Glutathione Contribution in Interactions between Turnip mosaic virus and Arabidopsis thaliana Mutants Lacking Respiratory Burst Oxidase Homologs D and F

Respiratory burst oxidase homologs (Rbohs) play crucial and diverse roles in plant tissue-mediated production of reactive oxygen species during the development, growth, and response of plants to abiotic and biotic stress. Many studies have demonstrated the contribution of RbohD and RbohF in stress signaling in pathogen response differentially modulating the immune response, but the potential role of the Rbohs-mediated response in plant–virus interactions remains unknown. The present study analyzed, for the first time, the metabolism of glutathione in rbohD-, rbohF-, and rbohD/F-transposon-knockout mutants in response to Turnip mosaic virus (TuMV) infection. rbohD–TuMV and Col-0–TuMV interactions were characterized by susceptible reaction to TuMV, associated with significant activity of GPXLs (glutathione peroxidase-like enzymes) and induction of lipid peroxidation in comparison to mock-inoculated plants, with reduced total cellular and apoplastic glutathione content observed at 7–14 dpi and dynamic induction of apoplast GSSG (oxidized glutathione) at 1–14 dpi. Systemic virus infection resulted in the induction of AtGSTU1 and AtGSTU24, which was highly correlated with significant downregulation of GSTs (glutathione transferases) and cellular and apoplastic GGT (γ-glutamyl transferase) with GR (glutathione reductase) activities. On the contrary, resistant rbohF–TuMV reactions, and especially enhanced rbohD/F–TuMV reactions, were characterized by a highly dynamic increase in total cellular and apoplastic glutathione content, with induction of relative expression of AtGGT1, AtGSTU13, and AtGSTU19 genes. Moreover, virus limitation was highly correlated with the upregulation of GSTs, as well as cellular and apoplastic GGT with GR activities. These findings clearly indicate that glutathione can act as a key signaling factor in not only susceptible rbohD reaction but also the resistance reaction presented by rbohF and rbohD/F mutants during TuMV interaction. Furthermore, by actively reducing the pool of glutathione in the apoplast, GGT and GR enzymes acted as a cell first line in the Arabidopsis–TuMV pathosystem response, protecting the cell from oxidative stress in resistant interactions. These dynamically changed signal transductions involved symplast and apoplast in mediated response to TuMV.


Introduction
Plant NADPH oxidases (NOXs) are known as respiratory burst oxidases and are encoded by Rboh genes [1,2]. Respiratory burst oxidases can transfer electrons to extracellular O 2 to produce O 2 -which can be converted to H 2 O 2 by superoxide dismutase [3]. Therefore, Rbohs are considered as important factors in the reactive oxygen species (ROS) signaling network. Analysis of the expression of Rbohs genes may help to comprehend The aim of our study was to verify whether glutathione and glutathione metab enzymes play a role in rbohD-, rbohF-, and rbohD/F double mutant-TuMV interac The study investigated, for the first time, glutathione localization and content, glu one-associated relative expression of selected genes, and enzyme activities in rb rbohF-, and rbohD/F-TuMV pathosystem. The results showed significant differences lular and apoplastic glutathione localization and content between susceptible an sistance responses to virus inoculation. Moreover, significant variations were noted relative expression of AtGGT and AtGSTUs, which clearly indicated the role of glutat and AtGGT, AtGSTUs in resistance as well as susceptible reaction to TuMV.

Significant Regulation of TuMV Concentration in Col-0-, rbohD-, rbohF-, and rbohD TuMV Interactions
Our previous research demonstrated that Col-0 and RbohD and RbohF trans mutants rbohD, rbohF, and rbohD/F exhibited different levels of susceptibility/resista TuMV isolate PV-0104 [25]. Moreover, rbohD plants displayed systemic virus infectio were even more susceptible than Col-0 plants. On the other hand, rbohF plants sh virus limitation and were more resistant to TuMV compared to Col-0. In turn, rb plants displayed even higher resistance than rbohF accompanied by local necrosis, is characteristic of HR-like reactions [25]. Previously, virus titers were checked ser cally (DAS-ELISA) and by ultrastructural analysis (TEM). Therefore, in the present s we performed a validation based on the relative expression of TuMV-CP (TuMV c protein) gene, in comparison to two plant host reference genes, AtEf1α (A. thaliana gation factor-1 alpha) and AtF-Box (A. thaliana F-box protein family), in order to u stand the changes in TuMV content in the inoculated leaves between 3 and 14 dpi (F 1). The results of TuMV-CP gene expression analysis clearly indicated the upregulat TuMV-CP in Col-0 (2.7-fold) and rbohD (13.58-fold) plants between 3 and 14 dpi. O contrary, in the same time after inoculation, TuMV-CP levels were decreased in (2.24-fold) and rbohD/F (5. 32-fold). This confirms that rbohD mutants are more susce to TuMV infection than the already susceptible Col-0 plants, while rbohF mutants are resistant to TuMV infection. Moreover, rbohD/F plants showed enhanced resistan TuMV to some extent. The high induction or reduction of virus content was clearly ev at the 7-dpi timepoint in different types of infected plants.   3,7, and 14 dpi based on normalized relative expression of TuMV-CP. Normalized relative expression of TuMV-CP was calculated based on the mean expression of AtEf1α and AtF-Box reference genes. Statistical significance of differences was assessed at the p < 0.05 level using ANOVA with post hoc Tukey's HSD and is indicated by different letters above the bars.

