VDAC1 Negatively Regulates Floral Transition in Arabidopsis thaliana

Voltage-dependent anion channels (VDACs) are the most important proteins in mitochondria. They localize to the outer mitochondrial membrane and contribute to the metabolite transport between the mitochondria and cytoplasm, which aids plant growth regulation. Here, we report that Arabidopsis thaliana VDAC1 is involved in the floral transition, with the loss of AtVDAC1 function, resulting in an early-flowering phenotype. AtVDAC1 is expressed ubiquitously in Arabidopsis. To identify the flowering pathway integrators that may be responsible for AtVDAC1′s function during the floral transition, an RNA-seq analysis was performed. In total, 106 differentially expressed genes (DEGs) were identified between wild-type and atvdac1-5 mutant seedlings. However, none were involved in flowering-related pathways. In contrast, AtVDAC1 physically associated with FLOWERING LOCUS T. Thus, in the floral transition, AtVDAC1 may function partly through the FLOWERING LOCUS T protein.

There are five VDAC isoforms in Arabidopsis, four of which have been characterized [12,13]. AtVDAC1 is involved in reproductive development. In particular, the atvdac1 knockout mutants exhibit slender and shorter siliques, resulting in a reduced seed set [13,14]. In atvdac1 plants, pollen development, including the number of pollen grains, pollen germination rate, and tube length of germinated pollen, is affected [13,15]. This suggests that zygotic or early embryo lethality may be the main reason for the decreased seed production in the atvdac1 mutant [12]. The atvdac1 mutant also has a deficiency in female development, resulting in undeveloped ovules in siliques. Furthermore, the mitochondrial transmembrane potential and ATP synthesis rate are impacted [2]. In Arabidopsis, VDAC1 negatively regulates cold responses, with the atvdac1 mutant showing a more rapid germination rate and enhanced tolerance under cold-stress conditions [16]. AtVDAC1 is also involved in Agrobacterium tumefaciens infection tolerance, given that AtVDAC1-overexpression seedlings demonstrate enhanced tolerance to Agrobacterium infections by suppressing apoptotic stress [17].
Although atvdac3 mutants are not distinguishable from wild type (WT) under standard conditions, AtVDAC3 interacts with several proteins in Arabidopsis [21][22][23]. It specifically interacts with kinesin-like protein 1 to stabilize ATP levels and balance the aerobic respiration pathways during seed germination at low temperatures [21]. AtVDAC3 also interacts with thioredoxin m2 (AtTrx m2) and Metallothionein 2b (AtMT2b), which are both involved in ROS production under salt stress. Overexpression of AtVDAC3 results in short primary roots and increased ROS accumulation levels under NaCl stress conditions, with AtTrx m2 and AtMT2b having opposite roles [22,23]. Furthermore, VDAC3 modifies the size and number of mitochondria and regulates root growth in phosphorus-and Fe-deficient conditions [24,25].
According to this research, VDACs appear to be vital in plant growth. However, the specific functions of each isoform, particularly those of VDACs implicated in flowering time, are still unknown. Here, we demonstrated that AtVDAC1 functioned as a repressor during floral transition, and the loss of AtVDAC1 function resulted in an early-flowering phenotype. However, none of the expression levels of key flowering pathway integrators were changed in atvdac1-5 mutant seedlings. In contrast, AtVDAC1 physically associated with FLOWERING LOCUS T (FT). Collectively, our results suggest that AtVDAC1 functions in the floral transition partly through the FT protein.

AtVDAC1 Participates in Flowering-Time Regulation in Arabidopsis
VDACs play important roles in plant growth, but whether VDACs are involved in flowering-time control is unknown. To identify the functions of AtVDAC1 during the floral transition, we used the atvdac1-5 mutant (SALK_058473) and atvdac1 mutant (SALK_034653), which contain an insertion in the fifth exon and sixth exon, respectively ( Figure 1A). They were identified as knock-out mutants ( Figure 1B). atvdac1-5 and atvdac1 exhibited early-flowering phenotypes compared to WT seedlings under long-day conditions ( Figure 1C,D), while atvdac1-5 and atvdac1 displayed comparable flowering times to WT seedlings under short-day conditions ( Figure 1D), suggesting that AtVDAC1 inhibits the floral transition.
To verify that the loss of AtVDAC1 function is responsible for the early-flowering phenotype of atvdac1-5, we transformed the atvdac1-5 line with the gAtVDAC1-3FLAG construct. Two independent atvdac1-5 gAtVDAC1-3FLAG transformants exhibited comparable flowering times to WT plants ( Figure 1E). These observations suggest that AtVDAC1 acts as a flowering repressor.

