Comparative transcriptomic analyses of normal and malformed flowers in sugar apple (Annona squamosa L.) to identify the differential expressed genes between normal and malformed flowers
- 1.6k Downloads
Sugar apple (Annona squamosa L.), a popular fruit with high medicinal and nutritional properties, is widely cultivated in tropical South Asia and America. The malformed flower is a major cause for a reduction in production of sugar apple. However, little information is available on the differences between normal and malformed flowers of sugar apple.
To gain a comprehensive perspective on the differences between normal and malformed flowers of sugar apple, cDNA libraries from normal and malformation flowers were prepared independently for Illumina sequencing. The data generated a total of 70,189,896 reads that were integrated and assembled into 55,097 unigenes with a mean length of 783 bp. A large number of differentially expressed genes (DEGs) were identified. Among these DEGs, 701 flower development-associated transcript factor encoding genes were included. Furthermore, a large number of flowering- and hormone-related DEGs were also identified, and most of these genes were down-regulated expressed in the malformation flowers. The expression levels of 15 selected genes were validated using quantitative-PCR. The contents of several endogenous hormones were measured. The malformed flowers displayed lower endogenous hormone levels compared to the normal flowers.
The expression data as well as hormone levels in our study will serve as a comprehensive resource for investigating the regulation mechanism involved in floral organ development in sugar apple.
KeywordsA. squamosa Normal flower Malformed flower Transcriptome Digital gene expression Phytohormone
Clusters of Orthologous Groups
differentially expressed genes
digital gene expression
Kyoto Encyclopedia of Genes and Genomes
quantitative real-time PCR
reads per kb per million reads
Sugar apple (Annona squamosa L.), a member of the Annonaceae family, is an important medicinal plant that is widely distributed in tropical South Asia and America . In addition to its pharmaceutical value, sugar apple is used in juices, jellies and compotes . Due to its medicinal and nutritional properties, sugar apple is a popular tropical fruit in China . The flowering of sugar apple is a complex process that is regulated by various environmental factors, such as improper shoot state, low temperature or short day conditions . Malformed flowers are the major reason for the low percentage of fertile fruit in cultivated sugar apple. However, the available information on the floral organ development-related genes in the genus Annona is very scarce.
Flowering is a vital developmental process in the life cycles of plants . The coordinated transition from vegetative growth to reproductive development is required for successful reproduction in higher plants . In the model plant Arabidopsis thaliana, the specification of floral organ identities is regulated by a large number of floral genes, most of which encode transcription factors. Based on the phenotypes of Arabidopsis mutants, a classical ABC model was developed . According to the model, most of the floral organ identity-related genes can be grouped into three gene classes: A, B and C. In detail, sepals are identified by Class A genes; petals are identified by Class A together with Class B genes; stamens are identified by Class B genes together with Class C genes; and carpels are identified by Class C genes . Recently, the ABC model was extended and modified by genes (Class E) that are essential for the specification of all types of floral organs . The functions of homeotic genes are highly conserved across different species, allowing us to analyze and identify potential genes that are related to flower development in sugar apple.
Various environmental and endogenous signals are associated with an array of biochemical and cellular processes during the formation of floral organs . Among the signals, phytohormones are endogenously occurring compounds involved in the floral transition, flowering and floral organ development . Several classical phytohormones, such as gibberellins (GAs), auxins, cytokinins (CKs), ethylene, abscisic acid (ABA), jasmonates (JAs), salicylic acid (SA) and brassinosteroids (BRs), have been implicated in the flowering-time pathway . GAs are a class of phytohormones that function not only to induce the transition to flowering, but also to control flowering-time . The role of GAs in the floral transition reflects their recognized function as integrators through DELLA-mediated pathways . GAs promote flowering by triggering the expression of a series of floral integrator encoding genes, including SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1, LEAFY and FLOWERING LOCUS T, (in the inflorescence and floral meristems [14, 15]. Auxins, another classical hormone with a floral-inductive signalling role, were reported as regulators of embryonic and postembryonic development. There is a close correlation between the endogenous auxin content and the flowering time . Several proteins, such as TRANSLOCATED PROMOTER REGION, SUPPRESSOR OF AUXIN RESISTANCE3 and HASTY, are involved in both auxin responses and flowering-time control . CKs, together with auxins, regulate both the division cycle and meristem homeostasis, and may promote the floral transition . In Arabidopsis, applications of exogenous CKs accelerate flowering, though the molecular mechanism underlying this is largely unknown .
