Accepted Manuscript Comparative transcriptome analysis reveals defense-related genes and pathways against downy mildew in Vitis amurensis grapevine Xinlong Li, Jiao Wu, Ling Yin, Yali Zhang, Junjie Qu, Jiang Lu PII:
S0981-9428(15)30045-0
DOI:
10.1016/j.plaphy.2015.06.016
Reference:
PLAPHY 4217
To appear in:
Plant Physiology and Biochemistry
Received Date: 13 March 2015 Revised Date:
12 June 2015
Accepted Date: 24 June 2015
Please cite this article as: X. Li, J. Wu, L. Yin, Y. Zhang, J. Qu, J. Lu, Comparative transcriptome analysis reveals defense-related genes and pathways against downy mildew in Vitis amurensis grapevine, Plant Physiology et Biochemistry (2015), doi: 10.1016/j.plaphy.2015.06.016. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACCEPTED MANUSCRIPT Comparative transcriptome analysis reveals defense-related genes and pathways against downy
2
mildew in Vitis amurensis grapevine
3
Xinlong Li1,†, Jiao Wu1,†, Ling Yin1, Yali Zhang1, Junjie Qu2, Jiang Lu1,*
4
1 The Viticulture and Enology Program, College of Food Science and Nutritional Engineering, China
5
Agricultural University, Beijing, China.
6
2 Guangxi crop genetic improvement and biotechnology key lab, Guangxi academy of Agricultural
7
science, Guangxi, China.
8
†
These authors contributed equally to this work
9
*
Corresponding author
E-mail address: [email protected] (J. Lu)
11
16 17 18 19 20 21 22
EP
15
AC C
14
TE D
12 13
SC
M AN U
10
RI PT
1
2
ACCEPTED MANUSCRIPT Abstract
24
Downy mildew (DM), caused by oomycete Plasmopara viticola (Pv), can lead to severe damage to
25
Vitis vinifera grapevines. V. amurensis has generally been regarded as a DM resistant species. However,
26
when V. amurensis ‘Shuanghong’ were inoculated with Pv strains ‘ZJ-1-1’ and ‘JL-7-2’, the former led
27
to obvious DM symptoms (compatible), while the latter did not develop any DM symptoms but
28
exhibited necrosis (incompatible). In order to underlie molecular mechanism in DM resistance,
29
mRNA-seq based expression profiling of ‘Shuanghong’ was compared at 12, 24, 48 and 72 hours post
30
inoculation (hpi) with these two strains. Specific genes and their corresponding pathways responsible
31
for incompatible interaction were extracted by comparing with compatible interaction. In the
32
incompatible interaction, 37 resistance (R) genes were more expressed at the early stage of infection
33
(12 hpi). Similarly, genes involved in defense signaling, including MAPK. ROS/NO, SA, JA, ET and
34
ABA pathways, and genes associated with defense-related metabolites synthesis, such as
35
pathogenesis-related genes and phenylpropanoids/stilbenoids/flavonoids biosynthesizing genes, were
36
also activated mainly during the early stages of infection. On the other hand, Ca2+ signaling and
37
primary metabolism, such as photosynthesis and fatty acid synthesis, were more repressed after ‘JL-7-2’
38
challenge. Further quantification of some key defense-related factors, including phytohormones,
39
phytoalexins and ROS, generally showed much more accumulation during the incompatible interaction,
40
indicating their important roles in DM defense. In addition, a total of 43 and 52 RxLR effectors were
41
detected during ‘JL-7-2’ and ‘ZJ-1-1’ infection processes, respectively.
42
Keywords: Vitis amurensis, downy mildew, interaction, comparative transcriptome, resistance
43
mechanism
44
AC C
EP
TE D
M AN U
SC
RI PT
23
3
ACCEPTED MANUSCRIPT 1.
