Advertisement

Proximity labeling facilitates defining the proteome neighborhood of photosystem II oxygen evolution complex in a model cyanobacterium

  • Author Footnotes
    4 Co-first author.
    Zhen Xiao
    Footnotes
    4 Co-first author.
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Author Footnotes
    4 Co-first author.
    Chengcheng Huang
    Footnotes
    4 Co-first author.
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Haitao Ge
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China
    Search for articles by this author
  • Yan Wang
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Xiaoxiao Duan
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Gaojie Wang
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Limin Zheng
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Jinghui Dong
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Xiahe Huang
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China
    Search for articles by this author
  • Yuanya Zhang
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China
    Search for articles by this author
  • Hongyu An
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Wu Xu
    Affiliations
    Department of Chemistry, University of Louisiana at Lafayette, Lafayette, LA 70504, USA
    Search for articles by this author
  • Yingchun Wang
    Correspondence
    Corresponding author: Yingchun Wang, Ph.D. . Phone: 86-10-6480-6149, Fax: 86-10-6480-6252
    Affiliations
    State Key Laboratory of Molecular Developmental Biology, Innovation Academy for Seed Design, CAS, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, No.1 West Beichen Rd., Beijing 100101, China

    University of Chinese Academy of Sciences
    Search for articles by this author
  • Author Footnotes
    4 Co-first author.
Open AccessPublished:November 07, 2022DOI:https://doi.org/10.1016/j.mcpro.2022.100440

      Highlights

      • A robust workflow was developed for APEX2-based proximity labeling in a model cyanobacterium, Synechocystis sp. PCC6803.
      • Proteins proximal to the photosystem II oxygen evolution complex were identified through PsbO-APEX2 catalyzed proximity labeling.
      • The luminal side proteome neighborhood of photosystem II was defined to include 38 integral thylakoid membrane proteins and 93 luminal proteins.

      ABSTRACT

      Ascorbate peroxidase (APEX)-based proximity labeling coupled with mass spectrometry have a great potential for spatiotemporal identification of proteins proximal to a protein complex of interest. Using this approach is feasible to define the proteome neighborhood of important protein complexes in a popular photosynthetic model cyanobacterium Synechocystis sp. PCC6803 (hereafter named as Synechocystis). To this end, we developed a robust workflow for APEX2-based proximity labeling in Synechocystis, and used the workflow to identify proteins proximal to the photosystem II (PS II) oxygen evolution complex (OEC) through fusion APEX2 with a luminal OEC subunit, PsbO. In total, 38 integral membrane proteins (IMPs) and 93 luminal proteins were identified as proximal to the OEC. A significant portion of these proteins are involved in PS II assembly, maturation, and repair while the majority of the rest were not previously implicated with PS II. The IMPs include subunits of PS II and cytochrome b6/f (Cyt b6/f), but not of photosystem I (PS I) (except for PsaL) and ATP synthases, suggesting that the latter two complexes are spatially separated from the OEC with a distance longer than the APEX2 labeling radius. Besides, the topologies of 6 IMPs were successfully predicted because their lumen-facing regions exclusively contain potential APEX2 labeling sites. The luminal proteins include 66 proteins with a predicted signal peptide and 57 proteins localized also in periplasm, providing important targets to study the regulation and selectivity of protein translocation. Together, we not only developed a robust workflow for the application of APEX2-based proximity labeling in Synechocystis and showcased the feasibility to define the neighborhood proteome of an important protein complex with a short radius, but also discovered a set of the proteins that potentially interact with and regulate PS II structure and function.

      Graphical abstract

      KEYWORDS

      ABBREVIATIONS:

      Synechocystis (Synechocystis sp. PCC 6803), WT (Wild type), APEX (Ascorbate peroxidase), BP (Biotin phenol), H2O2 (Hydrogen peroxide), OEC (Oxygen evolution complex), PS I (Photosystem I), PS II (Photosystem II), IMP (Integral membrane protein), GAP2 (Glyceraldaldehyde-3-phosphate dehydrogenase 2), WCL (Whole cell lysate), FT (Flow through), TM (Transmembrane domain), Cyt b6/f (cytochrome b6/f), Sec (Secretory), TAT (Twin-arginine translocation)

      INTRODUCTION

      Cyanobacteria are a group of gram-negative bacteria capable of oxygenic photosynthesis. Cyanobacteria thrive in almost all kinds of terrestrial and aquatic habitats on the earth, and are an important component of our ecosystem by providing a significant portion of oxygen in the atmosphere (
      • Kasting J.F.
      • Siefert J.L.
      Life and the evolution of Earth's atmosphere.
      ). Cyanobacteria highly resembles the chloroplasts of the higher plants in both structure and function, and it is believed that the latter are derived from the former through endosymbiosis (
      • Goksoyr J.
      Evolution of eucaryotic cells.
      ), and therefore cyanobacteria are frequently used as model organisms for studying photosynthesis. Cyanobacteria have recently caught more attention because of their capability of photoautotrophic growth and the potential to serve as cell factory to produce clean and renewable biofuels and high value compounds (
      • Lubner C.E.
      • Applegate A.M.
      • Knörzer P.
      • Ganago A.
      • Bryant D.A.
      • Happe T.
      • Golbeck J.H.
      Solar hydrogen-producing bionanodevice outperforms natural photosynthesis.
      ,
      • Varman A.M.
      • Xiao Y.
      • Pakrasi H.B.
      • Tang Y.J.
      Metabolic engineering of Synechocystis sp. strain PCC 6803 for isobutanol production.
      ,
      • Xue Y.
      • Zhang Y.
      • Cheng D.
      • Daddy S.
      • He Q.
      Genetically engineering Synechocystis sp. Pasteur Culture Collection 6803 for the sustainable production of the plant secondary metabolite p-coumaric acid.
      ).
      To better use cyanobacteria both as model organisms for photosynthesis research and as green E.coli for industrial production, it is important to understand how diverse types of biological and physiological processes and metabolic pathways are operated and coordinated in a single compartment. This is evident as photosynthesis must be coordinated with carbon and nitrogen metabolisms, and production of renewable biofuels and high value compounds need metabolic engineering that diverts metabolic flux to the target metabolites. Proteins are essentail players in these processes and pathways, and they locate and operate in a single or multiple subcellular compartments. Determining the cellular locations of proteins is critically important to the understanding how the processes and pathways they are involved are regulated and coordinated in time and space. Therefore, mounting efforts have been invested on the determination of protein locations in cyanobacteria, particularly at a proteome level (
      • Gao L.
      • Ge H.
      • Huang X.
      • Liu K.
      • Zhang Y.
      • Xu W.
      • Wang Y.
      Systematically ranking the tightness of membrane association for peripheral membrane proteins (PMPs).
      ,
      • Huang F.
      • Parmryd I.
      • Nilsson F.
      • Persson A.L.
      • Pakrasi H.B.
      • Andersson B.
      • Norling B.
      Proteomics of Synechocystis sp. strain PCC 6803: identification of plasma membrane proteins.
      ,
      • Huang F.
      • Hedman E.
      • Funk C.
      • Kieselbach T.
      • Schroder W.P.
      • Norling B.
      Isolation of outer membrane of Synechocystis sp. PCC 6803 and its proteomic characterization.
      ,
      • Baers L.L.
      • Breckels L.M.
      • Mills L.A.
      • Gatto L.
      • Deery M.J.
      • Stevens T.J.
      • Howe C.J.
      • Lilley K.S.
      • Lea-Smith D.J.
      Proteome mapping of a cyanobacterium reveals distinct compartment organization and cell-dispersed metabolism.
      ,
      • Wang Y.
      • Xu W.
      • Chitnis P.R.
      Identification and bioinformatic analysis of the membrane proteins of synechocystis sp. PCC 6803.
      ,
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ). To this end, the major subcellular compartments of cyanobacteria such as thylakoid membrane, plasma membrane, outer membrane, periplasmic space, and even the extracellular proteome have been biochemically isolated, and the protein components in each compartments have been identified using either 2D-MS or LC-MS approaches (
      • Gao L.
      • Ge H.
      • Huang X.
      • Liu K.
      • Zhang Y.
      • Xu W.
      • Wang Y.
      Systematically ranking the tightness of membrane association for peripheral membrane proteins (PMPs).
      ,
      • Huang F.
      • Parmryd I.
      • Nilsson F.
      • Persson A.L.
      • Pakrasi H.B.
      • Andersson B.
      • Norling B.
      Proteomics of Synechocystis sp. strain PCC 6803: identification of plasma membrane proteins.
      ,
      • Huang F.
      • Hedman E.
      • Funk C.
      • Kieselbach T.
      • Schroder W.P.
      • Norling B.
      Isolation of outer membrane of Synechocystis sp. PCC 6803 and its proteomic characterization.
      ,
      • Baers L.L.
      • Breckels L.M.
      • Mills L.A.
      • Gatto L.
      • Deery M.J.
      • Stevens T.J.
      • Howe C.J.
      • Lilley K.S.
      • Lea-Smith D.J.
      Proteome mapping of a cyanobacterium reveals distinct compartment organization and cell-dispersed metabolism.
      ,
      • Wang Y.
      • Xu W.
      • Chitnis P.R.
      Identification and bioinformatic analysis of the membrane proteins of synechocystis sp. PCC 6803.
      ,
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ). Using these approaches, the majority of proteins encoded by the genome of a model cyanobacterium Synechocystis have been determined for their subcellular locations, and a proteome atlas has been constructed based on such information (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ). Though the atlas provides subcellular localization information for the majority of proteins, the spatial resolution of the atlas is still insufficient for defining protein residents in a specific domain of a subcellular compartment. It is documented that thylakoid membrane, and plasma membrane as well, are not homogenous, and contain different domains with different protein and lipid compositions (
      • Zhang L.
      • Selao T.T.
      • Selstam E.
      • Norling B.
      Subcellular Localization of Carotenoid Biosynthesis in Synechocystis sp. PCC 6803.
      ,
      • Rast A.
      • Schaffer M.
      • Albert S.
      • Wan W.
      • Pfeffer S.
      • Beck F.
      • Plitzko J.M.
      • Nickelsen J.
      • Engel B.D.
      Biogenic regions of cyanobacterial thylakoids form contact sites with the plasma membrane.
      ,
      • Agarwal R.
      • Matros A.
      • Melzer M.
      • Mock H.P.
      • Sainis J.K.
      Heterogeneity in thylakoid membrane proteome of Synechocystis 6803.
      ). It is conceivable that the major macromolecular protein complexes such as PS I and PS II may reside at particular domains with specific proteins in the neighborhood, and thereby to maintain their structural stability and optimal function. Moreover, the atlas does not contain information for the proteins in thylakoid lumen, a subcellular compartment of cyanobacteria that has never been isolated for proteome analysis due to the lack of an effective approach for biochemical separation and purification of the compartment.
      Recent technological innovations in proximity labeling technologies allow labeling and capturing proteins in the nearby of target proteins (
      • Rhee H.W.
      • Zou P.
      • Udeshi N.D.
      • Martell J.D.
      • Mootha V.K.
      • Carr S.A.
      • Ting A.Y.
      Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging.
      ). Among them, peroxidase proximity labeling with APEX and its upgraded version APEX2 are of particular interest because their short reaction time and small labeling radius (
      • Rhee H.W.
      • Zou P.
      • Udeshi N.D.
      • Martell J.D.
      • Mootha V.K.
      • Carr S.A.
      • Ting A.Y.
      Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging.
      ,
      • Lam S.S.
      • Martell J.D.
      • Kamer K.J.
      • Deerinck T.J.
      • Ellisman M.H.
      • Mootha V.K.
      • Ting A.Y.
      Directed evolution of APEX2 for electron microscopy and proximity labeling.
      ), which are important for spatiotemporally defining the neighborhood proteome of the protein of interest. It is possible to genetically fuse APEX2 with a subunit of the major macromolecular protein complexes on the thylakoid membrane such as PS I and PS II or a protein in thylakoid lumen, and to identify proteins in the neighborhood of the corresponding protein complexes or in the thylakoid lumen. To date, APEX2 has been successfully used in different systems including yeast, animal cell lines, and tissues (
      • Hung V.
      • Lam S.S.
      • Udeshi N.D.
      • Svinkina T.
      • Guzman G.
      • Mootha V.K.
      • Carr S.A.
      • Ting A.Y.
      Proteomic mapping of cytosol-facing outer mitochondrial and ER membranes in living human cells by proximity biotinylation.
      ,
      • Singer-Kruger B.
      • Frohlich T.
      • Franz-Wachtel M.
      • Nalpas N.
      • Macek B.
      • Jansen R.P.
      APEX2-mediated proximity labeling resolves protein networks in Saccharomyces cerevisiae cells.
      ,
      • Chen C.L.
      • Hu Y.
      • Udeshi N.D.
      • Lau T.Y.
      • Wirtz-Peitz F.
      • He L.
      • Ting A.Y.
      • Carr S.A.
      • Perrimon N.
      Proteomic mapping in live Drosophila tissues using an engineered ascorbate peroxidase.
      ). In cyanobacteria, only one study was available regarding APEX2 application. In that report APEX2 was targeted to the thylakoid lumen of Synechococcus sp. PCC 7002 through fusion with a PS II subunit PsbU, and 123 biotinylated proteins were subsequently identified as luminal proteins (
      • Dahlgren K.K.
      • Gates C.
      • Lee T.
      • Cameron J.C.
      Proximity-based proteomics reveals the thylakoid lumen proteome in the cyanobacterium Synechococcus sp. PCC 7002.
      ).
      Herein, we firstly optimized a workflow that can serve as a general protocol for APEX2 application in cyanobacterium Synechocystis, a popular model organism for photosynthesis research because of its completely sequenced genome and the natural ability of transformation (
      • Ikeuchi M.
      • Tabata S.
      Synechocystis sp. PCC 6803 - a useful tool in the study of the genetics of cyanobacteria.
      ,
      • Kaneko T.
      • Sato S.
      • Kotani H.
      • Tanaka A.
      • Asamizu E.
      • Nakamura Y.
      • Miyajima N.
      • Hirosawa M.
      • Sugiura M.
      • Sasamoto S.
      • Kimura T.
      • Hosouchi T.
      • Matsuno A.
      • Muraki A.
      • Nakazaki N.
      • Naruo K.
      • Okumura S.
      • Shimpo S.
      • Takeuchi C.
      • Wada T.
      • Watanabe A.
      • Yamada M.
      • Yasuda M.
      • Tabata S.
      Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions.
      ). Then we used the protocol to identify proteins proximal to the extrinsic OEC of PS II through fusion of APEX2 with PsbO, an OEC subunit. A number of integral thylakoid membrane proteins and luminal proteins were identified. A small portion of the identified proteins are known to be proximal to PS II or in lumen, whereas the other majority are not previously known representing novel targets potentially involved in regulation of PS II structure and function.

