The Ve-mediated resistance response of the tomato to Verticillium dahliae involves H2O2, peroxidase and lignins and drives PAL gene expression

Background Verticillium dahliae is a fungal pathogen that infects a wide range of hosts. The only known genes for resistance to Verticillium in the Solanaceae are found in the tomato (Solanum lycopersicum) Ve locus, formed by two linked genes, Ve1 and Ve2. To characterize the resistance response mediated by the tomato Ve gene, we inoculated two nearly isogenic tomato lines, LA3030 (ve/ve) and LA3038 (Ve/Ve), with V. dahliae. Results We found induction of H2O2 production in roots of inoculated plants, followed by an increase in peroxidase activity only in roots of inoculated resistant plants. Phenylalanine-ammonia lyase (PAL) activity was also increased in resistant roots 2 hours after inoculation, while induction of PAL activity in susceptible roots was not seen until 48 hours after inoculation. Phenylpropanoid metabolism was also affected, with increases in ferulic acid, p-coumaric acid, vanillin and p-hydroxybenzaldehyde contents in resistant roots after inoculation. Six tomato PAL cDNA sequences (PAL1 - PAL6) were found in the SolGenes tomato EST database. RT-PCR analysis showed that these genes were expressed in all organs of the plant, albeit at different levels. Real-time RT-PCR indicated distinct patterns of expression of the different PAL genes in V. dahliae-inoculated roots. Phylogenetic analysis of 48 partial PAL cDNAs corresponding to 19 plant species grouped angiosperm PAL sequences into four clusters, suggesting functional differences among the six tomato genes, with PAL2 and PAL6 presumably involved in lignification, and the remaining PAL genes implicated in other biological processes. An increase in the synthesis of lignins was found 16 and 28 days after inoculation in both lines; this increase was greater and faster to develop in the resistant line. In both resistant and susceptible inoculated plants, an increase in the ratio of guaiacyl/syringyl units was detected 16 days after inoculation, resulting from the lowered amount of syringyl units in the lignins of inoculated plants. Conclusions The interaction between the tomato and V. dahliae triggered a number of short- and long-term defensive mechanisms. Differences were found between compatible and incompatible interactions, including onset of H2O2 production and activities of peroxidase and PAL, and phenylpropanoid metabolism and synthesis of lignins.