Introduction
46
The oomycete pathogen Plasmopara viticola (Pv) causes downy mildew (DM), one of the most
47
important grapevine diseases worldwide (Yu et al., 2012). Pv is a typical obligate biotrophic pathogen
48
that obtains nutrients from living cells of hosts to complete its life cycle through specialized structure
49
known as haustoria (Gessler et al., 2011). By releasing motile zoospores, Pv enters the grapevine leaves
50
through stomata (Kiefer et al., 2002). Colonization is achieved by the germination of zoospores,
51
intercellular mycelia growth in host cells, and the formation of haustoria. Similar to other oomycete
52
pathogens of plants, during the infection process Pv can secrete various effectors which are involved in
53
the establishment of compatible/incompatible interactions between grapevines and different Pv strains
54
(Casagrande et al., 2011). According to the ‘gene-for-gene’ theory, the incompatible interaction takes
55
place when the effectors secreted by the pathogens are specifically recognized by cognate resistance
56
proteins of hosts. Otherwise, the compatible interaction happens (Dangl and Jones, 2006).
TE D
M AN U
SC
RI PT
45
Although European V. vinifera varieties are highly susceptible to Pv, some American and Oriental
58
Vitis species show different levels of resistance (Yu et al., 2012). At least thirteen loci resistant to Pv
59
(Rpv) have been mapped on various chromosomes of the Vitis genome (Merdinoglu et al., 2003;
60
Fischer et al., 2004; Wiedemann-Merdinoglu et al., 2006; Welter et al., 2007; Bellin et al., 2009;
61
Marguerit et al., 2009; Blasi et al., 2011; Moreira et al., 2011; Schwander et al., 2012; Venuti et al.,
62
2013). Up to now, only one major resistance loci (Rpv1) has been cloned (Feechan et al., 2013). By
63
conventional hybridization, previous attempts to introgress these resistance traits into cultivated V.
64
vinifera have resulted in some resistant hybrids (http://www.vivc.de), while their acceptance by wine
65
makers is low due to disrupting the desired phenotype of these traditional varieties. Future breeding
66
efforts are to couple strong resistance with desirable wine attributes utilizing availability of
AC C
EP
57
4
ACCEPTED MANUSCRIPT high-throughput selection markers and the grapevine genome sequence (Di Gaspero et al., 2007; Jaillon
68
et al., 2007). In addition, expression profiling analysis of genes and proteins has been applied to study
69
the interactions between grapevine and Pv pathogen (Polesani et al., 2008; Polesani et al., 2010; Wu et
70
al., 2010; Legay et al., 2011; Figueiredo et al., 2012; Milli et al., 2012).
RI PT
67
Transcriptome and proteome analysis of grapevines during the infection by Pv will enable us to
72
understand the molecular mechanism through discovering important genes and pathways related to DM
73
resistance, eventually come up with novel strategies to breed DM-resistance grape cultivars. In this
74
regard, we chose V. amurensis ‘Shuanghong’, an intraspecific hybrid of V. amurensis ‘Tonghua-3’ and
75
‘Shuangqing’, for a transcriptome analysis to mine DM resistant genes and pathways. A very
76
cold-hardy ‘Shuanghong’ is characterized by having good wine quality and very strong resistance to
77
DM disease (Song et al., 1998; Yu et al., 2012). However, our recent study also revealed that this
78
grapevine had compatible interaction with certain Pv strains. The availability of incompatible and
79
compatible interactions between ‘Shuanghong’ and Pv strains will be extremely useful for us to identify
80
genes and pathways associated with DM resistance. Therefore, a genome-wide comparative
81
transcriptome analysis was performed by using mRNA-Seq method. Our long term goal is to
82
understand the resistance mechanism of grapevines to DM disease, and eventually to identify, isolate
83
and utilize the DM resistant genes for grape cultivar improvement.
84
2.
85
2.1. Plant materials, Pv strains and pathogen infection
86
V. amurensis ‘Shuanghong’ (DM resistant) and V. vinifera ‘Thompson Seedless’ (DM susceptible) were
87
grown in a greenhouse with a photoperiod of 16 h light/8 h darkness at 25 ℃. A total of 206 strains
88
were obtained from the main grape-growing regions in China using single sporangiophore transfer
AC C
EP
TE D
M AN U
SC
71
Materials and methods
5
ACCEPTED MANUSCRIPT method (Wong and Wilcox, 2000). These strains were inoculated on leaf discs of the five grapevine
90
hosts with different DM resistance and their virulence was determined according to Gómez-Zeledón et
91
al. (2013) (data not shown). According to the virulence assessment, the strains ‘JL-7-2’ and ‘ZJ-1-1’
92
were selected for the RNA-seq inoculation and were further axenically propagated on ‘Thompson
93
Seedless’.
