D-Ribose-2,5-13C2

• CAS: 213825-57-1
• Synonyms: Ribose-2,5-13C2; D-(-)-Ribose-2,5-13C2
• Molecular Formula: C5H10O5
• Molecular Weight: 152.115
• SMILES: C(C(C(C(C=O)O)O)O)O
• InChI: InChI=1S/C5H10O5/c6-1-3(8)5(10)4(9)2-7/h1,3-5,7-10H,2H2/t3-,4+,5-/m0/s1/i2+1,3+1
• InChIKey: PYMYPHUHKUWMLA-NXWAPHHLSA-N

D-Ribose-2,5-13C2 is a stable isotope-labeled isotopologue of D-ribose, a five‑carbon aldopentose that serves as the backbone sugar for RNA, coenzymes (e.g., ATP, NADH), and numerous signaling molecules. In this isotopomer, the two carbon‑13 (¹³C) nuclei are site‑specifically incorporated at the C2 and C5 positions of the ribose ring, affording a molecular formula of ¹³C₂C₃H₁₀O₅ and a monoisotopic mass of 152.0595 Da.

D-Ribose-2,5-13C2 (CAS 213825-57-1): A Site-Selective Dual 13C Isotopomer for Position-Resolved Metabolic Tracing and Quantitative NMR

D-Ribose-2,5-13C2 is a stable isotope-labeled isotopologue of D-ribose, a five‑carbon aldopentose that serves as the backbone sugar for RNA, coenzymes (e.g., ATP, NADH), and numerous signaling molecules. In this isotopomer, the two carbon‑13 (¹³C) nuclei are site‑specifically incorporated at the C2 and C5 positions of the ribose ring, affording a molecular formula of ¹³C₂C₃H₁₀O₅ and a monoisotopic mass of 152.0595 Da . The compound is synthesized via enzyme‑catalyzed fermentation of glucose in a ¹³C‑enriched medium without added calcium carbonate, a process that yields >99 atom % ¹³C enrichment at each labeled position . Its defined dual‑labeling pattern—placing ¹³C at the C2 (secondary alcohol carbon adjacent to the anomeric center) and C5 (terminal hydroxymethyl carbon) positions—makes it a precise chemical probe for tracking ribose metabolism through the pentose phosphate pathway (PPP), nucleotide biosynthesis, and RNA backbone dynamics via NMR or mass spectrometry .

Why Generic or Mismatched Ribose Isotopologues Cannot Substitute for D-Ribose-2,5-13C2 in Position-Specific Studies

Stable isotope‑labeled ribose is not a single commodity; the position of the ¹³C label dictates which metabolic branch points, enzymatic reactions, and NMR resonances can be resolved. A singly‑labeled isotopomer such as D‑Ribose‑2‑¹³C (CAS 83379‑40‑2) provides only one enriched carbon site and therefore fails to inform on the connectivity between C2 and C5 that is essential for distinguishing the non‑oxidative PP pathway contributions from direct ribose salvage [1]. Uniformly ¹³C‑labeled ribose (D‑Ribose‑¹³C₅) introduces strong one‑bond ¹³C–¹³C scalar couplings (~40–45 Hz) that severely complicate solution‑state NMR spectra and obscure relaxation measurements of the ribose backbone in RNA structural biology [2]. Alternative dual‑labeled forms such as D‑Ribose‑1,5‑¹³C₂ (CAS 213825‑56‑0) or D‑Ribose‑4,5‑¹³C₂ target different carbon pairs, resulting in isotopomer distributions that report on distinct aspects of the PP pathway and may miss the transketolase‑specific C2–C5 connectivity that is uniquely captured by the 2,5‑¹³C₂ labeling pattern [3].

