Integrated transcriptomic and proteomic analyses provide insights into potential molecular mechanisms underlying globulin content variation in maize embryos

Abstract

Background

Globulin in maize kernels is a high-quality protein, which is mainly enriched in the embryo. However, academic research focusing on the molecular mechanism underlying globulin synthesis remains relatively limited to date. Therefore, unveiling the molecular mechanism of globulin synthesis in maize embryos is of great significance for the breeding of high-protein maize varieties. To gain better insights into this mechanism, we integrated transcriptomic and proteomic analysis to identify differentially expressed genes (DEGs) and differentially expressed proteins (DEPs) between two types of materials with distinct globulin contents.

Results

Two maize inbred lines were selected for this study, including B73 (low globulin content) and HP434 (high globulin content). Integrated transcriptomic and proteomic analyses revealed that DEGs and DEPs were predominantly enriched in the endoplasmic reticulum (ER) protein processing pathway. Significantly, most of these DEGs and DEPs belonged to the heat shock protein (HSP) family. This indicates that globulin synthesis is closely related to the genes of this family, especially the small molecule heat shock proteins.

Conclusions

The high expression of HSP family genes in HP434 might alleviate the increased ER processing pressure caused by the aggregation of misfolded proteins during the high-abundance synthesis of globulin, thereby contributing to the synthesis of globulin in the ER.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08940-w.

Introduction

Maize (Zea mays L.), stands as the crop with the highest global total output and serves as a food source for more than half of the global population. The maize kernel contains abundant nutrients, including carbohydrates, proteins, fats, and vitamins. Seed storage proteins (SSPs) are the main component of proteins in maize kernels. They mainly consist of four types: prolamin, globulin, glutelin and albumin [1]. Current research on maize SSPs synthesis primarily focuses on prolamin, which lack lysine—an essential amino acid vital for humans, livestock, and poultry. Lysine is crucial for animal growth and participates directly in protein synthesis in animals [2]. Enhancing the nutritional quality of maize kernels has always been an important domain in maize quality breeding, and high-quality protein maize rich in lysine is an ideal material for feed.

Globulin is predominantly stored in the maize embryos, constituting approximately 10–20% of the protein therein, and its content serves as a critical trait determining the lysine content of the kernel [3]. Previous studies have shown that high-lysine mutants in rice and maize exhibit increased globulin and Binding protein (BiP) content [4, 5]. Zheng’s research demonstrated that suppressing α-zein in the endosperm via αRNAi increases ZmGLB1 levels in the embryos and significantly enhances Protein storage vacuoles (PSV) size and quantity, confirming that globulins are the primary protein responding to the redistribution of proteins from the endosperm to the embryo [6]. Globulin, with its high lysine content, is classified as a high-quality protein and represents a promising target for breeding high-quality protein maize. However, current research on the molecular mechanisms governing globulin accumulation in maize embryos remains limited.

The biosynthesis of globulin in higher plants primarily occurs on polyribosomes attached to the rough endoplasmic reticulum. Initially, globulin precursors are synthesized on ribosomes, enter the ER lumen for processing, and are transported through the Golgi apparatus to PSVs, where they mature and accumulate [6–8]. Globulin precursors undergo at least three processing steps to form mature proteins, including the conversion of ZmGLB1’ to ZmGLB1. The ER is crucial for protein folding and post-translational modifications, and maintaining ER homeostasis is essential for normal cellular function. High-abundance protein synthesis can lead to unfolded or misfolded proteins, which must be promptly eliminated via ER-associated degradation (ERAD) to prevent adverse cellular impacts. Since many unfolded proteins originate from high-abundance storage proteins, ERAD is particularly important for their accumulation [9]. Molecular chaperones assist in accurate protein folding and prevent aggregation within ERAD, maintaining a healthy intracellular protein environment. Heat shock proteins (HSPs), a large family of molecular chaperones, include small HSP, HSP60, HSP 70, HSP90, and HSP100 [10, 11]. HSPs are vital for plant growth, development, and stress resistance (heat, drought, salt). Initially identified during heat shock, HSPs are also expressed in unstressed cells during embryogenesis and seed maturation [12].

