Introduction

Alpha-linolenic acid (ALA, n:3/omega-3) and linoleic acid (LA, n:6/omega-6) serve as precursors for the synthesis of long chain fatty acids such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) (Tang et al. 2018). EPA and DHA can not only promote brain development, improve visual acuity, prolong life expectancy, but also help in eradicate heart disease, hypertension, diabetes and cancer (Bai et al. 2016). Since the rediscovery of ancient and sacred oil crop chia, it has become more and more attractive due to its high unsaturated fatty acid content such as ALA and other important nutraceuticals such as proteins, edible fiber, seed coat gel, vitamin E, and flavonoid antioxidants (Sreedhar et al. 2015). Among known crops, Chia as an oil plant contains the highest ALA level of the total fatty acids. It is grown in deserts below an altitude of 4,000 feet in Mexico and Southwest America, and has developed into one of the important staple crops by ancient Astek and Mayas (Ayerza and Coates 2005).

Plant flowering time is mainly regulated by five pathways: photoperiod pathway, vernalization pathway, gibberellin pathway, autonomous pathway and aging pathway. The MADS-box transcription factor SOC1 (AGL20) is an integrator that integrates flowering signals from multiple flowering regulatory pathways to promote flowering (Richter et al. 2013). Functional loss in Arabidopsis thaliana SOC1 mutants or transgenic silencing of SOC1 led to late flowering. When A. thaliana SOC1-like genes were transformed into Citrus reticulate, flowering time was significantly accelerated (Tan and Swain 2007). Transforming petunia SOC1-like/FBP21 into tobacco resulted in early flowering, without influence on flower number and seed quality (Ma et al. 2011). Surprisingly, overexpression of Gossypium hirsutum GhSOC1-like that is orthologous to AtAGL42/71/72 in Gerbera jamesonii did not advance the flowering time, but interfered with floral organ development (Ruokolainen et al. 2011). Rice FDRMADS8, which is an orthologue of AtAGL14, has certain influence on floral development (Jia et al. 2000) and AtAGL14 also promotes root development in Arabidopsis (Garay-Arroyo et al. 2013). AtAGL19 promoted early flowering in hda9 mutants, and was inhibited in HDA9 or in LDs (Kim et al. 2013). The photoperiod pathway enhances SOC1 and promotes flowering by promoting the expression of FT and CO (Yoo et al. 2005). FT and SOC1 are affected by the change of photoperiod from the leaf to the apical meristem, promoting the plant from vegetative to reproductive growth (Immink et al. 2012). Autonomous and vernalization pathways inhibit the expression of flowering repressor FLC and then promote flowering. FLC delays flowering by delaying SOC1 expression in meristems by inhibiting FT (Li et al. 2016a; Richter et al. 2019). The gibberellin pathway becomes the main flowering pathway that affects SOC1, rather than FLC and FT in SDs (Moon et al. 2003). In Arabidopsis, SOC1 and SOC1-like (AGL42, AGL71, AGL72) together constitute a subgroup of the MIKCC-type MADS-box transcription factors, and its main function is to participate in the control of the plants flowering time. Involved in the regulation of floral organ development, SOC1 directly controls SOC1-like to balance these SOC1-like expression levels (Dorca-Fornell et al. 2011), and thus precisely controls flowering. The SOC1 subgroup have been cloned in various plants, such as barley (Papaefthimiou et al. 2012), soybean (Zhong et al. 2012), mango (Wei et al. 2016), Brassica juncea (Li et al. 2019) etc. There is no report on cloning and research of SOC1 subgroup genes in the Lamiales order.

According to the latest Chile climate simulation research, the suitable cultivation area of traditional short-day chia genotypes in China is restricted to mountainous regions of southern Yunnan and Taiwan (Cortés et al. 2017). This research group took the lead in carrying out the chia research in China. Since 2016, it was found that sowing in spring and summer in Beibei and Hechuan (30° N), Chongqing, China, chia bloomed in October, and a small amount of mature seeds could be harvested in late autumn. Though a small proportion of the seeds on the ear could mature, most seeds could not reach full maturity, indicating that the southern part of the Sichuan Basin, which is warm and frost-free in the winter, is a planting northern margin of traditional chia. This means that if flowering time is promoted a little, chia can be grown not only in China's low-latitude climate-friendly regions but also in the entire Sichuan Basin and many other parts of China's mid-latitudes. Even for the whole world, short-day photoperiod habit and late flowering of chia is the crucial limiting factor for its crop spreading, thus dissecting chia flowering mechanism has become a necessary basic work. In this study, two SOC1 genes were cloned from chia, the basic characteristics of genes and proteins were analyzed, the plant SOC1 evolutionary characteristics were revealed, and their transcriptional organ-specificity and responsiveness to multiple hormones under long/short photoperiods, circadian rhythms, seasonal transitions and abiotic stresses were also investigated. It will promote the understanding of the chia flowering mechanism, and enrich the understanding of SOC1.

Materials and Methods

Plant materials, treatment, and nucleic acid extraction

Chia was grown at Hechuan Farm, Southwest University, sown on May 24, 2016. On August 21–22, September 5–6, September 20–21 and October 5–6, mature leaves were collected at 2:58, 5:58, 9:28, 12:58, 16:28, 19:58, 23:28 of the day. They were used for gene cloning and to detect diurnal styles of gene expression. Root (Ro), stem (St), young leaves (YL), mature leaves (ML), young buds (YB, about 5 days old), semi-mature buds (SMB, about 10 days old), mature buds (MB, about 15 days old), flowers (Fl), early-stage seeds (ES, about 10 days old), middle-stage seeds (MS, about 20 days old) and late-stage seeds (LS, about 30 days old) were sampled for detecting organ-specificity of the cloned genes.

The methods used to cultivate the seedlings of chia in the artificial climate chamber referred by Xue et al. (2017). The six-leaf stage seedlings were moved to the plant growth chamber for treatment with 2 styles of photoperiods. The long-day treatment was 16 h-light and 8h-dark, and the short-day treatment was 12h-day and 12-night, with constant temperature of 30ºC and relative humidity of 56%. Each photoperiod treatment lasted for one week. Four hormone treatments were carried out, i.e. 80 μmol L^-1 kinetin (KT), 2 μmol L^-1 brassinolide (BR), 200 μmol L^-1 gibberellin (GA₃) and 250 μmol L^-1 indole acetic acid (IAA). Each hormone was treated for 0 d (control/CK, basal level), 1 d, 3 d and 9 d. Mature leaves were sampled at each time point for characterization of phytohormone responsiveness of ShSOC1-1 and ShSOC1-2.

Chia seedlings were cultured in the artificial climate chamber and subjected to high temperature at 38ºC, low temperature at 4ºC, mechanical wounding, 100 μmol L^-1 methyl jasmonate (MeJA), 100 μmol L^-1 abscisic acid (ABA), 1 mmol L^-1 salicylic acid (SA), 300 mmol L^-1 sodium chloride (NaCl) and 10% polyethylene glycol 6000 (PEG6000). At 0 h, 0.5 h, 3 h, 9 h, 24 h, and 48 h time points after treatment, mature leaf samples were taken for characterization of stress responsiveness of cloned genes (Xue et al. 2017).

Each study had 3 biological replicates. Samples were kept in liquid nitrogen for transportation and stored at -80°C. Total cellular RNA was extracted using the Biospin Plant Total RNA Extraction Kit (BioFlux, China), and total gDNA was extracted from mature leaves using general CTAB method. Electrophoresis and spectrophotometric detection were adopted to detect the quality and quantity of the nucleic acids.

