1. Introduction

Manganese is an essential micronutrient for plants, as it is necessary for biochemical and physiological processes, specifically involved in photosynthesis, respiration, enzyme and protein synthesis (^{Marschner and Marschner., 2012}; ^{Bityutskii, et al., 2017}; ^{Shao et al., 2017}). However, Mn excess can induce phytotoxicity due to increased Mn solubility in an acidic soil solution (pH<5.5), increasing Mn uptake into the cell roots, mainly in cationic Mn²⁺ form (^{Kopittke et al., 2013}; ^{Blamey et al., 2015}). As result, visual shoot symptoms such as necrotic root tissues and brown necrotic spots have been observed in crops subjected to high doses of Mn. These are the first symptoms of Mn toxicity due to chlorophyll destruction and the oxidation of phenolic compounds, respectively (^{Demirevska-Kepova et al., 2004}; ^{St.Clair and Lynch., 2005}; ^{Rosas et al., 2007}; ^{Millaleo et al., 2010}; ^{Santos et al., 2017}). Accordingly, significant reductions of shoot biomass because of increased Mn doses are shown in plants, as consequence of a reduction in CO₂ assimilation and stomatal conductance (^{Santos et al., 2017}).

Excess of Mn causes imbalances in the production of reactive oxygen species (ROS), exacerbating oxidative stress through lipid peroxidation (LP) of membranes in crops such as barley, clover, soybean and perennial ryegrass (^{Demirevska-Kepova et al., 2004}; ^{Rosas et al., 2007}; ^{Mora et al., 2009}; ^{Reyes-Díaz et al., 2017}; ^{Santos et al., 2017}). To counteract Mn excess induced oxidative stress, plants are capable of improving antioxidant mechanisms such as enzymatic and non-enzymatic compounds. Non-enzymatic compounds such as phenols are able to donate electrons or generate protons which are effective ROS scavenging compounds against metal stresses (^{Ulloa-Inostroza et al., 2017}). Likewise, antioxidant enzymes such as superoxide dismutase (SOD) and guaiacol peroxidase (POD) increase their activity under Mn excess, raising values of antioxidants that are correlated with a higher ROS detoxification in plants (^{Shi et al., 2005}, ²⁰⁰⁶; ^{Mora et al., 2009}; ^{Gangwar et al., 2010}; ^{Ribera et al., 2013}; ^{Santos et al., 2017}). In contrast, other studies show that the antioxidant system decreases its activity due to the effects of Mn (^{Li et al., 2012}). This occurs when plants are able to counteract the toxic effects of excess Mn, therefore decreasing antioxidant enzyme activity such as SOD in cultivars like cucumber (Cucumis sativus) (^{Shi et al., 2006}), perennial ryegrass (Lolium perenne) (^{Mora et al., 2009}) and pea (Pisum sativum) (^{Gangwar et al., 2010}). These contrasting antioxidant responses when facing excess Mn are directly influenced by Mn resistance and sensitivity mechanisms in plants (^{Millaleo et al., 2010}).

Cereal crops are some of the most cultivated in southern Chile with a crop area of over 350,000 hectares (^{ODEPA, 2018}). Given the high acidity levels of Andisol, which can lead to Mn phytotoxicity, combined with the limited information available about Mn tolerance in commercial cereal cultivars, the objective of this study was to determine the early responses to Mn excess and its relation to antioxidant performance mechanisms and organic acid exudation in commercial barley cultivars.

Barley (Hordeum vulgare L.) seed from four cultivars (Barke, Tatoo, Scarlett, Sebastian) were germinated on filter paper moistened with 10 mL of deionized water (< 1 μS) for seven days. After germination, 10 seedlings per pot (200 mL) were transferred containing hydroponic system (^{Taylor and Foy, 1985}) composed of: 1270 µM Ca(NO₃)₂, 650 µM KNO₃, 120 µM MgSO₄, 100 µM K₂HPO₄, 2.4 μM MnSO₄, 1000 µM NH₄NO₃, 6.6 μM Na₂B₄O₇, 150 μM Mg(NO₃)₂, 0.1 μM (NH₄)₆Mo₇O₂₄, 0.6 μM ZnSO₄, 0.2 μM CuSO₄, 17.9 μM Fe- EDTA and 46 μM NaCl.

