membrane proteomic insights into the physiology and ... · metabolomics have become essential for...

Membrane Proteomic Insights into the Physiology andTaxonomy of an Oleaginous Green Microalga1

Adriana Garibay-Hernández, Bronwyn J. Barkla, Rosario Vera-Estrella, Alfredo Martinez, andOmar Pantoja*

Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, 62210 Mexico(A.G.-H., R.V.-E., A.M., O.P.); and Southern Cross Plant Science, Southern Cross University, Lismore, 2480New South Wales, Australia (B.J.B.)

ORCID IDs: 0000-0002-5695-356X (A.G.-H.); 0000-0002-4691-8023 (B.J.B.); 0000-0002-4804-6687 (A.M.); 0000-0002-6538-9059 (O.P.).

Ettlia oleoabundans is a nonsequenced oleaginous green microalga. Despite the significant biotechnological interest in producingvalue-added compounds from the acyl lipids of this microalga, a basic understanding of the physiology and biochemistry ofoleaginous microalgae is lacking, especially under nitrogen deprivation conditions known to trigger lipid accumulation. Usingan RNA sequencing-based proteomics approach together with manual annotation, we are able to provide, to our knowledge, thefirst membrane proteome of an oleaginous microalga. This approach allowed the identification of novel proteins inE. oleoabundans, including two photoprotection-related proteins, Photosystem II Subunit S and Maintenance of Photosystem IIunder High Light1, which were considered exclusive to higher photosynthetic organisms, as well as Retinitis Pigmentosa Type2-Clathrin Light Chain, a membrane protein with a novel domain architecture. Free-flow zonal electrophoresis of microalgalmembranes coupled to liquid chromatography-tandem mass spectrometry proved to be a useful technique for determining theintracellular location of proteins of interest. Carbon-flow compartmentalization in E. oleoabundans was modeled using thisinformation. Molecular phylogenetic analyses of protein markers and 18S ribosomal DNA support the reclassification ofE. oleoabundans within the trebouxiophycean microalgae, rather than with the Chlorophyceae class, in which it is currentlyclassified, indicating that it may not be closely related to the model green alga Chlamydomonas reinhardtii. A detailed survey ofbiological processes taking place in the membranes of nitrogen-deprived E. oleoabundans, including lipid metabolism, providesinsights into the basic biology of this nonmodel organism.

Ettlia oleoabundans (taxonomic synonym of Neochlorisoleoabundans) is a unicellular edaphic green microalgathat belongs to the Chlorophyta phylum and iscurrently placed within the Chlorophyceae class(Chantanachat and Bold, 1962; Deason et al., 1991). It isa nonsequenced microalga classified as oleaginous dueto its high lipid content (up to 56% [w/w] of its dryweight; Gouveia et al., 2009). Several abiotic stressconditions, such as high temperature (Yang et al., 2013),

high salinity (Arredondo-Vega et al., 1995), and nitro-gen deficiency (Tornabene et al., 1983; Li et al., 2008;Pruvost et al., 2009; Garibay-Hernández et al., 2013),trigger neutral lipid accumulation in this microalga.E. oleoabundans is a highly versatile organism, as it cangrow in freshwater, wastewater (Levine et al., 2011;Wang and Lan, 2011; Yang et al., 2011; Olguín et al.,2015), and in culture media with salt concentrations upto seawater levels (Arredondo-Vega et al., 1995;Baldisserotto et al., 2012; Popovich et al., 2012). More-over, it is able to grow under phototrophic, mixotrophic(Giovanardi et al., 2013; Baldisserotto et al., 2016), andheterotrophic (Wu et al., 2011; Morales-Sánchez et al.,2013) conditions. Owing to its high lipid content andgrowth versatility, E. oleoabundans is an organism ofbiotechnological interest. However, a basic under-standing of its physiology is currently lacking, as mostreports have focused on improving the lipid yield andproductivity of E. oleoabundans under nitrogen defi-ciency conditions through different culture strategiesand on evaluating how other environmental factorsadditionally control lipid production. At present, only afew reports have assessed the biology and biochemistrybehind nitrogen deficiency and lipid accumulation inE. oleoabundans (Rismani-Yazdi et al., 2012; Benvenutiet al., 2015; Baldisserotto et al., 2016; Matich et al., 2016).

Rapidly developing postgenomics, systems biologyapproaches such as transcriptomics, proteomics, and

1 This work was supported by CONACyT-Mexico (grant no.220085 to O.P., grant no. 178232 to R.V.-E., and a scholarship toA.G.-H.) and by the Universidad Nacional Autónoma de México-DGAPA (grant no. IN202514 to R.V.-E. and grant no. IT200312 toA.M.).

* Address correspondence to [email protected] author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policy de-scribed in the Instructions for Authors (www.plantphysiol.org) is:Omar Pantoja ([email protected]).

A.G.-H. and O.P. conceived the research; A.G.-H., B.J.B., R.V.-E.,and O.P. designed the experiments; A.G.-H. performed the experi-ments and analyzed and interpreted the data. B.J.B. and R.V.-E. pro-vided support in membrane fractionation and in proteomics dataanalysis and validation; A.G.-H., B.J.B., and O.P. discussed the re-sults; A.G.-H. wrote the article, and B.J.B. and O.P. edited and com-plemented the writing; all authors supervised the experiments andrevised the final article.

www.plantphysiol.org/cgi/doi/10.1104/pp.16.01240

390 Plant Physiology�, January 2017, Vol. 173, pp. 390–416, www.plantphysiol.org � 2017 American Society of Plant Biologists. All Rights Reserved. www.plantphysiol.orgon December 26, 2019 - Published by Downloaded from

Copyright © 2017 American Society of Plant Biologists. All rights reserved.

http://orcid.org/0000-0002-5695-356X

http://orcid.org/0000-0002-4691-8023

http://orcid.org/0000-0002-4804-6687

http://orcid.org/0000-0002-6538-9059

mailto:[email protected]

http://www.plantphysiol.org

mailto:[email protected]

http://www.plantphysiol.org/cgi/doi/10.1104/pp.16.01240


metabolomics have become essential for understand-ing the physiology of different organisms, includingmicroalgae (Jinkerson et al., 2011; Ndimba et al., 2013).Algal proteomics has been performed primarily withthe model green alga Chlamydomonas reinhardtii for theanalysis of subcellular compartments (Schmidt et al.,2006; Atteia et al., 2009; Terashima et al., 2010) and thecharacterization of the proteome under stress condi-tions (Chen et al., 2010; Baba et al., 2011; Castruita et al.,2011; Mühlhaus et al., 2011), including nitrogen star-vation (Schmollinger et al., 2014; Valledor et al., 2014;Wase et al., 2014). Although C. reinhardtii currentlyprovides the best model for microalgal lipid research(Liu and Benning, 2013), it is not an oleaginous speciesand may not represent the physiology of other speciesof biotechnological interest, as microalgae comprise anextremely diverse group of photosynthetic microor-ganisms (Hu et al., 2008). In recent years, a limitednumber of proteomics studies have been performed onnonmodel oleaginous strains such as Chlorella proto-thecoides (Gao et al., 2014), Chlorella sorokiniana (Maet al., 2013), Chlorella vulgaris (Guarnieri et al., 2011,2013), Isochrysis galbana (Song et al., 2013), and Nanno-chloropsis oceanica (Dong et al., 2013). However, pro-teomic analysis of nonmodel microalgae is stillchallenging, as the lack of a sequenced genome in mostof them compromises the quality and quantity of thegenerated data (Ndimba et al., 2013; Wang et al., 2014).Most of the proteomic studies performed on non-

model microalgae are limited to the analysis of totalprotein extracts; therefore, the most abundant solubleproteins will be overrepresented. These approachesoverlook the role played by microalgal membraneproteins, which have remained understudied despitethe fact that the majority of lipid metabolism proteinshave been proposed to be membrane associated (Natteret al., 2005; Joyard et al., 2010;Wang and Benning, 2012)and that nitrogen starvation is known to exert ultra-structural changes in microalgal cells (Moellering andBenning, 2010), including membrane lipid remodeling,turnover, and degradation (Li et al., 2012, 2014; Yoonet al., 2012).In this study,microsomalmembranes fromE. oleoabundans

cells submitted to nitrogen deprivation were ana-lyzed via an RNA sequencing (RNA-Seq)-based pro-teomics approach (Wang et al., 2014) using theE. oleoabundans transcriptome generated by Rismani-Yazdi et al. (2012). To overcome the typical limitationsof membrane proteomics due to the heterogenous, hy-drophobic, and low-abundance nature of membraneproteins (Tan et al., 2008), a gel-free shotgun proteomicsstrategy was employed. In parallel, free-flow zonalelectrophoresis (FFZE), a liquid-based matrix-free sep-aration technology (Barkla et al., 2007; Wildgruberet al., 2014), was coupled to shotgun proteomics anduniquely employed to assess the intracellular locationof novel identified proteins as well as to provide a de-tailed survey of biological processes related to energyand carbon flux in nitrogen-deprived E. oleoabundans,giving insights into the basic biology of this organism.

Molecular phylogenetic analysis of proteins identified inthis work and of 18S ribosomal DNA (rDNA) raised con-cerns regarding the taxonomic status of E. oleoabundans, asthey support a close alliance between E. oleoabundans andspecies of the Trebouxiophyceae class that is contrary toits current classification.

RESULTS AND DISCUSSION

E. oleoabundans Membrane Proteome

To study themembrane proteome of the nonsequencedoleaginous microalga E. oleoabundans, microsomes from4-d nitrogen-deprived cultures were isolated and subse-quently analyzed using gel-free liquid chromatography-tandem mass spectrometry (LC-MS/MS).

As a first approach, product ion data were searchedagainst the Viridiplantae protein database (TaxID33090, unknown version; 677,107 entries) using theMascot search program (Matrix Science). A total of45 proteins (1,057 spectra) were identified with two ormore unique peptides, from which only 30 weredetected in at least two of four biological replicates. Thesmall number of identified proteins can be attributed tothe typical limitations of studying membrane proteins(Tan et al., 2008) but mostly to the lack of sequence datafor nonsequenced organisms (Ndimba et al., 2013;Wang et al., 2014), such as E. oleoabundans. To overcomethese limitations, an RNA-Seq-based proteomics strat-egy was established, using as a guide the E. oleoabundansde novo sequenced transcriptome (Rismani-Yazdi et al.,2012). The E. oleoabundans transcriptome comprises56,550 nonredundant transcripts and was obtainedfrom cells cultured under both nitrogen-replete andnitrogen-deprived conditions (Rismani-Yazdi et al.,2012). In order to generate an E. oleoabundans proteindatabase, an in silico six-frame translation of the tran-scriptome was performed, yielding 54,652 nonredun-dant putative protein sequences. The E. oleoabundansprotein database (unknown version; 53,921 entries) wasmerged with the Viridiplantae database and subs-equently used for peptide and protein identification.Using this approach, 551 proteins (18,902 spectra) wereidentified, from which only 404 complied with thestringency described above. This was a 13.5-fold in-crease in identifications over using only the Viridiplantaedatabase. This result shows that use of theE. oleoabundanstranslated transcriptome significantly improved pro-tein identification and highlights the advantages ofintegrating de novo transcriptomic and proteomicanalyses to study nonmodel microalgae (Guarnieriet al., 2011).

In order to describe the composition of theE. oleoabundansmembrane proteome, the sequences from the 404 identi-fied proteins were analyzed with transmembranehelix (HMMTOP version 2.0 and TMHMM version 2.0)and beta-barrel membrane protein (MCMBB andTMBETADISC-RBF) prediction programs (Krogh et al.,2001; Tusnády and Simon, 2001; Bagos et al., 2004; Ouet al., 2008). Proteinspredicted topossess a transmembrane

Plant Physiol. Vol. 173, 2017 391

Oleaginous Microalga Membrane Proteomics

www.plantphysiol.orgon December 26, 2019 - Published by Downloaded from Copyright © 2017 American Society of Plant Biologists. All rights reserved.


region by any of the four prediction programs were con-sidered as integral membrane proteins. This analysisdemonstrated that 57% of the E. oleoabundans membraneproteome is composed of integral membrane proteins; theremaining 43% can be classified as peripheral membraneproteins that do not transverse the membrane but may beassociated with the membrane surface to varying extents(Tan et al., 2008).

Functional Annotation of the E. oleoabundansMembrane Proteome

To address the biological significance of the identifiedmembrane proteins, we initially performed a functionalannotation based on sequence similarity using theBlast2GO suite (Götz et al., 2008). From the total of404 proteins, 391 returned a significant BLASTP match(E value cutoff # 0.001) against the National Center forBiotechnology Information (NCBI) nonredundant proteinsequences database. Themajority of theproteins (92%) hadbest-hit homologs in species from the trebouxiophyceanclass, whereas only 3% possessed best-hit homologs withmembers of the chlorophycean class (Supplemental Fig.S1). This result resembles those obtained from the analysisof the E. oleoabundans transcriptome (Rismani-Yazdi et al.,2012), suggesting a closer proximity of E. oleoabundans tothe trebouxiophycean Chlorella spp.

Although automatic annotations have been shown tobe more reliable than generally believed (Škunca et al.,2012) and there is a demonstrated good performance ofBlast2GO (Götz et al., 2008), functional misannotationin computational analysis remains a significant concern(Schnoes et al., 2009). To improve protein annotation,each of the 404 identified proteins were manually cu-rated as described in “Materials and Methods.”

Following manual curation, the cellular locationof 85% of the identified proteins was predicted(Supplemental Fig. S2A). Themajority of themembraneproteins (41%) were located in the chloroplast, notsurprising as this organelle occupies most of themicroalga’s cell volume (Giovanardi et al., 2013). De-spite this, most of the cellular compartments wererepresented in the membrane proteome, including thelipid droplets (LDs), which are structures induced bynitrogen deprivation (Davis et al., 2012; Popovich et al.,2012; Giovanardi et al., 2013). Cytoplasmic pro-teins comprised only 9% of the membrane proteome(Supplemental Fig. S2A). The majority of identifiedproteins (82%) were categorized into known biologicalprocesses (Supplemental Fig. S2B), which were grou-ped into three broad categories: protein and nitrogenmetabolism (28%), energy production and homeostasis(23%), and carbon metabolism (15%). All the identifiedproteins are described in Supplemental Tables S1 to S15and classified according to the biological process towhich they are related. The corresponding protein ho-mologs in the model organisms C. reinhardtii and Ara-bidopsis (Arabidopsis thaliana) also are indicated, as wellas the predictions and the experimental evidence for thecellular locations of these proteins.

Novel Occurrence of the Photosynthesis-Related Proteins,Photosystem II Subunit S and Maintenance ofPhotosystem II under High Light1, in Green Microalgae

Photosystem II Subunit S (PSBS; or 22-kD protein)and Maintenance of Photosystem II under High Light1(MPH1) are proteins related to well-characterizedphotoprotective responses that have been consideredexclusive to higher photosynthetic organisms.

In this work, the E. oleoabundans PSBS protein(EoPSBS; m.395061; Supplemental Table S1) was iden-tified in all four biological replicates with up to sevenunique peptides and a maximum coverage of 29% ofthe predicted protein sequence. EoPSBS is composedof 252 amino acids with an estimatedmolecular mass of28 kD considering the predicted chloroplast transitpeptide (CTP; Fig. 1A) and of 25 kD following tran-sit peptide cleavage. Similar to Arabidopsis PSBS(AtPSBS), EoPSBS comprises a chlorophyll a/b-bindingprotein domain (Fig. 1A), four membrane-spanninga-helices (Kyte-Doolittle [Fig. 1B] and HMMTOP ver-sion 2.0 [Fig. 1A]; Supplemental Fig S3A), and a pre-dicted N-terminal CTP (28 amino acids; PredAlgo; Fig.1A; Supplemental Fig. S3A). The two symmetricallyarranged lumen-exposed Glu residues that arenecessary for the PSBS pH-sensing mechanism and,thus, its function in land plants (Glu-122 and Glu-226 inAtPSBS; Li et al., 2002, 2004) are both conserved inEoPSBS (Glu-120 and Glu-224; Fig. 1A; SupplementalFig. S3A). EoPSBS homologs from higher plants andgreen algae species were retrieved from the UniProtKBdatabase (Fig. 1C; Supplemental Fig. S3A), althoughPSBS transcripts (Miller et al., 2010; Gerotto andMorosinotto, 2013) but not the corresponding protein(Bonente et al., 2008) had been identified in greenmicroalgae until very recently, when PSBS was detectedin C. reinhardtii upon high light acclimation (Correa-Galvis et al., 2016; Tibiletti et al., 2016). EoPSBS shares75% identity with the predicted protein from Chlorellavariabilis (class Trebouxiophyceae) and 46% with thecorresponding proteins from chlorophycean species; incontrast, only 30% identity is sharedwith PSBS from landplants (Supplemental Fig. S3B). Phylogenetic analysisshowed that PSBS is conserved along the green line-age (Viridiplantae); however, PSBS from green algae andland plants clustered into two distinct clades (Fig. 1C).This result suggests that PSBSwas present in the commonancestor of extant green algae and land plants but that itevolved separately in these two phylogenetic groups, ashas been suggested in previous evolutionary analyses ofgenomic and transcriptomic PSBS sequences (Koziol et al.,2007; Bonente et al., 2008; Gerotto andMorosinotto, 2013).

Vascular plants rely on PSBS for the pH-regulatedactivation of the energy-dependent feedback deexcita-tion component (qE) of nonphotochemical quenching(NPQ) for photoprotection (Li et al., 2000; Niyogi andTruong, 2013). In contrast, eukaryotic algae, except forred algae, cryptophytes (Dittami et al., 2010), andperidinin-containing dinoflagellates (Boldt et al., 2012),commonly depend on LHC-Like Protein Stress Related

392 Plant Physiol. Vol. 173, 2017

Garibay-Hernández et al.


http://www.plantphysiol.org/cgi/content/full/pp.16.01240/DC1














(LHCSR; LI818 or LHCSX) for qE induction (Peers et al.,2009; Bailleul et al., 2010). Despite pH regulation ofqE being restricted to LHCSR in C. reinhardtii, recentevidence has shown that its full NPQ capacity is also de-pendent on PSBS, whose substoichiometric accumulationis a prerequisite for further activation of the LHCSR-dependent qE mechanism (Correa-Galvis et al., 2016).These two mechanisms apparently overlapped at somepoint during evolution, as both proteins, LHCSR andPSBS, also have been identified in organisms that representtransitional states betweengreen algae andvascular plants,where, contrary to what has been found in C. reinhardtii,these proteins function independently and additively in qEregulation (Alboresi et al., 2010; Gerotto et al., 2012; Mouet al., 2013; Zhang et al., 2013). In view of this, we searched

for the presence of LHCSR in E. oleoabundans to determineif both qE effector proteins were present. An LHCSR ho-molog was not identified by searching the tran-scriptome (Rismani-Yazdi et al., 2012) and thecorresponding in silico-translated proteome of E.oleoabundans. The absence of LHCSR was further con-firmed by western-blot analysis of microsomes fromnitrogen-deprivedE. oleoabundans (Fig. 1D). The absence ofan LHCSR homolog in E. oleoabundans questions the con-servation of qE mechanisms within green microalgae.

