Sequential release of transforming growth factor β1 and fibroblast growth factor 2 from nanofibrous scaffolds induces cartilage differentiation of mouse adipose-derived stem cells Open Access

Once damaged, cartilage displays poor intrinsic capacity for repair, possibly resulting in diseases such as airway obstruction, osteoarthritis (OA), or deformation of the ears and nose.^1–9 However, there are exceptions to this. Previous reports have indicated that minor defects or defects in the fetus or highly immature cartilage show some capacity to heal without intervention.^4,5 Some researchers have shown in a fetal lamb model that there is spontaneous repair in partial defects of articular cartilage.⁶ Mature animals seem not to have this repair process. Integration of restored cartilage with the host tissue is a significant condition for lesion repair.⁷ Several authors documented that poor integration can lead to failure of cartilage repair in the natural cartilage repair process.^2,8,9 During the transplantation of periosteal and perichondrial grafts,^10,11 natural,¹² and bioengineered grafts,¹³ this failure of repair was found as well. To enhance the success rate of repair, a modification step is required before graft tissue transplantation to achieve a proper fit into the defect.

Currently, various treatment strategies provided in the literature utilize scaffold materials primarily composed of biodegradable polymers, such as collagen, chitosan, and alginate, or synthetic polymers like polylactic acid, polyglycolic acid, and their copolymers. These scaffolds are either loaded with growth factors (GFs),^14–16 preseeded with chondrocytes or mesenchymal stem cells (MSCs),^17,18 or used in combination with both growth factors and cells.^19,20 Overall, these studies have achieved some promising results, but not all outcomes have met the anticipated expectations.¹⁹ A promising strategy is provided by biomimetic scaffolds to engineer cell/scaffold constructs which once implantated, can integrate into native tissues, and then the injured tissue is recovered to its original biological and mechanical state.²¹ 

In cell differentiation studies, the most widely used approach for induction involves release of growth factors from scaffolds.²² Researchers have developed various methods to incorporate growth factors into scaffolds. For instance, the recombinant human bone morphogenetic protein 7(rhBMP-7) is encapsulated in PLGA microspheres using double emulsion techniques and implanted onto the surface of nanofiber PLLA scaffolds.²³ A study revealed that the biomimetic PLLA scaffold combined with rhBMP-7 microspheres significantly promoted bone tissue formation. After six weeks of subcutaneous implantation in SD rats, the rhBMP-7 microsphere scaffold demonstrated a significantly higher total bone formation compared to the scaffold with rhBMP-7 simply surface-adsorbed.

An insulin-transferrin-selenium (ITS+) based medium supplemented with 10 ng/ml of transforming growth factor β1 (TGFβ1) is usually applied to induce chondrocytes toward a chondrogenic phenotype in 3D culture.²⁴ However, chondrogenic differentiation efficiency of the conventional culture medium is relatively low, and it is necessary to explore more effective combinations of growth factors to promote chondrogenic differentiation. A previous study reported that in wounded cartilage, maximal expression of FGF2 appeared at day 7, and a gradual decrease was followed; TGFβ1 expression was increased from day 3 and lasted until day 14.²⁵ Media supplemented with fibroblast growth factor 2 (FGF2) has been proved to facilitate chondrocyte monolayer expansion as well as chondrogenic capacity in the aspects of ECM synthesis and chondrocyte specific phenotype expression.^26–28 However, the application of TGFβ1 and FGF2 in different sequences and the effects of these sequences on chondrogenic differentiation have not been reported in the existing literature. Therefore, further experimentation is required to explore this aspect. In this research, the effect of these growth factors in different sequences for chondrogenesis differentiation was tested. The hypothesis of this study is that a proper spatial and temporal release of growth factors FGF2 and TGFβ1 will lead to an increased chondrogenic differentiation of stem cells and facilitate the deposition of extracellular matrix in cartilage tissues. The aim of this study is to determine the optimal release sequence of growth factors FGF2 and TGFβ1 and establish a microenvironment by using the biomimetic scaffold that can support chondrogenesis and promote in vivo cartilage regeneration, providing valuable insights for more effective promotion of cartilage tissue regeneration in translational models.