Experimental design was completely randomized. Four replicates per treatment were used, with a total of 80 barley seedlings and 10 plants per pot. Barley cultivars were subjected to control Mn concentrations using plants submitted to 2.4 μM MnCl₂ (Taylor and Foy nutrient solution) and increasing Mn concentrations of 150, 350, 750 and 1500 µM Mn. These Mn concentrations were added as MnCl₂ x 4H₂O at the beginning of the assays. Control plants were exposed to 2.4 µM Mn, as is the optimal dose of Mn for grass species (^{Rosas et al., 2007}). After an acclimation period of seven days, plants were subjected to Mn treatments during seven days. Later, seedlings were allocated at hydroponic a nutrient solution at pH 4.8, which was renewed every 2 days. A pH adjustment was performed daily. Solutions were aerated continuously with an aquarium pump and changed before and after the start of treatments.

The acclimation period was 7 days at semi controlled conditions with a 16/8 h photoperiod at 25/16 °C (day/night) and relative humidity range of 60-70% under greenhouse conditions. The study was carried out at the Scientific and Technological Bioresource Nucleus (BIOREN) at the Universidad de La Frontera, Temuco, Chile (38°45’S, 72°.40’W).

Finally, roots and shoots were carefully separated and dried at 65 °C during 48h for subsequent chemical analysis and growth determinations. Plant material was stored at -80 °C for later biochemical analysis.

Relative growth rate (RGR), total biomass and length: At the beginning of the Mn treatments, plant samples (separated in shoots and roots) from each barley cultivar were dried and weighed to determine dry weight (W1) at day 0 (t1). At harvest time (t2=7d), plant material was also collected for dry weight measurements (W2). Dry weight data were used to determine mean relative growth rate (RGR), defined as an increase of dry weight per unit of biomass and per day [g day^-1(g DW)^-1] using the formula from previous studies (^{Fernando et al., 2009}):

Where:

ln W2= media of ln transformed plant weight at the end of the experiment (day 7, t2)

ln W1= media of ln transformed plant weight at the start of the experiment (day 0, t1).

Biomass of shoots and roots was calculated based in dry weight (mg) at the end of each experiment (t2). Lengths of shoots and roots (cm) were taken from the barley plants to measure the last grown leaf (shoot) and last grown root.

Plant material was reduced in a muffle furnace at 500°C for 8 h and treated with 2M hydrochloric acid (HCl). Mn was quantified using a simultaneous multi-element atomic absorption spectrophotometer (Model 969; Unicam, Cambridge, UK) according to (^{Sadzawka et al. 2007}). To obtain Mn content, values of Mn concentrations were multiplied by the dry matter from the plant material.

The level of lipid peroxidation (LP) was assessed in fresh samples of roots and shoots from barley by monitoring the thiobarbituric acid reacting substances (TBARS) as an index of oxidative damage to the plant cells. Absorbance was measured at 532, 600 and 440 nm in order to correct the interference generated by TBARS-sugar complexes according to the modified method by ^{Du and Bramlage (1992}). This method determines the occurrence of malondialdehyde (MDA) as a secondary by product of the polyunsaturated fatty acid oxidation that can react with thiobarbituric acid (TBA) (^{Hodges et al., 1999}). The unit for LP was determined as equivalents of malondialdehyde (MDA) content (nmol g^-1 FW).

Samples were homogenized by applying ethanol (80% v/v) and then centrifuged at 13000 rpm at 4°C for 10 min. Supernatant (extract) was separated and stored at -20°C.

The AA was determined in shoots and roots using 2.2-diphenyl-1-picrylhydrazyl (DPPH·) stable free radical according to ^{Chinnici et al., (2004}) with minor modifications. A 250 µL aliquot of extract was mixed with 750 µL DPPH ethanolic solution (250 µM) in darkness for 30 min at room temperature. Later, samples were mixed and measured at 517 nm in a multimode microplate reader (SynergyTM HT, BIOTEK). Values were expressed as Trolox equivalents [µM g^-1 fresh weight (FW)] using standard curves.

Total phenols (TP) was determined by the Folin-Ciocalteau reagent, according to Slinkard and Singleton (1977). Aliquot extracts of 50 µL were mixed with 550 µL of deionized water and 100 µL of the Folin-Ciocalteau reagent. After 5 min, 300 µL of Na₂CO₃ (7%) was applied and mixed again. Mix extract was incubated for 10 min and measured at 760 nm using a multimode microplate reader (SynergyTM HT, BIOTEK). Results were expressed as micrograms of gallic acid equivalent per gram of fresh weight (µg GAE g^-1 FW).

