Introduction

Critical-sized bone defects present significant clinical challenges, necessitating innovative solutions for effective bone regeneration1,2,3. While autografts and allografts are commonly used, they have limitations and potential complications4. This issue has aroused interest in artificial bone grafts due to their unlimited supply and the ability to customize their physical and biological properties5,6,7. Artificial grafts not only offer viable options for bone replacement but also serve as valuable tools for studying osteogenesis and bone formation, including bone models and scaffold-on-a-chip systems8,9.

Additive manufacturing (3D printing) becomes an attractive choice for bone graft fabrication as it effortlessly creates the porous structures necessary for nutrient exchange during cell proliferation and bone formation10,11. Among a wide range of material options, polylactic acid (PLA) is widely recognized for its versatility, durability, functionalization capabilities, biocompatibility, and biodegradability12,13. However, its relatively low toughness has led to efforts aimed at enhancing its ductility. A promising approach involves the rational design of polymer blends to improve properties of PLA14,15,16. In this context, hydroxyapatite (HA) has emerged as a valuable additive, particularly for reinforceing PLA mechanical strength. Incorporating HA into the scaffold introduces two distinct phases that mimic the composition of natural bone. The inorganic (ceramic) phase provides structural support and toughness, while the organic (polymeric) phase enhances integration with surrounding tissues and the vascular system13,16,17.

Despite these advancements, the development of PLA-HA biocomposite materials for promoting cell growth is still in its infancy and the correlation between material surface properties and cellular behavior remains largely uncharted18,19,20. For instance, HA has been reported to enhance cell attachment21,22, but conflicting evidence suggests it may reduce cell proliferation and increase apoptosis23,24. This discrepancy highlights the incomplete understanding of the material properties that are most conducive to cell growth within the scaffold's microenvironment. Moreover, while nanoscale material properties could provide more detailed insights, they are often overlooked due to their complexity and undetectability at the macroscale21,25,26. To comprehend cellular responses to their surrounding environment, pinpointing the material properties at the surface of a sample, cellular level is crucial27,28. Atomic force microscopy (AFM) has emerged as a powerful tool for nanoscale investigations29,30,31. AFM employs a nanometer-sized tip on a cantilever to scan the sample’s surface, measuring parameters like stiffness, adhesion, and phase difference. This technique provides high-spatial-resolution data, allowing precise analysis of both surface morphology and mechanical properties simultaneously32,33,34.

In this study, we investigated the surface nanomechanical properties of 3D-printed PLA-based scaffolds functionalized with hydroxyapatite (HA). We examined scaffolds with two levels of HA addition—10% (P10) and 20% (P20)—as well as a surface-treated sample with HA coating on PLA (PC). AFM-based nanoscale quantitative mapping revealed that the P10 scaffold surface homogeneity in terms of stiffness and adhesion.These unique properties of the P10 scaffold significantly enhanced cell attachment and osteogenic differentiation of human adipose-derived mesenchymal stem cells (hADMSCs). This nanomechanical data-driven approach demonstrates potential for engineering scaffolds that effectively guide hADMSCs towards osteoblasts differentiation, thereby facilitating bone tissue regeneration. The study’s findings pave the way for innovative scaffold designs with the potential to revolutionize artificial bone grafts.

Material and methods

Preparation of PLA and PLA-composite scaffolds

Hydroxyapatite (HA, Sigma-Aldrich, USA) and Poly(lactic acid) (PLA, grade 3051D, Natureworks, USA) pellets were dried at 60 °C for 4 h before mixing. The PLA-HA composites with varying compositions of HA (0, 10, 20%) were blended for 6 min in a laboratory internal mixer (Brabender®, Germany). The relative melting points of the blend were set at 180, 185, and 190 °C. Subsequently, the composites were ground into pellets using a high-speed pelletizer and extruded into filaments (diameter = 1.75 mm) at 160 °C and 40 RPM using a filament maker (Filabot, USA). Fused deposition modeling (FDM) was used to fabricate cylindrical scaffolds with gyroid geometry (D = 10 mm, h = 5 mm) on a Sermoon V1 3D printer (Creality 3D, China). The layer height was set to 0.2 mm. The scaffolds were labeled according to the HA content as PLA, P10, and P20, corresponding to 0, 10, and 20% HA, respectively.