The primary links in gene expression are transcription and protein synthesis. mRNA serves as the intermediate, while proteins execute gene functions. To understand the specific regulatory processes of gene expression, it is essential to monitor both mRNA and proteins simultaneously. By integrating proteomics and transcriptomics data and analyzing them at both transcriptional and protein levels, a more comprehensive understanding of molecular mechanisms in various biological processes can be achieved [13]. Integrated transcriptome and proteome analysis can provide deeper insights into the mechanisms underlying globulin content differences among maize varieties and elucidate the molecular mechanisms of globulin synthesis and metabolism. This is crucial for increasing globulin and lysine levels in maize and provides a theoretical foundation and genetic resources for breeding high-quality protein maize varieties.

Materials and methods

Plant materials

Based on our previous work, a panel of 348 maize inbred lines was evaluated for protein content across multiple environments. Among these, HP434 and B73 were identified as exhibiting significant differences in total protein content, which belong to the SS heterotic group and share a similar growth period [14]. These two inbred lines were kindly provided by the Maize Quality and Stress Resistance Breeding Group, Chinese Academy of Agricultural Sciences. Subsequent analysis further revealed a significant difference in globulin content between the two lines. Therefore, Maize plants were cultivated at the experimental base of the Institute of Crop Sciences, Chinese Academy of Agricultural Sciences, located in Sanya, Hainan Province (18.27° N, 108.57° E), in November 2023. The planting density was maintained at 60,000 plants per hectare. Artificial self-pollination was carefully conducted by hand, and conventional field management was implemented during the whole growth period.

Morphological character assays

The embryos of maize HP434 and B73 at 15, 20, 25, 30, and 35 Days after pollination(DAP) were collected separately. The collected embryos were cryopreserved in liquid nitrogen and stored at -80 °C. Once all the embryos were collected, they were photographed uniformly. Fifty embryos of mature maize kernels were obtained, and the volume of 20 embryos was estimated with water at a precision of 0.1mL. Each time, 20 embryos were randomly measured, and the measurement was repeated three times before being converted to the volume of a single embryo. The method for measuring the weight of 100 embryos followed Li’s protocol. Embryos were dried at 40℃ for 48 h to remove moisture. The weight of 20 randomly selected embryos was measured using an analytical balance (precision: 0.001 g), and this process was repeated three times. The average weight of these 20 embryos was then multiplied by 5 to obtain the weight of 100 embryos [15].

Extraction of globulin from maize kernels and protein quantitative analysis

To compare globulin content, globulin was extracted from the embryos, and protein concentration was measured using the Bicinchoninic Acid (BCA) method [16]. Protein separation and analysis were conducted via SDS-PAGE [17]. Embryos of HP434 and B73 maize at 25 DAP and at maturity were dried at 40℃ for 48 h to remove moisture. After drying, 0.1 g of ground embryos was mixed with 600 µL of albumin extraction buffer, shaken for 2 h, and centrifuged at 12,000 rpm for 10 min. The supernatant, collected in a 1.5 mL EP tube, served as the albumin extract. This process was repeated twice to ensure complete depletion of albumin. Next, 300 µL of globulin extraction buffer was added to the remaining residue, shaken for 2 h, and centrifuged at 12,000 rpm for 10 min. The resulting supernatant, collected in a 1.5 mL EP tube, served as the globulin extract. A 20 µL aliquot of the globulin sample was mixed with an appropriate amount of 5X loading buffer, boiled at 95℃ for 5 min, and separated on a 10% SDS-PAGE gel (120 V, 85 min). Protein bands were visualized using Coomassie Blue R-250 staining. BCA protein concentration was determined using the PC0020-BCA kit from Beijing Solarbio Science & Technology Co., Ltd.

Transcriptome sequencing and analysis

Total RNA was extracted from maize embryos using TRIzol^® Reagent (Magen) according to the manufacturer’s instructions. RNA quality was assessed using a Nanodrop ND-2000 system (Thermo Scientific) based on the A260/A280 ratio, and RNA integrity (RIN) was determined using an Agilent Bioanalyzer 4150 (Agilent Technologies). Three biological replicates were performed for each sample.

Only qualified samples were used for library construction. Paired-end libraries were prepared using the ABclonal mRNA-seq Lib Prep Kit (ABclonal) following the manufacturer’s protocol. mRNA was purified with oligo(dT)-coated magnetic beads and fragmented in the First Strand Synthesis Reaction Buffer. Double-stranded cDNA was synthesized, followed by end repair, 3’ adenylation, and adapter ligation using exonuclease and polymerase. PCR amplification was performed, and the resulting products were purified to obtain the sequencing library. Library quality was evaluated using the Agilent Bioanalyzer 4150. Sequencing was conducted on the MGISEQ-T7 platform.

