The material employed was the non-heading Chinese cabbage ‘Suzhouqing,’ which was sown in a seedling tray containing a matrix-vermiculite (3:1) mixture and cultivated in a light incubator. The culture conditions were established at 24°C, 16 h of light/16°C, 8 h of darkness, and 65% relative humidity. The non-heading Chinese cabbage seedlings are moved to a turnover box with Hoagland nutrient solution for continuing culture during the four-leaf phase, and the nutrient solution is refreshed every 3 days. The experiment included a normal iron supply (control treatment) (CK, 0.02 mol/L) and an iron deficiency treatment (−Fe, 0 mol/L). On the 4th, 7th, 11th, and 14 days of culture in the turnover box, the phenotypes were observed. On the 14th day of cultivation, each treatment’s roots and leaves were collected simultaneously. The leaves were separated into three sections once they were harvested. The remaining samples were utilized to generate a transcriptome sequencing library, with one part used to determine chlorophyll content and one part used to determine Fe concentration. The remaining roots were similar to leaves and were used to generate the transcriptome sequencing library, with a portion of the root was utilized to determine the concentration of Fe.

We improved the prior Du et al. (2007) method for determining chlorophyll content in non-heading Chinese cabbage leaves. This is how it was done:

(1) Clean the surface dirt off of new non-heading Chinese cabbage leaves, remove the major veins, and cut the tissues; (2) Weigh three fresh cut samples of 0.1 g each for each treatment and deposit them in test tubes, respectively. To completely submerge the sample in an ethanol solution, 5 mL of 95% ethanol was added to each test tube. Soak it for 24 h in the dark; (3) The chlorophyll pigment extract was applied to the enzyme standard plate after 24 h in the dark, and 95% ethanol was employed as the control. The absorbance was measured at 665 nm and 649 nm; (4) Chlorophyll a and b concentrations (mg/L) were calculated using the formula, and the total chlorophyll (Chla + b) concentration was computed by adding the total sum of chlorophyll a (Chla) and chlorophyll b (Chlb).

The following formula was used for the chlorophyll quantification:

C_a = 13.95 A₆₆₅-6.88 A₆₄₉

C_b = 24.96 A₆₄₉-7.32 A₆₆₅

The concentrations of chlorophyll a, chlorophyll b, and total chlorophyll are represented by the formula Ca, Cb, and Ca + b. In 95% ethanol, the maximum absorption wavelengths of chlorophyll a and chlorophyll b were 665 nm and 649 nm, respectively. The absorbance of chlorophyll a and chlorophyll b were denoted by the letters A665 and A649, respectively.

(5) After determining the pigment concentration, the content of each pigment per unit fresh weight in the tissue was determined as follows:

Chloroplast pigment content (mg/g) = (pigment concentration × extract volume)/sample fresh weight

The materials were non-heading Chinese cabbage roots and leaves that had been cultured for 14 days. The samples were first destroyed at 105°C before being dried at 70°C. The samples were finally crushed using a stainless steel grinding machine. The mixed acid (HNO₃:HClO₄ = 3:1) was used to digest the milled drying samples (0.25–0.5 mm sieve) 0.02–0.2 g. Finally, the digested solution was diluted to 100 mL with water, and the Fe concentration was measured using an Inductively Coupled Plasma Emission Spectrometer (ICP-AES).

As samples, the roots and leaves from the control and iron deficiency treatments were obtained. Each treatment was carried out three times for a total of 12 samples. RNA was extracted using the RNA simple total RNA extraction kit, and the purity, concentration, and integrity of the RNA were determined using NanoDrop, a Qubit 2.0, and an Agilent 2100, respectively. The final cDNA library was created by reverse-transcribed total RNA of roughly 10 g. The Illumina NovaSeq™ 6000 was used to sequence the library after it had been qualified, and the sequencing read length was 2*150bp (PE150). Transcriptome data was submitted to the Sequence Read Archive (SRA) of the National Center for Biotechnology Information (NCBI) (Registration number PRJNA745363¹).

