Systematic Investigation and Expression Profiles of the Nitrate Transporter 1/Peptide Transporter Family (NPF) in Tea Plant (Camellia sinensis)

NRT1/PTR FAMILY (NPF) genes are characterized as nitrate and peptide transporters that played important roles in various substrates transport in plants. However, little is known about the NPF gene in tea plants. Here, a total of 109 CsNPF members were identified from the tea plant genome, and divided into 8 groups according to their sequence characteristics and phylogenetic relationship. Gene structure and conserved motif analysis supported the evolutionary conservation of CsNPFs. Many hormone and stress response cis-acting elements and transcription factor binding sites were found in CsNPF promoters. Syntenic analysis suggested that multiple duplication types contributed to the expansion of NPF gene family in tea plants. Selection pressure analysis showed that CsNPF genes experienced strong purifying selective during the evolution process. The distribution of NPF family genes revealed that 8 NPF subfamilies were formed before the divergence of eudicots and monocots. Transcriptome analysis showed that CsNPFs were expressed differently in different tissues of the tea plant. The expression of 20 CsNPF genes at different nitrate concentrations was analyzed, and most of those genes responded to nitrate resupply. Subcellular localization showed that both CsNPF2.3 and CsNPF6.1 were localized in the plasma membrane, which was consistent with the characteristics of transmembrane proteins involved in NO3- transport. This study provides a theoretical basis for further investigating the evolution and function of NPF genes.


Introduction
Nitrogen (N) is an essential nutrient for plant growth and development. Nitrate is a major N source for higher plants and also serves as a signal to regulate plant development [1][2][3]. The NO 3 − uptake system consists of low-affinity transport system (LATS) and high-affinity transport systems (HATS), which are mainly mediated by nitrate transporter 1/peptide transporter (NRT1/PTR) and nitrate transporter 2 (NRT2), respectively [4,5]. NRT1 belongs to the large PTR family, in which family members are also described as proton-coupled oligopeptide transporter (POT) or solute carrier 15 (SLC15) [5][6][7]. A study involving 31 fully sequenced plant genomes unified these proteins under the name NPF (NRT1/PTR FAMILY) [7].
The first functionally characterized NPF protein, AtNPF6.3/AtNRT1.1/CHL1, was identified in Arabidopsis thaliana as a nitrate transporter [8]. Since then, its homologs have been cloned and functionally characterized in various plants [9,10]. The NPF proteins well known for their roles in nitrate transport, are considered to be the main components of LATS of high NO 3 − . However, some NPF proteins, such as AtNPF6.3 in A. thaliana and MtNRT1.3 in Medicago truncatula, were dual-affinity transporters participating in both HATS and LATS [11,12]. These NPF proteins have a wide absorption range for both high and low nitrate concentrations. ZmNPF6. 6 in Zea mays and MtNIP/LATD in M. truncatula have been shown to be high-affinity nitrate transporters [13,14]. To date, at least 21 Arabidopsis NPF members have been shown to transport nitrate, and many of them are also able to transport other substrates [15,16].
Tea plant, Camellia sinensis (L.) O. Kuntze, is an economically important crop and its leaves are the raw material for producing the non-alcoholic beverage "tea". Tea is rich in tea polyphenols, theanine and other active substances beneficial to human health [17,18].
To improve the quality of tea, a large amount of nitrogen fertilizer was applied to increase the content of free amino acids in tea leaves. The N input in tea garden is about 2-5 times that of other crops [19]. Therefore, it is particularly important to understand the molecular mechanism of nitrogen uptake in tea plants. To date, NPF family genes have been identified in several species, such as Arabidopsis, rice [20], poplar [21], apple [22], sugarcane [23], wheat [16] and spinach [24]. However, research on NPF family gene in tea plants is still limited. Due to its important function, it is necessary to make further in-depth investigations of NPF genes in tea plants.
In this study, a total of 109 CsNPF genes were identified in the tea plant genome. The phylogenic distribution, chromosomal location, gene structure, conserved motifs, duplication pattern, and selection pressure of NPF genes in tea plant were systematically analyzed. Furthermore, we performed expression profile analysis of NPF genes in different tissues of tea plant. The expression of CsNPF genes under different nitrate concentrations was detected by reverse transcription-quantitative real-time PCR (RT-qPCR). This study provides a basis for studying the evolution and function of NPF genes in tea plant.

Identification of Tea Plant CsNPFs
A total of 109 CsNPF protein sequences were identified from the tea plant genome (Tables S1 and S2). To better distinguish these genes, we named CsNPF genes as CsNPF1.X~CsNPF8.X according to the previously reported rules [7]. The lengths of the CsNPF proteins ranged from 72 (CsNPF5.17) to 957 (CsNPF2.5) amino acids, with molecular weight ranging from 7.8 kD to 105.81 kD and theoretical pI values ranging from 4.43 (CsNPF2.9) to 9.98 (CsNPF5.17). Subcellular localization prediction indicated that most CsNPFs were localized to the plasma membrane (Table S1).

