Morindaparvin A

• CAS: 41621-32-3
• Synonyms: morindaparvin A
• Molecular Formula: C15H8O4
• Molecular Weight: 252.22 g/mol
• SMILES: C1OC2=C(O1)C3=C(C=C2)C(=O)C4=CC=CC=C4C3=O
• InChI: InChI=1S/C15H8O4/c16-13-8-3-1-2-4-9(8)14(17)12-10(13)5-6-11-15(12)19-7-18-11/h1-6H,7H2
• InChIKey: YBXYYSSRAZPEGN-UHFFFAOYSA-N

Morindaparvin A is a naturally occurring anthraquinone first isolated from the rhizome and roots of Morinda parvifolia (Rubiaceae) via bioassay-directed fractionation targeting antileukemic activity. Its structure was elucidated as 1,2-methylenedioxyanthraquinone (C₁₅H₈O₄, MW 252.22), a distinctive scaffold confirmed by total synthesis from alizarin and dibromomethane.

Morindaparvin A (CAS 41621-32-3): A 1,2-Methylenedioxyanthraquinone Natural Product with Documented Antileukemic Differentiation for Research Procurement

Morindaparvin A is a naturally occurring anthraquinone first isolated from the rhizome and roots of Morinda parvifolia (Rubiaceae) via bioassay-directed fractionation targeting antileukemic activity. Its structure was elucidated as 1,2-methylenedioxyanthraquinone (C₁₅H₈O₄, MW 252.22), a distinctive scaffold confirmed by total synthesis from alizarin and dibromomethane [1]. Unlike most common plant-derived anthraquinones that possess free phenolic hydroxyl groups, Morindaparvin A features a 1,2-methylenedioxy bridge that eliminates phenolic -OH redox liability while creating a planar, heterocycle-fused pharmacophore. The compound is catalogued in authoritative metabolite databases including ChEBI (CHEBI:6999), KEGG COMPOUND (C10374), and KNApSAcK (C00002843), and is classified under MeSH as an anthraquinone from M. parvifolia [2].

Cell model P-388 lymphocytic leukemia screening; bioassay-directed isolation from Morinda parvifolia

Pharmacophore 1,2-Methylenedioxy scaffold distinct from redox-active phenolic anthraquinones

Standard Cataloged natural product (ChEBI, KEGG) suitable for phytochemical and SAR research

• [1] Chang P, Lee KH, Shingu T, Hirayama T, Hall IH, Huang HC. Antitumor agents 50. 1 Morindaparvin-A, a new antileukemic anthraquinone, and alizarin-1-methyl ether from Morinda parvifolia, and the antileukemic activity of the related derivatives. J Nat Prod. 1982;45(2):206-210. DOI: 10.1021/np50020a017. PMID: 7097296. View Source
• [2] ChEBI: Morindaparvin A (CHEBI:6999). European Bioinformatics Institute. KEGG COMPOUND: C10374. MeSH Supplementary Concept: morindaparvin A (Unique ID: C035011). National Library of Medicine. View Source

Why Morindaparvin A Cannot Be Substituted by Generic Anthraquinones: Evidence from Bioassay-Directed Fractionation

Generic substitution with other Morinda-derived or common anthraquinones (e.g., alizarin, emodin, alizarin-1-methyl ether) is not scientifically valid for Morindaparvin A. The original bioassay-directed fractionation study demonstrated that alizarin-1-methyl ether — co-isolated from the identical M. parvifolia extract — was entirely devoid of cytotoxic activity, whereas Morindaparvin A exhibited potent P-388 leukemia cell growth inhibition (ED₅₀ = 1.85 μg/mL) [1]. Furthermore, multiple semisynthetic alizarin derivatives (mono- and di-acetates, cinnamates, senecioates) proved inactive in vivo, while the parent alizarin and its close analog 1,2-ethylenedioxyanthraquinone showed only marginal antileukemic activity (T/C = 126–136% at 10 mg/kg/day) [1]. These findings establish that the 1,2-methylenedioxy substitution pattern — rather than the anthraquinone core alone — is the critical pharmacophoric determinant of antileukemic potency. Even within the M. parvifolia anthraquinone family, activity varies dramatically: morindaparvin B, a structurally related congener with a free hydroxymethyl group replacing the methylenedioxy bridge, displays approximately 5.7-fold lower P-388 potency (ED₅₀ = 10.5 μg/mL) [2].

