This is the accepted version of the journal article: Kashyap, Anurag; Jiménez Jiménez, Álvaro Luis; Zhang, Weiqi; [et al.]. «In- duced ligno-suberin vascular coating and tyramine-derived hydroxycinnamic acid amides restrict Ralstonia solanacearum colonization in resistant tomato». New Phytologist, (February 2022). DOI 10.1111/nph.17982 This version is available at https://ddd.uab.cat/record/255144 under the terms of the license Page 1 of 59 New Phytologist 1Induced ligno-suberin vascular coating and tyramine-derived hydroxycinnamic acid amides restrict Ralstonia solanacearum colonization in resistant tomato Short title: A pathogen-induced ligno-suberin vascular coating Anurag Kashyap1,¥,, Álvaro Luis Jiménez-Jiménez1,#, Weiqi Zhang1,#, Montserrat Capellades1,2, Sumithra Srinivasan3, Anna Laromaine3, Olga Serra4, Mercè Figueras4, Jorge Rencoret5, Ana Gutiérrez5, Marc Valls1,6, Nuria S. Coll1,2,* 1 Centre for Research in Agricultural Genomics (CRAG), CSIC-IRTA-UAB-UB, Campus UAB, 08193 Bellaterra, Spain 2 Consejo Superior de Investigaciones Científicas (CSIC), 08001 Barcelona, Spain 3 Institute of Material Science of Barcelona (ICMAB), CSIC, Campus UAB, 08193 Bellaterra, Spain 4 Laboratori del Suro, Biology Department, Universitat de Girona, Campus Montilivi, 17003 Girona, Spain 5 Institute of Natural Resources and Agrobiology of Seville (IRNAS), CSIC, 41012 Seville, Spain 6 Department of Genetics, Universitat de Barcelona, 08028 Barcelona, Spain # These authors contributed equally to the work * Author for correspondence: Nuria S. Coll. e-mail: nuria.sanchez-coll@cragenomica.es Total word count: 6806 Word count Introduction: 1063 Word count Materials and Methods: 1244 Word count Results: 2490 Word count Discussion: 2009 Number of figures: 9 (1, 2, 3, 4, 5, 7, 8, 9 in color) Supporting information is included ORCID AUTHORS: New Phytologist Page 2 of 59 Anurag Kashyap 0000-0003-2622-8209 Álvaro Jiménez-Jiménez 0000-0002-9406-4595 Weiqi Zhang 0000-0002-2535-1398 Montse Capellades 0000-0001-9514-2885 Sumithra Srinivasan 0000-0002-0473-9801 Anna Laromaine 0000-0002-4764-0780 Olga Serra 0000-0002-1678-0932 Mercè Figueras 0000-0002-6288-1830 Jorge Rencoret 0000-0003-2728-7331 Ana Gutiérrez 0000-0002-8823-9029 Marc Valls 0000-0003-2312-0091 Núria S. Coll 0000-0002-8889-0399 TWITTER ACCOUNTS: @BactoDeath @anuragkashyap11 @cork_lab @alaromaine @RencoretJorge @sumithrayasaswi Page 3 of 59 New Phytologist 2 Summary 3 4  Tomato varieties resistant to the bacterial wilt pathogen Ralstonia solanacearum have 5 the ability to restrict bacterial movement in the plant. Inducible vascular cell wall 6 reinforcements seem to play a key role in confining R. solanacearum into the xylem 7 vasculature of resistant tomato. However, the type of compounds involved in such 8 vascular physico-chemical barriers remain understudied, while being a key component 9 of resistance. 10  Here we use a combination of histological and live-imaging techniques, together with 11 spectroscopy and gene expression analysis to understand the nature of R. 12 solanacearum-induced formation of vascular coatings in resistant tomato. 13  We describe that resistant tomato specifically responds to infection by assembling a 14 vascular structural barrier formed by a ligno-suberin coating and tyramine-derived 15 hydroxycinnamic acid amides. Further, we show that overexpressing genes of the 16 ligno-suberin pathway in a commercial susceptible variety of tomato restricts R. 17 solanacearum movement inside the plant and slows disease progression, enhancing 18 resistance to the pathogen. 19  We propose that the induced barrier in resistant plants does not only restrict the 20 movement of the pathogen, but may also prevent cell wall degradation by the pathogen 21 and confer anti-microbial properties, effectively contributing to resistance. 22 23 Key words: 24 25 Bacterial wilt, Feruloyltyramine, HCAAs, Lignin, Ralstonia solanacearum, Suberin, 26 Tomato, Vascular coating 3 New Phytologist Page 4 of 59 28 Introduction 29 30 In natural environments plants are constantly exposed to diverse microbiota, including 31 pathogenic organisms. In addition to pre-existing structural cell barriers that act as a first 32 line of defense (Serrano et al., 2014; Falter et al., 2015), pathogen perception results in 33 activation of a complex, multi-layered immune system in plants (Jones and Dangl, 2006). 34 As part of the suite of inducible defenses, de novo formation of physico-chemical barriers 35 prevents pathogen colonization and spread inside the plant. Despite their importance, the 36 exact composition of these barriers, as well as the mechanisms that lead to their formation 37 in the plant upon pathogen invasion remain largely unknown. 38 39 The interaction between the soil-borne bacterial wilt pathogen Ralstonia solanacearum and 40 tomato offers a paradigmatic scenario to study inducible physico-chemical barriers, because 41 of its agro-economic impact, and the well-developed genetic and molecular tools available 42 in both organisms. R. solanacearum enters the root system through wounds or at the points 43 of emergence of lateral roots, where the epidermal and endodermal barriers may be 44 compromised (Vasse et al., 1995; Álvarez et al., 2010; Ursache et al., 2021). After entering 45 the root, the bacterium moves centripetally towards the vasculature and once it reaches the 46 xylem, it multiplies and spreads vertically within the vessels and horizontally to other 47 vessels and the surrounding tissues (Digonnet et al., 2012). 48 49 The xylem tissue is a major battleground for the interaction between vascular wilt 50 pathogens and their hosts, where the outcome of the infection is at stake (Yadeta and 51 Thomma, 2013). To prevent the spread of pathogenic propagules, the xylem vasculature of 52 resistant plants undergoes intense structural and metabolic modifications, such as 53 reinforcing the walls of xylem vessels, pit membranes and surrounding xylem parenchyma 54 cells in response to pathogens (Street et al., 1986; Benhamou, 1995). This prevents 55 pathogen colonization of the surrounding parenchyma cells, nearby vessels and inter- 56 cellular spaces through degeneration of the vessel pit membranes or cell walls by the 57 pathogen (Nakaho et al., 2000; Digonnet et al., 2012; Liu et al., 2005; Pérez-Donoso et al., 58 2010; Lowe-Power et al., 2018). This vascular confinement is an effective strategy 4 Page 5 of 59 New Phytologist 59 commonly found among plants resistant to vascular wilt pathogens such as R. 60 solanacearum, which otherwise spread systemically and eventually kill the plant 61 (McGarvey et al., 1999; Vasse et al., 2000; Potter et al., 2011; Caldwell et al., 2017; 62 Scortichini, 2020; Kashyap et al., 2021). 63 64 Among the various tomato germplasms, the cultivar Hawaii 7996 (H7996) is the most 65 effective natural source of resistance against R. solanacearum (Nakaho et al., 2004; 66 Grimault et al., 1994). In this cultivar, resistance to R. solanacearum is a complex 67 polygenic trait (2000; Thoquet et al., 1996; Mangin et al., 1999; Wang et al., 2013). Our 68 previous study identified the bottlenecks through which H7996 is able to limit R. 69 solanacearum spread in planta (Planas-Marquès et al., 2019), namely: i) root colonization, 70 ii) vertical movement from roots to shoots, iii) circular invasion of the vascular bundle and 71 iv) radial apoplastic spread from the vessels into the cortex. Vascular cell wall 72 reinforcements seem to play a key role in confining R. solanacearum into the xylem 73 vascular bundles of resistant tomato H7996. Ultra-microscopic studies in resistant tomato 74 showed that the pit membranes, as well as xylem vessel walls and parenchyma cells form a 75 conspicuously thick coating in the form of an electron dense amorphous layer, as part of the 76 defense response against R. solanacearum (Nakaho et al., 2000; Kim et al., 2016;). 77 78 Among the polymers constituting vascular coating structures, lignin is the most typically 79 found, constituting an integral part of the secondary cell wall of the xylem vasculature. 80 Lignin is well documented as a common structural defense against vascular wilt pathogens 81 (Novo et al., 2017; Zeiss et al., 2019; Kashyap et al., 2021) and it is also emerging as an 82 important inducible defense component in other diseases/pests affecting the vasculature 83 (Jhu et al., 2021; Joo et al., 2021) Suberin has also been reported to be deposited in vascular 84 coatings as a defense response (Kashyap et al., 2021), although the mechanisms regulating 85 its synthesis, spatio-temporal dynamics and inducibility remain elusive. Interestingly, root 86 microbiota has been recently shown to shape suberin deposits in the plant, highlighting its 87 central role in plant-microbe interactions (Salas-González et al., 2021). Suberin is a 88 polyester containing long and very long chain fatty acids and derivatives and also some 89 aromatics, mainly ferulic acid. Cells that accumulate suberin also accumulate lignin, whose 5 New Phytologist Page 6 of 59 90 deposition has been described to precede that of suberin in phellem cells (Lulai and Corsini, 91 1998). This lignin is also known as a lignin-like polymer, which consists of 92 hydroxycinnamates and monolignols (Graça, 2015). The ligno-suberin heteropolymer 93 formed by the lignin-like polymer and suberin has been also referred to as the 94 poly(aromatic) and poly(aliphatic) domains of suberin, respectively. Commonly, suberized 95 cell walls also comprise soluble phenolic compounds, which share biosynthetic pathways 96 with suberin and lignin (Bernards, 2002). 97 98 Ferulic acid present in the suberin and lignin-like fractions is proposed to link both 99 polymers (Graça, 2010) and its continuous production has been demonstrated essential for 100 suberin deposition (Andersen et al., 2021). Ferulic acid amides, such as feruloyltyramine 101 and feruloyloctopamine, have been described as structural components of the lignin-like 102 polymer and in the phenolic soluble fraction of suberizing wound-healing potato tuber 103 (Negrel et al., 1996; Razem and Bernards, 2002). Ferulic acid amides belong to the 104 Hydroxycinnamic acid amide (HCAA) family, which present antimicrobial activity and are 105 considered biomarkers during plant-pathogen interactions (Zeiss et al., 2021). However, the 106 precise role of HCAAs in plant defense remains to be elucidated (Macoy et al., 2015). 107 Besides their direct antimicrobial activity as soluble phenols, HCAAs have also been 108 proposed to cross-link to cell wall structural polymers during infection, potentially 109 contributing towards the formation of a phenolic barrier that can make the cell wall resilient 110 to pathogenic degradation (Zeiss et al., 2021). 111 112 In the present study, we conducted a detailed investigation of the inducibility, structure and 113 composition of the xylem vascular wall reinforcements that restrict R. solanacerum 114 colonization in resistant tomato. Using a combination of histological and live-imaging 115 techniques, together with spectroscopy and gene expression analysis we provide important 116 new insights into the pathogen-induced formation of vascular coatings. In particular, we 117 show that a ligno-suberin vascular coating and tyramine-derived HCAAs contribute to 118 restriction of R. solanacearum in resistant tomato. In addition, we demonstrate that genes in 119 the ligno-suberin-associated pathways can be explored to engineer resistance against R. 120 solanacearum into commercial susceptible varieties of tomato. 6 Page 7 of 59 New Phytologist 121 122 Materials and Methods 123 Plant materials and growth conditions 124 Tomato (Solanum lycopersicum) varieties Marmande, Hawaii 7996 (H7996) and 125 Moneymaker (wild-type and 35S::THT 1-3, generated by Campos et al., (2014)), were 126 used. Plants were grown in controlled growth chambers at 60% humidity, 12 h day/night 127 and 27ºC (LED lighting) or 25ºC (fluorescent lighting). 128 129 Ralstonia solanacearum strains and growth conditions 130 R. solanacearum GMI1000 strain (Phylotype I, race 1 biovar 3) was used, including 131 luminescent and fluorescent reporter strains of R. solanacearum GMI1000 described in 132 (Cruz et al., 2014; Planas-Marquès et al., 2019). 133 134 Cloning and stable transformation in tomato 135 136 For generation of the 35S::FHT-HA construct the FHT (Solyc03g097500) coding sequence 137 was amplified from tomato H7996 cDNA using the forward and reverse primers 138 part7FHTF1 and part7FHTHAR1, respectively (Table S1). The amplified product was 139 cloned into the pJET1.2/blunt cloning vector using CloneJet PCR cloning kit 140 (Thermofisher) and then digested by SmaI and BamHI. The digested products were purified 141 using NZYGelpure (Nzytech) followed by ligation into the pART7 and later to pART27 142 vector (Gleave, 1992). pART27 containing 35S::FHT-HA was transformed into Marmande. 143 For this, the construct was transformed into Agrobacterium tumefaciens strain C58C1. 144 Cotyledon explant preparation, selection, and regeneration followed the methods described 145 by (Mazier et al., 2011). Transformants were selected on kanamycin-containing medium. 146 Accumulation of FHT-HA protein was assayed by immunoblot with a monoclonal HA 147 antibody (GenScript). 148 149 Bacterial inoculation in plants 150 Four to five week-old tomato plants were inoculated through roots with R. solanacearum 151 using the soil drenching method with a to 1x107 CFU ml-1 suspension of bacteria as 7 New Phytologist Page 8 of 59 152 described in (Planas-Marquès et al., 2018). Inoculated plants were kept in a growth 153 chamber at 27°C. For tomato leaf infiltration, plants were vacuum-infiltrated by 154 submerging the whole aerial part in a ~105 CFU ml-1 bacterial suspension as described in 155 Planas-Marquès et al., (2018). For inoculation directly onto the stem vasculature, 10 µl (5 156 µl at a time) of 105 CFU ml-1 bacterial suspension was placed at the node of the petiole and 157 pin-inoculated using a sterile needle (30G×½″, BD Microlance, Becton Dickinson). 