Ginger and turmeric expressed sequence tags identify signature genes for rhizome identity and development and the biosynthesis of curcuminoids, gingerols and terpenoids

Abstract

Background: Ginger (Zingiber officinale) and turmeric (Curcuma longa) accumulate important pharmacologically active metabolites at high levels in their rhizomes. Despite their importance, relatively little is known regarding gene expression in the rhizomes of ginger and turmeric.

Results: In order to identify rhizome-enriched genes and genes encoding specialized metabolism enzymes and pathway regulators, we evaluated an assembled collection of expressed sequence tags (ESTs) from eight different ginger and turmeric tissues. Comparisons to publicly available sorghum rhizome ESTs revealed a total of 777 gene transcripts expressed in ginger/turmeric and sorghum rhizomes but apparently absent from other tissues. The list of rhizome-specific transcripts was enriched for genes associated with regulation of tissue growth, development, and transcription. In particular, transcripts for ethylene response factors and AUX/IAA proteins appeared to accumulate in patterns mirroring results from previous studies regarding rhizome growth responses to exogenous applications of auxin and ethylene. Thus, these genes may play important roles in defining rhizome growth and development. Additional associations were made for ginger and turmeric rhizome-enriched MADS box transcription factors, their putative rhizome-enriched homologs in sorghum, and rhizomatous QTLs in rice. Additionally, analysis of both primary and specialized metabolism genes indicates that ginger and turmeric rhizomes are primarily devoted to the utilization of leaf supplied sucrose for the production and/or storage of specialized metabolites associated with the phenylpropanoid pathway and putative type III polyketide synthase gene products. This finding reinforces earlier hypotheses predicting roles of this enzyme class in the production of curcuminoids and gingerols.

Conclusion: A significant set of genes were found to be exclusively or preferentially expressed in the rhizome of ginger and turmeric. Specific transcription factors and other regulatory genes were found that were common to the two species and that are excellent candidates for involvement in rhizome growth, differentiation and development. Large classes of enzymes involved in specialized metabolism were also found to have apparent tissue-specific expression, suggesting that gene expression itself may play an important role in regulating metabolite production in these plants.

Figure 1

Proposed biosynthetic pathway from L-Phe to diarylheptanoids and gingerol-related compounds in ginger and turmeric. Enzymes are as follows: PAL = phenylalanine ammonia lyase; C4H = cinnamate 4-hydroxylase; 4CL = 4-coumarate:CoA ligase; CST = p-coumaroyl shikimate transferase; C3´H = p-coumaroyl 5-O-shikimate 3´-hydroxylase; OMT = O-methyltransferase; CCOMT = caffeoyl-CoA O-methyltransferase; SAMS = S-adenosylmethionine synthetase; SAHC = S-adenosylhomocysteine. All conversions have been demonstrated in other species, except for those catalyzed by the polyketide synthases, the reductases, and the hydroxylases and OMTs that would convert bisdemethoxycurcumin via demethoxycurcumin to curcumin (indicated by dashed arrows). The polyketide synthases indicated represent two distinct classes that work in tandem: diketide synthases and curcuminoid/gingerol synthases. Different combinations of these two classes appear to be responsible for production of different compounds in these plants. The large red X and the solid arrows associated with the DKS1 + CURS reactions indicate that formation of curcumin has been demonstrated to proceed directly from feruloyl-CoA, and not through the orthodiol intermediate B. Compounds A and B, therefore, are not likely to be intermediates in the pathway to curcumin but instead are likely to be products of a different pair of PKS enzymes.

Figure 2

Ginger and turmeric PKSs that are not bona fide chalcone synthases cluster into two major groups, WtPKS1-like (DKS-like) and WtPKS2-like (CURS-like), in a neighbor-joining similarity tree (parsimony and maximum likelihood approaches produce similar trees). Abbreviations are as follows: WtPKS, Wachendorfia thyrsiflora (Haemodoraceae) polyketide synthase; CURS, turmeric curcumin synthase; DKS, diketide synthass; other PKSs as indicated. In many cases, branches of the tree represent a large number of related sequences, for example, the branch called “Marchantia paleacea PKSs” consisted of four closely related sequences, whereas the branch called “Large Cluster of Rice PKSs” represented over 20 sequences, and the branch called “Other Chalcone Synthases” represented over 100 sequences.

Figure 3

Comparison of GC/MS-determined 1,8-cineole levels to cineole synthase gene expression profiles generated by DNA microarray analysis. 1,8-Cineole levels in either turmeric or ginger are shown as bar graphs. Gene expression levels from normalized microarray spot intensities are shown as the line graph. Error bars indicate standard errors of the mean. The sequence for 1,8-cineole synthase used in this comparison was identified from ginger rhizomes. Both metabolite profiling and gene expression profiling of the ginger 1,8-cineole synthase match very well for ginger. However, 1,8-cineole produced in turmeric appears to be produced by an enzyme that is not the same as that from ginger. Bottom axis abbreviations represent developmental time courses for three cultivars, G: white ginger, F: turmeric variety FMO (Fat Mild Orange), and T: turmeric variety TYA (Thin Yellow Aromatic), respectively. Rh: rhizome, R: root, L: leaf. 2 M, 3 M, 4 M, 5 M, 7 M represent plants of different ages in months, e.g., 2 M: 2 months, 3 M: 3 months, etc.

Figure 4

Fraction of ESTs (standardized per 10,000 ESTs) of different classes of transcriptional regulators (GO category 0003677) in ginger and turmeric libraries. Library descriptions are listed under the graph, followed by additional information per library in the parenthesis: (number of unitrans, number of ESTs, number of unitrans with GO:0003677, number of ESTs with GO:0003677). ArRESTs: EST collection of all ginger and turmeric libraries within ArREST. ZO_CL_Rh_ESTs_ArREST: combined ESTs of all ginger and turmeric rhizome libraries within ArREST. CL_Ea_ESTs_ArREST: turmeric rhizome library. ZO_Rh_ESTs_ArREST: combined EST collection of two ginger rhizome libraries. ZO_Ec_ArREST: White ginger rhizome EST library. ZO_Ed_ArREST: Yellow ginger rhizome library. Values per category are shown in Additional file 1: Table S8.

Figure 5

Comparisons of microarray signal intensities of several rhizome-enriched transcription factor unitrans in ginger and turmeric. Each of these selected genes possesses signal intensities with rhizome–enriched expression in at least one of the species compared to other tissues and/or time-points with coefficients ≥ 2 and p-values ≤ 0.05.

Figure 6

Associations between MADS box transcription factors from ginger and turmeric, sorghum rhizome-enriched ESTs, and rice rhizomatous QTLs on rice chromosomes 2, 3, 6, 7, and 10. Orange boxes represent rhizome QTLs, while triangles indicate the spatial locations of MADS box genes on the rice chromosomes. Rice MADS box genes are illustrated using white triangles; black triangles are ginger/turmeric Blastx hits (e-values < e^-20). * denotes the best Blastx hits for the ginger and turmeric contig/rice gene comparison. † denotes a MADS box gene shown by microarray analysis to be preferentially expressed in the rhizome when compared to leaf or root tissues (Figure 3D & E).