Experimental study on the repair of rabbit femoral condyle critical-sized bone defect using polyphosphoester-amino acid copolymer-composite nanohydroxyapatite filling material

Preparation and characterization of nHA/PPE-AA

The nanohydroxyapatite/polyphosphoester-amino acid (nHA/PPE-AA) bone-filling material was developed by collaborating institutions. The preparation procedure was as follows.

4-Aminocaproic acid, 6-aminobutyric acid, and caprolactam were weighed and dissolved in deionized water at a mass ratio of 2:1:1, and the solution was transferred into a reaction kettle (Weihai, Shandong, China). A 50% (w/w) phosphoric acid solution was added as the catalyst, and the mixture was stirred under nitrogen protection at 150–200°C to remove water. The temperature was then increased to 210°C, and melt copolymerization was conducted for 2.5 h. Subsequently, 2.5%–10% cyclic phosphate ester (molar fraction) was added, and the reaction was continued for 0.5 h. After cooling, crude PPE-AA was obtained.

The PPE-AA was then crushed and transferred into an automatically controlled reaction kettle. Subsequently, nHA was added to achieve a mass fraction of 30 ± 5% in the system, followed by the addition of an appropriate amount of purified water for mixing. The mixture was stirred at 210 ± 5°C for 2.5 h. Cyclic phosphate ester (molar fraction, 2.5%) was subsequently added, and the reaction was continued for 1 h. After cooling, a semi-finished nHA/PPE-AA product was obtained. The semi-finished product was extrusion-molded using a twin-screw extruder (SHJ-20; Nanjing, China) at a gradient extrusion temperature decreasing from 160°C to 110°C and a screw speed of 30–50 Hz. The extrudate was pelletized and sieved to obtain irregular granules with particle sizes ranging from 300 to 5000 μm.

The positive control material was a medical-grade calcium sulfate artificial bone graft (Osteoset® pellets, without DBM; Wright Medical, USA), which was processed into 2–4 mm granules before implantation. Calcium sulfate was selected as the positive control because it is a clinically used, commercially available, resorbable bone void filler with established biocompatibility and osteoconductive properties. All implant materials were sterilized with ethylene oxide, fully aerated, and stored in sealed packages until use.

For material characterization, morphology was observed by scanning electron microscopy (SEM; S-3400, HITACHI, Japan). Phase composition was analyzed by X-ray diffraction (XRD; D8 ADVANCE, Bruker, Germany; Cu Kα radiation; 2θ = 10°–80°). Functional groups were examined by Fourier-transform infrared spectroscopy (FTIR; Nicolet iS10, Thermo Fisher Scientific, USA; 4000–400 cm⁻¹). Porosity was determined by the Archimedes drainage method. Pore size distribution was measured statistically from SEM images. Compressive strength was determined by compression testing and reported as a range.

Animal model and surgery procedures

Thirty-six healthy adult New Zealand White rabbits of either sex, weighing 2.5–3.0 kg, were included in this study. All procedures involving animals were approved by the Institutional Animal Ethics Committee ([Information blinded for peer review]) and conducted in accordance with institutional and national guidelines for the care and use of laboratory animals. All efforts were made to minimize animal suffering and the number of animals used.

All animals underwent routine physical examination upon arrival. Normal food intake, mental status, activity, urination, and defecation were confirmed. The coat was smooth, with no evidence of trauma or infection. Routine blood tests and serum biochemistry showed normal liver and kidney function. After 2 weeks of acclimatization without abnormalities, the animals were enrolled in the experiment.

The rabbits were randomly assigned to three groups (n = 12 per group). A critical-sized bone defect measuring 10 × 10 × 10 mm³ was created in the femoral condyle of each rabbit. In the experimental group, the defect was filled with nHA/PPE-AA granules; in the positive control group, the defect was filled with calcium sulfate granules; and in the blank control group, no material was implanted after defect creation. Animals were euthanized at 1, 2, 3, and 6 months postoperatively for sample collection (n = 3 per group at each time point).