158 159 R. solanacearum pathogenicity assays and quantification of bacterial growth in planta 160 Infected plants were scored for wilting symptoms using a scale from 0 to 4: 0=healthy plant 161 with no wilt, 1=25%, 2=50%, 3=75%, and 4=100% of the canopy wilted as described by 162 Planas-Marquès et al., (2019). The relative light units per second (RLU·s–1) readings were 163 converted to CFU·g–1 tissue as described in Planas-Marquès et al., (2019). For bacterial 164 colonization assays using GFP reporter strain, transverse stem cross-sections were made at 165 the inoculation point and at a distance of 0.5 cm, 1 cm, 2 cm and 3 cm in both upward and 166 downward direction, using a sterile razor blade. Quantification of mean green fluorescence 167 was done using ImageJ software (Planas-Marquès et al., 2019). For leaf in planta 168 multiplication assays, 3 leaf discs of 0.8 cm2 were homogenized in 200 µl of sterile distilled 169 water. CFU cm-2 leaf tissue were calculated after dilution plating of samples with 170 appropriate selection antibiotics and CFU counting 24 hours later. 171 172 Histological methods 173 Thin (approximately 150 µm) transverse cross-sections were obtained with a sterile razor 174 blade from a 1.5 cm area of the taproot-to-hypocotyl transition zone located immediately 175 below the soil line (Fig. S1a). Inoculated plants were sectioned when bacterial colonization 176 level reached 105 CFU g-1 taproot-to-hypocotyl transition zone tissue. This corresponded to 177 4 dpi in Marmande and 9 dpi in H7996, at which stage only H7996 sections showed a 178 localized browning at one xylem pole indicative of infection. Sections were kept in 70 % 179 ethanol at room temperature for 5 days and examined using fluorescence microscopy using 180 a Leica DM6B-Z microscope under UV illumination (340-380 nm excitation and 410-450 181 nm barrier filters). Autofluorescence emitted from phenolic deposits was recorded using a 182 Leica-DFC9000GT-VSC07341 camera and the signal was pseudo-colored green. Sections 8 Page 9 of 59 New Phytologist 183 were also stained with Phloroglucinol-HCl for the detection of lignin and observed under 184 bright field (Pomar et al., 2004). Photographs were taken with a DP71 Olympus digital 185 camera. Cross-sections were also observed under UV with a Leica-DM6B-Z microscope 186 (340-380 nm excitation and 410-450 nm barrier filters). To detect the autofluoresecent 187 blue-to-green pH-dependent color conversion of wall-bound ferulic acid cross-sections 188 were first mounted in 70% ethanol (neutral pH) and then in 1N KOH (pH above 10) 189 adapting the protocol from (Harris and Trethewey, 2010; Carnachan and Harris, 2000; 190 Donaldson and Williams, 2018). A Leica DM6B-Z microscope was used to observe 191 autofluorescence (340-380 nm excitation and 410-450 nm barrier filters). Images were 192 recorded using a Leica MC190-HD-0518131623 digital camera. To visualize suberin 193 aliphatics, sections were treated with 5 % Sudan IV, dissolved in 70 % ethanol and 194 illuminated with UV light. These sections were subsequently treated with 1N KOH to 195 detect ferulic acid as described above. For both ferulic acid and suberin, the HC PL APO or 196 HC PL FLUOTAR objectives of the Leica DM6B-Z microscope were used and images 197 were captured using a Leica MC190-HD-0518131623 color digital camera. The UV 198 autofluorescence signal from xylem vessel walls and surrounding layers was measured 199 using the LAS X Leica software and changes in ferulate accumulation were quantified 200 using ImageJ software by selecting the area of xylem vessel walls showing autoflorescence. 201 Quantifications of fluorescence intensity were normalized per µm of region of interest 202 (ROI), which corresponded to a particular area of the vascular bundles, where main vessels 203 concentrate (represented in Fig. 1b). Basically the line is drawn at each of the 4 corners in 204 the whole image and then the fluorescence is normalized by the length of the lines. 205 206 2D-NMR 207 The samples of a pool of 15 tomato plants (taproot-to-hypocotyl region), water treated or 208 having a bacterial load of 105 CFU.g-1 were milled and extracted sequentially with water, 209 80% ethanol, and with acetone, by sonicating in an ultrasonic bath during 30 min each time, 210 centrifuging and eliminating the supernatant. Then, lignin/suberin fraction was 211 enzymatically isolated by hydrolyzing the carbohydrates fraction with Cellulysin 212 (Calbiochem), as previously described (Rico et al. 2014). Approximately 20 mg of 213 enzymatic lignin/suberin preparation was dissolved in 0.6 mL of DMSO-d6. Heteronuclear 9 New Phytologist Page 10 of 59 214 single quantum coherence (HSQC) spectra were acquired on a Bruker AVANCE III 500 215 MHz spectrometer equipped with a 5 mm TCI cryoprobe, using the experimental 216 conditions previously described (Rico et al., 2014). HSQC cross-signals were assigned and 217 quantified as described (Rencoret et al., 2018; del Río et al., 2018; Mahmoud et al., 2020). 218 In the aromatic region, the correlation signals of G2 and S2,6 were used to estimate the 219 content of the respective G- and S-lignin units. The Cα/Hα signals of the β–O–4′ ethers (Aα), 220 phenylcoumarans (Bα), and resinols (Cα) in the linkages region were used to estimate their 221 relative abundances, whereas the Cγ/Hγ signal was used in the case of cinnamyl alcohol 222 end-units (Iγ). 223 224 RNA extraction, cDNA synthesis and quantitative RT-PCR analysis 225 Taproot-to-hypocotyl transition zone sections of ~ 0.5 mm thickness were obtained and the 226 xylem vascular tissues (vascular bundles and surrounding parenchyma cells) were collected 227 and kept in liquid nitrogen. Each sample comprised taproot-to-hypocotyl transition zone 228 xylem tissues of 6 plants. RNA was extracted using the Maxwell RSC Plant RNA Kit 229 (Promega). cDNA was synthesized from 2 µg RNA using High Capacity cDNA Reverse 230 Transcription Kit (Applied Biosystems, USA). cDNA amplification and analysis was 231 performed using the LightCycler 480 System (Roche). The Elongation Factor 1 alpha 232 houskeeping gene (eEF1 α, Solyc06g005060) was used as a reference. All reactions were 233 run in triplicate for each biological replicates. Melting curves and relative quantification of 234 target genes were determined using the software LightCycler V1.5 (Roche). The level of 235 expression relative to the reference gene was calculated using the formula 2-ΔCT, where 236 ΔCT = (CT RNA target - CT reference RNA). 237 238 Statistical analysis 239 Statistical analyses were performed using Statgraphics software. All statistical tests are 240 indicated in the respective figure legends. 241 242 Results: 243 10 Page 11 of 59 New Phytologist 244 Resistant H7996 tomato restricts R. solanacearum colonization and induces a vascular 245 coating response involving wall-bound phenolics 246 247 In order to understand the mechanisms underscoring restriction of R. solanacearum spread 248 in resistant tomato varieties we used the resistant variety Hawaii 7996 (H7996) and 249 compared it to the susceptible cultivar Marmande. In our assay conditions, most Marmande 250 plants were wilted 10 days after inoculation with R. solanacearum GMI1000, while H7996 251 plants remained largely asymptomatic (Fig. 1a, S2a and (Planas-Marquès et al., 2019). 252 Accordingly, bacterial loads in the taproot-to-hypocotyl region were drastically reduced in 253 H7996 compared to Marmande, confirming the remarkable bacterial restriction ability of 254 this cultivar (Fig. S1b and (Planas-Marquès et al., 2019)). 255 256 To identify defense-associated anatomical and physico-chemical modifications in H7996 257 after infection with R. solanacearum compared to Marmande we performed histochemical, 258 spectroscopic and gene expression analysis. For this, plants were infected with a 107 R. 259 solanacearum solution or mock through their roots using the soil-drench method and then 260 we collected tissue containing 105 CFU g-1 tissue of bacteria at the taproot-to-hypocotyl 261 transition area, located approximately 1 cm below ground (Fig. S1a). Marmande reached 262 105 CFU g-1 tissue at around 4 dpi, while the resistant H7996 took approximately 9 days to 263 do so (Fig. S2). We have previously observed that the root-to-hypocotyl area constitutes a 264 key bottleneck for bacterial progression inside the plant (Zuluaga et al., 2015; Puigvert et 265 al., 2017; Planas-Marquès et al., 2019), being thus an ideal target zone for analysis of 266 structural defense responses. 267 268 We first analyzed ultraviolet (UV) autofluorescence of transverse cross-sections of the 269 taproot-to-hypocotyl region, indicative of phenolic compounds (Donaldson, 2020). To 270 focus on cell wall-deposited phenolic compounds, soluble phenolic compounds were 271 removed with ethanol prior to observation as reported (Pouzoulet et al., 2013; Araujo et al., 272 2014). Infection with R. solanacearum induced a strong UV signal emitted from the walls 273 of the vessels, and also from surrounding xylem parenchyma cells and tracheids in resistant 274 H7996 (Figs. 1b, c). This enhanced autofluorescence was not observed in the susceptible 11 New Phytologist Page 12 of 59 275 variety Marmande nor in mock-treated samples (Fig. 1b, c). In tissues outside the vascular 276 area, inoculation resulted in a decrease of autofluorescence in both susceptible and resistant 277 tomato lines. 278 279 Spectroscopic analysis reveals R. solanacearum-induced deposition of suberin and 280 accumulation of tyramine-derived amides in resistant H7996 tomato and lignin 281 structural modifications in susceptible Marmande tomato 282 283 In order to decipher the composition of the cell wall-deposited compounds we used nuclear 284 magnetic resonance (2D-HSQ NMR), one of the most powerful tools for plant cell wall 285 structural analysis providing information on the composition and linkages in lignin/suberin 286 polymers (Ralph and Landucci, 2010; Correia et al., 2020). 2D-HSQC spectra of infected 287 or mock-treated taproot-to-hypocotyl transition zones of H7996 and Marmande tomato 288 plants were obtained and the main lignin and suberin substructures identified are shown in 289 Fig. 2, while the chemical shifts of the assigned cross-signals are detailed in Table S2. 290 Importantly, the aliphatic region of the 2D-HSQC spectra revealed that H7996 infected 291 plants were more enriched in poly-aliphatic structures characteristic of suberin (magenta- 292 colored signals), compared to its mock control (Fig 2a). Related to this, an olefinic cross- 293 signal of unsaturated fatty acid structures (UF, δC/δH 129.4/5.31), typical of suberin, was 294 also found to be increased in the HSQC spectrum of the infected H7996 tomato. A rough 295 estimate based on the integration of lignin and suberin HSQC signals, revealed that the 296 suberin/lignin ratio in R. solanacearum-infected H7996 plants was doubled compared to 297 mock-treated plants, evidencing an increase in suberin deposition as a consequence of the 298 bacterial infection. Interestingly, signals compatible with feruloylamides (FAm7; δC/δH 299 138.6/7.31) and with tyramine-derived amides (Ty in orange; δC/δH 129.3/6.92, 114.8/6.64, 300 40.5/3.29 and 34.2/2.62) were exclusively found in the spectrum of infected H7996 plants, 301 suggesting the presence of feruloyltyramine exclusively in these samples (Fig. 2a). Since 302 tyramines have been found as structural components co-ocurring with suberin (Bernards et 303 al., 1995; Bernards and Lewis, 1998), which generates physically and chemically resistant 304 barriers (He and Ding, 2020), our results substantiate the hypothesis of suberin as an 305 important defense element against R. solanacearum infection in resistant tomato plants. On 12 Page 13 of 59 New Phytologist 306 the contrary, the 2D-HSQC spectra from the Marmande variety did not display notable 307 variations between mock and infected plants in the signals corresponding to suberin, 308 tyramine-related structures nor feruloylamides (Fig 2a). 309 310 Interestingly, 2D-HSQC NMR spectra also revealed significant structural modifications in 311 the composition of lignin and the distribution of linkages in tomato plants after infection. 312 Lignins with lower S/G ratios are more branched (condensed) and recalcitrant towards 313 pathogen attack (Iiyama et al., 2020). Therefore, lignin in inoculated H7996, with an S/G 314 ratio of 1.0 should be, a priori, more resistant than the lignin in inoculated Marmande plants 315 (S/G ratio of 1.5). 2D-HSQC analysis revealed that the infection of susceptible Marmande 316 plants resulted in an increase of the S/G ratio and a clear reduction of all major lignin 317 linkages (β–O–4´, β–5´ and β–β´; reduction in roughly 9%, 43% and 46%, respectively), 318 evidencing that a lignin depolymerization process took place (Fig. 2a). In contrast, infected 319 H7996 tomato displayed a slight decrease of the S/G ratio, and only β–O–4´ linkages (the 320 easiest to degrade in the lignin polymer) were significantly reduced, while the β–5´ and β– 321 β´ were not so affected as in the case of Marmande plants. In this context, the major 322 reduction in lignin linkages observed in Marmande after infection could explain, at least in 323 part, its higher susceptibility to the pathogen. 324 325 Histochemical analysis reveals the formation of structural vascular coatings 326 containing suberin and ferulate/feruloylamide in resistant H7996 tomato in response 327 to R. solanacearum infection 328 329 To confirm our spectroscopic data, we histochemically analyzed taproot-to-hypocotyl 330 transition zone samples of mock and infected H7996 and Marmande tomato plants. 