Global dissection of alternative splicing uncovers transcriptional diversity in tissues and associates with the flavonoid pathway in tea plant (Camellia sinensis)

Abstract

Background

Alternative splicing (AS) regulates mRNA at the post-transcriptional level to change gene function in organisms. However, little is known about the AS and its roles in tea plant (Camellia sinensis), widely cultivated for making a popular beverage tea.

Results

In our study, the AS landscape and dynamics were characterized in eight tissues (bud, young leaf, summer mature leaf, winter old leaf, stem, root, flower, fruit) of tea plant by Illumina RNA-Seq and confirmed by Iso-Seq. The most abundant AS (~ 20%) was intron retention and involved in RNA processes. The some alternative splicings were found to be tissue specific in stem and root etc. Thirteen co-expressed modules of AS transcripts were identified, which revealed a similar pattern between the bud and young leaves as well as a distinct pattern between seasons. AS events of structural genes including anthocyanidin reductase and MYB transcription factors were involved in biosynthesis of flavonoid, especially in vegetative tissues. The AS isoforms rather than the full-length ones were the major transcripts involved in flavonoid synthesis pathway, and is positively correlated with the catechins content conferring the tea taste. We propose that the AS is an important functional mechanism in regulating flavonoid metabolites.

Conclusion

Our study provides the insight into the AS events underlying tea plant’s uniquely different developmental process and highlights the important contribution and efficacy of alternative splicing regulatory function to biosynthesis of flavonoids.

Electronic supplementary material

The online version of this article (10.1186/s12870-018-1497-9) contains supplementary material, which is available to authorized users.

Background

The perennial tea plant (Camellia sinensis) is an important economic woody crop, and is cultivated from tropical to temperate regions [1]. The leaves of the plant are used for making tea beverage, which is popular due to good taste, attractive aroma and health-promoting effects [2]. Flavonoids are major secondary metabolites present in tea plant which closely relate to the rich flavors of tea infusions [3]. The levels of secondary metabolites in different tissues of tea plant change with different developmental stages. These changes were reported to be regulated by the expression of mRNA transcripts [1, 4, 5].

Alternative splicing (AS) is alternative splice site selection, which results in the generation of multiple isoforms from precursor-mRNA (pre-mRNA) transcripts [6, 7]; these AS transcripts are universally present in plants and animals [8]. AS greatly contributes to transcriptome and proteome diversities, and may result in differences in protein stability, sub-cellular localization or function [7, 9]. In plants, increasing numbers of studies have reported large numbers of genes that undergo extensive AS under environmental stresses and during plant developmental processes, such as heat stress, flowering timing and the circadian clock [10–14]. Furthermore, analyses of comprehensive transcriptomes from different tissues indicate that almost every intron-containing gene generates multiple tissue-specific splice variants [15, 16]. In Arabidopsis, an abnormally expressed flower-specific YUCCA4 transcript isoform could encode a protein that is localized to the endoplasmic reticulum (ER), differing from the cytosol localization of its ubiquitously expressed transcript [17]. In addition, alternative splicing of zinc-induced facilitator-like 1 (ZIFL1) yields two isoforms; ZIFL1.1 influences cellular auxin efflux and polar auxin transport in roots, whereas ZIFL1.3 regulates stomatal movement. Moreover, AS events are involved in secondary metabolite biosynthesis pathways during plant development, such as phenolic acid, phylloquinone, salicylic acid and shikimate biosynthesis pathways [18–20]. Isochorismate synthase (ICS) is an enzyme necessary for salicylic acid (SA) biosynthesis, and the emergence of AS isoforms affects its activity in vivo; splice variants of ICS could not complement the function of the full length enzyme in transgenic Arabidopsis [19]. Despite some studies having explored AS-mediated regulation of flavonoid biosynthesis [21, 22], very little is known of how and by which isoforms modulation occurs in the expression of related structural genes and transcription factors in the flavonoid biosynthesis pathway.

