Poly-l-lactic acid electrospun membrane with specific morphology promotes bone regeneration through macrophage reprogramming
Daiyuan Tang, Yunrong Xu, Zaitian Huang, Zhenping Xiao, Chenxun Sun, Jingyu Li, Fei He, Bing Han, Chen Zhu

TL;DR
A specific type of PLLA electrospun membrane promotes bone healing by changing macrophage behavior, offering new insights for regenerative biomaterials.
Contribution
The study reveals a novel mechanism of bone regeneration through macrophage reprogramming by PLLA electrospun membranes with specific morphology.
Findings
Aligned 600 nm PLLA electrospun membranes showed the best bone defect repair in rat models.
Macrophage subtype transformation was identified as a key factor in PLLA-induced bone regeneration.
A molecular pathway involving M2 macrophage polarization and osteogenic differentiation was confirmed.
Abstract
Host immune response to bone biomaterials plays a crucial role in influencing the outcome of bone regeneration. In this research, electrospun membranes composed of poly-l-lactic acid (PLLA) with different configurations (aligned and random) and sizes (nanoscale and microscale) were created, and the molecular mechanisms through which PLLA electrospun membranes facilitate bone regeneration were investigated. Firstly, when various material groups were inserted into the rat skull defect model, the bone defect repair was most obvious in the aligned 600 nm group compared with other groups. Then, single-cell RNA sequencing was performed on the bone tissues of the aligned 600 nm group to analyze the characteristics of the immune microenvironment, and the results showed that the effect of PLLA electrospinning membrane on bone regeneration was closely related to the transformation of macrophage…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 8- —Clinical Medical Technology Innovation Guidance Project of Hunan Province
- —Major Science and Technology Project of Yunnan Provincial Department of Science and Technology
- —China Postdoctoral Science Foundation
- —Henan Provincial Medical Science and Technology Research Joint Venture Project
- —National Natural Science Foundation of China
- —Anhui Provincial Key Research and Development Program-Clinical Medical Research Translation Specialization
- —Anhui Provincial Health Research Program
- —Anhui Provincial Scientific Research Compilation Project
- —USTC Research Funds of the Double First-Class Initiative
- —Research Funds of Centre for Leading Medicine and Advanced Technologies of IHM
- —Predictive modelling of regression after spinal cord injury by fusing machine learning with multimodal data
- —Innovative Research on Upper Limb Function Assessment and Rehabilitation Technology for Spinal Cord Injury Based on Multimodal Data Integration
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsBone Tissue Engineering Materials · Electrospun Nanofibers in Biomedical Applications · Mesenchymal stem cell research
Introduction
Bone defects that arise from iatrogenic causes, including trauma or the excision of tumors, present significant challenges in the field of clinical orthopedic surgery [1]. These conditions require effective intervention strategies to restore skeletal integrity and functionality, yet the complexities involved make treatment particularly demanding. Traditional treatment approaches, such as the use of autografts, allografts, and xenografts, have been widely employed to address such bone defects [2, 3]. However, these methods come with notable drawbacks. A primary concern is the limited availability of sufficient bone tissue, which can hinder the effectiveness of the repair process. Additionally, these treatments often result in high rates of morbidity for patients, as the procedures can be complicated and invasive. Furthermore, there may be issues with the compatibility of the grafts, as they do not always adapt well to the specific characteristics of the defects after surgery, leading to suboptimal outcomes [4].
Bone cement is frequently utilized in orthopedic surgeries [5]; however, its biological inertness, lack of degradability, and possible cytotoxic effects impede bone regeneration [6]. Additionally, titanium along with its alloys has been widely employed in both orthopedic [7] and dental treatments [8]. Poly-ether-ketone is a sophisticated polymer that demonstrates superior biocompatibility with lignin, making it a suitable choice for various medical applications, including cranial grafting [9–11]. Its favorable properties have attracted attention in the field of biomedical engineering. Nonetheless, a significant limitation of both metallic materials and poly-ether-ether-ketone polymers is their inability to integrate and remodel effectively with the host bone tissue. This lack of biological integration can result in complications such as aseptic loosening, a condition where the implant becomes unstable without the presence of infection. Addressing this issue typically requires strategies focused on biological regeneration, which underscores the need for developing materials that can better enhance the integration with bone [12]. Additionally, an acute response to foreign bodies or chronic inflammation at the interface between bone and implant may result in the unsuccessful integration of biomaterials with tissue [13].
In contrast to bioinert materials, polyesters like polylactic acid and polycaprolactone scaffolds exhibit a greater capability to regulate cell behaviors, including adhesion and proliferation [12]. This regulation can effectively assist in the healing of bone defects and enhance integration between the scaffold and surrounding tissue [14, 15]. Earlier research has concentrated on enhancing the osteogenic characteristics of polyester materials, whereas the natural immune reaction is often neglected and can result in unsuccessful bone healing [16]. Osteoimmunology posits that effective bone healing is rooted in the intricate interplay and communication between inflammatory cells and those responsible for bone formation [17]. This suggests that for optimal recovery of bone tissue, it is crucial to understand and facilitate the interactions among these different cell types. In this context, early regulation of the immune microenvironment during the initial phases of bone tissue repair is of significant scientific importance.
Once biomaterials are introduced into living organisms, the first response involves the attraction of macrophages to the implant location. Additionally, macrophages largely influence long-term immune reactions to biomaterials [18]. Macrophages adeptly adapt to varying microenvironments by skillfully transitioning between the classically activated M1 immunophenotype and the alternatively activated M2 immunophenotype [19, 20]. The M1 phenotype is distinguished by its capacity to generate inflammatory cytokines, including tumor necrosis factor α (TNF-α) and inducible nitric oxide synthase (iNOS), and plays a role in worsening inflammation. Conversely, the M2 phenotype, characterized by the secretion of arginase-1 (Arg-1) and interleukin (IL-10), is crucial in promoting tissue repair and sustaining tissue homeostasis [21]. The influence of a different macrophage polarization on tissue regeneration outcomes after scaffold implantation has been reported [22]. A transition towards M2 macrophage polarization is viewed as a beneficial adaptation [23]. For example, macrophages play an essential role in the mineralization of osteoblasts and the formation of bone in a living organism. Osteoinductive substances, such as transforming growth factor β (TGF-β) and bone morphogenetic protein 2, facilitate this process through secretion [24, 25]. The interactions between biomaterial scaffolds and macrophages have been widely investigated due to the significant role played by macrophages in the immune response associated with the material. Grasping how the scaffold’s structure interacts with macrophages is crucial for directing the immune response throughout the process of bone repair.
From the perspective of biomaterial design, the morphology of macrophages can be adjusted by altering the micromorphology of the matrix surface. This adjustment allows for the transformation of macrophage phenotype and function, making it an effective immune regulation strategy [26]. It has been observed that, for certain materials, the surface topography of the matrix has a greater impact on macrophage regulation compared to chemical signals in the microenvironment within 6–48 h of contact with biomaterials [27].
Compared to other scaffolds, the preparation process of electrospun fiber scaffolds is simpler, and their structure closely resembles the natural environment of cells. As a result, they are widely used in tissue repair. However, the impact of more complex electrospun fiber scaffolds on macrophage activation and immune response is still a subject of debate. Previous studies have demonstrated that electrospun fibers made of random polycaprolactone with a diameter of 5 μm have the ability to improve cell infiltration and facilitate macrophage polarization toward the M2 type when compared to fibers with a diameter of 600 nm [28]. Garg et al. further supported the notion that an increase in the diameter of electrostatic spun poly dioxanone random fiber scaffold, ranging from 350 nm to 2.8 μm, led to an upregulation in the expression of M2 macrophages. Moreover, these polarized macrophages exhibited the ability to enhance endothelial cell proliferation [29]. In contrast to this conclusion, experimental evidence has shown that filament nanofibers with a diameter of 500 nm induce a lesser inflammatory response when compared to photo synovial and micrometer fiber scaffolds with a diameter of 1.5 µm [30].
Taken together, the electrospinning technology is commonly used to prepare micro-nanofiber scaffolds for the repair of bone tissue engineering. Unfortunately, electrospinning technology is frequently employed to create micro-nanofiber scaffolds intended for the repair of bone tissue engineering. However, research examining how these fibrous scaffolds specifically influence changes in macrophage polarization and subsequently impact osteogenic differentiation is currently limited. The purpose of this study was to investigate the effect of poly-l-lactic acid (PLLA) electrospun membranes with different fiber diameters and orientations on bone regeneration and its potential regulatory mechanism through in vivo and in vitro experiments (Scheme 1). The findings from this research will aid in the creation and practical application of scaffolds employing spinning technology, offering a theoretical foundation for subsequent studies.
Scheme 1. Schematic illustration of the different morphologies of poly-l-lactic acid (PLLA) electrospun membranes manufactured by electrospinning technology and the mechanism of bone regeneration induced by bone immune environment with specific morphologies
Results
The PLLA electrospun membrane characterization
The aligned and random fiber membranes with different fiber diameters were fabricated by electrospinning technique. We had prepared PLLA electrospun membranes of different diameters by adjusting various parameters (as shown in Table S1). Ultimately, we chose 600 nanometers and 1200 nanometers precisely to finely characterize this possible biological switch. The diagrammatic sketch of the electrospinning process was shown in (Fig. 1K). Through the application of scanning electron microscopy (SEM), we analyzed the surface structure of the electrospun membrane composed of PLLA. The SEM images displayed a consistent, smooth, and uniform appearance across all four fiber groups. Specifically, the fiber orientations of random 600 nm (R600) and random 1200 nm (R1200) appeared to be randomly distributed, whereas the orientations of aligned 600 nm (A600) and aligned 1200 nm (A1200) exhibited alignment (Fig. 1A). The diameters of the fibers were measured and found to be 608.66 ± 33.04 nm for A600, 606.85 ± 36.41 nm for R600, 1223.83 ± 65.81 nm for A1200, and 1221.85 ± 62.95 nm for R1200. Statistical analysis showed no significant difference in fiber diameter between A600 and R600, A1200 and R1200 (Fig. 1I). In addition, the diameter distributions among the four groups exhibited a close approximation to the normal distribution (Fig. 1B). The centralization of fiber angle distributions was observed in groups A600 and A1200, whereas irregular distribution patterns were found in groups R600 and R1200 (Fig. 1C). Moreover, a water contact angle (WCA) meter was utilized for the analysis of the electrospun membrane’s hydrophilicity. Notably, the minimum angle measured in the A600 group was 92.86 ± 0.51°, displaying a statistically significant distinction from the remaining three groups. In the R600 group, the angle was 115.52 ± 0.40°; in the A1200 group it was 110.79 ± 0.62°; and in the R1200 group, it was 119.08 ± 0.35° (Fig. 1D and J).
