Spatiotemporal Changes in Xylan-1/Xyloglucan and Xyloglucan Xyloglucosyl Transferase (XTH-Xet5) as a Step-In of Ultrastructural Cell Wall Remodelling in Potato–Potato Virus Y (PVYNTN) Hypersensitive and Susceptible Reaction

One type of monitoring system in a plant cell is the cell wall, which intensively changes its structure during interaction with pathogen-stress factors. The wall plays a role as a dynamic and controlled structure, although it is not fully understood how relevant these modifications are to the molecular mechanisms during plant–virus interactions. In this work we localise the non-cellulosic polysaccharides such as xyloglucan, xylan (xylan-1) and xyloglucosyl transferase (XTH-Xet5), the enzyme that participates in the metabolism of xyloglucan. This provided us with information about the in situ distribution of the components of the hemicellulotic cell wall matrix in hypersensitive and susceptible potato–PVYNTN interactions. The loosening of the cell wall was accompanied by an increase in xylan depositions during susceptible interactions, whereas, during the hypersensitive response, when the cell wall was reinforced, the xylan content decreased. Moreover, the PVY inoculation significantly redirected XTH-Xet5 depositions, regardless of types of interactions, compared to mock-inoculated tissues. Furthermore, the immunogold localisation clearly revealed the domination of Xet5 in the cell wall and in vesicles in the susceptible host. In contrast, in the resistant host increased levels of Xet5 were observed in cytoplasm, in the cell wall and in the trans-Golgi network. These findings show that the hypersensitive reaction activated XTH-Xet5 in the areas of xyloglucan endo-transglycosylase (XET) synthesis, which was then actively transported to cytoplasm, cell wall and to vacuoles. Our results provide novel insight into cell wall reorganisation during PVYNTN infection as a response to biotic stress factors. These novel findings help us to understand the mechanisms of defence responses that are incorporated into the cell wall signalling network.


Introduction
Plant virus diseases are a major threat to crop production around the world. Viruses are widely known as a large, highly various but also economically important group of plant pathogens, and are the source of an enormous amount of pathological changes in plant tissues [1]. One of the most important plant viruses, Potato Y virus (PVY, genus Potyvirus, family Potyviridae), has been categorised as the fifth most economically damaging virus worldwide [2]. However, PVY is able to infect the wide spectrum In contrast, in the hypersensitive reaction the level of xyl-1 deposition decreased ( Figure 2). In the susceptible cultivar Irys, the green fluorescence signal was observed mainly in the phloem and at lower intensity also in the mesophyll, 10 days past virus inoculation (dpi), after the symptoms appeared ( Figure 1B). Unlike in mock-inoculated Irys, xylan-1 was detected only in xylem tracheary elements ( Figure 1A). During our investigation we analysed the localisation of xylan1/xyloglucan at all time points after mock inoculation-in hypersensitive potato Sárpo Mira at 7, 10 and 14 days post-mock-inoculation and for sensitive potato at 10, 14, 21 day post-mock-inoculation-to check the development of potential influence on xyl-1 deposition. However, no significant differences in localisation of xylan1/xyloglucan were noticed between time points in all mock-inoculated potato plants. The lack of the green fluorescence signal was observed in the potato tissue after incubation In contrast, in the hypersensitive reaction the level of xyl-1 deposition decreased ( Figure 2). In the susceptible cultivar Irys, the green fluorescence signal was observed mainly in the phloem and at lower intensity also in the mesophyll, 10 days past virus inoculation (dpi), after the symptoms appeared ( Figure 1B). Unlike in mock-inoculated Irys, xylan-1 was detected only in xylem tracheary elements ( Figure 1A). During our investigation we analysed the localisation of xylan1/xyloglucan at all time points after mock inoculation-in hypersensitive potato Sárpo Mira at 7, 10 and 14 days post-mock-inoculation and for sensitive potato at 10, 14, 21 day post-mock-inoculation-to check the development of potential influence on xyl-1 deposition. However, no significant differences in localisation of xylan1/xyloglucan were noticed between time points in all mock-inoculated potato plants. The lack of the green fluorescence signal was observed in the potato tissue after incubation with pre-immune serum as well as when primary antibodies were omitted ( Figures 1E and 2E).