This study used genome-wide analysis of Illumina and SMRT RNA-Seq data from eight developing tissues, and we observed that a large number of genes were alternatively spliced in tea plant. Furthermore, a number of tissue-specific AS genes were identified by WGCNA analysis, and these may play important roles in tissue development. Many AS events were observed in flavonoid biosynthesis pathway gene transcripts, and the majority were highly abundant in vegetative tissues. Unexpectedly, the expression analysis indicated some flavonoid-derived AS transcripts were predominantly expressed in leaf relative to their full-length transcripts; also, their expression levels were highly correlated with the accumulation of catechins, suggesting these AS transcripts may function as major transcripts instead of their full-length counterparts. Together, these results highlight the underlying alternative splicing based regulatory mechanisms in tea plant development, and improve our understanding of the diverse functions of AS.

Results

Identification and classification of AS events

To identify AS events in the transcriptome, we generated 522,811,493 clean Illumina RNA-seq reads and 80,217 SMRT subreads from eight tissues of tea plant. In our study, a total of 69,569 AS events were identified in 15,869 AS genes, which accounted for 46.8% of the total expressed genes from eight samples (Table 1). An additional 9503 AS events in 4227 genes were identified from 136 Mb SMRT Iso-Seq reads, which were generated from the same eight tissue-derived samples. Among these AS, 2875 AS genes were present in both sequencing data sets (Fig. 1a). Furthermore, approximately 81.9% of AS transcripts were uniquely identified by Illumina sequencing;the lower number of AS transcripts detected from SMRT data may have resulted from the low coverage of the SMRT sequencing reads generated here. Therefore, we focused on the AS analysis detected in Illumina data set and used the SMRT sequencing data to verify some of the AS isoforms.

Table 1.

Statistics of different AS events obtained from the Illumina and SMRT libraries generated in this study

	Illumina		SMRT
	Number	Percentage	Number	Percentage
Alternative donor site	9752	14.01%	836	8.79%
Alternative acceptor site	10,850	15.59%	1404	14.77%
Intron retention	14,140	20.32%	3079	32.40%
Exon skipping	10,477	15.05%	626	6.58%
Others	24,350	35.03%	3558	37.46%
Total events	69,569	100.00%	9503	100.00%

‘Illumina’ indicates the Illumina paired-end RNA-seq data generated from the eight tea plant tissues. ‘SMRT’ indicates the SMRT sequencing data generated from the eight pools of tea plant tissues. “Percentage” indicates the ratio of each AS type to the total events for each library

Fig. 1.

Comparative analysis of different AS isoforms among eight tissues of tea plant. a Venn diagram showing the common and unique AS genes determined with Illumina and SMRT sequencing. b Distribution of different types of alternative splicing events in eight tissues. c Overlap of AS genes among eight tissues. d KEGG pathway enrichment for AS genes from eight tissues. The area of each colored circle is proportional to the size scale of genes involved in each pathway. The color indicates the q-value, and the x-axis is the Rich Factor

Among the 69,569 AS events, 20.32% resulted from intron retention (IR), followed by alternative acceptor sites (AA, 15.59%), exon skipping (ES, 15.05%), and alternative donor sites (AD, 14.01%) (Table 1). A similar pattern of AS abundance was observed in each tissue (Fig. 1b, Additional file 1: Table S1). These patterns were in accordance with previous observations in Arabidopsis thaliana, Brachypodium distachyon and Glycine max [23–25]. In addition, a large proportion of “other” AS types were detected in all samples, which may result from multiple splicing modes occurring on a single transcript. For example, the CUFF.13109.2 transcript was classified into“other” type due to the combination of AD and IR in the bud.

Among 15,869 AS genes, 4969 genes were collectively identified in eight tissues via comparisons of transcript isoforms, and a small proportion of AS genes were detected within individual tissue (Fig. 1c). To understand the biological processes in which AS genes might be involved, all the AS genes were analyzed with KEGG enrichment by bi-directional best hit method, and were highly enriched in processes related to RNA, such as RNA transport, spliceosome, RNA degradation (Fig. 1d). Similar biological processes were enriched among the individual tissues (data not shown).In addition, multiple splicing factors were identified from these conserved AS genes, such as eight serine/arginine-rich (SR), five U2af-like and three 3B-likeprotein genes (Additional file 1: Table S2).