      EXPERIMENTAL PROCEDURES

      Antibodies

      The primary antibodies for GAP2, CpcB were provided by PhytoAB (San Jose, CA). The antibody for HA was purchased from MBL (Beijing, China). The antibody for D1 was purchased from Agrisera (Vännäs, Sweden). The antibody for GFP was provided by Abcam (Waltham, MA). The antibody for biotin was provided by Cell Signaling (Danvers, MA). The streptavidin-HRP was provided by Thermo Scientific (Rockford, IL).

      Cell Culture

      Synechocystis cells (the glucose-tolerant strain) were grown in liquid BG11 medium on a shaker (200 rpm) at 30°Cwith a photosynthetic photon flux density of 50 μmol m−2 s−1. The concentration of cells in liquid culture was estimated from the optical density at 730 nm (OD730) using a SmartSpec Plus spectrophotometer (Bio-Rad, Hercules, CA).

      Mutant Generation

      The knock-in strains of Synechocystis expressing APEX2/HA-tagged target proteins were generated as described previously (
      • Gao L.
      • Shen C.
      • Liao L.
      • Huang X.
      • Liu K.
      • Wang W.
      • Guo L.
      • Jin W.
      • Huang F.
      • Xu W.
      • Wang Y.
      Functional proteomic discovery of Slr0110 as a central regulator of carbohydrate metabolism in Synechocystis species PCC6803.
      ). Briefly, a ∼1 kb DNA fragment of target genes spanning from about 500 bp to the 5’ prime of the stop codon to about 500 bp to the 3’ prime of the stop codon were cloned by PCR using the primers listed in supplemental Table S1. A DNA fragment containing APEX2 or HA tag and a kanamycin resistance gene cassette was inserted into the target gene fragment before the stop codon. The final constructs were used to transform wild type (WT) Synechocystis, and the knock-in strains were selected on solid BG11 plates containing kanamycin. The fully segregated strains were confirmed by PCR as previously described (
      • Golden S.S.
      • Brusslan J.
      • Haselkorn R.
      Genetic engineering of the cyanobacterial chromosome.
      ).

      Biotin-phenol Labeling

      The cyanobacterial culture at the exponential phase (OD730 = 1.0) was diluted to a cell density of OD730 = 0.2 and grown until OD730 = 0.8. Cells were harvested by centrifugation at 4000 g for 5 min and resuspended in BG11 containing 2 mM BP, and incubated at 30°C for 10 min. H2O2 was then added to a final concentration of 2 mM and the labeling reaction was allowed to continue for 1 min. The reaction was quenched by replacing the medium with ice cold quenching buffer (TBS supplemented with 1 mM CaCl2, 10 mM sodium ascorbate, 1 mM sodium azide, and 1 mM Trolox).

      Enrichment of Biotin-labeled Proteins with Streptavidin

      Cells were lysed in a RM buffer (50 mM Tris, 150 mM NaCl, 1% NP40, 0.1% deoxycholate, 1mM EDTA, pH 7.5, and 0.5 mM phenylmethanesulfonyl fluoride (PMSF, Sigma-Aldrich, Saint Louis, MO)) with a bead beater, and the insoluble debris was removed by centrifugation with 5000 g for 30 min at 4°C. After precipitation with methanol and chloroform (
      • Wessel D.
      • Flugge U.I.
      A method for the quantitative recovery of protein in dilute-solution in the presence of detergents and lipids.
      ), total proteins were washed with methanol to remove pigments, lipids, and residual chloroform. The precipitated proteins were air-dried before re-solubilization with 0.1% sodium dodecyl sulfate (SDS) in 0.1 M Tris-HCl (pH 7.6). The protein concentrations were measured using a BCA protein assay kit (Thermo Scientific, Rockford, IL). To enrich biotinylated proteins, 4 mg of proteins were added to 100 μL of streptavidin-agarose beads (Thermo Scientific, Rockford, IL), which were then rotated and incubated overnight at 4°C. The beads were subsequently washed once with RM buffer, twice with PBST, once with 1 M KCl. For streptavidin blotting analysis, biotinylated proteins were then eluted by boiling the beads in 50 μL of 3×protein loading buffer supplemented with 20 mM DTT and 2 mM biotin and run on SDS-PAGE gel.

      On-bead Trypsin Digestion and Desalting

      On-bead trypsin digestion was performed in the same way as described previously (
      • Fukuyama H.
      • Ndiaye S.
      • Hoffmann J.
      • Rossier J.
      • Liuu S.
      • Vinh J.
      • Verdier Y.
      On-bead tryptic proteolysis: an attractive procedure for LC-MS/MS analysis of the Drosophila caspase 8 protein complex during immune response against bacteria.
      ). Briefly, proteins bound to streptavidin beads were first reduced by adding DTT to a final concentration of 10 mM and incubating at 37˚C for 1h with shaking, and then alkylated by adding iodoacetamide (IAA, Sigma-Aldrich, Saint Louis, MO) to a final concentration of 55 mM and incubating at 25˚C for 45 min with shaking. Finally, trypsin (Sigma-Aldrich, Saint Louis, MO) was added to digest proteins by overnight incubation at 37˚C with shaking. The peptides were acidified by acetic acid and desalted using C18 StageTips (
      • Rappsilber J.
      • Ishihama Y.
      • Mann M.
      Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics.
      ). Desalted peptides were dried with a Speed-Vac concentrator, stored at -20 ˚C, and resuspended in 0.1% formic acid (FA) immediately before LC-MS/MS.

      Mass Spectrometry

      For MS analysis, the peptides were analyzed by an LTQ Elite mass spectrometer (Thermo Fisher Scientific, Waltham, MA) coupled online to an Easy-nLC 1000 in a data-dependent mode. Briefly, peptide sample was injected into a 25 cm length, 75 μm inner diameter capillary analytic column packed with C18 particles of 5 μm diameter (SunChrom, Friedrichsdorf, Germany). The LC gradient was composed of 3% to 8% buffer B (buffer B contained 100% ACN and 0.1% FA, whereas buffer A contained 0.1% FA) for 10 min, 8% to 20% buffer B for 60 min, 20% to 30% buffer B for 8 min, 30% to 100% buffer B for 2 min, and 100% buffer B for 10 min. The flow rate was set at 600 nL/min. The source voltage was set at 2.5 KV, and the current was set at 100 μA. MS measurement was performed in positive ion mode. The precursors were measured by survey scans in the Orbitrap with a mass range of 300 to 1800 m/z. The 20 most abundant precursor ions from each survey scan were isolated and fragmented by collisional dissociation (CID) for MS/MS analysis. The duration of dynamic exclusion was set to 30 s to prevent repeat identification of peptide ions within the time duration.

      Database Search

      The raw MS files were searched against the Synechocystis proteome sequence database appended with 248 common contaminations using the software MaxQuant (version 1.6.0.16) (
      • Cox J.
      • Mann M.
      MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification.
      ). The database containing 3672 entries was downloaded from CyanoBase (ftp://ftp.kazusa.or.jp/pub/CyanoBase/Synechocystis, released on 5/11/2009). Trypsin/P was set as the protease for protein digestion, and up to two missed cleavages were allowed. N-terminal acetylation and methionine oxidation were chosen as the variable modifications and cysteine carbamidomethylation was chosen as the fixed modification. For precursor ions, the mass tolerance was specified at 20 ppm for the first search and 4.5 ppm for the main search. For fragment ions, the mass tolerance was set to 20 ppm. The 1% false discovery rate (FDR) was set at both the peptide and protein levels. The minimum scores for unmodified peptides and modified peptides were set to 15 and 40, respectively. The match between runs was enabled. All other parameters were set to default values of MaxQuant.

      Experimental Design and Statistical Rationale

      Four label-free quantitative experiments were performed to identify of proteins proximal to PsbO. The mutant cells expressing PsbO-APEX2 incubated with BP and H2O2 were used as the experimental group. The mutant cells expressing PsbO-APEX2 incubated with H2O2 but without BP, the WT and the mutant cells expressing Zwf-APEX2 incubated with BP and H2O2 were included as the three negative controls. Three biological replicates of each treatment were included. After searching the database, the label-free quantitative intensities were used to calculate the ratios for each protein between the experimental group and the controls. Only proteins identified with MS/MS count >1 in the experimental group were included for all quantitative analyses.

      Bioinformatics and Statistics

      Bioinformatic and statistical analyses were mainly performed using the software Perseus version 1.5.2.6 (
      • Cox J.
      • Mann M.
      1D and 2D annotation enrichment: a statistical method integrating quantitative proteomics with complementary high-throughput data.
      ). Enrichment analysis was performed with Fisher’s exact test. A p-value<0.05 was used as the cut-off for all statistical analyses.

      RESULTS

      Exogenous Expression of Active APEX2 in Synechocystis

      To test whether APEX2 is active for proximity labeling in Synechocystis, a mutant strain expressing APEX2-GFP under the control of the promoter of psbAII (PPsbAII-APEX2), a gene encoding the PS II D1 protein, was generated by inserting the exogenous DNA fragment into the chromosome through homologous recombination at a neutral site where the ORF slr0168 resides (Fig. 1A) (
      • Mohamed A.
      • Eriksson J.
      • Osiewacz H.D.
      • Jansson C.
      Differential expression of the psbA genes in the cyanobacterium Synechocystis 6803.
      ). The complete segregation of the mutant was confirmed by PCR and the expression of APEX2-GFP was detected by Western blotting (Fig. 1B). To test the peroxidase activity of APEX2, the mutant cells were incubated with or without biotin-phenol (BP) for 10 min, and then treated with 1 mM hydrogen peroxide (H2O2) to activate the catalytic biotinylation reaction (
      • Rhee H.W.
      • Zou P.
      • Udeshi N.D.
      • Martell J.D.
      • Mootha V.K.
      • Carr S.A.
      • Ting A.Y.
      Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging.
      ). The WT treated with both BP and H2O2 was also included as the control. The whole cell lysates (WCLs) were probed for biotinylated proteins by Western blotting. Strong biotinylation signals were only detected in the lane for PPsbAII-APEX2 cells treated with BP, but not in the other lanes (Fig. 1C). The data indicates that APEX2 can be exogenously expressed and activated, and could be used for proximity labeling in Synechocystis. APEX2 catalyzes the reduction of H2O2 to water by oxidizing ascorbic acid, and the reaction is important for regulating the level of toxic oxidative radicals (
      • Mandelman D.
      • Schwarz F.P.
      • Li H.
      • Poulos T.L.
      The role of quaternary interactions on the stability and activity of ascorbate peroxidase.
      ). To investigate whether the exogenous expression of APEX2 causes growth defects in Synechocystis, the growth phenotype of the mutant cells was examined. The PPsbAII-APEX2 showed barely observable growth defect under the autotrophic and mixotrophic growth conditions in comparison with the WT (Fig. 1D). Together, these results suggest that active APEX2 could be expressed for biotinylation of the endogenous proteins of Synechocystis without causing observable growth defect in the host cells.
      Figure thumbnail gr1
      Fig. 1Expression of active APEX2 in Synechocystis. A, the diagram depicts the plasmid construct used to express APEX2-GFP fusion protein at a neutral genomic site (slr0168 locus) under the control of the psbAII promoter in Synechocystis. B, confirmation of the complete segregation of the mutant by PCR (left panel) and the expression of APEX2-GFP by Western blotting (right panel). The same pair of primers were used to amplify DNA fragments from the WT and the mutant. GAP2 blot and Ponceau staining were included as the loading control. C, Western blotting detection of endogenous proteins biotin-labeled by APEX2-GFP. Samples were labeled for 1 min, the BP omitted sample and the WT were included as the negative controls. D, growth curves of the mutant and the WT strains measured in the presence (photomixotrophic) or absence (photoautotrophic) of 5 mM glucose under medium light intensity (50 μmol m-2s-1) (Left panel). The cultures were photographed at the indicated time points during the time course (Right panel).

