Molecular Characterization and Expression Profile of PaCOL1, a CONSTANS-like Gene in Phalaenopsis Orchid

CONSTANS (CO) and CONSTANS-like (COL) genes play important roles in coalescing signals from photoperiod and temperature pathways. However, the mechanism of CO and COLs involved in regulating the developmental stage transition and photoperiod/temperature senescing remains unclear. In this study, we identified a COL ortholog gene from the Taiwan native orchid Phalaenopsis aphrodite. The Phalaenopsis aphrodite CONSTANS-like 1 (PaCOL1) belongs to the B-box protein family and functions in the nucleus and cytosol. Expression profile analysis of Phalaenopsis aphrodite revealed that PaCOL1 was significantly expressed in leaves, but its accumulation was repressed during environmental temperature shifts. We found a differential profile for PaCOL1 accumulation, with peak accumulation at late afternoon and at the middle of the night. Arabidopsis with PaCOL1 overexpression showed earlier flowering under short-day (SD) conditions (8 h/23 °C light and 16 h/23 °C dark) but similar flowering time under long-day (LD) conditions (16 h/23 °C light and 8 h/23 °C dark). Transcriptome sequencing revealed several genes upregulated in PaCOL1-overexpressing Arabidopsis plants that were previously involved in flowering regulation of the photoperiod pathway. Yeast two-hybrid (Y2H) analysis and bimolecular fluorescence complementation (BiFC) analysis revealed that PaCOL1 could interact with a crucial clock-associated regulator, AtCCA1, and a flowering repressor, AtFLC. Furthermore, expressing PaCOL1 in cca1.lhy partially reversed the mutant flowering time under photoperiod treatment, which confirms the role of PaCOL1 function in the rhythmic associated factors for modulating flowering.


Introduction
Throughout their life span, plants are constantly exposed to changing environmental conditions. The sessile plants have to develop multitudinous strategies to survive the environmental challenges. The intimations of transition from vegetative to reproductive growth are determinant in the life cycle of plants. To ensure the inception of flowering at the correct time, plants have to coalesce a diverse range of external and internal signals [1]. The photoperiod oscillator, ambient temperature, gibberellic acid, vernalization, and autonomous/aging pathways are the main factors involved in the transition of plants from the vegetative to reproductive stage [2][3][4][5].
The use of genetic approaches and functional analyses has identified several key moderators in the complex signal regulatory network in Arabidopsis thaliana; most of these moderators belong to the regulatory protein family [6,7]. Among the regulatory protein families, B-box (BBX) proteins are a group of transcription factors with one or more B-box domains that can be stabilized by binding to zinc aphrodite (mRNA/CDS GeneBank accession no: FJ810423/ACY95395.1; promoter GeneBank accession no: HM068557; Figure S1A). PaCOL1 was the first member of the BBX family identified in the native P. aphrodite (diploid). Biochemical and molecular analyses revealed that PaCOL1, a nuclear localized protein, could interact with specific transcription factors involved in sensing day length and temperature-alteration signals. Ectopic overexpression analyses revealed that PaCOL1 could function in mediating flowering time in Arabidopsis plants.

PaCOL1 Encodes a COL Protein
The genomic DNA (gDNA) of PaCOL1 is 945 bp, which splices into a 744-bp sense RNA product and encodes a protein with 247 amino acids of 27.9 kDa ( Figure S1B). From the deduced amino acid sequence, we identified one B-box zinc finger domain near the N-terminal region and one CCT motif at the C-terminal region ( Figure S1C,D). Also, the CCT motif included a putative nuclear localization sequence (NLS). Multiple sequence alignment was carried out with the Clustal Omega program with default parameters in the European Bioinformatics Institute toolbox (https://www.ebi.ac.uk/Tools/msa/clustalo/) ( Figure 1A). The results revealed PaCOL1 with 69-78% and 60-81% similarity to Arabidopsis COL proteins and other plant COL proteins, respectively ( Figure 1B). To investigate the evolutionary relationship among plant COLs, we constructed a phylogenetic tree by using the Mega6 software (http://megasoftware.net/) and unweighted pair-group method with arithmetic means (UPGMA) method. Phylogenetic tree analysis with other COL/BBX proteins from Arabidopsis, rape, sesame, rice, corn, grass, barley, darnel, and orchid revealed the sequence association as three major clades in the evolution ( Figure 1C). The members of the PaCOL1 clade share a B-box domain and a CCT motif, which suggests that PaCOL1 shared the same sequence structures as the group 3 B-box protein family in Arabidopsis and rice.
Plants 2020, 9, x FOR PEER REVIEW 3 of 20 no: HM068557; Figure S1A). PaCOL1 was the first member of the BBX family identified in the native P. aphrodite (diploid). Biochemical and molecular analyses revealed that PaCOL1, a nuclear localized protein, could interact with specific transcription factors involved in sensing day length and temperature-alteration signals. Ectopic overexpression analyses revealed that PaCOL1 could function in mediating flowering time in Arabidopsis plants.