RI PT
89
The third to fifth unfolded leaves from the shoot apex were inoculated with a suspension of 1×105
95
sporangia per ml (Wu, et al., 2010). The leaves inoculated by ‘JL-7-2’ or ‘ZJ-1-1’ were harvested at 12,
96
24, 48, and 72 hpi, respectively, for microscopy observations (Koch and Slusarenko, 1990),
97
mRNA-Seq analysis and qualification of SA, JA, stilbenes and H2O2. ‘Shuanghong’ leaves treated with
98
water were sampled as control in the transcriptome analyses. For each strain at any time point, three
99
leaves collected from five individual grapevines were polled as a biological replicate. Two independent
M AN U
SC
94
biological replicates were sequenced for each treatment.
101
2.2. Construction of Illumina library and sequencing
102
Total RNA was extracted from leaves by a modified CTAB method (Iandolino et al., 2004). Three µg
103
RNA per sample was used as input material for library construction. Libraries were constructed using
104
NEBNext Ultra Directional RNA Library Prep Kit for Illumina (NEB, Ispawich, USA) and four index
105
codes were added to attribute sequences to each sample. The clustering of the index-coded samples was
106
performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-c Bot-HS (Illumina).
107
After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2000 platform
108
conducted by Novogene (Beijing, China, http://www.novogene.cn/) and 100 bp paired-end reads were
109
generated.
110
2.3. Data analysis
AC C
EP
TE D
100
6
ACCEPTED MANUSCRIPT 111
Raw sequenced reads were processed using Trimmomatic software (Bolger et al., 2014). All
112
high-quality
113
(ftp://ftp.jgi-psf.org/pub/compgen/phytozome/v9.0/Vvinifera) using the default setting of TopHat
114
2.0.11 (Trapnell et al., 2009). Uniquely mapped reads were counted in HTSeq 0.6.1 software (Anders et
115
al., 2014) and normalized with NOISeq package version 2.6.0 (Tarazona et al., 2011). Pearson’s
116
correlation coefficients between biological replicates for each sample were calculated in R (logiciel).
117
Likewise, the high-quality reads from each sample were also mapped to Pv genome (our unpublished
118
data) for detecting the expression of RxLR effectors. Differentially expressed genes (DEGs) were
119
identified using NOISeq package version 2.6.0. Absolute values of log2- (fold change) ≥ 1 and q-value
120
≥ 0.99 were set as the criteria for defining DEGs. All DEGs were subjected to agriGO database
121
(http://bioinfo.cau.edu.cn/agriGO/) to investigate the gene ontology (GO) enrichment (Fisher, P-value
122
< 0.05) with singular enrichment analysis. Kyoto Encylopedia of Genes and Genomics (KEGG)
123
enrichment analysis (hypergeometric test, P-value < 0.05) of DEGs was performed by KOBAS2
124
software (Xie et al., 2011). Transcription factors (TFs) were identified and classified based on plant
125
transcription factor database (http://planttfdb.cbi.pku.edu.cn/). Hierarchical clustering of DEGs was
126
computed in MeV software (www.tm4.org/mev.html).
127
2.4. Validation of real-time RT-PCR
128
Primers of all selected genes were designed using Primer 5 software (Premier Biosoft Interpairs, Palo
129
Alto, CA) (Table S1). All real-time RT-PCR reactions were conducted with the same RNA samples
130
taken for RNA-seq analysis using protocol established in our lab (Wu et al., 2010). Vitis elongation
131
factor1-α (EF1-α) was used as an internal control to normalize all data (Monteiro et al., 2013). The fold
132
change in gene expression was estimated using threshold cycles by the 2-△△CT method (Livak and
of
each
sample
were
mapped
to
grapevine
reference
genome
AC C
EP
TE D
M AN U
SC
RI PT
reads
7
ACCEPTED MANUSCRIPT Schmittgen, 2001).
134
2.5. Detection of SA and JA levels
135
The levels of SA and JA were measured referring to Pan et al. (2010). Each sample was ground to
136
powder in liquid nitrogen and transferred to a 2 ml Eppendorf tube containing 50 µl of 1 µg·ml-1
137
internal
138
(2-propanol/H2O/concentrated HCl = 2:1:0.002, vol/vol/vol). The tubes were kept on a shaker at a
139
speed of 100 rpm at 4 ℃ for 30 min, then added 1 ml dichloromethane and shaken for another 30 min.
