Germanium-doped calcium phosphate ceramics in combination with mesenchymal stem cells derived from the neural crest, obtained from rabbit vibrissae: an experimental model of rabbit jaw regeneration

Article information

Arch Craniofac Surg. 2026;27(4):180-188

Publication date (electronic) : 2026 August 20

doi : https://doi.org/10.7181/acfs.2026.0026

Anna Zhmurko^,1, Roman Palyvoda¹, Natalia Ulyanchich², Valeryi Chemerovskyi³, Tetyana Todosiuk³, Alona Zlatska⁴

¹Department of Maxillofacial Surgery, Bogomolets National Medical University, Kyiv, Ukraine

²I.M. Frantsevich Institute for Problems of Materials Science, National Academy of Sciences of Ukraine, Kyiv, Ukraine

³Department of Surgery and Diseases of Small Domestic Animals, Bila Tserkva National Agrarian University, Bila Tserkva, Ukraine

⁴Institute of Genetic and Regenerative Medicine, M. D. Strazhesko National Scientific Center of Cardiology, Clinical and Regenerative Medicine, National Academy of Medical Sciences of Ukraine, Kyiv, Ukraine

Correspondence: Anna Zhmurko Department of Maxillofacial Surgery, Bogomolets National Medical University, 13 Taras Shevchenko Boulevard, Kyiv 01601, Ukraine E-mail: annetkazhmurko@gmail.com

Received 2026 January 28; Revised 2026 March 23; Accepted 2026 July 25.

Abstract

INTRODUCTION

Synthetic bone substitutes, although widely used and commercially available, often lack optimal biomechanical strength and the capacity to achieve complete osseointegration [1]. Modern synthetic materials predominantly have osteoconductive properties [2]. Along with the use of updated surgical techniques, the problem of choosing bone graft material remains clinically relevant. It is known that osteoplasty opens up great opportunities for the complete rehabilitation of patients [3]. In addition, other properties influence the success of bone graft integration, including biocompatibility, bioresorption, sterility, structural integrity, sufficient porosity for vascularization, ease of processing, and compressive strength [4]. Therefore, finding the optimal solution in each specific clinical situation is a choice based on recommendations, financial considerations, or the empirical judgment of the surgeon [5]. To improve the clarity of choice, expand the scope of practical application, and summate biological effects, scientific research has focused on improving bone substitutes using osteoinductive growth factors [6] and mesenchymal stem cells (MSCs) [7]. Improved scaffold designs, including optimal pore size, interconnectivity, density, and resorption rates, are crucial for improving osseointegration and vascularization [8]. These innovations aim to mimic the properties of natural bone while ensuring cost-effectiveness, low immunogenicity, and clinical practicality [9]. Combinations of these innovative techniques enhance each other’s effects and may enable the development of novel bone substitute materials [10]. To our knowledge, no previous studies have evaluated germanium-doped biphasic calcium phosphate ceramics combined with neural crest-derived MSCs from vibrissae in a critical-size mandibular defect model, highlighting the novelty of the current work. Therefore, based on this hypothesis, we conducted this study.

MSCs are viable osteogenic progenitor cells derived from bone marrow that are capable of differentiating into osteoblasts and effectively regenerating large bone defects [11]. Their use in combination with bone replacement materials improves healing and the mechanical properties of newly formed bone, which is especially important when installing dental implants [12].

The aim of the study was to evaluate the effect of bioactive ceramic materials on the viability of neural crest-derived multipotent mesenchymal stem cells (NC-MSCs) derived from rabbit vibrissae, and to conduct a comparative analysis of the regenerative potential of rabbit mandibular bone during the repair of bone defects using biphasic calcium phosphate ceramics doped with 1% germanium (BCPdGe 1%) alone and BCPdGe 1% combined with NC-MSCs.

METHODS

To study the biocompatibility, adhesive properties, and potential cytotoxic effects on synthetic, ceramic, and osteoplastic material granules, combined fluorescein diacetate and propidium iodide staining was performed to detect adhesion to the granules, as well as to identify viable and dead cells 24 hours, 3 days, and 5 days after seeding.

Nine samples of bioactive ceramic materials were studied (sample no. 1, pure hydroxyapatite [HA]; sample no. 2, pure hydroxyapatite doped with 1% germanium [HAdGe 1%]; sample no. 3, pure hydroxyapatite doped with 1.5% germanium [HAdGe 1.5%]; sample no. 4, biphasic calcium phosphate ceramic [BCP]; sample no. 5, biphasic calcium phosphate ceramic, doped with 1% germanium [BCPdGe 1%]; sample no. 6, biphasic calcium phosphate ceramic, doped with 1.5% germanium [BCPdGe 1.5%]; sample no. 7, biphasic calcium phosphate ceramics doped with SiO₂ [BCPdSiO₂]; sample no. 8, biphasic calcium phosphate ceramics doped with SiO₂ [BCPdSiO₂ rinse]; sample no. 9, biphasic calcium phosphate ceramic doped with SiO₂ [BCPdSiO₂ elution]) (Table 1).