PaCOL1 Encodes a COL Protein
The genomic DNA (gDNA) of PaCOL1 is 945 bp, which splices into a 744-bp sense RNA product and encodes a protein with 247 amino acids of 27.9 kDa ( Figure S1B). From the deduced amino acid sequence, we identified one B-box zinc finger domain near the N-terminal region and one CCT motif at the C-terminal region ( Figure S1C, D). Also, the CCT motif included a putative nuclear localization sequence (NLS). Multiple sequence alignment was carried out with the Clustal Omega program with default parameters in the European Bioinformatics Institute toolbox (https://www.ebi.ac.uk/Tools/msa/clustalo/) ( Figure 1A). The results revealed PaCOL1 with 69-78% and 60-81% similarity to Arabidopsis COL proteins and other plant COL proteins, respectively ( Figure  1B). To investigate the evolutionary relationship among plant COLs, we constructed a phylogenetic tree by using the Mega6 software (http://megasoftware.net/) and unweighted pair-group method with arithmetic means (UPGMA) method. Phylogenetic tree analysis with other COL/BBX proteins from Arabidopsis, rape, sesame, rice, corn, grass, barley, darnel, and orchid revealed the sequence association as three major clades in the evolution ( Figure 1C). The members of the PaCOL1 clade share a B-box domain and a CCT motif, which suggests that PaCOL1 shared the same sequence structures as the group 3 B-box protein family in Arabidopsis and rice.

Subcellular Analysis of PaCOL1 Protein
To determine the subcellular localization of PaCOL1, the corresponding coding sequence was genetically fused with green fluorescence protein (GFP) in transient expression vectors. Recombinant constructs were expressed in onion epidermal cells by using particle bombardment and transient expression assay. The vector containing eIF4AIII [42] was used as a marker to indicate the nucleus region or even nuclear body. The full length of PaCOL1 protein localized in the nucleus (Figure 2). In addition, we analyzed the cellular localization of PaCOL1 protein without a B-box domain (deletion of amino acid residues 6-52; PaCOL1∆B) or CCT motif (deletion of amino acid residues 177-221; PaCOL1∆C). PaCOL1∆B was still localized in the nucleus, but PaCOL1∆C localized in the cytoplasm and nucleus. We further confirmed the cellular localization of PaCOL1, PaCOL1∆B, and PaCOL1∆C in protoplasts from P. aphrodite seedling root cells ( Figure S2). Thus, PaCOL1 may function in the nucleus.

Subcellular Analysis of PaCOL1 Protein
To determine the subcellular localization of PaCOL1, the corresponding coding sequence was genetically fused with green fluorescence protein (GFP) in transient expression vectors. Recombinant constructs were expressed in onion epidermal cells by using particle bombardment and transient expression assay. The vector containing eIF4AIII [42] was used as a marker to indicate the nucleus region or even nuclear body. The full length of PaCOL1 protein localized in the nucleus ( Figure 2). In addition, we analyzed the cellular localization of PaCOL1 protein without a B-box domain (deletion of amino acid residues 6-52; PaCOL1ΔB) or CCT motif (deletion of amino acid residues 177-221; PaCOL1ΔC). PaCOL1ΔB was still localized in the nucleus, but PaCOL1ΔC localized in the cytoplasm and nucleus. We further confirmed the cellular localization of PaCOL1, PaCOL1ΔB, and PaCOL1ΔC in protoplasts from P. aphrodite seedling root cells ( Figure S2). Thus, PaCOL1 may function in the nucleus.

