SALK_ SALK_ SALK_ GABI_692H09 G A G A SALK_ CL1.3 4CL1.1 4CL1.2 SM_3_ CL1.3 4CL1.1 4CL1.

Transcription

1 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc pal1-2 pal1-3 pal2-2 SALK_ SALK_ SALK_ GABI_692H9 residual expression pal2-3 c4h c4h-3* 4cl1-1 4cl1-2 4cl2-1 4cl2-3 ccoaomt1-3 ccoaomt1-5 ccr1-3 ccr1-6 f5h1-2* f5h1-4 comt1-1 comt1-4 cad6-1 cad G A G A 5 3 SALK_ CL1.3 4CL1.1 4CL SM_3_ CL1.3 4CL1.1 4CL SL 2_14_4 5 3 GABI_353A SALK_ GABI_7F2 5 3 SALK_ GABI_622C1 5 C T 5 3 SALK_ SALK_ SALK_ SALK_ SAIL_776_B scale (kb) Supplemental Figure 1. Schematic representation of the 2 mutant genes used in this study. All mutants are in Col-. Thick bar: open reading frame; purple: exon; gray intron; pink vertical line: point mutation; pink triangle: T-DNA insertion. For 4CL1, three splicing variants exist. The positions of their stop-codons are given. The residual expression is given as log 2 (abundance in the sample/ abundance in 24cm). *The c4h-3 and f5h1-2 mutant have been complemented by Schilmiller, et al., (29) and Meyer, et al., (1998) respectively. References Meyer, K., Shirley, A.M., Cusumano, J.C., Bell-Lelong, D.A. and Chapple, C.C.S (1998). Lignin monomer composition is determined by the expression of a cytochrome P45-dependent monooxygenase in Arabidopsis. Proc. Natl. Acad. Sci. U. S. A. 95: Schilmiller, A.L., Stout, J., Weng, J.-K., Humphreys, J., Ruegger, M.O., and Chapple, C. (29). Mutations in the cinnamate 4-hydroxylase gene impact metabolism, growth and development in Arabidopsis. Plant J. 6:

2 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc Supplemental Figure 2. Growth characteristics of the 2 Arabidopsis lines given in Table 1. Panel A: detailed growth characteristics. Negative values for slope indicate slower growth, whereas negative values for plateau points to smaller fully-grown plants. *.5>q-value>.1, **.1>q-value>.1, ***.1>q-value. The height of plants in the eight trays was measured daily. Two replicates of each mutant and 1 replicates of were grown in each tray. A tray effect was revealed, and therefore, trays were split in two groups, each consisting of four trays. Data from 4 until 12 days was used to detect lines with different growth rate. Data from 27 until 44 days was used to detect lines with different final height. Panel B and C: growth curves of the 2 Arabidopsis lines. A growth characteristics slope plateau mutant allele group1 group2 group1 group2 pal pal *** pal pal c4h *** -1.1*** *** *** c4h ** cl cl cl * cl ccoaomt *** -.39*** ccoaomt *** ccr *** -3.11*** *** -4.91*** ccr *** -.71*** *** *** f5h *** f5h ** comt *** comt *** cad6-1.25***.17** cad * Supplemental Figure 2 page 1/3 2

3 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc B traygroup 1 traygroup 2 pal1 Average Height[cm] pal1-2 pal pal1-2 pal pal pal2-2 pal pal2-2 pal c4h c4h-2 c4h c4h-2 c4h cl cl1-1 4cl cl1-1 4cl cl cl2-1 4cl cl2-1 4cl Supplemental Figure 2 page 2/3 3

4 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc C ccoa -omt ccoaomt1-3 coaomt1-5 traygroup 1 traygroup ccoaomt1-3 ccoaomt ccr ccr1-3 ccr ccr1-3 ccr f5h f5h1-2 f5h f5h1-2 f5h comt comt-1 comt comt-1 comt cad cad6-1 cad cad6-1 cad Supplemental Figure 2 page 3/3 4

