Introduction

Petunia (Petunia ×hybrida) is one of the most popular ornamental crops worldwide in landscaping. In the United States, petunia is the top bedding plant, and its value of petunia at wholesale was estimated at $160 million in 2020 (USDA 2021). Even though petunias are cultivated under optimum environmental conditions in a greenhouse, their quality can be significantly diminished if not properly managed during the postproduction period. Water deficit is a major environmental factor that negatively affects plant growth and production, reducing aesthetic and economic value. Therefore, understanding the responses of petunia to water deficit stress and breeding petunia cultivars with enhanced water deficit tolerance are crucial for maintaining their quality and marketability.

Plants, being sessile organisms, must endure and adapt to various harsh environmental conditions to survive. To cope with adverse environments, plants have evolved sophisticated networks to perceive stress and regulate the expression of stress-responsive genes (Farooq et al. 2009). Transcription factors (TFs) act as crucial controllers in these networks by regulating the expression of downstream genes, consequently enhancing plant stress tolerance (Hadiarto and Tran 2011). Recently, a wide range of TF families relevant to water deficit stress response have been identified. Among the TFs, the APETALA 2/ethylene-responsive factor (AP2/ERF) family is a large gene family of plant-specific TFs involved in responses to water deficit, salt, high and cold temperature stresses and disease resistance (Mizoi et al. 2012).

The AP2/ERF family can be classified into four major subfamilies: AP2, ERF, related to abscisic acid and viviparous 1 (RAV), and dehydration-responsive element-binding proteins (DREBs) (Xie et al. 2019). Many stress-inducible ERF subfamily members have been isolated and characterized, suggesting that ERF members play diverse roles in plant adaptation mechanisms to both abiotic and biotic stresses (Li et al. 2018;Nie et al. 2018;Quan et al. 2010;Rong et al. 2014). For example, overexpression of wheat TaERF3 exhibited enhanced tolerance to salt and water deficit stresses by increasing proline accumulation and inducing stomatal closure, while TaERF3-silencing plants displayed more sensitivity to both stresses (Rong et al. 2014). A tomato ERF protein, TSRF1, improved osmotic and water deficit tolerance in rice seedlings by increasing the amount of proline and soluble sugars and inducing the expression of pathogenesis-related genes (Quan et al. 2010). In contrast, a cotton GhERF38 reduced seed germination rate, chlorophyll content, and survival rate under salt and water deficit stresses (Ma et al. 2017).

In petunia, time-course transcriptomic analysis revealed water stress-related genes, including those involved in redox homeostasis processes and hormone signal transduction, especially abscisic acid and ethylene (Park et al. 2021). One hundred seven TFs in 34 TF families were differentially expressed, and among them, the AP2/ERF family was the most abundant with eight upregulated AP2/ERF TFs under water deficit. This comprehensive analysis highlights the complex regulatory networks activated in response to water deficit stress. Here, we investigated the responses of PhERF039 to a water deficit stress. PhERF039 is a member of the AP2/ERF family, and differentially expressed at the early stage of water deficit. By manipulating the expression of PhERF039, we aimed to observe changes in petunia’s growth, development, and water deficit stress phenotypes. Understanding the role of PhERF039 in stress response mechanisms can provide valuable insights for developing strategies to enhance water deficit tolerance in this economically important ornamental plant.