MPH1 is a Pro-rich intrinsic thylakoid protein thatparticipates in the protection and stabilization of PSIIagainst photooxidative damage in Arabidopsis underhigh-light stress (Liu and Last, 2015a, 2015b). Identifi-cation of an MPH1 homolog in the membranes of

Figure 1. Analysis of qE effector pro-teins (PSBS and LHCSR) in nitrogen-depleted E. oleoabundans. A, Proteinarchitecture of EoPSBS. The identifiedprotein domain signatures (InterPro;yellow), the predicted CTP (PredAlgo;green), and the transmembrane domains(TM; HMMTOP version 2.0; blue) areindicated. The conserved residues in-volved in the PSBS pH-sensing mech-anism are shown (red circles). B,Hydrophobicity comparison of PSBSfrom E. oleoabundans (Eol) and Arabi-dopsis (Ath). Probable transmembranedomains (values greater than 0) areshown in the Kyte-Doolittle hydropho-bicity plots (window size, 19). C, Phy-logenetic analysis of PSBS homologs.Aligned sequences (Supplemental Fig.S3) were submitted for maximum like-lihood (ML) analysis. The topology of theML tree with the highest log likelihood(23,122.6386) is shown. Bootstrapmaximum likelihood (MLb) values areshown next to the branches. UniProtKBaccession numbers for PSBS homologsare provided. D, Immunological detec-tion of LHCSR in microsomes fromE. oleoabundans and C. reinhardtii(Cre). The 12.5% (w/v) SDS-PAGE ac-rylamide gel was loaded with 20 mg ofprotein per lane.







nitrogen-deprived E. oleoabundans (EoMPH1;m.337748;Supplemental Table S1) also was unexpected, sinceMPH1 has been reported as a protein specific to landplants (Liu and Last, 2015a, 2015b). In this study,EoMPH1 was identified in three out of four biologicalreplicates, with up to three unique peptides and amaximum coverage of 19% of the predicted proteinsequence. EoMPH1 is composed of 164 amino acidswith an estimated molecular mass of 16.2 kD consid-ering the CTP (Fig. 2A) and of 13.6 kD following itscleavage. It shares 15% to 20% sequence identity withMPH1 sequences from higher plants (Supplemental Fig.S4B), including structural features predicted for AtMPH1

(Liu and Last, 2015b): a single transmembrane domain(Kyte-Doolittle [Fig. 2B] and TMHMM version 2.0 [Fig.2A]; Supplemental Fig. S4A), an N-terminal CTP(25 amino acids; PredAlgo; Fig. 2A; Supplemental Fig.S4A), and a high Pro content (6% of the protein), fromwhich some interspersed Pro residues are conserved (Fig.2A; Supplemental Fig. S4A). Based on the presence ofMPH1 in E. oleoabundans, we searched for unreported ho-mologs from other green microalgae species. The resultsidentified predicted sequences from C. variabilis (classTrebouxiophyceae) andC. reinhardtii (classChlorophyceae),which shared 68% and 28% identity, respectively, withEoMPH1 (Supplemental Fig. S4). Phylogenetic analysis

Figure 2. MPH1 sequence analysis andidentification in green microalgaemembranes. A, Protein architecture ofEoMPH1. The predicted CTP (PredAlgo;green) and transmembrane domain (TM;TMHMM version 2.0; blue) are indi-cated. Conserved Pro residues areshown (red circles). B, Hydropho-bicity comparison of MPH1 fromE. oleoabundans (Eol) and Arabidopsis(Ath). Probable transmembrane do-mains (values greater than 1.6) areshown in the Kyte-Doolittle hydropho-bicity plots (window size, 19). C, Phy-logenetic analysis of MPH1 homologs.Aligned sequences (Supplemental Fig.S4) were submitted for ML analysis. Thetopology of the ML tree with the highestlog likelihood (22,177.6951) is shown.MLb values are shown next to thebranches. UniProtKB accession num-bers for MPH1 homologs are provided.D, Immunological detection of MPH1 inmicrosomes from Arabidopsis (positivecontrol; red box), E. oleoabundans, andC. reinhardtii (Cre). The 12.5% (w/v)SDS-PAGE acrylamide gel was loadedwith 20 mg of protein per lane.















showed that MPH1 sequences from green algae and landplants clustered into two distinct clades (Fig. 2C). Toconfirm the presence of MPH1 in membranes of greenmicroalgae, microsomes from E. oleoabundans and C. rein-hardtii were probed using a polyclonal antibody raisedagainst AtMPH1 (Liu and Last, 2015b). A single band ofapproximately 25 kD was identified on the blot in bothcases (Fig. 2D). Considering that both homologs have apredicted molecular mass close to 13 kD, this result sug-gests thatmicroalgalMPH1may formdimers. Contrary tothis finding, a single band corresponding to the AtMPH1predicted molecular mass (20 kD) was identified (Fig. 2D)in Arabidopsis membranes, whose low intensity may beattributed to its low abundance.

Retinitis Pigmentosa Type 2-Clathrin Light Chain, aNovel Domain Architecture Protein Identified inE. oleoabundans Membranes

The proteomic analysis of E. oleoabundansmembranesresulted in the detection of several unknown proteins.Among them,we identified proteinm.397619 (SupplementalTable S9), which presents a domain architecture that hasnot been described previously. This protein comprises

368 amino acids and contains an N-terminal region cor-responding to the Retinitis Pigmentosa Type 2 (RP2)protein family and a C terminus that comprises a Clath-rin Light Chain (CLC) domain (Fig. 3A). Accordingly, wenamed this protein EoRP2-CLC. A survey of proteinswith this architecture in the UniProtKB database showedthat they are limited to certain unicellular eukaryotes,such as green microalgae from both TrebouxiophyceaeandChlorophyceae classes, several ciliated protozoa, andmold species from theOomycetes class, but are not presentin higher eukaryotes.

EoRP2-CLC is a Tubulin-Binding Cofactor C (TBCC)domain-containing protein, as it comprises a predictedTBCC-like domain within the N-terminal RP2 region(Fig. 3A). Three protein families with TBCC domains(Fig. 3B) have been described (Stephan et al., 2007). Thefirst (clade 1/TBCC) is the canonical TBCC, which isessential for de novo native a/b-tubulin heterodimerformation by stimulating GTP hydrolysis in b-tubulin(Lundin et al., 2010); this clade comprises proteins from adiverse range of eukaryotes. The second (clade 2/RP2)contains homologs of human RP2, which are apparentlyrestricted to eukaryotes capable of forming cilium/flagellum (Stephanet al., 2007). The third (clade3/TBCCd1)

Figure 3. EoRP2-CLC is a membraneTBCC domain-containing protein with anovel domain architecture. A, Proteinarchitecture of EoRP2-CLC. The identi-fied InterPro signatures and the Lys res-idue (red circle) that may correspond toa homologous substitution of the keyArg residue for GAP activity are indi-cated. B, ML analysis of TBCC domain-containing proteins. Clade 1/TBCC is inblue, clade 2/RP2 is in red, and clade 3/TBCCd1 is in yellow. RP2-CLC domainarchitecture proteins are highlighted inred text. Amino acid sequences werealigned with webPRANK. The topologyof the ML tree with the highest log like-lihood (28,226.6668) is shown. MLbvalues are shown next to the branches.Accession numbers are provided: Uni-ProtKB (Hsa, Tbr, and Cva), Phytozomeversion 10 (Cre and Vca), and The Ara-bidopsis Information Resource (Ath).Ath, Arabidopsis; Cre, C. reinhardtii;Cva, C. variabilis; Hsa, H. sapiens; Eol,E. oleoabundans (red dots); Tbr, Try-panosoma brucei; Vca, Volvox carteri.C, Immunological detection of RP2-likeproteins in microsomes from C. rein-hardtii and E. oleoabundans. The 10%(w/v) SDS-PAGE acrylamide gel wasloaded with 20 mg of protein per lane.Total protein extracts (10 mL) from hu-man (Hsa) C2BBe1 cells (clone of Caco-2) were analyzed as a positive control,where the 40-kD RP2 human proteinwas identified.







comprises noncanonical TBCC domain-containingproteins, which lack a conserved catalytic Arg respon-sible for GTPase-activating protein (GAP) activity(Bartolini et al., 2002). However, an Arg residue locatedclose to the Arg finger position in TBCC and RP2-likeproteins (Supplemental Fig. S5) has been suggested tosuffice for GAP activity in TBCCd1 proteins (Feldman andMarshall, 2009). E. oleoabundans has predicted protein ho-mologs for each clade (Fig. 3B); however, only EoRP2-CLCwas identified in the membrane proteome. EoRP2-CLC,together with the protein homolog from C. variabilis(UniProtKB no. E1ZN89), which is characterized byautosporic reproduction (Huss et al., 1999), clusteredwithin theRP2 clade (Fig. 3B). Thus, theRP2 clademaynotbe restricted to cilium/flagellum-forming eukaryotes, asproposed previously (Stephan et al., 2007). Sequenceanalysis showed that only the TBCC domains fromE. oleoabundans and C. variabilis RP2 sequences do notpresent the conserved catalytic Arg but instead showed ahomologous substitution with a Lys residue (Fig. 3A;Supplemental Fig. S5),whichmay suffice forGAPactivity.

The presence of RP2-CLC proteins in the microsomesfrom both E. oleoabundans and C. reinhardtii was con-firmed using a polyclonal antibody raised against Homosapiens RP2 (HsRP2; Fig. 3C). A single protein band witha molecular mass slightly higher than that predicted forEoRP2-CLC (39 kD) was identified in E. oleoabundans. InC. reinhardtii, two protein bands around the molecularmass predicted for CrRP2-CLC (Cre06.g293250.t1.1;58 kD) were detected, where one of these bands maycorrespond to a posttranslationally modified CrRP2-CLC, similar to HsRP2 that is known to be subjected todual N-terminal acylation (Chapple et al., 2000).

RP2-CLCproteinshaveadomainarchitecture thatdiffersfrom currently characterized RP2 proteins and are presentin both flagellated and nonflagellated microalgae (Fig. 3B),suggesting that they may be involved in other noncilia/flagella-specific functions. Contrary to microalgae of theChlorophyceae class, no obvious homologs to currentlyknown CLC proteins were identified in either theE. oleoabundans in silico-translated protein database or thepredictedproteins forC.variabilis (pico-PLAZA;Vandepoeleet al., 2013). CLC also has been shown to be absent in otherunicellular eukaryotes, such as Cyanidioschyzon merolae(Misumi et al., 2005), Entamoeba histolytica, and Giardialamblia (Manna et al., 2015); however, this might be due tothe high divergence of CLC sequences among eukaryotes(Wang et al., 2003). The existence ofNoRP2-CLCas theonlycandidate for a CLC-harboring protein in E. oleoabundans,together with current evidence that supports a role ofHsRP2 in post-Golgi trafficking (Evans et al., 2010), sug-gests that EoRP2-CLC may play a role in the formation/traffickingof clathrin-coatedvesicles inE. oleoabundans cells.

Determining the Subcellular Location of Novel Proteinsvia FFZE Membrane Fractionation Coupled to MassSpectrometry-Based Analysis

To assess the subcellular locations of the novelmicroalgal proteins identified in this work, we

employed a membrane fractionation approach. In thisstudy, we avoided traditional fractionation techniquesby using FFZE, a liquid-based high-resolution mem-brane separation technique based on surface chargethat has proved useful for subcellular proteome samplepreparation (Barkla et al., 2007; Wildgruber et al., 2014;de Michele et al., 2016).

E. oleoabundans microsomes from nitrogen-deprivedcultures were separated into 96 individual FFZE frac-tions (Fig. 4A), and each fraction was subjected to eitherdirect chlorophyll measurements (Fig. 4B) or western-blot analysis against protein markers for both thechloroplast and the plasma membrane (Fig. 4C). Chlo-roplast membranes were detected in fractions 51 to 58,while more positively charged fractions (56–67) com-prised the plasma membrane (Fig. 4C). IndividualFFZE fractions, from 45 to 67, were then subjected toshotgun proteomics analysis to gain a more compre-hensive overview of the protein profile of each of thesefractions. A spectral counting-based quantitative ap-proach, expressed in terms of the normalized spectralabundance factor (NSAF; Zhang et al., 2010), was usedto determine the distribution of proteinmarkers specificfor different subcellular compartments among theanalyzed FFZE fractions (Fig. 4D). Similar to theprotein-blot analysis, the mass spectrometry (MS)-based analysis confirmed the presence of two differentmembrane populations but showed that the chloroplastfractions also were enriched in mitochondrial mem-branes (fractions 45–58), whereas the plasma mem-brane fractions comigrated with vacuolar membranes(fractions 56–67; Fig. 4D).

The MS-based analysis of the FFZE fractions wasfurther employed to assess the subcellular location ofEoPSBS and EoMPH1. Peptides for these proteins weredetected, and their distribution profiles within theFFZE fractions were mapped with those from knownmarker proteins. These two photosynthesis-relatedproteins presented similar distributions to that shownby the chloroplast-enriched membrane fractions (Fig.4E). These results, together with the evidence that bothproteins possess a predicted N-terminal CTP (Figs. 1Aand 2A), confirmed that EoPSBS and EoMPH1 arechloroplastic membrane proteins, as has been demon-strated for their land plant homologs (Li et al., 2000;Ferro et al., 2010; Liu and Last, 2015b). Similar analysisfor the distribution of EoRP2-CLC among the FFZEfractions showed that this protein was more abundantin the plasmamembrane-enriched fractions (Fig. 4E), anobservation that agrees with the RP2 plasmamembranelocalization in vertebrates (Chapple et al., 2002;Grayson et al., 2002).

Molecular Phylogenetic Analysis of Identified ProteinsPlaces E. oleoabundans within Trebouxiophycean Algae

E. oleoabundans is a coccoid green microalga with acomplex taxonomic history. It was initially classifiedwithin the Neochloris genus (Sphaeropleales) but laterreclassified into the Ettlia genus (Chlamydomonadales),







Figure 4. Subcellular locations of novel proteins from E. oleoabundans via FFZE fractionation coupled to MS-based analysis.Microsomal membranes from nitrogen-deficient cultures were separated by FFZE. A, Protein profile of FFZE fractions. OD280,Optical density at 280 nm. B, Chlorophyll a and b concentrations in FFZE fractions. C, Immunological detection in the respectivefractions of ATPb (a chloroplast marker) and H+-ATPase (a plasma membrane marker). The 10% (w/v) SDS-PAGE acrylamide gel





which comprised uninucleate cells with thin-walledzoospores (Deason et al., 1991; Guiry and Guiry, 1996).Despite that most researchers commonly refer to thismicroalga as Neochloris oleoabundans, it is currently clas-sified as Ettlia oleoabundans, which is placed within theChlorophyceae class according to the classification of theEttlia genus type species, Ettlia carotinosa (Guiry andGuiry, 1996; Pegg et al., 2015).

The results from this work raised concerns regardingthe taxonomic status of E. oleoabundans. Most of theidentified proteins (greater than 90%) had best-hit ho-mologs in species from the trebouxiophycean class(Supplemental Fig. S1), and phylogenetic analysis ofidentified proteins, including PSBS (Fig. 1C), MPH1(Fig. 2C), TBCC, RP2-CLC (Fig. 3B), and enolase(Supplemental Fig. S7), showed a closer relationship be-tween E. oleoabundans andC. variabilis (Trebouxiophyceae)rather than with chlorophycean species. Moreover,evidence for zoospore formation in several Ettlia spp. islacking (Yoo et al., 2013), including E. oleoabundans, forwhich only autosporic reproduction has been observed,similar to Chlorella spp. of the trebouxiophycean class(Huss et al., 1999).

To assess the taxonomy of E. oleoabundans, a molec-ular phylogenetic analysis based on a multigene ap-proach was performed. This increased the power ofdiscrimination and robustness of the phylogeneticanalysis compared with single-gene analysis (Moreiraet al., 2000; Gontcharov et al., 2004; Tippery et al., 2012).Six proteins proposed as microalgal phylogeneticmarkers (Moreira et al., 2000; Tippery et al., 2012;Wei et al., 2013) were chosen, two per genome(Supplemental Table S16): nucleus-encoded Actin andElongation Factor1-a (EF-1a); plastid-encoded Photo-system II D1 (PSBA) and Rubisco Large Chain (RBCL);and mitochondria-encoded Cytochrome Oxidase Sub-unit1 (COX1) and COX2. Amino acid sequences rep-resentative for the three green microalgae classes,Chlorophyceae, Trebouxiophyceae, and Prasinophy-ceae (outgroup), were retrieved, aligned, and concate-nated. Phylogenetic analysis of the concatenatedmarkers placed E. oleoabundans as a close relative of theChlorella spp. within the Trebouxiophyceae class (Fig.5A), a relationship that was strongly supported by thebootstrap value from the maximum likelihood analy-sis (MLb = 0.79). The relationship of E. oleoabundanswith trebouxiophycean algae, rather than with thechlorophycean class, is reinforced by the analysisof E. oleoabundans COX2 (EoCOX2; m.110997;Supplemental Table S2). E. oleoabundans contains anorthodox intact mitochondria-encoded COX2, identi-fied in this study by a single polypeptide, which lacks a

predicted N-terminal mitochondrial targeting sequence(PredAlgo). This protein showed around 80% identityto orthodox COX2 homologs from members of thetrebouxiophycean class (Supplemental Fig. S6B). Phy-logenetic analysis confirmed the close relationship ofEoCOX2 with orthodox COX2 from Chlorella spp.rather than with homologs from chlorophycean algae,which clustered into a different clade (SupplementalFig. S6A). EoCOX2 clearly differs from its homologs inchlorophycean algae, characterized by exhibiting anatypical COX2 heterodimer as a consequence of alineage-specific fragmentation and nuclear relocation ofthe mitochondrial COX2 gene (Pérez-Martínez et al.,2001; Rodríguez-Salinas et al., 2012).