Approval of all experimental protocols was granted by the Medical Ethics Committee of Xiangya Hospital affiliated to the Center for Medical Ethics, Central South University and all experimental procedures were carried out in accordance with the guidelines by the Medical Ethics Committee of Xiangya Hospital affiliated to the Center for Medical Ethics, Central South University (Protocol No. 202103743). The reporting in this manuscript followed the recommendations in the ARRIVE guidelines. Fabricated porous NF scaffolds according to the method reported in a previous work.²⁹ Briefly, PLLA (EVONIK, Germany) with an intrinsic viscosity of approximately 1.6 dl/g was dissolved at 60 °C in tetrahydrofuran (THF) (10% w/v) and cast under mild vacuum into an assembled sugar template (composed of bound D-fructose spheres, 150–250 μm in diameter). Phase separate PLLA/D-fructose composites at −80 °C overnight and then immerse the composites in hexane for 2 days for exchange of THF. The resulting composites were freeze-dried and distilled water was used to leach out the D-fructose spheres to form a network with interconnected spherical pores. Highly porous scaffolds were obtained after freeze-drying again and cut into circular disks with a thickness of 1.5 mm and a diameter of 5 mm. A JEOL-7800FLV scanning electron microscope (JEOL, USA) was used to take the macroimage of the scaffold (Fig. 1).

A water-oil-water double emulsion process with a probe sonicator (VirTis VirSonic 100 Ultrasonic Cell Disruptor) was used to fabricate the loaded poly(lactic-coglycolic acid) (PLGA) particles. Briefly, 5% (w/v) of PLGA 4A polymer (Lakeshore Biomaterials) was dissolved in methylene chloride, forming the larger oil phase. 1000 ng/ml of TGFβ1 was prepared in deionized water as the smaller water phase, and a 1% (w/v) polyvinyl alcohol (PVA) solution was also prepared as the larger water phase. Single water-in-oil emulsification was first prepared by combining 1 ml of polymer in dichloromethane (DCM) with 100 μl of TGFβ1 solution and sonicating for 10 s on power 10. This phase was then used to prepare the second emulsification by slowly adding the contents of the first emulsification into 20 ml of PVA on a continuous sonication mode at a power of 20. Particles were stirred for 3 h for complete evaporation of methylene chloride. Then, centrifuged the particles for 6 min at a speed of 12 000 RPM which were then washed twice, lyophilized, and sterilized under ethylene oxide (EtO) gas before they were ready for use (Fig. 1).

According to the protocol, adipose tissue was harvested by removing the inguinal fatpad from wild type C57BL/6 mice and digested to isolate ADSCs.³⁰ Briefly, ADSCs are isolated from adipose tissue by washing the tissue sample extensively with PBS containing 5% penicillin/streptomycin (P/S) followed by digesting the tissue by immersion in 0.075% type I collagenase prepared in PBS containing 2% P/S. The sample was incubated for 30 min at 37°C and 5% CO₂. Collagenase type I activity was then neutralized through adding 5 ml of a-MEM (Mediatech, Herndon, VA) that contains 20% heat inactivated fetal bovine serum (FBS, Atlanta Biological, Atlanta, GA) to the tissue sample. The sample was transferred to a 50 ml tube after disintegration, avoiding solid aggregates, and then, centrifuged at 2000 RPM for 5 min three times. The cell pellet was resuspended in a maximum of 3 ml of stromal a-MEM medium supplemented with 20% FBS, 1% L-glutamine (Mediatech), and 1% P/S. The cells were subpassaged while they reached 80% confluence and the cells prior to passage 5 were applied in the subsequent researches.

Before being imaged with a JEOL-7800FLV SEM (JEOL, USA) at 15 kV, a stereo microscope was used to observe the blank PLLA scaffolds, and then, the scaffolds were sputter-coated with gold. After 1 day of culture, 2.5% phosphate-buffered glutaraldehyde (Sigma-Aldrich, St. Louis, MO) was used to fix the cell/scaffold constructs overnight at 4 °C, and then, the constructs were poststained for 1h with 1% osmium tetroxide (Sigma-Aldrich). After being rinsed three times using PBS, and dehydrated via a graded ethanol series (50% → 70% → 100%), these samples were dried with hexamethyldisilazane (HMDS, Sigma-Aldrich) as previously described.²⁹ After sputter-coated with gold, these samples were imaged with a JEOL-7800FLV SEM at 15 kV to observe cell behaviors such as adhesion, spreading, and ECM deposition on and within the scaffolds.

