Characterisation and expression profiles of the NPF gene family in Cannabis sativa L. under low nitrogen

Abstract

Background

The Nitrate transporter1/peptide transporter family (NPF) affects nitrate use efficiency (NUE) by regulating plant nitrate absorption and distribution. Many NPF-related genes identified across various crops have been shown to improve NUE. However, the characteristics of NPF in cannabis sativa L. and their functions remain unclear.

Results

In this study, 66 NPF genes of hemp (Cannabis sativa L.) were identified from the pink pepper genome. The phylogenetic analysis showed that these genes could be divided into eight subfamilies. The NPF genes of Cannabis sativa were renamed according to their chromosomal locations and characterised through bioinformatic analysis. Transcriptome analysis of roots from two hemp varieties with distinct NUE under two nitrogen concentrations (2-hour treatment) showed varied expression patterns across all NPF genes. Among those with higher expression levels, quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis successfully screened out two candidate genes, CsNPF4.3 and CsNPF5.14, as potentially involved in NUE regulation.

Conclusions

This work deepens our understanding of the gene structure and expression of NPF genes in hemp, and their roles in nitrate response, paving the way for improving hemp nitrogen efficiency through molecular breeding. Altogether, these findings suggest that the expansion of these gene families in hemp could provide valuable genetic variability for identifying NUE-related candidate genes in future breeding programs in the context of low-impact, sustainable agriculture.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08990-0.

Background

Nitrogen is an essential component of nucleic acids and amino acids, which build various molecules such as phospholipids, chlorophyll, enzymes, vitamins, and hormones. Nitrogen deficiency and overuse can severely affect plants’ growth and development. In plants, nitrogen utilisation includes absorption, transport, assimilation, and reuse. Plants absorb nitrogenin two forms, i.e. inorganic nitrogen (including nitrate and ammonium salts) and organic nitrogen (including peptides, amino acids, etc.) [1].

However, in most plants, nitrate nitrogen (NO₃⁻) is the main form in which nitrogen is absorbed and utilised in the soil [2]. Once the roots absorb it, most of the NO₃⁻ will be transported to the aboveground, with only a small portion being reduced in the roots.

After it is transported to the aboveground, NO₃⁻ is reduced to nitrite (NO₂⁻) in the cytoplasm under the action of nitrate reductase (NR). Nitrite is reduced to NH₄⁺ in chloroplasts and proplastids under the action of nitrate reductase (NiR). Finally, glutamate (Glu) is synthesized by the catalytic action of two key enzymes, i.e.,glutamate synthase (GOGAT) and glutamine synthase (GS). Then the resulting glutamate is synthesised into other amino acids and amides through transamination forming nitrogen-containing substances that can be utilised directly by plants and participate in various physiological and biochemical pathways [3].

Nitrogen uptake and utilisation are the most crucial factors for improving plant nitrogen use efficiency (NUE), aside from carbon metabolism and amino acid biosynthesis. To enhance NUE, it is essential to identify and characterise the genes involved in nitrogen metabolism that regulate nitrogen uptake and utilisation in plants [4]. Some genes related to nitrogen acquisition, accumulation, assimilation, and transport have been identified in plants such as OsNRT2.3a and OsNRT1.1b in rice [5], and GhNRT2.1e in Gossypium hirsutum [6]. Nitrogen uptake and utilisation in plants can be influenced by abiotic stresses such as drought, nutrient deficiency, and salinity [7]. The more N fertiliser is applied to crops like maize, soybean, and wheat (Triticum aestivum L.), the lower the NUE value. The highest NUE can be achieved under no- or low-N fertiliser treatments. Under low nitrogen (LN) conditions, the efficiency of uptake is more important than that of utilisation [1].

Plants mainly absorb nitrogen via ion transport proteins located in the root cell membrane. These nutrient transport proteins are key targets for genetic improvement aimed at increasing nutrient efficiency in crops. In plants, genes encoding these transport proteins are mostly grouped into polygene families. Currently, four major gene families related to nitrate absorption and transport have been identified in plants: SLAC/SLAH (slow ion channels and their genomes), CLC (chloride channel), NRT2 (nitrate transporter 2), and NRT1/PTR (NPF; nitrate transporter 1/peptide transporter family). Plants have developed various nitrogen uptake and transport mechanisms for different nitrogen sources. In Arabidopsis and rice, root nitrate transport systems include a low-affinity transport system (LATS) and a high-affinity transport system (HATS). The high-affinity transport system can be subdivided into the inducible high-affinity transport system (iHATS) and the constitutive high-affinity transport system (cHATS). The dominant system depends on the nitrate concentration in the environment. When the concentration exceeds 1 mmol/L, the low-affinity system is preferential; when it is below 1 mmol/L, the high-affinity system predominates [2].

