FNDC5/irisin-enriched sEVs conjugated with bone-targeting aptamer alleviate osteoporosis: a potential alternative to exercise

Abstract

Exercise maintains bone health and produces protective effects on bone loss. In this study, we investigated the potential protective effects of circulating small extracellular vesicles (sEVs) generated under endurance exercise training (Exe-sEVs) on ovariectomized (OVX)-induced bone loss. Inhibition of sEVs secretion by GW4869 partially reversed exercised-induced protection against OVX-induced bone loss. Importantly, Exe-sEVs was internalized by bone tissue and alleviated bone loss in OVX-mice. The increased levels of fibronectin type-III domain-containing protein 5 (FNDC5/irisin) in Exe-sEVs contributed to the promotion of osteogenesis in bone marrow mesenchymal stem cells (BM-MSCs). However, systemic knockdown of FNDC5, the precursor of irisin, abolished the exercise-induced protective effects against bone loss in OVX-mice. Moreover, incubation of irisin enhanced osteogenesis and attenuated adipogenesis in BM-MSCs. Intriguingly, implantation of BM-MSCs overexpressing FNDC5 accelerated osteogenesis and chondrogenesis in BALB/c immunodeficiency mice. Mechanistically, irisin promoted phosphorylation of p38MAPK and JNK, but not ERK. Blocking the JNK and p38MAPK signaling pathway with specific inhibitors abolished the pro-osteogenesis and anti-adipogenesis effects of irisin on BM-MSCs. However, inhibition of β-arrestin-2 rescued the irisin-induced activation of p38MAPK and JNK. Finally, aptamer-modified FNDC5-sEVs (Apt-FNDC5-sEVs) exhibited higher enrichment in bone tissues and enhanced bone formation. In conclusion, exercise-induced circulating FNDC5/irisin-enriched sEVs promote osteogenesis of mouse BM-MSCs both in vitro and in vivo, partially through a β-arrestin-2-dependent p38MAPK and JNK signaling pathway. Apt-FNDC5-sEVs represent a promising strategy for the treatment of osteoporosis.

Keywords: Aptamer; Bone marrow mesenchymal stem cells; Exercise; FNDC5/irisin; Osteoporosis; Small extracellular vesicles.

Fig. 1

Inhibition of sEVs secretion partially blocked exercise-induced protection on bone loss in OVX-induced mice. A-I Schematic flow diagram illustrating the experimental design for OVX-mice subjected to sedentary conditions, treadmill exercise, with or without GW4869 inhibition. Mice subjected to sham operation were used as control group (n = 6). Briefly, OVX-mice were randomly subjected to sedentary (Sed group, n = 6) or eight weeks treadmill training (Exe group, n = 6) with or without inhibition of sEVs secretion by GW4869 for every two days (Exe + GW4869 group, n = 6). B Representative views of micro-CT were presented. C Statistical analysis of bone mineral density (BMD). D Statistical analysis of trabecular bone volume to tissue volume (Tb.BV/TV). E Statistical analysis of trabecular bone number (Tb.N). F Statistical analysis of trabecular thickness (Tb.Th). G Statistical analysis of trabecular separation to spacing (Tb.Sp). H Quantification of cortical bone thickness (Ct.Th). I Representative images of calcein double labeling were displayed. Scale bar represents 100 μm. Quantitative analysis of bone formation rate (BFR) and mineral apposition rate (MAR) was shown in the right panel. J Representative immunohistochemical images of osteocalcin (OCN) in bone tissue. Scale bar represents 200 μm. Quantitative analysis of OCN positive area was presented in the right panel. K Serum level of P1NP in mice was detected by ELISA kit. (L-N) BM-MSCs were treated with serum isolated from sedentary (Sed) or exercised (Exe) mice, with or without deletion of sEVs through ultracentrifuging. L Alkaline phosphatase (ALP) activity was determined by a specific measurement kit (n = 5). M Expression of Runx2 were detected by western blotting (n = 4). Quantitative analysis was presented in the lower panel. N The osteogenic differentiation and mineralization level of BM-MSCs was assessed by Alizarin Red S (ARS) and ALP staining (n = 5). Scale bar represents 200 μm. Quantitative analysis was presented in the lower panel. Results are represented as mean ± SD with four to six replicates per group. Statistical significance was determined using two-way ANOVA with the Tukey’s HSD post hoc analysis. *p < 0.05. **p < 0.01. ***p < 0.001. NS, no significance

