Abstract

   graphic file with name ao0c06111_0009.jpg

   Terrestrial cyanobacteria, originated from aquatic cyanobacteria,
   exhibit a unique mechanism for drought adaptation during long-term
   evolution. To elucidate this diverse adaptive mechanism exhibited by
   terrestrial cyanobacteria from the post-translation modification
   aspect, we performed a global phosphoproteome analysis on the abundance
   of phosphoproteins in response to dehydration using Nostoc
   flagelliforme, a kind of terrestrial cyanobacteria having strong
   ecological adaptability to xeric environments. A total of 329
   phosphopeptides from 271 phosphoproteins with 1168 phosphorylation
   sites were identified. Among these, 76 differentially expressed
   phosphorylated proteins (DEPPs) were identified for each dehydration
   treatment (30, 75, and 100% water loss), compared to control. The
   identified DEPPs were functionally categorized to be mainly involved in
   a two-component signaling pathway, photosynthesis, energy and
   carbohydrate metabolism, and an antioxidant system. We concluded that
   protein phosphorylation modifications related to the reactive oxygen
   species (ROS) signaling pathway might play an important role in
   coordinating enzyme activity involved in the antioxidant system in N.
   flagelliforme to adapt to dehydration stress. This study provides deep
   insights into the extensive modification of phosphorylation in
   terrestrial cyanobacteria using a phosphoproteomic approach, which may
   help to better understand the role of protein phosphorylation in key
   cellular mechanisms in terrestrial cyanobacteria in response to
   dehydration.

1. Introduction

   With cyanobacteria’s evolution from aquatic to terrestrial
   cyanobacteria, water availability becomes the most important factor
   influencing its growth and evaluation. Over time, drought stress is
   being intensified due to extreme global environmental change.^[40]1
   Terrestrial cyanobacteria exhibit a unique mechanism of drought
   adaptation to acclimate water loss.^[41]2Nostoc flagelliforme belongs
   to a terrestrial nitrogen-fixing cyanobacterium class, with dark brown
   colonies when dry, while after absorbing water, it turns light brown or
   dark green. This species is found in both semiarid and arid steppes and
   is widely distributed in western and north-western China, where it is
   exposed to severe environmental stress from seasonal changes in
   temperature and precipitation.^[42]3 Consequently, N. flagelliforme has
   a high degree of ecological adaptability to xeric environments.
   Therefore, this species can survive in an extremely dry environment for
   decades and quickly regains its active physiological and metabolic
   state when it reabsorbs water.^[43]4

   In the past few years, an increasing trend is observed in investigating
   the transcriptomic and proteomic levels in cyanobacterium under
   different stress conditions, especially under drought
   stress.^[44]5−[45]7 In this regard, hundreds of genes and proteins are
   being screened out that are related to stress responses. Furthermore, a
   large number of protein post-translation modifications (PTMs) are also
   being reported to be associated with different plant stress
   responses.^[46]8,[47]9 Among them, one of the most studied
   modifications is protein phosphorylation, which is involved in various
   regulatory mechanisms, for instance, transcription and translation,
   metabolism, homeostasis, protein degradation, and cellular signaling
   and communication.^[48]10 Therefore, to have a better insight into the
   role of protein phosphorylation, numerous large-scale phosphoproteomic
   studies have been performed to elucidate its role in the growth,
   development, and diverse response mechanisms in various plants and
   cyanobacteria, such as Arabidopsis, Brachypodium distachyon L., and
   Synechococcus sp. Strain PCC 7002.^[49]11−[50]14 However, protein
   phosphorylation in other species is still poorly understood,
   particularly in terrestrial nitrogen-fixing cyanobacterium species, and
   specifically, the phosphoproteomic characterization of N. flagelliforme
   subjected to drought stress has received less attention so far.

   Here, we performed a systematic study on the identities and functional
   roles of the Ser/Thr/Tyr phosphoproteins in N. flagelliforme. First, to
   explore Ser/Thr/Tyr phosphoprotein in N. flagelliforme, we did a
   genome-wide and site-specific phosphoproteomic analysis of N.
   flagelliforme by employing high-accuracy mass spectrometry in
   conjunction with biochemical enrichment of phosphopeptides from
   digested cell lysates. Furthermore, dynamic changes of phosphorylation
   modification of N. flagelliforme phosphoproteome were investigated in
   response to different dehydration conditions. This high resolution for
   terrestrial cyanobacterial phosphorylated proteins on Ser/Thr/Tyr
   residues provides further evidence in support of the emerging view that
   protein phosphorylation is not just limited to eukaryotes but a general
   and fundamental regulatory mechanism, allowing for extensive functional
   and evolutionary study in cyanobacteria. In addition to this, we also
   functionally categorized the identified phosphoproteins involved in
   several important biological processes (BPs) into an interactive map.
   This provides an exquisite interaction network of phosphoproteins in N.
   flagelliforme, a terrestrial cyanobacterium. Our results may help us to
   better look into the role of phosphorylation in key cellular mechanisms
   in terrestrial cyanobacteria under dehydration stress.

2. Results

2.1. Phosphoproteomics Establishment in N. flagelliforme

   In our previous study, we used a gel-based proteomic analysis on N.
   flagelliforme in response to dehydration and rehydration.^[51]6 To
   collect a large-scale data set on Ser/Thr/Tyr phosphorylation sites in
   N. flagelliforme, we hence performed a systematic investigation to
   characterize the phosphoproteome in N. flagelliforme under different
   water statuses ([52]Figure [53]1
   A). To do this, different techniques were combined to generate a large
   set of data, including TiO[2] affinity chromatography, iTRAQ labeling,
   nLC–tandem mass spectrometry (MS/MS) analysis; phosphoproteomic methods
   and two different search algorithms (MASCOT and pFind) were used. The
   identified phosphopeptides were further validated by preliminary manual
   inspection of MS/MS data; our approach resulted in the identification
   of 786 N. flagelliforme proteins detected by liquid chromatography
   MS/MS (LC–MS/MS) and 329 unique phosphopeptides and 1168
   phosphorylation sites from 271 N. flagelliforme phosphoproteins
   ([54]Figure S1), with high confidence through the combined use of
   protein/peptide prefractionation, which represents an informative
   phosphorylation data set obtained from the cyanobacteria, and Cyanobase
   database of N. flagelliforme
   ([55]http://genome.kazusa.or.jp/cyanobase/Nostoc flagelliforme CCNUN1).

