Gracillin

Gracillin at a Glance

Gracillin is a spirostanol steroidal saponin — C₄₅H₇₂O₁₇, MW 885.1, PubChem CID 159861, CAS 19083-00-2. It is diosgenin carrying a branched three-sugar chain at C-3: a glucose bearing rhamnose at the 2-position and a second glucose at the 3-position. That trisaccharide is sometimes called gracillimatriose, and it is the only thing separating gracillin from dioscin.

The comparison is worth making precisely, because the two molecules are constantly reported side by side and constantly conflated:

	Aglycone	Sugar chain at C-3	Formula	MW
Dioscin	diosgenin	chacotriose — glc with rha(1→2) and rha(1→4)	C₄₅H₇₂O₁₆	869.1
Gracillin	diosgenin	glc with rha(1→2) and glc(1→3)	C₄₅H₇₂O₁₇	885.1

One rhamnose is replaced by a glucose and moved from the 4- to the 3-position of the inner sugar. Everything else — the steroid nucleus, the closed spiroketal F-ring, the 25R configuration — is identical.

Two facts frame the page. First, gracillin has a real, substantial, in-vivo pharmacology of its own — which makes it the exception in this corner of the site. Its 83-paper literature includes a mitochondrial complex II target identified by chemical screening, a glycolytic enzyme target, patient-derived xenograft work, and a dozen mouse tumour models. That is far more than protogracillin, its furostanol precursor, has ever accumulated. Second, almost all of it is oncology, and the single most important number in it is negative: when the National Cancer Institute screened gracillin as NSC-698787 across its 60-cell-line panel, the authors’ conclusion was that gracillin will not be pursued due to lack of selectivity against human cancer diseases ^{7Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →}.

Both things are true at once, and the rest is an attempt to hold them together honestly.

Gracillin & Protogracillin: The Furostanol Bond

Gracillin has a bidesmosidic furostanol precursor, protogracillin, and the relationship between them is the same one that links protodioscin to dioscin. Protogracillin carries an extra glucose at C-26 which holds the F-ring open as a 22-hydroxy furostane; remove that glucose and the ring snaps shut, giving gracillin. The arithmetic checks: protogracillin is C₅₁H₈₄O₂₃, MW 1,065 — gracillin plus one glucose plus the water taken up on ring opening.

The enzymology was characterised directly on this pair. Furostanol glycoside 26-O-β-glucosidase (F26G), purified to homogeneity from Costus speciosus rhizomes in 1996 and cloned the same year, is specific for the C-26 glucosidic bond and has a Kₘ of 50 µM for protogracillin ^{39,40Reference 39Inoue K · 1996Purification and characterization of furostanol glycoside 26-O-β-glucosidase from Costus speciosus rhizomesView study →Reference 40Inoue K et al. · 1996Molecular cloning and bacterial expression of a cDNA encoding furostanol glycoside 26-O-β-glucosidase of Costus speciosusView study →}. The plant stores the inert furostanol separated from the enzyme; tissue damage, drying or processing brings them together and generates gracillin on the spot. It is a two-component chemical defence with the same architecture as the mustard-oil bomb.

Three consequences run through everything below, and they agree with what the protogracillin and dioscin pages already say:

• The gracillin content of a dried herb is partly a record of processing, not only of genetics. A furostanol-rich harvest that was slowly dried is a spirostanol-rich sample.
• In head-to-head comparisons, gracillin is the active member of the pair. In a Dioscorea nipponica lipase screen, four saponins appeared to inhibit pancreatic lipase, but stability testing showed the furostanols convert under assay conditions and the authors named dioscin and gracillin as the true inhibitors ^{30Reference 30Jin P et al. · 2023Screening and identification of lipase inhibitors extracted from Dioscorea nipponica Makino by UV-vis and HPLC coupled to UPLC-Q-TOF-MS/MSView study →}. In a Paris polyphylla gastroprotection study, the spirostanols worked and the furostanols did not, and the authors concluded the closed F-ring was required. A 2003 structure–activity comparison run inside the NCI programme found the F-ring critical to the difference between the furostanol and spirostanol members of this very series ^{7Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →}.
• Both eventually converge on diosgenin. A single fungal β-glucosidase strips gracillin’s sugars to the free aglycone in one step ^{41Reference 41Lei J et al. · 2012A novel β-glucosidase from Aspergillus fumigates releases diosgenin from spirostanosides of Dioscorea zingiberensis CView study →}, and in rats given Paris saponins, diosgenin appeared in tissue even though none was present in the dose ^{33Reference 33Wang G et al. · 2021AnimalTissue distribution, metabolism and absorption of Rhizoma Paridis saponins in the ratsView study →}.

⚠️ And that is where the most persistent myth in this category lives. Diosgenin was the industrial feedstock for the semi-synthesis of progesterone and cortisone by the Marker degradation — a factory process involving acetic anhydride, chromium oxidation and heat. The human body has no enzyme that performs it. There is no metabolic route from gracillin, or from diosgenin, to progesterone, DHEA or testosterone in a person. Wild-yam “natural progesterone” creams that do anything do so because progesterone was added. Nothing here supports a hormonal claim of any kind, and gracillin has no demonstrated steroid-receptor pharmacology.

The Selectivity Problem

This section goes early because it governs how to read every number below.

Gracillin is a surfactant. Spirostanol saponins insert into cholesterol-containing membranes, concentrate in raft microdomains and destabilise them, which is both the reason plants make them and the reason they register as “cytotoxic at low micromolar” in almost any cell-based assay ^{55Reference 55Podolak I et al. · 2010ReviewSaponins as cytotoxic agents: a reviewView study →}. So the question that separates a drug lead from a detergent is always the same: was a normal-cell comparator run, and what was the ratio?

For gracillin the answer is largely no, and where the closest thing to an answer exists it is unflattering.

• The NCI screen is the strongest evidence available and it is negative on selectivity. Gracillin (NSC-698787) was cytotoxic against most of the 60 lines with GI₅₀, TGI and LC₅₀ at micromolar levels — but with no activity at all against EKVX, HT29, OVCAR-5 or SN12C, no coherent subpanel pattern, and no match to any compound in the NCI database by COMPARE analysis. The authors advanced its furostanol relative methyl protoneogracillin to hollow-fibre assay and explicitly dropped gracillin ^{7Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →}.
• Uniform potency across unrelated tumour types is a warning sign, not a strength. In Paris vietnamensis rhizome, gracillin gave IC₅₀ values of 1.17, 1.20 and 1.18 µM against a lung, a liver and a prostate line respectively ^{23Reference 23Hoang LTA et al. · 2025Diosgenin glycosides from Paris vietnamensis (Takht) H.Li with cytotoxic and inhibitory NO production activitiesView study →}. Three different tissues, three different driver mutations, three indistinguishable numbers. That is what a membrane effect looks like; a receptor-mediated effect usually does not behave that way.
• A PubMed search pairing gracillin with normal cells or a selectivity index returns exactly one record ^{57Reference 57Hatim MS et al. · 2025Hatim MS, Al-Saffar AZ, Al-Aadhami MAWS (2025). 5-Bromouracil–gracillin (5BrU-G) complex: an APOBEC3-activated therapeutic strategy exploiting cancer-specific enzymatic activity for selective cytotoxicity. Medical Oncology. https://pubmed.ncbi.nlm.nih.gov/40335833/View study →} — and that paper is not a selectivity study of gracillin. It uses gracillin as a carrier in a 5-bromouracil conjugate, and the selectivity it reports (against MDA-MB-231 and Caco-2, sparing MCF-10A and CCD 841 CoN) is attributed to APOBEC3-dependent release of the 5-bromouracil, not to gracillin.
• The one clean toxicity ratio that does exist is narrow. Against the fish ectoparasite Dactylogyrus intermedius, gracillin’s 48-hour EC₅₀ was 0.18 mg/L while its acute LD₅₀ for the goldfish host was 1.40 mg/L — a therapeutic index of 7.78 ^{38Reference 38Wang GX et al. · 2010Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis CView study →}. In a companion study against Ichthyophthirius multifiliis, EC₅₀ 0.53 mg/L against a grass carp LD₅₀ of 1.64 mg/L ^{37Reference 37Zheng W et al. · 2015Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliisView study →}. A compound that kills the host at three to eight times the dose that kills the parasite is behaving like a detergent, not like a targeted agent.

