Introduction

Wheat (Triticum aestivum L.) is an important crop worldwide. Purple-colored wheat has purple-colored grains that contain significant levels of anthocyanins, commonly in the pericarp (Zeven 1991; Liu et al. 2010). The anthocyanins have antioxidant activities, which benefit human health, including reducing the incidence of colon cancer, protecting blood vessels and having anti-inflammatory effects (Hirawan et al. 2011; Zilić et al. 2012; Cavalcanti et al. 2011). The correlation results between antioxidant activity and contents of bioactive phytochemicals in purple grains have shown a high correlation for cyanidin and pelargonidin. It is shown anthocyanins content changes depending on growth stage (Sytar et al. 2018). Therefore, the breeding of purple-colored wheat, and its nutritional value, have attracted attention (Bartl et al. 2015).

Anthocyanins, which are widely found in higher plants, such as vegetables, fruits, and cereal grains, are water-soluble pigments belonging to the flavonoid group (Chen et al. 2013). The anthocyanin content increases significantly in response to stresses (e.g., drought) (Ma et al. 2014) and phytohormones (Olivares et al. 2017). Both auxins, such as IAA, 2,4-D and NAA, and CKs regulate anthocyanin metabolism (Murthy et al. 2014; Zhou et al. 2008; Das et al. 2012). For auxins, certain IAA levels inhibit IAA signaling, resulting in the decreased expression of anthocyanin-related genes, which then leads to the suppression of anthocyanin biosynthesis in strawberry (Fragaria ananassa L.), radish (Raphanus sativus L.), grape (Vitis vinifera L.) and Arabidopsis thaliana (Guo et al. 2018; Chen et al. 2016; Jia et al. 2017; Shi and Xie 2011; Liu et al. 2014). Additionally, ARF7 and ARF19 mutants have overaccumulations of anthocyanins in Arabidopsis shoots (Huang et al. 2018). However, Park et al. reported recently that appropriate auxin concentrations increase anthocyanin accumulation through enhanced production of cyanidin 3-O-glucoside and cyanidin 3-O-rutinoside in Tartary buckwheat (Park et al. 2016). In addition, anthocyanins are regulators of polar auxin carriers, suggesting crosstalk between auxin- and anthocyanin-dependent processes (Lazar and Goodman 2006; Brown and Aaron 2001; Buer and Muday 2004). CKs treatments stimulate anthocyanin accumulation through the up-regulation of anthocyanin biosynthesis genes, which then affect the transcript levels of MYB-related regulatory genes in Arabidopsis (Hammer 1995; Das et al. 2012; Nguyen et al. 2015) and maize (Piazza et al. 2002). Ji et al. reported that certain CK levels promote anthocyanin accumulation, and if the CK level exceeds a certain concentration, then the anthocyanin accumulation decreases in apple calli (Ji et al. 2015). Exogenous CKs also enhance anthocyanin accumulation through activation of anthocyanin biosynthesis genes through type-B response regulatory genes in pear (Jun et al. 2017). Thus, the inhibitory effects of exogenous auxin on anthocyanin biosynthesis can be influenced by CKs, while co-treatments with auxins plus CKs at certain concentrations significantly enhance CK-induced increases of anthocyanin in apple calli (Ji et al. 2015). However, less is known about the interactions of auxins and CKs during wheat grain development.

In this study, we investigated the effects of variations in endogenous auxin and CK levels on anthocyanin accumulation during the natural development of purple wheat grains. Based on the transcriptome sequencing results, dynamic expression levels of eight auxin-related and seven CK-related genes during anthocyanin accumulation at different developmental periods of purple wheat grains were determined. Furthermore, the endogenous contents of auxins and CKs were detected. Our results demonstrated that endogenous auxins and CKs have antagonistic roles during anthocyanin accumulation in purple wheat grains.

Materials and Methods

Plant Materials and Sample Preparation

Triticum aestivum L. cv. Guizi 1 (GZ1, Certificate No. Qian 2015003) (Li et al. 2018), which was stored in the Guizhou Sub-Center of National Wheat Improvement Center at the College of Agriculture in Guizhou University, showed strong advantages in yield, quality and disease resistant. GZ1 was planted on the experimental farm in accordance with Li et al. (2018) and the field management (including watering, weeding and fertilization) was carried out in a unified manner.

