Evaluation of the significance of cell wall polymers in flax infected with a pathogenic strain of Fusarium oxysporum

Background Fusarium oxysporum infection leads to Fusarium-derived wilt, which is responsible for the greatest losses in flax (Linum usitatissimum) crop yield. Plants infected by Fusarium oxysporum show severe symptoms of dehydration due to the growth of the fungus in vascular tissues. As the disease develops, vascular browning and leaf yellowing can be observed. In the case of more virulent strains, plants die. The pathogen’s attack starts with secretion of enzymes degrading the host cell wall. The main aim of the study was to evaluate the role of the cell wall polymers in the flax plant response to the infection in order to better understand the process of resistance and develop new ways to protect plants against infection. For this purpose, the expression of genes involved in cell wall polymer metabolism and corresponding polymer levels were investigated in flax seedlings after incubation with Fusarium oxysporum. Results This analysis was facilitated by selecting two groups of genes responding differently to the infection. The first group comprised genes strongly affected by the infection and activated later (phenylalanine ammonia lyase and glucosyltransferase). The second group comprised genes which are slightly affected (up to five times) and their expression vary as the infection progresses. Fusarium oxysporum infection did not affect the contents of cell wall polymers, but changed their structure. Conclusion The results suggest that the role of the cell wall polymers in the plant response to Fusarium oxysporum infection is manifested through changes in expression of their genes and rearrangement of the cell wall polymers. Our studies provided new information about the role of cellulose and hemicelluloses in the infection process, the change of their structure and the expression of genes participating in their metabolism during the pathogen infection. We also confirmed the role of pectin and lignin in this process, indicating the major changes at the mRNA level of lignin metabolism genes and the loosening of the pectin structure. Electronic supplementary material The online version of this article (doi:10.1186/s12870-016-0762-z) contains supplementary material, which is available to authorized users.


Background
Flax (Linum usitatissimum) is a unique plant which is a valuable source of fibre and oil. Flax raw materials are applicable in many industrial branches: medicine, pharmacy and cosmetics. It is estimated that around 20 % of flax cultivation loss is a result of fusariosis.
These diseases caused by Fusarium species fungi contribute to the lowering of yield, grain and fibre quality. The highest pathogenicity towards flax was exhibited by F. oxysporum f. sp. linii, which causes flax wilt [1,2].
The plant cell wall is the first physical barrier to pathogen infection. During the first stages of infection, pathogens secrete enzymes that degrade the cell wall: pectinases, cellulases and hemicellulases. Their first aim is degradation of pectin, which results in loosening of cell wall structure and thereby enables digestion of the following polymers: cellulose and hemimcellulose [3,4]. During the colonization in plants, antifungal compounds are generated: phytoalexins, PR proteins, small antifungal peptides and reactive oxygen species [5][6][7]. PR proteins are locally accumulated in the infection sites, in the adjacent tissues and also in non-infected tissues, which provides plant resistance to subsequent infections. Being part of the plant systemic response, these genes are not activated immediately after the pathogen attack, though their analysis was sufficient to establish the initial stage of the infection [8,9].
The plant cell wall is a dynamic structure, composed of polysaccharide polymers (cellulose, hemicelluloses and pectin) and non-polysaccharide polymers (lignin) and proteins (structural and enzymatic) [10]. Cell wall composition is strictly regulated in the different types of cells during their growth, development and plant response to abiotic and biotic stress factors [11].
Cellulose consists of long, non-branched microfibrils composed of β-1,4-glucose chains, which are transversely connected with hydrogen bonds and van der Walls forces. There are two types of cellulose structure: highly polymerized and ordered and less polymerized, loose and amorphous [12]. The parameter that describes cellulose structure is the crystallinity index (CI), determining the content of crystalline form in the cellulose [13]. During cellulose biosynthesis, the key role is played by a large protein complex that is anchored in the cell membrane and consists of six subunits, each subunit consisting of six proteins (cellulose synthases; CESA) [14].
During pathogen infection pectin de-esterification plays a key role in the plant defence responses. Pectin de-esterification leads to the generation of free carboxyl groups, altering pH in the cell wall and enabling aggregation of polyuronates to the gel structure, which results in changes in the porosity of the cell wall [26]. Additionally, this process enables HG degradation by pectin polygalacturonases, pectin lyase and pectate lyase [27]. The level of methyl esterification of pectin determines the sensitivity of plants to the pathogen infection. The high content of methylated residues of galacturonic acid in HG corresponds to the increase of plant resistance [28,29]. Moreover, the level and pattern of methyl esterification of pectin influence the activity of polygalacturonases, which are responsible for the generation of short fragments of homogalacturonan chains, which are oligogalacturonides (OG), endogenous molecules of elicitor activity that play a crucial role in the pathogen defence response by enhancing the plant natural response [30][31][32].
By strengthening the cell wall, lignin provides a better barrier to pathogen attacks. The increased lignin synthesis resulting from biotic stress factors results from phenylpropanoid pathway stimulation and from lignin polymerization [38,39].
Development of genetic engineering enabled generation of genetically modified plants characterized by increased resistance to pathogen infections. Flax that was more resistant to F. oxysporum and F. culmorum infection was generated by overexpression of genes involved in pathogenesis (PR genes) [40] and genes of secondary metabolites [41][42][43].
It is justified to perform research on cell wall components in order to discover their significance for plant resistance to pathogens.
The aim of this study was to estimate the role of flax cell wall polymers in response to Fusarium oxysporum. The significance of cell wall polymers (cellulose, hemicelluloses, pectin and lignin) was elucidated by analysis of the expression level of genes implicated in the metabolism of these compounds and by the analysis of the corresponding metabolites in flax in response to pathogenic fungi.

