To identify GATA TFs genome-wide in B. distachyon we downloaded PF00320, a characteristic GATA domain, from the Pfam database (Mistry et al., 2021) and searched against the B. distachyon genome protein sequence. Twenty-nine Arabidopsis and 28 rice GATA protein sequences (Reyes et al., 2004) were used to BLAST (Basic Local Alignment Search Tool) against the genome protein sequence of B. distachyon with a threshold of <e^–5 and identity of 50%. Redundant genes were manually removed before the NCBI-CDD and SMART programs were used to confirm that GATA TFs without a GATA domain were completely removed.

We used the ExPASy ProtParam¹ to predict physicochemical properties of GATA TFs, and subcellular locations were predicted using CELLO v2.5². The sequences of cDNAs, coding sequence (CDS), proteins, and DNA genes were extracted from the Ensembl Plants database³.

Gene location information was obtained from Ensembl Plants, and tandem and segmental duplication events were obtained from the PGDD (Plant Genome Duplication Database) (Lee et al., 2017) and visualized using TBtools (Chen et al., 2020). Un-rooted neighbour joining (NJ) and maximum likelihood (ML) trees were constructed using MEGA 7 with 1,000 bootstrap replications and the Poisson model based on the full-length protein sequences (Kumar et al., 2016), and visualized by Evolview v3 (Subramanian et al., 2019).

Gene structures of BdGATA genes were displayed by GSDS (Gene Structure Display Server) (Hu et al., 2015) after submitting the CDS and DNA sequences (Hu et al., 2015). MEME Suite (Bailey et al., 2009) was used to predict conserved motifs with the following parameters: number of motifs set at six, and width of motifs set from 6 to 50. The structures were visualized using Evolview v3 (Subramanian et al., 2019).

Ten-day-old B. distachyon seedlings were planted in a growth chamber at 26/24°C (day/night) with a 14/10 h day/night photoperiod. Roots, stems, leaves, and spikes collected after heading were used for tissue expression analysis. To apply abiotic stresses 10-day-old seedlings were treated with salt (200 mM NaCl), drought (20% PEG), heat (45°C), cold (4°C), ABA (200 μM), and GA (10 μM) for 2 h in hydroponic culture to obtain whole plants for analysis. Materials were frozen in liquid nitrogen and stored at −80°C for further use. RNA extraction and cDNA synthesis were performed using the RNA Easy Fast Plant Tissue and FastKing gDNA Dispelling RT SuperMix (Tiangen Biotech, Beijing) Kits, respectively.

Real-time quantitative (qRT)-PCR was performed in triplicate using SuperReal PreMix Plus SYBR Green (Tiangen Biotech). Data collection and analyses were conducted using an ABI7900 system (Applied Biosystems, Germany). Data were normalized to BdGAPDH and Atactin 8 as described previously (Hong et al., 2008; Reichel et al., 2016) and calculated using the 2^–ΔΔCt analysis method (Livak and Schmittgen, 2001). Primers used for PCR are listed in Supplementary Table 1.

The full-length coding sequence of BdGATA13 was amplified by PCR and cloned into the pCambia-1301 vector harboring the CaMV35S promoter. The recombinant vector was transformed into Arabidopsis strain Col-0 using the GV3101-mediated floral dip method (Clough and Bent, 1998). Transgenic lines were screened using a 0.1% hygromycin B solution and further confirmed by PCR. The full-length BdGATA13 coding sequence without a stop codon was inserted into the pCambia-1301-GFP vector to produce construct 35S: BdGATA13-GFP. For subcellular localization assays this construct and the vector pCAMBIA1301-GFP were co-transformed into tobacco leaves. Subcellular localization in tobacco leaves using GFP and DAPI staining was expedited by confocal microscopy (Olympus IX83-FV1200, Japan).

