Genetic Perturbation of the Starch Biosynthesis in Maize Endosperm Reveals Sugar-Responsive Gene Networks

In maize, starch mutants have facilitated characterization of key genes involved in endosperm starch biosynthesis such as large subunit of AGPase Shrunken2 (Sh2) and isoamylase type DBE Sugary1 (Su1). While many starch biosynthesis enzymes have been characterized, the mechanisms of certain genes (including Sugary enhancer1) are yet undefined, and very little is understood about the regulation of starch biosynthesis. As a model, we utilize commercially important sweet corn mutations, sh2 and su1, to genetically perturb starch production in the endosperm. To characterize the transcriptomic response to starch mutations and identify potential regulators of this pathway, differential expression and coexpression network analysis was performed on near-isogenic lines (NILs) (wildtype, sh2, and su1) in six genetic backgrounds. Lines were grown in field conditions and kernels were sampled in consecutive developmental stages (blister stage at 14 days after pollination (DAP), milk stage at 21 DAP, and dent stage at 28 DAP). Kernels were dissected to separate embryo and pericarp from the endosperm tissue and 3′ RNA-seq libraries were prepared. Mutation of the Su1 gene led to minimal changes in the endosperm transcriptome. Responses to loss of sh2 function include increased expression of sugar (SWEET) transporters and of genes for ABA signaling. Key regulators of starch biosynthesis and grain filling were identified. Notably, this includes Class II trehalose 6-phosphate synthases, Hexokinase1, and Apetala2 transcription factor-like (AP2/ERF) transcription factors. Additionally, our results provide insight into the mechanism of Sugary enhancer1, suggesting a potential role in regulating GA signaling via GRAS transcription factor Scarecrow-like1.

Introduction

Maize (Zea mays) is one of the most important cereal crops in terms of economic and nutritional value and, together with rice and wheat, provides over 60% of calories consumed by people throughout the world (Fao.org). The major component of cereal seeds is the endosperm, which comprises approximately 85% of grain volume at maturity (Fao.org, 1998). The endosperm is the main site of starch synthesis and storage in the seed and serves as the source of nutrients to a germinating embryo. Starch also serves as the major calorie source for humans and livestock and is utilized in diverse industrial processes.

Starch is a storage carbohydrate primarily consisting of amylose, a linear polymer of α-(1→4)-linked glucose units, and amylopectin, a highly branched glucan polymer of α-(1→6)-linked glucose units connecting linear chains. Starch begins accumulating in the endosperm of maize kernels approximately 10 days after pollination (DAP) and continues through approximately 30 DAP, when starch production declines and kernels begin to enter the dehydration phase (reviewed in Hannah and Boehlein, 2017). Starch biosynthesis requires the coordination of multiple classes of enzymes, including adenosine 5′ diphosphate-glucose (ADP-Glc) pyrophosphorylase (AGPase), starch synthases (SSs), starch branching enzymes (SBEs), and starch debranching enzymes (DBEs) (Smith et al., 1997; Hannah and Boehlein, 2017). In maize, starch mutants have facilitated characterization of key genes involved in endosperm starch biosynthesis. Examples include large subunit of AGPase Shrunken2 (Sh2) and isoamylase type DBE Sugary1 (Su1). The AGPases catalyze the first committed step of starch biosynthesis and require both large and small subunits for enzymatic function. The sh2 mutation was caused by a complex structural variation (Hu Y. et al., 2021) and produces a non-functioning endosperm AGPase. The mutation results in kernels with a severe reduction in starch content to approximately 20% of wildtype levels. Starch biosynthesis is also impacted by mutations downstream in the pathway. The su1 allele is a mutation of Isoamylase I, which reduces starch content to approximately 50% of wildtype and shifts the ratio of amylose and amylopectin (Creech, 1968).

While many of the key enzymes involved in maize starch biosynthesis have been characterized, very little is understood about the regulation of starch biosynthesis. Some genes that affect starch biosynthesis have still not been functionally characterized. For example, the Sugary enhancer1 (se1) gene is commercially relevant in sweet corn production, but its role in starch and carbohydrate metabolism is still undefined. Although its exact mechanism is unclear, the Sugary enhancer1 mutant allele is caused by deletion of Zm00001eb115450 (Zhang X. et al., 2019). In conjunction with su1, the se1 mutation further reduces the starch content and increases sugar levels, including maltose, in the kernel (Ferguson et al., 1979).

Mutations throughout the starch biosynthetic pathway provide genetic perturbations that allow us to study the transcriptional response to the changing endosperm sugar and starch levels to identify coexpression networks controlling these responses. Sugars can act as signaling molecules as well as carbohydrate sources, with regulatory control over various processes including establishment of the basal endosperm transfer layer (BETL), inflorescence branching, and activation of specific transcription factors (reviewed in Doll et al., 2017). To characterize the impact of starch mutations on the endosperm transcriptome we performed comparative RNAseq analysis on near-isogenic lines (NILs) (Wt, sh2, and su1) in six genetic backgrounds. Kernels were harvested in three consecutive developmental stages to study the responses of gene networks over time. Transcriptomes were compared to identify both conserved and background-specific differentially expressed genes (DEGs) in response to disruptions in the starch biosynthesis pathway. We show here that responses to the sh2 mutation include increased expression of sugar (SWEET) transporters and an increase in genes involved in ABA signaling. We identify Class II trehalose 6-phosphate synthases (TPS synthetases), Hexokinase 1, and Apetala2 transcription factor-like (AP2/ERF) transcription factors as potential regulators impacting the transcriptional response to sh2. Additionally, we propose a potential role of ZmSE1 in regulating the hormone balance of GA and ABA to ultimately regulate starch breakdown and sugar metabolism.

Materials and Methods

Plant Material and Experimental Design

Eighteen sweet corn lines that included three near-isogenic lines (NILs) in six genetic backgrounds were obtained from Dr. William Tracy at the University of Wisconsin-Madison. Each inbred line, which were all originally su1, were backcrossed into wild type and sh2 lines. The inbreds were selected for the corresponding genotype (wild type or sh2) to create near isogenic lines. Plants were grown in 2018 in Belle Glade, Florida at the University of Florida Everglades Research and Education Center in Belle Glade, FL, United States. Soil type at the research center is Histosol. Plots of 12 plants per NIL were grown in a completely randomized design with five replicates. Plants were spaced at 18 cm; rows were 76 cm apart with 0.91 m alleys between ranges. All plants were self-pollinated, and two ears were harvested from each plot at blister stage, milk stage, and dent stage (14, 21, and 28 DAP, respectively) and immediately frozen. The kernels were removed from both frozen ears, bulked per plot, and stored at −80°C for RNA extraction and quantification of starch and sugars.

Sugar and Starch Quantification

Sugars were determined using the sucrose, D-fructose and D-glucose assay procedure from Megazyme (K-SUFRG). The assay was scaled to 96 well plate format. Briefly ears, harvested in the field, were frozen immediately in liquid nitrogen. Fifteen kernels were then taken from the equatorial region midway between the tip and base, weighed and lyophilized. The dried kernels were pulverized in a geno-grinder and 10–20 mg were transferred to each of three 1.5 mL tubes. One mL of cold water (4°C) was added to each tube and vortexed. Samples were then incubated at 80°C for 15 min. Samples were centrifuged at maximum speed for 10 min and transferred to clean 1.5 ml tubes. Appropriate dilutions adjusted sugar concentrations to within the linear range of the standard curve (0–5 μg). Typical dilutions were 1:2 for glucose and fructose, and 1:10 for sucrose. Enzymes, ATP and NADP⁺ were added according to the Megazyme assay procedure but were scaled to accommodate a 96 well format. Standard curves for glucose, fructose and sucrose were used to quantify sugars in each sample. The amount of sucrose was calculated by subtracting the amount of glucose obtained in each sample and subtracting it from the amount of sucrose plus glucose in the sample. Starch was quantified as described by Boehlein et al. (2018).

RNA Extraction and Sequencing

For RNA sequencing, two kernels were randomly selected from each pooled-kernel sample from each of the five biological replicates and frozen in liquid nitrogen. In all the treatments, endosperm was dissected from each frozen kernel by removing the embryo, pericarp, and pedicel. Total RNA was isolated from the dissected endosperms using a CTAB based extraction as previously described (Chang et al., 1993). Samples were DNase I treated and purified using the New England Biolabs Monarch RNA Cleanup Kit (NEB). Endosperm dissection and RNA extraction were performed on each biological replicate. Quality of the RNA samples was assessed using gel electrophoresis and the three highest-quality RNA samples were then pooled in equimolar ratios for library preparation and sequencing. If there were not at least three high quality mRNA samples where 16S and 18S rRNA bands were clearly visible, the one or two high quality samples were used moving forward. This occurred only in 4 of the 52 resulting libraries were a pool of fewer than three biological replicates (Supplementary Figure S1). Libraries were prepared by the Cornell University Institute for Biotechnology using the Lexogen QuantSeq 3′ RNA-Seq Library Prep Kit FWD for Illumina according to the manufacturer’s protocol. Paired-end 75 bp reads were generated using Illumina NextSeq500.