Figure 1.

   [56]Figure 1
   [57]Open in a new tab

   Workflow of the experiment to analyze the phosphoproteome, a
   representative MS/MS spectrum, and general description of the
   phosphopeptides identified. (A) Overview of the analytical workflow
   used in this study. Proteins were prefractionated using gel-free
   methods, followed by trypsin digestion and TiO[2] enrichment of
   phosphopeptides. Phosphopeptides were separated by nano-LC–MS/MS and
   mass measured and fragmented using the mass spectrometer. (B) Example
   of an MS/MS spectrum assigned to VSSKIGVIETLLEK from phosphoglycerate
   kinase (PGK) ([58]WP_100901575.1). The b and y ions including loss of
   ammonia and water were considered when we calculated the PTM score. (C)
   Distribution of singly, doubly, and triply phosphorylated peptides. The
   photos in panel (A) were taken by Wenyu Liang, an author listed in this
   manuscript.

   To better analyze the reprogrammed biological pathways of N.
   flagelliforme in response to dehydration, the accuracy of localization
   of these phosphorylation sites was further checked by estimating the
   probability-based post-translational modification (PTM) scores. There
   are 224 unique phosphopeptides from 271 phosphoproteins determined with
   localization probability >0.75 ([59]Table S2). Details of the
   identified phosphopeptides, including their protein IDs, sequences,
   search algorithm scores, PTM scores, and localized P-values, for the
   samples with dehydration treatments (30, 75, and 100%) and control (0%)
   are provided in [60]Tables S3–S5. A representative example for mass
   spectrum on a phosphopeptide sequence (VSSKIGVIETLLEK from
   phosphoglycerate kinase, [61]WP_100901575.1) was presented ([62]Figure
   [63]1
   B). Phosphorylation sites that were occupied with probability >0.75
   were categorized as class I phosphorylation sites and identified
   (“Localization p value” = 1), whereas those with localization
   probability <0.75 were classified as ambiguous phosphorylation sites
   ([64]Tables S3–S5). The raw data has been deposited in a publicly
   accessible database Peptideatlas
   ([65]http://www.peptideatlas.org)^[66]29 and can be accessed with
   identifier PASS00119 ([67]http://www.peptideatlas.org/PASS/PASS00119).
   The phosphoproteome analysis in the current study for N. flagelliforme
   contained a relatively high number of proteins; in total, 344
   phosphopeptides have been detected, including 329 phosphopeptides that
   are unique (having single phosphoprotein), while the remaining 15
   phosphopeptides are mapped to multiple proteins ([68]Figure [69]1C).

2.2. Bioinformatic Analysis on Differential Expression of Phosphopeptides

   Differentially expressed phosphoproteins (DEPPs) were identified (the
   screening method is presented in [70]Section 5). There are in total 76
   DEPPs in N. flagelliforme exposed to all combinations between
   dehydration treatments and control ([71]Figure [72]2
   A). These DEPPs show distinct expression patterns across different
   dehydration treatments ([73]Figure [74]2B). Most DEPPs exhibited
   gradually decreased trends when N. flagelliforme was exposed to
   exacerbating water loss ([75]Figure [76]2C). There were 46, 10, 10, 13,
   and 21% DEPPs localized in the cytoplasm, inner membrane, extracellular
   matrix, periplasm, and outer membrane, respectively ([77]Figure
   [78]2D).

Figure 2.

   [79]Figure 2
   [80]Open in a new tab

   Characterization of differentially expressed phosphopeptides (DEPs) in
   N. flagelliforme exposed to different dehydration treatments. (A)
   Comparison of differentially expressed phosphopeptides (DEPs) among the
   group samples with different dehydration treatments (30, 75, and 100%
   water loss) and control (0%) based on one-way analysis of variance
   (ANOVA) analysis. Significant change in abundance >1.2 fold, p value <
   0.05. (B) Heatmap representing the results of DEPs; hierarchical
   clustering results are represented by a treelike thermograph in which
   each row represents a phosphopeptide and each column represents a set
   of samples. The abundance represents the logarithmic values of
   expression amounts of the DEPs in different samples
   (log[2] expression). Different colors are shown in the thermogram, with
   red representing significantly upregulated DEPs, green representing
   significantly downregulated DEPs, and gray representing no quantitative
   information for DEP. (C) Cluster analysis of differential expression
   pattern for the 60 DEPs. (D) Subcellular location of DEPs.

   The 76 DEPPs were further analyzed with Gene Ontology (GO) and Kyoto
   Encyclopedia of Genes and Genomes (KEGG). The top 20 enriched terms
   from GO analysis shows that biological pathways related to
   protein–chromophore linkage, photosynthesis, oxidoreductase, protein
   modification process, cellular protein modification process, and
   macromolecule modification; molecular functions (MFs) related to
   phosphoglucosamine mutase activity and phosphotransferases; and
   cellular components related to light-harvesting complex, membrane part,
   chloroplast, protein complex, phycobilisome, vesicle, intracellular
   vesicle, cytoplasmic vesicle part, light-harvesting complex,
   cytoplasmic vesicle, chromophore, and membrane-bounded organelle are
   significantly enriched in the list of DEPPs ([81]Figure [82]3
   A).

Figure 3.

