Etude Des Sources De Carbone Et D'énergie Pour La Synthèse Des Lipides De Stockage Chez La Microalgue Verte Modèle Chlamydo

Home , Antoniewicz, Glyoxylate oxidase, Photosystem, Photosystem II

Aix Marseille Université

L'Ecole Doctorale 62 « Sciences de la Vie et de la Santé »

Etude des sources de carbone et d’énergie pour la synthèse des lipides de stockage chez la microalgue verte modèle Chlamydomonas reinhardtii

Yuanxue LIANG

Soutenue publiquement le 17 janvier 2019

pour obtenir le grade de « Docteur en biologie »

Jury

Professor Claire REMACLE, Université de Liège (Rapporteuse)

Dr. David DAUVILLEE, CNRS Lille (Rapporteur)

Professor Stefano CAFFARRI, Aix Marseille Université (Examinateur)

Dr. Gilles PELTIER, CEA Cadarache (Invité)

Dr. Yonghua LI-BEISSON, CEA Cadarache (Directeur de thèse)

ACKNOWLEDGEMENTS

First and foremost, I would like to express my sincere gratitude to my advisor Dr. Yonghua Li-Beisson for the continuous support during my PhD study and also gave me much help in daily life, for her patience, motivation and immense knowledge. I could not have imagined having a better mentor. I’m also thankful for the opportunity she gave me to conduct my PhD research in an excellent laboratory and in the HelioBiotec platform. I would also like to thank another three important scientists: Dr. Gilles Peltier (co- supervisor), Dr. Fred Beisson and Dr. Pierre Richaud who helped me in various aspects of the project. I’m not only thankful for their insightful comments, suggestion, help and encouragement, but also for the hard question which incented me to widen my research from various perspectives. I would also like to thank collaboration from Fantao, Emmannuelle, Yariv, Saleh, and Alisdair. Fantao taught me how to cultivate and work with Chlamydomonas. Emmannuelle performed bioinformatic analyses. Yariv, Saleh and Alisdair from Potsdam for amino acid analysis. My sincere thanks also go to Ismael, Stéphan, Adrien, Bertrand, Stephanie, Pascaline, Audrey, Solène, Hélène and Marie-Christine. Ismael helped me in many aspects, including solving daily problems, improving my oral English, and also giving me guidance in my experiments. Stéphan and Stephanie solved many technical or machine problems in the lab, in addition, they also taught me how to do Western blot and lipid extraction/TLC analyses. Adrien and Solène taught me how to measure mitochondrial respiration and photosynthetic performances. Bertrand helped me with the LC-MS analyses and also taught me how to run GC-MS. Pascaline taught me how to do qRT-PCR experiment. Audrey, Hélène and Marie-Christine taught me how to use Confocal microscopy. I would especially like to thank Véronique who did a lot of behind-the-scene’s work, such as keeping the algal strains healthy, cleaning and preparing the equipment for experiment. Thanks also go to Marie and Cyril for being friendly in the lab.

Thanks again to everyone in the “Laboratoire de bioénergétique et biotechnologie des bactéries et microalgues” for creating an excellent working environment, thank you all for being so friendly and kind to me.

Last but not the least, I would like to thank my family: my parents and my wife gave me moral support and care for my life.

OUTLINE “Avant-props” ABSTRACT ABBREVIATONS INTRODUCTION 1. Microalgae and microalgal biotechnology 1.1 The importance of biodiesel 1.2 Microalgal biotechnology 1.3 Conversion of algal lipid or biomass to biodiesel 2. Lipid definition and classes 2.1 Lipid definition 2.2 Lipid classes 2.3 Fatty acid (FA) and glycerolipids 3. Lipid metabolism 3.1 Carbon and energy supply for FA synthesis 3.2 De novo FA synthesis and export 3.3 Polar membrane lipid synthesis 3.4 Triacylglycerol (TAG) biosynthesis 3.5 Lipid droplets (LDs) 3.6 Lipid catabolism and the β-oxidation of FAs 4. Branched-chain amino acid (BCAA) metabolism 4.1 Occurrence and cellular functions 4.2 BCAA biosynthesis 4.3 BCAA degradation 5. A crosstalk between metabolism of lipids and BCAAs 6. Chlamydomonas reinhardtii OBJECTIVES OF THIS PHD THESIS MATERIALS AND METHODS 1. Strains and culture conditions

2. The use of Flow cytometry to evaluate cellular oil content 3. DNA and RNA isolation and cDNA preparation 4. Identification of the Aph8 insertion site in the Pb8C12 mutant by RESDA-PCR 5. Reverse transcription PCR (RT-PCR) 6. Cloning the full length gene BKDE1α and genetic complementation of Pb8C12 (i.e. bkdE1α) mutant 7. LD imaging 8. Protein extraction, quantification and SDS-PAGE 9. Antibody generation and immunoblot analysis 10. Lipid analyses 11. Chlorophyll and starch quantification 12. Isolation and validation of mutants from the Chlamydomonas library (CLIP) 13. Amino acid analyses by gas chromatography-mass spectrometry (GC-MS) 14. Respiration analysis using a Clark electrode 15. Total RNA extraction, RNA-seq library preparation and sequencing 16. Enzymatic activity assays RESULTS AND DISCUSSION Chapter 1: Branched-chain amino acid catabolism impacts triacylglycerol homeostasis in Chlamydomonas reinhardtii. Chapter 2: Chlamydomonas carries out fatty acid β-oxidation in ancestral peroxisomes using a bona fide acyl-CoA oxidase. Chapter 3: Interorganelle communication: peroxisomal MALATE DEHYDROGENASE 2 connects lipid catabolism to photosynthesis through redox coupling in Chlamydomonas. Chapter 4: Saturating light induces sustained accumulation of oil primarily stored in lipid droplets of plastidial origin in Chlamydomonas reinhardtii. CONCLUSION AND PERSPECTIVES REFERENCE Annex: CURRICULUM VITAE

“Avant-propos”

I was awarded a PhD studentship by the China Scholarship Council (CSC). I joined the “Laboratoire de bioénergétique et biotechnologie des bactéries et microalgues (LB3M)” at CEA Cadarache the 1st of March 2015. I finished my contract at the end of Febrary 2018. My research interests are to understand lipid biosynthesis and turnover processes in microalgae and investigate the possible occurrence of metabolic links between lipids and other subcellular metabolisms. To this end, I took advantage of the recent isolation of a few mutants defected in oil turnover in the model microalga Chlamydomonas reinhardtii in our laboratory (Cagnon et al., 2013). In the past three years, I have been involved in molecular genetic, biochemical and physiological characterization of three mutants, and contributed to one other work in the laboratory (detailed below).

1. A mutant defected in a gene encoding a putative E1α subunit of the branched chain -keto acid dehydrogenase (BCKDH). [Liang et al submitted to Plant Physiology] This is my major work. The mutant bkdE1α, deficient in the E1α subunit of the BCKDH complex, is found to accumulate less oil during nitrogen (N) starvation, and also is compromised in oil breakdown upon N resupply. We showed via quantification of the amount of branched chain amino acids (BCAAs) that BCKDH is indeed involved in BCAA degradation in Chlamydomonas reinhardtii. Through biochemical, genetic and physiological characterization of this mutant, we put on evidence that BCAA degradation contributes to TAG metabolism via the provision of carbon precursors and ATP, thus highlighting the complex interplay between distinct subcellular metabolisms for oil storage in green microalgae.

2. A mutant defected in an acyl-CoA oxidase (ACX) catalyzing the first step in the

-oxidation of fatty acid in peroxisomes. [Kong, Liang et al Plant J 2017 90:358] This work helped to bring peroxisome to a center stage on lipid homeostasis in a primitive alga, i.e. the Chlamydomonas reinhardtii. We showed for the first time that

H2O2 generating reaction occurs in algal peroxisomes and that Chlamydomonas carries out β-oxidation of fatty acids mainly in this subcellular organelle. Specifically, I: - Cloned the 11 kb genomic DNA encoding the ACX2 protein for genetic complementation of the acx2 mutant. - Did lipid extractions and fatty acid composition analysis. - Carried out quantitative RT-PCR analysis of the expression level of the gene under various cultivation conditions. - Tested the effect of perturbation in fatty acid turnover on senescence and in prolonged darkness.

3. Contributed to characterize the peroxisomal malate dehydrogenase 2 (MDH2) mutant. [Kong, Burlacot, Liang et al Plant Cell 2018 30:1824] This paper showed for the first time, the occurrence of a redox communication between peroxisome and chloroplast, and we showed further that this communication is not only important for carbon reserve accumulation but also is critical for cells to avoid photooxidative damage during changing environmental conditions such as nutrient deficiency or high light (HL). Specifically, I: - did RT-PCR. - performed MDH activity assays. - tested growth parameters under various cultivation conditions. - carried out lipid and starch analysis during a day/night cycle.

In addition:  Microscopy characterization of subcellular localization of lipid droplets in cells starved for N in comparison to cells exposed to HL. [Goold, Cuine, Legeret, Liang et al Plant Physiology 2016 171:2406] This work reported the synthesis of TAGs upon HL exposure, and HL can induce an overall higher lipid productivity than the most commonly used N starvation. Furthermore, by studying and comparing the lipidome and proteome of lipid droplets (LDs) isolated from HL-exposed cells versus N-starved cells, we infer a possibility that distinct sub-populations of LDs or different biogenesis origins of LDs could accumulate in HL-exposed cells compared to N-starved cells. This is further supported by microscopy observations.

RESUME

Etude des sources de carbone et d’énergie pour la synthèse des lipides de stockage chez la microalgue verte modèle Chlamydomonas reinhardtii

Les triacylglycérols d'algues (TAG) représentent une source prometteuse de biocarburants. Les principales étapes de la synthèse des acides gras et du métabolisme du TAG des algues ont été déduites de celles des plantes terrestres, mais on en sait peu sur les sources de carbones et d’énergie intervenant dans la synthèse de lipides de réserve. Pour répondre à cette question, nous avons étudié la synthèse des acides gras chez l’algue modèle Chlamydomonas reinhardtii en utilisant une combinaison d'approches génétiques, biochimiques et microscopiques. Plus précisément, j'ai d'abord examiné la localisation subcellulaire de gouttelettes de lipides dans des cellules d'algues exposées à une forte lumière, conditions où une plus grande quantité de pouvoir réducteur est produite. J'ai ensuite contribué à mettre en évidence que la bêta-oxydation des acides gras est un processus peroxysomal, et que pendant une carence en azote réalisée en conditions photoautotrophe, des mutants dépourvus de la malate déshydrogénase 2 peroxysomale (mdh2) accumulent 50% plus TAG que les souches parentales. Ces résultats nous ont permis de mettre en évidence l'importance du contexte redox cellulaire sur la synthèse lipidique. Cette étude a également permis de révéler l’existence d'un échange d’énergie entre le peroxysome et le chloroplaste. Enfin, en caractérisant des mutants déficients dans la dégradation des acides aminés à chaîne ramifiée (BCAA), j'ai montré que le catabolisme des BCAAs joue un double rôle dans la synthèse de TAG en fournissant des précurseurs carbonés et de l'ATP. L'ensemble de ces travaux a mis en évidence l'existence d’interactions complexes entre le métabolisme du carbone et le métabolisme énergétique dans les cellules photosynthétiques, et ouvert de nouvelles pistes pour l'amélioration génétique future de souches d'algues pour la production de biocarburants.

ABSTRACT

Study of carbon and energy sources for storage lipid synthesis in model green microalga Chlamydomonas reinhardtii

Algal triacylglycerols (TAG) represent a promising source for biofuel. The major steps for fatty acid synthesis and TAG metabolism have been deduced based on that of land plants, but little is known about carbon and energy sources. To address this question, we investigated fatty acid synthesis in algal cells using a combination of genetic, biochemical and microscopic approaches in the model microalga Chlamydomonas reinhardtii. Specifically, I first examined the subcellular localization of lipid droplets in algal cells exposed to high light, a condition favoring production of reducing power. Secondly, I contributed to put on evidence that the β-oxidation of fatty acids is a peroxisomal process, and that during photoautotrophic nitrogen starvation, knock-out mutants of the peroxisomal malate dehydrogenase 2 (mdh2) made 50% more TAG than parental strains, highlighting the importance of cellular redox context on lipid synthesis. This study also revealed for the first time the occurrence of an energy trafficking pathway from peroxisome to chloroplast. And finally, by characterizing mutants defected in degradation of branched-chain amino acids (BCAAs), we showed that BCAA catabolism plays a dual role in TAG synthesis via providing carbon precursors and ATP. Taken together, this work highlighted the complex interplay between carbon and energy metabolism in green photosynthetic cells, and pointed future directions for genetic improvement of algal strains for biofuel productions.

ABBREVIATIONS

ACCase, acetyl-CoA carboxylase; ACX, acyl-CoA oxidase; ACP, acyl carrier protein; BCAA, branched chain amino acid; BCCP, biotin carboxyl carrier protein; BCKDH, branched chain α-keto acid dehydrogenase; BKDE1α, the α-subunit of BCKDH complex; CoA, coenzyme A; CTS1, comatose 1; DAG, diacylglycerol; DGAT, diacylglycerol acyltransferase; DGDG, digalactosyldiacylglycerol; DHA, docosahexaenoic acid; DGTS, diacylglyceryl-3-O-4'-(N,N,N-trimethyl)-homoserine; ER, enoyl-ACP reductase; FA, fatty acid; FAD, fatty acid desaturase; FAT, fatty acyl- ACP thioesterase; FAX1, fatty acid export1; KAS, 3-ketoacyl-ACP synthase; G3P, glycerol-3-phosphate; G6PDH, glucose-6-phosphate dehydrogenase; GC-MS, gas chromatography-mass spectrometry; GPAT, glycerol-3-phosphate acyltransferase; HL, high light; LACS, long chain acyl-CoA synthetase; Lyso-PA, lysophosphatidic acid; LPAAT, lysophosphatidic acid acyltransferase; MCMT, malonyl-CoA:ACP malonyltransferase; MDH, malate dehydrogenase; ME, malic enzyme; MFP, multifunctional protein; MGDG, monogalactosyldiacylglycerol; N, nitrogen; OAA, oxaloacetate; PA, phosphatidic acid; PAP, phosphatidic acid phosphatase; PtdCho, phosphatidylcholine; PDAT, phospholipid:diacylglycerol acyltransferase; PDH, pyruvate dehydrogenase; PtdCho, phosphatidylcholine; PtdEtn, phosphatidylethanolamine; PtdIns, phosphatidylinositol; PtdGro, phosphatidylglycerol; PL, phospholipid; PPP, pentose phosphate pathway; PUFA, polyunsaturated fatty acid; SQDG, sulfoquinovosyldiacylglycerol; TAG, triacylglycerol; TCA, tricarboxylic acid

INTRODUCTION

1. Microalgae and microalgal biotechnology 1.1 The importance of biodiesel Considering the negative impact of fossil fuel consumption, and the finite underground reserve, the production of sustainable and renewable clean energies is a pressing matter, especially for those countries lacking conventional fuel resources. Thus, the term ‘biodiesel’ has appeared, which refers to diesel made from biomass, mostly from vegetable oils. There are many advantages associated to the biodiesel which made it an attractive energy resource. First, biodiesel is a renewable and sustainable energy resource, since they can be produced by many organisms. Second, biodiesel is carbon neutral, because plants or algae fixe CO2 from the atomosphere, therefore resulting in no net release of CO2. Third, biodiesel contains very low sulfur content, because biodiesel contains no or very little harmful aromatic compounds compared to fossil fuel (Antolin et al., 2002). Fourth, the number of carbon in a traditional diesel molecule is about 15, which is similar to that of plant and microalgal oil with 16-18 carbons (Huang et al., 2010). Finally, biodiesel is better than diesel in terms of flash point and biodegradability (Ma and Hanna, 1999).

Microalgae are placed with high expectations for a number of reasons, i.e. their short generation time, high lipid content (55% dry weight) and ease of being modified by biotechnological means, high photosynthetic efficiency, high biomass productivity compared to other energy crops (Hu et al., 2008; Wijffels and Barbosa, 2010). In addition, the fatty acid (FA) composition of algal oils is similar to vegetable oils (Huang et al., 2010). Their massive accumulation is possible in algal cells, but this is often observed with an offset in cell proliferation and biomass growth. This is partly because these lipophilic materials are not only useful storage materials under environmental stress, but also are essential part of cell membranes and precursors to lipid-derived signalling molecules, thus a redirect in their metabolism can, in most cases, alter

12 photosynthetic function, cell physiology and eventually cell fitness and survival. Understanding this dogma is a key to solve the economic issue of algal biomass production for fuel- or chemical- based applications.

1.2 Microalgal biotechnology Microalgae have been mass cultivated especially in Asia and Africa for centuries, and mostly as a food source. They provide rich sources of proteins, carbohydrates and lipids. From a biotechnology point of view, lipids are not only sources of biodiesel, but also are precursors to synthesis of many chemical materials for oleochemical industries. Synthesis and metabolism of lipids by microalgae have therefore been subjected to intensive investigations in the past 10 years (Merchant et al., 2012; Liu and Benning, 2013). Starch can be used to make bioethanol, therefore starch metabolism has also been studied in microalgae (Ball, 2002; Tunçay et al., 2013). Recent research highlighted complicated interactions between the synthesis of starch and that of lipids, especially during the transition to nitrogen (N) depletion (Li et al., 2010; Work et al., 2010; Siaut et al., 2011; Park et al., 2015). In order to get high-yield strains for lipids, currently, the genetic engineering of microalgae has been attempted to increase lipid productivity. For example, the overexpression of the first and rate-limiting enzyme of de novo fatty acid (FA) synthesis, i.e. acetyl-CoA carboxylase (ACCase), increased lipid content by 60% in laboratory conditions and above 40% in outdoor cultivation (Huang et al., 2010).

Microalgae can grow under phototrophic condition by fixing CO2 from the air and use only light energy to produce polysaccharides, proteins, lipids and hydrocarbons. In phototrophic culture, two systems can be used to cultivate algae, open ponds or closed photobioreactors. For production of bulk chemicals, open ponds are less expensive, therefore considered a future for bioenergy production (Huang et al., 2010). Under photoautotrophic condition, microalgae can use light and CO2, but have a relatively low cell density, and moreover, photoinhibition can occur under excessive light. Some

13 microalgae can also grow heterotrophically or mixotrophically. Under heterotrophic cultivation, a better control of the cultivation process can be reached, with relatively lower cost for biomass harvesting because high cell density cultures are often reached (Chen and Johns, 1991). Reduced carbon sources include ethanol, glycerol, and fructose depending on the microalgal species used (Yokochi et al., 1998). The drawback is that some of these carbon sources can be expensive, therefore adding to the total cost of production. There are potential advantages and disadvantages associated to each cultivation system, and the choice of the system or trophic style should be determined when all factors considered (the type and price of the product, the algal strains in question, etc).

1.3 Conversion of algal lipid or biomass to biodiesel The viscosities of microalgal oils are usually higher than that of traditional diesel (Fuls et al., 1984), therefore they can’t be applied to engines directly. For commericial use of algal oil for biodiesel, there are two major ways for their conversion to biofuel. First, oils need to be extracted and then converted to their fatty acid methyl esters (FAMEs). The transesterification step can reduce the viscosity and increase the fluidity of microalgal oils. Chemical catalysis and enzymatic catalysis are two major methods used for transesterification reactions. The advantages of chemical catalysis are: the reaction condition can be well controlled, the cost of the production process is cheap, the methanol can be recycled, with this method, large-scale production with high conversion efficiency is possible, but this process requires a lot of energy and the following disposal process is complex. For enzymatic catalysis, advantages are that only small amount of methanol is needed, less pollution to the environment, but also are performed under moderate reaction conditions; On the downside, enzymes used are often expensive and some chemicals required in the process can damage the enzyme therefore reducing the efficiency of conversion.

Another method, bypassing the lipid extraction process, is the conversion of algal

14 biomass directly to bio-crude via pyrolysis or liquefaction. Pyrolysis can be divided into “fast pyrolysis” or “slow pyrolysis”. The “fast pyrolysis” produces bio-fuel in the absence of air at atmospheric pressure, with a relatively low temperature (450-550°C) as well as short gas residence time to crack biomass into short chain molecules and be cooled to liquid rapidly (Bridgwater et al., 1999). And the products of fast pyrolysis are oil and gases with a yield of about 70%, whereas, the main products of “slow pyrolysis” are char-oils with 15-20% yield (Huang et al., 2010). Compared with “slow pyrolysis”, “fast pyrolysis” produces more bio-oils, with time-saving and energy saving properties. Another method for biomass conversion is called “liquefaction”, where high cellular water levels are not a problem, and rather it can be used for hydrogenolysis (Minowa et al., 1995). The advantage of the above methods for the prodution of “bio-crude” is that there is no need to remove water from microalgal biomass, which is often the most energy demanding step.

2. Lipid definition and lipid types 2.1 Lipid definition Lipids, together with carbohydrates and proteins, are the main constituents of all living cells. The term lipid refers to highly hydrophobic or amphipathic molecules that have a common physico-chemical property: they are soluble in organic solvents such as chloroform but poorly or not at all soluble in water. This definition encompasses molecules belonging to very diverse structural groups, including FAs, glycerolipids, sphingolipids, sterols, wax esters, triterpenes, hydrocarbons and also chlorophylls (Dennis, 2009). Strictly speaking, lipids are FAs and their derivatives, and substances related biosynthetically or functionally to these compounds.

2.2 Lipid classes By different methods, lipids can be divided into different categories. Depending on whether the lipids contain hydrolyzable functional groups, they can be classified into hydrolyzable or non-hydrolyzable lipids. Example, neutral fats, waxes, phospholipids

15 and glycolipids contain ester groups which are hydrolyzable in water, but steroids and fat-soluble vitamins (e.g. A, D, E and K) lack such functional groups. According to the biochemical structure of lipids, biological lipids including two distinct types of ‘building-blocks’ or biochemical subunits: ketoacyl and isoprene groups (Fahy et al., 2009). Using this approach, lipids can be divided into eight categories (Figure 1): FAs, glycerolipids, glycerophospholipids, sphingolipids, saccharolipids and polyketides (derived from condensation of ketoacyl subunits); and sterol lipids and prenol lipids (derived from condensation of isoprene subunits) (Fahy et al., 2005; Fahy et al., 2009).

Figure 1. Representative structures for each lipid category. Stuctures were taken from that of (Fahy et al., 2009).

Glycerolipids are one of the most common lipid classes, and they include not only storage lipid i.e. triacylglycerol (TAG), but also phospholipids and galactolipids which are major consitituents of cell membranes. Sphingolipids are a complicated family of compounds with a common structure, i.e. a sphingoid base backbone, ceramides (N- acyl-sphingoid bases) with an amide-linked FA, and a non-acyl headgroup. Saccharolipids refer to those FAs directly linked to a sugar backbone, instead of a glycerol. Polyketides have a wide variety of structural forms, most of which are cyclic molecules whose backbone is typically further modified by methylation, oxidation, glycosylation or other processes. Sterols are also important components of membrane lipids, all derived from the same fused four-ring core structure but have different biological roles for signal transduction or as hormones. Prenol lipids are synthesized from the 5-carbon precursors isopentenyl diphosphate and dimethylallyl diphosphate that are produced mainly by the mevalonic acid (MVA) pathway (Kuzuyama and Seto, 2003).

2.3 FA and glycerolipid classes FA is a carboxylic acid with a long saturated or unsaturated aliphatic chain. At least several hundreds of FAs have been reported in nature. FAs differ by chain length (number of carbon atoms), the number and position of unsaturation (double bond: C=C) and presence of substituent such as hydroxyl or epoxy groups. A simple shorthand notation based on these parameters has been developed to designate FAs. For example, oleic acid is written as C18:1 (9c) (Figure 2).

Figure 2. An example of shorthand nomination for a fatty acid. Abbreviations: C, for carbon; c for cis- configuration.

By the length of carbon chains, it can be categorized to short-chain FAs (FAs with aliphatic tails ≤5 carbons), medium-chain FAs (6-12 carbons), long-chain FAs (13-20 carbons) or very long chain FAs (≥22 carbons). One given species often contains specific types of FA, for example, in cells of Chlamydomonas reinhardtii, only long- chain FAs are present (Siaut et al., 2011). According to whether it contains carbon- carbon double bonds (C=C), it’s known as saturated FAs (without C=C) or unsaturated FAs (with C=C). Saturation level is also species- tissue- and membrane type- dependent, and also fluctuate to developmental stage and environmental conditions the organism is subjected to. For example, C. reinhardtii contains more than 60% of its total FAs as unsaturated ones during optimal growth, and this level decreases in response to N starvation (Siaut et al., 2011), or after being subjected to high light (HL) (Goold et al., 2016). For unsaturated FAs, depending on structural difference, it has two distinct types: cis- or trans-FA. Cis-FAs refer to those FAs with two hydrogen atoms adjacent to the double bond stick out on the same side of the chain, and conversely, are trans-FAs. The FA repertoire of edible vegetable oils are mostly cis- formation.

Free FAs have detergent properties, therefore are cytotoxic, and only trivial amount are present as this form. Majority of cellular FAs occur as part of glycerolipids, i.e. FAs are esterified to the OH group of a glycerol backbone. Membrane or neutral lipids consisting of FAs and a head group bound to a glycerol backbone are called glycerolipids. Glycerolipids are the most abundant types of lipids occur in nature, and can be classified into galactolipids, phospholipids, sulfolipids, and neutral lipid TAG (Boudiere et al., 2014). For example, membranes of Chlamydomonas are mainly made of monogalactosyldiacylglycerol (MGDG), digalactosyldiacylglycerol (DGDG), sulfoquinovosyldiacylglycerol (SQDG), diacylglyceroltrimethylhomoserine (DGTS), phosphatidylglycerol (PtdGro), phosphatidylethanolamine (PtdEtn) and phosphatidylinositol (PtdIns) (Giroud et al., 1988) (Figure 3). Among them, MGDG, DGDG and SQDG are the major components, account for about 70% of total membrane lipids in C. reinhardtii, which is important for photosynthesis (Vieler et al., 2007).

These lipid types share not only similarity in structure, but also are biosynthetically related. Dynamic turnover of membrane lipids and shuffling of acyl chains between different lipid classes, collectively called lipid remodeling, is a very active area of research (Benning, 2008; Hurlock et al., 2014). Lipid remodeling has been shown important to supply acyl-chains for TAG synthesis in N-starved Chlamydomonas reinhardtii (Li et al., 2012).

Figure 3. Common lipid species reported for C. reinhardtii cells. Labels R1, R2 and R3 represent different fatty acid acyl residues (Yang et al., 2015).

3. Lipid metabolism 3.1 Carbon and energy supply for FA synthesis FA synthesis requires a carbon source, phosphorylating power (ATP) and a reducing equivalents in the form of NAD(P)H. For example, to make a palmitic acid (C16:0), 8 molecules of acetyl-CoA, 14 NAD(P)H, and 7 ATP are required. In plants, de novo FA synthesis occurs in plastid, and the first committed step for FA synthesis is the convertion of an acetyl-CoA to a malonyl-CoA. Since acetyl-CoA can’t cross the plastid membranes, so it must be produced in the plastid or imported from cytosol and other subcellular compartments. The precursors of acetyl-CoA in many organisms include glucose-6-phosphate (Glc6P), dihydroxyacetone phosphate, phosphoenolpyruvate, pyruvate, malate, hexose, acetate, acetoacetate, or butyrate (Pleite et al., 2005). In photoautotrophs (like higher plants), it is generally accepted that pyruvate is a major source of acetyl-CoAs, and this reaction is catalyzed by pyruvate dehydrogenase (PDH) (Rawsthorne, 2002). Possible sources of pyruvate include glycolysis, malate and the pentose phosphate pathway (PPP). The exact contribution of each pathway is not clear, and this could change depending on the species, or tissues in question.

In addition to carbon sources, FA synthesis requires large amounts of reducing power (NADPH or NADH) (Slabas and Fawcett, 1992) and phosphorylating power (ATP). For example, ATP is required at the step of ACCase while NADH and NADPH are required in the β-ketoacyl reductase and enoyl reductase steps, respectively (Figure 4). For a heterotrophic plastid, the generation of reductant can be from Glc6P, malate and pyruvate (Smith et al., 1992; Kang and Rawsthorne, 1996) by glycolysis pathway, citric acid cycle (TCA cycle), glyoxylate bypass or via the plastid oxidative pentose phosphate pathway (OPPP) (Schwender et al., 2003). In chloroplasts, ATP is derived from photophosphorylation and NAD(P)H: ferredoxin reductase during photosynthetic electron transport and during glycolytic metabolism. So major sources of ATP and NAD(P)H are photosynthesis and respiration. Taken together, it is generally accepted that glycolysis is a main supplier of carbon, NAD(P)H and ATP for FA synthesis, and

20 this seems to be also true for non-photosynthetic oil storing tissues like oil seeds (Schwender et al., 2003). But little information is known about the sources or relative contributions of NAD(P)H or ATP for lipid synthesis in microalgae.

Figure 4. De novo FA synthesis in plants and algae. ACCase, acetyl-CoA carboxylase; ACP, acyl-carrier protein. (a), acetyl transacylase; (b), malonyl transacylase; (c), β-ketoacyl synthase; (d), β-ketoacyl reductase; (e), β- hydroxyacyl dehydratase; (f), Enoyl reductase. Abbreviations: ACP, acyl carrier protein; CoA, coenzyme A.

3.2 De novo FA synthesis and export FAs are important structural components of cells and energy storage substances, also are intermediates in the biosynthesis of hormones and other biologically important molecules. FA synthesis occurs in the plastid in green photosynthetic tissues such as plants and algae. The first committed step of de novo FA synthesis is the reaction catalyzed by acetyl-CoA carboxylase (ACCase), which is composed of α- carboxyltransferase, β-carboxyltransferase, biotin carboxylase (BC) and biotin carboxyl carrier protein (BCCP) (Sasaki and Nagano, 2004) and produce malonyl-CoA from acetyl-CoA. Then, malonyl-CoA enters the 2 carbon-carbon elongation processes, catalyzed by the enzymatic complexes collectively called fatty acid synthase (FAS), producing at the end a C16 or a C18 FA molecule in plants and in most algal species. The growing FA chain is carried between these active sites while attached covalently to the phosphopantetheine prosthetic group of an ACP. It requires 7 cycles to produce a C16:0 fatty acid (Figure 4).

In nature, there are two types of FAS, fatty acid synthase I (FASI) (Jenke-Kodama et al., 2005), which utilizes a single, large, multifunctional polypeptide enzyme system and fatty acid synthase II (FASII) which is characterized by the use of discrete, monofunctional enzymes for FA synthesis (Li-Beisson Y, 2010). In animals and yeasts, the FA synthesis reaction is catalyzed by FASI, whereas, in plants as well as in algae and in prokaryotes such as E. coli, the same reaction is catalyzed by FASII.

The C16- or C18-ACP produced is subjected to two fates in the plastid: where they could be used directly as substrate for plastidial acyltransferases to form galactolipids, or they could be subjected to cleavage by fatty acid thioesterases (FATs) which cleaves off the acyl carrier protein, and the free fatty acids formed are exported out of the plastid via the recently identified fatty acid export 1 (FAX1) (Li et al., 2015), and further made their way to the ER with the aid of an ABC transporter (ABCA9) (Kim et al., 2013), and finally being assembled to glycerolipids by ER-resident acyltransferases.

Glycerolipids are made through the action of these acyltransferases, and they could be further modified through the action of desaturases etc, forming at the end hundreds of glycerolipid molecular species, collectively named as lipidome of a cell.

3.3 Polar membrane lipid synthesis Glycerolipid syntheses in higher plants occur via two distinct pathways that are known as the chloroplast inner envelope-localized pathway (also called “prokaryotic pathway”) and the endoplasmic reticulum (ER)-localized pathway (also called the “eukaryotic pathway”) on the basis of their evolutionary origins (Ohlrogge and Browse, 1995)(Figure 5). The glycerolipid synthesis requires acyltransferases to esterify acyl- ACP to a glycerol molecule.

Figure 5. The prokaryotic and eukaryotic pathways of glycerolipid synthesis. The prokaryotic pathway occurs in plastids, uses acyl-ACPs as substrates, and esterifies predominantly palmitate (16:0) at position 2 of glycerol. The eukaryotic pathway occurs outside the plastid (primarily at the ER), uses acyl-CoAs as substrates, and positions 18-carbon fatty acids at position 2 of glycerol 3-phosphate. The figure was taken from that of (Ohlrogge and Browse, 1995). Abbreviations: ACP, acyl carrier protein; CoA, Coenzyme A ; G3P, glycerol 3- phosphate ; PA, phosphatidic acid; LPA, lyso PA; DAG, diacylglycerol; CDP-DAG, cytidine diphosphate-diacylglycerol; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PI, phosphatidylinositol; SL, sulfolipid.

The Kennedy pathway occurs in ER (Figure 6). G3P is the backbone for the assembly of glycerolipids. Acyl-CoAs are another substrates and they are attached to G3P via the reactions catalyzed by sn-glycerol-3-phosphate acyltransferase (GPAT), lysophosphatidic acid acyltransferase (LPAAT), phosphatidic acid phosphatase (PAP) and diacylglycerol acyltransferase (DGAT) to produce TAG (Kennedy, 1956).

Figure 6. Proposed pathways for TAG biosynthesis in plants. Three known pathways are outlined (Kennedy pathway, PDAT pathway and DAGTA pathway). DAG represents an important branch point between storage and membrane lipid synthesis. Abbreviations: G3P, glycerol 3-phosphate; GPAT, G3P acyltransferase; LPA, lysophosphatidic acid; LPAAT, LPA acyltransferase; PA, phosphatidic acid; PAP, PA phosphatase; DAG, diacylglycerol; TAG, triacylglycerol; PDAT, phospholipid:diacylglycerol acyltransferase; DGAT, diacylglycerol acyltransferase; DAGTA, diacylglycerol diacylglycerol acyltransferase; PtdCho, phosphatidylcholine; MAG, monoacylglycerol.

Although these two pathways occur in distinct subcellular compartments, but are similar in the following two enzymatic reactions from G3P to LPA and subsequentially

24 to PA. In the prokaryotic pathway, FAs are directly transferred from ACP (acyl carrier protein) to G3P, whereas, in the eukaryotic pathway FAs are first cleaved from ACP by FAT to form free fatty acids (FFAs) that are exported to cytoplasm and then esterified to CoA to form acyl-CoA. Finally, a PA phosphatase (PP) that converts PA to DAG. Diacylglycerol (DAG) is a central intermediate in membrane or TAG synthesis. In addition to the above pathway, DAGs can also be produced through hydrolysis of membrane lipids, such as via the reaction catalyzed by a choline phosphotransferase (CPT)(Voelker and Kinney, 2001). Afterward, specialty enzymes are then required for specific reactions related to the syntheses of galactolipid, phospholipids and betaine lipids for algae, as detailesd in (Li-Beisson et al., 2015).

Due to the specificities of the plastidial acyltransferases for certain acyl-ACP substrates, the phosphatidic acid (PA) made by the “prokaryotic pathway” has 16:0 at the sn-2 position and, in most cases, 18:1 at the sn-1 position. Whereas, the ER-located acyltransferases use acyl-CoA as substrates to produce PA with a C18 fatty acid at the sn-2 position (Ohlrogge and Browse, 1995). But recent data show the situation in C. reinhardtii is different. Two candidate genes for LPAAT were noted in C. reinhardtii (Li-Beisson et al., 2015). The CrLPAAT1 was reported to be located in the chloroplast with a preference for 16C at its sn-2 position (Yamaoka et al., 2016), whereas LPAAT2 has recently been identified as the ER isoform but also with a preference of 16C at its sn-2 position (Kim et al., 2018). Both are found to be implicated in TAG synthesis, because plastidial over-expression of CrLPAAT1 increased oil content (Yamaoka et al., 2016), and silencing of CrLPAAT2 reduced oil content during N starvation (Kim et al., 2018).

3.4 TAG biosynthesis TAGs are one of the most important lipids, and they are present in all eukaryotes including fungi, algae, plants and animals, and also in some prokaryotes such as Streptomyces species (Olukoshi and Packter, 1994) and Mycobacteria species (Akao

25 and Kusaka, 1976). TAGs are found in large amount in seeds or fruits of many plant species, including sunflower, oilseed rape, soybean, and oil palm (Graham, 2008; Bourgis et al., 2011). TAGs are not only important as energy source, but also represent an important source to supply acyl-chains for membrane lipid biosynthesis (Athenstaedt and Daum, 2006).

In plants, there are two major pathways for TAG synthesis: acyl-CoA dependent pathway (also known as Kennedy pathway) and acyl-CoA independent pathway (Zhang et al., 2009; Li-Beisson Y, 2010) (Figure 6). The relative contribution for each pathway in plants is still under debate. In Chlamydomonas there is one PDAT (Boyle et al., 2012; Yoon et al., 2012) but six diacylglycerol: acyl-CoA acyltransferases (DGATs) of two types, type 1 DGAT and type 2 DGTT. The DGAT genes are called DGAT1 and DGTT1-5 (Miller et al., 2010; Boyle et al., 2012; La Russa et al., 2012).

For the type-2 DGATs in C. reinhardtii, CrDGAT2A, B and C were investigated in overexpressing strains but total TAG accumulation was not changed significantly from wild-type under normal growth or after N- or sulfur (S)-depletion (La Russa et al., 2012). The substrate selectivity of CrDGTT1, 2 and 3 (DGAT2B, E and D, respectively) was assessed. CrDGTT1 preferred polyunsaturated fatty acids (PUFAs), CrDGTT2 preferred monounsaturated acyl-CoAs and DGTT3 preferred 16C acyl-CoAs (Liu et al., 2016). Knock-down of each of the three genes caused a 20-35% decrease in TAG together with a change in fatty acid composition of TAGs. Hung et al. (Hung et al., 2013) performed heterologous complementation assays for C. reinhardtii DGTT1-4 in yeast mutants and showed that DGTT1, 2 and 3 but not 4 complemented the TAG deficiency phenotype. Complementation with DGTT2 was the most effective. In agreement with previous reports, the authors could not detect transcripts for DGTT-5.

While most attention has been paid to the multiple DGAT2 genes/enzymes, a putative sequence for DGAT1 in C. reinhardtii (Li-Beisson et al., 2015) has been reported. For

P. tricornutum, a DGAT1 was cloned and its activity demonstrated in a yeast mutant (Guiheneuf et al., 2011) where a preference for saturated 16 or 18C fatty acids was displayed.

The “acyl-CoA independent pathway” occur in parallel to the Kennedy pathway. The acyl-CoA independent pathway can again occur in three possible ways. The first and most well characterized one is catalyzed by phospholipid: diacylglycerol acyltransferase (PDAT) using a phospholipid to produce a TAG and a LPA as co- product (Durrett et al., 2008). A second one is catalyzed by a DAG transacylase (DAGTA) which allows the formation of a TAG molecule from two DAGs and producing one MAG as a co-product (Stobart et al., 1997) (Figure 6). A gene encoding PDAT was found in C. reinhardtii and its activity demonstrated by expression in a TAG- deficient yeast mutant. MicroRNA silencing of PDAT in C. reinhardtii altered membrane lipid composition and reduced the growth rate (Yoon et al., 2012). The CrPDAT also had a strong lipase activity and was suggested to be functional during the log phase of growth under normal conditions but not during the large induction of TAG deposition on N-depletion (Yoon et al., 2012). In higher plants PDAT is thought to generally use PtdCho, which is not present in C.reinhardtii although Boyle et al. (Boyle et al., 2012) reported that the CrPDAT complemented TAG-deficient yeast (where it may well have used PtdCho). Moreover, the gene for CrPDAT is predicted to be chloroplast-located (Terashima et al., 2011) so the ability of the enzyme to use MGDG (but not DGDG or SQDG) as acyl donor in this alga is important (Yoon et al., 2012).

3.5 Lipid droplets (LDs) Since TAGs are inert carbon dense molecules, they can not form membranes, are therefore stored in a hydrophobic subcellular structure called LDs, also named as lipid bodies, oil bodies, oil globules, or oleosomes (Huang, 1996; Murphy, 2001). The choice of one name versus the other turns to be community-biased, and does not reflect any differences in their composition. LDs have been initially studied mostly in oil-rich

27 tissues/organs, like adipose tissue, oilseeds, or nitrogen starved cells (Huang, 1996; Zweytick et al., 2000; Murphy, 2001; Czabany et al., 2007; Goodman, 2008). With the development of new analytical and visualization techniques, it is now generally accepted that LDs are present in all eukaryotic cells.

LDs are an important subcellular organelle which are simple spherical compartment surrounded by a single layer of membranes (Huang, 1996). The size of LDs varies depending on species, growth stage and environmental conditions. It ranges from a dozen of nanometers to 30 µm. For example, Chlamydomonas reinhardtii make LDs around 1.5 µm (Moellering and Benning, 2010). The current LD structural model consists of a neutral lipid core (usually TAGs), which are surrounded by a monolayer of polar lipids (PL) and decorated with proteins (Figure 7). This model is generally accepted for LDs of diverse kingdoms (Zweytick et al., 2000; Czabany et al., 2008; Yang et al., 2012), but except the original work by the group of Anthony Huang, no experimental data has been obtained to validate this hypothetical model in any other species. TAG and PL can take up to 90% of total weight, whereas proteins weigh between 1-5%. Depending on species, some minor components including free fatty acids, carotenes, etc. are also present.

Figure 7. A structural model of lipid droplet (based on that of (Huang, 1992)).

LDs isolated from N-starved cells of C. reinhardtii contained 80-90% TAGs, some free fatty acids, and around 5-10% polar lipids (Wang et al., 2009; Nguyen et al., 2011). The 28 fatty acyl chains present in the TAG species are very similar to that of whole cell lysate, i.e. reduced polyunsaturated fatty acids and increases in saturated/monounsaturated fatty acid species as compared to non-stress conditions (Moellering and Benning, 2010; Fan et al., 2011; Siaut et al., 2011). The phospholipid monolayer of LDs differs in its composition from other cellular membranes such as ER, plasma membrane, or mitochondria, consisting with the general view that all membranes are unique in its composition (van Meer et al., 2008). Lipid class analyses by thin layer chromatography (TLC) have revealed that the betaine lipid diacylglycerol N’N’N’ - trimethyl homoserine (DGTS) and sulfoquinovosyldiacylglycerol (SQDG) are the two major lipid classes associated to isolated LDs from C. reinhardtii (Wang et al., 2009).

The first proteomics analysis of LDs isolated from microalgae was performed independently by several groups in the model C. reinhardtii. Most of these studies identified the presence of a novel protein named as MLDP (stands for ‘Major Lipid Droplet Protein’) (Moellering and Benning, 2010; James et al., 2011; Nguyen et al., 2011). MLDP has a molecular mass of 28 kDa and is the most abundant protein based on SDS-PAGE separation of total proteins extracted from purified lipid droplet, and has also shown the highest spectra count in LC-MS/MS analyses (Moellering and Benning, 2010; Nguyen et al., 2011). Based on primary amino acid sequences, MLDP is predicted as a very hydrophobic protein with no known function domains identified. It shares no primary sequence similarity to known LD proteins including that of plant oleosins (Huang, 1992) nor to that of mammalian perilipins (Londos et al., 1999).

Regarding LD biogenesis, due to the hydrophobicity property of the TAGs synthesized in the ER, the newly formed neutral lipids are unable to integrate into bilayer membranes, so they cluster and accumulate in the hydrophobic region inside ER membrane. More and more TAGs synthesized and accumulated into the LDs to form a bud, after the bud reaching a certain size, and then is released into the cytosol as a mature LD (Huang, 1996; Athenstaedt and Daum, 2006).

Recent data in studies carried out in Chlamydomonas has started to point out the contribution of chloroplast envelop to LD biogenesis (Figure 8) (Goodson et al., 2011). This data is further supported by bioinformatic prediction of a chloroplast location for some major proteins of TAG biosynthesis i.e. that of PDAT or GAPT1 (Li-Beisson et al., 2015); by biochemical analyses of signature acyl-groups at the sn-2 positions of TAG molecular species produced (Fan et al., 2011) and also by observations of LDs under transmission electron microscopy (TEM) (Fan et al., 2011). A plastidial location was nevertheless not observed in wild-type strains, but was limited to LDs present in the starchless mutant bafj5 (Fan et al., 2011). A possible chloroplast location for LDs in cells of Chlamydomonas was however challenged by two recent studies, i.e. another microscopic study on LD localization employing Confocal microscope and TEM (Moriyama et al., 2018), and also by a study on the substrate specificity of the ER LPAAT2 enzyme (Kim et al., 2018). The exact subcellular localizations of LDs in Chlamydomonas therefore remains controversial. However, as Moriyama et al. (Moriyama et al., 2018) argued that “Revisiting the Algal "Chloroplast Lipid Droplet": The Absence of an Entity That Is Unlikely to Exist”, latest literature have shown repeatedly that LDs rich in TAG are found to accumulate in LDs of diverse algal species (Balamurugan et al., 2017; Nobusawa et al., 2017; Wang and Dong, 2018), and the fact that the well known plastoglobules do contain TAG despite at lesser amount. Therefore to conclude, new techniques beyond the use of microscopy are needed to solve this problem eventually, or this should help to further increase our understanding on the biogenesis and heterogeneity of LDs accumulated.

Figure 8. A model for formation of LDs in Chlamydomonas reinhardtii. Cited from that of (Goodson et al., 2011). In green are chloroplast and in brune is ER. Abbreviations: TAG, triacylglycerol; DAGAT, diacylglycerol.

In addition to LDs present in cytosolic environment, LDs are also presented in the chloroplast of higher plants, usually known as plastoglobules. Plastoglobules usually have less TAGs, but with the occurrence of large amount of tocopherols, phytoesters, and other isoprenoid-derived metabolites (Brehelin and Kessler, 2008). Although plastoglobules from higher plants have been isolated, and proteomics/lipidomics tools have been used to identify its protein/lipid repertoires, nothing is yet known about the algal plastoglobules. In conclusion, be it cytosolic or plastidial, detailed molecular mechanisms of LD biogenesis in algal cells and LD heterogenity remain to be illucidated.

3.6 Lipid catabolism Lipid catabolism is essential to ensure cellular lipid homeostasis during vegetative growth, and is a major route to release carbon and energy contained in storage lipids for cell division and growth. Lipid degradation is comprised of two spatially and temporarily separated processes i.e. lipolysis and fatty acid -oxidation. Lipolysis is catalyzed by ester bond-breaking lipases at membranes or lipid droplets. A common product is free fatty acids, which can be used to make new membranes, or be broken down to acetyl-CoAs through -oxidation in peroxisome or mitochondria. A schematic presentation of lipolytic processes and FA degradation and their subcellular compartmentation is drawn in Figure 9.

Figure 9. Pathways and processes involved in lipid catabolism in microalgae. This figure is taken from (Kong et al., 2018). Fatty acids are made in the chloroplast, one part of it is being used to make plastidial membranes, and the other part is exported to the ER where they are assembled to glycerol to make the extra-plastidial membranes. In response to developmental signal or adverse conditions, cells accumulate some of these acyl-chains as TAGs in lipid droplets. Lipid turnover starts from lipolysis of 32 structural or storage lipids through action of various lipases. The released free fatty acids enter peroxisomes for complete degradation. Fatty acid β-oxidation operates as a spiral reaction consisting of four repeating enzymatic steps. Acetyl-CoA, the final product of fatty acid degradation, are then used to make a 4-carbon compounds by the glyoxylate cycle partly operating in the peroxisomes. Succinate produced are used to fuel the TCA cycle in the mitochondria and subsequently used to produce sucrose for growth through the gluconeogenesis pathway in the cytoplasm. It is worth mentioning here that this pathway is drawn mostly based on what we know in the model alga Chlamydomonas, and in some other algal species, FA β-oxidation is known to occur in mitochondria. Known enzymes are written in red and bold. Biosynthetic steps are in grey lines, whereas degradative steps are in black.

Abbreviations: ACP, acyl-carrier protein; ACX, acyl-CoA oxidase; CoA, coenzyme A; DAG, diacylglycerol; DGDG, digalactosyldiacylglycerol; DGTS, diacylglycerol- N,N,N- trimethylhomoserine; ER, endoplasmic reticulum; FAS, fatty acid synthase; FFA, free fatty acid; G3P, glycerol-3-phosphate; KAT, ketoacyl-CoA thiolase; ICL, isocitrate lyase; LACS, long chain acyl-CoA synthetase; LD, lipid droplet; CrLIP1, Chlamydomonas lipase 1; MAG, monoacylglycerol; MFP, multifunctional protein; MGDG, monogalactosyldiacylglycerol; PDAT, phospholipid:diacylglycerol acyltransferase; PEP, phosphoenoylpyruvate; PtdGro, phosphatidylglycerol; PtdEtn, phosphatidylethanolamine; PtdIns, phosphatidylinositol; SDP1, sugar-dependent1; SQDG, sulfoquinovosyldiacylglycerol; TAG, triacylglycerol; TCA, tricarboxylic cycle.

The most studied and notable example is germinating seeds where rapid breakdown of oil reserves is essential to provide carbon needed for seedling establishment before photosynthesis takes place, whereas a block in this process can compromise seed germination and seedling establishment therefore causing hugh loss in agriculture (Graham, 2008). Although less spectacular, lipid turnover is also essential in rapidly growing tissues/dividing cells where it is critical to remove harmful free fatty acids which can originate from both protein and lipid degradation. Lipid turnover has also been found to contribute to resource allocation during aging or senescence (Troncoso- Ponce et al., 2013), and plays a role in response to biotic/abiotic stresses, and to nutrient starvation.

Lipid degradation can be easily manipulated and observed in algal cells. For example, the triacylglycerols (TAGs) accumulated during N starvation can be rapidly remobilized upon N resupply (within 24h) to synthesize membranes and restore cell growth (Siaut et al., 2011; Li et al., 2012) (Figure 10). But relatively much less is known on lipid catabolism in algae. At the onset of this PhD thesis, only two lipases have been characterized at a molecular level related to lipolytic processes in the model alga Chlamydomonas reinhardtii. These two proteins are the CrLIP1 and PGD1. The CrLIP1 was able to complement the yeast TAG lipase mutant (tgl3Δtgl4Δ), and artificial mcriRNA silencing of the CrLIP1 in situ resulted in strains compromised in TAG degradation. Biochemical analyses of the CrLIP1 activity showed that CrLIP1 is a DAG lipase with broad substrate specificity (Li et al., 2012). Plastid Galactoglycerolipid Degradation 1 (PGD1) is the second lipase identified (Li et al., 2012). Mutants deficient in PGD1 accumulated 50% less TAGs than its parental line during N starvation. Using pulse-chase labelling assays, the authors demonstrated that PGD1 hydrolyzes MGDG at its sn-1 position, therefore producing lyso-MGDG and a free acyl-group. Based on these results, the authors proposed a route of acyl-fluxes from MGDG to TAG upon N starvation in Chlamydomonas.

Figure 10. Changes in lipid content in response to nitrogen status in the media. (A). Changes in cellular TAG levels as quantified by Flow cytometry after cells being stained with Nile red. (B). Changes in polar membrane lipid content. Data are means of three biological replicates, and error bars represent standard deviations.

Even less is known related to the β-oxidation of FAs in microalgae. FA beta-oxidation is known to occur in mitochondria of mammalian cells, and occur exclusively in peroxisome of higher plants and yeast (Poirier et al., 2006; Graham, 2008). At the onset of this PhD thesis, although candidate genes encoding proteins homologous to known plant peroxisomal proteins can be identified in the genome of Chlamydomonas, none of them has however been validated experimentally (Li-Beisson et al., 2015). In Chlamydomonas, it is not yet clear if classical peroxisomes occur. Nevertheless, during this thesis work, Plancke et al (Plancke et al., 2014) reported that isocitrate lyase plays a critical role in glyoxylate cycle, and later on, they further showed that major enzymes of glyoxylate cycle occur in the peroxisomes of Chlamydomonas reinhardtii (Lauersen

35 et al., 2016), presenting therefore the first metabolic pathway associated to peroxisomes in the model alga Chlamydomonas.

To identify molecular players implicated in lipid breakdown upon N resupply, two independent forward genetic screens have been carried out to look for mutants defected in oil remobilization (Cagnon et al., 2013; Tsai et al., 2014). The identity of enzymatic and regulatory proteins implicated in lipid turnover has just started to be unraveled. One of the major focus of this PhD thesis is to dissect the molecular mechanisms behind the compromise observed in the Chlamydomonas mutants defected in lipid catabolism.

4. Branched-chain amino acid (BCAA) metabolism 4.1 BCAA cellular functions Plants and algae synthesize 21 amino acids (AAs), which are major building blocks to assemble a functional protein. Leucine (Leu), isoleucine (Ile) and valine (Val) are referred to as BCAAs, and have in common a branched aliphatic chain (Figure 11), and they are also biosynthetically related (Binder, 2010). As a result of their aliphatic nature, BCAAs are typically located in transmembrane domains of proteins (Binder, 2010).

Figure 11. Structure of branched-chain amino acids. Images were taken from (Binder, 2010).

For two main reasons, the research on BCAAs has attracted intensive attention in recent years. First, BCAAs are among the nine amino acids considered “essential” because

36 they can not be synthesized de novo by mammals including humans (Harper et al., 1984). They account for 35% of the essential amino acids found in muscle proteins (Shimomura et al., 2006) (Shimomura et al., 2004). Second, acetohydroxyacid synthase (AHAS), catalyzing a committed step for Val biosynthesis, is a commercially important target for four classes of herbicides (Jander et al., 2003).

Although in most cases, storage proteins are major sources of amino acids, free amino acids can also account for 1 to 10% of the total, thus could make significant contributions to the levels of essential amino acids, and can have an impact on the nutritional value of the species concerned (Muehlbauer et al., 1994). The level of free amino acids depends on tissue types, developmental stage, or environmental conditions. A large amount of free amino acids can be cytotoxic, therefore their turnover is under tight regulation. For example, the lack of a key enzyme in the degradation of branched- chain α-keto acid dehydrogenase (BCKDH) result in maple syrup urine disease in human (Podebrad et al., 1999; Wang et al., 2011).

BCAAs are not only building blocks of protein and essential nutrients for human, but also participate in a wide variety of metabolic pathways and also play as critical regulators of a number of signaling pathways in animals, for example in the regulation of protein and lipid synthesis, cell growth and autophagy (Wang and Proud, 2009; Sengupta et al., 2010). In animal cells, the increased level of BCAAs is a significant risk factor for the development of insulin resistance and diabetes (Wang et al., 2011), and the supplementation of BCAAs helps to prevent oxidative damage and supports cardiac and skeletal muscle biogenesis (Valerio et al., 2011). In addition to their role as essential nutrients, in plants, BCAAs and their derivatives such as glucosinolates (GSL), fatty acids and acyl-sugars are important for plant growth, development, defense and flavoring (Binder, 2010; Gonda et al., 2010). For example, Leu-derived metabolites are important signaling molecules that can activate the target of rapamycin (TOR) signaling pathway in animals that in turn regulates several aspects of growth and development

(Kimball and Jefferson, 2006, 2006).

Besides a structural, storage and signaling role, BCAAs could be used as respiratory substrates to supply cells with carbon and energy (Araujo et al., 2010; Binder, 2010; Kochevenko et al., 2012; Hildebrandt et al., 2015). Although protein respiration is one of the least efficient, and it occurs essentially in cells experienced through adverse conditions especially carbon starvation. We want to emphasize that ATP production from BCAAs and lysine catabolism is relatively high when compared with other amino acids (Hildebrandt et al., 2015), this implies that BCAAs are very important when being used as respiratory substrates to supply ATP. For example, in the model plant Arabidopsis thaliana, the absence of some components of BCKDH showed early dark- induced senescence and accumulated more BCAAs in seeds (Peng et al., 2015). Moreover, BCAA metabolism can also provide precursors for specialized metabolism. For example, variants derived from isopropylmalate synthase (IPMS), the committed enzyme in Leu biosynthesis, contribute to the Met side chain elongation in Arabidopsis (de Kraker and Gershenzon, 2011).

4.2 BCAA biosynthesis BCAA levels are tightly controlled through the orchestration of their synthesis and degradation. This idea has been strengthened by the observation of strong correlations between the levels of free BCAAs in Arabidopsis. BCAAs are synthesized in the chloroplasts in higher plants where reducing power and ATP can be provided directly, and are degraded mostly in mitochondria and certain steps toward the end could also occur in peroxisomes (Hildebrandt et al., 2015). Catabolic pathways of BCAAs appear to play an important role in regulating the level of these essential amino acids (Galili et al., 2016). Despite their high nutritional value, and the long term biotechnological goal to enrich their content in edible plant species, the genes and proteins that constitute the full catabolic pathway and their regulation is not yet fully understood in Arabidopsis or in any other plant species (Binder, 2010; Hildebrandt et al., 2015; Galili et al., 2016).

Archaea, bacteria, fungi and plants are able to synthesize BCAAs with a conserved metabolic pathway, apart from the threonine independent citramalate pathway in some prokaryotes (Wu et al., 2010). Benefited from complete genome sequences and thorough bioinformatics analysis of these sequences, all enzymes and genes involved in the BCAA biosynthesis can be easily identified. The BCAA biosynthesis pathway is shown in Figure 12 (Binder, 2010). Briefly, biosynthesis of Val and Ile occurs in parallel pathways and share some common catalytic enzymes i.e. acetohydroxyacid synthase (AHAS), ketolacid reductoisomerase (KARI), dihydroxyacid dehydratase (DHAD) and branched-chain aminotransferase (BCAT) (Singh and Shaner, 1995). An exception is the threonine deaminase (TD) catalyzing the deamination and dehydratation of Thr to form α-ketobutyrate (KB) and ammonia, the initiating step toward the synthesis of Ile. Leu biosynthesis branches off from 3-Methyl-2- oxobutanoate which is before the final transamination step of Val biosynthesis and requires three additional enzymes i.e. isopropylmalate synthase (IPMS), isopropylmalate dehydratase (IPMDHT) and isopropylmalate dehydrogenase (IPMDH) (Binder et al., 2007).

