To identify the ACR gene sequences in maize, we used the Ensembl website (http://plants.ensembl.org/index.html) to download the B73 genome database, which included whole-genome files and gene annotation files (gff3). Based on the ACR domain (PF01842) in Pfam (http://pfam.xfam.org/), we employed a hidden Markov model (HMM) to query the ACR protein sequences throughout the entire B73 genome by using default parameters in HMMER software. The ZmACR proteins were identified using the Hummsearch program in the Linux system 5.2.38 (Jeanmougin et al., 1998). Subsequently, the Pfam, SMART (Simple Modular Architecture Research Tool, http://smart.embl-heidelberg.de) and Conserved domain (CDD, http://www.NCBI.nlm.NIH.gov/Structure/cdd/wrpsb) were utilized to confirm and eliminate redundant sequences. Subsequently, ExPASy Proteomics Server software (http://www.expasy.org) was used to predict the molecular weights (MWs), isoelectric points (PIs), and amino acid lengths of all identified ACR genes in maize.

A phylogenetic tree was constructed using the Maximum Composite Likelihood (MCL) model of neighbor-joining (NJ) algorithm in MEGA7 software, and 1,000 iterations of bootstrapping were performed. The Evolview (http://www.omicsclass.com/article/671) website was browsed to further modify and enrich the evolutionary tree. Motif analysis based on ZmACR protein sequences was carried out using MEME Suite 5.3.3, and the detailed information of the motif was subsequently viewed through the MEME website (http://meme-suite.org/). Finally, TBtools software was used to edit and save the retrieved images (Chen et al., 2020).

The Gene Structure Display Server (GSDS, http://gsds.gao-lab.org/) was used to analyze the gene structures of the ACR family genes. To explore the regulation of gene expression, the 1.5 kb sequence upstream of the initiation codon of each ZmACR gene was extracted to analyze the cis-acting elements of these genes. The Plant CARE website (Cis-acting Regulatory Element, http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) was used to further analyze the cis-acting elements according to the obtained 1.5 kb sequences, and the obtained cis-acting element information was subsequently mapped through the GSDS website. The distribution of cis-acting elements within the promoters was visualized with TBtools (Chen et al., 2020).

The chromosomal localization of ZmACR genes was determined from whole-genome files and gene annotation files, and the repeat pattern of each ZmACR gene was analyzed via BLAST. Then, we used MapChart and Adobe Illustrator CS6 software to construct a map of the chromosomes. Ka/Ks represents the ratio between the nonsynonymous substitution rate (Ka) and the synonymous substitution rate (Ks) of two protein-coding genes. After alignment, the synonymous substitution rate (Ks) and nonsynonymous substitution rate (Ka) of nucleotides in ZmACR family genes, as well as their ratios, were calculated by the KaKs_Calculator2.0 tool. The threshold for P-value of KA, KS and their ratio was set to 1×10^-10. MCScan toolkit (V1.1) (Wang et al., 2012; VanBuren et al., 2018) implemented in python (https://github.com/tanghaibao/jcvi/wiki/MCscan-(Python-version)) was used to analyze the collinearity of the ZmACR genes. The collinear genes were identified in the whole genome based on the sequence alignment of proteins. Finally, the results were visualized using Circos (http://circos.ca/) software.

The secondary structure of ZmACR proteins was predicted using the online website Prabi (https://npsa-prabi.ibcp.fr). In order to delve deeper into the protein structure of the ZmACR proteins, their tertiary structures were predicted based on their protein sequences. Subsequently, the online tools SWISS-MODEL (https://swissmodel.expasy.org/) was amployed to generate the 3D models of ZmACR proteins through homologous protein modeling method (Waterhouse et al., 2018).