   [83]Figure 3
   [84]Open in a new tab

   GO and KEGG pathway analysis of phosphoproteins in N. flagelliforme
   between 75% water loss dehydration treatment and control. (A)
   Functional enrichment analysis of GO in group one-way ANOVA. The
   abscissa indicates the GO functional classification, which is divided
   into three major categories, biological processes (BPs), molecular
   function (MFs), and cell components (CCs); ordinates denote the number
   of differentially expressed proteins under each functional
   classification; the color bar denotes the significance of enriched GO
   functional classification based on the P-value calculated from Fisher’s
   exact test; the color gradient from orange to red represents the size
   of the P value: the closer the red, the smaller the P value and the
   higher the corresponding GO functional category richness level. The
   label above the bar chart shows the enrichment factor (richFactor ≤1),
   which indicates that the number of differentially expressed proteins
   annotated to a GO functional class account for all of the identified
   proteins annotated to that GO functional class. (B) Ordinates in the
   graph represent the significantly enriched KEGG pathway; the abscissa
   represents the number of differentially expressed proteins contained in
   each KEGG pathway, and the bar color represents the significance of the
   enriched KEGG pathway, i.e., the P value is calculated based on
   Fisher’s exact test. The color gradient represents the size of the P
   value, and the color changes from orange to red. The closer the red,
   the smaller the P value and the higher the significant level of the
   corresponding KEGG pathway enrichment.

   The top 20 terms from KEGG pathway enrichment analysis on DEPPs
   indicate that photosynthesis-antenna proteins, the pentose phosphate
   pathway, carbon fixation in photosynthetic organisms,
   glycolysis/gluconeogenesis, RNA degradation, methane metabolism,
   glutathione metabolism, oxidative phosphorylation, photosynthesis,
   thymidine metabolism, aminoacyl-tRNA biosynthesis, amino sugar
   metabolism, fructose and mannose metabolism, glyoxylate and
   dicarboxylate metabolism, the two-component system, and purine
   metabolism are significantly enriched in the list of DEPPs ([85]Figure
   [86]3
   B). This result suggests that some phosphoproteins related to
   photosynthetic carbon metabolism are likely to play important roles in
   the biological response to dehydration in N. flagelliforme. As revealed
   by KEGG analysis on photosynthesis-antenna proteins, we found that
   allophycocyanin A, allophycocyanin B, allophycocyanin F, phycocyanin A,
   and phycocyanin B were downregulated, but photosynthetic electron
   transport subunit H was upregulated ([87]Figure S2). Moreover, other
   biological pathways were also found to be significantly enriched in the
   DEPPs, such as the two-component system, sucrose metabolic pathway,
   photosystem proteins, protein metabolism, protein kinase, and ABC
   transport system ([88]Figure S3 and Table S2).

   Carbohydrate and energy metabolism are generally assumed to be more
   susceptible to dehydration. To confirm whether the DEPPs related to the
   photosynthetic carbon metabolic pathway are involved in dehydration
   response, we tested the gene expression levels for GADPH, 6PGDH, G6PDH,
   PRK, FBA, and PGK and found that most genes showed a gradual decrease
   in the expression levels along with exacerbating water loss ([89]Figure
   [90]4
   ). In particular, most genes under the 100% dehydration condition were
   downregulated by 10–30% compared to control, except for PRK ([91]Figure
   [92]4A–F). Interestingly, we found that the metabolites related to the
   sucrose metabolic pathway were increased following severe dehydration
   stress, including fructose, sucrose, and glycogen contents ([93]Figure
   S4), suggesting that sucrose synthesis increased after dehydration
   treatment in N. flagelliforme.

Figure 4.

   [94]Figure 4
   [95]Open in a new tab

   Reverse transcription real-time polymerase chain reaction (RT-qPCR)
   testing for differential phosphoproteins related to carbon metabolism
   in N. flagelliforme on dehydration. Each bar data represents the
   average of three independent replicates (± standard error (SE)). (A–F)
   Dehydration-induced relative expression of glyceraldehyde-3-phosphate
   dehydrogenase (GAPDH), 6-phosphoglueonate dehydrogenase (6PGDH),
   glucose 6-phosphate dehydrogenase (G6PDH), phosphoribulosekinase (PRK),
   fructose-1,6-bisphosphatase (FBP), and phosphoglycerate kinase (PGK),
   respectively. The adjacent alphabetic letters to the bars reflect the
   significant levels based on one-way ANOVA (P < 0.05).

2.3. Changes in Reactive Oxygen Species (ROS) Levels and Antioxidant Enzyme
Activities

   As mentioned above, the oxidoreductase pathway was significantly
   enriched in the list of DEPPs ([96]Figure [97]3
   A). Therefore, we measured the superoxide anion content, H[2]O[2]
   level, and antioxidant enzyme activities in dehydrated N.
   flagelliforme. The levels of phosphorylation for DEPPs related to the
   antioxidant system were gradually decreased following dehydration
   treatments ([98]Figure [99]5A). Similarly, it showed a gradual decrease
   in activities of peroxidase (POD), superoxide dismutase (SOD), and
   glutathione-S-transferase (GST) upon dehydration, although the
   activities of catalase (CAT) and glutathione reductase (GR) increased
   upon dehydration. It leads to increased contents of oxygen free radical
   (OFR) and H[2]O[2] levels accordingly ([100]Figure [101]5B–H).

Figure 5.

   [102]Figure 5
   [103]Open in a new tab

   Accumulation of the antioxidant system and activities of antioxidant
   enzymes in N. flagelliforme on dehydration. (A) Heatmap of differently
   expressed phosphorylated proteins (DEPPs) related to the antioxidant
   system. (B–H) Enzymatic activities and substrates related to ROS
   signaling pathway involved in dehydration response, including catalase
   (CAT) activities, peroxidase (POD), superoxide dismutase (SOD)
   activities, and glutathione-S-transferase (GST) activities and
   glutathione reductase (GR), oxygen free radical (OFR), and H[2]O[2]
   contents. For panels (B–H), each bar data represents the average of
   three independent replicates (±SE). The alphabetic letters adjacent to
   the bars reflect the significant levels based on one-way ANOVA (P <
   0.05).