Thus, there is limited data on the molecular basis of floral initiation and differentiation in the genus Annona. Without reference genomes, de novo sequencing is an effective approach for the identification of candidate genes in most trees, including melon, litchi and Chinese cherry [20, 21, 22]. In our study, two independent cDNA libraries, one of a normal flower and one of a malformed sugar apple flower, were constructed for Illumina RNA sequencing (RNA-seq). The annotation of the transcriptome sequences allowed for the identification and analysis of potential genes related to floral organ development in sugar apple.
Plant material and sampling
In our study, 10-year-old adult trees of A. squamosa cv. ‘Bendi’ were cultivated in a 4 × 4 m arrangement with drip irrigation and fertilizer applications as required. The location of the trees was the Ling Nan Normal University field experimental station in Zhanjiang City (Guangdong Province, China). The flowers before open were collected from a normal flower and a malformed flower. All of the flower samples were frozen immediately in liquid nitrogen, and stored at −80 °C for further studies.
cDNA library preparation and Illumina sequencing for transcriptome analysis
Total RNA was extracted using a TRIzol Kit according to the manufacturer’s protocol (Promega, Beijing, China). Residual DNA contamination was removed by RNase-free DNase I (TaKaRa, Dalian, China). RNA quality was verified by RNase-free agarose gel electrophoresis, and the total RNA concentration was measured using a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) at 260 nm and 280 nm. RNA samples with 260 nm/280 nm ratios between 1.8 and 2.0 were used for subsequent analyses. Each cDNA library was constructed by mixing three independent replicate samples. The library of the normal flower was named ‘NF’, while the library of the malformed flower was named ‘MF’. The two libraries were used for the comparative analysis of transcriptome sequencing using the Illumina HiSeq™ 2500 platform by Gene Denovo Co. (Gene Denovo, Guangzhou, China). Raw reads were generated in a paired-end format, and the NF and MF transcriptome data sets were deposited in the GenBank Short Read Archive under accession number SRA508784.
De novo assembly and functional annotation of Illumina sequencing
Low-quality reads (with more than 5% unknown bases) and adaptor sequences were filtered and removed, and then the clean reads were assembled using Trinity software to generate unique consensus contigs . All of the contigs were calculated using a sequence clustering software, and the longest sequences were defined as unigenes. The assembled sugar apple unigenes were aligned to several protein databases, such as NCBI Nr protein (http://www.ncbi.nlm.nih.gov), Swiss-Prot protein (http://www.expasy.ch/sprot), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway (http://www.genome.jp/kegg) and Clusters of Orthologous Groups (COG) (http://www.ncbi.nlm.nih.gov/COG), using the BLASTX algorithm with an E-value <0.00001. For unigene annotation, the Blast2GO program was used to produce the gene ontology (GO) annotation of unigenes, and a BLAST algorithm-based search against the KEGG database (http://www.genome.jp/kegg/) was used to analyze protein products of metabolic processes and related gene functions in cellular processes .
Analysis and mapping of digital gene expression (DGE) tags
To map DGE tags, raw data were filtered to remove the low quality tags (sequences with a number of unknown ‘N’s), empty tags (sequences only containing the adaptors), and tags with only one copy number (sequences from sequencing errors). For annotation, cleaned tags containing CATG and the 21-bp tag sequence were mapped to our transcriptome reference database. Initially, the tags that mapped to multiple genes were filtered out. Then, the remaining tags were treated as unambiguous tags for the gene expression analysis. The number of unambiguous tags of each gene was calculated and then normalized to the number of transcripts per million clean tags. The differentially expressed tags between NF and MF were used for further analysis and mapping.
Identification of differentially expressed genes (DEGs)
Reads assoicated with each unigene were mapped to the transcriptome using the alignment software Bowtie 0.12.8 (https://sourceforge.net/projects/bowtie-bio/files/bowtie/0.12.8/). To calculate unigene expression, the number of mapped reads for each unigene was counted and normalized into a reads per kb per million reads (RPKM) value. To determine significant differences, a false discovery rate < 0.001 and an absolute value of log2 ratio > 1 were set as thresholds. DEGs between the two flower samples were calculated using the edgeR package . GO and KEGG enrichment analyses of the DEGs between two flower samples were performed according to a previous reported method .
