The low-temperature sensitive maize (SM) B283-1 and low-temperature resistance maize (RM) 04Qun0522-1-1maize inbred lines used in this study were bred in our laboratory. They were grown at the experimental station (36°90′N, 117°90′E) of Shandong Agricultural University, Shandong, China. Seeds were sown on 12 June 2019. The plant density was 67,500 plants/ha. Seeds used in this study were harvested 50 days after pollination.

Seed germination was evaluated according to a previous study with some modifications (Wen et al., 2018). The bottom of a sprouting bed consisted of 4 cm height silica sand (diameter of 0.05–0.8 mm) with 60% saturation moisture content in a germination box. Randomly selected 30 maize seeds were sown on the surface of the sprouting bed, and then they were covered with 2 cm height silica sand with 60% saturation moisture content. Subsequently, the germination boxes were placed in different germination conditions for various treatments. The germination boxes were placed in a growth chamber at 13°C for 4 days to detect the percentage of seeds showing radicle protrusion. The germination boxes were placed in a growth chamber at 25°C for 7 days or at 13°C for 7 days to measure germination percentage. A seed was considered as germinating seed when the radicle was similar to seed length and the germ was similar with half of the seed length. Moreover, some germination boxes were placed in a growth chamber at 25°C for 4 days (NT) as control, and some germination boxes were placed in a growth chamber at 13°C for 4 days followed by 25°C for 2 days (LNT) as low-temperature treatment. All tissues of the two inbred lines under NT and LNT were used for later experiments. Each treatment included three replicates.

TAC was measured by a 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) assay kit (Comin, Suzhou, China) following the manufacturer’s protocols. ABTS, a colorless substance, can be oxidized to a stable blue–green cationic radical form ABTS⁺ (Scalzo et al., 2005). The absorption peak of ABTS⁺ is at 734 nm. When the tested substance is added to the ABTS⁺ solution, the antioxidant components can react with ABTS⁺ and fade the reaction system. The change of absorbance at 734 nm was detected. The antioxidant capacity quantified with Trolox was used as a control. The results were expressed in μmol Ttoloox/g fresh weight (FW). Three biological replicates were used for each treatment.

The content of thiobarbituric acid (TBA) reactive substances (TBARS) can be used to assess lipid peroxidation (Checchio et al., 2021). In this study, lipid peroxidation was detected using a commercial kit (Comin, Suzhou, China). The results were expressed in nmol/g FW. Three biological replicates were used for each treatment.

The determination principle of sucrose content is that alkali is used to heat the sample together to destroy the reducing sugar (Handel, 1968). Then, sucrose was hydrolyzed to glucose and fructose under acidic conditions, and fructose reacted with resorcinol to form a colored substance with a characteristic absorption peak at 480 nm. In this study, sucrose content was detected using a sucrose content determination kit (Comin, Suzhou, China) according to the manufacturer’s protocols. The results were expressed in mg/g FW. Three biological replicates were used for each treatment.

SOD activity was determined by the photoinhibition of nitro blue tetrazole (NBT) according to a previous study (Giannopolitis and Ries, 1977). In this study, SOD activity was measured using a SOD determination kit (Comin, Suzhou, China) according to the manufacturer’s protocols. The results were expressed in U/g FW. Three biological replicates were used for each treatment.

POD activity was determined by the method of guaiacol oxidation (Zhang et al., 2015a). In this study, POD activity was detected using a POD determination kit (Comin, Suzhou, China) according to the manufacturer’s protocols. The results were expressed in U/kg FW. Three biological replicates were used for each treatment.

Chlorophyll contents of the two inbred lines germinated at 25°C for 6 days or germinated at 13°C for 4 days followed by 25°C for 4 days were measured by Soil and Plant Analyzer Development (SPAD) 502 chlorophyll meter (Konica Minolta Inc., Japan). Three biological replicates were used for each treatment.

