Genome-Wide Prediction and Analysis of Oryza Species NRP Genes in Rice Blast Resistance

Members of the N-rich proteins (NRPs) gene family play important roles in the plant endoplasmic reticulum stress in response, which can be triggered by plant pathogens’ infection. Previous studies of the NRP gene family have been limited to only a few plants, such as soybean and Arabidopsis thaliana. Thus, their evolutionary characteristics in the Oryza species and biological functions in rice defense against the pathogenic fungus Magnaporthe oryzae have remained unexplored. In the present study, we demonstrated that the NRP genes family may have originated in the early stages of plant evolution, and that they have been strongly conserved during the evolution of the Oryza species. Domain organization of NRPs was found to be highly conserved within but not between subgroups. OsNRP1, an NRP gene in the Oryza sativa japonica group, was specifically up-regulated during the early stages of rice-M. oryzae interactions-inhibited M. oryzae infection. Predicted protein-protein interaction networks and transcription-factor binding sites revealed a candidate interactor, bZIP50, which may be involved in OsNRP1-mediated rice resistance against M. oryzae infection. Taken together, our results established a basis for future studies of the NRP gene family and provided molecular insights into rice immune responses to M. oryzae.


Introduction
Rice is a crucial crop that is responsible for feeding more than half of the world's population [1]. However, a variety of extreme environmental conditions, (such as drought, salinity and extreme temperatures) negatively affect rice plant growth. Additionally, rice blast, which is caused by the hemi-biotrophic fungal pathogen Magnaporthe oryzae, can reduce rice yield by 30% [2].
To inhibit pathogens' invasion, host plants have evolved a two-layer immune system [3,4]. The first is determined by plant pattern recognition receptors (PRRs). Through recognizing pathogen-associated molecular patterns (PAMPs) of pathogens (such as PGN and chitin oligosaccharide), plant PRRs activate PAMP-triggered immunity (PTI) for the inhibition of pathogens' colonization. Pathogens can correspondingly release various effector proteins into host cells to evade or subvert host PTI defenses [5]. The second layer plant immune system is governed by resistance (R) proteins, which mainly have nucleotide-binding site leucine-rich repeat (NBS-LRR) domain architecture. Through directly or indirectly binding to avirulence (Avr) effectors of pathogens, R proteins are induced and activate a rapid response, also known as effector-triggered immunity (ETI) [6].
Rice resistance based on NBS-LRR proteins is often overcome within five years due to the rapidly evolving and highly variable effectors of blast fungus [7]. In contrast, multiple defense-regulator (DR) genes also confer partial but broad and durable resistance. Until now, dozens of DR genes were found that maintain this durable resistance by guaranteeing effective signal transduction and being responsible for downstream immune responses [8,9].
have reported the functions of NRP proteins in rice resistance to blast fungus. The roles and biological functions of NRP proteins in other rice defense responses also remain unclear. The present study aimed to shed light on the evolutionary history, putative interaction partners, expression patterns, and biological functions of NRPs in rice resistance. The results revealed conserved evolution of the NRP protein family and synergistic differentiation between Oryza species. We also found that OsNRP1, an NRP gene in the Oryza sativa japonica group, was specifically up-regulated at the early infection stage and enhanced rice resistance to M. oryzae attack. Furthermore, OsNRP1 was also predicted to interact with IRE1, bZIP TFs and NAC-domain containing proteins. Our findings provide a basis for further investigation of NRP protein functions and regulatory mechanisms in rice resistance.