Background
Verticillium wilt, caused by the vascular fungus Verticillium dahliae Kleb., limits the production of a wide range of economically important crops [1]. Once the fungus infects a field, it is very persistent because it colonizes such non-host plants as cereals, which then act as reservoirs for the fungus. Furthermore, the fungus develops resistant structures known as microsclerotia that are capable of survival in the soil for decades. Significant losses are caused by this pathogen, and currently there are no efficient management methods for its control. The Verticillium spp. are among the most damaging pathogens threatening cultivation of the tomato (Solanum lycopersicum), and are responsible for serious economic losses both in greenhouses and in the field. The cultivation of resistant varieties has proven to be an appropriate strategy for combating plant pathogens because of its efficacy, low cost, and limited environmental impact. Recent developments in molecular biology have made it possible to transfer resistance genes between unrelated species and have revealed that the molecular events in the resistance response elicited from recognition of the pathogen are often conserved among plants of the same family [2].
The only Verticillium resistance genes in Solanaceae that are now known are those of the tomato Ve locus. This locus is formed by two linked genes, Ve1 and Ve2, each capable of conferring resistance to different Verticillium species. The structure of the Ve genes suggests that they code for cell-surface glycoproteins with signals for receptor-mediated endocytosis and with leucine zipper motifs (in Ve1) or PEST sequences (in Ve2) [3]. Potato plants transformed with either of the Ve genes acquired resistance against Verticillium, demonstrating that the cell machinery required for the incompatible (no disease) interaction with Verticillium is present and functional in other Solanaceae species.
Events in the early stages of a plant's response to an infecting pathogen determine the degree of colonization and the damage caused. An incompatible interaction generally seems to require the presence in the plant of a cognate resistance gene against an avirulence factor of the pathogen [4]. However, incompatibility is probably related more to the timing of induction of defense genes and factors than to qualitative differences in the set of genes expressed when compared to compatible systems [5]. During an incompatible interaction, plant cells respond with such resistance strategies as (i) generation of reactive oxygen species (ROS), (ii) induction of a hypersensitive response, a localized cell-death reaction that confines the infection to its initial location, (iii) expression of pathogenesis-related genes and other toxic peptides, (iv) synthesis of phytoalexins, (v) stabilization of cell walls, and (vi) closure of the stomata [6].
The production of oxygen intermediates during the so-called oxidative burst is characteristic of the defensive response in plants [7]. Increased levels of ROS, notably the superoxide anion (O 2 -) and hydrogen peroxide (H 2 O 2 ), kill the pathogen or limit colonization by triggering a hypersensitive response in the infected plant tissue [8]. H 2 O 2 is the most stable oxygen intermediate and is involved in the cross-linking of cell wall components [9], regulation of pathogenesis-related gene expression [10], transduction of the hypersensitive response [7], and killing of invading pathogens [11]. H 2 O 2 also acts as a signaling molecule in the cellular collapse that occurs during the hypersensitive response, and in systemic acquired resistance [12].
The peroxidases (PODs; EC 1.11.1.7; donor H 2 O 2oxidoreductase) are heme-containing enzymes that catalyze the oxidation of different substrates using H 2 O 2 . They also produce ROS as a result of their peroxidative and hydroxylic catalytic cycles [13]. Peroxidases are widely distributed in the plant kingdom [14,15] and are active in such physiological processes as ferulate dimerization [16], phenol oxidation [17] and lignification [18,19]; these mechanisms may be activated in the defensive response against pathogens [20].
Some authors have reported the involvement of peroxidases in the formation of phenylpropanoid dimers using equimolar mixtures of hydroxycinnamic acid [21]. Peroxidases contribute to the construction of the cell wall. These actions include intervention in the possible covalent binding of tyrosine residues from extensin and other cell wall glycoproteins with dimers of hydroxycinnamic acid and p-hydroxybenzoic acid bound to pectins or some xylans. Peroxidases are also involved in the biosynthesis of lignins, in the deposition of lignin bound to cell wall glycoproteins, and in the process of suberization [22].
Phenylalanine ammonia lyase (PAL; L-phenylalanine ammonia lyase, EC 4.3.1.5) is the first enzyme in the phenylpropanoid metabolism pathway. PAL catalyzes the deamination of phenylalanine to trans-cinnamic acid, the common precursor for the synthesis of all phenol derivatives. A number of studies have reported increased PAL expression and activity in response to environmental stimuli, such as cold [23], wounding [24] and UV-B light [25]. Silencing of PAL genes in transgenic tobacco lines inhibits the salicylate production normally seen after tobacco mosaic virus infection, and abolishes systemic acquired resistance [26]. PAL over-expression in tobacco leads to production of large amounts of chlorogenic acid, and a marked reduction in sensitivity to infection by the fungus Cercospora nicotianae [27].
In higher plants, PAL is found as a family of homologous genes. The significance of this diversity is unclear, but is consistent with the complexity of metabolic pathways in phenylpropanoid metabolism. Four PAL genes have been described in Arabidopsis [28], five in pine and tomato [29,30], and a total of 16 in the genome of a diploid potato hybrid [31].
The onset of phenylpropanoid metabolism is another crucial defensive mechanism [32]. This results in hydroxycinnamic acids with a characteristic C 6 C 3 phenylpropane skeleton being produced from the primary metabolite phenylalanine. The functions of the phenylpropanoids are very diverse: some are pigments, others phytoalexins, phytoanticipins, UV-protectants or signals mediating the interaction between plants and microorganisms. Furthermore, some phenylpropanoids can polymerize and form defensive structures, such as lignin [33]. There is strong evidence to suggest that esterification of phenols, such as ferulic or p-coumaric acids, to cell walls is a common phenomenon in the expression of resistance [34][35][36]. It is generally thought that phenols play an important role in the modification of the mechanical properties of cell walls [37], limiting polysaccharide degradation by exogenous enzymes [38,39] and increasing cell wall rigidity by linking polysaccharides and lignin [40].
Lignins are amorphous heteropolymers that result from oxidative coupling of p-coumaryl, coniferyl and sinapyl alcohols, forming the subunits H (hydroxyphenyl), G (guaiacyl) and S (syringyl), respectively. Lignins are primarily deposited in cell walls of tissues as tracheids, veins, fibers of xylem and phloem, and schlereids. Lignin composition varies depending on the tissue. For example, lignins type G are predominant in Arabidopsis xylem, while in schlerenchyma cells lignins type S are more commonly found [41]. Lignification of cell walls is a key event in resistance against pests in herbaceous or woody plants and resistant genotypes possess a greater accumulation of lignins [42]. Ratios among lignin subunits change after a pathogen attack [43]. Several mechanisms have been postulated to explain the role of lignins in resistance, including sealing of cell walls [38] or direct biocidal effects of phenolic lignin precursors [44].
In this study we examine H 2 O 2 production, peroxidase and PAL activity, expression of PAL genes and lignin accumulation in two nearly isogenic tomato lines after exposure to V. dahliae, one line carrying the Ve resistance gene and the other not. This characterization of Ve-mediated resistance at the molecular level may allow identification of factors which, after induction, can trigger resistance responses in tomato and in other Solanaceae.

H 2 O 2 content
H 2 O 2 production in roots of control and V. dahliaeinoculated resistant and susceptible tomato plants was measured using the xylenol orange method, which is very sensitive for the detection of low levels of soluble hydroperoxides. Definite variations in the production of H 2 O 2 were found in both tomato lines, whether inoculated or not (Figure 1). At 2 hours post-inoculation (hpi) a dramatic increase in H 2 O 2 content was observed in the roots of inoculated resistant plants, with a maximum peroxide concentration found at 8 hpi, when H 2 O 2 content was three times higher than that in resistant controls or in inoculated susceptible plants. In resistant plants, a second lesser increase was detected between 24 hpi and 48 hpi. In inoculated susceptible plants, a single increase was observed starting at 2 hpi, reaching a peak at 16 hpi. At the final time point monitored, 192 hpi, H 2 O 2 content was similar in control and inoculated roots of the two lines.

Peroxidase (POD) activity
Because the binding of phenolic compounds to the cell wall is mediated by peroxidases and, during plant-pathogen interactions, this binding occurs at the expense of a massive generation of H 2 O 2 [27], we monitored peroxidase activity. In our experiments, the inoculation of resistant plants with V. dahliae led to a rapid increase in peroxidase activity in roots, detectable at 2 hpi ( Figure 2). Similar activity levels were then maintained from 4 and 8 hpi; subsequently another increase in peroxidase activity was observed, reaching a maximum between 24 hpi and 48 hpi. No appreciable changes were detected in inoculated susceptible roots throughout the experiment. At the end of the study period (192 hpi), similar peroxidase activities were found in susceptible, resistant, inoculated and control samples.

Phenylalanine ammonia lyase (PAL) activity
The crucial role of PAL within plant secondary metabolism reflects its function as a catalyst for the first step of phenylpropanoid metabolism. We found an increase in PAL activity in roots of inoculated resistant plants, detectable at 4 hpi, with a maximum at 8 hpi, when activity was approximately 3-fold higher than initial values ( Figure 3). After 16 hpi, PAL activity decreased, and by 24 hpi it was identical to that in controls and in inoculated susceptible plants. In inoculated susceptible plants, PAL activity peaked at 48 hpi, at around 70% of the maximum value seen in inoculated resistant plants, although it was 50% higher than in inoculated resistant plants at the same time point. At 192 hpi, PAL activity was similar in roots of both lines, whether inoculated or not.