The raw data from the BGI platform was filtered to remove adapter sequences, low-quality reads, and reads with an N ratio exceeding 5%, resulting in clean reads. HISAT2(http://daehwankimlab.github.io/hisat2/) was used to align the clean reads to the Zea mays NCBI B73 Ref Gen v4 (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_000005005.2/),generating mapped reads. FeatureCounts (http://subread.sourceforge.net/) was then employed to count the reads aligned to each gene, and FPKM values were calculated based on gene length [18].DESeq2 ( http://bioconductor.org/packages/release/bioc/html/DESeq2.html) was used for differential expression analysis between groups, with |log₂FC| >1 and Padj < 0.05 as the criteria for DEGs. The clusterProfiler R package was used for Gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis, with P < 0.05 indicating significant enrichment.

Quantitative real-time PCR (qRT-PCR) analysis

Total RNA was extracted from the samples using the FastPure Universal Plant Total RNA Isolation Kit (Nanjing Novozymes Biotech Co., Ltd.) and reverse-transcribed into cDNA using the TIANGEN cDNA First-Strand Synthesis Kit. Specific primers for the selected DEGs were designed using the NCBI Primer-BLAST tool. The relevant primers used for qRT-PCR were listed in Table S10. The qRT-PCR reactions were prepared using the SuperReal PreMix Plus (SYBR Green) FP205 kit from TIANGEN and analyzed on the BIO-IQ5 real-time PCR instrument. The internal reference gene was Ubi [19], and three biological replicates and three technical replicates were designed. Relative gene expression levels were calculated using the 2^−∆∆Ct method [20].

Proteome sequencing and analysis

The samples for proteome analysis were also corn embryos collected at the 25 DAP, consistent with the stage used for RNA-seq and three biological replicates were performed for each sample. The samples were ground in liquid nitrogen-soaked pestle and mortar, then precipitated with TCA/acetone buffer. SDT buffer (4% SDS, 100 mM Tris-HCl, pH 7.6) was added, followed by sonication and boiling for 15 min. After centrifugation at 14,000 g for 40 min, protein quantification was done using the BCA method. Each sample was mixed with 3.75 µg of 5X loading buffer, heated to boiling for 5 min, and resolved by SDS-PAGE electrophoresis (4%-20% precast gradient gel, 180 V, 45 min), then stained with Coomassie Brilliant Blue R-250. The QC sample was a pooled mixture of all individual samples. All samples were enzymatically digested using the FASP method [21]. The resulting peptides were desalted with a C18 Cartridge, lyophilized, and reconstituted in 40µL of 0.1% formic acid. Peptide concentration was measured at OD280. iRT standard peptides were added to the digested peptides before DIA (Data-Independent Acquisition) mass spectrometry analysis on an Astral high-resolution mass spectrometer.

DIA analysis was conducted using the Vanquish Neo system (Thermo Fisher) with nanolitre flow rates for chromatographic separation. The separated samples were then analyzed by DIA mass spectrometry using the Astral high-resolution mass spectrometer (Thermo Fisher Scientific). Precursor Ions were scanned at a mass range of 380–980 m/z, MS1 resolution was 240,000 at 200 m/z, Normalized AGC Target: 500%, Maximum IT: 5ms. 299 windows were set for DIA mode in MS2 scanning, Isolation Window: 2 m/z, HCD Collision Energy: 25ev, Normalized AGC Target: 500%, Maximum IT: 3ms.

DIA data were processed using DIA-NN software with the following parameters: trypsin as the enzyme, one missed cleavage site allowed, carbamidomethylation (C) as the fixed modification, and oxidation (M) and N-terminal acetylation as variable modifications. Proteins identified were filtered to maintain an FDR < 1%. To analyze the differences in DEPs between the groups, further screening was conducted using the following criteria: Fold Change (FC) > 1.5, P value < 0.05 (t-test). GO terms were assigned using Blast2GO (BLASTP 2.8.0+), and KEGG pathways were annotated using KOBAS (3.0). To assess enrichment, Fisher’s Exact Test was applied to compare the distribution of GO categories and KEGG pathways between the target protein set and the overall proteome. Protein domain analysis was conducted using the Pfam database, a comprehensive collection of protein families represented by multiple sequence alignments and hidden Markov models. We used InterProScan (version 5.25-64.0) to perform integrated scanning from the InterPro database, obtaining domain annotations for the target protein sequences.

