Previous PageTable Of ContentsNext Page

Manipulating the lipid resorcinol pathway to enhance allelopathy in rice

Franck E. Dayan, Daniel Cook, Scott R. Baerson, and Agnes M. Rimando

USDA-ARS Natural Products Utilization Research Unit, P. O. Box 8048, University, MS 38677. USA. Email


About 4% of the rice cultivars have demonstrated allelopathic potential against some of the most troublesome weed species in paddy fields, such as barnyardgrass (Echinochloa crus-galli), redstem (Ammannia species), Cyperus species and ducksalad (Heteranthera limosa). A tremendous international effort is underway to produce rice lines with enhanced allelopathy through breeding techniques. Quantitative Trait Loci (QTL) analysis has associated the allelopathic trait with several rice chromosomes, suggesting that these weed-repressing varieties may produce more than one phytotoxin. Recent publications have shed some light on the nature of these phytotoxins. Allelopathic rice produces glycosides of lipid resorcinols, flavones, and benzoxazinoids, as well as momilactones and cyclohexenones. Lipid resorcinols are of particular interest to our research group because these secondary metabolites have been associated with pathogen resistance and allelochemical traits of other monocotyledonous species. Furthermore, these compounds have been identified as markers to evaluate the allelopathic potential of rice varieties. Genes putatively involved in the ring formation of these unusual resorcinols in rice have been identified. These novel polyketide synthases accept long chain fatty acid-CoA substrates instead of the usual coumaroyl-CoA substrate used by the more common polyketide synthases such as chalcone and stilbene synthases. The substrate specificity of some of these rice enzymes indicate that they may be involved in the biosynthetic pathway of lipid resorcinols.

Media summary

Rice plants produce several defense compounds, including lipid resorcinols. Manipulation of their biosynthesis may lead to varieties with enhanced natural defense mechanisms.

Key Words

Metabolic engineering, polyketide synthase, allelopathy, phytoalexin


Rice (Oryza spp) originated in Southeast Asia (Figure 1) and has been grown in China for human consumption as far back as the Sheng Nung period (c. 2700 B.C.). Rice is now cultivated throughout the tropical and subtropical regions of world. They can inhabit a variety of ecosystems and tolerate extremes in sun exposure and in moisture levels (Vaughan 1994).

Figure 1. Map showing the distribution of rice around the world (green). Area in red shows the geographical point of origin of the species.

The genus Oryza is a member of the Poaceae (grass family) and consists of 22 species distributed in nine genomes. Oryza species have been divided in the O. sativa, O. officinalis, O. ridleyi, and O. meyeriana species complexes (Vaughan 1994).

A large majority (90%) of the world rice production originates in Asia, with China and India being the largest producers. Rice is one of the most commonly consumed cereal grains, and as a result, has considerable economic and agricultural importance. The worldwide annual rice production exceeds 500 million metric tons, equalling that of wheat and maize. However, unlike wheat and corn, rice is almost entirely consumed by humans. Models predict that annual rice production will need to increase 800 million tons over the current levels by the year 2025 to supply the ever increasing worldwide demand (Pingali et al. 1997).

Rice allelopathy has been extensively studied in part because many cultivars and ancestral lines have exhibited significant allelopathic potential in the field. Several international efforts to generate allelopathic rice varieties are underway (Olofsdotter et al. 2002; Dilday et al. 2001; Kong et al. 2002). Thousands of varieties have been screened for allelopathic potential, and up to 4% of varieties can suppress important weeds such as barnyardgrass (Echinochloa crus-galli) (e.g., Dilday et al. 2001; Jensen et al. 2001) and Cyperus difformis (Kong et al. 2004a). While the level of weed management by allelopathy is not as good as that obtained with herbicides, herbicide use rates can be substantially reduced in paddy field planted with such varieties (Gealy et al. 2003).

Figure 2. Field study at the Dale Bumpers National Rice Research Center in Stuttgart, Arkansas, showing the effect of allelopathic rice (center plot) on the growth of barnyardgrass.

