Using the CRISPR/Cas9 (Clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9) system, the ZmSPX1 gene was knocked out in the maize inbred line KN5585 by the Agrobacterium tumefaciens-mediated transformation method. The mutation site was detected using ZmSPX1-KO-F and R primers ( Supplementary Figure S1 ; Supplementary Table S1 ). The construction method for maize overexpressing lines utilizes the homologous recombination method to introduce the coding sequence (CDS) of the ZmSPX1 gene into the vector pCAMBIA3301. Subsequently, A. tumefaciens-mediated transformation of the maize recipient line KN5585 was employed to generate transgenic plants. The results for the detection of overexpression and the specific primers used for detection are provided in Supplementary Figure S2 and Supplementary Table S1 , respectively. Seeds of 178 (low-Pi tolerant maize inbred line), A. thaliana (Columbia), and tobacco (Nicotiana benthamiana) were provided by the Maize Research Institute of Sichuan Agricultural University. A relatively low-Pi field site in Wenjiang farm (WJ, Chengdu, Sichuan Province, plain region, available Pi 22.8 mg/kg) of Sichuan Agricultural University was selected for the experiments on low-Pi treatment of maize inbred line 178. Each plot received two levels of Pi treatments (low- and normal-Pi conditions), with each treatment replicated thrice. The plots measured 3 m in length with 0.8 m between rows. Prior to reaching the five-leaf stage, plant thinning reduced the number to 14 plants per plot (58,000 plants/ha). In the normal-Pi treatment plots, the following fertilizers (Stanley Agriculture Group Co., Ltd.) were applied: 150 kg of urea, 900 kg of calcium superphosphate, and 350 kg of potassium chloride per hectare before planting; 130 kg of urea per hectare at the six-leaf stage; and 210 kg of urea per hectare as a side dressing prior to tasseling. The low-Pi treatment plots received the same fertilizer combination as the normal-Pi plots, with the exception of the calcium superphosphate. The seeds of A. thaliana were planted in nutrient soil and soil sterilized at 121°C for 20 min and cooled for 30 min. Then, seeds were cultured in nutrient soil at 23°C for 12 h of light and darkness until the A. thaliana mature.

Based on our previous research (Nie et al., 2021), ZmSPX homologous genes were identified with high similarity sequence: ZmSPX1 (GRMZM2G035579), ZmSPX2 (GRMZM2G171423), ZmSPX3 (GRMZM2G024705), ZmSPX4 (GRMZ2G122108), ZmSPX5 (GRMZM5G828488), ZmSPX6 (GRMZM2G065989), and ZmSPX7 (GRMZM2G083655). The ZmSPX family members were identified in maize, and the reliability of the ZmSPX CDS was predicted using BioXM 2.6 software. The conserved domain of ZmSPXs was verified using the National Center for Biotechnology Information (NCBI) search database (Marchler-Bauer et al., 2015). Amino acid sequences query homologous amino acid sequences of SPXs in maize, wheat, soybean, rape, rice, and Arabidopsis using the BLASTP program from the NCBI (http://www.ncbi.nlm.nih.gov); phylogenetic relationships were aligned using Mega (version 11.0.13) (Kumar et al., 1994). The gene IDs of the SPX gene family in other species were acquired from previous research studies (Secco et al., 2012; Yao et al., 2014b; Du et al., 2017; Kumar et al., 2019). Finally, ZmSPXs were studied for cloning and functional analyses based on protein comparison and transcriptome sequencing, respectively. The promoter region (2,000 bp) of ZmSPXs was utilized to predict the functional region of these genes using the BDGP online database (http://www.fruitfly.org/seq_tools/promoter.html) ( Supplementary Table S2 ), and the cis-acting elements within the promoter were predicted using the plantCARE web server (http://bioinformatics.psb.ugent.be/webtools/plantcare/html).