Co-expression network identified across tissues

A co-expression network across eight tissues of tea plant was constructed with WGCNA (v1.29) based on pairwise correlation between common AS expression patterns [26]. Thirteen distinct modules were observed (Additional file 1: Figure S1), and the genes within these modules were partially correlated with distinct tissue types (Fig. 2). Notably, AS genes in six co-expression modules (black, blue, pink, green, red and brown) are highly expressed in a single tissue type (r > 0.8, P < 10^− 3); hence, these six modules could be regarded as tissue specific clusters of genes. For example, the red module contains 1645 AS genes identified as FL-specific (flower), and the brown module contains 2223 FR-specific (fruit) AS genes.The splicing mode was not uniform across the six modules. The predominant AS type IR was only the most prevalent in root and fruit modules, but remained common in the other modules (Fig. 3a, Additional file 1: Table S3).

Fig. 2.

Co-expression network analysis. WGCNA analysis of AS genes in eight tissues. Each column corresponds to a specific tissue (B, bud; YL, young leaf; SML, summer mature leaf; WOL, winter old leaf; S, stem; R, root; FL, flower; FR, fruit). Each row corresponds to a module. The number of genes in each module is indicated on the left box. The color and number of each cell at the row-column intersection indicates the correlation coefficient between the module and the tissue type. A high degree of correlation between a specific module and the tissue type is indicated by dark red

Fig. 3.

Characteristic analysis of tissue-specific AS genes. a Distribution of different types of alternative splicing events in eight tissues. b Heat map showing the transcript level in FPKM of each tissue-specific AS gene from the pink (stem) module, green (root) module, red (flower) module and brown (fruit), respectively. Red and blue colors represent high and low levels of transcript abundance, respectively

It’s remarkable that some AS genes within the turquoise module were co-expressed in the bud and the young leaf, and were enriched in the process of cell growth and death by KEGG enrichment analysis (data not shown). In addition, two structural genes involved in the flavonoid pathway with significant co-expression within the turquoise module are leucoanthocyanidin reductase (TEA027557.1) and anthocyanidin reductase (TEA022952.1). Conjecturally, as these AS genes are co-expressed in young leaf and bud, they are likely biologically relevant and their correlation may underscore the developmental relationship and molecular similarities between these two tissues. Notably, summer mature leaf (SML) and winter old leaf (WOL) are clearly divided into two distinct modules (black and blue), and no credible co-expression relationship was observed between the two tissues. Therefore, to certain degrees, the identification of SML-specific (black module) and WOL-specific (blue module) modules (Fig 3a) probably suggested that the developing divergence between AS expression patterns in summer mature leaf and winter old leaf.

Identification of tissue-specific AS gene expression

Characterization of tissue or cell-specific genes provides a foundation for unraveling their molecular mechanisms during tissue development. We further analyzed the expression patterns of other tissue-specific AS genes. Several key regulatory genes were identified and found to be highly expressed. Three class III HD-ZIP genes (ATHB15) and three KNOX1 transcription factors (KNAT2) were identified in the stem-specific module (Fig. 3), which were reported to be directly associated with vascular development and maintenance of meristematic identity in Arabidopsis [27]. The occurrences of AS events with these genes suggest that they could perform a redundant function during stem development. In Arabidopsis, the nuclear protein, ALF (ABERRANT LATERAL ROOT FORMATION) is required to maintain xylem pole pericycle cells in lateral root formation [28]. A paralog to ALF was found within the root-specific module, and had similar root-specific expression profiles. The ABCE classes of genes are the best characterized flower developmental genes, and their expression patterns were highly stereotypic in flower [29]. Not surprisingly, we found one APETALA1 (AP1) and three AGAMOUS (AG) splice variants which were highly expressed in the flower-specific module. Interestingly, AP1 and AG gene were detected in the fruit-specific module, but none of the A, B, C and E genes were expressed in other modules. This conservation of AS patterns in the ABCE model genes between flower and fruit suggest similar regulation in these two reproductive tissues.