Fig. 1. Characterization and mechanical evaluation of electrospun membranes with diverse morphologies. (A) Scanning electron microscope images displaying electrospun membranes with differing fiber diameters and orientations (n = 3); (B) Distribution of fiber diameters (n = 100); (C) Distribution of fiber orientations (n = 100); (I) Examination of fiber diameters (n = 100); (D, J) Analysis of water contact angle for electrospun membranes showing diverse morphologies (n = 3); (E) Stress-strain curves; (F) Young’s modulus; (G) Strain at failure; (H) Ultimate tensile strength; (K) Diagrammatic sketch of aligned electrospinning process. The results are expressed as Mean ± Standard Deviation (Mean ± SD). A one-way analysis of variance (ANOVA) test with Tukey’s post-hoc test was utilized. ^ns^p > 0.05, *p = 0.023, **p = 0.0041, ***p = 0.0003, and ****p < 0.0001
Uniaxial tensile tests were conducted to assess the mechanical properties of the PLLA electrospun membranes. A representative tensile stress-strain curve for these membranes is shown in Fig. 1E. The observed characteristics of these curves are consistent with Hooke’s law, which describes the elastic behavior of materials under stress. According to this principle, materials exhibit linear elasticity within a certain range, allowing them to return to their original shape after the applied force is removed. However, once the yield point is surpassed, the behavior of the material transitions into a phase where stretching and plastic deformation begin to occur. The trends observed in the ultimate tensile strength, strain at failure, and Young’s modulus of the membranes were comparable, suggesting that aligned fibers possess greater strength compared to random fibers. The Young’s modulus values for A600 and A1200 were found to be 869.12 ± 71.05 MPa and 604.75 ± 48.36 MPa, respectively, which were higher than the values of 263.68 ± 7.16 MPa and 115.30 ± 5.52 MPa recorded for R600 and R1200, respectively (Fig. 1F). The breakage strains recorded were 59.75 ± 3.83 MPa for A600 and 60.61 ± 5.46 MPa for A1200, which were lower than the values of 80.44 ± 2.69 MPa and 80.96 ± 5.21 MPa observed for R600 and R1200, respectively (Fig. 1G). The ultimate tensile strengths observed for A600 and A1200 were determined to be 20.90 ± 0.89 MPa and 16.64 ± 1.27 MPa, respectively. These values are notably greater than those recorded for R600 and R1200, which were 5.89 ± 2.42 MPa and 2.49 ± 1.15 MPa, respectively, indicating a statistically significant difference (Fig. 1H). This result shows that in the initial linear elastic stage, the stress-strain relationship is approximately linear. Meanwhile, this also indicates that it has a good shape recovery ability within the elastic deformation range.
In vivo repair of skull defect in rat by PLLA electrospun membranes
Before the bone regeneration study, we need to first detect the influence of the four groups of membrane implantation on various important organs. At the moment of collecting the rat skull specimen, additional organs such as the liver, colon, heart, lung, spleen, brain, and kidneys were also extracted for hematoxylin-eosin (HE) staining analysis. In comparison to the blank control group, there were no notable alterations in the histopathological structure among the four groups subjected to PLLA electrospun membranes implantation (Fig. 2A).
Fig. 2. In vivo performance of the four PLLA electrospun membranes on bone defect repair. (A) Representative histological examinations of hematoxylin-eosin staining (HE)-stained organs from rats receiving four kinds of PLLA electrospun membranes. Scale bar = 100 μm; (B) HE and Masson stainings of the engineered bone in the rat model at 4 and 8 weeks. NB: new bone, HB: home position bone, FT: fibrillar connective tissue, Scale bar = 150 μm; (C) Following the implantation of scaffolds for durations of 4 and 8 weeks, micro-CT scanning was used to obtain representative 3D reconstructed images as well as 2D topographical views of the calvarial bone. The hole diameter of the bone defect was 5 mm. The metrics of bone volume relative to tissue volume, bone mineral density, trabecular number, and trabecular separation bone mineral density, were calculated from the data obtained through the micro-CT analysis; (D) Immunohistochemical representative images and quantitative results of bone regeneration in the defect area at weeks 4 and weeks 8, scale bar = 100 μm; (E) Sirius red staining was applied to the tissue sections of bone defects observed at 4 and 8 weeks, with a scale bar of 500 μm. The image on the right provides a magnified perspective of the white outline, featuring a scale bar of 100 μm (n = 4 for each group in all experiments). The results are expressed as Mean ± SD. A one-way ANOVA test with Tukey’s post-hoc test was utilized. ^ns^p > 0.05, *p = 0.023, **p = 0.0041, ***p = 0.0003, and ****p < 0.0001
In order to assess if membranes facilitate bone fusion in Sprague-Dawley (SD) rats, various material groups were inserted into the surfaces of skull defects based on the previously mentioned grouping scheme. At 4 and 8 weeks post-implantation, the formation of new bone in the skull defects was assessed. Alongside histological analysis, 3D reconstruction of micro-CT images was conducted following the retrieval of the rat skull. The 3D visualization indicated a significant quantity of newly generated bone tissue in the A600 group at both 4 and 8 weeks, whereas minimal newly formed bone tissue was observed in the other groups. The quantitative analysis revealed that the A600 group showed the most elevated measurements for bone mineral density (BMD), bone volume relative to tissue volume (BV/TV), and trabecular number (Tb.N), concurrently exhibiting the lowest levels of trabecular separation (Tb.Sp) (Fig. 2C).
To assess the effects of different treatments on cranial defects in rats, histological staining was conducted on the collected samples at 4 and 8 weeks following the surgery. The images stained with HE and Masson depicting the cranial defects following various treatments revealed that the A600 groups exhibited a higher level of new bone formation compared to the other groups. Notably, the most significant bone value was recorded in the A600 group after 4 and 8 weeks, suggesting a more pronounced degree of osteogenesis (Fig. 2B). Furthermore, Masson staining revealed the presence of new bone tissue stained blue and cancellous bone that appeared red in the A600 group, suggesting a rise in calcification, maturation, and remodeling of the newly formed bone tissue (Fig. 2B).
Immunohistochemical (IHC) staining revealed that the cells that were recruited showed expression of CD146, which serves as a surface marker for BMSCs, in the defect region at 4 weeks, and this expression persisted for a duration of 8 weeks (Fig. 2D). The quantity of CD146 + cells found in the A600 group was considerably greater compared to the other groups at both time intervals. Additionally, a significant presence of runt-related transcription factor 2 (RUNX2, the primary transcription factor for osteogenesis), BMP2 (a promoter of osteogenic differentiation), and OPN (osteopontin) positive cells was noted in the defect region at 4 weeks, with a marked increase observed at 8 weeks in the A600 group. In summary, A600, when compared to the other groups, has the ability to attract a greater number of BMSCs to enhance osteogenesis within the defect region, aligning with our earlier results.
At 4 weeks, a small amount of collagen fibers could be observed in the immature area at the junction of scaffold material edge and old bone in all the tissue sections, while almost no new collagen fibers were found in the interior of scaffold material, with no noteworthy differences among the groups. As the duration of the experiment increased, a greater number of collagen fibers were observed dispersed within the A600 group at the 8-week mark. This observation indicated the impending formation of new bone, which aligned with the findings from HE and Masson staining. In the A600 group, a substantial quantity of reticular type I collagen was observed in the bone defect after 8 weeks, indicating a significant increase compared to the amount present at 4 weeks. The A600 group had obvious osteogenic induction potential in the indirect reaction of increased type I collagen production, indicating that the morphology parameters of A600 enhanced the osteogenic ability of scaffold materials (Fig. 2E). In conclusion, the A600 group had the strongest effect on promoting bone regeneration.
Single-cell RNA sequencing reveals bone immune modulation of PLLA electrospun membranes to promote bone regeneration
In order to explore the specific mechanism of PLLA electrospun membranes promoting bone regeneration, we selected skull defect rats without implanted material (model group) and skull defect rats implanted with A600 membrane (A600 group) for single-cell RNA sequencing. Quality control metrics for each sample, including cell counts, median UMIs and genes per cell, mitochondrial read percentage, and doublet rates, are summarized in Table S2. The detailed per-cell gene detection metrics across all samples are provided in Table S3. In order to explore the cell populations that are crucial for bone regeneration, eleven distinct cell clusters were identified using t-distributed stochastic neighbor embedding. These clusters include mural cell, endothelial cell, fibroblast, myeloid cell, proliferating cell, macrophage, neutrophil, monocytes, T cell, Schwann cell, and mast cell (Fig. 3A and B; Figure S1; Figure S2). Substantial variations were observed in the quantity of these nine cell clusters across the two groups (Fig. 3C and D). Notably, macrophages in the bone marrow microenvironment may be involved in bone regeneration (Fig. 3E).
