3.1. Cloning and sequence analysis of CiHY5 from C. indicum
To explore the functions of pivotal regulators of the light-mediated pathway in response to abiotic stresses in chrysanthemum, we isolated a member of the bZIP family, CiHY5, from C. indicum transcriptome data. Based on the cDNA library, the complete 477 bp open reading frame (ORF) sequence of CiHY5 (GenBank accession number No. OP589306), which encodes a 158 amino acid polypeptide, was cloned, and the predicted molecular weight of the CiHY5 protein was 17.51 kDa. Sequence alignment revealed that CiHY5 contains a highly flexible and disordered N-terminal structure and a conserved bZIP_HY5-like domain in the C-terminus from amino acids 90 to 141, which may directly bind to G-box elements in promoter regions to activate gene expression and regulate signaling networks (Fig. 1A). Additionally, CiHY5 contains a basic region from amino acids 77 to 109 to a stable tertiary structure and a leucine zipper (LZ) domain from amino acids 110 to 141, which are responsible for DNA binding and dimerization, respectively (Fig. 1A).
In the phylogenetic trees composed of CiHY5 and 10 HY5s in the other 10 species (Fig. 1B), CiHY5 was classified with CmHY5, AaHY5 and HaHY5 from C. morifolium, Artemisia annua and Helianthus annuus in the same branch, and it was also close to AtHY5 from Arabidopsis. The length of the evolutionary branch represented the genetic variability and evolutionary distance, which indicated that CiHY5 and AaHY5 had the highest sequence similarities with the shorter branch, implying that the divergence of the HY5 family occurred later in chrysanthemum and Artemisia annua. Additionally, CiHY5 was more closely related to dicotyledons than to monocotyledons, such as OsHY5 and ZmHY5 from Oryza sativa and Zea mays, indicating the consistency of phylogenetic relationships and evolutionary conservation among CiHY5 and HY5s in other species.
3.2. Nucleus subcellular localization of CiHY5 and tissue-specific expression
Agrobacterium-mediated transient transformation of N. benthamiana leaves with the 35S:CiHY5::GFP fusion plasmid resulted in a GFP signal only in the nucleus (Fig. 2A). As a positive control, a GFP signal was observed throughout the cell after transfection with the 35S:GFP plasmid. CiHY5 was localized to the nucleus.
Tissue-specific expression analysis revealed that the transcript levels of CiHY5 in the stems and leaves were greater than those in the other tissues (Fig. 2B). Relatively lower expression levels were detected in ray florets, and the lowest expression levels were detected in roots and buds. The expression level of CiHY5 in leaves was approximately 1.4, 64.6 and 37.6 times greater than that in ray florets, buds and roots, respectively. Taken together, the gene expression and subcellular localization data suggested the potential functions and regulation of the CiHY5 transcription factor.
3.3. CiHY5 expression pattern analysis
CiHY5 has received a great deal of attention due to its pivotal regulatory capacity in response to external conditions via the light signaling pathway. The qRT‒PCR results revealed that CiHY5 expression was strongly correlated with light conditions (Fig. S3). Both white light and UVB radiation significantly stimulated CiHY5 expression, but the expression of CiHY5 did not significantly change in darkness, suggesting that CiHY5 is an important gene involved in the response to the light signaling pathway. Under salt stress, the expression of CiHY5 significantly decreased and reached its lowest level at 1 h, which was 0.54 times that at 0 h (Fig. 2C). Then, the expression levels gradually increased and peaked at 12 h at 1.52 to 0 h. When plants are subjected to environmental stresses, the hormone metabolism balance is disturbed. Interestingly, the expression trends of CiHY5 under 100 µmol/L ABA were similar to those under salt stress. At 1 h, the expression of CiHY5 decreased to the lowest level (0.39 to 0 h) and then gradually increased, but there were no significant differences at 6 h, 12 h or 24 h compared to 0 h under ABA treatment (Fig. 2D). These findings suggested that CiHY5 may act as a pivotal component involved in light, salt stress and ABA signaling responses in C. indicum.