None of this means the mechanistic work below is wrong. The complex II and PGK1 findings ^{9,10Reference 9Min HY et al. · 2019The natural compound gracillin exerts potent antitumor activity by targeting mitochondrial complex IIView study →Reference 10Min HY et al. · 2020Potent anticancer effect of the natural steroidal saponin gracillin is produced by inhibiting glycolysis and oxidative phosphorylation-mediated bioenergeticsView study →} are properly controlled biochemistry with genetic and pharmacological epistasis. It means that the cell-viability numbers attached to gracillin should be treated as an upper bound on interest, and that the field still owes this molecule a study reporting matched normal-cell IC₅₀ values alongside the tumour ones.

What It's Used For

Gracillin’s evidence has an unusual shape for this site: a large in-vitro oncology literature that is also backed by a dozen independent mouse experiments, plus a small but real anti-inflammatory and cytoprotective set. Nothing has been tested in a human.

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. No row here is supported by human data, so the ceiling for the whole table is “replicated rodent model.”

01AnticancerShaky40%Very large cell-line set plus ~12 independent mouse models and two well-controlled metabolic targets — heavily discounted for the NCI selectivity failure and target incoherence ^{7,9,10,11Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →Reference 9Min HY et al. · 2019The natural compound gracillin exerts potent antitumor activity by targeting mitochondrial complex IIView study →Reference 10Min HY et al. · 2020Potent anticancer effect of the natural steroidal saponin gracillin is produced by inhibiting glycolysis and oxidative phosphorylation-mediated bioenergeticsView study →Reference 11Su JX et al. · 2025Gracillin suppresses cancer progression through inducing Merlin/LATS protein–protein interaction and activating Hippo signaling pathwayView study →}

02Anti-InflammatoryShaky32%Sub-micromolar isolate IC₅₀ from an independent QC group, plus a topical mouse dermatitis model and an LPS cardiac model ^{6,26,27Reference 6Yang G et al. · 2022Identification of anti-inflammatory components in Dioscorea nipponica Makino based on HPLC-MS/MS, quantitative analysis of multiple components by single marker and chemometric methodsView study →Reference 26Jegal J et al. · 2018AnimalAnti-atopic properties of gracillin isolated from Dioscorea quinqueloba on 2,4-dinitrochlorobenzene-induced skin lesions in miceView study →Reference 27Song YX et al. · 2020AnimalGracillin inhibits apoptosis and inflammation induced by lipopolysaccharide to alleviate cardiac injury in mice via improving miR-29aView study →}

03Antifibrotic (Lung)Shaky22%One bleomycin mouse model with a STAT3 rescue experiment ^{25Reference 25Xie M et al. · 2023Gracillin relieves pulmonary fibrosis by suppressing the STAT3 axisView study →}

04HepatoprotectiveShaky18%One mouse ischaemia–reperfusion study with Akt/GSK3β inhibitor epistasis ^{28Reference 28Shen Y et al. · 2025Gracillin protects liver ischemia-reperfusion injury from oxidative stress-induced apoptosisView study →}

05Antiparasitic (Fish)Shaky18%Two in-vivo fish trials with EC₅₀ and host LD₅₀ measured — the therapeutic index is narrow ^{37,38Reference 37Zheng W et al. · 2015Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliisView study →Reference 38Wang GX et al. · 2010Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis CView study →}

06Metabolic (Lipase, Urate)Shaky12%In-vitro enzyme work inside analytical papers ^{29,30Reference 29Wang Z et al. · 2024Efficient discovery of active isolates from Dioscorea spongiosa by the combination of bioassay-guided macroporous resin column chromatography and high-speed counter-current chromatographyView study →Reference 30Jin P et al. · 2023Screening and identification of lipase inhibitors extracted from Dioscorea nipponica Makino by UV-vis and HPLC coupled to UPLC-Q-TOF-MS/MSView study →}

Mechanisms

The last row is the one to keep in mind while reading the rest. Rows demonstrated below ~1 µM, or in an intact animal, carry most of the weight; a viability number at 10 µM carries almost none.

Pharmacokinetics

The defining number is “under 1%.” In rats given a Paris polyphylla extract, gracillin was one of nine steroidal saponins quantified by UPLC-MS/MS; its Cmax was below 10 µg/L, every component’s oral bioavailability was under 1%, all nine were hydrolysed by intestinal flora in anaerobic incubation, and tissue distribution was predominantly liver and lung ^{32Reference 32Lu W et al. · 2020AnimalThe pharmacokinetics and tissue distributions of nine steroidal saponins from Paris polyphylla in ratsView study →}. A dedicated HPLC-MS/MS method was later validated for gracillin alone in rat plasma over 0.065–800 ng/mL and applied to a three-dose oral study ^{31Reference 31Zhang X et al. · 2016Quantitative determination of gracillin by HPLC-MS/MS after oral administration and its application to a pharmacokinetic studyView study →}; a parallel five-saponin method did the same for Dioscorea zingiberensis extract ^{62Reference 62Zhang X et al. · 2015AnimalSimultaneous quantification of five steroid saponins from Dioscorea zingiberensis C.HView study →}.

Three further points shape interpretation:

• The aglycone, not the parent, is what persists. After oral Paris saponins, diosgenin was detected in tissues although there was none in the dose, with the highest levels in lung and brain ^{33Reference 33Wang G et al. · 2021AnimalTissue distribution, metabolism and absorption of Rhizoma Paridis saponins in the ratsView study →}. Whatever gracillin does after swallowing is partly a statement about diosgenin.
• Absorption is formulation-limited, and the limitation is fixable in principle. In an everted rat gut sac model, a poloxamer-407 solid dispersion of Paris saponins roughly doubled absorptive transport (48 → 104 µg) and raised solubility ~3.5-fold; a phytosome formulation did not help. Ranked by n-octanol/water partition coefficient, gracillin sat mid-pack among the six saponins tested, and diosgenyl saponins were absorbed better than pennogenyl ones ^{35Reference 35Liu Z et al. · 2013In vitroFormulation and in vitro absorption analysis of Rhizoma paridis steroidal saponinsView study →}.
• A curiosity worth noting. A 1987 study incorporated gracillin into liposome membranes and found the liposomes were taken up by rat liver through a glucose-specific, anomerically blind receptor, mostly by non-parenchymal cells ^{56Reference 56Das N et al. · 1987Plant glycosides in a liposomal drug-delivery systemView study →}. It is the earliest hint that the terminal glucose gracillin carries — the one sugar distinguishing it from dioscin — changes how the molecule is handled by tissue.

No human pharmacokinetic study of gracillin exists, and no absolute oral bioavailability figure has been published for it in any species.

Clinical Trials

None. A PubMed search for gracillin AND clinical trial[pt] returns zero records, and no study of gracillin is registered under any name. Every human exposure to date has been incidental, through Dioscorea, Paris or Smilax preparations.

Last checked: August 2026.

Isolate vs. Plant

Gracillin sits better on this axis than most saponins, and it is worth saying so plainly.