Pharmacophore dependency

The 1,2-methylenedioxy bridge is essential for P-388 cell-model response; common phenolic anthraquinones (alizarin, emodin) may not replicate activity.

Co-isolate inactivity

Alizarin-1-methyl ether from the same M. parvifolia extract showed no cytotoxicity in parallel testing; anthraquinone core alone is not predictive.

Congener variability

Morindaparvin B, differing only by a hydroxymethyl group, yields markedly different P-388 response; structural analog interchange cannot be assumed.

• [1] Chang P, Lee KH, Shingu T, Hirayama T, Hall IH, Huang HC. Antitumor agents 50. 1 Morindaparvin-A, a new antileukemic anthraquinone, and alizarin-1-methyl ether from Morinda parvifolia, and the antileukemic activity of the related derivatives. J Nat Prod. 1982;45(2):206-210. DOI: 10.1021/np50020a017. PMID: 7097296. View Source
• [2] Chang P, Lee KH. Cytotoxic antileukemic anthraquinones from Morinda parvifolia. Phytochemistry. 1984;23(8):1733-1736. DOI: 10.1016/S0031-9422(00)83480-9. Data compiled in: 百眼藤 [Baiyanteng / Morinda parvifolia]. 医学百科 [Medical Encyclopedia]; big5.wiki8.cn. View Source

Morindaparvin A Quantitative Differentiation Evidence: Comparator-Backed Data for Informed Procurement Decisions

P-388 Lymphocytic Leukemia In Vitro Cytotoxicity: Direct Head-to-Head vs Co-Isolated Alizarin-1-Methyl Ether

In the original bioassay-directed isolation study from Morinda parvifolia, Morindaparvin A (1) and alizarin-1-methyl ether (2) were co-isolated from the same antileukemic methanolic extract. When tested in parallel against the P-388 lymphocytic leukemia cell line in vitro, Morindaparvin A exhibited an ED₅₀ of 1.85 μg/mL, while alizarin-1-methyl ether demonstrated no detectable cytotoxic activity [1]. This represents a qualitative (active vs. inactive) differentiation established within a single experimental framework, eliminating cross-laboratory variability as a confounding factor. The structural distinction — a 1,2-methylenedioxy bridge in Morindaparvin A versus a 1-methoxy-2-hydroxy substitution in alizarin-1-methyl ether — is the sole variable accounting for this potency gap.

P-388 Head-to-Head

Head-to-head

ED₅₀ 1.85 μg/mL vs. no detectable cytotoxicity (same assay)

Supports methylenedioxy pharmacophore requirement for P-388 cell-model response.

Co-isolated comparison eliminates inter-study variability; Chang et al. 1982.

Evidence Dimension	In vitro cytotoxicity against P-388 murine lymphocytic leukemia cells
Target Compound Data	ED₅₀ = 1.85 μg/mL
Comparator Or Baseline	Alizarin-1-methyl ether: no cytotoxic activity (inactive)
Quantified Difference	Qualitative: active (ED₅₀ = 1.85 μg/mL) vs. inactive (no detectable cytotoxicity)
Conditions	In vitro P-388 lymphocytic leukemia tissue culture; same bioassay-guided fractionation study; Chang et al. 1982

Why This Matters

This direct co-isolation comparison eliminates any argument that the anthraquinone core alone confers activity — the specific 1,2-methylenedioxy pharmacophore is essential, making Morindaparvin A irreplaceable by common anthraquinone substitutes for P-388-targeted research.

• [1] Chang P, Lee KH, Shingu T, Hirayama T, Hall IH, Huang HC. Antitumor agents 50. 1 Morindaparvin-A, a new antileukemic anthraquinone, and alizarin-1-methyl ether from Morinda parvifolia, and the antileukemic activity of the related derivatives. J Nat Prod. 1982;45(2):206-210. DOI: 10.1021/np50020a017. PMID: 7097296. View Source

P-388 Potency Differential vs Closest Structural Congener Morindaparvin B: ~5.7-Fold Superiority

