Molecular Analysis of MgO Nanoparticle-Induced Immunity against Fusarium Wilt in Tomato

Fusarium wilt, caused by Fusarium oxysporum f. sp. lycopersici (FOL), is a devastating soilborne disease in tomatoes. Magnesium oxide nanoparticles (MgO NPs) induce strong immunity against Fusarium wilt in tomatoes. However, the mechanisms underlying this immunity remain poorly understood. Comparative transcriptome analysis and microscopy of tomato roots were performed to determine the mechanism of MgO NP-induced immunity against FOL. Eight transcriptomes were prepared from tomato roots treated under eight different conditions. Differentially expressed genes were compared among the transcriptomes. The Kyoto Encyclopedia of Genes and Genomes enrichment analysis revealed that in tomato roots pretreated with MgO NPs, Rcr3 encoding apoplastic protease and RbohD encoding NADPH oxidase were upregulated when challenge-inoculated with FOL. The gene encoding glycine-rich protein 4 (SlGRP4) was chosen for further analysis. SlGRP4 was rapidly transcribed in roots pretreated with MgO NPs and inoculated with FOL. Immunomicroscopy analysis showed that SlGRP4 accumulated in the cell walls of epidermal and vascular vessel cells of roots pretreated with MgO NPs, but upon FOL inoculation, SlGRP4 further accumulated in the cell walls of cortical tissues within 48 h. The results provide new insights into the probable mechanisms of MgO NP-induced tomato immunity against Fusarium wilt.


Introduction
Global yield losses of important staple crops to pathogens and pests can reach 30%, with estimated costs to the global economy owing to lost food production in the hundreds of billions of dollars [1]. Therefore, control of plant diseases is an important and essential technology for staple crop production. Although various plant disease control technologies have been devised and developed, chemical fungicides are the most effective, labor-saving, and easy to use [2]. However, the large-scale use of chemical fungicides has caused problems that cannot be ignored, such as adverse effects on human health, emergence of fungicide-resistant microbes, burden on ecosystems, and food contamination owing to residual chemical fungicides [3]. Therefore, Integrated Pest Management (IPM), a pest management system that reduces the use of chemical fungicides by combining various pest control technologies to minimize risks to food and environmental safety, is being promoted internationally [4].
To accomplish IPM, it is important to develop pest control technologies other than chemical fungicides that have less impact on the environment. In this respect, the induction of disease resistance by plant immunity inducers, which suppress disease by activating the plant's innate immune mechanism, is a promising disease control technology. In addition to their low environmental impact, plant immunity inducers have the advantage that there is almost no risk of the emergence of fungicide-resistant microbes, which is a problem with chemical fungicides [5]. Many plant immunity inducers, either chemically synthesized [5] or biologically derived [6], have been reported. Plant growth-promoting microorganisms (PGPR) found in the rhizosphere and root symbionts of plants are also known to induce plant immunity and have been applied in biocontrol [7].
Despite numerous reports, few plant immunity inducers have been put into practical use because of the large amount of time and expense required to develop plant immunity activators. In particular, there are very few plant immunity inducers developed and commercialized for vegetable diseases [8]. However, some known compounds and natural products, such as silicon [9] and chitosan [10], which are safe for humans and the environment, activate plant immunity. If these compounds can be applied for disease control, it would be advantageous for practical use because the development cost would be lower than that of new plant immunity activators. We found that magnesium oxide nanoparticles (MgO NPs), which are widely used as raw materials for fertilizers, soil conditioners, pharmaceuticals, and cosmetics, have an excellent plant immunity activation ability and have the potential to suppress bacterial wilt [11] and Fusarium wilt [12] in tomatoes.
Metal monoxides such as MgO NPs are known to produce reactive oxygen species (ROS) by interacting with biomolecules such as proteins, enzymes, and amino acids [13]. However, little is known about ROS production through interactions with plant-derived biomolecules. Imada et al. [11] showed that ROS are produced by the interaction of MgO NPs with plant polyphenols and speculated that the ROS generated in tomato roots treated with MgO NPs may be phenoxyl radicals produced by the deprotonation of phenolic hydroxyl radicals. In contrast, in Arabidopsis roots, ROS generation by MgO NPs has been suggested to involve NADPH oxidase (Rboh D), which acts in plant disease resistance responses [14].
MgO NPs are solid base catalysts of crystalline structured particles characterized by many edge/corner structures and reactive surface structures with a large surface area owing to their minute size [15]. MgO NPs with different surface structures can also be obtained depending on the calcination temperature [16]. We found that MgO NPs obtained by calcination at 700 • C induced significant ROS production in tomato and Arabidopsis roots [11,12,14]. It has been suggested that MgO NPs obtained by calcination at 700 • C have different surface structural characteristics from MgO NPs calcined at other temperatures (550, 900, 1000 • C), which may be involved in the interaction between MgO NPs and biomolecules [17].
Fusarium wilt of tomato is a soilborne disease caused by the filamentous fungus Fusarium oxysporum f. sp. lycopersici (FOL) that leads to significant economic losses to tomato production [18]. FOL survives for several years or more in the soil by forming thickwalled resting spores called chlamydospores, which germinate and invade the host through their roots and colonize xylems once the host is cultivated. The characteristic wilt symptoms appear as a result of severe water stress, mainly owing to vessel clogging caused by the growth of FOL. Wilting is most likely caused by a combination of pathogenic activities, such as the accumulation of fungal mycelia and/or toxin production [19]. Currently, control methods against Fusarium wilt mainly consist of soil fumigation with chemical fungicides and the use of resistant cultivars. Fumigation negatively affects the environment and human health. In addition, resistant varieties are losing their effectiveness owing to the continuous emergence of FOLs that overcome resistance. Therefore, it is highly desirable to establish alternative Fusarium wilt-control technologies. Biocontrol is a promising control method against Fusarium wilt, but it has the disadvantage that its effectiveness varies with the season, crop, and field [20].
In our previous report, we showed that, in tomato roots treated with MgO NPs, ROS are rapidly produced at the root tips where the FOL mycelium penetrates and at the vascular bundles through which FOL passes before entering the root xylem [12]. ROS strengthen the host cell wall [21] and serve as signals that mediate the activation of defense genes [22]. Therefore, ROS-mediated immunity against FOL may have been established in MgO NP-treated tomato roots before FOL infection. In fact, tomato seedlings irrigated at the base with a MgO NP suspension and grown for 7 days (pretreatment) did not develop Fusarium wilt until fruit harvest [12]. Moreover, because ROS play an important role in the adaptation process of plants to abiotic stresses [23], MgO NPs are suggested to be an abiotic stress for tomato roots. Furthermore, FOL infection is a biotic stressor in the tomato roots. Therefore, when tomato roots pretreated with MgO NPs were challenge-inoculated with FOL, they were subjected to both abiotic and biotic stresses. However, it is not known how tomato roots respond to these stressful conditions.
In this study, we aimed to determine the response of tomato roots to MgO NPs (shortand long-term treatments) as a nonbiotic stress and to FOL infection (with and without MgO NPs pretreatment) as a biotic stress. Transcriptome analysis was performed on tomato roots treated under eight different conditions. We selected the cell wall-localized glycinerich protein (SlGRP4) gene from the genes whose expression was significantly upregulated in tomato roots pretreated with MgO NPs and challenge-inoculated with FOL as revealed by the transcriptome analysis, and immunostaining analysis confirmed the accumulation of SlGRP4 in the root tissue. In addition, we also performed microscopic observations to reveal the presence or absence of FOL proliferation in the roots of tomato plants pretreated with MgO NPs.