The pharmacology is overwhelmingly isolate work. The mouse tumour studies dose purified gracillin at defined mg/kg ^{9,10,11,12,14,15,16,18Reference 9Min HY et al. · 2019The natural compound gracillin exerts potent antitumor activity by targeting mitochondrial complex IIView study →Reference 10Min HY et al. · 2020Potent anticancer effect of the natural steroidal saponin gracillin is produced by inhibiting glycolysis and oxidative phosphorylation-mediated bioenergeticsView study →Reference 11Su JX et al. · 2025Gracillin suppresses cancer progression through inducing Merlin/LATS protein–protein interaction and activating Hippo signaling pathwayView study →Reference 12Liao J et al. · 2026Gracillin overcomes osimertinib resistance in NSCLC via dual inhibition of EGFR and Mcl-1View study →Reference 14Lee W et al. · 2025Gracillin induces mitochondria-mediated apoptosis on pancreatic ductal adenocarcinoma through disruption of redox homeostasis via inhibiting NRF2/HO-1 antioxidant axisView study →Reference 15Li JK et al. · 2023In vitroGracillin exerts anti-melanoma effects in vitro and in vivo: role of DNA damage, apoptosis and autophagyView study →Reference 16Yang L et al. · 2021Gracillin shows potent efficacy against colorectal cancer through inhibiting the STAT3 pathwayView study →Reference 18Li Y et al. · 2022Gracillin shows potential efficacy against non-small cell lung cancer through inhibiting the mTOR pathwayView study →}. The macrophage IC₅₀ is a purified-standard figure ^{6Reference 6Yang G et al. · 2022Identification of anti-inflammatory components in Dioscorea nipponica Makino based on HPLC-MS/MS, quantitative analysis of multiple components by single marker and chemometric methodsView study →}. The dermatitis, fibrosis, liver-injury and cardiac studies all used the isolated compound ^{25,26,27,28Reference 25Xie M et al. · 2023Gracillin relieves pulmonary fibrosis by suppressing the STAT3 axisView study →Reference 26Jegal J et al. · 2018AnimalAnti-atopic properties of gracillin isolated from Dioscorea quinqueloba on 2,4-dinitrochlorobenzene-induced skin lesions in miceView study →Reference 27Song YX et al. · 2020AnimalGracillin inhibits apoptosis and inflammation induced by lipopolysaccharide to alleviate cardiac injury in mice via improving miR-29aView study →Reference 28Shen Y et al. · 2025Gracillin protects liver ischemia-reperfusion injury from oxidative stress-induced apoptosisView study →}. The fish trials used bioassay-guided isolates with structures confirmed by NMR ^{37,38Reference 37Zheng W et al. · 2015Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliisView study →Reference 38Wang GX et al. · 2010Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis CView study →}. Commercial reference standards at ≥98% purity are routinely used in the analytical work ^{5Reference 5Hwang JT et al. · 2019Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effectsView study →}.

But two qualifications matter.

1. The isolate does not survive contact with a gut. Sub-1% bioavailability with complete hydrolysis by intestinal flora ^{32Reference 32Lu W et al. · 2020AnimalThe pharmacokinetics and tissue distributions of nine steroidal saponins from Paris polyphylla in ratsView study →} means an oral dose of yam or Paris rhizome does not produce anything like the 1–3 µM concentrations the cell work uses. The mouse antitumour studies that dosed by intraperitoneal injection ^{15Reference 15Li JK et al. · 2023In vitroGracillin exerts anti-melanoma effects in vitro and in vivo: role of DNA damage, apoptosis and autophagyView study →} bypass this; the ones that dosed by gavage ^{11Reference 11Su JX et al. · 2025Gracillin suppresses cancer progression through inducing Merlin/LATS protein–protein interaction and activating Hippo signaling pathwayView study →} do not, and none of them reports the plasma concentration achieved. That gap — efficacy at 5 mg/kg orally against a bioavailability under 1% — is the single largest unexplained thing in gracillin’s file.
2. Extract-level findings get re-narrated as gracillin findings. Paris rhizome (Chonglou) has a large antitumour literature in which gracillin is one of a dozen named saponins, usually a minor one next to the polyphyllins ^{24,44Reference 24Bhuia MS et al. · 2023ReviewAnticancer potential of the plant-derived saponin gracillin: a comprehensive review of mechanistic approachesView study →Reference 44Sun K et al. · 2025UHPLC-Q-TOF/MS unveils metabolic diversity in Paris L. and highlights medicinal potential of five newly identified speciesView study →}; Polygonatum kingianum’s anti-hypoxia effect was attributed to gracillin by network pharmacology with only in-vitro confirmation ^{52Reference 52Li X et al. · 2023Protective effect and mechanism of Polygonatum kingianum against hypoxia-induced injuryView study →}; the Smilax nipponica anti-gout work predicted gracillin computationally and then tested different compounds ^{53Reference 53Yuan C et al. · 2025Potential anti-gout properties and determination by UPLC-Q/TOF-MS of total saponins from Smilax nipponica MiqView study →}. None of these is a gracillin experiment.

There is also a species caution specific to this site. In a direct UHPLC-QTOF-MS comparison, asperin and gracillin were found only in Dioscorea cayenensis, while dioscoreavillosides A and B and parvifloside were found only in D. villosa — and these markers are used to detect species substitution in commercial “Dioscorea” products ^{42Reference 42Avula B et al. · 2014Characterization of steroidal saponins from Dioscorea villosa and D. cayenensis using UHPLC/ESI-QTOF-MSView study →}. The dioscin page already carries this correctly. It follows that gracillin is not a marker of North American wild yam, and the widely repeated statement that gracillin is “extracted from Reineckia carnea, Dioscorea villosa, and other medicinal plants” ^{18Reference 18Li Y et al. · 2022Gracillin shows potential efficacy against non-small cell lung cancer through inhibiting the mTOR pathwayView study →} is not supported by the analytical record for D. villosa.

Prevalence in Nature

Gracillin is a saponin of monocot rhizomes and tubers, and its documented range is wide: Dioscoreaceae (Dioscorea, Tamus), Melanthiaceae (Paris, Trillium), Asparagaceae (Reineckia, Polygonatum, Dracaena), Smilacaceae (Smilax) and Costaceae (Costus speciosus). Two reports fall outside the monocots entirely and should be treated as unreplicated: soapwort (Saponaria officinalis, Caryophyllaceae) ^{50Reference 50Ashurova LN et al. · 2026A new withanolide from Saponaria officinalis and its antimicrobial and cytotoxic activitiesView study →} and the processed tubers of Aconitum carmichaeli (Ranunculaceae) ^{51Reference 51Shim SH et al. · 2005Norditerpenoid alkaloids and other components from the processed tubers of Aconitum carmichaeliView study →}.

Where it has been quantified properly, it sits in the 7–9 mg/g dry weight band in Dioscorea rhizomes — the same order as its furostanol precursor and roughly comparable to dioscin in the same tissue.