Morindaparvin B represents the closest structural congener to Morindaparvin A among all anthraquinones reported from M. parvifolia, differing primarily by replacement of the 1,2-methylenedioxy bridge with a 1-hydroxy-2-hydroxymethyl substitution pattern (C₁₅H₁₀O₅, MW 270.24). In the P-388 lymphocytic leukemia in vitro assay, Morindaparvin A demonstrated an ED₅₀ of 1.85 μg/mL, while Morindaparvin B yielded an ED₅₀ of 10.5 μg/mL — an approximately 5.7-fold potency advantage for Morindaparvin A [1][2]. Notably, the potency ranking reverses in the KB nasopharyngeal carcinoma cell line (Morindaparvin A ED₅₀ = 10 μg/mL vs. Morindaparvin B ED₅₀ = 4.0 μg/mL), indicating that the methylenedioxy substitution confers selective, rather than universally superior, cytotoxicity with a preference for the P-388 leukemia phenotype.

vs Morindaparvin B

Cross-study

~5.7-fold lower ED₅₀ (1.85 vs. 10.5 μg/mL) in P-388 assay

Morindaparvin A shows greater P-388 cell-model response among M. parvifolia congeners.

Rank reverses in KB cell line; selectivity context-dependent.

Evidence Dimension	In vitro ED₅₀ against P-388 lymphocytic leukemia cell growth
Target Compound Data	ED₅₀ = 1.85 μg/mL
Comparator Or Baseline	Morindaparvin B: ED₅₀ = 10.5 μg/mL
Quantified Difference	~5.7-fold lower ED₅₀ (greater potency) for Morindaparvin A against P-388
Conditions	P-388 murine lymphocytic leukemia cells in vitro; data from Chang et al. 1982 (Morindaparvin A) and Chang & Lee 1984 (Morindaparvin B); compiled in authoritative Chinese pharmacognosy references

Why This Matters

Among the two novel anthraquinones discovered in M. parvifolia, Morindaparvin A is markedly more potent against the P-388 leukemia target — the primary screening model that motivated the original bioassay-directed isolation — making it the preferred research tool for leukemia-focused mechanism-of-action studies.

• [1] Chang P, Lee KH, Shingu T, Hirayama T, Hall IH, Huang HC. Antitumor agents 50. 1 Morindaparvin-A, a new antileukemic anthraquinone, and alizarin-1-methyl ether from Morinda parvifolia, and the antileukemic activity of the related derivatives. J Nat Prod. 1982;45(2):206-210. DOI: 10.1021/np50020a017. View Source
• [2] Chang P, Lee KH. Cytotoxic antileukemic anthraquinones from Morinda parvifolia. Phytochemistry. 1984;23(8):1733-1736. DOI: 10.1016/S0031-9422(00)83480-9. ED₅₀ values compiled and verified in: 百眼藤 [Morinda parvifolia monograph]. 医学百科; big5.wiki8.cn; and 39药品通 [39 Medicine Reference]; wapypk.39.net. View Source

Pharmacophoric Distinction: 1,2-Methylenedioxy Substitution Eliminates Redox-Active Phenolic Hydroxyls Present in Generic Anthraquinones

Spectroscopic characterization confirmed that Morindaparvin A bears a 1,2-methylenedioxy group (δ 6.32, 2H singlet in ¹H NMR; absence of hydroxyl IR absorption), with no free phenolic -OH groups [1]. This contrasts sharply with the majority of bioactive natural anthraquinones — including alizarin (1,2-dihydroxyanthraquinone), emodin (1,3,8-trihydroxy-6-methylanthraquinone), and the co-isolated alizarin-1-methyl ether — all of which possess one or more free phenolic hydroxyls. Phenolic -OH groups in anthraquinones are well-established contributors to redox cycling, generation of reactive oxygen species (ROS), and attendant off-target cytotoxicity [2]. The replacement of two phenolic -OH groups (as in alizarin) with a single methylenedioxy bridge eliminates this redox liability while preserving the planar anthraquinone core geometry. Chang and Chen (1996) subsequently reported that the carcinostatic mechanism of Morindaparvin A involves DNA intercalation, consistent with the retention of a planar chromophore in the absence of redox-active substituents [3].

Redox-Liability Distinction

Class-level

0 free phenolic -OH vs. alizarin (2), emodin (3); 1,2-methylenedioxy confirmed

Eliminates redox-cycling confound; DNA intercalation mechanism implied.

Redox activity of phenolic anthraquinones reviewed in Bolton & Dunlap 2017.

