Peanut (line Gan hua-5) seeds were obtained from the Jiangxi Ecological Experimental Station of Red Soil, Chinese Academy of Science, surface sterilized (5 min in 70% EtOH; rinse sterile water), germinated in the dark at 28°C with autoclaved damp vermiculite until the radicle reached 2–3 cm and then transferred to 1/2 Hogland nutrient solution [6 mM/L KNO₃, 5 mM/L CaCl₂, 1 mM/L NaH₂PO₄, 2 mM/L MgSO₄, 0.025 mM/L H₃BO₃, 2 μM/L MnCl₂, 1 μM/L ZnSO₄, 0.1 μM/L (NH₄)₆Na₇MoO₂₄, 0.25 μM/L CuSO₄ and 100 μM/L Fe-EDDHA, pH 6.0] until different treatments, and kept in a growth chamber at a constant temperature of 28°C on a day/night cycle of 16/8 h.

The fungal strain P. liquidambaris was originally isolated from the inner bark of Bischofia polycarp and inoculated on potato dextrose agar, labeled with a green fluorescent protein (GFP) through a vector plasmid pCT74 by Chen et al. (2011). The fungal inoculum was centrifuged before use, and then the mycelium was washed and dissolved with sterilized double-distilled water for inoculation and irrigated to the roots. Total genomic DNA from plant roots was extracted 7 days after inoculation to detect colonization by qPCR, and GFP primers were used (Supplementary Table 1). The fungi treatment was expressed as E in the Figure. For the H₂O₂ treatment, different concentrations (20, 40, 80, 100, and 200 μmol/L) were used, for the catalase, CAT-1 (5 mKat/L), and CAT-2 (10 mKat/L) were chosen according to the previous study (Xie et al., 2017).

Once harvested, root and shoot weights were measured using a digital scale. The chlorophyll content of young developed leaves was measured by acetone extraction. Accurately, weighed plant samples (0.1 g) were homogenized in the presence of 1 ml 80% acetone and leached for 2 h. Each treatment was performed in six parallel samples. Subsequently, the chlorophyll content of the samples was determined using a spectrophotometer at 663 and 645 nm. The calculation of chlorophyll pigment concentrations was performed according to the following equation (Arnon, 1949):

Peanut root and shoot samples were dried in an oven at 80°C until a constant weight was achieved. Then, 0.1 g of dried sample was placed in a digestion tube filled with 5 ml of nitric acid and digested using a microwave digestion apparatus. The digested sample was diluted to 50 ml with distilled water, and the Fe content was then determined using an atomic spectrophotometery (Prodigy, Leeman, USA). To observe the localization of Fe in vetiver plants, fresh root tissues were processed and stained with Perls Prussian blue (Stacey et al., 2008). Fresh roots were infused with 4% (v/v) HCl and 4% (w/v) potassium ferrocyanide (Perls Prussian blue stain). Excess stain was washed with distilled water, and the slides were photographed using a digital camera.

Cell death was tested with minor modifications with the Evans blue method (Kabir et al., 2020). Briefly, at room temperature, shoots and roots were initially incubated in 0.25% Evans blue emulsion for ~15 min. The suspension was subsequently treated with 1 ml of 80% ethyl alcohol for 10 min. The plant tissues were next incubated at 50°C for 15 min in a water bath and then rotated at 12,000 rpm for 10 min. The supernatant's optical absorption was finally measured at 600 nm. H₂O₂ was determined using a kit purchased from Jiancheng (Nanjing, China).

For the measurement of the ·O2- generation rate, 0.3 g of fresh samples were ground in liquid N₂ and extracted in 3 ml of ice-cold 50 mM sodium phosphate buffer (PBS) (pH 7.0). About 1 ml of fresh leaf extract supernatant was added to 0.9 ml 65 mM phosphate buffer solution (pH 7.8) and 0.1 ml 10 mM hydroxylammonium chloride. PBS (pH 7.0) instead of 1 ml of fresh leaf extract supernatant was used as a blank. The reaction was incubated at 25°C for 35 min. The solution from the above reaction mixture (0.5 g) was then added to 0.5 ml of 17 mM sulfanic acid and 0.5 ml of 7.8 mM α-naphthylamine solution. After 20 min of reaction, 2 ml of ether was added to the above solution and then mixed well. The solution was centrifuged at 1,500 g at 4°C for 5 min. The absorbance of the pink supernatant was measured at 530 nm with a spectrophotometer. Absorbance values were calibrated to a standard curve generated with known concentrations of HNO₂ (Wang Q. H. et al., 2013).