Whole-Genome Characterization and Phylogenic Analysis of NRP Genes in Oryza Species
Two strategies were independently applied for the identification of NRP genes in Oryza species. First, a hidden Markov model approach (HMM) was used to search against 12 Oryza species and soybean proteomics, which were retrieved from the Ensembl plant database (https://plants.ensembl.org/, (accessed on 5 July 2022)), based on the NRP domain HMM model file (Pfam accession: PF10539). Second, 26 characterized plant NRP genes were used as queries in BLASTP searches against the 13 proteomes using an e-value cutoff of 1 × 10 −5 . To further verify the reliability of candidates NRP genes, the Conserved Domains Database (CDD) tool was used to validated the completeness of the NRP domain in each candidate. A total of 136 complete, non-redundant NRP genes were detected in 12 Oryza species and soybean (Table 1), with each Oryza species containing between nine and 13 homologs, and 15 in soybean (Table S1 and Figure 1a). This demonstrated wide and conserved distribution of NRP genes among Oryza species. We investigate the evolutionary history of NRP genes in Oryza species by constructing a maximum likelihood (ML) phylogenetic tree (Figure 1b). A total of 136 putative NRP genes in the Oryza species and soybean were identified. These genes were categorized into four subgroups (I-IV). NRP genes from each Oryza species and soybean were present in all four groups, suggesting that four ancestral NRP subgroups divided in the most recent common ancestor (MRCA) of Oryza species and soybean. No clustered pairs of NRP genes were detected in any Oryza species, indicating a lack of species-specific duplication events. In contrast, paralogous pairs of G. max NRP genes in each subgroup revealed that this family underwent several duplication events in soybean. After scanning the domain architecture of each protein, we found tandem Kelch motifs located in C-terminus of all subgroup II members except OINRP3, GmNRP1, and GmNRP11. The preference of C-terminus location  (a) Species tree of 12 Oryza species and soybean and distribution of NRP proteins. The species tree was constructed using the maximum likelihood method with 1000 bootstraps based on a concatenated alignment of housekeeping genes identified by CEGMA analysis. Numbers of NRP protein were listed on the right side of the species tree. (b) Phylogenic tree of NRP proteins identified in 12 Oryza species and soybean. Maximum likelihood tree, with 1000 bootstraps (values displayed per branch). NRP proteins identified in Oryza sativa japonica group are marked in red.

Synteny Analysis of the NRP Gene Family in Oryza Species
There were 12, 13, 11, 11, and 15 NRP genes in the O. sativa japonica group, the O. sativa indica group, Oryza punctata, Oryza barthii, and Glycine max, respectively. To further investigate the orthologous relationships, a synteny analysis was conducted for NRP genes in these species (Figure 2a and Table S2). The O. sativa japonica group genome shared 10, 11, and 11 syntenic NRP genes with the O. sativa indica group, O. barthii, and O. punctata, respectively. The O. sativa indica group genome had 11 and 10 syntenic genes with O. barthii and O. punctata, respectively, and there were 11 syntenic NRP genes between O. barthii and O. punctata. These results indicated a high level of conservation of NRP gene family after differentiation of Oryza species. In contrast, only GmNRP4 and GmNRP12 showed a syntenic relationship with corresponding genes in Oryza species, suggesting more extensive divergence between NRP genes in soybean and Oryza species. An intraspecific synteny analysis of the O. sativa japonica group revealed that only OsNRP3 and OsNRP8 had a syntenic relationship (Figure 2b), indicating that there was no species-specific expansion of the NRP gene family. These results together with the phylogenic analysis ( Figure 1b) indicated that the putative duplication occurred in the MRCA of Oryza species.
The ratio of nonsynonymous (Ka) to synonymous (Ks) nucleotide substitution rates is a common measurement used to distinguish between selective pressures on protein-coding genes and to assess their evolutionary rates [26]. When Ka/Ks ≈ 0, a gene is considered to be under neutral selection, whereas Ka/Ks < 1 indicates purifying selection and if Ka/Ks > 1 indicates positive selection [27]. Ka/Ks value of the syntenic NRP gene pairs detected above were nearly all <1 (Table S2), suggesting that the NRP gene families in these species had evolved under purifying selection. However, the syntenic gene pair OsNRP4/OsinNRP6 (which was also syntenic with OsNRP2) had a Ka/Ks ratio of 1.79, revealing that OsNRP4 may have been under positive selection and may have a novel function.