Phenotypic analysis of flax seedlings incubated with Fusarium oxysporum
In order to determine the role of cell wall polymers in flax in the response to a pathogenic strain of Fusarium oxysporum infected flax seedlings were incubated with the fungus for 6, 12, 24, 36 and 48 h. In the subsequent incubation period the transferred plants were photographed (Additional file 1: Figure S1). The first phenotypic changes of the flax seedlings were observed within 24 h after the transfer. Cotyledons of the seedlings remained green, while the adventitious root tips became necrotic and the necrosis progressed with the incubation time. Initially, after 24 h only a few root cells became necrotic, while after 48 h the necrotic changes were observed in most of the roots. Despite this, the cotyledons remained green and firm F. oxysporum mycelium was not observed on the surface of the MS medium. The last incubation period analyzed was 48 h after transfer, as at this stage the plants retained their green color and turgor, thus enabling activation of its defence mechanisms. In the consecutive hours of incubation the progress of the infection contributed to weakening and wilting of flax seedlings (data not shown); therefore their detailed analysis was abandoned.

Expression of PR genes increased in flax seedlings infected with Fusarium oxysporum
In order to determine the earlier stages of infection we investigated the changes occurring during 6 and 12 h of incubation with the pathogen. We determined the levels of mRNAs of PR genes, because it is known that the genes are strongly expressed in plants in response to pathogen infections. Changes in PR gene expression in flax infected with a pathogenic Fusarium oxysporum fungus strain are presented in Fig. 1. The analyzed genes were characterized by an unchanged expression level in 6 h of incubation (β-1,3-glucanase 2 and chitinase) or with lower expression (by 40 %) followed by an increase in the subsequent hours of incubation in the case of β-1,3-glucanase 1. The level of expression of β-1,3-glucanase 1 increased during the period of incubation (from 2.6-fold in 12 h to 11-fold in 48 h). A similar expression pattern was found for the chitinase gene. The level of its transcript increased from 2.6-fold in 12 h to 4.9-fold in 36 h and dropped to 2.5-fold in the control. The analysis of β-1,3-glucanase 2 revealed the smallest changes in the expression in comparison to the other PR genes tested. However, compared to the control the level of mRNA of this gene increased 1.7fold in 12 h, 2.3-fold in 24 h, 1.6-fold in 36 h and 2-fold in 48 h of incubation with F. oxysporum.

