F-Box Family Genes, LTSF1 and LTSF2, Regulate Low-Temperature Stress Tolerance in Pepper (Capsicum chinense)

The F-box proteins belong to a family of regulatory proteins that play key roles in the proteasomal degradation of other proteins. Plant F-box proteins are functionally diverse, and the precise roles of many such proteins in growth and development are not known. Previously, two low-temperature-sensitive F-box protein family genes (LTSF1 and LTSF2) were identified as candidates responsible for the sensitivity to low temperatures in the pepper (Capsicum chinense) cultivar ‘sy-2’. In the present study, we showed that the virus-induced gene silencing of these genes stunted plant growth and caused abnormal leaf development under low-temperature conditions, similar to what was observed in the low-temperature-sensitive ‘sy-2’ line. Protein–protein interaction analyses revealed that the LTSF1 and LTSF2 proteins interacted with S-phase kinase-associated protein 1 (SKP1), part of the Skp, Cullin, F-box-containing (SCF) complex that catalyzes the ubiquitination of proteins for degradation, suggesting a role for LTSF1 and LTSF2 in protein degradation. Furthermore, transgenic Nicotiana benthamiana plants overexpressing the pepper LTSF1 gene showed an increased tolerance to low-temperature stress and a higher expression of the genes encoding antioxidant enzymes. Taken together, these results suggest that the LTSF1 and LTSF2 F-box proteins are a functional component of the SCF complex and may positively regulate low-temperature stress tolerance by activating antioxidant-enzyme activities.


Introduction
In plants, low temperatures affect many physiological processes, including water and nutrient uptake, photosynthesis, immune responses, growth, and development, as well as the geographical distribution of plants [1,2]. Enhancing the low-temperature tolerance of temperature-sensitive crops is necessary to increase their productivity [3][4][5]. Pepper, an important fruit crop grown worldwide, is temperature-sensitive, with an optimum temperature of 25-30 • C [6,7]. Deviations from these temperatures can adversely affect growth and development, resulting in a variety of developmental and physiological disorders [6][7][8].
Many plant species overcome the adverse effects of low and non-freezing temperatures through complex adaptive mechanisms which are associated with the extensive reprogramming of the expression of a number of cold-responsive genes and subsequent changes in numerous physiological and biochemical processes, such as membrane stability, calcium ion fluxes, and changes in the properties of The LTSF1 and LTSF2 proteins share a 91% sequence similarity, with a single AA difference in their F-box domains. The protein sequence alignment of the F-box domains of LTSF1 and LTSF2 in pepper and other plant species, including Arabidopsis, rice, and tomato, revealed a high sequence conservation in the F-box domain (Figure 1). LTSF1 and LTSF2 showed a maximum sequence identity of 41-42% with the tomato F-box protein Solyc09g005480, and shared a sequence identity of 24-25% with the FBA motif-containing Arabidopsis At4g12560 protein (CPR1; CONSTITUTIVE EXPRESSER OF PR GENES 1); At3g06240 (a F-box/kelch-repeat protein); and rice Os02g54240, a stress-responsive F-box protein. A comparison of the genomic structures revealed that both the LTSF1 and LTSF2 genes contain two introns ( Figure 2A). The LTSF1 and LTSF2 proteins share a 91% sequence similarity, with a single AA difference in their F-box domains. The protein sequence alignment of the F-box domains of LTSF1 and LTSF2 in pepper and other plant species, including Arabidopsis, rice, and tomato, revealed a high sequence conservation in the F-box domain (Figure 1). LTSF1 and LTSF2 showed a maximum sequence identity of 41-42% with the tomato F-box protein Solyc09g005480, and shared a sequence identity of 24-25% with the FBA motif-containing Arabidopsis At4g12560 protein (CPR1; CONSTITUTIVE EXPRESSER OF PR GENES 1); At3g06240 (a F-box/kelch-repeat protein); and rice Os02g54240, a stress-responsive F-box protein. A comparison of the genomic structures revealed that both the LTSF1 and LTSF2 genes contain two introns ( Figure 2A).