Fig. 3. Single-cell RNA sequencing reveals bone immune modulation of membranes to promote bone regeneration. (A) T-distributed stochastic neighbor embedding visualization of different cell subsets in a rat model of bone defect; (B) Sanker plot showing the correlation between different cell clusters; (C) Cell composition of different samples; (D) Proportional expression heatmap of each cluster of differentially expressed genes; (E) Cell phone communication between macrophages and osteoblasts; (F) Targeted macrophages were analyzed specifically, and they were divided into 9 subgroups. Dot plots illustrating the marker expression for each macrophage cell cluster; (G) Violin plots depicting the marker expression associated with each macrophage cell cluster; (H) KEGG functional enrichment analysis of different subtypes of macrophages; (n = 3 for each group in all experiments); (I) Immunohistochemical representative images and quantitative results of macrophage polarization in the defect area at week 4 and week 8, Scale bar = 100 μm (n = 4). The results are expressed as Mean ± SD. A one-way ANOVA test with Tukey’s post-hoc test was utilized. ^ns^p > 0.05, *p = 0.023, **p = 0.0041, ***p = 0.0003, and ****p < 0.0001
Therefore, we conducted a subsequent subtype analysis of macrophages and carefully identified its surface markers. The Figure S3 visually presents the T-distributed stochastic neighbor embedding visualization plot showing these nine macrophage subclusters and the distribution of macrophage subclusters in each sample. The findings indicated that in the A600 group that received treatment, the surface markers of macrophage group Arg-1, CD206, VEGF, TGF-β, and IL-10 were significantly increased, while iNOS, CD86, TNF-α, IL-1β, and IL-6 were significantly decreased. In addition, KEGG functional enrichment analysis suggested that macrophage subtype transformation was most correlated with osteogenic differentiation. These results consistently indicate that electrospun PLLA membranes with specific micromorphology can promote M2 macrophage polarization and thus regulate bone regeneration (Fig. 3F-H). Subsequently, we conducted in vivo experiments, and the results from IHC staining indicated a significant enhancement in the levels of the M2 marker in the A600 group, while a marked decrease in the expression of the M1 marker was observed during both the 4th and 8th weeks (Fig. 3I). Furthermore, to directly validate the single-cell RNA findings, we performed immunofluorescence co-localization analysis of CD206 and BMP2 on calvarial defect tissues from the A600 group at 4 weeks post-implantation. The results revealed significant spatial co-localization of CD206⁺M2 macrophages with BMP2⁺osteogenic signals, particularly at the frontier of new bone formation and in perivascular regions (Figure S4). This spatial evidence demonstrates that A600-induced M2 macrophages are specifically localized within osteogenically active areas, providing key support for their direct role in regulating local bone formation. Therefore, we can conclude that the promotion effect of PLLA electrospun membrane on bone regeneration is closely related to the transformation of macrophage subtypes.
Macrophage depletion attenuates the pro-regenerative effect of the A600 membrane
To directly test the necessity of macrophages for the osteo-immunomodulatory effect of the A600 membrane, we performed systemic macrophage depletion using clodronate liposomes (CLO) in rats bearing A600 implants (Fig. 4A). The depletion protocol was designed to cover the critical early inflammatory phase (days 1–7 post-operation) and extend into the early repair phase (up to day 14). Flow cytometric analysis of peri-implant tissues at postoperative day 7-the peak of macrophage infiltration-confirmed a significant reduction in the proportion of macrophages in the A600 + CLO group compared to the A600 control group (Fig. 4B), validating the efficacy of the depletion protocol.
Fig. 4. Macrophage depletion attenuates the pro-regenerative effect of the A600 membrane. (A) The flowchart for clearing macrophages using CLO; (B) The clearance effect of macrophages was assessed using flow cytometry at day 7; (C) Micro-CT scanning was used to obtain representative 3D reconstructed images as well as 2D topographical views of the calvarial bone at weeks 4; (D) Quantitative results of bone regeneration in the defect area at weeks 4; (E) Quantitative bar graphs for BMD, BV/TV, Tb.N, and Tb.Sp; (F) Representative immunohistochemistry staining sections, scale bar = 100 μm; (G) Representative Masson’s trichrome staining sections, scale bar = 500, 100 μm; (H) Representative HE staining sections, scale bar = 500, 100 μm. (n = 4 for each group in all experiments). The results are expressed as Mean ± SD. A one-way ANOVA test with Tukey’s post-hoc test was utilized. ^ns^p > 0.05, *p = 0.023, **p = 0.0041, ***p = 0.0003, and ****p < 0.0001
Consistent with our hypothesis, macrophage depletion markedly attenuated the bone regenerative capacity of the A600 membrane. Micro-CT 3D reconstructions and quantitative analysis at 4 weeks revealed that the A600 + CLO group exhibited inferior bone healing compared to the A600 group (Fig. 4C). Quantitatively, key parameters including BMD, BV/TV, and Tb.N were significantly lower in the depletion group, while Tb.Sp was increased (Fig. 4E, all p < 0.01 vs. A600 group). Histological evaluations further corroborated these findings. Immunohistochemistry for the osteogenic transcription factor Osterix demonstrated a notable reduction in osteoblast-lineage cells within the regenerating tissue of the A600 + CLO group (Fig. 4F), indicating impaired osteogenic activity. Quantification of the new bone area confirmed a significant decrease upon macrophage depletion (Fig. 4D). Masson’s trichrome staining showed substantially less mature collagenous bone matrix deposition in the A600 + CLO group (Fig. 4G). HE staining also displayed a less organized tissue structure and reduced bone in growth in the defect area of the A600 + CLO group (Fig. 4H). Collectively, these results demonstrate that the pro-regenerative benefit of the aligned nanofibrous A600 membrane is critically dependent on macrophages. Their depletion during the key early phase disrupts the establishment of a regenerative immune microenvironment, leading to significantly compromised new bone formation and mineralization.
Effect of PLLA electrospun membrane on proliferation, adhesion, and biocompatibility of RAW264.7 and BMSCs
As it will be verified in vitro that PLLA electrospun membranes can promote bone regeneration by regulating bone immunity, it is necessary to test the biocompatibility of PLLA electrospun membranes with RAW264.7 and BMSCs. Figure 5A and B reveal that RAW264.7 and BMSCs treated with PLLA electrospun membranes were observed under bright field and scanning electron microscopy (1500 and 4000 magnification).
Fig. 5. Evaluation of the biocompatibility of electrospun membranes. (A, B) The morphology of RAW264.7 and bone mesenchymal stem cells (BMSCs) were observed under bright field and scanning electron microscope with different multiples; (C, D) Both RAW264.7 and BMSCs were evaluated using live/dead staining, Scale bar = 500 μm; (E-H) Visual depiction of RAW264.7 and BMSC cells adhered to electrospun membranes with different morphologies after 3, 6, and 12 h; (I, J) Cell Counting Kit-8 outcomes of RAW264.7 and BMSCs cultured on electrospun membranes with distinct morphologies at 1, 4, and 7 days (n = 4 for each group in all experiments). The results are expressed as Mean ± SD. A one-way ANOVA test with Tukey’s post-hoc test was utilized. ^ns^p > 0.05, *p = 0.023, **p = 0.0041, ***p = 0.0003, and ****p < 0.0001
The compatibility of biomaterials with biological systems is a crucial property. We conducted live/dead staining and adhesion tests to assess how electrospun membranes influence the proliferation and compatibility of RAW264.7. To assess the influence of electrospun membranes on the proliferation of RAW264.7, we employed the Cell Counting Kit-8 assay. Initially, no noteworthy variance in the effects of the four groups of membranes on cell proliferation was observed. Nevertheless, by the 4th day, the group of A600 exhibited the greatest activity in cell proliferation, displaying a significant disparity compared to the remaining three groups. Additionally, the R600 group exhibited a greater absorbance value compared to the R1200 group. Specifically, on the 7th day, the absorbance measurement in the A600 group surpassed both the R1200 and A1200 groups, while the R600 group again outperformed the R1200 group in terms of absorbance value (Fig. 5I). Following a 72-hour coculture of the RAW264.7 cells and electrospun membranes, we proceeded to evaluate the vitality of the RAW264.7 cells on the four groups of membranes using the live/dead staining technique. The findings demonstrate that most cells exhibited favorable survival rates, with just a handful of cell death occurrences noted among all groups (Fig. 5C). The results from the live/dead staining aligned with those from Cell Counting Kit-8 assay, implying that satisfactory viability was observed in all four cell groups. The same experimental method was used to detect BMSCs, and the experimental results were consistent with RAW264.7 (Fig. 5D and J). This indicates that the electrospun membranes have exceptional biocompatibility.
The effect of fiber diameter and alignment on cell adhesion was evaluated by observing 5-chloromethylfluorescein diacetate-labeled RAW264.7. Over time, the quantity of RAW264.7 adhering to the PLLA electrospun membranes displayed an upward trend. Cells exhibited distinct morphologies depending on the fiber alignment, with orderly arrangements on aligned fibers and disorderly arrangements on random fibers. However, no notable distinction was observed in the number of cells adhering when considering various morphologies simultaneously (Fig. 5E and G). The same experimental method was used to detect BMSCs, and the experimental results were consistent with RAW264.7 (Fig. 5F and H). The consistency of the above results indicated that PLLA had good biocompatibility with RAW264.7 and BMSCs.