2.4. Phosphorylation Motif Analysis in N. flagelliforme

   To determine whether a common sequence preference occurred for
   phosphopeptides at Ser/Thr/Tyr residues, we evaluated 1168
   phosphorylation sites ([104]Figure [105]6
   ). To identify the possible specific motifs flanking phosphorylated
   lysine, we extracted the over-represented motifs. Six major kinds of
   motifs were identified in N. flagelliforme, and the numbers of
   phosphopeptides for different phosphorylation motifs ranged from 11 to
   35 ([106]Figure [107]6A), among which four kinds of conserved motifs
   were found near to the phosphorylated serine site, including [SxE],
   [SxxQ], [SxxxxxY], and [SxxxxD] ([108]Figure [109]6B). Besides,
   significantly enriched phosphorylation motifs near threonine sites were
   [TxL] and [TG] ([110]Figure [111]6B). These findings suggest that the
   possible phosphorylated peptide motifs of N. flagelliforme are located
   downstream of serine or threonine. Interestingly, the [SP] motif was
   not found in N. flagelliforme. The amino acids surrounding the
   phosphorylation sites were shown in a heatmap. According to the
   heatmap, serine (S) and threonine (T) were significantly
   over-represented in positions −1 and −2, respectively, and other
   residues like glutamic acid (E), glutamine (Q), tyrosine (Y), aspartic
   acid (D), leucine (L), and glycine (G) were highly present in position
   +2, +3, +6, +5, +2, and +1, respectively. The conserved motifs found in
   this study not only help to predict the phosphorylation sites of
   unknown phosphorylated proteins but also provide a basis for the
   relationship between the phosphorylated protein and its kinase.

Figure 6.

   [112]Figure 6
   [113]Open in a new tab

   Phosphorylation motif analysis on all phosphorylated sites in N.
   flagelliforme across dehydration treatments and control. (A) Number of
   identified peptides containing phosphorylation sites in each motif. (B)
   Sequence motif analysis of phosphorylation sites. (C) Relative
   abundance of amino acid residues flanking the phosphorylation sites
   represented by an intensity map. The intensity map shows the relative
   abundance of six amino acids from the phosphorylation site. The colors
   in the intensity map represent the log[10] of the ratio of frequencies
   (red shows enrichment, green shows depletion). Default: occurrences =
   10, significance = 0.00018, background =
   P17036_NCBI_Nostoc_flagelliforme_18909_20171228.

2.5. Regulatory Network of Differentially Expressed Phosphoproteins

   To elucidate the interactions among DEPPs and also with their potential
   substrates, protein–protein interaction (PPI) analysis was conducted
   using the STRING database ([114]https://string-db.org/). The PPI
   network was built using 50 DEPPs from four functional categories:
   photosynthesis; carbohydrate and energy metabolism; ROS scavenging; and
   DNA, RNA, and protein metabolism ([115]Figure S5). The network revealed
   that the kinases and phosphatases, such as histidine kinase,
   nucleoside-diphosphate kinase, serine/threonine protein kinase, acetate
   kinase, and phosphoglycerate kinase, exhibited substantial interactions
   with the phosphoproteins from the four functional categories. These
   results suggest that they could be involved in the phosphorylation or
   dephosphorylation of protein substrates in N. flagelliforme upon
   dehydration.

   Collectively, dehydration stress-induced decreased levels of
   phosphorylation for most DEPPs are related to photosynthetic light
   reaction, Calvin cycle, and sucrose metabolism ([116]Figure [117]7
   ). These DEPPs include ATP synthase (ATPsynth), Rubisco,
   phosphoglycerate kinase (PGK), fructose-bisphosphate aldolase (FBA),
   and glycogen synthase (GlgA). The levels of phosphorylation for DEPPs
   related to the reactive oxygen species (ROS) signaling pathway were
   also decreased, including catalase (CAT), peroxidase (POD), ascorbate
   peroxidase (APX), glutathione-S-transferase (GST), and glutathione
   reductase (GR), while the phosphorylation levels for only a few DEPPs
   were increased, including cytochrome b6–f complex subunit H (PetH) and
   histidine kinase (HistK) involved in the two-component signaling
   pathway.

Figure 7.

   [118]Figure 7
   [119]Open in a new tab

   Schematic presentation of phosphoproteins related to photosynthesis,
   carbohydrate metabolism, energy conversion, and ROS scavenging of N.
   flagelliforme in response to drought stress. Dehydration stress induced
   a number of signaling molecules (SMs) that were accumulated in the
   cytoplasm. These SMs are then involved in the two-component signaling
   system, activating some protein kinases, such as histidine kinase
   (HistK) and downstream transcription factors (TFs). Under extremely
   server dehydration stress, some defensive proteins related to ROS
   scavenging were impaired and degraded by phosphorylation modification
   and some unknown pathways. The increased and decreased phosphorylation
   levels of proteins were depicted in red and blue, respectively. The
   dotted arrow was used to represent the potential cross talk between the
   photosynthetic light reaction and ROS scavenging through supplying
   reducing power by NADPH.

3. Discussion

3.1. Ser/Thr/Tyr Phosphorylation Involved in the Two-Component Signaling
Pathway

   Two-component systems are common signaling pathways in bacteria that
   mediate a wide range of adaptive cellular responses.^[120]30,[121]31
   Phosphorylation signaling proceeds via His–Asp phosphorelay cascades
   involving two central partners, i.e., histidine protein kinase and
   response regulator protein.^[122]32 Phosphorylation modification of
   proteins in bacteria is usually ascribed to the two-component system
   for signal transduction, whereas eukaryotic organisms engage Ser/Thr
   kinases and phosphatases for this kind of modification.^[123]33 In some
   cases, two systems are interactive with each other. Genetic and
   molecular studies demonstrated that some Ser/Thr/Tyr kinases or
   phosphatases could be linked to two-component systems in the same
   signal transduction pathways in some cyanobacterial strains such as
   Anabaena sp. PCC 7120,^[124]34Synechocystis sp. PCC
   6803^[125]35,[126]36 and Synechococcus sp. Strain PCC
   7002.^[127]14,[128]37,[129]38 In particular, Synechocystis has
   two-component system eukaryotic-type Ser/Thr kinases, playing a key
   role in regulating cyanobacterial physiology under abiotic
   stress.^[130]39 In this study, seven phosphorylated proteins (histidine
   kinase, Ser/Thr protein kinase, K(+)-transporting ATPase subunit B,
   alkaline phosphatase, nucleoside-diphosphate kinase, chemotaxis family,
   and hybrid sensor histidine kinase/response regulator) are involved in
   the two-component signaling pathway in N. flagelliforme ([131]Table
   S2). Importantly, the phosphorylation level of histidine kinase was
   upregulated in colonies upon dehydration ([132]Figure S3). These
   findings suggest that some Ser/Thr/Tyr kinases or phosphatases
   identified in our study could also be coupled to two-component systems
   in the signal transduction pathways in N. flagelliforme on dehydration.
   However, the function of these protein kinases and phosphatases
   involved in N. flagelliforme signal transduction in response to
   dehydration remains unclear and needs to be sorted out. Most probably,
   these proteins could participate in signal transduction via a cascade
   of Ser/Thr/Tyr phosphorylation or dephosphorylation, which is similar
   to that observed in eukaryotes.^[133]32,[134]40