The grain samples were prepared in accordance with Li et al. (2018). The developing caryopses were selected from spikes at identical flowering stages at 10-, 25- and 35-days post anthesis (dpa), and samples were stored in − 80 °C. The caryopses from three spikes of different plants were selected and used as repeats for RNA extraction, anthocyanin determination, phytohormone determination and enzyme activity determination.

RNA Isolation and Quantitative Real-Time PCR

Total RNA from wheat grains was extracted using an EASYspin Plus Complex Plant RNA Kit (Aidlab, Beijing, China), and then the first-strand cDNA was synthesized from 1 μg of RNA using a TRUEscript 1st Strand cDNA Synthesis Kit with gDNA Eraser (Aidlab, Beijing, China). The β-actin gene was used as the internal control, and the qRT-PCR analysis was performed using LightCycler480 SYBR Green (Bio-Rad, CFX 96 Touch, USA). And then, the Delta Delta CT relative quantitation (2−∆∆ct) was used to calculate the relative expression (Livak and Schmittgen 2001). The primers for qRT-PCR are listed in Table 1.

Table 1 Sequences of specific genes for the qRT-PCR analysis
Full size table

Extraction and Measurement of Total Anthocyanin

The extraction and quantitative determination of anthocyanin levels were performed in manners similar to those of Li et al. (2018). The frozen samples were weighed (0.5 g) and ground into a powder in liquid nitrogen, and then 4 mL acidified ethanol was added and mixed with the sample for anthocyanin extraction. Total anthocyanin was determined by measuring the absorbance at 527 with Evolution 220 (Thermo Fisher, USA) and calculated using the following formula: anthocyanin content (mg/Kg) = (C × V)/m, where C represents the anthocyanin concentration, V represents the volume , and m represents the sample weight.

The Differentially Expression of Auxin- and CK-Pathway Genes in the Anthocyanin Biosynthesis Pathway

The differentially expressed genes of auxin- and CK-pathway genes in 10 dpa, 25 dpa, and 35 dpa were filtered from the KEGG biological pathway annotation. The number of clusters was set to 30. Venn diagrams were generated using the online tool at https://www.bioinformatics.psb.ugent.be/webtools/Venn/. The expression levels of these genes were calculated using FPKM value (fragments per kb per million reads), and the value of log2Ratio (the ratio was from the FPKM values of 25 dpa VS 10 dpa, 35 dpa VS 10 dpa, and 35 dpa VS 25 dpa) was used to draw the heat map of gene expression (Excel 2013, USA). Furthermore, real-time PCR was performed to verify the accuracy of genes expression including auxin-pathway genes (GzPIN1, GzLAX3, GzIAA12, GzARF13, GzARF17, GzABP1, GzAKR1, and Gz5NG4) and CK-pathway genes (GzZOG2, GzZOG3, GzUGT, GzAHK3, GzCKX1, GzCKX2, and GzCKX2.4). The experiment was performed with ≥ 3 replicates to calculate the average value. Statistical significance was determined using Student’s t-test.

Determination of Antioxidant Enzymes

The extraction and determination of the antioxidant enzymes POD, CAT, and SOD were performed in accordance with the instructions in the Cominbio kit (Jiangsu, China). The frozen samples (0.1 g) were ground into powders in liquid nitrogen, and then 1 mL extraction solution was added. The mixture was homogenized on ice in an EP tube. Then, the tube was centrifuged at 8000×g for 10 min in 4 °C, and the supernatant was used for activity detection with a Multiskan FC Microplate reader (Thermo Fisher, USA). For the SOD content, 200 μL of the supernatant was placed in an ELISA plate and incubated for 30 min at room temperature. Then, the absorbance (A) was measured at 560 nm, and the sample and reagent blank were recorded as A1 and A0, respectively. The inhibition percentage (IP) was calculated using the following formula: IP = (A0–A1)/A0 × 100%. The SOD content was calculated using the following formula: SOD content (U/g) = 11.11 × IP × (1–IP)−1 × m−1, where m represents the sample weight. For the POD content, 200 μL of the supernatant was placed in an ELISA plate, and the POD content was measured at 1 min and 2 min at 470 nm and recorded as B1 and B2, respectively. Then, the following formula was applied: POD content (U/g) = 4000 × (B2−B1) × m−1, where m represents the sample weight. For the CAT content, 200 μL of the supernatant was placed in an ELISA plate, and the CAT content was measured at 0 min and 1 min at 240 nm and recorded as C1 and C2, respectively. Then, the following formula was applied: CAT content (U/g) = 918 × (C2–C1) × m−1, where m represents the sample weight.

