An Up-To-date Workflow For

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
An up-to-date workflow for plant (phospho)proteomics identifies
differential drought-responsive phosphorylation events in maize leaves
Lam Dai Vu, Elisabeth Stes, Michiel Van Bel, Hilde Nelissen, Davy Maddelein,
Dirk Inz, Frederik Coppens, Lennart Martens, Kris Gevaert, and Ive De Smet
J. Proteome Res., Just Accepted Manuscript DOI: 10.1021/acs.jproteome.6b00348 Publication Date (Web): 19 Sep 2016
Downloaded from http://pubs.acs.org on September 21, 2016
Just Accepted
Just Accepted manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides Just Accepted as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. Just Accepted manuscripts
appear in full in PDF format accompanied by an HTML abstract. Just Accepted manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI). Just Accepted is an optional service offered
to authors. Therefore, the Just Accepted Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the Just
Accepted Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these Just Accepted manuscripts.
Journal of Proteome Research is published by the American Chemical Society. 1155

Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 48 Journal of Proteome Research
1
2
3 An up-to-date workflow for plant (phospho)proteomics identifies differential drought-
4
5 responsive phosphorylation events in maize leaves
6
7
8
9
10 Lam Dai Vua,b,c,d,#, Elisabeth Stesa,b,c,d,#, Michiel Van Bela,b, Hilde Nelissena,b, Davy
11
12 Maddeleinc,d, Dirk Inza,b, Frederik Coppensa,b, Lennart Martensc,d, Kris Gevaertc,d,, Ive De
13
14 Smeta,b,,*
15
16
17
18 a
Department of Plant Systems Biology, VIB, 9052 Ghent, Belgium
19
20 b
21 Department of Plant Biotechnology and Bioinformatics, Ghent University, 9052 Ghent,
22
23 Belgium
24
25 c
Medical Biotechnology Center, VIB, 9000 Ghent, Belgium
26
27 d
Department of Biochemistry, Ghent University, 9000 Ghent, Belgium
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58 1
59
60
ACS Paragon Plus Environment
Journal of Proteome Research Page 2 of 48
1
2
3 ABSTRACT
4
5
6
7 Protein phosphorylation is one of the most common post-translational modifications (PTMs),
8
9
10 which can regulate protein activity and localization, as well as proteinprotein interactions in
11
12 numerous cellular processes. Phosphopeptide enrichment techniques enabled plant researchers
13
14 to acquire insight in phosphorylation-controlled signaling networks in various plant species.
15
16 Most phosphoproteome analyses of plant samples still involve stable isotope labeling, peptide
17
18
fractionation, and demand lots of mass spectrometry (MS) time. Here, we present a simple
19
20
21 workflow to probe, map and catalogue plant phosphoproteomes, requiring relatively low
22
23 amounts of starting material, no labeling, no fractionation, and no excessive analysis time.
24
25 Following optimization of the different experimental steps on Arabidopsis thaliana samples,
26
27 we transferred our workflow to maize, a major monocot crop, to study signaling upon drought
28
29
30 stress. In addition, we included normalization to protein abundance to identify true
31
32 phosphorylation changes. Overall, we identified a set of new phosphosites in both
33
34 Arabidopsis thaliana and maize, some of which are differentially phosphorylated upon
35
36 drought. All data are available via ProteomeXchange with identifier PXD003634, but to
37
38
provide easy access to our model plant and crop datasets, we created an online database, Plant
39
40
41 PTM Viewer (bioinformatics.psb.ugent.be/webtools/ptm_viewer/), where all phosphosites
42
43 identified in our study can be consulted.
44
45
46
47 Key Words: Phosphoproteomics, maize, Arabidopsis, drought stress, database
48
49
50
51
52
53
54
55
56
57
58 2
59
60
1
2
3 INTRODUCTION
4
5
6
7 The balanced action of protein kinases and phosphatases determines a proteomes
8
9
10 phosphorylation status. Protein phosphorylation may transiently modify protein properties,
11
12 such as enzymatic activity, subcellular localization, protein structure and stability, and
13
14 interactions with other proteins. As such, many cellular signaling processes, such as
15
16 transmembrane signaling, intracellular amplification of signals and cell cycle control, occur
17
18
via reversible protein phosphorylation.1 In plants, phosphorylation-mediated signaling is of
19
20
21 central importance in various physiological processes, including hormone signaling and stress
22
23 responses.2 However, only a limited number of plant kinases and phosphatases (and their
24
25 targets) have been studied in different levels of detail.3
26
27 Mass spectrometry (MS)-based proteomics became an essential tool for studying
28
29
30 protein phosphorylation and has enabled the identification of numerous phosphorylation sites
31
32 on plant proteins.3 Nevertheless, studies of phosphorylation events remain challenging, due to
33
34 their dynamic nature and the sub-stoichiometric levels of phosphorylated proteins. Therefore,
35
36 at least some level of enrichment for phosphorylation sites is needed and this is best done at
37
38
the peptide level to maximize the identification of phosphosites. The most productive
39
40
41 approach is based on metal (ion) chelation. By exploiting the interaction between negatively
42
43 charged phosphate groups and positively charged metal ions or metal oxides, immobilized
44
45 metal affinity chromatography (IMAC) and metal oxide affinity chromatography (MOAC)
46
47 methods, respectively, represent efficient ways to enrich phosphopeptides from complex
48
49
50 mixtures. Enrichment with TiO2 beads became a routine method in plant proteomics studies in
51
52 recent years.4-12 In an attempt to maximally cover phosphoproteomes, the majority of
53
54 phosphoproteomics approaches make use of peptide fractionation methods, such as strong
55
56 cation exchange chromatography, hydrophilic interaction chromatography or reversed-phase
57
58 3
59
60
1
2
3 chromatography.7, 13-17 These however result in far more LC-MS/MS measurement time per
4
5 sample to be analyzed and also require large(r) amounts of starting material.
6
7 With the ever-increasing number of phosphorylation sites being identified for nearly
8
9
10 every human cellular protein a phosphosite has been reported18 the functionality of these
11
12 post-translational modifications (PTMs) is questioned. Crowdedness effects are hypothesized
13
14 to give rise to non-functional transfer of a phosphate group by kinases upon encounter of a
15
16 random protein.19, 20 Merely profiling phosphorylation sites will hence likely lead to the large
17
18
scale identification of nonfunctional PTMs. To discriminate these noisy phosphosites from
19
20
21 sites with regulatory significance, it is vital to design experiments where differential
22
23 conditions are compared. Obviously, this requires assessing dynamics in the
24
25 phosphoproteome via quantitative methods. Methodologies for the quantitative analysis of
26
27 phosphoproteomes in plants are most frequently based on stable isotope labeling, like 15N/14N
28
29
30 metabolic labeling of proteins during plant growth4, 11, 16, 21, 22 or post-metabolic labeling of
31
32 peptides with iTRAQ.9, 12, 23, 24 As labeling imposes limitations on the number of conditions
33
34 that can be monitored, label-free methods represent a practical alternative. Two label-free
35
36 methods, spectral counting and precursor ion intensity-based quantification, have been
37
38
applied in plant phosphoproteome strategies.8, 10, 11, 14, 25-27
However, label-free approaches
39
40
41 often suffer from quantitative incompleteness due to stochastic data acquisition (MS/MS
42
43 sequencing) leading to numerous missing values in the dataset, which to some extent can
44
45 be avoided by matching data between LC-MS(/MS) runs.
46
47 Although missing in most published plant phosphoproteome studies [notwithstanding
48
49
50 some exceptions12, 15, 16, 28-30], parallel and in depth investigation of the overall proteome is
51
52 recommended for normalization of quantitative PTM studies. To determine if phosphopeptide
53
54 changes are the result of true phosphorylation changes or rather general abundance changes of
55
56 the phosphoprotein, phosphopeptide levels need to be normalized to overall protein
57
58 4
59
60
1
2
3 abundances. Ideally, such changes in overall protein levels should be derived from an analysis
4
5 of non-phosphopeptides of the same sample.31
6
7 In recent years, the number of phosphoproteome and proteome researches in monocot
8
9
10 crop species under abiotic stresses has been steadily growing.32, 33 Agricultural plants, such as
11
12 maize, routinely face drought stress, which is one of the worst environmental hazards that
13
14 impacts crop productivity.34, 35
As plants remodel their proteome in response to stress,
15
16 drought-adaptive traits are likely to be reflected at the proteome level.36 Moreover, as a
17
18
universal biochemical signal in cells, protein phosphorylation controls stress responses,
19
20
21 transmitting stress signals from the cell surface to the nucleus.32 However, parallel
22
23 phosphoproteome and proteome analysis, while existing,15, 29 are not yet standardized for such
24
25 monocot crop and model plant studies and often the change in phospho-status is not corrected
26
27 against the protein abundance.
28
29
30 Taken together, commonly used strategies in plant phosphoproteomics involve tedious
31
32 labeling approaches and fractionation steps, which are time consuming, expertise demanding
33
34 and negatively affect reproducibility, robustness and throughput. Here, we present a label-free
35
36 quantitative workflow for quick and reproducible phosphoproteome analysis of plant tissue,
37
38
requiring only small sample amounts and no costly expert software for data analysis and
39
40
41 integrating steps (such as normalization) that are not yet standard in plant phosphoproteomics.
42
43 We applied our workflow to maize, a major monocot crop, to study signaling upon drought
44
45 stress. In this case study we identified a set of new phosphosites in maize, some of which are
46
47 differentially phosphorylated upon drought.
48
49
50
51
52 MATERIALS AND METHODS
53
54
55
56
57
58 5
59
60
1
2
3 Plant growth
4
5 Seedlings of Arabidopsis thaliana (ecotype Columbia) were grown on vertically-held plates
6
7 with half-strength Murashige and Skoog medium solidified with 0.8% agar at 22C in
8
9
10 continuous light. Four days post germination (dpg), the plants were transferred to 10 M 1-
11
12 naphthaleneacetic acid (NAA)-containing plates. Roots were harvested 5 dpg. Maize plants
13
14 (inbred line B104) were grown in soil in a growth chamber with controlled relative humidity
15
16 (55%) and temperature (24C), in a 16h/8h (day/night) cycle. Drought was induced by
17
18
lowering the soil water capacity to 62.5% relative to that of the well-watered control plants.