Integrative analysis of transcriptome and proteome

A correlation analysis was conducted on the DEGs and DEPs data from both the transcriptome and proteome. The Spearman correlation coefficient was calculated to assess the expression levels of correlated genes [22]. Subsequently, GO and KEGG functional annotations were performed on the correlation results to further elucidate the functional information of these genes.

Data analysis

Data were presented as mean ± SD and analyzed with GraphPad Prism10.0. The numbers of biologically independent samples (n) for each experiment are noted in the figure legends. Independent samples t-test conducted in GraphPad Prism10.0 were used for statistical analyses.

Results

Phenotype and globulin comparisons of maize embryos

The morphology of maize HP434 and B73 embryos was compared at different days after pollination (15, 20, 25, 30, and 35 DAP). Results showed that HP434 embryos were consistently larger than B73 embryos at each stage (Fig. 1A). Further determination showed that embryo volume and 100-embryo weight of HP434 were increased by 53.85% and 32.80% respectively compared with those of B73 (P < 0.01) (Fig. 1B, C). SDS-PAGE analysis revealed significantly higher levels of the 63-kDa polypeptide encoded by ZmGlb1 and the intermediate ZmGLB1’ protein in HP434 embryos, especially at 25 DAP (Fig. 1D). BCA results indicated that the globulin concentration in HP434 embryos was higher at both 25 DAP and the mature stage (Fig. 1E). These findings suggest that HP434 embryos contain more globulin than B73. To compare the genetic structures of ZmGlb1 between the two materials, and found that ZmGlb1 contains 5 exons and 4 introns in both the materials (Fig. 1F). More specifically, it was completely identical the 2000 bp promoter region, 2138 bp full-length genomic sequence (Fig.S1), and 1719 bp CDS sequence (Fig.S2) of this gene. In contrast to the higher concentration of ZmGLB1 protein in HP434 embryos at 25 days after pollination compared to B73, qRT-PCR showed that the transcriptional level of this gene in HP434 embryos was significantly lower than that in B73 at the same stage (Fig. 1G). Therefore, it is speculated that the efficient accumulation of globulin in HP434 may be due to the influence of post-translational modifications, as well as protein processing and transport processes, rather than gene sequence variations. To elucidate the molecular basis of globulin difference, transcriptome and proteome sequencing were used at 25 DAP embryos.

Fig. 1.

Phenotype and globulin comparison the embryos of B73 and HP434. (A) The Phenotype of maize B73 and HP434 embryos at different stages. (B-C) The embryo volume (B) and the weight of 100 embryos (C) of B73 and HP434. (D-E) SDS-PAGE analysis (D) and BCA results (E) of B73 and HP434. M: Marker. Note: The markers are 180, 130, 100, 70, 55, 40, 35, 25, 15 k DA in sequence from top to bottom. (F) Analysis of the genetic architectures of ZmGlb1. (G) The transcriptional level of ZmGlb1 in B73 and HP434. Data are presented as means ± SDs. (t-test, ** P <0.01)

Transcriptome differential analysis

The heat map indicated a Pearson correlation coefficient R² ≥ 0.86 between samples, demonstrating strong sample correlation and effective selection (Fig. 2A). Comparative transcriptomics between HP434 and B73 identified 7481 DEGs, with 3235 upregulated and 4246 downregulated (Fig. 2B, Table S1). GO annotations covered 5222 terms: 3205 biological processes (BP), 568 cellular components (CC), and 1449 molecular functions (MF) (Table S2). GO enrichment analysis of all differentially expressed genes revealed that biological process terms were mainly enriched in response to heat and response to temperature stimulus. Molecular function terms were primarily enriched in heme binding and tetrapyrrole binding. Cellular component terms were mostly enriched in extracellular region and cell periphery (Fig.S3). To investigate the differences in functional characteristics and enrichment patterns between upregulated and downregulated differentially expressed genes, GO annotation was performed separately for the two gene sets.The results showed that upregulated DEGs were mainly enriched in terms including stress response, protein folding, and cytosolic ribosomes (Fig. 2C), whereas downregulated DEGs were significantly enriched in extracellular regions, oxidoreductase activity, and transcription regulator activity (Fig. 2D). KEGG pathway analysis identified 122 signaling pathways (Table S3). KEGG enrichment analysis of all differentially expressed genes showed that the most significantly enriched pathways were metabolic pathways, biosynthesis of secondary metabolites, plant hormone signal transduction, photosynthesis, and phenylpropanoid biosynthesis (Fig.S4). To investigate the differences in metabolic pathway enrichment between up-regulated and downregulated differentially expressed genes, KEGG pathway enrichment analysis was performed separately on upregulated and downregulated DEGs. The results demonstrated that upregulated DEGs were mainly enriched in pathways such as protein processing in endoplasmic reticulum, starch and sucrose metabolism, and biotin metabolism (Fig. 2E). In comparison, downregulated DEGs were more highly enriched in plant hormone signal transduction, glutathione metabolism, phenylpropanoid biosynthesis, and photosynthesis (Fig. 2F).

