Cytoskeletal regulation on polycaprolactone/graphene porous scaffolds for bone tissue engineering Authors: Hendrik Setia Budi, Silvia Anitasari, Yung-Kang Shen, Shuntaro Yamada Published: 2024 ABSTRACT AbstractUnderstanding cellular mechanics requires evaluating the mechanical and chemical cues that regulate the actin cytoskeleton, particularly filopodia and lamellipodia. Therefore, this study aims to investigate the effect of scaffolds properties on cell migration. The results showed that scaffolds toughness, strain, and strength played a key role in promoting cell movement by stimulating the dynamic formation of filopodia and lamellipodia. The test sample containing 3 wt% G significantly enhanced toughness, yield strength, and strain, leading to increase cell motility as well as enhanced development of longer filopodia and larger lamellipodia in MG-63 cells. These results provide valuable insights for optimizing scaffolds to promote bone tissue regeneration.Keywords: Cytoskeleton, Filopodia, Lamellipodia, Polycaprolactone, Graphene, ScaffoldsSubject terms: Biotechnology, Structural biology, Materials science Introduction The achievement of successful tissue repair and regeneration through tissue engineering approaches is largely dependent on the harmonious interplay among cells, scaffolding materials, and bioactive molecules. Among these crucial elements, optimizing cell motility on biomaterials is a fundamental and essential factor. Cell motility has been reported to be associated with several processes, including the extension of pseudopodia and the formation of new adhesions. In addition, cell migration is mechanically driven through actin-rich protrusions at the leading edge of the mobile cells, primarily by lamellipodia and filopodia1,2. Lamellipodia are broad, sheet-like structures that are filled with a network of branched actin filaments. These structures are primarily responsible for generating large pushing forces that help to propel cells forward during migration3. Within lamellipodia, there are finger-like structures filled with tightly packed bundles of actin filaments, which are widely known as microspikes. Upon extending beyond the boundary of the lamellipodium, these structures are denoted as filopodia. According to previous studies, filopodia play a crucial role in steering the cell toward its target. These structures can extend far beyond the cell edge and are associated with sensing the environment and interacting with distant targets4. Filopodia can be used by cells to probe their surroundings and to establish contact with other cells or extracellular matrix (ECM) components. The balance between filopodia and lamellipodia can vary greatly depending on the specific needs. For example, cells exhibiting a requirement for expeditious and proficient migration may manifest an augmented presence of lamellipodia. Meanwhile, those confronted with the necessity to traverse intricate environments may show an increased prevalence of filopodia, thereby enhancing their capacity to perceive and engage with the surroundings5,6. Although polycaprolactone (PCL) is broadly used for its biocompatibility and biodegradability in tissue engineering, the inherent hydrophobicity, lack of surface functional groups, and comparatively low surface energy hinder optimal cell adhesion and migration7. To improve cell-material interaction, functionalization with various compounds has been considered. Recently, graphene (G) has been incorporated as a filler to enhance the mechanical properties of PCL-based scaffolds for bone regeneration8,9. In addition to improving the strength of scaffolds, the incorporation of G into PCL at elevated concentrations improved biocompatibility as well as enhanced cell adhesion and proliferation8,10. The PCL/G combination has been investigated as a viable approach to enhance the mechanical characteristics of PCL scaffolds used in bone tissue engineering. Graphene has the ability to improve the mechanical strength of PCL composites, rendering the materials more appropriate for bone applications10. Several studies have also shown that it acts as a chemoattractant and promotes the formation of adhesion molecules within the cells, selectively extending filopodia toward the cue while decreasing filopodia and lamellipodia formation in other directions11,12. As proof of the concept, graphene was shown to increase pseudopodia extension when combined with scaffolds of metal, silicon, and hydrogels13. Although PLC/Graphene (PCL/G) scaffolds have been shown as a promising scaffolding material for bone tissue engineering, their mechanical and chemical cues on filopodia and lamellipodia of the loaded cells on 3D scaffolds remain a subject of ongoing investigation. The association between the presence of graphene and pore size has also not been comprehensively explored. Therefore, this study aims to investigate the mechanotransduction in cellular