Seeds of two barley genotypes with differential Mn efficiency, viz. the Mn-inefficient cv. Antonia and the Mn-efficient cv. Vanessa (Hebbern et al., 2005) were germinated for 5 days in vermiculite and subsequently transplanted to light-impermeable 4-L black cultivation units with four plants per unit. A chelate-buffered nutrient solution was prepared in double ionized water as specified in Pedas et al. (2005). The nutrient solution was renewed once a week and pH was adjusted every 3rd day to 6.0 ± 0.5 with 1 M HCl or 1 M NaOH. For the first 3 weeks, all plants received a Mn²⁺ concentration of 100 nM in the nutrient solution per week whereupon Mn-replete (control) plants received a Mn²⁺ concentration of 500 nM per week. In the remaining cultivation units, Mn was removed completely for different lengths of period (3–4 days between each Mn deficiency level) to obtain Mn deficiency levels at three different severities. The progression of Mn deficiency was continuously followed using Chlorophyll (Chl) a fluorescence as a tool to monitor the Mn status of the plants ensuring the establishment of three Mn deficiency levels of increasing intensity that were identical for both genotypes. The three Mn deficiency levels are hereafter designated as mild, moderate, and strong Mn deficiency. A re-supply treatment was performed by a single addition of 1,000 nM Mn²⁺ to plants exposed to strong Mn deficiency. At 8–10 days after Mn re-supply, Chl a fluorescence measurements demonstrated values comparable to control plants. All plants were grown in a controlled growth chamber with a 250–280 μmol m^-2 s^-1 photon flux density, 75–80% air humidity, and a 20°C/15°C (16 h/8 h) day/night temperature and light regime.

Chlorophyll a fluorescence induction kinetics was measured on the youngest fully emerged leaves by use of a hand-held portable fluorescence detector (Handy Plant Efficiency Analyser; Hansatech Instruments, King’s Lynn, UK) to determine the maximum quantum yield of PSII (F_v/F_m). The leaves were dark-adapted for 30 min using leaf clips before measurement. Fluorescence measurements were recorded by illumination for 1 s with 1,500 μmol photons m^-2 s^-1 at a wavelength of 650 nm.

Non-photochemical quenching (NPQ), effective quantum yield of PSII (Φ_PSII), and PSII excitation pressure (1-qP) were measured on the youngest fully emerged leaves using a pulse amplitude modulation fluorometer (DUAL-PAM-100, Walz, Germany). Leaves were dark-adapted for minimum 30 min before measurements. Data were analyzed using the software version, DUAL-PAM version 1.11. Fluorescence were measured on leaves either in continued growth light being 300 μmol photons m^-2 s^-1 (PAR) or increasing PAR levels from 0–1,976 μmol photons m^-2 s^-1 PAR with 1 min exposure per light level.

The youngest fully emerged leaves were freeze dried (Christ Alpha 2–4; Martin Christ GmbH, Osterode, Germany). Homogenized leaf material was digested with ultra-pure acids (HNO₃ and H₂O₂) as previously described (Pedas et al., 2005; Hansen et al., 2009). Subsequently, the element concentrations were analyzed by ICP-OES (Optima 5300 DV, PerkinElmer, USA). The data quality was evaluated by including certified reference material (apple leaf, NIST1515, National Institute of Standards and Technology, Gaithersburg, MD, USA) in each analytical run. Data was accepted if the accuracy was above 90% of certified reference values.

Isolation of thylakoid membranes were performed under dim green light at 5°C. Leaves were disrupted in a razor-blade blender in ice-cold homogenisation buffer [0.2 M sucrose, 10 mM NaCl, 5 mM MgCl₂, 20 mM Tricine (pH 7.9), 10 mM ascorbate and 10 mM NaF, 1mM Na-orthovanadate, and 1 tablet of protease inhibitor (Complete protease inhibitor, Roche) per 100 mL buffer]. The extract suspension was filtered through two layers of Miracloth and centrifuged at 6,000 × g for 10 min at 4°C. The supernatant was decanted and the pellet was re-suspended in Tricine buffer [5 mM Tricine (pH 7.9), 10 mM NaF, 1mM Na-orthovanadate, and 1 tablet of protease inhibitor per 100 mL buffer] followed by 10 min centrifugation at 11,200 × g, at 4°C. The thylakoid pellet was re-suspended in storage buffer [0.2 M sucrose, 10 mM NaCl, 5 mM MgCl₂, 20 mM Tricine, 10 mM NaF, 20% glycerol, 1 mM Na-orthovanadate, and 1 tablet of protease inhibitor per 100 mL buffer] and stored at -80°C until further analysis.