Fig. 2.

Transcriptomic analysis of embryos of HP434 and B73. (A) Heat map of the correlation coefficients between samples. (B) Volcano plot of DEGs. (C-D) GO enrichment bar plots for all upregulated and downregulated DEGs. BP: Biological Process; CC: Cellular Component; MF: Molecular Function. (E-F) KEGG enrichment bubble plots for upregulated and downregulated DEGs.

Proteomics characterization by DIA

Based on DIA proteomics technology, B73 and HP434 yielded a total of 114,000 peptides (Table S4) and 11,448 proteins (Table S5). A total of 1,287 DEPs were identified, including 614 upregulated and 673 downregulated proteins (Fig. 3A, Table S6). Domain prediction of DEPs was performed using the domain prediction software InterProScan. Through KEGG enrichment analysis, it was found that the domains of DEPs were significantly enriched in HSP20/alpha crystallin family, Cytochrome P450, and alpha/beta hydrolase fold, among others (Fig. 3B). GO enrichment analysis of all differentially expressed proteins showed that they were significantly enriched in molecular function terms including catalytic activity and transporter activity, cellular component terms including cellular anatomical entity and protein-containing complex, as well as biological process terms including response to stimulus (Fig.S5). In addition, upregulated DEPs were enriched mainly in protein folding chaperone and nutrient reservoir activity (Fig. 3C, Table S7), while downregulated DEPs were more enriched in metabolic processes (Fig. 3D, Table S7). KEGG pathway enrichment analysis of all differentially expressed proteins revealed that these proteins were mainly enriched in thiamine metabolism, phenylpropanoid biosynthesis, and diterpenoid biosynthesis (Fig.S6). Further enrichment analysis performed separately on up-regulated and down-regulated differentially expressed proteins showed that up-regulated proteins were primarily enriched in pathways including protein processing in endoplasmic reticulum and inositol phosphate metabolism(Fig. 3E, Table S8). In contrast, down-regulated proteins were more highly enriched in linoleic acid metabolism, photosynthesis, starch and sucrose metabolism, and ribosome pathways (Fig. 3F, Table S8).

Fig. 3.

Proteomic analysis of embryos of HP434 and B73. (A) Volcano plot of DEPs. (B) Domain pathways enrichment. (C-D) GO enrichment bar plots for all upregulated and downregulated DEPs. BP: Biological Process; CC: Cellular Component; MF: Molecular Function. (E-F) KEGG enrichment bubble plots for upregulated and downregulated DEPs.

Protein processing pathway is involved in globulin accumulation in maize embryos

To better understand the biological functions of coexpressed differentially expressed genes and differentially expressed proteins, GO functional and metabolic pathway annotations were performed for the proteome and transcriptome, followed by integrated correlation analysis of the annotated genes and proteins. Integrated GO enrichment analysis showed that differentially expressed genes and proteins were mainly enriched in protein folding, oxidoreductase activity, and extracellular region (Fig.S7). Integrated KEGG enrichment analysis revealed that differentially expressed genes and proteins were enriched in starch and sucrose metabolism, phenylpropanoid biosynthesis, and glutathione metabolism (Fig.S8). GO and KEGG enrichment integration analyses were conducted for upregulated and downregulated DEPs and DEGs. GO enrichment analysis showed that both were significantly upregulated in biological processes related to catalytic activity and protein folding (Fig. 4A), while functional modules associated with transport activity and extracellular region were significantly downregulated (Fig. 4C). KEGG enrichment analysis demonstrated that upregulated DEPs and DEGs were co-enriched in metabolic regulatory pathways such as protein processing in endoplasmic reticulum and glycolysis (Fig. 4B), whereas downregulated DEPs and DEGs were mainly concentrated in specific pathways including glutathione metabolism and phenylpropanoid biosynthesis (Fig. 4D).

Fig. 4.

