Introduction

Wheat (Triticum aestivum L.) is an important crop consumed as bread and noodles all over the world. Pre-harvest sprouting, seed germination before harvesting, causes severe damage to crop production because of starch degradation. Therefore, dormancy-related genes have been studied for a long time to breed higher dormant lines. Grain color is one factor that is closely linked to seed dormancy not only of wheat but also other plant species: darker seeds are more dormant than those lighter colored seeds (Debeaujon et al. 2000; Warner et al. 2000). In wheat, an MYB-type transcription factor, Tamyb10, is reportedly a strong candidate as a regulator of grain color, called R-1 gene (Himi et al. 2011; Himi and Noda 2005). Because wheat is hexaploid, the R-1 genes are triplicated, located on chromosomes 3A, 3B, and 3D, and respectively designated as R-A1, R-B1, and R-D1. White-grained lines have all recessive alleles (R-A1a/R-B1a/R-D1a), whereas red-grained lines have at least one of the dominant alleles (R-A1b/R-B1b/R-D1b). Nevertheless, the relation between grain pigmentation and dormancy remains unknown.

A lipid transfer protein (LTP) was first isolated from an active fraction that stimulates the exchange of phospholipids by gel filtration from potatoes (Kader 1975). Later, it was revealed that plant LTPs consist of a large protein family and have various functions (Carvalho and Gomes 2007). They are called non-specific lipid transfer protein (nsLTPs), which have a characteristic conserved motif, eight-cystein motif (8CM) generally formed by C-Xn-C-Xn-CC-Xn-CXC-Xn-C-Xn-C (Salminen et al. 2016). Kouidri et al. (2018) reported 461 putative nsLTPs (TaLTPs) identified in wheat and classified into five types: Types 1, 2, C, D, and G. Nevertheless, the functions of these TaLTPs are poorly understood.

Some reports have described that members of nsLTPs are involved in seed maturation and germination. Lee and Suh (2018) reported that Arabidopsis LTPG15, a member of glycosylphosphatidylinositol-anchored LTP, is involved in suberin monomer export in seed coats and consequently seed coat permeability. OsLTPL36 of rice, predominantly expressed in developing seed, might be involved in seed development and germination (Wang et al. 2015). Additionally, rice qLTG3-1, which was identified as QTL for low-temperature germinability on chromosome 3 and which has the conserved motif, 8CM, was reported as involved in seed germination (Fujino et al. 2008).

For this study, we generated transgenic red-grained wheat lines into which Tamyb10-D1 had been introduced into white-grained line, Fielder using its native promoter. Transgenic lines showed higher dormancy than wild type. Gene expressions of embryos were compared between them. A member of nsLTPs, TaLTP2.128, was found in differentially expressed genes. Expressions of TaLPT2.128 were strongly enhanced in white-grained lines of different genetic backgrounds during imbibition and also in the reduced seed dormancy mutant. We infer that TaLTP2.128 might be a novel candidate for the regulation of dormancy release.

Materials and methods

Production of transgenic wheat lines

For the generation of transgenic wheat, Tamyb10-D1 was selected from three orthologous Tamyb10 genes located in each chromosome of 3A, 3B and 3D. This was done because the longest upstream region sequence data (GenBank acc. no. AB191460) was available from the database at the time of cloning. PCR amplification was performed with Prime STAR*GXL DNA Polymerase (Takara Bio Inc., Japan) using primer set 1 (Table S1) from genomic DNA of Zenkojikomugi (red-grained); a 4119-bp DNA fragment was obtained. The resultant amplicon was first cloned into the vector pCR-Blunt II-TOPO (Thermo Fisher Scientific Inc., USA) and sequenced.

A fragment of the nos terminator with Sac I and EcoR I restriction enzyme sites at the ends was prepared because we were not sure whether the terminator contained the full length in the short 3′ untranslated region of the 131 bp of Tamyb10-D1. To the 4,119-bp length of Tamyb10-D1 genomic fragment HindIII and BamHI recognition sequences were added primer set 2 (Fig. S1 and Table S1). The nos terminator fragment and the genomic Tamyb10-D1 fragment were fused into pZH2B (Kuroda et al. 2010), a binary vector derived from pPZP202 (Hajdukiewicz et al. 1994) that contained hygromycin phosphotransferase as a selectable marker.

Agrobacterium-mediated transformation of wheat (Triticum aestivum L.) Fielder was performed using immature embryos as described in an earlier report (Abe et al. 2019; Ishida et al. 2015). The transformation efficiency was calculated as the percentage of wheat embryos that regenerated transformed seedlings relative to the number of embryos infected. Rooted regenerated plants were transplanted to soil in a growth chamber 20/15 °C (day/night) under a 12 h light/ 12 h dark photoperiod.

