hment-induced senescence, a 6-fold increase was evident following ABA-promoted senescence. Moreover, retardation of ABA-promoted senescence reduced PA levels. These results indicated that changes in PA levels were associated Suppression of PLDd Retards Leaf Senescence tightly with ABA-promoted senescence. Whereas levels of PA increased by 0.5 nmol/mg in WS plants, they increased by only 0.1 nmol/mg in PLDd-KO plants. This suggested that at concentrations of approximately 0.4 nmol/mg, the formation of PA is suppressed by approximately 85% in plants in which PLDd has been ablated. These results indicated that the additional PA in WS plants was derived mainly from PLDd-mediated hydrolysis. Therefore, limited availability of PA might account for the retardation of ABA-promoted senescence in PLDd-KO plants. Both PLDa1 and PLDd are key members of 
the PLD family in plants. Although 16824511 the symptoms of delayed leaf senescence from plants with reduced levels of PLDa1 was reported,if and how their lipids and PA in particular changed was not examined. The retardation of ABA-promoted senescence under both n-butanol and NAE treatments implied that PLDa1-mediated PA was involved in the process. We used TLC to measure PA levels in leaves from Col and PLDa1-antisense Arabidopsis plants and found that levels of PA were lower in leaves of PLDa1AS plants than in leaves from Col plants after ABA treatment. These lines of evidence indicated that PLDa1mediated PA plays important roles in ABA-promoted senescence and suggested that both PLDa1 and PLDd might function in ABA-promoted senescence through the same mechanism, which involves the regulation of PA formation. The results also demonstrated that limiting the formation of PA retards ABApromoted leaf senescence. Changes in PC Levels during ABA-promoted Senescence Given that PLD may use different phospholipids as substrates, we examined changes in phospholipids to identify potential substrates of PLDd during ABA-promoted senescence. After treatment with ABA for 5 days, the levels of PE and PI did not decrease in WS leaves and were comparable between leaves of WS and PLDd-KO plants. Given our HC-067047 chemical information demonstration that PS was not involved in ABA-promoted senescence, 12182951 we conclude that PE, PI, and PS cannot be the substrates of PLDd. However, the significant decrease in levels of PC in WS leaves and the observation that this decrease was suppressed in leaves from PLDd-KO plants together suggested that PC is hydrolyzed by PLDd to generate PA. Discussion The metabolism of organelle membranes plays a crucial role during leaf senescence. The general involvement of PLD and its product PA in the deterioration of senescing leaves has been reported widely, with PLDa1 having been the first PLD identified to function in ABA-promoted senescence. Here we report that the complementary use of biochemical and genetic approaches demonstrates a positive role for PLDd in ABApromoted leaf senescence through the regulation of lipid degradation. Our data suggest that PLDd uses PC as its preferred substrate to produce PA. Not only was PLDd responsible for most of the increase in levels of PA, but PLDa1 also contributed to the formation of PA during ABA-promoted senescence. We also performed detailed analysis of membrane lipid degradation during leaf senescence. Plastidic and extraplastidic lipids showed different patterns of degradation. The levels of all plastidic lipids decreased, with the most extensive degradation occurring in levels of MGDG