Major Natural Sources

Source	Part	Reported level	Basis
Dioscorea quinquelobata	Rhizome	9.01 ± 0.10 mg/g	Validated HPLC-UV, ≥98% standard 5Reference 5Hwang JT et al. · 2019Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effectsView study →
Dioscorea septemloba	Rhizome	7.12 ± 0.03 mg/g	Validated HPLC-UV 5Reference 5Hwang JT et al. · 2019Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effectsView study →; also a quantitative QC marker 59Reference 59Chen H et al. · 2022Strategy of combining offline 2D LC-MS with LC-DIA-MS/MS to accurately identify chemical compounds and for quality control of Dioscorea septemloba ThunbView study →
Dioscorea nipponica	Rhizome	Not detected 5Reference 5Hwang JT et al. · 2019Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effectsView study → — but quantified and assayed elsewhere 6Reference 6Yang G et al. · 2022Identification of anti-inflammatory components in Dioscorea nipponica Makino based on HPLC-MS/MS, quantitative analysis of multiple components by single marker and chemometric methodsView study →	See the discrepancy note below
Dioscorea hypoglauca / D. collettii var. hypoglauca	Rhizome	Not quantified	Isolated by activity-guided fractionation 8Reference 8Hu K et al. · 1996Antineoplastic agents; IView study →; NCI source material 7Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →
Dioscorea spongiosa	Rhizome	Not quantified	Purified to >95% by counter-current chromatography 29Reference 29Wang Z et al. · 2024Efficient discovery of active isolates from Dioscorea spongiosa by the combination of bioassay-guided macroporous resin column chromatography and high-speed counter-current chromatographyView study →
Dioscorea zingiberensis	Rhizome	Not quantified	Isolated for the antiparasitic work 38Reference 38Wang GX et al. · 2010Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis CView study →; in the five-saponin PK panel 62Reference 62Zhang X et al. · 2015AnimalSimultaneous quantification of five steroid saponins from Dioscorea zingiberensis C.HView study →
Dioscorea gracillima	Rhizome	Not quantified	The type source — isolated with dioscin in 1956 1Reference 1Tsukamoto T · 1956Saponins of Japanese DioscoreaceaeView study →
Dioscorea cayenensis	Rhizome	Not quantified	Present; absent from D. villosa in the same study — a substitution marker 42,43Reference 42Avula B et al. · 2014Characterization of steroidal saponins from Dioscorea villosa and D. cayenensis using UHPLC/ESI-QTOF-MSView study →Reference 43Ali Z et al. · 2013Two spirostan steroid glycoside fatty esters from Dioscorea cayenensisView study →
Dioscorea pseudojaponica	Tuber	Not quantified	Isolated with dioscin and the methyl furostanols 61Reference 61Yang DJ et al. · 2003Isolation and identification of steroidal saponins in Taiwanese yam cultivar (Dioscorea pseudojaponica Yamamoto)View study →
Paris polyphylla and relatives	Rhizome	Detected in all Guizhou provenances; total saponins 0.99–9.62%	HPLC across eight areas; gracillin one of five compounds present in every sample 64Reference 64Liang Y et al. · 2012Determination of nine steroidal saponins in Paris polyphylla from different areas of Guizhou province by HPLCView study →
Paris forrestii	Rhizome	A primary saponin; near-undetectable in P. polyphylla var. yunnanensis	Comparative metabolomics + transcriptomics 45Reference 45Xie Z et al. · 2025Comparative transcriptome analysis reveals key genes responsible for the differences in polyphyllin composition in two Paris polyphylla speciesView study →
Paris tengchongensis	Rhizome	Relatively high (with protogracillin)	26-species UHPLC-Q-TOF/MS survey 44Reference 44Sun K et al. · 2025UHPLC-Q-TOF/MS unveils metabolic diversity in Paris L. and highlights medicinal potential of five newly identified speciesView study →
Trillium spp.	Rhizome	Quantified alongside Paris	HPLC-ELSD and HPLC-MS/MS 46Reference 46Man S et al. · 2010Qualitative and quantitative determination of major saponins in Paris and Trillium by HPLC-ELSD and HPLC-MS/MSView study →
Costus speciosus	Rhizome	Not quantified	Bioassay-guided isolation 37Reference 37Zheng W et al. · 2015Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliisView study →; F26G substrate work 39Reference 39Inoue K · 1996Purification and characterization of furostanol glycoside 26-O-β-glucosidase from Costus speciosus rhizomesView study →
Reineckia carnea	Whole plant	Not quantified	Isolated by preparative RP-HPLC, NMR-confirmed 17Reference 17Yang J et al. · 2021Gracillin isolated from Reineckia carnea induces apoptosis of A549 cells via the mitochondrial pathwayView study →
Black bryony (Tamus communis)	Rhizome	Not quantified	First report of gracillin in the genus, 1984 47Reference 47Aquino R et al. · 1984Steroidal glycosides of Tamus communisView study →; also screened for antiviral activity, where the phenanthrenes rather than the saponins carried the interest 48Reference 48Aquino R et al. · 1991Antiviral activity of constituents of Tamus communisView study →
Soapwort (Saponaria officinalis)	Aerial parts	Not quantified	Single 2026 report — outside the usual families 50Reference 50Ashurova LN et al. · 2026A new withanolide from Saponaria officinalis and its antimicrobial and cytotoxic activitiesView study →
Tribulus (T. terrestris)	Supplement extracts	Never quantified	See below

Two Attribution Notes

The Dioscorea nipponica discrepancy is real and unresolved. A Korean validated-HPLC study used the absence of gracillin as one of the two features distinguishing D. nipponica from D. quinquelobata and D. septemloba, reporting it as not detected in D. nipponica while present at 9.01 and 7.12 mg/g in the other two ^{5Reference 5Hwang JT et al. · 2019Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effectsView study →}. A Chinese group in the same period quantified gracillin in D. nipponica rhizome as one of seven saponins, and measured its NO-inhibition IC₅₀ there ^{6Reference 6Yang G et al. · 2022Identification of anti-inflammatory components in Dioscorea nipponica Makino based on HPLC-MS/MS, quantitative analysis of multiple components by single marker and chemometric methodsView study →}; a spectrum–effect study of D. nipponica subsp. rosthornii likewise named gracillin as one of three antitumour actives ^{60Reference 60Liang M et al. · 2026Screening, validation, and mechanism study of antitumor components from Dioscorea nipponica Makino subsp. rosthornii based on spectrum-effect relationship analysisView study →}. The likeliest explanation is chemotype and processing — gracillin is a processing product of protogracillin, so a D. nipponica sample dried differently will have a different spirostanol load. But it means “gracillin content of D. nipponica” is not a fixed species property, and the discrimination method that depends on its absence should be used with that in mind.

The Tribulus claim is thin — thinner than it looks. A PubMed search for gracillin AND "Tribulus terrestris" returns exactly one record: a 2011 Russian-language paper in Voprosy Pitaniia that developed an HPLC method for steroidal saponins in dietary supplements containing T. terrestris extract, and lists gracillin among the compounds “found” — with no quantification, and in finished products rather than authenticated raw plant material ^{49Reference 49Kozlova OI et al. · 2011Determination by high performance chromatography of steroid saponins in biologically active food supplements containing the extract of Tribulus terrestrisView study →}. The definitive nine-country Tribulus chemotype survey did not include gracillin among its analytes at all — it measured protodioscin, prototribestin, pseudoprotodioscin, dioscin, tribestin, tribulosin and rutin ^{65Reference 65Dinchev D et al. · 2008Distribution of steroidal saponins in Tribulus terrestris from different geographical regionsView study →}. So gracillin is reported in Tribulus, once, weakly. It is not a Tribulus marker in any operational sense, and this mirrors — very nearly — what the protogracillin page found for its own molecule, where the count was zero.

Its Role in Plants

Gracillin is a defence saponin, and unusually for this site the defensive function is directly demonstrated rather than assumed. It kills a ciliate at 0.53 mg/L and a monogenean fluke at 0.18 mg/L ^{37,38Reference 37Zheng W et al. · 2015Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliisView study →Reference 38Wang GX et al. · 2010Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis CView study →}, it causes morphological abnormality in Pyricularia oryzae mycelia ^{8Reference 8Hu K et al. · 1996Antineoplastic agents; IView study →}, and its membrane-disrupting chemistry is the same one that makes it cytotoxic in a dish.