The NPF, also known as the NRT1/PTR family, is the largest nitrate transporter family in plants. NRT1, which belongs to the PTR (peptide transporters) family (in which family members are also described as proton-coupled oligopeptide transporter (POT) or solute carrier 15 (SLC15)) class, is mainly composed of low-affinity nitrate transporters. A member of the NRT1 family in rice, OsNPF3.1, can effectively improve nitrogen utilisation efficiency [8]. Under N-limiting fields, the overexpression of ZmNRT1.1B (ZmNPF6.6), which was identified as high-affinity NO₃⁻ transport and signalling in modern maize hybrids, improved grain yield [9]. The transporter OsNRT2.3b, found in rice, played an important role in improving growth, yield, and NUE in barley and rice [10].

NRTs can be induced by LN stress. LN induced the expression of most NRTs in Sophora japonica L SiNRT2.3 and SiNRT2.4 in foxtail millet, the members of high-affinity nitrate transporter genes, are key genes in LN tolerance variety [11]. LN stress tolerance genes CsNRT1.1 and CsNPF5.2 are all detected in cucumber [12]. TaNPF6.3-6D, TaNIA1–6B, TaNRT2.3–3D, TaAAP14–5 A/5D, TaUPS2–5 A, and TaGLN1;2–4B, which are associated with efficient transport and assimilation of nitrate, are all identified from the allohexaploid wheat [13]. Compared with the LN-sensitive variety, the LN-tolerant variety has higher expression of TaNRT2.2, TaNRT2.1, and TaNRT1.1 in wheat root, TaNRT1.8 and TaNRT1.5 in wheat leaves and roots [14]. The increasing expression of NRT2.3 and NPF8.3 in the LN-tolerant variety can promote the nitrogen transport ability from root to shoot in tomato [15].

The prevalence of hemp (Cannabis sativa L) is mainly attributed to environmental factors and the environmentally friendly traits of hemp cultivation. During cultivation, nitrogen plays a crucial role in growth, development, and production. Under intensive agricultural practices, nitrogen fertiliser is needed to compensate for nitrogen loss from the soil. Nitrogen fertiliser can increase the yield of stems, fibre, seeds, and protein content. Unfortunately, floral hemp is sensitive to nitrogen fertiliser [16]. However, up to 70% of nitrogen fertiliser that plants cannot sufficiently absorb and utilise will enter the environment, causing issues such as water eutrophication, soil acidification, and soil hardening [17]. Agricultural production should urgently reduce excessive reliance on nitrogen fertiliser and adopt cultivation of low-nitrogen-efficient varieties, as this is an effective way to address the issue. Identifying genes related to efficient utilisation of low nitrogen forms is the basis for developing such varieties. Currently, some genes from various crops have been identified and proven effective in improving NUE. However, none have been identified in hemp.

In this study, we conducted genome-wide analyses of the NPF gene family in hemp, along with several bioinformatic tests of the candidates, including chromosomal location, gene structure, conserved motifs, phylogenetic and collinearity relationships, and cis-element identification. We then examined the expression patterns of NPF candidates in two varieties under normal and low nitrogen conditions. Our findings lay the groundwork for further research into the molecular and physiological functions of NPF family genes in hemp and their potential application in hemp molecular breeding.

Methods

Genome-wide identification and characterisation of CsNPF proteins

The latest reference genome files of Cannabis sativa L. (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_029168945.1/) and the protein sequences of the NPF gene family of Arabidopsis thaliana were downloaded from the National Centre for Biotechnology Information (NCBI)( https://www.ncbi.nlm.nih.gov/) on 12 March 2025. All the AtNPF proteins were used as queries for BLAST analysis in TBtools-II (e-value, 1 × 10 ⁵) to identify potential NPF protein sequences in Cannabis sativa L [18]. Subsequently, the HMM files of the PTR2 domain (PF00854), downloaded from Pfam (https://www.ebi.ac.uk/interpro/entry/pfam/) within Interpro (https://www.ebi.ac.uk/interpro/) [19], were used to identify possible NPF gene family members in Cannabis sativa L. using the Simple HMM Search function in Tbtools II.

The results from both analyses were combined, with duplicate sequences removed.

Finally, candidate sequences were validated through the Conserved Domain Database (CDD) (https://www.ncbi.nlm.nih.gov/cdd/) [20], accessed on 21 March 2025.

Then, the physicochemical properties of these candidate cannabis sativa NPF (CsNPF) proteins were predicted by the “Protein Parameter Calc” function of TBtools-II. Finally, the subcellular localisations of these CsNPF proteins were predicted using the WoLF PSORT online website (https://wolfpsort.hgc.jp/).

