Tailoring weight ratio of PCL/PLA in electrospun three-dimensional nanofibrous scaffolds and the effect on osteogenic differentiation of stem cells

Abstract

Three-dimensional (3D) scaffolds as artificial ECMs have been extensively studied to mimic the critical features of natural ECMs. To develop more clinically relevant 3D scaffolds, electrospun nanofibrous scaffolds with different weight ratios of PCL/PLA (i.e., 100/0, 60/40, and 20/80) were fabricated via the thermally induced (nanofiber) self-agglomeration (TISA) method. The hypothesis was that, with the weight ratio increase of stiffer and more bioactive PLA in the 3D PCL/PLA blend scaffolds, the osteogenic differentiation of human mesenchymal stem cells (hMSCs) would be enhanced. The results indicated that, all of the 3D scaffolds were elastic/resilient and possessed interconnected and hierarchical pores with sizes from sub-microns to ~300 μm; therefore, the morphological structures of these scaffolds were similar to those of natural ECMs. The PLA80 scaffolds exhibited the best overall properties in terms of density, porosity, water absorption capacity, mechanical properties, bioactivity, and cell viability. Furthermore, with increasing the PLA weight ratio, the alkaline phosphatase (ALP) activity, calcium content, and gene expression level were also increased, probably due to the improved stiffness/bioactivity of scaffold. Hence, the novel 3D electrospun PLA80 nanofibrous scaffold might be desired/favorable for the osteogenic differentiation of hMSCs.

Graphical Abstract

1. Introduction

The worldwide incidence of bone disorders and repairs remains as a significant clinical challenge. The engineered bone tissue has been viewed as an alternative treatment strategy to replace autogenous and allogenous bone grafts owing to donor site morbidity, disease transmission, and immunoreactions [1–3]. In natural bone tissue, the extracellular matrix (ECM) composed of organic collagen nanofibers and inorganic nano-hydroxyapatite provides structural support for cells and regulates a variety of cell functions, such as assembling cells, regulating cell growth and cell-cell communication [4,5]. As a result, tissue engineering scaffolds have been extensively investigated to mimic the critical features of natural ECMs, including chemical composition, structural organization, and mechanical properties [6–8].

Among various scaffold-fabrication techniques, electrospinning has attracted growing interests because of its capability to convert a broad range of materials into fibers which are morphologically similar to the fibrous structures in natural ECMs [9–12]. The highly porous structure and large specific surface area of electrospun nanofibrous scaffolds might facilitate cell functions (e.g., adhesion, proliferation, migration, and differentiation) [13]; moreover, the nanofibers are able to be functionalized via incorporation of bioactive species (e.g., growth factor) to better control the cell functions. However, the major limitation of conventional electrospun nanofibrous membranes/mats is that the nanofibers are typically overlaid; hence, the membranes/mats generally behave as 2D scaffolds with apparent/equivalent pore sizes in sub-microns thus lack of necessary large pores for cell penetration and tissue formation. Therefore, to design innovative electrospun 3D scaffolds with appropriate fiber diameters, interconnected pores (including large pores), and good mechanical properties is of great significance. Although several approaches have been explored for making electrospun 3D structures/scaffolds [14,15], most of which have severe disadvantages. For example, (1) the mechanical properties are usually very low, due to the cotton-like stacks formed by random accumulation of nanofibers, without effective binding among nanofibers; (2) the pore structures (e.g., pore size and porosity) need to be improved; and (3) the fabrication methods are time-consuming, and the devices are complex. Recent research endeavors reported by the groups of Ding [16], Greiner [17], and Fong [18–21] have indicated that, upon breaking an electrospun nanofibrous mat/membrane into short nanofibers and/or tiny pieces followed by using them as the building materials to construct the 3D nanofibrous structure is an innovative and effective approach. The resulting 3D monolithic structures (i.e., aerogels/sponges/scaffolds) assembled from fragmented electrospun nanofibrous mats/membranes represent a rising and/or hot research topic in the electrospinning field. Owing to extremely low density, very high porosity, and excellent structural flexibility and stability, these 3D electrospun structures have attracted significant interests for various applications, especially for tissue engineering scaffolds.