Due to the reduced availability of completely se-quenced microalgal genomes, the multigene approachwas performed with a reduced taxon sampling. Toimprove the accuracy of the analysis, we performed an18S rDNA phylogenetic analysis with increased taxonsampling, which included three independent 18S par-tial sequences for E. oleoabundans, one of them obtainedin this work (Fig. 5B; Supplemental Table S17). Twomajor taxa within the Chlorophyceae class, wherespecies of the Neochloris and Ettlia genera are currentlyclassified (Sphaeropleales and Chlamydomonadales),were highly represented. Taxon sampling also was in-creased for the diverse trebouxiophycean class. The 18Sphylogeny confirmed the extremely close relation-ship between E. oleoabundans and trebouxiophyceaenmicroalgae, as all three 18S E. oleoabundans sequencesclustered together within this class, particularly in thewell-defined Chlorellales lineage supported by a robustbootstrapping score (MLb = 0.76; Fig. 5B). Other Ettliaspp. considered in this analysis were still placed inseveral groups within the chlorophycean class, sup-porting previous concerns regarding the classificationof the species from this genus (Pegg et al., 2015).

Altogether, our results provide compelling evidence forreclassifying E. oleoabundans into the trebouxiophyceanclass, close to the Chlorellales lineage, and indi-cate that it is not closely related to chlorophyceanmicroalgae.

Lipid Metabolism Represented in the E. oleoabundansMembrane Proteome

E. oleoabundans has shown potential for biotechno-logical applications, as its lipid acyl chains are consid-ered an energy-rich feedstock for the production ofbiofuels and value-added compounds (Hu et al., 2008;Garibay-Hernández et al., 2013; Liu and Benning, 2013).Analysis of the E. oleoabundans membrane proteome

Figure 4. (Continued.)was loaded with 15 mg of protein per lane. The approximate molecular masses of the detected proteins are shown. D, Graphicalrepresentation of the normalized spectral count (NSAF values) of protein markers specific for subcellular compartments amongthe FFZE fractions. Individual FFZE fractions were analyzed by LC-MS/MS. The identification numbers for the surveyed proteinmarkers are as follows: gi|416678 (ATPb), m.392881 (ATPg), m.378383 (VDAC), m.363780 (H+-ATPase), and m.395664 (VHA-B). E, Graphical representation of the NSAF values of PSBS, MPH1, and RP2-CLC among the FFZE fractions.













Figure 5. Molecular phylogenetic analysis of E. oleoabundans. A, ML analysis of concatenated nucleus-encoded (EF-1a andActin), plastid-encoded (PSBA and RBCL), and mitochondria-encoded (COX1 and COX2) amino acid sequences (SupplementalTable S16). The topology of the ML tree with the highest log likelihood (220,458.0427) is shown. B, ML analysis of 18S rDNAnucleotide sequences (Supplemental Table S17); Ettol2 corresponds to the sequence obtained in thiswork. The topology of theMLtree with the highest log likelihood (212,721.9945) is shown. MLb values are shown next to the branches. E. oleoabundanssequences are highlighted (red dots). Green microalgae classes are denoted as follows: Chlorophyceae (red), Trebouxiophyceae(blue), and Prasinophyceae (gray; outgroup). Major taxa represented within these classes are denoted in B.








under lipid accumulation conditions (nitrogen depri-vation) presents an opportunity to study proteins re-lated to lipid metabolism, since many of them aremembrane associated (Natter et al., 2005; Joyard et al.,2010; Wang and Benning, 2012). Table I lists the pro-teins involved in acetyl-CoA synthesis and lipid me-tabolism identified in this study.

Acetyl-CoA Synthesis

The direct carbon precursor for de novo fatty acidsynthesis in photosynthetic organisms is plastidicacetyl-CoA,which is synthesized directly by the activityof the chloroplastic pyruvate dehydrogenase (PDH)complex via the oxidative decarboxylation of glycolysis-derived pyruvate (Shtaida et al., 2015). We identifiedthe four subunits of the chloroplastic PDH complex inthe membranes of nitrogen-deprived E. oleoabundans(E1a, E1b, E2, and E3; Table I; Fig. 6). Cytoplasmicproduction of acetyl-CoA also was represented bythe identification of the ATP-citrate synthase a- andb-subunits (Table I; Fig. 6), whose cytoplasmic locationhas been demonstrated in Arabidopsis (Fatland et al.,2002). ATP-citrate synthase has been proposed as a keyenzyme for lipid accumulation inmammals, oleaginousyeast, fungi (Courchesne et al., 2009), and C. reinhardtii(Wase et al., 2014).

Lipid Metabolism

The membrane proteome of E. oleoabundans wascomposed of 3.5% of proteins related to lipid metabo-lism (Table I). A similar amount of lipid metabolism-related proteins (Table I, proteins highlighted withasterisks) were additionally identified in only one rep-licate of the total microsomal samples and/or throughMS analysis of the FFZE fractions where sample com-plexity was reduced. This demonstrates that the lowabundance of lipid biosynthetic proteins hindered theiridentification in E. oleoabundans membranes.

Among the proteins identified in E. oleoabundansmembranes, the committed step for fatty acid biosyn-thesis catalyzed by the heteromeric acetyl-CoA car-boxylase (ACC; Stern, 2009) was represented. TwoACC chloroplast-targeted components, a-carboxyltransferase and biotin carboxylase (Table I), wereidentified in all four biological replicates of total mi-crosomal samples, whereas the chloroplast-encodedb-carboxyl transferase subunit (Table I) was detectedexclusively through MS analysis of the FFZE fractions.The fourth ACC component, biotin carboxyl carrierprotein, was not detected, probably due to its lowabundance and low molecular mass (26 kD) predictedfrom its transcript (Rismani-Yazdi et al., 2012). Incom-plete detection of the ACC constituents ismore frequentthan expected, as only one or two subunits have beenidentified in several proteomics studies performed inArabidopsis (Ferro et al., 2003; Froehlich et al., 2003;Kleffmann et al., 2004; Peltier et al., 2006) and C. rein-hardtii (Bienvenut et al., 2011; Schmollinger et al., 2014).

The chloroplastic fatty acid synthesis also was repre-sented in thiswork by the identification of 3-hydroxyacyl-ACP dehydratase and enoyl-ACP reductase (Table I),which are components of the multipartite (type II) fattyacid synthase complex (Li-Beisson et al., 2015). In agree-ment with the existence of very-long-chain fatty acids(acyl chain length beyond 18C) in nitrogen-deprived E.oleoabundans (Tornabene et al., 1983; Garibay-Hernándezet al., 2013;Matich et al., 2016), we identified a homolog ofthe b-ketoacyl-CoA reductase (Table I), a component ofthe endoplasmic reticulum-bound multienzymatic fattyacid elongase complex (Haslam and Kunst, 2013). Threelong-chain acyl-CoA synthetase isoforms (Table I), re-quired for the activation of free fatty acids to acyl-CoAthioesters (Li-Beisson et al., 2015), also were identified.

Glycerolipid metabolism was represented inE. oleoabundans microsomes by glycerol-3-phosphateacyltransferase (GPAT) and 1-acyl-sn-glycerol-3-phosphate acyltransferase (Table I), which catalyzethe first two reactions common to glycerolipid synthe-sis leading to phosphatidic acid formation (Li-Beissonet al., 2015). The GPAT identified in this work is a ho-molog of aC. reinhardtii LD-associated protein (Nguyenet al., 2011) but also of plant GPAT9 proteins (Shockeyet al., 2016) and, thus, may be involved in E. oleoa-bundans triacylglycerol biosynthesis, as has been dem-onstrated for its homologs in Parietochloris incisa(Trebouxiophyceae) and Arabidopsis (Iskandarov et al.,2016; Shockey et al., 2016; Singer et al., 2016). Regardingthemetabolismofmembrane lipids, twophosphoinositide(phosphorylated derivatives of phosphatidylinositol) sig-naling proteins were identified in E. oleoabundans mem-branes: phosphatidylinositol 4-kinase-a and Sac1p-likephosphoinositide phosphatase (Table I). These proteinsare likely to be involved in the ultrastructural changesrequired for LD formation in nitrogen-deprived E.oleoabundans (Giovanardi et al., 2013). Evidence showsthat phosphatidylinositol levels are responsive tonitrogen depletion in E. oleoabundans (Matich et al.,2016), and changes in phosphoinositide dynamics havebeen observed in other microalgae species under envi-ronmental stress (Einspahr et al., 1988; Heilmann et al.,2001). Moreover, homologs of both phosphoinositidesignaling proteins have been identified in C. reinhardtiiLDs (Moellering and Benning, 2010; Nguyen et al.,2011), and a link between phosphoinositide signalingand LD homeostasis has been demonstrated in yeast(Saccharomyces cerevisiae; Ren et al., 2014).

In accordancewith LD formation in nitrogen-deprivedE. oleoabundans (Popovich et al., 2012; Giovanardi et al.,2013), homologs for LD structural proteins were detectedin this work (Table I), includingmembers of the probablePlastid-Lipid Associated Protein family (Singh andMcNellis, 2011), a rubber (Hevea brasiliensis) elongationfactor protein (Horn et al., 2013; Berthelot et al., 2014),and a putative Major Lipid Droplet Protein (MLDP) thatshares around 25% identity with currently character-ized MLDPs from Haematococcus pluvialis (Peled et al.,2011),C. reinhardtii (Moellering and Benning, 2010), andDunaliella spp. (Davidi et al., 2012).





Tab

leI.

Ace

tyl-CoAan

dlipid

metab

olism

proteinsiden

tified

inmem

branes

ofnitroge

n-dep

letedE.

oleoab

undan

s

Ace

tyl-CoAan

dlipid

metab

olism

proteinsiden

tified

withat

leasttw

ouniquepep

tides

intw

oormore

biologica

lreplica

tesoftotalmicrosomal

mem

branesamplesaredescribed

.Very-low-

abundan

ceproteinsthat

wereiden

tified

exclusively

inonebiologicalreplicate

and/orin

FFZEmem

branefrac

tionsarehighligh

tedwithasterisks.Theca

lculatedmolecu

larmassesareshown,

toge

ther

withtheco

rrespondingprotein

homologs

inthemodel

organ

ismsC.reinhardtii(Cr)an

dArabidopsis(At)an

dtheirco

rrespondingperce

ntage

iden

tity

values

(%ID

).Commonab

-breviations(Abbr.)oftheiden

tified

proteinsareprovided

.Su

bcellularloca

liza

tionswerepredicted(Pr)usingPredAlgo.Curatedce

llloca

tions(Cu)wereestablished

acco

rdingto

theprotein

homologs.Referen

cesareprovided

forprotein

homologs

whose

cellloca

tionshavebee

ndem

onstratedex

perim

entally(N

A,notavailable).C,Chloroplast;Cy,cytoplasm

;FA

E,fattyac

idelonga

seco

mplex;

FAS,

fattyac

idsynthaseco

mplex;

M,mitoch

ondria;

O,other;PM,plasm

amem

brane;

PX,peroxisome;

SP,sign

alpep

tide.

Protein

Iden

tifier

CuratedDescription

Abbr.

Molecu

lar

Mass

ECNo.

BestCrHomolog

BestAtHomolog

SubcellularLo

calization

JGIVersion

5.5

Iden

tifier

%ID

TheArabidopsis

Inform

ationResource

Version10Iden

tifier

%ID

Pr

Cu

Referen

ce

Acetyl-CoAsynthesis

m.73526

Pyruvate

deh

ydroge

naseE1

componen

tsubunita,

chloroplastic*a

PDH

47

1.2.4.1.

Cre02.g099850.t1.1

75

AT1G01090

70

CCb

Terashim

aet

al.(2010)

E1a

m.371437

Pyruvate

deh

ydroge

naseE1

componen

tsubunitb,

chloroplastic

PDH

41

1.2.4.1.

Cre03.g194200.t1.2

86

AT1G30120

69

CCb

Terashim

aet

al.(2010)

E1b

m.130266

Pyruvate

deh

ydroge

naseE2

componen

t,ch

loroplastic

PDH

E250

2.3.1.12

Cre03.g158900.t1.2

74

AT3G25860

54

CCb

Terashim

aet

al.(2010)

m.324850

Pyruvate

deh

ydroge

naseE3

componen

t,ch

loroplastic

PDH

E363

1.8.1.4.

Cre01.g016514.t1.1

72

AT3G16950

60

CCb

Terashim

aet

al.(2010)

m.396299

ATP-citrate

synthasea-chain

protein*c

ACLA

47

2.3.3.8.

Cre05.g241850.t1.2

62

AT1G10670

65

SPCyb

Fatlan

det

al.(2002)

m.391431

ATP-citrate

synthaseb-chain

protein

ACLB

76

2.3.3.8.

Cre02.g088600.t1.2

77

AT5G49460

75

OCyb

Fatlan

det

al.(2002)

Lipid

metab

olism

Acyl-lipid

biosynthesis

m.212887

Ace

tyl-CoAca

rboxy

lase

carboxy

ltran

sferasesubunit

a,ch

loroplastic

ACC

60

6.4.1.2.

Cre12.g519100.t1.2

73

AT2G38040

54

CCb

Terashim

aet

al.(2010)

a-CT

gi|3023244

Ace

tyl-CoAca

rboxy

lase

carboxy

ltran

sferasesubunit

b,ch

loroplastic*a

ACC

47

6.4.1.2.

Cre12.g484000.t1.1

75

ATCG00500

61

OCb

Terashim

aet

al.(2010)

b-CT

m.390372

Biotinca

rboxy

lase,

chloroplastic

ACCBCR

61

6.3.4.14

6.4.1.2.

Cre08.g359350.t1.2

80

AT5G35360

74

CCb

Terashim

aet

al.(2010)

m.332618

3-H

ydroxyacyl-[acyl-carrier-

protein]deh

ydratase,

chloroplastic*a

FASHAD

25

4.2.1.59.

Cre03.g208050.t1.2

70

AT5G10160

54

MCb

Terashim

aet

al.(2010)

m.179214

Enoyl-[acyl-carrier-protein]

reductase[N

ADH],

chloroplastic*c

FASEN

R39

1.3.1.9.

Cre06.g294950.t1.1

82

AT2G05990

76

CCb

Terashim

aet

al.(2010)

m.84123

b-Ketoacyl-CoAreductase*

cFA

EKCR

39

1.1.1.330.

Cre09.g392430.t1.1

32

AT1G67730

43

SPEb

Bea

udoin

etal.(2009)

m.368848

Long-ch

ainacyl-CoA

synthetaseA

LCS

71

6.2.1.3

Cre13.g566650.t2.1

58

AT4G23850

48

OLD

bMoelleringan

dBen

ning(2010)

m.224985

Long-ch

ainacyl-CoA

synthetaseB

LCS

76

6.2.1.3

Cre13.g566650.t2.1

47

AT4G11030

43

SPLD

bMoelleringan

dBen

ning(2010)

(Tab

leco

ntinues

onfollowingpage.)





Tab

leI.(Continued

from

previouspag

e.)

Protein

Iden

tifier

CuratedDescription

Abbr.

Molecu

lar

Mass

ECNo.

BestCrHomolog

BestAtHomolog

SubcellularLo

calization

JGIVersion

5.5

Iden

tifier

%ID

TheArabidopsis

Inform

ationResource

Version10Iden

tifier

%ID

Pr

Cu

Referen

ce

m.371326

Long-ch

ainacyl-CoA

synthetaseC

LCS

67

6.2.1.3

Cre12.g507400.t1.2

63

AT5G27600

53

MM

dNA

Isoprenoid

biosynthesisviathemevalonatepathway

m.149328

Ace

tyl-CoAac

etyltran

sferase*

aACAT

51

2.3.1.9.

Cre02.g146050.t1.2

64

AT5G47720

59

MCyb

Carrieet

al.(2007)

Glycerolipid

biosynthesis

m.241864

Glycerol-3-phosphate

acyltran

sferase*

eGPA

T54

2.3.1.15.

Cre06.g273250.t1.2

55

AT5G60620

60

OEb,

LDb

Giddaet

al.(2009);

Ngu

yenet

al.(2011)

m.357823

1-Acyl-sn-glyce

rol-3-phosphate

acyltran

sferase,

chloroplastic*a

LPAAT

37

2.3.1.51.

Cre09.g398289.t1.1

63

AT4G30580

56

CCb,

LDb

Ferroet

al.(2010);

Ngu

yenet

al.(2011)

m.250190

Acyltransferasefamilyprotein*c

54

NA

NA

NA

NA

NA

CU

NA

Lipid

sign

aling

m.34057

Phosphatidylinositol4-kinase

a*a

PI4Ka

213

2.7.1.67.

Cre05.g245550.t1.1

45

AT1G49340

39

OPM

b,

LDb

Ngu

yenet

al.(2011);

Zhan

gan

dPeck

(2011)

m.226782

Sac1

p-likephosphoinositide

phosphatase*

aSA

C1

69

3.1.1.-

Cre09.g388750.t1.2

43

AT3G51460

39

OEb,

LDb

Despreset

al.(2003);

Ngu

yenet

al.(2011)

LDstructuralproteins

m.413736

Majorlipid

dropletprotein

MLD

P28

NA

Cre12.g491550.t1.2

24

NA

NA

OLD

bDavidiet

al.(2012)

m.392627

Probab

leplastid-lipid

associated

protein

A,

chloroplastic

PLA

P59

NA

Cre07.g325736.t1.1

42

AT5G19940

32

MCb

Terashim

aet

al.(2010)

m.50827

Probab

leplastid-lipid

associated

protein

B,

chloroplastic

PLA

P42

NA

Cre03.g189300.t1.1

44

AT4G04020

48

MCb

Ferroet

al.(2010)

m.244306

Probab

leplastid-lipid

associated

protein

C,

chloroplastic

PLA

P23

NA

Cre03.g188650.t1.2

62

AT3G26070

58

CCb

Terashim

aet

al.(2010)

gi|132270

Rubber

elonga

tionfactor

protein*a

REF

15

NA

NA

NA

AT3G05500

48

OLD

bHorn

etal.(2013)

Lipid

trafficking

m.216464

Mem

brane-associated

30-kD

protein,ch

loroplastic

VIPP1

33

NA

Cre13.g583550.t1.2

54

AT1G65260

54

CCb

Nordhues

etal.(2012)

m.362261

Protein

trigalac

tosyldiacylglyce

rol2,

chloroplastic

TGD2

44

NA

Cre16.g694400.t1.2

57

AT3G20320

40

OCb

Terashim

aet

al.(2010)

m.116135

Phospholipid-transporting

ATPase*a

ALA

159

3.6.3.1.

Cre16.g656500.t1.1

32

AT1G59820

33

OEb,

PM

bPoulsen

etal.(2008);

Mitra

etal.(2009)

m.417181

ABCtran

sporter

Gfamily

mem

ber

AABCG

70

3.6.3.-

Cre07.g313250.t1.2

57

AT3G55100

31

OPM

dNA

m.306564

ABCtran

sporter

Gfamily

mem

ber

CABCG

69

3.6.3.-

Cre07.g313250.t1.2

55

AT3G55100

29

OPM

dNA

(Tab

leco

ntinues

onfollowingpage.)