For 2D cell culture, mouse ADSCs prior to passage 5 (50 000 cells per well) were plated in a 12-well plate. The concentrations of growth factors were according to the method reported in previous works,^26,31 and the monolayered cells were treated with (1) TGFβ1 (10 ng/ml) (Pepro Tech Ins., NJ, USA) for 1 week, followed by FGF2 (5 ng/ml) (Stemgent) (T → F); (2) FGF2 (5 ng/ml) for 1 week, followed by TGFβ1 (10 ng/ml) for 1 week (F → T); (3) FGF2 (5 ng/ml) alone for 2 weeks (F); (4) TGFβ1(10 ng/ml) alone for 2 weeks (T); (5) a mixture of FGF2(5 ng/ml) and TGFβ1(10 ng/ml) for 2 weeks (T + F); and (6) a growth medium as control (blank). Induction supplements [1X ITS, ascorbic acid (0.1 mM), L-proline (40 μg/ml), and dexamethasone (0.1 μM)] were included in all of the TGFβ1 and/or FGF2 treatment groups (groups 1–5).

For 3D cell culture, 70% ethanol was used to prewet the porous NF PLLA scaffolds for 30 min, and mild vacuum was applied to expel the internal air bubbles.³¹ After three washes with PBS and two washes with the culture medium, seeded each scaffold with 2 × 10⁵ mouse ADSCs. The cell/scaffold constructs were cultured with the following treatment groups as described above: (i) a growth medium (blank); (ii) TGFβ1 (10 ng/ml) for 2 weeks (T); and (iii) FGF2 (5 ng/ml) for 1 week, followed by TGFβ1 (10 ng/ml) for 1 week (F → T).

After 2 weeks of treatment, the harvested samples (monolayer cells and cell/scaffold constructs) were collected for western blot and PCR analysis. The cell number of the monolayer samples and cell/scaffold constructs were quantified using CellTiter-Blue (Promega) on the last day of the experiment.

Trizol reagent (Life Technologies Corporation) and disposable plastic pestles (Fisher Scientific, Pittsburg, PA) were used to extract the total RNA from the monolayer cell samples and the cell/scaffold constructs. RNA concentration determination was accomplished by testing the optical absorbance of the extract at 260 nm. The SuperScript II cDNA Synthesis kit (Invitrogen) was used to synthesize complementary DNA (cDNA) according to the protocol. Real-time PCR was carried out using SYBR Green PCR Master Mix (Applied Biosystems) with predesigned primers for Sox9 (Forward: GACTTCCGCGACGTGGAC, Reverse: GTTGGGCGGCAGGTACTG) and collagen type II (Forward: CCGTGGTGAGGCTGGTC, Reverse: GCACCAGGTTGGCCATCA). The reactions were performed using an ABI 7500 real-time PCR system (Applied Biosystems). The housekeeping gene 18S was used to normalize the gene expression (forward: TAGAGGGACAAGTGGCGTTC and reverse: CGCTGAGCCA-GTCAGTGT).

The total protein was extracted from the cultured cells and cell/scaffold constructions, and the protein concentration was determined with a BCA Protein Quantification Kit (Thermo Scientific). After being separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, proteins were transferred to a PVDF transfer membrane (Bio-Rad, USA). Used tris buffered saline-tween (TBS-T) with 5% BSA to block the membranes at room temperature for 1h. Then, the appropriate primary antibody was used to incubate the membranes: anticollagen II (Thermo Fisher) or anti-β-actin antibody (Santa Cruz Biotechnology) at a 1:1000 dilution overnight at 4 °C in TBS-T. Then, a horseradish peroxidase (HRP)-labeled secondary IgG antimouse or antirabbit (Santa Cruz Biotechnology, 1:5000) was used to incubate the membranes at room temperature for 1h. An ECL Prime Western Blotting detection reagent (GE Healthcare) was used to detect the immunoreactive bands and the bands were exposed to an X-posure film (Thermo Scientific) for 5–10 min.