RNA-seq Data Analysis
For transcriptome analysis, RNA was prepared from the roots of the following tomato plants and subjected to RNA sequencing: (i) those treated with distilled water for 1 h (H 2 O1h), (ii) those treated with MgO NPs for 1 h (MgO1h), (iii) those treated with distilled water for 7 days (H 2 O7d), (iv) those treated with MgO NPs for 7 days (MgO7d), (v) those treated with distilled water for 7 days and challenge-inoculated with FOL for 1 h (H 2 O7d + FOL1h), (vi) those treated with MgO NPs for 7 days and challenge-inoculated with FOL for 1 h (MgO7d + FOL1h), (vii) those treated with distilled water for 7 days and challenge-inoculated with FOL for 21 days (H 2 O7d + FOL21d), and (viii) those treated with MgO NPs for 7 days and challenge-inoculated with FOL for 21 days (MgO7d + FOL21d). After removing ambiguous nucleotides, 217,840,780 raw reads were obtained from the eight cDNA libraries. More than 96% of the high-quality reads in each sample could be mapped to the Solanum lycopersicum reference genome. vs. H 2 O7d + FOL21d"), 5923 DEGs with 3641 upregulated and 2282 downregulated genes were identified. In tomato roots treated with MgO NPs for 7 days and challenge-inoculated with FOL for 21 days ("MgO7d vs. MgO7d + FOL21d"), 4685 DEGs with 2557 upregulated and 2128 downregulated genes were identified. Expressed genes after FOL inoculation for 1 h with ("MgO7d + FOL1h") and without ("H 2 O7d + FOL1h") MgO NP pretreatment were also analyzed for DEGs. In this comparison group, 4078 DEGs with 2093 upregulated and 1985 downregulated genes were identified (Figure 2A). A comparison of genes whose expression changed significantly in response to FOL inoculation (1 h) with and without MgO NPs pretreatment revealed 716 genes specific to MgO NP-pretreated tomato roots ( Figure 2B).  (Figure 2A). A comparison of genes whose expression changed significantly in response to FOL inoculation (1 h) with and without MgO NPs pretreatment revealed 716 genes specific to MgO NP-pretreated tomato roots ( Figure 2B).
For GO analysis, the DEGs in the "H 2 O1h vs. MgO1h" group were classified into 151 functional terms, namely, 84 terms in biological process (BP), 9 in chemical component (CC), and 58 in molecular function (MF). The top 10 GO terms enriched for the upregulated genes in the BP category are listed in Table S1. The total DEGs in the "H 2 O7d vs. MgO7d" group were classified into 251 functional terms, namely, 116 terms in BP, 30 in CC, and 105 in MF. The top 10 GO terms enriched for the upregulated genes in the BP category are listed in Table Table S7. The "H 2 O1h vs. MgO1h" comparison group showed enrichment of GO terms related to terpenoid and isoprenoid biosynthesis and metabolic process (Table S1), whereas the other six comparison groups were enriched in similar GO terms (Tables S2-S7).