Crucial Cellular and Apoplastic Localization with Concentration of GSH (Reduced Glutathione) and GSSG (Oxidized Glutathione) Pools as an Element of Increased Susceptibility or Resistance Reaction in rbohD, rbohF, and rbohD/F Arabidopsis Mutants
The changes in the expression of TuMV-CP during TuMV infection in rbohD, rbohF, and rbohD/F mutant plants confirmed the differences in the reaction of these plants to the presence of TuMV. Moreover, our previous works confirmed that RbohD [24] and glutathione [39,40] act as modulators of reaction to different Potyviruses. In normal physiological conditions, the levels of Rbohs and glutathione are precisely regulated in plants. Furthermore, the changes in their levels are highly correlated. This leads to the question on how glutathione levels will be changed and whether glutathione levels will be correlated with the modulation of host reactions, namely susceptibility or resistance, to TuMV infection. To answer this question, we quantified immunofluorescence and immunogold localization of total glutathione content (Figures 2 and 3, respectively) with the use of corrected total cell fluorescence (CTCF) estimation ( Figure 4A The total glutathione immunogold localization ( Figure 3A-L) confirmed CTCF analyses data. TuMV-inoculated Col-0 and rbohD plants were characterized by a significantly decrease of glutathione in symplast as well as in apoplast between 7 and 14 dpi ( Figures 3A,B,D,E and 4C-E). Moreover, TuMV-inoculated Col-0 and rbohD plants characterized also decreased deposition in specific cell compartments (mitochondria and chloroplast) between 7 and 14 dpi. Whereas, deposition of immunogold particles associated with glutathione was statistically increased in symplast as well as in apoplast of TuMV-inoculated rbohF and rbohD/F plants was induced between 7 and 14 dpi (Figures 3G,H,J,K and 4C,D). In contrary to the Col-0 and rbohD plants, TuMV-inoculated rbohF and rbohD/F plants had also induced deposition in cytoplasm, mitochondria, and especially chloroplast between 7 and 14 dpi ( Figure 4E).    tathione was statistically increased in symplast as well as in apoplast of TuMV-inoculated rbohF and rbohD/F plants was induced between 7 and 14 dpi (Figures 3G,H,J,K and 4C,D). In contrary to the Col-0 and rbohD plants, TuMV-inoculated rbohF and rbohD/F plants had also induced deposition in cytoplasm, mitochondria, and especially chloroplast between 7 and 14 dpi ( Figure 4E).  ( Figure 5A,B). Most interestingly, in virus-inoculated rbohD plants, the decrease of GSH (4.45-fold in cell, 4.37-fold in apoplast) was more evident and intense than that observed in TuMV-susceptible Col-0 plants. Furthermore, the regulation of cellular and apoplastic GSH levels in these two susceptible plants was highly associated with the 7-dpi time point.  ity (rbohD), the cellular GSSG level was decreased by approximately 4-fold and 4.71-fold, respectively. In turn, the apoplast level of GSSG was systemically increased between 1 and 14 dpi by 1.59-fold and 1.68-fold, respectively. This suggests that virus-inoculated rbohD plants developed more severe reactions in response to the oxidative stress in cell, which may be related to increased levels of GSSG being directed to the apoplast region.  and rbohD/F plants, there was, respectively, a 1.5-fold (in cells) and 1.29-fold (in apoplast) and 1.78-fold (in cells) and 1.34-fold (in apoplast) increase of GSH+GSSG pool between 1 and 14 dpi. In contrast, in Col-0-and rbohD-TuMV plants, cellular and apoplastic GSH+GSSG levels dynamically decreased between 7 and 14 dpi, with the highest decrease in rbohD (4.37-fold in cells and 1.78-fold in apoplasts).  Further quantitative HPLC analyses confirmed significant and timepoint-specific regulation of both glutathione (GSH and GSSG) forms in cells and in the apoplast of Col-0, rbohD, rbohF, and rbohD/F during TuMV infection. During virus infection, we observed enhanced GSH content in cell and apoplast from 1 to 7 dpi in comparison to mock-inoculated Col-0 and mutant plants ( Figure 5A,B). However, this increase was more dynamic and significant in rbohF-TuMV (resistance) and rbohD/F-TuMV (enhanced resistance) pathosystems. The highest levels of GSH were always found in virus-inoculated rbohD/F plants ( Figure 5A,B). Moreover, GSH was continuously induced between 1 and 14 dpi in rbohF (1.47-fold in cell, 1.42-fold in apoplast) and rbohD/F (1.76-fold in cell, 1.56-fold in apoplast) plants. Unlike resistant rbohF and rbohD/F plants, in the virus-inoculated Col-0 (susceptibility) and rbohD (enhanced susceptibility) plants, enhanced cellular and apoplastic GSH content was observed only up to 7 dpi, after which the GSH level decreased ( Figure 5A,B). Most interestingly, in virus-inoculated rbohD plants, the decrease of GSH (4.45-fold in cell, 4.37-fold in apoplast) was more evident and intense than that observed in TuMV-susceptible Col-0 plants. Furthermore, the regulation of cellular and apoplastic GSH levels in these two susceptible plants was highly associated with the 7-dpi time point.
GSH content as well as the cellular and apoplastic levels of GSSG was differentially regulated during TuMV infection ( Figure 6A,B). In resistant TuMV-inoculated rbohF and rbohD/F plants, the cellular GSSG content was systemically induced between 1 and 14 dpi by 1.92-fold and 2.04-fold, respectively, compared to mock-inoculated plants. However, in the same TuMV-inoculated plants, the apoplast GSSG content was increased only between 1 and 7 dpi ( Figure 6B), and after 7 dpi, the GSSG content was decreased by 1.50-fold and 2.35-fold, respectively. This indicates that TuMV inoculation modulated GSH content, which was probably due to direct conversion of GSH to GSSG in cells, but in the apoplast, another factor was also involved in GSSG regulation at some point. On the other hand, in virus-inoculated susceptible (Col-0) plants and plants with enhanced susceptibility (rbohD), the cellular GSSG level was decreased by approximately 4-fold and 4.71-fold, respectively. In turn, the apoplast level of GSSG was systemically increased between 1 and 14 dpi by 1.59-fold and 1.68-fold, respectively. This suggests that virus-inoculated rbohD plants developed more severe reactions in response to the oxidative stress in cell, which may be related to increased levels of GSSG being directed to the apoplast region.
The analyses of CTCF, quantification of immunogold labeling, a and further evaluation of HPLC results revealed that TuMV infection modulated changes in summary glutathione (GSH+GSSG) levels in cell and apoplast ( Figure 7A,B). In virus-inoculated rbohF and rbohD/F plants, there was, respectively, a 1.5-fold (in cells) and 1.29-fold (in apoplast) and 1.78-fold (in cells) and 1.34-fold (in apoplast) increase of GSH+GSSG pool between 1 and 14 dpi. In contrast, in Col-0-and rbohD-TuMV plants, cellular and apoplastic GSH+GSSG levels dynamically decreased between 7 and 14 dpi, with the highest decrease in rbohD (4.37-fold in cells and 1.78-fold in apoplasts).
The modulation of GSH, GSSG pools, and total GSH+GSSG indicated that current oxidation of plant cells infected by TuMV could be also changed. The most important parameter which is associated with oxidation feature of cell is GSH/GSSG ratio. The TuMV inoculated Col-0 and rbohD were characterized by slight increase in GSH/GSSG ratio between 1 and 7 dpi ( Figure S1). Conversely, the GSH/GSSG ratio was drastically reduced at 7 dpi to the lowest levels in the case of rbohD plants. In contrast to that, TuMV-infected rbohF and rbohD/F plants were characterized by a systematic increase of the GSH/GSSG ratio between 1 and 14 dpi.

Relative Expression of AtGGT1 and Selected AtGSTU Genes during Col-0-, rbohD-, rbohF-, and rbohD/F-TuMV Interactions Correlated with Increased Susceptibility or Resistance Reaction
The results from CTCF quantification of immunolocalization of total glutathione indicated decreased deposition of glutathione near the apoplasts of infected cells in Col-0 and rbohD plants and increased deposition in rbohF and rbohD/F plants inoculated with virus. This suggests that modulation of cellular and apoplastic glutathione could have contributed to the reaction to TuMV infection. The main enzyme responsible for the catabolism/usage of glutathione in cell wall in A. thaliana is GGT (γ-glutamyl transferase) [41]. One isoform of this enzyme plays a major role in leaf cell apoplast and is encoded by AtGGT1 gene [41]. Therefore, we examined the normalized expression of AtGGT1 gene in cells in the Atrboh-TuMV pathosystem. The changes in the expression of this gene could explain the modulation of glutathione content in the cell wall. Analyses of AtGGT1 relative expression indicated significant changes in TuMV-inoculated plants ( Figure 7A). The gene was highly upregulated in virus-inoculated rbohF (1.41-fold) and rbohD/F (2.25-fold) plants between 3 and 14 dpi, and its expression was also much higher compared to mock-inoculated plants. Moreover, the rate of upregulation and the final level of AtGGT1 expression was the highest in rbohD/F plants. In contrast, virus inoculation induced AtGGT1 in Col-0 and rbohD plants only at 3 dpi compared to mock-inoculated plants. Moreover, the relative expression of AtGGT1 decreased systemically after 3 dpi by 4.45-fold in virus-inoculated Col-0 and by 21-fold in virus-inoculated rbohD plants. Our previous investigation showed that GSTs (glutathione S-transferases), especially class tau (GSTU), play an important role in the modulation of cell glutathione levels in TuMV infection [40]. Moreover, modification of the expression of different GST genes was found to be crucial for the precise modulation of the host's response to this viral pathogen [40]. Therefore, in this study, we analyzed the relative expression of four GST genes-AtGSTU1, AtGSTU13, AtGSTU19, and AtGSTU24-which were previously identified to be involved in response to TuMV infection.
