⚠️ Partially verified 2026-05-04. Singh 2022 ATLAS RCT verified against local PDF (corrections applied). Ryu 2016 (10.1038/nm.4132) and Andreux 2019 (10.1038/s42255-019-0073-4): green OA but DOI lookup failed — claims derived from these sources are marked no-fulltext-access and should be treated as approximate. Garcia 2020 (10.1039/d0fo01649g): species name corrected from title/abstract; full PDF not verified.

Urolithin A

A gut microbiome-derived dibenzopyranone (lactone) metabolite produced from dietary ellagitannins — polyphenols found in pomegranates, walnuts, and certain berries. Urolithin A (UA) is the most clinically advanced mitophagy inducer known, and the only one with completed Phase 2 randomized controlled trial data in middle-aged adults. Characterized mechanistically by the Auwerx group; commercialized by Amazentis under the brand name Mitopure (purified UA formulation that bypasses variable gut microbiome conversion).

Identity

PubChem CID: 5488186
InChIKey: RIUPLDUFZCXCHM-UHFFFAOYSA-N
ChEMBL: CHEMBL1836264
IUPAC name: 3,8-dihydroxybenzo[c]chromen-6-one
Molecular formula: C₁₃H₈O₄
Molecular weight: 228.20 g/mol
Class: dibenzopyranone (benzo[c]chromenone); polyphenol metabolite
Canonical SMILES: C1=CC2=C(C=C1O)C(=O)OC3=C2C=CC(=C3)O

Biosynthesis — the microbiome bottleneck

UA is not absorbed directly from food. The biosynthetic pathway requires sequential gut bacterial biotransformation:

Dietary ellagitannins (e.g., punicalagins in pomegranates) are hydrolyzed to ellagic acid in the gut lumen.
Ellagic acid undergoes reductive ring-opening and lactonization by specific gut bacteria to produce urolithins (A, B, C, D, etc.), with UA (3,8-dihydroxy) as the predominant end-product in high-producing individuals.
Key bacteria identified for urolithin production include Gordonibacter urolithinfaciens (reductase activity; earlier urolithin steps) and Ellagibacter isourolithinifaciens (more efficient catechol-dehydroxylase activity; full conversion to UA) ¹. no-fulltext-access — Garcia 2020 PDF not yet in archive; species names confirmed from paper title and abstract only; mechanistic detail (which step each species handles) not fully PDF-verified.

Individual variability is large. Population studies classify people into three gut metabotypes: non-producers (~40%), low-producers, and high-producers of urolithins. Non-producers and low-producers derive little UA from dietary ellagitannin intake. Direct UA supplementation (Mitopure) bypasses this variability entirely. dose-response-unclear — optimal dietary intake to support UA production in high-producing individuals is not established.

Mechanism of action

Mitophagy induction via PINK1/Parkin

The primary characterized mechanism is activation of mitophagy — selective autophagic clearance of damaged mitochondria — via the pink1-parkin-pathway ²:

In C. elegans, UA increased mitophagy (measured by fluorescent reporter) and extended median lifespan ~45% in wild-type worms; the lifespan extension was abrogated in dct-1 (C. elegans BNIP3L ortholog) and pink-1 mutants, indicating mitophagy dependence. no-fulltext-access — Ryu 2016 PDF unavailable (DOI lookup failed); lifespan % and genetic epistasis claims not verified against full paper.
In rodent models, UA improved skeletal muscle function and exercise capacity; in aged mice, oral UA preserved mitochondrial membrane potential and reduced mitochondrial ROS. no-fulltext-access — rodent dose and n not verified against Ryu 2016 full text.
In human primary myotubes, UA activated mitophagy flux and increased the PINK1/Parkin-dependent ubiquitin signal on mitochondria. no-fulltext-access — Ryu 2016 in-vitro claims not verified against full text.

Dimension	Status
Pathway conserved in humans?	yes — PINK1/Parkin pathway is identical
Phenotype conserved in humans?	partial — muscle mitochondrial markers improved in human trial; functional gains moderate
Replicated in humans?	in-progress — single Phase 2 RCT completed; further trials recruiting

Downstream effects

Mitochondrial biogenesis: UA upregulates PGC-1α and mitochondrial gene expression in muscle, consistent with improved mitochondrial turnover removing damaged organelles and allowing net biogenesis ³.
Inflammatory signaling: Reduces circulating CRP (plasma CRP reduction was a significant secondary endpoint in Singh 2022 ATLAS RCT at 1000 mg/day) ⁴. Mechanism of anti-inflammatory effect is not fully characterized — possibly secondary to reduced mitochondrial stress and SASP-like DAMP release. no-mechanism