Expression of cellulose metabolism genes changed in flax seedlings infected with Fusarium oxysporum
In the next step we analyzed changes in the levels of mRNAs of genes involved in cell wall polymer metabolism and determined their quantities in flax seedlings infected with a pathogenic strain of F. oxysporum. Polysaccharide (cellulose, hemicellulose, pectin) and non-polysaccharide (lignin) polymers were investigated.
Analysis of expression of genes of the synthesis and degradation of cellulose (5 isoforms of cellulose synthase and 2 isoforms of cellulase) in flax incubated for 48 h with a pathogenic F. oxysporum strain is presented in Fig. 2. Cellulose synthesis genes were characterized by a twofold expression pattern. In the first group (CSL1, CSL2 and CSL4) the expression levels were lowered (from 20 to 80 % depending on the gene analyzed and incubation time), while in the second group (CSL3 and CSL5) the expression was first reduced (to 60 % in 6 h and 26 % in 36 h for CSL3 and to 77 % in 12 h and 64 % in 36 h for CSL5) to increase 1.8-fold for both genes (CSL3 and CSL5). Among the levels of mRNAs of cellulose degradation genes, the expression of the cellulase 1 gene decreased (from 75 % in 6 h to 61 % in 12 and 36 h) and then increased 1.2-fold in 48 h of incubation with F. oxysporum. The expression of cellulase 2 initially increased 1.65-fold (in 6 h) and then decreased (to 46 % in 24 h and 60 % in 36 h).

Cellulose content increases in flax seedlings infected with Fusarium oxysporum
Analysis of cellulose content revealed a 20 % increase in flax seedlings incubated for 48 h with Fusarium oxysporum compared to non-infected seedlings (Fig. 3a). The amounts of cellulose did not change in the remaining hours of incubation of flax with the pathogen.

Expression of hemicellulose metabolism genes changed in flax seedlings infected with Fusarium oxysporum
Results depicting changes in the levels of expression of genes involved in hemicellulose synthesis (glucomannan 4β-mannosyltransferase -GMT, galactomannan galactosyltransferase -GGT, xyloglucan xylosyltransferase -XXT) and degradation (endo  Figure S2A) and total simple sugar (Additional file 2: Figure S2B) contents remained unchanged in the seedlings after F. oxysporum infection and the content was 1 mg/g FW and 17 mg/g FW, respectively. Changes in uronic acid (Fig. 3b) and simple sugar (Fig. 3c) contents were observed in particular hemicellulose fractions. Uronic acid content decreased in the K1SF fraction and increased in K4SF after the infection. A similar association was noted for simple Changes in expression levels of PR genes (β-glucanase 1, β-glucanase 2 and chitinase) in flax seedlings treated with pathogenic strains of F. oxysporum (F.ox.) at 6, 12, 24, 36 and 48 h after inoculation were presented as relative quantity (RQ) relative to reference gene (actin) in relation to control (C). The data were obtained from real-time RT-PCR analysis. The data represent the mean ± standard deviations from three independent experiments. The significance of the differences between the means was determined using Student's t test (*P < 0.05, **P < 0.01) sugars. In addition, analysis of the simple sugars in both fractions indicated an increased contribution of the K1SF fraction to total hemicellulose.  seedlings treated with pathogenic strains of F. oxysporum (F.ox.) at 6, 12, 24, 36 and 48 h after inoculation were presented as relative quantity (RQ) relative to reference gene (actin) in relation to control (C). The data were obtained from real-time RT-PCR analysis. The data represent the mean ± standard deviations from three independent experiments. The significance of the differences between the means was determined using Student's t test (*P < 0.05, **P < 0.01) changed over the time of incubation with F. oxysporum and in 24 h a 1.8-fold increase for PME1, 2.4-fold increase for PME3 and 2-fold increase for PME5 were observed. Moreover, the expression of PME5 increased 1.5-fold in 36 and 48 h, the expression of PME1 increased 1.4-fold in 6 h, but fell to 60 % in 12 h and the expression of PME3 decreased to 68 % in 6 h and increased 1.5-fold in 12 and 36 h. Expression of PG, PaL and PL genes initially decreased (67-43 % for PG and 82-32 % for PaL from 6 to 36 h and to 75 % for PL in 24 h) and then increased 1.4-fold for PG, 1.23-fold for PaL and 1.6-fold for PL in 48 h of incubation.