Phylogenetic Analysis of CsNPFs
To investigate the evolutionary relationship of CsNPFs, multi-sequence alignment of the 53 AtNPFs, 85 PtNPFs, and 109 CsNPFs was conducted by using the clustalx1.83 software and the result was used for the construction of an unrooted phylogenetic tree. Notably, 36 CsNPFs were removed from the phylogenetic tree construction due to C-terminal or N-terminal deletions. According to the phylogenetic tree (Figure 1), the CsNPFs were divided into 8 subfamilies (NPF1-NPF8). Among the 8 subfamilies, the NPF5 subfamily (23) contained the most NPF genes, while the NPF6 subfamily (6) contained the least ( Figure 2 and Table S2). All members of the NPF3, NPF6, and NPF7 subfamilies contained complete domains, while 70% of the members of the NPF4 subfamily contained incomplete domains.

Distribution of NPF Members in Different Plant Species
To investigate the evolution of CsNPF in plant species, a comparison of NPF members in different plant species was conducted. Information on NPF family members in other species was extracted from previous studies [7,25]. A total of 36 plant species were analyzed, including 28 eudicots, 5 monocots, 1 other angiosperm (Amborella trichopoda), 1 spike moss (Selaginella moellendorffii), and 1 bryophyte (Physcomitrella patens). As shown in Figure 3, eight NPF subfamily members are present in almost all species, while NPF1 and

Gene Structure and Motif Analysis of CsNPFs
To characterize the protein sequences of the 73 CsNPF family members, 10 distinct motifs were predicted by the MEME program ( Figure 4A). About 71% of CsNPF proteins contained all 10 conserved motifs, while other proteins appeared to be missing one

Gene Structure and Motif Analysis of CsNPFs
To characterize the protein sequences of the 73 CsNPF family members, 10 distinct motifs were predicted by the MEME program ( Figure 4A). About 71% of CsNPF proteins contained all 10 conserved motifs, while other proteins appeared to be missing one or more of these motifs. Among them, all members of the NPF1 subfamily contain 10 motifs. Most closely related members share common motif compositions and arrangement orders. For example, motifs 1, 7, and 9 were lost in CsNPF2.5, and CsNPF2.6 of NPF2 subfamily. Motifs 3, 4 and 10 were absent in CsNPF5.9, CsNPF5.10 and CsNPF5.11 of NPF5 subfamily. or more of these motifs. Among them, all members of the NPF1 subfamily contain 10 motifs. Most closely related members share common motif compositions and arrangement orders. For example, motifs 1, 7, and 9 were lost in CsNPF2.5, and CsNPF2.6 of NPF2 subfamily. Motifs 3, 4 and 10 were absent in CsNPF5.9, CsNPF5.10 and CsNPF5.11 of NPF5 subfamily. Gene structure is an important parameter for gene evolution that further supports phylogenetic trees [26]. The coding sequences (CDS) of CsNPF genes were further analyzed to determine the structural diversity of these genes ( Figure 4B). The numbers of introns of CsNPF genes ranged from 1 to 11. Of the 73 CsNPF genes, 53 (73%) had 3 or 4 introns. Fourteen CsNPF genes contained more than 4 introns, and 6 of them belonged to the NPF3 subfamily. CsNPF2.5 and CsNPF2. 6 have the largest number of introns, with 10 and 11 introns respectively. However, both CsNPF8.14 and CsNPF8.15 contain only one intron. Overall, the most closely related CsNPF genes shared similar gene structures in terms of intron number and exon length. Gene structure is an important parameter for gene evolution that further supports phylogenetic trees [26]. The coding sequences (CDS) of CsNPF genes were further analyzed to determine the structural diversity of these genes ( Figure 4B). The numbers of introns of CsNPF genes ranged from 1 to 11. Of the 73 CsNPF genes, 53 (73%) had 3 or 4 introns. Fourteen CsNPF genes contained more than 4 introns, and 6 of them belonged to the NPF3 subfamily. CsNPF2.5 and CsNPF2.6 have the largest number of introns, with 10 and 11 introns respectively. However, both CsNPF8.14 and CsNPF8.15 contain only one intron.
Overall, the most closely related CsNPF genes shared similar gene structures in terms of intron number and exon length.

Chromosome Distribution of CsNPFs
The chromosomal location of each CsNPF gene was analyzed based on tea plant genomic annotation information. As shown in Figure 5 and Table S1, 95 out of the 109 CsNPF were mapped on the 15 chromosomes, while the remaining 14 genes were located on Contig. The distribution of CsNPF on each chromosome was extremely uneven. For example, Chr10 possesses the most CsNPF genes (18), followed by Chr11 (15 genes), while Chr7 and Chr13 contain only one CsNPF gene. Notably, NPF3 subfamily genes have an obvious preference for Chr10, with 9 out of 10 CsNPF located on Chr10.