Expression Profiles of PaCOL1
To investigate the function of PaCOL1 in P. aphrodite, we tested expression profiles in vegetative and reproductive tissues. Under normal growth conditions (12 h/23 ℃ light and 12 h/23 ℃ dark, 58.39 W/m 2 ), PaCOL1 transcripts were significantly expressed in buds, floral stems, and leaves but were less accumulated in flowers and roots ( Figure 3A). Leaves are important for sensing light and temperatures. In Arabidopsis, AtBBX1/CO was synthesized in leaf cells, followed induced florigen production, and translocated to shoot apical for floral induction [14,22]. The orchid market often uses low ambient temperatures (i.e., 12 h/23 ℃ light and 12 h/18 ℃ dark for four weeks) for floral bud induction [43]. Thus, we examined the expression levels of PaCOL1 of mature P. aphrodite plants with floral bud induction or high temperature treatment (12 h/32 ℃ light and 12 h/23 ℃ dark). We found that PaCOL1 transcript level in leaves was significantly decreased after floral bud induction ( Figure  3BII) but recovered to a high level when the plants returned to normal growth conditions (Figure

Expression Profiles of PaCOL1
To investigate the function of PaCOL1 in P. aphrodite, we tested expression profiles in vegetative and reproductive tissues. Under normal growth conditions (12 h/23 • C light and 12 h/23 • C dark, 58.39 W/m 2 ), PaCOL1 transcripts were significantly expressed in buds, floral stems, and leaves but were less accumulated in flowers and roots ( Figure 3A). Leaves are important for sensing light and temperatures. In Arabidopsis, AtBBX1/CO was synthesized in leaf cells, followed induced florigen production, and translocated to shoot apical for floral induction [14,22]. The orchid market often uses low ambient temperatures (i.e., 12 h/23 • C light and 12 h/18 • C dark for four weeks) for floral bud induction [43]. Thus, we examined the expression levels of PaCOL1 of mature P. aphrodite plants with floral bud induction or high temperature treatment (12 h/32 • C light and 12 h/23 • C dark). We found that PaCOL1 transcript level in leaves was significantly decreased after floral bud induction ( Figure 3BII) but recovered to a high level when the plants returned to normal growth conditions ( Figure 3BIII). In contrast, PaCOL1 level in leaves did not increase when plants were transferred to high Plants 2020, 9, 68 6 of 20 temperature treatment followed by floral bud induction treatment ( Figure 3BIV). PaCOL1 transcript levels did not differ in roots and stems of P. aphrodite plants during treatments. 3BIII). In contrast, PaCOL1 level in leaves did not increase when plants were transferred to high temperature treatment followed by floral bud induction treatment ( Figure 3BIV). PaCOL1 transcript levels did not differ in roots and stems of P. aphrodite plants during treatments.
The expression profiles of COL genes from many plants are often differentially regulated by light and exhibit a diurnal rhythm [44][45][46]. Next, we investigated the daily expression pattern of PaCOL1 protein. On immunostaining with PaCOL1 antibody ( Figure S3), PaCOL1 protein (~28 kDa) was clearly detected from late afternoon to the middle of the night ( Figure 3C). These results revealed that PaCOL1 protein differentially accumulated in P. aphrodite under 12 h light/12 h dark conditions.  The expression profiles of COL genes from many plants are often differentially regulated by light and exhibit a diurnal rhythm [44][45][46]. Next, we investigated the daily expression pattern of PaCOL1 protein. On immunostaining with PaCOL1 antibody ( Figure S3), PaCOL1 protein (~28 kDa) was clearly detected from late afternoon to the middle of the night ( Figure 3C). These results revealed that PaCOL1 protein differentially accumulated in P. aphrodite under 12 h light/12 h dark conditions.