Fourteen days after PVY inoculation, the strongest signal was detected in xylem, but also in spongy mesophyll cells ( Figure 1C). After 21 days the xyl-1/xyloglucan signal was observed in both the vascular tissues and in the two types of mesophyll, alongside the tissue reorganisation as a consequence of virus infection ( Figure 1D). The quantitative measuring of the fluorescence confirmed a statistically significant increase of xyl-1 signal in susceptible potatoes at all time points dpi, alongside a decrease in xyl-1 during the hypersensitive response compared to mock-inoculated plants ( Figure 3). Moreover, as the infection progressed in the compatible interaction, the deposition of xylan-1 increased (Figures 1  and 3). On the contrary, in potato Sárpo Mira the xyl-1 signal gradually reduced starting at seven dpi, when the reaction symptoms developed (Figures 2 and 3). Seven dpi xyl-1 was found mainly in xylem cells ( Figure 2B), whereas after 10 days a weak signal also appeared in the epidermis ( Figure 2C), and the lowest signal was observed in phloem cells 14 days after inoculation ( Figure 2D). with pre-immune serum as well as when primary antibodies were omitted ( Figures 1E and 2E). Fourteen days after PVY inoculation, the strongest signal was detected in xylem, but also in spongy mesophyll cells ( Figure 1C). After 21 days the xyl-1/xyloglucan signal was observed in both the vascular tissues and in the two types of mesophyll, alongside the tissue reorganisation as a consequence of virus infection ( Figure 1D). The quantitative measuring of the fluorescence confirmed a statistically significant increase of xyl-1 signal in susceptible potatoes at all time points dpi, alongside a decrease in xyl-1 during the hypersensitive response compared to mock-inoculated plants ( Figure 3). Moreover, as the infection progressed in the compatible interaction, the deposition of xylan-1 increased (Figures 1 and 3). On the contrary, in potato Sárpo Mira the xyl-1 signal gradually reduced starting at seven dpi, when the reaction symptoms developed (Figures 2 and 3). Seven dpi xyl-1 was found mainly in xylem cells ( Figure 2B), whereas after 10 days a weak signal also appeared in the epidermis ( Figure 2C), and the lowest signal was observed in phloem cells 14 days after inoculation ( Figure 2D).   Immunogold labelling revealed that in mock-inoculated potato Irys the xyl-1 epitopes appeared in the endoplasmic reticulum and in the trans-Golgi network or in other vesicular and membranous structures ( Figure 4A,B). In compatible PVY-Irys interaction, xyl-1 appeared in the cell wall around plasmodesmata in the phloem tissue ( Figure 4C), as well as in the mesophyll ( Figure 4E,G). Gold deposition was often associated with vacuoles ( Figure 4C-G) and vesicular structures ( Figure 4E,F), also with paramular bodies between cell wall and plasmalemma ( Figure 4D). Additionally, xyl-1 epitopes were deposited in the area undergoing necrotisation 21 dpi, especially alongside PVY particles or inclusion bodies ( Figure 4F). Quantification by immunogold of xylan epitopes revealed an increase of xyl-1 in potato Irys infected with PVY NTN (Table 1). Moreover, during compatible interaction a statistically significant amount of xylan was detected in the cell wall, in vacuoles containing vesicles, as well as in the trans-Golgi network and the endoplasmic reticulum. In all these compartments the deposition was much higher than in mock-inoculated Irys plants. The above results clearly indicate that xylan/xyloglucan was activated as a result of compatible potato-PVY interaction, but also that its distribution and deposition were visibly changed compared to healthy plants.
The immunogold localisation of xylan in the hypersensitive potato Sárpo Mira confirmed the above fluorescence analyses. The gold deposition in mock-inoculated resistant plants was higher than in susceptible potatoes, similar to the fluorescence data ( Figures 4A,B and 5A,B, Table 1). In healthy Sárpo Mira tissues the xyl-1 was detected mainly in the cell wall and in vacuoles ( Figure  5A,B). When the symptoms of hypersensitive response to PVY were visible, the xylan was observed in the vascular tissues, in xylem tracheary elements with xylem parenchyma as well as in phloem sieve elements ( Figure 5C,D). After hypersensitive response, xylan was noticed in mesophyll tissue close to the plasmodesmata ( Figure 5E,F), but also associated with multivesicular bodies ( Figure 5E). Additionally, xylan accumulated in the collenchyma at lower intensity than in vascular bundles or the mesophyll ( Figure 5G). Regardless of the type of interaction, a lack of gold depositions was noticed in those tissue sections that were incubated with pre-immune serum or when primary antibodies were omitted ( Figures 4H and 5H  Immunogold labelling revealed that in mock-inoculated potato Irys the xyl-1 epitopes appeared in the endoplasmic reticulum and in the trans-Golgi network or in other vesicular and membranous structures ( Figure 4A,B). In compatible PVY-Irys interaction, xyl-1 appeared in the cell wall around plasmodesmata in the phloem tissue ( Figure 4C), as well as in the mesophyll ( Figure 4E,G). Gold deposition was often associated with vacuoles ( Figure 4C-G) and vesicular structures ( Figure 4E,F), also with paramular bodies between cell wall and plasmalemma ( Figure 4D). Additionally, xyl-1 epitopes were deposited in the area undergoing necrotisation 21 dpi, especially alongside PVY particles or inclusion bodies ( Figure 4F). Quantification by immunogold of xylan epitopes revealed an increase of xyl-1 in potato Irys infected with PVY NTN (Table 1). Moreover, during compatible interaction a statistically significant amount of xylan was detected in the cell wall, in vacuoles containing vesicles, as well as in the trans-Golgi network and the endoplasmic reticulum. In all these compartments the deposition was much higher than in mock-inoculated Irys plants. The above results clearly indicate that xylan/xyloglucan was activated as a result of compatible potato-PVY interaction, but also that its distribution and deposition were visibly changed compared to healthy plants.
The immunogold localisation of xylan in the hypersensitive potato Sárpo Mira confirmed the above fluorescence analyses. The gold deposition in mock-inoculated resistant plants was higher than in susceptible potatoes, similar to the fluorescence data ( Figures 4A,B and 5A,B, Table 1). In healthy Sárpo Mira tissues the xyl-1 was detected mainly in the cell wall and in vacuoles ( Figure 5A,B). When the symptoms of hypersensitive response to PVY were visible, the xylan was observed in the vascular tissues, in xylem tracheary elements with xylem parenchyma as well as in phloem sieve elements ( Figure 5C,D). After hypersensitive response, xylan was noticed in mesophyll tissue close to the plasmodesmata ( Figure 5E,F), but also associated with multivesicular bodies ( Figure 5E). Additionally, xylan accumulated in the collenchyma at lower intensity than in vascular bundles or the mesophyll ( Figure 5G). Regardless of the type of interaction, a lack of gold depositions was noticed in those tissue sections that were incubated with pre-immune serum or when primary antibodies were omitted ( Figures 4H and 5H).     Quantification of xylan-1 antigen by immunogold in the host cell compartments revealed a statistically significant decrease in xyloglucan deposition during hypersensitive reaction, compared to controls ( Figure 5, Table 1). As an effect of virus infection, the xyl-1 epitope reached the highest level in the cytoplasm and in vacuoles but was at a lower level in the cell wall and in the trans-Golgi network, whereas, in the control mock-inoculated plants the depositions took place mainly in vacuoles and in the cell wall.