Cloning of the conservative sequences of SOC1 genes from chia

In order to clone the conserved region of chia SOC1 genes, the A. thaliana SOC1 mRNA was retrieved from NCBI GenBank (NM_130128.4), and used as an electron probe for the in silico cloning the orthologous sequence from chia-relative species Sesamum indicum (sesame), Erythranthe guttatu, Salvia pomifera and S. miltiorrhiza, since there was no chia sequence in the GenBank. All SOC1 annotation mRNA, TSA, EST and gDNA tag sequences were downloaded and multiple sequence alignment was performed. According to the conservative sites of SOC1 alignments, degenerate primer combination FLSOC1C + RLSOC1C was designed (Table 1). One μg of total RNA mixed from all organs was subjected to gDNA deletion and reverse-transcription using the PrimeScript Reagent Kit with gDNA Eraser (TaKaRa Dalian, China) to obtain the first-strand total cDNA. Then it was used as a template for amplification of the conservative regions of chia SOC1 genes using conventional PCR (Annealed at 58ºC and extended for 2 min). Conventional electrophoresis, gel recovery, pMD19-T vector (TaKaRa Dalian, China) recombination and Escherichia coli DH5α transformation were performed. After PCR test for positive clones, batches of clones corresponding to insert length polymorphism were sent to Shanghai Lifei Information & Technology Company (China) for sequencing using M13F/M13R primers.

5-RACE and 3-RACE of chia SOC1 genes

The results of sequencing showed that the conservative regions of 2 chia SOC1 genes were obtained, which were named as ShSOC1-1 and ShSOC1-2 respectively. Then 5'-RACE and 3'-RACE primers of ShSOC1-1 and ShSOC1-2 were designed (Table 1), according to the conservative region sequences. One μg of total RNA of organ-mixture was used as start material for RACE handling using the SMARTer™ RACE Amplification Kit (Clontech, U.S.A.) to obtain the first-strand total cDNA template of the 5'-RACE and 3'-RACE, respectively. Primers FShSOC1-13-1/FShSOC1-23-1 and FShSOC1-13-2/FShSOC1-23-2 were used for pairing with the universal primer LUPM and NUP (Table 1) for 3'-RACE primary and nested amplifications of ShSOC1-1/ShSOC1-2, respectively. The PCR annealing temperature was 63°C, and the extension time was 1 min. Primers RShSOC1-15-1/RShSOC1-25-1 and RShSOC1-15-2/RShSOC1-25-2 were matched with the universal primers LUPM and NUP (Table 1) for primary and nested amplifications of 5'-RACE of ShSOC1-1/hSOC1-2, respectively. The PCR annealing temperature was 62°C and the extension time was 1 min. Electrophoresis, gel recovery, TA cloning and sequencing were performed.

Cloning of full-length sequences of chia SOC1 subfamily genes

Based on the conservative regions and 5'-RACE and 3'-RACE results, we can obtain the full-length cDNAs of ShSOC1-1 and ShSOC1-2 using Vector NTI assemblage function. Based on this, we designed the primer combinations FShSOC1-1 + RShSOC1-1 and FShSOC1-2 + RShSOC1-2 (Table 1). The two full-length cDNAs were amplified by PCR using 3'-RACE template, annealed at 55°C and extended for 2 min. Electrophoresis, gel recovery, TA cloning and sequencing were performed.

Bioinformatics analysis

GenBank sequence search, BLAST, in silico cloning, and CDD detection were performed at NCBI (http://www.ncbi.nlm.nih.gov). Vector NTI Advance 11.5.1 and DNAStar version 7.1.0 software were used for sequence creation, analysis, annotation, translation, comparison, assembly and other analysis. Protein analysis were performed at Expasy (http://www.expasy.org), GSDS2.0 (http://gsds.cbi.pku.edu.cn/), CBS (http://www.cbs.dtu.dk/ Services/). According to the NCBI BLASTp chia SOC1 subfamily results, completed genome sequencing and representative species in plant taxonomy were selected, and then all their SOC1 proteins sequences were electronically cloned, and then multiple comparisons were performed using ClustalX V2.0 to generate fst files. SeaView 4.0 uses the muscle pattern to perform multiple comparisons. Under Distance Method and BioNJ method, Distance=Poisson and Bootstrap=1000 are set to build the phylogenetic tree and display the tree in Squared format.

qRT-PCR detection of transcript expression of chia SOC1 genes

The transcriptional expression of ShSOC1-1 and ShSOC1-2 was detected by using FShSOC1-1RT + RShSOC1-1RT, FShSOC1-2RT + RShSOC1-2RT primer pairs, respectively. The 25SrRNA was detected by F25SRT + R25SRT as internal control (Table 1). qRT-PCR was performed on a CFX Connect™ Real-Time PCR Detection System (Bio-Rad, U.S.A.) with a program of 95°C for 10 min, and 45 cycles of amplification (95°C for 10 sec, 60°C for 20 sec, 72°C for 10 sec). When qRT-PCR was completed, the temperature was raised from 65°C to 95°C and the melting curve was detected to confirm the specificity of the amplification.

Results

Cloning of full-length cDNAs of ShSOC1-1 and ShSOC1-2 genes

Electrophoresis analysis of PCR product of amplification for the conservative sequences of the chia SOC1 genes showed a 0.6-kb specific band. Sequencing of 20 positive clones produced 2 member genes and NCBI BLASTn showed orthologs to the SOC1 (AtAGL20) and SOC1-like (AtAGL42) of plants, and they were named as ShSOC1-1 and ShSOC1-2, respectively. No significant band was found in the primary amplifications of 5'-RACE and 3'-RACE of ShSOC1-1 and ShSOC1-2, with only smear at the predicted size. The 5'-RACE nested PCRs of ShSOC1-1 and ShSOC1-2 each generated a band of about 350 bp. After TA cloning, 5'-RACE clones had insert length polymorphisms. The net length of ShSOC1-1 clones after batch sequencing was 206, 231, 418, 377 bp, with intron retention in some clones. The net length of ShSOC1-2 clones was 251, 286, 309, and 310 bp. The 3'-RACE of ShSOC1-1 and ShSOC1-2 nested PCR generated 2 bands of about 0.35 kb and 0.45 kb, respectively. All the 3'-RACE clones had polymorphic insert length after TA cloning. The net ShSOC1-1 clones were 298, 344, 349, and 397 bp, with a net length of 369, 430, 435, 461, and 485 bp for ShSOC1-2 (Poly A not included). Based on the RACE results, about 1.2 kb and 1 kb band identical to the expected size was obtained by amplifying the full-length cDNA of ShSOC1-1 and ShSOC1-2 using end-to-end PCR primer combinations. The sequences corresponded to the assembled ones. Sequence analysis revealed intron retention in the 5'-RACE of some of the mRNA molecules of ShSOCl-1, so ShSOCl-1 has 3 versions of mRNA, whereas ShSOCl-2 has only one version of mRNA. We chia total gDNA was used as template to amplify the full-length gDNAs of ShSOC1-1 and ShSOC1-2, which was failed even we replaced reagents such as enzymes and optimized the amplification cycle parameters, indicating that they either have very long introns or have very complex structures.