140
The mixture was centrifuged at 13,000 rpm at 4 ℃ for 5 min and approximately 900 µl of lower phase
141
was transferred into a new tube, concentrated by N-Evap system with nitrogen system and redissolved
142
in 100 µl methanol. The sample solutions (50 µl) were injected into HLPC-MS system to detect SA and
143
JA contents (Varian, CA, USA).
144
2.6. Stilbenes quantification
145
Stilbenes extraction and quantification were carried out with the same samples referring to Pezet et al.
146
(2003). Each sample (3 mg) was placed in a 1.5 ml Eppendorf tube containing 50 µl methanol,
147
extracted at 60 ℃ for 10 min on a shaker at a speed of 200 rpm, then placed on ice for 5 min. The
148
methanol extracts (30 µl) were detected with HLPC-MS system (Varian, CA, USA).
149
2.7. Analysis of H2O2 level
150
In order to analyze the accumulation of H2O2, the samples were incubated in 10 mg·ml-1 3,
151
3-diaminobenzi-dine (DAB) solution (Sigma, CA, USA) at 25 ℃ for 8 h, then transferred into 95%
152
ethanol and boiled for 15 min until the chlorophyll were completely decolorized (Thordal-Christensen
153
et al., 1997). The decolorized samples were further soaked in 2.5 g·ml-1 chloral hydrate solution to
154
reduce the background. The H2O2 level was quantified by Photoshop and Image J.
(D4-SA
and
D2-JA,
respectively)
and
500
µl
extraction
solvent
AC C
EP
TE D
M AN U
SC
standards
RI PT
133
8
ACCEPTED MANUSCRIPT 3.
Results
156
3.1. ‘Shuanghong’ inoculated with compatible and incompatible Pv strains
157
Pathogenicity was evaluated among a total of 206 Pv strains collected from major viticulture areas in
158
China. Among them, Pv strains ‘ZJ-1-1’ and ‘JL-7-2’ presented similar DM symptoms on susceptible
159
‘Thompson Seedless’ (Fig. 1A and 1C). However, these two strains gave absolute different results on
160
DM-resistant ‘Shuanghong’. The former led to obvious DM symptom (compatible) while the latter did
161
not develop DM symptoms but exhibited necrosis (incompatible) (Fig. 1D and 1B). Microscopic
162
observation showed that zoospores of the compatible strain ‘ZJ-1-1’were localized in the vicinity of
163
stomata at 12 hpi (Fig. 2A). Germ tubes were observed in the intercellular spaces at 24 hpi (Fig. 2B).
164
From 24 to 48 hpi, the germ tubes elongated along the intercellular spaces and formed primary hyphae
165
(Fig. 2C). A number of branched hyphae with many haustoria and sporangia were visible by 72 hpi (Fig.
166
2D and 2E). In contrast, the incompatible strain ‘JL-7-2’ did not develop germ tubes, hyphae and
167
haustoria during the entire infection and interaction processes (Fig. 2F, 2G and 2H). However, necrosis
168
was clearly observed at 48 hpi (Fig. 2H). Based on the observation of key developmental stage after Pv
169
infection, leaf samples at 12, 24, 48 and 72 hpi were collected for the subsequent RNA-Seq analysis.
170
3.1. Illumina sequencing and mapping to the reference genome
171
A survey using Illumina sequencing technology was carried out to analyze gene expression profiles of
172
‘Shuanghong’. The raw reads generated from Hiseq 2000 were filtered with Illumina passed filter call.
173
After discarding the low-quality reads, more than 21 million high-quality reads were obtained for each
174
sample. About 85% of these reads were mapped to the grapevine reference genome. Of which 76%
175
were mapped to single locations and 9% were mapped to multiple locations, respectively (Table 1).