Sequence alignment, phylogenetic analysis and chromosomal location

NPF protein sequences from four representative dicotyledonous species, including Arabidopsis thaliana (53 proteins, 53 genes),,Populus trichocarpa (88 proteins, 51 genes), Ricinus communis L. (46 proteins, 46 genes), and Vitis vinifera (81 proteins, 45 genes) were selected and aligned with Cannabis sativa to construct the phylogenetic tree. These species were chosen to represent diverse evolutionary characteristics, including a model herbaceous plant (Arabidopsis thaliana), a woody species with lineage-specific gene expansion (Populus trichocarpa), a relatively conserved genome (Vitis vinifera), and an intermediate species (Ricinus communis L.,). This selection enables a comprehensive evaluation of the evolutionary conservation and divergence of the NPF gene family within dicots. All candidate proteins of these species were download from Phytozome (https://phytozome-next.jgi.doe.gov/) and confirmed by the presence of conserved PTR2 domain (PF00854). The MUSCLE tool of the MEGA11 software was used to align the above protein sequences of the NPF gene family members. And then the maximum-likelihood (ML) method with default parameter was used to construct a phylogenetic tree. Finally, the online tool iTOL (https://itol.embl.de/upload.cgi (accessed on 5 April 2026)) was used to produce a high-quality tree map. And the chromosomal locations of these CsNPF proteins were visualised using TBtools-II. The Members within each subfamily were numbered sequentially (e.g., CsNPF1.1, CsNPF1.2…) according to their physical order on the chromosomes.

Structure, domain and conserved motif analyses

The phylogenetic tree of CsNPF proteins was constructed using the same method as mentioned in 1.2. Gene structure annotations files in GFF format for cannabis sativa were prepared to display the CsNPF gene structure. The conserved motifs of CsNPF proteins were analysed using MEME (https://meme-suite.org/meme/tools/meme (accessed on 15 April 2025)) [21]with the default parameters, except that the maximum number of motifs was set to 10. And the Batch-CD Search with the default parameters in Conserved Domain Database (CDD) (https://www.ncbi.nlm.nih.gov/cdd/(accessed on 15 April 2025)) was used to analyse the conserved domains. All results were combined and visualised in TBtools-II.

Cis-regulatory element analysis of CsNPF genes

The 2000 bp upstream of the initiation codon of each CsNPF gene was extracted using TBtools-II. The cis-regulatory elements were identified using PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/ (accessed on 19 April 2025)) [22]. The results were displayed using TBtools-II.

Collinearity synteny analysis

The collinearity maps were visualised using TBtools-II. The Arabidopsis reference genome files, downloaded from NCBI (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_000001735.4/), were prepared to conduct the collinearity and synteny analysis. The one-step MCScanX function in TBtools-II was used to identify gene duplication events and collinearity relationships using the default parameters. To explore the collinearity relationships of the orthologous NPF genes in cannabis with those in Arabidopsis, collinearity maps were constructed using the multiple synteny plot tool in TBtools-II. Then, the Ka/Ks values of gene pairs of cannabis sativa were calculated by the simple Ka/Ks Calculator of TBtoolsII.

Plant material and treatment for RNA-seq data analysis of NPF genes

Two hemp varieties identified in our previous research were selected as our experimental materials. Longma No.2, a low nitrogen (LN) tolerant variety, and Hanma No.10, an LN sensitive variety. They are all provided by the Daqing Branch of Heilongjiang Academy of Sciences. The experiment was also conducted in Daqing Branch of the Heilongjiang Academy of Sciences.

Firstly, the healthy, full seeds of the two varieties were germinated in a petri dish in a manual climatic box. Germinated seeds were picked out to cultivate in vermiculite substrate (vermiculite: water = 3:2). The seeds were covered with 1–2 cm vermiculite.

After 4 days, the seedlings with green cotyledons will be transplanted into a hydroponic box (150 mm diameter and 130 mm height) containing distilled water under 18 h/6 h light/dark conditions. Each hole was sown with 3 seedlings, and each box was sown with a density of 90 plants per box. After one week, the seedlings will be treated with modified Hoagland nutrient solution (pH 6.0–6.5). Per litre solution contains the following: 1.7 mM Ca (NO₃)₂·4H₂O, 2.0 mM KNO₃, 2.0 mM KCl, 1.7 mM CaCl₂, 0.5 mM KH₂PO₄, 0.5 mM MgSO₄·7H₂O, 11.6 µM H₃BO₃, 2.39 µM MnSO₄·4H₂O, 0.08 µM CuSO₄·5H₂O, 0.19 µM ZnSO₄·7H₂O, 12.5 µM FeSO₄·7H₂O, and 12.5 µM Na₂EDTA. The nutrient solution will be replaced every 7 days. When the cannabis seedlings reached the third-leaf stage, some seedlings were transferred to a low-nitrogen nutrient solution, while the others were kept under normal nitrogen conditions as controls. The normal nitrogen solution is 3.24 mmol L^− 1 and the low nitrogen solution is 1.08 mmol L^− 1. There are 4 treatments (Longma No.2 under normal nitrogen conditions (T_NN), Longma No.2 under low nitrogen conditions (T_LN), Hanma No.10 under normal nitrogen conditions (S_NN), Hanma No.10 under low nitrogen conditions (S_LN)). Each treatment has three biological repeats. After 2 h, the seedlings’ roots were sampled and rapidly frozen in liquid nitrogen and stored at -80℃ for the transcriptomic analysis by Majorbio (Shanghai, China).