Food was withheld for 8 h before surgery. General anesthesia was induced by intravenous injection of 3% sodium pentobarbital (0.03 g/kg) through the marginal ear vein. A longitudinal incision approximately 3 cm in length was made on the medial side of the knee, and the medial femoral condyle was exposed layer by layer. Under saline irrigation, the defect was created using an orthopedic drill and curette until only the anterior and posterior cortices, the articular cartilage surface, and the subchondral bone were preserved. After irrigation, nHA/PPE-AA granules were tightly packed into the defect in the experimental group, and calcium sulfate granules were implanted in the positive control group. No implant was placed in the blank control group. The wound was closed in layers, and the animals were allowed unrestricted activity after surgery. Penicillin was administered intramuscularly at 40,000 U/day for 3 consecutive days postoperatively to prevent infection.

Macroscopic observation and serum markers of bone metabolism

At 1, 2, 3, and 6 months after surgery, three rabbits from each group were randomly selected for cardiac blood collection to measure bone-specific alkaline phosphatase (B-ALP) and N-terminal midfragment osteocalcin (N-MID). After euthanasia, femoral condyle specimens were harvested, and the degradation of implanted materials and filling of the defects by newly formed tissue were recorded and photographed.

DR imaging evaluation and G/N radiopacity ratio

Anteroposterior and lateral radiographs of the operated femur were obtained using digital radiography (DR). Radiopacity values were measured in the grafting area and the adjacent non-grafting normal bone area using Siemens imaging software. The grayscale ratio (G/N) was then calculated, with the normal bone area defined as G/N = 1.0. Because nHA is inherently radiopaque, the G/N ratio was interpreted as a composite radiographic parameter reflecting both residual implanted material and newly mineralized bone. Therefore, longitudinal changes in G/N over time, rather than the absolute value at a single time point, were used as supportive evidence of progressive mineralization. These radiographic findings were further interpreted together with micro-CT, fluorochrome labeling, and histological results.

Micro-CT scanning

After X-ray examination, the femoral condyles were harvested and scanned using a micro-computed tomography system (Quantum GX, PerkinElmer, USA) under the following conditions: 57 min per specimen, slice interval 50 μm, scanning range 20 × 25 × 25 mm, 90 kV, and 88 μA. The scans were used to evaluate bone density, new bone formation, and the distribution of osteoid/high-density signals within the defect region. Residual material and newly formed bone were differentiated mainly according to the morphology and distribution of high-density signals. Residual material was identified as discrete irregular granular signals, whereas newly formed bone was identified by interconnected trabecular architecture continuous with the adjacent host bone.

Sequential fluorochrome labeling and dynamic histomorphometry

Fourteen days and 3 days before euthanasia, calcein (15 mg/kg) and alizarin red (30 mg/kg), respectively, were injected into the gluteus maximus muscle. After specimen collection, fluorescence imaging was performed using an upright two-photon laser confocal fluorescence microscope. Calcein was excited at approximately 470 nm and detected at approximately 516 nm, showing green fluorescence, whereas alizarin red was excited at approximately 543 nm and detected at approximately 600 nm, showing red fluorescence. Images were acquired at a thickness of 5–10 μm, with 50–100 images collected per specimen.

The images were imported into NIS-Elements Analysis 4.50.00.64 software for reconstruction and quantitative analysis. The following parameters were calculated:

(1) Interlabeled width (IL.WI), defined as the mean distance from the inner edge of the inner label to the inner edge of the outer label within the same trabecular structure, representing the thickness of mineralized bone formed during the interval between the two administrations.

Histological evaluation

Specimens were fixed in 10% neutral formalin for 48 h and decalcified in 10% EDTA for 4–8 weeks, with the decalcifying solution changed weekly. After gradient ethanol dehydration and paraffin embedding, serial sections of 5 μm thickness were prepared. The sections were stained as follows:

(1) Hematoxylin and eosin (H&E) staining was used to evaluate inflammatory response, morphology of newly formed bone, and material degradation.

SEM observation

After fixation in 10% formaldehyde solution, the specimens were dehydrated in graded ethanol, air-dried, and sputter-coated with gold by ion sputtering. Bone-material interfacial integration was examined by SEM (HITACHI S-3400N, HITACHI, Japan).

