Identification of TaSPL14 in wheat
We annotated 56 sequences containing SBP domain from wheat genome sequence (TGACv1 version) [23]. Since common wheat (BBAADD) is a typical hexaploid species that generated from three diploid ancestral species Triticumurartu (AA), Aegilopsspeltoides (BB), and Aegilopstauschii (DD) [24–27]. The homoeologous genes from three subgenomes among these 56 sequences were grouped to 19 unique TaSPL genes, among which 10 had been previously isolated, named as previous report [28] (Table S1). A phylogenetic tree was performed based on the SBP-domain of 19 OsSPLs from rice and 19 TaSPLs from wheat [29–31]. According to the phylogenetic tree, the remaining 9 genes were named by orthologous genes of rice (Fig. 1A). We found three homoeologous genes including TraesCS5A02G265900, TraesCS5B02G265600 and TraesCS5D02G273900, designated TaSPL14-A, TaSPL14-B and TaSPL14-D, which showed high amino acid similarity with OsSPL14. Previous studies suggested that OsSPL14 was the target gene of OsmiR156 [7]. To verify whether miR156 could regulate TaSPL14 in vivo, we performed modified RNA ligase-mediated 5'-rapid amplification of cDNA ends (RLM-5'-RACE), and the results indicated that the TaSPL14 occurred cleavage at the 10th nucleotide from the 5' end of miR156 (Fig. 1B).
Three homoeologues genes of TaSPL14 were observed to be located on chromosome 5A, 5B and 5D, respectively, through amplifying from genomic DNA of Chinese Spring (CS) nullisomic-tetrasomic of chromosome group 5. The results showed the PCR amplification binds disappeared when the corresponding chromosome was deleted from CS lines (Fig. 1C). The expression levels of TaSPL14-A, TaSPL14-B, and TaSPL14-D among different tissues including seeding roots (SR), stems in jointing stage (SJ), seeding leaves (SL), grains at 4, 15, and 25 days after pollination (GDAP), young spikes with 5 mm (YS5), 10 mm (YS10), 20 mm (YS20) and 30 mm (YS30) in length were measured through reverse transcription quantitative real-time PCR (RT-qPCR) (Fig. 1D). The results indicated that the expression levels of TaSPL14-B or TaSPL14-D were higher than TaSPL14-A in measured tissues, TaSPL14-A and TaSPL14-D were mostly abundant in stems and young spikes, TaSPL14-B were mostly abundant in stems, leaves, seeds and young spikes, suggesting that three homoeologous genes of TaSPL14 might differently contribute to wheat development.
Functional comparison of TaSPL14 and OsSPL14
In order to investigate the biological function of TaSPL14 in wheat, we generated TaSPL14 knock-out mutant using the CRISPR/Cas9 system. In order to simultaneously knock out three homoeologous genes TaSPL14-A/B/D, we designed a specific sgRNA to target conserved region within first exon of the TaSPL14-A, TaSPL14-B and TaSPL14-D (Fig. 2A). The CRISPR/Cas9 vector pBUE411::sgRNA was transformed into wild type Fielder by Agrobacterium-mediated transformation [32]. Three independent transgenic T0 plants were obtained. DNA sequencing for TaSPL14 revealed that three homoeologous genes mutated simultaneously with frameshift mutations in mutant line #5 and #12. The mutant line #13 showed a 3-base deletion of TaSPL14-5A without shifting mutation of protein coding region, while TaSPL14-5B and TaSPL14-5D were mutated simultaneously resulting code-shifting mutations (Fig. 2B). Homozygous T2 generation plants #5 − 3, #12 − 4 and #13 − 4 were generated and planted in field.
We found all of three taspl14 mutant lines showed significantly decreased plant height and spike length compared to wide-type Fielder (Fig. 2C). Further observations revealed that the average length of rachis internode and spikelet number of taspl14 mutants were significantly decreased compared with Fielder (Fig. 2D). These results indicated that TaSPL14 affected wheat spike development through affecting both rachis internode elongation and spikelet formation. Thousand-grain weight (TGW) of mutant line #5 − 3, #12 − 4 and #13 − 4 was markedly decreased 17.3%, 22.88% and 23.7%g respectively compared to wild-type, indicating that the TaSPL14 influenced the grain weight of wheat (Fig. 2E). osspl14 (ipa1) mutants with a frame shift in protein and inactivation of OsSPL14, which caused rice mutants with notable decreased plant height, panicle length and number of flowers [10]. Moreover, thousand-grain weight increased in NIL OsSPL14ipa1 with higher expression level of OsSPL14 [7]. Thus, our results indicated that TaSPL14 acts as a conserved regulator with OsSPL14 to pleiotropically regulate plant height, spike length, spikelet number and thousand grain weight.