Reads Processing and Differential Expression Analysis

Single end reads were demultiplexed, trimmed and quality filtered using Trimmomatic v0.32 (Bolger et al., 2014). The reads were then mapped to the B73v5 reference genome (Hufford et al., 2021) using Salmon v1.1 (Patro et al., 2017). Expression was measured by quantifying transcripts per kilobase of gene per million transcripts mapped (TPM). Genes with read counts < 10 across samples were removed to reduce background. The different genetic backgrounds for the same starch mutation were treated as the replications of the alleles of interest in order to find transcriptome changes caused by the perturbation in starch biosynthesis and that are not genotype specific. Gene expression was modeled as a function of genetic background, mutation, and age of the kernel (DAP) and DESeq2 (Love et al., 2014) was used to test for differential expression with a Bonferroni-corrected P-value threshold of 0.05. DEGs are defined as having an absolute value log₂ fold change > 2 unless otherwise specified.

Identification of Coexpression Networks

Gene coexpression modules were identified using the R package Weighted Correlation Network Analysis (WGCNA) (Langfelder and Horvath, 2008). Expression data for all genes in each sample was loaded into WGCNA. The similarity matrix was raised to the power of 6 to identify coexpression modules. Once modules were constructed, module eigengene values were determined using the first principal component of the module’s expression matrix. To test for module – phenotype associations, Pearson correlations were generated for all pairwise comparisons of each module eigengene expression value and measured kernel phenotype (sucrose, glucose, fructose, starch). Bonferroni adjustments corrected for multiple comparisons.

Gene Ontology and Pathway Enrichment Analysis

Gene Ontology (GO) terms for each gene were obtained from the B73v5eb annotation. The R package topGO v.2.38.1 was used to conduct GO enrichment analysis (Alexa and Rahnenfuhrer, 2020). GO terms with p-values ≤0.05 were considered over-represented. For pathway enrichment analysis, KOBAS3.0 software was used to test the statistical enrichment of DEGs in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (Xie et al., 2011). Pathways with a Bonferroni-corrected P-value of ≤0.05 were considered over-represented.

Results

Kernel Sugar Levels Are Affected by Mutation and Genetic Background

To assess the impact of starch mutations on endosperm sugar content in different maize backgrounds, we quantified sugar levels on the 18 isogenic lines (Figure 1). Milligrams of sucrose, fructose, and glucose per gram of dried tissue were quantified and summed for this analysis. At milk stage, the sh2 mutation led to a 255.9 mg g^–1 (P-value < 0.001) increase in all quantified sugars on average (Figure 1A and Supplementary Table S1). Contrastingly, the su1 mutation resulted in an average increase in sugars of 46.12 mg g^–1. However, this response was not consistent across genetic backgrounds. Although there is a pattern of increased sugar in su1 kernels at each time point, when all the genotypes are jointly analyzed, there is no significant difference from the Wt sugar levels (Figure 1B and Supplementary Table S1). In most backgrounds tested, su1 kernel sugar levels are also significantly higher than Wt by 21 DAP (P-value < 0.05) (Supplementary Figure S2 and Supplementary Table S1). Even taking developmental stage and mutation into account, genetic background has a significant impact on overall sugar levels (P-value < 0.05). Some backgrounds, most notably Ia5125, exhibit a stronger increase in sugars with the sh2 mutation (mean sugar levels increase by 365.6 mg g^–1) (Figure 1A and Supplementary Table S1). The sugar levels of other backgrounds are less responsive to the mutation. At 21 DAP, IL101T sh2 kernels only have 152.3 mg g^–1 more sugar than the wildtype kernels (Figure 1A and Supplementary Table S3). Sugar levels in the kernels decreased over time from 14 to 28 DAP (P-value < 0.001) (Figure 1B). By 21 DAP, sh2 kernels have significantly higher sugar levels than su1 or Wt kernels (P-value < 0.001) (Figure 1B and Supplementary Table S1).

FIGURE 1

Figure 1. Sugar levels across NILs. (A) Comparison of sugar levels in different NILs at milk stage (21 DAP). Values represent the mean sum of sucrose, glucose, and fructose in percent of the dry tissue weight ± SE. (B) Comparison of sugar level in different mutants at different developmental stages. Values represent the mean sum of sucrose, glucose, and fructose across genotypes in percent of the dry tissue weight ± SE.

Generation and Analysis of RNA-Seq Data

To investigate responses to perturbations of the starch biosynthesis pathway, we compared transcriptomes from mutant and wildtype endosperm and identified DEGs. We generated 489 million high-quality single-end Illumina reads from our 52 libraries (Supplementary Figure S1). Reads were uniquely mapped to the B73v5eb transcriptome (Hufford et al., 2021) with an average mapping rate of 68.9%. Uniquely mapped reads were used to estimate normalized transcript level as transcripts per kilobase of gene per million transcripts mapped (TPM). In total, 25,972 genes were carried forward for further analysis. Principal component analysis of the TPM counts indicates that developmental stage is the most important factor in transcriptome composition (Figure 2A). Thus, all further differential expression (DE) comparisons were performed within a given developmental stage. Each genotype (sh2, su1, Wt) was represented by combining results across backgrounds. By using different genetic backgrounds as replicates, we aim to identify conserved responses to these mutations in the maize endosperm. Using an adjusted p-value threshold of 0.05 and a log₂ fold change cutoff of two, 667 DEGs were identified between sh2 and Wt endosperm and 77 DEGs were identified between su1 and Wt endosperm in at least one stage of development (Figures 2B,C). In sh2 endosperm, 18 genes are DE at all developmental stages examined, while in su1 endosperm no genes are DE in all developmental stages. The 18 consistently DE genes in sh2 endosperm include Shrunken2, protein kinases and chromatin regulators. The lowest number of DEGs for both comparisons (sh2 vs. Wt and su1 vs. Wt) is found at blister stage (Figures 2B,C). In sh2 endosperm, the number of DEGs rapidly increases from 48 DEGs at blister stage to 401 DEGs milk stage and then slightly declines to 379 DEGs at dent stage (Figure 1B). In su1 endosperm, there is a slight increase in DEGs between blister stage (5 DEGs) and milk stage (12 DEGs), followed by a more dramatic increase to 60 DEGs at dent stage (Figure 2C).

FIGURE 2

Figure 2. Transcriptome/DE summary. (A) Clusters of samples from different mutant endosperm over developmental time identified by principal component analysis of gene expression levels. Colors represent timepoints while shapes indicate mutant. (B) Venn diagram of DEGs in sh2 endosperm at different timepoints. (C) Venn diagram of DEGs in su1 endosperm at different timepoints.

Mutant vs. Wildtype Endosperm – Gene Ontology and KOBAS Enrichment

To capture broad effects of sh2 and su1 mutations, genes that were at least moderately differentially expressed (Bonferroni corrected P-value ≤ 0.05, absolute value log₂ fold change > 1) were tested for KEGG pathway enrichment using KOBAS 3.0 (Xie et al., 2011) and GO term enrichment using the R package topGO (Alexa and Rahnenfuhrer, 2020). The sh2 endosperm showed a downregulation of genes in the pathways for carbon metabolism and starch and sucrose metabolism. There was an upregulation of genes involved in glycolysis and gluconeogenesis, the pentose phosphate pathway, the citric acid cycle, plant hormone signal transduction, and fatty acid metabolism (Table 1 and Supplementary Table S2). The su1 endosperm showed an upregulation of genes involved in butanoate and glutamate metabolism (Table 1 and Supplementary Table S2). GO enrichment analysis of DEGs revealed that genes up- or down- regulated in sh2 are enriched for terms related to polysaccharide catabolism, sucrose metabolism, trehalose metabolism, 1,3-beta-D-glucan synthesis, glycolysis, and lipid biosynthesis (Supplementary Table S3). Genes up- and down-regulated in su1 are enriched for terms related to peptidase activity, carboxylyase activity, carboxylic acid metabolic process, metal ion transport, and malate dehydrogenase activity are significantly enriched (Supplementary Table S3).

TABLE 1

Table 1. KOBAS pathway enrichment results.

Differential Expression in su1 Kernels

A total of 77 genes were significantly DE between su1 and wildtype endosperm, 18 of which overlap with DEGs in sh2 endosperm when compared to wildtype. The transcriptomic profiles between wildtype and su1 kernels were most different at dent stage (28 DAP), near the completion of starch synthesis (Supplementary Table S4). At this timepoint, out of the list of DEG, an alcohol dehydrogenase Zm00001eb317510 is upregulated with a log₂ fold-change of 2.9. Furthermore, transcription factors exhibit differential regulation in su1 endosperm. AP2/ERF transcription factor ZmEREB4 is upregulated with a log₂ fold-change of 3.1, while transcription factors ZmHB37 and ZmC3H39 are downregulated (log₂ fold-change of −2.2 and −2.5, respectively). An F-box containing protein (Zm00001eb067000) is significantly downregulated in su1 endosperm with a log2 fold-change of −7.1. Additionally, a number of auxin related genes including ZmIAA24 are downregulated in su1 endosperm (log₂ fold-change = −2.8). Interestingly, the aforementioned alcohol dehydrogenase and ZmEREB4 are also upregulated in sh2 endosperm (Supplementary Tables S4, S5).