RT-PCR of PAL genes from different organs
After detection of the rapid increase in PAL activity in roots of inoculated resistant plants, we studied possible differences in the level of expression of different tomato PAL genes. An extensive search was performed in the NCBI and TIGR databases for cDNA and contig sequences of the tomato PAL genes. As a result, the sequences of 6 tomato PAL genes that differed in their noncoding 3' ends were determined ( Figure 4). The b-tubulin gene was chosen as a constitutive gene for quantification experiments.
Following RNA extraction from roots, hypocotyls, epicotyls, cotyledons, leaves and flowers from resistant and susceptible tomato plants, the corresponding cDNAs were synthesized and PCR amplifications carried out ( Figure 5A). Amplicons from PAL2, PAL3, PAL4 and PAL6 were clearly visualized on ethidium bromidestained agarose gels when 5 ng RNA was used as the starting material. In contrast, for visualization of amplicons, 200 ng RNA was required for PAL5 and 600 ng RNA for PAL1 in all organs analyzed. This finding indicated the existence of different levels of expression of the PAL genes in the tomato. In summary, expression of the six PAL genes could be detected in all organs studied, albeit at different levels, with lower expression levels for PAL5 and lowest for PAL1.
Real-time RT-PCR analysis of PAL2, PAL3, PAL4 and PAL6 in tomato roots We next analyzed root samples from LA3030 and LA3038 tomato plants using real-time RT-PCR. Because of the extremely low levels of expression of the PAL1 and PAL5 genes, which led to inconsistent results from RT-PCR, we considered only the four remaining PAL genes. The results showed slight differences in expression among these genes ( Figure 5B). The highest relative expression value was found for PAL2, followed by PAL3 and PAL4, which had similar expression levels. The lowest level of expression was found for PAL6. This was the only gene that showed any significant difference between LA3030 and LA3038 plants: its expression level was three times higher in LA3038, the resistant plant.

Real-time RT-PCR analysis of the main PAL genes in V. dahliae-inoculated roots
We then used real-time RT-PCR to quantify the expression of PAL2, PAL3, PAL4 and PAL6 at specific intervals following inoculation of resistant and susceptible tomato roots with V. dahliae ( Figure 6).
PAL2 showed maximum expression in inoculated resistant plants at 4 hpi and then decreased to a level similar to the controls by 8 hpi. The maximum level was 6 times higher than that in the control plants of both lines. In contrast, in inoculated susceptible plants, PAL2 showed a significant decrease in expression as early as 4 hpi (P-value < 0.05).
PAL3 expression in inoculated resistant plants showed a detectable increase at 4 hpi, with the maximum level reached at 8 hpi. The maximum level was 9 times higher than that in the controls and in inoculated susceptible plants. In inoculated susceptible plants, there was an increase in gene expression of PAL3 at 48 hpi, a level 3.5 times higher than in the controls and double that seen in the inoculated resistant plants at that time. Interestingly, this was the only increase in expression of a PAL gene seen in susceptible plants.
PAL4 expression was slightly higher in all the inoculated resistant plants than in the inoculated susceptible plants.
PAL6 showed a dramatic increase in expression in the inoculated resistant plants at 8 hpi, with a level of expression approximately 60 times higher than that seen in the control plants of both lines and in the inoculated susceptible plants.

Phylogenetic relationships among plant PAL genes
The sequences of 48 PAL genes, belonging to 19 plant species, stored in the GenBank database (Table 1) were retrieved and compared to the six tomato cDNA sequences. A 116-nucleotide sequence from the 3' end of the coding region was chosen for the comparison. The resulting maximum parsimony phylogenetic tree ( Figure 7) was rooted in the sequences from the most ancient species, the pteridophyte Isoetes lacustis and the spikemoss Selaginella kraussiana. Sequences from all angiosperm species were grouped in four different clusters (A, B, C and D), with the Equisetum arvense (Pteridophyta) and Picea abies (Gymnosperma) sequences in two independent branches. The 6 PAL genes from S. lycopersicum were placed in two different clusters, A and B. In cluster A, besides the tomato genes PAL2 and PAL6, most of the gene sequences belonged to woody plants (Populus kitakamiensis, Populus tremuloides, Coffea canephora and Quercus suber); though sequences from Trifolium pratense, Nicotiana tabacum and Daucus carota were also included in this cluster. The four remaining tomato genes (PAL1, PAL3, PAL4 and PAL5) were placed in cluster B, together with sequences from other dicotyledonous species including N. tabacum, D. carota, Solanum tuberosum, Capsicum chinense and Ipoema batatas. A third cluster, C, was composed of the PAL sequences from monocotyledonous species (Hordeum vulgare, Oryza sativa and Triticum aestivum) and a fourth cluster, D, was exclusively formed by the four PAL sequences from A. thaliana.

Phenolic compounds in roots
To investigate the effects of V. dahliae inoculation on the contents of phenolic compounds, root samples were collected and analyzed. Some phenols, notably hydroxycinnamic acids, are involved in cell wall reinforcement, which enhances plant resistance to fungal colonization of the vascular system. Therefore, we determined whether there were any changes in the levels of specific phenols related to cell wall reinforcement. Two hydroxycinnamic acids, ferulic acid and p-coumaric acid, and their respective benzaldehydes, vanillin and p-hydroxybenzaldehyde, were analyzed using reverse-phase HPLC. Ferulic acid  levels showed only a small increase in inoculated resistant plants at 2 hpi, while p-coumaric acid level showed an increase between 16 and 96 hpi. The vanillin level was significantly higher in inoculated resistant plants between 96 and 192 hpi. p-Hydroxybenzaldehyde was found in inoculated resistant roots between 24 and 192 hpi. No changes in levels of these phenols were observed in susceptible plants ( Table 2).

