Biosynthesis of germination stimulants of parasitic weeds Striga and Orobanche

My research focused on the biosynthetic origin of germination stimulants of the root parasitic plants, Striga spp. and Orobanche spp., which have an increasing impact on cereal and other economically important crops in many regions of the world. The traditional control methods are not sufficient and therefore efficient and feasible control methods are urgently required. Control methods that target the early steps in the parasitic weed life cycle are of particular interest as they could prevent parasitic weed infection in an early stage. The first chemical signal exchange between the host and the parasitic plant, the secretion by the host of germination stimulants, is obviously interesting as a target for control methods. Recent research on the germination stimulants in our group has shown that the most important class of germination stimulants, the strigolactones, are derived from the carotenoid pathway. To explain this carotenoid origin, carotenoid cleavage enzymes were predicted to be involved in the strigolactone formation in maize. Therefore my research efforts were directed at cloning carotenoid cleavage enzymes from maize. The first and only characterised carotenoid cleavage enzyme in maize so far is VP14 that belongs to a small family of related genes. In this thesis I have shown that knocking out the NCED (VP14) gene family in maize using RNAi technology suppressed germination stimulant formation leading to reduce Striga seed germination (Chapter 3). This result confirms the previous results with maize vp14 that induced 40% lower Striga seed germination and the tomato mutant, notabilis, with a mutation in tomato NCED1 (VP14 ortholog) that also induced about 40% lower germination of O. ramosa. These results indicate that the NCEDs are directly involved in the strigolactone formation in maize or indirectly through the reduced formation of abscisic acid (ABA) through an as yet unknown mechanism. In my thesis work I cloned a second, as yet unknown maize carotenoid cleavage dioxygenase, ZmCCD1 (Chapter 2). I showed that ZmCCD1 is not involved in strigolactone formation but catalyses the formation of “yellow pigment” in mycorrhizal maize roots. Interestingly, mycorrhizal maize roots show a clear reduction in germination stimulant formation (Chapter 2). The possible explanation is that ZmCCD1 competes with other carotenoid cleavage enzymes for a common carotenoid precursor. A higher activity of ZmCCD1 therefore leads to a decreased germination stimulant production in AM root. These results suggest that the reduction of Striga infestation of maize by AM-fungi may be caused by a reduction in the formation of strigolactones. Using model plants may have a lot of advantages to dissect the germination stimulant biosynthetic pathway, because they take advantage of the extensive genetic, biochemical and physiological information and can be genetically engineered more easily and rapidly. However, the first important question of course is if these model plants produce germination stimulants, what kind of germination stimulants and from which pathway. Two model plants were used to study the germination stimulant biosynthesis, the dicotyledonous model plant Arabidopsis and the monocotyledonous model plant rice. Arabidopsis is a host of several Orobanche spp. and rice can be infected by several Striga spp. Interestingly, Arabidopsis is not a host of mycorrhizal fungi, whereas rice is. Arabidopsis is a host of Orobanche aegyptica, Orobanche ramosa and Orobanche minor. The germination stimulants of Arabidopsis have so far not been identified but could be supposed not to be strigolactones because Arabidopsis is not a host of arbuscular mycorrhizal fungi. Nevertheless, my research has shown that orobanchol (a strigolactone) is present in the Arabidopsis root exudates and in a hairy root culture under phosphate starvation (Chapter 4) and this suggests that Arabidopsis is therefore a suitable model for studying the early stage of host-parasite interaction. However, the concentration of orobanchol found is much lower than in other plant species (compare Chapter 5) and, in addition to orobanchol, I also found several root exudates fractions containing unknown, most likely non-strigolactone, compounds in the Arabidopsis exudates, which have germination stimulating activity on O. ramosa (Chapter 4). This could make it difficult to use fast screens to look for mutants. The biosynthetic origin of these other germination stimulants remains unknown with the exception perhaps of one of these that I showed is probably derived from the early plastidic pathway (Chapter 4). The question remains why Arabidopsis produces strigolactones considering it is not a host of arbuscular mycorrhizal fungi. Is it because strigolactones are involved in other symbiotic associations? Also it is possible that the strigolactones have