*2.5. Qualitative Epitope Changes in CWs Produced by the Pph Inoculation and INA Priming*

As the Pph inoculation, as well as INA priming, involved a displacement of polysaccharides populations extracted across all the fractions (Figures 5 and 6), and the increment in Mw of these polysaccharides (Figure 7), an epitope characterization analysis of these CWs was carried out (Figure 8). Therefore, a screening of CW qualitative changes was performed by immunodot assays (IDAs) against different CW proteins such as extensins (recognized by LM1 monoclonal antibody [66]) and arabinogalactan proteins (AGPs, recognized by LM2 [9]). HGs with different degree and pattern of methyl esterification were screened by means of JIM5 or JIM7 antibodies, which probe to low and high methyl esterified HGs respectively [67]. The β-1,4-galactan and α-1,5-arabinan side chains of RG-I were also probed by using LM5 or LM6 antibodies, respectively [9]). In addition, hemicellulose epitopes were analyzed with LM11, which detects xylans and AXs [68]), and LM15, which labels non-fucosylated XG [69]).

The upper side of the heatmap shows the relationship among antibodies binding profiles. Xylan/arabinoxylan (LM11) and extensin (LM1) antibodies clustered together (cluster A) and had less CW epitopes than pectic and hemicellulosic antibodies, which clustered in the other group (cluster B). Within group B, HG antibodies (JIM5 and JIM7) had the more intense binding profile (cluster B.1), while RG-I (LM5 and LM6), XG (LM15), and AGP (LM2) antibodies grouped in an independent sub-cluster (B.2).

Attending to the left part of the heatmap, two clusters were observed: cluster C, which included different fractions from INA-pretreatments (INA and INA + Pph), and cluster D, in which Na2CO3, KI, and SnCR fractions from Mock and Mock + Pph plants were included. Interestingly, Na2CO3 and KI fraction always grouped together although in different clusters depending on the treatment (cluster C.1a for INA and INA + Pph, and cluster D for Mock and Mock + Pph), whereas CDTA and KII fractions were clustered independently in cluster C.1b and cluster C.2, respectively. The clustering analysis after IDAs revealed a similar epitope profile for CDTA and 4N KOH (KII) extracted polysaccharides independently of the treatment. Besides, INA and INA + Pph linked to a differentiated labelling pattern in Na2CO3 and 0.1N KOH (KI) extracted polysaccharides when compared to Mock or Mock + Pph. As expected, in Mock plants the pectin fraction extracted with CDTA was enriched in HG with high and low degree of methyl esterification (JIM7 and JIM5) and RG-I (LM5 and LM6), but also non-fucosylated XG and AGPs were found. Mock + Pph produced only a slight increase in HG methylation degree (JIM7) in the CDTA fraction. However, INA and INA + Pph also increased the presence of extensins (LM1) and AGPs (LM2), and decreased galactan side chains of RG-I (LM5) and XG (LM15) epitopes, mainly in INA + Pph.

In Mock, CW polysaccharides extracted in Na2CO3 and KI fractions were enriched in RG-I (LM5 and LM6), HG (JIM5 and JIM7), XG (LM15) and AGPs (LM2), but the Na2CO3 fraction had only a small proportion of extensin epitopes (LM1), and KI had only a small proportion of xylan epitopes (LM11). Mock + Pph did not substantially change the epitope distribution in Na2CO3 and KI fractions. However, the epitope profile of these fractions dramatically changed after INA pretreatment, mainly due to the increase in extensin (LM1) and HG (JIM5 and JIM7) epitopes, especially in the Na2CO3 fraction after the Pph inoculation (INA + Pph).

The KII fraction was mainly composed of XG (LM15) and xylans (LM11) in Mock, although HG (JIM5 and JIM7), RG-I (LM5 and LM6) and AGP (LM2) epitopes were also detected. Mock + Pph itself did not substantially affect this profile, but the INA pretreatment produced a pectin increment, mainly HG with a high degree of methylation (JIM7) and RG-I (LM5 and LM6) and an increase in extensin epitopes.

Finally, the SnCR fraction was more variable in composition depending on the treatment. In Mock, SnCR was composed of pectins (RG-I and HG), hemicelluloses (XG and Xylans), and AGP. However, in Mock + Pph a decrease in hemicellulosic (LM11 and LM15) and HG (JIM5 and JIM7) epitopes was observed. The composition of this fraction was also different between INA pretreatments. The INA pretreatment increased, as in other fractions, the presence of HG (JIM5 and JM7) and extensin (LM1) epitopes, but the Pph inoculation (INA + Pph) also increase the abundance of hemicellulose (LM15) and AGP (LM2) epitopes compared to Mock.

**Figure 8.** Heatmap of data obtained from IDAs of fractioned CW from Mock, Mock + Pph, INA and INA + Pph conditions. The value assigned from 0 to 5, depending on the marked dilution after development, is represented in a color key for the 5 levels. Fractions of each treatment are clustered in the horizontal axis, while monoclonal antibodies are clustered in the vertical axis. For easy visualization, the next clusters were established with respect to antibodies clusterization: cluster A, grouping LM1 and LM11; cluster B, divided into B.1, containing JIM5 and JIM7, and B.2, where LM5, LM6, LM2 and LM15 were included. Regarding fractions and treatments, the clusters obtained were: cluster C, divided into C.1a (relative to Na2CO3 and KI fraction of INA treatments), C.1b (where all CDTA fractions were located), and C.2, with KII fractions mostly; cluster D, relative to Na2CO3 and KI fraction but of Mock and Mock + Pph.