AtVDAC1 Is Expressed Ubiquitously in Arabidopsis
To examine the expression patterns of AtVDAC1 in Arabidopsis, a gAtVDAC1-GUS reporter construct was generated. Almost all of the gAtVDAC1-GUS transgenic lines showed similar GUS staining patterns. We then used one representative gAtVDAC1-GUS transgenic line for further investigations. This line showed a strong staining pattern in leaves of developing seedlings before and during the floral transition, which occurred 3 to 13 days after germination (DAG) under long-day conditions (Figure 2A Figure 2I). This GUS staining pattern was consistent with the quantitative real-time PCR analysis, in which the highest VDAC1 expression level occurred in leaves ( Figure 2J). Thus, AtVDAC1 is active in the leaves of developing seedlings.

Transcriptome Profiles and Differentially Expressed Gene Identification between WT and atvdac1-5 Mutant Seedlings
To identify the downstream genes of AtVDAC1 that might be responsible for its role in repressing flowering, RNA-seq analyses were performed between the WT and atvdac1-5 mutant seedlings at 9 DAG. Six RNA-seq libraries from the WT and atvdac1-5 mutant seedlings were constructed for transcriptome sequencing. The raw data were qualified, and adapters were removed, yielding approximately 6.68 Gb of sequence data from each library (Table S1). Pair-wise, Pearson's correlation coefficients of the three biological replicates of each sample indicated that the sequencing data were highly repeatable ( Figure S1). The differentially expressed genes (DEGs) between WT and atvdac1-5 mutant seedlings at 9 DAG were assessed using calculated reads per kilobase per million (RPKM) values. As a result, 106 DEGs were identified, with 89 and 17 genes induced and repressed, respectively ( Figure 3A). The expression profiles of the 106 DEGs are shown using a heatmap ( Figure 3B). To understand the functions of the DEGs, a gene ontology (GO) term enrichment analysis was performed. The dominant GO terms in biological process were "glycolytic process", "response to salt stress", "drug transmembrane transport", "response to anoxia", and "inorganic anion transport"; in molecular function, the five dominant GO terms were "catalytic activity", "ATPase activity", "voltage-gated anion channel activity", "auxin efflux transmembrane transporter activity", and "drug transmembrane transporter activity"; and in cellular component, the dominant GO terms were "mitochondrion", "mitochondrial outer membrane", "mitochondrial inner membrane", "mitochondrial envelope", and "vacuolar membrane" ( Figure 3C).
We then attempted to identify the flowering-related DEGs between WT and atvdac1-5 mutant seedlings. Surprisingly, among the 106 DEGs that were identified (Figure 3), none were involved in flowering time-related pathways.
Although the expression levels of the floral meristem identity genes are unaffected in vdac1-5 mutant at 9 DAG, this could be because AP1 and LFY expression levels were all low at that time. The expression levels of these genes were subsequently measured at 13 DAG, and we found that the expression of AP1 and LFY were all increased in vdac1-5 and atvdac1 mutants ( Figure 4A,B).

AtVDAC1 Interacts with FT
To elucidate how AtVDAC1 affects flowering time, we performed a yeast two-hybrid screening to identify interacting partners of AtVDAC1, and one of the interactors en-coded FT. To confirm the interaction between FT and AtVDAC1, we performed several experiments. First, yeast two-hybrid assays confirmed that FT interacted with AtVDAC1 ( Figure 5A). Second, FT interacted with AtVDAC1 in Nicotiana benthamiana leaves, as assessed by a luciferase complementation imaging (LCI) assay ( Figure 5B). Third, we further performed bimolecular fluorescence complementation (BiFC) analyses and found the interaction between AtVDAC1 and FT in Arabidopsis protoplasts ( Figure 5C). Thus, AtVDAC1 interacts with FT in plant cells.
To further investigate the relationship between AtVDAC1 and FT, a genetic analysis was performed. As a result, atvdac1-5 enhanced the early-flowering phenotype of SUC2:FT-9myc, SUC2:FLAG-FT, and KNAT1:FT transgenic alleles, which indicated that VDAC1 is active in the leaves and SAM (shoot apical meristem). Furthermore, atvdac1-5 ft-10 double mutants flowered earlier than ft-10 ( Figure 5D), indicating that AtVDAC1 also affects flowering through other independent targets.
To further prove that FT protein movement may be the main reason for the earlyflowering phenotype of atvdac1-5 and atvdac1, we analyzed the FT protein localization in WT, atvdac1-5, and atvdac1 mutants. As a result, we found that FT-GFP protein was visible in the mitochondria in WT, atvdac1-5, and atvdac1 mutant plants ( Figure 6). However, it is likely that there was a reduced level of FT-GFP signals in the mitochondria of atvdac1-5 and atvdac1 mutants compared to that in WT ( Figure 6), indicating that AtVDAC1 may hold FT proteins in the outer membrane of mitochondria. Simultaneously, FT-GFP signals were still visible in the mitochondria in atvdac1-5 and atvdac1 mutants, indicating that other components such as AtVDAC2, 3, 4, homologs of AtVDAC1, may also interact with FT for holding it in mitochondria.

