Title: Genome-Wide Survey and Expression Analysis of NIN-Like Protein (NLP) Genes Reveals Its Potential Roles in the Response to Nutrition Deciency in Tomato

Background: Tomato (Solanum lycopersicum) is one of the most important horticultural crops, with a marked preference of nitrate as inorganic nitrogen source. The molecular mechanisms of nitrate uptake and assimilation are poorly understood in tomato. NIN-Like Proteins (NLPs) are conserved, plant-specic transcription factors that play crucial roles in nitrate signaling. Results: In this study, genome-wide analysis revealed six NLP members in tomato genome. They were clustered into three clades in a phylogenic tree. Comparative genomic analysis showed that SlNLP genes had collinear relationships to NLPs in Arabidopsis, canola, maize and rice, and that the expansion of the SlNLP family mainly resulted from segmental duplications in tomato genome. Tissue-specic expression analysis showed that the close homologues of AtNLP6/7, SlNLP3, was strongly expressed in roots during both seedling and owering stages; SlNLP4 and SlNLP6 exhibited preferential expression in stems and leaves; and SlNLP6 were expressed in high levels in fruits. Further, the nitrate uptake in tomato roots and expression patterns of SlNLP genes were measured under nitrogen/phosphate/potassium deciency and nitrate resupply conditions. The transcript abundance of SlNLP3 decreased to 70% under phosphate/potassium deciency. Most of SlNLPs were up-regulated after nitrogen starvation. SlNLP1 and SlNLP5 were induced rapidly and temporally by nitrate. Conclusions: These results provided signicant insights into the potential diverse functions of SlNLPs to regulate nitrate uptake.


Background
Nitrogen (N), one of essential macro-nutrients for plants, serves as the component of amino acids, nucleotides, chlorophyll, hormones and co-enzymes. The growth and development of plants depends on proper nitrogen supply. And the availability of N in agricultural eld affects crop yields signi cantly (Miller and Cramer, 2005). Plants absorb inorganic N from the soils mainly in two forms, nitrate (NO 3 − ) and ammonium (NH 4 + ). Under mild climatic conditions, nitrate is the main nitrogen source in dry land (Forde and Clarkson, 1999). The concentration of nitrate in the soils uctuates between 10 μM to 100 mM (Crawford, 1995). To sustain vigorous growth, high-a nity and low-a nity transport systems have been evolved in plants to absorb nitrate e ciently from the environment. Nitrate is also one of important signaling molecules for lateral root development, owering and synergistic absorption of the other nutrients (Vidal et al., 2020).
As one of the most important crops, tomato (Solanum lycopersicum) shows a marked preference of nitrate as inorganic nitrogen source (Errebhi et al., 1990).
In the present study, comparative bioinformatics analysis of the tomato NLP genes was performed. Further, the rate of root nitrate uptake and expressions of SlNLP genes under nutrition de ciency and nitrate resupply conditions were detected to evaluate their potential roles in nitrate uptake regulation in roots.

Results
Identi cation of NLP Genes in tomato Chromosomal distribution and syntenic analysis of SlNLP genes Six SlNLP genes were distributed unevenly in tomato genome (Fig. 2). SlNLP3, SlNLP4 and SlNLP5 were identi ed on chromosomes 8. The other three SlNLP genes, SlNLP1, SlNLP2 and SlNLP6 genes were identi ed on chromosomes 1, 4 and 11, respectively. Inter-chromosomal relationship of SlNLP genes showed two pairs of segmental duplications (SlNLP1 and SlNLP2, SlNLP3 and SlNLP5), indicating that tomato NLP genes were mainly generated by gene duplication during evolution.
Further, four comparative syntenic maps between tomato and Arabidopsis, canola, rice and maize, were constructed, to analyze the phylogenetic mechanisms of SlNLPs (Fig. 3). Tomato SlNLP genes showed 10 syntenic gene pairs with canola, 8 with Arabidopsis, 5 with maize and 3 with rice. Most background collinear blocks associated with NLP gene pairs identi ed between tomato and dicotyledon Arabidopsis/canola contained more genes than those between tomato and monocotyledon rice/maize (Supplementary Table 2). SlNLP1, SlNLP2 and SlNLP5 were found in the four comparative syntenic maps, suggesting that these orthologous pairs might already exist before evolutional divergence of monocotyledon and dicotyledon, and these three genes might have played fundamental roles in NLP gene family. The ratio of non-synonymous (Ka) to synonymous substitutions (Ks), presenting the selection type acting on the coding sequences, were also calculated (Supplementary Table 2). Two SlNLP gene pairs, SlNLP1 and SlNLP2, SlNLP3 and SlNLP5, had Ka/Ks ratio of 1.01 and 1.46, respectively, indicating positive selection during evolution for functional divergence occurring after duplication. Most of the orthologous NLP gene pairs had a Ka/Ks ratio less than 1 (ranging from 0.10 to 0.96), suggesting purifying selective pressure during NLP gene family evolution and conserved functions of these genes. Three orthologous gene pairs, SlNLP1 and AtNLP5, SlNLP2 and BnaNLP4-4, SlNLP1 and ZmNLP1, had a Ka/Ks ratio more than 1, indicating they have underwent positive selection pressure and might be evolved with some new functions to cope with their living environments.