Chromosome Distribution of CsNPFs
The chromosomal location of each CsNPF gene was analyzed based on tea plant genomic annotation information. As shown in Figure 5 and Table S1, 95 out of the 109 CsNPF were mapped on the 15 chromosomes, while the remaining 14 genes were located on Contig. The distribution of CsNPF on each chromosome was extremely uneven. For example, Chr10 possesses the most CsNPF genes (18), followed by Chr11 (15 genes), while Chr7 and Chr13 contain only one CsNPF gene. Notably, NPF3 subfamily genes have an obvious preference for Chr10, with 9 out of 10 CsNPF located on Chr10.

Expansion and Evolution of CsNPF Genes
To study the evolutionary processes of CsNPF family, the gene duplication pattern of CsNPF genes was analyzed ( Figure 6 and Table S1). Thirty-three of the 109 CsNPF genes were found to be segmentally duplicated, 31 of which were located on chromosomes 1, 2, 4, 5, 6, 9, 10, 11, 12, 13, 14, and 15, and 2 on Contig ( Figure 6A). There were 21 pairs of syntenic relationships in tea plant ( Figure 6B). A total of 13 tandem repeats containing 20 CsNPF genes were identified. In addition, 25 and 31 CsNPFs were also found to be dispersed and proximal duplicates, respectively, which may also contribute to the expansion of the CsNPF family genes.
To further study the selection pressure of tea plant NPF genes during evolution, the values of synonymous (Ks) and nonsynonymous (Ka) substitution rates, and Ka/Ks ratios for each duplicated pair from different subfamilies were calculated ( Figure 6C and Table S3). The mean values of Ka of gene pairs in NPF1-NPF8 subfamilies were 0.05, 0.12, 0.07, 0.12, 0.10, 0.08, 0.14, and 0.09, respectively. The mean values of Ks were 0.09, 0.27, 0.29, 0.34, 0.35, 0.39, 1.24, and 0.44, respectively. The mean values of Ka/Ks were 0.52, 0.59, 0.27, 0.42, 0.6, 0.39, 0.11, and 0.33, respectively. Among the results, the Ks of the NPF7 subfamily was much higher than that of other NPF subfamilies, suggesting that they may have originated from more ancient duplication events. Except for CsNPF5.18/CsNPF5. 19, the Ka/Ks ratio of almost all NPF duplicated pairs was less than

Expansion and Evolution of CsNPF Genes
To study the evolutionary processes of CsNPF family, the gene duplication pattern of CsNPF genes was analyzed ( Figure 6 and Table S1). Thirty-three of the 109 CsNPF genes were found to be segmentally duplicated, 31 of which were located on chromosomes 1, 2, 4, 5, 6, 9, 10, 11, 12, 13, 14, and 15, and 2 on Contig ( Figure 6A). There were 21 pairs of syntenic relationships in tea plant ( Figure 6B). A total of 13 tandem repeats containing 20 CsNPF genes were identified. In addition, 25 and 31 CsNPFs were also found to be dispersed and proximal duplicates, respectively, which may also contribute to the expansion of the CsNPF family genes. 1, especially in the NPF7 subfamily. The results indicate that tea plant NPF genes migh suffer from strong purifying selection throughout the long evolutionary events.

Synteny Analyses of NPF Genes between Tea Plant and Six Representative Species
To further explore the potential evolutionary mechanisms of the tea plant NPF fam ily, we selected 6 representative species, including 4 eudicots (A. thaliana, P. trichocarpa Vitis vinifera, and Solanum lycopersicu) and 2 monocots (O. sativa, and Sorghum bicolor), to construct comparative syntenic maps with tea plant (Figure 7 and Table S4). The result showed that the tea plant has more gene pairs with eudicots than with monocots. Ther were 75 pairs of syntenic pairs between tea plant and P. trichocarpa, followed by V. vinif era (65), S. lycopersicu (52), A. thaliana (45). However, there were only 7 and 6 pairs o syntenic relationships of tea plant with O. sativa, and S. bicolor, respectively. Among them, 3 NPF genes, CsNPF1.2, CsNPF2.2, and CsNPF7.4, were simultaneously identified between tea plant and the other six species, indicating that these genes may have vita roles in the evolution of the NPF gene family. To further study the selection pressure of tea plant NPF genes during evolution, the values of synonymous (Ks) and nonsynonymous (Ka) substitution rates, and Ka/Ks ratios for each duplicated pair from different subfamilies were calculated ( Figure 6C and Table S3).
The mean values of Ka of gene pairs in NPF1-NPF8 subfamilies were 0.05, 0.12, 0.07, 0.12, 0.10, 0.08, 0.14, and 0.09, respectively. The mean values of Ks were 0.09, 0.27, 0.29, 0.34, 0.35, 0.39, 1.24, and 0.44, respectively. The mean values of Ka/Ks were 0.52, 0.59, 0.27, 0.42, 0.6, 0.39, 0.11, and 0.33, respectively. Among the results, the Ks of the NPF7 subfamily was much higher than that of other NPF subfamilies, suggesting that they may have originated from more ancient duplication events. Except for CsNPF5.18/CsNPF5.19, the Ka/Ks ratio of almost all NPF duplicated pairs was less than 1, especially in the NPF7 subfamily. The results indicate that tea plant NPF genes might suffer from strong purifying selection throughout the long evolutionary events.

