Galdosol

Galdosol at a Glance

Galdosol is an abietane diterpene — C₂₀H₂₄O₅, molecular weight 344.4, CAS 52591-18-1, PubChem CID 13966127 ^{13Reference 13PubChem Compound Summary: Galdosol et al.National Library of MedicineView study →}. It is the last compound anyone has put a name to at the end of the oxidation cascade that begins with carnosic acid, the principal antioxidant of rosemary and sage.

Its place in that cascade is worth stating precisely, because it explains the molecule. Carnosic acid oxidises to carnosol, whose lactone runs C-20→C-7 as a six-membered ring. Carnosol opens onward to the δ-lactones rosmanol, epirosmanol and isorosmanol, which carry a hydroxyl at C-7. Galdosol is what happens when that hydroxyl is oxidised one step further, to a ketone ^{5,13,16Reference 5Marrero JG et al. · 2002In vitroSemisynthesis of rosmanol and its derivativesView study →Reference 13PubChem Compound Summary: Galdosol et al.National Library of MedicineView study →Reference 16Loussouarn M et al. · 2017In vitroCarnosic acid and carnosol, two major antioxidants of rosemary, act through different mechanisms — in vitro and in plantaView study →}. The consequence is neat: rosmanol and epirosmanol differ only in the configuration of the C-7 hydroxyl, and a ketone has no configuration. The two epimers converge on galdosol. The stereochemical distinction that the 1984 rosemary literature was built around simply disappears at this step.

Two facts should frame everything that follows.

First, this is one of the smallest literatures here. PubMed indexes five records that mention galdosol. Adding the primary chemistry papers that PubMed’s title-and-abstract index misses entirely — recovered here through Crossref and Europe PMC full-text mining — the whole literature is about ten papers, four of which put purified galdosol into a biological assay. There is no review of the molecule, no animal study, no pharmacokinetics, no toxicology, and no clinical trial. All ten are printed below.

Second, one of those five records is genuinely surprising. In a benzodiazepine-receptor binding assay-guided fractionation of sage, galdosol showed the strongest binding of any compound isolated — IC₅₀ 0.8 µM against ³H-flumazenil at the human brain benzodiazepine receptor ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}. That is tighter than the flavone hispidulin (1.3 µM) from the same experiment, tighter than apigenin (30 µM), and roughly 400-fold tighter than cirsimaritin. The same group took hispidulin forward to a full functional characterisation — receptor subtypes, an animal seizure model, blood–brain-barrier transit — and published it in the British Journal of Pharmacology the following year ^{11Reference 11Kavvadias D et al. · 2004AnimalThe flavone hispidulin, a benzodiazepine receptor ligand with positive allosteric properties, traverses the blood-brain barrier and exhibits anticonvulsive effects — gerbils and ratsView study →}. They never did any of that for galdosol. Twenty-three years later, nobody else has either.

The Whole Literature, Listed

Ten papers is small enough to print rather than summarise. Note how many of them PubMed’s galdosol search does not return — a reminder that in 1970s–90s Spanish and Japanese abietane chemistry, “no PubMed record” and “no evidence” are different statements.

Year	Study	What it did with galdosol	In PubMed’s galdosol search?	Ref
1973	González, Fraga, Luis & Ravelo	First isolation and structure, from Salvia canariensis	No	1Reference 1González AG et al. · 1973Constituents of Labiatae XIXView study →
1983	Escudero et al.	Reported among diterpenoids of S. oxyodon and S. lavandulifolia	No	22Reference 22Escudero J et al. · 1983Diterpenoids from Salvia oxyodon and Salvia lavandulifoliaView study →
1989	González et al.	Isolated from S. canariensis aerial parts; reaction with diazomethane gave unexpected homoditerpenes	No	6Reference 6González AG et al. · 1989Abietane diterpenes with antibiotic activity from the flowers of Salvia canariensisView study →
1990	Darias et al.	Cytostatic ID₅₀ 0.5 µg/mL — most active of 13 Canarian Lamiaceae isolates; moderate Gram-positive antibacterial	Yes	4Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →
1992	Djarmati et al.	NMR and X-ray crystallography of galdosol 5-methyl ether from S. officinalis	No	7Reference 7Djarmati Z et al. · 1992NMR studies and X-ray crystallography of galdosol 5-methyl ether from Salvia officinalis LView study →
2002	Miura, Kikuzaki & Nakatani	Isolated from sage leaf; antioxidant activity comparable to α-tocopherol by oil stability index	Yes	3Reference 3Miura K et al. · 2002In vitroAntioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitroView study →
2002	Marrero, Andrés & Luis	Semisynthesis from carnosol	Yes	5Reference 5Marrero JG et al. · 2002In vitroSemisynthesis of rosmanol and its derivativesView study →
2012	Lima et al.	Isolated from Medusantha carvalhoi (Lamiaceae, Brazil)	No	8Reference 8Lima KSB de et al. · 2012Abietane diterpenes from Hyptis carvalhoi HarleyView study →
2017	Bisio et al.	One of 13 diterpenes from a S. chamaedryoides surface extract; antibacterial and α-glucosidase panel	No	9Reference 9Bisio A et al. · 2017In vitroAntibacterial and hypoglycemic diterpenoids from Salvia chamaedryoides — in vitroView study →
2022	Ríos, Schinella & Moragrega	Review; nominates galdosol as a priority target for future GABAergic drug research	Yes	10Reference 10Ríos JL et al. · 2022ReviewPhenolics as GABA_A receptor ligands — an updated reviewView study →

