To screen the PIF proteins in pepper, all capsicum proteins were acquired from three public databases: PepperHub server,¹ Solanaceae Genome Database,² and NCBI.³ Then, the amino acid sequences of AtPIFs were obtained from TAIR,⁴ and used as queries to search for the homologous sequences from the capsicum proteome. The candidate pepper PIFs were submitted to SMART⁵ and InterPro⁶ for checking the presence of conserved bHLH and APB domains.

The molecular weight (MW) and isoelectric point (pI) of the candidate PIFs were calculated with ExPasy.⁷ In order to study the evolutionary relationships of the PIFs between pepper and other plant species, the protein sequences of PIFs from pepper, Arabidopsis, tomato, rice, and maize were aligned by Clustal Omega, and the alignment results were used to create an NJ (neighbor-joining) phylogenetic tree with the MEGA7.0 software using the pairwise deletion option, Poisson correction and 1,000 bootstrap replicates. The online MEME tool was used to search the conserved motifs of the pepper, Arabidopsis, tomato, rice, and maize, and the results were visualized by TBtools (Chen et al., 2020). The online tool Gene Structure Display Server v2.0⁸ was employed to analyze the gene structures of PIFs from different plant species.

The full-length coding region without the stop codon (1,416 bp) of CaPIF8 (XM_016689836.1) was cloned into the Super1300-GFP vector between the Xba I and Kpn I sites. The pSuper1300:CaPIF8:GFP vector was transformed into Agrobacterium tumefaciens strain GV3101. The pSuper1300:CaPIF8:GFP and pSuper1300:GFP (as control) were grown overnight, re-suspended in the induction medium (10 mM MgCl₂, 10 mM MES, and 200 μmol L^–1 acetosyringone), and injected into Nicotiana benthamiana. GFP fluorescence was photographed using a confocal laser-scanning microscope (Nikon C2-ER, Tokyo, Japan).

To obtain the expression profiles of CaPIF genes in various tissues including bud, flower, leaf, root and stem, transcriptomic data of pepper cultivar “Zunla” were obtained from publicly available GEO database under the accession numbers of GSE45037 and GSE45154 (Qin et al., 2014). The genome-wide transcriptome data of an elite breeding pepper (Capsicum annuum) line 6,421 treated with 200 mM NaCl and 10°C were obtained under PepperHub server (see text footnote 1) (Liu et al., 2017). The gene expression levels were measured following the TopHat/Cufflinks pipeline based on the FPKM (fragments per kilobase of exon per million fragments mapped) values according to the previous reports (Haq et al., 2019; Wu et al., 2020). The heatmaps of the gene log2 transformed values were generated with TBtools software.

Hangjiao12 (Capsicum annuum L.) seedlings were planted in a greenhouse under 25/18°C (day/night) temperature with a 12 h photoperiod. For cold stress, TRV2:CaPIF8 and TRV2:00 seedlings at the four-leaf stage were placed in 4°C incubator. RNA samples were collected from leaves after 6 h of cold treatment. For salt stress, 300 mM NaCl (90 mL per plant) was used to irrigate the seedlings. RNA samples were collected from the leaves after 12 h of salt treatment. Harvested samples were rapidly frozen by using liquid nitrogen and stored at –80°C.

Tobacco rattle virus (TRV) based VIGS was used to generate CaPIF8-silenced pepper plants. Partial cDNA fragment of CaPIF8 (300 bp) was amplified with the VIGS tool.⁹ The PCR-amplified fragment was cloned into the pTRV2 vector by using the restriction sites of EcoR I and Xho I. The empty vector of TRV2:00 served as the control, and the TRV2:CaPDS (phytoene desaturase gene) vector was set as positive control due to the photobleaching phenotype, and above TRV2 and TRV1 were transformed into Agrobacterium tumefaciens strain GV3101. Two-week-old pepper seedlings were used for silencing the PIF8 gene according to the previously described method (Yang et al., 2015).

Total RNA extraction was conducted with the total RNA Miniprep Kit (Axygen Biosciences, Union City, CA, United States). Approximately 5 μg of RNA was reverse transcribed with the ReverTra Ace qPCR-RT Kit (Toyobo, Japan) according to the manufacturer’s protocol. qRT-PCR was carried out in triplicate using the iCycler iQ^TM Real-time PCR Detection System (Bio-Rad, Hercules, CA, United States) with the procedures as previously described (Yang et al., 2019). The pepper actin gene was used as an internal control. The primers of actin and CaPIF genes are listed in Supplementary Table 1.