Synteny Analyses of NPF Genes between Tea Plant and Six Representative Species
To further explore the potential evolutionary mechanisms of the tea plant NPF family, we selected 6 representative species, including 4 eudicots (A. thaliana, P. trichocarpa, Vitis vinifera, and Solanum lycopersicu) and 2 monocots (O. sativa, and Sorghum bicolor), to construct comparative syntenic maps with tea plant (Figure 7 and Table S4). The results showed that the tea plant has more gene pairs with eudicots than with monocots. There were 75 pairs of syntenic pairs between tea plant and P. trichocarpa, followed by V. vinifera (65), S. lycopersicu (52), A. thaliana (45). However, there were only 7 and 6 pairs of syntenic relationships of tea plant with O. sativa, and S. bicolor, respectively. Among them, 3 NPF genes, CsNPF1.2, CsNPF2.2, and CsNPF7.4, were simultaneously identified between tea plant and the other six species, indicating that these genes may have vital roles in the evolution of the NPF gene family.

Regulatory Mechanism in the Promoter Regions of CsNPF Genes
To understand the potential functions of CsNPF genes, the cis-acting regulatory elements (CREs) in promoter regions of CsNPFs were predicted using PlantCARE. A total of 111 types of CREs were identified from 109 CsNPF gene promoters (Table S5). Each CsNPF contained 20 to 40 CREs in the promoter region, and some closely related genes showed similar CREs type. Several common CREs, such as CAAT-box and TATA-box and some light-responsive elements (G-box, Box 4, GT1-motif and TCT-motif) were obtained. Meanwhile, some CREs were shown to respond to stresses, such as drought-inducible MYB binding site elements (MBS, 50), low temperature-responsive elements (LTR, 67), wound-responsive elements (WUN-motif, 64), and anaerobic-inducible element (ARE, 100) ( Figure 8A). Large numbers of CREs were involved in hormone responses, such as MeJA-responsive elements (TGACG-motif and

Regulatory Mechanism in the Promoter Regions of CsNPF Genes
To understand the potential functions of CsNPF genes, the cis-acting regulatory elements (CREs) in promoter regions of CsNPFs were predicted using PlantCARE. A total of 111 types of CREs were identified from 109 CsNPF gene promoters (Table S5). Each CsNPF contained 20 to 40 CREs in the promoter region, and some closely related genes showed similar CREs type. Several common CREs, such as CAAT-box and TATA-box and some lightresponsive elements (G-box, Box 4, GT1-motif and TCT-motif) were obtained. Meanwhile, some CREs were shown to respond to stresses, such as drought-inducible MYB binding site elements (MBS, 50), low temperature-responsive elements (LTR, 67), wound-responsive elements (WUN-motif, 64), and anaerobic-inducible element (ARE, 100) ( Figure 8A). Large numbers of CREs were involved in hormone responses, such as MeJA-responsive elements (TGACG-motif and CGTCA-motif), gibberellin-responsive elements (TATC-box, P-box and GARE-motif), Auxin (TGA-element and AuxRR-core), abscisic acid-responsive elements (ABREs), Ethylene (ERE) and salicylic acid reaction (TCA-elements) elements. In addition, many transcription factor (TF) binding sites were identified, such as WRKY binding sites (W box) and MYB binding sites (MRE, MYB, MBS, and MBSII).
Transcription factors regulate the expression of target genes by binding to CR promoters. To further explore the transcriptional regulation of CsNPFs, we specu the potential TFs that bind to the CsNPFs promoters ( Figure 8B and Table S6). T sults showed that 612 TF genes from 43 TF gene families had potential target bi sites in the CsNPFs promoter regions. Among the 612 TFs, MYB (90) gene family w most abundant, followed by ERF (80), NAC (59), WRKY (43), bZIP (35), bHLH (32 C2H2 (25) gene family.