Organ-dependent expression of SlNLPs
To obtain evidence of physiological function, tissue-speci c transcript abundance of 6 SlNLP genes was analyzed by qRT-PCR at different developmental stages (Fig. 4). All of SlNLP genes had relatively low expression levels, 1/10000-4/100 of the level of internal control SlEF1α gene expression. SlNLP1 had the lowest expression. Therefore, SlNLP1 expression levels in roots or fruits were set to 1 for comparison of expression levels. At both the seedling and owering stages, SlNLP2 and SlNLP3 were preferentially expressed in roots ( Fig. 4A and 4B). SlNLP2 and SlNLP3 showed the highest transcript abundance in root at the seedling stage (Fig. 4A). When owering, SlNLP3 still showed the most abundance in roots, followed by SlNLP3 and SlNLP6 (Fig. 4B). At the owering stage, the transcript abundance of SlNLP4 and SlNLP6 increased signi cantly in all the test tissues, with preferential expression in stems and leaves. And SlNLP6 had highest transcript accumulation in leaves, stems and owers. Particularly, signi cantly higher expression of SlNLP6 was observed in fruits (Fig. 4C).
Expression of SlNLPs in response to nutrition de ciency Nitrate absorption in tomato roots were found to be in uenced by major mineral elements nutrition (nitrogen/phosphate/potassium) de ciency, indicated by 15 NO 3 in ux assay after different treatments ( Fig. 5). The results showed that the root high-a nity nitrate uptake ability was enhanced under nitrogen starvation, but repressed under potassium/phosphate starvation (Fig. 5A). And the root low-a nity nitrate uptake ability was enhanced under potassium starvation, but repressed under nitrogen/phosphate starvation (Fig. 5B).
To obtain evidence of possible roles of SlNLPs in root nitrate absorption regulation during nutrition de ciency, the transcript abundance of SlNLP genes in roots was examined by qRT-PCR after starvation treatments (Fig. 6). The expression of SlNLP1, SlNLP2, SlNLP4 and SlNLP6 were up-regulated for 6.2, 3.1, 17 and 1.5 times, respectively, after nitrogen starvation. In response to phosphate starvation, SlNLP3 showed expression decrease to 70% speci cally. And the expression level of SlNLP2, SlNLP3 and SlNLP6 decreased to around 70% in response to potassium starvation.

Nitrate-dependent expression of SlNLPs
Both the root high-a nity and low-a nity nitrate uptake rates were enhanced after nitrate resupply to the nitrogen-starved plants, showed by results of 15 NO 3 in ux assay (Fig. 7). The nitrate-dependent expression of SlNLP genes in roots were examined at 0.5 h, 1 h and 2 h after nitrate was resupplied to the starved seedlings. The results (Fig. 8) showed that the transcript abundance of SlNLP1 and SlNLP5 increased rapidly and temporally in response to nitrate. The expression of SlNLP1 and SlNLP5 reached the maximum levels, 4.1 and 2.8 times respectively, 0.5 h after nitrate was supplied. The expression of SlNLP2 and SlNLP4 was repressed signi cantly after nitrate resupply for 1 h. By contrast, SlNLP3 and SlNLP6 did not show any response to nitrate in transcription level.