systems, specifically focusing on the impact of PCL/G scaffolds. Results Improved mechanical properties of PCL/G The presence of graphene (G) within PCL scaffolds had an impact on its mechanical properties. When the concentrations of graphene were below 1.5 wt%, the toughness of scaffolds increased in a concentration-dependent manner, with 1.5 wt% exhibiting the highest toughness at approximately 50 kJ/m³ Fig. 1a. However, when the concentration reached 2 wt%, there was a notable decrease in toughness from 50 kJ/m³ to approximately 20 kJ/m3, likely due to increased brittleness. However, the introduction of 3 wt% graphene effectively restored the toughness, surpassing that of PCL. Yield strength and strain also showed an increase in a manner that depended on the concentration Fig. 1b. Mechanical properties. (a) Graphic toughness and yield strength; (b) Graphic yield strength and yield strain (c) Stress-Strain Curves; n = 3. The incorporation of 3 wt% graphene led to a yield strength that was more than 5 times higher compared to PCL. Similarly, yield strains also exhibited a concentration-dependent increase, with 3 wt% graphene resulting in a 5–6 times higher value compared to the naïve PCL. Furthermore, the 3 wt% G scaffolds exhibited the highest stress value, followed in descending order by 1.5 wt%, 1 wt%, 2.5 wt%, 0.5 wt%, and 2 wt% G, with PCL having the lowest stress. For strain, pure PCL had the highest value, followed by 0.5 wt%, 1 wt%, 2.5 wt%, 2 wt%, and 3 wt% G, which showed the lowest strain in Fig. 1c. Based on these mechanical properties, PCL/G scaffolds with 3 wt% graphene showed great potential as a material that could simultaneously exhibit strength and toughness. Graphene concentration-dependent enhancement of cell migration A wound healing assay was performed using MG63 cells and L929 cell lines in the presence of the extract from scaffolds Fig. 2a. The assay was carried out to assess the impact of varying concentrations of graphene on cell migration. Generally, the presence of graphene promoted cell migration, and the effect was robust at 3 wt% on MG63 cells, which showed a tendency of mild wound closure after the scratching (showed as 0 h). When cells encountered a chemoattractant, particularly at a higher concentration of graphene, there were extended filopodia and lamellipodia, showing enhanced cell motility under the condition. The quantification of the average velocity showed that, regardless of the cell type, the cell migration rate in the first 24 h exhibited a clear correlation with the concentration of graphene in scaffolds (Table 1; Fig. 2b, c. (a) The scratch assay of PCL/G scaffolds were observed by an inverted microscope at three distinct time intervals: 0, 24, and 48 h; (b) MG-63 cells migration rate; (c) L-929 cells migration rate; n = 10; #p < 0.05 compared to PCL. Table 1Wound healing assay of MG-63 and L-929 cells for 24 and 48 h.ScaffoldMG-63L-929Wound Closure (%)Wound Closure (%)24 h48 h24 h48 hPCL36.7282.679.7581.160.5 wt% G43.590.250.4371.231 wt% G49.4610067.1686.881.5 wt% G57.1710074.041002 wt% G69.3510090.691002.5 wt% G81.3210096.931003 wt% G93.4110097.78100 Wound healing assay of MG-63 and L-929 cells for 24 and 48 h. The highest velocity was observed for 3 wt% G in both MG-63 and L-929 cells (p < 0.05), while the lowest speed of edge advancement was seen in the PCL for both cell types. The highest velocity was observed for 3 wt% graphene (G) in both MG-63 and L-929 cells (p < 0.05), while the smallest speed of edge advancement was seen in the PCL for both cell types. The average velocity of the advancement of the cell edge for 3 wt% graphene in MG-63 was 9.64 ± 1.3 μm/hour, while for L-929, it was 8.86 ± 1.38 μm/hour. However, the velocities for PCL in MG-63 and L-929 were 4.67 ± 0.5 μm/hour and 6.36 ± 0.27 μm/hour, respectively. Extended pseudopodia on PCL/G scaffolds The impact of graphene on cell motility within the porous scaffolds was assessed, and an examination of cell morphology was conducted using SEM imaging and subsequent quantification. The results showed that an increased concentration of graphene within scaffolds led to the development of longer filopodia in MG63 cells, Fig. 3a-d. On day 7, the average length of filopodia in the 3 wt% G was measured as 5.5 μm, while it was 5.8 μm in the PCL. The length of filopodia of osteoblast-like (MG-63) cells at various graphene concentrations in PCL. (a) Edge detection of MG-63. (b) Length of MG-63 filopodia; (c) Surface area of MG-63 lamellipodia; (d) Density of MG-63 filopodia; n = 10; #p < 0.05 compared to PCL. After a period of 14 days, there was a significant change in the length of filopodia compared to day 7. The average length of filopodia in the 3 wt% G was 6 μm, while it was 3 μm in PCL (p < 0.05). However, different conditions were observed in L-292 cells. On day 7, the average length of filopodia was 7.5 μm for 3 wt% G, while a value of 1.5 μm