Total RNA is extracted using the TransZol up Plus RNA kit (TransGen Biotech, China) from all tissue samples. The concentration and purity are detected by Nanodrop2000.

The RNA integrity is tested by agarose gel electrophoresis. RQN value is tested by Agilent5300. All mRNA is enriched by Oligo dT and is fragmentated into small fragments. The mRNA fragment is used to synthesize cDNA by the action of reverse transcriptase. Then the sticky ends of the double-stranded cDNA are repaired using End Repair Mix, and an A base is added at the 3’ end. Then, the adapter is added to the product for purification and fragment sorting. The sorted product is used for PCR amplification, and the purified product is obtained as the final library. Paired-end sequencing (2 × 150 bp) of the mRNA libraries was performed on an Illumina NovaSeq X Plus platform. Raw reads were processed using fastp (version 0.19.5) to remove adapter sequences, poly-N sequences, and low-quality bases (Q-score < 20). Reads shorter than 50 bp after trimming were discarded. The resulting high-quality clean reads were then aligned to the cannabis sativa reference genome (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_029168945.1/) using HISAT2 (version 2.1.0). Finally, gene-level expression quantification was performed based on the alignment results using RSEM (version 1.3.3). The raw sequencing data have been deposited in the NCBI SRA under accession number PRJNA1267284 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1267284).

Then, based on the expression levels and lengths of each gene, the FPKM (fragments per kilobase of transcript per million mapped reads) value (The calculation formula FPKM= (count/gene length)/ (total counts/1000000); count represents the number of reads mapped to a specific gene in a given sample; gene length represents the length of the gene (in base pairs); total counts represents the total number of reads for all genes in the sample) for each gene in each sample was calculated. Subsequently, TPM (transcripts per million) (The calculation formula TPM= (FPKM/∑FPKM)*1000000; FPKM represents the FPKM value calculated previously; ∑ FPKM is the sum of FPKM values for all genes in the sample) was calculated by normalising the FPKM values. To facilitate further analysis, duplicate numbers in sample names were removed, and the average TPM value for each gene was calculated. The data were analysed on the online tool of Majorbio Cloud Platform (https://cloud.majorbio.com/page/tools/).

Quantitative real-time PCR analysis

The qRT–PCR was adopted to analyse the NPF gene expression pattern. Primers were designed using Primer Premier 5 (Table S6). An RNAiso Plus kit (TaKaRa, Dalian, China) was used to isolate total RNA from nitrogen-treated cannabis roots. cDNA was synthesised from 1 µg of total RNA using a PrimeScript™ RT Reagent Kit with gDNA Eraser (TaKaRa, Dalian, China). A real-time PCR analysis (qRT–PCR) was performed using a One Step SYBR PrimeScript RT–PCR kit (TaKaRa, Dalian, China) with a CFX96 Touch™ Real-Time PCR Detection System (Bio-Rad, Berkeley, CA, USA).

The cycling protocol was as follows: 3 min at 95 °C, followed by 40 cycles of 15 s at 95 °C, 30 s at 55 °C, and 15 s at 72 °C. Normalisation was performed using the GAPDH gene. The 2^−∆∆Ct method was applied to calculate the relative expression level. Three biological replicates were used in the analysis. Tukey’s HSD test was used to assess the significant difference. The data were analysed, and an image was generated using “ggplot2, dplyr, multcomp, and emmeans” packages in R.

Results

Identification of CsNPF Genes in Cannabis sativa L.