Molecular analysis of transgenic lines

Genomic DNAs of the regenerated plants were isolated from the seedling tips using a DNeasy Plant Kit (Qiagen Inc.) according to the manufacturer’s instructions. To confirm the presence of introduced genes of hygromycin phosphotransferase (htp) and Tamyb10-D1, primer sets 3 and 4 were used, respectively (Fig. S1 and Table S1).

To detect the junction between T-DNA and the host plant, inverse PCR was performed. After 10 μg of DNA was digested with 10 U of Alw44I (Toyobo Co. Ltd.) for 16 h at 37 °C or 10 U of TaqI (Toyobo) for 16 h at 65 °C, the digested DNA was purified using an UltraClean GelSpin DNA Extraction Kit (MO BIO). It was then ligated with 1000 U of T4 ligase (Nippon Gene Co. Ltd.) for 23 h at 16 °C. Self-ligated DNA was cleaned using the UltraClean GelSpin DNA Extraction Kit and was used as a template for inverse PCR. Genomic DNA (200 ng) of T3 lines was used as a template in a 10 µl reaction solution (1 × GC buffer I (Takara), 0.2 mM dNTP mix (Takara), 0.3 µM of primer set 5, and 0.5 U Ex Taq (Takara). PCR conditions were the following: 2 min denaturation at 94 °C, then 30 cycles of 30 s at 94 °C, 30 s at 60 °C, and 2 min at 72 °C. Nested PCR was subsequently performed with primer set 6, using the same conditions as those for the first PCR. Amplified fragments were cloned into a pGEM-T Easy vector (Promega Corp.). The DNA sequence was determined using a sequencer (ABI 3100; Applied Biosystems) and was analyzed with GENETYX ver.13.0.3 (Genetyx Corp.). The T-DNA insertions were verified with primers based on the vector and the adjacent sequences (primer sets 7–9).

Total RNA was extracted from whole grains of 5 DAP and leaves of shoots of T2 plants and their original line, Fielder, and red-grained Zenkojikomugi and white-grained Tamaizumi as controls using TRIzol Reagent (Invitrogen Corp.). After 10 µg RNA for each sample were treated with DNase I (Roche Diagnostics GmbH, Germany), random hexamers were used as primers for reverse transcription using the Transcriptor First Strand cDNA Synthesis Kit (Roche Diagnostics, Germany). For each reverse transcription reaction, 5 µg of RNA treated with DNase I was used. Then the remaining 5 µg of RNA was used as a negative control using distilled water instead of reverse transcriptase. Expression of the transfected Tamyb10-D1 gene and flavonoid biosynthetic gene of CHS, CHI, F3H, DFR were examined using primer sets 10–13 according to Himi et al. (2005) and primer set 15. A set of primer for actin (primer set 14) was used as an internal control. These are listed in Table S1.

Growth conditions and germination test for transgenic lines and fielder

Each seed was sown in a 5 L pot of soil. Ten plants each of T2 generation and Fielder were grown in a transgenic plant greenhouse at 14–29 °C (average temperature about 21 °C). The daylength was the natural daylength of Sapporo, Japan, located at 43°N, during March–July 2016. Each pot was given approximately the same amount of water daily. The position of each was rotated weekly to ensure that the climatic conditions of the 30 pots in the greenhouse were as equal as possible.

Spikes were tagged at pollination day and were harvested at 50 and 60 days after pollination (DAP) of Fielder and G8 and G12 lines of T2 generation derived from homozygous T1 plants. T3 grains of primary and secondary florets of the central spikes were hand-threshed. After-ripened (AR) grains were obtained by storing 60 DAP grains at room temperature for 55 days. Two 8.5 cm diameter filter papers (No.2 qualitative filter paper; Advantec) were placed in a 9 cm diameter petri dish. Then 25 seeds were placed in each dish; 6 ml of distilled deionized water was poured. The dishes were placed inside a plastic box with a lid. Germination tests were conducted with the fresh grains and AR grains at 15 and 20 °C in the dark for 10–14 days. Each replication petri dish was replaced in turn from top to bottom daily. The position in the box was rotated daily. The box was rotated in the incubator daily. Results are presented as the means of four replicates for 50 DAP and of five replicates for 60 DAP and AR. The germination Index (GI) for estimating germinability was calculated as GI = (14n1 + 13 × n2 + 12 × n3 + 11 × n4 + 10 × n5 + 9 × n6 + 8 × n7 + 7 × n8 + 6 × n9 + 5 × n10 + 4 × n11 + 3 × n12 + 2 × n13 + 1 × n14) × 100/(14 (days) × 25 (total number of grains)), where n1, n2, n3 … n14 respectively represent the numbers of grains that germinated on the first, second, third and subsequent days until the 14th day.