Quantitative Real-Time PCR (qRT-PCR) Analysis
To verify the accuracy of RNA-seq in tomato plants, four DEGs were randomly selected for qRT-PCR analysis using specific primers. The results showed that the expression patterns of the four genes detected using qRT-PCR are consistent with the RNA-seq results ( Figure 5).

Analysis of DEGs
The top 100 DEGs in each of the seven comparative groups are shown in the Supplementary Materials (Tables S8-S14).

Quantitative Real-Time PCR (qRT-PCR) Analysis
To verify the accuracy of RNA-seq in tomato plants, four DEGs were randomly selected for qRT-PCR analysis using specific primers. The results showed that the expression patterns of the four genes detected using qRT-PCR are consistent with the RNA-seq results ( Figure 5). To identify marker genes characterizing MgO NP-induced immunity against Fusarium wilt, we searched for genes that showed remarkable upregulation when MgO-pretreated tomato plants were inoculated with FOL compared to those before FOL inoculation. Four genes (Solyc03g098010.2, encoding acid phosphatase; Solyc08g006850.2, encoding a patatinlike phospholipase family protein; Solyc10g075150.1, encoding a nonspecific lipid-transfer protein; and Solyc09g097760.2, encoding a glycine-rich protein) showed such traits (Figure 6, asterisk). Among those genes, Solyc09g097760.2 encoding glycine-rich protein (accession XP004248032.1; gene name: SlGRP4 [24]), which has been suggested to be involved in Fusarium wilt resistance in tomato [25], was chosen for further analysis.

Quantitative Real-Time PCR (qRT-PCR) Analysis
To verify the accuracy of RNA-seq in tomato plants, four DEGs were randomly selected for qRT-PCR analysis using specific primers. The results showed that the expression patterns of the four genes detected using qRT-PCR are consistent with the RNA-seq results ( Figure 5).

Analysis of DEGs
The top 100 DEGs in each of the seven comparative groups are shown in the Supplementary Materials (Tables S8-S14).

Expression Analysis of the SlGRP4 Gene
The transcript levels of SlGRP4 were examined in the seven comparative groups. SlGRP4 gene was differentially expressed with upregulation in the three comparative groups: "H2O1h vs. MgO1h" (response of roots to MgO NPs) (log2FC = 4.020), "MgO7d vs. MgO7d + FOL1h" (response of MgO NP-pretreated roots to FOL) (log2FC = 4.686), and
To determine the expression pattern of SlGRP4 in tomato roots pretreated with MgO NPs and challenge-inoculated with FOL, qRT-PCR was conducted with specific primers. In MgO NP-pretreated tomato roots, SlGRP4 was highly expressed immediately (30 min) after FOL inoculation and decreased at 1 h. However, it then increased again at 3 h and 24 h and decreased at 72 h. In other words, multiple peaks of increased expression of SlGRP4 were observed when MgO NP-pretreated roots were inoculated with FOL ( Figure 7). To determine the expression pattern of SlGRP4 in tomato roots pretreated with MgO NPs and challenge-inoculated with FOL, qRT-PCR was conducted with specific primers. In MgO NP-pretreated tomato roots, SlGRP4 was highly expressed immediately (30 min) after FOL inoculation and decreased at 1 h. However, it then increased again at 3 h and 24 h and decreased at 72 h. In other words, multiple peaks of increased expression of SlGRP4 were observed when MgO NP-pretreated roots were inoculated with FOL ( Figure 7).

Prediction of Cis-Acting Elements in the SlGRP4 Gene
To understand the possible regulatory mechanisms of SlGRP4, the cis-acting elements encompassing 2000 bp 5′ upstream of the translation start site of the gene were analyzed using PlantCARE (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 10 December 2022). Thirty-five cis-acting elements potentially involved in the regulation of SlGRP4 were identified, including those that respond to phytohormones, such as abscisic acids, salicylic acid, and methyl jasmonate, and those recognized by MYC and MYB transcription factors ( Figure S9).