Fig. 2

Identification and tracking of sEVs. A Transmission electron micrographs depicted sEVs isolated from rat serum. Scale bar represents 50 nm. B Nanoparticle tracking analysis of sEVs purified from rat serum was conducted. C Expression of exosomal marker proteins CD9, CD81, TSG101, and the cytoplasmic marker BSA in sEVs were detected by western blotting. D-F The OVX-mice were administered with vehicle (sterile PBS), or DiR-labeled Sed-sEVs, or Exe-sEVs intravenously (100 μg per mouse, n = 4 per group). Representative in vivo images of mice (D), ex-vivo images of major organs (E), and ex-vivo images of bone tissues (F) were presented after treatment for 24 h. Quantification of fluorescence intensity was presented in the right panel. G BM-MSCs were incubated with Dio-labeled sEVs for 12 h and the uptake of sEVs by BM-MSCs was observed using fluorescence microscopy (Dio stained in green, nuclear staining with DAPI in blue). Three independent experiments were performed, and representative data are shown. Statistical significance was determined using two-way ANOVA with the Tukey’s HSD post hoc analysis. *p < 0.05. **p < 0.01. ***p < 0.001. NS, no significance

Fig. 3

Exe-sEVs ameliorated bone loss in OVX-induced mice. Rats were assigned to either a treadmill training group or a sedentary control group. Following the treatment period, serum was collected from the rats and subjected to ultracentrifugation to isolate sEVs. The sEVs isolated from serum of sedentary rats (Sed-sEVs) or exercise-trained rats (Exe-sEVs) were subsequently injected into OVX-mice. A-I OVX-mice were randomly assigned to receive either Sed-sEVs or Exe-sEVs (100 μg per mouse) via intravenous injection every three days for a duration of eight weeks. B Representative views of micro-CT were presented. C Statistical analysis of BMD. D Statistical analysis of Tb.BV/TV. E Statistical analysis of Tb.N. F Statistical analysis of Tb.Th. G Statistical analysis of Tb.Sp. H Quantification of Ct.Th. I Representative images of calcein double labeling. Scale bar represents 100 μm. Quantitative analysis of BFR and MAR was presented in the lower panel. J Representative immunohistochemical images of OCN in bone tissues. Scale bar represents 200 μm. Quantitative analysis of OCN positive area was presented in the right panel. K Serum level of P1NP in mice was detected by ELISA kit. Results are represented by mean ± SD with six replications (in vivo) for each group. Statistical significance was determined using Student’s t-test. *p < 0.05. **p < 0.01. ***p < 0.001

Fig. 4

Levels of FNDC5/irisin in circulation of exercise rats and roles of irisin in sEVs-induced promotion of bone formation. A Levels of serum FNDC5/irisin in rats was detected by commercial ELISA kit (n = 6). B Level of FNDC5/irisin in sEVs was detected by a commercial ELISA kit (n = 6). C-H BM-MSCs were incubated with Exe-sEVs or Sed-sEVs for the indicated time and the cell lysates were isolated. C qRT-PCR was used to detect mRNA expression of FNDC5 in BM-MSCs (n = 4). D Western blotting was used to detect protein expression of FNDC5 in BM-MSCs (n = 4). Quantitative analysis was shown in the right panel, as shown by percentage relative to the control group (100%). E BM-MSCs were treated with serum isolated from Sed-mice or Exe-mice with or without deletion of sEVs by ultracentrifuging. After the treatment, western blotting was used to detect protein expression of FNDC5 in BM-MSCs. Quantitative analysis was shown in the right panel. F–H BM-MSCs were incubated with Exe-sEVs or Sed-sEVs with or without FNDC5 neutralization antibodies. F Expression of Runx2 was detected by western blotting (n = 4). Quantitative analysis was shown in the right panel. G ALP activity was determined by a specific ALP measurement kit (n = 5). H The osteogenic differentiation and mineralization level of BM-MSCs was assessed by ARS and ALP staining (n = 5). Scale bar represents 200 μm. Quantitative analysis was presented in the right panel. Results are represented by mean ± SD with four or five replicates per group. Statistical significance was determined using Student’s t-test. *p < 0.05. **p < 0.01. ***p < 0.001