3.4. Identification of transgenic C. indicum
To investigate the biological functions of CiHY5 in response to salt stress, the complete ORF (477 bp) and specific RNA interference (RNAi) fragment (242 bp) sequences of CiHY5 were cloned and inserted into the overexpression and RNAi vectors pBI121 and pART27, respectively, and subsequently transformed into host cells (Fig. S1, S4). Through Agrobacterium-mediated leaf disc transformation, we screened stable transgenic lines (Fig. S5). RT‒PCR was used to amplify 675 bp, 1293 bp and 2051 bp products from the Kanamycin (Kana) resistance tag, universal primers from the pBI121 vector and 35S promoter region primers, respectively, to verify the OE-CiHY5 transgenic plants, and a 789 bp product from the Spectinomycin (Spec) resistance tag primers was used to verify the RNAi-CiHY5 transgenic plants (Table S1, Fig. S5). There were no amplification results for the above primers in the WT plants. Eventually, we obtained seven independent OE-CiHY5 transgenic lines (OE-1 to OE-7) and four RNAi-CiHY5 lines (Ri-1 to Ri-4) for further exploration. Compared with those in the WT plants, the relative expression levels of CiHY5 were significantly greater in the OE-CiHY5 lines and lower in the RNAi-CiHY5 lines, indicating the successful generation of transgenic C. indicum plants (Fig. S5). OE-2, OE-4, and OE-7 and the Ri-1, Ri-2, and Ri-4 lines with high, medium, and low expression levels of CiHY5 were selected for further investigation.
3.5. CiHY5 alters the growth and development of C. indicum
To provide an overview of the phenotypes of the transgenic plants, we observed the phenotypes of the C. indicum WT, OE-CiHY5 and RNAi-CiHY5 lines. The leaf area, internode length and diameter of 8-week-old plants with consistent growth from cuttings were calculated. As shown in Fig. 3A and Fig. 3B, compared with those of the WT plants, the leaves of the OE-CiHY5 plants were greater, and the OE-2 line had the greatest leaf area among all the transgenic lines. In terms of leaf number, the leaf area of the RNAi-CiHY5 plants did not differ from that of the WT plants, but the leaf area of the Ri-1, Ri-2 and Ri-4 plants significantly decreased compared with that of the WT plants (Fig. 3C). Further exploration revealed that the OE-CiHY5 plants exhibited more and stronger phenotypes than did the WT plants. In contrast, the RNAi-CiHY5 plants presented shorter heights and greater lignification at the bottom of the stems than did the WT plants (Fig. 3D). Compared with the WT and RNAi-CiHY5 lines, the OE-CiHY5 lines exhibited obviously greater increases in both the internode length and diameter and greater synchronous growth (Fig. 3F, G, H). Furthermore, the RNAi-CiHY5 plants had relatively more roots and earlier flowering than did the WT and OE-CiHY5 plants, which was in agreement with the findings in Arabidopsis (Fig. 3E, Fig. S6).
3.6. CiHY5 affects salt stress tolerance in C. indicum
Given the multifaceted roles of HY5 in the regulation of plant growth and development in Arabidopsis, we investigated the salt stress resistance of WT and transgenic plant lines subjected to 200 mmol/L NaCl for 10 days. During the first 5 days, the growth of the plants in the OE-CiHY5 and RNAi-CiHY5 lines did not exhibit obvious differences, but during the following days, the plants in the RNAi-CiHY5 lines were more severely injured by salt stress than were those in the WT and OE-CiHY5 lines; for example, the plants were curled and dehydrated with a brown color on a large range of leaves. In contrast to those in the WT and RNAi-CiHY5 lines, slight or limited injury occurred at the edge of the OE-CiHY5 leaves with erect stems during growth (Fig. 4A, B). After water recovery for 10 d, 100% of the OE-CiHY5 plants survived and continued to sprout new buds from the apex and lateral shoots, whereas the survival rates of the WT and RNAi-CiHY5 plants in the stagnant growth state were only 66.67% and 44.44%, respectively (Fig. 4C).