Figure 12. BCAA biosynthesis pathway in the plastid. Key enzymes (blue) and metabolites (black) of BCAA biosynthesis. Allosteric inhibition is indicated by red lines. The restoring effect of Val on threonine deaminase inhibition by Ile is depicted as a dotted green line. This image was taken from (Binder, 2010).

4.3 BCAA degradation Concerning BCAA degradation, here we focused particularly on the three common steps catalyzing initial steps in the degradation of all three BCAAs. The BCAA catabolic pathway and its major reactions are shown in Figure 13.

Figure 13. Pathway of BCAA catabolism in the mitochondria of higher plants. This figure is partly based on that of (Peng et al., 2015). Abbreviations: IVD, isovaleryl-CoA dehydrogenase; BCAT, Branched-chain aminotransferase; α-KG, α-ketoglutarate; BCKDH, branched-chain ketoacid dehydrogenase; ETC, electron transport chain; ETF, electron transfer flavoprotein; ETFQO, electron transfer flavoprotein ubiquinone oxidoreductase; ETC, electron transport chain; MCC, 3-methylcrotonyl-CoA carboxylase; E-CoAH, enoyl-CoA hydratase; HMG-CoA, 3-hydroxy-3-methyl-glutaryl-CoA; HML, 3-hydroxyl-3- methylglutaryl-CoA lyase; TCA, tricarboxylic acid.

Branched-chain aminotransferase (BCAT): BCAT (EC 2.6.1.42) catalyzes the first step of BCAA degradation, and this enzyme also acts at the last step of BCAA biosynthesis in the plastid. In Arabidopsis, seven BCAT genes have been identified: BCAT1 (At1g10060), BCAT2 (At1g10070), BCAT3 (At3g49680), BCAT4 (At3g19710), BCAT5 (At5g65780), BCAT6 (At1g50110), and BCAT7 (At1g50090). But so far, no evidence has been obtained for the transcription of BCAT7 in Arabidopsis

(Binder, 2010). Different isoforms are found to be present in the plastid (BCAT2, 3), cytosol (BCAT4, 6) or mitochondria (BCAT1,2,5). Among which, BCAT1, 2 and 5 are mitochondria located, suggesting their possible implication in BCAA degradation. Although similar enzymes are required for their synthesis and degradation, to avoid potential interference, this step is compartmentalized in different subcellular compartment, i.e. chloroplast for biosynthesis and mitochondria for degradation. Considering its important position in BCAA homeostasis, BCATs have been characterized in diverse species ranging from spinach, Arabidopsis to potato, barley and tobacco (Maloney et al., 2010; Kochevenko et al., 2012).

Branched chain α-keto acid dehydrogenase (BCKDH): The second step in the breakdown of BCAAs is the irreversible oxidative decarboxylation of α-keto acids to their respective acyl-CoAs (Binder, 2010). This reaction occurs in mitochondria (Taylor et al., 2004), and is catalyzed by an enzyme complex called branched-chain α-keto acid dehydrogenase (BCKDH). BCKDH belongs to a 2-oxoacid dehydrogenase multienzyme complex family, including also pyruvate dehydrogenase and α- ketoglutarate dehydrogenase (Yeaman, 1989). The BCKDH complex consists of three catalytic components: a heterotrimeric branched-chain α-ketoacid dehydrogenase (E1: EC 1.2.4.4), a dihydrolipoyl transacylase (E2: EC 2.3.1.168), and a dihydrolipoamide dehydrogenase (E3: EC 1.8.1.4). The protein complex is super large (up to 9 MDa) and consists of multiple copies of the three subunits (Binder, 2010). In Arabidopsis, for example, only E2 core of this complex is found to contain already 24 subunits, which produce a molecular mass of 0.95 MDa (Mooney et al., 2000). This reaction requires CoA and NAD+ as cofactors, thus in addition to acyl-CoA esters, it produces NADH. It has been hypothesized that the BCKDH complex contributes to mitochondrial electron transport by feeding NADH to the internal alternative NADH dehydrogenase of the complex I (Schertl and Braun, 2014).

The activity of BCKDH has also been analysed in the leaves of Arabidopsis compared

42 to that from rat muscle. Its activity was found to be higher in dark-treated leaves than from rat muscle, but no activity could be detected from light-grown plants (Fujiki et al., 2000). The authors also found dark treatment can increase accumulation of the transcripts and proteins of BCKDH E2 in Arabidopsis leaves, and found BCKDH E1β and E2 subunits are modulated by carbohydrate level, i.e. high carbohydrate concentration can suppress their expression. Application of a photosynthesis inhibitor strongly induced BCKDH E1 and E2 expression even under light illumination in detached Arabidopsis leaves. In short, most of the BCKDH subunit genes was only activated under carbohydrate deprivation condition. This condition can be a result of long-term maintenance in the dark for a photosynthetic organism. Taken together, these results suggest that BCKDH play a key role through degradation of BCAAs to supply an alternative energy source when photosynthesis is hampered (Fujiki et al., 2000).

Isovaleryl-CoA dehydrogenase (IVD): The third and the last common step requires the oxidation of the acyl-CoA esters to their respective dehydrogenated product. This reaction is catalyzed by isovaleryl-CoA dehydrogenase (IVD, EC1.3.99.10). Araujo et al. found that IVD is not only involved in BCAA degradation, but also plays an important role in the breakdown of phytol and Lys which is important for alternative respiration under dark-induced senescence (Araujo et al., 2010). The electrons of this redox reaction are transferred into the mitochondrial respiratory chain at the level of ubiquinone via FAD, the electron-transfer flavoprotein (ETF) and an electron-transfer flavoprotein: ubiquinone oxidoreductase (ETFQO, EC 1.5.5.1). In Arabidopsis, two subunits of ETF (ETFα: At1g50940 and ETFβ: At5g43430) and one ETFQO (At2g43400) have been identified which supports respiration during carbon starvation or extension of darkness (Araujo et al., 2010). The incorporation of BCAA catabolic pathway to the mitochondrial respiratory pathway at this step is reinforced by two observations: that the mutant defected in ivd1-2 senescenced faster than WT in prolonged darkness, and that mutants defected in ETF or ETFQO over-accumulated BCAAs (Ishizaki et al., 2005; Araujo et al., 2010).

The subsequent steps of BCAA breakdown took a different route depending on the amino acids in question. The reactions catalyzed by enoyl-CoA hydratase led to the formation of a range of hydroxyl acyl-CoAs, which are subsequently catabolized by a series of enzymes located either in mitochondria or peroxisomes (Reumann et al., 2009; Hu et al., 2012; Hildebrandt et al., 2015). The final products enter the TCA cycle for complete respiration, providing energies for metabolic processes.

With the advent of molecular genetics and insertional mutants, several proteins of the catabolic pathway have been investigated through mutant studies. Mutants blocked in BCAA catabolism that have been studied in detail include the bcat2, ivd, mcca, mccb and hml mutants. Common features of the defect in the catabolic pathway of these mutants are the over-accumulation of BCAAs and a sign of earlier senescence than WT during prolonged darkness. These studies also provided solid evidence that BCAA catabolic pathway interacts actively with mitochondria energy metabolism at multiple levels via feeding either electrons or carbon precursors (i.e. at the level of BCKDH, IVD, and also through TCA cycle). Transcripts encoding major enzymes of BCAA catabolic pathways are found to be upregulated in the dark, in prolonged darkness, and also in carbon-deficient conditions (Ge et al., 2014; Peng et al., 2015). They are also found to be correlated to the expression of genes of autophagy and ubiquitination processes (Galili et al., 2016).

5. A crosstalk between metabolism of lipids and BCAAs BCAA catabolism in mitochondria can provide precursors in the form of succinyl-CoA or acetyl-CoA for TCA cycle, and in the meantime, it also provides direct electron donors to mitochondrial electron transport chain (Ishizaki et al., 2005; Araujo et al., 2010). Acetyl-CoAs are essential building blocks for FA synthesis, and ATP generated through mitochondrial respiration could be an important source to sustain lipid synthesis, especially during stress conditions when ATP production in chloroplast is

44 low.

In mammalian cells, FA β-oxidation occur essentially in mitochondria, therefore the metabolic link between lipid synthesis and that of mitochondrial activity has been well established. It has often been observed that the level of BCAAs is a marker for obesity (Newgard; Crown et al., 2016; Green et al., 2016). For example, lately in mouse cell lines, Green et al. reported that inhibition of BCAA degradation decreased oil content, and the authors further report that BCAA could contribute up to 30% acetyl-CoA for lipogenesis in adipocytes (Green et al., 2016).

Considering the many possible connections between lipid synthesis and degradation of BCAAs, relatively few demonstrated evidence is available regarding this metabolic relationship occurring in plants and algae. Below we describe the few known examples of the occurrence of such a crosstalk in plants and algae.

The first report on a possible connection between metabolisms of lipids and BCAAs is the observation that an Arabidopsis mutant disrupted in Val catabolism (the CHY1 gene) is compromised in peroxisomal FA β-oxidation (Lange et al., 2004). The authors provide evidence that CHY1 encodes a peroxisomal protein that is 43% identical to the mammalian β-hydroxyisobutryl-CoA hydrolase of the Val catabolism. The authors analyzed levels of several metabolites in the mutant, and based on this result, they concluded that disruption of CHY1 results in over-accumulation of methacrylyl-CoA, a toxic intermediate, which inhibits the activity of 3-ketoacyl-CoA thiolase, thereby preventing normal functioning of FA β-oxidation.

A second report on this topic came from the study of lipogenesis processes in the model diatom Phaeodactylum triconumtum using a quantitative proteomics approach (Ge et al., 2014). Ge et al. found that in parallel to the onset of massive oil accumulation, proteins of BCAA catabolic pathways showed >2 fold upregulation in Phaeodactylum.

To firmly establish this connection, ruling out a mere coincidence, the authors created RNAi silenced lines for one of the most significantly upregulated genes the methylcrotonyl-CoA carboxylase (MCC2). The silenced strains over-accumulated BCAAs (especially Leu), and as a result of this perturbation of carbon fluxes, the silenced lines accumulated less neutral lipids. This study points out a possible carbon route from BCAAs to neutral lipids during N starvation in diatom. Later on, the same group published another article also related to the relationship between BCAA catabolism and lipid synthesis (Pan et al., 2017).

More lately, during my PhD thesis, through isolation of high-oil mutant in the green alga Dunaliena tertiolecta, Yao et al. suggested that the high-oil phenotype is related to an enhanced catabolic pathway of BCAA degradation (Yao et al., 2017). However, it remains to be worked out what are the genetic mutations present in the mutant that confer the higher oil content.

6. Chlamydomonas reinhardtii Chlamydomonas reinhardtii is a single-celled green alga, about 5-10 microns in diameter, with two flagella. The first laboratory strain of Chlamydomonas was isolated from a field near Amherst, Massachusetts, in 1945 by Gilbert Smith (https://en.wikipedia.org/wiki/Chlamydomonas_reinhardtii). It has a cell wall made of glycoproteins rich in hydroxyproline, a large cup-shaped chloroplast, a large pyrenoid, and a photosensitive eyespot (Figure 14). C. reinhardtii is widely distributed in soil and fresh water all over the world.

Figure 14. Structure of a vegetative cell of Chlamydomonas reinhardtii. The drawing was taken from (Sasso et al., 2018).

C. reinhardtii has emerged as a prime model for algal research due part to some of the features below. It is an unicellular eukaryote with relatively simple subcellular structure. Its vegetative cells are haploid. All three of its genomes can be transformed and are fully sequenced (Merchant et al., 2007). It has a short generation time (usually less than 8h). Finally it can be cultivated under autotrophic, heterotrophic or mixotrophic conditions.

Research field traditionally employ Chlamydomonas reinhardtii as a model including study of flagella biology and photosynthesis (Harris, 2001). More recently, Chlamydomonas has also emerged as a model for studying metabolism, especially that related to the synthesis and accumulation of energy-rich compounds for example starch, hydrogen and lipids (Chochois et al., 2009; Merchant et al., 2012; Liu and Benning, 2013; Li-Beisson et al., 2015).

Molecular tools available for basic research: Many molecular techniques for genetic manipulation of Chlamydomonas have been developed and libraries of tagged insertional mutants have been available since 2016 (Gonzalez-Ballester et al., 2011; Li et al., 2016). Like plants and other eukaryotic algae, Chlamydomonas possesses three genetic systems located in nucleus, chloroplast, and mitochondria. The three genomes can be transformed (Boynton et al., 1988; Kindle, 1990). Targetted gene manipulation tools developed for C. reinhardtii include the establishment of RNAi or amiRNAi for generation of knock-down lines (Molnar et al., 2007; Molnar et al., 2009; Schmollinger et al., 2010), and lately, the use of CRISPR/Cas9 for genome editing of this microalga (Greiner et al., 2017; Shin et al., 2019).

OBJECTIVES OF THIS PHD THESIS

Considering the negative impact of fossil fuel consumption, and the finite underground reserve, production of sustainable and renewable clean energies is a pressing matter. Algal TAGs represent an attractive source for clean fuel production. A major issue is the observation that in algal cells significant oil accumulation occurs only under conditions when growth is impaired (such as N deficiency, high salt, stationary phase, or HL). Several fundamental as well as engineering challenges need to be solved before the establishment of an economically viable sector on algal bioenergy platform. Although the major steps for FA synthesis and TAG metabolism have been deduced based on that of land plants, little is known about carbon and energetic context. Therefore, the major objectives of this Thesis was to investigate possible source(s) of acetyl-CoA, ATP and reducing equivalents used for TAG synthesis in algal cells using a combination of genetic, biochemical and microscopic approaches in the model microalga Chlamydomonas reinhardtii.

First, we examined TAG synthesis in cells exposed to HL, a condition favoring production of reducing power. We showed that HL-exposed cells made significant higher amount of TAGs than those grown under low light (LL). Furthermore, LDs isolated from cells exposed to HL contained a different lipid and protein composition compared to LDs present in N-starved cells. Combined with Confocal microscopy imaging, we suggest the occurrence of large heterogeneity in LDs present in algal cells.

Secondly, we are interested in FA catabolism in Chlamydomonas reinhardtii. To investigate the molecular details of this pathway, we characterized two mutants, i.e. the acyl-CoA oxidase 2 (acx2) and the malate dehydrogenase 2 (mdh2) mutants of Chlamydomonas reinhardtii. We provided for the first time evidence that the β- oxidation of FAs is a peroxisomal process. We further demonstrated that knocking out FA degradation during N starvation can be used to increase TAG content, and that FA

β-oxidation plays a role in FA turnover processes during a day/night cycle. Through detailed characterization of the mdh2 mutant under photoautotrophy, we revealed for the first time the occurrence of a redox transduction pathway from peroxisome to chloroplast. We showed further that this redox communication is important for reserve accumulation as well as important for cell fitness especially during N deficiency or under HL. Taken together, these two studies highlighted the importance of FA β-oxidation in overall cell metabolism and physiology.

A major work of this study is that we identified one mutant accumulated less oil during N starvation, and also showed a delay in oil remobilization upon N resupply. The underlying genetic lesion is identified in a gene encoding a putative α-subunit of BCKDH complex of BCAA catabolism. In this study, we characterized the mutants genetically, biochemically, and physiologically, and with the focus to establish the possible metabolic connection between the homeostasis of lipids and that of BCAAs. We showed that BCAA catabolism plays a dual role in TAG synthesis via providing carbon precursors and ATP, since mutants defected in BCKDH complex all showed reduced TAG amount under N starvation and with lower mitochondrial respiration. We showed further that BCAA catabolism plays a dual role in TAG synthesis via providing carbon precursors and ATP.

Taken together, this work highlighted the complex interplay between carbon and energy metabolism in green photosynthetic cells, and pointed future directions for genetic improvement of algal strains for biofuel productions.

MATERIALS AND METHODS

1. Strains and culture conditions Two different Chlamydomonas reinhardtii cell strains were used in this study: dw15.1 (nit1-305 cw15; mt+) and 137AH. The strain dw15.1 was the parental strain to generate the insertional mutant library (Cagnon et al., 2013). All strains were maintained on tris- acetate-phosphate (TAP) agar (1.5%) plates in a 25°C constant temperature culture room and under continuous illumination (about 30 μmol photons m-2s-1). Mutant strains were maintained on agar plates with appropriate amount of antibiotics, but when grown in liquid culture, antibiotics were omitted to avoid any potential side effect on cell physiology. Flasks with liquid cultures were incubated in incubation shakers (INFORS HT), at 25°C, 120 rpm shaking, with 100 μmol photons m-2s-1 illumination and with addition of 2% CO2 in the air if autotrophic growth is desired. For photoautotrophic growth, cells were cultivated in minimal media (MM), whose composition is similar to that of TAP except the absence of acetate. To induce TAG accumulation, media without nitrogen (TAP-N or MM-N) was used.

For routine analysis, cell numbers, sizes and total cellular volumes were quantified by the use of a Beckman Coulter Multisizer 4 (MultisizerTM3 Coulter Counter, Beckman Coulter, USA).

2. The use of Flow cytometry to evaluate cellular oil content Cellular oil content of independent transformants were first screened using Flow cytometry after cells being stained with BODIPY 505/515 (D-3921) (4,4-difluoro- 1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-s-indacene, Thermo Scientific, USA) (Mou et al., 2012). BODIPY was added to reach a final concentration of 0.25 µg mL-1 (from a stock solution of 50 μg mL-1 BODIPY solubilized in dimethyl sulfoxide, DMSO) and incubated in the dark for 5 min at room temperature before analyses. Neutral lipids stained with BODIPY showed an emission peak at 521 nm when excited by 488 nm

51 laser line. Cultivation conditions used were the same as detailed in (Cagnon et al., 2013).

3. DNA and RNA isolation and cDNA preparation Genomic DNA, or RNA were extracted from Chlamydomonas cells (4x107) grown to exponential phase following the protocol of Tolleter et al (Tolleter et al., 2011), or the protocol of Nguyen et al. (Nguyen et al., 2013), for DNA and RNA, respectively. Extracted total RNA was treated with the TURBOTM DNase (Life technologies) to remove any genomic DNA contamination. To obtain highly pure RNA, the total extracts were passed through a NucleoSpin RNA Clean-up column (Macherey-Nagel) following the manufactures’ instructions. The first strand cDNA was made using the SuperScript VILO cDNA Synthesis Kit (ThermoFisher Scientific).

4. Identification of the APH8 Insertion Site in the Pb8C12 Mutant by Restriction Enzyme Site-Directed Amplification (RESDA)-PCR The insertion site of the APH8 cassette in the Pb8C12 mutant was identified by RESDA-PCR method (Gonzalez-Ballester et al., 2005). The principal of RESDA-PCR is shown in Figure 15. Primers and PCR conditions were as described in Kong et al. (Kong et al., 2017). After Blast against the genome of Chlamydomonas reinhardtii (v5.5, Phytozome), the insertion cassette was found located in the 4th intron of the locus Cre12.g539900 which encoding the E1α subunit of a large enzymatic complex named branched chain α-keto acid dehydrogenase (BCKDH).

Figure 15. Outline of the principals for RESDA-PCR. Flashes indicate the position of primers siuated in the sequence. This is taken from (Kong and Li-Beisson, 2018).

5. Reverse transcription PCR (RT-PCR) Expression level of the different BCKDH subunits (BKDE1α, BKDE1β and BKDE2) was evaluated by semi-quantitative RT-PCR. The RACK1 (Cre06.g278222) was used as a house keeping gene. Primer pairs used for RT-PCR analyses were CBLP-F/CBLP- R (for RACK1), BCKDH-F/BCKDH-R (for BKDE1α), BKDE1β-F/BKDE1β-R (for BKDE1β) and BKDE2-F/BKDE2-R (for BKDE2). RT-PCR reactions were carried out on template cDNAs prepared from WT or mutant RNAs, and amplified using the KOD Xtreme™ Hot Start DNA Polymerase (Novagen). PCR conditions for amplification were 94°C for 2 min, followed by 28 cycles of 10 s at 98°C, 30 s at 60°C and 2 min at 68°C. The PCR condition to amplify CBLP was 94°C for 2 min, followed by 25 cycles of 10 s at 98°C, 30 s at 60°C and 30 s at 68°C. After 28 cycles, a portion of the PCR reaction product was deposited on an agarose gel to verify formation of amplified product.

6. Cloning the full-length genomic DNA encoding BKDE1α and genetic complementation of bkdE1α mutant The BKDE1α gDNA (Cre12.g539900) and its native promoter region (1028 bp upstream the 5’ untranslated region) together with 3’UTR of RbcS2 were cloned into the PAP22 vector (which was obtained by inserting a Hygromycin gene to the blunt end of the PCR-blunt II-TOPO vector). Due to the presence of some complicated secondary structures around this region and also due to their high GC content (67%), we cloned the full-length sequence by two over-lapping fragments: i. the first fragment, containing the promoter region until the end of the 4th exon, was amplified using the primer pair (CrBCKDH-XbaI-FP3/CrBCKDH-R6); ii. The second fragment was cloned from the beginning of the 4th exon to the stop codon using the primer pair CrBCKDH- F6/CrBCKDH-NdeI-R2. These two fragments were then ligated into one single fragment by PCR. Briefly, the above two fragments were purified and mixed in an equal molar ratio (500 ng each) to serve as a template for PCR amplifications. The PCR reaction was carried out following KOD Xtreme™ Hot Start DNA Polymerase

(Novagen) protocol without primers [94°C for 2 min, then 11× (98°C for 10 s, 74°C for 3 min 10 s) and finally 1 cycle at 74°C for 5 min]. Then, two µL of above reaction were used as template to allow PCR amplification using primers CrBCKDH-XbaI-FP3 and CrBCKDH-NdeI-R2. The entire gene sequence was then confirmed by DNA sequencing (at GATC® Ltd). Finally, the full-length sequence (5349 bp), together with the 3’UTR of RbcS2 [from pChalmy_4 vector (Life Technologies) using RbcS3Ter- NdeI and RbcS2-NotI primers], was ligated to PAP22 vector (digested by XbaI and NotI). The linearized vector cut with XmaI was electroporated to the bkdE1α mutant cells, and antibiotic resistant clones were selected on agar plates containing 17 µg mL- 1 hygromycin. All primer sequences used were presented in Chapter 1 of the result part.

7. LD imaging To observe LDs, Chlamydomonas cells were first enriched via a gentle centrifugation (850 g, 3 min), then re-suspended in a small amount (100-200 µL) of fresh media with 0.25% glutaraldehyde. Cells were stained with BODIPY (to a final concentration of 0.25 µg mL-1, from a stock of 50 µg mL-1 in DMSO) in the dark for 5 min. Cells were then excited with a laser line at 488 nm, and emission was collected between 500-545 nm for BODIPY and between 650-730 nm for chlorophyll autofluorescence. All were carried out under a 63xoil immersion objective in a Confocal laser scanning microscope (TCS SP2, Leica, Germany). Pseudo colors were applied using the ZEN (Carl Zeiss) software.

8. Protein extractions, quantification and SDS-PAGE To examine whether WT and bkdE1a mutant differ significantly in protein, we performed SDS-PAGE analysis. Twenty million cells from TAP, TAP-N (24, 48 and 72h) and MM (24 and 48h) conditions were harvested by centrifugation for 3 min at 850 g and resuspended in 1 mL lysis buffer (50 mM HEPES-KOH pH 7.5, 1 mM EDTA, 5 mM Dithiothreitol (DTT), 10% (V/V) glycerol, 0.1x protease inhibitor cocktail; Sigma P9599). Cells were sonicated on ice for 42 s with an alternating cycle of 7 s

54 pulse/3 s pause. Lysates were transferred into 1.5 mL Eppendorf tubes, and centrifuged at 13,000 g for 10 min at 4°C. The supernatant (300 µL) was transferred to a new tube, to which 1200 μL acetone (precooled at -20°C) was added. The mixture was vortexed and incubated for >1 h at -20°C. The sample was centrifuged again at maximum speed for 15 min at 4°C, decant the supernatant and air dry for 5 min, re-solubilise in 200 μL of 0.2% SDS to use for protein quantification following the protocol of Pierce Bicinchoninic acid (BCA) protein assay kit (thermos scientific, USA). Ten μg protein from each sample were used for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis following the NuPAGE Bis-Tris Mini gels protocol (Novex, Life technologies).

9. Antibody generation and immunoblot analysis Polyclonal antibodies against the synthetic peptide (AA380-394: Cys- SGGLLTEPAVGAVGK) were immunized in two rabbits (ProteoGenix, SAS, Schiltigheim, France). For immunoblot analysis, total protein extracts (10 µg) were separated on 10% Bis-Tris gel using MES running buffer, transferred to BioTrace NT nitrocellulose membrane (Sigma-Aldrich) and immunoblotted with specific primary polyclonal antibodies (1/500) from rabbit, at room temperature for 2 h. This was followed by incubation with a secondary anti-rabbit antibody (Invitrogen) for 1 h. Immobilon™ Western Chemiluminescent HRP (horseradish peroxidase) substrate (EMD Millipore) was used for detection and images were recorded using a G:BOX Chemi XL (Syngene).

10. Lipid analyses All hardwares used for lipid extraction and analyses were glass ware, and pre-rinsed with ethanol to avoid contamination from detergent. A general flow chart of lipid analyses at various levels are summarized in Figure 16 (Li-Beisson et al., 2016).

Figure 16. A summary of major lipid analytical tools. Abbreviations: GC-MS, gas chromatography – mass spectrometry; LC-MS/MS, liquid chromatography-tandem mass spectrometry. MGDG, monogalactosyldiacylglycerol; DGDG, digalactosyldiacylglycerol; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; SQDG, sulfoquinovosyldiacylglycerol; TAG, triacylglycerol. Images were taken from that of (Li-Beisson et al., 2016).

Lipid extractions: Depending on downstream needs, total lipids were extracted from exponentially-grown Chlamydomonas cells using either the method of Bligh and Dyer (Siaut et al., 2011), or MTBE (methyl-tert-butyl ether) method (Goold et al., 2016), for quantification of lipid classes by Higher Performance-Thin Layer Chromatography (HP-TLC), or for lipid molecular species analysis by ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS), respectively.

Lipid class identification and quantification by TLC: To allow quantification of lipid classes in a given sample, total lipid extracts were first migrated on a silica-coat glass plates, then developed under a mixture of solvent to allow separation by mass and polarity etc. Lipid amount in each lipid classes was then quantified based on densitometry via comparing to a standard curve generated from the same type of lipid class from a commercial source. In this study, we have used an automated “High- Performance Thin-Layer Chromatography” platform (HP-TLC, CAMAG) for this purpose. Briefly, a given amount of lipid extracts was deposited on an 20x10 cm silica gel 60 F254 TLC plate (Merck KGA, Germany) using an ATS5 automatic TLC sampler (Camag, Switzerland). The TLC plate was developed either in a mixture of hexane/diethyl ether/acetic acid (17/3/0.2, v/v/v) for neutral lipid separation, or developed in a mixture of acetone/toluene/water (91/30/8, v/v/v) for polar lipid separation. After development, the plates were air dried, and lipids were revealed after dipping in a mixture of (20 g CuSO4, 80 mL H3PO4 and filled with H2O to 1 L) and heated at 170°C for 20 min. The lipid standards used were triheptadecanoin (C17:0 TAG, Sigma-Aldrich, Saint-Louis, USA), monogalactosyl distearoylglyceride (MGDG, Larodan Fine Chemicals AB, Malmö, Sweden), digalactosyl distearoylglyceride (DGDG, Larodan), 1,2-dipalmitoyl-sn-glycerol-3-phospho-(1'-rac-glycerol) (PtdGro, Avanti) and 1,2-dipalmitoyl-sn-glycerol-3-phosphoethanolamine (PtdEtn, Avanti).

FA compositional analysis by Gas Chromatography-flame ionization detector (GC- FID): FA composition in a given lipid extract, or a whole cell, can be analyzed after being converted to their more volatile fatty acid methyl esters (FAMEs) by GC. Most modern GC is coupled either to a mass spectrometry (GC-MS), or a flame ion detector (GC-FID), to allow FA identification and quantification, respectively. In this study, we routinely analysed FA composition via direct transmethylation of cell pellets as described in detail in Siaut et al. (Siaut et al., 2011). Briefly, 1 mL 5% (v/v) sulfuric acid in methanol and 10 μL 1% butylated hydroxytoluene (BHT) was added to a glass tube containing 20 million freshly harvested cells. C17:0 TAG (10 μg) were added as

57 an internal standard for quantification and also as a control for the efficiency of transesterification reaction. The mixture was vortexed, and heated for 90 min at 85°C. After cooling down to room temperature, NaCl (1.5 mL, 0.9% (w/v)) was added to the reaction, and lipophilic products were extracted three times with hexane. To allow easier phase separation, the mixture was then centrifuged at 3000 g for 2 min. The upper hexane phase containing FAMEs were transferred to a clean glass tube, and evaporated to dryness under a gentle N2 stream.

FAMEs were then dissolved in 200 μL hexane, transferred to a GC vial, and analyzed by GC-FID (Agilent 7890A GC and Agilent 5975C MS, Agilent Technologies) using a polar TR-WAX column (30m x 0.25mm x 0.50μm). The GC condition was: split ratio of 1:20, injector and flame ionization detector temperature 240°C; oven temperature program 50°C for 2 min, then increasing at 15°C min-1 to 150°C, and then increasing again at 6°C min-1 to 240°C and holding at this temperature for 4 min. The flow rate of

-1 the carrier gas (H2) was 1 mL min .

Lipid molecular species analysis by LC-MS/MS: Lipid molecular species analyses can provide another layer of information about changes in lipidome, which could have otherwise been overlooked if only base the analysis on lipid classes or FA compositions. Lipid molecular species analysis relies on the use of modern state-of-the-art LC- MS/MS. Another advantage of this method is its non-destructive nature, i.e. lipids were injected and passed through LC, and can also be collected for other analyses if needed.

TM So, briefly, total lipid extracts were first separated on a Kinetex C18 2.1×150 mm 2.6 μm column (Phenomenex) connected to an Ultimate RS 3000 UPLC system (Thermo Fisher). The LC system is connected to a Triple TOF 5600 mass spectrometer (AB Sciex) equipped with a duo-spray ion source. Same set of samples was run both at positive and then negative mode for quantification of polar membrane lipids and neutral lipids, respectively. A given amount of TAG(17:0/17:0/17:0) or PtdEtn(17:0/17:0) were usually added to serve as internal standards for neutral or polar lipid quantification

58 respectively. Lipid identification was based on retention time in the LC, mass accuracy peaks from the MS survey scan, and also based on fragment-ions from the MS/MS scan. This was then corrected by internal standard. Relative quantification was achieved with MultiQuant software (AB Sciex) on the basis of intensity values of extracting masses of different lipids previously identified. Lipid molecular species were noted as lipid class (total number of carbon atoms: total number of double bonds) or as lipid class (sn- 1 fatty acid/sn-2 fatty acid) when the sn position of FAs was known.

11. Chlorophyll and starch quantification Chlorophyll and starch were quantified from cells harvested from 1-mL culture (enviro 5x106 cells). Briefly, cell culture (1 mL) were collected in a 2-mL screw top Eppendorf tube, centrifuged at 13 000 g for 10 min. Supernatant was removed by decanting, and methanol (1 mL) were added to the cell pellet, and vortexed strongly to re-suspend. After storing at least 1 h at -80°C, the suspension was centrifuged again (1 min at 13000 g at 4°C), and the upper methanol phase was transferred to a cuvette for chlorophyll quantification using a UV-visible spectrophotometer (Schimadzu UV260). Absorbance (A) at 653 nm, 666 nm and 750 nm were measured and total chlorophyll was calculated by the formula: (A653-A750) x 19.71 + (A666-A750) x 4.44 (Lichtenthaler, 1987).

The remaining cell pellet was dried in a fume hood to evaporate remaining traces of methanol. The pellet was then resuspended in 400 µL of distilled water, and samples were then autoclaved for 20 min at 121°C. Two hundred µL of amyloglucosidase (1 U mL-1 - Roche) was added and incubated for 1 h at 60°C, and this process converts starch polymer to soluble sugars. Samples were then again centrifuged to remove any particulate matter. The upper phase containing glucose was analyzed with an YSI2700 select sugar analyzer (YSI Life Sciences). Commercial glucose was used as a standard for calibration.

12. Isolation and validation of mutants from the Chlamydomonas Library (CLIP) Insertional mutant lines LMJ.RY0402.153671-1 (bkdE1β), LMJ.RY0402.045578 (bkdE2-1), LMJ.RY0402.192581 (bkdE2-2) and their parental line CC5325 were ordered from the Chlamydomonoas library (https://www.chlamylibrary.org/; CLIP) (Li et al., 2016). These strains were streaked on agar plates containing paromomycin (10 µg mL-1) to obtain a single colony. Genomic DNA as well as RNA was extracted from cell cultures grown up from a single colony. To validate predicted insertional event, the following primer pairs were used: Control locus-F/Control locus-R (for control PCR), LMJ-gBKDE1β-F1/LMJ-gBKDE1β-R1 (for bkdE1β), LMJ-gBKDE2-F1/LMJ- gBKDE2-R1 (for bkdE2-1), and LMJ-gBKDE2-F2/LMJ-gBKDE2-R2 (for bkdE2-2). The KOD Xtreme™ Hot Start DNA Polymerase (Novagen) was used for all PCR amplifications unless otherwise stated.

13. Amino acid analyses by GC-MS Chlamydomonas cells (60 million) were harvested quickly by centrifugation at 13000 g for 10 s at 4°C, and then immediately decant the supernatant and pellets were frozen in liquid nitrogen. Metabolite extraction and derivatization were as described previously (Lisec et al., 2006), and then quantified using an Agilent 7683 series auto- sample (Agilent Technologies), coupled to an Agilent 6890 gas-chromatograph-Leco Pegasus two time-of-flight mass spectrometer (Leco). The chromatogram parameters and analytical part were exactly as has been reported previously in (Cuadros-Inostroza et al., 2009; Kong et al., 2018).

14. Respiration analysis using a Clark electrode The mitochondrial respiration rate was determined in the dark using a Clark-type oxygen electrode which consist of a cathode (platinum electrode) and an anode (Ag/AgCl electrode) linked by a concentrated KCl solution (Hansatech Instruments, Norfolk, U.K.). The oxygen consumption of the cathode is stoichiometrically related to the electrical current, so a change in O2 concentration by respiration can be easily

60 determined by measuring the electrical current produced. Before measurement, the instrument was first calibrated to 0 by adding sodium dithionite (Na2S2O4) which can consume all the molecular O2 present in air-saturated water; and then set up the O2 concentration value of air-saturated water as 1000. Finally, the respiration rate of cell culture was calculated based on the data recorded on the PicoScope2202 recorder software (Pico Technology, Interworld Electronics, Point Roberts, UK).

15. Total RNA extraction, RNA-Seq library preparation, and sequencing RNA was extracted from dw15 cells harvested under three different N status (+N, -N, and NR; four replicates for each condition) using the TruSeq RNA Sample Preparation Kit (Illumina). Then a cDNA library was built from 1 µg of total RNA and Illumina HiSeq 2500 sequencing were performed by the Biopuces and Sequencage platform at Illkirch, generating 27.6-38.1 million 50-nt single-end reads for each replicate. Reads were aligned using Bowtie2 software (Langmead and Salzberg, 2012) onto the Chlamydomonas reinhardtii genome assembly version V5.5 and HTSeq count (Love et al., 2014) was used to identify reads uniquely mapped on the genome.

16. Enzymatic activity assays Cells at the logarithm phase (around 20 million cells total) were collected and resuspended in extraction buffer [50 mM HEPES-KOH, 1 mM EDTA, 10% glycerol v/v, 20 µL mL-1 protein protease inhibitor cocktail for plant cells (Sigma P9599) and 5 mM DTT at pH 7.5], and then lysed by sonication three times with a 10-s interval cycle. Protein concentraion was determined spectrophotometrically at 280 nm using a BCA protein assay kit (Bio-Rad) according to the manufacturer’s instructions.

The activity of MDH in cell lysates was determined photometrically by measuring the decrease in absorbance at 340 nm resulting from the oxidation of NADH to NAD＋ as previously described (Mekhalfi et al., 2014). Enzyme assays were performed in a 1-cm path-length cuvette containing 2×assay buffer (90 mM KH2PO4-KOH pH 7.4, 0.05%

Triton X-100 v/v, 5 mM MgCl2) buffer with 0.2 mM NADH at 30°C. The reaction was initiated by addition of 3 mM OAA (Sigma-Aldrich), and the rate of NAD＋ formation was monitored at 340 nm in a Kontron Uvikon 810 spectrophotometer (Thermo scientific). All solutions used for enzyme activity assays were freshly prepared prior to use. The enzyme activities were calculated based on the molar extinction coefficient of

-1 -1 NAD(P)H at 340 nm (ε340=6220 M cm ), and the enzymatic reaction rate was calculated using a linear regression. One unit of MDH activity is defined as 1 µmol of NADH oxidized per min per mg of protein.

RESULTS

Chapter 1

Branched-Chain Amino Acid Catabolism Impacts Triacylglycerol

Homeostasis in Chlamydomonas reinhardtii

Yuanxue Lianga, Fantao Konga,, Ismael Torres-Romeroa, Adrien Burlacota, Stéphan Cuinea, Bertrand Légereta, Emmanuelle Billona, Yariv Brotmanb, Saleh Alseekhb, Alisdair R Fernieb, Fred Beissona, Gilles Peltiera, Yonghua Li-Beissona,*

a Aix-Marseille University, CEA, CNRS, Institute of Biosciences and Biotechnologies of Aix-Marseille, UMR7265, CEA Cadarache, Saint-Paul-lez Durance F-13108, France b Max Planck Institute of Molecular Plant Physiology, 14476 Potsdam-Golm, Germany

Section: Biochemistry and Metabolism Branched-Chain Amino Acid Catabolism Impacts Triacylglycerol Homeostasis in Chlamydomonas reinhardtii Yuanxue Lianga, Fantao Konga,, Ismael Torres-Romeroa, Adrien Burlacota, Stéphan Cuinea, Bertrand Légereta, Emmanuelle Billona, Yariv Brotmanb, Saleh Alseekhb, Alisdair R Fernieb, Fred Beissona, Gilles Peltiera, Yonghua Li-Beissona,*

There are no conflicts of interest. Short title (50 charac.): BCAA and TAG homeostasis in Chlamydomonas One sentence summary: The degradation of branched-chain amino acids provides carbon precursors for lipid biosynthesis during nitrogen starvation, and provide ATP for lipid remobilization upon nitrogen resupply in green algae. Author contributions: Y.L-B, G.P. and F.B. designed research; Y.L. performed all research except detailed below; F.K. mapped the insertion site for bkdE1α; I.T.R. taught how to use flow cytometry; A.B. taught the use of Clark electrode; S.C. performed immuno-blot; B.L. helped some lipid analyses; E.B. performed bioinformatic analyses; S.A., Y.B. performed metabolomics analysis under the supervision of A.R.F. Y.L-B wrote the paper with inputs from Y.L., A.R.F, F.B. and G.P. All authors approved the manuscript. Funding: This project is funded by an Amidex from Aix-Marseille University. *Correspondence to: Yonghua Li-Beisson ([email protected])

ABSTRACT Nitrogen (N)-starvation induced triacylglycerol (TAG) synthesis and its complex relationship with starch metabolism in algal cells has been intensively studied, however studies on interaction between amino acid metabolism and TAG biosynthesis are scarce. Here, via a forward genetic screen for TAG homeostasis, we isolated a Chlamydomonas mutant (bkdE1α) that is deficient in the E1α subunit of the branched-chain ketoacid dehydrogenase (BCKDH) complex. Metabolomics analysis revealed a defect in the catabolism of branched-chain amino acids (BCAAs) in bkdE1α. Furthermore, this mutant accumulated 30% less TAG than the parental strain during N starvation, and was compromised in TAG remobilization upon N resupply. Intriguingly the rate of mitochondrial respiration was 20-35% lower in bkdE1α compared to the parental strains. Three additional knock-out mutants of the other components of the BCKDH complex exhibited similar phenotypes to bkdE1α. Transcriptional responses of BCKDH to different N status were consistent with its role in TAG homeostasis. Collectively, these results indicate that BCAA catabolism is involved in TAG metabolism via the provision of carbon precursors and ATP, thus highlighting the complex interplay between distinct subcellular metabolisms for oil storage in green microalgae.

Key words: branched-chain amino acids; triacylglycerol; mitochondrial respiration; ATP; acetyl-CoA; Chlamydomonas reinhardtii

INTRODUCTION Microalgae are fast-growing microorganisms that have developed efficient mechanisms to harvest and transform solar energy into energy-rich molecules such as lipids. They are thus promising cell factories for production of fuels and biomaterials for chemical industries. However, several fundamental as well as engineering challenges need to be resolved before the establishment of a sector on algal bioenergy. A major issue is the observation that in algal cells significant oil accumulation occurs only under conditions when growth is impaired (such as nitrogen deficiency, high sality, stationary phase, or high light) (Wang et al., 2009; Moellering and Benning, 2010; Siaut et al., 2011; Urzica et al., 2013; Goold et al., 2016). To uncouple the inverse relationship between TAG synthesis and cell division (i.e. biomass growth), a deeper and holistic understanding of the pathways for fatty acid (FA) synthesis and their assembly into oil, i.e. triacylglycerols (TAGs) as well as the regulatory mechanisms involved, is required. Nitrogen (N)-starvation induced oil accumulation in algal cells has been mostly studied through -omics studies, as well as the enzymatic steps and regulations involved (Work et al., 2010; Boyle et al., 2012; Chen and Smith, 2012; Li et al., 2012; Schmollinger et al., 2014; Tsai et al., 2014; Kajikawa et al., 2015; Warakanont et al., 2015; Schulz-Raffelt et al., 2016; Kong et al., 2017). Studies on carbon and energy sources required are more scarce and have mostly focused on competition with starch accumulation for carbon precursors (Wang et al., 2009; Li et al., 2010; Work et al., 2010; Siaut et al., 2011; Krishnan et al., 2015). Increasing evidence in plants suggests the control of TAG synthesis occurs at the earlier step of de novo FA synthesis (Bourgis et al., 2011). A positive correlation between the rate of de novo FA synthesis and the amount of carbon precursors has been found in both plants and algae (Fan et al., 2012; Ramanan et al., 2013; Goodenough et al., 2014; Avidan et al., 2015). N-starved cells are known to over-accumulate acetyl-CoA prior to TAG synthesis in the green alga Chlorella desiccate (Avidan et al., 2015). It has also been observed that feeding cells with additional amount of acetate (an “acetate boost”) enhances lipid synthesis in the model microalga Chlamydomonas reinhardtii (Goodson et al., 2011; Fan et al., 2012),

66 which highlights the importance of acetyl-CoA supply as carbon source for lipid synthesis. In addition to increased de novo FA synthesis, membrane lipids are another source of acyl-chains for TAG assembly during N starvation (Fan et al., 2011). This is supported by transcriptional, biochemical and genetic evidence. First, many genes encoding lipolytic enzymes were upregulated upon N starvation (Miller et al., 2010). Secondly, the degradation of major membrane lipids occurs simultaneously to TAG accumulation (Moellering and Benning, 2010; Siaut et al., 2011). Finally, the pgd1 mutant, deficient in a major galactolipid lipase i.e. Plastid Galactoglycerolipid Degradation 1 (PGD1), made less TAG than its parental strain providing a compelling demonstration of flux of acyl-chains from plastid lipid to TAG (Li et al., 2012). Furthermore, the result gained from the study of pgd1 mutant could also indicate that de novo synthesized FAs, at least partly, first incorporated into plastid lipids before entering TAG synthesis. Besides carbon precursors, lipid synthesis requires a stoichiometric supply of ATP and reducing equivalents NAD(P)H in a ratio of 1:2 (Ohlrogge and Browse, 1995; Li-Beisson et al., 2013). The role(s) of both energetic and redox considerations in governing sub-cellular metabolism has been frequently demonstrated (Geigenberger et al., 2005; Michelet et al., 2013; Kong et al., 2018). However, little is known concerning the source(s) and variations of ATP supply on lipid synthesis. Alongside lipid and starch, amino acids (AA) are known respiratory substrates (Araujo et al., 2010; Binder, 2010; Kochevenko et al., 2012; Hildebrandt et al., 2015). Among all AAs synthesized by plants and green algae, leucine (Leu), isoleucine (Ile) and valine (Val) have in common a branched aliphatic chain and their degradation products include an acetyl-CoA, potential substrates for de novo FA synthesis (Binder, 2010). These three AAs are collectively called “branched-chain amino acids (BCAAs)”. Thus far, relationship between BCAA catabolism and lipid synthesis has been studied in mammals (Green et al., 2016) and in the model diatom Phaeodactylum triconutum (Ge et al., 2014), and more recently in Dunaliella tertiolecta where Yao et al (Yao et al.,

2017) reported acetyl-CoAs produced through BCAA catabolism contributed to TAG synthesis. However, this interaction has not been investigated at molecular level in the green lineage notably plants or green algae because the genetic basis of the mutation in the mutant strain of Dunaliella tertiolecta has not been revealed. Moreover, the connection between BCAA and lipid is currently limited to the contribution of BCAA to lipid biosynthesis, and has not been explored beyond this. In this study, we isolated and characterized several mutants of Chlamydomonas reinhardtii impaired in the BCKDH complex. We show that during N starvation, BCAA degradation provides 10-30% of the carbon precursors required for TAG synthesis. In addition, we found that TAG remobilization following N resupply is also impaired in the mutants. This study thus provides genetic and biochemical evidence that BCAA catabolism contributes significantly to both biosynthesis and turnover of TAGs in Chlamydomonas reinhardtii, most likely by supplying reduced carbon precursors and ATP in these two processes.

RESULTS

Isolation of the Pb8C12 Mutant Impaired in Both TAG Accumulation and Remobilization To understand lipid synthesis and turnover processes in Chlamydomonas, a library of >4,500 mutants were screened for significant alterations in TAG contents under different N status (Cagnon et al., 2013). The mutant library was generated via random insertion of the APH8 gene encoding paromomycin resistance in the nuclear genome of C. reinhardtii (Cagnon et al., 2013). Among 80 mutants isolated, the Pb8C12 mutant accumulated less oil than the parental strain during mixotrophic N starvation (TAP-N, 2d), and was also impaired in oil degradation following N resupply (MM+N, 1d) (Supplemental Figure 1A). By contrast, no significant changes in the content and composition of polar lipids that are major components of membranes were found in this mutant (Supplemental Figure 1B). However, a reduction in total fatty acids was

68 observed in the mutant after N starvation (Supplemental Figure 1C), which paralleled changes in TAG amounts, suggesting that the reduction in TAG resulted from a defect in the de novo fatty acid synthesis rather than from membrane lipid remodelling.

A 4

dw15 cells

6 3 - Pb8C12 ** 2 ** 1 µg TAG 10 TAG µg * 0 TAP TAP-N (1d) MM+N (1d)

B TAP TAP-N (2d) MM+N (1d) std MGDG

DGDG PG DGTS PE PC Loading origin dw15 Pb8C12 dw15 Pb8C12 dw15 Pb8C12

30 C dw15 Pb8C12

cells ** 6

- 20

10 **

µg FAMEs 10 FAMEs µg 0 TAP TAP-N (2d) MM+N (2d)

Supplemental Figure 1. Lipid analyses during mixotrophic N starvation. (A). TAG content. (B). Polar lipid analyses by TLC. (C). Total fatty acid amount. Data are means of three biological replicates with standard deviation shown. Asterisks represent statistical significance with paired Student’s t-test (* P ≤ 0.01; ** P ≤ 0.05). TLC, thin layer chromatography; std, standard.

The initial screening was carried out in mixotrophic condition (Cagnon et al., 2013)(Supplemental Figure 1), and here we show that under photoautotrophic cultivation, a similar conclusion can be drawn (Figure 1A, B). The Pb8C12 mutant accumulated 30% less oil than WT during N starvation (MM-N 2d), and it was also impaired in oil degradation following N resupply, i.e. ~80% oil remaining 1 d after N resupply compared to only ~30% remaining in the WT (Figure 1B). The delay in oil remobilization during recovery was further validated by imaging of the lipid droplets (LDs) using confocal microscopy following cell- staining with BODIPY (Figure 1C).

Figure 1. Isolation of Pb8C12 mutant in a genetic screen for TAG homeostasis. (A). Cultivation conditions used for the genetic screen. (B). TAG quantification. (C). Lipid droplet imaging. Cells were cultivated until mid-log phase photoautotrophically in the air with a supply of 2% CO2, then being starved for N for two days, followed by N resupply. Cells were harvested at each stage for analysis of triacylglycerols (TAGs) as well as for imaging. Data are means of 4 biological replicates and error bars represent standard deviation. Asterisks indicate significant difference from wild type strain dw15 using paired- sample Student’s t-test (* P ≤ 0.05; ** P ≤ 0.01). Lipid droplets were stained with 70

BODIPY. Pseudo-colors were used: chlorophyll in red; lipid droplet in green. N, nitrogen; MM, minimal medium; d, day.

Growth of the mutant cells was not affected in either mixotrophic or heterotrophic cultures (Supplemental Figure 2A, B). In photoautotrophic conditions, the mutant reached a higher density at the stationery phase than its parental strain dw15, and we subsequently demonstrated that genetic complementation recovered the growth phenotype to wild-type levels (see later results on generation of complemented lines) (Supplemental Figure 2C).

Supplemental Figure 2. Growth kinetics. (A). Mixotrophic growth. (B). Heterotrophic growth. (C). Photoautotrophic growth. Cells were grown in standard conditions. Cell concentration was monitored every day using a Coulter counter. Measurements for three biological replicates were shown (n=3, SD). C1-3, complemented lines (see Figure 2 in the text).

The Pb8C12 Mutant Contains an Insertion of the APH8 Gene in the Locus Encoding the E1α Subunit of The BCKDH Complex To understand the underlying genetic lesion in the mutant, the APH8 insertion site was identified by Restriction Enzyme Site-Directed Amplification (RESDA)-PCR (Gonzalez-Ballester et al., 2005; Kong and Li-Beisson, 2018). A primer specific to the APH8 gene, together with other degenerate primers, was used for PCR, which amplified a 705 bp fragment. DNA sequencing of this fragment identified the presence of a 421 bp DNA sequence from APH8 and 284 bp DNA sequence from the bordering region. Blast searches, using this flanking DNA sequence (284 bp) as bait and carried out within the genome of C. reinhardtii housed in Phytozome (V5.5), revealed that this sequence matched a region in the 4th intron of the locus Cre12.g539900. It was therefore assumed that the APH8 gene was inserted into Cre12.g539900 encoding a putative E1α subunit of the BCKDH protein complex (named hereafter BKDE1α) (Figure 2A). In plants as well as animals, the BCKDH complex is known to catalyze the second step in the degradation of BCAAs, i.e. the irreversible oxidative decarboxylation of α-keto acids to their respective acyl-CoAs (Binder, 2010). The knockout of this locus in the mutant Pb8C12 was confirmed by semi- quantitative RT-PCR (Figure 2C) as well as by immunoblot analysis using antibodies raised against the BKDE1α subunit (Figure 2D). Although unspecific reactions were detected with the antibodies, the occurrence of a band of ~55 kDa, corresponding to the estimated molecular weight for BKDE1α protein (without the transit peptide) only in the WT, but not in the mutant confirmed the absence of BKDE1α protein in the mutant (Figure 2D). From here on, we renamed the Pb8C12 mutant bkdE1α.

Figure 2. Mapping of insertion site and genetic complementation. (A). Locus of insertion in Pb8C12 mutant. (B). Gene construct for complementation. (C). RT-PCR. (D). Immunoblot analyses. (E). TAG quantification. Cells were cultivated until mid-log phase photoautotrophically in the air with a supply of 2% CO2. Cells were harvested for RT-PCR as well as for immunoblot. Data are means of four biological replicates and error bars represent standard deviation. Asterisks indicate significant difference from wild type strain by paired-sample Student’s t-test (* P ≤ 0.05). For immunoblot, samples were loaded at equal total protein amount and stained by Ponceau red. kDa, kilo dalton; UTR, untranslated region; RACK1, Receptor of activated protein C kinase 1. C1, 2, 3 represent three independent complemented lines.

Complementation of the bkdE1α Mutant by Nuclear Expression of BKDE1α The bkdE1α mutant was complemented using the genomic DNA of BKDE1α (5567 bp)

73 including a 1028 bp promoter region, 5’UTR (119 bp) and the coding region (Figure 2B). After screening >100 hygromycin-resistant clones, three independent lines displayed recovered transcription of the BKDE1α gene (Figure 2C), produced BKDE1α protein (Figure 2D) and contained similar amounts of oil as the parental strain dw15 both during N starvation and upon N resupply (Figure 2E).

Metabolomics Analysis Points to a Defect in BCAA Catabolism in bkdE1α Mutants of the BCAA catabolic pathway often showed difference in free AA content (Peng et al., 2015). To test if this is also the case in the bkdE1a mutant, we analysed free AAs in the mutant, its parental strain dw15 and two independent complemented lines (C1 and C2). We observed that in wild-type cells of Chlamydomonas, BCAAs represented around 20% of total free AAs (Supplemental Figure 3) and all free AAs reduced in response to N depletion in all strains (Figure 3A; Supplemental Figure 4). The static level of free AAs correlated negatively with cellular TAG amount under varying N status, i.e. when TAG content is higher, free amino acids are lower and vice versa (Figure 3B), further suggesting that BCAAs could provide substrates for TAG synthesis.

Supplemental Figure 3. Free amino acid composition in the parental strain dw15. Cells were cultivated until mid-log phase photoautotrophically in the air with a supply of 2% CO2. Data are means of 4 biological replicates.

Supplemental Figure 4. Free amino acids before and after N starvation. Cells were cultivated until mid-log phase photoautotrophically in the air with a supply of 2% CO2. Data are means of 4 biological replicates and error bars represent standard deviation. Asterisks indicate significant difference from control strains by paired- sample Student’s t-test (* P ≤ 0.05). C1 and C2, two complemented lines.

Figure 3. BCAA remobilization is impaired in the bkdE1a mutant. (A). Free AA analyses by GC-MS. (B). BCAAs are in inverse relation to TAG content. (C). Percentage of BCAAs remained upon N starvation. Cells were cultivated until mid-log phase photoautotrophically in the air with a supply of 2% CO2. Data are means of 4 biological replicates and error bars represent standard deviation. Asterisks indicate significant difference from control strains by paired- sample Student’s t-test (* P ≤ 0.05; ** P ≤ 0.01). A.U. arbitrary unit. C1 and C2, two complemented lines. BCAA, branched-chain amino acid.

Interestingly, the bkdE1α mutant made 20-30% less BCAAs during optimal growth, and the level of BCAAs reached similar levels as that of the parental strain as well as complemented lines upon N depletion (Supplemental Figure 4). When the capacity to remobilize BCAAs upon N starvation was calculated as a percentage of the remaining BCAA, we could see clearly that 1 d after N starvation, the bkdE1α mutant kept significantly higher proportion of its BCAAs compared to the control strains (parental

76 strain and complemented lines). This suggests a defect in the remobilization of BCAAs in the mutant (Figure 3C). Taken together, these data confirm that BCKDH is involved in BCAA catabolism in Chlamydomonas reinhardtii.

Comparative RNA-seq Analysis Revealed the Activation of BCAA Catabolic Pathway during N Starvation and upon N Resupply In higher plants and animals the breakdown of BCAAs is known to occur in the mitochondria (Taylor et al., 2004; Angelovici et al., 2013). Major steps and proteins involved in the degradation of BCAAs have recently been described in the model plant Arabidopsis thaliana (Peng et al., 2015) as well as in tomato (Kochevenko and Fernie, 2011; Kochevenko et al., 2012). Homology search of proteins similar to the known Arabidopsis ones have identified similar sets of proteins for BCAA degradation in Chlamydomonas (Table 1). Compared to Arabidopsis, fewer proteins are present. For example, only two genes encoding BCKDH E1 subunits are present, whereas Arabidopsis possesses four such genes. Except the five putative branched-chain aminotransferase (BCATs), most other components of the pathway are predicted mitochondrial based on PredAlgo (Tardif et al., 2012) as well as TargetP (Emanuelsson et al., 2000). BCATs are predicted to be either chloroplast, mitochondria or other, consistent with a dual role of these enzymes acting either at the last step of AA synthesis or at the first step of AA degradation (Binder et al., 2007; Angelovici et al., 2013). To quantify changes in expression of genes of BCAA catabolism in response to N starvation and upon N resupply, we performed a comparative RNA-seq experiment in the parental strain dw15. Total RNAs were extracted from dw15 cells harvested during normal growth (+N), then 1 d after N starvation (-N) and 1 d after N resupply (NR). For each of the four biological replicates, we mapped 22.4-30.5 million reads to C. reinhardtii genome. The raw data was subjected to Principal Component Analysis (PCA) (Supplemental Figure 5), which indicates a very good overall quality for the data obtained by RNA-seq. DESeq2 package (Love et al., 2014) was used for the analysis of differential expression of the genes. Only genes with a p value adjusted

(p≤0.05) and an absolute log2 fold change (log2FC)≥1 were kept for further analysis. Over 19,000 transcripts were detected and annotated. Mapping and gene annotation revealed that N starvation or resupply impacts massively genome transcription and that many metabolic pathways are affected (Supplemental Table 1A). Since comparative transcriptome analyses of cells before and after N starvation have been previously published (Miller et al., 2010; Schmollinger et al., 2014; Tsai et al., 2014; Tsai et al., 2017), we focused our analyses here on the changes in gene expression levels for genes of BCAA catabolism.