was obtained for PCL. On the 14th day, the length of filopodia in PCL increased significantly (p < 0.05) compared to 3wt% G Fig. 4a-d. Furthermore, there was a correlation observed between the length of filopodia, the surface area of lamellipodia, and the density of filopodia in both days 7 and 14. Lastly, the results showed that the incorporation of 3 wt% G in PCL switches had a more significant effect on cell length and motility in the favorable outcomes of MG 63 than in L929 (p < 0.05). Actin cytoskeleton of mouse fibroblast (L-929) cells at various graphene concentrations in PCL. (a) Edge detection of L-929. (b) Length of L-929 filopodia; (c) Surface area of L-929 lamellipodia; (d) Density of L-929 filopodia; n = 10; #p < 0.05 compared to PCL. Interplay between pseudopodia development, graphene concentration, and pore size Pore sizes were separated into 4 ranges namely, < 100, 101 ~ 250, 251 ~ 500, and > 501μm. The results showed that there were notable differences among PCL/G scaffolds with pore sizes < 100 and 101–250μm. Higher concentrations of graphene (2, 2.5, and 3 wt% G) exhibited more pores greater than > 100μm compared to lower concentrations (0.5, 1, and 1.5 wt% G) and PCL, as shown in Fig. 5b, c, and d. Additionally, the 0.5 wt% G had more pore of < 100μm compared to the 3 wt% G, as shown in Fig. 5a (p > 0.05). Figure 6a-g showed that the 3 wt% G had a pore size that was more favorable for nutrient transport and waste metabolism, with a higher degree of pore interconnectivity compared to PCL and 0.5 wt% G. The 1 wt% G showed less clarity in pore structure compared to 1.5 wt% and 2 wt%. The 2.5 wt% G showed improved pore formation and interconnectivity, which further increased at 3 wt% G. PCL and the 0.5 wt% G exhibited fewer pores and lower interconnectivity. Pore sizes of PCL/G scaffolds. (a) Pore size < 100 μm; (b) Pore sizes 101–350 μm; (c) Pore sizes 351–500 μm; (d) Pore sizes > 501 μm; n = 5. The morphology of PCL/G scaffolds, illustrating the shape and size of the scaffolds. This suggested that the PCL/G scaffolds had a high probability of supporting the growth of long and wide filopodia. Based on the density data of filopodia in MG-63 and L-929 cells, 3 wt% G had a high density (p < 0.05). This information was crucial for understanding cell adhesion and how filopodia tips attached to the substrate. Discussions The study on the mechanical properties on of PCL composites revealed notable insights, especially regarding the increase in brittleness and decrease in toughness at a 2 wt% G concentration. This behavior may result from inadequate exfoliation of graphene within the PCL matrix, leading to agglomeration that adversely affects toughness. Understanding these mechanisms is essential for optimizing the performance of graphene-reinforced PCL composites6. At 2 wt% G graphene, poor dispersion may cause agglomerates to form, disrupting the polymer matrix continuity and creating stress concentration points. Studies on epoxy composites demonstrate that even low graphene concentrations can significantly improve toughness with proper dispersion; for example, 1 wt% G resulted in substantial toughness enhancement14. This highlights that the effectiveness of graphene as a reinforcement is highly dependent on its dispersion within the polymer matrix. Without adequate exfoliation, the expected mechanical improvements may not materialize, resulting in increased brittleness6. Moreover, the mechanical properties of polymer composites are greatly influenced by the size and shape of the dispersed phase. Effective toughening relies on the intermolecular interactions and morphology of the dispersed phase within the matrix15. In PCL composites, if graphene is not uniformly distributed, the resulting microstructure may reduce toughness by hindering stress transfer between the graphene and polymer. This is especially critical at 2 wt%, where reinforcement and agglomeration must be carefully balanced16. Interestingly, toughness recovery at 3 wt% graphene appears to be due to improved graphene dispersion within the PCL matrix. At this concentration, interfacial interactions between graphene and polymer may better facilitate load transfer, mitigating brittleness observed at lower concentrations. Li. et al.17, define toughness as the ability to resist crack propagation and absorb energy before fracturing, while strength is the maximum stress a material can withstand before failure. High-strength materials are typically brittle and prone to sudden fracture, whereas high-toughness materials tend to be more ductile, allowing significant deformation before fracture16. The mechanical properties of scaffolds also play a role in cellular interactions, as mechanical cues from the substrate influence mechanotransduction-where cells convert mechanical stimuli from scaffold surfaces into biochemical signals. These signals can prompt various biological responses, including