Expression Patterns of CsNPFs in Different Tissues
To investigate the expression patterns of NPF genes in tea plant, we analyze expression levels of CsNPFs in 8 tissues (apical buds, young leaves, mature leave leaves, immature stems, flowers, young fruits, and tender roots) (Figure 9 and Tab Of 109 CsNPF genes, 60 (55%) were ubiquitously expressed in all 8 tissues; 39 (36%) low expressed with meanFPKM < 1. CsNPF genes showed different expression ch teristics in different subfamilies. In NPF1, NPF6, and NPF8 subfamilies, most C genes were expressed at high levels in most tissues. In particular, all members of subfamily genes were highly expressed in almost all tissues. Most members of the and NPF7 subfamily genes showed lower expression levels. Some CsNPF gene played tissue-specific or preferential expression patterns. For example, in NPF3 su ily, CsNPF2.6, CsNPF2.9 and CsNPF2.10 were preferentially expressed in young mature leaves, while CsNPF2.7, CsNPF2.11 and CsNPF2.12 were preferentially expr in root. In the NPF2 subfamily, CsNPF3.7 was preferentially expressed in root, CsNPF3.10 was preferentially expressed in mature leaves. CsNPF5.13, CsNP CsNPF5. 21 and CsNPF5.22 in NPF4 subfamily were mainly expressed in root. I Transcription factors regulate the expression of target genes by binding to CREs in promoters. To further explore the transcriptional regulation of CsNPFs, we speculated the potential TFs that bind to the CsNPFs promoters ( Figure 8B and Table S6). The results showed that 612 TF genes from 43 TF gene families had potential target binding sites in the CsNPFs promoter regions. Among the 612 TFs, MYB (90) gene family was the most abundant, followed by ERF (80), NAC (59), WRKY (43), bZIP (35), bHLH (32), and C2H2 (25) gene family.

Expression Patterns of CsNPFs in Different Tissues
To investigate the expression patterns of NPF genes in tea plant, we analyzed the expression levels of CsNPFs in 8 tissues (apical buds, young leaves, mature leaves, old leaves, immature stems, flowers, young fruits, and tender roots) (Figure 9 and Table S7). Of 109 CsNPF genes, 60 (55%) were ubiquitously expressed in all 8 tissues; 39 (36%) were low expressed with meanFPKM < 1. CsNPF genes showed different expression characteristics in different subfamilies. In NPF1, NPF6, and NPF8 subfamilies, most CsNPF genes were expressed at high levels in most tissues. In particular, all members of NPF6 subfamily genes were highly expressed in almost all tissues. Most members of the NPF4 and NPF7 subfamily genes showed lower expression levels. Some CsNPF genes displayed tissue-specific or preferential expression patterns. For example, in NPF3 subfamily, CsNPF2.6, CsNPF2.9 and CsNPF2.10 were preferentially expressed in young and mature leaves, while CsNPF2.7, CsNPF2.11 and CsNPF2.12 were preferentially expressed in root. In the NPF2 subfamily, CsNPF3.7 was preferentially expressed in root, while CsNPF3.10 was preferentially expressed in mature leaves. CsNPF5. 13

Expression of CsNPFs under Different Nitrate Concentrations
Nitrate is the main substrate for NPF protein transport. Here, the expression profiles of 20 CsNPF genes were detected under different nitrate concentrations. The 20 CsNPF genes were selected from the 8 subfamilies, which were representative and can explain the expression profiles of genes from different subfamilies ( Figure 10

Expression of CsNPFs under Different Nitrate Concentrations
Nitrate is the main substrate for NPF protein transport. Here, the expression profiles of 20 CsNPF genes were detected under different nitrate concentrations. The 20 CsNPF genes were selected from the 8 subfamilies, which were representative and can explain the expression profiles of genes from different subfamilies ( Figure 10

Subcellular Localization of CsNPF2.3 and CsNPF6.1
The CsNPF2.3-GFP fusion gene, CsNPF6.1-GFP fusion gene and empty vecto taining GFP were transformed into tobacco epidermal cells. The results showed th fusion proteins CsNPF2.3 and CsNPF6.1 were specifically located in the plasma brane of tobacco leaves cells, and control GFP signals were observed in the nucleu toplasm, and cell membrane ( Figure 11). These results indicated that CsNPF2. CsNPF6.1 were located in the plasma membrane, which was consistent with the r predicted by bioinformatics and consistent with the characteristics of transmem proteins involved in NO3 -transport. The CsNPF2.3-GFP fusion gene, CsNPF6.1-GFP fusion gene and empty vector containing GFP were transformed into tobacco epidermal cells. The results showed that the fusion proteins CsNPF2.3 and CsNPF6.1 were specifically located in the plasma membrane of tobacco leaves cells, and control GFP signals were observed in the nucleus, cytoplasm, and cell membrane ( Figure 11). These results indicated that CsNPF2.3 and CsNPF6.1 were located in the plasma membrane, which was consistent with the results predicted by bioinformatics and consistent with the characteristics of transmembrane proteins involved in NO 3 transport.