A study conducted with 111 rice cultivars in parallel between the Philippines and Korea showed that the allelopathic rice cultivars behaved similarly in the two geographical locations, indicating that the traits were more influenced by genetics than the environment. Studies on rice allelopathy suggest that this trait is inherited (Courtois and Olofsdotter 1998). Broad-sense heritability, which characterizes the strength of the relationship between phenotype and genotype, is rated on a scale of 0 to 1, and the higher the score, the stronger the relationship. The broad-sense heritability for the ability of rice to reduce the root growth of barnyardgrass is 0.85, which is above the threshold required by breeders to screen for a trait. QTL (Quantitative Trait Loci) analysis of allelopathic cultivars show that the allelopathic trait is quantitatively inherited. Four QTL located on three separate chromosomes were found to account for 35% of the total variation for the allelopathic phenotype in a population of 142 recombinant inbred rice lines derived from a cross between IAC 165 and CO 39 (Jensen et al. 2001).

QTL analysis suggests that more than one gene and probably several compounds are responsible for the allelochemical trait of certain rice varieties. This is in agreement with other research reporting several allelochemicals in rice. It is difficult to differentiate allelopathic effects from competition. Factors such as tillering and root biomass often result in weed suppressive characteristics. In particular, some of the most allelopathic cultivars have larger root masses than non-allelopathic cultivars, However, none of the QTL identified to contribute to allelopathy were linked to QTL associated with root morphology.

Unlike most other crops, there is a relatively high occurrence of allelopathy in rice (about 4%). One factor that may contribute to this may be the fact that the highly allelopathic cultivar Taichung Native 1 is present in the genetic background of most of the post-Green Revolution rice cultivars (Olofsotter 2001). Since the allelopathic trait is inheritable, many of these cultivars may possess some of the allelopathic character of Taichung Native 1, or that of other allelopathic lines in their background.

There is still some ambiguity as to the primary compounds responsible for the allelopathic potential observed in some rice species. However, the following compounds are examples of the phytotoxin classes found in root exudates of allelopathic rice varieties: momilactone B (Kato-Noguchi and Ino 2004), resorcinols (Bouillant et al. 1994), flavone, benzoxazinoids, and their respective glycosides (Kong et al. 2002); and a cyclohexenone (Kong et al. 2004a) (Figure 3). It is likely that more than one type of phytotoxin may play a role in combating weeds in the most allelopathic varieties of rice.

Figure 3. Structures of some of the allelochemicals isolated from rice.

It should be noted that these compounds have other biological activities involved in plant defense responses. For example, momilactone and the structurally related oryzalexins and phytocassanes have well characterized antimicrobial activities and their synthesis is stimulated upon infection with pathogenic fungi (VanEtten et al. 1994). Similarly, the cyclohexenone and flavone both inhibit the germination of common rice fungal pathogens (e.g., Pyricularia oryzae and Rhizoctonia solani). The 5-alkyl resorcinols have antibacterial and fungistatic properties and have been associated with plants natural defense mechanisms against fungal infections (Kozubek and Tyman 1999). Furthermore, lipid resorcinol have molluscicidal activities, with their potency being inversely proportional to the level of unsaturation of the alkyl tail.

Additionally, the allelopathic potential of rice plants can be stimulated in the presence of other plants by increasing the production of allelochemicals that may help them repress the growth of competitors. The synthesis of some rice allelochemicals, such as the flavone and cyclohexenone, can be stimulated in the presence of barnyardgrass in the surrounding (Kong et al. 2004b). This has not been demonstrated for alkyl resorcinols. However, our interest in other lipid resorcinol-derived allelochemicals, such as those produced in sorghum, has led our research group to identify, clone and characterize the substrate specificity of the key enzymes involved in the synthesis of the rice lipid resorcinols.


Cloning of rice polyketide synthase (PKS)

Polyketide synthase candidate orthologues from rice were identified by protein-protein blast using a sorghum PKS enzyme that had been shown to use long chain acyl-CoA starter units to form 5-alkyl resorcinols. Full length cDNA were amplified from a rice cDNA library using primer designed from the corresponding genomic sequence. Full length cDNA’s were cloned into the pET15b vector.

Expression and purification of rice PKS

The rice PKS clone was expressed in BL21(DE3) E. coli cells. Cultures were grown with 500 g/ml of carbenicillin, and were induced with 0.5 mM IPTG. Bacterial pellets were washed with 0.1% NaCl and stored at –80C. For purification of the his-tagged protein the cells were disrupted using a French Press at 1,500 psi of internal pressure. Benzonase (Novagen) was used to precipitate nucleic acids. Following centrifugation, the clear lysate was passed through a Hi-Trap chelating column (Amersham). The bound his-tagged protein was eluted with imidazole. The eluent was passed through a PD-10 desalting column equilibrated with potassium phosphate buffer containing 10% glycerol. After determining protein concentration with the Bio-Rad Protein Assay, the protein was stored at –80C in small aliquots. The samples remained active after several weeks of storage.