Evidence Dimension	Presence of free phenolic hydroxyl groups (redox liability) vs. methylenedioxy substitution
Target Compound Data	Morindaparvin A: 0 free phenolic -OH groups; 1,2-methylenedioxy bridge confirmed by ¹H NMR (δ 6.32, 2H, s) and IR (no -OH absorption)
Comparator Or Baseline	Alizarin: 2 free phenolic -OH groups (1-OH, 2-OH); Emodin: 3 free phenolic -OH groups; Alizarin-1-methyl ether: 1 free phenolic -OH (2-OH)
Quantified Difference	Structural: methylenedioxy substitution removes 2 redox-active -OH sites present in the direct synthetic precursor alizarin
Conditions	Spectroscopic structural elucidation (IR, UV, ¹H NMR, MS); Chang et al. 1982; mechanism studies from Chang & Chen 1996

Why This Matters

For researchers concerned with redox-mediated off-target effects that confound many anthraquinone bioassays, Morindaparvin A offers a cleaner pharmacological probe — antileukemic activity attributable to DNA intercalation rather than non-specific oxidative stress.

• [1] Chang P, Lee KH, Shingu T, Hirayama T, Hall IH, Huang HC. Antitumor agents 50. 1 Morindaparvin-A, a new antileukemic anthraquinone, and alizarin-1-methyl ether from Morinda parvifolia, and the antileukemic activity of the related derivatives. J Nat Prod. 1982;45(2):206-210. DOI: 10.1021/np50020a017. (IR, UV, NMR, MS characterization; structure confirmation by synthesis from alizarin). View Source
• [2] Bolton JL, Dunlap T. Formation and biological targets of quinones: cytotoxic versus cytoprotective effects. Chem Res Toxicol. 2017;30(1):13-37. DOI: 10.1021/acs.chemrestox.6b00256. (Review: redox cycling mechanisms of phenolic anthraquinones and quinone toxicity.) View Source
• [3] Chang P, Chen CF. Antineoplastic anthraquinones. II. Design and synthesis of 1,2-heteroannelated anthraquinones. J Heterocyclic Chem. 1996;33(2):367-371. DOI: 10.1002/jhet.5570330223. (Reports DNA intercalation as carcinostatic mechanism of morindaparvin-A.) View Source

Validated Lead Structure Status: Morindaparvin A as the Design Template for Heterocycle-Fused Anthraquinone Antineoplastic Agents

Morindaparvin A (designated 2a) was explicitly selected as the lead structure in a systematic medicinal chemistry program based on the '2-phenylnaphthalene-type' structural pattern hypothesis. Using Morindaparvin A as the template scaffold, Chang and Chen (1996) designed and synthesized multiple series of 1,2-heteroannelated anthraquinones, including 1,2-alkylenedioxyanthraquinones, naphtho[2,3-f]quinoxaline-7,12-diones, anthra[1,2-d]imidazole-6,11-diones, and naphtho[2,3-f]quinoxaline-7,12-dione derivatives [1]. Most synthesized analogs retained varying degrees of anticancer activity, and one derivative — ω-chloromethyl-1H-anthra[1,2-d]imidazole-6,11-dione (4b) — exhibited cytotoxic activity against all tested human carcinoma cell lines [1]. This lead structure validation is a distinction not shared by the vast majority of natural anthraquinones, including the more widely studied alizarin, emodin, or rhein — none of which have been comparably employed as rational design templates for a focused heterocycle-fused analog series in the peer-reviewed medicinal chemistry literature.

Lead Structure Validation

Supporting evidence

Used as template (2a) for 4 heterocycle-fused anthraquinone series; active analogs generated

Precedent for rational diversification; de-risks hit-to-lead campaigns.

Derivative 4b active against multiple human carcinoma lines; Chang & Chen 1996.

Evidence Dimension	Utilization as a rational drug design lead structure in a published medicinal chemistry program
Target Compound Data	Morindaparvin A used as lead structure (2a) for design of 1,2-heteroannelated anthraquinone series; multiple active analogs generated
Comparator Or Baseline	Alizarin, emodin, rhein, and other common natural anthraquinones: no equivalent systematic heterocycle-fused analog design program identified in peer-reviewed literature
Quantified Difference	Qualitative: validated lead template vs. no comparable systematic lead optimization literature
Conditions	Chang & Chen 1996; compounds tested for antineoplastic activity; derivative 4b active against all tested human carcinoma cell lines

Why This Matters

For medicinal chemistry groups seeking a pre-validated natural product scaffold for anthraquinone-based lead optimization, Morindaparvin A is among the very few with published proof-of-concept that its methylenedioxy pharmacophore can be rationally diversified into active heterocycle-fused analogs — significantly de-risking early-stage hit-to-lead campaigns.