5 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc A B C transcripts GC-MS compounds UPLC-MS compounds 32 4 c 7 8 # b a # # # mutants # mutants # mutants Supplemental Figure 3. Number of transcripts (A), compounds detected by GC-MS (B) and compounds detected by UPLC-MS (C) that are differential in one or multiple mutants. Red and blue bars represent the number of transcripts or compounds that are increased and reduced in abundance, respectively, in the mutants where they were found to be different. Green bars represent the number of transcripts or compounds that are oppositely affected in different mutants. The dark colors represent the measured differentials, the faded colors the expected number of differentials by chance alone. (a), 14 genes are up in five of the set of 1 mutants; (b), 12 genes are down in five of the set 1 mutants; (c), 16 genes are oppositely regulated in at least 2 of the five mutants of the set of 1. The number of differential transcripts and metabolites that was shared between multiple mutants was much higher than expected by chance alone. For example, for a transcript that is significantly differential in 5 mutants, the probability that it is up (or down) in all 5 mutants by chance alone, is one in 32 (C). Thus, for the 42 transcripts that were found to be differential in 5 mutants in our dataset, ~ one is expected to be up in all 5 mutants, ~ one down and ~ 4 in opposite directions (up and down). For this particular example, we found 14 up, 12 down and 16 in opposite directions and, hence, a significant enrichment of transcripts that are differential in the same direction (C). Also for GC-MS detected compounds, the shared differentials differed from those that would be expected by chance alone (B). Remarkably, there was enrichment on both sides of the distribution: there were more compounds that differed in one mutant only and there were more that were shared by four to seven mutants than expected by chance alone. No compounds were shared between eight or more mutants. In contrast to the microarray results, most compounds were up in abundance (B). The distribution of UPLC-MS detected compounds that were differential in multiple mutants also deviates from the one expected by chance alone (C). Interestingly, compounds that were differential in only one or two mutants were increased in abundanced. This shows that certain mutations resulted in the accumulation of specific compounds that were not found to be differential in many other mutants of the set. On the other hand, UPLC-MS compounds that were shared by six, seven or eight mutants were mainly decreased in abundance. This can logically be explained by the fact that UPLC-MS targets phenolics (i.e. the intermediates and products of the phenylpropanoid pathway) and because the different mutations analyzed often block the pathway, the same sets of compounds were reduced. 5

6 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc MS2 benzoyl (=343, RT=9.9 min) -acetate benzoyl benzoic nm acetate MS2 p-hydroxybenzoyl (=299, RT=5. min) 137 fragmentation fragmentation MS2 vanilloyl (=329, RT=5.9 min) MS2 syringoyl (=359, RT=6.5 min) vanilloyl fragmentation syringoyl vanillic syringic fragmentation nm MS2 cinnamic (=147, RT=16.4 min) nm MS2 p-coumaric (=163, RT=7.7 min) nm nm MS2 5-hydroxy (=29, RT=3. min) MS2 cinnamoyl (=369, RT=15.6 min) -acetate acetate nm Supplemental Figure 4 page 1/15 6

7 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc MS2 p-coumaroyl (=325, RT=8.3 min) 1 MS2 diferuloyl (=531, RT=23.2 min) 1-18 and nm fragmentation and MS2 feruloyl sinapoyl (=561, RT=22.7 min) and and and MS2 p-coumaroyl (=479, RT=16.1 min) and MS2 disinapoyl (=591, RT=21.7 min) MS2 feruloyl (=59, RT=17.2 min) disinapoyl sinapic nm and fragmentation nm and fragmentation MS2 sinapoyl (=539, RT=16.7 min) MS2 cinnamoyl (=263, RT=7.8 min) fragmentation and nm Supplemental Figure 4 page 2/15 7

8 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc MS2 p-coumaroyl (=279, RT=3.6 min) MS2 caffeyl 3/ (=327, RT=5.7 min) malic if R 1 =Hex 2 malic MS2 coniferin (=41, RT=7. min) acetate -acetate MS2 (=399, RT=9.5 min) nm MS2 syringin (=431, RT=8. min) nm 371 -acetate MS2 G(8 5)feruloyl (=533, RT=18.2 min) acetate acetate A -15 B fragmentation acetate [M-H + -CO 2-1,2 A] [M-H + -H 2 O- 1,2 A] , B MS2 G(8 O 4)caffeoyl (=519, 17.4 min) MS2 G(8 5)G glucoside, DCG (=579, RT= 15.2 min) fragmentation B and A A - : 179 fragmentation B -44 B - : 339, [B+2H] - : nm A B 1,2 B -: 221 -acetate and Supplemental Figure 4 page 3/15 8