Localization of SlGRP4 in Tomato Roots
To localize the SlGRP4 protein expressed in tomato root tissues, immunostaining of tomato roots was performed using an anti-SlGRP4 antibody. SlGRP4 protein was detected in the cell walls of epidermal and vascular vessel cells of tomato roots treated with MgO NPs for 7 days (pretreatment only) ( Figure 8C). In the root pretreated with MgO NPs and challenge-inoculated with FOL, SlGRP4 protein heavily accumulated in the cell walls of cortical tissue, epidermal tissue, and vascular vessel cells ( Figure 8D). No accumulation of SlGRP4 protein was observed in tomato roots that were not treated with MgO NPs

Prediction of Cis-Acting Elements in the SlGRP4 Gene
To understand the possible regulatory mechanisms of SlGRP4, the cis-acting elements encompassing 2000 bp 5 upstream of the translation start site of the gene were analyzed using PlantCARE (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 10 December 2022). Thirty-five cis-acting elements potentially involved in the regulation of SlGRP4 were identified, including those that respond to phytohormones, such as abscisic acids, salicylic acid, and methyl jasmonate, and those recognized by MYC and MYB transcription factors ( Figure S9).

Localization of SlGRP4 in Tomato Roots
To localize the SlGRP4 protein expressed in tomato root tissues, immunostaining of tomato roots was performed using an anti-SlGRP4 antibody. SlGRP4 protein was detected in the cell walls of epidermal and vascular vessel cells of tomato roots treated with MgO NPs for 7 days (pretreatment only) ( Figure 8C). In the root pretreated with MgO NPs and challenge-inoculated with FOL, SlGRP4 protein heavily accumulated in the cell walls of cortical tissue, epidermal tissue, and vascular vessel cells ( Figure 8D). No accumulation of SlGRP4 protein was observed in tomato roots that were not treated with MgO NPs followed by FOL inoculation (Figure 8B). In the root pretreated with MgO NPs and challenge-inoculated with FOL, reactive oxygen species (ROS) were rapidly produced in cortical tissue and vascular vessel cells ( Figure 8H).

Microscopic Observation of Tomato Root Tissue Pretreated with MgO NPs and Challenge-Inoculated with FOL
Tomato plants treated with and without MgO NPs were inoculated with FOL and grown for 21 days. Similar to the results reported in a previous study [12], tomato plants pretreated with MgO NPs were completely disease-free, even 21 days after FOL inoculation. In contrast, all tomato plants without MgO NPs pretreatment showed yellowing and wilting of leaves. Microscopy analysis revealed mycelial masses of FOL in the main roots and stems of the tomato plants that were not pretreated with MgO NPs (Figure 7, arrows). In addition, dark brown discoloration of vascular bundles, which is one of the symptoms of Fusarium wilt, was observed in the stems (Figure 9, asterisk). Few mycelia of FOL were observed in the roots and stems, and no discoloration of vascular bundles in the stem was observed in tomato plants pretreated with MgO NPs.

Microscopic Observation of Tomato Root Tissue Pretreated with MgO NPs and Challenge-Inoculated with FOL
Tomato plants treated with and without MgO NPs were inoculated with FOL and grown for 21 days. Similar to the results reported in a previous study [12], tomato plants pretreated with MgO NPs were completely disease-free, even 21 days after FOL inoculation. In contrast, all tomato plants without MgO NPs pretreatment showed yellowing and wilting of leaves. Microscopy analysis revealed mycelial masses of FOL in the main roots and stems of the tomato plants that were not pretreated with MgO NPs (Figure 7, arrows). In addition, dark brown discoloration of vascular bundles, which is one of the symptoms of Fusarium wilt, was observed in the stems (Figure 9, asterisk). Few mycelia of FOL were observed in the roots and stems, and no discoloration of vascular bundles in the stem was observed in tomato plants pretreated with MgO NPs.

Discussion
In our previous report, we showed that when a MgO NPs suspension was drenched into the roots of tomato seedlings, ROS were produced in the tomato root tissue immediately after irrigation. In addition, tomato seedlings grown for 7 days after drenching (MgO NPs pretreatment) showed long-term immunity to FOL up to harvest, and the JA signaling pathway is essential for induction of this immunity [12]. However, gene expression in tomato roots immediately after MgO NPs treatment and in MgO NP-pretreated tomato roots is not known. We also did not know the gene expression patterns when MgO NPpretreated tomato roots were inoculated with FOL, or the gene expression patterns in tomato roots with established immunity to FOL. Furthermore, how immunity to FOL is established in tomato roots is also unknown. This study aims to clarify these issues.

Discussion
In our previous report, we showed that when a MgO NPs suspension was drenched into the roots of tomato seedlings, ROS were produced in the tomato root tissue immediately after irrigation. In addition, tomato seedlings grown for 7 days after drenching (MgO NPs pretreatment) showed long-term immunity to FOL up to harvest, and the JA signaling pathway is essential for induction of this immunity [12]. However, gene expression in tomato roots immediately after MgO NPs treatment and in MgO NP-pretreated tomato roots is not known. We also did not know the gene expression patterns when MgO NPpretreated tomato roots were inoculated with FOL, or the gene expression patterns in tomato roots with established immunity to FOL. Furthermore, how immunity to FOL is established in tomato roots is also unknown. This study aims to clarify these issues.