VIGS Analysis
To test the functional roles of LTSF1 and LTSF2 in low-temperature sensitivity, we silenced their expression in wild-type pepper ('No.3341') using tobacco rattle virus (TRV)-mediated VIGS (Figure 2A,B). VIGS vectors were constructed targeting the UTR or coding (CDS) regions. To confirm the efficient silencing of the genes, a qPCR was performed in the silenced plants. No significant changes were observed in the expression levels of the LTSF1 and LTSF2 genes in plants infiltrated with constructs targeting either LTSF1 or LTSF2 gene UTR regions due to the unsuccessful gene silencing, possibly due to shorter VIGS target region of 135 bp and 127 bp, respectively for LTSF1 and LTSF2. However, in plants infiltrated with VIGS construct (LTSF1/2-double-silenced plants) targeting CDS regions of both LTSF1 and LTSF2 with 305 bp target length (Figure 2A), the expression levels of these genes were approximately four times lower compared with the mock plants ( Figure 2C). The double silencing of the LTSF1 and LTSF2 genes by targeting their CDS region resulted in an abnormal phenotype and severe growth retardation under low-temperature conditions (20 • C), consistent with the suppressed expression of LTSF1 and LTSF2 in these plants ( Figure 2B,C). The LTSF1/2-silenced plants produced small, irregularly shaped, pale, and thick leaves, similar to the phenotype observed in the 'sy-2' pepper mutant, which exhibits thick irregularly shaped leaves when grown under low-temperature

VIGS Analysis
To test the functional roles of LTSF1 and LTSF2 in low-temperature sensitivity, we silenced their expression in wild-type pepper ('No.3341') using tobacco rattle virus (TRV)-mediated VIGS ( Figure  2A,B). VIGS vectors were constructed targeting the UTR or coding (CDS) regions. To confirm the efficient silencing of the genes, a qPCR was performed in the silenced plants. No significant changes were observed in the expression levels of the LTSF1 and LTSF2 genes in plants infiltrated with constructs targeting either LTSF1 or LTSF2 gene UTR regions due to the unsuccessful gene silencing, possibly due to shorter VIGS target region of 135 bp and 127 bp, respectively for LTSF1 and LTSF2. However, in plants infiltrated with VIGS construct (LTSF1/2-double-silenced plants) targeting CDS regions of both LTSF1 and LTSF2 with 305 bp target length (Figure 2A), the expression levels of these genes were approximately four times lower compared with the mock plants ( Figure 2C). The double silencing of the LTSF1 and LTSF2 genes by targeting their CDS region resulted in an abnormal phenotype and severe growth retardation under low-temperature conditions (20 °C), consistent with the suppressed expression of LTSF1 and LTSF2 in these plants ( Figure 2B,C). The LTSF1/2-silenced plants produced small, irregularly shaped, pale, and thick leaves, similar to the phenotype observed