All tissues of the two inbred lines under NT and LNT conditions were used for RNA extraction. After quick freezing with liquid nitrogen, samples from each replication were pooled and stored at −80°C. Three biological replicates were used for each treatment. Total RNA was extracted using a biospin plant total RNA extraction kit (Bioflux, Beijing, China) according to the manufacturer’s protocols. Libraries were generated using NEBNext^® UltraTM RNA Library Prep kit for Illumina^® (NEB, United States) following the manufacturer’s protocols. Illumina sequencing was performed as described previously (Yu et al., 2021). Hisat2 (v2.0.5) was used to build an index of the reference genome (B73 v4),¹ and then paired-end clean reads were aligned to the reference genome (Yu et al., 2021). Differential expression analyses of two groups (three biological replicates per treatment) were performed using the DESeq2 R package (v1.16.1). The resulting values of p were adjusted (padj) using Benjamini and Hochberg’s approach for controlling the false discovery rate. Differentially expressed genes (DEGs) with padj <0.05 and |log₂Fold change| ≥ 1were used for further analyses.

To understand the main biological functions of DEGs in maize, we carried out an enrichment analysis of DEGs. ClusterProfiler R package (v3.4.4) was used to achieve Gene Ontology (GO) enrichment analysis of DEGs (Yu et al., 2012). GO terms with corrected p < 0.05 were considered as significantly enriched GO terms. To understand the primary metabolic pathway of DEGs in maize, we carried out the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of DEGs. ClusterProfiler R package (v3.4.4) was used to test the statistical enrichment of DEGs in the KEGG pathway (Yu et al., 2012). Similarly, the KEGG pathways with corrected p < 0.05 were assigned as significantly enriched pathways.

To validate the DEGs identified from transcriptome analysis, six genes were randomly selected for a qRT-PCR assay: Zm00001d002611 (SOD 13), Zm00001d025103 (Amine oxidase1), Zm00001d027557 (Glutathione transferase 31), Zm00001d027422 (PsbP-like protein 1), Zm00001d021906 (Light harvesting complex A2), and Zm00001d046170 (phosphoenolpyruvate carboxylase1). Primers were designed using the primer-premier software (v6.0; Supplementary Table S6). The maize Actin gene (Zm00001d010159) was used as an internal control (Zhang et al., 2015b). An ABI StepOne Plus Real-time PCR System (Applied Biosystems, CA, United States) was used to perform qRT-PCR according to the instructions of SYBR^® Green Real-time PCR Master Mix (Takara, Dalian, China). The 2^−ΔΔCT method was used to calculate the relative expression of genes (Livak and Schmittgen, 2001).

Multiple comparisons were performed using Duncan’s test at the 0.05 significance level. All the tests were conducted using SPSS Version 21.0 for Windows (SPSS, Chicago, IL, United States).

To investigate the effects of maize seed germination under low-temperature stress on subsequent seedling growth under normal temperature, we first measured some traits of two maize inbred lines with different low-temperature resistance. When maize seeds germinated at 13°C for 4 days, the percentage of seeds showing radicle protrusion in the RM line was higher compared to the SM line (Figure 1A). In seed testing, maize seeds germinating at 25°C for 7 days are usually used for evaluating seed germination percentage. The germination percentage of the SM line decreased about 60% at 13°C for 7 days, while there was no significant difference in germination percentage between seeds germinated at 13°C for 7 days and seeds germinated at 25°C for 7 days in the RM line (Figures 1B,C). Maize seeds germinating at 25°C for 4 days (NT) are usually used for evaluating seed germination energy in seed testing. In this study, the two inbred lines had similar germination energy and seedling length (Figure 1D). To investigate the effects of maize seed germination under low-temperature stress impacting seedling growth under normal temperature, we set a similar accumulated temperature between NT and low-temperature stress followed by normal temperature (LNT) treatment. Given the same accumulated temperature, maize seeds first germinated at 13°C for 4 days followed by 25°C for 2 days under LNT treatment. At this time, the RM line had notably higher seedlings than the SM line (Figure 1E). Therefore, these results suggested that the SM line was more vulnerable to low-temperature stress. Subsequently, we detected the changes in lipid peroxidation, expressed as malondialdehyde (MDA) content, TAC, and sucrose content. Although MDA contents of the two lines under LNT were about twice as high as those under NT, the RM line under LNT had a lower MDA content (115 nmol/g FW) than the SM line under LNT (195 nmol/g FW; Figure 1F). TAC decreased in the SM line but increased in the RM line under LNT treatment compared with NT treatment (Figure 1G). TAC under LNT (2.6 μmol Trolox/g FW) decreased to about half of that under NT (5.0 μmol Trolox/g FW) in the SM line. Moreover, the RM line had notably higher sucrose content under LNT than under NT (Figure 1H). Taken together, seeds germinated under low-temperature stress increased TAC and sucrose content in the RM line, which might be related to low-temperature resistance.

