Identification and analysis of RcAP2/ERF
We used BLASTP and HMMER3.1 in the entire rose genome to discover AP2/ERF members. All potential rose proteins were then subjected to domain analysis to confirm the presence of the AP2 protein domain (PF00847). Eventually, a total of 135 AP2/ERF family members were identified in rose. According to the distribution order of their chromosomal locations, they were named consecutively from RcAP2/ERF1 to RcAP2/ERF135. Their characteristics including gene name, locus ID, physical position, and other properties are shown in Table S1. RcAP2/ERF segregated to 7 rose chromosomes and mainly localized in the nucleus, chloroplast, cytoplasm, and mitochondria. The longest sequence with 832 amino acid residues was RcAP2/ERF52, whereas the shortest having 125 amino acids was RcAP2/ERF132. The predicted isoelectric points ranged from 4.52 (RcAP2/ERF54) to 10.28 (RcAP2/ERF99). The protein molecular weight varied between 13.94 kDa to 91.01 kDa. Our findings revealed that different biological roles were played by individual coding regions of RcAP2/ERFs.
RcAP2/ERF gene structure
The gene structure of RcAP2/ERF is closely related to its function, and together with their phylogenetic analysis, could reflect the phylogenetic relationships among the RcAP2/ERF. As shown in Fig 1, the RcAP2/ERF family was divided into seven groups. Analysis of RcAP2/ERF structure for exon/intron organizations revealed that each gene had 0 to 9 introns (Fig 1), the genes clustered into the same branch on the phylogenetic tree were found with similar exon-intron structure. The results provided support validating our analysis regarding gene structure. Further, most RcAP2/ERF (94, 69.62%) contained no introns, only 19 (14.07%) contained one intron whereas RcAP2/ERF30 contained nine introns. It was surprising yet interesting that none of the RcAP2/ERF contained three introns. AP2 gene configuration was reasonably conserved among family members, and 51% of AP2/ERF genes had no intron (Fig 1). Typically, closest family members had a common exon/intron configuration concerning intron number and phase as well as the length of exons.
In an attempt to carry out the in-depth examination of typical RcAP2/ERF protein sequences, we assessed 135 typical arrangements from the MEME website. We predicted that the conserved sequences would amount to 10 in number their sequences were sought (Fig 1). Besides, we also observed that common patterns with the same orientation and location were most apparent among phylogenetically close relatives. This suggests common roles for AP2 members located in similar subgroups. Motifs 1 and 7, with a net presence of 61.48 percent (83/135) and 60.74 percent (82/135) respectively were frequent among RcAP2/ERF. Several preserved motifs have been identified in various subgroups. Phylogenetic investigations show similarity in gene composition and pattern arrangements of subgroup members adding confidence to classification accuracy.
Phylogenetic and evolutionary analysis of RcAP2/ERF
To assess phylogenetic associations of rose AP2/ERF, we aligned AP2 domain sequences of rose,apple (Malus demostica), and Arabidopsis to construct a parallel phylogenetic tree . 135 RcAP2/ERF proteins of rose, 260 members from apple, and 167 members from Arabidopsis were selected for phylogenic analysis (Fig 2). Using the same criteria for classification as in Arabidopsis, the AP2/ERF proteins of rose were classified into DREB, Solosist, AP2, and ERF segments containing 44, 6, 17, and 68 RcAP2/ERF proteins, respectively (Table 1). The RcAP2/ERF gene members within groups were mostly the same except ERF-B1, ERF-B6, and ERF-B3. DREB segment is further split into six subgroups, from A1 to A6, containing 7, 11, 1, 15, 7, 3 AP2/ERF proteins in rose and 3, 25, 2, 18, 18, 10 AP2/ERF members in apple, respectively. ERF subfamily being the most abundant type in rose, accounted for 50.37% of all RcAP2/ERF proteins, which was closely similar to apple, with a percentage of 51.54%. Arabidopsis has the lowest percentage of ERF subfamily members, which accounted for 46.11%. ERF subfamily was also further segregated to B1-6 subgroups, containing 12, 3, 13, 15, 5, and 20 RcAP2/ERF proteins, respectively. AP2 subfamily contains 17 members of RcAP2/ERF proteins, whereas the Solosist subfamily is the smallest, consisting of only 6 members, indicating that RcAP2/ERF genes were distributed in different clades unevenly.