• [1] Chang P, Chen CF. Antineoplastic anthraquinones. II. Design and synthesis of 1,2-heteroannelated anthraquinones. J Heterocyclic Chem. 1996;33(2):367-371. DOI: 10.1002/jhet.5570330223. View Source

Total Synthesis Accessibility: Khanapure–Biehl Route Enables Laboratory-Scale Resupply Independent of Botanical Extraction

A total synthesis of Morindaparvin A was achieved by Khanapure and Biehl (1989), employing a benzyne-mediated cycloaddition strategy that constructs the anthraquinone core with regioselective installation of the 1,2-methylenedioxy functionality [1]. The synthesis confirms the assigned structure and provides a published, reproducible route for laboratory-scale preparation. This contrasts with the procurement reality for many botanical anthraquinones (e.g., morindone, soranjidiol, lucidin derivatives), which remain accessible solely through extraction from plant material — a process subject to seasonal variation, low yields (Morindaparvin A was originally isolated in only 0.00081% yield from M. parvifolia rhizome [2]), and supply chain dependency on specific Morinda species. The existence of a validated synthetic route means researchers can, if needed, bypass botanical sourcing constraints and generate analogs through modification of the published procedure.

Total Synthesis Route

Supporting evidence

Khanapure–Biehl (1989) benzyne cycloaddition; structure confirmed; lab-scale resupply feasible

Synthesis-enabled procurement; bypasses botanical extraction (~0.00081% yield).

Route adaptable for analog generation; J Nat Prod 1989.

Evidence Dimension	Availability of published total synthesis route vs. exclusive reliance on natural extraction
Target Compound Data	Total synthesis published: Khanapure & Biehl 1989; benzyne cycloaddition approach; structure confirmed by direct comparison with natural material
Comparator Or Baseline	Morindone, soranjidiol, lucidin-ω-methyl ether, and many other Morinda anthraquinones: no published total synthesis route identified; accessible only via extraction
Quantified Difference	Synthesis-enabled resupply vs. extraction-only procurement; natural isolation yield: ~0.00081% from M. parvifolia rhizome
Conditions	Published synthetic methodology; Khanapure SP, Biehl ER, J Nat Prod 1989; extraction yield data from Chang et al. 1982

Why This Matters

For procurement planning, the published total synthesis provides an insurance mechanism against supply discontinuity: if natural sourcing becomes unavailable, laboratory synthesis remains a viable fallback — an option not available for extraction-only anthraquinone standards.

• [1] Khanapure SP, Biehl ER. Synthesis of Morindaparvin A, an Antitumor Agent, and Related Anthraquinones. J Nat Prod. 1989;52(6):1357-1359. DOI: 10.1021/np50066a033. View Source
• [2] Chang P, Lee KH, Shingu T, Hirayama T, Hall IH, Huang HC. Antitumor agents 50. 1 Morindaparvin-A, a new antileukemic anthraquinone, and alizarin-1-methyl ether from Morinda parvifolia, and the antileukemic activity of the related derivatives. J Nat Prod. 1982;45(2):206-210. (Reports natural isolation yield of ~0.00081%.) View Source

Morindaparvin A Research Procurement Application Scenarios: Where the Evidence Supports Prioritized Selection

P-388 Leukemia Model Mechanistic Studies: Selecting the Most Potent M. parvifolia-Derived Anthraquinone Probe

For laboratories investigating antileukemic mechanisms using the P-388 murine lymphocytic leukemia model, Morindaparvin A (ED₅₀ = 1.85 μg/mL) is the most potent anthraquinone isolated from M. parvifolia, outperforming its congener Morindaparvin B by approximately 5.7-fold and dramatically exceeding the co-isolated alizarin-1-methyl ether (completely inactive) [1][2]. The demonstrated in vivo antileukemic activity of the parent M. parvifolia methanolic extract (T/C = 152% at 50 mg/kg/day, I.P.) and Morindaparvin A alone (T/C = 129% at 10 mg/kg/day) provides additional confidence for in vivo follow-up studies [1]. Researchers focused on leukemia-specific anthraquinone mechanisms should prioritize Morindaparvin A over generic anthraquinone standards (alizarin, emodin) that lack comparable P-388 potency data.