Response of Tomato Roots to MgO NPs
In this study, KEGG enrichment analysis for "MAPK signaling pathway-plant" of DEGs in the comparison group "H 2 O1h vs. MgO1h" showed that ROS-induced serine/threonine kinase OXI1, a serine/threonine kinase induced by ROS [26], was upregulated. OXI1 is required for the full activation of MPK3 and MPK6; OXI1 is an essential part of the signal transduction pathway linking ROS signals to diverse downstream responses [26]. This suggests that OXI1 induced by ROS is involved in the upregulation of MPK3 and MPK6, which was shown in the roots 1 h after MgO NPs treatment, and further triggers downstream responses in which MPK3/6 acts. Interestingly, the gene (Rboh) for the enzyme (NADPH oxidase) that generates ROS [27] was not upregulated in "H 2 O1h vs. MgO1h". However, the CNGCs gene, which triggers Ca 2+ influx that activates Rboh [28], was upregulated, suggesting that Ca 2+ influx may have activated Rboh and induced a ROS burst in the roots treated with MgO NPs. Among the DEGs of roots that were treated with MgO NPs for 1 h ("H 2 O1h vs. MgO1h") (Table S8), Solyc10g081740.1 encoding "calcium dependent protein kinase 1," which is upregulated depending on Ca 2+ [29], and Solyc03g113940.2 encoding "calmodulin-binding protein 60-like G (CBP60g)" [30], which is involved in Ca 2+ signaling in tomato plants [31], also suggest a Ca 2+ influx in roots treated with MgO NPs. GO enrichment analysis of DEGs in "H 2 O1h vs. MgO1h" showed enrichment of GO terms involved in terpenoid production and metabolism. Because terpenoids mitigate the effects of oxidative stress by modulating the oxidative status of plants [32], a mitigation system for oxidative stress induced by MgO NPs would be activated in roots treated with MgO NPs.
In roots after 7 days of MgO NPs treatment (i.e., MgO NPs pretreatment: "H 2 O7d vs. MgO7d" group), genes involved in "Pathogen attack" and "Reactive oxygen species" pathways were downregulated in the "MAPK signaling pathway-plant" (Figure 3, blue frame). In addition, many genes involved in "Plant hormone signal transduction" and "Plant-pathogen interaction" pathways were downregulated. In contrast, Rcr3, which encodes a papain-like apoplastic protease, was upregulated. Rcr3 (required for Cladosporium resistance-3) is a secreted papain-like cysteine protease of tomato that acts as a coreceptor for the Cf-2 resistance protein to detect Avr2, a secreted apoplastic effector of the fungal pathogen Cladosporium fulvum [33]. Recognition of Avr2 by Cf-2 results in localized programmed cell death known as the hypersensitivity response (HR) [33]. Rcr3 is upregulated by salicylic acid (SA) [34]. In fact, KEGG enrichment pathway "plant hormone signal transduction" analysis showed that the SA pathway was upregulated only in the "H 2 O7d vs. MgO7d" group among the seven comparison groups in this study, suggesting the production of SA in tomato roots pretreated with MgO NPs.

