Chinese fir (Cunninghamia lanceolata), a commercially important tree for the timber and pulp industry, is widely distributed in southern China and northern Vietnam, but its large and complex genome has hindered the development of genomic resources. Few efforts have focused on analysis of the modulation of transcriptional networks in vascular cambium during the transition from active growth to dormancy in conifers.
Here, we used Illumina sequencing to analyze the global transcriptome alterations at the different stages of vascular cambium development in Chinese fir.
By analyzing dynamic changes in the transcriptome of vascular cambium based on our RNA sequencing (RNA-Seq) data at the dormant, reactivating and active stages, many potentially interesting genes were identified that encoded putative regulators of cambial activity, cell division, cell expansion and cell wall biosynthesis and modification. In particular, the genes involved in transcriptional regulation and hormone signaling were highlighted to reveal their biological importance in the cambium development and wood formation.
Our results reveal the dynamics of transcriptional networks and identify potential key components in the regulation of vascular cambium development in Chinese fir, which will contribute to the in-depth study of cambial differentiation and wood-forming candidate genes in conifers.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Wood represents one of the most important sources of energy on earth and is an environmentally acceptable future alternative to fossil fuel resources. The perennial stem growth habit of most tree species is characterized by secondary growth that results in a cumulative increase in girth during each growth cycle. This is achieved by the cell division activity of the vascular cambium, with subsequent differentiation of secondary xylem toward the inside of the cambium and secondary phloem toward the outside (Schrader et al., 2004; Wang et al., 2007). Recent advances in the understanding of these processes have revealed that wood formation is under highly regulated genetic control, notably at the transcriptional level (Hertzberg et al., 2001; Wang et al., 2007). Therefore, a better understanding of the regulation of cambial activity and wood formation is essential for improving the quality of wood using molecular biological techniques.
The recent development of novel high-throughput sequencing technologies (i.e. next-generation sequencing (NGS), such as Solexa/Illumina RNA-Seq (RNA sequencing) and digital gene expression (DGE)) have provided an opportunity to address wood formation-associated genes in tree species by de novo assembly or mapping and quantification of transcriptomes. Transcriptome sequencing is an efficient means to generate functional genomic data for nonmodel organisms or those with genome characteristics prohibitive to whole-genome sequencing (Raherison et al., 2012). Production and analysis of expressed sequence tags (ESTs) from wood-forming tissues have increased our understanding of the gene regulation involved in wood formation (Allona et al., 1998; Kirst et al., 2003; Pavy et al., 2005; Foucart et al., 2006; Wang et al., 2010a). In particular, cloning and identification of genes from the cambium of woody plants are important strategies in the investigation of wood formation (Hertzberg et al., 2001; Schrader et al., 2004; Wang et al., 2010a), because the secondary growth in woody plants is initiated from meristematic cambium cells that retain a perpetual cell division ability. In recent years, genomic approaches (e.g. microarray analyses) have been used to elucidate transcriptional networks associated with various stages of activity–dormancy transitions in the model plant poplar (Schrader et al., 2004; Druart et al., 2007; Ruttink et al., 2007; Larisch et al., 2012). Schrader et al. (2004) and Druart et al. (2007) reported a comprehensive analysis to investigate cambial dormancy utilizing cryosection-isolated cambial cells from the woody plant Populus tremula during active growth and dormancy using microarray experiments. However, few transcriptome-level studies have been carried out to date on the transition from active growth to dormancy in the vascular cambium in coniferous species. Such studies would bridge the physiological and anatomical changes during wood formation with the molecular data.
Chinese fir (Cunninghamia lanceolata) is an evergreen conifer in the family of Taxoidiaceae, which is the most commercially important coniferous species widely distributed in southern China and northern Vietnam. Furthermore, Chinese fir is the most important fast-growing timber tree of the warm regions south of the Yangtze River. Because of the limited number of EST sequences and the total absence of genomic sequences available, our knowledge, at the molecular level, of regulation of wood formation in Chinese fir remains poorly understood. In particular, genes expressed only during the transition between stages are likely to control the entire process of wood formation. This prompted us to analyze the transcriptome of isolated cambial tissues during three sequential developmental stages in the nonmodel tree Chinese fir growing under natural conditions. This approach has allowed us to reveal the dynamic changes in the key transcriptional and development networks during the distinct stages of the dormancy–activity cycle. These results provide an insight into the molecular basis underlying the physiological and anatomical changes in dormant, reactivating, and active cambium cells. Furthermore, using cryosectioning to isolate cambial meristem for analysis has allowed a much higher cellular resolution in defining the transcriptional and development profiles in wood formation in gymnosperm. To the best of our knowledge, this study is the first to characterize the transcriptome dynamics of Chinese fir using RNA-Seq, which may serve as a gene expression profile blueprint for cambium development and wood formation in the conifer.
