Combinatorial Effect of Mesenchymal Stem Cells and Extracellular Vesicles in a Hydrogel on Cartilage Regeneration

Abstract

Background:

Mesenchymal stem cells (MSCs) are used for tissue regeneration due to their wide differentiation capacity and anti-inflammatory effects. Extracellular vesicles (EVs) derived from MSCs are also known for their regenerative effects as they contain nucleic acids, proteins, lipids, and cytokines similar to those of parental cells. There are several studies on the use of MSCs or EVs for tissue regeneration. However, the combinatorial effect of human MSCs (hMSCs) and EVs is not clear. In this study, we investigated the combinatorial effect of hMSCs and EVs on cartilage regeneration via co-encapsulation in a hyaluronic-acid (HA)-based hydrogel.

Methods:

A methacrylic-acid-based HA hydrogel was prepared to encapsulate hMSCs and EVs in hydrogels. Through in vitro and in vivo analyses, we investigated the chondrogenic potential of the HA hydrogel-encapsulated with hMSCs and EVs.

Results:

Co-encapsulation of hMSCs with EVs in the HA hydrogel increased the chondrogenic differentiation of hMSCs and regeneration of damaged cartilage tissue compared with that of the HA hydrogel loaded with hMSCs only.

Conclusion:

Co-encapsulation of hMSCs and EVs in the HA hydrogel effectively enhances cartilage tissue regeneration due to the combinatorial therapeutic effect of hMSCs and EVs. Thus, in addition to cartilage tissue regeneration for the treatment of osteoarthritis, this approach would be a useful strategy to improve other types of tissue regeneration.

Introduction

Osteoarthritis (OA), commonly referred to as degenerative arthritis, is one of the most common chronic diseases, affecting nearly 300 million people worldwide. Aging and frequent joint use favor OA, which is one of the main causes of joint pain in adult patients [1–4]. Another cause of OA is exposure of subchondral bone due to wear of articular cartilage [5, 6]. In addition, OA-related inflammation of the synovial membrane around the joint is accompanied by pain and cartilage destruction [7, 8]. Therefore, cartilage tissue in OA patients exhibits physiological and pathological changes, such as cartilage calcification, osteogenesis, synovial membrane inflammation, gradual loss of articular cartilage, etc. [9–14]. Unfortunately, the self-healing of cartilage is limited due to the lack of blood vessels around the cartilage [15]. Therefore, many researchers in this field have attempted to develop treatment techniques using external stimuli or the introduction of substances to regenerate cartilage [16–20].

Mesenchymal stem cells (MSCs) are multipotent progenitor cells that have the ability to self-renew and differentiate into various cell lineages such as adipocytes, osteoblasts, and chondrocytes [21]. MSCs can be isolated from various tissues such as bone marrow, adipose tissue, synovial membrane, and umbilical cord blood [22]. Recently, MSCs have been used to improve the progression of OA by implantation into cartilage tissue [23–25]. Implanted MSCs are not only directly involved in cartilage tissue regeneration by differentiating into chondrocytes, but also regulate the biological processes of cells residing in cartilage tissue, such as cell migration, proliferation, differentiation, extracellular matrix (ECM) synthesis, etc. [26, 27]. In addition, various secretomes of MSCs have been reported to induce anti-inflammatory effect and tissue regeneration by immune cells such as monocytes, macrophages, dendritic cells, T cells, and B cells in degenerated cartilage tissue [28, 29]. However, MSCs therapy for OA patients has some limitations, such as poor survival, rapid degradation, immune rejection, etc. [30]. To improve them, many studies have recently proposed a different strategy using extracellular vesicles (EVs) secreted by MSCs. Indeed, administration of MSC-EVs into OA cartilage reduced inflammation through immunomodulation, followed by enhanced regeneration of cartilage tissue [31–35].