Chromosomal location and gene duplication of RcAP2/ERF
Mapchart software was used to identify the RcAP2/ERF chromosomal location. All RcAP2/ERFs had precise positions on the chromosomes. Each rose chromosome contains more than 11 RcAP2/ERFs (Fig. 3A). The RcAP2/ERF genes on 7 chromosomes are randomly and unevenly distributed. Our maximum predicted the number of RcAP2/ERF was 38 genes on chromosome 2 and can be compared to rather fewer, 11 genes located on chromosomes 5 and 3. Although chromosome 3 is the shortest and chromosome 5 is the longest in rose they contain the least RcAP2/ERF. Therefore, there is no apparent correlation between chromosome length and RcAP2/ERF gene distribution.
Segmental and tandem duplication are evolutionary mechanisms that result in generating groups of genes . Hence, Synteny and MCScanX  were used to analyze the duplication and segmental events of RcAP2/ERF. As shown in Fig. 3B, 20 genes were confirmed as tandem duplicated genes. Chromosomes 7 and 4 had three groups of two duplicate tandem genes, while chromosomes 6 and 1 had five and four duplicate tandem genes respectively. Segmental duplication events were also detected in each chromosome and accounted for 24.4% of the RcAP2/ERFs. Taken in combination, these outcomes indicated that segmental and tandem duplication extensively contributes to the expansion of the RcAP2/ERF family, whereas the former is being more deeply involved in particular.
Analyzing cis-regulatory motifs in promoters of RcAP2/ERF genes
To establish the gene expression control of RcAP2/ERF, a bioinformatics review was performed to detect apparent cis-regulatory regions in the RcAP2/ERF gene promoter region. This promoter zone was identified by the ~1500 bp upstream area of the initial transcript sites. Fig. 4 illustrates that different promoter cis-elements were grouped into three groups, including plant growth and development (MBSI, E2FB, MSA-like, P-box, TCA-element, AT-rich element,), phytohormone responsive (ABRE, CGTCA-motif, ERE, TGACG-motif, AuxRR, GARE-motif, TATC-box), and abiotic and biotic stresses (MBS, TC-rich repeats, LTR, DRE, WUN-motif, MYC, W-box, MYB, STRE, WRE3). These cis-regulatory elements were found to be randomly dispersed among various promoters of RcAP2/ERF genes. ERF subfamily B1 group member RchiOBHmChr5g0008991 has the largest number of cis-elements, accounting for 43. MYB and MYC elements were present in 79.4% of RcAP2/ERF gene promoters, whereas 77.9% were found with ABRE. Moreover, 66.2%, 50.7%, 45.6%, and 44.9% of RcAP2/ERF gene promoters were presenting MeJA, gibberellin, W box and MBS, respectively, which indicating important contributions to gene expression of the RcAP2/ERF gene. However, RcAP2/ERF gene promoters had a lower abundance of some motifs such as LTR (33.8%), DREB (32.4%), and salicylic acid responsiveness (31.6%), suggesting that these regions controlled the expression of particular genes only under appropriate circumstances. These analyses of cis-regulatory elements implied that RcAP2/ERF genes may not only be involved in hormone signaling but also participated in reacting to external stresses.
Isolation, structure, and promoter analysis of RcDREB2B
As DREB A2 subgroup members are considerably involved in abiotic stress tolerance in different plant species, we select RcDREB2B, which belongs to DREB A2 subgroup member, for further analysis. The primers of RcDREB2B were designed for PCR amplification, and full-length cDNA sequences were cloned. The sequences of RcDREB2B cDNA and deduced amino acid were submitted to the National Center for Biotechnology Information (NCBI) GenBank (MH152409). It was revealed by sequence analysis that an open reading frame (ORF) of 585 bp was contained in the full-length cDNA and encoded a putative protein of 194 amino acids with a pI of 9.03 and a predicted molecular weight of 21.2 kDa.
It is evident from multiple sequence alignment between reported DREB A2 subgroup members (AtDREB2B, TaDREB1, ZmDREB2A, and HvDRF1) and RcDREB2B that all proteins demonstrate the characteristics features of DREB proteins, specifically a conserved DNA-binding domain (AP2-domain) comprising 64 amino acids, and conserved YRG, WLG, and RAYD motifs. Glutamic acid (E19) was conserved at the 19th position in the AP2 domain in RcDREB2B, HvDRF1 and TaDREB1 (Fig. S2). Moreover, the AP2 domain manifested a greater degree of amino acid identity than the N-termini and C-termini of the proteins. There was also a typical nuclear localization signal (NLS) in the N-terminal region (Fig. S1). The sequence analysis revealed that RcDREB2B possessed a high degree of sequence homology to AtDREB2B (30% identity) in Arabidopsis thaliana, TaDREB1 (36.49% identity) in Triticum aestivum, HvDFR1 (26.61% identity) in Hordeum vulgare, and ZmDREB2A (30.09% identity) in Zea mays.