NRP Genes in the O. sativa japonica Group were Differentially Expressed during M. oryzae Infection
To analyze the expression patterns of O. sativa japonica group NRP genes (OsNRPs) during M. oryzae infection, we calculated expression values using the data from our previously RNA sequencing (RNA-seq) [28]. In that study, O. sativa L. ssp. japonica cv. 'Nipponbare' (Nip) plants were infected with three M. oryzae strains (A248, B235, and C162). Seven OsNRP genes were found to be significantly up-regulated during the interactions between Nip and M. oryzae. As Figure 3a showed, OsNRP2, OsNRP3, OsNRP4, OsNRP5 and Os-NRP8 were significantly up-regulated at 24 h post infection (hpi). In addition, OsNRP1 and OsNRP6 were significantly up-regulated at 8 hpi with M. oryzae A248 or B235. Due to our rice the genetic transformation system relied on wild-type cultivar 'Zhonghua 11' (ZH11), which was used as the background cultivar. We performed quantitative reverse transcription (qRT)-PCR assays to validate the expression patterns of seven OsNRPs candidates during interactions between ZH11 and Guy11 ( Figure 3b). OsNRP2, OsNRP3, and OsNRP4 were down-regulated during infection. OsNRP6 and OsNRP8 were up-regulated at 8, 48, 72, and 96 hpi, whereas OsNRP6 was up-regulated at 24, 72, and 96 hpi. Notably, OsNRP1 (LOC_Os01g36950) was specifically up-regulated at eight hpi but was barely detected at other timepoints. These results suggested that OsNRP1 may play a role in the plant defense response at 8 hpi.

OsNRP1, a NRP Protein in the O. sativa japonica Group, may Enhance Blast Fungus Resistance
The high expression levels of OsNRP1 during the early infection stage suggested that OsNRP1 may be involved in blast resistance. To validate this hypothesis, a transgenic OsNRP1-overexpression line (OsNRP1 OX ) was constructed. We obtained a total of 22 independent transgenic T 1 lines. Through qRT-PCR validation of each T 1 lines, we found that OsNRP1 was expressed at significantly level (around four-to 12-fold higher than in the wild-type) in three T 1 lines: PXQ8-4, PXQ8-5, and PXQ8-17 ( Figure 4d). These three lines were selected for inoculation assays by M. oryzae Guy11. Seven-days after in vitro inoculation, the total lesion length on ZH11 was 2.7 cm (1.5 cm + 1.2 cm), which was significantly longer than the total lesion lengths on the OsNRP1 OX plant (1.5, 1.4, and 1.7 cm on PXQ8-4, PXQ8-5, and PXQ8-17, respectively) ( Figure 4a). Similar results were observed after in vivo inoculation; the diseased leaf area was significantly smaller on PXQ8-4, PXQ8-5 and PXQ8-17 than on ZH11 (Figure 4b,c). These results suggested that OsNRP1 might be activated by blast pathogen infection and could enhance rice resistance against M. oryzae.

Predicted Protein-Protein Interaction (PPI) Network of OsNRP1
NRPs require interacting partners to activate the NRP-mediated cell-death signaling pathway [20]. To further analyze the mechanisms of OsNRP1-mediated resistance to rice blast, a predicted interaction network was constructed for OsNRP1 using STRING website ( Figure 5a). We identified 10 candidate proteins that may interact with OsNRP1 (Table S3), including a bZIP TF (bZIP50), Serine/threonine-protein kinase (IRE1), Zinc finger domaincontaining stress-associated proteins (SAP16, SAP3 and SAP5), a NAC domain-containing protein (Q5Z7Q4), and cold shock domain-containing protein (CSP1). We also investigated the PPI network of an OsNRP1 ortholog in Arabidopsis thaliana, AT5G42050 (Figure 5b). Notably, a bZIP TF (bZIP60) and NAC domain-containing proteins (NAC062 and NAC036) were also detected in that network.
Gene Ontology (GO) annotation enrichment analysis was conducted for the putative OsNRP1 interaction partners. This analysis revealed that 'Response to stress', 'Response to organic substance', and 'Endoplasmic reticulum unfolded protein response' were the major enriched terms, which contributed to understanding of the NRP-mediated response to rice blast. For example, soybean NRPs and their orthologs in Arabidopsis were found to be induced by endoplasmic reticulum (ER) stress, and triggered PCD response [12]. Ref. [29] reported Arabidopsis thaliana NRP genes were up-regulated in response to cold and drought stress, uncovering in their contribution in signal transduction. Thus, we inferred that OsNRP1 may enhance rice defense through contributing to mediate ER stress.  OsNRP1 may be involved in blast resistance. To validate this hypothesis, a transgenic Os-NRP1-overexpression line (OsNRP1 OX ) was constructed. We obtained a total of 22 independent transgenic T1 lines. Through qRT-PCR validation of each T1 lines, we found that OsNRP1 was expressed at significantly level (around four-to 12-fold higher than in the wild-type) in three T1 lines: PXQ8-4, PXQ8-5, and PXQ8-17 ( Figure 4d). These three lines were selected for inoculation assays by M. oryzae Guy11. Seven-days after in vitro inoculation, the total lesion length on ZH11 was 2.7 cm (1.5 cm + 1.2 cm), which was significantly longer than the total lesion lengths on the OsNRP1 OX plant (1.5, 1.4, and 1.7 cm on PXQ8-4, PXQ8-5, and PXQ8-17, respectively) ( Figure 4a). Similar results were observed after in vivo inoculation; the diseased leaf area was significantly smaller on PXQ8-4, PXQ8-5 and PXQ8-17 than on ZH11 (Figure 4b,c). These results suggested that OsNRP1 might be activated by blast pathogen infection and could enhance rice resistance against M. oryzae. ( Figure 5a). We identified 10 candidate proteins that may interact with OsNRP1 (Table  S3), including a bZIP TF (bZIP50), Serine/threonine-protein kinase (IRE1), Zinc finger domain-containing stress-associated proteins (SAP16, SAP3 and SAP5), a NAC domain-containing protein (Q5Z7Q4), and cold shock domain-containing protein (CSP1). We also investigated the PPI network of an OsNRP1 ortholog in Arabidopsis thaliana, AT5G42050 (Figure 5b). Notably, a bZIP TF (bZIP60) and NAC domain-containing proteins (NAC062 and NAC036) were also detected in that network.

