Figure 1.
Major volatile compounds from ginger and turmeric rhizomes as analyzed by GC/MS.
(A–F) Total ion chromatograms of 7 month old yellow ginger (GY) and turmeric (T) tissues are shown; (A) yellow ginger rhizome, (B) root, (C) leaf, (D, E) turmeric rhizome, (F) root and (G) leaf. Both ginger and turmeric produce (−)-α-zingiberene (green) and (−)-β-sesquiphellandrene (blue). However, only turmeric produces (+)-α-turmerone (orange) and (+)-β-turmerone (red). The total amounts of (+)-α-turmerone and (+)-β-turmerone are variable according to sample (D, E), but their ratios remain relatively constant. β-Bisabolene is shown as yellowish green. *: standard compounds (from early retention time to later retention time, p-chloro toluene, 1,2,4-trimethyl benzene and 2,4,5,6-tetrachloro-m-xylen). Compounds identified: 1, tricyclene; 2, α-thujene (3-thujene); 3, α-pinene; 4, camphene; 5, sabinene (4(10)-thujene); 6, β-pinene; 7, myrcene; 8, α-phellandrene; 9, 3-carene; 10, α-terpinene; 11, limonene; 12, β-phellandrene; 13, 1,8-cineole (eucalyptol); 14, (E)-β-ocimene; 15, γ-terpinene; 16, (Z)-sabinene hydrate; 17, p-mentha-1,4(8)-diene (terpinolene); 18, linalool; 19, (E)-pinene hydrate ((E)-pinan-2-ol); 20, α-campholenal; 21, (E)-pinocarveol (2(10)-pinen-3-ol); 22, camphor; 23, (E)-verbenol ((E)-2-pinen-4-ol); 24, β-citronellal; 25, pinocarvone (2(10)-pinen-3-one); 26, borneol (endo-borneol); 27, p-menth-1-en-8-ol (α-terpineol); 28, myrtenal; 29, verbenone (2-pinen-4-ol); 30, β-citronellol (3,7-dimethyl-6-octen-1-ol); 31, neral (β-citral); 32, (E)-geraniol; 33, geranial (α-citral); 34, bornyl acetate; 35, myrtenyl acetate (2-pinen-10-ol acetae); 36, δ-elemene; 37, α-terpinyl acetate; 38, citronellyl acetate; 39, nerol acetate; 40, unknown ((+)-cyclosativene-like); 41, (+)-cyclosativene; 42, α-copaene; 43, geraniol acetate; 44, β-elemene; 45, unknown (7-epi-sesquithujene-like); 46, (E)-caryophyllene (β-caryophyllene); 47, β-copaene; 48, γ-elemene; 49, (E)-α-bergamotene; 50, α-humulene (α-caryophyllene); 51, (E)-β-farnesene; 52, allo-aromadendrene; 53, γ-curcumene; 54, germacrene D; 55, ar-curcumene; 56, β-selinene (eudesma-4(14),11-diene); 57, (−)-α-zingiberene; 58, γ-amorphene; 59, α-muurolene; 60, (E,E)-α-farnesene; 61, β-bisabolene; 62, γ-cadinene; 63, β-curcumene; 64, 7-epi-α-selinene; 65, (−)-β-sesquiphellandrene; 66, (E)-γ-bisabolene; 67, α-elemol; 68, unknown (cis-sesquisabinene hydrate-like); 69, germacrene B; 70, (E)-nerolidol; 71, caryophyllene oxide; 72, unknown (trans-sesquisabinene hydrate-like1); 73, humulene oxide II; 74, unknown (trans-sesquisabinene hydrate-like2); 75, unknown (trans-sesquisabinene hydrate-like3); 76, (+)-α-turmerone; 77, epi-α-bisabolol/α-bisabolol; 78, (+)-β-turmerone; 79, (Z)-α-atlantone; 80, α-oxobisabolene; 81, (E)-α-atlantone.
Figure 2.
Similarity tree of cloned full length mono- and sesquiterpene synthases and unitrans sequences of di-, tri- and tetra-terpene synthases from ginger and turmeric.
The neighbor-joining tree was generated by ClustalX with ginger and turmeric sequences and 181 additional TPSs from GenBank. The non-ginger/turmeric TPSs were removed from the tree for clarity. The STPS cluster is separate from other terpene synthases. Linalool/nerolidol synthase (MT00) is located outside of the MTPS cluster. Major product(s) of each corresponding recombinant protein is/are shown next to each gene name, as is the tissue used for cloning.
Table 1.
Monoterpenes produced by ginger and turmeric terpene synthases.
Table 2.
Sesquiterpenes produced by ginger and turmeric terpene synthases.
Figure 3.
Analysis of ST00A and ST00B functions when proteins were expressed in the yeast strain, EPY219.
(A–E) Total ion chromatograms are displayed: (A) EPY219 without pESC-URA-ST00A or pESC-URA-ST00B, 18°C, 2 days; (B) EPY219 expressing ST00A, 18°C, 2 days; (C) EPY219 expressing ST00B, 18°C, 2 days; (D) EPY219 without pESC-URA-ST00A or pESC-URA-ST00B, 18°C, 8 days; and (E) EPY219 expressing ST00A, 18°C, 4 days. D2 and E2 are boxed regions of D and E panels to show very small peaks (expanded y-axes). Additional mass spectra comparisons are in Figure S4. Products/compounds identified: 1, (−)-α-zingiberene; 2, (−)-β-sesquiphellandrene; 3, β-bisabolene; 4, unknown (trans-sesquisabinene hydrate-like2); 5, unknown (7-epi-sesquithujene-like); 6, trans-α-bergamotene; 7, γ-curcumene; 8, ar-curcumene; 9, (E)-γ-bisabolene; 10, unknown (α-eudesmol-like); 11, γ-eudesmol; 12, unknown (trans-sesquisabinene hydrate-like3); 13, α-acorenol; 14, (E)-β-farnesene; 15, (E)-nerolidol.