176
Among the reads with unique locations, 79% distributed in exon regions, while 16% located in intron
AC C
EP
TE D
M AN U
SC
RI PT
155
9
ACCEPTED MANUSCRIPT and 5% were in intergenic regions (Fig. S1). High Pearson’s correlation coefficients (R2) of FPKM
178
distribution between the two biological replicates for each sample were detected (R2=0.92-0.97,
179
p
S0981-9428(15)30045-0
DOI:
10.1016/j.plaphy.2015.06.016
Reference:
PLAPHY 4217
To appear in:
Plant Physiology and Biochemistry
Received Date: 13 March 2015 Revised Date:
12 June 2015
Accepted Date: 24 June 2015
Please cite this article as: X. Li, J. Wu, L. Yin, Y. Zhang, J. Qu, J. Lu, Comparative transcriptome analysis reveals defense-related genes and pathways against downy mildew in Vitis amurensis grapevine, Plant Physiology et Biochemistry (2015), doi: 10.1016/j.plaphy.2015.06.016. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACCEPTED MANUSCRIPT Comparative transcriptome analysis reveals defense-related genes and pathways against downy
2
mildew in Vitis amurensis grapevine
3
Xinlong Li1,†, Jiao Wu1,†, Ling Yin1, Yali Zhang1, Junjie Qu2, Jiang Lu1,*
4
1 The Viticulture and Enology Program, College of Food Science and Nutritional Engineering, China
5
Agricultural University, Beijing, China.
6
2 Guangxi crop genetic improvement and biotechnology key lab, Guangxi academy of Agricultural
7
science, Guangxi, China.
8
†
These authors contributed equally to this work
9
*
Corresponding author
E-mail address: [email protected] (J. Lu)
11
16 17 18 19 20 21 22
EP
15
AC C
14
TE D
12 13
SC
M AN U
10
RI PT
1
2
ACCEPTED MANUSCRIPT Abstract
24
Downy mildew (DM), caused by oomycete Plasmopara viticola (Pv), can lead to severe damage to
25
Vitis vinifera grapevines. V. amurensis has generally been regarded as a DM resistant species. However,
26
when V. amurensis ‘Shuanghong’ were inoculated with Pv strains ‘ZJ-1-1’ and ‘JL-7-2’, the former led
27
to obvious DM symptoms (compatible), while the latter did not develop any DM symptoms but
28
exhibited necrosis (incompatible). In order to underlie molecular mechanism in DM resistance,
29
mRNA-seq based expression profiling of ‘Shuanghong’ was compared at 12, 24, 48 and 72 hours post
30
inoculation (hpi) with these two strains. Specific genes and their corresponding pathways responsible
31
for incompatible interaction were extracted by comparing with compatible interaction. In the
32
incompatible interaction, 37 resistance (R) genes were more expressed at the early stage of infection
33
(12 hpi). Similarly, genes involved in defense signaling, including MAPK. ROS/NO, SA, JA, ET and
34
ABA pathways, and genes associated with defense-related metabolites synthesis, such as
35
pathogenesis-related genes and phenylpropanoids/stilbenoids/flavonoids biosynthesizing genes, were
36
also activated mainly during the early stages of infection. On the other hand, Ca2+ signaling and
37
primary metabolism, such as photosynthesis and fatty acid synthesis, were more repressed after ‘JL-7-2’
38
challenge. Further quantification of some key defense-related factors, including phytohormones,
39
phytoalexins and ROS, generally showed much more accumulation during the incompatible interaction,
40
indicating their important roles in DM defense. In addition, a total of 43 and 52 RxLR effectors were
41
detected during ‘JL-7-2’ and ‘ZJ-1-1’ infection processes, respectively.
42
Keywords: Vitis amurensis, downy mildew, interaction, comparative transcriptome, resistance
43
mechanism
44
AC C
EP
TE D
M AN U
SC
RI PT
23
3
ACCEPTED MANUSCRIPT 1.
Introduction
46
The oomycete pathogen Plasmopara viticola (Pv) causes downy mildew (DM), one of the most
47
important grapevine diseases worldwide (Yu et al., 2012). Pv is a typical obligate biotrophic pathogen
48
that obtains nutrients from living cells of hosts to complete its life cycle through specialized structure
49
known as haustoria (Gessler et al., 2011). By releasing motile zoospores, Pv enters the grapevine leaves
50
through stomata (Kiefer et al., 2002). Colonization is achieved by the germination of zoospores,
51
intercellular mycelia growth in host cells, and the formation of haustoria. Similar to other oomycete
52
pathogens of plants, during the infection process Pv can secrete various effectors which are involved in
53
the establishment of compatible/incompatible interactions between grapevines and different Pv strains
54
(Casagrande et al., 2011). According to the ‘gene-for-gene’ theory, the incompatible interaction takes
55
place when the effectors secreted by the pathogens are specifically recognized by cognate resistance
56
proteins of hosts. Otherwise, the compatible interaction happens (Dangl and Jones, 2006).