Total RNA was extracted from the leaves and roots of 178 inbred lines following the manufacturer’s protocol of Thermo Fisher Scientific, Life Technologies (TRIzol Reagent^®; the protocol can be found at https://www.thermofisher.com). The cDNA reverse transcription was performed using the PrimeScript™ II 1st Strand cDNA Synthesis Kit (Takara Bio Inc., Otsu, Japan). The CDS of ZmSPXs was then amplified using the cDNA as a template. The polymerase chain reaction (PCR) primers were designed based on the CDS of the B73 reference, targeting the specific sequences of the ZmSPX family members using Primer Premier 5 software. The forward and reverse primers are listed in Supplementary Table S3 . The PCR was performed in a 25-µL volume containing phanta max buffer 12.5 μL (Phanta Max Super-Fidelity DNA Polymerase, Nanjing Vazyme Biotech Co., Ltd., Nanjing, China), phanta max super fidelity 0.5 μL, dNTP 4 μL, cDNA 1 μL, 1 μL of each forward and reverse primer, and 7 μL ddH₂O. PCR was programmed for 3 min at 95°C followed by 30 cycles of 95°C for 15 s, annealing for 15 s, 72°C for 45 s, extension for 5 min at 72°C, and final 12°C for preservation. Then, PCR products were separated using 1% agarose gel, purified using a DNA purification kit, and cloned into pEASY^®-Blunt zero Cloning Vector (TransGen Biotech Co., Ltd., Beijing, China) according to the manufacturer’s protocol. Finally, positive clones were selected for sequencing by TsingKe Biological Technology Co., Ltd. (Beijing, China). The coding sequence of ZmSPXs was compared with that of the B73 reference inbred line using DNAMAN v6.0 software.

Different maize tissues were harvested in the silking stage from normal-Pi and low-Pi treatments. Total RNA was extracted from different maize tissues of 178 inbred lines including leaf, stem, anther, cornsilk, root, ear, and ear bract according to the TRIzol Reagent^® (Invitrogen, Carlsbad, CA, USA) manufacturer’s instructions, and reverse transcription was performed using PrimeScript™ II 1st Strand cDNA synthesis kit (Takara Bio Inc., Otsu, Japan). Relative quantitative results were calculated by normalization to the reference gene (GAPDH). At least three independent experiments were performed, and each experiment was performed in technical triplicate. All primers used in the qRT-PCR assay were designed in BEACON DESIGNER 7 and are listed in Supplementary Table S4 . qRT-PCR data were analyzed using the 2^−ΔΔCT method.

The full-length CDS region was cloned into plant expression vector pCAMBIA2300 for the ZmSPXs. The homologous recombination primers with enzyme cutting sites (SmaI and XbaI) were designed by CEDesignV1.04 software and are shown in Supplementary Table S5 . The subcellular localization was assayed in tobacco leaves according to the previously reported transit transformation method (Zhou et al., 2018; Sahito et al., 2020a, b). The transformed tobacco epidermal cells were observed using the A1R-si laser confocal microscope (LSCM, Nikon, Tokyo, Japan) to visualize the green fluorescent protein (GFP) fluorescent signals. The experiment was repeated at least three times to ensure consistency of results.

The full-length CDSs of ZmSPX family members ZmPHR1 (GRMZM2G006477) and ZmPHR2 (GRMZM2G162409) were amplified using gene-specific primers from the cDNA samples. The primers were designed using CE design v1.04 and Primer 5.0 to incorporate (EcorR I, BamH I, and Nde I) restriction sites, as specified in Supplementary Table S6 . Phanta Max high-fidelity DNA polymerase (Nanjing Vazyme Biotech Co., Ltd.) was used to amplify the targeted fragments, and then the amplified fragments were inserted into bait (pGBKT7) and prey (pGADT7) vectors. All possible combinations were co-transferred into the Y2HGold yeast-competent cell. pGBKT7-53 and pGBKT7-lam were used as positive and negative controls, respectively. To exclude the possible autoactivation of ZmSPX members, a control experiment was carried out by transformation of loaded bait and prey plasmids with empty prey and bait plasmids, respectively. After screening on solid DDO (−Leu/−Trp) medium for 2–4 days at 28°C, selected monoclonals were inoculated in liquid DDO (−Leu/−Trp) medium until OD600 = 0.3–0.5. These cultures were inoculated on QDO (SD/−Ade/−His/−Trp/−Leu/X-α-Gal) medium after 10-fold dilution. Results were observed after 3–5 days of incubation at 30°C.

Cloning of the full open reading frame (ORF) of ZmSPX1 into the vector pCAMBIA3300-35s-PROII MCS-bar and subsequent transformation into wild-type (WT) Arabidopsis produced several transformants through the floral dip method previously described (Clough and Bent, 1998). The seeds of the T0 generation were harvested and sown in the soil to select positive transgenic seedlings. Two-week-old seedlings of T1 plants were screened by spraying with Basta (1/1,000) solution. After the transgenic plants were harvested, DNA was extracted, and PCR was performed to confirm the positive ZmSPX1 transgenic lines.

