Cell Wall Modifications in Giant Cells Induced by the Plant Parasitic Nematode Meloidogyne incognita in Wild-Type (Col-0) and the fra2 Arabidopsis thaliana Katanin Mutant

Meloidogyne incognita is a root knot nematode (RKN) species which is among the most notoriously unmanageable crop pests with a wide host range. It inhabits plants and induces unique feeding site structures within host roots, known as giant cells (GCs). The cell walls of the GCs undergo the process of both thickening and loosening to allow expansion and finally support nutrient uptake by the nematode. In this study, a comparative in situ analysis of cell wall polysaccharides in the GCs of wild-type Col-0 and the microtubule-defective fra2 katanin mutant, both infected with M. incognita has been carried out. The fra2 mutant had an increased infection rate. Moreover, fra2 roots exhibited a differential pectin and hemicellulose distribution when compared to Col-0 probably mirroring the fra2 root developmental defects. Features of fra2 GC walls include the presence of high-esterified pectic homogalacturonan and pectic arabinan, possibly to compensate for the reduced levels of callose, which was omnipresent in GCs of Col-0. Katanin severing of microtubules seems important in plant defense against M. incognita, with the nematode, however, to be nonchalant about this “katanin deficiency” and eventually induce the necessary GC cell wall modifications to establish a feeding site.


Introduction
Root-knot nematodes (RKN; Meloidogyne spp.) are obligatory parasites that inhabit plant roots, nursing from specially modified host cells to complete their life cycle [1]. Of major interest is the Meloidogyne incognita species that infests crops of high economic interest, causing worldwide agricultural yield production reduction [2][3][4]. M. incognita stage 2 juveniles (J2) enter the host plant close to the root tip; they migrate towards the elongation zone anchor to the root central cylinder, become sedentary and after bypassing host defenses, establish a feeding site [5]. Nematode signals direct root cylinder parenchyma cells to differentiate into multinucleate and metabolically active giant cells (GCs) that resemble transfer cells [1], allowing M. incognita to withdraw nutrients from the plants' conducting tissues. During root invasion, M. incognita juveniles express numerous genes encoding cell wall-degrading enzymes and virulence effectors to enter, migrate into roots, establish feeding site and eventually induce GC formation [6,7]. At the end, the feeding site consists of several GCs enclosed in a macroscopically visible gall, a typical characteristic of RKN infestation. The second stage juveniles after entering the root, hatch to pass through several developmental stages (J3, J4) and finally reach to used to locate the major wall components in giant cells induced by RKN. The role of fra2 mutation in GCs cell wall modifications and in nematode infection is discussed.

Cell Wall Hemicellulose and Pectin Epitope Analysis of Uninfected Roots
Toluidine blue staining further confirmed the already observed [24] differences of fra2 and wild type root anatomical features. Generally, the primary A. thaliana root anatomy is similar to fra2 and Col-0 roots ( Figure 1) as it has been already reported [22]. Cell division zone of fra2 roots is shorter than that of the wild type. Furthermore, due to an increase in cell width, an overall increase in root diameter could be observed ( Figure 1B cf. Figure 1A). The mean diameter of Col-0 roots was 81.5 ± 0.9 µm, while mean root diameter of fra2 plants was 146.3 ± 0.5 µm (± standard error; n = 10). All epitopes analyzed were present in the uninfected roots. Cellulose staining (calcofluor-white staining) was more intense in fra2 ( Figure 1D cf. Figure 1C).
antibodies was used to locate the major wall components in giant cells induced by RKN. The role of fra2 mutation in GCs cell wall modifications and in nematode infection is discussed.