EVs, spherical and 30 to 150 nm in diameter, are cellular vesicles composed of bilayer lipids and cellular components [36]. EVs act as paracrine mediators and contain a variety of cargoes such as lipids, mRNAs, miRNAs, proteins, and amino acids to exchange information and signaling molecules between cells [37]. In addition, EVs have the therapeutic advantage of having a long circulating half-life and posing no risk of tumor formation compared with living cells [38–40]. It is well known that MSC-derived EVs (MSC-EVs) regulate the biological activities of cells such as energy metabolism, cell proliferation, migration, ECM synthesis, and differentiation, which are involved in tissue regeneration [41, 42]. In addition, there are several studies on the regenerative effect of MSC-EVs on OA cartilage. MSC-EVs have also been found to stimulate immune cells as immunomodulators and suppressors in inflammatory environments [43–47]. However, EVs are not fully suitable for tissue regeneration for therapeutic purposes. For example, EVs are easily removed by phagocytosis and aggregation with biological molecules in tissues and blood vessels. Moreover, there are low therapeutic effects and side effects caused by off-targeting of EVs [48–51]. As described above, both MSCs therapy and EVs therapy for the treatment of OA have therapeutic effects and limitations at the same time. However, interestingly, there is no research on the combinatorial effect of MSCs and EVs in the treatment of OA. Therefore, in this study, we hypothesized that treating MSCs and EVs together may enhance the therapeutic effect while balancing the disadvantages of each.

Hyaluronic acid (HA) was chosen for the simultaneous introduction of MSC and EV because HA is one of the most important natural polymers that form the ECM of cartilage tissue and have biocompatible and biodegradable properties [52]. Moreover, a hydrogel based on HA is known to promote chondrogenic differentiation of MSCs and regeneration of cartilage tissue [53, 54]. In this study, we aimed to investigate the combinatorial effect of MSCs and EVs on chondrogenic differentiation and cartilage regeneration by encapsulating them in HA-based hydrogel (Fig. 1). For this purpose, we prepared a three-dimensional (3D) hydrogel system by preparing a methacrylated hyaluronic acid (MAHA) and then encapsulated MSCs and EVs into the hydrogel. In vitro and in vivo analyses showed that the HA hydrogel loaded with MSCs and EVs increased chondrogenic differentiation of MSCs and intensified cartilage tissue regeneration. Therefore, we believe that co-treatment of MSCs and EVs in HA hydrogel would be a useful therapeutic strategy for cartilage tissue regeneration and could be applied to other types of tissue regeneration.

Fig. 1.

Schematic illustration. hMSCs and EVs were encapsulated in MAHA hydrogel by UV-crosslinking, and then the hydrogel plug was implanted to a chondral defect site that was 2 mm wide and 2 mm long

Materials and methods

Isolation and culture of MSC

The hMSCs were isolated from human adipose tissue of the patients with the approval of the Ethics Committee of Dongguk University. Adipose tissue was treated with Dulbecco's phosphate-buffered saline (DPBS) containing 2% (v/v) penicillin/streptomycin (P/S; HyClone, Logan, UT, USA), washed and digested with 0.5 mg/mL collagenase type II (Sigma-Aldrich, St. Louis, MO, USA) in Dulbecco's minimal essential medium/low glucose (DMEM/LG; Hyclone) containing 1% (v/v) P/S at 37 °C for 40 min in a shaking incubator. Floating lipids were removed, and the digested tissues were filtered with a 40-μm cell strainer after centrifugation at 1 300 rpm for 10 min. Filtered tissues were washed with DMEM/LG by centrifugation at 1 300 rpm for 10 min. hMSCs were cultured in growth media [(DMEM/LG) supplemented with 10% (v/v) fetal bovine serum (FBS; Hyclone) and 1% (v/v) P/S] at 37 °C in a 5% CO₂ incubator.

Isolation and characterization of EVs

hMSCs were seeded at a density of 1 × 10⁴ cells/cm² on cell culture dishes and cultured in growth media at 37 °C in a 5% CO₂ incubator. After 12 h, the growth medium was replaced and thereafter the growth media were replaced every 2 d. After 7 d, hMSCs were washed with DPBS and incubated with serum-free media (DMEM/LG) for 24 h. Collected media containing EVs were centrifuged at 1,300 rpm for 3 min removing debris and the supernatant was filtered using a 0.2 μm syringe filter. Then, the filtered supernatant was purified through AMICON filtering system. First, a filtered supernatant was loaded onto the Amicon® Ultra-15 10 kDa Centrifugal Filter (Merck Millipore, Billerica, MA, USA) and centrifuged at 4,000 × g, 4 °C for 20 min. Second, DPBS was added onto the centrifugal filter with a supernatant to bring the volume to 15 mL and centrifugation was performed at 4,000 × g, 4 °C for 20 min. Third, the resulting solution was loaded onto the Amicon® Ultra-0.5 10 kDa Centrifugal Filter (Merck Millipore) and centrifuged at 13,000 rpm, 4 °C for 20 min. Fourth, DPBS was added onto the centrifugal filter with a resulting solution to bring the volume to 500 μL and centrifugation was performed at 13,000 rpm, 4 °C for 20 min. Finally, the EVs were isolated and resuspended with 100 μL of DPBS per 1.5 × 10⁶ cells/cm² of seeding density. The size distribution of EVs was confirmed using nanoparticle tracking analysis (NTA; LM10, Malvern Panalytical, Westborough, MA, USA) and dynamic light scattering (DLS; Nano ZS, Malvern Panalytical). The resuspended EVs’ solution was diluted 10 times in 1 mL of DPBS and measured at room temperature. The morphology of EVs was observed using energy-filtering transmission electron microscopy (EF-TEM; LIBRA 120, Carl Zeiss, Oberkochen, Germany). EVs of 1 × 10⁸ particles were loaded carbon film grid (Electron Microscopy Sciences, Hatfield, PA, USA) and incubated at room temperature for 3 min. The EVs loaded grid was washed with DW and removed remained DW. Afterward, a 1% (w/v) uranyl acetate solution in DW was treated on a grid for negative staining for 3 secs and washed out using DW. After drying at room temperature for 5 min, EVs were visualized EF-TEM image.

