Investigating the regulatory role of HvANT2 in anthocyanin biosynthesis through protein–motif interaction in Qingke

Plant materials

The purple barley variety Nierumuzha and the white barley variety Kunlun 10 were procured from the College of Agriculture and Forestry at Qinghai University (Xining, Qinghai, China). Seeds from both varieties were harvested at different growth stages following Zadoks’ growth scale (Zadok, Chang & Konzak, 1974), including the early milk ripening, late milk ripening, and soft dough stages, for subsequent RNA-seq analysis.

Gene identification

The bHLH TF families were retrieved from the barley genome (http://doi.org/10.5447/ipk/2021/3). A comprehensive genome-wide search (E-value cutoff = 0.0001 threshold filter) was executed using HMMER 3.2.1 to identify the bHLH gene families. Subsequently, the extracted protein sequences were manually verified to confirm the presence of conserved structural domains. This verification process involved consulting the SMART database to cross-reference the obtained homologous family genes.

Gene location mapping

The positional information of all bHLH TFs in the barley genome was extracted, and the chromosomal locations of barley bHLH family genes were visualized and mapped using TBtools (version 1.12).

Structure and motif analysis of genes

The online MEME web server (https://meme-suite.org/meme/tools/meme) was utilized to search for conserved motifs within the barley bHLH protein. The search criteria involved motifs with lengths ranging from 6 to 50, with an optimal length of 10, and with other parameters set to the system defaults. Following motif discovery, the relevant masthead files were downloaded, and the gene structure and motif visualization of the bHLH gene family members were performed using the gene structure view program available in TBtools.

Analysis of covariance between species of the bHLH gene family

Genome-wide covariance analyses of barley with Arabidopsis thaliana, Oryza sativa, and Zea mays were carried out using TBtools “One Step MCScanX” (Threshold: E-value = 1e⁻¹⁰, Num of Blast Hits = 5), and the results of inter-genome gene pairings were obtained. Subsequently, the gene pairing information of the gene families was extracted by using in-home Perl scripts, and finally, covariance maps were plotted using TBtools. BLASTp and MCScanX were used to analyze gene duplication events.

Multi-species evolutionary tree construction

To conduct sequence alignment, the bHLH TFs of barley and the bHLH family genes of Arabidopsis were downloaded from the http://planttfdb.gao-lab.org/index.php database. These sequences were subjected to alignment using MUSCLE software (v3.8.31), followed by further processing and alignment using TBtools. Subsequently, a phylogenetic tree was analyzed with FastTree (FastTree 2.1), after which it was visualized and drawn using iTOL (https://itol.embl.de).

Gene expression analysis in RNA-seq data

Utilizing RNA-seq data from our prior research on Qingke with different grain colors (Yao et al., 2022), we retrieved the number of enzyme fragments per million (FPKM) values for the bHLH family at each developmental stage using the OmicStudio platform. The obtained expression data were collated and subjected to analysis using Microsoft Excel 2010 and SPSS 22.0 statistical software.

Analysis of the expression pattern of the HvANT2 gene

Quantitative primers were designed using Primer 5.0. The upstream primer sequence was GAAGAAAGCTTTGGCCGGTG, and the downstream primer sequence was TCCACCTTGTGAATGGACGG. cDNAs from tissues (roots, stems, leaves, awns, and seed coats) of Nierumuzha and Kunlun 10 cDNAs at three different stages (early milk, late milk and early dough) after sowing were used as templates, and 18SrRNA was selected as an internal reference gene. TaKaRa’s TB GreenpremixExTaq II fluorescent dye was used for RT-qPCR, which was conducted on a LightCycler 480 system. The qRT-PCR reaction procedure and system setup were conducted with reference to the method of Yao et al. (2022). To calculate the relative expression of genes in different stages for Qingke, the 2^−ΔΔCt formula was applied (Yao et al., 2022). It is worth noting that all experiments were conducted with three biological replicates, and the experimental data were statistically analyzed through ANOVA using SPSS 22.0.

Identification of motifs that interact with the transcription factor HvANT2

To identify motifs interacting with HvANT2, the HvANT2 gene itself, constructed on the pGADT7 vector (ProNet Biotech Co., Ltd., Nanjing, China), was used as bait to screen a Y1H motif library (ProNet Biotech Co., Ltd., Nanjing, China). Motifs interacting with pGADT7-HvANT2 were identified by DNA sequencing and comparative analysis of positive clones. The sequences of structural genes containing these motifs (Table S1) were obtained from the online website Ensembl Plants (https://plants.ensembl.org/index.html).