A 1:1 ethanol and 0.25 M NaOH aqueous solution was prepared for alkali treatment and HA coating. The scaffolds were submerged into the solution and subjected to magnetic stirring at 550 RPM for 1 h. Subsequently, the scaffolds were subjected to ultrasonic cleaning with 100% ethanol for 3 min. The 3D-printed PLA scaffolds were coated with the 3% (w/v) HA in 100% ethanol under magnetic stirring at 550 RPM for 2 h. The coated scaffolds were washed multiple times using deionized water until the pH of the solution reached 7.0. The scaffolds were further soaked in pure ethanol and air-dried at room temperature for additional characterization and cell culture experiments. These scaffolds were identified as PC.

Scanning electron microscopy (SEM) and energy dispersive X-ray spectrometry (EDS) analysis

For scaffold with cells, the scaffolds were immersed in 2.5% glutaraldehyde (Loba Chemie, India) and fixed overnight at 4 °C. The scaffolds underwent a 45 min dehydration process using a graded series of ethanol concentrations (30, 50, 70, 95, and 100%). Scaffolds without cells were prepared directly for coating. Then, the scaffolds were sputter-coated with gold using a sputter coater (BALZERS SCD 040, Balzers, Lichtenstein), and analyzed using a scanning electron microscope (JSM-IT300LV, Jeol, Japan) with an accelerating voltage of 10 kV. Furthermore, the chemical composition of the constructs was characterized using energy dispersive spectroscopy (EDS) with the assistance of computer-controlled software (Aztec EDS system, Oxford Instruments, Abingdon, UK).

AFM-based nanomechanical characterization

Nanomechanical properties were assessed using AFM. The scaffold was mounted on a glass slide with double-sided tape and examined on a JPK NanoWizard 3 AFM (Bruker, USA) in quantitative imaging (QI) mode, where the surface morphology and the nanomechanical data were retrieved simultaneously35. The cantilever was a non-contact probe with a spring constant of 37 N/m and a resonance frequency of 300 kHz (ACTA model, AppNano, USA). Force calibration with a thermal tune preceded data collection. Measurements covered ten randomly selected 20 µm × 20 µm surface areas (two areas per scaffold from five scaffolds) at a resolution of 256 × 256 pixels and a force threshold of 100 nN. Stiffness and adhesive force data were analyzed by plotting distribution histograms, with Gaussian or bimodal Gaussian functions used for fitting depending on the distribution. Peak center and standard deviation values were extracted from the fitting, and distribution plots were generated using the Matplotlib library in Python36.

Cultivation of human adipose-derived mesenchymal stem cells (hADMSCs)

Human adipose-derived mesenchymal stem cells (hADMSCs) at the fifth passage obtained from PromoCell (Germany) were used in this study. Briefly, 5 × 105 hADMSCs were seeded onto 3D-printed scaffolds, including PLA, P10, P20, and PC. The cells were cultured in an hMSC growth medium (alpha-MEM supplemented with 4% (v/v) FBS, 1% (v/v) antibiotic–antimycotic, 1% (v/v) HEPES buffer, 1% (v/v) GlutaMAX, and 1 ng/ml bFGF) for seven days and then shifted to an osteogenic medium (DMEM with 4% (v/v) FBS, 1% (v/v) antibiotic–antimycotic, 1% (v/v) HEPES buffer, 10 mM β-glycerolphosphate, 50 μM ascorbic acid, and 10 μM dexamethasone) for 21 days, with 3–4 days medium changes. Cell viability was quantified using PrestoBlue™ (Thermo Fisher Scientific, USA), and live/dead cells were visualized using the live/dead Cell Double Staining kit (Sigma-Aldrich, USA).