331 Observation of Phloroglucinol-HCl stained sections under brightfield microscopy (Wiesner 332 staining) (Pomar et al., 2002; Pradhan Mitra and Loqué, 2014), showed that mock and 333 infected H9776 (resistant) as well as mock Marmande (susceptible) samples showed a red- 334 purple color characteristic of the reaction of phloroglucinol-HCl in vessels and fibers, 335 indicative of lignin (Fig. 3a). In contrast, infected Marmande sections exhibited reduced 336 phloroglucinol-HCl staining, suggesting a change in composition of xylem lignin upon 13 New Phytologist Page 14 of 59 337 infection (Fig. 3a). This observation is in agreement with the structural changes specifically 338 detected in the lignin structure of infected Marmande plants by 2D-HSQC NMR (Fig. 2a), 339 which suggest lignin depolymerization and may partly underscore the high susceptibility of 340 this tomato variety. 341 342 UV illumination of phloroglucinol-HCl-stained samples allows quenching the 343 autofluorescence from lignin and hence detect residual cell wall autofluorescence, which 344 has been associated with suberin deposits (Baayen and Elgersma, 1985; Rioux et al., 1998; 345 Pouzoulet et al., 2013). Under these conditions, the increased autofluorescence observed in 346 the vascular coating regions of infected H7996 tomato plants was not quenched in 347 phloroglucinol-HCl stained samples (Fig. 3a, b). A more detailed observation revealed that 348 this non-quenched autofluorescence was localized in specific regions compatible with (i) 349 intervessel and vessel-parenchyma pit membranes or pit chamber walls and (ii) parenchyma 350 coatings with fluorescent signals enriched in intracellular spaces (Fig 3c). 351 352 Next, we analyzed whether the pathogen-induced coating of vessels observed in H7996 353 correlated also with an increase in ferulates, a major suberin component. We performed 354 KOH treatment of plant tissues, which specifically shifts the UV fluorescence of 355 ferulate/feruloylamide to green, allowing its detection (Carnachan and Harris, 2000; Harris 356 and Trethewey, 2010; Donaldson and Williams, 2018). UV autofluorescence of vascular 357 coatings in response to R. solanacearum infection in resistant H7996 shifted from blue to a 358 strong green color upon treatment with alkali (1N KOH) (Fig. S3a). This indicated that the 359 R. solanacearum-induced xylem vasculature feruloylation was specific to resistant H7996, 360 as the fainter blue autofluorescence observed in mock-treated resistant H7996 or 361 susceptible Marmande tissues did not change to green at high pH in either early (Fig. S3a, 362 b) or late (Fig. S3c) stages of infection. 363 364 To corroborate that the ferulate/feruloylamide accumulation in infected H7996 tomato was 365 related with vascular suberization, we combined the ferulate-specific UV-alkali treatment 366 described above with Sudan IV staining, which binds aliphatic components of suberin to 367 produce a reddish-brown coloration. This revealed suberization in the taproot-to-hypocotyl 14 Page 15 of 59 New Phytologist 368 area of R. solanacearum-infected H7996 plants, xylem vessel walls as well as the layers of 369 vessels, parenchyma cells and tracheids in the immediate vicinity (Fig. 4). In the periphery 370 of suberized cells, a green signal from UV-alkali was observed (Fig. 4), which may indicate 371 ferulate/feruloylamide deposition indicative of a preceding stage towards suberization in 372 this cell layer. In comparison, no positive Sudan IV or UV-alkali staining was detected in 373 infected Marmande or mock-treated tomato plants. Together, suberized and feruloylated 374 layers of parenchyma cells, vessels and tracheids might form a “suberization zone” creating 375 a strong physico-chemical barrier to limit R. solanacearum spread from the colonized 376 xylem vessel lumen. 377 378 R. solanacearum infection activates the biosynthesis of aliphatic suberin precursors 379 and feruloylamide, and aliphatic esterification of ferulic acid in the vasculature of 380 resistant H7996 381 382 Since a differential accumulation of suberin-compatible compounds was specifically 383 observed in infected H7996, we surmised that genes related to suberin and feruloylamide 384 synthesis, as well as ferulic acid esterification to aliphatics may be upregulated in resistant 385 tomato in response to R. solanacearum invasion. To test this hypothesis, we analyzed: i) 386 expression of genes in the phenylpropanoid and suberin biosynthesis pathways, which 387 provide the necessary precursors for the ligno-suberin heteropolymer; ii) the feruloyl 388 transferase FHT (ASFT/HHT in Arabidopsis), which is involved in the formation of 389 ferulate esters of fatty acyl compounds necessary to form suberin and soluble waxes 390 (Molina et al., 2009; Gou et al., 2009; Serra et al., 2010); and iii) N-hydroxycinnamoyl 391 transferases (THT), which are involved in the synthesis of HCAAs such as 392 feruloyltyramine, which is found on the lignin-like polymer and in the soluble phenolic 393 fraction of some suberized tissues (Negrel et al., 1993; Schmidt et al., 1999) . 394 395 Quantitative RT-PCR from xylem vascular tissue obtained from the taproot-to-hypocotyl 396 zone in R. solanacearum- or mock-treated H7996 and Marmande plants showed specific 397 upregulation of all genes analyzed from the suberin biosynthetic pathway in H7996 infected 398 plants compared to the mock controls (Fig. 5, S4). These included essential suberin 15 New Phytologist Page 16 of 59 399 biosynthesis genes such as CYP86A1 and CYP86B1 (fatty acid oxidation), FAR (primary 400 alcohol generation), KCSs (fatty acid elongases) and GPAT5 (acylglycerol formation). In 401 addition, feruloyl transferase FHT (ASFT/HHT in Arabidopsis), was also strongly 402 upregulated in infected H7996 plants (Fig. 5 and S5b). Regarding THT, in tomato we 403 identified five putative homologs (Fig S6a), all induced by infection in the vascular tissue 404 of H7996 (Fig. 5 and S6b). Among them, SlTHT1-3 showed the strongest upregulation in 405 H7996 after infection (Fig. 5 and S6b). In comparison, R. solanacearum infection had only 406 a modest effect on genes related to phenylpropanoid pathway as only upregulation was 407 detected in the first enzyme of the pathway (PAL) (Fig. 5 and S7). 408 409 Together, these data indicate that upregulation of genes involved in the formation of 410 aliphatic suberin precursors, ferulic acid esterification to aliphatics (FHT) and production of 411 HCAAs, such as feruloyltyramine (THT), constitute a very specific response of H7996 412 plants that takes place in the vasculature upon R. solanacearum infection. Further, these 413 data are in agreement with NMR data of infected H7996, which showed a specific increase 414 in insoluble fatty acid structures typical of suberin as well as the appearance of signals from 415 structural tyramine-derived amides and feruloylamides (Fig. 2a). 416 417 Overexpression of SlTHT1-3 in a susceptible tomato cultivar confers resistance to R. 418 solanacearum 419 420 Based on our results, we set to determine whether overexpressing genes involved in ferulic 421 acid esterification to suberin aliphatics and feruloylamide biosynthesis, such as SlFHT and 422 SlTHT1-3, respectively, would increase resistance against R. solanacearum in a susceptible 423 tomato background. First, we obtained transgenic tomato lines stably overexpressing 424 SlFHT on a susceptible Marmande background (Fig. S8). Under normal growth conditions 425 these lines are morphologically undistinguishable from wild-type, although they display a 426 subtle increase in fresh weight (Fig. S9). We analyzed symptom progression and bacterial 427 colonization. SlFHT overexpression lines showed a slight delay in disease progression (Fig. 428 6a) and moderately milder symptoms. The taproot-to-hypocotyl of SlFHT overexpressors 16 Page 17 of 59 New Phytologist 429 displayed a slight reduction in bacterial loads after soil-soak inoculation in comparison to 430 Wt tomato (Fig. 6b). 431 432 Regarding SlTHT1-3, the corresponding tomato overexpressing line was readily available 433 on a Moneymaker background (Campos et al., 2014). This line has been shown to 434 overaccumulate soluble HCAA such as feruloyltyramine and also the hormone salicylic 435 acid (SA) upon infection with Pseudomonas syringae pv. tomato (Campos et al., 2014). It 436 is worth noting that tomato plants overexpressing SlTHT1-3 display a slight decrease in 437 fresh weight compared to wild-type plants, although with the naked eye they appear 438 undistinguishable (Fig. S10). As expected, the Moneymaker tomato cultivar showed similar 439 susceptibility to R. solanacearum as Marmande (Fig. 7a, b). In contrast, overexpression of 440 SlTHT1-3 resulted in a dramatic increase of resistance against R. solanacearum, with 441 disease progressing remarkably slower in this line compared to wild type (Fig. 7a, b). 442 Importantly, bacterial loads were significantly lower in the taproot-to-hypocotyl and 443 hypocotyl of the SlTHT1-3 overexpressor after soil inoculation in comparison to Wt tomato 444 (Fig. 7c). Similarly, direct leaf inoculation also showed severe bacterial growth restriction 445 in the THT1-3 overexpressing line (Fig. S11a). Further, we monitored the colonization 446 patterns of a R. solanacearum GFP reporter strain after stem inoculation of the SlTHT1-3 447 overexpressing line compared to Wt. In transverse stem cross-sections of 6 dpi plants, 448 bacteria stayed confined near the inoculation point in the 35S::SlTHT1-3 line whereas they 449 spread unrestrictedly in susceptible wild type stems from the inoculation point and at least 3 450 cm up and downwards (Fig. 7d, e and S11b). 451 452 Concomitant with the observed restriction of R. solanacearum colonization, an increase in 453 autofluorescence around the vasculature was observed in the SlTHT1-3 overexpressor (Fig. 454 8a). At similar bacterial loads, Wt did not display such enhanced vascular fluorescence. 455 Phloroglucinol-HCl staining did not quench the paravascular autofluorescence in SlTHT1-3 456 (Fig. 8a, d), indicating that similar to what was previously observed for H7996, the 457 observed increase in wall-bound phenolic deposits did not only correspond to lignin. To 458 gain a deeper insight into the composition of the R. solanacearum-induced vascular 459 deposits in SlTHT1-3 overexpressing plants we performed combined Sudan IV-alkali 17 New Phytologist Page 18 of 59 460 staining (Fig. 8b, c, e). Treatment with alkali resulted in a clear blue-to-green shift of UV 461 autofluorescence around xylem vessels occurring specifically in the SlTHT1-3 462 overexpressor upon infection, which reveals the presence of ferulates/feruloylamides as 463 part of the observed vascular deposits. In contrast, no positive Sudan IV staining was 464 detected, indicating that a canonical suberin polyester does not seem to be part of vascular 465 coatings in SlTHT1-3 overexpressing plants. Since Sudan IV only stains specific moieties 466 of the complex suberin heteropolymer, we cannot rule out that the observed vascular 467 deposits in SlTHT1-3 are formed by a non-canonical ligno-suberin heteropolymer that does 468 not react with Sudan IV. Further investigation will be needed in order to ascertain the exact 469 nature of the R. solanacearum-induced vascular deposits in the SlTHT1-3 overexpressor. In 470 conclusion, our data clearly show that StTHT1-3 ectopic expression provides a very 471 effective resistance mechanism against R. solanacearum -potentially mediated by 472 accumulation of elevated amounts of HCAAs such as feruloyltyramine-, which drastically 473 restricts vascular colonization, preventing bacterial spread and blocking the onset of 474 disease. 475 476 Discussion 477 Ligno-suberin deposits in vascular cell walls and feruloyltyramine accumulation acts 478 as a resistance mechanism restricting R. solanacearum colonization in resistant tomato 479 480 In our study, resistant tomato (H7996) was observed to react aggressively to R. 481 solanacearum infection by reinforcing the walls of vessels and the surrounding parenchyma 482 cells with UV autofluorescent phenolic deposits (Fig. 1). An increase in autofluorescence 483 had been previously reported in another resistant tomato variety, LS-89, although its 484 composition was not precisely defined (Ishihara et al., 2012). Histochemical analysis of 485 vascular coatings in resistant tomato upon R. solanacearum infection showed that suberin- 486 associated autofluorescence was prominent in the vasculature, in line with previous reports 487 using TEM that showed thickening of the pit membranes accumulating electron-dense 488 material in tomato plants resistant to R. solanacearum (Nahako et al., 2000 and 2004). The 489 suberin nature of these coatings was further supported by the positive Sudan IV staining of 490 vessels and surrounding parenchyma cells of H7996 taproot-to-hypocotyl transition zone 18 Page 19 of 59 New Phytologist 491 upon infection (Fig. 4). These results are in agreement with coatings detected in tomato 492 plants resistant to Verticillium albo-atrum (Robb et al., 1991), where suberin and lignin 493 were both deposited in intercellular spaces between parenchyma cells adjoining a xylem 494 vessel or infusing and occluding pit membranes coatings (Robb et al., 1991; Street et al., 495 1996). Besides, inhibition of the phenylpropanoid pathway inhibited the formation of both 496 lignin and suberin coatings (Street et al., 1996), in agreement with the ferulic acid 497 requirement to correctly deposit suberin (Andersen et al., 2021) and reinforcing our 498 observations of the presence of a ferulate/feruloylamide-derived polymer detected in H7996 499 R. solanacearum (Fig. 4). In line with this, NMR data of resistant H7996 tomato vascular 500 tissue revealed the presence of tyramine-derived amides and feruloylamides incorporated 501 into the cell wall and also an enrichment in poly-aliphatic structures characteristic of 502 suberin (Fig. 2) (Graça, 2015; Legay et al., 2016; Figueiredo et al., 2020). 