• Aucun résultat trouvé

State transition7-dependent phosphorylation is modulated by changing environmental conditions, and its absence triggers remodeling of photosynthetic protein complexes

N/A
N/A
Protected

Academic year: 2022

Partager "State transition7-dependent phosphorylation is modulated by changing environmental conditions, and its absence triggers remodeling of photosynthetic protein complexes"

Copied!
78
0
0

Texte intégral

(1)

Article

Reference

State transition7-dependent phosphorylation is modulated by changing environmental conditions, and its absence triggers

remodeling of photosynthetic protein complexes

BERGNER, Sonja Verena, et al.

Abstract

In plants and algae, the serine/threonine kinase STN7/STT7, orthologous protein kinases in Chlamydomonas reinhardtii and Arabidopsis (Arabidopsis thaliana), respectively, is an important regulator in acclimation to changing light environments. In this work, we assessed STT7-dependent protein phosphorylation under high light in C. reinhardtii, known to fully induce the expression of light-harvesting complex stress-related protein3 (LHCSR3) and a nonphotochemical quenching mechanism, in relationship to anoxia where the activity of cyclic electron flow is stimulated. Our quantitative proteomics data revealed numerous unique STT7 protein substrates and STT7-dependent protein phosphorylation variations that were reliant on the environmental condition. These results indicate that STT7-dependent phosphorylation is modulated by the environment and point to an intricate chloroplast phosphorylation network responding in a highly sensitive and dynamic manner to environmental cues and alterations in kinase function. Functionally, the absence of the STT7 kinase triggered changes in protein expression and photoinhibition of photosystem [...]

BERGNER, Sonja Verena, et al. State transition7-dependent phosphorylation is modulated by changing environmental conditions, and its absence triggers remodeling of photosynthetic protein complexes. Plant Physiology, 2015, vol. 168, no. 2, p. 615-634

DOI : 10.1104/pp.15.00072 PMID : 25858915

Available at:

http://archive-ouverte.unige.ch/unige:86430

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

1 Running head: STT7 is tuning acclimation 1

2

Corresponding Author: Prof. Dr. Michael Hippler 3

4

Address:

5

Institut für Biologie und Biotechnologie der Pflanzen 6

Westfälische Wilhelms-Universität Münster 7

Schlossplatz 8 8

48143 Münster 9

10

Telephone: +49 251 83-24790 11

12

E-mail: mhippler@uni-muenster.de 13

14

Research Area: Membranes, Transport and Biogenetics: Associate Editor Anna Amtmann 15

(Glasgow) 16

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

(3)

2 Title of article:

33

State transition7-dependent phosphorylation is modulated by changing environmental 34

conditions and its absence triggers remodeling of photosynthetic protein complexes 35

36

Sonja Verena Bergnera, Martin Scholza, Kerstin Trompelta, Johannes Bartha, Philipp Gäbeleina, 37

Janina Steinbecka, Huidan Xuea, Sophie Clowezb, Geoffrey Fucilec, Michel Goldschmidt- 38

Clermontc, Christian Fufezana and Michael Hipplera*

39

40 a Institute of Plant Biology and Biotechnology, University of Münster, Münster 48143, Germany 41

b Institut de Biologie Physico-Chimique, UMR 7141 CNRS-UPMC, 13 rue P et M Curie, 75005 42

Paris, France.

43

c Department of Botany and Plant Biology and Institute of Genetics and Genomics in Geneva 44

University of Geneva, CH-1211 Geneva 4, Switzerland 45

46

S.V.B. and M.S. contributed equally to this work 47

48 49

One sentence summary:

50 51

Lack of STT7-dependent protein phosphorylation in Chlamydomonas triggers stress-induced 52

changes in protein expression, photosystem II core phosphorylation and association of LHCSR3 53

with photosystem I 54

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69

(4)

3

Footnotes: Support by the Swiss National Foundation (grant 31003A_146300) to M.G.C. and 70

from the “Deutsche Forschungsgemeinschaft” (DFG, HI 739/7.2 in FOR964 and HI 739/8.1) to 71

M.H. is acknowledged.

72 73

*Corresponding author: Michael Hippler (mhippler@uni-muenster.de) 74

75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115

(5)

4 Abstract

116 117

In plants and algae the Ser/Thr kinase STN7/STT7 is an important regulator in acclimation to 118

changing light environments. In the present work we assessed STT7-dependent protein 119

phosphorylation under high light, known to fully induce expression of LHCSR3 and non- 120

photochemical quenching mechanism, in relationship to anoxia where activity of cyclic electron 121

flow (CEF) is stimulated. Our quantitative proteomics data revealed numerous novel STT7 122

protein substrates and STT7-dependent protein phosphorylation variations that were reliant on the 123

environmental condition. These results indicate that STT7-dependent phosphorylation is 124

modulated by the environment and point to an intricate chloroplast phosphorylation network 125

responding in a highly sensitive and dynamic manner to environmental cues and alterations in 126

kinase function. Functionally, the absence of the STT7 kinase triggered changes in protein 127

expression, photoinhibition of PSI and resulted in remodeling of photosynthetic complexes. This 128

remodeling initiated a pronounced association of LHCSR3 with PSI-LHCI-FNR-supercomplexes.

129

Lack of STT7 kinase strongly diminished PSII-LHCII supercomplexes while PSII core complex 130

phosphorylation and accumulation was significantly enhanced. In conclusion, our study provides 131

strong evidence that regulation of protein phosphorylation is critical for driving successful 132

acclimation to high light and anoxic growth environments and gives new insights into 133

acclimation strategies to these environmental conditions.

134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154

(6)

5 Introduction

155 156

Oxygenic photosynthesis converts solar energy into chemical energy. This energy is utilized for 157

carbon dioxide assimilation, allowing the formation of complex organic material. Plant 158

photosynthesis is performed by a series of reactions in and at the thylakoid membrane resulting in 159

light-dependent water oxidation, NADP reduction and ATP formation (Whatley et al., 1963).

160

These light reactions are catalyzed by two photosystems (PSI and PSII). A third multi-protein 161

complex, also embedded in the thylakoid membrane, is the cytochrome b6f complex (cyt b6f 162

complex) that links photosynthetic electron transfer processes between the two photosystems and 163

functions in proton translocation. The ATP synthase takes advantage of the proton-motive force 164

that is generated by the light reactions (Mitchell, 1961) to produce ATP. ATP and NADPH, 165

generated through linear electron flow from PSII to PSI, drive the Calvin Benson Bassham cycle 166

(Bassham et al., 1950) to fix CO2. Alternatively cyclic electron flow (CEF) between PSI and the 167

cyt b6f complex solely produces ATP (Arnon, 1959).

168

Under normal growth conditions, CEF provides additionally required ATP for CO2

169

fixation (Lucker and Kramer, 2013), counteracts over-reduction of the PSI acceptor side under 170

stressful environmental cues and re-adjusts the ATP poise leading to increased lumen 171

acidification important for photo-protection (Alric, 2010; Peltier et al., 2010; Leister and 172

Shikanai, 2013; Shikanai, 2013). In microalgae and vascular plants, CEF relies on the NAD(P)H 173

dehydrogenase (NDH)-dependent and/or PROTONGRADIENT REGULATION5 (PGR5)- 174

related pathways (Munekage et al., 2002; Munekage et al., 2004; Petroutsos et al., 2009; Tolleter 175

et al., 2011; Johnson et al., 2014). For both pathways, supercomplexes consisting of PSI-LHCI 176

and components of the respective electron transfer routes have been identified. In Arabidopsis 177

thaliana, a novel NDH-PSI supercomplex with a molecular mass of >1000 kDa was discovered 178

(Peng et al, 2008). From Chlamydomonas reinhardtii, Iwai et al. (Iwai et al, 2010) isolated a 179

protein supercomplex composed of PSI-LHCI, LHCII, the cytochrome b6/f complex, ferredoxin 180

(Fd)-NADPH oxidoreductase (FNR), and PGRL1.