      Optimization of the Experimental Conditions for APEX2-catalyzed Protein Biotinylation

      To achieve an optimal efficiency of APEX2-catalyzed protein biotinylation in Synechocystis, we tested the concentration and incubation time for BP and H2O2 in the reaction. For mammalian cell culture, 0.5 mM BP was sufficient to label proteins (
      • Martell J.D.
      • Deerinck T.J.
      • Sancak Y.
      • Poulos T.L.
      • Mootha V.K.
      • Sosinsky G.E.
      • Ellisman M.H.
      • Ting A.Y.
      Engineered ascorbate peroxidase as a genetically encoded reporter for electron microscopy.
      ). Thus we tested BP concentrations below or above 0.5 mM, ranging from 0.25 to 4 mM. Western-blotting detection of biotinylated proteins from the WCLs of 10 min BP-treated cells shows that biotinylation is detectable for the cells treated with 0.5 mM BP, and increases along with the increasing concentration of the BP. Because BP is expensive and excess BP may affect the subsequent affinity purification as shown below (Fig. 2E), we chose the concentration 2 mM instead of a higher one for further experiments (Fig. 2, A and H). To optimize the time for BP incubation, the cells were incubated with BP from 10 min to 90 min, a time scale that covers the time usually used for mammalian cells (i.e., 30 min) (
      • Lobingier B.T.
      • Huttenhain R.
      • Eichel K.
      • Miller K.B.
      • Ting A.Y.
      • von Zastrow M.
      • Krogan N.J.
      An approach to spatiotemporally resolve protein interaction networks in living cells.
      ). Western-blotting of the WCLs showed that there was no apparent change of the labeled proteins regardless of the incubation time with BP. Thus we chose 10 min as the BP incubation time hereafter (Fig. 2, B and H).
      Figure thumbnail gr2
      Fig. 2Optimization of the experimental conditions for APEX2 catalyzed labeling and affinity purification of biotinylated endogenous proteins. A-D, optimization working conditions for APEX2 including BP concentration (A) and incubation time (B), H2O2 concentration (C) and labeling time (D) at room temperature. Activity and expression of APEX2-GFP fusion proteins were analyzed by streptavidin or anti-biotin Western blotting as indicated. Ponceau staining was used as the loading control. E-G, optimization of the affinity purification of biotinylated proteins by streptavidin beads. WCLs of PPsbAII-APEX2 were directly incubated with streptavidin agarose (E), or filtered with 3 kDa ultra centrifugal filters and followed by incubation with streptavidin agarose (F), or precipitated with methanol/chloroform before incubation with streptavidin agarose (G). The input, FT and the eluates were separated by SDS-PAGE and probed for biotinylated proteins by Western blotting. H, the bar graph shows the relative intensity of the biotinylated proteins in different conditions shown in (A, B, C and D). Image J was used for the quantitative densitometry analysis on a representative image from each triplicated experiments.
      The appropriate H2O2 concentration is also an important factor for efficient APEX2 catalyzed protein biotinylation. Low H2O2 may be not sufficient for the reaction and high H2O2 may produce free radicals leading to oxidative damage to proteins and membrane lipids. Therefore, we tested H2O2 concentration ranging from 0.5 to 16 mM to determine the optimal concentration. Weak protein biotinylation can be detected for the cells treated with 0.5 or 1 mM H2O2, and strong biotinylation was detected for the cells treated with 2 mM H2O2. Further increase of H2O2 concentration did not increase the level of protein biotinylation (Fig. 2, C and H). Thus, we chose 2 mM as the working concentration for H2O2. The optimal working time for H2O2 was also tested because long time H2O2 incubation could cause damage to the cells resulting in experimental artifacts. Encouragingly, one minute of H2O2 treatment is sufficient for efficient APEX2-catalyzed biotinylation as detected by the immunoblotting (Fig. 2, D and H). Together, the experimental condition was optimized for APEX2-GFP catalyzed protein biotinylation in Synechocystis, and the same condition will be used hereafter for the APEX2-based proximity labeling.

      Optimization of the Affinity Purification of Proteins Labeled by APEX2-catalyzed Biotinylation

      Affinity purification of the biotinylated proteins was tried using streptavidin beads. Unexpectedly, the purification was not successful and no biotin signal was detected in the eluates (Fig. 2E). Increasing the amount of the starting material, i.e. the WCLs, did not solve the problem. It is suggested that the existence of free BP in the WCLs, which is the leftover of the AEPX2-catalyzed reaction, could competitively inhibit the binding of the biotinylated proteins to the streptavidin beads (
      • Mair A.
      • Xu S.L.
      • Branon T.C.
      • Ting A.Y.
      • Bergmann D.C.
      Proximity labeling of protein complexes and cell-type-specific organellar proteomes in Arabidopsis enabled by TurboID.
      ). Existing evidence shows that for removal of free BP from the WCLs, thorough washing of the cells is sufficient for mammalian cell culture or rice protoplasts, whereas an additional filtration step is needed for plant tissue (
      • Mair A.
      • Xu S.L.
      • Branon T.C.
      • Ting A.Y.
      • Bergmann D.C.
      Proximity labeling of protein complexes and cell-type-specific organellar proteomes in Arabidopsis enabled by TurboID.
      ). Therefore, we filtered the WCLs using ultrafilters with a 3 kDa molecular weight cutoff before performing streptavidin affinity purification. As expected, the biotinylated proteins in the filtered WCLs successfully bound to streptavidin beads and showed strong signals in the eluates when probed by streptavidin. In contrast, the biotinylation signal was barely displayed in the flow through (FT) indicating high efficiency of affinity purification (Fig. 2F). Since the preparation of the WCLs and the subsequent affinity purification were performed in a non-denaturing condition, and many non-biotinylated proteins could interact and co-precipitate with the biotinylated proteins. To address this problem, the WCLs were denatured by precipitation with methanol/chloroform and re-solubilized in a buffer containing 0.1% SDS before streptavidin immunoprecipitation. It is notable that the precipitation can also effectively remove the free BP from the sample, making the filtration step avoidable. Successful biotin labeling and purification of biotinylated proteins were confirmed by immunoblotting (Fig. 2G). Taking the above in consideration, we chose the precipitation approach to prepare further samples for affinity purification of biotinylated proteins.

      Proximity Labeling of Proteins in the Neighborhood of PsbO-APEX2

      To express APEX2 in the thylakoid lumen, a knock-in mutant was generated to express PsbO-APEX2, with an HA tag inserted between the C-terminus of PsbO and N-terminus of APEX2 (Fig. 3A). PsbO is a subunit of the OEC and resides in the thylakoid lumen, and PsbO-APEX2 is expected to label proteins proximal to the PS II OEC complex, which should include luminal proteins and integral thylakoid membrane proteins with domains exposed to lumen (Fig. 3B). The complete segregation of the mutant was confirmed by PCR and the expression of the fusion protein was confirmed by Western-blotting using anti-HA antibody (Fig. 4A).
      Figure thumbnail gr3
      Fig. 3Proximity labeling of proteins by PsbO-APEX2. A, the diagram depicts the plasmid construct used to transform Synechocystis for the expression of PsbO-APEX2. B, diagram illustration of PsbO-APEX2 catalyzed biotin-labeling of proteins proximal to the PSII OEC complex. Note that both luminal proteins and integral thylakoid membrane proteins with domains exposed to the lumen can be labeled.
      Figure thumbnail gr4
      Fig. 4Affinity purification of proteins labeled by PsbO-APEX2. A, the complete segregation and expression of PsbO-APEX2 was confirmed by PCR (left panel) and Western blotting (right panel), respectively. Equal loading was shown by Ponceau staining. B, affinity purification and Western-blotting detection of the biotinylated proteins from the indicated cells incubated with or without BP. The streptavidin-HRP was used detect biotinylated proteins in the input, FT, and the eluates. C, Western-blotting detection of PsbO-HA-APEX2, D1, and CpcB in both the inputs and the eluates. D, same as in (B) except that PsbO-APEX2 is replaced by Zwf-APEX2, an additional negative control.
      PsbO-APEX2-catalyzed proximity labeling and affinity purification of the biotinylated proteins were performed using the optimized condition as described above (Fig. 2). The biotinylated proteins in the WCLs and the eluates of the affinity purification were probed with streptavidin. Again, successful protein biotinylation was catalyzed by PsbO-APEX2, and the affinity purification of biotinylated proteins was highly efficient as strong signals were displayed in both input and the eluates, but not in the FT (Fig. 4B). Note that the WT cells treated with BP and the PsbO-APEX2 cells without BP treatment were also included as the control to evaluate the background protein biotinylation (Fig. 4B). In the case of the WT cells treated with BP, weak protein biotinylation was detected due to the existence of the endogenous peroxidase activity (
      • Mironov K.S.
      • Sinetova M.A.
      • Shumskaya M.
      • Los D.A.
      Universal molecular triggers of stress responses in cyanobacterium Synechocystis.
      ). The almost negligible weak biotinylation signal of the control indicates that the PsbO-APEX2-catalyzed protein biotinylation is highly specific. The efficiency of affinity purification was further confirmed with the visualization of the enriched proteins stained with coomassie blue or silver staining (supplemental Fig. S1), and much less and weaker protein bands were detected in the controls.
      PsbO locates at the thylakoid lumen and PsbO-APEX2 is expected to catalyze the biotinylation of luminal proteins or the thylakoid membrane proteins with domains exposed to the lumen, but not the cytoplasmic proteins or the thylakoid membrane proteins without a domain exposed to the lumen. To examine such a spatial specificity of the labeling, the eluates of the affinity purification were probed for three proteins including PsbO, D1, and CpcB. The result showed that high spatial specificity was achieved as indicated by the detection of PsbO and D1 protein but not CpcB in the eluates for PsbO-APEX2 treated with BP (Fig. 4C). D1 is an integral membrane protein with multiple transmembrane domains (TMs), and therefore have multiple inter-TM loops as well as the C-terminus exposed to the thylakoid lumen (
      • Chang W.
      • Li C.
      • Cui Z.
      • Li W.
      • Song H.
      • Chang H.
      • Fu W.
      • Wang C.
      • Huang T.
      • Luo Y.
      • Shan Y.
      • Wang Y.
      • Wang F.
      • Xu M.
      • Fu A.
      Diverged early from CtpB and CtpC, CtpA has evolved to process D1 precursor in oxygenic photosynthetic organisms.
      ), and is accessible for proximity labeling by PsbO-APEX2. CpcB is a subunit of phycobilisome that locates at the cytoplasmic side of the thylakoids, and is not accessible for PsbO-APEX2 catalyzed labeling.
      All thylakoid luminal proteins are translated outside of the lumen and transported into the lumen via different mechanisms (
      • Frain K.M.
      • Gangl D.
      • Jones A.
      • Zedler J.A.
      • Robinson C.
      Protein translocation and thylakoid biogenesis in cyanobacteria.
      ). Although PsbO, and PsbO-APEX2 as well, mainly locate at the thylakoid lumen, existence of low amount of newly synthesized PsbO-APEX2 in the cytoplasm is possible. To further exclude the proteins labeled by such a particular fraction of PsbO-APEX2, a knock-in mutant with APEX2 fused at the C-terminus of glucose-6-phosphate, a known cytoplasmic protein encoded by the gene zwf (Zwf-APEX2) (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ), was generated. The proteins biotinylated by Zwf-APEX2 were effectively purified by affinity purification and detected by streptavidin blotting, which show a distinct pattern from those biotinylated by the PsbO-APEX2 (Fig. 4, B and D), suggesting that a different set of proteins were labeled. Therefore, the Zwf-APEX2-labeled proteins were also included as a negative control to remove potential cytoplasmic background. Using the combination of multiple negative controls would ensure high confident identification of thylakoid lumen specific proteins within the proximity of PsbO.

      Quantitative Identification of the Proteins Proximal to the OEC

      The strategy for the identification of proteins proximal to PsbO is illustrated in a diagram (Fig. 5A). The mutant cells expressing PsbO-APEX2 or Zwf-APEX2 were incubated with 2 mM BP for 10min and followed by incubation with 2 mM H2O2 for 1 min. The WT treated in the same way was included as the control, and the mutant expressing PsbO-APEX2 incubated with H2O2 but without BP was included as an additional control as described above. Three biological replicates of each treatment were included for label-free quantitative proteomic comparison (
      • Cox J.
      • Hein M.Y.
      • Luber C.A.
      • Paron I.
      • Nagaraj N.
      • Mann M.
      Accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction, termed MaxLFQ.
      ). The biotinylated proteins in each treatment were purified by streptavidin affinity purification as described above (Figs. 2 and 4), and then on-bead digested by trypsin (
      • Fukuyama H.
      • Ndiaye S.
      • Hoffmann J.
      • Rossier J.
      • Liuu S.
      • Vinh J.
      • Verdier Y.
      On-bead tryptic proteolysis: an attractive procedure for LC-MS/MS analysis of the Drosophila caspase 8 protein complex during immune response against bacteria.
      ). The tryptic peptides were subsequently analyzed with LC-MS (Fig. 5A).
      Figure thumbnail gr5
      Fig. 5Quantitative identification of the proteins proximal to OEC. A, schematic representation of the workflow for quantitative proteomic identification of the proteins proximal to the OEC. Three negative controls (Ctrls 1-3) were included. B, the diagramed workflow illustrates the step-by-step filtration to remove false positive identifications using the indicated criteria. The 867 proteins identified in the target experiment were sequentially filtered and only 131 proteins remained. The x-axis in each volcano plot indicates the logarithm-transformed ratio of the relative abundance of a protein between the two indicated samples. The y-axis: the logarithm-transformed p-value of the Student’s t-test. The subunits of PS II were used as the positive landmarks which remained in the list after the filtering, and the subunits of phycobilisome were used as the negative landmarks which were removed in the first two filtering steps. The number of remaining proteins in each step are shown in parenthesis.
      The raw MS files were searched against the Synechocystis proteome database using MaxQuant (version 1.6.0.16). In total, 1202 proteins were identified with a 1% FDR (supplemental Table S2). Only proteins identified in at least two replicates of the target experiment and with MS/MS count greater than 1 were included for further analysis. Filtered with this criterion 867 proteins were identified as the biotinylated proteins labeled by PsbO-APEX2 (Fig. 5B). The reproducibility of the quantitation was evaluated by the pairwise comparison of the MS intensities among the triplicate samples. In all comparisons, high correlation coefficients were obtained with a minimum R2 equal to 0.94, indicative of high reproducibility of the quantitation (supplemental Fig. S2).
      To narrow down the number of positive identifications, the 867 proteins were sequentially compared with the proteins identified in the three negative controls to remove background identifications step by step. Student’s t-test p < 0.05 and a fold-change of 2 were used as the criteria to filter for positive identifications in each comparison (Fig. 5B). In each filtering step a significant portion of proteins were removed as illustrated in the 2D-scatter plots. Such a multi-step sequential filtering with the stringent criteria resulted in 131 remaining proteins (Fig. 5B and supplemental Table S3). All the three subunits of the OEC (PsbO, PsbU, and PsbV) were included in the list which can be used as the positive controls. All the subunits of the phycobilisome, which were used as the negative controls, were excluded from the list. As such the 131 proteins were determined as the proteins proximal to PsbO-APEX2 with high confidence.