Preparation of methacrylated HA hydrogel encapsulated with hMSCs and EVs

Methacrylated hyaluronic acid (MAHA) was fabricated conjugation methacrylic anhydride (Sigma-Aldrich) and hyaluronic acid (Bioland, Mainz, Germany; MW 500 kDa). Hyaluronic acid was dissolved in distilled water (1% w/v; D.W.). Methacrylic anhydride was added to the hyaluronic acid solution to yield MAHA. The mixture was adjusted to about pH 8.0 with 5 N NaOH for 24 h at 4 °C under constant stirring in the dark. Then the mixture was dialyzed for 3 d with D.W. to remove remain methacrylic anhydride. The MAHA solution was filtered with a 0.45 μm syringe filter and frozen at − 80 °C. Then the frozen MAHA solution was lyophilized by a freeze dryer. The methacrylation ratio of MAHA was analyzed with proton nuclear magnetic resonance spectroscopy (¹H-NMR; 500 MHz FT-NMR Spectrometry, Bruker, Billerica, MA, USA). The ratio was calculated from the relative peak integration ratio of the acryl protons (peaks at ~ 6.1, 5.6, and 1.85 ppm) and methyl protons of HA.

To prepare MAHA hyaluronic acid hydrogel, MAHA was first dissolved in 1% DPBS. To make a hydrogel of 0.5% hyaluronic acid, the remaining volume was added with DPBS and finally photo-initiator that Irgacure (I2959) to make a mixed solution. In the case of I2959, the final concentration is made to be 0.2%. To compare the stiffness according to EVs’ containing, the shear stress of MAHA hydrogel was measured by rheometer (HAKKE MARS, Thermo Electron Corporation, Waltham, MA, USA). The oscillatory amplitude sweep measurements were conducted between 0.001 and 10 Pa and the oscillatory frequency sweep measurements were carried out during the frequency of 0.05–20 Hz in the oscillation mode, OSC. The Gap distance was 0.5 mm at 37 °C.

In the case of cells, 8 × 10⁴ cells were loaded in 40 μL MAHA solution. In the case of EVs, 8 × 10⁷ EVs were loaded in 40 μL MAHA solution. After all the mixture was mixed, photo-crosslinking was performed through UV-C radiation (365 nm, 60 mW/cm², Sei Myung Vactron Co., Ltd., Bucheon, Korea). The reaction was allowed to proceed for 10 secs to form a hydrogel. The polydimethylsiloxane (PDMS) mold is used to adjust the size and shape of the hydrogel. For hydrogel culture, cultivated in 24 well culture plate and changed 1 mL of chondrogenic media per 2 d (Table 1). To evaluate cell viability, hASCs encapsulated HA gel was placed on cell culture plate and incubated at 37 °C in a humidified atmosphere of 5% CO₂. Afterward, live and dead cells in each group were detected using Cytation 3 (Biotek Instruments, Winooski, VT, USA) after staining with Calcein AM (Invitrogen, Carlsbad, CA, USA)/ethidium homodimer-1 (Invitrogen).

Table 1.