3.2. Phosphorylation Modifications Play Important Roles in the Response of
the Photosynthetic Pathway to Dehydration Stress

   Combining phosphorylation events with the key photosynthetic pathway
   could help to facilitate the integration of phosphoproteomic data with
   biological function, and it is well established that protein
   phosphorylation is involved in the regulation of the photosynthesis
   process in cyanobacteria,^[135]41 which is consistent with the evidence
   observed in the current study that there are many DEPPs related to the
   photosynthetic carbon and energy metabolic pathway during dehydration
   response in N. flagelliforme ([136]Table S2 and [137]Figure [138]7
   ). Besides, some identified phosphoproteins involved in photosynthetic
   light reaction pathways are always interactive with each other, such as
   cpcA, cpcB, apcA, apcB, and apcF in N. flagelliforme exposed to severe
   dehydration stress (100% water loss) ([139]Table S2 and Figure S5).
   Among these DEPPs, the presented phosphorylation of cpcA, cpcB, apcA,
   and apcB was previously reported in Synechococcus sp. Strain PCC
   7002^[140]14 and Synechocystis sp. PCC 6803.^[141]42

   In higher plants, it is widely excepted that reversible phosphorylation
   of light-harvesting complex II (LHCII) makes a balance of the
   excitation pressure between the two photosystems.^[142]43 Therefore,
   phosphorylation of light-harvesting complexes phycobilisome in N.
   flagelliforme are likely to be involved in the regulation of the
   distribution of light energy between photosystem I (PSI) and
   photosystem II (PSII) in response to the environmental stress, similar
   to that of higher plants. Although many proteins related to
   light-harvesting complexes phycobilisome such as cpcA, cpcB, apcA,
   apcB, and apcF, as well as PsbO, as a component of reaction center in
   PSII, photosystem II oxygen-evolving enhancer protein1 and DUF2382
   domain-containing protein as photosystem reaction center subunit were
   phosphorylated; while only a few phosphoproteins involved in PSI and
   PSII show substantial changes for phosphorylation levels ([143]Figure
   [144]7
   ), which indicates the integrity of PSI and PSII photosynthetic system
   could allow very quick response to rehydration for N.
   flagelliforme.([145]44) Several photosynthetic proteins are reported to
   be involved in various abiotic stress, including water shortage stress
   as reported by Zhang et al.,^[146]45 however, the phosphoproteins,
   identified in our study are involved in the response to water stress,
   suggesting the protein degradation or activity could play important
   role in this response, explained by Parry and his co-workers,^[147]46
   this could be due to regulation of some upstream protein kinase as
   suggested by Chen et al.,^[148]47 Therefore, in this study, we
   confirmed that some proteins involved in the photosynthetic pathway can
   be phosphorylated, and the modification of phosphorylation is important
   for plant adaption to abiotic stress.

3.3. Phosphorylated Proteins Are Involved in the Process of Carbohydrate
Metabolism and Energy Conversion

   Metabolic flux and carbon source utilization in cyanobacterium and
   eukaryotes are coordinated by phosphorylation-targeting metabolic
   proteins.^[149]12,[150]48 Previous studies have found that many protein
   phosphorylation modifications are involved in starch synthesis in
   plants,^[151]49,[152]50 suggesting that phosphorylation modification of
   proteins plays an important role in regulating carbohydrate metabolism,
   and this modification resulting in sucrose accumulation is beneficial
   to N. flagelliforme to adapt to the arid environment.^[153]5 In this
   study, we confirmed that extensive phosphorylation modifications were
   present in photosynthetic carbon metabolism such as ribulose
   bisphosphate carboxylase large subunit, carbon dioxide-concentrating
   mechanism protein CcmM, UDP-glucose 6-dehydrogenase, glycogen synthase
   GlgA, and NAD(P)-dependent oxidoreductase ([154]Table S2).
   Interestingly, phosphorylation levels for most proteins, such as type I
   glyceraldehyde-3-phosphate dehydrogenase, fructose-bisphosphate
   aldolase class II, ribulose bisphosphate carboxylase small subunit,
   6-phosphogluconolactonase, PGK, and GAPDH, were downregulated in
   colonies on dehydration ([155]Table S2 and Figure S3). Furthermore, the
   changes of mRNA levels of rbcS, PGK, FBP, 6PGDH, G6PDH, and GlgA were
   consistent with their phosphorylation levels except for PRK
   ([156]Figure [157]4
   A–F). Since carbon metabolism is a core metabolic pathway that affects
   photosynthetic rates and TCA energy metabolism, many reports show that
   proteins involved in carbon metabolism are highly abundant and feasible
   to be modified by various chemicals and post-translational
   modifications.^[158]51 In this study, we found the proteins, such as
   PGK, ATPsynth, GlgA, Rubisco, that can be phosphorylated; these
   proteins are differently abundant among different dehydration stresses,
   and these proteins are closely linked with both carbon metabolism and
   energy conversion metabolic pathway, suggesting the important roles of
   these phosphorylated proteins in these reactions under abiotic stress
   as reported by Wingler et al. and Li et al.^[159]52,[160]53 Therefore,
   we speculated that phosphorylated proteins played an important role in
   carbohydrate metabolism and energy conversion and that downregulation
   of some phosphorylated proteins in colonies on dehydration might help
   to promote energy conversion.