9 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc MS2 G 4-O-glucoside(red8 5)G, IDDDC glucoside ( 581, RT=15.4 min) -acetate MS2 G 7/9-O-hexoside(8 O 4)G (= 537, RT=8.2 min) nm 375 A B 521 MS2 G(8 O 4)caffeyl hexoside (= 55, RT=11.8 min) if R=Hex A - : MS2 G (8 O 4)G (= 375 perfect co-elution with = 597, RT=1.3 min) A B [B + 2H] - : 163, 165 if R 2 =H 323, 325 if R 2 =Hex A A - : 357 B nm A 195 B 179 acetate 2 29 MS2 G(8 O 4)G 9-O-hexoside (= 537, RT=15.6 min) MS2 p-coumaric (=325, RT=3.6 min) A B and and Time (min) [A-CH 2 O] A nm 2 [A-CH 2 O] A B MS2 G (8 O 4)S(8 5)G (=583 perfect co-elution with 85, RT=19.2 min) A - : B - : PC,I: 8 B - -H 2 O PC,I: 8 B - -CH 2 O AE,II: A - AE,II: 8 B MS2 dihydro-p-coumaric + (=327, RT=4.8 min) =Hex fragmentation 312 Supplemental Figure 4 page 4/15 9

10 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc MS2 dihydrocaffeic + (=343, RT=4.6 min) =Hex MS2 dihydro+ (=357, RT=5.7 min) =Hex nm fragmentation Supplemental Figure 4 page 5/15 1

11 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc Supplemental Figure 4. MS 2 spectra and the reasoning for the (tentative) structural identification of 34 metabolites of which the MS 2 spectra were not published before. Further evidence for the identities is derived from UV/Vis and high resolution FT-ICR-MS, fragments in FT-ICR-MS are derived via in source fragmentation. Compound 1, benzoyl In the chromatograms of the c4h mutants, a compound with 343 eluted at 9.9 min. In the MS 2 spectrum a neutral loss of 6 Da, which is typical for an acetic adduct, led to a base peak of 283. Further neutral losses of 6, 9, 12 and 162 Da (leading to fragments of 223, 193, 163 and 121, respectively) indicate that 283 holds a -moiety.the exact massesof the ions measured in the full-ms derived by in-source fragmentation, were , and with best hits for the chemical formulas of C 15 H 19 O 9 -,C 13 H 15 O 7 - and C 7 H 5 O 2 -, respectively. The UV/Vis absorption spectrum at 9.9minshowed a large peak around nm and a smaller one at nm (see insert). The UV-Vis absorption spectrum of benzoic (Sigma aldrich, Steinheim, Germany) had a similar shape but with peaks at nm and nm, respectively. The slightly higher wavelengths of the absorption maxima of compound 1 as compared to benzoic suggest that compound 1 is an esterified benzoate. Therefore, this compound is tentatively identified as the benzenoid benzoyl, i.e., benzoic esterified to a sugar. Compound 2, p-hydroxybenzoyl A compound with a mass of 3 Daeluted at a retention time of 5. min in the chromatograms of 4cl1 mutants. MS 2 fragmentation led to a typical loss, i.e., a neutral loss of 162 Da that resulted in an ion of 137. An ion of 93 that originates from a CO 2 loss ( 44 Da) of the aglycone ( 137), is indicative for a carboxylic.the exact for this ion is (C 13 H 15 O 8 - ) and the for the aglycone ion (C 7 H 5 O 3 - ). Given the biological origin of the sample, the aglycone is most likely p hydroxybenzoic. Furthermore, the high abundance of 239, 29 and 179 (neutral losses of 6, 9 and 12 Da, respectively) indicates that the is esterified to the 7-O-, and not etherified on the 4- O-position. Thus, compound 2 is tentatively identified as p-hydroxybenzoyl. Compound 3, vanilloyl The MS 2 spectrum of the parental ion with 329 (exact : , corresponding with C 14 H 17 O 9 - ) that eluted at 5.9min in the chromatogram of ccr1 mutants, has many similarities with the MS 2 spectrum of compound 2, p-hydroxybenzoyl: neutral losses that are typical for hexosylated molecules (6, 9, 12 and 162 Da) result in high abundant second generation ions. The additional mass of 3 Da of compound 3 as compared to compound 2, and several fragments that result from a neutral loss of 15 Da (-CH 3 ), suggest the presence of an methoxyl group. Taking into account the biological origin of the sample, a methoxylation on the ortho position of the phenolic function is the most plausible. In addition, the UV/Vis absorption spectrum at 5.9 min showed an absorption maximum around nm and a shoulder around3 nm (see insert). The UV/Vis absorption spectrum of vanillic (Sigma aldrich, Steinheim, Germany) had a major absorption peak at 26.6 nm and a smaller one at nm. The shift toward higher wavelengths of the absorption maxima of compound 3 as compared to vanillic further supports compound 3 to be an esterified vanillate. Consequently, this molecule is tentatively identified as vanilloyl, i.e., vanillic esterified to a sugar. Supplemental Figure 4 page 6/15 11