Based on the 53 AtNPFsprotein queries, 70 NPF proteins (66 NPF genes) were identified from the cannabis sativa genome through BLAST search and conserved domain validation. The physicochemical properties of the NPF protein sequences are shown in Table S1. The NPF proteins had an average of 575 amino acid residues, ranging from 259aa (XP_060975047.1) to 681aa (XP_030487746.2), with a mean value of 63.92 kD and a difference of 46.36 kD between the maximum and minimum values. The isoelectric point (pI) ranged from 5.06 (XP_060974473.1) to 9.48 (XP_060975047.1), of which 14 were acidic proteins (pI less than seven) and 56 were basic proteins (pI higher than seven). The instability index ranged from 23.81 (XP_030480342.1) to 45.71 (XP_030491762.2); of these, 11 were greater than 40, indicating unstable proteins. The grand average of hydropathy (GRAVY) ranged from 0.142 (XP_030500572.1) to 0.556 (XP_030490259.2), indicating that all CsNPF proteins were hydrophobic proteins. Subcellular localisation predictions showed that the great majority of the CsNPF proteins are located at the plasma membrane, except XP_060963844.1, which is located in the vacuole. It can be inferred that the majority of NPF proteins are membrane proteins, while XP_060963844.1 is a vacuole membrane-integrated protein. These properties of CsNPF proteins can guarantee their stability within cell membranes and their activity and efficiency in transmembrane transport and signal transduction.

Phylogenetic analysis and chromosomal location of CsNPF proteins in Cannabis sativa L.

Based on the phylogenetic analysis, the NPF proteins from different species were classified into eight subfamilies (NPF1–NPF8). Notably, compared with other species, Cannabis sativa contained a significantly higher number of NPF genes, suggesting a potential expansion of this gene family. This expansion was primarily concentrated in the CsNPF5, CsNPF7, and CsNPF8 subfamilies, indicating possible subfamily-specific diversification and functional specialization during the evolution of Cannabis sativa. Among these subfamilies, CsNPF5 subfamily has the largest number of members, while CsNPF3 has the fewest (Fig. 1,Table S2). The CsNPFs, which are closely related to the AtNPFs shown in the phylogenetic tree, may share similar physiological functions in nitrate uptake and transport.

Fig. 1.

Phylogenetic analysis of the NPF gene family. In total, 338 protein sequences, including 53 from Arabidopsis thaliana,70 from cannabis sativa, 88 from Populus trichocarpa, 46 from Ricinus communis L., 81 from Vitis vinifera were involved in the analysis. The clades representing different subfamilies (NPF 1–8) are indicated by different colours and different shapes representing different species

All CsNPF proteins were renamed according to their chromosome locations. The NPF genes in cannabis sativa are widely distributed across the chromosomes, with Chr1, Chr7, and ChrX exhibiting the highest gene density (Fig. 2). As shown in Fig. 2, except for one CsNPF gene (CsNPF5.24) located in the unplaced genome scaffold, ASM2916894v1 Contig3. The other 65 CsNPF genes were heterogeneously located on 10 chromosomes of cannabis sativa. Chromosome (Chr) 7 contained 12 CsNPF genes, followed by 10 genes located in Chr1 and ChrX, and 7 genes located in Chr4 and Chr9, 6 genes located in Chr2, 5 genes located in Chr8, 4 genes located in Chr3, 3 genes located in Chr5, and 1 gene located in Chr6.

Fig. 2.

Chromosomal distribution of all CsNPF proteins

Structure, domain and conserved motif analyses

As shown in Fig. 3 that they all have the same conserved domain MFS, and ten conserved motifs (motifs one to ten) were identified from 70 CsNPF proteins. Among them, 60 CsNPFs have all ten motifs. This indicates that these CsNPF proteins are highly conserved. 10 CsNPF proteins, which belong to NPF5, possess a different number of motifs. This indicates these proteins may have a different function compared with other proteins.

Fig. 3.

Phylogenetic relationships, conserved protein motifs, domain analysis, and gene structure of CsNPF proteins. (a) Phylogenetic tree of the CsNPF proteins. (b) Conserved motif analysis of CsNPF proteins. Different colour boxes represent motifs 1–10. (c) Domain analyses of CsNPF proteins, with green colour indicating the NPF structural domain. (d) UTR and CDS structures of CsNPF proteins. Green squares indicate UTR structure. Yellow squares indicate CDS structures

Motif 4, 6, 7 and 8 can be identified from most of the CsNPF proteins. Motif 4 and motif 8 are present in 69 CsNPF proteins except CsNPF5.15.2. Motif 6 and motif 7 are all present in 68 CsNPF proteins except CsNPF5.10 and CsNPF5.11. Compared with other motifs, the above motifs may play a key role in the NPF gene family’s core function.

The examination of the exon-intron structure showed that all CsNPF genes had exons (ranging from one to six) and introns, except three genes (CsNPF5.10, CsNPF5.20, and CsNPF5.24) had no introns.

The NPF proteins exhibited a relatively high level of sequence conservation, particularly within the same subfamilies, as supported by above analysis.

Analysis of Cis-acting regulatory elements

To reveal the potential roles of NPF genes within regulatory networks, the cis-acting regulatory elements (CAREs) were analysed using the 2000 bp upstream promoter sequences of CsNPF genes in PlantCARE. The results showed that the promoter regions of these CsNPF subfamily genes were generally enriched in light responsiveness, abscisic acid responsiveness and anaerobic induction (Fig. 4, Table S3). Thus, it is hypothesised that the expression of CsNPF genes can be influenced by multiple signals, such as stress-responsive, light signals, and hormones.