Fig. 5

Role of FNDC5/irisin in exercise-induced protection against bone loss in OVX-induced mice. A-I Schematic flow diagram representing the OVX-mice subjected to sedentary, treadmill, with or without AAV9-shFNDC5 (n = 6). Briefly, OVX-mice were randomly subjected to sedentary and AAV9-shCtrl (Sed + AAV9-shCtrl group, n = 6), eight weeks treadmill training and AAV9-shCtrl (Exe + AAV9-shCtrl group, n = 6), or eight weeks treadmill training and AAV9-shFNDC5 (Exe + AAV9-shFNDC5 group, n = 6). B Representative views of micro-CT were shown. C Statistical analysis of mineral density BMD. D Statistical analysis of Tb.BV/TV. E Statistical analysis of Tb.N. F Statistical analysis of Tb.Th. G Statistical analysis of Tb.Sp. H Quantification of Ct.Th. I Representative images of calcein double labeling. Scale bar represents 100 μm. Quantitative analysis of BFR and MAR was shown in the right panel. J Representative immunohistochemical images of OCN in bone tissues. Scale bar represents 200 μm. Quantitative analysis of OCN positive area was presented in the right panel. Results are represented by mean ± SD with six replications (in vivo) per group. Statistical significance was determined using two-way ANOVA with the Tukey’s HSD post hoc analysis. *p < 0.05. **p < 0.01. ***p < 0.001. NS, no significance

Fig. 6

FNDC5/irisin increased osteogenesis and decreased adipogenesis in BM-MSCs and promoted osteogenesis in vitro and in vivo. A-C BM-MSCs were incubated with irisin at concentration of 5 nM to 40 nM or vehicle control, accompanying with osteogenic medium. A The ALP activity was determined by a specific ALP measurement kit (n = 5). B The osteogenic differentiation and mineralization level of BM-MSCs was detected by ARS and ALP staining (n = 5). Scale bar represents 200 μm. C Expression of Runx2 was detected by western blotting (n = 4). D BM-MSCs were incubated with irisin at concentration of 5 nM to 40 nM or vehicle control, accompanying with adipogenic medium. Expression of PPARγ was detected by western blotting. E Lentivirus carrying FNDC5 was used to over-express FNDC5 in BM-MSCs, and then these cells loading on beta-tricalcium phosphate stent complex scaffold were implanted into nude mice for 12 weeks. F H&E staining of the isolated tissues. G Osteoid tissue formation was detected by Masson’s trichome staining. Quantitative analysis was shown in the right panel. H Representative immunohistochemical images of OCN, Runx2 and β-arrestin-2 in isolated tissues. I Cartilaginous tissue was detected by safranin O-fast green staining. Quantitative analysis was shown in the right panel. Green scale bar represents 1 mm. Red scale bar represents 400 μm. Black scale bar represents 100 μm. Quantitative analysis was shown in the right panel. Results are represented by mean ± SD with four or five replicates per group. Statistical significance was determined using Student’s t-test or two-way ANOVA with the Tukey’s HSD post hoc analysis. *p < 0.05. **p < 0.01. ***p < 0.001. NS, no significance