12 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc Compound 4, syringoyl The neutral losses in the MS 2 fragmentation of the molecule thateluted with 359 at 6.5 min in the chromatogram of ccr1 mutants, are similar to those of compound 2 and 3. The exact mass of , corresponding with a chemical formula ofc 15 H 19 O 1 -, and the neutral losses of two times 15 Da (indicative for two methoxy groups) further suggest this molecule to be the benzenoid syringoyl. This identity is supported by the UV/Vis absorption spectrum at 6.5 min with an absorption maximum at nm (see insert). This maximum is about 5 nm higher compared to the UV/Vis absorption maximum of syringic (274.7 nm; Sigma aldrich, Steinheim, Germany). The shift toward a higher wavelengths of the absorption maxima of compound 4 as compared to syringic further supports compound 3 to be syringoyl (see also compound 1 and 3). Compound 5, cinnamic A compound with molecular mass of 148 Da eluted at 16.4 min in the chromatograms of c4h mutants. The exact (C 9 H 7 O 2 - ) hinted the compound to be cinnamic. The identity was further authenticated by the standard (The British drug houses Ltd, Poole, England), that eluted at the same RT with the same UV/Vis absorption spectrum. Notably, MS 2 (of both the compound in the chromatogram as the standard) resulted in a low abundance of the unfragmented parental ion, but did not result in reproducible fragments. However, in one occasion, a neutral loss of 44 Da (CO 2 ) was observed, which is in line with the presence of a carboxylic function. Compound 6, p-coumaric A compound with 163 eluted at 7.7 min in the chromatograms of 4cl1 mutants. The MS 2 showed only one reproducible fragment: a neutral loss of 44 Da (CO 2 ), resulting in 119. The UV/Vis absorption spectrum at 7.7 min had an absorption maximum at 39 nm. A standard (Sigma aldrich, Steinheim, Germany) proved this compound to be p-coumaric. Compound 7, 5-hydroxy At 3. min, a compound with 29 appeared in the chromatograms of comt mutants. The MS 2 showed neutral losses of -15 (CH 3 ), -44 (CO 2 ). The former neutral loss hints the presence of a methoxy group, while the latter is typically derived from a carboxylic function. The exact of , corresponding with a chemical formula of C 1 H 9 O 5 -, further suggest compound 5 to be 5-hydroxyferulic. Compound 8, cinnamoyl Compound 8 eluted with 369 at 15.6 and 16.7 min in the chromatograms of c4h mutants. A neutral loss of 6 Da in the MS 2 spectrum (leading to the base peak of 39), showed that the parental ion is the acetate adduct of a molecule with molecular mass of 31 Da. Further neutral loss of 162 Da, indicate the molecule to be hexosylated. The exact of the ions were , and which corresponds to ions with chemical formulas of C 17 H 21 O - - 9, C 15 H 17 O 7 and C 9 H 7 O - 2, Supplemental Figure 4 page 7/15 12