Discussion
The VDAC proteins play important roles in metabolite transport between the mitochondria and cytoplasm, which is crucial for both mitochondrial functions and cellular energy transactions [2,3]. They also play important roles in plant growth, leaf and pollen development, and mitochondrial membrane potential maintenance [12][13][14]. Here, we showed that AtVDAC1 acts as a repressor during the floral transition, the atvdac1-5 and atvdac1 mutants have early-flowering phenotypes, and AtVDAC1 is expressed ubiquitously in Arabidopsis. To identify the flowering pathway integrators that may be responsible for AtVDAC1 functions during the floral transition, an RNA-seq analysis was performed, and DEGs were identified between WT and atvdac1-5 mutant seedlings during the floral transition. A GO term enrichment analysis of these DEGs showed that the top five dominant GO terms in cellular component were "mitochondrion", "mitochondrial outer membrane", "mitochondrial inner membrane", "mitochondrial envelope", and "vacuolar membrane" (Figure 3C), which indicated the important roles of AtVDAC1 in mitochondria.
FT protein movement may be the main reason for the early-flowering phenotype of atvdac1-5. First, transcriptomic analyses between WT and atvdac1-5 mutant seedlings revealed that none of the DEGs were involved in floral transition. Second, multiple protein -protein interaction assays demonstrated that AtVDAC1 interacts with the FT protein in vivo.
In Arabidopsis, the FT protein is generated in the companion cells of vascular tissues. The translocation of FT into sieve elements and its movement to the shoot apical meristem through the phloem requires FT to interact with multiple proteins. When FT protein moves to the shoot apical meristem cells, FT interacts with 14-3-3 proteins in the cytoplasm, and then, the FT-14-3-3 complex enters the nucleus, where it interacts with FD to form florigen activation complexes. The complexes bind to the promoter regions of floral meristem identity genes (e.g., AP1), thereby activating their gene expression to promote flowering [26,27]. Because VDAC proteins play important roles in metabolite transport between mitochondria and the cytoplasm [2,3], we speculated that AtVDAC1 interacts with the FT protein in the cytoplasm and then translocates it into mitochondria. In the mitochondria, the FT protein cannot interact with the 14-3-3 protein or FD protein, thereby failing to activate the floral meristem identity genes. When the function of AtVDAC1 was interrupted, more FT protein interacted with the 14-3-3 protein and then entered the nucleus to interact with FD. More florigen activation complexes formed in the nucleus and then increased the floral meristem identity genes' expression levels, resulting in the early-flowering phenotype of the atvdac1-5 mutant. Although the expression levels of the floral meristem identity genes were not altered in the vdac1-5 mutant at 9 DAG, the expression of AP1 and LFY were all increased in vdac1-5 and atvdac1 mutants at 13 DAG ( Figure 4A,B), possibly because at the early stage the expression of AP1 and LFY were all at a low level.
Here, we speculate that AtVDAC1 prevented flowering by translocating the FT protein into mitochondria. However, atvdac1-5 ft-10 double mutants flowered earlier than ft-10 ( Figure 5D); consequently, we believe that AtVDAC1 may also interact with other components to regulate flowering, such as TWIN SISTER OF FT (TSF), the closest homolog of FT. TSF and FT probably perform redundant roles, and AtVDAC1 may also translocate the TSF protein into mitochondria to prevent flowering. Further studies should focus on the other interacting partners of AtVDAC1 that are associated with floral-transition regulation.

Plant Materials and Growth Conditions
The Arabidopsis plants were grown under long-day (16-h/8-h, light/dark) conditions at 23 • C with a relative humidity of 75%. The plants were grown under cool-white fluorescence lamps, with a light intensity of 100 µmol·m −2 ·s −1 at the soil surface. Transgenic plants were generated through Agrobacterium tumefaciens-mediated transformation using the floral dipping method. Briefly, the gAtVDAC1-3FLAG and gAtVDAC1-GUS constructs were transformed independently into A. tumefaciens strain GV3101. Then, A. tumefaciens was cultured overnight in a 28 • C incubator at 200 rpm to reach OD600 = 1.8. Afterward, they were transformed independently into atvdac1-5 mutants using the floral dipping method. The atvdac1-5 mutants being used for transformation were approximately 4 weeks old with plenty of inflorescences. Developing floral tissues were dipped into an Agrobacterium solution containing 5% sucrose and Silwet-77 (500 µL L −1 ). After inoculation, the plants were placed under a plastic dome for 24 h and then grown under normal conditions. Seeds were obtained following self-pollination. Transformants containing gAtVDAC1-3FLAG and gAtVDAC1-GUS, independently, were selected on MS medium supplemented with hygromycin (30 mg L −1 ). Two independent homozygous T3 lines with hygromycin resistance were chosen for further studies.