Seeds of wild type Arabidopsis and transgenic lines were surface-sterilized and sown on 1/2 MS plates and incubated in darkness at 4°C for 48 h before germination. Chlorophyll content was measured according to a previous study (Zhang et al., 2011). For phenotypic assessment under drought stress, 5-day-old seedlings were transplanted to 1/2 MS plates containing 0, 100 and 200 mM mannitol and cultured at 23°C in a 16/8 h (light/darkness) photoperiod for 10 days. Five-day-old seedlings for GA treatment were transplanted to 1/2 MS plates containing 0, 0.25 and 0.5 μmol GA₃ and cultured under the above conditions for 5 days. There were six replicates of each treatment; root lengths for each sample were measured by ImageJ (Rueden et al., 2017); and data were analyzed using Microsoft Excel 2010. Error bars in all graphs represent means ± S.D and “^∗” (P < 0.05) or “^∗∗” (P < 0.01) were used to indicate significant differences found by Student’s t-tests.

Twenty-seven GATAs were identified in the B. distachyon genome. To facilitate subsequent analysis, these GATA genes were named BdGATA1 to BdGATA27 according to their chromosomal location. The BdGATA loci were unevenly distributed and there were 2 to 8 genes on each chromosome (Figure 1). Eight BdGATA genes were present as four tandem duplications; and eight were in four segmental duplications.

BLAST of the CDS of each gene against the B. distachyon EST database in NCBI⁴ revealed that seven BdGATAs had no EST validation. Characteristic features of each gene were further examined. Average molecular weights, isoelectric points, and grand average hydropathicities were 35.478 KDa, 7.68, and −0.63, respectively. Detailed information for each gene is provided in Supplementary Table 2.

To elucidate the phylogenetic relationships among plant GATAs the 27 full-length GATA protein sequences in B. distachyon, 29 in Arabidopsis, and 28 in rice were extracted to build an un-rooted NJ tree. As shown in Figure 2 they were divided into four subgroups designated I, II, III, and IV based on bootstrap support. To further validate the reliability of the NJ tree the ML tree was also generated and formed the same subgroups (Supplementary Figure 1). The phylogenetically related genes were functionally conserved. For example, subgroup II members AT5G56860 (GNC), AT4G26150 (GNL), and LOC_Os06g37450 (OsGATA16) were found to participate in response to cold stress (Richter et al., 2013a; Zhang H. et al., 2021); and LOC_Os05g50270 (SNFL1) and LOC_Os02g12790 (Cga1) regulated plant architecture (Hudson et al., 2013; He et al., 2018); Overexpression of subgroup II members AT3G06740, AT3G16870, AT5G56860 (GNC), AT4G26150 (GNL), BdGATA15, and BdGATA18 in Arabidopsis showed increased chlorophyll accumulation and delayed flowering (Behringer et al., 2014). An unrooted NJ phylogenetic tree for B. distachyon placed 11, 8, 6, and 2 BdGATAs into subgroups I, II, III, and IV, respectively (Figure 3A).

Analysis of the gene structures and conserved motifs in BdGATAs (Figure 3B) showed that the numbers of exons in BdGATA genes ranged from one to eight; exon numbers in subgroups I and II varied from three to eight whereas subgroups III and IV had one to three. Six types of motifs were identified in BdGATA proteins (Figure 3C); motif one was present in all BdGATAs whereas motifs two, three, and six were specific to subgroups IV, I, and III, respectively. Motif one formed the GATA domain.

Analysis of the expression patterns of the GATA genes in root, stem, leaf, and spike tissues at different stages of plant development (Figure 4) showed that all BdGATA genes except BdGATA15 had much higher expression levels in the leaves than in other tissues. There were much lower differences among roots, stems, and spikes with some genes showing higher expression levels in roots but others with higher expression levels in stems or spikes.

Gene expression under different abiotic stresses was also evaluated using RT-PCR (Figure 5). BdGATA genes also participated in abiotic stress responses. PEG treatment showed the greatest impact on expression of all the BdGATA genes. Other treatments caused changes in expression levels of some genes. For example, the expression levels of BdGATA6, BdGATA9, BdGATA13, BdGATA14, BdGATA15, BdGATA19, BdGATA23, and BdGATA27 were significantly upregulated by GA treatment; BdGATA6 and BdGATA15 were significantly upregulated and downregulated by cold treatment, respectively; and genes 12 and 6 were significantly upregulated and downregulated by salt treatment, respectively. Some genes, such as BdGATA26, showed consistent down-regulation under most treatments while others, such as BdGATA5 and BdGATA9, showed consistent up-regulation under most treatments.