Lignin content, monomer composition and cross-linking in roots
The total lignin content of cell walls, as measured by acetyl bromide is shown in Table 3. An increase in lignin content was observed in inoculated resistant roots at 16 and 28 days post-inoculation (dpi). In inoculated susceptible roots, the increase was not found until 28 dpi. Nitrobenzene oxidation in an alkaline medium degrades lignins forming p-hydroxybenzaldehyde from hydroxyphenyl (H), vanillin from guaiacyl (G) and syringyl aldehyde from syringyl (S). Table 3 also shows the relative monomeric composition of cell walls of susceptible and resistant roots at 16 and 28 dpi, calculated after nitrobenzene oxidation. In inoculated plants at 16 dpi, there was an increase in percentage of G-units at the expense of S groups. The increase in G groups produces higher G/S ratios at 16 dpi after V. dahliae inoculation. At 28 dpi, the G/S ratio was the same among control and inoculated resistant or susceptible plants, but an increase in H subunits was detected.
The possibility of other qualitative changes in composition and structure of cell walls was assessed using the thioacidolysis degradative method. Fragments resulting from thioacidolysis were identified by GC/MS. We found primarily thioethylated monomers (erythro-and threo-isomers) resulting from aryl-glycerol-β-aryl ether structures derived from coniferyl and sinapyl alcohols and from their respective cinnamyl aldehydes. Terminal O-4 structures from coniferyl alcohol, p-coumaric acid, ferulic acid, vanillin and dihydroconiferyl alcohol were also found. Table 4 shows the values of ionic current for all fragments. A general increase in abundance of a large majority of fragments was found at 28 dpi in inoculated plants, resistant or not, in concordance with the results from acetyl bromide. At 28 dpi, inoculated resistant plants had significant increases in b-O-4 monomers and also showed a 3.3-fold increase in coniferyl aldehyde, a 8.4-fold increase in the terminal O-4 monomers of dihydroconiferyl alcohol, and a 4.2 increase in vanillin.
Comparing the ratios among different groups of monomers, considering their chemical nature and type of bond, at 16 dpi an increase in the ratio guaiacyl/syringyl was found in inoculated plants from around 1.4 to more than 2.0 (Table 5). At 16 dpi, G/S ratios were lower than those found after nitrobenzene oxidation, in agreement with the higher level of G groups of the condensed nucleus of lignins, as compared to the linear fraction susceptible to be degraded by thioacidolysis. This increase disappeared at 28 dpi, when ratios in both control and inoculated plants were similar. The ratio between b-O-4 and O-4 terminal monomers, indicating the level of polymerization of the linear fraction of lignins, decreased at 16 dpi from values close to 100 in the controls to 13.1 in inoculated susceptible and 6.1 in inoculated resistant plants. This decrease in the ratio was primarily due to a dramatic increase in O-4 terminal groups, particularly of ferulic acid, in the inoculated plants.
Increases of the aldehydes, vanillin and cinnamaldehydes, detected by thioacidolysis were confirmed by FT-IR spectroscopy analyses. The FT-IR spectroscopy of root cell walls showed an absorption band at 1650 cm -1 ( Table  6). This band is clearly attributable to the C = O stretching vibration of conjugated/aromatic aldehydes, in which the carbonyl oxygen atom sustains either intramolecular or intermolecular H-bonds [43]. At 16 dpi, inoculated plants showed an increased level of conjugated/aromatic aldehydes in cell walls, which increased further at 28 dpi.