The seeds of both WT and ZmSPX1 transgenic lines were sterilized with 75% (v/v) ethanol for 60–90 s and 2% sodium hypochlorite for 8–15 min, followed by several rinses with distilled water. Fifty seeds of both WT and transgenic lines were cultured on 1/2 normal MS agar medium, initially kept at 4°C for 2 days, and subsequently transferred to a greenhouse with a light cycle of 16 hours at 22°C and a dark cycle of 8 hours at 18°C for 7 days. After 7 days, seedlings were transferred to 1/8 normal MS agar medium and subjected to different Pi concentrations: 0 mmol/L (control), low Pi (0.1 mmol/L), normal Pi (1 mmol/L), and high Pi (10 mmol/L). The seedlings were kept in a controlled environment for 2 weeks. Phenotypic observations under different Pi concentrations were recorded after 2 weeks, and their roots were scanned using a WinRHIZO root-scanning method. All Arabidopsis seedlings were dried and digested for the detection of P and nitrogen concentration through a chemical continuous flow analyzer.

A total of 211 out of 360 inbred lines were screened from the current Southwest China breeding program (Zhang et al., 2016). These lines were used for association analysis of ZmSPX1 in maize. The genomic DNA was extracted from leaves at the seedling stage according to the cetyltrimethylammonium bromide (CTAB) method (Porebski et al., 1997). The genomic sequence of ZmSPX1 from the B73 inbred line was utilized as the reference sequence and obtained from the Maize Genomic Database (http://www.maizegdb.org). The targeted fragment was amplified using high-fidelity phanta max enzyme polymerase (Vazyme Biotech Co., Ltd.). Amplified fragments were sequenced by TsingKe Biological Technology Co., Ltd. Then, the targeted sequence of ZmSPX1 was compared with the reference genomic sequence of the B73 using DNAMAN v6.0 software, and the sequences of 211 maize inbred lines were trimmed neatly using Bio-Edit 7.1 software (Strable and Scanlon, 2009). Single-nucleotide polymorphisms (SNPs) and insertions/deletions (InDels) were identified in all tested inbred lines with a <0.05 minor allele frequency (MAF), and linkage disequilibrium (LD) between two polymorphic sites was generated using Tassel software v2.1 (Bradbury et al., 2007). Association analysis was carried out between SNPs and InDels and 22 phenotypic traits using Tassel software (Bradbury et al., 2007; Luo et al., 2019). The standard mixed linear model (MLM) including a population structure (Q) and kinship matrix (K) was chosen to detect the significant association of SNPs and InDels as described previously (Zhang et al., 2016; Luo et al., 2019). The association of sites was considered significant at p < 0.05, and the calculated p-values were converted into −log₁₀ ^{( p value)}.

The SPX homologous genes ZmSPX1 (GRMZM2G03557), ZmSPX2 (GRMZM2G171423), ZmSPX3 (GRMZM2G024705), ZmSPX4 (GRMZM2G122108), ZmSPX5 (GRMZM5G828488), ZmSPX6 (GRMZM2G065989), and ZmSPX7 (GRMZM2G0836555) were predicted in our previous study, and genes were located on different chromosomes ( Supplementary Table S7 ). However, in the fourth edition of the maize genome, ZmSPX1 and ZmSPX7 were merged into a single gene. Additionally, preliminary transcriptome results indicated that ZmSPX7 (GRMZM2G0836555) did not respond to low-Pi stress ( Supplementary Table S7 ). Hence, in subsequent experiments, the ZmSPX1–6 in the third version of the maize genome were exclusively analyzed. Furthermore, ZmSPX1, ZmSPX2, ZmSPX3, ZmSPX4, ZmSPX5, and ZmSPX6 were amplified in maize according to the targeted CDS by PCR. Specifically, the CDS length of ZmSPX1 was 354 bp, encoding a protein comprising 117 amino acids. Correspondingly, ZmSPX2, ZmSPX3, ZmSPX4, ZmSPX5, and ZmSPX6 have CDS lengths of 846 bp, 687 bp, 996 bp, 765 bp, and 759 bp, encoding proteins of 281, 228, 331, 254, and 252 amino acids, respectively. Simultaneously, SPX genes of wheat, soybean, rape, rice, and Arabidopsis were also searched for in the database (Secco et al., 2012; Yao et al., 2014b; Du et al., 2017; Kumar et al., 2019). Among them, Arabidopsis has four SPX genes (AtSPX1–4), rice has six SPX genes (OsSPX1–6), soybean has nine SPX genes (GmSPX1–9), and rape and wheat have 11 and 15 SPX genes, respectively ( Supplementary Table S8 ). To deepen our understanding of the evolutionary relationships between these SPX proteins, a phylogenetic tree based on the SPX proteins in these plants was constructed ( Figure 1 ). The six SPX genes of maize were distributed across different clades. Notably, ZmSPX5 and ZmSPX6 displayed the closest evolutionary relationship, clustering within the same clade but occupying distinct branch points ( Figure 1 ). Interestingly, the closest evolutionary relationship existed between the SPX genes in maize and rice. Specifically, ZmSPX5 and OsSPX5, ZmSPX3 and OsSPX3, and ZmSPX1 and OsSPX1 were all located at the same branch point ( Figure 1 ). Furthermore, while ZmSPX4 and OsSPX4 were positioned at different branch points on the evolutionary tree, they belonged to the same clade and were situated at adjacent branch points ( Figure 1 ). Remarkably, ZmSPX2 showed a close evolutionary relationship with BnaA3SPX3 ( Figure 1 ).

