Plant Physiology Preview. Published on July 5, 2015, as DOI:10.1104/pp.15.00414
1
Running title: Dynamic DNA methylation in rice seeds
2 3
Corresponding author:
4
Hong-Wei Xue
5
300 Fenglin Road, Shanghai, China 200032
6
Phone: 86-21-54924330
7
Fax: 86-21-54924331
8
Email:
[email protected] 9 10
Research Area: Genes, Development and Evolution
11
1 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
Copyright 2015 by the American Society of Plant Biologists
12
Global Analysis Reveals the Crucial Roles of DNA Methylation during Rice Seed
13
Development
14 15
Mei-Qing Xing2, Yi-Jing Zhang2, Shi-Rong Zhou2, Wen-Yan Hu2, Xue-Ting Wu, Ya-Jin
16
Ye, Xiao-Xia Wu, Yun-Ping Xiao, Xuan Li, and Hong-Wei Xue1,*
17 18
National Key Laboratory of Plant Molecular Genetics, Institute of Plant Physiology and
19
Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 300
20
Fenglin Road, 200032 Shanghai, China (M.-Q.X., W.-Y.H., S.-R.Z., Y.-J.Z., X.-T.W.,
21
Y.-J.Y., X.L., H.-W.X.); College of Bioscience and Biotechnology, Yangzhou University,
22
88 South University Ave, 225009 Yangzhou, China (X.-X.W.); and Shanghai OEbiotech
23
Co., Ltd, 138, Lane 1505, Zuchongzhi Road, 201210 Shanghai, China (Y.-P.X.)
24 25
2
These authors contribute equally to the study.
26 27 28
One-sentence summaries:
29 30
Characterization of the dynamics DNA methylome in developing rice seeds reveals the
31
crucial roles of DNA methylation in regulating seed development and global DNA
32
de-methylation of early endosperm.
33
2 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
34
This study was supported by the State Key Project of Basic Research (2012CB944804),
35
the Ministry of Science and Technology of China (2012AA10A302).
36 37
Address correspondence to
[email protected] 38 39
M.-Q.X., S.-R.Z., and H.-W.X. designed the projects. Y.-J.Y., X.-X.W., and S.-R.Z.
40
collected the materials. Y.-J.Z., W.-Y.H., X.-T.W., X.L., M.-Q.X., and S.-R.Z. analyzed
41
the data. Y.-P.X. helped the sequencing. M.-Q.X., Y.-J.Z., and S.-R.Z. drafted the
42
manuscript and H.W.X. wrote the paper.
43
3 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
44
ABSTRACT
45 46
Seed development is an important process of reproductive development and consists of
47
embryo and endosperm development, both comprise several key processes. To determine
48
and investigate the functions of dynamic DNA methylome during seed development, we
49
profiled the DNA methylation genome-wide in a series of developmental stages of rice
50
embryo and endosperm by MeDIP-seq. Results showed that embryo is hyper-methylated
51
predominantly around non-TE genes, short DNA-TEs and short SINEs compared to
52
endosperm, and non-TE genes have the most diverse methylation status across seed
53
development. In addition, lowly expressed genes are significantly enriched of
54
hyper-methylated genes, but not vice versa, confirming the crucial role of DNA
55
methylation in suppressing gene transcription. Further analysis revealed the significantly
56
decreased methylation at early developing stages (from 2 to 3 day after pollination),
57
indicating a predominant role of de-methylation during early endosperm development
58
and that genes with a consistent negative correlation between DNA methylation change
59
and expression change maybe potentially directly regulated by DNA methylation.
60
Interestingly, comparative analysis of the DNA methylation profiles revealed that both
61
rice indica and japonica subspecies showed robust fluctuant profile of DNA methylation
62
levels in embryo and endosperm across seed development, with highest methylation level
63
at 6 DAP (2 DAP of endosperm of japonica subspecies as well), indicating a complex
64
and fine-controlled methylation pattern is closely associated with the seed development
65
regulation. The systemic characterization of the dynamics DNA methylome in
66
developing rice seeds will help to study the effects and mechanism of epigenetic
67
regulation in seed development.
