Selenium maintains cytosolic Ca2+ homeostasis and preserves germination rates of maize pollen under H2O2-induced oxidative stress

Selenium (Se) displays antioxidant properties that can be exploited, in plants, to counteract abiotic stresses caused by overly-produced reactive oxygen species (ROS). Here, we show that fertigation of maize crops with sodium selenate effectively protects pollen against oxidative stress. Pollen isolated from Se-treated plants (Se1) and untreated controls (Se0) was incubated in vitro with H2O2 to produce oxidative challenge. Given the impact of ROS on Ca2+ homeostasis and Ca2+-dependent signaling, cytosolic Ca2+ was measured to monitor cellular perturbations. We found that H2O2 disrupted Ca2+ homeostasis in Se0 pollen only, while Se1 samples were preserved. The same trend was observed when Se0 samples were treated with sodium selenate or Se-methionine, which recapitulated in vitro the protective capacity of Se-fertigation. Furthermore, we found that germination rates were much better retained in Se1 as compared to Se0 (46% vs 8%, respectively) after exposure to 20 mM H2O2. The same was observed with Se0 pollen treated with Se-methionine, which is the organic form of Se into which most fertigated sodium selenate converts in the plant. These results, together, show a close correlation between ROS, Ca2+ homeostasis and pollen fertility, and provide strong evidence that Se-fertigation is an excellent approach to preserve or enhance agricultural productivity.

microgametophyte followed by rapid elongation of the pollen tube are both essential to ensure the sexual reproduction of plants 21,22 . To this end, ion exchanges mediate cell signaling events that guide the male microgametophyte towards the female egg 19 .
In particular, a number of studies pointed to the physiological role of Ca 2+ as a secondary messenger in plant cells [23][24][25][26][27] . In the plasma membrane of the plant cell, several channels regulate Ca 2+ entry, however, despite the ion's permeability, none of these channels appeared to be Ca 2+ specific 17 . Ca 2+ enters the cell through these channels and activates vertical cell growth 19 . Additionally, the formation of the pollen tube is Ca 2+ -dependent in the pollination process 19,25,28,29 .
ROS-mediated oxidative stress can cause infertility of pollen with obvious negative repercussions on plant reproduction and, in turn, agricultural production 14 . High concentrations of ROS can alter the molecular signals of the cell, including that mediated by cytosolic Ca 2+ . ROS act like agonists by either stimulating the mobilization of the ion from internal Ca 2+ stores such as endoplasmic reticulum, Golgi vesicles and vacuoles 30 , or activating the Ca 2+ entry from the extracellular medium, thereby disrupting the thousands-fold gradient between nanomolar and micromolar concentrations that the cell tightly maintains in the cytosolic and the extracellular space, respectively 25,[31][32][33][34][35][36] . A large body of literature describes the interaction between ROS and Ca 2+ , however only scant information exists about the effect of either inorganic or organic Se in the relationship between ROS, Ca 2+ signals and pollen germination 25,31,37 .
Here, we provide evidence that Se, under oxidative stress, can maintain cytosolic Ca 2+ and preserve germination of corn pollen grains, suggesting that Se should be included in soil fertilizers to enhance the plants' response to ROS formation. Furthermore, results point to a strong correlation between levels of cytosolic Ca 2+ and the mechanistic events that underlie the germination process.