Moreover, the distribution of xyl-1/xyloglucan was generally different after compatible as opposed to incompatible interactions. The content of xyl-1 in susceptible potatoes was induced by PVY (Figures 1, 3 and 4, Table 1). On the contrary, in hypersensitive potato the level of xyl-1 decreased (Figures 2, 3 and 5, Table 1). Also, after compatible interaction the most intense deposition occurred in cell walls and in vacuoles, and was associated with the membranous compartments such as ER, trans-Golgi network or vesicular structures ( Figure 4, Table 1), whereas in hypersensitive reaction the highest level of xyloglucan was found in the cytoplasm and in vesicular structures ( Figure 5, Table 1). These data clearly indicate that virus infection can cause varied xylan-1 distribution in potato tissues, depending upon the level of anti-PVY NTN resistance.

207) during Compatible and Incompatible PVY NTN -Potato Interactions
For compatible PVY-potato (cv. Irys) interaction, a strong green fluorescence signal of XTH-Xet5 was observed predominantly in the vascular bundle in mock-inoculated plants, whereas a weaker signal was seen from mesophyll cells 10 days post-inoculation ( Figure 6A,B). The localisation of XTH/Xet5 at all time points after mock-inoculation-in hypersensitive potato Sárpo Mira at 7, 10 and 14 days post-mock-inoculation and for sensitive potato at 10, 14, 21 days post-mock-inoculation-were analysed to check the potential developmental influence on XTHE/Xet5 deposition. However, there were no significant differences in the localisation of Xet5 between time points of all mock-inoculated potato plants. Fourteen dpi a XTH-Xet5 signal was detectable almost exclusively in xylem tracheary elements, similar to 21 days post-inoculation, where besides xylem cells the green fluorescence was noticed in the epidermis outer cell wall (Figure 6 C,D).
The immunofluorescence analyses of the hypersensitive reaction tissue provided a significantly different picture. In mock-inoculated hypersensitive cv. Sárpo Mira, the XTH-Xet5 antigen was detected only in the cell wall of xylem tracheary elements, a different pattern than in cv. Irys ( Figure 7A). Seven days after PVY inoculation of Sárpo Mira, the green fluorescence signals were detected not only in both xylem and in phloem tissue, but also in the cell wall of mesophyll cells ( Figure 7B,C). Moreover, starting from 10 days post-inoculation the XTH-Xet5 antigen signals were noticed in all leaf and petiole cells, the most intensely visible in both vascular tissues ( Figure 7D,E). The fluorescence signal was not detectable in tissue sections incubated with the pre-immune serum or when primary antibodies were omitted ( Figures 6E and 7F). The immunofluorescence-based quantitative measurement of total cell fluorescence (CTCF) clearly indicated the statistically significant stepwise increase of the XTH-Xet5 antigen signal during hypersensitive response. To the contrary, during susceptible potato-PVY interaction the XTH-Xet5 signal gradually decreased (Figure 8).     In cv. Irys, the XTH-Xet5 antigen was detected by immunogold labelling mainly in cell wall, along endoplasmic reticulum and the vacuoles ( Figure 9A). As a result of PVY inoculation, the XTH-Xet5 antigen was deposited firstly in a loosened cell wall around plasmodesmata alongside the virus inclusions, and thereafter in vesicular structures ( Figure 9B,C,E). In vascular tissues the gold was mainly associated with sieve elements, phloem parenchyma as well as with xylem tracheary elements and xylem parenchyma ( Figure 9C,D), especially when virus inclusions and particles were present. Moreover, gold granules were found not only in the cell wall or along the plasmalemma, but even in the intercellular space ( Figure 9F). In resistant potato Sárpo Mira, the transferase epitopes were detected mostly in the cell wall area, but also in vacuoles ( Figure 10A). After hypersensitive response to PVY infection, similar to compatible interaction, the presence of XTH-Xet5 antigen localised to the cell wall, between plasma membranes and the cell wall of the necrotised area, or even in the intercellular space ( Figure 10B,C,E). In the phloem the preferable places of gold deposition were vesicular/membranous structures and the plasmodesmata rounded areas ( Figure 10E). In xylem the depositions were noticed inside xylem tracheary elements and in xylem parenchyma ( Figure 10D). The gold granules were absent in control sections (Figures 9G and  10F). The immunogold quantification clearly demonstrated a significantly higher deposition of XTH-Xet5 after hypersensitive response to PVY infection than after the compatible interaction (Figures 9 and 10, Table 1). Interestingly, the xyloglucan transferase protein reached much higher levels of XTH in mock-inoculated Irys potato than in resistant Sárpo Mira. A statistically significant deposition of gold granules in mock-inoculated Sárpo Mira was detected only in the cell wall and in vacuoles. After PVY NTN inoculation, the highest level of XTH-Xet5 in Sárpo Mira was in the cytoplasm, cell wall and trans-Golgi network, whereas in Irys the XTH-Xet5 was found in vesicular structures and in the cell wall (Table 1)  In cv. Irys, the XTH-Xet5 antigen was detected by immunogold labelling mainly in cell wall, along endoplasmic reticulum and the vacuoles ( Figure 9A). As a result of PVY inoculation, the XTH-Xet5 antigen was deposited firstly in a loosened cell wall around plasmodesmata alongside the virus inclusions, and thereafter in vesicular structures ( Figure 9B,C,E). In vascular tissues the gold was mainly associated with sieve elements, phloem parenchyma as well as with xylem tracheary elements and xylem parenchyma ( Figure 9C,D), especially when virus inclusions and particles were present. Moreover, gold granules were found not only in the cell wall or along the plasmalemma, but