Impacts of the sh2 Mutation on Carbohydrate and Sugar Metabolism Gene Expression

Many genes involved in sugar and carbohydrate metabolism were differentially expressed in response to the sh2 mutation (Table 1 and Supplementary Tables S2, S3, S5). To further characterize changes to the pathway, we investigated expression changes of genes in the KEGG starch and sucrose metabolism pathway (Figure 3). The strongest downregulation in response to the sh2 mutation is that of the Shrunken2 gene itself with an average log₂ fold change of −7.15. Other AGPases including ZmAGP2 (log₂ fold change = 2.06) and Brittle2 (log₂ fold change = 1.34) are upregulated in the sh2 mutant (Figures 3, 4A). A UDP glucose pyrophosphorylase (UDPGP) is also upregulated (log₂ fold change = 2.3) in sh2 endosperm (Figure 3 and Supplementary Table S5). Genes involved in starch synthesis and starch degradation, including Granule-bound starch synthase2 (ZmGBSS2), Pullulanase (ZmZPU1), and multiple β-amylases show significantly decreased expression in sh2 endosperm. Starch phosphorylase ZmPHO1 is slightly upregulated in sh2 endosperm (log₂ fold change = 1.1) (Figure 3 and Supplementary Table S5). Notably, genes involved in trehalose metabolism show differential expression in response to the sh2 mutation. Trehalose-6-phosphate (T6P) synthetase 11 (ZmTRPS11) is downregulated in sh2 endosperm while T6P phosphatase2 (ZmTRPP2) is upregulated (Figures 3, 4B). In addition to sugar and starch metabolism genes, sugar transporters are differentially expressed in sh2 endosperm (Figure 4C). Sugar equilibrators including ZmSWEET15a, ZmSWEET11a, and ZmSWEET2a are significantly upregulated in sh2 endosperm. Contrastingly, sucrose transporter ZmSUT7 is significantly downregulated in sh2 endosperm (Figure 4C).

FIGURE 3

Figure 3. Expression changes of genes in the starch biosynthetic pathway in maize endosperm. Enzymes with gene expression changes in sh2 endosperm are indicated with colored backgrounds. Blue boxes indicate increased expression in sh2 endosperm, while red boxes indicate decreased expression in sh2 endosperm. Suc, sucrose; Fru, fructose, Glc, glucose; Tre, trehalose; Glc-1-P, glucose 1-phosphate; Glc-6-P, glucose 6-phosphate, Fru-6-P, fructose 6-phosphate; Fru-1,6-BP, fructose 1,6-bisphosphate; Tre-6-P, trehalose 6-phosphate; Gly-3-P, glycerol 3-phosphate; DHAP, dihydroxyacetone phosphate; PPi, pyrophosphate; ADP-Glc, ADP glucose; UDP-Glc, UDP glucose; SUS, sucrose synthase; UGPase, UDP-glucose pyrophosphorylase; AGPase, ADP-glucose pyrophosphorylase; HXK, hexokinase; FK, fructokinase; PFK, phosphofructokinase; FBP, fructose 1,6 bisphosphatase; TPI, triose phosphate isomerase; PGM, phosphoglucomutase; PGI, phosphoglucoisomerase; PHO, starch phosphorylase; SS, starch synthase; SBE, starch branching enzyme; DBE, debranching enzyme; TPS, trehalose-6-phosphate synthase; TPP, trehalose-6-phosphate phosphatase.

FIGURE 4

Figure 4. Differential expression in maize endosperm. Relative expression levels of DEGs in each mutant/timepoint combination. Values shown represent the log₁₀ of the TPM summed across transcripts and averaged across backgrounds. (A) Relative expression of AGPase-encoding genes. (B) Relative expression of trehalose-6-phosphate metabolizing enzymes. (C) Relative expression of sugar-transporting genes. (D) Relative expression of genes encoding glycolytic enzymes. (E) Relative expression of AP2/ERF domain transcription factors.

The sh2 Mutation Increases Glycolysis and Pentose-Phosphate Pathway Gene Expression

As presented above, the large increase in sucrose and hexose content in sh2 endosperm results in a significant enrichment of DEGs in the glycolysis and gluconeogenesis pathway and in the pentose phosphate pathway (Table 1 and Supplementary Table S2). Most DEGs in the glycolysis and gluconeogenesis pathway are upregulated. DEGs include alcohol dehydrogenase ZmADH2 (log₂ fold change = 2.1), aldehyde dehydrogenases ZmALDH3 and ZmALDH13 (log₂ fold change = 2.5, 2.8, respectively), a pyruvate kinase (Zm00001eb433740; log₂ fold change = 2.42), a pyruvate-phosphate dikinase (Zm00001eb278140; log₂ fold change = 2.15), and a moderate upregulation (average log₂ fold change = 1.2) of the three annotated maize pyruvate decarboxylases (ZmPDC1, ZmPDC2, ZmPDC3) (Figure 4D and Supplementary Table S5). Interestingly an enolase (Zm00001eb098760) is the only part of the glycolytic pathway significantly downregulated (log₂ fold change −2.1) (Figure 4D). The pentose phosphate pathway is also significantly enriched with DEGs. Twelve genes in the pentose phosphate pathway are DE in sh2 endosperm. Eleven of these genes are moderately but significantly upregulated (average log₂ fold change = 1.34) (Supplementary Table S5). These include transketolases, a transaldolase, pyruvate dehydrogenase, phosphofructokinases, and phosphogluconate dehydrogenase ZmPGD4. The exception to this trend is a ribose-5 phosphate isomerase, which is downregulated in sh2 endosperm (log₂ fold change = −1.4) (Supplementary Table S5). Taken together, this indicates that the upregulation of genes in the pentose phosphate and glycolytic pathways are important components of the response to sh2 and/or the increased sucrose levels in the endosperm.

Differential Expression of Apetala2 Transcription Factor-Like Transcription Factors in sh2 Endosperm

Several genes involved hormone response are DE in sh2 endosperm. Notably, eight Apetala2 transcription factor-like (AP2/ERF) domain transcription factors are differentially expressed in the sh2 endosperm (Figure 4E). Six AP2/ERF TFs are upregulated (ZmWri1, ZmWri2, ZmEREB4, ZmEREB31, ZmEREB136, ZmEREB180) in the sh2 endosperm while two (ZmEREB150, ZmDBP4) are downregulated (Figure 4E). An additional AP2/ERF TF, ZmEREB156, exhibited moderate downregulation in sh2 endosperm with a log₂ fold-change of −1.8 at 28 DAP (Figure 4E). Previously identified target genes of ZmWri1 (Pouvreau et al., 2011) show moderate (1 < log₂ fold-change < 2) but significant upregulation in sh2 endosperm. These targets include ZmPYK3, pyruvate dehydrogenases (Zm00001eb050410, Zm00001eb031900, Zm00001eb138800), acyl carrier proteins (Zm00001eb331920, Zm00001eb102360), ZmTHL1, and ZmGLPDH4 (Supplementary Table S5).

Zein-Encoding Genes and Nitrogen Related Genes Have Altered Expression in sh2 Endosperm

Zeins compose the majority of storage proteins in the maize endosperm and are encoded by a large gene family (reviewed in Li and Song, 2020). To further explore the dynamics of protein biosynthesis in sh2 endosperm, we evaluated the expression of zein-encoding and other protein-related gene families. Overall, expression changes in zeins and amino acid transporters indicate differential regulation of grain filling in sh2 endosperm. Twelve zeins, including ZmFloury4, are downregulated in sh2 endosperm (log₂ fold change between −2.1 and −5.6) (Supplementary Table S5). Most (8 of 12) of the differentially regulated zeins are annotated 19 kD or 22 kD alpha zeins, except for Zm00001eb313790, a 50 kD gamma zein. An endosperm-expressed nitrate transporter1/peptide transporter family (NPF) protein (Zm00001eb112670) and a transmembrane amino acid transporter (Zm00001eb257490) are also downregulated in sh2 endosperm (log₂ fold change −2.2, −2.6, respectively). However, a homolog of AtAAP1 (Zm00001eb384590) and nitrate-reductase ZmNNR2 are upregulated (log₂ fold change 2.6, 2.3, respectively) (Supplementary Table S5). The GABA shunt also plays a role in maintaining the carbon:nitrogen balance in the cells (Fait et al., 2008; Batushansky et al., 2014). Two glutamate decarboxylases (Zm00001eb399220, Zm00001eb081550), which catalyze first step in the GABA shunt, are upregulated in sh2 endosperm (log₂ fold change 1.6, 2.1, respectively) (Supplementary Table S5).

Genetic Background Affects Sugar Accumulation in sh2 Kernels

To gain insight into the background effect of sugar accumulation in the sh2 endosperm, DEGs were compared across genetic backgrounds. Backgrounds were classified as “high-sugar-accumulating” or “low-sugar accumulating” based on whether the increase in sugars of sh2 kernels was significantly above or below the average change (+255.9 mg g^–1) (Figure 1A, Supplementary Figure S2, and Supplementary Table S1). Most notably, core histone ZmHIS2B5 is downregulated in the low-sugar-accumulating backgrounds IL101T and C40 (log₂ fold change = −3.7, −6.2, respectively) and either slightly upregulated (Ia453, log₂ fold change = 1.6) or not significantly DE in high-sugar-accumulating genotypes (Supplementary Table S6). A helix-loop-helix transcription factor (Zm00001eb101410; ZmBHLH157) is specifically upregulated in low-sugar-accumulating backgrounds (Table 2 and Supplementary Table S6). Other expression changes are specific to high-sugar-accumulating backgrounds, including the downregulation of a protein of unknown function (Zm00001eb250800) and a protein tyrosine kinase (Zm00001eb158770; ZmMKKK55) and the upregulation of a glutathione S-transferase (Zm00001eb397280) (Table 2 and Supplementary Table S6).

TABLE 2

Table 2. Differential expression of putative modifiers.

