Molecular regulation of drought tolerance in rice

Abiotic stresses are the primary cause of crop failure worldwide, reducing average yields by more than 50%. Among the various forms of abiotic stress, drought is the most limiting factor for rice productivity. Drought affects about 20% of the total rice cultivation area in Asia. Understanding the various aspects of drought stress, the response and resistance mechanisms in relation to plant growth is therefore of fundamental importance to improve sustainable agriculture. Drought tolerance is usually controlled by complex gene networks and engineering of a single gene is unlikely to improve this trait. However, altering the expression of Transcription Factors (TFs) may be a tool for improvement of drought tolerance since they have been shown to activate the expression of multiple genes in a coordinated manner and they are therefore attractive and promising targets for application in molecular breeding or genetic engineering. In addition, studies on TFs will improve our understanding of the physiological and molecular mechanisms of drought tolerance. The overall objective of the work presented in this thesis was to get more detailed insight in the molecular regulation of drought tolerance in rice, with a particular focus on the role of TFs of the homeobox class and two groups of plant hormones, abscisic acid and strigolactones. In Chapter 2, I described the isolation and characterisation of the rice <em>Oshox22</em> gene which is an homeobox gene of the HD-Zip I family. I showed that the expression of <em>Oshox22</em> is strongly induced by salt stress, abscisic acid (ABA) and polyethylene glycol (PEG) treatment, and weakly by cold stress. Trans-activation assays in yeast and transient expression analyses in rice protoplasts demonstrated that Oshox22 is able to bind to the CAAT(G/C)ATTG element and acts as a transcriptional activator that requires both the HD and Zip domains. Rice plants homozygous for a T-DNA insertion in the promoter region of <em>Oshox22 </em>showedreduced <em>Oshox22 </em>expression and ABA content, decreased sensitivity to ABA, and enhanced tolerance to drought and salt stress in the seedling stage. In contrast, transgenic rice over-expressing <em>Oshox22</em> showed increased sensitivity to ABA, increased ABA content, and decreased drought and salt tolerances. These results support the conclusion that <em>Oshox22</em> acts as a negative regulator in stress response. Since reporter gene studies in yeast and rice cells suggested that <em>Oshox22</em> acts as a transcriptional activator, its function as a negative regulator in stress responses might be explained via activation of other repressors. As <em>Oshox22</em> is highly expressed in developing panicles and grains, in Chapter 4 I investigated the role of <em>Oshox22</em> in controlling grain length (GL) in rice. We found a stable quantitative trait locus (QTL) for GL on this position in four mapping populations. Sequence analysis of <em>Oshox22</em> in rice cultivars Bala, Azucena and Nipponbare revealed an extra A base in the Azucena promoter, which is a long grain type rice. Using a PCR-based insertion/deletion (InDel) CAPS maker assay in rice populations and collections, I found an association between the A InDel in the <em>Oshox22</em> promoter with GL. Furthermore, expression of <em>Oshox22</em> under the control of a promoterwith the A InDelin Zhonghua 11 (which does not have the A InDel) resulted in a significant increase in GL in Zhongua 11. Scanning electron microscopy revealed that the enhanced GL was caused by an increased cell length in the inner epidermal cells of the lemma. In addition, the data show that there is a tendency for lower expression of <em>Oshox22</em> when GL increases which would suggest that Oshox22 functions as a repressor of GL.These findings suggest that natural variation in the <em>Oshox22 </em>promoter can be exploited in breeding programmes to modify GL using molecular marker-assisted selection. However, the exact mechanism of regulation of GL by <em>Oshox22</em> is still not clear. Since <em>Oshox22</em> is a homeobox gene, it will exert its function via regulation of downstream target genes which we do not know yet. Therefore, more research is needed to elucidate the genetic and biochemical pathways to understand the molecular mechanisms underlying rice GL development and to determine if there are interactions with other known regulators of GL. The strigolactones are a relatively new class of plant hormones and a possible role in drought tolerance is unknown. In Chapter 4 of this thesis, I reviewed the various roles that strigolactones (SLs) play both in the rhizosphere and as endogenous plant hormone. In addition, the current knowledge on the SL biosynthetic and downstream signalling pathways and the interactions of SLs with other plant hormones, such as ABA, is described. It has been reported that there seems to be a functional link between ABA and SLs but the mechanism of that link remained unknown. In Chapter 5, I studied the intimate relationship between ABA and SL biosynthesis through the further characterisation of β-carotene isomerase D27 in rice. The results show that the ABA content was increased in SL-deficient and -insensitive <em>dwarf </em>(<em>d</em>) rice mutants, <em>d10</em>, <em>d17</em> and <em>d3</em> compared with wild type, while it was reduced in <em>d27</em>. In addition, this difference was significantly enhanced by exposure to drought. Interestingly, as a consequence of their enhanced ABA levels, <em>d10</em>, <em>d17 </em>and <em>d3</em> plants displayed an increased tolerance to drought compared with wild-type plants, while the ABA deficient <em>d27</em> plants were more drought sensitive. Transient over-expression of <em>OsD27</em> in <em>Nicotiana benthamiana</em>enhanced both ABA and SL production. However, constitutive over-expression of <em>OsD27</em> in rice plants showed no significant changes in ABA and SL levelsunder normal conditions. Still, <em>OsD27</em> over-expression did result in higher SL levels, compared with wild-type plants, under phosphate starvation. This suggests that likewise, <em>OsD27</em>over-expression may only result in increased ABA levels during drought stress conditions. I concluded that the <em>OsD27</em> gene is involved in SL as well as ABA biosynthesis, and that, depending on the environmental conditions, the expression of the more downstream SL and ABA specific biosynthetic genes determines which of the two and how much is being produced. In Chapter 6, I discussed the main findings of this thesis and presented the future perspective of how the knowledge generated in this thesis can contribute to the improvement of drought tolerance and GL in rice.