Embed Size (px)
Citation preview
Subject: PP2016-01546D: Decision LetterFrom: [email protected]: 2016/10/13 23:11To: [email protected]
13‐Oct‐2016
Prof. Takehito InabaUniversity of MiyazakiDepartment of Agricultural and Environmental Sciences, Faculty of AgricultureMiyazaki 8892192Japan
RE: Ubiqui n‐proteasome dependent regula on of GLK1 in response to plas d signals in Arabidopsis
Dear Dr. Takehito Inaba:
We are pleased to accept your manuscript for publica on in the Plant Physiology. This acceptance iscon ngent on revision based on the comments below. In par cular, please consider the following:
You have convincingly revised the manuscript according to the referees' comments. Two points s ll needa en on: You addressed the issue of significant which should only be used in conjunc on with asta s cal test. This is good. But please indicate sta s cal significance of difference in all figures. Second Iask you to explain GLK1 in the tle and in the one sentence summary.
To submit your revised manuscript, click:
h p://pphys.msubmit.net/cgi‐bin/main.plex?el=A5It5Clr5A4DbC1I2A9 dk2aho3EXqDh91TNBBLnAQZ
If you cannot return the revised manuscript within 8 weeks of receipt, please let us know. Otherwise, wewill assume that you have elected not to revise the manuscript and are withdrawing it.
Thank you for allowing us to review your work. We look forward to hearing from you soon.
Sincerely,
Karl‐Josef DietzMonitoring Editor, Plant Physiology
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ Reviewer comments:
PP2016-01546D: Decision Letter imap://mail.cc.miyazaki-u.ac.jp:993/fetch>UID>.INBOX>42406?hea...
1 / 1 2016/10/14 9:37
1
1
Running Head: Proteasome-dependent regulation of GLK1 protein 2
3
Corresponding Author: Takehito Inaba 4
Department of Agricultural and Environmental Sciences, Faculty of Agriculture, 5
University of Miyazaki, Miyazaki 889-2192, Japan 6
Tel/Fax: +81-985-58-7899 7
E-mail: [email protected] 8
9
Research Area: Cell Biology 10
11
12
13
14
15
2
Ubiquitin-proteasome dependent regulation of GLK1 in 16
response to plastid signals in Arabidopsis 17
18
Mitsuaki Tokumaru1,5, Fumi Adachi1,5, Makoto Toda2, Yasuko Ito-Inaba1,3, Fumiko 19
Yazu1, Yoshihiro Hirosawa1, Yoichi Sakakibara2, Masahito Suiko2, Tomohiro 20
Kakizaki4, and Takehito Inaba1,5* 21
22
1Department of Agricultural and Environmental Sciences, Faculty of Agriculture, 23
University of Miyazaki, Miyazaki 889-2192, Japan 24
2Department of Biochemistry and Applied Biosciences, Faculty of Agriculture, University 25
of Miyazaki, Miyazaki 889-2192, Japan 26
3Organization for Promotion of Tenure Track, University of Miyazaki, Miyazaki 889-2192, 27
Japan 28
4Institute of Vegetable and Floriculture Science, NARO, 360 Kusawa, Ano, Tsu, Mie 29
514-2392, Japan 30
5These authors contributed equally to this work. 31
32
Summary of important findings 33
Accumulation of the transcription factor AtGLK1 in response to plastid signals is 34
regulated by the ubiquitin-proteasome system. 35
36
3
Footnotes 37
Author Contributions 38
M.T., F.A., M.T., Y.I.I., F.Y., Y.H., T.K. and T.I. performed experiments; Y.I.I., Y.S., M.S., 39
T.K. and T.I. designed the experiments and analyzed the data; T.I. conceived the project 40
and wrote the manuscript with assistance of Y.I.I. and T.K. 41
42
Funding 43
This work was supported by the Program to Disseminate Tenure Tracking System from 44
the Japanese Ministry of Education, Culture, Sports, Science, and Technology (MEXT), 45
a grant for Scientific Research on Priority Areas from the University of Miyazaki (to T.I. 46
and Y.I.I.), the Inamori Foundation (to Y.I.I.), the Strategic Young Researcher Overseas 47
Visits Program for Accelerating Brain Circulation (to Y.S., M.S. and T.I.), and 48
Grant-in-Aid for Young Scientists (25850073 to T.I. and 26850065 to Y.I.I.) and 49
Grant-in-Aid for Scientific Research (15K07843 to T.I.) from MEXT. 50
51
*Corresponding author: Takehito Inaba 52
E-mail: [email protected] 53
54
55
4
Abstract 56
Arabidopsis thaliana GOLDEN2-LIKE (GLK) transcription factors promote chloroplast 57
biogenesis by regulating the expression of photosynthesis-related genes. Arabidopsis 58
GLK1 is also known to participate in retrograde signaling from chloroplasts to the 59
nucleus. To elucidate the mechanism by which GLK1 is regulated in response to plastid 60
signals, we biochemically characterized Arabidopsis GLK1 protein. Expression analysis 61
of GLK1 protein indicated that GLK1 accumulates in aerial tissues. Both tissue-specific 62
and sucrose-dependent accumulation of GLK1 were regulated primarily at the 63
transcriptional level. In contrast, norflurazon- or lincomycin-treated gun1-101 mutant 64
expressing normal levels of GLK1 mRNA failed to accumulate GLK1 protein, suggesting 65
that plastid signals directly regulate the accumulation of GLK1 protein in a 66
GUN1-independent manner. Treatment of the glk1glk2 mutant expressing functional 67
GFP-GLK1 with a proteasome inhibitor, MG-132, induced the accumulation of 68
polyubiquitinated GFP-GLK1. Furthermore, the level of endogenous GLK1 in plants with 69
damaged plastids was partially restored when those plants were treated with MG-132. 70
Collectively, these data indicate that the ubiquitin-proteasome system participates in the 71
degradation of Arabidopsis GLK1 in response to plastid signals. 72
73
5
Introduction 74
Plant cells have evolved complex mechanisms to coordinate nuclear gene 75
expression depending on the functional state of plastids. Nuclear-encoded plastid 76
proteins are required for plastid biogenesis (Inaba and Schnell, 2008). Furthermore, 77
light-absorbing plastids are subjected to oxidative stress, requiring plants to finely tune 78
their cellular activities to adapt to environmental conditions (Szechynska-Hebda and 79
Karpinski, 2013). Thus, retrograde signals from plastids to the nucleus, referred to as 80
plastid signals, have been proposed to play key roles in coordinating nuclear gene 81
expression with the functional and metabolic states of plastids (Inaba et al., 2011; Chi et 82
al., 2013; Jarvis and Lopez-Juez, 2013). 83
Regulation of nuclear gene expression by plastids is divided into two 84
mechanisms: biogenic and operational control (Pogson et al., 2008). Biogenic control is 85
attributed to the regulation of genes necessary for the construction of the photosynthetic 86
apparatus. This mechanism is critical for proper assembly of the photosynthetic 87
apparatus and chloroplast biogenesis (Pogson et al., 2008; Inaba et al., 2011; Chi et al., 88
2013; Jarvis and Lopez-Juez, 2013). In contrast, operational control enables plastids to 89
regulate the expression of nuclear genes in response to environmental cues, enabling 90
plants to optimize photosynthetic performance. To date, several molecules, including 91
reactive oxygen species (ROS) (Karpinski et al., 1999; Wagner et al., 2004), 92
methylerythritol cyclodiphosphate (Xiao et al., 2012), and 3′-phosphoadenosine-5′93
-phosphate (Estavillo et al., 2011; Chan et al., 2016), have been shown to participate in 94
operational control. 95
Transcriptional activator GOLDEN2-LIKE (GLK) proteins play key roles in 96
biogenic control of nuclear gene expression by plastid signals (Jarvis and Lopez-Juez, 97
6
2013). The GLK genes positively regulates the expression of photosynthesis-related 98
genes in numerous plants (Yasumura et al., 2005; Waters et al., 2009). In Arabidopsis, 99
there are two copies of GLK genes, designated as GLK1 and GLK2, which function 100
redundantly to regulate chloroplast biogenesis. The glk1glk2 double mutant exhibits a 101
pale-green phenotype (Fitter et al., 2002). Furthermore, overexpression of GLK has 102
been shown to be sufficient to induce chloroplast development in rice calli (Nakamura et 103
al., 2009) and Arabidopsis root cells (Kobayashi et al., 2012). When Arabidopsis plants 104
are subjected to treatments that induce plastid signals, expression of GLK1 is 105
suppressed (Kakizaki et al., 2009; Waters et al., 2009; Kakizaki et al., 2012). GLK genes 106
appear to regulate chloroplast biogenesis positively and are involved in biogenic control; 107
however, to date, the biochemical nature of GLK1 protein has not been characterized. 108
Chimeric GLK genes fused to green fluorescent protein (GFP) and introduced into a 109
glk1glk2 double mutant complemented a pale-green phenotype (Waters et al., 2008), 110
but chimeric proteins have not been detected by fluorescence microscopy or 111
immunoblotting. This may be likely because GLK proteins are highly unstable, or 112
because the level of GLK proteins is strictly regulated in vivo. 113
Transcription factors involved in plastid-to-nucleus signaling are regulated by 114