503 504 Beyond histochemistry and spectroscopic signature detections, further evidence supporting 505 the nature of these ligno-suberin coatings as responsible of the resistance observed in 506 H7796 to R. solanacearum was unequivocally provided transcriptionally using gene 507 markers. Tissues undergoing suberization have to go through a complex reprogramming 508 involving a network of metabolic pathways, in order to produce the precursors of the 509 polymer and their polymerization into the matrix (Lashbrooke et al., 2016). Transcriptional 510 reprogramming associated to suberin biosynthesis was clearly observed in the taproot-to- 511 hypocotyl transition zone vascular tissue of resistant H7996 tomato upon infection with R. 512 solanacearum (Fig. 5). Interestingly, PAL, which showed modest upregulation in resistant 513 H7996, had been previously defined as a rate-limiting enzyme of phenylpropanoid pathway 514 (Faragher and Brohier, 1984; Howles et al., 1996). Considering this, the observed 515 upregulation could provide more tyramine and feruloyl-CoA, which together with the 516 upregulation of THT would be in agreement with the increased presence of 517 feruloyltyramine detected by 2D-HSQC NMR (Figure 2a). 518 519 2D-HSQC NMR also revealed differences in the composition and structure of lignin 520 between resistant and susceptible tomato cultivars after infection (Fig. 2). The amounts and 521 the level of lignin of a particular tissue affect wall strength, degradability and pathogen 19 New Phytologist Page 20 of 59 522 resistance (Cho et al., 2012; Mnich et al., 2020). However, its role in 523 resistance/susceptibility responses is not fully understood. Part of the challenge lies in the 524 fact that its composition seems to be less static than what was previously established. A 525 large variety of lignin-like polymers may co-exist in plants depending on the developmental 526 or environmental context. This becomes particularly relevant in plant-pathogen 527 interactions, where a large variety of compounds linked to lignin differentially accumulate 528 upon infection (Cho et al., 2012; Zeiss et al., 2019). The observed lignin structural 529 differences after infection indicate that i) under basal conditions the two tomato varieties 530 display differences in the composition and structure of lignin and ii) R. solanacearum 531 infection affects very differently the lignin fraction in the two varieties: resistant tomato 532 shows only a slight decrease in the S/G ratio that may be linked to an accumulation of the 533 ligno-suberin heteropolymer, while susceptible Marmande undergoes pronounced 534 depolymerization that correlates with a decrease in ferulate/feruloylamide (Fig. 3A). 535 Although R. solanacearum has not been shown to be able to specifically depolymerize 536 lignin, the pathogen secretes enzymes that can degrade cell wall polysaccharides and could 537 participate in the observed Marmande stem collapse phenotype (Fig. 1A). In resistant 538 H7996, on the other hand, vascular ligno-suberin-containing coatings would allow to create 539 a hydrophobic barrier to prevent enzymes from accessing the cell wall substrates and at the 540 same time create reinforcements, contributing to resistance to the pathogen. The fact that 541 these reinforcements are rich in tyramine/feruloyltyramine, may further reinforce the 542 structural barrier, providing rigidity and hampering cell wall digestibility by the pathogen’s 543 hydrolytic enzymes (Macoy et al., 2015; Zeiss et al., 2020). In addition to that, resistant 544 H7996 may have evolved yet undiscovered mechanisms that directly prevent lignin 545 degradation by the pathogen. 546 547 Overall, our data indicate that vascular coating with wall-bound ligno-suberized 548 compounds may restrict horizontal spread of the bacterium (Fig 1). In comparison, 549 susceptible tomato is either not able to induce such vascular coating upon R. solanacearum 550 infection or induces a very weak response (Figs. 1, 3), potentially predisposing its vascular 551 walls to disruption by the pathogen’s cell wall degrading enzymes. Considering that both 552 varieties seem to possess the metabolic components to build such barriers, the difference in 20 Page 21 of 59 New Phytologist 553 response may be a direct effect of the differential transcriptional activation of the pathway 554 in vascular tissue of H7996 compared to Marmande. The fact that varieties with moderate 555 resistance to R. solanacearum show intermediate restriction of colonization (Planas- 556 Marquès et al., 2019), indicate that the formation of these barriers may be a quantitative 557 trait. However, this also opens the possibility that the differential transcriptional activation 558 of the ligno-suberin pathway observed in resistant tomato may have evolved as an effective 559 mechanism to execute defense responses triggered by activation of an immune receptor 560 upon R. solanacearum recognition. Very few immune receptors involved in perception of 561 vascular wilt pathogens have been identified so far, and the mechanisms involved in 562 translating this recognition into effective defense responses in the vasculature remain vastly 563 unknown. Considering that the xylem is a dead tissue, it is expected that the surrounding 564 parenchyma cells will have a pivotal role in perception of the pathogen as well as the 565 signaling leading to the synthesis and wall-binding of the metabolites involved in vascular 566 coating structures, such as the one described here. In fact, xylem parenchyma cells have 567 been shown to synthesize vascular coating components in response to the wilt pathogen V. 568 albo-atrum (Street et al., 1986). However, how suberin is synthesized and deposited in the 569 xylem is still poorly defined. Exciting research currently ongoing in this area will certainly 570 help understanding the origin and transport of ligno-suberin components to form inducible 571 vascular deposits in response to pathogens. This will also help determining the exact point 572 of perception of the pathogen (at a cell type or tissular level). Identification of pathogen- 573 inducible pathways specifically occurring in resistant varieties such as the one presented 574 here open new avenues of research to shed light on this biologically and agronomically 575 relevant question. 576 577 Based on the above observations, we propose the following model (Fig. 9). When reaching 578 the xylem vessels of resistant H7996, R. solanacearum multiplies and tries to invade the 579 surrounding healthy vessels and parenchyma cells by degradation of the xylem pit 580 membranes and walls. Resistant tomato plants respond to R. solanacearum vascular 581 invasion depositing feruloyltyramine and other HCAA-tyramine derived compounds, and 582 suberin. These deposits would block the pit membrane access and serve as coatings of the 583 vessel walls and parenchyma cells present in the immediate vicinity of colonized vessels, 21 New Phytologist Page 22 of 59 584 compartmentalizing the infection. These ligno-suberized layers together form a “zone of 585 ligno-suberization” creating a strong physico-chemical barrier to limit R. solanacearum 586 spread. 587 588 Engineering tomato resistance against R. solanacearum by inducing the tyramine- 589 HCAA pathway 590 591 Considering the observed accumulation of ligno-suberin and cell wall-linked 592 feruloyltyramine in resistant H7996 tomato in response to R. solanacearum infection, we 593 sought to understand the implications of overexpressing genes involved in the synthesis of 594 these compounds in susceptible tomato cultivars upon R. solanacearum infection. We 595 focused on FHT and THT and because their corresponding transcripts are upregulated in the 596 xylem vasculature of resistant tomato upon R. solanacearum infection (Fig. 5) and they are 597 the enzymes related with the synthesis of suberin ferulates and ether linked 598 feruloyltyramine, respectively. 599 600 SlFHT overexpression had a small effect on the responses of susceptible tomato against R. 601 solanacearum, showing a slight delay in wilting symptoms together with a slight decrease 602 of bacterial loads in the plant (Fig. 6). The fact that increasing the levels of FHT in 603 Marmande does only result in a marginal increase in resistance might be linked to a 604 shortfall of aliphatic precursors in this variety (Fig. 5), which constrain a subsequent 605 increase in suberin synthesis. In contrast, transgenic tomato overexpressing SlTHT1-3 on a 606 susceptible background was highly resistant to R. solanacearum (Fig. 7). Importantly, this 607 transgenic line was previously shown to accumulate elevated amounts of soluble HCAAs 608 such as feruloyltyramine and also SA upon infection with the bacteria Pseudomonas 609 syringae pv. tomato (Pto) and to slightly but significantly restrict bacterial growth (Campos 610 et al., 2014). Since SA does not seem to play a major role in defense responses against R. 611 solanacearum (Hirsch et al., 2002; Hernández-Blanco et al., 2007; Hanemian et al., 2016), 612 accumulation of this hormone in SlTHT1-3 overexpressing line may not be the major 613 underlying cause of the observed increase in resistance in this line. Alternatively, enhanced 614 production of tyramine-derived HCAAs may constitute an important defense strategy 22 Page 23 of 59 New Phytologist 615 against R. solanacearum. Feruloyltyramines exhibit antimicrobial activity (Fattorusso et al., 616 1999; Novo et al., 2017) and that they can be involved in plant priming or an adaptive 617 strategy where plants are in a physiological state with improved defensive capacity (Zeiss 618 et al., 2021). These tyramine-derived HCAAs overproduced in SlTHT1-3 overexpressing 619 plants may interfere with R. solanacearum colonization by becoming incorporated into the 620 vascular and perivascular cell walls, providing a stronger cross-linking and restricting the 621 movement of the pathogen inside the plant (Figure 8) but also partly by remaining soluble 622 and acting as direct antimicrobial agents against the pathogen. R. solanacearum possesses a 623 hydroxycinnamic acid degradation pathway and it has been shown that mutants that cannot 624 degrade hydroxycinnamic acids are less virulent on tomato (Lowe-Power et al., 2015; 625 Zhang et al., 2019), which clearly underscores the importance of HCAAs in the arms race 626 taking place in this pathosystem. Considering that the ligno-suberin pathway and HCAAs 627 are well-conserved across the plant kingdom (Philippe et al., 2020; Kashyap et al., 2021; 628 Zeiss et al., 2021), these findings open the possibility to engineer disease resistance in other 629 R. solanacearum hosts by manipulating these pathways. Interestingly, ligno-suberin 630 deposits and accumulation of HCAAs have been reported in response to drought (Macoy et 631 al., 2015; Zhang et al., 2020). Therefore, engineering these pathways could have a double 632 impact both on biotic and abiotic stress responses, improving plant performance in the field 633 under adverse conditions. 634 635 In conclusion, we have provided evidence of the formation of a “ligno-suberization zone” 636 enriched in ether-linked feruloyltyramine and possibly related amides as an effective 637 strategy to confine R. solanacaerum into infected vessels of resistant tomato plants, 638 preventing horizontal spread of the pathogen into healthy tissues and delaying disease 639 symptoms. Resistance against R. solanacearum can be attained in susceptible tomato 640 background by stably overexpressing THT, potentially contributing In the future, it will be 641 interesting to investigate the contribution of HCAAs and suberin to resistance against the 642 pathogen, the mechanisms whereby R. solanacearum perception leads to the formation of a 643 ligno-suberin coatings around the vasculature in resistant tomato varieties. Increasing the 644 spatio-temporal resolution of the tomato-R. solanacearum interaction will be instrumental 645 to reach a deeper insight into structural resistance mechanisms. Also, since vascular 23 New Phytologist Page 24 of 59 646 confinement has been reported in different plant species as a means of resistance against 647 various vascular wilt pathogens (De Ascensao and Dubery, 2000; Martín et al., 2008; Xu et 648 al., 2011; Sabella et al., 2018), the level of conservation of vascular ligno-suberin 649 deposition and HCAAs as a constituent of vascular coatings and part of a resistance 650 mechanism remains to be determined. 651 652 Acknowledgements 653 The authors would like to thank Gabriel Castrillo and Nico Geldner for inspiring 654 discussions. We also thank Marc Planas-Marquès for helpful comments and María Pilar 655 López Gresa (IBMCP-UPV) for kindly sharing the tomato THT seeds. Research is funded 656 by MCIN/AEI/10.13039/501100011033 (NSC, MV), MCIN/AEI/PID2019-110330GB-C21 657 (MF, OS), through the “Severo Ochoa Programme for Centres of Excellence in R&D” 658 (SEV-2015-0533 and CEX2019-000902-S funded by MCIN/AEI/ 659 10.13039/501100011033). , and by the Spanish National Research Council (CISC) pie- 660 201620E081 (JR, AG). AK is the recipient of a Netaji Subhas - Indian Council of 661 Agricultural Research International Fellowship. This work was also supported by the 662 CERCA Programme/Generalitat de Catalunya. We acknowledge support of the publication 663 fee by the CSIC Open Access Publication Support Initiative through its Unit of Information 664 Resources for Research (URICI). Author Contribution AK designed and performed experiments, interpreted data and wrote the manuscript. ALJ-J performed the experiments required for the second submission of the manuscript. WZ performed experiments. MC performed experiments, interpreted data and reviewed the manuscript. SS conducted preliminary spectroscopy experiments and reviewed the manuscript. JR isolated the lignin/suberin fractions and conducted the 2D-HSQC NMR analysis, including data interpretation. AG isolated the lignin/suberin fractions and conducted the 2D-HSQC NMR analysis, including data interpretation. AL conducted preliminary spectroscopy experiments and reviewed the manuscript. 