Substrate specificity

Assays were done microfuge tubes with a total of 200 l volume of 0.1 M potassium phosphate (pH 7) with 1 g PKS protein per assay, 40 M malonyl-CoA (a mix of “cold” and 14C malonyl-CoA), and 15 M “starter molecule” CoA. The protein and starter molecule were combined and the tube pre-warmed to 30C. The reaction was started with the addition of malonyl-CoA. The reaction was stopped with the addition of 1 ml of ethyl acetate and 5 l of 20% HCl. The tube was vortexed, centrifuged, and a 800 l aliquot of the ethyl acetate layer was transferred to another tube and dried under N2. The residue was suspended in 25 l of chloroform, a 5 l aliquot was placed in a scintillation tube for scintillation counting, and 10 l was applied on a aluminium-backed silica gel TLC plate and developed in the presence of 70% chloroform / 30% ethyl acetate solvent system for 15 min. Afterward, the plate was dried, covered with a thin layer of plastic and placed against a phosphor screen for 24 h. The image was visualized on a Cyclone phosphoimager system (Perkin Elmer Life Sciences).


Genetic information related to the allelopathy of rice, such as quantitative trait loci mapping of allelopathic traits (Jensen et al. 2001), have been developed, but no direct link between this genetic information to production of any particular allelochemical. The availability of the complete rice genome (Goff et al. 2001; Yu et al. 2001) now enables the identification of the genes involved in the production of the more important allelochemicals in rice. For example, this genomic information was used by Xu et al. (2004) to find that the gene for syn-copalyl diphosphate synthase plays a regulatory role in the synthesis of the momilactones and structurally related phytoalexins.

Using a similar approach, the rice genome available from the Institute for Genomic Research ( was surveyed for the presence of genes encoding for putative PKS enzymes. The analysis showed that there are 33 PKS-like genes in rice. Of these, 8 sequences were selected as candidates genes because they had significant similarity to the PKS genes involved in the synthesis of lipid resorcinol in sorghum.

Enzymes involved in the formation of alkyl resorcinol belong to the Type-III class PKS. These enzymes are typically homodimeric enzymes that possess a conserved Cys-His-Asn catalytic triad buried inside the active site. Chalcone synthase and stilbene synthase are the most studied Type-III PKS. Each monomer is typically approximately 400 amino acids and there is a high level of similarity between chalcone and stilbene synthases (75 to 90% amino acid sequence identity). It has been postulated that stilbene synthase has evolved from chalcone synthase (Tropf et al. 1994).

Both of these enzymes use coumaroyl-CoA as a starter unit and catalyze the sequencial condensation of acetyl units that are derived from malonyl-CoA. They also produce an identical linear tetraketide intermediate. The primary difference between the two enzymes rests on the way the linear tetraketide is cyclized. Chalcone synthase catalyze an intramolecular C6 to C1 Claisen condensation resulting in phloracetophenone-type ring, whereas stilbene synthase catalyzes a C2 to C7 aldol intramolecular cyclization resulting in orsellinic acid-type rings (Figure 4). The enzyme reaction is usually accompanied by the decarboxylation of orsellinic ring, yielding a resorcinol type ring.

Figure 4. Phloracetophenone and orsellininic type rings resulting from Claisen or aldol condensation typically catalyzed by chalcone and stilbene synthases, respectively.

Mechanistically, the rice PKS involved in the production of 5-alkyl resorcinols catalyzes the aldol condensation of the linear tetraketide intermediate to yield orsellinic acid-type rings similar to that catalyzed by stilbene synthase (Figure 5), rather than claisen condensation typically associated with chalcone synthases (Tropf et al. 1995; Rawlings, 1999).

Figure 5. Biosynthesis of lipid resorcinol begins with the formation of the starter unit, in this case, a fatty acid-CoA originating from fatty acid synthesis. The rice polyketide synthase catalyzes the condensation of 3 malonyl-CoA units to the fatty acyl-CoA starter unit followed by the cyclization, reduction and decarboxylation steps, yielding the lipid resorcinol. Subsequently, an arabinosyl glycoside is obtained via the action of glycosyl transferase.