Cell Wall Hemicellulose and Pectin Epitope Analysis of Uninfected Roots
Toluidine blue staining further confirmed the already observed [24] differences of fra2 and wild type root anatomical features. Generally, the primary A. thaliana root anatomy is similar to fra2 and Col-0 roots ( Figure 1) as it has been already reported [22]. Cell division zone of fra2 roots is shorter than that of the wild type. Furthermore, due to an increase in cell width, an overall increase in root diameter could be observed ( Figure 1B; cf. 1A). The mean diameter of Col-0 roots was 81.5 ± 0.9 μm, while mean root diameter of fra2 plants was 146.3 ± 0.5 μm (± standard error; n = 10). All epitopes analyzed were present in the uninfected roots. Cellulose staining (calcofluor-white staining) was more intense in fra2 ( Figure 1D; cf. 1C). However, some epitopes displayed different distribution patterns between Col-0 and fra2 plants. In particular, callose staining (aniline blue) was more intense in fra2 ( Figure 2B), while in wild-type plants callose signal was strong in the xylem elements ( Figure 2A). The xyloglucans epitope recognized by LM25 had a more expanded cortex cell distribution in fra2, in contrast to the vascular cylinder cell distribution pattern of it in the Col-0 plants ( Figure 2D; cf. 2C). However, some epitopes displayed different distribution patterns between Col-0 and fra2 plants.
In particular, callose staining (aniline blue) was more intense in fra2 ( Figure 2B), while in wild-type plants callose signal was strong in the xylem elements ( Figure 2A). The xyloglucans epitope recognized by LM25 had a more expanded cortex cell distribution in fra2, in contrast to the vascular cylinder cell distribution pattern of it in the Col-0 plants ( Figure 2D cf. Figure 2C). Pectic polysaccharide and arabinogalactan-protein epitopes also had a different distribution in uninfected Col-0 and fra2 root sections. The demethylesterified HG (DeSPHG) epitope recognized by JIM5 did not seem to differ between Col-0 and fra2 ( Figure 3A; cf. 3B). Highly MHGs recognized by LM20 were present only in the vascular cylinder of the Col-0 while in the fra2 root sections their distribution expanded also to the cortex cells ( Figure 3C; cf. 3D). Arabinan epitopes (LM6) had a more expanded distribution in wild-type roots, found in all of the root tissues, while in fra2 their distribution was mostly restricted to the vascular cylinder ( Figure 3E; cf. 3F). Pectic polysaccharide and arabinogalactan-protein epitopes also had a different distribution in uninfected Col-0 and fra2 root sections. The demethylesterified HG (DeSPHG) epitope recognized by JIM5 did not seem to differ between Col-0 and fra2 ( Figure 3A cf. Figure 3B). Highly MHGs recognized by LM20 were present only in the vascular cylinder of the Col-0 while in the fra2 root sections their distribution expanded also to the cortex cells ( Figure 3C cf. Figure 3D). Arabinan epitopes (LM6) had a more expanded distribution in wild-type roots, found in all of the root tissues, while in fra2 their distribution was mostly restricted to the vascular cylinder ( Figure 3E cf. Figure 3F).     Figure 4C), a trend visible in all gall cross sections. The GCs located in the vascular cylinder bared numerous nuclei, were highly vacuolated and appeared to be of similar morphology in both Col-0 and fra2 plants ( Figure 4E cf. Figure 4F). GCs located in the vascular cylinder bared numerous nuclei, were highly vacuolated and appeared to be of similar morphology in both Col-0 and fra2 plants ( Figure 4E cf. 4F).     Callose was omnipresent in the GC cell walls of the Col-0 plants ( Figure 5A,C) contrary to the GC cell walls of fra2, which lacked callose signal ( Figure 5B,D) as revealed by both aniline blue staining and anti-β-1,3-glucan localization. On the other hand, no significant differences in the distribution pattern of the LM25 bound epitope were observed ( Figure 5E,F).
MPHGs signal was more intense in the GC cell walls of fra2 compared to the wild type ( Figure  6D), while DeSPHG signal did not seem to differ among the roots of Col-0 and fra2 ( Figure 6B; cf. 6A). LM6 recognized epitopes were omnipresent in fra2 nemadote-infected roots ( Figure 6F) while a weak signal was observed in the corresponding infected Col-0 roots ( Figure 6E). Table 1 summarizes all epitopes present in GCs. Callose was omnipresent in the GC cell walls of the Col-0 plants ( Figure 5A,C) contrary to the GC cell walls of fra2, which lacked callose signal ( Figure 5B,D) as revealed by both aniline blue staining and anti-β-1,3-glucan localization. On the other hand, no significant differences in the distribution pattern of the LM25 bound epitope were observed ( Figure 5E,F).
MPHGs signal was more intense in the GC cell walls of fra2 compared to the wild type ( Figure 6D), while DeSPHG signal did not seem to differ among the roots of Col-0 and fra2 ( Figure 6B cf. Figure 6A). LM6 recognized epitopes were omnipresent in fra2 nemadote-infected roots ( Figure 6F) while a weak signal was observed in the corresponding infected Col-0 roots ( Figure 6E). Table 1 summarizes all epitopes present in GCs.