Determination of Phytohormones

The extraction and determination of phytohormone levels were performed in accordance with Shao et al. (2019). The frozen samples (80 mg) were ground to a powder in liquid nitrogen. The powder was added to 1 mL ethyl acetate (containing 50 μL internal standards) and shaken for 2 min to mix well. Then, phytohormones were extracted from the powder at 4 °C for 12 h in the dark. The supernatant (800 μL) was collected after centrifugation (14,000×g for 10 min at 4 °C) and was then evaporated to dryness under N2. The residue was resuspended in 100 μL 50% acetonitrile (v/v). After being centrifuged (14,000×g for 10 min at 4 °C), the supernatant was then analyzed by HPLC–ESI–MS/MS at the Shanghai Applied Protein Technology Company (Shanghai, China). The following steps were the same as described Shao et al. (2019).

Results

The Phenotype and Anthocyanin Accumulation During the Natural Development of GZ1 Grains

Natural grain development was observed and the change in caryopsis size was calculated. The width increased 1.4-fold and the length increased 1.71-fold from 5 to 20 dpa, respectively. Then, the increment rate of grain size being decreased gradually until it reached its final size (Fig. 1b). Visible effects of anthocyanin accumulation appeared at 20 dpa and then gradually spread to almost all of the grain surface (Fig. 1a). Subsequently, the anthocyanin content began to rapidly increase at 20 dpa and reached a maximum level by 35 dpa, which was consistent with phenotypic observations (Fig. 1c).

Fig. 1
figure 1

Phenotypic observations and anthocyanidin content determinations during the natural development of GZ1 grains. a The phenotypic observation at 5, 10, 15, 20, 25, 30, and 35 dpa. b Dynamic changes in grain size during natural grain development. c Dynamic changes in anthocyanidin content during natural grain development

Full size image

Dynamic Expression of Auxin- and CK-Pathway Genes During Anthocyanin Biosynthesis at Different Grain Development Periods

Based on the transcriptome sequencing results of three important periods (10, 25, and 35 dpa) for anthocyanin biosynthesis, the upregulated expression of structural and regulatory genes may result in anthocyanin biosynthesis and the resulting coloration of GZ1 grains (Li et al. 2018). In order to identify transcripts common at different grain developmental periods, Venn diagrams of auxin-pathway genes (Fig. 2a–c) and cytokinin-pathway genes (Fig. 2d–f) were generated separately for the upregulated and for the downregulated transcripts at 10, 25 and 35 dpa. For the auxin-pathway genes, the highest overlap was the upregulated transcripts. For the cytokinin -pathway genes, the highest overlap was the downregulated transcripts.

Fig. 2
figure 2

Venn diagram analysis of downregulated and upregulated transcripts at different grain development stages 10dpa, 20dpa, and 35dpa. a–c Venn diagram analysis of auxin-pathway genes. d–f Venn diagram analysis of cytokinin-pathway genes

Full size image

Then, we studied dynamic changes in the expression of auxin- and CK-related genes during anthocyanin biosynthesis at different grain development periods in GZ1 grains. Herein, 22 auxin-related genes that underwent significant expression changes were selected from transcriptome sequencing results (Li et al. 2018). The expression of the auxin efflux transporter gene GzPIN1 and auxin influx transporter gene GzLAX3 was significantly higher at 25 dpa and 35 dpa than at 10 dpa, and exhibited dynamic increased from 10 to 25 dpa and then decreased from 25 to 35 dpa. However, the expression level of the auxin influx transporter gene GzLAX2 gradually decreased during anthocyanin accumulation. Here, nine AUX/IAA genes and six ARF genes related to auxin signaling were selected. The expression levels of GzIAA1, -3, -4, -12, GzIAA14–16, and GzIAA30 were significantly higher at 25 dpa and 35 dpa than at 10 dpa, and GzIAA1, -3, -15, -16 and -30 exhibited dynamic increased from 10 to 25 dpa and then decreased from 25 to 35 dpa. However, GzIAA4, -12 and -14 gradually increased. Additionally, the expression of GzIAA19 decreased from 10 to 25 dpa and then increased from 25 to 35 dpa. The six ARF genes showed transcriptomic diversity and complexity during grain development, among which GzARF4, -5 and -17 decreased from 10 to 25 dpa and then increased from 25 to 35 dpa in contrast to GzARF8. GzARF6 gradually increased and GzARF13 gradually decreased during grain development. Overall, our data indicated that AUX/IAA genes may positively correlated with anthocyanin accumulation during the natural development of GZ1 grains, while ARF genes may play negative roles. The auxin-binding protein gene (GzABP1) and auxin-induced protein genes (GzAKR1, Gz5GN4, and GzARP) also showed transcriptomic diversity and complexity during grain development. GZ1 is a polyploid wheat, and a heat map of these genes, including copies and alternatively spliced transcripts, was constructed (Fig. 3).