The plant does not store it in that active form. It stores protogracillin — the C-26-glucosylated, F-ring-open, biologically quiet precursor — physically separated from F26G, the β-glucosidase that removes the C-26 glucose ^{39,40Reference 39Inoue K · 1996Purification and characterization of furostanol glycoside 26-O-β-glucosidase from Costus speciosus rhizomesView study →Reference 40Inoue K et al. · 1996Molecular cloning and bacterial expression of a cDNA encoding furostanol glycoside 26-O-β-glucosidase of Costus speciosusView study →}. Damage brings the two together and gracillin is generated at the wound. This is why the furostanol/spirostanol ratio of a dried rhizome is a record of post-harvest handling rather than a stable species fingerprint, and it is the reason the two D. nipponica results above can both be right.

One further glimpse of the chemistry comes from an unexpected direction. In Dracaena cochinchinensis, trunk injury triggers metabolic reprogramming in which the original steroidal saponins are depleted and transformed into rearranged aglycones as the plant shifts to polyphenol production. The parent saponins — dioscin and gracillin — were significantly cytotoxic to HepG2 cells; the rearranged degradation products were devoid of cytotoxicity, which the authors read as a deliberate detoxification step ^{54Reference 54Dai B et al. · 2026Dracaenogenins C and D, two new 12(13→14)-abeo-spirostanols from the red resin of Dracaena cochinchinensisView study →}. The plant, in other words, treats gracillin as something to be switched off once it has served its purpose.

Discovery & Synthesis

Gracillin was isolated in 1956 by Tsukamoto and Kawasaki, working through the “Saponins of Japanese Dioscoreaceae” series, from the rhizome of Dioscorea gracillima Miq. — the species that gave the compound its name. It was found together with dioscin in the water-insoluble saponin fraction and characterised as diosgenin combined with one rhamnose and two glucoses ^{1Reference 1Tsukamoto T · 1956Saponins of Japanese DioscoreaceaeView study →}. The full structure, and its relationship to dioscin and kikuba-saponin, was settled in 1962 ^{2Reference 2Kawasaki T · 1962Structures of dioscin, gracillin, and kikuba-saponinView study →}. The proto- form, protogracillin, was named afterwards on the same convention that gives protodioscin its name from dioscin.

Unlike its furostanol precursor, gracillin has been totally synthesised — twice, and by design rather than accident. A Shanghai group assembled it by stepwise glycosylation of diosgenin in 1998 ^{3Reference 3Li C et al. · 1998Synthesis of diosgenyl α-L-rhamnopyranosyl-(1→2)-[β-D-glucopyranosyl-(1→3)]-β-D-glucopyranoside (gracillin) and related saponinsView study →}, and a Peking Union group published a general approach yielding both gracillin and dioscin in 2003, using guanidine to selectively deblock acetyl groups while retaining benzoyl protection ^{4Reference 4Zou CC et al. · 2003The synthesis of gracillin and dioscin: two typical representatives of spirostanol glycosidesView study →}. Neither route is commercially relevant — an 885 Da triglycoside is far more cheaply obtained from a rhizome — but they matter because they establish the structure unambiguously and make analogues accessible.

Research-grade material is purified from plant extract by counter-current chromatography, macroporous resin followed by preparative RP-HPLC, or microwave-assisted extraction coupled to CCC; commercial reference standards are sold at ≥95–98% purity for analytical use ^{5,17,29Reference 5Hwang JT et al. · 2019Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effectsView study →Reference 17Yang J et al. · 2021Gracillin isolated from Reineckia carnea induces apoptosis of A549 cells via the mitochondrial pathwayView study →Reference 29Wang Z et al. · 2024Efficient discovery of active isolates from Dioscorea spongiosa by the combination of bioassay-guided macroporous resin column chromatography and high-speed counter-current chromatographyView study →}. One practical warning attaches to the isolation of the whole series: extracting the furostanol saponins with methanol produces 22-O-methyl artefacts — methyl protogracillin, methyl protodioscin — which have their own separate literature and must not be credited to the natural products ^{7Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →}. Gracillin itself, being a spirostanol, is not subject to that particular artefact.

Patents: not yet researched.

Toxicity & Safety

Gracillin carries a moderate flag, and the reasoning needs stating carefully because it is not the usual one.

There has never been a mammalian toxicology study of gracillin. No acute-toxicity work, no repeat-dose study, no genotoxicity panel, no reproductive study, no haemolysis assay — a PubMed search pairing gracillin with haemolysis returns zero records, which is a striking gap for a spirostanol saponin. The badge is therefore not based on an adverse finding in a safety study, because no safety study exists.

It is based on three things that were measured:

1. Non-selective cytotoxicity, established by the most rigorous screen ever run on this molecule. The NCI 60-cell-line panel found gracillin cytotoxic at micromolar levels across most lines with no selectivity pattern, and the investigators dropped it for that reason ^{7Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →}.
2. A narrow measured therapeutic index in a live vertebrate. Acute LD₅₀ for goldfish 1.40 mg/L and for grass carp 1.64 mg/L, against parasite EC₅₀ values of 0.18 and 0.53 mg/L — an index between roughly 3 and 8 ^{37,38Reference 37Zheng W et al. · 2015Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliisView study →Reference 38Wang GX et al. · 2010Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis CView study →}.
3. Its nearest structural neighbour has a documented repeat-dose adverse effect. Dioscin differs from gracillin by one sugar and, in a 90-day rat study, produced haemolytic anaemia and gastrointestinal distension in males at 300 mg/kg/day ^{66Reference 66Xu T et al. · 2012AnimalA 90-day subchronic toxicological assessment of dioscin, a natural steroid saponin, in Sprague-Dawley ratsView study →}. That is class reasoning rather than direct evidence, and it is listed third for that reason — but it is the most likely shape of gracillin’s own toxicity if anyone ever looks.

Against all of that sits the mouse record, which is reassuring as far as it goes: across roughly a dozen independent tumour studies at 1–10 mg/kg/day for two to three weeks, by gavage and by injection, investigators consistently report no significant adverse effects and no overt change in body weight ^{9,10,11,15,16Reference 9Min HY et al. · 2019The natural compound gracillin exerts potent antitumor activity by targeting mitochondrial complex IIView study →Reference 10Min HY et al. · 2020Potent anticancer effect of the natural steroidal saponin gracillin is produced by inhibiting glycolysis and oxidative phosphorylation-mediated bioenergeticsView study →Reference 11Su JX et al. · 2025Gracillin suppresses cancer progression through inducing Merlin/LATS protein–protein interaction and activating Hippo signaling pathwayView study →Reference 15Li JK et al. · 2023In vitroGracillin exerts anti-melanoma effects in vitro and in vivo: role of DNA damage, apoptosis and autophagyView study →Reference 16Yang L et al. · 2021Gracillin shows potent efficacy against colorectal cancer through inhibiting the STAT3 pathwayView study →}. Those are efficacy studies with efficacy-study endpoints, not toxicology — but a dozen groups reporting nothing is not nothing.