Detection and characterization of AS in genes involved in the flavonoid pathway

To better understand the putative influence of alternative splicing on secondary metabolism, we investigated AS transcripts of genes which are directly or indirectly involved in characteristic secondary metabolism pathways including the flavonoid, theanine and caffeine (Additional file 1: Figures S2, S3 and S4) [2, 3, 30, 31]. Interestingly, nearly all the structural flavonoid pathway genes generated multiple splicing variants, and some transcripts were tissue-specific (Additional file 1: Table S4). For instance, three different AS isoforms were produced for dihydroflavonol 4-reductase (TEA024764.1), which performs a vital rate-limiting function; these were produced in three different tissues (bud, young leaf and stem). Notably, the occurrence of flavonoid biosynthesis related AS were frequently observed in vegetative organs such as bud and leaf (Fig. 4). Furthermore, an AS type called mutually exclusive exon (MXE) was found in some transcripts, which were classified into the “other” category in the previous transcriptome analysis. To exclude the possibility that these transcripts were experimental or informatic artefacts, RT-PCR was performed for validation using primers flanking the AS site (Fig. 4).

Fig. 4.

Alternatively spliced isoforms of flavonoid-related genes in tea plant. The red asterisk indicates the position of PTC. The full-length and AS isoforms on the gel images are denoted with red and black triangles, respectively. The orange solid dot indicates the targeted AS transcripts were identified in specific tissues (B, bud; YL, young leaf; SML, summer mature leaf; WOL, winter old leaf; S, stem; R, root; FL, flower; FR, fruit)

The MYB transcription factors are key regulators of the synthesis of flavonoid-derived compounds [32]. We identified 12 homologs of flavonoid-related MYB genes in tea plant and constructed a phylogenetic tree (Additional file 1: Figure S5). Meanwhile, among these CsMYBs, four were regulated by AS, which may serve as an activator to mediate down-stream genes. Similarly, AS of two CsMYBs (TEA012130.1 and TEA029017.1) was detected in the Illumina data set and verified by RT-PCR (Fig. 4). Additionally, considering of possibility of unstable AS transcripts, five key flavonoid-related AS isoforms were selected to validate by RT-PCR, using independent samples (bud, young leaf, root and flower). The results showed that the five isoforms were detected, although the expression levels varied with different tissues (Additional file 1: Figure S6).

Sequence analysis unexpectedly revealed that most of the detected AS events resulted in introduction of a premature stop codon (PTC). Therefore they could be degraded through the nonsense-mediated decay (NMD) mechanism (Fig. 4) [33]. However, the non-PTC AS transcripts could be translated into truncated proteins despite gaining or losing parts of conserved domains, which could also lead to variants with functional differences.

Expression of flavonoid-related AS genes was correlated with catechins

To explore the extent and conservation of AS among flavonoid-related genes, we evaluated the expression patterns of these genes in different tissues using RNA-Seq data. Among eight flavonoid-related genes, six full-length transcripts were highly expressed in vegetative tissues, while the corresponding AS transcripts also followed similar expression patterns (Fig. 5a). The AS transcripts of two genes, CsPAL (TEA003146.1) and CsMYB (TEA012130.1), were the predominant isoforms in vegetative organs while their full-length transcripts exhibited completely opposite expression patterns, suggesting a potential functional divergence between these two transcripts. Furthermore, Sanger sequencing, IGV visualization and semi-quantitative RT-PCR confirmed the expression patterns of predicted alternative splicing events in the CsPAL and CsMYB transcripts (Fig. 5b).

Fig. 5.