We utilized the CELLO online website (http://cello.life.nctu.edu.tw/) to predict the subcellular localization of the ZmACR genes based on their protein sequences (Yu et al., 2006). By submitting the protein sequences to the website, we obtained potential organelle locations. The results were then visualized using TBtools. To verify the accuracy of the bioinformatic prediction results, we randomly selected 4 ZmACR genes and investigated their expression localization. We amplified the full-length CDSs of the genes using gene-specific primers ( Supplementary Table S1 ) and KOD One TM PCR Master Mix (TOYOBO, Osaka, Japan). The vector was linearized using HindIII and BamHI (NEB, Nebraska, USA). Subsequently, the CDSs of the 4 ZmACR genes were subcloned and inserted into the destination vector under the control of the CaMV 35S promoter using the ClonExpress^® II One Step Cloning Kit (Vazyme, Nanjing, China). Finally, the vectors containing the target genes and GFP were subsequently transformed into Agrobacterium GV3101, which was subsequently used to infect tobacco leaves. The fluorescence was observed using a Leica TCS SP8 (Mannheim, Germany) confocal microscope imaging system.

Total RNA was extracted from leaves using a CWBIO Ultrapure RNA Kit (DNase I) (CWBIO, Beijing, China), and the integrity of the RNA was evaluated via 1.2% agarose gel electrophoresis. First-strand cDNA synthesis was performed using HiScript^® III RT SuperMix (+gDNA WIper) (Vazyme, Nanjing, China). Quantitative real-time polymerase chain reaction (qRT-PCR) was conducted in triplicate for each sample using ChamQ SYBR qPCR Master Mix (Low ROX Premixed) (Vazyme, Nanjing, China). The gene-specific primers for qRT-PCR listed in Supplementary Table S1 were used to amplify ZmACR genes by PCR under the following conditions: predenaturation at 95°C for 30 s, denaturation at 95°C for 10 s, annealing at 60°C for 30 s, and 40 cycles (Ge et al., 2022). An ABI 7500 FAST Real-Time PCR System was used for qRT-PCR. Maize ACTIN was used as a control for normalization between samples ( Supplementary Table S1 ). Comparative threshold cycling was used to calculate relative transcription levels (Livak and Schmittgen, 2001).

After exposure to salt treatments, ZmACR5 expression in leaves increased the most, approximately 15-fold. These findings suggested that ZmACR5 may play a crucial role in the response of maize to salt stress. To further investigate the function of ZmACR5, we ectopically overexpressed ZmACR5 in Arabidopsis using the PWM101 vector and Agrobacterium-mediated floral-dip method (Clough and Bent, 1998). Two stable overexpression lines (OE1 and OE2) were identified after hygromycin screening and expression identification. The growth conditions and stress treatments were applied to both the transgenic and wild-type plants as previously described (Dong et al., 2012), with minor modifications. Specifically, surface-sterilized seeds harvested from both T3 transgenic lines and the wild type were grown on 1/2 MS agar media for approximately ten days. The uniform plants were transferred to small pots, filled with nutrient soil and vermiculite, in a greenhouse, each containing ten plants. The temperature and light cycle were set at 22°C with a 16 h/8 h (light and dark) photoperiod. After another 7 days of growth, we divided the wild-type and overexpressing plants into three groups. The plants in the control group were watered with regular water, while those in the two treatment groups were watered with 100 or 150 mmol/L salt solution, with 30 ml of solution added to each pot each day. This process continued for 10 consecutive days. Finally, the leaf length and width, as well as the fresh and dry weights above ground, of each plant were measured to evaluate salt tolerance.

To grow the maize plants, plump seeds were picked and treated with 75% ethanol for 1 min, followed by sterile water for 6 h. The seeds were subsequently placed on moist sterile filter paper and incubated in the dark at 28°C for approximately 2 days. Once the radicles reached a length of 1-2 cm, the plants were transplanted into small pots filled with nutrient soil and vermiculite (at a ratio of 3:1) for growth in a greenhouse. The temperature and light cycle were set at 26°C with a 14 h/10 h (light and dark) photoperiod. When the maize plants had grown to the two-leaf stage, they were transferred to a hydroponic device and grown in improved Hoagland nutrient solution until the plants reached the four-leaf stage (the nutrient solution was changed every two days). Maize plants with consistent growth were selected and divided into two groups: one group served as a blank control without any treatment, while the other was treated with 250 mmol/L NaCl solution. The maize leaves from both groups were collected at 24 h after salt treatment, immediately frozen in liquid nitrogen and stored in an ultralow temperature refrigerator at -80°C for subsequent experiments.