Quantitative real-time PCR (qRT-PCR) validation
Total RNA was extracted from the same sample that were used for sequencing. In total, 3 μg RNA of each sample was used for reverse transcription. The cDNA was synthesized using ReverAid First Strand cDNA Synthesis Kit (Thermo Scientific. Shanghai, China). QRT-PCR was performed using the SYBR Premix Ex Taq Kit (TaKaRa, Dalian, China) and an ABI PRISM 7700 DNA Sequence Detection System (Applied Biosystems, Shanghai, China). Three independent cDNA samples from “NF” and “MF” were used for qRT-PCR validation. The primer sequences were designed using Primer Premier 5 software (Premier Biosoft International, Palo Alto, CA, USA). The sugar apple Actin gene was used as an internal standard to calculate relative fold-differences based on comparative cycle threshold (2−ΔΔCt ) values. Then, ddH2O was used as no-template control. The qRT-PCR procedure was as follows: 1 μL of a 1/10 dilution of cDNA in H2O was added to 5 μL of 2× SYBR® Green buffer, with 0.1 μM of each primer and H2O to a final volume of 10 μL. The reactions were run as follows: 50 °C for 2 min and 95 °C for 10 min, followed by 40 cycles of 95 °C for 30 s, 56 °C for 30 s and 72 °C for 30 s in 96-well optical reaction plates . Each real-time PCR reaction was performed three times.
Measurements of various hormones
For the exogenous hormone contents analysis, independent samples from normal and malformed sugar apple flowers were harvested, immediately frozen in liquid nitrogen and stored at −80 °C for further extractions. Exogenous indole-3-acetic acid (IAA) contents were determined using a FOCUS GC-DSQII (Thermo Fisher Scientific Inc., Austin, TX, USA) as described previously . Endogenous GA was measured using the nano-LC-ESI-Q-TOF-MS analysis method described previously . In additionally, ABA and zeatin ribosides (ZRs) were detected using a UFLC-MS/MS system described by Kasote et al. .
Significant differences between different samples were calculated using a one-way analysis of variance with a Tukey’s test (at a significance level of α = 0.01) in Excel software. All of the expression analyses were performed for three biological replicates. All reported values represent the averages of three replicates, and data are expressed as the mean plus or minus the standard deviation (mean ± SD).
Sequencing, de novo assembly and sequence annotation
In our study, a large number of raw reads, including 98.15% of clean reads, 0.88% of only adaptors and 0.97% of N-containing reads, were obtained (Fig. 1b). A total of 70,189,896 reads were integrated and assembled into 55,097 unigenes with a mean length of 783 bp. The detailed information of the obtained reads is listed in Additional file 1. The size distributions of clean reads and unigenes in sugar apple are shown in Fig. 1c and d. For clean reads, 3% of the reads were >3000 bp in length and a majority of the reads (58%) were <500 bp in length. Only 1% of the unigenes were >10,000 bp in length, and the majority of unigenes were between 11 bp and 50 bp.
For the COG classification, 13,548 unigenes were grouped into 25 functional classifications. The largest term, “General function prediction only”, contained 2415 unigenes. “Replication, recombination and repair”, “Cell wall/membrane/envelope”, “Posttranslational modification” and “Signal transduction mechanisms” shared a large percentage of unigenes. In addition, only nine unigenes were classified into “Cell motility”, and two unigenes were classified into “Nuclear structure”. Furthermore, two flower development-related COG terms, “signaling transduction mechanisms” and “Cell cycle control, cell division”, were analyzed in detail (Fig. 2c).
GO and KEGG classifications of unigenes
In sugar apple, most of the unigenes could be assigned to 46 functional terms that belonged to three GO categories, biological process, cellular component and molecular function. For biological process, “metabolic processes” (6128 unigenes) and “cellular processes” (5721 unigenes) were dominant terms; for cellular component, the dominant terms were “cell” (6377 unigenes), “cell part” (6227 unigenes) and “organelle” (6128 unigenes); and for molecular function, a large percentage of unigenes were related to “metabolic process” (6707 unigenes), “catalytic activity” (5942 unigenes) and “binding” (5168 unigenes) (Additional file 2).