Effects on aging hallmarks

Hallmark	Effect	Evidence
mitochondrial-dysfunction	Mitophagy induction clears damaged mitochondria; improves membrane potential and respiratory function in aged muscle	²
disabled-macroautophagy	UA is a selective mitophagy activator; acts upstream of autophagosome formation via PINK1/Parkin	²
chronic-inflammation	Reduces plasma CRP in human RCT at 1000 mg/day	⁴
immunosenescence	Expands naive-like CD8+ T cells (~0.5 pp shift, p=0.0437); increases CD8 fatty acid oxidation; modulates NK cells, monocyte subsets, mitochondrial biogenesis in T cells	⁵

Pharmacokinetics and bioavailability

UA from direct supplementation (Mitopure) is absorbed orally; rapidly conjugated (glucuronidated, sulfated) in intestinal mucosa and liver. Circulating UA is primarily in conjugate form.
First-in-human PK study (Andreux 2019, n=60 healthy sedentary elderly 61–85 yr): UA had a favourable safety profile (primary endpoint) and was bioavailable at all doses ³. no-fulltext-access — Andreux 2019 PDF unavailable; details from ClinicalTrials.gov NCT02655393 and PubMed abstract only. Per the abstract: plasma acylcarnitines and skeletal muscle mitochondrial gene expression were modulated at 500mg and 1000mg/day (4-week multiple-dose period).
Half-life of free UA: not precisely reported; conjugates circulate longer. Unlike fisetin, UA does not appear to require a “hit-and-run” senolytic model — chronic daily dosing is used.
Mitopure formulation specifically developed to achieve consistent plasma exposure independent of gut microbiome composition.

Dose-response evidence

Study	Organism	n	Dose	Route	Duration	Effect
Ryu 2016 ²	C. elegans (wild-type N2)	~35–50/group	50 µM in media	In media	Adult lifespan	~45% median lifespan extension; mitophagy-dependent
Ryu 2016 ²	Aged C57BL/6 mice	n not specified	400 mg/kg/day	Oral gavage	6 weeks	Improved grip strength, running distance; increased mitophagy markers in muscle
Andreux 2019 ³	Healthy sedentary elderly (61–85 yr)	n=60 total (Part A single-dose: 24; Part B multi-dose 28 days: 36)	Part A: single doses 250/500/1000/2000 mg; Part B: 250/500/1000 mg/day repeated	Oral	Part B: 4 weeks daily	Favorable safety profile (primary). At 500mg and 1000mg/day: modulated plasma acylcarnitines + skeletal muscle mitochondrial gene expression. no-fulltext-access — PDF unverifiable; design confirmed from ClinicalTrials.gov NCT02655393 and PubMed abstract.
Singh 2022 ATLAS RCT ⁴	Untrained, overweight middle-aged adults (40–64 yr; VO₂max <35 mL/kg/min)	n=88 randomized (placebo n=29, 500mg n=29, 1000mg n=30), 79 completed	500 mg/day or 1000 mg/day vs placebo (3-arm)	Oral	4 months	Primary endpoint (peak power output): NOT significant vs placebo. Significant muscle strength gains (hamstring) at both doses. Aerobic endurance gains (peak VO₂, 6MWT +33.4m) primarily at 1000 mg/day. CRP reduction significant at 1000 mg/day.

Human clinical evidence — ATLAS RCT (Singh 2022)

The ATLAS trial (doi:10.1016/j.xcrm.2022.100633; NCT03464500) is the canonical Phase 2 RCT of UA in untrained, overweight middle-aged adults ⁴:

Design: randomized, double-blind, placebo-controlled, 3-arm parallel
Population: untrained, overweight, middle-aged adults (40–64 yr) with low physical endurance (VO₂max <35 mL/kg/min); n=88 randomized (placebo n=29, 500mg n=29, 1000mg n=30), 79 completed (9 dropouts; ITT n=88, per-protocol n=74 with >80% compliance)
Arms: placebo / 500 mg UA/day / 1000 mg UA/day (Mitopure)
Duration: 4 months
Primary endpoint: peak power output — NOT significant vs placebo at either dose
Secondary endpoints: Hamstring skeletal muscle average peak torque was significantly improved at both doses vs placebo (500mg: +12%, p=0.027; 1000mg: +9.8%, p=0.029). Maximum torque during knee flexion also significantly improved at both doses (500mg: +10.6%, p=0.017; 1000mg: +10.5%, p=0.022). Aerobic endurance (peak VO₂) improved dose-dependently, reaching significance at 1000mg vs placebo (p=0.058 trend). 6-minute walk distance increased by 33.4m in the 1000mg group (p=0.008 within-group). Plasma CRP was significantly reduced at 1000mg/day. Skeletal muscle mitochondrial gene expression improved (at 500mg dose; GSEA).
Safety: well-tolerated; 102 total adverse events across 45 participants; no serious adverse events; adverse event profile similar across arms (mostly musculoskeletal/connective tissue AEs from muscle biopsy procedure)