The Impact of Katanin Mutation on the Infection of A.thaliana by M. incognita
The total nematode burden between Col-0 and fra2 expressed as female number/dry root weight (mg) significantly differed (p < 0.05). Root of fra2 seemed to bear more nematodes per mg of their dry root weight compared to Col-0 ( Figure 7) indicating that root invasion success was higher in fra2 mutants.

The Impact of Katanin Mutation on the Infection of A.thaliana by M. incognita
The total nematode burden between Col-0 and fra2 expressed as female number/dry root weight (mg) significantly differed (p < 0.05). Root of fra2 seemed to bear more nematodes per mg of their dry root weight compared to Col-0 ( Figure 7) indicating that root invasion success was higher in fra2 mutants.

Discussion
The fragile fiber 2 (fra2) mutant, as its name states, displays increased fragility in all the plant organs and particularly in the stems [22]. The reduced cellulose content and the altered lignin deposition, leads to shorter and thinner fibers and to lower mechanical resistance [22]. In the present study, we additionally observed that hemicellulose and pectin cell wall components have a different deposition pattern in the fra2 mutant root compared to Col-0 (Figures 2 and 3). Xyloglucans ( Figure  2) but also MPHGs, DeSPHG and arabinans ( Figure 3) had a differential distribution between Col-0 and fra2.
Generally it was reported that the fra2 root cell division zone is shorter than that of Col-0 [24]. This difference was not the outcome of a decrease in cell divisions rate, but was due to the distorted shape of the root cells of the mutant. The root cells at the division zone of the mutant are unable to elongate, contrary to those of the wild type, a fact that can be seen by checking the height/width ratio of the protodermal root cells. So, the failure of MT alignment in fra2 results in a dwarf root phenotype and due to the cell width increase, an increase in root diameter was observed [24]. Although cell shape is definitely controlled by MTs arrangement, there is evidence on the