For antioxidant enzyme extraction, samples of fresh shoot material were frozen in liquid nitrogen and stored at -80°C before extraction. Later, subsamples were ground in liquid nitrogen with 0.1 M potassium phosphate buffer (K₂HPO₄-KH₂PO₄) at pH 7.0, then centrifuged at 13000 rpm at 4 °C during 15 min. The supernatant was collected and stored for subsequent enzymatic analysis. All enzymatic activities were expressed on the basis of total protein, according to Bradford (1976).

i) Superoxide dismutase (SOD) activity. The SOD activity was assayed by monitoring the photochemical inhibition of nitroblue tetrazolium (NBT) according to Giannopolitis and Ries (1977) with minor modifications (^{Mora et al., 2009}). Summarizing, 20 µL of crude extracts were exposed to a reaction mixture containing 0.1 M potassium phosphate buffer pH 7.0, 10 mM ethylenediaminetetraacetic acid (EDTA), 260 mM methionine and 4200 µM NBT. The reaction started by adding 130 µM riboflavin. Sample mixtures were illuminated for 15 min and blanks were kept in darkness during course of the assay, at the same time. Absorbance was measured at 560 nm on a multimode microplate reader (Synergy TM HT, BIOTEK). One unit of SOD activity (U g^-1) was defined as the amount of enzyme required to cause 50% inhibition of the NBT reduction.

ii) Guaiacol peroxidase (POD) activity. The enzymatic activity consisted of measuring the formation of tetraguaiacol at 470 nm during 60 s, according to (^{Mora et al., 2009}). Sample extract (2 µL) was added to the reaction mixture, which contained 1mL of extraction buffer, 5µL of H₂O₂ (30% v/v) and 5 µL of guaiacol. Over time, absorbance was measured on a multimode microplate reader (Synergy TM HT, BIOTEK).

Exudates were collected using the methodology described by ^{Rosas et al., (2007}) with minor modifications (^{Meier et al., 2012}). Generally, entire plants were extracted and roots were washed and rinsed with deionized water. Later, samples were placed in 50 mL of HPLC water with aeration during 2 h, to avoid microbial degradation of organic acids (^{Jones and Darrah, 1994}). For anion organic acid exudation, samples were taken from the control (2.4 µM Mn), and the highest Mn dose (1500 µM Mn). Oxalate, citrate, malate and succinate concentration in exudates were evaluated.

Samples were lyophilized, resuspended in 800-1000 µL of HPLC water, filtered (0.22 µm) and analyzed by HPLC chromatography system (Merck-Hitachi, Primaide 1000 series modules) with a UV-visible detector and diode array detector. Analysis of HPLC was conducted on a 250x4.6 mm reverse phase column (Purospher STAR 120 RP-18, 5 µm particle size, Merck, Darmstadt, Germany). The mobile phase was H₂O-CH₃OH with orthophosphoric acid (200 mM, at pH 2.1) buffer. The gradient started with 80/20 H₂O-CH₃OH (v/v) during 2.5 min and then 77/23 H₂O-CH₃OH (v/v) during 2.6-4.7 min and returning to 80/20 H₂O-CH₃OH (v/v) to complete 7 min total. The flow rate used was 1 mL min^-1, 20 µL the injection volume and the detection wavelength was λ 210 nm.

The presented results are the mean values with standard errors based on 4 replicates, where all data passed the normality and equal variance tests according to Kolmogorov-Smirnov. Differences between treatments and cultivars were carried out by two-way ANOVA. Tukey’s test was used to identify means with significant differences (P≤0.05 and P≤0.01). The statistical software Sigma Plot 11.0 was used. Furthermore, to group and determine significant differences between samples based on antioxidant activity, productivity and biochemical parameters, data were imported into the PRIMER 7 software (PRIMER-E Ltd, Ivybridge, UK), transformed by a log (Xþ1) and normalized. Then, a distance matrix was generated based on Euclidean distances and samples were grouped by hierarchical clustering (group average), then visualized by Principal Component Analysis (PCA). Results of the correlation analysis are given according to their levels of significance as ** and * for P < 0.01 and < 0.05, respectively.

Plant shoots belonging to Barke and Scarlett cultivars showed less biomass when major Mn doses were applied, whereas Tatoo showed no difference in biomass with respect to the control. Sebastian cultivars showed a difference in biomass only at in the highest doses of Mn (1500 µM). In the roots, only Scarlett cultivars showed less biomass at increasing doses (Table 1).

Regarding biomass increase in comparison to the start of the experiment (RGR), all cultivars showed less biomass increase at elevated doses in the shoots. Thus, at the highest Mn doses, Barke showed the highest RGR reduction with respect to control (45%), Tatoo and Scarlett had a 24% RGR reduction and Sebastian 23%. In roots, a similar tendency was observed. Only Sebastian did not show significant differences in root RGR during all Mn doses. RGR reduction was 59% in Barke, 62% in Tatoo, 59% in Scarlett and 27% in Sebastian (Table 1).