Generally, the results showed that ZmSPX gene family members were associated with abscisic acid response, light reaction, methyl jasmonate (MeJA) reaction, low-temperature response, auxin response, and regulation of zein metabolism cis-acting elements ( Supplementary Table S9 ). Specifically, ZmSPX gene family members also possess unique cis-acting regulatory elements. For instance, ZmSPX1 contains response elements for gibberellin and salicylic acid. The ZmSPX2 gene responds to light and is also involved in defense and stress responses. ZmSPX3 is involved in the light responsiveness. ZmSPX4 contains a conserved DNA module (ATCT-motif) involved in photoreaction. ZmSPX5 is involved in anaerobic induction. ZmSPX6 contains DNA-binding protein binding sites and responds to light. It also contains MYB binding sites involved in the regulation of flavonoid biosynthesis genes. Previous studies have demonstrated that SPXs can interact with the MYB domain and the CC domain of PHRs (Jia et al., 2023). Additionally, our results also indicated that ZmSPX gene family members also contain MYB binding sites or MYB recognition sites ( Supplementary Table S9 ). All of these findings suggested that ZmSPXs may play a role in the regulation of ZmPHRs in association with certain hormones.

Expression patterns of ZmSPXs were analyzed under low- and normal-Pi treatments in different tissues of 178 maize inbred lines, including ear leaf, stem, anther, cornsilk, root, ear, and ear bract ( Figure 2 ). Notably, the ZmSPXs exhibited higher expression levels in anthers compared to other tissues but were significantly inhibited under low-Pi conditions. Meanwhile, ZmSPX1–5 were induced by low-Pi stress in roots ( Figure 2 ). Collectively, these results showed that the ZmSPX family members were involved in response to the low-Pi stress and play a specific function in different tissues of maize.

ZmSPX1, ZmSPX3, ZmSPX4, ZmSPX5, and ZmSPX6 were found to be localized in the nucleus and cytoplasm ( Supplementary Figure S3 ). ZmSPX2 was mainly localized in the nucleus ( Supplementary Figure S3 ). In a previous study (Xiao et al., 2021), subcellular localization studies of ZmSPXs in maize protoplasts revealed that ZmSPX1 (equivalent to ZmSPX5 in this study) and ZmSPX3 (equivalent to ZmSPX4 in this study) were localized in both the nucleus and cytoplasm. ZmSPX5 (equivalent to ZmSPX2 in this study) was exclusively localized in the nucleus. Moreover, ZmSPX4 (equivalent to ZmSPX6 in this study) and ZmSPX5 (equivalent to ZmSPX2 in this study) were detected in both the nucleus and cytoplasm, while ZmSPX6 (equivalent to ZmSPX1 in this study) was found in the chloroplast. The results showed partial consistency in the subcellular localization between tobacco tissue and maize protoplasts. However, some discrepancies were observed, presumably attributed to variations in cell types across different species.