Expression patterns of flavonoid-related AS transcripts. a Heat map showing the FPKM of non-AS and AS transcripts in eight tissues (B, bud; YL, young leaf; SML, summer mature leaf; WOL, winter old leaf; S, stem; R, root; FL, flower; FR, fruit). Red and blue colors represent high and low transcript abundance, respectively. b Read coverage graphs of TEA003146.1 and TEA012130.1 derived from eight tissues viewed in IGV and validated by semi-quantitative RT-PCR. The red dotted boxes indicate the positions of alternative splicing. The RT-PCR bands are shown in the gel images beneath the graphs. The CsGAPDH gene was used as the internal control

To further explore the relationship between transcript expression of flavonoid-related genes and metabolites, a pearson’s correlation analysis between the expression levels in FPKM of the AS genes and total catechins’ contents indicated significant correlations (P < 0.05) between total catechins and CsFLSα (TEA016596.1), CsLARα (TEA026475.1), CsMYBα (TEA029017.1), CsLARβ (TEA026475.1), CsMYBβ (TEA012130.1) and CsMYBβ (TEA029017.1) (Table 2). Surprisingly, CsANRα (TEA029026.1) and β were not correlated with total catechins, but were significantly correlated (P < 0.05) with epigallocatechin (EGC). In addition, we observed high correlation (P < 0.01) of epicatechin-3-gallate (ECG) with AS transcripts of CsMYBβ (TEA012130.1) and CsPALβ (TEA003146.1), but no correlation with the full-length transcripts.

Table 2.

Correlation analysis between flavonoid-related transcripts and catechins

Pearson Correlation
		2,3-trans-catechin		2,3-cis-catechin
Annotation	Gene ID	C	GC	EC	EGC	ECG	EGCG	Total catechins
	Non-AS transcripts
CsPALα	TEA003146.1	0	0.05	− 0.25	− 0.25	− 0.53	− 0.57	− 0.58
Cs4CLα	TEA031662.1	− 0.16	− 0.2	0.05	0.26	0.49	0.80*	0.69
CsFLSα	TEA016596.1	0.3	−0.03	0.2	0.04	0.92**	0.77*	0.71*
CsDFRα	TEA024764.1	−0.09	− 0.25	− 0.22	0.03	0.46	0.44	0.35
CsLARα	TEA026475.1	0.26	0.25	0.51	0.43	0.84**	0.92**	0.93**
CsANRα	TEA029026.1	−0.09	0.53	0.32	0.81*	−0.32	0	0.2
CsMYBα	TEA012130.1	0.04	0.36	0.16	0.44	−0.62	−0.39	−0.25
CsMYBα	TEA029017.1	0.17	0.09	0.42	0.29	0.81*	0.93**	0.89**
	AS transcripts
CsPALβ	TEA003146.1	0.41	0.08	0.27	0.08	0.87**	0.71*	0.69
Cs4CLβ	TEA031662.1	0.14	0.74*	0.65	0.91**	0.06	0.3	0.53
CsFLSβ	TEA016596.1	−0.03	−0.21	0.16	−0.04	0.84**	0.78*	0.68
CsDFRβ	TEA024764.1	0.14	0.02	−0.02	0.04	0.22	0.35	0.3
CsLARβ	TEA026475.1	0.32	0.06	0.32	0.11	0.87**	0.85**	0.79*
CsANRβ	TEA029026.1	−0.02	0.63	0.41	0.86**	−0.29	− 0.01	0.22
CsMYBβ	TEA012130.1	0.31	0.03	0.32	0.12	0.92**	0.87**	0.82*
CsMYBβ	TEA029017.1	0.3	0.38	0.58	0.56	0.78*	0.87**	0.93**

The numbers indicate the Pearson correlation coefficient. The asterisks indicate the significance level (*P < 0.05, **P < 0.01) based on a Tukey’s honestly significant difference test

Discussion

Alternative splicing (AS) is known as an alternative expressional regulatory mechanism, which commonly occurs in plants to control gene expression, and potentially results in modified protein function [6, 7]. Little is known about underlying AS events in tea plant (Camellia sinensis). Although Illumina based RNA-Seq transcriptome in different tissues of tea plant had been reported [4, 34], the crucial and more flexible regulatory roles of AS were never addressed. There have been limitations considering that many previous analyses were performed without genome data and with a low depth of sequencing data; AS events in SMRT RNA-Seq data had been proposed in three secondary metabolism pathways [18]. In this study, we investigated the total AS events in different tissues of tea plant by combining Illumina and SMRT RNA-Seq technologies. AS events of key tissue-specific and flavonoid-derived transcripts were observed, and they could play important roles in producing characteristic tea flavor and for regulation of developmental processes. Thus, this study provides novel understanding of the roles of alternative splicing in tea plant.