even in the intercellular space ( Figure 9F). In resistant potato Sárpo Mira, the transferase epitopes were detected mostly in the cell wall area, but also in vacuoles ( Figure 10A). After hypersensitive response to PVY infection, similar to compatible interaction, the presence of XTH-Xet5 antigen localised to the cell wall, between plasma membranes and the cell wall of the necrotised area, or even in the intercellular space ( Figure 10B,C,E). In the phloem the preferable places of gold deposition were vesicular/membranous structures and the plasmodesmata rounded areas ( Figure 10E). In xylem the depositions were noticed inside xylem tracheary elements and in xylem parenchyma ( Figure 10D). The gold granules were absent in control sections (Figures 9G and 10F). The immunogold quantification clearly demonstrated a significantly higher deposition of XTH-Xet5 after hypersensitive response to PVY infection than after the compatible interaction (Figures 9 and 10, Table 1). Interestingly, the xyloglucan transferase protein reached much higher levels of XTH in mock-inoculated Irys potato than in resistant Sárpo Mira. A statistically significant deposition of gold granules in mock-inoculated Sárpo Mira was detected only in the cell wall and in vacuoles. After PVY NTN inoculation, the highest level of XTH-Xet5 in Sárpo Mira was in the cytoplasm, cell wall and trans-Golgi network, whereas in Irys the XTH-Xet5 was found in vesicular structures and in the cell wall (Table 1).  in cytoplasm at almost the same level as in the cell wall and around the trans-Golgi network. This is unlike in the susceptible host Irys, where transferase was mainly deposited in the cell wall and in vesicular structures, but at a much lower level then in HR. Moreover, the location in the cytoplasm in Irys was at the lowest level. The quantification statistics implied the PVY infection to modify the distribution of xyloglucan transferase in the tissues and cell compartments of Sárpo Mira and Irys plants, depending on the type of virus-host interactions.  Moreover, the deposition of XTH-Xet5 epitopes after compatible interaction was significantly lower, compared to mock-inoculated plants. During hypersensitive response, XTH-Xet5 dominated in cytoplasm at almost the same level as in the cell wall and around the trans-Golgi network. This is unlike in the susceptible host Irys, where transferase was mainly deposited in the cell wall and in vesicular structures, but at a much lower level then in HR. Moreover, the location in the cytoplasm in Irys was at the lowest level. The quantification statistics implied the PVY infection to modify the distribution of xyloglucan transferase in the tissues and cell compartments of Sárpo Mira and Irys plants, depending on the type of virus-host interactions.

Discussion
The cell wall components contribute to plant growth, development and cell interaction with different stimuli such as the cascade of signal transduction factors, both biotic and abiotic [17]. The cell wall acts as a first line of defence. The composition of the cell wall can serve as a molecular signature of the environmental monitoring as many components undergo synthesis and/or hydrolysis during stress [9]. Therefore, the function of different cell wall components and the question of how they interact with each other and with exogenous factors (such as pathogens) have been a subject of extensive research for many years [18,19].
Some of these questions may be answered based on the first published transcriptome dataset of the response to different groups of pathogens [18][19][20]. These analyses revealed the effects of plant-virus interactions [14,21], especially about the reinforcing of the cell wall structure. Our knowledge of the mechanisms of resistance to plant pathogens was extended by RNAseq analyses of the whole transcriptomes [14]. In general, the expression of genes that coded for key protein/enzymes participating in the synthesis of cell wall components can be significantly repressed by virus infection [22,23].
More recently, it has been reported that expression of genes modifying cell wall could be actively enhanced, as response to the stress-related cell wall signalling [21,24]. The current view is that the plant cell wall forms a functional network that is able to resist via a dynamic extracellular complex of polysaccharides, together with glycoproteins and the modifying enzymes [25]. Moreover, the metabolism of cell wall polysaccharides can regulate the balance between biosynthesis and degradation, and the shift of this balance can lead to structural changes in the cell wall [25].
The unique composition of hemicellulosic polysaccharides, including xyloglucans, xylans and the polysaccharide-modifying enzymes, can be modified due to the response to the pathogen-related stress. Bacete et al. [9] demonstrated that the metabolism of xyloglucan plays an important role in the expansion of cell wall, affecting the pathogen invasion. The microbial pathogens not only break the cell wall, but the β-1,4 xylanase is induced to degrade xylans producing endoxylanases, and this mechanism is also relevant to viral infections [26]. Recent microarray analysis and the transcriptome data demonstrate the regulation of xyloglucan metabolism during different types of plant-virus interactions [14,27]. Indeed, xylan biosynthesis as well as the metabolism of glucuroxylan seem to represent the most common down-regulated functional categories after the transcriptomics analyses of the Tobacco etch virus (Potyvirus, TEV) infected Arabidopsis plants [27]. Rice stripe virus (RSV) suppressed the hypersensitive reaction in rice resistant varieties, with downregulated pathogenesis-related proteins, whereas changes in the xyloglucan endo transglycosylase/hydrolase could lead to cell wall strengthening, conditioning the resistance mechanism to RSV in rice [14].