Coexpression Networks in the Maize Endosperm Identified Nine Modules With an Overrepresentation of Carbohydrate Metabolism Genes

Expressed genes were classified into 48 coexpression modules using the WGCNA R Software package (Langfelder and Horvath, 2008) and modules were assigned a color for naming and visualization purposes (as seen in Figure 5A). Module sizes ranged from 50 to 2643 genes with an average size of 541. Analysis for GO Term enrichment using topGO (Alexa and Rahnenfuhrer, 2020) revealed nine modules with an overrepresentation of carbohydrate metabolism genes. These modules were carried forward for further analysis using Cytoscape’s Network Analyzer (Assenov et al., 2008). Module eigengenes were calculated, and pairwise correlations with sugar- and starch-level phenotypes were generated (Figure 5A). Module Black, which contains an overrepresentation of genes related to trehalose metabolism (GO:0005992), pyruvate kinase activity (GO:0004743), G3P metabolism (GO:0004367; GO:0046168), oligosaccharide biosynthesis (GO:0009312), and cellulose metabolism (GO:0016760; GO:0030244) (Supplementary Table S7), has its eigengene positively associated with sugar levels (sucrose = 0.42, glucose = 0.29, fructose = 0.28) and negatively associated with starch levels (−0.37) (Figure 5A). In this module, ZmTRPS10 has the highest betweenness centrality, suggesting a central role in regulating this group of genes (Supplementary Table S8). Other modules including Lightsteelblue1 and Grey60 show strong correlations with sugar and starch levels as well. Lightsteelblue1’s eigengene is associated with sucrose (0.5), starch (−0.37), and adjusted sugars (0.28) (Figure 5A). Lightsteelblue1 is a module consisting of 75 genes (Supplementary Table S9). Sugary enhancer1 (ZmSE1, Zm00001eb115450) is a key regulator of this network with the highest closeness centrality and betweenness centrality values. The hub gene, or gene with the highest number of connections to other genes in the module, is the GRAS transcription factor Scarecrow-like 1 (ZmSCL1, Zm00001eb344400). Module Lightsteelblue1 appears to be involved in the regulation of genes at the protein level with enriched GO Terms related to protein phosphatase regulatory activity and an overrepresentation of genes in the KEGG RNA transport pathway (specifically translation initiation) (Supplementary Table S3). Lightsteelblue1 contains a number of genes related to ubiquitin-mediated protein degradation including 26 proteasome component Zm00001eb232220, ubiquitin conjugating enzyme Zm00001eb295710, and F-box protein ZmARABIDILLO1 (Zm00001eb023660) (Figure 5B). Module Grey60 has the strongest association with sugar levels – its eigengene is significantly associated with sucrose (−0.85), glucose (−0.79), fructose (−0.78), starch (0.76), adjusted sugars (−0.37) and adjusted starch (0.29) (Figure 5A). Module Grey60 consists of 265 genes and is enriched with genes related to glycolysis (GO:0006096) and the non-oxidative branch of the pentose-phosphate pathway (GO:0009052) (Supplementary Tables S7, S8). The gene with highest betweenness centrality in this module is ZmHXK1. The hub gene, or most connected gene, of Module Grey60 is a 50 kD Gamma zein Zm00001eb313790. There are 15 additional zein genes in this module (Figure 5C and Supplementary Table S8).

FIGURE 5

Figure 5. WGCNA results. (A) Pairwise correlations between module eigengenes and phenotypic values. Significance (P-value < 0.05) is indicated with an asterisk. (B) Visual representation of Module Lightsteelblue1 using the yFiles organic layout format. SCF-complex genes are indicated in green, genes that impact grain filling and starch metabolism are indicated in pink, and transcription factors in purple. Connection weight is indicated by line color, where darker lines represent higher-weighted connections. (C) Module Grey60 represented by Cytoscape Network Analyzer using the yFiles organic layout format. Genes encoding zein proteins are indicated in yellow, genes involved in carbohydrate metabolism are indicated in pink, transcription factors in purple and glycolytic genes in orange. Connection weight is indicated by line color, where darker lines represent higher-weighted connections.

Differential Expression in se1 Kernels

To further investigate the role of Se1 in transcriptional regulation, we compared RNAseq profiles from two replicates of wildtype and se1 kernels sampled at 11, 15, and 19 DAP (Zhang X. et al., 2019). We downloaded RNAseq data from via the National Center for Biotechnology Information Sequence Read Archive BioProject PRHNA287265 (Zhang X. et al., 2019). Reads were trimmed and uniquely mapped to the B73v5 transcriptome (Hufford et al., 2021) with an average mapping rate of 83.8%. Low-count transcripts (transcripts with <10 counts total) were removed, leaving 25964 genes for further analysis. The R package DESeq2 was used to identify 116 total DEGs (log₂ fold change > 2, Bonferroni-corrected P-value < 0.05) between se1 and Wt kernels over the three timepoints. Six DEGs were identified between Wt and se1 kernels at 11 DAP, 12 DEGs at 15 DAP, and 112 DEGs at 19 DAP (Supplementary Table S9). Certain regulatory genes are oppositely DE (compared to Wt) in se1 kernels and sh2 endosperm. A homolog of AtSNF4 (Zm00001eb139880) is upregulated in sh2 and downregulated in se1 kernels, while the GRAS transcription factor ZmSCL1 is upregulated in se1 kernels and downregulated in sh2 endosperm. In addition to ZmSCL1, other gibberellin-related genes are DE in se1 kernels. A gibberellin degradation enzyme, GA4 2β-hydroxylase (Zm00001eb412670), is the most upregulated gene in se1 kernels (log₂ fold change = 7.07). B-glucosidase 6, a gene involved in gibberellin-ester hydrolysis, is also upregulated in se1 kernels (log₂ fold change = 2.7). Additionally, starch degradation genes known to be activated by GA (MacLeod and Millar, 1962) are upregulated in se1 kernels (Endoglucanase, Zm00001eb015719, log₂ fold change = 1.67) (Supplementary Table S9). It is also notable that genes related to galactose and raffinose metabolism are DE in se1 kernels. ZmRAFS3 is downregulated (log₂ fold change = −8.9) at the 19 DAP timepoint in se1 kernels while another raffinose synthase (Zm00001eb005800) is upregulated at this timepoint (log₂ fold change = 2.7). Additionally, two β-galactosidases (Zm00001eb152540, Zm00001eb383110) are upregulated in se1 kernels at 19 DAP (log₂ fold change = 5.3, 4.5, respectively) (Supplementary Table S9).

Discussion

Mutation and Timepoint Impact the Endosperm Transcriptome

Shrunken2 encodes the large subunit of the endosperm AGPase, responsible for synthesis of ADP-glucose used for starch production in the kernel. The mutant, sh2, is a complete knock out of this gene, and as such creates a drastic kernel phenotype where kernels have ∼25% wildtype starch content and a shriveled (shrunken) pericarp (Hannah, 1997). Sugary1 encodes maize isoamylase I. Mutations in su1 have a less severe impact on starch production and grain filling in the kernels. Kernels have approximately 50% of their wildtype starch content. The total number of DEGs between sh2 and wildtype endosperm is much larger than DEGs between su1 and wildtype (667 DEGs vs. 77 DEGs) (Figures 2B,C and Supplementary Tables S4, S5). This result is consistent with expectations due to the severity of the impact sh2 has on the kernel development and endosperm composition. Starch biosynthesis and sugar accumulation begins at ∼10 DAP and continues until ∼30 DAP. We isolated mRNA from wildtype and mutant (sh2 and su1) endosperm at multiple developmental stages spanning starch biosynthesis and sugar accumulation (blister stage at 14 DAP, milk stage at 21 DAP, and dent stage at 28 DAP). Developmental timepoint had a strong impact on the overall transcriptome profile (Figure 2A). As such, we analyzed each developmental stage separately to identify DEGs between mutant and wildtype kernels. Interestingly, only 18 genes were differentially regulated in sh2 endosperm at all time points. These genes include protein kinases and chromatin remodeling proteins, suggesting that part of the changes in sh2 endosperm may be post-translationally and/or epigenetically regulated (Figure 2B). Between sh2 and wildtype endosperm, the number of DEGs increased from 48 at 14 DAP to 401 at 21 DAP (Figure 2B and Supplementary Table S5). In wildtype kernels, activity of the Shrunken2 enzyme activity peaks at approximately 22 DAP (Tsai et al., 1970). During this developmental stage, we also see the largest difference in sugar content between sh2 and wildtype kernels (Figure 1B). The differences in the transcriptomes are also more pronounced at this timepoint. Contrastingly, DEGs between su1 and wildtype endosperm peaked later at 28 DAP (Figure 2C and Supplementary Table S4).

Genetic Background Effects Impact Phenotypic Response to Mutations

Mutations in the starch biosynthetic pathway have variable effects depending on the genetic background (Shuler et al., 2017). To assess the degree of difference between our isogenic lines, sugar levels were quantified in kernels at three key stages of development. The sh2 mutation resulted in a strong increase in sugars compared to both wildtype and su1 kernels in all genetic backgrounds at 21 and 28 DAP (Figures 1A,B, Supplementary Figure S2, and Supplementary Table S1). The su1 mutation, however, resulted in a more subtle sugar increase. In fact, when all genotypes are taken together, the su1 mutation does not significantly increase sugar levels (Figure 1B and Supplementary Table S1). This is due to the variability in response to the mutation between genetic backgrounds (Supplementary Figure S2). The large effect of genetic background potentially explains the lack of significant DEGs in su1 endosperm when results are combined across backgrounds. Genotypic background variability is observed in the sh2 response as well. In some genotypes, the increase in sugars is less dramatic and not significant until later in development (Supplementary Figure S2 and Supplementary Table S1). In addition to the sh2 and su1 mutations, many different regulators or modifiers of the starch biosynthetic pathway have been identified, including the transcription factor, ZmNAC130 (Zhang Z. et al., 2019). Variability of backgrounds in their sugar accumulation (Supplementary Figure S2 and Supplementary Table S1) allowed us to further pursue such putative modifiers of endosperm composition in sh2 kernels. To identify putative regulators and modifiers, we searched for genes whose expression changes correlate with sugar accumulation in response to sh2 (Supplementary Table S6). Notably, histone ZmHIS2B5 and MAPKKK protein ZmMKKK55 expression profiles correlate to low- and high-sugar-accumulating backgrounds, respectively, indicating a potential regulatory role modulating the impact of sh2 (Table 2 and Supplementary Table S6).