Results
Morphology of maize pollen grains. Initially, to assess any morphological alterations of pollen following treatment with Se, we generated images (500X) by Field Emission Scanning Electron Microscopy, as shown in Fig. 1. These revealed that the size and shape of Se-fertilized (Se1) and control (Se0) pollen grains were typical of maize species; 38 however, Se1 samples displayed an especially rough, grainy appearance, which differed from a much smoother outer wall surface found in Se0 counterparts. content and speciation of selenium in pollen and pollen supernatant. Selenium buildup in soil and subsequent uptake by crops, as well as distribution and accumulation of Se in different chemical forms can largely depend on fertilization protocols 5,6,8,9 . Thus, we decided to investigate abundance and chemical fate of Se-treated pollen. In untreated maize pollen, Se mainly accumulated in organic form such as Se-methionine (SeMet), which represented more than 95% of the organic pool, while Se(IV) accounted for about 2/3 of the total inorganic fraction (Table 1). On the other hand, Se-fertilization stimulated a marked increase (9.1-fold) in total selenium and, more particularly, its organic forms. Notably, while inorganic Se (Se [IV] and Se[VI]) showed, approximately, a 4-fold increase in Se1compared to Se0 samples, organic compounds such as SeCys, Se-Met, and  MetSeCys displayed 8.8-,10.9-and >700-fold differences, respectively, with SeMet accounting for approximately 88% of the total amount of selenium (Table 1).
In addition, to further discriminate differences between Se0 and Se1 samples, we analyzed chemical contents in pollen supernatant, collected after the centrifugation step performed to isolate pollen grains. In Se1 supernatant, total protein and total Se levels were higher (35% and 53%), respectively than in Se0.
Since H 2 O 2 was subsequently employed to challenge pollens with oxidative stress in order to test the antioxidant potential of Se treatment, we also analysed the supernatant obtained from pollen treated with 10 mM H 2 O 2 . In this regard, virtually identical results were found in both Se1 and Se0 samples ( Table 2).

Effect of oxidative stress on cytosolic calcium ([Ca 2+ ] cp ) in control and Se-fertilised pollen grains.
With the overall objective of assessing the antioxidant potential of Se, we initially tested the impact of oxidative stress on levels of cytosolic Ca 2+ . To this end, 1-20 mM H 2 O 2 , a range previously found to be optimal for cell-based procedures 37 , was included with resuspended aliquots of pollen.
To monitor the entry of Ca 2+ from the extracellular space, we exogenously included 1 mM CaCl 2 in the reaction mixture 250 s after the start of the fluorometric measurement (37). As shown in Fig. 2A

Effects of sodium selenate and Se-methionine on ([Ca 2+ ] cp ) under H 2 o 2 -induced oxidative stress.
To assess the effect of inorganic and organic Se under oxidative stress, we carried out analyses in vitro using  Pre-incubating the pollen suspension with SeO 4 -2 or SeMet mitigated the effect of 10 mM H 2 O 2 , which could not be detected any longer when SeO 4 -2 and SeMet doses achieved 3.4 μM and 1 μM, respectively (Fig. 3A). Additionally, pre-treatment with SeO 4 -2 or SeMet prevented the increase of Ca 2+ -entry when 1 mM CaCl 2 was included in the medium (Fig. 3B).
For comparison purposes, we tested SeO 4 -2 and SeMet in the absence of H 2 O 2 and, as expected, did not find any effects on both internal Ca 2+ levels and Ca 2+ -entry (data not shown).
Germination of maize pollen under oxidative stress. Germination rates of maize pollen under normal conditions compared well with results obtained by different authors 39,40 . As shown in Fig. 4, in the absence of oxidative stress, pollen germination was similar in Se1(39% ± 1.1) and Se0 treated pollen (36% ± 0.8), whereas, in both pollen groups, H 2 O 2 reduced rates in a dose-dependent fashion. However, the inhibitory effect of H 2 O 2 was much more pronounced in Se0 than Se1 pollen, as germination curves at increasing H 2 O 2 concentration progressively diverged. At the highest H 2 O 2 concentration employed (20 mM), Se-treated pollen germinated at 18% ± 0.5 rate, which represented 46% efficiency in relation to normal conditions. Instead, the same amount of H 2 O 2 almost abolished (3% rate, 8% efficiency) germination in untreated pollen. Interestingly, Se0 pollen exposed in vitro to 1 μM SeMet exhibited germination rates similar to Se1 (Fig. 4).

Germination rates of control maize pollen grains (Se0) in the presence of SeO 4 -2
and SeMet. In the absence of oxidative stress, incubation of Se0 samples in vitro with increasing concentrations (1-30 μM) of SeO 4 -2 or SeMet, caused a dose-dependent reduction of germination ability. Figure 5 shows that the effect of the inorganic SeO 4 -2 form was slightly stronger than that displayed by organic SeMet.