3.4. Phosphoproteins Involved in Stress Defense and ROS Scavenging

   ROS scavenging or cell detoxification constitutes an important defense
   strategy in plants that involves several enzymes and proteins that
   reduce oxidative damage induced by water stress.^[161]54 Many
   biological processes produce ROS signaling intermediates such as
   superoxide radicals, hydrogen peroxide, and hydroxyl radicals,
   including the photorespiratory pathway.^[162]55 External environmental
   stimuli such as drought stress could induce cells to produce a large
   amount of ROS, and excessive ROS production may lead to oxidative
   stress and result in cell component and cell structure damage, cell
   function loss, and ultimately programmed cell death (PCD) or
   necrosis.^[163]56,[164]57 Therefore, balancing oxidant and antioxidant
   intracellular systems is critical for higher plants or cyanobacteria to
   adapt to diverse growth conditions through the synthesis of a large
   number of responsive and defensive proteins to protect cells from
   damage.^[165]58 In this study, we observed that some proteins related
   to stress defense, osmotic protection, and ROS scavenging were
   phosphorylated with decreased levels of phosphorylation together with
   decreased activities of these proteins following severe dehydration
   stress in N. flagelliforme, including superoxide dismutase, peroxidase,
   and glutathione-S-transferase, except for catalase and glutathione
   reductase ([166]Table S2 and [167]Figures [168]5
   , [169]7). This indicates that a decrease in their phosphorylation
   levels together with a decrease in their activity is likely not
   sufficient to eliminate ROS, which causes a significant increase in the
   superoxide anion and H[2]O[2] level (P < 0.01) ([170]Figure [171]5G–H).
   This might reflect a stressful response mechanism through
   phosphorylation modification to coordinate between oxidant and
   antioxidant intracellular pathways in response to dehydration in N.
   flagelliforme. Interestingly, this coordination between phosphorylation
   and enzymatic activities seems also related to protein expression
   levels, as observed in a previous report that contents of these
   antioxidant enzymes were decreased along with metabolic and
   photosynthetic proteins when N. flagelliforme was exposed to
   dehydration stress;^[172]6 one possibility is that these phosphorylated
   proteins may be replaced to suffer damage during the desiccation
   process. Therefore, these proteins might be impaired and quickly
   degraded as a result of hydrolyzation during the dehydration process,
   but more studies are needed to further explore the potential mechanism
   during this process ([173]Figure [174]7).

3.5. Possible Model of Phosphorylation Response in N. flagelliforme on
Dehydration

   Based on the results of present and past studies,^[175]5,[176]6 a
   proposed key responsive model to dehydration stress in N. flagelliforme
   was proposed ([177]Figure [178]7
   ). When cells are subjected to dehydration, it is first supposed to
   perceive the external stimulus from the membrane and cell wall through
   a specific transporter on the plasma membrane, the signal molecules
   will, in turn, accumulate in the cytoplasm of the cells, some signal
   molecules are transported into the two-component signaling system, and
   the signal molecule then activates its corresponding protein kinase(s),
   such as histidine kinase (HistK).^[179]59 Subsequently, activated
   protein kinase hence phosphorylates its downstream transcription
   factor(s) to facilitate their binding to specific genes and then induce
   the production of stress-related proteins; moderate dehydration stress
   (30% water loss) inhibited phosphorylation levels of proteins such as
   PGK, while this stress event stimulated the gene expression of PGK;
   however, it is under extremely severe dehydration stress (100% water
   loss), some stress-related proteins are impaired and consequently
   degraded through phosphorylated modification by specific protein
   kinase(s) especially in ROS scavenging pathway. This suggests that
   PGK1, acting a protein kinase is involved in response to dehydration
   stress probably due to the fact that this phosphorylation can inhibit
   mitochondrial pyruvate metabolism and ROS production in some reports
   for instance Li et al.^[180]60 Interestingly, some stress-related
   proteins are impaired and consequently degraded through phosphorylated
   modification by specific protein kinase(s), especially in the ROS
   scavenging pathway. Importantly, the phosphorylation modification by
   kinase signaling transduction participates in various biological
   pathways, such as ROS scavenging, Calvin cycle, sucrose metabolism, and
   light reaction to regulate energy supply and osmotic balance to adapt
   to moderate dehydration stress in N. flagelliforme.

4. Conclusions

   This study presents a systematic survey on phosphoproteome in
   terrestrial cyanobacteria, N. flagelliforme, in response to dehydration
   processes and provides a comprehensive infrastructure of
   phosphoproteins for understanding the mechanistic aspect of drought
   tolerance of terrestrial cyanobacteria related to protein
   phosphorylation. Here, a total of 76 out of 271 identified
   phosphoproteins were found with significantly differential expression
   for dehydration treatment compared to control. These phosphoproteins
   are involved in different dehydrated-induced signaling processes, such
   as photosynthesis, sucrose metabolism, the two-component signaling
   system, the ROS scavenging system, and other aspects. We concluded that
   protein phosphorylation is a key post-translational modification
   regulating protein function in almost all cellular processes and an
   important regulatory mechanism for coordinating enzyme activity related
   to the antioxidant system in N. flagelliforme to adapt to dehydration
   stress. The results obtained here would not only extend current
   knowledge related to protein phosphorylation of terrestrial
   cyanobacteria in response to dehydration but also can serve as a
   sourcing hub for future studies of evolution and ecological adaptation
   mechanisms of terrestrial cyanobacteria in the xeric environment from
   the perspective of protein phosphorylation.