Investigation of Cis-Element within OsNRP1 Promoter and Conserved Motifs in NRP Genes
OsNRP5, OsNRP2, OsNRP8, OsNRP3, OsNRP4, OsNRP1, and OsNRP6 were found to be significantly up-regulated during interactions with rice blast fungus. To investigate the regulatory mechanisms associated with the up-regulated OsNRPs, we analyzed predicted cis-elements within their promoter regions (defined as the 1500-bp region upstream of the transcription start site) (Figure 6a). In total, 175 cis-elements were detected (Table S4); of the 20 most abundant (Table 2), 8 were likely to be involved in abiotic or biotic stress responses. For example, WRKY71OS and WBOXNTERF3 are putative binding sites of WRKY proteins, suggesting that DCD/NRP genes may be required for WRKY-mediated regulation of plant defense against pathogens (Table 2). In addition, CGCGBOXAT, the binding site of bZIP proteins, was also present in the promoters of NRP genes. This result was consistent with PPI network predictions, which suggested that bZIP proteins maybe important components of the NRP-mediated cell-death signaling pathway. Moreover, the cis-elements OSE2ROOTNODULE and GT1GMSCAM4, which are involved in activation of pathogen-and salt-induced genes, were present; MYBCORE, ACGTATERD1, and ACGTATERD1 were found, and may participate in responses to diverse abiotic stressors, such as dehydration These findings suggested that DCD/NRP genes were involved not only in abiotic stress but also in plant pathogen defense, during which multiple TFs play important roles.  We also used MEME to predict conserved motifs among NPRs in Oryza species. In total, five conserved motifs that formed the NRP domain were found. For example, FGLP and PLxE motifs, the typical motifs of NRP proteins, were displayed as Motif_1 and Mo-tif_5, which are present at the N-and C-terminus of NRPs, respectively, and are composed of sequences that conserve the secondary structure [17]. Thus, conserved motif organization seems to shape the NRP domain.

Discussion
The NRPs are a family of genes-encoding proteins that contain the NRP domain; this domain is asparagine-rich, and proteins containing it are therefore called N-rich proteins (NRPs) [17]. Ref. [16] first discovered that soybean NRP genes were induced when inoculated by Pseudomonas syringae pv. Glycinea, which contain the avirulence gene avrA [16]. However, genomic identification of NRP gene family members has been limited to only a few plants, such as Arabidopsis and soybean [29,30].
Common cultivated rice (O. sativa) is one of the most essential crops for food security, but its production is threatened by rice blast. Rice breeding for pathogen resistance has historically depended on a narrow range of genetic diversity, but other Oryza species may provide a broader range of genetic resources for breeding pathogen resistance in rice [31]. Previous reports clarified that soybean NRP proteins participate in ER stress so that soybean resistance is enhanced against several pathogens, such as Pseudomonas syringae pv. We also used MEME to predict conserved motifs among NPRs in Oryza species. In total, five conserved motifs that formed the NRP domain were found. For example, FGLP and PLxE motifs, the typical motifs of NRP proteins, were displayed as Motif_1 and Motif_5, which are present at the N-and C-terminus of NRPs, respectively, and are composed of sequences that conserve the secondary structure [17]. Thus, conserved motif organization seems to shape the NRP domain.