To determine the release profile of TGFβ1 from PLGA nanoparticles, 5 mg of nanoparticles were put into 1 ml of PBS, pH 7.4 in an Eppendorf tube. At designated timepoints (2 h, 6 h, 8 h, 24 h, 2 d, 3 d, 5 d, 7 d, 10 d, 13 d, 21 d, and 29 d), centrifuged the samples at 12 000 RPM for 6 min and collected the supernatant for protein quantification. Nanoparticles were then resuspended into new PBS for the next timepoint. The factor concentration was detected with a BCA Protein Quantification Kit (Thermo Scientific).

In order to study the capacity of the scaffolds to improve chondrogenesis in seed cells and investigate whether the FGF2 to TGFβ1 sequential treatment showed positive effects in vivo, the precultured constructs were implanted subcutaneously into a model of mice in vivo. Subcutaneous implantation of the constructs was performed in ten C57BL/6 6–8week old female mice (Charles River Laboratories, Wilmington, MA) after one week of culture either in growth medium or FGF2 supplemented medium in vitro according to the protocol that was approved by the Animal Ethics Committee of Central South University. Inhalation of isoflurane was used for general anesthesia when the surgery was performed. After making a midsagittal incision on the dorsa, a subcutaneous pouch was formed by blunt dissection on either side of the incision. The cell/PLLA scaffold constructs described above were implanted into each pouch subcutaneously. Three different samples were implanted in each mouse: (1) blank group (7 days preculture with a growth medium followed by no TGFβ1 (PLGA particle)) on the left side of the mouse, (2) TGFβ1 (PLGA particle) group [7 days preculture with a growth medium followed by TGFβ1 (PLGA particle)] in the top right position of the mouse, and (3) FGF2 to TGFβ1 (PLGA particle) group [7 days preculture with FGF2 medium followed by TGFβ1(PLGA particle)] in the bottom right position of the mouse. PLGA particles containing TGFβ1 were injected into the right side [see Fig. 4(b)]. Surgical staples were used to close the incisions after implantation. Four weeks postsurgery, CO₂ asphyxiation, and cervical dislocation were used to euthanize the mice, and the implanted constructs were collected for the subsequent histological, immunohistochemical, and extracellular matrix studies.

4% paraformaldehyde was used to fix the harvested cell/scaffold constructs (n = 7) at 4 °C overnight, and a graded ethanol series was used for dehydration afterward. After being cleared by xylene, the constructs were embedded in paraffin blocks. Xylene and a graded ethanol series were used to dewax the cross sections acquired at 10-μm thickness, and then, the sections were stained with Alcian Blue and Safranin O to test sulfated glycosaminoglycans (sGAG), von Kossa staining to evaluate ossification, and immunohistochemical staining (IHC) to detect the expression of chondrogenic markers. For immunohistochemistry, pepsin solution (Fisher Scientific) was used to treat the dewaxed sections for 20 min at room temperature to retrieve antigens. Polyclonal mouse anti-Col 2A1 (1:200) was purchased from Thermo Fisher and mouse anti-Col 1A1 (1:200) was purchased from Santa Cruz Biotechnology. An antimouse HRP-AEC Cell and Tissue Staining Kit (R&D Systems) was used to visualize all the above primary antibodies. An Olympus BX53 microscope (Olympus Corporation, Tokyo, Japan) was used to take histological images.

In each group, the harvested cell/scaffold constructs (n = 3) were homogenized as mentioned previously. Following this, papain solution [125 μg/ml papain, 100 mM phosphate, 10 mM ethylene-diaminetetraacetic acid (EDTA), 10 mM cysteine, and pH 6.3] was used to digest the mixture at 60 °C overnight. Measured the GAG content under absorbance at 525 nm with dimethylmethylene blue (DMMB, Sigma-Aldrich) assay. Bovine chondroitin sulfate (Sigma) was used as a standard. The total DNA amount was used to normalize the GAG content and Hoechst 33 258 assay was used to measure the DNA amount with calf thymus DNA (Sigma-Aldrich) as a standard.

SPSS was used for analysis. A mean value ± standard deviation is used for the presentation of all data. In order to ensure reproducibility, experiments were conducted in duplicate. A Student’s t-test was used between the two groups that fit the normal distribution. A p-value of less than 0.05 was considered statistically significant.