Cell viability assay

PrestoBlueTM (Thermo Fisher Scientific, USA) measured cell viability under the manufacturer's instructions. Briefly, the cell-seeded scaffolds were incubated in 1 ml of medium containing 10% (v/v) PrestoBlue™ in hMSC growth medium at 37 °C for an hour in the dark. The fluorescence intensity was measured using a microplate reader at λex 560 and λem 590 nm (Varioskan Flash Multimode, Thermo Scientific, USA).

For live/dead cell viability assay, the cells/scaffold constructs were incubated in live/dead Cell Double Staining kit (Sigma-Aldrich, USA) for 30 min in the dark. The constructs were observed and analyzed using an inverted fluorescence microscope (Nikon Eclipse Ti, USA) to determine live cells (green) and dead cells (red) after 1, 2, and 4 weeks of culture.

Assessment of calcium deposition by alizarin red staining

Alizarin Red S staining, the scaffolds were incubated in 2% (w/v) Alizarin Red S (pH 4.1–4.3) for 1 h, and repeatedly washed with DI water. The amount of Alizarin Red S bound to calcium was quantified by eluting the stain with a 10% (w/v) hexadecyltrimethylammonium chloride solution. The absorbance of the eluted solution was measured at 557 nm with a spectrophotometer (Multiskan Go, Thermo Fisher Scientific, USA).

Statistical analysis

Statistical analysis, employing analysis of variance (ANOVA) in GraphPad Prism version 9.0, revealed group differences. The data are presented as mean ± standard deviation (S.D.).

Results

3D-printed PLA and PLA-composite scaffolds

3D printing has gained significant popularity as a method for the design of customized scaffolds. Fused Deposition Modeling (FDM) has emerged as one of the most practical choices among various 3D printing techniques due to its versatility, environmental friendliness, and simplicity, making it ideal for printing a wide range of plastic materials37.

The printed scaffold surface was analyzed using SEM (Fig. 1). High-magnification images provided a close look at the surface morphology of the scaffolds. The surface of the PLA group appeared smooth, whereas the P10 and P20 groups exhibited a rough surface due to HA particles mixed with PLA13. Additionally, scratch marks and micropores were observed on the surface of the PC scaffold following alkali treatment and HA coating. It is possible that carbon chain cleavage resulted in the release of carbon radicals, which could bind with hydroxyl groups from the alkali solution, thus creating an ideal surface for the absorption of HA particles38. The Ca distribution images, as revealed by EDS mapping, indicate the presence of Ca in the P10, P20, and PC scaffolds. While P10 and P20 show localized variations in calcium density, PC demonstrates a more uniform calcium distribution across its surface. The calcium content in P10, P20, and PC was measured at 6.76% ± 0.12, 12.37% ± 0.12, and 3.76% ± 0.08, respectively.

Fig. 1
figure 1

SEM micrographs of the 3D-printed PLA, P10, P20, and PC scaffolds at magnification of 200x, 1000 × and 5000x. EDS mapping images illustrate the distribution of calcium (Ca) on the scaffold surfaces. Scale bars represent 200 µm at 200x, 50 µm at 1,000x, and 5 µm at 5000x.