Discussion
The NRPs are a family of genes-encoding proteins that contain the NRP domain; this domain is asparagine-rich, and proteins containing it are therefore called N-rich proteins (NRPs) [17]. Ref. [16] first discovered that soybean NRP genes were induced when inoculated by Pseudomonas syringae pv. Glycinea, which contain the avirulence gene avrA [16]. However, genomic identification of NRP gene family members has been limited to only a few plants, such as Arabidopsis and soybean [29,30].
Common cultivated rice (O. sativa) is one of the most essential crops for food security, but its production is threatened by rice blast. Rice breeding for pathogen resistance has historically depended on a narrow range of genetic diversity, but other Oryza species may provide a broader range of genetic resources for breeding pathogen resistance in rice [31]. Previous reports clarified that soybean NRP proteins participate in ER stress so that soybean resistance is enhanced against several pathogens, such as Pseudomonas syringae pv. Glycinea and Phytophthora sojae Race 1 [16,17]. Meanwhile, the NRP domain is only present in the plant kingdom and conserved region. Therefore, we inferred that NRP proteins may also play a role in rice resistance against blast fungal, which is still unclear. To this end, we performed a genome-wide analysis of the NRP gene repertoires in 12 Oryza species; these analyses were designed to shed light on the evolutionary histories of these genes and to infer putative biological functions in rice resistance against rice blast.
As Figure 7 shows, we here conducted a hidden Markov model (HMM) search of NRP genes in 12 different Oryza species, yielding a total of 136 NRP genes. The number of NRP genes in each species ranged from nine in O. brachyantha to thirteen in the O. sativa indica group, indicating a wide distribution of NRP genes in Oryza species. Fifteen NRP genes were identified in soybean, which was more than were found in the Oryza species. This may have resulted from the previously proposed retention of extended blocks of duplicated genes in the soybean genome [32]. There was no obvious evidence of NRP gene family expansion in Oryza species. To clarify the phylogenic relationships between Oryza NRP genes, we constructed a phylogenetic tree from NRP genes in soybean and Oryza species. The NRP genes segregated into four subgroups, reminiscent of the results obtained in a previous analysis of soybean and Arabidopsis NRP genes [12]. No subgroup-specific gene family expansion was observed. Each subgroup contained members from all Oryza species included in the analysis. Taken together, these findings indicated that NRPs originated from the MRCA of soybean and Oryza species and were highly conserved during evolution.
Patterns of NRP protein domain organization were highly conserved within subgroups, but more diversified across subgroups. For example, the DCD/NRP domain of NRP proteins in subgroup III and I were preferentially found in the N-and C-, respectively. NRP proteins in subgroup II also contained Kelch motifs; these are present in the Arabidopsis Kelch repeat-containing F-box (KFB) protein SAGL1 and negatively regulate salicylic acid (SA) biosynthesis during immune responses [33]. Highly conserved syntenic NRP genes were also observed in NRPs in four different Oryza species, which was in contrast to results from the other Oryza species and soybean. Thus, the evolutionary trajectories of NRP genes appear to be influenced by species differentiation. In the O. sativa japonica group genome, only OsNRP3 and OsNRP8 showed a syntenic relationship, implying a putative paralogous relationship between OsNRP3 and OsNRP8. This result also supports the finding that NRP gene duplication events rarely occurred in Oryza species. In addition, the Ka/Ks value of OsNRP3/OsNRP8 suggested that these genes have undergone positive selection, which may have played a role in biological functional divergence throughout evolution. Patterns of NRP protein domain organization were highly conserved within subgroups, but more diversified across subgroups. For example, the DCD/NRP domain of NRP proteins in subgroup III