19
20
21 Twenty-one days after sowing, the first 4 cm of growing leaf 7 was harvested.
22
23
24
25 Protein Extraction and Tryptic Digestion
26
27 Plant material was harvested in three biological replicates. One g of fresh weight material was
28
29
30 flash-frozen in liquid nitrogen, and manually ground into a fine powder with a pestle and
31
32 mortar. Proteins were extracted in homogenization buffer containing 50 mM Tris-HCl buffer
33
34 (pH 8), 0.1 M KCl, 30% sucrose, 5 mM EDTA, and 1 mM DTT in milliQ water, and the
35
36 appropriate amounts of the Complete protease inhibitor mixture and the PhosSTOP
37
38
phosphatase inhibitor mixture (both from Roche) were added. The samples were sonicated on
39
40
41 ice and centrifuged at 4C for 15 min at 2,500g to remove debris. Supernatants were
42
43 collected and a methanol/chloroform precipitation was carried out by adding 3, 1 and 4
44
45 volumes of methanol, chloroform and water, respectively. Samples were centrifuged for
46
47 10 min at 5,000g, and the aqueous phase was removed. After addition of 4 volumes
48
49
50 methanol, the proteins were pelleted via centrifugation for 10 min at 2,500g. Pellets were
51
52 washed with 80% acetone and re-suspended in 6 M guanidinium hydrochloride in 50 mM
53
54 triethylammonium bicarbonate (TEAB) buffer (pH 8). Alkylation of cysteines was carried out
55
56 by adding a combination of tris(carboxyethyl)phosphine (TCEP, Pierce) and iodoacetamide
57
58 6
59
60
1
2
3 (Sigma-Aldrich) to final concentrations of 15 mM and 30 mM respectively, and the reaction
4
5 was allowed for 15 min at 30C in the dark. Before digestion, the samples were buffer
6
7 exchanged on Illustra NAP columns (GE Healthcare Life Sciences) to 50 mM TEAB buffer
8
9
10 (pH 8) and the protein concentration was measured using the Bio-Rad Protein Assay. One mg
11
12 of the proteins was pre-digested with EndoLysC (Wako Chemicals) for 4 h, followed by a
13
14 digestion with trypsin overnight (Promega Trypsin Gold, mass spectrometry grade), both
15
16 digestions occurring at 37C at an enzyme-to-substrate ratio of 1:100 (w:w). The digest was
17
18
acidified to pH 3 with trifluoroacetic acid (TFA) and desalted with SampliQ C18 SPE
19
20
21 cartridges (Agilent) according to the manufacturers guidelines. The eluates were split into
22
23 two and dried in a vacuum centrifuge. One half of the samples served for proteome analyses
24
25 and were re-dissolved in 30 L of 2% (v/v) acetonitrile and 0.1% (v/v) TFA right before LC-
26
27 MS/MS analysis.
28
29
30
31
32 Phosphopeptide Enrichment
33
34 The dried eluates were re-suspended in 100 l of loading solvent (80% acetonitrile, 5% TFA)
35
36 and incubated with 1 mg MagReSyn Ti-IMAC microspheres (ReSyn Biosciences) for 20
37
38
min at room temperature. The microspheres were next washed once with wash solvent 1 (80%
39
40
41 acetonitrile, 1% TFA, 200 mM NaCl) and two times with wash solvent 2 (80% acetonitrile,
42
43 1% TFA). The bound phosphopeptides were eluted with three volumes (80 l) of a 1%
44
45 NH4OH solution, followed immediately by acidification to pH 3 with formic acid. Prior to
46
47 MS analysis, the samples were vacuum-dried and re-dissolved in 50 L of 2% (v/v)
48
49
50 acetonitrile and 0.1% (v/v) TFA.
51
52
53
54
55
56
57
58 7
59
60
1
2
3 Mass Spectrometry
4
5 Each sample was analyzed twice (i.e. in two technical replicates) via LCMS/MS on an
6
7 Ultimate 3000 RSLC nano LC (Thermo Fisher Scientific) in-line connected to a Q Exactive
8
9
10 mass spectrometer (Thermo Fisher Scientific). The sample mixture was first loaded on a
11
12 trapping column (made in-house, 100 m internal diameter (I.D.) 20 mm, 5 m beads C18
13
14 Reprosil-HD, Dr. Maisch, Ammerbuch-Entringen, Germany). After flushing from the
15
16 trapping column, the sample was loaded on an analytical column (made in-house, 75 m I.D.
17
18
150 mm, 3 m beads C18 Reprosil-HD, Dr. Maisch). Peptides were loaded with loading
19
20
21 solvent A (0.1% TFA in water) and separated with a linear gradient from 98% solvent A
22
23 (0.1% formic acid in water) to 55% solvent B (0.1% formic acid in water/acetonitrile, 20/80
24
25 (v/v)) in 170 min at a flow rate of 300 nL/min. This was followed by a 5 min wash reaching
26
27 99% solvent B. The mass spectrometer was operated in data-dependent, positive ionization
28
29
30 mode, automatically switching between MS and MS/MS acquisition for the 10 most abundant
31
32 peaks in a given MS spectrum. The source voltage was 3.4 kV, and the capillary temperature
33
34 was 275C. One MS1 scan (m/z 4002000, AGC target 3 106 ions, maximum ion injection
35
36 time 80 ms) acquired at a resolution of 70000 (at 200 m/z) was followed by up to 10 tandem
37
38
MS scans (resolution 17500 at 200 m/z) of the most intense ions fulfilling predefined
39
40
41 selection criteria (AGC target 5 104 ions, maximum ion injection time 60 ms, isolation
42
43 window 2 Da, fixed first mass 140 m/z, spectrum data type: centroid, underfill ratio 2%,
44
45 intensity threshold 1.7xE4, exclusion of unassigned, 1, 5-8, >8 charged precursors, peptide
46
47 match preferred, exclude isotopes on, dynamic exclusion time 20 s). The HCD collision
48
49
50 energy was set to 25% Normalized Collision Energy and the polydimethylcyclosiloxane
51
52 background ion at 445.120025 Da was used for internal calibration (lock mass).
53
54
55
56
57
58 8
59
60
1
2
3 Data Analysis
4
5 For the Arabidopsis samples, MS/MS spectra were searched against a UniProt database
6
7 containing A. thaliana sequences (34,509 entries, version November, 2014) with the
8
9
10 MaxQuant software (version 1.5.3.8). For the maize samples, the searches were done against
11
12 a Zea mays database downloaded from PLAZA Monocots 3.037 containing sequences (39,305
13
14 entries, version 2014) with the MaxQuant software (version 1.5.0.30). For all searches, a
15
16 precursor mass tolerance was set to 20 ppm for the first search (used for nonlinear mass re-
17
18
calibration) and to 4.5 ppm for the main search. Trypsin was selected as enzyme setting.
19
20
21 Cleavages between lysine/arginine-proline residues and up to two missed cleavages were
22
23 allowed. Carbamidomethylation of cysteine residues was selected as a fixed modification and
24
25 oxidation of methionine residues was selected as a variable modification. For the samples
26
27 enriched for phosphopeptides phosphorylation of serine, threonine and tyrosine residues were
28
29
30 set as additional variable modifications. The false discovery rate for peptide and protein
31
32 identifications was set to 1%, and the minimum peptide length was set to 7. The minimum
33
34 score threshold for both modified and unmodified peptides was set to 30. The MaxLFQ
35
36 algorithm allowing label-free quantification38 and the Matching Between Runs feature were
37
38
enabled. All mass spectrometry proteomics data have been deposited to the ProteomeXchange
39
40
41 Consortium via the PRIDE39 partner repository with the dataset identifier PXD003634. For
42
43 the quantitative maize proteome and phosphoproteome analyses, the ProteinGroups and
44
45 Phospho(STY)sites output files, respectively, generated by the MaxQuant search was loaded
46
47 into Perseus, the data analysis software available in the MaxQuant package.40 Only proteins or
48
49
50 phosphosites which were quantified in at least two of the three biological replicates of at least
51
52 one sample were retained. Log2 transformed protein LFQ intensities or phosphosites
53
54 intensities were centered by subtracting the median of the entire set of protein/phosphosite
55
56 intensities per sample. A two-sample test with p<0.05 was carried out to test the differences
57
58 9
59
60
1
2
3 between groups. The statistically significant hits were then Z-scored and clustered into groups
4
5 by a hierarchical clustering analysis based on Euclidean distance.
6
7 To identify novel phosphosites (not previously reported ones), we compared our data
8
9
10 to the PhosPhAt 4.0 full dataset of experimentally identified phosphosites41 (data from
11
12 04.04.2016) and to the retrieved database of phosphosites identified in maize seed and leaf
13
14 tissues from the Maize Protein Atlas17, 42 (data from 04.04.2016).
15
16
17
18
19
Normalization of phosphoproteome data
20
21 We accounted for protein expression changes to allow proper interpretation of the maize
22
23 quantitative phosphoproteomics data. After log2 transformation and centralization, the
24
25 intensities of the phosphosites were normalized by subtracting the log2 transformed and
26
27 centralized LFQ intensities of the corresponding proteins. The latter dataset of protein LFQ
28
29
30 intensities resulted from the parallel protein expression study of all maize samples.
31
32
33
34 Gene Ontology Analysis
35
36 GO enrichment analysis was performed in the PLAZA 3.0 workbench37. For the maize
37
38
proteome dataset, 234 proteins with significant changes in abundance were analyzed, using
39
40
41 the dataset of 2299 identified proteins as background model. A FDR cutoff 0.04 was used to
42
43 score significantly overrepresented or depleted GO terms (Supplementary Table S9). For the
44
45 maize phosphoproteome dataset, the identified proteins in each conditions were used for the
46
47 enrichment, and the whole theoretical proteome (based on the genome annotation of Z. mays)
48
49
50 was used at background. P-value cutoff was set at 0.05 and only terms enriched in either
51
52 condition were presented (Supplementary Table S15).
53
54
55
56 Motif-X analysis
57
58 10
59
60
1
2
3 The Motif-X algorithm43 was used to extract significantly enriched amino acid motifs
4
5 surrounding the identified phosphosites. The sequence window was limited to 13 amino acids
6
7 and foreground peptides were pre-aligned with the phosphosite centered. Zea mays proteome
8
9
10 data set from PLAZA was used as the background database. The occurrence threshold was set
11
12 at the minimum of 20 peptides and the P-value threshold was set at < 10-6.
13
14
15
16 STRING analysis of protein-protein interaction networks
17
18
Protein-protein interactions were analyzed by STRING (http://string-db.org/),44 using the
19
20
21 sequences of differentially phosphorylated proteins and proteins with significant abundance
22
23 changes as input. The required confidence score was set as > 0.700 for highly confident
24
25 interactions (STRING protein-protein interaction prediction is based on data available for
26
27 genomic homology, gene fusion, occurrence in the same metabolic pathways, co-expression,
28
29
30 experiments, database and text mining. A combined score is calculated based on the score of
31
32 all the methods that were used for the protein-protein interaction prediction. The higher the
33
34 score is, the more confident the interaction). The results were visualized using the Cytoscape
35
36 package.
37
38
39
40
41 PubMed search
42
43 A PubMed search (www.ncbi.nlm.nih.gov/pubmed/) was performed on 21-22/06/2016 using
44
45 maize proteomics 2016, wheat proteomics 2016 or proteomics arabidopsis 2016 to
46
47 identify relevant papers [only research papers (no reviews or opinions) with the correct focus
48
49
50 (namely those that actually investigated the indicated species, that were available, and that
51
52 actually applied proteomics) were retained].
53
54
55
56
57
58 11
59
60
1
2
3 RESULTS AND DISCUSSION
4
5
6
7 Optimized quantitative workflow for proteomics and phosphoproteomics in plants
8
9
10
11
12 To facilitate efficient proteome analyses of plants, we developed a simple workflow, which
13
14 maximizes the coverage and reproducibility of protein and phosphorylation site quantification
15
16 in single LC-MS/MS runs, thus without requiring peptide fractionation steps (Figure 1). To
17
18
optimize the pipeline, we used the fully sequenced and well-annotated model plant
19
20
21 Arabidopsis thaliana.
22
23 First, we provide a brief overview of the key steps in the protocol and the
24
25 improvements that were introduced step-by-step to robustly survey plant proteomes and
26
27 phosphoproteomes (more details can be found in the Materials and Methods). To reproducibly
28
29
30 capture a comprehensive spectrum of proteins, we opted for a protein precipitation approach.