5. Experimental Section

5.1. Sample Treatment

   N. flagelliforme strain was collected from the Helan Mountain in the
   east region of Ningxia, China. Fresh samples were collected and
   cultured in BG11 medium as reported earlier,^[181]15 in which we added
   sodium hydroxide for the alkalinity of their natural habitats (pH 8.5).
   Samples were cultured in an incubator at 25 °C with continuous
   illumination at 40 μmol photon m^–2·s^–1 for 6–7 h, which was long
   enough for the photosynthetic activity to fully recover.^[182]16

   Each sample was spread on a plastic net with a biomass density of
   approx. 5 mg·cm^–2, 0.5 g dry weight spread equably forming square area
   ∼100 cm^2. Water drops on the sample were removed with filter paper
   before the initial wet weight was determined. The algal mats were
   exposed to air at 25 °C to ensure fast water loss and were sampled at
   intervals of 1 h to assess water loss of 0, 30, 75, and 100%. Water
   loss (WL, %) = (Ww – Wt)/(Ww – Wd) × 100%, where Ww is the initial wet
   weight, Wt is the instantaneous weight of samples measured at certain
   intervals, and Wd is the dry weight.^[183]16

5.2. Protein Extraction

   The extraction of total proteins was performed with minor modifications
   by following Liang et al.^[184]6 through removing 2% Pharmalyte 3–10
   from sample buffer. Protein concentrations were determined based on the
   Bradford assay. The protein solution was used immediately for further
   experiments.

5.3. Protein Digestion and iTRAQ Labeling

   Protein digestion was performed according to the FASP procedure as
   described previously,^[185]17 and the resulting peptide mixture was
   labeled using the 8-plex iTRAQ reagent according to the manufacturer’s
   instructions (Applied Biosystems). Briefly, ∼200 μg of proteins for
   each sample was incorporated into 30 μL of STD buffer, including 4%
   sodium dodecyl sulfate (SDS), 100 mM dithiothreitol (DTT), and 150 mM
   Tris–HCl (pH 8.0). The detergent, DTT, and other low-molecular-weight
   components were removed using UA buffer (8 M urea, 150 mM Tris–HCl pH
   8.0) by repeated ultrafiltration three times (Microcon units, 30 kD).
   Then, 100 μL of 0.05 M iodoacetamide in UA buffer was added to block
   reduced cysteine residues and the samples were then incubated for 20
   min in darkness. The filters were washed with 100 μL of UA buffer three
   times and then washed with 100 μL of DS buffer (50 mM triethylammonium
   bicarbonate at pH 8.5) twice. Finally, the protein suspensions were
   digested with 2 μg of trypsin (Promega) in 40 μL of DS buffer overnight
   at 37 °C, and the resulting peptides were collected as a filtrate. The
   peptide content was estimated by UV light spectral density at 280 nm
   using an extinction coefficient of 1.1 of 0.1% (g·L^–1) solution that
   was calculated based on the frequency of tryptophan and tyrosine in
   vertebrate proteins. For peptide labeling, each iTRAQ reagent was
   dissolved in 70 μL of ethanol and added to the respective peptide
   mixture. The samples were labeled, multiplexed, and vacuum-dried.

5.4. Enrichment of Phosphorylated Peptides by the TiO[2] Beads

   The labeled peptides were then mixed, concentrated by a vacuum
   concentrator, and resuspended in 500 μL of loading buffer (2% glutamic
   acid/65% ACN/2% trifluoroacetyl (TFA)). Then, TiO[2] beads were added
   and agitated for 40 min. Centrifugation was carried out for 1 min at
   5000g, producing the first beads. The supernatant from the first
   centrifugation was mixed with another TiO[2] bead, resulting in the
   second beads, which were collected as before. Both beads were combined
   and washed with 50 μL of washing buffer I (30% ACN/3% TFA) three times,
   and then, 50 μL of washing buffer II (80% ACN/0.3% TFA) for another
   three times to remove the remaining nonabsorbed material. Finally,
   phosphopeptides were eluted with 50 μL of elution buffer (40% ACN/15%
   NH[4]OH),^[186]18 followed by lyophilization and LC–MS/MS analysis.

5.5. LC–MS/MS Analysis

   The phosphopeptide solution (5 μL) mixed with 15 μL of 0.1% (v/v)
   trifluoroacetic acid and then 10 μL of solution mixture was injected
   for nano-LC–MS/MS analysis using a Q Exactive MS (Thermo Scientific)
   equipped with an Easy nLC (Proxeon Biosystems, now Thermo Scientific).
   The peptide mixture was loaded onto a C[18] reversed-phase column (15
   cm long, 75 μm inner diameter, RP-C18 3 μm, packed in-house) in buffer
   A (0.1% formic acid) and separated with a linear gradient of buffer B
   (80% acetonitrile and 0.1% formic acid) at a flow rate of 250 nL·min^–1
   controlled by IntelliFlow technology over 240 min. The peptides were
   eluted with a gradient of 0–60% buffer B from 0 to 200 min, 60 to 100%
   buffer B from 200 to 216 min, and 100% buffer B from 216 to 240 min.

   For MS analysis, peptides were analyzed in positive ion mode. MS
   spectra were acquired using a data-dependent top10 method dynamically
   choosing the most abundant precursor ions from the survey scan
   (300–1800 m/z) for higher-energy collisional dissociation (HCD)
   fragmentation. Determination of the target value is based on predictive
   Automatic Gain Control (pAGC). Dynamic exclusion duration was 40.0 s.
   Survey scans were acquired at a resolution of 70 000 at m/z 200, and
   the resolution for HCD spectra was set to 17 500 at m/z 200. The
   normalized collision energy was 27 eV, and the underfill ratio, which
   specifies the minimum percentage of the target value likely to be
   reached at maximum fill time, was defined as 0.1%. The instrument was
   run with peptide recognition mode enabled.