TE D
M AN U
SC
RI PT
45
Although European V. vinifera varieties are highly susceptible to Pv, some American and Oriental
58
Vitis species show different levels of resistance (Yu et al., 2012). At least thirteen loci resistant to Pv
59
(Rpv) have been mapped on various chromosomes of the Vitis genome (Merdinoglu et al., 2003;
60
Fischer et al., 2004; Wiedemann-Merdinoglu et al., 2006; Welter et al., 2007; Bellin et al., 2009;
61
Marguerit et al., 2009; Blasi et al., 2011; Moreira et al., 2011; Schwander et al., 2012; Venuti et al.,
62
2013). Up to now, only one major resistance loci (Rpv1) has been cloned (Feechan et al., 2013). By
63
conventional hybridization, previous attempts to introgress these resistance traits into cultivated V.
64
vinifera have resulted in some resistant hybrids (http://www.vivc.de), while their acceptance by wine
65
makers is low due to disrupting the desired phenotype of these traditional varieties. Future breeding
66
efforts are to couple strong resistance with desirable wine attributes utilizing availability of
AC C
EP
57
4
ACCEPTED MANUSCRIPT high-throughput selection markers and the grapevine genome sequence (Di Gaspero et al., 2007; Jaillon
68
et al., 2007). In addition, expression profiling analysis of genes and proteins has been applied to study
69
the interactions between grapevine and Pv pathogen (Polesani et al., 2008; Polesani et al., 2010; Wu et
70
al., 2010; Legay et al., 2011; Figueiredo et al., 2012; Milli et al., 2012).
RI PT
67
Transcriptome and proteome analysis of grapevines during the infection by Pv will enable us to
72
understand the molecular mechanism through discovering important genes and pathways related to DM
73
resistance, eventually come up with novel strategies to breed DM-resistance grape cultivars. In this
74
regard, we chose V. amurensis ‘Shuanghong’, an intraspecific hybrid of V. amurensis ‘Tonghua-3’ and
75
‘Shuangqing’, for a transcriptome analysis to mine DM resistant genes and pathways. A very
76
cold-hardy ‘Shuanghong’ is characterized by having good wine quality and very strong resistance to
77
DM disease (Song et al., 1998; Yu et al., 2012). However, our recent study also revealed that this
78
grapevine had compatible interaction with certain Pv strains. The availability of incompatible and
79
compatible interactions between ‘Shuanghong’ and Pv strains will be extremely useful for us to identify
80
genes and pathways associated with DM resistance. Therefore, a genome-wide comparative
81
transcriptome analysis was performed by using mRNA-Seq method. Our long term goal is to
82
understand the resistance mechanism of grapevines to DM disease, and eventually to identify, isolate
83
and utilize the DM resistant genes for grape cultivar improvement.
84
2.
85
2.1. Plant materials, Pv strains and pathogen infection
86
V. amurensis ‘Shuanghong’ (DM resistant) and V. vinifera ‘Thompson Seedless’ (DM susceptible) were
87
grown in a greenhouse with a photoperiod of 16 h light/8 h darkness at 25 ℃. A total of 206 strains
88
were obtained from the main grape-growing regions in China using single sporangiophore transfer
AC C
EP
TE D
M AN U
SC
71
Materials and methods
5
ACCEPTED MANUSCRIPT method (Wong and Wilcox, 2000). These strains were inoculated on leaf discs of the five grapevine
90
hosts with different DM resistance and their virulence was determined according to Gómez-Zeledón et
91
al. (2013) (data not shown). According to the virulence assessment, the strains ‘JL-7-2’ and ‘ZJ-1-1’
92
were selected for the RNA-seq inoculation and were further axenically propagated on ‘Thompson
93
Seedless’.
