Tretinoin

All-trans retinoic acid (ATRA), the biologically active form of vitamin A. The best-evidenced topical anti-aging pharmaceutical — the only retinoid with an FDA approval specifically for photoaging reversal (Renova, 1995). Its molecular mechanism is established by direct human in vivo data: topical tretinoin pretreatment reduces UVB-induced AP-1 DNA binding by 70% and suppresses MMP-1 and MMP-9 mRNA, protein, and enzymatic activity by 50–80%, via nuclear RAR-mediated transrepression of c-Jun/c-Fos rather than a sunscreen or collagen-synthesis-first mechanism ¹. Decades of RCT evidence confirm clinical reversal of photoaging signs; tolerability remains the primary barrier to use.

Identity

• PubChem CID: 444795
• InChIKey: SHGAZHPCJJPHSC-YCNIQYBTSA-N
• CAS: 302-79-4
• ChEMBL: CHEMBL38
• DrugBank: DB00755
• Molecular formula: C₂₀H₂₈O₂
• Molecular weight: 300.4 Da
• Canonical SMILES: CC1=C(C(CCC1)(C)C)/C=C/C(=C/C=C/C(=C/C(=O)O)/C)/C
• Class: diterpenoid retinoid (β-ionone ring + polyene chain + carboxylic acid head group)
• LogP (XLogP3): 6.3 (highly lipophilic; penetrates lipid-rich stratum corneum)
• Topological polar surface area: 37.3 Ų
• Hydrogen bond donors / acceptors: 1 / 2

Brand names (historical): Retin-A (0.025%, 0.05%, 0.1% cream/gel; acne indication, 1971); Renova (0.02%, 0.05% cream; photoaging indication, 1995); Vesanoid (10 mg oral capsules; acute promyelocytic leukemia, 1995); multiple generics.

FDA approval timeline

Year	Product	Indication	Route
1971	Retin-A (Johnson & Johnson)	Acne vulgaris	Topical cream/gel
1995	Vesanoid (Roche)	Acute promyelocytic leukemia (APL; PML-RARα fusion)	Oral
1995	Renova (Ortho-McNeil)	Adjunctive treatment of fine facial wrinkles + mottled hyperpigmentation in photoaged skin	Topical cream

Renova’s 1995 approval is the aging-context-direct indication — making tretinoin the only retinoid with formal FDA approval for a skin-aging endpoint (not just acne). This distinction matters for the clinical-stage: fda-approved classification here (not “supplement”).

Mechanism

(a) RAR-α/β/γ binding and RARE-driven transcription

All-trans retinoic acid binds nuclear retinoic acid receptors (RAR-α, RAR-β, RAR-γ) with nanomolar affinity ². The RAR/ATRA complex heterodimerizes with retinoid X receptor (RXR) and binds retinoic acid response elements (RAREs) — direct repeat sequences — in target gene promoters. Depending on coregulator recruitment (co-activators vs co-repressors), the complex either activates (e.g., RAR-β induction) or represses target gene transcription. RAR-β itself is a transcriptional target of tretinoin, creating a positive feedback loop for retinoid responsiveness in treated skin.

(b) AP-1 transrepression — the canonical photoaging mechanism (Fisher 1996)

The primary mechanism through which tretinoin prevents UV-induced collagen degradation is ligand-dependent AP-1 transrepression ¹. The RAR/ATRA complex physically interacts with c-Jun and c-Fos proteins (the AP-1 heterodimer), forming complexes that lack transactivation capacity for AP-1 response elements in MMP gene promoters.