Thylakoid samples (equivalent of 2 μg of total protein) were subjected to SDS-PAGE as described in Powikrowska et al. (2014) or by using the following conditions: 12% Criterion TGX Stain-Free precast gels (Bio-Rad, USA) and Tris/Glycine/SDS running buffer (Bio-Rad, USA) at 250 V constant for 32 min. The separated proteins were transferred to a 0.2 μm PVDF membrane using Trans-Blot Turbo transfer system (Bio-Rad, USA) according to the protocol of the manufacturer. Subsequently, the membrane was blocked for 1 h in 5% (w/v) skimmed milk in PBS-T buffer and incubated overnight in primary antibody. All primary antibodies were obtained from Agrisera, AB, Sweden. CF1-ATPase was used as a loading control. The blot was washed three times 5 min in PBS-T buffer and incubated with a secondary swine anti-rabbit horseradish HRP-conjugated antibody (Pierce, USA) at a 1:5,000 dilution in PBS-T for 1 h. Again the blot was washed three times 5 min in PBS-T buffer and the secondary antibody was detected using Clarity Western ECL chemiluminescent substrate (Bio-Rad, USA) and developed with a ChemiDoc Touch Imaging System (Bio-Rad, USA). Quantification of the protein bands was performed using Image Lab software (version 5.2.1, Bio-Rad laboratories, USA). Data was normalized to the CF1-ATPase signal, and the effects of the Mn treatments were interpreted by calculating the relative ratio between treatment and control.

Statistical analysis was undertaken using SAS (SAS Institute; version 9.3.) or R (R version 3.2.1) for variance analysis for comparison of mean values of F_v/F_m values, leaf Mn concentrations, and protein abundances between the treatments. The genotypic effect of Mn deficiency on F_v/F_m and leaf Mn concentrations were analyzed with two-way ANOVA (Table 1) and the effect of Mn deficiency on protein abundance was analyzed with one-way ANOVA (Figure 5) for each genotype. Both analyses were followed by LSD test and P-values < 0.05 were considered significant. Mean values (X) are listed with the associated SE values (X ±).

Two barley genotypes with contrasting Mn efficiency were grown at different levels of Mn deficiency, designated as control, mild, moderate, or strong Mn deficiency. Chl a fluorescence was continuously measured to monitor the progression of Mn deficiency and revealed the induction of Mn deficiency below a critical threshold limit of about 15 μg Mn g^-1 dry weight (DW) in barley leaves (Table 1) (Reuter et al., 1997), although no visual leaf Mn deficiency symptoms could be observed in any of the treatments. However, it should be noted that prevailing Mn deficiency eventually will bring about leaf interveinal chlorosis and reductions in biomass (Hebbern et al., 2009; Schmidt et al., 2016). The recorded F_v/F_m values of control plants were maintained close to the theoretical optimum of 0.83 throughout the experimental period (Table 1). However, F_v/F_m decreased drastically in response to Mn deficiency, reaching values of 0.7, 0.6, and 0.5 in the mild, moderate, and strong Mn deficiency treatments, respectively. The measured F_v/F_m values corresponded to Mn leaf concentrations of around 9 μg Mn g^-1 DW and around 6.5 μg Mn g^-1 DW under mild and moderate Mn deficiency, respectively, whereas strong Mn deficiency correlated with values of 3.8 and 5.3 μg Mn g^-1 DW for Antonia and Vanessa, respectively (Table 1). Furthermore, the multi-element ICP-MS analysis showed sufficient concentrations of all other essential plant nutrients as evaluated on the basis of established threshold concentrations (data not shown).

After having established that chlorophyll fluorescence is a reliable tool to monitor the degree of Mn deficiency, we used this tool to cultivate plants with similar levels of Mn deficiency in both genotypes. These plants were used for a more detailed characterization of the photosynthetic apparatus in response to Mn deficiency by evaluating the corresponding OJIP fluorescence transients. Furthermore, NPQ responses under various light intensities were analyzed together with the PSII subunit composition at three stages of Mn deficiency. The results are described in the following sections.