Materials and Methods

Expression of PhERF039 under water deficit stress

Petunia ×hybrida ‘Mitchell Diploid’ (MD) seedlings were grown in 11-cm pots with soilless media (Sunshine^® Mix #1; Sun Gro Horticulture, Agawam, MA, USA) at a greenhouse (Morgantown, WV, USA). Nine-week-old petunias were randomly divided into two groups: control and water-stressed. Half plants were daily watered (control), while the other half received no water during water deficit treatment for up to five days (water-stressed). Each group had three biological replications (n = 3). The expression level of PhERF039 was analyzed using a Real-Time PCR. Total RNA was isolated with Trizol (Invitrogen, USA) and treated with DNase I (Thermo Fisher Scientific, Waltham, MA, USA). cDNA was prepared from 1 μg of total RNA with qScript^TM cDNA Synthesis Kit (Quanta Biosciences, Gaithersburg, MD, USA). The transcript levels of PhERF039 were quantified using QuantStudio^TM 3 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA) with PowerUp^TM SYBR^TM Green Master Mix (Thermo Fisher Scientific, Waltham, MA, USA). The amplification conditions were as follows: an initial incubation at 50°C for 2 m and 95°C for 2 m, followed by 40 cycles of 95°C for 15 s, 60°C for 15 s, and 72°C for 1 min. Primers used to determine PhERF039 expression were 5’-TGACCGTTATCAATACCCAACTT-3’ as a forward primer (F) and 5’-CCACTAATAGCACTCCCACAAA-3’ as a reverse primer (R). PhEF1α was used as an internal control and the primers were 5’-CCTGGTCAAATTGGAAACGG-3’ (F) and 5’-CAGATCGCCTGTCAATCTTGG-3’ (R). A melting curve (55 – 95°C with a heating rate of 0.15°C‧s^-1 and a continuous fluorescence measurement) was performed for each quantitative real-time PCR (qRT-PCR) reaction. The 2^-ΔCt method was used for the analysis. All the samples were measured in three technical replications. The transcript levels of PhERF039 in non-transgenic MD and a transgenic line were quantified as described above.

Cloning, construction of plasmid, and plant transformation

Using the total RNA extracted from the leaf tissue, the first strand cDNA containing the open reading frame (ORF) of PhERF039 was generated using a pair of primers: 5’-TGTGGTCTCAAATGGAAGATCATCATCTTCATCATG-3’ (F) and 5’-TGTGGTCTCAAAGCTAATAGCACTCCCACAAAAATG-3’ (R)(the underlined bases are for a restriction site of BsaI). The deduced amino acid sequences of PhERF039 were aligned with homologs from other plant species. The PCR fragment was digested by BsaI and ligated into the binary expression vector pICH86988 (Addgene plasmid 48076). The expression of the gene was driven by the CaMV35S promoter. This construct was transformed into P. ×hybrida ‘Mitchell Diploid’ leaf tissue using Agrobacterium (A. tumefaciens LBA4404)-mediated transformation according to the established protocol with modification (Conner et al. 2009). Positive transgenic plants were verified by using a primer set spanning the PhERF039 insert to identify the transgene. The verification primer set was 5’-GAGGACACGCTCGAGTATAA-3’ (F) and 5’-GTAAGGATCTGAGCTACACATG-3’ (R).

Plant growth conditions

Plants were blocked in two growth chambers with two plants per treatment. The average temperatures of one chamber were 20.7/18.6 ± 0.2/1.3°C day/night (mean ± SD) with a relative humidity of 65.8% ± 5.9%, photosynthetic photon flux density (PPFD) of 204 μmol·m^-2·s^-1 from 0600 to 2000 HR daily (daily light integral = 10.3 mol·m^-2·d^-1), and CO₂ at 470 ± 52 μmol·m^-2·s^-1 (mean ± SD). For the other chamber, the average temperatures were 20.5/18.6 ± 1.0/0.7°C day/night (mean ± SD) with a relative humidity of 67.6% ± 4.1%, PPFD of 208 μmol·m^-2·s^-1 from 0600 to 2000 HR daily, and CO₂ at 472 ± 60 μmol·m^-2·s^-1 (mean ± SD). Data were the means of measurements from two replications (n = 2).

Measurements of growth and development

Plant growth and development were evaluated in wild type (WT) and T₁ generation of the PhERF039 transgenic line at weeks 15 and 21 after sowing. The growth index (GI) was calculated using the following equation:

Relative chlorophyll content of fully expanded leaves (seventh or eighth leaf from the top) was measured using a SPAD-502 Chlorophyll Meter (Konica Minolta Sensing, Inc., Osaka, Japan). The flowering time was determined as the number of days from the sowing date until the first flower was fully opened. The number of flowers per plant was counted at weeks 24 and 28 after sowing. The flower size was determined as the average of a flower’s widest and perpendicular widths. The flower longevity was determined as the time from anthesis until the corolla was completely wilted (Chang et al. 2014). Five flowers from different plants were used for flower size and longevity measurements at week 21 after sowing.