To further elucidate the molecular mechanism involved in the response to salt stress influenced by CiHY5, we measured physiological parameters and indexes in WT and transgenic plants. The RWC of leaves did not obviously differ between OE-CiHY5 and RNAi-CiHY5 plants before and after 10 d of salt stress but slightly decreased in the RNAi-CiHY5 line (Fig. S7). Compared with those of the WT plants, the electrolyte leakage rates of the OE-7, Ri-1 and Ri-2 lines decreased by 27.32% and increased by 19.56% and 21.36%, respectively, confirming that the OE-CiHY5 plants could maintain the integrity of the cell membrane under salinity stress but that the RNAi-CiHY5 plants could not (Fig. 4D). Chlorophyll degradation indicates that photosynthesis is seriously impaired in plants under salinity stress. Under normal growth conditions, the total chlorophyll content of the OE-CiHY5 and RNAi-CiHY5 lines was distinctly different from that of the WT, but there was no apparent difference. However, under salt stress, a significant decrease in the total chlorophyll content in the Ri-1 (p < 0.05) and Ri-4 lines (p < 0.01) is shown in Fig. 4E, indicating severe chlorophyll degradation and sensitivity to salt stress in the RNAi-CiHY5 plants.
In terms of the antioxidant defense system, the SOD activity in the OE-CiHY5 lines significantly increased by 35.78%, 44.34% and 78.60% compared with that in the WT, and the CAT activity in the OE-7 line significantly increased by 36.31% compared with that in the WT (Fig. 4F, G). Free proline accumulation and MDA reduction are closely related to the extent of injury caused by external environmental stresses in plants. As shown in Fig. 4H and I, compared with those of the WT plants, the MDA content of the OE-4 and OE-7 lines significantly decreased, and the free proline content of all the RNAi-CiHY5 lines significantly decreased, implying that the OE-CiHY5 plants had a greater capacity to withstand salt stress and that the RNAi-CiHY5 plants had the opposite effect. Taken together, these results revealed that compared with the WT plants, the CiHY5-overexpressing and RNAi-CiHY5-overexpressing plants exhibited greater and lower salinity resistance, respectively.
3.7. CiHY5 affects the homeostasis of Na+ and K+ under salt stress
The influx and efflux of Na+ and K+ can reflect the metabolic capacity of cells and further infer salt stress resistance in plants. Under salt stress, the Na + content in the leaves of the three RNAi-CiHY5 lines significantly decreased by 18.77% and 27.84% in the OE-2 and OE-7 lines, respectively, and increased by 41.24%, 32.57% and 25.52%, respectively, compared with that in the WT (Fig. 5A). However, no significant difference was found in either the K+ or Na+ content in roots or the K+ content in leaves between the WT and transgenic lines (Fig. 5B, C, D). Intriguingly, the ratio of Na+ to K+ significantly increased in both the roots and leaves of the Ri-1 and Ri-4 lines but decreased in the roots of the OE-2 and OE-7 lines under saline conditions (Fig. 5E, F), although there was no obvious difference in the Na+ or K+ concentration in the roots. For the above six indicators, the trend of absorbed Na+ and reduced K+ in the OE-CiHY5 lines was opposite to that in the RNAi-CiHY5 lines under salt stress, with a decrease in Na+/K+ in the OE-CiHY5 plants.
3.8. CiHY5 modulates the expression of stress-responsive genes related to the ABA signaling pathway
Exogenous ABA treatment significantly downregulated the expression of CiHY5 from 1 h to 3 h (Fig. 2D). We attempted to confirm whether the expression levels of related ABA-dependent or ABA-independent pathway genes were altered in the transgenic plants. As shown in Fig. 6, compared with those in the WT, the expression levels of the ABA-dependent pathway genes CiRAB18 and CiERD7 significantly increased in the OE-CiHY5 line but decreased in the RNAi-CiHY5 line. ABA-independent pathway genes, such as CiDREB1D and CiERF1, were significantly upregulated in the OE-CiHY5 plants but relatively downregulated in the RNAi-CiHY5 lines. Moreover, the expression levels of genes encoding a series of protein kinases and protein phosphatases involved in ABA signaling, such as CiPP2C and CiSnRK2, were significantly variable. In addition, the expression of CiABF4 in the OE-CiHY5 plants was significantly greater than that in the WT plants, and the expression of CiABF4 in the OE-CiHY5 lines increased by 3.48-, 3.59- and 3.39-fold, respectively (Fig. 6). It has been reported that ABF4 is involved in the ABA signaling pathway and specifically binds to ABA-responsive elements (ABREs) through SnRK2 in response to salt stress. Additionally, a genome-wide study of HY5 target genes in Arabidopsis suggested that HY5 binds to ABF1, ABF3 and ABF4 but not to ABF2 (Fernando et al., 2018). These findings suggested that CiHY5 likely affects salt stress tolerance by modulating the accumulation of CiABF4 in an ABA-dependent manner.