Structure and features of ShSOC1-1 and ShSOC1-2 genes

Table 1: Primers used in cloning and qRT-PCR detection of SOC1 genes from Chia

Primers name	Primers sequence (5’→3’)	Application
FLSOC1C	AGAAATGGGCTGYTGAAGAARGC	Forward primer for Chia SOC1 conservative regions amplification
RLSOC1C	GGBRGNCCDATGAACAATTCNGTCDCNAC	Reverse primer for Chia SOC1 conservative regions amplification
FShSOC1-13-1	TTGAGCGCAGTGTCACCACCATTCGT	GSP for ShSOC1-1 3'-RACE primary amplification
FShSOC1-13-2	TTGGACTTCAAACACAAGGTGGAGG	GSP for ShSOC1-1 3’-RACE nested amplification
FShSOC1-23-1	TCCAGCGAAGCCTACACAATGTCAGG	GSP for ShSOC1-2 3’-RACE primary amplification
FShSOC1-23-2	GTGAAGTTAGGGAAACAGAAAGAGAGAG	GSP for ShSOC1-2 3’-RACE nested amplification
RShSOC1-15-1	CTGCATATTATGCTCCGAAGGTGGATTG	GSP for ShSOC1-1 5'-RACE primary amplification
RShSOC1-15-2	TGAGCTTGCAAATTCATGGAGCTTGCC	GSP for ShSOC1-1 5’-RACE nested amplification
RShSOC1-25-1	CTTCATGCGTTGTTCGACTTCATTGCC	GSP for ShSOC1-2 5’-RACE primary amplification
RShSOC1-25-2	TTGGAGCTTGAGAACTCATAAAGTCTTCC	GSP for ShSOC1-2 5’-RACE nested amplification
LUPM	CTAATACGACTCACTATAGGGCAAGCAGTGGTATCAACGCAGAGT	Anchor primer for 5'-and 3'-RACE primary amplification
NUP	AAGCAGTGGTATCAACGCAGAGT	Anchor primer for 5'-and 3'-RACE nested amplification
FShSOC1-1	ACTGTGAATATTACTCCTAGTACTACTAC	ShSOC1-1 full-length forward primer
RShSOC1-1	AATGCATAAAAAAGTTGTCATTAGTTAATAAATA	ShSOC1-1 full length reverse primer
FShSOC1-2	CCTCCCTCTCTCTCTCTCTCTCTCACAT	ShSOC1-2 full-length forward primer
RShSOC1-2	AGATTAAAGATGCATCCAAAAGAATTTCCAG	ShSOC1-2 full length reverse primer
F25SRT	GATTTCTGCCCAGTGCTCTGAA	25SrRNA qRT-PCR forward primer
R25SRT	TCTGCCAAGCCCGTTCCCTT	25SrRNA qRT-PCR reverse primer
FShSOC1-1RT	GGCTACTTGGTGAAGGGTTAGG	ShSOC1-1 qRT-PCR forward primer
RShSOC1-1RT	CTCTCCTCGTTCGATCCTCCT	ShSOC1-1 qRT-PCR reverse primer
FShSOC1-2RT	GCAATGAAGTCGAACAACGCAT	ShSOC1-2 qRT-PCR forward primer
RShSOC1-2RT	CATTGTGTAGGCTTCGCTGGA	ShSOC1-2 qRT-PCR reverse primer

Fig. 1 shows that ShSOC1-1 has 3 versions of mRNA (GenBank Accession Numbers MF577048, MF577049 and MF577050). The longest standard mRNA is 1103 bp (Poly A not included), the longest mRNA of 5'-UTR intron-retention is 1299 bp, and the longest mRNA is 1112 bp which has a 9-bp alternative splicing at the right border of the second intron. The longest mRNA of ShSOC1-2 is 1003 bp (Poly A not included, GenBank Accession Number MF577051). The normal 5′-UTR, ORF and 3′-UTR of ShSOC1-1/ShSOC1-2 are 196/128 bp, 666/642 bp, and 241/233 bp, respectively. The G+C content of 5′-UTR, ORF, and 3′-UTR of ShSOC1-1/ShSOC1-2 were 36.2/41.4%, 52.0/44.8% and 29.5/29.2%, respectively. There were 3 and 4 transcription initiation sites for ShSOC1-1 and ShSOC1-2, with 4 and 6 poly A tail sites, respectively. The identity percentages between ShSOC1-1 and ShSOC1-2 on mRNA and ORF levels were 54.1 and 58.7%, respectively. BLASTn showed that ShSOC1-1 has higher homology with sesame SOC1, followed by E. guttata SOC1; and ShSOC1-2 has higher homology with sesame AGL42, followed by E. guttata SOC1-like. The phylogenetic tree of coding regions indicates that ShSOC1-1 and ShSOC1-2 are orthologous to the SOC1 and SOC1-like in the plant kingdom, respectively.

Characterization of deduced ShSOC1-1 and ShSOC1-2 proteins

The ShSOC1-1 and ShSOC1-2 proteins are 221 aa and 213 aa in length, with theoretical MWs of 25.04 and 24.97 kD, pIs of 9.06 and 9.55, respectively, and which are alkaline. The identity percentage between ShSOC1-1 and ShSOC1-2 is 50.2% and the positives percentage is 61.8%. BLASTp showed that ShSOC1-1 has higher homology with E. guttata and sesame SOC1, followed by Plantago major SOC1; ShSOC1-2 has higher homology with the sesame SOC1 and AGL42, followed by the E. guttata SOC1.

Fig. 1: mRNA and encoded amino acid of ShSOC1-1 and ShSOC1-2 genes

The start codon ATG and the stop codon TGA/TAG are in underlined bold face. Transcription start sites and polyadenylation sites are in underlined and italic bold face. The possible polyadenylation signals are in italic bold face. 5’UTR-intron in ShSOC1-1 is labeled with grey background. Poly A tail is in lower case

NCBI BLASTp CDD analysis showed that ShSOC1-1 and ShSOC1-2 have MADS and Coiled coil domains. SignalP 4.1 predicted that ShSOC1-1 and ShSOC1-2 do not contain a signal peptide. BaCelLo (http://gpcr.biocomp.unibo.it/bacello/pred.htm), EpiLoc (http://epiloc.cs.queensu.ca/), Plant-mPLoc (www.csbio.sjtu.edu.cn/bioinf/plant-multi/), and YLoc (www.multiloc.org/YLoc) predicted ShSOC1-1 and ShSOC1-2 are located in the nucleus. SLP-Local predicts that they are located to the mitochondria or nucleus. Taken together, they are most probably located to the nucleus, consistent with the identity of the MADS-box transcription factors. TMHMM2.0 (www.cbs.dtu.dk/services/TMHMM/) and TOPCONS (http://topcons.net/) predicted that ShSOC1-1 and ShSOC1-2 have no transmembrane domain. NetPhos3.1 (www.cbs.dtu.dk/services/NetPhos/) predicted that ShSOC1-1 has 23 potential phosphorylation sites, including 15 S (serine), 7 T (threonine), and 1 Y (tyrosine), and ShSOC1-2 have 21 potential phosphorylation sites, with 11 S, 9 T, and 1 Y.

In the secondary structures predicted by SOPMA (https://npsa-prabi.ibcp.fr/cgi-bin/secpred_sopma.pl), α-helix of ShSOC1-1/ShSOC1-2 as high as accounted for 59.73/68.08%, β-sheet (extended strand) accounted for 7.69/14.08%, β-turn accounted for 7.24/5.16%, and random coil accounted for 25.34/12.68% (Fig. 2). ShSOC1-1 has several large α-helix in middle sites, while ShSOC1-2 has a very large α-helix throughout the middle sites and the α-helix occupies a large proportion in these two genes. However, the N-terminal α-helix of ShSOC1-1 is apparent, whereas the N-terminus of ShSOC1-2 has no α-helix.