Full size image

AFM-based nanomechanical mapping

The hADMSCs are responsive to the microenvironments in which they reside39,40,41. It has been shown that even nanoscale alteration can impact cell viability and fate42,43. Therefore, we investigated the surface properties of the scaffolds at the nanometer resolution. The AFM-based quantitative imaging (QI) mode simultaneously collected the height, stiffness, and adhesive force maps. The exemplary data maps for the PLA and PLA/HA composite scaffolds (P10, P20, and PC) were displayed (Fig. 2). The data showed that the PLA and PC scaffolds had a relatively smooth surface. On the other hand, the P10 and P20 scaffolds possessed sharp and rough surfaces. The rough appearance observed from the AFM measurement agreed well with the SEM micrographs (Fig. 1 1000x–5000x). The QI mode allowed simultaneous extraction of the height, stiffness, and adhesive force of the interfaces. The QI mapping data were collected for the PLA, P10, P20, and PC scaffolds. For the PLA scaffolds, the height of the PLA surface fluctuated within a 400-nm range (Fig. 2a). The stiffness, the slope of the approaching force-distance curve, of the PLA surface was about 0.5–1.5 N/m (Fig. 2b). Lastly, the adhesive force, which is the attraction between the AFM tip and the PLA surface, varied within the 40-nN range (Fig. 2c). By comparing three QI maps of the PLA surface, it appeared that the rough regions (Fig. 2a) have high stiffness (Fig. 2b) and high adhesive force (Fig. 2c). The result agrees with previous studies that indicate that the stiffness depends on surface roughness44.

Fig. 2
figure 2

Exemplary AFM micrographs of (a, d, g, j) topography or height, (b, e, h, k) stiffness, and (c, f, i, l) adhesive force of the (a-c) PLA, (d-f) P10, (g-i) P20, and (j-l) PC scaffold surfaces. The scale bars (5 µm) on the right bottom corner showed the lateral (XY) dimensions. The color gradient bars on the right showed the values of height, stiffness, and adhesive force.

Full size image

Stiffness and adhesion distributions

We analyzed the height, stiffness, and adhesive force distributions across different scaffold types. Height, stiffness, and adhesive forces were extracted and plotted as distributions, revealing distinct patterns (Fig. 3). PLA and PC scaffolds displayed bimodal Gaussian distributions for height, stiffness, and adhesive forces. P20 scaffolds displayed a Gaussian distribution of height but bimodal distributions of stiffness and adhesive forces. In contrast, the P10 scaffold exhibited Gaussian distributions of all three parameters, indicating surface homogeneity. For height (Fig. 3a–d), the PLA scaffolds showed an average height of 1.01 µm, with a small part having an average height of 1.24 µm. The P20 scaffolds had an average of 1.92 µm, with a small portion having an average of 2.42 µm. The surfaces of the P10 and PC scaffolds exhibited average heights of 6.56 and 0.26 µm, respectively. For stiffness (Fig. 3e–h), the PLA scaffolds predominantly showed an average stiffness of 1.20 N/m, with a small portion having an average stiffness of 2.56 N/m. In contrast, the P20 scaffolds featured a majority with an average stiffness of 4.54 N/m, and a small portion had an average stiffness of 1.65 N/m. The PC scaffolds had a dominant stiffness of 1.75 N/m, with a small portion having an average stiffness of 3.71 N/m. Notably, the P10 scaffolds displayed the most uniform stiffness, averaging 1.69 N/m. These variations in stiffness are attributed to the presence of hydroxyapatite (HA) particles in P10 (1.69 N/m), P20 (4.54 N/m), and PC (1.75 N/m) scaffolds, resulting in higher stiffness compared to PLA (1.20 N/m) presumably due to the addition of the HA particles45. The surfaces of the PC scaffolds also had higher stiffness than the PLA scaffolds, possibly due to an HA layer. Some studies also suggest that the HA presence contributes to the reinforcement of the scaffold, which may be associated with the increasing stiffness in our case26.

Fig. 3
figure 3

Height (Z), stiffness (k), and adhesive force (Fadh) distributions of the surfaces of the (a, e, i) PLA, (b, f, j) P10, (c, g, k) P20, and (d, h, l) PC scaffolds. The centers and the standard deviation (S.D.) extracted from the distribution fitting were displayed as center ± S.D.