The molecular pathway involving “PLLA electrospun membrane-M2 macrophage polarization-osteogenic differentiation” can serve as a potential mechanism for bone regeneration
To investigate how PLLA electrospun membranes affects the transformation of macrophage subtypes, RAW264.7 derived from mice were grown on the surface of PLLA electrospun membranes, after which the phenotypes of the macrophages were analyzed. Scanning electron microscopy revealed that RAW264.7 were evenly distributed on various electrospun membranes and the A600 stimulated the growth of pseudopodia in macrophages, leading to M2 polarization (Fig. 6A). Then, with the use of immunofluorescence (IF) staining and flow cytometry (FC) methods, the polarization of macrophages caused by the four different electrospun membranes with different morphologies was characterized. After 3 days of culture in the A600 and A1200 groups, the results of IF staining indicated the occurrence of M2 macrophages, characterized by CD68 + CD206+ expression. By contrast, the macrophages in the R600 and R1200 groups showed M1 macrophage indicators CD68 and iNOS (Fig. 6B). FC analysis revealed an increased expression level of the CD206 marker, indicative of M2 macrophages, in the A600 group. Specifically, the percentage of CD206 positive cells was observed to be 20.5% on day 1, while in comparison, it was 11.7%, 10.2%, and 9.49% in the A1200, R600, and R1200 groups, respectively. This observation of increased CD206 expression persisted on day 3 as well (Fig. 6C and Figure S5). These findings collectively suggest that the A600 group has the most pronounced effect on directing macrophages toward M2 phenotype polarization.
Fig. 6. Potential regulatory mechanism of PLLA electrospun membranes promoting bone regeneration. (A) The morphology and nanostructured characteristics of macrophages placed on electrospun membranes as observed via scanning electron microscope; (B) CD68 (green), CD206 (red), and nuclei (blue) were utilized for immunofluorescent staining of macrophages on electrospun membranes for 1 and 3 days. Scale bar = 5 μm; (C) Flow cytometry analysis of the CD206 marker, indicative of M2 macrophages, in the A600 group at 1 and 3 days; (D) The conditioned medium was prepared, as illustrated in Figdraw; (E, F) BMSCs were stained for alkaline phosphatase staining at 7 days and 14 days, while alizarin red S staining of BMSCs was performed at 14 days and 21 days; qualitative images of optical staining are shown, with a scale bar of 200 μm; quantitative analysis of ALP and activity (n = 3 for each group in all experiments). The results are expressed as Mean ± SD. A one-way ANOVA test with Tukey’s post-hoc test was utilized. ^ns^p > 0.05, *p = 0.023, **p = 0.0041, ***p = 0.0003, and ****p < 0.0001
To investigate the effect of immune microenvironment on osteogenic differentiation, we established a conditional medium co-culture system of RAW264.7 and BMSCs. Initially, the RAW264.7 were cultured for a single day using the A600 group electrospun membrane. Following this, conditional media were collected from macrophages to filter the cells for further culture of BMSCs. This was designated as the RAW264.7 + PLLA group. Pure RAW264.7 and membrane were also cultured in identical medium for one day, and then their supernatants were extracted respectively as conditioned media groups of different controls. These were designated as the RAW264.7 group and the PLLA group. At the same time, we cultured pure BMSCs as a blank group (Fig. 6D). The evaluation of the osteogenic differentiation of BMSCs cultured in various conditioned mediums was conducted using alkaline phosphatase (ALP) staining and lizarin red s (ARS) staining. After 7 and 14 days of culturing, the ALP staining color of BMSCs in the RAW264.7 + PLLA group appeared darker compared to the other groups. Quantitative analysis of ALP activity indicated that the RAW264.7 + PLLA group at day 7 (8.735 ± 0.480) and day 14 (12.630 ± 0.665) displayed the highest level of ALP activity (Fig. 6E). The assessment of osteogenic activity in BMSC cultures was conducted by evaluating extracellular matrix (ECM) mineralization using ARS. A comparable pattern emerged from the ARS analysis, indicating that the highest calcium levels were observed in cells cultured with RAW264.7 + PLLA for 14 days (2.232 ± 0.160) and 21 days (2.797 ± 0.138) in the respective conditioned medium (Fig. 6F). Furthermore, conditioned medium prepared from A600 alone failed to induce osteogenic differentiation in BMSCs, whereas conditioned medium secreted by A600-reprogrammed macrophages robustly promoted osteogenesis. This contrast directly ruled out the possibility that the material acts directly on BMSCs via soluble factors, confirming the indispensable role of macrophages as signal transducers and the necessary step for converting the topological cues into effective biological instructions. Thus, the co-culture of RAW264.7 and BMSCs identified that the molecular pathway involving “PLLA electrospun membrane-M2 macrophage polarization-osteogenic differentiation” could serve as a potential mechanism for bone regeneration.
To further validate the sufficiency of the M2-polarized microenvironment, we performed a conditioned medium rescue assay (Figure S6). The results showed that BMSCs treated with conditioned medium from macrophages co-cultured with the R600, A1200, and R1200 membrane exhibited low levels of ALP activity (day 14) and mineralized nodule formation (day 21). In contrast, BMSCs treated with conditioned medium from the A600 group showed a significant increase. This indicates that the microenvironment created by A600 via macrophage reprogramming is, by itself, sufficient to drive robust osteogenic differentiation of BMSCs.
Discussion
Electrospinning facilitates the straightforward production of fibers at the microscale and nanoscale [31]. By modifying specific factors, the diameter and alignment of the fibers can be regulated [32–34]. These factors are classified as solution, process, and environmental parameters. In this research, membranes produced through electrospinning were created with varying fiber diameters and orientations. A 16% solution of PLLA was utilized, while modifications were made to the applied voltage, flow rate, receiving distance, and the kind of receiving apparatus.
The important physicochemical characteristics of biomaterials, including their mechanical and hydrophilic properties, have a direct influence on their biological properties. The mechanical properties and hydrophilicity of the electrospun membranes were significantly affected by both the orientation and diameter of the fibers. Jin and his team developed three types of electrospun membranes with aligned, random, and lattice configurations, demonstrating that the mechanical properties of the aligned fibers exceeded those of the other two types [35]. According to He et al., the mechanical characteristics of PLLA electrospun membranes containing aligned fibers were found to be better than those featuring random fibers [36]. In a similar vein, our results align with earlier research, indicating that fibers arranged in an aligned manner exhibit greater tensile strength. This occurrence may arise from the organized configuration of the aligned fibers, which restricts tensile movement. Previous studies have shown that membranes made from aligned fibers demonstrate better hydrophilicity than those composed of randomly arranged fibers [37, 38]. Furthermore, it has been observed that there exists an inverse relationship between the diameter of the fibers and their hydrophilicity, indicating that fibers with greater diameters tend to be less hydrophilic [39]. Our research indicates that, due to the hydrophobic characteristics of PLLA, the WCA for all four categories of electrospun membranes to exceed 90°. Nevertheless, the A600 group exhibited a lower WCA compared to the other three groups, likely attributable to the fibers’ smaller diameter and organized alignment. Under the condition of a basically constant porosity, we observed that fiber diameter and alignment have significant effects on cell behavior and tissue regeneration, suggesting that they may independently function through direct nanoscale topological signals, independent of the macroscopic pore structure.
To investigate the impact of PLLA electrospun spinning membranes on the local microenvironment and their role in endogenous bone regeneration, we implanted PLLA electrospun membranes with different morphologies into classical rat skull defect models to observe the changes in their effects. However, before we do this, we perform histocompatibility tests to understand the impact of the material on important organs in the body. After conducting the HE staining test, no notable alterations were detected in the histopathological structure among the four groups within the spinning membrane implantation category in comparison to the blank control group. This shows that after our material is implanted in the body, there is no inflammation, cancer and rejection of the tissue.
Bone tissue exhibits significant anisotropy due to the systematic organization of collagen fibers. Research indicates that BMSCs are more likely to differentiate into an osteogenic phenotype when this specific arrangement is maintained [40]. According to Gao et al., hMSCs exhibited enhanced osteogenic differentiation when exposed to aligned fibers compared to random fibers [41]. Similarly, Xie et al. discovered that BMSCs demonstrated greater osteogenesis on nanofibers than on microfibers, and that the osteogenic induction potential of aligned nanofibers surpassed that of random nanofibers [37]. These experiments have strongly proved that our spinning membrane can be compared to an effective periosteum, which plays a certain role in repairing bone defects. In our investigation of skull defect models, micro-CT analysis revealed that the bone defect repair efficacy in the A600 group was markedly superior compared to the other groups. Additionally, subsequent experimental findings, including HE staining, Masson staining, and Sirius scarlet staining, corroborated this observation. Targeting host stem cells to areas of deficiency plays a crucial role in effective bone regeneration. In cases of pathological injury to bone tissue, BMSCs are primarily mobilized from the bone marrow and attracted to the injury site, where they later transform into osteoblasts to aid in the repair of bone [42, 43]. Research has demonstrated that biomaterials with osteoinductive properties have the ability to capture endogenous growth factors present in the bloodstream, which can subsequently encourage the recruitment of BMSCs for the regeneration of new bone44. Here, immunohistochemistry proves this point again. To sum up, the effect of membranes with different morphologies on bone defect repair after implantation in vivo has already appeared, but whether there is a relationship between this and the immune microenvironment in vivo, we tried to solve this mystery by using single-cell RNA sequencing.
Recent advancements in single-cell RNA sequencing have transformed the investigation of the bone marrow microenvironment and associated diseases [45, 46]. The examination of different pathogenic T-helper 17 cell types was conducted through the integration of single-cell RNA sequencing and mouse models of experimental autoimmune encephalomyelitis [47]. Furthermore, single-cell RNA sequencing was employed to examine T cells that infiltrate tumors, uncovering exhaustion programs in human metastatic melanoma [48]. Nevertheless, studies on the combination of single-cell RNA sequencing with bone defect repair are still scarce. Our research identified interactions among various cell types within the bone marrow microenvironment using bioinformatics analysis, thereby aiding in the comprehension of the significance of these interactions. In addition, according to the results of macrophage subtype analysis, PLLA electrospun membranes promoted osteoblast differentiation through M2 macrophage polarization, expanding the use of single-cell technology. These findings further establish a foundation for comprehending how the immune microenvironment influences the application of biomaterials.