24 Page 25 of 59 New Phytologist OS conducted histopathology staining experiments, interpreted data and reviewed the manuscript. MF interpreted data and reviewed the manuscript. MV designed experiments, interpreted data and reviewed the manuscript. NSC conceptualized the research, designed experiments, interpreted data and wrote the manuscript. Data Availability The data that support the findings of this study are available from the corresponding author upon reasonable request. References Álvarez, B., Biosca, E.G., and López, M.M. (2010). On the life of Ralstonia solanacearum, a destructive bacterial plant pathogen. In Technology and education topics in applied microbiology and microbial biotechnology, A. Méndez-Vilas, ed (Badajoz: Formatex), pp. 267–279. Andersen, T.G., Molina, D., Kilian, J., Franke, R.B., Ragni, L., and Geldner, N. (2021). Tissue-autonomous phenylpropanoid production is essential for establishment of root barriers. Curr. Biol. 31: 965-977. Araujo, L., Bispo, W.M.S., Cacique, I.S., Moreira, W.R., and Rodrigues, F.A. (2014). Resistance in mango against infection by Ceratocystis fimbriata. Phytopathology 104: 820–833. De Ascensao, A.R.D.C.F. and Dubery, I.A. (2000). Panama disease: cell wall reinforcement in banana roots in response to elicitors from Fusarium oxysporum f. sp. cubense race four. Phytopathology 90: 1173–1180. Baayen, R.P. and Elgersma, D.M. (1985). Colonization and histopathology of susceptible and resistant carnation cultivars infected with Fusarium oxysporum f. sp. dianthi . Netherlands J. Plant Pathol. 91: 119–135. Benhamou, N. (1995). Ultrastructural and cytochemical aspects of the response of eggplant parenchyma cells in direct contact with Verticillium-infected xylem vessels. Physiol. 25 New Phytologist Page 26 of 59 Mol. Plant Pathol. 46: 321–338. Bernards, M., Lopez, M., Zajicek, J., and Lewis, N. (1995). Hydroxycinnamic acid- derived polymers constitute the polyaromatic domain of suberin. J. Biol. Chem. 270: 7382–7386. Bernards, M.A. (2002). Demystifying suberin. Can. J. Bot. 80: 227–240. Bernards, M.A. and Lewis, N.G. (1998). The macromolecular aromatic domain in suberized tissue: a changing paradigm. Phytochemistry 47: 915–933. Biggs, A. (1984). Intracellular suberin: occurrence and detection in tree bark. IAWA Bull. 5: 243–248. Caldwell, D., Kim, B., and Iyer-Pascuzzi, A.S. (2017). Ralstonia solanacearum differentially colonizes roots of resistant and susceptible tomato plants. Phytopathology 107: 528–536. Campos, L., Lisón, P., López-Gresa, M.P., Rodrigo, I., Zacarés, L., Conejero, V., and Bellés, J.M. (2014). Transgenic tomato plants overexpressing tyramine N - hydroxycinnamoyltransferase exhibit elevated hydroxycinnamic acid amide levels and enhanced resistance to Pseudomonas syringae . Mol. Plant-Microbe Interact. 27: 1159–1169. Carnachan, S.M. and Harris, P.J. (2000). Ferulic acid is bound to the primary cell walls of all gymnosperm families. Biochem. Syst. Ecol. 28: 865–879. Cho, K., Kim, Y., Wi, S.J., Seo, J.B., Kwon, J., Chung, J.H., Park, K.Y., and Nam, M.H. (2012). Nontargeted metabolite profiling in compatible pathogen-inoculated tobacco (Nicotiana tabacum L. cv. Wisconsin 38) using UPLC-Q-TOF/MS. J. Agric. Food Chem. 60: 11015–11028. Correia, V.G., Bento, A., Pais, J., Rodrigues, R., Haliński, P., Frydrych, M., Greenhalgh, A., Stepnowski, P., Vollrath, F., King, A.W.T. et al. (2020). The molecular structure and multifunctionality of the cryptic plant polymer suberin. Mater. Today Bio 5: 100039. Cruz, A.P.Z., Ferreira, V., Pianzzola, M.J., Siri, M.I., Coll, N.S., and Valls, M. (2014). A novel, sensitive method to evaluate potato germplasm for bacterial wilt resistance using a luminescent Ralstonia solanacearum reporter strain. Mol. Plant-Microbe Interact. 27: 277–285. 26 Page 27 of 59 New Phytologist Digonnet, C., Martinez, Y., Denancé, N., Chasseray, M., Dabos, P., Ranocha, P., Marco, Y., Jauneau, A., and Goffner, D. (2012). Deciphering the route of Ralstonia solanacearum colonization in Arabidopsis thaliana roots during a compatible interaction: Focus at the plant cell wall. Planta 236: 1419–1431. Donaldson, L. (2020). Autofluorescence in plants. Molecules 25: 2393. Donaldson, L. and Williams, N. (2018). Imaging and spectroscopy of natural fluorophores in pine needles. Plants 7: 10. Dorado, J., Almendros, G., Field, J.A., and Sierra-alvarez, R. (2001). Infrared spectroscopy analysis of hemp (Cannabis sativa) after selective delignification by Bjerkandera sp . at different nitrogen levels. Enzyme Microb. Technol. 28: 550–559. Falter, C., Ellinger, D., Von Hulsen, B., Heim, R., and Voigt, C.A. (2015). Simple preparation of plant epidermal tissue for laser microdissection and downstream quantitative proteome and carbohydrate analysis. Front. Plant Sci. 6: 194. Faragher, J.D. and Brohier, R.L. (1984). Anthocyanin accumulation in apple skin during ripening: regulation by ethylene and phenylalanine ammonia-lyase. Sci. Hortic. 22: 89–96. Fattorusso, E., Lanzotti, V., and Taglialatela-Scafati, O. (1999). Antifungal N- feruloylamides from roots of two Allium species. Plant Biosyst. 133: 199–203. Figueiredo, R., Portilla Llerena, J.P., Kiyota, E., Ferreira, S.S., Cardeli, B.R., de Souza, S.C.R., dos Santos Brito, M., Sodek, L., Cesarino, I., and Mazzafera, P. (2020). The sugarcane ShMYB78 transcription factor activates suberin biosynthesis in Nicotiana benthamiana. Plant Mol. Biol. 104: 411–427. Gleave, A.P. (1992). A versatile binary vector system with a T-DNA organisational structure conducive to efficient integration of cloned DNA into the plant genome. Plant Mol. Biol. 20: 1203–1207. Gou, J.-Y., Yu, X.-H., and Liu, C.-J. (2009). A hydroxycinnamoyltransferase responsible for synthesizing suberin aromatics in Arabidopsis. Proc. Natl. Acad. Sci. 106: 18855– 18860. Graça, J. (2010). Hydroxycinnamates in suberin formation. Phytochem. Rev. 9: 85–91. Graça, J. (2015). Suberin: the biopolyester at the frontier of plants. Front. Chem. 3: 62. Grimault, V., Anais, G., and Prior, P. (1994). Distribution of Pseudomonas 27 New Phytologist Page 28 of 59 solanacearum in the stem tissues of tomato plants with different levels of resistance to bacterial wilt. Plant Pathol. 43: 663–668. Hanemian, M. et al. (2016). Arabidopsis CLAVATA1 and CLAVATA2 receptors contribute to Ralstonia solanacearum pathogenicity through a miR169-dependent pathway. New Phytol. 211: 502–515. Hao, Z. et al. (2014). Loss of arabidopsis GAUT12/IRX8 causes anther indehiscence and leads to reduced G lignin associated with altered matrix polysaccharide deposition. Front. Plant Sci. 5: 357. Harris, P.J. and Trethewey, J.A.K. (2010). The distribution of ester-linked ferulic acid in the cell walls of angiosperms. Phytochem. Rev. 9: 19–33. He, M. and Ding, N. (2020). Plant unsaturated fatty acids : multiple roles in stress response. Front. Plant Sci. 11: 562785. Hernández-Blanco, C. et al. (2007). Impairment of cellulose synthases required for Arabidopsis secondary cell wall formation enhances disease resistance. Plant Cell 19: 890–903. Hirsch, J., Deslandes, L., Feng, D.X., Balagué, C., and Marco, Y. (2002). Delayed symptom development in ein2-1, an Arabidopsis ethylene-insensitive mutant, in response to bacterial wilt caused by Ralstonia solanacearum. Phytopathology 92: 1142–1148. Howles, P.A., Sewalt, V.J.H., Paiva, N.L., Elkind, Y., Bate, N.J., Lamb, C., and Dixon, R.A. (1996). Overexpression of L-phenylalanine ammonia-lyase in transgenic tobacco plants reveals control points for flux into phenylpropanoid biosynthesis. Plant Physiol. 112: 1617–1624. Iiyama, K., Lam, T.B.T., and Stone, B. (2020). Covalent cross-links in the cell wall. Plant Physiol. 104: 315–320. Ishihara, T., Mitsuhara, I., Takahashi, H., and Nakaho, K. (2012). Transcriptome analysis of quantitative resistance-specific response upon Ralstonia solanacearum infection in tomato. PLoS One 7(10): e46763. Jhu, M., Farhi, M., Wang, L., Philbrook, R.N., Belcher, M.S., Nakayama, H., Zumstein, K.S., Rowland, S.D., Ron, M., Shih, P.M., et al. (2021). Lignin-based resistance to Cuscuta campestris parasitism in Heinz resistant tomato cultivars. 28 Page 29 of 59 New Phytologist bioRxiv. https://doi.org/10.1101/706861. Jones, J.D.G. and Dangl, J.L. (2006). The plant immune system. Nature 444: 323–329. Joo, Y., Kim, H., Kang, M., Lee, G., Choung, S., Kaur, H., Oh, S., Choi, J.W., Ralph, J., Baldwin, I.T., et al. (2021). Pith-specific lignification in Nicotiana attenuata as a defense against a stem-boring herbivore. New Phytol. 232: 332–344. Kashyap, A., Planas-marquès, M., Capellades, M., Valls, M, and Coll, N.S. (2021). Blocking intruders : inducible physico-chemical barriers against plant vascular wilt pathogens. J. Exp. Bot. 72: 184–198. Kim, S.G., Hur, O.S., Ro, N.Y., Ko, H.C., Rhee, J.H., Sung, J.S., Ryu, K.Y., Lee, S.Y., and Baek, H.J. (2016). Evaluation of resistance to Ralstonia solanacearum in tomato genetic resources at seedling stage. Plant Pathol. J. 32: 58–64. Kutscha, N.P. and Gray, J.R. (1972). The suitability of certain stains for studying lignification in balsam fir, Abies balsamina (L.) Mill. Tech. Bull. 53: 1–51. Lahlali, R., Song, T., Chu, M., Yu, F., Kumar, S., Karunakaran, C., and Peng, G. (2017). Evaluating changes in cell-wall components associated with clubroot resistance using fourier transform infrared spectroscopy and RT-PCR. International J. Mol. Sci. 18: 2058. Lashbrooke, J., Cohen, H., Levy-Samocha, D., Tzfadia, O., Panizel, I., Zeisler, V., Massalha, H., Stern, A., Trainotti, L., Schreiber, L., et al. (2016). MYB107 and MYB9 homologs regulate suberin deposition in angiosperms. Plant Cell 28: 2097– 2116. Legay, S., Guerriero, G., André, C., Guignard, C., Cocco, E., Charton, S., Boutry, M., Rowland, O., and Hausman, J.F. (2016). MdMyb93 is a regulator of suberin deposition in russeted apple fruit skins. New Phytol. 212: 977–991. Liu, H., Zhang, S., Schell, M.A., and Denny, T.P. (2005). Pyramiding unmarked deletions in Ralstonia solanacearum shows that secreted proteins in addition to plant cell-wall-degrading enzymes contribute to virulence. Mol. Plant-Microbe Interact. 18: 1296–1305. Lopes, M. H., Neto, C. P., Barros, A. S., Rutledge, D., Delgadillo, I., and Gil, A. M. (2000). Quantitation of aliphatic suberin in Quercus suber L. cork by FTIR spectroscopy and solid-state 13C-NMR spectroscopy. Biopolymers 57: 344-351. 29 New Phytologist Page 30 of 59 Lowe-Power, T.M., Ailloud, F., and Allen, C. (2015). Hydroxycinnamic Acid Degradation, a Broadly Conserved Trait, Protects Ralstonia solanacearum from Chemical Plant Defenses and Contributes to Root Colonization and Virulence. Mol. Plant-Microbe Interact. 28: 286–297. Lowe-Power, T.M., Khokhani, D., and Allen, C. (2018). How Ralstonia solanacearum exploits and thrives in the flowing plant xylem environment. Trends Microbiol. 26: 929–942. Lulai, E.C. and Corsini, D.L. (1998). Differential deposition of suberin phenolic and aliphatic domains and their roles in resistance to infection during potato tuber (Solanum tuberosum L.) wound-healing. Physiol. Mol. Plant Pathol. 53: 209–222. Macoy, D.M., Kim, W.Y., Lee, S.Y., and Kim, M.G. (2015). Biosynthesis, physiology, and functions of hydroxycinnamic acid amides in plants. Plant Biotechnol. Rep. 9: 269–278. Macoy, D.M., Kim, W.Y., Lee, S.Y., and Kim, M.G. (2015). Biotic stress related functions of hydroxycinnamic acid amide in plants. J. Plant Biol. 58: 156–163. Mahmoud, A.B., Danton, O., Kaiser, M., Han, S., Moreno, A., Algaffar, S.A., Khalid, S., Oh, W.K., Hamburger, M., and Mäser, P. (2020). Lignans, amides, and saponins from Haplophyllum tuberculatum and their antiprotozoal activity. Molecules 25: 2825. Mangin, B., Thoquet, P., Olivier, J., and Grimsley, N.H. (1999). Temporal and multiple quantitative trait loci analyses of resistance to bacterial wilt in tomato permit the resolution of linked loci. Genetics 151: 1165–1172. Martín, J.A., Solla, A., Coimbra, M.A., and Gil, L. (2005). Metabolic distinction of Ulmus minor xylem tissues after inoculation with Ophiostoma novo-ulmi. Phytochemistry 66: 2458–2467. Martín, J.A., Solla, A., Domingues, M.R., Coimbra, M.A., and Gil, L. (2008). Exogenous phenol increase resistance of Ulmus minor to dutch elm disease through formation of suberin-like compounds on xylem tissues. Environ. Exp. Bot. 64: 97– 104. Mazier, M., Flamain, F., Nicolaï, M., Sarnette, V., and Caranta, C. (2011). Knock- down of both eIF4E1 and eIF4E2 genes confers broad-spectrum resistance against potyviruses in tomato. PLoS One 6(12): e29595. 30 Page 31 of 59 New Phytologist McGarvey, J.A., Denny, T.P., and Schell, M.A. (1999). Spatial-temporal and quantitative analysis of growth and EPS I production by Ralstonia solanacearum in resistant and susceptible tomato cultivars. Phytopathology 89: 1233–1239. Mnich, E. et al. (2020). Phenolic cross-links: building and de-constructing the plant cell wall. Nat. Prod. Rep. 37: 919-961. Molina, I., Li-Beisson, Y., Beisson, F., Ohlrogge, J.B., and Pollard, M. (2009). Identification of an Arabidopsis feruloyl-coenzyme a transferase required for suberin synthesis. Plant Physiol. 151: 1317–1328. Nakaho, K., Hibino, H., and Miyagawa, H. (2000). Possible mechanisms limiting movement of Ralstonia solanacearum in resistant tomato tissues. J. Phytopathol. 148: 181–190. Nakaho, K., Inoue, H., Takayama, T., and Miyagawa, H. (2004). Distribution and multiplication of Ralstonia solanacearum in tomato plants with resistance derived from different origins. J. Gen. Plant Pathol. 70:115–119. Negrel, J., Javelle, F., and Paynot, M. (1993). Wound-induced tyramine hydroxycinnamoyl transferase in Potato (Solanum tuberosum) tuber discs. J. Plant Physiol. 142: 518–524. Negrel, J., Pollet, B., and Lapierre, C. (1996). Ether-linked ferulic acid amides in natural and wound periderms of potato tuber. Phytochemistry 43: 1195–1199. Novaes, E., Kirst, M., Chiang, V., Winter-sederoff, H., and Sederoff, R. (2010). Lignin and biomass: A negative correlation for wood formation and lignin content in trees. Plant Physiol. 154: 555–561. Novo, M., Silvar, C., Merino, F., Martínez-Cortés, T., Lu, F., Ralph, J., and Pomar, F. (2017). Deciphering the role of the phenylpropanoid metabolism in the tolerance of Capsicum annuum L. to Verticillium dahliae Kleb. Plant Sci. 258: 12–20. Pérez-Donoso, A.G., Sun, Q., Caroline Roper, M., Carl Greve, L., Kirkpatrick, B., and Labavitch, J.M. (2010). Cell wall-degrading enzymes enlarge the pore size of intervessel pit membranes in healthy and Xylella fastidiosa-infected grapevines. Plant Physiol. 