The sh2 Mutation Alters Expression Profiles of Genes in the Starch Metabolism Pathway

Pathway and GO term enrichment analysis reveal a significant overrepresentation of starch biosynthesis and degradation genes that are differentially regulated in sh2 kernels. As sh2 disrupts the first committed step of starch biosynthesis, many of the DEGs involved in starch synthesis are downregulated (Table 1 and Supplementary Table S2). Starch synthases and starch degradation genes, including amylases, are also downregulated (Figure 3), likely due to the increase in sugars in the endosperm enhancing sink activity (Bihmidine et al., 2013; Guan and Koch, 2015). Shrunken2 transcripts are observed at much lower levels with an average log2 fold change of −7.15 (Figure 4A). Other AGPases, including Brittle2 (ZmBT2) are upregulated in the sh2 endosperm (Figure 4A), which corroborates previous findings (Giroux et al., 1994). In addition to other AGPases, a UDP-glucose pyrophosphorylase (UDPGP) gene is upregulated in sh2 endosperm (Supplementary Table S5), possibly due to an increased concentration of glucose 1-phosphate. Interestingly, certain starch synthases have demonstrated activity using UDP-Glucose as a substrate, although at much lower levels than ADP-Glucose (Macdonald and Preiss, 1983). Starch phosphorylase 1 is upregulated in sh2 endosperm (Figure 3 and Supplementary Table S5). Starch phosphorylase1 (ZmPHO1) is the plastidial starch phosphorylase and is typically active in the endosperm amyloplasts (Yu et al., 2001; MaizeGDB.org, 2021). The function of starch phosphorylase in maize kernels is unclear, as evidence has suggested a role in both degradative and synthetic processes (Subasinghe, 2013). Under specific conditions, disruptions in starch phosphorylase activity have resulted in reductions in starch accumulation (Satoh et al., 2008). ADP-glucose typically inhibits starch phosphorylase activity in the synthetic direction during later developmental stages (Hwang et al., 2010). Upregulation of ZmPHO1 in sh2 endosperm in later stages of kernel development further supports the hypothesis that in the absence of AGPase function, ZmPHO1 could be responsible for the remaining 25% starch production in the endosperm (Boehlein et al., 2018).

Sugar-Responsive Genes for Sink Strength

Among the genes for source-sink adjustment affected here are those aiding sugar transport, with upregulation of mRNAs for sugar equilibration in sinks versus downregulation of those related to phloem loading in sources (SWEETs and SUTs, respectively). Members of the SWEET (Sugars Will Eventually be Exported Transporter) protein family transport sucrose and/or hexoses in the direction of the concentration gradient (Chen et al., 2010). Members of the SUT (sucrose transporter) protein family transport sucrose against the concentration gradient and are often involved in phloem loading (reviewed in Braun and Slewinski, 2009). The SWEET genes upregulated in sh2 endosperm (Figure 4C) include SWEET2, a member of the glucose-equilibrating Clade I, and two members of the sucrose-equilibrating Clade III (SWEET11a and SWEET15a; Eom et al., 2015). In contrast, the SUT7 downregulated in sh2 endosperm (Figure 4C) has a close homolog, SUT1, with a demonstrated role in phloem loading (reviewed in Julius et al., 2017).

It is likely that the increase of sucrose content in sh2 endosperm activates several signaling pathways, including ABA pathways. Sucrose has been shown to impact ABA signaling, and together they have a demonstrated synergistic impact on gene expression (Rook et al., 2001; Akihiro et al., 2006). Expression profiles change for many transcription factors in sh2 endosperm, most notably those containing AP2/ERF domains (Figure 4E). Transcription factors with this domain, are responsive to sugars, ABA, and/or ethylene, and frequently have roles integrating these sensing pathways (Weigel, 1995; Finkelstein et al., 1998; Fujimoto et al., 2000). Most are upregulated in sh2 endosperm (Figure 4E). Some are downregulated, such as an EREB156 that otherwise enhances starch production in response to sucrose and ABA (Huang et al., 2016). Other sugar modulated AP2/ERF genes upregulated here that would favor sink strength are ZmWri1 and ZmWri2, orthologs of Wrinkled1 from Arabidopsis thaliana (Figure 4E). This AP2/ERF transcription factor is a known regulator of both glycolysis and fatty acid biosynthesis (Baud et al., 2009). In addition, overexpression of the maize Wri1 increases the oil content of kernels (Shen et al., 2010). The upregulation of ZmWri1 and ZmWri2 in sh2 endosperm correlates with enhanced expression of glycolytic genes, genes for fatty acid biosynthesis, metabolism, and elongation (Table 1, Figure 4D, and Supplementary Table S2), as well as the increased fatty acid levels of sh2 kernels (Ratcliff et al., 1993).

A Potential Role for Hexokinase1 in Grain Filling and Composition

Several lines of evidence indicate prominent input from the hexokinase-sensing system in the sh2 endosperm. Among these is the abundance of hexokinase-responsive genes in the transcriptome (Figures 4D, 5C and Supplementary Tables S5, S8). Most follow a classical feast or famine framework (Koch, 1996) that adjusts source-sink balance, with sugar abundance enhancing mRNAs for sink strength and repressing those for source activity (Supplementary Tables S2, S5 and Figure 4C; Bihmidine et al., 2013; Guan and Koch, 2015). Coexpression network analysis reveals multiple roles for HXK1 in the maize endosperm. Structure of this module links the breadth of previously known HXK1-responsive genes to those for zein storage proteins and to transcription factors that contribute to grain filling (Figure 5C; Aguilera-Alvarado and Sánchez-Nieto, 2017). In the present work, the HXK1-coexpression module also correlates most strongly with the sugar and starch content of developing kernels, further emphasizing its importance to grain filling (Figure 5A). As noted above, HXK1 modulates diverse genes (Reviewed in Aguilera-Alvarado and Sánchez-Nieto, 2017), with those upregulated by sugar abundance typically enhancing sink strength relative to source supplies (Bihmidine et al., 2013; Guan and Koch, 2015). In contrast, opposite results, as observed here for N-responses, are expected if a sugar-rich system senses disproportionally limited N-resources (Koch, 1997). Genes for acquisition of N-assimilates are upregulated (those for amino-acid biosynthesis and the GABA shunt), while those for storage in N-sinks are downregulated (zein storage proteins). The decreased expression of genes for zein storage proteins seen here (Supplementary Table S5) is corroborated by other studies of the sh2 mutant (Sandrine et al., 2020) and of the high-sugar metabolic environments of low-AGPase kernels (Zhang Z. et al., 2019). Work here reveals that ZmHXK1 is centrally relevant to control of the C:N balance prominent in response to the sh2 mutation. A still stronger integration of ZmHXK1 into the control network for C:N balance is evident in the presence of two NAC transcription factors (ZmNAC12 and ZmNAC130) in the same module (Figure 5C and Supplementary Table S8). Among these, previous work by Zhang Z. et al. (2019) indicated that ZmNAC130 is a key regulator of grain filling and could bind the promoter motif of 50 kD gamma zeins (one of these being a hub gene in the coexpression network identified here) (Figure 5C and Supplementary Table S8). Furthermore, when ZmNAC130 was downregulated, so was ZmHXK1. Taken together, these results indicate that dual action of ZmHXK1 as a catalytic enzyme and transcription factor upregulates sink metabolism and coordinates zein expression, C:N balance, and the impacts of other transcription factors like ZmNAC130.

Sugar Signaling Plays a Role in the Molecular Response to the sh2 Mutation

Data here indicate that endosperm perturbation with a sh2 mutation affects both the sucrose- and hexose-based sensing systems. Differential impacts of sucrose and hexose “states” (Wobus and Weber, 1999) are now recognized as effects of hexokinase (HXK) and trehalose-based (TPS-TPP) signaling systems (Bihmidine et al., 2013; Guan and Koch, 2015). Hexose abundance favors cell division and expansion whereas sucrose enhances differentiation and maturation (Guan and Koch, 2015). In sh2 endosperm, many hexose sensing pathways including glycolysis and the pentose-phosphate pathway are differentially regulated (Table 1 and Supplementary Table S2). Overall, we see an increase in expression of the enzymes involved in glycolysis and the pentose phosphate pathway in sh2 endosperm (Table 1, Figure 4D, and Supplementary Table S2). This is likely due to the dramatic increase of sugars and/or changes in flux through these sensing pathways in the kernel (Guan and Koch, 2015). Flux through these pathways is important to many secondary processes as well as regulatory mechanisms. Glycolysis in particular is centrally important in the low-oxygen endosperm of maize (Rolletschek et al., 2005), as well as for sugar signals generated in the first step of this path by hexokinase (Bihmidine et al., 2013). The hexokinase responsiveness of glycolytic genes seen here and elsewhere provides a feed-forward mechanism for enhancing capacity of both glycolytic flux and sensing by hexokinase (Guan and Koch, 2015). In addition to glycolysis, other genes related to oxidative stress and excess respiration are upregulated in sh2 endosperm (Supplementary Table S5). Upregulation of these pathways is typical of sweet corn backgrounds (de Sousa et al., 2008) as a response to the high-sugar, low oxygen environment of the maize endosperm (Sanclemente et al., 2021). In non-photosynthetic plastids, the main function of glycolysis is to generate carbon skeletons and ATP for processes including fatty acid synthesis (reviewed in Plaxton, 1996). The reducing power for fatty acid synthesis could also be provided by the oxidative pentose phosphate pathway (Kang and Rawsthorne, 1996). Increased flux of sucrose through these pathways probably contributes to the four-fold increase of fatty acids (by weight) in sh2 endosperm (Ratcliff et al., 1993).