At the cellular scale, the surface architecture of electrospun membranes plays a crucial role in modulating cell adhesion, migration, and morphology, thereby greatly enhancing interactions with cells [32]. The survival, proliferation, and growth of cells attached to electrospun membranes depend on adhesion. Various morphologies have different impacts on cell adhesion, mainly determined by the surface’s roughness, fiber orientation, and the lateral arrangement of nanoscale characteristics [49]. Park et al. conducted a study on the adhesion of MSCs to oriented titanium dioxide nanotubes. They observed that the extent of adhesion was strongly influenced by the diameter of the nanotubes, which ranged from 15 to 100 nm. Specifically, the cells exhibited strong adhesion when the nanotube diameter was between 15 and 30 nm. However, when the diameter exceeded 50 nm, both adhesion and cell spreading were found to be significantly reduced [50]. In this study, the adhesion of two kinds of cells were not significantly affected by the four morphologies. The absence of noteworthy variations in fiber roughness and lateral spacing may be a contributing factor. However, material morphology played a crucial role in guiding cell migration. The investigation revealed that scaffolds possessing suitable porousness and porous architecture brought advantages to the growth and differentiation of cells, along with their adhesion characteristics. These factors stand as crucial determinants for the macrophage activation result. Particularly, the nanofiber scaffolds with well-arranged structures exhibited noticeable amplification in the proliferation and differentiation of macrophages in comparison to the unordered fibers. Furthermore, it stimulates macrophage differentiation toward the M2 phenotype.
In our investigation, it was observed that the A600 group demonstrated increased expression of typical M2 markers on the surface of macrophages in the early stages. Conversely, the R600 group exhibited elevated levels of M1 marker molecules. Additionally, the macrophages in the A600 group displayed elongated morphology. Previous research has indicated a strong correlation between their activation state and macrophage shape, suggesting that feedback from the shape of macrophages might play a role in regulating their phenotype [51–53]. Notably, the orientation difference of PLLA electrospun membranes surfaces with different orientations would not be greatly reduced with the degradation of PLLA electrospun membranes in the early stages [54]. To sum up, the arrangement of fibrous membranes has the ability to stimulate macrophages to differentiate into an anti-inflammatory M2 phenotype.
In addition, this study found that the A600 group recruited more CD146 + BMSCs in the bone defect area. This suggests that its regenerative mechanism not only lies in regulating the fate of BMSCs that have been in place but may also actively enhance the homing efficiency of BMSCs to the damaged site. Our macrophage clearance experiments further demonstrated that macrophages play a key role in this process. We speculate that activated M2 cells are likely to secrete specific chemokines, such as SDF-1/CXCL12, which play a core role in BMSCs homing [55], thereby forming a chemical concentration gradient at the defect site and guiding BMSCs in the peripheral or bone marrow to migrate directionally to the repair area. This mechanism of material morphology guiding immune cells to secrete specific signals and then remotely recruit stem cells more closely couples immune regulation with tissue regeneration, clarifying the deep logic by which biomaterials achieve efficient regeneration by regulating the endogenous repair system.
To explore more about the role of PLLA electrospun membranes in enhancing endogenous neo-bone formation and to determine if macrophage polarization has an impact on this process. We set up a conditioned medium cell co-culture mode. Given that the differentiation of osteoblasts is essential for enhancing the effectiveness of bone regeneration [56], the osteogenic properties of different scaffolds were systematically assessed through ALP staining and ARS staining. ALP, recognized as a key biomarker for osteogenic differentiation, is crucial in the process of bone mineralization through the stimulation of pyrophosphate hydrolysis [57]. ARS staining was employed to investigate the mineralization stages of osteogenic-differentiated BMSCs in different groups by coloring calcium nodules red. No matter ALP staining or ARS staining, it is not difficult to see that the maximum osteogenic effect can be stimulated by co-culture of PLLA electrospun membranes with the supernatant of macrophage culture and BMSCs. The in vitro dual-control experiment provides the most direct evidence for this immune-mediated mechanism. We designed stringent controls to parse the origin of the pro-osteogenic signal: the conditioned medium from the A600 material itself had no activating effect on the osteogenic differentiation of BMSCs, which definitively rules out the possibility that the material exerts a direct chemical stimulus through leachables or surface physicochemical properties. However, when the same material acted on macrophages, the resulting conditioned medium exhibited potent pro-osteogenic efficacy. This indicates that macrophages are not passive carriers but active biological signal processors, responsible for receiving the material’s topological information and outputting a precise secretome program that guides tissue regeneration. In summary, the development of a co-culture model using conditioned media suggests that the transformation of macrophage subtypes due to host immune responses apparently enhances biomaterial-mediated bone regeneration.
Up to this point, even though we have experienced a degree of success, our study does have certain limitations. First and foremost, it is important to emphasize that the local immune reaction following the implantation of biomaterials might extend beyond just the activation of macrophages. A prior investigation demonstrated that collagen scaffolds with silica promote bone healing and the formation of blood vessels through the immunomodulation of monocytes [58]. The current study illustrates that the hierarchical nanointerface, inspired by biological models, facilitates the polarization of host M2 macrophages and enhances the recruitment of BMSCs while supporting their osteogenic differentiation, thereby offering valuable insights into the role of biomaterials in promoting endogenous bone regeneration. Future studies will concentrate on different immune cells, including neutrophils involved in the innate response and T cells associated with the adaptive response, as well as their roles in the hierarchical nanointerface-guided regeneration of neo-bone. In addition, it is essential to develop various animal models to investigate the characteristics of this spun membrane more thoroughly and enhance it in the future. Crucially, we will also examine if this material has the ability to stimulate angiogenesis and suppress osteoclast activity, alongside its role in modulating the immune microenvironment.
Collectively, our results indicate that membranes made from PLLA electrospun membranes are essential for bone regeneration by facilitating the polarization of M2 macrophages (Fig. 6). These results have significant implications for scaffold design, understanding immunoregulatory mechanisms, and potential clinical applications of electrospinning technology in the future. The PLLA electrospun membranes, by establishing a specific early immune tone, creates a permissive niche that guides these downstream events. Validating this hypothesis will not only strengthen the mechanistic narrative but also provide key parameters for designing next-generation biomaterials aimed at coordinately optimizing the immune, vascular, and remodeling phases of bone healing. Bone repair is a finely orchestrated process involving not only osteogenesis but also concomitant angiogenesis and coupled bone remodeling. Our single-cell RNA sequencing data offer promising clues, suggesting that the M2 microenvironment induced by PLLA electrospun membranes may favor these processes-for instance, through the upregulation of pro-angiogenic transcripts in endothelial cells and the presence of a regulated osteoclast precursor population without signs of excessive activation. However, precise spatiotemporal coordination among immune cells, nascent blood vessels, and bone-resorbing cells is critical for functional outcomes [72]. Therefore, in future studies, direct quantitative assessment of angiogenesis and osteoclast-mediated remodeling-coupled with validation of specific pathways via genetic knockdown or pharmacological inhibition in vitro-will be paramount for translating the immunomodulatory principle identified here into a comprehensive understanding of the full regenerative sequence.
Conclusion
The current study elucidates the molecular mechanisms through which PLLA electrospun membranes facilitate bone regeneration. Our findings indicate that the A600 group is crucial for endogenous bone regeneration, as it fosters M2 macrophage polarization. This process ultimately facilitating the osteogenic differentiation of BMSCs and thereby aiding in bone formation. Using the electrospinning technique, we effectively fabricated PLLA electrospun membranes with diverse morphologies. Then, based on single-cell RNA sequencing technology, we further found that the molecular axis of “PLLA electrospun membranes-M2 macrophage polarization-osteogenic differentiation” may serve as a potential mechanism for bone regeneration. To be specific, the A600 group showed a higher tendency to polarize RAW264.7 cells toward alternatively activated macrophages (M2) compared to other types. This was evident from the rise in M2 phenotypic markers and the drop in classically activated macrophage (M1) markers in the cells. In summary, the polarization of M2 macrophages induced by PLLA electrospun membranes significantly aids in the natural regeneration of bone. The findings could offer valuable perspectives for the development of biomimetic materials focused on tissue regeneration, particularly concerning the modulation of bone immunity.
Materials and methods
Fabrication and characterization of electrospun membrane
Bio-grade PLLA with an average molecular weight of 16 KDa was obtained from Daigang Biomaterial (Jinan, China). Electrospinning equipment from Yong Kang Le Ye (Beijing, China) was utilized to create both aligned and randomly dispersed fibers that possessed diverse diameters. Initially, PLLA was dissolved in a hexafluoroisopropanol solvent (Mclean Biochemical Technology, Shanghai, China) and then converted into a spinning solution with a concentration of 16% w/v. Subsequently, the homogeneous solution was carefully injected into a syringe and modified by adjusting various factors, including the distance between the injection source and the receiver, voltage, receiver speed, and injection speed. The study’s pertinent parameters are displayed in Table 1. The aligned fiber receiver had a rotational speed of 2800 rpm, while the random fiber was rotated at a speed of 100 rpm. To ensure the removal of any residual organic solvent, prepared electrospun membranes were kept at room temperature for a period of 3 days. In this study, the researchers classified the aligned nanofibers and microfibers as A600 and A1200, respectively. Similarly, the random nanofibers and microfibers were identified as R600 and R1200.
Table 1. The diameters of fibres and fabrication parametersGroupVoltage(kV)Supply rate(mm/min)Needle type(G)Receiving distance(cm)Rotating speed(rpm)A60060.0528142800R60090.033020100A120070.252122800R120060.052814100
To analyze the surface morphological features of the four groups of membranes, it was first cut into 1 × 1 cm^2^. The next step involved spraying the surfaces of the membrane samples with gold. Finally, the samples were observed using the scanning electron microscope (Thermo Fisher, MA, USA), then the figures were marked the scale with high definition and high contrast to obtain more relevant information. Furthermore, 100 fibers were randomly selected from each group, and the ImageJ software was used to calculate the diameter and orientation angle of these fibers. The Origin software was then used for mapping and statistical analysis. The WCA meter (Thermo Fisher) was employed to assess the hydrophilicity of the sample. To determine the WCA, a flat test bench was utilized to secure the sample, and subsequently, a drop of 5 µl of deionized water was released from a 5 cm altitude onto the surface of the sample. An image of the water droplet on the membrane surface was immediately captured, and utilizing tools for image analysis, the WCA was determined. This process was repeated three times for accuracy.