Quantitative results (text-stated in Fisher 1996, pp.338-339; verified R39 against local PDF):

• Topical 0.1% t-RA applied 48 h before 2 MED UVB irradiation reduced UVB-induced AP-1 DNA binding by 70% (n=9, p<0.01 vs vehicle) ¹.
• t-RA pretreatment reduced UVB-induced MMP-9 (92K gelatinase) mRNA, protein, and activity by 50–80% (n=10/9/9, p<0.01 vs vehicle) ¹.
• t-RA pretreatment reduced UVB-induced MMP-1 (interstitial collagenase) mRNA, protein, and activity by 50–80% (n=10/8/7, p<0.01 vs vehicle) ¹.
• MMP-2 (72K gelatinase) was NOT induced by UVB and was NOT altered by t-RA — explicitly stated and confirmed ¹.
• The affected MMPs are MMP-1 (interstitial collagenase), MMP-3 (stromelysin I, not stromelysin II — confirmed by probe hybridization), and MMP-9 (92K gelatinase). MMP-2 is constitutively expressed but UV/t-RA-independent ¹.
• t-RA did not reduce UVB-induced skin reddening (erythema), ruling out a UV-absorption/sunscreen mechanism for the MMP suppression ¹.

What the mechanism is NOT: AP-1 transrepression via protein-protein interaction (Jun/Fos) is the established mechanism. The commonly-cited alternative hypotheses — CBP/p300 coactivator competition and TIMP-1 induction — are NOT described in Fisher 1996 and are not the primary mechanism supported by this study.

(c) RAR-β-driven collagen synthesis stimulation

Post-Fisher-1996 laboratory work from the Voorhees/Fisher group documented that tretinoin also stimulates procollagen I synthesis in dermal fibroblasts, partially via upregulation of TGF-β/Smad3 signaling downstream of RAR-β. This is a separate mechanism from AP-1 transrepression and was established in subsequent studies, not in Fisher 1996 itself. needs-replication — the specific molecular pathway (RAR-β → TGF-β → Smad3 → COL1A1) has not been independently replicated in large human in vivo studies.

Effects on aging hallmarks

Hallmark	Mechanism	Evidence
genomic-instability	Reduces UV-induced MMP-1/MMP-9 → less collagen degradation; reduces actinic keratosis burden (UV-mutant p53+ clone clearance via increased keratinocyte turnover)	1, 3
loss-of-proteostasis	Suppresses MMP-driven ECM protein degradation; stimulates procollagen I synthesis (RAR-β/TGF-β axis); net shift in MMP:collagen balance	1

Pharmacokinetics

Topical (primary aging-context route)

• Penetration: Lipophilic character (logP 6.3) drives effective penetration through the stratum corneum into the viable epidermis. Accumulates in epidermis; dermal penetration is lower.
• Systemic absorption: Minimal — estimated 1–2% of topically applied dose reaches systemic circulation in adult skin under normal use conditions. Topical tretinoin does not meaningfully elevate plasma retinoid levels above endogenous baseline in adults.
• Local metabolism: Tretinoin is oxidized in keratinocytes to 4-oxo-retinoic acid and other polar metabolites; some local recycling to retinol. The active form (ATRA) is the predominant intracellular species after application.

Oral (APL context)

• Bioavailability: 50% (highly variable; food increases absorption)
• Plasma t½: ~0.7–1.4 h (the endogenous production rate means the drug clears rapidly; autoinduction of CYP26 further shortens t½ with repeated dosing — a major obstacle for chronic oral use)
• Protein binding: >95% to albumin and α1-acid glycoprotein
• Oral tretinoin is NOT used for photoaging; its PK characteristics (rapid clearance, autoinduction, teratogenicity risk) make it unsuitable for chronic dermatological use. The oral form is restricted to APL treatment.

Clinical indications

1. Acne vulgaris (FDA 1971)

The original indication. Tretinoin normalizes follicular keratinization (primary pathogenesis in comedonal acne), reduces comedone formation, and has anti-inflammatory effects via RAR-β. Concentrations: 0.025%, 0.05%, 0.1% cream; 0.01%, 0.025% gel. Photosensitivity and retinoid dermatitis are dose-limiting.