Discussion
One of the objectives of this study was to monitor any variations in H 2 O 2 content and peroxidase and PAL enzyme activities during the tomato-V. dahliae interaction. Both enzymes are frequently considered to be key players in the development of plant resistance against pathogens [45]. Their coordinated actions results in changes in the relative amounts of phenylpropanoid compounds, commonly regarded as defensive compounds themselves [46]. To analyze this, two nearly isogenic tomato lines differing in the presence of the Ve gene conferring resistance to Verticillium were used. H 2 O 2 production is one of the markers of the oxidative burst that is one of the most rapid events associated with the hypersensitive response in plant-pathogen interactions [12,47]. After V. dahliae inoculation of resistant tomato plants, we observed a rapid increase in H 2 O 2 content in roots at 2 hpi, with a second smaller peak at 48 hpi. The increase in H 2 O 2 content was slightly delayed in roots of susceptible plants, reaching its peak at 16 hpi. The rapid increase seen in resistant plants is probably related to several known defense mechanisms. H 2 O 2 has been proposed to act directly as a toxic compound for microbes [11]; to contribute to cell wall reinforcement in plants [48]; and to be responsible for lipid peroxidation and salicylic acid synthesis [49]. In addition, H 2 O 2 may also play a role in the signal transduction cascade, triggering the coordinate expression of different genes involved in the defensive response, such as those responsible for the hypersensitive response or for the synthesis of pathogenesis-related proteins [50,51]. The second increase in H 2 O 2 content seen in resistant roots at 48 hpi may reflect the onset of a systemic response, and is in agreement with the timing of systemic pathogenesis-related (PR) gene expression and the beginning of salicylic acid accumulation in the tomato [52]. When we monitored changes in peroxidase activity following V. dahliae inoculation, we found a small early increase in roots of resistant plants. By 96 hpi, peroxidase activity was the same as in the control samples. Notably, the similar induction of H 2 O 2 production in susceptible and resistant plants was not paralleled by comparable increases in peroxidase activity in susceptible plants. Peroxidase activity is the result of the action of a large number of enzymes with similar functions. H 2 O 2 detoxification in roots of resistant plants is probably due to simultaneous actions of peroxidases and catalase, while the latter is probably predominant in susceptible plants. In any case, H 2 O 2 detoxification is an efficient mechanism in both resistant and susceptible roots. There was a general decline in peroxidase activity over time in all samples, probably related to the aging processes.
Considering the essential role of phenylalanine ammonia lyase (PAL) in phenolic metabolism, we decided to   Relationships among the monomeric degradation products resulting from thioacidolysis of root cell walls from control and V. dahliae-inoculated LA3030 (susceptible) and LA3038 (resistant) tomato plants at 16 [54]. Our results seem to support the involvement of H 2 O 2 in the induction of the expression of the PAL genes in resistant tomato plant roots, although this effect was less apparent in susceptible plants.
Using RT-PCR, we attempted to detect any changes in gene expression to identify which PAL genes are responsible for the observed increases in enzymatic activity. PAL is a multigenic system composed of a variable number of highly homologous genes. We found sequences corresponding to at least 6 different PAL genes that differed in their noncoding 3' ends in the Sol-Genes tomato EST database.
In our first approach, we were able to detect expression of all six PAL genes in tomato roots, cotyledons, hypocotyls, epicotyls, leaves and flowers using RT-PCR, although some differences in the levels of expression were observed.
To assess whether the differences in PAL activity in V. dahliae-challenged roots from resistant and susceptible tomato lines were due to a coordinated increase in the expression of the different PAL genes, or whether there was differential regulation of these genes, we assessed changes in expression of PAL genes in inoculated roots using real-time RT-PCR. Because PAL1 and PAL5 amplification from root cDNA samples was barely detectable after even 30 PCR cycles, we considered only the remaining PAL genes. This is somewhat contradictory with a previous report in which the expression of individual tomato PAL genes was analyzed [30]; in that paper, most of the findings referred to PAL1 and PAL5. In more recent works, a predominant expression of PAL5 has been detected in tomato roots and leaves using RT-PCR [55,56]. In our RT-PCR study, PAL2 was the most abundant transcript, followed by PAL3, PAL4 and PAL6.
Our results revealed differing patterns of expression of the PAL genes following V. dahliae inoculation. Most of the total increase in PAL expression in resistant roots in the first 4 hours after inoculation was from increased PAL2 transcription in the initial moments of the interaction. By 8 hpi, however, expression of PAL2 had returned to its original level and PAL3 and PAL6 expression was increased in the roots of resistant plants. PAL6 showed a dramatic increase in expression at 8 hpi, which together with the increase in PAL3 expression, could explain the maximum peak in PAL activity observed at this time point in inoculated resistant roots. This may indicate that PAL6 was the main gene responsible for the increase in PAL activity detected in roots of resistant plants at 8 hours after V. dahliae inoculation.
Apart from the 4-fold change in PAL3 expression at 48 hpi, no clear changes in PAL expression were  PAL isoforms AtPAL1, 2 and 4 indicate that all three followed standard Michaelis-Menten kinetics. However, AtPAL3 was estimated to have a catalytic efficacy 500to 1000-fold lower than that of the other PAL isoforms, based on its higher K M and very low K cat values [28].
We next decided to analyze the structural relationship of tomato PAL genes with PAL genes from other plant species, specifically to compare them with isoforms of known biological functions and relevance from other plants. To check for phylogenetic relationships among the different PAL genes, 48 gene sequences located in the National Center for Biotechnology Information (NCBI) database and The Institute for Genomic Research (TIGR) annotated database were aligned. Because four of the tomato sequences came from partial cDNAs, a total of 116 nucleotides corresponding to the 3' end of the coding region were considered. Preliminary analysis of the alignment of these sequences revealed extensive homology, therefore this region was used for phylogenetic reconstruction.
The resulting topology grouped PAL2 and PAL6 within cluster A, which includes most of the isoforms and homologues from other woody dicotyledonous species. One of the genes in this cluster, PAL1 from Populus tremuloides, is found in non-lignified cells showing accumulation of condensed tannins, while PAL2 from the same species, also included in cluster A, appears in lignification structures and conducting elements from the xylem and the phloem; its expression decreases once the lignification process is complete [57]. Osakabe [58] measured levels of four PAL genes in Populus kitakamiensis stems developing secondary xylem, and found the highest levels for the palg2b transcript, which is also in cluster A.
Cluster B includes the other four tomato PAL genes (PAL1, PAL3, PAL4 and PAL5) and sequences from dicotyledonous plants including N. tabacum, D. carota, S. tuberosum, C. chinense and I. batatas. IPBPAL from I. batatas was induced after mechanical damage [59]; Tag 402 from N. tabacum was induced 4-fold 2 hours after a methyl jasmonate treatment; AJ539006 from N. tabacum is positively regulated by H 2 O 2 [60]; expression of D. carota DcPAL1 in a cell suspension could be induced by a fungal elicitor, UV-B irradiation or a dilution effect [25].