Plus the 2003 benzodiazepine paper ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}, which is the fifth PubMed record and the subject of the next section.

Four of those ten are isolate pharmacology ^{2,3,4,9Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →Reference 3Miura K et al. · 2002In vitroAntioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitroView study →Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →Reference 9Bisio A et al. · 2017In vitroAntibacterial and hypoglycemic diterpenoids from Salvia chamaedryoides — in vitroView study →}. Four are pure chemistry ^{1,5,6,7Reference 1González AG et al. · 1973Constituents of Labiatae XIXView study →Reference 5Marrero JG et al. · 2002In vitroSemisynthesis of rosmanol and its derivativesView study →Reference 6González AG et al. · 1989Abietane diterpenes with antibiotic activity from the flowers of Salvia canariensisView study →Reference 7Djarmati Z et al. · 1992NMR studies and X-ray crystallography of galdosol 5-methyl ether from Salvia officinalis LView study →}. Two are occurrence reports ^{8,22Reference 8Lima KSB de et al. · 2012Abietane diterpenes from Hyptis carvalhoi HarleyView study →Reference 22Escudero J et al. · 1983Diterpenoids from Salvia oxyodon and Salvia lavandulifoliaView study →}. Nothing has been replicated by an independent laboratory except the observation that the molecule exists.

The Benzodiazepine-Receptor Result

This is the reason the page exists, and it deserves both the enthusiasm and the discipline.

Kavvadias and colleagues at Würzburg ran a benzodiazepine-receptor binding assay-guided fractionation of a methanol extract of sage leaves — meaning the receptor affinity, not a chemical hunch, led them to the compounds. Five active constituents emerged: three flavones and two abietane diterpenes. They then measured competitive inhibition of ³H-flumazenil binding to the human brain benzodiazepine receptor ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}:

Compound	Class	IC₅₀
Galdosol	Abietane diterpene	0.8 µM
Hispidulin	Flavone	1.3 µM
7-Methoxyrosmanol	Abietane diterpene	7.2 µM
Apigenin	Flavone	30 µM
Cirsimaritin	Flavone	350 µM

Galdosol was the most potent binder in the study. A 2022 review of phenolics as GABA_A ligands, surveying the whole field, singled it out by name as one of a short list of compounds that “deserve to be considered as future subjects of research” ^{10Reference 10Ríos JL et al. · 2022ReviewPhenolics as GABA_A receptor ligands — an updated reviewView study →}.

Now the discipline, because four separate caveats sit on that number and none of them has been resolved.

Binding is not function. Displacing flumazenil tells you a molecule occupies the benzodiazepine site. It does not tell you what the molecule does there — positive allosteric modulator, antagonist and inverse agonist all displace the same radioligand, and their effects on a person run in opposite directions. No electrophysiology, no chloride-current measurement and no behavioural experiment has ever been performed with galdosol. Nothing here supports calling it sedative, anxiolytic or anticonvulsant, and actions: here deliberately omits all three.

The comparison inside the same paper is the tell. Hispidulin, the second-place binder, was carried forward the following year: expressed at recombinant α₁₋₃,₅,₆β₂γ₂S GABA_A receptors in Xenopus oocytes, where it stimulated GABA-induced chloride currents from 50 nM (a positive allosteric profile); fed to seizure-prone Mongolian gerbils at 10 mg/kg/day, where it cut seizure incidence from 80% to 30%; and tracked across the blood–brain barrier by ¹⁴C in-situ perfusion, with an uptake rate approaching diazepam’s ^{11Reference 11Kavvadias D et al. · 2004AnimalThe flavone hispidulin, a benzodiazepine receptor ligand with positive allosteric properties, traverses the blood-brain barrier and exhibits anticonvulsive effects — gerbils and ratsView study →}. That is what a followed-up hit looks like. Galdosol — the tighter binder — received none of it. The literature does not say why.