Table 1.

Assessment of the cytotoxic effect of bioactive ceramics on NC-MSCs from rabbit vibrissae (% of viable cells)

The BCPdGe 1% group without stem cells was previously evaluated in our earlier experiment using the same rabbit mandibular defect model. In accordance with ethical principles of animal research and the 3R concept (Reduction), these data were used as the reference control for comparison with the cellloaded construct in the present study.

For the in vivo experiment, 30 certified Californian breed laboratory rabbits (weight: 1,995±150.4 g), equally divided between males and females and free of visible pathological signs, were used. The animals were housed individually in a controlled vivarium in individual cages under combined lighting with daily cage cleaning. They were provided with a balanced pelleted diet (200 g/day) and had ad libitum access to water.

The sample size of rabbits was calculated based on the statistical power of the experiment and allowed for reliable results.

All procedures for obtaining NC-MSCs from rabbit vibrissae, isolation, and cell cultivation were performed in accordance with national legislation and relevant European guidelines. Primary cell lines were obtained in a certified biotechnology laboratory operating under GMP/GLP (Good Manufacturing Practice/Good Laboratory Practice) standards.

All studies were approved by the institutional Ethics Committee for animal research and were carried out in accordance with national legislation and relevant European guidelines for the protection of vertebrate animals used in scientific research.

The animals were divided into a control group (group 1) and a main group (group 2), with 15 rabbits in each group. The experimental study consisted of creating a bicortical critical-size mandibular defect (7 mm in diameter) in the mandibular angle using a trephine, followed by implantation of BCPdGe 1% granules in group 1 and BCPdGe 1% combined with NC-MSCs obtained from rabbit vibrissae in group 2. In the postoperative period, rabbits were injected intramuscularly with Bicillin-3 to prevent postoperative infection. The animals underwent X-ray examination of the skull in lateral projection on days 7, 21, 42, and 90. All animals were euthanized on days 21, 42, and 90 by an overdose of intravenous thiopental anesthesia (5 animals per observation point). After that, we took material from the replaced defect site for histological examination of each animal (Fig. 1).

Fig. 1.

Design of the experimental study. Group 1 represents BCPdGe 1% granules, and Group 2 represents BCPdGe 1%+NC-MSCs granules. BCPdGe 1%, biphasic calcium phosphate ceramics doped with 1% germanium; NC-MSCs, neural crest-derived multipotent mesenchymal stem cells.

Statistical data processing was performed using EZR software. The normality of data distribution was assessed using the Shapiro– Wilk test. Descriptive statistics are presented as mean± standard deviation (SD).

The study design included independent groups, with five animals per group at each time point (21, 42, and 90 days). Comparisons between groups were conducted using independentsamples Student t-tests at each time point. To account for multiple comparisons, Bonferroni correction was applied, and the adjusted level of statistical significance was set at p<0.0167.

RESULTS

BCPdGe 1% on the first day of cultivation (24 hours after seeding) showed a significant number of red-stained cells (Fig. 2A, bottom row), indicating acute cytotoxicity of the material for cells, while dense, isolated clusters of fibroblast-like cells actively dividing and migrating to form strands were observed (Fig. 2B, top row). Analysis 5 days after seeding revealed only a few dead cells on the surface of the granules (Fig. 2C, lower row), while numerous viable fibroblast-like cells continued to proliferate and migrate. The granule surface was uniformly covered by cells, reaching approximately 60%–70% confluence (Fig. 2C, upper row). The highest viability after seeding was observed in NC-MSCs from rabbit vibrissae when co-cultured with samples no. 3 (HAdGe 1.5%), no. 5 (BCPdGe 1%), no. 7 (BCPdSiO₂).

Fig. 2.

Adhesion and viability of NC-MSCs from rabbit vibrissae on BCPdGe 1% granules (sample no. 5) after 24 hours (A), 3 days (B), and 5 days (C) of cultivation. Live cells are shown in the top row; dead cells in the bottom row. Cells were stained using combined FDA/PI staining and visualized by fluorescence microscopy (×50). Green indicates live cells; red indicates dead cells. NC-MSCs, neural crest-derived multipotent mesenchymal stem cells; BCPdGe 1%, biphasic calcium phosphate ceramics doped with 1% germanium; FDA, fluorescein diacetate; PI, propidium iodide.

Histologically, in group 2 on day 21, the central area of the defect was filled with dense connective tissue surrounding the granules, with layers of loose connective tissue observed in some regions. Around the granules, dense clusters of cells with rounded nuclei and elements of the hemomicrocirculatory system were present. Near the native bone, coarse-fibrous bone tissue beams were formed, sometimes containing single lamellae of lamellar bone, which developed into osteons, semi-osteons, and bone beams. Individual granules were in direct contact with the trabeculae (Fig. 3).