68
4 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
69
INTRODUCTION
70 71
DNA methylation is essential for repression of transposable elements (TEs) and
72
regulation of gene expression. In plants, DNA methylation occurs in cytosine residue
73
within three different sequence contexts (CG, CHG, CHH) (Cokus et al., 2008), of which
74
methylation at CHG and CHH is predominantly found in TEs, while CG methylation
75
occurs in TEs and genes (Feng et al., 2010; Zemachet al., 2010). Recent studies
76
demonstrated the crucial roles of DNA methylation in plant development, including
77
gametogenesis, seed development and response to stresses (Chen and Zhou, 2013; Boyko
78
and Kovalchuk, 2008).
79
Seed development is an important and fine-controlled complex process of
80
reproductive development. Seed development involves in several key processes and
81
initiates with zygote and fertilized central cell. Early embryo development mainly
82
involves the apico-basal polarity establishment, epidermis differentiation and shoot/root
83
meristem formation (Goldberg et al., 1994). Endosperm development mainly involves the
84
coenocytic,
85
developmental program of rice embryo is different from that of Arabidopsis. An ordered
86
cell division is observed in Arabidopsis embryo at early stage, while cell divisions are not
87
regular even at very early stage of rice embryo (Itoh et al., 2005). In addition,
88
Arabidopsis endosperm is consumed by developing embryo and disappears when seed
89
matures, while that of rice reserves, reflecting the fundamental difference of embryo and
90
endosperm development between rice and Arabidopsis. Studies have shown that many
91
factors are involved in the regulation of embryo and endosperm development, and
92
genetics studies, functional genomics studies have identified some key regulators of seed
93
development. Recent high-throughput transcriptome analysis using microarray and
94
RNA-Seq provided detailed transcriptional profile of genes at different developmental
95
stages of developing seeds (Xue et al., 2012; Xu et al., 2012; Gao et al., 2013), which
96
help to illustrate the regulatory networks of seed development.
cellularization,
differentiation
and
maturation.
The
structure
and
97
In addition to the genetic approaches, recent studies also indicated a crucial role of
98
epigenetic modification in developing seeds (Ikeda, 2012; Sreenivasulu and Wobus,
99
2013). The central cell chromatin is less methylated than that of somatic cells (Jullien and
100
Berger, 2010) and then reprogrammed in the offspring after double fertilization. The
101
endosperm DNA of Arabidopsis and rice is globally hypo-methylated (Zemach et al., 5 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
102
2010; Hsieh et al., 2009; Gehring et al., 2009). Analysis of the quantified methylation in
103
rice embryo and endosperm at milky stage by bisulfite-sequencing indicated the globally
104
reduced non-CG methylation and local CG hypo-methylation in endosperm compared to
105
that of in embryo, resembling the phenomenon in Arabidopsis (Zemachet al., 2010). In
106
addition, comparative analysis of transcriptome and methylome between two inbred
107
parents Nipponbare and indica varieties and their hybrid offspring showed the variable
108
epigenetic heritability and strong association between siRNA differences and epimutation
109
levels (Chodavarapu et al., 2012).
110
Like most angiosperms, endosperm development of both rice and Arabidopsis follow
111
the nuclear type development manner, which consists of two main phases: an initial
112
syncytial phase (free nuclear divisions without cytokinesis) followed by a cellularized
113
phase (cell walls form around the free nuclei) (Brown et al., 1996; Hehenberger et al.,
114
2012). Endosperm cellularization shares multiple basic components with somatic
115
cytokinesis (Sorensen et al., 2002), and analysis of the endosperm defective1 (ede1)
116
(Pignocchi et al., 2009) and spätzle mutants (Sorensen et al., 2002) indicated that
117
endosperm cellularization may also have some unique components. However, the key
118
components and regulatory mechanism of endosperm cellularization are still largely
119
unknown.
120
Integrative analysis of epigenetics, especially genome-wide DNA methylation and
121
histone modification, and transcriptomics will help to elucidate the molecular mechanism
122
underlying seed development and may contribute to the yield improvement by molecular
123
breeding. To further determine and investigate the functions of dynamic DNA methylome
124
during seed development, we profiled the DNA methylation genome-wide in a series of
125
developmental stages of rice embryo and endosperm. Results showed that endosperm is
126
predominantly hypo-methylated around short DNA-TEs, short SINEs and non-TE genes,
127
and methylation of non-TE genes showed higher variation during early seed development.
