Ethylene responsive element binding factors contributes to waterlogging tolerance by regulating photosynthetic and physiological parameters in petunia

ABSTRACT: Ethylene-responsive element binding factors (ERFs) are widely involved in the regulation of plant responses to different abiotic stresses. In petunia (Petunia × hybrida), PhERF2 belonging to the subfamily Ⅶ of ERF transcription factors participates in the response to waterlogging stress. In this study, we investigated waterlogging tolerance variation of WT and transgenic petunia plants with RNAi silencing and overexpression of PhERF2 through photosynthetic and physiological performance. Chlorophyll content and root vigor declined continuously in both WT and PhERF2 transgenic lines under waterlogging stress, but the extent of the fall in PhERF2-overexpressing lines was less than that in WT and PhERF2-RNAi lines. At the end of waterlogging treatment, soluble protein levels in PhERF2-overexpressing lines were significantly higher than those in WT and PhERF2-RNAi lines, while the latter showed a higher malondialdehyde content overall. Different degrees of reductions in Pn, Gs, and Tr levels occurred in both WT and PhERF2 transgenic lines upon exposure to waterlogging. The Ci levels of PhERF2-overexpressing lines decreased after 3 hours of waterlogging treatment, and the Ci levels of WT and PhERF2-RNAi lines gradually increased from 6 to 72 hours of waterlogging treatment. These data suggested that non-stomatal factors were the primary limiting factors for Pn in WT and PhERF2-RNAi lines under severe stress, while the stomatal opening was the main factor limiting Pn in PhERF2-overexpressing lines. Our results demonstrated that the contribution of PhERF2 to the waterlogging tolerance of petunia appears to depend on the regulation of physiological and photosynthetic responses. PhERF2 represents a hopeful candidate gene for enhancing waterlogging tolerance of ornamental plants.


INTRODUCTION
Petunia, cultivated all over the world, is one of the most important Solanaceae plants utilized for ornamental purpose (GERATS & STROMMER, 2011).It is of great significance in global floriculture production, and usually planted in outdoor gardens in many countries (XIA et al., 2006).Waterlogging has become one of the main abiotic stresses restricting common planting and landscape cultivation of Ciência Rural, v.53, n.5, 2023.Yin et al. petunia.It is urgent to discover the wild excellent germplasm with stress resistance for developing antistress breeding by molecular technology.
The ethylene-responsive element binding factors (ERFs) are essential transcriptional regulators in response to different biotic and abiotic stresses in plants.ERFs belong to the plant-specific AP2/ERF transcription factor superfamily, which has great potential in transgenic breeding (DAOYANG et al., 2016).Previous research has reported the biological role of ERFs in the modulation of waterlogging stress.Overexpression of an ERF gene Sub1A in rice leads to the enhancement of waterlogging tolerance (FUKAO et al., 2006;XU et al., 2006;FUKAO & BAILEY-SERRES, 2008).HRE1, a hypoxiainducible ERF gene, confers waterlogging tolerance in transgenic Arabidopsis plants (LICAUSI et al., 2010).Petunia is greatly susceptible to waterlogging; and is therefore, a promising landscape flower for studying the waterlogging responses.We have reported that PhERF2 plays an important role in antiviral RNA silencing (DAOYANG et al., 2016), and the expression levels of PhERF2 were significantly increased by abiotic stress such as cold, salt, and waterlogging treatments.
The changes of photosynthetic parameters Photosynthetic rate (Pn), Stomatal conductance (Gs), Transpiration rate (Tr), and Intercellular carbon dioxide concentration (Ci) reflect the effects of waterlogging stress on photosynthetic characteristics of plants.Previous research has indicated that, compared with waterlogging-sensitive plants, photosynthesis was less affected in the tolerant ones.The native plants in wetland show better adaption to flooding without restraint on photosynthesis (PEREIRA et al., 2014;WARING & MARICLE, 2012;KUMAR et al., 2013;SAGREDO et al., 2014;ZHU et al., 2016;STRIKER & COLMER, 2017).The decrease in stomatal conductance (Gs) leads to the reduction in CO 2 availability, which is the primary negative effect of waterlogging on Pn during short term waterlogging stress (IRFAN et al., 2010;HERZOG et al., 2016;PÉREZ-JIMÉNEZ et al., 2018).After long-term waterlogging stress, lowered CO 2 assimilation, decreased chlorophyll content, and excessive ROS production lead to non-stomatal inhibition on photosynthesis, which is the main limiting conditions for Pn (ZHENG et al., 2009).Photosynthesis is increasingly suppressed by the changes in leaf wilting and sagging, and the decrease in chlorophyll levels (SMETHURST et al., 2005).
Nowadays, it has been widely characterized that waterlogging stress increases the ROS production and leaf lipid peroxidation in various plants (XIAO et al., 2016;ZHANG et al., 2016).The final product of membrane lipid peroxidation is malondialdehyde (MDA), that combines with the specific proteins on the cell membrane to inactivate or destroy the structure and function of biofilm (SONG, 2012).Soluble protein focuses on the ability of mediating osmotic adjustment, stabilizing subcellular structures, and scavenging free radicals (DELAUNEY & VERMA, 1993).Under waterlogging stress, soluble protein as an osmolyte functions to maintain plant water status and hydraulic conductivity (BARICKMAN et al., 2019).When soil waterlogging takes place, the root system is directly affected by the stress.Oxygen deprivation in the root severely inhibits the growth and development of plants due to the degradation of chlorophyll (DREW, 1997).According to previous studies, we reported that PhERF2 plays an essential role in petunia anoxia tolerance (YIN et al., 2019).The regulatory mechanism of waterlogging response by PhERF2 requires a further dissection in the aspects of photosynthesis and physiology.Thus, in this study, wild-type (WT) and transgenic petunia plants with RNAi silencing and overexpression of PhERF2 were selected as experimental materials to measure the levels of leaf photosynthetic parameters, chlorophyll, MDA, soluble protein, and root vigor under waterlogging stress.Our results supported an essential role of PhERF2 in waterlogging tolerance of petunia.