We also identified proteins involved in lipid traf-ficking (Table I), including the membrane-associated30-kD Vesicle Inducing Protein in Plastids (VIPP1),the trigalactosyldiacylglycerol chloroplastic protein(TG2), and an aminophoshopholipid-transportingATPase (ALA). VIPP1 has been suggested to play arole in thylakoid membrane formation via membranevesicles (Nordhues et al., 2012), whereas TG2 is a ho-molog of the substrate-binding component of aprokaryote-type ATP-binding cassette (ABC) trans-porter located in the chloroplast envelope that is pro-posed to participate in the chloroplast import of lipidsderived from the endoplasmic reticulum (Awai et al.,2006; Li et al., 2016). ALAs are P4-ATPases implicatedin the translocation of specific phospholipids within thetwo leaflets of biological membranes, a process pro-posed to generate the local curvature that precedesvesicle budding (Poulsen et al., 2008; Zhou andGraham, 2009). Two homologs of half-sized ABCGtransporters also were identified (Table I), which maybe involved in lipid trafficking in E. oleoabundans, asABCG transporters have been related to the transport oflipophilic molecules in other organisms (Verrier et al.,2008; Li et al., 2016).

Our results suggest that lipid degradation is still ac-tive in nitrogen-deprived E. oleoabundans despite themassive oil accumulation triggered by this stress con-dition (Tornabene et al., 1983; Li et al., 2008; Pruvostet al., 2009; Garibay-Hernández et al., 2013). We iden-tified a putative triacylglycerol lipase probably in-volved in the release of fatty acids from neutralglycerolipids as well as homologs of both core andauxiliary plant peroxisomal proteins that participatein the b-oxidation reactions for the degradation ofsaturated and unsaturated fatty acids (Table I). InE. oleoabundans, b-oxidation probably takes place inunspecialized peroxisomes (microbodies) lackingthe glycolate metabolic enzymes, similar to what hasbeen proposed for Eremosphaera (Stabenau et al., 1984)and Dunaliella (Stabenau et al., 1993) green microalgaespecies. The identification of lipid catabolism-relatedproteins in nitrogen-deprived E. oleoabundans supportsthe idea that fatty acid turnover is constitutive and thata continuous balance between oil synthesis and deg-radation exists even under nitrogen stress (Li-Beissonet al., 2015).

The coverage of the lipid metabolic pathways in thiswork was still limited, although additional lipidmetabolism-related proteins were identified in theFFZE fractions and/or in only a single total microsomalsample. Incomplete detection of the entire lipid me-tabolism machinery may be attributed to their lowabundance or lack of similarity with current annotatedsequences, which prevents their positive identification,but also may reflect their complete absence in the ana-lyzed samples. Our results suggest that the amount ofmembrane-associated lipid metabolism proteins maybe lower than has been proposed (Natter et al., 2005;Joyard et al., 2010; Wang and Benning, 2012), as manyhave been identified to a major extent in total solubleT

able

I.(Continued

from

previouspag

e.)

Protein

Iden

tifier

CuratedDescription

Abbr.

Molecu

lar

Mass

ECNo.

BestCrHomolog

BestAtHomolog

SubcellularLo

calization

JGIVersion

5.5

Iden

tifier

%ID

TheArabidopsis

Inform

ationResource

Version10Iden

tifier

%ID

Pr

Cu

Referen

ce

m.117336

ChloroplastJ-like

domain-

containingprotein

CJD

130

NA

Cre03.g171100.t1.1

34

AT1G08640

26

CCb

Ajjaw

iet

al.(2011)

Lipases

andfattyac

idb-oxidation

m.419400

Putative

triacylglyce

rollipase*

aTGL

47

3.1.1.3.

Cre07.g348550.t1.1

42

AT5G67050

37.3

SPCyd

NA

m.225854

Acyl-CoAoxidaseA,

peroxisomal*a

ACX

75

1.3.3.6.

Cre05.g232002.t1.1

66

AT5G65110

57.9

OPXb

Stab

enau

etal.(1984)

m.366023

Acyl-CoAoxidaseB,

peroxisomal*e

ACX

74

1.3.3.6.

Cre11.g467350.t1.2

62

AT1G06290

36

OPXb

Stab

enau

etal.(1984)

m.420224

Fattyac

idb-oxidation

multifunctional

protein,

peroxisomal*c

MFP

77

4.2.1.17.

1.1.1.35.

Cre16.g695050.t1.2

66

AT3G06860

55

SPPXb

Stab

enau

etal.(1984)

m.356362

2,4-D

ienoyl-CoAreductase,

peroxisomal*c

RED

35

1.3.1.4.

Cre17.g731850.t1.2

64

AT3G12800

53

SPPXb

Reu

man

net

al.(2009)

aProtein

iden

tified

exclusively

inFF

ZEmem

branefrac

tions

bEx

perim

entalev

iden

ceofsubce

llularloca

tionisavailable.

c Protein

iden

tified

inonebiologica

lreplicate

oftotalmicro-

somal

samplesan

din

FFZEmem

branefrac

tions.

dEx

perim

entalev

iden

ceofsubce

llularloca

tionisnotavailable.

e Protein

iden

tified

inonly

onebiologicalreplicate

oftotalmicrosomal

samples.





Figure 6. Carbon metabolism in nitrogen-deficient E. oleoabundans. Graphical representation is shown for the carbon metab-olism proteins identified by LC-MS/MS in the membrane proteome of E. oleoabundans. All proteins were identified in FFZEfractions except hexokinase, which was identified exclusively in total microsome samples. Very-low-abundance proteins thatwere identified exclusively in FFZE fractions and not in total microsomes are highlightedwith asterisks. Not identified proteins areshown in a clear gray color. Subcellular locations were predicted using PredAlgo together with experimental evidence availablefor the corresponding homologs. Identified proteins are described in Supplemental Table S4. Protein abbreviations are as follows:ACL, ATP-citrate synthase; AGP, Glc-1-P adenylyltransferase; ALD, aldolase; BASS, Bile Acid:Na+ Symporter, sodium/pyruvatecotransporter; CAH1, carbonic anhydrase, periplasmic; CAH3, carbonic anhydrase, chloroplastic; ENO, enolase; FBP, Fru-1,6-bisphosphatase; G6P, Glc-6-phosphatase; GBSS, granule-bound starch synthase; GPDHc, glyceraldehyde-3-phosphate dehy-drogenase, cytosolic; GPDHp, glyceraldehyde-3-phosphate dehydrogenase A, chloroplastic; GPI, Glc-6-P isomerase; HLA3,probable inorganic carbon transporter HLA3; HK, hexokinase; HPT, UhpC-type hexose phosphate translocator; LCIA, putativeinorganic carbon transporter LCIA; LCIB, LCIB family protein; MDH, malate dehydrogenase, cytoplasmic; PDH, pyruvate de-hydrogenase; PEPC, phosphoenolpyruvate carboxylase; PFK, phosphofructokinase; PGAM, phosphoglycerate mutase; PGK,phosphoglycerate kinase; PGM, phosphoglucomutase; PPT, phosphoenolpyruvate/phosphate translocator; PYK, pyruvate kinase;TPI, triose phosphate isomerase; TPT, triose phosphate/phosphate translocator; RuBisCO, ribulose-1,5-biphosphate carboxylase.Compound abbreviations are as follows: ADP-GLU, ADP-Glc; CIT, citrate; DHAP, dihydroxyacetone phosphate; FBP, Fru-1,6-bisphosphate; F6P, Fru-6-P; G3P, glyceraldehyde-3-phosphate; GLU, Glc; G1P, Glc-1-P; G6P, Glc-6-P; MAL, malate; OAA,oxaloacetate; PEP, phosphoenolpyruvate; 1,3-PG, 1,3-bisphosphoglycerate; 2-PG, 2-phosphoglycerate; 3-PG,3-phosphoglycerate; PYR, pyruvate; RuBP, ribulose-1,5-biphosphate; WSP, water-soluble polysaccharide.






protein extracts from other microalgae (Guarnieri et al.,2011; Gao et al., 2014).

Carbon Metabolism Proteins in E.oleoabundans Membranes

Additional biological processes were covered almostin their entirety in the E. oleoabundans membraneproteome, including central carbon metabolism andelectron transport. A comprehensive description isprovided in this study to better understand the majormetabolic constraints of carbon partitioning in thismicroalga. Other biological processes typically relatedto nitrogen deprivation, including nitrogen acquisition,protein turnover, and oxidative stress responses, alsowere represented in the E. oleoabundans membraneproteome, as well as processes related to pigment me-tabolism, transcription regulation, and signaling, all ofwhich are described in Supplemental Information S1.Carbon metabolism was represented in the mem-

branes of nitrogen-depleted E. oleoabundans by 13% ofthe total proteome (Supplemental Fig. S2; SupplementalTable S4). Not all carbon metabolism-related proteinswere identified in the analysis of total microsomal sam-ples; however, the remainder were identified throughMSanalysis of the FFZE fractions due to the decrease insample complexity by fractionation (Fig. 6; SupplementalTable S4, proteins highlighted with asterisks).

Inorganic Carbon Acquisition and Assimilation

The CO2-concentrating mechanism appears to beactive in nitrogen-limited E. oleoabundans (Fig. 6,orange; Supplemental Table S4), as we identified pro-teins involved in active inorganic carbon uptake (HighLight Activated3 [HLA3]) as well as in its interconver-sion (CO2/HCO3

2), recapture, and concentration withinthe cell (a-type carbonic anhydrases, Limiting CO2-Inducible B [LCIB]-like proteins). Key regulators of theCalvin cycle (CP12, Rubisco activase) and almost half(five of 11) of its chloroplastic enzymes were identified(Supplemental Table S4), including phosphoglyceratekinase (PGK), glyceraldehyde-3-phosphate dehydro-genase (GPDH), and aldolase, enzymes that are sharedwith the chloroplast glycolytic pathway (Fig. 6, green).Photorespiration and one carbon metabolism proteinsalso were present in membranes of nitrogen-deprivedE. oleoabundans (Supplemental Table S4).

Biosynthesis of Photosynthetic Carbon Precursors via aCompartmentalized Glycolytic Pathway

The complete glycolytic pathway, with the exceptionof Glc-6-P isomerase, was identified in this work (Fig. 6,green and blue; Supplemental Table S4), demonstratingthat this pathway is active in nitrogen-deprived E.oleoabundans. This suggests that glycolysis may be themajor contributor for pyruvate production, which ispresumed to be the primary carbon source for fatty acidbiosynthesis (Chapman et al., 2013; Shtaida et al., 2015).

The identification of most of the glycolytic proteins inE. oleoabundans membranes is not as surprising as itmay at first appear. The membrane association of someor all of the glycolytic pathway components and theirsequestration within different organelles are the mostusual forms of glycolytic compartmentalization forregulating central carbon metabolism (Ginger et al.,2010; Johnson and Alric, 2013). Several modes of gly-colytic compartmentalization have emerged in micro-algae (Ginger et al., 2010; Smith et al., 2012), where acomprehensive view exists only for C. reinhardtii (Klein,1986; Johnson and Alric, 2013) and some diatoms(Smith et al., 2012). To assess carbon flow compart-mentalization in E. oleoabundans, we manually curatedthe subcellular location of glycolytic proteins (Fig. 6;Supplemental Table S4). Accordingly, the enzymesfrom the upper part of glycolysis (from hexokinase toPGK; Fig. 6, green) are apparently targeted to the chlo-roplast, and those from the lower part (from phospho-glyceratemutase to pyruvate kinase; Fig. 6, blue)may beassociated with the cytoplasmic face of membranes dueto the lack of a predicted target peptide.

In order to confirm the compartmentalization of theglycolytic pathway, we analyzed the distribution of theglycolytic enzymes in the FFZE fractions using an MS-based approach (Fig. 7). Enzymes from the lower partof glycolysis separated similar to the fractions enrichedin tonoplast and plasma membrane (fractions 56–67),indicating their possible association with the cytoplas-mic side of these membranes. In contrast, chloroplasticGPDH and PGK, two enzymes of the upper part ofglycolysis, were highly abundant in fractions 56 to 59,which alsowere enrichedwith the proteinmarker of theouter chloroplast envelope (TOC75), and to a lesser extentwith markers of the thylakoid (ATP synthase subunit g)and inner chloroplast membranes (TIC110; Fig. 7). Thissuggests that chloroplastic GPDH and PGK may be as-sociated with the chloroplast envelope, as demonstratedfor their counterparts in Arabidopsis (Ferro et al., 2010).The two other enzymes from the upper part of glycolysis,phosphofructokinase and aldolase, showed adistributionprofile similar to that from the tonoplast and plasmamembrane but still were detected in envelope-enrichedfractions (Fig. 7). Thus, we suggest that they may betargeted to both chloroplast and cytoplasmic locations,similar to their C. reinhardtii homologs (Klein, 1986;Johnson and Alric, 2013). Additional targeting of glyco-lytic proteins to other cell locations such as the tonoplastis possible, as demonstrated for enolase and aldolase inthe salt-tolerant plant Mesembryanthemum crystallinum,where they are targeted to the tonoplast to performmoonlighting functions (Barkla et al., 2009).

Glycolysis compartmentalization in E. oleoabundans isreinforced by the identification of triose phosphate andhexose phosphate translocators that may communicatethe two parts of glycolysis across the chloroplast enve-lope (Fig. 6; Supplemental Table S4). Phosphoenolpyr-uvate and pyruvate transporters were not identified inthis work; however, homologs of a phosphoenolpyr-uvate/phosphate translocator and a sodium/pyruvate

















Figure 7. FFZE profiles of glycolytic enzymes suggest their targeting to multiple cellular locations. Individual FFZE fractions wereanalyzed by LC-MS/MS and surveyed for proteins of interest. A, Graphical representation of NSAF values of protein markersspecific for subcellular compartments among the FFZE fractions. The surveyed compartment markers are as follows: m.363780(H+-ATPase), m.392881 (ATPg), m.227792 (TIC110), and m.134654 (TOC75). B, Graphical representation of NSAF values ofglycolytic enzymes among the FFZE fractions. According to their predicted cellular locations, proteins are grouped into upper andlower glycolytic pathway enzymes. These proteins are described in Supplemental Table S4. FFZE fractions enrichedwith the outerchloroplast membrane are enclosed in the box (fractions 56–59).






cotransporter (BASS [Bile Acid:Na+ Symporter]; Fig. 6)have been shown to be transcribed in nitrogen-deprivedE. oleoabundans (Rismani-Yazdi et al., 2012). Altogether,our results support the compartmentalization of theglycolytic proteins in E. oleoabundans; however, addi-tional experiments to determine their specific localizationand dynamics are necessary.Among the identified glycolytic proteins, it is worth

highlighting enolase (EoENO; Supplemental Table S4;Supplemental Fig. S7), whose presence in E. oleoabundansmembraneswas confirmed bywestern blot (SupplementalFig. S7D). Sequence analysis of EoENO and its homologsrevealed that enolases from trebouxiophycean and chloro-phycean microalgae present an additional N-terminalregion (InterPro no. IPR003117) that corresponds to theRIIadomain (Canaves andTaylor, 2002; Supplemental Fig.S7, A and B). Phylogenetic analysis confirmed that RIIa-containing enolases are exclusive to trebouxiophycean andchlorophycean species, clustering separately from eno-lases of land plants and prasinophyceaen microalgae(Supplemental Fig. S7C). The RIIa domain mediateshomodimerization of the regulatory subunit of cAMP-dependent protein kinases (PKA) and high-affinity bind-ing toA-kinase anchoringprotein (AKAP) scaffold proteins,required for the integration of signaling pathways and forthe subcellular compartmentalization of its components(Newlon et al., 2001). This suggests that, in green micro-algae, the RIIa domain may be involved in enolase di-merization and/or anchoring to AKAPs for intracellulartargeting and regulatory purposes, similar to the non-PKARIIa-containing Radial Spoke Protein11 (Yang et al., 2006).

Starch Synthesis

Starch synthesis was represented in the E. oleoabundansmembrane proteome byGlc-1-P adenylyltransferase, oneof the major rate-controlling enzymes, and by other keyenzymes involved in green algal starch metabolism, in-cluding the plastidial phosphoglucomutase and thegranule-bound starch synthase (Fig. 6, red; SupplementalTable S4). This suggests that starch synthesis may still beactive in this microalga after prolonged nitrogen stress,which agrees with the increase in both lipid and starchcontent that was reported previously for nitrogen-deprived E. oleoabundans (Rismani-Yazdi et al., 2012;Garibay-Hernández et al., 2013).

Photosynthetic and Mitochondrial Electron Transport inE. oleoabundans

Analysis of the E. oleoabundans membrane proteomeallowed us to provide a survey of the components ofthe photosynthetic and respiratory electron transportchains (Fig. 8; Supplemental Tables S1–S3), essential forsupplying energy to the processes taking place duringnitrogen deprivation.

The Mitochondrial Respiratory Chain

Each of the five complexes of the mitochondrial re-spiratory chain were represented in the E. oleoabundans

membrane proteome (Fig. 8; Supplemental Table S2).However, comparison of the identified subunits withthe oxidative phosphorylation proteome in C. rein-hardtii (Stern, 2009) shows that important differencesdo exist. Contrary to chlorophycean microalgae,E. oleoabundans has an orthodox intact mitochondria-encoded COX2 (Supplemental Fig. S6) and a classicalmitochondrial ATP synthase (complex V). The latteris supported by the identification of subunit d(Supplemental Table S2), which is known to be absentfrom the noncanonical mitochondrial ATP synthases ofthe chlorophycean lineage (Vázquez-Acevedo et al.,2006, 2016). In addition, sequence analysis of the iden-tifiedmitochondrial ATPase constituents (a, b, g, d, andd; Fig. 8; Supplemental Table S2) suggests that they areencoded by both nuclear and mitochondrial genomes,contrary to the chlorophycean lineage, where allATPase subunits are nucleus encoded (Vázquez-Acevedo et al., 2006, 2016).

