Pattern of ROS generation and interconversion on wet stigmas in basal and divergent angiosperms

The active participation of ROS in the reproductive physiology of flowering plants is well-documented in model species, mainly plants with dry stigmas. Here we studied the release of ROS into stigma exudate—a fluid involved in pollen acceptance on a wet stigma. The study involved plants from different systematic groups: Paeonia suffruticosa, Nicotiana tabacum, Lilium hybr., Aristolochia manshuriensis and Berberis vulgaris; in most cases, exudate was collected from intact flowers on the whole plant to minimize the effect of experimental intervention on the sensitive redox balance system. The dynamics of total oxidizing capacity of stigma exudate was assessed by EPR spectroscopy and different stigma staining. The level of hydrogen peroxide and the activity of superoxide dismutase, which catalyzes H2O2 formation, was also evaluated. In all the plants studied, total ROS level increased with stigma maturation, however, H2O2 concentration decreased in studied representatives of eudicots, except for peony. In lily it increased, while in birthwort it remained low, which was apparently due to the absence of SOD activity. In all divergent plants studied, the enzyme was active on stigmas with two isoenzymes. During the fertile period, activity was higher than at an earlier stage. Based on the obtained data, we speculate on primitive and progressive patterns of ROS generation and transformation on stigma and its possible evolution.


Introduction
Major milestones in biological evolution, such as the emergence of eukaryotic organisms, as well as chloroplast and mitochondrion formation, presumably, occurred in the presence of ROS (Inupakutika et al. 2016).Studies of ROS in plants, as well as in other living systems, developed along a sinusoid, falling into extremes.Thus, ROS were initially considered as stress agents (Baker and Orlandi 1995;Bartosz 1997); the focus then shifted from stress towards signaling/regulatory role of ROS in various physiological and developmental responses (Mittler 2017).We now understand ROS as agents that can either block growth and lead to cell death, or direct and support growth processes and cellular interactions.Interestingly, ROS appear to play both roles in plant reproduction (Breygina and Klimenko 2020;Franklin-Tong and Bosch 2021;Xie et al. 2022).
Several groups have reported ROS production on plant stigma and found it characteristic of all species studied (McInnis et al. 2006;Hiscock et al. 2007;Zafra et al. 2016).The studies were carried out in model species, and they resulted in relevant concept of ROS regulating pollen germination in vivo; other possible functions of ROS such as defense and metabolism are discussed as well (Zhang et al. 2020;Sankaranarayanan et al. 2020;Breygina et al. 2021).No comparison of redox balance systems on plant stigmas from different phylogenetic groups has been reported to date.
For plants with dry stigmas important discoveries have been made recently: PCP-B (POLLEN COAT PROTEIN B-class) peptides from a pollen grain compete with stigmatic RALFs to displace them from FERONIA-ANJEA receptor complex, leading to a decrease in ROS production (Liu et al. 2021).Thus, in Arabidopsis, ROS production on stigma first increases, since stigma peptides provoke its activation in the fertile phase, and then, after pollination, decreases, which is important for supporting germination.In Brassica rapa, ROS levels in stigmas decrease after cross pollination and increase after self-pollination (Zhang et al. 2021a).By manipulating stigmatic ROS the authors showed that they affect pollen germination and pollen tube growth in both self-incompatible and compatible responses (Zhang et al. 2021a).However, the role of ROS is not so unambiguous: for example, they were found to be necessary for pollen germination (Zhou et al. 2021).Thus, lowering ROS levels on stigma of Brassica oleracea drastically reduced the attachment and germination of compatible pollen indicating an opposite role of ROS for pollen hydration and germination in this species (Lan et al. 2017).In Arabidopsis ROS synthesis is very important at the stage of pollen hydration (Gao et al. 2016).It turned out that the mutant lacking subunit KINβγ of SnRK1 (SNF1 (sucrose nonfermenting 1)-related kinase1) protein kinase protein complex simultaneously exhibit a reduced level of endogenous ROS and hydration disorders.
The dynamics of ROS production and its significance for pollen germination on a wet stigma are not understood, indirect data are contradictory (Zhou et al. 2021).Thus, an inverse relationship between the level of ROS production after pollination and pollen tube growth in vivo has been reported for lily in the studies of gametophytic self-incompatibility (SI).In case of SI pollination, ROS production was higher; exposure to antioxidants (Tezuka et al. 1997) and inhibition of NADPH oxidase by cAMP (Tsuruhara et al. 1999) allowed SI tubes to grow further into the style.On the other hand, in vitro pollen tube growth in lily was stimulated by superoxide radical and suppressed by hydroxyl radical (Podolyan et al. 2021).
Another group of studies was conducted on tobacco.Moderate levels of H 2 O 2 stimulated tobacco pollen germination in vitro (Smirnova et al. 2013).The application of superoxide dismutase (SOD) inhibitor to receptive tobacco stigmas markedly reduced the efficiency of pollen germination and seed formation (Breygina et al. 2022).Thus, in tobacco, the conversion of superoxide to peroxide, a SODcontrolled step in redox metabolism, is one of the keys to maintaining pollen germination in vivo.However, is this true for other plants?
We have recently developed a methodological approach to semi-quantitatively assess dynamics of different ROS, which is applicable to wet stigmas.It is based on EPR spectroscopic and spectrophotometric analysis of stigma exudate (Breygina et al. 2022) and has been so far applied to one model species -tobacco.Here we use this approach to compare flowering plants from different taxonomic groups, phylogenetically distant from each other.The peculiarity of this work is that we took plants with wet stigmas, while most of the studies have been carried out on dry stigmas.Our working hypothesis was that SOD-controlled ROS interconversion is a key step in redox metabolism on wet stigma, and an evolutionary trend for it can be assumed.Certainly, we cannot draw conclusions about the whole group based on the analysis of one or two species; we nevertheless decided to share these observations, discuss them with caution, and further involve more species in the study.