LTSF1 and LTSF2 Interaction with SKP1
LTSF1 and LTSF2 were predicted to encode F-box proteins containing a C-terminal FBA domain ( Figure 1). To determine whether the LTSF1 and LTSF2 proteins are functional components of the SCF complex, protein-protein interaction studies were carried out ( Figure 3A). SKP1 interacts with F-box proteins by binding to their conserved F-box domain [20]; therefore, we performed yeast two-hybrid (Y2H) assays to assess whether pepper SKP1 can interact with the Sy-2 candidates LTSF1 and LTSF2. No significant growth was detected in yeast cells carrying the control vectors on the selection medium (SD/-Leu/-Trp/-His/3-AT); however, the yeast cells carrying the pGAD-T7 vector containing the F-box domain sequence of LTSF1 or LTSF2 (LTSF1/2-F2) and pGBK-T7:SKP1 grew well ( Figure 3A), suggesting that LTSF1/LTSF2 can interact with SKP1. Consistent with these results, β-galactosidase (GUS) assay raveled a strong interaction between LTSF1/LTSF2 and SKP1 ( Figure 3B). The growth of the yeast cells carrying the pGAD-T7 vector with the full-length sequence of LTSF1 or LTSF2 (LTSF1/2-F1) and the pGBK-T7:SKP1 was inhibited on the SD plate, however ( Figure 3A). This weak or absent interaction Plants 2020, 9, 1186 5 of 15 between the full-length LTSF1/LTSF2 proteins and SKP1 may suggest that the C-terminal FBA domains may have an inhibitory effect in yeast cells. Alternatively, it is possible that the low effectivity of interaction could also be caused by different expression levels of the proteins.
(Y2H) assays to assess whether pepper SKP1 can interact with the Sy-2 candidates LTSF1 and LTSF2. No significant growth was detected in yeast cells carrying the control vectors on the selection medium (SD/-Leu/-Trp/-His/3-AT); however, the yeast cells carrying the pGAD-T7 vector containing the F-box domain sequence of LTSF1 or LTSF2 (LTSF1/2-F2) and pGBK-T7:SKP1 grew well ( Figure 3A), suggesting that LTSF1/LTSF2 can interact with SKP1. Consistent with these results, β-galactosidase (GUS) assay raveled a strong interaction between LTSF1/LTSF2 and SKP1 ( Figure 3B). The growth of the yeast cells carrying the pGAD-T7 vector with the full-length sequence of LTSF1 or LTSF2 (LTSF1/2-F1) and the pGBK-T7:SKP1 was inhibited on the SD plate, however ( Figure 3A). This weak or absent interaction between the full-length LTSF1/LTSF2 proteins and SKP1 may suggest that the C-terminal FBA domains may have an inhibitory effect in yeast cells. Alternatively, it is possible that the low effectivity of interaction could also be caused by different expression levels of the proteins.  To confirm the protein-protein interaction between LTSF1/LTSF2 and SKP1 in plant cells, a bimolecular fluorescence complementation (BiFC) assay was performed. LTSF1 and LTSF2 were fused with the C-terminus of YFP, and SKP1 was fused with the N-terminus of YFP. These constructs were then co-bombarded into onion epidermal cells. The negative control failed to produce a BiFC signal, while a strong fluorescent signal was observed throughout the cells transformed with the positive controls. A strong fluorescent signal was observed in the nuclei, cytoplasm, and cell membranes of epidermal cells co-bombarded with the YFP fusion proteins of LTSF1 and SKP1 ( Figure 3C), indicating a strong interaction between LTSF1 and SKP1, whereas a weak fluorescent signal was observed in the epidermal cells co-bombarded with the YFP fusion proteins of LTSF2 and SKP1, indicating a weak interaction between LTSF2 and SKP1 ( Figure 3C). Overall, these results indicate that both LTSF1 and LTSF2 interact with SKP1 in plant cells, and the detection of the BiFC signals in the nucleus, cytoplasm, and plasma membrane indicates that LTSF1 and LTSF2 may be involved in proteasome degradation in both the nucleus and the cytoplasm.