Total RNA was extracted from the root samples using a total RNA isolation reagent (Sangon, Shanghai) following the manufacturer's instructions, and the final RNA yield was quantified using a Nanodrop Spectrophotometer. First-strand cDNA synthesis was performed from RNA with a cDNA synthesis kit (Vazyme, China). Quantitative real-time PCR was performed in a real-time PCR system to detect the expression of AhIRT1, AhFIT, and AhFRO2 using specific primers (Supplementary Table 1). The levels of relative gene expression were analyzed using the 2^−ΔΔCt method, and the housekeeping gene AhActin was used as an internal control. Six independent replicates were considered for each sample.

Superoxide dismutase (SOD) activity was assayed by measuring its capacity to inhibit the photochemical reduction of nitroblue tetrazolium following the method of Stewart and Bewley (1980). Peroxidase (POD) activity was measured by the increase in absorbance at 470 nm due to guaiacol oxidation (Nickel, 1969). Catalase (CAT) activity was measured as the decline in absorbance at 240 nm due to the decrease in the extinction of H₂O₂ according to the method of Patra et al. (1978). The reduced glutathione (GSH) content assays according to the previous study (Li et al., 2013). VC was determined by titration with 2,6-dichlorophenolindophenol (Bessey, 1933). For the nicotinamide adenine dinucleotide phosphate (NADPH) detection, fresh tissues were ground into powder in liquid nitrogen and dissolved in an extraction solution (50 mM Tris-HCl, 0.25 M sucrose, 1 mM ASC, 1 mM EDTA, 0.6 % PVP, 1 mM PMSF), and then the homogenate was centrifuged at 12,000 rpm at 4°C for 20 min. A 20 μl of an aliquot of the supernatant was mixed by adding 0.5 ml of 10 μM NADPH and 50 mM XTT. The level of lipid peroxidation in fresh leaves was measured in terms of MDA con centration by the thiobarbituric acid reaction method (Heath and Packer, 1968). Fresh leaves and roots were homogenized in 0.05 mol/L phosphate buffer (pH 7.8) with a mortar and pestle under chilled conditions with liquid nitrogen. The homogenate was filtered through a four-layer muslin cloth filter and centrifuged at 12,000 g for 10 min at 4°C. To estimate .OH production, the reaction was performed according to the method reported by Gómez-Toribio (Cheeseman et al., 1988). The reaction mixture contained 500 μM quinone, 100 μM Fe³⁺- 110 μM EDTA, 0–25 μM IlLPMO9A, 100–2500 U/L GDH, 5 mM glucose, and 2.8 mM 2-deoxyribose in 100 mM phosphate buffer (pH 5.0). The production of .OH was estimated as the conversion of 2-deoxyribose into thiobarbituric acid (TBA) reactive substances (TBARS). The absorbance of TBARS was determined at 532 nm.

Total RNA was extracted from peanut roots 7 days after P. liquidambaris cocultivation. Briefly, RNA was quantified and qualified using an Agilent 2100 Bioanalyzer, NanoDrop (Thermo Fisher Scientific, Waltham, MA, USA), and 1% agarose gel. A total of 1 μg of RNA with an RNA integrity number value >7 was used for subsequent library preparation. Next-generation sequencing library preparations were constructed according to the manufacturer's protocol. Transcripts in FASTA format were converted from a known GFF annotation file and indexed properly. The DESeq Bioconductor package was used for differential expression analysis. The p-value was set below 0.05 to detect differentially expressed genes. Gene Ontology (GO) TermFinder was used to identify GO terms, and a list of significantly enriched genes (p < 0.05) was annotated. All Illumina sequence data have been deposited in Sequence Read Archive with the project ID PRJNA779473 (https://www.ncbi.nlm.nih.gov/).

All experiments were performed in at least six replicates except RNA-seq (3 replicates). The data were analyzed using R-Studio, the t-test was used to analyze differences between two groups, and Duncan was used when more groups existed. Statistical significance was determined using one-way ANOVA. Different letters represent differences between different groups, and significance is defined as p < 0.05. Graphs and images were assembled using Adobe Photoshop 7.0.