Safety at a Glance

⚠️ Moderate toxicity

• Long dietary and traditional exposure. Dioscorea rhizomes and yams have been eaten and used medicinally for centuries with this saponin present, and oral bioavailability is under 1% ^{32Reference 32Lu W et al. · 2020AnimalThe pharmacokinetics and tissue distributions of nine steroidal saponins from Paris polyphylla in ratsView study →}.
• Tolerated in a dozen independent mouse studies at 1–10 mg/kg/day for 2–3 weeks, with no overt toxicity reported ^{9,10,11,15,16Reference 9Min HY et al. · 2019The natural compound gracillin exerts potent antitumor activity by targeting mitochondrial complex IIView study →Reference 10Min HY et al. · 2020Potent anticancer effect of the natural steroidal saponin gracillin is produced by inhibiting glycolysis and oxidative phosphorylation-mediated bioenergeticsView study →Reference 11Su JX et al. · 2025Gracillin suppresses cancer progression through inducing Merlin/LATS protein–protein interaction and activating Hippo signaling pathwayView study →Reference 15Li JK et al. · 2023In vitroGracillin exerts anti-melanoma effects in vitro and in vivo: role of DNA damage, apoptosis and autophagyView study →Reference 16Yang L et al. · 2021Gracillin shows potent efficacy against colorectal cancer through inhibiting the STAT3 pathwayView study →}.
• The NCI dropped it for lack of selectivity ^{7Reference 7Hu K · 2003In vitroThe cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitroView study →}. Gracillin kills most cell types it is put on; that is the property, not a side effect of it.
• Acutely lethal to fish at 1.4–1.6 mg/L, only 3–8× above its antiparasitic concentration ^{37,38Reference 37Zheng W et al. · 2015Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliisView study →Reference 38Wang GX et al. · 2010Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis CView study →}. Do not discharge saponin-rich extracts into water.
• No mammalian toxicology of any kind. No acute, repeat-dose, genotoxicity or haemolysis study has ever been performed on this molecule. Silence is not clearance.
• No reproductive or developmental data. Nothing is known about gracillin in pregnancy or lactation; the parent herbs carry their own cautions on their own pages.
• Saponins are surfactants. As a class they permeabilise cholesterol-containing membranes, which is both the source of the in-vitro cytotoxicity numbers and a reason not to assume a concentrated or injected preparation would be benign ^{55Reference 55Podolak I et al. · 2010ReviewSaponins as cytotoxic agents: a reviewView study →}.
• P-glycoprotein inhibition is documented ^{36Reference 36Nguyen VT et al. · 2009Selective modulation of P-glycoprotein activity by steroidal saponines from Paris polyphyllaView study →}. Gracillin inhibited P-gp-mediated drug efflux in both drug-selected and transfected cell lines. No CYP, absorption-transporter or clinical interaction study has been done, so no interaction can be ruled in or out.
• Do not infer a hormonal effect from the diosgenin backbone. The Marker degradation is industrial chemistry; humans cannot perform it, and gracillin has no demonstrated steroid-receptor activity.

Dosage

There is no dose. Gracillin has never been administered to a human being as an isolate, there is no clinical trial, and no product is standardised to it. Every published figure is an animal dose, an in-vitro concentration or a plasma curve.

Real-world exposure is incidental. At around 7–9 mg per gram of dried rhizome in the richest quantified sources ^{5Reference 5Hwang JT et al. · 2019Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effectsView study →}, a conventional Dioscorea decoction delivers tens of milligrams of gracillin — as one of dozens of saponins, mostly unabsorbed, in a preparation whose effects have never been attributed to it. Scaling the mouse figures to a person is not meaningful and is not attempted here. These are doses studied in research and are not a personal recommendation.