Supplemental Figure 5. Data quality analysis for RNAseq. Principal component analysis (PCA) plot of the RNAseq data. The PCA plot shows the variance of the four experimental replicates of each of the 3 different samples. The percentages on each axis represent the percentages of variation explained by the principal components.

Interestingly, the BKDE1α gene was upregulated (log2FC>3) in response to N starvation, as well as during oil remobilisation phase i.e. upon N resupply (log2FC>3) (Table 1). RNA-seq analyses of all putative genes of the BCAA degradation pathway indicate that, except the E3 subunit (encoded by Cre18.g749847) of BCKDH and the five putative BCATs, all the others showed a similar up-regulation as BKDE1α both under N starvation as well as upon N resupply (Table 1). Most of them have also been 78 found to be upregulated in the night compared to the day in a diurnal cycle (Zones et al., 2015). The collective transcriptional response of the BCAA catabolic genes points to a common function of this catabolic pathway under nutrient stress, in the dark, or during carbon starvation, supporting the buffering role of BCAAs in supplying an alternative energy source when carbon or N levels are low.

Mutants Defected in Other Components of the BCKDH Complex were also Impaired in TAG Homeostasis The BCKDH complex consists of three catalytic components: a heterotrimeric branched-chain α (β)-ketoacid dehydrogenase (E1: EC1.2.4.4), a dihydrolipoyl transacylase (E2: EC2.3.1.168), and a dihydrolipoamide dehydrogenase (E3: EC1.8.1.4) (Figure 4A). To further validate the link between the BCKDH complex and TAG homeostasis, we identified insertional mutants for E1β and E2 subunits from the

Chlamydomonas library (CLiP, https://www.chlamylibrary.org/) (Li et al., 2016), i.e. LMJ.RY0402.153671-1 for the BKDE1β subunit, LMJ.RY0402.045578 and LMJ.RY0402.192581 for the BKDE2 subunit (Figure 4B). The insertional event in each mutant was confirmed by PCR following the protocol given by the CLiP homepage (Supplemental Figure 6), then the knock-out of the corresponding gene was confirmed by RT-PCR analyses (Figure 4C). These mutants were thus named as bkdE1β for LMJ.RY0402.153671-1, bkdE2-1 for LMJ.RY0402.045578, and bkdE2-2 for LMJ.RY0402.192581, respectively. TAG quantification revealed that all three mutants made ~10% less oil than their parental strain during N starvation as well as showed a defect in oil remobilization following N resupply (Figure 4D). The reduction in oil content (~10%) is smaller compared to what was observed in the bkdE1α mutant (~30%) (Figure 1). Various factors could explain such a difference including the difference in the parental strain or in the contribution of these different subunits to activity of the BCKDH complex. Nevertheless, these data collectively point to a contribution of the BCKDH complex in TAG homeostasis in C. reinhardtii.

Figure 4. Mutants deficient in other components of the BCKDH complex are also impaired in TAG homeostasis. (A). A model for the BCKDH protein complex and putative reactions associated to it. (B). The site of insertion(s) of APH8 in each individual genome. (C). RT-PCR. (D). TAG quantification. Cells were cultivated until mid-log phase photoautotrophically in the air with a supply of 2% CO2. Data are means of 4 biological replicates and error bars represent standard deviation. Asterisks indicate significant difference from control strains by paired- sample Student’s t-test (* P ≤ 0.05; ** P ≤ 0.01). BCAT, branched-chain amino transferase; BCKDH, branched-chain keto-acid dehydrogenase.

Mitochondrial Respiration is Reduced in Mutants of the BCKDH Complex BCAAs can serve as substrates for respiration, and to evaluate the impact on mitochondrial respiration, the rate of oxygen (O2) consumption in the dark was measured. The dark O2 consumption rate was 20-35% lower in the bkdE1α mutant than in the control strains (dw15, and the two complemented lines C1, C2) before and after N starvation as well as upon N resupply (Figure 5). A similar defect in mitochondrial respiration was also observed in the other three mutants defected in the distinct subunits

80 of the BCKDH complex i.e. the BKDE1β and BKDE2, respectively (Supplemental Figure 7). This data supports the conclusion that a defect in BCAA catabolism affects mitochondrial respiration, providing evidence that BCAAs as respiratory substrates.

Figure 5. Mitochondrial respiration analyses of the bkdE1α mutant. Cells were cultivated until mid-log phase photoautotrophically in the air with a supply of 2% CO2, then starved for N. Data are means of 4 biological replicates and error bars represent standard deviation. Asterisks indicate significant difference from control strains by paired-sample Student’s t-test (* P ≤ 0.05; ** P ≤ 0.01). C1 and C2 refer to three complemented lines.

DISCUSSION Due to the promise microalgae have as a next generation platform for bio-products, various ways to increase energy density in algal biomass have been explored in the past 10 years. Among which, N-starvation is one of the most potent triggers for initiating TAG accumulation, therefore N starvation response has been investigated at multiple levels (Miller et al., 2010; Moellering and Benning, 2010; Siaut et al., 2011; Schmollinger et al., 2014). The outcome of this research is that in response to N

81 starvation, significant metabolic changes occur: a massive accumulation of oil and starch, increased membrane lipid remodelling, a decrease in photosynthesis and photosynthetic pigments and an increased rate of protein degradation and autophagy (Miller et al., 2010; Siaut et al., 2011; Schmollinger et al., 2014; Couso et al., 2018) (Davey et al., 2014). Whilst the effect of some pathways (notably starch) on oil accumulation has been studied, the interaction between AA metabolism and TAG synthesis remain less clear. In the current study, we demonstrate that BCAAs contribute significantly to TAG homeostasis. We further show that through its close interaction with the electron transport chain in mitochondria, BCAA catabolism also interacts with lipid degradation through the contribution of ATP for the initial steps of lipid catabolism. Thus several potential metabolic interactions between BCAA catabolism and fatty acid synthesis and lipid degradation are drawn and we propose that BCAA degradation provides not only carbon sources but also ATP for de novo fatty acid synthesis (Figure 6).

Figure 6. A model explains the interaction between BCAA catabolism and TAG homeostasis in Chlamydomonas. Blue lines indicate potential flow of reducing equivalents; red lines indicate flow of ATP. When N-starved, cells deplete their free amino acids pool. The degradation of some of these free amino acids, namely BCAAs, produces acetyl-CoA as well as NADH which is fed into the mitochondrial respiratory chain. The production of acetyl-CoA and ATP from BCAA catabolism contributes to TAG synthesis during N starvation. In addition, upon N resupply in the dark, the production of ATP through the BCAA catabolic pathway become essential to initiate the β-oxidation of fatty acids as its earlier steps require 2 molecules of ATP, i.e. the transport of free fatty acids into peroxisome is an ATP-requiring reaction (catalyzed by the ABC transporter called comatose - CTS); and secondly, the activation of free fatty acids to their CoA esters is an ATP-dependent reaction catalyzed by long chain acyl-CoA synthetase (LACS). FAS, fatty acid synthase; TAG, triacylglycerols; BCAA, branched-chain amino acid; TCA, tricarboxylic acid cycle; ETC, electron transport chain.

BCAA Degradation and Lipid Biosynthesis The bkdE1α mutant made 30% less TAG than control strains during N starvation (Figure 1 and 2), implicating BCAA acts as a carbon or energy source for fatty acid

83 synthesis. A connection between BCAA catabolism and lipid synthesis has been evidenced recently in diatoms. Ge et al found that in parallel to the onset of massive oil accumulation in the model diatom Phaeodactylum triconumtum, proteins of BCAA catabolic pathways showed >2 fold upregulation (Ge et al., 2014), moreover, silencing the gene encoding the methylcrotonyl-CoA carboxylase (MCC2) resulted in strains making 27-43% less neutral lipids. This study suggests that BCAAs could act as substrates for TAG biosynthesis during N starvation in diatoms. Moreover, the occurrence of a relationship between BCAA catabolism and lipid synthesis is not limited to algae, and has also been observed in mammalian cell lines. For example, it was found that inhibition of BCAA degradation in mouse cell lines decreased oil content, and the authors further report that BCAA contributed up to 30% acetyl-CoA required for lipogenesis in adipocytes (Green et al., 2016). For this reason, levels of BCAAs in mammalian cells have been used as a marker for obesity (Newgard; Crown et al., 2016; Green et al., 2016). Taken together, these studies point out the existence of a route for carbon flux from BCAAs to TAG in organisms of diverse evolutionary origin. Nevertheless, in the current literature, the metabolic interaction between BCAA catabolism and TAG accumulation has not been explored beyond the contribution of carbon skeletons such as acetyl-CoAs. In addition to acetyl-CoA esters, the reaction catalyzed by the BCKDH complex produces NADH, which is believed to be fed to the internal alternative NADH dehydrogenase of complex I of the mitochondrial respiratory chain (Schertl and Braun, 2014), thereby influencing ATP production. Therefore besides being a source of reduced carbon precursors, BCAAs are also respiratory substrates and their degradation are tightly connected to the mitochondrial respiratory chain. The impairment in the mitochondrial respiration of the bkdE1a mutant as well as the other mutants of the BCKDH complex provides clear evidence for this in algae (Figure 5). Taken together, BCAA catabolism in mitochondria provides carbon precursors for lipid synthesis, and in parallel, it provides direct electron donors to the mitochondrial electron transport chain by so doing generating ATP. Acetyl-CoAs are

84 essential building blocks for fatty acid synthesis, and ATP generated through mitochondrial respiration is an important source to sustain lipid synthesis, especially during stress conditions when the ATP production by the chloroplast photosynthesis is likely repressed (Schmollinger et al., 2014).

BCAA Degradation and Lipid Catabolism In mammalian cells, since fatty acid β-oxidation occur essentially in mitochondria (Eaton et al., 1996), the metabolic link between storage lipid degradation and mitochondrial activity has been well established (Kujala et al., 2016). Based on our analysis and also based on literature, BCAA degradation generally occur in mitochondria, whereas the major steps for β-oxidation of fatty acids occur in peroxisome of Chlamydomonas (Kong et al., 2017; Kong et al., 2018; Kong et al., 2018). This raises an interesting question namely how a defect in BCAA catabolism can affect lipid degradation. Reduction in mitochondrial respiration observed in the bkdE1α mutant probably provides an explaination (Figure 5). Thus it seems reasonable to postulate that less ATP is produced in the BCKDH-deficient mutants upon N resupply in the dark. The initial steps of β-oxidation of fatty acids require two ATPs both to fuel the long chain acyl-CoA synthetase (LACS) (Fulda et al., 2004; Kong et al., 2018), as well as to energize fatty acid import into the peroxisome via an ATP-binding cassette transporter (CTS), the plant homolog of which is known to transfer fatty acids from cytosol into peroxisomes (Footitt et al., 2002). We thus propose that when mitochondrial respiration is impaired in the bkdE1α mutant, less ATP is produced, consequently fatty acid activation and import into peroxisomes is reduced thereby impacting TAG breakdown (Figure 6).

Defects in BCKDH Complex on AA Metabolism and Cell Physiology The level of free AAs depends on tissue types, developmental stage, or environmental conditions. In this study, we showed that AAs degrade massively during N-starvation when TAGs accumulate (Figure 3). This is consistent with the upregulation of most

85 genes of the BCAA catabolic pathway during N starvation reponse (Table 1). This also correlates with an increase in most enzymes of BCAA catabolism in a quantitative proteomics study of Phaeodactylum tricornutum during N starvation response (Ge et al., 2014). It is worth noting here that metabolic changes were not limited to BCAAs in the bkdE1α mutant, and the level of a few other AAs also altered (Supplemental Figure 4). The perturbation in BCAA catabolism has been observed to perturb the levels of other AAs in Arabidopsis thaliana (Peng et al., 2015) and in diatoms (Ge et al., 2014), which suggests a tight co-ordination in the synthesis and degradation of different types of amino acids. Blocking BCAA catabolism in higher plants has resulted in overaccumulation of BCAAs (Peng et al., 2015); however, this seems not to be the case in Chlamydomonas (this study). Many reasons could explain this discrepancy. It could be due to differences in carbon and energy metabolism in the green algae compared to plants, as has been witnessed before (Grossman et al., 2007; Liu and Benning, 2013). This could also be a result of the choice of sampling time point, and in this study, we analyzed only cells 2-d after N starvation (Figure 3). Indeed, leucine over-accumulation was observed in MCC2-silenced diatom cells after 10-d N starvation, but not prior to this time point (Ge et al., 2014). Intriguingly, we observed that the level of BCAAs as well as several other AAs is low compared to WT (dw15) during optimal growth in the presence of nitrogen (Figure 3; Supplemental Figure 4). How a defect in BCAA catabolism has altered cellular levels of BCAAs or their biosyntheses is currently unknown, and require further investigation through the use of carbon isotope flux analysis. During optimal growth, we observed that the bkdE1α mutant reached much higher cell density at the stationary phase than the parental strain dw15 as well as the two complemented strains (C1 and C2) (Supplemental Figure 2). This contrasts with the lower cell density observed in strains silenced in the MCC2 gene in diatom (Ge et al., 2014), reflecting the divergence in the evolutionary origin of these two groups of algae. Besides playing structural, storage and respiratory roles, BCAAs are signalling molecules linking carbon to energy

86 and to nitrogen metabolism (Araujo et al., 2010; Binder, 2010; Kochevenko et al., 2012; Hildebrandt et al., 2015), and the differences in growth behaviour could be related to the use and uptake of nitrogen. Indeed, MCC2-silenced Phaeodactylum strains showed incomplete use of nitrogen from the medium. Further work is required to address this and other questions in the green alga Chlamydomonas rienhardtii.

MATERIALS AND METHODS

Strains, Mutant Isolation and Culture Conditions Chlamydomonas reinhardtii cell wall-less strain dw15.1 (nit1-305 cw15; mt+) was the parental strain to generate the mutant library from which the mutant Pb8C12 (bkdE1α) was derived (Cagnon et al., 2013). The strains were maintained on Tris-Acetate- Phosphate (TAP) agar plates supplemented with 10 µg mL-1 paromomycin at 25°C under continuous illumination (about 30 μmol photons m-2s-1). Liquid cultures were kept in incubation shakers (INFORS HT), at 25°C, 120 rpm shaking, with 75 μmol photons m-2s-1 illumination. Cells were cultivated either mixotrophically (TAP) or photoautotrophically in minimal media (MM) with an addition of 2% CO2 in the air, as specified in the text. To induce TAG accumulation, media without nitrogen (TAP-N, or MM-N) were used. For TAG remobilization, N was added back to the cell culture while cells were being kept in the dark. Therefore a routine analysis of TAG homeostasis consists of: first, cells were cultivated in TAP or MM medium with addition of 2% CO2 in the air, transferred to TAP-N or MM-N media (for 2 d) and then changed to a minimal medium (MM) that contains N but without acetate for another couple of days in the dark (Figure 1A). Cell concentration was followed using a Beckman Coulter Multisizer 4 (MultisizerTM3 Coulter Counter, Beckman Coulter, USA).

Isolation of Genomic DNA and RNA, and cDNA Preparation Genomic DNA (gDNA) or RNA were extracted from exponentially grown Chlamydomonas cells following the protocol of (Tolleter et al., 2011), or (Nguyen et

87 al., 2013), respectively. Extracted total RNA was treated with the TURBOTM DNase (Life technologies) to remove any genomic DNA contamination. To obtain highly pure RNA, the total extracts were passed through a NucleoSpin RNA Clean-up column (Macherey-Nagel) following the manufactures’ instructions. The first strand cDNA was made using the SuperScript VILO cDNA Synthesis Kit (ThermoFisher Scientific).

Identification of the APH8 Insertion Site in the Pb8C12 Mutant by RESDA-PCR The insertion site of APH8 cassette in the Pb8C12 mutant was identified by the RESDA-PCR method (Gonzalez-Ballester et al., 2005) and also in Bioprotocols (Kong and Li-Beisson, 2018). Primers and PCR conditions were previously described in Kong et al (Kong et al., 2017). After Blast searches of the Chlamydomonas genome (v5.5, Phytozome), the insertion cassette was found located in the 4th intron of the locus Cre12.g539900 which encodes the E1α subunit of a large enzymatic complex named BCKDH. The KOD Xtreme™ Hot Start DNA Polymerase (Novagen) was used for all PCR amplifications unless otherwise stated. All primer sequences used in this study are provided in Supplemental Table 2.

Isolation and Validation of Mutants from the Chlamydomonas Library (CLIP) Insertional mutant lines LMJ.RY0402.153671-1 (bkdE1β), LMJ.RY0402.045578 (bkdE2-1), LMJ.RY0402.192581 (bkdE2-2) and their parental line CC5325 were ordered from the Chlamydomonas library (https://www.chlamylibrary.org/; CLIP) (Li et al., 2016). These strains were streaked on agar plates containing paromomycin (10 µg mL- 1) to obtain a single colony. Genomic DNA as well as RNA were extracted from cell cultures inoculated from a single colony. To validate predicted insertional event, the following primer pairs were used: Control locus-F/Control locus-R (for control PCR), LMJ-gBKDE1β-F1/LMJ-gBKDE1β-R1 (for bkdE1β), LMJ-gBKDE2-F1/LMJ- gBKDE2-R1 (for bkdE2-1), and LMJ-gBKDE2-F2/LMJ-gBKDE2-R2 (for bkdE2-2).

Reverse Transcription PCR (RT-PCR) Expression level of the different BCKDH subunits (BKDE1α, BKDE1β and BKDE2) was evaluated by semi-quantitative RT-PCR. The RACK1 (Cre06.g278222) was used as a house keeping gene. Primer pairs used for RT-PCR analyses were CBLP-F/CBLP- R (for RACK1), BCKDH-F/BCKDH -R (for BKDE1α), BKDE1β-F/BKDE1β-R (for BKDE1β) and BKDE2-F/BKDE2-R (for BKDE2).

Cloning the Full Length BKDE1α and Genetic Complementation of bkdE1α Mutant The BKDE1α gDNA (Cre12.g539900) and its native promoter region (1028 bp upstream the 5’ untranslated region) together with 3’UTR of RbcS2 were cloned into the PAP22 vector (which was obtained by inserting a Hygromycin gene to the blunt end of the PCR-blunt II-TOPO vector). Due to the presence of some complicated secondary structures around this region and also due to their high GC content (67%), we cloned the full-length sequence by two over-lapping fragments: i. the first fragment, containing the promoter region until the end of the 4th exon, was amplified using the primer pair (CrBCKDH-XbaI-FP3/CrBCKDH-R6); ii. The second fragment was cloned from the beginning of the 4th exon to the stop codon using the primer pair CrBCKDH- F6/CrBCKDH-NdeI-R2. These two fragments were then ligated into one single fragment by PCR. Briefly, the above two fragments were purified and mixed in an equal molar ratio (500 ng each) to serve as a template for PCR amplifications. The PCR reaction was carried out following KOD Xtreme™ Hot Start DNA Polymerase (Novagen) protocol without primers [94°C for 2 min, then 11× (98°C for 10 s, 74°C for 3 min 10 s) and finally 1 cycle at 74°C for 5 min]. Then, two µL of above reaction were used as template to allow PCR amplification using primers CrBCKDH-XbaI-FP3 and CrBCKDH-NdeI-R2. The entire gene sequence was then confirmed by DNA sequencing (at GATC® Ltd). Finally, the full-length sequence (5349 bp), together with the 3’UTR of RbcS2 [from pChalmy_4 vector (Life technologies) using RbcS3Ter- NdeI and RbcS2-NotI primers], was ligated to the PAP22 vector (digested by XbaI and

NotI). The linearized vector cut with XmaI was electroporated to the bkdE1α mutant cells, and antibiotic resistant clones were selected on agar plates containing 17 µg mL- 1 hygromycin.

The Use of Flow Cytometry to Screen for Complemented Lines Cellular oil content of independent transformants was first screened using Flow cytometry after cells being stained with BODIPY 505/515 (D-3921) (4,4-difluoro- 1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-s-indacene, Thermo Scientific, USA)(Mou et al., 2012). BODIPY was added to cell cultures at a final concentration of 0.25 µg mL-1 (from a stock solution of 50 μg mL-1 BODIPY solubilized in dimethyl sulfoxide, DMSO) and incubated in the dark for 5 min at room temperature before analyses. Neutral lipids stained with BODIPY show an emission peak at 521 nm when excited by 488 nm laser line. Cultivation conditions for 96-well plate cultures used were the same as detailed in (Cagnon et al., 2013). Clones showing recovered oil content as wild-type strain based on Flow cytometry analyses were then confirmed by TAG quantification using Thin Layer Chromatograph (TLC).

Lipid Extraction and Quantification Total lipids were extracted from exponentially-grown Chlamydomonas cells using a modified method of Bligh and Dyer (Siaut et al., 2011). TAG amount as well as amount of other polar lipids were quantified using the high performance-thin layer chromatography (HP-TLC; CAMAG, Switzerland). Briefly, a given amount of lipid extracts were deposited on an 10x20 cm silica gel 60 F254 TLC plate (Merck KGA, Germany) using an ATS5 automatic TLC sampler. The TLC plate was developed either in a mixture of hexane/diethyl ether/acetic acid (17/3/0.2, v/v/v) for neutral lipid separation, or developed in a mixture of acetone/toluene/water (91/30/8, v/v/v) for polar lipid separation. After development, the plates were air dried, and lipids were revealed after dipping in a mixture of (20 g CuSO4 and 80 mL H3PO4) and heated at 170°C for 20 min. The lipid standards used were triheptadecanoin (C17:0 TAG, Sigma-Aldrich,

Saint-Louis, USA), monogalactosyl distearoylglyceride (MGDG, Larodan Fine Chemicals AB, Malmö, Sweden), digalactosyl distearoylglyceride (DGDG, Larodan), 1,2-dipalmitoyl-sn-glycerol-3-phospho-(1'-rac-glycerol) (PtdGro, Avanti) and 1,2- dipalmitoyl-sn-glycerol-3-phosphoethanolamine (PtdEtn, Avanti).

Fatty Acid Compositional Analysis by Gas Chromatography (GC) Fatty acid composition in a given lipid extract, or a whole cell, can be analyzed by GC after being converted to their more volatile fatty acid methyl esters (FAMEs). In this study, we routinely analyzed fatty acid composition via direct transmethylation of cell pellets. Briefly, one mL 5% (v/v) sulfuric acid in methanol and 10 μL 1% butylated hydroxytoluene (BHT) were added to a glass tube containing 20 million freshly harvested cells. C17:0 TAG (10 μg) were added as an internal standard for quantification and as a control for the efficiency of transesterification reaction. The mixture was vortexed, and heated for 90 min at 85°C. After cooling down to room temperature, 1.5 mL 0.9% (w/v) NaCl was added to the reaction, and lipophilic products were extracted three times with hexane. To allow easier phase separation, the mixture was then centrifuged at 3000 g for 2 min. The upper hexane phase containing FAMEs were transferred to a clean tube, and evaporated to dryness under a gentle nitrogen stream. The FAMEs were then dissolved in 200 μL hexane, transferred to a GC vial, and analyzed by a GC-FID apparatus (Agilent 7890A GC and Agilent 5975C MS, Agilent technologies) using a polar TR-WAX column (30 m x 0.25 mm x 0.50 μm). The GC conditions were: split ratio of 1:20, injector and flame ionization detector temperature 240°C; oven temperature program 50°C for 2 min, then increasing at 15°C min-1 to 150°C, and then increasing again at 6°C min-1 to 240°C and holding at this

-1 temperature for 4 min. The flow rate of the carrier gas (H2) was 1 mL min .

LD Imaging To observe LDs, Chlamydomonas cells were first enriched via a gentle centrifugation (850 g, 3 min), then re-suspended in a small amount (100-200 µL) of fresh media with

0.25% glutaraldehyde. Cells were stained with BODIPY (to a final concentration of 0.25 µg mL-1, from a stock of 50 µg mL-1 in DMSO) in the dark for 5 min. Cells were then excited with a laser line at 488 nm, and emission was collected between 500-545 nm for BODIPY and between 650-730 nm for chlorophyll autofluorescence. All were carried out under a 63xoil immersion objective in a Confocal laser scanning microscope (TCS SP2, Leica, Germany). Pseudo colors were applied using the ZEN (Carl Zeiss) software.

Antibody Generation and Immunoblot Analysis Polyclonal antibodies against the synthetic peptide (AA380-394: Cys- SGGLLTEPAVGAVGK) were immunized in two rabbits (ProteoGenix, SAS, Schiltigheim, France). For immunoblot analysis, total protein extracts (10 µg) were separated on 10% Bis-Tris gel using MES running buffer, transferred to BioTrace NT nitrocellulose membrane (Sigma-Aldrich) and immunoblotted with specific primary polyclonal antibodies (1/500) from rabbit, at room temperature for two hours. This was followed by incubation with a secondary anti-rabbit antibody (Invitrogen) for 1 h. Immobilon™ Western Chemiluminescent HRP (horseradish peroxidase) Substrate (EMD Millipore) was used for the detection and images were recorded using a G:BOX Chemi XL (Syngene).

Amino Acid Analyses by GC-MS Chlamydomonas cells (60 million) were harvested quickly by centrifugation at 13000 g for 10 s at 4°C, and then immediately frozen in liquid nitrogen. Metabolite extraction and derivatization were as described previously (Lisec et al., 2006), and then quantified using an Agilent 7683 series auto-sample (Agilent Technologies), coupled to an Agilent 6890 gas-chromatograph-Leco Pegasus two time-of-flight mass spectrometer (Leco). The chromatogram parameters and analytical part were exactly as has been reported previously in (Cuadros-Inostroza et al., 2009; Kong et al., 2018).

Respiration Analysis using a Clark Electrode The mitochondrial respiration rate was determined in the dark using a Clark-type oxygen electrode which consist of a cathode (platinum electrode) and an anode (Ag/AgCl electrode) linked by a concentrated KCl solution (Hansatech Instruments, Norfolk, U.K.). The oxygen consumption of the cathode is stoichiometrically related to the electrical current, so a change in the O2 concentration by respiration can be easily determined by measuring the electrical current produced. Before measurement, the instrument was first calibrated to 0 by adding sodium dithionite (Na2S2O4) which can consume all the molecular O2 present in the air-saturated water; and then set up the O2 concentration value of air-saturated water as 1000. Finally, the respiration rate of cell culture was calculated based on the data recorded on the PicoScope2202 recorder software (Pico Technology, Interworld Electronics, Point Roberts, UK).

Total RNA Extraction, RNA-Seq Library Preparation, and Sequencing RNA was extracted from dw15 cells harvested under three different N status (+N, -N, and NR; four replicates for each condition) using the TruSeq RNA Sample Preparation Kit (Illumina). Then a cDNA library was built from 1 µg of total RNA and Illumina HiSeq 2500 sequencing were performed by the Biopuces and Sequencage platform at Illkirch, generating 27.6-38.1 million 50-nt single-end reads for each replicate (see Supplemental Table 1B). Reads were aligned using Bowtie2 software (Langmead and Salzberg, 2012) onto the Chlamydomonas reinhardtii genome assembly version V5.5 and HTSeq count (Love et al., 2014) was used to identify reads uniquely mapped on the genome.

ACKNOWLEDGEMENTS We thank Pascaline Auroy, Stephanie Blangy and Audrey Beyly for technical help. This project is funded by an Amidex from Aix-Marseille University. Y.L. thanks China Scholarship Council (CSC) for a PhD student stipend. I.T.R. and A.B. acknowledge the CEA for an international PhD studentship (Irtelis). We also acknowledge the European

Union Regional Developing Fund (ERDF), the Région Provence Alpes Côte d’Azur, the French Ministry of Research and the CEA for funding the HelioBiotec platform. S.A. and A.R.F. thank the European Union for funding in the framework of the European Union 2020 TEAMING Project (SGA-CSA No 664621 and No 739582 under FPA No. 664620).

TABLES Table 1. Putative proteins involved in BCAA catabolism in Arabidopsis and Chlamydomonas, their predicted subcellular localization and transcriptional responses. PredAlgo was performed as explained in the paper by Tardif et al (Tardif et al., 2012). Abbreviations: N, night; D, day; N, nitrogen; NR, N recovery; C, chloroplast; M, mitochondria; O, other; -, not available. *data are based on Zones et al (Zones et al., 2015).

Log2F Pre Phytozome C Log2FC Target Annotation Gene Tair ID N/D* dAl ID (- (NR/+N) P go N/+N) Cre02.g081 BCAT1 At1g10060 0.6 -0.2 0.6 C C 400 Cre05.g245 BCAT2 At1g10070 -0.9 -0.9 1.0 C M

900 Cre10.g458 Branched-chain BCAT3 At3g49680 1.2 -0.9 0.2 O O 050 aminotransferas Cre13.g576 e BCAT4 At3g19710 -0.9 0.0 1.8 C C 400 Cre02.g110 BCAT5 At5g65780 -1.7 -1.8 0.9 O O 700 BCAT6 At1g50110 ------BCAT7 At1g50090 ------BCKD ------α subunit of At1g21400 H E1A1 branched-chain BCKD ketoacid H Cre12.g539 dehydrogenase At5g09300 3.3 4.3 4.7 M M

E1A2 900 E1 (BKDE 94

1α) β subunit of BCKD Cre06.g311 At1g55510 3.4 3.5 9.0 M M branched-chain H E1B1 050 ketoacid ------BCKD dehydrogenase At3g13450 H E1B2 E1 Branched-chain ketoacid BCKD Cre04.g228 At3g06850 1.6 2.0 8.8 M M dehydrogenase H E2 350 E2 BCKD - - H E3 Cre18.g749 At1g48030 0.8 M M

Branched-chain mtLPD 847 ketoacid 1 dehydrogenase BCKD ------E3 H E3 At3g17240 mtLPD 2 Isovaleryl-CoA Cre06.g296 IVD1 At3g45300 2.1 1.9 7.8 M M dehydrogenase 400 α subunit of 3- methylcrotonyl MCCA Cre06.g278 At1g03090 2.3 2.9 12.7 O M

-CoA 1 098 carboxylase β subunit of 3- methylcrotonyl MCCB Cre03.g181 At4g34030 2.3 2.7 4.6 M C

-CoA 1 200 carboxylase Hydroxymethyl Cre12.g485 glutaryl-CoA HML1 At2g26800 - -0.8 1.1 M M

550 lyase

SUPPLEMENTAL TABLES Supplemental Table 1. RNA-seq analyses. (available upon request) (A) Change in transcriptomes in cells from normal growth, N starvation and N resupplied conditions. (B). Bowtie and HTSeq count results with the percentage of reads uniquely mapped on the Chlamydomonas reinhardtii's genome v5.5. +N, during normal growth; -N, N starvation; NR, N recovery. A, B, C, D represent four biological replicates.

Supplemental Table 2. Primers used in this study.

Purpose Primer name Sequence RT-PCR CBLP-F GAGTCCAACTACGGCTACGCC CBLP-R CTCGCCAATGGTGTACTTGCAC BCKDH-F CGCTTCTCCTTCTACCTGACCTGC BCKDH-R CGCTCACCCTTGTACTGCTCGT CrBCKDH-XbaI-FP3 AATTGGGCCCTCTAGACCCGCACCCGCATCTGCAA CrBCKDH-R6 CCCACCTGGTTGGCGAACTG CrBCKDH-F6 CGCAGGTGTTTGCGCAGTAC CrBCKDH-NdeI-R2 AATCATATGTCACCGCAGTGGCACGTCGG

RbcS3Ter-NdeI TGCGGTGACATATGATTCCGCTCCGTGT RbcS2-NotI ATCACACTGGCGGCCGCCGCTTCAAATACGCCCAGCC CrBCKDH-TP-NdeI AGCCCATATGCATTCACCAGCGCTTG

localization CrBCKDH-TP- CCATGAATTCGCACAGTAGCTCGCTCAAGC 20AA-EcoRI

Complementation and subcellular 20AA-EcoRI-GFP-F GTGCGAATTCATGGCCAAGGGCGAGGAGCTG GFP-XbaI-R TGCCATCTAGATTACTTGTACAGCTCGTCCATGCC Control locus-F ATGCTTCTCTGCATCCGTCT Control locus-R ATGTTTTACGTCCAGTCCGC LMJ-gBKDE1β-F1 CGCTTGATATCTGCTGTCCA LMJ-gBKDE1β-R1 ACCGATAACAACGGACTTGC

LMJ-gBKDE1β-F2 GTGTATGCCACACAGGATGC LMJ-gBKDE1β-R2 TTTCTTCAACGCAAAAGGCT LMJ-gBKDE2-F1 ACCCCATGGTCTCCTCTCTT LMJ-gBKDE2-R1 CGTGTGTCGCAAACTGTCTT LMJ-gBKDE2-F2 GCCCAACATAAAGCAGGTGT LMJ-gBKDE2-R2 TAGTTGGGGGCGTGTTTTAG LMJ-gBKDE2-F3 ACCCCATGGTCTCCTCTCTT LMJ-gBKDE2-R3 CGTGTGTCGCAAACTGTCTT LMJ-gBKDE2-F4 ACACACACACACACACACGC LMJ-gBKDE2-R4 CCGTAAGGAAACAGTCGCTC LMJ-cBKDE1α-F GCGAGGAGGCCACCAACATC

Genomic forDNA PCR Mutants from CLIP LMJ-cBKDE1α-R CACCGCCGCGAAGTTGAAG BKDE1β-F AACCAAATGTTGCCTTGTTTAGGAC

BKDE1β-R TCAGAAGTCCAGCGCCTG

PCR for

- BKDE2-F GCACTGCCCAACTTCCACTT

R CLIP mutants BKDE2-R CTGACCGGCATCACTGACACA

Chapter 2

Chlamydomonas carries out fatty acid β-oxidation in ancestral

peroxisomes using a bona fide acyl-CoA oxidase

Fantao Konga, Yuanxue Lianga, Bertrand Légereta, Audrey Beyly-Adrianoa, Stéphanie Blangya, Richard P. Haslamb, Johnathan A. Napierb, Fred Beissona, Gilles Peltiera, Yonghua Li-Beissona,1

a Commissariat à l’Energie Atomique et aux Energies Alternatives, CNRS, Aix Marseille Université, UMR7265, Institut de Biosciences et Biotechnologies Aix Marseille, 13108 Cadarache, France b Department of Biological Chemistry and Crop Protection, Rothamsted Research, Harpenden, UK

In this work, my contributions are: I - Cloned the 11 kb genomic DNA encoding the ACX2 protein for genetic complementation. - Did lipid extractions and fatty acid composition analysis. - Carried out quantitative RT-PCR analysis of the expression level of the gene under various cultivation conditions. - Tested the effect of perturbation in fatty acid turnover on senescence and in prolonged darkness.

The Plant Journal (2017) 90, 358–371 doi: 10.1111/tpj.13498 Chlamydomonas carries out fatty acid b-oxidation in ancestral peroxisomes using a bona ﬁde acyl-CoA oxidase

Fantao Kong1, Yuanxue Liang1, Bertrand Legeret 1, Audrey Beyly-Adriano1,Stephanie Blangy1, Richard P. Haslam2, Johnathan A. Napier2, Fred Beisson1, Gilles Peltier1 and Yonghua Li-Beisson1,* 1Commissariat a l’Energie Atomique et aux Energies Alternatives, CNRS, Aix Marseille Universite, UMR7265, Institut de Biosciences et Biotechnologies Aix Marseille, 13108 Cadarache, France, and 2Department of Biological Chemistry and Crop Protection, Rothamsted Research, Harpenden, UK

Received 9 September 2016; revised 25 January 2017; accepted 27 January 2017; published online 31 January 2017. *For correspondence (e-mail [email protected]).

SUMMARY Peroxisomes are thought to have played a key role in the evolution of metabolic networks of photosynthetic organisms by connecting oxidative and biosynthetic routes operating in different compartments. While the various oxidative pathways operating in the peroxisomes of higher plants are fairly well characterized, the reactions present in the primitive peroxisomes (microbodies) of algae are poorly understood. Screening of a Chlamydomonas insertional mutant library identiﬁed a strain strongly impaired in oil remobilization and defective in Cre05.g232002 (CrACX2), a gene encoding a member of the acyl-CoA oxidase/dehydrogenase superfamily. The puriﬁed recombinant CrACX2 expressed in Escherichia coli catalyzed the oxidation of fatty

acyl-CoAs into trans-2-enoyl-CoA and produced H2O2. This result demonstrated that CrACX2 is a genuine acyl-CoA oxidase, which is responsible for the ﬁrst step of the peroxisomal fatty acid (FA) b-oxidation spiral. A ﬂuorescent protein-tagging study pointed to a peroxisomal location of CrACX2. The importance of peroxisomal FA b-oxidation in algal physiology was shown by the impact of the mutation on FA turnover during day/night cycles. Moreover, under nitrogen depletion the mutant accumulated 20% more oil than the wild type, illustrating the potential of b-oxidation mutants for algal biotechnology. This study provides experi-

mental evidence that a plant-type FA b-oxidation involving H2O2-producing acyl-CoA oxidation activity has already evolved in the microbodies of the unicellular green alga Chlamydomonas reinhardtii.

Keywords: acyl-CoA oxidase, microbodies, lipid catabolism, oil content, hydrogen peroxide, lipid homeostasis, nitrogen starvation, catalase, lipid droplet, Chlamydomonas reinhardtii.

INTRODUCTION The b-oxidation of fatty acids (FAs) plays a pivotal role in have been studied intensively in mammalian cells (Eaton, eukaryotic cells. This catabolic pathway generates acetyl- 2002), in germinating oilseeds (Graham, 2008) and in fun- CoAs via breakdown of FAs acquired from the environment gal species that can utilize FAs as a carbon source (Daum or released upon hydrolysis of membrane structural lipids et al., 2007), and have also recently been explored in plant and storage triacylglycerols (TAGs) by lipolytic enzymes (li- leaves (Kunz et al., 2009; Troncoso-Ponce et al., 2013; Fan pases), and could also be de novo synthesized FAs when et al., 2014). In mammalian cells b-oxidation of FAs occurs the rate of synthesis bypass downstream metabolic needs in both mitochondria and peroxisomes (Eaton, 2002), but (Marchesini and Poirier, 2003; Poirier et al., 2006; Graham, in plant/fungal cells it occurs exclusively in the peroxisome 2008). Lipid degradation therefore provides cells with car- (Purdue and Lazarow, 2001; Poirier et al., 2006; Graham, bon skeletons and energy to drive anabolic processes 2008). whilst also ensuring membrane function and cell ﬁtness Peroxisomes (also called microbodies) are small single through the elimination of oxidized, toxic or unusual FAs membrane-bound entities, and were originally deﬁned as produced either after being exposed to adverse growth organelles that carry out oxidative reactions leading to the conditions or via transgenic means (Marchesini and Poir- production of hydrogen peroxide (H2O2); thus the occur- ier, 2003; Moire et al., 2004; Poirier et al., 2006; Napier, rence of peroxisomes allows the separation of otherwise 2007). Lipid catabolic processes and the enzymes involved dangerous oxidative reactions from the remaining cellular

358 © 2017 The Authors The Plant Journal © 2017 John Wiley & Sons Ltd Fatty acid degradative processes in microalgae 359 metabolism. Since the discovery of peroxisomes in the dehydrogenase superfamily. We show that this protein 1950s (Rhodin, 1954), the metabolic processes occurring encodes an enzyme with a genuine acyl-CoA oxidase activ- within them have been well studied in a few model organ- ity (producing H2O2) that is required for breakdown of FAs isms including human and Saccharomyces cerevisiae (Pur- during lipid remobilization. We provide experimental evi- due and Lazarow, 2001), Pichia pastoris (Gasser et al., dence for the physiological roles of FA b-oxidation in algal 2013), Yarrowia lipolytica (Ledesma-Amaro and Nicaud, diurnal growth. Finally we demonstrate that shutting down 2016) and the higher plant Arabidopsis thaliana (Poirier core enzymes of the FA b-oxidation spiral increases oil et al., 2006; Graham, 2008; Hu et al., 2012). content in green microalgae. Plant peroxisomes are the best characterized in the green lineage, and are known to perform a plethora of RESULTS functions including lipid metabolism, detoxification, nitro- Isolation of a Chlamydomonas mutant compromised in oil gen metabolism, photorespiration and synthesis of some degradation hormones (Kaur et al., 2009; Hu et al., 2012). Most of the metabolic processes in the peroxisome are a part of large To understand lipid turnover processes in green microal- metabolic networks spanning several other subcellular gae, we screened an insertional mutant library for strains organelles, notably plastids and mitochondria. Indeed, perturbed in their capacity to remobilize oil reserves. The physical associations of peroxisomes with mitochondria or screening procedure has previously been described in plastids have been observed in 7-day-old Arabidopsis detail (Cagnon et al., 2013). Briefly, cells were cultivated in seedlings under transmission electron microscopy (Kaur acetate-containing medium (TAP) and then boosted to et al., 2009). Essential roles of peroxisomes in coordinating accumulate oils via removal of nitrogen (N; TAP-N72 h). In plant metabolism can also be seen through their dynamic the wild-type (WT) strains oil is usually used rapidly to nature (increases in both size and number), and that plants power regrowth upon resupply of N (Siaut et al., 2011). Oil without peroxisomes are not viable (Kaur et al., 2009; Hu content was determined 24 h after addition of N back to et al., 2012). Moreover, the peroxisome has been shown to the culture (MM24h). Several mutants were found to be be useful target for re-programming plant metabolism to impaired in their capacity to remobilize oil upon N resup- produce bioplastics and to increase plant biomass produc- ply, including Nb7D4 (Figure 1a,b). Oil quantification based tivity (Poirier, 2002; Moire et al., 2004; Poirier et al., 2006; on densitometry showed that in this mutant 50–70% of oils Kessel-Vigelius et al., 2013). accumulated at the height of the oil accumulation phase Despite the essential functions attributed to peroxisomes were retained in the mutant 24 h after N resupply when in higher plants, and the growing interest in microalgae for kept in the dark, contrasting with WT cells where 20–30% green biotechnology, little is known about the metabolic of oil is retained (Figure 1c). The retention of oils in the repertoire of algal peroxisomes or about degradation of lipid droplet (LD) was observed using confocal microscopy FAs in microalgae. Earlier literature suggests that, depend- (Figure 1d). Flow cytometry, chemical lipid analysis and ing on algal species, FA degradation can occur in mitochon- microscopy therefore pointed to a severe defect in oil dria, in peroxisomes or in both organelles (Stabenau et al., degradation in the Nb7D4 mutant. 1984, 1989; Winkler et al., 1988). Differences in the compart- Nb7D4 is disrupted in Cre05.g232002 encoding an acyl- mentalization of the enzymes of the FA oxidative pathway CoA oxidase (ACX) have been suggested as a consequence of different phylo- genetic development. Of particular note, it is reported that a The insertion site of the paromomycin resistance gene few algal species in the genera Mougeotia, Pyramimonas (AphVIII)inNb7D4 was identified by restriction enzyme and Eremosphaera harbor a peroxisomal acyl-CoA oxidiz- site-directed amplification (RESDA)-PCR (Gonzalez-Balles- ing (ACX) enzyme which uses O2 as an electron acceptor ter et al., 2011). Using a combination of specific and and produces H2O rather than H2O2 (Winkler et al., 1988; degenerate primers, a fragment of 1000 bp was amplified, Stabenau et al., 1989). This feature is often observed in per- including 426 bp of the AphVIII gene and 574 bp of flank- oxisomes that lack catalase (Stabenau et al., 1989). Thus ing sequences. The fragment was sequenced and then certain algal species may harbor ACXs with activities differ- Blasted against the genome of C. reinhardtii v5.5 (Phyto- ent from the H2O2-producing ones present in higher plants. zome); AphVIII was found inserted in the third intron of the In this study, we used the green unicellular microalga locus Cre05.g232002 (Figure 2a). A RT-PCR analysis using Chlamydomonas reinhardtii as a model to uncover factors gene-specific primers demonstrated that the insertion involved in lipid hydrolysis. To this end, we employed a resulted in null expression of Cre05.g232002 in the mutant forward genetic approach to screen for mutants compro- background (Figure 2b). Cre05.g232002 encodes a protein mised in oil remobilization. We report here the detailed of 76 kDa and is annotated as acyl-CoA oxidase in v5.5 of genetic, biochemical and cell biological characterization of the C. reinhardtii genome (Merchant et al., 2007). In many a mutant defective in a member of the acyl-CoA oxidase/ organisms, acyl-CoA oxidase (EC1.1.1.3) catalyzes the first

Figure 1. The mutant Nb7D4 (cracx2) is compromised in oil turnover. (a) Oil content screening by flow cytometry based on Nile red fluorescence. (b) Chemical quantification of oil content by thin layer chromatography (TLC). (c) Defect in triacylglycerol (TAG) remobilization. Cells were starved nitrogen for 3 days, and samples were taken for analysis during the recovery phase (+N24 h). Each lane represents one biological replicate from that genotype. Total lipids were extracted from a fixed number of cells, then deposited onto a TLC plate, and TAGs were revealed after staining

with a CuSO4 containing solution (see Experimental procedures for details). This is a representative of at least four or six biological replicates done at different times. Error bars represent standard deviation. (d) Retention of lipid droplet in the mutant cells as revealed by staining with Nile red. Lipid droplets are colored green and chlorophyll autofluorescence is in red. Bars = 10 lm. AU, artificial unit. [Colour figure can be viewed at wileyonline library.com].

committed step in FA b-oxidation, thus exerting major con- A second allele (LMJ.SG0182.014586) was identified in trol on this pathway (Klein et al., 2002; Haddouche et al., the mutant library made by the Jonikas group (Li et al., 2010; Theodoulou and Eastmond, 2012). BlastP analysis of 2016). It harbors an insertion in the first intron (Figure 2a). this protein on The Arabidopsis Information Resource The null expression of CrACX2 in this mutant was con- (TAIR) website revealed that its amino acid sequence is firmed by RT-PCR (Figure 2b). Analyses of oil content dur- mostly similar to Arabidopsis AtACX2 (At5g65110) (54% ing the N recovery phase showed impairment in oil identity, and 68% similarity). Based on this sequence remobilization, i.e. the same defect as observed for the homology, we named the protein encoded by the locus cracx2-1 mutant (Figure 2e). We thus named this line Cre05.g232002 as CrACX2, and the mutant Nb7D4 as cracx2-2. Dynamic changes in TAG content in both mutant cracx2-1. alleles (cracx2-1, cracx2-2) and their corresponding WT strains are shown in Figure S3. Taken together, these data Genetic complementation and isolation of a second allele firmly establish that the impairment in oil remobilization in of the mutant cracx2-2 the mutants is due to disruption in the normal expression To confirm that the cracx2-1 mutant phenotype was a result of CrACX2. of the disruption of the gene Cre05.g232002, complementa- cracx2 is defective in b-oxidation of FAs tion of the mutant with a cDNA of the WT locus Cre05.g232002 was conducted. To this, we first cloned the To determine if the failure of cracx2 to utilize TAG is full-length transcript (2025 bp; corresponding to caused by impaired b-oxidation, we tested the growth of Cre05.g232002.t2.1) into the vector pChlamy4 in frame to Chlamydomonas on minimal medium (MM) containing the 30 end of the epitope V5. Despite several trials, we could oleic acid as the sole carbon source in the dark. Indeed, not clone the cDNA corresponding to Cre05.g232002.t1.1. Chlamydomonas cells are able to utilize oleic acid supplied The promoters and gene structure information are shown in the medium to drive TAG synthesis in the presence of in Figure S1(a) in the Supporting Information. After screen- light and acetate (Fan et al., 2011). The utilization of oleic ing about 100 independent zeocin-resistant clones, one acid as a source of carbon requires a functional b-oxidation clone (cracx2-1;V5-CrACX2) recovered almost its full capac- cycle in which oleic acid is reduced to acetyl-CoA then to ity to remobilize oil (Figure 2c), and the presence of the sugars through the glyoxylate and gluconeogenesis path- expressed protein is validated by immunoblot against the ways (Graham, 2008). We reasoned that mutants defective anti-V5 antibodies (Figure 2d). Results for some representa- in the b-oxidation of FAs should display reduced growth tive clones possessing varying degrees of complementation when cultivated in the presence of oleic acid as the only are shown in Figure S2. Due to the notorious low expres- carbon source. sion of transgenes in the nuclear genome of C. reinhardtii, To test this, cells were grown under strict photoau- only a few complemented lines were obtained here. totrophic conditions, transferred to oleic acid-

Figure 2. Nb7D4 is defective in an acyl-CoA oxidase. (a) The insertion site of the cassette AphVIII in the mutant Nb7D4 (cracx2-1), and in the second line cracx2-2 (background strain CC-4533), respectively. (b) RT-PCR valid zero expression of CrACX2 in the cracx2-1 and cracx2-2 mutants, respectively. (c) Genetic complementation of cracx2-1. C1: a representative of the complemented strain. (d) Immunoblot detection of the presence of CrACX2 protein in the C1 line. (e) Defect in oil remobilization in the mutant cracx2- 2 line. (c,e) Cells were first starved of N for 3 days, then transferred to an N-containing medium to initiate oil remobilization. Cells were harvested for lipid analysis 24 h after the onset of oil degradation. The triacylglycerol (TAG) quantification data by TLC are representative of three biological replicates done at different times. Data are means of two biological replicates with two technical replicates each, and error bars indicate standard deviations. (b) CBLP (Cre06.g278222) codes for a receptor of activated protein kinase C; (d) V5 is the recombinant protein fused at its N-terminus to the epitope V5. [Colour figure can be viewed at wileyonlinelibrary.- com].

supplemented MM medium and then cell density was functional FA b-oxidation is therefore essential in the redis- monitored. We ﬁrst evaluated the optimal oleic acid contribution of carbon skeletons occurring under strict carbon centration for such a test in the WT. We observed that starvation in C. reinhardtii. when the oleic acid concentration exceeds 0.8 mM cells Turnover of FAs during diurnal growth is compromised in started to bleach and eventually died (Figure S4), most cracx2 mutant cells likely due to the detergent property of FAs. An oleic acid concentration of 0.5 mM is optimal when added to a cell It has been observed in the marine unicellular stra- culture of 2 9 106 cells ml1. We observed that 24 h after menopile Nannochloropsis oceanica that total FA content addition of 0.5 mM oleic acid, WT cells grew at a rate twice varies during a day/night cycle, i.e. FAs accumulate during that of the mutant cracx2-1 (Figure 3). Growth was arrested the day and degrade at night (Poliner et al., 2015). This eventually, probably due to the exhaustion of oleic acid in phenomenon was also observed here in C. reinhardtii the medium, because when additional oleic acid (0.5 mM) when it was cultivated photoautotrophically under a day/ was then added to the same culture, regrowth was night cycle (12 h/12 h) (Figure 4a). At the end of night per- observed with the WT strain but not the cracx2 mutant iod, the mutant retained >80% of total FAs accumulated at (Figure 3). We also observed a slower growth of the the end of day, in contrast to 60% in the WT (Figure 4b). A mutant compared with the WT in the control experiment role of CrACX2 in lipid turnover during the day/night cycle (dotted lines in Figure 3), probably due to the fact that is consistent with high transcription of CrACX2 during the under strict carbon starvation, b-oxidation of FAs released night (Figure 4c, adapted from Zones et al., 2015). There- from membrane lipids could provide another source of the fore, this study provides experimental evidence that func- carbon skeletons required for maintenance of growth. This tional FA b-oxidation is involved in lipid homeostasis test allows us to attribute the defect in oil utilization in the during nutrient stress but also plays a role in lipid turnover mutant to a block in FA b-oxidation. These data show that following natural diurnal cycles.

ScPOX1 cannot grow when oleic acid is present as the sole carbon source, thus demonstrating its essential role in the oxidative degradation of FAs in yeast peroxisomes (Dmo- chowska et al., 1990). To avoid a potential effect of protein mis-targeting or the occurrence of ‘untypical’ targeting sequences at its N- or C- terminus, we made two constructs, one with mCherry protein fused at the N-terminus and the other with mCherry fused at the C-terminus of CrACX2 (Figure S1b,c). CrMDH2 contains a typical PTS2 signal at its N-terminus and has previously been localized to peroxisomes in C. reinhardtii (Hayashi et al., 2015); it was used here as a positive marker for peroxisomes. Either of the two constructs was co-transformed, independently, into the WT strain with the PTS2 sequence from CrMDH2 fused to GFP (Figure S1d). Protein ﬂuorescence analyses using confocal microscopy demonstrated that the co-transformation of mCherry-CrACX2 and PTS2(MDH2)-GFP co-localize to peroxisomes in C. reinhardtii (Figure 5). Despite several attempts, no signals could be detected when mCherry was fused at the C-terminus of CrACX2 (CrACX2-mCherry). One 1st 2nd feeding of oleic acid of the reasons for this could be that the fusion of mCherry

Figure 3. Oleic acid feeding test in Chlamydomonas reinhardtii. to the C-terminus of the CrACX2 could potentially interfere Cells were grown to mid log phase, then diluted to around 2 million cells with correct protein targeting due to the likely presence of 1 ml . Either ethanol (as a control) or 0.5 mM oleic acid was added to each PTS1(-like) internal signal sequences close to the C-termi- culture, and cells were then kept in the dark. The growth kinetics were followed up for a few days after the addition of oleic acid. Potential growth is nus (Kaur et al., 2009). seen as a result of oleic acid utilization. This is a representative ﬁgure for three independent biological replicates. CrACX2 catalyzes the conversion of acyl-CoA to trans-2- Data are means of three technical replicates, and error bars indicate stan- enoyl-CoA and produces H2O2 dard deviations. [Colour ﬁgure can be viewed at wileyonlinelibrary.com]. Previous studies reported the occurrence of acyl-CoA oxidases in certain algal peroxisomes which do not produce H O but instead transfer the energy into water (Stabenau CrACX2 localizes to peroxisomes 2 2 et al., 1989). The absence of catalase in the peroxisomes of

No subcellular localization could be assigned for the Chlamydomonas raised the question of whether H2O2-pro- CrACX2 using the PredAlgo prediction tool. This is not sur- ducing activities were present in its peroxisomes. To prising, because in the design of the PredAlgo program, understand the molecular mechanism of this oxidative step peroxisomes/microbodies are not included due to lack of in Chlamydomonas, we characterized the catalytic activity known peroxisome protein sequences in algae (Tardif of CrACX2. A codon-optimized version of CrACX2 was et al., 2012). In order to determine in which compartment cloned into an E. coli expression vector (Figure S5). In par- (peroxisomes or mitochondria) FA b-oxidation occurs in allel, we also expressed AtACX2 in E. coli (Figure 6a). The

Chlamydomonas, we determined the subcellular localiza- AtACX2 protein is known to produce H2O2 while oxidizing tion of CrACX2, the ﬁrst enzyme of the pathway. Sequence long chain fatty acyl-CoAs (Hooks et al., 1999; Eastmond examination of the C-terminus or N-terminus of CrACX2 et al., 2000). Puriﬁed recombinant CrACX2 protein cat- did not reveal obvious sequence similarity to either the alyzed the conversion of acyl-CoAs to their respective peroxisome-targeting sequence (PTS1) [(S/C/A)(K/R/H)(L/ trans-enoyl-CoA products, and produced H2O2. CrACX2 is

M)] or the PTS2 [(R/K)(L/V/I)X5(H/Q)(L/A)] consensus more active toward long chain acyl-CoAs (C18:1-, C18:0-, sequence (Klein et al., 2002; Hu et al., 2012). This is differ- C20:0-, C16:0-CoAs) than to medium chain acyl-CoA (C12:0- ent from the homolog of the two Arabidopsis proteins – CoA) (Figure 6b). AtACX2 showed higher activity toward AtACX2 contains a PTS2 sequence, whereas AtACX1 con- C18:1-CoA followed by C12:0-CoA, and had lower activity tains a typical PTS1 signal (Eastmond et al., 2000) – but with C16:0-, C18:0- and C20:0-CoA (Figure 6b). The prefer- similar to ScPOX1 from S. cerevisiae, where no apparent ence for mono-unsaturated CoAs over saturated CoA is PTS sequence is present but the protein is known to be consistent with a previous ﬁnding (Hooks et al., 1999), imported into the peroxisomes through a novel non-PTS1 whereas in our assay the Arabidopsis protein can also uti- pathway (Klein et al., 2002). The knockout mutant of lize C12:0-CoA.

Figure 4. The cracx2 mutant is impaired in fatty acid (FA) turnover during day/night cycles. (a) Fluctuation of FA content during the day and night cycle in the wild type and the mutant. (b) Percentage of FA retained during the night. (c) Expression proﬁle of CrACX2 within a day/night (12 h/12 h) cycle (data are based on Zones et al., 2015). (a,b) Data are means of three biological replicates and with two technical replicate each. Error bars represent standard deviations. (c) RPKM stands for ‘reads per kilobase per million mapped reads’. [Col- our ﬁgure can be viewed at wileyonlinelibrary.- com].