Mouse ADSCs were isolated from fat tissue and cultured in vitro. As Sox9 and collagen type II are commonly expressed genes in chondrocytes, and the presence of collagen type II protein indicates successful chondrocyte differentiation, these two genes were tested. The cells were treated with fibroblast growth factor 2 (FGF2) for one week, followed by transforming growth factor β1 (TGFβ1) for another week [FGF2 → TGFβ1 in Fig. 2(A)]. This sequential treatment induced ADSCs to differentiate into chondrocytelike cells that produced collagen type II. The gene expression levels of Sox9 and collagen type II were significantly higher in the FGF2 to TGFβ1 group compared to the reverse group (TGFβ1 → FGF2), the combination group (TGFβ1 + FGF2), and the TGFβ1 alone group [Fig. 2(A)(a, b)]. (P < 0.01 for FGF2 → TGFβ1 versus TGFβ1 → FGF2; P < 0.01 for FGF2 → TGFβ1 versus TGFβ1 + FGF2; P < 0.01 for FGF2 → TGFβ1 versus TGFβ1). FGF2 alone failed to yield chondrocytelike cells [Fig. 2(A)(a, b)]. The FGF2 to TGFβ1 treated ADSCs showed higher Sox9 and collagen II mRNA expression compared to untreated ADSCs [Fig. 2(A)(a, b)] (P < 0.01 for FGF2 → TGFβ1 versus blank). In order to eliminate the influence of cell numbers, the overall gene expression was normalized to the cell count, resulting in the calculation of average gene expression. The cell number in the FGF2 to TGFβ1 treated group did not significantly increase [Fig. 2(A)(c)] (P < 0.01 for FGF2 → TGFβ1 versus FGF2; P < 0.01 for FGF2 → TGFβ1 versus blank) and the average gene expression followed the same trend as the absolute gene expression results [Fig. 2(A)(d, e)] (P < 0.01 for FGF2 → TGFβ1 versus TGFβ1 → FGF2; P < 0.01 for FGF2 → TGFβ1 versus TGFβ1 + FGF2; P < 0.01 for FGF2 → TGFβ1 versus TGFβ1; P < 0.01 for FGF2 → TGFβ1 versus blank). Additionally, the FGF2 to TGFβ1 group exhibited collagen II protein expression in monolayer culture, consistent with the gene expression results [Fig. 2(B)]. In conclusion, the sequential application of FGF2 and TGFβ1 was necessary and sufficient to induce chondrogenic differentiation of mouse ADSCs in vitro.

To continue the in vitro studies, nanofibrous scaffolds were used to complete a 3D study. In order to promote tissue regeneration, the scaffolds were developed to provide pores with the suitable size and an overall structure that mimics the ECM of the native tissue.²⁹ In previous studies, these scaffolds were used to promote and conduct cell behaviors such as attachment, migration, proliferation, and differentiation. They can also be used to foster tissue formation and organization for targeted commitment toward the desired lineage (i.e., cardiac tissue, osteogenic, etc.).²⁹ A uniform porous structure was shown in the scaffolds [Fig. 3(A)(a)] and interconnectivity was shown among the pores [Fig. 3(A)(b, c)]. Pore walls consist of nanofibers with an average diameter between 100 and 200 nm [Fig. 3(A)(d)] and this structure simulates the collagen fibers of the ECM.³² Because of these properties, cells were uniformly retained in the pores 24h after seeding and attached to the pore walls of the scaffolds [Fig. 3(a)(e, f)], suggesting an environment conducive to regeneration.

Before in vivo application of the sequential release strategy from the nanofibrous scaffolds, it was ascertained whether the sequential delivery of FGF2 and TGFβ1 induced ADSC differentiation into cells which generated type II collagen on NF scaffolds. Porous NF scaffolds were seeded with ADSCs and placed in a cell culture plate. After a two-week culture in vitro, sequentially released FGF2 and TGFβ1 induced ADSCs showed type II collagen gene expression, suggesting that ADSCs were differentiated into chondrocytelike cells [Fig. 3(b)(a)]. Also, cell numbers in the FGF2 to TGFβ1 treated group did not significantly increase [Fig. 3(B)(b)] and in order to eliminate the influence of cell numbers, the average gene expression was calculated as well. The average collagen II gene expression showed the same trend as the absolute values for gene expression [Fig. 3(B)(c)]. Moreover, collagen II protein expression has the same trend in the nanofibrous scaffolds as shown through Western Blot [Fig. 3(C)].