RI PT
89
The third to fifth unfolded leaves from the shoot apex were inoculated with a suspension of 1×105
95
sporangia per ml (Wu, et al., 2010). The leaves inoculated by ‘JL-7-2’ or ‘ZJ-1-1’ were harvested at 12,
96
24, 48, and 72 hpi, respectively, for microscopy observations (Koch and Slusarenko, 1990),
97
mRNA-Seq analysis and qualification of SA, JA, stilbenes and H2O2. ‘Shuanghong’ leaves treated with
98
water were sampled as control in the transcriptome analyses. For each strain at any time point, three
99
leaves collected from five individual grapevines were polled as a biological replicate. Two independent
M AN U
SC
94
biological replicates were sequenced for each treatment.
101
2.2. Construction of Illumina library and sequencing
102
Total RNA was extracted from leaves by a modified CTAB method (Iandolino et al., 2004). Three µg
103
RNA per sample was used as input material for library construction. Libraries were constructed using
104
NEBNext Ultra Directional RNA Library Prep Kit for Illumina (NEB, Ispawich, USA) and four index
105
codes were added to attribute sequences to each sample. The clustering of the index-coded samples was
106
performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-c Bot-HS (Illumina).
107
After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2000 platform
108
conducted by Novogene (Beijing, China, http://www.novogene.cn/) and 100 bp paired-end reads were
109
generated.
110
2.3. Data analysis
AC C
EP
TE D
100
6
ACCEPTED MANUSCRIPT 111
Raw sequenced reads were processed using Trimmomatic software (Bolger et al., 2014). All
112
high-quality
113
(ftp://ftp.jgi-psf.org/pub/compgen/phytozome/v9.0/Vvinifera) using the default setting of TopHat
114
2.0.11 (Trapnell et al., 2009). Uniquely mapped reads were counted in HTSeq 0.6.1 software (Anders et
115
al., 2014) and normalized with NOISeq package version 2.6.0 (Tarazona et al., 2011). Pearson’s
116
correlation coefficients between biological replicates for each sample were calculated in R (logiciel).
117
Likewise, the high-quality reads from each sample were also mapped to Pv genome (our unpublished
118
data) for detecting the expression of RxLR effectors. Differentially expressed genes (DEGs) were
119
identified using NOISeq package version 2.6.0. Absolute values of log2- (fold change) ≥ 1 and q-value
120
≥ 0.99 were set as the criteria for defining DEGs. All DEGs were subjected to agriGO database
121
(http://bioinfo.cau.edu.cn/agriGO/) to investigate the gene ontology (GO) enrichment (Fisher, P-value
122
< 0.05) with singular enrichment analysis. Kyoto Encylopedia of Genes and Genomics (KEGG)
123
enrichment analysis (hypergeometric test, P-value < 0.05) of DEGs was performed by KOBAS2
124
software (Xie et al., 2011). Transcription factors (TFs) were identified and classified based on plant
125
transcription factor database (http://planttfdb.cbi.pku.edu.cn/). Hierarchical clustering of DEGs was
126
computed in MeV software (www.tm4.org/mev.html).
127
2.4. Validation of real-time RT-PCR
128
Primers of all selected genes were designed using Primer 5 software (Premier Biosoft Interpairs, Palo
129
Alto, CA) (Table S1). All real-time RT-PCR reactions were conducted with the same RNA samples
130
taken for RNA-seq analysis using protocol established in our lab (Wu et al., 2010). Vitis elongation
131
factor1-α (EF1-α) was used as an internal control to normalize all data (Monteiro et al., 2013). The fold
132
change in gene expression was estimated using threshold cycles by the 2-△△CT method (Livak and
of
each
sample
were
mapped
to
grapevine
reference
genome
AC C
EP
TE D
M AN U
SC
RI PT
reads
7
ACCEPTED MANUSCRIPT Schmittgen, 2001).
134
2.5. Detection of SA and JA levels
135
The levels of SA and JA were measured referring to Pan et al. (2010). Each sample was ground to
136
powder in liquid nitrogen and transferred to a 2 ml Eppendorf tube containing 50 µl of 1 µg·ml-1
137
internal
138
(2-propanol/H2O/concentrated HCl = 2:1:0.002, vol/vol/vol). The tubes were kept on a shaker at a
139
speed of 100 rpm at 4 ℃ for 30 min, then added 1 ml dichloromethane and shaken for another 30 min.