Fig. 7

FNDC5/irisin promoted BM-MSCs osteogenic differentiation and reduced their adipogenesis by activating p38MAPK and JNK signaling pathways. A BM-MSCs were incubated with FNDC5/irisin for indicated time and p-p38MAPK, p-JNK, p-ERK1/2 as well as their total proteins were detected by western blotting. B-D BM-MSCs were pretreated with 10 µM SB203580 (a p38MAPK inhibitor) or 10 µM SP600125 (a JNK inhibitor) for 30 min, and then incubated with FNDC5/irisin for indicated time. B The ALP activity was assessed by a specific ALP measurement kit (n = 5). C Expression of Runx2 was detected by western blotting (n = 4). D The osteogenic differentiation and mineralization level of BM-MSCs was assessed by ARS and ALP Staining (n = 5). Scale bar represents 200 μm. E BM-MSCs were pretreated with 10 µM SB203580 or 10 µM SP600125 for 30 min and then incubated with FNDC5/irisin, accompanying with adipogenic medium. Expression of PPARγ was assessed by western blotting (n = 4). Quantitative analysis was presented in the right or lower panel. Results are represented by mean ± SD with four or five replicates per group. Statistical significance was determined using two-way ANOVA with the Tukey’s HSD post hoc analysis. *p < 0.05. **p < 0.01. ***p < 0.001. NS, no significance

Fig. 8

β-arrestin-2 mediated FNDC5/irisin-induced activation of p38MAPK and JNK and promoting effects on BM-MSCs osteogenic differentiation. A-E BM-MSCs were pretreated with β-arrestin-2 specific siRNA (si-β-arrestin-2) or control siRNA, and then incubated with irisin for indicated time, accompanying by osteogenic medium. A Expression of Runx2 and β-arrestin-2 was assessed by western blotting (n = 4). Quantitative analysis was shown in the right panel. B p-p38MAPK and p-JNK as well as their total proteins were detected by western blotting (n = 4). Quantitative analysis was shown in the right panel. C The ALP activity was assessed by a specific ALP measurement kit (n = 5). D The osteogenic differentiation and mineralization level of BM-MSCs was assessed by ARS and ALP staining (n = 5). Scale bar represents 200 μm. E BM-MSCs were pretreated with si-β-arrestin-2 or control siRNA, and then incubated with FNDC5/irisin for indicated time, accompanying by adipogenic medium. Expression of PPARγ was detected by western blotting (n = 4). Quantitative analysis was presented in the lower panel. Results are presented by mean ± SD with four or five replicates per group. Statistical significance was determined using Student’s t-test or two-way ANOVA with the Tukey’s HSD post hoc analysis. *p < 0.05. **p < 0.01. ***p < 0.001. NS, no significance

Fig. 9

Aptamer-modified FNDC5-sEVs effectively alleviated bone loss in OVX-induced mice. A Schematic illustration of the bone targeting aptamer-modified sEVs. B Rd-sEVs or Apt-sEVs, either with or without the aptamer modification, were labeled with DiR and subsequently administered intravenously to mice (100 μg per mouse, n = 4 per group). Representative ex-vivo images of major organs (B, left panel), and ex-vivo images of bone tissues (B, middle panel) were captured at 6 and 12 h post-treatment. Quantitative analysis was presented in the right panel. C-J OVX-mice were randomly assigned to receive intravenous injections of Rd-Ctrl-sEVs, Apt-Ctrl-sEVs, Rd-FNDC5-sEVs, or Apt-FNDC5-sEVs (100 μg per mouse) every three days for a duration of eight weeks. C Representative views of micro-CT were displayed. D Statistical analysis of BMD. E Statistical analysis of Tb.BV/TV. F Statistical analysis of Tb.N. G Statistical analysis of Tb.Th. H Statistical analysis of Tb.Sp. I Quantification of Ct.Th. J Representative images of calcein double labeling. Scale bar represents 100 μm. Quantitative analysis of BFR and MAR was presented in the right panel. K Representative immunohistochemical images of OCN in bone tissues. Quantitative analysis was presented in the right panel. Scale bar represents 200 μm. L Serum level of P1NP in mice was detected by ELISA kit. Results are represented by mean ± SD with six replications (in vivo) per group. Statistical significance was determined using Student’s t-test. *p < 0.05. **p < 0.01. ***p < 0.001. NS, no significance

Fig. 10

A proposed model of sEVs-containing FNDC5/irisin ameliorates bone loss. Exercise stimulates the secretion of sEVs enriched with FNDC5/irisin into the circulation, which are then taken up by bone tissues. Irisin activates the phosphorylation of p38MAPK and JNK in a β-arrestin-2-dependent manner, resulting in enhanced bone formation and reduced adipogenesis. Finally, OVX-induced bone loss is abrogated