Before the in vivo test, a release test in vitro was performed to determine the release profile of TGFβ1 from PLGA nanoparticles. The PLGA (TGFβ1) nanoparticles released curve showed sustainable factor release until day 30 [Fig. 4(A)]. The PLGA (TGFβ1) nanoparticles were used for in vivo test.

Finally, the precultured constructs were implanted subcutaneously into a model of mice in vivo to study the capacity of the scaffolds to improve chondrogenesis in seed cells and investigate whether the FGF2 to TGFβ1 sequential treatment showed positive effects in vivo. Safranin O and Alcian Blue stainings showed that the treatment with TGFβ1 (PLGA particle) led to partial GAG expression in the implanted constructs [Fig. 4(C), TGFβ1(PLGA)]. The constructs precultured with FGF2 medium and followed by induction with TGFβ1 (PLGA particle) successfully increased the expression of GAG in vivo [Figs. 4(C) and 4(F) → TGFβ1(PLGA)]. Likewise, immunohistochemistry staining indicated that the constructs treated with FGF2 to TGFβ1(PLGA particle) had increased Col2A1 expression after implantation [Fig. 4(C), Col2A1]. Additionally, FGF2 to TGFβ1 (PLGA particle) did not increase fibrotic or osteogenic tissue formation, which was proved by the test of Col1A1 [Fig. 4(C), Col1A1]. Also, in von Kossa staining, weak ossification was shown in the FGF2 to TGFβ1 (PLGA particle) treatment group. In contrast, significant ossification was observed in the blank and TGFβ1 (PLGA particle) groups [Fig. 4(C), von Kossa]. GAG quantification results [Fig. 4(D)] showed a significant difference between all three groups. As indicated in Fig. 4(D), the sequence FGF2 to TGFβ1 (PLGA particle) supported GAG expression that is much higher than TGFβ1 (PLGA particle) and blank groups after four weeks of testing.

This work describes a study determining the effect of the sequential release of FGF2 and TGFβ1 on nanofibrous constructs for chondrogenic differentiation of mouse ADSCs and the production of chondrogenic ECM deposition. The main objective was to identify the most effective sequence in the six different sequences tested that can stimulate chondrogenic ECM production and chondrogenic differentiation of mouse ADSCs on a nanofibrous scaffold. Among the six different sequences tested, it was observed that treating cells with FGF2 during the first week, followed by TGFβ1 for the duration of the experiment, yielded the best results for chondrogenic differentiation.

The increased chondrogenic phenotype of FGF2 to TGFβ1 treated samples was reflected by the increased expression of collagen type II and Sox9. By observing the other treatment groups, it was found that FGF2 treatment alone led to a decrease in gene expression of collagen type II and Sox9, indicating a certain inhibitory effect on chondrogenesis in mouse ADSCs. The expressions of collagen type II and Sox9 in the FGF2 to TGFβ1 treatment group were higher than those in the TGFβ1 alone group, revealing that this sequence may be the most effective in promoting chondrogenesis among all the groups. Although FGF2 was used in this sequence, the cell number did not increase significantly in the FGF2 to TGFβ1 treatment group, reflecting that the differences in cell expression was not due to differences in cell proliferation, but rather is affected by the coordination between FGF2 and TGFβ1. In order to eliminate the influence of cell numbers, the overall gene expression was normalized to the cell count, resulting in the calculation of average gene expression. Average gene expressions of collagen type II and Sox9 in the FGF2 to TGFβ1 group were higher than those in the TGFβ1 alone group. This further supports the notion that the induction of chondrogenic phenotype in FGF2 to TGFβ1 sequence is not caused by an increase in cell number.

Protein expression was consistent with gene expression. Collagen type II protein expression in the FGF2 to TGFβ1 group was obviously higher compared with that in the TGFβ1 alone group and the other experimental groups. Collectively, these results prove that the sequential treatment of FGF2 to TGFβ1 increased the corresponding levels of gene expression and protein expression in 2D culture.

The mechanism of chondrogenic differentiation promoted by the above sequences was discussed. During chondrogenesis, which is the process of cartilage formation, both fibroblast growth factor 2 (FGF2) and transforming growth factor beta 1 (TGFβ1) play important roles. However, their specific roles and the order in which they act can vary depending on the context and stage of chondrogenesis. FGF2 is involved in the early stages of chondrogenesis. It is known to promote the proliferation and differentiation of mesenchymal stem cells into chondrocytes, which are the cells responsible for cartilage formation. FGF2 helps to initiate the chondrogenic differentiation process by stimulating the expression of genes involved in cartilage development. TGFβ1, on the other hand, acts later in chondrogenesis and is involved in the maturation and maintenance of chondrocytes. It stimulates the production of extracellular matrix components, such as collagen and proteoglycans, which are essential for cartilage structure and function.