In this study, a yeast two-hybrid assay was conducted to identify the interaction between ZmSPX and ZmPHR proteins. Initially, the self-activation of ZmSPX and ZmPHR genes was assessed. Our findings showed that ZmPHR1 and ZmPHR2 demonstrated normal growth on SD/−Ade/−His/−Trp-deficient medium, while the ZmSPXs transformed with yeast cells did not grow on the selective medium ( Figure 3A ), indicating that the ZmSPXs had no self-activation effect. Therefore, SPX protein was chosen as bait and PHR protein as prey to verify the interaction. Furthermore, protein interaction was examined in various combinations of ZmSPXs, ZmPHR1, and ZmPHR2. Results revealed that combinations such as ZmSPX1 and ZmPHR2, ZmSPX2 and ZmPHR1, ZmSPX3 and ZmPHR1, ZmSPX5 and ZmPHR1, and ZmSPX6 and ZmPHR1 all exhibited robust growth on SD/−Ade/−His/−Leu/−Trp/+X-α-gal medium, indicating their protein interactions ( Figure 3B ). It is suggested that ZmSPXs and ZmPHRs may modulate the maize response to low-Pi stress at the post-transcriptional level.

To investigate the function of ZmSPXs in response to low-Pi stress, overexpression lines of ZmSPXs were created in Arabidopsis. However, only the overexpression of ZmSPX1 served to enhance root sensitivity to Pi deficiency and high-Pi conditions in A. thaliana ( Supplementary Figures S4 , S5 ). In order to further explore the function and role of ZmSPX1, we generated knockout and overexpressing maize transgenic lines for this gene. Our experimental results demonstrated a significant reduction in the hundred-grain weight and grain weight per year in the overexpressing lines compared to the wild type, while the knockout lines exhibited a marked increase ( Figures 4A, B, D, E ). Moreover, analysis of the P concentration in the grains of the knockout and overexpressing lines revealed no significant difference compared to the wild type ( Figures 4C, F ). We also assessed the P concentration in the internode, sheath, and leaf of the transgenic plants in the field trials. The findings indicated a higher P concentration in the knockout lines in these tissues than in the wild-type lines, whereas the overexpressing lines showed lower P concentrations compared to the wild-type lines ( Figures 4G–I ). These results suggested that ZmSPX1 affected the P concentration in maize stems and leaves and exerted a certain impact on the yield of maize, but not on the grains.

To identify significant variations associated with phenotypic traits, we performed genome sequence amplification on 211 inbred lines and conducted multiple sequence alignments. We obtained a total of 1,390 bp of sequences, comprising 780 bp upstream of the initiation codon, 354 bp of the coding region, and 256 bp downstream of the initiation codon ( Supplementary Table S10 ). We identified a total of 41 variants, which included 34 SNPs and seven InDels. The average distances between SNPs and InDels were 40.89 bp and 198.57 bp, respectively. The sequence variation frequencies varied across regions, with the highest frequency observed in the upstream sequence of the initiation codon (0.038) and the lowest in the coding region (0.006) ( Supplementary Table S10 ). By utilizing a 200-bp sliding window with a 50-bp step size, we analyzed the nucleotide diversity (π × 1,000) of the ZmSPX1. We observed the overall nucleotide diversity to be 0.009, with the upstream sequence of the initiation codon exhibiting the highest diversity (0.013) among all regions and the coding region displaying the lowest diversity (0.002). Additionally, we investigated the selection pressure of ZmSPX1 using Tajima’s D, Fu and Li’s D*, and F* tests ( Supplementary Table S10 ). The results showed that the upstream sequences displayed positive values, indicating a mode of balanced selection in their sequence evolution, with both Tajima’s D and Fu and Li’s F* tests being significant ( Supplementary Table S10 ). Conversely, the coding region and downstream showed negative values, suggesting that these regions have experienced either negative selection or population expansion ( Supplementary Table S10 ).