Discussion
In the present study, genome-wide analysis revealed six tomato NLPs (Table 1). The Solanum lycopersicumNLP family size is similar with Arabidopsis thaliana (9), Oryza sativa (5) and Zea mays (9), much smaller than Brassica napus (31). Phylogenetic analysis showed that every NLP family has members belongs to three groups (Fig. 1A). All of SlNLPs has conversed RWP-RK and PB1 domains. SlNLP5 is special for double RWP-RK and PB1 domains (Fig. 1B). The expansion of tomato NLP gene family was mainly generated by gene duplication in genome (Fig. 2). Orthologous gene pairs associated with SlNLP1,SlNLP2 or SlNLP5 were indicated existence before the ancestral divergence of dicotyledonous and monocotyledonous plants (Fig. 3). It is worth noting that Ka/Ks ratio of two paralogous SlNLP gene pairs (SlNLP1 and SlNLP2, SlNLP3 and SlNLP5) and three orthologous NLP gene pairs (SlNLP1 and AtNLP5, SlNLP2 and BnaNLP4-4, SlNLP1 and ZmNLP1) were more than 1 (Supplementary Table 2), representing positive selection and fast evolutionary rates in these SlNLPs at the protein level. Therefore, it is implied that NLPs in tomato might evolve some new functions to meet their growth and development demands.
As one of fundamental regulatory elements at the transcriptional level, NLPs play important roles in nitrate uptake and assimilation regulation (Guan, 2017;Gaudinier et al., 2018). Tissue-dependent expression pattern showed that all 6 SlNLP genes were expressed in all tested tissues including roots, stems, leaves, owers and fruits (Fig. 4), which is similar with NLPs in Arabidopsis (Chardin et al., 2014), maize (Ge et al., 2018) and Brassica napus (Chardin et al., 2014). SlNLP3, one of the close homologues of AtNLP6/7 (Fig. 1A), the key component of nitrate signaling (Liu et al., 2017), has the highest expression level in roots at both seedling and owering stages. Besides SlNLP3, SlNLP2 and SlNLP6 were also expressed in high levels in roots, at different stages of development, implying their different functions in nitrate uptake regulation, rather than simple functional redundancy. Two SlNLPs from Clade III, SlNLP4 and SlNLP6, showed preferentially expressed in aboveground tissues and were strongly up-regulated in their transcription abundance when owering, suggesting that they might probably regulate nitrogen translocation and assimilation to support ower and fruit development. Different from SlNLP4, SlNLP6 had higher transcript abundance both in roots and aboveground tissues. What is more, SlNLP6 showed extremely higher expression level than all the other ve SlNLPs in fruits. The close homologue of SlNLP6 is AtNLP8 (Fig. 1A). AtNLP8 has been reported as a master regulator of nitrate-promoted seed germination (Yan et al., 2016), which might provide some hints for functional research on SlNLP6.
Nitrate is more favorable inorganic nitrogen source form for tomato. The nitrate uptake in tomato roots must be under precise regulation with complex interactions between nitrogen and the other essential macro-nutrients phosphate and/or potassium availability (Vidal et al., 2020). When environmental nitrogen source is depleted, the root low-a nity nitrate in ux rate decreased, but high-a nity nitrate in ux rate increased (Fig. 5). Similar results have been reported that higher nitrate in ux was detected in tomatoes growing in nutrient solutions containing 5 mM nitrate than 0.1 mM (Abenavoli et al., 2016). Both low-a nity and high-a nity nitrate uptake in roots increased after nitrate was resupplied to the nitrogen-starved tomato seedlings (Fig. 7). Distinct from nitrogen starvation, potassium de ciency led to enhanced low-a nity nitrate in ux rate and deceased high-a nity nitrate rate in roots (Fig. 5), which is reasonable because some published data show that strong expression increase of the nitrate transporters SlNRT1.2 and SlNRT2.1 had been induced by potassium deprivation (Wang et al., 2001). Slow-down of both low-a nity and high-a nity nitrate uptake rate were observed under phosphate de ciency (Fig. 5), which is consistent with the recent study in Arabidopsis .
In Arabidopsis, nlp7 mutants show features of a nitrogen-starved plant (Castaings et al., 2009); AtNLP7 overexpression increases plant biomass under both nitrogen-poor and -rich conditions (Yu et al., 2016). Expression of rice NLPs (OsNLP1, OsNLP4 and OsNLP5) was promoted by nitrogen de ciency as well as nitrate supply (Jagadhesan et al., 2020). Overexpression of OsNLP1 could enhance rice nitrogen use e ciency (Alfatih et al., 2020). Here, the transcript abundance of SlNLPs in roots has been detected under various nutrition conditions ( Fig. 6 and Fig. 8). Most of SlNLPs (SlNLP1, SlNLP2,SlNLP4 and SlNLP6) showed up-regulated expression after nitrogen starvation for 2 days. When nitrate was resupplied, the temporal expression of SlNLP2 and SlNLP4 was repressed, but SlNLP1 was still showed rapidly upregulated. One of the two close homologues of AtNLP6/7, SlNLP5, was induced rapidly and temporally by nitrate. However, the other close homologue of AtNLP6/7, SlNLP3, which showed the highest expression level in roots during both seedling and owering stages (Fig. 4), did not show any response to nitrate. It is noteworthy that AtNLP6/7 responds to nitrate signaling not in transcription level either (Liu et al., 2017). Under phosphate de ciency or potassium de ciency, SlNLP3 could be down-regulated in transcript abundance. After 2-days' phosphate starvation, SlNLP3 was the only SlNLP gene to show altered expression level, 70% of control. SlNLP3 also showed decreased transcript abundance to 70% after potassium starvation for 2 days, together with another two SlNLP genes, SlNLP2 and SlNLP6. Therefore, it is interesting to gure out how SlNLP3 participate in various nutrition de ciency signaling and/or nitrate signaling pathways.

