-
The relationship between MaMADS33 and dormancy was investigated by comparing MaMADS33 expression in flower buds from three mulberry varieties with different dormancy durations (LJ109, JQ63, and ZZB) using primers targeting the conserved amino acid sequence of the MADS-box domain (primer I, Supplemental Table S1). MaMADS33 expression peaked at different times in the three varieties: early November in LJ109, late November in JQ63, and January in ZZB (Fig. 1). Notably, the relative timing of these peaks was consistent with the relative timing of bud break in the three varieties (earliest in LJ109 and latest in ZZB). Expression levels of MaMADS33 also differed among the mulberry varieties. In October, November, and December, MaMADS33 expression was highest in ZZB, moderate in JQ63, and lowest in LJ109, again consistent with their relative durations of dormancy. In January, MaMADS33 expression remained high in ZZB but was low in LJ109 and JQ63. By February, no unopened buds remained on LJ109, and MaMADS33 expression was low in both JQ63 and ZZB.
Figure 1.
MaMADS33 expression is positively associated with dormancy in mulberry. MaMADS33 expression in the mulberry varieties 'Lunjiao109' (LJ109), 'Jinqiang63' (JQ63), and 'Zhenzhubai' (ZZB) was measured throughout dormancy by qRT–PCR using primers targeting the conserved sequence of the MADS-box domain. MaRPL15 was the internal control gene for normalization of the expression data (n = 3; mean ± measurement range). Significant differences are indicated by different lowercase letters (ANOVA and Duncan's multiple range test; p < 0.05).
AS affects transcriptions of MaMADS33 during dormancy
-
To understand the possible roles of AS during dormancy, MaMADS33 transcripts were next examined in flower buds of JQ63, which exhibited an intermediate dormancy duration. Using 5' RACE and 3' RACE with specific primers targeting the conserved MADS-box domain sequence (Fig. 2a & b), four distinct complete ORFs of MaMADS33: MaMADS33-AS1, MaMADS33-AS2, MaMADS33-AS3, and MaMADS33-AS4, with lengths of 327, 555, 633, and 675 bp, respectively were identified. MaMADS33-AS1 and MaMADS33-AS2 were truncated isoforms that arose from alternative last exons; MaMADS33-AS1 terminated in the second intron and MaMADS33-AS2 in the sixth intron (Fig. 2c). MaMADS33-AS3 was derived from the skipping of the seventh exon. Finally, MaMADS33-AS4 represented the complete sequence, including all eight exons.
Figure 2.
Alternative splicing (AS) of mulberry MaMADS33 from endodormancy through ecodormancy. (a) 5' RACE, and (b) 3' RACE methods were used to amplify MaMADS33 transcripts in cDNA pools from JQ63 flower buds collected on five dates from October 2020 through February 2021. Primers were designed to target the conserved sequence of the MADS-box domain. Bands corresponding to MaMADS33-AS1, MaMADS33-AS2, MaMADS33-AS3, and MaMADS33-AS4 are indicated by arrows. (c) Schematic of MaMADS33 AS isoforms. Exons are represented by green boxes and introns by lines.
MaMADS33-AS1 is a long non-coding RNA
-
To investigate the functions of the four MaMADS33 AS isoforms, their coding potential was examined and it was found that all contained a complete ORF. However, the coding potential of MaMADS33-AS1 was 0.23, whereas those of MaMADS33-AS2, MaMADS33-AS3, and MaMADS33-AS4 were all greater than 0.95. Based on their coding potentials, MaMADS33-AS1 was classified as a noncoding sequence, and the other three isoforms were classified as coding sequences (Table 1). The low coding potential of MaMADS33-AS1 was further validated by transient expression of a MaMADS33-AS1-GFP construct in N. benthamiana leaves, which resulted in no GFP signal (Fig. 3a). These findings confirmed that MaMADS33-AS1 was a long noncoding RNA.
Table 1. Coding potential of the four MaMADS33 AS isoforms were predicted using Coding Potential Calculator 2 (CPC2).
ID Label Coding probability Peptide length (aa) ORF
integrityMaMADS33-AS1 noncoding 0.232883 109 complete MaMADS33-AS2 coding 0.977475 185 complete MaMADS33-AS3 coding 0.969709 211 complete MaMADS33-AS4 coding 0.985029 225 complete Figure 3.
Assessment of MaMADS33-AS1 coding potential and alignment of amino acid sequences. (a) The coding potential of MaMADS33-AS1 was examined by transient expression in Nicotiana benthamiana leaves. Scale bar = 20 μm. (b) Amino acid sequences of MaMADS33-AS2, MaMADS33-AS3, and MaMADS33-AS4. The MADS-box and K-domains are indicated by red and black boxes, respectively. The I-domain and C-terminal domain are marked above the sequences.