The clustering of genes from plants like T. pratense, A. thaliana and the monocot species suggests that PAL duplication and divergence has occurred independently within some plant lineages. On the other hand, the presence of genes from tomato and tobacco in different clusters seems to reflect ancient duplication of other PAL genes and divergent evolution. Further sequencing of additional PAL genes from these and other species may enable more detailed elucidation of the phylogenetic history of the PAL gene family.
PAL2 and PAL6 genes were grouped with genes from other species in which the process of lignification is very active. At the same time, increases in PAL2 and PAL6 gene expression were only found in inoculated resistant plants. PAL3, clustered with those genes from other species involved in the resistance to various stressors, showed increased expression in both inoculated tomato lines. The distribution of the tomato PAL genes in two different clusters may reflect functional differences among isoforms, with possible involvement of PAL6 and, very probably, PAL2 in lignification, and roles for PAL1, PAL3, PAL4 and PAL5 in other biological processes.
While some phenolic compounds occur constitutively and serve as pathogen inhibitors in non-host resistance [61,62], others are synthesized de novo in response to fungal infection and act as part of an active defense response [63]. In our study, the only detectable change in total phenolic content after V. dahliae inoculation was a small but statistically significant increase in total phenolics in roots of inoculated resistant plants at 2 hpi.
To identify differences in the relative amounts of different phenolics, we analyzed the content of bound phenolics by HPLC. We saw a slight increase in ferulic acid levels in roots of inoculated resistant plants at 2 hpi. We also found differences in p-coumaric acid, vanillin and p-hydroxybenzaldehyde contents at later post-infection times. These phenols are related to cell wall esterification, which enhances plant resistance against fungal enzymes [35]. The most marked changes were observed in the content of p-coumaric acid where we found increased levels of this compound in roots of inoculated resistant tomato plants between 16 and 96 hpi. This increase coincided with maximum peroxidase activity in resistant roots, and may be related to the decline seen in the initial increase in H 2 O 2 at these times points in inoculated resistant plants. p-Coumaric acid has a very important role in the maintenance of cell walls in plants. It mediates the cross-linking of lignins to polysaccharides in cell walls of gramineous plants [64]. Vanillin content increased from 96 hpi on, when peroxidase activity was higher in inoculated resistant plants than in control plants of both lines or in inoculated susceptible plants. To summarize, V. dahliae infection had a clear influence on phenolic metabolism in the tomato.
Lignification and reinforcement of cell walls are important processes in the response of plants against fungal infection [65][66][67]. A lignified cell wall is waterresistant and thus less accessible to fungal cell walldegrading enzymes [68]. Smit and Dubery [69] observed an increase in synthesis and deposition of lignins and similar polymers after exposure of cotton hypocotyls to an elicitor of V. dahliae. The active lignification phase was preceded by increased activity of PAL, cinnamyl alcohol dehydrogenase and cell-bound peroxidases. They also found that the response of a resistant cultivar was faster and more intense than that of a susceptible one. Pomar et al. [43] found that inoculation of pepper varieties differing in their degree of resistance against V. dahliae triggered a significant increase in the amount of lignin. In our study, inoculation with V. dahliae induced a significant increase in the total amount of lignin in tomato roots in both the susceptible line LA3030 and in the resistant line LA3038, although at earlier times in the latter.
Differences in monolignol composition between the lignin of healthy plants and resistant-related lignin in infected plants have been described in several studies of interaction with a variety of pathogens [66,70]. It has also been observed that the monomeric composition and degree of crossing-over in lignins from inoculated pepper plants were closely related to the maintenance of the integrity of the photosynthetic system and thus with tolerance to the presence of the pathogen [43]. From nitrobenzene oxidation, we found that at 16 dpi there was an increase in the proportion of G groups compared to S groups. In addition, at 28 dpi, both cultivars showed an increased proportion of H groups. Presumably, this increase was due to the incorporation of p-coumaryl alcohol in the condensed nucleus. However, the possibility that some of the benzaldehyde units quantified by nitrobenzene oxidation could come from hydroxycinnamic acids bound to cell walls, because they share the aromatic skeleton with monolignols, cannot be ruled out.
The ratio between guaiacyl and syringyl moieties has an important influence on the type and frequency of lignin interunit linkages and, consequently, on lignin structure [71]. In our thioacidolysis analyses of tomato lignins, there were remarkable changes in both resistant and susceptible roots. At 16 dpi there was an increase in the G/S ratio in all inoculated plants, consequent to the reduction of S units. This kind of lignification could be analogous to that found in primary walls, where polymerization is rapid and rich in β-5, β-1, β-β, 5-5 y 5-O-4 bonds. These young lignins are rich in hydroxyphenyl (H) and guaiacyl (G) groups and poor in syringyl (S) groups. At 28 dpi, the differences between control and inoculated plants disappeared, as a result of a large incorporation of S groups, characteristic of slow polymerization. The chemistry of sinapyl alcohol radicals predicts that their only coupling modes are β-β and β-O-4, because the possible resonance structures are R O4 and R β [72]. The β-β mode is less favored than β-O-4 at low concentrations, therefore most sinapyl alcohol sources used in lignin biosynthesis are incorporated in polymers rich in β-O-4 bonds. Thus, the higher incorporation of syringyl groups observed at 28 dpi in inoculated plants should be accompanied by a higher number of β-O-4 bonds and, consequently, by a higher proportion of the linear fraction of lignins.
Another change observed after V. dahliae inoculation was the increase in O-4 terminal units of DHCA, p-coumaric acid, coniferyl alcohol, vanillin and ferulic acid. The increases were greater in the resistant line and were quantitatively highest for ferulic acid. The O-4 terminal unit can act as nucleation points for the growth of lignins after coupling of with a monolignol radical in position β [73]. There was also a remarkable increase in aldehyde groups in inoculated plants, especially in b-O-4 coniferyl aldehyde in the resistant line at 28 dpi. The increase observed in the quantity of aldehyde groups in inoculated cell walls was confirmed by FT-IR analysis. The presence of carbonyl groups could have an inhibitory effect on fungal enzymes, because these groups can react with the amino groups of enzymes, inactivating them. In addition, these aldehydes confer a hydrophobic character to lignins, protecting them against the action of cellulolytic enzymes [74].
In summary, our analysis of lignins indicates that there was an increase in lignin synthesis following inoculation with V. dahliae. This increase was greater and faster in the resistant line, where two phases were detected. Initially, there was an accumulation of lignins with a high degree of crossing-over, apparently rich in G and H groups, particularly in its condensed nucleus. This increase is accompanied by deposition of phenolic units that possibly act as initiation points that would allow the growth of highly polymerized linear lignins rich in S units in the slower second phase.