In addition to scaffolds, stem cells and growth factors are also essential in tissue engineering [22,23]. Bone marrow mesenchymal stem cells (MSCs) have shown enormous potential in bone tissue engineering, because they can be differentiated into osteoblasts, chondrocytes, adipocytes, and smooth muscle cells [24–26]. It is important to note that the interaction between ECM and MSCs plays an important role in cell differentiation. An ideal scaffold can provide a favorable micro-environment for stem cell attachment, proliferation, and differentiation. Besides biochemical signals, the stem cell maintenance and differentiation can also be influenced by biophysical aspects of micro-environment, including mechanical properties, loading amount, and cell shape [27–31].

To develop more clinically relevant 3D nanofibrous scaffolds, we have recently invented the method of thermally induced (nanofiber) self-agglomeration (TISA) [18]. Polycaprolactone (PCL) has been selected as the base biomaterial, because it is a biocompatible and biodegradable polymer with low-cost, and it is approved by the US Food and Drug Administration (FDA) for biomedical applications. The resulting electrospun PCL-based nanofibrous scaffolds possess highly interconnected (porosity > 96%) and hierarchically structured pores (with sizes from sub-microns to hundreds of microns); hence, they are 3D scaffolds (instead of 2D substrates). Mouse MSCs (mMSCs) could readily penetrate and proliferate in these TISA scaffolds. Specifically, in vitro results indicated that the PCL-based TISA scaffolds could promote chondrogenic (rather than osteogenic) differentiation in a BMP-2 induced manner; while in vivo results demonstrated that new bone could be formed on the superficial layer of these scaffolds through the physiological endochondral ossification. In the follow-up work, human MSCs (hMSCs) were studied; additionally, the PCL-based TISA scaffolds were chemically modified through blending PCL with stiffer and more bioactive poly(lactic acid) (PLA) [19]. Although both PCL and PLA are linear aliphatic polyesters, the difference on macromolecular compositions makes PCL a more flexible, hydrophobic, and crystalline polymer that is slower to degrade than PLA; on the other hand, PLA has higher stiffness/modulus and strength than PCL. Compared to 3D PCL scaffolds, 3D PCL/PLA blend (with weight ratio of 80/20) scaffolds had substantially higher mechanical properties and in vitro bioactivity; consequently, they not only enhanced the cell viability of hMSCs but also promoted the osteogenic differentiation. Furthermore, in vivo studies revealed that these 3D PCL/PLA scaffolds considerably facilitated new bone formation in a critical-sized cranial bone defect mouse model. In other words, the results suggested that the stiffness and chemical properties of PLA might be beneficial to osteogenic differentiation. Nevertheless, only one weight ratio of PCL/PLA (i.e., 80/20) was investigated [19]; whereas the effect of PCL/PLA weight ratio (in electrospun 3D blend scaffolds) on stem cell differentiation needs to be further clarified.

In this work, we prepared a series of electrospun 3D PCL/PLA blend nanofibrous scaffolds with PCL/PLA weight ratios of 100/0, 60/40, and 20/80 (denoted as PCL, PLA40, and PLA80, respectively) via the TISA method; subsequently, the effect of PCL/PLA weight ratio on hMSCs differentiation was investigated in vitro. The hypothesis was that, with increasing the weight ratio of PLA in 3D PCL/PLA scaffolds, the osteogenic differentiation of hMSCs would be enhanced. The porosity, density, morphological structure, mechanical properties, and water absorption capacity of these PCL/PLA scaffolds were characterized. Moreover, the in vitro hydroxyapatite formation on different 3D nanofibrous scaffolds was examined using a simulated body fluid (SBF); while the cell viability and osteogenic differentiation of hMSCs were studied through measuring the ALP activity, calcium content, and gene expression.

2. Experimental

2.1. Materials

PLA resin 4042D (made from 95.8% L-lactide and 4.2% D-lactide, MW = 66,000 g/mol) was purchased from NatureWorks LLC (Minnetonka, MN). PCL (MW = 80,000 g/mol), gelatin (G1890), ethanol, dichloromethane (DCM), and N,N-dimethylformamide (DMF) were purchased from Sigma-Aldrich (St. Louis, MO). All of the chemicals/materials were used without further purifications.