Composition of chondrogenic medium

Component	Company	Amount
DMEM/high glucose (DMEM/HIGH)	Cytiva
Fetal bovine serum	Cytiva	10% (v/v)
100X insulin transferrin selenium A	Gibco	1×
L-ascorbic acid	Sigma-Aldrich	50 μg/mL
Dexamethasone	Sigma-Aldrich	100 nM
Transforming growth factor beta	Sigma-Aldrich	10 ng/mL
Penicillin/streptomycin	Cytiva	1% (v/v)

Chondrogenic medium was changed every 2 d

Quantitative real-time polymerase chain reaction (qRT-PCR) analysis

To isolate RNA, MAHA hydrogels were frozen with liquid nitrogen and broken with a pestle homogenizer in 200 μL TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA). The broken samples were added 800 μL TRIzol and 200 μL chloroform. After vortexing the sample and centrifuging at 13,000 rpm for 20 min at 4 °C, the RNA supernatant was mixed with isopropyl alcohol at the same volume. After centrifuging in the same conditions, the pellet was washed with 75% ethyl alcohol. The pellet was dried and dissolved in RNA free water (Life Technologies, Carlsbad, CA, USA, AM9930). RNA quantification was performed through Cytation3 (Biotek). cDNA was synthesized with a TAKARA cDNA synthesis kit (Takara, Shiga, Japan) (Table 2). qRT-PCR was performed in a final reaction mixture (20 μL) containing 10 μL of 2 × Power SYBR Green PCR Master Mix (Applied Biosystems, Warrington, UK), 1 μL of gene-specific primers (5 μM), 5 μL of template, and 3 μL of nuclease-free water using a Step One Plus Real-Time PCR system (Applied Biosystems) with the following conditions: an initial denaturation at 95 °C for 1 min, followed by 45 cycles of 95 °C for 15 s, 56 °C for 15 s, and 72 °C for 15 s, and a final extension at 72 °C for 5 min.

Table 2.

Primers for qPCR

Gene	Forward primer (5′–3′)	Reverse primer (5′–3′)
ALK5	CGA CGG CGT TAC AGT GTT TCT	CCC ATC TGT CAC ACA AGT AAA ATT G
TGFbR2	AGG ACT GCC CAT CCA CTG AGA CA	GTG GAA ACT TGA CTG CAC CGT TGT
SOX9	GTA CCC GCA CTT GCA CAA C	TCT CGC TCT CGT TCA GAA GTC
COL2A1	CAC GTA CAC TGC CCT GAA GGA	CGA TAA CAG TCT TGC CCC ACT T
ACAN	GCC TGC GCT CCA ATG ACT	ATG GAA CAC GAT GCC TTT CAC
18s rRNA	GTA ACC CGT TGA ACC CCA TT	CCA TCC AAT CGG TAG TAG CG

Western blot analysis

Western blotting was performed by using ‘simple western’ (WES) through the manufacturer’s protocol (Protein Simple, San Jose, CA, USA). Before WES, every sample (EVs, hMSC, and MAHA hydrogel) was washed using DPBS and was lysed. For lysis of EVs, 10 μL of EVs solution was added 190 μL of 1 × RIPA buffer (Sigma-Aldrich). To prepared whole cell lysates (WCL), cultured hMSCs were collected by trypsinization and washed using DPBS. Then, the hMSCs were frozen with liquid nitrogen and added 50 μL of 5 × RIPA buffer. MAHA hydrogel was cultured in chondrogenic induction media at 1 h and 14 d. The hydrogel was washed three times with DPBS and frozen with liquid nitrogen. Then the hydrogel was added 50 μL of 5 × RIPA buffer. Subsequently, every sample was broken using a pestle homogenizer. Extracts were collected and centrifuged at 13,000 rpm for 20 min. Finally, the supernatant was collected, and protein samples were prepared. The concentration of total protein was quantified by Pierce BCA protein assay kit (Thermo Fisher Scientific). The protein samples, blocking buffer, primary and secondary antibodies, chemiluminescent substrate reagent and wash buffer were loaded on the WES assay plate. Then the WES assay plate was placed in the WES instrument, and western blotting was performed automatically. Primary and secondary antibodies were purchased from commercial company for the protein detection (Table 3).

Table 3.