152: 1748–1759. Planas-Marquès, M., Bernardo-Faura, M., Paulus, J., Kaschani, F., Kaiser, M., Valls, M., Van Der Hoorn, R.A.L., and Coll, N.S. (2018). Protease activities triggered by 31 New Phytologist Page 32 of 59 Ralstonia solanacearum infection in susceptible and tolerant tomato lines. Mol. Cell. Proteomics 17: 1112–1125. Planas-Marquès, M., Kressin, J.P., Kashyap, A., Panthee, D.R., Louws, F.J., Coll, N.S., and Valls, M. (2019). Four bottlenecks restrict colonization and invasion by the pathogen Ralstonia solanacearum in resistant tomato. J. Exp. Bot. 71: 2157–2171. Philippe, G., Sørensen, I., Jiao, C., Sun, X., Fei, Z., Domozych, D.S., and Rose, J.K. (2020). Cutin and suberin: assembly and origins of specialized lipidic cell wall scaffolds. Curr. Opin. Plant Biol. 55: 11–20. Pomar, F., Merino, F., and Barceló, A.R. (2002). O-4-linked coniferyl and sinapyl aldehydes in lignifying cell walls are the main targets of the Wiesner (phloroglucinol- HCl) reaction. Protoplasma 220: 17–28. Pomar, F., Novo, M., Bernal, M.A., Merino, F., Barceló, A.R., and Barceló, A.R. (2004). Changes in stem lignins ( monomer composition and crosslinking ) and peroxidase are related with the maintenance of leaf photosynthetic integrity during Verticillium wilt in Capsicum annuum. New Phytol. 163: 111–123. Potter, C., Harwood, T., Knight, J., and Tomlinson, I. (2011). Learning from history, predicting the future: The UK dutch elm disease outbreak in relation to contemporary tree disease threats. Philos. Trans. R. Soc. B Biol. Sci. 366: 1966–1974. Pouzoulet, J., Jacques, A., Besson, X., Dayde, J., and Mailhac, N. (2013). Histopathological study of response of Vitis vinifera cv. Cabernet Sauvignon to bark and wood injury with and without inoculation by Phaeomoniella chlamydospora. Phytopathol. Mediterr. 52: 313–323. Pradhan Mitra, P. and Loqué, D. (2014). Histochemical staining of Arabidopsis thaliana secondary cell wall elements. J. Vis. Exp.: 87: e51381. Puigvert, M., Guarischi-Sousa, R., Zuluaga, P., Coll, N.S., Macho, A.P., Setubal, J.C., and Valls, M. (2017). Transcriptomes of Ralstonia solanacearum during root colonization of Solanum commersonii. Front. Plant Sci. 8: 370. Ralph, J. and Landucci, L. (2010). NMR of lignins. In Heitner JA, Dimmel C, Scmidt DR eds. Lignin and Lignans: Advances in chemistry. Boca Raton, FL, USA: CRC Press, Taylor & Francis, 137–243.Razem, F.A. and Bernards, M.A. (2002). Hydrogen peroxide is required for poly(phenolic) domain formation during wound-induced 32 Page 33 of 59 New Phytologist suberization. J. Agric. Food Chem. 50: 1009–1015. Rencoret, J., Kim, H., Evaristo, A.B., Gutiérrez, A., Ralph, J., and Del Río, J.C. (2018). Variability in lignin composition and structure in cell walls of different parts of macaúba (Acrocomia aculeata) Palm Fruit. J. Agric. Food Chem. 66: 138–153. Rico, A., Rencoret, J., Del Río, J.C., Martínez, A.T., and Gutiérrez, A. (2014). Pretreatment with laccase and a phenolic mediator degrades lignin and enhances saccharification of Eucalyptus feedstock. Biotechnol. Biofuels 7: 6. del Río, J.C., Rencoret, J., Gutiérrez, A., Kim, H., and Ralph, J. (2018). Structural characterization of lignin from Maize (Zea mays L.) fibers: evidence for diferuloylputrescine incorporated into the lignin polymer in Maize kernels. J. Agric. Food Chem. 66: 4402–4413. Rioux, D., Blais, M., Nadeau-Thibodeau, N., Lagacé, M., Des Rochers, P., Klimaszewska, K., and Bernier, L. (2018). First extensive microscopic study of butternut defense mechanisms following inoculation with the canker pathogen Ophiognomonia clavigignenti-juglandacearum reveals compartmentalization of tissue damage. Phytopathology 108: 1237–1252. Rioux, D., Nicole, M., Simard, M., and Ouellette, G.B. (1998). Immunocytochemical evidence that secretion of pectin occurs during gel (gum) and tylosis formation in trees. Phytopathology 88: 494–505. Rittinger, P.A., Biggs, A.R., and Peirson, D.R. (1986). Histochemistry of lignin and suberin deposition in boundary layers formed after wounding in various plant species and organs. Can. J. Bot. 65: 1886–1892. Robb, J., Lee, S.W., Mohan, R., and Kolattukudy, P.E. (1991). Chemical characterization of stress-induced vascular coating in tomato. Plant Physiol. 97: 528– 536. Sabella, E., Luvisi, A., Aprile, A., Negro, C., Vergine, M., Nicolì, F., Miceli, A., and De Bellis, L. (2018). Xylella fastidiosa induces differential expression of lignification related-genes and lignin accumulation in tolerant olive trees cv. Leccino. J. Plant Physiol. 220: 60–68. Salas-González, I., Reyt, G., Flis, P., Custódio, V., Gopaulchan, D., Bakhoum, N., Dew, T.P., Suresh, K., Franke, R.B., Dangl, J.L., et al. (2021). Coordination 33 New Phytologist Page 34 of 59 between microbiota and root endodermis supports plant mineral nutrient homeostasis. Science. 371: eabd0695. Schmidt, A., Grimm, R., Schmidt, J., Scheel, D., and Strackt, D. (1999). Cloning and expression of a potato cDNA encoding hydroxycinnamoyl- CoA:tyramine N- (hydroxycinnamoyl)transferase. J. Biol. Chem. 274: 4273–4280. Scortichini, M. (2020). The multi-millennial olive agroecosystem of salento (Apulia, Italy) threatened by Xylella fastidiosa subsp. Pauca: A working possibility of restoration. Sustain. 12: 6700. Serra, O., Figueras, M., Franke, R., Prat, S., and Molinas, M. (2010). Unraveling ferulate role in suberin and periderm biology by reverse genetics. Plant Signal. Behav. 5: 953–958. Serrano, M., Coluccia, F., Torres, M., L’Haridon, F., and Métraux, J.P. (2014). The cuticle and plant defense to pathogens. Front. Plant Sci. 5: 274. Smith, R.A., Schuetz, M., Roach, M., Mansfield, S.D., Ellis, B., and Samuels, L. (2013). Neighboring parenchyma cells contribute to Arabidopsis xylem lignification, while lignification of interfascicular fibers is cell autonomous. Plant Cell 25: 3988–3999. Street, P.F.S., Robb, J., and Ellis, B.E. (1986). Secretion of vascular coating components by xylem parenchyma cells of tomatoes infected with Verticillium albo-atrum. Protoplasma 132: 1–11. Thoquet, P., Olivier, J., Sperisen, C., Rogowsky, P., Laterrot, H., and Grimsley, N. (1996). Quantitative trait loci determining resistance to bacterial wilt in tomato cultivar Hawaii7996. Mol. Plant-Microbe Interact. 9: 826–836. Türker-Kaya, S. and Huck, C.W. (2017). A review of mid-infrared and near-infrared imaging: principles, concepts and applications in plant tissue analysis. Molecules 22: 168. Underwood, W. (2012). The plant cell wall: a dynamic barrier against pathogen invasion. Front. Plant Sci. 3: 85. Ursache, R. et al. (2021). GDSL-domain proteins have key roles in suberin polymerization and degradation. Nat. Plants 7: 353–364. VanderMolen, G.E., Beckman, C.H., and Rodehorst, E. (1987). The ultrastructure of tylose formation in resistant banana following inoculation with Fusarium oxysporum 34 Page 35 of 59 New Phytologist f.sp. cubense. Physiol. Mol. Plant Pathol. 31: 185–200. Vasse, J., Frey, P., and Trigalet, A. (1995). Microscopic studies of intercellular infection and protoxylem invasion of tomato roots by Pseudomonas solanacearum. Mol. Plant- Microbe Interact. 8: 241–251. Vasse, J., Genin, S., Frey, P., Boucher, C., and Brito, B. (2000). The hrpB and hrpG regulatory genes of Ralstonia solanacearum are required for different stages of the tomato root infection process. Mol. Plant-Microbe Interact. 13: 259–267. Wang, J.F., Ho, F.I., Truong, H.T.H., Huang, S.M., Balatero, C.H., Dittapongpitch, V., and Hidayati, N. (2013). Identification of major QTLs associated with stable resistance of tomato cultivar “Hawaii 7996” to Ralstonia solanacearum. Euphytica 190: 241–252. Wang, J.F., Olivier, J., Thoquet, P., Mangin, B., Sauviac, L., and Grimsley, N.H. (2000). Resistance of tomato line Hawaii7996 to Ralstonia solanacearum Pss4 in Taiwan is controlled mainly by a major strain-specific locus. Mol. Plant-Microbe Interact. 13: 6–13. Xu, L., Zhu, L., Tu, L., Liu, L., Yuan, D., Jin, L., Long, L., and Zhang, X. (2011). Lignin metabolism has a central role in the resistance of cotton to the wilt fungus Verticillium dahliae as revealed by RNA-seq-dependent transcriptional analysis and histochemistry. J. Exp. Bot. 62: 5607–5621. Yadeta, K.A. and Thomma, B.P.H.J. (2013). The xylem as battleground for plant hosts and vascular wilt pathogens. Front. Plant Sci. 4: 97. Zeiss, D.R., Mhlongo, M.I., Tugizimana, F., Steenkamp, P.A., and Dubery, I.A. (2019). Metabolomic profiling of the host response of tomato (Solanum lycopersicum) following infection by Ralstonia solanacearum. Int. J. Mol. Sci. 20: 3945. Zeiss, D.R., Piater, L.A., and Dubery, I.A. (2021). Hydroxycinnamate amides: intriguing conjugates of plant protective metabolites. Trends Plant Sci. 26: 184–195.. Zhang, L., Merlin, I., Pascal, S., Bert, P.F., Domergue, F., and Gambetta, G.A. (2020). Drought activates MYB41 orthologs and induces suberization of grapevine fine roots. Plant Direct 4: 278 Zhang, Y., Zhang, W., Han, L., Li, J., Shi, X., Hikichi, Y., and Ohnishi, K. (2019). Involvement of a PadR regulator PrhP on virulence of Ralstonia solanacearum by 35 New Phytologist Page 36 of 59 controlling detoxification of phenolic acids and type III secretion system. Mol. Plant Pathol. 20: 1477–1490. Zuluaga, A.P., Solé, M., Lu, H., Góngora-Castillo, E., Vaillancourt, B., Coll, N., Buell, C.R., and Valls, M. (2015). Transcriptome responses to Ralstonia solanacearum infection in the roots of the wild potato Solanum commersonii. BMC Genomics 16: 246. Supplemental data: Table S1: List of primers used in this study. Table S2: Assignments of the correlation signals in the 2D HSQC spectra. Figure S1: Tissue used for analysis and bacterial dynamics. Figure S2: H7996 plants show mild symptoms upon challenge inoculation of R. solanacearum. Figure S3: Vascular coating response to R. solanacearum infection with wall bound phenolics. Figure S4: Expression of suberin biosynthetic genes in xylem vasculature of taproots upon infection of R. solanacearum. Figure S5: Phylogeny of Feruloyl transferase (FHT) orthologues in different plant species and expression of the putative tomato FHT ortholog in response to Ralstonia solanacearum infection. Figure S6: Phylogeny of tyramine hydroxycinnamoyl transferase (THT) orthologues in different plant species and expression of the tomato THT gene family members in response to R. solanacearum infection. Figure S7: Expression of phenylpropanoid pathway genes in xylem vasculature of taproots upon invasion of R. solanacearum. Figure S8: Immunoblot of 35S::SlFHT-HA in independent Marmande tomato lines expressing 35S::SlFHT-HA (Marmande). Figure S9: Fresh weight of 35S::SlFHT-HA plants. Figure S10: Fresh weight of 35S::SlTHT1-3 plants. 36 Page 37 of 59 New Phytologist Figure S11: Overexpression of SlTHT1-3 in tomato results in restricted colonization by R. solanacearum. Figure Legends Figure 1: Resistant H7996 tomato restricts Ralstonia solanacearum colonization and induces a vascular coating response with wall bound phenolics. Susceptible (Marmande) and resistant (H7996), 5-week old tomato plants were inoculated through roots by soil-soak with ~1x107 colony forming units (CFU)/ml of R. solanacearum GMI1000 and incubated at 28°C. (a) At 12 days post-inoculation (dpi) most Marmande plants showed severe wilting symptoms, whereas H7996 remained mostly symptomless. (b) Cross- sections of the taproot-to-hypocotyl area containing 105 CFU g −1 of R. solanacearum were analyzed by ultraviolet (UV) microscopy. To focus on cell wall-deposited phenolic compounds, soluble phenolic compounds were removed with ethanol prior to observation. A strong autoflorescence signal emitted from the walls of vessels and surrounding parenchyma cells in infected H7996 plants compared to Marmande or the mock controls can be observed. Fluorescence signal in white was green colored. Images from a representative experiment out of 3 with n=5 plants per cultivar. Scale bar = 500 µm. (c) The UV auto-fluorescence signal in (b) was measured using the Leica Application Suite X (LAS X) software. A representative region of interest (ROI) is highlighted in (b) and corresponded to a line traversing the selected vascular bundles. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Different letters indicate statistically significant differences (α=0.05, Fisher’s least significant difference test). Figure 2: Feruloylamides, tyramine-derived amides and suberin-compatible compounds are specifically enriched in resistant H7996 tomato after infection with R. solanacearum. (a) Two dimensional heteronuclear single-quantum correlation–nuclear magnetic resonance (2D-HSQC NMR) spectra of enzymatically isolated lignin/suberin fractions from mock-treated and R. solanacearum-infected taproots (containing 105 CFU g- 1 taproot-to-hypocotyl transition tissue) of H7996 and Marmande tomato plants. The 37 New Phytologist Page 38 of 59 experiment was performed twice with similar results. (b) Main lignin/suberin structures identified: β–O–4´ alkyl aryl ethers (A), β–5´ fenylcoumarans (B), β–β´ resinols (C), cinnamyl alcohols end-groups (I), feruloylamides (FAm), tyramine-derived amides (Ty), guaiacyl lignin units (G), syringyl lignin units (S), as well as unassigned aliphatic signals from suberin. The structures and contours of the HSQC signals are color coded to aid interpretation. 