2.2. Electrospinning

The solutions of PCL and PCL/PLA blends (with the PCL/PLA weight ratios being 60/40, and 20/80, respectively) were prepared using a mixture solvent of DCM/DMF (with the weight ratio of 2/1). The electrospinning parameters are shown in Table 1. Each solution was then put in a 30 mL BD Luer-Lok tip plastic syringe having a stainless-steel needle with 18 gauge 90° blunt end. The electrospinning setup consisted of a high-voltage power supply purchased from Gamma High Voltage Research (Ormond Beach, FL) and a digitally controlled syringe pump purchased from KD Scientific (Holliston, MA). The electrospun nanofibrous mat/membrane was collected on a laboratory-produced roller (with the diameter of 25 cm) covered with aluminum foil. During electrospinning, the distance between the syringe tip and the roller surface was set at 25 cm, while the rotational speed of roller was set at 100 rpm.

Table 1.

Electrospinning parameters for making nanofibers with different PCL/PLA weight ratios.

PCL/PLA (g/g)	Polymer concentration (wt.%)	Polymer weight (g)		Solvent weight (g)		Conditions
		WPCL	WPLA	WDCM	WDMF	Rate (mL/h)	Voltage (kV)
100/0	8	7.83	0	60	30	3	13
60/40	12	7.362	4.908	60	30	2.5	13
20/80	14	3.316	13.264	60	30	2	13

2.3. Preparation of short nanofibers and/or tiny pieces

Each electrospun nanofibrous mat/membrane was first cut into small pieces (with length and width of ~1 cm) and then soaked with cold ethanol; subsequently, the soaked nanofibrous pieces were fragmented into short nanofibers and/or tiny pieces by using a high-speed blender (i.e., a Waring Laboratory Blender) according to the following procedure: (1) after being blended for 10 min (at the no-load rpm of 19,600), the short nanofibers and/or tiny pieces (dispersed in ethanol) that went through a mesh/sieve (with pore size of ~1 mm) were collected; (2) the leftover large pieces were placed back into the blender for further breaking; (3) the above procedure was repeated for several times for collection of short nanofibers and/or tiny pieces. Thereafter, the acquired material (dispersed in ethanol) was transferred into a glass flask. After the system was stored under ambient conditions for 48 h, the short nanofibers and/or tiny pieces would precipitate at the bottom of flask; while the clear ethanol at the top of flask was carefully removed using a glass Pasteur pipette.

2.4. Fabrication of 3D nanofibrous scaffolds

Each type of the acquired short fibers and/or tiny pieces was first mixed with a gelatin solution. The obtained uniform suspension was then submerged into a water bath for the TISA process (with parameters shown in Table 2). The 3D agglomerate (soaked with water) was first placed at −15 °C for 1 h to turn water into ice and then submerged into liquid nitrogen for 3 min; finally, the sample was put into a pre-cooled glass flask followed by being freeze-dried at 25 °C for 24 h. For the PLA80 structures/scaffolds, after freeze-drying, the 3D samples were further stabilized at 105 °C for 5 min to improve the mechanical properties.

Table 2.

TISA processing parameters for making 3D nanofibrous scaffolds with different PCL/PLA weight ratios.

PCL/PLA (g/g)	Medium (VEthanol/VWater/VGelatin)	Temperature (°C)	Time (min)
100/0	4/2/1	52	2
60/40	4/2/0.5	52	3.5
20/80	4/2/0.25	65	30

2.5. Characterization

A Zeiss Supra 40VP field-emission scanning electron microscope (SEM) was employed to characterize the morphological structures of various samples.

The weight-based absorption capacity was calculated from the following equation:

$$\text{Capacity}=\frac{m_1-m_0}{m_0}\times 100\%$$

where m₀ and m₁ are the weights of scaffold before and after absorption of water, respectively.

The density and porosity were calculated from the following equation:

$${\rho }_{\text{scaffold}}=\frac{M}{V}\times 100\%$$

$$P_{\text{scaffold}}=\frac{V-V_P}{V}\times 100\%$$

where ρ_scaffold is the density of scaffold, P_scaffold is the porosity of scaffold, M is the total weight of scaffold, V is the total volume of scaffold, and V_p is the volume of PCL or PCL/PLA nanofibers (i.e., the weight divided by the density of PCL or PCL/PLA blend).

The Young’s moduli of electrospun nanofibrous scaffolds were measured by using an atomic force microscope (AFM, Model: MFP-3D BIO, Asylum Research, Santa Barbara, CA) mounted on an inverted microscope (Model: IX73, Olympus America Inc.). A silicon nitride tip (0.6 N/m) attached with a 5 μm diameter borosilicate sphere was used to indent each scaffold at the speed of 2 μm/s. Three different locations were measured for each scaffold. The Young’s modulus of a scaffold was determined upon fitting a modified Hertz model onto the AFM indentation curve using the built-in software (Asylum Research) [32].