181

PGRL1 and PGR5 interact physically in A. thaliana and associate with PSI to allow the 182

operation of CEF (DalCorso et al, 2008). Functional data suggest that PGRL1 might operate as a 183

ferredoxin-plastoquinone reductase (Hertle et al., 2013). The PGRL1 containing CEF- 184

supercomplex isolated from C. reinhardtii is capable of CEF under in vitro conditions in the 185

presence of exogenously added soluble plastocyanin and ferredoxin (Iwai et al, 2010). Terashima 186

(7)

6

et al. isolated a CEF supercomplex of similar composition from anaerobic growth conditions that 187

was active in vitro and contained proteins such as the chloroplast localized Ca2+ sensor CAS and 188

ANR1 (anaerobic response 1), which were also shown to be functionally important for efficient 189

CEF in the alga (Terashima et al., 2012). Notably it was suggested that the onset of CEF in C.

190

reinhardtii is redox-controlled (Takahashi et al., 2013).

191

It has been demonstrated that efficient CEF is crucial for successful acclimation to excess 192

light (Munekage et al., 2004; Dang et al., 2014; Johnson et al., 2014; Kukuczka et al., 2014). The 193

most rapid response to excess light, however, relies on a mechanism called non-photochemical 194

quenching (NPQ). The fastest constituent of NPQ is qE, which operates at a time scale of seconds 195

to minutes and regulates the thermal dissipation of excess absorbed light energy, thereby 196

providing effective photo-protection. In vascular plants, the PSII protein PSBS is essential for qE 197

(Li et al, 2000), whereas qE induction in the green alga Chlamydomonas reinhardtii is mediated 198

by LHCSR3, an ancient light-harvesting protein that is missing in vascular plants (Peers et al., 199

2009). CEF and qE are complementary for acclimation to excess light as double mutants deficient 200

in both mechanisms possess additive phenotypes and are highly sensitive to light (Kukuczka et 201

al., 2014). Another constituent of NPQ is qT, which is related to state transitions. State transitions 202

are important to balance the excitation energy between PSI and PSII (Bonaventura and Myers, 203

1969; Murata, 1969). Under light conditions where PSII is preferentially excited, both PSII core 204

and LHCII proteins become phosphorylated (Lemeille and Rochaix, 2010). As a consequence 205

phosphorylated LHCII proteins detach from PSII and partly connect to PSI (state 2). Under 206

conditions where PSI excitation is predominant, this process is reversed. LHCII proteins are de- 207

phosphorylated and associate with PSII (state 1). The extent of state transition between vascular 208

plants such as A. thaliana and C. reinhardtii differs significantly. The proportion of mobile 209

LHCII antenna is about 80 % in the alga, whereas in A. thaliana only 15-20 % of LHCII are 210

transferred to PSI under state 2 conditions (Lemeille and Rochaix, 2010). However, the large 211

increase in PSI antenna size in C. reinhardtii has recently been challenged (Nagy et al., 2014;

212

Unlu et al., 2014): While 70 to 80 % of mobile LHCII detached from PSII in response to 213

transition to state 2 conditions, only a fraction of about 20 % functionally attached to PSI.

214

Phosphorylation of LHC proteins requires the function of the STT7 kinase or its ortholog 215

STN7 in C. reinhardtii or A. thaliana, respectively. In the absence of the STT7/STN7 kinase, the 216

initiation of state transitions is blocked (Depege et al., 2003; Bellafiore et al., 2005). The mobile 217

LHCII fraction of Chlamydomonas includes the two monomeric LHCII proteins CP26 and CP29 218

(8)

7

and the major LHCII protein LHCBM5 (Takahashi et al., 2006), but also the LHCSR3 protein 219

was suggested to migrate during state transitions (Allorent et al., 2013). Takahashi and co- 220

workers (Takahashi et al., 2014) suggested that only CP29 and LHCBM5 directly associate with 221

PSI to form the PSI-LHCI-LHCII supercomplex, while the binding of CP26 could occur 222

indirectly or via the other two proteins. However, it is not yet known whether STT7 directly 223

phosphorylates the LHCII proteins, or if this takes place as part of a kinase cascade (Rochaix, 224

2007). Nevertheless, the direct interaction between STT7 and the LHCII proteins is quite likely, 225

since none of the other chloroplast kinases was found to be specifically required for LHCII 226

phosphorylation (Rochaix, 2014). The activity of the STT7 kinase is mainly determined by the 227

redox status of the plastoquinone pool (Vener et al., 1997; Zito et al., 1999). The identification of 228

a PP2C-type phosphatase responsible for the dephosphorylation of the LHCII proteins in 229

Arabidopsis has been described by two studies in parallel pointing to the fact that this enzyme 230

called PPH1/TAP38 acts directly on phosphorylated LHCII proteins, in particular when they are 231

associated with the PSI-LHCI supercomplex (Pribil et al., 2010; Shapiguzov et al., 2010).

232

Moreover, it is not known whether these phosphatases are constitutively active or if they are 233

regulated by other means, for example through the redox state of the PQ-pool. Nonetheless, both 234

enzymes are conserved in land plants and exhibit orthologous proteins in C. reinhardtii (Rochaix 235

et al., 2012).

236

Another kinase related to STN7/STT7 is encoded in the Arabidopsis and Chlamydomonas 237

genomes and named STN8 and STL1, respectively. STN8 is involved in PSII core subunit 238

phosphorylation and influences repair of PSII after photodamage (Bonardi et al., 2005; Vainonen 239

et al., 2005). Remarkably, the disassembly of the PSII holocomplex is inhibited in STN7/STN8 240

double mutants (Tikkanen et al., 2008; Fristedt et al., 2009; Dietzel et al., 2011; Nath et al., 241

2013), suggesting that phosphorylation of core subunits is required for PSII disassembly. It was 242

further suggested that STN8 controls the transition between LEF and CEF by phosphorylation of 243

PGRL1 in Arabidopsis (Reiland et al., 2011). As described for STN7, activity of STN8 is 244

probably regulated via the redox state of the PQ pool (Bennett, 1991; Fristedt et al., 2009).

245

Notably, the action of STN8 is counteracted by a chloroplast PP2C phosphatase (PBCP) (Samol 246

et al., 2012), allowing for fast reversibility of STN8 mediated acclimation responses. Thus it 247

appears that an intricate regulatory network of chloroplast protein kinases and phosphatases 248

evolved in vascular plants and algae that drives the acclimation response to various 249

environmental cues including excess and changing light settings (Rochaix et al., 2012). As 250

(9)

8

STN7/STT7 and STN8/STL1 kinase activities appear to be controlled by the redox poise of the 251

PQ pool, the PQ pool would be a central player in these acclimation responses. On the other hand 252

the kinases themselves are subjected to phosphorylation (Reiland et al., 2009; Lemeille et al., 253

2010; Reiland et al., 2011; Wang et al., 2013). However, the functional consequences of this 254

phosphorylation are unknown.

255

Recent comparative analyses revealed the presence of at least 15 distinct chloroplast protein 256

kinases, suggesting an intricate kinase phosphorylation network in the chloroplast (Bayer et al., 257

2012). Generally phosphorylation of proteins is one of the most abundant post-translational 258

modifications (PTM). In complex eukaryotic systems protein phosphorylation occurs most 259

frequently on serine (Ser, pS) followed by threonine (Thr, pT) residues, whereas protein 260

phosphorylation of tyrosine (Tyr, pY) residues (1800:200:1) is comparatively rare (Hunter, 1998;

261

Mann et al., 2002). Protein phosphorylation is a general phenomenon in vivo; it is assumed that 262

about one third of all proteins are phosphorylated at a given time (Cohen, 2000; Ahn and Resing, 263

2001; Venter et al., 2001; Manning et al., 2002; Knight et al., 2003). A recent large-scale 264

quantitative evaluation of human proteomic data strengthened the importance of protein 265

phosphorylation for cellular function and human biology (Wilhelm et al., 2014). Chlamydomonas 266

and Arabidopsis genomes encode large kinase families (Arabidopsis, 2000; Kerk et al., 2002;

267

Merchant et al., 2007), supporting the view that protein phosphorylation also plays an important 268

role in a plant’s life cycle. It is thus evident that the understanding of protein phosphorylation, 269

including the specificity of residues phosphorylated or dephosphorylated in response to cellular 270

as well as environmental factors, is one key to the understanding of the complex functional 271

biological networks at a whole system level. Likewise, it is crucial to design experimental setups 272

allowing the linkage between phosphorylation events and particular physiological consequences 273

to be elucidated.