We used Pfam (PF01842) to determine the putative members of the ACR gene family in maize. A total of 28 candidate ZmACR genes were identified from the B73 genome database. Subsequently, we analyzed various characteristics of these genes, including amino acid length, molecular weight and isoelectric point. The amino acid length of the identified ZmACRs ranged from 105 (ZmACR11) to 620 (ZmACR15). The molecular weights of the products varied from 11513.8 Da (ZmACR11) to 67475 Da (ZmACR15). Additionally, most of the predicted isoelectric points of these 28 ZmACR genes were acidic proteins, except for ZmACR15, ZmACR17 and ZmACR21, whose values ranged from 3.92 to 9.22; these values provided important information for subsequent protein separation, purification and electrophoresis. The detailed information is listed in Supplementary Table S2 .

Twenty-eight ZmACR genes were unevenly distributed on nine chromosomes, with the exception of chromosome 9 ( Supplementary Figure S1 ). We assigned numbers to the 28 ZmACR genes based on their physical positions, from ZmACR1 to ZmACR28. Most of the ZmACR genes were located at the proximal or distal end of chromosomes. Chromosome 4 contained the highest number of ZmACR genes (6, approximately 21.4%), followed by chromosomes 1 and 2 (5 each, approximately 17.9%). Chromosome 5 housed 3 ZmACR genes, while Chromosome 8 possessed the fewest ZmACR genes (1, approximately 3.6%). The remaining chromosomes (3, 6, 7, and 10) contained 2 ZmACR genes (approximately 7.1%).

To gain further insight into the evolutionary relationship and categorization of ACR genes across different species, 58 ACR genes, including 28 in maize, 11 in Arabidopsis, 10 in rice and 9 in sorghum, were used to construct an unrooted phylogenetic tree via the neighbor-joining (NJ) method in MEGA7 software ( Figure 1 ). These ACR sequences were subsequently clustered into three subgroups. The largest subgroup, Group III, comprised 13 ACR genes of maize, 9 ACR genes of rice, all 9 ACR genes of sorghum and 7 ACR genes of Arabidopsis. Group I consisted of 12 ACR family members of maize and 2 ACR family members of Arabidopsis. Group II included 3 ACR genes from maize, 2 ACR genes from Arabidopsis and 1 ACR gene from rice. In the evolutionary tree, ACR genes consistently exhibited closer evolutionary relationships between sorghum and maize, both of which were C4 plants. In addition, compared to those in other species, the ACR genes in Arabidopsis tended to have separate branches, suggesting that there was a relatively distant evolutionary relationship between Arabidopsis and the other three species. This finding suggested that there was divergence of ACR genes between monocotyledonous and dicotyledonous plants during evolution. Furthermore, the number of ACR members in maize was significantly greater than that in the other three species, likely due to the large genome of maize, which led to the expansion of ACR genes during genome evolution.

To understand the structural features of the ZmACR genes, gene structure and conserved motif analyses were also conducted. The results suggested that genes within the same group had similar organization profiles, and most genes contained multiple introns, indicating a particular phylogenetic state. The number and length of untranslated regions (UTRs), exons and introns exhibited significant divergence ( Supplementary Figure S2A ). The genes in Groups I and II contained a greater number of exons, ranging from 5 to 16, while the number of exons in Group III ranged from 3 to 9, except for ZmACR7 (18) and ZmACR15 (13), which displayed relatively simple genetic structures. Most of the ZmACR genes contained UTRs, although some genes in Groups I (ZmACR9 and ZmACR11) and III (ZmACR15) lacked UTRs, which may affect the regulation of gene expression.