References

1. Tsukamoto T, Kawasaki T (1956). Saponins of Japanese Dioscoreaceae. VI. The structure of gracillin. Pharmaceutical Bulletin 4:104–108. https://pubmed.ncbi.nlm.nih.gov/13335473/
2. Kawasaki T, Yamauchi T (1962). Structures of dioscin, gracillin, and kikuba-saponin. Chemical & Pharmaceutical Bulletin 10:703–708. https://pubmed.ncbi.nlm.nih.gov/14042124/
3. Li C, Yu B, Liu M, Hui Y (1998). Synthesis of diosgenyl α-L-rhamnopyranosyl-(1→2)-[β-D-glucopyranosyl-(1→3)]-β-D-glucopyranoside (gracillin) and related saponins. Carbohydrate Research. https://pubmed.ncbi.nlm.nih.gov/9691446/
4. Zou CC, Hou SJ, Shi Y, Lei PS, Liang XT (2003). The synthesis of gracillin and dioscin: two typical representatives of spirostanol glycosides. Carbohydrate Research. https://pubmed.ncbi.nlm.nih.gov/12668091/
5. Hwang JT, Park KS, Ryuk JA, Kim HJ, Ko BS (2019). Development of an oriental medicine discrimination method through analysis of steroidal saponins in Dioscorea nipponica Makino and their anti-osteosarcoma effects. Molecules. https://pubmed.ncbi.nlm.nih.gov/31698850/
6. Yang G, Liu P, Shi H, Fan W, Feng X, Chen J, Jing S, Wang L, Zheng Y, Zhang D, Guo L (2022). Identification of anti-inflammatory components in Dioscorea nipponica Makino based on HPLC-MS/MS, quantitative analysis of multiple components by single marker and chemometric methods. Journal of Chromatography B. https://pubmed.ncbi.nlm.nih.gov/36371964/
7. Hu K, Yao X (2003). The cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitro. Phytotherapy Research. https://pubmed.ncbi.nlm.nih.gov/12820229/
8. Hu K, Dong A, Yao X, Kobayashi H, Iwasaki S (1996). Antineoplastic agents; I. Three spirostanol glycosides from rhizomes of Dioscorea collettii var. hypoglauca. Planta Medica. https://pubmed.ncbi.nlm.nih.gov/9000889/
9. Min HY, Jang HJ, Park KH, Hyun SY, Park SJ, Kim JH, Son J, Kang SS, Lee HY (2019). The natural compound gracillin exerts potent antitumor activity by targeting mitochondrial complex II. Cell Death & Disease. https://pubmed.ncbi.nlm.nih.gov/31649278/
10. Min HY, Pei H, Hyun SY, Boo HJ, Jang HJ, Cho J, Kim JH, Son J, Lee HY (2020). Potent anticancer effect of the natural steroidal saponin gracillin is produced by inhibiting glycolysis and oxidative phosphorylation-mediated bioenergetics. Cancers (Basel). https://pubmed.ncbi.nlm.nih.gov/32276500/
11. Su JX, Zhou HX, Zhang ZJ, et al. (2025). Gracillin suppresses cancer progression through inducing Merlin/LATS protein–protein interaction and activating Hippo signaling pathway. Acta Pharmacologica Sinica. https://pubmed.ncbi.nlm.nih.gov/40055528/
12. Liao J, Liang Q, Wu X, Li D, Liu Y, Han S, Wang R, Guo P, Yi M, Li W (2026). Gracillin overcomes osimertinib resistance in NSCLC via dual inhibition of EGFR and Mcl-1. Cancer Gene Therapy. https://pubmed.ncbi.nlm.nih.gov/41857145/
13. Huang S, Lin W, Ding X, Shi J, Liu F, Wang G, Gao C, Su X (2026). CCL24 recruits CCR3+ TAMs to promote immunosuppression via YAP1 activation and serves as a therapeutic target for gracillin in colorectal cancer. International Journal of Biological Sciences. https://pubmed.ncbi.nlm.nih.gov/41694595/
14. Lee W, Song G, Bae H (2025). Gracillin induces mitochondria-mediated apoptosis on pancreatic ductal adenocarcinoma through disruption of redox homeostasis via inhibiting NRF2/HO-1 antioxidant axis. Bioorganic Chemistry. https://pubmed.ncbi.nlm.nih.gov/40505322/
15. Li JK, Zhu PL, Wang Y, Jiang XL, Zhang Z, Zhang Z, Yung KK (2023). Gracillin exerts anti-melanoma effects in vitro and in vivo: role of DNA damage, apoptosis and autophagy. Phytomedicine. https://pubmed.ncbi.nlm.nih.gov/36334389/
16. Yang L, Zhu T, Ye H, et al. (2021). Gracillin shows potent efficacy against colorectal cancer through inhibiting the STAT3 pathway. Journal of Cellular and Molecular Medicine. https://pubmed.ncbi.nlm.nih.gov/33259114/
17. Yang J, Cao L, Li Y, Liu H, Zhang M, Ma H, Wang B, Yuan X, Liu Q (2021). Gracillin isolated from Reineckia carnea induces apoptosis of A549 cells via the mitochondrial pathway. Drug Design, Development and Therapy. https://pubmed.ncbi.nlm.nih.gov/33505158/
18. Li Y, Liu H, Liu X, Xiao B, Zhang M, Luo Y, Li M, Yang J (2022). Gracillin shows potential efficacy against non-small cell lung cancer through inhibiting the mTOR pathway. Frontiers in Oncology. https://pubmed.ncbi.nlm.nih.gov/35392243/
19. Xiao B, Zhang M, Liu H, et al. (2025). Study on the mechanism of gracillin inhibiting the proliferation of lung cancer NCI-H1299 cells based on MAPK signaling pathway. Journal of Cancer. https://pubmed.ncbi.nlm.nih.gov/40740238/
20. Cui A, Tian Y, Chen H, et al. (2026). Study on the anti-lung cancer effects of gracillin from Reineckia carnea by targeting STC1. BMC Cancer. https://pubmed.ncbi.nlm.nih.gov/42135683/
21. Liu W, Wang Y, Chen J, Lin Z, Lin M, Lin X, Fan Y (2021). Beneficial effects of gracillin from Rhizoma Paridis against gastric carcinoma via the potential TIPE2-mediated induction of endogenous apoptosis and inhibition of migration in BGC823 cells. Frontiers in Pharmacology. https://pubmed.ncbi.nlm.nih.gov/34630074/
22. Chen CR, Zhang J, Wu KW, Liu PY, Wang SJ, Chen DY, Ji ZN (2015). Gracillin induces apoptosis in HL60 human leukemic cell line via oxidative stress and cell cycle arrest of G1. Pharmazie. https://pubmed.ncbi.nlm.nih.gov/25980181/
23. Hoang LTA, Pham VC, Ngo VD, et al. (2025). Diosgenin glycosides from Paris vietnamensis (Takht) H.Li with cytotoxic and inhibitory NO production activities. Natural Product Research. https://pubmed.ncbi.nlm.nih.gov/41021975/
24. Bhuia MS, Chowdhury R, Sonia FA, Kamli H, Shaikh A, El-Nashar HAS, El-Shazly M, Islam MT (2023). Anticancer potential of the plant-derived saponin gracillin: a comprehensive review of mechanistic approaches. Chemistry & Biodiversity (review). https://pubmed.ncbi.nlm.nih.gov/37547969/
25. Xie M, Yang L, Cheng J, Qu H, Gu Y, Ding C, Xu X, Zhao C, Huang X, Wang L (2023). Gracillin relieves pulmonary fibrosis by suppressing the STAT3 axis. Journal of Ethnopharmacology. https://pubmed.ncbi.nlm.nih.gov/37257706/
26. Jegal J, Park NJ, Jo BG, Bong SK, Jegal H, Yang MH, Kim SN (2018). Anti-atopic properties of gracillin isolated from Dioscorea quinqueloba on 2,4-dinitrochlorobenzene-induced skin lesions in mice. Nutrients. https://pubmed.ncbi.nlm.nih.gov/30200442/
27. Song YX, Ou YM, Zhou JY (2020). Gracillin inhibits apoptosis and inflammation induced by lipopolysaccharide to alleviate cardiac injury in mice via improving miR-29a. Biochemical and Biophysical Research Communications. https://pubmed.ncbi.nlm.nih.gov/31941605/
28. Shen Y, Shi R, Wang G, Liu H, Li C, Liu F, Cai J (2025). Gracillin protects liver ischemia-reperfusion injury from oxidative stress-induced apoptosis. Drug Design, Development and Therapy. https://pubmed.ncbi.nlm.nih.gov/41262334/
29. Wang Z, Yu J, Zhao L, Niu T, Wang X (2024). Efficient discovery of active isolates from Dioscorea spongiosa by the combination of bioassay-guided macroporous resin column chromatography and high-speed counter-current chromatography. Journal of Separation Science. https://pubmed.ncbi.nlm.nih.gov/38356225/
30. Jin P, Chen L, Zhong J, Yuan T, Gan L, Huang J, Wang L, Fan H, Lin C (2023). Screening and identification of lipase inhibitors extracted from Dioscorea nipponica Makino by UV-vis and HPLC coupled to UPLC-Q-TOF-MS/MS. International Journal of Biological Macromolecules. https://pubmed.ncbi.nlm.nih.gov/36706882/
31. Zhang X, Xue X, Zhao J, Guo Z, Ito Y, Sun W (2016). Quantitative determination of gracillin by HPLC-MS/MS after oral administration and its application to a pharmacokinetic study. Steroids. https://pubmed.ncbi.nlm.nih.gov/27394958/
32. Lu W, Pan M, Zhang P, Zheng T, Huang L, Ye F, Lei P (2020). The pharmacokinetics and tissue distributions of nine steroidal saponins from Paris polyphylla in rats. European Journal of Drug Metabolism and Pharmacokinetics. https://pubmed.ncbi.nlm.nih.gov/32661907/
33. Wang G, Hao R, Liu Y, Wang Y, Man S, Gao W (2021). Tissue distribution, metabolism and absorption of Rhizoma Paridis saponins in the rats. Journal of Ethnopharmacology. https://pubmed.ncbi.nlm.nih.gov/33746004/