MaMADS33-AS2, MaMADS3-AS3, and MaMADS03-AS4 were composed of 185, 211, and 225 amino acids, respectively (Table 1). Alignment of their amino acid sequences revealed that MaMADS33-AS2 lacked the C-terminal domain, MaMADS33-AS3 lacked 14 amino acids, mainly in the C-terminal domain, and MaMADS33-AS4 contained all the characteristic domains of MADS family proteins, including the MADS-box, I-domain, K-domain, and C-terminal domain (Fig. 3b).
MaMADS33-AS3 and MaMADS33-AS4 are positively associated with endodormancy
-
To investigate the potential functions of MaMADS33 in endodormancy, the endodormancy stages of JQ63 flower buds were first characterized(Fig. 4a). From October to December, temperatures gradually decreased (Fig. 4b), and the bud-break percentage remained below 50%, indicating that the flower buds were in an endodormant state. Next, the expression of the four MaMADS33 AS isoforms in bud tissues were measured by RT–PCR with specific primers during endodormancy and ecodormancy (Fig. 4c). Transcript levels of MaMADS33-AS1 remained low in October, increased significantly from November to December, and then decreased markedly in February. By contrast, expression of MaMADS33-AS2 increased gradually after October and remained high throughout ecodormancy. Expression of MaMASD33-AS3 and MaMADS33-AS4 was relatively high throughout the endodormancy period, peaking around November, then decreased significantly during ecodormancy. To further analyze the expression of these isoforms during endodormancy, nucleic acid bands of MaMASD33-AS3 and MaMADS33-AS4 were counted by cloning and sequencing. A total of 44 clones were sequenced, 16 (36.36%) of which were MaMASD33-AS3 and 28 (63.64%) of which were MaMADS33-AS4 (Supplemental Table S2). MaMADS33-AS4 was therefore the predominant isoform in the mixed nucleic acid bands during endodormancy. Expression of these two long protein-coding mRNAs, particularly MaMADS33-AS4, thus exhibited a positive association with endodormancy in mulberry flower buds.
Figure 4.
Expression profiles of four MaMADS33 AS isoforms during dormancy. Dormancy stages of mulberry flower buds. Flower buds of JQ63 were collected from October 2020 through February 2021, and the (a) bud break percentage (n = 3, mean ± measurement range), and (b) temperature were recorded. (c) Expression profiles of four MaMADS33 AS isoforms in flower buds measured by RT–PCR with MaRPL15 as the reference gene. PCR products were separated on 1% agarose gels.
Endodormancy induces AS of MaMADS33, resulting in accumulation of MaMADS33-AS1
-
To gain a deeper understanding of the relationship between dormancy and MaMADS33 expression patterns, the splicing efficiency of intron 2 was analyzed using qRT–PCR data. A schematic representation of the MaMADS33 gene and the primers used for amplification is provided in Fig. 5a. The amplification products generated by primer I, which targeted the conserved MADS-box domain, corresponded to all MaMADS33 transcripts (Fig. 5b). Overall transcription of MaMADS33 was high during endodormancy, then decreased significantly during ecodormancy. The amplification products of primer II, which specifically targeted transcripts from which intron 2 had been spliced, corresponded to MaMADS33-AS2, MaMADS33-AS3, and MaMADS33-AS4. Notably, the abundance of these spliced transcripts decreased after October (Fig. 5c). The amplification products generated by primer III represented unspliced pre-mRNA containing intron 2 (Fig. 5d). To quantify the splicing efficiency of intron 2 in the isoforms MaMADS33-AS2, MaMADS33-AS3, and MaMADS33-AS4, the spliced/unspliced ratio was calculated based on qRT–PCR data obtained with primer II vs primer III. The results revealed a significant decrease in splicing efficiency during endodormancy, followed by a gradual increase during ecodormancy (Fig. 5e).
Figure 5.
Splicing efficiency of MaMADS33 intron 2. (a) Schematic of the MaMADS33 gene. The locations of primers I, II, and III are indicated with arrows. (b) Primer I, located in the conserved MADS-box domain, targeted all spliced and unspliced isoforms of MaMADS33. (c) Primer II targeted mRNA from which intron 2 had been spliced, corresponding to transcripts of MaMADS-AS2, MaMADS-AS3, and MaMADS-AS4. (d) Primer III targeted unspliced pre-mRNA for intron 2. (e) Splicing efficiency was calculated as the spliced/unspliced ratio. Relative expression was measured by qRT–PCR using MaRPL15 as a reference gene (n = 3; mean ± measurement range). Flower buds of JQ63 were collected from October 2020 through February 2021. Significant differences are indicated by different lowercase letters (ANOVA and Duncan's multiple range test; p < 0.05).