The mineralization process of 3D PCL and PCL/PLA nanofibrous scaffolds was studied using the Kokubo’s SBF [33]. In specific, each scaffold was immersed in 15 mL SBF and placed in an incubator at 37 °C with constant shaking at 60 rpm for 7 days; and the SBF was refreshed twice a week. At the end of each incubation time, the scaffold was taken out of SBF, rinsed with deionized water, frozen at −20 °C overnight, and then freeze-dried.

The weight-increase percentages of PCL and PCL/PLA scaffolds were calculated from the following equation:

$$\text{Percentage}=\frac{m_1-m_0}{m_0}\times 100\%$$

where m₀ and m₁ are the weights of scaffold before and after being immersed in SBF for 7 days.

Fourier transform infrared (FT-IR) spectra of different scaffolds were acquired from a Tensor 27 FT-IR spectrophotometer (Bruker, Germany) equipped with a Smart Orbit diamond attenuated total reflection (ATR) accessory. The wavenumber range was from 4000 to 400 cm^–1, and each specimen was scanned for 32 times.

2.6. Cell seeding

The 3D scaffolds with different PCL/PLA weight ratios were first cut into disc samples (5 mm diameter × 2 mm thickness) with tissue punch, the samples were then immersed in 70% ethanol for 30 min followed by being rinsed with PBS for 3 times. Thereafter, these samples were incubated in minimum essential medium α (α-MEM, Gibco, Waltham, MA) for 30 min. The residual medium on a scaffold was removed with sterile gauze before hMSCs were seeded into scaffold (1 × 10⁵ cells per scaffold). All of the cells/scaffolds were cultured in 24-well plate on an orbital shaker (30 rpm) at 37 °C with 5% CO₂.

2.7. Cell viability

Cell viability was quantitatively analyzed using CellTiter 96 AQueous One Solution Cell Proliferation Assay (MTS, Promega, USA) according to the manufacture’s instruction. In brief, after culturing for 1 and 3 days, the culture medium was removed; fresh medium with 10% MTS was then added and incubated at 37 °C with 5% CO₂ in dark for 1 h. The absorbance was measured at 490 nm using a microplate reader (Infinite M200, Tecan, USA). The relative cell viability (%) was expressed as percentage relative to control group.

Morphologies of hMSCs on different 3D scaffolds were visualized upon staining with Texas Red-X Phalloidin (Life Technologies, Carlsbad, CA) and DAPI (Southern Biotech, Birmingham, AL), which were able to label F-actin and cell nuclear, respectively. The cells/scaffolds were examined using a laser scanning microscope (FV1200, Olympus, Japan) as reported previously [19].

2.8. ALP activity and calcium content

ALP activity was examined using an EnzoLyte pNPP Alkaline Phosphatase Assay Kit (AnaSpec, San Jose, CA), as we previously described (with some minor modifications) [34]. In brief, cells/scaffolds were rinsed with PBS solution and lysed with lysis buffer for 2 min at room temperature. The lysate was then transferred into a tube followed by being centrifuged (at 2,500 g) for 15 min at 4 °C. The collected supernatant or standard solution (50 μL) was then mixed with p-nitrophenyl phosphate and incubated for 30 min at 37 °C. After the incubation, the reaction was stopped by adding 100 μL termination liquid. ALP activity was measured at 405 nm and normalized against total protein content. The total protein content was measured with a BCA Kit (Thermo Scientific^™, Waltham, MA) according to the manufacturer’s instruction. In brief, 25 μL of the collected supernatant (the same from ALP activity) or standard solution was mixed with 200 μL BCA working reagent and incubated for 30 min at 37 °C. Following the incubation, the protein content was measured at 562 nm. The cell-scaffold constructs were also examined for calcium deposition using a Total Calcium LiquiColor® Kit (Stanbio laboratory, TX). After 3 weeks of culture, cells/scaffolds were rinsed with DPBS and cut into small pieces with a sharp blade. The calcium was extracted by using 1 mL 6 M hydrochloric acid. Thereafter, 10 μL extraction solution or 10 μL standard solution was added into 1 mL working solution that was prepared according to the manufacturer’s instruction. The absorbance was measured at 550 nm, and the calcium content was calculated from the following equation:

$$\text{Calcium}(\mathit{mg}∕\mathit{dL})=\frac{\mathit{Au}}{\mathit{As}}\times 10$$

where Au and As are the absorbance values of sample and standard, respectively.