Based on the amino acid sequences of RcDREB2B and other plants' DREB proteins, a phylogenetic tree was constructed. Phylogenetic analyses revealed that DREB proteins can be categorized into six sub-groups (A1–A6), among which RcDREB2B along with AtDREB2A, AtDREB2B, TaDREB1, ZmDREB2A, and HvDRF1 belong to the A2 sub-group (Fig. 5A). Moreover, RcDREB2B exhibited high similarity to AtDREB2A and AtDREB2B. Therefore, this gene was named RcDREB2B instead. Collectively, the data implies that RcDREB2B is an A2 sub-group member of the DREB subfamily.
Expression, nuclear localization, and transcriptional activity of RcDREB2B
We analyzed RcDREB2B expression in roots and leaves of the rose seedlings under normal drought stress (ND), medium drought stress (MD), and severe drought stress (SD) by real-time quantitative PCR (RT-qPCR). RcDREB2B was constitutively expressed in leaves tissues under the treatments we tested. In the root tissue, expression levels of RcDREB2B were significantly down-regulated under MD and SD treatment, in comparison to the ND controls (Fig. 5 B). These findings indicated that RcDREB2B might be involved in mediating drought stress signaling transduction.
Investigation of the subcellular localization of RcDREB2B was conducted by constructing GFP and RcDREB2B-GFP and transformed them with Arabidopsis protoplast cells. Fig. 5C shows that GFP fluorescence is equally divided in the nucleus and the cytoplasm with the control plasmid GFP, whereas the RcDREB2B-GFP fusion protein, was only located in the nucleus (Fig. 5C). A yeast assay system was used to estimate whether RcDREB2B was capable of activating transcription. Yeast strain Y2HGold was transfected individually with pGBKT7-RcDREB2B, pGAL4 (positive control), and pGBKT7 (negative control) plasmids. All transformed yeasts were able to grow in SD/Trp- media. The yeast transfected with pGBKT7-RcDREB2B and pGAL4 was able to grow on the SD/Trp-His-Ade-, however the negative control was not able to grow. Besides, blue color was observed upon incubation of the yeast extract of yeast transfected with pGBKT7-RcDREB2B and pGAL4 with X-gal (Fig. 5D). The outcome implies that RcDREB2B is a nuclear-localized transcriptional activator.
Overexpression of RcDREB2B reduced the high salinity tolerance in transgenic Arabidopsis
Driven by a constitutive super promoter , RcDREB2B was overexpressed in Arabidopsis for further functional characterization. The T3 generation of three independent RcDREB2B-overexpression lines (OE#4, 5, and 16) with different expression levels was subsequently chosen for analysis (Fig. 6A). No apparent morphological changes in terms of flower diameter, petal length, petal width, and single petal area were observed between VC and RcDREB2B-OE plants (Fig. S2). MS media with 0, 50, 100, 150, and 200 mM NaCl were used to sow seeds of the RcDREB2B-OE and VC (Fig. 6B). No significant morphological differences were observed in the rate of seed germination between RcDREB2B-OE and control lines showed when grown on 0, 50, and 100 mM NaCl plates. However, when supplemented with 150 and 200 mM NaCl, the germination rate of all plants reduced, accordingly. Upon supplementation with 150 mM NaCl, the germination rates of OE# 4, 5, and 16 were 52.8, 54.1, and 16.2 %, respectively, significantly lower than VC controls (91.7 %) (Fig. 6B, C).
To further analyze the root phenotypes of RcDREB2B-OE plants under salt stress, we placed 6-day-old seedlings in the MS media plates containing 0, 100, 150, and 200 mM NaCl, and treated for 10 days (Fig. 6C). The increment root length between RcDREB2B-OE and VC exhibited no significant variation when grown on 0 and 200 mM NaCl plates. But, at NaCl concentration of 100, or 150 mM, the VC displayed a significantly longer increment of root length than RcDREB2B-OE lines. For instance, when the growth was carried out on MS plates containing 150 mM NaCl, the increment in root length of the VC plants was 0.88 cm, while the increment in root lengths of the OE# 4, 5, and 16 was 0.70, 0.61, and 0.70 cm, respectively (Fig. 6E). Without NaCl treatment, there were no significant differences between RcDREB2B-OE lines and VC controls in terms of lateral root number. However, the lateral root number the same in RcDREB2B-OE lines was significantly more than VC in plants when exposed to the 150 and 200 mM NaCl condition (Fig. 6C, F).