2. Photoaging / fine facial wrinkles (Renova, FDA 1995)

On-label aging indication. Evidence base includes Fisher 1996 (mechanism), Griffiths 1995 (clinical reversal of wrinkles/hyperpigmentation in vehicle-controlled RCT; the canonical clinical trial), and Sumita 2018 (RCT, formulation comparison). The Renova labeling specifies adjunctive use with sunscreen and moisturizer; tretinoin does not eliminate photoaging changes. Approved concentrations: 0.02% and 0.05% emollient cream formulations.

3. Acute promyelocytic leukemia (oral Vesanoid, FDA 1995)

APL carries the PML-RARα fusion gene. Pharmacological doses of oral ATRA overcome PML-RARα-mediated differentiation block, driving leukemic blasts to differentiate to granulocytes (differentiation therapy). ATRA + arsenic trioxide (ATO) is now first-line treatment in APL with >90% remission rate. This mechanism is distinct from the photoaging mechanism and is not covered further here.

4. Off-label dermatology

Hyperpigmentation and melasma (reduces melanin production via RAR-mediated transcriptional effects on melanogenesis); keratosis pilaris; striae distensae; other photoaged skin manifestations. Evidence for some off-label uses is robust (hyperpigmentation), others weaker.

Aging-context clinical evidence

Fisher 1996 — canonical mechanism in human skin in vivo

The foundational mechanistic study. Single-dose UVB applied to buttock skin of adult Caucasians; topical 0.1% t-RA applied 48 h prior. Direct human in vivo measurements of AP-1 binding, MMP-1/MMP-3/MMP-9 mRNA, protein, and activity. All quantitative figures (70% AP-1 reduction; 50–80% MMP-1/MMP-9 reduction) are TEXT-STATED in the paper body (pp.338-339, verified R39) ¹.

Dimension	Status
Pathway conserved in humans?	N/A — direct human study
Phenotype conserved in humans?	N/A — direct human study
Replicated in humans?	Yes — mechanism confirmed in multiple subsequent human in vivo studies by the same group and others

Sumita 2018 — RCT formulation comparison (n=24)

Randomized, evaluator-blinded RCT in 24 women (48 forearms; bilateral within-person design) comparing 0.05% tretinoin cream (3×/week) vs 5% tretinoin peel (every 2 weeks) for photoaged forearms. Both formulations produced ~20% reduction in photoaging scores and ~60% reduction in actinic keratosis counts from baseline (no between-arm difference on these primary endpoints). Three secondary endpoints diverged at p<0.05: dermis echogenicity and Ki67 favored the cream; field cancerization (AK pattern) favored the peel. Study page: sumita-2018-tretinoin-photoaging. Note: paper is closed-access; all quantitative claims are abstract-derived ³. no-fulltext-access

Chien 2022 — JAMA Dermatology RCT (n=20; tretinoin vs retinoid precursors)

Randomized clinical trial (n=20 White women, 24 weeks) comparing topical tretinoin precursors (retinol and derivatives) vs tretinoin in moderate-to-severe facial photodamage. No significant difference in Griffiths photoaging scores between groups. Tretinoin caused erythema 6× more frequently than precursors. Notably, MMP-2 mRNA was significantly reduced in precursor-treated samples and correlated with fine-wrinkle improvement — a mechanistically unexpected finding given Fisher 1996’s documentation that MMP-2 is UV/t-RA-independent. This warrants a dedicated follow-up. needs-replication ⁴

Siddiqui 2024 — systematic review (25 studies; tretinoin vs alternative topicals)

Systematic review of 25 studies comparing tretinoin to alternative topical photoaging therapies ⁵. Key findings: tretinoin is confirmed as “the gold standard therapy for photoaging.” Of 25 comparator studies: alternatives were more effective in 7, equivalent in 13, less effective in 3. Tretinoin’s primary disadvantage is tolerability — most alternatives have better short-term tolerability profiles. Retinaldehyde and pro-retinal nanoparticle formulations were identified as second-line options for tretinoin-intolerant patients.