The tertiary structures of ShSOC1-1 and ShSOC1-2 were predicted by SWISS-MODEL (https://swissmodel.expasy.org), which are similar to each other. In many flowering plants, the SOC1 genes are important flowering regulatory genes, known as the signal integrator of flowering pathway (Chen et al. 2017). In combination with the protein structures, key sites in the conserved region, and physicochemical properties of ShSOC1-1 and ShSOC1-2, it was concluded that ShSOC1-1 and ShSOC1-2 genes may be involved in the regulation of flowering traits in Chia.

Fig. 2: Predicted secondary structures of ShSOC1-1 and ShSOC1-2

Fig. 3: Phylogenetic relationship of SOC1 subgroup proteins from plant kingdom

Ac, Adiantum capillus-veneris; Ao, Asparagus officinalis; At, Arabidopsis thaliana; Atr, Amborella trichopoda; Bd, Brachypodium distachyon; Dc, Dendrobium catenatum; Eg, Erythranthe guttatus; Fv, Fragaria vesca; Th, Tarenaya hassleriana; Gm, Glycine max; Mn: Monoraphidium neglectum; Os: Oryza sativa; Pa: Picea abies; Ps: Picea sitchensis; Sh, Salvia hispanica; Si, Sesamum indicum; Th: Tarenaya hassleriana; Vv, Vitis vinifera

The phylogenetic relationship of ShSOC1 proteins and SOC1 evolution in plant kingdom

In order to systematically explore the phylogenetic relationships of the ShSOC1 genes and reveal some new features of the evolution of the plants SOC1 subgroup in the plant kingdom, we selected different taxa from the plant kingdom (chlorophytas, ferns, gymnosperms, monocots, and dicots) with whole genomes being sequenced to carry out phylogenetic analysis. Their SOC1 proteins were electronically cloned and then ShSOC1-1 and ShSOC1-2 were used together to construct a phylogenetic tree of SOC1 proteins (Fig. 3). The phylogenetic relationships among the species showed by this tree conform to taxonomy positions established by traditional community. Some new phenomena can be observed. There is no SOC1 in non-seed plants (green aglae, ferns, etc.). Although there are MADS-box transcription factors in Monoraphidium neglectum and Adiantum capillus-veneris, there is a non-orthologous relationship with SOC1. The origins of SOC1 is accompanied by the birth of sexual reproduction through “flowering” and seed setting.

The plant SOC1 subfamily phylogenetic tree has two clusters: gymnosperms (Picea abies, Picea sitchensis) and angiosperms. The gymnosperm SOC1 cluster occupies the basal sites, and the angiosperm SOC1 cluster is divided into 4 branches. The monocots plant has 1 branch, and the dicots plant have 3 branches corresponding to Arabidopsis SOC1/AGL20, AGL42/71/72/SOC1-like, and AGL14/19/SOC1-like respectively. Within gymnosperms and monocots, it seems that SOC1 duplication events widely occurred at order or family levels, resulting in 2 or more SOC1 genes in most analyzed species. The ancestor of dicots has experienced SOC1 twice duplication events, making the 3-SOC1-gene basic status of all eudicots. In Brassicales, more than one families have experienced their own AGL14 and AGL42 duplications, causing the AGL14/19 double gene and AGL42/71/72 triple gene status. At protein sequence level, AGL42 is conserved, while AGL71/72 diverged rapidly. In non-Brassicales eudicots, there is only one gene corresponding to Brassicales AGL14/19 and AGL42/71/72 respectively. Besides Brassicales AGL71/72, soybean GmSOC1-like/XP_003536488.1 and GmSOC1-like/XP_006606087.1, strawberry FvAGL14/XP_011463796.1, Brachypodium distachyon BdMADS50-like/NP_001289808.1, rice OsSOC1-like/XP_015625203.1 etc. also evolved rapidly at protein sequence level. It is noteworthy that the basal angiosperm Amborella trichopoda (common ancestor of monocots and dicots) has only one SOC1 gene, indicating that angiosperms SOC1 duplication events occurred after separation from basal angiosperms, and the non-duplicated and non-diversified status of important floral genes such as SOC1 might be associated with the primitive and simple floral traits of basal angiosperms.

The phylogenetic tree also showed that ShSOC1-1 and ShSOC1-2 were orthologous to A. thaliana SOC1/AGL20 and AGL42/71/72, respectively. They are highly homologous to orthologous genes from Salvia, S. indicum and E. guttata, followed by other dicots plants. Since the orthologous gene corresponding to AtAGL14/19 is common in dicots, it is speculated that this orthologous gene also exists in chia and needs to be cloned.

Organ specificity of ShSOC1-1 and ShSOC1-2 expression

The results of qRT-PCR showed (Fig. 4) that the expression of ShSOC1-1 and ShSOC1-2 in all organs was significantly different from each other. ShSOC1-1 was highest in stems and buds, and high or significant in mature leaves, semi-mature buds, mature buds and flowers, but lowest in seeds. ShSOC1-2 was highest in young buds and significantly lower in other organs, especially in seeds. According to the relative quantitative preliminary judgments, the expression of ShSOC1-1 in various organs is generally higher than that of ShSOC1-2, especially in stems, leaves, buds and flowers, implying that ShSOC1-1 may be more effective than ShSOC1-2 in regulating whole plant overall functions as well as the basal role during the flowering process. ShSOC1-2 is mainly synergistically up-regulated at the key time-point for flowering determination.

Fig. 4: Relative expression of ShSOC1-1 and ShSOC1-2 genes in different chia organs

Ro: root; St: stem; YL: young leaf; ML: mature leaf; YB: young bud; SMB: semi-mature bud; MB: mature bud; Fl: flower; ES: early-stage seed; MS: middle-stage seed; LS: late-stage seed

Circadian rhythm of ShSOC1-1 and ShSOC1-2 expression and its response to long-short photoperiod seasonal shift

The qRT-PCR was used to examine the circadian rhythms of ShSOC1-1 and ShSOC1-2 in mature leaves and the response to the seasonal change from long to short photoperiod. The results showed that the expression of ShSOC1-1 and ShSOC1-2 was significantly different (Fig. 5). On August 21–22 (LD, sunny, 28–38 °C), ShSOC1-1 was lower in the morning and did not change very much, upregulated from the afternoon and increased significantly at night, but fluctuated slightly at midnight; ShSOC1-2 was lower throughout the day, and there was almost no significant fluctuation. On September 5–6 (LD, rainy, 20–24°C), affected by the rainy weather, ShSOC1-1 was very low with less fluctuation within a whole day, while ShSOC1-2 was low in morning, with a peak in afternoon, and then gradually down-regulated until maintaining a relatively stable low level. On September 20–21 (Autumnal equinox, sunny, 20–28°C), the two genes were very similar, with very weak and small changes during the daytime, but increased at night, peaked at midnight, and then decreased. On October 5–6 (SD, cloudy to overcast, 20–29°C), the two genes within a whole day were significantly similar to those of August 21–22 and September 20–21; ShSOC1-1 was low during the daytime and high at night, peaked at midnight, and then gradually decreased; ShSOC1-2 was relatively lower with little change within a whole day. Taken together, ShSOC1-1 is characterized by predominant expression at nights on sunny days but suppressed by rainy weather, while ShSOC1-2 is inhibited by high temperatures sunny days in summer, and dominantly expressed in rainy afternoon and in midnight of autumnal equinox (critical time for determining flowering and early floral differentiation).