Materials and Methods
Plant material and total RNA isolation
Chinese fir (Cunninghamia lanceolata (Lamb.) Hook, the wide type) trees were grown under standard glasshouse conditions (70% relative humidity and 25 : 18°C, day : night). Fresh leaves, stems and roots of 2-month-old young seedlings were harvested and stored at −80°C until use. Dry seeds were stored at 4°C. Total RNA was extracted from seeds and roots using RNAiso-mate for plant tissue and RNAiso plus (Takara, Dalian, Liaoning, China) according to the manufacturer's instructions. Total RNA from leaves and stems was isolated using Concert Plant RNA Reagent (Invitrogen), following the supplied protocol. RNA quality was verified using a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA), and all four samples had an RNA integrity number (RIN) value of > 8.0. A total of 20 μg RNA was pooled from the four tissues equally for cDNA preparation.
cDNA library preparation and transcriptome sequencing
Illumina sequencing using the GAII platform was performed at the Beijing Genomics Institute (BGI)-Shenzhen, Shenzhen, China (http://www.genomics.cn/index.php) according to the manufacturer's instructions (Illumina, San Diego, CA, USA). Briefly, magnetic beads with oligo(dT) were utilized to isolate poly(A) mRNA following collection of total RNA from Chinese fir tissues. A fragmentation buffer was added to break the mRNA into short fragments. Using these short fragments as templates, a random hexamer primer was used to synthesize first-strand cDNA. Second-strand cDNA was synthesized using buffers, dNTPs, RNase H, and DNA polymerase I. Short fragments were purified with a QiaQuick PCR extraction kit (Qiagen) and resolved with an elution buffer for end repair and by addition of poly(A). Thereafter, the short fragments were connected with sequencing adapters. For PCR amplification, we selected suitable fragments as templates based on the results of agarose gel electrophoresis. Finally, the library was sequenced using an Illumina HiSeq™ 2000. The sequencing data were deposited in the US National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA, http://www.ncbi.nlm.nih.gov/Traces/sra; Wheeler et al., 2008) under accession number SRA053525.
Analysis of Illumina transcriptome sequencing results
De novo assembly was carried out using SOAPdenovo-Trans (version 1.04; http://soap.genomics.org.cn/soapdenovo.html) with the default parameters, except for the K-mer value (Li et al., 2010). After assessing different K-mer sizes, 29-mer yielded the best assembly for the desired application and was chosen to construct the de Bruijn graph. Although this higher value reduced the number of assembled contigs, it increased the reliability and resulted in longer contigs. The contigs without N were obtained by conjoining the K-mers in an unambiguous path. The reads were then mapped back to contigs to construct scaffolds using the paired end information. SOAPdenovo connected the contigs using N to represent unknown sequences between each pair of contigs, and thus scaffolds were made. Paired-end reads were again used for gap-filling of scaffolds to obtain sequences with the least Ns (unknown sequences) and that could not be extended at either end. Such sequences were defined as unigenes used for blast search and annotation against an NCBI nonredundant protein (Nr) database (http://www.ncbi.nlm.nih.gov) and Swiss-Prot protein database (http://www.expasy.ch/sprot) using an E-value cutoff of 10−5. Functional annotation by gene ontology terms (GO, http://www.geneontology.org) was analyzed using the Blast2GO program (Conesa et al., 2005). Unigene sequences were also aligned to the Cluster of Orthologous Groups (COG) database (http://www.ncbi.nlm.nih.gov/COG) to predict and classify possible functions. The Kyoto Encyclopedia of Genes and Genomes Pathway (KEGG; http://www.genome.jp/kegg) annotation were carried out according to the KEGG database (Ogata et al., 1999) using BLASTx with an E-value threshold of 10−5.