Overexpression of PaCOL1 in Arabidopsis Conferred SD Floral Inductivity
To further investigate the roles of PaCOL1, we developed transgenic A. thaliana overexpressing PaCOL1 in ecotype Col-0 plants. We used T3 homozygous transgenic lines COL1-ox, COL1∆B-ox, and COL1∆C-ox constitutively expressing PaCOL1, PaCOL1∆B, and PaCOL1∆C, respectively, for further study. We measured and compared plant height, leaf morphology, and flowering time between wild type (WT) and transgenic plants.  Figure S4). In addition, COL1-ox and COL1∆B-ox plants showed an early-flowering phenotype (18% and 24% decrease in number of rosette leaves, respectively) as compared with WT plants, whereas the flowering time of COL1∆C-ox plants was similar to that of the WT ( Figure 4A and Figure S4).
To further evaluate the abundance of target mRNAs mediated by PaCOL1 and unravel the molecular mechanism regulated by this gene, we analyzed transcriptome of shoot apex from 40-day-old SD-grown transgenic plants with the Illumina HiSeq 4000 platform (Novogene, Hong Kong). After filtering, we obtained more than 42,000,000 clean reads ( Figure 4B). Relative fold change was calculated by comparing the RNA sequencing (RNAseq) data for COL1-ox1 and WT plants grown under SD conditions, with relative change of ≥2-fold considered significantly different. We identified 1370 upregulated and 1269 downregulated genes in COL1-ox1 versus WT plants (Table S1). On Gene Ontology (GO) analysis (http://geneontology.org/), the genes upregulated in COL1-ox1 were classified into 22 groups ( Figure 4C). For annotated genes (n = 234) involved in the developmental process, we found 48 (3% in upregulated dataset) corresponding to the flower-development-associated process ( Figure 4D). In addition, we found 78 upregulated genes including MYBs, MADS-box proteins, bZIPs, ATPases, helicases, ubiquitination enzymes (UEs), and zinc-finger proteins as putative flowering-regulation candidate genes [47][48][49][50]. Thus, PaCOL1 may affect transcript levels of flowering-related Arabidopsis genes under SD conditions. Furthermore, we used qRT-PCR to investigate the expression of AtCCA1 (At2G46830), AtCO, AtSVP (At2G22540), AtSOC1 (At2G45660), AtFLC (At5G10140), AtAGL24 (At4G24540), and AtTSF (At4g20370), which are important floral modulator genes but were not found in RNAseq data. Under SD conditions, the expression of AtCO was higher in 40-day-old COL1-ox1, COL1∆B-ox1, and COL1∆C-ox1 plants than in the WT, whereas the expression of AtCCA1, AtFLC, AtSOC1, AtSVP, and AtAGL24 genes was nearly equal in WT and transgenic plants ( Figure 4E). In addition, we found the expression of CDF1 (At5G62430), a repressor of CO, was decreased in COL1-ox1 transgenic plants but the accumulation of MAF5 (At5G65080), an FLC-like gene, was slightly increased in COL1-ox1 transgenic plants. A similar expression level of TFL1 (At5G03840), an antagonist of FT, was found in WT and transgenic plants. We hardly detected the accumulation levels of AtTSF transcripts in this assay (data not shown). We also found no difference in accumulation of NFYB2 (At5g47640) and NFYC3 (At1g54830) transcripts, the positive regulators for flowering under LD conditions. Of note, GI (At1G22770), involved in regulation of circadian rhythm and photoperiodic flowering, was shown to upregulate in the RNAseq dataset but to downregulate in qRT-PCR analysis. Our results suggest that ectopic expression of PaCOL1 in Arabidopsis may disturb the circadian output gene expressions.

PaCOL1 Can Interact with AtCCA1 and AtFLC
The interaction between PaCOL1 and other proteins was predicted by using the STRING v 11 online program [51]. The results suggest a possible link between PaCOL1 and AtCCA1/AtLFY, BBXs, or transcriptional factors in clock control in Arabidopsis ( Figure 5A). We thus used yeast two-hybrid (Y2H) analysis to examine the protein-protein physiological interaction between PaCOL1 and AtSOC1, AtSVP, AtFLC, and AtCCA1. Hybrid yeast cells containing PaCOL1 with AtCCA1 and AtFLC could survive on triple amino acid dropout selection medium ( Figure 5B and Figure S5). In addition, two truncated PaCOL1 proteins, PaCOL1ΔB and PaCOL1ΔC, could interact with AtFLC and AtCCA1, which was confirmed by using the quanta dropout medium ( Figure 5C). We did not find an interaction between PaCOL1 and AtSOC1 or AtSVP ( Figure 5B).
Furthermore, we used bimolecular fluorescence complementation (BiFC) analysis to examine the PaCOL1-AtCCA1 and PaCOL1-AtFLC interactions in onion epidermal cells and orchid protoplast cells. PaCOL1, PaCOL1ΔB, and PaCOL1ΔC interacted with AtFLC ( Figure 5D). In addition, we detected AtCCA1 interacting with PaCOL1, PaCOL1ΔB, and PaCOL1ΔC ( Figure 5E). We observed the yellow fluorescence protein (YFP) fluorescent signal for all combinations tested in the nucleus. We did not detect any YFP fluorescence in the combination of PaCOL1 with AtSVP and AtSOC1 (data not shown). No interactions of PaCOL1 or AtFLC with vector only were observed ( Figure 5F). In addition, for co-transfection of the PaCOL1 full length fused with each half of the YFP gene, we observed the YFP florescent signal in orchid protoplasts, but there were no signals in YFP C -PaCOL1∆B/YFP N -PaCOL1∆B, YFP C -PaCOL1∆C/YFP N -PaCOL1∆C, or YFP C /PaCOL1-YFP N (Figure