In contrast to AtGGT1, the expression of AtGSTU1 gene was induced between 3 and 14 dpi in TuMV-inoculated Col-0 and rbohD plants compared to mock-inoculated plant. The induction was found to be 1.42-fold in virus-inoculated Col-0 and 1.33-fold in virusinoculated rbohD plants ( Figure 8B). The final expression level at the 14-dpi timepoint was the highest in TuMV-inoculated rbohD plants. On the contrary, in rbohF-TuMV and rbohD/F-TuMV interactions, AtGSTU1 expression was upregulated only between 3 and 7 dpi (in comparison to mock-inoculated plants). After 7 dpi, the level of AtGSTU1 decreased by 2.67-fold and 16-fold, respectively, in virus-inoculated rbohF and rbohD/F plants. The relative expression of AtGSTU13 was also significantly modulated between 3 and 14 dpi in Col-0, rbohD, rbohF, and rbohD/F plants after TuMV inoculation ( Figure 8C). In the rbohFand rbohD/F-TuMV interaction, we observed induction of AtGSTU13 expression compared to mock-inoculated plants, and the induction was the highest in rbohD/F plants ( Figure 8C). In contrast, the expression of AtGSTU13 was induced in virus-inoculated Col-0 and rbohD only between 3 and 7 dpi, and after 7 dpi, it significantly reduced. Further analyses of the relative expression of AtGSTU19 and AtGSTU24 ( Figure 8D,E) showed significant changes in the analyzed plants. The timepoint changes in AtGSTU19 expression were similar to the modulation of AtGSTU13 expression. The normalized expression of AtGSTU19 was highly upregulated in rbohF and rbohD/F plants during interaction with TuMV by 1.55-fold and 1.64-fold, respectively, between 3 and 14 dpi. On the other hand, in Col-0 and rbohD plants, the expression of AtGSTU19 was downregulated by 7.24-fold (Col-0) and 13.85-fold (rbohD) after 7 dpi. The time of regulation of AtGSTU24 was similar to that of AtGSTU1. AtGSTU24 was upregulated between 3 and 14 dpi in Col-0 (2-fold) and rbohD (3.03-fold) plants during interaction with TuMV ( Figure 8D), whereas its expression was only slightly induced between 3 and 7 dpi in rbohF and rbohD/F plants during TuMV interaction. After 7 dpi, AtGSTU24 was downregulated in virus-infected rbohF and rbohD/F plants.
the other hand, in Col-0 and rbohD plants, the expression of AtGSTU19 was downregulated by 7.24-fold (Col-0) and 13.85-fold (rbohD) after 7 dpi. The time of regulation of AtGSTU24 was similar to that of AtGSTU1. AtGSTU24 was upregulated between 3 and 14 dpi in Col-0 (2-fold) and rbohD (3.03-fold) plants during interaction with TuMV ( Figure  8D), whereas its expression was only slightly induced between 3 and 7 dpi in rbohF and rbohD/F plants during TuMV interaction. After 7 dpi, AtGSTU24 was downregulated in virus-infected rbohF and rbohD/F plants.   The modulation of total glutathione localization, with changed levels of cellular and apoplastic GSH/GSSG forms, and differences in the expression of AtGTT1 and AtGSTU genes, revealed the precise control of glutathione in different types of host reactions to TuMV. Therefore, we analyzed the cellular and/or apoplastic activity of glutathione metabolism enzymes in reaction to viral infection. For validation, we selected enzymes associated with whole cell (GSTU, GR, and glutathione peroxidase-like enzymes (GPXLs)) as well as two enzymes associated mainly or partially with the apoplast of host cells (GGT and GR). The analyses of the activity of GGT ( Figure 9A,B) and GR ( Figure 9C,D) indicated other types of relationship with different reactions to TuMV. The activity of cellular GGT was stably upregulated between 3 and 14 dpi in virus-inoculated rbohF and rbohD/F mutants, respectively, by 1.81-fold and 1.87-fold ( Figure 9A). Similarly, upregulated GGT activity was also observed in apoplasts in rbohF and rbohD/F mutants ( Figure 9B) In these plants, GGT activity was induced, respectively, by 1.8-fold and 1.87-fold. In contrast, in virus-inoculated Col-0 and rbohD plants, cellular GGT activity was induced only at 3 dpi, compared to mockinoculated plants. After 3 dpi, GGT activity was found to be decreased in these plants by 4.17-fold (Col-0) and 5.43-fold (rbohD). The same pattern was also observed for the activity of GTT in apoplast, with depletion of enzyme activity by 3.71-fold (Col-0) and 4.83-fold (rbohD). In Col-0 (susceptibility) and rbohD (enhanced susceptibility) plants, GR activity was upregulated between 3 and 7 dpi during interaction with TuMV ( Figure 9C,D). On the other hand, the upregulation of GR activity in the virus-inoculated rbohD plants was lower compared to Col-0 plants and the enzyme activity was only induced by 1.04-fold between 3 and 7 dpi. After 7 dpi, the cellular and apoplastic GR activity was downregulated in Col-0 as well as rbohD plants. Moreover, the most significant reduction in cellular (9.72-fold) and apoplastic GR activity (28-fold) was observed in rbohD plants. In contrast, rbohF and rbohD/F plants were characterized by increased cellular and apoplastic GR activity between 3 and 14 dpi ( Figure 9C,D). Moreover, the highest induction of enzymatic activity was observed in rbohD/F plants during TuMV interaction, with 1.28-fold and 1.5-fold induction in cell and apoplast, respectively. Further validation of GST ( Figure 10A) and GPXL ( Figure 10B) activity was also performed. In general, the pattern of GST activity changes was quite similar to that of GR in cell. Plants showing susceptibility (Col-0)/enhanced susceptibility (rbohD) during interaction with TuMV had slightly upregulated GST between 3 and 7 dpi ( Figure 10A). In the virus-inoculated rbohD plants, the activity of GST was lower than in Col-0 ( Figure 10A). Further validation of GST ( Figure 10A) and GPXL ( Figure 10B) activity was also performed. In general, the pattern of GST activity changes was quite similar to that of GR in cell. Plants showing susceptibility (Col-0)/enhanced susceptibility (rbohD) during interaction with TuMV had slightly upregulated GST between 3 and 7 dpi ( Figure 10A). In the virus-inoculated rbohD plants, the activity of GST was lower than in Col-0 ( Figure 10A). After 7 dpi, GST activity was downregulated in Col-0 and rbohD plants, especially in the latter, in which reduction was 4-fold. In contrast, plants showing resistance (rbohF) and enhanced resistance (rbohD/F) were characterized by increased activity of GST between 3 and 14 dpi ( Figure 10A). The induction of enzymatic activity was the highest (2.12-fold) in rbohD/F plants during TuMV interaction. GPXL was upregulated in Col-0 and rbohD plants inoculated with TuMV between 3 and 14 dpi, by 1.55-fold and 1.72-fold, respectively ( Figure 10B). In turn, GPXL activity was downregulated between 3 and 14 dpi in rbohFand rbohD/F-TuMV interaction by 2.77-fold and 5.33-fold, respectively. Changes in the cellular and apoplastic activity of glutathione metaboli suggested that they perform a potential protecting function against oxidativ Changes in the cellular and apoplastic activity of glutathione metabolism enzymes suggested that they perform a potential protecting function against oxidative stress and lipid peroxidation in infected plants. To verify this hypothesis, we validated lipid peroxidation based on the amount of malondialdehyde (MDA; Figure 11), as described by Hoges et al. [42]. For this analysis, we selected two time points-7 and 14 dpi-which were highly associated with changes in virus levels during infection.
were calculated and analyzed by ANOVA and Tukey's HSD test at p < 0.05. Stat values are indicated by different letters above the bars.