Plants showed leaf chlorosis in alkaline soil (Figures 1A,B), which seriously affects the chlorophyll synthesis and plant fresh weight (Figures 1D–F). Through the pot experiments (pH 8.0, DTPA-Fe, 3 mg/kg), we learned that in alkaline soil, the addition of P. liquidambari significantly promoted a 14% increase in chlorophyll (Figure 1D), a 27.8% increase in fresh weight of the aboveground part (Figure 1E), 50.68% and 81.01% increase in Fe concentration of the root and shoot (Figure 1G and Supplementary Figure 1A), indicating a significant promotion in Fe absorption following P. liquidambaris inoculation. At the same time, we also observed changes in plant appearance. P. liquidambaris makes the plant greener and grows better in the seedling stage (Figures 1A–C). The Fe concentration, fresh weight, and dry weight of the roots were also improved significantly, and the results were shown in the figure (Supplementary Figures 1A–C). We also compared peanuts grown in alkaline soil (pH 8.0) and acidic soil (pH 6.0) in the early stage. It can be seen from Supplementary Figure 1D that the color of peanut leaves in alkaline soil showed chlorotic, and the chlorophyll is also significantly lower than in acidic soil (pH6.0) (Supplementary Figure 1E). This suggested that peanuts grown in alkaline soils also exhibit chlorophyll synthesis disorders.

To examine whether P. liquidambaris attenuated Fe deficiency-induced damage in peanuts, we analyzed cell death, H₂O₂, and ·O2-. According to the results, Fe deficiency significantly increased H₂O₂ (33.7%, 31.8%, Figures 2A,B), ·O2- (128%, 30.95% Figures 2C,D), and cell death (129%, 116%, Figures 2E,F) in leaves and roots, indicating serious cell damage. However, the addition of P. liquidambaris to Fe-deficient plants showed a significant decrease in these indicators, which were similar to those cultivated under Fe-sufficient conditions. At the same time, the Fe concentration and chlorophyll content are also consistent with the results of Fe staining. The Fe concentration, chlorophyll, and fresh weight of the E-Fe group are significantly higher than that of the -Fe group (Figures 2G–J). The growth of the plant in the seedling stage shows a better growth state, and the leaves show a deeper green (Figures 2K,L). In addition, compared with the CK, Fe staining in roots showed a significantly weak blue color due to Fe deficiency, P. liquidambaris supplementation significantly strengthened the blue color of E-Fe plants (Figure 2M) in the root. In Prussian blue dyeing, dark blue represents a higher Fe concentration (Banerjee et al., 2019). The plants showed similar staining levels in roots in the CK and CK+E plants. Compared with the CK group, the Fe concentration has no obvious difference in CK+E, and there is a small change in chlorophyll and fresh weight.

To test the correlation between FIT, IRT1, FRO2, and Fe concentration, we designed experiments with different Fe concentrations and quantified the expression of FIT, IRT1, and FRO2 using RT–qPCR after Fe-deficiency. First, the H₂O₂ content in Fe-deficient plants was significantly higher than that in the Fe-sufficient treatment (100 μmol/L, Figure 3A). Then, we found that FIT was negatively correlated with Fe concentration. When the Fe concentration was 2 μmol/L, the expression of FIT was 89.5 times higher than that at 100 μmol/L (Figure 3B). IRT1 and FRO2 were also strongly induced 202 times and 172 times higher than normal Fe supply (100 μmol/L), respectively (Figures 3C,D). Meanwhile, IRT1 and FRO2 showed similar expression patterns with FIT. The fresh weight and chlorophyll of the shoot were also an indicator of Fe concentration. Therefore, chlorophyll and shoot fresh weight are negatively correlated with Fe concentration (Figures 3E,F). Fresh weight of roots, total fresh weight, and phenotypic changes of plants are shown in Supplementary Figure 2 and they were negatively correlated with Fe concentration, too.