Other plant materials, growth conditions, and germination tests

Other wheat lines used for this study are portrayed in Table 4: EMS-AUS is a white-grained mutant derived from EMS treatment with red-grained AUS1490 (Himi et al. 2011; Mares et al. 2005). CS-W1 and -W2 are also white-grained mutants generated by sodium azide treatment with red-grained Chinese Spring (Warner et al. 2000). An F2 population was generated by crossing between Chinese Spring and Novosibirskaya 67, a white-grained line. RSD32 is a non-dormant mutant derived from sodium azide treatment with Norin 61 (Rikiishi et al. 2021); both are red-grained lines. These lines were grown in a field of the Institute of Plant Science and Resources, Okayama University, under plastic roof to avoid rain.

Spikes were tagged at pollination day and were harvested at 30 DAP of Chinese Spring, CS-W1, -W2 and F2 individuals, 40 DAP of Norin 61 and RSD32, and 20, 30, and 40 DAP of AUS1490 and EMS-AUS. Grains of primary and secondary florets of the central spikes were hand-threshed. For germination tests, fresh grains were used immediately according to our earlier report (Himi et al. 2002).

Determination of grain color

Grain colors of transgenic lines, control cultivars, and individuals of the F2 population were determined by incubating grains in 5% sodium hydroxide (NaOH) solution according to an earlier report (Chmelar and Mostovoj 1938). The incubation time was 1 h, longer than that used for that earlier study. The NaOH treatment emphasized the reddish color of the grains. The white grains became a distinct cream color.

Tamyb10-D1 genotyping

Tamyb10-D1 genotyping was performed with each of the F2 population according to our earlier report (Himi et al. 2011, 2015).

RNA sequencing and analysis

Fielder and T3 generation of G8 were grown in a biosafety greenhouse under natural lighting conditions with 15–20 °C air temperature. Grains of 30 DAP of Fielder and G8 were used for RNA sequencing. Collected grains from a spike were divided into two groups. Whole grains of the first group were imbibed with distilled water. Grains of the second group were cut into embryo-half grains and were imbibed with 10 μM ABA solution. Grains of both groups were imbibed at 20 °C for 24 h in dark conditions. Three biological replicates were prepared with three spikes. Total RNAs of embryos were extracted with an RNA extraction kit following the manufacturer’s procedures (NucleoSpin RNA; Machery-Nagel). Each cDNA library was prepared using a NEBNext Ultra RNA Library Prep Kit for Illumina following the manufacturer’s procedures (New England Biolabs). High-throughput sequencing was run by a paired-end read 2 × 100-bp on a HiSeq 2500 (Illumina Inc.). The raw reads of RNA-seq datasets were filtered with FastQC (ver. 0.11.5) (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Low-quality reads were removed using the FASTX Toolkit (ver. 0.0.13) (http://hannonlab.cshl.edu/fastx_toolkit/). The clean RNA-seq data were aligned to the reference genome (Triticum_aestivum.TGACv1.dna.toplevel.fa) using HISAT2 (ver. 2.1.0) (Kim et al. 2019). Fragments per kilobase per million fragments mapped (FPKM) values were calculated using StringTie (ver. 1.3.3) (Pertea et al. 2016). Differential gene expression analysis was performed using the package DESeq2 (ver. 1.16) in R (v3.4.0).

$$\text{DESeqDataSetFromMatrix(countData}=\text{countdata, colData}=\text{coldata, design}=\sim \text{genotype}+\text{condition}).$$

Quantitative reverse-transcription PCR analysis

After total RNA was extracted from embryos with NucleoSpin RNA (Machery-Nagel), it was reverse transcribed with PrimeScript Reagent Kit Perfect Real Time (TaKaRa) according to the manufacturer’s instructions. Quantitative reverse-transcription PCR was performed with TB Green Premix Ex Taq II (TaKaRa) using LightCycler 2.0 (Roche) under the following conditions: 30 s denaturation at 95 °C, then 40 cycles of 5 s at 95 °C, 20 s at 60 °C. Primer sets15–17 used for qRT-PCR are listed in Supplemental Table 1. Each qRT-PCR analysis was performed in triplicate. Cell division control protein (CDCP; primer set 18) was used as a reference gene for data normalization (Paolacci et al. 2009). Reverse transcription PCR (RT-PCR) was also performed with CS, CS-W1 and -W2 using Quick Taq (Takara) under the following conditions: 2 m denaturation at 94 °C, then 30 or 35 cycles of 30 s at 94 °C, 30 s at 60 °C and 30 s at 68 °C. Primer sets 19 and 20 used for RT-PCR are listed in Table S1.