PaCOL1 Can Interact with AtCCA1 and AtFLC
The interaction between PaCOL1 and other proteins was predicted by using the STRING v 11 online program [51]. The results suggest a possible link between PaCOL1 and AtCCA1/AtLFY, BBXs, or transcriptional factors in clock control in Arabidopsis ( Figure 5A). We thus used yeast two-hybrid (Y2H) analysis to examine the protein-protein physiological interaction between PaCOL1 and AtSOC1, AtSVP, AtFLC, and AtCCA1. Hybrid yeast cells containing PaCOL1 with AtCCA1 and AtFLC could survive on triple amino acid dropout selection medium ( Figure 5B and Figure S5). In addition, two truncated PaCOL1 proteins, PaCOL1∆B and PaCOL1∆C, could interact with AtFLC and AtCCA1, which was confirmed by using the quanta dropout medium ( Figure 5C). We did not find an interaction between PaCOL1 and AtSOC1 or AtSVP ( Figure 5B). Furthermore, we used bimolecular fluorescence complementation (BiFC) analysis to examine the PaCOL1-AtCCA1 and PaCOL1-AtFLC interactions in onion epidermal cells and orchid protoplast cells. PaCOL1, PaCOL1∆B, and PaCOL1∆C interacted with AtFLC ( Figure 5D). In addition, we detected AtCCA1 interacting with PaCOL1, PaCOL1∆B, and PaCOL1∆C ( Figure 5E). We observed the yellow fluorescence protein (YFP) fluorescent signal for all combinations tested in the nucleus. We did not detect any YFP fluorescence in the combination of PaCOL1 with AtSVP and AtSOC1 (data not shown). No interactions of PaCOL1 or AtFLC with vector only were observed ( Figure 5F). In addition, for co-transfection of the PaCOL1 full length fused with each half of the YFP gene, we observed the YFP florescent signal in orchid protoplasts, but there were no signals in YFP C -PaCOL1∆B/YFP N -PaCOL1∆B, YFP C -PaCOL1∆C/YFP N -PaCOL1∆C, or YFP C /PaCOL1-YFP N ( Figure S6), which suggests that PaCOL1 protein may form dimers for their physiological functions. Thus our Y2H and BiFC results suggest specific domain(s) other than the B-box domain and CCT motif for the interaction of PaCOL1 with AtCCA1 and AtFLC.

Discussion
Flowering is a seasonal response that involves distinct signaling pathways. BBX proteins have positive or negative effects on the complex regulation of flowering [31][32][33]. Under SD conditions,