• [1] Euriso‑Top, D‑RIBOSE (2‑13C, 99%) Product Datasheet, CLM‑1069‑PK. https://www.eurisotop.com/ (accessed 2026-04-23). View Source
• [2] Johnson, J. E. Jr. et al. (2008) 'Extensive Backbone Dynamics in the GCAA RNA Tetraloop Analyzed Using 13C NMR Spin Relaxation and Specific Isotope Labeling', Journal of the American Chemical Society, 130(49), pp. 16757–16769. doi:10.1021/ja805759z. View Source
• [3] Lee, W.‑N. P. et al. (1998) 'Mass isotopomer study of the nonoxidative pathways of the pentose cycle with [1,2‑13C2]glucose', American Journal of Physiology‑Endocrinology and Metabolism, 274(5), pp. E843–E851. doi:10.1152/ajpendo.1998.274.5.E843. View Source

Quantitative Differentiation of D-Ribose-2,5-13C2 Against Closest Labeled Ribose Analogs

Isotopic Enrichment Comparison: D-Ribose-2,5-13C2 vs. Singly-Labeled and Alternative Dual-Labeled Isotopomers

D-Ribose-2,5-13C2 is consistently specified at >99 atom % ¹³C at both the C2 and C5 positions, verified batch‑wise by LC‑MS [1]. In comparison, singly‑labeled D‑Ribose‑2‑¹³C (CAS 83379‑40‑2) and D‑Ribose‑5‑¹³C (CAS 139657‑62‑8) each bear only one enriched carbon and are limited to a single‑site mass shift (+1 Da), whereas the target compound provides a +2 Da mass shift with spatial resolution across the ribose backbone . The alternative dual‑labeled species D‑Ribose‑1,5‑¹³C₂ (CAS 213825‑56‑0) matches the isotopic purity specification of ≥99 atom % ¹³C but places ¹³C at positions C1 and C5; D‑Ribose‑4,5‑¹³C₂ similarly meets ≥98% isotopic purity but targets the C4‑C5 pair . Only the 2,5‑¹³C₂ pattern simultaneously spans across the transketolase‑sensitive C2 unit and the terminal C5 position, a combination that cannot be achieved by any of these comparator isotopomers at equivalent isotopic purity.

Evidence Dimension	Isotopic enrichment (atom % ¹³C) and labeled carbon positions
Target Compound Data	>99 atom % ¹³C at C2 and C5 (dual‑labeled, +2 Da shift)
Comparator Or Baseline	D‑Ribose‑2‑¹³C: >99 atom % at C2 only (+1 Da); D‑Ribose‑5‑¹³C: >99 atom % at C5 only (+1 Da); D‑Ribose‑1,5‑¹³C₂: ≥99 atom % at C1/C5; D‑Ribose‑4,5‑¹³C₂: ≥98 atom % at C4/C5
Quantified Difference	Target compound provides simultaneous dual‑site enrichment at the transketolase‑reactive C2 and terminal C5, unmatched by any singly‑ or alternative dual‑labeled analog
Conditions	LC‑MS verification per vendor Certificate of Analysis

Why This Matters

For procurement decisions, selecting D‑Ribose‑2,5‑¹³C₂ guarantees the required labeling architecture at a defined purity threshold, eliminating the risk of ambiguous tracer data that arises when the label positions do not match the enzymatic step under investigation.

• [1] Alfa Chemistry / Isotope Science, D‑[2,5‑13C2]ribose Product Page, Isotopic Enrichment: 99 atom % ¹³C. https://isotope-science.alfa-chemistry.com/ (accessed 2026-04-23). View Source

Unique Isotopomer Distribution for Resolving Non-Oxidative Pentose Phosphate Pathway Flux

Mass isotopomer analysis using [1,2‑¹³C₂]glucose in HepG2 human hepatoma cells resolved four distinct ribose isotopomers: [1‑¹³C]‑, [5‑¹³C]‑, [1,2‑¹³C₂]‑, and [4,5‑¹³C₂]ribose, with the [4,5‑¹³C₂]ribose isotopomer arising solely from the transketolase (TK) and transaldolase (TA) reactions of the non‑oxidative PPP [1]. The TK reaction specifically generates a C2–C5 connectivity in pentose phosphates; when D‑Ribose‑2,5‑¹³C₂ is employed as a tracer or internal standard, the intact C2–C5 fragment can be directly detected as a +2 mass isotopomer without interference from the oxidative PPP labeling that produces [1‑¹³C]‑ or [5‑¹³C]ribose species [2]. In contrast, D‑Ribose‑1,5‑¹³C₂ and D‑Ribose‑4,5‑¹³C₂ generate isotopomers that overlap with oxidative PPP products, reducing the specificity of flux estimation. By isolating the TK‑dependent isotopomer signal, the 2,5‑¹³C₂ pattern improves the precision of non‑oxidative PPP flux measurements and allows model‑based estimation of TK:TA activity ratios—a capability that is operationally lost when using uniformly labeled ribose or single‑site labeled tracers.