Quantitative RT-PCR analysis of BMP-2 mRNA expression

At 1, 2, 3, and 6 months after surgery, newly formed tissue from the defect region was collected. Total RNA was extracted using a BIOFIT kit, and cDNA was synthesized using an Aidlab kit. RT-qPCR amplification was performed using a SYBR Green I system. Relative BMP-2 mRNA expression in each group was normalized to β-actin as the internal control, and relative expression levels were calculated using the Pfaffl method. Primer sequences and annealing temperatures are listed in Table 1.

OPEN IN VIEWER

Table 1.

Primer sequences and annealing temperatures used in RT-qPCR.

Gene	Primer	Sequence (5′ to 3′)	Annealing temperature (°C)
β-actin	Forward (F)	GTGGCGCTGGACTTCGA	58
β-actin	Reverse (R)	CGTTGCCGATGGTGATGA	58
BMP-2	Forward (F)	GCC​TTT​GCT​CGT​AAC​TTT​TGG	58
BMP-2	Reverse (R)	TGT​GTT​TCG​CTT​GAC​GCT​TT	58

Material safety evaluation

At 6 months after surgery, the liver, lung, kidney, brain, heart, and implant-related tissues (femoral condyle implantation site and intramuscular implantation site) were harvested for H&E staining. In addition, Masson’s trichrome staining was performed for the femoral condyle implantation site. Blood samples were collected at 1 day, 3 days, 7 days, 2 weeks, 1 month, 3 months, and 6 months postoperatively for measurement of erythrocyte sedimentation rate (ESR), white blood cell count (WBC), and C-reactive protein (CRP).

Statistical analysis

Quantitative data are presented as mean ± standard deviation. Statistical analyses were performed using SPSS 19.0 software (SPSS Inc., Chicago, IL, USA). At each time point, one-way analysis of variance (one-way ANOVA) was used to compare the experimental, positive control, and blank groups, followed by pairwise comparisons. A value of p < 0.05 was considered statistically significant.

Characterization of nHA/PPE-AA porous granules

As shown in Figure 1, PPE-AA copolymer was successfully synthesized and combined with nHA to form a composite system. The nHA/PPE-AA material appeared as gray-white irregular granules with particle sizes of approximately 300–5000 μm (Figure 1(b)). SEM revealed a distinct porous architecture, with rough pore walls and interconnected pores. The pore size was mainly distributed within the range of 20–500 μm, and the initial porosity of the material was 35%–55% (Figure 1(c)).OPEN IN VIEWER

XRD analysis showed a broad diffuse peak for PPE-AA at approximately 2θ = 20.32°, indicating that the polymer was predominantly amorphous. nHA exhibited characteristic crystalline diffraction peaks, particularly a prominent peak cluster around 2θ = 31.84°. In the nHA/PPE-AA composite, the characteristic diffraction peaks of nHA, including the peak cluster at 31.84°, were retained together with the amorphous background of PPE-AA, indicating that nHA was successfully incorporated into the PPE-AA matrix without disruption of its major crystalline phase. No additional strong diffraction peaks were detected, suggesting that no detectable new crystalline impurities formed during composite fabrication (Figure 1(d)).

FTIR further confirmed the coexistence of the two components in the composite system. nHA showed a characteristic -OH absorption peak near 3573.3 cm⁻¹ and multiple characteristic phosphate vibration peaks at approximately 1095.2, 1041, 993.7, and 962.3 cm⁻¹ (Figure 1(e)). PPE-AA exhibited an aliphatic -CH₂- stretching vibration peak at 2922.7 cm⁻¹, together with clear absorption peaks in the carbonyl- and amide-related regions at 1730 cm^-1 (C = O) and 1649.6 and 1541.1 cm⁻¹ (amide-related peaks). In the nHA/PPE-AA composite, characteristic absorption peaks from both the inorganic and organic phases were present simultaneously, indicating the stable coexistence of nHA and PPE-AA within the composite material (Figure 1(e)). Compression testing showed that the compressive strength of the material ranged from 3 to 7 MPa.