Photosynthesis and Photoprotective Mechanisms

All the photosynthetic complexes, PSI-LHCI, cyto-chrome b6f, PSII-LHCII, and ATP synthase, wererepresented in membranes of nitrogen-deprivedE. oleoabundans (Fig. 8; Supplemental Table S1). Atleast 45% of the protein subunits comprising each of thecomplexes were identified, including several low-molecular-mass PSII proteins (PSBE, PSBH, PSBR,PSB27, PSB29, PSB32, and PSB33; Supplemental TableS1), which have proven difficult to detect owing to theirlow abundance, small size, and hydrophobicity (Shiand Schröder, 2004; Shi et al., 2012). Compared withglycolysis- and lipid metabolism-related proteins,whose low abundance hindered their detection in E.oleoabundans membranes, the identification of low-molecular-mass PSII proteins in this work can be at-tributed to the high proportion of chloroplast-localizedproteins that constituted the E. oleoabundans membraneproteome (41%; Supplemental Fig. S2A). The identifiedphotosynthesis-related proteins are homologs of the cor-responding proteins in the model green alga C. reinhardtii(Stern, 2009; Minagawa and Tokutsu, 2015), suggestingthat, contrary to what is observed for the mitochondrialrespiratory chain, the composition of the core photosyn-thetic complexes is highly conserved among green algae.

In agreement with the diminished integrity of thephotosynthetic apparatus that has been reported fornitrogen-deprived E. oleoabundans (Benvenuti et al.,2015), proteins involved directly in the synthesis andturnover of the D1 subunit (Filamentous Temperature-Sensitive H [FTSH] ATP-dependent zinc metalloproteases,Low PSII Accumulation1, atypical short-chain dehydro-genase HCF244), as well as in PSI (YCF4) and PSIIassembly, stability, and/or repair (peptidyl-prolyl cis-trans-isomerase CYP38, M-Enriched Thylakoid1,MPH1, rubredoxin), were identified in this work(Supplemental Table S1). Additional photoprotective re-sponses appeared to be active in nitrogen-deprivedE. oleoabundans, as we were able to identify key
























Figure 8. Photosynthesis and oxidative phosphorylation in the membrane proteome of nitrogen-deficient E. oleoabundans.Graphical representation is shown for the proteins identified by LC-MS/MS in the membrane proteome of E. oleoabundans in-volved in energy conversion and homeostasis. Identified proteins are described in Supplemental Table S1 (photosynthesis),Supplemental Table S2 (oxidative phosphorylation), and Supplemental Table S3 (energy and reducing power homeostasis). Thenumber of proteins detected in E. oleoabundans from each of the complexes of the chloroplastic and mitochondrial electrontransfer chains is shown in colored boxes and comparedwith the number of proteins identified in the complexes of themodel algaC. reinhardtii. In complex IV, CrCOX2 is considered as a nonsplit subunit. In the E. oleoabundans ATP synthase mitochondrialcomplex, the N- and C-terminal peptides of subunit d are considered as a unique nonsplit protein. Black dashed lines indicateelectron transfer, and red dashed lines indicate proton translocation. C, Cytoplasmic; M,mitochondrial; P, chloroplastic. Asterisksindicate proteins associated with any part of the PSII-LHCII supercomplex. AAA, ADP/ATP carrier protein, chloroplastic; AAC,ADP/ATP carrier protein, mitochondrial; ADK, adenylate kinase; Cytc, cytochrome c; DIT1, dicarboxylate transporter 1, chlo-roplastic; DTC, mitochondrial dicarboxylate/tricarboxylate carrier; Fd, ferredoxin; FNR, ferredoxin-NADP reductase; FUM, fu-marate; ICM, inner chloroplast membrane; IMM, inner mitochondrial membrane; LHCI, light-harvesting complex of PSI; LHCII,light-harvesting complex of PSII; LMMS, low-molecular-mass subunits; MAL, malate; MDH, malate dehydrogenase; MPT, mi-tochondrial phosphate carrier protein; OAA, oxaloacetate; OCM, outer chloroplast membrane; OEC, oxygen-evolving complex;OMM, outer mitochondrial membrane; PC, plastocyanin; Pi, inorganic phosphate; PQ, plastoquinone; Q, ubiquinone; SUC,succinate; THY, thylakoid membrane.








molecular effectors of the short-term components of NPQ(Supplemental Table S1): energy-dependent feedbackdeexcitation quenching (qE; calcium-sensing receptorCAS, PSBS), zeaxanthin-dependent quenching (qZ; vio-laxanthin deepoxidase), and state transition-dependentquenching (qT; Ser/Thr protein kinase STT7) (Ericksonet al., 2015; Minagawa and Tokutsu, 2015). Key compo-nents of the two proposed cyclic electron flow pathways(Iwai et al., 2010; Johnson and Alric, 2013), the NADPHdehydrogenase-dependent pathway (Type-II NAD(P)Hdehydrogenase) and the ferredoxin-dependent path-way (ferredoxin-NADP reductase, PGR5-like protein1),also were identified. (Supplemental Tables S1 and S3).This supports that alternative electron pathways maybe active in nitrogen-deprived E. oleoabundans forphotoprotection and for satisfying the varying demandfor ATP/NADPH under abiotic stress conditions.Additional mechanisms known to modulate the redox

potential and ATP concentration in different cellularcompartments, particularly under fluctuating environ-mental conditions (Cardol et al., 2003; Johnson and Alric,2013; Erickson et al., 2015), were represented in nitrogen-deprived E. oleoabundans. We were able to identify themain effectors of the malate shunt (malate dehydrogen-ase isoforms, dicarboxylate transporter, and mitochon-drial dicarboxylate/tricarboxylate carrier) and proteinsnecessary for regulating ATP concentrations within thechloroplast and the mitochondria (adenylate kinase iso-forms, mitochondrial and chloroplastic ADP/ATP car-rier proteins, and mitochondrial phosphate carrierprotein; Fig. 8; Supplemental Table S3).

CONCLUSION

The results from this work provide a detailed surveyof the membrane proteome of an oleaginous microalga.Combining gel-free shotgun proteomics with searchingagainst an organism-specific RNA-Seq-based proteindatabase considerably improved protein identification.This approach overcame both the typical limitations ofstudying membrane proteins and the difficulty ofworking with nonsequenced organisms for which thequality and quantity of the data available in referencedatabases are neither complete nor specific. Althoughmanual annotation may be a time-consuming strategy,we demonstrated its usefulness for analyzing non-sequenced organisms, as it significantly improved thenumber of identified proteins as well as the accuracyand reliability of their annotations.This approach allowed the novel identification in E.

oleoabundans of the photosynthesis-related proteinsMPH1 and PSBS, both thought to be exclusive to higherphotosynthetic organisms. These findings suggest thatphotoprotective mechanisms, including NPQ, are ac-tive after prolonged nitrogen deprivation and indicatethat, in E. oleoabundans, these mechanisms are moreclosely related to higher photosynthetic organisms thanwas proposed previously. The identification of PSBSand the presumed absence of an LHCSR homolog in E.oleoabundans are contrary to what has been observed in

C. reinhardtii. In C. reinhardtii, a light-inducible PSBSwas identified recently and was demonstrated to beessential for the activation of an LHCSR-dependent qEmechanism to which most of the microalgal NPQ ca-pability has been attributed (Peers et al., 2009; Niyogiand Truong, 2013; Correa-Galvis et al., 2016; Tibilettiet al., 2016). This result questions the conservation of qEmechanisms within green microalgae, where the spe-cific role played by PSBS in E. oleoabundans NPQ mustbe determined. In addition to the photosynthesis-related proteins, we also detected RP2-CLC, a noveldomain architecture protein that is likely involved inthe intracellular trafficking of clathrin-coated vesicles inlower eukaryotes, a process that apparently has its ownpeculiarities in these understudied organisms. UsingFFZE fractionation of membranes, we confirmed thechloroplastic location of PSBS and MPH1 together withthe enrichment of RP2-CLC in the plasma membrane.Using this strategy also contributed to the identificationof very-low-abundance proteins related to E. oleoabundanslipid metabolism, allowing us to identify a detailedlist of proteins involved in the major steps of acyl-lipid metabolism, lipid trafficking, lipid signaling,and LD formation in E. oleoabundans. An MS-basedanalysis of FFZE fractions additionally supported thecompartmentalization of glycolytic proteins in E.oleoabundans, which is an important constraint thatappears to govern central carbon metabolism andpartitioning.

Finally, through molecular phylogenetic approaches,we provide compelling evidence for the phylogeneticgrouping of this microalga with the Chlorellales lineageof the trebouxiophycean class of green microalgaerather than with the chlorophycean class in which it iscurrently classified. Our results provide an importantplatform for studying E. oleoabundans and underscorethe importance of studying nonmodel organisms, as theanalysis of specific features in E. oleoabundans demon-strates that its biology differs from that of non-oleoaginous model organisms.

MATERIALS AND METHODS

Microalgae Strains and Culture Conditions

Ettlia oleoabundans UTEX 1185 was grown under phototrophic conditions in2.8-L Fernbach glass flasks with a working volume of 40% using modifiedBold’s Basal Medium (Garibay-Hernández et al., 2013). Axenic cultures withan initial cell density of 1 to 23 106 cells mL21 weremaintained for 7 d at 25°C60.5°C under continuous orbital agitation (300 rpm) and white fluorescent lightillumination (100 mE m22 s21). To induce nitrogen deprivation, 7-d cultureswere centrifuged individually for 10 min (10,000g at 4°C), washed once with200 mL of nitrogen-free modified Bold’s Basal Medium, resuspended in 1.12 Lof nitrogen-free modified Bold’s Basal Medium, and transferred into Fernbachflasks. Axenic nitrogen-deprived cultures with an initial cell density of 10 to15 3 106 cells mL21 were maintained during 4 d under the aforementionedconditions.

The Chlamydomonas reinhardtii wall-less strain cw15 mt+ was grown undermixotrophic conditions in 0.5-L Erlenmeyer glass flasks with a 40% workingvolume using Tris-acetate-phosphate medium (Harris, 2009). Axenic cultureswith an initial cell density of 1 to 23 106 cells mL21 were maintained for 4 d at24°C 6 2°C under white fluorescent light illumination (50–100 mE m22 s21 and16/8-h light/dark cycle) and continuous orbital agitation (80 rpm).








Cell density was determined by direct microscopic cell count using a Neu-bauer chamber. Cultureswere tested for the absence of bacteria in Glc-free Luriabroth agar (1.5%, w/v) medium plates (Bertani, 1951) incubated at 37°C for atleast 24 h.

Microsomal Membrane Isolation

All the operations in this protocolwere performed at 4°C.Nitrogen-deprivedE. oleoabundans cultures (1.12 L) or C. reinhardtii cultures (0.4 L) were used formicrosome isolation. Cultures were centrifuged for 10 min (10,000g at 4°C),washed once with 0.1 M HEPES-KOH, pH 7.5 (50/500 mL of centrifuged cul-ture), and resuspended in 2.5 mL of homogenization medium (400 mM man-nitol, 10% [w/v] glycerol, 5% [w/v] polyvinylpyrrolidone-10, 0.5% [w/v]bovine serum albumin, 1 mM phenylmethylsulfonyl fluoride, 30 mM Tris, 2 mM

dithiothreitol, 5 mM EGTA, 5 mM MgSO4, 0.5 mM butylated hydroxytoluene,0.25 mM dibucaine, 1 mM benzamidine, and 26 mMK+-metabisulfite; adjusted topH 8 with NaOH). Cells were homogenized by passing the cell suspension fivetimes through a French press (Thermo Spectronic; model FA-078) at 20 k.p.s.i.using a mini pressure cell. Microsomal membranes were isolated as describedby Barkla et al. (1995). Briefly, the homogenate was centrifuged for 20 min(10,000g at 4°C) and the pellet was discarded. To concentrate the microsomes,the supernatant was centrifuged (80,000g, 50 min, at 4°C; Beckman 45 Ti rotor,L8-M ultracentrifuge). The microsomal pellet was resuspended in suspensionmedium (400 mM mannitol, 10% [w/v] glycerol, 6 mM Tris/MES, pH 8, and2 mM dithiothreitol) using a 10-mL glass Teflon homogenizer. Samples werefrozen in liquid N2 for storage at 280°C. Microsomes from Arabidopsis (Ara-bidopsis thaliana) were isolated as described by Barkla et al. (2007).

FFZE

Microsomal membranes from nitrogen-deprived E. oleoabundans were frac-tionated by FFZE using the BD FFE System (BD Proteomics) as described byBarkla et al. (2007). Briefly, prior to fractionation, the microsomal sample wasdiluted 2:1 (v/v) in separation medium (10 mM triethanol amine, 10 mM aceticacid, 2 mM KCl, and 250 mM Suc) and centrifuged for 20 min (14,000g at 4°C).The sample was supplemented with MgATP (3 mM final concentration) to en-hance membrane separation during FFZE (Barkla et al., 2007). The sample wasinjected continuously via a peristaltic pump at 1.2 mL h21 using the anodicsample inlet. Media inlets 2 to 6 and counter-flow inlets C1, C2, and C3 con-tained separation medium, whereas inlets 1 and 7 contained stabilization me-dium (40 mM triethanol amine, 40 mM acetic acid, 8 mM KCl, and 180 mM Suc).The cathodic and anionic circuit electrolyte solutions consisted of 100 mM

triethanol amine, 100 mM acetic acid, and 20 mM KCl adjusted to pH 7.4(NaOH); 0.4% (v/v) formaldehyde was added to the anodic solution to preventthe loss of chloride by anodic oxidation.

FFZE was performed in horizontal mode at 5°C and 750 V (150 mA) withmedia and counter-flow rates of 250 mL h21. Following separation in thechamber, membrane fractions were collected continually on 96-deep-wellmicrotiter plates (4 mL per well). Fractions from sequential plates corre-sponding to the same well were pooled; 1.2-mL aliquots were collected perpooled fraction, frozen in liquid N2, and stored at 280°C for LC-MS/MSanalysis. The remaining volume of each collected fraction was ultracentrifuged(100,000g, 50 min, at 4°C; Beckman 55.2 Ti rotor, L8-M ultracentrifuge) formembrane concentration. Membrane pellets corresponding to each fractionwere resuspended in 25 to 200 mL of suspension buffer (250 mM mannitol, 10%[w/v] glycerol, 10 mM Tris/MES, pH 8, and 2 mM dithiothreitol) and frozen inliquid N2 for storage at 280°C. FFZE separation was monitored by collectingmicrotiter plates (250 mL per well) at several time points andmeasuring protein(optical density at 280 nm) using a microplate scanning spectrophotometer(Power WaveX; Bio-Tek Instruments).

Protein and Chlorophyll Concentration Measurements

Protein in microsomal and concentrated FFZE fractions was measured by amodification of the Bradford method (Bradford, 1976). Membrane protein waspartially solubilizedwith 0.5% (v/v) Triton X-100 for 5min prior to dilution andaddition of the dye reagent concentrate; the final Triton X-100 concentration inthe assay was 0.05% (w/v). Bovine serum albumin was used as the proteinstandard.

Chlorophyll in FFZE fractions prior to concentration was measuredaccording to the method of Arnon (1949), with some modifications, using amicroplate scanning spectrophotometer (Power WaveX; Bio-Tek Instruments).

Absorbance was measured directly at 645 and 663 nm, and calculations weremade according to the following equations:

Chla�mg mL-1

� ¼ 12:7A663 2 2:69A645

Chlb�mg mL-1

� ¼ 22:9A645 2 4:68A663

SDS-PAGE, Staining, and Immunoblotting

For protein precipitation, total microsome samples and concentrated FFZEfractions were diluted 50-fold in 1:1 (v/v) ethanol:acetone and incubatedovernight at 220°C according to Parry et al. (1989). Samples were then centri-fuged for 20 min (14,000g at 4°C). Pellets were air dried, resuspended inLaemmli buffer (Laemmli, 1970), and heated (60°C for 2 min) before loading(15–20mg of protein per lane) onto 10% to 12.5% (w/v) linear acrylamide gels asindicated. After electrophoresis, SDS-PAGE separated proteins were eitherfixed and stained with Coomassie Blue R250 or electrophoretically transferredonto nitrocellulosemembranes (enhanced chemiluminescence; GE Lifesciences)for immunoblot analysis as described by Vera-Estrella et al. (2004). Digitalimages were captured using the Gel Doc XR+ System (Bio-Rad). Primary anti-bodies, either commercially available or custom made, and dilutions employedin this study were as follows: Arabidopsis anti-MPH1 (1:200; Liu and Last, 2015b);C. reinhardtii anti-LHCSR3 (1:1,000; Agrisera; AS14 2766);Homo sapiens anti-enolase(1:1,000; Santa Cruz Biotechnology; sc-7455); H. sapiens anti-RP2 (1:3,000; Chappleet al., 2000); Spinacia oleracea anti-AtpB (1:10,000;McCormac andBarkan, 1999); andglobal anti-H+-ATPase (1:1,000; Agrisera; AS07 260).

Shotgun Proteomics Analysis

For sample preparationprior to proteomic analysis, totalmicrosome samples(100mg per replicate; four independent biological replicates) and FFZE fractions(1.2 mL per fraction; 23 individual FFZE fractions) were suspended in TE buffer(10 mM Tris/HCl, pH 7.6, 1 mM EDTA pH 8), and 0.3% [w/v] sodium deoxy-cholate), precipitated with 72% (w/v) TCA (9% final concentration; 4°C for 1 h),and, after recovering the precipitated protein, submitted to an additional pre-cipitation step with 90% (v/v) acetone (230°C overnight). The preparation(solubilization, reduction, alkylation, and trypsin digestion) of the vacuum-dried protein extracts and their further manipulation (resolubilization anddesalting) prior to LC-MS/MS analysis were performed as described in detailby Barkla et al. (2012).

LC-MS/MS analysis was performed at the proteomics discovery platform ofthe Institut de Recherches Cliniques de Montréal using the LC-MS/MSequipment described in detail by Barkla et al. (2012). Chromatography bufferswere 0.2% (v/v) formic acid (buffer A) and 100% (v/v) acetonitrile/0.2% (v/v)formic acid (buffer B). Peptide samples were loaded on column (600 nL min21)and eluted (250 nL min21) with a two-slope gradient: buffer B first increasedfrom 2% to 40% (85 min) and then from 40% to 80% (15 min). LC-MS/MS dataacquisition was accomplished using an 11-scan event cycle, composed of a fullMS scan for scan event 1 acquired in the Orbitrap. Mass resolution for MS wasset to 60,000 (at mass-to-charge ratio of 400) and used to trigger the 10 addi-tional tandem mass spectrometry (MS/MS) events acquired in parallel in thelinear ion trap for the top 10 most intense ions. The mass-to-charge ratio rangewas from 360 to 1,700 for MS scanning with a target value of 1,000,000 chargesand from approximately one-third of the parent mass-to-charge ratio to 2,000for MS/MS scanning with a target value of 10,000 charges. Data-dependent scanevents used a maximum ion fill time of 100 ms and one microscan. Target ionsalready selected for MS/MS were dynamically excluded for 31 s after two counts.Nanospray and S-lens voltages were set to 1.3 to 1.8 kV and 50 V, respectively.Capillary temperature was set to 250°C. MS/MS conditions were as follows:normalized collision energy, 35 V; activation Q, 0.25; activation time, 10 ms.