      Enriched Functions in the Proteins Proximal to the OEC

      The enriched functions among the 131 proteins were examined using Fisher’s exact test against gene ontology (GO) terms and CyanoBase functional categories (
      • Nakao M.
      • Okamoto S.
      • Kohara M.
      • Fujishiro T.
      • Fujisawa T.
      • Sato S.
      • Tabata S.
      • Kaneko T.
      • Nakamura Y.
      CyanoBase: the cyanobacteria genome database update 2010.
      ) (Fig. 6). For the CyanoBase functions, only two categories are enriched, namely, photosynthesis and respiration and PS II. For GO terms, one term in biological process (GOBP) and one term in molecular function (GOMF) are enriched, which are protein repair and peptidase activities, respectively (Fig. 6). The data suggests that proteases are highly needed in the proximity to the OEC, either for protein processing such as cleavage of the c-terminal tail of the precursor of the PS II reaction center D1 protein (pD1) by the protease CtpA (
      • Anbudurai P.R.
      • Mor T.S.
      • Ohad I.
      • Shestakov S.V.
      • Pakrasi H.B.
      The ctpA gene encodes the C-terminal processing protease for the D1 protein of the photosystem II reaction center complex.
      ), or for protein quality control such as degradation of photodamaged D1 by FtsH2/FtsH3 (
      • Boehm M.
      • Yu J.
      • Krynicka V.
      • Barker M.
      • Tichy M.
      • Komenda J.
      • Nixon P.J.
      • Nield J.
      Subunit organization of a synechocystis hetero-oligomeric thylakoid FtsH complex involved in photosystem II repair.
      ). In contrast, a number of terms in cellular component (GOCC) are enriched. The majority of the GOCC terms are related with thylakoid, thylakoid lumen, and particularly PS II but not PSI (Fig. 6). This result strongly suggests that the specificity of the current identification of proteins proximal to PS II OEC is high. Remarkably, periplasmic space is one of the most significantly enriched GOCC terms (Fig. 6), indicating that many proteins proximal to OEC are also residents in periplasm.
      Figure thumbnail gr6
      Fig. 6Functions enriched in the identified proteins proximal to the OEC. The enrichment assay was performed using Fisher’s exact test, and an enrichment factor >2 and a p-value < 0.05 were used together as the cutoffs. Bars show the significance of enrichment (p-value), and the enrichment factor was also shown to the right of each bar. CyanoBase: the functional categories in CyanoBase. GOBP, GOCC, and GOMF are the three types of gene ontologies.

      Integral Membrane Proteins Proximal to the OEC

      Since the identified proteins contain both bona fide luminal proteins and membrane proteins with domains exposed to the lumen, it is important to separate these two types of proteins. Topology analysis using the program TMHMM predicts that 70 proteins contain at least one TM, including 20 proteins with 2 or more TMs (Fig. 7A). Signal peptide prediction for proteins with less than 2 TMs reveals that 66 proteins contain signal peptide at the N-terminal, including 26 proteins that were also predicted as 1-TM containing (Fig. 7B). For 26 overlapping proteins that are predicted as both signal peptide-containing and 1-TM-containing proteins, the predicted regions for signal peptide and TM are almost completely overlapping for each protein. Since predictions of both signal peptide and TM are essentially based on the existence of a hydrophobic core, despite there may be some differences in details such as the special amino acids needed at the N-terminus of signal peptide-containing proteins (
      • Frain K.M.
      • Gangl D.
      • Jones A.
      • Zedler J.A.
      • Robinson C.
      Protein translocation and thylakoid biogenesis in cyanobacteria.
      ,
      • Krogh A.
      • Larsson B.
      • von Heijne G.
      • Sonnhammer E.L.
      Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes.
      ), it is not easy to determine whether they are IMPs or bona fide luminal proteins solely based on computer prediction. Therefore, we also include literature evidence for the determination of true IMPs among those containing one TM or a signal peptide at N-terminal (
      • Baers L.L.
      • Breckels L.M.
      • Mills L.A.
      • Gatto L.
      • Deery M.J.
      • Stevens T.J.
      • Howe C.J.
      • Lilley K.S.
      • Lea-Smith D.J.
      Proteome mapping of a cyanobacterium reveals distinct compartment organization and cell-dispersed metabolism.
      ,
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ). Together, 38 proteins were determined as IMPs, and 93 proteins were determined as luminal proteins and collectively referred to as luminal proteins. These also include 21 proteins that contain neither predicted TM nor signal peptide. The latter may be transported to the lumen via an unknown pathway (Fig. 7B and supplemental Table S3) (
      • Frain K.M.
      • Gangl D.
      • Jones A.
      • Zedler J.A.
      • Robinson C.
      Protein translocation and thylakoid biogenesis in cyanobacteria.
      ).
      Figure thumbnail gr7
      Fig. 7Prediction of integral membrane proteins and luminal proteins that are proximal to the OEC. A, the bar graph shows the distribution of the number of proteins containing the indicated numbers of TM. The prediction was preformed using the software TMHMM 1.0 (
      • Moller S.
      • Croning M.D.
      • Apweiler R.
      Evaluation of methods for the prediction of membrane spanning regions.
      ). B, the Venn diagram shows the numbers of TM-containing proteins, signal peptide-containing proteins, and proteins contain neither TM nor signal peptide. Signal peptide prediction was performed using the software SignalP 5.0 (
      • Almagro Armenteros J.J.
      • Tsirigos K.D.
      • Sonderby C.K.
      • Petersen T.N.
      • Winther O.
      • Brunak S.
      • von Heijne G.
      • Nielsen H.
      SignalP 5.0 improves signal peptide predictions using deep neural networks.
      ).
      It is conceivable that IMPs are less mobile than soluble proteins as their horizontal movements on the membrane are restrained by the surrounding lipids. Therefore, the identified IMPs better represent the in vivo residents in the proximity to the OEC complex, within a small APEX-labeling radius (< 20 nm) (
      • Rhee H.W.
      • Zou P.
      • Udeshi N.D.
      • Martell J.D.
      • Mootha V.K.
      • Carr S.A.
      • Ting A.Y.
      Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging.
      ). The OEC neighborhood IMPs can be categorized into four groups (supplemental Table S3). First, the PS II subunits. These include D1, D2, and PsbL. Second, proteins involved in PS II assembly, maturation, and damage-degradation-repair cycle. These include HliD, YidC, Slr0575, FtsH2, FtsH3, and Psb32. HliD is a small one-TM containing membrane protein belonging to the high-light induced protein (Hlip) family, and was found to be associated with RCII complex and is involved in early stage assembly of PS II (
      • Knoppová J.
      • Sobotka R.
      • Tichy M.
      • Yu J.
      • Konik P.
      • Halada P.
      • Nixon P.J.
      • Komenda J.
      Discovery of a chlorophyll binding protein complex involved in the early steps of photosystem II assembly in Synechocystis.
      ,
      • Staleva H.
      • Komenda J.
      • Shukla M.K.
      • Slouf V.
      • Kana R.
      • Polivka T.
      • Sobotka R.
      Mechanism of photoprotection in the cyanobacterial ancestor of plant antenna proteins.
      ). Moreover, HliD forms a complex together with chlorophyll synthase and the general insertase YidC, and potentially involved in delivering chlorophyll to nascent apoproteins of PS II pigment-protein complex (
      • Chidgey J.W.
      • Linhartova M.
      • Komenda J.
      • Jackson P.J.
      • Dickman M.J.
      • Canniffe D.P.
      • Konik P.
      • Pilny J.
      • Hunter C.N.
      • Sobotka R.
      A cyanobacterial chlorophyll synthase-HliD complex associates with the Ycf39 protein and the YidC/Alb3 insertase.
      ). YidC (Slr1471) is the Synechocystis Albino3/Oax1 homolog and is essential for membrane integration of pD1. YidC directly interacts with D1 (
      • Ossenbuhl F.
      • Inaba-Sulpice M.
      • Meurer J.
      • Soll J.
      • Eichacker L.A.
      The synechocystis sp PCC 6803 oxa1 homolog is essential for membrane integration of reaction center precursor protein pD1.
      ), and is essential for thylakoid biogenesis (
      • Spence E.
      • Bailey S.
      • Nenninger A.
      • Moller S.G.
      • Robinson C.
      A homolog of Albino3/OxaI is essential for thylakoid biogenesis in the cyanobacterium Synechocystis sp. PCC6803.
      ). The hypothetical protein Slr0575 is also involved in PS II assembly. Slr0575 locates primarily on the thylakoid membrane (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ), and was identified as a subunit of the PS II assembly intermediate RCII (
      • Zhou L.
      • Ye Y.
      Assembly of the photosystem II reaction center, the heart of photosynthesis.
      ). FtsH2 and FtsH3 are two proteases forming hetero-oligomeric complex that is essential for D1 degradation (
      • Boehm M.
      • Yu J.
      • Krynicka V.
      • Barker M.
      • Tichy M.
      • Komenda J.
      • Nixon P.J.
      • Nield J.
      Subunit organization of a synechocystis hetero-oligomeric thylakoid FtsH complex involved in photosystem II repair.
      ,
      • Silva P.
      • Thompson E.
      • Bailey S.
      • Kruse O.
      • Mullineaux C.W.
      • Robinson C.
      • Mann N.H.
      • Nixon P.J.
      FtsH is involved in the early stages of repair of photosystem II in Synechocystis sp PCC 6803.
      ). FtsH2, FtsH3, and FtsH4 locate mainly on the thylakoid membrane, whereas FtsH1 locates mainly on the plasma membrane (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ). FtsH4 was also identified as a OEC neighborhood protein, though its function and substrates are not well known (
      • Yao D.
      • Kieselbach T.
      • Komenda J.
      • Promnares K.
      • Prieto M.A.
      • Tichy M.
      • Vermaas W.
      • Funk C.
      Localization of the small CAB-like proteins in photosystem II.
      ). In contrast, FtsH1 was not identified. Psb32 (Sll1390) is associated with PS II and involved in protection of PS II from photodamage and acceleration of its repair (
      • Wegener K.M.
      • Bennewitz S.
      • Oelmüller R.
      • Pakrasi H.B.
      The Psb32 protein aids in repairing photodamaged photosystem II in the cyanobacterium Synechocystis 6803.
      ).
      Third, proteins functionally connected with PS II (supplemental Table S3). These include PetA, PetC1, CcsA, CcsB, Slr0404, Slr0814, and Hik33. PetA, PetC1, CcsA, CcsB are related with Cyt b6/f complex, one of the main protein complexes of the photosynthetic electron transport locating at the downstream of PS II and the plastoquinone pool and the upstream of PS I. The apocytochrome f (PetA) and the major Rieske protein PetC1 are two components of the Cyt b6/f complex (
      • Schneider D.
      • Berry S.
      • Volkmer T.
      • Seidler A.
      • Rogner M.
      PetC1 is the major Rieske iron-sulfur protein in the cytochrome b6f complex of Synechocystis sp. PCC 6803.
      ). PetC1 is predicted to contain a signal peptide and was experimentally determined as a twin-arginine translocation (Tat) substrate (
      • Barnett J.P.
      • Robinson C.
      • Scanlan D.J.
      • Blindauer C.A.
      The Tat protein export pathway and its role in cyanobacterial metalloprotein biosynthesis.
      ,
      • Aldridge C.
      • Spence E.
      • Kirkilionis M.A.
      • Frigerio L.
      • Robinson C.
      Tat-dependent targeting of Rieske iron-sulphur proteins to both the plasma and thylakoid membranes in the cyanobacterium Synechocystis PCC6803.
      ). In higher plants the signal peptide protein is not cleaved off after translocation and remained as a TM (
      • Karnauchov I.
      • Herrmann R.G.
      • Klösgen R.B.
      Transmembrane topology of the Rieske Fe/S protein of the cytochrome b6/f complex from spinach chloroplasts.
      ). The proteins CcsA and CcsB are important for cytochrome f biogenesis (
      • Hubschmann T.
      • Wilde A.
      • Elanskaya I.
      • Shestakov S.V.
      • Borner T.
      A putative cytochrome c biogenesis gene in Synechocystis sp. PCC 6803.
      ,
      • Tichy M.
      • Vermaas W.
      Accumulation of pre-apocytochrome f in a Synechocystis sp. PCC 6803 mutant impaired in cytochrome c maturation.
      ). This data strongly suggests that the Cyt b6/f complex is closely connected with PS II not only in function, but also in space. The hypothetical protein Slr0404 is the Synechocystis homolog of Arabidopsis Fluctuating-light acclimation protein1 (FLAP1), and is involved in regulation of pH homeostasis and proper induction of NPQ in chloroplasts. Inactivation of slr0404 impaired the growth of Synechocystis in light-dark cycles (
      • Inago H.
      • Sato R.
      • Masuda S.
      Regulation of light-induced H(+) extrusion and uptake by cyanobacterial homologs of the plastidial FLAP1, DLDG1, and Ycf10 in Synechocystis sp. PCC6803.
      ). Sll0814 is also a hypothetical protein, its homolog in Synechococcus sp. PCC 7002 is involved in synthesis of Myxoxanthophyll, a carotenoid glycosides acting as natural surfactants, which can stabilize membranes and regulate the permeability of membranes to oxygen (
      • Graham J.E.
      • Bryant D.A.
      The biosynthetic pathway for myxol-2' fucoside (myxoxanthophyll) in the cyanobacterium Synechococcus sp. strain PCC 7002.
      ), and presumably facilitates localized release of photosynthetically produced O2 from the lumen. The two-component sensory histidine kinase Hik33 is a major sensor for various abiotic stresses (
      • Ge H.
      • Fang L.
      • Huang X.
      • Wang J.
      • Chen W.
      • Liu Y.
      • Zhang Y.
      • Wang X.
      • Xu W.
      • He Q.
      • Wang Y.
      Translating divergent environmental stresses into a common proteome response through the histidine kinase 33 (Hik33) in a model cyanobacterium.
      ). Hik33 locates mainly on the thylakoid membrane and we previously proposed that the TM-residing Hik33 is activated not directly by the extracellular signals, but indirectly by the intracellular signals such as the redox change of the PQ pool (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ). Residing in the neighborhood of PS II, and of the PQ as well, may allow Hik33 to keenly sense the redox change of the PQ pool, a consequence of the electron transport from PS II.
      Fourth, proteins without known functions related with PS II (supplemental Table S3). The majority of these proteins are hypothetical proteins and their membrane localization is supported by predicted TM-containing as well reported experimental evidence (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ). The functions of these hypothetical proteins are largely unknown, and it is reasonable to infer that they might be functionally related with PS II because of their short distance. Remarkably, PsaL is the only PS I subunit that was identified in the proximity to PS II. PsaL is not an integral component of PS I monomer but is essential for formation of PS I trimer (
      • Chitnis V.P.
      • Chitnis P.R.
      PsaL subunit is required for the formation of photosystem I trimers in the cyanobacterium Synechocystis sp. PCC 6803.
      ). Freely existing PsaL but not as a subunit of PS I trimer has the chance to locate proximal to PS II. The fact no other PS I subunit was identified as the PS II neighborhood protein strongly suggests that PS II and PS I are spatially separated, with a distance longer than the radius limited by APEX2. Indeed, in chloroplast of higher plants, PS I locates mainly on the stroma lamellae, whereas PS II locate mainly on the inner stacks of grana (
      • Lawlor D.W.
      Ke, B. Photosynthesis: photobiochemistry and photobiophysics.
      ). In addition, in the Synechococcus sp. PCC 7942, the heterogeneous distribution of the four photosynthetic complexes is determined by live-cell fluorescence imaging (
      • Casella S.
      • Huang F.
      • Mason D.
      • Zhao G.Y.
      • Johnson G.N.
      • Mullineaux C.W.
      • Liu L.N.
      Dissecting the native architecture and dynamics of cyanobacterial photosynthetic machinery.
      ). Moreover, in Synechocystis, using atomic force microscopy (AFM), large areas of featureless membranes were observed around the PS I that separate PS I from the PS II and the Cyt b6/f complexes (
      • MacGregor-Chatwin C.
      • Sener M.
      • Barnett S.F.H.
      • Hitchcock A.
      • Barnhart-Dailey M.C.
      • Maghlaoui K.
      • Barber J.
      • Timlin J.A.
      • Schulten K.
      • Hunter C.N.
      Lateral segregation of photosystem I in cyanobacterial thylakoids.
      ).