31
32 Proteins from ground plant material were extracted with a sucrose buffer containing protease
33
34 and phosphatase inhibitors. The extract was subsequently purified through a
35
36 chloroform/methanol precipitation step to eliminate compounds (e.g. polysaccharides,
37
38
phenolic compounds, lipids and secondary metabolites) that may hinder preparation and
39
40
41 analysis of proteome samples.45 The pelleted proteins were reconstituted in a buffer
42
43 containing guanidinium hydrochloride. Cysteine disulfide bonds were reduced with tris(2-
44
45 carboxyethyl)phosphine hydrochloride (TCEP-HCl), allowing the alkylation reaction with
46
47 iodoacetamide to simultaneously take place.46 Next, we pre-digested the proteins with
48
49
50 endoproteinase-LysC, followed by a full digestion with trypsin. The pre-digestion step was
51
52 previously shown to substantially improve the proteolytic efficiency of trypsin.47 The
53
54 resulting peptides were desalted, and split into two. One part was used for the proteome
55
56 analysis, leaving 500 g of digest material as input for phosphopeptide enrichment. We opted
57
58 12
59
60
1
2
3 for a Ti4+-IMAC-based method, as it was found to perform extremely well in terms of
4
5 reproducibility and provides even greater selectivity and sensitivity than the more commonly
6
7 used TiO2 chromatography.48, 49
Both the proteome and phosphoproteome samples were
8
9
10 analyzed by 3 hour gradients on a quadrupole Orbitrap instrument [Q Exactive50].
11
12 Second, peptide identification and quantification are important steps following LC-
13
14 MS/MS analysis. We chose a label-free quantitation approach over a labeling approach, as it
15
16 is cost-effective, does not restrict the numbers of samples that can be compared, and can span
17
18
several orders of magnitude of protein concentrations.38 In most label-free studies of plant
19
20
21 phosphoproteomes, the raw data are analyzed by a combination of expensive expert peptide
22
23 identification software, like Proteome Discoverer or Mascot, and in-house developed
24
25 algorithms to facilitate label free quantification10, 11, 26, 27
(Supplementary Table S1). Here,
26
27 peptide identification was carried out by the freely available and easy-to-use software package
28
29
30 MaxQuant.51 Simultaneously, the label-free quantitation is carried out by MaxQuant, in an ion
31
32 intensity-based manner.38 The missing value issue, due to stochastic peptide sequencing
33
34 inherent to mass spectrometry, was tackled by using the match between runs feature in
35
36 MaxQuant, which can transfer MS/MS identifications between measurements based on a
37
38
peptide retention time correlation approach.38
39
40
41 Taken together, compared to published methods for (phospho)proteomics in plants12,
42
14-16, 23
43 , we reduced the number of sample preparation steps, MS time and data analysis
44
45 complexity, due to the lack of labeling, gel-based steps and pre-fractionation steps and the
46
47 introduction of MaxQuant in our workflow (Figure 1). With respect to the latter, this
48
49
50 facilitates and standardizes data analysis, but does (not yet) seem to be routinely integrated in
51
52 plant proteomics (Supplementary Table S1). Further, for differential studies it is also
53
54 important to correct the phosphosite intensities against the protein abundance to pinpoint
55
56 protein level-independent phosphorylation events (Figure 1).
57
58 13
59
60
1
2
3 Validating the optimized workflow on Arabidopsis thaliana roots
4
5
6
7 The proteome analysis of the non-enriched samples amounted to a total of 18 hours of MS
8
9
10 time, leading to the cumulative identification of 34,216 unique peptides that could be mapped
11
12 on 4,903 protein groups (Supplementary Table S2). The latter can be defined as protein
13
14 entries distinguishable on the basis of identified peptides.52 Via the MaxLFQ algorithm, 4,847
15
16 of those could be quantified in at least one biological replicate and 2,992 in all replicates.
17
18
Seeing that label-free methods are very replicate dependent, reproducibility of the
19
20
21 chromatographic separation must be very high. The data from the replicate experiments
22
23 clearly show highly accurate quantitative reproducibility with an average Pearson correlation
24
25 of 0.978 (Supplementary Figure S1A).
26
27 A common challenge for plant proteomics studies is the difficulty of isolating proteins
28
29
30 from the different subcellular organelles with sufficient efficiency. Membrane proteins
31
32 represent an additional hurdle, as their large size and hydrophobicity render them difficult to
33
34 isolate. To obtain a wide-ranging snapshot of cellular signaling processes it is vital to capture
35
36 proteins from not only the cytosol, but also from membranes and organelles. GO analysis
37
38
shows that the applied protocol extracted proteins from cytosol, nucleus, plasma membrane
39
40
41 and other organelles (Figure 2 and Supplementary Table S3). This evidences that our
42
43 approach is not limited by particular experimental difficulties and recovers proteins from all
44
45 subcellular membranes and organelles.
46
47 Next, we monitored the phosphorylation events in the Arabidopsis samples. The total
48
49
50 of six LC-MS/MS runs of the Ti4+-IMAC enriched samples resulted in the identification of
51
52 1,051 unique phosphopeptides, corresponding to 1,331 phosphosites on 706 protein groups
53
54 (Supplementary Table S4). The vast majority of these sites occurred on serine and threonine
55
56 residues (90.3% and 9.1%, respectively), whereas phosphotyrosines accounted for less than
57
58 14
59
60
1
2
3 1% of the identified sites, which is in agreement with other reports.9, 13
Accurate site
4
5 localization (probability > 0.75) was achieved for 799 of these phosphosites on 552 proteins.
6
7 Of the 1,331 unique phosphosites, we could accurately quantify 1,022 and 711 in at least one
8
9
10 and in all biological replicates, respectively. To evaluate the quality of the experiment, we
11
12 assessed the correlation of all phosphopeptide intensities between the three biological
13
14 replicates. An average Pearsons correlation of 0.818 illustrates the high reproducibility of the
15
16 phosphopeptide enrichment strategy (Supplementary Figure S1B). All the identified
17
18
Arabidopsis phosphosites were used to search against the PhosPhAt 4.0 full dataset of
19
20
21 experimentally identified phosphosites.41 This resulted in 169 phosphosites (13% of the
22
23 dataset) uniquely identified in our study (Supplementary Table S5).
24
25 In summary, we have experimental evidence that our workflow successfully detects a
26
27 large portion of the (phospho)proteome and can thus be applied to understanding biological
28
29
30 processes.
31
32
33
34 Applying the (phospho)proteomics workflow to maize leaves under drought stress
35
36
37
38
Following the validation of our pipeline in Arabidopsis roots, we applied our workflow to a
39
40
41 monocot crop under stress. Given the importance of drought-related research,32, 53-55
we
42
43 profiled the proteome and phosphoproteome of maize leaves subjected to drought stress.
44
45 Since the growth zone of the maize leaf determines to a great extent the final leaf length56 and
46
47 drought affects cell division and cell expansion in the growth zone of the maize leaf,57 we
48
49
50 harvested the growth zone of the growing leaf 7 of 21 day old plants. Drought treatment was
51
52 applied by preventing irrigation upon sowing and when the soil water content reached 62.5%
53
54 of that of the well-watered controls, the plants were maintained at the respective watering
55
56 regime by daily watering. At the moment leaf 7 appeared, the effects of the drought were
57
58 15
59
60
1
2
3 quantified by measuring the final leaf length of the youngest fully grown leaf, leaf 4. This leaf
4
5 showed a significant length reduction compared to the control (data not shown), supporting
6
7 that the applied drought affected leaf growth.
8
9
10 Proteome and phosphoproteome data were obtained for growth zones of maize leaves
11
12 as described above, further emphasizing the importance of moving away from gel-based
13
14 approaches, also in maize and wheat where this is not standard yet (Supplementary Table S1).
15
16 All biological samples were analyzed twice by nanoLC-MS/MS using three hour gradients.
17
18
19
20
21 Changes in abundance of stress regulators under drought revealed through proteome
22
23 analysis
24
25
26
27 In the non-enriched samples, a total of 22,093 peptides were identified originating from 4,409
28
29
30 protein groups. 4,361 protein groups could be accurately quantified, of which 2,299 in at least
31
32 two of the three biological replicates (Supplementary Table S6). The data from the replicate
33
34 experiments show quantitative reproducibility with an average Pearson correlation of 0.856
35
36 and 0.892 for control and drought samples, respectively (Supplementary Figure S2).
37
38
Statistical testing (p<0.05) pinpointed 234 of these proteins to be differently regulated, with
39
40
41 156 up- and 78 down-regulated proteins upon drought stress (Figure 3A and Supplementary
42
43 Table S7-S8).
44
45 The protein that exhibited the largest fold-change is the RuBisCO large subunit (21-
46
47 fold). The abundance of RuBisCO during drought stress has been debated with contrasting
48
49
50 results in different studies.58 While it might be the result of impaired growth during prolonged
51
52 stress treatment rather than an immediate response to drought, here we showed that an
53
54 increase in RuBisCO is a distinct read-out for maize growing under drought conditions.
55
56 Furthermore, amongst the proteins with increased abundance we observed proteins associated
57
58 16
59
60
1
2
3 with water deprivation, like lipoxygenase (GRMZM2G015419),59 keto reductase family 4
4
5 member C9 (GRMZM2G059314),60 fructose-biphosphate aldolase (GRMZM2G057823)61
6
7 and protein-L-isoaspartate methyltransferase (GRMZM2G423027)62. Interestingly, BRI1-
8
9
10 associated receptor kinase (BAK1) (GRMZM2G089819), known for its role in plant defense
11
12 responses and brassinosteroid signaling,63, 64
was found to be down regulated in water
13
14 deprived conditions. The brassinosteroid receptors, BRIs, to which BAK1 binds upon
15
16 brassinosteroid induction, play an important role in maize leaf growth.65 Moreover,
17
18
brassinosteroid signaling has previously been linked to abiotic stress66-68 and brassinosteroid
19
20
21 application is reported to improve drought tolerance in wheat and the resurrection grass
22
23 Sporobolus stapfianus.69
24
25 GO analysis of these proteins showed an overrepresentation of proteins involved in
26
27 carbohydrate metabolism (Figure S4 and Table S9). This result is in agreement with previous
28
29
30 transcriptomic and proteomic studies in other species70-72 that show the abundance of proteins
31
32 involved in carbohydrate metabolism pathways such as glycolysis and pentose phosphate
33
34 pathway to be greatly induced by drought (Figure 4). In contrast, proteins involved in
35
36 chromatin organization such as several histones showed an overall reduced expression, which
37
38
may be a result of remodeling of chromatin, an important epigenetic regulation of stress-
39
40
41 induced genes.73
42
43 Analysis of protein-protein interactions between the 234 proteins with significant
44
45 abundance changes resulted in a network of 141 proteins and 277 interactions, whereas each
46
47 interaction has a combined score of all prediction methods > 0.7 (Figure 4). The network is
48
49
50 approximately centralized around the DNA topoisomerase II (GRMZM2G021270/PLAZA
51
52 identifier ZM05G37510). From the resulted network, we identified different groups of
53
54 interaction between proteins involved in different cellular processes. These included the
55
56 categories DNA/chromatin organization, photosynthesis and glucose metabolism, of which
57
58 17
59
60
1
2
3 the corresponding GO terms were enriched in the dataset. Photosynthesis is very sensitive to
4
5 environmental stresses.74, 75
Unexpectedly, while abiotic stresses are known to reduce
6
7 photosynthesis rate,75 our data exhibited an up-regulation of proteins belonging to
8
9
10 photosynthetic apparatus. This apparent contradiction is in agreement with recent microarray
11
12 data that showed an upregulation in photosynthetic transcripts.76 While in mature leaves, the
13
14 down-regulation or unchanged transcription possibly help maintaining the fully developed
15
16 photosynthetic machinery, in leaves developing under drought, the increase of photosynthetic
17
18
proteins might play a role in resuming the photosynthetic capacity in the later recovery