Table 1 Up-regulated genes found on DAP30 in G8
Full size table

Quantification of plant hormones in embryos

Grains of 30 DAP collected from a spike were divided into three groups for fresh grains, imbibed whole grains, and imbibed embryo-half grains. The imbibed grains were prepared by imbibition with distilled water at 20 °C for 24 h in a dark condition. Three biological replicates were prepared with three spikes of Fielder and G8. Plant hormones, ABA, DHZ, GA1, GA4, IAA, iP, JA, SA and tZ, of the embryos were extracted, purified and quantified with LC–ESI–MS/MS as described in an earlier report (Matsuura et al. 2019).

Results

Characteristics of transgenic lines with introduced Tamyb10-D1

Genomic Tamyb10-D1 fragment isolated from Zenkojikomugi (a red grained and dormant cultivar) contained a 2 kb promoter including the 5’ untranslated region (5’ UTR), three exons, two introns, and short 3′ untranslated region 131 bp. Comparison of the isolated sequences with data from IWGSC Chinese Spring RefSeq_v2.1_chromosome 3D (https://wheat-urgi.versailles.inra.fr/Seq-RepositoryBLAST) showed 99.6% identity in the 2 kb promoter including 5’ UTR and 100% identity from the start codon to 131 bases downstream of the stop codon. Results show that the fragment Tamyb10-D1 cloned from Zenkojikomugi has no polymorphism with Chinese Spring in amino acid sequence. A white-grain cultivar, Fielder, was transformed with the genomic Tamyb10-D1 fragment (Fig. 1a). The transformation efficiency was 2.9%. We obtained 24 hygromycin-resistant T0 plants, G1 to G25. These regenerated plants were confirmed to harbor hygromycin phosphotransferase (htp) gene (Fig. S2a). The inserted Tamyb10-D1 gene was found in these regenerated plants except for 1 plant (G13); the grains of the G13 appeared white (Fig. 1b, Fig. S2b and c). From the 24 regenerated plants, 2 plants (G8 and G12) that were assumed to have single T-DNA insertion were selected. For T-DNA insertion, T1 progenies of these T0 plants were segregated at a ratio of 3:1. We decided to use G8 and G12 progeny for germination tests.

Fig. 1
figure 1

Characteristics of transgenic lines with Tamyb10-D1. a. Diagrams of an induced vector (upper), a T-DNA inserted site on chromosome 7D (lower) and partial sequences of the insertion site (middle). b. Grain color of Fielder and G8 before NaOH treatment (upper) and after treatment (lower). Bar represents 1 cm. c. PCR products of Fielder and transgenic line G8. Three individuals were used: G8-1, 2, and 3

Full size image

The T-DNA/plant DNA junction sequence was found using inverse PCR with two restriction enzymes: Alw44I and TaqI. In G8, flanking sequences of the amplified fragments of both digested and self-ligated DNA were found to be identical, which suggests that the inserted T-DNA is a single copy onto the distal region of chromosome 7D (Fig. 1a). The insertion site was verified by PCR using primers based on T-DNA and the flanking sequence (Fig. 1c). The T-DNA insertion site was found between predicted genes, TraesCS7D02G490100 and TraesCS7D02G490200, respectively separate from about 160 kb and 50 kb. Therefore, the insertion is not expected to have affected the functions of other genes. The right-border of T-DNA was truncated (Fig. 1a), but the loss of this region might not affect transgene expression as described in reports of earlier studies (Bartlett et al. 2014). Similarly, the insertion site of G12 was surveyed. The T-DNA was inserted onto an upper region of a putative gene, TraesCS5A02G166400 on chromosome 5A (Fig. S3a–c). We selected the G8 line and used it for additional experiments to avoid unanticipated effects of insertion.

Flavonoid biosynthetic genes, chalcone synthase (CHS), chalcone isomerase (CHI), flavanone 3-hydroxylase (F3H), and dihydroflavonol 4-reductase (DFR), are essential genes for proanthocyanidin synthesis. These genes are known to be expressed in immature grains of red-grained lines (Himi et al. 2005). Expressions of these genes were strongly upregulated in immature T3 grains of G8 and G12 as well as the red-grained cultivar, Zenkojikomugi (Fig. S4a, b). Expression of the inserted Tamyb10-D1 gene driven by its own promoter, probably present in the 2 kb-5′UTR of the inserted genomic fragment, was observed in G8 and G12 in immature grains but not in leaves (Fig. S4c). Roots of hydroponically grown Fielder and G12 were also subjected to testing, but the expression was not observed (data not shown). These data demonstrated that the promoter of this gene engenders testa tissue specific expression in immature grains.