Response of Tomato Roots to Fusarium oxysporum
Many pathways and genes in the three KEGG enrichment pathways, "MAPK signaling pathway-plant", "Plant hormone signal transduction", and "Plant-pathogen interaction" were enriched in both the "H 2 O + FOL1h" and "MgO + FOL1h" groups. However, in "MAPK signaling pathway-plant" pathways, tomato roots pretreated with MgO NPs followed by FOL inoculation for 1 h ("MgO7d + FOL1h") showed that only MAPK genes were upregulated among MAPKKK, MAPKK, and MAPK genes, whereas tomato roots pretreated with H 2 O followed by FOL inoculation ("H 2 O7d vs. H 2 O7d + FOL1h" group) showed that MAPKKK as well as MAPK genes were upregulated (Figure 3, red frame). As mentioned above, MAP3/6 genes were upregulated by the OXI1 transcription factor. OXI1 was not upregulated in tomato roots pretreated with MgO NPs followed by FOL inoculation for 1 h, suggesting that OXI1-independent mechanisms for upregulation of MAPK3/6 genes were induced by MgO NPs pretreatment in the roots. In tomato roots pretreated with MgO NPs followed by FOL inoculation for 1 h ("MgO7d + FOL1h"), both Rboh encoding NADPH oxidase and Rcr3 encoding a papain-like apoplastic protease were upregulated. In contrast, no upregulation of these genes was observed in tomato roots pretreated with H 2 O followed by FOL inoculation for 1 h ("H 2 O7d + FOL1h"). Corroborating the elevated expression of Rboh, ROS were rapidly produced in cortical and vascular tissues in roots pretreated with MgO NPs in response to FOL inoculation ( Figure 8H); no ROS were detected when roots that had not been pretreated with MgO NPs were inoculated with FOL, suggesting that roots pretreated with MgO NPs were hypersensitive to FOL infection. The roles of NADPH oxidase and Rcr3 in suppressing FOL infection in tomato roots are unknown; however, both are involved in extracellular immunity [35]. Thus, their rapid expression within 1 h of inoculation seems to be important for extracellular immunity (apoplastic immunity) in MgO NP-induced immunity against FOL in tomatoes.
In tomato roots with no MgO NPs pretreatment 21 days after FOL inoculation, FOL had reached the stem xylems, and tomato plants showed symptoms of Fusarium wilt. In other words, the "H 2 O7d vs. H 2 O7d + FOL21d" group represents DEGs in the susceptible response. On the other hand, the "MgO7d vs. MgO7d + FOL21d" group represents DEGs in which MgO NPs-induced immunity has been established.
Comparing the two groups, the "Pathogen attack" pathway was enriched in the "MAPK signaling pathway-plant" in the "H 2 O7d vs. H 2 O7d + FOL21d" group (susceptible response). MAP3/6 and OXI1 were also upregulated ( Figure 3). In addition, SA-inducible PR1 was upregulated in the "Plant hormone signal transduction" pathways ( Figure S8), suggesting that SA is produced in the roots with no MgO NPs pretreatment 21 days after FOL inoculation. This result is consistent with a report by Hernandez-Aparico et al. [36] that the activation of SA signaling occurs in susceptible responses. In contrast, in "MgO7d vs. MgO7d + FOL21d" (resistance response), MAP3/6 and OXI1 were not upregulated in the "MAPK signaling pathway-plant" pathways, but the apoplastic protease Rcr3 was upregulated. This suggests the involvement of apoplastic immunity in the MgO NPinduced immunity against FOL.

SlGRP4
In the present study, we revealed that the SlGRP gene was rapidly upregulated within 1 h when tomato roots were treated with MgO NPs and when tomato roots pretreated with MgO NPs were inoculated with FOL, suggesting that SlGRP is upregulated in response to both abiotic (MgO NPs) and biotic (FOL) stresses. Cis-acting element analysis of the SlGRP gene revealed 35 cis-acting elements, including those that respond to phytohormones such as abscisic acids, SA, and methyl jasmonate, and those recognized by MYC and MYB transcription factors. We concluded that the SlGRP gene is involved in many stress responses. Increased GRP expression has been reported in response to abiotic stresses, such as salt, cold, heat, and injury, as well as biotic stress, including fungal infection [37].
GRPs are characterized by high glycine content and the presence of conserved segments including glycine-containing structural motifs of repetitive amino acid residues [38] and are classified into five classes (I to V) based on their general structure, taking into consideration the arrangement of the glycine repeats as well as the presence of conserved motifs [39]. The predicted amino acid sequence of SlGRP contains the Class V feature GGX/GXGX (where G represents glycine and X any amino acid). However, the function of class V GRPs is currently unknown.
GRPs localized in the cell walls of potato vascular bundles have been suggested to be involved in resistance to Ralstonia solanacearum, which causes vascular wilt disease such as FOL [24]. In the present study, it was shown that SlGRP4 accumulated in the cell walls of tomato vascular bundles and epidermal cells after MgO NPs treatment. Furthermore, inoculation of tomato roots pretreated with MgO NPs with FOL resulted in rapid and significant SlGRP4 upregulation, demonstrating that SlGRP4 protein accumulates in the cell walls of root cortex cells in addition to the cell walls of tomato vascular bundles and epidermal cells. After FOL invades tomato roots, the mycelium extends into the intercellular space of the cortex to reach the xylem vessels, during which many host defense reactions were observed [40]. Thus, SlGRP4 may play an important role in this process. In fact, in tomato roots pretreated with MgO NPs and challenge-inoculated by FOL, FOL mycelia were barely found in the tissues (Figure 9), suggesting that FOL infection of the root cortex and subsequent mycelial extension into xylems was strongly inhibited in tomato plants pretreated with MgO NPs even 21 days after FOL inoculation. However, because FOL is re-isolated from the roots of such plants [12], it is expected to be present and survive on the root surface. Further studies are warranted to examine the colonization of FOL on the surface of tomato roots pretreated with MgO NPs using GFP-labeled FOL.