A catechol in a binding assay is a flag. Galdosol carries an ortho-dihydroxy (catechol) ring, the same group that drives its antioxidant activity ^{3Reference 3Miura K et al. · 2002In vitroAntioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitroView study →}. Catechols are redox-active, oxidise to reactive quinones, and are a recognised source of frequent-hitter behaviour in binding screens. This is not a reason to dismiss the result — it is a reason the confirmatory experiments were needed and never done.

And the compound may partly have been made during the experiment. The extract was methanolic ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}, and methanol is precisely the solvent class in which the carnosic acid cascade runs forward during workup ^{14,15Reference 14Paloukopoulou C · 2022In vitroA validated method for the determination of carnosic acid and carnosol in the fresh foliage of Salvia rosmarinus and Salvia officinalis from Greece — in vitroView study →Reference 15Birtić S et al. · 2015ReviewCarnosic acid — review of distribution, biosynthesis, accumulation and role in plantaView study →}. See the next section.

Is Galdosol Actually in the Plant?

Because galdosol sits at the far end of an oxidation cascade, the honest first question is whether the plant makes it or the chemist does. This site has now found both answers in the same family: rosmaricine turned out to be an outright ammonia extraction artefact, while carnosol — suspected on identical grounds — was confirmed at 1.0–4.5 mg/g in liquid-nitrogen-frozen fresh rosemary leaf ^{14Reference 14Paloukopoulou C · 2022In vitroA validated method for the determination of carnosic acid and carnosol in the fresh foliage of Salvia rosmarinus and Salvia officinalis from Greece — in vitroView study →}.

For galdosol the answer is: unresolved, and nobody has run the experiment that would resolve it. What we have is evidence pulling in both directions.

The verdict taken here is that galdosol is a real molecule with a real, if thin, occurrence record, and that its level in any given preparation is almost certainly a function of how that preparation was made — more so than for carnosol, because it sits two steps further along the same one-way reaction. It is best understood the way epirosmanol is: a marker of how much oxidative processing the tissue and the extract have undergone.

The experiment that would settle it is small and obvious: quantify galdosol alongside carnosic acid, carnosol and rosmanol in liquid-nitrogen-frozen fresh sage leaf extracted in an aprotic solvent, and again in the same leaf after drying, storage and methanolic extraction. Nobody has done it. Until somebody does, treat every galdosol result — including the 0.8 µM one — as a result obtained on a compound of partly uncertain provenance.

What It's Used For

Four studies have put purified galdosol into a biological assay. They found four different things, none of which has been checked by anyone else.

Evidence by IndicationStrength of support · Beta

Support is an experimental score — weighted by study type (human > animal > in vitro) and volume. A beta, not a validated metric. With a ten-paper literature, zero animal studies and zero replication, every score here is capped low by design. Each application links to its write-up.

01AntioxidantShaky26%One isolate panel, plus a well-understood catechol mechanism shared with the whole series ^{3Reference 3Miura K et al. · 2002In vitroAntioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitroView study →}

02Benzodiazepine-Receptor BindingShaky18%One isolate binding assay; potent, unreplicated, no functional data ^{2,10Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →Reference 10Ríos JL et al. · 2022ReviewPhenolics as GABA_A receptor ligands — an updated reviewView study →}

03AntibacterialShaky12%Panel-level 1990 report plus one MIC range known only through a review table ^{4,9,12Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →Reference 9Bisio A et al. · 2017In vitroAntibacterial and hypoglycemic diterpenoids from Salvia chamaedryoides — in vitroView study →Reference 12Saha P et al. · 2021ReviewAntimicrobial diterpenes — recent development from natural sources. [Review.] Frontiers in Pharmacology, 12, 820312. https://pubmed.ncbi.nlm.nih.gov/35295739/View study →}

04CytostaticShaky10%A single 1990 ID₅₀ with no cell line named and no selectivity control ^{4Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →}

Mechanisms

Pharmacokinetics

There is none. No absorption figure, no oral bioavailability, no plasma concentration, no half-life, no tissue distribution, no metabolism study, in any species. galdosol AND pharmacokinetics returns zero records anywhere.