An E. oleoabundans protein database was generated at the Unidad Uni-versitaria de Apoyo Bioinformático-Universidad Nacional Autónoma deMéxico through in silico six-frame translation of the E. oleoabundans nonre-dundant transcriptome database (56,550 transcripts; Rismani-Yazdi et al.,2012). TransDecoder (Haas et al., 2013) was used to identify the candidateprotein-coding regions based on nucleotide composition and open readingframe length. For database search, peak list files were generated with ProteomeDiscoverer (version 1.4) using the following parameters: minimummass, 500 D;maximum mass, 6,000 D; no grouping of MS/MS spectra; precursor charge,auto; minimum number of fragment ions, five. All MS/MS spectra were ana-lyzed with Mascot 2.3 (Matrix Science) against the Viridiplantae (TaxID 33090,unknown version; 677,107 entries) and/or the E. oleoabundans (unknown





version; 53,921 entries) protein databases assuming trypsin digestion. The masstolerances for precursor and fragment ions were set to 10 ppm and 0.6 D, re-spectively. Cys carbamidomethylation was specified as a fixed modificationand Met oxidation as a variable modification.

Scaffold 4.3.4 (Proteome Software) was used to validate MS/MS-basedpeptide and protein identifications. Scaffold parameters were set to a minimumof two peptides per protein, with minimum probabilities of 99% at the proteinlevel (Protein Prophet algorithm; Nesvizhskii et al., 2003) and 95% at the cor-responding peptide level (Scaffold Local FDR algorithm). Proteins containingsimilar peptides that could not be differentiated based on MS/MS analysisalone were grouped to satisfy the principles of parsimony.

Protein Functional Annotation

Proteinswere annotated if they compliedwith the above-mentioned Scaffoldparameters and if they were identified with two or more unique peptides in atleast two of four biological replicates, unless stated otherwise. Automatic an-notation based on sequence similarity was initially performed using theBlast2GO suite (version 2.8; April 2014) with the default parameters (Götz et al.,2008). To improve protein annotation, each of the identified proteins was in-dividually curated during the period from May 2014 and May 2015. Manualcuration was initially based on the results provided by the Blast2GO analysisand additionally supported by the following resources: experimental evidenceavailable in the literature for protein homologs and detailed sequence analysisin the PredAlgo subcellular localization prediction tool (Tardif et al., 2012) andin multiple databases, including ARAMEMNON version 8.0 (Schwacke et al.,2003), InterPro (Mitchell et al., 2015), Phytozome version 10.2 (Goodstein et al.,2012), SUBA3 (Tanz et al., 2013), TCDB (Saier et al., 2014), UniProtKB (Swis-sProt and TrEMBL entries; UniProt Consortium, 2015), ARALIP (Li-Beissonet al., 2010), AT_CHLORO (Ferro et al., 2010), MEROPS version 9.9(Rawlings et al., 2014), and PlnTFDB (Pérez-Rodríguez et al., 2010), dependingon the requirements. Functional assignments were additionally verified usingthe IntEnz (Fleischmann et al., 2004) and QuickGO (Binns et al., 2009) data-bases. A protein sequence from E. oleoabundans was manually annotated as aspecific protein only if it contained the domains required for the function thatwas being assigned or at least the signatures necessary to be allocated to aspecific protein family; otherwise, proteins were annotated as unknown even ifthey were electronically inferred as a known protein.

Genomic DNA Extraction and 18S rDNA Sequencing

Genomic DNA was extracted from 25 mL of nitrogen-sufficient E. oleoabundanscultures as follows. Cells were pelleted by centrifugation (10,000g, 10 min,at room temperature), frozen in liquid N2, and ground with mortar and pestle.Ground cells were suspended in TE buffer to an initial working volume of600mL towhich 180mL of lysozyme (50mgmL21) was added and incubated for30 min at room temperature. Then, 120 mL of 10% (w/v) SDS and 3 mL ofproteinase K (20 mg mL21) were added to the sample and incubated for 1 h at37°C with agitation. After this, 100 mL of 5 M NaCl and 80 mL of cetyl-trimethyl-ammonium bromide/NaCl (10% [w/v]/0.7 M) were added to the sample,which was mixed and incubated for a further 10 min at 65°C with agitation.DNA was extracted with 1 volume of CH3Cl:isoamyl alcohol (24:1, v/v). Theupper phase was recovered by centrifugation (14,000g, 5 min, at room tem-perature) and combined with 1 volume of phenol:CH3Cl:isoamyl alcohol(25:24:1, v/v/v). For DNA precipitation, the upper phase was recovered bycentrifugation, combined with 0.6 volume of isopropanol, and incubated for15 min at 230°C. The precipitate was collected by centrifugation (14,000g,10 min, at 4°C), washed with 1 mL of cold 75% (v/v) ethanol, dried, andresuspended in 50 mL of TE buffer. Then, genomic DNA was treated with 3 mLof RNase A (20mgmL21; 30min, 37°C, and agitation), taken to a final volume of300 mL by adding water, and reextracted with 1 volume of phenol:CH3Cl:iso-amyl alcohol (25:24:1, v/v/v). The upper phase was recovered by centrifuga-tion, and DNA was precipitated with 0.1 volume of 7.5 M ammonium acetateplus 2.5 volume of absolute ethanol (15 min at 230°C). The precipitate wascollected, washed, dried, and dissolved as described above. Genomic DNAquality was assessed by agarose gel electrophoresis and by A260/A280 and A260/A230 ratios (NanoDrop spectrophotometer).

The E. oleoabundans 18S rRNA gene was amplified by PCR (35 cycles; LongPCR Enzyme Mix; ThermoFisher Scientific) using genomic DNA as a templateand the oligonucleotide primers (59-ACCTGGTTGATCCTGCCAG-39 and59-TGATCCTTCYGCAGGTTCAC-39) according to Moon-van der Staay et al.(2001). A PCR product of approximately 2.5 kbwas amplified and purified from

an agarose gel using the High Pure PCR Product Purification Kit (Roche Di-agnostics), whose quality was verified as described above. A partial 18S rDNAsequence was obtained by sequencing (Applied Biosystems, model 3130xl) anddeposited in the NCBI GenBank database under the accession numberKX350066.

Molecular Phylogenetic Analysis

To perform a molecular phylogenetic analysis using a multigene approach,two nuclear (Actin and EF-1a), two chloroplastic (PSBA and RBCL), and twomitochondria-encoded (COX1 and COX2) protein markers were selected. Theamino acid sequences retrieved for the analysis are described in SupplementalTable S16. Each set of proteins was aligned individually using Clustal Omega(Sievers et al., 2011). The resulting alignments were manually inspected usingMEGA7 (Kumar et al., 2016), where all positions with less than 70% site cov-erage were removed. The verified alignments for each of the six proteins wereconcatenated using CLC Genomics Workbench 9.0 (Qiagen Bioinformatics).The evolutionary analysis was conducted in MEGA7 using the ML methodbased on the JTT matrix-based model (Jones et al., 1992). The initial tree wasmade automatically using the default method (Neighbor Joining/BiologicalNeighbor Joining). The nodal support was estimated by bootstrap values(Felsenstein, 1985) based on 1,000 replications. Additional phylogenetic anal-yses of proteins identified in this work were conducted in MEGA7 as describedabove, where all positions with less than 70% site coverage were eliminated.

Phylogenetic analysis of 18S rDNA data was performed with an increasedtaxon sampling, mostly based on Pegg et al. (2015), that comprised 62 different18S rDNAnucleotide sequences (complete and partial) retrieved from the NCBInucleotide database (Supplemental Table S17). Sequences were aligned usingClustal Omega. The evolutionary analysis was conducted in MEGA7 using theML method based on the model of Tamura and Nei (1993), where all positionswith less than 70% site coverage were eliminated. The initial tree constructionand the estimation of nodal support were performed as described above.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL datalibraries under accession number KX350066 [18S ribosomal RNA gene, partialsequence]. The E. oleoabundans protein database and the MS proteomics datahave been deposited to the ProteomeXchange Consortium (http://www.proteomexchange.org) via the MassiVE partner repository (http://massive.ucsd.edu) with the MassiVE data set identifier MSV000079977. Additionalprotein and peptide information from LC-MS/MS analysis of FFZE membranefractions is provided in Table S18.

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. BLASTP best hits for E. oleoabundans membraneproteins.

Supplemental Figure S2. Cellular component and biological process dis-tributions of identified proteins.

Supplemental Figure S3. Sequence analysis of EoPSBS.

Supplemental Figure S4. Sequence analysis of EoMPH1.

Supplemental Figure S5. Sequence analysis of TBCC domain-containingproteins.

Supplemental Figure S6. Sequence analysis of EoCOX2.

Supplemental Figure S7. Analysis of EoENO and its homologs in thegreen lineage.

Supplemental Tables S1 to S15. Detailed description of proteins identifiedin the membranes of nitrogen-deprived E. oleoabundans according to thebiological process in which they are involved.

Supplemental Table S16. Molecular phylogenetic analysis based on amultigene approach.

Supplemental Table S17. 18S rDNA molecular phylogenetic analysis.

Supplemental Table S18. Proteins surveyed in FFZE membrane fractionsvia LC-MS/MS analysis.







http://www.proteomexchange.org

http://www.proteomexchange.org

http://massive.ucsd.edu

http://massive.ucsd.edu














Supplemental Information S1. Additional biological processes repre-sented in the E. oleoabundans membrane proteome.

ACKNOWLEDGMENTS

We thank Dr. Denis Faubert and Marguerite Boulos at the Institut de Recher-ches Cliniques de Montréal-Proteomics Discovery Platform for LC-MS/MSanalysis; Dr. Jordan Peccia (Yale University) and Dr. Berat Z. Haznedaroglu(Boğaziçi University) for providing the E. oleoabundans transcriptome; Dr. AlejandroSanchez-Flores and Karel Estrada at the Unidad Universitaria de ApoyoBioinformático-Universidad Nacional Autónoma de México (UNAM) forgenerating the E. oleoabundans protein database; Dr. Diego Gonzalez-Halphen (UNAM) for the gift of the C. reinhardtii strain, Dr. Carlos F. Arias(UNAM) for the gift of C2BBe1 cells, Dr. Robert L. Last (Michigan StateUniversity) for the gift of the anti-MPH1 antibody, Dr. Michael Cheetham(University College London) for the gift of the anti-RP2 antibody, and Dr.Patricia Leon (UNAM) for the gift of the anti-AtpB antibody; Luz MaríaMartínez, Mario A. Caro Bermúdez, Mercedes Enzaldo Cruz, and M.Guadalupe Muñoz García for technical assistance; and Gustavo Rodriguez-Alonso for support in phylogenetic analysis.

Received August 8, 2016; accepted November 3, 2016; published November 8,2016.

LITERATURE CITED

Ajjawi I, Coku A, Froehlich JE, Yang Y, Osteryoung KW, Benning C, LastRL (2011) A J-like protein influences fatty acid composition of chloro-plast lipids in Arabidopsis. PLoS ONE 6: e25368

Alboresi A, Gerotto C, Giacometti GM, Bassi R, Morosinotto T (2010)Physcomitrella patens mutants affected on heat dissipation clarify theevolution of photoprotection mechanisms upon land colonization. ProcNatl Acad Sci USA 107: 11128–11133

Arnon DI (1949) Copper enzymes in isolated chloroplasts: poly-phenoloxidases in Beta vulgaris. Plant Physiol 24: 1–15

Arredondo-Vega BO, Band CJ, Vazquez-Duhalt R (1995) Biochemicalcomposition of Neochloris oleoabundans adapted to marine medium.Cytobios 83: 201–205

Atteia A, Adrait A, Brugière S, Tardif M, van Lis R, Deusch O, Dagan T,Kuhn L, Gontero B, Martin W, et al (2009) A proteomic survey ofChlamydomonas reinhardtii mitochondria sheds new light on the met-abolic plasticity of the organelle and on the nature of the a-proteo-bacterial mitochondrial ancestor. Mol Biol Evol 26: 1533–1548

Awai K, Xu C, Tamot B, Benning C (2006) A phosphatidic acid-bindingprotein of the chloroplast inner envelope membrane involved in lipidtrafficking. Proc Natl Acad Sci USA 103: 10817–10822

Baba M, Suzuki I, Shiraiwa Y (2011) Proteomic analysis of high-CO2-inducible extracellular proteins in the unicellular green alga, Chlamy-domonas reinhardtii. Plant Cell Physiol 52: 1302–1314

Bagos PG, Liakopoulos TD, Hamodrakas SJ (2004) Finding beta-barrelouter membrane proteins with a Markov chain model. WSEAS TransBiol Biomed 2: 186–189

Bailleul B, Rogato A, de Martino A, Coesel S, Cardol P, Bowler C,Falciatore A, Finazzi G (2010) An atypical member of the light-harvesting complex stress-related protein family modulates diatom re-sponses to light. Proc Natl Acad Sci USA 107: 18214–18219

Baldisserotto C, Ferroni L, Giovanardi M, Boccaletti L, Pantaleoni L,Pancaldi S (2012) Salinity promotes growth of freshwater Neochlorisoleoabundans UTEX 1185 (Sphaeropleales, Chlorophyta): morphophy-siological aspects. Phycologia 51: 700–710

Baldisserotto C, Popovich C, Giovanardi M, Sabia A, Ferroni L,Constenla D, Leonardi P, Pancaldi S (2016) Photosynthetic aspects andlipid profiles in the mixotrophic alga Neochloris oleoabundans as usefulparameters for biodiesel production. Algal Res 16: 255–265

Barkla BJ, Vera-Estrella R, Hernández-Coronado M, Pantoja O (2009)Quantitative proteomics of the tonoplast reveals a role for glycolyticenzymes in salt tolerance. Plant Cell 21: 4044–4058

Barkla BJ, Vera-Estrella R, Pantoja O (2007) Enhanced separation of membranesduring free flow zonal electrophoresis in plants. Anal Chem 79: 5181–5187

Barkla BJ, Vera-Estrella R, Pantoja O (2012) Protein profiling of epidermalbladder cells from the halophyte Mesembryanthemum crystallinum.Proteomics 12: 2862–2865

Barkla BJ, Zingarelli L, Blumwald E, Smith J (1995) Tonoplast Na+/H+

antiport activity and its energization by the vacuolar H+-ATPase in thehalophytic plant Mesembryanthemum crystallinum L. Plant Physiol 109:549–556

Bartolini F, Bhamidipati A, Thomas S, Schwahn U, Lewis SA, Cowan NJ(2002) Functional overlap between retinitis pigmentosa 2 protein and thetubulin-specific chaperone cofactor C. J Biol Chem 277: 14629–14634

Beaudoin F, Wu X, Li F, Haslam RP, Markham JE, Zheng H, Napier JA,Kunst L (2009) Functional characterization of the Arabidopsis beta-ketoacyl-coenzyme A reductase candidates of the fatty acid elongase.Plant Physiol 150: 1174–1191

Benvenuti G, Bosma R, Cuaresma M, Janssen M, Barbosa MJ, WijffelsRH (2015) Selecting microalgae with high lipid productivity and pho-tosynthetic activity under nitrogen starvation. J Appl Phycol 27: 1425–1431

Bertani G (1951) Studies on lysogenesis. I. The mode of phage liberation bylysogenic Escherichia coli. J Bacteriol 62: 293–300

Berthelot K, Lecomte S, Estevez Y, Peruch F (2014) Hevea brasiliensis REF(Hev b 1) and SRPP (Hev b 3): an overview on rubber particle proteins.Biochimie 106: 1–9

Bienvenut WV, Espagne C, Martinez A, Majeran W, Valot B, Zivy M,Vallon O, Adam Z, Meinnel T, Giglione C (2011) Dynamics of post-translational modifications and protein stability in the stroma of Chla-mydomonas reinhardtii chloroplasts. Proteomics 11: 1734–1750

Binns D, Dimmer E, Huntley R, Barrell D, O’Donovan C, Apweiler R(2009) QuickGO: a web-based tool for Gene Ontology searching. Bio-informatics 25: 3045–3046

Boldt L, Yellowlees D, Leggat W (2012) Hyperdiversity of genes encodingintegral light-harvesting proteins in the dinoflagellate Symbiodinium sp.PLoS ONE 7: e47456

Bonente G, Passarini F, Cazzaniga S, Mancone C, Buia MC, Tripodi M,Bassi R, Caffarri S (2008) The occurrence of the psbS gene product inChlamydomonas reinhardtii and in other photosynthetic organisms andits correlation with energy quenching. Photochem Photobiol 84: 1359–1370

Bradford MM (1976) A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dyebinding. Anal Biochem 72: 248–254

Canaves JM, Taylor SS (2002) Classification and phylogenetic analysis ofthe cAMP-dependent protein kinase regulatory subunit family. J MolEvol 54: 17–29

Cardol P, Gloire G, Havaux M, Remacle C, Matagne R, Franck F (2003)Photosynthesis and state transitions in mitochondrial mutants of Chla-mydomonas reinhardtii affected in respiration. Plant Physiol 133: 2010–2020

Carrie C, Murcha MW, Millar AH, Smith SM, Whelan J (2007) Nine3-ketoacyl-CoA thiolases (KATs) and acetoacetyl-CoA thiolases(ACATs) encoded by five genes in Arabidopsis thaliana are targetedeither to peroxisomes or cytosol but not to mitochondria. Plant Mol Biol63: 97–108

Castruita M, Casero D, Karpowicz SJ, Kropat J, Vieler A, Hsieh SI, YanW, Cokus S, Loo JA, Benning C, et al (2011) Systems biology approachin Chlamydomonas reveals connections between copper nutrition andmultiple metabolic steps. Plant Cell 23: 1273–1292

Chantanachat S, Bold HC (1962) Phycological Studies II. Some Algae fromArid Soils. University of Texas, Austin, Texas

Chapman KD, Dyer JM, Mullen RT (2013) Commentary: why don’t plantleaves get fat? Plant Sci 207: 128–134

Chapple JP, Hardcastle AJ, Grayson C, Spackman LA, Willison KR,Cheetham ME (2000) Mutations in the N-terminus of the X-linked ret-initis pigmentosa protein RP2 interfere with the normal targeting of theprotein to the plasma membrane. Hum Mol Genet 9: 1919–1926

Chapple JP, Hardcastle AJ, Grayson C, Willison KR, Cheetham ME (2002)Delineation of the plasma membrane targeting domain of the X-linkedretinitis pigmentosa protein RP2. Invest Ophthalmol Vis Sci 43: 2015–2020