140
The mixture was centrifuged at 13,000 rpm at 4 ℃ for 5 min and approximately 900 µl of lower phase
141
was transferred into a new tube, concentrated by N-Evap system with nitrogen system and redissolved
142
in 100 µl methanol. The sample solutions (50 µl) were injected into HLPC-MS system to detect SA and
143
JA contents (Varian, CA, USA).
144
2.6. Stilbenes quantification
145
Stilbenes extraction and quantification were carried out with the same samples referring to Pezet et al.
146
(2003). Each sample (3 mg) was placed in a 1.5 ml Eppendorf tube containing 50 µl methanol,
147
extracted at 60 ℃ for 10 min on a shaker at a speed of 200 rpm, then placed on ice for 5 min. The
148
methanol extracts (30 µl) were detected with HLPC-MS system (Varian, CA, USA).
149
2.7. Analysis of H2O2 level
150
In order to analyze the accumulation of H2O2, the samples were incubated in 10 mg·ml-1 3,
151
3-diaminobenzi-dine (DAB) solution (Sigma, CA, USA) at 25 ℃ for 8 h, then transferred into 95%
152
ethanol and boiled for 15 min until the chlorophyll were completely decolorized (Thordal-Christensen
153
et al., 1997). The decolorized samples were further soaked in 2.5 g·ml-1 chloral hydrate solution to
154
reduce the background. The H2O2 level was quantified by Photoshop and Image J.
(D4-SA
and
D2-JA,
respectively)
and
500
µl
extraction
solvent
AC C
EP
TE D
M AN U
SC
standards
RI PT
133
8
ACCEPTED MANUSCRIPT 3.
Results
156
3.1. ‘Shuanghong’ inoculated with compatible and incompatible Pv strains
157
Pathogenicity was evaluated among a total of 206 Pv strains collected from major viticulture areas in
158
China. Among them, Pv strains ‘ZJ-1-1’ and ‘JL-7-2’ presented similar DM symptoms on susceptible
159
‘Thompson Seedless’ (Fig. 1A and 1C). However, these two strains gave absolute different results on
160
DM-resistant ‘Shuanghong’. The former led to obvious DM symptom (compatible) while the latter did
161
not develop DM symptoms but exhibited necrosis (incompatible) (Fig. 1D and 1B). Microscopic
162
observation showed that zoospores of the compatible strain ‘ZJ-1-1’were localized in the vicinity of
163
stomata at 12 hpi (Fig. 2A). Germ tubes were observed in the intercellular spaces at 24 hpi (Fig. 2B).
164
From 24 to 48 hpi, the germ tubes elongated along the intercellular spaces and formed primary hyphae
165
(Fig. 2C). A number of branched hyphae with many haustoria and sporangia were visible by 72 hpi (Fig.
166
2D and 2E). In contrast, the incompatible strain ‘JL-7-2’ did not develop germ tubes, hyphae and
167
haustoria during the entire infection and interaction processes (Fig. 2F, 2G and 2H). However, necrosis
168
was clearly observed at 48 hpi (Fig. 2H). Based on the observation of key developmental stage after Pv
169
infection, leaf samples at 12, 24, 48 and 72 hpi were collected for the subsequent RNA-Seq analysis.
170
3.1. Illumina sequencing and mapping to the reference genome
171
A survey using Illumina sequencing technology was carried out to analyze gene expression profiles of
172
‘Shuanghong’. The raw reads generated from Hiseq 2000 were filtered with Illumina passed filter call.
173
After discarding the low-quality reads, more than 21 million high-quality reads were obtained for each
174
sample. About 85% of these reads were mapped to the grapevine reference genome. Of which 76%
175
were mapped to single locations and 9% were mapped to multiple locations, respectively (Table 1).
176
Among the reads with unique locations, 79% distributed in exon regions, while 16% located in intron
AC C
EP
TE D
M AN U
SC
RI PT
155
9
ACCEPTED MANUSCRIPT and 5% were in intergenic regions (Fig. S1). High Pearson’s correlation coefficients (R2) of FPKM
178
distribution between the two biological replicates for each sample were detected (R2=0.92-0.97,
179
p