Fig. 3
figure 3

Heat map of auxin-pathway genes, including copies and alternatively spliced transcripts, during anthocyanin accumulation in GZ1 grains. Different combinations of numbers and letters (7A, 7B, 7D, 1B, 1D etc.) represent different chromosomal locations. “New” indicates that the chromosomal location is presently not clear. The same chromosome with different numbers represents corresponding copies of a gene located on that chromosome. Numbers after a capital “T” (T1, T2, T3, etc.) represent alternatively spliced transcripts of corresponding genes or copies

Full size image

In total, 10 significantly changed CK-related genes, including CK-O-glucosyltransferases (ZOGs), uridine diphosphate glycosyltransferases (UGTs), Arabidopsis histidine kinases (AHKs), CK-N-glucosyltransferase (ZNGs), and CK oxidases (CKXs), were selected for heat map construction (Fig. 4). The expression levels of GzZOG2, GzZOG3 and GzZNG2 were significantly higher at 25 dpa and 35 dpa than at 10 dpa, it increased from 10 to 25 dpa and then decreased from 25 to 35 dpa. The UGT on chromosome 3 gradually decreased during grain development in contrast to the gene on chromosome 6A. The CK receptor GzAHK3 decreased from 10 to 25 dpa and then increased from 25 to 35 dpa. The five GzCKX genes showed transcriptomic diversity and complexity during grain development, with GzCKX1-NEW, GzCKX2.4-3A, GzCKX2.4-3B1, GzCKX2.4-3D1, GzCKX2.4-3D3, GzCKX2-3A, and GzCKX3-7A gradually decreased; GzCKX1-2D, GzCKX2.4-3B2, GzCKX2.4-3D2, GzCKX3-7B, GzCKX3-U1, GzCKX8-2A, and GzCKX8-2B increased from 10 to 25 dpa and then decreased from 25 to 35 dpa; and GzCKX8-1B decreased from 10 to 25 dpa and then increased from 25 to 35 dpa. Our results demonstrated that different copies or alternative splicing transcripts showed different expression pattern, which may be responsible for different biological function.

Fig. 4
figure 4

Heat map of cytokinin-pathway genes, including copies and alternatively spliced transcripts during anthocyanin accumulation in GZ1 grains. Different combinations of numbers and letters (7A, 7B, 7D, 1B, 1D, etc.) represent different chromosomal locations. “New” indicates that the chromosomal location is presently not clear. The same chromosome with different numbers represents corresponding copies of a gene located on that chromosome. Numbers after a capital “T” (T1, T2, T3, etc.) represent alternatively spliced transcripts of corresponding genes or copies

Full size image

The qRT-PCR Validation of Auxin- and CK-Pathway Gene Expression Levels During Anthocyanin Accumulation in GZ1 Grains

To confirm the expression patterns of auxin- and CK-pathway genes related to anthocyanin biosynthesis in GZ1 grains, eight auxin- and CK-pathway genes, copies, or their alternatively spliced transcripts were selected for qRT-PCR validation. Among the former, GzIAA12-5A (Fig. 5c) and Gz5NG4-7A1 (Fig. 5h) were upregulated during anthocyanin biosynthesis. GzARF13-7D (Fig. 5d) and GzAKR1-7B2 (Fig. 5g) were downregulated. GzPIN1-7B (Fig. 5a), GzLAX3-1D (Fig. 5b), and GzABP1-5D (Fig. 5f) were significantly increased at 25 dpa and then reduced by 35 dpa compared to 10 dpa, in contrast to GzARF17-7A-T1 (Fig. 5e), which was significantly reduced at the 25 dpa and then significantly increased by 35 dpa. Among the latter, GzZOG2-2D (Fig. 6a) and GzAHK3-3D-T1 (Fig. 6e) were upregulated during anthocyanin biosynthesis. GzUGT-3D1 (Fig. 6d), GzCKX2-3A (Fig. 6g), and GzCKX2.4-3D3 (Fig. 6h) were downregulated. GzZOG2-7D (Fig. 6b) and GzZOG3-7A (Fig. 6c) were significantly increased at 25 dpa and then significantly reduced by 35 dpa compared to 10 dpa, while GzCKX1-2D significantly increased and then slightly decreased.