Grain dormancy of red-grained transgenic lines

The phenotype and fertility of transgenic lines G8 and G12 (T2 plants) were not clearly different from those of Fielder except for grain testa color. Germination tests were conducted at 15 and 20 °C using 50 and 60 DAP and after-ripened. Although after-ripened grains of all lines which had broken dormancy showed higher germination indices, the germination index of G8 was lower than those of wild type (Fig. 2a, b). G12 also showed lower germination index at 20 °C treatment, but no significant difference was found at 15 °C (Fig. 2b). These results suggest that Tamyb10-D1 gene regulated both grain color and grain dormancy.

Fig. 2
figure 2

Germination indices of Fielder, G8, and G12. Grains of 50 and 60 DAP and after-ripened (AR) were used. Germination assays were performed at 15 °C. (a) and 20 °C (b). White, black and gray bars respectively denote Fielder, G8 and G12. Bars represent standard errors. Different letters above the bars were found to be significantly different according to Tukey’s test (p < 0.05)

Full size image

Quantification of plant hormones in embryos

Plant hormones of both lines were quantified using liquid chromatography – electrospray tandem mass spectrometry (LC-ESI-MS/MS) to ascertain whether grain color influenced the synthesis and/or metabolism of plant hormones. Embryos of 30 DAP were collected from freshly harvested grains (FH), 24-h imbibed whole grains (IW), and 24-h imbibed embryo-half grains (IEH) from the same spike and were used for quantification as depicted in Fig. 3a. Contents of indole acetic acid (IAA) tended to increase by imbibition, but no significant difference was found (Fig. 3b). ABA levels of both lines were rapidly decreased by imbibition. Significant differences (p < 0.05) were found between non-imbibed whole grain and imbibed embryo-half grain in both lines, but no significant difference were found between the two lines (Fig. 3c). Amounts of jasmonic acid (JA) in non-imbibed whole grains of Fielder were significantly higher than in the others (Fig. 3d). No significant differences were found in the amounts of salicylic acid (SA), trans-zeatin (tZ), or isopentenyladenine (iP) (Table S2). It is particularly interesting that gibberellin A1 (GA1), gibberellin A4 (GA4), and dihydrozeatin (DHZ) were not detected even in imbibed grains.

Fig. 3
figure 3

Plant hormone quantification of Fielder and G8. a. Grains of a same 30 DAP spike were divided into three groups: (1) freshly harvested grains (FH), (2) 24-h imbibed grains with water (IW), and (3) 24-h imbibed embryo-half grains with water (IEH) and embryos of each group were used for plant hormone quantification. b. Contents of indole-3-acetic acid (IAA). c. Contents of abscisic acid (ABA). d. Contents of jasmonic acid (JA). Different letters above error bars (i.e., ± SE) denote significant differences (p < 0.05) among groups (Tukey's test)

Full size image

Differentially expressed genes between fielder and G8

Gene expressions in embryos of Fielder and G8 (T4 grains) were compared by RNA-seq analysis using the grains of 30 DAP. Embryos were collected from imbibed whole grain in water and imbibed embryo-half grain in ABA solution (Fig. 4a). RNA-seq analysis showed that 79 and 45 differential expressed genes (DEGs) were up-regulated, respectively, in water absorption and ABA absorption of G8. Moreover, 30 genes were common (Fig. 4b). Table 1 presents a list of the common up-regulated genes, except for uncharacterized protein genes and non-coding RNA. Protein kinase and Ras-related protein RABC1, which were located on homeologous chromosomes, were involved.

Fig. 4
figure 4

RNA-seq analysis of Fielder and G8. a. Grains of a same 30 DAP spike were divided into two groups: (1) 24-h imbibed grains with water and (2) 24-h imbibed embryo-half grains with ABA and embryos of each group were used for RNA-seq analysis. b. Venn diagram of the number of up-regulated differentially expressed genes of G8 (upper) and down-regulated genes of G8 (lower)

Full size image

Numbers of down-regulated genes in G8 were 79 and 14, respectively, in water and ABA; 6 genes were common (Fig. 4b). Table 2 shows the down-regulated genes in G8 except for an uncharacterized protein gene. Lipid transfer proteins (LTPs) located on homeologous chromosomes were involved in down-regulated DEGs. These genes were strong candidates as novel dormancy-related genes. Expression analysis was performed in other pairs of wheat lines that had different dormancy levels.