Differential gene expression library preparation
Dormant, reactivating, and active cambium samples were obtained from the stems of Chinese fir trees (the wide type, c. 16 yr old, 15 m tall, 12 cm diameter) growing under natural conditions in the Forestry Station at Fuzhou, China (45.44′N, 126.36′E). Samples were collected on 25 December 2010, 20 March 2011, and 1 June 2011, to cover the major stages of the dormancy–activity cycle. Small blocks (4 × 2 × 2 cm3) containing secondary phloem, vascular cambium, and secondary xylem were sampled c. 1.5 m above ground from 20 independent trees at each time point. Samples were flash-frozen in liquid nitrogen and stored at −80°C. Frozen stem segments were trimmed to blocks of c. 30 mm length (axial) and 3.0 mm width (tangential). Tangential sections through the cambial region of the stem were obtained using a cryomicrotome and frozen in liquid nitrogen, as described by Uggla et al. (1996) with modification. Briefly, 30 μm sections were isolated by tangential cryosectioning at −24°C with a Leica CM1850 Cryostat (Leica Microsystems Nussloch GmbH, Nussloch, Germany) equipped with a steel knife. Transverse sections taken from both ends of the specimen and stained with aniline blue were used to locate the position of tangential sections. For each time point, 90 sections were used to prepare total RNA for the construction of cDNA libraries and sequencing using Illumina GA II. Total RNAs from dormant, reactivating, and active cambial meristems, respectively were isolated using Concert Plant RNA Reagent (Invitrogen) following the supplied protocol. RNA integrity was confirmed using a 2100 Bioanalyzer (Agilent Technologies). The short-read data sets are available at the NCBI SRA with the accession number GSE37152.
Anatomical observations of the vascular cambium
Blocks of c. 5 mm3 including secondary phloem, vascular cambium, and secondary xylem cut from the stems of Chinese fir were collected and fixed in 2.5% glutaraldehyde in 100 mM phosphate buffer (pH 7.2). After dehydration through an ethanol series, the samples were embedded into Spurr's resin. Sectioning was performed using a Leica microtome. Sections (1 μm) were mounted on slides, stained with 0.25% (w/v) toluidine blue O (Sigma), and observed under a Zeiss Axioskop 2 Plus microscope equipped with a computer-assisted digital camera. Images were processed using Photoshop (Adobe, San Jose, CA, USA).
Analysis and mapping of Illumina short reads
The RNA-Seq bioinformatic analysis pipeline is shown in the Supporting Information (Fig. S1). Before mapping reads to the reference database, all reads were filtered to remove adaptors, low-quality reads, and reads with unknown bases. The clean read expression distribution was used to evaluate the normality of the whole data. All clean reads were mapped to our transcriptome reference sequences using SOAPaligner/soap2, and mismatches of only 1 bp were considered. The number of unambiguous clean reads for each gene was calculated and then normalized to the number of reads per kilobase per million clean reads (RPKM; Mortazavi et al., 2008), uniquely aligning within each sample.
Bioinformatics for functional annotation of differential gene expression
A rigorous algorithm to identify differentially expressed genes was developed based on the method of Audic & Claverie (1997). The false discovery rate (FDR) was used to determine the threshold of P-value in multiple test and analysis. We used FDR < 0.001 and the absolute value of log2(ratio) ≥ 2 as thresholds to determine the significance of gene expression difference (Benjamini & Yekutieli, 2001). Furthermore, an additional criterion, which involved using only differentially expressed genes (DEGs) with a minimum of fourfold change, was adopted for further analysis.
Total RNAs were isolated from dormant, reactivating, and active cambium, as described earlier. First-strand cDNA synthesis was performed using Superscript II reverse transcriptase (Invitrogen) according to the manufacturer's instructions, using 1 μg of total RNA and oligo(dT) primers. qRT-PCR was performed using a Rotor-Gene 3000 real-time PCR detection system (Qiagen) using SYBR® qPCR Mix (Toyobo, Tokyo, Japan) according to the manufacturer's protocol. Gene-specific primers were designed using the Primer Express software version 3.0 (Applied Biosystems, Foster City, CA, USA; Table S1). Quantitative PCR reactions were conducted in 20 μl volumes containing 2 μl diluted cDNA, 300 nM of each primer, and 10 μl of the Thunderbird SYBR Green PCR Master Mix with the following cycling conditions: 95°C for 2 min, 40 cycles at 95°C for 15 s, 60°C for 15 s, and 72°C for 15 s. After amplification, a thermal denaturing cycle at 95°C for 15 s, 60°C for 15 s, and 95°C for 15 s was carried out to determine the dissociation curves and verify the specificity of the amplifications. All reactions were performed in biological triplicates, and the results were expressed relative to the expression levels of an internal reference gene, glyceraldehyde 3-phosphate dehydrogenase (GAPDH, CV170251), in each sample using the method (Livak & Schmittgen, 2001).