Discussion
NPF genes comprise a large family of members widely distributed in eukaryote An increasing number of studies has confirmed that NPF genes are involved in t transport of nitrate/nitrite [8], di/tri-peptides [27], and various hormones [28][29][30]. As important cash crop, tea plant is widely planted in China and all over the world becau of its functional activity and important economic value. Tea plant is a plant that needs lot of nitrogen fertilizer [19]. So far, functional studies of NPF gene have been carried o in Arabidopsis, rice, and other plants, but almost no reports on tea plant. In this stud 109 CsNPF genes were identified from the tea plant genome and divided into 8 su groups based on the similarity of structure.
It is widely believed that the variation in gene structure is one of the representati traces of gene family evolution. According to the analysis of gene structure and co served motifs, CsNPF was relatively conserved structurally. Most CsNPFs contain all motifs, and no specific motifs belonging to a subfamily have been found. In the sam subfamily, some closely related genes show the same motif deletion. Analysis of ge structure showed that most CsNPF genes (73%) contained 3 or 4 introns. Similar resu have been reported in Arabidopsis, P. trichocarpa, soybean [31], and B. napus [25]. On t whole, CsNPFs within the same subfamily shared similar gene structures and moti Changes in gene structure and conserved motifs may be important reasons for the fun tional diversification of tea plant NPF family genes.
Gene duplication and loss events are the main drivers of species evolution, pr moting the expansion of gene families and the generation of novel genes [32]. NPF gen are widely distributed in all subfamilies of 36 species, although NPF1 and NPF2 subfam ilies are absent in lower plants S. Moellendorffii and P. Patens. In original angiosperms, 8 subfamilies were present, suggesting that the duplication and diversification of NP genes may have occurred before the differentiation of eudicots and monocots, and th NPF gene duplication in eudicots and monocots was an independent process. Mo NPF1 and NPF2 subfamily members were found in eudicots, while more NPF3, NP and NPF8 subfamily members were found in monocots. The number of NPF1, NPF NPF4 and NPF8 subfamily members in tea plant was higher than that in eudicots, ind