The Transgenic Expression of LTSF1 Promotes Low-Temperature Stress Tolerance in N. benthamiana Plants
To confirm the role of the pepper F-box gene LTSF1 in low-temperature stress tolerance, transgenic N. benthamiana lines overexpressing LTSF1 were developed ( Figure 4A,B). The genotypes of the T 0 transgenic LTSF1-expressing lines were confirmed using reverse-transcription PCR (RT-PCR) analysis ( Figure S2). Two transgenic lines (LTSF1-TG1 and -TG2) and their corresponding vector control (VC) plants were grown under normal (25 • C) and low (14 • C) temperatures for one month. No significant morphological changes were observed between the LTSF1-overexpressing N. benthamiana lines and the control plants under the normal conditions; however, the reduced plant growth and shorter heights induced in the control plants by the low-temperature conditions were less severe in the transgenic lines, which grew approximately 1.5 times taller than the control plants ( Figure 4B,C). Thus, the transgenic expression of LTSF1 alleviated the effect of low-temperature stress in the LTSF1-TG N. benthamiana plants.  To determine whether LTSF1 regulates low-temperature stress through the antioxidant pathway, we analyzed the transcript abundances of the genes encoding ROS-scavenging antioxidant enzymes, such as ascorbate peroxidase (APX), catalase (CAT), glutathione S-transferase (GST), and superoxide dismutase (SOD), in the control and transgenic plants subjected to low-temperature stress. We observed a significant increase in the accumulation of APX transcripts in the transgenic LTSF1-TG1 and -TG2 plants compared with the control plants both under the stressed and non-stressed conditions ( Figure 5). Similarly, the expression of GST and CAT was also significantly increased in the LTSF1-TG1 and -TG2 plants under the stress conditions. The accumulation of SOD transcripts was higher in the LTSF1-TG1 and -TG2 transgenic plants subjected to the stress conditions than in those grown in the non-stressed conditions. Overall, the expression levels of the genes encoding the ROS-scavenging enzymes were considerably altered by the low-temperature stress, and their levels were higher in the transgenic plants than in the control plants. These results indicate that the transgenic expression of To determine whether LTSF1 regulates low-temperature stress through the antioxidant pathway, we analyzed the transcript abundances of the genes encoding ROS-scavenging antioxidant enzymes, such as ascorbate peroxidase (APX), catalase (CAT), glutathione S-transferase (GST), and superoxide dismutase (SOD), in the control and transgenic plants subjected to low-temperature stress. We observed a significant increase in the accumulation of APX transcripts in the transgenic LTSF1-TG1 and -TG2 plants compared with the control plants both under the stressed and non-stressed conditions ( Figure 5). Similarly, the expression of GST and CAT was also significantly increased in the LTSF1-TG1 and -TG2 Plants 2020, 9, 1186 7 of 15 plants under the stress conditions. The accumulation of SOD transcripts was higher in the LTSF1-TG1 and -TG2 transgenic plants subjected to the stress conditions than in those grown in the non-stressed conditions. Overall, the expression levels of the genes encoding the ROS-scavenging enzymes were considerably altered by the low-temperature stress, and their levels were higher in the transgenic plants than in the control plants. These results indicate that the transgenic expression of LTSF1 may influence the antioxidant system and, thus, confers low-temperature tolerance.

Discussion
The SCF complex is an E3 ubiquitin ligase that mediates the ubiquitination of proteins destined for proteasomal degradation in several plant developmental processes, including hormone signaling, morphogenesis, embryo development, circadian rhythms, and senescence [26]. F-box proteins are adaptor subunits of the SCF complex; the N-terminal F-box domain and the SKP subunit protein physically interact to form a core complex and recruit target proteins for proteasomal degradation through the C-terminal protein-protein interaction domain of the F-box protein [28][29][30].

