The mechanism of suppression by the legume Desmodium uncinatum of the parasitic weed Striga

In sub-Saharan savannah agriculture, African farmers have a range of pest problems with which to contend and yet without access to modern pesticides. During the course of work on controlling insect pests, particularly the stem-boring larvae of indigenous and introduced moths, by intercropping with repellent plants and trap cropping with plants that attract the adult moths, an intercrop was found that provided dramatic control of the African witchweed, Striga species. These parasitic weeds in the snapdragon family are a major threat to the staple food crops of over 100 million people in the region. The parasite-controlling intercrops were cattle forage legumes, commonly called desmodium, and when grown on-farm as a one-to-one intercrop between maize and other subsistence cereals such as sorghum, the parasitic Striga was controlled to the extent that a yield of well under 1 tonne/hectare of maize could be raised to over 5 tonnes/hectare. Experimentally, it was demonstrated that the effect could be provided by passing water over the roots of desmodium and then into soil containing the Striga and maize seeds. Furthermore, when desmodium was grown in water with nutrients but without soil, the water captured the activity, which could then be transferred to the Striga and maize seeds in soil, and again conferred control of the parasite. Fractionation of the chemicals from the water effecting this control identified a fraction that affected development, which was recorded simply by measuring the reduction in radicle growth of the parasite. Chemicals from these this fractions accounting for the activities described, have been identified. However, the exact way in which these chemicals work to prevent Striga infestation has not been determined. Thus, the aim of this project is to identify the stage(s) of the Striga lifecycle affected by the inhibitory compounds present in desmodium root exudates, to purify some of these compounds and to use them to test the hypothesis that they are responsible for the suppression of Striga seen in the field. Finally, the root exudates of another legume, that is commonly used in molecular studies, Lotus japonicus, that we know also produces a root exudates that inhibits Striga radicle growth, will be examined to determine whether they contain compounds that suppress Striga infection. Although L. japonicus would not be agronomically suitable for practical control of Striga, the wealth of genomic information and tools available for this model legumes would enable the biosynthetic pathways involved in the synthesis of the novel flavones to be elucidated in the future. This work will provide the first elucidation of how a valuable weed suppression occurs and will provide the basis for attempts to control weeds in UK agriculture.

Desmodium uncinatum, when intercropped with maize and other cereals in soil with a high level of seeds of the parasite Striga hermonthica (Scrophulariaceae), gives yield increases of 5-fold and when repeatedly used, helps to reduce the Striga seed bank. Activity has been demonstrated to be in both root exudate and extracts of D. uncinatum and acts without the need for soil or rhizobial inoculation. Such suppressive activity is not present in the root exudates of other commonly used intercrops such as cowpea, beans, soybean. Bioassay-guided fractionation of D. uncinatum root exudate, from hydroponic culture, or extract has identified, by NMR spectroscopy, lipophilic components conferring high germination of the Striga seed and hydrophilic components that cause a reduction in the rate of elongation of the Striga radicle. The main inhibitory compound identified is an unusual di-C-linked glycoside of a flavonoid. Although Striga radicle elongation is clearly slower in the presence of D. uncinatum root exudate, this is unlikely to be the sole basis of the suppression of Striga seen in the field and, although intercropping with Desmodium species is being adopted as a key Striga control strategy, the mechanistic basis of the phenomenon is unknown. In order to optimise, extend the control strategy in the short term and, in the longer term, to use biotechnological approaches to engineer the production of the Striga suppressive compounds in legume crops, we will identify the stage(s) of the Striga lifecycle affected by the inhibitory compounds present in D. uncinatum root exudates, purify the main inhibitory compounds and test the hypothesis that they are responsible for the suppression of Striga. Finally, the root exudates of a legume that is commonly used in molecular studies, Lotus japonicus, that we know also produces a root exudates that inhibits Striga radicle growth, will be examined to determine whether it produces compounds that suppress Striga infection. (Joint with BB/E015794/1)