A universal testing machine (EZ-LX, Shimadzu, Japan) calibrated to a load of 50 N was utilized to perform uniaxial tensile testing. Rectangular specimens, measuring 5 × 1 cm^2^, were prepared, with sandpaper securing both ends. The sample length was recorded before setting a constant crosshead speed of 1 mm/min. Comprehensive documentation of stress-strain curves was conducted, encompassing five samples within the analysis. These curves act as fundamental instruments for assessing the mechanical characteristics of materials when subjected to stress. Utilizing the information gathered from these observed stress-strain connections, we went on to compute vital material parameters, such as ultimate tensile strength, failure strain, and elastic modulus. Such computations yield significant understanding of the materials’ strength and ductility, along with their capacity to undergo elastic deformation when loads are applied.
Fig. 7. Schematic diagram of mechanism of PLLA electrospun membranes promoting bone regeneration. PLLA electrospun membranes (especially in the A600 group) can induce the transformation of macrophages into M2 subtype and finally promote the osteogenic differentiation of BMSCs to achieve bone defect repair. Therefore, we compare the material itself, PLLA electrospun membranes, to the “gold-encrusted staff” held by the Monkey King “Sun Wukong” in the classic Chinese novel Journey to the West. It can be made small or large, symbolizing the freedom to change the surface morphology of the material. Monkey King also has a magical skill: the Cloud Raft, with one segment representing the early stage of macrophage reprogramming and the other segment representing the later stage of osteogenic differentiation. In conclusion, Sun Wukong, with the “gold-encrusted staff” (PLLA electrospun membranes) in his hand, and with the help of the bone immune microenvironment, defeated the demon “White Bone Spirit” that caused bone defects through macrophage reprogramming, and finally promoted bone regeneration
Porosity measurement
The porosity of the electrospun membranes was determined using the liquid displacement method (also known as the density method). Briefly, a rectangular sample with a known area (S) and thickness (h) was precisely cut and weighed for its dry mass (m) using an analytical balance (precision: 0.1 mg). The sample was then immersed in a liquid with a known density (ρ_liquid, e.g., absolute ethanol) and placed in a vacuum desiccator. Vacuum was applied until no bubbles emerged, ensuring the liquid thoroughly filled the pores within the membrane. The saturated sample was removed, and any residual liquid on the surface was quickly blotted with filter paper before immediately weighing to obtain the wet mass (m_wet). The true density of PLLA material (ρ_PLLA) was taken as 1.25 g/cm³. The porosity (P) was calculated using the following formula:
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$P(\%) = \frac{(m_{wet}-m)/\rho_{liquid}}{S\times h}\times 100\%$$\end{document}or
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$P(\%)=(1-\frac{m}{\rho_{PLLA}\times S \times h})\times 100\%$$\end{document}For each sample group, at least five parallel specimens were measured, and the results are presented as the mean ± standard deviation. Furthermore, the pore morphology of the membranes was observed via SEM.
Experimental animals
The experimental methods, housing conditions, and animal care are regulated and approved by the Animal Ethics Committee at Kunming Medical University (kmmu20230620). The SD rats, with a weight range of 150–180 g, were acquired from the Department of Animal Experimentation at Kunming Medical University. The animals were kept in standard environmental conditions, which included 12 h of light and 12 h of darkness, a temperature range of 18 to 22 °C, and humidity levels of 55 ± 5%. They had access to a standard diet for rats and were allowed to drink water freely. Initial experiments were carried out for all animal research to ascertain the necessary sample size. All surgical procedures were executed under anesthesia, and measures were taken to reduce any discomfort. For all surgical procedures, rats were deeply anesthetized by intraperitoneal injection of sodium pentobarbital (50 mg/kg). Depth of anesthesia was confirmed by loss of pedal withdrawal reflex. Post-operative analgesia was provided via subcutaneous injection of buprenorphine (0.05 mg/kg) to ensure animal welfare. At the experimental endpoints (4- and 8-weeks post-surgery), all rats were euthanized exclusively by CO₂ asphyxiation. Animals were placed in a dedicated euthanasia chamber and exposed to a controlled flow of 100% CO₂ at a fill rate of 30–70% of the chamber volume per minute until respiratory arrest, followed by confirmatory bilateral thoracotomy. This protocol is in full compliance with the AVMA Guidelines for Euthanasia. Moreover, detailed sample size design for each animal experiment is provided in Table S4.
Rat bone defects
To create a model of cranial defects, rats were anesthetized using sodium pentobarbital administered via intraperitoneal injection. Post-operative analgesia was provided via subcutaneous injection of buprenorphineto ensure animal welfare. The head of the SD rat was first subjected to hair removal and then disinfected with iodophor. An incision approximately 5–8 mm long was created in the skin, and symmetrical incisions were executed on both the left and right sides of the calvaria using a 5 mm diameter ring bone drill, with the skull’s midline serving as a reference point. This approach created two circular bone defects, each with a diameter of 5 mm. The cranial defects in the bones were complete, with the dura mater remaining undamaged [59, 60]. Throughout the drilling procedure, sterile saline is used to flush the drill bit and keep it cool. The left skull defect was filled with the ordered PLLA fiber material, whereas the disordered PLLA fiber material was placed in the right skull defect. The layers of skin and subcutaneous tissue were stitched together in layers using 4 − 0 silk thread. Following the procedure, the rats were maintained at a warm temperature until they became alert again and were provided with regular food in individual cages.
In vivo biocompatibility
To evaluate the biocompatibility of PLLA in vivo, all rats were euthanized by CO₂ asphyxiation at the 4th and 8th weeks after PLLA implantation. After euthanasia, a variety of vital organs were collected for analysis, including the hearts, livers, spleens, lungs, colons, kidneys, and brains. To preserve the integrity of the collected tissues, they were fixed in 4% paraformaldehyde. This fixation process is essential for maintaining cellular structure and morphology, which is crucial for the subsequent histological examination. The tissues were then sectioned into 5 μm thick slices, prepared for HE staining. The detailed methodology for performing HE staining will be described in the following section, providing a framework for the examination and analysis of the histological characteristics of the tissues.
Micro-CT
At weeks 4 and 8 post-cranial defect surgery, the rats were euthanized via CO₂ asphyxiation, and the surrounding soft tissues of the skull were removed to retrieve the complete calvarial bone. Bone tissues were stored in a 4% paraformaldehyde solution. The samples underwent scanning via micro-CT (Xuanzun Bioscience, Chongqing, China) with a tube voltage of 90 kV and a current of 60 µA. The frame rate for the scanning process was established at a rate of 20 frames per second, ensuring a smooth capture of data during the imaging procedure. Additionally, the parameter was specifically calibrated to 395 mm, which facilitated both the scanning process and the subsequent three-dimensional reconstruction of the skull. To analyze the intricate microstructure of the bone tissue, we employed Scanco software, which is renowned for its capability in processing and evaluating medical imaging data. The evaluation included measurements of BMD, BV/TV, Tb.N, and Tb.Sp.
Histological and IHC analysis
For the histological assessment of bone, the skulls were preserved in a 4% paraformaldehyde solution for 48 h. This was succeeded by decalcification using a 10% ethylenediaminetetraacetic acid solution for either 4 or 8 weeks. Subsequently, the samples were embedded in paraffin wax and sliced into 5-µm-thick sections for various staining procedures, including HE staining, Masson staining, Sirius scarlet staining, and IHC analysis [61].
For HE staining, the sections underwent dewaxing with water, followed by a 5-minute staining with hematoxylin. After washing with water, they were then treated briefly with 1% hydrochloric alcohol. Following a rinse with water, eosin was applied to stain the cytoplasm for 1 min, after which it was dehydrated and sealed with neutral resin for examination. In the trichrome staining process developed by Masson, the previously prepared thin sections were dewaxed and then immersed in a Wiegert iron hematoxylin solution (Sigma-Aldrich, MO, USA) for 5 min. The sections were distinguished in a differentiation solution of acidic ethanol (1% ethanol hydrochloride solution; Sigma-Aldrich) for 10 s and subsequently rinsed with water. The sections were subsequently treated with a 1% aqueous solution of lithium carbonate (Solarbio, Beijing, China), stained blue for 3 min. The sections were later subjected to staining using a carmine acidic magenta buffer (Guidechem, Hangzhou, China) for 5 min, followed by a rinse with a dilute acid solution (Sigma-Aldrich) for 1 min. The sections were first exposed to a solution of phosphomolybdic acid (Sigma-Aldrich) for a duration of 2 min, after which they were rinsed with a diluted working solution of acid for 1 min. Subsequently, the sections were submerged in a staining solution of aniline blue (Sigma-Aldrich) for 2 min. Following this, the sections were rinsed with a mildly acidic working solution for 1 min. After the dehydration process, a neutral resin was used to seal the sections. For the Sirius scarlet staining procedure, decalcified bone tissue that had been embedded was sliced into 6 μm thick sections using a paraffin microtome. These sections were then stained with a Sirius red dyeing solution (Sigma-Aldrich) for 1 h at room temperature. Rinse the section with water once and thoroughly remove the stain indicated by the section. The nucleus was stained with Mayer hematoxylin staining solution. After staining, rinse the staining solution with water once. Collagen formation was observed under an inverted microscope and photographed. A two-step detection kit (Zhongshan Golden Bridge Biotechnology, Beijing, China) was employed for IHC as previously described [62]. In brief, samples were placed in a solution designed for antigen retrieval for a duration of 20 min, subsequently undergoing a 30-minute blocking procedure that utilized 5% bovine serum albumin (Beyotime Biotechnology, Shanghai, China). Following this, the specimens were incubated for 24 h at 4 °C with primary antibodies at a dilution of 1:100. The antibodies employed included those that target rat CD146 (#ab75769; Abcam, MA, USA) as markers for stem cells, along with OPN (#88742; Cell Signaling Technology, MA, USA), RUNX2 (#12556; Cell Signaling Technology), and BMP2 (#ab284387; Abcam) which are involved in osteogenesis. Moreover, CD68 (#ab283654; Abcam) was utilized as a pan-macrophage marker, while CD206 (#ab64693; Abcam) and iNOS (#ab178945; Abcam) acted as markers for M2 and M1 macrophages, respectively. Following a thorough rinse in phosphate buffer saline, secondary antibodies conjugated with horseradish peroxidase (Zhongshan Golden Bridge Biotechnology) were applied onto the slides. Each group consisted of more than three slides, and observations were made using a Zeiss light microscope, focusing on the area of the defects that included the scaffolds.