A BdGATA13-eGFP fusion driven by the 35S promoter was transformed into tobacco leaves to investigate subcellular localization. The 35S:BdGATA13-eGFP fusion protein was detected in the nucleus (Figure 6), consistent with its predicted function as a TF.

All BdGATA genes were highly expressed in leaves (Figure 4), indicating that BdGATA genes have an important role in leaf growth and development. As one example, we investigated the function of BdGATA13, a gene predominantly expressed in leaf tissue but not previously studied. This gene is phylogenetically close to subgroup II genes LOC_Os05g50270 (SNFL1), GNC (AT5G56860), and GNL (AT4G26150). To further analyze whether their functions were conserved in the phylogeny, the CDS of BdGATA13 driven by the 35S promoter was transformed into Arabidopsis, and two transgenic lines (ox-5 and ox-15) showing different gene expression levels (Figure 7A) were generated and used for phenotypic analyses. Overexpression of BdGATA13 produced dark green seedling leaves (Figures 7B,C) by accumulation of chlorophyll (Figure 7D) under both dark and light conditions. Flowering time of the transgenic lines was also delayed (Figures 7E,F).

The expression level of BdGATA13 was increased by PEG and GA treatments (Figure 5). Root growth of the transgenic lines on 1/2 MS medium was similar to wild type plants but was clearly increased relative to the WT under drought treatment (P < 0.01) (Figure 8A). When grown in 100 mM mannitol solution, the root lengths of the transgenic lines were increased by 30.70 and 38.08% compared to wild type plants (P < 0.01) (Figure 8B). When grown in 200 mM mannitol, the root lengths of the transgenic lines were increased by 42.41 and 38.19%, respectively (P < 0.01) (Figure 8B). These results clearly demonstrated that BdGATA13 enhanced drought tolerance in Arabidopsis.

The expression levels of several drought-responsive genes, including RD29A, ABI5, LEA, NCED3, SnRK2.3, and DREB2A, were also determined. Except for SnRK2.3, expression levels of these genes were induced by drought treatment (Figure 8C). These results indicated potential links between BdGATA13 and drought-related genes in Arabidopsis.

The expression level of BdGATA13 was upregulated by GA₃ treatment. As shown in Figure 9A, the root lengths of the two transgenic lines did not differ from that of wild type plants under control conditions. However, under 2.5 μM GA₃ treatment, the root lengths of the transgenic lines were 38.32–47.26% higher than the wild type (P < 0.01) and under 5 μM GA₃ treatment, the root lengths of transgenic lines were 28.69% higher than the wild type (P < 0.01) (Figure 9B).

The expression levels of gibberellin-related genes GA3ox1, GA2ox, GA20ox1, and GA20ox2 were also assayed. Expression levels of all four genes in the transgenic lines were higher than those in the WT without GA₃ treatment (P < 0.01) (Figure 9C) and under 5 μM GA₃ treatment expression levels of the four genes in transgenic plants were similar to the wild type (Figure 9C). These results suggested that BdGATA13 affects plant growth by negative regulation of GA signaling.

We identified 27 GATA TFs in the B. distachyon genome, a similar number to Arabidopsis (29) (Reyes et al., 2004), rice (28) (Gupta et al., 2017), pepper (Capsicum tetragonum L.) (28) (Yu et al., 2021), but less than in Brassica napus (96) (Zhu et al., 2020), apple (Melus pumila L.) (35) (Chen et al., 2017), and soybean [Glycine max (L.) Merill] (64) (Zhang et al., 2015). Among the 27 BdGATA genes, 16 were duplicated, including eight tandemly duplicated and eight segmently duplicated. The expression levels of the segmently duplicated genes were similar in all tissues. For example, BdGATA8 and BdGATA19 were segmently duplicated and showed low expression in stem and splike tissues but high expression in leaf tissue. These results indicated that gene duplication contributed to expansion of the GATA gene family in B. distachyon. This corresponds with findings in apple (Chen et al., 2017), Gossypium hirsutum L. (Zhang et al., 2019), B. napus (Zhu et al., 2020), and pepper (Yu et al., 2021).