Analysis of DNA and protein interaction

Based on the motifs that interacted with HvANT2, three key structural genes (CHI, F3′H, and GT) involved in the anthocyanin synthesis pathway were selected (Yao et al., 2022). The full-length CDS of HvANT2 was inserted into pGADT7 (ProNet Biotech Co., Ltd., Nanjing, China) to generate the AD-HvANT2 structure. The promoter fragment of these selected genes were cloned into the pHIS2 vector (ProNet Biotech Co., Ltd., Nanjing, China) to generate pHIS2- (pHIS2-CHI, pHIS2- F3′H, and pHIS2-GT), and it was then transformed into the Y1HGold yeast strain (ProNet Biotech Co., Ltd., Nanjing, China). The yeast colonies were subsequently transferred to plates containing SD/-Trp/-Leu and SD/-His/-Trp/-Leu media supplemented with 50 mM 3-AT and allowed to grow at 30 °C for 3–5 days.

Chromosome location of HvbHLH genes

The barley genome and proteome databases were searched using the HMM model to screen and identify genes encoding the barley HvbHLH proteins. After eliminating redundant proteins, a total of 161 HvbHLH TFs were identified (Table S2). To illustrate the chromosomal distribution of the HvbHLH TFs, their locations were mapped based on their starting positions on the chromosomes (Fig. 1). It was found that 161 barley HvbHLH TFs were successfully localized on the seven chromosomes of barley, displaying a heterogeneous distribution of HvbHLH TFs. Notably, chromosome 7 exhibited the highest number, with 31 HvbHLH TFs, while chromosome 1 contained the fewest, with only 12 HvbHLH TFs.

Phylogenetic analysis of the HvbHLH protein family

An unrooted phylogenetic tree was constructed using full-length amino acid sequences to analyze the evolutionary relationships among the bHLH TFs in Arabidopsis and barley (Fig. 2). The bHLH family was divided into 21 distinct groups, denoted as A-U. The R subfamily had the largest number of bHLH members, comprising 29 bHLH units, while the B subfamily had the lowest number, with only four members. Moreover, subfamilies B, C, and L exclusively contained genes from barley. The phylogenetic tree of barley HvbHLH TFs revealed that proteins with similar sequence lengths and structural domain positions formed distinct clusters.

Analysis of HvbHLH gene structure and protein conserved motifs

We examined the intron structure to gain insight into the gene structure of HvbHLH TFs. The analysis revealed that the intron numbers ranged from 1 to 12. Specifically, HvbHLH57 (HORVU.MOREX.r3.5HG0488610) had the highest count of 12 (Fig. 3C). Moreover, HvbHLH from the same subfamily showed similarities in their intron structures.

To elucidate the structural diversity of bHLH proteins in barley, the 161 bHLH proteins of barley were subjected to evolutionary tree construction (Fig. 3A) and conserved motif analysis (Fig. 3B) using TBtools software, which led to the identification of 10 conserved motifs (Fig. S1). The results showed that members of the bHLH TF family within the same subfamily exhibited similar types and quantities of motifs; despite this, variations in motif patterns were observed among members of the same subfamily. For example, subfamilies III, IV, and VI all have the same motifs and are aligned. Subfamily I has motifs that are largely similar, but some motifs are contiguous while others are separated. Overall, barley bHLH proteins typically displayed relatively simple motif configurations, with most sub-families comprising 2–3 members each. Notably, HvbHLH54 (HORVU.MOREX.r3.4HG0362890), HvbHLH92 (HORVU.MOREX.r3.2HG0110130), HvbHLH115 (HORVU.MOREX.r3.7HG0750050), and HvbHLH157 (HORVU.MOREX.r3.7HG0642120) had the simplest motif configurations, each with only one motif. Conversely, HvbHLH15 (HORVU.MOREX.r3.5HG0493040) and HvbHLH147 (HORVU.MOREX.r3.7HG0703220) had the most abundant motifs, with each having five distinct motifs.

Furthermore, certain motifs were specific to particular subfamilies, such as motif nine, which exclusively appeared within a single subfamily, functioning as a hallmark conserved motif for that subfamily. Similarly, Pattern 6 is only found in Subfamily V. Motif 3 was found in 38 HvbHLH proteins, predominantly originating from evolutionary branches I, II, III, IV, V, VI, and VII. Motif 4 was found in 12 HvbHLH proteins (HvbHLH12, 44, 91, 115, 81, 73, 39, 6, 42, 3, 148, and 89). Motif 5 was found in 54 HvbHLH proteins that primarily originated from evolutionary branches X, XI, XII, XIII, XIV, XV, and XVI. Motif 7 was found in 18 HvbHLH proteins originating primarily from evolutionary branches III and VIII. Motif 8 was found in seven HvbHLH proteins (HvbHLH82, 7, 68, 77, 123, 78, and 122) that were mainly derived from evolutionary branches XV and XIII. Motif 10 was found in five HvbHLH proteins (HvbHLH15, 2, 94, 16, and 147) predominantly derived from evolutionary branches V and XIV. Additionally, we found that Motif 1 and Motif 2 were characteristic conserved motifs situated within the bHLH conserved structural domain, and almost all genes include Motif 1 and Motif 2, with only HvbHLH157 containing only Motif 1, suggesting their potential significance in influencing the functionality of barley HvbHLH TFs.