multiple mechanisms (Chi et al., 2013). As stated above, the expression of GLK1 has 115
been shown to respond to treatments that induce plastid signals (Kakizaki et al., 2009). 116
In contrast, post-translational activation of another transcription factor, ABSCISIC ACID 117
INSENSITIVE 4 (ABI4), prevents the binding of G-box binding factors (GBFs) to the 118
light-harvesting chlorophyll a/b-binding protein (LHCB) promoter upon inhibition of 119
plastid biogenesis, thereby suppressing expression of LHCB in the nucleus 120
(Koussevitzky et al., 2007). The activation of ABI4 involves a plant homeodomain 121
7
transcription factor with transmembrane domains (PTM), which localizes to the nucleus 122
and chloroplasts. When plastids are subjected to stress, the N-terminus of PTM is 123
cleaved by proteolysis and moves into the nucleus, thereby activating transcription of 124
ABI4 and allowing plant cells to suppress photosynthesis-related genes (Sun et al., 125
2011; Chi et al., 2013). Hence, regulation of transcription factors at both transcriptional 126
and post-translational levels is important in plastid-to-nucleus retrograde signaling. 127
In this study, we demonstrate that ubiquitin-proteasome dependent 128
post-translational regulation plays a key role in the accumulation of GLK1 protein in 129
response to plastid signals. We raised antibodies against GLK1 and successfully 130
detected GLK1 protein. The level of GLK1 protein was decreased by treatments that 131
induce plastid damage, regardless of the level of GLK1 mRNA. Furthermore, this 132
decrease of GLK1 was attenuated by treatment with a proteasome inhibitor, MG-132. 133
Our results show that plastid signals down-regulate the accumulation of GLK1 through 134
the ubiquitin-proteasome pathway. 135
136
137
8
Results 138
Production of specific antibodies against GLK1 protein 139
Both genetic and transgenic studies have demonstrated that GLK1 participates 140
in the induction of photosynthesis-related genes and plastid-to-nucleus signaling 141
(Kakizaki et al., 2009; Waters et al., 2009). However, to date, stable, high-yield 142
purification of GLK1 has been unsuccessful and has prevented biochemical 143
characterization of the protein. To investigate the mechanism by which GLK1 protein 144
accumulation is regulated, we first purified NusA-TEV-GLK1-His fusion protein 145
expressed in Escherichia coli (E. coli) (Supplemental Fig. S1) and raised antibodies 146
against GLK1. As shown in Fig. 1, we confirmed the specificity of our antibodies to 147
GLK1 using glk1, glk2 and glk1glk2 mutants, and a GLK1ox line overexpressing GLK1 148
(Fig. 1). The antibodies detected a ~60-kDa protein in WT, glk2, and GLK1ox plants (Fig. 149
1). The apparent molecular mass of this band was slightly higher than the predicted 150
molecular mass of GLK1 (~47 kDa), and most abundant in GLK1ox plants (Fig. 1). 151
Further, this band was not detected in either glk1 or glk1glk2 mutants. Hence, we 152
concluded that the ~60-kDa band is indeed GLK1 and that the antibodies we raised can 153
detect endogenous GLK1. 154
155
Levels of GLK1 protein are attributable to tissue-specific expression of GLK1 156
gene 157
A previous study showed that GLK1 is expressed in aerial tissues but not in 158
roots (Fitter et al., 2002). This finding is consistent with the fact that GLK1 participates in 159
the coordinated expression of photosynthesis-related genes (Waters et al., 2009). To 160
determine whether mRNA transcript levels are correlated with GLK1 protein 161
9
accumulation, we examined tissue-specific accumulation of GLK1 using anti-GLK1 162
antibodies. As shown in Fig. 2, GLK1 accumulates in aerial tissues but not in roots. 163
When signal intensities of GLK1 were normalized to those of actin, it was apparent that 164
GLK1 protein was most abundant in leaf tissue (Fig. 2B). This distribution of GLK1 is 165
consistent with GLK1 mRNA accumulation (Fig. 2C). These data indicate that 166
tissue-specific accumulation of GLK1 protein is primarily regulated at the transcriptional 167
level. 168
169
Sucrose up-regulates the expression of GLK1 in a GUN1-independent manner 170
During the detection of GLK1 protein in each tissue, we noted that the level of 171
GLK1 in rosette leaves was lower for plants grown in soil than on MS plates. Because 172
we routinely supply sucrose to MS plates, we suspected that sucrose may up-regulate 173
the expression of GLK1. Consistent with this idea, it has been shown that carbohydrate 174
metabolism in chloroplasts affects the expression of GLK1 in the nucleus (Paparelli et 175
al., 2012). Hence, we tested the effect of sucrose on the accumulation of GLK1. We 176
extracted crude proteins from wild-type plants grown on MS plates containing different 177
concentrations of sucrose. To control for osmotic pressure, we also included plants 178
grown on mannitol. Wild-type plants accumulated only small amounts of GLK1 in the 179
10
absence of sucrose (Fig. 3). However, addition of sucrose increased the accumulation 180
of GLK1 (Fig. 3A). This increase was dose dependent, as plants grown on 2% sucrose 181
11
accumulated more GLK1 protein. The accumulation of GLK1 protein was consistent 182
with that of GLK1 mRNA (Fig. 3B). In contrast, the expression of LHCB1 did not respond 183
12
to sucrose dramatically in our experimental conditions (Fig. 3C). These data indicate 184
that sucrose regulates the expression of GLK1 mRNA, thereby affecting the 185
accumulation of GLK1 protein. 186
A previous study suggested that plastid-localized PPR protein, GUN1, 187
participates in the sucrose-dependent regulation of anthocyanin accumulation (Cottage 188
et al., 2010). Furthermore, our previous study indicated that GUN1 regulates expression 189
of GLK1 in the nucleus when plastids are damaged (Kakizaki et al., 2009). Therefore, 190
we also tested if GUN1 regulates GLK1 expression in a sucrose dependent manner. 191
Compared to the wild type, the gun1-101 mutant accumulated slightly higher levels of 192
GLK1 in the absence of sucrose (Fig. 3B). However, the gun1-101 mutation did not 193
affect sucrose-dependent induction of GLK1 expression (Fig. 3B). GLK1 transcript 194
levels correlate with GLK1 protein levels in gun1-101 mutants (Fig. 3A and 3B). These 195
results suggest that sucrose regulates the expression of GLK1 in a GUN1-independent 196
manner. 197
198
Damaged plastids directly regulate the accumulation of GLK1 in a 199
GUN1-independent manner 200
We next investigated if plastid signals also regulate GLK1 protein accumulation 201
at the post-transcriptional level. Previous studies demonstrated that expression of GLK1 202
is significantly down-regulated when plastid biogenesis is inhibited (Kakizaki et al., 203
2009; Waters et al., 2009). This regulation is GUN1-dependent, as the gun1-101 204
mutation has been shown to abolish the down-regulation of GLK1 (Kakizaki et al., 2009). 205
We examined if the level of GLK1 mRNA correlates with the amount of GLK1 protein 206
when plastids are damaged (Fig. 4). When wild-type Arabidopsis is treated with 207
13
norflurazon (NF), the accumulation of GLK1 decreased dramatically (Fig. 4A and 4B). 208
This is in part attributable to the fact that the expression of GLK1 is down-regulated in 209
14
NF-treated wild-type plants (Fig. 4C). In contrast, expression of GLK1 is active when the 210
gun1-101 mutant was treated with NF (Fig. 4C). Intriguingly, the gun1-101 mutant failed 211
to accumulate GLK1 protein in the presence of NF even though GLK1 mRNA is fully 212
expressed (Fig. 4A and 4B). We also investigated the effects of lincomycin treatment on 213
the accumulation of GLK1 protein. When wild-type and gun1-101 plants grown in 214
submerged culture system were treated with lincomycin for 5 days, the accumulation of 215
GLK1 was dramatically decreased in both plants (Fig. 4D and 4E). Similar to NF-treated 216
wild-type plants, the effects of lincomycin on the expression of GLK1 mRNA was 217
moderate compared to those on the accumulation of GLK1 protein (Fig. 4D-4F). In 218
contrast, the GLK1 mRNA was fully expressed in gun1-101 mutant, suggesting that 219
inhibitor-treated gun1-101 mutant down-regulates the level of GLK1 protein other than 220
the transcriptional regulation of GLK1 mRNA. 221
We also examined if damaged plastids caused by the plastid protein import2-2 222
(ppi2-2) mutation affect the accumulation of GLK1. The homozygous ppi2-2 mutant 223
lacks the major protein import receptor of chloroplasts and exhibits a severe albino 224
phenotype (Kakizaki et al., 2009). Consistent with our previous observation, the level of 225
GLK1 mRNA was down-regulated in ppi2-2 mutant (Fig. 4I). However, the level of GLK1 226
protein in ppi2-2 mutant appeared to be much lower than that of GLK1 mRNA found in 227