      Topology Prediction for IMPs Based on the Distribution of Potential APEX2 Targeting Sites

      The phenoxyl radicals can covalently react with electron-rich amino acids such as Tyr, Trp, His and Cys (
      • Minamihata K.
      • Goto M.
      • Kamiya N.
      Protein heteroconjugation by the peroxidase-catalyzed tyrosine coupling reaction.
      ,
      • Rogers M.S.
      • Hurtado-Guerrero R.
      • Firbank S.J.
      • Halcrow M.A.
      • Dooley D.M.
      • Phillips S.E.
      • Knowles P.F.
      • McPherson M.J.
      Cross-link formation of the cysteine 228-tyrosine 272 catalytic cofactor of galactose oxidase does not require dioxygen.
      ,
      • Bhaskar B.
      • Immoos C.E.
      • Shimizu H.
      • Sulc F.
      • Farmer P.J.
      • Poulos T.L.
      A novel heme and peroxide-dependent tryptophan–tyrosine cross-link in a mutant of cytochrome c peroxidase.
      ). It was suggested that tyrosine is the primary site of biotinylation labeled by APEX2, the other three residues account for less than 2% of the APEX2-labeled residues (
      • Udeshi N.D.
      • Pedram K.
      • Svinkina T.
      • Fereshetian S.
      • Myers S.A.
      • Aygun O.
      • Krug K.
      • Clauser K.
      • Ryan D.
      • Ast T.
      • Mootha V.K.
      • Ting A.Y.
      • Carr S.A.
      Antibodies to biotin enable large-scale detection of biotinylation sites on proteins.
      ). It is possible to predict transmembrane topology for certain IMPs based on the exclusive existence of such biotinylation-targeting residues in the regions exposed to the lumen. For instance, the mature cytochrome f has a TM locating between a long N-terminus and a short C-terminus (
      • Anderson C.M.
      • Gray J.
      Cleavage of the precursor of pea chloroplast cytochrome f by leader peptidase from Escherichia coli.
      ). There are multiple potential biotin-tagging sites in the N-terminal region but not in the C-terminal (Fig. 8A). Therefore, it is inferred that the N-terminal region of the mature cytochrome f locates in the thylakoid lumen while the C-terminal region locates in the cytoplasm (Fig. 8B). Such a topology ensures that the potential biotin-tagging sites in the N-terminal region are accessible to PsbO-APEX2 catalyzed labeling. The prediction is consistent with earlier experimental evidence (
      • Gray J.C.
      Cytochrome f: Structure, function and biosynthesis.
      ). Similarly, we successfully predicted the topology for five other IMPs identified in the present study and two of these contain 2 TMs (Fig. 8B).
      Figure thumbnail gr8
      Fig. 8Transmembrane topology prediction for IMPs. A, schematic representation of apocytochrome f structure. The N-terminal signal peptide will be cleaved when the protein become mature. The large region to the N-terminal of the TM contains a plethora of potential biotinylation sites, whereas the small region to the C-terminal of the TM does not contain such a site. B, predicted transmembrane topology for six indicated IMPs.

      Thylakoid Luminal Proteins in the Neighborhood of the PS II OEC Complex

      The 93 luminal proteins can also be categorized into three groups based on their functional relationship with PS II (supplemental Table S3). First, the peripheral subunits of PS II or its assembly factors. These include PsbO, PsbU, PsbV, Psb27, PsbP2, Ycf48. PsbO, PsbU, and PsbV are components of the OEC complex. Psb27 is a luminal extrinsic subunit of PS II that facilitates manganese cluster assembly (
      • Roose J.L.
      • Pakrasi H.B.
      The Psb27 protein facilitates manganese cluster assembly in photosystem II.
      ). Psb27 also binds to CP43 complex (
      • Komenda J.
      • Knoppová J.
      • Kopečná J.
      • Sobotka R.
      • Halada P.
      • Yu J.
      • Nickelsen J.
      • Boehm M.
      • Nixon P.J.
      The Psb27 assembly factor binds to the CP43 complex of photosystem II in the cyanobacterium Synechocystis sp. PCC 6803.
      ,
      • Liu H.
      • Huang R.Y.
      • Chen J.
      • Gross M.L.
      • Pakrasi H.B.
      Psb27, a transiently associated protein, binds to the chlorophyll binding protein CP43 in photosystem II assembly intermediates.
      ). PsbP2 is a peripheral component of PS II, but tightly associated with the thylakoid membrane (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Sveshnikov D.
      • Funk C.
      • Schroder W.P.
      The PsbP-like protein (sll1418) of Synechocystis sp. PCC 6803 stabilises the donor side of Photosystem II.
      ,
      • Ishikawa Y.
      • Schroder W.P.
      • Funk C.
      Functional analysis of the PsbP-like protein (sll1418) in Synechocystis sp. PCC 6803.
      ). Ycf48 is a thylakoid lumen residing PS II assembly factor that is essential for PS II assembly or stability in cyanobacteria (
      • Komenda J.
      • Nickelsen J.
      • Tichy M.
      • Prasil O.
      • Eichacker L.A.
      • Nixon P.J.
      The cyanobacterial homologue of HCF136/YCF48 is a component of an early photosystem II assembly complex and is important for both the efficient assembly and repair of photosystem II in Synechocystis sp. PCC 6803.
      ). The Arabidopsis homolog of Ycf48 is also located in the thylakoid lumen and essential for PS II assembly (
      • Meurer J.
      • Plucken H.
      • Kowallik K.V.
      • Westhoff P.
      A nuclear-encoded protein of prokaryotic origin is essential for the stability of photosystem II in Arabidopsis thaliana.
      ).
      Second, proteins functionally related with PS II (supplemental Table S3). These include PratA, CtpA, HhoA, PetE, MncA, and CucA. PratA is a well-documented protein important for PS II biogenesis and interacts with D1 protein (
      • Stengel A.
      • Gügel I.L.
      • Hilger D.
      • Rengstl B.
      • Jung H.
      • Nickelsen J.
      Initial steps of photosystem II de novo assembly and preloading with manganese take place in biogenesis centers in Synechocystis.
      ,
      • Schottkowski M.
      • Gkalympoudis S.
      • Tzekova N.
      • Stelljes C.
      • Schunemann D.
      • Ankele E.
      • Nickelsen J.
      Interaction of the periplasmic PratA factor and the PsbA (D1) protein during biogenesis of photosystem II in Synechocystis sp. PCC 6803.
      ,
      • Rengstl B.
      • Oster U.
      • Stengel A.
      • Nickelsen J.
      An intermediate membrane subfraction in cyanobacteria is involved in an assembly network for Photosystem II biogenesis.
      ,
      • Klinkert B.
      • Ossenbuhl F.
      • Sikorski M.
      • Berry S.
      • Eichacker L.
      • Nickelsen J.
      PratA, a periplasmic tetratricopeptide repeat protein involved in biogenesis of photosystem II in Synechocystis sp. PCC 6803.
      ). It is of note that PratA was also reported as a periplasmic protein (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ). CtpA is a C-terminal processing peptidases for the D1 protein and functions in the thylakoid lumen (
      • Anbudurai P.R.
      • Mor T.S.
      • Ohad I.
      • Shestakov S.V.
      • Pakrasi H.B.
      The ctpA gene encodes the C-terminal processing protease for the D1 protein of the photosystem II reaction center complex.
      ). In most oxygenic photosynthetic organisms, D1 is synthesized as a precursor protein (pD1) with a short C-terminal extension. After integration into the thylakoid membrane, the pD1 C-terminal extension is removed by CtpA to generate mature D1 (mD1), which provides a docking site for the manganese cluster of the OEC (
      • Inagaki N.
      • Yamamoto Y.
      • Satoh K.
      A sequential two-step proteolytic process in the carboxyl-terminal truncation of precursor D1 protein inSynechocystissp. PCC68031.
      ). The other two carboxyl-terminal peptidases (CtpB and CtpC) were also identified in the lumen. The physiological function of the two peptidase is yet unknown, and inactivation of ctpC gene seems to be lethal for the cyanobacterium (
      • Ivleva N.B.
      • Sidoruk K.V.
      • Pakrasi H.B.
      • Shestakov S.V.
      Investigation of the functional role of Ctp proteins in the cyanobacterium Synechocystis sp PCC 6803.
      ). HhoA (Sll1679) is a Deg family peptidase and is responsible for PsbO degradation in a thioredoxin dependent manner (
      • Roberts I.N.
      • Lam X.T.
      • Miranda H.
      • Kieselbach T.
      • Funk C.
      Degradation of PsbO by the Deg protease HhoA is thioredoxin dependent.
      ). HhoA is co-transcribed with a downstream open reading frame (sll1680) that was also identified as a luminal protein (
      • Dong W.
      • Wang J.
      • Niu G.
      • Zhao S.
      • Liu L.
      Crystal structure of the zinc-bound HhoA protease from Synechocystis sp. PCC 6803.
      ). PetE (plastocyanin) is a luminal copper protein and one of the soluble electron carriers for photosynthetic and respiratory electron transport (
      • Bertini I.
      • Ciurli S.
      • Dikiy A.
      • Fernandez C.O.
      • Luchinat C.
      • Safarov N.
      • Shumilin S.
      • Vila A.J.
      The first solution structure of a paramagnetic copper(II) protein: the case of oxidized plastocyanin from the cyanobacterium Synechocystis PCC6803.
      ,
      • Duran R.V.
      • Hervas M.
      • De La Rosa M.A.
      • Navarro J.A.
      The efficient functioning of photosynthesis and respiration in Synechocystis sp. PCC 6803 strictly requires the presence of either cytochrome c6 or plastocyanin.
      ). The heme protein PetJ (cytochrome c553) is the other soluble electron carriers in the lumen. PetJ was not identified possibly due to its low expression in the growth condition used for the present study. MncA (Sll1358) is the only known Mn2+ binding protein in cytoplasm and presumably involved in Mn2+ storage and homeostasis (
      • Stengel A.
      • Gügel I.L.
      • Hilger D.
      • Rengstl B.
      • Jung H.
      • Nickelsen J.
      Initial steps of photosystem II de novo assembly and preloading with manganese take place in biogenesis centers in Synechocystis.
      ,
      • Tottey S.
      • Waldron K.J.
      • Firbank S.J.
      • Reale B.
      • Bessant C.
      • Sato K.
      • Cheek T.R.
      • Gray J.
      • Banfield M.J.
      • Dennison C.
      • Robinson N.J.
      Protein-folding location can regulate manganese-binding versus copper- or zinc-binding.
      ). MncA is predicted to contain a TM by TMHMM. Existence of this protein in lumen indicates its role in luminal Mn2+ storage and homeostasis that is critical for optimal functioning of the OEC complex. Similarly, Sll1785 (CucA) is a major copper binding protein that might be important for luminal copper homeostasis, and supplies copper to the other copper proteins such as PetE (
      • Tottey S.
      • Waldron K.J.
      • Firbank S.J.
      • Reale B.
      • Bessant C.
      • Sato K.
      • Cheek T.R.
      • Gray J.
      • Banfield M.J.
      • Dennison C.
      • Robinson N.J.
      Protein-folding location can regulate manganese-binding versus copper- or zinc-binding.
      ). Notably, both MncA and CucA were previously identified as periplasmic proteins (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ), indicating that the two proteins locate in multiple subcellular compartments.
      Third, proteins without determined functional relationship with PS II (supplemental Table S3). All proteins except those in the above-described two groups are assigned to this group. Notwithstanding their functional relationship with PS II remains to be determined, a few proteins have the potential to functionally interplay with PS II. One example is the existence of multiple redox proteins in the proximity to the OEC including the thioredoxin-like proteins Slr1796, Slr1980, and a methionine sulfoxide reductase Sll1680 (MsrB). Existence of these redox proteins might be required for keen redox regulation of PS II subunits, which are highly susceptible to oxidative stresses (
      • Tarrago L.
      • Laugier E.
      • Rey P.
      Protein-repairing methionine sulfoxide reductases in photosynthetic organisms: gene organization, reduction mechanisms, and physiological roles.
      ,
      • Holmgren A.
      Antioxidant function of thioredoxin and glutaredoxin systems.
      ). Notably, Sll1980 is closely related to a thioredoxin-like protein HCF164 of Arabidopsis that is located in the thylakoid lumen and has disulfide reductase activity (
      • Lennartz K.
      • Plucken H.
      • Seidler A.
      • Westhoff P.
      • Bechtold N.
      • Meierhoff K.
      HCF164 encodes a thioredoxin-like protein involved in the biogenesis of the cytochrome b(6)f complex in Arabidopsis.
      ). Sll1680 is possibly co-expressed with HhoA under the control of the same operon (
      • Dong W.
      • Wang J.
      • Niu G.
      • Zhao S.
      • Liu L.
      Crystal structure of the zinc-bound HhoA protease from Synechocystis sp. PCC 6803.
      ), and co-existence of the two protein in the proximity to OEC is a strong indicator for their functional connection, and is probably involved in protection against an oxidative stress and protein quality control. The other example is the existence of multiple peptidyl-prolyl cis-trans isomerases (PPI) including Sll0227, Sll0408, and Slr1761. The PPI proteins accelerate protein folding by catalyzing the cis-trans isomerization of proline (
      • Schiene-Fischer C.
      Multidomain Peptidyl Prolyl cis/trans Isomerases.
      ). Synechocystis genome encodes only four PPI proteins (
      • Kaneko T.
      • Sato S.
      • Kotani H.
      • Tanaka A.
      • Asamizu E.
      • Nakamura Y.
      • Miyajima N.
      • Hirosawa M.
      • Sugiura M.
      • Sasamoto S.
      • Kimura T.
      • Hosouchi T.
      • Matsuno A.
      • Muraki A.
      • Nakazaki N.
      • Naruo K.
      • Okumura S.
      • Shimpo S.
      • Takeuchi C.
      • Wada T.
      • Watanabe A.
      • Yamada M.
      • Yasuda M.
      • Tabata S.
      Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions.
      ), and three of them were identified here. Consistently, the majority of the signal peptide-containing PPI proteins were also identified in the chloroplast lumen of Arabidopsis thaliana (
      • Kieselbach T.
      • Schroder W.P.
      The proteome of the chloroplast lumen of higher plants.
      ). Existence of the PPI proteins in the thylakoid lumen is an expected observation because they are needed to accelerate protein folding, considering that many luminal proteins including PsbO are transported across the thylakoid membrane via the secretory (Sec) pathway in an unfolded state (
      • Frain K.M.
      • Gangl D.
      • Jones A.
      • Zedler J.A.
      • Robinson C.
      Protein translocation and thylakoid biogenesis in cyanobacteria.
      ). Probably for the same reason, Sll0227 and Slr1761 also locate in the periplasmic space (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ).