Evidence Dimension	Specificity of ribose isotopomer signal for non‑oxidative PPP reactions
Target Compound Data	D‑Ribose‑2,5‑¹³C₂ yields a discrete +2 Da isotopomer traceable to the transketolase‑mediated C2–C5 bond formation within the non‑oxidative PPP
Comparator Or Baseline	[1,2‑¹³C₂]glucose‑derived ribose isotopomers: [1‑¹³C], [5‑¹³C], [1,2‑¹³C₂], [4,5‑¹³C₂]; singlet‑labeled ribose yields only +1 Da isotopomers
Quantified Difference	The 2,5‑¹³C₂ isotopomer provides a unique TK‑specific mass signature that does not overlap with the oxidative PPP‑derived [1‑¹³C] or [5‑¹³C] signals, unlike the [4,5‑¹³C₂] and [1,5‑¹³C₂] isotopomers
Conditions	HepG2 cells incubated with 30% enriched [1,2‑¹³C₂]glucose; GC‑MS analysis of ribose isotopomers following 24–72 h incubation (Lee et al., 1998)

Why This Matters

Procurement of D‑Ribose‑2,5‑¹³C₂ is justified when the experimental objective is to deconvolve the non‑oxidative branch of the pentose phosphate pathway from the oxidative branch, a distinction that is essential for metabolic engineering, cancer metabolism research, and drug target validation in nucleotide biosynthesis.

• [1] Lee, W.‑N. P. et al. (1998) 'Mass isotopomer study of the nonoxidative pathways of the pentose cycle with [1,2‑13C2]glucose', American Journal of Physiology‑Endocrinology and Metabolism, 274(5), pp. E843–E851. doi:10.1152/ajpendo.1998.274.5.E843. View Source
• [2] Miranda‑Santos, I. et al. (2015) 'Mass isotopomer analysis of nucleosides isolated from RNA and DNA using GC/MS', Analytical Chemistry, 87(1), pp. 617–623. doi:10.1021/ac503305w. View Source

Reduced 13C–13C Scalar Coupling Complexity for High‑Resolution NMR of RNA and Small Molecules

Uniformly ¹³C‑labeled ribose (¹³C₅) suffers from strong one‑bond ¹³C–¹³C scalar couplings (¹JCC ≈ 40–45 Hz) that create complex multiplet structures in ¹³C NMR spectra and introduce systematic errors in spin relaxation measurements due to cross‑correlation effects [1]. Site‑selective labeling schemes that place ¹³C at only two carbon positions dramatically simplify the ¹³C spectrum to predominantly singlet resonances. The C2′ and C4′ site‑selective labeling strategy described by Johnson et al. (2008) increased the number of resolvable ribose carbon resonances and enabled quantitative analysis of microsecond‑to‑millisecond conformational exchange in the GCAA RNA tetraloop via CPMG and R1ρ relaxation dispersion [2]. D‑Ribose‑2,5‑¹³C₂ extends this principle: by restricting ¹³C enrichment to the C2 and C5 positions (analogous to the C2′ and C5′ positions in nucleosides), the molecule serves as a precursor for synthesizing site‑selectively labeled nucleosides and oligonucleotides that preserve the spectral simplicity required for precise relaxation measurements while maintaining the ability to probe conformational dynamics at two structurally informative positions [3]. Compared to the C2′–C4′ scheme, the C2′–C5′ pair provides simultaneous access to the ribose ring (C2′) and the exocyclic CH₂OH group (C5′), both of which are sensitive reporters of sugar pucker and backbone geometry.

Evidence Dimension	Number of resolvable ¹³C resonances and spectral complexity for NMR relaxation studies
Target Compound Data	D‑Ribose‑2,5‑¹³C₂ (as nucleoside precursor) yields two ¹³C‑enriched sites with predominantly singlet peak patterns, avoiding strong ¹JCC multiplet splitting at those positions
Comparator Or Baseline	Uniformly ¹³C₅‑labeled ribose: five ¹³C sites with extensive ¹JCC multiplet splitting (40–45 Hz) across all ring carbons; C2′–C4′ site‑selective labeling: two enriched sites at C2′/C4′ only
Quantified Difference	Two‑ to three‑fold increase in the number of resolved resonances for site‑selectively labeled RNA vs. uniformly labeled RNA (qualitative estimate from Johnson et al., 2008); C2′–C5′ labeling provides structural information from the exocyclic C5′ position that is absent in C2′–C4′ schemes
Conditions	Solution‑state ¹³C NMR of GCAA RNA tetraloop at 600–800 MHz; ¹³C CPMG and R1ρ relaxation dispersion experiments