Conclusions
In summary, this study provided genome-wide analysis of NLP genes in tomato. NLP genes are highly conserved among tomato, Arabidopsis, canola, maize and rice. Segmental duplication was the major driving force of SlNLP genes evolution. Some SlNLP genes had undergone positive selection during evolution, probably leading to functional divergence in gene family. The expression patterns of SlNLP genes provided hints for their diverse physiological roles in tomato growth and development, especially in nitrate uptake regulation. Further functional analysis for each SlNLP, especially SlNLP3 and SlNLP6, will be necessary to explore their regulatory functions. It is believed that a comprehensive understanding of the roles of SlNLP under uctuating nutrition conditions is an essential step towards deciphering the molecular mechanism of nitrogen utilization and promoting nitrogen use e ciency in tomato.  (Gasteiger et al., 2005). Subcellular localizations of SlNLP proteins were predicted using CropPAL2020 (https://www.croppal.org) (Hooper et al., 2020).

Multiple sequences alignment and phylogenetic analysis
Clustal W (version 2.1) was employed for the multiple sequences alignment and sequence identity matrix of the proteins (Larkin et al., 2007). Then, the deduced amino acid sequences in RWP-RK and PB1 domains were adjusted manually using GeneDoc software. Phylogenetic tree was constructed with MEGAX program (http://www.megasoftware.net/) using the Neighbor-Joining method. Proportions of amino acid differences were computed using Poisson correction distance to estimate evolutionary distance. The pairwise deletion option was used to circumvent the gaps and missing data. The conserved protein motifs of SlNLP proteins were analyzed using MEME server v5.3.0 (http://memesuite.org/tools/meme) (Bailey et al., 2015). The parameters for the search were as follows: max motif number to nd is 5 and min-max motif width to nd is 2-40. The matched motifs with low quality were manually removed based on an e-value of less than 1e −15. Sequences of NLP proteins of tomato (Solanum lycopersicum), Arabidopsis (Arabidopsis thaliana), canola (Brassica napus), rice (Oryza sativa) and maize (Zea mays) were downloaded from Phytozome (https://phytozome.jgi.doe.gov/).

Chromosomal distribution and gene duplication
All SlNLP genes were mapped to chromosomes based on physical location information using Circos (Krzywinski et al., 2009). Then, chromosome distribution was plotted with MapChart2.0 (https://mapchart.net/). The gene duplication events were analyzed using Multiple Collinearity Scan toolkit MCScanX. The syntenic analysis maps of orthologous NLP genes were constructed using the Dual Systeny Plotter software (https://github.com/CJ-Chen/TBtools) . Non-synonymous (Ka) and synonymous (Ks) substitution of each duplicated NLP genes were calculated using KaKs_Calculator 2.0 (Wang et al., 2010).

Plant materials and treatments
Tomato ecotype Micro-Tom was used in this study. The seeds were germinated and grown on vermiculite for 7 d before transferred to hydroponics.  (Zou et al., 2020). Tomato roots were washed in CaSO 4 for 1 min and then submerged in medium containing 1 mM or 0.1 mM K 15 NO 3 for 5 min. 15 N concentration was measured using an isotope ratio mass spectrometer (IRMS; DELTA plus XP).

Statistical analysis
Data were processed using the statistics program SPSS version 21. The statistical signi cance of differences in 15 N influx and gene expression was examined by student's t-test (*p < 0.05, **p < 0.01).

Declarations
Ethics approval and consent to participate The experimental research on plants performed in this study complies with institutional, national and international guidelines.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.