The measurements of fluorescence induction kinetics showed that Mn deficiency caused an increase in the minimal fluorescence, F₀ and a concomitant decrease in the maximal fluorescence, F_m (Figure 1). This indicates a defect in the electron transfer within PSII, leading to declining F_v/F_m values (Table 1).

Also the general shape of the OJIP fluorescence transient curves showed distinct alterations in response to Mn deficiency (Figure 1). With intensified Mn deficiency, a pronounced K step developed between 0.2 and 0.4 ms. A similar development of the K step was observed after heat-treatment of plants and was attributed to inhibition of the OEC of PSII (Toth et al., 2007). The K step became increasingly prominent with increasing severity of Mn deficiency (Figure 1) implying more damage to the OEC under these conditions. While the I step plateau remained unchanged, a D dip appearing between 75 and 90 ms after the I step became deeper as Mn deficiency became stronger (Figure 1).

No clear differences in any of the kinetic steps could be observed between the two genotypes except that the fluorescence induction transients declined faster for the Mn-efficient cv. Vanessa than for the Mn-inefficient cv. Antonia (Figure 1). Notably, a normal OJIP transient curve could be restored by re-supplying Mn to plants experiencing strong Mn deficiency (Figure 1), demonstrating that fully functional OEC’s were present in the newly developed leaves at about 8–10 days after Mn addition to the nutrient solution.

The impact of Mn deficiency on photoprotection mechanisms was examined by conducting a saturating pulse (SP) quenching analysis under growth light conditions, i.e., 300 PAR (μmol photons m^-2 s^-1). A marked reduction in the Φ_PSII was observed, constituting up to 70% under strong Mn deficiency (Figures 2A,B). This supports the drastic decrease in PSII efficiency observed in the OJIP transients (Figure 1). Plants with moderate and strong Mn deficiency were not able to increase the Φ_PSII as shown by the decreasing response as a function of time (Figures 2A,B).

Also NPQ was strongly affected by Mn, leading to declining NPQ values in Mn deficient plans, while in Mn replete (control) plants NPQ initially increased with the onset of illumination (Figures 2C,D). In general for both genotypes, the more Mn deficient the plants became, the lower NPQ values were recorded (Figures 2C,D). Strong Mn deficiency resulted in NPQ values reaching only 10% of control levels, indicating that Mn deficient plants were unable to dissipate excess absorbed energy. Notable, the steady-state levels of NPQ were maintained longer under mild Mn deficiency in leaves of Vanessa compared to Antonia although under moderate and strong Mn deficiency the genotypes were equally affected as the plants were unable to induce NPQ under these stress conditions (Figures 2C,D).

After 2 min of illumination with growth light there was a clear increase in PSII excitation pressure (1-qP) for plants exposed to moderate and strong Mn deficiency (Figures 2E,F), indicating a strongly reduced plastoquinone (PQ) pool. By contrast, the PSII excitation pressure for plants grown under mild Mn deficiency or control conditions continued to decrease showing the ability of these plants to better handle the absorbed light in a controlled way and with a greater PSII efficiency (Figures 2E,F). In general, the excitation pressure of PSII was slightly higher for the Mn-efficient genotype Vanessa compared to the Mn-inefficient genotype Antonia across all treatments (Figures 2E,F), suggesting that the ability to reduce the PQ pool was less affected in the Mn-efficiency genotype Vanessa.

Photosynthetic performance under various light intensities was measured (Figure 3), in order to investigate how increasing light intensities in combination with Mn deficiency influence the photosystem functionality, including regulation of NPQ mechanisms and the PSII excitation pressure (1-qP). The recorded light response curves revealed that for both genotypes the Φ_PSII declined faster and more pronouncedly with increasing Mn deficiency than was the case for control plants (Figures 3A,B). When comparing the light induced responses for the two genotypes in more detail, only minor differences were detected. However, the mild Mn deficiency and control treatments resulted in similar quantum yield-response curves for the Mn-inefficient genotype Antonia whereas these treatments differed for the Mn-efficient genotype Vanessa where the mild Mn deficient plants revealed a generally lower Φ_PSII than the control plants at all light intensities.