To examine the relationship between H₂O₂ and Fe absorption, peanuts were treated with different concentrations of H₂O₂ (20, 40, 80, 100, and 200 μmol/L) under sufficient Fe supply, and then RT–qPCR was performed to detect the expression levels of FIT, IRT1, and FRO2. The Fe concentration, H₂O₂ content, chlorophyll, and fresh weight of peanut roots were also recorded. The H₂O₂ content was shown in Figure 4A. Figures 4B–D shows that a low concentration of H₂O₂ had no inhibiting effect on FIT, IRT1, FRO2, rather, it had a stimulating effect. However, high H₂O₂ (80-200 μmol/L) inhibited the expression of FIT, IRT1, and FRO2, which is consistent with previous experimental results, and the expression of IRT1 and FRO2 was repressed, too (Figures 4A–C). Due to the addition of H₂O₂ at a low concentration of (20 μmol/L), the H₂O₂ content detected in roots is low (29.19 μmol/g FW), while that under Fe-deficiency is 40.53 μmol/g FW (Figure 4A). Therefore, it may be that the addition of H₂O₂ at a low concentration does not increase the H₂O₂ content in roots to a high level, so a stimulating effect was observed. However, a high concentration of H₂O₂ (80, 100, 200 μmol/L) significantly increased H₂O₂ content to 39.19-52.99 μmol/g FW in roots and inhibited the expression of FIT, IRT1, and FRO2 significantly. This also implies that the H₂O₂ content in the plant reaches a certain value to inhibit Fe absorption. In addition, the Fe concentration and chlorophyll content were decreased dependent on H₂O₂ content (Figures 4E,F). When H₂O₂ addition is low, it has no obvious effect on Fe concentration and chlorophyll. Increasing H₂O₂ (100 and 200 μmol/L) addition significantly reduced Fe concentration and chlorophyll. This shows that a low external concentration of H₂O₂ will not inhibit the absorption of Fe, while a high concentration has an obvious inhibitory effect. The shoot fresh weight and total fresh weight were all significantly reduced (100 and 200 μmol/L, Supplementary Figures 3B,C).

Since our results showed that the H₂O₂ decreased with the addition of P. liquidambaris under Fe deficiency, we explored whether this phenomenon was related to the gene expression of Fe absorption. Considering that P. liquidambaris can reduce H₂O₂, we increase the concentration of added H₂O₂. After peanuts colonized by P. liquidambaris were transferred to Fe-deficiency solution, external H₂O₂ was added. As shown in Figure 5A, the addition of external H₂O₂ (80, 100, 200, 400 μmol/L,) significantly increased H₂O₂ in peanut root. When the H₂O₂ concentration was artificially changed by external addition, the expression of FIT, IRT1, and FRO2 in peanuts under Fe deficiency also was inhibited, and the inhibition effect was dose-dependent (Figures 5B–D). Furthermore, high FIT, IRT1, and FRO2 expression corresponded to high Fe concentration, and chlorophyll content (Figures 5E,F). Although the expression of FIT, IRT1, and FRO2 was higher when 80 μmol/L H₂O₂ than without addition. We have not observed a decrease in Fe concentration and chlorophyll. This may be the same reason as above, that is, low exogenous H₂O₂ will stimulate the expression of FIT, IRT1, and FRO2. The root fresh weight, shoot fresh weight and total fresh weight were all reduced significantly (200 and 400 μmol/L, Supplementary Figures 4B,C), too.

To further prove that H₂O₂ could affect the expression of the Fe-absorption gene under Fe deficiency, we used CAT to eliminate H₂O₂ from plant roots, then performed qRT-PCR to detect the expression of FIT, IRT1, and FRO2. As shown in Figure 6A, the addition of CAT-2 (10 mKat/L) significantly reduced the H₂O₂ content in plant roots. In addition, Fe concentration and chlorophyll under Fe-deficient were increased, along with the high expression of FIT, IRT1, and FRO2 (Figures 6B–D). At the same time, with the addition of 5 mKat/L, we observed no significant change in H₂O₂ (Figure 6A) and no change in FIT and FRO2 expression, and a small change in IRT1 (Figures 6B–D). The Fe concentration and chlorophyll of plants are also consistent with the Fe absorption of plants, 10 mKat/L CAT makes Fe concentration and chlorophyll maximum (Figures 6E,F). At the same time, the shoot fresh weight and total fresh weight were also the largest at this concentration (10 mKat/L, Supplementary Figures 5B,C).

We detected the contents of different stress indictors to observe whether the addition of P. liquidambaris can help peanuts eliminate oxidative damage. SOD in the root was significantly up-regulated in the E-Fe group compared with E, however, it was not significant compared with -Fe. There was no obvious phenomenon in the leaves (Figures 7A,B). No obvious changes were detected in POD (Figures 7C,D). The CAT of roots and leaves decreased significantly in the Fe-deficiency group, while E-Fe increased significantly (Figures 7E,F), however, in the root, the CAT in E-Fe was higher than CK and E while in leaf it's a little lower. The GSH of the E-Fe group increased significantly in leaves and roots compared to the -Fe group (Figures 7G,H), but we did not detect a significant change between the -Fe group and CK. Finally, we observed a significant increase in VC content in the E-Fe group compared to other groups (Figure 7I). But VC can only be detected at the leaves. No obvious changes were detected in NADPH (Figures 7J,K). Compared with CK, the root MDA in E-Fe group decreased significantly, but compared with the -Fe group, the change was not obvious (Figures 7L,M). The MDA of the Fe-deficiency group of the leaf was significantly higher than that of the other groups. No obvious changes were detected in .OH (Figures 7N,O).