Discussion
Flowering is a seasonal response that involves distinct signaling pathways. BBX proteins have positive or negative effects on the complex regulation of flowering [31][32][33]. Under SD conditions, overexpression of OsBBX27/OsCO3 in rice had opposing roles in the expression of FT homologs (Hd3a and FT-like) and resulted in a late-flowering phenotype [55]. Expression of SiCOL1, SiCOL2, CO homologs of sesame, enhanced FT expression in transgenic Arabidopsis but could repress the expression of FT homologous genes in sesame under LD conditions [56]. With knowledge of the conserved sequence of the B-box domain and CCT motif, we screened an orchid COL protein from P. aphrodite, PaCOL1, which contains a B-box domain and a CCT motif ( Figure 1A). Recently, PaCOL1 was reported in the Orchidstra database open source (http://orchidstra2.abrc.sinica.edu.tw/orchidstra2/index.php) as the CONSTANS-like 1 gene (accession no. PATC130360) [57]. Phylogenetic tree analysis of BBXs indicated that PaCOL1 and some group-3 COL proteins of rice and Arabidopsis are in the same clade ( Figure 1C), but have little physiological research for group-3 BBX proteins. Our subcellular analysis agreed that the CCT motif is important for PaCOL1 localization to the nucleus (Figure 2 and Figure S2). CCT-initiated nuclear localization has been found necessary for the BBX24 protein function in light signaling [58]. From our research, PaCOL1 may function as an integrator for flowering by modulating the input signals of the diurnal rhythmic and stress-induced pathways in P. aphrodite. Also, our results suggest that PaCOL1 may be involved in the plant development ( Figure 6). However, these require further investigation.
The circadian clock is an approximately 24 h biological oscillator, which generally enables organisms to adjust their physiological activities corresponding to the external light/dark cycles by expecting daily environmental changes. Expression profile analysis of PaCOL1 revealed suppressed PaCOL1 accumulation in leaves during the temperature shift ( Figure 3B). On immunoblotting analysis, PaCOL1 protein accumulation showed a differential profile, with a peak during late afternoon and at the middle of the night in P. aphrodite ( Figure 3C).
Vernalization is a key role for many plants to adapt to a temperature cue for flowering and germinating under suitable conditions [59,60]. VRN2 and its orthologs, with high conserved sequences during evolution, are often repressors of flowering in monocot cereals under LD conditions [61]. In this study, continuous expression of PaCOL1 in Arabidopsis plants caused early flowering under noninductive SD conditions. However, WT and transgenic plants with a long period of cold temperature during seed imbibition ( Figure S8A) or seedling stage ( Figure S8B) showed similar flowering time under SD conditions. Therefore, PaCOL1 may not participate directly in the cold-induced flowering regulatory pathway. From our RNAseq results, we found that the expression of PaCOL1 could upregulate (≥2-fold increase) genes for the flowering regulatory network in Arabidopsis under SD conditions (Table S1). Among these genes, ATPases, helicases, DEAD-box proteins, WDRs, SETs, and PHD proteins are involved in chromatin-associated transcriptional regulation [62][63][64]; MADS-box proteins, MYBs, AP2s, C2H2 zinc finger proteins, bZIPs, ABI3s, bHLHs, UEs, hydrolases, and IQ domain proteins (IQDs) are important for the transition from the vegetative to reproductive stage [38,48,65,66]. Expression of an IQD gene in Arabidopsis may stimulate floral transition by activating AtCO [67]. Our real-time PCR results revealed an increased AtCO transcript level in transgenic Arabidopsis overexpressing PaCOL1 variants under SD conditions ( Figure 4E). The expression of circadian clock output gene GI or CDF also interfered. Thus, PaCOL1 may function in the photoperiodic flowering control pathway in Arabidopsis.
The circadian network is conserved in higher plants [68]. Plants adapt to diurnal changes with regulation via a number of negative and positive feedback loops. CCA1, a key clock-associated modulator in Arabidopsis, functions with LHY in a negative feedback loop. Light transiently induces CCA1 and LHY transcription, with the peak at dawn [69,70]. CCA1 and LHY proteins, MYB transcription factors, repress the expression of the evening-expressed gene Timing of Cab expression 1 (TOC1) by directly binding to the TOC1 promoter. Synergistic functioning of CCA1 and LHY to TOC1 forms a negative feedback loop in the plant circadian regulation network [71,72]. In the early morning, CCA1 and LHY transiently increased Pseudo-response regulator 7 (PRR7) and PRR9 expression [73]; however, the PRR5, PRR7, and PRR9 protein complex repressed CCA1 and LHY transcription, which resulted in another negative feedback loop to each other [74,75].
Using Y2H and BiFC analysis, we identified that PaCOL1 and its truncated forms were able to interact with AtCCA1 and AtFLC in the nucleus ( Figure 5). In addition, our complementary test revealed that expression of PaCOL1 in the cca1.lhy mutant partially rescued the flowering time of the mutants under LD and SD conditions ( Figure 6). However, PaCOL1, CCA1, and LHY show much divergence in amino acid sequence. Using the NetPhos 3.1 server (http://www.cbs.dtu.dk/services/NetPhos/) to predict the eukaryotic protein phosphorylation sites, we identified a PaCOL1 16-amino acid region with a serine phosphorylation site showing similarity with a conserved domain for phosphorylation activity of AtCCA1 [76]. Also, we found similar regions in AtBBX1, OsHd1, and HvCO1 ( Figure S9). Phosphorylation of the conserved amino acids in CCA1 and LHY was necessary for their stability to control flowering time [27,76]. However, we need further study to identify the role of the 16-amino acid region. In addition, AtCCA1 is a master regulator of reactive oxygen species (ROS) homeostasis [77], so PaCOL1 may also be involved in the ROS response induced by light length or temperature shifting.
Floral transition is an important stage for plants to overcome environmental challenges to survive. FT-related genes are crucial for accelerated flowering under LD conditions [78,79] and are important to accelerate flowering in some plants under SD or stress conditions [38,47,80]. Thus, PaCOL1 may play a key role in the photo-rhythmical flowering pathways in P. aphrodite. Of note, deletion of the CCT motif or B-box domain did not lose the interaction between PaCOL1 and AtCCA1 and AtFLC ( Figure 5), whereas expression of PaCOL1∆C could not affect the flowering time of Arabidopsis plants under SD conditions ( Figures 4A and 6). Hence, PaCOL1 might interact with these two flowering modulators via other specific domain(s) although CCT motif and B-box domain are well known for protein-protein interaction in BBX proteins. Identification of the specific region needs further study.