This gap matters more here than it usually does, because the one interesting finding is a central nervous system binding result ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}. A benzodiazepine-site ligand is only interesting if it reaches the brain, and nothing is known about whether galdosol survives ingestion, let alone crosses the blood–brain barrier. The instructive comparison is again hispidulin, which was measured directly by ¹⁴C in-situ rat brain perfusion and shown to enter at a rate approaching diazepam’s ^{11Reference 11Kavvadias D et al. · 2004AnimalThe flavone hispidulin, a benzodiazepine receptor ligand with positive allosteric properties, traverses the blood-brain barrier and exhibits anticonvulsive effects — gerbils and ratsView study →}. That measurement is exactly what galdosol lacks.

What can be said indirectly: galdosol is more oxidised, more polar and lower in molecular weight than its carnosic acid parent, which is directionally favourable for passive membrane transit — the same argument made for epirosmanol, which showed ~50% permeability in an artificial-membrane blood–brain-barrier model where carnosic acid and carnosol showed almost none. That is an inference about a sibling, not a measurement of this molecule.

Clinical Trials

None. A ClinicalTrials.gov search for galdosol returns 0 studies (checked August 2026). Human exposure is entirely incidental, through culinary sage and concentrated Salvia extracts.

Last checked: August 2026.

Isolate vs. Plant

The boundary is unusually clean here, because there is almost nothing on either side.

Genuine isolate evidence: all four pharmacology papers tested purified galdosol ^{2,3,4,9Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →Reference 3Miura K et al. · 2002In vitroAntioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitroView study →Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →Reference 9Bisio A et al. · 2017In vitroAntibacterial and hypoglycemic diterpenoids from Salvia chamaedryoides — in vitroView study →}. In three of the four it was one entry in a comparative table; in the fourth ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →} it was the winner. That is better isolate discipline than most plant molecules manage, and the page credits it.

What must not be transferred in. Galdosol inherits nothing from upstream. Carnosic acid’s Keap1/Nrf2 neuroprotection, its selective mPGES-1 inhibition, its AMPK and lipase effects, and the whole E 392 food-additive regulatory identity belong to carnosic acid and carnosol. The EU specification for rosemary extract is written as the sum of carnosic acid plus carnosol — galdosol is not named in it, is not measured in commerce, and has no regulatory status of any kind.

What must not be transferred out. Nothing here supports a claim about sage or rosemary as herbs. In particular, sage’s traditional and modern reputation in cognition and mood has been attributed by some reviews to its benzodiazepine-receptor-active constituents — but the compound with the functional and in-vivo evidence behind that story is the flavone hispidulin ^{11Reference 11Kavvadias D et al. · 2004AnimalThe flavone hispidulin, a benzodiazepine receptor ligand with positive allosteric properties, traverses the blood-brain barrier and exhibits anticonvulsive effects — gerbils and ratsView study →}, not galdosol. Galdosol has an affinity number and nothing else.

One near-name entity that is not this molecule. Galdosol 5-methyl ether, isolated from S. officinalis and solved by X-ray crystallography ^{7Reference 7Djarmati Z et al. · 1992NMR studies and X-ray crystallography of galdosol 5-methyl ether from Salvia officinalis LView study →}, is a distinct compound. Its data do not transfer, in either direction.

Prevalence in Nature

Galdosol is a Lamiaceae compound and, within the family, almost entirely a Salvia compound. Its distribution follows carnosic acid’s, because it is made from it.

Source	Part	Basis
Canary Island sage (Salvia canariensis)	Aerial parts, flowers	Type source — first isolation and structure 1973 1Reference 1González AG et al. · 1973Constituents of Labiatae XIXView study →; re-isolated from aerial parts 1989 6Reference 6González AG et al. · 1989Abietane diterpenes with antibiotic activity from the flowers of Salvia canariensisView study →
Sage (Salvia officinalis)	Leaf	Isolated from leaf among 23 constituents 3Reference 3Miura K et al. · 2002In vitroAntioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitroView study →; recovered by receptor-binding-guided fractionation 2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →; the 5-methyl ether solved by X-ray 7Reference 7Djarmati Z et al. · 1992NMR studies and X-ray crystallography of galdosol 5-methyl ether from Salvia officinalis LView study →
Salvia chamaedryoides	Aerial-part surface exudate	One of 13 diterpenes from a surface extract 9Reference 9Bisio A et al. · 2017In vitroAntibacterial and hypoglycemic diterpenoids from Salvia chamaedryoides — in vitroView study →
Medusantha carvalhoi (syn. Hyptis carvalhoi)	Aerial parts	Single Brazilian report, with rosmanol, methylrosmanol, 7α-ethoxyrosmanol and epi-isorosmanol 8Reference 8Lima KSB de et al. · 2012Abietane diterpenes from Hyptis carvalhoi HarleyView study →
Spanish sage (Salvia lavandulifolia)	Aerial parts	Single 1983 report 22Reference 22Escudero J et al. · 1983Diterpenoids from Salvia oxyodon and Salvia lavandulifoliaView study →; repeated in a 2021 review with a broken citation, so treat as provisional

No concentration figure exists. Not a mg/g, not a % dry weight, not a peak-area percentage, for any source.