Integrative multi-omics analysis of embryos of HP434 and B73. (A) GO enrichment bar plots for all consistently upregulated DEGs and DEPs. BP: Biological Process; MF: Molecular Function; (B) KEGG pathway annotation of consistently up-regulated DEGs and DEPs. (C) GO enrichment bar plots for all consistently downregulated DEGs and DEPs. BP: Biological Process; CC: Cellular Component; MF: Molecular Function. (D) KEGG pathway annotation of consistently downregulated DEPs and DEPs.

Based on the results of integrated transcriptomic and proteomic analyses, it was found that protein folding and protein processing in the endoplasmic reticulum were significantly upregulated in both omics datasets. Therefore, we propose that protein processing in the endoplasmic reticulum may play a crucial role in globulin accumulation in maize embryos. Specifically, key protein processing pathways in the endoplasmic reticulum, including HSP70s, HSP40s, sHSPs and the ubiquitin ligase complex UBE4 of heat shock proteins, exhibited significant differences in both omics. Additionally, genes such as HSP90s, GRP94, Derlin, and RMA1 showed significant differences only at the transcriptomic level. A total of 55 DEPs and 35 DEGs related to the endoplasmic reticulum protein folding pathway were identified, among which 22 DEPs and DEGs could be correlated. Analysis of these 22 differential genes and proteins revealed that 19 differential proteins belonged to the HSP family, including 14 sHSPs, 4 HSP70s, and 1 HSP40 (Table S9). We emphasize that the elevated HSP70 expression in HP434, together with the enriched endoplasmic reticulum protein processing pathway, may contribute to the alleviation of endoplasmic reticulum stress and thereby provide a favorable condition for stable globulin accumulation during seed development.

Analysis of cis-acting elements in sHSP promoters

To investigate the mechanism underlying the differential expression of sHSP genes between HP434 and B73, 2 kb promoter sequences of 14 differentially expressed sHSP genes were extracted from both the B73 reference genome database and the self-assembled HP434 database developed by our research group. Sequence alignment revealed that except for the promoters of Zm00001d015777 and Zm00001d052194, which showed no differences between the two materials, the promoter sequences of other genes exhibited significant variations. Subsequent cis-acting element analysis using TBtools [23] demonstrated that, compared with B73, most sHSP genes in HP434 contained an increased number of cis-acting elements capable of binding bZIP transcription factors (including O2-site, GCN4_motif, and G-box) as well as endoplasmic reticulum stress-related elements (Box 4 and ARE) (Fig. 5). These findings suggest that the increased abundance of these cis-acting elements in sHSP genes of HP434 may facilitate the binding of corresponding transcription factors, thereby promoting the upregulation of their expression levels.

Fig. 5.

Analysis of cis-acting elements in sHSP promoters of B73 and HP434.

Validation of RNA-seq data

To verify the reliability of RNA-seq data, qRT-PCR analysis was performed on 10 DEGs selected from the “Protein processing in endoplasmic reticulum” pathway, including 6 upregulated genes (Zm00001d042922, Zm00001d024903, Zm00001d044874, Zm00001d039566, Zm00001d033058, Zm00001d019481) (Fig. 6A-F) and 4 downregulated genes (Zm00001d030725, Zm00001d007301, Zm00001d029561, Zm00001d004139) (Fig. 6G-J). The expression changes detected by qRT-PCR were consistent with those from RNA-seq, confirming the accuracy of the RNA-seq data.

Fig. 6.

Validation of quantitative real time PCR. (A-F) Verification of the transcriptional level of upregulated genes (Zm00001d042922,Zm00001d024903,Zm00001d044874,Zm00001d039566,Zm00001d033058 and Zm00001d019481). (G-J) Verification of the transcriptional level of downregulated genes (Zm00001d030725, Zm00001d007301, Zm00001d029561 and Zm00001d004139). Data are presented as means ± SDs. (t-test, ** P < 0.01; * P < 0.05)