Pectin composition changed in flax seedlings infected with Fusarium oxysporum
Pectin content was evaluated based on the analysis of uronic acid and simple sugar contents. Because uronic acids are the main structural components, the pectin assay is often based on the analysis of these constituents. In order to perform detailed evaluation of pectin content, all simple sugars in consecutive pectin fractions of cell wall (WSFwater soluble fraction, CSF -CDTA soluble fraction, NSF -Na 2 CO 3 soluble fraction) must be assayed. Uronic acid content should be analyzed and need not be omitted because of partial qualitative analysis of pectin.
Total uronic acid content (about 5 mg/g FW) and simple sugars (about 12 mg/g FW) did not change in flax incubated for 48 h (Additional file 2: Figure S2C and D). However, uronic acid contents in particular pectin fractions of cell wall differ, indicating their higher content in the CSF fraction in the control seedlings (37.5 % of total pectin) and in the WSF fraction in the seedlings infected with F. oxysporum (43.6 % of total pectin) (Fig. 3e). The content of uronic acids in the NSF fraction did not change after the infection (26 % of total pectin). Differences were observed in simple sugars in pectin fractions between the infected and control flax seedlings (Fig. 3e). After infection with F. oxysporum the content of simple sugars in the WSF fraction increased by 10 % compared to the control, but did not change in the CSF fraction and decreased in the NSF fraction.
Among all the analyzed genes, four (HCT, CCoAOMT, COMT and SAD) showed the same expression pattern during the incubation with F. oxysporum compared to the control. Their expression increased from 6 h of the incubation to reach the maximum in 24 h (3.9-fold increase in HCT expression and 2.7-fold increase in expression of CCoAOMT, COMT and SAD) and decreased in 48 h to a level equal to the control (CCoAOMT, COMT), a level below the control (40 % lower expression of SAD) and a level above the control (HCT expression level of 1.8-fold of the control). Similarly, although shifted in time, the pattern of expression was characteristic for the GT gene, whose expression reached a maximum in 36 h (6.8-fold). PAL gene expression was initially lowered to 40 % in 6 h, and increased gradually to reach a maximum in 48 h (7.4-fold increase compared to the control). 4CL gene expression was decreased to 40 % in 6 and 12 h and to 80 % in 48 h. h after inoculation were presented as relative quantity (RQ) relative to reference gene (actin) in relation to control (C). The data were obtained from real-time RT-PCR analysis. The data represent the mean ± standard deviations from three independent experiments. The significance of the differences between the means was determined using Student's t test (*P < 0.05, **P < 0.01) Analysis of the last of the genes of lignin metabolism (CAD) showed a 2.6-and 3.1-fold increase in mRNA level in 6 and 48 h, respectively, and a 30 % decrease in 12 h of incubation with F. oxysporum.

Lignin content decreases in flax seedlings infected with Fusarium oxysporum
Lignin content was assayed in flax seedlings incubated with F. oxysporum for 48 h, and the results are presented seedlings treated with pathogenic strains of F. oxysporum (F.ox.) at 6, 12, 24, 36 and 48 h after inoculation were presented as relative quantity (RQ) relative to reference gene (actin) in relation to control (C). The data were obtained from real-time RT-PCR analysis. The data represent the mean ± standard deviations from three independent experiments. The significance of the differences between the means was determined using Student's t test (*P < 0.05, **P < 0.01) in Fig. 3f. The infection caused a decrease of lignin content in flax seedlings by about 20 % in comparison to the non-infected seedlings.