High-throughput transcriptome sequencing and read assembly
To maximize the number of genes included in the transcriptome, a cDNA sample was prepared from an equal mixture of total RNA isolated from seeds, roots, stems, and leaves, and sequenced using the Illumina high-throughput sequencing platform. After stringent quality checks and data cleaning, we obtained 27 666 672 reads containing a total of 2490 000 480 nucleotides (2.49 Gb). The average read size, Q20 percentage (sequencing error rate < 1%), and GC (guanine + cytosine) percentage were 90 bp, 94.51%, and 45.04%, respectively. Based on the high-quality reads, 417 105 contigs were assembled with an average length of 144 bp. With paired-end joining and gap-filling, the contigs were further assembled into 84 980 scaffolds with an average length of 385 bp, including 6548 scaffolds larger than 1000 bp. After employed local assembly with the unmapped end to fill in the small gaps within the scaffolds, the de novo assembly yielded 59 669 unigenes with an average length of 497 bp (Table 1). The size distribution of these scaffolds and unigenes is shown in Fig. S2 and Table S2. To demonstrate the quality of sequencing data, we randomly selected six unigenes and designed six pairs of primers for RT-PCR, products that were confirmed by Sanger sequencing.
Table 1. Summary for the Chinese fir (Cunninghamia lanceolata) transcriptome
Total number of reads
27 666 672
Total nucleotide length
2490 000 480 bp
Average read length
Total number of contigs
Mean length of contigs
Total number of scaffolds
Mean length of scaffolds
Total number of unigenes
Mean length of unigenes
Gene annotation and functional classification
For validation and annotation of assembled unigenes, sequence similarity searches were conducted against the NCBI Nr database and the Swiss-Prot protein database using the BLASTx algorithm with an E-value threshold of 10−5. The results indicated that of 59 669 unigenes, 34 749 (58.23%) had a significant similarity to known proteins in the Nr database (Table 2). Owing to the lack of Chinese fir genome and EST information, 41.77% of unigenes could not be matched to known genes. Similarly, up to 35 782 unigenes (59.97% of the total) had no Swiss-Prot annotation (Table 2).
Table 2. Annotation of unigene sequences in Chinese fir (Cunninghamia lanceolata)
Number of annotated unigene sequences
Percentage of annotated unigene sequences
Nr, NCBI nonredundant protein database; Swiss-Prot, Swiss-Prot protein database; KEGG, Kyoto Encyclopedia of Genes and Genomes Pathway; GO, gene ontology; COG, clusters of orthologous groups.
To further evaluate the completeness of our transcriptome library and the effectiveness of our annotation process, we searched the annotated sequences for genes involved in COG classifications. Of 34 749 Nr hits, 18 713 sequences had a COG classification. Among the 25 COG categories, the cluster for ‘general function prediction only’ (3138, 16.77%) represented the largest group, followed by ‘transcription’ (1487, 7.95%) and ‘replication, recombination, and repair’ (1432, 7.92%). The following categories, ‘nuclear structure’ (2, 0.01%), ‘extracellular structures’ (4, 0.02%), and ‘cell motility’ (79, 0.42%), represented the smallest groups (Fig. 1).
Gene ontology (GO) assignments were also used to classify the functions of the predicted Chinese fir genes. Based on sequence homology, 12 365 sequences can be categorized into 44 functional groups (Fig. S3). In each of the three main GO classifications (biological process, cellular component, and molecular function), the ‘metabolic process’, ‘cell part’, and ‘catalytic activity’ terms, respectively, were dominant. We also noticed a high percentage of genes from the ‘cellular process’, ‘organelle’, and ‘binding’ categories, but few from ‘biological adhesion’, ‘extracellular region part’, and ‘electron carrier activity’ (Fig. S3). The GO analysis indicated that the identified genes are associated with various biological processes. Five thousand five hundred and eighty sequences were annotated as belonging to the ‘metabolic process’ category, which may allow for the identification of novel genes involved in the secondary metabolism pathways in wood formation.
The KEGG Pathway database records the networks of molecular interactions in the cells, and variants of them specific to particular organisms. Based on a comparison against the KEGG database using BLASTx with an E-value cutoff of < 10−5, of the 59 669 unigenes, 16 517 (27.68%) had significant matches in the database and were assigned to 119 KEGG pathways (Table 2). The most represented pathways were ‘metabolism pathways’ (4213 members), ‘biosynthesis of secondary metabolites’ (2484 members), ‘plant–pathogen interaction’ (1283 members), ‘spliceosome’ (893 members), and ‘phenylpropanoid biosynthesis’ (663 members). These annotations provide a valuable resource for investigating the processes, functions, and pathways involved in wood formation.