19
20
21 phase.76
22
23 Further, we identified a small cluster of seven proteins involved in protein folding.
24
25 Four of them, a member of the heat shock protein HSP70 family
26
27 (GRMZM2G415007/PLAZA identifier ZM04G41380) and three chaperone proteins
28
29
30 belonging to the Clp protease family (GRMZM2G110023/PLAZA identifier ZM01G09650;
31
32 GRMZM2G123922/PLAZA identifier ZM10G15640; GRMZM2G162968/ PLAZA identifier
33
34 ZM09G19730). are upregulated upon drought. It is known that the control of protein folding
35
36 state is crucial for the survival of plants during abiotic stress.77 The chaperonin subunit alpha
37
38
60 (AC215201.3_FG005/PLAZA identifier ZM06G23100), involved in the structural
39
40
41 assembly of chloroplasts, was shown to be transcriptionally stimulated by various abiotic
42
43 stresses,78 and is here also induced on the protein level by drought stress. All in all, our data
44
45 greatly overlap with the observations in other studies, including transcriptional findings that
46
47 we can now be confirmed on the protein level.
48
49
50
51
52 Analysis of the maize leaf phosphoproteome under drought stress
53
54
55
56
57
58 18
59
60
1
2
3 Ti4+ IMAC enrichment of maize phosphopeptides led to the detection of a total of 980 unique
4
5 phosphosites on 686 phosphopeptides, which could be mapped on 536 phosphoproteins
6
7 (Supplementary Table S10). The data from the replicate experiments show quantitative
8
9
10 reproducibility with an average Pearson correlation of 0.887 and 0.856 for control and
11
12 drought samples, respectively (Supplementary Figure S3). The number of identified
13
14 phosphoproteins lies in between the range published in recent maize phosphoproteomics
15
16 studies: 282,23 858,9 2,852,14 and 3,557 phosphoproteins17. Important to note is that the latter
17
18
two studies fractionated the enriched phosphopeptides via extensive SCX chromatography,
19
20
21 hereby greatly increasing MS analysis time per sample to two days.14, 17 In our work, six hours
22
23 analysis time was used per sample (2 technical replicates), hence yielding a relative high
24
25 number of phosphoprotein identifications. However, the number of identifications also
26
27 depends on the experimental design, such as the tissue-specificity or the number and the
28
29
30 efficiency of treatment conditions, resulting in the above-mentioned large range of
31
32 identification across the studies.
33
34 Overall, the majority (97.4%) was mono-phosphorylated peptides, while around 2.6%
35
36 of the phosphopeptides carried two phosphorylated residues. There were 84.0%
37
38
phosphoserine, 15.2% phosphothreonine and 0.8% phosphotyrosine containing peptides
39
40
41 identified, sharing a similar distribution pattern to other maize phosphoproteomics studies.9, 14,
42
17, 23
43 All the identified maize phosphosites were searched against the retrieved set of
44
45 phosphosites identified in maize seed from the Atlas of Maize Proteotypes and a dataset of
46
47 phosphosites garnered from different developmental zones of maize leaves.17, 42 This resulted
48
49
50 in 359 phosphosites (37% of the dataset) uniquely identified in our study (Supplementary
51
52 Table S11).
53
54 Overrepresentation of amino acid motifs surrounding the identified phosphosites was
55
56 analyzed using Motif-X (Table 1). Phosphorylated tyrosine sites were excluded from the
57
58 19
59
60
1
2
3 analysis due to their low abundance in the dataset. Similar to other studies in Arabidopsis and
4
5 other monocots,79-82 [sP] is the most enriched motif for the S-phosphorylation as well as its
6
7 phosphorylated threonine counterpart [tP] for the T-phosphorylation dataset. Further, 20
8
9
10 peptides are enriched with the proline-rich motif [sxSP]. Peptides containing the proline-
11
12 directed [sP] and [tP] motifs are suggested to be substrates for MAP-kinases (MAPK),
13
14 sucrose non-fermenting1-related protein kinase 2 (SnRK2), receptor-like kinases (RLKs),
15
16 AGC family protein kinases PKA, PKG and PKC, CDKs (cyclin-dependent kinases),
17
18
calcium-dependent protein kinases (CDPKs) and STE20-like kinases (SLKs).79 Only one
19
20
21 common acidic motif [sDxE] resulted from the analysis, belonging to 22 peptides that
22
23 might be potential substrates for casein kinase II (CKII) and CDPKs. Further, three basophilic
24
25 motifs are overrepresented in the dataset, [Rxxs] and the subtype [RSxs], which are
26
27 recognized by MAPK kinases (MKKs), and [Kxxs], which is targeted by PKA and PKC. No
28
29
30 specific protein kinases are found for the T-phosphorylation motif [tS].
31
32
33
34 Differential analysis of phosphorylation sites after correcting against the protein level
35
36
37
38
Earlier differential phosphoproteomics studies of maize leaf tissue, identifying differences
39
40 9, 15, 23
41 between stress conditions, lack normalization to the protein abundance. Here, because
42
43 of our extensive analysis, we can simultaneously take into account protein and
44
45 phosphorylation site profiles.
46
47 In total, 615 phosphosites on 445 phosphoproteins were quantified, of which 536
48
49
50 phosphosites in at least two biological replicates of one condition. Taking into account that
51
52 differences in protein levels can influence the outcome of the differential phosphorylation
53
54 data, we set out to normalize the intensities of the phosphopeptides to the protein intensities.
55
56 For 224 phosphosites, matching proteins were quantified in the proteome experiment allowing
57
58 20
59
60
1
2
3 normalization. This result demonstrated that phosphopeptide enrichment facilitated the
4
5 identification of low abundance proteins, of which non-phosphorylated peptides were likely
6
7 missed in the proteome scans due to different dynamic ranges and crowdedness. A two
8
9
10 sample test (p<0.05) on the normalized phosphosites intensities showed that 18 of those were
11
12 differentially regulated by drought (Figure 3B, Table 2 and Supplementary Table S12). Some
13
14 of those phosphosites, S470 on HISTONE DEACETYLASE 6 (GRMZM2G457889) (8.3-
15
16 fold down regulated upon drought) and S247 stem-specific protein TSJT1
17
18
(GRMZM2G169671) (2.5-fold up regulated upon drought), are mapped on proteins
19
20
21 previously shown to be regulated during drought signaling.73, 83
In mammalian systems,
22
23 phosphorylation of HISTONE DEACETYLASE 6 was shown to correlate with enzyme
24
25 activity and consequent tubulin deacetylation and microtubule destabilization.84A similar
26
27 mechanism could take place during drought signaling, as plant microtubules are known to
28
29
30 function as sensors for abiotic stress.85
31
32 As protein levels for many phosphosites could not be inferred, because no non-
33
34 phosphorylated peptides of the corresponding proteins were detected, we also subjected the
35
36 not-normalized phosphosites dataset to a two sample test (p<0.05). Based on the
37
38
phosphoproteome data alone, we found 44 phosphosites to be statistically significant between
39
40
41 two conditions, of which 32 were up-regulated upon drought stress and 12 down-regulated
42
43 (Figure 3C and Supplementary Tables S13-14). Interestingly, four microtubule-associated
44
45 proteins were found to be differentially phosphorylated upon drought stress:
46
47 MICROTUBULE-ASSOCIATED PROTEIN 70-2 (GRMZM2G017525), DYNEIN LIGHT
48
49
50 CHAIN 1 (GRMZM2G472231), MICROTUBULE-ASSOCIATED PROTEIN
51
52 (GRMZM2G026309), and KINESIN-LIKE PROTEIN KIN12A (GRMZM2G034828).
53
54 Microtubule-associated proteins are involved in microtubuli organization and their binding
55
56 affinity to microtubules is known to be controlled via phosphorylation.86, 87 Furthermore, the
57
58 21
59
60
1
2
3 phosphorylation of a putative MAP kinase (MAPK) superfamily protein
4
5 (GRMZM2G044557)88, 89
and a PROTEIN PHOSPHATASE 2C (PP2C) 64
6
7 (GRMZM2G107565; GRMZM2G021610)90, 91
increased and decreased upon water
8
9
10 deprivation, respectively. Since PP2Cs are known to negatively regulate stress-activated
11
12 MAPK pathways,92 our data suggested the similar regulation in maize leaves during drought
13
14 stress. Further, in stressed conditions, a SPS1-related proline-alanine-rich protein kinase
15
16 (GRMZM2G413544; GRMZM2G073399) was found to be less phosphorylated. This
17
18
serine/threonine kinase is a part of the Sterile 20 (STE20)-related kinase family that is
19
20
21 conserved across the fungi, plant and animal kingdom93 and its mammalian homologs are
22
23 known to act as a mediator of stress-activated signals.94 Up to our knowledge this protein has
24
25 not been related to drought stress.
26
27 Because of the small dataset of significantly regulated phosphosites, no GO classes
28
29
30 showed a significant over- or underrepresentation versus the dataset of all identified
31
32 phosphorylation sites (p <0.05). Alternatively, we investigated which GO classes were only
33
34 enriched in the control or the drought stressed samples versus a background set of all maize
35
36 proteins. This analysis showed that pathways involved in sodium transport, immune response
37
38
and chromatin silencing are exclusively overrepresented in the drought samples (Figure 5 and
39
40
41 Supplementary Table S15).
42
43
44
45 A crop PTM database
46
47
48
49
50 Comprehensive information of Arabidopsis protein phosphorylation sites can be found at the
51
52 PhosPHAt database41 and P3DB.95 However, information on detected phosphorylation sites in
53
54 crop species is only scarcely available in these resources. A very specific database, Atlas of
55
56 Maize Proteotypes, holds proteomics data for maize seed tissue and can also be consulted to
57
58 22
59
60
1
2
3 query phosphorylation sites, but only in limited, seed-specific datasets.42 To serve as a general
4
5 tool for PTMs in all plants, we generated a searchable database for plant protein
6
7 posttranslational modifications, called Plant PTM viewer
8
9
10 (bioinformatics.psb.ugent.be/webtools/ptm_viewer/). This database will function as an
11
12 important resource for future functional studies of plant protein posttranslational
13
14 modifications, including protein phosphorylation.