Furthermore, all of the unigenes mapped to canonical pathways in the KEGG database. Our data showed that 7191 unigenes from sugar apple were assigned to 124 KEGG pathways (Additional file 3). The most represented KEGG pathway was “metabolic pathways” with 1940 unigenes (ko01100). Additionally, 959 unigenes were associated with the KEGG pathway “biosynthesis of secondary metabolites”, 335 unigenes were mapped to “ribosome”, and 219 unigenes were assigned to “protein processing in endoplasmic reticulum”.
Screening and classification of DEGs between normal and malformed flowers
Identification of the floral organ development-associated TF-encoding genes and MADS-box genes
In total, 701 TF genes were identified as DEGs (Additional file 4). The top five largest differential expressed TF families were: ERF (53 members), bHLH (42 members), MYB-related (38 members), FAR1 (37 members) and NAC (36 members). Moreover, several hormone-related TFs also were identified as DEGs, including 24GRASs and 9 ARFs. In our study, 21 MADS-box genes were revealed from sugar apple flowers (Additional file 5). Among these MADS-box genes, four genes, such as AGL9, SEP1, AP1-like, were identified in MIKC family and 17 genes, such as AP3-like, AGL63 and PISTILLATA-like.
Identification of key flowering- and flower development-related DEGs
A large number of flowering- and flower development-related genes play important roles in the entire flower developmental cycle . In total, 18 DEGs in sugar apple showed homology to known flowering- and flower development-related genes in the NCBI and UniProt databases (Additional file 6). For example, two flowering promoting factor genes (Unigene0041725 and Unigene0054870) showed higher expression levels in the malformed flowers than in the normal flowers. Additionally, four early flowering protein genes (Unigene0023978, Unigene0023976, Unigene0033968 and Unigene0008711), one flowering locus T gene (Unigene0000043), one flowering time control protein FY-like gene (Unigene0025538), three flowering time control protein FPA-like genes (Unigene0005475, Unigene0019814 and Unigene0005474), and one flowering time control protein FCA-like gene (Unigene0024514), were also identified as DEGs.
Based on their GO terms (GO:0009908), 41 DEGs were identified as flower development-related genes (Additional file 7). Interestingly, two genes, a probable histone H2A variant 3-like gene (Unigene0017118) and ODORANT1-like gene (Unigene0047135), were largely induced (> 5 fold) in the malformed flowers. Most of other identified flower development-related DEGs were reduced in the malformed flowers.
Identification of hormone-related DEGs
Based on the RPKM values, the expression levels of the hormone-related unigenes were analyzed and shown in Fig. 5b. The expression levels of most hormone-related unigenes in the normal flowers were greater than those in the malformed flowers.
Validation of the expression of several key hormone-related genes
Endogenous hormone measurements
Sugar apple is a popular fruit tree with a high commercial value in subtropical and tropical areas . Without a public genome database, detailed information on the genes involved in their growth and development is unavailable for sugar apple studies. A DEG analysis is a good tool for studying the temporal regulation of gene expression . Our previous work constructed independent cDNA libraries of four different flower stages in A. squamosa for Illumina RNA-seq. A large number of flower developmental stage-related genes have been identified . In the present study, transcriptome data from normal and malformed flowers at the same developmental stage (the mature flowers with partially opened petals) were used to uncover the differences between normal and malformed flowers and provide a more adequate resource to study sugar apple.
In model plants, various TF families, such as MYB, MADS-box, NAC, ARF and bHLH, have been reported to be involved in floral development [34, 35, 36, 37, 38]. Therefore, the differential expression levels of TF genes between normal and malformed flowers were analyzed in our study. Our previous study identified a large number of flower development-related TF families, including bHLH, NAC, B3, MYB-related and bZIP. MYB factors promote petal and stamen growth in Arabidopsis. For example, MYB21 fed back negatively on the expression of JA biosynthesis-related genes to control flower development by decreasing the JA level . Our previous study identified a large MYB-related TFs, suggesting an important role for MYB factors in the floral organ development of sugar apple . In the present study, the MYB family contained the largest number of differentially expressed TFs (59 members). Additionally, bHLH is another large TF family that regulates various flower developmentally related processes . In sugar apple, the bHLH family had the second largest number of differentially expressed TFs (35 members) between normal and malformed flowers. Our previous study showed that bHLH is the largest TF family involved in flower development of sugar apple, indicating a involvement of MYB and bHLH families in the differences between normal and malformed flowers.