In this work we analysed the effects of Potyvirus infection on selected hemicelluloses, non-cellulosic cell wall polysaccharides in both susceptible and resistant potato hosts. Indeed some papers mentioned above concentrated on gene regulation and transcriptomes being triggered by host-virus interactions. To gather complex information about the role of xylan/xyloglucan in cell wall integrity during virus infection, here we have analysed in situ localisation and the distribution of selected components of xyloglucan metabolism such as xylan-1/xyloglucan and XTH-Xet5 after compatible and hypersensitive responses during PVY NTN -potato infection.
Our findings reveal that xyl-1/xyloglucan was induced after PVY NTN -infection in susceptible potato cv. Irys. Moreover, this tendency correlated closely with the appearance of the necrotic symptoms, being fully developed 21 days after virus inoculation. Unexpectedly, the observed dynamic changes in deposition of the major non-cellulosic polymer correlated well with the observed distribution of xyloglucans after reconstruction of the cell wall during the formation of syncytia, induced by nematodes [28]. The most intense deposition of xylan/xyloglucan was in vascular bundles (xylem and phloem) and in mesophyll, but also in the virus-induced necrotising areas. Similar data were reported by Northcote et al. [29], detecting xylan in the growing cell wall and in the xylem differentiating and thickening cell walls plus in mesophyll cells. Along these lines Northcote et al. [29] observed gold location in the cell wall, and a relatively similar intensity in and around vesicles, as well as in the trans-Golgi networks.
Xyloglucan can be found in almost every plant species and is the most abundant hemicellulose of the primary wall [30]; it is closely linked to pathogen-induced changes. Our observations confirmed that cell wall loosening was accompanied by an increase in xylan deposition in the PVY-infected potato Irys. These results are also linked to the increase of viral presence measured by ELISA (Table S1). In contrast, during the hypersensitive response, the cell wall was reinforced, while the xylan content decreased, similar to the interactions of nematodes on potatoes or on Arabidopsis. The latter has roots deprived of the secondary cell wall and thus does not carry xylans [31]. Moreover, our findings clearly indicate that the distribution of xyl-1/xyloglucan depends upon the types of reaction to the PVY NTN infection. That is, after compatible interaction the xylan epitope dominated in cell wall and in vacuoles, whereas after hypersensitive response the epitope was redistributed mainly to the cytoplasm and to vesicles.
The plant cell wall contains numerous enzymes that modify polysaccharides [32], and xyloglucan endotransglycosylase/hydrolase (XET, XTH) is an essential constituent, especially in the primary cell wall, participating in wall construction and elongation [33]. XET cleaves the xyloglucan chain endolytically and forms a covalent polysaccharide-enzyme complex. XTH/Xet is commonly thought to participate in cell wall loosening during plant development and wall expansion as well as increase rigidity during pathogen intrusion [34]. Transgenic Arabidopsis expressing Capsicum annuum XTH revealed distorted leaves, carrying irregular cell pattern in cross sections [35]. Further analysis suggested the role of XTH in cell wall remodelling. Zea mays XET1 is likely involved in the wall extension via hydrolysis and rejoining the xyloglucan molecules [36]. Additionally, xyloglucan holds seven glucosidic linkages and their formation requires different enzymes, including a complex of one xyloglucansynthase and fucosyltransferase, two galactosyltransferases and three xylosyltransferases [37].
Our analyses of compatible and incompatible interactions of PVY with potato cultivars focused on the deposition of xyloglucan xylosyl transferase (XTH-Xet5). Based on seven Arabidopsis genes encoding the xyloglucan xylosyltransferase, which are not fully characterised [38], XTH-Xet5 allegedly belongs to group GH16 of glycoside hydrolases. Moreover, Xet5 catalyses in vitro the formation of covalent linkages between xyloglucans and cellulosic substrates or even between xyloglucan and (1,3-and 1,4-) D-glucan [39,40]. It is thus possible that XTH-Xet5 is responsible for the linking of different polysaccharides in vivo, which consequently affects the strength of cell walls, but also decreases their flexibility and porosity. Our data follow these general considerations. Just as we observed the most intense XTH-Xet5 deposition during hypersensitive reaction, it has also been observed during the strengthening of cell walls in the cv. Sárpo Mira [11]. The analysis of cell wall metabolism during infection of papaya with PMeV (Papaya meleira virus) showed that this interaction upregulated xyloglucan endotransglycosylase activity [41]. Also, similar to our observations and to the data on PMeV infection, xyloglucan endotransglycosylases were affected by Potato leafroll virus (PLRV) infection in potatoes [42]. Moreover, higher deposition of XTH-Xet5 was correlated with a decrease in the virus presence measured by ELISA (Table S1).
The results from two other systems are also relevant to PVY NTN interaction with susceptible and resistant potatoes. Namely, in the susceptible A. thaliana-Turnip mosaic virus (TuMV, Potyvirus) system the dramatic downregulation of XTH6 was observed in the wild-type plant or even in a defective silencing mutant, with both the qRT-PCR and the microarray assays [43]. Our results on the lowered level of XTH-Xet5 in susceptible potatoes were also observed in Arabidopsis, compared to mock-inoculated plants. As in Rose et al. [34], the transfer of xyloglucan was observed, also shifting the function of XTH toward the breakdown of the xyloglucan-cellulose network [43]. Our data from the PVY NTN -Sárpo Mira hypersensitive interaction also suggest the potential involvement of XTH in the hydrolysis of the xyloglucan-cellulose network. We have shown that the higher activity of XTH-Xet5 paralleled the lower level of xyl-1/xyloglucan. We have also previously observed a lowered level of CesA4 during HR than for susceptible interaction.