      DISCUSSION

      In the present study we targeted APEX2 to the Synechocystis thylakoid lumen through fusion with PsbO, and labeled proteins in the neighborhood of the fusion protein with biotin for subsequent affinity purification and proteomic identification. In total, we identified 131 proteins proximal to PsbO-APEX2, which includes 38 integral thylakoid membrane proteins and 93 luminal proteins. It is worth noting that the identified proteins do not necessarily include either the entire neighborhood integral thylakoid membrane proteome or the entire luminal proteome. For IMPs proximal to PsbO-APEX2, only those with lumen-facing domains that contain potential sites for biotinylation were identified. For luminal proteins, labeling by PsbO-APEX2 catalyzed biotinylation could be affected by free-diffusing ability in the lumen except for those tightly associated with the OEC. Nevertheless, our results unveiled the luminal side neighborhood proteome, for the first time, for PS II in Synechocystis. Besides, our results showcase the feasibility to define the proteome in a subdomain of interest in this model organism. In addition to PsbO-APEX2, we also included Zwf-APEX2 in this study, though as a negative control to remove cytoplasmic background. In our practice, many APEX2-fused Synechocystis proteins can be expressed and efficiently catalyze proximity labeling, while some others are difficult to express. Therefore, expressing APEX2-fused proteins in Synechocystis is still a technical challenge. Removing the background is another technical challenge. In the present study we included three different negative controls to remove background from different origins, and successfully removed backgrounds resulting from endogenous peroxidase activity, endogenous substrates, cytoplasmic background, and even non-specific binding to streptavidin beads. Among these, the most challenging is to remove the cytoplasmic background labelled by the residual cytoplasm-locating PsbO-APEX2, because the fusion protein is translated in the cytoplasm and might stay in the cytoplasm for a short time before translocated across the thylakoid membrane into the luminal destination. Because of this, it is conceivable that it is much more difficult to identify proteins proximal to a freely-diffusing soluble protein than an integral or peripheral membrane protein that has a relatively stable cellular location.
      During the progress of the current study, Dahlgren et al. published their work on the proximity labeling-based proteomic identification of luminal proteins in Synechococcus sp. PCC 7002 (
      • Dahlgren K.K.
      • Gates C.
      • Lee T.
      • Cameron J.C.
      Proximity-based proteomics reveals the thylakoid lumen proteome in the cyanobacterium Synechococcus sp. PCC 7002.
      ). Both studies chose an extrinsic subunit of PS II as the bait in coincidence. Among the 131 proteins identified by us, 95 have homologs in Synechococcus and 51 of these were also identified by the other study. These include but not limited to PS II subunits and assembly factors, multiple proteases, and a number of proteins existing also in periplasmic space (supplemental Table S4). The overlapping proteins in the two studies presumably represent the consensus neighborhood proteins of the PS II OEC, and corroborate the feasibility of defining the proteome within a particular domain of interest in cyanobacteria using APEX-based proximity labeling. Nevertheless, significant difference exists between the two studies in both experiment design and the results. First, we used three negative controls and the other study used only one to exclude background. The two extra controls used in the present study were attempted to exclude the background resulting from endogenous peroxidase activity (the WT strain treated with H2O2 and BP) and the background from endogenous biotinylated proteins (PsbO-APEX strain treated with H2O2 only), the exclusion of cytoplasmic background was both attempted by using different fusion proteins, i.e., Zwf-APEX2 in the present study and GFP-APEX2 in the other study. Although the background signals from the control of WT strain are hardly detectable by anti-biotin blotting (Fig. 1C), the signals are obviously detectable when the biotinylated proteins were enriched by streptavidin affinity purification (Fig. 4B). Hence it is important to include these negative controls. Second, because of the difference in experiment design, it is expectable that the result is different. The present study identified 44 proteins with homologs in Synechococcus that were not identified by Dahlgren et al.(
      • Dahlgren K.K.
      • Gates C.
      • Lee T.
      • Cameron J.C.
      Proximity-based proteomics reveals the thylakoid lumen proteome in the cyanobacterium Synechococcus sp. PCC 7002.
      ), the majority of these are functionally uncharacterized proteins (supplemental Table S4). Similarly, Dahlgren et al. identified 55 Synechococcus proteins with homologs in Synechocystis that were excluded in our identification list (supplemental Table S4). These include five PS I subunits (PsaA, PsaB, PsaC, PsaD, and PsaF) and two ATP synthase subunits (AtpF and AtpG). PsaC and PsaD are the two extrinsic PS I subunits at the cytoplasmic side, they are not expected to be labeled by the luminal side APEX2 activity. This further underlines the importance to use more stringent controls to exclude background. Third, both studies uniquely identified a set of organism-specific proteins (supplemental Table S4). For example, we identified plastocyanin (PetE) but not cytochrome c553 (PetJ) whereas the other study identified PetJ but not PetE. Both PetE and PetJ are known soluble electron carriers in thylakoid lumen in Synechocystis, but in Synechococcus only PetJ is expressed (
      • Dahlgren K.K.
      • Gates C.
      • Lee T.
      • Cameron J.C.
      Proximity-based proteomics reveals the thylakoid lumen proteome in the cyanobacterium Synechococcus sp. PCC 7002.
      ). Fourth, experimental procedures. We found that it is impossible to effectively enrich biotinylated proteins without removing the excess biotin-phenol from the lysate of the cells underwent APEX2-catalyzed proximity labeling through filtering or precipitation of the samples (Fig. 2), and this is consistent with a previous report (
      • Mair A.
      • Xu S.L.
      • Branon T.C.
      • Ting A.Y.
      • Bergmann D.C.
      Proximity labeling of protein complexes and cell-type-specific organellar proteomes in Arabidopsis enabled by TurboID.
      ). In the other study removal of excess biotin-phenol was omitted (
      • Dahlgren K.K.
      • Gates C.
      • Lee T.
      • Cameron J.C.
      Proximity-based proteomics reveals the thylakoid lumen proteome in the cyanobacterium Synechococcus sp. PCC 7002.
      ). This is the only experimental procedure with major difference and the differences of the other procedures are minor and compared side by side (supplemental Table S5). Last, the difference in the bait proteins. In the present study PsbO is used as the bait and in the other study PsbU was used as the bait. PsbO and PsbU are both the subunits of the OEC and the difference might be minor. Notwithstanding these differences, both studies generated a protein catalog that is important to define the protein neighborhood of PS II OEC in the respective model organism.
      The multiple negative controls and the stringent filtering criteria allow us to resolve the PS II OEC neighborhood proteome with a resolution that can exclude the other major macromolecular protein complexes on the thylakoid membrane such as PS I and ATP synthase. This result suggests that PS I, and ATP synthase as well, are spatially separated from PS II with a distance beyond the labeling radius of APEX2 (∼ 20 nm) (
      • Rhee H.W.
      • Zou P.
      • Udeshi N.D.
      • Martell J.D.
      • Mootha V.K.
      • Carr S.A.
      • Ting A.Y.
      Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging.
      ), which supports the notion that thylakoid membrane is not homogenous (
      • Casella S.
      • Huang F.
      • Mason D.
      • Zhao G.Y.
      • Johnson G.N.
      • Mullineaux C.W.
      • Liu L.N.
      Dissecting the native architecture and dynamics of cyanobacterial photosynthetic machinery.
      ). Using the same approach, the neighborhood proteome of the other membrane or membrane-associated proteins can also be spatiotemporally resolved with a similar resolution. This information will be important for structural, functional, and physiological studies involved in both basic and applied research, particularly in discovering novel proteins that interact with and play a significant role in regulating the structure and function of a protein complex of interest.
      A remarkable finding is that 57 luminal proteins were also previously identified from isolated periplasm by us and the others (supplemental Table S3) (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ), and similar finding was also reported by Dahlgren et al. (
      • Dahlgren K.K.
      • Gates C.
      • Lee T.
      • Cameron J.C.
      Proximity-based proteomics reveals the thylakoid lumen proteome in the cyanobacterium Synechococcus sp. PCC 7002.
      ). Such a large portion cannot be simply explained by contamination. Some of these proteins are associated with the periplasmic side of the plasma membrane, and it is not surprising that they are also associated with the luminal side of the thylakoid membrane because it was proposed that thylakoid membrane might be originated from the plasma membrane (
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ). In fact, a recent study also showed that the biogenic regions of the thylakoid membrane forms contacts with the plasma membrane (
      • Rast A.
      • Schaffer M.
      • Albert S.
      • Wan W.
      • Pfeffer S.
      • Beck F.
      • Plitzko J.M.
      • Nickelsen J.
      • Engel B.D.
      Biogenic regions of cyanobacterial thylakoids form contact sites with the plasma membrane.
      ). PratA is one of the proteins locates at this region (
      • Schottkowski M.
      • Gkalympoudis S.
      • Tzekova N.
      • Stelljes C.
      • Schunemann D.
      • Ankele E.
      • Nickelsen J.
      Interaction of the periplasmic PratA factor and the PsbA (D1) protein during biogenesis of photosystem II in Synechocystis sp. PCC 6803.
      ,
      • Klinkert B.
      • Ossenbuhl F.
      • Sikorski M.
      • Berry S.
      • Eichacker L.
      • Nickelsen J.
      PratA, a periplasmic tetratricopeptide repeat protein involved in biogenesis of photosystem II in Synechocystis sp. PCC 6803.
      ). The other proteins may be transported to the two different compartments by the endogenous protein translocation systems. In cyanobacteria, the Sec and the Tat are two major pathways transporting proteins to both periplasmic space and thylakoid lumen (
      • Frain K.M.
      • Gangl D.
      • Jones A.
      • Zedler J.A.
      • Robinson C.
      Protein translocation and thylakoid biogenesis in cyanobacteria.
      ). Synechocystis has only one set of Sec and Tat transport systems that locate on both the plasma membrane and the thylakoid lumen (
      • Wang J.
      • Huang X.
      • Ge H.
      • Wang Y.
      • Chen W.
      • Zheng L.
      • Huang C.
      • Yang H.
      • Li L.
      • Sui N.
      • Wang Y.
      • Zhang Y.
      • Lu D.
      • Fang L.
      • Xu W.
      • Jiang Y.
      • Huang F.
      • Wang Y.
      The quantitative proteome atlas of a model cyanobacterium.
      ,
      • Fulda S.
      • Huang F.
      • Nilsson F.
      • Hagemann M.
      • Norling B.
      Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
      ), therefore it is possible a particular protein can be transported across the plasma membrane to the periplasmic space or across the thylakoid membrane to the lumen. If this is the case, an interesting question would be raised regarding the selectivity of the translocation because some proteins are translocated to a single destination while the others are translocated to the both destinations. The other intriguing question is what the functional difference is when a particular protein locates at a different compartment.
      In summary, we developed an effective workflow for APEX2-based proximity labeling in Synechocystis and successfully resolved neighborhood proteome of the PS II OEC. The information is particularly useful for the mechanistic study regarding the PS II structure and function.

      Data availability

      The raw mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (
      • Perez-Riverol Y.
      • Csordas A.
      • Bai J.
      • Bernal-Llinares M.
      • Hewapathirana S.
      • Kundu D.J.
      • Inuganti A.
      • Griss J.
      • Mayer G.
      • Eisenacher M.
      • Perez E.
      • Uszkoreit J.
      • Pfeuffer J.
      • Sachsenberg T.
      • Yilmaz S.
      • Tiwary S.
      • Cox J.
      • Audain E.
      • Walzer M.
      • Jarnuczak A.F.
      • Ternent T.
      • Brazma A.
      • Vizcaino J.A.
      The PRIDE database and related tools and resources in 2019: improving support for quantification data.
      ) partner repository with the dataset identifier PXD036113.

      Supplemental data

      This article contains supplemental data.

      Acknowledgments

      We thank Dr. Jinlong Wang and Dr. Weiyang Chen for the constructive suggestions in data analysis.