Figure 4.
Analysis of linalools produced by MT00 and MT17A2 in chiral column.
MT00 produces (S)-(+)-linalool and MT17A2 synthesizes (R)-(−)-linalool.
Figure 5.
Proposed synthesis of (−)-neointermedeol, α-elemol and β-elemene from FPP.
These are major products of ST02A4, ST02B and ST02C respectively and share the same mechanistic pathway. Δ: heat during GC run.
Figure 6.
Proposed mechanism for the formation of caryophyllene related compounds, the proposed major products of ST07/ST07A.
The mass spectrum of the major product of ST07A is very similar to caryophyllenyl alcohol, caryolan-1-ol and caryolan-8-ol. However, it seems unlikely that caryophyllenyl alcohol would be produced by a single terpene synthase enzyme.
Figure 7.
Comparison of ST07A major product to (−)-caryolan-1-ol and (+)-caryolan-1-ol.
(A) The major product when ST07A is expressed in yeast has the same retention time as both (−)- and (+)-caryolan-1-ol in a general purpose GC column. (B) However, use of a chiral column confirmed that the major product of ST07A is the (−)-caryolan-1-ol enantiomer. (C) Mass spectra of the major product of ST07A and (−)- and (+)-caryolan-1-ol are the same.
Figure 8.
Proposed reaction of α-zingiberene/β-sesquiphellandrene hydroxylase.
ST00A/ST00B produces (−)-α-zingiberene and (−)-β-sesquiphellandrene, which are hydroxylated by α-zingiberene/β-sesquiphellandrene hydroxylase, leading to the requirement of a dehydrogenase to make the ketone forms, (+)-α- and (+)-β-turmerones. The percentages shown inside the box represent ST00A/ST00B product percentage. The color used in this figure is compatible with the color used in Figure 1. Substrate of curlone: Cyclohexene, 3-[(1S)-1,5-dimethyl-4-hexen-1-yl]-6-methylene-, (3S)-.
Figure 9.
Analysis of the products of the putative α-zingiberene/β-sesquiphellandrene hydroxylase.
P1A, P1A2, P4, P4A and P4A2 are P450 monooxygenase candidates for α-zingiberene/β-sesquiphellandrene hydroxylase. (A–G) Total ion chromatograms are shown for yeast cell line INVSc1 with or without plasmids: (A) INVSc1 with no plasmid; (B) INVSc1 with pESC-TRP-Ob_CPR and pESC-URA-ST00A; (C) INVSc1 with pESC-URA-ST00A-P1A; (D) INVSc1 with pESC-URA-ST00A-P1A2; (E) INVSc1 with pESC-URA-ST00A-P4; (F) INVSc1 with pESC-URA-ST00A-P4A; and (G) INVSc1 with pESC-URA-ST00A-P4A2. (H) Peaks 1, 2, 3 and 4 are new peaks resulting from the action of the P450 that yield mass spectra and are similar to (−)-α-zingiberene and (−)-β-sesquiphellandrene. (I) Proposed fragmentation pathway for (−)-α-zingiberene, (+)-α-turmerone, and the proposed hydroxylated intermediate.
Figure 10.
Comparison of MT06 and MT06B modeled structures.
(A–C) MT06 and MT06B structures were modeled based on the structure of (4S)-limonene synthase from Mentha spicata as a template. (A) Alignment of all three structures in ribbon, (B) alignment of MT06 and MT06B in ribbon, and (C) side chains of MT06 and MT06B near bound ligand are shown. An additional amino acid (E573) in MT06B lies in the loop indicated with an arrow in B and does not appear to affect the structure. Y576 in MT06 and Y577 in MT06B are right after the loop and are aligned very well in C. Green: (4S)-limonene synthase from M. spicata, Blue: MT06, Orange: MT06B, Ligand with hashed surface: 2-fluorogeranyl diphosphate.
Figure 11.
Comparison of ST02A4, ST02B and ST02C modeled structures.
ST02A4, ST02B and ST02C structures were modeled with the structure of (+)-δ-cadinene synthase from Gossypium arboreum as a template. (A) Alignment of all three structures in ribbon, (B) alignment of ST02B and ST02C in ribbon, and (C) atoms/bonds of ST02B and ST02C are shown. C-terminal region and N-terminal regions of ST02B and ST02C are indicated with arrows in B. (D–G) Only the C-terminal region and N-terminal regions are shown. (D) The loop and helix structures with side chains are shown with W28 from the conserved RRX8W motif labeled. (E–G) Surfaces of (E) ST02A4, (F) ST02B and (G) ST02C are shown. Purple: (+)-δ-cadinene synthase from G. arboreum, Green: ST02A4, Blue: ST02B, Orange: ST02C, Ligand with hashed surface: 2-fluorofarnesyl diphosphate from aligned (+)-δ-cadinene synthase from G. arboreum.