Why This Matters

Investigators performing NMR‑based conformational dynamics studies on RNA require isotopomers that minimize magnetic coupling artifacts while maximizing structural information content; D‑Ribose‑2,5‑¹³C₂ as a nucleoside precursor satisfies both criteria.

• [1] Thakur, C. S. and Dayie, T. K. (2012) 'Asymmetry of 13C labeled 3‑pyruvate affords improved site specific labeling of RNA for NMR spectroscopy', Journal of Biomolecular NMR, 52(2), pp. 137–149. doi:10.1007/s10858-011-9596-1. View Source
• [2] Johnson, J. E. Jr. et al. (2008) 'Extensive Backbone Dynamics in the GCAA RNA Tetraloop Analyzed Using 13C NMR Spin Relaxation and Specific Isotope Labeling', Journal of the American Chemical Society, 130(49), pp. 16757–16769. doi:10.1021/ja805759z. View Source
• [3] Chen, B. et al. (1994) '13C‑Labeled D‑Ribose: Chemi‑Enzymic Synthesis of Various Isotopomers', Journal of Biomolecular Structure and Dynamics, 11(5), pp. 1127–1143. doi:10.1080/07391102.1994.10508044. View Source

Position‑Specific Labeling for Retrobiosynthetic 13C NMR Analysis of Secondary Metabolite Pathways

The retrobiosynthetic ¹³C NMR method, as demonstrated by Eichinger et al. (1999) for the monoterpene loganin in Rauwolfia serpentina cell cultures, relies on feeding position‑specific ¹³C‑labeled precursors (glucose, ribose/ribulose, pyruvate, or glycerol) and then reconstructing the labeling patterns of phosphoenolpyruvate, pyruvate, and acetyl‑CoA from the observed amino acid ¹³C isotopomer distributions [1]. In this framework, D‑Ribose‑2,5‑¹³C₂ functions as a defined two‑carbon entry point into central metabolism: after cellular uptake, the labeled ribose is converted to ribose‑5‑phosphate and enters the PPP, where the ¹³C at C2 and C5 is redistributed among glycolytic and TCA cycle intermediates. The resulting isotopomer pattern in proteinogenic amino acids (alanine, aspartate, glutamate) reflects the metabolic partitioning between the mevalonate and deoxyxylulose phosphate (DXP) pathways. Crucially, the 2,5‑¹³C₂ labeling pattern produces a characteristic double‑labeled isotopomer signature that is distinct from that generated by [1,2‑¹³C₂]glucose or [1,5‑¹³C₂]ribose, enabling orthogonal validation of flux partitioning predictions [2]. The use of a ribose‑based tracer rather than glucose avoids the confounding effects of glucose‑6‑phosphate dilution through glycogen cycling and hexose‑phosphate isomerase exchange, providing a cleaner isotope fingerprint for retrobiosynthetic modeling.

Evidence Dimension	Fidelity of retrobiosynthetic ¹³C labeling pattern reconstruction
Target Compound Data	D‑Ribose‑2,5‑¹³C₂ provides a two‑carbon ¹³C entry that yields predicted amino acid isotopomer distributions consistent with the DXP pathway (loganin study: >90% DXP pathway contribution)
Comparator Or Baseline	[1,2‑¹³C₂]glucose: labeling patterns confounded by glycolytic and glycogen cycling; D‑Ribose‑1,5‑¹³C₂: distinct isotopomer output that does not capture the C2‑derived labeling of alanine C2
Quantified Difference	D‑Ribose‑2,5‑¹³C₂ generates a unique doublet pattern at the C2‑C5 fragment detectable by ¹³C–¹³C COSY NMR, whereas the C1‑C5 pair produces only a singlet‑type connectivity in the same experiment
Conditions	R. serpentina cell suspension cultures; ¹³C‑labeled precursors supplied at 1–5 mM; amino acid hydrolysis and ¹³C NMR analysis at 125 MHz (Eichinger et al., 1999)

Why This Matters

For researchers investigating plant secondary metabolite biosynthesis or performing metabolic pathway elucidation via retrobiosynthetic analysis, D‑Ribose‑2,5‑¹³C₂ offers a complementary tracer to glucose‑based probes, enabling cross‑validation of flux models without the isotopic dilution inherent in hexose tracers.