Effect of phytohormones on the expression of ShSOC1-1 and ShSOC1-2 under long- and short-photoperiods

In the present study, KT, BR, GA₃, and IAA treatments were performed on 6-leaf chia seedlings under long/short-day conditions (LD/SD) respectively, and the expression of ShSOC1-1 and ShSOC1-2 was detected by qRT-PCR (Fig. 6). There are similarities and differences among hormones as well as between genes and photoperiods. After BR treatment, ShSOC1-2 and ShSOC1-2 were slightly inhibited under LDs or SDs, with ShSOC1-2 being more sensitive. After KT treatment, ShSOC1-1 and ShSOC1-2 were significantly inhibited under SDs, with ShSOC1-2 being more sensitive, and ShSOC1-1 restored earlier. However, the expression trends of ShSOC1-1 and ShSOC1-2 were consistent under LDs, firstly upregulated and then fluctuated, but ShSOC1-1 was more sensitive. After IAA treatment, ShSOC1-1 and ShSOC1-2 showed a slowly inhibition under SDs, but the two genes had the same tendency under LDs, firstly up-regulated and then fluctuated, with ShSOC1-1 being more sensitive. After GA₃ treatment, ShSOC1-1 had little effect on under SDs but ShSOC1-2 was significantly upregulated; however, ShSOC1-2 had little effect but ShSOC1-1 was upregulated under LDs. Taken together, the expression of ShSOC1 was down-regulated by BR, KT, and IAA under SDs, but upregulated by GA₃, and ShSOC1-2 was more sensitive than ShSOC1-1. Under LDs, the expression of ShSOC1 genes were down-regulated by BR and upregulated by KT, IAA and GA₃, and ShSOC1-1 was more sensitive than ShSOC1-2.

Effect of abiotic stresses on the expression of ShSOC1-1 and ShSOC1-2

The MADS-box transcription factor SOC1 subfamily genes regulate plant growth as well as control flowering time and have important basic functions, but reports on abiotic stresses influence on their expression are limited. We used 5-week old chia seedlings for treatments with a variety of abiotic stresses and examined changes in the expression of ShSOC1-1 and ShSOC1-2 based on qRT-PCR (Fig. 7). After cold treatment at 4°C, ShSOC1-1 was continuously upregulated within 48 h, while ShSOC1-2 was quickly and significantly inhibited, then slowly restored to basal level at 48 h. After treatments with heat stress at 38°C and NaCl, ShSOC1-1 and ShSOC1-2 showed fluctuation and a rough down-regulation, though ShSOC1-1 was transiently upregulated at 0.5 h under high temperature and at 9 h under NaCl. ABA quickly inhibited their expression, but the recovery was also quick. Although there were small fluctuations in the expression of ShSOC1-1 and ShSOC1-2 after PEG treatment, the overall change was not clear. After MeJA, SA, and mechanical injury treatments, ShSOC1-1 and ShSOC1-2 showed a wavy up-regulation, especially when SA treatment was at 48 h, ShSOC1-1 and ShSOC1-2 were upregulated by 5–10 folds.

Fig. 5: Circadian rhythm of ShSOC1-1 and ShSOC1-2 expression, and response to seasonal change from long to short photoperiod

Fig. 6: Influence of phytohormones on the expression of ShSOC1-1 and ShSOC1-2 under long-photoperiod (-L) and short-photoperiod (-S) respectively

Fig. 7: Influence of abiotic stresses on the expression of ShSOC1-1 and ShSOC1-2

Discussion

In A. thaliana, all SOC1 subfamily genes AGL20/SOC1, AGL42, AGL71, AGL72, AGL14 and AGL19 participate in the regulation of flowering, and are integrators of photoperiod, vernalization, gibberellin, autonomy, aging pathway and other flowering pathway signals, which occupy a central position in regulating flowering. SOC1 is dominant within the subfamily, which regulates flowering time by regulating the same family members AGL42, AGL71, AGL72, etc. (Lee et al. 2000; Moon et al. 2003; Dorca-Fornell et al. 2011; Immink et al. 2012). The Arabidopsis flowering suppressor gene FLC normally inhibits expression of SOC1 and rapidly upregulates SOC1 in short-day flc mutants (Hepworth et al. 2002; Moon et al. 2003). Functional studies have shown that the function of the SOC1 subfamily to promote flowering is at least conserved in angiosperms. In dicots, deletion of Arabidopsis SOC1 and SOC1-like leads to late flowering (Hepworth et al. 2002). The overexpression of cabbage BrAGL20 into Brassica napus resulted in early floral phenotype (Hong et al. 2013). Mangifera indica MiSOC1 was transformed into Arabidopsis, which resulted in an early flowering stage (Wei et al. 2016). G. hirsutum GhSOC1 also promotes flowering after A. thaliana transformation, and it can regulate APETALA1/FRUITFULL-like gene GhMADS42 to regulate flower organ morphology (Zhang et al. 2016). In monocot plants, overexpression of rice OsSOC1 greatly advanced flowering time, OsSOC1 was transformed to Arabidopsis soc1 mutants to normalize flowering time (Andersen et al. 2004), and RNA interference of maize ZmSOC1 gene postponed flowering time, while ZmSOC1 overexpression or heterologous expression in Arabidopsis promoted early flowering (Alter et al. 2016). Phyllostachys praecox SOC1 orthologous gene PvMADS56 was transformed to Arabidopsis and the flowering time was also accelerated (Liu et al. 2016b). Dendrobium nobile DnAGL19 gene transformed into A. thaliana promoted flowering through the HOS1-FT pathway (Liu et al. 2016c).

With a few exceptions, FvSOC1 gene is an inhibitor of the flowering time in the perennial short-day plant wild strawberry, which regulates vegetative growth and reproductive growth respectively, through independent pathways (Mouhu et al. 2013). Overexpression of G. hirsutum GhSOC1 in G. jamesonii does not advance the flowering time, but results in the decline of floral organs (Ruokolainen et al. 2011). Actinidia chinensis SOC1-like subfamily may not be involved in controlling the flowering time, but may affect the length of dormancy (Voogd et al. 2015). The I domain and C domain are lost in the protein encoded by Kalanchoe daigremontiana KdSOC1 gene, and its function is to play an important role in the vegetative propagation of adventitious buds through the auxin signaling pathway, and its overexpression affects plant morphology (Liu et al. 2016a).

Chia ShSOC1-1 and ShSOC1-2 encode proteins with typical SOC1 full domain features, which have also typical conserved domains and conserved sites. Their transcripts are highest in the early stage of flower bud differentiation. Given that higher plants SOC1 subfamily generally regulates flowering and floral organ differentiation, it is speculated that the ShSOC1 subfamily may be a positive regulator of chia flowering and early floral organ differentiation.

A large number of literatures have shown that phytohormones are involved in regulating flowering time, especially the gibberellin signaling pathway is one of the five major pathways of flowering induction, and the hormone pathway is intertwined with the photoperiodic and vernalization reactions (Shi et al. 2019). Although SOC1 is one of the most important integrators of the five major pathways of flowering induction, there are not many reports on its response to the phytohormones. In view of this, four phytohormones were sprayed on chia plants under LDs and SDs in the present study. The results showed that under SDs ShSOC1 subfamily was down-regulated by BR, KT and IAA, but upregulated by GA₃. ShSOC1-2 was more sensitive than ShSOC1-1. ShSOC1 subfamily was down-regulated under LDs, while ShSOC1-1 was up-regulated by KT, IAA, and GA₃. ShSOC1-1 was more sensitive than ShSOC1-2. Arabidopsis SOC1 was slowly up-regulated after GA₃ treatment under both SDs and LDs (Moon et al. 2003), which is consistent with the trend in chia of this study. However, wheat SOC1 and SOC1-like are down-regulated when treated with GA₃ under SDs (Pearce et al. 2013), which is contrary to the trend in chia of this study. It seems that low-temperature long-day plants and low-temperature short-day plants remain the same, and it is the opposite of high-temperature long-day plants. This study systematically revealed the response expression characteristics of chia SOC1 subfamily to four phytohormones under LDs and SDs.