Interpretation: The null primary endpoint is important and should not be downplayed. However, significant muscle-strength gains (hamstring torque at both doses) and aerobic-endurance improvements (at 1000mg) were observed as secondary endpoints. These are clinically meaningful results but remain secondary endpoints in a single, industry-sponsored, proof-of-concept trial in a specific overweight/low-endurance population. needs-replication — independent replication of muscle strength and endurance gains is needed, particularly in a leaner or sarcopenic population, before broad human-efficacy claims can be made.

Dimension	Status
Pathway conserved in humans?	yes
Phenotype conserved in humans?	partial — biomarkers improved; primary power endpoint null
Replicated in humans?	no — one completed Phase 2 RCT; further trials in progress

Human clinical evidence — additional RCTs (2024–2025)

Three additional published RCTs since ATLAS expand the human evidence base in distinct populations and indications. None has yet replicated ATLAS’s secondary muscle-strength/endurance findings independently in an aging population, but Denk 2025 (immune-aging) is the first non-skeletal-muscle proof-of-concept in humans.

Denk 2025 — Immune aging (Nature Aging)

The first RCT testing UA against age-related immune decline ⁵:

Design: randomized, double-blind, placebo-controlled
Population: n=50 healthy middle-aged adults
Arms: UA 1000 mg/day vs placebo (Mitopure)
Duration: 4 weeks
Primary outcomes: phenotypical changes in peripheral CD3+ T cell subsets and immune metabolic remodeling
Findings: UA expanded peripheral naive-like, less terminally exhausted CD8+ T cells (treatment difference 0.50 percentage points; 95% CI 0.16–0.83; p=0.0437); increased CD8+ fatty acid oxidation capacity (treatment difference 14.72 percentage points; 95% CI 6.46–22.99; p=0.0061). Secondary: increased mitochondrial biogenesis in CD8+ cells, increased peripheral CD56^dim^CD16^bright^ NK cells and nonclassical CD14^lo^CD16^hi^ monocytes; improved activation-elicited TNF secretion; improved bacterial uptake by monocytes. Exploratory single-cell RNA-seq showed UA-driven transcriptional shifts in immune populations modulating inflammation- and metabolism-linked pathways.
NCT: NCT05735886
Significance: First clinical demonstration of UA effects on the immunosenescence hallmark; supports the mitochondrial-dysfunction → T-cell-aging mechanistic chain. Industry-sponsored (Amazentis co-authors). needs-replication — single trial, short duration (4 weeks), and the primary endpoint is a surrogate (CD8 phenotype shift) rather than infection-incidence or vaccine-response. Whether the immune phenotype changes translate to clinically meaningful protection from age-related infection susceptibility is unknown.

Whitfield 2025 — Highly trained distance runners (Sports Medicine)

UA tested in young athletes (not an aging population) for ergogenic effects ⁶:

Design: double-blind, parallel-group, placebo-controlled
Population: n=42 competitive male distance runners, mean age 27.2 ± 1.0 yr
Arms: UA vs placebo
Duration: 4 weeks
Primary outcome: 3000 m time trial — NOT significant in either arm
Secondary: UA significantly lowered ratings of perceived exertion (p=0.02) and reduced indirect markers of post-exercise muscle damage (CK total AUC, p<0.0001). Aerobic capacity time × treatment interaction not significant (p=0.138).
NCT: NCT04783207
Interpretation: UA does not enhance running performance in already-highly-trained young athletes — consistent with the ATLAS interpretation that UA’s benefits are most plausible in populations with substantial mitochondrial-quality reserve to recover (untrained, overweight, sarcopenic). Supports a recovery/anti-inflammatory benefit even in trained populations.