Discussion
The SCF complex is an E3 ubiquitin ligase that mediates the ubiquitination of proteins destined for proteasomal degradation in several plant developmental processes, including hormone signaling, morphogenesis, embryo development, circadian rhythms, and senescence [26]. F-box proteins are adaptor subunits of the SCF complex; the N-terminal F-box domain and the SKP subunit protein physically interact to form a core complex and recruit target proteins for proteasomal degradation through the C-terminal protein-protein interaction domain of the F-box protein [28][29][30].
In this study, the low-temperature stress-responsive LTSF1 and LTSF2 genes were predicted to encode F-box proteins containing a C-terminal FBA domain, which participates in protein-protein interactions associated with a wide range of plant physiological processes [31][32][33][34]. We cloned the F-box genes LTSF1 and LTSF2 from pepper and performed BiFC and Y2H assays to verify that LTSF1 and LTSF2 are functional F-box proteins. Protein-protein interaction studies revealed a weak interaction between LTSF2 and SKP1 compared with the LTSF1 and SKP1. The differences interaction between these proteins could be due to a single AA change (valine (V) in LTSF1 to alanine (A) in LTSF2) in their F-box domain (Figure 1).
To examine the subcellular localization of LTSF1 and LTSF2, we fused the LTSF1 and LTSF2 genes to the N-terminus of a green fluorescent protein (GFP) reporter gene and transformed the fusion protein expression vectors into N. benthamiana leaves for the transient expression of LTSF1/LTSF2:GFP fusion proteins. As shown in Figure S3, the green fluorescence signal was distributed throughout the cells infiltrated with the 35S::GFP control vector ( Figure S3). Similarly, the fluorescence signal of the 35S::LTSF1/2-GFP fusion was detected throughout the cells, including in the nucleus, cytoplasm, and plasma membrane ( Figure S3). In previous studies, SCF complex proteins, such as the tomato E3 ubiquitin ligase SlBAH1, the Arabidopsis SKP1-like protein13 (AtASK13), the wheat (Triticum aestivum) F-box protein TaFBA1, and the rice F-box protein OsMAIF1, were found to be involved in the tolerance of biotic and abiotic stresses, and were also documented to be localized to the nucleus, cytoplasm, and plasma membrane [35][36][37][38]. The dispersed subcellular localizations of LTSF1, LTSF2, and other SCF complex proteins may, therefore, contribute to their diverse roles in plant cells.
Various components of SCF complexes have been reported to be associated with plant growth, development, and the responses to biotic and abiotic stresses [19,[34][35][36]39,40] while a number of stress conditions have been documented to activate the expression of genes encoding the subunits of the SCF complex [28,41]. In the present study, to confirm the functional role of LTSF1 and LTSF2 genes in conveying low-temperature tolerance, we silenced the expression of these genes using VIGS. Silencing of the LTSF1 or LTSF2 gene by targeting their UTR region was not successful in our case. The double silencing of both LTSF1 and LTSF2 genes by targeting their CDS region resulted in an abnormal phenotype and severe growth retardation under low-temperature conditions (20 • C). Consistent with the suppressed expression of LTSF1/LTSF2, these double-silenced plants produced small, irregularly shaped, thick, and pale leaves (Figure 2), similar to those observed in the low-temperature-sensitive 'sy-2' cultivar grown under low temperatures [19]. Furthermore, VIGS of the LRR domain-containing Capsicum annuum F-box (CaF-box) gene reduced the cold-stress tolerance of pepper seedlings [42] supporting the role of the F-box genes in regulating the temperature sensitivity of plants.
To further confirm the role of LTSF1 and LTSF2 in low-temperature stress, we developed LTSF1 overexpressing transgenic N. benthamiana lines. We observed that the growth of the vector control plants in terms of their height was more severely inhibited by low-temperature stress than in the transgenic lines overexpressing LTSF1. The transgenic plants showed enhanced transcriptional activity for the genes encoding the antioxidant enzymes, including SOD, APX, CAT, and GST ( Figure 5), which could be associated with the improved temperature tolerance shown by the transgenic plants. Similar results were observed in transgenic wheat and tobacco plants overexpressing the TaFBA1 F-box gene, in which increased antioxidant-enzyme activity was correlated with increased abiotic stress [37,43]. In another study, the overexpression of SKP1 improved the resistance of Arabidopsis to various abiotic stresses, including heat, salinity, and oxidative stress [35], whereas the T-DNA insertion Plants 2020, 9, 1186 9 of 15 mutants and RNAi mutant lines lacking the related ASK13 gene showed a low tolerance to abiotic stress [35]. Our previous study revealed that the cell death and leaf deformation observed in the 'sy-2' mutant grown under low temperatures is caused by the excessive accumulation of ROS, and is correlated with the downregulation of the genes encoding ROS-scavenging enzymes at 20 • C [19]. In the present study, N. benthamiana plants overexpressing LTSF1 showed a markedly increased expression of ROS-scavenging enzyme-encoding genes, such APX, CAT, GST, and SOD, further substantiating the role of the ROS-scavenging pathway in low-temperature stress tolerance ( Figure 5). The F-box protein LTSF1, which is a key subunit of the SCF complex, may, therefore, play an important role in positively regulating low-temperature tolerance by modulating the stress-related genes, possibly through the repression of the negative regulators of cold stress.
A previous study revealed that the cell death and leaf deformation observed in the 'sy-2' mutant grown under low temperatures is caused by the excessive accumulation of ROS, and is correlated with the downregulation of the genes encoding ROS-scavenging enzymes, such as APX6, APXT1, and GST1, at 20 • C [6]. Furthermore, salt-tolerant transgenic tobacco plants overexpressing the TaFBA1 gene showed increased antioxidant-enzyme activity along with increased E3 ligase enzyme activity [44]. In the present study, N. benthamiana plants overexpressing LTSF1 showed a markedly increased expression of ROS-scavenging enzyme-encoding genes, such APX, CAT, GST, and SOD, further substantiating the role of the ROS-scavenging pathway in low-temperature stress tolerance ( Figure 5). However, further experiments will be needed to determine whether the increased transcriptional activity of antioxidant-enzyme genes in transgenic N. benthamiana plants overexpressing LTSF1 could be associated with increased E3 ligase activity.