To assess membrane permeability, electrolyte leakage assays were performed as described previously by using a conductivity meter (ST3100C, Changzhou, China). Relative electrolyte leakage (REL) was calculated based on the previously described method (Haq et al., 2019; Zhang H. et al., 2020). To determine the ABA content, 0.25 g of frozen leaves was ground into powder in liquid nitrogen. ABA was extracted in 1.5 mL of 80% aqueous methanol overnight at 4°C. After centrifugation, the supernatant was removed and extracted twice with 2 mL petroleum ether, the pH was adjusted to 2–3 with citric acid, and 2 mL ethyl acetate was used to extract the supernatant twice. The upper organic phase was transferred to a new EP tube, blown dry under the flow of N₂ (gas). The dried samples were dissolved and mixed with 0.2 mL of mobile phase, and filtered through a needle filter. ABA content was analyzed using a high-performance liquid chromatography (Waters 2695, Milford, MA, United States).

Significant differences (P < 0.05) were tested by Duncan’s test in the statistical product and service solution (SPSS) software between the data from the control and treatments, and different letters were used to indicate differences.

In total, six PIF genes were identified from the pepper genome, which were designated as CaPIF1, CaPIF3, CaPIF4, CaPIF7a, CaPIF7b, and CaPIF8. Table 1 presents the physiological and biochemical properties of these six CaPIF genes. The gene lengths of CaPIFs varied from 1,950 to 11,887 bp, and the amino acid sequence lengths were from 342 to 633 aa. The predicted pI of various proteins ranged from 5.55 to 7.31 (Table 1).

To analyze the residue conservation features in these CaPIFs, the full-length CaPIFs and AtPIFs were aligned by Clustal Omega. As a result, the bHLH and APB domains were found in all six CaPIF proteins, with the exception of CaPIF7b, which contained an incomplete Helix2 in the bHLH domain (Table 1 and Figure 1). In addition, CaPIF1 and CaPIF3 contained an additional APA domain, which was also observed in PIF proteins from other plants, such as tomato SlPIF1a, SlPIF1b, and SlPIF3 (Rosado et al., 2016), Arabidopsis AtPIF1 and AtPIF3 (Lee and Choi, 2017), and apple MdPIF1–MdPIF5 (Zheng et al., 2020).

To elucidate the evolution of PIFs in pepper and other plant species, a NJ phylogenetic tree was created on the basis of multiple sequence alignment of full-length PIF protein sequences from pepper, Arabidopsis (Lee and Choi, 2017), tomato (Rosado et al., 2016), rice (Nakamura et al., 2007), and maize (Wu et al., 2019). In the phylogenetic tree, these PIF proteins could be divided into three distinct groups, namely PIF1/4/5, PIF2/3/6, and PIF7/8, and CaPIFs displayed a closer relationship with tomato SlPIFs and Arabidopsis AtPIFs than with rice OsPIFs and maize ZmPIFs (Figure 2A).

Conserved motif analysis was also conducted with the MEME tool, resulting in the identification of 10 conserved motifs (Figure 2B). Among these motifs, motif 1, and motif 2 were annotated as the bHLH and APB domains, which are present in all PIF proteins. In addition, motifs 3, 5, and 7 were also widely distributed in these PIF proteins. However, several motifs were unique to specific groups. For example, motif 9 was identified specifically in PIF4/5 members. Motif 8 was only observed in ZmPIF3.3 and all PIF7/8 members, and ZmPIF3.3 was lack of motif 10, which was widely found in PIF7/8 group (Figure 2B). In addition, PIF7 proteins were lack of motif 4, which was widely present in PIF proteins of other groups, except for OsPIL14, AtPIF2, and AtPIF6 (Figure 2B).

We also analyzed the structure of the PIF genes from pepper and other plant species. As a result, most of the PIF genes possessed 4–6 introns, especially those from dicots. The number of introns in CaPIF genes ranged from five to seven, while that of PIF genes in other plants varied more greatly, ranging from one (SlPIF1a) to nine (ZmPIF5.1) (Figure 3). In addition, some closely related PIF genes exhibited similar gene structures in intron number and CDS length, such as ZmPIF4.2 and OsPIL12, SlPIF8a and CaPIF8, SlPIF7a and CaPIF7a (Figure 3).