the same plants (Fig. 4G and 4H). This observation was similar to NF- or Lin-treated 228
wild-type plants. 229
Taken together, these data suggest that damaged plastids directly regulate the 230
accumulation of GLK1 protein at the post-transcriptional level. Furthermore, this 231
regulation caused by NF or Lin treatment does not seem to require GUN1. 232
233
15
GLK1 protein forms high molecular weight complexes in the nucleus 234
To further investigate the regulation of GLK1 at the protein level, we attempted to 235
generate a glk1glk2 line complemented with GFP-GLK1. A previous study suggested 236
that GFP-GLK1 is highly unstable (Waters et al., 2008). To avoid degradation due to 237
undesirable conformation of the fusion protein, we inserted a flexible linker 238
(Gly-Gly-Ser-Gly-Gly-Ser) between GFP and GLK1 (Evers et al., 2006). The GFP-GLK1 239
chimeric construct was transformed into a glk1glk2 double mutant. We obtained 240
transgenic lines with a phenotype similar to the wild type (Fig. 5), indicating that the 241
GFP-GLK1 gene complements the glk1glk2 phenotype. When total protein was 242
extracted from these plants and probed with antibodies against GFP, a band was 243
detected with the apparent molecular mass of approximately 100 kDa (Fig. 5B), 244
indicating that the protein was expressed as the chimeric protein. Although these 245
transgenic plants accumulated a small amount of degradation products (an asterisk in 246
Fig. 5B), the major product appeared to be the full length GFP-GLK1 (an arrow in Fig. 247
5B). Similar to the observation of a GLK1 overexpressing line (Leister and Kleine, 2016), 248
the expression of LHCB1 in these lines were comparable to that in wild-type (line #1) or 249
even higher than that in wild-type (line #2). In these transgenic plants, the levels of 250
chlorophyll content were comparable to those in the wild type (Fig. 5D). Hence, we 251
concluded that the GFP-GLK1 protein expressed in these transgenic plants was 252
functional. 253
We next investigated the nature of GFP-GLK1 in more detail, by performing 254
gel-filtration chromatography using the protein extract isolated from the glk1glk2 mutant 255
complemented with GFP-GLK1 (Fig. 6). When total proteins were resolved by 256
gel-filtration chromatography, GFP-GLK1 appears to form high molecular weight 257
16
complexes (Fig. 6A, upper panel). The high molecular weight complexes were also 258
observed in the GLK1ox line, and the apparent molecular masses of GLK1 complexes 259
17
were similar to those of GFP-GLK1 complexes (Fig. 6A, lower panel). This finding 260
suggests that GLK1 forms heteromeric complexes with other proteins such that both 261
18
GLK1 and GFP-GLK1 complexes exhibit similar molecular masses. Furthermore, NF 262
treatment did not affect the apparent molecular masses of GFP-GLK1 complexes 263
19
(Supplemental Fig. S2). 264
A previous study showed that GLK2 localized in the nucleus of Arabidopsis 265
protoplasts (Rauf et al., 2013). Hence, we also investigated the intracellular localization 266
of GFP-GLK1 protein in the complemented line. As shown in Fig. 6B, GFP-GLK1 267
localizes in the nucleus. Consistent with the previous observation in GLK1 268
overexpressing plants (Kobayashi et al., 2012), GFP-GLK1 plants also exhibited excess 269
accumulation of chlorophyll in root plastids (Fig 6B). This observation is attributable to 270
the function of GLK1, as GFP control plants do not show chlorophyll fluorescence under 271
the same conditions (Supplemental Fig. S2B). Furthermore, we sometimes observed 272
green fluorescence from plastids of GFP-GLK1 plants (arrowheads in Fig. 6B). Under 273
the same conditions, we did not observe plastid-derived green fluorescence in GFP 274
control plants (Supplemental Fig. S2B). However, chloroplasts isolated from GFP-GLK1 275
plants did not contain GFP-GLK1 sufficient to demonstrate localization in chloroplasts 276
(Fig. 6C). Hence, we concluded that the majority of GFP-GLK1 localizes to the nucleus. 277
Green fluorescence on plastids appears to be derived from enhanced auto-fluorescence 278
of plastids due to the excess accumulation of chlorophyll within plastids of GFP-GLK1 279
expressing plants (Fig. 5D). 280
281
GFP-GLK1 is polyubiquitinated in vivo 282
The fact that NF treatment did not affect the context of GFP-GLK1 complexes 283
(Supplemental Fig. S2) prompted us to hypothesize that GLK1 is most likely regulated 284
by degradation rather than partitioning within the cell in response to plastid signals. To 285
further confirm this hypothesis, we investigated if GFP-GLK1 accumulation is regulated 286
by damaged plastids. The glk1glk2 mutants complemented with GFP-GLK1 were 287
20
treated with NF. In addition to lines #1 and #2, we also included additional two lines 288
which exhibited moderate complementation phenotype with larger stature and 289
21
light-green coloration (Supplemental Fig. S3). Using those four independent transgenic 290
lines, we confirmed that the expression of GFP-GLK1 mRNA in NF-treated plants was 291
moderately up-regulated regardless of the degree of complementation (Fig. 7A and 292
Supplemental Fig. S3). In contrast, the accumulation of GFP-GLK1 protein was 293
consistently decreased in the NF-treated plants (Fig. 7B and Supplemental Fig. S3). 294
These results further substantiate the idea that the GLK1 levels are directly regulated by 295
degradation of this protein in response to plastid signals. 296
We next explored the possible mechanism by which GLK1 degradation is 297
regulated. One possible regulatory mechanism is polyubiquitination of GLK1 and its 298
subsequent degradation by a proteasome. To test the involvement of 299
ubiquitin-proteasome system in the regulation of GLK1, we examined the 300
ployubiquitination of GFP-GLK1 in vivo using an antibody against the multiubiquitin 301
chain. The glk1glk2 plants complemented with GFP-GLK1 were grown in a whole-plant 302
submerged culture system (Ohyama et al., 2008). After cultivation for 2 weeks, plants 303
were treated with either DMSO or a proteasome inhibitor, MG-132. Addition of MG-132 304
induced the accumulation of polyubiquitinated proteins (Fig. 7C, left panel). When 305
affinity purified GFP-GLK1 was probed with antibody against multi-ubiquitin chain, we 306
observed that MG-132-tretaed plants indeed accumulated polyubiquitinated GFP-GLK1 307
(Fig. 7C, right panel). Specific polyubiquitination of GFP-GLK1 was further confirmed by 308
immunoprecipitation using anti-GLK1 antibodies (Fig. 7D). 309
Taken together, these data indicate that GLK1 is subjected to 310
ubiquitin-proteasome dependent regulation. 311
312
Proteasome-dependent pathway regulates the accumulation of GLK1 protein in 313
22
response to plastid signals 314
To further prove that ubiquitin-proteasome system directly regulates the 315
accumulation of GLK1 in response to plastid signals in vivo, we investigated the effects 316
of MG-132 treatment on the accumulation of GLK1 in WT and gun1-101. WT and 317
gun1-101 plants grown in a whole-plant submerged culture system were treated with NF 318
for 6 days. Then, plants were further treated with DMSO or MG-132 for 18 hours. 319
Consistent with the data obtained from plants grown on MS plates (Fig. 4A), 320
NF-treatment in submerged culture dramatically decreased the level of GLK1 in WT and 321
gun1-101 (Fig. 8). However, those plants partially restored GLK1 accumulation when 322
they were subsequently treated with MG-132 (Fig. 8A). Unlike transcriptional regulation 323
of GLK1 in response to plastid signals (Fig. 4C), it appeared that this regulation of GLK1 324
did not require GUN1 (Fig. 8A). We also investigated the effects of MG-132 on GLK1 325
accumulation using ppi2-2 mutant overexpressing GLK1 (GLK1ox ppi2-2). The 326
expression of GLK1 in GLK1ox ppi2-2 is ~10 times higher than that in wild-type 327
(Kakizaki et al., 2009), allowing us to detect GLK1 protein easily. As shown in Fig. 8B, 328
the level of GLK1 protein in the homozygous GLK1ox ppi2-2(-/-) was dramatically 329
decreased compared to that in heterozygous GLK1ox ppi2-2(+/-) (Fig. 8B, compare 330
lanes 1 and 2). However, the decrease of GLK1 accumulation caused by homozygous 331
ppi2-2 mutation was attenuated by MG-132 treatment (Fig. 8B, lane 3). These data 332
indicate that the decrease of GLK1 caused by multiple plastid signals can be attenuated 333
by MG-132 treatment. MG-132 treatment slightly up-regulated the level of GLK1 mRNA 334
in norflurazon-treated wild-type and gun1-101 plants (Fig. 8C). Nonetheless, this 335
increase was insufficient to fully explain the increase of GLK1 protein in those plants 336
(Fig. 8C, compare mRNA with protein). The level of GLK1 mRNA was unaffected in the 337
