Mono- and digalactosyldiacylglycerol (MGDG and DGDG) are the most abundant lipids of photosynthetic membranes in chloroplasts of algae and plant cells. These galactolipids constitute therefore the most profuse lipid class on earth. Due to their high levels in food of plant origin, they are a primary source of galactose and fatty acids in human diet. Galactolipids have long been considered to be strictly localized in plastids (a plant specific organelle), where they constitute 80% of membrane lipids, in contrast with the endomembrane system and mitochondria which are phospholipid-rich. However, DGDG can relocate outside plastids to replace phospholipids in response to phosphate (Pi) deficiency. Galactolipid and phospholipid metabolisms are thus coupled, although the respective biosynthetic pathways are in distinct organelles, i.e. the plastid envelope for galactolipids and endoplasmic reticulum for phospholipids. The question addressed here is to understand how MGDG synthesis can be controlled by a molecular tuning of MGD enzymes, in relation with the phospholipid metabolism. The proposed project falls in the scope of the SVSE5 axis of the call (enzymology, structural biology, membrane systems, biomimetic systems). In Arabidopsis, a multigenic family of MGDG synthases (MGD1, 2, 3) can catalyze the galactosylation of diacylglycerol (DAG). MGD1 is the most abundant, localized in the inner envelope membrane of chloroplasts and is essential for the expansion of thylakoids. Expression of MGD2 and 3 increases in response to Pi shortage, indicating that the lipid-remodelling observed in this condition implies MGD2 and 3 isoforms in a genetic regulation. Partners 1 and 2 have recently examined the possible role of phosphatidic acid (PA) in a metabolic regulation since PA is known to act as a signalling molecule, is a precursor of phospholipids and a product of phospholipid breakdown by phospholipases D. PA was shown to be a strong and specific activator of MGD1 and could be a key regulator in the galactolipid vs phospholipid dialogue. Possible mechanisms for tuning MGD1 activity could imply the binding of effectors, conformational changes, modulation of MGD1 association to membranes, etc. No microsystem is currently available to allow a correlation between MGD1 activity, its association to lipids and its activation or inhibition. The ReGal project aims at dissecting the molecular mechanisms of MGD1 regulation based (1) on structural analyses of MGD1 interaction with effectors, (2) on mechanistic studies of MGD1 activation and inhibition in a biomimetic membrane and (3) on the monitoring of MGD1 substrate (DAG) and activator (PA) in the chloroplast of Arabidopsis cells in relevant physiological contexts and genetic backgrounds. Two models of effectors have been selected. Model of activator is PA. Model of inhibitor is a synthetic molecule named galvestine, obtained by Partner 1 following a high throughput screening and chemical optimization program. Galvestine competes with the binding of DAG to MGD1, 2 and 3 in vitro, and allows a dose-dependent control of MGDG level in planta. Partner 2 has improved the purification of MGD1 in such manner that it is now feasible to attempt crystallographic resolution and functional analyses in reconstituted membranes. Membrane systems will be reconstituted by assembly of lipids including MGDG, DGDG, and lipid effectors like PA, forming Langmuir monolayers, a technology mastered by Partner 3. Living cell dynamics of DAG and PA in chloroplasts, monitored by fluorescent indicators, should allow to correlate information gained from structural and functional studies with cellular measurements in various physiological and environmental contexts. Analyses in Arabidopsis mutated at the level of PLDz and/or NPC phospholipases will be carried out to integrate the deduced metabolic regulation model with knowledge on the genetic regulation, in the context of Pi deficiency.