cell migration, regulated by the dynamic movement of pseudopodia. Recent studies suggest cells are more sensitive to the strain energy and strength of their surrounding material than stiffness alone. High strength and toughness are often incompatible in materials; strong materials typically exhibit low toughness. Therefore, developing materials that are both strong and tough remains a critical goal18. During cellular mechanotransductions, the substrate’s strain energy acts as a force sensor, transmitting mechanical tension to focal adhesions on scaffold suefaces. Focal adhesions, protein complexes that connect the extracellular matrix (ECM) with the actin cytoskeleton, initiate cellular responses. Tension-induced polarization and elongation of filopodia-a sensing structure-detect substrate strength. In favorable environments, filopodia activate signaling pathways, prompting actin cytoskeleton movement toward these regions. Subsequently, lamellipodia are formed, aiding cellular migration to target areas. Scaffolds with suitable mechanical properties support cell types such as MG-63 and L-929 by mimicking the ECM’s mechanical properties, allowing cells to interact with scaffolds in a manner similar to their natural surroundings19. In addition to mechanical cues, cells also possess receptors on their surface that bind to chemokines, and this triggered a signaling cascade that led to the formation of filopodia, which enabled the cell to probe its environment and detect the gradient direction of chemokine16. In the presence of a chemokine gradient, the concentration of the chemokine was greater at one end of the gradient than at the other20. The cells responded to this concentration difference by extending their filopodia in the direction of the higher concentration, allowing it to move in that direction21. However, when the chemical cue inhibited the formation of filopodia, the cells could be less able to detect and respond to the chemokine gradient22. Therefore, the cells could be unable to migrate toward the higher concentration of the chemokine, resulting in a decrease in migration speed, as shown in Fig. 2b and Fig. 2c 16,21. Filopodial protrusion was a dynamic process that involved multiple steps, including elongation, stabilization, and maturation10. The length of filopodia was likely determined by the specific set of molecules involved in each cell type and function, as different cells could have different requirements for protrusion length and dynamics6. One of the key regulators of filopodial elongation was the actin-bundling protein fascin, which promoted the formation of long and stable filopodia by crosslinking actin filaments and enhancing their rigidity2. After reaching their maximal length, filopodia entered the stabilization phase, during which the phase exhibited a slow rearward movement and established focal complexes with the ECM. This process enhanced force generation by connecting the filopodium to contractile actin fibers, which stabilized the filopodium. The duration of stabilization could vary among filopodia and likely depended on the specific cellular context23. The maturation phase was necessary for filopodia to re-enter the elongation phase, and this process involved adherence, contraction, and further development of filopodial focal complexes. The interplay between these processes determined the overall morphology and function of filopodia in different cell types. According to Gruening, et al.24, it appeared that during the interaction between the lamellipodium and filopodium, an optimal distance was reached that allowed for efficient communication between the 2 structures. The lamellipodium underwent sequential movements of protrusion and retraction during both the persistence and stabilization phases25. Additionally, it was suggested that during this state, the tip complex was either reassembled or reactivated to initiate the growth of filopodium26. A tip complex was a group of proteins located at the leading edge of a growing filopodium that was responsible for regulating the dynamics of actin filaments and promoting their elongation27. When lamellipodia and filopodia come into contact, the tip complex, which was responsible for the extension and guidance of filopodia, could become disrupted or inactive due to the mechanical forces and signaling events occurring at the contact site25. However, once the optimal distance between the 2 structures was reached, the tip complex could be reassembled or reactivated, allowing for the growth of a new filopodium28,29. Filopodia could transform into lamellipodia to initiate the directional formation and extension of lamellipodia and could also serve as scaffolds to guide the extension of lamellipodia. However, the exact sequence of events that led to the formation of filopodia and lamellipodia during cell migration was not fully understood. In addition, it was possible that filopodia