Water deficit stress treatment and measurements

At week 22, wild type and the transgenic line of petunia were evaluated under the various levels of volumetric water content (θ): 0.25, 0.15, 0.10, or 0.05 m³·m^-3 in growing media. The moisture levels of the growing media were maintained by an automated irrigation system (Nemali and van Iersel 2006). The visual wilt status was measured daily based on the ratings described by Waterland et al. (2010a). Wilt status ratings were from 1 to 5, with 5 = completely turgid, 4 = soft to the touch but still upright, 3 = starting to wilt, 2 = severely wilted, and 1 = wilted to the point that leaves are dried and desiccated (Waterland et al. 2010a). Stomatal conductance (g_s) was measured daily with a portable photosynthesis system (LI-6400XT; LI-COR, Lincoln, NE). A leaf was clamped into an extended chamber with clear top and bottom covers (Extended Reach 1 cm Chamber LI6400-15; LI-COR). Environmental conditions in the chamber were set at 400 mmol·m^-2·s^-1 CO₂, and 25°C as the block temperature with a relative humidity 42.3% ± 2.0%. Readings were conducted from 1000 to 1400 HR.

Statistical analysis

The experiment design was a randomized complete block design. Plants were blocked in two growth chambers (n = 2), and each chamber had two plants per water deficit treatment (θ). Statistical analyses were performed using SAS (version 9.3; SAS Institute, Cary, NC). Differences between WT and the transgenic line were assessed using Student’s t-test at p ≤ 0.05.

Results and Discussion

Isolation and sequence analysis of PhERF039

In previous research, RNA sequencing was performed with petunia grown under water deficit conditions to identify the genes regulating water deficit stress responses (Park et al. 2021). The transcriptome revealed 6,679 unigenes differentially expressed under water deficit, and one member of the AP2/ERF family, PhERF039, was selected for functional analysis. The length of ORF of PhERF039 is 816 bp, and it encodes a predicted protein of 271 amino acids (30.2 KDa). The amino acid sequence alignment analysis showed that PhERF039 contains a highly conserved AP2 domain and a high degree of sequence homology to NtERF039-like in Nicotiana tabacum (tobacco, 71% identity), SlERF038-like in Solanum lycopersicum (tomato, 67% identity), CaERF034-like in Capsicum annuum (pepper, 62% identity), AtERF38 in Arabidopsis thaliana (52% identity), and GhERF38 in Gossypium hirsutum (cotton, 52% identity)(Fig. 1). Among them, GhERF38 were characterized under abiotic stresses (Ma et al. 2017).

Expression of PhERF039 under water deficit stress.

The temporal expression pattern of PhERF039 was analyzed by qRT-PCR. Petunias were exposed to stress by withholding water for five days, and the stressed plants reduced g_s five days after the water deficit (Park et al. 2021). The transcript level of PhERF039 was markedly elevated three days after the water deficit compared to that of the control (Fig. 2). There was no difference in PhERF039 transcript levels between the control and the stressed plants one and five days after the water deficit. The results of g_s indicated that plants perceived water deficit and induced stomatal closure to limit water loss and tolerate the stress. The increased transcript level of PhERF039 suggested the possible role of PhERF039 in the responses to water deficit stress. Övernäs (2010) reported that AtERF38 was induced under salt and cold stresses but not by abscisic acid (ABA) treatment, suggesting AtERF38 is involved in stress responses through an ABA-independent pathway. GhERF38 was upregulated under abiotic stresses (salt and water deficit) and by ABA treatment (Ma et al. 2017).