Plant materials and treatments
Petunia × hybrida seeds of wild-type (WT) 'Mitchell Diploid', PhERF2-RNAi (1A, 1B, and 4B) and PhERF2-overexpressing (C, D, and I) lines were harvested in University of California Davis.Three experimental materials were taken for PhERF2 overexpression and PhERF2-RNAi respectively, in order to avoid chance, reduce experimental errors, and make experimental results more accurate.Seeds were sown in the botanical garden of Shanghai Institute of Technology.Two weeks post germination, the seedlings were transferred into plastic pots (12cm in height, and 10.5cm in diameter) filled with soil mixture (vermiculite: peat: perlite, 1:1:1) and grown in an artificial climatic room (50~60% relative humidity, 23/19 ℃ day/night, and 12/12h light/dark).
Eight weeks later, uniform petunia plants were selected for waterlogging treatments to examine the impact of waterlogging on WT and homozygous T2 generation transgenic petunia lines.
Potted plants were put in plastic containers, and the containers were filled with tap water to keep the water levels at 2-3 cm above soil surface during the whole waterlogging treatments.WT, PhERF2-RNAi, PhERF2-overexpressing lines and the seven waterlogging treatments (0, 3, 6, 12, 24, 48 and 72h of waterlogging) were respectively arranged as subplot and main-plot.All treatments were repeated 3 times and 6 plants were used for each replication; and therefore, there were 18 plants for each treatment.The control plants were untreated with waterlogging.We hypothesized that PhERF2 plays an essential role in improving petunia waterlogging tolerance.To test the hypothesis, the leaf and root samples were collected for measuring photosynthetic parameters and physiological indicators at 0, 3, 6, 12, 24, 48, and 72h after waterlogging stress.

Measurement of photosynthesis and related parameters
Net photosynthetic rate (Pn), stomatal conductance (Gs), transpiration rate (Tr) and intercellular CO 2 concentration (Ci) were measured on the expanded leaves by the LI-6400XT portable photosynthesis system (LI-COR, Lincoln, NE, USA).CO 2 concentration, actinic light intensity and temperature were set at 400μmol mol -1 , 1000μmol m -2 s -1 and 25 ℃, respectively.For each treatment, the values of Pn, Gs, Tr and Ci were recorded for six leaves at each time point.

Measurement of chlorophyll and soluble protein contents
Approximately 0.2 g of fresh leaf samples were sliced and incubated in 95% ethanol in the dark for 24h at 25 ℃, until the tissues became white.The absorbance of the extract was measured at 663 and 645nm for calculating chlorophyll content (ARNON, 1949).Soluble protein content of leaf was measured following SHI et al. ( 2007), bovine serum albumin was used as a standard.Specific enzyme activities were expressed as units (U) per g of protein.