However, different from OsSPL14, which is a regulator of rice tillering [5], TaSPL14 is not associated with tiller development in wheat. We found that tasp14 mutant line #5 − 3, #12 − 4, #13 − 4 showed similar tiller number as wild type Fielder in heading stage (Fig. 2F). These data indicated that TaSPL14 showed functional variation from OsSPL14 in tiller development.
Downstream genes affected by TaSPL14
To understand how TaSPL14 regulated spike development, RNA sequencing (RNA-seq) experiment was performed with young spikes (20–30 mm) of wild-type Fielder and mutant line taspl14-#13 − 4 with two biological replications. Totally 1103 and 228 genes were down- and up-regulated in taspl14-#13 − 4 compared with Fielder, respectively (Table S2). Further analysis of Gene Ontology of differentially expressed genes showed that upregulated genes were mainly enriched in the pathways related to red or far red light response and jasmonic acid response, while the downregulated genes were mainly enriched in the pathways related to meristem maintenance, meristems growth, cell proliferation and ethylene response (Fig. 3A, Table S3).
Among the down-regulated genes, we observed that 7 genes including TaERF1 (ETHYLENE-RESPNSIVE TRANSCRIPTION FACTOR 1), TaRAP2.11 (ETHYLENE-RESPNSIVE TRANSCRIPTION FACTOR 2), TaTCP (TEOSINTE BRANCHED1, CYCLOIDEA AND PCF TRANSCRIPTION FACTOR), TaRAP2-1A (ETHYLENE-RESPNSIVE TRANSCRIPTION FACTOR 1), TaEIN4 (ETHYLENE INSENSITIVE 4), TaRAP2-1D (ETHYLENE-RESPNSIVE TRANSCRIPTION FACTOR 1), TaEIL1 (EIN3-LIKE 1), which associated with ethylene response exhibited a statistically significant decreased expression in taspl14-#13 − 4 compared to Fielder (Fig. 3B). The down regulation of these genes in taspl14-#13 − 4 was further validated by performing RT-qPCR analysis (Fig. 3C-3I). Therefore, these results indicated that TaSPL14 regulated the expression of genes related to ethylene response.
It was reported that OsSPL14 was able to bind GTAC or TGGGCC/T core motif in promoter region of its target genes [14], which reminded us to detect whether TaSPL14 regulate genes related to ethylene response through binding to their promoters. The core GTAC motif was identified from promoter sequence of TaEIL1, TaRAP2.11 and TaERF1, we employed the electrophoretic mobility shift assays (EMSAs) to determine whether the TaSPL14 binds on these genes. Specific labeled and unlabeled probes were synthesized from promoter region of TaEIL1, TaRAP2.11 and TaERF1 respectively. As the results showed, the specificity bindings were observed in the combinations between TaSPL14- GST fusion proteins and labeled probes, and excessive molar of unlabeled probes prevented the band shift, but mutated unlabeled probes did not, indicating the TaSPL14 bound the GTAC core motif from the promoter of TaEIL1, TaRAP2.11and TaERF1 respectively (Fig. 4A, 4B, 4C). To confirm the regulatory effect of the TaSPL14 on TaEIL1, TaRAP2.11 and TaERF1 in vivo, the transcriptional activity assays by Agrobacterium tumefaciens-mediated transient infiltration were carried out in Nicotiana benthamiana leaves. We generated three plasmids containing the 2 kb promoter of TaEIL1 (pTaEIL1::LUC), TaRAP2.11 (pTaRAP2.11::LUC), TaERF1 (pTaERF1::LUC) driving the expression of the luciferase (LUC) reporter gene respectively (Fig. 4D, 4E and 4F). At the same time, we constructed a plasmid overexpressing TaSPL14 (35S::TaSPL14) as effector. These results confidently illustrated that co-expression of 35S::TaSPL14 with pTaEIL1::LUC, pTaRAP2.11::LUC and pTaERF1::LUC led to significantly elevates the LUC reporter activity compared with that in the empty vector controls. The above results suggested that TaSPL14 could directly and dramatically elevate expression of TaEIL1, TaRAP2.11 and TaERF1 (Fig. 4D, 4E and 4F).
OsDEP1, directly and positively regulated by OsSPL14, is an important regulatory gene that affects panicle architecture [12, 14, 33] In order to detect whether SPL14-DEP1 regulatory module is conserved between rice and wheat, we detected the expression of wheat TaDEP1 between taspl14-#13 − 4 and wild type Fielder. TaDEP1-A (TraesCS5A01G215100), TaDEP1-B (TraesCS5B01G208700) and TaDEP1-D (TraesCS5D01G216900) has been identified in wheat [34, 35]. Based on the results of RNA-seq, these three homoeologous genes were not differentially expressed in young spikes between mutant line taspl14-#13 − 4 and wild type Fielder, implying that TaSPL14 might regulate spike development through other genes such as related to ethylene response not TaDEP1 (Figure S1).