23
homozygous GLK1ox ppi2-2(-/-) mutant in response to MG-132 treatment (Fig. 8D). 338
This further supports the idea that the increase of GLK1 in MG-132 treated plants 339
24
results from inhibition of GLK1 degradation by ubiquitin-proteasome pathway. 340
Taken together, these data suggest that the plastid signals regulate the 341
accumulation of GLK1 through the ubiquitin-proteasome system. Furthermore, unlike 342
the transcriptional regulation of GLK1 in response to plastid signals, this regulation does 343
not require GUN1 function. 344
345
346
25
Discussion 347
The GLK1 family of transcription factors is a key component in signal 348
transduction from plastids to the nucleus (Inaba et al., 2011; Chi et al., 2013; Jarvis and 349
Lopez-Juez, 2013). Previous studies have demonstrated that the expression of GLK1 is 350
regulated by plastid signals through GUN proteins (Kakizaki et al., 2009; Waters et al., 351
2009). To elucidate how GLK1 protein levels are regulated within the cell, we 352
characterized the regulatory mechanism of GLK1 using specific antibodies and a 353
chimeric GLK1 protein fused to GFP. Sucrose-dependent and tissue-specific 354
accumulation of GLK1 protein was primarily regulated at the transcriptional level. In 355
contrast, alteration of GLK1 accumulation in response to plastid signals was also 356
regulated at the post-transcriptional level (Fig. 4). We demonstrated that the 357
ubiquitin-proteasome system regulates accumulation of GLK1 protein (Fig. 7 and 8). 358
Our results suggest that plastids regulate the accumulation of GLK1 at multiple levels, 359
allowing the cell to optimize plastid development in response to developmental and 360
environmental cues. We summarize a possible mechanism by which the accumulation 361
of GLK1 is regulated in response to developmental and metabolic cues (Fig. 9). 362
At least two mechanisms involving the ubiquitin-proteasome pathway have been 363
shown to regulate chloroplast protein import and biogenesis (Jarvis and Lopez-Juez, 364
2013). The first mechanism is the direct regulation of protein translocation machinery. In 365
this mechanism, a RING-type ubiquitin ligase of the chloroplast outer membrane, SP1, 366
directly regulates the context of the TOC complex (Ling et al., 2012). The second 367
mechanism involves degradation of unimported chloroplast precursor proteins. This 368
mechanism involves the cytosolic heat shock cognate, Hsc70-4, and its interacting 369
E3-ubiquitin ligase, CHIP (Lee et al., 2009). Those proteins together mediate the 370
26
degradation of unimported chloroplast precursor proteins through the 371
ubiquitin-proteasome system. This quality control mechanism may also involve 372
N-acetylation of unimported precursors (Bischof et al., 2011). Now, we propose a third 373
mechanism regulating chloroplast biogenesis through the ubiquitin-proteasome pathway. 374
That is, the transcription factor indispensable in the accumulation of photosynthetic 375
proteins, GLK1, is degraded by the ubiquitin-proteasome system. In this mechanism, 376
we propose that the damaged plastids send some signal to activate the 377
ubiquitin-proteasome system, thereby degrading GLK1 (Fig. 9). Taken together, these 378
results suggest that the ubiquitin-proteasome system simultaneously regulates the level 379
of GLK1 and precursors of photosynthetic proteins, thereby optimizing the rate of 380
photosynthetic protein import into chloroplasts in response to intracellular or 381
environmental signals. 382
27
According to a previous observation, fluorescence of GLK1-GFP chimeric 383
protein was undetectable even though the construct complemented the glk1glk2 384
phenotype (Waters et al., 2008). Furthermore, GLK1-GFP fusion protein could not be 385
detected by immunoblotting (Waters et al., 2008). To overcome these challenges, we 386
introduced a flexible linker (Gly-Gly-Ser-Gly-Gly-Ser) between GFP and GLK1. This 387
made it possible to detect GFP-GLK1 protein and to analyze the behavior of GLK1 in 388
vivo. Our GFP-GLK1 construct fully complemented LHCB1 expression and chlorophyll 389
accumulation in the glk1glk2 mutant (Fig. 5C and 5D). Because the transcription factor 390
PTM localizes to the chloroplast envelope and targets to the nucleus in response to 391
plastid signals (Sun et al., 2011), we suspected that GFP-GLK1 also localizes to the 392
outer envelope membrane of chloroplasts. However, our results indicate that 393
GFP-GLK1 localizes to the nucleus primarily (Fig. 6B). Hence, unlike PTM, 394
nuclear-plastid partitioning of GLK1 does not play an indispensable role in plastid 395
signaling. 396
As summarized in Fig. 9, the level of GLK1 protein appears to be regulated at 397
the transcriptional level under normal conditions. Consistent with a previous observation 398
(Fitter et al., 2002), GLK1 transcripts predominantly accumulate in aerial tissues but not 399
in roots (Fig. 2C). This expression pattern is reflected by the accumulation of GLK1 400
protein in each tissue (Fig. 2B). We also observed that sucrose positively regulates the 401
expression of GLK1 in Arabidopsis. In a previous study, we showed that plastid signals 402
suppress the transcription of GLK1 in a GUN1-dependent manner (Kakizaki et al., 2009). 403
However, it appears that the ubiquitin-proteasome dependent regulation predominates 404
over transcriptional regulation. Transcription of GLK1 failed to respond to NF treatment 405
in the absence of GUN1, but the ubiquitin-proteasome system degraded GLK1 406
28
regardless of the GLK1 transcript level (Fig. 4 and Fig. 7). Meanwhile, it should be noted 407
that the addition of MG-132 did not fully restore the level of GLK1 in all the plants tested 408
(Fig. 8). Hence, we do not exclude the possibility that mechanisms other than 409
ubiquitin-proteasome system also participate in the regulation of GLK1. Intriguingly, a 410
gun1-101 mutant treated with NF still exhibited down-regulation of LHCB1 expression, 411
whereas the expression of GLK1 was virtually unaffected by NF treatment (Kakizaki et 412
al., 2009). Hence, our finding that GLK1 is degraded by the ubiquitin-proteasome 413
system in response to plastid signals nicely explains the moderate repression of LHCB1 414
observed in the gun1-101 mutant. 415
Whether or not transcription factors other than GLK1 are regulated by the 416
ubiquitin-proteasome system in response to plastid signals remains unclear. It is well 417
known that the transcription factor HY5 is subjected to ubiquitin-proteasome dependent 418
regulation during photomorphogenesis (Osterlund et al., 2000). In the dark, COP1 419
protein interacts with HY5 in the nucleus to target HY5 for degradation. COP1 420
redistributes to the cytosol upon illumination, thereby allowing HY5 to facilitate 421
photomorphogenesis. Coincidently, it has been shown that genetic interaction between 422
HY5 and GLKs plays a key role in coordinating gene expression required for chloroplast 423
biogenesis (Kobayashi et al., 2012). Furthermore, the amount of ABI4 also seems to be 424
in part regulated by proteasome degradation (Finkelstein et al., 2011). It remains to be 425
characterized if HY5 and ABI4 are also regulated by the ubiquitin-proteasome system in 426
response to plastid signals. Nonetheless, the genetic and functional link between GLKs 427
and these transcription factors suggests that the ubiquitin-proteasome system may be 428
important in regulating other transcription factors in plastid signaling pathways. 429
Our gel filtration chromatography results reveal that GLK1 forms high molecular 430
29
weight complexes with other proteins. The components involved in high molecular 431
weight complexes remain to be identified. To date, a few proteins have been shown to 432
associate with GLKs. A search for proteins that interact with G-box binding factors 433
(GBFs) led to the identification of GLKs as possible interacting partners (Tamai et al., 434
2002). Likewise, yeast two-hybrid screening of proteins interacting with the NAC 435
transcription factor ORE1 identified GLKs as interacting partners (Rauf et al., 2013). 436
Interaction of GLKs with other transcription factors appears to affect the transcription of 437
target genes. In addition to transcription factors, proteins involved in GLK1 degradation, 438
such as ubiquitin conjugating enzymes, are likely to participate in these molecular 439
complexes containing GLK1. As illustrated by the role of COP1 in HY5 degradation 440
(Osterlund et al., 2000), ubiquitin-conjugating enzymes that associate with GLKs may 441
determine specific degradation of GLK1 in response to plastid signals. Hence, it is of 442
great interest to identify proteins involved in high molecular weight complexes of GLK1. 443