128
There is an obvious global de-methylation of non-TE genes and LTRs during
129
cellularization. Methylation of LTR is most stable, which shows typical negative
130
association with expressions of surrounding genes. The candidate non-TE genes
131
regulated by DNA methylation are involved in multiple processes. Moreover, genes
132
related to cell cycle and hormone response, especially those responses to auxin and
133
abscisic acid stimulus, are up-regulated during endosperm cellularization.
134 6 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
135
7 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
136
RESULTS
137 138
Global DNA Methylation Patterns of Developing Rice Seeds
139
To characterize the dynamic DNA methylation during rice seed development,
140
methylcytosine immunoprecipetation followed by Illumina sequencing (MeDIP-seq) was
141
performed using Zhonghua11 (ZH11) embryo at three developmental stages (3, 6, 9 days
142
after pollination, DAP) and endosperm at four stages (2, 3, 6, 9 DAP) (Supplemental Fig.
143
S1). These developmental stages correspond to the key stages/events during embryo and
144
endosperm development. Embryonic shoot apical meristem (SAM) differentiates after 3
145
DAP (late globular stage) and first leaf primordium emerges at 5-6 DAP. Seed organs
146
enlarge and morphogenesis completes at 9-10 DAP (Itoh et al., 2005). In endosperm,
147
following fertilization, free nuclear division without cytokinesis occurs at initial syncytial
148
phase and then cell wall forms at 3 DAP. Cellularization process completes at 6 DAP
149
followed by endoreduplication at 8-10 DAP (Agarwal et al., 2011).
150
Genome-wide DNA methylation profiles were obtained and calculation of the read
151
intensity around all annotated genes revealed the higher level of DNA methylation in
152
embryo compared to endosperm across all examined developmental stages (Fig. 1A and
153
Supplemental Fig. S2), which is similar as the previous studies from Arabidopsis
154
torpedo-stage seeds (Gehring et al., 2009) or rice seed at milky stage (Zemach et al.,
155
2010). Analysis of the methylation status at promoter region of two randomly selected
156
genes (LOC_Os01g59300 and LOC_Os04g18610) by bisulfite sequencing confirmed the
157
results by MeDIP-seq (Supplemental Fig. S3). In addition, the differentially methylated
158
genes identified by MeDIP-seq dataset were compared with the data of bisulfite
159
sequencing using rice embryo and endosperm at milky stage and results showed that
160
among 69 embryo-hypermethylated genes (Zemach et al., 2010), 51 genes present similar
161
methylation status (high methylation in embryo and significantly decreased methylation
162
level in endosperm), further confirming the consistency and accuracy of the MeDIP-seq
163
data.
164
DNA methylation regions characterized by enriched MeDIP-seq reads were firstly
165
detected to analyze the distribution of DNA methylation in relation to gene annotations.
166
Genome is divided into consecutive 1 kb bins and those overlapping with read enriched
167
regions are recorded. Results showed that embryo has more methylation regions (56,145
168
regions on average) compared to endosperm (47,400 regions on average) (Supplemental 8 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
169
Table S1). In addition, number of methylated regions in intragenic regions is similar in
170
endosperm and embryo, while more intergenic regions are methylated in embryo (Fig.
171
1B). Analysis of the read distribution around TSS (transcriptional start site) and TES
172
(transcriptional end site) showed the high methylation level in embryo, which is more 9 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
173
obvious in intergenic regions (Fig. 1C). The hyper-methylation in intergenic region of
174
embryo possibly contributes to maintaining the high stability of embryo genome.
175 176
Hyper-Methylation of TE Genes in Embryo and Hypo-Methylation of Small
177
DNA-TEs in Endosperm
178
To compare the DNA methylation profile around TE genes and non-TE genes, patterns
179
of read distribution around TSS and TES were plotted for 15,839 TE genes and 40,147
180
non-TE genes, annotated by MSU version 7, respectively. Analysis showed that in both
181
endosperm and embryo, TE genes are highly methylated compared with non-TE genes
182
across all examined developmental stages. Gene body of TE genes are higher methylated
183
compared to intergenic regions, while non-TE genes showed lower methylation in gene
184
body than intergenic regions (Supplemental Fig. S4A). Hyper-methylation of TE genes is
185
responsible for suppression of active transposons.