Interestingly, the algal protein showed slightly broader and for long-chain acyl specificity (KWWI/PG/G/N) in activity toward different acyl-chain lengths than the plant CrACX2. protein (Figure 6b). This may at least partly explain why The cracx2 mutant does not over-accumulate acyl-CoAs a strong phenotype can already be observed in the single mutant of Chlamydomonas (cracx2), whereas the We further asked if a block in the first step of the FA b-oxida- single Arabidopsis mutant atacx2 did not show any phe- tion spiral resulted in over-accumulation of its precursors, notype (Pinfield-Wells et al., 2005). A defect in oil remo- i.e. free acyl-CoAs. Total acyl-CoAs were extracted and ana- bilization is only detectable in the Arabidopsis double lyzed from the mutant as well as the WT by liquid chro- mutant when both AtACX1 and AtACX2 are absent (Pin- matography coupled to mass spectrometry (LC-MS/MS) field-Wells et al., 2005). To conclude, ACX activity mea- (see Methods S1). No significant difference in acyl-CoAs surement showed that both the algal and plant proteins could be found between the WT and cracx2-1 under any of have wide substrate specificities toward a range of med- the three conditions of optimal growth, N starvation and N ium- to long-chain CoAs (Hooks et al., 1999; Pinfield- resupply (Figure S6). Free FAs were not detected in the Wells et al., 2005), thus further supporting their essential mutant either. This is consistent with the lack of growth role in FA b-oxidation. As previously demonstrated for defects in the mutant (Figure S7) because large amounts of the plant protein (AtACX2), CrACX2 requires flavin free FAs are cytotoxic and can have negative impact on cell adenine dinucleotide (FAD) as a cofactor (Figure 6). This growth (Fan et al., 2013a). This suggests that, in algae, there is supported by the presence of signature amino acid are tight metabolic regulations that coordinate cytosolic sequences for FAD-binding (the GGGHGY motif) lipolysis with b-oxidation of FAs in the peroxisomes.

PTS2(MDH2)-GFP mCherry-CrACX2 Merge Figure 5. CrACX2 localizes to peroxisomes in Chlamydomonas reinhardtii. Peroxisomal localization of mCherry-CrACX2 by confocal microscopy in a representative of transgenic lines expressing mCherry-tagged CrACX2. This line was created by co-transformation with a GFP-tagged PTS2 signal from the protein CrMDH2. Pseudo-colors are used. Chlorophyll autofluorescence is shown as magenta. GFP, green fluorescent protein; PTS, peroxisomal targeting signal; TRITC, tetramethylrhodamine isothiocyanate. [Colour figure can be viewed at wileyonlinelibrary.com].

Chlorophyll Transmission

Cellular oil content is increased by 20% in cracx2 mutants proteins involved in b-oxidation of FAs have so far been during N starvation studied in detail in microalgae. Here, through a forward genetic screen in C. reinhardtii, we isolated the mutant Quantitative RT-PCR analysis of the transcriptional expres- cracx2 defective in the first step of the core FA b-oxidation sion level of CrACX2 in the WT revealed that CrACX2 pathway. In-depth characterization of the mutant revealed expression is upregulated two-fold during the N recovery that Chlamydomonas, an ancestral eukaryotic cell, harbors phase (Figure 7a), in accordance with its role in FA degra- its FA degradation machinery in its peroxisomes, and dation. Moreover, we observed that it is repressed during defects in the normal functioning of this pathway increase N starvation, when active oil synthesis occurs. Therefore, cellular oil content under N starvation. Oil remobilization is we examined the changes in storage lipid content during severely, but not entirely, blocked in the cracx2 mutant the N starvation phase. As shown in Figure 7(b), a 20% (Figures 1 and 2), which could be explained by the pres- increase in total TAGs was observed in the cracx2-1 and ence of at least four other ACX isozymes in C. reinhardtii cracx2-2 mutants after 3 days of N starvation. Strikingly, (Table S1) (Merchant et al., 2007; Li-Beisson et al., 2015). this increase in TAG content is largely due to an increase These isozymes show at least 35% sequence identity to in TAG52 molecular species (>55% increases) while there their closest Arabidopsis ACX proteins, where functional are no significant changes in TAG50 or TAG54 species (Fig- overlaps are well known because oil breakdown and seed- ure 7c). The preferential increase in TAG52 lipid species ling establishment are largely unaffected in any of the sin- suggests that during N starvation the cellular TAG pool is gle mutants of Arabidopsis (Eastmond et al., 2000; Graham not metabolically inert but highly dynamic, conforming to and Eastmond, 2002; Pinfield-Wells et al., 2005). Presum- the observations made previously either in germinating ably, therefore, in the cracx2 mutant the presence of the seeds of Arabidopsis FA b-oxidation mutants where TAG four other isozymes makes a partial contribution to the content and composition were altered due to simultaneous continued operation of the pathway. turnover and synthesis (Hernandez et al., 2012), or in actively growing vegetative tissues (Fan et al., 2013b). Dur- b-Oxidation of FAs occurs in Chlamydomonas ing normal cultivation under standard mixotrophic condi- peroxisomes tion, no significant differences were observed in terms of b-Oxidation of FAs is one of the major lipid catabolic path- oil and polar lipid content and FA composition between ways, essential for converting fats to sugars in many WT and the mutant (Figure S8). organisms. Since its discovery over 100 years ago, the biochemistry and subcellular compartment harboring this DISCUSSION pathway in microalgae has remained enigmatic. In this Increasing evidence suggests the importance of lipid cata- study, we demonstrate that Chlamydomonas houses the bolism in metabolically active cells, but no genes or major reactions of FA b-oxidation in peroxisomes. Two

Figure 6. CrACX2 is a bona ﬁde acyl-CoA oxidase. (a) (b) (a) Production of recombinant acyl-CoA oxidizing (ACX) protein in Escherichia coli. Purified enzymes (b) ACX enzymatic activities. Data are means of four independent assays. Error AtACX2 bars represent standard deviations. *Denotes the CrACX2 test including C18:1-CoA but with the absence of FAD. [Colour ﬁgure can be viewed at wileyonlineli- (kDa) brary.com]. 115 80 77 kDa 65 74 kDa

40 30 *

lines of evidence support this claim: ﬁrstly the protein namely the two multifunctional proteins (enoyl-CoA hydra- (CrACX2) catalyzing the initial step of the pathway is tase, 3-hydroxyacyl-CoA dehydrogenase) and the ketoacyl- located in peroxisomes and not mitochondria; and sec- CoA thiolase (Li-Beisson et al., 2015), and their expressions ondly, biochemical analysis showed that CrACX2 is a bona are also found to be higher in the dark than during the day

ﬁde oxidase, producing H2O2, rather than a dehydroge- (Zones et al., 2015), suggesting a similar role in lipid turn- nase. All Chlamydomonas ACX proteins contain two acyl- over to that of CrACX2. Nevertheless, their exact contribu- CoA dehydrogenase domains and one acyl-CoA oxidase tions to degradation of FAs and their subcellular localization domain, a general feature of all known peroxisomal acyl- in algal cells remain to be validated experimentally. CoA oxidases (Eastmond et al., 2000). Acyl-CoA oxidases The ﬁnding that Chlamydomonas employs a peroxisomal and acyl-CoA dehydrogenases are two closely related pathway to degrade FAs raised the question of why Chlamy- enzyme families that are present in the peroxisome and domonas has adopted a peroxisomal pathway where energy mitochondria, respectively, and require FAD. They both is lost (through production of H2O2 and its immediate con- catalyze the dehydrogenation of acyl-CoA to a trans-2- version to H2O) in contrast to a mitochondrial pathway enoyl-CoA, but differ in the oxidative half of the reaction, where FA degradation is tuned to energy production. One of i.e. acyl-CoA oxidase uses molecular oxygen to re-oxidize the possible explanations is that Chlamydomonas,foundin

FADH2, thereby producing H2O2, whereas acyl-CoA dehy- soil where large fluctuations in nutrient supply occur, opts drogenase transfers the electrons to the mitochondrial res- for high metabolic fluxes at the cost of energy loss, rather piratory chain and is thus coupled to ATP production (Kim than favoring energy conservation. This would be consistent and Miura, 2004; Poirier et al., 2006). Protein structural with the extensive metabolic flexibility that Chlamydomonas studies have shed light on the similarities and differences displays, most likely as a result of adaption to inhabit distinct in the functional mode of these two closely related pro- environmental niches and to survive fluctuations in nutrient teins (Kim and Miura, 2004). No homologs of mammalian availability (Grossman et al., 2007). acyl-CoA dehydrogenase (Lea et al., 2000) are encoded in As well as acetyl-CoAs, the other end products of FA b- the genome of C. reinhardtii. The protein showing the oxidation are H2O2 and NADH. H2O2-producing activity is a highest sequence similarity to mammalian acyl-CoA dehy- key feature and is behind the naming this organelle a ‘per- drogenase in fact encodes a homolog of the Arabidopsis oxisome’ (Kaur et al., 2009). H2O2 is highly oxidative, and isovaleryl-coenzyme A dehydrogenase (AtIVD, At3g45300) potentially damaging to cellular components including known to be involved in breakdown of branched chain DNA, proteins and lipids. Due to the absence of catalase in amino acids in the mitochondrion (Gu et al., 2010). Based the peroxisomes of Chlamydomonas (Kato et al., 1997), on this evidence, it is highly likely that C. reinhardtii other reactive oxygen species (ROS)-detoxifying enzymes employs a peroxisomal pathway for degradation of its oil in the peroxisome must function to quench H2O2 produced reserves. during the period of active oil degradation. An alternative As well as the first reaction (i.e. ACX), the genome of is the ascorbate peroxidase (APX)/monodehydroascorbate Chlamydomonas also encodes all the other three enzymes reductase (MDAR) electron transfer system. Indeed disrup- catalyzing the subsequent steps of the FA b-oxidation spiral, tion of the MDAR system in Arabidopsis results in

Figure 7. Oil content analysis of cells starved of nitrogen. (a) Quantitative RT-PCR analysis of transcriptional responses of CrACX2 to differing N status in the wild type. Data are fold changes compared with the expression level determined for optimally grown cells (in acetate containing medium, TAP). The housekeeping gene used is RACK1. Data are means of three biological replicates, and with two technical replicates each. This is a representative of three independent experiments. (b) Comparison of triacylglycerol (TAG) content. (c) Analysis of molecular species of TAGs by LC-MS/MS. Data are means of three biological replicates with two technical replicates each; error bars represent standard deviations. Detector response has been normalized based on cell numbers. *Signiﬁcant difference between strains. Statistical analysis was carried out using the Student’s t-test (P < 0.05). [Colour ﬁgure can be viewed at wileyonlinelibrary.com].

impaired oil catabolism through inhibition of the sugar proteins are encoded in the genome of Chlamydomonas dependent-1 (SDP1) lipase activity by escaped H2O2 (East- (Merchant et al., 2007). The contribution of these proteins mond, 2007). This APX/MDAR membrane-bound system to lipid catabolism in algal peroxisomes needs to be tested has been shown to possess a much higher afﬁnity than in the future once corresponding mutants are available. catalase for H O (Lisenbee et al., 2003, 2005). Another 2 2 Physiological roles of FA b-oxidation in Chlamydomonas possible route includes hydroxypyruvate reductase (HPR) and also a peroxisomal NADH transporter. Deﬁciencies in The isolation of the cracx2 mutants in this study provided these proteins have resulted in plants impaired in oil us with a means not only to probe the role of FA b-oxida- breakdown (Pracharoenwattana et al., 2010; Bernhardt tion in oxidative degradation of lipids, but also to experi- et al., 2012). Enzymes homologous to known Arabidopsis mentally test the physiological functions of peroxisomes in

C. reinhardtii. The increase in oil content in the cracx2 glyoxylate cycle, apart from isocitrate lyase (ICL), are mutants under N starvation supports the idea that FA turn- located in the peroxisomes of C. reinhardtii (Lauersen over is ubiquitous, occurs simultaneously with active oil et al., 2016). Among the putative glyoxylate enzymes, only synthesis in green algae and also plays a role in carbon ICL, found to be located in the cytosol, has been shown to management during day/night cycles. be essential for a functional glyoxylate cycle because the Peroxisomal b-oxidation in fungi is primarily devoted to null mutant icl cannot grow in the dark when only acetate the degradation of extracellular FAs supplied in the diet for is available as a carbon source (Plancke et al., 2014; Lauer- the subsequent use of acetyl-CoA as a carbon source for sen et al., 2016). Thus, there had previously been no func- growth. In this study, we showed that when supplied at tional demonstration of the involvement of the low concentrations, and in the absence of other carbon peroxisome-based glyoxylate cycle proteins in acetate sources (during carbon starvation condition), Chlamy- metabolism and little was known about metabolism occur- domonas can convert FAs to sugars to power growth, ring in the peroxisomes of C. reinhardtii. whereas mutant defective in the normal function of the This study provides experimental data to demonstrate b-oxidation cycle cannot. This might not be a major cata- the active involvement of peroxisomes in subcellular metabolic reaction, but it is significant enough to sustain bolism in C. reinhardtii – highlighting its central place as growth under strict carbon starvation, which might be criti- an ‘organelle at the crossroads’ in green microalgae. It cal for survival under extreme environmental conditions. would be interesting to determine in future studies if con- In plant tissues, the most studied physiological function version of fats to sugars in Chlamydomonas requires the has been the utilization of internal neutral lipids in storage coupling of FA b-oxidation to the peroxisomal glyoxylate tissues (e.g. seeds); this allows cells to convert stored cycle as occurs in plants (Pracharoenwattana et al., 2005), TAGs to sugars. Intensive research in the last 10 years has or if it can function independently as is the case in S. cere- also started to uncover the essential functions of FA beta- visiae, due to the presence of a carnitine shuttle (Vanroer- oxidation in other aspects of growth and development, mund et al., 1995; Graham and Eastmond, 2002). especially in active photosynthetic cells, including their Shutting down FA b-oxidation led to augmented levels vital role under prolonged carbohydrate starvation in of TAGs, further demonstrating the dynamic nature of lipid extended darkness (Kunz et al., 2009), namely the synthe- turnover in Chlamydomonas. Impairment in lipid catabo- sis of FA-based signaling molecules, degradation of lism has previously been shown to increase lipid content branched chain amino acids and their roles in flower devel- (Slocombe et al., 2009; Fan et al., 2013b; Trentacoste et al., opment (Gerhardt, 1992; Hu et al., 2012). 2013), but this study provides the first such example for To prevent lipotoxicity, cells have developed sophisti- green microalgae. Considering the increasing interest in cated mechanisms to coordinate the synthesis and meta- developing algae as a platform for the production of bio- bolism of their FAs. To avoid a large flux of acyl-chain s materials, and the central place the peroxisome occupies entering the cytoplasm, the sequestration of acyl-chains in cellular metabolism, the capacity to manipulate metabo- into TAGs stored in LDs has been shown to be an effective lism in peroxisomes seems essential. For example, one of way to protect cells in plant vegetative tissues (Fan et al., the often neglected functions of FA b-oxidation is its 2014), as well as in mammalian cells (Rambold et al., ‘house-keeping’ function during normal development, i.e. 2015), from lipotoxic damage. This study suggests that preventing the integration of ‘exotic’ FAs into membranes. TAGs could also serve as a temporary reservoir for acyl- This house-keeping role is particularly amplified in trans- groups in the green alga C. reinhardtii. This is evidenced genic plants synthesizing ‘unusual’ FAs where FA b-oxida- by the fact that during N starvation, massive degradation tion is enhanced (Jaworski and Cahoon, 2003; Moire et al., of membrane lipids occurs (Siaut et al., 2011) and the extra 2004). Thus, knowledge gained through this study should fluxes of acyl-chains were therefore sequestered in TAGs. be useful not only for production of biodiesel and other The amount of TAGs produced is even higher in the functional FAs in microalgae, but will also have implica- mutant cells where FA b-oxidation has been shut down. tions for algal fitness and biomass productivity. Peroxisome biology and algal biotechnology EXPERIMENTAL PROCEDURES Research on algal peroxisomes has been limited compared Chlamydomonas reinhardtii strains and culture conditions with that on other organelles due to their elusive nature and associated technical challenges. Only fairly recently Chlamydomonas reinhardtii strain dw15.1 (nit1-305 cw15; mt+) have peroxisomes been visualized and shown to occur in was used to generate the mutant library (Cagnon et al., 2013). l 1 Chlamydomonas (Shinozaki et al., 2005, 2009; Hayashi and Independent paromomycin-resistant (10 gml ) clones were screened for oil content by flow cytometry after Nile Red staining. Shinozaki, 2012; Hayashi et al., 2015). Since then, micro- Chlamydomonas reinhardtii can be cultivated photoautotrophi- scopy and fluorescence protein tagging studies have cally (MM medium supplied with 2% CO2 in air) (Cagnon et al., shown that homologs of the known proteins of the 2013) or mixotrophically in TAP medium (Harris, 2001). For all

© 2017 The Authors The Plant Journal © 2017 John Wiley & Sons Ltd, The Plant Journal, (2017), 90, 358–371 368 Fantao Kong et al. cultures, cells were cultivated in an incubator (Infors, http://www. CBLP-F1 and CBLP-R1, and was used as a housekeeping gene for infors-ht.co.uk/en) (25°C, 100 rpm shaking, and illuminated at normalization (Kong et al., 2015). 2 1 100 lmol photons m sec ). To induce N starvation, exponen- The cDNA sequences of CrACX2 (Cre05.g232002.t2.1) was tially grown cells were centrifuged at 600 g for 5 min, washed amplified using the gene-specific primers KpnI-ACX2-F2 and XbaI- twice in N-free medium and finally resuspended in TAP-N or MM- ACX2-R2. The PCR reaction was carried out using high-fidelity N. Cell concentrations were determined using a Multisizer 3 Coul- KOD Hot Start DNA Polymerase (Merck Millipore, http://www.merc ter counter (Beckman Coulter, https://www.beckmancoulter.com/). kmillipore.com/). The amplified DNA fragment was cloned as a Exponentially grown cells were harvested for all analyses. KpnI–XbaI fragment into the pChlamy4 vector (Life Technologies) The strain LMJ.SG0182.014586 harboring an insertion in the which contains the ble gene conferring zeocin resistance (Stevens gene encoding CrACX2 was ordered, together with its parental et al., 1996; Kong et al., 2015), generating the plasmid pChlamy4- line CC-4533, from the Martin Jonikas collection (Li et al., 2016; cACX2. The target gene was cloned in-frame with a V5 tag at its 30 https://www.chlamylibrary.org/). Null expression in the CrACX2 end, allowing screening of complemented lines by anti-V5 gene was verified by RT-PCR using the primers ACX-F2 and ACX- (GKPIPNPLLGLDST) antibodies. R2, and control PCR was done via amplification of the CBLP gene pChlamy4-cACX2 was integrated into the cracx2-1 genome by using primers CBLP-F2 and CBLP-R2. The sequences for these and electroporation (Shimogawara et al., 1998). Briefly, exponentially all the other primers used are given in Table S2. grown cells (about 20 million) were harvested by centrifugation and suspended in 250 ll of TAP medium supplemented with Identification of the insertion site by RESDA-PCR 40 mM sucrose. Electroporation was performed by applying an l The genomic region(s) flanking the inserted DNA (AphVIII) were electric pulse of 0.7 kV at a capacitance of 50 F (GENE PULSER, l characterized by RESDA-PCR (Gonzalez-Ballester et al., 2005), Bio-Rad, http://www.bio-rad.com/) to cells mixed with 0.3 gof SspI-linearized plasmids. The transgenic strains were selected which allows amplification from primers with a known sequence 1 into adjacent regions using degenerate primers containing restric- directly on TAP agar plates containing zeocin (25 mg L ), and the l tion enzyme site sequences that are frequent in the Chlamy- plates were incubated under continuous light (50 mol photons 2 1 ° domonas genome (AluI, PstI, SacII or TaqI). Specific primers (RB1 m sec )at25C. Colonies resistant to zeocin were visible after 0 around 7 days. The presence of the cassette in the transformants and RB4) at the 3 regions of AphVIII, and the degenerate primers (DegPstI and DegTaqI) and specific primer (Q0) targeted to degen- in some selected strains was determined by PCR with amplifica- erate primers, were used for RESDA-PCR to amplify the DNA from tion by the primers ACXcloneF and RBcS2 R. the possible regions affected. The amplified PCR product (around Oleic acid feeding test 0.8–1.0 kb) was sequenced. Blast searches of the amplified flanking sequences against the genome of C. reinhardtii (v5.5) (Mer- Oleic acid (Sigma-Aldrich, http://www.sigmaaldrich.com/) at a final chant et al., 2007) located at Phytozome (v10.3) identified the concentration of 0.5 mM from a 1 M stock solution in ethanol was insertion site. added to pre-cultures grown photoautotrophically. The same volume of ethanol was added to the controls. The capacity to use Quantification of gene expression by real-time PCR oleic acid as a carbon source was determined by cell growth in the dark at 25°C. Quantification of CrACX2 expression by real-time PCR (qRT-PCR) was performed on the LightCycler 480 System (Roche, http:// Protein extraction and immunoblot analysis www.roche.com/) using 2.5 ll of SYBR Premix Ex Taq II (Takara, www.clontech.com) in a final volume of 5.0 ll with 2.0 ll first Total proteins were extracted from Chlamydomonas cells follow- cDNA strand synthesized as described above and 0.5 ll of 3 pmol ing the method described in Nguyen et al. (2013). Protein concen- of each primer. The specific primer set (qACX2 F and qACX2 R) trations were determined spectrophotometrically at 280 nm using was used to amplify a 135-bp CrACX2 cDNA fragment. The ampli- a BCA protein assay kit (Bio-Rad). For immunoblot, approximately fication conditions were as follows: 95°C for 10 min, followed by 15 lg of proteins were separated on 12% SDS-PAGE, transferred 45 cycles of 10 sec at 95°C, 15 sec at 60°C and 10 sec at 72°C. The to a nitrocellulose membrane (Sigma-Aldrich) using the semidry specificity of PCR amplifications was checked by a melting curve transfer technique and immunoblotted with specific polyclonal program (95°C for 5 sec, 65°C for 10 sec, continuous acquisition at rabbit anti-V5 primary antibodies (1/5000) (eBioscience, https:// 95°C and cooling at 40°C for 30 sec) and analyzed by electrophore- www.ebioscience.com/) for detecting the V5-tagged proteins. The sis on a 2.0% agarose gel. The data were normalized as relative chemiluminescent substrate CDP-Star (Roche) was used to detect values with respect to the housekeeping gene RACK1 immunoreactive proteins by utilizing secondary antibodies conju- (Cre06.g278222) (Nguyen et al., 2013). gated to the anti-rabbit IgG-fluorescein isothiocyanate (FITC) (1/ 20 000). Blots were imaged using the G:BOX Chemi XRQ system RNA extraction, reverse transcription (RT)-PCR, gene (Syngene, http://www.syngene.com/). cloning and Chlamydomonas transformation Lipid extraction and analyses Total RNA was isolated as previously described (Nguyen et al., 2013). Purified total RNA was treated with DNase I (Ambion, Invit- Freshly grown cells were harvested and either quenched immedi- rogen, www.thermofisher.com) to remove the contaminated resid- ately in boiling isopropanol or kept frozen in liquid nitrogen. A hot ual genomic DNA, and was then purified with Nucleospin RNA isopropanol method was used to extract the total lipids (Legeret Clean Up (Macherey Nagel, http://www.mn-net.com/). First-strand et al., 2016). Extracted lipids were dried under a stream of N2 and cDNA was synthesized from total RNA (1 lg) with the SuperScript then dissolved either in a mixture of chloroform:methanol (2:1 by VILO cDNA Synthesis Kit (Life Technologies, www.thermofisher. volume) for TLC or in a solvent mixture of acetonitrile:iso- com). For RT-PCR, a fragment of the cDNA coding region of propanol:10 mM ammonium acetate (65:30:5 by volume) for LC- CrACX2 was amplified using the gene-specific primers ACX-F1 MS/MS. Depending on the analysis required, lipid standards and ACX-R1. CBLP (Cre06.g278222) was amplified using primers TAG51:0(17:0/17:0/17:0) and PtdEtn34:0 (17:0/17:0) (Sigma-

Aldrich), were added just before the solvent extraction step. centrifugation at 4000 g at 22°C for 30 min and re-suspended in Detailed TLC procedures and quantifications have been described 25 ml of lysis buffer [50 mM 2-amino-2-(hydroxymethyl)-1,3-propa- in Legeret et al. (2016) and Siaut et al. (2011). nediol (TRIS) pH 8.0, 500 mM NaCl, 10 mM imidazole, 10% glyc- 1 1 erol, 10 lgml DNase, 20 mM MgSO4 and 0.25 mg ml Nile red staining of LDs and confocal microscopy lysozyme]. The cells were lysed by sonication three times with a 10-sec interval cycle. The cell debris was centrifuged at 21 000 g Lipid droplets in live cells were first stained with a Nile red solu- for 10 min in a microcentrifuge and the protein purified with a His tion (1 lgml1), kept for 10 min in the dark and then imaged with GraviTrap kit (GE Healthcare, http://www3.gehealthcare.com/) a confocal laser scanning microscope (TCS SP2, Leica, http:// according to the manufacturer’s instructions. www.leica-microsystems.com/). A 639 oil immersion objective was used throughout all the imaging work. Cells were excited The eluate was used for determining ACX activity as described using a laser excitation line at 488 nm, emission for the Nile red previously (Hryb and Hogg, 1979; Hooks et al., 1999). Briefly, the reaction mixture (50 lM an acyl-CoA substrate, 50 lM FAD, 5 lgof signal was collected between 554 and 599 nm and the chlorophyll purified enzyme and 175 mM TRIS pH 7.4) were incubated at 30°C autofluorescence signal was collected between 650 and 714 nm. for 30 min. The acyl-CoA substrates tested are lauroyl-CoA (C12:0- Pseudo colors were obtained for all images using ZEN (Carl Zeiss, http://www.zeiss.com/) software. CoA), palmitoyl-CoA (C16:0-CoA), stearoyl-CoA (C18:0-CoA), oleoyl-CoA (C18:1-CoA) and arachidoyl-CoA (C20:0-CoA). Except Subcellular localization of CrACX2 for the C20:0-CoA (from Avantis Polar, https://avantilipids.com/), all other acyl-CoAs were ordered from Sigma-Aldrich. Stock solu- Constructs for protein subcellular localization. CrACX2 was tions were made for each acyl-CoA by dissolving it into a 100 mM cloned in frame to the C-terminus of an mCherry cassette. The 2-(N-morpholino)ethanesulfonic acid (MES) buffer adjusted to pH full-length CrACX2 open reading frame (ORF) was amplified by 5.8 (for long-term storage); for reactions, pH was adjusted to 7.4. KpnI-ACXf and XbaI-ACXr primers, and inserted into the The final concentration of the acyl-CoA in the stock solution was pChlamy3 plasmid as a KpnI/XbaI fragment, resulting in the determined by measuring the absorbance at 260 nm (E260 = 16.4 â recombinant plasmid pChlamy3-ACX2. The mCherry ORF was at pH 7.4). For H2O2 measurement, an Amplex red hydrogen per- amplified by InFUmCherryF and InFUmCherryR from the pBR9 oxide/peroxidase assay kit (Invitrogen) was used according to the mCherry Cr plasmid (Rasala et al., 2013). The mCherry ORF was manufacturer’s instructions, and a Xenius XC spectrofluorometer fused in frame to the N-terminus or C-terminus of CrACX2 to con- (SAFAS, http://www.safas.com/) was used to measure the fluores- struct the recombined plasmid mCherry-ACX2, or ACX2-mCherry, cence emission at 580 nm (excitation at 540 nm) during the linear â by an In-Fusion HD cloning kit (Clontech, http://www.clontech.c kinetic range. om/). The known PTS2 (MADPLNRIQKIASHLDPAKPRKFKVA) for CrMDH2 was cloned using the primers NdeI-PTSf and XhoI-PTSr ACCESSION NUMBERS and integrated into pBR9 GFP plasmid as a NdeI/XhoI fragment (Rasala et al., 2013). The AphVIII gene was amplified from the Sequence data presented in this article can be found for pSI103 plasmid (Sizova et al., 2001) by Aph8-infusF and Aph8- C. reinhardtii genes in Phytozome (https://phytozome.jgi.d infusR primers and cloned into PTS2(MDH2)-GFP plasmid. The oe.gov/pz/portal.html#!info?alias=Org_Creinhardtii) with mCherry-ACX2, or ACX2-mCherry, and PTS2(MDH2)-GFP plasmids gene identifications as: CrACX2 (Cre05.g232002.t2.1), were co-transformed to the dw15 strain, and the double transfor- CrMDH2 (Cre10.g423250) and RACK1 (Cre06.g278222). The mants were selected on agar plates containing both hygromycin B (15 lgml1) and paromomycin (10 lgml1). Arabidopsis AtACX2 protein (At5g65110) and AtACX1 (At4g16760) can be found at The Arabidopsis Information Fluorescence microscopy. For live cell fluorescence micro- Resource (https://www.arabidopsis.org/). scopy, representative clones of co-transformants were grown in ACKNOWLEDGEMENTS TAP medium to log phase. Images were captured on a confocal microscope (TCS SP2; Leica). The cells were excited with a 488-nm We thank Pascaline Auroy, Cyril Aselmeyer, Audrey Beynel and laser line. The following filters were used: for mCherry, excitation Marie-Christine Thibaud for excellent technical assistance. We at 561 nm and emission 610–630 nm; for GFP, excitation at also thank Dr Peter Eastmond for useful discussions on acyl- 458 nm and emission at 500–530 nm; and for chlorophyll autofluo- CoA oxidase activity assays. Yuanxue Liang acknowledges the rescence, excitation at 645 nm and emission at 685–720 nm. China Scholarship Council (CSC) for a Postgraduate Award. Work in the authors’ laboratory is supported by A*MIDEX pro- Protein expression in E. coli and in vitro assays for ACX ject and ANR MUsCA. Support for the microscopy equipment activity was provided by the Region Provence Alpes Cote^ d’Azur, the Conseil Gen eral des Bouches-du-Rhone,^ the French Ministry of The codon-optimized CrACX2 cDNA was synthesized by GenArts Research, the CNRS and the CEA. We thank the European Union (Invitrogen). It was then amplified using Lic07eCrACXf and Lic07e- Regional Developing Fund (ERDF), the Region Provence Alpes CrACXr primers. The fragment was cloned into plasmid Lic07 with Cote^ d’Azur, the French Ministry of Research and the CEA for â an In-Fusion HD cloning kit (Clontech) and then transformed into funding the HelioBiotec platform. The authors declare no con- E. coli strain BL21. The AtACX2 was cloned using the primers flicts of interest. AtACX2-F and AtACX2-R. For protein expression, pre-cultures were grown in 50 ml of standard Luria broth overnight at 37°C. A AUTHOR CONTRIBUTIONS small aliquot (10 ml) was added to 500 ml of fresh Terrific Broth FK, FB, GP and YL-B designed the research; FK, YL, BL, AB- and the cells grown at 37°C to an optical density (OD600 nm)of approximately 1.0. Isopropyl b-D-1-thiogalactopyranoside (IPTG) A, SB and RPH performed experiments and analyzed data; was added to a final concentration of 0.5 mM and cultures were JAN contributed new analytical tools; and FK, FB, GP and ° incubated overnight at 18 C. The cells were harvested by YL-B wrote the paper.

SUPPORTING INFORMATION Graham, I.A. (2008) Seed storage oil mobilization. Annu. Rev. Plant Biol. 59, 115–142. Additional Supporting Information may be found in the online ver- Graham, I.A. and Eastmond, P.J. (2002) Pathways of straight and sion of this article. branched chain fatty acid catabolism in higher plants. Prog. Lipid Res. – Figure S1. Constructs used for genetic complementation and pro- 41, 156 181. Grossman, A., Croft, M., Gladyshev, V., Merchant, S., Posewitz, M., Proch- tein subcellular localization studies. nik, S. and Spalding, M. (2007) Novel metabolism in Chlamydomonas Figure S2. Examples of some of the screening results for genetic through the lens of genomics. Curr. Opin. Plant Biol. 10, 190–198. complementation. Gu, L., Jones, A.D. and Last, R.L. (2010) Broad connections in the Arabidop- Figure S3. Dynamic changes in absolute triacylglycerol content in sis seed metabolic network revealed by metabolite profiling of an amino the two mutant alleles and their corresponding wild-type strain. acid catabolism mutant. Plant J. 61, 579–590. Figure S4. Oleic acid feeding test. Haddouche, R., Delessert, S., Sabirova, J., Neuveglise, C., Poirier, Y. and Figure S5. Constructs for expression in Escherichia coli. Nicaud, J.-M. (2010) Roles of multiple acyl-CoA oxidases in the routing Figure S6. Quantification of acyl-CoAs by LC-MS/MS. of carbon flow towards b-oxidation and polyhydroxyalkanoate biosyn- – Figure S7. Growth kinetics under mixotrophic and photoau- thesis in Yarrowia lipolytica. FEMS Yeast Res. 10, 917 927. Harris, E. (2001) Chlamydomonas as a model organism. Annu. Rev. Plant totrophic conditions. Physiol. Plant Mol. Biol. 52, 363–406. Figure S8. Fatty acid and lipid composition of the wild type and Hayashi, Y. and Shinozaki, A. (2012) Visualization of microbodies in mutant cells during optimal growth. Chlamydomonas reinhardtii. J. Plant. Res. 125, 579–586. Table S1. The family of acyl-CoA oxidases in Chlamydomonas Hayashi, Y., Sato, N., Shinozaki, A. and Watanabe, M. (2015) Increase in reinhardtii. peroxisome number and the gene expression of putative glyoxysomal Table S2. All primer sequences used in this study. enzymes in Chlamydomonas cells supplemented with acetate. J. Plant. Res. 128, 177–185. Methods S1. Acyl-CoA profiling. Hernandez, M.L., Whitehead, L., He, Z., Gazda, V., Gilday, A., Kozhevnikova, REFERENCES E., Vaistij, F.E., Larson, T.R. and Graham, I.A. (2012) A cytosolic acyltransferase contributes to triacylglycerol synthesis in sucrose-rescued Bernhardt, K., Wilkinson, S., Weber, A.P.M. and Linka, N. (2012) A peroxiso- Arabidopsis seed oil catabolism mutants. Plant Physiol. 160, 215–225. mal carrier delivers NAD+ and contributes to optimal fatty acid degrada- Hooks, M.A., Kellas, F. and Graham, I.A. (1999) Long-chain acyl-CoA oxi- – tion during storage oil mobilization. Plant J. 69,1–13. dases of Arabidopsis. Plant J. 20,1 13. Cagnon, C., Mirabella, B., Nguyen, H.M., Beyly-Adriano, A., Bouvet, S., Hryb, D.J. and Hogg, J.F. (1979) Chain length specificities of peroxisomal Cuine, S., Beisson, F., Peltier, G. and Li-Beisson, Y. (2013) Development and mitochondrial beta-oxidation in rat liver. Biochem. Biophys. Res. – of a forward genetic screen to isolate oil mutants in the green microalga Commun. 87, 1200 1206. Chlamydomonas reinhardtii. Biotechnol. Biofuels, 6, 178. Hu, J., Baker, A., Bartel, B., Linka, N., Mullen, R.T., Reumann, S. and Zol- Daum, G., Wagner, A., Czabany, T. and Athenstaedt, K. (2007) Dynamics of man, B.K. (2012) Plant peroxisomes: biogenesis and function. Plant Cell, – neutral lipid storage and mobilization in yeast. Biochimie, 89, 243–248. 24, 2279 2303. Dmochowska, A., Dignard, D., Maleszka, R. and Thomas, D.Y. (1990) Struc- Jaworski, J. and Cahoon, E.B. (2003) Industrial oils from transgenic plants. – ture and transcriptional control of the Saccharomyces cerevisiae POX1 Curr. Opin. Plant Biol. 6, 178 184. gene encoding acylcoenzyme A oxidase. Gene, 88, 247–252. Kato, J., Yamahara, T., Tanaka, K., Takio, S. and Satoh, T. (1997) Characteri- Eastmond, P.J. (2007) Monodehydroascorbate reductase 4 is required for zation of catalase from green algae Chlamydomonas reinhardtii. J. Plant – seed storage oil hydrolysis and postgerminative growth in Arabidopsis. Physiol. 151, 262 268. Plant Cell, 19, 1376–1387. Kaur, N., Reumann, S. and Hu, J. (2009) Peroxisome biogenesis and func- Eastmond, P.J., Hooks, M. and Graham, I.A. (2000) The Arabidopsis acyl- tion. Arabidopsis Book, 7, e0123. CoA oxidase gene family. Biochem. Soc. Trans. 28, 755–757. Kessel-Vigelius, S.K., Wiese, J., Schroers, M.G., Wrobel, T.J., Hahn, F. and Eaton, S. (2002) Control of mitochondrial beta-oxidation flux. Prog. Lipid Linka, N. (2013) An engineered plant peroxisome and its application in Res. 41, 197–239. biotechnology. Plant Sci. 210, 232–240. Fan, J.L., Andre, C. and Xu, C.C. (2011) A chloroplast pathway for the de Kim, J.J. and Miura, R. (2004) Acyl-CoA dehydrogenases and acyl-CoA oxi- novo biosynthesis of triacylglycerol in Chlamydomonas reinhardtii. FEBS dases. Structural basis for mechanistic similarities and differences. Eur. – Lett. 585, 1985–1991. J. Biochem. 271, 483 493. Fan, J., Yan, C. and Xu, C. (2013a) Phospholipid:diacylglycerol acyltrans- Klein, A.T.J., van den Berg, M., Bottger, G., Tabak, H.F. and Distel, B. (2002) ferase-mediated triacylglycerol biosynthesis is crucial for protection Saccharomyces cerevisiae acyl-CoA oxidase follows a novel, non-PTS1, against fatty acid-induced cell death in growing tissues of Arabidopsis. import pathway into peroxisomes that is dependent on Pex5p. J. Biol. Plant J. 76, 930–942. Chem. 277, 25011–25019. Fan, J., Yan, C., Zhang, X. and Xu, C. (2013b) Dual role for phospholipid:dia- Kong, F., Yamasaki, T., Kurniasih, S.D., Hou, L., Li, X., Ivanova, N., Okada, cylglycerol acyltransferase: enhancing fatty acid synthesis and diverting S. and Ohama, T. (2015) Robust expression of heterologous genes by fatty acids from membrane lipids to triacylglycerol in Arabidopsis leaves. selection marker fusion system in improved Chlamydomonas strains. J. Plant Cell, 25, 3506–3518. Biosci. Bioeng. 120, 239–245. € Fan, J., Yan, C., Roston, R., Shanklin, J. and Xu, C. (2014) Arabidopsis lipins, Kunz, H.-H., Scharnewski, M., Feussner, K., Feussner, I., Flugge, U.-I., Fulda, PDAT1 acyltransferase, and SDP1 triacylglycerol lipase synergistically M. and Gierth, M. (2009) The ABC transporter PXA1 and peroxisomal b- direct fatty acids toward b-oxidation, thereby maintaining membrane oxidation are vital for metabolism in mature leaves of Arabidopsis dur- lipid homeostasis. Plant Cell, 26, 4119–4134. ing extended darkness. Plant Cell, 21, 2733–2749. Gasser, B., Prielhofer, R., Marx, H., Maurer, M., Nocon, J., Steiger, M., Pux- Lauersen, K.J., Willamme, R., Coosemans, N., Joris, M., Kruse, O. and baum, V., Sauer, M. and Mattanovich, D. (2013) Pichia pastoris: protein Remacle, C. (2016) Peroxisomal microbodies are at the crossroads of production host and model organism for biomedical research. Future acetate assimilation in the green microalga Chlamydomonas reinhardtii. – Microbiol. 8, 191–208. Algal Res. 16, 266 274. Gerhardt, B. (1992) Fatty acid degradation in plants. Prog. Lipid Res. 31,417–446. Lea, W., Abbas, A.S., Sprecher, H., Vockley, J. and Schulz, H. (2000) Long- Gonzalez-Ballester, D., Pootakham, W., Mus, F. et al. (2011) Reverse genet- chain acyl-CoA dehydrogenase is a key enzyme in the mitochondrial ics in Chlamydomonas: a platform for isolating insertional mutants. Plant beta-oxidation of unsaturated fatty acids. Biochim. Biophys. Acta, 1485, – Methods, 7,24–37. 121 128. Gonzalez-Ballester, D., de Montaigu, A., Higuera, J.J., Galvan, A. and Ledesma-Amaro, R. and Nicaud, J.-M. (2016) Yarrowia lipolytica as a Fernandez, E. (2005) Functional genomics of the regulation of the nitrate biotechnological chassis to produce usual and unusual fatty acids. Prog. – assimilation pathway in Chlamydomonas. Plant Physiol. 137, 522–533. Lipid Res. 61,40 50.

Legeret, B., Schulz-Raffelt, M., Nguyen, H.M., Auroy, P., Beisson, F., Peltier, Rambold, A.S., Cohen, S. and Lippincott-Schwartz, J. (2015) Fatty acid traf- G., Blanc, G. and Li-Beisson, Y. (2016) Lipidomic and transcriptomic anal- ficking in starved cells: regulation by lipid droplet lipolysis, autophagy, yses of Chlamydomonas reinhardtii under heat stress unveil a direct and mitochondrial fusion dynamics. Dev. Cell, 32, 678–692. route for the conversion of membrane lipids into storage lipids. Plant Rasala, B.A., Barrera, D.J., Ng, J., Plucinak, T.M., Rosenberg, J.N., Weeks, Cell Environ. 39, 834–847. D.P., Oyler, G.A., Peterson, T.C., Haerizadeh, F. and Mayfield, S.P. (2013) Li, X., Zhang, R., Patena, W. et al. (2016) An indexed, mapped mutant Expanding the spectral palette of fluorescent proteins for the green library enables reverse genetics studies of biological processes in microalga Chlamydomonas reinhardtii. Plant J. 74, 545–556. Chlamydomonas reinhardtii. Plant Cell, 28, 367–387. Rhodin, J.A.G. (1954) Correlation of Ultrastructural Organization: and Func- Li-Beisson, Y., Beisson, F. and Riekhof, W. (2015) Metabolism of acyl-lipids tion in Normal and Experimentally Changed Proximal Convoluted Tubule in Chlamydomonas reinhardtii. Plant J. 82, 504–522. Cells of the Mouse Kidney: An Electron Microscopic Study. Stockholm: Lisenbee, C.S., Heinze, M. and Trelease, R.N. (2003) Peroxisomal ascorbate Dept. of Anatomy, Karolinska Institutet. peroxidase resides within a subdomain of rough endoplasmic reticulum Shimogawara, K., Fujiwara, S., Grossman, A. and Usuda, H. (1998) High- in wild-type Arabidopsis cells. Plant Physiol. 132, 870–882. efficiency transformation of Chlamydomonas reinhardtii by electropora- Lisenbee, C.S., Lingard, M.J. and Trelease, R.N. (2005) Arabidopsis peroxi- tion. Genetics, 148, 1821–1828. somes possess functionally redundant membrane and matrix isoforms Shinozaki, A., Sato, N. and Hayashi, Y. (2005) Reporter gene assay of target- of monodehydroascorbate reductase. Plant J. 43, 900–914. ing signal-like elements and the expression pattern of peroxisomal Marchesini, S. and Poirier, Y. (2003) Futile cycling of intermediates of fatty enzymes in Chlamydomonas reinhardtii. Plant Cell Physiol. 46, 823–829. acid biosynthesis toward peroxisomal b-oxidation in Saccharomyces Shinozaki, A., Sato, N. and Hayashi, Y. (2009) Peroxisomal targeting signals cerevisiae. J. Biol. Chem. 278, 32596–32601. in green algae. Protoplasma, 235,57–66. Merchant, S.S., Prochnik, S.E., Vallon, O. et al. (2007) The Chlamydomonas Siaut, M., Cuine, S., Cagnon, C. et al. (2011) Oil accumulation in the model genome reveals the evolution of key animal and plant functions. Science, green alga Chlamydomonas reinhardtii: characterization, variability 318, 245–250. between common laboratory strains and relationship with starch Moire, L., Rezzonico, E., Goepfert, S. and Poirier, Y. (2004) Impact of unu- reserves. BMC Biotechnol. 11,7. sual fatty acid synthesis on futile cycling through b-oxidation and on Sizova, I., Fuhrmann, M. and Hegemann, P. (2001) A Streptomyces rimosus gene expression in transgenic plants. Plant Physiol. 134, 432–442. aphVIII gene coding for a new type phosphotransferase provides stable Napier, J.A. (2007) The production of unusual fatty acids in transgenic antibiotic resistance to Chlamydomonas reinhardtii. Gene, 277, 221–229. plants. Annual Review of Plant Biology Palo Alto: Annual Reviews, pp. Slocombe, S., Cornah, J., Pinfield-Wells, H., Soady, K., Zhang, Q., Gilday, 295–319. A., Dyer, J. and Graham, I. (2009) Oil accumulation in leaves directed by Nguyen, H.M., Cuine, S., Beyly-Adriano, A., Legeret, B., Billon, E., Auroy, P., modification of fatty acid breakdown and lipid synthesis pathways. Plant Beisson, F., Peltier, G. and Li-Beisson, Y. (2013) The green microalga Biotechnol. J. 7, 694–703. Chlamydomonas reinhardtii has a single x-3 fatty acid desaturase that Stabenau, H., Winkler, U. and Saftel, W. (1984) Enzymes of beta-oxidation localizes to the chloroplast and impacts both plastidic and extraplastidic in different types of algal microbodies. Plant Physiol. 75, 531–533. membrane lipids. Plant Physiol. 163, 914–928. Stabenau, H., Winkler, U. and Saftel, W. (1989) Compartmentation of perox- Pinfield-Wells, H., Rylott, E.L., Gilday, A.D., Graham, S., Job, K., Larson, T.R. isomal enzymes of the group of Prasinophyceae – occurrence of possible and Graham, I.A. (2005) Sucrose rescues seedling establishment but not microbodies without catalase. Plant Physiol. 90, 754–759. germination of Arabidopsis mutants disrupted in peroxisomal fatty acid Stevens, D.R., Purton, S. and Rochaix, J.D. (1996) The bacterial phleomycin catabolism. Plant J. 43, 861–872. resistance geneble as a dominant selectable marker in Chlamydomonas. Plancke, C., Vigeolas, H., Hohner, R. et al. (2014) Lack of isocitrate lyase in Mol. Gen. Genet. 251,23–30. Chlamydomonas leads to changes in carbon metabolism and in the Tardif, M., Atteia, A., Specht, M. et al. (2012) PredAlgo, a new subcellular response to oxidative stress under mixotrophic growth. Plant J. 77, 404– localization prediction tool dedicated to green algae. Mol. Biol. Evol. 29, 417. 3625–3639. Poirier, Y. (2002) Polyhydroxyalknoate synthesis in plants as a tool for Theodoulou, F.L. and Eastmond, P.J. (2012) Seed storage oil catabolism: a biotechnology and basic studies of lipid metabolism. Prog. Lipid Res. 41, story of give and take. Curr. Opin. Plant Biol. 15, 322–328. 131–155. Trentacoste, E.M., Shrestha, R.P., Smith, S.R., Gle, C., Hartmann, A.C., Poirier, Y., Antonenkov, V.D., Glumoff, T. and Hiltunen, J.K. (2006) Peroxi- Hildebrand, M. and Gerwick, W.H. (2013) Metabolic engineering of lipid somal beta-oxidation – a metabolic pathway with multiple functions. Bio- catabolism increases microalgal lipid accumulation without compromis- chim. Biophys. Acta, 1763, 1413–1426. ing growth. Proc. Natl Acad. Sci. USA, 110, 19748–19753. Poliner, E., Panchy, N., Newton, L., Wu, G., Lapinsky, A., Bullard, B., Zien- Troncoso-Ponce, M.A., Cao, X., Yang, Z. and Ohlrogge, J.B. (2013) Lipid kiewicz, A., Benning, C., Shiu, S.H. and Farre, E.M. (2015) Transcriptional turnover during senescence. Plant Sci. 205–206,13–19. coordination of physiological responses in Nannochloropsis oceanica Vanroermund, C.W.T., Elgersma, Y., Singh, N., Wanders, R.J.A. and Tabak, CCMP1779 under light/dark cycles. Plant J. 83, 1097–1113. H.F. (1995) The membrane of peroxisomes in Saccharomyces cerevisae Pracharoenwattana, I., Cornah, J.E. and Smith, S.M. (2005) Arabidopsis per- is impermeable to NAD(H) and acetym-CoA under in vivo conditions. oxisomal citrate synthase is required for fatty acid respiration and seed EMBO J. 14, 3480–3486. germination. Plant Cell, 17, 2037–2048. Winkler, U., Saftel, W. and Stabenau, H. (1988) Beta-oxidation of fatty acids Pracharoenwattana, I., Zhou, W.X. and Smith, S.M. (2010) Fatty acid beta- in algae - localization of thiolase and acyl-CoA oxidizing enzymes in 3 dif- oxidation in germinating Arabidopsis seeds is supported by peroxisomal ferent organisms. Planta, 175,91–98. hydroxypyruvate reductase when malate dehydrogenase is absent. Plant Zones, J.M., Blaby, I.K., Merchant, S.S. and Umen, J.G. (2015) High-resolu- Mol. Biol. 72, 101–109. tion profiling of a synchronized diurnal transcriptome from Chlamy- Purdue, P.E. and Lazarow, P.B. (2001) Peroxisome biogenesis. Annu. Rev. domonas reinhardtii reveals continuous cell and metabolic Cell Dev. Biol. 17, 701–752. differentiation. Plant Cell, 27, 2743–2769.

Interorganelle Communication: Peroxisomal MALATE

DEHYDROGENASE 2 Connects Lipid Catabolism to Photosynthesis

through Redox Coupling in Chlamydomonas

Fantao Konga, Adrien Burlacota,1, Yuanxue Lianga,1, Bertrand Légereta, Saleh Alseekhb,c, Yariv Brotmanb, Alisdair R Fernieb,c, Anja Krieger-Liszkayd, Fred Beissona, Gilles Peltiera, Yonghua Li-Beissona,2* aCEA, CNRS and Aix-Marseille University, Institute of Biosciences and Biotechnologies of Aix-Marseille, UMR 7265, CEA Cadarache, Saint-Paul-lez Durance F-13108, France b Max Planck Institute of Molecular Plant Physiology, 14476 Potsdam-Golm, Germany c Center of Plant System Biology and Biotechnology, 4000 Plovdiv, Bulgaria d Institute for Integrative Biology of the Cell, CEA Saclay, CNRS, University Paris-Sud, University Paris-Saclay, 91191 Gif-sur-Yvette Cedex, France

In this paper, my contributions are: I - did RT-PCR - performed malate dehydrogenase activity assays - tested growth parameters under various cultivation conditions - carried out lipid analysis during a day/night cycle

Interorganelle Communication: Peroxisomal MALATE DEHYDROGENASE2 Connects Lipid Catabolism to Photosynthesis through Redox Coupling in Chlamydomonas[OPEN]

Fantao Kong,a Adrien Burlacot,a Yuanxue Liang,a Bertrand Légeret,a Saleh Alseekh,b,c Yariv Brotman,b Alisdair R. Fernie,b,c Anja Krieger-Liszkay,d Fred Beisson,a Gilles Peltier,a and Yonghua Li-Beissona,1 aAix Marseille University, CEA, CNRS, BIAM, Laboratoire de Bioénergétique et Biotechnologie des Bactéries et Microalgues, F-13108 Saint Paul-Lez-Durance, France bMax Planck Institute of Molecular Plant Physiology, 14476 Potsdam-Golm, Germany cCenter of Plant System Biology and Biotechnology, 4000 Plovdiv, Bulgaria dInstitute for Integrative Biology of the Cell, CEA Saclay, CNRS, University Paris-Sud, University Paris-Saclay, 91191 Gif-sur-Yvette Cedex, France ORCID IDs: 0000‑0003‑2847‑2838 (F.K.); 0000‑0001‑7434‑6416 (A.B.); 0000‑0002‑1690‑8094 (Y.L.); 0000‑0002‑0957‑4700 (B.L.); 0000‑0003‑2067‑5235 (S.A.); 0000‑0001‑9000‑335X (A.R.F.); 0000‑0001‑7141‑4129 (A.K.-L.); 0000‑0001‑9995‑7387 (F.B.); 0000‑0002‑2226‑3931 (G.P.); 0000‑0003‑1064‑1816 (Y.L.-B.)

Plants and algae must tightly coordinate photosynthetic electron transport and metabolic activities given that they often face fluctuating light and nutrient conditions. The exchange of metabolites and signaling molecules between organelles is thought to be central to this regulation but evidence for this is still fragmentary. Here, we show that knocking out the peroxisome-located MALATE DEHYDROGENASE2 (MDH2) of Chlamydomonas reinhardtii results in dramatic alterations not only in peroxisomal fatty acid breakdown but also in chloroplast starch metabolism and photosynthesis. mdh2 mutants accumulated 50% more storage lipid and 2-fold more starch than the wild type during nitrogen deprivation. In parallel, mdh2 showed increased photosystem II yield and photosynthetic CO2 fixation. Metabolite analyses revealed a >60% reduction in malate, together with increased levels of NADPH and H2O2 in mdh2. Similar phenotypes were found upon high light exposure.

Furthermore, based on the lack of starch accumulation in a knockout mutant of the H2O2-producing peroxisomal ACYL-COA

OXIDASE2 and on the effects of H2O2 supplementation, we propose that peroxisome-derived H2O2 acts as a regulator of chloroplast metabolism. We conclude that peroxisomal MDH2 helps photoautotrophs cope with nitrogen scarcity and high light by transmitting the redox state of the peroxisome to the chloroplast by means of malate shuttle- and H2O2-based redox signaling.