Fig. 5
figure 5

The qRT-PCR validation of auxin-pathway genes involved in anthocyanin biosynthesis in GZ1 grains. Each bar shows the mean ± SD of triplicate assays. A-H) the expression pattern of GzPIN1-7B, GzLAX3-1D, GzIAA12-5A, GzARF13-7D, GzARF17-7A-T1, GzABP1-5D, GzAKR1-7B2, and Gz5NG4-7A1 at different grain development stages 10dpa, 20dpa, and 35dpa, respectively. Different letters represent significant differences (P < 0.05)

Full size image
Fig. 6
figure 6

The qRT-PCR validation of cytokinin-pathway genes involved in anthocyanin biosynthesis in GZ1 grains. a–h the expression pattern of GzZOG2-2D, GzZOG2-7D, GzZOG3-7A, GzUGT-3D1, GzAHK3-3D-T1, GzCKX1-2D, GzCKX2-3A, and GzCKX2.4-3D3 at different grain development stages 10dpa, 20dpa, and 35dpa, respectively. Each bar shows the mean ± SD of triplicate assays. Different letters represent significant differences (P < 0.05)

Full size image

Dynamic Changes in the Auxin and CKs Contents During Anthocyanin Accumulation in GZ1 Grains

We determined the dynamic changes in the endogenous auxin and CK contents during anthocyanin accumulation in GZ1 grains. The endogenous auxin (IAA) content significantly increased at 25 dpa and then decreased by 35 dpa compared to 10 dpa (Fig. 7a). The contents of five individual active CKs, iPR, cZ, czR, tZ, and tZR, were detected. During anthocyanin accumulation, the iPR and czR contents significantly decreased (Fig. 7b and d). The tZ contents at 25 dpa and 35 dpa, were significantly lower than the level at 10 dpa (Fig. 7e). In contrast to the IAA content, the cZ and tZR contents were significantly decreased at 25 dpa and then increased by 35 dpa (Fig. 7c and f). In summary, the auxin and CKs may play antagonistic roles during anthocyanin accumulation in GZ1 grains.

Fig. 7
figure 7

Auxin and cytokinin contents during anthocyanin accumulation in GZ1 grains. a Changes of IAA content during anthocyanin accumulation; b–f Changes of CKs (iPR, cZ, czR, tZ, and tzR, respectively) content during anthocyanin accumulation. Each bar shows the mean ± SD of three biological replicates

Full size image

Antioxidant Enzyme Contents During Anthocyanin Accumulation in GZ1 Grains

Furthermore, we detected the contents of the antioxidant enzymes POD, CAT and SOD (Fig. 8). The POD and CAT contents were significantly increased at 25 dpa and then decreased by 35 dpa. However, the SOD content was significantly decreased at 25 dpa and then increased by 35 dpa.

Fig. 8
figure 8

Contents of antioxidant enzymes during anthocyanin accumulation in GZ1 grains. Each bar shows the mean ± SD of three biological replicates. a Content of POD during anthocyanin accumulation; b Content of CAT during anthocyanin accumulation; c Content of SOD during anthocyanin accumulation. Different letters represent significant differences (P < 0.05)

Full size image

Discussion

Exogenous auxins and CKs regulate anthocyanin accumulation in plants (Liu et al. 2014; Park et al. 2016; Nguyen et al. 2015; Piazza et al. 2002; Ji et al. 2015; Jun et al. 2017). However, no information is available on the roles of auxin and CKs in the natural grain development of wheat. In this study, we explored the phenotypes of GZ1 grains at different development stages by observing and recording changes in the caryopsis size and anthocyanin content. By 20 dpa, GZ1 grains had basically reached their final size and the anthocyanin content started to gradually increase up to its maximum level (Fig. 1), which is in agreement with a previous study (Li et al. 2018).