Conclusions
Ve-mediated resistance again Verticillium spp. is a complex process that triggers molecular responses at several levels, including H 2 O 2 accumulation, increased peroxidase activity, differential production of phenylpropanoids, specific regulation of PAL genes and differential deposition of lignins. These events are most likely the result of the coordinated activation of different defensive responses, resulting in the production of ROS, the induction of expression of defense genes, the production of antimicrobial compounds and the reinforcement of cell walls. The comparison of sequences from PAL genes also seems to reveal the involvement of different PAL isoforms in different biological processes. The generation of specific tomato PAL mutants, or the identification by TILLING screening of tomato plants with defective PAL alleles, together with an analysis of their resistance against pathogens, their relative phenylpropanoid content and the structure of their cell walls, may elucidate the actual function of the various PAL isoforms in lignification or other cellular mechanisms.

Plant material
Seeds from the near-isogenic tomato (S. lycopersicum cv Gardener) lines LA3030 and LA3038 were provided by the C.M. Rick Tomato Genetics Resource Center (UC Davis, CA, USA). LA3038 carries the Ve gene conferring resistance against Verticillium spp. and the I gene for resistance to Fusarium oxysporum f. sp. lycopersici. The seeds were surface-sterilized by immersion in 10% bleach for 30 minutes and thoroughly rinsed before sowing in sterile perlite. One-week-old plantlets were transplanted into individual pots, placed on heating mats and kept in the greenhouse through September and October with a 16:8 h photoperiod at 342 μmol m -2 s -1 at temperatures ranging from 18 to 25°C. The plants were supplemented once a week with a water soluble fertilizer (N-P-K: 15-2.2-9 ).

Fungal material
The virulent V. dahliae Kleb. isolate VD53 was used [75]. To ensure virulence, the pathogen was freshly isolated from infected plants before each inoculation. After isolation it was cultured on potato dextrose agar (PDA) plates.

Fungal inoculation
Four-week-old plants were inoculated with inoculum prepared from the V. dahliae cultures grown on PDA plates. After 25 days of culture at 25°C in the dark, 5 ml of sterile distilled water was added to each plate and the mycelia were brushed away with a rubber spatula. The suspension was filtered through a double layer of cheesecloth. The conidia were counted in a Thoma chamber and the concentration adjusted to 10 7 conidia ml -1 . One ml of the suspension was directly pipetted onto the soil surface of each pot. Plants in the control group received 1 ml of sterile water. After inoculation, all plants were kept in the greenhouse conditions described above.

Measurement of H 2 O 2
Root samples were homogenized in extraction buffer (Tris-acetate 50 mM, pH 5.0) using a mortar and pestle.
The mixture was filtered through a double layer of cheesecloth and centrifuged at 14,000 g for 30 minutes at 4°C. The supernatant was transferred into a clean tube and the pellet was discarded.
H 2 O 2 was quantified in the roots of control and inoculated LA3030 and LA3038 plants using the xylenol orange method [76] that is based on the oxidation of Fe 2+ ions by peroxide, followed by colorimetric detection of the reaction of Fe 3+ with the sodium salt of xylenol orange. Five hundred μl of the reaction mixture (500 μM ferrous ammonium sulfate, 50 mM H 2 SO 4 , 200 μM xylenol orange and 200 mM sorbitol) was added to 500 μl of root crude extract. After 45 minutes, absorbance by the Fe 3+ -xylenol orange complex was measured at 560 nm. Data were normalized with reference to fresh weight and are presented as H 2 O 2 concentrations (μM).

Measurement of peroxidase activity
Sample extraction was performed as described above. Peroxidase activity was determined at 25°C in 50 mM Tris-acetate at pH 5.0 and 0.5 mM H 2 O 2 , supplemented with 1 mM 4-methoxy-α-naphthol as the electron donor [43].

Measurement of phenylalanine ammonia lyase activity
Total PAL enzyme was extracted by the method of El Ghaouth et al. [77]. The root samples were homogenized at 4°C in 50 mM sodium acetate, pH 5.0; the lysate was then centrifuged at 10,000 g for 15 minutes and the supernatant was collected. PAL activity was measured in this fraction using the method of Beaudoin-Eagan and Thorpe [78]. The extract was incubated for 2 h at 37°C in 10 mM L-phenylalanine, 0.5 M Tris-HCl, pH 8.0. The reaction was stopped by adding 5 M HCl. The mixture was centrifuged and the amount of trans-cinnamic acid formed in the supernatant was measured spectrophotometrically at 290 nm. PAL activity was expressed as μg of cinnamic acid formed per μg of protein.
To confirm PAL activity, an inhibition assay using different amounts of the cinnamic acid derivatives ferulic acid, trans-cinnamic acid, coumaric acid and caffeic acid was performed. Total inhibition of commercial PAL (Sigma-Aldrich, Madrid) was found at concentrations of 10 -10 μg μl -1 .

RNA extraction and cDNA synthesis
Samples from roots, hypocotyls, epicotyls, cotyledons, leaves and flowers were taken from 6-week-old LA3030 and LA3038 plants. Samples from LA3030 and LA3038 roots were harvested at different times after fungal induction and stored at -80°C for further use. Total RNA was extracted from frozen samples using the Aurum™ Total RNA Mini Kit (Bio-Rad, Barcelona) following the manufacturer's instructions. RNA quantity was measured spectrophotometrically and its integrity was confirmed using 1.2% agarose-formaldehyde gel electrophoresis [79]. First-strand cDNA was synthesized from 100 ng of total RNA using the iScript cDNA Synthesis Kit (Bio-Rad), following the protocol supplied by the manufacturer.