Postoperative general condition

All animals in the three groups resumed feeding and drinking after recovery from anesthesia. As the surgical incisions healed, daily activity gradually returned to normal. The animals maintained good appetite and general condition, and body weight increased progressively. All incisions healed by primary intention, with no necrosis, infection, suppuration, material rejection, or material exposure observed. No postoperative complications, including swelling, joint effusion, deformity, dislocation, or fracture, were detected in the operated knees (Figure 2).OPEN IN VIEWER

Gross repair status and changes in serum bone metabolism

Gross specimens are shown in Figure 3(a)–(f). In the experimental group, fibrous callus formation covering the implant surface was observed at 1 month postoperatively, and the bone callus progressively covered the defect surface over time (Figure 3(a)). At 6 months, the implanted region was grossly similar to normal bone tissue, the defect surface was covered by cortical bone, and the contour of the femoral condyle had been restored (Figure 3(b)).OPEN IN VIEWER

In the positive control group, the defect was enclosed by fibrous callus at 1 month, and the implanted material was prone to detachment on sectioning (Figure 3(c)). At 6 months, the defect had also undergone repair, and cortical continuity was largely re-established; however, the surface morphology remained slightly irregular (Figure 3(d)). In the blank group, fibrous tissue ingrowth was evident within the defect at 1 month (Figure 3(e)), and incomplete repair was still observed at 6 months, with the defect area being occupied predominantly by fibrous scar tissue (Figure 3(f)).

Serum bone metabolic markers were measured at each postoperative time point. In all three groups, an active osteogenic phase was observed after implantation. B-ALP increased during the early postoperative period, peaked at 2 months, and then gradually declined. At 1 month, the B-ALP level in the experimental group was significantly higher than those in the positive control and blank groups (Figure 3(g)). N-MID peaked at 3 months, and the level in the experimental group was significantly higher than those in the other two groups (Figure 3(h)), suggesting that nHA/PPE-AA promoted bone matrix formation and maturation.

X-ray observation and detection of G/N ratio

DR images showed a time-dependent increase in radiographic density within the defect area in all groups. In the experimental group, the defect area had already shown an increase in radiographic density at 1 month postoperatively (Figure 4(a)). At 6 months, the density had increased further and approached that of normal bone (Figure 4(b)). The positive control group also exhibited a gradual increase in density over time, but the overall density at 6 months remained lower than that in the experimental group (Figure 4(c) and (d)). In the blank group, the defect area remained markedly radiolucent at 1 month (Figure 4(e)). By 6 months, only mild density increase was observed at the defect margins, whereas repair of the central defect area remained limited (Figure 4(f)).OPEN IN VIEWER

To quantify radiographic changes within the defect area, the grayscale ratio between the defect area and adjacent normal bone (G/N) was calculated. Because both nHA/PPE-AA granules and calcium sulfate are radiopaque, the G/N ratio was interpreted as a composite radiographic parameter rather than a direct stand-alone measure of newly formed bone. The G/N ratio increased progressively over time in both the experimental and positive control groups, suggesting increased mineralized tissue within the defect region when interpreted together with micro-CT, fluorochrome labeling, and histological findings. In the intergroup comparison, the G/N ratio in the positive control group was significantly lower than that in the experimental group at 2 and 3 months postoperatively (p = 0.037 and p = 0.041, respectively). In addition, the G/N ratio in the experimental group was significantly higher than that in the blank group at all time points (all p < 0.001).

Micro-CT evaluation of bone regeneration

Micro-CT demonstrated marked differences in new bone formation and defect repair among the three groups. In the experimental group, a high-density signal was visible in the defect area at 1 month, with a clear bone-material interface, although the defect area was still dominated by low-density signals and trabecular structures were not yet evident (Figure 5(a)). At this stage, the high-density signal was mainly interpreted as residual nHA/PPE-AA material because it appeared as discrete irregular granular signals. At 3 months, trabeculae had increased in number and become thicker, extending into the defect in a woven network-like pattern (Figure 5(b)). The interconnected trabecular architecture continuous with adjacent host bone supported progressive new-bone formation rather than residual material signal alone. At 6 months, cortical continuity was largely restored, and the trabecular structure became denser and more clearly defined, with progressive integration of bone and material (Figure 5(c)).OPEN IN VIEWER

In the positive control group, high-density signals corresponding to residual calcium sulfate granules were observed in the defect area at 1 month, and trabecular structures were not obvious (Figure 5(d)). At 3 months, trabeculae extended toward the center of the defect, while the high-density signal had markedly decreased, consistent with calcium sulfate resorption and progressive tissue/bone replacement within the defect region (Figure 5(e)). At 6 months, cortical continuity was restored, and the trabecular structure became relatively clear (Figure 5(f)).