Table 2 Down-regulated genes found on DAP30 in G8
Full size table

Results of the plant hormone quantification suggest that the expression of genes related to hormone synthesis/metabolism might be of equal levels in both lines. Actually, the expressions of ABA 8′-hydroxylase (which catalyzes the first step in the oxidative degradation of ABA), Gibberellin 3-beta-dioxygenase 2–2 (GA3ox2-2, which converts the inactive gibberellin precursors GA9 and GA20 in the bioactives gibberellins GA4 and GA1), Jasmonic acid-amido synthetase (JAR1, which catalyzes the synthesis of jasmonates-amino), and Allene oxide cyclase 1 (AOC1, which is involved in the production of 12-oxo-phytodienoic acid (OPDA), a precursor of jasmonic acid) were not different between G8 and Fielder (Table 3). These results support that the contents of plant hormones between the lines were not different.

Table 3 Expression of phytohormone and dormancy-related genes
Full size table

Additionally, no significant differences were found in the expression levels of already known dormancy-related genes, MOTHER OF FT AND TFL1 (MFT-3A), Viviparous-1 (Vp1), mitogen-activated protein kinase kinase 3 (TaMKK3), TaDOG1L4, Protein phosphatase 2C (PP2C), and dehydration-responsive element-binding protein 1 (DREB1). However, G8 showed higher grain dormancy (Table 3). Tamyb10-D1 showed no effects on the expression of known dormancy-related genes.

Expression patterns of DEGs during imbibition

Protein kinase (PK), ras-related protein (Ras), and lipid transfer protein (LTP) were regarded as strong candidates for novel dormancy-related genes because these genes were found in duplicate, derived from homeologous genes (Tables 1 and 2). By contrast, F-box genes consisted of multifamilies; the sequences of the two genes were quite dissimilar, suggesting that these F-box genes were not homeologous. For this study, we specifically examined LTP, PK, and Ras by reliability of multiple detection. We designed sets of primers for expression analysis (Table S1).

Expressions of candidate genes were examined in other lines showing different genetic backgrounds. Chinese Spring and its white-grained mutant, CS-W1, were used for expression analysis (Fig. 5a). Tamyb10-D1 was disrupted in CS-W1 (Table 4). This mutant also showed less dormancy than wild type (Fig. 5b), as reported from an earlier study (Warner et al. 2000). Figure 5C–E show the expression levels of LTP on chromosome 4B (LTP-B), PK on chromosome 2A (PK-A), and Ras on chromosome 2D (Ras-D) of 30 DAP embryos from freshly harvested grains and 24-h imbibed grains. Significant difference (P < 0.05) was found for the expressions of LTP-B between imbibed Chinese Spring and CS-W1 (Fig. 5c). However, PK-A, -D and Ras-B, -D showed no significant differences between these lines (Figs. 5d, e, S5).

Fig. 5
figure 5

Germination indices and gene expression in Chinese Spring and white-grained mutant. a. Chinese Spring (CS), red-grained line, and its white-grained mutant, CS-W1. b. Germination Index of 40 DAP grains with CS and CS-W1. Error bars represent the standard error (SE). ** denotes significant difference (P < 0.01). c-e. Relative expression of LTP-B (c), PK-A (d), and Ras-D (e) per CDCP gene of embryos from freshly harvested grain (0 h; white bar) and 24 h-imbibed grain (24 h; black bar). Different letters above error bars (i.e., ± SE) denote significant differences (p < 0.05) among groups (Tukey's test)

Full size image
Table 4 Tamyb10 genotype of the wheat lines used for this study
Full size table

The F2 population derived from a cross between Chinese Spring and Novosibirskaya 67, white-grained line, was used for another set for verifying expression analysis. Dormancy levels of 16 red-grained lines and 5 white-grained lines are shown in Fig. S6a. Lower dormancy of white-grained lines was observed. Expression levels were analyzed using selected 5 red-grained and 4 white-grained individuals. Higher expression levels were observed in white-grained lines than in red-grained lines (Fig. S6b). Moreover, the expression of LTP-B gene was accelerated by imbibition in both lines. These results indicate that LTP-B was acting on germination processes.

Characteristics of LTP-B

LTP-B located on the distal region of the long arm of chromosome 4B (Fig. 6a). The putative amino acid sequences had characteristic eight cysteine motif (8-CM) as do other plant nsLTPs (Fig. 6b). Kouidri et al. (2018) reported 461 putative wheat nsLTP identified in Chinese Spring and classified them into five types (Types 1, 2, C, D, and G). The LTP-B in this study is a member of Type 2, the largest group, consisting of 274 genes. It was designated as TaLTP2.128 according to Kouidri et al. (2018). It is particularly interesting that Type 2 of nsLTP in rice comprises only 13 genes. The C-term inal region including 8-CM of TaLTP2.128 showed high similarity to OsHyPRP3. However, several amino acid residues of the N-terminal region were shared with OsHyPRP5 (Fig. 6b).