Plasmid Construction
To construct gAtVDAC1-3FLAG and gAtVDAC1-GUS, a 4.5-kb AtVDAC1 genomic fragment (gAtVDAC1) containing the 2.6-kb upstream sequence and 1.9-kb full-coding sequence plus introns was amplified and subsequently cloned into binary vector pCAM-BIA1300, which already contained the 3FLAG and GUS tags, respectively. The coding sequences of FT and AtVDAC1 were amplified and subsequently cloned into the binary vectors pCAMBIA1300-35S:nLUC and pCAMBIA1300-35S:cLUC, respectively, to create 35S:FT-nLUC and 35S:cLUC-AtVDAC1. The primers used for plasmid construction are listed in Table S2.

Yeast Two-Hybrid Assay
To construct the vectors for the yeast two-hybrid assays, the coding regions of AtV-DAC1 and FT were amplified and cloned into pGADT7 and pGBKT7, respectively. The yeast two-hybrid assay was performed using the Yeastmaker Yeast Transformation System 2, in accordance with the manufacturer's instructions (Clontech, Palo Alto, CA, USA). The yeast two-hybrid assay was performed three times. All the primers used for generating these constructs are listed in Table S2.

LCI Assays
Agrobacterium tumefaciens GV3101, carrying the indicated plasmids, was prepared in infiltration medium (MS, 10 mM MES, and 150 µM acetosyringone) at OD600 = 0.5. The cultures were transiently infiltrated into N. benthamiana leaves, and the LUC luminescence intensity was detected using a ROPER CA2048B (ROPER Scientific). The LCI assays were performed with three biological replicates.

GUS Staining
GUS staining was carried out using a published protocol [28] with minor modifications. Seedlings were incubated with GUS staining solution (100 mM sodium phosphate buffer, pH 7.0, 10 mM ethylenediaminetetraacetic acid disodium salt, pH 8.0, 1 mM potassium ferrocyanide, 1 mM potassium ferricyanide, 0.5% (v/v) Triton X-100, 20% (v/v) methanol, and 0.5 mg/mL X-Gluc) under vacuum conditions and then stained at 37 • C for 6 h. The stained tissues were cleared of chlorophyll using an ethanol series, placed in a clear solution (chloral hydrate/water/glycerol = 8 g/3 mL/1 mL), and observed under a light microscope.

RNA-seq Analyses
Total RNA was extracted from the aerial part of seedlings with an RNAprep Pure Plant Kit (TIANGEN, Beijing, China) and analyzed by a NanoDrop and an Agilent 2100 bioanalyzer (Thermo Fisher Scientific, MA, USA). Oligo (dT) magnetic beads were used to purify mRNA, and then, the mRNA was sheared into small pieces in a fragmentation buffer. The first-strand cDNA was synthesized using random hexamer-primed reverse transcription, and then, second-strand cDNA was synthesized. Afterward, adapters were added to the double-stranded cDNA, and the resulting cDNA fragments were amplified by PCR. The PCR products were purified and dissolved in EB solution. The PCR products from the previous step were heat-denatured and circularized to produce the final library. The final library was loaded and the reads were generated on the BGIseq500 platform (BGI-Shenzhen, China). The transcriptome datasets have been submitted to the NCBI (accession number PRJNA695037). The sequencing data were filtered and analyzed in accordance with a previous article [29]. The significance levels of genes and GO-term enrichment analyses were set at Q-value ≤ 0.05.

Expression Analysis
In total, 1 µg total RNA was reverse transcribed using a FastKing gDNA Dispelling RT SuperMix kit (TIANGEN, Beijing, China). Quantitative real-time PCR was performed on the CFX96 real-time PCR detection system (Bio-Rad) using the UltraSYBR Mixture (CWBio, Beijing, China). The expression of TUBULIN 2 (TUB2) was used as an internal control. The expression analysis was performed with three biological replicates. All of the primers used for the expression analysis are listed in Table S2.

Flowering Time Measurement
Flowering time was calculated and represented as the average number of rosette leaves when the first flower appeared.

BiFC Analysis
The cDNAs of the proteins tested were cloned into pDOE-03 vector containing N-and C-terminal yellow fluorescence protein mVenus fragments for BiFC analysis as previously published [33].

Statistical Analysis
The experimental data were analyzed using two-tailed paired Student's t tests with SPSS 12.0 software.