Primer design, PCR and real-time PCR
Sequences for the different tomato PAL genes were retrieved from the databases of the National Center for Biotechnology Information (NCBI) http://www.ncbi.nlm. nih.gov/ (PAL1 and PAL5) and The Institute for Genomic Research (TIGR) http://www.tigr.org/tdb/agi/ (PAL2, PAL3, PAL4 and PAL6). Specific primers were designed using the program Primer-3 [80]; primer sequences are detailed in Table 7. The amplification conditions of the different PAL genes were optimized. The thermal cycling conditions consisted of an initial denaturation step at 95°C for 2 minutes followed by 30 cycles at 95°C for 30 s, 60°C for 25 s, 72°C for 50 s, and a final step at 72°C for 5 minutes. The respective PCR products were sequenced and corresponded to the expected amplicons. PCR products were separated on 1% agarose gels and visualized after staining with ethidium bromide.
Real-time RT-PCR was performed in 50 μl of a reaction mixture composed of 2.5 μl cDNA, 1X iQ SYBR Green Supermix (Bio-Rad) and 0.3 μM of each gene-specific primer, using an iCycler iQ system (Bio-Rad) and the same thermal cycling conditions described above. The Optical System Software 3.0 (Bio-Rad) was used to analyze the results. RT-PCR specificity was confirmed by identification of a single peak in the melting curve analysis.
The β-tubulin gene was used as a constitutively expressed endogenous control. To determine the amplification efficiencies for the PAL and tubulin genes, we used five-fold serial dilutions of cDNA. Efficiencies greater than 95% were obtained in all cases. For quantification, an efficiency-corrected C t model was used [81]. For the direct comparison of levels of expression among PAL genes in roots, the expression of each gene was related to that of the gene with the lowest C t (PAL2). Each test was repeated twice and each measurement was performed in duplicate.

Phylogenetic analyses
All sequences representing PAL genes from different organisms were extracted from the National Center for Biotechnology Information (NCBI) and The Institute for Genomic Research (TIGR) annotated databases. We included 48 nucleotide sequences, restricted to a homologous region of 116 nucleotides from the 3' end of the coding region.
Multiple alignments of the nucleotide sequences were conducted using the BioEdit [82] and ClustalX [83] programs with default parameters as specified by each program. The trees were produced using maximum parsimony with the MEGA program, version 3.1 [84]. Reliability of the resulting topologies was tested by bootstrap (1,000 replicates) for each interior branch of the trees.

Analysis of phenolic compounds
Root samples were lyophilized and ground in liquid nitrogen, after which 100-300 mg was homogenized in 70% methanol and incubated for 30 minutes at 80°C. After cooling to room temperature, water was added to a volume of 2 ml and samples were centrifuged at 1,300 g for 5 minutes. The pellet was then resuspended in 2 ml 70% methanol and centrifuged again using the same conditions. The supernatants from both extractions were combined and extracted twice with ethyl acetate after methanol evaporation under vacuum. The resulting supernatants were evaporated and resuspended in 1 ml methanol. This supernatant was retained as the soluble phenols fraction. Two ml of 4 N NaOH was added to the pellet from the soluble phenols extraction, saturated with nitrogen, and incubated at 170°C for 2 hours. After cooling, 2 ml H 2 O and 1 ml 35% HCl were added. The reaction mixture was centrifuged for 10 minutes at 3,200 g and the supernatant was retained. Phenolic compounds were extracted 3 times with ethyl acetate, then anhydrous sodium sulfate was added and the samples were evaporated to dryness in a Rotavapor R-205 (Buchi, Postfach, Switzerland) and resuspended in 1 ml methanol; this fraction contained linked phenols.
The quantitative determination of free and linked phenolics was accomplished using Folin-Ciocalteu reagent [85], with ferulic acid as the standard. The sum of both fractions was considered to be the total phenols content.
For identification and quantification of individual compounds, the linked phenol samples were analyzed using reverse-phase HPLC on an Alliance system (Waters, Barcelona) equipped with a Waters 996 photodiode detector. The reverse-phase employed a Spherisorb ODS2 C18 analytical column (Waters) with a Spherisorb ODS2 C18 precolumn. Ten μl of each sample was injected and run at a flow of 1 ml min-1 at 25°C . Compounds were detected between 225 and 400 nm, and quantification was performed at 290 nm, using the corresponding standards.

Cell wall isolation and lignin analysis
Cell walls were prepared using a Triton X-100 washing procedure that included as the last steps three washes with ethanol and three washes with diethyl ether [86]. Lignin quantification was performed using acetyl bromide [87]. Alkaline nitrobenzene oxidation of lignifying cell walls and HPLC analyses were performed essentially as described in Pomar et al. [88]. Quantification of p-hydroxybenzaldehyde, vanillin and syringaldehyde was accomplished at 280 nm, using the corresponding standards. Thioacidolysis, which solubilizes the β-O-4 lignin core, and gas chromatography-mass spectrometry (GC-MS) analyses were performed using the Thermo Finnigan Trace GC gas chromatograph, Thermo Finnigan Polaris Q mass spectrometer and DB-XLB, J&W (60 m × 0.25 mm I.D.) column [86]. Mass spectra were recorded at 70 eV. Quantification of chromatographic peaks utilized total ionic current (TIC) chromatograms. Fourier transform infrared spectra of finely ground cell wall samples were recorded on a Bruker Vector 22 FT-IR spectrophotometer (Bruker Optics Madrid Spain).

Statistics
All experiments and measurements were performed in triplicate. The Student t-test was used for two-group comparisons and ANOVA followed by an unpaired Student t-test with Bonferroni's correction was used for multiple group comparisons. The differences were considered significant when P-value was < 0.05