Preprocessing the original data, including the elimination of sequencing joints (added during database development) and low-quality sequencing data (caused by the error of the sequencing instrument itself), is necessary to produce accurate and reliable analytical results. The effective data (Valid Data) was compared to the Chinese cabbage genome² and the gene position information specified in the genome annotation file (gtf and gff) was statistically assessed. By counting the original sequencing, effective sequencing, Q20, Q30, and GC content (Ewing et al., 1998) for a comprehensive evaluation, we produce a large number of high-quality data through the Illumina high-throughput sequencing platform called raw data based on Sequencing By Synthesis (SBS) technology. StringTie was used to assemble the aforementioned reads in this study (Pertea et al., 2015). The expression levels of all transcripts were estimated using StringTie and DESeq2 (Love et al., 2014). Gene expression levels were determined using the Fragments Per Kilobase of Transcript per Million Fragments Mapped (FPKM) method (Florea et al., 2013), and differential expression genes (DEGs) were screened using the Fold Change ≥2 and False Discovery Rate (FDR) < 0.01 criteria (Florea et al., 2013). The FDR is calculated by subtracting the difference manifestness p-value from the error detection rate (Benjamini and Hochberg, 1995).

Genes that have similar functions are often found together. The R language package is used to cluster the genes in the Java language environment, and the genes are then submitted to NCBI for GO item analysis. Biological process, molecular function, and cell component are the three major components of the GO annotation system. Enrichment in two ways p-value and adjusted q-value were used to analyze KEGG pathways. The KEGG database was compared to the NCBI’s BLAST search engine (Kanehisa et al., 2008).

The co-expression network is built using functions from the WGCNA package in R to describe the correlation patterns among genes in different samples (Langfelder and Horvath, 2008). To construct the adjacent matrix, all gene expression data is normalized using the log₂ (1 + FPKM) value, and the soft threshold = 8 is chosen using the scale-free topology standard. The adjacency matrices were then turned into topological overlap matrices (TOM). Module required a minimum of 30 genes and a merge cut height of 0.25. Intergene variability was used to classify genes into clusters (Langfelder and Horvath, 2008). The correlation between the gene co-expression module and physiological indices was used to generate the correlation analysis (Liu et al., 2021). The protein interaction network of specific genes was constructed in the STRING database and visualized using Cytoscape 3.6.0 software (Shannon et al., 2003). Venn plots and heatmaps of DEGs were plotted using TBtools (Chen et al., 2020).

Nine genes were selected at random from DEGs in RNA-seq data to test the accuracy and reliability of transcriptome data. Primer 5.0 software (Ren et al., 2004) was used to design specific primers (Supplementary Table 9). RNA was reverse transcribed to cDNA from 1 μg of RNA using the PrimerScript RT Reagent Kit (Takara, Dalian, China). For quantitative real-time PCR (qRT-PCR), the cDNA was diluted a certain number of times. The reaction system included 10 μL of SYBR Green (Vazyme, Nanjing, China), 0.4 μL of forward primer, 0.4 μL of reverse primer, 0.4 μL of Rox, 2 μL of cDNA template, and 6.8 μL of re-evaporation. In each sample, three technical repeat tests were performed on nine potential genes selected from the transcriptome data. For each treated sample, three biological replicates were selected. The ABI StepOnePlus real-time fluorescence quantitative PCR machine was used to perform the qRT-PCR reaction. The reaction conditions were 95°C for 10 s, 95°C for 5 s, and 60°C for 30 s, for a total of 40 cycles. The dissolution curve conditions were reported by Yuan et al. (2019). To normalize the data, the GAPDH (glyceraldehyde 3-phosphate dehydrogenase) gene of non-heading Chinese cabbage was employed as an internal reference gene. The 2^–(ΔΔCT) method (Livak and Schmittgen, 2001) was used to calibrate the expression level of the target gene.

The non-heading Chinese cabbage ‘Suzhouqing’ was utilized as experimental material, with seedlings switched from matrix culture to nutrient solution culture at the four-leaf stage. The hydroponic experiment revealed that on the fourth day of hydroponics, there was no significant difference in plant height between the two treatments [iron deficiency treatment: 0 mol/L Fe²⁺; control (normal culture): 0.02 mol/L Fe²⁺], but the leaves under iron deficiency had a slight green loss compared to the control leaves (Figure 1A). The size and color of plant leave gradually changed as the amount of hydroponic time increased (Figures 1A–D). On the 14th day, the leaf area of non-heading Chinese cabbage was much larger than on the 4th and 7th days. We also discovered that at 14 days of culture, the leaf area of non-heading Chinese cabbage plants treated with iron deficiency was smaller than the control plants, and the leaf color was yellower than the control plants (Figure 1D). As a result, we hypothesized that iron deficiency promotes non-heading Chinese cabbage plant dwarfing and leaf yellowing.