Discussion

Current research on improving maize grain protein quality primarily focuses on reducing the content of alcohol-soluble proteins in the endosperm, thereby indirectly increasing the levels of non-alcohol-soluble proteins such as globulins through proteome rebalancing [24, 25]. In our laboratory, Near-Infrared Spectroscopy (NIR) measurements revealed that the crude protein content of HP434 was higher than that of B73. SDS-PAGE analysis further showed that both alcohol-soluble and non-alcohol-soluble protein contents were elevated in HP434. This suggests that the high globulin content in HP434 is not due to proteome rebalancing caused by reduced alcohol-soluble protein content but rather results from other molecular mechanisms. However, the molecular mechanisms underlying globulin accumulation have been rarely studied. To investigate the molecular basis of high globulin content in maize embryos, we examined gene expression at both transcriptional and protein levels. Zheng’s study demonstrated that Glb1 gene expression initiates at 20 DAP, peaks at 28 DAP, and subsequently declines [6]. Therefore, we selected 25 DAP maize embryos for our experiments. Transcriptome and proteome data identified Glb1, the most important globulin-coding gene in maize embryos. At the transcriptional level, Glb1 expression in HP434 was lower than in B73, but its protein expression level was significantly higher in HP434 compared to B73. These findings were further validated by qRT-PCR and SDS-PAGE analyses. Based on these observations, we speculate that the elevated globulin content in HP434 is primarily attributed to changes in post-transcriptional regulation, translation, post-translational modifications, processing, and transport mechanisms.

Gene expression is a complex and intricate process from RNA to protein to phenotype, involving regulatory activities such as post-transcriptional and post-translational modifications that occur after mRNA formation [26]. Therefore, examining only the transcription level cannot fully elucidate the gene expression and regulation mechanisms. Instead, integrated analysis of transcriptome and proteome data can provide a comprehensive understanding of gene expression levels and thoroughly explore the molecular mechanisms underlying protein accumulation. We compared the differential expression of genes at both mRNA and protein levels using transcriptomic and proteomic data. The results showed that the number of DEGs was higher than that of DEPs. Correlation analysis revealed a weak correlation between the proteome and transcriptome, indicating low concordance between transcription levels and protein abundance. These findings are consistent with previous studies on the integration of transcriptome and proteome data, which have shown that changes in most mRNAs are often opposite to changes in protein levels, resulting in low correlation [27–29]. These observations suggest that post-transcriptional and post-translational regulation, reversible phosphorylation, cellular splicing events, and translation efficiency play crucial roles in regulating protein accumulation [30]. Therefore, gene translation and post-translational modifications may significantly influence protein accumulation.

Numerous studies have demonstrated that the protein folding and secretion pathways in the endoplasmic reticulum represent the major rate-limiting steps in protein secretion [31–33]. In this study, 55 DEPs and 35 DEGs involved in the endoplasmic reticulum protein folding pathway were identified, among which 22 DEPs and DEGs were correlated. Analysis of these 22 differential genes and proteins revealed that 19 differential proteins belong to the HSP family, including 14 sHSPs, 4 HSP70s, and 1 HSP40. In the present study, sHSPs were the most prominent differential proteins, with all differential sHSPs showing upregulated expression at the protein level. sHSPs inhibit abnormal protein aggregation, protect normal proteins, maintain cellular integrity, and enhance resistance under various adverse conditions [34]. Furthermore, current research indicates that sHSPs are not only present under heat stress but also in mature seeds of many higher plants grown under normal temperature conditions [35–37]. In tobacco and Arabidopsis, it has been found that the expression of sHSPs and globulins starts simultaneously and disappears concurrently during seed germination. Transgenic tobacco plants lacking any protein bodies or storage proteins showed no sHSP accumulation in their embryos, suggesting that sHSP accumulation in plants is associated with globulin accumulation [35–38]. Additional studies have shown that sHSP expression increases significantly during embryo formation and maturation in maize seeds [39]. Moreover, during maize kernel filling, the content of storage proteins increases alongside the accumulation of heat shock proteins, indicating their important roles in responding to heat stress and protein folding. To investigate the causes underlying the differential expression of sHSP genes between HP434 and B73, we compared the coding sequences (CDS) and promoter sequences of 14 differentially expressed sHSP genes in the two materials. It was found that their CDS sequences either showed no differences or only minor variations, whereas the promoter sequences exhibited substantial differences. Therefore, we propose that the differences in promoter sequences are likely responsible for the different expression of these genes. Based on the results of cis-acting element analysis, we observed an increased number of bZIP-binding cis-acting elements (O2-site, GCN4_motif, and G-box) in HP434. bZIP transcription factors play crucial roles in plant growth and development, including seed maturation, germination, and floral development. Opaque2 (O2), an endosperm-specific bZIP transcription factor, can activate zein genes by specifically binding to the O2-box [40]. bZIP transcription factors are recruited to drive the expression of storage protein genes; for instance, in rice, there may exist bZIP proteins functionally similar to maize O2, which participate in regulating the endosperm-specific expression of GluB-1 through the GCN4 motif [41]. A similar mechanism has been identified in Arabidopsis: AtbZIP10 and AtbZIP25, by binding to G-boxes containing the ACGT core sequence, cooperatively regulate the expression of 2 S albumin and 12 S globulin genes with the B3 family transcription factor ABI3 [42]. In this study, several genes with consistent changes at both transcriptional and protein levels were selected for validation, taking Zm00001d044874 and Zm00001d039566 as examples. These two genes showed significant differential expression in the transcriptomic analysis, and the changes in protein abundance detected by proteomic profiling were highly consistent with their transcript levels. Quantitative real‑time PCR (qRT‑PCR) further verified the expression trends observed in the omics analyses. In addition, cis‑element analysis also revealed variations in the promoters of these two genes, which may be an important reason for the differences in their transcriptional and translational levels.