Expression level of PhERF039 in transgenic plants

To analyze the function of PhERF039, Agrobacteriumediated transformation was used to overexpress PhERF039 in petunia. The gene expression was driven by CaMV 35S promoter. One transgenic line (T₀) was obtained, and the integration of PhERF039 was confirmed by PCR (Fig. 3). The T₁ seeds of the transgenic line were sown, and 17 T₁ plants were verified to have the integration of PhERF039 out of 24 T₁ plants (70.8%). Sixteen T₁ plants with the PhERF039 integration were used for this experiment (Fig. 3). The introduction of PhERF039 was expected to increase its expression level; however, the expression of PhERF039 was reduced in T₁ generations compared to wild-type (WT) (Fig. 4). It is suspected that the expression of the transgene (PhERF039) degraded mRNA of both endogenous and the transgene. This phenomenon has been well documented as RNA interference (RNAi) (Hannon 2002). In petunias, white flowers and/or a mixture of white and purple flowers were produced by introducing a chimeric chalcone synthase expected to produce darker purple flowers (Napoli et al. 1990). Later, the molecular mechanism revealed that the double-stranded RNAs (dsRNAs) from transgene led to the destruction of mRNAs containing similar sequences, and such dsRNA-mediated silencing is employed for functional analysis of target genes (Wilson and Doudna 2013). Instead of overexpressing, an under-expression line of PhERF039 transgenic plants (UE) was obtained for functional analysis.

Effects of under-expression of PhERF039 on plant growth and development

The UE produced more rosette branches than the WT (Fig. 5A and B). After 21 weeks of growth, WT plants displayed approximately six branches, while the UE produced approximately nine branches (p = 0.0002) (Fig. 5B). Growth index (GI) of transgenic plants was smaller than that of WT on week 15 (p < 0.0001) and 21 (p < 0.0001) (Fig. 5C). The leaf chlorophyll concentrations of the transgenic plants significantly increased compared to WT on week 15 and 21 (p < 0.0001 and p = 0.0048, respectively) (Fig. 5D). The UE plants exhibited delayed flowering by 20 days compared to WT (p < 0.0001) (Fig. 6A). The number of flowers were significantly higher in UE than WT on week 24 (p < 0.0433), but the difference became insignificant on week 28 (Fig. 6B). The flower size and longevity were not different between WT and UE (Fig. 6C and D).

Wild-type plants displayed strong apical dominance, showing high canopy height and larger GI, while UE plants produced more lateral branches. The shoot morphology of UE suggested that PhERF039 could inhibit the development of shoot branching. It has been known that plant hormones, auxin and cytokinins (Burbidge et al. 1999), regulate bud outgrowth. Auxin induces apical dominance and inhibits shoot branching, while CKs antagonize auxin in apical dominance by triggering outgrowth (Müller and Leyser 2011). Some studies have revealed that the AP2/ERF is involved in plant shoot morphology, possibly through the dynamic regulation of CKs homeostasis and responses (Gupta and Rashotte 2014;Raines et al. 2016). Haver et al. (2002) reported that the exposure of ethylene inhibited apical dominance and promoted the outgrowth of lateral shoots by decreasing the auxin/cytokinin ratio in P. ×hybrida ‘Orchid’. These results indicated that PhERF039 might play the opposite role to ethylene effects regulating shoot branching through auxin and CK signaling. The relative chlorophyll content was increased in UE, suggesting that PhERF039 might be involved in triggering leaf senescence through chlorophyll degradation and its impaired synthesis (Iqbal et al. 2017). The UE also delayed flowering compared to WT. Flowering is regulated by complex gene regulation networks that integrate multiple environmental conditions, endogenous signals, and hormonal signaling (Campos-Rivero et al. 2017). The effects of ethylene in the regulation of flower development are complicated to investigate. The mutation of CTR1, a key negative regulator in ethylene signaling, delayed flowering, suggesting that ethylene inhibited flowering in Arabidopsis (Achard et al. 2007). However, ethylene-induced a reproductive transition in rice (Wuriyanghan et al. 2009). The number of flowers was also affected by ethylene, but the sensitivity to ethylene on flower number was unclear, and neither was the flowering time (Iqbal et al. 2017). Plant hormone networks and their cross-talk should be considered in future research to identify the functional role of PhERF039 in growth and flowering time.