A reduction in shoot lengths was observed in all barley cultivars with the increasing Mn treatments (Table 1). The Sebastian cultivar was the only one which did not show significant differences in shoot length when Mn was applied, in relation to the control. In the roots, a similar tendency was observed in which Sebastian remained stable, showing no differences in either tissues.

The Mn content in the barley cultivars increased proportionally with the applied Mn treatments, being greater in roots than in shoots for all cultivars (Figure 1). Scarlett showed major and significant values of Mn content in roots with the higher Mn treatments (750 and 1500 µM Mn), compared with the other barley cultivars.

Oxidative damage is represented by lipid peroxidation (LP). These parameters showed a strong positive correlation to the increasing Mn content in shoots of Tatoo, Scarlett and Barke (r= 0.759; r=0.744, r=0.742, respectively), indicating higher oxidative damage with an increase Mn concentration. Oppositely, roots showed no correlation between LP and Mn content when increasing doses of Mn were applied, except in Scarlett (Figure 2). Scarlett showed a significant increment of oxidative damage to shoots (above of 60 nmol of MDA), with the highest Mn treatment compared with the other barley cultivars.

Increasing Mn treatments were greater in shoots than in roots for barley cultivars (Figure 3A). In barley shoots, Barke had the highest antioxidant activity values compared with the other cultivars, although it showed no differences with higher Mn doses. Tattoo increased from 350 µM Mn, Sebastian decreased AA with intermediate Mn treatments, although there was no change with the last Mn treatment (1500 µM Mn), compared with the control (2.4 µM Mn). Scarlett cultivars showed no significant changes with Mn treatments. In the roots, Barke, Tatoo and Scarlett remained unaltered, and in contrast, Sebastian had a significant increment with higher Mn doses (750 and 1500 µM Mn) compared to the control of 2.4 µM Mn.

Total phenols (TP) in the shoots were higher than in roots for the barley cultivars. Barke and Scarlett significantly decreased TP with higher Mn concentrations (Figure 3B). In the roots, there was a strong decreased of TP only in Tatoo (r= 0.509), although it increased with the highest Mn treatment, similar to control values. Scarlett and Sebastian increased TP values only at the highest Mn treatment.

Values of SOD enzyme activity in roots were higher than shoots for the barley cultivars (Figure 4A). In the roots, Tatoo showed an increase of SOD activity 1.8-fold higher in the last Mn treatments, compared with the control. Barke increased SOD activity with all Mn treatments, and Sebastian only increased with the highest Mn dose. Nevertheless, Scarlett had a decrease of SOD activity with the intermediate Mn dose. However, there were no differences with the highest Mn treatment.

Likewise, POD activity also showed higher activity in roots than in shoots in the studied barley cultivars (Figure 4B). The POD activity of shoots incremented from the minimal Mn treatment applied (Tatoo and Scarlett). Barke and Sebastian had significant differences only with the highest Mn treatment, compared with the control. In roots, Tatoo increased its POD activity with the higher Mn treatments compared to the control. Scarlett and Sebastian exhibited a decrease of POD activity with the increasing Mn dose, but there were no significant differences with the 1500 µM Mn.

According to PCA, in the shoot samples we observed that control (green circles) were grouped independently at highest Mn doses (1500 µM) revealing differences among treatments at distance 2.5 (Figure 5). For example, in general control samples were most related with yield parameters such as length, RGR and biomass. In contrast, treatments with the highest Mn doses showed more relation with antioxidant performance, i.e Scarlett and Sebastian with LP and Mn content, Tatoo with SOD and AA. In roots, we found similar behavior, where control as Sebastian, Barke and Scarlett were associated to chemical parameters but control of Tatoo showed a major relation to antioxidant activity as is the case of cultivars grown under the highest Mn doses (Figure 5).

Carboxylate exudates by roots were exhibited among the control plants and by the barley cultivars under the highest Mn treatment (Figure 6A). Barley oxalate exudation in Sebastian, Scarlett and Barke was significantly higher with 1500 µM Mn. Contrarily, malate exudation was the highest for the control roots compared with Mn treatments for the same barley cultivars. Only Tatoo showed an increase of malate with the highest Mn dose. For citrate exudation, Tatoo and Barke displayed strong differences between the control and 1500 µM Mn. Thus, during last Mn treatment, the highest values of citrate exudation were obtained. For succinate exudation, Sebastian and Scarlett exhibited greater exudation rates with 1500 µM Mn.