Plant Materials
For the functional characterization of the two F-box candidate genes using VIGS, the pepper (C. chinense) 'No.3341' (wild-type) and 'sy-2' (mutant) cultivars [19] were used. For the functional studies of the F-box genes using heterologous expression, tobacco (N. benthamiana) was used. Plants were maintained under a 16/8 h light-dark regime at 24 ± 2 • C.

Phylogenetic Analysis and Identification of Conserved Motifs
Homologs of LTSF1 and LTSF2 (Table S1) were identified using a Basic Local Alignment Search tool (BLAST) search of other plant genomes in the National Center for Biotechnology Information (NCBI) database with default parameters using the LTSF1 and LTSF2 protein sequences as queries. To verify the conserved domains, the F-box homologous protein sequences were compared with known sequences using the ClustalW2 alignment tool with default parameters [45]. A multiple-sequence alignment of the LTSF1 and LTSF2 proteins and their homologs from other plant species, including Arabidopsis, tomato (Solanum lycopersicum), and rice, was created using the ClustalW2 program with default parameters [28]. The InterPro protein sequence analysis and classification tool [46] was used for the identification of the functional domains.

VIGS
Total RNA was extracted from the pepper ('No.3341') leaf samples using an MG RNAzol kit (MGmed, Seoul, Korea), according to the manufacturer's instructions. Complementary DNA (cDNA) was synthesized from 1 µg RNA using EasyScript Reverse Transcriptase kit (TransGen Biotech, Beijing, China). VIGS target sequences in the LTSF1 and LTSF2 genes were amplified from the cDNA sequences of pepper using Pfu DNA Taq polymerase (Bioneer, Seoul, Korea) and gene-specific primers ( Table 1). The target sequence for silencing of LTSF1 and LTSF2 genes were predicted by sol genomics VIGS tool [47] via BLAST against the Capsicum annuum v1.55 genome. Ligation-independent cloning (LIC) was used to clone the VIGS target sequences (LTSF1 (135 bp), LTSF2 (127 bp), and LTSF1/2 (305 bp)) into the TRV VIGS vector (TRV2), as previously described [48]. Plasmids from the positively transformed clones were isolated, purified, and sequenced at the National Instrumentation Center for Environmental Management (NICEM; Seoul National University, Seoul, Korea). The Agrobacterium tumefaciens strain GV3101 was transformed with the pTRV2 (containing the target sequences) and pTRV1 vectors, then used for the coinfiltration of 3-week-old pepper cotyledons using a needleless syringe, as previously described [48]. Control plants infiltrated with empty vector (TRV2:00) and TRV1 were used as mock. Three weeks after coinfiltration of TRV constructs, quantitative measurements including leaf area and plant height were recorded. The VIGS experiment was repeated two times with least three biological replicates in each trail. ImageJ software was used for the analysis of leaf area.