Primary and secondary antibodies used for the western blotting (WES) and immunofluorescence (IF) analyses

Purpose	Name of antibody	Company	Cat. no.
WES	Anti-pSMAD2	Cell signaling technology	3108
	Anti-SMAD2/3	Cell signaling technology	9520
	Anti-phospho-p38 MAPK	Cell signaling technology	4631S
	Anti-p38 MAPK	Cell signaling technology	9212S
	Anti-phospho-p44/42 MAPK	Cell signaling technology	4377S
	Anti-p44/42 MAPK	Cell signaling technology	9102S
	Anti-phospho-SAPK/JNK	Cell signaling technology	4668S
	Anti-JNK2	Cell signaling technology	9258S
	Anti-SOX9	Cell signaling technology	82630
	Anti-ACAN	Abcam	ab3778
	Anti-collagen type II	Milipore	MAB8887
	Anti-β-actin	Abm	G043
	Anti-mouse secondary antibody	Bio-techne	042-205
	Anti-rabbit secondary antibody	Bio-techne	042-206
IF	Anti-chondroitin sulfate	Abcam	ab11570
	Anti-collagen type II	Millipore	MAB8887
	Alexa Fluor 488 goat anti-mouse IgG (H + L)	Invitrogen	A11001
	Texas Red-X goat anti-mouse IgG (H + L)	Invitrogen	T862
	Anti-SOX9	Millipore	AB5535

Immunofluorescence (IF) staining

hMSCs encapsulated MAHA hydrogels were washed with DPBS three times and fixed with 4% paraformaldehyde (Biosesang, Seongnam, Korea) at room temperature for 1 h. The fixed hydrogel was washed with DPBS, and then permeabilized with 0.3% (v/v) Triton X-100 in DPBS (PBS-T) at room temperature for 30 min. Then, the hydrogel was blocked with 1% (w/v) BSA in PBS-T at room temperature for 1 h. Chondroitin sulfate and Type II collagen were diluted 1:100 with 1% (w/v) BSA in PBS-T, incubated at 4 °C for 24 h, and washed with PBS-T three times. The primary antibodies were purchased from a commercial company for IF staining (Table 3) Fluorescein-conjugated secondary antibodies were diluted 1:200 and incubated at room temperature for 2 h and washed with PBS-T three times. The secondary antibodies were purchased from a commercial company for IF staining (Table 3).

In vivo study using chondral defect model in rat

All animal experiments were performed in accordance with all relevant ethical regulations and used approved study protocols. For the in vivo studies, we used a sample size of five per group based on our published studies using a preclinical rat model. We used male animals to exclude variability from the influence of female hormones in bone tissue regeneration. In addition, the details of experimental groups were blinded to the investigators for the biochemical and histological evaluations. Then, we investigated using 8 weeks old male SD rat chondral defect model to confirm the synergistic effect hMSCs and EVs for cartilage regeneration. Defects of 2 mm in diameter were generated on the knee using a drill and the hydrogel was implanted into the defective site using forceps. After 8 weeks, all the experimental animals were sacrificed. And the left knee joints were harvested and then fixed in 4% paraformaldehyde. To confirm cartilage regeneration, we prepared a paraffin block section slide of the knees and performed histological staining. The fixed tissues were treated with a decalcification solution (BBC Biochemical, Mount Vernon, WA, USA), dehydrated with ethanol and xylene, and embedded in paraffin. The paraffinized samples were sliced into 5 μm sections using a semi-motorized rotary microtome (Leica, Wetzlar, Germany, RM2245). The specimens were de-paraffinized with xylene and ethanol and washed with DPBS, then stained using safranin-O or hematoxylin and eosin (H&E) staining solution. COL2 and SOX9 protein were observed by IF staining (Table 3). The stained sections were washed with DPBS three times and visualized using a microscope (Olympus, Tokyo, Japan).

Statistical analysis

All the statistical analyses were performed using Prism ver. 8.0 (Graphpad Software inc., San Diego, CA, USA). A one-way ANOVA test using Tukey’s-multiple comparison post-test was performed to compare the samples. The data were considered statistically significant if p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Results

Characterization of EVs and MAHA hydrogels

The EF-TEM image of EVs showed that the EVs had spherical morphology and a size of approximately 150 nm (Fig. 2A). The expression of representative EV marker proteins (CD9, CD63, CD81, and HSP70) on the isolated EVs was observed by western blot analysis (Fig. 2B). According to NTA (Fig. 2C). The size distribution of EVs was measured in the range of 100–200 nm. These results indicated that EVs with intrinsic characteristics were successfully isolated from hMSCs.

Fig. 2.