Galdosol is not, on the primary record, a rosemary compound. A PubMed search for galdosol AND ("Rosmarinus officinalis" OR "Salvia rosmarinus") returns zero records, and every full-text co-mention of galdosol with rosemary is a review or an introductory sentence restating the carnosic acid cascade — not a detection ^{17,18Reference 17Luca SV et al. · 2023In vitroChemical profile and bioactivity evaluation of Salvia species from Eastern Europe — in vitroView study →Reference 18Etsassala NGER et al. · 2021ReviewPotential application of some Lamiaceae species in the management of diabetes. [Review.] Plants, 10(2), 279. https://pubmed.ncbi.nlm.nih.gov/33535455/View study →}. Given that galdosol forms from carnosol, and rosemary is the richest source of the parent, its presence in rosemary is chemically likely; it has simply never been reported. Anyone writing “galdosol, a rosemary diterpene” is extrapolating.

Two annotations to discount. Untargeted metabolomics has recently placed galdosol in two implausible plants. A 2025 study annotated it in Ephedra foeminea branches by exact mass ^{19Reference 19Bennici G et al. · 2025Medicinal effects of Ephedra foeminea aqueous extracts — metabolomic characterization, biological evaluation, and molecular dockingView study →} — a gnetophyte with no known abietane catechol pathway. And a 2024 multi-omics study annotated it in sweet pepper (Capsicum annuum), a Solanaceae ^{20Reference 20Wang J et al. · 2024Sweet pepper extract reduces fat storage in Caenorhabditis elegans by the SREBP-SCD axis based on multiomics analysisView study →}; that table’s own numbers give it away, reporting an observed and calculated m/z of 343.228 for the [M−H]⁻ ion of C₂₀H₂₄O₅, which is 343.155. A 0.073 Da discrepancy is not a measurement error, it is a mis-assignment. Both are database false positives of the kind that follow obscure natural products around, and neither should be cited as an occurrence.

Its Role in Plants

Galdosol has no dedicated biosynthetic enzyme, and almost certainly no dedicated function. It is a terminal product of the sacrificial antioxidant chemistry that carnosic acid performs in the chloroplast.

The pathway upstream of it is fully solved: geranylgeranyl diphosphate → miltiradiene → ferruginol → carnosic acid, by way of four Salvia cytochrome P450s. What happens next is not enzymology but chemistry. Carnosic acid quenches reactive oxygen species and is consumed in the act, its catechol oxidising to an ortho-quinone that is intercepted by the C-20 carboxyl to give carnosol, and thence onward to the rosmanol series and to galdosol ^{15,16Reference 15Birtić S et al. · 2015ReviewCarnosic acid — review of distribution, biosynthesis, accumulation and role in plantaView study →Reference 16Loussouarn M et al. · 2017In vitroCarnosic acid and carnosol, two major antioxidants of rosemary, act through different mechanisms — in vitro and in plantaView study →}. Oxidised carnosic acid derivatives accumulate in leaves under stress and are detectable even under low light, a signature of chronic background turnover ^{16Reference 16Loussouarn M et al. · 2017In vitroCarnosic acid and carnosol, two major antioxidants of rosemary, act through different mechanisms — in vitro and in plantaView study →}.

Galdosol therefore sits at the far end of a one-way reaction that measures oxidative history — of the living leaf, and then of everything that happened to it afterwards. That is the best available explanation for why the compound is so scarce, so patchily reported, and so resistant to quantification: its level is not a plant trait so much as a running total.

Discovery & Synthesis

Galdosol was isolated and its structure established in 1973 by Antonio González, Braulio Fraga, Javier Luis and Ángel Ravelo at the Instituto Universitario de Bio-Orgánica Antonio González, University of La Laguna, Tenerife, from the Canary Islands endemic sage Salvia canariensis ^{1Reference 1González AG et al. · 1973Constituents of Labiatae XIXView study →}. The report appeared in Experientia as the nineteenth paper in the group’s long Constituents of Labiatae series — a body of Canarian phytochemistry that also gave the neighbouring diterpene salviol its modern description.