Lin 2025 — network meta-analysis (23 RCTs; 3,905 participants)

Network meta-analysis of 23 RCTs (3,905 participants) comparing topical retinoids and other photoaging agents ⁶. Tretinoin and isotretinoin emerged as “the most balanced treatments across efficacy and safety.” Tretinoin and retinol were superior for hyperpigmentation; isotretinoin ranked highest for fine wrinkles overall. Tretinoin had the most favorable adverse-event profile overall; tazarotene and glycolic acid carried higher adverse-event risk. Note: Scientific Reports (2025); download pending; findings are from PubMed abstract. no-fulltext-access

Dose and concentration

Formulation	Concentration	Common schedule	Tolerability
Cream (Retin-A, generics)	0.025%, 0.05%, 0.1%	Nightly to every-other-night	Increases with concentration; 0.025% best tolerated
Emollient cream (Renova)	0.02%, 0.05%	Nightly; with moisturizer	Better than Retin-A at equivalent concentration due to vehicle
Gel	0.01%, 0.025%	Nightly	More drying than cream; acne-preferred vehicle
Lotion	0.05%	Nightly	Tested for chest photoaging in RCT (Wood 2022) 7

Induction protocol: typically start with lowest available concentration on alternating nights; advance to nightly and then higher concentrations as tolerance develops over 4–12 weeks.

Side effects and tolerability

Retinoid dermatitis (erythema, peeling, dryness, photosensitivity) affects the majority of new users; severity correlates with concentration and application frequency. Typically peaks at 4–8 weeks (“retinization period”) and partially resolves with continued use as epidermal adaptation occurs.

Teratogenicity. Tretinoin is a known teratogen:

• Topical: FDA pregnancy category D (risk not excluded; systemic absorption is minimal from topical use but not zero; FDA labeling advises avoidance in pregnancy)
• Oral Vesanoid: FDA pregnancy category X (major teratogen at pharmacological doses; APL treatment in pregnancy requires specialist counseling)

Women of childbearing potential using topical tretinoin should use effective contraception; the very low systemic absorption from topical application likely represents minimal teratogenic risk compared to oral vitamin A supplements at high doses, but formal safety data in pregnancy are unavailable.

Photosensitivity: Tretinoin increases UV sensitivity by thinning the stratum corneum. Daily broad-spectrum SPF ≥30 sunscreen is mandatory with topical tretinoin use for photoaging.

Drug interactions (topical): Concurrent benzoyl peroxide, salicylic acid, or glycolic acid increases irritation. Concurrent sulfur-containing products may inactivate tretinoin.

Autoinduction (oral only): Oral ATRA induces its own metabolic clearance via CYP26A1 upregulation with repeated dosing — relevant to APL therapy management, not to topical use.

Comparison to other retinoids

Retinoid	Mechanism	Conversion steps to ATRA	Potency vs tretinoin	Key feature
Tretinoin (ATRA)	Direct RAR agonist	0 (active form)	Reference standard	Gold standard; most evidence; most irritating
retinol (vitamin A)	RAR agonist (prodrug)	2 (retinol → retinal → ATRA)	~20× lower (OTC; 2-step conversion is rate-limiting)	OTC accessible; well-tolerated; less potent
retinaldehyde (retinal)	RAR agonist (prodrug)	1 (retinal → ATRA)	Intermediate; better tolerated than tretinoin	1-step conversion; better potency-tolerability tradeoff than retinol
adapalene	RAR-β/γ selective agonist	0 (synthetic; binds RAR directly)	Comparable to tretinoin 0.025%; less AP-1 transrepression	More selective receptor profile; better tolerability at equivalent clinical efficacy for acne; OTC in US since 2016
tazarotene	RAR-β/γ selective (prodrug: tazarotenic acid)	1 (hydrolysis to active acid)	Most potent topical retinoid; irritant	Most photodamage-efficacious per some comparisons; least tolerated
bakuchiol	Functional retinoid analog (not a retinoid structurally)	N/A (structurally distinct; activates retinoid gene expression program via distinct mechanism)	Approximately equivalent to retinol at tested doses per some RCTs	Plant-derived; suitable for retinoid-intolerant patients; not an RAR agonist structurally

See retinoids (R41 class MOC, forward reference) for the full retinoid family comparison.