To explore genes and metabolic pathways that control maize seed germination in response to low-temperature stress, we selected samples of RM and SM under NT and LNT for transcriptome analysis. After removing adapters and sequences with low-quality regions, there remained approximately 40–50 million clean reads (Supplementary Table S1). Then, about 35–46 million clean reads were mapped to the maize genome. These clean reads included 83–88% uniquely mapped reads and 2.5–3.0% multiple mapped reads. DESeq2 R package was used to identify DEGs using padj <0.05 and |log₂Fold change| ≥ 1 as the cutoff. The results displayed that 3,186 genes were significantly up-regulated and 4,281 genes were significantly downregulated in the SM line under LNT compared with those under NT (SM_LNTvsNT). Moreover, 2,797 genes were significantly up-regulated and 3,918 genes were significantly downregulated in the RM line under LNT compared with those under NT (RM_LNTvsNT; Figure 2A). Venn diagram showed that common up-regulated genes (54) were fewer than common down-regulated genes (608), most DEGs were specific in SM_LNTvsNT and RM_LNTvsNT (Figures 2B,C).

GO enrichment analysis displayed that there was no significantly enriched GO term for common up-regulated genes, which might be due to the low number of DEGs. For these SM-specific up-regulated DEGs, there were only two significantly enriched GO terms, i.e., myosin complex (GO: 0016459, p = 4.82 × 10⁻⁴) in the cellular component group and ADP binding (GO: 0043531, p = 2.32 × 10⁻⁵) in the molecular function group (Figure 3A). For these RM-specific up-regulated DEGs, there were nine significantly enriched GO terms. Among them, the most significantly enriched GO terms were superoxide metabolic (GO: 0006801, p = 8.55 × 10⁻⁶) in the biological process group, vitamin binding (GO: 0019842, p = 5.35 × 10⁻⁷) in the molecular function group, which might play essential roles in RM resistant to low-temperature stress (Figure 3B).

Compared with the up-regulated GO terms, there were more down-regulated GO terms. For these common down-regulated DEGs, the most significantly enriched GO terms were photosynthesis (GO: 0015979, p = 2.16 × 10⁻⁴⁴) in the biological process group, thylakoid (GO: 0009579, p = 5.33 × 10⁻⁵⁴) in the cellular component group, iron–sulfur cluster binding (GO: 0051536, p = 4.38 × 10⁻⁸) in the molecular function group (Figure 3C). Moreover, many photosynthesis-related GO terms were also enriched. The results indicated that maize seed germination under low-temperature stress caused the down-regulation of photosynthesis-related GO terms in the two inbred lines. Translation (GO: 0006412, p = 6.79 × 10⁻¹⁶) in the biological process group, ribosome (GO: 0005840, p = 7.64 × 10⁻¹⁸) in the cellular component group, structural constituent of ribosome (GO: 0003735, p = 6.47 × 10⁻¹⁸) in the molecular function group, represented the most markedly enriched GO terms in SM-specific down-regulated DEGs, suggesting that ribosomes of SM line might be damaged by low-temperature stress (Figure 3D). For RM-specific down-regulated DEGs, the most prominently enriched GO terms were response to oxidative stress (GO: 0006979, p = 2.41 × 10⁻¹⁷) in the biological process group, apoplast (GO: 0048046, p = 4.58 × 10⁻⁷) in the cellular component group, peroxidase activity (GO: 0004601, p = 1.14 × 10⁻¹⁷) in the molecular function group (Figure 3E). The results indicated that peroxidase activity might not be involved in the enhanced antioxidant capacity of the RM line in response to low-temperature stress.