Input into the TPS-TPP system may differ here despite the elevated levels of sucrose in sh2 mutants. Genes involved in this sensing path and those responsive to its signals are clearly affected (Figures 3, 4B and Supplementary Table S5). In the TPS-TPP sensing system, trehalose is synthesized from UDP-glucose and glucose-6-phosphate by trehalose-6-phosphate synthase (TPS) generating the intermediary, trehalose-6-phosphate (T6P) which is subsequently dephosphorylated by trehalose-6-phosphate phosphatase (TPP) (Henry et al., 2014). This pathway, specifically T6P levels, are implicated in many signaling and regulatory roles and generally indicative of sucrose abundance (Yadav et al., 2014; Oszvald et al., 2018). Here, a kernel-expressed TPP (ZmTRPP2) is upregulated in sh2 endosperm and a class-II TPS (ZmTRPS11) is downregulated (Figure 4B). Class-II T6P synthases are non-catalytic proteins that have been conserved throughout evolution of land plants although their function has not yet been identified. Recent studies have implicated their role in transcriptional regulation (Vandesteene et al., 2010; Barraza et al., 2016). Interestingly, one of the hub genes identified here via WGCNA analysis was also a class-II TPS (ZmTRPS10) (Supplementary Table S8, black module). This module is highly correlated with overall starch and sugar levels and is enriched for genes with roles in carbohydrate metabolism (Figure 5A and Supplementary Table S7, black module). In addition, Hu S. et al. (2021) identified ZmTRPS10 (Zm00001eb192680; called ZmTPS9 in their work) as a QTL for overall starch content and, using mutant analysis, validated this role plus a contribution to seed yield. Taken together, these results indicate that Class-II TPS’s, and specifically ZmTRPS10, are central players in the transcriptomic response to the sh2 perturbation of maize endosperm.

Impacts of Sugary Enhancer1 Implicate Prominent Roles for Gibberellin Signaling

Coexpression network analysis identified Sugary Enhancer1 (SE1) as a key regulator of the maize endosperm transcriptome. The SE1 gene is a long-known effector of starch and sugar metabolism in maize kernels (Gonzales et al., 1976; La Bonte and Juvik, 1990) and is revealed here to be a central regulator of a coexpression module (Figure 5C). This module also includes key regulatory genes such as the GRAS transcription factor, Scarecrow-like1 (SCL1). Also present are multiple genes related to protein-level regulation including a 26S proteasome, a ubiquitin conjugating enzyme, and an F-box protein ARABIDILLO. Notably, the 26S proteasome-ubiquitin pathway has a role in gibberellin signaling (Wang and Deng, 2011). Also, the F-box protein, ARABIDILLO1, has been shown to repress effects of GA₃ in Arabidopsis roots (Mu et al., 2010) and aid regulation of seed germination (Moody et al., 2016). To further explore the relationship between SE1 and gibberellins, we analyzed DEGs in RNA-seq data from se1 kernels (Zhang X. et al., 2019). In se1 kernels, SCL1 and other GA related genes are differentially expressed. In contrast to sh2 kernels where SCL1 is downregulated, SCL1 is upregulated in se1 kernels. The downregulation of SCL1 in sh2 kernels compared to Wt could result from increased ABA response to excess sucrose, as ABA and GA are known to counter one another. Multiple genes for GA-metabolism are upregulated in se1, as are GA-activated genes for starch degradation. Taken together, our results indicate that SE1 could impact starch metabolism via regulation of GA, potentially through control of SCL1. The suggested mechanism is further supported by a spectrum of typical GA responses observed here in se1 kernels, including the increased sugars (specifically maltose), decreased starch (Zhang X. et al., 2019), and upregulation of germination genes such as endoglucanases (MacLeod and Millar, 1962) and β-galactosidases (Giannakouros et al., 1991; Chantarangsee et al., 2007).

Conclusion

In this study, we utilized near isogenic lines (NILs) of starch biosynthetic mutants su1 and sh2 to perturb the maize endosperm transcriptome during progressive developmental stages. Mutation of the Su1 gene led to minimal changes in the endosperm transcriptome, most of which were took place 28 DAP. In the sh2 endosperm, we see changes in the regulation of starch and sugar metabolism. Additionally, we identify ZmHXK1 as a key controller of a coexpression network containing transcription factors (namely ZmNAC130) and genes related to grain filling and C:N balance in the endosperm, such as zein-encoding genes. Through a comparison of genetic backgrounds between NILs we also identified putative modifiers of the sh2 response, including maize BHLH157 and MKKK55, that can potentially be used in sweet corn breeding or for targeted starch production. Coexpression analysis has also provided insight into the role of Class II TPS’s and the ZmSE1 gene in the maize endosperm. We propose that the previously identified starch content QTL TRPS10 (Hu S. et al., 2021) functions as a key transcriptional regulator of starch production. We also propose that ZmSE1 impacts starch metabolism through its function as a key regulator of altering GA signaling in the kernels. It is possible that se1 kernels begin aspects of germination prematurely through altering GA signaling and derepression of ZmSCL1. These results will guide future research to determine the usefulness of the putative starch pathway modifiers and validate the proposed TRPS10 and SE1 functions.

Data Availability Statement

The datasets generated for this study can be found in the NCBI Sequence Read Archive under BioProject PRJNA790930.

Author Contributions

MR and CF designed the experiments and wrote the manuscript. CF isolated and prepared RNA for sequencing, performed transcriptomic analyses, and performed statistical analysis of kernel carbohydrate data. SB extracted and quantified kernel carbohydrates. WT, KK, LH, KL, and GM contributed to interpretation, discussion, and manuscript review. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the National Institute of Food and Agriculture (SCRI 2018-51181-28419 and AFRI 2019-05691 to MR).

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Acknowledgments

We thank Chase Wasson and Allyson Jones for help with sample collection and processing.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2021.800326/full#supplementary-material

References

Aguilera-Alvarado, G., and Sánchez-Nieto, S. (2017). Plant hexokinases are multifaceted proteins. Plant Cell Physiol. 58, 1151–1160. doi: 10.1093/pcp/pcx062

PubMed Abstract | CrossRef Full Text | Google Scholar

Akihiro, T., Umezawa, T., Ueki, C., Lobna, B., Mizuno, K., Ohta, M., et al. (2006). Genome wide cDNA-AFLP analysis of genes rapidly induced by combined sucrose and ABA treatment in rice cultured cells. FEBS Lett. 580, 5947–5952. doi: 10.1016/j.febslet.2006.09.065

PubMed Abstract | CrossRef Full Text | Google Scholar

Alexa, A., and Rahnenfuhrer, J. (2020). topGO: Enrichment Analysis for Gene Ontology. R Package Version 2.42.0. Available online at: https://bioconductor.org/packages/release/bioc/html/topGO.html#:~:text=Enrichment%20Analysis%20for%20Gene%20Ontology&text=Different%20test%20statistics%20and%20different,can%20be%20implemented%20and%20applied.&text=Citation%20(from%20within%20R%2C%20enter,Enrichment%20Analysis%20for%20Gene%20Ontology

Google Scholar

Assenov, Y., Ramírez, F., Schelhorn, S., Lengauer, T., and Albrecht, M. (2008). Computing topological parameters of biological networks. Bioinformatics 24, 282–284. doi: 10.1093/bioinformatics/btm554

PubMed Abstract | CrossRef Full Text | Google Scholar

Barraza, A., Contreras-Cubas, C., Estrada-Navarrete, G., Reyes, J., Juárez-Verdayes, M., Avonce, N., et al. (2016). The class II Trehalose 6-phosphate synthase gene PvTPS9 modulates trehalose metabolism in Phaseolus vulgaris nodules. Front. Plant Sci. 7:1589. doi: 10.3389/fpls.2016.01589

PubMed Abstract | CrossRef Full Text | Google Scholar

Batushansky, A., Kirma, M., Grillich, N., Toubiana, D., Pham, P., Balbo, I., et al. (2014). Combined transcriptomics and metabolomics of Arabidopsis thaliana seedlings exposed to exogenous GABA suggest its role in plants is predominantly metabolic. Mol. Plant 7, 1065–1068. doi: 10.1093/mp/ssu017

PubMed Abstract | CrossRef Full Text | Google Scholar

Baud, S., Wuillème, S., To, A., Rochat, C., and Lepiniec, L. (2009). Role of WRINKLED1 in the transcriptional regulation of glycolytic and fatty acid biosynthetic genes in Arabidopsis. Plant J. 60, 933–947. doi: 10.1111/j.1365-313X.2009.04011.x