Druggability and class context

Druggability tier: 1 (FDA-approved drug for an aging-indication — photoaging is on-label for Renova). This is a straightforward tier-1 assignment: Renova (tretinoin 0.05% cream) received FDA approval in 1995 specifically for photoaging, making it the only drug in this wiki with aging as a direct on-label indication (vs rapamycin, which is FDA-approved for transplant/cancer but not aging). No aging-context qualification needed.

Class MOC: retinoids — covers all retinoid family members, shared RAR/RXR mechanism, and family-level comparisons. (Forward reference — page not yet seeded; planned R41.)

SENS alignment: Tretinoin does not map cleanly to a SENS damage category — it modulates the signaling consequences of UV-induced damage rather than repairing a SENS-defined damage lesion. Closest alignment is sens-damage-categories extracellular crosslinks/junk outside cells (ECM integrity), though the mechanism is not direct crosslink removal.

Limitations and gaps

• Fisher 1996 limitations: Buttock skin only (photoprotected site); single UV dose per session; short time course (≤72 h); all-Caucasian cohort; small n per endpoint (n=7–10). Long-term prevention of cumulative collagen loss from repeat UV exposure not tested. long-term-unknown
• Sumita 2018: n=24; women only; Fitzpatrick II-III; forearm only (not face); closed-access PDF; abstract-derived quantitative claims only; no untreated control arm. no-fulltext-access needs-replication
• RAR-β-collagen-synthesis mechanism: The specific downstream pathway (RAR-β → TGF-β → Smad3 → COL1A1) has not been independently replicated in large human in vivo studies beyond the Voorhees/Fisher lab. needs-replication
• Chien 2022 MMP-2 finding: Unexpected reduction of MMP-2 by retinoid precursors (contradicting Fisher 1996’s documentation that MMP-2 is UV/t-RA-independent) requires follow-up. contradictory-evidence
• Epigenetic-age readouts absent: No published study has used a validated epigenetic clock (Horvath, DunedinPACE, GrimAge) as a secondary endpoint in a tretinoin RCT. The DHM/Qi 2026 precedent suggests this is feasible and would provide a cross-hallmark efficacy signal. needs-human-replication
• Long-term safety: Decades of clinical use support safety. No dedicated long-term (>5 year) RCT of continuous topical tretinoin use with hard structural endpoints (dermal biopsy collagen quantification) has been conducted. Carcinogenicity of topical use is not considered a concern (UV-protection mechanism is anti-mutagenic). long-term-unknown
• dose-response-unclear — optimal concentration and schedule for anti-aging (vs acne) endpoints are not definitively established; most RCT evidence uses 0.05% cream but comparative-concentration RCTs for aging-specific endpoints are limited.

Cross-references

• fisher-1996-photoaging-ap1-mmp — canonical mechanism; R39-verified; local PDF available
• sumita-2018-tretinoin-photoaging — RCT formulation comparison; n=24; closed-access
• skin-aging — phenotype page; photoaging mechanism and retinoid therapy section
• dermis — ECM aging context; MMP-driven collagen degradation
• epidermis — keratinocyte UV response
• dermal-fibroblasts — UV-SIPS and collagen synthesis context
• keratinocytes — AP-1/NF-κB → MMP cascade; actinic keratosis context
• mmp-1 — MMP-1 (interstitial collagenase); primary target of AP-1 transrepression
• mmp-3 — MMP-3 (stromelysin I, not II)
• mmp-9 — MMP-9 (92K gelatinase)
• col1a1 — procollagen I alpha-1; downstream synthesis target
• rara, rarb, rarg — receptor targets (forward refs; protein pages not yet seeded)
• retinol — OTC retinoid precursor; 2-step prodrug
• retinaldehyde — 1-step retinoid precursor; intermediate potency
• adapalene — RAR-β/γ-selective synthetic retinoid; acne + photoaging comparator
• tazarotene — RAR-β/γ-selective prodrug; most potent topical retinoid
• bakuchiol — structural retinoid analog; RCT-validated functional analog for retinoid-intolerant patients
• retinoids — R41 class MOC for the full retinoid family (forward ref; not yet seeded)
• genomic-instability — UV-induced MMP cascade + field cancerization context
• loss-of-proteostasis — ECM structural-protein degradation context