and I were preferentially found in the N-and C-, respectively. NRP proteins in subgroup II also contained Kelch motifs; these are present in the Arabidopsis Kelch repeat-containing F-box (KFB) protein SAGL1 and negatively regulate salicylic acid (SA) biosynthesis during immune responses [33]. Highly conserved syntenic NRP genes were also observed in NRPs in four different Oryza species, which was in contrast to results from the other Oryza species and soybean. Thus, the evolutionary trajectories of NRP genes appear to be influenced by species differentiation. In the O. sativa japonica group genome, only OsNRP3 and OsNRP8 showed a syntenic relationship, implying a putative paralogous relationship between OsNRP3 and OsNRP8. This result also supports the finding that NRP gene duplication events rarely occurred in Oryza species. In addition, the Ka/Ks value of OsNRP3/OsNRP8 suggested that these genes have undergone positive OsNRP1 was found that specifically up-regulated at an early stage during O. sativa japonica group interacting with three M. oryzae strains proposed by [28]. Inoculation assays of overexpression transgenic line (OsNRP1 OX ) revealed that OsNRP1 slowed rice disease. A similar phenomenon was also verified in soybean. For example, GmNRP-A and GmNRP-B, the NRP proteins in soybean, was induced by the binding protein (BiP) under activation of salicylic acid (SA) signaling, and mediated programmed cell death [34].
We identified 10 candidate protein-protein interactors with OsNRP1. These interactors were found that enriched in GO terms such as 'Response to stress', 'Endoplasmic reticulum unfolded protein response', and 'Response to organic substance', which indicate that Os-NRP1 may play a role in these biological processes. Interestingly, NRPs have been proposed to act as mediators of ER-and osmotic-stress-induced cell death in soybean [19]. ER capacity on protein procession is required for activation of the plant immune response, which relies on ER stress networks [11]. Notably, IRE1 is a critical ER stress sensor/transducer in Arabidopsis [35]. The TF bZIP50 was predicted as a potential interaction partner of OsNRP1. Interestingly, OsbZIP50 is regulated by IRE1-mediated splicing and is required to regulate known marker genes involved in ER stress after pathogen recognition [36]. This result suggested that OsNRP1 may inhibit M. oryzae infection through involvement in regulation of ER stress. We also established a protein-protein interaction network for AT5G42050, an Arabidopsis ortholog of OsNRP1; the Arabidopsis protein was predicted to interact with bZIP60, the Arabidopsis ortholog of OsbZIP50 [37]. We also found that OsNRP1 may interact with three stress-associated proteins (SAP16, SAP5, and SAP3) that have been shown to participate in multiple abiotic stress responses [38]. OsSAP1, another stress-associated protein, enhances disease resistance against virulent bacterial pathogens [39]. In addition, TF binding-site prediction revealed the presence of the CGCGBOXAT motif in the OsNRP1 promoter. This suggested that OsNRP1 was downstream of bZIP TFs, indirectly supporting the potential OsNRP1-bZIP50 interaction.
In this study, we report genome-wide analysis of NRPs across 11 Oryza species. Our results demonstrated conserved evolution of the NRP gene family, although the domain organization exhibits divergence at the subgroup level. OsNRP1, an NRP gene in the O. sativa japonica group, was found to be specifically up-regulated during the early stage of the rice-M. oryzae interaction and to inhibit M. oryzae infection. Moreover, predicted proteinprotein interaction networks and TF binding site predictions indicated that OsNRP1 may be downstream of bZIP TF, and may participate in the ER stress response after M. oryzae infection. Taken together, this study revealed OsNRP1-enhanced rice resistance against M. oryzae infection. Due to the conserved domain region and evolutionary history of NRP proteins in Oryza genus, this result gave an insight that OsNRP1 can provide partial but durable rice resistance against blast fungal, which increases our understanding of biological functions of NRP genes in Oryza genus.