186
Notably, non-TE genes in endosperm are significantly less methylated than those in
187
embryo (Supplemental Fig. S4B, C) and a total of 5,048 non-TE genes are consistently
188
highly methylated in embryo compared to endosperm across all examined 3
189
developmental stages. Functional analysis of these highly methylated genes indicated
190
that majority of the enriched biological processes are closely associated with biosynthetic
191
and metabolic processes (Supplemental Table S2). Lipids act as important components of
192
membrane, genes with low methylation status in endosperm are enriched in steroid
193
biosynthetic process (25 genes), lipid biosynthetic process (18 genes) and cell wall
194
modification (12 genes), suggesting the involvement of lipid metabolism in endosperm
195
cellularization. In addition, enrichment of amino acid biosynthetic process (tryptophan,
196
phenylalanine and valine), ER to Golgi vesicle-mediated transport (8 genes) and
197
extracellular polysaccharide biosynthetic process implies a relationship between
198
de-methylation of endosperm and synthesis/accumulation of storage compounds during
199
seed development.
200
Four major types of TEs are identified by scanning rice genome, i.e. 40,126
201
DNA-TEs whose transposition does not involve an RNA intermediate, and 3 classes of
202
retro-transposons, including 17,267 LTRs, 2,609 SINES and 550 LINES (Supplemental
203
Table S3). Among these TEs, LTR and DNA-TE are most abundant and account for
204
12.11% and 5.60% of rice genome. Analysis showed that LTR has correspondingly more
205
methylation signals compared to other TEs (Fig. 2A), and that only DNA-TEs and SINEs 10 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
206
present significant hypo-methylation in endosperm, while other TEs didn’t show obvious
207
difference in global methylation level between endosperm and embryo (Fig.2B),
208
suggesting the distinct mechanism of de-methylation in endosperm for different types of
209
TEs. Previous studies showed that short TEs are predominantly de-methylated in 11 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
210
Arabidopsis endosperm (Ibarra et al., 2012), and consistently, short DNA-TEs and short
211
SINEs have significantly lower methylation level across all examined developmental
212
stages of rice endosperm (Fig. 3). Short DNA-TE accounts for 87.28% of all DNA-TEs,
213
hypo-methylation of which indicates that de-methylation of DNA-TE is wide-spread in 12 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
214
endosperm.
215 216
Dynamic DNA Methylomes during Early Seed Development
217
Investigation of the dynamic methylation by comparing DNA methylation level between
218
each consecutive developmental stage using MAnorm (Shao et al., 2012) showed that
219
numbers of differentially methylated regions in embryo are only 1/4 - 1/2 of those in
220
endosperm (Fig. 4A), indicating the relatively stable methylome of embryo. Of the
221
differentially methylated regions during embryo development, ~24% occurs in exon,
222
while DNA methylation regions detected in embryo samples is only about 14% on
223
average (Figs. 1B and 4B). Additionally, a slightly higher percentage of differential
224
methylation regions in intron is also observed. Previous study in rice showed that DNA
225
methylation in gene body has a larger effect on transcription than methylation in
226
promoter region (Li et al., 2008). High percentage of DNA methylation in exon suggests
227
a more crucial role of DNA methylation in regulating gene transcription during embryo
228
development.
229
Further examination of the dynamic DNA methylation levels by calculating the
230
variance coefficient showed that in both embryo and endosperm, non-TE genes have the
231
most significant changes, while LTR is the most stable element with slight changes of
232
DNA methylation across developmental stages (Fig. 4C).
233
In Arabidopsis, it has been assumed that de-methylation of DNA-TEs in central cell
234
could facilitate the production of 24-nt small RNAs, which are transported to egg cell to
235
suppress transposons by increasing the methylation in embryo (Hsieh et al., 2009; Ibarra
236
et al., 2012). If this is the case, the hypo-methylated TEs in early stage of rice endosperm
237
should result in the higher methylation level of TEs in embryo at later developmental
238
stages. Indeed, statistical analysis revealed the significant enrichment of highly
239
methylated regions of all TEs in embryo (3DAP) compared to that in endosperm (3DAP),
240
or at late embryo (9DAP) compared to that at early embryo (3DAP) (Supplemental Table
241
S4).