In summary, we demonstrated that a proteasome-dependent pathway regulates 444
the accumulation of GLK1 protein in response to plastid signals in Arabidopsis. In 445
addition to transcriptional regulation of GLK1 through GUN1, it appears that the 446
ubiquitin-proteasome dependent degradation of GLK1 plays a pivotal role in plastid 447
signaling. Taken together, these results suggest that plastids have evolved multiple 448
mechanisms to optimize the import rate of nuclear-encoded plastid proteins in response 449
to plastid signals, thereby regulating chloroplast biogenesis appropriately. 450
451
452
Materials and Methods 453
Plant Material and Growth Conditions 454
30
All experiments were performed on Arabidopsis thaliana accessions Columbia 455
(Col-0). The glk1, glk2, and glk1glk2 mutants were elsewhere (Fitter et al., 2002) and 456
obtained from ABRC. The gun1-101 mutant is described elsewhere (Kakizaki et al., 457
2009). GLK1ox , GLK1ox ppi2-2 (+/-) and GLK1ox ppi2-2(-/-) plants were segregated 458
from a PPI2/ppi2 line overexpressing GLK1 (Kakizaki et al., 2009). Plants were grown 459
on soil (expanded vermiculite) or on 0.5% agar medium containing 1% sucrose and 0.5x 460
MS salts. To synchronize germination, all seeds were kept at 4 °C for 2-3 days after 461
sowing. Unless specified, plants were grown under continuous white light (60~80 462
μmol·m-2·s-1) at 22 °C and 50% relative humidity in a growth chamber (LHP-220S or 463
LHP-350S, NK systems). 464
Tissue specific expression of GLK1 mRNA and GLK1 protein was examined 465
using Arabidopsis grown on soil under continuous light. Each tissue was harvested at 466
day 6 (cotyledon), day 28 (rosette leaves and roots), and day 42 (stem, cauline leaves 467
and flower bud) after transfer to a growth chamber. To test the effect of sucrose on 468
expression (Fig. 3), wild-type and gun1-101 mutant were grown on 0.5x MS plates 469
containing either sucrose (1% or 2%) or mannitol (1.06%, equivalent to 2% sucrose in 470
terms of molar concentration). Because plants grown on sucrose grew faster than those 471
grown on mannitol-containing or sucrose-depleted plates, we harvested each sample 472
12 days (1% and 2% sucrose) or 13 days (sucrose-depleted and mannitol) after the 473
transfer of plates to the growth chamber. For the norflurazon (NF) treatment in Fig. 4 474
and Fig. 7, plants were grown under normal conditions under continuous white light 475
(~100 μmol·m-2·s-1) for 3 days and then treated with 1 μM norflurazon or DMSO under 476
continuous light for another 5 days. 477
For lincomycin treatment, WT and gun1-101 were grown on MS plates for 7 478
31
days and then transferred to liquid MS medium according to the literature (Ohyama et 479
al., 2008). After cultivation in liquid MS medium for 3 days, plants were treated with 480
either 0.5mM lincomycin or distilled water (control) for 5 days. The extracted protein and 481
RNA were analyzed by immunoblotting and real-time PCR, respectively. 482
483
Construction of vectors and Arabidopsis transformation 484
To create the NusA-TEV-GLK1-His construct (Supplemental Fig. S1A), GLK1 485
cDNA was inserted into SpeI and XhoI sites of a pET43.1a vector (Novagen). To 486
introduce the TEV protease cleavage site, a nucleotide sequence that encodes the TEV 487
cleavage site was added at the 5′ end of the primer containing the GLK1 start codon. 488
To create the GFP-GLK1 construct, GFP and GLK1 cDNAs were amplified 489
separately. To increase conformational flexibility of the chimeric protein, flexible linker 490
DNA encoding Gly-Gly-Ser-Gly-Gly-Ser was added between GFP and GLK1. Both GFP 491
and GLK1 fragments were subcloned simultaneously into a pUC vector using the 492
infusion HD cloning system (Takara, Otsu, Japan). The resulting plasmid, 493
pUC-GFP-GGSGGS-GLK1, was used as the template to further amplify GFP-GLK1 494
using PCR. The amplified fragment was inserted into NcoI and NheI sites of the 495
pCAMBIA1301 vector. The pCAMBIA1301 construct was introduced into Arabidopsis 496
thaliana cv. Columbia via Agrobacterium tumefaciens-mediated transformation using 497
the floral dip method (Clough and Bent, 1998). Primer sequences used for vector 498
construction are shown in Supplemental Table S1. 499
500
Purification of GLK1-His 501
For bacterial expression, pET43.1a-GLK1-His was transformed into E. coli BL21(DE3). 502
32
Expression was induced with 0.2 mM IPTG overnight at 20°C, and both soluble and 503
insoluble fractions were recovered to obtain the NusA-GLK1-His fusion protein. The 504
insoluble NusA-GLK1-His was purified using Ni2+ affinity chromatography and refolded 505
by gel filtration. The fusion protein was then cleaved into NusA and GLK1-His 506
(Supplemental Fig. S1B, lane 1). The GLK1-His protein (Supplemental Fig. S1B, lane 2) 507
was further purified using preparative gel electrophoresis (Hayakawa et al., 2001) using 508
a NA-1800 device (Nihon Eido, Tokyo, Japan). 509
510
Antibodies and Immunoblotting 511
Rabbit polyclonal antibodies against GLK1 were produced using either 512
GLK1-His excised from the gel (Supplemental Fig. S1B, lane 1, arrow) or purified 513
GLK1-His (Supplemental Fig. S1B, lane 2) as the antigen. To obtain purified antibodies 514
against GLK1, serum was first applied to NusA-sepharose to deplete NusA antibodies, 515
and further applied to NusA-TEV-GLK1-His-Sepharose. The bound antibodies were 516
eluted with 0.2M Glycine-HCl buffer, pH 2.2. Monoclonal antibody against actin was 517
purchased from CHEMICON. Immuno-labeled proteins were detected with a 518
lumino-image analyzer (AE-6972C, ATTO Co. Ltd., Japan) using chemiluminescence 519
reagents. Signal was quantified using image acquisition software (CS Analyzer, ATTO 520
Co. Ltd.). 521
522
Purification of polyubiquitinated GFP-GLK1 523
The glk1glk2 mutant complemented with GFP-GLK1 was grown in liquid MS 524
medium according to the literature (Ohyama et al., 2008). After cultivation for 2 weeks, 525
plants were treated with either DMSO or a proteasome inhibitor, 50μM MG-132, for 12 526
33
hours. GFP-GLK1 was recovered using the μMACS GFP-Tagged Protein Isolation Kit 527
(Miltenyi Biotec K.K., Germany) or the affinity-purified antibodies against GLK1 bound to 528
protein A sepharose. Total protein extract and purified GFP-GLK1 protein were then 529
resolved by SDS-PAGE and probed with an antibody against the multiubiquitin chain 530
(clone FK2, Cayman Chemical). 531
532
Detection of GLK1 in MG-132 treated plants 533
WT and gun1-101 were grown on MS plates for 7 days and then transferred to 534
liquid MS medium. After cultivation in liquid MS medium for 3 days, plants were treated 535
with either DMSO for 4 days or 1μM norflurazon for 6 days. Those plants were further 536
treated with either DMSO or 50μM MG-132 for 18 hours. The extracted protein and RNA 537
were analyzed by immunoblotting and real-time PCR, respectively. 538
For MG-132 treatment of ppi2-2 mutant overexpressing GLK1 (Fig. 8B and 8D), 539
the progeny of heterozygous ppi2-2 carrying homozygous GLK1 transgene was grown 540
on MS plates for 7 days. Then, heterozygous ppi2-2 and homozygous ppi2-2 mutants 541
were separately transferred to liquid MS medium and allowed to grow for 6 days 542
(heterozygous) or 10 days (homozygous). Those plants were further treated with either 543
DMSO or 50μM MG-132 for 18 hours. The extracted protein and RNA were analyzed by 544
immunoblotting and real-time PCR, respectively. 545
546
Real-Time PCR Analysis 547
cDNA was synthesized using the PrimeScriptTM RT reagent kit (TaKaRa) with 548
random hexamer and oligo d(T) primers. Real-time PCR (RT-PCR) was performed on a 549
Thermal Cycler DiceTM Real-Time System (TaKaRa) using SYBR Premix ExTaq II 550
34
(TaKaRa) as described previously (Kakizaki et al., 2009). All the real-time PCR analyses 551
were performed using biological triplicate samples. Primers used for real-time PCR are 552
listed in Supplemental Table S1. Transcript levels of each gene were normalized to that 553
of ACTIN2. 554
555
Gel filtration chromatography 556
True leaves of Arabidopsis plants were homogenized in a buffer containing 50 557
mM Tricine-KOH (pH 7.5), 2 mM EDTA, and 400 mM sucrose, and the homogenate was 558
centrifuged at 1,000xg for 5 min. Then, the supernatant was mixed with an equal 559
volume of buffer containing 2% Triton X-100 to solubilize membranes, and centrifuged 560
at 20,000xg for 10 min. The resulting supernatant was resolved on a Sephacryl 561
S-300HR column equilibrated with TES buffer containing 0.1% Triton X-100 (Okawa et 562
al., 2008) using an AKTA-prime chromatography system (GE Healthcare). The first 30 563
mL of eluate were discarded as void volume because we did not observe any 564