The contents of chlorophyll and Fe²⁺ were examined to see how iron deficiency stress affected physiological indicators in non-heading Chinese cabbage. The materials utilized in this investigation were non-heading Chinese cabbage seedlings treated with iron deficiency and control for 14 days (Figure 2A). Under iron deficiency treatment, the contents of chlorophyll a, chlorophyll b, and chlorophyll a + b in seedling leaves dropped by 71.36, 80.83, and 75.88%, respectively (Figure 2B). Under iron deficiency treatment, the iron content of non-heading Chinese cabbage leaves and roots was reduced by 59.33 and 94.62%, respectively (Figure 2C). By analyzing the aboveground and underground parts of non-heading Chinese cabbage under two treatments, we discovered that iron deficiency not only inhibited the chlorophyll content and iron content in the aboveground tissue of non-heading Chinese cabbage, but also inhibited root growth and absorption of iron content (Figure 2).

We conducted the transcriptome analysis of 12 samples (roots and leaves of control and iron deficiency treatment, each treatment repeated three times) using Illumina paired-end sequencing, yielding a total of 80.29 GB of clean data. Each sample yielded 6.31 GB of data, with a Q30 base percentage of 93.82% or higher (Supplementary Table 1). The alignment efficiency of each sample’s Clean Reads with the defined reference genome ranged from 66.56 to 87.53% (Supplementary Table 2). Twelve samples were given the name ‘treatment-tissue-number’ to make the analysis easier. For example, CK-Leaf-1 is the first sample in the leaves under normal conditions, whereas Fe-Root-2 is the second sample in the root under iron deficiency conditions (Supplementary Table 1).

We discovered that principal component 1 (PCA1) described 93.0% of the content and principal component 2 (PCA2) explained 4.2% of the content of all genes in 12 libraries (Figure 3A). Each treatment’s three biological replicates were essentially clustered together, while CK Leaf, Fe Leaf, CK Root, and Fe Root were totally separated (Figure 3A). The results of the cluster analysis were similar to those of the main component analysis, indicating that the biological replication of the sample is consistent (Figure 3B).

Transcriptome sequencing was performed on the samples of the two treatments to acquire DEGs in the leaves and roots of non-heading Chinese cabbage under iron deficiency stress and control treatment. We discovered 9213 DEGs in this investigation (Figure 4). In the roots of non-heading Chinese cabbage, 2168 genes were up-regulated and 3380 genes were down-regulated under iron deficiency stress (Figure 4A), whereas 2604 genes were up-regulated and 2988 genes were down-regulated in the leaves (Figure 4B). Under iron stress, a Venn diagram revealed that 1927 genes, including 897 up-regulated genes and 1030 down-regulated genes, were co-expressed in roots and leaves, revealing a link between DEGs in roots and leaves (Figure 4C). These findings imply that under iron deficiency stress, the root and leaf of non-heading Chinese cabbage may have a common or unique response mechanism.

GO categorization analysis of DEGs in roots and leaves was performed to define the key functional categories of DEGs. The majority of DEGs were found to be involved in cell processes, metabolic processes, cell/cell components, binding sites, and catalytic activities, according to the findings (Figures 5A,B and Supplementary Table 3).

Differentially expressed genes were also examined for KEGG enrichment in this research (Supplementary Table 4). DEGs were found to be confined in plant hormone signal transduction (Ko04075), phenylpropane-like biosynthesis (Ko00940), starch and sucrose metabolism (Ko00500), plant-pathogen interaction (Ko04626), cysteine and methionine metabolism (Ko00270), and other pathways in the roots of non-heading Chinese cabbage under iron deficiency stress (Figure 5C). In the leaves of non-heading Chinese cabbage under iron deficiency stress, DEGs are mainly concentrated in plant hormone signal transduction (Ko04075), carbon metabolism (Ko01200), starch and sucrose metabolism (Ko00500), phenylpropanoid biosynthesis (Ko00940), amino acid biosynthesis (Ko01230) and other pathways (Figure 5D).