Results
Plant species that we chose for this study had to meet certain criteria: (a) have a wet stigma, (b) bloom in Moscow in open ground or in a climate chamber under similar conditions.Out of all the plants that met these requirements, we chose five species phylogenetically distant from each other, based on Angiosperm Phylogeny Group classification APG IV (Chase et al. 2016;Byng et al. 2018) (Fig. 1, scheme).Thus, we opted for the following genera: Aristolochia (birthwort), Lilium (lily), Berberis (barberry), Paeonia (peony) and Nicotiana (tobacco) (Fig. 1).All studied species bloomed in June 2022, and experiments were conducted during this period.
Stigma maturity was assessed according to flower appearance and was divided into 3 stages.Closed flowers that reached the final size or at least 4/5 of it were classified as juvenile (J).Flowers that had just opened but had immature anthers were classified as pre-mature (PM).Flowers with a fully open perianth and open anthers were considered mature (M) (Fig. 1).We used non-pollinated stigmas at both stages.To qualitatively assess the timing of redox metabolism activation, we stained pistils at different stages of maturity with benzidine solution.All stigmas (except barberry) were fully stained at the fertility stage (Fig. 1b, e, h).However, there were differences in the staining of non-mature stigmas.Thus, before the onset of fertility, there was no staining in birthwort (Fig. 1a), while in lily and tobacco pre-mature stigmas were colored as intensely as the fertile ones (Fig. 1d,  g).At the same time, juvenile stigmas of lily were almost uncolored (Fig. 1c), while light coloring appeared in tobacco (Fig. 1f).In barberry, intensive staining was not visible at any stage; however, at the fertile stage, faint staining was visible at the stigma edges (Fig. 1i).The most accessible objects, tobacco and lily, for which there was no shortage of plant material, were stained additionally by two conventional methods-the fluorescent dye DCFH and nitro blue tetrazolium (NBT).
We found that both ROS-sensitive dyes in terms of signal dynamics largely repeat each other.In this regard, NBT staining repeats the dynamics of staining with benzidine, except that in lily at the stage of maturity, the stigmas were stained weaker than at the stage of full fertility (Fig. 1j, m, n = 5).The dynamics of DCFH stigma staining was the same as for NBT: the fluorescence intensity gradually increased with maturation in lily (Fig. 1k, n = 4), while in tobacco it slightly increased upon transition to the stage of pre-maturity, then the signal was stable (Fig. 1l, n = 5).However, stigma staining does not provide sufficient information on the level of various ROS.To do this, we used a combined approach to measure ROS in stigma exudate.
We used cell-impermeable hydroxylamine spin probe CAT1H for ROS quantification in stigma exudates, for reviews on application of hydroxylamine spin probes in plants see (Steffen-Heins and Steffens 2015).The advantage of hydroxylamine spin probes is that they react with ROS much faster than nitrone spin traps and are very stable, but they are not selective (Dikalov et al. 2011(Dikalov et al. , 2018)).The EPRsilent spin probe CAT1H can be oxidized to stable nitroxide CAT1 by various reactive oxygen species, mainly superoxide radical, hydroxyl radical and hydrogen peroxide, revealing the total oxidative capacity of the extracellular medium.However, since superoxide radical is the primary ROS during extracellular generation, and the spin probe reacts very quickly, it can be assumed that O • 2 ¯ is the main form that oxidizes CAT1H.
The average EPR signal intensity in the species studied (Fig. 2a, d) showed different patterns of dynamics: (1) a slight decrease at the initial stage of maturation (significant, A.m., insignificant, P.s.) followed by a sharp increase towards the stage of full fertility (birthwort, A.m., n = 5 peony, P.s., n = 5); (2) a gradual increase through the stages (lily, L.l., n = 7); (3) almost complete absence of signal at the first and second stages with moderate radical production in mature stigmas (barberry, B.v., n = 7); (4) a gradual increase at the initial stage and stable production from pre-maturity to maturity (tobacco, N.t., n = 6).The characteristic EPR spectra for some species are presented in Fig. 2c.A very sharp increase in oxidizing capacity at the stage of fertility is typical for birthwort and lily: in the second case, EPR signal increases by 3.6 times, and in the first case, by 4.5 times.For comparison, in studied Eudicots, the maximum increase was found in peony (2 times); for tobacco and barberry, the ratio of signal M to J is 1.5.