242 243
DNA Hyper-Methylation is Crucial for Suppressing Surrounding Genes during
244
Seed Development
245
Further, DNA methylome was compared with the gene transcriptome of embryo and
246
endosperm at different developmental stages (Xue et al., 2012) to assess the effects of 13 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
247
DNA methylation on gene expression during seed development. Calculation of the
248
average methylation read intensity showed that the extent of DNA methylation in gene
249
body is obviously negatively correlated with the gene expression (Supplemental Fig.
250
S5A). 14 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
251
Four genesets including genes stably hypo-methylated or hyper-methylated, and genes
252
with stably high or low expression across seed development were extracted to investigate
253
the long-term role of DNA methylation in transcription/expression of surrounding genes.
254
Being consistent with the previous report in rice (He et al., 2010), Fisher Exact test
255
showed that in both embryo and endosperm, hyper-methylated genes are enriched in low
256
expressed genes, but not vice versa (Supplemental Fig. S5B). Further analysis showed
257
that hyper-methylated genes account for ~20% of low expressed genes in embryo and 17%
258
low expressed genes in endosperm, while the hyper-methylated genes are not enriched
259
among the highly-expressed genes (Supplemental Fig. S5C), indicating that DNA
260
hyper-methylation is responsible for gene expression suppression during seed
261
development at different stages, but the hypo-methylation is not directly related to gene
262
expression. Meanwhile, 1847 TE genes were identified in microarray dataset and
263
comparison of the expression and methylation status showed that hyper-methylation of
264
TE genes is obviously correlated with whose low expression. Analysis by Fisher Exact
265
test showed that in both embryo and endosperm, the hyper-methylated TE genes are
266
enriched in low expressed genes.
267
The DNA methylation level in promoter region (1-kb upstream of TSS) does not show
268
obviously negative correlation as in other regions (Supplemental Fig. S5A). We
269
hypothesized that this may due to the effect from some outliers, and thus plotted the
270
methylation level for all genes on microarray ranked by expression intensity (from low to
271
high). The read intensity gradually increased in gene bodies but not promoter regions
272
(Fig. 4D). It’s possible that promoter regions are the major sites for transcriptional
273
regulation by diverse mechanisms, while the methylation status in regions downstream of
274
TSS is responsible for maintenance of the expression level. Inspired by a similar pattern
275
previously reported for MITE TEs in rice (Zemachet al., 2010), we plotted the DNA
276
methylation profile around TSS and TES for different types of TEs. Analysis revealed
277
that DNA-TEs showed negative correlation with the transcription of genes surrounding
278
DNA-TEs in gene body and downstream region, but not with the regions upstream of
279
TSS (Supplemental Fig. S5D). In contrast, methylation of LTRs is negatively correlated
280
with the expression of genes surrounding LTRs around the whole intragenic region
281
(Supplemental Fig. S5E). Given that LTR is hyper-methylated (Fig.2A), and the
282
hyper-methylated genes are enriched in low expressed genes (Supplemental Fig. S5B), it
283
is hypothesized that LTR is closely associated with stable repression of surrounding 15 Downloaded from www.plantphysiol.org on July 18, 2015 - Published by www.plant.org Copyright © 2015 American Society of Plant Biologists. All rights reserved.
284
genes.
285 286
Genes Regulated by DNA Methylation Involve in Multiple Processes
287
Genes that are potentially negatively regulated by DNA methylation were identified by
288
analyzing the differential gene expression and methylation based on comparisons
289
between each consecutive stage. After calculating the Pearson correlation between
290
expression change and methylation change of each gene, 318 genes with negative
291
association (Pearson correlation coefficient20 were kept for further analysis. Syntenic regions between ZH11 and
581
93-11 were detected for comparison (Supplemental Fig. S8). Briefly, homologous
582
regions of sequences shared by two genome were obtained by pairwise alignment using
583
BLASTN (Version 2.2.25, Altschul et al., 1997) and alignments with E-value