first appeared and then guided the formation of lamellipodia25,30. The results showed that an increased concentration of graphene within scaffolds led to the development of longer filopodia, as shown in Fig. 3b and Fig. 4b. The observed phenomenon could be explained by a combination of 2 factors, namely, (1) decreased rates of both retraction and protrusion. The slowed rates of retraction and protrusion had the potential to cause elongated filopodia due to the additional time offered for extension and subsequent exploration of the surrounding environment, (2) increased force was required to overcome scaffolds friction. The wrinkles and ripples on scaffolds could increase the friction that filopodia experienced as these frictions moved along scaffolds. This increased friction could make it more difficult for filopodia to retract, resulting in a longer minimal length that filopodia could reach at the end of retraction, which could be prolonged due to the increased friction31,32. Filopodia initially surveyed the substrate to detect the surface structures of scaffolds. Upon encountering a pore, it established cell stability and attachment to the surface structure by using specific points along filopodia and their tips. The tips of filopodia then anchor into the pore, broaden, and branch out to become localized adhesive structures (footpads), which enabled subsequent cell spreading. However, when filopodia extended beyond a pore, their trailing tip adhesion could be pulled across the pore, potentially spanning33,34. Filopodia could divert away from the pore when the adhesion tips preferred to extend within the same plane rather than bending over an edge into a different plane. This behavior could be influenced by contact guidance, which was the force exerted on filopodia and the size of the pores in scaffolds. The surface topography could reshape cytoskeleton actin filaments, which could cause filopodia to either extend along or withdraw from pore edges instead of crossing over into the channel below33,35. A limitation of this study is the dispersion techniques used; although somewhat effective in distributing graphene, they may not be optimal in preventing graphene agglomeration36. Studies indicate that advanced methods, such as ultrasonic-assisted mixing or chemical surface modification, may enhance exfoliation, potentially yielding better mechanical properties. Future research could employ these alternative dispersion techniques to assess their effects on toughness and brittleness. Additionally, this study primarily examines mechanical behavior and cellular response using MG-63 and L-929 cell lines, while informative, these do not represent the full range of cell types that could interact with PCL/G composites in a physiological context. Testing other cell types could provide a more comprehensive understanding of how the material might perform in different biological environments. Finally, the study’s in vitro nature may not fully capture the in vivo conditions that would influence the PCL/G scaffolds/s mechanical properties and cellular interactions, especially as cells respond differently to substrates under physiological stresses and biochemical cues. Future work should consider in vivo studies to validate these findings and evaluate long-term stability, biodegradability, and biocompatibility of the composite in real-world applications. Methods Fabrication of scaffolds The fabrication of PCL (Mn: 80,000, Sigma, USA) and PCL/G scaffolds was achieved using a solvent casting and particle leaching technique37. The PCL was dissolved in chloroform at a temperature of 20 to 25°C for a duration of 12 h. The mixture was subsequently combined with different concentrations of graphene and sodium chloride (< 1μm in size) for a duration of 2 h. Graphene was produced by transferring a graphite intercalation compound into a crucible that had been preheated to a temperature of 700 °C. This process took place in a regular furnace situated in front of a fume cupboard to ensure safety against inhaling nanoparticles. The compound was then kept in the crucible for 60 s. The layers were expanded through the use of ultrasonication, resulting in the dispersion of graphene within the solvent. After the fabrication process, the mixture was poured and then allowed to cure at ambient temperature for the duration of the night. The chloroform underwent a 24-hour evaporation process at a temperature of 37 °C in a vacuum oven for drying purposes and was evaporated for 24 h at 37 °C in a drying vacuum oven (Deng Yng, Taiwan). To remove the porogen from scaffolds, deionized water and a water bath (BH-130D, Taiwan) were used. Furthermore, the water that was deionized was changed every 2 h and then dried in an oven at a temperature of 50 °C for a period of 12 h38. Cell culture Scaffolds used for cell culture had dimensions of 10 × 10 × 2 mm3 and were composed of different graphene weight