      REFERENCES

        • Kasting J.F.
        • Siefert J.L.
        Life and the evolution of Earth's atmosphere.
        Science. 2002; 296: 1066-1068
        • Goksoyr J.
        Evolution of eucaryotic cells.
        Nature. 1967; 214: 1161
        • Lubner C.E.
        • Applegate A.M.
        • Knörzer P.
        • Ganago A.
        • Bryant D.A.
        • Happe T.
        • Golbeck J.H.
        Solar hydrogen-producing bionanodevice outperforms natural photosynthesis.
        Proc. Natl. Acad. Sci. U. S. A. 2011; 108: 20988-20991
        • Varman A.M.
        • Xiao Y.
        • Pakrasi H.B.
        • Tang Y.J.
        Metabolic engineering of Synechocystis sp. strain PCC 6803 for isobutanol production.
        Appl. Environ. Microbiol. 2013; 79: 908-914
        • Xue Y.
        • Zhang Y.
        • Cheng D.
        • Daddy S.
        • He Q.
        Genetically engineering Synechocystis sp. Pasteur Culture Collection 6803 for the sustainable production of the plant secondary metabolite p-coumaric acid.
        Proc. Natl. Acad. Sci. U. S. A. 2014; 111: 9449-9454
        • Gao L.
        • Ge H.
        • Huang X.
        • Liu K.
        • Zhang Y.
        • Xu W.
        • Wang Y.
        Systematically ranking the tightness of membrane association for peripheral membrane proteins (PMPs).
        Mol. Cell. Proteomics. 2015; 14: 340-353
        • Huang F.
        • Parmryd I.
        • Nilsson F.
        • Persson A.L.
        • Pakrasi H.B.
        • Andersson B.
        • Norling B.
        Proteomics of Synechocystis sp. strain PCC 6803: identification of plasma membrane proteins.
        Mol. Cell. Proteomics. 2002; 1: 956-966
        • Huang F.
        • Hedman E.
        • Funk C.
        • Kieselbach T.
        • Schroder W.P.
        • Norling B.
        Isolation of outer membrane of Synechocystis sp. PCC 6803 and its proteomic characterization.
        Mol. Cell. Proteomics. 2004; 3: 586-595
        • Baers L.L.
        • Breckels L.M.
        • Mills L.A.
        • Gatto L.
        • Deery M.J.
        • Stevens T.J.
        • Howe C.J.
        • Lilley K.S.
        • Lea-Smith D.J.
        Proteome mapping of a cyanobacterium reveals distinct compartment organization and cell-dispersed metabolism.
        Plant Physiol. 2019; 181: 1721-1738
        • Wang Y.
        • Xu W.
        • Chitnis P.R.
        Identification and bioinformatic analysis of the membrane proteins of synechocystis sp. PCC 6803.
        Proteome Sci. 2009; 7: 11
        • Wang J.
        • Huang X.
        • Ge H.
        • Wang Y.
        • Chen W.
        • Zheng L.
        • Huang C.
        • Yang H.
        • Li L.
        • Sui N.
        • Wang Y.
        • Zhang Y.
        • Lu D.
        • Fang L.
        • Xu W.
        • Jiang Y.
        • Huang F.
        • Wang Y.
        The quantitative proteome atlas of a model cyanobacterium.
        J Genet Genomics. 2022; 49: 96-108
        • Fulda S.
        • Huang F.
        • Nilsson F.
        • Hagemann M.
        • Norling B.
        Proteomics of Synechocystis sp strain PCC 6803 - Identification of periplasmic proteins in cells grown at low and high salt concentrations.
        Eur. J. Biochem. 2000; 267: 5900-5907
        • Zhang L.
        • Selao T.T.
        • Selstam E.
        • Norling B.
        Subcellular Localization of Carotenoid Biosynthesis in Synechocystis sp. PCC 6803.
        PloS one. 2015; 10e0130904
        • Rast A.
        • Schaffer M.
        • Albert S.
        • Wan W.
        • Pfeffer S.
        • Beck F.
        • Plitzko J.M.
        • Nickelsen J.
        • Engel B.D.
        Biogenic regions of cyanobacterial thylakoids form contact sites with the plasma membrane.
        Nat Plants. 2019; 5: 436-446
        • Agarwal R.
        • Matros A.
        • Melzer M.
        • Mock H.P.
        • Sainis J.K.
        Heterogeneity in thylakoid membrane proteome of Synechocystis 6803.
        J Proteomics. 2010; 73: 976-991
        • Rhee H.W.
        • Zou P.
        • Udeshi N.D.
        • Martell J.D.
        • Mootha V.K.
        • Carr S.A.
        • Ting A.Y.
        Proteomic mapping of mitochondria in living cells via spatially restricted enzymatic tagging.
        Science. 2013; 339: 1328-1331
        • Lam S.S.
        • Martell J.D.
        • Kamer K.J.
        • Deerinck T.J.
        • Ellisman M.H.
        • Mootha V.K.
        • Ting A.Y.
        Directed evolution of APEX2 for electron microscopy and proximity labeling.
        Nat. Methods. 2015; 12: 51-54
        • Hung V.
        • Lam S.S.
        • Udeshi N.D.
        • Svinkina T.
        • Guzman G.
        • Mootha V.K.
        • Carr S.A.
        • Ting A.Y.
        Proteomic mapping of cytosol-facing outer mitochondrial and ER membranes in living human cells by proximity biotinylation.
        Elife. 2017; 6https://doi.org/10.7554/eLife.24463
        • Singer-Kruger B.
        • Frohlich T.
        • Franz-Wachtel M.
        • Nalpas N.
        • Macek B.
        • Jansen R.P.
        APEX2-mediated proximity labeling resolves protein networks in Saccharomyces cerevisiae cells.
        FEBS J. 2020; 287: 325-344
        • Chen C.L.
        • Hu Y.
        • Udeshi N.D.
        • Lau T.Y.
        • Wirtz-Peitz F.
        • He L.
        • Ting A.Y.
        • Carr S.A.
        • Perrimon N.
        Proteomic mapping in live Drosophila tissues using an engineered ascorbate peroxidase.
        Proc. Natl. Acad. Sci. U. S. A. 2015; 112: 12093-12098
        • Dahlgren K.K.
        • Gates C.
        • Lee T.
        • Cameron J.C.
        Proximity-based proteomics reveals the thylakoid lumen proteome in the cyanobacterium Synechococcus sp. PCC 7002.
        Photosynth. Res. 2021; 147: 177-195
        • Ikeuchi M.
        • Tabata S.
        Synechocystis sp. PCC 6803 - a useful tool in the study of the genetics of cyanobacteria.
        Photosynth. Res. 2001; 70: 73-83
        • Kaneko T.
        • Sato S.
        • Kotani H.
        • Tanaka A.
        • Asamizu E.
        • Nakamura Y.
        • Miyajima N.
        • Hirosawa M.
        • Sugiura M.
        • Sasamoto S.
        • Kimura T.
        • Hosouchi T.
        • Matsuno A.
        • Muraki A.
        • Nakazaki N.
        • Naruo K.
        • Okumura S.
        • Shimpo S.
        • Takeuchi C.
        • Wada T.
        • Watanabe A.
        • Yamada M.
        • Yasuda M.
        • Tabata S.
        Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions.
        DNA research. 1996; 3: 109-136
        • Gao L.
        • Shen C.
        • Liao L.
        • Huang X.
        • Liu K.
        • Wang W.
        • Guo L.
        • Jin W.
        • Huang F.
        • Xu W.
        • Wang Y.
        Functional proteomic discovery of Slr0110 as a central regulator of carbohydrate metabolism in Synechocystis species PCC6803.
        Mol. Cell. Proteomics. 2014; 13: 204-219
        • Golden S.S.
        • Brusslan J.
        • Haselkorn R.
        Genetic engineering of the cyanobacterial chromosome.
        Methods Enzymol. 1987; 153: 215-231
        • Wessel D.
        • Flugge U.I.
        A method for the quantitative recovery of protein in dilute-solution in the presence of detergents and lipids.
        Anal. Biochem. 1984; 138: 141-143
        • Fukuyama H.
        • Ndiaye S.
        • Hoffmann J.
        • Rossier J.
        • Liuu S.
        • Vinh J.
        • Verdier Y.
        On-bead tryptic proteolysis: an attractive procedure for LC-MS/MS analysis of the Drosophila caspase 8 protein complex during immune response against bacteria.
        J Proteomics. 2012; 75: 4610-4619
        • Rappsilber J.
        • Ishihama Y.
        • Mann M.
        Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics.
        Anal. Chem. 2003; 75: 663-670
        • Cox J.
        • Mann M.
        MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification.
        Nat. Biotechnol. 2008; 26: 1367-1372
        • Cox J.
        • Mann M.
        1D and 2D annotation enrichment: a statistical method integrating quantitative proteomics with complementary high-throughput data.
        BMC Bioinformatics. 2012; 13: S12
        • Mohamed A.
        • Eriksson J.
        • Osiewacz H.D.
        • Jansson C.
        Differential expression of the psbA genes in the cyanobacterium Synechocystis 6803.
        Mol. Gen. Genet. 1993; 238: 161-168
        • Mandelman D.
        • Schwarz F.P.
        • Li H.
        • Poulos T.L.
        The role of quaternary interactions on the stability and activity of ascorbate peroxidase.
        Protein Sci. 1998; 7: 2089-2098
        • Martell J.D.
        • Deerinck T.J.
        • Sancak Y.
        • Poulos T.L.
        • Mootha V.K.
        • Sosinsky G.E.
        • Ellisman M.H.
        • Ting A.Y.
        Engineered ascorbate peroxidase as a genetically encoded reporter for electron microscopy.
        Nat. Biotechnol. 2012; 30: 1143-1148
        • Lobingier B.T.
        • Huttenhain R.
        • Eichel K.
        • Miller K.B.
        • Ting A.Y.
        • von Zastrow M.
        • Krogan N.J.
        An approach to spatiotemporally resolve protein interaction networks in living cells.
        Cell. 2017; 169 (e312): 350-360
        • Mair A.
        • Xu S.L.
        • Branon T.C.
        • Ting A.Y.
        • Bergmann D.C.
        Proximity labeling of protein complexes and cell-type-specific organellar proteomes in Arabidopsis enabled by TurboID.
        Elife. 2019; 8https://doi.org/10.7554/eLife.47864
        • Mironov K.S.
        • Sinetova M.A.
        • Shumskaya M.
        • Los D.A.
        Universal molecular triggers of stress responses in cyanobacterium Synechocystis.
        Life (Basel). 2019; 9https://doi.org/10.3390/life9030067
        • Chang W.
        • Li C.
        • Cui Z.
        • Li W.
        • Song H.
        • Chang H.
        • Fu W.
        • Wang C.
        • Huang T.
        • Luo Y.
        • Shan Y.
        • Wang Y.
        • Wang F.
        • Xu M.
        • Fu A.
        Diverged early from CtpB and CtpC, CtpA has evolved to process D1 precursor in oxygenic photosynthetic organisms.
        Front Plant Sci. 2021; 12676036
        • Frain K.M.
        • Gangl D.
        • Jones A.
        • Zedler J.A.
        • Robinson C.
        Protein translocation and thylakoid biogenesis in cyanobacteria.
        Biochim. Biophys. Acta. 2016; 1857: 266-273
        • Cox J.
        • Hein M.Y.
        • Luber C.A.
        • Paron I.
        • Nagaraj N.
        • Mann M.
        Accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction, termed MaxLFQ.
        Mol. Cell. Proteomics. 2014; 13: 2513-2526
        • Nakao M.
        • Okamoto S.
        • Kohara M.
        • Fujishiro T.
        • Fujisawa T.
        • Sato S.
        • Tabata S.
        • Kaneko T.
        • Nakamura Y.
        CyanoBase: the cyanobacteria genome database update 2010.
        Nucleic Acids Res. 2010; 38: D379-D381
        • Anbudurai P.R.
        • Mor T.S.
        • Ohad I.
        • Shestakov S.V.
        • Pakrasi H.B.
        The ctpA gene encodes the C-terminal processing protease for the D1 protein of the photosystem II reaction center complex.
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 8082-8086
        • Boehm M.
        • Yu J.
        • Krynicka V.
        • Barker M.
        • Tichy M.
        • Komenda J.
        • Nixon P.J.
        • Nield J.
        Subunit organization of a synechocystis hetero-oligomeric thylakoid FtsH complex involved in photosystem II repair.
        Plant Cell. 2012; 24: 3669-3683
        • Krogh A.
        • Larsson B.
        • von Heijne G.
        • Sonnhammer E.L.
        Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes.
        J. Mol. Biol. 2001; 305: 567-580
        • Knoppová J.
        • Sobotka R.
        • Tichy M.
        • Yu J.
        • Konik P.
        • Halada P.
        • Nixon P.J.
        • Komenda J.
        Discovery of a chlorophyll binding protein complex involved in the early steps of photosystem II assembly in Synechocystis.
        Plant Cell. 2014; 26: 1200-1212
        • Staleva H.
        • Komenda J.
        • Shukla M.K.
        • Slouf V.
        • Kana R.
        • Polivka T.
        • Sobotka R.
        Mechanism of photoprotection in the cyanobacterial ancestor of plant antenna proteins.
        Nat. Chem. Biol. 2015; 11: 287-291
        • Chidgey J.W.
        • Linhartova M.
        • Komenda J.
        • Jackson P.J.
        • Dickman M.J.
        • Canniffe D.P.
        • Konik P.
        • Pilny J.
        • Hunter C.N.
        • Sobotka R.
        A cyanobacterial chlorophyll synthase-HliD complex associates with the Ycf39 protein and the YidC/Alb3 insertase.
        Plant Cell. 2014; 26: 1267-1279
        • Ossenbuhl F.
        • Inaba-Sulpice M.
        • Meurer J.
        • Soll J.
        • Eichacker L.A.
        The synechocystis sp PCC 6803 oxa1 homolog is essential for membrane integration of reaction center precursor protein pD1.
        Plant Cell. 2006; 18: 2236-2246
        • Spence E.
        • Bailey S.
        • Nenninger A.
        • Moller S.G.
        • Robinson C.
        A homolog of Albino3/OxaI is essential for thylakoid biogenesis in the cyanobacterium Synechocystis sp. PCC6803.