Full size image

Similarly, adhesive force distributions showed bimodal Gaussian patterns in PLA, P20, and PC scaffolds, while P10 exhibited a Gaussian distribution (Fig. 3i–l). Most PLA scaffolds displayed an average adhesive force of 57.3 nN, with a small portion having an average adhesive force of 23.9 nN. For P20, the dominant adhesive force was 166.3 nN, while a small portion had an average adhesive force of 114.7 nN. PC scaffolds mainly exhibited an average adhesive force of 38.3 nN, with a small portion having an average adhesive force of 26.2 nN. Lastly, the P10 scaffolds demonstrated uniform adhesive force distribution, averaging 58.8 nN. We found that both P10 and P20 scaffolds exhibited high adhesive forces. Our result suggests that HA incorporation enhances surface stickiness, agreeing with other studies that adding reinforcing filler can increase adhesion interfaces and, thus, create adhesive force between the surface and the AFM tip46. However, despite having HA as a top layer, the PC scaffolds showed a lower average adhesion force than PLA. The result implies that the HA corporation method with PLA can affect the nanomechanical properties of the scaffold. The presence of the HA layer in the PC scaffolds did not increase the surface adhesion, but mixing the HA as reinforcing filler allows the co-existence of two phases—HA and PLA, which, in turn, increase the surface adhesion. These nanomechanical characterizations provide critical insights into scaffold surface properties. Next, we explore how these properties influence the behavior of hADMSCs, which have clinical relevance and are instrumental in in vitro bone formation studies.

Cell attachment and metabolic activity

AFM results indicated that HA addition or coating altered the nanoscale surface properties of PLA scaffolds. To understand their impact on cellular behavior, hADMSCs were seeded and differentiated into osteoblasts, serving as a proof of concept of bone tissue engineering. Fluorescence of PrestoBlue™ assay results (Fig. 4a) reflected cellular metabolic activity. After three days, hADMSCs showed robust attachment to all scaffolds, as indicated by strong fluorescence signals. However, by day seven, the PC scaffold group exhibited significantly lower metabolic activity compared to the other groups. These findings suggest that PLA, P10, and P20 scaffolds support cell attachment and viability. SEM analysis (Fig. 4b) after seven days revealed spread-out cell morphologies on PLA, P10, and P20 scaffolds, indicating favorable cell adhesion and proliferation. In contrast, cells appeared rounded on the PC scaffold, correlating with reduced viability. These observations suggest that the surface properties of scaffolds play a crucial role in influencing cell adhesion and viability. PLA/HA composite materials have shown promise for bone scaffolds with osteogenic and osteoconductive potential12,40. Based on cell attachment and viability results, we selected P10 and P20 scaffolds to investigate long-term hADMSCs osteogenic differentiation. Both scaffolds exhibited sustained cell viability and even distribution on the scaffold surface for up to 4 weeks (Fig. 4c).

Fig. 4
figure 4

The hADMSCs (a) viability cultured on the PLA, P10, P20, and PC scaffolds for three days and seven days. All data are expressed as mean ± SD (n = 3) and *indicates a statistically significant difference with p-value < 0.05, (b) morphology by SEM micrographs cultured in the PLA, P10, P20, and PC scaffolds for seven days at different magnifications. The scale bars are 200 µm for 200x, 50 µm for 1000x, and 10 µm for 5000x, (c) images of live/dead staining of P10 and P20 scaffolds for 1, 2, and 4 weeks (green: live cells, red: dead cells). The white scale bars are 100 µm.

Full size image

The mineralization and calcium deposition in the cell/scaffold constructs were visualized using Alizarin Red S staining (Fig. 5a,b), which has a high affinity for calcium ions. This staining method is commonly used to assess calcium deposition during osteogenic differentiation. The non-cell seeded scaffolds were initially stained with Alizarin Red S, and the resulting staining intensity reflected the baseline levels of hydroxyapatite incorporated in the scaffolds. Alizarin Red staining of the P10 scaffold significantly increased after 2 weeks of osteogenic differentiation, and the staining reached its peak at week 4. The staining intensity of the P20 scaffold gradually increased until week 4. We found that osteogenic differentiation at 4 weeks in the P10 group was significantly higher than in the P20 group. This is related to the quantitative analysis of the cell viability test at 4 weeks, where P10 showed higher cell viability than P20, as shown in (Figure S1). Together, the results suggest that the P10 scaffold appeared to be the most suitable for hADMSCs culture and osteoblast differentiation.