First of all, fibroblast growth factor 2 (FGF2) is known for its potent growth-stimulating properties, crucial for long-term proliferation and self-repair of mesenchymal stem cells. Existing literature suggests that FGF2 can induce differentiation of retinal epithelial cells into neuroretinal cells,³³ and is involved in the genesis and development of various embryonic epithelial cells.³⁴ It has been shown that FGF2 plays a crucial part in the development of mesenchymal stem cells through the extracellular signal-related kinase 1/2 signaling pathway.³⁵ 

Second, it has been shown from many previous researches in multimodel systems that TGFβs can stimulate early chondrogenic differentiation and inhibit the aberrant differentiation that leads to osteoarthritis. It is crucial for sustaining the homeostasis of articular cartilage.³⁶ TGFβ1 initiated and maintained chondrogenesis of bone marrow stromal cells through the differential activation of C-Jun N-terminal kinase (JNK), P38 and extracellular signal-regulated kinase-1 (ERK-1). Chondrogenesis involves the regulation of N-cadherin expression levels and it is regulated by mitogen-activated protein (MAP) kinase. MAP kinase may control intercellular interactions during cell polymerization and subsequent chondroblast differentiation by sequentially down-regulating and overground N-cadherin. TGFβ1-mediated activation of MAP kinase also regulates the gene expression of Wnt and Wnt-mediated signal transduction, thereby regulating N-cadherin expression and the formation of cellular adhesion complexes in the early stages of cartilage formation in bone marrow mesenchymal stem cells.^37,38

Finally, some scholars studied the signal transduction relationship between FGF2 and TGFβ1. Studies have shown that FGF2 regulates proliferation through interaction with its receptor and mitosis induced by MAPKs pathways.^39,40 Phosphorylation of ERK1/2 protein plays a part of importance in mitogen-mediated proliferation. TGFβ1 promotes the release of large amounts of FGF2 in cells, which induces cell proliferation through phosphorylation of p38 MAPK and JNK proteins. Further studies have shown that TGFβ1 can mediate the phosphorylation of protein MAPKs by stimulating the production of fibroblasts, which then release FGF2. These findings suggest that TGFβ1 and FGF2 have mutually reinforcing signaling pathways in signal transduction, in which phosphorylation of P38, MAPK, and JNK plays a critical role.

Based on the analysis in this study, the sequential use of growth factors FGF2 to TGFβ1 is believed to promote and enhance the chondrogenic differentiation effect. The underlying mechanism may involve these two growth factors acting on the same signaling pathway’s upstream and downstream components, leading to a synergistic effect.

Before entering the stage of animal experiment, the optimal growth factor sequences under the two-dimensional environment culture condition need to be verified under the three-dimensional culture condition. Whether the sequence use of the FGF2 and TGFβ1 method can make adipose stem cells toward the direction of chondrogenic differentiation was observed. The results of the three-dimensional cell experiment demonstrate that utilizing growth factors FGF2 and TGFβ1 sequentially yields a substantial increase in the expression of type II collagen genes and proteins, indicating a favorable cartilage phenotype. Furthermore, the expression levels of genes and proteins surpassed those observed when TGFβ1 was used as a standalone growth factor. These findings highlight the effectiveness of the sequential application of FGF2 and TGFβ1 in adipose stem cells during the three-dimensional cartilage differentiation process.