Cambial activity of Chinese fir at three developmental stages
The cambium shows seasonal variation in cell division, termed the dormant, reactivating, and active seasons. We initially performed anatomical observations of the cambial zone to identify the timing of the key events of the activity–dormancy cycle, induction, and the cessation of cambial cell division. The vascular cambium is defined as the actively dividing layer of cells that lies between secondary xylem and secondary phloem cells. As shown in Fig. 2, it is obvious that dormant cambium cells were arranged very compactly and were easily distinguished from terminal latewood tracheids and differentiated secondary phloem cells in the samples collected on 25 December 2010 (Fig. 2a,b). In samples collected on 20 March 2011, cambial reactivation was evident, with thin-walled cells displaying some cell expansion but remaining at two to three cells (Fig. 2c,d), and on 1 June 2011, a dramatic increase in the width of the cambial zone reflected a fully activated cambium (Fig. 2e,f). Based on the anatomical observation, we used cryosectioning to isolate cambial meristem cells, which allowed a higher cellular resolution for defining transcriptional and development profiles, and overcame the limitations arising from the mixed samples in many previous investigations. Importantly, use of this technique provides new information about the genetic regulation involved in wood formation and facilitates the dynamic analysis by separating samples.
A snapshot of Chinese fir mRNA profiling in dormant, reactivating, and active vascular cambium
To investigate the annotated transcriptome assembly served as a reference for RNA-Seq profiling of stage-specific expression, we conducted a small RNA-Seq experiment using tangential cryosections of dormant, reactivating, and actively growing Chinese fir vascular cambium and mapped the resulting reads to our reference transcriptome. Using Illumina high-throughput sequencing, 11.6–12.1 million clean reads were generated in the analyzed samples and a mean of 5.8 million unique reads in each cambium sample were mapped uniquely to the transcriptome of Chinese fir, corresponding to 49.3% of the high-quality reads (Table S3).
Filtered with an FDR ≤ 0.001 and |log2(ratio)| ≥ 2, the expression of 4415 DEGs was found to be significantly changed between the dormant and active cambium libraries. The 883 DEGs in both the reactivating and active cambium libraries showed quantitative differences. In addition, we compared the dormant and reactivating cambium libraries, and 4018 variant genes were found, of which 1548 were up-regulated and 2470 were down-regulated (Fig. 3). Larger numbers of genes were up- or down-regulated at the active and dormant stages than at the reactivating stage, indicating that the transcript abundance changed dramatically at these key switches among developmental stages of vascular cambium.
At the active and dormant stages, genes whose transcript abundance exhibited highly dynamic changes (|log2(ratio)| ≥ 4, Fig. 4) included genes encoding transcription regulators (auxin-induced protein, PIN-like auxin efflux carrier, thaumatin-like protein, putative nodulin like-protein, receptor-like protein kinase, somatic embryogenesis receptor kinase), transcription factors (R2R3-MYB transcription factor, class III homeodomain leucine zipper protein, transcription factor AP2-EREBP, zinc finger protein; |log2(ratio)| ≥ 4; Fig. 4), cell cycle-associated proteins (mitotic cyclin A1-like protein, putative A-like cyclin, minichromosome maintenance protein, plant mitotic spindle assembly checkpoint protein mad2), cell wall-associated proteins (beta-glucosidase, endo-beta-1,3-glucanase, celullose synthase, xyloglucan endotransglucosylase, pectin methylesterase), and stress-associated proteins (late embryogenesis abundant protein, disease resistance associated protein, peroxidase, low temperature induced-like protein; Fig. 4, Table S4). Therefore, the change in expression patterns of distinct transcripts suggests the requirement of different development events from dormant to active growth. For example, 12 preferentially expressed transcription factor-related transcripts accumulated to a higher level in the active phase than in the reactivating and dormant phases, which indicates that the active phase may need more transcription factors than the dormant and reactivating phases (Fig. 4, Table S4). In addition, seven genes that are involved in cell cycle and expansion and 15 genes that are involved in cell wall remodeling showed the highest accumulation in the reactivating and active stages (Fig. 4, Table S4), as expected for actively dividing cells. Interestingly, > 85% of the DEGs were classified as either ‘no hits’ or ‘proteins of unknown function’ (Table S5), suggesting that these genes are only present in trees and might play important roles in cambium development and wood formation.