The same group returned to the molecule in 1989, isolating it again from the aerial parts of S. canariensis alongside the new diterpenes canariquinone, rosmaquinone and 11-acetoxycarnosic acid from the flowers, and reporting that its reaction with diazomethane unexpectedly gave two new homoditerpenes rather than the expected methyl ether ^{6Reference 6González AG et al. · 1989Abietane diterpenes with antibiotic activity from the flowers of Salvia canariensisView study →}. In 1992 a Serbian group isolated galdosol 5-methyl ether from Salvia officinalis and solved it by NMR and X-ray crystallography ^{7Reference 7Djarmati Z et al. · 1992NMR studies and X-ray crystallography of galdosol 5-methyl ether from Salvia officinalis LView study →} — the only crystallographic work in this molecule’s literature.

Structurally, galdosol is the point at which the rosmanol epimers converge. Rosmanol and epirosmanol are C₂₀H₂₆O₅ and differ only in the configuration at C-7 (7α versus 7β). Galdosol is C₂₀H₂₄O₅ — the same skeleton with that carbon oxidised from a secondary alcohol to a ketone, which erases the stereocentre altogether ^{13Reference 13PubChem Compound Summary: Galdosol et al.National Library of MedicineView study →}. It retains the catechol on the aromatic ring and the δ-lactone bridge, which is why it keeps its siblings’ antioxidant chemistry ^{3Reference 3Miura K et al. · 2002In vitroAntioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitroView study →}.

There is no total synthesis. There is a practical semisynthesis: Marrero, Andrés and Luis obtained rosmanol, rosmaquinone, 7-methoxyrosmanol, 7-ethoxyrosmanol, galdosol and epirosmanol by partial synthesis from carnosol, an abundant natural product, with physical and spectroscopic data identical to authentic natural samples ^{5Reference 5Marrero JG et al. · 2002In vitroSemisynthesis of rosmanol and its derivativesView study →}. Their stated motivation — that these compounds occur in Salvia “in low quantities” — remains the clearest published statement about galdosol’s natural abundance, and it has no number attached to it.

Commercially, galdosol is sold only as a milligram-scale analytical reference standard. It is not produced at scale, is not a component of any commercial product, and does not appear in any regulatory specification.

Unusually for this batch’s molecules, galdosol has no synonym problem: PubChem lists no alternative chemical name for CID 13966127 beyond registry identifiers ^{13Reference 13PubChem Compound Summary: Galdosol et al.National Library of MedicineView study →}, and no second name for it appears anywhere in the literature. A searcher who misses this molecule misses it because there is so little to find, not because it is filed under something else.

Patents: not yet researched.

Toxicity & Safety

The honest verdict is not characterised, and the unknown grading is deliberate rather than reassuring. There has never been a toxicology study of galdosol of any description: no acute or repeat-dose study, no LD₅₀, no genotoxicity panel, no reproductive or developmental work, no drug-interaction assessment, and no administration to a living animal by any route. low would imply a clean record; what exists is an absent one.

Two in-vitro observations are worth holding, and both point in the cautious direction rather than the benign one.

It reduces cell viability at low micromolar concentrations. The 1990 cytostatic ID₅₀ of 0.5 µg/mL (~1.5 µM) is the most potent number attached to this molecule, and it was measured as an anticancer endpoint — but with no normal-cell comparator, a viability figure that low is equally readable as general cytotoxicity ^{4Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →}. Its close relative epirosmanol was later found cytotoxic to normal rat myoblasts at ~14 µM with a selectivity index of 1.6, which is the more informative precedent.

A benzodiazepine-site ligand with unknown directionality is not a neutral finding. Galdosol binds the benzodiazepine site tightly ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}. Because nobody has determined whether it modulates the receptor positively, negatively or not at all, the possibility of a CNS effect — in either direction, including a pro-convulsant one — is open rather than excluded. This is a reason to be cautious about concentrated Salvia diterpene preparations in anyone taking benzodiazepines or antiepileptics, and it is a reason nothing here should be read as a sedative claim.

Reproductive and developmental effects have never been investigated for this molecule. Decisions about sage or rosemary in pregnancy belong on the sage and rosemary monographs, where intake is actually decided, and turn on thujone and on whole-preparation data — not on this compound.