Jamialahmadi 2024 — Heart failure with reduced ejection fraction (HFrEF)

The first RCT in a clinical-disease population ⁷:

Design: randomized, double-blind, placebo-controlled 2×2 crossover
Population: n=10 HFrEF patients
Arms: UA 500 mg BID (1000 mg/day total) vs placebo
Duration: two 4-week periods, separated by 2-week washout
Outcomes: NULL for echocardiographic indices (LVEF, LVEDD, LVESV, TAPSE all unchanged). NULL for plasma pro-BNP, glucose, and CRP. Serum HDL-C increased modestly (+6.46 ± 2.33 mg/dL, p=0.026) on UA vs placebo. Other lipids unchanged.
Interpretation: Underpowered (n=10), short duration. The negative cardiac result does not support cardioprotective benefit at this dose/duration in established HFrEF, contrasting with preclinical mouse cardioprotection literature. needs-replication — larger, longer trials at higher doses needed before drawing conclusions about UA cardiac efficacy in disease.

Human trials summary (active + recently completed)

ClinicalTrials.gov v2 API (queried 2026-05-08): 17 trials in RECRUITING or ACTIVE_NOT_RECRUITING status; 32 completed; 1 withdrawn; 1 terminated.

Trial	NCT	Phase	Status	Population	Primary Outcome
ATLAS (Singh 2022)	NCT03464500	Phase 2	Completed	Untrained overweight middle-aged adults (40–64 yr)	Peak power output (null); muscle strength significant
ENERGIZE	NCT03283462	Phase 2	Completed	Elderly skeletal muscle	6-min walk, ATP synthesis
Denk 2025 immune-aging	NCT05735886	N/A	Completed	Healthy middle-aged adults (n=50)	CD3+ T-cell subset shift (significant)
Whitfield 2025 (runners)	NCT04783207	N/A	Completed	Competitive male distance runners	3000m time trial (null); RPE + CK reduced
Jamialahmadi 2024 HFrEF	not registered	N/A	Completed	n=10 HFrEF crossover	Echo measures (null); HDL-C up
ATLAS 2	NCT07231783	N/A	Recruiting	Healthy middle-aged adults	Knee flexion strength at 6 months
COPD supplement study	NCT06324214	Phase 2	Recruiting	COPD patients	Exercise endurance capacity
MitoEM	NCT06556706	N/A	Completed	Frail older adults	Mitochondrial ultrastructure (EM)
Brain Longevity (Mitopure)	NCT07060898	N/A	Active not recruiting	Cognitive function	Cognition
UA + ICI in solid tumors	NCT07161310	N/A	Recruiting	Untreated solid tumors on immune checkpoint inhibitors	Immune/anti-tumor response
UA + fisetin (sleep + aging biomarkers)	NCT06990256	N/A	Recruiting	Adults	Sleep, aging biomarkers
UA prostate-cancer radical prostatectomy	NCT06022822	N/A	Recruiting	Men with prostate cancer	Tissue mitochondrial markers
UA glucose metabolism in 55+	NCT06274749	N/A	Recruiting	Healthy adults ≥55 yr	Glucose metabolism
UA + protein single-leg immobilization	NCT05814705	N/A	Active not recruiting	Disuse atrophy model	Muscle preservation

long-term-unknown — No trial to date has followed participants beyond 6 months. Long-term safety and efficacy in older, frailer, or disease populations requires further study. The Brain-Longevity (NCT07060898) and ATLAS 2 (NCT07231783) trials are the first to extend dosing to ≥6 months and may begin to address this gap.

Classification

SENS strategy: MitoSENS — supports mitochondrial quality via mitophagy-mediated turnover
Hallmark targets: mitochondrial-dysfunction, disabled-macroautophagy
Clinical category: dietary supplement / investigational compound (Mitopure formulation)

Relationship to dietary ellagitannin sources

Source	Ellagitannin content	UA production likelihood
Pomegranate juice	High (punicalagins dominant)	Depends on gut metabotype
Walnuts	Moderate (pedunculagin)	Depends on gut metabotype
Raspberries, strawberries	Lower	Depends on gut metabotype
Mitopure (direct UA)	N/A — bypasses conversion	Consistent regardless of microbiome

The Amazentis commercial rationale for Mitopure is precisely that dietary ellagitannin intake is unreliable as a UA source for ~40% of the population.