INTRODUCTION flow, the water-to-water cycle, 2O photoreduction processes, and chlororespiration) have been identified to play roles in dissi- Photoautotrophs convert light energy into reducing equivalents pation of photoreductant and/or reequilibration of the NADPH/ (NADPH) and phosphorylating power (ATP), which are used to ATP ratio (Peltier et al., 2010; Curien et al., 2016; Saroussi et al., drive the metabolic reactions of CO assimilation. Coordinating 2 2017). Recently, it has been shown that chloroplast redox bal- photosynthetic electron transport activities to downstream met- ance can also be achieved through export of excess reducing abolic needs is essential for cell survival and growth because equivalents to mitochondria in green algae (Dang et al., 2014) excess production of reducing power may result in an overre- and diatoms (Bailleul et al., 2015) yet the detailed molecular duction of the photosynthetic electron transport chain, which mechanisms remain to be elucidated. Moreover, alterations of may lead to photooxidative damage (Niyogi, 2000). Multiple mitochondrial metabolism have also been shown to influence strategies have therefore evolved to facilitate fine-tuning of pho- photosynthetic performance in plants and algae (Sweetlove tosynthesis and allow plants and algae to rapidly acclimate to et al., 2006; Nunes-Nesi et al., 2011; Massoz et al., 2015; Larosa natural environments where nutrient, light, and temperature can et al., 2018). Thus current knowledge on chloroplast redox poise change frequently (Saroussi et al., 2017). Several chloroplast- is centered on the dissipation of excess reducing equivalents located alternative electron pathways (notably, cyclic electron through chloroplast-based processes or through collaboration between chloroplast and mitochondria. 1Address correspondence to [email protected]. Alongside chloroplasts and mitochondria, peroxisomes are a The author responsible for distribution of materials integral to the findings further subcellular compartment involved in energetic metabo- presented in this article in accordance with the policy described in lism. The peroxisome was originally defined as an organelle that the Instruction for Authors (www.plantcell.org) is: Yonghua Li-Beisson carries out oxidative reactions leading to production of H O , a ([email protected]). 2 2 [OPEN]Articles can be viewed without a subscription. reactive oxygen species (ROS) that can cause oxidative damage www.plantcell.org/cgi/doi/10.1105/tpc.18.00361 if present in excess (Erickson et al., 2015; Dietz et al., 2016). In Peroxisome Redox Status and Chloroplast Metabolism 1825

plant peroxisomes, the major sources of H2O2 are photorespira- 2004). MDHs are ubiquitous enzymes, and each subcellular tion and β-oxidation of fatty acids (FAs). However, in algae such compartment usually contains at least one isoform (Mettler and as Chlamydomonas reinhardtii that possess a CO2 concentrating Beevers, 1980; Scheibe, 2004). With the exception of the ma- mechanism, photorespiration is negligible and does not produce jor chloroplast isoform which is NADP+-dependent (EC1.1.1.82)

H2O2 due to the absence of glyoxylate oxidase (Aboelmy and and so far the best characterized (Musrati et al., 1998; Miginiac- Peterhansel, 2014; Hagemann et al., 2016). Algal FA β-oxidation, Maslow and Lancelin, 2002; Lemaire et al., 2005), all the other which has recently been demonstrated to be the major path- MDHs require NAD+ as a cofactor (EC1.1.1.37). Despite their way for FA catabolism in Chlamydomonas (Kong et al., 2017), potential importance in cellular redox homeostasis, none of the + is therefore the major contributor to H2O2 formation in peroxi- algal NAD -dependent MDH has been studied. The Chlamydo- + somes. H2O2 is produced at the first step ofβ -oxidation, which monas genome encodes four putative NAD -dependent MDHs is catalyzed by acyl-CoA oxidase (ACX). H2O2 can also be pro- (Merchant et al., 2007), and MDH2 has been localized to per- duced in chloroplasts through the Mehler reaction as a “safety oxisomes (Hayashi and Shinozaki, 2012; Lauersen et al., 2016; valve” to photochemical reactions and has been shown to play Kong et al., 2017). a signaling role if present at sublethal levels (Dietz et al., 2016). In this study, through isolation and characterization of two However, it remains to be established whether peroxisome- Chlamydomonas mutants deficient in peroxisomal NAD+- derived H2O2 can also play a signaling role. dependent MALATE DEHYDROGENASE2 (MDH2), we show

In addition to H2O2 and its end-product acetyl-CoA, FA that MDH2 plays a key role in the reverse coupling of the

β-oxidation generates one molecule of NADH (1:1:1) through the redox/H2O2 signal from the peroxisome to the chloroplast. The 3-hydoxyacyl-CoA dehydrogenase activity of the multifunctional mdh2 mutant cells are compromised in triacylglycerol (TAG) protein (MFP-DH). Studies in yeasts and plants show that reox- breakdown and display increased TAG and starch accumula- idation of peroxisomal NADH must occur inside the organelle tion during both photoautotrophic nitrogen (N) deprivation and because the peroxisomal membrane is impermeable to NAD+ high light (HL) exposure. We propose that the increased carbon (van Roermund et al., 1995). NADH is an essential electron donor storage is a result of more active photosynthesis and increased for numerous biochemical reactions occurring in different cellu- levels of NADPH and H2O2. MDH2 thus connects peroxisomal lar compartments; therefore, NADH trafficking and homeostasis FA catabolism to synthesis of starch and lipid reserves and pho- requires tight regulation and coordination (Mettler and Beevers, tosynthetic electron transport in the chloroplast by transmitting 1980; Scheibe, 2004). By catalyzing the reversible conversion of the redox state of the peroxisome to the chloroplast. Therefore, NADH to NAD+ via reduction of oxaloacetate to malate, which this study provides unanticipated evidence for the role of peroxi- can be shuttled across subcellular membranes, malate dehy- somes in helping photoautotrophs to cope with N scarcity and drogenases (MDHs) play a key role in intracellular trafficking HL, which are widespread phenomena in many terrestrial and of reducing equivalents (Mettler and Beevers, 1980; Scheibe, aquatic environments. 1826 The Plant Cell

RESULTS

Mutants Knocked out for MDH2 Are Compromised in Lipid Catabolism

In Chlamydomonas, significant levels of TAG accumulate during N deprivation, and upon N resupply, TAGs are degraded to release carbon and energy contained in acyl groups to support cell division and growth (Liu and Benning, 2013; Schmollinger et al., 2014; Li-Beisson et al., 2015). To investigate the pathways and regulatory mechanisms of TAG homeostasis, a library of insertional mutants (∼4500) was screened for alterations in TAG content based on Nile red staining and flow cytometry (Cagnon et al., 2013). Among 80 mutants isolated, one mutant (Lb9G9) was found to be severely defective in TAG breakdown during the recovery phase after a period of mixotrophic N deprivation (Figure 1A). One day after N resupply while cells kept in the dark, the mutant cells retained ∼80% of the TAG amount accumulated before N resupply, whereas the parental strain (dw15) kept only ∼20% of the TAG accumulated. The observation that the mutant cells were 20% bigger than the wild type during the remobilization phase could not explain the fivefold higher TAG present in the mutant cells compared with the wild type at that stage (Figure 1F). The defect in TAG degradation could also be visualized by stronger BODIPY staining under confocal microscopy (Figure 1B). The experimental setup used in this study is depicted in Supplemental Figure 1A. The antibiotic resistance gene APHVIII was found to be inserted in the 4th exon of the Cre10.g423250 locus (Figure 1C), which codes for an isoform of NAD+-dependent MDH (MDH2/MDY2) (Merchant et al., 2007). The mutant was therefore designated as mdh2-1. DNA gel blot analyses revealed a single insertion in mdh2-1 (Supplemental Figure 2). RT-PCR performed on total Figure 1. Isolation of the mdh2-1 (Lb9G9) Mutant Impaired in TAG Break- RNAs isolated from exponentially grown cells showed the ab- down during N Recovery. sence of MDH2 transcripts in mdh2-1 (Figure 1D). A polyclonal (A) TAG content 2 d after N deprivation and 1 d after N resupply in mix- antibody raised against a synthetic peptide recognized in its pa- otrophically grown cells. rental strain (dw15) one polypeptide corresponding in size to the (B) Visualization of LDs in cells upon N resupply (1 d) by staining with predicted molecular mass of MDH2 (∼36 kD), which was absent BODIPY. in mdh2-1 (Figure 1E). Taken together, these data suggest that (C) The insertion site of the cassette APHVIII in mdh2-1. mdh2-1 is a null mutant. Activity measurements performed on (D) RT-PCR analysis. Chlamydomonas total protein extracts showed an ∼24% reduc- (E) Immunoblot analysis using anti-MDH2 antibodies. tion in the activity of NAD+-dependent MDH in mdh2-1 in com- (F) Cell volume determination. parison to dw15 during optimal growth (Supplemental Figure 3). Cells were first grown mixotrophically in N-deprived conditions. Af- The residual NAD+-dependent MDH activity is likely due to ex- ter N resupply, cells were kept in the dark in MM medium (1 d) then istence of at least three other NAD+-requiring MDH isoforms in were stained with BODIPY, and pseudocolors were used: lipid droplet in green and chlorophyll in red. The housekeeping gene used for RT- Chlamydomonas (Merchant et al., 2007) (Supplemental Table 1). PCR is RACK1. For immunoblots, samples were loaded at equal total To confirm that the mdh2-1 mutant phenotype is due to a de- protein amount and stained by Coomassie blue. RACK1, Receptor of fect in the expression of MDH2, the mutant was complemented activated protein C kinase 1. Values are the mean of biological replicates by transformation with a construct carrying a genomic copy (i.e., independent shaking flask cultures; n = 5, sd). Biological replicates of the full-length gene cloned in frame with the epitope tag refer to cells that were grown in independent flasks. Asterisks indicate V5 (Figure 2A). After screening ∼100 antibiotic resistant clones, statistically significant changes compared with the parental line dw15 by we obtained two independent complemented lines that showed paired-sample Student’s t test (*P ≤ 0.05 and **P ≤ 0.01). restoration in their capacity to remobilize oil (Figure 2B). The presence of the expressed protein MDH2 in these complemented lines was confirmed by immunoblot analysis using anti-V5 as g423250) in the CC4533 background from the Chlamydomo- well as anti-MDH2 antibodies (Figure 2C). nas Mutant Library (CLiP; https://www.chlamylibrary.org) (Li In parallel to the genetic complementation experiment, et al., 2016). This mutant, which we named mdh2-2, harbored an we also identified a second allele of the same locus (Cre10. insertion of the antibiotic resistance gene in the first exon of the Peroxisome Redox Status and Chloroplast Metabolism 1827

Figure 2. Genetic Complementation of mdh2-1 Mutant. (A) The construct used for genetic transformation. (B) Restoration of TAG levels to that of the wild type in the complemented lines. (C) Immunoblot detection of MDH2 in complemented lines using anti-V5 and anti-MDH2 antibodies. (D) TAG analysis during mixotrophic N deprivation. (E) Total FA quantification. (F) Starch analysis during mixotrophic N deprivation. TAG and starch contents were determined in N-deprived cells (2–3 d) and in N-resupplied cells (1 d). Two independent transformants (C1 and C2) were analyzed, and each line with three biological replicates (i.e., independent shaking flask cultures) and two technical replicates (i.e., different sampling from the same flask). Values are the mean of biological replicates (n = 3, sd). Asterisks indicate statistically significant changes compared with the control strains (dw15 and C1) by paired-sample Student’s t test (*P ≤ 0.05 and **P ≤ 0.01). Cells were grown in mixotrophic conditions under constant light. PHsp70A/RbcS2, the heat shock protein 70A and Rubisco small subunit promoter; ble, antibiotic resistance marker gene; TRbcS2, RbcS2 terminator; 2A, FMDV 2A self-cleaving peptide. For immunoblots, samples were loaded at equal total protein amounts. Loading controls were stained by Ponceau red for the upper panel and Coomassie blue for the lower panel.

coding sequence (Figure 3A). Immunoblot analysis revealed the in plants (Pracharoenwattana et al., 2010). In peroxisomes, in absence of the full-length protein and the presence of a smaller addition to FA β-oxidation, the glyoxylate cycle also requires the hybridizing signal likely due to the formation of a truncated pro- activity of MDH. Defects in enzymes of glyoxylate cycle often tein (Figure 3B). Oil content analyses showed a similar oil remo- result in strains unable to grow in the dark when acetate is sup- bilization phenotype (upon N resupply) in mdh2-2 (Figure 3C), as plied as the only carbon source (Plancke et al., 2014). It was was observed in mdh2-1. observed here that mdh2-1 and mdh2-2 mutants grew nor- To summarize, based on genetic complementation and isola- mally in the dark either in liquid culture or on agar plates (Sup- tion of an independent allelic mutant, we conclude that the de- plemental Figures 4A and 4B), ruling out that MDH2 is required fect in oil remobilization observed in mdh2-1 and mdh2-2 results for acetate utilization. Moreover, the mutants grew as well as from the absence of a functional MDH2 protein. MDH2 has been dw15 when cultivated photoautotrophically in the presence of previously shown to be located in peroxisomes of Chlamydo- CO2-enriched air, but reached a lower cell density at stationary monas (Hayashi and Shinozaki, 2012; Kong et al., 2017) and is phase when cultivated mixotrophically (Supplemental Figures expected to play a role in FA β-oxidation based on its function 4C and 4D). 1828 The Plant Cell

total fatty acids remained similar between the three genotypes (dw15, mdh2-1, and C1) (Figure 2E). Based on immunoblot analysis, we observed that photoautotrophically grown wild-type cells made twice as much MDH2 protein than those grown mixotrophically (Supplemental Figure 5). Unless otherwise specified, cells were grown throughout this work under photoautotrophic conditions using air supplemented

with 2% CO2 (Supplemental Figure 1B). To study the role of MDH2 under photoautotrophic condition, we followed TAG accumulation in N-starved cells cultivated under this condition. Interest- ingly, we found a much more dramatic increase in TAG when cultivated photoautotrophically, i.e., the mutant cells accumulated >50% more TAGs than dw15 following 2 d of N deprivation (Fig- ure 4A). This is considerably greater than the 20% increase in mixotrophically grown N-deprived cells (Figures 1A and 2D). Ac- cordingly, an increased number and size of lipid droplets (LDs) was observed in mdh2-1 by confocal microscopy in comparison to the control strains (Figure 4C). In addition, although mdh2-1 mutant cells were smaller during photoautotrophic N-sufficient growth (i.e., before the onset of N starvation), they grew bigger than the wild type after 2-d N starvation (Supplemental Figure 6A). To determine whether observed increases in TAG and starch content in mdh2-1 mutant result from metabolic effects or from an increase in cell size, we also plotted TAG and starch content on cellular volume basis (Supplemental Figures 6B and 6C). Despite the changes being less dramatic compared with the per cell basis, TAG and starch content per cellular volume were still 30% higher in the mdh2-1 mutant as compared with its control strains (dw15 and C1). Similar phenotypes (i.e., in- Figure 3. mdh2-2 Mutant Characterization. crease in TAG and LDs) were also observed in mdh2-2 in com- (A) The insertion site of the cassette APHVIII in mdh2-2. parison to its parental strain (CC4533) on a cell basis (Figures (B) Immunoblot analysis using anti-MDH2 antibodies. 5A and 5D) or based on cellular volume (Supplemental Figures (C) TAG content analysis. 7A and 7B). Cells were grown in mixotrophic conditions under constant light. TAG Gas chromatography-mass spectrometry (GC-MS) analysis contents were determined in N-deprived cells (2 d) and in N-resupplied of total cellular FAs as their methyl esters (FAMEs) showed that cells (1d and 2d). Values are the mean of biological replicates (i.e., in- there was a net increase in total cellular FAs in mdh2-1 following dependent shaking flask cultures; n = 3, sd). Asterisks indicate statistically significant changes compared with parental strain (CC4533) by 2 d of N deprivation (Figure 4D), but interestingly there was no paired-sample Student’s t test (**P ≤ 0.01). Note: We have screened six difference found when cells were cultivated under mixotrophic independent lines, and all of them possessed a hybridizing signal just conditions (Figure 2E). It is worth noting here that prior to N depri- below the expected size, suggesting the likely formation of a truncated vation, slightly smaller amounts of total FAs were present in the protein (black arrow). For immunoblot, samples were loaded at equal to- mdh2-1 mutant mostly due to its smaller cell size compared with tal protein amounts. Loading controls were stained by Coomassie blue. the wild type. With the exception of an increase in 18:2(9,12), no significant change occurred in FA composition in comparison to mdh2 Mutants Overaccumulate TAG and Starch during the control strains (dw15 and C1) 2 d after photoautotrophic Photoautotrophic N Deprivation (Supplemented with 2% N starvation (Supplemental Figure 8). That said, during N-sufficient CO in the Air) growth, the mdh2-1 mutant contains relatively higher amount of 2 16:0 and slightly lower amount of 18:3(9,12,15) than the control Increasing evidence suggests the occurrence of lipid remobi- strains (dw15 and C1). Chlorophyll content in all strains was re- lization simultaneous to synthesis (Miller et al., 2010). For ex- duced upon N starvation, and the reduction was similar between ample, knocking down lipid catabolism has been demonstrated all three strains (Supplemental Figures 6D and 6E). to lead to strains that overaccumulate TAGs (Trentacoste et al., Metabolic changes in mdh2 mutants were not limited to stor- 2013; Kong et al., 2017). Disruption in the first step of FA age lipids. It was also found that the mutants accumulated two- β-oxidation (i.e., ACX2) in Chlamydomonas resulted in a mutant fold more starch than control strains (dw15 and C1) following strain (acx2) that accumulated 20% more TAG than its paren- 2 d of N deprivation (Figures 4B and 5B). By contrast, during mix- tal strain (dw15) during mixotrophic N deprivation (Kong et al., otrophic N deprivation, starch accumulated to similar amounts 2017). Furthermore, in this study, a ∼20% increase was detected in dw15 and the mdh2 mutants (Figure 2F). We therefore con- under mixotrophic culture conditions in the two mutants in which clude that during photoautotrophic N starvation, higher TAG MDH2 was knocked out (Figures 1A, 2D, and 3C), although and starch accumulation observed in the mdh2 mutants are the Peroxisome Redox Status and Chloroplast Metabolism 1829

in PSII yield was stronger in their respective parental lines than in the two mutants (Figures 5C and 6A). No significant difference in PSII yield could be observed when cells were cultivated in the presence of acetate (TAP condition) (Supplemental Figure 9).

Measurement of O2 evolution and CO2 uptake using membrane inlet mass spectrometry (MIMS) demonstrated that gross and

net photosynthetic O2 evolution and net CO2 uptake rates, although similar in N-replete conditions (Figure 6B), were significantly higher in the mdh2-1 mutant than in the control lines following 2 d of N deprivation (dw15 and C1) (Figure 6C).

In addition to the increased rates of O2 production and CO2 fixation during N starvation, we also measured the production of NADPH upon the transition from dark to light by following changes in NAD(P)H fluorescence (Roach et al., 2015). This analysis revealed that while no significant difference between the strains was observed in N-replete conditions, both mdh2 mutants displayed a higher level of NADPH fluorescence under illumination in comparison to their corresponding parental strain following 2 d of photoautotrophic N deprivation (Figures 7A and 7B; Supple- mental Figure 10). We conclude from these measurements that photosynthesis (measured as PSII efficiency, photosynthetic gas exchange rates, or NADPH accumulation) is less affected by N deprivation in mdh2 mutants than in the control lines. To better understand the higher photosynthetic activities in mdh2 mutants, we next determined the levels of representative photosynthetic proteins by immuno-analysis. While photosystem I (PSI) and PSII amounts probed by PSAD and D1 subunits, Figure 4. The mdh2-1 Mutant Overaccumulates TAG and Starch during respectively, strongly decreased following 2 d of N deprivation, Photoautotrophic N Deprivation. the decrease was less pronounced in the mdh2-1 mutant (Fig- (A) TAG content. ure 8). Similar effects were observed on the light-harvesting (B) Starch content. complex stress-related 3 protein (LHCSR3), on the cytochrome (C) Confocal microscopy image of LDs in N-deprived cells (2 d). b6f complex (probed by “the cytochrome f subunit”), and on the (D) Total FA content. ATPase complex (probed by the ATPB subunit) albeit to a lesser Cells were cultivated under constant light in photoautotrophic conditions extent than the photosystem proteins (Figure 8). Levels of other with additional supply of 2% CO in the air. Values are the mean of bi- 2 proteins, including the chloroplast type II NADPH dehydrogenase ological replicates (i.e., independent shaking flask cultures; n = 6, sd). (NDA2), proton gradient regulation like 1 (PGRL1), flavodiiron Cells were stained with BODIPY, and pseudocolors were used: lipid droplet in green and chlorophyll in red. Asterisks represent statistically protein (FLVB), and Rubisco large subunit (RBCL1) remained es- significant difference from both control strains (the parental line dw15 sentially unchanged. We conclude from this experiment that the and C1) by paired-sample Student’s t test (*P ≤ 0.05 and ** P ≤ 0.01). C1, general decrease in components of the photosynthetic machin- one representative complemented line. ery occurring during N deprivation in the control strains, and particularly in PSII and PSI amounts, is less pronounced in mdh2, which may at least partly explain the maintenance of higher consequence of two cumulative events, a metabolic effect and photosynthetic performance in the mutant during N deprivation. an increase in the cell size. Metabolomics Analyses Reveal Reduction of Malate and mdh2 Mutants Maintain a Higher Photosynthetic Activity Readjustment of Primary Metabolism in mdh2 during during Photoautotrophic N Deprivation Photoautotrophic N Starvation

While the overaccumulation of TAG in mdh2 may not be surpris- To determine if major changes in primary carbon metabolism ing given the involvement of MDH2 in lipid catabolism, the effect occurred in mdh2 mutants, we performed metabolomics using on starch accumulation was more puzzling. To gain insight into a GC-MS-based method (Lisec et al., 2006). A total of 60 me- the overaccumulation of storage compounds in mdh2 mutants, tabolites including organic acids, free amino acids, sugars, and we measured photosynthetic parameters under photoautotro- their derivatives were quantified. In comparison to N-replete phic conditions during optimal growth and N deprivation. Pho- condition, over 60% of the total polar metabolites decreased in tosynthetic performances (measured here as the photosystem their cellular amount in N-deprived cells (Figure 9A; Supplemen- II [PSII] operating efficiency) decreased in wild-type cells during tal Data Set 1). Furthermore, most organic acids and free amino N deprivation as previously reported (Figure 6A) (Saroussi et al., acids were significantly more reduced in the mdh2-1 mutant than 2016; Schulz-Raffelt et al., 2016), but interestingly, the reduction control strains (dw15 and C1) following 2 d of N deprivation, 1830 The Plant Cell

Figure 5. mdh2-2 Overaccumulates TAG and Starch, and Possesses Higher Photosynthetic Activity during Photoautotrophic N Deprivation. (A) TAG content. (B) Starch content. (C) PSII operating efficiency. (D) LD imaging after cells being stained by BODIPY.

Cells were cultivated under constant light in photoautotrophic conditions with additional supply of CO2 at 2% in the air. Values are the mean of biological replicates (i.e., independent shaking flask cultures; n = 6, sd). Actinic light (200 µmol m−2 s−1) supplied by a red LED source was used for PSII yield measurement. Asterisks indicate statistically significant changes compared with the parental strain (CC4533) by paired-sample Student’s t test (**P ≤ 0.01). Cells were stained with BODIPY, and pseudocolors were used: lipid droplet in green and chlorophyll in red. mag., magnification.

whereas the changes in sugars were more varied. No signifi- Supplemental Data Set 1). For example, with the exception of cant difference in the total content of malate, the product of the two nonproteogenic amino acids (β-alanine and ornithine), MDH, was detected prior to N deprivation between mdh2-1 and which did not change significantly, most other free amino acids control strains. In all strains, the malate content was reduced were present at levels that are 200 to 500% higher in mdh2-1 as a response to N deprivation, but the extent of reduction in during N-replete growth. Yet, following 2 d of N deprivation, most mdh2-1 was far greater than that in control strains (dw15 and free amino acid levels were present at much lower amount in the C1) (Figure 9B). A >60% reduction in intracellular malate con- mdh2-1 mutant compared with its control strains (Figure 9A). tent was observed in mdh2-1 in comparison to control strains Taken together, these results suggest that during response to (dw15 and C1) following 2 d of N deprivation. It is worth noting N deficiency, greater metabolic adjustment occurs in the mutant that, maltose, a breakdown product of starch degradation, was compared with its parental strain. increased by 200% following 2 d of N deprivation in mdh2-1 compared with its control strains (Figure 9C). In addition, we The mdh2-1 Mutant Is Delayed in Cell Division during observed a ∼80% decrease in sucrose content in the mdh2-1 N Deprivation but Overall Biomass Productivity Is Not mutant compared with its control strains following 2 d of N depri- Affected vation (Figure 9D). The observed metabolic shift during N deprivation is in clear Upon N starvation, the mdh2-1 mutant showed delayed cell di- contrast to cells grown under optimal conditions, where most vision (1.5-fold increase in cell concentration) compared with the metabolites were found to be present in higher quantities in control strains (dw15 and C1), which typically show a doubling mdh2-1 than in their control strains (dw15 and C1) (Figure 9A, of the cell concentration 1 d after the onset of N deprivation Peroxisome Redox Status and Chloroplast Metabolism 1831

biomass basis (Figure 10C). The increase in cell volume in the mdh2-1 mutant could be partly due to an impairment in cell division, due to the accumulation of carbon reserves, and partly due to the considerable increases in sugars and organic acids (Figure 9), some of which are known to alter cellular osmotic potential (Centeno et al., 2011; Araújo et al., 2012). The lack of an increase in dry biomass, in contrast to the increase in total cell volume, supports the latter hypothesis. Taken together, in spite of a slight retardation in cell division, there is no compromise in overall biomass productivity, possibly because the biomass of the mdh2-1 mutant is enriched in oil and starch (Figure 4).

mdh2 Overaccumulates Starch upon HL Exposure and during Diurnal Growth

To determine if MDH2 has a function in other situations beyond N deprivation, we examined mutant behavior under HL, a condition frequently encountered by many algae. We observed that although cells grew at a similar rate in liquid cultures (Figure 11A), mdh2-1 is more sensitive to HL (500 µmol m−2 s−1) during photoautotrophic growth on agar plates than the control strains (dw15 and C1) visible 6 d after being exposed to HL, while the growth was similar under low light (LL; 50 µmol m−2 s−1) (Figure 11B). Such a difference in growth performance observed between solid and liquid cultures may be surprising but not un- precedented and have indeed also been observed in the pgrl1 mutant (Dang et al., 2014). Multiple physiological parameters could account for this difference among which the shading

Figure 6. mdh2-1 Displays Higher Photosynthetic Yield during Photoau- totrophic N Deprivation.

(A) Measurement of PSII operating efficiency.

(B) and (C) Measurements of O2 production and net CO2 fixation using MIMS in N-replete (B) and in N-deprived cells (C). Figure 7. Measurement of NAD(P)H Fluorescence during Photoautotro- Cells were cultivated under constant light in photoautotrophic conditions phic N Starvation. with additional supply of 2% CO in the air. Values are the mean of biolog- 2 (A) NADPH fluorescence in mdh2-1 and its parental strain dw15. ical replicates (i.e., independent shaking flask cultures; n = 3, sd). Actinic (B) NADPH fluorescence inmdh2-2 and its parental strain CC4533. −2 −1 light (200 µmol m s ) supplied by a red LED was used for PSII yield Cells were grown to a constant OD under continuous light in photoauto- measurement, whereas for MIMS analyses, light was supplied by a green trophic condition (+2% CO2 supplemented in the air), then NADPH fluo- LED source. Asterisks indicate statistically significant difference from both rescence was measured before (MM) and after N deprivation (MM-N 2 d). control strains (dw15 and C1) by paired-sample Student’s t test (*P ≤ 0.05 Cells were kept in the dark for 1 min before an actinic light exposure (at and **P ≤ 0.01). C1, one representative complemented line. −2 −1 70 µmol photons m m provided by a red LED). Values are the mean of the NAD(P)H fluorescence level obtained across the 20-s light exposure under photoautotrophy (Figure 10A). However, on a cellular (i.e., independent shaking flask cultures; n = 3, sd). Original NAD(P)H volume basis, growth was higher in the mdh2-1 mutant than fluorescence traces are shown in Supplemental Figure 10. Asterisks in control strains (Figure 10B). Interestingly, there was no sig- indicate significant difference compared with their respective parental nificant difference found when growth was evaluated on a dry strains by paired-sample Student’s t test (*P ≤ 0.05). A.U., arbitrary unit. 1832 The Plant Cell

a 12-h-light/12-h-dark cycle. We observed that starch accumulation followed the typical “bell shape” in all three strains, i.e., starch reached the highest level at the end of day and was degraded at night (Figure 12), but strikingly much higher amounts of starch (>300% increase at the end of day) were accumulated in the mutant. Intriguingly, at the end of night, the starch content returned to a similar level as found in the parental strain, which showed that mdh2-1 mutant degraded a greater amount of starch within the same time period, indicating a possible increase in catalytic activities of starch-degrading enzymes and implying that the circadian regulation of starch degradation is similar in Chlamydomonas as that previously reported in Arabi- dopsis thaliana (Smith and Stitt, 2007).

mdh2-1 Produces Higher Amounts of H2O2 under N Starvation and HL as Well as during Diurnal Growth

The absence of MDH2 is expected to hinder reoxidation of NADH produced by the 3-hydoxyacyl-CoA dehydrogenase activity of the β-oxidation spiral, therefore likely resulting in an overreduced peroxisome. The observed increases in de novo FA synthesis, photosynthetic activity, and starch accumulation suggested that the imbalance in peroxisomal redox status can

be transmitted to chloroplast. H2O2 is a known cellular mes-

senger. We therefore determined extracellular levels of H2O2 using the Amplex Red method (Allorent et al., 2013; Dang

et al., 2014). The H2O2 level was 20% higher in mdh2 compared with control lines (dw15 and C1) following photoautotrophic N

starvation (Figure 13A). The increase in the level of H2O2 in the mdh2-1 mutant triggered the production of higher amount of Figure 8. Immunoblot Analyses of Photosystem Proteins during Photo- catalase 1 protein (CAT1) compared with wild-type strain during autotrophic N Deprivation. photoautotrophic N starvation (Supplemental Figure 11). To further investigate the overproduction of H O by the mutant, we (A) Representative images of immunoblot analysis. 2 2 (B) Quantification of signal intensities from cells being N-starved for 2 d. used the oxidant-sensing fluorescent probe 2',7'-dichlorodihy-

Proteins were collected from cells before (MM) and after photoautotro- drofluorescein diacetate (H2DCFDA), which is converted to the green fluorescent dichlorofluorescein (DCF) upon exposure to phic (+2% CO2 supplemented in the air) N deprivation (MM-N, 2d) under constant light. Representative images of immunoblot analysis are conditions high in ROS. DCF fluorescence can be captured with shown. Signals for (B) were averaged from three biological replicates a confocal microscope, and intensity of signal reflects the in- (i.e., independent shaking flask cultures) for N-starved cells (n = 3; sd). tracellular level of ROS including H2O2. Increased fluorescence Asterisks indicate significant difference from the parental strain dw15 by was observed in all strains (dw15, mdh2, and C1) upon N star- paired-sample Student’s t test (*P ≤ 0.05). Samples were loaded at equal vation, conforming to previous observations on the increase total protein amounts and stained using Coomassie blue. in ROS level in N-deprived Chlorella sorokiniana (Zhang et al., 2013), Dunaliella salina (Yilancioglu et al., 2014), and Chlamydo- monas (Du et al., 2018). Furthermore, the increase in ROS level effect encountered quickly in liquid cultures as a result of cell upon N deprivation was even greater in mdh2 as observed by doubling could explain a lack of growth difference between measurement of DCF fluorescence using a spectrofluorometer mdh2-1 and the wild type (i.e., dw15). Another reason might be (Figure 13B) as well as under confocal imaging (Figure 13C). due to the length of incubation, i.e., 2 d in liquid cultures versus Nonuniform patches of green fluorescent spots were observed 5 to 7 d in solid agar plate. Higher amounts of TAG and starch in cells of all genotypes, and although we could not be sure that accumulated in all strains exposed to HL, which was consistent they are present in the peroxisomes, it was clear that they are with previous observations in Chlamydomonas wild-type strains not limited to the chloroplasts. Furthermore, we also determined

(Goold et al., 2016), yet the increase in TAG and starch content extracellular H2O2 levels in the parental strain and mutants upon was even greater in mdh2-1 (Figures 11C and 11D). HL exposure and during a diurnal growth. Whereas no differ- Changes in starch metabolism are not limited to N deprivation ence was observed between strains under LL or in the dark, a or HL. During diurnal growth, starch is accumulated during the 20 to 30% increase in H2O2 was observed in mdh2-1 compared day and consumed at night (Ball et al., 1990). Therefore, we also with its control strains (dw15 and C1) after 1 d of HL exposure followed starch content in mdh2-1 and the control strains during or during the light period (Supplemental Figure 12). Peroxisome Redox Status and Chloroplast Metabolism 1833

Figure 9. Metabolomics Analysis of Polar Metabolites in Photoautotrophically Grown Cells before and 2 d after N Deprivation. (A) A heat map view of metabolic changes. (B) Intracellular malate content. (C) Intracellular maltose content. (D) Intracellular sucrose content.

Cells were cultivated under constant light in photoautotrophic conditions with additional supply of 2% CO2 in the air. Values are the mean of biological replicates (i.e., independent shaking flask cultures; n = 8, sd). Asterisks indicate significant difference from control strains by paired-sample Student’s t test (*P ≤ 0.05 and **P ≤ 0.01). A.U., arbitrary unit.

In plant/algal peroxisomes, H2O2 is produced by ACX, the ini- of FA β-oxidation (Kong et al., 2017). We observed that under N tial step of β-oxidation (Graham, 2008; Kong et al., 2017). To deprivation, the acx2-1 mutant did not overaccumulate H2O2 and identify the source of the observed increase in H2O2, we de- didn’t overproduce starch or show a difference in PSII operating termined the level of ACX activity in cell-free protein extracts efficiency from the control strains (Figures 14A to 14C). The lack from mdh2-1 and its parental strain (dw15) using stearoyl-CoA of a reduction in H2O2 level in acx2-1 compared with its parental (18:0) as a substrate given that it represents one of the major strain dw15 is not surprising given the occurrence of four other acyl-chain lengths in Chlamydomonas. ACX activity (18:0) was ACX isoforms in Chlamydomonas (Kong et al., 2017). Further- found to be almost twice higher in mdh2 than in dw15 following more, we also examined the response of the acx2-1 mutant to N-deprivation (2 d) (Supplemental Figure 13), consistent with HL exposure. The acx2-1 mutant grew as well as dw15 under HL the increased H2O2 production. To further explore a possible link (Figure 14D). Unlike the mdh2-1 mutant, where starch is overac- between peroxisomal production of H2O2 by ACX and reserve cumulated during a diurnal growth period (+N), the acx2-1 mu- formation, we also analyzed the acx2-1 mutant, which we have tant did not overaccumulate, or made even less, starch during previously shown to be blocked in the H2O2-generating reaction the same period (Figure 14E). Overall, the observation on the 1834 The Plant Cell

increase of ACX (18:0) activity in mdh2-1 mutant together with the lack of starch overaccumulation in acx2 mutant therefore supported the idea that peroxisomal FA β-oxidation is a major

source of H2O2 in mdh2.

H2O2 Supplementation Results in TAG and Starch Overaccumulation in the mdh2-1 Mutant

To further evaluate the idea that oil and starch over-accumulation

observed in mdh2 may result from an increased H2O2, we determined TAG and starch contents in cells supplemented with

H2O2. Considering the dual role of H2O2 on cell physiology and

metabolism, we first tested the effect of various 2H O2 concentrations on growth of the wild-type strain. We chose to use a

final concentration of 0.5 mM H2O2 because at this concentration, cells were still growing and did not show any obvious del- eterious effects seen with higher concentrations (Supplemental

Figure 14A). Cell growth of all strains in the presence of H2O2 supplementation was similar in liquid cultures (Supplemental Figure 14B), whereas when cells were plated out in solid agar

Figure 11. HL Response of the mdh2-1 Mutant during Photoautotrophic Growth.

(A) Growth kinetics in liquid cultures under HL. Figure 10. Cell Growth during Photoautotrophic N Deprivation. (B) Growth comparison on agar plates. (A) Relative growth based on cell number per milliliter of culture. (C) TAG content on a per cell basis. (B) Relative growth based on cell volume per milliliter of culture. (D) Starch content on a per cell basis. (C) Relative growth based on dry biomass per milliliter of culture. Cells were cultivated under constant light (either HL or LL) in photoau-

Cells were cultivated under constant light in photoautotrophic N starva- totrophic conditions with additional supply of 2% CO2 in the air. Light was provided by a cool LED white light. The same number of cells were tion conditions with additional supply of CO2 at 2% in the air. Cell growth was monitored every day using a Coulter counter. Then, cell concentra- inoculated on MM agar plate and kept under continuous light to monitor tion was normalized to that before N deprivation (which is set as 1). Val- cell growth. Images were taken 6 d after cells being deposited. Values ues are the mean of biological replicates (i.e., independent shaking flask are the mean of biological replicates (i.e., independent shaking flask cul- cultures; n = 11, sd). Asterisks indicate statistically significant difference tures; n = 4, sd). Asterisks indicate significant difference from control from control strains by paired-sample Student’s t test (*P ≤ 0.05). Cell strains by paired-sample Student’s t test (*P ≤ 0.05 and **P ≤ 0.01). LL, vol, cellular volume. 50 µmol m−2 s−1; HL, 500 µmol m−2 s−1. Peroxisome Redox Status and Chloroplast Metabolism 1835

posure; and (4) MDH2 plays a role in modulating starch metabolism during diurnal growth. This study thus reveals regulation of chloroplast-based activities by factors derived from peroxi-

somes. We further identify malate and H2O2 as important players in this process. Below, we discuss and provide a working model explaining metabolic reorientations occurring in the absence of MDH2 in Chlamydomonas (Figure 16). The logics and reasoning in supporting this model are provided below.

MDH2 Is a Major Contributor to NAD+ Regeneration in Peroxisomal FA β-Oxidation MDH2 has been localized to peroxisomes in Chlamydomonas by three independent studies (Hayashi and Shinozaki, 2012; Lauersen et al., 2016; Kong et al., 2017) and has further been observed that MDH2 colocalized with ACX2, the first enzyme of the β-oxidation cycle (Kong et al., 2017). The defect in TAG hydrolysis reported here together with a peroxisomal localization Figure 12. Starch Accumulation during Photoautotrophic Diurnal strongly supports the involvement of MDH2 in FA β-oxidation Growth. (Figure 16, event 1). Moreover, MDH2 showed the highest amino Cells were cultivated in a diurnal cycle of 12 h light/12 h dark in a con- acid sequence similarity (>55% identity, BLAST) to the two Ara- bidopsis peroxisomal proteins AtpMDH1 and AtpMDH2 (Sup- trolled incubation chamber supplied with 2% CO2 in air. Light was supplied via fluorescent tubes at an intensity of 100µ mol m−2s−1. Values are plemental Table 1). Interestingly, the double Arabidopsis mutant the mean of biological replicates (i.e., independent shaking flask cul- (pmdh1 pmdh2) lost the capacity to remobilize TAG during seed tures; n = 4, sd). Shaded area refers to the night period. germination (Pracharoenwattana et al., 2007). A defect in lipid catabolism has also been observed in a mutant of Saccharomyces cerevisiae deficient in the peroxisomal MDH3 (van Roermund plate in serial dilutions, the mdh2-1 mutant showed increased et al., 1995). Taken together, these studies demonstrate that sensitivity in growth compared with its parental line visible 7 d the metabolism of NADH generated via FA β-oxidation in peroxi- after being deposited (Figure 15A). This might be due to a cu- somes is conserved between fungi, plants, and algae, and sim- mulative effect of light and H2O2 supplementation encountered ilar to the other two systems, the algal peroxisomal membranes more severely in solid cultures. are not permeable to NAD(H), thus prohibiting free exchange We then compared the capacity of all three strains in accu- with cytoplasm where de novo NAD+ synthesis occurs (Noctor mulating starch and TAG following H2O2 supplementation. In the et al., 2006). presence of exogenously added H2O2, it was found that starch Oil hydrolysis is severely but not completely blocked in the increased >200%, whereas TAG increased by 50% in the wild mdh2 mutants (Figure 1). This suggests the possible occurrence + type 2 d after H2O2 supplementation (Figures 15B and 15C). We of compensatory pathways in providing NAD for β-oxidation. further noticed that the increases in starch and TAG contents in Indeed, a number of other mechanisms controlling NAD+ ho- the mdh2-1 mutant were much greater (>30% more) than those meostasis have been reported to operate in plant peroxisomes. of control strains (dw15 and C1) (Figures 15B and 15C). Studies in Arabidopsis have shown that FA β-oxidation can be supported by peroxisomal hydroxypyruvate reductase (HPR) DISCUSSION when AtpMDH1 and AtpMDH2 are absent (Pracharoenwattana et al., 2010). Plant peroxisomes are also known to contain In all eukaryotic cells, biochemical reactions are compartmen- high amounts of the ascorbate peroxidase (APX)/monodehy- talized in specific subcellular organelles rendering the coordina- droascorbate reductase (MDAR1) electron transfer system, tion of metabolism across different compartments essential for which has been reported to play a role in NADH reoxidation as cell functioning (Sweetlove and Fernie, 2013). The chloroplast, well as in detoxifying H2O2 (Eastmond, 2007). In addition, an the major power house of photoautotrophs, is often found located Arabidopsis peroxisomal NAD+ carrier (PXN) has been demon- in close proximity to peroxisomes (Hayashi and Shinozaki, strated to operate in the import of NAD+ into the peroxisomes 2012; Hu et al., 2012; Schwarz et al., 2017), but little is known from the cytoplasm (Bernhardt et al., 2012; van Roermund about the interaction and exchange of information between et al., 2016), and knockout mutants for PXN are defective in oil them in algae. Here, via characterization of two knockout mu- breakdown during seedling establishment. Genes homologous tants of the peroxisomal MDH2 in the unicellular green alga to HPR (Cre06.g295450), APX (Cre09.g401886), MDAR1 (Cre17. Chlamydomonas, we provide evidence that (1) FA β-oxidation g712100), and PXN (Cre07.g353300) are encoded in the genome requires a functional MDH2; (2) lipid catabolism is connected to of Chlamydomonas (Merchant et al., 2007), and their potential photosynthesis and chloroplast metabolism through MDH2; (3) involvement in lipid catabolism in algae is further supported by the reverse coupling of redox/H2O2 from peroxisome to chloro- a recent transcriptomic study where their transcriptions were plast is essential for cell division during N deprivation or HL ex- dysregulated in a mutant defective in oil hydrolysis following 1836 The Plant Cell

Figure 13. Determination of H2O2 Level during Photoautotrophic N Deprivation.

(A) Extracellular H2O2 level determined by Amplex Red. (B) Relative DCF fluorescence level. (C) Intracellular ROS level determined using H2DCFDA staining.

Cells were cultivated under constant light in photoautotrophic conditions with additional supply of 2% CO2 in the air. Values are the mean of biological replicates (i.e., independent shaking flask cultures; n = 4, sd). Asterisks indicate significant difference compared with control strains (dw15 and C1) by paired-sample Student’s t test (*P < 0.05). C1, the complemented line.

N resupply (i.e., the cht7 mutant, compromised in hydrolysis of the medium (Schmollinger et al., 2014; Park et al., 2015; TAG) (Tsai et al., 2014, 2018). That said, none of the algal proteins Saroussi et al., 2016). Structural changes and transcrip- has yet been functionally characterized. Nevertheless, these al- tomic responses during this adaptation have been reported ternative mechanisms in any case do not appear to play a major (Schmollinger et al., 2014; Saroussi et al., 2016), but little role, at least under the conditions tested in this study, since mdh2 is known concerning the underlying regulatory mecha- mutants lost >80% of their capacity to remobilize TAG reserves. nisms. Findings in this study suggest that the peroxisomal MDH2 through participation in peroxisomal FA β-oxidation + The Role of Peroxisomal MDH2 in Fine-Tuning of and NAD metabolism is, at least partly, involved in down- Photosynthesis during Photoautotrophic N Deprivation regulating photosynthesis in N-deprived Chlamydomonas. This conclusion is evidenced by the fact that in the absence Chlamydomonas reduces its electron transport activities to of MDH2, cells sustained a higher photosynthetic electron adapt to a reduced metabolic need when N is depleted from transport activity and generated higher amounts of NADPH Peroxisome Redox Status and Chloroplast Metabolism 1837

Figure 14. Effect of Silencing ACX2 on Starch Accumulation and HL Response.

(A) Fold change in H2O2 production in acx2-1. (B) Operating PSII efficiency. (C) Starch content before and after N deprivation (2 d). (D) Light sensitivity test. (E) Starch content during diurnal growth.

Cells were grown in liquid culture under constant light with additional supply of 2% CO2 and then the same number of cells was inoculated on MM agar plate and kept under continuous light (supplied by cool LED white light) to monitor cell growth at 25°C. Images were taken 7 d after cells being deposited. Actinic light (200 µmol m−2 s−1) supplied by a red LED source was used for PSII yield measurement. LL, 50 µmol m−2 s−1; HL, 500 µmol m−2 s−1. Shown are the parental strain dw15, acx2-1, and one complemented line (Comp1). Shaded area refers to the night period. Values are the mean of biological replicates (i.e., independent shaking flask cultures; n = 4, sd). during photoautotrophic N starvation than control strains By contrast, in the absence of MDH2, the malate level is lower (Figures 6, 7, and 16, event 2). and more NADP+ is available to accept electrons from PSI, there- Two major metabolic changes could result in more sustained by enhancing electron transport activities and producing more photosynthesis in mdh2, namely, a reduction in malate and an in- NADPH. A lower malate level in the chloroplast of mdh2 mutants crease in starch. First, malate is a recognized electron carrier and may result from impaired shuttling of malate between peroxi- can be transported across subcellular membranes by dicarbox- somes and chloroplasts, such shuttling being possible thanks ylate transporters (Mettler and Beevers, 1980). In wild-type cells to a concerted functioning of both peroxisomal and plastidial when malate level is high, less NADP+ is available as acceptor of MDHs. Although the redox-regulated plastidial MDH5 has been electrons released from PSI; therefore, photosynthesis is down- often reported to function as a malate valve exporting reducing regulated to match to the reduced metabolic needs of the cell. power out of the chloroplast (Scheibe, 2004; Hebbelmann et al., 1838 The Plant Cell

Figure 15. Effect of H2O2 Supplementation on TAG and Starch Content during Photoautotrophic Growth. (A) Cell growth. (B) Starch content per cell. (C) TAG content per cell.

Cells were cultivated under constant light in photoautotrophic conditions with additional supply of 2% CO2 in the air. Cells were collected before and

2 d after H2O2 addition (0.5 mM). A given number of cells was deposited on MM agar plates with or without addition of H2O2. Images were taken 7 d after cells being deposited. Values are the mean of biological replicates (i.e., independent shaking flask cultures; n = 4, sd). Asterisks indicate statistically significant difference from control strains (dw15 and C1) by paired-sample Student’s t test (**P ≤ 0.01).

2012; Heyno et al., 2014), the enzymatic activity of NADP-MDH in starch synthesis in mdh2 will consume a large amount of is highly reversible, and it is therefore conceivable that depending NADPH, therefore regenerating NADP+ for photochemical re- on the metabolic conditions (for instance, N availability and/or actions, resulting in more sustained photosynthesis during light intensity) and depending on the cellular redox poise the photoautotrophic N starvation compared with the wild type. shuttle may work in one direction or in the other. Moreover, in In addition, the chloroplast redox state has been reported to addition to the redox-regulated NADP-MDH, plants and algal induce the expression of proteins of the photosynthetic ap- chloroplasts contain a NAD-MDH (MDH1) (Supplemental Table 1), paratus (Oswald et al., 2001), which might explain the higher which is not redox regulated and not involved in the function- levels of these proteins observed in the mutant than the wild ing of the malate valve, but supports other metabolic functions type during N starvation (Figure 8). The lower sucrose level (Beeler et al., 2014; Selinski and Scheibe, 2014). present in the mdh2-1 mutant (Figure 9) could also be involved Second, the higher photosynthetic activities of mdh2 (in in sustaining photosynthesis, via relieving feedback regulation comparison to the wild type) occurring during photoautotro- of photosynthesis, as has been reported previously in land phic N starvation could be due to increased sink strength since plants (Pfannschmidt et al., 1999; Lobo et al., 2015). Together, starch accumulates in mdh2 mutants (Figure 4), and starch is the different metabolic adjustments occurring in mdh2 during an excellent electron sink. When facing harsh conditions (nu- photoautotrophic N starvation favor a higher photosynthetic trient deprivation or HL), the limiting factor for photosynthesis electron flow and higher CO2 fixation into organic carbon as is the availability of electron acceptors. The observed increase compared with the parental strains. Peroxisome Redox Status and Chloroplast Metabolism 1839

Figure 16. Tentative Model Explaining Redirection of Metabolism in the Absence of MDH2 during Photoautotrophic N Deprivation. FA degradation starts with hydrolysis of TAGs and membrane lipids by lipases. FAs released enter the peroxisomes via an ABC transporter and subsequently are degraded to acetyl-CoAs by the core FA β-oxidation spiral, which consists of four enzymatic activities (ACX, MFP-hydratase, MFP-DH, and KAT). MDH2 plays a role in oxidation of NADH, which is generated by hydroxylacyl-CoA dehydrogenase (MFP-DH) at the third step of β-oxidation spiral. In the absence of MDH2, NADH is likely accumulated, thereby increasing the reduction state of peroxisome. This would result in two metabolic changes in peroxisome: a decrease in malate export and an increase in H2O2 level. These two metabolites would in turn significantly alter photosynthesis and chloroplast metabolism. Activated pathways in the mutant are indicated by blue arrows, whereas downregulated pathways are indicated by red arrows. Based on results obtained from this study and current literature, we propose a cascade of events leading to the observed phenotypes in mdh2 mutants: “High oil” phenotype (events 1, 2, and 5): A block in FA β-oxidation combined with increased de novo FA synthesis contributes to higher TAG accumulation. “High starch” phenotype (events 2 to 4): During N starvation, the more active CO2 fixation and photosynthesis in themdh2-1 mutant provides more NADPH and carbon precursors for starch synthesis. The increased level of NADPH in chloroplast activates AGPase for starch synthesis and also activates several starch-degrading enzymes; therefore, it results in an increased carbon flux into and out of the starch route. Higher amounts of starch indicate high sink capacity, therefore sustaining photosynthesis. As a consequence, more NADPH is produced by photochemical reactions, and this further activates AGPase, together resulting in the 100 to 300% increase in starch content. “High H2O2” phenotype (event 3): The higher level of H2O2 in the mutant is supported by the 2-fold increase in the H2O2-generating reaction catalyzed by ACX activity. Once it is transmitted to chloroplast, H2O2 activates the starch synthesis pathway, as supported further by the observation of starch overaccumulation in wild-type cells supplemented with exogenous H2O2. “Sustained photosynthesis” (events 1 and 2): A blockage in FA β-oxidation at the step of MDH2 will result in less malate being produced in the peroxisome. Malate, a recognized electron carrier, can be transported to other compartments via the dicarboxylate transporter. When malate level is high, less NADP+ is available as electron acceptor at PSI, therefore downregulating photosynthesis. Conversely, in the absence of MDH2, the malate level is decreased, and more NADP+ is available to accept electrons from PSI. ACX, acyl-CoA oxidase; CAT, catalase; CBC, Calvin-Benson-Bassham cycle; CTS1, comatose 1; DH, dehydrogenase; FAD, flavin adenine dinucleotide; FFA, free fatty acid; Fdx, ferredoxin; FNR, ferredoxin-NADP+ reductase; KAT, ketoacyl-CoA thiolase; LHC, light-harvesting complex; Mal, malate; MFP, multifunctional protein; NTRC, NADP:thioredoxin reductase C; OAA, oxaloacetate; PQ, plastoquinone; SDP1, sugar dependent 1; SOD, superoxide dismutase.

Malate, H2O2 and Relationships to Starch Metabolism an increase in the levels of maltose, NADPH, and H2O2 was during Photoautotrophic N Deprivation observed (Figures 7, 9, and 13). Together, these results point + to a link between malate content, H2O2 level, NADP reduction In addition to lipids, we detected large amounts of starch in the state, and starch metabolism (Figure 16, event 3). A negative mutants under N deprivation, HL, and during diurnal growth correlation between malate and starch content has been pre- (Figures 4, 5, 11, and 15). In parallel, a reduction in malate and viously observed in potato tubers (Solanum tuberosum) and 1840 The Plant Cell

tomato fruits (Solanum lycopersicum) when mitochondrial ma- reveals the occurrence of starch degradation activities simulta- late metabolism is disturbed (Jenner et al., 2001; Centeno et al., neous to starch synthesis in the light, in specific cell types, or 2011), and elevated photosynthesis has been reported following under particular environmental conditions (Sparla et al., 2006; downregulation of the mitochondrial malate dehydrogenase in Baslam et al., 2017; Daloso et al., 2017). tomato (Nunes-Nesi et al., 2005). However, the identity of in- In parallel to energetic changes, the increase in CO2 assimi- tracellular signal(s) involved in such a connection has not been lation in mdh2 results in higher production of precursors for the established. Compared with previous reports, our work brings synthesis of carbon reserves i.e., starch and TAG. Together, the in two new aspects: (1) It provides an example in which chlo- increased synthesis and turnover of starch in mdh2 led to the roplast starch metabolism can be modulated by malate metab- spectacular “high starch high maltose” phenotype. Adding to olism in peroxisomes; and (2) it provides evidence that diurnal the already complex regulatory mechanisms of starch synthesis, starch metabolism can be modulated by a peroxisome-located our work provides a demonstration of the regulation of chloro- protein. plast starch synthesis and turnover by signal (and metabolite) Regulation of starch synthesis is known to occur at the level derived from the peroxisomes. of the first and rate-limiting step catalyzed by ADP-glucose pyrophosphorylase (AGPase), which is under multiple regulations Increased TAG Amount in N-Deprived mdh2 including principally allosteric and redox regulation (Geigenberger, 2011). Reductive activation of AGPase has been shown to oc- In this study, we found that significant amounts of TAGs are cur in tomato as well as Chlamydomonas (Iglesias et al., 1994; made in N-deprived mdh2 cells (Figures 1, 4, and 5). Upon N Libessart et al., 1995; Buléon et al., 1997) and is mediated by deprivation, a block in β-oxidation diverts otherwise toxic free a chloroplast-located NADPH-dependent thioredoxin reductase FAs released from membrane lipids to make storage lipid. This C (NTRC) (Michalska et al., 2009; Geigenberger, 2011; Lepistö is supported by two recent observations in Chlamydomonas, et al., 2013; Naranjo et al., 2016; Thormählen et al., 2017). The i.e., in acx2 mutants (Kong et al., 2017), as well as in the icl increased reduction state of mdh2 chloroplasts (i.e., higher (isocitrate lyase) mutant that is blocked in glyoxylate cycle that NADPH level) could activate AGPase via NTRC (Figure 16, event functions downstream of β-oxidation allowing conversion of FAs

3). In parallel, the higher CO2 fixation of mdh2 (Figure 6) results to sugars (Plancke et al., 2014). However, this can only partly in an increased production of photosynthates through the CBC explain the increased TAG content in mdh2 under photoautotro- cycle; these two effects (enzyme activation and substrate avail- phic N deprivation because we found that the increase in TAG ability) result in a boost of starch synthesis in mdh2 mutants. content in mdh2 is much more dramatic under photoautotro-

Moreover, a positive correlation between H2O2 and starch phic N deprivation (50–100%) than under mixotrophic N depri- amounts was observed, which is supported by three lines of vation (∼20%). The dramatic increase in TAG content in mdh2 evidence. (1) The H2O2 level was increased in mdh2 mutants under photoautotrophic conditions can partly be accredited to together with enhanced starch accumulation. 2) When the H2O2 enhanced de novo FA synthesis, as suggested by the increase level remained similar to the wild type, i.e., in the case of the in total FA amount under photoautotrophy but not mixotrophy acx2-1 mutant, starch was not overaccumulated. (3) When wild- (Figures 2E and 4D). The difference in the photosynthetic re- type cells were supplied with exogenous H2O2, starch was over- sponses during N starvation under mixo- versus photoautotro- accumulated. This report thus establishes a positive correlation phic conditions supports an increased production of NADPH between starch amount and H2O2. The biochemical observation (and potentially total FA) only in photoautotrophically N-starved made here is supported by a recent H2O2-induced transcrip- cells (Figure 4; Supplemental Figure 9). Thus, the additional in- tomic response study showing that mRNA abundance of sev- crease in TAG content in photoautotrophically N-starved mdh2-1 eral starch synthesis genes is increased in H2O2-supplemented cells is likely due to an increased de novo FA synthesis (Figure cells (Blaby et al., 2015). Nevertheless, the underlying molecular 16, event 5). mechanisms linking H2O2 and starch still need to be worked out. In mdh2, enhanced de novo FA synthesis can be attributed Metabolomics analysis revealed that maltose and glucose, to the increased supply of carbon precursors as well as of the the major products of starch degradation, were dramatically in- reducing equivalent NADPH. De novo FA synthesis occurs in creased in mdh2 in comparison to its parental strain. Moreover, the chloroplast of algae and plants and is catalyzed by fatty despite the fact that more starch was made by the end of day acid synthase (FAS), and this reaction requires a stoichiometric in mdh2, the level of starch at the end of night period fell to supply of carbon precursors, ATP and NAD(P)H (Ohlrogge and the same level as those of control strains, indicating acceler- Browse, 1995; Li-Beisson et al., 2013). Enhanced CO2 assimi- ated activities of starch-degrading enzymes. These two sets of lation in mdh2 will provide more organic carbon precursors fed data thus point to an increase in starch degradation in mdh2 into the FAS complex for FA synthesis. This conforms to the (Figure 16, event 4). This could be explained by the elevated previous finding that carbon precursor supply for FA synthesis levels of NADPH present in mdh2 because several enzymes of controls TAG accumulation in Chlamydomonas (Goodson et al., the starch degradation pathway, notably, an α-glucan water dik- 2011; Fan et al., 2012). inase, a β-amylase, and a phosphatase, have been shown to The FAS complex is known to be activated by NADPH, which be under redox regulation (Mikkelsen et al., 2005; Sparla et al., was present at higher levels in mdh2 as a result of enhanced 2006; Comparot-Moss et al., 2010; Geigenberger, 2011). The photosynthetic activities and increased starch turnover. Indeed, concomitant increase in synthesis and degradation of starch a positive link between the level of NADPH and FA amount has seems, at first sight, puzzling. Actually, increasing knowledge been established in fungi, plants, and algae (Zhang et al., 2007; Peroxisome Redox Status and Chloroplast Metabolism 1841

Xue et al., 2016). For instance, it has been observed that de novo used as the parental strain for mutant library generation (Li et al., 2012). FA synthesis can be enhanced via transgenic overexpression of The mutant mdh2-1 was isolated following the previously published pro- an NADPH-producing malic enzyme in the diatom Phaeodac- tocol (Cagnon et al., 2013). The mdh2-2 mutant (LMJ.RY0402.208051) tylum tricornutum and in the green alga Chlorella pyrenoidosa together with its background strain CC4533 was purchased from the CLiP collection at (https://www.chlamylibrary.org) (Li et al., 2016). Un- (Xue et al., 2015, 2016). By contrast, reducing the NADPH supply less otherwise specified, cells were routinely cultivated in an incubation via inhibition of malic enzyme activities using sesamol in Hae- shaker (INFORS Multitron pro) maintained at 25°C, with 100 rpm shaking matococcus pluvialis, Nannochloropsis sp, and the filamentous −2 −1 and constant illumination at 100 µmol m s . Except the HL experi- fungus Mucor circinelloides led to reduction in total FAs (Wynn ment (500 μmol m−2 s−1), which was provided by a cool LED white light, et al., 1997; Recht et al., 2012). all other lightings used for cell cultivation in the INFORS were supplied In addition, the increased TAG level in mdh2 could also be by fluorescent tubes (Fluora Osram). Photoautotrophic growth refers to cells grown in MOPS-buffered minimal medium (MM) supplemented with due to overproduction of H2O2. Indeed, a study in Arabidopsis 2% CO in the air (Harris, 2009; Schulz-Raffelt et al., 2016), whereas has shown that H2O2 released from the peroxisomes defec- 2 tive in the APX/MDAR system inhibits the major TAG lipase mixotrophic growth refers to cells cultivated in Tris-acetate-phosphate SUGAR-DEPENDENT1, therefore resulting in impaired TAG hy- (TAP) medium (Harris, 2009). With the exception of TAG remobilization experiment (performed in the dark in air), all other cultures were grown in drolysis during seed germination and postgerminative growth MM-MOPS supplemented with 2% CO in the air (Supplemental Figure (Eastmond, 2007). Taken together, the dramatic increase in 2 1). Cell concentration, cell size, and cellular volume were monitored with TAG content in photoautotrophic N-deprived mdh2 is a com- a Multisizer 3 Coulter counter (Beckman Coulter). For drop test, a series binatorial result of downregulating FA β-oxidation and TAG lip- of diluted samples was either spotted on TAP plate for heterotrophic olysis and enhancing their de novo synthesis due to increased growth or spotted on MM agar plates and incubated for 5 to 10 d under availability of carbon precursors as well as reducing equivalent either LL (50 μmol m−2 s−1) or HL (500 μmol m−2 s−1) supplied with a cool

NADPH. LED white light at 25°C. For growth test in the presence of H2O2, a given

number of cells were spotted on MM agar plates with or without H2O2 supplementation (at a final concentration of 0.5 mM). For N deprivation, Conclusion exponentially grown cells were centrifuged at 600g for 5 min, and cell pellets were washed twice with N-free media (TAP-N or MM-N) before In summary, although the peroxisome was described >60 years being resuspended in N-free media for starvation experiments. Unless ago in kidney (Rhodin, 1954), found later in plants (Beevers, otherwise stated, biological replicates refer to cells grown in independent 1979) and only fairly recently confirmed in Chlamydomonas Erlenmeyer flasks. (Hayashi and Shinozaki, 2012), the role of the peroxisomes as signaling organelle has just started to be recognized in mammals (Tripathi and Walker, 2016). We still know comparatively Starch and Chlorophyll Quantification little about the role of peroxisomes in photoautotrophs beyond Starch and chlorophyll were quantified from 1 mL of culture containing photorespiration in land plants. In addition to revealing previ- ∼5 million cells. Briefly, cell culture was centrifuged in a microfuge tube at ously unreported functions of peroxisome-chloroplast inter- 13,000g for 10 min. Pellet was resuspended into 1 mL of methanol, mixed actions in algal lipid metabolism (β-oxidation and de novo FA vigorously, and stored frozen at −80°C before analyses. The supernatants synthesis), this study uncovered an important function of the were used to quantify chlorophyll photometrically (Lichtenthaler, 1987), peroxisome in photoautotrophs in exerting control on photo- and the remaining pellets were left at room temperature in a fume hood synthesis and chloroplast metabolism. The proper exchange of to evaporate the residual methanol. Pellets were then resuspended in 400 μL of distilled water and autoclaved for 15 min at 120°C to solubi- information (malate/H2O2) from the peroxisome to the chloroplast is essential for cell growth especially when facing harsh lize the starch polymer. Total starch was quantified using an enzymatic starch assay kit (Sigma-Aldrich; ref. SA-20) following the manufacturer’s environmental conditions. The peroxisomal H O is likely the 2 2 instructions. Glucose converted from the starch was quantified using an missing link between environmental stress, metabolism, and automated YSI 2700 sugar analyzer (YSI Life Sciences) using a known redox balance. This study also points out the importance of amount of commercial glucose as a standard. peroxisomes in the already complex landscape of chloroplast- and mitochondrion-based electron dissipation mechanisms. Findings from this study on the unicellular green model alga Biomass Determination Chlamydomonas should be extendable to photosynthetic tissues Biomass was determined by dry weight measurements. Briefly, cells of land plants where photosynthesis occurs alongside the forma- in 10 mL of culture medium (N replete) and 5 mL of culture medium tion of carbon reserves in the same cell. Our results therefore may (N deplete) were collected on the preweighted glass fiber filter (VWR; suggest additional strategies for engineering plants and algae for ref. 611-0739) and dried overnight at 80°C in an incubator. The dried improved biomass production and stress tolerance. biomass was then gravimetrically measured.