This study showed that, in short-day plant chia, ShSOC1-1 and ShSOC1-2 are very different from each other in response to circadian rhythm and seasonal transition of the long-short photoperiod during the summer to autumn. ShSOC1-1 is characterized by predominant expression at nights on sunny days but suppressed by rainy weather, while ShSOC1-2 is inhibited by high temperatures sunny days in summer, and is dominantly expressed in rainy afternoon and in midnight of autumnal equinox (critical time for determining flowering and early floral differentiation). In SDs, short-day plant soybean SOC1 and SOC1-like family genes were low during the daytime, high at night, peaked at midnight, and then decreased, with soybean SOC1 consistent with chia SOC1 genes while SOC1-like largely different from chia SOC1 genes. In LDs, soybean SOC1 and SOC1-like are dynamic and stable, with SOC1-like being similar to chia SOC1 genes while SOC1 being totally different from the chia SOC1 trend of low-in-daytime and high-at-night (Na et al. 2013). Short-day plant Zea mays ZmMADS1 as a homologue of A. thaliana SOC1, was low during the daytime and high at night (Alter et al. 2016), which consistent with chia SOC1, regardless of LDs or SDs.

At 37°C heat treatment, the expression of ShSOC1-1 was firstly rapidly upregulated, but then it was greatly inhibited, while ShSOC1-2 was immediately down-regulated and was successively inhibited. A. thaliana SOC1 in long-day heat treatment was down-regulated, which was similar to chia (Takato et al. 2013). At 4°C cold treatment, ShSOC1-1 was continuously upregulated, while ShSOC1-2 was rapidly down-regulated, then increased continuously, and returned to basal level at 48 h, consistent with the trend of sustained up-regulation of Arabidopsis SOC1 at 4°C cold treatment under LDs (Li et al. 2017). After PEG treatment, ShSOC1-1 and ShSOC1-2 had no major change and remained stable. After ABA and NaCl treatments, ShSOC1-1 and ShSOC1-2 were firstly down-regulated and even fluctuated in midway, but returned to near basal levels by 48 h, which was slightly different from that of Arabidopsis. SOC1 was promoted in low and medium concentrations of NaCl, and inhibited in high concentrations of NaCl under LDs (Liu et al. 2013). After treatments with MeJA, SA, and mechanical injury, ShSOC1-1 and ShSOC1-2 was upregulated, especially when SA treatment was performed for 48 h, ShSOC1-1 and ShSOC1-2 was upregulated by 5–10 folds. This shows that SOC1 is an integrator gene that responds to the photoperiod, vernalization, autonomous, and gibberellin pathway, and its expression characteristics have certain fluctuations in the adverse conditions. ShSOC1-1 is continuously upregulated in response to cold treatment, suggesting that it may also be affected by vernalization pathway. Inhibition of flowering genes are inhibited under low temperature conditions, thereby promoting expression of SOC1 (Amasino 2005).

In this study, representative species with genome being completely sequenced in a typical taxonomic unit were selected (while a few representative species did not have genome sequencing, but the SOC1 subfamily was cloned), and the phylogenetic tree of SOC1 proteins in the plant kingdom was constructed. The phylogenetic trees using the SOC1 proteins are in good agreement with the established plant community classifications in the academic community (Li et al. 2012; Zhong et al. 2012; Zhang et al. 2018). It also reveals some new features of SOC1 gene evolution in the plant kingdom. There is no SOC1 in non-seed plants (green aglae, moss, ferns, etc.), and all seed plants (gymnosperms, angiosperms) have SOC1. Because the angiosperms have a true floral structure, the “flowers” of gymnosperms are only spore-containers instead of a true flower structure, so the SOC1 gene origin exactly accompanied with the origin of seed reproduction, and its function is to initiate sexual reproductive growth (strobile/flower differentiation) in gymnosperms and angiosperms in response to the best seasonal conditions, and creates conditions for follow-up reproductive behavior (strobile/flower opening, pollination, fertilization, and seed setting). Gymnosperm Cryptomeria japonica SOC1-like gene CjMADS15 and AGL6-like gene CjMADS14 regulate the development of male and female strobili (Katahata et al. 2014). Barley research also suggests that SOC1 subfamily genes not only respond to vernalization and regulate flowering but also participate in regulating seed development (Papaefthimiou et al. 2012). The plants SOC1 phylogenetic tree have two clusters: gymnosperms and angiosperms. The gymnosperm SOC1 cluster occupies the basal site (Zhong et al. 2012), and the angiosperm SOC1 cluster is divided into 4 branches. The monocots have 1 branch and the dicots have 3 branches corresponding to Arabidopsis SOC1/AGL20, AGL42/71/72/SOC1-like, and AGL14/19/SOC1-like, respectively. Within gymnosperms and monocots, it seems that SOC1 duplication events widely occurred at order or family levels, resulting in 2 or more SOC1 genes in most analyzed species. The ancestors of dicots have twice SOC1 duplication events, making the 3-SOC1-gene basic status of all eudicots. The AGL14/19 and AGL42/71/72 duplication phenomena occurred in more than one families in the Brassicales order. However, only one orthologous gene corresponds to AGL14/19 and AGL42/71/72, respectively, in non-Brassicales dicots. AGL71/72 originated from AGL42 duplication. At protein sequence level, AGL42 is conserved, while AGL71/72 diverged rapidly. It is noteworthy that the basal angiosperm Amborella trichopoda (common ancestor of monocots and dicots) has only one SOC1 gene, indicating that angiosperms SOC1 duplication events occurred after separation from basal angiosperms, and the non-duplicated and non-diversified status of important floral genes such as SOC1 might be associated with the primitive and simple floral traits of basal angiosperms. The phylogenetic tree also showed that ShSOC1-1 and ShSOC1-2 were orthologous to A. thaliana SOC1/AGL20 and AGL42/71/72, respectively. Since the orthologous gene corresponding to AtAGL14/19 is common in dicots, it is speculated that this orthologous gene also exists in chia and needs to be cloned.