Many DEGs are linked to antioxidant enzyme activity, such as peroxidase (POD), superoxide dismutase (SOD), catalase (CAT), and ascorbate peroxidase (APX), were discovered during iron deficiency stress (Table 1). Under iron deprivation, the transcript profiles of 58 POD genes altered considerably in both leaves and roots, with 7 genes up-regulated and 32 genes down-regulated. Only in leaves were 5 POD genes highly up-regulated and 5 genes significantly down-regulated. In both roots and leaves, four POD genes were considerably up-regulated and five genes were significantly down-regulated (Supplementary Figure 1A). Three SOD genes were found to be considerably up-regulated exclusively in leaves, whereas one gene (Bra013863) was found to be strongly down-regulated in both leaves and roots (Supplementary Figure 1B). In addition, two CAT genes (Bra034674 and Bra034674) were highly up-regulated in leaves, while one CAT gene (Bra034674) was strongly down-regulated. In both leaves and roots, one CAT gene (Bra025832) was considerably down-regulated (see Supplementary Figure 1C). Only one APX gene (Brassica rapa_ newGene_2429) was down-regulated significantly in roots, and one APX gene (Bra031598) was down-regulated strongly in both roots and leaves (Supplementary Figure 1D).

Iron deficiency in plant seedlings influence the signaling system for plant hormones. Under iron deficiency stress, 1 SAUR gene, 10 AUX/IAA genes, 3 GH3 genes, and 4 Auxin transporter genes were considerably down-regulated in both roots and leaves, whereas 11 SAUR genes and 1 AUX/IAA gene were significantly down-regulated exclusively in leaves. Eight GH3 genes were strongly down-regulated in roots under iron deficiency stress, but only one gene (Bra022725) was highly up-regulated in leaves, and three genes’ expression was dramatically down-regulated (Supplementary Table 5). Three ARR genes were considerably up-regulated and four of them were significantly down-regulated in the cytokinin signaling pathway in both roots and leaves. Similarly, only one ARR gene (Bra002512) was highly up-regulated in roots, whereas two others were significantly down-regulated. Also, three ARR genes were highly upregulated exclusively in leaves, while nine were considerably downregulated only in leaves (Supplementary Table 5). Under iron deficiency stress, the expression patterns of 10 PYL genes associated with the abscisic acid signal pathway changed considerably. One gene (Bra005115) was significantly down-regulated in both roots and leaves, and seven genes were exclusively significantly down-regulated in leaves. Under iron deficiency stress, only one JAR gene (Bra034205) associated with jasmonic acid signal was discovered to be significantly down-regulated in both roots and leaves (Supplementary Table 5). According to the findings, iron deficiency in Chinese cabbage seedlings had an impact on plant hormone signaling pathways, particularly auxin (IAA), cytokinin (CTK), abscisic acid (ABA), and jasmonic acid (JA).

To better understand the co-expression network of associated genes in non-heading Chinese cabbage roots and leaves under iron deficiency stress, 9213 DEGs were screened by transcriptome sequencing, and 8148 DEGs were selected for WGCNA analysis (Supplementary Table 6). Gene expression cluster tree (Figure 6A) and hierarchical cluster tree analysis (Figure 6B) in WGCNA yielded a total of 11 modules. The largest module, with 2472 genes, accounted for 30.34% of all DEGs, whereas the smallest module, with 99 genes, accounted for 1.22% of all DEGs (Figure 6C). The physiological index of iron content (FeCon) was strongly positively correlated with the MEturquoise module, according to correlation analysis between these modules and iron content or chlorophyll a + b content (chlab). A total of 2319 genes were within the correlation coefficient >0.75, p < 0.01. Chlab was positively correlated with MEbrown and MEgreen modules, and negatively correlated with MEred and MEpurple modules. There were 968, 95, 118, and 18 genes in MEbrown, MEgreen, Mered, and MEpurple modules, respectively, with correlation coefficients >0.75, p < 0.01 (Figure 6D and Supplementary Table 6).

To further screen genes involved in iron uptake and transport, 40 DEGs were obtained by searching the keyword ‘iron’ in 8148 DEGs (Table 2), including iron reductase protein (6 genes), bivalent iron transporter (6 genes), vacuolar iron transporter (5 genes), metal nicotinamide transporter (5 genes), ferritin (4 genes), proton ATPase (4 genes), iron deficiency-induced transcription factors (3 genes), zinc transporter (3 genes), metal transporter (3 genes), and oligopeptide transporter (1 gene) (Table 2). Among the 40 DEGs, we found that they came from many modules in WGCNA, of which 9 genes from MEturquoise module, were FRO8 (Bra023239), VIT2 (Bra008277, Bra015718), VIT1 (Bra012271, Bra025879), YSL2 (Bra009756), HA (Bra027752, Bra040512), and ZIP2 (Bra020314). As a result, we assume that these nine genes are important in iron uptake and transport.