To elucidate the possible functions of CaPIF genes, their tissue expression profiles were analyzed with publicly available transcriptomic data. As a result, CaPIF1, CaPIF3, CaPIF4, CaPIF7a, and CaPIF8 were constitutively expressed in all the tissues tested, while CaPIF7b was only expressed in leaves and stems (Figure 4). Notably, all CaPIF genes had high expression in the leaves and stems, but there were some differences among CaPIF genes. For example, the expression of CaPIF8 was much higher than that of other PIF genes in leaves and stems. Only one gene, CaPIF7a, showed a higher expression level in the leaves than in the stems, and other genes exhibited no obvious difference in expression between the two tissues (Figure 4).

To investigate the possible functions of CaPIFs under abiotic stresses, we examined their expression profiles under cold and salt stress based on public RNA-seq data. Under cold stress treatment, CaPIF4 and CaPIF8 displayed up-regulated expression at certain time points (3 h and/or 6 h), but the expression decreased at 24 h (Figure 5A). The expression of CaPIF7a and CaPIF7b was down-regulated at the later time points (24 h) (Figure 5A). Under salt stress treatment, CaPIF4, CaPIF7a, and CaPIF7b were down-regulated at 12 h, while CaPIF8 was up-regulated at 3 h (Figure 5B). Since CaPIF8 showed noticeably high expression under cold and salt stress, and its homolog in citrus (CsPIF8) was found to enhance the tolerance of citrus to cold stress (He et al., 2020), we then chose CaPIF8 for further research in cold and salt stress.

For subcellular localization analysis of CaPIF8, the pSuper1300:CaPIF8:GFP and pSuper1300:GFP (as control) vectors were infiltrated into Nicotiana benthamiana epidermal cells by transient Agrobacterium-mediated transformation. Confocal laser scanning microscopy revealed that the GFP fluorescence of CaPIF8-GFP fusion protein was exclusively located in the nucleus, while that of 35S::GFP control was ubiquitously distributed in the whole cells (Figure 6), indicating that CaPIF8 is localized to the nucleus.

Since the expression of CaPIF8 markedly increased at specific time points under cold and salt treatments, it can be speculated that CaPIF8 plays certain roles in stress response. Hence, we silenced CaPIF8 in pepper plants by VIGS to study its possible role in cold and salt stress tolerance. After approximately 1 month of virus injection, the leaves of TRV2:CaPDS plants were photo-bleached, while those of plants infected with TRV2:00 and TRV2:CaPIF8 showed no obvious symptoms (Figure 7A). However, the TRV2:CaPIF8 plants had significantly lower height than the TRV2:00 plants (Figure 7B). The qRT-PCR results showed that the CaPIF8 transcription in TRV2:CaPIF8 plants was reduced by 39% compared with that in TRV2:00 plants (Figure 7C), suggesting the successful silencing of CaPIF8.

To investigate the response of TRV2:CaPIF8 plants to cold stress, the TRV2:00 and TRV2:CaPIF8 plants were subjected to 4°C cold treatment for 3 days. After the treatment, the TRV2:CaPIF8 plants displayed severe growth retardation compared with the TRV2:00 plants, including serious wilting of leaves (Figure 7D). We also measured the REL and ABA level in leaves of the TRV2:00 and TRV2:CaPIF8 plants. Before cold stress treatment, no difference in REL was found between TRV2:00 and TRV2:CaPIF8 plants, while the TRV2:CaPIF8 plants exhibited markedly higher REL than the TRV2:00 plants after cold stress (Figure 7E). The ABA content in the TRV2:CaPIF8 plants was significantly lower than that in the TRV2:00 plants before and after cold stress treatment, though cold stress treatment led to increases in the ABA content of both plants (Figure 7F).

To elucidate the possible function of CaPIF8 in cold stress response, the expression profiles of five stress-related genes were determined by qRT-PCR. As a result, cold treatment did not significantly change the expression of ABA responsive genes (CaABF1 and CaABI5) in both TRV2:00 and TRV2:CaPIF8 plants (Figure 7G). However, the TRV2:CaPIF8 plants showed observably lower expression levels of CaRBOHB, CaCBF1, and ABA biosynthesis gene (CaNCED1) than the TRV2:00 plants with or without cold treatment (Figure 7G).