In general, similar amino acid sequences could lead to similar gene functions within the same group. To better understand the conserved amino acid sequences of ZmACR proteins, we conducted protein sequence alignment and motif analyses. A total of 10 motifs were identified and labeled after motifs 1 to 10 ( Supplementary Figures S2B, C ). With the exception of ZmACR7 and ZmACR15 in Group III, most of the ZmACR proteins contained conserved motifs, with almost all the genes containing motif 2 and motif 5. The genes in Group II had the fewest motifs, with only Motifs 2 and 9, which was quite different from the other subgroups. All genes in Group I had Motif 2 and Motif 5, which was similar to most genes in Group III. Genes in Group III possessed varying amounts of Motif 2, 5 and 10 combinations. Interestingly, although genes in different subgroups contained different motifs, genes within the same subgroup appeared to share similar motifs. In summary, the conserved motif compositions and distributions of ZmACR proteins exhibited significant differences among the different subgroups, which suggested that these ZmACR proteins might have undergone functional group divergence during evolution.

Genome duplication has long been considered as one of the main drivers of evolution. During gene evolution, tandem repetition and large fragment duplication are the driving forces behind the expansion of gene families within genomes. Previous studies have shown that gene pairs with a sequence similarity of more than 70% and within a distance of 100 kb are considered tandem repeats (Wei et al., 2022a). Interestingly, no tandem duplication events were detected in the ACR gene family in maize. In contrast, 4 pairs of ZmACR genes, ZmACR8/ZmACR25, ZmACR12/ZmACR27, ZmACR17/ZmACR21 and ZmACR23/ZmACR26, were identified as segmented repeats ( Figure 2A ). These results suggested that the primary driver of ZmACR gene diversity was segmental duplication events.

The ratio of Ka (nonsynonymous substitution) to Ks (synonymous substitution) is a critical parameter that reflects the selection pressure of species (Zhang et al., 2006). To clarify the selection pressure on ZmACR genes, we calculated Ka, Ks and their Ka/Ks ratio using KaKs_Calculator2.0 software. The results showed that only two pairs of genes were significantly selected ( Supplementary Table S3 ). The Ka/KS ratios of the two gene pairs were significantly less than 1, indicating negative selection and suggesting that the ACR gene family in maize was possibly subjected primarily to purifying selection during the evolutionary process.

To further investigate the evolutionary history of the ACR genes between maize and 3 other species, one dicotyledon (Arabidopsis) and two monocotyledons (rice and sorghum), collinearity analysis was performed ( Figures 2B-D ). The collinearity relationships revealed 3, 7 and 8 pairs of ACR genes between maize and Arabidopsis, rice, and sorghum, respectively, and 8 ZmACR genes were involved, which further confirmed the close evolutionary relationship between maize and sorghum. The number of collinear gene pairs between maize and rice, sorghum was greater than that between maize and Arabidopsis, which was similar to the evolutionary relationships between dicots and monocots. These results clearly suggested that ACR genes may have existed before the divergence of monocotyledons and dicotyledons and further verified the evolutionary differences between monocotyledons and dicotyledons.

The role of proteins is closely linked to their structure. In order to gain a more profound insight into ZmACR proteins, we carefully selected 8 ZmACRs from distinct subgroups for the prediction of their secondary structure and the construction of 3D models ( Supplementary Figure S3 , Supplementary Table S4 ). Our analysis revealed that alpha helices, extended strands, beta turns, and random coils constitute the secondary structure of ZmACR proteins ( Supplementary Table S4 ). However, the distribution of these secondary structures varied among different protein subgroups. In Group I (e.g., ZmACR26 and ZmACR27), random coils represented the largest proportion (approximately 40%), followed by alpha helices (32.81%-39.10%) and extended strands (14.05% - 20.73%). Conversely, proteins in Group II, such as ZmACR17 and ZmACR21, predominantly exhibited alpha helices and random coils, accounting for 32.08% ~ 34.99% and 33.1% ~ 38.05%, respectively. The proportion of extended strands in Group II proteins was notably higher compared to Group I and III. In contrast, Group III (e.g., ZmACR3, ZmACR5, ZmACR15, and ZmACR22) showed alpha helices as the major component (36.11% ~ 40.9%), followed by random coils. Beta turns were the least abundant among all secondary structures of ZmACR proteins ( Supplementary Table S4 ). Variations in the secondary structure compositions within these ZmACR protein subgroups contributed to the alterations in their three-dimensional structures ( Supplementary Figure S3 ). The homology modeling of ZmACR proteins provided a clear view of the composition and arrangement of their secondary structures ( Supplementary Figure S3A ). The GMQE (global model quality estimate) values ranged from 0.66 (ZmACR21) to 0.83 (ZmACR26), with sequence identity between an ACR protein and its homologous model varying from 85.52% to 100%, affirming the accuracy of 3D structure prediction for these ZmACR proteins. The secondary structures of ZmACR proteins, including α-helices, random coils, β-turns, and extended chains, are further intricately folded through the interaction of side chain groups, leading to the formation of a compact spherical spatial structure maintained by various secondary bonds. The stability of the tertiary structure is essential for the biological activity of proteins.