Based on the functional annotation of 8148 DEGs, 38 DEGs implicated in photosynthesis were searched, as yellowing of seedlings under iron deficiency stress could impair photosynthesis. These genes include 10 chlorophyll a/b binding protein, 3 carotenoids, 3 glutamine-tRNA reductase, 2 vetiver diphosphate reductase, and 19 light system I and II subunit protein genes. We discovered 31 genes that are part of the MEbrown module among these 38 genes (Supplementary Table 6). Thus, we hypothesize that these 31 genes may play a critical role in controlling leaf color in seedlings under iron deficiency stress, based on the highly significant positive correlation between Chlab and MEbrown modules (Figure 6).

Sixty-seven gene families were discovered to be involved in the control of iron deficiency stress by examining differentially expressed transcription factors, including the WRKY, MYB, AP2/ERF-ERF, bHLH, NAC, C2H2, bZIP, HB-HD-ZIP, and GARP-G2-like families (Supplementary Figure 2). The WRKY, MYB, ERF, bHLH, NAC, C2H2, and bZIP families account for 50.35% of the genes. Among them, 67 bHLH family genes were shown to be involved in iron deficiency stress (Supplementary Figure 2 and Supplementary Table 8). Given the importance of bHLH family genes in iron stress, we utilized STRING to examine the interaction network of 67 bHLH transcription factors, 40 iron absorption and transport-related proteins in Table 2, and the link between the two categories of proteins. According to the results of the protein interaction network prediction, these proteins were split into five categories, each of which was represented by a distinct color (Figure 7). One category was formed by seventeen iron absorption and transport proteins and one bHLH transcription factor (Bra039926) (purple). Four iron absorption and transport proteins, as well as 10 bHLH genes, were grouped (orange). Nine iron absorption and transport proteins were grouped in one category (dark green), whereas the remaining bHLH family genes were split into two groups (light blue and green). Three FIT genes (Bra000495, Bra011972, and Bra034392) belonging to the bHLH family transcription factors, which are iron deficiency-induced transcription factors, were discovered in the orange interaction network diagram. Three FIT genes had 83.75 to 86.88% sequence similarity with Arabidopsis FIT (bHLH029). Bra014657, Bra014658, Bra014659, Bra014971, Bra016946, and Bra004609 are the three FIT proteins that interact with Bra014657, Bra014658, Bra014659, Bra014971, Bra016946, and Bra004609 (Supplementary Figure 3). Since Arabidopsis FIT (bHLH029) interacts with bHLH038 and bHLH100 to form heterodimers. Similarly, Arabidopsis FIT (bHLH029) starts the expression of FRO2 and IRT1 genes in roots to promote iron absorption (Colangelo and Guerinot, 2004; Wang et al., 2013). While Bra014657, Bra014658, Bra014659, and Bra014971 are homologous genes of Arabidopsis bHLH38 and Bra016946 and Bra004609 are homologous genes of bHLH100. As a result, we hypothesized that the three FIT proteins, in conjunction with other bHLH transcription factors, would help maintain the iron balance in non-heading Chinese cabbage.

To ensure the accuracy of the transcriptome data, we used qRT-PCR to confirm the 9 candidate genes. The results revealed that these 9 genes were consistently up-regulated and down-regulated in roots and leaves, which matched the transcriptome sequencing findings (Supplementary Table 9). The Pearson correlation coefficient (R²) was used to analyze the correlation between transcriptome data and qRT-PCR data, and the results showed that the R² was 0.8025, indicating that the RNA sequencing data and qRT-PCR data were positively correlated (Supplementary Figure 4). We conclude that these differentially expressed genes may play an important part in iron deficiency stress based on our findings.