Nanoparticles and Fusarium Wilt
The suppression of Fusarium wilt by nanoparticles, other than MgO NPs, has been demonstrated in metal oxide NPs such as CuO [41], ZnO [42], CaO [43], and Fe2O3 [44], and in metal NPs including silver [45] and manganese [46]. ROS induction after interaction with NPs has also been observed stably across plant species [47]. ROS induced by NPs acts as a signal for secondary metabolism, and secondary metabolites can act as antimicrobial compounds to protect plants from pathogenic microbes [47]. In the present study, GO enrichment analysis of DEGs in "H 2 O1h vs. MgO1h" showed that GO terms involved in the production and metabolism of terpenoids, which are materials for phytoalexins, were enriched (Table S1), suggesting that secondary metabolites accumulated in tomato roots pretreated with MgO NPs. This also suggests that tomato roots pretreated with MgO NPs may have already been ready to produce phytoalexins and other antimicrobial compounds before inoculation with FOL.
Present study revealed that rapid ROS production and SlGRP accumulation were induced in cortical tissue and vascular vessel cells of the roots pretreated with MgO NPs when challenge-inoculated with FOL (Figure 8), suggesting that ROS production may be associated with GRP accumulation. It is not known whether such rapid ROS production occurs in the roots pretreated with other NPs when challenge-inoculated with FOL. It is also unclear whether the "apoplastic immunity" suggested in the present study also occurs in other NPs-induced immunity against Fusarium wilt. Further studies are warranted to elucidate these points, which will provide important insights into Fusarium wilt resistance in tomato.

Plant Material
Tomato plants ('Momotaro'; Takii Seed, Kyoto, Japan) were grown in small pots (diameter, 9 cm; height, 7.6 cm) containing a vermiculite and perlite (1:1) mixture at 25 • C under a 12 h photoperiod with a photon flux density of 100 µmol·m −2 ·s −1 for 30 days. The 30-day-old tomato plants were subjected to MgO NP treatment, FOL inoculation experiments, and RNA sequencing, as described below.

Fungal Strains and Growth Conditions
The pathogen causing Fusarium wilt in tomatoes, FOL CK3-1 [48], was obtained from Dr. Takashi Tsuge (Nagoya University, Nagoya, Japan). A pure FOL culture was maintained on potato dextrose agar plates at 25 • C for 7 days. After incubation, fungal hyphal tips were obtained from the growing edge using a sterile sharp needle and transferred to 100 mL of potato dextrose broth (PDB). The broth was incubated at 25 • C with shaking at 120 rpm for seven days. The samples were filtered through three layers of gauze cloth and the filtrate was centrifuged at 3000× g for 5 min. The pellets were washed thrice and suspended in 20 mL of sterile distilled water. Conidial spore load was calculated using a hemocytometer and adjusted to 1 × 10 6 conidia/mL for inoculation.

Inoculation Experiments
For treatment with MgO NPs, tomato plant pots were drenched by irrigation with 50 mL of 1% MgO NPs (KF-37) suspension or water (control), after which the plants were left to grow for seven days under the conditions described in Section 4.2. (MgO NPs pretreatment). Following the MgO NPs treatment, tomato roots were carefully washed with water to remove all surface MgO NPs. For FOL inoculation, roots were dipped in a fungal spore suspension or water (control) for 60 min. The inoculated plants were transferred to pots containing vermiculite and grown under preinoculation conditions. More than three tomato plants were used for each treatment, and all experiments were repeated at least thrice.

RNA Extraction
For RNA sequencing library construction, total RNA was extracted using Sepasol-RNA I Super G (Nacalai Tesque, Kyoto, Japan) from the roots of tomato plants with eight different treatments; (i) 30 (MgO7d + FOL21d). The total RNA concentration was determined using a Biospec-nano spectrophotometer (Shimadzu, Kyoto, Japan). More than three tomato plants were used for each treatment, and all experiments were repeated twice.

Preparation of cDNA Library and Sequencing
Aliquots (500 ng) of RNA were used for cDNA library preparation using an NEBNext Ultra II RNA Library Prep Kit for Illumina (NEB, Ipswich, MA, USA). cDNA libraries were sequenced using the Illumina MiniSeq platform to generate 75 bp single-end raw reads. Prior to assembly, raw data were filtered using a range of quality control measures. Highquality sequences were obtained from the original offline data sequences by removing the reads containing adapters and low-quality reads. All subsequent analyses were based on filtered clean data, which were aligned against the reference genome (GenBank accession number: GCA_000188115.3) using the CLC Genomics Workbench software ver. 12.0.3.