Analysis of covariance between barley and the bHLH family of Arabidopsis, rice, and maize

In this study, we analyzed the covariance of the barley bHLH TF with Arabidopsis, rice, and maize using TBtools (Fig. 4A). The results showed that the size of covariance fragments was closely related to the timing of gene evolution (McCouch, 2001). Analyses of covariance between genes of different species could inform studies of their evolutionary relationships. The results revealed that barley had 18 covariant gene pairs with Arabidopsis, 145 covariant gene pairs with rice, and 186 covariant gene pairs with maize (Table S3). These findings implied that the barley bHLH TF covaried with Arabidopsis in a shorter timeframe compared to rice and maize. Consequently, it was hypothesized that the barley bHLH TFs share a higher degree of homology with rice and maize than with Arabidopsis.

Furthermore, our investigation delved into the role of segmental and tandem duplications within the HvbHLH gene family. Our findings revealed the existence of 24 homologous gene pairs among the 161 HvbHLH TFs, comprising 16 segmental duplicate pairs and eight tandem duplicate pairs (Fig. 4B). Simultaneously, we analyzed the covariance of genes within the same species. It is noteworthy that tandem and segmental duplications have played a pivotal role in the development of the gene family members and the evolution of plant genomes.

Expression profiling of the HvbHLH genes in kernels of white and purple grain colors of Qingke

Using RNA-seq data obtained from the seed coats of white-grained Qingke (Kunlun 10) and purple-grained Qingke (Nierumuzha), we analyzed the transcription expression levels within the HvbHLH gene family at different growth stages. The results showed that 142 HvbHLH genes were expressed and that 19 genes had no detectable expression in both varieties. The unexpressed genes in the seed coat of these species were hypothesized to either have no function or to be expressed in other tissues (Fig. 5A).

Furthermore, we identified 17 TFs that displayed significant differences between the two varieties (P < 0.01) (Fig. 5C). Notably, the TF HvbHLH78, also known as HvAnt2, exhibited minimal expression during the first three stages of white grains and the first two stages of purple grains. However, its expression significantly increased during the late milk and early dough stages of purple grains compared to each period of white grains (P < 0.01) (Fig. 5C). In the early dough stage, the expression level of HvAnt2 in purple grains was approximately 100 times higher than that in white grains. Additionally, in the late milk stage, the expression level of HvAnt2 in purple grains was approximately 40 times higher than that in white grains (Fig. 5C). This observation led to the hypothesis that the HvAnt2 TF plays a pivotal role in the regulation of grain color.

To investigate the expression pattern of the HvAnt2 gene in color formation in barley, we employed quantitative real-time polymerase chain reaction (qRT-PCR) to measure the relative expression levels of HvAnt2 in various tissues (roots, stems, leaves, awns, and seed coats) and at different stages in two different varieties. The results revealed low or almost negligible expression levels of HvAnt2 in all tissues of Kunlun 10. However, in Nierumuzha, HvAnt2 showed tissue-specific expression, with extremely low expression levels in roots, stems, leaves, and awns, but significantly high expression levels in seed coats. The expression levels in seed coats were tremendously elevated during the late milk stage and the early dough stage (P < 0.01), reaching their highest level in the late milk stage, which was approximately 16 times higher than that in the early milk stage (Fig. 5B). This result caused us to take great interest in the HvAnt2 gene.

Motifs interacting with the transcription factor HvANT2

We successfully obtained 41 motifs (Table 1). Subsequent re-verification experiments confirmed that all 41 motifs identified through the screening were able to grow on SD-TL, SD-TLH, and SD-TLH+50 mM 3AT plates (Fig. 6). Eventually, we identified three enzymes encoding anthocyanin structural genes, namely, chalcone isomerase (CHI), dihydroflavonol-3′-hydrogenase (F3′H), and UDP-glucosyltransferase (GT). These genes contain motifs that interact with HvANT2. Two motifs (TGCAAGC and TGCGGCC) were found in the promoter region of the CHI gene, two motifs (GAGGAAC and AGTGCAG) were found in the F3′H gene, and six motifs (GGTACTC, TCGGTCC, GTATCAT, TGCGGCC, ACAAAA, and AAGTACG) were found in the GT gene.

Motif interactions of HvANT2 with promoters of HvCHI, HvF3′H, and HvGT

The motifs obtained from the library were screened for motifs present in the promoter regions of structural genes involved in the anthocyanin synthesis pathway. The co-transformation process involved introducing the pHIS2 vector containing the CHI, F3′H, and GT genes, along with the pGADT7 vector containing the HvANT2 protein, was performed in the yeast strain Y187. In the Y1H experiment, the combinations pHIS2-CHI + pGADT7-HvANT2, pHIS2-F3′H + pGADT7-HvANT2, and pHIS2-GT + pGADT7-HvANT2 exhibited normal growth on selective media (SD-TL, SD-TLH, and SD-TLH + 50 mM 3AT), indicating positive growth (Fig. 7). These results demonstrate that pHIS2-CHI, pHIS2-F3′H, and pHIS2-GT interact with pGADT7-HvANT2 (Fig. 8).