Changes in the cellular and apoplastic activity of glutathione meta suggested that they perform a potential protecting function against oxi lipid peroxidation in infected plants. To verify this hypothesis, we valid dation based on the amount of malondialdehyde (MDA; Figure 11), as de et al. [42]. For this analysis, we selected two time points-7 and 14 dpi-w associated with changes in virus levels during infection.  This validation showed increased levels of MDA (increased lipid peroxidation) in all types of virus-inoculated plants in comparison to mock-inoculated plants at 7 and 14 dpi. However, in virus-inoculated Col-0 (susceptibility) and rbohD (increased susceptibility), the levels of MDA were systemically increased by, respectively, 1.38-fold and 1.33-fold at these timepoints, with the highest increase observed in TuMV-inoculated rbohD plants. In contrast, MDA levels were decreased between 7 and 14 dpi in TuMV-inoculated rbohF (1.75-fold) and rbohD/F (1.53-fold) plants and the levels were slightly higher than in mock-inoculated plants. The presented results of cellular and apoplastic activity of glutathione-associated enzymes and lipid peroxidation indicate that precise control of glutathione in both cell regions is crucial for the modulation of host reaction to TuMV infection. Furthermore, it appears that increased levels of GSH and cellular and apoplastic activity of GGT and GR could be responsible for lower lipid peroxidation in plants exhibiting resistance, such as rbohF and rbohD/F, during ongoing TuMV infection.
To summarize and confirm the relationship between the response of Arabidopsis RbohDand RbohF-deficient mutant to TuMV infection and glutathione metabolism-associated findings, a correlation analysis (based on Pearson's correlation coefficients-PCCs) was performed (Tables S1-S3). PCCs confirmed a high positive correlation between the relative expression of AtGSTU1 and AtGSTU24 genes (Table S1) with GPXL activity (Table S2) and the TuMV concentration in Col-0 and rbohD plants between 7 and 14 dpi. For the same plants, PCCs indicated a high negative correlation between virus concentration and cellular/apoplastic total glutathione localization and content (Tables S1 and S3), associated with changes in the relative expression of AtGGT1, AtGSTU13, and AtGSTU19 genes (Table S1), followed by apoplastic GGT and GR activity with cellular GGT, GSTU, and GR activity (Tables S2 and S3). Unlike rbohFand rbohD/F-TuMV interactions, where the virus limitation had a high negative correlation with the relative expression of AtGSTU1 and AtGSTU24 genes (Table S1) and the activity of GPXL (Table S2), a high positive correlation of all other glutathione-associated parameters was observed at 7 and 14 dpi (Tables S1-S3). Moreover, PCCs data indicated the importance of glutathione content with the expression of glutathione-associated genes and the activity of selected glutathione metabolism enzymes in cellular as well as apoplastic element of resistance and enhanced resistance reaction in A. thaliana rbohFand rbohD/F-TuMV pathosystems.

Discussion
We analyzed the correlation between the response of rbohD and rbohF Arabidopsis thaliana mutants to TuMV and glutathione metabolism. RbohD and RbohF displayed differential contribution to A. thaliana reaction to virus inoculation. The present study confirmed reduced relative expression of TuMV during TuMV-rbohD/F and TuMV-rbohF interactions. In contrast, we observed that rbohD supported TuMV infection even more than Col-0, which led to systemic development of virus infection. Therefore, we postulated that rbohD mutants are characterized by opposite effects in A. thaliana-TuMV interactions in comparison to the and rbohF and rbohD/F. Moreover, when mutants with RbohD defect reacted as more susceptible, RbohD appears to play an important role in the resistance response to potyviruses such as TuMV and PVY [24,25]. However, Rboh can have different contributions in different pathosystems [43]. In accordance with our results, Lukan et al. [44] indicated that RbohD ortholog limited the systemic transport of PVY N -GFP to uninoculated potato leaves. In contrast, RbohD did not show play any role as a resistance factor in rbohD-Alternaria brassicola interactions [45]. Moreover, the response of rbohD and rbohF mutants was not associated with bacteria or necrotrophic fungi promoting spread [15,46]. On the other hand, we demonstrated that rbohF and rbohD/F mutant exhibited strong TuMV limitation. It can be assumed that RbohF promotes virus infection and induces susceptible response to TuMV in Arabidopsis. A similar association between rboh mutants and activated resistance was reported in comparison to the wild type [46,47], but often in the context of immunity establishment. This dominant effect of rbohF mutation clearly indicates that RbohF can act as a susceptibility factor in TuMV infection. The functions of Rbohs as well as the mechanism of the signaling modulation process have been analyzed in studies on plant responses to pathogens, but this topic still seems to be poorly understood in terms of plant-virus interactions. The glutathione induction was found to be much more dynamic in HR than in mock-inoculated plants and compared to susceptible potato response and was accompanied by statistically significant GSH induction and deposition in cell wall in HR reaction [39]. Moreover, the response of Atgstu19and Atgstu24-knockout mutants to TuMV demonstrated that glutathione and glutathione metabolism enzymes play an important role and can be differentially modulated in the plant-virus pathosystem [40]. A quite similar tendency was reported by Hakmaoui et al. [48], who observed that during Nicotiana benthamiana-Pepper mild mottle virus interactions the GSH content decreased and systemic infection was suppressed. This finding is also in line with that of Hernandez [49], who highlighted that glutathione was downregulated to efficiently detoxify ROS in a susceptible reaction to prevent the development of symptoms. Therefore, in our experiment, we examined glutathione content and immunolocalization in Col-0, RbohD-deficient mutant during susceptible TuMV interactions. Analyses of fluorescence signal and immunogold labeling indicating the role of glutathione in rbohD-TuMV and Col-0-TuMV interactions clearly showed that GSH was induced only up to 7 dpi and was located mainly in symplast (usually in cytoplasm). Conversely, between 7 and 14 dpi, its content was significantly decreased. Cellular GSH pool more intensely increased between 1 and 7 dpi in Col-0-TuMV interaction and then in rbohD-TuMV interaction compared to mock-inoculated plants and a dynamic decrease was seen between 7 and 14 dpi in cells as well as in apoplasts. On the other hand, GSSG pool content in apoplast steady increased (1-14 dpi) in rbohD-TuMV and Col-0-TuMV interactions, but a steady decrease in total cell pool was evident in both susceptible interactions. This reaction reflected also in GSH/GSSG ratio which decreased after 7 dpi indicating the high level of oxidative stress. Decrease of GSSG in rbohD-TuMV as well as in Col-0-TuMV interaction was similar to tendency observed in Atgstu19-TuMV interactions. It can be assumed that some part of glutathione pool in apoplast constitutes GSSG during susceptible rbohD-TuMV and Col-0-TuMV interactions. Therefore, we postulated that the susceptible reaction of Col-0 and especially rbohD to TuMV was correlated with the upregulation of TuMV expression, decrease of whole glutathione content (7-14 dpi), and cellular GSSG between 1 and 14 dpi, in contrast to the induction of GSSG, but only in apoplasts. Our observation was similar to the findings from analyses from other pathosystems, in which GSH or/and GSSG pool decrease in mutant or nonmutant plants caused susceptibility reaction to Pseudomonas [50][51][52][53].