After 7 days, post P. liquidambaris addition, we used q-PCR to detect the colonization of P. liquidambaris. As shown in Supplementary Figure 6, P. liquidambaris can be detected in peanut root, it indicated that P. liquidambaris successfully symbiosis with peanuts. Then to verify our above experimental results, we performed transcriptomic analysis. Consistent with the above experiment, the time point we selected was 7 days after Fe-deficiency, that is, the 14^th day of P. liquidambaris colonization. These genes were hierarchically clustered according to similar functions (Figure 8A). According to the Venn diagram (Figure 8B), the addition of P. liquidambaris upregulated 79 genes and downregulated 96 genes compared with the CK group, while Fe-deficiency treatment upregulated 3886 genes and downregulated 2008 genes. In the presence of P. liquidambaris, Fe deficiency upregulated 62 genes and downregulated 424 genes compared with Fe-deficiency. The early colonization of P. liquidambaris reduced the number of different expression genes number in plants caused by Fe deficiency (Figure 8C). For the GO analysis (Supplementary Figures 7A–C), it was found that the phenylpropanolamine metabolic process was the most significantly enriched in the Fe-deficiency group relative to the CK group, followed by glutathione metabolism and metal iron ion metabolism, and secondary metabolism and phenylpropanolamine synthesis were also significantly enriched. When P. liquidambaris-colonized peanut was exposed to Fe deficiency, significant enrichment was observed in the regulation process of transcription factors, DNA binding transcription factor activity and specific DNA sequence binding activity. In the P. liquidambaris group compared with CK treatment, the cell junctions were significantly enriched (Supplementary Figure 7C). The KEGG results showed that compared with the CK group (Supplementary Figure 7D), the main enrichment pathways in the Fe-deficiency group were flavonoid synthesis, genes regulating plant circadian rhythm, and plant–pathogen interactions. At the same time, the protein process in the endoplasmic reticulum and the MAPK signaling pathway in plants were also significantly activated. In the P. liquidambaris treatment group compared with CK peanut (Supplementary Figure 7E), the most significantly enriched signaling pathway was the plant rhythm regulation signaling pathway, followed by the plant–pathogen interaction and RNA degradation pathways, and obvious changes were also observed in ribosome biosynthesis and ion transport. We also found significant changes in the process of photosynthesis. Under Fe deficiency stress after the addition of P. liquidambaris with Fe-deficiency (Supplementary Figure 7F), significant changes were observed in phenolic acid synthesis, plant circadian rhythm regulation, plant–pathogen interaction, and phenolic acid biosynthesis. The transport of metal ions, protein synthesis of the endoplasmic reticulum, and MAPK signaling were also activated.

To verify the DEGs identified in RNA-Seq, 8 DEGs were selected for qRT-PCR. qRT-PCR results were compared with the FOLD change of FPKM in RNA-sequencing expression analysis (Supplementary Figure 8). As shown in Supplementary Figure 8, 6 genes showed similar expression patterns in qRT-PCR analysis to RNA-Seq analysis. These results confirm the reliability and accuracy of our transcriptome data.

Reactive oxygen species (ROS) signaling is the key determinant of the plant's environmental response (He et al., 2018). H₂O₂ has been shown to negatively regulate Fe absorption in many plants. It can increase in amount under stress and damage proteins, nucleic acids, and lipids, eventually triggering cell death (Ranieri et al., 2003; von der Mark et al., 2021). H₂O₂ is not only an oxidative damage substance in vivo, but also an important hormone signal (Van Leene et al., 2016). Our results show that the addition of P. liquidambaris significantly reduces the H₂O₂ of plants under Fe-deficiency, suggesting that Fe-deficiency stress is alleviated. A low H₂O₂ content can maintain the normal growth state of plants. Some studies have confirmed that the reduction of H₂O₂ content depends on the activation of cat2 as demonstrated in an Arabidopsis cat2 mutant (von der Mark et al., 2021), indicating that cat2 can help metabolize excess H₂O₂ in plants. The CAT enzyme activity was higher than that of Fe-deficiency (Figure 7) after P. liquidambaris addition. In addition, IRT1 functions as a transporter that helps CAT2 transport Fe, as the activity of CAT2 requires the participation of Fe. Fe-deficiency will prompt endogenous H₂O₂ to produce .OH in the Fenton reaction. This is also the reason why Fe-deficiency causes oxidative damage.