On the same species, we measured the concentration of hydrogen peroxide in stigma exudates.Since the size of the stigmas, as well as the volume of liquid for exudates washout, was different for each plant species (although we kept the approximate ratio of the washout volume to the stigma surface area), on Fig. 2b we presented the obtained data as a ratio of each point to the starting point (juvenile stigmas); absolute values of H 2 O 2 concentration are presented in Table 1.
Plants from different phylogenetic groups showed different patterns of H 2 O 2 dynamics.Thus, there were no significant changes in birthwort (Fig. 2b, A.m.); the production was very low at all stages (Table 1).Lily showed a clear increase in H 2 O 2 as the stigma matured, with a maximum at the stage of full fertility (Fig. 2b, L.l.).In peony, the maximum concentration was observed at the fertile stage, the lowest-on pre-fertile stigmas (Fig. 2b, P.s.).Tobacco was characterized by small but significant decrease in production as the stigma matured (Fig. 2b, N.t.); dynamics seems very similar to the one of birthwort.However, in terms of absolute values, H 2 O 2 production in tobacco was high, especially at the juvenile stage (Table 1).Although the pistil of barberry almost did not stain with benzidine, H 2 O 2 concentration on it was sufficient for analysis (Table 1).The spread of values was very large, which is apparently due to the small pistil size; however, the trend was similar to tobacco: hydrogen peroxide concentration was maximum in juvenile flowers (Fig. 2b,B.v.).
Superoxide dismutase (SOD) is an enzyme controlling an important step of ROS interconversion, so its activity is important to maintain an optimal balance between ROS at each stage of flower development.We used convenient and illustrative zymographic method to assess SOD activity in all studied stigmas except peony.We failed to use peony stigma extract due to additional substances interacting with the components of the reaction mixture.We had too little material obtained during the flowering season to purify the protein extract.To isolate the protein from the stigmas of birthwort, a closely related species of Aristolochia, A. contorta, was used, since at that moment the flowering of A. manshuriensis had already ended.
The molecular weight of isoenzymes differed from species to species, however, in three of the four species studied, an increase in SOD activity was observed as the stigma matured (Fig. 2e).Birthwort differed from other species in the absence of SOD activity at the both stages studied.Other plants had two isoforms: in barberry the activity of both isoenzymes was detected only at stage of full maturity; in lily, at the third stage, two active isoenzymes could be seen: major and minor; at the second stage, the activity of one isoenzyme was noticeably weaker, while the other was not clearly visible; in tobacco, both isoenzymes increased their activity during the transition to maturity.et al. 2006).With a similar reagent, we studied stigma maturation in plants from distant phylogenetic groups and showed that the overall ROS level rises during stigma development in all cases; moreover, in divergent angiosperms it occurs earlier, at pre-mature or even juvenile stage (tobacco and lily), while in basal angiosperm birthwort it occurs relatively late, directly at the transition to fertility stage.Oxidation of benzidine and its derivatives is conventionally used to stain stigmas, but interpretations vary: some authors use it to determine fertility (Dafni 1992;Souza et al. 2016), which is believed to correlate strongly with peroxidase activity (Galen and Plowright 1987;Shivanna 2020).Other use similar staining to localize ROS, and more specifically hydrogen peroxide (Ros Barceló 1998;McInnis et al. 2006).It is clear that both peroxidase and hydrogen peroxide are involved in the color reaction.Considering, however, benzidine staining in combination with EPR and spectrophotometric data, here we used it to assess the total level of redox metabolism.The relevance of benzidine staining was confirmed on two objects -tobacco and lily -dynamical staining was additionally carried out with DCFH fluorescent dye and NBT.The first dye is mostly used as a non-specific one (Lai et al. 2022), however, according to indirect data, on stigmas it shows hydrogen peroxide to a greater extent (McInnis et al. 2006), while the second one is used to visualize superoxide radical in flower tissues (Cui et al. 2022).Both stains generally repeated benzidine dynamics showing a significant increase in ROS in lily and a small increase in tobacco.