Fig. 5
figure 5

Evaluation of osteogenic differentiation of hADMSCs in P10 and P20 scaffolds. (a) Alizarin Red S staining detected calcium deposition in scaffolds at 2 and 4 weeks. A non-seeded scaffold was a control group without cells. The white scale bars are 1 cm. (b) The corresponding quantitative analysis of Alizarin red S staining for calcium deposition. The data are shown as mean ± SD. ** and *** indicate a statistically significant difference with p-value < 0.01 and 0.0001.

Full size image

Discussion

The surface properties where cells reside can influence cell functions, including attachment, proliferation, and differentiation47,48,49,50. Eukaryotic cells respond to their surroundings through mechanosensing signals that trigger metabolic pathways regulating cell behavior and fate51,52. For example, cells tend to become neurons on soft surfaces and transform into muscle and bone cells at intermediate and high stiffness ranges, respectively50. To investigate the effect of surface properties on cell behavior, we analyzed the nanomechanical properties of 3D-printed scaffold surfaces. Nanomechanical mapping revealed that the PLA, P20, and PC scaffolds displayed bimodal distributions of stiffness and adhesive forces, while the P10 scaffold exhibited homogeneous surface properties, characterized by a Gaussian distribution of both stiffness and adhesion. Previous studies showed that the bimodal distributions normally reflect two composition phases26,34,44. In this study, the presence of a sufficient amount of HA and PLA in P20 and PC scaffolds seems to give rise to the bimodal stiffness and adhesion distributions, agreeing with the previous work. The bimodal distributions of PLA scaffold may be caused by printing traits, which may occur for pure PLA. Since HA has a much higher modulus than PLA, adding 10% of HA as a reinforcing filler may help obscure such traits, resulting in a Gaussian distribution of the stiffness and adhesion in the P10 scaffold. Adding more HA (i.e., 20%) could cause HA agglomeration within the PLA matrix, thus displaying the two-phase behavior and the bimodal distributions46. Lastly, the coating of HA in the PC scaffold directly introduces a separated phase of the HA, resulting in bimodal distributions.

The scaffold surface properties, such as roughness, wettability, and surface energy, have been demonstrated to be crucial for cell attachment, growth, and differentiation53,54. Several studies showed that high surface roughness increased space for cells residing13 and thus promoted cell attachment during the initial stage of the cells55,56,57. Regarding MSC differentiation, Adipocytes, in contrast to osteoblasts, are highly sensitive to topographical features58. High surface wettability is also shown to control the adhesion of cellular proteins and modulate cell adhesion59.

In this study, hADMSCs exhibited better attachment to P10 and P20 scaffolds than PLA and PC scaffolds. Higher surface roughness of P10 and P20 scaffolds, providing more contact points for cell anchoring58,60. Mineralization analysis revealed higher calcium deposition on the P10 scaffold than on the P20 scaffold. Since calcium mineralization is linked to cell differentiation, the result indicates that hADMSCs can differentiate into bone cells on the P10 scaffold better than on the P20 scaffold. The hADMSCs cell differentiation result suggests that the P10 scaffold provides a more conducive mechanical environment for cellular function61,62.