Discussion
The fragile fiber 2 (fra2) mutant, as its name states, displays increased fragility in all the plant organs and particularly in the stems [22]. The reduced cellulose content and the altered lignin deposition, leads to shorter and thinner fibers and to lower mechanical resistance [22]. In the present study, we additionally observed that hemicellulose and pectin cell wall components have a different deposition pattern in the fra2 mutant root compared to Col-0 (Figures 2 and 3). Xyloglucans (Figure 2) but also MPHGs, DeSPHG and arabinans ( Figure 3) had a differential distribution between Col-0 and fra2.
Generally it was reported that the fra2 root cell division zone is shorter than that of Col-0 [24]. This difference was not the outcome of a decrease in cell divisions rate, but was due to the distorted shape of the root cells of the mutant. The root cells at the division zone of the mutant are unable to elongate, contrary to those of the wild type, a fact that can be seen by checking the height/width ratio of the protodermal root cells. So, the failure of MT alignment in fra2 results in a dwarf root phenotype and due to the cell width increase, an increase in root diameter was observed [24]. Although cell shape is definitely controlled by MTs arrangement, there is evidence on the participation of the cell wall matrix materials in the morphogenesis of various cell types [25,26]. There is strong evidence of interplay between MTs and various cell wall components [25]. MTs that cannot be correctly aligned may have an impact on the differential cell wall component organization, but also the establishment of a specific cell wall micro domain can possibly affect MT arrangement, indicating the tight connection between MTs and cell wall components [25]. So, the inability of the cells to acquire their shape could be the result not only of the altered MT order, but because also of the changes observed in the structure of the cell wall architecture.
During root development, root cell walls must endure both the internal forces created by the growing protoplast and the external forces due to root soil penetration; consequently, root cell walls must be strong enough to withstand the applied forces but at the same time they must be flexible enough to support root growth. Therefore, cell wall synthesis, softening and maturation processes are co-ordinated [27] in order to support cell maturation and expansion. Especially cell expansion requires normal cellulose synthesis [28], but also pectin modifications [29,30]. Since we focused our study onn the root differentiation zone (visible xylem elements) [31], it is expected that, by the action of PMEs, the methyl-esterified homogalacturonans are demethylesterified. As already reported, DeSPHGs are interconnected by calcium bridges, forming cell wall matrices of high viscosity to support the cell wall stiffening in the differentiation zone [32]. Indeed, in Col-0 roots LM20 (MPHGs) signal was restricted to the vascular cylinder, "missing" from the other root tissues, in which cells could have acquired their final size. On the contrary, the extensive presence of the LM20 epitope in the root sections of fra2 ( Figure 3D), could possibly be a side-effect of the hampered development of fra2 root [24,33]. Generally, MPHGs characterize cell walls which elongate (i.e cell walls of hypocotyl cells) [34]. In fra2 root cells, the extensive presence of the LM20 epitope could indicate that the cells have not terminated their elongation process, since the desired cell shape has not yet been acquired.
The notion that cell elongation of fra2 root cells is hampered is further on confirmed by the differential LM6 epitope distribution observed in fra2 root sections compared to Col-0 ( Figure 3E). Usually, arabinans are required for maintenance of cell wall flexibility [29] and their presence intensifies in the differentiation root zone [29]. This local cell wall remodeling reflects a change in the cell wall mechanical properties [29]. In fra2 mutant this shift could be delayed indicating a hindered cell differentiation. Xyloglucans signal is distributed more broadly in fra2 root sections (Figure 2), further defining the problematic cell expansion of fra2 root cells. Xyloglucans binding to cellulose microfibrils seems to reinforce the cell wall, turning it more rigid and less expandable [35]. In fra2 roots, the increased presence of xyloglucans, that interact with cellulose can result in increased cell wall rigidness, opposing flexibility ( Figure 2D). Except for the disturbed root development, the altered hemicellulose and pectin deposition patterns observed in fra2 could be the outcome of the plants compensatory efforts to overcome the potential stiffness reduction because of cellulose loss. Generally no important change in the elasticity modulus of the cell walls in katanin mutants was recorded, suggesting that other cell wall components compensate for the hampered cellulose synthesis observed in katanin mutants [36,37]. In this context, xyloglucans were essential when katanin expression was hindered [37].
The unique cells induced by the nematode infection in the feeding structure must be capable to withstand the turgor pressure within, whilst maintaining the necessary flexibility to cope with the periodic demands of nematode feeding [5]. To facilitate this, the walls of the cells produced in the feeding site are specifically modified. In previous studies, abundant MPHGs, xyloglucans and arabinans were found in syncytial cell walls of a cyst nematode [16,17]. In particular, it is reported that the cell walls in syncytia cells formed in A. thaliana are rich in MPHGs and in arabinans, while galactans were not detected. The cell wall modifications continue to take place even after the syncytium reaches its final size at 14 dpi. The abundance of MPHGs in the vascular cylinder of infected roots implies that the syncytial cell wall is flexible. Since the syncytia are highly metabolically active and represent a nutrient sink for solutes, their flexibility is essential for their functionality [38]. Apart from that, the cell walls of the GCs of M. incognita infected roots in three different hosts [11] display the same characteristics. The GCs are metabolically active and act as transfer cells for the nematode nourishment. A unique feature of the giant cells is their ability to grow isotropically. This kind of cell expansion requires extensive and coordinated cell wall remodeling. The loosening of the cellulose/xyloglucan network is mediated through the action of specific enzymes, like glucanases, expansins and endotransglycosylases. These subtle modifications provide the GCs cell walls with the necessary flexibility during the changes in cell volume and shape. In the present study, we found that callose was also present in the GCs cell walls, (and seemed to increase its abundance upon nematode infection Figure 2 cf. Figure 5) as confirmed by both aniline staining and anti-β-1,3-glucan immunolabelling ( Figure 5). It is commonly accepted that callose controls the permeability of plasmodesmata and/or provides mechanical support to the cell wall against different environmental stresses. This polysaccharide is synthesized upon injury, low temperature, heavy metals and various pathogens [39][40][41][42]. Another biotic stress upon which callose is synthesized, seems to be nematode infection adding the necessary mechanical support to the GCs cell walls structure, along with the other cell wall changes ( [11]; this study). However, one cannot exclude the fact that callose could also act as a defense barrier as reported for other nematode species [10].
In order to elucidate if the observed differences in GC cell wall composition had any functional consequences on the plant-nematode interaction, we used the fra2 mutant, a mutant with established cellulose production defects [22]. Cellulose microfibrils are the scaffold of the cell wall and together with hemicelluloses they provide the cell wall with the necessary rigidity [43]. The fra2 mutant cell walls are extremely fragile; this fragility, due to reduced cellulose production, could lead to an increased susceptibility towards M. incognita infection, since fragile cell walls could lead to easiest nematode infection and penetration. In fact, susceptibility experiments proved that fra2 is more vulnerable to M. incognita infection than Col-0 ( Figure 7). M. incognita females were able to induce a feeding site in the fra2 roots more successfully than they did in Col-0. Conclusively, it seems possible that katanin MT severing is necessary for plant defense against nematode infection, even in an indirect way, adding to the existing microtubule associated proteins (MAPs; e.g., MAP65-3; [44]) that have a pivotal role in RKN invasion. Moreover, it was reported that MT reorganization is essential for feeding site establishment and juvenile maturation [45]. In accordance to this, katanin, a MT arrangement mediator, [33] seems to also have an important contribution towards M. incognita infection.
The established/observed ( [11]; this study) cell wall modifications in GC cell walls differed in fra2 plants ( Figures 5 and 6). Table 1 summarizes the noticed cell wall modifications that took place upon nematode infection in both Col-0 and fra2 roots. A noticeable difference refers to the levels of callose in fra2 GC cell walls. As confirmed by aniline staining and anti-β-1,3-glucan immunolabelling callose was absent from fra2 GC cell walls ( Figure 5). Generally, callose is responsible for cell wall flexibility supporting tenseness and contraction during mechanical stress. In other words, this polysaccharide decreases firmness and increases elasticity of cell walls [46]. Accordingly, fra2 GC cell walls lack elasticity, due to callose absence, becoming more vulnerable to turgor pressure. However, this loss seems to be compensated by the increased levels of arabinan (LM6; Figure 6) and MPHGs, which both generally provide cell wall with the necessary flexibility ( [11,17]; this study). Therefore, the wall flexibility of the GCs is maintained, even in a mutant with reduced cell wall elasticity (fra2). Cell wall alterations are mediated by the nematode-induced regulation of host genes [6]. It seems therefore plausible that the nematode recognizes the peculiarities of different hosts and is capable to manipulate host gene expression to successfully establish a feeding site. This notion is further supported by the already reported differential response of different hosts to M. incognita infection [11].