Spatial co-localization Immunofluorescence
To validate the spatial relationship between M2-polarized macrophages and osteogenic activity suggested by single-cell RNA sequencing, co-localization immunofluorescence was performed on fresh-frozen cryosections of calvarial defect tissue harvested at postoperative week 4 from the A600 group. Briefly, tissue Sect. (8 μm thickness) were fixed, permeabilized, and blocked. Sections were then incubated overnight at 4 °C with primary antibodies against CD206 (#ab64693; Abcam) and BMP2 (#ab284387; Abcam). After washing, sections were incubated with appropriate secondary antibodies conjugated to Alexa Fluor 488 (for CD206) A Alexa Fluor 594 (for BMP2) for 1 h. Nuclei were counterstained with DAPI. Fluorescent images were captured using a confocal laser scanning microscope (Leica TCS SP8) under consistent acquisition settings. Co-localization analysis was performed using ImageJ software.
Generation and processing of single-cell RNA sequencing data
Libraries for single-cell RNA sequencing were created utilizing a microfluidics-based method on the Chromium Single-Cell Controller (10X Genomics), following the guidelines provided by the manufacturers and employing the Chromium Single Cell 3’ Reagent Kit v3.1. Briefly, Single cells were enclosed in gel beads within an emulsion and then lysed, after which RNA barcoding, reverse transcription, and PCR amplification (between 13 and 15 cycles) were performed. Sequencing of the libraries was conducted using an Illumina NovaSeq 6000 machine (paired-end, with a read length of 150 bp). The Fastq files underwent processing via Cell Ranger (version 4.0.0), employing the default settings [63]. The reads were aligned to the Rnor_6.0 and mRatBN7.2 genomes from Ensembl using STAR v2.7.8a. To generate a gene expression matrix indicating the number of unique molecular identifiers for each cell and gene, only the confidently mapped reads with valid barcodes and unique molecular identifiers were preserved. Gene counts were brought into the R environment (version 4.0.3) and analyzed using Seurat (version 4.0.3). In this analysis, we set specific doublet rates (dbr) to effectively filter out potential doublets present in our samples. For the samples, the doublet rate was imputed at 0.07 [64]. Dbr were created in accordance with the quantity of loaded cells, following the guidelines provided by 10X Genomics. Cells with unique molecular identifier counts below 1,000 were removed from the analysis. Additionally, cells that displayed fewer than 200 genes in the samples weas eliminated. Normalized expression data were obtained through log2 transformation using the NormalizeData function. Finally, the data were scaled with the ScaleData function, which were calculated earlier using the CellCycleScoring function. The FindVariableFeatures function was utilized to calculate the top 3,000 genes that exhibited the highest standardized variance.
Cell clusters were identified through the Louvain algorithm using the FindCluster function. Uniform Manifold Approximation and Projection was employed for two-dimensional visualization [65]. The identification of cluster-specific genes was performed utilizing the FindAllMarkers function with the parameters only.pos = TRUE and min.pct = 0.1, establishing a threshold of FDR < 0.01. Single-cell copy number variations were estimated with the CopyKAT R package (version 1.0.5) [66]. Cell clusters were established using a marker-based manual annotation approach that had been previously applied to each dataset. Cell clusters were annotated into major cell types based on the expression of established canonical marker genes, detailed in Table S5. Following this initial step, the Seurat object underwent conversion into the Scanpy format (version 1.6.0) through the use of Seurat Disk (version 0.0.0.9019) [67]. Subsequently, all further analyses were conducted within a Python environment (version 3.6.10). The BAM files, which had been sorted using cell barcodes and generated from the output of Cell Ranger, underwent further processing through the Velocyto pipeline, specifically version 0.17.17. This processing was aimed at annotating the reads, distinguishing between spliced and unspliced RNA sequences. This crucial step allows for a more detailed analysis of the transcriptional dynamics within the sampled cells by providing clear differentiation of the various RNA types of present [68]. RNA velocity vectors were computed using the scVelo Python package (version 0.2.2), which applies dynamical modeling to assess splicing kinetics [69]. The CellRank package (version 1.2.0) was utilized to merge RNA velocity data with transcriptomic similarity information into a unified kernel for the calculation. The estimator for generalized Perron cluster analysis was utilized to recognize macrostates [70]. By examining the coarse-grained transition matrix, terminal states were deduced and subsequently applied to calculate absorption probabilities.
In vivo macrophage depletion experiment
To verify the necessity of macrophages in A600-mediated bone regeneration, two experimental groups were established: A600 membrane implantation (A600) and A600 implantation with macrophage depletion (A600 + CLO). Macrophage depletion was achieved via intravenous injection of CLO. Control groups received equivalent PBS liposomes. The early inflammatory phase of bone defect repair (postoperative days 1–7) is the critical stage for macrophage polarization and function. Therefore, the depletion protocol must cover this window and extend into the early repair phase. The first dose of 2 ml is administered one day before surgery to establish the bone defect and implant the A600 membrane. The purpose is to deplete pre-existing tissue-resident macrophages in preparation for observing the early inflammatory response post-implantation. On the day of surgery and for the following 4 consecutive days postoperatively, a dose of 2 ml is administered daily. This period is crucial for massive monocyte recruitment, and sustained depletion effectively blocks the replenishment of new macrophages. Consolidation doses of 1 ml each are administered on postoperative days 7, 10, and 14 to address potentially sustained low-level monocyte infiltration. Systematic administration is stopped on postoperative day 14 because the 4-week observation endpoint (postoperative day 28) primarily reflects mid-to-late stage repair (soft callus formation, bone remodeling), by which time the direct effects of macrophages are diminished. Administering for the entire 4 weeks may compromise animal health due to excessive immunosuppression or introduce non-specific effects. Additionally, on postoperative day 7 (the peak of macrophage infiltration), extra animals (n = 3) from each experimental group are sacrificed. The tissue surrounding the bone defect (including the implant interface) is harvested to validate macrophage depletion efficiency via flow cytometry. Bone regeneration was evaluated at 4 weeks post-surgery.
Cell cultures
The RAW264.7 cell line is a well-known model derived from murine macrophages, commonly utilized in various research studies. This cell line was obtained from the Typical Culture Preservation Commission Cell Bank (Shanghai, China). The process of culturing the cells was conducted using Dulbecco’s Modified Eagle’s Medium, a widely utilized growth medium that provides essential nutrients and supports cellular metabolism. This medium was enriched with 100 U/mL of penicillin/streptomycin, which serves as an antibiotic agent to prevent bacterial contamination during cell culture. Additionally, the medium was supplemented with 10% fetal bovine serum, a source of growth factors, hormones, and other nutrients vital for promoting optimal cell growth and proliferation. The reagents mentioned above were acquired from Sigma-Aldrich. The culture environment’s temperature was set to 37 °C, maintaining a relative humidity of 95% and a CO2 level of 5%. To ensure proper passaging, the cell density was kept between 80% and 90%.
The rats were obtained from Kunming Medical University for this study. Following their arrival, the animals were euthanized in accordance with ethical guidelines, and their biospecimens were collected for further analysis. BMSCs were then isolated from the harvested samples and cultured using the whole bone marrow adherence technique. This process was performed in an α-MEM culture medium that was enriched with 10% fetal bovine serum and supplemented with 1% penicillin-streptomycin to promote cell growth and prevent bacterial contamination. The cultural environment for the experiments was maintained at an optimal temperature of 37 °C, along with a relative humidity of 95% and a carbon dioxide concentration of 5%. To ensure the health and viability of the cells, the growth medium was refreshed every two days. At the time of passaging, the cell density was consistently observed to be between 80% and 90%. For the various experimental procedures involving the cells, the passage numbers utilized ranged from P3 to P5 [71].
In vitro cytotoxicity study
The PLLA electrospun membranes was cut to correspond to the size of the bottom of the petri dish plate and then laid on the bottom of the petri dish plate. Next, the serum-free medium was placed in the incubator and cultured for 24 h as a pretreatment. Subsequent cell inoculation on the membrane is equivalent to inoculation in a petri dish. RAW264.7 cells were initially seeded into 96-well plates at a density of 2 × 10^4^ cells for each well. Following a one-day incubation, the cells were then cocultured with membranes across various sets. Equal concentrations of sulfosalicylic acid solutions were also added to the cocultures. The cells were then evaluated using a Cell Counting Kit-8 assay, which was provided by Thermo Fisher. To quantify the results, a microplate reader, specifically the TECAN Infinite 200 PRO, was utilized to measure the absorbance of the samples at a wavelength of 450 nm. To determine the relative cell viability (%), the equation provided in the Supporting Information was employed.