Discussion
NPF genes comprise a large family of members widely distributed in eukaryotes. An increasing number of studies has confirmed that NPF genes are involved in the transport of nitrate/nitrite [8], di/tri-peptides [27], and various hormones [28][29][30]. As an important cash crop, tea plant is widely planted in China and all over the world because of its functional activity and important economic value. Tea plant is a plant that needs a lot of nitrogen fertilizer [19]. So far, functional studies of NPF gene have been carried out in Arabidopsis, rice, and other plants, but almost no reports on tea plant. In this study, 109 CsNPF genes were identified from the tea plant genome and divided into 8 subgroups based on the similarity of structure.
It is widely believed that the variation in gene structure is one of the representative traces of gene family evolution. According to the analysis of gene structure and conserved motifs, CsNPF was relatively conserved structurally. Most CsNPFs contain all 10 motifs, and no specific motifs belonging to a subfamily have been found. In the same subfamily, some closely related genes show the same motif deletion. Analysis of gene structure showed that most CsNPF genes (73%) contained 3 or 4 introns. Similar results have been reported in Arabidopsis, P. trichocarpa, soybean [31], and B. napus [25]. On the whole, CsNPFs within the same subfamily shared similar gene structures and motifs. Changes in gene structure and conserved motifs may be important reasons for the functional diversification of tea plant NPF family genes.
Gene duplication and loss events are the main drivers of species evolution, promoting the expansion of gene families and the generation of novel genes [32]. NPF genes are widely distributed in all subfamilies of 36 species, although NPF1 and NPF2 subfamilies are absent in lower plants S. Moellendorffii and P. Patens. In original angiosperms, all 8 subfamilies were present, suggesting that the duplication and diversification of NPF genes may have occurred before the differentiation of eudicots and monocots, and that NPF gene duplication in eudicots and monocots was an independent process. More NPF1 and NPF2 subfamily members were found in eudicots, while more NPF3, NPF7 and NPF8 subfamily members were found in monocots. The number of NPF1, NPF3, NPF4 and NPF8 subfamily members in tea plant was higher than that in eudicots, indicating these subfamilies seem to be more active in duplication and may have more function in the tea plant.
Gene family expansion mainly depends on tandem, segmental, or whole-genome duplications [33,34]. Recently completed genome sequencing of the tea plant indicated that tea plant has undergone two rounds of whole-genome duplications (WGD) [35][36][37]. Moreover, the large number of genes in the CsNPF family suggests that it evolved through many duplication events in tea plant. In B. napus, 137 (71%) of 193 BnaNPFs were found to be segmentally duplicated, and only 2 genes were identified as tandem duplication, suggesting that segmentation duplication was the main driver of NPF gene family expansion in B. napus [34]. Here, 33 (30%) and 20 (18%) of 109 CsNPFs were identified involved in segmental duplication and tandem events, respectively. Dispersed and proximal duplicates were also identified in the tea plant NPF gene family. These results indicated that the expansion forms of the NPF gene family in tea plant were more diversified. The number of NPF3 subfamily in tea plant was 3.3 times that in eudicots. Notably, there are two segmental and two tandem duplications in the NPF3 subfamily of tea plant, which may account for the higher number of NPF3 subfamily genes in tea plant than in other species. Few gene pairs were identified between tea plant and two monocots; while numerous colinearity gene pairs were found between tea plant and other 4 eudicots because of their close relationship. The results showed that the NPF gene differentiated rapidly in monocots and eudicots. In addition, three CsNPF genes, CsNPF1.2, CsNPF2.2, and CsNPF7.4, were colinear paired with all six species, suggesting that these colinear pairs already existed before the differentiation of monocots and eudicots ancestors.
The expression pattern of a gene is usually closely related to its CREs [38]. Many CREs related to development, stress, and phytohormone were detected in the CsNPF promoter regions, suggesting that CsNPFs may play different roles in biological processes. Numerous studies have demonstrated that NPF can transport hormone substrates. In Arabidopsis, at least 9 NPF members have ABA transport function and 18 NPF members have GA transport function [28]. In B. napus, 32.66% of BnaNPF expressions (65/199 gene) were regulated by at least one hormone [39]. In this study, most cis-elements involved in phytohormone responses, such as MeJA-(80/109 genes), Ethylene-(81/109), and ABAresponsive CRE (74/109 genes), were found in a series of CsNPF promoters, suggesting their potential hormone-inducing characteristics. The absorption and transport of nitrogen plays an important role in plant stress adaptation [40,41]. AtNPF6.3/AtNRT1.1 is involved in regulating drought tolerance of Arabidopsis [42]. AtNRT1.5 plays a negative role in salt tolerance and drought resistance in Arabidopsis, regulating the long-distance transport and spatial distribution of Na + , and affecting the expression of stress-responsive genes [43]. Many stress-related elements related to low temperature (LTR), drought (MBS), anaerobic (ARE) and wound (WUN-Motif) were also identified in the CsNPF promoters. Previous studies have shown the transcription factor AtERF59 binds directly to the AtNRT1.8 promoter and ethylene insensitive 3 (EIN3) binds to the AtNRT1.5 promoter [44]. Arabidopsis AtMYB59 positively regulates AtNRT1.5 expression to coordinate root-to-shoot K + /NO 3 transport [45]. Many transcription factor binding sites were found in the CsNPF promoter regions. The results suggested that the expression of CsNPF genes might be regulated by multiple transcription factors.
Genes expressed in given tissues may have specific tissue functions, which provides crucial clues to understanding gene functions [46]. The CsNPF genes showed complex expression patterns, and closely related genes in the same subfamily showed similar expression patterns, reflecting the consistency of structure and function. Some CsNPF genes showed obvious tissue preference. For example, CsNPF2.6, CsNPF2.9, CsNPF2.10 were mainly expressed in leaf. CsNPF4.5, CsNPF7.3, CsNPF7.6, CsNPF8.2 showed preferred expression in flower. Most of the tissue-specific expression genes are preferentially expressed in root, such as CsNPF2.7, CsNPF2.11, CsNPF2.12, CsNPF3.7, CsNPF5.13, CsNPF5.21, CsNPF5.22, CsNPF7.3, CsNPF7.4. Similarly, most rice tissue-specific genes were preferentially expressed in root [20]. AtNPF6.3/AtNRT1.1 was a dual-affinity nitrate transporter mainly expressed in Arabidopsis root and involved in nascent organ development [47]. In T. aestivum, TaNPF6.2 was predominantly expressed in root; TaNPF6.1 was highly expressed in different tissues; while TaNPF6.4 was preferentially expressed in spike and node, but low in root and leaf [16]. All 4 CsNPF6 genes were highly expressed in tea plant except for the root and old leaf. The variation in expression patterns suggested the divergence of AtNPF6.3 orthologues in different species.
Nitrate is the main substrate for NPF protein transport [15,16]. In Arabidopsis, more than a third of the NPF genes have been shown to transport nitrates [16]. AtNPF4.6/AtNRT1.2 has been reported to be expressed mainly in the root of Arabidopsis and is involved in constitutive nitrate uptake [48]. In rice, OsNPF4.5 plays a key role in mycorrhizal NO 3 acquisition [49]. Here, the transcripts of CsNPF4.1 and CsNPF4.8 in tea plant were significantly up-regulated under low nitrate conditions while down-regulated under high nitrate conditions. AtNPF7.3/AtNRT1.5 has been reported to be involved in nitrate loading in root xylem and transport nitrate from root to shoot [50]. OsNPF7.3 was mainly expressed in lateral root and stem of rice, and could improve nitrogen utilization efficiency in rice paddy [51]. Its homologous genes in tea plant were CsNPF7.3 and CsNPF7.4, which were preferentially expressed in root, and first decreased and then increased with the increase of nitrate concentration. In this study, the expression of several CsNPF genes changed significantly, indicating that these genes may be involved in nitrate uptake.

Plant Materials and Treatments
Fourteen-month-old cutting seedlings from tea plant, Zhongming 6 hao, were selected for this study. As previously [52], tea seedlings were cultured in full nitrogen hydroponic solution for 2 months. Then the seedlings were transferred to the nitrogen-deficient (without N) nutrient solution for 2 weeks of N-starvation treatment. Afterward, the seedlings were cultivated in the nutrient solution containing 0.1, 1, and 10 mM nitrate, respectively. Nitrogen starvation was used as the control (CK). The white roots were harvested after 2 h of treatment, rapidly frozen in liquid nitrogen, and then stored at −80 • C.