Infrared spectroscopy of cell wall of flax infected with Fusarium oxysporum confirms results of biochemical analysis of cell wall components
Analysis of infra-red spectroscopy of the cell wall of flax infected with F. oxysporum was performed to determine the structure of the cell wall and verify the results of cell wall polymer content assay obtained with spectrophotometric methods.
The infection of flax seedlings with F. oxysporum influenced the composition and structure of the cell wall. Infrared spectroscopy spectra of the infected and noninfected flax seedlings are presented in Fig. 7. Based on the spectra changes in the cellulose, pectin and lignin contents and changes in cellulose structure were determined in the studied samples. The cellulose content was 40 % higher after the infection with F. oxysporum compared to the non-infected seedlings (Fig. 7a). Cellulose structure was determined based on the analysis of appropriate bands. Integral intensities of bands at 1058 and 988 cm −1 corresponding to asymmetric vibrations of ν(C-O-C) indicate changes in the length of cellulose Fig. 6 Relative expression of selected genes of phenylpropanoid pathway transcripts in flax seedlings infected with Fusarium oxysporum. Changes in expression levels of phenylpropanoid metabolism genes: phenylalanine ammonia lyase -PAL, 4-hydroxycinnamoyl : CoA ligase -4CL, chalcone synthase -CHS, p-hydroxycinnamoyl CoA : quinic/shikimic acid transferase -HCT, caffeoyl-CoA O-methyltransferase -CCoAOMT, caffeic acid/ 5-hydroxyferulic acid 3/5-O-methyltransferase -COMT, synaptic acid dehydrogenase -SAD, hydroxycinnamic alcohol dehydrogenase -CAD, glucosyltransferase -GT in flax seedlings treated pathogenic strains of F. oxysporum (F.ox.) at 6, 12, 24, 36 and 48 h after inoculation were presented as relative quantity (RQ) relative to reference gene (actin) in relation to control (C). The data were obtained from real-time RT-PCR analysis. The data represent the mean ± standard deviations from three independent experiments. The significance of the differences between the means was determined using Student's t test (*P < 0.05, **P < 0.01) chains, which were shorter in the infected flax seedlings (Fig. 7b). Integral intensities of bands in the range of 3400-3425 cm −1 correspond to intra-molecular vibrations O-H × × × O and in the range of 3290-3315 cm −1 correspond to inter-molecular vibrations ν(O-H × × × O) (Fig. 7c). The analyzed results show changes in the arrangement of cellulose chains in flax seedlings after infection with F. oxysporum and a 60 % higher number of hydrogen bonds compared to the non-infected seedlings. Infected flax seedlings display increased the crystallinity index by 16 %, indicating a more organized cellulose structure in the cell wall of the infected flax seedlings. Higher crystallinity also suggests lower reactivity of cellulose, lower water absorption and higher plasticity of cell walls.
Changes in pectin and lignin contents determined by the analysis of differences in the integral intensities of bands at 1735, 1655 and 1609 cm −1 for pectin (Fig. 7d) and at 1337, 1260 and 1245 cm -1 for lignin (Fig. 7e) confirm a significant decrease of pectin and lignin contents after infection with F. oxysporum.