The protein expression site is closely associated with its function. Therefore, we used the online software CELLO to predict the subcellular localization of ZmACR proteins to gain insight into the potential functions of ZmACR genes ( Supplementary Figure S4 ). The results showed that the genes in different subgroups exhibited varying expression locations. Specifically, the ZmACR genes in Group I were predominantly expressed in the cytoplasm, followed by the nucleus and chloroplast. The three ZmACR genes in Group II exhibited a high likelihood of being expressed in chloroplasts and mitochondria. The genes in Group III were predicted to be chiefly expressed in the nucleus and cytoplasm, with chloroplasts and mitochondria being secondary sites. Overall, genes within the same subgroup tended to be expressed at the same sites, suggesting functional similarity.

To validate the prediction results, 4 randomly selected ZmACR genes from each of the three subgroups were individually attached to a GFP tag and driven by the CaMV 35S promoter. These constructs were then transiently expressed in the epidermal cells of Nicotiana benthamiana leaves to determine their subcellular localization. In N. benthamiana leaf epidermal cells, fluorescence signals corresponding to the GFP control were observed in almost all parts of the cells, suggesting normal expression of the GFP ( Supplementary Figure S5 ). The GFP fluorescent signals of the ZmACR10 and ZmACR20 fusion proteins were exclusively detected in the chloroplast and were not observed in the cytoplasm, indicating that these two proteins may be localized in the chloroplast ( Figure 3 ). The GFP signals of the ZmACR13 and ZmACR23 fusion proteins not only colocalized with the cyan membrane-anchored marker PIP 2A-CFP after transient expression in tobacco leaves but also extended to the inner part of the cell membrane, suggesting that both genes were likely localized to the cytoplasm ( Figure 3 ). In addition, the ZmACR13 fusion protein was also found to be localized to the nucleus, as evidenced by the presence of several dot distributions of the GFP fluorescence signal. These experimental results were largely consistent with the predicted results, thereby confirming the reliability of the predictions.

Cis-acting elements play pivotal roles in regulating transcription during numerous physiological processes in plants, such as plant growth, development, and biotic and abiotic stress responses. To explore the potential molecular function of the ZmACR gene family, the 1.5 kb promoter sequence of each ZmACR gene was extracted and submitted to PlantCARE to predict the cis-acting elements. In total, 81 distinct cis-acting regulatory elements associated with plant growth, development and abiotic stress responses were identified ( Supplementary Figure S6 , Supplementary Table S5 ). Several fundamental elements, such as the CAAT-box and TATA-box, were present in the majority of ZmACR promoters. Moreover, many ZmACR promoters contained cis-acting elements associated with light reactions (e.g., AE-box, G-box, GA-Motif, GATA motif, GT1-motif, and TCT-motif), hormone response elements (e.g., ABRE, STRE and MYB) and drought induction elements (e.g., MBS and W-boxes). These findings suggested that ZmACR genes might participate in hormone signaling pathways, biotic or abiotic stress responses, and other physiological processes.