Verification of the gene expression profiles by qRT-PCR
To further verify the expression profiles of genes in our Illumina sequencing analyses, we have selected 12 DEGs by qRT-PCR using the same samples originally used for RNA-Seq, including genes encoding class III homeodomain leucine zipper protein, zinc finger protein, R2R3-MYB transcription factor, ARF-L1 protein, PIN-like auxin efflux carrier, auxin-induced protein 5NG4, putative beta-1,3-glucanase, endo-beta-1,4-glucanase, putative A-like cyclin, xyloglucan endotransglycosylase hydrolase, pectate lyase and histone H4, respectively. These genes were selected for their key roles in regulating cambial activity, cell division, and cell expansion. The results presented in Fig. 5 showed that the expression levels of 10 genes were higher in active and reactivating cambium than in dormant cambium, including genes encoding class III homeodomain leucine zipper protein, ARF-L1 protein, PIN-like auxin efflux carrier, auxin-induced protein 5NG4, putative beta-1,3-glucanase, endo-beta-1,4-glucanase, putative A-like cyclin, xyloglucan endotransglycosylase hydrolase, pectate lyase and histone H4, whereas two genes encoding zinc finger protein and R2R3-MYB transcription factor were more highly expressed in active cambium than in reactivating and dormant cambium. These results indicated that there was a close correlation between the expression changes (fold difference) measured by RNA-Seq and those by qRT-PCR (Fig. 5), further indicating the reliability of our sequencing data as well as confirming the differences in regulation of cambial activity, cell division, and cell expansion during vascular cambium development.
Illumina paired-end sequencing, assembly, and functional annotation
The transcriptome is the complete set and quantity of transcripts in a cell at a specific developmental stage or under a physiological condition. Therefore, transcriptome analysis is of importance in interpreting the functional elements of the genome and elucidating the molecular constituents of cells and tissues. In the present study, we sampled the pooled transcriptomes of roots, stems, leaves, and seeds of Chinese fir using Illumina paired-end sequencing technology to generate a large-scale EST database. Approximately 2.5 Gb of data was generated and assembled into 59 669 unigenes. This large number of reads with paired-end information produced much longer unigenes (mean, 497 bp) than those in previous studies (Novaes et al., 2008; Meyer et al., 2009; Wang et al., 2010b). This increased coverage depth of transcriptome facilitated de novo assembly, enhanced sequencing accuracy, and avoided possible contamination.
Of the Chinese fir unigenes, 58.23% (34 749 of 59 669) had homologs in the Nr databases, whereas only 38.50, 16.20, and 46.21% unigenes had homologs in the Nr database in Epimedium sagittatum (Zeng et al., 2010), whitefly (Wang et al., 2010b), and sweet potato (Wang et al., 2010c), respectively. The higher percentage of hits found in our paired-end sequencing was largely the result of the greater number of long sequences in the unigene database of Chinese fir, in agreement with the earlier findings in Taxus (Hao et al., 2011) and Siraitia grosvenorii (Tang et al., 2011). Importantly, we can assign a number of these unigenes to a wide range of GO categories and COG classifications (Figs 1, S3), indicating that a wide diversity of transcripts involved in wood formation are represented in the sequence data of this species, reflecting the complexity of differential developmental stages in woody plants. Furthermore, most representative unigenes were annotated to specific pathways, such as metabolic pathways, biosynthesis of secondary metabolites, plant–pathogen interactions, spliceosome, and phenylpropanoid biosynthesis using the KEGG database (Table 2), leading us to conclude that most of the genes we identified are involved in cambial differentiation and wood formation.
Global changes of gene expression during the transition from active growth to dormancy
The vascular cambium forms a continuous cylinder of meristematic cells in the stem, producing secondary phloem on the outside and secondary xylem or wood on the inside. Druart et al. (2007) demonstrated that extensive changes in the transcriptome took place in the cambial zone of the model tree aspen during the course of their activity–dormancy cycle. By utilizing microarray technology, Larisch et al. (2012) uncovered significant differences between gene profiles in cambial and ray cells from wood samples of poplar (Populus × canescens) during early spring (reactivation) and active growth. Other studies have observed similar large shifts in gene expression by utilizing microarray technology on large-scale transcript profiling to examine gene expression changes in cambial tissues of forest trees (Schrader et al., 2004; Holliday et al., 2008; Ko et al., 2011; Leonardo et al., 2012), but little attention has been paid to gene expression of cambial cells from the woody plant Chinese fir during dormancy, reactivation and active growth. In the present study, using RNA-Seq to examine global gene expression profiles over a time course of cambium development in Chinese fir, we found that the expression of 4415 DEGs was up- or down-regulated (|log2(ratio)| ≥ 2) between the dormant and active cambium libraries, indicating considerable changes of gene expression in vascular cambium during the transition from active growth to dormancy. Although the genes identified here showed similar functions to those reported previously (Hertzberg et al., 2001; Zhang et al., 2011), it is of interest to note that > 5% of the genes have not been reported in previous studies. Since we analyzed the transcriptional profiles following the dynamic process of cambium development from active growth to dormancy, we obtained more genes with dynamic changes in transcript abundance or those that are only strongly transcribed during the transitions.