Safety at a Glance

❔ Toxicity unknown

• No adverse effect has ever been reported. It has also never been looked for; the unknown grading reflects the second statement, not the first.
• Dietary exposure is long-standing and incidental. Anyone who cooks with sage has consumed trace amounts indefinitely, with no signal attached — though nobody has ever measured how much.
• Cytostatic at ~1.5 µM in vitro with no selectivity control ^{4Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →}. For a redox-active catechol, that is as easily read as general cytotoxicity as it is as anticancer activity.
• Binds the benzodiazepine site at 0.8 µM with unknown functional direction ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}. Positive modulation, antagonism and inverse agonism all look identical in that assay and do not have identical consequences.
• No toxicology, no pharmacokinetics, no animal data, no human data. Every statement here comes from a dish or a binding tube.
• No reproductive or developmental data exist. Nothing is known about use in pregnancy or breastfeeding.
• Drug interactions have never been assessed — no CYP, no transporter, no protein-binding work of any kind.
• Provenance is unresolved. No study has established how much galdosol is in fresh plant material versus how much forms during extraction and storage.

Dosage

There is no dose. Galdosol has never been administered to a person or an animal, in any amount, by any route. No pharmacokinetic study has established whether an oral dose would reach the bloodstream, and no toxicology study has established what would happen if it did. Every published figure is an in-vitro concentration — 0.8 µM at a receptor ^{2Reference 2Kavvadias D et al. · 2003In vitroConstituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitroView study →}, ~1.5 µM at a cell ^{4Reference 4Darias V et al. · 1990In vitroCytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitroView study →}, 93–186 µM against a bacterium ^{9,12Reference 9Bisio A et al. · 2017In vitroAntibacterial and hypoglycemic diterpenoids from Salvia chamaedryoides — in vitroView study →Reference 12Saha P et al. · 2021ReviewAntimicrobial diterpenes — recent development from natural sources. [Review.] Frontiers in Pharmacology, 12, 820312. https://pubmed.ncbi.nlm.nih.gov/35295739/View study →} — which is the concentration in a well, not an intake.

Real-world exposure comes from sage leaf and from concentrated Salvia extracts, in which galdosol is an unquantified trace member of a diterpene fraction defined, regulated and standardised around carnosic acid and carnosol. The plant, not the molecule, is what anyone is actually taking. Nothing here is a recommendation.