2.9. Gene expression analysis

Quantitative gene expression analysis was carried out as we previously described (with some minor modifications) [35]. In brief, total RNA was extracted using the GeneJET^™ RNA Purification Kit (Thermo Scientific^™, Waltham, MA) by following the manufacturer’s instruction. RNA concentration was measured with UV-Vis spectroscopy (DU 730, Beckman coulter) at 260 nm, an equivalent amount of RNA was processed to generate cDNA using the High Capacity cDNA Reverse Transcript Kit purchased from Applied Biosystems (Forster City, CA). Quantitative PCR was performed with Taqman gene expression assays (Applied Biosystems, Forster City, CA) using the Applied Biosystems 7500 Fast Real-Time PCR System (Applied Biosystems, Carlsbad, CA). Triplicates were performed for each sample, and the acquired results were normalized to β-actin. Gene primers of β-actin (Hs01060665), ALP (Hs00758162), and Bone sialoprotein (BSP) (Hs00173720) were purchased from Applied Biosystems (Forster City, CA).

2.10. Statistical analysis

To determine statistical significance of observed differences between the study groups, a two-tailed homoscedastic t-test was applied. A value of p < 0.05 was considered to be statistically significant, while 0.05 < p < 0.10 was considered to represent a non-significant while clear trend in cell or tissue response. Values are reported as mean ± one standard deviation.

3. Results and discussion

3.1. Preparation of 3D scaffolds via the TISA method

Nanofiber mats/membranes were prepared by electrospinning of PCL and PCL/PLA blend solutions (Fig. 1A). As shown in Fig. S1, the electrospun PCL and PCL/PLA mats/membranes consisted of overlaid nanofibers with diameters being hundreds of nanometers. These nanofibers had relatively uniform morphological structures, and the mats/membranes contained no microscopically identifiable beads and/or beaded nanofibers. The obtained mats/membranes were then cut into small pieces followed by being mechanically fragmented in ethanol using a high-speed blender. Thereafter, short nanofibers and tiny pieces were obtained (Fig. S2). It was observed that, with more brittle PLA being blended with PCL, the resulting nanofibers became easier to break. To acquire the uniform suspension and to control the subsequent nanofiber self-agglomeration, gelatin was adopted to increase the viscosity of suspension system; while deionized water was added to improve the solubility of gelatin (Fig. 1B). Note that the short nanofibers and tiny pieces could be stabilized by gelatin molecules; as a result, the obtained suspension systems were reasonably uniform. Upon submerging a glass bottle (containing a suspension system) into a water bath at the pre-determined temperature for certain time, the short nanofibers and tiny pieces of PCL or PCL/PLA could gradually self-agglomerate to form a 3D nanofibrous structure (Fig. 1C). This was because, at the pre-determined temperature, the surface of PCL or PCL/PLA nanofibers would become soft/sticky, since the melting point of PCL is ~60 °C. Immediately afterward, the bottle was taken out of the water bath and submerged into ice water to prevent the 3D structure from further shrinkage. Finally, the obtained 3D nanofibrous structure was rinsed with deionized water and then freeze-dried to acquire the shape-stable 3D scaffold (Fig. 1D).

Fig. 1.

Schematic diagram showing the preparation of electrospun 3D PCL and PCL/PLA nanofibrous scaffolds: (A) electrospinning setup, (B) nanofiber suspension containing ethanol, water, and gelatin, (C) nanofiber agglomeration upon TISA process, (D) freeze-dried 3D scaffold.

The TISA temperatures for making the PCL and PLA40 scaffolds were the same at 52 °C. This was because the amount of PCL was higher than that of PLA, thus the TISA process was primarily attributed to PCL; whereas the TISA time had to be prolonged with the increase of PLA weight ratio in the blend nanofibers. For making the PLA80 scaffold, in which the amount of PCL was lower than that of PLA thus PCL would only act as binding agent, the TISA temperature had to be higher at 65 °C; and the TISA time was also longer. Note that the melting point of PLA is ~175 °C. During the TISA process, all of the samples except for PLA80 had similar shrinkages. While for the PLA80 samples, the nanofiber agglomeration occurred slowly even at the higher temperature of 65 °C, resulting in the lowest volume shrinkage. Concomitantly, PLA 80 samples had the highest porosity while lowest density (Fig. S3). Experimental results also indicated that further increasing the temperature and/or time did not distinguishably facilitate the nanofiber agglomeration. In this case, the PLA80 nanofibers were not able to be tightly bound/fused; hence, the subsequent thermal treatment at 105 °C for 5 min (after the freeze drying process) became necessary to further stabilize the scaffold.