As mentioned previously, lipid resorcinol are biologically active molecules. Rice glycosylate lipid resorcinols to yield arabinosyl derivatives (Kong et al. 2002). Plants often conjugate bioactive molecules to sugars, amino acids or glutathione to reduce potential autotoxicity problems and enhance compartmentalization of the products.

Expression and purification of the PKSs followed standard protocols and provided highly active PKS preparations. At least one of those PKS genes from the rice genome used acyl-CoA as starter unit as substrate. Analysis of its substrate specificity indicates that this enzyme preferred long chain alkyl-CoA starter units (Figure 6). Most polyketide synthase cannot accept substrate with such long and lipophilic chains. The size of this enzyme is within the range common to other type III PKSs.

The enzyme appears to be somewhat promiscuous in accepting starter units with carbon chains ranging between 10 to 16 carbons (Figure 6). However, it prefers longer substrate, which distinguishes this enzyme from other Type III PKS known to date (other than those identified by our laboratory in sorghum). While the total amount of product made was lower with the C16 substrate (palmitoyl-CoA) than the C14 (Myristoyl-CoA), there were no detectable derailment products with this substrate, whereas derailment products are evident with shorter substrates. Interestingly, the enzyme were not able to use oleoyl-CoA (C18:0) as a starter unit, although the primary resorcinol identified in rice is a heptyl-resorcinol. The reason for this remains to be determined. However, substrate availability in the cell and even the cellular compartmentalization may influence the products being made.

Figure 6. Substrate specificity of a rice polyketide synthase showing affinity for the longer fatty acid-CoA starter units. The numbers at the bottom of the TLC plate represent the number of carbon in the acyl-CoA starter unit. p = indicate the lipid resorcinol product of the reaction as shown in figure 5. The other spots represent derailment products.


A relatively large number of the rice cultivars have allelopathic traits that repress the growth of problem weed species in paddy fields. A tremendous effort is underway to produce rice lines with enhanced allelopathy through breeding techniques. Quantitative Trait Loci (QTL) analysis has associated the allelopathic trait with several chromosomes, suggesting that these weed-repressing varieties may produce more than one phytotoxin. Recent publications report that allelopathic rice cultivars produce glycosides of lipid resorcinols, a flavone, and benzoxazinoids, as well as momilactone B and a cyclohexenone. Lipid resorcinols are of particular interest to our research group because these secondary metabolites have been associated with pathogen resistance and allelochemical traits of other monocotyledonous species. Genes putatively involved in the ring formation of these unusual resorcinols in rice have been identified. The substrate specificity of at least one of these rice enzymes indicates that it may be involved in the biosynthetic pathway of lipid resorcinols. The potential for reduction of herbicide use in paddy field is promising with the use of allelopathic rice. Cultivars with enhanced allelopathic traits, either via traditional breeding or biotechnological means, should be available in the future. Their use, in conjunction with other allelopathic ground cover or mulches (Hong et al. 2004) may suppress more than 80% of the weeds in paddy field. However, much remains to be done to fully understand the synthesis of rice allelochemicals as well as the factors regulating their production. It has been shown that the synthesis of allelochemicals can be induced in rice plants by the presence of weeds such as barnyardgrass (Kong et al. 2004b), or by the application of phytotoxins (Tamogami et al. 1997) or jasmonic acid (Koga et al. 1998). The environmental and hormonal regulation associated with the synthesis of inducible allelochemicals (as opposed to constitutive) should be considered when altering allelochemical production of a crop.


Bouillant ML, Jacoud C, Zanella I, Favre-Bonvin J, Bally R (1994). Identification of 5-(12-heptadecenyl)-resorcinol in rice root exudates. Phytochemistry 35, 769-771.

Courtois B, Olofsdotter M (1998). Incorporating the allelopathy trait in upland rice breeding programs. In ‘Proceedings of the Workshop on Allelopathy in Rice’ (Eds M Olofsdotter) pp. 57-68, Manila, Philippines.

Dayan FE, Kagan IA, Rimando AM (2003). Elucidation of the biosynthetic pathway of the allelochemical sorgoleone using retrobiosynthetic NMR analysis. Journal of Biological Chemistry 278, 28607-28611.