15
16
17
18
19
CONCLUSIONS
20
21
22
23 Phosphoproteomics workflows traditionally involve tedious labeling and fractionation steps
24
25 for comprehensive quantitative analysis of phosphorylation profiles. Here, we present a
26
27 streamlined and reproducible platform for quantitative phosphoproteomics, which i) does not
28
29
30 require specialized equipment nor expert software and can be easily implemented in any
31
32 molecular biology lab with access to a mass spectrometer, ii) involves limited sample prep
33
34 time due to the lack of labeling, fractionation and gel steps, and requires relatively low
35
36 amounts of starting material [which varies a lot across studies17, 96, 97] needed because of the
37
38
straightforward and sensitive pipeline, iii) does not require excessive MS analysis time as the
39
40
41 samples are not fractionated into multiple fractions, and iv) can be applied to model plants,
42
43 such as A. thaliana, and economically important crops, such as maize. Wherever possible,
44
45 protein levels inferred from non-phosphorylated peptides should be used to normalize
46
47 phosphopeptide intensities, so that true phosphorylation events rather than changes in
48
49
50 phosphoprotein amounts can be monitored. This has so far been poorly implemented in plant
51
52 phosphoprotemics, but is absolutely essential when reporting differential changes in the
53
54 phosphoproteome.
55
56
57
58 23
59
60
1
2
3 The largest hurdle for MS-based (phospho)proteomics for many crop species, such as
4
5 wheat, currently is the data analysis. This is mainly due to a lack of fully sequenced genomes
6
7 and also sufficiently annotated databases of crop species with large and complex genomes,
8
9
10 leading to decreasing efficiency in protein identification and subsequently, protein
11
12 quantification.98 However, crop (phospho)proteomics is continuously being boosted by the
13
14 rapid advancement of crop genomic resources. This will lead to increased knowledge on the
15
16 crop (phospho)proteome and will help understanding better the link between a desired
17
18
physiological or developmental trait and a genotype, and consequently improve breeding
19
20
21 strategies.99 Moreover, advanced molecular engineering techniques have been successfully
22
23 applied in crop species, and detailed insight in the dynamic (phospho)proteome will yield new
24
25 opportunities for crop improvement in a more targeted manner, e.g. using CRISPR/Cas9.100-
26
27 102
28
29
30 As a consequence of climate change, drought stress has become a severe limiting
31
32 factor in plant growth and productivity throughout the world.103, 104 Showcased in drought-
33
34 stressed maize leaves, our workflow enabled the in-depth quantitative comparison of
35
36 phosphorylation patterns. Finally, the data generated, comprising novel (phospho)protein
37
38
candidates implicated in drought stress signaling, contributes to our understanding of the
39
40
41 molecular and cellular mechanisms utilized by crops to survive unfavorable environmental
42
43 conditions.
44
45
46
47 ASSOCIATED CONTENT
48
49
50
51
52 Supporting Information
53
54 Figure S1. Pearson correlation coefficient for Arabidopsis proteome and phosphoproteome
55
56 data.
57
58 24
59
60
1
2
3 Figure S2. Pearson correlation coefficient for maize proteome data.
4
5 Figure S3. Pearson correlation coefficient for maize phosphoproteome data.
6
7 Figure S4. GO enrichment analysis (p<0.04) of significantly differential maize proteins, with
8
9
10 all identified maize proteins as background dataset.
11
12 Table S1. Pubmed search.
13
14 Table S2. Identified proteins in whole proteome analysis in Arabidopsis root.
15
16 Table S3. GO categorization in Arabidopsis proteome dataset.
17
18
Table S4. Total phosphosites identified in Arabidopsis roots.
19
20
21 Table S5. List of phosphosites in Arabidopsis uniquely identified in this study.Table S6.
22
23 Identified proteins in whole proteome analysis in maize leaves under drought and control
24
25 conditions.
26
27 Table S7. Maize proteins after filtering and Student's t-test.
28
29
30 Table S8. Significantly different protein groups upon drought stress in maize leaves (p<0.05).
31
32 Table S9. GO enrichment analysis of significantly differential maize proteins, with all
33
34 identified maize proteins as background dataset.
35
36 Table S10. Total phosphosites identified in maize leaves under control and drought
37
38
conditions.
39
40
41 Table S11. List of phosphosites in maize leaves uniquely identified in this study.
42
43 Table S12. Significantly different phosphosites upon drought stress in maize leaves (p<0.05),
44
45 after normalization to protein levels.
46
47 Table S13. Maize phosphosites after filtering and Student's t-test.
48
49
50 Table S14. Significantly different phosphosites upon drought stress in maize leaves (p<0.05),
51
52 without normalization to protein levels.
53
54 Table S15. GO classes exclusively overrepresented in one of the two samples, either control
55
56 or drought, with the whole maize annotated genome as background set.
57
58 25
59
60
1
2
3
4
5 AUTHOR INFORMATION
6
7
8
9
10 Corresponding Author
11
12 * E-mail: ive.desmet@psb.vib-ugent.be; Phone 003293313930.
13
14
15
16 Author Contributions
17
18
E.S., D.I., K.G., L.M. and I.D.S. designed research; E.S., L.D.V., M.V.B, H.N, D.M. and F.C.
19
20
21 performed research; E.S. and L.D.V. analyzed data; E.S., L.D.V., K.G., and I.D.S. wrote the
22
23 paper. All authors have given approval to the final version of the manuscript.
24
25 # and These authors contributed equally.
26
27
28
29
30 Notes
31
32 The authors declare no competing financial interest.
33
34
35
36 ACKNOWLEDGMENTS
37
38
39
40
41 E.S. was a Postdoctoral Fellow of the Research Foundation-Flanders. L.D.V. is the recipient
42
43 of a VIB International PhD program fellowship.
44
45
46
47 REFERENCES
48
49
50
51
52 1. Cohen, P., The origins of protein phosphorylation. Nat Cell Biol 2002, 4, (5), E127-30.
53
54 2. Xu, J.; Zhang, S., Mitogen-activated protein kinase cascades in signaling plant growth
55
56 and development. Trends Plant Sci 2015, 20, (1), 56-64.
57
58 26
59
60
1
2
3 3. Li, J.; Silva-Sanchez, C.; Zhang, T.; Chen, S.; Li, H., Phosphoproteomics technologies
4
5 and applications in plant biology research. Front Plant Sci 2015, 6, 430.
6
7 4. Benschop, J. J.; Mohammed, S.; O'Flaherty, M.; Heck, A. J.; Slijper, M.; Menke, F. L.,
8
9
10 Quantitative phosphoproteomics of early elicitor signaling in Arabidopsis. Mol Cell
11
12 Proteomics 2007, 6, (7), 1198-214.
13
14 5. Fan, S.; Meng, Y.; Song, M.; Pang, C.; Wei, H.; Liu, J.; Zhan, X.; Lan, J.; Feng, C.;
15
16 Zhang, S.; Yu, S., Quantitative phosphoproteomics analysis of nitric oxide-responsive
17
18
phosphoproteins in cotton leaf. PLoS One 2014, 9, (4), e94261.
19
20
21 6. Stecker, K. E.; Minkoff, B. B.; Sussman, M. R., Phosphoproteomic Analyses Reveal
22
23 Early Signaling Events in the Osmotic Stress Response. Plant Physiol 2014, 165, (3), 1171-
24
25 1187.
26
27 7. Yang, F.; Melo-Braga, M. N.; Larsen, M. R.; Jorgensen, H. J.; Palmisano, G., Battle
28
29
30 through signaling between wheat and the fungal pathogen Septoria tritici revealed by
31
32 proteomics and phosphoproteomics. Mol Cell Proteomics 2013, 12, (9), 2497-508.
33
34 8. Reiland, S.; Finazzi, G.; Endler, A.; Willig, A.; Baerenfaller, K.; Grossmann, J.;
35
36 Gerrits, B.; Rutishauser, D.; Gruissem, W.; Rochaix, J. D.; Baginsky, S., Comparative
37
38
phosphoproteome profiling reveals a function of the STN8 kinase in fine-tuning of cyclic
39
40
41 electron flow (CEF). Proc Natl Acad Sci U S A 2011, 108, (31), 12955-60.
42
43 9. Wu, L.; Hu, X.; Wang, S.; Tian, L.; Pang, Y.; Han, Z.; Wu, L.; Chen, Y., Quantitative
44
45 analysis of changes in the phosphoproteome of maize induced by the plant hormone salicylic
46
47 acid. Sci Rep 2015, 5, 18155.
48
49
50 10. Qiu, J.; Hou, Y.; Tong, X.; Wang, Y.; Lin, H.; Liu, Q.; Zhang, W.; Li, Z.; Nallamilli,
51
52 B. R.; Zhang, J., Quantitative phosphoproteomic analysis of early seed development in rice
53
54 (Oryza sativa L.). Plant Mol Biol 2015.
55
56
57
58 27
59
60
1
2
3 11. Hoehenwarter, W.; Thomas, M.; Nukarinen, E.; Egelhofer, V.; Rohrig, H.;
4
5 Weckwerth, W.; Conrath, U.; Beckers, G. J., Identification of novel in vivo MAP kinase
6
7 substrates in Arabidopsis thaliana through use of tandem metal oxide affinity
8
9
10 chromatography. Mol Cell Proteomics 2013, 12, (2), 369-80.
11
12 12. Roitinger, E.; Hofer, M.; Kocher, T.; Pichler, P.; Novatchkova, M.; Yang, J.;
13
14 Schlogelhofer, P.; Mechtler, K., Quantitative phosphoproteomics of the ataxia telangiectasia-
15
16 mutated (ATM) and ataxia telangiectasia-mutated and rad3-related (ATR) dependent DNA
17
18
damage response in Arabidopsis thaliana. Mol Cell Proteomics 2015, 14, (3), 556-71.
19
20
21 13. Nakagami, H.; Sugiyama, N.; Mochida, K.; Daudi, A.; Yoshida, Y.; Toyoda, T.;
22
23 Tomita, M.; Ishihama, Y.; Shirasu, K., Large-scale comparative phosphoproteomics identifies
24
25 conserved phosphorylation sites in plants. Plant Physiol 2010, 153, (3), 1161-74.
26
27 14. Marcon, C.; Malik, W. A.; Walley, J. W.; Shen, Z.; Paschold, A.; Smith, L. G.;
28
29
30 Piepho, H. P.; Briggs, S. P.; Hochholdinger, F., A high-resolution tissue-specific proteome
31
32 and phosphoproteome atlas of maize primary roots reveals functional gradients along the root
33
34 axes. Plant Physiol 2015, 168, (1), 233-46.
35
36 15. Bonhomme, L.; Valot, B.; Tardieu, F.; Zivy, M., Phosphoproteome dynamics upon
37
38
changes in plant water status reveal early events associated with rapid growth adjustment in
39
40
41 maize leaves. Mol Cell Proteomics 2012, 11, (10), 957-72.
42
43 16. Zhang, H.; Zhou, H.; Berke, L.; Heck, A. J.; Mohammed, S.; Scheres, B.; Menke, F.
44
45 L., Quantitative phosphoproteomics after auxin-stimulated lateral root induction identifies an
46
47 SNX1 protein phosphorylation site required for growth. Mol Cell Proteomics 2013, 12, (5),
48
49
50 1158-69.
51
52 17. Facette, M. R.; Shen, Z.; Bjornsdottir, F. R.; Briggs, S. P.; Smith, L. G., Parallel
53
54 proteomic and phosphoproteomic analyses of successive stages of maize leaf development.