Additionally, genomic variations between the two materials (HP434 and B73) are specifically reflected in significant variations in the promoter sequences of differentially expressed genes, which in turn lead to differences in the transcriptional levels of these genes. Variations in the number of cis-acting elements (such as endoplasmic reticulum stress-related elements) in the promoter region can regulate the transcriptional activity of corresponding genes and promote their up-regulation to cope with changes in the intracellular environment—such as endoplasmic reticulum stress caused by the massive synthesis of globulins in maize embryos at 25 days after pollination. Represented by HSP family genes, the up-regulation of their transcriptional levels is essentially driven by genomic variations (variations in cis-acting elements of promoters). The enhanced expression of such genes can assist in the correct folding and accumulation of proteins by exerting molecular chaperone functions and participating in the maintenance of endoplasmic reticulum homeostasis, ultimately supporting the formation of high-abundance globulins in HP434. This process clearly indicates that promoter sequence variations at the genomic level are an important molecular basis for regulating transcriptional differences of genes and further affecting related physiological processes and trait formation.

Based on the above research findings and discussions, we found that the association between sHSPs and the synthesis of high-abundance globulins cannot be ignored, which will be explored through further studies. Additionally, we hypothesize that during the seed maturation of maize HP434, the massive synthesis of globulins in the embryo exacerbates the processing pressure on the endoplasmic reticulum, leading to an increase in unfolded and misfolded proteins. In response, the endoplasmic reticulum synthesizes a large number of HSPs to inhibit the abnormal aggregation of misfolded proteins, promote the degradation of abnormal proteins, and alleviate endoplasmic reticulum stress. Furthermore, HSPs facilitate the correct folding of proteins and accelerate the folding rate, thereby increasing the globulin content.

Conclusions

Through integrated transcriptomic and proteomic data analysis, it was found that genes from the HSP family, particularly sHSPs, in the protein processing pathway of the endoplasmic reticulum in maize HP434 were significantly and highly enriched at both the transcriptional and protein levels, functioning as molecular chaperones for protein folding. We propose that the high expression of HSP family genes in HP434 might alleviate the increased endoplasmic reticulum processing pressure caused by the aggregation of misfolded proteins during the high-abundance synthesis of globulins, thereby contributing to the synthesis of globulins in the endoplasmic reticulum. Furthermore, this study reveals the potential molecular mechanisms regulating globulin content variation in maize embryos, and provides genetic resources and a theoretical basis for breeding high-quality protein maize varieties.

Abbreviations

DAP

Days after pollination

DEGs

Differentially expressed genes

DEPs

Differentially expressed proteins

ER

Endoplasmic Reticulum

HSPs

Heat shock proteins

sHSP

Small Heat shock protein

GO

Gene ontology

KEGG

Kyoto encyclopedia of genes and genomes

SSPs

Seed storage proteins

BiP

Binding protein

PSVs

Protein storage vacuoles

ERAD

ER-associated degradation

DIA

Data-Independent Acquisition

Authors’ contributions

The experimental design is conceived by Changhong Guo and Xinhai Li. Yan Cui and MeiYu E performed the experiments and wrote the paper. Meijie Cao, Zhiqiang Zhou, Xin Lu, Yutong Zhang and Ruiqi Wang helped correct the manuscript. All the authors have read and approved the final manuscript.

Funding

This work was supported by grants from the General Program of the National Natural Science Foundation of China (32572311) and Modern Agro-Industry Technology Research System of Maize (CARS-02-02).

Data availability

Raw data that support the findings of this study have been deposited in the SRA database with the accession number PRJNA1391915 and ProteomeXchange Consortium via the iProX partner repository with the dataset identifier PXD072287.

Declarations

Ethics approval and consent to participate

The plant materials used in this study are in full compliance with relevant institutional, national and international guidelines and legislation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.