A completely different situation was observed for the resistance reaction during rbohF-TuMV and rbohD/F-TuMV interactions. Both interactions were characterized by a highly dynamic increase in glutathione localization (7-14 dpi) in mitochondria and chloroplast, as well as a glutathione content between 1 and 14 dpi after TuMV infection. A similar upregulation of the whole glutathione content was observed during potato resistance reaction to PVY NTN [39] as well as in Atgstu24-knockout mutant-TuMV interactions. Moreover, reduced glutathione pool was significantly induced in whole cells and in apoplast. In turn, despite the upregulation of cellular GSSG, in apoplasts, the oxidized glutathione pool increased only to 7 dpi. Furthermore, even based only on fluorescence signal or immunogold labeling, it seems clear that in resistance reaction of rbohF and rbohD/F to TuMV a strong glutathione localization was associated with whole cell of all Arabidopsis leaf tissues, as well as with apoplast. Our previous ultrastructural observations indicated that Atgstu24-TuMV interaction as resistance reaction was characterized by steady stable significant glutathione deposition in cell walls, which was in line with the findings for rbohD/Fand rbohF-TuMV interactions. Moreover, this statement is consistent with those of Vanacker et al. [54] and Tolin et al. [55], who indicated that in apoplast glutathione acts as a sensor signaling stress conditions, and if the apoplast content is more oxidized, glutathione can take part in adaptation to biotic stress. Our findings from two Potyvirus pathosystems are consistent with the reports of Singh et al. [56], who postulated that the GSSG content was higher in resistant plants compared to susceptible ones. Moreover, Kunstler et al. [57] and Kiraly et al. [58] suggested that a high GSSG concentration can indicate the role of glutathione in the suppression of defense response in oxidative stress during resistance reaction to TMV. As stated by Han et al. [59], stress conditions often change the concentration of glutathione as well as causing its shift toward the GSSG form. Therefore, it can be assumed that resistant rbohFand rbohD/F-TuMV interactions are characterized by intensively reduced virus expression which is strictly correlated with strong induction of total glutathione pool and GSH/GSSG ratio.
Significant differences in the expression profile of selected glutathione metabolism genes were demonstrated in susceptible rbohD-TuMV and resistant rbohFand rbohD/F-TuMV interactions. The presented results of GSH localization with quantification indicated apoplast association, especially in resistance reaction. Therefore, we examined the relative expression level of AtGGT1 gene in resistant and susceptible rbohF-, rbohD-, and rbohD/F mutants during TuMV interactions. The apoplastic glutathione pool is degraded by GGTs [E.C. 2.3.2.2], which cleave the γ-glutamyl moiety of GSH and GSSG [60][61][62]. Arabidopsis genome encodes four GGT genes, but only AtGGT1 is expressed in all plant tissues, but predominantly in leaves and the vascular system [63,64]. Finally, AtGGT1 encoding product, which is localized in apoplast, where it participates in the degradation of GSSG and the recycling of its amino acids which are then used in the synthesis of GSH in symplast [65]. As reported by Masi et al. [65], these amino acids can be translocated to the cytosol for protein synthesis or resynthesis of GSH through round export degradation in the apoplast. In our experiment, we observed the most dynamic and steady induction of the relative expression of AtGGT1 in rbohD/F-TuMV and rbohF-TuMV interaction. These findings indicate that AtGGT1 plays a role in redox signaling from apoplast to internal compartments in symplast. Moreover, the analysis of GGT activity clearly indicated that almost all activity was associated with the apoplast region and was highly upregulated during the most resistant rbohD/F-TuMV interaction and resistant rbohF-TuMV interaction. Although virus inoculation activated GGT at 3-dpi timepoint, the enzyme activity was dynamically downregulated between 3 and 14 dpi during susceptible interaction in rbohD and Col-0. Our findings on resistance reaction are consistent with those of Ohkama-Ohtsu et al. [66], who postulated that GGT1 metabolizes extracellular GSSG to protect from oxidative stress. The activity of GGT was higher in apoplast, while the lower GSSG content was noted in apoplast especially between 7 and 14 dpi.
Several studies have investigated the potential role of GSTs, especially one of the largest Arabidopsis enzymes class tau, but our knowledge regarding their contribution in plant-virus interactions still remains poor. In our experiment, normalized relative expression levels of selected AtGSTU genes were found to be significantly changed in susceptible rbohD-TuMV and Col-0-TuMV as well as resistant rbohFand rbohD/F-TuMV interactions. Moreover, in susceptible Col-0-TuMV and rbohD-TuMV interactions, where virus infection was developed, a highly dynamic induction of AtGSTU1 and ATGSTU24. On the other hand, AtGSTU13 and AtGSTU19 were downregulated especially at 7-14 dpi. Furthermore, AtGSTU13 was significantly induced only till symptoms development-up to 7 dpi, as was reported by Otulak-Kozieł et al. [40]. Therefore, we assume that AtGSTU1 and AtGSTU24 play a role in susceptible reaction to TuMV, and most importantly in enhanced susceptible rbohD-TuMV interaction. In contrast, AtGSTU13 and AtGSTU19 were the most dynamically correlated with resistant rbohF-TuMV interaction, and especially enhanced resistant rbohD/F-TuMV interaction. The differentially regulated relative expression of AtGSTU genes was strictly associated with the modulation of glutathione levels and was correlated with the activity of total GSTs pool. The activity of GSTs was intensively induced during resistant rboh-TuMV reactions, and the most in rbohD/F-TuMV reaction, whereas in susceptible Col-0-TuMV and rbohD-TuMV reactions, their activity decreased. Some individual and limited information about the engagement of certain GSTUs in plant-pathogen interactions is available. The role of GSTU13 in another pathosystem in Glycine max was reported by Zhang et al. [67] in terms of the plant-virus interaction. These authors suggested the contribution of GSTUs in the induction of symptoms development at transcriptional and protein level during Soybean mosaic virus infection. Piślewska-Bednarek et al. [68] documented that GSTU13 plays a role in GSH conjugation, inducing an immune response in A. thaliana against fungal pathogens. Moreover, Pavan Kumar et al. [69] highlighted the accumulation of GST proteins during systemic infection in susceptible interaction between soybean and Mungbean yellow mosaic virus as well as Mungbean yellow mosaic Indian virus. Furthermore, some GSTU representatives were also documented in plant-virus interactions. Chen et al. [70] confirmed that GSTU4 from N. benthamiana upregulated the plant interactions with Bamboo mosaic virus. Méndez-López et al. [71] reported that S. lycopersicon GSTU38 functioned as a Pepino mosaic virus (PepMV) susceptibility factor. Moreover, these authors postulated the dual role of proviral SlGSTU38 in interaction with PepMV capsid protein, and indicated that it may also delay PepMV infection sensing by participating in redox intracellular homeostasis in a nonspecific manner. On the other hand, soybean GSTU10-10 was identified to be specifically induced after systemic infection by SMV. There are also a few reports indicating the role of selected induced GSTUs in resistance reactions. Kiraly et al. [72] observed increased NtGSTU1 between 3 and 6 h after TMV inoculation, accompanied by limited virus replication process. Moreover, Satoh et al. [73] confirmed that all GST genes were upregulated in the resistance response of rice to Rice tungro spherical virus. Furthermore, Wang et al. [74] postulated that overexpression of Triticum aestivum GSTU6 in Arabidopsis induced resistance response to P. syringae pv. Tomato DC300. In contrast to our observations, Chaouch et al. [17] reported that in rboh mutants (rbohF and rbohD single mutants), no significant changes in the relative expression of GSTU24 to actin were found, despite the accumulation of a higher level of glutathione in rbohF than rbohD mutants in response to P. syringae.