Interestingly, the ability to induce NPQ, and thereby the ability to regulate light-harvesting, was in both genotypes markedly decreased at moderate and strong Mn deficiency, the reduction reaching 90% at strong Mn deficiency (Figures 3C,D). The induction of NPQ in control and mild Mn deficiency plants showed a saturating response in relation to increasing PAR levels, whereas moderate and strong Mn deficiency resulted in a linear response implying a defect in establishment of NPQ under these conditions. In particular, more drastic reductions in NPQ induction levels were observed for the Mn-inefficient Antonia under all light intensities compared to Vanessa under mild Mn deficiency, indicating a reduced electron flow through PSII (Figures 3C,D). More specifically, Vanessa maintained around 20% higher NPQ values compared to Antonia under mild Mn deficiency, indicating an increased resistance of Vanessa plants toward photoinhibition under these conditions. No genotypic differences were observed between plants grown at control conditions or under strong Mn deficiencies, but under moderate Mn deficiency Vanessa was more affected (up to 25% reduction in NPQ) compared to Antonia (Figures 3C,D).

The impact of high light intensities on photosynthetic performance was further elucidated by calculation of the PSII excitation pressure (1-qP; Figures 3E,F). A significant increase (up to 50%) in 1-qP was observed for both genotypes at moderate and strong Mn deficiency. However, at mild Mn deficiency the excitation pressure for Antonia was lower than that for both Vanessa and for control plants (Figures 3E,F).

High light intensities clearly challenged the photosynthetic apparatus and increased the degree of photoinhibition under Mn deficiency (Figure 3). Hence, we investigated the ability of plants to restore PSII functionality when exposed to high light induced photoinhibition at increasing levels of Mn deficiency (Figure 4). As expected, the Φ_PSII was drastically reduced in leaves exposed to 1000 μmol photons m^-2 s^-1 for 10 min (Figures 4A,B). The reduction relative to control plants was most pronounced in strongly Mn deficient plants (Figures 4A,B). Subsequently, control plants were able to regain up to 90% of the initial PSII quantum yield measured prior to the light treatment, whereas plants with strong Mn deficiency only reached to approximately 60% of the initial value (Figure 4C). Interestingly, for Vanessa plants exposed to mild and moderate Mn deficiency the kinetics of recovery was slightly faster which resulted in 5–10% higher Φ_PSII values compared to Antonia (Figure 4C). Mn deficiency (mild and moderate levels) thus resulted in less overall damage to the photosynthetic apparatus in the Mn-efficient genotype Vanessa than in the Mn-inefficient genotype Antonia, suggesting differences in coping with light stress between the two genotypes under Mn deficiency.

The PSII subunit composition was investigated by western blot analysis in the two genotypes under increasing Mn deficiency (Figure 5). The abundance of the extrinsic proteins PsbP and PsbQ was dramatically reduced, whereas only minor alterations in PsbO abundance were observed (Figures 5A,B). More specifically, the accumulation of PsbP and PsbQ was reduced to 16 and 13% for PsbP and to 22 and 18% for PsbQ of control levels, for the Mn-inefficient genotype Antonia and the Mn-efficient genotype Vanessa, respectively. The strong response in PsbP and PsbQ abundance toward Mn deficiency correlates with leaf Mn concentrations (Table 1). However, the abundance of PsbP and PsbQ seem highly variable at incipient Mn deficiency rendering it difficult to demonstrate clear differences between genotypes (Supplementary Figure S1). Even so, a notable observation was that the Mn-efficient genotype Vanessa appears to react and adjust more sensitively in response to increasing Mn deficiency compared to the Mn-inefficient genotype Antonia (Figures 5A,B). To further investigate the effects of Mn deficiency on PSII subunit composition, the levels of representative PSII core proteins and LHC subunits were investigated. Interestingly, only minor, but significant, reductions in abundance of the PSII core protein D1 were observed in both genotypes as a response to increasing Mn deficiency (Figures 5C,D). Likewise, the abundance of the minor antenna complexes found closest to the PSII core complexes (Lhcb4-6) (Figures 5E,F) seem to adjust to Mn deficiency in accordance with the D1 core protein of PSII (Figures 5C–F). By contrast, the levels of the proteins Lhcb1 and Lchb2, subunits of the major antenna complexes, LHCII, remain unaffected, even in the case of strong Mn deficiency (Figures 5E,F). These results support the visual observation that the plants stayed green without any visual leaf symptoms no matter the degree of Mn deficiency throughout the experimental period. Finally, changes in selected PSI subunits (PsaA, PsaD, and PsaF) were recorded with about 20% decrease in the abundance going from control to strong Mn deficiency for both genotypes (data not shown). It is important to note that no genotypic differences in the composition of PSII core or peripheral antenna complex proteins could be observed.