Fig. 4.

Cis-acting regulatory element (CRE) analysis in the 2000 bp promoter regions of CsNPF genes

Synteny analysis of CsNPF genes

Synteny analysis can reveal the evolutionary history and functional conservation of gene families; we therefore investigated the syntenic relationships of CsNPF genes to explore their functions and regulatory mechanisms in cannabis and Arabidopsis.

Through synteny analysis, 6 pairs of syntenic genes (CsNPF 5.3/ CsNPF 5.4, CsNPF 5.1/ CsNPF 5.14, CsNPF 2.4/ CsNPF 2.6, CsNPF 4.2/ CsNPF 4.5, CsNPF 7.4/ CsNPF 7.7, CsNPF 5.20/ CsNPF 5.24) were identified among all CsNPF gene family (Fig. 5a).

Fig. 5.

Synteny analysis of CsNPF genes (a) Synteny analysis of CsNPF genes in cannabis sativa. Grey lines in the background indicate all syntenic blocks in the cannabis genome, and the black lines indicate syntenic NPF gene pairs. b Synteny analysis of NPF genes between cannabis and Arabidopsis. The grey lines indicate all syntenic blocks, and the red lines indicate syntenic NPF gene pairs

To determine these gene-duplication events, the Ka/Ks ratios between the duplicated genes were evaluated (Table S4). The Ka value for gene pairs CsNPF 5.3/CsNPF 5.4 is 0. They have the same sequences. However, one is on the positive strand of the chromosome, and the other is located on the negative strand of the same chromosome. It is hypothesised that they are functionally conserved paralogous genes that originated from ancient duplication events (e.g., WGD) and are retained due to their functional importance, with reverse orientation potentially reducing the risk of homologous recombination. The Ks value of gene pairs CsNPF 2.4/ CsNPF 2.6 is NaN. This was caused by sequence incomparability or can be explained by a very recent evolutionary duplication event (the absence of synonymous substitutions (Ks = 0) alongside the presence of nonsynonymous substitutions (Ka>0)). We discovered that the Ka/Ks ratios in the other 4 duplicated events were less than 0.50, indicating that the evolutionary process of these genes was predominantly driven by purifying selection. These gene pairs are highly conserved.

The syntenic map of cannabis relative to Arabidopsis was compared. Figure 5b shows that 27 NPF collinear gene pairs were identified between cannabis and Arabidopsis. Among 66 CsNPF genes, 18 CsNPF genes can be paired with Arabidopsis. The remaining CsNPF genes do not occur as syntenic gene pairs with Arabidopsis, suggesting that these genes may have been specific to cannabis during its evolution. CsNPF1.1.1, CsNPF2.6, CsNPF4.5, CsNPF4.2, CsNPF5.18, CsNPF5.22, CsNPF5.14, CsNPF6.5, and CsNPF8.2 correspond to two syntenic gene pairs in Arabidopsis, and it is hypothesised that these CsNPF genes may play significant roles in the evolution of the NPF gene family. The remaining 9 paired genes were found to be associated with one syntenic gene pair between cannabis and Arabidopsis, suggesting that these NPF genes may have evolved from the same ancient NPF genes.

Expression profiles of NPF genes under different nitrate concentrations

GO annotation of these CsNPFs genes mainly enriches the membrane of cellular components, transmembrane transporter activity in molecular functions, and transmembrane transport in biological processes (Fig. 6a).

Fig. 6.

(a) GO annotation of CsNPF genes. b Expression profiles of 62 CsNPF genes under different N treatments in cannabis roots. The expression data were obtained from RNA-seq data and are shown as TPM values. S_LN, low nitrogen treatments for sensitive variety; S_NN, normal nitrogen treatments for sensitive variety; T_NN, normal nitrogen treatments for tolerant variety; T_LN, low nitrogen treatments for tolerant variety

By analysing transcriptomic data from different nitrogen-tolerant varieties under low and normal nitrogen conditions, the results showed that none of the genes were significantly regulated. Among all the NPF genes, the expression level of CsNPF6.2 and CsNPF7.4 is far higher than that of other genes (Table S5). Most CsNPF genes have low expression levels, and the expression levels of NPF family members vary (Fig. 6b, Table S5). The relatively highly expressed genes from each subfamily were selected as our candidate genes for further analysis. The candidate genes include CsNPF1.2 (LOC115702148), CsNPF2.1 (LOC115708189), CsNPF3.1 (LOC115712669), CsNPF4.3 (LOC115714632), CsNPF5.14 (LOC115697513), CsNPF6.1 (LOC115714073), and CsNPF7.7 (LOC115710913).