References

1. González AG, Fraga BM, Luis JG, Ravelo AG. (1973). Constituents of Labiatae XIX. Structure of galdosol, a new diterpene from Salvia canariensis L. Experientia, 29, 1471. https://doi.org/10.1007/BF01943861
2. Kavvadias D, Monschein V, Sand P, Riederer P, Schreier P. (2003). Constituents of sage (Salvia officinalis) with in vitro affinity to human brain benzodiazepine receptor — in vitro. Planta Medica, 69(2), 113–117. https://pubmed.ncbi.nlm.nih.gov/12624814/
3. Miura K, Kikuzaki H, Nakatani N. (2002). Antioxidant activity of chemical components from sage (Salvia officinalis L.) and thyme (Thymus vulgaris L.) measured by the oil stability index method — in vitro. Journal of Agricultural and Food Chemistry, 50(7), 1845–1851. https://pubmed.ncbi.nlm.nih.gov/11902922/
4. Darias V, Bravo L, Rabanal R, Sánchez-Mateo CC, Martín-Herrera DA. (1990). Cytostatic and antibacterial activity of some compounds isolated from several Lamiaceae species from the Canary Islands — in vitro. Planta Medica, 56(1), 70–72. https://pubmed.ncbi.nlm.nih.gov/17221371/
5. Marrero JG, Andrés LS, Luis JG. (2002). Semisynthesis of rosmanol and its derivatives. Easy access to abietatriene diterpenes isolated from the genus Salvia with biological activities — in vitro. Journal of Natural Products, 65(7), 986–989. https://pubmed.ncbi.nlm.nih.gov/12141857/
6. González AG, Andrés LS, Herrera JR, Luis JG, Ravelo AG. (1989). Abietane diterpenes with antibiotic activity from the flowers of Salvia canariensis. Reaction of galdosol with diazomethane. Canadian Journal of Chemistry, 67(2), 208–212. https://doi.org/10.1139/v89-035
7. Djarmati Z, Jankov RM, Csanadi J, Djordjević A, Ribar B, Lazar D. (1992). NMR studies and X-ray crystallography of galdosol 5-methyl ether from Salvia officinalis L. Journal of Crystallographic and Spectroscopic Research, 22(5), 585–590. https://doi.org/10.1007/BF01161344
8. Lima KSB de, Ávila Pimenta AT, Silva Guedes ML, Sousa Lima MA, Silveira ER. (2012). Abietane diterpenes from Hyptis carvalhoi Harley. Biochemical Systematics and Ecology, 44, 240–242. https://doi.org/10.1016/j.bse.2011.12.001
9. Bisio A, De Mieri M, Milella L, Schito AM, Parricchi A, Russo D, et al. (2017). Antibacterial and hypoglycemic diterpenoids from Salvia chamaedryoides — in vitro. Journal of Natural Products, 80(2), 503–514. https://pubmed.ncbi.nlm.nih.gov/28186413/
10. Ríos JL, Schinella GR, Moragrega I. (2022). Phenolics as GABA_A receptor ligands — an updated review. Molecules, 27(6), 1770. https://pubmed.ncbi.nlm.nih.gov/35335130/
11. Kavvadias D, Sand P, Youdim KA, Qaiser MZ, Rice-Evans C, Baur R, et al. (2004). The flavone hispidulin, a benzodiazepine receptor ligand with positive allosteric properties, traverses the blood-brain barrier and exhibits anticonvulsive effects — gerbils and rats. British Journal of Pharmacology, 142(5), 811–820. https://pubmed.ncbi.nlm.nih.gov/15231642/
12. Saha P, Rahman FI, Hussain F, Rahman SMA, Rahman MM. (2021). Antimicrobial diterpenes — recent development from natural sources. [Review.] Frontiers in Pharmacology, 12, 820312. https://pubmed.ncbi.nlm.nih.gov/35295739/
13. PubChem Compound Summary: Galdosol, CID 13966127 (C₂₀H₂₄O₅, MW 344.4, CAS 52591-18-1). National Library of Medicine. https://pubchem.ncbi.nlm.nih.gov/compound/13966127
14. Paloukopoulou C, Karioti A. (2022). A validated method for the determination of carnosic acid and carnosol in the fresh foliage of Salvia rosmarinus and Salvia officinalis from Greece — in vitro. Plants, 11(22), 3106. https://pubmed.ncbi.nlm.nih.gov/36432835/
15. Birtić S, Dussort P, Pierre FX, Bily AC, Roller M. (2015). Carnosic acid — review of distribution, biosynthesis, accumulation and role in planta. Phytochemistry, 115, 9–19. https://pubmed.ncbi.nlm.nih.gov/25639596/
16. Loussouarn M, Krieger-Liszkay A, Svilar L, Bily A, Birtić S, Havaux M. (2017). Carnosic acid and carnosol, two major antioxidants of rosemary, act through different mechanisms — in vitro and in planta. Plant Physiology, 175(3), 1381–1394. https://pubmed.ncbi.nlm.nih.gov/28916593/
17. Luca SV, Skalicka-Woźniak K, Mihai CT, Gradinaru AC, Mandici A, Ciocarlan N, et al. (2023). Chemical profile and bioactivity evaluation of Salvia species from Eastern Europe — in vitro. Antioxidants, 12(8), 1514. https://pubmed.ncbi.nlm.nih.gov/37627509/
18. Etsassala NGER, Hussein AA, Nchu F. (2021). Potential application of some Lamiaceae species in the management of diabetes. [Review.] Plants, 10(2), 279. https://pubmed.ncbi.nlm.nih.gov/33535455/
19. Bennici G, Al Younis I, Sharfalddin A, Akkawi M, Al-Rimawi F, Sawalha K, Emwas AH. (2025). Medicinal effects of Ephedra foeminea aqueous extracts — metabolomic characterization, biological evaluation, and molecular docking. PLoS One. https://pubmed.ncbi.nlm.nih.gov/40705749/
20. Wang J, Xu P, Liu X, Cao C, Sheng Y, Wang J. (2024). Sweet pepper extract reduces fat storage in Caenorhabditis elegans by the SREBP-SCD axis based on multiomics analysis. Food Science & Nutrition, 12(9). https://pubmed.ncbi.nlm.nih.gov/39554335/
21. Hernández-Álvarez E, Rodríguez-Sabina S, González-Delgado G, Reyes CP, Giménez C, Llaría-López MÁ, et al. (2024). Bioassay-guided fractionation networking for discovery of biofungicides from cultivated Salvia canariensis — in vitro. International Journal of Molecular Sciences, 25(24), 13323. https://pubmed.ncbi.nlm.nih.gov/39769088/
22. Escudero J, Perez L, Rabanal RM, Valverde S. (1983). Diterpenoids from Salvia oxyodon and Salvia lavandulifolia. Phytochemistry, 22(2), 585–587. https://doi.org/10.1016/0031-9422(83)83052-0

View all 22 references▾