Exosomes Derived from Hypoxia-Cultured Human Adipose Stem Cells Alleviate Articular Chondrocyte Inflammaging and Post-Traumatic Osteoarthritis Progression

Abstract

Osteoarthritis (OA) is the most common age-related degenerative joint disease. Inflammaging, linking inflammation and aging, is found in senescent cells with the secretions of matrix-degrading proteins and proinflammatory cytokines. The senescence-associated secretory phenotype (SASP) plays a very important role in OA progression. However, there remains no effective way to suppress OA progression, especially by suppressing inflammaging and/or the chondrocyte SASP. Recent studies have shown that exosomes derived from hypoxia-cultured BMSCs can regenerate cartilage in OA animal models. Some reports have further indicated that exosomes secreted from MSCs contribute to the efficacy of MSC therapy in OA. However, whether hypoxia-cultured ADSC-secreted exosomes (hypoxia-ADSC-Exos) can alleviate the chondrocyte SASP or OA progression remains unclear. Accordingly, we hypothesized that hypoxia-ADSC-Exos have a beneficial effect on the normal functions of human articular chondrocytes (HACs), can attenuate the SASP of OA-like HACs in vitro, and further suppress OA progression in rats. Hypoxia-ADSC-Exos were derived from ADSCs cultured in 1% O₂ and 10% de-Exo-FBS for 48 h. The molecular and cell biological effects of hypoxia-ADSC-Exos were tested on IL1-β-induced HACs as OA-like HACs in vitro, and the efficacy of OA treatment was tested in ACLT-induced OA rats. The results showed that hypoxia-ADSC-Exos had the best effect on GAG formation in normal HACs rather than those cultured in normoxia or hypoxia plus 2% de-Exo-FBS. We further found that hypoxia-ADSC-Exos alleviated the harmful effect in OA-like HACs by decreasing markers of normal cartilage (GAG and type II collagen) and increasing markers of fibrous or degenerative cartilage (type I or X collagen), matrix degradation enzymes (MMP13 and ADAMT5), and inflammatory cytokines (TNFα and IL-6). More importantly, intra-articular treatment with hypoxia-ADSC-Exos suppressed OA progression, as evidenced by the weight-bearing function test and cartilage GAG quantification in ACLT rats. Moreover, through NGS and bioinformatic analysis, seven potential miRNAs were found in hypoxia-ADSC-Exos, which may contribute to regulating cellular oxidative stress and attenuating cell senescence. In summary, we demonstrated that hypoxia-ADSC-Exos, carrying potent miRNAs, not only improve normal HAC function but also alleviate HAC inflammaging and OA progression. The results suggest that hypoxia-ADSC-Exo treatment may offer another strategy for future OA therapy.

Keywords: SASP (senescence-associated secretory phenotype); articular chondrocytes; exosomes; hypoxia; inflammaging; micro ribonucleic acid (miRNA); osteoarthritis (OA).

Figure 1

Exosomes induced by normoxia/hypoxia and normal/low serum culture conditions in cultured ADSCs. (A) Transmission electron microscopy (TEM) image of ADSC-Exos (scale bar: 50 nm). (B) Western blot analysis of the protein levels of ADSC-Exo markers in ADSCs and ADSC-Exos was performed, including CD9, CD63, CD81, ALIX, TSG101, and α-tubulin levels, which were used as negative controls. (C) Nanoparticle tracking analysis (NTA) evaluating the ADSC-Exo size and (D) concentration after four conditions induced ADSCs, including 20% O₂ (normaxia) + 2% de-Exo-FBS, (2 N), 1% O₂ (hypoxia) + 2% de-Exo-FBS (2H), 20% O₂ (normaxia) + 10% de-Exo-FBS (10 N), and 1% O₂ (hypoxia) + 10% de-Exo-FBS (10H) for 2 days. Data represent the mean exosome size, and each column represents the mean ± SEM of six replicate cultures. *, p < 0.05 compared to 2 N induction. (E) 10H-ADSC-Exo uptake by HACs. The CM-DiI-labeled exosomes (red fluorescence-stained) were taken up by cell tracker-labeled chondrocytes (green fluorescence-stained; DAPI-stained nucleus: blue fluorescence) on Day 12.

Figure 2

We compared different culture conditions and found the optimal function of exosomes derived from ADSCs on chondrocytes. (A) Alcian blue staining and (B) DMMB assay analysis of GAG synthesis in HACs treated with normal culture medium (as a positive control), basal medium (only DMEM), and 4-condition-induced ADSC-Exos (2 N, 2H, 10 N, 10H) supplemented in basal medium. Data represent the relative GAG/DNA content, and each column represents the mean ± SEM of six replicate cultures. ***, p < 0.001 compared to the basal control culture. (C) MTS assay evaluating chondrocyte viability after treatment with 10H-ADSC-Exos (10⁷~10⁹ particles/mL) for 5 days. Data represent the relative cell viability, and each column represents the mean ± SEM of six replicate cultures. There was no difference compared with the basal control culture (Ctrl). (D) GAG synthesis of HACs treated with different doses of 10H-ADSC-Exos (10⁷~10⁹ particles/mL) by Alcian blue staining and (E) DMMB assay. Data represent the relative GAG/DNA content, and each column represents the mean ± SEM of six replicate cultures. **, p < 0.01, compared to the basal control culture.