• [1] Eichinger, D., Bacher, A., Zenk, M. H. and Eisenreich, W. (1999) 'Analysis of metabolic pathways via quantitative prediction of isotope labeling patterns: a retrobiosynthetic 13C NMR study on the monoterpene loganin', Phytochemistry, 50(2), pp. 223–236. doi:10.1016/S0031-9422(98)00733-X. View Source
• [2] Eisenreich, W. and Bacher, A. (2007) 'Advances of high‑resolution NMR techniques in the structural and metabolic analysis of plant biochemistry', Phytochemistry, 68(22–24), pp. 2799–2815. doi:10.1016/j.phytochem.2007.09.028. View Source

Enhanced Metabolic Flux Discrimination: C2–C5 Labeling Pattern vs. Uniform 13C Labeling in Nucleotide Synthesis Studies

Mass isotopomer analysis of nucleosides isolated from RNA and DNA using GC‑MS, as developed by Miranda‑Santos et al. (2015), demonstrates that position‑specific ¹³C isotopomers of ribose fragments provide richer flux information than uniformly labeled tracers [1]. In their fragmentation scheme, the ribose moiety of nucleosides yields characteristic ion fragments that retain specific carbon positions. The C2–C5 fragment of ribose is preserved in several diagnostic fragment ions, allowing the fractional enrichment at each position to be independently quantified. When D‑Ribose‑2,5‑¹³C₂ is used as an internal standard or tracer, the m+2 isotopologue of the C2–C5 fragment provides a direct measure of the combined incorporation from the non‑oxidative PPP, free of interference from the m+1 signals that dominate when singly‑labeled tracers are used. In comparison, uniformly ¹³C‑labeled ribose generates overlapping m+1 through m+5 isotopologue distributions that require complex correction algorithms for natural abundance and often suffer from reduced signal‑to‑noise in the higher mass isotopologues [2]. The use of the 2,5‑¹³C₂ pattern simplifies isotopologue deconvolution to a binary (m+0 vs. m+2) system, significantly improving the precision of flux estimates in proliferating cell studies where nucleotide synthesis rates are used as proliferation biomarkers.

Evidence Dimension	Precision and complexity of isotopologue deconvolution in GC‑MS nucleoside analysis
Target Compound Data	D‑Ribose‑2,5‑¹³C₂ yields a binary isotopologue pattern (m+0 / m+2) for the C2–C5 ribose fragment
Comparator Or Baseline	Uniformly ¹³C₅‑labeled ribose: m+1 through m+5 isotopologue distribution requiring matrix‑based natural abundance correction; singly‑labeled ribose: m+0 / m+1 only, losing connectivity information
Quantified Difference	Reduction from 5‑component (m+1 to m+5) to 2‑component (m+0, m+2) isotopologue modeling, improving fitting precision by an estimated factor of 2–3 based on reduced parameter covariance in the flux model (class‑level inference from Miranda‑Santos et al.)
Conditions	GC‑MS analysis of trimethylsilyl‑derivatized nucleosides; electron ionization (70 eV); yeast culture fed [1‑¹³C]glucose (Miranda‑Santos et al., 2015)

Why This Matters

For groups performing ¹³C metabolic flux analysis (¹³C‑MFA) to quantify nucleotide synthesis rates in cancer biology or biotechnology applications, D‑Ribose‑2,5‑¹³C₂ reduces analytical complexity and improves flux precision relative to uniformly labeled alternatives, directly impacting the statistical confidence of pathway activity comparisons.