Additionally, we tested the response of TRV2:CaPIF8 plants to salt stress. After salt stress treatment for 7 days, the TRV2:CaPIF8 plants exhibited more serious symptoms than TRV2:00 plants, including leaf shrinkage, yellowing, and even wilting (Figure 8A). Also, no significant difference was observed in REL and ABA content between TRV2:00 and TRV2:CaPIF8 plants without salt stress. Upon salt stress treatment, the TRV2:CaPIF8 plants exhibited remarkably higher REL (Figure 8B) and a nearly 50% decrease in ABA content (Figure 8C) relative to the TRV2:00 plants. As for the stress-related genes, CaABI5 and CaNHX2 showed no significant changes in expression with and without salt treatment in both TRV2:00 and TRV2:CaPIF8 plants (Figure 8D). Salt treatment increased the expression of CaRBOHA, CaCAT2, and ABA biosynthesis genes (CaNCED1, CaNCED3, and CaNCED6) in both TRV2:CaPIF8 and TRV2:00 plants, and the expression was much lower in TRV2:CaPIF8 plants than in TRV2:00 plants (Figure 8D).

In this work, six PIF genes were identified from the pepper genome, and the number was comparable to that in many other plants, such as P. patens (six genes) (Possart et al., 2017), O. sativa (six genes) (Nakamura et al., 2007), Z. mays (seven genes) (Zhang et al., 2018; Wu et al., 2019), M. domestica (seven genes) (Zheng et al., 2020), A. thaliana (eight genes) (Lee and Choi, 2017), and S. lycopersicum (eight genes) (Rosado et al., 2016). Amino acid sequence alignment revealed that all the six CaPIF proteins harbor a conserved bHLH and an APB domain (Figures 1A,B), which is a typical feature of PIF proteins. However, CaPIF7b has an incomplete Helix2 in its bHLH domain, which may affect its function to interact with other proteins. Notably, an amino acid substitution (Q to E) in the APB domain was observed in CaPIF4 (Figure 1B), which was also identified in MdPIF8 (Zheng et al., 2020), suggesting that they possibly have similar functions. Similar to tomato SlPIF1s and SlPIF3 (Rosado et al., 2016), maize ZmPIF3s (Wu et al., 2019), Arabidopsis AtPIF1 and AtPIF3 (Leivar and Quail, 2011), CaPIF1 and CaPIF3 also contain the conserved functional APA domain (Figure 1C), which was found to play a vital role in modulating the interaction with phyA.

In the phylogenetic tree, PIF proteins from different plants (pepper, Arabidopsis, tomato, rice, and maize) could be divided as three distinct groups, and members of the same group have similar conserved proteins motif arrangements (Figure 2). PIF proteins did not cluster in a species-specific manner; however, CaPIFs had a much closer phylogenetic relationship with PIFs from dicots than with those from monocots (Figure 2A). Additionally, the PIF1/4/5 and PIF2/3/6 groups comprise PIF proteins from both dicots and monocots, while the PIF7/8 group exclusively includes PIF proteins from dicots (Figure 2A), indicating the evolution of the PIF genes after the monocotyledon–dicotyledon differentiation, and the PIF genes in the PIF7/8 group might have been derived from those in the PIF1/4/5 and PIF2/3/6 groups. Compared with Arabidopsis, pepper is lack of PIF2, PIF5, and PIF6 homologs (Table 1 and Figure 2A), and the intron numbers of CaPIF genes are relatively stable compared with those of PIFs in other plants (Figure 3). It can be speculated that pepper PIF genes might have undergone relatively small variations and gene loss during evolution.