The evaluation of live/dead staining was conducted in 24-well plates, with each well containing a cell density of 1 × 10^5^ cells. Following a three-day culture period, an acetoxymethyl ester/propidium iodide solution (2/4 µM) reagent (Solarbio) was applied to the cells for 40 min to facilitate coloring. To analyze cellular morphology, the progression began by placing the cells onto 12-well plates with 1 × 10^5^ cells per well and cultivating them alongside diverse sets of electrospun membranes. Afterward, the cells were treated with 4% paraformaldehyde for a duration of 30 min to immobilize them, and subsequently, they were stained using phalloidin Fluor 488 conjugate, adhering to the guidelines provided. The obtained outcomes were observed with the utilization of the Olympus IX73 fluorescence microscope.
The cells were replaced with BMSCs and tested again using the same experimental method to verify the effect on cytotoxicity.
Cell adhesion
RAW264.7 cells were subjected to a labeling procedure using 5 µM Celltracker Green 5-chloromethylfluorescein diacetate, which is a product from Thermo Fisher. The labeling process was carried out for a total duration of 15 min to ensure the proper attachment and identification of the targeted cells. Following this initial procedure, 1 ml of culture medium, which contained a concentration of 5 × 10^4^ cells, was introduced onto each of the four groups of membranes that had been strategically positioned within 24-well culture plates. Following incubation times of 3, 6, and 12 h, the RAW264.7 cells were, fixed with 4% paraformaldehyde and examined under a fluorescence microscope. The quantitative assessment of the cells present on the membranes was conducted through a manual counting method. This counting was performed within five randomly selected regions of interest, with the analysis being carried out at a magnification of 40×. The cells were replaced with BMSCs and tested again using the same experimental method to verify the effect on cytotoxicity.
Anti-inflammation and macrophages polarization research in vitro
The morphology of cells cultivated on the electrospun membrane was examined in this study. Each scaffold was populated with 10,000 RAW264.7 cells in 24-well plates, allowing for detailed observation of cellular behavior in response to the specific conditions provided by the scaffolding material. The RAW264.7 cells were cultured for 1 and 3 days. To fix the cells, buffered saline with 2.5% glutaraldehyde was added. Then, the samples underwent dehydration using increasing ethanol concentrations of 75%, 85%, 95%, and finally 100%. Finally, the morphology of RAW264.7 cells that were seeded on various scaffolds was examined using a SEM. This analysis was performed at a magnification of 10 kV and was conducted by Thermo Fisher.
FC was utilized to investigate RAW264.7 cells, which were seeded at a density of 4 × 10^5^ cells per well. To assess different macrophage phenotypes, the M2 marker CD206 (#141714; BioLegend, CA, USA) was identified with an FC device (BD Biosciences, CA, USA) on days 1 and 3. The Alexa 647 CD206 of the mouse was incubated individually according to the instructions provided by the manufacturer. The data analysis adopted a step-by-step gate strategy. Firstly, cell debris is excluded by using the area parameters of forward scattering and lateral scattering, and adhered cells are distinguished and excluded by the height and width parameters of forward scattering to obtain a single cell population. Subsequently, the living cells were identified and delineated using fluorescent viability dyes. Within the living cell phylum, target lymphocyte subsets are further identified based on specific surface markers. Ultimately, the expression levels of function-related molecules were analyzed within the target subpopulation. The analysis of the data was carried out with FlowJo v10.8.1 software.
To further assess the expression levels of inducible nitric oxide synthase and the CD206 cluster of differentiation, IF staining was carried out utilizing specific antibodies, including CD206 (#ab64693; Abcam) and the pan marker CD68 (#ab283654; Abcam), along with 4,6-diamidino-2-phenylindole (#C1002; Beyotime Biotechnology) for highlighting nuclear structures. To analyze the macrophages, a 4% paraformaldehyde solution was used to immobilize them, followed by treatment with 0.25% Triton X-100 (Sigma-Aldrich) for permeabilization of the cells, and then blocked with 2% bovine serum albumin. The subsequent step involved the addition of primary and secondary antibodies. In order to visualize and seize images, laser scanning confocal microscopy (Olympus, Tokyo, Japan) was employed.
Osteogenic differentiation BMSCs cultured with conditioned medium in vitro
The collection and preparation of conditioned medium were conducted as follows (Fig. 6A). RAW264.7 were grown in a specialized growth medium that was enhanced with PLLA electrospun membranes (A600) for a duration of three days. After this incubation period, the culture supernatants were harvested and subsequently diluted with α-MEM that was supplemented with 10% fetal bovine serum in a 1:1 ratio prior to utilization.
ALP Analysis. BMSCs were placed in a cell culture dish at a density of 1.5 × 10^5^ cells per dish. The various conditioned media were renewed every three days. The ALP staining procedure was conducted on the 7th day utilizing a BCIP/NBT color development kit (Leagene, Beijing, China), and photographic images were obtained using a stereomicroscope (Olympus). To quantify the ALP staining, the absorbance was measured at a wavelength of 560 nm.
ARS Staining. BMSCs were maintained in various conditioned media for durations of 14 and 21 days within the culture dish. The cells underwent preservation at room temperature for 20 min utilizing a 4% paraformaldehyde solution. Following three washes with phosphate buffer saline, the cells were incubated for ten minutes in a 1% (w/v) solution of ARS at pH 4.2 (Haixing, Suzhou, China). The stained samples were examined using an Olympus IX73 inverted fluorescence microscope. The orange-red color of alizarin red was quantified through extraction by introducing a 70% ethanol solution with 10 mM HCl to the stained Petri dishes, followed by an incubation period of 30 min. The absorbance of the extract was recorded at 450 nm.
Macrophage conditioned medium collection and osteogenic rescue assay
To test the sufficiency of the M2-polarized microenvironment, macrophage-conditioned medium was prepared. Briefly, RAW264.7 cells were co-cultured with R600, A600, A1200, or R1200 PLLA membranes for 48 h. Serum-free conditioned medium was collected, centrifuged, and filtered before being used to stimulate BMSCs. BMSCs were divided into four groups treated with the respective conditioned medium and cultured in osteogenic induction medium for 14 days (ALP assay) and 21 days (Alizarin Red S staining) to assess osteogenic differentiation.
Statistical analysis
Statistical analyses were performed using GraphPad Prism software (version 9.0). Data are presented as the mean ± standard deviation. A one-way analysis of variance (ANOVA) was first conducted. If the ANOVA indicated a statistically significant difference, Tukey’s honestly significant difference post-hoc test was subsequently applied for all pairwise comparisons between groups. The Tukey’s test inherently controls for the family-wise error rate due to multiple comparisons. The threshold for statistical significance was set at p < 0.05 Fig. 7.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Wright ZM, Pandit AM, Karpinsky MM, Holt BD, Zovinka EP, Sydlik SA, Bioactive. Ion-Releasing PMMA bone cement filled with functional graphenic materials. Adv Healthc Mater. 2021;10(2):e 2001189. 10.1002/adhm.202001189.10.1002/adhm.20200118933326158 · doi ↗ · pubmed ↗
- 2Baryawno N, Przybylski D, Kowalczyk MS, Kfoury Y, Severe N, Gustafsson K, Kokkaliaris KD, Mercier F, Tabaka M, Hofree M, Dionne D, Papazian A, Lee DJ, Ashenberg O, Subramanian A, Vaishnav ED, Rozenblatt-Rosen O, Regev. A. David T Scadden A Cellular Taxonomy of the Bone Marrow Stroma in Homeostasis and Leukemia. Cell. 2019;177:1915–32.10.1016/j.cell.2019.04.040PMC 657056231130381 · doi ↗ · pubmed ↗
- 3Oetjen KA, Lindblad KE, Goswami M, Gui G, Dagur PK, Lai C, Dillon LW, Mc Coy JP, Hourigan CS. Human bone marrow assessment by Single-Cell RNA Sequencing, mass cytometry, and flow cytometry. J Clin Invest Insight. 2018;3(23):e 124928. 10.1172/jci.insight.124928.10.1172/jci.insight.124928 PMC 632801830518681 · doi ↗ · pubmed ↗
- 4Tang D, Han B, He C, Xu Y, Liu Z, Wang W, Huang Z, Xiao Z, He F. Electrospun Poly-l‐Lactic acid membranes promote M 2 macrophage polarization by regulating the PCK 2/AMPK/m TOR signaling pathway. Adv Healthc Mater. 2024;13(22):e 2400481. 10.1002/adhm.202400481.10.1002/adhm.20240048138650356 · doi ↗ · pubmed ↗
- 5Cheng L, Chen Z, Cai Z, Zhao J, Lu M, Liang J, Wang F, Qi J, Cui W, Deng L. Bioinspired functional black phosphorus electrospun fibers achieving recruitment and biomineralization for staged bone regeneration. Small. 2020;16(50):e 2005433. 10.1002/smll.202005433.10.1002/smll.20200543333230977 · doi ↗ · pubmed ↗
- 6Yang C, Ma H, Wang Z, Younis MR, Liu C, Wu C, Luo Y, Huang P. 3D printed wesselsite nanosheets functionalized scaffold facilitates NIR-II photothermal therapy and vascularized bone regeneration. Adv Sci. 2021;8(20):e 2100894. 10.1002/advs.202100894.10.1002/advs.202100894 PMC 852944434396718 · doi ↗ · pubmed ↗
- 7Zheng Z, Chen Y, Wu D, Wang J, Lv M, Wang X, Zhang ZY. Development of an Accurate and Proactive Immunomodulatory Strategy to Improve Bone Substitute Material-Mediated Osteogenesis and Angiogenesis. Theranostics. 2018;8(19):5482–500.10.7150/thno.28315 PMC 627609130555559 · doi ↗ · pubmed ↗
- 8Bressan E, Ferroni L, Gardin C, Pinton P, Stellini E, Botticelli D, Sivolella S, Zavan B. Donor Age-Related biological properties of human dental pulp stem cells change in nanostructured scaffolds. P Lo S ONE. 2012;7(11):e 49146. 10.1371/journal.pone.0049146.10.1371/journal.pone.0049146 PMC 350912623209565 · doi ↗ · pubmed ↗