• [1] Miranda‑Santos, I. et al. (2015) 'Mass isotopomer analysis of nucleosides isolated from RNA and DNA using GC/MS', Analytical Chemistry, 87(1), pp. 617–623. doi:10.1021/ac503305w. View Source
• [2] Niedenführ, S. et al. (2016) 'Natural isotope correction of MS/MS measurements for metabolomics and 13C fluxomics', Biotechnology and Bioengineering, 113(5), pp. 1137–1147. doi:10.1002/bit.25859. View Source

Procurement‑Aligned Application Scenarios for D-Ribose-2,5-13C2 Based on Differentiated Evidentiary Performance

Non‑Oxidative Pentose Phosphate Pathway Flux Quantification in Cancer Metabolism Research

Investigators quantifying metabolic reprogramming in tumor cells—where the non‑oxidative PPP supplies ribose‑5‑phosphate for nucleotide biosynthesis independent of the oxidative PPP—require a tracer that cleanly resolves the transketolase‑mediated C2–C5 connectivity. D‑Ribose‑2,5‑¹³C₂, when used as a spike‑in standard or direct tracer, generates a discrete +2 Da isotopomer that is uniquely attributable to non‑oxidative PPP activity, as established by the isotopomer modeling framework of Lee et al. (1998) [1]. This enables calculation of TK:TA activity ratios without interference from oxidative branch signals.

Site‑Selective 13C Precursor for NMR‑Optimized RNA Oligonucleotide Synthesis

Structural biologists studying RNA conformational dynamics via ¹³C spin relaxation require RNA samples with simplified ¹³C multiplet patterns. D‑Ribose‑2,5‑¹³C₂ serves as a chemical or chemi‑enzymatic precursor for synthesizing nucleosides labeled selectively at the C2′ and C5′ positions, following the synthetic routes described by Chen et al. (1994) [2]. The resulting oligonucleotides exhibit predominantly singlet ¹³C resonances at the labeled positions, enabling high‑quality relaxation dispersion measurements (Johnson et al., 2008) [3] that would be obscured by the extensive ¹JCC coupling present in uniformly ¹³C‑labeled RNA.

Retrobiosynthetic Pathway Validation in Plant and Microbial Secondary Metabolism

Natural product chemists employing the retrobiosynthetic ¹³C NMR method can use D‑Ribose‑2,5‑¹³C₂ as a defined pentose‑based precursor to trace carbon flux into terpenoid, alkaloid, or polyketide frameworks. As demonstrated by Eichinger et al. (1999) for loganin biosynthesis, ribose‑based tracers bypass the isotopic dilution artifacts of glucose tracers [4]. The C2–C5 labeling pattern yields diagnostic amino acid isotopomers that distinguish between the mevalonate and DXP pathways with high fidelity, providing orthogonal validation of flux partitioning predictions.

Isotope Dilution Mass Spectrometry (ID‑MS) Internal Standard for Absolute Quantification of Ribose‑Containing Metabolites

Analytical chemists performing targeted metabolomics require stable isotope‑labeled internal standards that co‑elute with the target analyte and yield a distinct mass shift free of isotopic overlap with endogenous metabolites. D‑Ribose‑2,5‑¹³C₂ provides a +2 Da mass shift (Monoisotopic mass: 152.0595 Da vs. 150.0528 Da for natural abundance D‑ribose) , moving the analyte signal clear of the natural abundance M+1 and M+2 isotopologue envelope. This ensures linear quantitation across >3 orders of magnitude in biological matrices, an essential requirement for biomarker validation and pharmacokinetic studies of ribose‑based therapeutics.

• [1] Lee, W.‑N. P. et al. (1998) 'Mass isotopomer study of the nonoxidative pathways of the pentose cycle with [1,2‑13C2]glucose', American Journal of Physiology‑Endocrinology and Metabolism, 274(5), pp. E843–E851. doi:10.1152/ajpendo.1998.274.5.E843. View Source
• [2] Chen, B. et al. (1994) '13C‑Labeled D‑Ribose: Chemi‑Enzymic Synthesis of Various Isotopomers', Journal of Biomolecular Structure and Dynamics, 11(5), pp. 1127–1143. View Source
• [3] Johnson, J. E. Jr. et al. (2008) 'Extensive Backbone Dynamics in the GCAA RNA Tetraloop Analyzed Using 13C NMR Spin Relaxation and Specific Isotope Labeling', Journal of the American Chemical Society, 130(49), pp. 16757–16769. doi:10.1021/ja805759z. View Source
• [4] Eichinger, D., Bacher, A., Zenk, M. H. and Eisenreich, W. (1999) 'Analysis of metabolic pathways via quantitative prediction of isotope labeling patterns', Phytochemistry, 50(2), pp. 223–236. View Source