Phylogenetic analysis of rice, Arabidopsis, and B. distachyon GATAs identified four subgroups (Figure 2). In the case of B. distachyon there were 11, 8, 6, and 2 members in each subgroup and subgroups shared similar gene structures and conserved motifs, and implying conserved functions among members within each subgroup (Figure 3). Among conserved motifs, motif one was present in all BdGATA members and formed a GATA domain. Expression pattern analysis showed that all BdGATA genes were highly expressed in leaf tissues. Members in the same subgroup had similar functions. For example, subgroup II members GNC (AT5G56860) and GNL (AT4G26150) contribute to chlorophyll biosynthesis as evidensed by chlorophyll accumulation in grown A. thaliana seedlings grown in light (Bastakis et al., 2018). Overexpression of Arabidopsis subgroup II members AT3G06740, AT3G16870, AT5G56860 (GNC), AT4G26150 (GNL), BdGATA15, and BdGATA18 showed increased chlorophyll accumulation and delayed flowering (Behringer et al., 2014). Subgroup II member BdGATA13 also accumulated chlorophyll when grown in light, confirming that BdGATA genes in the same subgroups have similar functions.

All BdGATA genes had relatively high expression levels in leaf tissues indicating a significant role of GATA genes in leaf development. Overexpression of BdGATA13 in Arabidopsis caused plants to be greener, due to higher chlorophyll content than in wild type controls (Figures 4, 6). It was shown previously that overexpression of BdGATA4 (named BdGATA15 in this study), BdGATA6 (named BdGATA18 in this study), and SlGATA4, SlGATA5, and SlGATA7 from S. lycopersicon in Arabidopsis produced dark green leaves and accumulated high levels of chlorophyll when grown in light (Behringer et al., 2014). These results indicate that BdGATA genes have essential roles in chlorophyll biosynthesis.

In addition to regulating chlorophyll biosynthesis and chloroplast development GATA genes also function in seed germination, flowering time, and response to abiotic stress (Richter et al., 2010; Zhang et al., 2013). For example, the Arabidopsis gnc mutant flowered earlier than the wild type and overexpression of GNC showed a late-flowering phenotype (Richter et al., 2010); whereas in wheat, overexpression of TaZIM-A1 caused delayed flowering under long-day conditions (Liu et al., 2019). Our results also showed that overexpression of BdGATA13 caused delayed flowering in transgenic overexpression lines.

Real-time quantitative-PCR results showed that all BdGAGA genes were upregulated by PEG treatment, suggesting they function in response to drought. Overexpression of BdGATA13 promoted drought tolerance in transgenic plants by regulating the expression of drought-related genes such as RD29A, ABI5, LEA, NCED3, and DREB2A. Consistent with BdGATA13 results, overexpression of OsGATA8 in rice increased tolerance to drought stress (Nutan et al., 2020), and overexpression of SlGATA17 improved drought tolerance in tomato (Zhao et al., 2021).

Previous studies showed that Arabidopsis GNC is a transcriptional target downstream of GA (Richter et al., 2010). Some plant GATAs were induced by GA₃. For example, the expression of GmGATA58 was promoted by GA₃ treatment (Zhang et al., 2020). In the present study, the expression levels of four GA-related genes in transgenic lines grown under normal conditions, namely GA3ox1, GA2ox, GA20ox1, and GA20ox2, were higher than in the WT. With GA₃ treatment, expression levels of these genes in BdGATA13 transgenic lines returned to normal levels, indicating that BdGATA13 plays a role in plant growth by negatively regulating GA signaling. The overall results suggest that BdGATA13 transcribes a TF that regulates various functions involved in abiotic stress and development in plants.