Discussion
The EPR data are consistent with the staining of pistils with a certain refinement.First, the level of oxidation capacity increases significantly during the transition from juvenile to mature stigmas.Similar dynamics had been previously described for olive, but in this case the level of superoxide was assessed by staining pistil with dihydroethidium (Zafra et al. 2010).It seems that high production of ROS at the stage of fertility is a universal pattern, but the sharpest increase in superoxide generation is found in basal angiosperms (birthwort) and monocots (lily).Dicots studied show a smoother increase.We also revealed significant differences in ROS dynamics preceding the stage of maturity.So, the stage of pre-maturity can be either below juvenile (birthwort, peony), above it (tobacco, lily, olive) or absent at both stages (barberry).It can be assumed that the increase in stigmatic ROS production as preparation for fertility is a relatively progressive feature characteristic of divergent angiosperms; the sharp increase of oxidating capacity immediately before the onset of fertility, on the contrary, can be considered a rather primitive feature.At first glance, these data do not quite agree with results obtained on another group of plant species using non-specific staining of pistils (Zafra et al. 2016).Although most species tended to increase ROS production as the pistil matures, there were also species for which ROS production was lower in mature pistils than at earlier stages (Zafra et al. 2016).Curiously, this trend was slightly pronounced in early diverging angiosperm poppy and more pronounced in basal angiosperm magnolia.Others-more divergent-flowering plants were characterized by the trend similar to that we report here.A possible reason why we did not see such conflicting trends in our sample is the selection of plants with wet stigmas.Thus, our data obtained on another selection of plants using an alternative set of methodological approaches complement the results of colleagues and clarify the processes occurring on a wet stigma.Summarizing our results obtained by two different methods, on wet stigma total ROS production always increases during maturation, but in basal plants it occurs more abruptly and starts relatively late, while more divergent plants prepare for fertility in advance by producing ROS.
Considering the obtained data on SOD activity, several trends can be traced.First, the most basal of the plants in this study was birthwort, and it had no SOD activity in stigma extracts.Since SOD genes have been found in organisms that appeared on Earth about 4.1-3.5 billion years ago (Inupakutika et al. 2016) and isoforms of this enzyme, according to bioinformatics, are numerous in higher plants (Zhang et al. 2021b), birthwort, of course, also has this enzyme, but it, apparently, is not active in stigma, which can be speculatively attributed to primitive properties.Based on this result together with EPR data and colorimetric determination of peroxide, it can be concluded that the superoxide radical makes the main contribution to the EPR signal in this species.In more divergent plants, SOD activity is present at least at the stage of full fertility (barberry).In the most divergent representatives of monocots and eudicots, lily and tobacco, SOD activity is present at both stages studied.
In petunia, SOD activity has been previously reported to increase during stigma maturation; thus in anthers and ovary it was higher than on stigma (Leung et al. 2006).The strongest Cu/ZnSOD expression in tobacco, assessed by SodCc promoter-β-glucuronidase fusion, was observed, among generative organs, in ovules, stigma and pollen grains (Van Camp et al. 1997).In stigmas of stress-resistant cotton cultivar, SOD activity was higher, and the level of ROS was lower than in the sensitive one (Hu et al. 2020).At high temperatures in pearl millet, SOD activity on stigma decreased, and the level of ROS increased (Djanaguiraman et al. 2018).In both cases, the decrease in SOD activity correlated with a decline in reproductive success.The importance of SOD for successful tobacco pollination was confirmed by the application of SOD inhibitor to stigma, which led to a decrease in both the number of tubes ingrown into the style and the seed set (Breygina et al. 2022).It can be assumed that SOD activity on stigma is important both for ROS interconversion in preparation for pollination and oxidative stress management.It should be noted that other enzymes also work on stigma, such as peroxidases and catalase, which together with SOD determine the level of ROS, mainly hydrogen peroxide.Studying the activity of these enzymes is an interesting task for future research.
As for hydrogen peroxide dynamics, in birthwort its level is stably low, at the limit of sensitivity, which is consistent with the absence of SOD in this basal angiosperm.In the only representative of monocots-lily-the level of H 2 O 2 is growing, and peony has a complex dynamics with a maximum on pre-mature stigmas.The decrease of hydrogen peroxide level was observed in two of the three dicots studied: tobacco and barberry.Olive is also characterized in general by a decrease in H 2 O 2 concentration on stigma (Zafra et al. 2010;Aslmoshtaghi and Shahsavar 2016).In general, it is currently impossible to trace an unambiguous relationship between phylogeny and H 2 O 2 dynamics.We hypothesized that the overall level of hydrogen peroxide might be more indicative.Since the size of the stigmas in different species varies significantly, we cannot make a quantitative comparison of H 2 O 2 levels on the stigma based on measurements of its concentration in washout from stigmas, but some observations are still possible.Thus, for example, juvenile pistils of tobacco and peony gave a H 2 O 2 signal an order of magnitude greater than that of lily and birthwort.This is indirectly confirmed by staining experiments: for NBT staining of tobacco stigma, we had to reduce the concentration by 4 times compared to lily, and for fluorescent imaging of DCF we significantly reduced exposure.With some approximation, we can assume that in more divergent plants the level of hydrogen peroxide is generally higher than in basal ones.It can be assumed that the perception of peroxide as a signaling molecule is a more advanced property, since, in addition to its direct effect on ion channels (Breygina et al. 2016), extracellular sensors have already been described in somatic tissues (Mishra et al. 2022).
One possible explanation of the findings (summing total ROS level, H 2 O 2 concentration and SOD activity) is that superoxide radical could initially be the main signaling molecule in stigma exudate; apparently, ROS synthesis occurred just before the onset of maturity.Due to increasing SOD activity, H 2 O 2 concentration and its role as a germination regulator became higher in more divergent angiosperms.In parallel, the onset of ROS generation shifted towards younger stigmas.The pattern of ROS balance on stigmas is very complex, since it depends on many factors, such as the phenotype of a particular plant, environmental factors, and pollen accessibility.Our study traces some important patterns, but there is still much to be explored.