Phylogenetic Analysis and Structural Characterization
Sequence alignments of NPF proteins from tea plant, A. thaliana, and P. trichocarpa were performed clustalx1.83 software with default parameters. The unrooted phylogenetic tree was constructed using MEGA 5.0 software with the neighbor-joining (NJ) methods, bootstrapping with 1000 replicates. NPF sequences of A. thaliana and P. trichocarpa were obtained from previous reports (Table S8) [53]. The exon/intron structures of CsNPF genes were obtained from tea plant genome annotation [36]. Conserved motifs were analyzed using the Multiple Em for Motif Elicitation 5.4.1 (MEME, https://meme-suite.org/meme/ (accessed on 9 November 2021)) program.

Cis-Element Analysis
The sequence of 2000 bp upstream of the CsNPF gene transcription start site (ATG) was considered as promoter sequence. These promoter sequences were used to evaluate the putative cis-acting regulatory elements (CREs) using the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/ (accessed on 9 November 2021)). The transcription factor (TF) binding sites in CsNPF promoter sequences were predicted using the PlantTFDB database (http://planttfdb.gao-lab.org/ (accessed on 9 November 2021)).

Chromosomal Arrangement and Gene Duplication of CsNPF Genes
Physical positions of 109 CsNPF genes were extracted from the tea plant genome database. Mapchart software was used to visualize the chromosome localization [54]. Multiple Collinear Scan Toolkit (MCScanX) was used to determine the gene duplication events and collinearity relationships of tea plant NPF genes [55]. A chromosome region containing two or more homologous genes within a 200 kb range is defined as a tandem duplication event. The syntenic analysis maps of tea plant and other selected plants were constructed and visualized using the MCScanX and TBtools software [55,56]. The synonymous (Ks) and nonsynonymous (Ka) substitution rates, and Ka/Ks ratios for each duplicated pair (similarity of aligned regions > 70%) were calculated based on CDS alignments of the CsNPF genes using TBtools [56].

Transcript Abundance Analysis
For expression profiling of tea plant NPF genes, we utilized the transcriptome sequencing data of eight tissues of tea plant that were derived from the Tea Plant Information Archive (TPIA) (http://tpia.teaplant.org/ (accessed on 20 July 2021)) [35]. The eight tea plant tissues were apical buds (AB), young leaves (YL), mature leaves (ML), old leaves (OL), immature stems (ST), flowers (FL), young fruits (FR), and tender roots (RT). The expression cluster of CsNPF was analyzed and displayed by HemI 1.0 software.

RNA Isolation, Reverse Transcription, and RT-qPCR Detection
Total RNAs were extracted from samples using the EASYspin plant RNA extraction kit (Aidlab Biotechnologies, Beijing, China), and first-strand cDNA synthesis was performed using the PrimeScript RT reagent kit (TaKaRa, Dalian, China). RT-qPCR was performed using a LightCycler 480 machine (Roche Diagnostics) with SYBR Green reagents (Takara, Japan). RT-qPCR was performed in a 10 µL reaction mixture consisting of 5 µL of SYBR Premix Ex, 1 µL of diluted template cDNA, 0.3 µL of each primer, and 3.4 µL of ddH 2 O. The PCR amplification profile was as follows: denaturation at 95 • C for 30 s, followed by 40 cycles of 95 • C for 10 s, 58 • C for 10 s and 72 • C for 15 s. Three independent PCR reactions were performed on each gene, and the relative gene expression was calculated using the 2 −∆∆CT method [57]. Tea plant GAPDH was selected as the reference gene [52]. All primers used are listed in Table S9.

Subcellular Localization
Subcellular localization analysis of tobacco epidermis was according to Zhang et al.'s method [58]. The CsNPF2.3 and CsNPF6.1 genes of the deletion termination codon were amplified using specific primers (Table S9) and inserted into the pBWA(V)HS-GFP vector. The constructed vector plasmid was transformed into Agrobacterium GV3101. The Agrobacterium cells were injected into the lower epidermis of one-month-old tobacco leaves and cultured under low light for 2 days. The GFP signals were observed and imaged under the confocal laser microscope (Zeiss, Germany).

Conclusions
This study performed a comprehensive analysis of NPF genes in tea plant. A total of 109 CsNPF genes were identified from the tea plant genome, and their gene structure, conserved motifs, chromosome location, cis-elements, gene duplication, evolutionary relationships, and expression patterns were systematically analyzed. Gene structure and motif analysis indicated that the CsNPF proteins were highly conserved in each subfamily. Syntenic analysis revealed that the expansion of tea plant NPF gene family was caused by multiple duplication types. The CsNPF family genes of tea plant experienced strong purifying selective pressure during evolution. Most of the selected CsNPF genes were induced by nitrate treatments. In addition, we proved that CsNPF2.3 and CsNPF6.1 were localized in the plasma membrane, which was consistent with the characteristics of transmembrane proteins involved in NO 3 − transport. These results provide a valuable basis for studying the evolution and function of CsNPF genes, and aid in improving N absorption in tea plant breeding.