Effects of the under-expression line of PhERF039 on water deficit tolerance

Twenty-one-week-old plants were subjected to water treatments for seven days with various θs: 0.25, 0.15, 0.10, or 0.05 m³·m^-3. The WT plants seemed to reach the set θ points earlier than UE plants, approximately 1 to 2 days earlier (Fig. 7). Both WT and UE at 0.25 and 0.15 m³·m^-3 maintained a visual wilt status rating of 5 (Fig. 8A and B). The wilt status values were not significantly different between WT and UE at all θs (Fig. 8). However, UE plants had lower g_s than WT plants up to three days after treatment at all treatment conditions (Fig. 9). When plants were growing in the drier media (θs at 0.01 and 0.05 m³·m^-3), UE plants maintained lower stomatal conductance up to five days after treatment (Fig. 9C and D). Overall, the under-expression line of PhERF039 exhibited lower stomatal conductance than WT, indicating that PhERF039 might be negatively involved in stomatal closure.

Regression analyses were performed using the data obtained from wilt status (Fig. 8D) and g_s (Fig. 9D) measured at the lowest volumetric water content treatment (0.05 m³·m^-3) in Fig. 7. Plant wilt status was positively correlated with θ. There is no distinct difference in slopes and intercepts of the two regression models between WT and UE (p = 0.7554 and 0.5258, respectively) (Fig. 10). When subject to water deficit stress, both WT and UE exhibited a decrease in g_s as volumetric water content in the media gradually decreased to the set point at 0.05 m³·m^-3 (Fig. 11). The slope of WT was higher than that of UE (p < 0.0001) (Fig. 11). According to the regression models of θ in substrate and wilt status, UE did not increase the wilt status compared to WT (Fig. 10). In contrast, stomatal conductance was decreased in UE (Fig. 11). These results indicated that UE plants did not delay plant wilting compared to WT plants. However, the volumetric water content of the media for UE reached the set threshold later than that for WT. In general, UE plants maintained lower g_s than WT (Fig. 11). Plants lose about 97% of total water absorbed from roots through transpiration. The transpirational rate is determined largely by g_s (Li et al. 2018). Stomatal closure or blocking stomata has been shown to enhance water deficit tolerance by limiting transpirational water loss (Park et al. 2016;Waterland et al. 2010a;Waterland et al. 2010b). In this study, stomatal closure observed in UE appeared insufficient to delay wilting symptoms.

Overall, downexpression of PhERF039 delayed plant wilting, exhibited low stomatal conductance, and reached the set volumetric water content of soilless media slowly. Overexpression of GhERF38, a homolog of PhERF039, reduced salt and water deficit stress tolerance due to ABA-insensitive stomatal movements and large stomatal aperture (Ma et al. 2017). This result indicated that PhERF039 could play a negative role in stomatal closing and water deficit tolerance. The low stomatal conductance might result from induction of stomatal closure and/or low stomatal density. Arabidopsis applied with ethylene precursor (ACC) increased stomatal density, while impaired ethylene signaling pathways reduced stomatal density, suggesting ethylene plays a positive role in stomatal development (Serna and Fenoll 2001). Suppression of PhERF039 also promoted outgrowth of branching and delayed flowering. Further research with additional lines of over- and under-expression lines is needed to identify the exact roles of PhERF039 in stomatal closure, density and any other phenotypes.

Conclusion

This study characterized the petunia mutant, PhERF039, a gene encoding AP2/ERF, under water deficit stress. Transgenic plants, under-expression line of PhERF039, promoted growth of the axillary shoots and delayed flowering time. Down-regulation of PhERF039 delayed substrate drying by limiting transpirational water loss with decreased g_s, suggesting its negative role in stomatal closure. However, there was no visual phenotype in relation to the water deficit stress. It can’t be ruled out the possibility that the effect of down-regulation of PhERF039 could have been compensated by the gene redundancy and/or functional redundancy. Further research with multiple transgenic lines (overand under-expression) is needed to characterize the function of PhERF039 on stress responses, hormone networks, stomatal development and closure, and plant morphology.