Sequence Retrieval and Identification of NRP Proteins
The whole protein sequences of 12 Oryza species and soybean were downloaded from Ensembl Genomes (release 97; http://www.ensembl.org (accessed on 5 June 2022)) and were integrated into an initial data set for homolog identification. The hidden Markov model (HMM) file of the NRP domain was obtained from PFAM database (PF10539) [40], which was used to screen the whole proteome with cutoff E-value of 1 × 10 −5 . In parallell, known soybean NRP proteins, reported by [17], were defined as query sequences and applied BLASTP search against whole proteomes of species mentioned above (E-value cutoff = 1 × 10 −5 ). Then, total putative NRP candidates identified by two approaches were validated by the Conserved Domains Database (CDD) database and manually removed redundant sequences.

Phylogenic and Evolution Analysis of NRP Family in Oryza Species
MUSCLE version 3.8.31 (Mill Valley, USA) was used to apply multiple alignment analysis of obtained NRP proteins [41]. The maximum-likelihood (ML) phylogeny trees of NRP proteins in species of this study were built with IQ-TREE version 1.6.12 (Vienna, Austria) with automatic selection of optimal model for protein substitution [42]. A bootstrap analysis with 1000 replicates was conducted to evaluate the reliability of the tree. The visualization and modification of phylogenetic trees were performed using the iTOL server [43].

Synteny Analysis
Genome annotations of species in this study were retrieved from Ensembl Genomes website. We used DIAMOND v0.8.25 (Tübingen, Germany) to conduct all-vs-all comparisons of corresponding protein sequences (-max-target-seqs 5 -E-value 1 × 10 −5 ) [44]. Genome annotations and DIAMOND output file were input into MCScanX for synteny detection with default parameters [45]. The visualization of synteny results was performed by TBtools toolkit (Guangdong, China) [46].

Construction of OsNRP1 OX Transgenic Lines
We amplified coding sequence of OsNRP1 (LOC_Os01g36950) using 2X Phata master mix (Vazyme Biotech Co., Ltd., Nanjing, China). The amplified coding sequences were cloned into the rice transformation PXQ vector. The integrated construct PXQ::OsNRP1 was introduced into Agrobacterium strain EHA105 and then transformed into wild type (Zhonghua11, ZH11). Hygromycin-containing media were used to screen transgenic plants (40 mg/L).

Inoculation Assays
Two-week-old seedlings and leaf strips of wild-type (ZH11) and overexpression transgenic lines of OsNRP1, OsNRP1 OX (PXQ8-4, PXQ8-5 and PXQ8-17) were inoculated by Magnaporthe oryzae strains Guy11. 5 mL conidia suspension were sprayed onto leaves of each seedling (5 × 10 5 spores/mL), and 5 uL of Guy11 spore suspension was added to the wounds of each leaf strip. The disease symptoms were assessed seven days after inoculation. Leaves of ZH11 at 0, 8, 24, 48, 72 and 96 hpi were collected for real-time RT-PCR validation.

Real-Time RT-PCR
Qiagen RNAeasy Mini kit (Qiagen Inc., Valencia, CA, USA) was used to isolate total RNA from collected inoculation ZH11 leaves. isolated RNA samples were then converted into cDNA by the Superscript IV Reverse transcriptase cDNA synthesis kit (TB Green ® Premix Ex Taq TM II, Takara Bio Inc, Kusatsu, Japan). We diluted a 20-µL aliquot of cDNA to 100 µL with water, which was used to established real-time RT-PCR reactions. cDNA sample of 0 hpi was used as control samples, and the actin gene of rice (LOC_Os03g50885) was used as an internal reference gene with a stable expression pattern as 10 housekeeping genes proposed by [47]. Bio-Rad Real-Time PCR cycler was applied for the relative gene expression level estimation using SYBG as the fluorescent dye. All primers used to real-time RT-PCR reactions were designed in Primer3 website (https://bioinfo.ut.ee/primer3-0.4.0/ (accessed on 14 June 2022)).

PPI Networks Prediction and Cis-Element Analysis
Protein sequence of OsNRP1 was submitted to the online server STRING (version 11.5; http://string-db.org (accessed on 23 June 2022)), with specified organism as "Oryza. sativa japonica group". Online BLAST search was used to interacting partners by detecting hits with the highest scores (Bitscore), of which Gene Ontology (GO) annotations were also provided and displayed by TBtools [46]. The cis-element of 8 highly-expressed OsNRPs in infection stage were predicted through the PLACE website (http://www.dna.affrc.go. jp/PLACE/ (accessed on 29 June 2022)), and their annotations as well. The locations of cis-elements were displayed by TBtools toolkit (Guangdong, China) [46].

Analyzing Conserved Motifs of NRP Proteins in Oryza Species
We performed multiple alignment of NRP protein sequences in Oryza species mentioned above by MUSCLE version 3.8.31 (Mill Valley, USA) with default parameters. Conserved motifs in NRP proteins were identified using MEME 5.0.5 according to the following parameters: -protein, -nmotifs 10, -minw 8, and -maxw 80 [48].