Secondly, the five plant genes involved in cell wall metabolism were identified to operate during PVY infection in the resistant VAM seedlings of Nicotiana tabacum [44]. The genes associated with the cell wall structure were downregulated, while those participating in the remodelling tend to be upregulated. On the contrary, the identified xyloglucan endotransglucosylase hydrolase (JZ 897688) was downregulated at 12 h post-inoculation, but upregulated starting from day one after infection [44]. Our findings, similar to Chen et al.'s [44], reveal a gradually increasing level of XTH-Xet5 during the resistant potato-PVY NTN interaction. However, during the susceptible interaction there was some decrease in the deposition of XTH-Xet5. Zheng et al. [14] came to a similar conclusion based on the comparative transcriptome analyses of the XTH message. The xyloglucan xylosyltransferase in Arabidopsis is potentially expressed in all plant tissues, with a strong presence in roots, stems, but also leaves [37]. In Arabidopsis, the immunolocalisation detected this activity in roots, vascular bundles, and in the root epidermis and hairs. Similarly, by using immunolocalisation we demonstrated a strong XTH-Xet5 signal in both vascular tissues and the epidermis in the PVY-infected potato, but during hypersensitive response all leaves showed strong deposition starting 10 days after inoculation. A similar conclusion was made by Antosiewicz et al. [45] based on an abiotic stress stimulation experiment, where the xyloglucan endotransglycosylase was mainly detected in vascular tissues and the epidermis. Like during PVY-potato interaction, the wind stimulated the deposition of XET mainly in the xylem, but also in the mesophyll of A. thaliana. XET was also localised in middle lamella or even in the intercellular space [45,46].
Ultrastuctural analysis localised XTH-Xet5 in the susceptible potato cultivar in the cell wall and in membranous structures, but also in the ER, trans-Golgi network, and in vacuoles with vesicles. Zabotina et al. [47] also observed high levels of the enzyme in the trans-Golgi network. Accumulation of XTH-Xet5 in membranous structures seems to be logical, considering that the endomembrane complex is known to function as an orchestrated system that delivers the Golgi-derived and endocytic vesicles that carry the cell wall and the cell membrane components [48].
Additionally, PVY infection significantly redirected XTH-Xet5 depositions, regardless of the type of interaction, versus the mock-inoculated tissues. The immunogold analysis clearly indicates that the deposition of XTH-Xet5 in resistant and susceptible interactions significantly differed from each other. In the susceptible host XTH-Xet5 dominated in the cell wall and vesicles, whereas in the resistant host it dominated in the cytoplasm, cell wall and trans-Golgi network, generally at a much higher level. Our findings demonstrate that the hypersensitive reaction induces the XTH-Xet5 more actively in the areas of XET synthesis and transports the enzyme more actively to the cytoplasm, cell wall and vacuoles.

Plant Material and Virus Inoculation
Potato plants (Solanum tuberosum) of two cultivars with different resistance levels (Irys (PVY NTN resistance score 5.5 in a 1-9 scale) and Sárpo Mira (resistance score 9) [49]) were acquired from IHAR-PIB, Plant Breeding and Acclimatisation Institute, Bonin Research Center, Bonin, Poland. Plants were grown and inoculated mechanically as previously presented [11,50] at the four-leaf stage with the NTN strain of PVY. Potato cv. Sárpo Mira developed a hypersensitive necrotic response visible at 7 days post-inoculation. This reaction is conferred by the Ny-Smira gene located on the long arm of the potato IX chromosome [51]. Hypersensitive reaction symptoms on inoculated leaves appeared 7 days post-inoculation. Cultivar Irys developed systemic necrosis visible at 10 days post-inoculation. Leaves from both PVY NTN -infected plants were collected at three different time intervals to categorise the reaction as susceptible or resistant. In the case of susceptible potatoes (cv. Irys), the leaves were collected 10, 14 and 21 days post-PVY NTN inoculation (dpi), whereas the resistant potato cv. Sárpo Mira leaves were collected after 7, 10 and 14 dpi. Different starting points for collecting the plant material were chosen because of differences in the course of viral infection on either cultivar. Healthy leaves of both cultivars (used as controls) were mock-inoculated with phosphate buffer, at 10 dpi (susceptible) and 7 dpi (resistant). To double-check the potato cultivars, all analysed potato plants were tested for the presence of PVY using ELISA [52]. ELISA testing was performed according to [53], and the results are presented in Table S1. Absorbance was measured at 405 nm. Mean values for ELISA titres were assessed for 50 leaves from each combination (Table S1).