Figure 3

Hypoxia-induced ADSC-Exos enhanced cartilaginous matrix synthesis in IL-1β-induced OA-like inflammaging articular chondrocytes. (A) The mRNA expression of the normal functional genes aggrecan, col2a1, and PTHrP in vitro in HACs treated with IL1-β with or without ADSC-Exos (10⁹ particles/mL) for 12 days. The mRNA expression of GAPDH in HACs was used as the internal control. (B) Type II collagen ELISA was performed to quantify the protein level when HACs were treated with IL1-β and with or without ADSC-Exos (10⁷~10⁹ particles/mL) for 12 days. (C) Alcian blue staining ((a) Ctrl, (b) IL-1β, (c) IL-1β+ADSC-Exo 10⁷ particles/mL, (d) IL-1β+ADSC-Exo 10⁸ particles/mL, (e) IL-1β+ADSC-Exo (10⁹ particles/mL) and (D) DMMB assay analysis of GAG synthesis in HACs treated with IL1-β and with or without ADSC-Exo (10⁷~10⁹ particles/mL) for 12 days. Each column represents the mean ± SEM of six replicate cultures. Data from each group were compared with those of the IL-1β only group, *, p < 0.05, **, p < 0.01, ***, p < 0.001.

Figure 4

Hypoxia-induced ADSC-Exos inhibited fibrous and degenerated cartilage in IL-1β-induced OA-like articular chondrocytes. (A) Fibrous cartilage genes, col1a1, and (B) degeneration marker gene, col10a1, mRNA expression in vitro in HACs treated with IL1-β and with or without ADSC-Exos (10⁹ particles/mL) for 12 days. GAPDH mRNA was used as the internal control. (C) Type I and (D) type X collagen ELISAs were performed to quantify the protein levels when HACs were treated with IL1-β and with or without ADSC-Exos (10⁷~10⁹ particles/mL) for 12 days. Each column represents the mean ± SEM of six replicate cultures. Data from each group were compared with those of the IL-1β only group, *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 5

Hypoxia-induced ADSC-Exos suppressed inflammatory cytokines and degradation enzymes in inflammaging articular chondrocytes. (A). Inflammatory-related gene (CEBPβ, COX-2, IL-6, and TNF-α) and (B) catabolic gene (MMP13 and ADAMT5) mRNA expression in HACs treated with IL1-β with or without ADSC-Exos (10⁹ particles/mL) for 12 days in vitro. GAPDH mRNA expression was used as the internal control. Data from each group were compared with those of the IL-1β only group, *, p < 0.05; **, p < 0.01. (C) IL-6, (D), TNF-α, (E), and MMP13 ELISAs were performed to quantify the protein levels when HACs were treated with IL1-β with or without ADSC-Exos (10⁹ particles/mL) for 12 days. Each column or scatter plot represents the mean ± SEM of six replicate cultures. Data from each group were compared with those of the control group, *, p < 0.05; **, p < 0.01 and were compared with those of the IL-1β group, #, p < 0.05; ##, p < 0.01.

Figure 6

Hypoxia−induced ADSC−Exos and ADSC treatment ameliorate ACLT−OA in rats. (A) Schematic of the experimental protocol. The rats were followed for 8 weeks after ACLT or sham surgery, and static weight-bearing was assessed before ACLT and every week post-surgery. (B) Comparison of the weight distribution analysis of ADSCs or ADSC-Exo treatment in ACLT-OA rats using the weight-bearing test. Sham control and OA control group rats received PBS in the right knee. Each scatter plot presents the means ± SEM of 7 to 8 replicate cultures. Data from each group are compared with those of the sham control group (** p < 0.01) or with those of the OA group (# p < 0.05, ## p < 0.01). (C) Gross evaluation of tibia in knee joints. The photograph represents the surfaces of tibial plateaus (scale bar: 4 mm). (D) Safranin O staining analysis of GAG content. The upper panel shows the Safranin O staining of tibia sections (scale bars: 2 mm), and the second panel shows the higher magnification of the boxed area (scale bar: 200 µm). Comparison of the density of the Safranin O stained area (red color) in articular cartilage of tibia sections in all groups after ADSC and ADSC-Exo treatment. Each column shows the mean and SEM of 7–8 samples. *, p < 0.05, **, p < 0.01 compared to the OA group. Immunohistochemical analysis of type II collagen (E), MMP13 (F), ADAMT5 (G), and IL1-β (H) in articular cartilage of tibia sections in the Sham Ctrl, OA, OA+ADSC-Exo and OA+ADSCs groups. The upper panel shows the IHC staining of ADAMT5, type II collagen, MMP13, and IL1-β (scale bar: 200 µm), and higher magnification of the box area (Scale bar: 50 µm) is shown in the bottom in all groups.

Figure 7

Next generation sequencing (NGS) and bioinformatic analysis of miRNA profiles in 4 culture conditions induced by ADSC-Exos. (A) Comparison and intersection of miRNAs from 4 culture conditions induced ADSC−Exos (2H, 2N, 10H, 10N group). (B) Heatmap of the differentially expressed miRNAs from 4 groups. Hierarchically represented the miRNAs from the 10N group and compared them with the other groups (p value < 0.05 and fold change > 2). (C) GO and (D) KEGG pathway analyses. There were 3358 target genes of 7 selected miRNAs in ADSC-Exos.