Two duplication events of SOC1 in early dicots resulted in three SOC1 genes in eudicots, and functional divergence occurred, e.g. some paralogs were also involved in regulating other traits besides flowering time. Arabidopsis SOC1 subfamily controls flowering via regulating the expression of genes from the same family, i.e. AGL42, AGL71, AGL72 etc. (Dorca-Fornell et al. 2011), regulates leaf stomata opening, and prevents dark-induced chlorosis and senescence in leaves (Chen et al. 2017). AGL14 regulates auxin polarity transport and root growth in addition to flowering (Garay-Arroyo et al. 2013). The 3 P. hybrida SOC1-like genes regulate flowering redundantly, but FBP21 and UNS are related to developmental age, whereas FBP28 is more related to short-day flowering habits (Preston et al. 2014). All of the 3 Prunus mume SOC1-like genes promoted flowering in transgenic Arabidopsis, but PmSOC1-1 and PmSOC1-2 also changed flower morphology, while PmSOC1-3 did not (Li et al. 2016b). Daucus carota DcSOC1-1 was related to early bolting and flowering with significant variations among different materials, while DcSOC1-2 expression was low. In perennial short-day wild strawberry, FvSOC1 regulates both vegetative and reproductive growth (Mouhu et al. 2013). The daily expression rhythm of ParSOC1 in leaves of perennial trees P. armeniaca is related to cold demand and dormancy disruption (Trainin et al. 2013). P. mume SOC1 interacts with DAM6 to regulate both vegetative and floral bud differentiation (Kitamura et al. 2016). Divergence of expression patterns between ShSOC1-1 and ShSOC1-2 is similar to those observed in wild strawberry (Mouhu et al. 2013), Orchid Dendrobium (Ding et al. 2013) and P. praecox (Liu et al. 2016b). ShSOC1-1 expression is high in stems and buds, considerable in mature leaves, semi-mature buds, mature buds and flowers, and low in seeds. ShSOC1-2 expression is high only in young buds, but low in all other organs especially in seeds. Expression of ShSOC1-1 in various organs is generally higher than that of ShSOC1-2, implying that ShSOC1-1 is more important than ShSOC1-2 in regulating whole-plant functions and plays a basic role in determining flowering. ShSOC1-2 is sharply up-regulated in early stage of bud formation, which might play a decisive role as the "last straw" in flowering time, i.e. elevating the cellular SOC1 level above a threshold in order to initiate flower bud differentiation.

Conclusion

In this study two SOC1 genes, ShSOC1-1 and ShSOC1-2, from chia were isolated and characterized. They had typical structural, molecular and expressional features, while distinct organ-specificity and responses to diverse physiological and environmental factors indicated their functional divergence. The effect of phytohormones on chia SOC1 expression varied depending on the photoperiod. Chia SOC1 expression also changed in response to circadian rhythms, climate and seasons, and was affected by a variety of abiotic stresses. This study also revealed some new evolutionary features, especially the origin and duplication, of plant-type SOC1 genes. This study is the first report of SOC1 subfamily genes of the order Lamiales. It will promote the flowering mechanism dissection of Lamiales, and also enrich the evolution and expression characteristics of plant SOC1. Our results will promote the study photoperiodic influence on flowering from the interaction between the photoperiod and the hormone pathways, and shed light on the molecular basis of flowering induction pathways in Chia and other short-day plants.

Acknowledgements

This study was supported by the Chongqing Research Program of Basic Research and Frontier Technology (cstc2015jcyjBX0143), National Key R&D Program of China (2016YFD0100506) and the Fundamental Research Funds for the Central Universities (No. XDJK2014D009).

References

Alter P, S Bircheneder, LZ Zhou, U Schlüter, M Gahrtz, U Sonnewald, T Dresselhaus (2016). Flowering time regulated genes in maize include the transcription factor ZmMADS1. Plant Physiol 172:389‒404

Amasino RM (2005). Vernalization and flowering time. Curr Opin Biotechnol 16:154‒158

Andersen CH, CS Jensen, K Petersen (2004). Similar genetic switch systems might integrate the floral inductive pathways in dicots and monocots. Trends Plant Sci 9:105‒107

Ayerza R, W. Coates (2005). Chia: Rediscovering a Forgotten Crop of the Aztecs, pp: 1‒197. The University of Arizona Press, Tucson, Arizona, USA

Bai S, JG Wallis, P Denolf, J Browse (2016). Directed evolution increases desaturation of a cyanobacterial fatty acid desaturase in eukaryotic expression systems. Biotechnol Bioeng 113:1522‒1530

Chen J, X Zhu, J Ren, K Qiu, Z Li, Z Xie, J Gao, X Zhou, B Kuai (2017). Suppressor of Overexpression of CO 1 negatively regulates dark-induced leaf degreening and senescence by directly repressing pheophytinase and other senescence-associated genes in Arabidopsis. Plant Physiol 173:1881‒1891

Cortés D, HS Robledo, C Baginsky, L Morales-Salinas (2017). Climatic zoning of chia (Salvia hispanica L.) in Chile using a species distribution model. Span J Agric Res 15; Article e0302

Ding L, Y Wang, H Yu (2013). Overexpression of DOSOC1, an ortholog of Arabidopsis SOC1, promotes flowering in the orchid Dendrobium Chao Parya Smile. Plant Cell Physiol 54:595‒608

Dorca-Fornell C, V Gregis, V Grandi, G Coupland, L Colombo, MM Kater (2011). The Arabidopsis SOC1-like genes AGL42, AGL71 and AGL72 promote flowering in the shoot apical and axillary meristems. Plant J 67:1006‒1017

Garay-Arroyo A, E Ortiz-Moreno, MDLP Sánchez, AS Murphy, B García-Ponce, N Marsch-Martínez, SD Folter, A Corvera-Poiré, F Jaimes-Miranda, MA Pacheco-Escobedo, JG Dubrovsky, S Pelaz, ER Álvarez-Buylla (2013). The MADS transcription factor XAL2/AGL14 modulates auxin transport during Arabidopsis root development by regulating PIN expression. EMBO J 32:2884‒2895

Hepworth SR, F Valverde, D Ravenscroft, A Mouradov, G Coupland (2002). Antagonistic regulation of flowering-time gene SOC1 by CONSTANS and FLC via separate promoter motifs. EMBO J 21:4327‒4337

Hong JK, SY Kim, KS Kim, SJ Kwon, JS Kim, AK Jin, SI Lee, YH Lee (2013). Overexpression of a Brassica rapa MADS-box gene, BrAGL20, induces early flowering time phenotypes in Brassica napus. Plant Biotechnol Rep 7:231‒237

Immink RGH, D Posé, S Ferrario, F Ott, K Kaufmann, FL Valentim, SD Folter, FVD Wal, ADJV Dijk, M Schmid, GC Angenent (2012). Characterization of SOC1’s central role in flowering by the identification of its upstream and downstream regulators. Plant Physiol 160:433‒449

Jia HW, D Luo, B Cong, ZZ Gao, JY Shao, KM Cao, CR Sun (2000). Cloning and expression of a new MADS-box gene from rice. Acta Bot Sin 42:490‒495

Katahata SI, N Futamura, T Igasaki, K Shinohara (2014). Functional analysis of SOC1-like and AGL6-like MADS-box genes of the gymnosperm Cryptomeria japonica. Tree Genet Genomics 10:317‒327

Kim W, D Latrasse, C Servet, DX Zhou (2013). Arabidopsis histone deacetylase HDA9 regulates flowering time through repression of AGL19. Biochem Biophys Res Commun 432:394‒398

Kitamura Y, T Takeuchi, H Yamane, R Tao (2016). Simultaneous down-regulation of DORMANCY-ASSOCIATED MADS-BOX6 and SOC1 during dormancy release in japanese apricot (Prunus mume) flower buds. J Hortic Sci Biotechnol 91:476‒482

Lee H, SS Suh, E Park, E Cho, JH Ahn, SG Kim, JS Lee, YM Kwon, I Lee (2000). The AGAMOUS-like 20 MADS domain protein integrates floral inductive pathways in Arabidopsis. Gene Dev 14:2366‒2376

Li C, H Gu, W Jiang, C Zou, D Wei, Z Wang, Q Tang (2019). Protein interactions of SOC1 with SVP are regulated by a few crucial amino acids in flowering pathways of Brassica juncea. Acta Physiol Plantarum 41:43

Li H, K Ye, Y Shi, J Cheng, X Zhang, S Yang (2017). BZR1 positively regulates freezing tolerance via CBF-dependent and CBF-independent pathways in Arabidopsis. Mol Plant 10:545‒559