FA and Lipid Analyses METHODS FAs were analyzed by GC-MS after all acyl-lipids were converted to their Strains and Culture Conditions methyl esters. TAG content was quantified after being separated from the bulk membrane lipids on a thin-layer chromatography plate. Unless The cell wall-less strain dw15.1 (CC4619 cw15, nit1, mt+; abbreviated as otherwise stated, all methods related to lipid analyses were performed dw15 throughout this article), kindly provided by Christoph Benning, was as previously described (Kong et al., 2017). 1842 The Plant Cell

Insertion Site Identification by Restriction Enzyme Site-Directed Proteogenix SAS (Schiltigheim). Specificities of the antibodies were then Amplification-PCR tested on total proteins extracted from whole-cell Chlamydomonas parental lines (dw15 and CC4533) and the two mdh2 mutants. The position of the insertion of the cassette APHVIII in the genome of Chlamydomonas reinhardtii was determined according to the method of restriction enzyme site-directed amplification-PCR (González-Ballester et al., 2005; Kong et al., 2017). These and all other primer sequences Protein Extraction and Immunoblot Analysis used in this study are listed in Supplemental Table 2. Exponentially grown cells were collected by centrifugation for 2 min at 1789g and resuspended in 1 mL lysis buffer (20 mM HEPES-KOH, pH

DNA Extraction, Gene Cloning, and Vector Construction for 7.2, 10 mM KCl, 1 mM MgCl2, 154 mM NaCl, and 0.1× protease inhibitor Complementation cocktail; Sigma-Aldrich). Cells were then sonicated for 90 s with an alternating cycle of 1-s pulse/1-s pause. The homogenates were centrifuged Genomic DNA was isolated from exponentially grown cells by the CTAB at 14,000g at 4°C for 10 min, and the supernatant was used as total pro- method (Schroda et al., 2001). The genomic DNA coding for MDH2 tein extracts. Protein concentrations were determined spectrophotomet- was cloned using primers EcoRI-MDH2-F1 and XbaI-MDH2-R1. The rically at 280 nm using a bicinchoninic acid protein assay kit (Bio-Rad). PCR reaction was performed using the high fidelity KOD Hot Start DNA The loading controls were visualized either by Ponceau red staining or Polymerase (Merck Millipore). The amplified DNA fragment was cloned Coomassie Brilliant Blue staining as detailed in the legend. For immuno- as an EcoRI-XbaI fragment into the pChlamy_4 vector (Life Technolo- blots, a given amount of total proteins were separated on SDS-PAGE and gies), which contains the ble gene conferring zeocin resistance (Stevens then transferred to a nitrocellulose membrane using the semidry tech- et al., 1996; Kong et al., 2015), generating pChlamy4-gMDH2. The target nique. Primary antibodies (at a dilution of 1/1000) included the specific gene was cloned in frame with a V5 tag at its 3′ end, allowing screen- polyclonal rabbit anti-V5 antibodies (Invitrogen; catalog no. PA1-993), ing complemented lines by anti-V5 (GKPIPNPLLGLDST) antibodies. The the rabbit anti-MDH2 antibodies (this study), anti-PGRL1 (Tolleter et al., plasmids of pChlamy4-gMDH2 were transformed into the mdh2-1 mu- 2011), anti-NDA2 (Baltz et al., 2014), anti-FLVB (Chaux et al., 2017), and −1 tant using electroporation, and independent zeocin (25 mg L ) resistant the anti-PSAD (catalog no. AS09-461), anti-PSBA (catalog no. AS01- clones were screened for TAG content by thin-layer chromatography 016), anti-Cytb6f (subunit f; catalog no. AS06-119), anti-LHCSR3 (catalog (TLC). Briefly, a given number of cells (usually 20 million) was collected no. AS14-2766), anti-ATPB (catalog no. AS05-085), anti-RBCL (catalog from each transformant grown to exponential phase, and total lipids no. AS03-037), and anti-CAT1 (catalog no. AS15-2991) antibodies were were extracted following the method described (Kong et al., 2017). purchased (Agrisera). Anti-rabbit horseradish peroxidase-conjugated an- TAGs were separated out from other lipid classes by migration on a TLC tibodies (Sigma-Aldrich; catalog no. AQ132P) (1/20,000) were used as plate in a solvent mixture of hexane/diethyl ether/acetic acid (17/3/0.2, the secondary antibody. The detection was performed with the G:BOX v/v/v) as described (Siaut et al., 2011). TAG51:0 (17:0/17:0/17:0) (Sigma- Chemi XRQ system (Syngene) using ECL detection reagents (GE Health- Aldrich) was used as a standard to allow TAG quantification by densitom- care). The images were captured using a CCD camera by GeneSys Im- etry. Detailed methods for TLC and quantification can be found in Siaut age Acquisition Software (Syngene), and chemiluminescent fluorescence et al. (2011). was quantitated by GeneTools Analysis Software (Syngene) according to the manufacturer’s instructions. DNA Gel Blot Analysis

To determine the number of insertions in the mutant genome, genomic Total Protein Extraction and Enzymatic Activity Assays DNA was digested with NotI or StuI. The digested fragments were separated and blotted onto positively charged nylon membranes (GE Healthcare). Cells at the logarithm phase (around 20 million cells total) were collected The membrane containing DNA was then hybridized with a digoxigenin and resuspended in extraction buffer (50 mM HEPES-KOH, 1 mM EDTA, (DIG)-labeled probe made with a fragment of the APHVIII gene. The probe 10% glycerol [v/v], 20 μL mL−1 protein protease inhibitor cocktail for plant was made using a PCR DIG probe synthesis kit (Sigma-Aldrich) by PCR cells [Sigma-Aldrich P9599], and 5 mM DTT at pH 7.5), and then lysed amplification with primers APHVIII-probe-F and APHVIII-probe-R. Imag- by sonication three times with a 10-s interval cycle. The concentrations ing was done using the luminescent image analyzer G:Box Chemi XRQ of extracted protein were determined spectrophotometrically at 280 nm system (Syngene). using a bicinchoninic acid protein assay kit (Bio-Rad) according to the manufacturer’s instructions.

RT-PCR MDH Activity Assay Total RNA was extracted as described before (Nguyen et al., 2013). The RNA was treated with DNase I (Ambion, Invitrogen) and then purified The activity of MDH was determined photometrically by measuring the with Nucleospin RNA Clean Up (Macherey Nagel). First-strand cDNA decrease in absorbance at 340 nm resulting from the oxidation of NADH was synthesized from 1 μg of the total RNA with the SuperScript VILO to NAD+ as previously described (Mekhalfi et al., 2014). Enzyme assays cDNA synthesis kit (Life Technologies) according to the manufacturer’s were performed in a 1-cm path-length cuvette containing 2× assay buffer instructions. For RT-PCR, the cDNA fragment of MDH2 was amplified by (90 mM KH2PO4-KOH, pH 7.4, 0.05% Triton X-100 [v/v], and 5 mM MgCl2) PCR using gene-specific primers MDH2-F2 and MDH2-R2. The RACK1 buffer with 0.2 mM NADH at 30°C. The reaction was initiated by addition (Cre06.g278222) gene was used as a housekeeping gene and amplified of 3 mM oxaloacetate (Sigma-Aldrich), and the rate of NAD+ formation with primers RACK1-F1 and RACK1-R1. was monitored at 340 nm in a Kontron Uvikon 810 spectrophotometer (Thermo Fisher Scientific). All the solutions for enzyme activity assays were freshly prepared prior to use. The enzyme activities were calculated Generation of Anti-MDH2 Antibodies based on the molar extinction coefficient of NAD(P)H at 340 nm ε( 340 = Antipeptide antibodies against MDH2 were made by immunizing rabbits 6220 M−1 cm−1), and the enzymatic reaction rate was calculated using a with the synthetic peptide (PVSEYAYIRHPPRL). Synthesis of the peptide, linear regression. One unit of MDH activity is defined as 1µ mol of NADH rabbit immunization, and purification of antibodies were performed by oxidized per min per mg of protein. Peroxisome Redox Status and Chloroplast Metabolism 1843

ACX Activity Assay chamber filled with a 1.5-mL sample). The bottom of the chamber was sealed by a Teflon membrane (13-µm thickness) allowing dissolved gases Stearoyl-CoA (C18:0-CoA; Sigma-Aldrich) at 50 mM was used as the to be directly introduced through a vacuum line into the ion source of substrate for determining ACX activity in total cell-free extracts by fol- the mass spectrometer (model Prima δB; Thermo Fisher Scientific) after lowing the production of H O , detected using the Amplex Red hydrogen 2 2 passing through a −65°C cooled water trap. Cells of Chlamydomonas peroxide/peroxidase assay kit (Invitrogen). A Xenius XC spectrofluorom- were grown to mid-log phase and then centrifuged at 450g for 4 min and eter (SAFAS) was used to measure the fluorescence emission at 580 nm resuspended in fresh high salt medium at pH 6.2 in a final concentration (excitation at 540 nm). Detailed protocols and quantification have been of 20 µg chlorophyll mL−1. The cell suspension was placed in the mea- described previously (Kong et al., 2017). suring chamber under constant stirring and bicarbonate (0.5 mM final

concentration) was added to saturate CO2. Before gas exchange mea- LD Imaging 18 18 surement, [ O]-enriched O2 (99% O2 isotope content; Euriso-Top) was A confocal laser scanning microscope (TCS SP2; Leica) was used to bubbled at the top of the suspension until reaching approximately equal concentrations of 16O and 18O . Gas abundances (N , 16O , 18O , and CO ) observe LDs in cells first fixed in 0.25% glutaraldehyde and subsequently 2 2 2 2 2 2 stained with BODIPY (4,4-difluoro-1,3,5,7-tetramethyl-4-bora-3a,4a-dia- were recorded adjusting the magnet current to the corresponding mass za-s-indacene; Thermo Fisher Scientific; D3921) at a final concentration peaks (m/z = 28, 32, 36, and 44, respectively). The measuring chamber of 10 g mL−1. A 63× oil immersion objective was used for all imaging was then sealed and gas exchanges recorded for 2 min until light was µ −2 −1 work, cells were excited with a 488-nm laser line, and emission was col- provided with three green LEDs at 600 μmol photons m s for 10 min. lected between 500 and 540 nm for BODIPY signal and between 650 and Gas concentrations were calculated according to Cournac et al. (2002). O exchange rates were calculated using equations from Radmer and 714 nm for chlorophyll autofluorescence. Pseudo colors were applied 2 using ZEN software (Carl Zeiss). Kok (1976).

Determination of Extracellular H2O2 Levels NAD(P)H Fluorescence Measurement

Chlamydomonas secrets H2O2 outside cell (Michelet et al., 2013). Extra- NAD(P)H fluorescence was measured as previously described (Kauny cellular H O produced can be determined using the Amplex Red reagent 2 2 and Sétif, 2014; Roach et al., 2015). Briefly, cells (OD684 = 0.26) were (the Amplex Red hydrogen peroxide/peroxidase assay kit; Invitrogen). centrifuged and resuspended in fresh medium, then dark-adapted for For this, 1.5 mL culture was centrifuged at 700g for 3 min in the dark. 1 min at room temperature inside the spectrophotometer before start- One milliliter of supernatant was mixed with the Amplex Red following ing the measurement. Light-induced measurements were performed at the manufacturer’s instructions. The fluorescence emission spectra was room temperature using NADPH/9-AA module of the DUAL-PAM (Walz). read at 580 nm (with excitation at 540 nm) using a SAFAS Xenius XC NAD(P)H fluorescence was first measured in the dark, then the actinic fluorescence spectrophotometer. A known amount of commercial H O −2 −1 2 2 light (light intensity 70 µmol photons m m provided by a red LED) was (VWR chemicals) was first used to generate a standard curve. Concentra- switched on for 20 s. After switching off actinic light, fluorescence was tion was calculated based on the standard curve and normalized by cell recorded again in the dark for 10 s. Thirty measurements were averaged concentrations of the culture media at the time of sampling. to obtain a good signal-to-noise ratio for each replicate.

Intracellular ROS Level Imaging Polar Metabolite Analysis by GC-MS Confocal microscopy images of cellular ROS were captured following the method described (Rastogi et al., 2010). Briefly, Chlamydomonas Chlamydomonas cells were grown to mid log phase, and 60 million cells cells (∼20 million) were collected by centrifugation, washed once, and were collected quickly by centrifugation at 13,000g for 10 s at 4°C. Cell resuspended in 1× PBS buffer containing 5.0 μM of the oxidant-sensing pellets were quickly frozen in liquid nitrogen until analysis. Metabolites fluorescent probe H2DCFDA (Thermo Fisher Scientific). After incubation extraction and derivatization using GC-MS were performed as described at room temperature in the dark for 30 min with gentle shaking, the sam- (Lisec et al., 2006). The GC-MS data were obtained using an Agilent 7683 ples were washed three times with 1× PBS buffer. The cells were exam- series auto-sampler (Agilent Technologies), coupled to an Agilent 6890 ined by a confocal laser scanning microscope (TCS SP2; Leica), for the gas chromatograph-Leco Pegasus two time-of-flight mass spectrometer detection of DCF fluorescence (excitation at 488 nm and emission 510 to (Leco). Identical chromatogram acquisition parameters were applied to 530 nm) and for chloroplast autofluorescence (excitation at 645 nm and those previously used (Cuadros-Inostroza et al., 2009). Chromatograms emission at 685 to 720 nm). The DCF fluorescence of 2 million cells was were exported from LECO CHROMATOF software (version 3.34) to also collected with a Cary Eclipse fluorescence spectrophotometer using R software. Ion extraction, peak detection, retention time alignment, excitation at 504 nm and emission at 524 nm. and library searching were obtained using the Target Search package from Bioconductor (Cuadros-Inostroza et al., 2009).

Electron Transport Rate Measurement Statistical Analysis Chlorophyll fluorescence was measured using a Dual Pulse Amplitude Modulated Fluorimeter (DUAL-PAM-100; Walz) as previously described Student’s t test was performed on biological replicates using Excel soft- (Schulz-Raffelt et al., 2016). Actinic light (supplied by a red LED light) ware (2013 version). increased stepwise (every 2 min) from 50 to 500 µmol photons m−2 s−1, and saturating flashes (10,000 mol photons m−2 s−1, 200-ms duration) µ Accession Numbers were applied to determine the PSII operating efficiency. Sequence data presented in this article can be found for Chlamydo- monas genes in Phytozome (https://phytozome.jgi.doe.gov/pz/portal. Gas Exchange Measurement by MIMS html#!info?alias=Org_Creinhardtii) with gene identifications as fol- Gas exchanges were monitored inside a water-jacketed and thermo- lows: MDH1 (Cre03.g194850), MDH2 (Cre10.g423250), MDH3 (Cre02. regulated (25°C) measuring chamber (using a modified Hansatech electrode g145800), MDH4 (Cre12.g483950), MDH5 (Cre09.g410700), RACK1 1844 The Plant Cell

du-Rhône, the French Ministry of Research, the CNRS, and the CEA. (g6364), FLVB (Cre16.g691800), Cytb6f (Cre16.g650100), PSAD (Cre05. g238332), LHCSR3 (Cre08.g367400), CAT1 (Cre09.g417150), NDA2 S.A. and A.R.F. thank the European Union for funding in the framework of (Cre19.g750547), PGRL1 (Cre07.g340200), ATPB (Cre17.g698000), the European Union 2020 TEAMING Project (SGA-CSA No 664621 and No 739582 under FPA No. 664620). RBCL (Cre06.g298300), psbA (ChreCp.021), APX (Cre09.g401886), HPR (Cre06.g295450), MDAR1 (Cre17.g712100), PXN (Cre07.g353300), and ACX2 (Cre05.g232002). For proteins of Arabidopsis, protein IDs can be AUTHOR CONTRIBUTIONS found in TAIR (www.arabidopsis.org). Y.L.-B. designed the project. F.K. performed all experiments except the Supplemental Data ones detailed below. A.B. performed MIMS measurements with the supervision of G.P. Y.L. analyzed enzymatic activities and RT-PCR. B.L. su- Supplemental Figure 1. Major Experimental Setup Used in This pervised lipid analysis. S.A. and Y.B. performed metabolomics analysis Study. with the supervision of A.R.F. A.K.-L. performed NAD(P)H measurement. Y.L.-B., F.B., G.P., and F.K. wrote the article with comments from A.R.F. Supplemental Figure 2. DNA Gel Blot of the APHVIII Gene Insertion and A.K.-L. Events in mdh2-1. Supplemental Figure 3. NAD-Dependent Total MDH Activity Analyses during Mixotrophic Growth. Received May 7, 2018; revised June 12, 2018; accepted July 10, 2018; Supplemental Figure 4. Cell Growth under Heterotrophic, Mixotrop- published July 11, 2018. hic, and Photoautotrophic Conditions. Supplemental Figure 5. Comparison of MDH2 Protein Level in Mix- otrophically and Photoautotrophically Grown Wild-Type (dw15) Cells. REFERENCES Supplemental Figure 6. The mdh2-1 Mutant Overaccumulates TAG Aboelmy, M.H., and Peterhansel, C. (2014). Enzymatic characterization and Starch 2 d after N Starvation during Photoautotrophic Growth of Chlamydomonas reinhardtii glycolate dehydrogenase and its near- When Expressed Based on Cell Volume. est proteobacterial homologue. Plant Physiol. Biochem. 79: 25–30. Supplemental Figure 7. The mdh2-2 Mutant Overaccumulates TAG Allorent, G., et al. (2013). A dual strategy to cope with high light in and Starch 2 d after N Starvation during Photoautotrophic Growth Chlamydomonas reinhardtii. Plant Cell 25: 545–557.

(+2% CO2 Supplemented in the Air) When Expressed Based on Cell Araújo, W.L., Tohge, T., Osorio, S., Lohse, M., Balbo, I., Krahnert, Volume. I., Sienkiewicz-Porzucek, A., Usadel, B., Nunes-Nesi, A., and Supplemental Figure 8. Fatty Acid Compositional Analysis by Gas Fernie, A.R. (2012). Antisense inhibition of the 2-oxoglutarate dehy- Chromatography-Mass Spectrometry. drogenase complex in tomato demonstrates its importance for plant respiration and during leaf senescence and fruit maturation. Plant Cell Supplemental Figure 9. PSII Operating Yield in mdh2-1 during Mix- 24: 2328–2351. otrophic N Deprivation. Bailleul, B., et al. (2015). Energetic coupling between plastids and mi- Supplemental Figure 10. Traces for Measurement of NAD(P)H Fluo- tochondria drives CO2 assimilation in diatoms. Nature 524: 366–369. rescence in Photoautotrophically Grown Cells. Ball, S., Dirick, L., Decq, A., Martiat, J., and Matagne, R. (1990). Phys- Supplemental Figure 11. The Level of Catalase 1 Protein Revealed by iology of starch storage in the monocellular alga Chlamydomonas re- Immunoblot Using Anticatalase Antibodies. inhardtii. Plant Sci. 66: 1–9. Baltz, A., Dang, K.-V., Beyly, A., Auroy, P., Richaud, P., Cournac, L., Supplemental Figure 12. Measurement of Extracellular H O Level in 2 2 and Peltier, G. (2014). Plastidial expression of Type II NAD(P)H dehy- mdh2-1 and Control Strains. drogenase increases the reducing state of plastoquinones and hydro- Supplemental Figure 13. ACX (18:0) Activity Analysis. gen photoproduction rate by the indirect pathway in Chlamydomonas reinhardtii. Plant Physiol. 165: 1344–1352. Supplemental Figure 14. H2O2 Supplementation on Cell Growth. Supplemental Table 1. Bioinformatic Comparison of the MDH Family Baslam, M., Baroja-Fernández, E., Ricarte-Bermejo, A., Sán- in Chlamydomonas to That of Land Plant Arabidopsis thaliana. chez-López, A.M., Aranjuelo, I., Bahaji, A., Muñoz, F.J., Almagro, G., Pujol, P., Galarza, R., Teixidor, P., and Pozueta-Romero, J. Supplemental Table 2. Primer Sequences Used in This Study. (2017). Genetic and isotope ratio mass spectrometric evidence for Supplemental Data Set 1. Polar Metabolite Profiling of the Wild Type, the occurrence of starch degradation and cycling in illuminated Ara- mdh2-1, and the Complementation Line C1 by GC-MS. bidopsis leaves. PLoS One 12: e0171245. Beeler, S., Liu, H.C., Stadler, M., Schreier, T., Eicke, S., Lue, W.L., Truernit, E., Zeeman, S.C., Chen, J., and Kötting, O. (2014). Plas- ACKNOWLEDGMENTS tidial NAD-dependent malate dehydrogenase is critical for embryo development and heterotrophic metabolism in Arabidopsis. Plant We thank John Ohlrogge for critical reading of the manuscript. We thank Physiol. 164: 1175–1190. Isabelle Thé, Audrey Beyly, Stéphan Cuiné, and Pascaline Auroy for Beevers, H. (1979). Microbodies in higher-plants. Annu. Rev. Plant laboratory assistance. Adrien Burlacot is a recipient of a CEA (Irtelis) Physiol. Plant Mol. Biol. 30: 159–193. international PhD studentship. Yuanxue Liang acknowledges the China Bernhardt, K., Wilkinson, S., Weber, A.P., and Linka, N. (2012). A per- Scholarship Council for a PhD studentship. Work in the authors’ lab- oxisomal carrier delivers NAD+ and contributes to optimal fatty acid oratory is supported by the French Agence Nationale pour la Recher- degradation during storage oil mobilization. Plant J. 69: 1–13. che (ANR JCJC MUsCA) and by the A*MIDEX project funded by the Blaby, I.K., Blaby-Haas, C.E., Pérez-Pérez, M.E., Schmollinger, “Investissements d’Avenir.” We acknowledge the European Union Re- gional Developing Fund, the Région Provence Alpes Côte d’Azur, the S., Fitz-Gibbon, S., Lemaire, S.D., and Merchant, S.S. (2015). French Ministry of Research, and the CEA for funding the HelioBiotec Genome-wide analysis on Chlamydomonas reinhardtii reveals the platform. Support for the microscopy equipment was provided by the impact of hydrogen peroxide on protein stress responses and over- Région Provence Alpes Côte d’Azur, the Conseil Général des Bouches- lap with other stress transcriptomes. Plant J. 84: 974–988. Peroxisome Redox Status and Chloroplast Metabolism 1845

Buléon, A., Gallant, D.J., Bouchet, B., Mouille, G., D’Hulst, C., Goold, H.D., Cuiné, S., Légeret, B., Liang, Y., Brugière, S., Auroy, Kossmann, J., and Ball, S. (1997). Starches from A to C. Chlamydomo- P., Javot, H., Tardif, M., Jones, B., Beisson, F., Peltier, G., and Li- nas reinhardtii as a model microbial system to investigate the biosyn- Beisson, Y. (2016). Saturating light induces sustained accumulation thesis of the plant amylopectin crystal. Plant Physiol. 115: 949–957. of oil in plastidal lipid droplets in Chlamydomonas reinhardtii. Plant Cagnon, C., Mirabella, B., Nguyen, H.M., Beyly-Adriano, A., Bouvet, Physiol. 171: 2406–2417. S., Cuiné, S., Beisson, F., Peltier, G., and Li-Beisson, Y. (2013). De- Graham, I.A. (2008). Seed storage oil mobilization. Annu. Rev. Plant Biol. velopment of a forward genetic screen to isolate oil mutants in the 59: 115–142. green microalga Chlamydomonas reinhardtii. Biotechnol. Biofuels 6: Hagemann, M., Kern, R., Maurino, V.G., Hanson, D.T., Weber, A.P., 178. Sage, R.F., and Bauwe, H. (2016). Evolution of photorespiration from Centeno, D.C., et al. (2011). Malate plays a crucial role in starch metab- cyanobacteria to land plants, considering protein phylogenies and olism, ripening, and soluble solid content of tomato fruit and affects acquisition of carbon concentrating mechanisms. J. Exp. Bot. 67: postharvest softening. Plant Cell 23: 162–184. 2963–2976. Chaux, F., Burlacot, A., Mekhalfi, M., Auroy, P., Blangy, S., Richaud, Harris, E.H. (2009). The Chlamydomonas Sourcebook: Introduction to P., and Peltier, G. (2017). Flavodiiron proteins promote fast and Chlamydomonas and Its Laboratory Use. (Oxford, UK: Elsevier). transient O2 photoreduction in Chlamydomonas. Plant Physiol. 174: Hayashi, Y., and Shinozaki, A. (2012). Visualization of microbodies in 1825–1836. Chlamydomonas reinhardtii. J. Plant Res. 125: 579–586. Comparot-Moss, S., et al. (2010). A putative phosphatase, LSF1, is re- Hebbelmann, I., et al. (2012). Multiple strategies to prevent oxidative quired for normal starch turnover in Arabidopsis leaves. Plant Physiol. stress in Arabidopsis plants lacking the malate valve enzyme NADP- 152: 685–697. malate dehydrogenase. J. Exp. Bot. 63: 1445–1459. Cournac, L., Latouche, G., Cerovic, Z., Redding, K., Ravenel, J., and Heyno, E., Innocenti, G., Lemaire, S.D., Issakidis-Bourguet, E., and Peltier, G. (2002). In vivo interactions between photosynthesis, mi- Krieger-Liszkay, A. (2014). Putative role of the malate valve enzyme torespiration, and chlororespiration in Chlamydomonas reinhardtii. NADP-malate dehydrogenase in H2O2 signalling in Arabidopsis. Phi- Plant Physiol. 129: 1921–1928. los. Trans. R. Soc. Lond. B Biol. Sci. 369: 20130228. Cuadros-Inostroza, A., Caldana, C., Redestig, H., Kusano, M., Hu, J., Baker, A., Bartel, B., Linka, N., Mullen, R.T., Reumann, S., and Lisec, J., Peña-Cortés, H., Willmitzer, L., and Hannah, M.A. (2009). Zolman, B.K. (2012). Plant peroxisomes: biogenesis and function. TargetSearch--a Bioconductor package for the efficient preprocessing Plant Cell 24: 2279–2303. of GC-MS metabolite profiling data. BMC Bioinformatics10: 428. Iglesias, A.A., Charng, Y.Y., Ball, S., and Preiss, J. (1994). Character- Curien, G., Flori, S., Villanova, V., Magneschi, L., Giustini, C., Forti, G., ization of the kinetic, regulatory, and structural properties of ADP- Matringe, M., Petroutsos, D., Kuntz, M., and Finazzi, G. (2016). The glucose pyrophosphorylase from Chlamydomonas reinhardtii. Plant water to water cycles in microalgae. Plant Cell Physiol. 57: 1354–1363. Physiol. 104: 1287–1294. Daloso, D.M., Medeiros, D.B., Dos Anjos, L., Yoshida, T., Araújo, Jenner, H.L., Winning, B.M., Millar, A.H., Tomlinson, K.L., Leaver, W.L., and Fernie, A.R. (2017). Metabolism within the specialized C.J., and Hill, S.A. (2001). NAD malic enzyme and the control of guard cells of plants. New Phytol. 216: 1018–1033. carbohydrate metabolism in potato tubers. Plant Physiol. 126: Dang, K.-V., Plet, J., Tolleter, D., Jokel, M., Cuiné, S., Carrier, P., 1139–1149. Auroy, P., Richaud, P., Johnson, X., Alric, J., Allahverdiyeva, Y., Kauny, J., and Sétif, P. (2014). NADPH fluorescence in the cyanobacte- and Peltier, G. (2014). Combined increases in mitochondrial cooper- rium Synechocystis sp. PCC 6803: a versatile probe for in vivo mea- ation and oxygen photoreduction compensate for deficiency in cyclic surements of rates, yields and pools. Biochim. Biophys. Acta 1837: electron flow in Chlamydomonas reinhardtii. Plant Cell 26: 3036–3050. 792–801. Dietz, K.-J., Turkan, I., and Krieger-Liszkay, A. (2016). Redox- and Kong, F., Yamasaki, T., Kurniasih, S.D., Hou, L., Li, X., Ivanova, N., reactive oxygen species-dependent signaling into and out of the pho- Okada, S., and Ohama, T. (2015). Robust expression of heterologous tosynthesizing chloroplast. Plant Physiol. 171: 1541–1550. genes by selection marker fusion system in improved Chlamydomo- Du, Z.Y., Lucker, B.F., Zienkiewicz, K., Miller, T.E., Zienkiewicz, A., nas strains. J. Biosci. Bioeng. 120: 239–245. Sears, B.B., Kramer, D.M., and Benning, C. (2018). Galactoglycer- Kong, F., Liang, Y., Légeret, B., Beyly-Adriano, A., Blangy, S., Haslam, olipid lipase PGD1 is involved in thylakoid membrane remodeling in R.P., Napier, J.A., Beisson, F., Peltier, G., and Li-Beisson, Y. (2017). response to adverse environmental conditions in Chlamydomonas. Chlamydomonas carries out fatty acid β-oxidation in ancestral per- Plant Cell 30: 447–465. oxisomes using a bona fide acyl-CoA oxidase. Plant J.90: 358–371. Eastmond, P.J. (2007). MONODEHYROASCORBATE REDUCTASE4 is Larosa, V., Meneghesso, A., La Rocca, N., Steinbeck, J., Hippler, M., required for seed storage oil hydrolysis and postgerminative growth Szabò, I., and Morosinotto, T. (2018). Mitochondria affect photosyn- in Arabidopsis. Plant Cell 19: 1376–1387. thetic electron transport and photo-sensitivity in a green alga. Plant Erickson, E., Wakao, S., and Niyogi, K.K. (2015). Light stress and pho- Physiol. 176: 2305–2314. toprotection in Chlamydomonas reinhardtii. Plant J. 82: 449–465. Lauersen, K.J., Willamme, R., Coosemans, N., Joris, M., Kruse, O., Fan, J., Yan, C., Andre, C., Shanklin, J., Schwender, J., and Xu, C. and Remacle, C. (2016). Peroxisomal microbodies are at the cross- (2012). Oil accumulation is controlled by carbon precursor supply for roads of acetate assimilation in the green microalga Chlamydomonas fatty acid synthesis in Chlamydomonas reinhardtii. Plant Cell Physiol. reinhardtii. Algal Res. 16: 266–274. 53: 1380–1390. Lemaire, S.D., Quesada, A., Merchan, F., Corral, J.M., Igeno, M.I., Geigenberger, P. (2011). Regulation of starch biosynthesis in response Keryer, E., Issakidis-Bourguet, E., Hirasawa, M., Knaff, D.B., and to a fluctuating environment. Plant Physiol.155: 1566–1577. Miginiac-Maslow, M. (2005). NADP-malate dehydrogenase from González-Ballester, D., de Montaigu, A., Galván, A., and Fernández, unicellular green alga Chlamydomonas reinhardtii. A first step toward E. (2005). Restriction enzyme site-directed amplification PCR: a tool to redox regulation? Plant Physiol. 137: 514–521. identify regions flanking a marker DNA. Anal. Biochem.340: 330–335. Lepistö, A., Pakula, E., Toivola, J., Krieger-Liszkay, A., Vignols, F., Goodson, C., Roth, R., Wang, Z.T., and Goodenough, U. (2011). Struc- and Rintamäki, E. (2013). Deletion of chloroplast NADPH-dependent tural correlates of cytoplasmic and chloroplast lipid body synthesis in thioredoxin reductase results in inability to regulate starch synthesis Chlamydomonas reinhardtii and stimulation of lipid body production and causes stunted growth under short-day photoperiods. J. Exp. with acetate boost. Eukaryot. Cell 10: 1592–1606. Bot. 64: 3843–3854. 1846 The Plant Cell

Li, X., Moellering, E.R., Liu, B., Johnny, C., Fedewa, M., Sears, B.B., Naranjo, B., Mignée, C., Krieger-Liszkay, A., Hornero-Méndez, Kuo, M.-H., and Benning, C. (2012). A galactoglycerolipid lipase is D., Gallardo-Guerrero, L., Cejudo, F.J., and Lindahl, M. (2016). required for triacylglycerol accumulation and survival following nitro- The chloroplast NADPH thioredoxin reductase C, NTRC, controls gen deprivation in Chlamydomonas reinhardtii. Plant Cell 24: 4670– non-photochemical quenching of light energy and photosynthetic 4686. electron transport in Arabidopsis. Plant Cell Environ. 39: 804–822. Li, X., Zhang, R., Patena, W., Gang, S.S., Blum, S.R., Ivanova, N., Yue, Nguyen, H.M., Cuiné, S., Beyly-Adriano, A., Légeret, B., Billon, E., R., Robertson, J.M., Lefebvre, P.A., Fitz-Gibbon, S.T., Grossman, Auroy, P., Beisson, F., Peltier, G., and Li-Beisson, Y. (2013). The A.R., and Jonikas, M.C. (2016). An indexed, mapped mutant library green microalga Chlamydomonas reinhardtii has a single ω-3 fatty enables reverse genetics studies of biological processes in Chlam- acid desaturase that localizes to the chloroplast and impacts both ydomonas reinhardtii. Plant Cell 28: 367–387. plastidic and extraplastidic membrane lipids. Plant Physiol. 163: Li-Beisson, Y., et al. (2013). Acyl-lipid metabolism. The Arabidopsis 914–928. Book 11: e0161, doi/10.1199/tab.0161. Niyogi, K.K. (2000). Safety valves for photosynthesis. Curr. Opin. Plant Li-Beisson, Y., Beisson, F., and Riekhof, W. (2015). Metabolism of Biol. 3: 455–460. acyl-lipids in Chlamydomonas reinhardtii. Plant J. 82: 504–522. Noctor, G., Queval, G., and Gakière, B. (2006). NAD(P) synthesis and Libessart, N., Maddelein, M.L., Koornhuyse, N., Decq, A., Delrue, B., pyridine nucleotide cycling in plants and their potential importance in Mouille, G., D’Hulst, C., and Ball, S. (1995). Storage, photosynthe- stress conditions. J. Exp. Bot. 57: 1603–1620. sis and growth - the conditional nature of mutations affecting starch Nunes-Nesi, A., Carrari, F., Lytovchenko, A., Smith, A.M., Loureiro, synthesis and structure in Chlamydomonas. Plant Cell 7: 1117–1127. M.E., Ratcliffe, R.G., Sweetlove, L.J., and Fernie, A.R. (2005). En- Lichtenthaler, H. (1987). Chlorophylls and carotenoids - pigments of hanced photosynthetic performance and growth as a consequence photosynthetic biomembranes. Methods Enzymol. 148: 350–382. of decreasing mitochondrial malate dehydrogenase activity in trans- Lisec, J., Schauer, N., Kopka, J., Willmitzer, L., and Fernie, A.R. genic tomato plants. Plant Physiol. 137: 611–622. (2006). Gas chromatography mass spectrometry-based metabolite Nunes-Nesi, A., Araújo, W.L., and Fernie, A.R. (2011). Targeting mito- profiling in plants. Nat. Protoc.1: 387–396. chondrial metabolism and machinery as a means to enhance photo- Liu, B., and Benning, C. (2013). Lipid metabolism in microalgae distin- synthesis. Plant Physiol. 155: 101–107. guishes itself. Curr. Opin. Biotechnol. 24: 300–309. Ohlrogge, J., and Browse, J. (1995). Lipid biosynthesis. Plant Cell 7: Lobo, A.K., de Oliveira Martins, M., Lima Neto, M.C., Machado, E.C., 957–970. Ribeiro, R.V., and Silveira, J.A. (2015). Exogenous sucrose supply Oswald, O., Martin, T., Dominy, P.J., and Graham, I.A. (2001). Plas- changes sugar metabolism and reduces photosynthesis of sugarcane tid redox state and sugars: interactive regulators of nuclear-encoded through the down-regulation of Rubisco abundance and activity. J. photosynthetic gene expression. Proc. Natl. Acad. Sci. USA 98: Plant Physiol. 179: 113–121. 2047–2052. Massoz, S., Larosa, V., Horrion, B., Matagne, R.F., Remacle, C., and Park, J.-J., Wang, H., Gargouri, M., Deshpande, R.R., Skepper, J.N., Cardol, P. (2015). Isolation of Chlamydomonas reinhardtii mutants Holguin, F.O., Juergens, M.T., Shachar-Hill, Y., Hicks, L.M., and with altered mitochondrial respiration by chlorophyll fluorescence Gang, D.R. (2015). The response of Chlamydomonas reinhardtii to ni- measurement. J. Biotechnol. 215: 27–34. trogen deprivation: a systems biology analysis. Plant J. 81: 611–624. Mekhalfi, M., Amara, S., Robert, S., Carrière, F., and Gontero, B. Peltier, G., Tolleter, D., Billon, E., and Cournac, L. (2010). Auxiliary (2014). Effect of environmental conditions on various enzyme activ- electron transport pathways in chloroplasts of microalgae. Photo- ities and triacylglycerol contents in cultures of the freshwater diatom, synth. Res. 106: 19–31. Asterionella formosa (Bacillariophyceae). Biochimie 101: 21–30. Pfannschmidt, T., Nilsson, A., and Allen, J.F. (1999). Photosynthetic Merchant, S.S., et al. (2007). The Chlamydomonas genome reveals control of chloroplast gene expression. Nature 397: 625. the evolution of key animal and plant functions. Science 318: 245– Plancke, C., et al. (2014). Lack of isocitrate lyase in Chlamydomonas 250. leads to changes in carbon metabolism and in the response to oxida- Mettler, I.J., and Beevers, H. (1980). Oxidation of NADH in glyoxysomes tive stress under mixotrophic growth. Plant J. 77: 404–417. by a malate-aspartate shuttle. Plant Physiol. 66: 555–560. Pracharoenwattana, I., Cornah, J.E., and Smith, S.M. (2007). Arabidop- Michalska, J., Zauber, H., Buchanan, B.B., Cejudo, F.J., and sis peroxisomal malate dehydrogenase functions in beta-oxidation Geigenberger, P. (2009). NTRC links built-in thioredoxin to light but not in the glyoxylate cycle. Plant J. 50: 381–390. and sucrose in regulating starch synthesis in chloroplasts and amy- Pracharoenwattana, I., Zhou, W., and Smith, S.M. (2010). Fatty acid loplasts. Proc. Natl. Acad. Sci. USA 106: 9908–9913. beta-oxidation in germinating Arabidopsis seeds is supported by per- Michelet, L., Roach, T., Fischer, B.B., Bedhomme, M., Lemaire, S.D., oxisomal hydroxypyruvate reductase when malate dehydrogenase is and Krieger-Liszkay, A. (2013). Down-regulation of catalase activ- absent. Plant Mol. Biol. 72: 101–109. ity allows transient accumulation of a hydrogen peroxide signal in Radmer, R.J., and Kok, B. (1976). Photoreduction of O2 primes and Chlamydomonas reinhardtii. Plant Cell Environ. 36: 1204–1213. replaces CO2 assimilation. Plant Physiol. 58: 336–340. Miginiac-Maslow, M., and Lancelin, J.M. (2002). Intrasteric inhibition Rastogi, R.P., Singh, S.P., Häder, D.P., and Sinha, R.P. (2010). Detec- in redox signalling: light activation of NADP-malate dehydrogenase. tion of reactive oxygen species (ROS) by the oxidant-sensing probe Photosynth. Res. 72: 1–12. 2′,7′-dichlorodihydrofluorescein diacetate in the cyanobacterium Mikkelsen, R., Mutenda, K.E., Mant, A., Schürmann, P., and Blennow, Anabaena variabilis PCC 7937. Biochem. Biophys. Res. Commun. A. (2005). Alpha-glucan, water dikinase (GWD): a plastidic enzyme 397: 603–607. with redox-regulated and coordinated catalytic activity and binding Recht, L., Zarka, A., and Boussiba, S. (2012). Patterns of carbohydrate affinity. Proc. Natl. Acad. Sci. USA102: 1785–1790. and fatty acid changes under nitrogen starvation in the microalgae Miller, R., et al. (2010). Changes in transcript abundance in Chlamydo- Haematococcus pluvialis and Nannochloropsis sp. Appl. Microbiol. monas reinhardtii following nitrogen deprivation predict diversion of Biotechnol. 94: 1495–1503. metabolism. Plant Physiol. 154: 1737–1752. Rhodin, J.A.G. (1954). Correlation of Ultrastructural Organization and Musrati, R.A., Kollárová, M., Mernik, N., and Mikulásová, D. (1998). Function in Normal and Experimentally Changed Proximal Convoluted Malate dehydrogenase: distribution, function and properties. Gen. Tubule Cells of the Mouse Kidney: An Electron Microscopic Study. Physiol. Biophys. 17: 193–210. (Stockholm, Sweden: Karolinska Institutet). Peroxisome Redox Status and Chloroplast Metabolism 1847

Roach, T., Na, C.S., and Krieger-Liszkay, A. (2015). High light-induced Thormählen, I., Zupok, A., Rescher, J., Leger, J., Weissenberger, S., hydrogen peroxide production in Chlamydomonas reinhardtii is in- Groysman, J., Orwat, A., Chatel-Innocenti, G., Issakidis-Bourguet, creased by high CO2 availability. Plant J. 81: 759–766. E., Armbruster, U., and Geigenberger, P. (2017). Thioredoxins play Saroussi, S.I., Wittkopp, T.M., and Grossmann, E.R. (2016). The type a crucial role in dynamic acclimation of photosynthesis in fluctuating II NADPH dehydrogenase facilitates cyclic electron flow, energy- light. Mol. Plant 10: 168–182. dependent quenching, and chlororespiratory metabolism during ac- Tolleter, D., et al. (2011). Control of hydrogen photoproduction by the climation of Chlamydomonas reinhardtii to nitrogen deprivation. Plant proton gradient generated by cyclic electron flow in Chlamydomonas Physiol. 170: 1975–1988. reinhardtii. Plant Cell 23: 2619–2630. Saroussi, S., Sanz-Luque, E., Kim, R.G., and Grossman, A.R. (2017). Trentacoste, E.M., Shrestha, R.P., Smith, S.R., Glé, C., Hartmann, Nutrient scavenging and energy management: acclimation respons- A.C., Hildebrand, M., and Gerwick, W.H. (2013). Metabolic engi- es in nitrogen and sulfur deprived Chlamydomonas. Curr. Opin. Plant neering of lipid catabolism increases microalgal lipid accumula- Biol. 39: 114–122. tion without compromising growth. Proc. Natl. Acad. Sci. USA 110: Scheibe, R. (2004). Malate valves to balance cellular energy supply. 19748–19753. Physiol. Plant. 120: 21–26. Tripathi, D.N., and Walker, C.L. (2016). The peroxisome as a cell signal- Schmollinger, S., et al. (2014). Nitrogen-sparing mechanisms in ing organelle. Curr. Opin. Cell Biol. 39: 109–112. Chlamydomonas affect the transcriptome, the proteome, and photo- Tsai, C.-H., Warakanont, J., Takeuchi, T., Sears, B.B., Moellering, synthetic metabolism. Plant Cell 26: 1410–1435. E.R., and Benning, C. (2014). The protein Compromised Hydrolysis Schroda, M., Vallon, O., Whitelegge, J.P., Beck, C.F., and Wollman, F.-A. of Triacylglycerols 7 (CHT7) acts as a repressor of cellular quiescence (2001). The chloroplastic GrpE homolog of Chlamydomonas: two in Chlamydomonas. Proc. Natl. Acad. Sci. USA 111: 15833–15838. isoforms generated by differential splicing. Plant Cell 13: 2823–2839. Tsai, C.H., Uygun, S., Roston, R., Shiu, S.H., and Benning, C. (2018). Schulz-Raffelt, M., Chochois, V., Auroy, P., Cuiné, S., Billon, E., Recovery from N deprivation is a transcriptionally and functionally Dauvillée, D., Li-Beisson, Y., and Peltier, G. (2016). Hyper-accumulation distinct state in Chlamydomonas. Plant Physiol. 176: 2007–2023. of starch and oil in a Chlamydomonas mutant affected in a plant- van Roermund, C.W.T., Elgersma, Y., Singh, N., Wanders, R.J.A., and specific DYRK kinase. Biotechnol. Biofuels9: 55. Tabak, H.F. (1995). The membrane of peroxisomes in Saccharomyces Schwarz, V., Andosch, A., Geretschläger, A., Affenzeller, M., and Lütz- cerevisiae is impermeable to NAD(H) and acetyl-CoA under in vivo Meindl, U. (2017). Carbon starvation induces lipid degradation via au- conditions. EMBO J. 14: 3480–3486. tophagy in the model alga Micrasterias. J. Plant Physiol. 208: 115–127. van Roermund, C.W., Schroers, M.G., Wiese, J., Facchinelli, F., and Selinski, J., and Scheibe, R. (2014). Lack of malate valve capacities Kurz, S. (2016). The peroxisomal NAD carrier from Arabidopsis im- lead to improved N-assimilation and growth in transgenic A. thaliana ports NAD in exchange with AMP. Plant Physiol. 171: 2127–2139. plants. Plant Signal. Behav. 9: e29057. Wynn, J.P., Kendrick, A., and Ratledge, C. (1997). Sesamol as an in- Siaut, M., Cuiné, S., Cagnon, C., Fessler, B., Nguyen, M., Carrier, P., hibitor of growth and lipid metabolism in Mucor circinelloides via its Beyly, A., Beisson, F., Triantaphylidès, C., Li-Beisson, Y., and Peltier, action on malic enzyme. Lipids 32: 605–610. G. (2011). Oil accumulation in the model green alga Chlamydomonas Xue, J., Niu, Y.-F., Huang, T., Yang, W.-D., Liu, J.-S., and Li, H.-Y. reinhardtii: characterization, variability between common laboratory (2015). Genetic improvement of the microalga Phaeodactylum strains and relationship with starch reserves. BMC Biotechnol. 11: 7. tricornutum for boosting neutral lipid accumulation. Metab. Eng. 27: Smith, A.M., and Stitt, M. (2007). Coordination of carbon supply and 1–9. plant growth. Plant Cell Environ. 30: 1126–1149. Xue, J., Wang, L., Zhang, L., Balamurugan, S., Li, D.-W., Zeng, H., Sparla, F., Costa, A., Lo Schiavo, F., Pupillo, P., and Trost, P. (2006). Yang, W.-D., Liu, J.-S., and Li, H.-Y. (2016). The pivotal role of malic Redox regulation of a novel plastid-targeted beta-amylase of Arabi- enzyme in enhancing oil accumulation in green microalga Chlorella dopsis. Plant Physiol. 141: 840–850. pyrenoidosa. Microb. Cell Fact. 15: 120. Stevens, D.R., Rochaix, J.D., and Purton, S. (1996). The bacterial Yilancioglu, K., Cokol, M., Pastirmaci, I., Erman, B., and Cetiner, S. phleomycin resistance gene ble as a dominant selectable marker in (2014). Oxidative stress is a mediator for increased lipid accumulation Chlamydomonas. Mol. Gen. Genet. 251: 23–30. in a newly isolated Dunaliella salina strain. PLoS One 9: e91957. Sweetlove, L.J., and Fernie, A.R. (2013). The spatial organization of me- Zhang, Y., Adams, I.P., and Ratledge, C. (2007). Malic enzyme: the tabolism within the plant cell. In Annual Review of Plant Biology, Vol. 64, controlling activity for lipid production? Overexpression of malic en- S.S. Merchant, ed (Palo Alto, CA: Annual Reviews), pp. 723–746. zyme in Mucor circinelloides leads to a 2.5-fold increase in lipid accu- Sweetlove, L.J., Lytovchenko, A., Morgan, M., Nunes-Nesi, A., Tay- mulation. Microbiology 153: 2013–2025. lor, N.L., Baxter, C.J., Eickmeier, I., and Fernie, A.R. (2006). Mito- Zhang, Y.M., Chen, H., He, C.L., and Wang, Q. (2013). Nitrogen starva- chondrial uncoupling protein is required for efficient photosynthesis. tion induced oxidative stress in an oil-producing green alga Chlorella Proc. Natl. Acad. Sci. USA 103: 19587–19592. sorokiniana C3. PLoS One 8: e69225. Interorganelle Communication: Peroxisomal MALATE DEHYDROGENASE2 Connects Lipid Catabolism to Photosynthesis through Redox Coupling in Chlamydomonas Fantao Kong, Adrien Burlacot, Yuanxue Liang, Bertrand Légeret, Saleh Alseekh, Yariv Brotman, Alisdair R. Fernie, Anja Krieger-Liszkay, Fred Beisson, Gilles Peltier and Yonghua Li-Beisson Plant Cell 2018;30;1824-1847; originally published online July 11, 2018; DOI 10.1105/tpc.18.00361

This information is current as of November 28, 2018

Supplemental Data /content/suppl/2018/07/10/tpc.18.00361.DC1.html References This article cites 110 articles, 46 of which can be accessed free at: /content/30/8/1824.full.html#ref-list-1 Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X eTOCs Sign up for eTOCs at: http://www.plantcell.org/cgi/alerts/ctmain CiteTrack Alerts Sign up for CiteTrack Alerts at: http://www.plantcell.org/cgi/alerts/ctmain Subscription Information Subscription Information for The Plant Cell and Plant Physiology is available at: http://www.aspb.org/publications/subscriptions.cfm

Saturating Light Induces Sustained Accumulation of Oil Primarily

Stored in Lipid Droplets of Plastidial Origin in Chlamydomonas

reinhardtii

Hugh Douglas Goold1,2, Stéphan Cuiné1, Bertrand Légeret1, Yuanxue Liang1, Sabine Brugière3, Pascaline Auroy1, Hélène Javot1, Marianne Tardif3, Brian Jones2, Fred Beisson1, Gilles Peltier1, Yonghua Li-Beisson1*

1CEA, CNRS, Aix Marseille Université, UMR7265, Institut de Biosciences et Biotechnologies, Cadarache, 13108, France 2Faculty of Agriculture and the Environment, University of Sydney, Australia 3CEA, INSERM, Université Grenoble Alpes, Institut de Biosciences et Biotechnologies de Grenoble, U1038, Grenoble, 38000, France

In this work, my contributions are:

I performed microscopic characterization fo lipid droplets in cells exposed to high light or after being N starved.

100

Saturating Light Induces Sustained Accumulation of Oil in Plastidal Lipid Droplets in Chlamydomonas reinhardtii1

Hugh Douglas Goold, Stéphan Cuiné, Bertrand Légeret, Yuanxue Liang, Sabine Brugière, Pascaline Auroy, Hélène Javot, Marianne Tardif, Brian Jones, Fred Beisson, Gilles Peltier, and Yonghua Li-Beisson* Commissariat à l’Energie Atomique, Centre National de la Recherche Scientiﬁque Aix Marseille Université, Unité Mixte de Recherche 7265, Institut de Biosciences et Biotechnologies, Cadarache 13108, France (H.D.G., S.C., B.L., Y.L., P.A., H.J., F.B., G.P., Y.L.-B.); Faculty of Agriculture and the Environment, University of Sydney, Sydney, New South Wales 2006, Australia (H.D.G., B.J.); and Commissariat à l’Energie Atomique, INSERM, Université Grenoble Alpes, Institut de Biosciences et Biotechnologies de Grenoble, Grenoble 38000, France (S.B., M.T.) ORCID IDs: 0000-0002-3000-3355 (S.C.); 0000-0002-4577-2747 (H.J.); 0000-0001-9995-7387 (F.B.); 0000-0003-1064-1816 (Y.L.-B.).

Enriching algal biomass in energy density is an important goal in algal biotechnology. Nitrogen (N) starvation is considered the most potent trigger of oil accumulation in microalgae and has been thoroughly investigated. However, N starvation causes the slow down and eventually the arrest of biomass growth. In this study, we show that exposing a Chlamydomonas reinhardtii culture to saturating light (SL) under a nonlimiting CO2 concentration in turbidostatic photobioreactors induces a sustained accumulation of lipid droplets (LDs) without compromising growth, which results in much higher oil productivity than N starvation. We also show that the polar membrane lipid fraction of SL-induced LDs is rich in plastidial lipids (approximately 70%), in contrast to N starvation-induced LDs, which contain approximately 60% lipids of endoplasmic reticulum origin. Proteomic analysis of LDs isolated from SL-exposed cells identiﬁed more than 200 proteins, including known proteins of lipid metabolism, as well as 74 proteins uniquely present in SL-induced LDs. LDs induced by SL and N depletion thus differ in protein and lipid contents. Taken together, lipidomic and proteomic data thus show that a large part of the sustained oil accumulation occurring under SL is likely due to the formation of plastidial LDs. We discuss our data in relation to the different metabolic routes used by microalgae to accumulate oil reserves depending on cultivation conditions. Finally, we propose a model in which oil accumulation is governed by an imbalance between photosynthesis and growth, which can be achieved by impairing growth or by boosting photosynthetic carbon ﬁxation, with the latter resulting in higher oil productivity.

