone defect model.

• These results support miR-200c as a safe, potent and restorative osteogenic agent that when used in combination with CaCO₃ nanoparticles enhances in vivo bone formation in large craniofacial defects.

• These findings attest to the robust osteogenic properties of miR-200c for enhancing bone regeneration and further illustrate the potential clinical application of miR-based therapeutics in treating large craniofacial bone defects.