Measurement of malondialdehyde
The levels of lipid peroxidation were determined in terms of MDA content according to thiobarbituric acid (TBA) reaction method (KUN et al., 2010).0.5g of fresh leaf tissues were ground by ice rubbing and homogenized in 5 ml 5% trichloroacetic acid (TCA).After centrifugation at 4000×g for 10 min, 2 ml of supernatant was dissolved in 2 ml of 0.6% TBA.The mixture was heated to 99 ℃ for 30 min and then cooled quickly on ice.After that, the mixture was centrifuged at 3000×g for 15 min, and then the supernatant was used to measure the absorbance at 600, 532, and 450nm.

Measurement of root vigor
About 0.1g of root samples were harvested at the end of each waterlogging treatment, and extracted following the method as previously described (MIN & DE -XIAN, 2011).The material was homogenized in 5 ml of 0.4% TTC and 5 ml of 0.1mol/L phosphate buffer solution (pH 7.5), and heated to 37 ℃ for 1.5h in darkness.The reaction was ended with 2 ml of 1mol/L sulphuric acid.After that, the mixture was ground and homogenized in 10 ml of ethyl acetate.After precipitation, the absorbance of the supernatant was measured at 485nm.All the spectrophotometric analyses above were conducted on a 752 UV-Vis spectrophotometer ("GUANG PU", Shanghai, China), and all the chemicals were purchased from Sangon Biotech.

Statistical analysis
Statistical analysis was conducted by the Excel2007 and SPSS17.0, and graphs were drawn by Origin 2017.Data were expressed as means ± errors.One-way ANOVAs followed by Duncan's multiple range test was employed to assess the statistical significance of the treatment difference (P ≤ 0.05).

Photosynthetic parameters
After 3h of waterlogging stress, Pn in WT and PhERF2-RNAi (1A, 1B, and 4B) lines reached the highest point, increasing to 1.2-, 1.4-, 1.1-, and 1.3-fold of the control levels, respectively.After that, Pn declined continuously by prolonged periods of waterlogging stress.In PhERF2-overexpressing (C, D, and I) lines, Pn was constantly reduced during the whole period of waterlogging treatment.Both at 12 and 48h, Pn in overexpressing lines was obviously higher than that in silenced lines.On day 3, a reduction in Pn was reported in both WT and transgenic petunia lines, but the extent of the decrease in PhERF2overexpressing lines was less than that in WT and PhERF2-RNAi lines, decreasing to 56.28%, 62.02%, 71.73%, 1.89%, 41.2%, 49.13%, and 48.83% of the background levels, respectively.Meanwhile, Pn in overexpressing lines was significantly higher than that in the other lines, and the value of line I was the highest of others (Figure 1a).
As shown in figure 1b, Gs in PhERF2-RNAi (1A, 1B, and 4B) lines continued to decrease Ciência Rural, v.53, n.5, 2023.Yin et al. with the prolonged waterlogging treatment.Gs in WT and PhERF2-overexpressing (D and I) lines reached a peak at 3h of waterlogging, and the respective increases were 1.8-, 1.2-, and 1.6-fold of their non-waterlogged controls.When waterlogging stress continued for 48h, Gs in WT and PhERF2-RNAi (1B, 4B) lines was remarkably higher than that in PhERF2-overexpressing (C and D) lines.After 3 days of waterlogging stress, compared with the control levels, Gs in WT, PhERF2-RNAi (1A, 1B, and 4B) and PhERF2-overexpressing (C, D, and I) lines dropped to 31.87%, 29.73%, 36.69%,30.12%, 37.51%, 31.51%, and 42.18%, respectively, and meanwhile the degree of the fall in line I was the least.
As displayed in figure 1d, the Ci of PhERF2-RNAi (1A, 1B, and 4B) lines dropped to the lowest point in response to 3h of waterlogging.After that, there was a progressive increase of Ci, by day 3, Ci in PhERF2 silencing lines peaked at 1.4-, 2.4-, and 2.2-fold of the initial value, respectively.With the prolongation of waterlogging treatment time, the Ci of PhERF2-overexpressing (C, D, and I) lines continued to decrease, and dropped to 65.01%, 66.82%, and 83.95% of the control levels on day 3.For 6, 48, and 72h of waterlogging stress, the Ci of WT and PhERF2-RNAi lines were significantly higher than that of PhERF2-overexpressing lines.