Genes involved in cell division and cell expansion during vascular cambium development
Cell division is one of the key processes taking place in the cambial zone and the majority of cell cycle genes were up-regulated in Chinese fir vascular cambium during the active stage (Fig. 4), as expected for actively dividing cells. These genes include cyclin A1 (Unigene57815 and Unigene56058) and cyclin (Unigene4080), which are similar to several genes involved in cell cycle control, like cyclins A and B on maximum gene expression in poplar cambium zone using microarray approach by Hertzberg et al. (2001). The qRT-PCR analysis of one cell cycle gene, cyclins A, indicated that the abundance of this mRNA was higher in active and reactivating cambium than in dormant cambium (Fig. 5). The high abundance of cyclin homolog transcripts in active and reactivating cambium also reflected a positive correlation between cambium cell division and key cell cycle gene expression. In contrast to the active stage, the decline in the transcript abundance of the core cell cycle genes in vascular cambium during the dormant stage correlates well with the cessation of cambial cell division (Li et al., 2009). As many of those genes involved in cell division are down-regulated in the transition from the active to the dormant stage (Brown et al., 2005; Druart et al., 2007; Leonardo et al., 2012), we can conclude that the key cell cycle protein transcripts expressed preferentially at the active stage may be essential for cambial cell division, as expected changes during the growth cessation and reactivation cycles.
The expression of histone H4, as a marker for cell division, is associated with DNA replication (Hertzberg et al., 2001). In our experiment, the expression of the histone H4 gene was maximal during the active and reactivating stages and declined at the dormant stage, in support of the conclusions drawn from the cell cycle genes that cell proliferation is at its highest in active cambium. After cell division, cells in the cambial zone expand in the axial and radial directions. Cell wall expansion plays a crucial role in shaping the morphology of plants. During cell extension, modifications in the structure and composition of the cross-linked pectin xyloglucans occur. Xyloglucan endotransglycosylases (XETs) are responsible for cell wall remodeling during primary cell wall biosynthesis by cutting and rejoining the xyloglucan chains (Wang et al., 2007). Previous studies have shown that several genes encoding xyloglucan endotransglycosylase (XTH), pectin methylesterase and expansin are involved in the cell wall remodeling and expansion processes in coniferous species (Wang et al., 2007; Maurice et al., 2011). In this study, we found that the mRNA levels of several genes were more than fourfold higher in reactivating and active vascular cambium than in dormant vascular cambium, including those encoding XTH (Unigene28011 and Unigene47961), pectin methylesterase (Unigene38919, Unigene35308 and Unigene227), pectin esterase (Unigene34411, Unigene57279, Unigene52430), pectate lyase (Unigene42642 and Unigene52045) and expansin (Unigene7691, Unigene57087 and Unigene25630), suggesting that the expression of these genes is responsible for the cell wall loosening and cell elongation during the process. Furthermore, three putative genes encoding beta-1,3-glucanase and seven putative genes encoding endo-1,4-beta-glucanase, which are involved in cell wall loosening and elongation, displayed higher expression levels (more than fourfold differential expression) in active cambium than in dormant cambium. These results suggest that these highly differentially expressed genes are involved in a broad range of physiological functions, especially in cell division and cell expansion during vascular cambium development.
Genes involved in the auxin response pathway during vascular cambium development
Auxin plays a pivotal role in various aspects of plant growth and developmental processes, such as cambial cell division, cell wall loosening and cell elongation, vascular tissue differentiation, and secondary xylem development in trees, through regulation of the expression of early auxin response factors (ARFs; Schrader et al., 2004; Nilsson et al., 2008; Hayashi, 2012). It was demonstrated that auxin can be transported between cells in a directional manner, which is mediated by several families of auxin transporters. AUX1/LAX symporters function as auxin influx carriers, whereas the PIN-formed protein (PIN) family of auxin efflux facilitators export auxin from the cell (Hayashi, 2012). By comparing the changes in the components of the auxin signaling and transport machinery in dormant, reactivating, and active cambia, we found that genes encoding the PIN-like auxin efflux carrier (Unigene48697 and Unigene35220) and auxin-induced protein (Unigene54524, Unigene45986 and Unigene11714) were highly expressed (more than fourfold differential expression) in vascular cambium at the reactivating and active stages, when more cambial cells are actively expanding and elongating. These results suggest that the expression changes of auxin transport genes may contribute to the setting up of auxin concentration gradients once again during vascular cambium development and wood formation.