Fig. 3.

Defect area filled with BCPdGe 1% combined with NC-MSCs, on day 21 (hematoxylin and eosin, ×100). (A, B) Roman numerals indicate the following regions: I, granule; II, dense connective tissue; III, loose connective tissue. BCPdGe 1%, biphasic calcium phosphate ceramics doped with 1% germanium; NC-MSCs, neural crest-derived multipotent mesenchymal stem cells.

On day 42, the regenerate adjacent to the native bone contained lamellar bone tissue with remnants of coarse-fibrous bone. Some granules were in contact with the bone, while others were not and were surrounded by cell nuclei. In regions closer to the center of the defect, islands of bone tissue were observed, partially associated with the granules, while the surrounding areas consisted of fibrous connective tissue. Capillaries were present between the granules. A significant area of the defect was occupied by granules surrounded by dense fibrous tissue containing numerous cell nuclei and in a state of destruction (Fig. 4).

Fig. 4.

Defect area filled with BCPdGe 1% combined with NC-MSCs on day 42 (hematoxylin and eosin, ×100). The Roman numerals indicate the following structures in each panel: (A) I, granules; II, coarse-grained bone tissue; III, lamellar bone tissue. (B) I, granules; II, dense connective tissue. (C) I, granules; II, dense connective tissue infiltrated with cell nuclei. (D) I, granules; II, dense connective tissue; III, lamellar bone tissue. BCPdGe 1%, biphasic calcium phosphate ceramics doped with 1% germanium; NC-MSCs, neural crest-derived multipotent mesenchymal stem cells.

On day 90, the regenerate adjacent to the native bone consisted of lamellar bone surrounding granules in a state of degradation, with remnants of coarse-fibrous tissue in the beams. Towards the center of the defect, the granules were densely surrounded by connective tissue. In some regions, the bone tissue was in contact with several layers of rounded nuclei (probably stem cells) that formed the border of the central zone. Bundles of collagen fibers extended toward the defect center, and in certain areas, altered connective tissue with disoriented cells and fibers was observed, along with acellular zones in some regions (Fig. 5).

Fig. 5.

Defect area filled with BCPdGe 1% combined with NC-MSCs on day 90 (hematoxylin and eosin, ×100). The Roman numerals indicate the following structures in each panel: (A) I, native bone; II, layers of cell nuclei; III, granule; IV, bundles of first-order collagen fibers. (B) I, granule; II, lamellar bone; III, remnants of coarse-fibrous bone tissue. BCPdGe 1%, biphasic calcium phosphate ceramics doped with 1% germanium; NC-MSCs, neural crest-derived multipotent mesenchymal stem cells.

During the experiment, it was observed that on day 21, the difference in the width of newly formed trabeculae between groups (group 1: 71.22±14.57 μm; group 2: 48.60±14.09 μm) was not statistically significant after Bonferroni correction (p=0.037) (Fig. 6A). On day 42, this parameter was 84.7% greater in group 1 than in group 2 (126.86±27.64 μm vs. 68.69±11.80 μm; p=0.003), and the difference remained significant after correction, with a large effect size (η²=0.701) (Fig. 6B). On day 90, it remained 56.7% higher in group 1 (109.24±11.90 μm vs. 69.73±24.52 μm; p=0.012), also maintaining statistical significance after correction, with a large effect size (η²=0.568) (Table 2, Fig. 6C). Overall, intergroup differences increased over time, peaking on day 42 and persisting on day 90.

Fig. 6.

Comparison of trabecular width (μm) between the BCPdGe 1% and BCPdGe 1%+NC-MSCs groups at day 21 (A), day 42 (B), and day 90 (C). Data are presented as mean±standard deviation. BCPdGe 1%, biphasic calcium phosphate ceramics doped with 1% germanium; NCMSCs, neural crest-derived multipotent mesenchymal stem cells.

Table 2.