274

In this regard we designed experiments to investigate STT7 kinase-dependent 275

phosphorylation dynamics in C. reinhardtii in response to high light and anoxia, employing 276

quantitative proteomics in conjunction with in-depth physiological characterization. These 277

conditions are particularly interesting, as high light conditions are known to fully induce 278

LHCSR3 protein expression and qE while anoxia promotes CEF activity. Recently it was 279

demonstrated, that qE and CEF are complementary and crucial in acclimation to these 280

environmental cues (Kukuczka et al., 2014). Notably, LHCSR3 phosphorylation was suggested to 281

depend on STT7 function (Bonente et al., 2011), while CEF-supercomplex formation was found 282

(10)

9

to be independent of STT7 kinase function (Takahashi et al., 2013), indicating that STT7 function 283

might impact acclimation to high light and anoxia in different ways. However, our quantitative 284

proteomics and physiological data reveals that STT7-dependent variations in protein 285

phosphorylation profiles have similar dramatic phenotypic consequences in both conditions, 286

strongly suggesting that regulation of protein phosphorylation is critical for driving successful 287

acclimation to high light and anoxic growth environments.

288 289 290 291

(11)

10 Results

292 293

Acclimation responses of plants towards adverse environmental conditions require and correlate 294

with phosphorylation of chloroplast proteins, implying functional importance of this post- 295

translational modification for a successful survival strategy. The STT7 kinase is an important 296

chloroplast kinase involved in protein phosphorylation events driving acclimation processes. To 297

gain deeper insights into STT7-dependent chloroplast protein phosphorylation in acclimation to 298

high light and anaerobic growth conditions, we analyzed two STT7 mutants (stt7-1 (Depege et 299

al., 2003) and stt7-9 (Cardol et al., 2009)) and wild type for protein phosphorylation by 300

employing phosphoproteomics and quantitative mass spectrometry.

301

(12)

11

To this end, thylakoid membranes were isolated from the three strains under the 302

respective conditions and digested with trypsin (Figure 1, Supplement Tables 1 and 2) (Turkina et 303

al., 2006). Resulting peptides were subjected to SIMAC and using titanium dioxide (TiO2) 304

(13)

12

chromatography for the enrichment of phosphopeptides (Thingholm et al., 2008). In parallel, 305

thylakoid membranes from wild type (14N-labeled) and stt7-1 (15N-labeled) were isolated from 306

cells grown for 24h under high light. These membranes were solubilized with detergent and 307

(14)

13

fractionated by sucrose density centrifugation according to previously published work (Tokutsu 308

and Minagawa, 2013). In total six distinct sucrose fractions were collected (see below, Figure 309

4B). These fractions contained LHCII monomers and trimers, PSII core complexes, PSI-LHCI 310

(15)

14

and PSII-LHCII complexes as well as putative CEF supercomplexes. Sucrose gradient fractions 311

were tryptically digested according to the FASP method (Wisniewski et al., 2009) followed by 312

TiO2 phosphopeptide enrichment. Subsequently, peptides were analyzed by tandem mass 313

(16)

15

spectrometry (MS/MS). For peptide and phosphopeptide identification, OMSSA and X!

314

TANDEM algorithms were used to evaluated the MS/MS data incorporated as pipeline in 315

Proteomatic (Specht et al., 2011). In total, 884 phosphoproteins and 2,434 distinct 316

(17)

16

phosphopeptides were identified from the thylakoid membranes (Supplemental Table 1). To 317

“enhance” identification and quantification, LC-MS/MS runs of enriched (after SIMAC 318

enrichment) and non-enriched peptide samples (after tryptic digestion) were aligned in intensity 319

(18)

17

and retention time domains using piqDB (Hohner et al., 2013; Barth et al., 2014). This 320

database/platform integrated information of peptide identification and quantification qualities into 321

the definition of accurate isotopologue retention time windows (IRTW), thereby allowing high 322

(19)

18

quality alignments to be performed. As a result confident quantifications of all peptides in all LC- 323

MS/MS runs is possible, an approach very similar to that of SuperHirn (Mueller et al., 2007).

324

Subsequent label-free quantitation was performed using pyQms, a tool for MS1-based 325

(20)

19

quantitation (Hohner et al., 2013; Barth et al., 2014). Protein and peptide quantitation and peptide 326

phosphorylation events are listed in Supplemental Table 2. For quantitative arguments, these data 327

will be solely considered in the following. Although these criteria are very stringent, hundreds of 328

peptide phosphorylations can be described for the first time in a quantitative way and can be 329

partially related to STT7 kinase dependent action. Correspondingly, the focus of this work was to 330

dissect the phosphorylation events dependent on STT7 kinase activity.

331

Surprisingly, quantitative data presented in Figure 1A demonstrated that the STT7 kinase 332

is present in strain stt7-9, since an STT7 phosphopeptide, corresponding to a previously 333

recognized STT7 phosphopeptide (Lemeille et al., 2010), could be identified from stt7-9 334

thylakoid membranes. The presence of STT7 in stt7-9 is verified by SDS-PAGE and immunoblot 335

analyses using STT7-specific antibodies, showing that stt7-9 accumulates as a ~100 kDa fusion 336

protein of STT7 and, probably, ARG7, which can be explained by the in-frame insertion in the C- 337

terminal domain of the arg7 cassette used for genetic disruptions of the stt7 gene (Depege et al., 338

2003) (Supplemental Figure 1). Thus stt7-9 can be considered a leaky STT7 mutant. Due to the 339

presence of the phosphopeptide DAGLA(p)S533MEEAILK and peptide 340

VNQGALTEAQLMEELGLQEPAPVAPR we conclude that the kinase domain in stt7-9 possibly 341

intact while the C-terminus with yet unknown function might be absent. In contrast, no STT7 342

peptides could be identified from stt7-1 neither from thylakoids (Figure 1A) nor in sucrose 343

gradient fractions (Supplemental Table 2), in line with the fact that no STT7 protein could be 344

observed in the immunoblot assay (Supplemental Figure 1). Taking advantage of the label free 345

quantitation event of peptide 16 (VNQGALTEAQLMEELGLQEPAPVAPR) present in wild type 346

and stt7-9 thylakoids from high light conditions (Supplemental Table 2), we estimate that the 347

STT7 kinase is about 6-fold more abundant in wild type as compared to stt7-9. Notably, several 348

proteins in this dataset show STT7-dependent phosphorylation (Supplemental Table 1, 2 and 3).

349

In many cases proteins that lacked phosphorylation in stt7-1 were phosphorylated in stt7-9. In 350

more rare cases protein phosphorylation was present in wild type but absent in both stt7-1 and 351

stt7-9.

352

Here we identified and quantified for the first time LHCSR3 peptides that were 353

phosphorylated in an STT7 dependent fashion. Quantification results for LHCSR3 demonstrated 354

the lack of phosphorylation of S26, S28, T32 and T33 in the stt7-1 mutant (Figure 1B), whereas 355

phosphorylation was observed in stt7-9 (phosphopeptide enriched fraction) and wild type (non- 356

enriched and SIMAC enriched fractions). The absence of phosphorylation in LHCSR3 from stt7- 357

(21)

20

1 was confirmed by analyses of the sucrose gradient fractions. More importantly, it became 358

evident that non-phosphorylated peptides harbouring S26, S28, T32 and T33 were more abundant in 359

the mutant compared to the wild type (see peptide 22, 23, 29 and 30). At the same time the 360

average abundance of peptides, which do not contain known phosphorylation sites and which 361

were quantified in both strains is slightly higher in the wild type (see peptide 4, 5, 11 and 16).