ratios. The specimens underwent sterilization in a 95% ethanol solution for 24 h and were subsequently washed 3 times in a 1× PBS solution to eliminate any traces of ethanol that could be left behind. Before placing the cells, scaffolds were immersed in Dulbecco’s Modified Eagle Medium (DMEM; Gibco-Invitrogen) for a duration of 3 h. Osteoblast-like (MG-63) and mouse fibroblast (L-929) cells were obtained from 3D Global Biotech Inc, Taipei, Taiwan. The cells were grown in DMEM containing 10% fetal bovine serum (FBS) and 1% penicillin, and were incubated at 37 °C in a 5% CO2 environment. However, the culture medium was replaced at intervals of 2 to 3 days. After reaching 80% confluency, the cells were detached and transferred to a new culture using a solution of 0.25% trypsin-EDTA (Gibco, USA)8. Extract of scaffolds Scaffolds surface area was determined with the following equation:39 In the given formula, π was used to represent the value of 3.14, while r stood for the radius, and h represented the height. Furthermore, the DMEM solution was enhanced by including 10% FBS and 1% penicillin/streptomycin according to the following formula: The samples were transferred in a conical centrifuge tube with a capacity of 50 mL and then agitated in a shaking water bath set at a temperature of 37 °C and a speed of 100 rpm for a duration of 24 h. The extracts underwent filtration using a Millipore filter unit (Sartorius, France) equipped with a 0.22 μm pore size polyethersulfone membrane. Preparation of scaffolds and cell seeding The scaffold used for cell culture were 10 × 10 × 2 mm3 in dimension and contained varying G weight ratios. They were sterilized by immersion in a 95% ethanol solution for 24 h, then rinsed three times with 1x PBS to remove any ethanol residue. Prior to cell seeding, the scaffolds were incubated in Dulbecco’s modified Eagle medium (DMEM; Gibco- Invitrogen) for 3 h38. For cell culture, osteoblast-like (MG-63) and L-929 cells at passage 5 (provided by 3D Global Biotech Inc, Taipei, Taiwan) were grown in culture plates with DMEM containing 10% fetal bovine serum (FBS) and 1% penicillin. Incubation was carried out at 37°C with 5% CO2, and the medium was refreshed every 2 ~ 3 days. When cells reached 80% confluency, they were detached using 0.25% of trypsin-EDTA (Gibco, USA), and each sample was seeded with 0.5 mL at a concentration of 104 cells/mL in 24-well plates, then incubated for 14 days. The medium was refreshed every 2–3 days. Cell samples were collected from the incubator for evaluation on days 7 and 1436,40. To assess cell adhesion on the scaffold surface, scanning electron microscopy (SEM) was employed. After the culture medium was removed, samples were washes with 1 x PBS, then fixed in 0.6 mL of 2.5% glutaraldehyde in PBS for 30 min at 4 °C. Following two 1 x PBS washes, scaffolds were dehydrated in a graded ethanol series (30, 50, 70, 90, and 100%). The scaffolds were then placed in a critical point drying (CPD) chamber. Ethanol was replaced with liquid CO2 by pressurizing the chamber to dissolve the ethanol in the CO2 liquid phase, a process taking around 30 min. The chamber was gradually heated to the CO2 critical temperature (31.1 °C ) while maintaining pressure above the critical threshold (approximately 73.8 bar). After the transition, the chamber was carefully depressurized, allowing CO2 to convert from liquid to gas without passing through an intermediate liquid phase, thus preventing ice crystal formation. Once the pressure stabilized, the scaffolds were retrieved and stored in a desiccator until SEM analysis. For SEM visualization, the samples were gold-coated via sputter coating and observed at an accelerating voltage of 5 kV41. Quantification of lamellipodia area (lamellipodia morphometry) Lamellipodia region of MG-63 and L-929 cells was assessed at 7 and 14 days of incubation using Fiji software (NIH, USA). A total of samples was analyzed for each group, and after background subtraction, individual cells were selected using the wand tool, and their area and perimeter were calculated through the application of a threshold function. Lamellipodia were quantified by manually delineating regions of interest (ROIs) using a freehand tool as shown previously42. Filopodia analysis using FiloQuant A total of 5 scanning electron microscopy (SEM) images were obtained to assess cell morphology with a total of 5 samples for each group. The density and length of filopodia were analyzed using the FiloQuant plugin (version 1.52i) in ImageJ. Filopodia in the maximum intensity projections were analyzed through manual tracking and the single image tool. The filoQuant was used for filopodia length measurement. Additionally, filoQuant analysis defined the density of filopodia as the ratio of the value of identified filopodia to the length of the cell edge22,42. Scratch wound assay (migration) A volume of 1 mL of mouse fibroblasts (L-929) and