The Arabidopsis TFs ARF6 and ARF8 are expressed in multiple flower tissues, such as sepals, petals and stamen filaments . Moreover, these two genes function in different organs to promote the transition from closed buds to mature fertile flowers . ARF1 and ARF2 regulate senescence and floral organ abscission in Arabidopsis . In sugar apple, nine ARFs were identified as differentially expressed TFs between normal and malformed flowers, indicating that ARF-mediated auxin signaling participated in flower development.
It has been reported that morphogenesis of floral organs was under control of several MADS-box genes. Loss-of-function of any essential MADS-box genes may result in homeotic conversion of floral organs . For example, four orchid AGL9-like MADS-box genes play roles in floral transition and formation in Arabidopsis . In the present study, four MIKC-type MADS-box genes were identified, including an AGL9 gene, which may involved in floral transition of sugar apple. Besides, SEPALLATA subfamily MADS-box protein was reported to positively control spikelet meristem identify in rice . A SEPALLATA 1-like protein encoding gene also was identified in sugar apple. The data provides a more comprehensive information of the floral organ development process in sugar apple.
Based on the DEGs analysis, a large number of auxin-related genes, including ARFs, Aux/IAAs, and GHs, were identified as DEGs. The phytohormone auxin is involved in regulating many aspects of plant growth and development [47, 48]. A classical role of auxin in the formation of flowers at the periphery of the reproductive shoot apex was revealed by genetic evidence . For example, mutations in PIN1 or PID, auxin efflux carrier-encoding genes, lack flowers . At the early stage, the development of female gametophytes is controlled by the AtPIN1-mediated auxin flux . In our study, the identified auxin influx carrier-encoding genes and most of the identified auxin efflux carrier-encoding genes were down-regulated in the malformed flowers. In Arabidopsis, ARF4 is associated with flower patterning . Another ARF gene, AtARF3, integrates the functions of AGAMOUS and APETALA2 in floral meristem determinacy [53, 54]. Here, the expression of most of the identified ARFs and Aux/IAAs was also reduced in the malformed flowers. Thus, auxin and auxin transport may be required for floral meristem determinacy and flower patterning in sugar apple.
CK plays pivotal roles in many aspects of plant development, including the formation of male and female functions . In Arabidopsis, the floral homeotic gene APETALA1 directly reduces the biosynthesis of endogenous CK by promoting the expression of the CK degradation gene CYTOKININ OXIDASE/DEHYDROGENASE3 and inhibiting the expression of the CK biosynthetic gene LONELY GUY1 . In our study, a number of CK biosynthesis-related genes, including 3 CYPs and 10 CKXs, were identified as DEGs in sugar apple. Decreases in the biosynthesis of CK may cause the malformed flower phenotype.
In additional, GA and ABA play important roles in flower development. For example, effects of exogenous GA3 applications on grape flowers have been analyzed. An RNA-seq transcriptome analysis suggested that the morphology of grape inflorescences may be controlled by the biosynthesis and signaling of GA3 . In Arabidopsis, a GA-deficient mutant, ga1–3, displayed the retarded growth of four whorls in the floral organs, and its flower phenotypes could be rescued by the application of exogenous GA . GA-INSENSITIVE DWARF1 (GID1), a soluble protein with a high similarity to hormone-sensitive lipases, was first cloned in rice as a GA receptor . In our study, two GID-encoding genes were identified as DEGs, suggesting an involvement of GA signaling in sugar apple’s flower development. For GA biosynthesis, the GA2ox family genes were identified in various plant species . Based on our transcriptome data, some GA3OX1 family genes, such as Unigene0016441, were largely reduced in malformed flowers, while several GA2OX2 family genes, such as Unigene0010774 and Unigene0004102, were induced in malformed flowers. This suggested that a diversity of regulatory mechanisms are involved in GA biosynthesis during flower development. A number of transcriptomes revealed that there is a close relationship between ABA signaling and flower development in various plant species, such as Gerbera hybrida . Our data confirmed that ABA-related genes also showed significant expression changes between normal and malformed flowers.