Chen M, Zhao L, Sun YL, Cui SX, Zhang LF, Yang B, Wang J, Kuang TY,Huang F (2010) Proteomic analysis of hydrogen photoproduction insulfur-deprived Chlamydomonas cells. J Proteome Res 9: 3854–3866

Correa-Galvis V, Redekop P, Guan K, Griess A, Truong TB, Wakao S,Niyogi KK, Jahns P (2016) Photosystem II subunit PsbS is involved inthe induction of LHCSR protein-dependent energy dissipation inChlamydomonas reinhardtii. J Biol Chem 291: 17478–17487






Courchesne NMD, Parisien A, Wang B, Lan CQ (2009) Enhancement oflipid production using biochemical, genetic and transcription factorengineering approaches. J Biotechnol 141: 31–41

Davidi L, Katz A, Pick U (2012) Characterization of major lipid dropletproteins from Dunaliella. Planta 236: 19–33

Davis RW, Volponi JV, Jones HDT, Carvalho BJ, Wu H, Singh S (2012)Multiplex fluorometric assessment of nutrient limitation as a strategy forenhanced lipid enrichment and harvesting of Neochloris oleoabundans.Biotechnol Bioeng 109: 2503–2512

Deason T, Silva PC, Watanabe S, Floyd GL (1991) Taxonomic status of thespecies of the green algal genus Neochloris. Plant Syst Evol 177: 213–219

de Michele R, McFarlane HE, Parsons HT, Meents MJ, Lao J, GonzálezFernández-Niño SM, Petzold CJ, Frommer WB, Samuels AL, HeazlewoodJL (2016) Free-flow electrophoresis of plasma membrane vesicles enrichedby two-phase partitioning enhances the quality of the proteome fromArabidopsis seedlings. J Proteome Res 15: 900–913

Despres B, Bouissonnié F, Wu HJ, Gomord V, Guilleminot J, Grellet F,Berger F, Delseny M, Devic M (2003) Three SAC1-like genes showoverlapping patterns of expression in Arabidopsis but are remarkablysilent during embryo development. Plant J 34: 293–306

Dittami SM, Michel G, Collén J, Boyen C, Tonon T (2010) Chlorophyll-binding proteins revisited: a multigenic family of light-harvesting andstress proteins from a brown algal perspective. BMC Evol Biol 10: 365

Dong HP, Williams E, Wang DZ, Xie ZX, Hsia RC, Jenck A, Halden R, LiJ, Chen F, Place AR (2013) Responses of Nannochloropsis oceanica IMET1to long-term nitrogen starvation and recovery. Plant Physiol 162: 1110–1126

Einspahr KJ, Peeler TC, Thompson GA Jr (1988) Rapid changes in poly-phosphoinositide metabolism associated with the response of Dunaliellasalina to hypoosmotic shock. J Biol Chem 263: 5775–5779

Erickson E, Wakao S, Niyogi KK (2015) Light stress and photoprotection inChlamydomonas reinhardtii. Plant J 82: 449–465

Evans RJ, Schwarz N, Nagel-Wolfrum K, Wolfrum U, Hardcastle AJ,Cheetham ME (2010) The retinitis pigmentosa protein RP2 links peri-centriolar vesicle transport between the Golgi and the primary cilium.Hum Mol Genet 19: 1358–1367

Fatland BL, Ke J, Anderson MD, Mentzen WI, Cui LW, Allred CC,Johnston JL, Nikolau BJ, Wurtele ES (2002) Molecular characterizationof a heteromeric ATP-citrate lyase that generates cytosolic acetyl-coenzyme A in Arabidopsis. Plant Physiol 130: 740–756

Feldman JL, Marshall WF (2009) ASQ2 encodes a TBCC-like protein re-quired for mother-daughter centriole linkage and mitotic spindle ori-entation. Curr Biol 19: 1238–1243

Felsenstein J (1985) Confidence limits on phylogenies: an approach usingthe bootstrap. Evolution 39: 783–791

Ferro M, Brugière S, Salvi D, Seigneurin-Berny D, Court M, Moyet L,Ramus C, Miras S, Mellal M, Le Gall S, et al (2010) AT_CHLORO, acomprehensive chloroplast proteome database with subplastidial lo-calization and curated information on envelope proteins. Mol CellProteomics 9: 1063–1084

Ferro M, Salvi D, Brugière S, Miras S, Kowalski S, Louwagie M, Garin J,Joyard J, Rolland N (2003) Proteomics of the chloroplast envelopemembranes from Arabidopsis thaliana. Mol Cell Proteomics 2: 325–345

Fleischmann A, Darsow M, Degtyarenko K, Fleischmann W, Boyce S,Axelsen KB, BairochA, Schomburg D, Tipton KF, Apweiler R (2004) IntEnz,the integrated relational enzyme database. Nucleic Acids Res 32: D434–D437

Froehlich JE, Wilkerson CG, Ray WK, McAndrew RS, Osteryoung KW,Gage DA, Phinney BS (2003) Proteomic study of the Arabidopsisthaliana chloroplastic envelope membrane utilizing alternatives to tra-ditional two-dimensional electrophoresis. J Proteome Res 2: 413–425

Gao C, Wang Y, Shen Y, Yan D, He X, Dai J, Wu Q (2014) Oil accumulationmechanisms of the oleaginous microalga Chlorella protothecoides re-vealed through its genome, transcriptomes, and proteomes. BMC Ge-nomics 15: 582

Garibay-Hernández A, Vazquez-Duhalt R, Serrano-Carreón L, MartinezA (2013) Nitrogen limitation in Neochloris oleoabundans: a reassess-ment of its effect on cell growth and biochemical composition. ApplBiochem Biotechnol 171: 1775–1791

Gerotto C, Alboresi A, Giacometti GM, Bassi R, Morosinotto T (2012)Coexistence of plant and algal energy dissipation mechanisms in themoss Physcomitrella patens. New Phytol 196: 763–773

Gerotto C, Morosinotto T (2013) Evolution of photoprotection mechanismsupon land colonization: evidence of PSBS-dependent NPQ in lateStreptophyte algae. Physiol Plant 149: 583–598

Gidda SK, Shockey JM, Rothstein SJ, Dyer JM, Mullen RT (2009) Ara-bidopsis thaliana GPAT8 and GPAT9 are localized to the ER and possessdistinct ER retrieval signals: functional divergence of the dilysine ERretrieval motif in plant cells. Plant Physiol Biochem 47: 867–879

Ginger ML, McFadden GI, Michels PAM (2010) Rewiring and regulationof cross-compartmentalized metabolism in protists. Philos Trans R SocLond B Biol Sci 365: 831–845

Giovanardi M, Ferroni L, Baldisserotto C, Tedeschi P, Maietti A, Pantaleoni L,Pancaldi S (2013) Morphophysiological analyses of Neochloris oleoabundans(Chlorophyta) grown mixotrophically in a carbon-rich waste product. Proto-plasma 250: 161–174

Gontcharov AA, Marin B, Melkonian M (2004) Are combined analysesbetter than single gene phylogenies? A case study using SSU rDNA andrbcL sequence comparisons in the Zygnematophyceae (Streptophyta).Mol Biol Evol 21: 612–624

Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, MitrosT, Dirks W, Hellsten U, Putnam N, et al (2012) Phytozome: a com-parative platform for green plant genomics. Nucleic Acids Res 40:D1178–D1186

Götz S, García-Gómez JM, Terol J, Williams TD, Nagaraj SH, Nueda MJ,Robles M, Talón M, Dopazo J, Conesa A (2008) High-throughputfunctional annotation and data mining with the Blast2GO suite. Nu-cleic Acids Res 36: 3420–3435

Gouveia L, Marques AE, da Silva TL, Reis A (2009) Neochloris ole-abundans UTEX #1185: a suitable renewable lipid source for biofuelproduction. J Ind Microbiol Biotechnol 36: 821–826

Grayson C, Bartolini F, Chapple JP, Willison KR, Bhamidipati A, LewisSA, Luthert PJ, Hardcastle AJ, Cowan NJ, Cheetham ME (2002) Lo-calization in the human retina of the X-linked retinitis pigmentosaprotein RP2, its homologue cofactor C and the RP2 interacting proteinArl3. Hum Mol Genet 11: 3065–3074

Guarnieri MT, Nag A, Smolinski SL, Darzins A, Seibert M, Pienkos PT(2011) Examination of triacylglycerol biosynthetic pathways via de novotranscriptomic and proteomic analyses in an unsequenced microalga.PLoS ONE 6: e25851

Guarnieri MT, Nag A, Yang S, Pienkos PT (2013) Proteomic analysis ofChlorella vulgaris: potential targets for enhanced lipid accumulation. JProteomics 93: 245–253

Guiry MD, Guiry GM (1996) AlgaeBase. http://www.algaebase.org.Accessed May, 2016.

Haas BJ, Papanicolaou A, Yassour M, Grabherr M, Blood PD, Bowden J,Couger MB, Eccles D, Li B, Lieber M, et al (2013) De novo transcriptsequence reconstruction from RNA-seq using the Trinity platform forreference generation and analysis. Nat Protoc 8: 1494–1512

Harris EH, editor (2009) The Chlamydomonas Sourcebook: Introduction toChlamydomonas and Its Laboratory Use, Ed 2. Elsevier Academic Press,Toronto, Ontario, Canada

Haslam TM, Kunst L (2013) Extending the story of very-long-chain fattyacid elongation. Plant Sci 210: 93–107

Heilmann I, Perera IY, Gross W, Boss WF (2001) Plasma membranephosphatidylinositol 4,5-bisphosphate levels decrease with time in cul-ture. Plant Physiol 126: 1507–1518

Horn PJ, James CN, Gidda SK, Kilaru A, Dyer JM, Mullen RT, OhlroggeJB, Chapman KD (2013) Identification of a new class of lipid droplet-associated proteins in plants. Plant Physiol 162: 1926–1936

Hu Q, Sommerfeld M, Jarvis E, Ghirardi M, Posewitz M, Seibert M,Darzins A (2008) Microalgal triacylglycerols as feedstocks for biofuelproduction: perspectives and advances. Plant J 54: 621–639

Huss VAR, Frank C, Hartmann EC, Hirmer M, Kloboucek A, Seidel BM,Wenzeler P, Kessler E (1999) Biochemical taxonomy and molecularphylogeny of the genus Chlorella sensu lato (Chlorophyta). J Phycol 35:587–598

Iskandarov U, Sitnik S, Shtaida N, Didi-Cohen S, Leu S, Khozin-Goldberg I,Cohen Z, Boussiba S (2016) Cloning and characterization of a GPAT-likegene from the microalga Lobosphaera incisa (Trebouxiophyceae): over-expression in Chlamydomonas reinhardtii enhances TAG production. J ApplPhycol 28: 907–919

Iwai M, Takizawa K, Tokutsu R, Okamuro A, Takahashi Y, Minagawa J(2010) Isolation of the elusive supercomplex that drives cyclic electronflow in photosynthesis. Nature 464: 1210–1213

Jinkerson RE, Subramanian V, Posewitz MC (2011) Improving biofuelproduction in phototrophic microorganisms with systems biology. Bio-fuels 2: 125–144




http://www.algaebase.org


Johnson X, Alric J (2013) Central carbon metabolism and electron transportin Chlamydomonas reinhardtii: metabolic constraints for carbon parti-tioning between oil and starch. Eukaryot Cell 12: 776–793

Jones DT, Taylor WR, Thornton JM (1992) The rapid generation of mu-tation data matrices from protein sequences. Comput Appl Biosci 8: 275–282

Joyard J, Ferro M, Masselon C, Seigneurin-Berny D, Salvi D, Garin J,Rolland N (2010) Chloroplast proteomics highlights the subcellularcompartmentation of lipid metabolism. Prog Lipid Res 49: 128–158

Kleffmann T, Russenberger D, von Zychlinski A, Christopher W, SjölanderK, Gruissem W, Baginsky S (2004) The Arabidopsis thaliana chloroplastproteome reveals pathway abundance and novel protein functions. Curr Biol14: 354–362

Klein U (1986) Compartmentation of glycolysis and of the oxidativepentose-phosphate pathway in Chlamydomonas reinhardtii. Planta 167:81–86

Koziol AG, Borza T, Ishida K, Keeling P, Lee RW, Durnford DG (2007)Tracing the evolution of the light-harvesting antennae in chlorophyll a/b-containing organisms. Plant Physiol 143: 1802–1816

Krogh A, Larsson B, von Heijne G, Sonnhammer ELL (2001) Predictingtransmembrane protein topology with a hidden Markov model: appli-cation to complete genomes. J Mol Biol 305: 567–580

Kumar S, Stecher G, Tamura K (2016) MEGA7: Molecular EvolutionaryGenetics Analysis version 7.0 for bigger datasets. Mol Biol Evol 33:1870–1874

Laemmli UK (1970) Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature 227: 680–685

Levine RB, Costanza-Robinson MS, Spatafora GA (2011) Neochlorisoleoabundans grown on anaerobically digested dairy manure for con-comitant nutrient removal and biodiesel feedstock production. BiomassBioenergy 35: 40–49

Li J, Han D, Wang D, Ning K, Jia J, Wei L, Jing X, Huang S, Chen J, Li Y,et al (2014) Choreography of transcriptomes and lipidomes of Nanno-chloropsis reveals the mechanisms of oil synthesis in microalgae. PlantCell 26: 1645–1665

Li N, Xu C, Li-Beisson Y, Philippar K (2016) Fatty acid and lipid transportin plant cells. Trends Plant Sci 21: 145–158

Li X, Moellering ER, Liu B, Johnny C, Fedewa M, Sears BB, Kuo MH,Benning C (2012) A galactoglycerolipid lipase is required for triacyl-glycerol accumulation and survival following nitrogen deprivation inChlamydomonas reinhardtii. Plant Cell 24: 4670–4686

Li XP, Björkman O, Shih C, Grossman AR, Rosenquist M, Jansson S,Niyogi KK (2000) A pigment-binding protein essential for regulation ofphotosynthetic light harvesting. Nature 403: 391–395

Li XP, Gilmore AM, Caffarri S, Bassi R, Golan T, Kramer D, Niyogi KK(2004) Regulation of photosynthetic light harvesting involves intra-thylakoid lumen pH sensing by the PsbS protein. J Biol Chem 279:22866–22874

Li XP, Phippard A, Pasari J, Niyogi KK (2002) Structure-function analysisof photosystem II subunit S (PsbS) in vivo. Funct Plant Biol 29: 1131–1139

Li Y, Horsman M, Wang B, Wu N, Lan CQ (2008) Effects of nitrogensources on cell growth and lipid accumulation of green alga Neochlorisoleoabundans. Appl Microbiol Biotechnol 81: 629–636

Li-Beisson Y, Beisson F, Riekhof W (2015) Metabolism of acyl-lipids inChlamydomonas reinhardtii. Plant J 82: 504–522

Li-Beisson Y, Shorrosh B, Beisson F, Andersson MX, Arondel V, BatesPD, Baud S, Bird D, Debono A, Durrett TP, et al (2010) Acyl-lipidmetabolism. The Arabidopsis Book e0133, doi/10.1199/tab.0133

Liu B, Benning C (2013) Lipid metabolism in microalgae distinguishes it-self. Curr Opin Biotechnol 24: 300–309

Liu J, Last RL (2015a) MPH1 is a thylakoid membrane protein involved inprotecting photosystem II from photodamage in land plants. PlantSignal Behav 10: e1076602

Liu J, Last RL (2015b) A land plant-specific thylakoid membrane proteincontributes to photosystem II maintenance in Arabidopsis thaliana. Plant J82: 731–743

Lundin VF, Leroux MR, Stirling PC (2010) Quality control of cytoskeletalproteins and human disease. Trends Biochem Sci 35: 288–297

Ma Q, Wang J, Lu S, Lv Y, Yuan Y (2013) Quantitative proteomic profilingreveals photosynthesis responsible for inoculum size dependent varia-tion in Chlorella sorokiniana. Biotechnol Bioeng 110: 773–784

Manna PT, Gadelha C, Puttick AE, Field MC (2015) ENTH and ANTHdomain proteins participate in AP2-independent clathrin-mediated en-docytosis. J Cell Sci 128: 2130–2142

Matich EK, Butryn DM, Ghafari M, del Solar V, Camgoz E, Pfeifer BA,Aga DS, Haznedaroglu BZ, Atilla-Gokcumen GE (2016) Massspectrometry-based metabolomics of value-added biochemicals fromEttlia oleoabundans. Algal Res 19: 146–154

McCormac DJ, Barkan A (1999) A nuclear gene in maize required for thetranslation of the chloroplast atpB/E mRNA. Plant Cell 11: 1709–1716

Miller R, Wu G, Deshpande RR, Vieler A, Gärtner K, Li X, MoelleringER, Zäuner S, Cornish AJ, Liu B, et al (2010) Changes in transcriptabundance in Chlamydomonas reinhardtii following nitrogen depriva-tion predict diversion of metabolism. Plant Physiol 154: 1737–1752

Minagawa J, Tokutsu R (2015) Dynamic regulation of photosynthesis inChlamydomonas reinhardtii. Plant J 82: 413–428

Misumi O, Matsuzaki M, Nozaki H, Miyagishima SY, Mori T, Nishida K,Yagisawa F, Yoshida Y, Kuroiwa H, Kuroiwa T (2005) Cyanidioschyzonmerolae genome: a tool for facilitating comparable studies on organellebiogenesis in photosynthetic eukaryotes. Plant Physiol 137: 567–585

Mitchell A, Chang HY, Daugherty L, Fraser M, Hunter S, Lopez R,McAnulla C, McMenamin C, Nuka G, Pesseat S, et al (2015) The In-terPro protein families database: the classification resource after15 years. Nucleic Acids Res 43: D213–D221

Mitra SK, Walters BT, Clouse SD, Goshe MB (2009) An efficient organicsolvent based extraction method for the proteomic analysis of Arabi-dopsis plasma membranes. J Proteome Res 8: 2752–2767

Moellering ER, Benning C (2010) RNA interference silencing of a majorlipid droplet protein affects lipid droplet size in Chlamydomonas rein-hardtii. Eukaryot Cell 9: 97–106

Moon-van der Staay SY, De Wachter R, Vaulot D (2001) Oceanic 18SrDNA sequences from picoplankton reveal unsuspected eukaryotic di-versity. Nature 409: 607–610

Morales-Sánchez D, Tinoco-Valencia R, Kyndt J, Martinez A (2013)Heterotrophic growth of Neochloris oleoabundans using glucose as acarbon source. Biotechnol Biofuels 6: 100