osteoblast-like cells (MG-63) at a concentration of 3 × 104 was cultured in 12-well plates and incubated until these cells reached 90% confluency within 24 h. After the medium was discarded, the surface was scraped with a 200-µL tip, followed by rinsing with phosphate-buffered saline (PBS) to eliminate dislodged cells and debris. The extract from scaffolds was then added to each well as a chemoattractant reagent. The assay was performed using an Olympus IX73 inverted microscope at 100× magnification and processed with Gen 5.0 software. The wound edges’ distance was measured at 24 and 48 h and analyzed using ImageJ software. The calculation for wound closure was performed using the previously showed equation:43. In this equation, At=0h represented the initial measurement of the wound area immediately after scratching, and represented the measurement of the wound area h hours after the scratch. The migration rate could be determined using the following equation:43 In this equation denoted the migration rate of cells, referred to the initial width of the wound, represented the final width of the wound, and t signified the duration of migration. The study used 7 samples for each group. Tensile test The Shimadzu universal testing machine from Japan, equipped with a 250 N load cell was used to conduct the tensile test. The experiments were conducted at ambient temperature using a crosshead, and 5 samples were created by slicing scaffolds with dimensions of 40 × 20 × 10 mm3. The stress-strain curves were used to calculate the toughness, yield strength, and yield strain of each material. The toughness was determined using an equation:18 where T represented the toughness, σ was stress, and ε was strain. The yield strength was determined using the equation:44 where represented the yield strength, was the force, and was the cross-sectional area, and the yield strain was calculated using the following equation:44 where was the yield strain, was the total elongation, and was the original length. Pore size analysis Scaffolds’s pore diameters and morphology were analyzed through SEM (Hitachi, Japan) at an acceleration voltage of 15 kV. ImageJ was used to analyze the pores in the SEM images. Pore diameters were quantified on the SEM image using scale bars that represented a standard distance. The micrometer measurement of pore size was conducted using scaffolds of varying cross-sections as showed previously45. Statistical analysis Statistical analysis was frequently used to present experimental data for individual sample groups, using the mean ± standard error (SE) format. Each experiment consisted of a minimum of 3 replicates. The data underwent analysis using SAS software, and to identify any statistically significant differences in the data, ANOVA and Tukey’s test were used. For non-normal and non-homogenous distributions, the Kruskal-Wallis test and Mann-Whitney post hoc test were used to evaluate intergroup differences. The significance levels were set at a threshold of p < 0.0546. Conclusions In conclusion, mechanical and chemical cues on the surface of scaffolds played an important role in regulating the dynamics of filopodia and lamellipodia, which were essential for cell migration, adhesion, and other cellular processes. Additionally, it was possible that filopodia first appeared and guided the formation of lamellipodia after sensing and responding to the external environment by gathering spatial, topographical, mechanical, and chemical information from the material surface, which could reshape the cytoskeleton actin filaments and caused the materials to extend or withdraw from pore edges. Therefore, the dynamic regulation between filopodia and lamellipodia was of great interest in understanding cellular behavior on scaffolds surface. Future studies must focus on investigating the complex between filopodia and lamellipodia, expanding beyond osteoblast and fibroblast to include other relevant cell types. By elucidating the molecular and cellular mechanisms underlying this regulation, scientists could better understand how scaffolds could be optimized to guide cell behavior and enhance tissue regeneration. Acknowledgements We thank 3D Global Biotech Inc., Taipei, Taiwan, for providing the MG-63 and L-929 cell lines, which were invaluable to our research. Author contributions HSB, SA, YKS contributed to the Conceptualization, Formal analysis, Software, Validation, Visualization, Writing – original draft. S Y contributed to the Visualization, Proofreading. All authors reviewed the manuscript. Data availability The data analyzed to support the original study are included in the article. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Silvia Anitasari, Email: silvia.anitasari@fk.unmul.ac.id. Yung-Kang Shen, Email: ykshen@tmu.edu.tw. References Associated Data Data Availability Statement The data analyzed to support the original study are included in the article.