Discussion
Nowadays the research on host-pathogen interaction is of great interest to enable amelioration of plant defence mechanisms.
The current literature describes only the importance of pectin during plant infection and omits other polysaccharide polymers of the cell wall. The research aim of this study was to determine the significance of polysaccharide polymers and lignin during different stages of infection of a pathogenic strain of Fusarium oxysporum. It was suggested to approach this whole scientific problem and pay attention to different polymers and not to omit the possible interaction in the described process.
In order to examine the different stages of flax infection by F. oxysporum, flax seedlings incubated with the pathogen were collected after 6, 12, 24, 36 and 48 h. The choice of incubation period was established experimentally based on the phenotype of infected plants, determining the last incubation time and the PR gene expression whose level significantly increased in plants infected with pathogens [44][45][46]. Expression of genes of β-1,3-glucanase and chitinase in flax incubated with F. oxysporum were induced after 12 h, and their level increased in time. The results indicated that in the 12 h of incubation in spite of the lack of phenotype changes, the pathogen infected the plant and induced a systemic response. In order to examine the first stage of infection comprising pathogen penetration to the root cells and activation of the defence mechanism by the host, the 6th hour of incubation was chosen as the first hour.
The first analyzed polymer in flax infected with F. oxysporum was pectin, the main aim for pathogen in the initial stages of infection. Pectin methylesterases remove the methyl group from homogalacturonan, resulting in loosening of cell wall structures, enabling pectin degradation by polygalacturonases and pectin lyases, and also cellulose and hemicellulose by cellulases and hemicellulases [47]. It is then suggested that during the first stage of infection the methylation level of pectin plays the main role. Highly methylated pectin causes higher resistance to plant infection. In plants, there are endogenous pectin methylesterases that take part in many physiological processes, in which rearrangement of the cell wall is necessary.
Analysis of three isoforms of pectin methylesterases showed that during the first stage of infection expression of PME1 increased, PME3 decreased and PME5 did not change. These results indicate that PME5 probably does not take part in the infection process, but is involved in other physiological processes, although its level increased after 24 h of incubation. PME1 is involved in infection, and the alteration of the expression level of PME1 is putatively caused by a pathogen. A previous study confirmed that infecting fungi can influence the increase in PME gene expression [48], and silencing of a specific PME gene, which depends on the pathogen, led to an increase in plant resistance [49]. Nevertheless, the initial increase in PME1 expression can be explained by the purposeful plant strategy which aims to loosen the structure of pectin to release oligogalacturonides (OG) in order to elicit a plant defence response. Transgenic strawberries with overexpression of one of the isoforms of pectin methylesterase were characterized by increased resistance to Botrytis cinerea infection, which was explained by the release of oligogalacturonides from the cell wall [50]. Summing up, the observed decreases and increases in expression of PME1 and PME3 genes are probably caused by plant defence mechanisms in response to the ongoing infection: initially by maintaining high pectin methylation and then by its rearrangement and/or release of elicitors (OG).
Biochemical analysis of pectin in seedlings after flax infection with F. oxysporum confirmed its rearrangement, which was observed after 48 h in spite of no changes in the total content. Possibly, the activity of endogenous pectin methylesterases of fungi and endogenous pectin methylesterases of flax induced by the pathogen lead during the progressing infection to demethylation of pectin, and then to loosening of its structures by exogenous polygalacturonases and pectin lyases. Pectin rearrangement revealed an increase in the water-soluble fraction and lower content in the CDTA fraction with a simultaneous lack of change in its content in the fraction bound with the cell wall. Loosening of pectin structure is a result of breaking ion bonds by the activity of fungal enzymes [51].
Lowered mRNA levels of genes of polygalacturonase and pectin lyase from the first stage of infection (6 h) until the moment when the cell necrosis of roots is more advanced results from the presence of exogenous fungal genes (PG and PaL), which comprise a form of defence against excessive pectin degradation. In contrast, the decrease in the expression level of most genes of pectin synthesis during the whole infection process might result from redirection of transcription to other genes, which are necessary in abundance during infection. Loosening and rearrangement of pectin made accessible other polymers of the cell wall to exogenous fungal enzymes: hemicellulases and cellulases. Pathogens degrade polysaccharides in order to ensure themselves the nutrients, mainly glucose. Hemicellulose is responsible for the strengthening of the cell wall [52], but until now no participation in the plant response to pathogen infection has been documented. Fusarium oxysporum caused a reduction of the expression level of genes of hemicellulose synthesis at the first stage of infection of flax seedlings. Similarly, most genes of hemicellulose degradation were observed to exhibit a lowered expression level. Such changes might be explained by the altered transcription profile in plants that undergo pathogen infection. Probably, plants redirect their transcription potential to other genes necessary during pathogen-derived stress. Despite the lowering of expression of genes of hemicelluloses metabolism, their content did not change, and only slight lowering of the hemicellulose fraction soluble in 1 M KOH was observed.