Discussion
Here, we show that pollen obtained from maize crops exposed to a Se-supplemented fertilization regimen has a superior capacity to counteract ROS-mediated effects that may negatively impact germination rates. Our experimental setup implied field maize fertigation with sodium selenate, isolation of pollen from plants and, subsequently, application of H 2 O 2 in pollen preparations to raise oxidative stress in vitro. Conditions related to fertigation, and the setup of the cellular oxidative model were previously established (shown in refs 9,37 ., respectively).
Pollen samples, from plants either treated (Se1) or untreated (Se0) with sodium selenate, were used to measure cytosolic levels of Ca 2+ in the presence and absence of H 2 O 2 . Cytosolic Ca 2+ plays a major role as a second messenger, and alterations of Ca 2+ homeostasis may prompt molecular switches in the regulation and signalling networks of the pollen tube 24,29 . Furthermore, germinated pollen grains and elongated pollen tubes require an internal Ca 2+ gradient that is maintained by an extracellular ion supply 12,19,25 . Thus, in light of the well-documented, two-way relationship between ROS, which can modulate calcium-dependent cellular networks, and calcium signaling, which plays a key role in ROS assembly [31][32][33][34][35][36] , we decided to use cytosolic Ca 2+ as the end-point for monitoring perturbations caused by oxidative stress.
We found that H 2 O 2 -induced oxidative stress caused a significant alteration of Ca 2+ homeostasis, shown by an increase in cytosolic Ca 2+ and a markedly enhanced Ca 2+ entry when 1 mM CaCl 2 was exogenously added to the pollen resuspension medium. By contrast, Ca 2+ homeostasis was fully maintained in both Se1 pollen and Se0 pollen (i.e., pollen from plants not fertigated with sodium selenate) exposed to either sodium selenate or ) and selenium methionine (SeMet) on pollen grains subject to oxidative stress. In all instances, oxidative stress was induced with 10 mM H 2 O 2 . Changes in cytosolic Ca 2+ were assessed in Se0 pollen exposed to  or SeMet employed at doses ranging from 1 to 15 μM. The addition of SeO 4 -2 or SeMet to the incubation medium was performed 50 s prior to the treatment with H 2 O 2 , after which fluorometric measurements were immediately started. CaCl 2 (1 mM) was included (right panel) to assess the extent of Ca 2+ entry. Data are expressed as means ± SEM from five independent tests. In both panels, at any given concentration of SeO 4 -2 or SeMet, statistically significant differences between paired columns are indicated by different letters, whereas identical letters highlight non-significant trends. SEM, standard error of the mean. www.nature.com/scientificreports www.nature.com/scientificreports/ Se-methionine in vitro for 50 s prior to the addition of H 2 O 2 and, immediately after, the start of measurements. Given the short incubation times, it is possible to rule out, at least in vitro, the implication of any metabolic mechanism and conclude that Se, in either chemical form, acts basically as a mere ROS scavenger that ultimately prevents ROS-mediated dysfunction of Ca 2+ channels. Our results, however, indicate that the pollen obtained from Se-fertilized maize plants, as compared to Se0 controls, consisted of >95% organic Se, thereby suggesting that the inorganic form fertigated into crops was metabolically, and almost quantitatively, converted in Se-methionine (88%) and other organic forms such as SeCys and MetSeCys. to induce oxidative stress. The graph shows data obtained from Se0 (•, control), Se0 exposed to 1 μM SeMet (Δ), and Se1 (○) pollen. Tests were performed as described in the Materials and Methods. Germination was confirmed when the pollen tube was found to be longer than the pollen diameter. Results are reported as % of germinated pollen within a population of one hundred grains, and are expressed as means ± SEM from five independent tests, each of which included three technical replicates. As in Figs 3, 4, statistical significance of each set of data corresponding to a given dose of H 2 O 2 is indicated by different letters. SEM, standard error of the mean.  Fig. 4) and represent the means ± SEM from five independent measurements (each of which supported by three technical replicates). Statistical significance of each set of data corresponding to a given dose of Se forms is indicated by different letters. SEM, standard error of the mean.
The capacity of Se1 pollen to tolerate oxidative stress (46% efficiency in Se1 vs 8% in Se0 with 20 mM H 2 O 2 ) is especially important for agricultural productivity, based on multiple abiotic factors that can ultimately lead to an excess of ROS production 13,14,20 , and the observation made by other groups showing that abiotic stresses of different nature may inevitably lead to excessive ROS accumulation and, consequently, pollen sterility 14 .
Additionally, it must be noted that Se-fertigation, despite resulting in a total content of Se in pollen that was approximately 10-fold that of untreated crops, did not impact germination rates in the absence of oxidative stress. Rather, robust germination rates (39% vs 36% in Se1 and Se0, respectively), that were found to be within a range previously reported in the literature 39,40 , ruled out toxicity concerns. However, as shown in Fig. 5, higher doses (>1 μM) of inorganic or organic Se may ultimately impact fertility, although a comparison between data from Se-fertigation and in vitro Se treatments should be cautiously taken, given the capacity of fertilized plants to adapt to Se over a much longer term. In any case, any Se-fertigation protocol must be thoroughly validated considering the multiple environmental factors, both biotic and abiotic, that may influence agriculture.
Biochemical elucidation of the effects of Se in pollen are likely required in order to optimize conditions and address possible unexpected events or, perhaps, uncover mechanistic events that could further improve germination rates. For example, it will be interesting to understand why the surface of Se1 pollen is morphologically different compared to normal Se0 samples, as shown by Field Emission Scanning Electron Microscopy. The rough aspect of Se1 pollen may relate to a higher permeability of the outer wall to proteins and Se itself (reported in Table 1), suggesting the possibility of thinner and/or more fragile wall layer(s).
In conclusion, we provide strong evidence, for the first time, about the beneficial effects of Se-fertigation in maize pollen. Additionally, we demonstrate that measurement of cytosolic Ca 2+ is an easy, quick determination that can be used to monitor the occurrence of oxidative stress and the efficacy of antioxidative measures.