PubMed Abstract | CrossRef Full Text | Google Scholar

Bihmidine, S., Hunter, C., Johns, C., Koch, K., and Braun, D. (2013). Regulation of assimilate import into sink organs: update on molecular drivers of sink strength. Front. Plant Sci. 4:177. doi: 10.3389/fpls.2013.00177

PubMed Abstract | CrossRef Full Text | Google Scholar

Boehlein, S., Shaw, J., Boehlein, T., Boehlein, E., and Hannah, L. (2018). Fundamental differences in starch synthesis in the maize leaf, embryo, ovary and endosperm. Plant J. 96, 595–606. doi: 10.1111/tpj.14053

PubMed Abstract | CrossRef Full Text | Google Scholar

Bolger, A., Lohse, M., and Usadel, B. (2014). Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120. doi: 10.1093/bioinformatics/btu170

PubMed Abstract | CrossRef Full Text | Google Scholar

Braun, D., and Slewinski, T. (2009). Genetic control of carbon partitioning in grasses: roles of sucrose transporters and Tie-dyed Loci in phloem loading. Plant Physiol. 149, 71–81. doi: 10.1104/pp.108.129049

PubMed Abstract | CrossRef Full Text | Google Scholar

Chang, S., Puryear, J., and Cairney, J. (1993). A simple and efficient method for isolating RNA from pine trees. Plant Mol. Biol. Rep. 11, 113–116. doi: 10.1385/MB:19:2:201

PubMed Abstract | CrossRef Full Text | Google Scholar

Chantarangsee, M., Tanthanuch, W., Fujimura, T., Fry, S., and Ketudat Cairns, J. (2007). Molecular characterization of β-galactosidases from germinating rice (Oryza sativa). Plant Sci. 173, 118–134. doi: 10.1016/j.plantsci.2007.04.009

CrossRef Full Text | Google Scholar

Chen, L., Hou, B., Lalonde, S., Takanaga, H., Hartung, M., Qu, X., et al. (2010). Sugar transporters for intercellular exchange and nutrition of pathogens. Nature 468, 527–532. doi: 10.1038/nature09606

PubMed Abstract | CrossRef Full Text | Google Scholar

Creech, R. (1968). Carbohydrate synthesis in maize. Adv. Agron. 20, 275–322. doi: 10.1016/s0065-2113(08)60859-7

CrossRef Full Text | Google Scholar

de Sousa, S., Paniago, M., Arruda, P., and Yunes, J. (2008). Sugar levels modulate sorbitol dehydrogenase expression in maize. Plant Mol. Biol. 68, 203–213. doi: 10.1007/s11103-008-9362-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Doll, N., Depège-Fargeix, N., Rogowsky, P., and Widiez, T. (2017). Signaling in early maize kernel development. Mol. Plant 10, 375–388. doi: 10.1016/j.molp.2017.01.008

PubMed Abstract | CrossRef Full Text | Google Scholar

Eom, J., Chen, L., Sosso, D., Julius, B., Lin, I., Qu, X., et al. (2015). SWEETs, transporters for intracellular and intercellular sugar translocation. Curr. Opin. Plant Biol. 25, 53–62. doi: 10.1016/j.pbi.2015.04.005

PubMed Abstract | CrossRef Full Text | Google Scholar

Fait, A., Fromm, H., Walter, D., Galili, G., and Fernie, A. (2008). Highway or byway: the metabolic role of the GABA shunt in plants. Trends Plant Sci. 13, 14–19. doi: 10.1016/j.tplants.2007.10.005

PubMed Abstract | CrossRef Full Text | Google Scholar

Fao.org. Dimensions of Need–Staple foods: What do People Eat?. Available online at: http://www.fao.org/3/u8480e/u8480e07.htm (accessed September 10, 2021).

Google Scholar

Fao.org (1998). Maize in Human Nutrition. Available online at: http://www.fao.org/3/T0395E/T0395E00.htm#Contents (accessed February 22, 2021).

Google Scholar

Ferguson, J., Dickinson, D., and Rhodes, A. (1979). Analysis of endosperm sugars in a sweet corn inbred (Illinois 677a) which contains the sugary enhancer (se) gene and comparison of se with other corn genotypes. Plant Physiol. 63, 416–420. doi: 10.1104/pp.63.3.416

PubMed Abstract | CrossRef Full Text | Google Scholar

Finkelstein, R., Li Wang, M., Lynch, T., Rao, S., and Goodman, H. (1998). The Arabidopsis abscisic acid response locus ABI4 encodes an APETALA2 domain protein. Plant Cell 10, 1043–1054. doi: 10.1105/tpc.10.6.1043

PubMed Abstract | CrossRef Full Text | Google Scholar

Fujimoto, S., Ohta, M., Usui, A., Shinshi, H., and Ohme-Takagi, M. (2000). Arabidopsis ethylene-responsive element binding factors act as transcriptional activators or repressors of GCC box–mediated gene expression. Plant Cell 12, 393–404. doi: 10.1105/tpc.12.3.393

PubMed Abstract | CrossRef Full Text | Google Scholar

Giannakouros, T., Karagiorgos, A., and Simos, G. (1991). Expression of beta-galactosidase multiple forms during barley (Hordeum vulgare) seed germination. Separation and characterization of enzyme isoforms. Physiol. Plant. 82, 413–418.

Google Scholar

Giroux, M., Boyer, C., Feix, G., and Hannah, L. (1994). Coordinated transcriptional regulation of storage product genes in the maize endosperm. Plant Physiol. 106, 713–722. doi: 10.1104/pp.106.2.713

PubMed Abstract | CrossRef Full Text | Google Scholar

Gonzales, J., Rhodes, A., and Dickinson, D. (1976). Carbohydrate and enzymic characterization of a high sucrose sugary inbred line of sweet corn. Plant Physiol. 58, 28–32. doi: 10.1104/pp.58.1.28

PubMed Abstract | CrossRef Full Text | Google Scholar

Guan, J., and Koch, K. (2015). A time and a place for sugar in your ears. Nat. Biotechnol. 33, 827–828. doi: 10.1038/nbt.3315

PubMed Abstract | CrossRef Full Text | Google Scholar

Hannah, L., and Boehlein, S. (2017). “Starch biosynthesis in maize endosperm,” in Maize Kernel Development, ed. B. Larkins (Wallingford: CAB International), 149–160. doi: 10.1079/9781786391216.0149

CrossRef Full Text | Google Scholar

Hannah, L. C. (1997). “Starch synthesis in the maize seed,” in Cellular and Molecular Biology of Plant Seed Development, Vol. 4, eds B. A. Larkins and I. K. Vasil (Dordrecht: Springer). doi: 10.1007/978-94-015-8909-3_10

CrossRef Full Text | Google Scholar

Henry, C., Bledsoe, S., Siekman, A., Kollman, A., Waters, B., Feil, R., et al. (2014). The trehalose pathway in maize: conservation and gene regulation in response to the diurnal cycle and extended darkness. J. Exp. Bot. 65, 5959–5973. doi: 10.1093/jxb/eru335

PubMed Abstract | CrossRef Full Text | Google Scholar

Hu, S., Wang, M., Zhang, X., Chen, W., Song, X., Fu, X., et al. (2021). Genetic basis of kernel starch content decoded in a maize multi-parent population. Plant Biotechnol. J. 19, 2192–2205. doi: 10.1111/pbi.13645

PubMed Abstract | CrossRef Full Text | Google Scholar

Hu, Y., Colantonio, V., Müller, B., Leach, K., Nanni, A., Finegan, C., et al. (2021). Genome assembly and population genomic analysis provide insights into the evolution of modern sweet corn. Nat. Commun. 12:1227. doi: 10.1038/s41467-021-21380-4

PubMed Abstract | CrossRef Full Text | Google Scholar

Huang, H., Xie, S., Xiao, Q., Wei, B., Zheng, L., Wang, Y., et al. (2016). Sucrose and ABA regulate starch biosynthesis in maize through a novel transcription factor, ZmEREB156. Sci. Rep. 6:27590. doi: 10.1038/srep27590

PubMed Abstract | CrossRef Full Text | Google Scholar

Hufford, M., Seetharam, A., Woodhouse, M., Chougule, K., Ou, S., Liu, J., et al. (2021). De novo assembly, annotation, and comparative analysis of 26 diverse maize genomes. Science 373, 655–662. doi: 10.1126/science.abg5289

PubMed Abstract | CrossRef Full Text | Google Scholar

Hwang, S., Nishi, A., Satoh, H., and Okita, T. (2010). Rice endosperm-specific plastidial α-glucan phosphorylase is important for synthesis of short-chain malto-oligosaccharides. Arch. Biochem. Biophys. 495, 82–92. doi: 10.1016/j.abb.2009.12.023

PubMed Abstract | CrossRef Full Text | Google Scholar

Julius, B., Leach, K., Tran, T., Mertz, R., and Braun, D. (2017). Sugar transporters in plants: new insights and discoveries. Plant Cell Physiol. 58, 1442–1460. doi: 10.1093/pcp/pcx090

PubMed Abstract | CrossRef Full Text | Google Scholar

Kang, F., and Rawsthorne, S. (1996). Metabolism of glucose-6-phosphate and utilization of multiple metabolites for fatty acid synthesis by plastids from developing oilseed rape embryos. Planta 199, 321–327.