5.6. Data Processing

   MS/MS spectra were searched using Mascot 2.2 (Matrix Science, London,
   U.K.) embedded in Proteome Discoverer 1.4 against Cyanobase
   ([187]http://genome.kazusa.or.jp/cyanobase) containing 3186 protein
   sequences and the decoy database. For protein identification, the
   following options are used: peptide mass tolerance = 20 ppm, MS/MS
   tolerance = 0.1 Da, enzyme = trypsin, missed cleavage = 2, fixed
   modification = carbamidomethyl (C), iTRAQ4-/8-plex (K), iTRAQ4-/8-plex
   (N-term), and variable modification = oxidation (M), phosphorylation
   (S/T/Y). The score threshold for peptide identification was set at 1%
   false discovery rate (FDR), and the PhosphoRS site probability cutoff
   was 0.75.^[188]19

5.7. Bioinformatic Analysis

   To obtain information on cellular function and localization of the
   identified phosphoproteins, which were categorized by biological
   process and molecular function, we used an in-house PERL script
   according to Gene Ontology (GO) terms extracted from the Cyanobase
   database ([189]http://genome.kazusa.or.jp/cyanobase/N. flagelliforme
   CCNUN1).^[190]20 The enrichment of GO categories was analyzed using
   Cytoscape plugin BiNGO with the default parameters. The reference GO
   ontology in Cytoscape ontology format was created using N.
   flagelliforme CCNUN1 GO terms. All identified phosphoproteins were also
   analyzed with the PSORTb program, which is a web-based tool to predict
   bacterial protein subcellular localization. A cutoff of 7.5 or above is
   used to return a final prediction; otherwise, a result of “Unknown” is
   returned. To investigate the motif specificities between eukaryotic and
   prokaryotic kinases, a reliable online searching algorithm of SCANSITE
   ([191]http://scansite.mit.edu) was used to search for the identified
   phosphorylated sites in N. flagelliforme with the default settings.

   The DEPPs, defined as differentially expressed phosphoproteins, were
   determined based on the changes of abundance in each protein that was
   phosphorylated. One-way ANOVA was used to analyze significant
   differences among different dehydration treatments in N. flagelliforme.
   Differentially expressed phosphoproteins (DEPPs) via one-way ANOVA
   analysis were analyzed by P value < 0.05 screening as significant
   levels. Furthermore, the FASTA protein sequences of differentially
   changed phosphorylated proteins were blasted against the online Kyoto
   Encyclopedia of Genes and Genomes (KEGG) database
   ([192]http://geneontology.org/) to retrieve their KOs and subsequently
   mapped to pathways in KEGG.^[193]21 The corresponding KEGG pathways
   were extracted.

   The studied protein relative expression data was used to perform
   hierarchical clustering analysis. For this purpose, Cluster3.0
   ([194]http://bonsai.hgc.jp/~mdehoon/software/cluster/software.htmcluste
   r/software.htm) and Java Treeview software
   ([195]http://jtreeview.sourceforge.net) were used. The Euclidean
   distance algorithm for similarity measure and the average linkage
   clustering algorithm (clustering uses the centroids of the
   observations) for clustering were selected when performing hierarchical
   clustering. Heatmap is often presented as a visual aid in addition to
   the dendrogram.

   In addition, a custom-made PERL script was written to extract the
   prealigned phosphopeptide sequences, and the Motif-X algorithm was used
   to derive significant motifs based on all identified phosphorylation
   sites and their surrounding ±6 residues with the parameters as
   previously described.^[196]22 We also uploaded the identified
   phosphopeptides and phosphoproteins to the motif tool section of the
   PHOSIDA database ([197]www.phosida.com) to extract the specific motifs.

   The protein–protein interaction information of the studied proteins was
   retrieved from the IntAct molecular interaction database
   ([198]http://www.ebi.ac.uk/intact/) by their gene symbols or STRING
   software ([199]http://string-db.org/). The results were downloaded in
   the XGMML format and imported into Cytoscape software
   ([200]http://www.cytoscape.org/, version 3.2.1) to visualize and
   further analyze functional protein–protein interaction networks.
   Furthermore, the degree of each protein was calculated to evaluate the
   importance of the protein in the PPI network.

5.8. Measurement of Physiological and Biochemical Parameters

   The activities of enzymes involved in the ROS signaling pathway were
   determined, including superoxide dismutase (SOD) following the
   p-nitro-blue tetrazolium chloride (NBT) photoreduction method,^[201]23
   peroxidase (POD) according to the method described previously,^[202]24
   catalase (CAT) from the method described by Aebi,^[203]25 and
   glutathione-S-transferase (GST) and glutathione reductase (GR)
   according to the method by Bender.^[204]26 In addition, the contents of
   glucose, fructose, glycogen, oxygen free radicals (OFRs), and H[2]O[2]
   were also detected according to the instructions given in the kit of
   Suzhou Comin Biotechnology Co. Lid (Suzhou, China). Protein
   concentration was determined as previously described by
   Bradford.^[205]27

5.9. qRT-PCR Analysis

   The total RNA was extracted using the TIANGEN RNAprep Pure Plant Kit
   (Polysaccharides&Polyphenolics-rich) according to the supplier’s
   instructions.^[206]6 Total RNA (1 μg) was used for cDNA synthesis using
   the PrimeScript RT Reagent Kit with gDNA Eraser (TakaRa) according to
   the manufacturer’s instructions. qRT-PCR analysis was performed using
   16s rRNA expression as an internal standard to normalize the amount of
   cDNA. The detected genes are as follows: glyceraldehyde-3-phosphate
   dehydrogenase (GAPDH), 6-phosphoglueonate dehydrogenase (6PGDH),
   glucose6-phosphate dehydrogenase (G6PDH), phosphoribulosekinase (PRK),
   phosphoglycerate kinase (PGK), and fructose-1,6-bisphosphatase (FBP).
   The primer sequences of detected genes and PCR conditions are shown in
   [207]Table S1. Transcript levels of genes were quantified by qRT-PCR
   using the CFX96 Touch detection system. Amplifications were performed
   using UltraSYBR mixture (Shanghai Sangon Biotech Co., Ltd.). All
   qRT-PCR expression assays were performed and analyzed at least three
   times in independent experiments. The relative mRNA level (normalized
   to the level of 16s RNA gene) of each specific transcript was
   determined with Bio-Rad software and calculated using the 2^–ΔΔCT
   method.^[208]28

5.10. Statistical Analysis

   All experiments were performed with three independent biological
   replicates. Data from repeated measurements were shown as the mean ±
   SE. A comparison of differences among the groups was carried out by
   using one-way analysis of variance; P ≤ 0.05 was considered
   significant.

Acknowledgments