The expression of selected genes

Among all candidate genes, the expression patterns of CsNPF5.14 (LOC115697513) and CsNPF4.3 (LOC115714632) may be involved in nitrogen utilisation. Under low nitrogen conditions, the expression of CsNPF5.14 decreased slightly in the sensitive variety, while it increased in the tolerant variety. Low nitrogen induces this gene expression in a low-nitrogen-tolerant variety. Under low nitrogen conditions, the expression of CsNPF4.3 decreased significantly in both varieties, while the gene expression level in the tolerant variety was far higher than that of the sensitive variety.

Compared with the normal nitrogen condition, the expression pattern changed. The two expression patterns above suggest that they may be involved in nitrogen uptake and primary transport, particularly in maintaining stable nitrogen levels under low-nitrogen conditions (Fig. 7).

Fig. 7.

Relative expression levels of CsNPF5.14 (LOC115697513) and CsNPF4.3 (LOC115714632) using qRT–PCR in different nitrogen-sensitive varieties’ roots. A significant difference between the different treatments was indicated by Tukey’s HSD test (* p < 0.05, ** p < 0.01, *** p < 0.001). The expression levels of each gene under normal nitrogen treatment in the nitrogen-sensitive variety were set as 1. LN represents low nitrogen, NN represents normal nitrogen

Discussion

Excessive nitrogen fertilisation has drawn widespread global attention. The application and promotion of low-nitrogen, high-efficiency varieties can reduce the nitrogen fertiliser dosage. It has been reported that NPF gene family members are involved in nitrogen high-efficiency utilisation. So far, the NPF gene family has been identified and characterised in a variety of plant species, such as spinach [23], cigar tobacco [24], Gossypium hirsutum [25], and cucumber [26]. However, there is no report on hemp.

Composition of NPF gene family in Cannabis sativa

The number of NPF gene families in different crops varies, for example, 57 NPF genes, 9 NRT2 genes, 1 NRT3 gene in spinach [23], 54 NPF genes in cucumber (Cucumis sativus L.) [26], 49 TaNRT2 genes in wheat [27], and 117 NPF genes in cigar tobacco [24]. Even within the same plant, the number of NPF gene family members varies across species; for example, 14 NRT2 in G. hirsutum, 14 NRT2 in G. barbadense, seven NRT2 in G. raimondii, and seven NRT2 in G. arboreum [6]. In Brassica napus, 31 NRT2 in ‘zhongshuang11’ and 19 in ‘Darmor-bzh’ were characterised [28].

Moreover, the selection of the reference genome will impact the result of gene family characterisation and expression. While using another cannabis sativa variety, CS10, as a reference genome, 96 NPF proteins can be detected. By selecting the genome sequence of pink pepper as a reference.70 NPF proteins have been identified from cannabis sativa. Phylogenetic analysis revealed that they could be grouped into eight clades, similar to NPF proteins from Arabidopsis. This conforms to one classification system of the NPF gene family. In the other reported classification system, the NPF gene family can be divided into 10 supergroups and 32 groups [26]. Recently, a new classification method was proposed for Brassica napus [28]. Most studies conformed to the first division, as did our study.

The composition of the NPF gene family is complicated. Through investigation on NPF subfamilies in 36 plant species, the results showed that subfamily members’ distribution varied between eudicots and monocots. It shows that eudicots had a higher proportion of the NPF1 and NPF2 subfamilies, whereas monocots had a higher proportion of the NPF3, NPF7, and NPF8 subfamilies [24]. While in our study, cannabis sativa possesses a higher proportion of NPF5, NPF2 and NPF8 (Table S2, Fig. 1). This is different from previous reports in eudicots. These subfamilies may perform additional functions in cannabis sativa and appear to be more active in duplication.

Evolutionary and structural characteristics of the Cannabis sativa NPF family genes

The NPF gene family in most crops has a high degree of structural conservation and similarity in their sequences. One aspect of gene family evolution is variation in gene structure. Compared with most structurally reasonably conserved CsNPF proteins, of NPF5 members with structural variation in NPF5 indicate functional variation (Fig. 3).

The quality of the reference genome impacts the accuracy of gene family identification. It is challenging for high-heterozygosity plant genome assembly, such as cannabis sativa, even if the sequencing and analysis technologies have advanced a lot. Such genome assembly is often accompanied by high variations in intron length and sequence. Assembly fragmentation can prevent genes containing long introns from being fully assembled, resulting in incomplete gene models. This may explain the irregular exon-intron structure in cannabis sativa (Fig. 3d).