Alongside polar auxin transport, auxin signaling also plays an essential role in several developmental processes, from root and shoot development to flower and fruit development in plants. It was found that one IAA-induced transcription factor is up-regulated during poplar secondary vascular tissue regeneration after bark girdling (Zhang et al., 2011). Wang et al. (2007) reported several members of the auxin signaling pathway in Chinese fir. In the present investigation, we identified homologs of genes involved in auxin response, such as ARF in active cambium, using RNA-Seq and qRT-PCR, suggesting that the ARF genes are involved in the regulation of cambium formation and wood formation. Further studies on the roles of additional auxin polar transport carriers and auxin response factors will help us to understand the roles of auxin in cambial cell development and wood formation.
Genes involved in transcriptional regulation during vascular cambium development
Transcription factors that are expressed predominantly during vascular development and secondary growth are of considerable interest because of the economic importance of wood and wood fibers. Ingraham et al. (1988) found that the homeodomain-containing superfamily of transcription factors participated in a wide variety of plant developmental processes. Increasing evidence indicates that the leucine zipper-associated HD-Zip, knotted-related KNOX, and zinc finger-associated ZF-HD homeodomain-containing transcription factors are associated with processes related to meristem functions in both shoot and root apical meristems, polarity of lateral organs, and development of primary vascular tissues in several species (Federico et al., 2007; Ilegems et al., 2010). Previous studies indicated that the poplar homolog to Arabidopsis, ATHB-8, was highly expressed in the cambial meristem (Hertzberg et al., 2001). ATHB-8 is a member of the HD-Zip III class of transcription factors that is expressed in provascular cells of Arabidopsis, where it has been proposed to regulate vascular development (Kim et al., 2008). Our present results, which were obtained from the RNA-Seq and qRT-PCR, showed that these genes encoding class III homeodomain-leucine zipper protein and zinc finger transcription factor were specifically expressed in active cambium from Chinese fir, suggesting that these genes are required to balance cell proliferation and the maintenance and specification of undifferentiated cells during active growth, but are not necessary for dormancy.
Although most attention has been focused on the homeodomain-containing transcription factors, transcription factors belonging to other families are involved in regulating cambium development and wood formation. For instance, it has been shown that some members of the R2R3-MYB family are essential for controlling lignin deposition and secondary wall formation in tree species including pine and spruce by interacting with other R2R3-MYB genes, activated by NAC (NAM/ATAF/CUC) transcription factor master switches and binding to AC (ACCTACC) elements (Bedon et al., 2007; Zhong et al., 2008; Zhong & Ye, 2009). The genes MYB1 and MYB8 from Pinus taeda are potentially important players in conifers in the regulation of secondary cell wall biosynthesis, including lignin deposition (Bomal et al., 2008). Moreover, it has recently been shown that transcriptional regulators, including genes encoding zinc finger and AP2-EREBP transcription factors, were more highly expressed during cell wall thickening in cotton fiber development than during the earlier elongation stage (Al-Ghazi et al., 2009). A regulatory role for AP2-EREBPs in poplar secondary cell wall metabolism has also been suggested (van Raemdonck et al., 2005). Our RNA-Seq and qRT-PCR data indicated that these genes encoding R2R3-MYB and AP2-EREBP transcription factors were up-regulated in active cambium from Chinese fir. Given that these transcription factors may regulate, directly or indirectly, secondary wall biosynthesis and vascular development, characterization of the DEGs encoding transcription factors might shed light on the regulation of wood formation in conifer.
Vascular cambium is the lateral meristem producing xylem cells inwards and phloem cells outwards in plant stems. The genome-wide transcriptome and RNA-Seq analysis presented in this study has expanded our knowledge of this process by identifying dramatically expressed genes involved in crucial biological processes such as cell division, cell expansion, secondary cell wall biosynthesis and hormone pathways. The approach of combining tangential cryosectioning and transcriptome analysis is a valuable tool for the investigation of tissue-specific expression in Chinese fir. Importantly, the high-resolution expression patterns presented here highlight our understanding of the molecular mechanisms involved in vascular cambium development and wood formation in the conifers.
This work is supported by the Major Science Foundation of Ministry of Education of China (no. 313008), the National Basic Research Program of China (973 Program 2009CB118500), the Ministry of Agriculture of China (2011ZX08002-003 and 2009ZX08009-095B) and the China Postdoctoral Science Foundation (2012M520457). The authors declare that there is no conflict of interest.