Analysis of gene expression of cellulose metabolism in flax seedlings infected with F. oxysporum revealed a decrease in mRNA level of three isoforms of cellulose synthesis genes (CSL1, CSL2 and CSL4) in all studied incubation times. Two other isoforms of cellulose synthase and two isoforms of cellulose degradation genes responded to infection weakly or unspecifically. A decrease in expression of genes of cellulose synthesis was previously described only in Arabidopsis thaliana infected with Botrytis cinerea and revealed the involvement of genes of this polymer in the response to pathogen infection. Such a phenomenon was explained by modulation of transcription, but the hypothesis has not been confirmed yet [53]. A previous study showed that Arabidopsis thaliana with silenced cellulose synthase genes is characterized by increased resistance to Ralstonia solanacearum and Plectosphaerella cucumerina infections [54]. Regarding the difficulties in explaining the role of cellulose synthases in plant resistance to pathogens, the content and structure of cellulose was studied. The quantitative spectrophotometric analysis revealed an increase in cellulose content in seedlings infected with F. oxysporum for 48 h. Similarly, infrared spectroscopic analysis of the cell wall of infected seedlings showed a significant increase of cellulose with a 60 % increase of hydrogen bonds in cellulose and a 16 % increase of cellulose crystallinity. Increased crystallinity of cellulose means a more orderly structure, which is characterized by lower activity, lower water absorbance and increased plasticity of the cell wall. It is then deduced that the increase of cellulose crystallinity is an additional barrier of plants generated during infection, which impedes host cell wall polymer digestion by pathogen enzymes.
Alteration of cellulose crystallinity in the cell wall of seedlings infected with the pathogen is difficult to explain. It is suggested that it was due to rearrangements of hemicelluloses. Despite the lack of expression of most genes of hemicellulose metabolism, one of them exhibited strong activation during infection. The mRNA level of α-galactosidase increased at the first stage of infection, and then was increasing during the period of infection. Putatively, this enzyme that cuts off galactose from galactomannan indirectly enabled generation of hydrogen bonds in cellulose and thereby led to the increase of its crystallinity. This hypothesis needs further research and confirmation. According to other authors [55], enzymes that degrade hemicelluloses in the cell wall and release galactose can directly contribute to the alteration in cellulose crystallinity.
Lignin's involvement in pathogen infections is well known and described in the literature. Increased lignin biosynthesis results from stimulation of the phenylpropanoid pathway, while the lignification process is induced by reactive oxygen species, present in great amounts during pathogen infection [38,39]. Increase in expression of genes of the phenylpropanoid pathway and direct genes of lignin synthesis was observed in wheat infected with Fusarium graminearum and Puccinia triticina [55] and in cotton infected with Verticillium dahliae [56].
Expression of genes of the phenylpropanoid pathway (PAL, HCT, CCoAOMT, COMT and GT) was strongly induced during the later stages of seedling infection with F. oxysporum and increased during the time of infection. These changes were observed after 12 h of incubation, similarly in the case of PR genes, suggesting their involvement in the systemic flax response. At the initial stage of infection of F. oxysporum (after 6 h), genes of lignin synthesishydroxycinnamic alcohol dehydrogenase and synaptic alcohol dehydrogenase -were characterized by a significant increase of mRNA level. The results suggest that lignin synthesis occurs quite fast in order to strengthen the cell wall, but may be limited by the accessibility of substrates. Biochemical analysis and IR spectrometry of the cell wall of flax seedlings after 48 h of infection with F. oxysporum revealed a slight decrease in lignin content. Perhaps this short time from the infection onset is not sufficient to show alteration in this polymer. In cotton infected by pathogenic fungi, despite the significant increase of expression level of genes of lignin metabolism, from 12 h, changes in lignin content were observed after several days following infection onset [56].
Summing up, analysis of the expression level of genes of polymer cell wall metabolism in flax infected with F. oxysporum enabled us to distinguish two group of genes that responding differently to infection. The first group consisted of strongly reacting genes, whose expression levels were increased over 5-10 times: genes of lignin synthesis (phenylalanine ammonia lyase [PAL], glucosyltransferase [GT]) and genes involved in pathogenesis (β-1,3-glucanase and chitinase). They were characterized as systemic response genes, because despite the significant expression level increase, genes responded later (after 12 h of incubation with pathogens), and the maximum expression increase was mostly at 48 h. The second group of genes comprised genes slightly induced. They participate in the metabolic response, and their activation is connected with activity of elicitors (oligogalacturonides) generated as a result of cell wall digestion by fungal enzymes. In this group there are genes whose expression level increased (up to five times), decreased, and genes with different, often unspecific expression alterations. The first subgroup consisted of other lignin synthesis genes, the βglycosidase gene (GLS), and β-1,3-glucanase 2. In the second subgroup of genes with lowered expression level were genes of pectin metabolism (GAUT1, GAUT7, RGXT, PMT, PME3, PG and PaL), hemicellulose metabolism (GMT, GGT, XXT, XYN, XYLb and GS) and cellulose synthesis (cellulose synthases: CSL1, CSL2 and CSL4). The last subgroup, which exhibited different changes in expression, consists of other genes including pectin, hemicellulose and cellulose metabolism. It is suggested that genes that respond weakly to pathogen infection are only transcriptional noise and they do not take part in plant defence mechanisms.

