Carnosic Acid

Cheap, daily-safe, BBB-crossing rosemary diterpene that is the cleanest "pro-electrophilic" Nrf2 activator we have — it's inert until oxidative stress oxidizes it, then it switches on the master antioxidant gene program only in damaged tissue. A-tier preclinical neuroprotection across multiple labs and models (Lipton 2025 diAcCA + 5xFAD; Schubert/Salk lineage; Satoh 2008 Keap1 mechanism); zero published human cognitive RCTs. For Dylan: optional cheap-insurance layer for MMA subconcussive-impact concern, stacks clean with curcumin (already V4), astaxanthin (V5), NAC (V4). 60-180 mg standardized rosemary extract or pure carnosic acid with breakfast (fat-containing meal); subjective effects subtle-to-undetectable; this is rust-proofing, not a felt nootropic.

Carnosic acid (CA, MW 332.4 Da) is a phenolic abietane-type diterpene that accounts for most of the antioxidant activity of rosemary (Rosmarinus officinalis) and common sage (Salvia officinalis). It is the major reason rosemary extract has been a permitted food-preservation antioxidant in the EU and US for decades, and the reason rosemary leaf is on the FDA GRAS list. Mechanistically, it is the most pharmacologically interesting compound in the food-derived antioxidant family — and the one with the cleverest pharmacology.

The headline mechanism: carnosic acid is a proelectrophilic drug (PED). This is unusual and important enough to spell out:

• A normal "electrophilic" Nrf2 activator (sulforaphane, dimethyl fumarate, bardoxolone methyl) is electrophilic as administered. It will alkylate cysteines on Keap1 wherever it goes — including healthy tissue — driving generic Nrf2 activation and consuming glutathione system-wide. That's why high-dose sulforaphane and bardoxolone-class drugs have GI tolerability and off-target toxicity issues.
• Carnosic acid as administered is a catechol (two adjacent -OH groups on its aromatic ring), which is not electrophilic. It cannot alkylate cysteines. It just sits there.
• When CA encounters reactive oxygen species (ROS) — i.e., in tissue undergoing oxidative or inflammatory damage — the catechol gets oxidized to an ortho-quinone form that IS electrophilic. The quinoid form then S-alkylates targeted cysteines (most notably C151, C273, C288) on Keap1, the cytoplasmic repressor of Nrf2.
• Once Keap1 is alkylated on those cysteines, it can no longer bind Nrf2 for proteasomal degradation. Nrf2 escapes, translocates to the nucleus, and turns on the antioxidant response element (ARE) gene battery: heme oxygenase-1 (HO-1), NQO1, glutamate-cysteine ligase catalytic subunit (GCLC), glutathione peroxidase, glutathione S-transferases, and ~200 other phase-2 detoxification and antioxidant genes.

The clinical implication is the elegant part: CA is essentially inert in healthy tissue and turns on only where damage is occurring. It's a damage-targeted antioxidant amplifier rather than a generic radical-scavenger. This is why preclinical work shows efficacy without the off-target Nrf2-activator concerns that plagued bardoxolone (BEACON trial, cardiovascular signal in CKD).

Beyond the proelectrophilic Keap1/Nrf2 trick, CA has several layered mechanisms relevant to brain health:

1. HSF-1 (heat shock factor 1) activation — CA also induces the chaperone response (HSP70, HSP90), which helps refold misfolded proteins and clear protein aggregates. This is mechanistically connected to the amyloid-β and phospho-tau reductions seen in 5xFAD mice.

2. NLRP3 inflammasome suppression — CA suppresses NLRP3 assembly and IL-1β release in microglia and macrophages. This is the canonical pathway by which neuroinflammation accelerates neurodegeneration; CA's NLRP3 effect is the proposed link to long-COVID, Parkinson's, and AD applications (de Oliveira 2022 review, PMC8772720).

3. Microglial NF-κB attenuation — CA reduces pro-inflammatory cytokine production (TNF-α, IL-6) in activated microglia, complementing the Nrf2-driven antioxidant arm.

4. Lipid-membrane antioxidant action — CA's lipophilicity allows it to partition into membrane lipid bilayers and quench lipid peroxyl radicals directly. This overlaps astaxanthin's mechanism but at a different chemical level (CA is not membrane-spanning; it acts as a mobile membrane antioxidant similar to vitamin E or idebenone).

5. BBB crossing — CA crosses the blood-brain barrier passively due to its small molecular weight (332 Da) and lipophilicity. Confirmed in rat oral PK studies: detectable in brain, with cerebellum and cortex levels rising dose-proportionally. This distinguishes it from rosmarinic acid (the other major rosemary polyphenol), which crosses the BBB much more poorly.

6. Mild mitochondrial biogenesis / AMPK activation — CA upregulates PGC-1α, TFAM, and AMPK in adipocyte and hepatocyte models (Bahaeddin 2024 brown-fat phenotype paper). The brain-specific magnitude of this effect is uncertain.

7. Glutathione synthesis support — Via Nrf2-driven GCLC induction, CA increases cellular glutathione pools. Documented protection in 6-OHDA-treated SH-SY5Y cells (Park 2012, dopaminergic neurodegeneration model).

8. 6-OHDA / α-synuclein / MPTP protection (Parkinson's models) — Multiple labs have shown CA protects dopaminergic neurons in Parkinson's-relevant models. Mechanism overlaps: Nrf2-driven glutathione restoration + lipid peroxidation prevention.

One important nuance: dose-response is not strictly linear at the cellular level. Like many electrophilic Nrf2 activators, CA has a dose window — too low fails to push the redox threshold for activation; too high (in vitro >50 μM) becomes pro-oxidant via the quinone form. The proelectrophilic mechanism makes the in vivo therapeutic window much wider than for direct electrophiles, but the bell-shape exists.

Schubert / Salk lineage: David Schubert (Salk Institute, deceased 2020) led a body of work in the 2000s-2010s screening natural products for "geroneuroprotectors" — compounds that protect aged neurons. His group identified CA and curcumin derivatives (CNB-001, J147) as priority leads, and his screening framework is the lineage that produced the Nrf2/proelectrophilic mechanistic story. Stuart Lipton (then at Scripps/Sanford-Burnham, now at Scripps Research) was a longtime collaborator on the Nrf2 mechanism work and is the senior author of the 2025 diAcCA / 5xFAD paper that triggered the current wave of attention (see §Evidence). Lipton's broader career includes co-discovery of memantine's mechanism — that's the "memantine team" angle in the original brief.