Besides their crucial functions in photomorphogenesis, PIFs were also found to either positively or negatively regulate the response to different environmental stresses. For instance, both AtPIF1 and AtPIF4 were upregulated under heat and cold stress conditions; AtPIF6 was upregulated by ABA and salt stress; while AtPIF7 was suppressed by cold stress. AtPIF1, AtPIF4, and AtPIF5 were destabilized by cold stress depending on phytochromes, and they negatively regulated freezing tolerance in Arabidopsis (Jiang et al., 2020). AtPIF4 functions as negative regulator of salt tolerance by directly regulating the expression of genes associated with salt stress response, including ORE1, SAG29, and JUNGBRUNNEN1 (JUB1)/ANAC042 (Sakuraba et al., 2017). ZmPIF3.1 was significantly induced by drought and salt stress, and rice overexpressing maize ZmPIF3.1 showed enhanced tolerance to drought and salt stress (Gao et al., 2015). OsPIL14 overexpression or loss-of-function of the DELLA protein SLENDER RICE1 (SLR1) could also enhance the salt tolerance of rice seedlings by integrating light and GA signals (Mo et al., 2020). In this study, CaPIF4, CaPIF7a, CaPIF7b, and CaPIF8 were differentially regulated by cold or salt stress, indicating their critical and different roles in regulating the response to the two stresses (Figure 5). Among these four genes, CaPIF8 was induced under both cold and salt stress, suggesting that this gene may positively regulate plant resistance to cold or salt stress. To verify this inference, we generated CaPIF8-silenced pepper plants by using the VIGS technique. As a result, the TRV2:CaPIF8 plants exhibited more serious symptoms under cold and salt stress with increased REL than the TRV2:00 plants (Figures 7, 8). Compared with the TRV2:00 plants, the TRV2:CaPIF8 plants exhibited observable decreases in the expression of CaRBOHB under cold treatment, and CaRBOHA and CaCAT2 under salt treatment, respectively (Figures 7G, 8D), suggesting that reactive oxygen species (ROS) homeostasis was affected in TRV2:CaPIF8 plants under cold or salt stress. As the citrus homolog of CaPIF8 and AtPIF8, CsPIF8 can enhance the tolerance of citrus to cold stress through the regulation of SOD activity and the expression of CsSOD by binding to the E-box in its promoter (He et al., 2020). Therefore, CaPIF8 may participate in the regulation of the ROS homeostasis under cold and salt stress.

Various adverse stresses can result in ABA accumulation in plants, which can regulate stress-related genes involved in plant adaptation to environmental stimuli (Vishwakarma et al., 2017; Zhou et al., 2018). Some PIF genes were reported to participate in plant response to multiple stresses by activating or inhibiting downstream genes in the ABA-dependent signaling pathway. For example, overexpression of either ZmPIF3.1 or ZmPIF3.2 in rice conferred ABA-dependent drought tolerance (Gao et al., 2015, 2018). Myrothamnus flabellifolia MfPIF1 also positively regulates the response to drought and salt stress by activating the expression of both ABA biosynthesis and ABA-responsive genes (Qiu et al., 2020). In this study, the TRV2:CaPIF8 plants showed lower tolerance to salt stress with a significant decrease in ABA content (Figure 8). qRT-PCR analysis revealed that compared with those in the TRV2:00 plants, the ABA biosynthesis genes (CaNCED1, CaNCED3, and CaNCED6) were down-regulated but the ABA responsive gene showed no obvious change in TRV2:CaPIF8 plants (Figure 8D). Hence, we speculate that CaPIF8 has a crucial function in salt stress tolerance by activating ABA biosynthesis genes and then enhancing the ABA level under salt stress.

The TRV2:CaPIF8 plants showed higher sensitivity to cold stress, but the expression of ABA responsive genes (CaABF1 and CaABI5) had no obvious change when compared with those in the control. However, CaCBF1 was observably down-regulated in TRV2:CaPIF8 plants with and without cold stress (Figure 7G). The C-repeat binding factor (CBF) genes were found to enhance low temperature tolerance of plants. A previous study has suggested that AtPIF4 and AtPIF7 can reduce freezing tolerance through direct repression of the CBF genes under long day conditions (Lee and Thomashow, 2012). Arabidopsis CBF1 can interact with PIF3 but not with PIF1, PIF4, and PIF5, and stabilize PIF3 and phyB protein under cold stress (Jiang et al., 2020). In tomato, SlPIF4 can positively regulate low R/FR light-induced cold tolerance by inducing the CBF genes (Wang F. et al., 2020). Therefore, the decrease in cold stress tolerance of the TRV2:CaPIF8 plants may be mainly attributed to the significant decrease in the expression of the CBF genes.

In this study, six PIF members were identified in pepper, which were then classified into three distinct groups together with PIFs from other plants. Subsequently, their conserved domains, gene structures and expression profiles were analyzed. Amongst the identified CaPIF genes, CaPIF8 was upregulated under cold and salt stress, and silencing of CaPIF8 reduced the tolerance of pepper to cold and salt stress, suggesting that CaPIF8 can positively regulate tolerance to cold and salt stress. CaPIF8 enhances cold stress tolerance probably by inducing the CBF genes, while mediates salt stress tolerance by activating the ABA biosynthesis genes. Therefore, it can be inferred that CaPIF8 plays differential roles in mediating cold and salt tolerance of pepper. Overall, these findings provide important information for further revealing the mechanisms through which CaPIF8 controls the cold and salt stress response of pepper.