55
56 Plant Cell 2013, 25, (8), 2798-812.
57
58 28
59
60
1
2
3 18. Sharma, K.; D'Souza, R. C.; Tyanova, S.; Schaab, C.; Wisniewski, J. R.; Cox, J.;
4
5 Mann, M., Ultradeep human phosphoproteome reveals a distinct regulatory nature of Tyr and
6
7 Ser/Thr-based signaling. Cell Rep 2014, 8, (5), 1583-94.
8
9
10 19. Lienhard, G. E., Non-functional phosphorylations? Trends Biochem Sci 2008, 33, (8),
11
12 351-2.
13
14 20. Niu, S.; Wang, Z.; Ge, D.; Zhang, G.; Li, Y., Prediction of functional phosphorylation
15
16 sites by incorporating evolutionary information. Protein Cell 2012, 3, (9), 675-90.
17
18
21. Kline, K. G.; Barrett-Wilt, G. A.; Sussman, M. R., In planta changes in protein
19
20
21 phosphorylation induced by the plant hormone abscisic acid. Proc Natl Acad Sci U S A 2010,
22
23 107, (36), 15986-91.
24
25 22. Qing, D.; Yang, Z.; Li, M.; Wong, W. S.; Guo, G.; Liu, S.; Guo, H.; Li, N.,
26
27 Quantitative and Functional Phosphoproteomic Analysis Reveals that Ethylene Regulates
28
29
30 Water Transport via the C-Terminal Phosphorylation of Aquaporin PIP2;1 in Arabidopsis.
31
32 Mol Plant 2015.
33
34 23. Hu, X.; Wu, L.; Zhao, F.; Zhang, D.; Li, N.; Zhu, G.; Li, C.; Wang, W.,
35
36 Phosphoproteomic analysis of the response of maize leaves to drought, heat and their
37
38
combination stress. Front Plant Sci 2015, 6, 298.
39
40
41 24. Lan, P.; Li, W.; Wen, T. N.; Schmidt, W., Quantitative phosphoproteome profiling of
42
43 iron-deficient Arabidopsis roots. Plant Physiol 2012, 159, (1), 403-17.
44
45 25. Engelsberger, W. R.; Schulze, W. X., Nitrate and ammonium lead to distinct global
46
47 dynamic phosphorylation patterns when resupplied to nitrogen-starved Arabidopsis seedlings.
48
49
50 Plant J 2012, 69, (6), 978-95.
51
52 26. Wang, P.; Xue, L.; Batelli, G.; Lee, S.; Hou, Y. J.; Van Oosten, M. J.; Zhang, H.; Tao,
53
54 W. A.; Zhu, J. K., Quantitative phosphoproteomics identifies SnRK2 protein kinase substrates
55
56
57
58 29
59
60
1
2
3 and reveals the effectors of abscisic acid action. Proc Natl Acad Sci U S A 2013, 110, (27),
4
5 11205-10.
6
7 27. Xue, L.; Wang, P.; Wang, L.; Renzi, E.; Radivojac, P.; Tang, H.; Arnold, R.; Zhu, J.
8
9
10 K.; Tao, W. A., Quantitative measurement of phosphoproteome response to osmotic stress in
11
12 arabidopsis based on Library-Assisted eXtracted Ion Chromatogram (LAXIC). Mol Cell
13
14 Proteomics 2013, 12, (8), 2354-69.
15
16 28. Ma, Q. F.; Wu, C. H.; Wu, M.; Pei, W. F.; Li, X. L.; Wang, W. K.; Zhang, J.; Yu, J.
17
18
W.; Yu, S. X., Integrative transcriptome, proteome, phosphoproteome and genetic mapping
19
20
21 reveals new aspects in a fiberless mutant of cotton. Sci Rep 2016, 6, 24485.
22
23 29. Lv, D. W.; Subburaj, S.; Cao, M.; Yan, X.; Li, X.; Appels, R.; Sun, D. F.; Ma, W.;
24
25 Yan, Y. M., Proteome and phosphoproteome characterization reveals new response and
26
27 defense mechanisms of Brachypodium distachyon leaves under salt stress. Mol Cell
28
29
30 Proteomics 2014, 13, (2), 632-52.
31
32 30. Margaria, P.; Abba, S.; Palmano, S., Novel aspects of grapevine response to
33
34 phytoplasma infection investigated by a proteomic and phospho-proteomic approach with
35
36 data integration into functional networks. BMC Genomics 2013, 14, 38.
37
38
31. Wu, R.; Dephoure, N.; Haas, W.; Huttlin, E. L.; Zhai, B.; Sowa, M. E.; Gygi, S. P.,
39
40
41 Correct interpretation of comprehensive phosphorylation dynamics requires normalization by
42
43 protein expression changes. Mol Cell Proteomics 2011, 10, (8), M111 009654.
44
45 32. Wu, X.; Gong, F.; Cao, D.; Hu, X.; Wang, W., Advances in crop proteomics: PTMs of
46
47 proteins under abiotic stress. Proteomics 2016, 16, (5), 847-65.
48
49
50 33. Gong, F.; Hu, X.; Wang, W., Proteomic analysis of crop plants under abiotic stress
51
52 conditions: where to focus our research? Front Plant Sci 2015, 6, 418.
53
54 34. McKersie, B., Planning for food security in a changing climate. J Exp Bot 2015, 66,
55
56 (12), 3435-50.
57
58 30
59
60
1
2
3 35. Shanker, A. K.; Maheswari, M.; Yadav, S. K.; Desai, S.; Bhanu, D.; Attal, N. B.;
4
5 Venkateswarlu, B., Drought stress responses in crops. Funct Integr Genomics 2014, 14, (1),
6
7 11-22.
8
9
10 36. Kosova, K.; Vitamvas, P.; Urban, M. O.; Klima, M.; Roy, A.; Prasil, I. T., Biological
11
12 Networks Underlying Abiotic Stress Tolerance in Temperate Crops--A Proteomic
13
14 Perspective. Int J Mol Sci 2015, 16, (9), 20913-42.
15
16 37. Proost, S.; Van Bel, M.; Vaneechoutte, D.; Van de Peer, Y.; Inze, D.; Mueller-Roeber,
17
18
B.; Vandepoele, K., PLAZA 3.0: an access point for plant comparative genomics. Nucleic
19
20
21 Acids Res 2015, 43, (Database issue), D974-81.
22
23 38. Cox, J.; Hein, M. Y.; Luber, C. A.; Paron, I.; Nagaraj, N.; Mann, M., Accurate
24
25 proteome-wide label-free quantification by delayed normalization and maximal peptide ratio
26
27 extraction, termed MaxLFQ. Mol Cell Proteomics 2014, 13, (9), 2513-26.
28
29
30 39. Vizcaino, J. A.; Csordas, A.; Del-Toro, N.; Dianes, J. A.; Griss, J.; Lavidas, I.; Mayer,
31
32 G.; Perez-Riverol, Y.; Reisinger, F.; Ternent, T.; Xu, Q. W.; Wang, R.; Hermjakob, H., 2016
33
34 update of the PRIDE database and its related tools. Nucleic Acids Res 2016, 44, (D1), D447-
35
36 56.
37
38
40. Tyanova, S.; Temu, T.; Sinitcyn, P.; Carlson, A.; Hein, M. Y.; Geiger, T.; Mann, M.;
39
40
41 Cox, J., The Perseus computational platform for comprehensive analysis of (prote)omics data.
42
43 Nat Methods 2016.
44
45 41. Durek, P.; Schmidt, R.; Heazlewood, J. L.; Jones, A.; MacLean, D.; Nagel, A.;
46
47 Kersten, B.; Schulze, W. X., PhosPhAt: the Arabidopsis thaliana phosphorylation site
48
49
50 database. An update. Nucleic Acids Res 2010, 38, (Database issue), D828-34.
51
52 42. Walley, J. W.; Shen, Z.; Sartor, R.; Wu, K. J.; Osborn, J.; Smith, L. G.; Briggs, S. P.,
53
54 Reconstruction of protein networks from an atlas of maize seed proteotypes. Proc Natl Acad
55
56 Sci U S A 2013, 110, (49), E4808-17.
57
58 31
59
60
1
2
3 43. Chou, M. F.; Schwartz, D., Biological sequence motif discovery using motif-x. Curr
4
5 Protoc Bioinformatics 2011, Chapter 13, Unit 13 15-24.
6
7 44. Szklarczyk, D.; Franceschini, A.; Wyder, S.; Forslund, K.; Heller, D.; Huerta-Cepas,
8
9
10 J.; Simonovic, M.; Roth, A.; Santos, A.; Tsafou, K. P.; Kuhn, M.; Bork, P.; Jensen, L. J.; von
11
12 Mering, C., STRING v10: protein-protein interaction networks, integrated over the tree of
13
14 life. Nucleic Acids Res 2015, 43, (Database issue), D447-52.
15
16 45. Rose, J. K.; Bashir, S.; Giovannoni, J. J.; Jahn, M. M.; Saravanan, R. S., Tackling the
17
18
plant proteome: practical approaches, hurdles and experimental tools. Plant J 2004, 39, (5),
19
20
21 715-33.
22
23 46. Herbert, B. R.; Molloy, M. P.; Gooley, A. A.; Walsh, B. J.; Bryson, W. G.; Williams,
24
25 K. L., Improved protein solubility in two-dimensional electrophoresis using tributyl
26
27 phosphine as reducing agent. Electrophoresis 1998, 19, (5), 845-51.
28
29
30 47. Glatter, T.; Ludwig, C.; Ahrne, E.; Aebersold, R.; Heck, A. J.; Schmidt, A., Large-
31
32 scale quantitative assessment of different in-solution protein digestion protocols reveals
33
34 superior cleavage efficiency of tandem Lys-C/trypsin proteolysis over trypsin digestion. J
35
36 Proteome Res 2012, 11, (11), 5145-56.
37
38
48. Zhou, H.; Ye, M.; Dong, J.; Corradini, E.; Cristobal, A.; Heck, A. J.; Zou, H.;
39
40
41 Mohammed, S., Robust phosphoproteome enrichment using monodisperse microsphere-based
42
43 immobilized titanium (IV) ion affinity chromatography. Nat Protoc 2013, 8, (3), 461-80.
44
45 49. de Graaf, E. L.; Giansanti, P.; Altelaar, A. F.; Heck, A. J., Single-step enrichment by
46
47 Ti4+-IMAC and label-free quantitation enables in-depth monitoring of phosphorylation
48
49
50 dynamics with high reproducibility and temporal resolution. Mol Cell Proteomics 2014, 13,
51
52 (9), 2426-34.
53
54 50. Michalski, A.; Damoc, E.; Hauschild, J. P.; Lange, O.; Wieghaus, A.; Makarov, A.;
55
56 Nagaraj, N.; Cox, J.; Mann, M.; Horning, S., Mass spectrometry-based proteomics using Q
57
58 32
59
60
1
2
3 Exactive, a high-performance benchtop quadrupole Orbitrap mass spectrometer. Mol Cell
4
5 Proteomics 2011, 10, (9), M111 011015.
6
7 51. Cox, J.; Neuhauser, N.; Michalski, A.; Scheltema, R. A.; Olsen, J. V.; Mann, M.,
8
9
10 Andromeda: a peptide search engine integrated into the MaxQuant environment. J Proteome
11
12 Res 2011, 10, (4), 1794-805.
13
14 52. Cox, J.; Mann, M., MaxQuant enables high peptide identification rates, individualized
15
16 p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat Biotechnol 2008,
17
18
26, (12), 1367-72.