KEGG enrichment analysis showed photosynthesis-antenna proteins, photosynthesis, phenylpropanoid biosynthesis, flavonoid biosynthesis, glutathione metabolism, porphyrin and chlorophyll metabolism, stilbenoid, diarylheptanoid, and gingerol biosynthesis were markedly enriched KEGG pathways for common DEGs (Figure 4). Therefore, photosynthesis and antioxidant metabolism pathways might play essential roles in response to low-temperature stress at the germination stage.

Only the ribosome pathway was significantly enriched for SM-specific DEGs, which was also enriched in GO enrichment analysis. In the ribosome pathway, there were 78 DEGs in SM_LNTvsNT (Supplementary Table S2). Of the 78 DEGs, most genes were down-regulated except for three genes (Zm00001d022111, Zm00001d022197, and Zm00001d002462), indicating that low-temperature stress might have a significant influence on the ribosomal pathway of low-temperature sensitive inbred lines. For RM-specific DEGs, the significantly enriched KEGG pathways were related to phenylpropanoid biosynthesis, phenylalanine metabolism, ubiquinone, and other terpenoid-quinone biosynthesis, tropane, piperidine and pyridine alkaloid biosynthesis (Supplementary Table S3). Only one gene was down-regulated among five DEGs involved in tropane, piperidine and pyridine alkaloid biosynthesis.

The DEGs enriched in photosynthesis-related pathways were mainly located in the chloroplast and annotated to function as antenna proteins, photosystems I and II components, porphyrin, and chlorophyll metabolism-related proteins (Figure 5; Supplementary Table S4). Although DEGs involved in photosynthetic-antenna proteins were all down-regulated in both SM and RM lines, the degree of down-regulation of the DEGs was lower in the RM line than in the SM line. Moreover, most of the DEGs involved in photosynthesis and porphyrin and chlorophyll metabolism pathway were also down-regulated in both SM and RM lines, and the RM line also had a lower degree of down-regulation of the DEGs than the SM line (Figure 5; Supplementary Table S4). Taken together, the photosynthetic system of the SM line was even more damaged when seed germinated under low-temperature stress.

All the DEGs showed similar expression patterns in the qRT-PCR assays as their changes of relative expression level identified by RNA-seq, suggesting the transcriptome data were credible (Figure 6). By extending the seedling growth at 25°C for 2 days, we observed that maize seed germination under low-temperature markedly inhibited subsequent seedling growth under normal temperature, especially in the SM line (Figures 7A,B). Subsequently, we further detected the changes in SPAD value, SOD, and POD activities. Both SM and RM lines displayed a significant decrease of SPAD value under LNT, and the level of reduction in the SM line was larger than that in the RM line, which was consistent with the changes of the photosynthetic system from transcriptome analysis (Figure 7C). The SM line had lower SOD activity and higher POD activity under LNT than NT treatment, while the RM line showed the opposite trend of SOD and POD activities, which were consistent with the results of GO enrichment analysis (Figures 7D,E). Moreover, the changes in SOD activities in both SM and RM lines were similar to TAC trends (Figures 1F, 7D). Therefore, SOD activity might play a key role in TAC in seed germination under low-temperature stress.