Materials and methods
Reagents. Crop was sown on April 24 th , 2018 in single rows with 0.75 m spacing in between. The emergence process of the plants proceeded as expected; density was 6.90 ± 0.13 m -2 at the time of maize V2 stage 42 .
At the time of seeding, fertilizers were applied with broadcast distribution of urea (150 kg N ha -1 ). To prevent crops from the occurrence of weeds, we employed pre-emergence herbicides such as Terbuthylazine 17.4% + S-Metolachlor 28.9% (4.0 L ha -1 ) and hand hoeing.
To control evapotranspiration, crops were irrigated using drip lines (streamline plus 16080, Netafim srl, Italy) placed on soil surface beside each seeding row.
Experimental procedures, aimed at evaluating fertilization conditions, implied block randomization with four replicates. Se treatments included i) no external Se input (Se0, used as the reference), and ii) Se fertilization (Se1). In Se1, 100 g sodium selenate ha -1 (1449 mg plant -1 ) was applied by fertigation on June 25 th 2018. Elementary plots consisted of 24 maize rows, each of which was 20 m long.

Maize pollen collection.
Pollen was sampled at maize silk (R1 Stage) on June 3 rd 2018. Male inflorescences were shaken and pollen rain harvested with a steel funnel (20 cm diameter) outfitted with a steel filter (0.5 mm) to reduce debris (chaff and stamens) and keep insects out.
Pollen harvesting was carried out early in the morning with stable weather conditions (wind speed <1.9 ms-1; air temperature ranging between 26.7 and 28.7 °C, and relative humidity between 69% and 58%).
To minimize cross contamination, sampling of each replicate was performed by a dedicated group of technicians, and harvesting devices were thoroughly and consistently cleaned after each sampling procedure.
Within ten minutes from collection, pollen from each plot was cleaned up by hand, placed in polypropylene tubes wrapped with aluminum foil, and immediately transferred into a refrigerated container (5 °C). pollen preparations and measurement of Selenium and total protein content. Fresh pollen (200 mg/sample) was resuspended in PBS (8 ml) for 48 h in the dark, after which pollen supernatant was obtained by centrifugation (2000 g, 3 min). Both pollen grains and the correspondent supernatant were used to determine total selenium content. To this end, we employed a previously described protocol 37,43 , with minor modifications related to the biological material used, as previously described. Briefly, pollen (0.5 g/sample) was microwave-digested (ETHOS One high-performance microwave digestion system; Milestone Inc., Italy) with 8 mL of nitric acid (65%w/w) and 2 mL of hydrogen peroxide (30%w/w), after which digests were diluted with 20 mL water, passed through 0.45 μm filters, and tested by graphite furnace atomic absorption spectroscopy in a Shimadzu AA-6800 apparatus (GF-AAS; GFA-EX7, Shimadzu Corp., Japan) equipped with deuterium lamp background correction and a matrix modifier (Pd(NO 3 ) 2 , 0.5 mol L -1 in HNO 3 . Determinations were carried out in triplicate.
Total protein content in supernatant was assessed by the Bradford's method, with bovine serum albumin used as the standard 44  www.nature.com/scientificreports www.nature.com/scientificreports/ Selenium speciation in pollen by HpLc icpMS. The procedure employed to breakdown Selenium content in its different chemical forms was already described by Fontanella et al. in 43 . Briefly, fresh pollen (0.25 g/ sample in 10 mL PBS) was treated with pronase (20 mg), sonicated for 2 min, and stirred in a waterbath at 37 °C for 4 h. Samples were then centrifuged, and the supernatant 0.22 