Neutral lipid accumulation by microalgae has re- derivatives (Rosenberg et al., 2008; Wijffels and Barbosa, cently regained intensive interest because these orga- 2010; Khozin-Goldberg and Cohen, 2011). Most micro- nisms are considered promising as a feedstock for algal species do not accumulate large amounts of neutral the production of renewable fuels and fatty acid lipids (i.e. triacylglycerols [TAGs]) when grown under optimal conditions (Sheehan et al., 1998). Neutral lipid accumulation, however, can be induced by exposing 1 This work was supported by the French Agence Nationale pour cells to unfavorable culture conditions, such as removing la Recherche (grant nos. ANR–12–BIME–0001–02 and ANR–10– – – ’ nutritional elements (nitrogen [N], sulfur, iron, phos- INBS 08 01), the Région Provence Alpes Côte d Azur, the Conseil phate, etc.) from the media; increasing salinity or growth Général des Bouches-du-Rhône, the French Ministry of Research, the Centre National de la Recherche Scientiﬁque, the Commissariat à temperature (Moellering and Benning, 2010; Siaut et al., l’Energie Atomique, the European Union Regional Developing Fund, 2011; Urzica et al., 2013; Abida et al., 2015; Légeret et al., the University of Sydney for an Australian Postgraduate Award and 2016); or exposing cells to small chemically active mol- the Paul Finlay Scholarship to H.D.G., and the China Scholarship ecules (Kato et al., 2013; Kim et al., 2013, 2015). Most Council for a postgraduate award to Y.L. of the current understanding of TAG metabolism in * Address correspondence to [email protected]. Chlamydomonas reinhardtii has been gained through the The author responsible for distribution of materials integral to the study of molecular mechanisms occurring during the N ﬁndings presented in this article in accordance with the policy de- starvation response (Fan et al., 2011; Goodson et al., scribed in the Instructions for Authors (www.plantphysiol.org) is: 2011; Siaut et al., 2011; Tsai et al., 2014, 2015). It is un- Yonghua Li-Beisson ([email protected]). certain if the mechanisms of TAG accumulation upon N H.D.G., G.P., B.J., F.B., and Y.L.-B. designed research; H.D.G., S.C., B.L., Y.L., H.J., S.B., and P.A. performed research; H.D.G., M.T., B.L., starvation are generally applicable or whether different Y.L., H.J., M.T., F.B., G.P., and Y.L.-B. analyzed data; H.D.G., F.B., mechanisms are employed under other types of condi- M.T., G.P., and Y.L.-B. wrote the article; all authors agreed on the tions. A major limitation of the use of microalgae to article. produce oil is the fact that N deprivation, as well as most www.plantphysiol.org/cgi/doi/10.1104/pp.16.00718 other TAG-inducing conditions, provoke impairments

Ò 2406 Plant Physiology , August 2016, Vol. 171, pp. 2406–2417, www.plantphysiol.org Ó 2016 American Society of Plant Biologists. All Rights Reserved. Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. High Light Induces Reserve Accumulation in C. reinhardtii in protein synthesis and cell division, thus limiting pro- cellular oil content and to much higher oil productivity ductivity (Hu et al., 2008; Scott et al., 2010). than observed in response to N depletion. A general Biomass productivity is the result of highly coordi- model in which oil accumulation is the result of an im- nated cellular processes, starting with the capture of balance between photosynthetic activity and demand for fi light by photosystems, the xation of CO2 through the growth is discussed. We also provide insights into the Calvin-Benson cycle, and cell growth and division. mechanism of oil accumulation under SL by reporting the Light is one of the most variable environmental pa- proteomic and lipidomic analyses of SL-induced LDs. rameters during the growth of photoautotrophs in fi natural environments. In nonsaturating light, CO2 x- ation and biomass productivity increase linearly as a RESULTS function of light intensity. Above a certain threshold, light saturation occurs. A considerable body of work Experimental Setup and Choice of Light Intensity has documented the effects of high light on photosyn- To investigate the formation of carbon reserves in re- thesis, including effects on the pigment content sponse to high light under photoautotrophic conditions, (Bonente et al., 2012), on the induction of dissipation or it is paramount that light supply and penetrance be kept protection mechanisms (Peers et al., 2009), and on the uniform throughout all culture stages. When cultivat- production of reactive oxygen species (Fischer et al., ing algae in batch cultures under current laboratory 2006; Förster et al., 2006; Erickson et al., 2015; Sato et al., conditions, the biomass increase leads to a decrease of 2015). The effect of light intensity on carbon allocation light perceived per cell due to shading in dense cultures. and reserve formation also has been explored (Pal et al., To achieve constant light perception, we cultivated 2011; Fan et al., 2012; Klok et al., 2013; He et al., 2015). C. reinhardtii in continuous culture in photobioreactors For example, increasing light intensity has been shown (PBRs) monitored as turbidostats. This allows for the to increase the cellular neutral lipid content in a num- cultures to be maintained at constant biomass levels (i.e. ber of microalgal species, including Haematococcus under a defined physiological condition) throughout the pluvialis (Zhekisheva et al., 2002), Tichocarpus crinitus course of the experiment with minimal differences be- (Khotimchenko and Yakovleva, 2005), and C. reinhardtii tween cultures in terms of light exposure and other pa- (Mettler et al., 2014). Molecular factors involved in TAG rameters, such as pH, nutrient, or CO2 supply. storage under high light are still to be uncovered. We first measured biomass productivity as a function Oil accumulation is associated with the formation of fl of photon ux density (Fig. 1). A nonlimiting CO2 lipophilic droplets, called lipid droplets (LDs [or oil condition was achieved by bubbling air containing bodies or oleosomes]; Jolivet et al., 2013). LDs are spe- 1.8% CO2. The growth rate increased progressively cialized intracellular organelles made of a neutral lipid with rising photon flux density in the range of 0 to core surrounded by a membrane lipid coat in which proteins are embedded (Huang, 1996). LDs serve as a temporary storage site for neutral lipids and also participate in the active synthesis and metabolism of these non-membrane-forming lipids (Goodman, 2008; Farese and Walther, 2009; Chapman et al., 2012; Goold et al., 2015; Tsai et al., 2015). The current model of LD biogenesis suggests that these lipid-rich subcellular structures arise from membrane budding or blistering; thus, the lipid molecules present in the LD lipid coat suggest its origin of biogenesis. For example, oil bodies in the oilseed are coated by a monolayer of lipids of mostly endoplasmic reticulum (ER) origin (Huang, 1992; Tzen and Huang, 1992), whereas plastoglobules are covered by lipids usually found as part of the thylakoid membranes (Austin et al., 2006). Thus, the protein and lipid composition of LDs can shed light on the likely subcellular location of TAG synthesis and LD biogenesis. Compared with oilseeds, the unicellular microalga C. reinhardtii offers an excellent platform in which to study LD formation, due to the ease with which researchers Figure 1. Growth rate of biomass as a function of photon flux density in can induce LD biogenesis or degradation by simply turbidostatic PBRs. Strain CC-124 was cultivated in PBRs, with a supply of 1.8% CO2 in air. Growth rate was calculated as the rate of injection of changing the culture medium (Goodson et al., 2011; fresh medium to an existing culture while maintaining an optical den- Goold et al., 2015; Tsai et al., 2015). sity at 880 nm (OD880) of 0.4 (equivalent to approximately 2 million 2 Here, we used a turbidostatic continuous cultivation cells mL 1). The light intensities used in the rest of this study are indi- 2 2 system to study the effect of light intensity on oil accu- cated by blue arrows. Low light (LL) refers to 37 mmol m 2 s 1, and SL 2 2 mulation and productivity in C. reinhardtii cells. We report refers to 200 mmol m 2 s 1. Data are means of two biological replicates that exposure to a saturating light (SL) leads to increased and two technical replicates each. Error bars represent SD.

Plant Physiol. Vol. 171, 2016 2407 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. Goold et al.

2 2 100 mmol m 2 s 1, but when light reached more than 2 2 120 mmol m 2 s 1, a plateau in biomass growth was observed. Light intensities above that threshold are thus SL. To evaluate the effects of light intensity on carbon reserve formation, based on this first experiment, cells of C. reinhardtii were cultivated first at a sub- 2 2 saturating LL (37 mmol m 2 s 1) and then exposed to SL 2 2 of 200 mmol m 2 s 1 for 48 h (blue arrows in Fig. 1). It is worth noting that the light intensity at which the saturation of photosynthesis and growth was obtained un- 2 2 der this experimental setup (around 200 mmol m 2 s 1) may appear low compared with what is generally used in most laboratory conditions (i.e. shake flasks with monodirectional illumination). Such an LL saturation level resulted from the use of a relatively low-density culture (OD880 = 0.4) and radial illumination (creating a homogenous light inside the PBR). Since high light is a relative term, to avoid ambiguity, in this study we use the term SL throughout. To dissect mechanistic differences in carbon reserve formation between cells exposed to SL and those cultivated under N-starvation conditions, we set up four 1-L PBRs, two for SL experiments and two for the N-depletion experiments. To reflect real production conditions, we employed a gradual N-depletion process rather than an abrupt N starvation. The experimental setups used and the parameters analyzed in the rest of the work are outlined in Figure 2.

Exposure to SL Increases Intracellular Oil and Starch Contents Figure 2. Experimental setup for the comparison of N starvation and SL

fi conditions. Strain CC-124 was maintained at a constant OD880 of 0.4 Quanti cation of intracellular oil and starch showed 2 (equivalent to approximately 2 million cells mL 1). SL refers to a light that SL-exposed cells accumulated both starch and TAGs m 22 21 in a parallel manner (Fig. 3A). Oil and starch accumu- intensity of 200 mol m s . N starvation was achieved through progressive limitation, and culture N level was measured. MM, Minimal lation also was observed in N-starved cells under pho- medium. toautotrophic PBR conditions (Fig. 3B). Maximal oil accumulation measured in SL-exposed cells (approxi- 2 2 mately 6 mgmm 3 at 9 h) reached almost 50% of the reached at 0.12 h 1 (Fig. 4A). In sharp contrast, N de- level determined for N-starved cells (approximately 2 pletion led to a progressive decrease in the growth rate 12 mgmm 3 at 24 h after N depletion), and maxi- until it halted completely 6 h after full N depletion (Fig. mal starch accumulation was similar (approximately 2 4B). Calculation of oil and starch productivity using 80 mgmm 3) under both conditions. A transient peak of specific growth rates with cellular oil and starch con- TAG/starch accumulation was seen around 9 h upon SL tents revealed that, compared with N starvation, SL- exposure, followed by a drop and stabilization at 24 h. exposed cells achieved a 15 times higher productivity in 2 2 2 2 The stabilized level of reserves was still much higher than TAG (approximately 6 mg L 1 d 1 versus 0.4 mg L 1 d 1) before SL exposure but lower than the maximal level and a 7 times higher productivity in starch (approxi- 2 2 2 2 observed under N depletion. Moreover, increases in total mately 70 mg L 1 d 1 versus 10 mg L 1 d 1;Fig.4,Cand fatty acids and also in carbon/nitrogen (C/N) ratio (from D). Therefore, these results clearly show that sustained 4.8 to 5.7) were observed in cells exposed to SL compared productivity of oil and starch could be achieved in PBRs with those cultivated under LL (Supplemental Fig. S1), under SL. supporting the TAG quantification data (Fig. 3).

SL-Exposed Cells Accumulate Oil in LDs Oil and Starch Productivities Are Higher in SL-Exposed Cultures Than in N-Starved Cultures Oil accumulation under SL exposure was found to form distinct LDs that were visible after Nile Red Upon SL exposure, the speciﬁc growth rate of the C. staining (Fig. 5, A and B). Interestingly, a signiﬁcant reinhardtii culture grown in PBRs increased progres- portion of LDs in SL-exposed cells were smaller than 2 sively from 0.05 h 1 to a new steady state, which was LDs under N starvation. This contrasted to those

2408 Plant Physiol. Vol. 171, 2016 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. High Light Induces Reserve Accumulation in C. reinhardtii

further examined the difference in the nature of LD populations under SL and N starvation by analyzing the protein and lipid composition of isolated LDs.

LDs in SL-Exposed Cells Contain TAGs and Polar Lipids of Mostly Plastidial Origin To determine lipids and proteins present in LDs, we isolated LDs from both SL-exposed cells and N-starved cells cultivated in PBRs following 9 h of SL exposure or 24 h of N starvation, respectively. These time points were chosen because they correspond to the moment when the maximal cellular oil content was reached. After lipids were extracted from isolated LDs, polar and neutral fractions were separated by solid-phase extraction, and the proportion of each fraction was quantiﬁed after being converted to their fatty acid methyl esters. LDs of N-starved cells contained 97% neutral lipids and 3% polar lipids, whereas those of SL- exposed cells contained only 81% neutral lipids and 19% polar lipids (Fig. 6A). In a typical LD topology, neutral lipids are covered by a polar membrane shell (Jolivet et al., 2013). Since smaller LDs contain a higher proportion of polar membrane lipids than larger LDs, this suggests that the LDs in SL-exposed cells are

Figure 3. Exposure to SL triggers oil and starch accumulation. Cells were cultivated in PBRs in photoautotrophic conditions. Time zero in the SL and N-starvation experiments corresponds to LL and +N controls, respectively. Data are means of two biological replicates with two technical replicates. Error bars represent SD. MM, Minimal medium. formed under N starvation, which were bigger and more abundant (Fig. 5C). It is known that N-starved C. reinhardtii wild-type cells accumulate LDs mostly at the junction of the ER and plastid envelopes (the so-called b-cytoplasm [cyto]-LDs; Goodson et al., 2011). Here, we thus focused only on the subcellular location of the LDs present in the SL-treated cells. This was achieved by obtaining confocal microscopy images of several focal planes. For example, when moving the focal plane from Figure 4. Growth rates and oil and starch productivity under SL or N the top all the way to the bottom of the cell, we observed starvation. A, Growth rates under SL. B, Growth rates under N starva- the presence or absence of LDs (Fig. 5, D–F). This con- tion. C, Productivity of oil and starch under SL. D, Productivity of oil and trasted with the detection of chlorophyll auto- starch under N starvation. PBRs were first stabilized for 4 d and then fl exposed either to SL irradiance or N depletion. The growth rate was uorescence. These observations suggested that at least calculated based on the accumulated medium added to the PBR culture some of the LDs formed were likely present inside the to maintain a constant optical density during cell culture. Red dotted plastid (Fig. 5D; the white arrow points to one example lines indicate the moment SL is switched on, and green dotted lines of this LD type). LDs that appeared to be cytosolic or indicate when N is completely depleted from the medium. Data are associated with plastid membranes in the same cell also means of two biological replicates and two technical replicates. Error could be seen. To confirm these observations, we bars denote SD.

Plant Physiol. Vol. 171, 2016 2409 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. Goold et al.

Figure 5. Visualization of LDs by Nile Red staining and microscopy. A, C. reinhardtii cells cultivated under normal conditions (minimal medium [MM] + N/LL). B, C. reinhardtii cells after 9 h of SL (SL + N). C, C. reinhardtii cells after 24 h of N depletion (MM 2 N/LL). D, The top focal plane of an image of a Nile Red- stained SL-exposed cell. E, The middle focal plane of an image of a Nile Red-stained SL- exposed cell. F, The bottom focal plane of an image of a Nile Red-stained SL-exposed cell.

smaller than those induced by N starvation, which was LDs of SL-Exposed Cells Contained Higher Proportions of consistent with microscopic examinations (Fig. 5). Proteins of Plastidial Origin The polar lipids of LDs isolated from SL-exposed cells showed high contents of monogalactosyldiacylglycerol In order to gain further insights into LD formation (MGDG) and sulfoquinovosyldiacylglycerol (SQDG), under SL, we performed a proteomic analysis to com- followed by phosphatidylglycerol and diacyl- pare the protein compositions of LDs accumulating in glycerol N,N,N-trimethylhomoserine (DGTS), then SL with those isolated from N-starvation conditions. digalactosyldiacylglycerol (DGDG), phosphatidylino- Total proteins associated with LDs isolated under SL or sitol, and phosphatidylethanolamine (PtdEtn; Fig. 6B). N-starvation conditions were loaded and separated on This composition clearly contrasts with the LDs of an SDS-PAGE device. Protein extracts were loaded N-starved cells, where DGTS and SQDG are the major based on an equal amount of total lipids. LDs formed polar lipid species (Fig. 6B), which corroborates previ- under SL showed a distinct protein composition com- ous observations (Wang et al., 2009). With the exception pared with those of N-induced LDs (Fig. 7A). Proteomic fi of phosphatidylinositol, molecular species within a analyses by LC-MS/MS identi ed 222 proteins associ- particular lipid class also differed between these two ated with LDs in SL-exposed cells and 303 proteins in LD types (Supplemental Fig. S2). LDs induced by SL LDs formed under N starvation (Fig. 7B; Supplemental contained a high proportion (approximately 70%) of Table S1, A and B). To be stringent, we kept only those lipids unique to photosynthetic membranes (i.e. proteins for which two or more unique peptides and MGDG, DGDG, and SQDG), whereas LDs of cells in the two or more SSCs were identified. Proteins were ranked N-starvation condition contained approximately 60% according to their estimated relative abundance within extraplastidial lipids (DGTS and PtdEtn). each LD biogenesis condition. Comparisons between TAG molecular species present in the two LD types the two proteomes showed that 148 proteins were also were found to differ. Compared with TAG species identified in both proteomes (Supplemental Table S1C), present in cells under optimal growth conditions (i.e. LL whereas 74 unique proteins were identified in SL- and nutrient replete), SL-induced LDs contained sig- induced LDs (Supplemental Table S1D), a large portion nificantly higher proportions of all TAG50 species, yet of which (31 of 74) were of plastidial origin based on a with significant reductions in TAG species rich in prediction using the green algae protein-adapted algo- polyunsaturated fatty acids (TAG52:10 and TAG52:11; rithm PredAlgo (Tardif et al., 2012). We detected sig- Fig. 6C). Interestingly, these polyunsaturated TAG nificant amounts of plastidial proteins in both LD species are present in higher proportions in LDs induced preparations, but we could not tell if these proteins are by SL than in those triggered by N starvation (Fig. 6C). contaminations or are integral parts of the LD proteome.

2410 Plant Physiol. Vol. 171, 2016 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. High Light Induces Reserve Accumulation in C. reinhardtii

Figure 6. Lipid composition of LDs isolated from SL-exposed cells and N-starved cells. A, Proportion of membrane lipids versus TAG in isolated LDs. Lipid amounts were determined by the quantiﬁcation of total fatty acids in each lipid fraction. B, Proportion of major membrane lipid classes in isolated LDs. The content for each lipid class was determined by thin-layer chromatography (TLC). C, Distribution of TAG molecular species in isolated LDs. LDs were isolated from cells N starved for 24 h (corresponding to 2 2 48 h in Fig. 3) or from cells exposed to SL (200 mmol m 2 s 1) for 9 h. These time points were chosen because they represent the peak of oil accumulation under either treatment. Data are means of two biological and two technical replicates, with SD shown. FA, Fatty acid; MM-N, minimal medium under N starvation; PtdGro, phosphatidylglycerol; PtdIns, phosphatidylinositol.

These questions should be addressed in the future emPAI values and can be used to evaluate the relative through studies of protein subcellular localizations. abundance of each protein in a sample. The identities and annotations for the top 30 ranked Among the 74 proteins found only in LDs of SL ori- proteins associated with LDs isolated from SL-exposed gin, of particular note is the presence of the relatively cells are listed in Table I. Proteins common to both LD large proportion (approximately 45%) of proteins of isolates include the major lipid droplet protein (MLDP), plastidial origin, including Proton Gradient Regula- the DGTS synthesis protein (BTA1), the cyclopropane- tion Like1 (PGRL1), Plastid Lipid Associated Protein8 fatty acyl-phospholipid synthase (CFA2), an a/b- (PLP8; Cre14.g611700), and a 2-methyl-6-phytyl-1,4- hydrolase, a glycosyl hydrolase (GHL1), a long-chain benzoquinone/2-methyl-6-solanyl-1,4-benzoquinone acyl-CoA synthetase (LCS2), a b-tubulin (TUB1), a methyltransferase (VTE3; Supplemental Table S1D). lysophospholipid acyltransferase (LPLAT), and one of the components of the plastid lipid reimport machinery (TGD2; Table II; Supplemental Table S1C). These pro- DISCUSSION teins were found to be present in all three previously reported LD proteomics studies (Moellering and In this study, we showed that SL can increase cel- Benning, 2010; Nguyen et al., 2011; Tsai et al., 2015). With lular oil content while maintaining cell turnover and the exception of TGD2, all of the above-mentioned top- biomass growth, thus leading to a sustainable pro- ranked proteins in N-induced LDs are not the same duction of carbon reserves. This clearly contrasts with major proteins in SL-induced LDs (i.e. their ranks in SL- the well-characterized N starvation-induced oil accu- induced LDs are relatively inferior to those in N-induced mulation, where oil productivity remained low de- LDs; Table II). These rankings were calculated based on a spite achieving a higher cellular TAG content. We computed calculation of the total number of SSCs and further showed that SL induces the formation of

Plant Physiol. Vol. 171, 2016 2411 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. Goold et al.

In addition to the increased carbon ﬂow to fatty acid synthesis, the composition of total polar lipids was altered following exposure to SL. SL-treated cells contained a reduced proportion of MGDG and increased proportions of DGDG and DGTS, with the other major lipids remaining constant (Supplemental Fig. S1C). Taken together, these data suggest that, under SL, both de novo fatty acid synthesis and membrane lipid remodeling contribute to TAG formation, a phenomenon similar to what occurs in cells starved for N.

Biogenesis and Physiological Functions of LDs under SL Most of our knowledge of LD biology in C. reinhardtii is gained from studies of LDs induced by N starvation. Figure 7. Comparison of the proteomic composition of LDs isolated Three types of LDs (a-cyto-LDs, b-cyto-LDs, and chlo- from cells subjected to SL or N starvation. A, Silver nitrate-stained SDS- roplast [cpst]-LDs) have been distinguished in C. rein- PAGE gel of total proteins extracted from purified LDs. B, Venn diagram hardtii upon N starvation; the type and relative showing the total number of proteins associated with LDs under each abundance of each LD population differs depending on condition. Only proteins with two or more peptides and two or more the cultivation conditions and on the genotype (Fan specific spectral counts (SSCs) were retained for analyses. LDs were et al., 2011; Goodson et al., 2011). For example, it has purified from cultures after 24 h of N depletion or from cultures after been observed that optimally growing cells contain a being exposed to SL for 9 h. LC-MS/MS, Liquid chromatography-tandem a mass spectrometry; MM-N, minimal medium under N starvation. few -cyto-LDs (i.e. free cytosolic LDs), whereas upon N starvation, wild-type strains of C. reinhardtii store oil mostly in b-cyto-LDs (i.e. LDs formed at the junction smaller LDs than those formed under N starvation. between the ER and the plastid envelope), whereas the The SL-induced LDs had a lipid and protein coat starchless mutant bafJ5 accumulates cpst-LDs (i.e. enriched in components of plastidial origin; thus, LDs chloroplast-located LDs) in addition to b-cyto-LDs formed under SL are distinct from those formed under (Goodson et al., 2011). Here, we observed that SL also N starvation. This serves as the firstreportonLD can trigger LD formation. Part of the LDs seem to be composition beyond the model of N starvation- inside the plastids (cpst-LDs) and part of them outside induced LDs in C. reinhardtii. the plastid, but in most cases they seem to be associated with the plastid envelopes (b-cyto-LDs; Fig. 8). Origin of TAGs under SL Lipidomics analyses showed that LDs formed under SL exposure are made of a mixture of saturated and An increase in total fatty acids was seen together polyunsaturated TAGs and covered by a protein and with an increase in C/N ratio in cells exposed to SL lipid coat rich in components of plastidial origin. The (Supplemental Fig. S1). The positive correlation be- presence of a plastidial TAG biosynthesis pathway is tween TAG levels and C/N ratio also was observed not surprising; it has been observed previously in the previously in the seeds of Arabidopsis (Arabidopsis N-starved starchless mutant bafJ5 (Fan et al., 2011; thaliana) when the plant was grown under increasing Goodson et al., 2011) and also can be deduced by the irradiance (Li et al., 2006). Thus, after being exposed to subcellular targeting of key TAG enzymes in the plastid SL, the cells readjust their metabolism to store the extra (Li-Beisson et al., 2015). carbon and energy as energy-dense materials (i.e. starch The protein and lipid composition of the LD coat and oil). These storage compounds provide a tempo- further demonstrated that SL-induced LDs consist of rary reservoir to accommodate the extra fluxes of ATP components of mostly plastidial origin. For example, and reducing powers, thereby ensuring cell energetic besides common proteins, SL-induced LDs were found homeostasis. This increased de novo fatty acid synthe- to contain a significant subset of plastid-resident pro- sis is further reflected in the changes in total fatty acid teins, including PGRL1, PLP8, and VTE3. PGRL1 is a composition before and after SL exposure (Fig. 6C; major component of cyclic electron flow in C. reinhardtii Supplemental Fig. S1D; i.e. increased proportion of and has been proposed to supply additional ATP for saturated C16 fatty acid and reduction of polyunsatu- photosynthesis, particularly in conditions of high ATP rated fatty acid species). Increased carbon flow to TAG demand (Tolleter et al., 2011; Dang et al., 2014). The synthesis under high light was reported recently in the PGRL1-deficient mutant (pgrl1) was reported recently heterokont alga Nannochloropsis oculata, in which the to accumulate less neutral lipids under N starvation, activities of acetyl-CoA carboxylase and diacylglycerol and it was suggested that cyclic electron flow supplies acyltransferase are increased significantly following a ATP for lipid biosynthesis during N starvation (Chen transition from low-light to high-light cultivation con- et al., 2015). The association of PGRL1 with LDs of SL- ditions (Ma et al., 2016). exposed cells expands its role in supplying ATP for

2412 Plant Physiol. Vol. 171, 2016 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. High Light Induces Reserve Accumulation in C. reinhardtii

Table I. Identity and abundance of the top 30 proteins found to be associated with LDs isolated from cells exposed to SL for 9 h emPAI, Exponentially modiﬁed protein abundance index; \N, no annotation available. The rankings were calculated based on a computed calculation of the total number of SSCs and emPAI values and can be used to evaluate the relative abundance of each protein in a sample. Identiﬁer (Phytozome or Joint Genome Institute) Symbol Name SSC emPAI Rank Cre61.g792450.t1.2 \N \N 48.5 0.765 1 Cre07.g324200.t1.1 BTA1 DGTS synthesis protein 40.5 2.071 2 Cre01.g013700.t1.1 \N \N 24.5 7.566 3 gi|41179057|ref|NP_958414.1| \N ATP synthase CF1 b-subunit (chloroplast encoded) 24 1.978 4 gi|41179032|ref|NP_958388.1| \N PSII 47-kD protein (chloroplast encoded) 24 1.452 5 Cre17.g698000.t1.1 ATP2 Mitochondrial F1F0 ATP synthase, b-subunit 21 1.683 6 Cre12.g540550.t1.2 a/b-hydrolase \N 21 1.523 7 Cre05.g241950.t1.1 \N \N 20.5 6.462 8 Cre41.g786600.t1.1 AAA1 Plastidic ADP/ATP translocase 20.5 1.239 9 gi|41179063|ref|NP_958420.1| \N PSII protein D2 (chloroplast encoded) 20 1.713 10 Cre04.g229300.t1.2 RCA1 Rubisco activase 19 2.027 11 Cre12.g498600.t1.1 \N Eukaryotic translation elongation factor 1a 18 1.612 12 gi|41179065|ref|NP_958422.1| \N PSII 44-kD protein (chloroplast encoded) 17.5 1.394 13 Cre09.g405500.t1.2 MLDP Major LD protein 17 4.425 14 gi|41179003|ref|NP_958358.1| \N Cytochrome f (chloroplast encoded) 16.5 2.316 15 Cre09.g398700.t1.1 CFA2 Cyclopropane-fatty acyl-phospholipid synthase 16 1.365 16 Cre09.g386650.t1.1 ANT1 ADP/ATP carrier protein, mitochondrial 14.5 2.047 17 Cre16.g672650.t1.1 \N Mitochondrial substrate carrier protein, possible 14 2.618 18 2-oxoglutarate/malate carrier Cre02.g081050.t1.1 FAP24 Flagella-associated protein 14 0.965 19 Cre29.g778950.t1.1 FMG1-B Flagellar membrane glycoprotein, major form 14 0.104 20 Cre03.g171050.t1.2 GHL1 Glycosyl hydrolase 13 0.943 21 Cre02.g130650.t1.1 \N \N 13 0.583 22 Cre01.g002500.t1.1 COP2 Chlamyopsin, light-gated proton channel rhodopsin 12.5 2.709 23 Cre17.g738050.t1.1 AGG4 Flagellar membrane protein, paralog of AGG2 12 2.590 24 Cre17.g734300.t1.2 \N \N 12 0.411 25 c c Cre15.g638500.t1.1 CYC1 Ubiquinol:cytochrome oxidoreductase cytochrome 1 11.5 1.561 26 Cre03.g172300.t1.1 \N Mitochondrial phosphate carrier protein 10.5 0.901 27 Cre01.g025350.t1.1 FAP235 Flagella-associated protein 10 2.777 28 Cre03.g196900.t1.1 \N \N 10 1.813 29 Cre23.g766250.t1.1 LHCBM1 Chlorophyll a/b-binding protein of LHCII 10 1.396 30

TAG synthesis under SL conditions. Interestingly, ho- vitamin E synthesis and recycling (Martinis et al., 2013). mologs of the two other proteins (PLP8 and VTE3) in Vitamin E synthesis is essential for photoadaptation to the model plant Arabidopsis have been found to be survive in higher irradiance. The presence of VTE3 in associated with plastoglobules, a type of LD speciﬁcto isolated SL-induced LDs supports the view that LDs plant plastids (Kessler and Vidi, 2007). PLP8 is known play a photoprotective role via synthesizing oil-soluble to play a structural role, whereas VTE3 is required for vitamins.

Table II. Short list of key proteins consistently found in all three previously published proteomic studies of LDs isolated from C. reinhardtii These proteins have demonstrated roles related to LD biology, and the cited previous proteomics studies are Moellering and Benning (2010), Nguyen et al. (2011), and Tsai et al. (2015). MM-N, Minimal medium under N starvation. Score (Mascot score) and SSC are averages from two replicates. LDs (MM-N) LDs (SL) Symbol Name Peptides SSC Score Rank Peptides SSC Score Rank BTA1 DGTS synthesis protein 41 94 2,120 1 24 41 1,198 2 MLDP Major LD protein 19 75 1,467 2 13 17 982 14 a/b-Hydrolase a/b-Hydrolase 28 63 1,832 3 13 21 970 7 LCS2 Long-chain acyl-CoA synthetase 32 54 1,646 4 7 7 323 59 GHL1 Glycosyl hydrolase 23 46 1,268 5 11 13 597 21 CFA2 Cyclopropane-fatty acyl-phospholipid synthase 19 34 1,147 6 10 16 597 16 LPLAT Lysophospholipid acyltransferase 20 33 966 7 4 4 213 129 TUB1 b-Tubulin1 15 16 673 23 6 6 272 85 TGD2 Permease-like component of an ATP-binding cassette transporter 3 3 152 252 5 5 270 98

Plant Physiol. Vol. 171, 2016 2413 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. Goold et al.

Figure 8. Model of LD biogenesis in cells exposed to SL or N starvation. Three types of LDs were drawn, partly based on observations made in this study and partly based on those of Goodson et al. (2011).

We conclude that a significant subpopulation of SL- and cyto-LDs. The difference in protein relative abun- induced LDs is associated with the chloroplast (cpst- dance (i.e. as shown by rankings; Table II) could be be- LDs and b-cyto-LDs), consistent with our microscopic cause these proteins are indeed present in much lower observations. Under SL exposure, wild-type strains of concentrations in all subpopulations of LDs induced by C. reinhardtii would mostly form cpst-LDs. The forma- SL than in those triggered by N starvation; alternatively, tion of cpst-LDs was observed previously in the it could be that these proteins were associated only with starchless mutant bafJ5 during N starvation (Goodson a specific subpopulation of LDs in the SL-exposed cells. et al., 2011). Therefore, cpst-LDs are formed either in This and other possibilities can be answered only once the absence of the starch (bafJ5 mutant) or under a high the subcellular localization information is available for fluence rate, extra photosynthetic products being these proteins under different culture conditions. diverted to the formation of LDs inside the plastid. Both situations create an imbalance in the supply and demand of reduced carbon (discussed in the next section) The Imbalance between Photosynthesis and Growth and favor the formation of plastidial LDs, a conve- Governs Reserve Accumulation nient and local solution. However, in the absence of the protein/lipid composition data for cpst-LDs of the Previous studies have shown that oil accumulation starchless mutant, we cannot assert that the two cpst- is triggered by conditions that impair cell growth LDs are of the same nature, and this question should be (Merchant et al., 2012; Liu and Benning, 2013). In- investigated in the future. creasing light intensity also has been reported as a One should bear in mind that LD isolates in each means to increase oil content when combined with N condition most likely represent a mixture of cpst-LDs starvation (Klok et al., 2013; Kandilian et al., 2014). The

Figure 9. Carbon reserve formation in C. reinhardtii in response to energy imbalance: a model. During normal growth (LL, nutrient-replete conditions), C. reinhardtii cells convert most of the solar energy they have captured through the photosystems and ﬁxed as organic compounds to support biomass growth (cell division); only a small fraction is used to make temporary carbon reserves (starch and oil). Creating an energy imbalance either through an increased input of energy (SL with respect to biomass growth) or a decrease in energy use (through blocking growth via N starvation) results in reallocation of the excess energy to the synthesis of reserve compounds (oil and starch).

2414 Plant Physiol. Vol. 171, 2016 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. High Light Induces Reserve Accumulation in C. reinhardtii

latter two studies have shown that, in conditions of N identical conditions with additional supply of 1.8% CO2 in air (Cagnon et al., starvation, high light serves to increase the energy 2013). Cell numbers were counted using a Beckman-Coulter multisizer. Cellular concentration was determined as the number of cells per milliliter of culture or imbalance, resulting in higher TAG storage in the total cellular volume (mm3) per milliliter of culture. heterokont N. oculata and the green microalga Neo- chloris oleoabundans. Our study here provides a detailed report on the effect of light on reserve accumulation Cultivation in PBRs under optimal growth conditions for the model PBR cultures were cultivated in minimal medium in automated Biostat A microalga C. reinhardtii. We found that SL-exposed cells Plus PBR systems (Sartorius Stedium) as described previously (Dang et al., accommodate an excess of light energy by storing re- 2014). Cells were maintained at a constant OD880 = 0.4 throughout experi- mentation. Specific growth rates were determined from the measurement of duced carbon in the forms of starch and neutral lipids. fresh medium added to the turbidostat to maintain a constant biomass con- The impact of high light on storage reserve accu- centration. All PBRs were stirred at a constant 300 rpm. mulation is sometimes considered as a consequence of a light-stress effect. We observed a burst in oil/starch Quantification of Starch content following SL exposure. This suggests that cells respond to SL in two phases: an initial strong response, Starch was isolated and quantified based on the protocol described by Klein followed by an acclimated phase. This initial burst is and Betz (1978). C. reinhardtii cells were harvested by centrifugation at 1,000g for 3 min. Cellular pellets were resuspended in 1 mL of methanol and centri- most likely due to a combined stress response, with an fuged again for 3 min at 1,000g to remove all chlorophyll. The methanol-washed increased carbon flux toward reserve formation at the pellet was air dried under a fume hood and subsequently resuspended in beginning, followed by a metabolic adaptation to the 400 mL of distilled water; samples were then autoclaved for 20 min at 121°C. increased irradiance. As cultivation under SL pro- Amyloglucosidase (0.2 units; Roche) was added, and samples were incubated for 1 h at 60°C. Samples were then centrifuged to pellet particulate matter. Glc ceeded, acclimated cells synthesized higher amounts of was measured with a YSI2700 select sugar analyzer (YSI Life Sciences) using carbon reserves simultaneous to an increased growth commercial Glc as a standard. rate. Thus, SL combined with a nonlimiting CO supply fi 2 induces signi cant oil accumulation in C. reinhardtii in Isolation of LDs conditions that support maximal growth. Therefore, we conclude that oil accumulation is not necessarily linked In order to isolate low-abundance LDs, a previously established C. reinhardtii to a stress effect inducing growth impairment but may LD isolation protocol (Nguyen et al., 2011) was used with the following minor changes. A desktop ultracentrifuge replaced the standard ultracentrifuge, and occur in conditions of maximal growth. smaller 5-mL ultracentrifuge tubes were used to decrease the surface area of the Upon steady-state SL exposure, the photosynthetic lipid pad and increase the ease with which the lipid pad was completely re- capacity may exceed the cell division and growth ca- covered. Additionally, to avoid the loss of material, the hexane wash was pacity, thus creating an imbalance between the forma- omitted, and during the final wash, a Suc-free buffer was used to replace the 0.4 M Suc buffer. In the final step, the Suc-free buffer floats above all other tion of photosynthetic products and the capacity for buffers and allows for easier isolation of the final washed LDs. their direct consumption. The cellular metabolism would thus be readjusted to store the extra energy as fi energy-dense materials (i.e. starch and oil). These stor- Lipid Extraction, Quanti cation of Lipid Classes by TLC, age compounds provide a temporary reservoir to ac- and Lipid Molecular Species Analysis Using LC-MS/MS fl commodate the extra uxes of ATP and reducing A modified method of Bligh and Dyer (1959) was used to extract lipids. power, thereby ensuring cell energetic homeostasis. Samples were initially dissolved by vortexing in 1 mL of quenching solution Based on these data, we propose a more general model (1 mM EDTA and 0.15 M acetic acid) in an 8-mL glass tube with a Teflon screw in which starch and oil accumulations are triggered cap. Three milliliters of methanol:chloroform (2:1, v/v) was added, and if fi samples were to be used for LC-MS/MS at this stage, internal standards when an imbalance between photosynthetic CO2 xa- TAG51:0 (composed of 17:0/17:0/17:0 fatty acids) and PtdEtn34:0 (composed tion and growth occurs (Fig. 9). Such an imbalance may of 17:0/17:0 fatty acids) were added. Samples were then vortexed for 10 min, occur in conditions where growth is primarily affected 1 mL of chloroform and 0.8 mL of KCl (0.8%, w/v) were added, and samples (such as nutrient starvation, heat, salinity, or chemical were vortexed again for 10 min. Phases were separated by centrifugation at treatment) but also may occur in conditions of SL where 1,000g for 2 min at 4°C. The lower phase was then isolated and transferred to a fi new clean glass tube. To the remaining phase, 1 mL of hexane was added to photosynthetic CO2 xation exceeds the ability of algae extract again the remaining lipids. Phases were separated by centrifugation at to grow and replicate. The large increase in neutral lipid 1,000g for 2 min at 4°C. The upper phase containing lipids was then transferred and starch productivity achieved under SL clearly to the chloroform extract using a glass Pasteur pipette. The combined chloro- fl points out the biotechnological advantage of increasing form and hexane extracts were then dried under a ow of N and resuspended in a mixture of chloroform:methanol (2:1, v/v) for subsequent analyses. reserve accumulation under conditions not compro- To analyze TAG content in cells and major polar lipid species, total lipids mising growth in green microalgae. extracted by the above protocol were processed using an automated high- performance TLC platform (CAMAG). The detailed TLC method has been described (Siaut et al., 2011). Lipid molecular species were analyzed by LC- MATERIALS AND METHODS MS/MS, and conditions were detailed by Légeret et al. (2016). C. reinhardtii Strains and Precultures Generation of Fatty Acid Methyl Esters and Quantification The wild-type Chlamydomonas reinhardtii strain CC-124 (mt2 nit1 nit2)was by Gas Chromatography-Flame Ionization Detection used throughout this study. Cells were usually maintained in Tris-acetate- phosphate medium (Harris, 2001) in Erlenmeyer flasks at 25°C with shaking The conversion of total lipids to their fatty acid methyl esters was carried out 2 2 at 120 rpm under constant illumination (150 mmol photons m 2 s 1). For pho- as described previously (Siaut et al., 2011). A fraction of extract was mixed in an toautotrophic growth, cultures grown in minimal medium were kept under 8-mL glass tube with 1 mL of methanol containing 5% (v/v) H2SO4, to which

Plant Physiol. Vol. 171, 2016 2415 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. Goold et al.

10 mL of butylated hydroxytoluene (1%, v/v) and 5 to 25 mg of TAG Austin JR II, Frost E, Vidi PA, Kessler F, Staehelin LA (2006) Plastoglobules (17:0/17:0/17:0) standard were added. Each sample was vortexed for 30 s and are lipoprotein subcompartments of the chloroplast that are permanently heated to 85°C for 90 min. After cooling to room temperature, 1.5 mL of 0.9% coupled to thylakoid membranes and contain biosynthetic enzymes. Plant (w/v) NaCl was added together with 1 mL of hexane. Samples were vortexed Cell 18: 1693–1703 for 10 min and centrifuged at 1,000g for 2 min at 4°C. The upper (organic) phase Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and was then isolated with a clean glass Pasteur pipette and evaporated to dryness purification. Can J Biochem Physiol 37: 911–917 under a stream of N. Samples were finally resuspended in hexane. Fatty acid Bonente G, Pippa S, Castellano S, Bassi R, Ballottari M (2012) Acclimation methyl ester species were separated by the use of a TRACE GC Ultra gas of Chlamydomonas reinhardtii to different growth irradiances. J Biol Chem chromatograph (Thermo Fisher Scientific) using a polar TR-WAX column 287: 5833–5847 (Thermo Fisher Scientific; 30 m 3 0.25 mm). Detailed gas chromatography Cagnon C, Mirabella B, Nguyen HM, Beyly-Adriano A, Bouvet S, Cuiné conditions have been described (Nguyen et al., 2013). S, Beisson F, Peltier G, Li-Beisson Y (2013) Development of a forward genetic screen to isolate oil mutants in the green microalga Chlamydo- monas reinhardtii. Biotechnol Biofuels 6: 178 Protein Extraction, SDS-PAGE, and Proteomic Analysis Chapman KD, Dyer JM, Mullen RT (2012) Biogenesis and functions of lipid droplets in plants. J Lipid Res 53: 215–226 Proteins were extracted from isolated LDs by the addition of cold acetone Chen H, Hu J, Qiao Y, Chen W, Rong J, Zhang Y, He C, Wang Q (2015) Ca2+- (80%, v/v) and SDS (0.3%, v/v). Extracted proteins were separated by SDS- regulated cyclic electron flow supplies ATP for nitrogen starvation-induced PAGE following standard protocols (Sambrook and Russell, 2001). Proteins lipid biosynthesis in green alga. Sci Rep 5: 15117 were then stained with a silver nitrate solution for 30 min. For proteomics Dang KV, Plet J, Tolleter D, Jokel M, Cuiné S, Carrier P, Auroy P, analysis, samples in biological duplicates were concentrated at the top of a RichaudP,JohnsonX,AlricJ,etal(2014) Combined increases in mi- NuPage precasted gel (Invitrogen; stacking migration). Bands were cut and tochondrial cooperation and oxygen photoreduction compensate for submitted to LC-MS/MS analysis following the same procedure as reported deficiency in cyclic electron flow in Chlamydomonas reinhardtii. Plant Cell previously (Nguyen et al., 2011) except for the following: the LC-MS/MS ex- 26: 3036–3050 periment was performed on an LTQ-Orbitrap Velos Pro (Thermo Fisher Sci- Dupierris V, Masselon C, Court M, Kieffer-Jaquinod S, Bruley C (2009) A entific) device coupled to an Ultimate 3000 LC system. The C. reinhardtii protein toolbox for validation of mass spectrometry peptides identification and database used for Mascot (version 2.4.4; Matrix Science) searches was consti- generation of database: IRMa. Bioinformatics 25: 1980–1981 tuted of the models downloaded from Phytozome (release 169_v4.3; http:// Erickson E, Wakao S, Niyogi KK (2015) Light stress and photoprotection in phytozome.jgi.doe.gov/) plus mitochondrion and plastid-encoded proteins Chlamydomonas reinhardtii.PlantJ82: 449–465 from the National Center for Biotechnology Information (a total of 17,191 en- FanJ,AndreC,XuC(2011) A chloroplast pathway for the de novo bio- tries). Acetyl (protein N-a-acetylation) and Met oxidation were set as varia- synthesis of triacylglycerol in Chlamydomonas reinhardtii.FEBSLett585: ble modifications and carbamidomethyl Cys as a fixed modification. Two 1985–1991 miscleavages were allowed. Mascot search results were automatically filtered Fan J, Yan C, Andre C, Shanklin J, Schwender J, Xu C (2012) Oil accu- as described previously (Nguyen et al., 2011) with the IRMa 1.31.1 version mulation is controlled by carbon precursor supply for fatty acid syn- software (Dupierris et al., 2009). thesis in Chlamydomonas reinhardtii. Plant Cell Physiol 53: 1380–1390 Farese RV Jr, Walther TC (2009) Lipid droplets finally get a little R-E-S-P- – Microscopy E-C-T. Cell 139: 855 860 Fischer BB, Wiesendanger M, Eggen RIL (2006) Growth condition-dependent Samples were stained first with Nile Red (Sigma) at a final concentration sensitivity, photodamage and stress response of Chlamydomonas reinhardtii 21 21 – of 1 mgmL (from a stock solution of 1 mg mL in methanol) for 10 min in exposed to high light conditions. Plant Cell Physiol 47: 1135 1145 the dark and then imaged with a confocal microscope (TCS SP2) using a 633 Förster B, Mathesius U, Pogson BJ (2006) Comparative proteomics of high oil-immersion objective (Leica). The Nile Red signal was captured using a laser light stress in the model alga Chlamydomonas reinhardtii.Proteomics6: – excitation line at 488 nm, and emission was collected between 554 and 599 nm. 4309 4320 – Chlorophyll autofluorescence was captured between 650 and 714 nm. Goodman JM (2008) The gregarious lipid droplet. J Biol Chem 283: 28005 Pseudocolors for images were obtained using ZEN software (Carl Zeiss). 28009 Goodson C, Roth R, Wang ZT, Goodenough U (2011) Structural correlates Supplemental Data of cytoplasmic and chloroplast lipid body synthesis in Chlamydomonas reinhardtii and stimulation of lipid body production with acetate boost. – The following supplemental materials are available. Eukaryot Cell 10: 1592 1606 Goold H, Beisson F, Peltier G, Li-Beisson Y (2015) Microalgal lipid Supplemental Figure S1. Changes in total fatty acid, C/N ratio, propor- droplets: composition, diversity, biogenesis and functions. Plant Cell tion of lipid classes, and fatty acid composition during normal growth in Rep 34: 545–555 LL and under SL. Harris EH (2001) Chlamydomonas as a model organism. Annu Rev Plant Supplemental Figure S2. Polar membrane lipid molecular species pre- Physiol Plant Mol Biol 52: 363–406 sent in LDs isolated from SL-exposed cells compared with those from He Q, Yang H, Wu L, Hu C (2015) Effect of light intensity on physiological N-starved cells. changes, carbon allocation and neutral lipid accumulation in oleaginous microalgae. Bioresour Technol 191: 219–228 Supplemental Table S1. Proteomic analyses of LDs isolated from either Hu Q, Sommerfeld M, Jarvis E, Ghirardi M, Posewitz M, Seibert M, SL-exposed cells or cells starved of N for 24 h. Darzins A (2008) Microalgal triacylglycerols as feedstocks for biofuel – Supplemental Materials and Methods. production: perspectives and advances. Plant J 54: 621 639 Huang AHC (1992) Oil bodies and oleosins in seeds. Annu Rev Plant Physiol Plant Mol Biol 43: 177–200 ACKNOWLEDGMENTS Huang AHC (1996) Oleosins and oil bodies in seeds and other organs. Plant Physiol 110: 1055–1061 We thank Patrick Carrier for assistance in running and maintaining PBRs. Jolivet P, Acevedo F, Boulard C, d’Andréa S, Faure JD, Kohli A, Nesi N, Received May 9, 2016; accepted June 10, 2016; published June 13, 2016. Valot B, Chardot T (2013) Crop seed oil bodies: from challenges in protein identification to an emerging picture of the oil body proteome. Proteomics 13: 1836–1849 LITERATURE CITED Kandilian R, Pruvost J, Legrand J, Pilon L (2014) Influence of light ab- sorption rate by Nannochloropsis oculata on triglyceride production Abida H, Dolch LJ, Meï C, Villanova V, Conte M, Block MA, Finazzi G, during nitrogen starvation. Bioresour Technol 163: 308–319 Bastien O, Tirichine L, Bowler C, et al (2015) Membrane glycerolipid Kato N, Dong T, Bailey M, Lum T, Ingram D (2013) Triacylglycerol mo- remodeling triggered by nitrogen and phosphorus starvation in Phaeo- bilization is suppressed by brefeldin A in Chlamydomonas reinhardtii. dactylum tricornutum. Plant Physiol 167: 118–136 Plant Cell Physiol 54: 1585–1599

2416 Plant Physiol. Vol. 171, 2016 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. High Light Induces Reserve Accumulation in C. reinhardtii

Kessler F, Vidi PA (2007) Plastoglobule lipid bodies: their functions in Chlamydomonas reinhardtii has a single v-3 fatty acid desaturase that chloroplasts and their potential for applications. In A Fiechter, C Sautter, localizes to the chloroplast and impacts both plastidic and extraplastidic eds, Green Gene Technology: Research in an Area of Social Conflict. membrane lipids. Plant Physiol 163: 914–928 Springer, Berlin, pp 153–172 Pal D, Khozin-Goldberg I, Cohen Z, Boussiba S (2011) The effect of light, Khotimchenko SV, Yakovleva IM (2005) Lipid composition of the red alga salinity, and nitrogen availability on lipid production by Nannochloropsis Tichocarpus crinitus exposed to different levels of photon irradiance. sp. Appl Microbiol Biotechnol 90: 1429–1441 Phytochemistry 66: 73–79 Peers G, Truong TB, Ostendorf E, Busch A, Elrad D, Grossman AR, Khozin-Goldberg I, Cohen Z (2011) Unraveling algal lipid metabolism: Hippler M, Niyogi KK (2009) An ancient light-harvesting protein is recent advances in gene identification. Biochimie 93: 91–100 critical for the regulation of algal photosynthesis. Nature 462: 518–521 KimH,JangS,KimS,YamaokaY,HongD,SongWY,NishidaI,Li- Rosenberg JN, Oyler GA, Wilkinson L, Betenbaugh MJ (2008) A green Beisson Y, Lee Y (2015) The small molecule fenpropimorph rapidly light for engineered algae: redirecting metabolism to fuel a biotechnol- converts chloroplast membrane lipids to triacylglycerols in Chlamydo- ogy revolution. Curr Opin Biotechnol 19: 430–436 monas reinhardtii.FrontMicrobiol6: 54 Sambrook J, Russell DW (2001) Molecular Cloning: A Laboratory Manual. Kim S, Kim H, Ko D, Yamaoka Y, Otsuru M, Kawai-Yamada M, Ishikawa Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY T, Oh HM, Nishida I, Li-Beisson Y, et al (2013) Rapid induction of lipid Sato R, Ito H, Tanaka A (2015) Chlorophyll b degradation by chlorophyll b droplets in Chlamydomonas reinhardtii and Chlorella vulgaris by brefeldin reductase under high-light conditions. Photosynth Res 126: 249–259 A. PLoS ONE 8: e81978 Scott SA, Davey MP, Dennis JS, Horst I, Howe CJ, Lea-Smith DJ, Smith Klein U, Betz A (1978) Fermentative metabolism of hydrogen-evolving AG (2010) Biodiesel from algae: challenges and prospects. Curr Opin Chlamydomonas moewusii.PlantPhysiol61: 953–956 Biotechnol 21: 277–286 Klok AJ, Martens DE, Wijffels RH, Lamers PP (2013) Simultaneous Sheehan J, Dunahay T, Benemann J, Roessler PG (1998) A Look Back at growth and neutral lipid accumulation in microalgae. Bioresour Technol the U.S. Department of Energy’s Aquatic Species Program: Biodiesel 134: 233–243 from Algae. National Renewable Energy Laboratory, Golden, CO Légeret B, Schulz-Raffelt M, Nguyen HM, Auroy P, Beisson F, Peltier G, Siaut M, Cuiné S, Cagnon C, Fessler B, Nguyen M, Carrier P, Beyly A, BlancG,Li-BeissonY(2016) Lipidomic and transcriptomic analyses of Beisson F, Triantaphylidès C, Li-Beisson Y, et al (2011) Oil accumu- Chlamydomonas reinhardtii under heat stress unveil a direct route for the lation in the model green alga Chlamydomonas reinhardtii: characteriza- conversion of membrane lipids into storage lipids. Plant Cell Environ 39: tion, variability between common laboratory strains and relationship 834–847 with starch reserves. BMC Biotechnol 11: 7 Li Y, Beisson F, Pollard M, Ohlrogge J (2006) Oil content of Arabidopsis Tardif M, Atteia A, Specht M, Cogne G, Rolland N, Brugière S, Hippler seeds: the influence of seed anatomy, light and plant-to-plant variation. M, Ferro M, Bruley C, Peltier G, et al (2012) PredAlgo: a new subcel- Phytochemistry 67: 904–915 lular localization prediction tool dedicated to green algae. Mol Biol Evol Li-Beisson Y, Beisson F, Riekhof W (2015) Metabolism of acyl-lipids in 29: 3625–3639 Chlamydomonas reinhardtii.PlantJ82: 504–522 Tolleter D, Ghysels B, Alric J, Petroutsos D, Tolstygina I, Krawietz D, Liu B, Benning C (2013) Lipid metabolism in microalgae distinguishes it- Happe T, Auroy P, Adriano JM, Beyly A, et al (2011) Control of hy- self. Curr Opin Biotechnol 24: 300–309 drogen photoproduction by the proton gradient generated by cyclic Ma X, Liu J, Liu B, Chen T, Yang B, Chen F (2016) Physiological and bio- electron flow in Chlamydomonas reinhardtii. Plant Cell 23: 2619–2630 chemical changes reveal stress-associated photosynthetic carbon partitioning Tsai CH, Warakanont J, Takeuchi T, Sears BB, Moellering ER, Benning C into triacylglycerol in the oleaginous marine alga Nannochloropsis oculata. (2014) The protein Compromised Hydrolysis of Triacylglycerols 7 (CHT7) Algal Research 16: 28–35 acts as a repressor of cellular quiescence in Chlamydomonas.ProcNatl Martinis J, Glauser G, Valimareanu S, Kessler F (2013) A chloroplast Acad Sci USA 111: 15833–15838 ABC1-like kinase regulates vitamin E metabolism in Arabidopsis. Plant Tsai CH, Zienkiewicz K, Amstutz CL, Brink BG, Warakanont J, Roston R, Physiol 162: 652–662 Benning C (2015) Dynamics of protein and polar lipid recruitment Merchant SS, Kropat J, Liu B, Shaw J, Warakanont J (2012) TAG, you’re it! during lipid droplet assembly in Chlamydomonas reinhardtii.PlantJ83: Chlamydomonas as a reference organism for understanding algal tri- 650–660 acylglycerol accumulation. Curr Opin Biotechnol 23: 352–363 Tzen JTC, Huang AHC (1992) Surface structure and properties of plant Mettler T, Mühlhaus T, Hemme D, Schöttler MA, Rupprecht J, Idoine A, seed oil bodies. J Cell Biol 117: 327–335 Veyel D, Pal SK, Yaneva-Roder L, Winck FV, et al (2014) Systems Urzica EI, Vieler A, Hong-Hermesdorf A, Page MD, Casero D, Gallaher analysis of the response of photosynthesis, metabolism, and growth SD, Kropat J, Pellegrini M, Benning C, Merchant SS (2013) Remod- to an increase in irradiance in the photosynthetic model organism eling of membrane lipids in iron-starved Chlamydomonas.JBiolChem Chlamydomonas reinhardtii. Plant Cell 26: 2310–2350 288: 30246–30258 Moellering ER, Benning C (2010) RNA interference silencing of a major Wang ZT, Ullrich N, Joo S, Waffenschmidt S, Goodenough U (2009) Algal lipid droplet protein affects lipid droplet size in Chlamydomonas rein- lipid bodies: stress induction, purification, and biochemical characteri- hardtii. Eukaryot Cell 9: 97–106 zation in wild-type and starchless Chlamydomonas reinhardtii.Eukaryot Nguyen HM, Baudet M, Cuiné S, Adriano JM, Barthe D, Billon E, Bruley Cell 8: 1856–1868 C, Beisson F, Peltier G, Ferro M, et al (2011) Proteomic profiling of oil Wijffels RH, Barbosa MJ (2010) An outlook on microalgal biofuels. Science bodies isolated from the unicellular green microalga Chlamydomonas 329: 796–799 reinhardtii: with focus on proteins involved in lipid metabolism. Pro- Zhekisheva M, Boussiba S, Khozin-Goldberg I, Zarka A, Cohen Z (2002) teomics 11: 4266–4273 Accumulation of oleic acid in Haematococcus pluvialis (Chlorophyceae) under Nguyen HM, Cuiné S, Beyly-Adriano A, Légeret B, Billon E, Auroy P, nitrogen starvation or high light is correlated with that of astaxanthin esters. Beisson F, Peltier G, Li-Beisson Y (2013) The green microalga JPhycol38: 325–331

Plant Physiol. Vol. 171, 2016 2417 Downloaded from www.plantphysiol.org on September 13, 2016 - Published by www.plantphysiol.org Copyright © 2016 American Society of Plant Biologists. All rights reserved. CONCLUSIONS AND PERSPECTIVES The original objective of this PhD Thesis was to dissect the source(s) of acetyl-CoA, ATP and NAD(P)H for TAG biosynthesis, and also explore the interaction or contributions of various metabolic pathways (i.e. that of BCAA, or the β-oxidation of fatty acids, or high light) with/to lipid biosynthesis in the green model microalga Chlamydomonas reinhardtii. This work took advantage of the large number of oil mutants previously isolated from the laboratory (Cagnon et al., 2013). In the past 3- years, and in collaboration with many members of our team, and especially with Dr. Fantao Kong (Postdoc at the time in our group), we brought new insights into understanding of FA and lipid biosynthesis in green microalgae, notably:

- We provided for the first time evidence that the β-oxidation of FAs is a peroxisomal process in the model green microalga Chlamydomonas reinhardtii. - We demonstrated that shutting down FA β-oxidation can be used to boost cellular TAG content, and under mixotrophic condition, the increased TAG content is mostly a result of acyl-remodelling. - We also provided evidence that FA β-oxidation plays a role in FA turnover during a day/night cycle and under strict carbon starvation. - We revealed for the first time a redox communication from peroxisome to chloroplast mediated by malate or peroxisomal H2O2. We show that this communication is essential to maintain a lower chloroplast redox state during HL or nutrient deficiency. - Furthermore, we provided evidence that BCAA catabolism intersects with lipid biosynthesis as well as lipid turnover in Chlamydomonas. This interaction occurs at two different levels, i.e. contribution of acetyl-CoA as well as ATP through mitochondrial respiration. In addition, this is a first study of an enzyme related to the BCAA catabbolic pathway in green microalgae.

Taken together, this work highlighted the complex interplay between carbon and energy metabolism in green photosynthetic cells, and revealed the importance of inter-

101 organelle communications on TAG metabolism, i.e. the importance of the interaction between the three most energetic subcellular organelle, i.e. peroxisome, chloroplast and mitochondria.

Although considerable efforts have been put to study the molecular mechanisms of TAG accumulation in algal cells, majority of the current funded projects focused on – omics study of changes in gene, protein, lipid and metabolite levels accompanying TAG synthesis, or focused on engineering specific step(s) of the predicted pathways, the work presented in this Thesis highlighted the importance of redox balance on TAG metabolism. Ongoing research and furture perspectives are to continue our investigation on the articulation between photosynthetic carbon fixation and their chanelling to carbon researve formation, especially TAGs. In particular, we will focus on the study of energy trafficking between three energetic organelles i.e. peroxisome, chloroplast and mitochondria. Specifically,

- We aim to disect the possible interaction between peroxisome and mitochondria in Chlamydomonas, and investigate the importance of this interaction on reserve deposition. - We plan to provide metabolic and proteomic maps for peroxisomes of Chlamydomonas. - We will also like to develop redox sensors for different subcellular compartments.