Dilday RH, Mattice JD, Moldenhauer KA, Yan W (2001). Allelopathic potential in rice germplasm against ducksalad, redstem and barnyardgrass. Journal of Crop Production 4, 287-301.

Goff SA, Ricke D, Lan TH, Presting G, Wang R, Dunn M, Glazebrook J, Sessions A, Oeller P, Varma H et al. (2002). A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science 296, 92-100.

Gealy DR, Wailes EJ, Estorninos LE Jr, Chavez RSC (2003). Rice cultivar differences in suppression of barnyardgrass (Echinochloa crus-galli) and economics of reduced propanil rates. Weed Science 51, 601-609.

Hong NH, Xuan TD, Eiji T, Khanh TD (2004). Paddy weed control by higher plants from Southeast Asia. Crop Protection 23, 255-261.

Jensen LB, Courtois B, Shen L, Li Z, Olofsdotter M, Mauleon RP (2001). Locating genes controlling allelopathic effects against barnyardgrass in upland rice. Agronomy Journal 93, 21-26

Kato-Noguchi H (2004). Allelopathic substances in rice root exudates: Rediscovery of momilactone B as an allelochemical. Journal of Plant Physiology 161, 271-276.

Koga J, Yamauchi T, Shimura M, Ogawa N, Oshima K, Umemura K, Kikuchi M, Ogasawara N (1998). Cerebrosides A and C, sphingolipid elicitors of hypersensitive cell death and phytoalexin accumulation in rice plants. Journal of Biololgical Chemistry 273, 31985-31991.

Kong C, Xu X, Hu F, Chen X, Ling B, Tan Z (2002). Using specific secondary metabolites as markers to evaluate allelopathic potentials of rice varieties and individual plants. Chinese Science Bulletin 47, 839-43.

Kong C, Liang W, Xu X, Hu F, Jiang Y (2004a). Release and activity of allelochemicals from allelopathic rice seedlings. Journal of Agricultural and Food Chemistry 19, 2861-2865.

Kong C, Xu X, Zhou B, Hu F, Zhang C, Zhang M (2004b). Two compounds from allelopathic rice accession and their inhibitory activity on weeds and fungal pathogens. Phytochemistry 65, 1123-1128.

Kozubek A, Tyman JHP (1999). Resorcinolic lipids, the natural non-isoprenoid phenolic amphiphiles and their biological activity. Chemical Reviews 99, 1-25.

Olofsdotter M (2001). Rice - A step toward use of allelopathy. Agronomy Journal 93, 3-8.

Olofsdotter M, Jensen LB, Courtois B (2002). Improving crop competitive ability using allelopathy - an example from rice. Plant Breeding 121, 1-9.

Rawlings BJ (1999). Biosynthesis of polyketide (other than actinomycete macrolides). Natural Product Report 16, 425-484.

Tamogami S, Rakwal R, Kodama O (1997). Phytoalexin production elicited by exogenously applied jasmonic acid in rice leaves (Oryza sativa L.) is under the control of cytokinins and ascorbic acid. FEBS Letters 412, 61-64.

Tropf S, Karcher B, Schroder G, Schroder J (1995). Reaction mechanisms of homodimeric plant polyketide synthases (stilbene and chalcone synthase) Journal of Biological Chemistry 270, 7922-7928.

Tropf S, Lanz T, Rensing SA, Schrder J and Schrder G (1998) Evidence that stilbene synthases have developed from chalcone synthases several times in the course of evolution. Journal of Molecular Evolution 38, 610-618.

VanEtten HD, Mansfield, JW, Bailey, JA, Farmer EE (1994). Two classes of plant antibiotics: phytoalexins versus ‘phytoanticipins’. Plant Cell 6, 1191-1192.

Xu M, Hillwig ML, Prisic S, Coates RM, Peters RJ (2004). Functional identification of rice syn-copalyl diphosphosphate synthase and its role in initiating biosynthesis of diterpenoid phytoalexin/allelopathic products. Plant Journal 39, 309-18.

Yu J, Hu S, Wang J, Wong GK, Li S, Liu B, Deng Y, Dai L, Zhou Y, Zhang X, et al. (2002). A draft sequence of the rice genome (Oryza sativa L. ssp. indica). Science 296, 79-92.

Previous PageTop Of PageNext Page