Among the 15,869 detected genes alternatively spliced in eight tissues, the characterization of AS splicing modes and KEGG enrichment analysis were similar (Fig. 1b, d), indicating the high conservation of AS dynamics during tea plant development; A similar phenomenon was observed in soybean [25]. In contrast, we found that approximately 31.3% of AS genes to be tissue conserved, which was a greater percentage than in soybean [25]. Among these conserved AS genes, three representative types of splicing factors were found, such as eight SR genes (Additional file 1: Table S2). Previous analyses indicated that the SR protein family plays an important role in regulating AS in a tissue-specific and stress-responsive manner [35, 36]. The AS of these highly conserved splicing factors among all tissues implied that they may act as vital regulatory genes in not only self-regulation by AS, but also in regulating AS profiles of other genes during tea plant developmental stages.

Tissue specific AS may have potential functions. Studies in multiple plants have indicated that each organ or tissue has its specific AS transcripts [37, 38], which have been proven to mediate tissue differentiation and promote specialized characteristics; for example, tissue specific AS events affect polyploidy in endosperm and microgametogenesis in pollen [39–41]. In our study, although AS dynamics were conserved between leaves of different developmental stages, a set of AS genes with considerable overlap between bud and young leaf was observed, while large differences were found between SML-specific and WOL-specific AS genes (Fig. 2). This may imply that molecular features are similar between bud and young leaf, whereas SML and WOL have distinct molecular characteristics. Moreover, this speculation was strengthened by previous studies on comparative transcriptome analysis of tea leaf at different developmental stages, which examined expression of full-length transcripts [5, 42]. In addition, multiple tissue-specific AS transcripts were detected by WGCNA analysis, such as those of stem-specific ATHB15 and flower-specific ABCE genes. A tissue-specific non-functional isoform of cyclin in maize played a role in endoreduplication [43]. Thus, the occurrence of tissue specific AS in tea plant could perform various functions during development. Our future investigations will aim to test the potential functions of tissue-specific AS.

The roles of flavonoids in plants include responses to environmental challenges, and flavonoids are contribute to the quality of tea leaves [44, 45]. Great advances had been made in understanding transcriptional regulation of the flavonoid biosynthesis pathway in tea plant, which include the characterization of structural genes and transcription factors. However, the influence of alternative splicing to flavonoid biosynthesis has not been explored. In this study, abundant AS events of structural genes and MYB transcription factors were characterized in the flavonoid biosynthesis pathway (Fig. 4). Interestingly, except for up-regulation of these full-length transcripts during different stages of leaf development, the majority of AS transcripts were enriched in vegetative organs. This unexpected observation may result from active transcription of flavonoid biosynthesis pathway genes and the rapid accumulation of flavonoid in leaf; this observation is in agreement with previous studies that found that the frequency of AS occurrence closely associated with the degree of tissue development and location during plant development [25, 46, 47].