In principle, the TISA method is applicable to electrospun nanofiber mats of all thermoplastic polymers. However, the melting points of many polymers are higher than 150 °C; in this case, high boiling point compounds/solvents (rather than water) are needed to partially melt the nanofibers. It is important to note that these solvents are usually toxic to cells and also difficult to be thoroughly removed. Furthermore, even short nanofibers can self-agglomerate together upon using high boiling point solvents, the resulting 3D nanofibrous scaffolds are often unstable due to high rigidity/stiffness of these nanofibers. Therefore, 80% was chosen as the highest weight ratio of PLA (in the blend nanofibers) in this study.

3.2. Morphological structure

Morphological structures of the 3D scaffolds were characterized by SEM. All of the scaffolds had interconnected and hierarchically structured pores with sizes ranging from sub-microns to ~300 μm, as shown in Fig. 2. Such morphological structures would be highly favorable for cell growth. Specifically, the large pores (with sizes in hundreds of microns) would maintain the structural stability of scaffold, support cell proliferation, ECM deposition, and tissue formation; the medium pores (with sizes in several to tens of microns) would facilitate the diffusion of nutrients and promote the formation of vascularization, while the small pores (with sizes in microns or smaller) would have positive impacts on some cell behaviors such as seeding and gene expression [36–38]. Compared to porous aerogels/sponges prepared via the ice-template/freeze-drying method [16,17,39], the pore structures in our TISA scaffolds were formed through random assembly of short nanofibers and/or tiny pieces. The PCL/PLA nanofibers were in situ bound/fused together (Fig. 2A3–C3, marked by arrows), which could not only avoid toxic binding agents but also provide good mechanical properties.

Fig. 2.

SEM images showing the morphological structures of electrospun 3D nanofibrous scaffolds of (A1–A3) PCL, (B1–B3) PLA40, and (C1–C3) PLA80 under different magnifications. Insets (in A1, B1, and C1) are photographs of respective 3D scaffolds.

3.3. Water absorption capacity

During the TISA process, gelatin was added to increase the viscosity of suspension system and to stabilize the nanofibers; additionally, the gelatin on surfaces of hydrophobic PCL or PCL/PLA nanofibers could lead to substantial improvement on hydrophilicity of the scaffolds. Although the samples were rinsed with deionized water for several times, some gelatin could be retained on the nanofiber surface, making the scaffolds hydrophilic. Taking PLA80 scaffolds as examples, without gelatin during the TISA process, the 3D scaffolds could still be obtained; whereas they were hydrophobic (with the water contact angle of 127.2°, as shown in Fig. S4).

The prepared 3D scaffolds possessed low density, high porosity, hydrophilic surface, and interconnected pore structure; hence, they could be used as water absorbents. For example, the PLA80 scaffold would absorb water immediately upon immersion (Fig. 3A1). After squeezing, some residual water could remain in the scaffold; and the squeezed scaffold could not fully recover into its original shape (Fig. 3A2), unless it was re-immersed in water (Fig. 3A3). This was intriguing, suggesting that the prepared 3D nanofibrous scaffolds possessed the shape-memory characteristic. As shown in Fig. 3B, the weight-based absorption capacities of all scaffolds were in the range from 17~55 times of their own weights, depending upon the porosity and/or density of the scaffolds. Among the three types of samples, the PLA80 sample had the highest absorption capacity, due to the highest porosity and/or the lowest density. The high water absorption capacity would enable the scaffolds with the capability to quickly absorb cell culture medium. Note that even repeating the absorption-squeezing process for five times, the water absorption capacity was not changed appreciably (Fig. 3C).

Fig. 3.

(A1–A3) Photographs showing one water absorption-squeezing cycle of a PLA80 scaffold. Inset: (A1) the scaffold before immersion in water, and (A3) the scaffold after re-immersion in water. (B) Weight-based absorption capacities and residual amounts of different 3D scaffolds on water. (C) Absorption capacities and residual amounts of PLA80 scaffolds over 5 water absorption-squeezing cycles.