Next, two distinct histochemical staining assays were carried out on Arabidopsis seedlings to detect O2− and H2O2 using DAB and NBT, respectively. Ten-day-old seedlings of RcDREB2B transgenic and VC plants were exposed to 200 mM NaCl or water (CK) solutions for 1 h. No significant differences in histochemical staining of seedlings leaves and roots were observed in both RcDREB2B-OE and VC plants under CK conditions (Fig. 6 G). Compared with VC, RcDREB2B-OE lines showed significantly stronger blue and brown color under 200 mM NaCl. Quantities analyses of H2O2 and O2− indicated RcDREB2B-OE exhibited significantly higher content than VC controls (Fig. S3). The results demonstrate that RcDREB2B-OE accumulated higher H2O2 and O2− levels compared with VC plants under salt stress, thus implying that RcDREB2B plays a negative role in controlling reactive oxygen species. These results demonstrate that overexpressing RcDREB2B in Arabidopsis decreased salt tolerance during the seedling stage as well as germination.
Overexpression of RcDREB2B leads to ABA sensitivity in Arabidopsis
To elucidate the involvement of RcDREB2B in ABA signaling, we conducted a phenotypic analysis of RcDREB2B-OE plants at the germination and root development stages in response to ABA (Fig. 7). Seeds were sown on MS medium supplemented with different concentrations of ABA (0, 1.2, and 1.6 µM) for 9 days (Fig. 7 A). In the absence of ABA, the germination rates of RcDREB2B-OE and VC plants were 100%. However, with 1.2 µM ABA, the germination rates decreased to 70.05%-77.22% for RcDREB2B-OE lines, and 91.98% for VC. Similar inhibition was observed in these plants grown on a 1.6 μM ABA plate, the germination rates of OE#4, OE#5, and OE#16 were 37.4%, 64.1%, and 61.5%, respectively, whereas VC germination rate was 90.0% (Fig. 7B). We also tested the root phenotype of RcDREB2B-OE and VC exposed to 0 and 100 μM ABA (Fig. 7 C). However, no significant differences were observed in the root phenotype of RcDREB2B-OE and VC seedlings in the absence of ABA. As shown in Fig. 7 D, the RcDREB2B-OE plants exhibited a 0.97-1.20 cm increment in root length under 100 μM ABA in comparison to 2.4 cm in VC controls. RcDREB2B transgenic plants had a significantly lesser lateral root number as compared to VC in the presence of 100 μM ABA (Fig. 7E). Such results suggest that RcDREB2B causes enhancement in ABA sensitivity of plants during germination and thereafter.
Overexpression of RcDREB2B exhibit sensitivity to inhibition of germination by osmotic stress in transgenic Arabidopsis
Since the sensitivity of plants towards ABA at the seed germination stage of RcDREB2B-OE has been increased, we predicted that overexpression of RcDREB2B may also influence plant tolerance to stress. In an attempt to investigate this, the seed germination of RcDREB2B-OE and VC controls under 0, 4, 8, 12, and 16% PEG treatment was analyzed (Fig. S4). RcDREB2B-OE and VC lines when grown on 0, 4, and 8% PEG plates showed no significant differences. The germination rate of RcDREB2B-OE and VC controls were significantly inhibited by 12 and 16% PEG, but the inhibition in the VC was less severe. When supplemented with 12% PEG, VC germination rates declined to ~92.28%, whereas RcDREB2B-OE lines displayed 75.6–78.6% germination (Fig. S4). Furthermore, RcDREB2B-OE lines displayed a 67.5–71.7% germination ratio in MS medium supplement with 16% PEG whereas VC germination was reduced to 93.0% (Fig. S4). It is evident from these results that RcDREB2B-OE seeds display a sensitivity to inhibition of germination upon stress due to osmosis.
Overexpression of RcDREB2B in Arabidopsis altered the expression of osmotic and ABA-responsive genes
To further explore the molecular mechanisms through which RcDREB2B controls the sensitivity of germination, RT-qPCR analysis was carried out to examine the expression pattern of four osmotic stress-responsive genes (AtDREB2A, AtWRKY33, and AtERF5, AtKIN1), seven ABA-responsive genes (AtABI2, AtKAI2, AtADH1, AtCHS, AtABF3, AtABF4, and AtICK1), one ABA biosynthesis gene (AtNCED3) in RcDREB2B-OE and VC plants under normal growth condition. Relative expression of AtDREB2A, AtWRKY33, AtERF5, AtKINI, AtABI2, AtKAI2, AtADH1, and AtNCED3 was significantly increased in the RcDREB2B-OE plants under normal conditions compared with VC controls. Besides, the chalcone synthase gene CHS,  which is classified as the stress-responsive biosynthetic genes, was up-regulated. But, two abscisic acid-responsive element-binding factors (AtABF3, AtABF4) exhibited a significant reduction under normal conditions (Fig. 8). The enhanced or attenuated expression of these genes in RcDREB2B might contribute to the ABA-induced sensitivity to inhibition of germination as well as salinity and drought stress during the period of seed germination and early seedling development.