While the scaffold properties are extensively studied, surface homogeneity at the nanoscale is rarely investigated and discussed. The single distributions of the stiffness and the adhesive force of the P10 scaffold indicate a high degree of surface homogeneity compared to the PLA and PLA-composite materials. Our result suggests that homogenous surface properties of the P10 scaffold promote cell attachment and differentiation. The finding agrees with the previous study showing that a uniform β-phase poly(vinylidene fluoride) (PVDF) surface promotes uniform distribution of the fibroblast cells on the surface63. A study by Beijer et al. also showed that non-uniform surface topography caused hMSC cells to create an adapted cellular state where the shrinkage of cell and nucleus, protein condensation, and mitochondria reduction can be observed60. Furthermore, Chun Yang and colleagues confirmed that hMSC osteogenesis is influenced by surface pattern regularity. Cells on regularly patterned scaffolds were more spread, larger in area, and exhibited darker staining with alkaline phosphatase (ALP), indicating higher expression of the osteogenic marker ALP, like hMSCs cultured on uniformly stiff gels. In contrast, minimal staining was observed in hMSCs cultured on randomly patterned and uniformly soft surfaces, suggesting significantly lower osteogenesis64. Therefore, it is evident that the hADMSCs cells prefer the P10 scaffold due to its surface homogeneity properties.

The P10 scaffold also has unique properties. Its stiffness (1.69 N/m) and adhesive force (58.8 nN) were relatively in the middle range of the stiffness (1.20–4.54 N/m) and adhesive force values (23.9–116.3 nN). Surface stiffness in this work is essentially the surface energy, which has been shown to promote cell retention and adhesion53. Adhesive force in this work is directly proportional to wettability and surface charge, which help with cell attachment. Our results imply that hADMSCs favor distinct stiffness (surface energy) and adhesion force (surface charge) for the best growth condition. Such specific surface characteristics may differ for different cell types. Studies showed that different parts of the neuron cells favor different surface stiffness65,66. Fibroblasts prefer soft over hard surfaces54. The surface energy can influence the interactions between cells and the extracellular matrix, broadly impacting cell adhesion responses. Our work emphasizes the need to understand cell-surface interaction and cellular response to microenvironments to improve the design and engineering of bone-replacement material synthesis67.

Furthermore, the preparation method may affect the material properties. In this case, we employed two methods of incorporating HA into PLA, blending and surface coating. Despite the PC scaffold having HA content on the surface, hADMSCs exhibited poor cell viability when cultured on the PC scaffold, suggesting that contact with HA from surface modification may have a toxic effect on hADMSCs68,69. This outcome may be due to HA particle agglomeration38 and residual alkali or other chemicals70,71. HA agglomeration creates stress points for cells, adversely affecting their morphology and potentially leading to cell death. Residual alkali from surface treatment can increase local pH, producing cell toxicity. These changes in surface chemistry may render the material less conducive to cell attachment and proliferation72. Therefore, it is crucial to thoroughly evaluate the preparation method of these scaffolds before considering their in vivo usage.

Conclusion

The present study investigated the nanomechanical mapping of HA-composite PLA scaffolds and the potential association with in vitro cellular responses. The P10 scaffold modified by adding 10% of HA showed the highest cell attachment, viability, and mineralization among the four scaffold types. The AFM-based nanomechanical mapping provided high-spatial-resolution interface data of the PLA-composite scaffolds. Notably, the P10 scaffold exhibited uniform distributions of stiffness and adhesive force within an intermediate range. The result suggests an association between optimal conditions for cell growth and homogeneity of the surface properties. The study highlights the importance of understanding the material properties at a nanoscale resolution to reveal distinct properties that may not be evident in low-resolution data. The surface homogeneity is critical for interactions between cells and the scaffold interface. It must be further investigated for its potential correlation with other osteogenesis properties, such as osteoinductivity, osteointegrity, and osteoconductivity16. The insight into these correlations used as a design guideline will allow us to synthesize, functionalize, and tailor the scaffolds with specific requirements to maximize cell growth and enhance scaffold in vivo performance. Furthermore, the high-spatial-resolution data from AFM opens up the potential for future application of machine-learning analysis approaches73,74.