1H and 13C NMR chemical shifts of the assigned signals are detailed in Table S2. To detect FAm7 signal, the spectrum scaled-up to 2-fold (×2) intensity. The abundances of the main lignin linkages (A, B and C) and cinnamyl alcohol end-groups (I) are referred to as a percentage of the total lignin units (S + G = 100%). Figure 3: Resistant H7996 tomato shows vascular autofluorescence not-quenched with phloroglucinol and susceptible Marmande shows a decrease in phloroglucinol-HCl lignin signal. Susceptible (Marmande) and resistant (H7996) 5-week-old tomato plants were root-inoculated with a R. solanacearum GMI1000 strain at a concentration of ~1x107 CFU/ml or water mock. (a) Cross-sections of the taproot-to-hypocotyl area containing 105 CFU g −1 of R. solanacearum were stained with phloroglucinol-HCl and observed under UV to visualize other autofluorescent compounds different from lignin (not quenched with phloroglucinol-HCl) (right) and under brightfield to visualize lignin deposition (left). In infected H7996 strong UV autofluorescence could be observed in the walls of xylem vessels surrounding xylem parenchyma cells and tracheids, indicating reinforcement of walls of vascular tissue with phenolics formed de novo upon infection. In infected Marmande the red phlorogucinol stain was reduced especially in the intervessel areas. (b) The UV auto-fluorescence signal in (a) was measured using the LAS X Leica software after the Phloroglucinol-HCl treatment. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. (c) Detailed observation of infected H7996 xylem after the Phloroglucinol-HCl treatment shows the strong UV fluorescence concentrated in specific areas possibly corresponding to intervessel and vessel-parenchyma bordered pit membranes and/or pit chambers (yellow and white arrows, respectively). Fluorescence was also observed in parenchyma cells, specially enriched at intercellular cell corners (green arrow). (b) correspond to a representative experiment out of 3 each with n=6 38 Page 39 of 59 New Phytologist plants per variety. Different letters indicate statistically significant differences (α=0.05, Fisher’s least significant difference test). (a) and (c) were representative images. Scale bars = 100 µm in (a, left), 500 µm in (a, right) and 50 µm in (c). Figure 4: Resistant H7996 tomato shows cell wall ferulate/feruloylamide and suberin deposition in restricted zones of vascular tissue upon R. solanacearum infection. Susceptible Marmande or resistant H7996 tomato plants were soil-inoculated with a ~1x107 CFU/ml suspension of Ralstonia solanacearum GMI1000 or mock-inoculated with water and incubated at 28°C. Cross-sections were obtained from taproot-to-hypocotyl transition tissue containing 105 CFU g −1 of R. solanacearum. Sections were stained with Sudan IV to visualize suberin aliphatics and subsequently treated with 1N KOH (pH above 10) to visualize ferulic acid bound to cell wall. Sudan IV positive staining (reddish-brown coloration) was observed around xylem vessels specifically in infected H7996, indicating accumulation of suberin aliphatics. Accumulation of ferulic acid bound to cell wall (blue- green coloration) appears also specifically in infected H7996 resistant tomato, surrounding Sudan IV-stained areas. White arrowheads indicate the sites of accumulation of ferulates and aliphatic compounds. Representative images from one experiment out of three with n=6 plants each were taken. Scale bar= 100 µm. Figure 5: Genes of the ligno-suberin heteropolymer biosynthesis pathway are specifically induced in the xylem vasculature of resistant H7996 tomato upon R. solanacearum infection. The levels of expression of genes belonging to metabolic pathways relevant for suberin, lignin and feruloyltyramine and related amides biosynthesis were analyzed by quantitative PCR (qPCR) of taproot vascular tissue in infected or mock- treated H7996 or Marmande tomato plants. Plants containing an R. solanacearum inoculum of 105 CFU g−1 were selected and xylem vascular tissue from the taproot-to-hypocotyl transition zone, comprising of metaxylems and surrounding parenchyma cells was collected for RNA extraction and cDNA synthesis. In parallel, xylem tissue was collected from mock plants. Heatmaps show log2 fold change RTA (relative transcript abundance) values of infected vs. mock for Marmande (left) and Hawaii (right). The tomato gene encoding for the alpha-subunit of the translation elongation factor 1 (SleEF1 α) was used as endogenous 39 New Phytologist Page 40 of 59 reference. Three biological replicates (n=3) were used, and taproots of 6 plants were used in each replicate. All the original qPCR results can be found in Figs. S3, S4, S5 and S6. The scheme represents the phenylpropanoid and suberin biosynthesis pathways providing lignin-like and suberin precursors for the ligno-suberin heteropolymer. Abbreviations: PAL: Phenylalanine ammonia–lyase; C4H: Cinnamate–4–hydroxylase; C3H: Coumarate 3- hydroxylase; 4CL: 4–Coumarate–CoA ligase; HCT: Hydroxycinnamoyl–CoA shikimate/quinate hydroxycinnamoyl transferase; COMT: Caffeic acid 3-O- methyltransferase; CCoAOMT: Caffeoyl CoA 3-O-methyltransferase; CYP86A1 and CYP86B1: cytochrome P450 fatty acid ω-hydroxylases; KCS1/2: 3-ketoacyl-CoA synthase; FAR 1/3/4: Fatty acyl-CoA reductase; GPAT5: glycerol-3-phosphate acyltransferase 5; THT: Tyramine hydroxycinnamoyl transferase; TyDC: Tyrosine decarboxylase; FHT: feruloyl transferase. The question mark (?) denotes a hypothetical reaction. Figure 6: Overexpression of SlFHT-HA in susceptible tomato slightly restricts colonization by R. solanacearum. (a, b) A pathogenicity assay was performed comparing Wt and 3 independent 35S::SlFHT-HA Marmande tomato lines (a, c and d) after infection with R. solanacearum GMI1000 lux reporter strain. Five-week-old plants were soil-soak inoculated with ~1x107 CFU/ml or mock and grown at 28°C. (a) Wilting progress was monitored by rating plants daily on a 0 to 4 disease index scale where 0 = healthy and 4 =100% wilted. Plotted values correspond to means ± standard error of 24 independent plants (n=24) from a representative experiment out of a total of 3. Asterisks indicate statistically significant differences between Wt and each of the 35S::FHT-HA analyzed using a paired Student’s t-test (* p<0.05). (b) The level of R. solanacearum colonization in the taproot and hypocotyl was calculated as colony forming units per gram of fresh taproot tissue (CFU·g −1) at 12 dpi. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Data presented are of a representative experiment out of a total of 3 experiments. In (a) a Kruskall-Wallis test at each day post infection (dpi) was conducted to examine differences in disease index among different genotypes. Significant differences among genotypes were confirmed by applying a pairwise Wilcox test. Asterisks 40 Page 41 of 59 New Phytologist indicate statistically significant differences between wild type and 35S::FHT-HA tomato lines in a (* corresponds to a p-value of p <0.05). Figure 7: Overexpression of SlTHT1-3 in susceptible tomato confers resistance to R. solanacearum. (a, b) A pathogenicity assay was performed comparing Wt and 35S::SlTHT1-3 tomato lines (Moneymaker background) after infection with R. solanacearum lux reporter GMI1000 strain. Five-week-old plants were soil-soak inoculated with ~1x107 CFU/ml and grown at 28°C. (a) Wilting progress was monitored by rating plants daily on a 0 to 4 disease index scale where 0 = healthy and 4 =100% wilted. Plotted values correspond to means ± standard error of 24 independent plants (n=24) from a representative experiment out of a total of 3. Asterisks indicate statistically significant differences between Wt and 35S::SlTHT1-3 using a Kruskall-Wallis test at each day post infection (dpi) was conducted to examine differences in disease index among different genotypes. Significant differences among genotypes were confirmed by applying a pairwise Wilcox test”. (* corresponds to a p-value of p <0.05 and *** to p < 0.0001). (b) Pictures were taken 12 days post-infection. Wt plants were arranged according to the degree of symptom severity (from 4 to 0). (c) Transgenic 35S::SlTHT1-3 tomato significantly restricted R. solanacearum colonization in both the taproot-to-hypocotyl transition zone and hypocotyl compared to Wt. Five-week-old tomato plants were root-inoculated with a R. solanacearum GMI1000 luciferase reporter strain at a concentration of ~1x107 CFU/ml or water mock. The level of in planta colonization by R. solanacearum was calculated as colony forming units per gram of fresh taproot tissue (CFU·g −1) at 12 dpi. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Box-and- whisker plots show data from a representative experiment out of 3 (n =14 to 16). (*** corresponds to a p-value of p < 0.0001) (d) Transverse stem cross-sections of Wt and transgenic 35S::SlTHT1-3 tomato lines were imaged under a confocal microscope 6 days after infection with a R. solanacearum GMI1000 green fluorescent protein (GFP) reporter strain. R. solanacearum at a concentration of 105 CFU ml −1 was injected directly into the xylem vasculature of the first internode thorough the petiole. Orange arrow points the site of inoculation. Representative images of R. solanacearum colonization progress at the point 41 New Phytologist Page 42 of 59 of inoculation are shown. Scale bar= 2 mm. (e) Mean green fluorescence of the GFP signal emitted from R. solanacearum at cross-sections obtained as described in (d) at the point of inoculation (0), below the point of inoculation (-0.5 cm) and above the point of inoculation (+0.5 cm) was measured using ImageJ. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Data from a representative experiment out of a total of 3, with n=5 plants per condition. In (a) a Kruskall-Wallis test at each day post infection (dpi) was conducted to examine differences in disease index among different genotypes. Significant differences among genotypes were confirmed by applying a pairwise Wilcox test. Asterisks indicate statistically significant differences between wild type and 35S::SlTHT1-3 tomato lines in (a) (* corresponds to a p-value of p <0.05 and ** to p < 0.001). Figure 8: Overexpression of SlTHT1-3 in susceptible tomato results in vascular autofluorescence not-quenched with phloroglucinol and cell wall ferulate/feruloylamide deposition in restricted zones of vascular tissue upon R. solanacearum infection. Susceptible Moneymaker (Wt) or resistant 35S::SlTHT1-3 overxpressing tomato plants were soil-inoculated with a ~1x107 CFU/ml suspension of R. solanacearum GMI1000 lux reporter strain. Cross-sections were obtained from taproot-to- hypocotyl of both genotypes tissue containing containing 105 CFU g−1 of R. solanacearum. (a) Cross-sections were stained with phloroglucinol-HCl and observed under brightfield to visualize lignin deposition (left) and under UV to visualize other auto-fluorescent compounds different from lignin (not quenched with phloroglucionol- HCl) (right). (b) Combined Sudan IV+KOH treatment showed no positive suberin aliphatic signal in SlTHT1-3, but a significant increase in ferulate/feruloylamide accumulation upon infection. (c) Close-ups (10X) of the vascular bundles of Wt and 35S::SlTHT1-3 infected plants pointed with a white arrow in (b) are also shown. Images from a representative experiment out of 3 with n=6 plants per cultivar. (d) Quantification of UV fluorescence after phloroglucinol-HCl staining as shown in (a, right) were performed with LAS X software by selecting the vascular areas surrounding main vessels with strong localized fluorescence or green signal. (e) Quantification of UV green fluorescence from ferulate deposits after 42 Page 43 of 59 New Phytologist Sudan IV+KOH staining as shown in (a, right) were performed with LAS X software by selecting the vascular areas surrounding main vessels with strong localized fluorescence or green signal. Data in (d) and (e) are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Different letters indicate statistically significant differences (α=0.05, Fisher’s least significant difference test). Scale bar = 500 µm. Figure 9: Schematic representation of the vascular ligno-suberization process potentially taking place in infected vessels of resistant H7996 tomato upon R. solanacearum infection. Colonization of the vasculature by R. solanacearum in resistant tomato plants induces a ligno-suberization process in the walls of the infected vessel (V) and of the adjacent tracheids (T) and parenchyma cells (XP) (red). The lignin-like polymer accompanying suberin would be enriched in structural feruloyltyramine and related amides. The signal of structural ferulic acid (ester or as amide) would extend to the walls of peripheral parenchyma cells, vessels and tracheids (green), indicating a stage preceding suberization or a final layered pattern, still to be resolved. Together, the red and green areas, would form a “zone of ligno-suberization” (black dashed line) potentially creating a physico-chemical barrier to limit R. solanacearum spread from the colonized xylem vessel lumen. 43 New Phytologist Page 44 of 59 Figure 1: Resistant H7996 tomato restricts Ralstonia solanacearum colonization and induces a vascular coating response with wall bound phenolics. Susceptible (Marmande) and resistant (H7996), 5-week old tomato plants were inoculated through roots by soil-soak with ~1x107 colony forming units (CFU)/ml of R. solanacearum GMI1000 and incubated at 28°C. (a) At 12 days post-inoculation (dpi) most Marmande plants showed severe wilting symptoms, whereas H7996 remained mostly symptomless. (b) Cross-sections of the taproot-to-hypocotyl area containing 105 CFU g −1 of R. solanacearum were analyzed by ultraviolet (UV) microscopy. To focus on cell wall-deposited phenolic compounds, soluble phenolic compounds were removed with ethanol prior to observation. A strong autoflorescence signal emitted from the walls of vessels and surrounding parenchyma cells in infected H7996 plants compared to Marmande or the mock controls can be observed. Fluorescence signal in white was green colored. Images from a representative experiment out of 3 with n=5 plants per cultivar. Scale bar = 500 µm. (c) The UV auto-fluorescence signal in (b) was measured using the Leica Application Suite X (LAS X) software. A representative region of interest (ROI) is highlighted in (b) and corresponded to a line traversing the selected vascular bundles. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate Page 45 of 59 New Phytologist second quartile, median and third quartile. Different letters indicate statistically significant differences (α=0.05, Fisher’s least significant difference test). 