19
20
21 53. Budak, H.; Hussain, B.; Khan, Z.; Ozturk, N. Z.; Ullah, N., From Genetics to
22
23 Functional Genomics: Improvement in Drought Signaling and Tolerance in Wheat. Front
24
25 Plant Sci 2015, 6, 1012.
26
27 54. Gilbert, M. E.; Medina, V., Drought Adaptation Mechanisms Should Guide
28
29
30 Experimental Design. Trends Plant Sci 2016.
31
32 55. Claeys, H.; Inze, D., The agony of choice: how plants balance growth and survival
33
34 under water-limiting conditions. Plant Physiol 2013, 162, (4), 1768-79.
35
36 56. Nelissen, H.; Rymen, B.; Jikumaru, Y.; Demuynck, K.; Van Lijsebettens, M.; Kamiya,
37
38
Y.; Inze, D.; Beemster, G. T., A local maximum in gibberellin levels regulates maize leaf
39
40
41 growth by spatial control of cell division. Curr Biol 2012, 22, (13), 1183-7.
42
43 57. Avramova, V.; Sprangers, K.; Beemster, G. T., The Maize Leaf: Another Perspective
44
45 on Growth Regulation. Trends Plant Sci 2015, 20, (12), 787-97.
46
47 58. Demirevska, K.; Zasheva, D.; Dimitrov, R.; Simova-Stoilova, L.; Stamenova, M.;
48
49
50 Feller, F., Drought stress effects on Rubisco in wheat: changes in the Rubisco large subunit.
51
52 Acta Physiol Plant 2009, 31, 1129.
53
54
55
56
57
58 33
59
60
1
2
3 59. De La Fuente, G. N.; Murray, S. C.; Isakeit, T.; Park, Y. S.; Yan, Y.; Warburton, M.
4
5 L.; Kolomiets, M. V., Characterization of genetic diversity and linkage disequilibrium of
6
7 ZmLOX4 and ZmLOX5 loci in maize. PLoS One 2013, 8, (1), e53973.
8
9
10 60. Sengupta, D.; Naik, D.; Reddy, A. R., Plant aldo-keto reductases (AKRs) as multi-
11
12 tasking soldiers involved in diverse plant metabolic processes and stress defense: A structure-
13
14 function update. J Plant Physiol 2015, 179, 40-55.
15
16 61. Riccardi, F.; Gazeau, P.; Jacquemot, M. P.; Vincent, D.; Zivy, M., Deciphering genetic
17
18
variations of proteome responses to water deficit in maize leaves. Plant Physiol Biochem
19
20
21 2004, 42, (12), 1003-11.
22
23 62. Thapar, N.; Kim, A. K.; Clarke, S., Distinct patterns of expression but similar
24
25 biochemical properties of protein L-isoaspartyl methyltransferase in higher plants. Plant
26
27 Physiol 2001, 125, (2), 1023-35.
28
29
30 63. Vert, G., Plant signaling: brassinosteroids, immunity and effectors are BAK ! Curr
31
32 Biol 2008, 18, (20), R963-5.
33
34 64. Yamada, K.; Yamashita-Yamada, M.; Hirase, T.; Fujiwara, T.; Tsuda, K.; Hiruma, K.;
35
36 Saijo, Y., Danger peptide receptor signaling in plants ensures basal immunity upon pathogen-
37
38
induced depletion of BAK1. EMBO J 2015.
39
40
41 65. Kir, G.; Ye, H.; Nelissen, H.; Neelakandan, A. K.; Kusnandar, A. S.; Luo, A.; Inze,
42
43 D.; Sylvester, A. W.; Yin, Y.; Becraft, P. W., RNA Interference Knockdown of
44
45 BRASSINOSTEROID INSENSITIVE1 in Maize Reveals Novel Functions for
46
47 Brassinosteroid Signaling in Controlling Plant Architecture. Plant Physiol 2015, 169, (1),
48
49
50 826-39.
51
52 66. Marshall, A.; Aalen, R. B.; Audenaert, D.; Beeckman, T.; Broadley, M. R.; Butenko,
53
54 M. A.; Cano-Delgado, A. I.; de Vries, S.; Dresselhaus, T.; Felix, G.; Graham, N. S.; Foulkes,
55
56 J.; Granier, C.; Greb, T.; Grossniklaus, U.; Hammond, J. P.; Heidstra, R.; Hodgman, C.;
57
58 34
59
60
1
2
3 Hothorn, M.; Inze, D.; Ostergaard, L.; Russinova, E.; Simon, R.; Skirycz, A.; Stahl, Y.;
4
5 Zipfel, C.; De Smet, I., Tackling drought stress: receptor-like kinases present new approaches.
6
7 Plant Cell 2012, 24, (6), 2262-78.
8
9
10 67. Janeczko, A.; Oklestkova, J.; Pociecha, E.; Koscielniak, J.; Mirek, M., Physiological
11
12 effects and transport of 24-epibrassinolide in heat-stressed barley. Acta Physiologiae
13
14 Plantarum 2011, 33, (4), 1249-1259.
15
16 68. Anjum, S. A.; Wang, L. C.; Farooq, M.; Hussain, M.; Xue, L. L.; Zou, C. M.,
17
18
Brassinolide Application Improves the Drought Tolerance in Maize Through Modulation of
19
20
21 Enzymatic Antioxidants and Leaf Gas Exchange. Journal of Agronomy and Crop Science
22
23 2011, 197, (3), 177-185.
24
25 69. Bajguz, A.; Hayat, S., Effects of brassinosteroids on the plant responses to
26
27 environmental stresses. Plant Physiology and Biochemistry 2009, 47, (1), 1-8.
28
29
30 70. Pan, L.; Zhang, X.; Wang, J.; Ma, X.; Zhou, M.; Huang, L.; Nie, G.; Wang, P.; Yang,
31
32 Z.; Li, J., Transcriptional Profiles of Drought-Related Genes in Modulating Metabolic
33
34 Processes and Antioxidant Defenses in Lolium multiflorum. Front Plant Sci 2016, 7, 519.
35
36 71. Inomata, N.; Oomura, Y.; Akaike, N.; Edwards, C., The anion selectivity of the
37
38
gamma-aminobutyric acid controlled chloride channel in the perfused spinal ganglion cell of
39
40
41 frog. Neurosci Res 1986, 3, (5), 371-83.
42
43 72. Merewitz, E. B.; Gianfagna, T.; Huang, B., Protein accumulation in leaves and roots
44
45 associated with improved drought tolerance in creeping bentgrass expressing an ipt gene for
46
47 cytokinin synthesis. J Exp Bot 2011, 62, (15), 5311-33.
48
49
50 73. Kim, J. M.; Sasaki, T.; Ueda, M.; Sako, K.; Seki, M., Chromatin changes in response
51
52 to drought, salinity, heat, and cold stresses in plants. Front Plant Sci 2015, 6, 114.
53
54 74. Chaves, M. M.; Flexas, J.; Pinheiro, C., Photosynthesis under drought and salt stress:
55
56 regulation mechanisms from whole plant to cell. Ann Bot 2009, 103, (4), 551-60.
57
58 35
59
60
1
2
3 75. Gururani, M. A.; Venkatesh, J.; Tran, L. S., Regulation of Photosynthesis during
4
5 Abiotic Stress-Induced Photoinhibition. Mol Plant 2015, 8, (9), 1304-20.
6
7 76. Avramova, V.; AbdElgawad, H.; Zhang, Z.; Fotschki, B.; Casadevall, R.; Vergauwen,
8
9
10 L.; Knapen, D.; Taleisnik, E.; Guisez, Y.; Asard, H.; Beemster, G. T., Drought Induces
11
12 Distinct Growth Response, Protection, and Recovery Mechanisms in the Maize Leaf Growth
13
14 Zone. Plant Physiol 2015, 169, (2), 1382-96.
15
16 77. Ghosh, D.; Xu, J., Abiotic stress responses in plant roots: a proteomics perspective.
17
18
Front Plant Sci 2014, 5, 6.
19
20
21 78. Schluter, U.; Colmsee, C.; Scholz, U.; Brautigam, A.; Weber, A. P.; Zellerhoff, N.;
22
23 Bucher, M.; Fahnenstich, H.; Sonnewald, U., Adaptation of maize source leaf metabolism to
24
25 stress related disturbances in carbon, nitrogen and phosphorus balance. BMC Genomics 2013,
26
27 14, 442.
28
29
30 79. van Wijk, K. J.; Friso, G.; Walther, D.; Schulze, W. X., Meta-Analysis of Arabidopsis
31
32 thaliana Phospho-Proteomics Data Reveals Compartmentalization of Phosphorylation Motifs.
33
34 Plant Cell 2014, 26, (6), 2367-2389.
35
36 80. Zhang, M.; Lv, D.; Ge, P.; Bian, Y.; Chen, G.; Zhu, G.; Li, X.; Yan, Y.,
37
38
Phosphoproteome analysis reveals new drought response and defense mechanisms of seedling
39
40
41 leaves in bread wheat (Triticum aestivum L.). J Proteomics 2014, 109, 290-308.
42
43 81. Lv, D. W.; Li, X.; Zhang, M.; Gu, A. Q.; Zhen, S. M.; Wang, C.; Li, X. H.; Yan, Y.
44
45 M., Large-scale phosphoproteome analysis in seedling leaves of Brachypodium distachyon L.
46
47 BMC Genomics 2014, 15, 375.
48
49
50 82. Hou, Y.; Qiu, J.; Tong, X.; Wei, X.; Nallamilli, B. R.; Wu, W.; Huang, S.; Zhang, J.,
51
52 A comprehensive quantitative phosphoproteome analysis of rice in response to bacterial
53
54 blight. BMC Plant Biol 2015, 15, 163.
55
56
57
58 36
59
60
1
2
3 83. Sircar, S.; Parekh, N., Functional characterization of drought-responsive modules and
4
5 genes in Oryza sativa: a network-based approach. Front Genet 2015, 6, 256.
6
7 84. Pugacheva, E. N.; Jablonski, S. A.; Hartman, T. R.; Henske, E. P.; Golemis, E. A.,
8
9
10 HEF1-dependent Aurora A activation induces disassembly of the primary cilium. Cell 2007,
11
12 129, (7), 1351-63.
13
14 85. Nick, P., Microtubules, signalling and abiotic stress. Plant J 2013, 75, (2), 309-23.
15
16 86. Beck, M.; Komis, G.; Muller, J.; Menzel, D.; Samaj, J., Arabidopsis homologs of
17
18
nucleus- and phragmoplast-localized kinase 2 and 3 and mitogen-activated protein kinase 4
19
20
21 are essential for microtubule organization. Plant Cell 2010, 22, (3), 755-71.
22
23 87. Komis, G.; Illes, P.; Beck, M.; Samaj, J., Microtubules and mitogen-activated protein
24
25 kinase signalling. Curr Opin Plant Biol 2011, 14, (6), 650-7.
26
27 88. Sinha, A. K.; Jaggi, M.; Raghuram, B.; Tuteja, N., Mitogen-activated protein kinase
28
29
30 signaling in plants under abiotic stress. Plant Signal Behav 2011, 6, (2), 196-203.
31
32 89. Cowan, K. J.; Storey, K. B., Mitogen-activated protein kinases: new signaling
33
34 pathways functioning in cellular responses to environmental stress. J Exp Biol 2003, 206, (Pt
35
36 7), 1107-15.