These findings indicate that different GSTU genes can be differentially expressed in various pathosystems, resulting in differential regulation of oxidized and reduced glutathione content accompanied by changes in the activity of glutathione metabolism enzymes. Similar to our observations, Wu et al. [75] observed induced activity of GST in response to Sugarcane mosaic virus (ScMV) during interaction with sorghum plants. These authors noted increased GST activity at 3-dpi timepoint of resistance reaction, but in susceptible plant infected with ScMV, the enzyme activity was significantly decreased. In our experiment, strong induction of GST activity was correlated with resistance response in rbohFand especially rbohD/F-TuMV interactions. Díaz-Vivancos et al. [76] also postulated that systemic infection of Plum pox virus was correlated with the downregulation of GST activity. This finding is in line with our observations as well as with the decrease of total GR activity. Similarly, Amari et al. [77] observed a significant decrease in GR in the Prune necrotic ringspot virus (PNRSV)-apricot interaction, while Clarke et al. [78] made the same observation in susceptible White clover mosaic virus-bean interactions. Moreover, compared to that observed in our previous research, resistant rbohFand rbohD/F-TuMV interactions, similar to Atgstu24-TuMV interaction, resulted in higher induction of GR activity [40]. In contrast, rbohDand Col-0-TuMV interactions, similar to the susceptible Atgstu19-TuMV interaction, significantly downregulated GR activity. Therefore, it can be postulated that resistance reaction displayed by rbohF and rbohD/F mutants was associated with the induction of cellular and apoplastic GR activity. Moreover, similar to GGT activity, GR activity was induced only until 7-dpi timepoint in susceptible responses. Therefore, it can be assumed that GST and GR can contribute in symptoms development and the establishment of reaction. Furthermore, the activity of these enzymes in rbohD-TuMV pathosystem was even lower compared to Col-0 plants.
Glutathione peroxidases (GPXs) [E.C.1.11.1.9, E.C. 1.11.1.12, E.C. 1.11.1.15] play a pivotal role in peroxide detoxification [79][80][81]. Plant GPXs are closely related to mammalian ones than to fungal GPXs; thus, they are named "glutathione peroxidase-like enzymes". GPXLs were assumed to act as a link between glutathione-based and tioredoxin (TRX)based detoxifying system but they can also use glutathione [82]. Moreover, as reported by Riyazuddin et al. [83,84], plant GPXLs not only protect cell from stress-induced oxidative damage, but also take part in the regulation process associated with development and plant growth. AtGPXLs may function both as ROS scavengers and as redox transducers, and can link GSH with the TRX redox system. Analysis of GPXLs overexpression showed enhanced tolerance especially under abiotic stress, but decreased resistance under biotic stress [85,86]. In our experiment, the activity of AtGPXLs was found to be significantly downregulated in resistant rbohD/F-TuMV as well as rbohF-TuMV interactions. Moreover, the increase of GPXLs activity was accompanied by the induction of lipid peroxidation in susceptible Col-0-TuMV and especially in rbohD-TuMV interactions. Therefore, lower GPXLs activity was associated with lower lipid peroxidation in resistant plants compared to susceptible rbohdand Col-0-TuMV interactions. Furthermore, Igbal et al. [87] and Navrot et al. [88] pointed out that AtGPXLs efficiently reduced lipid peroxides and plasma membranelocalized AtGPXLs play an important role in maintaining membrane integrity. Similar to our reports, Singh et al. [56] stated that GPX activity was more dynamically induced in a susceptible cultivar during Yellow mosaic virus infection at 20 dpi. Additionally, in contrast to our findings, Kalapos et al. [89] reported high induction of GST genes accompanied by markedly suppressed GPXL, based on the results of transcriptome profiling during ObPV-C. annuum HR. Considering these findings, it can be stated that abiotic stress reaction as well as increased activity of antioxidant enzymes, such as ascorbate peroxidase, GR, and GPXL, is associated with reduced levels of stress markers and MDA [90].
It can be summarized that susceptible rbohD-TuMV and Col-0-TuMV reactions resulting from systemic virus infection were associated with reduced total glutathione localization and content between 7 and 14 dpi. GSH content increased only up to 7 dpi, when symptoms were established, while the apoplastic level of GSSG dynamically increased between 1 and 14 dpi. In susceptible response, glutathione metabolism was highly correlated with the induction of AtGSTU1 and AtGSTU24 genes, which was followed by the decrease of GST and cellular and apoplastic GGT with GR activities between 7 and 14 dpi. In contrast, significant upregulation of GPXL activity was associated with a high rate of lipid peroxidation.
On the contrary, different scenarios were observed in resistant rbohF-TuMV and especially enhanced rbohD/F-TuMV interaction, which revealed significant reduction of virus expression as well as highly dynamic increase of glutathione localization in all leaf tissues with cellular and apoplastic total glutathione content. The resistant reactions were associated with the upregulation of AtGGT1, AtGSTU13, and AtGSTU19 genes, whereas AtGSTU1 participated in symptoms development at the 7-dpi timepoint. The relative expression of these genes and glutathione changes in resistance reactions were followed by highly dynamic increase of GSTs, and cellular as well as apoplastic GR with GGT activities. The presented results clearly indicate that glutathione can act as a significant factor in susceptible rbohD transposon Arabidopsis mutant and in resistant rbohF and rbohD/F mutants during TuMV interaction.

Plant Material, Virus Inoculation, and DAS-ELISA and Molecular Test for TuMV Content
Changes occurring during viral infections induced by TuMV and related to glutathione content were examined in A. thaliana (L.) Heynh wild-type (Col-0) plants and specific mutants-A. thaliana rbohD (rbohD, knockout mutant carrying a single dSpm transposon insertion in RbohD), A. thaliana rbohF (rbohF, knockout mutant carrying a single dSpm transposon insertion in RbohF), and A. thaliana rbohD/F (rbohD/F, double knockout mutant obtained by crossing rbohF and rbohD single mutants) [48]. All homozygous mutant seeds were kindly provided by the Miguel-Angel Torres Laboratory. The plants were sown on peat rings in pods and grown in growth chamber photon flux density of 100 µmol m −2 s −1 (10/14 day/night period), at a relative humidity of 70% and 21 • C. For inducing TuMV inoculation, 19-day-old plants without any lesions and/or alterations were used. The Col-0 and mutant Arabidopsis plants were Wlash inoculated as described by Otulak-Kozieł et al. [40], Tomilson [91], and Walsh and Jenner [92] using the TuMV inoculum (isolate PV-0104 was kindly provided by the Leibniz Institute, Braunschweig, Germany) in phosphate buffer [93]. The leaves of mock-and TuMV-inoculated Col-0 and mutant plants were validated for the presence of virus using qPCR (quantitative polymerase chain reaction). The level of TuMV was determined by performing qPCR analysis of the TuMV-CP gene fragment using the primers presented by Arous et al. [94] according to the procedure of Otulak-Kozieł et al. [40]. TuMV-CP expression was compared with the mean expression of the plant host reference genes, AtEf1α and AtF-Box, as described by Otulak-Kozieł et al. [40]. A total of 60 plants (30 virus-inoculated and 30 mock-inoculated Col-0 and mutant plants of different types) were used for determining TuMV-CP expression and for other analyses (immunofluorescence microscopy, CTCF estimation, HPLC, AtGSTUs and AtGGT1 gene expression profiling, determination of GST/GR/GPXL enzymatic activity). All the analyses were performed in triplicate using a new set of plants every time.