μmfiltered. Se analyses were conducted by anion-exchange HPLC using an Agilent 1100 instrument equipped with a Hamilton PRP-X100, 250 × 4.6 mm column. Standard solutions (1,5,10, and 20 μg L -1 ) of inorganic [i.e., selenite, (SeO 3 -2 ) and selenate, (SeO 4 -2 )] and organic [i.e., selenocysteine, (SeCys); Se-(methyl)selenocysteine, (SeMeSeCys); selenomethionine, (SeMet)] forms of Se were prepared using ultrapure (18.2 MΩ cm) water 43 . images of maize pollen grains by electron Scanning Microscopy. The morphology of the pollen samples was examined at 500x by Field Emission Gun Electron Scanning Microscopy using a LEO 1525 Gemini workstation (ZEISS) following Chromium metallization.
Sample preparation, experimental oxidative stress, and measurement of cytosolic ca 2+ .
Oxidative stress was induced with variable doses (between 1 and 20 mM, shown in the Results) of H 2 O 2 , which was directly added to pollen suspensions immediately prior to measurements.
Intracellular levels of calcium were measured using the fluorometric FURA-2AM indicator dye according to a protocol previously described by our group 37 . Samples were incubated with FURA-2AM (2 μL of a 2 mM solution in DMSO) for 120 min, then centrifuged (1000 g, 4 min). Pollens were then resuspended in ~10 mL of Ca 2+ -free HBSS, supplemented with 0.1 mM EGTA to reduce background signal caused by contaminating ions, to obtain a suspension containing 10 6 pollen grains/mL.
Fluorometric measurements were performed in a Perkin-Elmer LS 50 B spectrofluorometer (ex. 340 and 380 nm by dual-view wavelength splitter, em. 510 nm) with a 10 nm and a 7.5 nm slit width in the excitation and emission windows, respectively. Signals were taken after 300-350 s, a time range previously found to be optimal 37 . The signal ratio 340-380/510 nm was finally used to determine concentrations.
To this end, measurements were performed in two steps: initially, without calcium in the incubation medium and, 250 s later, in the presence of 1 mM CaCl 2 to evaluate the accumulation in the cell of the extracellular ion 36 . Cytosolic calcium concentrations were calculated as reported by Grynkiewicz 45 . pollen germination studies. Fresh pollen samples from each plot were hydrated in a humid chamber at room temperature for 30 min 46 , and then transferred to 6-well culture Corning plates (1 mg of pollen per plate) containing 3 mL of an agar-solidified growing medium composed of 1.2% agar, 10%, sucrose, 0.03% boric acid and 0.15% calcium chloride (pH 5.5) 39 .
Pollen suspensions were incubated for 24-48 h in a growth chamber at 27 °C with gentle shaking to ensure homogeneous distribution of the samples in the wells. To induce oxidative stress, H 2 O 2 was directly dispensed in the culture wells.
Germinated and non-germinated pollen grains were counted under a 10X magnification microscope. Germination rates were calculated based on three replicates, each of which consisted of one hundred grains. Germination of grains was confirmed when the pollen tube had grown longer than the grain's diameter 39 . Statistical analysis. Statistical evaluations were performed using the GraphPad Prism 6.03 software. Variance assessments included homogeneity analysis by the Levene's test, and the normality test by D' Agostino-Pearson. Significance of differences was assessed by the Fisher's least significant differences test. Differences with p < 0.05 were considered statistically significant.

Data Availability
The Authors confirm that the dataset generated and used to compile this manuscript can be disclosed and made available by the corresponding author upon request.