Based on our previous transcriptome sequencing results, we determined that auxin-related genes (such as GzPIN1, GzLAX3, GzIAA12, and GzABP1) and CKs-related genes (such as GzZOG2, GzUGT, and GzCKX2) may play important roles in anthocyanin accumulation during grain development (Figs. 3 and 4). Because of gene expression redundancy owing to polyploidy, the genes often contain multiple copies and alternative splice forms, which greatly increase transcript diversity (Himi and Noda 2004; Egawa et al. 2006; Yoo et al. 2014). Different copies or transcripts of genes, such as GzIAA16 and GzZOG2, did not exhibit exactly the same expression patterns, which demonstrated that these copies or transcripts may play different functions in the precise regulation of anthocyanin accumulation. The expression levels of auxin biosynthesis genes (such as YUC, TAA, and TAR) are low or even absent during the growth and ripening process (Yuan et al. 2019). Here, the TAA and TAR gene families were not detected, and only the GzYUCCA2 gene, at a low expression level, was detected during anthocyanin biosynthesis at different developmental periods, which indicated that the caryopsis is not the site of IAA biosynthesis in wheat (data not shown). In general, the expression patterns of GzIAA12-5A, Gz5NG4-7A1, GzARF13-7D, GzARF17-7A-T1, GzABP1-5D, GzAKR1-7B2, GzUGT-3D1, GzCKX2-3A, GzCKX2.4-3D3, GzZOG2-7D, and GzZOG3-7A as assessed by qRT-PCR were consistent with our previous transcriptome sequencing results (Li et al. 2018). Furthermore, polar auxin carriers (GzPIN1 and GzLAX3), auxin receptors (GzABP and GzARP) and signaling components (most of the GzARFs and GzAUX/IAAs) at 25 dpa and 35 dpa showed significantly higher expression levels than at 10 dpa, having patterns that increased from 10 to 25 dpa and then decreased from 25 to 35 dpa (Figs. 3 and 5), which was consistent with our dynamic anthocyanin accumulation rate results (Fig. 1c) and the dynamic changes in the endogenous auxin content during anthocyanin accumulation in GZ1 grains (Fig. 7a). The high level of auxin at 25 dpa demonstrated that a rapid anthocyanin accumulation required a higher auxin content, which was in agreement with results in Tartary buckwheat (Park et al. 2016). However, our results were similar to those of Li et al. (2018) in which anthocyanin accumulation had normal levels of variation among different years owing to the environmental conditions (Liu et al. 2005; Knievel et al. 2009). CK homeostasis is precisely regulated owing to distinct developmental requirements. CK degradation genes (such as CKX) and inactivation genes (such as UGT) play important roles in maintaining CK homeostasis (Smehilova et al. 2016). Here, our results demonstrated that GzCKX (GzCKX1, GzCKX2, GzCKX2.4, and GzCKX8) and GzUGT genes gradually decreased (Figs. 4 and 6), which means the enhanced degradation and/or inactivation of endogenous CK is gradually eliminated. The differential expression of GzCKX3 may be related to its diverse roles in CK homeostasis maintenance (Koellmer et al. 2014). Nguyen et al. confirmed that the repression of CK signaling leads to increased anthocyanin biosynthesis in Arabidopsis (Nguyen et al. 2016). Combined with the decreasing endogenous CK content (Fig. 7b–f), these results led us to propose that CK negatively regulates anthocyanin accumulation in natural GZ1 grain development. In addition, GzZOGs, GzZNGs, and GzAHKs also participate in this process. Auxins and CKs interact during anthocyanin biosynthesis in plants (Ji et al. 2015; Teribia et al. 2016). Recently, a comparative transcriptome analysis of purple grain wheat demonstrated that auxin- and CK-related genes participate in the light-promoted anthocyanin accumulation process (Wang et al. 2018). We determined that auxins and CKs may play antagonistic roles in anthocyanin accumulation by regulating related genes in purple wheat grain development.

Both anthocyanins and antioxidant proteins (such as POD, CAT, and SOD) can act as ROS-scavenging antioxidants to protect cells from stress by reducing the ROS level (Mittler et al. 2011; Nakabayashi et al. 2014; Lotkowska et al. 2015). Here, the antioxidant protein contents were detected and had different trends (Fig. 8). In particular, the dynamic changes in POD and CAT content were consistent with the dynamic change in the anthocyanin accumulation rate (Fig. 1c), which implied that ROS signaling participates in the regulation of anthocyanin accumulation. Thus, the coordinated promotion of auxin signaling and repression of CK signaling lead to increased anthocyanin biosynthesis during the natural grain development of purple wheat. Interestingly, different copies or transcripts of the same auxin- and CK-related genes showed diverse expression patterns. Thus, further research is required to elucidate this complex network underlying the auxin and CK signaling involved in anthocyanin accumulation.