362

Consequently, the absence of detectable phosphopeptides in the mutant is not due to a lack of 363

LHCSR3 expression. Moreover, the differences between wild type and mutant with respect to 364

abundance of the non-phosphorylated peptides allowed the estimation of the degree of 365

phosphorylation. Accordingly we could detect 60 % more non-phosphorylated peptides 366

harbouring the phosphorylatible residues S26, S28, T32 and T33 for the mutant compared to the wild 367

type, likely reflecting the fraction of wild type-specific phosphorylated peptides. While the 368

STT7-dependent phosphorylation was observed in the N-terminal part of the protein, there was a 369

STT7-independent phosphorylation found in a peptide closer to the C-terminal that is not 370

proteotypic and shared between LHCSR1 and LHCSR3. Thus it remains unclear whether this 371

phosphorylation is present in only one LHCSR protein or in both but it demonstrates that a 372

second chloroplast kinase, other than STT7, is phosphorylating LHCSR protein(s) at their C- 373

terminal part. Besides the above mentioned non-proteotypic phosphorylation site, LHCSR1 was 374

not found to be phosphorylated at other proteotypic sites.

375

Other noteworthy proteins that showed STT7 dependent phosphorylation as revealed by 376

MS/MS analyses of both, thylakoid and sucrose density fractions, were PETO, LHCBM5 (Figure 377

1C, D) and TEF23 (Supplemental Table 2, 3). Phosphorylation of LHCBM5 has been reported 378

before (Lemeille et al., 2010). Interestingly, phosphorylation events of PETO and TEF23 were 379

also detected in stt7-9 whereas LHCBM5 phosphorylation could neither be detected in stt7-1 nor 380

in stt7-9. For Cre12.g540700.t1.2, a protein predicted to localize to the chloroplast by PredAlgo 381

(Tardif et al., 2012), phosphorylated peptides could be detected in the wild type and stt7-9 but not 382

in stt7-1 in non-enriched high light thylakoid samples, while phosphorylated peptides could be 383

observed in stt7-1 after SIMAC based enrichment, indicating that the amount of phosphorylation 384

of Cre12.g540700.t1.2 was drastically reduced in the stt7-1 mutant particularly under high light 385

growth conditions. This is supported by the fact that only one phosphorylation quantitation event 386

(peptide 5) could be detected for Cre12.g540700.t1.2 after TiO2 based phosphopeptide 387

enrichment in sucrose density fractions stemming from high light thylakoids of stt7-1 in contrast 388

to the wild type, where four phosphorylation events were recognized. In the case of STL1 (Figure 389

(22)

21

2), LHCA6 and PGRL1 (Supplemental Table 2, 3), no phosphorylation was seen under high light 390

conditions in stt7-1 (thylakoids and sucrose density fractions) in contrast to the wildtype.

391

However, phosphorylation was present in stt7-1 under anaerobic conditions as observed in the 392

wild type, implying that phosphorylation of these proteins is independent of STT7 under anoxia.

393

On the other hand, STT7 is involved in phosphorylation under high light settings. Also PSAF 394

was not phosphorylated in high light thylakoids from stt7-1, while it was in wild type. However 395

under anaerobic conditions phosphorylation of PSAF was enhanced in stt7-1 in respect to wild 396

type. The absence of phosphorylation for STL1 and PGRL1 under high light is confirmed by the 397

MS/MS analyses of the sucrose fractions (Figure 2, Supplemental Table 2, 3). In contrast, PSAF 398

was found phosphorylated in sucrose density fractions of stt7-1. Subunit IV of the cyt b6f 399

complex appeared to be phosphorylated in a STT7 dependent manner. In thylakoids from high 400

light conditions, its phosphorylation was three-times less abundant in stt7-1 as compared to wild 401

type, while it was missing in the mutant under anaerobic conditions. After sucrose density 402

centrifugation of detergent solubilized thylakoids treated with high light, no phosphorylation of 403

subunit IV was detectable in stt7-1 but present in wild type, confirming the STT7-dependent 404

phosphorylation. Other proteins, that were phosphorylated in a STT7-dependent way 405

(Supplemental Table 2, 3) and only found in phosphopeptide enriched and/or non-enriched 406

(23)

22

fractions from isolated thylakoids are CDS1 (Cre12.g561550.t1.1), HSP70B 407

(Cre06.g250100.t1.1), Cre17.g737450.t1.2, Cre06.g284100.t1.1, Cre09.g398400.t1.1 and 408

Cre17.g724150.t1.1. Cre01.g071450.t1.2 shares similarity with STN8 and STN7 kinases from 409

Arabidopsis. This protein was found phosphorylated in enriched fractions of high light thylakoids 410

from wild type and stt7-9 but not in stt7-1. As observed for STL1, this putative kinase was found 411

phosphorylated in stt7-1 under anaerobic conditions. However, according to PredAlgo (Tardif et 412

al., 2012) prediction, it is not chloroplast-localized.

413

In contrast to proteins showing lack of phosphorylation in the absence of the STT7 kinase, 414

the reaction center building PSII subunit PSBD (D2) displayed increased phosphorylation in the 415

absence of the kinase (Figure 3, Supplemental Table 2). The amounts of phosphorylated PSBD 416

peptides stemming from thylakoids isolated from high light conditions increased in stt7-1 5.6 and 417

4.4 fold as compared to wild type and stt7-9, respectively (peptides 21, 22, 25, 26). At the same 418

time the overall amount of PSBD non-phosphorylated peptides only slightly augmented 1.08 and 419

1.03 fold in stt7-1 in comparison to wild type and stt7-9, respectively. In thylakoids from 420

anaerobic conditions, the differences in amounts of phosphorylated PSBD peptides between stt7- 421

1 versus wild type and stt7-9 were even more pronounced. In this regard, a 73 and 44 fold 422

induction of phosphorylation was observed in stt7-1 while the quantities of non-phosphorylated 423

PSBD increased 1.6 fold in relation to wild type and stt7-9. A similar picture was observed for 424

phosphorylation of PSBC (CP43) (Supplemental Table 2, 3). Phosphopeptides (38, 39) were 2.3 425

and 1.7 fold more abundant in stt7-1 thylakoids isolated from high light conditions in comparison 426

to wild type and stt7-9, respectively. These differences further increased to 10.4 and 2.4 for wild 427

type and stt7-9, respectively, when thylakoids stemming from anaerobic conditions were 428

analyzed. Comparing the amounts of non-phosphorylated peptides of stt7-1 under high light and 429

anaerobic conditions revealed that PSBC was slightly enriched with ratios of 1.2 and 1.07 and 1.1 430

and 1.4 for wild type and stt7-9, respectively. In conclusion, deletion of the STT7 kinase caused a 431

significant increase in PSBD and PSBC phosphorylation, whereas depletion of the kinase in stt7- 432

9 led to only slightly higher PSBD and PSBC phosphorylation as compared to wild type, 433

suggesting that the amount of PSII core phosphorylation is not linked to STT7 in a strict 434

concentration dependent manner but rather to a threshold in STT7 activity. Notably, other core 435

subunits of PSII such as PSBA (D1) and PSBB (CP47) were not found phosphorylated.

436

Beside differences in protein phosphorylation in stt7-1 versus wild type and stt7-9, many 437

proteins were not altered in their phosphorylation status. As examples, it can be affirmed that 438

(24)

23

photosynthetic proteins such as PSAH, PSBR and PSBO showed a similar protein 439

phosphorylation profile in the three strains as already reported for PSAH and PSBR in stt7-1 and 440

wild type (Lemeille et al., 2010). In case of LHCB4 no qualitative changes in phosphorylation 441

events could be detected between the three strains. Moreover, LHCB4 peptide VATSTGTR, 442

which was previously found to be differentially phosphorylated between stt7-1 and wild type 443

(Lemeille et al., 2010), was detected to be phosphorylated in stt7-1 (Table 1, Supplemental Table 444

1), demonstrating that this peptide can be phosphorylated in an STT7 independent manner.

