Molecular genetic and physiological analysis of Rhizobium-rice interactions

Abstract

Recently, it was found that strains of the soil bacterium Rhizobium leguminosarum biovar trifolii, which normally infect and nodulate clovers, can also associate and colonise different rice culutivars. When rice seedlings were inoculated with these rhizobia, there was a strain-specific response in the growth of the seedlings. This thesis examines the interaction of Rhizobium strains with rice. The following approaches were investigated to increase our understanding of the effects of Rhizobium on the early growth and development of this rice-Rhizobium association by: (a) using of a reporter GFP construct in colonisation studies and following the progress of colonisation of GFP-labelled rhizobia; (b) locating the rhizobia genes involved in the interaction by deleting and curing the plasmids and comparing these changes with the genomic sequence of Sinorhizobium meliloti Sm1021. These studies were further refined by; (c)the use of R-primes, cosmid libraries and Tn5-Mob constructs; (d) the application of exogenous phytohormones and compounds regulating the auxin phytohormone pathway; (e) GC/MS quantification of IAA produced by different bacterial strains; and (f) the use of a plant bioassay to mimic rice growth under paddy field conditions. It was found that legume-associated Rhizobium strains can intimately associate with and enter rice roots within 48 h. However, the rice-Rhizobium association is a complex interaction between the medium, the non-legume host and rhizobia. Under certain conditions, rhizobia could either promote (eg., strain R4), inhibit (eg., strains ANU843 and E4) or have no effect on rice growth. By using a series of plasmid-cured strains of rhizobia, it was found that genes associated with plasmids of Rhizobium strains were involved in the inhibition or the stimulation of rice seedling growth and at least four replicons were needed to cause the inhibition of rice growth. High concentrations of auxin, cytokinin and nitrate mimicked the effect of inhibitory Rhizobium strains on rice root growth and development by forming short lateral roots. Similar observations were made with rice seedlings treated with the sequenced wildtype strain S. meliloti Sm1021 and its closely related strain Rm2011 when they were inoculated onto rice cv. Pelde and cv. Calrose. All pSymA deleted and cured derivatives of strain Rm2011 had no effect on rice growth while pSymB deleted derivatives partially inhibited rice seedling growth. To examine the effect of genes associated with the megaplasmids pSymA and pSymB a series of transconjugant experiments were done. By mobilising the megaplasmids pSymA and pSymB into the non-inhibitory Rhizobium strains of rice growth, it was found that the inhibition of the rice seedlings was associated with the pSymA. Complementation studies with RP4 plasmids containing regions of the pSymA plasmid suggested that genes involved in rice growth inhibition were located in the 450Kb deleted region of pSymA that is deleted from strain SmA146. This region has genes involved in nitrate reduction and IAA biosynthesis. However, Tn5-mutagenesis of the nitrate transport and nitrate reduction genes in megaplasmid pSymA of Rm2011 did not abolish rice growth inhibition. Use of GC/MS demonstrated that IAA was synthesised by all S. meliloti strains, Sm1021 (Rm2011) and pSymA and pSymB deleted and cured derivatives of Rm2011, and rice-associated strains R4 and E4. In addition, it was shown that the addition of agar, the form of nitrogen in the plant medium, and the light could affect the association of Rhizobium with rice seedlings cv. Pelde. This thesis will present evidence from these studies suggesting that IAA and nitrate are not the cause for rice growth inhibition in cv. Pelde. It is proposed that (a) the plasmid-associated inhibition phenomenon in Rhizobium-rice interaction is linked to the pattern of root growth and development; (b) a high level of nitrite in the medium may be the cause of inhibition due to the formation of NO in the plant medium and rice root tissues, and (c) IAA may indirectly be involved in rice-Rhizobium inhibition phenomenon. Therefore, the association of Rhizobium with rice seedlings i.e., the colonization of rice roots by rhizobia and the strain-specific response in the growth of the seedlings are controlled by plasmid-associated genes. Such plasmid-associated genes, under laboratory conditions affected the growth of the rice seedlings and under field conditions may be an important factor in stimulating or inhibiting rice growth and subsequently rice yields.