Footnotes

Footnotes

1. fisher-1996-photoaging-ap1-mmp · n=6–17 per experiment (varies by endpoint; see study page for per-assay n) · in-vivo · p<0.01 (t-RA vs vehicle for AP-1 and MMP endpoints) · model: human adult buttock skin (Caucasian); topical 0.1% t-RA pretreatment 48 h prior to 2 MED UVB · doi:10.1038/379335a0 · PMID 8552187 · Nature 1996;379:335-339 · locally available PDF (R39-verified end-to-end; all quantitative figures are text-stated, not figure-estimated) ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹

2. doi:10.1016/j.pharmthera.2020.107600 · Larange A, Cheroutre H · review of RAR isoform biology and retinoid receptor pharmacology — cited for RAR-α/β/γ binding affinities of ATRA; nanomolar Kd confirmed in multiple radioligand-binding studies · needs-canonical-id — confirm precise binding affinities from ChEMBL CHEMBL38 bioactivity page ↩

3. sumita-2018-tretinoin-photoaging · n=24 enrolled (23 per-protocol; bilateral within-person design; 48 forearms) · rct (evaluator-blinded) · p<0.05 (three secondary endpoints diverged between arms) · model: human women >60 yr, Fitzpatrick II-III, forearm · doi:10.1111/jdv.15020 · PMID 29704456 · J Eur Acad Dermatol Venereol 2018;32(10):1819-1826 · closed access; quantitative claims abstract-derived no-fulltext-access ↩ ↩²

4. doi:10.1001/jamadermatol.2022.1891 · Chien AL, Kim DJ, Cheng N, et al. · JAMA Dermatology 2022 · rct · n=20 · model: human women; facial photoaging, White; 24 weeks · tretinoin vs retinoid precursors; no photoaging-score difference; tretinoin 6× more erythema; unexpected MMP-2 mRNA reduction by precursors · PMC9178500 (OA); download pending ↩

5. doi:10.1007/s40257-024-00893-w · Siddiqui Z, Zufall A, Nash M, Rao D, Hirani R, Russo M · Am J Clin Dermatol 2024 · systematic-review · 25 studies · tretinoin confirmed as gold standard; more effective than alternatives in 3/25, equivalent in 13/25, less effective in 7/25; tolerability is primary limitation · PMID 39348007 · closed access; claims abstract-derived no-fulltext-access · FWCI 13.1; 99th percentile ↩

6. doi:10.1038/s41598-025-12597-0 · Lin L et al. · Scientific Reports 2025 · network meta-analysis · 23 RCTs; 3,905 participants · tretinoin + isotretinoin most balanced on efficacy + safety; tretinoin + retinol superior for hyperpigmentation; isotretinoin highest for fine wrinkles; tretinoin most favorable adverse-event profile overall (tazarotene + glycolic acid higher risk) · PMID 40707570 · OA (gold); download pending · claims abstract-derived no-fulltext-access ↩

7. doi:10.36849/JDD.6658 · Wood E, Fabi S et al. · J Drugs Dermatol 2022 · rct · tretinoin 0.05% lotion for chest photoaging · model: human (chest skin) · PMID 35674757 · claims abstract-derived no-fulltext-access ↩