MaMADS33-AS4 interacts with MaWAP18 during endodormancy
-
To further investigate the potential mechanism by which MaMADS33 functions in endodormancy, a Y2H screen of pooled RNA from dormant flower buds, mature fruits, and flower buds at the floral formation stage was performed. All interacting proteins identified are listed in Supplemental Table S3; among them was the winter-accumulating protein MaWAP18. We next performed a Y2H and β-galactosidase activity assay to further examine the interaction between MaMADS33-AS4 and MaWAP18 (Fig. 6a). The interaction between MaMADS33-AS4 and MaWAP18 was strong when MaMADS33-AS4 served as the bait and MaWAP18 as the prey. To confirm this interaction in vivo, a BiFC assay in N. benthamiana was performed. Transient co-expression of MaMADS33-AS4-YFPN and MaWAP18-YFPC in tobacco leaves resulted in a distinct yellow fluorescence signal in the guard cells, conclusively demonstrating the interaction between MaMADS33-AS4 and MaWAP18 in planta (Fig. 6b). To determine whether the expression of MaWAP18 was also associated with endodormancy, its expression profile was analyzed through time. Expression of MaWAP18 was high during endodormancy and low during ecodormancy (Fig. 7a). Subcellular localization analysis suggested that MaWAP18 was localized primarily in the cell membrane and nucleus, although it was also detected in the guard cells of N. benthamiana leaves (Fig. 7b).
Figure 6.
Interaction between MaMADS33-AS4 and MaWAP18. (a) Interaction of MaMADS33-AS4 and MaWAP18 in yeast. pGBKT7-p53 was mated with pGADT7-T as a positive control, and pGBKT7-Lam was mated with pGADT7-T as a negative control. Yeast was diluted 1, 10, and 100 fold before plating onto quadruple dropout (QDO) medium. Corresponding measurements of β-galactosidase activity are shown. Three independent experiments were performed with similar results. (b) BiFC assay in 4-week-old Agrobacterium-infiltrated N. benthamiana leaves. MaMADS33-AS4 and MaWAP18 were independently fused to the N-terminal and C-terminal halves of yellow fluorescent protein (YFP), respectively. Images of YFP fluorescence were obtained using a confocal microscope. Scale bar = 10 μm. Three independent experiments were performed with similar results.
Figure 7.
Expression and subcellular localization analyses of MaWAP18. (a) Relative expression was measured by qRT–PCR using MaRPL15 as the reference gene (n = 3; mean ± measurement range). Flower buds of JQ63 were collected from October 2020 through February 2021. (b) Subcellular localization of MaWAP18. Images of 4-week-old Agrobacterium-infiltrated N. benthamiana leaves expressing the MaWAP18-GFP fusion protein driven by the CaMV 35S promoter were obtained under green fluorescence, merged light, and visible light. 35S:GFP was used as a positive control. Scale bar = 10 μm. Significant differences are indicated by different lowercase letters (ANOVA and Duncan's multiple range test; p < 0.05).
-
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
-
About this article
Cite this article
Luo Y, Liu H, Han Y, Li W, Wei W, et al. 2024. Alternative splicing of the FLOWERING LOCUS C-like gene MaMADS33 is associated with endodormancy in mulberry. Forestry Research 4: e029 doi: 10.48130/forres-0024-0027
Alternative splicing of the FLOWERING LOCUS C-like gene MaMADS33 is associated with endodormancy in mulberry
- Received: 28 September 2023
- Revised: 02 August 2024
- Accepted: 14 August 2024
- Published online: 04 September 2024
Abstract: Alternative splicing (AS) is an important post-transcriptional process that generates multiple mRNA isoforms. FLOWERING LOCUS C (FLC) is a pivotal gene in both the vernalization and autonomous pathways of flowering plants, and MaMADS33 is one of the FLC homologs in white mulberry (Morus alba). Recent studies have revealed that MaMADS33 is involved in endodormancy, but the underlying molecular mechanism remains to be characterized. Here, a comparison of MaMADS33 expression among three mulberry cultivars with different degrees of dormancy revealed a positive association between MaMADS33 expression and dormancy. Further 3' and 5' rapid amplification of cDNA ends (RACE) analyses led to identifying four MaMADS33 isoforms derived from AS and designated MaMADS33-AS1–4. Analysis of their coding potential revealed that MaMADS33-AS1 was a long non-coding RNA. Expression profiling and splicing-efficiency analyses showed that cold stress during endodormancy induced AS of MaMADS33, resulting in a predominance of truncated isoforms, especially MaMADS33-AS1. MaMADS33-AS2 expression was upregulated during both endodormancy and ecodormancy, whereas MaMADS33-AS3 and MaMADS33-AS4 were endodormancy-associated isoforms that were upregulated during endodormancy and then downregulated during ecodormancy. MaMADS33-AS4 was used as bait for a yeast two-hybrid screen because its gene expression was higher than that of MaMADS33-AS3, and mulberry winter-accumulating 18 kDa protein (MaWAP18) was identified as an MaMADS33-AS4 interaction partner. The interaction between MaWAP18 and MaMADS33-AS4 was confirmed by a bimolecular fluorescence complementation assay. These findings offer insight into the role of FLC homologs in the endodormancy of woody plants.
-
Key words:
- FLOWERING LOCUS C /
- Mulberry /
- Alternative splicing /
- Endodormancy