Subcellular Localization of the LTSF1 and LTSF2 Proteins
The LTSF1 coding sequence was amplified in a total volume of 50 µL using PrimeStar GXL DNA polymerase (Takara Bio, Kusatsu, Japan) with four-fold-diluted cDNA and 10 pmol of each gene-specific primer ( Table 1). The PCR cycling conditions were as follows: initial denaturation at 95 • C for 3 min; followed by 34 cycles of denaturation at 98 • C for 10 s, annealing at 60 • C for 15 s, and extension at 68 • C for 1 min; with a final extension at 68 • C for 5 min. The LTSF1 and LTSF2 coding sequences, without the stop codon, were independently inserted into the pMDC83 expression vector at the SpeI-AscI restriction sites using gene-specific primer pairs (Table 1), resulting in the fusion of green fluorescent protein (GFP) to the C-termini of LTSF1 and LTSF2. The recombinant GFP constructs were agro-infiltrated into N. benthamiana leaves. Cells harboring an empty pMDC83-GFP vector were used as a control. At 48 h after agro-infiltration, the GFP signals were detected using a fluorescence microscope (Axioskop 2; Carl Zeiss, Oberkochen, Germany).

Y2H Assays
The coding sequences of LTSF1, LTSF2, and SKP1 were cloned in frame into the EcoRI-BamHI sites in the pGADT7 and pGBKT7 vectors to generate preys and baits, respectively, using the primers listed in Table 1. The paired constructs were co-transformed into the yeast strain AH109 and grown on a -Leu/-Trp selection medium at 28 • C for 2-5 days. Positive yeast transformants were grown on SD/-Trp/-Leu/-His/3-AT selection medium for the interaction studies. The GUS assay was performed as described in the Yeast Protocols Handbook.

Bimolecular Fluorescence Complementation (BiFC) Analysis
BiFC analyses were conducted as described by Walter et al. [49]. For the BiFC vector constructs, SKP1, LTSF1, and LTSF2 cDNAs lacking the termination codons were amplified using gene-specific PCR primers (listed in Table 1) and cloned into the EcoRI-BamHI sites of the binary vectors pSPYNE and pSPYCE, which harbor sequences encoding the YFP N and YFP C (YFP N-and C-termini) protein fragments, respectively. The resulting constructs were named SKP1-YFP N , SKP1-YFP C , LTSF1-YFP N , LTSF1-YFP C , LTSF2-YFP N , and LTSF2-YFP C . The designated plasmid combinations were co-bombarded into onion (Allium cepa) epidermal cells using the Bio-Rad PDS 1000/He Biolistic Delivery System (Bio-Rad Laboratories, Hercules, CA, USA). On days 1 to 3 after transformation, the onion epidermal cells were visualized using an LSM5 Exciter confocal laser-scanning microscope (LSM 510; Carl Zeiss, Jena, Germany) at NICEM. YFP was excited using a 514 nm laser beam, and the emissions were detected at 525-600 nm. The BiFC experiment was repeated twice with six biological replicates in each experiment.

Genomic DNA Extraction
Genomic DNA was extracted from young N. benthamiana leaves using the cetyltrimethylammonium bromide method [48]. The concentration and quality of genomic DNA were analyzed using a Nanodrop spectrophotometer (BioTek, Winooski, VT, USA).