Conclusion
The results of this study confirm that polymers of the cell wall participate in the flax response to Fusarium oxysporum infection through the changes in expression of their genes and rearrangement of cell wall structure. Although the role of pectin and lignin has been widely studied and the literature provides much information, our studies confirmed the contribution of pectin and lignin to the infection process through the loosening of the pectin structure and the increase at mRNA levels of genes participating in lignin synthesis. We provided new information about the role in the pathogen infection of cellulose and hemicellulose, which are expressed by changes in their structure rather than their content, as well as in the profile of their transcriptome. Our preliminary studies encourage further research to precisely determine the mechanism of participation of cellulose and hemicelluloses in the infection process.

Plant material preparation
All experiments were performed on a fibrous flax variety (Linum usitatissimum L. cv. Nike Seed germination and seedling development proceeded under controlled conditions. After 7 days the seedlings were transferred (with the medium) onto PDA medium overgrown with a pathogenic strain of Fusarium oxysporum or control PDA (without the fungus). The fungus was grown on the PDA medium for 5 days prior to the experiment. The seedlings (100 for each stage) were collected after 6, 12, 24, 36 and 48 h, frozen in liquid nitrogen and stored at −70°C.

Gene expression analysis
For total RNA isolation TRIzol reagent (Life Technologies, USA) was used. The isolation was performed according to the producer's protocol in triplicate. In order to purify the isolated RNA from the remaining DNA, a DNaseI kit (Invitrogen, Germany) was used according to the manufacturer's protocol. For reverse transcription reaction, that is cDNA synthesis on an RNA matrix, the High Capacity cDNA Reverse Transcription Kit (Life Technologies, USA) was used according to the manufacturer's instructions.
Real-time PCR was performed in a StepOnePlus Real-Time PCR Systems thermocycler (Applied Biosystems, USA) using the DyNAmo SYBR Green qPCR Kit (Thermo Scientific, USA), according to the manufacturer's protocol. Primers were prepared to not amplify fungal sequences. Their annealing temperature was 57°C and their sequences are presented in Additional file 3: Table S1. Real-time PCR reactions were performed in three repetitions for each of the analyzed samples. Changes in gene expression levels were calculated as relative quantities (RQ) of the reference gene (actin) and presented as x-fold of gene expression in relation to non-treated plants.

Determination of cellulose content
Cellulose content was determined with the anthrone method described by Ververis [57]. Tissue samples (100 mg) were incubated with a mixture of 65 % nitric acid and 80 % acetic acid (1:8 v/v) for 1 h at 100°C and then centrifuged for 5 min at 14 000 rpm. The pellet was then washed twice with water and dissolved in 1 ml of 67 % H 2 SO 4 (v/v) by shaking at room temperature for 1 h. 100 μl of extract was added to 900 μl of cooled solution of 0.2 % anthrone in 67 % sulfuric acid, mixed, heated at 100°C for 15 min and cooled down on ice. Cellulose content in the sample was determined by spectrophotometric measurement of absorbance at 620 nm. Commercially available cellulose was used for standard curve preparation.

Isolation and fractionation of cell wall polysaccharides
Isolation and fractionation of cell wall constituents were performed using modified methods described by Manganaris and Vicente [51,58]. Flax tissue was incubated with 96 % ethanol for 30 min at 100°C in order to inactivate enzymes, to extract components of low molecular mass and to prevent autolysis. Next, the pellet was incubated with the following: 80 % of ethanol for 20 min at 80°C, chloroform:methanol mixture (1:1 v/v) for 1 h at 40°C and acetone for 5 min. The samples were centrifuged (for 5 min at 5000 x g) between each solvent treatments and the supernatant was discarded. The pellet was dried overnight at 37°C and its constituents insoluble in alcohol (AIR) were used for further analysis.
The pellet (AIR) was dissolved in 1 ml of water and was shaken for 12 h at room temperature, centrifuged (6000 × g, 4°C, 10 min) and rinsed with water. Supernatants derived from two centrifugations were combined and used for analysis as a water soluble fraction (WSF). The following fractions of cell wall were obtained by the same procedures, but water was substituted with other solvents: CSF (fraction soluble in CDTA) -50 mM CDTA (1,2-cyclohexylenedinitrilotetraacetic acid) pH 6.5; NSF (fraction soluble in sodium carbonate) -50 mM Na 2 CO 3 with 20 mM NaBH 4 , additionally, neutralization with acetic acid of collected supernatants; K1SF (fraction soluble in 1 M KOH) -1 M KOH with 20 mM NaBH 4 , neutralized with hydrochloric acid; K4SF (fraction soluble in 4 M KOH) -4 M KOH, neutralization with hydrochloric acid. Supernatants of fractions CSF, NSF, K1SF and K4SF were dialyzed to water (membranes for dialysis 3.5-kDa), and all fractions were lyophilized before use for further analysis.

Determination of uronic acid content
The content of uronic acids was determined with the biphenol method [59] with the prior hydrolysis of polysaccharides with sulfuric acid [60]. Lyophilized samples (around 10 mg) were dissolved in 0.1 ml of concentrated sulfuric acid in a cooling bath with shaking for 5 min. Then, 0.1 ml of sulfuric acid, 0.05 ml of water, 0.05 ml of water and 0.7 ml of water were added with shaking between each portion. After the centrifugation (10 min, 2000 x g, RT), 0.1 ml of the supernatant was taken, and 10 μl of 4 M aminosulfonic acid of pH 1.6 was added and 600 μl of 75 mM sodium tetraborate in concentrated sulfuric acid was added. After careful mixing, the samples were incubated for 20 min at 100°C, cooled down and 20 μl of m-hydroxybiphenyl (0.15 %) in 0.5 % NaOH was added, and they were left at room temperature for 10 min. The content of uronic acids was measured spectrophotometrically at 525 nm. As a standard for the calibration curve, glucuronic acid was used.