Isolation of EVs and preparation of EVs-encapsulated HA hydrogel. HA hydrogel was prepared by UV-crosslinking with MAHA. A EF-TEM image of EVs. B The protein expression of EV marker genes (CD9, CD63, and HSP70) and house-keeping gene GAPDH of whole cell lysate (WCL) and EVs. C Size distribution of EVs measured by NTA. D ¹H-NMR spectra of MAHA and HA. E The shear stress test of HA hydrogel in the presence of EVs. Data are mean ± S.E.M. (unpaired t test, n.s.; no significance). F Photograph of HA hydrogel in the presence of EVs

In this study, we aimed to encapsulate hMSCs and EVs in HA hydrogel and then investigate their combinatorial effect on cartilage regeneration through in vitro and in vivo studies. To this end, first, MAHA was prepared through covalent conjugation of methacrylic anhydride to HA, and the successful methacrylation was confirmed by the double bond peak of the acryl group (assigned “a”) in ¹H-NMR spectrum (Fig. 2D). Next, the HA hydrogel containing hMSCs and/or EVs was prepared by mixing hMSCs and/or EVs to MAHA solution right before UV-C radiation. Through photo-crosslinking, the hydrogel with diameter of 4 mm diameter and height of 2 mm height was fabricated and the overall morphology of the hydrogel was observed by a photograph (Fig. 2). Regarding the physical properties of the HA hydrogel (HA gel) and EVs-loaded HA hydrogel (HA gel + EVs), shear stress representing hydrogel stiffness did not show a significant difference between them (Fig. 2E). This demonstrates that EVs encapsulation did not affect the stiffness and morphology of HA hydrogel.

Chondrogenic effect of hMSCs without/with EVs in MAHA hydrogel

To investigate the combinatorial effect of hMSCs and EVs in HA hydrogel for chondrogenic differentiation, we prepared two groups which are hMSCs-loaded HA hydrogel (hMSCs HA) and hMSCs-loaded HA hydrogel with EVs (hMSCs + EVs HA). Next, we examined the chondrogenic differentiation of hMSCs and signaling pathways in the presence of EVs by qPCR, western blotting, and immunofluorescence staining (IF). The UV-crosslinking was performed to fabricate HA hydrogel. The morphology and viability of hMSCs did not show a big difference between hMSCs HA hydrogel and hMSCs + EVs HA hydrogel (Fig. 3A) demonstrating additional encapsulation of EVs did not affect the morphology and viability of hMSCs. The mRNA levels of transforming growth factor β receptor type I (ALK5), transforming growth factor β receptor type II (TGFbR2), and SPY-Box transcription factor 9 (SOX9), which are related to chondrogenic differentiation were analyzed at 1 h after the induction of chondrogenic differentiation. As shown in Fig. 3B, the gene expressions of ALK5, TGFbR2, and SOX9 in a group of hMSCs + EVs HA were 1.6-, 1.2-, and 1.5-fold higher than hMSCs HA, respectively. Furthermore, after 14 d of the chondrogenic differentiation, the mRNA expression of the chondrogenic marker genes (SOX9, Collagen type II alpha 2 chain; COL2A1 and Aggrecan; ACAN) were increased 2.4-, 2.7-, and 2.0-fold higher in the group of hMSCs + EVs HA (Fig. 3C). These observations suggest that EVs would be able to stimulate chondrogenic differentiation of hMSCs in HA hydrogel.

Fig. 3.

Chondrogenic differentiation of hMSCs with EVs in HA hydrogel. hMSCs and EVs were co-encapsulated in HA hydrogel. A Live and dead assay of the hMSCs-encapsulated HA hydrogel (hMSCs HA) and the hMSCs and EVs co-encapsulated HA hydrogel (hMSCs + EVs HA). Right panel indicates quantitative data of fluorescence images. Data are mean ± S.E.M. (unpaired t-test, n.s.; no significance). B The receptors of TGF-β, ALK5, and TGFbR2, and transcription factor SOX9 mRNA expression at 1 h after the differentiation. Data are mean ± S.E.M. (unpaired ttest, *p < 0.05, **p < 0.01, ****p < 0.0001). C The mRNA expression of cartilage ECM, ACAN, and COL2, at 14 d after the differentiation. Data are mean ± S.E.M. (unpaired t test, *p < 0.05, ***p < 0.001, ****p < 0.0001). D The phosphorylation of signal molecules, SMAD2, p38, ERK, and JNK at 1 d after the differentiation, and protein expression of chondrogenic marker genes, ACAN, COL2, and SOX9 at 14 d after the differentiation. E IF staining of CS and COL2 at 14 d after the differentiation. The graph of right panel indicates the quantitative data of fluorescence intensity in images. Data are mean ± S.E.M. (unpaired t test, *p < 0.05, **p < 0.01)