Immunofluorescence Localisation and the Assessment of the Quantitative Fluorescence Signal by Using the Corrected Total Cell Fluorescence Method (CTCF)
Fragments of leaves from PVY NTN and mock-inoculated potato plants (at the abovementioned time intervals) were fixed and embedded in butyl-methyl-methacrylate (BMM) resin according to a procedure described previously [50], with the following modifications. Acetone was used to remove the BMM from 2 µm sections and stuck to silane slides (Thermo-Fisher Scientific, Warsaw, Poland). A further immunofluorescence procedure/analysis was carried out exactly as described in [11]. During analyses we have used two sets of primary and secondary antibodies. For localisation of XTH-Xet5 the primary rabbit antibodies were acquired from Agrisera ( Int. J. Mol. Sci. 2018, 19, x intervals to categorise the reaction as susceptible o Irys), the leaves were collected 10, 14 and 21 da resistant potato cv. Sárpo Mira leaves were collecte for collecting the plant material were chosen becaus either cultivar. Healthy leaves of both cultivars phosphate buffer, at 10 dpi (susceptible) and 7 dpi all analysed potato plants were tested for the prese performed according to [53], and the results are pre 405 nm. Mean values for ELISA titres were assessed

Immunofluorescence Localisation and the Assessme the Corrected Total Cell Fluorescence Method (CTCF)
Fragments of leaves from PVY NTN and mock-i time intervals) were fixed and embedded in butylprocedure described previously [50], with the follo the BMM from 2 µm sections and stuck to silane sl A further immunofluorescence procedure/analysi During analyses we have used two sets of prima XTH-Xet5 the primary rabbit antibodies were acqu Vӓnӓs , Sweden), whereas the secondary anti-rab provided by Jackson ImmunoResearch Europ xylan1/xyloglucan, the primary mouse antibodie recognises glycan group of xylan-1, binds to xyla Sweden), while the secondary anti-mouse IgG con ImmunoResearch Europe Ltd. The controls consist serum. An Olympus AX70 Provis (Olympus Poland an Olympus UC90 HD camera (Olympus Poland) w acquired using Olympus Cell Sense Standard Softw 1.18). After gaining florescent images, further quan performed. As a first step, regions (cell wall or p xylan-1 or Xet5 epitopes were marked and outli program (Version 1.52e, National Institutes of Heal of green immunofluorescence signal were done wi plants, and the infected Irys and Sárpo Mira plant signal were calculated in the form of corrected tot with 1.00 zoom factor [54,55] using the following fo CTCF = Integrated Density − (Area of Backgro Estimated CTCF values were then analysed sta of reaction to PVY NTN by using the one-factor analy analyses enabled us to find the values of statistical sig Xet5. Furthermore, the mean CTCF values were eval hoc Tukey HSD testing in STATISTICA software (S USA, version 13.0).

Quantitative Immunogold Localisation by Direct E
Potato leaves of both cultivars were fixed and transmission electron microscopy (TEM) according PVY NTN -infected or mock-inoculated plants were mo , Sweden), whereas the secondary anti-rabbit IgG conjugated with AlexaFluor ® 488 was provided by Jackson ImmunoResearch Europe Ltd. (Cambridgeshire, UK). To localise xylan1/xyloglucan, the primary mouse antibodies were obtained from Agrisera (CCRC-M108, recognises glycan group of xylan-1, binds to xylans and to non-fucosylated xyloglucans, 19, x intervals to categorise the reaction as susceptible or resistant. In the case of susceptible po Irys), the leaves were collected 10, 14 and 21 days post-PVY NTN inoculation (dpi), w resistant potato cv. Sárpo Mira leaves were collected after 7, 10 and 14 dpi. Different star for collecting the plant material were chosen because of differences in the course of viral in either cultivar. Healthy leaves of both cultivars (used as controls) were mock-inocu phosphate buffer, at 10 dpi (susceptible) and 7 dpi (resistant). To double-check the potat all analysed potato plants were tested for the presence of PVY using ELISA [52]. ELISA t performed according to [53], and the results are presented in Table S1. Absorbance was m 405 nm. Mean values for ELISA titres were assessed for 50 leaves from each combination

Immunofluorescence Localisation and the Assessment of the Quantitative Fluorescence Signal the Corrected Total Cell Fluorescence Method (CTCF)
Fragments of leaves from PVY NTN and mock-inoculated potato plants (at the above time intervals) were fixed and embedded in butyl-methyl-methacrylate (BMM) resin acc procedure described previously [50], with the following modifications. Acetone was used the BMM from 2 µm sections and stuck to silane slides (Thermo-Fisher Scientific, Warsaw A further immunofluorescence procedure/analysis was carried out exactly as describ During analyses we have used two sets of primary and secondary antibodies. For loca XTH-Xet5 the primary rabbit antibodies were acquired from Agrisera ( Vӓnӓs , Sweden), whereas the secondary anti-rabbit IgG conjugated with AlexaFluor provided by Jackson ImmunoResearch Europe Ltd. (Cambridgeshire, UK). T xylan1/xyloglucan, the primary mouse antibodies were obtained from Agrisera (CC recognises glycan group of xylan-1, binds to xylans and to non-fucosylated xylogluca Sweden), while the secondary anti-mouse IgG conjugated with AlexaFluor ® 488 was fro ImmunoResearch Europe Ltd. The controls consisted of mock-inoculated tissue and a p serum. An Olympus AX70 Provis (Olympus Poland, Warsaw, Poland) with a UM61002 fi an Olympus UC90 HD camera (Olympus Poland) were used for fluorescence imaging. Im acquired using Olympus Cell Sense Standard Software (Olympus, Center Valley, PA, US 1.18). After gaining florescent images, further quantitative measuring of the fluorescence performed. As a first step, regions (cell wall or protoplast) with green fluorescence s xylan-1 or Xet5 epitopes were marked and outlined by using a special marker in th program (Version 1.52e, National Institutes of Health, Bethesda, MD, USA). Next, the mea of green immunofluorescence signal were done within the outlined cell regions of mock-plants, and the infected Irys and Sárpo Mira plants with the use of Image J. Levels of fl signal were calculated in the form of corrected total cell fluorescence (CTCF) on magnif with 1.00 zoom factor [54,55] using the following formula:

CTCF = Integrated Density − (Area of Selected Cell Region × Mean Fluoresce
Background Readings) Estimated CTCF values were then analysed statistically at selected time intervals for of reaction to PVY NTN by using the one-factor analysis of variance method (ANOVA). Th analyses enabled us to find the values of statistical significance when quantifying the levels of Xet5. Furthermore, the mean CTCF values were evaluated at the p < 0.05 level of significance hoc Tukey HSD testing in STATISTICA software (StataSoft and TIBCO Software Inc., Palo USA, version 13.0).