Li W, X Liu, Y Lu (2016a). Transcriptome comparison reveals key candidate genes in response to vernalization of Oriental lily. BMC Genomics 17; Article 664

Li Y, Z Xu, W Yang, T Cheng, J Wang, Q Zhang (2016b). Isolation and functional characterization of SOC1-like genes in Prunus mume. J Amer Soc Hortic Sci 141:315‒326

Li ZN, JG Zhang, GF Liu, XM Li, C Lu, JW Zhang, MZ Bao (2012). Phylogenetic and evolutionary analysis of A-, B-, C- and E-Class MADS-box genes in the basal eudicot Platanus acerifolia. J Plant Res 125:381‒393

Liu C, C Zhu, HM Zeng (2016a). Key KdSOC1 gene expression profiles during plantlet morphogenesis under hormone, photoperiod, and drought treatments. Genet Mol Res 15:1–24

Liu SN, TT Qi, JJ Ma, TF Ma, LY Ma, XC Lin (2016b). Ectopic expression of a SOC1 homolog from Phyllostachys violascens alters flowering time and identity of floral organs in Arabidopsis thaliana. Trees 30:2203‒2215

Liu T, Y Li, J Ren, Y Qian, X Yang, W Duan, X Hou (2013). Nitrate or NaCl regulates floral induction in Arabidopsis thaliana. Biologia 68:215‒222

Liu XR, T Pan, WQ Liang, L Gao, XJ Wang, HQ Li, S Liang (2016c). Overexpression of an orchid (Dendrobium nobile) SOC1/TM3-like ortholog, Dnagl19, in Arabidopsis regulates HOS1-FT expression. Front Plant Sci 7; Article 00099

Ma G, G Ning, W Zhang, J Zhan, H Lv, M Bao (2011). Overexpression of Petunia SOC1-like gene FBP21 in tobacco promotes flowering without decreasing flower or fruit quantity. Plant Mol Biol Rep 29:573‒581

Moon J, SS Suh, H Lee, KR Choi, CB Hong, NC Paek, SG Kim, I Lee (2003). The SOC1 MADS-box gene integrates vernalization and gibberellin signals for flowering in Arabidopsis. Plant J 35:613‒623

Mouhu K, T Kurokura, EA Koskela, VA Albert, P Elomaa, T Hytönen (2013). The Fragaria vesca homolog of SUPPRESSOR OF OVEREXPRESSION OF CONSTANS1 represses flowering and promotes vegetative growth. Plant Cell 25:3296‒3310

Na X, B Jian, W Yao, C Wu, W Hou, B Jiang, Y Bi, T Han (2013). Cloning and functional analysis of the flowering gene GmSOC1-like, a putative SUPPRESSOR OF OVEREXPRESSION CO1/AGAMOUS-like 20 (SOC1/AGL20) ortholog in soybean. Plant Cell Rep 32:1219‒1229

Papaefthimiou D, A Kapazogloua, AS Tsaftaris (2012). Cloning and characterization of SOC1 homologs in barley (Hordeum vulgare) and their expression during seed development and in response to vernalization. Physiol Plantarum 146:71‒85

Pearce S, LS Vanzetti, J Dubcovsky (2013). Exogenous gibberellins induce wheat spike development under short days only in the presence of VERNALIZATION1. Plant Physiol 163:1–13

Preston JC, SA Jorgensen, SG Jha (2014). Functional characterization of duplicated SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1-Like genes in Petunia. PLoS One 9; Article e96108

Richter R, A Kinoshita, C Vincent, R Martinez-Gallegos, H Gao, ADV Driel, Y Hyun, JL Mateos, G Coupland (2019). Floral regulators FLC and SOC1 directly regulate expression of the B3-type transcription factor TARGET OF FLC AND SVP 1 at the Arabidopsis shoot apex via antagonistic chromatin modifications. PLoS Genet 15; Article e1008065

Richter R, E Bastakis, C Schwechheimer (2013). Cross-repressive interactions between SOC1 and the Gatas GNC and GNL/CGA1 in the control of greening, cold tolerance, and flowering time in Arabidopsis. Plant Physiol 162:1992‒2004

Ruokolainen S, YP Ng, VA Albert, P Elomaa, TH Teeri (2011). Over-expression of the Gerbera hybrida At-SOC1-like1 gene Gh-SOC1 leads to floral organ identity deterioration. Ann Bot 107:1491‒1499

Shi C, L Zhao, X Zhang, G Lv, Y Pan, F Chen (2019). Gene regulatory network and abundant genetic variation play critical roles in heading stage of polyploidy wheat. BMC Plant Biol 19; Article 6

Sreedhar RV, P Kumari, SD Rupwate, R Rajasekharan, M Srinivasan (2015). Exploring triacylglycerol biosynthetic pathway in developing seeds of chia (Salvia hispanica L.): A transcriptomic approach. PLoS One 10; Article 0123580

Takato H, M Shimidzu, Y Ashizawa, H Takei, S Suzuki (2013). An acyl-CoA-binding protein from grape that is induced through ER stress confers morphological changes and disease resistance in Arabidopsis. J Plant Physiol 170:591‒600

Tan FC, SM Swain (2007). Functional characterization of AP3, SOC1 and WUS homologues from citrus (Citrus sinensis). Physiol Plantarum 131:481‒495

Tang Y, Y Jiang, J Meng, J Tao (2018). A brief review of physiological roles, plant resources, synthesis, purification and oxidative stability of Alpha-linolenic Acid. Emir J Food Agric 30:341‒356

Trainin T, I Bar-Ya’Akov, D Holland (2013). ParSOC1, a MADS-box gene closely related to Arabidopsis AGL20/SOC1, is expressed in apricot leaves in a diurnal manner and is linked with chilling requirements for dormancy break. Tree Genet Genomics 9:753‒766

Voogd C, T Wang, E Varkonyi-Gasic (2015). Functional and expression analyses of kiwifruit SOC1-like genes suggest that they may not have a role in the transition to flowering but may affect the duration of dormancy. J Exp Bot 66:4699‒4710

Wei JY, DB Liu, GY Liu, J Tang, YY Chen (2016). Molecular cloning, characterization, and expression of Misoc1: A homolog of the flowering gene SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 from mango (Mangifera indica L). Front Plant Sci 7; Article 1758

Xue YF, N Yin, B Chen, F Liao, AN Win, J Jiang, R Wang, X Jin, N Lin, Y Chai (2017). Molecular cloning and expression analysis of two FAD2 genes from chia (Salvia hispanica). Acta Physiol Plantarum 39; Article 95

Yoo SK, KS Chung, J Kim, JH Lee, SM Hong, SJ Yoo, SY Yoo, JS Lee, JH Ahn (2005). CONSTANS activates SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 through FLOWERING LOCUS T to promote flowering in Arabidopsis. Plant Physiol 139:770‒778

Zhang X, J Wei, S Fan, M Song, C Pang, H Wei, C Wang, S Yu (2016). Functional characterization of GhSOC1 and GhMADS42 homologs from upland cotton (Gossypium hirsutum L.). Plant Sci 242:178‒186

Zhang Y, D Tang, X Lin, M Ding, Z Tong (2018). Genome-wide identification of MADS-box family genes in moso bamboo (Phyllostachys edulis) and a functional analysis of PeMADS5 in flowering. BMC Plant Biol 18; Article 176

Zhong X, X Dai, J Xv, H Wu, B Liu, H Li (2012). Cloning and expression analysis of Gmgal1, SOC1 homolog gene in soybean. Mol Biol Rep 39:6967‒6974