Limitations and gaps

Null primary endpoint in ATLAS: Peak power was not improved vs placebo. Aerobic endurance gains were secondary endpoints; positive results in secondary endpoints should be interpreted cautiously (#gap/needs-replication).
Single-company dominance: Most published RCT evidence comes from Amazentis-sponsored trials (Auwerx group + Rinsch = company co-founders). Independent replication is needed. needs-replication
Mechanism not fully resolved in humans: UA’s effect on mitophagy flux in human muscle biopsies has been inferred from gene-expression signatures and acylcarnitine ratios; direct mitophagy flux measurement (e.g., mitophagy reporters in human muscle) is lacking. no-mechanism
Population specificity: ATLAS enrolled specifically untrained, overweight adults with low baseline endurance (VO₂max <35 mL/kg/min) — a population with more room for mitochondrial improvement. Translation to already-active, lean, sarcopenic elderly, frail patients, or disease states is not established.
Microbiome metabotype and dietary UA: Whether high-producing individuals who achieve comparable UA plasma levels from diet alone show similar benefits is unknown. needs-replication
Long-term safety: No trial beyond 6 months. long-term-unknown
Urolithin family interactions: UA is the most studied; urolithins B, C, D have different bioactivities and may modulate UA effects when co-produced. Contribution of other urolithins to observed effects is not characterized. unsourced

Footnotes

garcia-2020-gordonibacter-ellagibacter-urolithin · doi:10.1039/d0fo01649g · observational (in vitro bacterial culture) · model: Gordonibacter urolithinfaciens + Ellagibacter isourolithinifaciens pure cultures + mixed human stool · no-fulltext-access — PDF not yet in archive; species names confirmed from paper title/abstract only ↩
ryu-2016-urolithin-a-mitophagy · doi:10.1038/nm.4132 · n=~35–50/group (worm); rodent n not specified · in-vivo (C. elegans + mouse) · p<0.05 (lifespan) · model: C. elegans N2 wild-type + aged C57BL/6 mice ↩ ↩² ↩³ ↩⁴ ↩⁵
andreux-2019-urolithin-a-human-safety · doi:10.1038/s42255-019-0073-4 · n=60 · rct (single + multiple ascending dose) · model: healthy sedentary elderly 61–85 yr (NCT02655393; Part B: 250/500/1000 mg/day × 4 weeks) · no-fulltext-access ↩ ↩² ↩³
singh-2022-atlas-rct-urolithin-a · doi:10.1016/j.xcrm.2022.100633 · n=88 randomized (placebo n=29 / 500mg n=29 / 1000mg n=30) / 79 completed · rct · primary endpoint (peak power) ns; muscle strength significant at both doses · model: untrained overweight middle-aged adults (40–64 yr; VO₂max <35 mL/kg/min), 4-month, 3-arm · local PDF: DOI lookup 10.1016/j.xcrm.2022.100633 ↩ ↩² ↩³ ↩⁴
denk-2025-urolithin-a-immune-aging · doi:10.1038/s43587-025-00996-x · n=50 · rct (double-blind, placebo-controlled) · primary CD8+ T-cell phenotype shift p=0.0437; CD8 FAO p=0.0061 · model: healthy middle-aged adults, 1000 mg/day × 4 weeks (NCT05735886) · journal: Nature Aging · authors: Denk D, Singh A, Kasler HG, D’Amico D, Rey J, Alcober-Boquet L, Gorol JM, Steup C, Tiwari R, Kwok R, Argüello RJ, Faitg J, Sprinzl K, Zeuzem S, Nekljudova V, Loibl S, Verdin E, Rinsch C, Greten FR · industry-sponsored (Amazentis co-authors) · needs-replication · abstract-verified via PubMed PMID 41174221 / Europe PMC PMC12618261 · author correction: doi:10.1038/s43587-025-01060-4 ↩ ↩²
whitfield-2025-urolithin-a-runners · doi:10.1007/s40279-025-02292-5 · n=42 · rct (double-blind, parallel) · primary 3000 m time trial ns; RPE p=0.02; CK total AUC p<0.0001 · model: competitive male distance runners (mean 27.2 yr; 4 weeks; NCT04783207) · journal: Sports Medicine · abstract-verified via PubMed PMID 40839339 / PMC12628386 ↩
jamialahmadi-2024-urolithin-a-hfref · doi:10.2174/0115748871279354240209101604 · n=10 · rct (2×2 crossover, double-blind, placebo-controlled) · primary echo measures (LVEF, LVEDD, LVESV, TAPSE) all ns; HDL-C +6.46 mg/dL p=0.026 · model: HFrEF patients, UA 500 mg BID × 4 weeks per period · journal: Reviews on Recent Clinical Trials · abstract-verified via PubMed PMID 38415449 ↩

🧬 Aging Wiki

Explorer

urolithin-a