3.4. Mechanical properties

Electrospun nanofibrous mats/membranes were soft, while the prepared 3D scaffolds were elastic/resilient. As shown in Fig. 4A1–A3, all of the scaffolds with varied PCL/PLA weight ratios could be bent and then recovered to their original shapes upon releasing the bending force. The hierarchically structured pores and nanofiber binding/fusing resulted in robust mechanical properties under large bending strain. To quantitatively measure the mechanical properties of individual/local nanofibers in these 3D scaffolds, AFM was employed. As shown in Fig. 4B, no significant difference was observed between PLA40 and PLA80 scaffolds. However, the Young’s moduli of both PLA40 scaffold (45.21 ± 6.82 KPa) and PLA80 scaffold (54.63 ± 5.23 KPa) were significantly higher than that of PCL scaffold (8.34 ± 1.06 KPa). AFM results indicated that the incorporation of PLA into PCL could substantially increase modulus/stiffness of the resulting scaffolds [19,40], and the critical weight ratio of PLA in the blend nanofibers might be below 40%.

Fig. 4.

A series of photographs showing the elasticity/resilience of different 3D scaffolds: (A1) PCL, (A2) PLA40, and (A3) PLA80. (B) Young’s moduli of different 3D scaffolds measured by AFM. Data are expressed as mean ± one standard deviation (n = 3).

3.5. Bioactivity

To study in vitro bioactivity of 3D PCL and PCL/PLA blend scaffolds, the prepared samples were immersed in SBF for 7 days. The formation of apatite-like deposits was noticeable on PCL scaffolds (Fig. 5A1 & A2), while more deposits were uniformly dispersed on those PCL/PLA blend scaffolds with higher PLA weight ratio (Fig. 5B–C), suggesting that the incorporation of more PLA into PCL was able to provide improved bioactivity of the resulting PCL/PLA blend nanofibers. After being immersed in SBF for 7 days, the weight-increase percentages of PCL and two PCL/PLA blend scaffolds were 15%, 38%, and 81%, respectively (Fig. S5). These results were further supported by the FT-IR spectra acquired from PCL and PCL/PLA blend scaffolds before and after SBF treatment. With the decrease of PCL amount and the corresponding increase of PLA amount in blend scaffolds, the band at 1724 cm⁻¹ (attributing to the vibration of C=O in PCL) decreased, while the band at 1759 cm⁻¹ (attributing to the vibration of C=O in PLA) became pronounced (Fig. S6A) [41]. After 7 days of SBF immersion, the vibration bands at 631 and 1650 cm⁻¹ (attributing to −OH groups in hydroxyapatite [42,43]) were detected (Fig. S6B); whereas no such characteristic bands were observed on the scaffolds prior to SBF immersion. Moreover, the intensities of these bands increased with the increase of PLA weight ratio in the blend scaffolds.

Fig. 5.

SEM images showing the morphological structures of different 3D scaffolds of (A1, A2) PCL, (B1, B2) PLA40, and (C1, C2) PLA80 after being immersed in SBF for 7 days. Arrows indicate the apatite-like deposits.

3.6. Cell viability

Morphologies of hMSCs on different 3D scaffolds were studied after culturing for 24 h. As shown in Fig. 6A–D, cells were able to adhere and spread well with typical fibroblastic morphologies on all types of scaffolds. No distinguishable differences were observed among cell morphologies on three types of scaffolds. Furthermore, cell viabilities of hMSCs on different 3D scaffolds were measured quantitatively using the MTS assay after cultured for 1 and 3 days. As shown in Fig. 7, no obvious discrepancy was detected on all of the 3D scaffolds with different PCL/PLA weight ratios after 1 day of cell culture. However, on day 3, the cell viabilities on PLA80 scaffolds were significantly higher than those on neat PCL scaffolds (*P < 0.05). This might be attributed to the improved mechanical properties and bioactivity of PCL/PLA scaffolds. It was reported that higher mechanical properties might result in better support for cell growth by providing enough oxygen/nutrient exchanges, which has been commonly known as a major challenge for cell culture in thick 3D scaffolds [44]. Additionally, it has been demonstrated that the mechanical properties of ECM are capable of modulating cell behaviors, such as spread area, morphology, and gene expression profile. For example, the MSCs on soft substrates/scaffolds had less spread, fewer stress fiber, and less proliferation rate than the MSCs on stiff substrates/scaffolds [45,46].

Fig. 6.

Morphologies of hMSCs on different scaffolds after culturing for 1 day: (A) PCL, (B) PLA40, and (C) PLA80. (Left panel: top view, right panel: 3D view).