When plants are exposed to abiotic stress, reactive oxygen species (ROS) build up in the cells, causing catastrophic damage. ROX in plants, for example, damage the photosynthetic organs of apple leaves when there is an iron deficiency (Cheng et al., 2020). Antioxidant enzyme systems (such as SOD, POD, and CAT) were discovered to be able to scavenge ROS (Liu et al., 2016; Aung and Masuda, 2020; Fan et al., 2021). Under iron deficiency stress, DEGs for POD, SOD, and CAT were also identified in the roots and leaves of non-heading Chinese cabbage (Table 1). There were 58 POD DEGs in the roots of non-heading Chinese cabbage under iron shortage stress, with 47 POD genes considerably repressed and 11 genes significantly upregulated. Iron deficiency stress generated three SOD genes and two CAT genes in the leaves (Table 1). These genes were found to be involved in peroxisome biosynthesis (ko00940) and phenylpropanoid biosynthesis (ko00940) in the KEGG pathway (ko04146). Hence, this is assumed that these antioxidant enzymes or secondary metabolites interact with ROS in many tissues, improving iron deficiency stress tolerance.

According to research findings, when individuals are iron deficient, the amount of IAA in their sunflower roots increases dramatically (Marschner et al., 1986). Exogenous IAA increased ethylene and NO levels in plants, whereas the ethylene signal transcription factors EIN3 (ETHYLENE IN-SENSITIVE3) and EIL1 (EIN3-LIKE1) worked together with FIT to positively regulate AtFIT1 transcription, boosting iron absorption (Brumbarova et al., 2014). EIN3-like3 (Bra003831) expression was also shown to be highly up-regulated in leaves in this investigation (Supplementary Table 5). Furthermore, we discovered that numerous auxin-related genes were implicated in the response in roots and leaves under iron shortage stress (Supplementary Table 5). As a result, we hypothesized that IAA may play a signal transduction role in the iron deficiency stress of non-heading Chinese cabbage leaves. Furthermore, we discovered that under iron deficiency stress, the expression of one JAR gene (Bra034205) related to jasmonic acid signal was significantly down-regulated in the roots and leaves of non-heading Chinese cabbage, which is similar to JA as a negative regulator involved in regulating iron uptake in Arabidopsis (Brumbarova et al., 2014; Kobayashi et al., 2016; Cui et al., 2018). We deduce from the above research that the iron nutrition regulatory signal in non-heading Chinese cabbage under iron deficiency stress is highly complex and that a range of hormone signaling molecules is involved.

The control of iron homeostasis in plants necessitates the involvement of several iron absorption and transport-related proteins and regulatory factors (Li et al., 2021). Ferric reduction oxidase (FRO2), iron transporter gene (IRT1), and H⁺-ATPase (HA) protein are among the proteins involved in iron uptake and transport on the cell membrane, as shown in Figure 8. FRO2 and IRT1 were the first Arabidopsis thaliana genes to be cloned in the 1990s. Under iron stress, they are engaged in iron intake and transport (Eide et al., 1996; Robinson et al., 1999). Under iron deficient conditions, cucumber and Arabidopsis induced a significant number of HA gene expressions, which played an important role in response (Santi et al., 2005; Santi and Schmidt, 2009). The expression levels of four FRO, five IRT, and one HA genes in the roots of non-heading Chinese cabbage were likewise found to be considerably up-regulated under iron deficiency stress in this study. Under iron deficiency stress, however, the expression levels of these genes in the leaves of non-heading Chinese cabbage were not significantly elevated or inhibited (Figure 8). Given that plants move the iron from underground to aboveground portions, we predicted that under iron deficiency, the leaves of non-heading Chinese cabbage were inhibited through the FRO2, IRT1, and HA genes, reducing iron absorption.

AtVIT1 was discovered to be involved in the transport of iron ions to vacuoles in Arabidopsis (Kim et al., 2006), whereas AtNRAMP3 was found to be involved in the transport of iron ions from vacuoles to cytoplasm and was activated under iron deficient stress (Lanquar et al., 2005). Under iron deficiency stress, 5 VITs and 3 NRAMPs were discovered in this study (Figure 8 and Table 2). Thus, we suggested that the plant would use VIT and NRAMP on the vacuole membrane in roots and leaves to cooperatively regulate the steady state of iron ions.