Fig. 6
figure 6

LTP-B gene as TaLTP2.128. a. Location of TaLTP2.128 gene on chromosome 4B. b. Alignment of putative amino acid sequences of TaLTP2.128 and, OsHyPRP3, OsHyPRP4, and OsHyPRP5 of rice. Cysteine residues shown in grey are components of the eight cysteine motif (8-CM). Bold letters signify common residues among TaLTP2.128 and OsHyPRP5

Full size image

Expression patterns of TaLTP2.128 in developing grains of wheat

TaLTP2.128 gene was found by RNA-seq with 30 DAP grains (Table 2) and was verified with other lines at the same developmental stage (Figs. 5c and S6b). Then, expression levels of TaLTP2.128 during grain development were examined in AUS1490 and EMS-AUS. The former, AUS1490, is a red-grained and dormant line; EMS-AUS is a white-grained and non-dormant mutant of AUS1490 (Himi et al. 2002; Mares and Himi 2021). Tamyb10-A1 was disrupted in EMS-AUS (Table 4). In immature 20 DAP grains, TaLTP2.128 was expressed in EMS-AUS. The expression level increased by imbibition, whereas low expression levels in AUS1490 were stable during imbibition (Fig. 7a). In 30 DAP grains, TaLTP2.128 expression was identified in imbibed AUS1490 grains, but the expression level was significantly higher in EMS-AUS (Fig. 7b). Figure 7c presents the expression levels in 40 DAP. The figure suggests that extremely high expression was observed in 48 h imbibition of EMS-AUS. Two-day (48 h) imbibition caused some grains of EMS-AUS to germinate (Fig. S7). Low and stable expression levels were also found with 40 DAP grains of AUS1490. These results indicate that the expression of TaLTP2.128 is closely linked with germination.

Fig. 7
figure 7

Relative expressions of LTP-B gene in AUS1490 and EMS-AUS. Embryos were extracted from freshly harvested whole grains (0 h) and 6, 24, and 48 h-imbibed grains. a. 20 DAP. b. 30 DAP. c. 40 DAP. Black and white circles respectively denote AUS1490 and EMS-AUS. Expression levels of 0, 6, and 24 h-imbibed grains are presented in the inset

Full size image

TaLTP2.128 expression of red-grained non-dormant mutant

These results presented up to this point were compared between red-grained and white-grained lines. To investigate the effects of the dormancy level on TaLTP2.128, expression patterns were determined in non-dormant mutant RSD32: a single recessive mutant that showed reduced grain dormancy derived from the Norin 61 by NaN3 treatment (Rikiishi et al. 2021). Norin 61 and RSD32 were a red-grained dormant line that possesses all dominant R-1b genes (Kaneko et al. 1994). Then, the effects of dormancy level on TaLTP2.128 were evaluated independently of the grain color. TaLTP2.128 expression was strongly induced by imbibition with RSD32 (Fig. 8). The reduced dormancy level of RSD32 affected the expression of TaLTP2.128 similarly to grain color.

Fig. 8
figure 8

Relative expressions of LTP-B gene in Norin 61 and RSD32. Embryos were extracted from freshly harvested whole grains (0 h) and 6, 24, and 48 h-imbibed grains of 40 DAP. Black and white circles respectively stand for Norin 61 and RSD32

Full size image

Discussion

Tamyb10 regulates grain color and dormancy

It has long been known that a relation exists between grain color and grain dormancy in wheat. Although the reason for this relation remains unclear, it has been suggested that the R-1 gene, which determines the grain color, might regulate not only pigment synthesis but also genes related to grain dormancy. Our earlier studies (Himi et al. 2011; Himi and Noda 2005) have indicated Tamyb10-1 gene as a strong candidate of the grain color regulator, R-1 gene. In this study, grain color and dormancy were modified in transgenic lines into which Tamyb10-D1 had been introduced. These results demonstrated directly that the Tamyb10-1 gene is the R-1 gene which is expressed predominantly in grains and which regulates dormancy.

Because ABA sensitivities differ among red-grained lines and white-grained lines (Himi et al. 2002), it has been predicted that the R-1 gene interacts with genes related to ABA sensitivity and biosynthesis. Matsuura et al. (2019) have already reported that no difference exists in plant hormone contents among different grain colors. Our results are consistent with their results. By contrast, Lang et al. (2021) reported that the function of Tamyb10-D1 confers pre-harvest sprouting resistance by enhancing ABA biosynthesis to delay germination. They introduced the same gene into the same host cultivar and produced transgenic red grain lines. However, their transgenic wheat was a high-expression system in which the cDNA is driven by a ubiquitin promoter, whereas our material is driven by its own promoter of the genomic fragment. The difference of promoter might affect ABA biosynthesis in transgenic lines. Our results indicate that sensitivity might be important rather than endogenous content in R-1 regulation pathway in an ABA-dependent manner of R-1 regulation for grain dormancy. As shown in Fig. 3, the amount of jasmonic acid is higher in white-grained grains in freshly harvested grains. The reason for this is unknown, but Matsuura et al. (2019) also reported no correlation between jasmonic acid content and grain dormancy, suggesting that jasmonic acid content does not cause the germinability of white-grained grains.