absorbance at 280 nm. Fractions of 1.5 mL each were collected and the protein in each 565
fraction was precipitated with 10% trichloroacetic acid. Even number fractions of the 566
original samples were then resolved by SDS-PAGE and probed with antibodies 567
indicated in each panel (Fig. 6A and Fig. S1A). Among four glk1glk2 lines 568
complemented with GFP-GLK1, line #2 was used for this analysis. 569
570
Quantification of Chlorophyll 571
Quantification of chlorophyll was performed as described previously (Kakizaki et 572
al., 2009). 573
574
35
Confocal microscopy 575
Leaf and root samples were prepared from the transgenic lines expressing 576
GFP-GLK1 or GFP. GFP control plant is described previously (Okawa et al., 2008). 577
Images were captured with a confocal laser-scanning microscope LSM 700 (Carl Zeiss, 578
Germany). To directly compare GFP fluorescence and chlorophyll auto-fluorescence 579
obtained from GFP-GLK1 transgenic plants with those obtained from GFP expressing 580
plants, laser power and detection gain were fixed during a series of analyses. 581
582
583
36
Figure Legends 584
Figure 1. Detection of GLK1 protein by affinity purified anti-GLK1 antibodies. 585
Total proteins were extracted from wild-type, glk1, glk2, glk1glk2, and GLK1 586
over-expressing (GLK1ox) Arabidopsis plants, resolved by SDS-PAGE, and probed with 587
antibodies against GLK1. Asterisks indicate nonspecific proteins detected by antibodies. 588
589
Figure 2. Tissue-specific accumulation of AtGLK1 protein in Arabidopsis 590
A. Accumulation of GLK1 protein in various tissues of Arabidopsis. Total protein was 591
extracted from various tissues, resolved by SDS-PAGE, and probed with antibodies 592
against GLK1 (upper panel) or actin (lower panel). The amount of protein loaded in each 593
lane was as follows: 30 μg for root and 80 μg for other tissues. 594
B. Quantification of GLK1 protein levels in each tissue. Protein levels were quantified 595
using image acquisition software and normalized to actin levels. Each black bar 596
represents the average of two independent samples. The GLK1 protein level in the 597
cotyledon was set to 1. 598
C. Quantitative analysis of GLK1 mRNA abundance in various tissues. Total RNA was 599
extracted from various tissues of wild-type plants. The mRNA levels were analyzed by 600
real-time PCR, and the expression levels were normalized to that of ACTIN2. The 601
expression level in the cotyledon was set to 1. Error bars represent 1 SE of the mean 602
(n=3). 603
604
Figure 3. GLK1 accumulation in response to sucrose 605
A. Immunoblot analysis of GLK1 protein in the wild-type (WT) and gun1-101 mutant 606
plants treated with sucrose (1% or 2%) or mannitol (1.06%, equivalent to 2% sucrose in 607
37
terms of molar concentration). Total protein was extracted from sucrose- or 608
mannitol-treated plants, resolved by SDS-PAGE, and probed with antibodies against 609
GLK1 or actin (upper panel). The protein levels in each image were quantified using 610
image acquisition software and normalized to actin levels (lower panel). The GLK1 611
protein level in untreated WT plants was set to 1. 612
B and C. Quantitative analysis of GLK1 (B) and LHCB1 (C) expression in wild-type and 613
gun1-101 mutant plants treated with sucrose or mannitol. The mRNA levels were 614
analyzed by real-time PCR, and the expression levels were normalized to that of 615
ACTIN2. Error bars represent 1 SE of the mean (n=3). The expression level in untreated 616
WT plants was set to 1. 617
618
Figure 4. Effects of inhibitor treatment and ppi2-2 mutation on GLK1 accumulation 619
A. Immunoblot analysis of GLK1 protein in wild-type (WT) and gun1-101 mutant plants 620
treated with norflurazon (NF). After growth under normal conditions for 3 days, plants 621
were treated with 1μM NF (+) or dimethyl sulfoxide (DMSO, -) under continuous light for 622
5 days. Extracted proteins were then resolved by SDS-PAGE, and proteins were probed 623
with antibodies against GLK1 (upper panel) and actin (lower panel), respectively. 624
B. Quantification of GLK1 protein levels in each sample shown in A. GLK1 protein levels 625
were quantified using image acquisition software and normalized to actin levels. The 626
GLK1 protein level in DMSO-treated WT plants was set to 1. 627
C, Quantitative analysis of GLK1 mRNA expression in WT and gun1-101 mutant plants 628
treated with NF (+) or DMSO (-). WT and gun1-101 plants were treated with either NF or 629
DMSO as described for panel A. The mRNA levels were analyzed by real-time PCR, 630
and expression levels were normalized to that of ACTIN2. Error bars represent 1 SE of 631
38
the mean (n=3). The expression level in DMSO-treated WT plants was set to 1. 632
D. Immunoblot analysis of GLK1 protein in WT and gun1-101 mutant plants treated with 633
lincomycin (Lin). Plants grown in a liquid MS medium (see materials and methods) were 634
treated with 0.5mM Lin (+) or distilled water (DW, -) under continuous light for 5 days. 635
Extracted proteins were then resolved by SDS-PAGE, and proteins were probed with 636
antibodies against GLK1 (upper panel) and actin (lower panel), respectively. 637
E. Quantification of GLK1 protein levels in each sample shown in D. The GLK1 protein 638
level in DW-treated WT plants was set to 1. 639
F. Quantitative analysis of GLK1 mRNA expression in the WT and gun1-101 mutant 640
plants treated with Lin (+) or DW (-). The mRNA levels were analyzed by real-time PCR, 641
and expression levels were normalized to that of ACTIN2. Error bars represent 1 SE of 642
the mean (n=3). The expression level in DW-treated WT plants was set to 1. 643
G. Immunoblot analysis of GLK1 protein in WT and homozygous ppi2-2. Plants were 644
grown under continuous light for 15 days. Extracted proteins were then resolved by 645
SDS-PAGE, and proteins were probed with antibodies against GLK1 (upper panel) and 646
actin (lower panel), respectively. 647
H. Quantification of GLK1 protein levels in each sample shown in G. The GLK1 protein 648
level in WT plants was set to 1. 649
I, Quantitative analysis of GLK1 mRNA expression in WT and ppi2-2 mutant. The mRNA 650
levels were analyzed by real-time PCR, and expression levels were normalized to that 651
of ACTIN2. Error bars represent 1 SE of the mean (n=3). The expression level in WT 652
plants was set to 1. 653
654
Figure 5. Expression of GFP-GLK1 in the glk1glk2 mutant 655
39
A. Representative phenotype of the glk1glk2 mutant transformed with GFP-GLK1 656
construct. Transformed plants were first grown on MS plates containing Hygromycin B 657
for 2 weeks, transferred to soil, and allowed to grow for another 2 weeks. Wild-type 658
(WT) and glk1glk2 plants were grown on MS plates without antibiotics and then 659
transferred to soil. 660
B. Immunoblot analysis of GFP-GLK1. Total proteins were extracted from WT, glk1glk2, 661
and transformed plants, and probed with antibodies against GFP. GFP-GLK1 is 662
indicated by an arrow. An asterisk indicates the position of degraded GFP-GLK1. 663
C. Expression analysis of LHCB1 in WT, glk1glk2, and transformed plants. The mRNA 664
levels were analyzed by real-time PCR, and expression levels were normalized to that 665
of ACTIN2. Error bars represent 1 SE of the mean (n=3). The expression level in WT 666
plants was set to 1. 667
D. Total chlorophyll content in WT, glk1glk2, and transformed plants. Error bars 668
represent 1 SE of the mean (n=3). FW, Fresh weight. 669
670
Figure 6. Identification of GFP-GLK1 complexes by gel-filtration chromatography and 671
intracellular localization of GFP-GLK1 672
A. Gel-filtration chromatography of GFP-GLK1 and GLK1 proteins expressed in 673
Arabidopsis. Total protein extracts from GFP-GLK1 transformed glk1glk2 plants (upper 674
panel) and GLK1ox plants (lower panel) were resolved by gel-filtration chromatography 675
on a Sephacryl S-300 HR column. The molecular masses of the standard proteins are 676
indicated by arrowheads. To verify the validity of chromatography, the distribution of 677
RubisCO (~550 kDa) was also investigated. Proteins in each fraction were precipitated 678
with trichloroacetic acid and analyzed by immunoblotting with antibodies against GLK1 679
40
or RubisCO holoenzyme. An arrow indicates the position of GLK1, and asterisks 680
indicate non-specific bands. 681
B. Localization of GFP-GLK1 protein in Arabidopsis leaf and root cells. Leaf (upper) and 682
root (lower) tissues of transgenic plants expressing GFP-GLK1 in glk1glk2 background 683
were observed using a confocal laser-scanning microscope LSM 700. Dashed lines in 684
DIC images indicate the location of the nucleus. GFP, GFP fluorescence; Chl., 685