445

However quantitative changes in LHCB4 phosphorylation were observed. Phosphopeptides 20 446

and 21 (NNKGS103VEAIVQATPDEVSSENR, Charge 2 and 3, respectively) of LHCB4 in stt7-1 were 447

diminished three-fold in regard to wild type and stt7-9, whereas phosphopeptides 5 and 6 and the 448

overall amount of non-phosphorylated peptides in stt7-1 were slightly more abundant as 449

compared to wild type and stt7-9. MS/MS analysis of non-enriched samples derived from FASP- 450

digested sucrose density gradient fractions showed that the same phosphopeptide (18 and 19 451

corresponding to phosphopeptides 20 and 21 found in thylakoids) was present in wild type 452

(monomeric LHCII) but not in stt7-1, therefore underpinning the notion that this phosphopeptide 453

is diminished in stt7-1 under high light growth conditions (Supplemental Table 2).

454

(25)

24

In the quantitative analyses of STT7 dependent phosphorylation, the comparison of 455

phosphopeptides from thylakoids and sucrose density fractionated thylakoids determined 456

differences independently. However, the sucrose density fractionation of detergent solubilized 457

thylakoids offered another advantage, namely the possibility to analyze co-migration of proteins 458

with known multi-protein complexes such as PSII and PSI, the cyt b6f complex, the ATPase or 459

the CEF-supercomplex. Thus the aim of these experiments was to determine whether STT7 460

deficiency alters the dynamic association of proteins in the thylakoid membrane. To define 461

amounts of PSII, PSI, the cytochrome b6f complex and the ATPase in thylakoids isolated after 24 462

h high light treatment, SDS-PAGE fractionation of wild type and stt7-1 thylakoids and 463

immunoblotting experiments were performed (Figure 4A). Using anti-ATPB antibodies 464

demonstrated that ATPB, a subunit of the chloroplast ATPase, was similar in amounts in wild 465

type and stt7-1, while the use of anti-PSBA and anti-CYTF antibodies showed that the amounts 466

of PSBA and CYTF are increased in stt7-1. In contrast, LHCSR3, PSAD and FNR were slightly 467

decreased in stt7-1 as revealed with specific antibodies directed against these proteins. Overall 468

these differences were also reflected in quantitative mass spectrometric data stemming from 469

isotopic labeling experiments (see below, Figure 5). In a next step, thylakoid membranes isolated 470

from wild type and stt7-1 cells grown in high light were solubilized with detergent and 471

fractionated by sucrose density centrifugation according to previously published work (Tokutsu 472

and Minagawa, 2013). Figure 4B shows the corresponding sucrose gradients after ultra- 473

centrifugation. Six green bands could be distinguished. These green bands corresponded, as 474

revealed by SDS-PAGE fractionation and immunoblotting experiments (Figure 4 C, D), to LHCII 475

monomers, LHC trimers, PSII core complexes, PSI-LHCI, PSII-LHCII complexes and putative 476

CEF supercomplexes. The data clearly showed that the intensity of the LHCII monomer and the 477

PSII core band increased in stt7-1 versus wild type. On the other hand, PSI-LHCI and PSII- 478

LHCII supercomplexes were diminished in stt7-1. The immunoblotting experiments after SDS- 479

PAGE fractionation of the sucrose gradients fractions using various specific antibodies revealed 480

some very interesting results. The amounts of PSBA in PSII-LHCII- (fractions 9-11) and PSAD 481

in PSI-LHCI- supercomplexes (fractions 11-13) were slightly diminished in stt7-1 compared to 482

the wild type, as seen in the band pattern of the sucrose gradients. Fraction five in wild type and 483

stt7-1 corresponded to the peak fraction of a putative CEF supercomplex, where PSAD, ANR1, 484

CYTF and FNR co-migrated as previously reported (Iwai et al., 2010; Terashima et al., 2012;

485

Takahashi et al., 2013). However, the CYTF signal peaked in stt7-1 in fraction 7 and was not 486

(26)

25

detectable for wild type. Thus we designated the putative protein complex in fraction 5, present in 487

wild type and stt7-1, as PSI-LHCI-FNR supercomplex as described previously (Takahashi et al., 488

2014). Moreover, the intensity of the LHCSR3 signal in fraction 5 was also stronger in stt7-1 489

(27)

26

than in wild type. The fact that LHCSR3 co-migrated with this PSI and FNR containing 490

supercomplex has not been described before. Association of LHCSR3 with PSI-LHCI in state 2 491

and high light had been described instead (Allorent et al., 2013; Xue et al., 2015). Notably, the 492

amounts of LHCSR3 in the LHCII monomeric and trimeric fractions in stt7-1 decreased in 493

comparison to wild type (Figure 4D). Thus the increase of LHCSR3 in the PSI-LHCI-FNR- 494

supercomplex fraction of stt7-1 is likely due to a change in association caused by a lack in STT7 495

rather than an overall increase in LHCSR3 abundance. This result is further underpinned when 496

considering the LHCSR3 amount in stt7-1 thylakoids (Figure 4A). On the other hand, the 497

amounts of CYTF significantly increased in fractions 4-16 in stt7-1 as compared to the 498

corresponding fractions in wild type.

499

To independently assess the amounts of thylakoid membrane proteins in the distinct 500

protein complexes in stt7-1 and wild type we performed quantitative mass spectrometric analyses 501

based on isotopic labeling. To this end, thylakoid membranes from 14N-labeled wild type and 502

15N-labeled stt7-1 cells grown for 24h in high light were isolated and subjected to detergent- 503

mediated solubilization and sucrose density gradient centrifugation and investigated by mass 504

spectrometry. The data confirmed that CYTF was more abundant in stt7-1 than in wild type in all 505

sucrose fractions analyzed (Figure 5, Supplemental Table 2). Moreover, FNR was enriched in 506

stt7-1 PSI-LHCI-FNR supercomplex fractions. PSAD amounts were found to be diminished in 507

stt7-1, in agreement with immunoblot data shown in Figure 4C. Furthermore, these data showed 508

that LHCSR3 was not only significantly more abundant in PSI-LHCI-FNR supercomplexes in 509

stt7-1 but also in PSII core complexes. At the same time the abundance of PSBD in stt7-1 was 510

increased considerably in PSII core complexes and decreased in PSII-LHCII supercomplexes.

511

The latter notion is in line with the fact that PSII-LHCII supercomplexes were decreased in the 512

stt7-1 sucrose density band pattern (Figure 4B), in agreement with the observation that PSBD and 513

PSBC phosphorylation was higher in stt7-1 PSII core complexes (Supplemental data Table 2).

514

To provide an experimental prove for a link between LHCSR3 and the PSI-LHCI-FNR 515

supercomplex, we took advantage of a C. reinhardtii His6-tagged PSAA strain (Gulis et al., 2008) 516

and purified the supercomplex from cells grown for 24 h in high light (Figure 6A) using 517

detergent solubilization of isolated thylakoids and sucrose density ultracentrifugation. In a next 518

step, the PSI supercomplex from sucrose density fractions 5-7 were further purified by the tagged 519

PSI-subunit (PSAA) via Ni-NTA chromatography. Fractions 5-7 before and after Ni-NTA 520

purification were separated by SDS-PAGE and analyzed by immunoblotting, employing specific 521

(28)

27

antibodies directed against LHCSR3, PSAD and FNR (Figure 6B). The immunoblotting data 522

revealed that PSAD and LHCSR3 were similarly co-purified while FNR was lost via Ni-NTA 523

separation. Therefore we can conclude that LHCSR3 is associated with PSI in the supercomplex.

