Exercise & FitnessResearch PaperPaywall

Exercise Boosts Joint Health in Aging via microRNA-29 Extracellular Vesicles

A 3-month aerobic exercise program raised microRNA-29 in blood vesicles of older adults, epigenetically restoring youthful cartilage biology.

Thursday, October 1, 2026 6 views
Published in Nat Aging
An older adult on a treadmill in a clinical exercise lab, with a close-up inset of a knee joint cross-section showing cartilage tissue

Summary

Exercise is a first-line treatment for knee osteoarthritis, but how it sends protective signals to the joint has been unclear. Researchers tested whether extracellular vesicles, tiny packets that circulate in blood, carry those signals. Vesicles collected from ten older adults after three months of moderate aerobic exercise were richer in microRNA-29. Applied to aged human cartilage cells grown on a stiff, age-mimicking surface, the post-exercise vesicles restored a more youthful cell shape, boosted cartilage-building proteins, lowered cartilage-degrading enzymes, and switched the longevity gene Klotho back on by reducing repressive DNA methylation. Lab-engineered vesicles loaded with microRNA-29 reproduced much of the effect, and injecting post-exercise vesicles into the knees of aged mice reduced cartilage damage. The work is small and preclinical, but it offers a concrete mechanism for how exercise protects joints.

Detailed Summary

Knee osteoarthritis is a leading cause of disability worldwide, and current treatments mostly relieve symptoms rather than slow cartilage aging. Exercise is recommended first-line, yet how it signals to a distant tissue like cartilage is poorly understood. Extracellular vesicles (EVs), nanosized packets carrying microRNAs, proteins and lipids, have been proposed as messengers. Aged cartilage is also much stiffer (about 50–100 kPa versus 1–30 kPa when young), and this stiffness epigenetically represses the longevity gene KLOTHO. This study asked what exercise puts into EVs and whether that cargo can rejuvenate aged chondrocytes.

The team first reanalyzed public small-RNA sequencing data (GSE144627) from ten older men (mean age 69.3) who did a 40-minute cycling bout at about 70% heart rate. Of 445 microRNAs, 16 changed (|log2 fold change| > 1, nominal P < 0.05), and 15 rose. Predicted targets (43 genes) clustered in cell–matrix adhesion, including ITGB1 and COL4A1. Network propagation on a cartilage-specific network linked them to integrin, focal adhesion, TGF-β and PI3K–Akt signaling, with stronger enrichment for age-related than post-traumatic osteoarthritis. Computationally deleting miR-29a-3p and miR-29b-3p cut predicted targets from 43 to 7 and removed over 75% of the effect on key pathways.

Next, ten adults aged 65–85 (80% women) completed three months of moderate aerobic exercise, with stable body weight (83.2 versus 83.4 kg). EVs were isolated before (EVpre) and after (EVpost) by size-exclusion chromatography. EV size and number did not change. On a 100 kPa stiff substrate, aged human chondrocytes treated with EVpost moved toward the shape of cells on soft substrate (12 of 19 altered features overlapped; FormFactor P = 0.009) and had less ITGB1 (P = 0.048). EVpost raised type II collagen (P = 0.008), aggrecan (P = 0.002) and SOX9 (P = 0.031), and lowered MMP13 (P = 0.001) and ADAMTS4 (P = 3.2 × 10⁻⁵).

Mechanistically, EVpost reduced the methylating enzyme DNMT1 (P = 0.003) and raised α-Klotho (P = 9.5 × 10⁻⁵); the two changes were inversely correlated (P = 0.009). KLOTHO mRNA rose (P = 0.015), and KLOTHO promoter methylation fell (P = 0.038). Digital PCR showed EV miR-29b-3p increased after exercise (P = 0.043), while miR-29a-3p showed a trend (P = 0.076). Individual rises in both microRNAs correlated with Klotho gains (P = 0.015 and 0.047), hinting at a responder biomarker.