3.9. CiHY5 directly combined with the promoter of CiABF4
Given that OE-CiHY5 transgenic plants significantly induced CiABF4 expression, we investigated whether CiHY5 directly bound to the promoter of CiABF4. According to the instructions of the Genome Walking Kit, we utilized TAIL-PCR technology to amplify the proCiABF4 fragment. Ultimately, we obtained a 1404 bp aligned sequence from the spliced products after three rounds of amplification; this sequence was regarded as proCiABF4 in this study (Fig. S8). The activity of proCiABF4 was verified by evaluation of the Gus gene expression of proCiABF4:Gus following visible blue in tobacco leaves (Fig. S8). Analysis of cis-regulatory elements revealed that a conserved G-box (light-responsive element) was present in the − 1104 bp to -1110 bp region in proCiABF4 (Fig. 7A). As previously described, HY5 predominantly acts as a pivotal transcription factor by binding G-box cis-acting elements downstream of target genes to regulate physiological and developmental processes in plants. In addition, the proCiABF4 region also contained MeJA-responsive elements (CGTCA motif and TGACG motif), abscisic acid responsiveness elements (ABREs), MYB and MYC recognition sites, an auxin-responsive element (TGA element) and other light responsiveness elements (AE box, ATCT motif, Box 4) (Fig. S8).
Transactivation activity analysis revealed that after cotransformation into Y2HGold cells, all the fusion plasmids grew normally on SD-Leu/-Trp media, but only the positive control (BD-53 + AD-T) grew and produced blue spots on SD/-Trp/-Leu/-His/-Ade/X-α-gal media (Fig. S9). This finding indicated that there was no transcription of downstream reporter genes through the combination of the BD DNA binding domain and the upstream activation sequence UAS in GAL4. These results indicated that CiHY5 could be used for further verification because it has no transcriptional activation activity. Yeast one-hybrid (Y1H) assay analysis showed that both the AD-CiHY5 + pHIS2-proCiABF4 (experimental group) and AD-Rec2 + pHIS2-proCiABF4 (control group) yeast strains grew normally on SD-Leu/-Trp plates, indicating that the bait and prey vectors were successfully cotransformed into yeast cells (Fig. 7B). However, the yeast strains transformed with AD-CiHY5 + pHIS2-proCiABF4 obviously grew on SD-Leu/-Trp/-His selection media supplemented with 50 mmol/L 3-AT, but at the same time, the AD-Rec2 + pHIS2-proCiABF4 yeast strains failed to grow (Fig. 7C). This result suggested that the GAL4 transcriptional domain activated PminHIS3 to express the His reporter gene in the experimental group, but the leakage of the reporter gene was inhibited by 50 mmol/L 3-AT, with no expression in the control group. Taken together, these results suggested that CiHY5 directly binds to the promoter of CiABF4 to activate downstream gene expression.
3.10. CiABF4 enhances salt tolerance in C. indicum
The complete ORF sequence of CiABF4, 1248 bp in length (GenBank accession number no. OP589307), encodes a 415-amino-acid polypeptide (Fig. S10). The phylogenetic analysis revealed that CiABF4 clustered with AtABF4, which possesses a conserved bZIP domain in the C-terminal region (Fig. S10).
To clarify whether CiABF4 plays a pivotal role in the salt stress response in chrysanthemum, we performed Agrobacterium-mediated instantaneous transformation with the pBI121-CiABF4 vector to observe the phenotypes of C. indicum leaf discs under salt stress. As shown in Fig. S11A, no obvious phenotypic variations were detected in the leaves of the WT and overexpressing-CiABF4 (OE-CiABF4) plants before salt stress treatment. When 50 mmol/L NaCl was exposed to media for 5 days, in contrast with the green and complete leaf plates of OE-CiABF4, the leaves of the WT plants were tawny, injured and susceptible to bacterial infection (Fig. S11B). After 5 days of normal growth, the OE-CiABF4 leaf plates gradually recovered and continued to differentiate, but most of the WT leaf plates still remained wilted and even putrid (Fig. S11C). These findings preliminarily verified that overexpression of CiABF4 increased the resistance of C. indicum to salt stress.