Ferritin (FER) is a polymeric iron storage protein that can store up to 4500 iron atoms in its central cavity and is essential for cell iron homeostasis (Zielińska-Dawidziak, 2015). In the early stages of photosynthesis, ferritin in leaves serves as an iron source for the synthesis of iron-containing proteins (Briat and Lobréaux, 1997). In Arabidopsis, four ferritin genes (AtFER1-4) were discovered for the first time. Only AtFER2 was found in seeds, while AtFER1, AtFER3, and AtFER4 were found in vegetative and reproductive tissues (Petit et al., 2001). Ferritin deficiency in Arabidopsis leaves reduces leaf development and CO₂ fixation (Ravet et al., 2009). Under iron deficiency stress, the expression of three FER genes in roots and leaves was dramatically down-regulated, implying that iron deficit had a considerable impact on plant iron storage. To manage internal iron homeostasis, multi-gene co-expression was necessary to store and distribute iron ions between vacuoles and cytoplasm.

FIT (FER-LIKE FE DEFICIENCY-INDUCED TRANSCRIPTION FACTOR) is a key player in keeping iron levels in plant roots stable and limited (Colangelo and Guerinot, 2004; Yuan et al., 2005). FIT activity is produced not only by ZAT12, DELLA, and jasmonic acid but also by the interaction between bHLH and BTSL2 proteins, according to research (Birte and Petra, 2020). In Arabidopsis thaliana, FIT (bHLH029) formed heterodimers with the bHLH Ib subfamily members bHLH038, bHLH039, bHLH100, and bHLH101, and subsequently induced the expression of FRO2 and IRT1 genes in roots to enhance iron absorption (Colangelo and Guerinot, 2004; Wang et al., 2013). The level of Fe²⁺ in stems rose after overexpression of bHLH039 in Arabidopsis thaliana, which was linked to the inhibition of FRO3 and NRAMP4 and the significant induction of IRT1, FRO2, At3g07720, and IRT2 (Naranjo-Arcos et al., 2017; Birte and Petra, 2020). This finding implies that the bHLH Ib protein responds in a cascade to Fe-deficiency signals and that iron acquisition and dynamic balance in plants are regulated by a complex network. In this study, we discovered three FIT genes in the roots and leaves of non-heading Chinese cabbage under iron deficiency stress. These three FIT proteins interacted with bHLH38 analogous proteins (Bra014657, Bra014658, Bra014659, and Bra014971) and bHLH100 homologous proteins (Bra016946 and Bra004609) in Arabidopsis, according to the protein interaction network (Figure 7 and Supplementary Figure 3). OsbHLH064 and bHLH IVc transcription factors (OsbHLH057/PRI4, OsbHLH058/PRI2, OsbHLH059/PRI3, and OsbHLH060/PRI1) are also regulatory factors for iron absorption in rice (OsbHLH057/PRI4, OsbHLH058/PRI2, OsbHLH059/PRI3, and OsbHLH060/PRI1) (Kobayashi et al., 2019; Zhang et al., 2020). Iron absorption in tomatoes is controlled by Ib SlbHLH068 and FER (Du et al., 2015). Furthermore, under iron deficiency stress, the transcription levels of these six bHLH transcription factors in roots and leaves were dramatically up-regulated. Given the discovery of 67 bHLH transcription factors, we predicted that bHLH transcription factors played a key regulatory role in t iron deficiency stress.

Plants with an iron deficit turn green due to a drop in chlorophyll levels (Donnini et al., 2009). Although the chlorophyll a-oxygenase (CAO) gene was dramatically upregulated in the chlorophyll metabolic pathway under iron deficiency stress, the glutamine-tRNA reductase (Glu-TR), co-proporphyrinogen oxidase (CPOX) gene, and Mg-chelating enzyme H subunit (MgCh) gene were all strongly suppressed (Figure 8). As a result, we hypothesized that the combined effect of these genes resulted in a considerable drop in chlorophyll a/chlorophyll b in non-heading Chinese cabbage leaves under iron deficiency stress. Photosynthetic activity and leaf chlorophyll were both reduced. The reduction of the light capture complex (LHC) is linked to this phenomenon (Donnini et al., 2009). Light capture of chlorophyll a/b binding protein (antenna protein) has been shown to improve light absorption and excitation energy transfer in studies (Li et al., 2008). Under iron deficiency stress, the expression of four chlorophyll a/b binding protein genes was dramatically down-regulated in non-heading Chinese cabbage leaves. In addition, 19 photosynthesis-related genes, including subunit I and subunit II, were shown to be considerably down-regulated in leaves (Supplementary Table 7).