524

Now, what is the function of LHCSR3 associated with PSI within the PSI-LHCI-FNR 525

supercomplex? A possible scenario might be that LHCSR3 is involved in photo-protection of PSI 526

within such supercomplexes. As recently reported, PSI from pgr5, pgrl1 and pgrl1/npq4 mutant 527

strains (npq4 is lacking LHCSR3 (Peers et al., 2009)) was susceptible to photoinhibition in high 528

light (Johnson et al., 2014; Kukuczka et al., 2014). To investigate whether the PSI 529

photoinhibition occurred faster in pgrl1/npq4 due to a lack of LHCSR3, we performed a time 530

course experiment under high light conditions. As reported earlier (Kukuczka et al., 2014), the 531

amount of oxidizable P700 was comparable in the wild type, pgrl1, npq4 and pgrl1/npq4 at low- 532

(29)

28

light intensities (Figure 7A). Interestingly after 4 h high light the extent of oxidizable P700 was 533

already further diminished in pgrl1/npq4 as compared to pgrl1. The same was true after 24 h of 534

high light, while the amounts of oxidizable P700 in the wild type and npq4 remained comparable 535

to the low light control. Fractionation of wild type, pgrl1, npq4 and pgrl1/npq4 cells by SDS- 536

PAGE and immunoblotting using anti-PSAD and anti-ATPB antibodies revealed that the amount 537

(30)

29

of PSAD diminished in pgrl1 and pgrl1/npq4 after 4 h and even more after 24 h high light, 538

whereas the amount of ATPB remained similar over the time course (Figure 7B). In line with the 539

observation made for the decrease in amounts of oxidizable P700, the immunoblotting data 540

confirmed that the decrease in PSAD was more pronounced in pgrl1/npq4 than in pgrl1. This 541

supports the notion that LHCSR3 has a functional impact, either direct or indirect, on photo- 542

protection of PSI under conditions where PGRL1 function is absent. The deficiency of PGRL1 543

affects CEF and will alter the redox poise and lumen ΔpH of the chloroplast, implying that PSI is 544

particularly prone to photoinhibition under conditions when CEF is impaired as suggested before 545

(Johnson et al., 2014; Kukuczka et al., 2014). We next investigated PSI photoinhibition under 546

(31)

30

high light conditions in stt7-1. Wild type and stt7-1 cells were shifted from low light to high light 547

conditions and the amount of oxidizable P700 was measured after 1, 4 and 24h high light and 24h 548

low light. The data shown in Figure 8A revealed that the amounts of oxidizable P700 decreased for 549

stt7-1 gradually to about 50% oxidizable P700 after 24h in high light in comparison to cells grown 550

for 24h in low light. SDS-PAGE fractionation of wild type and stt7-1 cells followed by 551

immunoblotting using anti-PSAD- anti-LHCSR3 and anti-ATPB antibodies revealed that 552

amounts of PSAD diminished in stt7-1 after 24 h high light to about 50% of the low light control 553

cells (Figure 8B). In addition the comparison to PSAD amounts between stt7-1 cells harvested 554

after 1 h and after 24 h high light, revealed a 50% reduction of PSAD after 24 h, while no change 555

in abundance was observed for PSAD when comparing wild type cells after 1 and 24 h high light.

556

Thus the conclusion that PSI in stt7-1 is also prone to photoinhibition, although not to the same 557

extent as observed for pgrl1 and pgrl1/npq4, was further supported. Moreover, the immunoblot 558

clearly revealed that in stt7-1 LHCSR3 expression is, already present in low light and faster 559

induced in high light in contrast to wild type (Figure 8B), underpinning the assumption that stt7-1 560

is already stressed under conditions where the wild type is not and possibly explaining why the 561

PSI-LHCI-FNR complex contains more LHCSR3 in stt7-1.

562 563 564 565

(32)

31 Discussion

566 567

In the present work we investigated STT7 related protein phosphorylation dynamics in C.

568

reinhardtii in response to high light and anoxia. We intended to assess STT7-dependent protein 569

phosphorylation under conditions where LHCSR3 and qE were fully induced in comparison to 570

conditions known to promote CEF activity. Our quantitative proteomics data revealed that the 571

protein phosphorylation status in STT7-deficient cells was dependent on the environmental 572

condition examined. Furthermore, the phosphoproteome differed between STT7-deficient cells 573

(stt7-1) and cells with diminished expression of a mutant form of STT7 kinase (stt7-9). These are 574

indications that the intricate chloroplast phosphorylation network is highly sensitive and dynamic 575

towards environmental cues and alterations in STT7 kinase function. Variations in chloroplast 576

protein phosphorylation profiles due to the absence of STT7 kinase caused changes in protein 577

expression, photoinhibition of PSI and resulted in remodeling of photosynthetic complexes and 578

triggering a pronounced association of LHCSR3 with PSI-LHCI-FNR supercomplexes.

579

Moreover, absence of STT7 kinase strongly diminished PSII-LHCII supercomplexes while PSII 580

core complex phosphorylation and formation was enhanced. In conclusion, our study provides 581

strong evidence that regulation of protein phosphorylation is critical for driving successful 582

acclimation to high light and anoxic growth environments. Furthermore, numerous new potential 583

STT7 kinase substrates were identified and quantified.

584 585

Expression of cyt b6f complex subunits and association of LHCSR3 with PSI-LHCI-FNR is 586

enhanced in the absence of STT7 function 587

The formation of CEF-supercomplexes has been described for state 2 and anoxic conditions (Iwai 588

et al., 2010; Terashima et al., 2012; Takahashi et al., 2013). Herein we provide evidence that such 589

supercomplexes might be formed in high light under aerobic phototrophic conditions as CYTF 590

co-migrates with PSI, FNR, and ANR1 in stt7-1 (Figure 4C). However, the peak fractions of 591

CYTF in comparison to ANR1, PSAD and FNR are different and it is currently unknown 592

whether the putative CEF-supercomplex is functional under high light conditions. The co- 593

migration of PSI, FNR, and ANR1 in wild type and stt7-1 points to the formation of a PSI-LHCI- 594

FNR supercomplex in high light as previously described (Takahashi et al., 2014). Notably, 595

expression of cyt b6f complex subunits in stt7-1 was enhanced (Figure 4A, C, Supplemental 596

Table 2, 3). This increased expression was also observed under anoxic conditions (Supplemental 597

(33)

32

Table 2, Supplemental Figure 3). Interestingly, as observed for high light, under anoxic 598

conditions more CYTF and PETO co-migrated with the CEF-supercomplex peak fraction (Figure 599

4C, Supplemental Table 3, Supplemental Figure 2). It is currently unclear how this increase of 600

expression is achieved. As the STT7 kinase is associated with the cyt b6f complex (Lemeille et 601

al., 2009), its absence might impact the assembly-controlled biosynthesis of the complex 602

(Choquet et al., 2001). On the other hand, the increase in complex abundance could be indirectly 603

modulated via e.g. changes in the physiological status of the cells. Depletion in CEF 604

supercomplex associated proteins such as ANR1 or PGRL1 combined with a reduced CEF 605

capacity caused a severe growth defect under high light and/or anoxic conditions (Terashima et 606

al., 2012; Kukuczka et al., 2014), indicating a requirement of efficient CEF for acclimation to 607

these conditions.

608

At the same time a remodeling of PSI-LHCI-FNR supercomplexes occurred by increasing 609

the amount of LHCSR3 associated with PSI within the supercomplex, as revealed by affinity 610

purification using the His6-tagged PSAA strain (Figure 6). The function of LHCSR3 bound to 611

PSI is likely to protect the complex from photoinhibition (Figures 7 and 8). The support for this 612

idea is that a pgrl1/npq4 mutant strain displayed a more pronounced PSI photoinhibition in high 613

light as compared to pgrl1 (herein, (Kukuczka et al., 2014)). However, these data cannot exclude 614

an indirect protective function of LHCSR3 in regard to shielding PSI from photoinhibition under 615

high light stress.

616

Nonetheless these data suggest that non-phosphorylated LHCSR3 was bound 617

preferentially to PSI-LHCI-FNR in stt7-1. In agreement only a small amount of phosphorylated 618

LHCSR3 was found associated with the respective supercomplex in the wild type. In conclusion, 619

the binding of LHCSR3 to PSI-LHCI-FNR is negatively correlated with the activity of the STT7 620

kinase. In this scenario, the inactivation of STT7 dependent phosphorylation via reduction of the 621

luminal STT7 disulfide bridge due to prolonged high light, as outlined by Lemeille and co- 622

workers (Lemeille et al., 2009) would increase the amount of non-phosphorylated LHCSR3, 623

which in turn would increase its binding to PSI-LHCI-FNR, thereby protecting PSI and 624

resembling the situation as in stt7-1.