        J. Biol. Chem. 2004; 279: 55792-55800
        • Zhou L.
        • Ye Y.
        Assembly of the photosystem II reaction center, the heart of photosynthesis.
        Plant Physiol. 2022; 189: 438-440
        • Silva P.
        • Thompson E.
        • Bailey S.
        • Kruse O.
        • Mullineaux C.W.
        • Robinson C.
        • Mann N.H.
        • Nixon P.J.
        FtsH is involved in the early stages of repair of photosystem II in Synechocystis sp PCC 6803.
        Plant Cell. 2003; 15: 2152-2164
        • Yao D.
        • Kieselbach T.
        • Komenda J.
        • Promnares K.
        • Prieto M.A.
        • Tichy M.
        • Vermaas W.
        • Funk C.
        Localization of the small CAB-like proteins in photosystem II.
        J. Biol. Chem. 2007; 282: 267-276
        • Wegener K.M.
        • Bennewitz S.
        • Oelmüller R.
        • Pakrasi H.B.
        The Psb32 protein aids in repairing photodamaged photosystem II in the cyanobacterium Synechocystis 6803.
        Mol. Plant. 2011; 4: 1052-1061
        • Schneider D.
        • Berry S.
        • Volkmer T.
        • Seidler A.
        • Rogner M.
        PetC1 is the major Rieske iron-sulfur protein in the cytochrome b6f complex of Synechocystis sp. PCC 6803.
        J. Biol. Chem. 2004; 279: 39383-39388
        • Barnett J.P.
        • Robinson C.
        • Scanlan D.J.
        • Blindauer C.A.
        The Tat protein export pathway and its role in cyanobacterial metalloprotein biosynthesis.
        FEMS Microbiol. Lett. 2011; 325: 1-9
        • Aldridge C.
        • Spence E.
        • Kirkilionis M.A.
        • Frigerio L.
        • Robinson C.
        Tat-dependent targeting of Rieske iron-sulphur proteins to both the plasma and thylakoid membranes in the cyanobacterium Synechocystis PCC6803.
        Mol. Microbiol. 2008; 70: 140-150
        • Karnauchov I.
        • Herrmann R.G.
        • Klösgen R.B.
        Transmembrane topology of the Rieske Fe/S protein of the cytochrome b6/f complex from spinach chloroplasts.
        FEBS Lett. 1997; 408: 206-210
        • Hubschmann T.
        • Wilde A.
        • Elanskaya I.
        • Shestakov S.V.
        • Borner T.
        A putative cytochrome c biogenesis gene in Synechocystis sp. PCC 6803.
        FEBS letters. 1997; 408: 201-205
        • Tichy M.
        • Vermaas W.
        Accumulation of pre-apocytochrome f in a Synechocystis sp. PCC 6803 mutant impaired in cytochrome c maturation.
        J. Biol. Chem. 1999; 274: 32396-32401
        • Inago H.
        • Sato R.
        • Masuda S.
        Regulation of light-induced H(+) extrusion and uptake by cyanobacterial homologs of the plastidial FLAP1, DLDG1, and Ycf10 in Synechocystis sp. PCC6803.
        Biochim. Biophys. Acta, Bioenerg. 2020; 1861148258
        • Graham J.E.
        • Bryant D.A.
        The biosynthetic pathway for myxol-2' fucoside (myxoxanthophyll) in the cyanobacterium Synechococcus sp. strain PCC 7002.
        J. Bacteriol. 2009; 191: 3292-3300
        • Ge H.
        • Fang L.
        • Huang X.
        • Wang J.
        • Chen W.
        • Liu Y.
        • Zhang Y.
        • Wang X.
        • Xu W.
        • He Q.
        • Wang Y.
        Translating divergent environmental stresses into a common proteome response through the histidine kinase 33 (Hik33) in a model cyanobacterium.
        Mol. Cell. Proteomics. 2017; 16: 1258-1274
        • Chitnis V.P.
        • Chitnis P.R.
        PsaL subunit is required for the formation of photosystem I trimers in the cyanobacterium Synechocystis sp. PCC 6803.
        FEBS letters. 1993; 336: 330-334
        • Lawlor D.W.
        Ke, B. Photosynthesis: photobiochemistry and photobiophysics.
        Annals of Botany. 2003; 91 (106-106)
        • Casella S.
        • Huang F.
        • Mason D.
        • Zhao G.Y.
        • Johnson G.N.
        • Mullineaux C.W.
        • Liu L.N.
        Dissecting the native architecture and dynamics of cyanobacterial photosynthetic machinery.
        Mol. Plant. 2017; 10: 1434-1448
        • MacGregor-Chatwin C.
        • Sener M.
        • Barnett S.F.H.
        • Hitchcock A.
        • Barnhart-Dailey M.C.
        • Maghlaoui K.
        • Barber J.
        • Timlin J.A.
        • Schulten K.
        • Hunter C.N.
        Lateral segregation of photosystem I in cyanobacterial thylakoids.
        Plant Cell. 2017; 29: 1119-1136
        • Minamihata K.
        • Goto M.
        • Kamiya N.
        Protein heteroconjugation by the peroxidase-catalyzed tyrosine coupling reaction.
        Bioconjug. Chem. 2011; 22: 2332-2338
        • Rogers M.S.
        • Hurtado-Guerrero R.
        • Firbank S.J.
        • Halcrow M.A.
        • Dooley D.M.
        • Phillips S.E.
        • Knowles P.F.
        • McPherson M.J.
        Cross-link formation of the cysteine 228-tyrosine 272 catalytic cofactor of galactose oxidase does not require dioxygen.
        Biochemistry. 2008; 47: 10428-10439
        • Bhaskar B.
        • Immoos C.E.
        • Shimizu H.
        • Sulc F.
        • Farmer P.J.
        • Poulos T.L.
        A novel heme and peroxide-dependent tryptophan–tyrosine cross-link in a mutant of cytochrome c peroxidase.
        J. Mol. Biol. 2003; 328: 157-166
        • Udeshi N.D.
        • Pedram K.
        • Svinkina T.
        • Fereshetian S.
        • Myers S.A.
        • Aygun O.
        • Krug K.
        • Clauser K.
        • Ryan D.
        • Ast T.
        • Mootha V.K.
        • Ting A.Y.
        • Carr S.A.
        Antibodies to biotin enable large-scale detection of biotinylation sites on proteins.
        Nat. Methods. 2017; 14: 1167-1170
        • Anderson C.M.
        • Gray J.
        Cleavage of the precursor of pea chloroplast cytochrome f by leader peptidase from Escherichia coli.
        FEBS letters. 1991; 280: 383-386
        • Gray J.C.
        Cytochrome f: Structure, function and biosynthesis.
        Photosynth. Res. 1992; 34: 359-374
        • Roose J.L.
        • Pakrasi H.B.
        The Psb27 protein facilitates manganese cluster assembly in photosystem II.
        J. Biol. Chem. 2008; 283: 4044-4050
        • Komenda J.
        • Knoppová J.
        • Kopečná J.
        • Sobotka R.
        • Halada P.
        • Yu J.
        • Nickelsen J.
        • Boehm M.
        • Nixon P.J.
        The Psb27 assembly factor binds to the CP43 complex of photosystem II in the cyanobacterium Synechocystis sp. PCC 6803.
        Plant Physiol. 2012; 158: 476-486
        • Liu H.
        • Huang R.Y.
        • Chen J.
        • Gross M.L.
        • Pakrasi H.B.
        Psb27, a transiently associated protein, binds to the chlorophyll binding protein CP43 in photosystem II assembly intermediates.
        Proc. Natl. Acad. Sci. U. S. A. 2011; 108: 18536-18541
        • Sveshnikov D.
        • Funk C.
        • Schroder W.P.
        The PsbP-like protein (sll1418) of Synechocystis sp. PCC 6803 stabilises the donor side of Photosystem II.
        Photosynth. Res. 2007; 93: 101-109
        • Ishikawa Y.
        • Schroder W.P.
        • Funk C.
        Functional analysis of the PsbP-like protein (sll1418) in Synechocystis sp. PCC 6803.
        Photosynth. Res. 2005; 84: 257-262
        • Komenda J.
        • Nickelsen J.
        • Tichy M.
        • Prasil O.
        • Eichacker L.A.
        • Nixon P.J.
        The cyanobacterial homologue of HCF136/YCF48 is a component of an early photosystem II assembly complex and is important for both the efficient assembly and repair of photosystem II in Synechocystis sp. PCC 6803.
        J. Biol. Chem. 2008; 283: 22390-22399
        • Meurer J.
        • Plucken H.
        • Kowallik K.V.
        • Westhoff P.
        A nuclear-encoded protein of prokaryotic origin is essential for the stability of photosystem II in Arabidopsis thaliana.
        EMBO J. 1998; 17: 5286-5297
        • Stengel A.
        • Gügel I.L.
        • Hilger D.
        • Rengstl B.
        • Jung H.
        • Nickelsen J.
        Initial steps of photosystem II de novo assembly and preloading with manganese take place in biogenesis centers in Synechocystis.
        Plant Cell. 2012; 24: 660-675
        • Schottkowski M.
        • Gkalympoudis S.
        • Tzekova N.
        • Stelljes C.
        • Schunemann D.
        • Ankele E.
        • Nickelsen J.
        Interaction of the periplasmic PratA factor and the PsbA (D1) protein during biogenesis of photosystem II in Synechocystis sp. PCC 6803.
        J. Biol. Chem. 2009; 284: 1813-1819
        • Rengstl B.
        • Oster U.
        • Stengel A.
        • Nickelsen J.
        An intermediate membrane subfraction in cyanobacteria is involved in an assembly network for Photosystem II biogenesis.
        J. Biol. Chem. 2011; 286: 21944-21951
        • Klinkert B.
        • Ossenbuhl F.
        • Sikorski M.
        • Berry S.
        • Eichacker L.
        • Nickelsen J.
        PratA, a periplasmic tetratricopeptide repeat protein involved in biogenesis of photosystem II in Synechocystis sp. PCC 6803.
        J. Biol. Chem. 2004; 279: 44639-44644
        • Inagaki N.
        • Yamamoto Y.
        • Satoh K.
        A sequential two-step proteolytic process in the carboxyl-terminal truncation of precursor D1 protein inSynechocystissp. PCC68031.
        FEBS Lett. 2001; 509: 197-201
        • Ivleva N.B.
        • Sidoruk K.V.
        • Pakrasi H.B.
        • Shestakov S.V.
        Investigation of the functional role of Ctp proteins in the cyanobacterium Synechocystis sp PCC 6803.
        Microbiology. 2002; 71: 433-437
        • Roberts I.N.
        • Lam X.T.
        • Miranda H.
        • Kieselbach T.
        • Funk C.
        Degradation of PsbO by the Deg protease HhoA is thioredoxin dependent.
        PLoS ONE. 2012; 7e45713
        • Dong W.
        • Wang J.
        • Niu G.
        • Zhao S.
        • Liu L.
        Crystal structure of the zinc-bound HhoA protease from Synechocystis sp. PCC 6803.
        FEBS Lett. 2016; 590: 3435-3442
        • Bertini I.
        • Ciurli S.
        • Dikiy A.
        • Fernandez C.O.
        • Luchinat C.
        • Safarov N.
        • Shumilin S.
        • Vila A.J.
        The first solution structure of a paramagnetic copper(II) protein: the case of oxidized plastocyanin from the cyanobacterium Synechocystis PCC6803.
        J Am Chem Soc. 2001; 123: 2405-2413
        • Duran R.V.
        • Hervas M.
        • De La Rosa M.A.
        • Navarro J.A.
        The efficient functioning of photosynthesis and respiration in Synechocystis sp. PCC 6803 strictly requires the presence of either cytochrome c6 or plastocyanin.
        The Journal of biological chemistry. 2004; 279: 7229-7233
        • Tottey S.
        • Waldron K.J.
        • Firbank S.J.
        • Reale B.
        • Bessant C.
        • Sato K.
        • Cheek T.R.
        • Gray J.
        • Banfield M.J.
        • Dennison C.
        • Robinson N.J.
        Protein-folding location can regulate manganese-binding versus copper- or zinc-binding.
        Nature. 2008; 455: 1138-1142
        • Tarrago L.
        • Laugier E.
        • Rey P.
        Protein-repairing methionine sulfoxide reductases in photosynthetic organisms: gene organization, reduction mechanisms, and physiological roles.
        Mol. Plant. 2009; 2: 202-217
        • Holmgren A.
        Antioxidant function of thioredoxin and glutaredoxin systems.
        Antioxid. Redox Signal. 2000; 2: 811-820
        • Lennartz K.
        • Plucken H.
        • Seidler A.
        • Westhoff P.
        • Bechtold N.
        • Meierhoff K.
        HCF164 encodes a thioredoxin-like protein involved in the biogenesis of the cytochrome b(6)f complex in Arabidopsis.
        Plant Cell. 2001; 13: 2539-2551
        • Schiene-Fischer C.
        Multidomain Peptidyl Prolyl cis/trans Isomerases.
        Biochim. Biophys. Acta. 2015; 1850: 2005-2016
        • Kieselbach T.
        • Schroder W.P.
        The proteome of the chloroplast lumen of higher plants.
        Photosynth. Res. 2003; 78: 249-264
        • Perez-Riverol Y.
        • Csordas A.
        • Bai J.
        • Bernal-Llinares M.
        • Hewapathirana S.
        • Kundu D.J.
        • Inuganti A.
        • Griss J.
        • Mayer G.
        • Eisenacher M.
        • Perez E.
        • Uszkoreit J.
        • Pfeuffer J.
        • Sachsenberg T.
        • Yilmaz S.
        • Tiwary S.
        • Cox J.
        • Audain E.
        • Walzer M.
        • Jarnuczak A.F.
        • Ternent T.
        • Brazma A.
        • Vizcaino J.A.
        The PRIDE database and related tools and resources in 2019: improving support for quantification data.
        Nucleic Acids Res. 2019; 47: D442-D450
        • Moller S.
        • Croning M.D.
        • Apweiler R.
        Evaluation of methods for the prediction of membrane spanning regions.
        Bioinformatics. 2001; 17: 646-653
        • Almagro Armenteros J.J.
        • Tsirigos K.D.
        • Sonderby C.K.
        • Petersen T.N.
        • Winther O.
        • Brunak S.
        • von Heijne G.
        • Nielsen H.
        SignalP 5.0 improves signal peptide predictions using deep neural networks.
        Nat. Biotechnol. 2019; 37: 420-423