In the blank group, only marginal changes were observed at 1 month (Figure 5(g)). At 3 months, sclerosis gradually developed at the defect margins, whereas the defect area remained predominantly occupied by soft tissue (Figure 5(h)). At 6 months, the defect remained incompletely healed, cortical continuity was not restored, and low-density soft tissue-like regions were still visible within the defect (Figure 5(i)).

Fluorochrome labeling and dynamic mineralization

Sequential fluorochrome labeling showed distinct double labeling with red and green fluorescent bands in both the experimental and positive control groups. These bands gradually extended from the defect margins toward the center and became increasingly continuous over time. In contrast, fluorescence signals in the blank group were weak and failed to form continuous bands (Figure 6).OPEN IN VIEWER

In the experimental group, abundant red and green fluorescent bands were observed at 1 month along the defect margins and the bone-material interface, with signal intensity gradually decreasing toward the center of the defect (Figure 6(a)). At 2 months, the number of fluorescent bands increased, and the bands became thicker and broader, extending further toward the center of the defect (Figure 6(b)). At 3 months, the bands exhibited a woven distribution, and fluorescence signals were also visible in the center of the defect. At higher magnification, trabecular structures with osteoblast attachment could be observed (Figure 6(c)). At 6 months, fluorescent labeling was distributed more uniformly throughout the defect area, and although the bands were relatively fewer, the structure was clearer (Figure 6(d)).

In the positive control group, fluorescent bands also gradually increased from 1 to 3 months and extended toward the defect center; however, their overall continuity and uniformity were inferior to those in the experimental group, with varying degrees of interruption still present (Figure 6(e)–(g)). At 6 months, extensive fluorescent bands and concentric ring-like structures were observed within the defect area (Figure 6(h)). In the blank group, only a small number of scattered fluorescence signals were observed at 1 month (Figure 6(i)). At 2, 3, and 6 months, the field was dominated by a dark background with scattered punctate red and green fluorescence, without obvious continuous bands (Figure 6(j)–(l)).

Dynamic histomorphometric analysis showed that IL.WI, MAR, and MOSW were all at relatively high levels in both the experimental and positive control groups during the first 1–2 months after surgery, and these values were significantly higher in the experimental group than in the positive control group, indicating more active bone formation and mineral apposition during the early and middle stages of repair in the experimental group (Figure 6(m)–(o)). MLT was shorter in the experimental group than in the positive control group, and the differences reached statistical significance at 2 and 3 months postoperatively (Figure 6(p)). After 3 months, these parameters generally declined or stabilized. By 6 months, no obvious differences in IL.WI, MAR, MOSW, or MLT were observed between the two groups, suggesting that osteogenic activity had gradually entered a plateau phase during late-stage repair.

Histological results

Experimental group: At 1 month postoperatively, osteoid tissue and/or trabecula-like structures were already observed interwoven with the surrounding tissue in the defect area (Figure 7(a)). Masson’s trichrome staining showed predominantly blue-stained collagenous matrix, with focal areas of red-stained bone matrix (Figure 8(a)). At 3 months, trabeculae were increased in number and became more continuous (Figure 7(b)). Correspondingly, Masson’s staining showed a marked increase in red-stained bone matrix with greater continuity, whereas the proportion of blue-stained matrix decreased (Figure 8(b)). At 6 months, the bone tissue became denser and showed a remodeling trend (Figure 7(c)), and Masson’s staining was dominated by extensive red-stained bone matrix (Figure 8(c)).OPEN IN VIEWER

OPEN IN VIEWER

Figure 8.

Masson’s trichrome staining of the femoral condyle defect area. (a–c) Staining results of the experimental group at 1, 3, and 6 months postoperatively. (d–f) Staining results of the positive control group at 1, 3, and 6 months postoperatively. (g–i) Staining results of the blank control group at 1, 3, and 6 months postoperatively. Scale bar = 100 μm.