625 626

Absence of STT7 function causes HL-induced photoinhibition of PSI and increased PSII 627

core phosphorylation and formation 628

(34)

33

Despite the binding of LHCSR3 to PSI, a sensitivity of PSI to photoinhibition was observed in 629

stt7-1. The extent of photoinhibition was not as pronounced as seen for the mutant strains pgrl1 630

and pgr5 (Figure 7, 8 (Johnson et al., 2014; Kukuczka et al., 2014)) but significant. It has been 631

reported that STT7/STN7 function is essential for state transitions in algae and vascular plant, 632

respectively (Depege et al., 2003; Bellafiore et al., 2005), respectively. State transitions are 633

important for an efficient balancing of excitation energy between PSI and PSII (Bonaventura and 634

Myers, 1969; Murata, 1969). They involve the phosphorylation of LHCII proteins by 635

STT7/STN7, which then detach from PSII and in part migrate to PSI (state 2). As reported 636

before, LHCBM5 is phosphorylated in state 2 settings under anaerobic and high light conditions 637

in C. reinhardtii. Moreover, LHCBM5 phosphorylation was shown to be STT7-dependent 638

(Lemeille et al., 2010). Notably, as found for stt7-1 in our study, no LHCBM5 phosphorylation 639

was detectable in stt7-9. Like stt7-1, stt7-9 is unable to perform state transitions (Cardol et al., 640

2009). So far phosphorylation of LHCBM5 has only been detected when the protein was 641

associated with PSI-LHCI complexes (Takahashi et al., 2006). Hence, the inability to 642

phosphorylate LHCBM5 might explain the fact that both mutants are locked in state 1. On the 643

other hand, no increased expression of cyt b6f complex subunits, augmented PSII core formation 644

or enhanced binding of LHCSR3 to PSI-LHCI-FNR was observed in stt7-9 (Supplemental Table 645

2, Supplemental Figure 2). Moreover, phosphorylation profiles of stt7-9 and stt7-1 showed clear 646

differences, likely explaining these phenotypic differences. These findings indicate that triggering 647

the association of LHCII with PSI requires far more active STT7 kinase than triggering the 648

acclimation process, which permits full PSII-LHCII supercomplex formation and circumvents 649

PSI photoinhibition in high light conditions. We cannot exclude, that STT7 from stt7-9 is 650

differently regulated due to its altered protein sequence (see also below). However, the lack of a 651

functional interconnection between PSI and LHCII cannot explain the phenotypic differences 652

between stt7-1 and stt7-9. Requirement of STN7 for effective retrograde signaling has been also 653

described for Arabidopsis, as growth phenotypes and diminished relative amounts of PSI had 654

been found in stn7 Arabidopsis plants grown in fluctuating light conditions (Tikkanen et al., 655

2010). Along the same line, growth phenotypes were also observed in plants treated with 656

fluctuating PSII and PSI lights (Bellafiore et al., 2005).

657

Recently the process of state transitions was revisited in C. reinhardtii (Nagy et al., 2014;

658

Unlu et al., 2014). Interestingly these new data indicated that although 70 to 80 % of LHCII 659

detached from PSII in state 2 conditions, only a fraction of about 20 % was attached to PSI. It 660

(35)

34

was postulated that the disconnected antenna complexes form energetically quenched complexes 661

that are protected against photodamage and exhibit a shortened excited-state fluorescence lifetime 662

(Unlu et al., 2014). In addition it was suggested that phosphorylated LHCII may also remain with 663

PSII but become quenched (Nagy et al., 2014). Importantly, these studies were conducted under 664

low light in photoheterotrophic conditions where expression of LHCSR3 is not expected. Thus 665

the observed quenching should be independent of LHCSR3 and qE. In this regard it is important 666

to note that phosphorylation of LHCB4 phosphopeptides 20 and 21 667

(NNKGS103VEAIVQATPDEVSSENR, Charge 2 and 3, respectively) were diminished two to 668

three-fold in stt7-1 thylakoids isolated from high light conditions in comparison to stt7-9 and 669

wild type (Supplemental Table 2), respectively. However, herein a postulated STT7-dependent 670

LHCB4 phosphorylation (Lemeille et al., 2010) was found to be present in the absence of STT7.

671

In total 12 different phosphorylation sites were detected, five more than described previously 672

(Turkina et al., 2006). Our data indicate that under high light and photoautotrophic conditions, 673

phosphorylation of LHCB4 is changed merely quantitatively by the absence of STT7. LHCB4 is 674

a minor core antenna protein of PSII that is suggested to be a linker between PSII core and 675

LHCBM proteins and likely involved in excitation energy transfer between the trimeric LHCBM 676

proteins and PSII core (Dainese and Bassi, 1991; Dekker and Boekema, 2005; van Amerongen 677

and Croce, 2013) as well as in state transitions in C. reinhardtii (Tokutsu et al., 2009). Moreover 678

it has been suggested that phosphorylation of LHCB4 and LHCB5 is responsible for the 679

detachment of all LHCBM polypeptides from PSII in the transition from state 1 to state 2.

680

Notably, we were unable to detect LHCB5 phosphorylation in the quantitative data. Since 681

changes in LHCB4 phosphorylation status between wild type and stt7-9 were similar, it is 682

possible that the larger reduction in LHCB4 phosphorylation in stt7-1 directly affected unbinding 683

of LHCBM proteins from PSII. Additionally, phosphorylated LHCB4 was not detected bound to 684

PSII core and suggested to be associated with other free LHCBM (Iwai et al., 2008), thereby 685

decreasing the amount and possibly the organization of unbound LHCII polypeptides. It is 686

tempting to speculate that this altered phosphorylation status might affect the formation of 687

energetically quenched LHCII complexes, since differences in LHCII complex formation became 688

apparent as more monomeric LHCII and less trimeric LHCII was observed in stt7-1 as compared 689

to wild type (Figure 4B).

690

Moreover, the amounts of LHCSR3, which co-migrated with the other free LHCII 691

polypeptides, was severely diminished in stt7-1 (Figure 4D) due to enhanced co-migration with 692

(36)

35

PSI in the PSI-LHCI-FNR fraction, which in turn would certainly have an impact on the 693

formation of LHCII complexes, that are energetically quenched by LHCSR3 as it has been shown 694

to quench excited LHC protein bound chlorophyll molecules (Bonente et al., 2011). Therefore a 695

lack in LHCSR3 association with aggregated LHCII complexes would likely promote 696

photooxidative stress, because quenching capacity should be lower.

697

It had been suggested that PSII core subunit phosphorylation functions as a mechanism to 698

protect photodamaged PSII complexes under high light conditions from proteolytic degradation 699

(Koivuniemi et al., 1995; Kruse et al., 1997; Turkina et al., 2006). Additionally, PSII core 700

phosphorylation has been implicated in the regulation of the repair cycle of photodamaged PSII 701

(Aro et al., 1993; Tikkanen et al., 2008; Fristedt et al., 2009; Dietzel et al., 2011; Nath et al., 702

2013) by facilitating the migration of damaged PSII centers from grana to stroma thylakoid 703

where damaged PSBA is replaced with a newly synthesized copy. Phosphorylation of PSBD and 704

PSBC had been also proposed to facilitate dissociation of peripheral light-harvesting antenna 705

(Turkina et al., 2006; Iwai et al., 2008). Our quantitative data clearly showed that PSBD and 706

PSBC were significantly phosphorylated to a higher degree in stt7-1 thylakoids and particularly 707

in PSII core complexes compared to the wild type. This is in line with recent anti-phospho- 708

threonine immunoblot data on isolated stt7-1 and stn7 thylakoids (Fristedt et al., 2010; Lemeille 709

and Rochaix, 2010). Our quantitative data also demonstrated that the formation of PSII core 710

complexes was highly enhanced in stt7-1 and that at the same time the amount of PSII-LHCII 711

supercomplexes was significantly decreased in comparison to wild type. Thus indicating that the 712

absence of STT7 was responsible for this transition and suggesting that PSII-LHCII complexes 713

were prone to photoinhibition in HL-acclimated stt7-1.