A negative correlation between ROS and Fe acquisition, particularly H₂O₂, has been reported (Ranieri et al., 2003), and their relationship has previously been shown to affect FIT activity through interaction with the zinc finger, an oxidative stress response transcription factor of Arabidopsis thaliana AtZAT12 (Brumbarova et al., 2016), thereby negatively regulating Fe absorption. Many studies have reported the relationship between FIT and IRT1 and FRO2 (Seguela et al., 2008; Yuan et al., 2008; Maurer et al., 2011; Sivitz et al., 2011; Nishida et al., 2012; Wang N. et al., 2013; Matsuoka et al., 2014). It is agreed that FIT is upstream of IRT1 and FRO2 in the regulation of Fe absorption. Even under Fe-deficiency induced by transitional Ni, the overexpression of FIT, IRT1, and FRO1 has been detected (Nishida et al., 2012). Our results also show that IRT1 and FRO2 are consistent with FIT under Fe-deficiency, which also supports our speculation. As the downstream result of H₂O₂ regulation, FIT further regulates the Fe absorption in plants. Although their expression fold changes are not the same. This may be because the low FIT can induce high expression of IRT1 and FRO2 or because posttranscriptional modification occurs (Schwarz et al., 2020). It has been reported that phosphorylation of FIT regulates the expression of IRT1 and FRO2, thereby regulating Fe absorption. Some studies suggest that FIT may exist in two forms, an active form and an inactive form (Wang et al., 2007). A number of studies have demonstrated that in most plants, FIT acts as the upstream of IRT1 and FRO2 to regulate Fe uptake in plants (Ling et al., 2002; Ogo et al., 2007; Kim et al., 2019). Although we did not observe the direct effect of FIT and IRT1 or FRO2 in peanuts, our qPCR results demonstrated the consistency of their expression, which is consistent with studies in other dicotyledonous plants (Seguela et al., 2008; Yuan et al., 2008; Maurer et al., 2011; Sivitz et al., 2011; Nishida et al., 2012; Wang N. et al., 2013; Matsuoka et al., 2014), we believe that the same regulation pattern also exists in peanut, that is, FIT as a transcription factor to regulate the expression of IRT1 and FRO2 and participates in Fe absorption in peanut. At the same time, our experimental results also showed the strong inhibitory effect of H₂O₂ on FIT, IRT1, and FRO2.

Plants exposed to various biotic and abiotic stresses are compelled to generate higher levels of ROS, such as .OH, ·O2- leading to an alteration in the cellular redox homeostasis, therefore, acquiring resistance to neutralizing the excessive oxidative damage (Jin et al., 2021). In the present study, we found that Fe deficiency stress-induced the generation of ·O2- and H₂O₂ in the leaves of the plants (Figures 2B,C). These results agreed with previous reports by Ranieri et al. (2001) for sunflower and Sun (Sun et al., 2007) for maize. The onset of oxidative damage in plants is more prominent due to Fe deficiency as Fe is the central constituent or factor of major antioxidant enzymes (Kabir et al., 2015). Plants generate ROS when facing the outside stress itself and antioxidant enzymes to resist the damage of the outside world, plant stress relief is usually dependent on the antioxidant enzymes in vivo, when Fe deficiency plants showed severe oxidative stress reaction, including H₂O₂, ·O²⁻, cell death increased significantly, it is usually harmful to the growth of plants. Our results show that the addition of P. liquidambaris can help plants relieve this oxidative stress and tend to a normal growth state, which may be attributed to the increase of antioxidant enzymes in plants. Under the condition of Fe deficiency, the levels of ROS scavenge enzymes (CAT, SOD, POX) of onion seedlings were reported to decrease. CAT could decompose H₂O₂ into H₂O and O₂ in vivo. CAT is an enzyme containing heme, so their activity is correlated with Fe concentration (Sevilla et al., 1984). The elevated Fe concentration enhanced the enzymatic activity of CAT, helping the plants to scavenge more ROS. Although our experimental results did not observe a significant increase in SOD and POD, this may be due to the inappropriate timing of our detection. Transcriptome data also demonstrated that the addition of P. liquidambaris helped the plants to far shrink the gene transcriptional changes in the absence of Fe, helping the plants to approach normal growth.