Furthermore, malformed flowers displayed lower endogenous hormone levels compared with the normal flowers. This was in agreement with the decreases in the expression levels of hormone-related genes in malformed flowers. In Arabidopsis, increased auxin levels attribute the initiation of flower formation in the shoot apex . Addition to auxin, cytokinin accumulation also causes alteration of flower development in Arabidopsis . In sugar apple, low auxin and cytokinin levels may affect downstream processes of flower development and lead to malformed flowers. ABA alters the number of carpels in Arabidopsis . Compared with normal flowers, lower ABA level may affect the formation of carpels in malformed flowers. Recent publications have established that GA is involved in stamen development. For example, mutations in GA receptors affect the elongation of stamens . GA concentration may play a role in stamen development in sugar apple.
In our previous work, genes associated with floral transition and flower development were identified in sugar apple . In the present study, most selected DEGs showed significantly expression changes during the flower development process, suggesting a causal relationship to the defects associated with the malformed flowers.
In summary, two independent cDNA libraries from normal and malformed flowers of sugar apple were separately constructed and sequenced. A large number of DEGs were identified between normal and malformed flowers. The expression changes of hormone-related unigenes were analyzed and validated using a qRT-PCR analysis. The expression analysis, together with the hormone contents, provides a comprehensive resource for investigating the differential expressed genes between normal and malformed flowers.
We are grateful to Qian Liu and Yang Liu (from Guangzhou Genedenovo Biotechnology Co., Ltd) for technical support.
This work was supported by the National Natural Science Foundation of China (grant no. 31201586); Science and Technology Program of Guangdong, China (grant no. 2014A020208138 and 2015A020208018); Science and Technology Project of Zhanjiang City, China (grant no.2016A03016); Natural Science Foundation of Lingnan Normal University (grant no. LZL1507). All these funding play roles in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
Availability of data and materials
The datasets supporting the conclusions of this article are included within the article and its additional files.
KL, HL, WL and JZ carried out the molecular studies, participated in the analysis and drafted the manuscript. YC carried out the qRT-PCR analysis. HL performed the statistical analysis. KL and CY conceived of the study, and participated in its design. KL acquired of funding. CS helped to draft the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The adult trees of A. squamosa cv. ‘Bendi’ are widely cultivated in China. This project uses plant materials and does not utilize transgenic technology. It does not require ethical approval.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 3.Liu K, Li H, Yuan C, Huang Y, Chen Y, Liu J. Identification of phenological growth stages of sugar apple (Annona squamosa L.) using the extended BBCH-scale. Sci Hortic. 2015;181:76–80.Google Scholar
- 20.Corbacho J, Romojaro F, Pech JC, Latche A, Gomez-Jimenez MC. Transcriptomic events involved in melon mature-fruit abscission comprise the sequential induction of cell-wall degrading genes coupled to a stimulation of endo and exocytosis. PLoS One. 2013;8(3):e58363.CrossRefPubMedPubMedCentralGoogle Scholar
- 30.Kasote DM, Ghosh R, Chung JY, Kim J, Bae I, Bae H. Multiple reaction monitoring mode based liquid chromatography-mass spectrometry method for simultaneous quantification of brassinolide and other plant hormones involved in abiotic stresses. Int J anal chem. 2016;2016:7214087.CrossRefPubMedPubMedCentralGoogle Scholar
- 33.Hao J, Guo H, Shi X, Wang Y, Wan Q, Song Y et al. Comparative proteomic analyses of two Taxus species (Taxus × media and Taxus mairei ) reveals variations in the metabolisms associated with paclitaxel and other metabolites. Plant and Cell Physiology. 2017. doi: 10.1093/pcp/pcx128.
- 34.Yu C, Guo H, Zhang Y, Song Y, Pi E, Yu C, Zhang L, et al. Identification of potential genes that contributed to the variation in the taxoid contents between two Taxus species (Taxus media and Taxus mairei). Tree Physiol. 2017; doi: 10.1093/treephys/tpx091.
- 35.Ning YQ, Ma ZY, Huang HW, Mo H, Zhao TT, Li L et al. Two novel NAC transcription factors regulate gene expression and flowering time by associating with the histone demethylase JMJ14. Nucleic acids research. 2015;43(3):1469-1484Google Scholar
- 58.Goto N, Pharis RP. Role of gibberellins in the development of floral organs of the gibberellin-deficient mutant, ga1-1, of Arabidopsis thaliana. Can J Bot. 1999;77(7):944–54.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.