Plant material
Experiments were carried out on living flowers growing in the Botanical garden of Moscow State University (Paeonia × suffruticosa Andrews), in a climatic chamber (Nicotiana tabacum L. var.Petit Havana SR1 ) and in the Main Botanical Garden of the Russian Academy of Sciences (Lilium, LA-hybrids var.Royal Trinity, Formose) or on cut branches of plants no later than 2 h after cutting (Aristolochia manshuriensis Kom., A. contorta Bunge, Berberis vulgaris L., Lilium, OT-hybrids var.Zambesi).Aristolochia was collected in the Botanical Garden of Moscow State University, Berberis -on the territory of Moscow State University.

Collection of stigma exudate
To determine the concentration of hydrogen peroxide and level of ROS generation on stigma, exudate was collected by the "cap method", which was developed by us earlier (Breygina et al. 2022).Briefly, a pipette tip containing a certain amount (corresponding to pistil size) of distilled water and/or spin trap solution [0.5 mM САТ1Н (see below), 0.1 EDTA] was put on the pistil without damaging any parts of the flower (except anthers, which in some cases were removed) and incubated for 30 min (except for peony, as mentioned further).Then the tip containing the drop was carefully removed, and drops from different flowers of the same stage were placed in an Eppendorf tube and analyzed immediately.
Since the pistils of different plants differ in size and shape, we selected the volume of liquid and the method of immersing the pistil individually.So, for tobacco and barberry, a pipette tip (10-15 µl and 50-60 µl, respectively) was convenient, for a lily and birthwort-an Eppendorf test tube (200-300 µl).The minimum volume for measurement was 200 µl for spectrophotometry and 300 µl for EPR, so for tobacco and barberry, one sample was the sum of washes from a certain number of flowers.For peony, special caps had to be made from the bottom of plastic centrifuge tubes, as their pistil is very large.In this case the incubation took place in a volume of 600 µl for 3 min.A longer incubation in peony was technically inconvenient because the cap had to be held by hand.