another physiological significance that we are currently not aware of. For the host plants of Striga, strigolactones have been found in the root exudates of cereals in tropical areas such as maize, sorghum, and millet. The cereal model rice could be good candidate for our research. Indeed we were able to detect orobanchol and another strigolactone, possibly an epimer of 5-deoxystrigol, in the root exudates of rice grown under phosphate starvation. The production of these compounds was completely inhibited by fluridone. These results are confirmed by the germination bioassay, in which root exudates from fluridone treated rice induced much lower Striga seed germination than that of non-treated plants grown under phosphate starvation. In addition, in a mycorrhizal branching assay fluridone treated root exudates induced much lower branching of mycorrhizal hyphae than the exudates of non-treated, phosphate-starved, plants. These results indicate that strigolactones are probably the major germination stimulants in rice root exudates. Thus, compared with Arabidopsis, rice has a great advantage to study the biosynthetic pathway of strigolactones. My work has contributed new insight in the chemical regulation of the host-parasitic plant lifecycle and as such can contribute to the development of new control methods. Once the strigolactone biosynthetic pathway has been further elucidated, it would also become feasible to make low-stimulant producing plants through the inactivation of one or more steps in the pathway. For the time being, as suitable targets enzymes of the primary carotenoid pathway could be used but better targets would be the dedicated pathway enzymes, i.e. the postulated enzymes involved in cleavage and further conversion of the cleavage product to the strigolactones. For example, we have designed RNAi constructs to block the carotenoid cleavage dioxygenases (VP14 cluster enzymes) using both root specific and universal promoters and these constructs were transformed to maize and the transgenic plants induced lower germination of Striga seeds than wildtype maize plants (Chapter 3). As an alternative to knocking out enzymes from the germination stimulant pathway, overexpression of key enzymes of competing pathways to channel away substrate can also be considered as a strategy to reduce germination stimulant formation. Possible candidates are the cleavage enzymes that are responsible for apocarotenoid formation upon mycorrhizal colonization, ZmCCD1 (Chapter 2). Also the suppressive effect of AM fungi on Striga infection of sorghum and maize may be developed into a Striga control strategy. The mechanism of this reduction was so far unknown, and therefore the possibilities to optimize and exploit this phenomenon for practical use were limited. However, I have shown that this reduction is - in any case partly - due to a decrease in germination stimulant formation after mycorrhizal colonisation (Chapter 2). A possible explanation is that due to the formation of mycorrhiza specific apocarotenoids or abscisic acid the formation of the Striga germination stimulant is reduced. Apocarotenoid formation may be competing for substrate with germination stimulant formation. Alternatively, mycorrhizal colonisation may be down-regulating the strigolactone production pathway directly. Therefore, research could now be aimed to optimize the use of AM fungi for controlling parasitic plants including Orobanche through reduced germination, for example through the selection of suitable AM fungus – host (variety) combinations. In rice, we found that irrigation but also spraying with a low concentration of fluridone significantly reduced the number of germinated/attached Striga seeds even with very low concentrations (Chapter 5). In the spraying experiment, the concentration of fluridone used was so low that bleaching of the leaves did not occur. This effect of fluridone suggests that the rice germination stimulants are strigolactones and indeed we have confirmed that they are (Chapter 5). These experiments show that herbicides that inhibit carotenoid biosynthesis can be used to significantly reduce the germination of parasitic seeds and that spraying or irrigating plants with such herbicides at one or more time intervals may be an effective and cheap method to reduce parasitic-weed induced yield losses of crop plants. Our research has shown that under increased phosphate levels, plants (maize, rice, Arabidopsis) produce less germination stimulants. Therefore increasing the phosphate level in deficient soils would probably inhibit Striga infection. However, strigolactones may not be the only germination stimulants of Striga and they are also the signal for mycorrhizal symbionts. Inhibition of production and exudation of strigolactones may also negatively effect the colonization by mycorrhizal fungi. Thus the effective control method should depend on the specific situation in the soil.