Leaf chlorophyll and MDA content
As shown in figure 2a, on the whole, chlorophyll content was followed by a progressive fall in both WT and PhERF2 transgenic lines with the extension of waterlogging treatment.After 3 days of waterlogging stress, chlorophyll content of WT plants, lines 1A, 1B, 4B, C, D, and I decreased to 57.58%, 75.95%, 63.35%, 64.91%, 75.46%, 58.21%, and 64.57% of the non-waterlogged levels, respectively.Meanwhile, the chlorophyll content in PhERF2overexpressing (C, D, and I) lines was significantly higher than that in WT and PhERF2-RNAi (1A, 1B, and 4B) lines.Particularly, the value of line I was the highest and line 4B was the lowest.
Under waterlogging stress, the MDA content remarkably increased followed by a decline (Figure 2b).By 6h of waterlogging, MDA content in WT, PhERF2-RNAi and -overexpressing lines peaked with 4.6-, 4.1-, 4.4-, 5.4-, 1.3-, 1.1-, and 1.4fold of the non-waterlogged controls, respectively.It was substantially higher in WT and RNAi silencing lines (1A, 1B, and 4B) than in overexpressing lines (C, D, and I) of PhERF2, and thereafter, MDA content decreased sharply.At 24 and 48h of waterlogging stress, the content of MDA in WT and PhERF2-RNAi lines was significantly higher than that in PhERF2-overexpressing lines as well.After 72h of waterlogging, in WT and RNAi silencing lines, MDA content rose to 2.3-, 1.8-, 1.7-, and 3.3-fold, while in PhERF2-overexpressing lines, the equivalent levels decreased to 33.33%, 45.1%, and 58.5%, respectively, compared to corresponding unstressed control.

Soluble protein content and root vigor
The soluble protein content of PhERF2overexpressing lines was significantly higher than that of WT and PhERF2-RNAi lines during the waterlogging period from 24h to72h.Compared with other lines, silenced line 4B maintained the lowest value, and conversely, overexpressing line I stayed at the highest level.By 24h of waterlogging, soluble protein content in line I was 0.37 mg/g FW, a level about twice than that in line 4B (0.19 mg/g FW).After 48h of exposure to waterlogging stress, line I (0.55 mg/g FW) was 2.4 times that of line 4B (0.23 mg/g FW).At the end, line I was 0.2 mg/g FW; and however, line 4B decreased sharply to approach the zero (Figure 3a).
With the prolongation of waterlogging treatment time, the root vigor of WT, PhERF2-RNAi (1A, 1B, and 4B) and PhERF2 overexpressing (C, D, and I) lines continued to decrease.At 72h of waterlogging treatment, root vigor in overexpressing line I was 967.2 µg TTC/g•h, which was about twice the average of the three silenced lines (488.8 µg TTC/g•h), By comparison with control lines without waterlogging stress, root vigor was decreased by 25.8% in WT plants, by 26.4%, 31.2%, and 25.1% in three silencing lines, and by 26.5%, 31.6%, and 45.8% in three overexpressing lines in response to 72h of waterlogging.In particular, the decrease of overexpressing lines was smaller than that of PhERF2 silenced lines, among which line I had the smallest decrease and line 4B had the largest decrease (Figure 3b).

DISCUSSION
There is a large demand for petunia with the development of the landscaping industry.The ability of a plant to maintain healthy under waterlogging, drought, and salt stresses is a criterion for the choice of these horticultural plants.Sometimes, the landscaping plants with elite traits have been developed in conventional breeding programs, while these plants do not express as good phenotypes as expected.The wide application of molecular technology together with the discovery of the ERF transcription factors has provided an alternative way to improve stress tolerance for the landscaping plants.In this study, we confirmed the function of PhERF2 in waterlogging tolerance using WT and its transgenic plants.Under waterlogging stress, hypoxia is the primary influencing factor (SHIONO et al., 2008).When tissues are anoxic, the aerobic energygenerating system declines rapidly and then affects the roots directly (ELSE et al., 2001).We reported that root vigor in WT and PhERF2-overexpressing lines was greater than that in PhERF2-RNAi lines under waterlogging treatment (Figure 3b).The result is consistent with a previous report in wheat (CHEN et al., 1998).Some scholars have shown a coordination between above-and below-ground parts of plant, involving signal transduction from the root to Ciência Rural, v.53, n.5, 2023.Yin et al. the shoot and subsequent physiological responses.GAO et al. ( 2018) reported that both salt and drought treatments induced oxidative stress response in Chrysanthemum morifolium 'White Snow' leaves.
When plants are exposed to waterlogging stress, oxidative stress will induce a large amount of toxic active oxygen which causes damages to chlorophyll, proteins, membrane lipids, and subcellular organelles (MUNNÉ-BOSCH & ALEGRE, 2000;RONG-HUA et al., 2006).After waterlogging treatment, chloroplast ultrastructure is damaged, resulting in a decrease of chlorophyll content (AHMED et al., 2002;KUMUTHA et al., 2008).For example, leaf chlorophyll content in both waterlogging-tolerant and -susceptible plants of pigeon pea decreases under waterlogging (KUMUTHA et al., 2009).In the current study, chlorophyll content declined in both WT and all transgenic lines, but the chlorophyll levels were significantly higher in PhERF2-overexpressing lines than in PhERF2-RNAi lines at the final stage of waterlogging stress (Figure 2a).It suggested that these plants probably exerted the PhERF2-regulated pathways to make overexpressing lines to suffer less from flooding stress than RNAi silencing lines.Furthermore, the changes in MDA and soluble protein content also concur with the physiological traits for waterlogging resistance (YIN et al., 2009;YIN et al., 2010).The greater accumulation of ROS in waterlogged tissues elevated the extent of lipid peroxidation.MDA level representing the degree of lipid peroxidation is a classic parameter to reflect oxidative injury (WU et al., 2003;YAN et al., 2013).In our study, MDA content in WT and three silencing lines was higher than three overexpressing lines throughout the waterlogging period (Figure 2b), indicating that overexpression of PhERF2 in plants seems to reduce oxidative injury.The synthesis and degradation of soluble proteins are associated with the responses of plants to stressful conditions (GUAN et al., 2010).The results from our studies showed that soluble protein content increased first followed by a remarkable reduction.During the waterlogging phases from 24h to 72h, the content of soluble protein in PhERF2-overexpressing lines was remarkably higher than in WT and PhERF2-RNAi lines (Figure 3a).These results of physiological indicators indicated that transgenic petunia plants overexpressing PhERF2 were more tolerant to waterlogging stress than WT and PhERF2-silenced plants, which is similar to the findings in transgenic tobacco plants with overexpression of the cotton CBF gene GbCBF1 (GUO et al., 2011).
The decrease in Pn demonstrates the vulnerability of photosynthesis in Jerusalem artichoke under waterlogging.According to the collective decline of Ci, Gs and Tr, the waterlogging stress forced the stomata of plant leaves to close, Gs decreased, and the ability of leaves to absorb CO2 decreased, resulting in a decrease in Ci, which in turn led to a decrease in Pn, which further led to the stomatal limitation of photosynthesis.(YAN et al., 2018).In PhERF2overexpressing lines, Ci, Gs, and Tr all decreased after 6h of waterlogging treatment, compared with their control levels (Figure 1), indicating that Pn was mainly affected by limited leaf gas exchange.The decrease in stomatal conductance is the primary limiting factor for Pn, while non-stomatal limitation can also affect Pn under severe waterlogging stress, such as the decreased chlorophyll content and increased MDA levels.Severe waterlogging causes a large amount of ROS to accumulate in plants and damage the chloroplasts, resulting in a decrease in the activity of photosynthetic enzymes, a decrease in the ability of plants to carry out photosynthesis, and a decrease in the assimilation ability of mesophyll cells to CO2, which directly leads to a decrease in Pn (ZHENG et al., 2009).In WT and PhERF2-RNAi lines, Gs and Tr decreased but Ci gradually increased during the periods from 6h to 72h, compared to the control plants without waterlogging (Figure 1), suggesting that non-stomatal limitation factors are the major factors for Pn.Previous studies have reported that stomatal closure would be helpful to keep water status in leaves by reducing the water loss from transpiration (TAKAHASHI & MURATA, 2008;GILL & TUTEJA, 2010).In both WT and transgenic lines, Gs declined continuously during the whole experiment, while Gs in WT and PhERF2-RNAi lines was higher than that in PhERF2-overexpressing lines at the late stages of waterlogging (Figure 1b), implying that overexpression of PhERF2 would make these transgenic plants more tolerant to waterlogging stress.

CONCLUSION
In summary, a set of photosynthetic and physiological parameters were variable in PhERF2silenced and -overexpressing transgenic petunia plants under waterlogging stress, compared to WT control.The overexpressing lines exhibited a better maintenance of chlorophyll content and root vigor than the WT and silenced lines upon waterlogging treatment.The waterlogging resulted in a reduction in Pn, Gs, and Tr levels in all lines, but an increase in Ci levels in PhERF2-silenced lines.Our data support an important positive role of PhERF2 in the modulation of waterlogging tolerance in petunia.