EPR spectroscopy
Total oxidation capacity of exudate was estimated by electron-paramagnetic resonance (EPR) spectroscopy.The EPR spectroscopy is based on interaction between the magnetic moment of an unpaired electron and external magnetic field.We used 1-Hydroxy-2,2,6,6-tetramethyl-4-(trimethylammonio)-piperidinium dichloride (CAT1H) (Dikalov et al. 2011(Dikalov et al. , 2018) ) as a spin probe.To quantitatively characterize oxidation capacity, we measured the intensity of the central line in CAT1H EPR spectra.
The EPR spectra were recorded at 21-22 ○ C with a RE-1307 EPR spectrometer (Russia) at microwave power 22 mW and time constant of 0.1 s.Each characteristic spectrum presented in the figures is an average of 10 replicates.

Benzidine staining of stigmas
The color reaction involves benzidine, enzymes with peroxidase activity, and hydrogen peroxide.To determine the total oxidative capacity (including both the production of hydrogen peroxide and the activity of peroxidases) (Josephy 1985), a 1% solution of benzidine in 50 mM Tris-acetate buffer (pH 5.0) was applied to the stigmas and left for 30 min.To ensure that intrinsic pigments did not contribute to the observed coloration, all pistils were photographed before and after staining.All, except for the pollinated pistils of birthwort, were practically uncolored before the start of incubation.Since injury to the pistil with a pipette tip led to a sharp increase of ROS production and, accordingly, to local intense staining, stigmas were not touched when applying the solution.Colored pistils were photographed using a binocular (for small pistils) and/or using a camera (for large ones).In the first case, pistils were cut from the flowers and placed in wet sand; otherwise the stigma could not be brought into focus.This cutting did not lead to hyperproduction of ROS.