Next, we examined the phosphorylation of Mothers against decapentaplegic homolog2 (SMAD2), Mitogen-activated protein kinase 11 (p38), extracellular-regulated kinase (ERK), and c-Jun N-terminal kinase (JNK) which are components of signaling pathways related to chondrogenic differentiation. The phosphorylation of signal molecules was investigated at 1 d after chondrogenic differentiation. Indeed, phosphorylation of SMAD2 and JNK were increased in the presence of EVs, respectively. However, phosphorylation of p38 and ERK have no big difference in the presence of EVs (Fig. 3D). Next, we checked ACAN, COL2, and SOX9 protein levels after 14 d of chondrogenic differentiation. ACAN, COL2, and SOX9 proteins were upregulated in the presence of EVs. The protein levels of chondroitin sulfate (CS) and COL2 were observed and quantified by IF staining at 14 d after the differentiation. The fluorescence intensity of IF images of CS and COL2 indicate that protein expression of CS and COL2 proteins were increased 1.46- and 1.50-fold in the group of hMSCs + EVs HA (Fig. 3E). These results demonstrate that the chondrogenic properties of hMSCs in HA hydrogel can be upregulated in the presence of EVs. Based on in vitro data, it is highly possible that co-encapsulation of hMSCs and EVs in HA gel also increases in vivo cartilage tissue regeneration of damaged cartilage due to the combinatorial effect of hMSCs and EVs.

Combinatorial effect of hMSCs and EVs for the cartilage tissue regeneration

Next, we investigated the in vivo cartilage tissue regeneration effect by using a chondral defect rat model. The experimental groups were compared with a sham-operated group (Sham; n = 3), a HA hydrogel-implanted group (HA; n = 5), a hMSCs-loaded HA hydrogel-implanted group (hMSCs HA; n = 5), and a hMSCs and EVs co-loaded HA hydrogel-implanted group (hMSCs + EVs HA; n = 6). First, the HA hydrogels of each group were prepared as described above and implanted into the defect site (2 mm diameter and 2 mm height). To evaluate the efficiency of cartilage regeneration, we performed histological analysis by safranin-O and H&E staining (Fig. 4A). The hMSCs + EVs HA group showed apparent tissue filling in defect lesions and similar cartilage properties with the Sham group. As shown in Fig. 4B, cartilage repair was evaluated using histological cartilage repair grading system of the ICRS (International Cartilage Repair Society) (Fig. 4B). As the result, the chondrogenic properties of the hMSCs + EVs HA group was significantly higher than those of HA and hMSCs HA group. In addition, the hMSCs + EVs HA group exhibited significantly higher expressions of COL2 and SOX9 compared with the HA group and hMSCs HA group, which is confirmed by the quantification of IF staining image (Fig. 4C, D). Therefore, we concluded that the co-encapsulation of hMSCs and EVs in HA hydrogel has a combinatorial effect on in vivo the cartilage regeneration following an increase in chondrogenic differentiation of MSCs.

Fig. 4.

Histological analysis of the reconstructed cartilage in chondral defect rat model at 8 weeks after implantation. A Safranin-O and Fast green and H and E staining of the Sham, HA, hMSCs HA, and hMSCs + EVs HA groups. B Quantitative data of cartilage regeneration by ICRS scoring system. Data are mean ± S.E.M. (one-way ANOVA, ***p < 0.001, ****p < 0.0001). C IF staining of COL2 and SOX9. The white dot line indicates the margin of cartilage tissue. D) Quantitative data of fluorescence intensity in panel C measured by image J. Data are mean ± S.E.M. (one-way ANOVA, n.s.; no significance, *p < 0.05, **p < 0.01)

Discussion

For successful cartilage regeneration, it is necessary to enhance cell viability, differentiation, and production of cartilage ECM in damaged cartilage site [5]. In addition, suppression of the inflammatory environment is essential for the tissue regeneration because inflammatory cytokines can induce cell apoptosis and degradation of cartilage ECM [6, 8, 21]. Many studies have been attempted to directly transplant cells or cell-derived biomolecules into damaged cartilage tissue for cartilage regeneration and suppression of inflammatory environment [23, 24]. In particular, it has been reported that the implantation of MSCs or MSC-EVs into damaged cartilage site stimulate the cartilage remodeling by increasing the activity of recipient cells and reducing the production of inflammatory cytokine by immunomodulation [30, 31, 36, 41, 45–47]. However, some reports indicate that MSCs implantation to damaged cartilage showed lower survival and differentiation into chondrocyte. The use of EVs for tissue regeneration also has several huddles such as low efficiency of tissue regeneration. To overcome these deficiencies, new strategies using MSCs and EVs are required for cartilage regeneration.