Moreira D, Le Guyader H, Philippe H (2000) The origin of red algae andthe evolution of chloroplasts. Nature 405: 69–72

Mou S, Zhang X, Dong M, Fan X, Xu J, Cao S, Xu D, Wang W, Ye N (2013)Photoprotection in the green tidal alga Ulva prolifera: role of LHCSR and PsbSproteins in response to high light stress. Plant Biol (Stuttg) 15: 1033–1039

Mühlhaus T, Weiss J, Hemme D, Sommer F, Schroda M (2011) Quanti-tative shotgun proteomics using a uniform 15N-labeled standard tomonitor proteome dynamics in time course experiments reveals newinsights into the heat stress response of Chlamydomonas reinhardtii.Mol Cell Proteomics 10: M110.004739

Natter K, Leitner P, Faschinger A, Wolinski H, McCraith S, Fields S,Kohlwein SD (2005) The spatial organization of lipid synthesis in theyeast Saccharomyces cerevisiae derived from large scale green fluores-cent protein tagging and high resolution microscopy. Mol Cell Proteo-mics 4: 662–672

Ndimba BK, Ndimba RJ, Johnson TS, Waditee-Sirisattha R, Baba M,Sirisattha S, Shiraiwa Y, Agrawal GK, Rakwal R (2013) Biofuels as asustainable energy source: an update of the applications of proteomicsin bioenergy crops and algae. J Proteomics 93: 234–244

Nesvizhskii AI, Keller A, Kolker E, Aebersold R (2003) A statistical modelfor identifying proteins by tandem mass spectrometry. Anal Chem 75:4646–4658

NewlonMG, RoyM,Morikis D, Carr DW,Westphal R, Scott JD, Jennings PA(2001) A novel mechanism of PKA anchoring revealed by solution structuresof anchoring complexes. EMBO J 20: 1651–1662

Nguyen HM, Baudet M, Cuiné S, Adriano JM, Barthe D, Billon E, BruleyC, Beisson F, Peltier G, Ferro M, et al (2011) Proteomic profiling of oilbodies isolated from the unicellular green microalga Chlamydomonasreinhardtii: with focus on proteins involved in lipid metabolism. Pro-teomics 11: 4266–4273

Niyogi KK, Truong TB (2013) Evolution of flexible non-photochemicalquenching mechanisms that regulate light harvesting in oxygenic pho-tosynthesis. Curr Opin Plant Biol 16: 307–314

Nordhues A, Schöttler MA, Unger AK, Geimer S, Schönfelder S,Schmollinger S, Rütgers M, Finazzi G, Soppa B, Sommer F, et al (2012)Evidence for a role of VIPP1 in the structural organization of the pho-tosynthetic apparatus in Chlamydomonas. Plant Cell 24: 637–659

Olguín EJ, Castillo OS, Mendoza A, Tapia K, González-Portela RE,Hernández-Landa VJ (2015) Dual purpose system that treats anaerobiceffluents from pig waste and produce Neochloris oleoabundans as lipidrich biomass. N Biotechnol 32: 387–395





Ou YY, Gromiha MM, Chen SA, Suwa M (2008) TMBETADISC-RBF:discrimination of beta-barrel membrane proteins using RBF networksand PSSM profiles. Comput Biol Chem 32: 227–231

Parry RV, Turner JC, Rea PA (1989) High purity preparations of higherplant vacuolar H+-ATPase reveal additional subunits: revised subunitcomposition. J Biol Chem 264: 20025–20032

Peers G, Truong TB, Ostendorf E, Busch A, Elrad D, Grossman AR,Hippler M, Niyogi KK (2009) An ancient light-harvesting protein iscritical for the regulation of algal photosynthesis. Nature 462: 518–521

Pegg C, Wolf M, Alanagreh L, Portman R, Buchheim MA (2015) Mor-phological diversity masks phylogenetic similarity of Ettlia and Hae-matococcus (Chlorophyceae). Phycologia 54: 385–397

Peled E, Leu S, Zarka A, Weiss M, Pick U, Khozin-Goldberg I, Boussiba S(2011) Isolation of a novel oil globule protein from the green algaHaematococcus pluvialis (Chlorophyceae). Lipids 46: 851–861

Peltier JB, Cai Y, Sun Q, Zabrouskov V, Giacomelli L, Rudella A, YtterbergAJ, Rutschow H, van Wijk KJ (2006) The oligomeric stromal pro-teome of Arabidopsis thaliana chloroplasts. Mol Cell Proteomics 5:114–133

Pérez-Martínez X, Antaramian A, Vázquez-Acevedo M, Funes S, TolkunovaE, d’Alayer J, Claros MG, Davidson E, King MP, González-Halphen D(2001) Subunit II of cytochrome c oxidase in chlamydomonad algae is aheterodimer encoded by two independent nuclear genes. J Biol Chem276: 11302–11309

Pérez-Rodríguez P, Riaño-Pachón DM, Corrêa LGG, Rensing SA, KerstenB, Mueller-Roeber B (2010) PlnTFDB: updated content and new fea-tures of the plant transcription factor database. Nucleic Acids Res 38:D822–D827

Popovich CA, Damiani C, Constenla D, Martínez AM, Freije H, GiovanardiM, Pancaldi S, Leonardi PI (2012) Neochloris oleoabundans grown inenriched natural seawater for biodiesel feedstock: evaluation of its growthand biochemical composition. Bioresour Technol 114: 287–293

Poulsen LR, López-Marqués RL, McDowell SC, Okkeri J, Licht D, SchulzA, Pomorski T, Harper JF, Palmgren MG (2008) The Arabidopsis P4-ATPase ALA3 localizes to the Golgi and requires a b-subunit to func-tion in lipid translocation and secretory vesicle formation. Plant Cell 20:658–676

Pruvost J, Van Vooren G, Cogne G, Legrand J (2009) Investigation of bio-mass and lipids production with Neochloris oleoabundans in photo-bioreactor. Bioresour Technol 100: 5988–5995

Rawlings ND, Waller M, Barrett AJ, Bateman A (2014) MEROPS: the da-tabase of proteolytic enzymes, their substrates and inhibitors. NucleicAcids Res 42: D503–D509

Ren J, Pei-Chen Lin C, Pathak MC, Temple BRS, Nile AH, Mousley CJ,Duncan MC, Eckert DM, Leiker TJ, Ivanova PT, et al (2014) A phos-phatidylinositol transfer protein integrates phosphoinositide signalingwith lipid droplet metabolism to regulate a developmental program ofnutrient stress-induced membrane biogenesis. Mol Biol Cell 25: 712–727

Reumann S, Quan S, Aung K, Yang P, Manandhar-Shrestha K, HolbrookD, Linka N, Switzenberg R, Wilkerson CG, Weber APM, et al (2009)In-depth proteome analysis of Arabidopsis leaf peroxisomes combinedwith in vivo subcellular targeting verification indicates novel metabolicand regulatory functions of peroxisomes. Plant Physiol 150: 125–143

Rismani-Yazdi H, Haznedaroglu BZ, Hsin C, Peccia J (2012) Transcriptomicanalysis of the oleaginous microalga Neochloris oleoabundans revealsmetabolic insights into triacylglyceride accumulation. Biotechnol Biofuels5: 74

Rodríguez-Salinas E, Riveros-Rosas H, Li Z, Fucíková K, Brand JJ, LewisLA, González-Halphen D (2012) Lineage-specific fragmentation andnuclear relocation of the mitochondrial cox2 gene in chlorophyceangreen algae (Chlorophyta). Mol Phylogenet Evol 64: 166–176

Saier MH Jr, Reddy VS, Tamang DG, Västermark A (2014) The trans-porter classification database. Nucleic Acids Res 42: D251–D258

Schmidt M, Gessner G, Luff M, Heiland I, Wagner V, Kaminski M,Geimer S, Eitzinger N, Reissenweber T, Voytsekh O, et al (2006)Proteomic analysis of the eyespot of Chlamydomonas reinhardtii providesnovel insights into its components and tactic movements. Plant Cell 18:1908–1930

Schmollinger S, Mühlhaus T, Boyle NR, Blaby IK, Casero D, Mettler T,Moseley JL, Kropat J, Sommer F, Strenkert D, et al (2014) Nitrogen-sparing mechanisms in Chlamydomonas affect the transcriptome, theproteome, and photosynthetic metabolism. Plant Cell 26: 1410–1435

Schnoes AM, Brown SD, Dodevski I, Babbitt PC (2009) Annotation errorin public databases: misannotation of molecular function in enzymesuperfamilies. PLOS Comput Biol 5: e1000605

Schwacke R, Schneider A, van der Graaff E, Fischer K, Catoni E, DesimoneM,Frommer WB, Flügge UI, Kunze R (2003) ARAMEMNON, a noveldatabase for Arabidopsis integral membrane proteins. Plant Physiol 131:16–26

Shi LX, Hall M, Funk C, Schröder WP (2012) Photosystem II, a growingcomplex: updates on newly discovered components and low molecularmass proteins. Biochim Biophys Acta 1817: 13–25

Shi LX, Schröder WP (2004) The low molecular mass subunits of thephotosynthetic supracomplex, photosystem II. Biochim Biophys Acta1608: 75–96

Shockey J, Regmi A, Cotton K, Adhikari N, Browse J, Bates PD (2016)Identification of Arabidopsis GPAT9 (At5g60620) as an essential geneinvolved in triacylglycerol biosynthesis. Plant Physiol 170: 163–179

Shtaida N, Khozin-Goldberg I, Boussiba S (2015) The role of pyruvate hubenzymes in supplying carbon precursors for fatty acid synthesis inphotosynthetic microalgae. Photosynth Res 125: 407–422

Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, Lopez R,McWilliam H, Remmert M, Söding J, et al (2011) Fast, scalable gen-eration of high-quality protein multiple sequence alignments usingClustal Omega. Mol Syst Biol 7: 539

Singer SD, Chen G, Mietkiewska E, Tomasi P, Jayawardhane K, Dyer JM,Weselake RJ (2016) Arabidopsis GPAT9 contributes to synthesis of in-tracellular glycerolipids but not surface lipids. J Exp Bot 67: 4627–4638

Singh DK, McNellis TW (2011) Fibrillin protein function: the tip of theiceberg? Trends Plant Sci 16: 432–441

Škunca N, Altenhoff A, Dessimoz C (2012) Quality of computationallyinferred Gene Ontology annotations. PLOS Comput Biol 8: e1002533

Smith SR, Abbriano RM, Hildebrand M (2012) Comparative analysis ofdiatom genomes reveals substantial differences in the organization ofcarbon partitioning pathways. Algal Res 1: 2–16

Song P, Li L, Liu J (2013) Proteomic analysis in nitrogen-deprived Iso-chrysis galbana during lipid accumulation. PLoS ONE 8: e82188

Stabenau H, Winkler U, Säftel W (1984) Enzymes of b-oxidation in dif-ferent types of algal microbodies. Plant Physiol 75: 531–533

Stabenau H, Winkler U, Säftel W (1993) Localization of glycolate dehy-drogenase in two species of Dunaliella. Planta 191: 362–364

Stephan A, Vaughan S, Shaw MK, Gull K, McKean PG (2007) An essentialquality control mechanism at the eukaryotic basal body prior to intra-flagellar transport. Traffic 8: 1323–1330

Stern D, editor (2009) The Chlamydomonas Sourcebook: Organellar and Meta-bolic Processes, Ed 2. Elsevier Academic Press, Toronto, Ontario, Canada

Tamura K, Nei M (1993) Estimation of the number of nucleotide substi-tutions in the control region of mitochondrial DNA in humans andchimpanzees. Mol Biol Evol 10: 512–526

Tan S, Tan HT, Chung MCM (2008) Membrane proteins and membraneproteomics. Proteomics 8: 3924–3932

Tanz SK, Castleden I, Hooper CM, Vacher M, Small I, Millar HA (2013)SUBA3: a database for integrating experimentation and prediction todefine the SUBcellular location of proteins in Arabidopsis. Nucleic AcidsRes 41: D1185–D1191

Tardif M, Atteia A, Specht M, Cogne G, Rolland N, Brugière S, HipplerM, Ferro M, Bruley C, Peltier G, et al (2012) PredAlgo: a new subcel-lular localization prediction tool dedicated to green algae. Mol Biol Evol29: 3625–3639

Terashima M, Specht M, Naumann B, Hippler M (2010) Characterizingthe anaerobic response of Chlamydomonas reinhardtii by quantitativeproteomics. Mol Cell Proteomics 9: 1514–1532

Tibiletti T, Auroy P, Peltier G, Caffarri S (2016) Chlamydomonas rein-hardtii PsbS protein is functional and accumulates rapidly and tran-siently under high light. Plant Physiol 171: 2717–2730

Tippery NP, Fu�cíková K, Lewis PO, Lewis LA (2012) Probing the mono-phyly of the Sphaeropleales (Chlorophyceae) using data from five genes.J Phycol 48: 1482–1493

Tornabene TG, Holzer G, Lien S, Burris N (1983) Lipid composition of thenitrogen starved green alga Neochloris oleoabundans. Enzyme MicrobTechnol 5: 435–440

Tusnády GE, Simon I (2001) The HMMTOP transmembrane topologyprediction server. Bioinformatics 17: 849–850

UniProt Consortium (2015) UniProt: a hub for protein information. NucleicAcids Res 43: D204–D212





Valledor L, Furuhashi T, Recuenco-Muñoz L, Wienkoop S, Weckwerth W(2014) System-level network analysis of nitrogen starvation and recov-ery in Chlamydomonas reinhardtii reveals potential new targets forincreased lipid accumulation. Biotechnol Biofuels 7: 171

Vandepoele K, Van Bel M, Richard G, Van Landeghem S, Verhelst B,Moreau H, Van de Peer Y, Grimsley N, Piganeau G (2013) pico-PLAZA, a genome database of microbial photosynthetic eukaryotes.Environ Microbiol 15: 2147–2153

Vázquez-Acevedo M, Cardol P, Cano-Estrada A, Lapaille M, Remacle C,González-Halphen D (2006) The mitochondrial ATP synthase ofchlorophycean algae contains eight subunits of unknown origin in-volved in the formation of an atypical stator-stalk and in the dimer-ization of the complex. J Bioenerg Biomembr 38: 271–282

Vázquez-Acevedo M, Vega-deLuna F, Sánchez-Vásquez L, Colina-Tenorio L,Remacle C, Cardol P, Miranda-Astudillo H, González-Halphen D (2016)Dissecting the peripheral stalk of the mitochondrial ATP synthase of chlor-ophycean algae. Biochim Biophys Acta 1857: 1183–1190

Vera-Estrella R, Barkla BJ, Bohnert HJ, Pantoja O (2004) Novel regulationof aquaporins during osmotic stress. Plant Physiol 135: 2318–2329

Verrier PJ, Bird D, Burla B, Dassa E, Forestier C, Geisler M, Klein M,Kolukisaoglu U, Lee Y, Martinoia E, et al (2008) Plant ABC proteins: aunified nomenclature and updated inventory. Trends Plant Sci 13: 151–159

Wang B, Lan CQ (2011) Biomass production and nitrogen and phosphorusremoval by the green alga Neochloris oleoabundans in simulatedwastewater and secondary municipal wastewater effluent. BioresourTechnol 102: 5639–5644

Wang J, Virta VC, Riddelle-Spencer K, O’Halloran TJ (2003) Compromiseof clathrin function and membrane association by clathrin light chaindeletion. Traffic 4: 891–901

Wang X, Liu Q, Zhang B (2014) Leveraging the complementary nature ofRNA-Seq and shotgun proteomics data. Proteomics 14: 2676–2687

Wang Z, Benning C (2012) Chloroplast lipid synthesis and lipid traffickingthrough ER-plastid membrane contact sites. Biochem Soc Trans 40: 457–463

Wase N, Black PN, Stanley BA, DiRusso CC (2014) Integrated quantitativeanalysis of nitrogen stress response in Chlamydomonas reinhardtii us-ing metabolite and protein profiling. J Proteome Res 13: 1373–1396

Wei L, Xin Y, Wang D, Jing X, Zhou Q, Su X, Jia J, Ning K, Chen F, Hu Q,et al (2013) Nannochloropsis plastid and mitochondrial phylogenomesreveal organelle diversification mechanism and intragenus phylotypingstrategy in microalgae. BMC Genomics 14: 534

Wildgruber R, Weber G, Wise P, Grimm D, Bauer J (2014) Free-flowelectrophoresis in proteome sample preparation. Proteomics 14: 629–636

Wu N, Li Y, Lan CQ (2011) Production and rheological studies of micro-algal extracellular biopolymer from lactose using the green alga Neo-chloris oleoabundans. J Polym Env 19: 935–942

Yang P, Diener DR, Yang C, Kohno T, Pazour GJ, Dienes JM, Agrin NS,King SM, Sale WS, Kamiya R, et al (2006) Radial spoke proteins ofChlamydomonas flagella. J Cell Sci 119: 1165–1174

Yang Y, Mininberg B, Tarbet A, Weathers P (2013) At high temperaturelipid production in Ettlia oleoabundans occurs before nitrate depletion.Appl Microbiol Biotechnol 97: 2263–2273

Yang Y, Xu J, Vail D, Weathers P (2011) Ettlia oleoabundans growth andoil production on agricultural anaerobic waste effluents. BioresourTechnol 102: 5076–5082

Yoo C, Choi GG, Kim SC, Oh HM (2013) Ettlia sp. YC001 showing highgrowth rate and lipid content under high CO2. Bioresour Technol 127:482–488

Yoon K, Han D, Li Y, Sommerfeld M, Hu Q (2012) Phospholipid:diacyl-glycerol acyltransferase is a multifunctional enzyme involved in mem-brane lipid turnover and degradation while synthesizing triacylglycerolin the unicellular green microalga Chlamydomonas reinhardtii. Plant Cell24: 3708–3724

Zhang X, Ye N, Mou S, Xu D, Fan X (2013) Occurrence of the PsbS andLhcSR products in the green alga Ulva linza and their correlation withexcitation pressure. Plant Physiol Biochem 70: 336–341

Zhang Y, Wen Z, Washburn MP, Florens L (2010) Refinements to label freeproteome quantitation: how to deal with peptides shared by multipleproteins. Anal Chem 82: 2272–2281

Zhang ZJ, Peck SC (2011) Simplified enrichment of plasma membraneproteins for proteomic analyses in Arabidopsis thaliana. Proteomics 11:1780–1788

Zhou X, Graham TR (2009) Reconstitution of phospholipid translocaseactivity with purified Drs2p, a type-IV P-type ATPase from buddingyeast. Proc Natl Acad Sci USA 106: 16586–16591





membrane proteomic insights into the physiology and ... · metabolomics have become essential for...

Documents