561x810mm (96 x 96 DPI) New Phytologist Page 46 of 59 Figure 2: Feruloylamides, tyramine-derived amides and suberin-compatible compounds are specifically enriched in resistant H7996 tomato after infection with R. solanacearum. (a) 2D-HSQC NMR spectra of enzymatically isolated lignin/suberin fractions from mock-treated and R. solanacearum-infected taproots (containing 105 CFU g-1 taproot-to-hypocotyl transition tissue) of H7996 and Marmande tomato plants. The experiment was performed twice with similar results. (b) Main lignin/suberin structures identified: β–O–4´ alkyl aryl ethers (A), β–5´ fenylcoumarans (B), β–β´ resinols (C), cinnamyl alcohols end-groups (I), feruloylamides (FAm), tyramine-derived amides (Ty), guaiacyl lignin units (G), syringyl lignin units (S), as well as unassigned aliphatic signals from suberin. The structures and contours of the HSQC signals are color coded to aid interpretation. 1H and 13C NMR chemical shifts of the assigned signals are detailed in Table S1. To detect FAm7 signal, the spectrum scaled-up to 2-fold (×2) intensity. The abundances of the main lignin linkages (A, B and C) and cinnamyl alcohol end-groups (I) are referred to as a percentage of the total lignin units (S + G = 100%). 248x244mm (300 x 300 DPI) Page 47 of 59 New Phytologist Figure 3: Resistant H7996 tomato shows vascular autofluorescence not-quenched with phloroglucinol and susceptible Marmande shows a decrease in phloroglucinol-HCl lignin signal. Susceptible (Marmande) and resistant (H7996) 5-week-old tomato plants were root-inoculated with a R. solanacearum GMI1000 strain at a concentration of ~1x107 CFU/ml or water mock. (a) Cross-sections of the taproot-to-hypocotyl area containing 105 CFU g −1 of R. solanacearum were stained with phloroglucinol-HCl and observed under UV to visualize other autofluorescent compounds different from lignin (not quenched with phloroglucinol-HCl) (right) and under brightfield to visualize lignin deposition (left). In infected H7996 strong UV autofluorescence could be observed in the walls of xylem vessels surrounding xylem parenchyma cells and tracheids, indicating reinforcement of walls of vascular tissue with phenolics formed de novo upon infection. In infected Marmande the red phlorogucinol stain was reduced especially in the intervessel areas. (b) The UV auto-fluorescence signal in (a) was measured using the LAS X Leica software after the Phloroglucinol-HCl treatment. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. (c) Detailed observation of infected H7996 xylem after the Phloroglucinol-HCl treatment shows the strong UV fluorescence New Phytologist Page 48 of 59 concentrated in specific areas possibly corresponding to intervessel and vessel-parenchyma bordered pit membranes and/or pit chambers (yellow and white arrows, respectively). Fluorescence was also observed in parenchyma cells, specially enriched at intercellular cell corners (green arrow). (b) correspond to a representative experiment out of 3 each with n=6 plants per variety. Different letters indicate statistically significant differences (α=0.05, Fisher’s least significant difference test). (a) and (c) were representative images. Scale bars = 100 µm in (a, left), 500 µm in (a, right) and 50 µm in (c). 190x275mm (283 x 283 DPI) Page 49 of 59 New Phytologist Figure 4: Resistant H7996 tomato shows cell wall ferulate/feruloylamide and suberin deposition in restricted zones of vascular tissue upon R. solanacearum infection. Susceptible Marmande or resistant H7996 tomato plants were soil-inoculated with a ~1x107 CFU/ml suspension of Ralstonia solanacearum GMI1000 or mock- inoculated with water and incubated at 28°C. Cross-sections were obtained from taproot-to-hypocotyl transition tissue containing 105 CFU g −1 of R. solanacearum. Sections were stained with Sudan IV to visualize suberin aliphatics and subsequently treated with 1N KOH (pH above 10) to visualize ferulic acid bound to cell wall. Sudan IV positive staining (reddish-brown coloration) was observed around xylem vessels specifically in infected H7996, indicating accumulation of suberin aliphatics. Accumulation of ferulic acid bound to cell wall (blue-green coloration) appears also specifically in infected H7996 resistant tomato, surrounding Sudan IV-stained areas. White arrowheads indicate the sites of accumulation of ferulates and aliphatic compounds. Representative images from one experiment out of three with n=6 plants each were taken. Scale bar= 100 µm. 190x275mm (283 x 283 DPI) New Phytologist Page 50 of 59 Page 51 of 59 New Phytologist Figure 5: Genes of the ligno-suberin heteropolymer biosynthesis pathway are specifically induced in the xylem vasculature of resistant H7996 tomato upon R. solanacearum. The levels of expression of genes belonging to metabolic pathways relevant for suberin, lignin and feruloyltyramine and related amides biosynthesis were analyzed by qPCR of taproot vascular tissue in infected or mock-treated H7996 or Marmande tomato plants. Plants containing an R. solanacearum inoculum of 105 CFU g−1 were selected and xylem vascular tissue from the taproot-to-hypocotyl transition zone, comprising of metaxylems and surrounding parenchyma cells was collected for RNA extraction and cDNA synthesis. In parallel, xylem tissue was collected from mock plants. Heatmaps show log2 fold change RTA (relative transcript abundance) values of infected vs. mock for Marmande (left) and Hawaii (right). The tomato gene encoding for the alpha- subunit of the translation elongation factor 1 (SleEF1 α) was used as endogenous reference. Three biological replicates (n=3) were used, and taproots of 6 plants were used in each replicate. All the original qPCR results can be found in Figs. S3, S4, S5 and S6. The scheme represents the phenylpropanoid and suberin biosynthesis pathways providing lignin-like and suberin precursors for the ligno-suberin heteropolymer. Abbreviations: PAL: Phenylalanine ammonia–lyase; C4H: Cinnamate–4–hydroxylase; C3H: Coumarate 3- New Phytologist Page 52 of 59 hydroxylase; 4CL: 4–Coumarate–CoA ligase; HCT: Hydroxycinnamoyl–CoA shikimate/quinate hydroxycinnamoyl transferase; COMT: Caffeic acid 3-O-methyltransferase; CCoAOMT: Caffeoyl CoA 3-O- methyltransferase; CYP86A1 and CYP86B1: cytochrome P450 fatty acid ω-hydroxylases; KCS1/2: 3- ketoacyl-CoA synthase; FAR 1/3/4: Fatty acyl-CoA reductase; GPAT5: glycerol-3-phosphate acyltransferase 5; THT: Tyramine hydroxycinnamoyl transferase; TyDC: Tyrosine decarboxylase; FHT: feruloyl transferase. The question mark (?) denotes a hypothetical reaction. 190x275mm (283 x 283 DPI) Page 53 of 59 New Phytologist Figure 6: Overexpression of SlFHT-HA in susceptible tomato slightly restricts colonization by R. solanacearum. (a, b) A pathogenicity assay was performed comparing Wt and 3 independent 35S::SlFHT-HA Marmande tomato lines (a, c and d) after infection with R. solanacearum GMI1000 lux reporter strain. Five- week-old plants were soil-soak inoculated with ~1x107 CFU/ml or mock and grown at 28°C. (a) Wilting progress was monitored by rating plants daily on a 0 to 4 disease index scale where 0 = healthy and 4 =100% wilted. Plotted values correspond to means ± standard error of 24 independent plants (n=24) from a representative experiment out of a total of 3. Asterisks indicate statistically significant differences between Wt and each of the 35S::FHT-HA analyzed using a paired Student’s t-test (* p<0.05). (b) The level of R. solanacearum colonization in the taproot and hypocotyl was calculated as colony forming units per gram of fresh taproot tissue (CFU·g −1) at 12 dpi. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Data presented are of a representative experiment out of a total of 3 experiments. In (a) a Kruskall-Wallis test at each day post infection (dpi) was conducted to examine differences in disease index among different genotypes. Significant differences among genotypes were confirmed by applying a pairwise New Phytologist Page 54 of 59 Wilcox test. Asterisks indicate statistically significant differences between wild type and 35S::FHT-HA tomato lines in a (* corresponds to a p-value of p <0.05). 190x275mm (283 x 283 DPI) Page 55 of 59 New Phytologist Figure 7: Overexpression of SlTHT1-3 in susceptible tomato confers resistance to R. solanacearum. (a, b) A pathogenicity assay was performed comparing Wt and 35S::SlTHT1-3 tomato lines (Moneymaker background) after infection with R. solanacearum lux reporter GMI1000 strain. Five-week-old plants were soil-soak inoculated with ~1x107 CFU/ml and grown at 28°C. (a) Wilting progress was monitored by rating plants daily on a 0 to 4 disease index scale where 0 = healthy and 4 =100% wilted. Plotted values correspond to means ± standard error of 24 independent plants (n=24) from a representative experiment out of a total of 3. Asterisks indicate statistically significant differences between Wt and 35S::SlTHT1-3 using a Kruskall-Wallis test at each day post infection (dpi) was conducted to examine differences in disease index among different genotypes. Significant differences among genotypes were confirmed by applying a pairwise Wilcox test”. (* corresponds to a p-value of p <0.05 and *** to p < 0.0001). (b) Pictures were taken 12 days post-infection. Wt plants were arranged according to the degree of symptom severity (from 4 to 0). (c) Transgenic 35S::SlTHT1-3 tomato significantly restricted R. solanacearum colonization in both the taproot-to-hypocotyl transition zone and hypocotyl compared to Wt. Five-week-old tomato plants were root- inoculated with a R. solanacearum GMI1000 luciferase reporter strain at a concentration of ~1x107 CFU/ml New Phytologist Page 56 of 59 or water mock. The level of in planta colonization by R. solanacearum was calculated as colony forming units per gram of fresh taproot tissue (CFU·g −1) at 12 dpi. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Box-and-whisker plots show data from a representative experiment out of 3 (n =14 to 16). (*** corresponds to a p-value of p < 0.0001) (d) Transverse stem cross-sections of Wt and transgenic 35S::SlTHT1-3 tomato lines were imaged under a confocal microscope 6 days after infection with a R. solanacearum GMI1000 green fluorescent protein (GFP) reporter strain. R. solanacearum at a concentration of 105 CFU ml −1 was injected directly into the xylem vasculature of the first internode thorough the petiole. Orange arrow points the site of inoculation. Representative images of R. solanacearum colonization progress at the point of inoculation are shown. Scale bar= 2 mm. (e) Mean green fluorescence of the GFP signal emitted from R. solanacearum at cross-sections obtained as described in (d) at the point of inoculation (0), below the point of inoculation (-0.5 cm) and above the point of inoculation (+0.5 cm) was measured using ImageJ. Data are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Data from a representative experiment out of a total of 3, with n=5 plants per condition. In (a) a Kruskall-Wallis test at each day post infection 190x275mm (283 x 283 DPI) Page 57 of 59 New Phytologist Figure 8: Overexpression of SlTHT1-3 in susceptible tomato results in vascular autofluorescence not- quenched with phloroglucinol and cell wall ferulate/feruloylamide deposition in restricted zones of vascular tissue upon R. solanacearum infection. Susceptible Moneymaker (Wt) or resistant 35S::SlTHT1-3 overxpressing tomato plants were soil-inoculated with a ~1x107 CFU/ml suspension of R. solanacearum GMI1000 lux reporter strain. Cross-sections were obtained from taproot-to-hypocotyl of both genotypes tissue containing containing 105 CFU g−1 of R. solanacearum. (a) Cross-sections were stained with phloroglucinol-HCl and observed under brightfield to visualize lignin deposition (left) and under UV to visualize other auto-fluorescent compounds different from lignin (not quenched with phloroglucionol- HCl) (right). (b) Combined Sudan IV+KOH treatment showed no positive suberin aliphatic signal in SlTHT1-3, but a significant increase in ferulate/feruloylamide accumulation upon infection. (c) Close-ups (10X) of the vascular bundles of Wt and 35S::SlTHT1-3 infected plants pointed with a white arrow in (b) are also shown. Images from a representative experiment out of 3 with n=6 plants per cultivar. (d) Quantification of UV fluorescence after phloroglucinol-HCl staining as shown in (a, right) were performed with LAS X software by selecting the vascular areas surrounding main vessels with strong localized fluorescence or green signal. (e) New Phytologist Page 58 of 59 Quantification of UV green fluorescence from ferulate deposits after Sudan IV+KOH staining as shown in (a, right) were performed with LAS X software by selecting the vascular areas surrounding main vessels with strong localized fluorescence or green signal. Data in (d) and (e) are represented with box and whiskers plots: whiskers indicate variability outside the upper and lower quartiles and boxes indicate second quartile, median and third quartile. Different letters indicate statistically significant differences (α=0.05, Fisher’s least significant difference test). Scale bar = 500 µm. 190x275mm (283 x 283 DPI) Page 59 of 59 New Phytologist Figure 9: Schematic representation of the vascular ligno-suberization process potentially taking place in infected vessels of resistant H7996 tomato upon R. solanacearum infection. Colonization of the vasculature by R. solanacearum in resistant tomato plants induces a ligno-suberization process in the walls of the infected vessel (V) and of the adjacent tracheids (T) and parenchyma cells (XP) (red). The lignin-like polymer accompanying suberin would be enriched in structural feruloyltyramine and related amides. The signal of structural ferulic acid (ester or as amide) would extend to the walls of peripheral parenchyma cells, vessels and tracheids (green), indicating a stage preceding suberization or a final layered pattern, still to be resolved. Together, the red and green areas, would form a “zone of ligno-suberization” (black dashed line) potentially creating a physico-chemical barrier to limit R. solanacearum spread from the colonized xylem vessel lumen. 190x275mm (283 x 283 DPI)