Google Scholar

Koch, K. E. (1996). Carbohydrate-modulated gene expression in plants. Annu. Rev. Plant Biol. 47, 509–540. doi: 10.1146/annurev.arplant.47.1.509

PubMed Abstract | CrossRef Full Text | Google Scholar

Koch, K. E. (1997). “Molecular crosstalk and the regulation of C-and N-responsive genes,” in A Molecular Approach to Primary Metabolism in Higher Plants, ed. C. Foyer (London: Taylor & Francis), 105–124.

Google Scholar

La Bonte, D., and Juvik, J. (1990). Characterization of sugary-1 (su-1) sugary enhancer (se) kernels in segregating sweet corn populations. J. Am. Soc. Hortic. Sci. 115, 153–157.

Google Scholar

Langfelder, P., and Horvath, S. (2008). WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9:559. doi: 10.1186/1471-2105-9-559

PubMed Abstract | CrossRef Full Text | Google Scholar

Li, C., and Song, R. (2020). The regulation of zein biosynthesis in maize endosperm. Theor. Appl. Genet. 133, 1443–1453. doi: 10.1007/s00122-019-03520-z

PubMed Abstract | CrossRef Full Text | Google Scholar

Love, M., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15:550. doi: 10.1186/s13059-014-0550-8

PubMed Abstract | CrossRef Full Text | Google Scholar

Macdonald, F., and Preiss, J. (1983). Solubilization of the starch-granule-bound starch synthase of normal maize kernels. Plant Physiol. 73, 175–178. doi: 10.1104/pp.73.1.175

PubMed Abstract | CrossRef Full Text | Google Scholar

MacLeod, A., and Millar, A. (1962). Effects of gibberellic acid on barley endosperm. J. Instit. Brew. 68, 322–332. doi: 10.1002/j.2050-0416.1962.tb01873.x

CrossRef Full Text | Google Scholar

MaizeGDB.org (2021). MaizeGDB Gene Record Page: 292668. Available online at: https://www.maizegdb.org/gene_center/gene/292668 (accessed September 10, 2021).

Google Scholar

Moody, L., Saidi, Y., Gibbs, D., Choudhary, A., Holloway, D., Vesty, E., et al. (2016). An ancient and conserved function for Armadillo-related proteins in the control of spore and seed germination by abscisic acid. New Phytol. 211, 940–951. doi: 10.1111/nph.13938

PubMed Abstract | CrossRef Full Text | Google Scholar

Mu, C., Chen, N., Li, X., Jia, P., Wang, Z., and Liu, H. (2010). F-box protein arabidillo-1 promotes lateral root development by depressing the functioning of GA3 in Arabidopsis. J. Plant Biol. 53, 374–380. doi: 10.1007/s12374-010-9125-8

CrossRef Full Text | Google Scholar

Oszvald, M., Primavesi, L., Griffiths, C., Cohn, J., Basu, S., Nuccio, M., et al. (2018). Trehalose 6-phosphate regulates photosynthesis and assimilate partitioning in reproductive tissue. Plant Physiol. 176, 2623–2638. doi: 10.1104/pp.17.01673

PubMed Abstract | CrossRef Full Text | Google Scholar

Patro, R., Duggal, G., Love, M., Irizarry, R., and Kingsford, C. (2017). Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14, 417–419. doi: 10.1038/nmeth.4197

PubMed Abstract | CrossRef Full Text | Google Scholar

Plaxton, W. (1996). The organization and regulation of plant glycolysis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 47, 185–214. doi: 10.1146/annurev.arplant.47.1.185

PubMed Abstract | CrossRef Full Text | Google Scholar

Pouvreau, B., Baud, S., Vernoud, V., Morin, V., Py, C., Gendrot, G., et al. (2011). Duplicate maize wrinkled1 transcription factors activate target genes involved in seed oil biosynthesis. Plant Physiol. 156, 674–686. doi: 10.1104/pp.111.173641

PubMed Abstract | CrossRef Full Text | Google Scholar

Ratcliff, S., Wilson, D., Knott, E., and Mohan, S. (1993). Free fatty acids in shrunken-2 sweet corn seed. Crop Sci. 33, 871–873.

Google Scholar

Rolletschek, H., Koch, K., Wobus, U., and Borisjuk, L. (2005). Positional cues for the starch/lipid balance in maize kernels and resource partitioning to the embryo. Plant J. 42, 69–83. doi: 10.1111/j.1365-313X.2005.02352.x

PubMed Abstract | CrossRef Full Text | Google Scholar

Rook, F., Corke, F., Card, R., Munz, G., Smith, C., and Bevan, M. (2001). Impaired sucrose-induction mutants reveal the modulation of sugar-induced starch biosynthetic gene expression by abscisic acid signalling. Plant J. 26, 421–433. doi: 10.1046/j.1365-313x.2001.2641043.x

PubMed Abstract | CrossRef Full Text | Google Scholar

Sanclemente, M. A., Ma, F., Liu, P., Della Porta, A., Singh, J., Wu, S., et al. (2021). Sugar modulation of anaerobic-response networks in maize root tips. Plant Physiol. 185, 295–317. doi: 10.1093/plphys/kiaa029

PubMed Abstract | CrossRef Full Text | Google Scholar

Sandrine, A., Zhao, H., Qin, Y., Sun, Q., Gong, D., Pan, Z., et al. (2020). 22KD Zein content coordinates transcriptional activity during starch synthesis in maize endosperm. Agronomy 10:624.

Google Scholar

Satoh, H., Shibahara, K., Tokunaga, T., Nishi, A., Tasaki, M., Hwang, S., et al. (2008). Mutation of the plastidial α-glucan phosphorylase gene in rice affects the synthesis and structure of starch in the endosperm. Plant Cell 20, 1833–1849. doi: 10.1105/tpc.107.054007

PubMed Abstract | CrossRef Full Text | Google Scholar

Shen, B., Allen, W., Zheng, P., Li, C., Glassman, K., Ranch, J., et al. (2010). Expression of ZmLEC1 and ZmWRI1 increases seed oil production in maize. Plant Physiol. 153, 980–987. doi: 10.1104/pp.110.157537

PubMed Abstract | CrossRef Full Text | Google Scholar

Shuler, S., Boehlein, S., Hannah, L., and Tracy, W. (2017). Endosperm carbohydrates and debranching enzyme activity in five native sugary1 alleles in maize. Crop Sci. 57, 3068–3074.

Google Scholar

Smith, A., Denyer, K., and Martin, C. (1997). The synthesis of the starch granule. Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 67–87.

Google Scholar

Subasinghe, R. (2013). Role and Regulation of Starch Phosphorylase and Starch Synthase IV in Starch Biosynthesis in Maize Endosperm Amyloplasts. Ph.D. thesis. Guelph, ON: The University of Guelph.

Google Scholar

Tsai, C., Salamini, F., and Nelson, O. (1970). Enzymes of carbohydrate metabolism in the developing endosperm of maize. Plant Physiol. 46, 299–306. doi: 10.1104/pp.46.2.299

PubMed Abstract | CrossRef Full Text | Google Scholar

Vandesteene, L., Ramon, M., Le Roy, K., Van Dijck, P., and Rolland, F. (2010). A single active trehalose-6-P synthase (TPS) and a family of putative regulatory TPS-like proteins in Arabidopsis. Mol. Plant 3, 406–419. doi: 10.1093/mp/ssp114

PubMed Abstract | CrossRef Full Text | Google Scholar

Wang, F., and Deng, X. (2011). Plant ubiquitin-proteasome pathway and its role in gibberellin signaling. Cell Res. 21, 1286–1294. doi: 10.1038/cr.2011.118

PubMed Abstract | CrossRef Full Text | Google Scholar

Weigel, D. (1995). The APETALA2 domain is related to a novel type of DNA binding domain. Plant Cell 7, 388–389. doi: 10.1105/tpc.7.4.388

PubMed Abstract | CrossRef Full Text | Google Scholar

Wobus, U., and Weber, H. (1999). Sugars as signal molecules in plant seed development. Biol. Chem. 380, 937–944. doi: 10.1515/BC.1999.116

PubMed Abstract | CrossRef Full Text | Google Scholar

Xie, C., Mao, X., Huang, J., Ding, Y., Wu, J., Dong, S., et al. (2011). KOBAS 2.0: a web server for annotation and identification of enriched pathways and diseases. Nucleic Acids Res. 39(Suppl. 2), W316–W322. doi: 10.1093/nar/gkr483

PubMed Abstract | CrossRef Full Text | Google Scholar

Yadav, U., Ivakov, A., Feil, R., Duan, G., Walther, D., Giavalisco, P., et al. (2014). The sucrose–trehalose 6-phosphate (Tre6P) nexus: specificity and mechanisms of sucrose signalling by Tre6P. J. Exp. Bot. 65, 1051–1068. doi: 10.1093/jxb/ert457

PubMed Abstract | CrossRef Full Text | Google Scholar

Yu, Y., Mu, H., Wasserman, B., and Carman, G. (2001). Identification of the maize amyloplast stromal 112-kD protein as a plastidic starch phosphorylase. Plant Physiol. 125, 351–359. doi: 10.1104/pp.125.1.351

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, X., Mogel, K., Lor, V., Hirsch, C., De Vries, B., Kaeppler, H., et al. (2019). Maize sugary enhancer1 (se1) is a gene affecting endosperm starch metabolism. Proc. Natl. Acad. Sci. U.S.A. 116, 20776–20785. doi: 10.1073/pnas.1902747116

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, Z., Dong, J., Ji, C., Wu, Y., and Messing, J. (2019). NAC-type transcription factors regulate accumulation of starch and protein in maize seeds. Proc. Natl. Acad. Sci. U.S.A. 116, 11223–11228.

Google Scholar