Previous literature has shown that the nanofibrous PLLA scaffold can provide sufficient mechanical properties to support cell growth and adhesion. Its components have been clearly introduced in the literature and proved to have good biocompatibility, and it has been applied in the repair of bone and heart tissues.²⁹ Scanning electron microscopy showed that the porous structure of the nanofiber scaffolds prepared in this paper was uniform and that there was high interconnectivity between the pores. The wall of each pore is made of nanofibers which imitate the natural collagen fibers found in the extracellular matrix. We analyze that the nanofiber PLLA scaffold mimics the extracellular matrix structure of natural tissue by providing appropriate size holes and overall structure, which is essential for tissue regeneration. Therefore, this scaffold can support the attachment of cells in the pores, promote cell adhesion to the pore wall in the pores, and provide a good environment for cell regeneration, so as to promote cell growth, differentiation, and finally, tissue repair. Other research reported that a porous nanofiber structure can support cell seeding, interaction between cells and scaffolds, and can cultivate and guide the cell adhesion, migration, growth, differentiation, promote new extracellular matrix formation, integrate to a particular organization, finally along the desired system development, so as to get the desired tissue (such as the heart tissue and ossification).²⁹ 

In histological staining, Safranin O and Alcian blue staining could make glycosaminoglycan positive in the matrix. In the early stage of cartilage formation, undifferentiated mesenchymal cells gathered and secreted the cartilage extracellular matrix.⁴¹ GAG could accumulate in the culture medium before type II collagen was expressed.⁴² The staining results showed a similar trend to that of gene and protein expressions. In all the groups, the experimental group which was pretreated with growth factor FGF2 and then treated with PLGA particles containing TGFβ1[FGF2 to TGFβ1 (PLGA particle) group] showed a significantly enlarged staining range and depth. Compared with the experimental group which was pretreated with a basic medium and then treated with PLGA particles containing TGFβ1 [TGFβ1 (PLGA particle) group], and the experimental group treated with a basic medium and then treated with unloaded PLGA particles (blank group), there were obvious advantages in the FGF2 to TGFβ1 (PLGA particle) group. This result reflects that the method of sequential use of growth factors FGF2 and TGFβ1 can promote chondrogenic differentiation in the mouse subcutaneous model.

Type II collagen, a main constituent of cartilage, is an important index for the detection of chondrogenic differentiation. The type II collagen expression in IHC staining also showed a similar trend to Safranin O and Alcian blue staining. On the contrary, immunohistochemical staining and silver nitrate (von Kossa) staining results suggested that the expression of type I collagen in the FGF2 to TGFβ1 (PLGA particle) group was the lowest among all experimental groups, which indicated that the sequential use of growth factors FGF2 and TGFβ1 did not significantly improve the effect of subcutaneous mineralization.

The results of the standardized extracellular matrix glycosaminoglycan test showed that the tissue in the FGF2 to TGFβ1 (PLGA particle) group expressed higher glycosaminoglycan deposition than the other experimental groups. At the same time, the amount of glycosaminoglycan deposition in the tissue of the TGFβ1 (PLGA particle) group was compared with that of the blank group. Glycosaminoglycan deposition in the TGFβ1 (PLGA particle) group increased significantly, indicating that the traditional treatment with growth factor TGFβ1 alone can promote the chondrogenic differentiation of adipose stem cells, which is consistent with previous studies.⁴³ The above results show that the sequential treatment of adipose stem cells with growth factors FGF2 and TGFβ1 is more effective than the traditional method in promoting adipose stem cells to chondrogenic differentiation.

The lack of an appropriate combination of a cell/controlled release system and an appropriate treatment strategy is the main challenge to restrict the clinical transformation of cartilage tissue repair. Therefore, the culture strategy provided in this paper is of great significance to prepare cartilage tissue by sequentially using growth factors FGF2 and TGFβ1 on a three-dimensional cell scaffold and a microsphere controlled release system to slowly release the growth factor from time and space, so as to achieve the purpose of cartilage tissue regeneration. At the same time, the research shows that the combination of the predesigned drug delivery system and the stent used in this paper can also combine the carriers with different release rates, and the carriers can contain different drugs. This method has the potential to regulate multiple signal molecules and can play an important role in different fields of tissue engineering.⁴⁴ 

While the findings in this paper exhibit promise, further experimental studies are required to precisely identify the specific location of this positive stimulus sequence within the signal pathway. Simultaneously, it holds great significance in assessing the growth of newborn cartilage in animal models.

The authors would like to thank the Natural Science Foundation of Hunan Province for Young Scholar of China (No. 2021JJ41032) for supporting this research project.

The authors have no conflicts to disclose.

Ethics approval has been obtained by the University Committee on Use and Care of Animals (UCUCA) at Central South University (Protocol No.: 202103743).

Yun-Qi Wu: Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Writing – original draft (equal). Jun Wang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.