Quantification of Immunofluorescence and Immunogold Localization of Total Glutathione in
TuMV-Inoculated Col-0, rbohD, rbohF, and rbohD/F Arabidopsis Leaves Mock-and TuMV-inoculated Col-0, rbohD, rbohF, and rbohD/F leaves were collected, fixed, and treated as described by Kozieł et al. [95] and Kozieł et al. [96] for immunofluorescence localization. To validate the distribution of total glutathione, quantification of immunolocalization based on CTCF (corrected total cell fluorescence) was performed according to the procedure of Kozieł et al. [97] without DAPI (4 ,6-diamidino-2-phenylindole) staining, as described by Otulak-Kozieł et al. [98]. For detection of glutathione content in immunofluorescent localizations, primary polyclonal rabbit antibodies targeting the total glutathione content (1:100 dilution; Merck, Darmstadt, Germany, catalog number: AB5010) were used as described by Otulak-Koziel et al. [40]. Visualization was performed using respectively secondary anti-rabbit antibodies conjugated with AlexaFluor ® 488 (Jackson ImmunoResearch Europe Ltd., Cambridgeshire, UK, catalog number: 111-545-144). Slides with leaf fragments were imaged on a PROVIS AX70 fluorescence microscope with an Olympus UP90 high-definition camera (Olympus, Warsaw, Poland) using Olympus Cell Sense Standard Software (version 1.18; Olympus, Center Valley, PA, USA). The strength of green fluorescence signal indicating total glutathione localization was validated using a quantitative measuring method, CTCF, as previously described [97,98] with one modification. CTCF was measured for whole cells and separately for apoplast (CTCF apop) regions of the cell. For each combination the cell CTCF and CTCF apop, 30 selected areas of each sample were analyzed based on 50 photos. The levels of fluorescence signal were measured using Fiji (version 2.9.0/1.53 t; National Institutes of Health, Bethesda, MD, USA). Measurements of green immunofluorescence signals obtained using Fiji were calculated to determine CTCF or CTCF apop at 20× magnification with 1.00 zoom factor using a formula previously presented by Otulak-Kozieł et al. [24]. The estimated cell CTCF and CTCF apop values were then analyzed statistically at selected time intervals for all plants by one-factor analysis of variance (ANOVA) [24].
Parallel to immunofluorescence we performed the quantified immunogold localization of total glutathione as was described in [40]. Mock-and TuMV-inoculated Col-0, rbohD, rbohF, and rbohD/F leaves were collected, fixed, and treated as described by Kozieł et al. [40]. For detection of glutathione content, we used the same type of primary polyclonal rabbit antibodies targeting the total glutathione content as in immunofluorescence (1:100 dilution; Merck, Darmstadt, Germany, catalog number: AB5010) [40]. Visualization was performed using secondary antirabbit antibodies conjugated with anti-rabbit antibodies conjugated with 18 nm nanogold particles (Jackson ImmunoResearch Europe Ltd., Cambridgeshire, UK, catalog number: 711-215-152). The immunogold-labeled sections on the grids were examined using a transmission electron microscope [99]. Then, labeling was quantified following the method of Luschin-Ebengreuth and Zechmann [100] in specific cell sections in the case of glutathione and globally in the case of TuMV. Statistical analyses were performed, as described by Otulak-Kozieł et al. [99]. The concentrations of gold particles in specific cell sections and globally were validated using ANOVA and post hoc Tukey's HSD (honestly significant difference) test using Statistica software (version 13.0; StatSoft and TIBCO Software Inc., Palo Alto, CA, USA). ANOVA was used to estimate gold labeling. For the statistical estimation of immunogold labeling, infected and mock-inoculated materials were compared at the 7 and 14 dpi time point. The number of gold particles globally or in-cell compartments was counted in 35 fields (10 µm 2 ) per image. For each combination (mock-inoculated plants and TuMV-inoculated Col-0 and mutant plants), gold particles from 200 photographs were counted to determine the presence of glutathione or TuMV content.

HPLC Analysis of Cellular and Apoplastic Content of GSH, GSSG, and Total Glutathione Accompanied by Estimation of Cytoplasmic Contamination
The cellular and apoplastic GSH, GSSG, and total glutathione (GSH+GSSG) contents in mock-and TuMV-inoculated Col-0, rbohD, rbohF, and rbohD/F plants were measured by reversed-phase HPLC with fluorescence detection, as reported by Kranner [101], using the exact procedure presented by Otulak-Kozieł et al. [24]. Briefly, leaf samples for HPLC were collected as described by Otulak-Kozieł et al. [25]. Some parts of samples were used directly for HPLC analyses [101] and for estimating the activity of leaf glucose-6phosphate dehydrogenase (G6PDH; EC 1.1.1.49), while other parts were used for the extraction of apoplastic components (apoplastic washing fluid, AWF), and then G6PDH activity was determined and HPLC analyses were performed, respectively, as described by Borniego et al. [102] and Ohkama-Ohtsu et al. [41]. The extraction of apoplastic components was performed according to the procedure of Vanacker et al. [103,104] with modifications of Ohkama-Ohtsu et al. [41]. For this, a solution consisting of 50 mM MES-KOH buffer (pH 6.0), 40 mM KCl, and 2 mM CaCl 2 was vacuum-infiltered into leaves. After recovering the apoplastic solution (AWF) by centrifugation at 2900× g, the solution was divided into two parts. One was used for measuring the G6PDH activity, which was used as a marker for estimating the cytoplasmic contamination of samples according to the procedures of Weimar and Rothe [105] and Vanacker et al. [103,104]. The second part was used for determining GSH and GSSG by HPLC after sample deproteination with an equal volume of 10% metaphosphoric acid. G6PDH activity was estimated in AWF and leaf exactly as described by Borniego et al. [102] at room temperature in a 100 µL solution containing 100 mM Tris-HCl (pH 8), 6.7 mM MgCl 2 , 12 mM glucose-6-phosphate, 0.4 mM NADP, and AWF or leaf extracts obtained from all types of plants at 1, 3, 7, and 14 dpi. NADP reduction was followed at 340 nm. The activity was estimated in unit (U), where 1 U is equal to 1 nmol of reduced NADP/min/mg total protein. After the estimation of G6PDH activity in AWF and leaves, we compared the values to validate cytoplasmic contamination of AWF, as described by Ohkama-Ohtsu et al. [43]. It was assumed that if the ratio of G6PDH activity in AWF to that in total leaf extracts was less than 0.25%, then there was no statistically significant cytoplasmic contamination (Table S4). GSH, GSSG, and GSH+GSSG contents were estimated using the results of standards and presented as nmol g −1 FW. The ratio of GSH/GSSG was calculated for particular combination of TuMV infected Col-0 and mutant plants.
For HPLC analyses of G6PDH activity, a total of 60 plants were used (30 virusinoculated and 30 mock-inoculated Col-0 plants and mutant plants of all three types). All analyses were performed in triplicate using a new set of plants every time. The cellular and apoplastic GSH, GSSG, and GSH+GSSG concentrations were analyzed statistically at selected time intervals for all plants by ANOVA.

Isolation of RNA and Genomic DNA (gDNA) for AtGGT1 and Selected AtGSTU Genes in
TuMV-Infected Col-0, rbohD, rbohF, and rbohD/F Plants To estimate the expression of A. thaliana GSTU genes in the plant host, molecular analyses were performed on the samples collected at 3, 7, and 14 dpi. Briefly, leaf samples (0.1 g of each sample) were collected from 30 mock-(buffer) or virus-infected Col-0 plants and mutant plants of different types. Analyses of RNA isolation, purification, and quality as well as confirmation of lack of RNA contamination were performed following previously described procedures [24,40]. The absence of RNA contamination was also verified again by performing reverse transcription PCR using AtEf1α and AtF-Box as reference standards [40], which confirmed the absence of contaminating gDNA. Then, cDNA was synthesized using the NG dART RT Kit (EURx Sp. z o.o., Gdansk, Poland), according to the manufacturer's instructions. Reverse transcription reactions were performed as described by Otulak-Kozieł et al. [40].