34. Wang G, Yan M, Hao R, Lv P, Wang Y, Man S, Gao W (2022). Q-marker identification of Paris polyphylla var. yunnanensis in pulmonary metastasis of liver cancer mice. Journal of Ethnopharmacology. https://pubmed.ncbi.nlm.nih.gov/35461989/
35. Liu Z, Wang J, Gao W, Man S, Guo H, Zhang J, Liu C (2013). Formulation and in vitro absorption analysis of Rhizoma paridis steroidal saponins. International Journal of Pharmaceutics. https://pubmed.ncbi.nlm.nih.gov/23107795/
36. Nguyen VT, Darbour N, Bayet C, Doreau A, Raad I, Phung BH, Dumontet C, Di Pietro A, Dijoux-Franca MG, Guilet D (2009). Selective modulation of P-glycoprotein activity by steroidal saponines from Paris polyphylla. Fitoterapia. https://pubmed.ncbi.nlm.nih.gov/18940238/
37. Zheng W, Yan CM, Zhang YB, Li ZH, Li Z, Li XY, Wang ZW, Wang X, Chen WQ, Yu XH (2015). Antiparasitic efficacy of gracillin and zingibernsis newsaponin from Costus speciosus against Ichthyophthirius multifiliis. Parasitology. https://pubmed.ncbi.nlm.nih.gov/25140457/
38. Wang GX, Jiang DX, Li J, Han J, Liu YT, Liu XL (2010). Anthelmintic activity of steroidal saponins from Dioscorea zingiberensis C. H. Wright against Dactylogyrus intermedius (Monogenea) in goldfish (Carassius auratus). Parasitology Research. https://pubmed.ncbi.nlm.nih.gov/20689967/
39. Inoue K, Ebizuka Y (1996). Purification and characterization of furostanol glycoside 26-O-β-glucosidase from Costus speciosus rhizomes. FEBS Letters. https://pubmed.ncbi.nlm.nih.gov/8549824/
40. Inoue K, Shibuya M, Yamamoto K, Ebizuka Y (1996). Molecular cloning and bacterial expression of a cDNA encoding furostanol glycoside 26-O-β-glucosidase of Costus speciosus. FEBS Letters. https://pubmed.ncbi.nlm.nih.gov/8766714/
41. Lei J, Niu H, Li T, Huang W (2012). A novel β-glucosidase from Aspergillus fumigates releases diosgenin from spirostanosides of Dioscorea zingiberensis C. H. Wright. World Journal of Microbiology and Biotechnology. https://pubmed.ncbi.nlm.nih.gov/22805852/
42. Avula B, Wang YH, Wang M, Ali Z, Smillie TJ, Zweigenbaum J, Khan IA (2014). Characterization of steroidal saponins from Dioscorea villosa and D. cayenensis using UHPLC/ESI-QTOF-MS. Planta Medica. https://pubmed.ncbi.nlm.nih.gov/24510365/
43. Ali Z, Smillie TJ, Khan IA (2013). Two spirostan steroid glycoside fatty esters from Dioscorea cayenensis. Natural Product Communications. https://pubmed.ncbi.nlm.nih.gov/23678801/
44. Sun K, Shao Y, Li R, Wang Z, Wang J, Luan X, Wu C, Wang L, He C, Tang Q (2025). UHPLC-Q-TOF/MS unveils metabolic diversity in Paris L. and highlights medicinal potential of five newly identified species. Frontiers in Pharmacology. https://pubmed.ncbi.nlm.nih.gov/40771935/
45. Xie Z, Yu H, Peng S, Zhang B, Liu G, Wei C, Lai J, Cai C, Xu F (2025). Comparative transcriptome analysis reveals key genes responsible for the differences in polyphyllin composition in two Paris polyphylla species. Gene. https://pubmed.ncbi.nlm.nih.gov/39938760/
46. Man S, Gao W, Zhang Y, Wang J, Zhao W, Huang L, Liu C (2010). Qualitative and quantitative determination of major saponins in Paris and Trillium by HPLC-ELSD and HPLC-MS/MS. Journal of Chromatography B. https://pubmed.ncbi.nlm.nih.gov/20870470/
47. Aquino R, Behar I, De Simone F, Pizza C (1984). Steroidal glycosides of Tamus communis. Bollettino della Società Italiana di Biologia Sperimentale. https://pubmed.ncbi.nlm.nih.gov/6529500/
48. Aquino R, Conti C, De Simone F, Orsi N, Pizza C, Stein ML (1991). Antiviral activity of constituents of Tamus communis. Journal of Chemotherapy. https://pubmed.ncbi.nlm.nih.gov/1667189/
49. Kozlova OI, Perederiaev OI, Ramenskaia GV (2011). Determination by high performance chromatography of steroid saponins in biologically active food supplements containing the extract of Tribulus terrestris. Voprosy Pitaniia (in Russian). https://pubmed.ncbi.nlm.nih.gov/22379868/
50. Ashurova LN, Khurramov AR, Bobakulov KM, et al. (2026). A new withanolide from Saponaria officinalis and its antimicrobial and cytotoxic activities. Natural Product Research. https://pubmed.ncbi.nlm.nih.gov/40029964/
51. Shim SH, Lee SY, Kim JS, Son KH, Kang SS (2005). Norditerpenoid alkaloids and other components from the processed tubers of Aconitum carmichaeli. Archives of Pharmacal Research. https://pubmed.ncbi.nlm.nih.gov/16350849/
52. Li X, Mei M, Pu X, et al. (2023). Protective effect and mechanism of Polygonatum kingianum against hypoxia-induced injury. Heliyon. https://pubmed.ncbi.nlm.nih.gov/36967867/
53. Yuan C, Chen H, Zhao Z, Wang Z, Zhong L, Xing Y, Jin C (2025). Potential anti-gout properties and determination by UPLC-Q/TOF-MS of total saponins from Smilax nipponica Miq. Natural Product Research. https://pubmed.ncbi.nlm.nih.gov/41051190/
54. Dai B, Wang L, Su LH, Li HZ (2026). Dracaenogenins C and D, two new 12(13→14)-abeo-spirostanols from the red resin of Dracaena cochinchinensis. Molecules. https://pubmed.ncbi.nlm.nih.gov/41828837/
55. Podolak I, Galanty A, Sobolewska D (2010). Saponins as cytotoxic agents: a review. Phytochemistry Reviews. https://pubmed.ncbi.nlm.nih.gov/20835386/
56. Das N, Bachhawat BK, Mahato SB, Basu MK (1987). Plant glycosides in a liposomal drug-delivery system. Biochemical Journal. https://pubmed.ncbi.nlm.nih.gov/3426542/
57. Hatim MS, Al-Saffar AZ, Al-Aadhami MAWS (2025). 5-Bromouracil–gracillin (5BrU-G) complex: an APOBEC3-activated therapeutic strategy exploiting cancer-specific enzymatic activity for selective cytotoxicity. Medical Oncology. https://pubmed.ncbi.nlm.nih.gov/40335833/
58. Rui G, Qin ZY, Chang YQ, Zheng YG, Zhang D, Yao LM, Guo L (2023). Chemical comparison and identification of xanthine oxidase inhibitors of Dioscoreae Hypoglaucae Rhizoma and Dioscoreae Spongiosae Rhizoma by chemometric analysis and spectrum-effect relationship. Molecules. https://pubmed.ncbi.nlm.nih.gov/38138603/
59. Chen H, Wang L, Wang C, Zhang Y, Yu H, Fu Z, Fu X, Han L (2022). Strategy of combining offline 2D LC-MS with LC-DIA-MS/MS to accurately identify chemical compounds and for quality control of Dioscorea septemloba Thunb. Phytochemical Analysis. https://pubmed.ncbi.nlm.nih.gov/35841277/
60. Liang M, Shen M, Wang Y, Fan P, Zheng Y, Wang H, Sun J, Jing S (2026). Screening, validation, and mechanism study of antitumor components from Dioscorea nipponica Makino subsp. rosthornii based on spectrum-effect relationship analysis. Frontiers in Oncology. https://pubmed.ncbi.nlm.nih.gov/42137146/
61. Yang DJ, Lu TJ, Hwang LS (2003). Isolation and identification of steroidal saponins in Taiwanese yam cultivar (Dioscorea pseudojaponica Yamamoto). Journal of Agricultural and Food Chemistry. https://pubmed.ncbi.nlm.nih.gov/14558759/
62. Zhang X, Li J, Ito Y, Sun W (2015). Simultaneous quantification of five steroid saponins from Dioscorea zingiberensis C.H. Wright in rat plasma by HPLC-MS/MS and its application to the pharmacokinetic studies. Steroids. https://pubmed.ncbi.nlm.nih.gov/25201262/
63. Qin Z, Du Y, Zhong Y, Gu Z, Zheng Q, Zheng Y, Zhang D, Guo L (2025). Purification of steroidal saponins from Dioscoreae hypoglaucae rhizoma by macroporous resin and evaluation of their anti-inflammatory effect. Journal of Chromatography B. https://pubmed.ncbi.nlm.nih.gov/40829486/
64. Liang Y, Liu Z, Gao W, Man S (2012). Determination of nine steroidal saponins in Paris polyphylla from different areas of Guizhou province by HPLC. Zhongguo Zhong Yao Za Zhi (in Chinese). https://pubmed.ncbi.nlm.nih.gov/23189740/
65. Dinchev D, Janda B, Evstatieva L, Oleszek W, Aslani MR, Kostova I (2008). Distribution of steroidal saponins in Tribulus terrestris from different geographical regions. Phytochemistry. https://pubmed.ncbi.nlm.nih.gov/17719068/
66. Xu T, Zhang S, Zheng L, Yin L, Xu L, Peng J (2012). A 90-day subchronic toxicological assessment of dioscin, a natural steroid saponin, in Sprague-Dawley rats. Food and Chemical Toxicology. Erratum published 2019. https://pubmed.ncbi.nlm.nih.gov/22386816/

View all 66 references▾