In this study, we prepared a 3D hydrogel system from methacrylated hyaluronic acid and co-encapsulated hMSCs and EVs in it for efficient cartilage tissue regeneration. In this study, we hypothesized that co-encapsulation of hMSCs and EVs in HA hydrogel will maximize their strength and overcome their deficiencies. HA is one of the cartilage ECM and has known as a useful factor to regenerate damaged cartilage by supplying the ECM. Because HA can interact with MSCs and EVs through the CD44 binding, they can be successfully encapsulated in HA hydrogel [55]. Based on these features, we suggested that the encapsulated hMSCs could be stimulated effectively by the neighboring encapsulated EVs in HA hydrogel during the chondrogenic differentiation. Indeed, co-encapsulation of hMSCs and EVs in HA hydrogel significantly increased the chondrogenic properties in in vitro and in vivo as well.

TGF-β is a component of the chondrogenic medium and necessary to the chondrogenic differentiation of hMSCs. The first step of the chondrogenic differentiation is the interaction with TGF-β and TGF-β receptors on the cell surface (ALK5 and TGFbR2). Therefore, we investigated that the expression of ALK5, TGFbR2, and downstream signal molecules such as SMAD2, p38, ERK, and JNK [56, 57]. After differentiation induction, mRNA expression of ALK5 and TGFbR2 and the phosphorylation of SMAD2 and p38 were distinctly higher in the hMSCs + EVs HA group than those in HA hydrogel (Fig. 3B, D). In addition, the expression of chondrogenic transcription factor (SOX9) and ECMs (ACAN, COL2, and CS) were also higher in the hMSCs + EVs HA group than that in HA hydrogel (Fig. 3B–E). Histological analysis by safranin-O, H&E, and IF staining of COL2 and SOX9 also confirmed that co-encapsulation of hMSCs and EVs in HA hydrogel is able to increase cartilage tissue regeneration of damaged tissue.

As described above, we found that the EVs in HA hydrogel intensified the chondrogenic differentiation of hMSCs in the same HA hydrogel. However, it is still not clear which factors in EVs are mainly involved in stimulating hMSCs for cartilage regeneration. As potential factors, we could presumably suggest several candidates such as chondrogenic transcription factors, miRNA, anti-inflammatory cytokines and so on. The mRNA or protein of SOX9 is secreted from cells by EVs and transferred into recipient cells to induce the chondrogenic properties of the cells [58]. The delivery of SOX9 into the cells induces the expression of cartilage ECM such as COL2, ACAN, and CS. In addition, the cartilage ECM increases the release of secretomes, which leads homing effect of cells neighboring damaged cartilage tissue for regeneration [26]. Therefore, the transcription factors in EVs could be the main regulator for the stimulation of MSCs. On the other hand, miRNA, inside of MSC-EVs, is also involved in the cartilage regeneration. For example, it was reported that miR-140 and miR-17 enhance anabolism and inhibit catabolism of cartilage ECM. In addition to them, it was known that many miRNAs are engaged in cartilage regeneration [59–61]. Moreover, miRNA play a role in inhibiting mitochondrial dysfunction, so they can inhibit apoptosis and accelerate cell activities following cell survival and differentiation [62–64]. Aspect of immunomodulation, it was reported that MSCs can influence immune cells that regulate inflammation through carrying cargoes of EVs containing anti-inflammatory cytokines [28, 65, 66] For example, macrophages exist as m1 and m2 phenotypes under inflammation conditions. It is known that when m2 polarization increases by anti-inflammatory cytokines, it induces releasing certain regulatory cytokines and exerting an immunomodulatory effect on cartilage regeneration.

In this study, we suggest a 3D hydrogel system containing both hMSCs and EVs for effective cartilage tissue regeneration because EVs exerted beneficial stimulations on MSCs. We believe the co-encapsulation of cells and EVs in a hydrogel system could be a useful strategy applied to various types of tissue regeneration organized with various types of cells and EVs such as immune cells, muscle cells, retinal cells, fibroblasts, cardiocytes and so on.

Declarations

Conflict of interest

The authors have no financial conflicts of interest.

Ethical statement

The study protocol was approved by the institutional review board of Dongguk University (IRB No. DUIH2020-03-012-022). Informed consent was confirmed by the IRB. The animal studies were performed after receiving approval of the Institutional Animal Care and Use Committee (IACUC) in Dongguk University (IACUC approval no. IACUC-2021-018-2).