102

REFERENCES

Akao T, Kusaka T (1976) Solubilization of diglyceride acyltransferase from the membrane of Mycobacterium smegmatis. J Biochem 80: 723-728

Angelovici R, Lipka AE, Deason N, Gonzalez-Jorge S, Lin H, Cepela J, Buell R, Gore MA, Dellapenna D (2013) Genome-wide analysis of branched-chain amino acid levels in Arabidopsis seeds. Plant Cell 25: 4827-4843

Antolin G, Tinaut FV, Briceno Y, Castano V, Perez C, Ramirez AI (2002) Optimisation of biodiesel production by sunflower oil transesterification. Bioresour Technol 83: 111-114

Araujo WL, Ishizaki K, Nunes-Nesi A, Larson TR, Tohge T, Krahnert I, Witt S, Obata T, Schauer N, Graham IA, Leaver CJ, Fernie AR (2010) Identification of the 2-hydroxyglutarate and isovaleryl-CoA dehydrogenases as alternative electron donors linking lysine catabolism to the electron transport chain of Arabidopsis mitochondria. Plant Cell 22: 1549-1563

Athenstaedt K, Daum G (2006) The life cycle of neutral lipids: synthesis, storage and degradation. Cellular and Molecular Life Sciences 63: 1355-1369

Avidan O, Brandis A, Rogachev I, Pick U (2015) Enhanced acetyl-CoA production is associated with increased triglyceride accumulation in the green alga Chlorella desiccata. Journal of Experimental Botany 66: 3725-3735

Balamurugan S, Wang X, Wang HL, An CJ, Li H, Li DW, Yang WD, Liu JS, Li HY (2017) Occurrence of plastidial triacylglycerol synthesis and the potential regulatory role of AGPAT in the model diatom Phaeodactylum tricornutum. Biotechnol Biofuels 10: 97

Ball SG (2002) The intricate pathway of starch biosynthesis and degradation in the monocellular alga Chlamydomonas reinhardtii. Australian Journal of Chemistry 55: 49-59

Benning C (2008) A role for lipid trafficking in chloroplast biogenesis. Progress in Lipid Research 47: 381-389

Binder S (2010) Branched-chain amino acid metabolism in Arabidopsis thaliana. The Arabidopsis Book: e0137

Binder S, Knill T, Schuster J (2007) Branched-chain amino acid metabolism in higher plants. Physiologia Plantarum 129: 68-78

Boudiere L, Michaud M, Petroutsos D, Rebeille F, Falconet D, Bastien O, Roy S, 103

Finazzi G, Rolland N, Jouhet J, Block MA, Marechal E (2014) Glycerolipids in photosynthesis: composition, synthesis and trafficking. Biochim Biophys Acta 1837: 470-480

Bourgis F, Kilaru A, Cao X, Ngando-Ebongue GF, Drira N, Ohlrogge JB, Arondel V (2011) Comparative transcriptome and metabolite analysis of oil palm and date palm mesocarp that differ dramatically in carbon partitioning. Proceedings of the National Academy of Sciences of the United States of America 108: 12527-12532

Boyle NR, Page MD, Liu B, Blaby IK, Casero D, Kropat J, Cokus SJ, Hong- Hermesdorf A, Shaw J, Karpowicz SJ, Gallaher SD, Johnson S, Benning C, Pellegrini M, Grossman A, Merchant SS (2012) Three acyltransferases and nitrogen-responsive regulator are implicated in nitrogen starvation-induced triacylglycerol accumulation in Chlamydomonas. Journal of Biological Chemistry 287: 15811-15825

Boynton J, Gillham N, Harris E, Hosler J, Johnson A, Jones A, Randolph- Anderson B, Robertson D, Klein T, Shark K, et a (1988) Chloroplast transformation in Chlamydomonas with high velocity microprojectiles. Science 240: 1534-1538

Brehelin C, Kessler F (2008) The plastoglobule: A bag full of lipid biochemistry Tricks. Photochemistry and Photobiology 84: 1388-1394

Bridgwater AV, Meier D, Radlein D (1999) An overview of fast pyrolysis of biomass. Organic Geochemistry 30: 1479-1493

Cagnon C, Mirabella B, Nguyen HM, Beyly-Adriano A, Bouvet S, Cuine S, Beisson F, Peltier G, Li-Beisson Y (2013) Development of a forward genetic screen to isolate oil mutants in the green microalga Chlamydomonas reinhardtii. Biotechnology for Biofuels 6: 178

Chen F, Johns MR (1991) Effect of C/N ratio and aeration on the fatty acid composition of heterotrophic Chlorella sorokiniana. Journal of Applied Phycology 3: 203-209

Chen JE, Smith AG (2012) A look at diacylglycerol acyltransferases (DGATs) in algae. J Biotechnol 162: 28-39

Chochois V, Dauvillee D, Beyly A, Tolleter D, Cuine S, Timpano H, Ball S, Cournac L, Peltier G (2009) Hydrogen production in Chlamydomonas: photosystem II- dependent and -independent pathways differ in their requirement for starch metabolism. Plant Physiology 151: 631-640

Couso I, Perez-Perez ME, Martinez-Force E, Kim HS, He Y, Umen JG, Crespo JL

104

(2018) Autophagic flux is required for the synthesis of triacylglycerols and ribosomal protein turnover in Chlamydomonas. J Exp Bot 69: 1355-1367

Crown SB, Marze N, Antoniewicz MR (2016) Catabolism of branched chain amino acids contributes significantly to synthesis of odd-chain and even-chain fatty acids in 3T3-L1 adipocytes. PLOS ONE 10: e0145850

Cuadros-Inostroza Á, Caldana C, Redestig H, Kusano M, Lisec J, Peña-Cortés H, Willmitzer L, Hannah MA (2009) TargetSearch - a Bioconductor package for the efficient preprocessing of GC-MS metabolite profiling data. BMC Bioinformatics 10: 428

Czabany T, Athenstaedt K, Daum G (2007) Synthesis, storage and degradation of neutral lipids in yeast. Biochimica Et Biophysica Acta-Molecular and Cell Biology of Lipids 1771: 299-309

Czabany T, Wagner A, Zweytick D, Lohner K, Leitner E, Ingolic E, Daum G (2008) Structural and biochemical properties of lipid particles from the yeast Saccharomyces cerevisiae. Journal of Biological Chemistry 283: 17065-17074

Davey MP, Horst I, Duong G-H, Tomsett EV, Litvinenko ACP, Howe CJ, Smith AG (2014) Triacylglyceride production and autophagous responses in Chlamydomonas reinhardtii depend on resource allocation and carbon source. Eukaryotic Cell 13: 392-400 de Kraker JW, Gershenzon J (2011) From amino acid to glucosinolate biosynthesis: protein sequence changes in the evolution of methylthioalkylmalate synthase in Arabidopsis. Plant Cell 23: 38-53

Dennis EA (2009) Lipidomics joins the omics evolution. Proceedings of the National Academy of Sciences 106: 2089

Durrett TP, Benning C, Ohlrogge J (2008) Plant triacylglycerols as feedstocks for the production of biofuels. Plant Journal 54: 593-607

Eaton S, Bartlett K, Pourfarzam M (1996) Mammalian mitochondrial beta-oxidation. Biochem J 320 ( Pt 2): 345-357

Emanuelsson O, Nielsen H, Brunak S, von Heijne G (2000) Predicting subcellular localization of proteins based on their N-terminal amino acid sequence. Journal of Molecular Biology 300: 1005-1016

Fahy E, Subramaniam S, Brown HA, Glass CK, Merrill AH, Murphy RC, Raetz CRH, Russell DW, Seyama Y, Shaw W, Shimizu T, Spener F, van Meer G, VanNieuwenhze MS, White SH, Witztum JL, Dennis EA (2005) A comprehensive classification system for lipids. Journal of Lipid Research 46: 839-862 105

Fahy E, Subramaniam S, Murphy RC, Nishijima M, Raetz CRH, Shimizu T, Spener F, van Meer G, Wakelam MJO, Dennis EA (2009) Update of the LIPID MAPS comprehensive classification system for lipids. Journal of Lipid Research 50: S9-S14

Fan J, Yan C, Andre C, Shanklin J, Schwender J, Xu C (2012) Oil accumulation is controlled by carbon precursor supply for fatty acid synthesis in Chlamydomonas reinhardtii. Plant and Cell Physiology 53: 1380-1390

Fan JL, Andre C, Xu CC (2011) A chloroplast pathway for the de novo biosynthesis of triacylglycerol in Chlamydomonas reinhardtii. FEBS Letters 585: 1985-1991

Footitt S, Slocombe SP, Larner V, Kurup S, Wu YS, Larson T, Graham I, Baker A, Holdsworth M (2002) Control of germination and lipid mobilization by COMATOSE, the Arabidopsis homologue of human ALDP. EMBO Journal 21: 2912-2922

Fujiki Y, Sato T, Ito M, Watanabe A (2000) Isolation and characterization of cDNA clones for the E1β and E2 subunits of the branched-chain α-ketoacid dehydrogenase complex in Arabidopsis. Journal of Biological Chemistry 275: 6007-6013

Fulda M, Schnurr J, Abbadi A, Heinz E, Browse J (2004) Peroxisomal Acyl-CoA synthetase activity is essential for seedling development in Arabidopsis thaliana. Plant Cell 16: 394-405

Fuls J, Hawkins CS, Hugo FJC (1984) Tractor engine performance on sunflower oil fuel. Journal of Agricultural Engineering Research 30: 29-35

Galili G, Amir R, Fernie AR (2016) The regulation of essential amino acid synthesis and accumulation in plants. Annual Review of Plant Biology 67: 153-178

Ge F, Huang W, Chen Z, Zhang C, Xiong Q, Bowler C (2014) Methylcrotonyl-CoA carboxylase regulates triacylglycerol accumulation in the model diatom Phaeodactylum tricornutum. Plant Cell 26: 1681-1697

Geigenberger P, Kolbe A, Tiessen A (2005) Redox regulation of carbon storage and partitioning in response to light and sugars. Journal of Experimental Botany 56: 1469-1479

Giroud C, Gerber A, Eichenberger W (1988) Lipids of Chlamydomonas reinhardtii - analyses of molecular sepcies and intracellular site(s) of biosynthesis. Plant and Cell Physiology 29: 587-595

Gonda I, Bar E, Portnoy V, Lev S, Burger J, Schaffer AA, Tadmor Y, Gepstein S, Giovannoni JJ, Katzir N, Lewinsohn E (2010) Branched-chain and aromatic amino acid catabolism into aroma volatiles in Cucumis melo L. fruit. J Exp Bot 106

61: 1111-1123

Gonzalez-Ballester D, de Montaigu A, Galvan A, Fernandez E (2005) Restriction enzyme site-directed amplification PCR: a tool to identify regions flanking a marker DNA. Anal Biochem 340: 330-335

Gonzalez-Ballester D, Pootakham W, Mus F, Yang WQ, Catalanotti C, Magneschi L, de Montaigu A, Higuera JJ, Prior M, Galvan A, Fernandez E, Grossman AR (2011) Reverse genetics in Chlamydomonas: a platform for isolating insertional mutants. Plant Methods 7: 24

Goodenough U, Blaby I, Casero D, Gallaher SD, Goodson C, Johnson S, Lee J-H, Merchant SS, Pellegrini M, Roth R, Rusch J, Singh M, Umen JG, Weiss TL, Wulan T (2014) The path to triacylglyceride obesity in the sta6 strain of Chlamydomonas reinhardtii. Eukaryotic Cell 13: 591-613

Goodman JM (2008) The gregarious lipid droplet. Journal of Biological Chemistry 283: 28005-28009

Goodson C, Roth R, Wang ZT, Goodenough U (2011) Structural correlates of cytoplasmic and chloroplast lipid body synthesis in Chlamydomonas reinhardtii and stimulation of lipid body production with acetate boost. Eukaryotic Cell 10: 1592-1606

Goold HD, Cuine S, Legeret B, Liang Y, Brugiere S, Auroy P, Javot H, Tardif M, Jones B, Beisson F, Peltier G, Li-Beisson Y (2016) Saturating light induces sustained accumulation of oil in plastidal lipid droplets in Chlamydomonas reinhardtii. Plant Physiol 171: 2406-2417

Graham IA (2008) Seed storage oil mobilization. Annual Review of Plant Biology 59: 115-142

Green CR, Wallace M, Divakaruni AS, Phillips SA, Murphy AN, Ciaraldi TP, Metallo CM (2016) Branched-chain amino acid catabolism fuels adipocyte differentiation and lipogenesis. Nat Chem Biol 12: 15-21

Greiner A, Kelterborn S, Evers H, Kreimer G, Sizova I, Hegemann P (2017) Targeting of photoreceptor genes in Chlamydomonas reinhardtii via zinc-finger nucleases and CRISPR/Cas9. Plant Cell 29: 2498-2518

Grossman A, Croft M, Gladyshev V, Merchant S, Posewitz M, Prochnik S, Spalding M (2007) Novel metabolism in Chlamydomonas through the lens of genomics. Curr Opin Plant Biol 10: 190 - 198

Guiheneuf F, Leu S, Zarka A, Khozin-Goldberg I, Khalilov I, Boussiba S (2011) Cloning and molecular characterization of a novel acyl-CoA:diacylglycerol acyltransferase 1-like gene (PtDGAT1) from the diatom Phaeodactylum 107

tricornutum. FEBS Journal 278: 3651-3666

Harper AE, Miller RH, Block KP (1984) Branched-chain amino acid metabolism. Annu Rev Nutr 4: 409-454

Harris E (2001) Chlamydomonas as a model organism. Annu Review Plant Physiol Plant Mol Biol 52: 363 - 406

Hildebrandt Tatjana M, Nunes Nesi A, Araújo Wagner L, Braun H-P (2015) Amino acid catabolism in Plants. Molecular Plant 8: 1563-1579

Hu J, Baker A, Bartel B, Linka N, Mullen RT, Reumann S, Zolman BK (2012) Plant peroxisomes: biogenesis and function. The Plant Cell 24: 2279-2303

Hu Q, Sommerfeld M, Jarvis E, Ghirardi M, Posewitz M, Seibert M, Darzins A (2008) Microalgal triacylglycerols as feedstocks for biofuel production: perspectives and advances. The Plant Journal 54: 621-639

Huang AHC (1992) Oil bodies and oleosins in seeds. Annual Review of Plant Physiology and Plant Molecular Biology 43: 177-200

Huang AHC (1996) Oleosins and oil bodies in seeds and other organs. Plant Physiology 110: 1055-1061

Huang G, Chen F, Wei D, Zhang X, Chen G (2010) Biodiesel production by microalgal biotechnology. Applied Energy 87: 38-46

Hung C-H, Ho M-Y, Kanehara K, Nakamura Y (2013) Functional study of diacylglycerol acyltransferase type 2 family in Chlamydomonas reinhardtii. FEBS Letters 587: 2364-2370

Hurlock AK, Roston RL, Wang K, Benning C (2014) Lipid trafficking in plant cells. Traffic 15: 915-932

Ishizaki K, Larson TR, Schauer N, Fernie AR, Graham IA, Leaver CJ (2005) The critical role of Arabidopsis electron-transfer flavoprotein: Ubiquinone oxidoreductase during dark-induced starvation. Plant Cell 17: 2587-2600

James GO, Hocart CH, Hillier W, Chen H, Kordbacheh F, Price GD, Djordjevic MA (2011) Fatty acid profiling of Chlamydomonas reinhardtii under nitrogen deprivation. Bioresource Technology 102: 3343-3351

Jander G, Baerson SR, Hudak JA, Gonzalez KA, Gruys KJ, Last RL (2003) Ethylmethanesulfonate saturation mutagenesis in Arabidopsis to determine frequency of herbicide resistance. Plant Physiol 131: 139-146

Jenke-Kodama H, Sandmann A, Muller R, Dittmann E (2005) Evolutionary

108

implications of bacterial polyketide synthases. Mol Biol Evol 22: 2027-2039

Kajikawa M, Sawaragi Y, Shinkawa H, Yamano T, Ando A, Kato M, Hirono M, Sato N, Fukuzawa H (2015) Algal dual-specificity tyrosine phosphorylation- regulated kinase, triacylglycerol accumulation regulator1, regulates accumulation of triacylglycerol in nitrogen or sulfur deficiency. Plant Physiology 168: 752-764

Kang F, Rawsthorne S (1996) Metabolism of glucose-6-phosphate and utilization of multiple metabolites for fatty acid synthesis by plastids from developing oilseed rape embryos. Planta 199: 321-327

Kennedy EP (1956) The biological synthesis of phospholipids. Canadian Journal of Biochemistry and Physiology 34: 334-348

Kim S, Yamaoka Y, Ono H, Kim H, Shim D, Maeshima M, Martinoia E, Cahoon EB, Nishida I, Lee Y (2013) AtABCA9 transporter supplies fatty acids for lipid synthesis to the endoplasmic reticulum. Proc Natl Acad Sci U S A 110: 773-778

Kim Y, Terng EL, Riekhof WR, Cahoon EB, Cerutti H (2018) Endoplasmic reticulum acyltransferase with prokaryotic substrate preference contributes to triacylglycerol assembly in Chlamydomonas. Proc Natl Acad Sci U S A 115: 1652-1657

Kimball SR, Jefferson LS (2006) New functions for amino acids: effects on gene transcription and translation. Am J Clin Nutr 83: 500s-507s

Kimball SR, Jefferson LS (2006) Signaling pathways and molecular mechanisms through which branched-chain amino acids mediate translational control of protein synthesis. J Nutr 136: 227s-231s

Kindle KL (1990) High-efficiency nuclear transformation of Chlamydomonas reinhardtii. Proceedings of the National Academy of Sciences of the United States of America 87: 1228-1232

Kochevenko A, Araújo WL, Maloney GS, Tieman DM, Do PT, Taylor MG, Klee HJ, Fernie AR (2012) Catabolism of branched chain amino acids supports respiration but not volatile synthesis in tomato fruits. Molecular Plant 5: 366- 375

Kochevenko A, Fernie AR (2011) The genetic architecture of branched-chain amino acid accumulation in tomato fruits. J Exp Bot 62: 3895-3906

Kong F, Burlacot A, Liang Y, Legeret B, Alseekh S, Brotman Y, Fernie AR, Krieger-Liszkay A, Beisson F, Peltier G, Li-Beisson Y (2018) Interorganelle communication: peroxisomal MALATE DEHYDROGENASE 2 connects lipid catabolism to photosynthesis through redox coupling in Chlamydomonas. Plant 109

Cell 30:1824-1847

Kong F, Li-Beisson Y (2018) Identification of insertion site by RESDA-PCR in Chlamydomonas mutants generated by AphVIII random insertional mutagenesis. Bio-protocol 8: e2718

Kong F, Liang Y, Legeret B, Beyly-Adriano A, Blangy S, Haslam RP, Napier JA, Beisson F, Peltier G, Li-Beisson Y (2017) Chlamydomonas carries out fatty acid beta-oxidation in ancestral peroxisomes using a bona fide acyl-CoA oxidase. Plant J 90: 358-371.

Kong F, Romero IT, Warakanont J, Li-Beisson Y (2018) Lipid catabolism in microalgae. New Phytol 218: 1340-1348

Krishnan A, Kumaraswamy GK, Vinyard DJ, Gu HY, Ananyev G, Posewitz MC, Dismukes GC (2015) Metabolic and photosynthetic consequences of blocking starch biosynthesis in the green alga Chlamydomonas reinhardtii sta6 mutant. Plant Journal 81: 947-960

Kujala UM, Peltonen M, Laine MK, Kaprio J, Heinonen OJ, Sundvall J, Eriksson JG, Jula A, Sarna S, Kainulainen H (2016) Branched-chain amino acid levels are related with surrogates of disturbed lipid metabolism among older men. Frontiers in Medicine 3: 1-9

Kuzuyama T, Seto H (2003) Diversity of the biosynthesis of the isoprene units. Nat Prod Rep 20: 171-183

La Russa M, Bogen C, Uhmeyer A, Doebbe A, Filippone E, Kruse O, Mussgnug JH (2012) Functional analysis of three type-2 DGAT homologue genes for triacylglycerol production in the green microalga Chlamydomonas reinhardtii. Journal of Biotechnology 162: 13-20

Lange PR, Eastmond PJ, Madagan K, Graham IA (2004) An Arabidopsis mutant disrupted in valine catabolism is also compromised in peroxisomal fatty acid β- oxidation. FEBS Letters 571: 147-153

Langmead B, Salzberg SL (2012) Fast gapped-read alignment with Bowtie 2. Nat Meth 9: 357-359

Lauersen KJ, Willamme R, Coosemans N, Joris M, Kruse O, Remacle C (2016) Peroxisomal microbodies are at the crossroads of acetate assimilation in the green microalga Chlamydomonas reinhardtii. Algal Research 16: 266-274

Li-Beisson Y, Beisson F, Riekhof W (2015) Metabolism of acyl-lipids in Chlamydomonas reinhardtii. Plant Journal 82: 504-522

Li-Beisson Y, Nakamura Y, Harwood J (2016) Lipids: from chemical structures, 110

biosynthesis, and analyses to industrial applications. In Y Nakamura, Y Li- Beisson, eds, Lipids in Plant and Algae Development. Springer International Publishing, Cham, pp 1-18

Li-Beisson Y, Shorrosh B, Beisson F, Andersson MX, Arondel V, Bates PD, Baud S, Bird D, DeBono A, Durrett TP, Franke RB, Graham IA, Katayama K, Kelly AA, Larson T, Markham JE, Miquel M, Molina I, Nishida I, Rowland O, Samuels L, Schmid KM, Wada H, Welti R, Xu C, Zallot R, Ohlrogge J (2013) Acyl-lipid metabolism. The Arabidopsis Book: e0161

Li N, Gugel IL, Giavalisco P, Zeisler V, Schreiber L, Soll J, Philippar K (2015) FAX1, a novel membrane protein mediating plastid fatty acid export. PLoS Biol 13: e1002053

Li X, Moellering ER, Liu B, Johnny C, Fedewa M, Sears BB, Kuo M-H, Benning C (2012) A galactoglycerolipid lipase is required for triacylglycerol accumulation and survival following nitrogen deprivation in Chlamydomonas reinhardtii. The Plant Cell 24: 4670-4686

Li X, Zhang R, Patena W, Gang SS, Blum SR, Ivanova N, Yue R, Robertson JM, Lefebvre PA, Fitz-Gibbon ST, Grossman AR, Jonikas MC (2016) An indexed, mapped mutant library enables reverse genetics studies of biological processes in Chlamydomonas reinhardtii. The Plant Cell 28: 367-387

Li XB, Benning C, Kuo MH (2012) Rapid triacylglycerol turnover in Chlamydomonas reinhardtii requires a lipase with broad substrate specificity. Eukaryotic Cell 11: 1451-1462

Li Y, Han D, Hu G, Dauvillee D, Sommerfeld M, Ball S, Hu Q (2010) Chlamydomonas starchless mutant defective in ADP-glucose pyrophosphorylase hyper-accumulates triacylglycerol. Metabolic Engineering 12: 387-391

Lichtenthaler H (1987) Chlorophylls and carotenoids - pigments of photosynthetic biomembranes. Method Enzymol 148: 350 - 382

Lisec J, Schauer N, Kopka J, Willmitzer L, Fernie AR (2006) Gas chromatography mass spectrometry-based metabolite profiling in plants. Nat Protoc 1: 387-396

Liu B, Benning C (2013) Lipid metabolism in microalgae distinguishes itself. Current Opinion in Biotechnology 24: 300-309

Liu J, Han D, Yoon K, Hu Q, Li Y (2016) Characterization of type 2 diacylglycerol acyltransferases in Chlamydomonas reinhardtii reveals their distinct substrate specificities and functions in triacylglycerol biosynthesis. Plant J 86: 3-19

Londos C, Brasaemle DL, Schultz CJ, Segrest JP, Kimmel AR (1999) Perilipins, 111

ADRP, and other proteins that associate with intracellular neutral lipid droplets in animal cells. Seminars in Cell & Developmental Biology 10: 51-58

Love M, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology 15: 550

Ma F, Hanna MA (1999) Biodiesel production: a review1Journal Series #12109, Agricultural Research Division, Institute of Agriculture and Natural Resources, University of Nebraska–Lincoln.1. Bioresource Technology 70: 1-15

Maloney GS, Kochevenko A, Tieman DM, Tohge T, Krieger U, Zamir D, Taylor MG, Fernie AR, Klee HJ (2010) Characterization of the branched-chain amino acid aminotransferase enzyme family in tomato. Plant Physiol 153: 925-936

Mekhalfi M, Amara S, Robert S, Carrière F, Gontero B (2014) Effect of environmental conditions on various enzyme activities and triacylglycerol contents in cultures of the freshwater diatom, Asterionella formosa (Bacillariophyceae). Biochimie 101: 21-30

Merchant SS, Kropat J, Liu B, Shaw J, Warakanont J (2012) TAG, You’re it! Chlamydomonas as a reference organism for understanding algal triacylglycerol accumulation. Current Opinion in Biotechnology 23: 352-363

Merchant SS, Prochnik SE, Vallon O, Harris EH, Karpowicz SJ, Witman GB, et al. (2007) The Chlamydomonas genome reveals the evolution of key animal and plant functions. Science 318: 245-250

Michelet L, Zaffagnini M, Morisse S, Sparla F, Pérez-Pérez ME, Francia F, Danon A, Marchand CH, Fermani S, Trost P, Lemaire SD (2013) Redox regulation of the Calvin–Benson cycle: something old, something new. Frontiers in Plant Science 4: 470

Miller R, Wu GX, Deshpande RR, Vieler A, Gartner K, Li XB, Moellering ER, Zauner S, Cornish AJ, Liu BS, Bullard B, Sears BB, Kuo MH, Hegg EL, Shachar-Hill Y, Shiu SH, Benning C (2010) Changes in transcript abundance in Chlamydomonas reinhardtii following nitrogen deprivation predict diversion of metabolism. Plant Physiology 154: 1737-1752

Minowa T, Yokoyama S-y, Kishimoto M, Okakura T (1995) Oil production from algal cells of Dunaliella tertiolecta by direct thermochemical liquefaction. Fuel 74: 1735-1738

Moellering ER, Benning C (2010) RNA interference silencing of a major lipid droplet protein affects lipid droplet size in Chlamydomonas reinhardtii. Eukaryotic Cell 9: 97-106

Molnar A, Bassett A, Thuenemann E, Schwach F, Karkare S, Ossowski S (2009) 112

Highly specific gene silencing by artificial microRNAs in the unicellular alga Chlamydomonas reinhardtii. Plant J 58: 165-74

Molnar A, Schwach F, Studholme DJ, Thuenemann EC, Baulcombe DC (2007) miRNAs control gene expression in the single-cell alga Chlamydomonas reinhardtii. Nature 447: 1126-U1115

Mooney BP, Henzl MT, Miernyk JA, Randall DD (2000) The dihydrolipoyl acyltransferase (BCE2) subunit of the plant branched-chain alpha-ketoacid dehydrogenase complex forms a 24-mer core with octagonal symmetry. Protein Sci 9: 1334-1339

Moriyama T, Toyoshima M, Saito M, Wada H, Sato N (2018) Revisiting the algal "chloroplast lipid droplet": the absence of an entity that is unlikely to exist. 176: 1519-1530

Mou SL, Xu D, Ye NH, Zhang XW, Liang CW, Liang Q, Zheng Z, Zhuang ZM, Miao JL (2012) Rapid estimation of lipid content in an Antarctic ice alga (Chlamydomonas sp.) using the lipophilic fluorescent dye BODIPY505/515. Journal of Applied Phycology 24: 1169-1176

Muehlbauer GJ, Gengenbach BG, Somers DA, Donovan CM (1994) Genetic and amino-acid analysis of two maize threonine-overproducing, lysine-insensitive aspartate kinase mutants. Theor Appl Genet 89: 767-774

Murphy DJ (2001) The biogenesis and functions of lipid bodies in animals, plants and microorganisms. Progress in Lipid Research 40: 325-438

Newgard Christopher B Interplay between lipids and branched-chain amino acids in development of insulin resistance. Cell Metabolism 15: 606-614

Nguyen HM, Baudet M, Cuiné S, Adriano J-M, Barthe D, Billon E, Bruley C, Beisson F, Peltier G, Ferro M, Li-Beisson Y (2011) Proteomic profiling of oil bodies isolated from the unicellular green microalga Chlamydomonas reinhardtii: With focus on proteins involved in lipid metabolism. Proteomics 11: 4266-4273

Nguyen HM, Cuiné S, Beyly-Adriano A, Légeret B, Billon E, Auroy P, Beisson F, Peltier G, Li-Beisson Y (2013) The green microalga Chlamydomonas reinhardtii has a single ω-3 fatty acid desaturase that localizes to the chloroplast and impacts both plastidic and extraplastidic membrane lipids. Plant Physiology 163: 914-928

Nobusawa T, Hori K, Mori H, Kurokawa K, Ohta H (2017) Differently localized lysophosphatidic acid acyltransferases crucial for triacylglycerol biosynthesis in the oleaginous alga Nannochloropsis. Plant J 90: 547-559

113

Ohlrogge J, Browse J (1995) Lipid biosynthesis. Plant Cell 7: 957-970

Olukoshi ER, Packter NM (1994) Importance of stored triacylglycerols in Streptomyces: possible carbon source for antibiotics. Microbiology 140 ( Pt 4): 931-943

Pan Y, Yang J, Gong Y, Li X, Hu H (2017) 3-Hydroxyisobutyryl-CoA hydrolase involved in isoleucine catabolism regulates triacylglycerol accumulation in Phaeodactylum tricornutum. Philos Trans R Soc Lond B Biol Sci 372: pii: 20160409

Park J-J, Wang H, Gargouri M, Deshpande RR, Skepper JN, Holguin FO, Juergens MT, Shachar-Hill Y, Hicks LM, Gang DR (2015) The response of Chlamydomonas reinhardtii to nitrogen deprivation: a systems biology analysis. The Plant Journal 81: 611-624

Peng C, Uygun S, Shiu SH, Last RL (2015) The impact of the branched-chain ketoacid dehydrogenase complex on amino acid homeostasis in Arabidopsis. Plant Physiology 169: 1807-1820

Plancke C, Vigeolas H, Hohner R, Roberty S, Emonds-Alt B, Larosa V, Willamme R, Duby F, Onga Dhali D, Thonart P, Hiligsmann S, Franck F, Eppe G, Cardol P, Hippler M, Remacle C (2014) Lack of isocitrate lyase in Chlamydomonas leads to changes in carbon metabolism and in the response to oxidative stress under mixotrophic growth. Plant J 77: 404-417

Pleite R, Pike MJ, Garces R, Martinez-Force E, Rawsthorne S (2005) The sources of carbon and reducing power for fatty acid synthesis in the heterotrophic plastids of developing sunflower (Helianthus annuus L.) embryos. J Exp Bot 56: 1297-1303

Podebrad F, Heil M, Reichert S, Mosandl A, Sewell AC, Bohles H (1999) 4,5- dimethyl-3-hydroxy-2[5H]-furanone (sotolone)--the odour of maple syrup urine disease. J Inherit Metab Dis 22: 107-114

Poirier Y, Antonenkov VD, Glumoff T, Hiltunen JK (2006) Peroxisomal beta- oxidation - a metabolic pathway with multiple functions. Biochimica Et Biophysica Acta-Molecular Cell Research 1763: 1413-1426

Ramanan R, Kim BH, Cho DH, Ko SR, Oh HM, Kim HS (2013) Lipid droplet synthesis is limited by acetate availability in starchless mutant of Chlamydomonas reinhardtii. FEBS Lett 587: 370-377

Rawsthorne S (2002) Carbon flux and fatty acid synthesis in plants. Progress in Lipid Research 41: 182-196

Reumann S, Quan S, Aung K, Yang P, Manandhar-Shrestha K, Holbrook D, Linka 114

N, Switzenberg R, Wilkerson CG, Weber AP, Olsen LJ, Hu J (2009) In- depth proteome analysis of Arabidopsis leaf peroxisomes combined with in vivo subcellular targeting verification indicates novel metabolic and regulatory functions of peroxisomes. Plant Physiol 150: 125-143

Sasaki Y, Nagano Y (2004) Plant acetyl-CoA carboxylase: structure, biosynthesis, regulation, and gene manipulation for plant breeding. Biosci Biotechnol Biochem 68: 1175-1184

Sasso S, Stibor H, Mittag M, Grossman AR (2018) From molecular manipulation of domesticated Chlamydomonas reinhardtii to survival in nature. eLife 7: e39233

Schertl P, Braun HP (2014) Respiratory electron transfer pathways in plant mitochondria. Front Plant Sci 5: 163

Schmollinger S, Muhlhaus T, Boyle NR, Blaby IK, Casero D, Mettler T, Moseley JL, Kropat J, Sommer F, Strenkert D, Hemme D, Pellegrini M, Grossman AR, Stitt M, Schroda M, Merchant SS (2014) Nitrogen-sparing mechanisms in Chlamydomonas affect the transcriptome, the proteome, and photosynthetic metabolism. Plant Cell 26: 1410-1435

Schmollinger S, Strenkert D, Schroda M (2010) An inducible artificial microRNA system for Chlamydomonas reinhardtii confirms a key role for heat shock factor 1 in regulating thermotolerance. Current Genetics 56: 383-389

Schulz-Raffelt M, Chochois V, Auroy P, Cuiné S, Billon E, Dauvillée D, Li-Beisson Y, Peltier G (2016) Hyper-accumulation of starch and oil in a Chlamydomonas mutant affected in a plant-specific DYRK kinase. Biotechnology for Biofuels 9: 55

Schwender J, Ohlrogge JB, Shachar-Hill Y (2003) A flux model of glycolysis and the oxidative pentosephosphate pathway in developing Brassica napus embryos. Journal of Biological Chemistry 278: 29442-29453

Sengupta S, Peterson TR, Sabatini DM (2010) Regulation of the mTOR complex 1 pathway by nutrients, growth factors, and stress. Mol Cell 40: 310-322

Shimomura Y, Honda T, Shiraki M, Murakami T, Sato J, Kobayashi H, Mawatari K, Obayashi M, Harris RA (2006) Branched-chain amino acid catabolism in exercise and liver disease. J Nutr 136: 250s-253s

Shimomura Y, Murakami T, Nakai N, Nagasaki M, Harris RA (2004) Exercise promotes BCAA catabolism: effects of BCAA supplementation on skeletal muscle during exercise. J Nutr 134: 1583s-1587s

Shin YS, Jeong J, Nguyen THT, Kim JYH, Jin E, Sim SJ (2019) Targeted knockout of phospholipase A2 to increase lipid productivity in Chlamydomonas 115

reinhardtii for biodiesel production. Bioresour Technol 271: 368-374

Siaut M, Cuine S, Cagnon C, Fessler B, Nguyen M, Carrier P, Beyly A, Beisson F, Triantaphylides C, Li-Beisson Y, Peltier G (2011) Oil accumulation in the model green alga Chlamydomonas reinhardtii: characterization, variability between common laboratory strains and relationship with starch reserves. BMC Biotechnology 11: 7

Singh BK, Shaner DL (1995) Biosynthesis of branched chain amino acids: From Test Tube to Field. Plant Cell 7: 935-944

Slabas AR, Fawcett T (1992) The biochemistry and molecular biology of plant lipid biosynthesis. In RA Schilperoort, L Dure, eds, 10 Years Plant Molecular Biology. Springer Netherlands, Dordrecht, pp 169-191

Smith RG, Gauthier DA, Dennis DT, Turpin DH (1992) Malate- and pyruvate- dependent fatty acid synthesis in leucoplasts from developing castor endosperm. Plant Physiology 98: 1233

Stobart K, Mancha M, Lenman M, Dahlqvist A, Stymne S (1997) Triacylglycerols are synthesised and utilized by transacylation reactions in microsomal preparations of developing safflower (Carthamus tinctorius L.) seeds. Planta 203: 58-66

Tardif M, Atteia A, Specht M, Cogne G, Rolland N, Brugière S, Hippler M, Ferro M, Bruley C, Peltier G, Vallon O, Cournac L (2012) PredAlgo, a new subcellular localization prediction tool dedicated to green algae. Molecular Biology and Evolution 29: 3625-39.

Taylor NL, Heazlewood JL, Day DA, Millar AH (2004) Lipoic acid-dependent oxidative catabolism of alpha-keto acids in mitochondria provides evidence for branched-chain amino acid catabolism in Arabidopsis. Plant Physiol 134: 838- 848

Terashima M, Specht M, Hippler M (2011) The chloroplast proteome: a survey from the Chlamydomonas reinhardtii perspective with a focus on distinctive features. Current Genetics 57: 151-168

Tolleter D, Ghysels B, Alric J, Petroutsos D, Tolstygina I, Krawietz D, Happe T, Auroy P, Adriano JM, Beyly A, Cuine S, Plet J, Reiter IM, Genty B, Cournac L, Hippler M, Peltier G (2011) Control of hydrogen photoproduction by the proton gradient generated by cyclic electron flow in Chlamydomonas reinhardtii. Plant Cell 23: 2619-2630

Troncoso-Ponce MA, Cao X, Yang Z, Ohlrogge JB (2013) Lipid turnover during senescence. Plant Science 205: 13-19

116

Tsai C-H, Warakanont J, Takeuchi T, Sears BB, Moellering ER, Benning C (2014) The protein Compromised Hydrolysis of Triacylglycerols 7 (CHT7) acts as a repressor of cellular quiescence in Chlamydomonas. Proceedings of the National Academy of Sciences 111: 15833-15838

Tsai CH, Uygun S, Roston R, Shiu SH, Benning C (2017) Recovery from N deprivation is a transcriptionally and functionally distinct state in Chlamydomonas. Plant Physiol 176: 2007-2023

Tunçay H, Findinier J, Duchêne T, Cogez V, Cousin C, Peltier G, Ball SG, Dauvillée D (2013) A forward genetic approach in Chlamydomonas reinhardtii as a strategy for exploring starch catabolism. PLoS ONE 8: e74763

Urzica EI, Vieler A, Hong-Hermesdorf A, Page MD, Casero D, Gallaher SD, Kropat J, Pellegrini M, Benning C, Merchant SS (2013) Remodeling of membrane lipids in iron-starved Chlamydomonas. Journal of Biological Chemistry 288: 30246-30258

Valerio A, D'Antona G, Nisoli E (2011) Branched-chain amino acids, mitochondrial biogenesis, and healthspan: an evolutionary perspective. Aging (Albany NY) 3: 464-478 van Meer G, Voelker DR, Feigenson GW (2008) Membrane lipids: where they are and how they behave. Nature Reviews Molecular Cell Biology 9: 112-124

Vieler A, Wilhelm C, Goss R, Sub R, Schiller J (2007) The lipid composition of the unicellular green alga Chlamydomonas reinhardtii and the diatom Cyclotella meneghiniana investigated by MALDI-TOF MS and TLC. Chemistry and Physics of Lipids 150: 143-155

Voelker T, Kinney AJ (2001) Variations in the biosynthesis of seed-storage lipids. Annu Rev Plant Physiol Plant Mol Biol 52: 335-361

Wang J, Liu H, Chen G, Tsuei SH, Yu T, Fu Q (2011) Identification of two novel BCKDHA mutations in a Chinese patient with maple syrup urine disease. J Pediatr Endocrinol Metab 24: 827-829

Wang X, Dong HP (2018) Dual expression of plastidial GPAT1 and LPAT1 regulates triacylglycerol production and the fatty acid profile in Phaeodactylum tricornutum. 11: 318

Wang X, Proud CG (2009) Nutrient control of TORC1, a cell-cycle regulator. Trends Cell Biol 19: 260-267

Wang ZT, Ullrich N, Joo S, Waffenschmidt S, Goodenough U (2009) Algal lipid bodies: stress induction, purification, and biochemical characterization in wild- type and starchless Chlamydomonas reinhardtii. Eukaryot Cell 8 : 1856-68 117

Warakanont J, Tsai C-H, Michel EJS, Murphy GR, Hsueh PY, Roston RL, Sears BB, Benning C (2015) Chloroplast lipid transfer processes in Chlamydomonas reinhardtii involving a TRIGALACTOSYLDIACYLGLYCEROL 2 (TGD2) orthologue. The Plant Journal 84: 1005-1020

Wijffels RH, Barbosa MJ (2010) An outlook on microalgal biofuels. Science 329: 796-799

Work VH, Radakovits R, Jinkerson RE, Meuser JE, Elliott LG, Vinyard DJ, Laurens LML, Dismukes GC, Posewitz MC (2010) Increased lipid accumulation in the Chlamydomonas reinhardtii sta7-10 starchless isoamylase mutant and increased carbohydrate synthesis in complemented strains. Eukaryotic Cell 9: 1251-1261

Wu B, Zhang B, Feng X, Rubens JR, Huang R, Hicks LM, Pakrasi HB, Tang YJ (2010) Alternative isoleucine synthesis pathway in cyanobacterial species. Microbiology 156: 596-602

Yamaoka Y, Achard D, Jang S, Legeret B, Kamisuki S, Ko D, Schulz-Raffelt M, Kim Y, Song WY, Nishida I, Li-Beisson Y, Lee Y (2016) Identification of a Chlamydomonas plastidial 2-lysophosphatidic acid acyltransferase and its use to engineer microalgae with increased oil content. Plant Biotechnol J 14: 2158- 2167

Yang D, Song D, Kind T, Ma Y, Hoefkens J, Fiehn O (2015) Lipidomic analysis of Chlamydomonas reinhardtii under nitrogen and sulfur deprivation. PLoS One 10: e0137948

Yang L, Ding YF, Chen Y, Zhang SY, Huo CX, Wang Y, Yu JH, Zhang P, Na HM, Zhang HN, Ma YB, Liu PS (2012) The proteomics of lipid droplets: structure, dynamics, and functions of the organelle conserved from bacteria to humans. Journal of Lipid Research 53: 1245-1253

Yao L, Shen H, Wang N, Tatlay J, Li L, Tan TW, Lee YK (2017) Elevated acetyl- CoA by amino acid recycling fuels microalgal neutral lipid accumulation in exponential growth phase for biofuel production. Plant Biotechnology Journal 15: 497-509

Yeaman SJ (1989) The 2-oxo acid dehydrogenase complexes: recent advances. Biochem J 257: 625-632

Yokochi T, Honda D, Higashihara T, Nakahara T (1998) Optimization of docosahexaenoic acid production by Schizochytrium limacinum SR21. Applied Microbiology and Biotechnology 49: 72-76

Yoon K, Han D, Li Y, Sommerfeld M, Hu Q (2012) Phospholipid:diacylglycerol

118

acyltransferase is a multifunctional enzyme involved in membrane lipid turnover and degradation while synthesizing triacylglycerol in the unicellular green microalga Chlamydomonas reinhardtii. The Plant Cell 24: 3708-3724

Zhang M, Fan J, Taylor DC, Ohlrogge JB (2009) DGAT1 and PDAT1 acyltransferases have overlapping functions in Arabidopsis triacylglycerol biosynthesis and are essential for normal pollen and seed development. Plant Cell 21: 3885-3901

Zones JM, Blaby IK, Merchant SS, Umen JG (2015) High-resolution profiling of a synchronized diurnal transcriptome from Chlamydomonas reinhardtii reveals continuous cell and metabolic differentiation. The Plant Cell 27: 2743-2769

Zweytick D, Athenstaedt K, Daum G (2000) Intracellular lipid particles of eukaryotic cells. Biochimica Et Biophysica Acta-Reviews on Biomembranes 1469: 101- 120

119

CURRICULUM VITAE

Yuanxue LIANG

Institute of Biosciences and Biotechnologies Aix-Marseille, CEA, Aix-Marseille University, UMR7265, Saint-Paul-Lez-Durance, France Mobile: +33683106993 Email: [email protected] Date of Birth: September 1988 Gender: Male

OBJECTIVES: looking for a postdoc position in an English speaking country

EDUCATION: Major in Bioscience and Biotechnologies, College of doctoral 2015.03-2019.01 school ‘Ecole Doctorale Sciences de la vie et de la santé’, Aix- PhD Marseille University, Marseille, 13288, France. 2011.08-2014.12 Major in Biological Chemical Engineering, College of Master of Materials and Chemical Engineering, Hainan University, Engineering Haikou, Hainan, 570228, P.R. China Major in Biochemistry and Molecular Biology, College of Life 2012.11-2013.01 Sciences, Sun Yat-sen University, Guangzhou, 510275, P.R. China 2007.09-2011.08 Major in Bio-engineering, College of Materials and Chemical Bachelor of Engineering, Hainan University, Haikou, Hainan, 570228, P.R. Engineering China

AWARDS: Since 2008, I have the honor to win National Encouraged Scholarship for one time, academy Miyoshi Outstanding Students for three times, First Grade Integrated Scholarship for one time and Third Grade Integrated Scholarship twice. - Award of Excellent Master Thesis from Hainan University, 2014. 120

- Award Chinese Government Scholarship to sponsor study abroad, 2014.

INVOLVED FUNDING PROGRAMMES: - Identification and functional analysis of microRNA involved in oil palm fatty acid biosynthesis, NSFC, No. 31260193, 2013. - Construction of oil palm fruit normalized full-length cDNA library and screening and cloning of fatty acid metabolism related genes. NSFC, No. 31160171, 2012. - Identification and cloning of genes associated with endosperm development in Coconut (Cocos nucifera L.) NFSC, No. 31060259, 2011. - Cloning and functional analysis of WRI1-like transcription factor of coconut endosperm in regulation of fatty acid metabolism, NFSC, No. 31360476, 2010.

PARTICIPATION TO SCIENTIFIC MEETINGS: Oral presentation - 25th annual meeting of doctoral school in Aix-Marseille University (2017.06.01-02) Liang Y. Interplay between catabolism of branched-chain amino acids and triacylglycerol homeostasis in Chlamydomonas reinhardtii. - PhD student day, BIAM, CEA Cadarache (2017.12.19) Liang Y. Interplay between catabolism of branched-chain amino acids and triacylglycerol homeostasis in Chlamydomonas reinhardtii. Poster presentation - 2017.07.02-05 8th European symposium on plant lipids in Suede Malmö of Sweden Yuanxue Liang, Fantao Kong, Adrien Burlacot, Ismael Torres Romero, Bertrand Legeret, Audrey Beyly, Fred Beisson, Gilles Peltier, Yonghua Li-Beisson. Interplay between Catabolism of Branched-chain Amino Acids and Triacylglycerol Homeostasis in Chlamydomonas reinhardtii - Imaging in life Sciences! From functions to molecular structures in Marseille centre d’Immunologie de Marseille Luminy (CIML) and Marseille (2017.10.12)

121

Yuanxue Liang, Fantao Kong, Adrien Burlacot, Ismael Torres Romero, Bertrand Legeret, Audrey Beyly, Fred Beisson, Gilles Peltier, Yonghua Li-Beisson Interplay between Catabolism of Branched-chain Amino Acids and Triacylglycerol Homeostasis in Chlamydomonas reinhardtii - DOC2AMU Interdisciplinary Doctoral Day, in conference room of Campus Saint- Charles (2017.10.13) Yuanxue Liang, Fantao Kong, Adrien Burlacot, Ismael Torres Romero, Bertrand Legeret, Audrey Beyly, Fred Beisson, Gilles Peltier, Yonghua Li-Beisson Interplay between Catabolism of Branched-chain Amino Acids and Triacylglycerol Homeostasis in Chlamydomonas reinhardtii - Biogenesis and Fate of Lipid Droplets, 14th GERLI Lipidomics meeting, in St Maximin la Saint-Baume of Aix-en-provence (2018.09.30-10.03) Yuanxue Liang, Fantao Kong, Ismael Torres-Romero, Adrien Burlacot, Stéphan Cuine, Bertrand Légeret, Saleh Alseekh, Alisdair R Fernie, Fred Beisson, Gilles Peltier, Yonghua Li-Beisson Interplay between catabolism of branched-chain amino acids and triacylglycerol homeostasis in Chlamydomonas reinhardtii

PUBLICATIONS: Work from PhD thesis: 1. Yuanxue Liang, Fantao Kong, Ismael Torres-Romero, Adrien Burlacot, Stéphan Cuine, Bertrand Légeret, Emmanuelle Billon, Saleh Alseekh, Alisdair R Fernie, Fred Beisson, Gilles Peltier, Yonghua Li-Beisson. Branched Chain Amino Acid Catabolism Impacts Triacylglycerol Metabolism in Chlamydomonas reinhardtii. Plant physiology (2018) (submitted) 2. Fantao Kong, Yuanxue Liang, Bertrand Légeret, Audrey Beyly‐Adriano, Stéphanie Blangy, Richard P. Haslam, Johnathan A. Napier, Fred Beisson, Gilles Peltier, and Yonghua Li‐Beisson. Chlamydomonas carries out fatty acid β‐ oxidation in ancestral peroxisomes using a bona fide acyl‐CoA oxidase. The Plant

122

Journal 90, no. 2 (2017): 358-371. 3. Fantao Kong, Adrien Burlacot, Yuanxue Liang, Bertrand Légeret, Saleh Alseekh, Yariv Brotman, Alisdair R. Fernie, Anja Krieger-Liszkay, Fred Beisson, Gilles Peltier, Yonghua Li-Beisson. Interorganelle Communication: Peroxisomal MALATE DEHYDROGENASE2 Connects Lipid Catabolism to Photosynthesis through Redox Coupling in Chlamydomonas. The Plant Cell 30, no. 8 (2018): 1824-1847. 4. Hugh Douglas Goold, Stéphan Cuiné, Bertrand Legeret, Yuanxue Liang, Sabine Brugière, Pascaline Auroy, Helene Javot, Marianne Tardif, Brian J Jones, Fred Beisson, Gilles Peltier, Yonghua Li-Beisson. Saturating light induces sustained accumulation of oil in plastidal lipid droplets in Chlamydomonas reinhardtii. Plant physiology (2016): pp-00718.

Work from Master thesis: 5. Yusheng Zheng, Chongjian Chen, Yuanxue Liang, Ruhao Sun, Lingchao Gao, Tao Liu, and Dongdong Li. Genome-wide association analysis of the lipid and fatty acid metabolism regulatory network in the mesocarp of oil palm (Elaeis guineensis Jacq.) based on small noncoding RNA sequencing. Tree physiology (2018). 6. Yijun Yuan, Lingchao Gao, Ruhao Sun, Tian Yu, Yuanxue Liang, Dongdong Li, and Yusheng Zheng. "Seed-specific expression of an acyl-acyl carrier protein thioesterase CnFatB3 from coconut (Cocos nucifera L.) increases the accumulation of medium-chain fatty acids in transgenic Arabidopsis seeds. Scientia Horticulturae 223 (2017): 5-9. 7. Yijun Yuan, Yuanxue Liang, Linchao Gao, Ruhao Sun, Yusheng Zheng, and Dongdong Li. Functional heterologous expression of a lysophosphatidic acid acyltransferase from coconut (Cocos nucifera L.) endosperm in Saccharomyces cerevisiae and Nicotiana tabacum. Scientia Horticulturae 192 (2015): 224-230. 8. Yuanxue Liang, Yijun Yuan, Tao Liu, Wei Mao, Yusheng Zheng, Dongdong Li. Identification and computational annotation of genes differentially expressed in

123

pulp development of Cocos nucifera L. by suppression subtractive hybridization. BMC Plant Biology, 2014, 14:205. 9. Yijun Yuan, Yuanxue Liang, Baozhu Li, Yusheng Zheng, Xianqun Luo, and Li Dongdong. Cloning and Function Characterization of a β-Ketoacyl-acyl-ACP Synthase I from Coconut (Cocos nucifera L.) Endosperm. Plant molecular biology reporter 33, no. 4 (2015): 1131-1140. 10. Lingchao Gao, Ruhao Sun, Yuanxue Liang, Mengdan Zhang, Yusheng Zheng, Dongdong Li. Cloning and functional expression of a cDNA encoding stearoyl- ACP Δ9-desaturase from the endosperm of coconut (Cocos nucifera L.). Gene, 2014. 11. Yijun Yuan, Yuanxue Liang, Yusheng Zheng, Dongdong Li. Cloning, characterization and expression analysis of a 7S globulin gene in mesocarp of oil palm (Elaeis guineensis jacq.). Scientia Horticulturae, 2012, 143:167-175. 12. Yijun Yuan, Yinhua Chen, Shan Yan, Yuanxue Liang, Yusheng Zheng, Dongdong Li. Molecular cloning and characterization of an ACP thioesterase gene (CocoFatB1) expressed in the endosperm of coconut (Cocos nucifera L.) and its heterologous expression in Nicotiana tabacum to engineer the accumulation of different fatty acids. Functional Plant Biology, 2014, 41:80-86.

Work from Master thesis (publications in Chinese) 13. Xiaoli Li, Yuanxue Liang, Lingchao Gao, Dongdong Li and Yusheng zheng. Changes of carotenoids in the fruit of oil palm (Elaeis guineensis) at different stages of development. Journal of Huazhong Agricultural University, 2015, 34(1):23-27. (In Chinese) 14. Liang Fang, Yuanxue Liang, Dongdong Li, Xianying Cao, Yusheng Zheng. Dynamic expression analysis of miRNAs during the development process of oil palm mesocarp. Plant Science Journal, 2013, 31:304-312. (In Chinese) 15. Yuanxue Liang, Yijun Yuan, Yujia Bao, Dongdong Li. Dynamic expression of one cytochrome P450-like gene in mesocarp of oil palm nut. Journal of Huazhong

124

Agricultural University, 2012, 31:419-422. (In Chinese) 16. Junfeng Zhang, Mingnan Zhang, Yuanxue Liang, Gangcheng Zuo, Youquan Lin. Chemical constituents in the roots of Evodia lepta (Spreng.) Merr (II). Natural Science Journal of Hainan University, 2011, 29:39-41. (In Chinese) 17. Junfeng Zhang, Zhifeng Dou, Yang Bai, Yuanxue Liang, Gangcheng Zuo, Youquan Lin. Chemical Constituents in the Roots of Evodia lepta (Spreng.) Merr. Natural Product Research and Development, 2011, 23:1061-1063. (In Chinese)

125

Etude des sources de carbone et d’énergie pour la synthèse des lipides de stockage chez la microalgue verte modèle Chlamydomonas reinhardtii Les triacylglycérols d'algues (TAG) représentent une source prometteuse de biocarburants. Les principales étapes de la synthèse des acides gras et du métabolisme du TAG des algues ont été déduites de celles des plantes terrestres, mais on en sait peu sur les sources de carbones et d’énergie intervenant dans la synthèse de lipides de réserve. Pour répondre à cette question, nous avons étudié la synthèse des acides gras chez l’algue modèle Chlamydomonas reinhardtii en utilisant une combinaison d'approches génétiques, biochimiques et microscopiques. Plus précisément, j'ai d'abord examiné la localisation subcellulaire de gouttelettes de lipides dans des cellules d'algues exposées à une forte lumière, conditions où une plus grande quantité de pouvoir réducteur est produite. J'ai ensuite contribué à mettre en évidence que la bêta-oxydation des acides gras est un processus peroxysomal, et que pendant une carence en azote réalisée en conditions photoautotrophe, des mutants dépourvus de la malate déshydrogénase 2 peroxysomale (mdh2) accumulent 50% plus TAG que les souches parentales. Ces résultats nous ont permis de mettre en évidence l'importance du contexte redox cellulaire sur la synthèse lipidique. Cette étude a également permis de révéler l’existence d'un échange d’énergie entre le peroxysome et le chloroplaste. Enfin, en caractérisant des mutants déficients dans la dégradation des acides aminés à chaîne ramifiée (BCAA), j'ai montré que le catabolisme des BCAAs joue un double rôle dans la synthèse de TAG en fournissant des précurseurs carbonés et de l'ATP. L'ensemble de ces travaux a mis en évidence l'existence d’interactions complexes entre le métabolisme du carbone et le métabolisme énergétique dans les cellules photosynthétiques, et ouvert de nouvelles pistes pour l'amélioration génétique future de souches d'algues pour la production de biocarburants.

Study of carbon and energy sources for storage lipid synthesis in model green microalga Chlamydomonas reinhardtii Algal triacylglycerols (TAG) represent a promising source for biofuel. The major steps for fatty acid synthesis and TAG metabolism have been deduced based on that of land plants, but little is known about carbon and energy sources. To address this question, we investigated fatty acid synthesis in algal cells using a combination of genetic, biochemical and microscopic approaches in the model microalga Chlamydomonas reinhardtii. Specifically, I first examined the subcellular localization of lipid droplets in algal cells exposed to high light, a condition favoring production of reducing power. Secondly, I contributed to put on evidence that the β-oxidation of fatty acids is a peroxisomal process, and that during photoautotrophic nitrogen starvation, knock-out mutants of the peroxisomal malate dehydrogenase 2 (mdh2) made 50% more TAG than parental strains, highlighting the importance of cellular redox context on lipid synthesis. This study also revealed for the first time the occurrence of an energy trafficking pathway from peroxisome to chloroplast. And finally, by characterizing mutants defected in degradation of branched-chain amino acids (BCAAs), we showed that BCAA catabolism plays a dual role in TAG synthesis via providing carbon precursors and ATP. Taken together, this work highlighted the complex interplay between carbon and energy metabolism in green photosynthetic cells, and pointed future directions for genetic improvement of algal strains for biofuel productions.

126