Tandem and segmental duplications lead to gene family expansion. Tandem duplication occurs mainly in regions of chromosomal recombination. The gene family members formed by tandem duplication are usually closely located on the same chromosome, forming a cluster of genes with similar sequences and functions. While segmental duplication occurs in duplicated genes that are distant or even on distinct chromosomes. The synteny analysis indicates that all 6 syntenic gene pairs (Fig. 5b) are associated with segmental duplication events. This gene family is characterised by significant evolutionary stability, without evidence of large-scale expansion, suggesting it performs an essential and sensitive biological function.

CsNPF genes function critically in nitrogen uptake and utilisation of Cannabis sativa

It has been shown that NPF gene families are heavily involved in the complex processes of N utilisation across various plants. Especially in the model plant Arabidopsis, one review has summarised the roles of NPF nitrate transporters. And different members and subfamilies are responsible for different functions. AtNPF4.6 of the NPF4 subfamily in Arabidopsis, which is located in root hairs and root epidermis, exhibits nitrate uptake under low nitrogen environment and can be responsible for abscisic acid (ABA) transport. Most members of the NPF5 family, which are located in the vascular stele of roots and leaves, exhibit the function of vacuolar nitrate efflux [29]. This may explain our finding of candidate key genes in cannabis sativa. These findings provide theoretical support for our hypothesis.

A gene’s expression pattern is typically determined by its cis-acting regulatory elements. Our analysis of the CsNPF gene promoters revealed the presence of various cis-elements associated with light responsiveness, abscisic acid responsiveness, and anaerobic induction. Similarly, except for the nitrogen metabolism-related cis-elements, cis-elements associated with abscisic acid responsiveness and anaerobic induction have also been detected in cigar tobacco [24]. These results suggest that CsNPF transporters may have diverse functions beyond nitrogen response.

Overall, CsNPF genes play key roles in N uptake and utilisation in cannabis sativa and have diverse functions in regulating other metabolic activities. The identification and characterization of CsNPF genes in this study provides a foundation for future molecular breeding and functional genomics studies. The identified gene (CsNPF5.14 and CsNPF4.3) could be utilized for targeted gene editing (e.g., CRISPR/Cas systems), as well as overexpression and knockout experiments, to elucidate their biological functions and improve nitrogen use efficiency in Cannabis sativa. However, more experimental varieties and diverse nitrogen treatment times are needed to further investigate the nitrogen-responsive mechanism of the NPF gene family. Through algorithm optimisation and the development of bioinformatics tools, the genome assembly process must be continuously refined to improve genome quality. And more gene families may participate in the complex nitrogen utilisation process; bioinformatics methods, such as gene network analysis, may facilitate the identification of key regulatory gene networks.

Conclusions

70 CsNPF proteins have been identified from the reference genome of Cannabis sativa. These proteins can be classified into eight subfamilies based on phylogenetic analysis and are distributed across 10 chromosomes and one unplaced genome scaffold. Moreover, the results of gene structure, conserved motifs, and collinearity analyses reveal that the CsNPF genes have highly conserved structures. The promoter cis-acting element analysis showed that the promoter of CsNPFs contains a large number of light-responsive, stress-responsive, and hormone-responsive elements. The expression patterns of candidate CsNPFs screened from transcriptome data across different genotypes under low-nitrogen conditions have been verified by qRT–PCR. And there are two genes (CsNPF4.3 (LOC115714632), CsNPF5.14 (LOC115697513)) that may be involved in maintaining nitrogen absorption or utilisation under low nitrogen stress. Our study will facilitate research into the function of the CsNPF gene.

Abbreviations

NPF

Nitrate Transporter 1/Peptide Transporter

LN

Low nitrogen

CAREs

Cis-acting regulatory elements

ABA

Abscisic acid

TPM

Transcripts per million

FPKM

Fragments per kilobase of transcript per million mapped reads

qRT-PCR

Quantitative reverse transcription polymerase chain reaction

NUE

Nitrate use efficiency

Authors' contributions

ZY was major contributor in writing-original draft and conceptualisation, WHQ and XL were major contributor in conceptualisation, CHPY was major contributor in performing the experiments and investigate experiment data, WXN was major contributor in preparing resources, JV, SH and ZAMH were major contributors in writing, review and editing draft.

Funding

This work was supported by Heilongjiang Provincial finance basic operating expenses special [No. CZKYF2024-1-C033]. Scientific Research Fund of Heilongjiang Academy of Sciences [KY2023DQ01]. 2024 Annual Education and Scientific Research Project for Young and Middle-aged Teachers of Fujian Provincial Department of Education [JZ240089]. 2025 Annual Education and Scientific Research Project for Young and Middle-aged Teachers of Fujian Provincial Department of Education [JAT251288]. Integration and Promotion of the Technology System for the Seed Industry of Horticultural Crops in Fujian and Taiwan [ZZXPT11525014]. And Grant TAP-K007217.

Data availability

The data used to support the findings of this study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.