To further investigate the potential functions of the ZmACR genes in maize growth and development, we analyzed the expression levels of 26 ZmACR genes (excluding ZmACR11 and ZmACR28) in 10 different tissues, namely, the embryo, endosperm, seed, root, shoot, leaf, ear, tassel, silk and anther tissues, using the transcriptome data of maize B73 lines from a previous study (Chen et al., 2014). Our findings revealed three distinctive expression patterns among the ZmACR genes ( Figure 4 ). First, the genes in the red box, such as ZmACR9, ZmACR26, and ZmACR17, were highly expressed in multiple tissues. Second, the genes in the purple box, such as ZmACR12 and ZmACR4, were moderately expressed in different tissues. Third, the genes in the green box were expressed in only a few specific tissues at moderate or low levels. For instance, ZmACR1 was moderately expressed in the roots but exhibited very low expression in other tissues ( Figure 4 ). The first two categories exhibited constitutive expression characteristics. Considering that many ACR genes in plants have been shown to be involved in regulating amino acid metabolism, which is crucial for the growth and development of different plant tissues, ZmACR genes might play a role in regulating plant growth and development in various tissues. Notably, the last category of ZmACRs exhibited specific tissue expression patterns, indicating that these genes might have unique functions in specific tissues.

Previous studies have demonstrated that ZmACR genes are involved in the response to abiotic stresses and plant hormones in Arabidopsis (Singh et al., 2018). The identification of cis-acting elements associated with the abiotic stress response at ZmACR promoters also indicated the potential role of these elements in the response to abiotic stresses. To further understand the function of ZmACR genes under abiotic stress conditions, we randomly selected 8 ZmACR genes from different subgroups and performed qRT-PCR to quantify their expression levels under salt stress ( Figure 5 ). Compared with those in the control group, 5 out of the 8 ACR genes (ZmACR4, ZmACR5, ZmACR13, ZmACR16 and ZmACR26) exhibited increased expression in roots, stems and leaves after 12 hours of salt treatment, suggesting their potential involvement in the response to salt stress. In particular, the expression level of ZmACR5 in leaves was upregulated the most, approximately 15-fold, indicating that ZmACR5 plays a significant role in the response to salt stress. Conversely, the expression level of ZmACR9 in roots and leaves significantly decreased under salt conditions. Interestingly, the expression of ZmACR9 in the stems significantly increased eightfold, suggesting that it may play a potential role in maize stem development. In addition, while the expression levels of ZmACR10 in roots and ZmACR20 in roots and stems did not significantly differ after salt treatment, their expression levels in stems or leaves significantly increased. In conclusion, these findings revealed that most ZmACR genes were involved in the response to salt stress but might operate through distinct mechanisms due to their varied expression patterns.

In light of the notable increase in ZmACR5 expression under salt stress conditions, we ectopically overexpressed this gene in Arabidopsis to further investigate its role in the response to salt stress. Through hygromycin screening and qRT-PCR, we successfully identified two stable T3 generation overexpression plants (OE1 and OE2), whose expression levels were approximately 620- to 259-fold greater than those of the wild type ( Supplementary Figure S7 ). We subsequently assessed the salt tolerance of the wild-type, OE1 and OE2 plants at two NaCl concentrations, 100 and 150 mmol/L. Under normal conditions, there were no significant differences in growth between wild-type Arabidopsis thaliana and the two overexpression lines ( Figure 6C ; Supplementary Figure S8 ). However, when exposed to salt stress, both the OE1 and OE2 plants exhibited diminished salt tolerance compared to that of the wild type plants at both 100 mmol/L and 150 mmol/L ( Figures 6A, B ). Specifically, compared with those of the control plants, the leaf length and width of both the OE1 and OE2 plants were significantly lower under the 100 mmol/L salt treatment, with leaf length decreasing by 0.8 and 0.6 cm and leaf width decreasing by 0.34 and 0.26 cm, respectively. Additionally, the fresh weight and dry weight of the above-ground plants also exhibited significant decreases under the 100 mmol/L salt treatment, with decreases of 0.165 g and 0.087 g for fresh weight, and 0.015 g and 0.010 g for dry weight, respectively ( Figure 6C ). The same phenotypes were observed at a salt concentration of 150 mmol/L, albeit to varying degrees ( Figure 6C ). These observations indicated that plants overexpressing the ZmACR5 gene were more susceptible to salt stress than the wild type plants. Therefore, we confirmed that ZmACR5 can negatively regulate salt stress in Arabidopsis.