chlorophyll auto-fluorescence; DIC, differential interference contrast image; GFP+Chl., 686
overlap of the GFP and Chl. images. 687
C. Detection of GFP-GLK1 in chloroplasts. Total protein extracts (lane 1) and 688
chloroplast proteins (lane 2) were resolved by SDS-PAGE and probed with antibodies 689
against GLK1 (top) and Toc75 (middle). As the negative control, the membrane was 690
also probed with an anti-actin antibody (bottom). 691
692
Figure 7. Effects of norflurazon and MG-132 on GFP-GLK1 in the transformed plants 693
A. Effects of norflurazon (NF) on the expression of GFP-GLK1. After growth under 694
normal conditions for 3 days, plants were treated with 1μM NF (+) or dimethyl sulfoxide 695
(DMSO, -) under continuous light for 5 days. The mRNA levels were analyzed by 696
real-time PCR, and expression levels were normalized to that of ACTIN2. The graph 697
shows the average of four independent transgenic lines shown in Supplemental Figure 698
S3. Error bars represent 1 SE of the mean (n=4). 699
B. Accumulation of GFP-GLK1 in the glk1glk2 double mutant transformed with 700
GFP-GLK1. Plants were treated with either 1μM NF or DMSO as described for panel A. 701
Extracted proteins were then resolved by SDS-PAGE, and proteins were probed with 702
antibodies against GLK1 or actin (Supplemental Figure S3B). GLK1 protein levels were 703
41
quantified using image acquisition software and normalized to actin levels. The graph 704
shows the average of four independent transgenic lines shown in Supplemental Figure 705
S3. Error bars represent 1 SE of the mean (n=4). 706
C. Accumulation of poly-ubiquitinated GFP-GLK1 in plants treated with MG-132. 707
Arabidopsis plants expressing GFP-GLK1 were cultured in a liquid MS medium for 2 708
weeks. Plants were then treated with 50μM MG-132 (+) in a liquid culture for 12 hours. 709
As the control, plants were also treated with DMSO (-) for the same duration. Total 710
proteins were extracted from MG-132 or DMSO-treated plants, and GFP-GLK1 protein 711
was affinity purified using monoclonal antibody against GFP. The starting material (1% 712
of the total) and eluates were then resolved by SDS-PAGE, and proteins were probed 713
with monoclonal antibody against the multiubiquitin chain. Actin amounts confirmed 714
equal loading of starting material. 715
716
Figure 8. Effects of MG-132 on GLK1 accumulation in vivo 717
A. Effects of MG-132 on the accumulation of GLK1 in wild-type (WT) and gun1-101 718
plants treated with norflurazon (NF). Plants were treated with DMSO (lanes 1 and 4) for 719
4 days or with NF for 6 days. The NF-treated plants were further treated with additional 720
DMSO (lanes 2 and 5) or 50μM MG-132 (lanes 3 and 6) for 18 hours. Extracted proteins 721
were then resolved by SDS-PAGE, and proteins were probed with antibodies against 722
GLK1 or actin. Note that the exposure time to capture GLK1 signals in WT lanes was 723
longer than that in gun1-101 lanes. 724
B. Effects of MG-132 on the accumulation of GLK1 in ppi2-2 mutant overexpressing 725
GLK1 (GLK1ox ppi2-2). Note that only the plants carrying homozygous ppi2 genotype 726
(ppi2 -/-) exhibit damaged plastids and albino phenotype. The GLK1ox ppi2-2 mutant 727
42
was grown on plates for 7 days. Then, plants were transferred to liquid MS medium. The 728
heterozygous GLK1ox ppi2-2(+/-) plants (green phenotype) were grown in liquid MS 729
medium for 6 days under continuous light and then subjected to DMSO treatment for 18 730
hours. The homozygous GLK1ox ppi2-2(-/-) plants (albino phenotype) were grown in 731
liquid MS medium for 10 days under continuous light and then treated with either DMSO 732
or 50μM MG-132 for 18 hours. Extracted proteins were then resolved by SDS-PAGE, 733
and proteins were probed with antibodies against GLK1 or actin. 734
C. Quantification of GLK1 mRNA and GLK1 protein levels in NF-treated plants (NF) and 735
both NF and MG-132-treated plants (NF + MG-132) shown in A. GLK1 protein levels 736
were quantified using image acquisition software and normalized to actin levels. The 737
mRNA levels were analyzed by real-time PCR, and the expression levels were 738
normalized to that of ACTIN2. The expression levels in WT (left) or gun1-101 (right) 739
plants treated with both NF and MG-132 was set to 1. Error bars shown in mRNA data 740
represent 1 SE of the mean (n=3). 741
D. Quantification of GLK1 mRNA and GLK1 protein levels in the GLK1ox ppi2-2 (-/-) 742
mutant treated with DMSO or MG-132 shown in B. GLK1 protein levels were quantified 743
using image acquisition software and normalized to actin levels. The mRNA levels were 744
analyzed by real-time PCR, and the expression levels were normalized to that of 745
ACTIN2. The expression level in GLK1ox ppi2-2(-/-) plants treated with MG-132 was set 746
to 1. Error bars shown in mRNA data represent 1 SE of the mean (n=3). 747
748
Figure 9. Model for the regulation of GLK1 by multiple mechanisms 749
Both developmental signals and sucrose regulate the accumulation of GLK1 protein 750
through transcriptional regulation of GLK1 gene. Plastid signals derived from damaged 751
43
plastids also regulate the expression of GLK1 through the activity of GUN1. Meanwhile, 752
plastid signals regulate the accumulation of GLK1 at protein level. Although the 753
components involved in this regulation remain to be identified, ubiquitin-proteasome 754
system appears to participate in this regulation. 755
756
757
Supplemental Figure Legends 758
Supplemental Figure S1. Expression of GLK1 fusion protein in E. coli 759
A. Schematic diagram of the construct used for bacterial expression. 760
B. Purification of GLK1-His protein. The purified NusA-TEV-GLK1-His protein was 761
cleaved into NusA (arrowhead) and GLK1-His (arrow) by TEV protease (lane 1). The 762
GLK1-His protein was further purified from the protease treated fraction using 763
preparative gel electrophoresis (lane 2). 764
765
Supplemental Figure S2. Localization of GFP-GLK1 and GFP in Arabidopsis 766
A. Gel-filtration chromatography of GFP-GLK1 proteins in Arabidopsis treated with NF. 767
Total protein extracts from the GFP-GLK1-transformed glk1glk2 mutants treated with NF 768
were resolved by gel-filtration chromatography on a Sephacryl S-300 HR column. The 769
molecular masses of the standard proteins are indicated by arrowheads. Proteins in 770
each fraction were precipitated with trichloroacetic acid and analyzed by immunoblotting 771
with antibodies indicated at the left. The level of GFP-GLK1 in the complemented line 772
was high such that it was detectable even in the plants treated with NF. 773
B. Localization of GFP in Arabidopsis leaf and root cells. Leaf (upper) and root (lower) 774
tissues of transgenic plants expressing GFP in wild-type background were observed 775
44
using a confocal laser-scanning microscope LSM 700. The images were taken as the 776
control of GFP-GLK1 shown in Fig. 6B. GFP, GFP fluorescence; Chl., chlorophyll 777
auto-fluorescence; DIC, differential interference contrast image; GFP+Chl., overlap of 778
the GFP and Chl. images. 779
780
Supplemental Figure S3. Analysis of glk1glk2 mutants complemented with GFP-GLK1 781
gene 782
A. Representative phenotype of the additional glk1glk2 lines transformed with 783
GFP-GLK1 construct. Transformed plants were first grown on MS plates for 2 weeks, 784
transferred to soil and continued to grow for another 2 weeks. Wild-type and glk1glk2 785
plants were grown on MS plates without antibiotics and then transferred to soil. 786
B. Effect of norflurazon (NF) on the accumulation of GFP-GLK1 protein. After growth 787
under normal conditions for 3 days, plants were treated with 1μM NF (+) or dimethyl 788
sulfoxide (DMSO, -) under continuous light for 5 days. Extracted proteins were then 789
resolved by SDS-PAGE, and proteins were probed with antibodies against GFP or actin 790
(top panel). The lower panel shows quantification of GLK1 protein level in each sample. 791
GLK1 protein levels were quantified using image acquisition software and normalized to 792
actin levels. The GLK1 protein level in DMSO-treated plants was set to 1 in each line. 793
C. Effects of NF on the expression of GFP-GLK1. Plants were treated with either 1μM 794
NF (+) or DMSO (-) as described for panel B. The mRNA levels were analyzed by 795
real-time PCR, and expression levels were normalized to that of ACTIN2. The 796
expression level in plants treated with DMSO was set to 1. Error bars represent 1 SE of 797
the mean (n=3). 798
799
Parsed CitationsBischof S, Baerenfaller K, Wildhaber T, Troesch R, Vidi PA, Roschitzki B, Hirsch-Hoffmann M, Hennig L, Kessler F, Gruissem W,Baginsky S (2011) Plastid proteome assembly without Toc159: photosynthetic protein import and accumulation of N-acetylatedplastid precursor proteins. Plant Cell 23: 3911-3928