Recent studies on protein LSm5 and SKIP indicated that the presence of AS transcripts generally decrease the levels of the corresponding full-length transcripts [48, 49]. Further expression analysis of flavonoid-derived AS genes showed they exhibit a pattern with low expression levels of AS transcripts relative to those of the full-length transcripts (Fig. 5a), which is agreement with previous reports [50, 51]. However, in contrast, the AS transcripts of CsPAL (TEA003146.1) and CsMYB (TEA012130.1) were predominant and displayed the opposite expression pattern relative to full-length transcripts in different tissues (Fig. 5b); this implies that the two AS transcripts likely serve as largely regulated or functional transcripts instead of the corresponding full-length transcripts. As a precedent for this idea, the circadian clock gene AtZTL was shown to have its AS transcripts translated into authentic functional protein instead of its full-length transcripts [52, 53]. In addition, this hypothesis is supported by correlation analysis with the accumulation of catechins, which showed the two AS transcripts of CsMYBβ and CsPALβ are significantly correlated with epicatechin-3-gallate (ECG) levels (Table 2). Collectively, our data propose and support the hypothesis and that achievement of gene function not only depends on conventional full-length transcripts, but also on the extraordinary AS transcripts, which may act as major mediators at the post-transcriptional level during plant development (Fig. 6). The key determinants are the specific functions that the predominant AS transcripts confer and the dynamic equilibrium of abundance between the full-length and AS transcripts.

Fig. 6.

Hypothetical model for functions of alternative splicing in tea plant. The biosynthesis of flavonoid is active in vegetative tissues (bud, young leaf, summer mature leaf and winter old leaf) but passive in other tissues (stem, root, flower and fruit). As details of the processes of flavonoid biosynthesis, the full-length and AS transcripts synergistically functions and positive and negative relationship are exhibited between them at transcription level

Conclusions

The genome-scaled approach to investigation of tea plant development is surprisingly informative in uncovering transcriptional diversity mediated by alternative splicing. Based on the results of this study, we propose that the tissue-specific gene function and flavonoid biosynthesis could potentially be affected by an alternative splicing regulatory mechanism. This study provides the insight into the AS events underlying tea plant’s uniquely different developmental process and highlights the important contribution and efficacy of alternative splicing regulatory function to biosynthesis of flavonoids.

Methods

Plant resources

Tea plants (C. sinensis cv. Shuchazao) were grown in the 916 Tea Plantation in Shucheng County (latitude 31.3 N, longitude 117.2E above sea level), Anhui Province, China under natural conditions. Eight samples of different tissues were collected from exactly the same tea plant. The tissues sampled were taken as follows: bud, young leaf, summer mature leaf, winter old leaf, stems, flowers, fruits and roots. Bud, young leaf, summer mature leaf and stems were collected on June 15,2015; Winter old leaf were collected on November 13, 2015; Flowers, fruits and roots were collected on October 12, 2015.

At the same periods in 2016, the four tissues (bud, young leaf, root and flower) were collected again from exactly the same tea plant. All samples in this study were performed with three biological replicates and all collected samples were immediately frozen in liquid nitrogen and stored at − 80 °C until use.

Transcriptomic data

Single molecule real-time DNA sequencing data was generated from individual pools of eight tea plant tissues (bud, young leaf, mature leaf in summer, old leaf in winter, stem, root, flower and seed) using the PacBio RS II platform (Pacific Biosciences of California, USA; Accession, SRR5460108) and data were reported in our previous study [18]. Additional RNA sequencing data was generated from 90-bp paired-end RNA-Seq reads from the aforementioned eight tissue samples using the Illumina Hiseq 2000 platform (Illumina, USA; Accession, PRJNA274203).

Alignment of reads to the reference tea plant genome and gene expression estimation

We used Bowtie v2.0.6 to build an index of the tea plant (Camellia sinensis) reference genome [54]. The clean RNA sequence data from each sample were mapped to the tea plant (Camellia sinensis) genome using Tophat (v.2.0.11) [55].The maximum alignment number for each read was set to 2, and the inner distance between mating pairs was set to 230 bp. Mapped reads were assembled with Cufflinks v.2.1.1 to generate a genome-based Cufflinks assembly [56]. The rest of the parameters were set as default. We calculated reads per kilobase of exon model per million mapped fragments (FPKM) of genes based on gene length and the number of fragments mapped to corresponding genes [56].