Width of newly formed trabeculae in the area of implantation of BCPdGe 1% granules and BCPdGe 1%+NC-MSCs granules into jaw bone defects

DISCUSSION

MSCs are multipotent progenitor cell populations capable of differentiating into osteoblast precursors in vitro under certain conditions and are therefore widely used for bone tissue regeneration [13]. When combined with a scaffold carrier, MSCs can differentiate into osteogenic cells and effectively contribute to the repair of large bone defects [14]. However, implanted cells are highly sensitive to local biological and mechanical conditions throughout the entire process, from cultivation to transplantation. Critical factors influencing cell survival and function include the duration of in vitro culture [15], mechanical stress during implantation, local hypoxia, and insufficient nutrient availability, all of which may adversely affect cell viability, proliferation, and differentiation [16]. For this reason, a germanium-containing biphasic calcium phosphate material was selected as the scaffold in the present study, given its reported bioactivity and osteoconductive potential. Among adult stem cells, MSCs are considered preferred candidates for transplantation as progenitor cells for bone tissue regeneration due to their multipotency, immunomodulatory properties, and ability to secrete trophic factors that support tissue repair [17]. In addition, MSCs can modulate the local immune response, thereby contributing to the formation of a regenerative microenvironment favorable for bone healing [18]. Nevertheless, the expected regenerative advantage of MSC-based constructs is not always achieved in vivo, which may be attributed to multiple limiting factors related to both the cells and the scaffold.

The enzymatic method of isolating MSCs from tissue is attractive because it allows variation of digestion time, which directly influences cell yield and viability at the isolation stage. However, this method inevitably produces a relatively heterogeneous cell population and does not allow precise separation of specific progenitor subtypes [19]. In addition to enzymatic isolation, alternative approaches such as mechanical methods (centrifugation, filtration, tissue mincing) are widely used. These techniques are rapid, inexpensive, and compliant with GMP requirements, but typically yield fewer cells and show greater variability in outcomes [20]. Combined mechanical– enzymatic protocols may improve cell yield and viability by reducing digestion time, although they require more complex procedures. Gravity settling or filtration methods are simple and cost-effective but result in a limited number of harvested MSCs [21]. Explantation techniques minimize enzymatic exposure but are time-consuming and associated with an increased risk of cellular aging, while automated closed systems (e.g., Celution, Lipogems) provide high cell yields and enhanced process safety at the expense of high cost and specialized equipment requirements [22].

The primary cell population obtained after enzymatic processing of rabbit vibrissae biomaterial included thin spindleshaped cells, typical fibroblast-like cells, large cloak-shaped cells with multiple processes, as well as a substantial number of small round cells (Fig. 7). Vibrissae represent a well-established source of NC-MSCs in animals [23], characterized by unique morphofunctional properties and the capacity for neurogenic differentiation. The enzymatic isolation approach allows the acquisition of a relatively high number of adherent cells exhibiting stellate morphology with shortened processes and pronounced proliferative activity, as evidenced by frequent mitotic figures. Despite these advantages, the resulting NC-MSC population remains heterogeneous. Nevertheless, enzymatic isolation followed by selection based on plastic adherence and colonyforming ability enables the establishment of a more homogeneous population with fibroblast-like morphology characteristic of MSCs. Typical MSC growth patterns, including the formation of circular vortex-like structures, were observed as early as passage 3 (Fig. 8). Importantly, heterogeneity within the implanted cell population may represent a critical factor limiting regenerative outcomes, as not all cells may possess equivalent osteogenic potential or adaptability to adverse in vivo conditions.

Fig. 7.

Morphology of NC-MSCs in vitro. Phase-contrast microscopy, passage 0. NC-MSCs, neural crest-derived multipotent mesenchymal stem cells.

Fig. 8.

Morphology of NC-MSCs in vitro. Phase-contrast microscopy, passage 3. NC-MSCs, neural crest-derived multipotent mesenchymal stem cells.

In addition to cellular factors, scaffold-related limitations play a crucial role. The adhesive properties of BCPdGe 1% granules directly influence cell attachment and retention at the implantation site. These properties are markedly inferior to those of standard tissue culture plastic, which is specifically optimized for substrate-dependent cell adhesion. Under classical seeding conditions, the proportion of cells failing to adhere to culture plastic does not exceed 5%, whereas for osteoplastic materials this value is substantially higher, ranging from 30% to 50%, depending on surface characteristics, granule size, and material composition [24]. Insufficient cell adhesion to ceramic granules, combined with mechanical stress during implantation and early postoperative hypoxia, may result in significant cell loss, thereby reducing the potential regenerative contribution of transplanted NC-MSCs.

In conclusion, it was found that the most viable cells after seeding were NC-MSCs from rabbit vibrissae when co-cultured with samples no. 3 (HAdGe 1.5%), no. 5 (BCPdGe 1%), no. 7 (BCPdSiO₂). Therefore, these samples are the most promising candidates for creating tissue-engineered structures based on bioactive ceramic granules. However, in the rabbit experiment, the synthetic material BCPdGe 1% alone demonstrated significantly better performance as a bone substitute than BCPdGe 1% combined with NC-MSCs. In vivo, BCPdGe 1% granules co-cultured with NC-MSCs from rabbit vibrissae did not effectively replace bone defects in the rabbit jaw and failed to produce the anticipated improvement. These findings highlight that implanted cells are highly sensitive to local biological conditions and mechanical stress, underscoring the need for further research on stem cell sources and cultivation methods.

Article information Continued