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Chan KX, Mabbitt PD, Phua SY, Mueller JW, Nisar N, Gigolashvili T, Stroeher E, Grassl J, Arlt W, Estavillo GM, Jackson CJ, PogsonBJ (2016) Sensing and signaling of oxidative stress in chloroplasts by inactivation of the SAL1 phosphoadenosine phosphatase.Proc Natl Acad Sci U S A 113: E4567-4576
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Chi W, Sun X, Zhang L (2013) Intracellular signaling from plastid to nucleus. Annu Rev Plant Biol 64: 559-582Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. PlantJ 16: 735-743
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Cottage A, Mott EK, Kempster JA, Gray JC (2010) The Arabidopsis plastid-signalling mutant gun1 (genomes uncoupled1) showsaltered sensitivity to sucrose and abscisic acid and alterations in early seedling development. J Exp Bot 61: 3773-3786
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Estavillo GM, Crisp PA, Pornsiriwong W, Wirtz M, Collinge D, Carrie C, Giraud E, Whelan J, David P, Javot H, Brearley C, Hell R,Marin E, Pogson BJ (2011) Evidence for a SAL1-PAP chloroplast retrograde pathway that functions in drought and high lightsignaling in Arabidopsis. Plant Cell 23: 3992-4012
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Evers TH, van Dongen EM, Faesen AC, Meijer EW, Merkx M (2006) Quantitative understanding of the energy transfer betweenfluorescent proteins connected via flexible peptide linkers. Biochemistry 45: 13183-13192
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Finkelstein R, Lynch T, Reeves W, Petitfils M, Mostachetti M (2011) Accumulation of the transcription factor ABA-insensitive (ABI)4is tightly regulated post-transcriptionally. J Exp Bot 62: 3971-3979
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Fitter DW, Martin DJ, Copley MJ, Scotland RW, Langdale JA (2002) GLK gene pairs regulate chloroplast development in diverseplant species. Plant J 31: 713-727
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Hayakawa M, Hosogi Y, Takiguchi H, Saito S, Shiroza T, Shibata Y, Hiratsuka K, Kiyama-Kishikawa M, Abiko Y (2001) Evaluation ofthe electroosmotic medium pump system for preparative disk gel electrophoresis. Anal Biochem 288: 168-175
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Inaba T, Schnell DJ (2008) Protein trafficking to plastids: one theme, many variations. Biochem J 413: 15-28Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Inaba T, Yazu F, Ito-Inaba Y, Kakizaki T, Nakayama K (2011) Retrograde signaling pathway from plastid to nucleus. Int Rev Cell MolBiol 290: 167-204
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Jarvis P, Lopez-Juez E (2013) Biogenesis and homeostasis of chloroplasts and other plastids. Nat Rev Mol Cell Biol 14: 787-802Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Kakizaki T, Matsumura H, Nakayama K, Che FS, Terauchi R, Inaba T (2009) Coordination of plastid protein import and nuclear geneexpression by plastid-to-nucleus retrograde signaling. Plant Physiol 151: 1339-1353
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Kakizaki T, Yazu F, Nakayama K, Ito-Inaba Y, Inaba T (2012) Plastid signalling under multiple conditions is accompanied by acommon defect in RNA editing in plastids. J Exp Bot 63: 251-260
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Karpinski S, Reynolds H, Karpinska B, Wingsle G, Creissen G, Mullineaux P (1999) Systemic signaling and acclimation in responseto excess excitation energy in Arabidopsis. Science 284: 654-657
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Kobayashi K, Baba S, Obayashi T, Sato M, Toyooka K, Keranen M, Aro EM, Fukaki H, Ohta H, Sugimoto K, Masuda T (2012)Regulation of root greening by light and auxin/cytokinin signaling in Arabidopsis. Plant Cell 24: 1081-1095
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Koussevitzky S, Nott A, Mockler TC, Hong F, Sachetto-Martins G, Surpin M, Lim J, Mittler R, Chory J (2007) Signals fromchloroplasts converge to regulate nuclear gene expression. Science 316: 715-719
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Lee S, Lee DW, Lee Y, Mayer U, Stierhof YD, Lee S, Jurgens G, Hwang I (2009) Heat shock protein cognate 70-4 and an E3ubiquitin ligase, CHIP, mediate plastid-destined precursor degradation through the ubiquitin-26S proteasome system inArabidopsis. Plant Cell 21: 3984-4001
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Leister D, Kleine T (2016) Definition of a core module for the nuclear retrograde response to altered organellar gene expressionidentifies GLK overexpressors as gun mutants. Physiol Plant 157: 297-309
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Ling Q, Huang W, Baldwin A, Jarvis P (2012) Chloroplast biogenesis is regulated by direct action of the ubiquitin-proteasomesystem. Science 338: 655-659
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Nakamura H, Muramatsu M, Hakata M, Ueno O, Nagamura Y, Hirochika H, Takano M, Ichikawa H (2009) Ectopic overexpression ofthe transcription factor OsGLK1 induces chloroplast development in non-green rice cells. Plant Cell Physiol 50: 1933-1949
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Ohyama K, Ogawa M, Matsubayashi Y (2008) Identification of a biologically active, small, secreted peptide in Arabidopsis by insilico gene screening, followed by LC-MS-based structure analysis. Plant J 55: 152-160
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Okawa K, Nakayama K, Kakizaki T, Yamashita T, Inaba T (2008) Identification and characterization of Cor413im proteins as novelcomponents of the chloroplast inner envelope. Plant Cell Environ 31: 1470-1483
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Osterlund MT, Hardtke CS, Wei N, Deng XW (2000) Targeted destabilization of HY5 during light-regulated development ofArabidopsis. Nature 405: 462-466
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Paparelli E, Gonzali S, Parlanti S, Novi G, Giorgi FM, Licausi F, Kosmacz M, Feil R, Lunn JE, Brust H, van Dongen JT, Steup M,Perata P (2012) Misexpression of a chloroplast aspartyl protease leads to severe growth defects and alters carbohydratemetabolism in Arabidopsis. Plant Physiol 160: 1237-1250
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Pogson BJ, Woo NS, Forster B, Small ID (2008) Plastid signalling to the nucleus and beyond. Trends Plant Sci 13: 602-609Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Rauf M, Arif M, Dortay H, Matallana-Ramirez LP, Waters MT, Gil Nam H, Lim PO, Mueller-Roeber B, Balazadeh S (2013) ORE1balances leaf senescence against maintenance by antagonizing G2-like-mediated transcription. EMBO Rep 14: 382-388
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Sun X, Feng P, Xu X, Guo H, Ma J, Chi W, Lin R, Lu C, Zhang L (2011) A chloroplast envelope-bound PHD transcription factormediates chloroplast signals to the nucleus. Nat Commun 2: 477
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Szechynska-Hebda M, Karpinski S (2013) Light intensity-dependent retrograde signalling in higher plants. J Plant Physiol 170:1501-1516
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Tamai H, Iwabuchi M, Meshi T (2002) Arabidopsis GARP transcriptional activators interact with the Pro-rich activation domainshared by G-box-binding bZIP factors. Plant Cell Physiol 43: 99-107
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Wagner D, Przybyla D, Op den Camp R, Kim C, Landgraf F, Lee KP, Wursch M, Laloi C, Nater M, Hideg E, Apel K (2004) The geneticbasis of singlet oxygen-induced stress responses of Arabidopsis thaliana. Science 306: 1183-1185
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Waters MT, Moylan EC, Langdale JA (2008) GLK transcription factors regulate chloroplast development in a cell-autonomousmanner. Plant J 56: 432-444
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Waters MT, Wang P, Korkaric M, Capper RG, Saunders NJ, Langdale JA (2009) GLK transcription factors coordinate expression ofthe photosynthetic apparatus in Arabidopsis. Plant Cell 21: 1109-1128
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Xiao Y, Savchenko T, Baidoo EE, Chehab WE, Hayden DM, Tolstikov V, Corwin JA, Kliebenstein DJ, Keasling JD, Dehesh K (2012)Retrograde signaling by the plastidial metabolite MEcPP regulates expression of nuclear stress-response genes. Cell 149: 1525-1535
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
Yasumura Y, Moylan EC, Langdale JA (2005) A conserved transcription factor mediates nuclear control of organelle biogenesis inanciently diverged land plants. Plant Cell 17: 1894-1907
Pubmed: Author and TitleCrossRef: Author and TitleGoogle Scholar: Author Only Title Only Author and Title