NBT staining of stigmas
Superoxide radical generation on Lilium stigma was assessed by NBT staining (10 µM, 10 min) in 50 mM sodium citrate buffer solution pH 6.0 followed by boiling (15 min), method adapted from staining anthers (Cui et al. 2022).The duration of boiling (5 min) and concentration (2.5 µM) was reduced for Nicotiana to avoid excessive staining.

DCF-based detection of extracellular ROS on stigmas
Total ROS generation on stigmas of Lilium and Nicotiana was assessed by DCFH staining (30 min, 5 µM).DCFH-DA (Sigma, Darmstadt, Germany) was deesterified in an alkaline medium (10 mM NaOH, 1 h) (Smirnova et al. 2009), which made it suitable for detecting extracellular ROS (Lai et al. 2022).Pistils at different stages of maturity were stained simultaneously with the same dye solution, the upper part of the stigma was cut off and immediately observed with a widefield fluorescence microscope Axioplan 2 imaging MOT (Carl Zeiss, Jena, Germany) equipped with AxioCam a mercury lamp and Pro08 camera (ADF Optics, Zhejiang, China).FITC filter set was used.Samples were photographed with standard exposure.Fluorescence intensity was assessed with ImageJ software.

Zymographic detection of SOD activity
Superoxide dismutase (EC 1.5.1.1)activity was detected as described earlier (Breygina et al. 2022).Briefly, stigmas were collected from fresh flowers and homogenized at 0 °C in Tris-HCl buffer (50 mM, pH 7.0) with 50 mM NaCl, 0.05% Tween-20 and 0.1% protein inhibitor cocktail (Sigma, Darmstadt, Germany).The homogenates were centrifuged at 10,000 g, 4 °C, for 20 min, the supernatants were mixed 1 3 with non-reducing Laemmli sample buffer and loaded on 10 or 15% PAAG.Concentration of protein in supernatants were quantified by conventional Bradford method (Bradford 1976) using SmartSpec spectrophotometer (BioRad, Hercules, CA, USA).Vertical gel electrophoresis was performed at 180 V for 2 h at 4 °C.The gel was washed and subsequently soaked in 0.5 mM nitro blue tetrazolium (Roche, Penzberg, Germany) in the dark.After 30 min, the gel was transferred to 50 mM phosphate buffer (pH 7.8) containing 28 µM riboflavin and 28 mM TEMED, incubated for 20 min and then exposed to light.During illumination, the gel became dark blue except the positions containing SOD activity.

Statistical analysis
Experiments were performed on three to eight independent flowers of each stage.Results in the text and figures, except the original pictures and characteristic spectra, are presented as means ± standard errors.Significant difference was evaluated by OriginLab software (Northampton, USA) according to Mann-Whitney test.(*p < 0.05, **p < 0.01).

Fig. 1
Fig. 1 Systematic position of the studied species according to plant phylogeny (APG IV) (scheme) and dynamics of their stigma staining with benzidine (a-i), NBT (j, m) and DCF (k, l) reflecting ROS production in different flowers.Different stages of flower development that were taken for analysis are shown for large flowers on separate images (birthwort, lily, peony): J -juvenile stage, PM -pre-maturity, M -maturity; for small ones they are indicated by arrows (barberry, tobacco): blue arrows -juvenile stage, green -pre-maturity, redmaturity.a-i Staining of stigmas with benzidine (30 min): a, b -Aristolochia manshuriensis, c-e -Lilium hybr., f-h -Nicotiana tabacum, i -Berberis vulgaris; a, c, f -juvenile stigmas, b,d,g,i (green arrows in b and i) -pre-mature stigmas, b, e, h, i (red arrows in b and i)mature stigmas; j, m Staining of stigmas of lily (j) and tobacco (m) with NBT (10 min, 10 µM); k, l Fluorescence intensity of stigmas of lily (k) and tobacco (l) stained with DCF (30 min, 5 µM); * p < 0.05, ** p < 0.01, Mann-Whitney test.For each of the objects, flowers used in the experiment (characteristic images) are shown ◂

In 2006 ,
McInnis et al., in their pioneering research, used tetramethylbenzidine stigma staining in different plants and detected ROS production on all stigmas studied (McInnis