In the principal component analyses (PCA), Sebastian, which was more tolerant to elevated Mn concentrations, shower a higher production of oxalate and succinate compared with the rest of cultivars. Scarlett also showed major production of succinate, and given that oxalate and succinate exudation was similar between, both Sebastian and Scarlett were grouped in the same cluster. Scarlett was affected with increasing doses of Mn. Control samples were related with malate exudation, whereas Barke and Tatoo with citrate exudation (Figure 6B).

In Chilean volcanic soil, cereal production is limited by a combination of low available P, low pH and high concentrations of toxic aluminum (Al³⁺) and manganese (Mn²⁺) (^{Mora et al., 2009}). Excess Mn²⁺ in soil can be readily transported into the root cells and translocate to the shoots, causing a negative effect on crop productivity (^{Marschner and Marschner, 2012}). In this study, we determined the early responses to excess Mn in four commercial barley cultivars (Barke, Tatoo, Scarlett and Sebastian). We found that Mn content was higher in roots than in shoots, and proportional to increasing Mn treatments in all cultivars. This finding is similar to that reported by ^{Rosas et al., (2007}) and ^{Inostroza-Blancheteau et al., (2017}) in ryegrass from Chilean Andisol. In contrast, recent studies in maize (^{Silva et al., 2017}) and barley genotypes (^{Huang et al., 2015}) showed that some Mn-tolerant genotypes exhibited higher Mn concentrations in shoots than in roots due to internal tolerance mechanisms.

In general, Tatoo and Sebastian showed similar biomass parameters when grown at high Mn content; whereas Barke and Scalett decreased their biomass parameters, as was reported for Mn-barley sensitive cultivars (^{Mora et al., 2009}; ^{Ribera et al., 2013}). Despite previous studies showing that lipid peroxidation (LP) is an important indicator to metal toxicity in ryegrass (^{Mora et al., 2009}; ^{Inostroza-Blancheteau et al., 2017}), cucumber (^{Shi et al., 2005}) and maize (^{Srivastava and Dubey 2011}), we observed only an increase of LP in shoot of all cultivars. In contrast, antioxidant performance was given by antioxidant enzymes, where mainly SOD activity was activated. Thus, Barke, Tatoo and Sebastian showed higher SOD production in roots at increasing Mn doses. This mechanism could be involved in tolerance against oxidative stress, to convert superoxide radicals into hydrogen peroxide and water, playing a principal role in the defense against excess Mn in tolerant plants (^{Sharma et al., 2012}; ^{Silva et al., 2017}). Meanwhile, POD increased in the shoots of all cultivars. ^{Demirevska-Kepova et al., (2004}) showed that apoplastic POD activity in barley is strongly associated to brown spots on leaves at increasing Mn content. Our results showed that brown spots on shoots and necrotic roots were evidenced at higher Mn treatments (Figure 7). Interestingly, antioxidant activity (AA) in tolerant cultivars of Tatoo and Sebastian was increased at high Mn doses, whereas total phenols (TP) were decreased in shoots of Barke and Scarlett. This is in concordance with ^{Ribera (2013}) in ryegrass, suggesting that AA and TP play a role in radical scavenging ability. However, no clear tendency in TP production was observed in this study.

Regarding organic acid production in roots, we found that tolerant Sebastian cultivars showed high oxalate production, similar to sensitive Barke and Scarlett. Several studies showed that oxalate is related to sequestration of soluble Mn in woody plants and legume species, decreasing Mn²⁺ free ions in soil. (^{Liu et al. 2017}) This is similar to that reported about microorganisms associated barley plant roots grown in acidic Andisol (Mora et al., 2017).

In summary, Tatoo and Sebastian were tolerant to high Mn doses, whereas Barke and Scarlett were sensitive. Also, higher oxalate exudation was found in Sebastian, supporting that it is a Mn-tolerant cultivar. Furthermore, Tatoo and Sebastian cultivars showed a tolerant early response against Mn stress, which could increase for longer periods under field conditions.

We determined the early responses to excess Mn in four barley cultivars (Barke, Tatoo, Scarlett, Sebastian). Tatoo and Sebastian are proposed as Mn tolerant given that the biomass parameters were not affected by increasing Mn doses. The mechanisms associated to Mn alleviation could be attributed to SOD, AA and organic acid production, mainly oxalate with a concomitant diminution of total phenols in shoot of sensitive cultivars (Barke and Scarlett). Non-enzymatic barriers were not related to early responses, and only an enzymatic barrier and oxalate exudation were considered as early indicators of Mn stress. Further studies are needed to elucidate these plant effects under field conditions, as well as genetic mechanisms involved in Mn tolerance.