Anthraquinone Pharmacophore Mapping: Using the Methylenedioxy Scaffold as a Redox-Silent DNA Intercalator Probe

For structure-activity relationship (SAR) studies seeking to decouple DNA intercalation from redox cycling in anthraquinone pharmacology, Morindaparvin A provides a uniquely clean probe. Its 1,2-methylenedioxy substitution eliminates the two free phenolic hydroxyl groups present in its synthetic precursor alizarin — hydroxyls that are well-established contributors to quinone redox cycling and ROS generation [1]. The reported DNA intercalation mechanism [2] in the absence of redox-active -OH groups makes Morindaparvin A a superior control compound for distinguishing intercalation-driven cytotoxicity from oxidative stress-driven cytotoxicity, a distinction that cannot be made with alizarin, emodin, or other hydroxyl-bearing anthraquinones.

Heterocycle-Fused Anthraquinone Lead Optimization: Building on a Pre-Validated Natural Product Template

Medicinal chemistry programs targeting the anthraquinone chemotype for anticancer lead discovery can leverage Morindaparvin A as a structurally characterized, biologically validated starting point. The Chang and Chen (1996) study provides direct precedent: Morindaparvin A (2a) served as the lead template from which four distinct heterocycle-fused anthraquinone series were designed and synthesized, with derivative 4b demonstrating pan-carcinoma cytotoxic activity [1]. The published total synthesis by Khanapure and Biehl (1989) further provides a modular route that can be adapted for analog generation [2]. This combined validation — biological, structural, and synthetic — is not available for most other natural anthraquinones, reducing the uncertainty associated with selecting a starting scaffold for anthraquinone-focused lead optimization.

Natural Product Reference Standard Procurement for Morinda Species Chemotaxonomy and Quality Control

Morindaparvin A is a taxonomically specific secondary metabolite first characterized from Morinda parvifolia and subsequently reported from Galianthe thalictroides and Paederia foetida [1]. Its distinctive 1,2-methylenedioxyanthraquinone structure — readily distinguishable from hydroxyl-bearing anthraquinones by the absence of phenolic -OH signals in ¹H NMR and IR — makes it an ideal chemotaxonomic marker. For quality control laboratories performing phytochemical authentication of Morinda or related Rubiaceae botanical materials, Morindaparvin A offers a structurally unambiguous reference standard with published spectroscopic fingerprints (¹H NMR: δ 6.32, 2H, s, -OCH₂O-; δ 7.15, d, J=9 Hz, H-3; δ 7.98, d, J=9 Hz, H-4; IR: 1675, 1587, 1455 cm⁻¹; UV λmax: 248, 273 nm) [2].

Application

Selection Property

Validation Focus

P-388 leukemia cell-model studies

M. parvifolia-derived anthraquinone with highest reported P-388 response

Cell-model endpoint response vs. congener and co-isolate comparators

DNA intercalation vs. redox cycling SAR studies

Redox-silent methylenedioxy scaffold without free phenolic -OH

Mechanism-of-action decoupling (intercalation vs. oxidative stress)

Anthraquinone lead diversification programs

Pre-validated natural product template with published synthetic route

Analog generation feasibility and prior active-derivative precedent

Rubiaceae phytochemical authentication

Taxonomically specific 1,2-methylenedioxyanthraquinone marker

Spectroscopic fingerprint match (¹H NMR, IR, UV) and absence of phenolic -OH signals

• [1] Chang P, Lee KH, Shingu T, Hirayama T, Hall IH, Huang HC. Antitumor agents 50. 1 Morindaparvin-A, a new antileukemic anthraquinone, and alizarin-1-methyl ether from Morinda parvifolia, and the antileukemic activity of the related derivatives. J Nat Prod. 1982;45(2):206-210. DOI: 10.1021/np50020a017. View Source
• [2] Chang P, Lee KH. Cytotoxic antileukemic anthraquinones from Morinda parvifolia. Phytochemistry. 1984;23(8):1733-1736. DOI: 10.1016/S0031-9422(00)83480-9. View Source