Plant Materials
P. aphrodite (diploid) was grown in the culture room under a 12 h light/12 h dark (23 ± 1 • C) cycle. The plants at five-leaf stage (mature stage) were used for bolting induction. The bolting induction plants were cultured in an incubator under 23 • C/18 • C (day/night) oscillation with about 100 µmolm −2 s −1 light for four weeks. After bolting induction, the plants were grown in a 12 h light/12 h dark (23 ± 1 • C) culture room. The leaves for PCR amplification/protein extraction were collected from independent plants with at least three biological replicates for each sampling and analysis.
A. thaliana ecotype Col-0 was used as the wild type (WT). The cca1.lhy mutant in the Wassilewskija (Ws) background was obtained from The Arabidopsis Information Resource (TAIR) (https://www. arabidopsis.org/). Arabidopsis plants were grown at 23 • C in a 16 h light/8-h dark (LD) or 8 h light/16 h dark (SD) cycle in a growth chamber with 100 µmolm −2 s −1 light and 60% relative humidity. Arabidopsis seeds were sterilized with 1.2% bleach and sown on half-strength Murashige and Skoog (1/2 MS) medium containing 1% sucrose (w/v), 0.8% agar (w/v), buffered to pH 5.7 with 25 µg/mL hygromycin for homozygous transgenic line selection. Flowering time was determined as the number of rosette leaves when the bolt was at 1 cm length.

RT-PCR and qRT-PCR
Transcript levels in different tissues were analyzed by RT-PCR. Total RNA was extracted as described [81]. The first-strand cDNA was synthesized from 2 µg total RNA in a 20 µL reaction volume with use of the SuperScriptIII First-Strand Synthesis system(Invitrogen, Thermo Scientific, Waltham, MA, USA). RNA and cDNA were quantified by using a spectrophotometer (NanoDrop 1000, Thermo Scientific, Waltham, MA, USA). PCR amplification conditions were 27 cycles of 30 s at 95 • C, 30 s at 55 • C, and 50 s at 68 • C, then 5 min at 68 • C. Primers used for analyzing gene expression levels were designed with the use of Primer3 (http://bioinfo.ut.ee/primer3-0.4.0/) and are shown in Table S2. DNA from 15 µL of each PCR reaction underwent 1.2% (w/v) agarose gel electrophoresis. ImageJ (http://rsbweb.nih.gov/ij/) was used to quantify the ethidium-bromide-stained DNA bands. The expression of Phalaenopsis actin9 (PaACT9) and Arabidopsis β-tubulin (AtTUB) was the internal control.
qRT-PCR involved using the iCycler-iQ5 Multicolor Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) and KAPA SYBR FAST qPCR Master Mix (2x) Kit (KAPA Biosystem, Wilmington, MA, U.S.A.) with the primer sequences listed in Table S2. qRT-PCR reaction involved 20 ng cDNA with each set of primers and the KAPA SYBR FAST qPCR buffer Master Mix. PCR cycling included an initial step at 95 • C for 5 min, then 40 cycles of 10 s at 95 • C, 30 s at 55 • C, and 20 s at 68 • C. The comparative C T method was used to measure the relative amount of each sample. The expression of PaACT9 was the internal control.

Preparation of DNA Constructs and Transformation
The DNA fragment of CaMV35S-Nos with cloning sites for XbaI/BamHI/KpnI was cut from pBI121 shuttle vector and subcloned into pCAMBIA1301. Then, the XbaI/BamHI fragments of PaCOL1, PaCOL1∆B, and PaCOL1∆C were separately cloned into the vector in the sense orientation downstream of the CaMV35S promoter to generate the overexpression vectors pCOL1-ox, pCOL1∆B-ox, and pCOL1∆C-ox. Sequences for all primers are in Table S2. All constructs were verified by sequencing analysis. The recombinant vectors were transferred to Agrobacteria (GV3101) and used to transform Arabidopsis by the floral-dip technique [82]. Transformed Arabidopsis seedlings were selected as described [81]. Homozygous T2 transgenic plants were selected by growth of the T3 generations on 1/2 MS medium containing 50 mg/l hygromycin, 1% sucrose (w/v), and 0.8% agar. T3 transgenic seeds derived from homozygous T2 transgenic plants were used for subsequent tests.