Engineered EVs enriched in miR-29a/b lowered ITGB1 (western blot, P = 0.003), raised α-Klotho (P = 0.044), and increased type II collagen (P = 0.015) and aggrecan (P = 0.003). Klotho knockdown partly blunted these effects, so the dependence is partial. In male mice aged 21–22 months, twice-weekly intra-articular injections for four weeks, with EVpre in one knee and EVpost in the other (n = 9), lowered OARSI cartilage damage scores (P = 0.026), increased type II collagen (P = 0.033), and reduced MMP13 (P = 0.032) and ADAMTS4 (P = 0.049). Body mass stayed stable.

The caveats are substantial. The exercise cohort was ten people with no non-exercising control group. All human EVs were tested on chondrocytes from one 65-year-old male donor, and the mouse work used only males. The in silico discovery data came from acute exercise in ten men, not chronic training. P values were nominal, and the miR-29a-3p PCR change did not reach significance. Direct EV injection is also not what exercise does physiologically. Still, the miR-29–Klotho axis is a testable target for biomarker-guided, EV-based osteoarthritis therapy.

Key Findings

  • Reanalysis of acute-exercise EV sequencing in 10 older men found 16 of 445 microRNAs changed, 15 of them upregulated, with predicted targets concentrated in cell–matrix adhesion (43 target genes).
  • In silico deletion of miR-29a-3p and miR-29b-3p cut predicted EV target genes from 43 to 7 and accounted for more than 75% of the effect on integrin, PI3K–Akt, focal adhesion, TGF-β and ECM–receptor pathways.
  • After 3 months of moderate aerobic exercise in 10 adults aged 65–85 (80% women), EV miR-29b-3p rose significantly (P = 0.043); miR-29a-3p showed a trend (P = 0.076).
  • Post-exercise EVs on a stiff 100 kPa substrate raised type II collagen (P = 0.008), aggrecan (P = 0.002) and SOX9 (P = 0.031) and lowered MMP13 (P = 0.001) and ADAMTS4 (P = 3.2 × 10⁻⁵) in aged human chondrocytes.
  • Post-exercise EVs reduced DNMT1 (P = 0.003), cut KLOTHO promoter methylation (P = 0.038) and raised α-Klotho protein (P = 9.5 × 10⁻⁵) and KLOTHO mRNA (P = 0.015).
  • Engineered EVs enriched in miR-29a/b lowered ITGB1 (P = 0.003) and raised α-Klotho (P = 0.044); Klotho knockdown only partly blunted the anabolic effects, showing partial dependence.
  • In 21–22-month-old male mice, post-exercise EVs injected into one knee gave lower OARSI cartilage damage than pre-exercise EVs in the opposite knee (P = 0.026, n = 9), with less MMP13 (P = 0.032) and ADAMTS4 (P = 0.049).

Methodology

The study combined in silico analysis, human exercise samples, cell culture and a mouse model. Public small-RNA-seq data (GSE144627; 10 older men, acute cycling) were analyzed with DESeq2 and mapped onto a HumanBase cartilage-specific network using random-walk-restart propagation. EVs were isolated by size-exclusion chromatography from 10 adults aged 65–85 before and after 12 weeks of moderate aerobic exercise, then applied to aged human chondrocytes on 4 kPa or 100 kPa substrates; outcomes were 53 morphometric features, immunofluorescence, KLOTHO expression and methylation, and paired t-tests. Engineered miR-29a/b EVs and KLOTHO siRNA tested causality, and aged male mice (21–22 months) received intra-articular injections twice weekly for 4 weeks in a within-animal design (EVpre versus EVpost, n = 9).

Study Limitations

The exercise cohort was small (n = 10), had no non-exercising control group, and was 80% women, while the in silico discovery data came from ten men after a single acute bout. Human EVs were tested on chondrocytes from a single 65-year-old male donor, mouse experiments used only males, and many analyses relied on nominal P values without multiple-testing correction. Engineered EV effects depended only partly on Klotho, and the miR-29a-3p change by digital PCR was not significant (P = 0.076). No conflict-of-interest statement appeared in the supplied text.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: