Ethylene and polyamines interact in rice spikelet degeneration in response to water stress during meiosis

This study aimed to investigate whether an interaction existed between ethylene and polyamines (PAs), which mediated the effects of water stress (WS) during meiosis on rice (Oryza sativa L.) spikelet degeneration. Two indica hybrid rice cultivars, HY113 (a drought-resistant rice cultivar) and ZZY8 (a drought-susceptible rice cultivar), were used in the present study, which were exposed to two soil moisture treatments: well water (WW) and WS. The spikelet degeneration rate (SDR) in WS plants increased by 12.1% for HY113 and 29.5% for ZZY8 compared with WW plants. The concentration of free spermidine (Spd) and free spermine (Spm) and the activities of S-adenosyl-l-methionine decarboxylase and Spd synthase were all significantly reduced in young panicles by the WS, with more reduction for ZZY8 than for HY113. The ethylene evolution rate, the concentrations of putrescine (Put), 1-aminocylopropane-1-car-boxylic acid (ACC), and H2O2, and the activities of ACC synthase and ACC oxidase in young panicles significantly increased for both cultivars under WS treatment, with more increase for ZZY8 than for HY113. Furthermore, when Spd or aminoethoxyvinylglycine (an inhibitor of ethylene synthesis) was applied to the young panicles of WS plants in the pollen mother cell meiosis prophase, the SDR significantly reduced. An opposite effect was observed when ACC or methylglyoxal-bis(guanylhydrazone) (an inhibitor of Spd and Spm synthesis) was applied to young panicles of WS plants. The results of the present study suggested the existence of an antagonistic interaction between ethylene and PA biosynthesis, which likely mediated the effect of WS during meiosis on rice spikelet degeneration.


Introduction
Water stress (WS) has been considered as one of the most important abiotic stresses seriously affecting crop growth and development and ultimately severely affecting crop yields (Atkinson and Urwin 2012;Barnabas et al. 2008;Basu et al. 2023;Farooq et al. 2009).In the last 40 years, WS has reduced crop production by 10% annually on average (van Ittersum et al. 2013).Moreover, it is predicted that WS will be more severe in the next 10 years due to the rapid climate change (Boyer and Westgate. 2004).Therefore, it is essential to understand whether and how field crops respond to WS for drought-resistant crop breeding in the future.
Rice (Oryza sativa L) is the most important paddy crop sensitive to WS (Lanceras et al. 2004;Luo 2010;Nguyen et al. 1997).The growth stage of meiosis in rice was the most stress-sensitive period, which could cause spikelet degeneration and pollen sterility, and finally reduce the grain yield (Yang et al. 2007; Barnabas et al. 2008;Kato et al. 2008).Although several previous studies reported that the most sensitive growth period to WS of rice was meiosis (Yang et al. 2007;Zhang et al. 2020), the underlying physiological and biochemical mechanisms are yet to be understood.
Polyamines (PAs), including putrescine (Put), spermidine (Spd), and spermine (Spm), and ethylene are generally accepted as two important endogenous plant growth regulators in higher plants; they play critical roles in response to different abiotic stresses and regulate several physiological and biochemical processes (Hu et al. 2020;Hussain et al. 2011;Kusano et al. 2008;Alcazar et al. 2010;Kazan 2015;Lin et al. 2009;Wilkinson and Davies 2010;Yang et al. 2007).The reduction in rice inferior grain weight had a significant negative correlation with the significant increase in the ethylene evolution rate and the 1-aminocylopropane-1-carboxylic acid (ACC, the precursor of ethylene) concentration (Chen et al. 2013;Yang et al. 2004Yang et al. , 2007)).Furthermore, the application of aminoethoxyvinylglycine (AVG), an ACC synthase inhibitor, to rice plants significantly increased the inferior grain weight (Chen et al. 2013;Yang et al. 2004Yang et al. , 2007)).Previous studies indicated that the greater concentration of free Spd/Spm in inferior spikelets significantly positively correlated with the higher inferior grain filling rate and grain weight (Wang et al. 2012;Chen et al. 2013;Yang et al. 2008).As ethylene and PAs share the same synthetic precursor (S-adenosyl-L-methionine, SAM), numerous previous studies indicated that a significant interaction existed between ethylene and PAs in response to different abiotic stress and regulated plant growth and development (Wang et al. 2012;Yang et al. 2014;Zhang et al. 2017;Feng et al. 2011).However, prior to this study, few studies focused on whether a significant interaction was present between ethylene and PAs, which affected rice spikelet degeneration.
The present study investigated the hypothesis that an interaction existed between ethylene and PAs, which might mediate the effects of WS during meiosis on spikelet degeneration in rice.In the present study, two cultivars with apparent differences in drought resistance were used to investigate the synchronous changes in the PAs concentration and the ethylene evolution rate in young rice panicles subjected to WS during meiosis.Furthermore, chemical regulators were used to verify the roles of ethylene and PAs on spikelet degeneration under WS.

Plant materials and cultivation
The present study was carried out at Fuyang Experimental Station (30°45'N, 119°55'E), which belonged to China National Rice Research Institute, from May to October in 2019 and 2020.Two indica hybrid cultivars, Hanyou-113 (HY113, a drought-resistant rice cultivar) and Zhongzheyou-8 (ZZY8, a local high-yielding and drought-susceptible rice cultivar), were used in the present study.The seeds of HY113 and ZZY8 were manually sown in the field on May 21 in both study years.After 25 days, the seedlings of the two rice cultivars were manually transplanted into polyethylene plastic pots.The polyethylene plastic pot was 25.5 cm in diameter, 34 cm in height, and 18.2 L in volume, and each polyethylene plastic pot was filled with 22.5 kg soil.The soil used in the present study was taken from a nearby field.The organic matter concentration of the soil used in the study was 38.4 g kg -1 , and the available concentrations of nitrogen (N), phosphorus (P), and potassium (K) were 74.2, 25.0, and 98.5 mg kg -1 , respectively.Three hills with two seedlings per hill were planted in each polyethylene plastic pot.Urea was used as an N source.The N fertilizer was divided into three different applications.For each pot, 2.0 g urea was used as the base 1 day before transplanting.Then, 1.2 g urea was applied 7 days after transplanting; and 0.8 g urea was used during the growth period of panicle initiation.Single superphosphate was used as a P source, and potassium chloride was used as a K source.For each pot, 1.5 g P 2 O 5 and 1.0 g K 2 O were applied 1 day before the manual transplant, and another 1.0 g K 2 O was applied during the growth period of panicle initiation.From transplanting to the onset of pollen mother cell meiosis (PMC), each pot was kept irrigated by maintaining a 2-to 3-cm water level, and then the treatment was started.The heading dates for both cultivars were from August 18 to 20, and the two cultivars were harvested on October 10 in 2019 and 2020.

Treatments
The present pot experiment was a 2 × 2 (two rice cultivars and two water treatments) factorial design with four-treatment combinations.Each treatment had 200 pots as replicates.Two soil moisture treatments were used in the present study.One was well water (WW), and another was WS.Both treatments were imposed from the onset of PMC to the pollen completion period.In the WW treatment, each pot was continuously flooded with 2-to 3-cm depth water, and the soil water potential (SWP) was maintained at 0 KPa.In the WS treatment, the SWP was maintained at − 50 ± 5 KPa.A tension meter was installed in each pot to monitor the SWP.The data of soil water potential were recorded every 3 h from 6:00 to 18:00.When the SWP was lower than the threshold in WS treatment, 0.15 L of tap water was manually supplied into each pot.All of these pots were placed in the field.A removable polyethylene shelter was used to cover the experimental field when it rained to avoid the interference of rainfall in the experiment.The pots of WS treatment were re-watered and flooded with a water depth of 2-3 cm after the water stress treatment was finished until 10 days before the final harvest.

Sampling
Six hundred main stems were chosen and manually marked in each plot during the tilling period.Then, 100 young panicles from the marked main stems were sampled in PMC meiosis prophase (PMCp), metaphase (PMCm), and telophase (PMCt).The sampling time was at 12:00 noon on each sampling day.Further, 20 panicles were used to measure the ethylene evolution rate, and the remaining 80 panicles were rapidly frozen in liquid nitrogen for 2 min and then transferred to an ultra-low-temperature freezer (-80 °C) for storage.These panicles were used to measure the concentrations of PAs, ACC, and H 2 O 2 , and the activities of the enzymes involved in PAs and ethylene biosynthesis.Using the method described by Zhang et al. (2017), 100 young panicles were sampled at the heading time to measure the number of differentiated and degenerated spikelets per panicle and spikelet degeneration rate (SDR).

Plant water status
For both cultivars, the leaf water potential and panicle water potential were both measured from 6:00 to 18:00, every 2 h, during the PMCm period.The upmost fully expanded leaves and panicles were chosen randomly to measure both leaf water and panicle water.The leaf panicle water potentials were measured using a Model-3000 pressure chamber (Soil Moisture Equipment Corp., CA, USA).

Ethylene, ACC, and PA analysis
Ethylene evolved from young panicles were measured by the method of Zhang et al. (2017).The sampled young panicles were placed between two sheets of wet paper for 60 min at 25 °C in the dark to allow the evolution of ethylene from the wound to subside.After that, each sampled young panicle was transferred to a 25-mL glass vial.Then, the glass vial was immediately sealed and incubated for 24 h at 25 °C in the dark.Ethylene evolving from young panicles was measured using a gas chromatograph (HP5890 Series II, Hewlett Packard, CA, USA) as described by Yang et al. (2007).The ACC concentration in young panicles was measured using the method of Chen et al. (2013); the transformation rate from ACC to ethylene for young panicles was 91.5% on average.
The concentrations of free PA fractions in panicles were determined according to Minocha et al. (1990), and the concentrations of soluble-and insoluble-conjugated fractions were determined by the method introduced by Liu et al. (2004).Benzoyl chloride was used to extract PAs from the nonhydrolyzed supernatant, hydrolyze the supernatant, and hydrolyze the pellet.PAs were analyzed using a high-performance liquid chromatography (HPLC) system (Waters 1525 Binary HPLC Pump/2489 UV Detector, Waters, MA, USA) using the method of Chen et al. (2013).The concentration of soluble-conjugated PA was calculated by subtracting the free PA concentration in the nonhydrolyzed supernatant from that in the hydrolyzed supernatant, and 1,6-hexanediamine was used as the internal standard.The PA levels were determined as the average of three replicates per independent sample.

Chemical applications
The drought-susceptible rice cultivar ZZY8 was used in the chemical application experiment.ZZY8 was pot grown, and both WW and WS treatments were imposed during the growth stage of meiosis, as described earlier.In each treatment, 120 pots were used as replicates.The details of treatments were the same as for the aforementioned experiment.Synthetic Put (50µM), Spd (50µM), MGBG (50µM), ACC (10µM), AVG (10µM), and 50µM Spd + 50µM MGBG were applied to panicles at the onset of PMCp, and the chemicals at the rate of 0.5 mL per panicle were applied daily for 2 days, using the method described by Zhang et al. (2020).As a control, 0.5 mL of deionized water was also injected.Each treatment had 80 main stems as replicates.During the PMCm period, the level of ethylene, ACC, free PAs (Put, Spd, and Spm), and H 2 O 2 in the young panicles were measured; the determination methods were in good agreement with those described earlier.The SDR was determined at the heading time.

Statistical analysis
The results of this study were evaluated using the analysis of variance with SPSS software Version 16.0 for Windows.Data from each sampling date were analyzed separately.The means were compared using the least significant difference at the P = 0.05 level.The data were averaged because the differences in data across the two study years and in the interaction between treatments and years were not significant (F < 1).

Leaf and panicle water potentials
Figure S1 indicates the changes in leaf and panicle water potentials during the PMCm period.Under WW treatment, the leaf water potential for both cultivars showed a small change during the day, which ranged from − 0.35 MPa (6:00) to − 0.85 MPa (12:00) (Fig. S1A).However, under WS treatment, the leaf water potential for both cultivars significantly reduced, which ranged from − 0.65 MPa (6:00) to − 1.78 MPa (12:00) (Fig. S1A).
Under both treatments, the panicle water potential for HY113 and ZZY8 showed a small change during the day, ranging from − 0.35 MPa at predawn (6:00) to − 0.93 MPa at midday (12:00) (Fig. S1B).Furthermore, no significant differences were found between HY113 and ZZY8 in leaf and panicle water potential under both treatments (Fig. S1).

Changes in the levels of PAs in young panicles
Figure 1 shows the changing pattern of free PAs in young panicles for HY113 and ZZY8 under different treatments during meiosis.For both cultivars, compared with WW treatment, WS treatment significantly increased the free Put concentration but significantly reduced the concentrations of free Spd and free Spm in young panicles (Fig. 1).Furthermore, the free Put concentration was significantly higher and the concentrations of free Spd and free Spm were significantly lower for ZZY8 than for HY113 under WS treatment (Fig. 1).However, the difference in the concentrations of free PAs (including free Put, free Spd, and free Spm) was not significant between the two cultivars under WW treatment (Fig. 1).
No significant change was found in the concentrations of soluble-conjugated PAs and insoluble-conjugated PAs in the

Changes in the levels of ethylene and ACC in young panicles
Figure 2 shows the changing pattern of the ethylene evolution and the concentration of ACC in young panicles for HY113 and ZZY8 under different treatments during meiosis.The ethylene evolution rate was significantly higher for ZZY8 than for HY113 under WS treatment (Fig. 2A).However, the difference in the ethylene evolution rate was not significant between HY113 and ZZY8 under WW treatment (Fig. 2A).The change trends of the concentration of ACC in young panicles were similar to that of the ethylene evolution rate (Fig. 2B).The concentration of ACC significantly correlated with the ethylene evolution rate (r = 0.989**, P = 0.001) (Fig. 2), implying that the increase in ethylene evolution was attributed to the enhancement of the ACC concentration in the young panicles of WS plants.

Changes in the activities of the enzymes involved in PA biosynthesis in young panicles
The changes in the activities of ADC in young panicles during meiosis were in good agreement with the free Put concentration (Fig. 3A).However, the activity of ODC was significantly lower than that of ADC; the difference in the activity of ODC between WW and WS treatments or between HY113 and ZZY8 was not significant (Fig. 3B).Furthermore, the changes in the activities of SAMDC and Spd synthase in young panicles during meiosis were similar to the changes in the concentrations of both free Spd and free Spm (Fig. 3C and D).For both cultivars, WS treatment significantly increased the activities of ADC but reduced the activities of SAMDC and Spd synthase in young panicles during the growth period of meiosis (Fig. 3A and C, and 3D).The activities of ADC were significantly greater, while the activities of both SAMDC and Spd synthase were significantly lower for ZZY8 than for HY113 under WS treatment (Fig. 3A and C, and 3D).However, the differences in the activities of ADC, SAMDC, and Spd synthase were not significant between the two cultivars under WW treatment (Fig. 3A and C, and 3D).

Changes in the activities of the enzymes involved in ethylene biosynthesis in young panicles
The activities of ACC synthase and ACC oxidase were in good agreement with the ethylene evolution and ACC concentration (Fig. 4).The activities of ACC synthase and ACC oxidase for both cultivars increased under WS treatment than under WW treatment (Fig. 4).The activities of ACC synthase and ACC oxidase were significantly greater for ZZY8 than for HY113 under WS treatment (Fig. 4).
young panicles during the growth period of meiosis.Furthermore, we observed that the concentrations of soluble-conjugated PAs and insoluble-conjugated PAs were significantly lower than those of free PAs.The concentration of solubleconjugated Put, Spd, and Spm in the young panicles during the growth period of meiosis was in the range of 0.21-0.25,0.11-0.18,and 0.04-0.09µmol g − 1 DW, respectively.The concentration of insoluble-conjugated Put, Spd, and Spm in the young panicles during the growth period of meiosis was in the range of 0.15−0.20,0.12−0.18,and 0.05−0.11µmol g − 1 DW, respectively.No significant difference was found between treatments and cultivars (data not shown).

Spikelet degeneration
The difference in the number of differentiated spikelets per panicle was not significant either between HY113 and ZZY8 or between WW and WS treatment (Fig. 6A).For both cultivars, WS treatment significantly increased the degenerated spikelet rate and reduced the number of spikelets per panicle compared with WW treatment (Fig. 6B and C).The number of spikelets per panicle showed no significant difference between HY113 and ZZY8 under WW treatment, and was significantly higher for HY113 than for ZZY8 under WS treatment (Fig. 6B).Furthermore, the degenerated spikelet rate of HY113 was less affected by WS and was only 12.1% higher than that of the WW plant (Fig. 6C).However, the However, the differences in the activities of ACC synthase and ACC oxidase in young panicles were not significant between the two cultivars under WW treatment (Fig. 4).

Changes in the H 2 O 2 concentration in young panicles
Figure 5 illustrates the changes in the H 2 O 2 concentration in young panicles.For both cultivars, the H 2 O 2 concentration in young panicles was significantly higher under WS treatment than under WW treatment (Fig. 5).Under WW treatment, the difference in the H 2 O 2 concentration between ZZY8 and HY113 was not significant (Fig. 5).However, the H 2 O 2 concentration in young panicles was significantly higher for ZZY8 than for HY113 under WS treatment during meiosis (Fig. 5).However, the application of AVG (an inhibitor of ACC synthase) or Spd could significantly increase the concentrations of free Spd and free Spm, and reduce the concentrations of free Put and ACC, the ethylene evolution rate, and the H 2 O 2 concentration (Table 2).Interestingly, compared with.degenerated spikelet rate of ZZY8 was 29.5% higher than that of the WW plant (Fig. 6C).

Relationships of PA and ethylene biosynthesis with spikelet degeneration
The correlation analysis indicated that during meiosis, the SDR significantly and inversely correlated with the mean concentrations of free Spd and free Spm, the activities of SAMDC and Spd synthase, and the ratio of Spd concentration to ACC concentration (Spd/ACC) and Spm concentration to ACC concentration (Spm/ACC) in young panicles.In contrast, it significantly and positively correlated with the mean concentration of free Put, the mean ethylene evolution rate, the ACC concentration, the activities of ACC synthase, ACC oxidase, and ADC, the H 2 O 2 concentration, and the ratio of free Put concentration to ACC concentration (Put/ ACC) (Table 1).

Effects of chemical regulators
As shown in Table 2, compared with control, the application of Put, MGBG (an inhibitor of SAM decarboxylase), and ACC could significantly increase the concentrations of free Put and ACC, the ethylene evolution rate, the H 2 O 2 concentration, and the SDR, and reduced the concentrations of free Spd and free Spm under both WW and WS treatments.

Discussion
The growth period of meiosis is one of the most sensitive periods to WS in rice.WS during this period can cause a significant increase in the SDR, finally leading to a significant reduction in both the number of spikelets per panicle and grain yield (Zhang et al. 2017(Zhang et al. , 2020;;Yang et al. 2007;Boyer and Westgate 2004).The results of the present study were very similar to those of previous studies.We found that WS during meiosis could cause serious rice spikelet degeneration.However, the SDR varied with cultivars.WS treatment significantly increased the SDR by 12.1% for HY113 and 29.5% for ZZY8 compared with WW treatment.The genotypic differences in drought resistance could provide a rare opportunity to research the underlying physiological and biochemical mechanisms involved in WS in rice (Chu et al. 2018;Anupama et al. 2018;Kato et al. 2007).
This study indicated that WS treatment significantly reduced the leaf water potential compared with WW treatment for both cultivars.However, no significant difference was observed between the two treatments in young panicle water potential for both cultivars, indicating that WS during meiosis could not affect the panicle water potential.Previous studies found that the panicle water status unaffected by WS might contribute to the enclosing leaf sheaths, limiting transpiration (Saini and Westgate 2000;He and Serraj 2012;Yang et al. 2007).Herein, we suspected that the difference in the SDR between the two cultivars under WS treatment was not attributed to the panicle water status.control, the application of AVG or Spd to WS plants could significantly reduce the SDR, but the spikelet degeneration remained constant when AVG or Spd was applied to the WW plants (Table 2).presented as averages between 2 years because the differences in them across years and in the interaction between treatments and years were not significant (F < 1).Data are means ± standard error of the eight independent measurements.Data within the same column and the same soil moisture treatment followed by dissimilar letters differ significantly at P = 0.05 ethylene promoted spikelet degeneration under WS by promoting the generation of ROS.
Prior to this study, few studies reported about how PAs regulated rice spikelet degeneration when plants were subjected to WS during meiosis.Previous studies showed that the higher concentrations of free PAs (Spd and Spm) were closely connected with the higher grain filling rate in rice and wheat (Yang et al. 2008(Yang et al. , 2014;;Liu et al. 2016;Xu et al. 2016Xu et al. , 2022)).In the present study, we observed that the mean concentrations of free Spd and free Spm and activities of SAMDC and Spd synthase during meiosis were very significantly and inversely correlated with the SDR.The application of Spd or Spm to the panicles of WS plants could significantly increase the concentrations of free Spd and free Spm and reduce the SDR.The opposite effects were observed when MGBG, an inhibitor of SAMDC, was applied to young panicles of these plants.Therefore, it was speculated that free Spd and free Spm and their biosynthetic activities in young rice panicles played an important role in responding to WS and regulating spikelet degeneration.However, we also observed that the change in spikelet degeneration was not significant when Spd or Spm was applied to the panicles of WW plants.A possible explanation was that PA levels in young panicles under WW treatment were sufficient during meiosis, and young panicles were less sensitive or even did not respond to exogenous PAs under such a condition.
The excessive accumulation of free Put in the plant might inhibit plant growth and development.A previous study reported that the excessive accumulation of free Put in young rice panicles could increase the SDR (Zhang et al. 2017).Our study indicated that WS treatment significantly increased the free Put concentration in young panicles for both cultivars, with more increase in ZZY8 than in HY113.Furthermore, the correlation analysis showed that the free Put concentration in young panicles significantly and positively correlated with the SDR.When Put was applied to young panicles of WS plants, the free Put concentration in young panicles and SDR both significantly increased.Therefore, we suspected that WS during the growth period of meiosis could increase the free Put accumulation in young panicles, and the higher free Put concentration might promote spikelet degeneration.
In higher plants, two major pathways are involved in the biosynthesis of free Put: ODC pathway and ADC pathway (Walden et al. 1997;Papadakis and Roubelakis-Angelakis 2005).Chen et al. (2013) indicated that the free Put concentration in rice grains was closely associated with the activities of ADC and the ADC1 expression levels, but no correlation was found between the free Put concentration and the activities of ODC and ODC expression levels.The present study indicated that WS treatment could significantly Various previous studies reported that PAs and ethylene were two important endogenous plant growth regulators in higher plants, which played very important roles in response to different abiotic stresses and regulated several physiological and biochemical processes (Feng et al. 2011;Hussain et al. 2011;Kusano et al. 2008;Pal et al. 2015;Xu et al. 2021;Chen et al. 2013).However, prior to this study, little information was available about the existence of an interaction between ethylene and PAs, which mediated the effects of WS during meiosis on spikelet degeneration in rice.Herein, our results implied that WS during meiosis could significantly enhance the ethylene evolution rate, concentration of ACC, and activities of ACC synthesis and ACC oxidase in young panicles for ZZY8 than for HY113.Furthermore, the ethylene evolution rate, concentration of ACC, and activities of ACC synthase and ACC oxidase in young panicles significantly and positively correlated with the SDR.The application of AVG to WS plants could significantly reduce the SDR, while the application of ACC showed the opposite effects.
Therefore, it was concluded that ethylene played an important role in promoting spikelet degeneration under WS during meiosis.When AVG was applied to the young panicles of WW plants, the ethylene levels and ACC concentration significantly reduced, but the spikelet degeneration remained constant.One possible explanation was that for rice plants under WW conditions, the ethylene level in panicles might not be at the damaging level, resulting in no negative impact of the suppression of ethylene production on spikelet development.
The underlying mechanism of the effect of ethylene on spikelet degeneration under WS during meiosis is yet to be understood.Numerous previous studies showed that ethylene synthesis was closely associated with reactive oxygen species (ROS) levels, such as promoting H 2 O 2 generation (Choudhury et al. 2013;Hussain et al. 2020;Xia et al. 2015;Overmyer et al. 2003).ROS overproduction and accumulation of ROS could damage the cellular membrane and cause lipid peroxidation and protein and nucleic acid degradation (Asad et al. 2021;Choudhury et al. 2013;Xia et al. 2015;Overmyer et al. 2003).A report showed that high levels of ethylene in rice inferior spikelets could cause ROA overproduction and accumulation, which reduced the activities of enzymes involved in the sucrose to starch conversion in grains and finally decreased the grain weight (Chen et al. 2013).In the present study, the WS treatment significantly increased the H 2 O 2 concentration of young panicles compared with WW treatments for ZZY8 than for HY113.Applying AVG could significantly reduce the H 2 O 2 concentration and the SDR, while applying ACC or ethephon showed the opposite effects.Thus, it was proposed that ethylene under WS significantly contributed to the increase in the SDR.The antagonistic interactions between ethylene and PA biosynthesis likely mediated the effect of WS during meiosis on spikelet degeneration in rice.
increase the ADC activities in young panicles for both cultivars, especially for ZZY8.However, no significant difference was found between WS and WW treatments in ODC activities in young panicles for both cultivars.Our results suggested that free Put synthesis in young rice panicles was primarily via ADC, rather than via ODC.
The interactions between plant hormones were widely reported by previous studies (Liao and Bassham 2020;Munne-Bosch and Mueller 2013); the balance between promoting and inhibiting agents could regulate plant growth and development (Wilkinson et al. 2012;Peleg and Blumwald 2011).However, prior to this study, few studies reported the synchronous change in PAs and ethylene levels in rice young panicles subjected to WS during meiosis and their relationship with spikelet degeneration.In higher plants, PAs and ethylene share a biosynthetic precursor SAM.Several previous studies indicated that the increases in Spd and Spm biosynthesis might reduce the ethylene synthesis rate (Wang et al. 2012;Yang et al. 2014;Zhang et al. 2017;Feng et al. 2011).We found that the change patterns of the concentrations of free Spd and free Spm in young panicles showed opposing patterns to those of ethylene production under WS.Furthermore, we observed that the SDR significantly negatively correlated with the ratio of free Spd concentration and ACC concentration (Spd/ACC), and free Spm and ACC (Spm/ACC), and significantly positively correlated with the ratio of free Put concentration and ACC concentration (Put/ACC) in young panicles.The application of Spd, Spm, or AVG to panicles of WS plants could significantly increase the concentrations of free Spd and free Spm in young panicles, significantly reduce the concentration of ACC and ethylene evolution rate, and finally significantly reduced the SDR.When MGBG or ACC was applied, opposite effects were observed.The present results implied that PAs (Spd and Spm) and ethylene exhibited an antagonistic relationship, and the potential metabolic competition or interaction between free PAs and ethylene biosynthesis might mediate the effects of WS on the spikelet degeneration of rice.Similar results were obtained in wheat (Triticum aestivum L) (Li et al. 2004;Liu et al. 2016), maize (Zea mays L.) (Freitas et al. 2018), Rosa damascena Mill.(Sood and Nagar 2008), and olive (Olea europaea L.) (Gil-Amado and Gomez-Jimenez 2012).

Conclusions
Rice spikelet degeneration induced by WS during meiosis varies largely with the drought resistance of the cultivar.The panicle water status has no connection with spikelet degeneration.Higher levels of free Spd and free Spm in plants under WS could reduce the SDR, and the overproduction of

Fig. 1
Fig. 1 Changes in concentrations of free Put (A), free Spm (B), and free Spd (C) in young panicles of the rice cultivars HY113 and ZZY8 under WW and WS treatments during meiosis.PMCp, PMCm, and PMCt represent pollen mother cell meiosis prophase, metaphase, and telophase, respectively.Data are presented as averages between 2 years because the differences in them across years and in the interaction between treatments and years were not significant (F < 1).Vertical bars represent ± standard error of the mean (n = 6) where these exceed the size of the symbol.Dissimilar letters above bars differ significantly at P = 0.05 on the same measurement date

Fig. 2
Fig. 2 Levels of ethylene (A) and ACC (B) in young panicles of the rice cultivars HY113 and ZZY8 under WW and WS treatments during meiosis.PMCp, PMCm, and PMCt represent pollen mother cell meiosis prophase, metaphase, and telophase, respectively.Data are presented as averages between 2 years because the differences in them across years and in the interaction between treatments and years were not significant (F < 1).Vertical bars represent ± standard error of the mean (n = 6) where these exceed the size of the symbol.Dissimilar letters above bars differ significantly at P = 0.05 on the same measurement date

Fig. 4 Fig. 3
Fig.4Changes in the activities of 1-aminocylopropane-1-carboxylic acid (ACC) synthase (A) and ACC oxidase (B) in young panicles of the rice cultivars HY113 and ZZY8 under WW and WS treatments during meiosis.PMCp, PMCm, and PMCt represent pollen mother cell meiosis prophase, metaphase, and telophase, respectively.Data are presented as averages between 2 years because the differences in them across years and in the interaction between treatments and years were not significant (F < 1).Vertical bars represent ± standard error of the mean (n = 6) where these exceed the size of the symbol.Dissimilar letters above bars differ significantly at P = 0.05 on the same measurement date

Fig. 6
Fig. 6 Spikelet degeneration of the rice cultivars HY113 and ZZY8 under WW and WS treatments during meiosis.Data are presented as averages between 2 years because the differences in them across years and in the interaction between treatments and years were not significant (F < 1).Vertical bars represent ± standard error of the mean (n = 6) where these exceed the size of the symbol.Dissimilar letters above bars differ significantly at P = 0.05 on the same measurement date

Table 1
Correlations of the spikelet degeneration rate with the mean concentrations of free PAs, ACC, and H 2 O 2 , ethylene evolution rate, and activities of the enzymes involved in PA and ethylene biosynthesis in rice young panicles during meiosis Correlation significance at P = 0.01 levels (n = 6)

Table 2
Effects of applied chemical regulators on the concentration of free PAs, ethylene evolution rate, ACC and H 2 O 2 in rice young panicles and the SDR of ZZY8 under WW and WS treatments