Additionally, to further investigate the relationship between sequence variations of ZmSPX1 and the phenotype in the maize seedling stage, we employed an MLM for candidate gene association analysis. We identified, under normal-Pi treatment, 37 markers (30 SNPs and seven InDels) as significantly associated with 17 traits, explaining 2.6% to 8.6% of the phenotypic variation ( Supplementary Table S11 ). We found, under low-Pi treatment, a total of 10 markers to be significantly correlated with 13 traits, with R ² ranging from 2.8% to 5.4% ( Supplementary Table S12 ). Moreover, 37 markers were significantly correlated with 16 traits of low-Pi tolerance index, explaining 2.8% to 20.2% of the phenotypic variation ( Figures 5A, B ; Supplementary Table S13 ). As the low-Pi tolerance index integrated traits under low- and normal-Pi conditions, we utilized all sequence variant sites significantly associated with the low-Pi tolerance index for haplotype division. First, LD analysis revealed the presence of multiple LD blocks with high linkage relationships ( Figure 5C ). Then, we categorized all variant sites into five haplotypes (MAF > 0.05) and conducted comparisons among the different haplotypes ( Figure 5D ; Supplementary Table S14 ). The findings indicated that in the low-Pi tolerance index of root traits, the fresh weight of the crown root index of Hap5 was significantly higher than that of Hap1, Hap2, and Hap4, while the fresh weight of the seminal root index was higher than that of Hap3 and Hap1. Additionally, the root volume of Hap5 was significantly higher than that of Hap1. Furthermore, we observed that the dry weight of the whole plant index of Hap5 was significantly higher than that of other haplotypes. Based on these results, it can be inferred that Hap5 enhances biomass production by promoting root development.

Pi homeostasis is crucial for plant growth and development, with ZmSPX1 playing a pivotal role in Pi uptake by plants from both the source and reservoir. Consequently, we first analyzed the phenotypes in ZmSPX1 overexpression transgenic A. thaliana and WT under different Pi concentration treatments. Under both Pi deficiency and high-Pi conditions, we observed a more robust root system development in the WT plants, whereas the growth of the root system was inhibited in the overexpressing (OE) lines. These results indicated that overexpression of the ZmSPX1 gene in A. thaliana enhanced the sensitivity of roots to both Pi deficiency and high-Pi conditions. Therefore, we proceeded to construct transgenetic lines for ZmSPX1 in maize. The knockout ZmSPX1 in maize led to an increase in P content in the stems and leaves. Additionally, a significant rise in hundred-grain weight and grain yield per ear was also observed in the knockout maize lines. Conversely, overexpression of ZmSPX1 in maize demonstrated contrasting phenotypes. Studies in dicotyledonous model plants have demonstrated that SPXs, functioning as negative regulators, can impede the central regulator PHR1 from acting on downstream low-Pi response genes (Jia et al., 2023). Conversely, research reports in monocotyledonous model plants such as A. thaliana and soybean suggest that SPXs serve as positive regulators. Overexpression of SPX1 induced the expression of low-Pi response genes and thus increased plant P concentration (Duan et al., 2008; Yao et al., 2014a). These results indicated that the expression patterns and functions of the SPX1 gene in maize and A. thaliana are different, which may also explain why the P concentration in A. thaliana overexpressing lines of SPX was higher than that in wild-type plants under normal-Pi conditions.

Additionally, a previous study demonstrated that OsPHO1;2 and ZmPHO1;2 play a crucial role in Pi allocation during grain filling. The activity of ADP-glucose pyrophosphorylase (AGPase) was inhibited in the knockout mutants of OsPHO1;2 and ZmPHO1;2, resulting in a defect in grain filling and a reduction in hundred-grain weight (Ma et al., 2021). Regarding the plant Pi signaling pathway (Wang et al., 2021b), the E2 ubiquitin-binding enzyme PHO2 can degrade PHO1 through its N-terminal SPX domain. The mRNA level of PHO2 was regulated by the cleavage of miR399. PHR1 can bind to the promoter region of miRNA399 to modulate its expression. Therefore, the relationship between SPX proteins and PHO1 was established through this network. However, further research is needed to investigate whether the impact of ZmSPX1 on maize grain yield, as observed in this study, is related to ZmPHO1.

To further investigate the relationship between the ZmSPX1 gene and low-Pi stress response-related traits, we conducted an association analysis utilizing sequence variations of the gene in the population and integrated the corresponding phenotypes. The results revealed that variant sites in ZmSPX1 are significantly associated with numerous low-Pi-related traits. This finding complements the findings from the phenotypic identification experiments in Arabidopsis involving the overexpression of the ZmSPX1 gene, providing strong evidence for the pivotal role of the ZmSPX1 gene in the low-Pi stress response pathway in maize. Additionally, we identified beneficial haplotypes based on 37 markers significantly associated with low-Pi tolerance indexes. Among these, Hap5 manifested a notable low-Pi tolerance phenotype, presenting an opportunity for the development of low-Pi-tolerant maize materials based on this outcome, and the identification of novel markers for screening purposes.