Quantitative Immunogold Localisation by Direct Estimation of the Relative Labelling Index (
, Sweden), while the secondary anti-mouse IgG conjugated with AlexaFluor ® 488 was from Jackson ImmunoResearch Europe Ltd. The controls consisted of mock-inoculated tissue and a pre-immune serum. An Olympus AX70 Provis (Olympus Poland, Warsaw, Poland) with a UM61002 filter set and an Olympus UC90 HD camera (Olympus Poland) were used for fluorescence imaging. Images were acquired using Olympus Cell Sense Standard Software (Olympus, Center Valley, PA, USA, version 1.18). After gaining florescent images, further quantitative measuring of the fluorescence signal was performed. As a first step, regions (cell wall or protoplast) with green fluorescence signal from xylan-1 or Xet5 epitopes were marked and outlined by using a special marker in the Image J program (Version 1.52e, National Institutes of Health, Bethesda, MD, USA). Next, the measurements of green immunofluorescence signal were done within the outlined cell regions of mock-inoculated plants, and the infected Irys and Sárpo Mira plants with the use of Image J. Levels of fluorescence signal were calculated in the form of corrected total cell fluorescence (CTCF) on magnification 20× with 1.00 zoom factor [54,55] using the following formula: CTCF = Integrated Density − (Area of Selected Cell Region × Mean Fluorescence of Background Readings) Estimated CTCF values were then analysed statistically at selected time intervals for both types of reaction to PVY NTN by using the one-factor analysis of variance method (ANOVA). The ANOVA analyses enabled us to find the values of statistical significance when quantifying the levels of Xylan-1 or Xet5. Furthermore, the mean CTCF values were evaluated at the p < 0.05 level of significance using post hoc Tukey HSD testing in STATISTICA software (StataSoft and TIBCO Software Inc., Palo Alto, CA, USA, version 13.0).

Quantitative Immunogold Localisation by Direct Estimation of the Relative Labelling Index (RLI)
Potato leaves of both cultivars were fixed and embedded and treated step by step to prepare for transmission electron microscopy (TEM) according to [11]. Then 50-70 nm thick leaf sections from PVY NTN -infected or mock-inoculated plants were mounted on Formvar-coated nickel grids and treated exactly as described by Otulak et al. [50]. Grids were rinsed with primary antibodies for XTH (Xet5) or xylan-1/xyloglucan in PBS and washed in PBS-Tween 20. After that the grids with leaf sections were treated for 1 h with gold-conjugated secondary antibody, with anti-rabbit 15 nm (Sigma-Aldrich, Warsaw, Poland) for XTH localisation or with 10 nm for xylan-1/xyloglucan detection (Sigma-Aldrich, Warsaw, Poland), rinsed for 5 min in PBS and then in distilled water. Labelling specificity was checked by incubating grids with material from mock-inoculated plants and by omitting the primary antibody in the incubation solution [11]. The grids were counterstained with 1% uranyl acetate for 5 min and washed 5 × 2 min with distilled water. The immunogold-labelled sections were examined by transmission electron microscope (as described above). The results of immunogold labelling of xylan-1 and Xet5 in mock-inoculated, susceptible or resistant potato plants were further analysed as follows. The quantitative assessment of preferential labelling of specific structures/organelles was carried out by using a reliable estimation method called relative labelling index (RLI). RLI was determined as described by Mayhew [56] and Otulak et al. [57]. The direct estimation method of RLI was selected by comparing the number of observed gold particles (G0) within selected compartments with the expected gold particles (Ge) of the appropriate reference structure or organelles in a leaf [56]. For estimation of G0 and Ge, gold particles were scored in 40 of 10 µm 2 fields per photo. When there is random labelling, RLI equalled 1, but where there is preferential labelling, RLI was higher than 1. Statistical significance of preferential labelling was assessed by partial X 2 analysis according to Mayhew [56]. The statistically significant RLI values were >1, with the corresponding partial X 2 values accounted for a significant proportion (at least 10%) of total X 2 .

Conclusions
In this work we have addressed a general question about the function of different cell wall components-particularly, how these elements interact with each other, but also how they change due to interactions with pathogenic viruses. By using precise in situ fluorescence and the ultrastructural localisation of non-cellulosic polysaccharides, we present for the first time the trends of accumulation of the major cell wall matrix hemicelluloses, i.e., xylan-1/xyloglucan together with xyloglucan xyloglucosyl-transferase (XTH-Xet5) in both the symplast and the apoplast during either compatible or incompatible interactions with PVY NTN . The function of xyloglucan in the plant cell wall seems to correlate closely with the pathogen-induced changes. Additionally, the loosening of cell wall accompanied the increase in xylan deposition during susceptible Irys-PVY interaction, whereas, during the hypersensitive response, with the cell wall strengthening, the xylan content extensively decreased. Moreover, the level of XTH-Xet5 gradually increases after the resistant potato-PVY NTN interaction; during susceptible interaction the deposition of XTH-Xet5 somewhat decreases. The presented data provide novel insight into the cell wall reorganisation induced by viruses, specifically by PVY NTN infection, illustrating the processes that take place during biotic stress. Our findings increase the understanding of the mechanisms of defence that are actively incorporated into cell wall signalling.