Identification and characterization of alternative splicing

The AS events were extracted from assembly files of different tissues with the software tool AStalavista and visualized by IGV browser [57, 58]. To minimize the mismatch rate, we required that the overhang size had to be greater than 25 bp with at least two reads spanning the junctions. The alternative splicing events were classified by using the gtf files assembled from the Illumina RNA-seq and SMRT sequencing data based on the tool AStalavista [57]. Four major types of AS events, namely IR (AS code: 1^2-,0), ES (AS code:1–2^,0), AA (AS code: 1-,2-), and AD (AS code: 1^,2^), were identified from the output files and counted, respectively.

Co-expression network analysis

WGCNA (v1.29) package in R [26] was employed to construct co-expression networks. All AS genes were used for WGCNA unsigned co-expression network analysis; the corresponding FPKM of each gene was imported into WGCNA. The automatic network construction function block wise Modules values were set as default to construct the modules, except that the power was set to 18, minModuleSize to 30, and merge CutHeight to 0.35. The eigengene value was calculated for each module and used to test the association with each tissue type.

Experimental validation of alternative splicing

Total RNA was isolated from the different tissues of tea plant as described above for RNA-seq. First-strand cDNA was synthesized from total RNA using a PrimeScript RT Reagent Kit (Takara, Japan) referred to the manufacturer’s instructions. Total AS events were validated by RT-PCR according to a reported protocol using P1-P42 primers that were designed based on each AS event.All specific primers are listed in Additional file 1: Table S5.

Gene structures and phylogenetic analysis

The alternative splicing isoforms were viewed using the IGV software and were corrected with the SMRT sequencing reads. Gene structures were analyzed with online website Gene Structure Display Server (GSDS 2.0, http://gsds.cbi.pku.edu.cn/index.php). The phylogenetic relationships of candidate MYBs from Camellia sinensis and other species were analyzed by MEGA 6.0using NJ (Neighbor Joining) method with the following parameters:bootstrap method (1000 replicates), Poisson model, Uniform rates, Complete deletion [32].

Extraction and HPLC analysis of catechins

Catechins were extracted from the eight tissues according to the reported method [59, 60]. 0.1 g of the freeze-dried tissues were ground in liquid nitrogen with a mortar and pestle and extracted with 3 mL 80% methanol in an ultrasonic sonicator for 10 min at 4 °C. After centrifugation at6,000 rpm for 10 min, the residues were re-extracted twice as described above. The supernatants were combined and diluted with 80% methanol to a volume of 10 mL. The obtained supernatants were filtered through a 0.22 μm organic membrane before HPLC analysis. Catechins contents were determined using Waters 2695 HPLC system with a reverse phase C18 column (Phenomenex 250 mm × 4.6 mm, 5 μm). The eluent was composed of 0.17% (v/v) acetic acid (A) in water and 100% acetonitrile (B) at 1 mL/min and the effluent was monitored at 278 nm [60]. The filtered sample (10 μL) was injected into the HPLC system for analysis. Samples from the eight tissues were analyzed in triplicate.

Abbreviations

AA

Alternative acceptor sites

AD

Alternative donor sites

AS

Alternative splicing

B

Bud

ECG

Epicatechin-3-gallate

EGC

Epigallocatechin

ES

Exon skipping

FL

Flower

FR

Fruit

HPLC

High performance liquid chromatography

IR

Intron retention

KEGG

Kyoto Encyclopedia of Genes and Genomes

MXE

Mutually exclusive exon

NMD

Nonsense-mediated decay

Pre-mRNA

Precursor-mRNA

PTC

Premature stop codon

R

Root

RNA-Seq

RNA sequencing

RT-PCR

Reverse transcriptase polymerase chain reaction

S

Stem

SML

Summer mature leaf

SR

Serine/arginine-rich

WGCNA

Weighted Gene Co-Expression Network Analysis

WOL

Winter old leaf

YL

Young leaf

Authors’ contributions

WCL, ZJY, WXW designed this study, performed the experiments, interpreted the results, and wrote the manuscript. XQS and ZSQ participated in the data analyse. TYL designed and conducted the RT-PCR experiment and HPLC analysis.

Ethics approval and consent to participate

Not applicable

Consent for publication

Not Applicable

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.