37
38
90. Vranova, E.; Langebartels, C.; Van Montagu, M.; Inze, D.; Van Camp, W., Oxidative
39
40
41 stress, heat shock and drought differentially affect expression of a tobacco protein
42
43 phosphatase 2C. J Exp Bot 2000, 51, (351), 1763-4.
44
45 91. Schweighofer, A.; Hirt, H.; Meskiene, I., Plant PP2C phosphatases: emerging
46
47 functions in stress signaling. Trends Plant Sci 2004, 9, (5), 236-43.
48
49
50 92. Kondoh, K.; Nishida, E., Regulation of MAP kinases by MAP kinase phosphatases.
51
52 Biochim Biophys Acta 2007, 1773, (8), 1227-37.
53
54 93. Gagnon, K. B.; Delpire, E., Molecular physiology of SPAK and OSR1: two Ste20-
55
56 related protein kinases regulating ion transport. Physiol Rev 2012, 92, (4), 1577-617.
57
58 37
59
60
1
2
3 94. Johnston, A. M.; Naselli, G.; Gonez, L. J.; Martin, R. M.; Harrison, L. C.;
4
5 DeAizpurua, H. J., SPAK, a STE20/SPS1-related kinase that activates the p38 pathway.
6
7 Oncogene 2000, 19, (37), 4290-7.
8
9
10 95. Yao, Q.; Ge, H.; Wu, S.; Zhang, N.; Chen, W.; Xu, C.; Gao, J.; Thelen, J. J.; Xu, D.,
11
12 P(3)DB 3.0: From plant phosphorylation sites to protein networks. Nucleic Acids Res 2014,
13
14 42, (Database issue), D1206-13.
15
16 96. Pi, E.; Qu, L.; Hu, J.; Huang, Y.; Qiu, L.; Lu, H.; Jiang, B.; Liu, C.; Peng, T.; Zhao,
17
18
Y.; Wang, H.; Tsai, S. N.; Ngai, S.; Du, L., Mechanisms of Soybean Roots' Tolerances to
19
20
21 Salinity Revealed by Proteomic and Phosphoproteomic Comparisons Between Two Cultivars.
22
23 Mol Cell Proteomics 2016, 15, (1), 266-88.
24
25 97. Choudhary, M. K.; Nomura, Y.; Wang, L.; Nakagami, H.; Somers, D. E., Quantitative
26
27 Circadian Phosphoproteomic Analysis of Arabidopsis Reveals Extensive Clock Control of
28
29
30 Key Components in Physiological, Metabolic, and Signaling Pathways. Mol Cell Proteomics
31
32 2015, 14, (8), 2243-60.
33
34 98. Barkla, B. J.; Vera-Estrella, R.; Pantoja, O., Progress and challenges for abiotic stress
35
36 proteomics of crop plants. Proteomics 2013, 13, (12-13), 1801-15.
37
38
99. Hu, J.; Rampitsch, C.; Bykova, N. V., Advances in plant proteomics toward
39
40
41 improvement of crop productivity and stress resistancex. Front Plant Sci 2015, 6, 209.
42
43 100. Wang, Y.; Cheng, X.; Shan, Q.; Zhang, Y.; Liu, J.; Gao, C.; Qiu, J. L., Simultaneous
44
45 editing of three homoeoalleles in hexaploid bread wheat confers heritable resistance to
46
47 powdery mildew. Nat Biotechnol 2014, 32, (9), 947-51.
48
49
50 101. Bortesi, L.; Fischer, R., The CRISPR/Cas9 system for plant genome editing and
51
52 beyond. Biotechnol Adv 2015, 33, (1), 41-52.
53
54
55
56
57
58 38
59
60
1
2
3 102. Svitashev, S.; Young, J. K.; Schwartz, C.; Gao, H.; Falco, S. C.; Cigan, A. M.,
4
5 Targeted Mutagenesis, Precise Gene Editing, and Site-Specific Gene Insertion in Maize Using
6
7 Cas9 and Guide RNA. Plant Physiol 2015, 169, (2), 931-45.
8
9
10 103. Cairns, J. E.; Sanchez, C.; Vargas, M.; Ordonez, R.; Araus, J. L., Dissecting maize
11
12 productivity: ideotypes associated with grain yield under drought stress and well-watered
13
14 conditions. J Integr Plant Biol 2012, 54, (12), 1007-20.
15
16 104. Harrison, M. T.; Tardieu, F.; Dong, Z.; Messina, C. D.; Hammer, G. L., Characterizing
17
18
drought stress and trait influence on maize yield under current and future conditions. Glob
19
20
21 Chang Biol 2014, 20, (3), 867-78.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58 39
59
60
1
2
3
4
5 TABLES
6
7
8
9
10 Table 1. Motif-X analysis for overrepresented phosphorylation motifs of all identified phosphosites in maize leaves.
11
12
13
14 Motif Foreground Foreground Background Background Fold
15 Motif
16 score matches size matches size increase
17
18 . . . . . .S . . . . . 16.00 210 784 63261 975876 4.13
19 . . . . . .S. SP. . . 13.24 20 415 6516 850139 6.29
20
21 . . . . . .SD. E . . . 19.40 22 437 3450 853589 12.46
22
23 . . .R . .S. . . . . . 16.00 98 535 52077 905666 3.19
24
. . .RS. S. . . . . . 22.98 39 574 6949 912615 8.92
25
26 . . .K . .S. . . . . . 6.33 43 395 39892 843623 2.30
27
28 . . . . . .TP. . . . . 6.60 27 140 35595 556423 3.01
29 6.01 29 113 50857 520828 2.63
. . . . . .TS. . . . .
30
31
32
33
34
35
36
37
38
39
40
40
41
42
43
44
45
46 ACS Paragon Plus Environment
47
48
1
2
3
4
5 Table 2. Significantly deregulated phosphosites upon drought treatment. Statistical test (p<0.05) were performed on phosphosites that were
6
7 corrected against the protein level.
8
9
10 PLAZA identifier MaizeSequence ID Amino acid Position Protein description
11
12
13 Upregulated phosphosites
14
15 ZM01G32530 AC234203.1_FG004 S 878 Oligomycin resistance ATP-dependent permease YOR1
16 ZM01G33500 GRMZM2G096806 S 1016 No description available
17 ZM03G34030 GRMZM2G157317 T 222 RANBP2-like and GRIP domain-containing protein
18 ZM10G23890 GRMZM2G114182 T 1042 No description available
19 ZM02G06750 GRMZM2G080274 S 82 Histone one (H1) 101
20 ZM02G38420 GRMZM2G457889 S 470 Histone deacetylase 6, RPD3 histone deacetylase homolog;probable histone
21 deacetylase 19
22 ZM03G40140 GRMZM2G078826 S 730 RNA helicase DBP2
23
24
Downregulated phosphosites
25
26 ZM02G47600 GRMZM2G034572 S 741 BRASSINOSTEROID INSENSITIVE 1-associated receptor kinase 1
27 ZM03G34030 GRMZM2G157317 S 444 RANBP2-like and GRIP domain-containing protein
28 ZM03G34030 GRMZM2G157317 T 440 RANBP2-like and GRIP domain-containing protein
29 ZM01G20150 GRMZM2G083173 S 457 Monosaccharide-sensing protein 3
30 ZM01G52420 GRMZM2G169671 S 247 Stem-specific protein TSJT1
31 ZM04G05250 GRMZM2G015295 S 465 adenosylhomocysteinase
32
ZM04G15030 AC207890.3_FG002 S 970 Staphylococcal nuclease domain-containing protein 1
33
ZM05G30080 GRMZM2G451314 S 4 Histidine decarboxylase
34
ZM05G37350 GRMZM2G027741 S 35 protein transporter, TOM1-like protein 2
35
ZM08G18360 GRMZM2G142640 S 84 60S ribosomal protein L24
36
ZM10G16510 GRMZM2G050647 S 499 Exocyst complex component SEC5A
37
38
39
40
41
41
42
43
44
45
46 ACS Paragon Plus Environment
47
48
1
2
3 FIGURE LEGENDS
4
5
6
7 Figure 1. Workflow and key improved steps.
8
9
10
11
12 Figure 2. Overrepresentation of cellular components of identified proteins in Arabidopsis
13
14 thaliana roots.
15
16
17
18
19
Figure 3. Differential (phospho)proteome response upon drought stress in maize leaves.
20
21 Heat map showing average log2 values of MaxLFQ intensities of the significantly
22
23 differentially expressed proteins (A), phosphosite intensities of the significantly regulated
24
25 phosphosites after normalization to protein levels (B), or phosphosite intensities of the
26
27 significantly regulated phosphosites without correction for protein abundance (C). The log2
28
29
30 values of the intensities were Z-scored for graphical representation.
31
32
33
34 Figure 4. Protein-protein interaction networks resulted from STRING analysis of maize
35
36 proteins with significant abundance changes. Cytoscape was used for visualization.
37
38
Subnetworks with fewer than 6 interactors are excluded from the representation. Nodes in red
39
40
41 represent upregulated proteins, in green downregulated proteins upon drought. Interaction
42
43 groups are indicated with black circles. The PLAZA identifiers for maize proteins were used
44
45 because of space constraints.
46
47
48
49
50 Figure 5. GO enrichment for biological processes and molecular functions of the
51
52 phosphoproteins in maize. The analysis was performed separately for control (blue) and
53
54 drought (red) conditions (see Supplementary Table S15 for details).
55
56
57
58 42
59
60
1
2
3 FOR TOC ONLY
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18 Photograph courtesy of Lam Dai Vu. Copyright 2016.
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58 43
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 Figure 1. Workflow and key improved steps.
27
28 94x52mm (300 x 300 DPI)
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Figure 2. Overrepresentation of cellular components of identified proteins in Arabidopsis thaliana roots.
26
27 40x23mm (300 x 300 DPI)
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Figure 3. Differential (phospho)proteome response upon drought stress in maize leaves.
47
Heat map showing average log2 values of MaxLFQ intensities of the significantly differentially expressed
48 proteins (A), phosphosite intensities of the significantly regulated phosphosites after normalization to protein
49 levels (B), or phosphosite intensities of the significantly regulated phosphosites without correction for
50 protein abundance (C). The log2 values of the intensities were Z-scored for graphical representation.
51
52
53 195x366mm (300 x 300 DPI)
54
55
56
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29 Figure 4. Protein-protein interaction networks resulted from STRING analysis of maize proteins with
30 significant abundance changes. Cytoscape was used for visualization. Subnetworks with fewer than 6
31 interactors are excluded from the representation. Nodes in red represent upregulated proteins, in green
32 downregulated proteins upon drought. Interaction groups are indicated with black circles. The PLAZA
33 identifiers for maize proteins were used because of space constraints.
34
106x70mm (300 x 300 DPI)
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24 Figure 5. GO enrichment for biological processes and molecular functions of the phosphoproteins in maize.
25 The analysis was performed separately for control (blue) and drought (red) conditions (see Supplementary
26 Table S15 for details).
27
79x39mm (300 x 300 DPI)
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

An Up-To-date Workflow For

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

An Up-To-date Workflow For

Загружено:

Авторское право:

Доступные форматы

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Journal of Proteome Research is published by the American Chemical Society. 1155

Вам также может понравиться