714

It is conceivable that strong PSII core phosphorylation in stt7-1, possible due to higher 715

requirement in PSII repair cycle, is linked to enhanced PSII core formation, as it is suggested to 716

trigger dissociation of LHCII proteins from PSII. Like in Arabidopsis (Bonardi et al., 2005;

717

Vainonen et al., 2005), it is proposed that STL1, the C. reinhardtii STN8 ortholog, is required for 718

phosphorylation of PSII core proteins PSBD and PSBC (Rochaix et al., 2012). Notably STL1 719

phosphorylation was absent in stt7-1 under high light but present in anoxic conditions. Yet PSBD 720

and PSBC phosphorylation was detected under high light and anoxic conditions. Thus STL1 721

phosphorylation is not required for PSII core phosphorylation in C. reinhardtii and STL1 722

phosphorylation is not strictly dependent on STT7.

723 724

(37)

36

The intricate chloroplast phosphorylation network is highly dynamic and sensitive towards 725

environmental cues and alterations in kinase function 726

Phosphorylation of STL1, LHCSR3 and PETO, for the latter in particular under anoxia, were 727

present in stt7-9, but reduced in regard to wild type, suggesting that the level of decline 728

corresponds and correlates to the diminishment in amounts of STT7 kinase in stt7-9. On the other 729

hand, no phosphorylation was observed for LHCBM5, another abundant substrate, in stt7-1 and 730

stt7-9. Here it is tempting to speculate that beside the depletion in kinase abundance, the 731

proposed regulatory C-terminal part of STT7, which is missing in stt7-9, is also involved in 732

substrate recognition and phosphorylation efficiency and could also be responsible for changes in 733

phosphorylation detected between stt7-1 and stt7-9.

734

As revealed for STL1, phosphorylation of PGRL1 was absent under high light but present 735

under anoxic conditions in stt7-1, thus pointing to highly dynamic chloroplast protein 736

phosphorylation that is dependent on environmental cues and alterations in kinase function. It has 737

been suggested that STN8 is responsible for PGRL1 phosphorylation (Reiland et al., 2011).

738

Phosphorylation of PGRL1 by STN8 slowed down the transition from CEF to LEF in 739

Arabidopsis, thereby providing a link between phosphorylation and control of CEF activity 740

(Reiland et al., 2011). In C. reinhardtii phosphorylation of PGRL1 was observed under anoxia in 741

stt7-1 but it was not increased compared to wild type or stt7-9. In contrast under the same 742

conditions the capacity of CEF was not diminished in stt7-1 in regard to the two other strains 743

(Supplemental Figure 3), implying that the CEF activity per se is not dependent on PGRL1 744

phosphorylation, as CEF rates and PGRL1 phosphorylation are not proportional. In the same line, 745

CEF activity is also not dependent on PETO phosphorylation, as it is absent in stt7-1 under 746

anoxia. It is tempting to speculate that there is a link between PGRL1and STL1 phosphorylation, 747

since their phosphorylation patterns coincide. Currently it is unknown which other chloroplast 748

kinase could phosphorylate STL1 or STN8. The protein Cre01.g071450.t1.1 shares similarity 749

with STN7/STN8 from Arabidopsis and showed the same phosphorylation dynamics as STL1.

750

However, this protein is probably not chloroplast localized (Tardif et al., 2012). Interestingly 751

phosphorylation of subunit IV of the cyt b6f complex was strongly diminished in stt7-1 under 752

high light but completely missing in anoxia. Yet, this is another example showing that protein 753

phosphorylation is not only kinase-dependent but also related to the respective environmental 754

conditions. TEF23 was the other STT7-dependent substrate that was not phosphorylated under 755

high light or anoxia in stt7-1. TEF23 is an ortholog of EGY1 (ethylene-dependent gravitropism- 756

(38)

37

deficient and yellow-green 1), a zinc metalloprotease required in Arabidopsis and tomato 757

(Solanum lycopersicum) for proper chloroplast development (Chen et al., 2005; Barry et al., 758

2012). Plants that are deficient in the protease possessed reduced grana stacks, poorly developed 759

thylakoid lamella networks and low amounts of chlorophyll a/b binding proteins. Furthermore it 760

has been suggested that EGY1 participates in abscisic acid/reactive oxygen species (ROS)- 761

dependent plastid retrograde signaling networks, in particular with regard to the regulation of 762

ammonium stress (Li et al., 2012; Li et al., 2013). Mutant plants deficient in EGY1 are 763

characterized by diminished H2O2 accumulation in guard cell chloroplasts, likely caused by an 764

impaired turnover and assembly of membrane-associated components of PSI and PSII (Li et al., 765

2012). In Arabidopsis STN7 was proposed to be involved in ROS homoeostasis and ROS-based 766

retrograde signaling through controlling LHCII distribution and the redox balance in the electron 767

transfer chain (Tikkanen et al., 2012). Hence, the STT7 dependency of TEF23 phosphorylation is 768

striking and may reveal that STT7 affects ROS-based retrograde signaling on multiple levels.

769

Beside the control of redox balance, STT7 might also be involved in the regulation of transport 770

processes in the chloroplast and/or the cell. Accordingly, another STT7 substrate is the protein 771

Cre13.g571500.t1.1 that is phosphorylated in wild type and stt7-9 but not in stt7-1. This protein 772

shows similarity to various transporters and permeases (Merchant et al., 2007), though its 773

location is unknown. Other transport proteins that were found phosphorylated in wild type and 774

stt7-9 are CDS1 (Cre12.g561550.t1.1), RHP1 (Cre06.g284100.t1.1) and TRP5 775

(Cre09.g398400.t1.1). In stt7-1, however, there was no evidence for protein phosphorylation nor 776

presence of non-phosphorylated peptides, indicating that either phosphorylation was absent 777

and/or protein expression vastly diminished. CDS1, a half-size ABC transporter with similarity to 778

Arabidopsis ATM3, suggested to be involved cadmium resistance (Kim et al., 2006) and iron- 779

homeostasis (Chen et al., 2007) is supposedly mitochondrion-localized, although proteomic data 780

also suggest chloroplast localization in Chlamydomonas (Terashima et al., 2011). RHP1 on the 781

other hand is a CO2 channel (Soupene et al., 2002). TRP5 is a member of the transient receptor 782

potential (TRP) multigene superfamily encoding integral membrane proteins that function as ion 783

channels. Notably, this multigene family of channels is absent in vascular plants and represented 784

by 14 potential TRP genes in C. reinhardtii (Fujiu et al., 2011). Currently it is unknown whether 785

TRP5 is chloroplast-localized. However, the absence of STT7-dependent phosphorylation 786

suggests direct or indirect control of TRP5 function and/or presence via STT7. The same holds 787

for the two other discussed transport proteins, since indirect impact cannot be excluded. As TRP 788

Références

Documents relatifs

Rongé par le doute, détruit par l’alcool, il ne savait plus ce qu’il avait fait la nuit précédant la mort de Marco, puisqu’il était ivre comme tous les soirs depuis que

Of these five cameras two view the divertor region and are used to obtain two-dimensional emission profiles, two view ICRF antennas and are used to determine impurity

Shillman, "Low Error Rate Optical Character Recognition of Unconstrained Handprinted Letters Based on a Model of Human Perception" (to appear in IEEE

Publications of the Division of Building Research may be obtained by mailing the appropriate remittance, (a Bank, Express, or Post Office Money Order or a cheque made

Direction des renseignements militaires.. ﺖــﺳرﺎﻣ ةدﺎــﺑإ بﺮــﺣ ﺔﻴــﺴﻧﺮﻔﻟا تﺎﻄﻠــﺴﻟا ﺔــﻤﻈﻨﻣ ﺎــﻫﺮﻫﺎﻈﻣ ﺖــﻠﲡ يﺮــﺋاﺰﳉا ﺐﻌــﺸﻟا ﺪــﺿ ﻞــﺘﻘﻟا ﰲ

The persistence of SEK and UST was assessed accounting for the potential confounding factors by using propensity score matching method (propensity score was used to

As expected from our data using GPC3 as target, trans- fection of HuH7 cells with miR-96 led to a slight decrease of FOXO1 protein, whereas transfection with miR-182 had no