Plant Growth Conditions, Nematode Rearing and Collection
Seeds of Arabidopsis thaliana L. (Heynh) (as previously reported; [47]) wild type (ecotype Columbia; Col-0) and fragile fiber 2 (fra2) mutant (purchased from the NASC European Arabidopsis Stock Centre, Nottingham, UK) were surface sterilized with a 30% (v/v) bleach solution and kept at 4 • C for 48 h. Subsequently, the seeds were transferred in soil and left to germinate and grow in a growth chamber under a constant 16 h day/8 h night regime at an ambient temperature of 21 ± 1 • C, with light intensity set at 120 µmol m −2 ·s −1 .
Seeds of tomato plants Solanum lycopersicum L. cv Belladonna were directly planted in pots filled with soil and grown in the same conditions as stated above for two months prior to nematode infection. Populations of M. incognita of Greek origin were reared on the above tomato plants [48]. Freshly hatched (24-48 h) J2 were extracted from egg masses according to a modified Baermann technique [49] from 40 day-old (d) infested roots, to be used for the bioassays.

Light and Fluorescence Microscopy
All chemicals and reagents used were purchased from Sigma (St. Louis, MO, USA), Merck (St. Louis, MO, USA), and Applichem (Darmstadt, Germany), unless otherwise stated. Twenty one days post infection (21 dpi), root tissue, of infected and non-infected (control) Col-0 and fra2 plants, was carefully cleared from soil. Given the incubation conditions, at 21 dpi giant cells are fully formed [50], a fact also verified in A. thaliana plants infected with M. incognita [6]. Both root knots and uninfected root pieces were prepared for light microscopy according to [51]. Each experiment was conducted in three replicates and about 10-15 roots from each experiment were selected and further on processed. In short, root tissue pieces were fixed in 2% w/v paraformaldehyde (PFA) and 0.5% v/v glutaraldehyde (GA) in PEM buffer (50 mM PIPES, 5 mM EGTA, 5 mM MgSO 4 , pH 6.8) for 2 h at 4 • C. Fixed roots were rinsed twice in the same buffer for 10 min. Root tissue samples were then dehydrated in a graded ethanol series, subjected to post-fixation in 0.25% OsO 4 in 30% ethanol and embedded in LR White resin. Finally, the samples were sectioned using an ULTROTOME III TYPE 8801A ultramicrotome (LKB, Stockhom, Sweden) equipped with a glass knife. Semi-thin sections (0.5-2 µm) were stained with 0.5% (w/v) toluidine blue O and observed by light microscopy. Especially for the uninfected roots, special care was taken to obtain sections at the differentiating zone [52], in both Col-0 and fra2. Since this particular root zone for fra2 is observed closer to the meristem than that of Col-0 [24] we used the visible xylem elements as a marker of root differentiation. Considering root knots special care was taken to obtain transverse sections through the middle of the gall in order to obtain vision of the biggest giant cells [53].
Detection of callose by aniline blue fluorochrome was performed by semithin sections remaining in aniline blue solution (0.05% (w/v) aniline blue) in 0.07 M K 2 HPO 4 buffer, pH 8.5) [54,55] and observed under an epifluoresence microscope. For callose detection via immunolabeling, semithin sections (approximately 2-3 µm thick) of material embedded in LRW resin were transferred to glass slides and blocked with 5% (w/v) BSA in PBS for 5 h. After washing with PBS, anti-callose antibody [56] diluted 1:40 in PBS containing 2% (w/v) BSA was applied overnight. Following rinsing with PBS and blocking again with 2% (w/v) BSA in PBS, the sections were incubated for 1 h, at 37 • C in FITC anti-mouse IgG diluted 1:40 in PBS containing 2% (w/v) BSA. After rinsing with PBS, the sections were mounted using an anti-fade medium or with calcofluor white [57].
All of the fluorescence specimens were examined with an Axioplan microscope (Zeiss, Berlin, Germany) equipped with a UV source, a Differential Interference Contrast (DIC) optical system. All the photos were taken with a Zeiss Axiocam MRc5 digital camera using the ZEN 2.0 software (Zeiss, Berlin, Germany) according to the manufacturers' instructions [58]. Two filter sets were used for specimen observation: a filter set provided with exciter solid glass filter 365 nm and barrier long-wave pass band filter 420 nm, and another set provided with exciter pass band filter 450-490 nm and barrier pass band filter 515-565 nm. All samples were checked for UV autofluorescence using the above filters. Digital micrographs were processed with Adobe Photoshop (Adobe Inc., San Jose, CA, USA) with only linear settings.