Agrobacterium-Mediated Transformation of N. benthamiana
The pMDC83-LTSF1:GFP vector constructed above was used for the Agrobacterium-mediated transformation of N. benthamiana, which was performed following the method described by Benvenuto et al. [50]. Putative transformants were confirmed using RT-PCR analyses with LTSF1 and Hpt (hygromycin phosphotransferase) gene sequence-specific primers (Table 1). Total RNA was extracted from young N. benthamiana leaf samples and cDNA was synthesized as described below. RT-PCR was performed in a total volume of 25 µL containing 4 µL diluted cDNA template, 2.5 µL 10 × PCR buffer, 2.0 µL 2.5 mM dNTP mix, 0.5 µL each of 10 µM forward and reverse primers, 0.25 µL homemade Taq polymerase, and 15.25 µL of water. The PCR cycling conditions were as follows: initial denaturation at 94 • C for 5 min; followed by 32 cycles of denaturation at 94 • C for 60 s, annealing at 55 • C for 60 s, and extension at 72 • C for 60-90 s; with a final extension at 72 • C for 5 min. The identified T 0 plants were micro-propagated using Murashige and Skoog [51] media to obtain multiple plants for the subsequent analysis.

Low-Temperature Stress Treatments
Approximately two-week-old transgenic and vector control (VC) N. benthamiana plants with a uniform appearance were grown in a walk-in growth chamber at 14 • C and 25 • C. Five plants per treatment were used in the stress experiments. After one month, the plant heights were recorded and leaf samples were collected for the qPCR analysis.

Total RNA Extraction and Reverse-Transcription Quantitative PCR (RT-qPCR)
To analyze the expression of LTSF1 and LTSF2, systemic leaves were collected from the VIGS pepper plants. The confirmation of the expression of the transgenes and antioxidant enzyme-encoding genes in the N. benthamiana plants were also performed using leaf tissues. Total RNA was extracted from these samples using an MG RNAzol kit (MGmed, Seoul, Korea), according to the manufacturer's instructions. The concentration and integrity of RNA were verified using 1% agarose gel electrophoresis and a Nanodrop spectrophotometer (BioTek, Winooski, VT, USA). A 1-µg aliquot of total RNA was used to produce cDNA using an EasyScript Reverse Transcriptase kit (TransGen Biotech, Beijing, China), according to the manufacturer's instructions. The resulting cDNAs were used for further expression studies.
The qPCR analysis was performed using the primers listed in Table S2. The expression levels of genes encoding the antioxidant enzymes were analyzed using a Light Cycler 2.0 instrument (Roche, Basel, Switzerland) using the following conditions: 95 • C for 4 min; followed by 42 cycles of 95 • C for 10 s, 58 • C for 20 s, and 72 • C for 20 s. Three biological replicates and three technical replicates were used for each gene. The relative expression levels were calculated by normalizing the expression levels to the endogenous reference gene Actin. The experiment was repeated twice.

Statistical Analysis
The statistical significance of the data was tested using Student's t-test, and the analyses were performed using Microsoft Excel 2016 (Microsoft, Redmond, WA, USA).

Conclusions
In conclusion, to elucidate the role of F-box genes in regulating the low-temperature sensitivity of pepper, we knocked down the expression of LTSF1 and LTSF2 genes using VIGS. The pepper plants lacking transcripts for both of these F-box genes exhibited an abnormal phenotype under low-temperature conditions, suggesting that they play an important function in plant adaptation to low-temperature stress. Furthermore, LTSF1 and LTSF2 function as part of the SCF complex, and could, therefore, modulate the degradation of the cellular regulatory proteins and enhance the plant response to low-temperature stress. Transgenic N. benthamiana plants overexpressing LTSF1 showed less growth inhibition and the enhanced expression of genes encoding ROS-scavenging enzymes. The further functional analysis of the LTSF1 and LTSF2 genes and the identification of their novel protein-protein interaction partners and regulatory sites will be essential for the elucidation of their precise functional consequences.