Analysis of Relative Expression of AtGGT1 and Selected AtGSTU Genes in TuMV-Infected
Col-0, rbohD, rbohF, and rbohD/F Plants Using qPCR Real-time qPCR was performed using the Bio-Rad CFX96Touch TM apparatus (Bio-Rad, Hercules, CA, USA) and Fast SG qPCR Master Mix (2x) (EURx Sp. z o.o., Gdansk, Poland), as previously described by Otulak-Kozieł et al. [40] for AtEf1α and AtF-Box as reference genes. All qPCR tests were calibrated using previously prepared five-point calibration curves (based on cDNA and gDNA). The analyzed A. thaliana GSTU genes GSTU1 (AtGSTU1, AT2G29490), GSTU13 (AtGSTU13, AT1G27130), GSTU19 (AtGSTU19, AT1G78380), and GSTU24 (AtGSTU24, AT1G17170) were selected based on involvement in TuMV and other pathogen host reaction [40,68,106]. AtGGT1 (γ-glutamyl transpeptidases gene, AT4G39640) was selected because of its location and role in GSH catabolism in cell wall. AtGGT1, AtGSTU1, AtGSTU13, ATGSTU19, and AtGSTU24 were analyzed by qPCR in comparison to reference genes. The expression of these genes in A. thaliana was validated, and complete sequences were acquired from the TAIR database [107]. Based on previously published papers, the primers were chosen for AtGGT1 [41] and AtGSTU [40]. All the primers used in the experiments are shown in Table S5. The starting cDNA solution (used for generating calibration curves) was a four-fold-diluted mix of 10 randomly selected cDNA mixes. An eight-fold-diluted cDNA mix was used to construct the calibration curve for gDNA, while the subsequent calibration points were measured at fourfold dilutions in a 15-µL volume. A 5-µL solution of eight-fold-diluted cDNA mix was added to the reaction mixture. The conditions used for qPCR analyses are presented in Table S6. The molecular  analyses were performed on 60 plants (30 virus-inoculated and 30  2) in Col-0, rbohD, rbohF, and rbohD/F plants, the leaves of these plants were collected at 3, 7, and 14-dpi timepoints after the inoculation of mock or TuMV. A total of 60 plants (30 virus-inoculated and 30 mock-inoculated Col-0 and mutant plants of all three types) were used. All analyses were performed in triplicate using a new group of plants every time. The activity of GST was validated as described by Islam et al. [108] and Otulak-Kozieł [39], and determined based on the enzyme's ability to conjugate GSH and 1-chloro-2,4-dinitrobenzene (CDNB) at 344 nm [109]. The results were presented in U, where 1 U is equal to 1 µmol of CDNB conjugated/min/mg total protein. The activity of GR was determined by measuring the increase in absorbance at 412 nm when 5,5 -dithio-bis(2-nitrobenzoic acid) (DTNB) was reduced by GSH, generated from GSSG, as proposed by Bela et al. [110]. The enzyme activity was calculated as the amount of reduced DTNB in U, where 1 U is equal to 1 µmol of DTNB conjugated/min/mg total protein (ε420 = 13.6 mM −1 cm −1 ). The activity of GPXLs in leaves was also measured spectrophotometrically using cumene hydroperoxide (CHP; Sigma-Aldrich, Darmstadt, Germany) as substrate, as described by Riyazuddin et al. [83]. The compositions of GPXL reaction mixtures were as follows: 4 mM GSH, 0.2 mM NADPH, 0.05 U of GR (from baker's yeast; Sigma-Aldrich, Darmstadt, Germany), 100 µL enzyme extract, and 0.5 mM substrate in a phosphate buffer (0.1 M, pH 7.0) in a total volume of 1 mL. GPXL activity was expressed in U, where 1 U is equal to 1 nmol of converted NADPH/min/mg total protein (ε340 = 6.22 mM −1 cm −1 ). For determining GGT activity, extracts from mock-and virusinoculated leaves were prepared using a modified version of the protocol described by Martin and Slovin [61] with modifications presented by Ohkama-Ohtsu et al. [41]. Briefly, 200 mg of leaf tissue was ground in 400 µL of a buffer containing 1 mM NaCl, 100 mM Tris-HCl (pH 8.0), 1 mM benzamidine, 1 mM 6-amino-n-hexanoic acid, 1 mM phenylmethylsulfonyl fluoride, 1 mM leupeptin, and 0.1% Triton X-100. During extraction, 1 mg of polyvinylpolypyrrolidone (PVPP) was added, and the resulting mixture was incubated on ice for 5 min and then centrifuged at 16,000× g for 15 min at 4 • C. The obtained supernatant was used to determine GGT activity which was measured spectrophotometrically at 410 nm using γ-GPNA (Sigma-Aldrich, Darmstadt, Germany) as substrate and glycyl-glycine (Sigma-Aldrich, Darmstadt, Germany) as acceptor. The determination was performed as described by Tate and Meister [111] and Ohkama-Ohtsu et al. [41] in a 1 mL solution containing 100 mM Tris-HCl (pH 8.0), 5 mM γ-GPNA, and 100 mM glycyl-glycine, with protein added for initiating the reaction. The enzyme activity was measured over a 10-min interval. In accordance with Ohkama-Ohtsu et al. [41], 1 U of GGT activity was assumed to be equal to the formation of 1 nmol p-nitroanilide/min/mg total protein [41]. To check the modulation of the activity of glutathione-associated enzymes, lipid peroxidation in leaf tissues was validated in mock-and virus-inoculated plants at 7 and 14 dpi, as described by Burian et al. [112] and according to the methodology proposed by Hoges et al. [42]. Lipid peroxidation was determined based on the level of the specific reaction product malondialdehyde (MDA). MDA content was measured in the presence of thiobarbituric acid (TBA, Sigma-Aldrich, Darmstadt, Germany). Control reactions were simultaneously performed without the addition of TBA. The absorbance of each sample was measured at λ = 440 nm, 532 nm, and 600 nm to correct nonspecific product contamination. The results of lipid peroxidation were presented as µmol MDA/g FW. All enzymatic activities and lipid peroxidation were analyzed statistically at selected time intervals for all plants by ANOVA.

Validation of Apoplastic GR and GGT Activities in
TuMV-Infected Col-0, rbohD, rbohF, and rbohD/F Leaves Apoplastic solutions were extracted according to the procedure of Ohkama-Ohtsu et al. [41] from mock inoculated Col-0 and mutant plants without the deproteination step. Validation of GR and GGT activities was performed as presented in Section 4.6.

Estimation of PCCs for Elements of Host Reaction
To validate the relationship between cell and apoplast, correlation analyses were performed with the use of Pearson's correlation coefficients (PCCs). We compared the likelihood with levels of TuMV for the following variables: cellular/apoplastic total glutathione localization, cellular/apoplastic content of GSH/GSSG, expression of host genes (AtGGT1, AtGSTU1, AtGSTU13, AtGSTU19, AtGSTU24), and cellular/apoplastic glutathione-associated enzyme activity. Pairwise correlations between the abovementioned variables and the levels of TuMV were determined at 7 and 14 dpi in Col-0 and all mutant plants as described by Wu et al. [113] and Manders et al. [114] using Excel 2019 software (Microsoft, Poland, Warsaw). The results were presented in the form of a heat map generated using the PCC values, where values higher than 0.70 were considered to reflect a strong positive correlation between the analyzed pairs.