TaLTP2.128 is an ortholog of qLTG3-1

LTP has a common motif, 8CM, on the C-terminus, whereas the N-terminal side is more varied. The N-terminal region of the putative TaLTP2.128 amino acid sequence has similarity to that of rice OsHyPRP5 gene, which is a synonym of qLTG3-1, known as a factor that controls low-temperature germinability in rice (Fujino et al. 2008). OsHyPRP3, 4, and 5 were tandemly located on chromosome 3 spanning about 2 and 8 kb intervals. That finding might indicate that these genes were triplicated, but that the temporal and special gene expressions of these genes varied, i.e. root (OsHyPRP3), root and germinating seed (OsHyPRP4), and specifically expression in germinating seed (OsHyPRP5) (Fujino et al. 2014). These findings indicate that the LTP-B, TaLTP2.128, might also play a role in controlling seed germination in wheat similar to OsHyPRP5. The function might be determined by the N-terminal motif (Fig. 6b).

qLTG3-1 locates on rice chromosome 3, which has synteny with wheat chromosome 4, supporting the possibility that these genes are orthologous. Because qLTG3-1 promotes grain germination at low temperatures (Fujino et al. 2008), the expression of wheat TaLTP2.128 might also be temperature-dependent. It is particularly interesting that Yazdanpanah et al. (2017) reported that Arabidopsis AT2G45180, a member of nsLTP, was found by microarray analysis using freshly harvested (dormant) and after-ripened (non-dormant) seeds and the AT2G45180 expressed higher in non-dormant seeds. Figure S8 presents a comparison of the putative amino acid sequence of AT2G45180 with that of TaLTP2.128, revealing that some residues at the N-terminal regions were common, which are also found in qLTG3-1 of rice. The findings suggest that this gene of Arabidopsis might have a similar function in rice and wheat.

Both rice and Arabidopsis have 13 genes of type-2 nsLTP genes, but there are 274 genes of type-2 nsLTPs in wheat (Kouidri et al. 2018). This redundancy suggests that other nsLTPs might also be associated with grain dormancy. Fujino et al. (2014) reported collinearity of chromosomal regions around qLTG3-1 and its orthologous genes as conserved among rice, Brachypodium, sorghum, and maize. They also discussed that structural changes i.e., deletion, insertion and duplication in this area were thought to be useful for analyzing evolution and diversity. The redundancy of wheat nsLTPs is expected to be useful in this study as well.

Interconnections between grain color, TaLTP2.128, and dormancy

As mentioned above, ABA sensitivity is thought to vary with grain color. We used grains of 30 DAP and treated them with water or ABA solution to search for differentially expressed genes. Although no differences were found between lines, several peroxidase genes were down-regulated by ABA treatment in both lines (data not shown). This result may suggest that ROS is occurred in imbibed grains. The grain pigment accumulated in wheat testa is proanthocyanidins which are composed of (+)-catechin mainly, and both (+)-catechin and proanthocyanidins have been known as antioxidant. Miyamoto and Everson (1958) also showed that catechins have an inhibitory effect on wheat germination. Additionally, it was reported that reactive oxygen species (ROS) act as dormancy-breaking substances in grain germination in wheat and barley (Ishibashi et al. 2012, 2008), and it is possible that grain pigments may have an effect on removing ROS leading to inhibit germination. ABA might promote to delete ROS for inhibition of germination.

A lipid transfer protein gene was found to be an up-regulated gene in white-grained line in both ABA and water treatments, and identified as TaLTP2.128 according to an earlier report (Kouidri et al. 2018). R-1 not only regulated grain color but also modified the expression of TaLTP2.128 in imbibed grains. Furthermore, TaLTP2.128 showed high expression levels in other white-grained mutant lines, increasing with the duration of water absorption (Fig. 7). A similar expression pattern was also observed in non-dormant mutant RSD32, irrespective of grain color. TaLTP2.128 might be indirectly regulated by the R-1 gene through the modification of dormancy level. In addition, the expression of TaLTP2.128 was detected before germination, suggesting that TaLTP2.128 might act as a trigger of a transition from the dormancy stage to the germination stage.