Subcellular Localization Assays
Fragments of PaCOL1, PaCOL1∆B, and PaCOL1∆C were PCR-amplified with the primers listed in Table S2. PCR products were individually fused upstream of the GFP at the XbaI/BamHI sites in the p326GFP vector (35S::GFP) [83], which resulted in constructs under control of the CaMV35S promoter. The plasmids were transiently expressed in onion epidermal cells by particle bombardment as described [81] and further confirmed in orchid protoplast cells by PEG transfection [84]. A helium biolistic particle-delivery system (model PDS-1000, Bio-Rad, Hercules, CA, USA) was used for particle bombardment. In total, 5 µg recombinant plasmid DNA was used per transformation, and all target materials were bombarded twice. The bombarded onion epidermal cells were recovered at 25 • C in the dark for at least four hours. The samples were observed by using an Olympus U-LH 100 HG inverted fluorescence microscope.

Yeast Two-Hybrid (Y2H) Analysis
For Y2H assay, the bait DNA fragments PaCOL1, PaCOL1∆B, and PaCOL1∆C were PCR-amplified and cloned into the pGBKT7 expression vector (Clontech, Mountain View, CA, USA). The prey DNA fragments AtFLC, AtSVP, AtSOC1, AtCCA1, PaAPX, and PaFT were PCR-amplified and cloned into the pGADT7 expression vector (Clontech, Mountain View, CA, USA). All constructs were verified by sequencing analysis. The bait and prey recombinant vectors were transformed into yeast strain Y2HGold and Y187, respectively. Yeast mating and selection were performed according to the manufacturer's protocols.

Transcriptome Sequencing
Total RNA was extracted from the shoot apex of 40-day-old Arabidopsis plants grown under SD conditions with sampling at ZT7 as described [81]. The RNA sequencing library was constructed by using the TruSeq PE Cluster kit (Illumina, Santiago, CA, USA) according to the manufacturer's instructions. Transcriptome sequencing and assembly involved an Illumina HiSeq 4000 platform (Novogene, Hong Kong, China).

Electrophoresis and Immunoblotting Analysis
The orchid protein extraction buffer contained 2% CTAB, 2% PVP40, 100 mM Tris (pH 8.0), 25 mM EDTA (pH 8.0), 68.8 mM spermidine, 1 mM PMSF, and 1 mM DTT, 1D SDS-PAGE involved using 12.5% (w/v) polyacrylamide gels as described [85]. To generate an antibody against PaCOL1, a DNA construct encoding the 126th to 247th amino acids of PaCOL1 was prepared by PCR and subcloned into the BamHI site of the pMAL-c5X expression vector (New England Biolabs, Ipswich, MA, USA) to produce maltose binding protein (MBP) fusion proteins. The MBP fusion proteins were purified according to the manufacturer's protocols and used to raise a rabbit polyclonal antiserum by LTK BioLaboratories, Taoyuan, Taiwan. The immunoblotting was performed with anti-PaCOL1 antibodies as described [85]. The proteins were quantified by using Bradford assay [86].

Statistical Analysis
Data are shown as mean ± SE from three independent experiments. Statistical differences were analyzed by Duncan's multiple range test or Student's t test. P < 0.05 was considered statistically significant.

Conclusions
We identified a BBX protein, PaCOL1, as an important regulator for flowering in Arabidopsis transgenic plants. PaCOL1 interacted and functioned with CCA1 and LHY in the diurnal rhythmic pathway for flowering. Thus, PaCOL1 plays an important role in integrating signals from photoperiod and stress pathways for flowering in Arabidopsis. PaCOL1 may be a good candidate for understanding the signaling pathways for flowering in orchids.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2223-7747/9/1/68/s1, Figure S1: Gene structure of PaCOL1, Figure S2: Subcellular localization of PaCOL1 and two truncated mutant proteins in orchid protoplasts visualized by fluorescence microscopy, Figure S3: Fusion protein purification and identification for antibody production, Figure S4: The phenotypes of PaCOL1 transgenic plants and the WT, Figure  S5: The negative control tests of AtSOC1 and AtFLC with PaCOL1 were grown on double dropout (-Trp -Leu) medium, Figure S6: Interactions of PaCOL1, PaCOL1∆B, and PaCOL1∆C in orchid petal protoplasts visualized by fluorescence microscopy, Figure S7: The complementary tests of PaCOL1, PaCOL1∆B, and PaCOL1∆C overexpression in mutant lines under LD condition, Figure S8: Cold treatment of transgenic Arabidopsis plants at seed imbibition and seedling stage, Figure S9: Phosphorylation site prediction of PaCOL1 and AtCCA1, OsHd1, HvCO1, and AtBBX1, Table S1: Genes showing significant differential expression between the wild type (WT) and COL-ox1 Arabidopsis. Table S2: Oligonucleotides used in this study.