Fig. 7.

Viabilities of hMSCs on different 3D scaffolds after 1 and 3 days of culture. Data are expressed as mean ± one standard deviation (n = 3). *P < 0.05.

3.7. Osteogenic differentiation of hMSCs

ALP activities of hMSCs on different 3D scaffolds in either growth medium (GM) or osteoinductive medium (OI) were studied after 7 days of culture. In GM, cells on the PLA80 scaffold had significantly higher ALP activity than those on the PCL scaffold (P < 0.01), while no such results were observed on the PLA40 scaffolds. Compared to GM, OI substantially elevated the ALP activity of hMSCs cultured on all types of 3D nanofibrous scaffolds. Note that the ALP activity exhibited a dose-dependent manner with the increase of PLA weight ratio in the blend scaffolds (Fig. 8A). Moreover, the mineralization (the late marker of osteogenic differentiation) was also analyzed by measuring the calcium content. After 3 weeks of culture, a similar amount of calcium was detected on all types of nanofibrous scaffolds in GM. In OI, the PLA 40 and PLA 80 scaffolds showed similar calcium contents, which were significantly higher than the calcium content of PCL scaffold (Fig. 8B). In addition to the ALP activity and mineralization, the quantitative gene expression was also studied. As shown in Fig. 8C, the PLA-containing scaffolds (i.e., the PLA40 and PLA80 scaffolds) expressed significantly higher levels of osteogenic gene RUNX2 than the neat PCL scaffold. In BSP, PLA80 scaffolds showed significantly higher level of gene expression than both PCL and PLA40 scaffolds (Fig. 8D), which could be explained by that BSP would act as nucleus for the first apatite crystallite and regulate the bone matrix mineralization [47,48].

Fig. 8.

(A) ALP activities of hMSCs cultured on different 3D scaffolds for 7 days in growth medium (GM) and osteoinductive medium (OI). The ALP activity was normalized by total protein content. (B) Calcium contents from different 3D scaffolds were measured after culturing in GM and OI for 3 weeks. (C and D) Osteogenic marker gene expressions (C) RUNX2 and (D) BSP were studied by real-time PCR assay after 7 days of culture in OI. Data are expressed as mean ± one standard deviation (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001.

It is well known that ECM can deliver biochemical and biophysical signals to direct MSC differentiation. Among biophysical signals, the ECM stiffness is a potent regulator of stem cell differentiation. AFM was employed to quantitatively measure the mechanical properties of individual/local nanofibers in the 3D scaffolds with varied PCL/PLA weight ratios, and the results indicated that the incorporation of PLA into PCL could substantially increase modulus/stiffness of the blend scaffolds. Therefore, the improved osteogenic differentiation might be attributed to the increase of scaffold stiffness, because stiffer ECM/scaffold has been demonstrated to direct MSCs towards the osteogenic (rather than the chondrogenic) differentiation pathway [45,46].

4. Conclusions

A series of electrospun PCL and PCL/PLA blend 3D scaffolds (with different PLA weight ratios) were prepared via the TISA method. All of the 3D scaffolds were elastic/resilient and possessed interconnected and hierarchical pores with sizes from sub-microns to ~300 μm; hence, the morphological structures of these scaffolds were similar to those found in natural ECM. The PLA80 scaffold exhibited the best overall properties, i.e., low density (12.6 mg/cm³), high porosity (99.0%), water absorption capacity (53.5 g/g), bioactivity, Young’s modulus (55.2 kPa), and cell viability (117%). Moreover, two PCL/PLA blend scaffolds significantly promoted more potent osteogenic differentiation of hMSCs than neat PCL scaffold. With the increase of PLA weight ratio in PCL/PLA blend scaffolds, the ALP activity, calcium content, and gene expression level were also increased, probably due to the improved stiffness and bioactivity. Therefore, both the overall properties and in vitro results suggested that the 3D electrospun PLA80 nanofibrous scaffold might be desired/favorable for osteogenic differentiation of hMSCs.

Highlights.

• ▶ Electrospun 3D scaffolds with different weight ratios of PCL/PLA were fabricated.

• ▶ These 3D scaffolds were elastic and possessed interconnected and hierarchical pores.

• ▶ The PLA80 nanofibrous scaffold exhibited the best overall properties.

• ▶ They might be highly desired/favorable for the osteogenic differentiation of hMSCs.