Screening of DEGs, Annotation, and GO and KEGG Analyses
The sequencing results were analyzed using CLC genomic workbench version 10.1.1. The reference genome sequence and annotation information of Solanum lycopersicum was obtained from NCBI (https://www.ncbi.nlm.nih.gov/genome/, accessed on 14 January 2020) and the used version was SL2.50. The value of unique exon read, which was obtained from the counted all mapped reads on exons, was used for gene expression profile analysis. Gene expression profiles were compared to find differentially expressed genes (DEGs) among different conditions using edgeR R package [49]. Genes with log 2 (fold change) values greater than 1 or lower than −1 were assigned as DEGs. Gene ontology (GO) number assignment of each gene of S. lycopersicum SL2.50 was performed by blastp program [50] with the protein sequence data set of UniProt [uniprot_sprot, release-2019_10] (https: //www.uniprot.org/, accessed on 25 October 2022). The cut-off criteria for assignment of a gene to homologous protein of UniProt database were >70% coverage and >25% identity, respectively. GO enrichment analysis of DEGs was performed using topGO R package [51]

Gene Expression Analysis
Total RNA was extracted as described in Section 4.5. Reverse transcription was performed with 1 µg total RNA in a 20 µL reaction volume using ReverTra Ace qPCR RT Master Mix with gDNA Remover (Toyobo, Osaka, Japan) according to the manufacturer's instructions. The primers used are listed in Table S15. cDNA was diluted (1:1), and 1 µL was used as a template in 20 µL of THUNDERBIRD SYBR qPCR Mix (Toyobo). The gene expression in each sample was normalized to that of tomato actin (GenBank accession No. FJ532351). Gene-specific primers were designed using Primer Express (Applied Biosystems, Foster City, CA, USA), based on sequences deposited in GenBank, to amplify 70-150 bp fragments (Table S10). Real-time quantitative polymerase chain reaction (RT-qPCR) analysis was performed using the Applied Biosystems' QuantStudio Real-Time PCR system (Applied Biosystems). Relative quantification was performed using the ∆∆C T method (Livak and Schmittgen, 2001 [52]).

Cis-Element Prediction
The generic file format of Solanum lycopersicum DC9.1 sequence (SlGRP4) [24] (accession: XP004248032.1) encompassing 2000 bp 5 upstream of the translation start site was used to search the promoter sequence. Afterwards, the online program Plant CARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 10 December 2022) was used to identify the cis-acting elements of the sequence.

Microscopy
To detect FOL in root tissues, tomato roots pretreated with MgO NPs and challengeinoculated with FOL were cut into 3 mm-long segments with a blade, which were prepared into sections for light microscopy using a freezing microtome (Nihon Khoki, Osaka, Japan). The sections were then stained with 1% cotton blue in lactophenol and observed under a BH2 light microscope (Olympus, Tokyo, Japan).
For immunostaining, tomato roots were cut into 3 mm-long segments with a blade and fixed in 5% glutaraldehyde in cacodylate buffer (pH 7.0). Fixed samples were immersed in t-butyl alcohol after dehydration through a graded series of t-butyl alcohol-ethanol and then embedded in paraffin. Transverse sections (14-18 m thick) were cut using a rotary microtome. Paraffin was removed using Histo-Clear (National Diagnostics, Atlanta, GA, USA), and sections were hydrated in a graded series of ethanol to water.
For immunohistochemical analysis for the localization of SlGRP4 in tomato root tissue, antibodies against a synthesized peptide fragment (CGGGYKPPHGEYKSPGGGYKPPH), designed based on the amino acid sequence of Solanum lycopersicum DC9.1 sequence (SlGRP4) [24] (accession: XP004248032.1), were raised in rabbits. The Vectastain ABC horseradish peroxidase staining kit (rabbit IgG) (Vector Laboratories, Burlingame, CA, USA) was used for immunohistochemical localization of SlGRP4 in tomato hypocotyl tissues with 3,3 -diaminobenzidine tetrahydrochloride (Wako Pure Chemical, Osaka, Japan) as the substrate following the manufacturer's instructions. The sections were observed under an optical microscope (Olympus BH-2, Tokyo, Japan).
For ROS production in tomato root tissue, tomato roots were histochemically stained with 3 3 -diaminobenzidine (DAB) to visualize in situ ROS (hydrogen peroxide) accumulation. Tomato roots were immersed in DAB solution (2 mg/mL) for 15 min under vacuum, then root sections were cut with a razor blade and observed under a BH2 light microscope (Olympus).

Conclusions
Tomato seedlings grown for 7 days after drenching with MgO NPs suspension showed long-term immunity to Fusarium wilt of tomato up to harvest. The molecular mechanisms underlying this immune response are largely unknown. In this study, we attempted to identify the genes involved in MgO NP-induced immunity against Fusarium wilt using RNA-seq analysis. Our study revealed that Rboh encoding NADPH oxidase and Rcr3 encoding a papain-like apoplastic protease were upregulated within 1 h after inoculation of FOL on roots pretreated with MgO NPs. This study also revealed that the SlGRP4 gene was rapidly expressed in tomato roots treated with MgO NPs and its product SlGRR4 profusely accumulated in the cell walls of epidermal and vascular cells. In addition, upon FOL inoculation, SlGRR4 rapidly and greatly accumulated in the cell walls of cortical cells. Because Rboh, Rcr3, and cell wall-located GRPs are all involved in extracellular immunity (apoplastic immunity) in tomato plants, apoplastic immunity plays an important role in MgO NP-induced immunity against Fusarium wilt.