Positive control group: At 1 month, scattered irregular osteoid tissue and/or fine trabeculae were observed in the defect area (Figure 7(d)), and Masson’s staining showed a relatively high proportion of blue-stained matrix with only focal red-stained areas (Figure 8(d)). At 3 months, trabeculae increased, but a considerable amount of stromal/fibrous tissue was still interspersed (Figure 7(e)). Masson’s staining showed mixed red and blue areas, with increased red-stained bone matrix but insufficient continuity (Figure 8(e)). At 6 months, the bone tissue became more mature and tended to be continuous (Figure 7(f)), and the extent of red staining further increased (Figure 8(f)).

Blank group: At 1 month, the defect area was dominated by fibrous/granulation-like tissue, with limited mineralized bone structure (Figure 7(g)). Masson’s staining showed predominantly blue-stained loose matrix (Figure 8(g)). At 3 and 6 months, the defect area still consisted mainly of fibrous connective tissue, without formation of continuous trabecular structures (Figure 7(h) and (i)). Red-stained bone matrix remained sparse and lacked broad continuous distribution on Masson’s staining (Figure 8(h) and (i)).

SEM results

Experimental group: At 1 month, abundant tissue-/matrix-like deposits were observed on the material surface and in the interfacial region, extending inward along the pore walls, with interfacial gaps still visible (Figure 9(a)). At 2 months, the coverage of these deposits increased, more continuous deposition was observed within the pores, and the interfacial gaps were reduced (Figure 9(b)). At 3 months, interfacial deposits became more abundant and denser, the interface boundary became less distinct, and focal fragmentation-like changes were observed in the pore walls of the material (Figure 9(c)). At 6 months, extensive tissue-/matrix-like filling was observed within the pores, and the bone tissue was more tightly apposed to the material surface, forming a continuous interfacial transition (Figure 9(d)).OPEN IN VIEWER

Positive control group: At 1 month, tissue-/matrix-like deposits were observed on the material surface and within the pores, and interfacial gaps were present (Figure 9(e)). At 2 months, the material surface appeared more fragmented than before, and the integrity of the pore structure was reduced, although the area covered by deposits had increased (Figure 9(f)). At 3 months, the interface boundary became less distinct, and deposition increased with a trend toward relative continuity (Figure 9(g)). At 6 months, the interfacial surface morphology became denser and smoother (Figure 9(h)).

BMP-2 mRNA expression (RT-qPCR)

RT-qPCR showed that the relative expression of BMP-2 mRNA in both the experimental and positive control groups was significantly higher than that in the blank group at 1–3 months postoperatively (Figure 10). In the experimental group, BMP-2 expression increased over time and reached its highest level at 3 months, at which point it was significantly higher than that in the positive control group (Figure 10). By 6 months, BMP-2 expression had declined to a low level in all three groups, with no obvious intergroup differences (Figure 10).OPEN IN VIEWER

Safety results

At 6 months postoperatively, H&E staining showed intact tissue architecture in the liver, lung, kidney, brain, and heart of the experimental group, without obvious necrosis, hemorrhage, or marked inflammatory cell infiltration (Figure 11(a)–(e)). In the femoral condyle implantation area, trabecular remodeling was observed; partial collapse of the material pore structure was accompanied by tissue ingrowth, and the trabeculae were in close contact with the material (Figure 11(f)). Masson’s staining showed increased collagen deposition and improved tissue maturation (Figure 11(g)). In the intramuscular implantation area, local calcium salt deposition and osteoid-/trabecula-like tissue formation were observed, without obvious myocyte lysis or necrosis (Figure 11(h)).OPEN IN VIEWER

With respect to inflammatory blood markers, ESR, WBC, and CRP showed transient elevations in all three groups during the early postoperative period, peaked at postoperative day 3, and then gradually declined (Figure 11(i)–(k)). The peak ESR, WBC, and CRP values in the experimental group did not differ significantly from those in the blank group (p = 0.473, 0.134, and 0.162, respectively) (Figure 11(i)–(k)).