Cell Wall Epitope Immunolabeling
Another set of semi-thin sections were incubated with rat monoclonal antibodies directed to cell wall matrix polysaccharides according to [11] ( Table 2). The distribution of highly esterified HGs recognized by LM20 antibody [59], the partially de-esterified HGs recognized by JIM5 antibody [60], the motifs of xyloglucans recognized by LM25 [61] and arabinan motifs with the LM6 antibody [59]. Pectate lyase treatment before immunolabeling with LM25 was performed as described by [11]. Semi-thin sections (approximately 2-3 µm thick) of material embedded in LRW resin were transferred to glass slides and blocked with 2% (w/v) BSA in PBS for 2 h. After washing with PBS, a specific antibody against a cell wall matrix material was added and applied overnight. All antibodies were diluted 1:40 in PBS containing 2% (w/v) BSA. Following rinsing with PBS and blocking again with 2% (w/v) BSA in PBS, the sections were incubated for 2 h, at 37 • C with a FITC anti-rat IgG (Sigma) diluted 1:40 in PBS containing 2% (w/v) BSA. After three washes with PBS, the sections were mounted using an anti-fade medium containing p-phenylenediamine. Table 2. List of the monoclonal antibodies used in the current study, of epitopes they recognize and of references.

Arabidopsis Wild-Type and fra2 Mutant Nematode Susceptibility
Artificial infestation with M. incognita was performed on A. thaliana wild-type and fra2 mutant plants using 1000 J2/plant. The plants were maintained for the completion of one biological cycle, which is 40 dpi, at the same conditions as stated above. After that, roots were stained with acid fuchsin [62] and the following variables were assessed: root weight (dry) and total number of female nematodes per gram of root at 10× magnification under uniform illumination by transparent light within tissue sample. The experiment was performed twice, and the treatments were always arranged in a completely randomized design with five replicates. Because ANOVA indicated no significant treatment by time interaction, means were averaged over experiments. Nematode infestation levels were compared using Tukey's test at p ≤ 0.05. Statistical analysis was performed using SPSS 20 (IBM, Armonk, NY, USA).

Conclusions
The MT katanin severing defects of fra2 mutants lead to altered pectin and hemicellulose distribution, compared to Col-0. Xyloglucans and MPHGs were more abundant in fra2 a fact that could lead to an increased M. incognita infection rate. Following M. incognita infection in fra2 roots, the GC cell wall modifications (highly-esterified pectic homogalacturonan and pectic arabinan) were present, maybe to compensate for the low amount of callose, which was omnipresent in GC cell walls of Col-0.
Therefore with this study we conclude for the first time that katanin MT severing seems important in plant defense against M. incognita. The nematode however, seems to be able to overcome this "katanin deficiency" and eventually induce the necessary GC cell wall modifications and to establish a feeding site.