13 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc respectively. In addition, the UV/Vis absorption spectrum at 15.6 min had an absorption maximum at nm (see insert). This maximum is 13 nm higher as compared to the UV/Vis absorption maximum of cinnamic (compound 5; The British drug houses Ltd, Poole, England). A shift toward a higher wavelength of the absorption maxima of compound 8 as compared to sinapic further hints compound 8 to be a cinnamic ester. Therefore this compound is identified as cinnamoyl. Compound 9, p-coumaroyl A compound with 325 eluted at 8.3 and 9.2 min in the chromatograms of 4cl1 mutants. The neutral losses of 6, 9, 12, 138 (=12+18), 162, 18 and 28 (= ) Da are typical for p-hydroxy phenylpropanoic s esterified to a sugar (see Dauwe et al. (27) for feruloyl and sinapoyl and Vanholme et al. (21) for 5-hydroxyferuloyl ). The UV/Vis absorption spectrum at 8.3 min had an absorption maximum at nm (see insert) and is indicative for an esterified p-coumaric (compound 6). The exact of compound 9 is (C 15 H 17 O 8 - ), which further suggests this compound to be p-coumaroyl. Compound 1, diferuloyl Compound 1 eluted at 23.2 minin the chromatograms of ccr1 mutants, and had a 531. Although the abundance of the compound was relative low, the MS 2 spectrum clearly showed the presence of second generation ions with 337 and 193, and two ions with a lower abundance: 323 and 175. All of these fragments are exactly 3 Da (CH 2 O) lower in mass compared to the main fragments of compound 12, disinapoyl, whereas the parental ion is 6 Da (2 times CH 2 O) lower. This suggest compound 1 to be diferuloyl, which is further supported by the exact of the parental ion (C 26 H 27 O 12 - ) and in-source fragmentation ions (C 16 H 17 O 8 - ) and (C 1 H 9 O 4 - ). Compound 11, feruloyl sinapoyl Compound 11 eluted at 22.7 min with a 561 (exact , corresponding to C 27 H 29 O 13 - ) in the chromatograms of ccr1 mutants. The MS 2 of compound 11 showed many similarities with the MS 2 of compound 12, disinapoyl.a neutral loss of 224 led to the basepeak of 337 in the MS 2 spectrum, and could be assigned to the loss of sinapic. On the other hand, a neutral loss of 194 (resulting in second generation ion with 367) could be assigned to the loss of. Both the sinapic and moiety were detected as second generation ions with a 223 and 197, respectively. Taken together, compound 11 could be tentatively assigned as feruloyl sinapoyl. Compound 12, disinapoyl glucose The MS 2 spectrum of compound 12, eluting at 21.7 min with a 591 had a base peak of 367, corresponding with a neutral loss of 224 Da, that could be attributed to a loss of sinapic. Further neutral loss of 144, likely originated from a minus water (162 Da- 18 Da), gave riseto the ion with 223. The latter ion is probably a second sinapic moiety. The exact mass of the parental ion and fragments, as measured by in-source fragmentation in FT-ICR-MS, are , and Supplemental Figure 4 page 8/15 13

14 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc which corresponds to C 28 H 31 O 14 -, C 17 H 19 O 9 - and C 11 H 11 O 5 -, respectively.the UV/Vis absorption spectrum at 21.7 min had absorption maxima at nm and nm (see insert) which are indicative for an esterified sinapic (Sigma aldrich, Steinheim, Germany). Therefore, compound 12 is tentatively assigned as disinapoyl glucose, a known secondary metabolite of Arabidopsis (Fraser et al., 27). Compound 13, p-coumaroyl At 16.1, 17.9 and 19.7 minutes in the chromatograms of 4cl1 mutants, three compounds eluted with equal masses (48 Da) and equal MS 2 spectra. A neutral loss of 154 Da resulted in an ion of 325. Almost all smaller fragments arealso present in the MS 2 of compound 9 (i.e.,p-coumaroyl ), albeit in different ratios, the second generation ion with 371 is likely p-coumaroyl. The exact mass of the ions measured in the full MS chromatogram by FT-ICR-MS was , with best hit for the chemical formula of the neutral molecule C 23 H 28 O 11. Because the p-coumaroyl part of the molecule has a chemical formula of C 15 H 17 O 8, the unknown extra mass of 155 is C 8 H 11 O 3, a structure with 3 instaurations. The UV/Vis absorption spectrum at 16.1 min is near the detection limit, but had an absorption maxima at (see insert) that suggests compound 13 to be esterified p-coumarica cid (see compounds 6 and 9). As the structure of 155 could not yet completelybe resolved, the identity of compound 13 is described as p-coumaroyl. Compound 14, feruloyl Another 155-adduct eluted as 59 at 17.3, 19. and 19.8 min in the chromatograms of ccr1 mutants. The MS 2 spectrum has many similarities to p coumaroyl (compound 13), sinapoyl (15) and 5-hydroxyferuloyl (Vanholme et al., 211). Here, the phenylpropanoid part could be assigned to ( 193). The accurate corresponds to the chemical formula for the neutral molecule C 24 H 3 O 12 ) and confirmed the 155-part to be C 8 H 11 O 3. Therefore, compound 14 was named feruloyl. Compound 15, sinapoyl Compound 15 eluted with a molecular mass of 54 Da at a retention time of 16.7 and 19.1 min. The exact of , as measured via by FT-ICR-MS, hint C 25 H 32 O 13 to be the chemical formula of the molecule. The MS 2 spectrum showed a neutral loss of 154 Da, resulting in 385, the mass of sinapoyl glucose. Also further fragments in the MS 2 spectrum are the same as those in the MS 2 of sinapoyl glucose (Dauwe et al., 27). Therefore, compound 14 was partly identified as sinapoyl. As in the case of compounds 13 and 14 (p coumaroyl and feruloyl, respectively) the 155-part could be assigned to C 8 H 11 O 3 (i.e., C 25 H 32 O 13 C 17 H 21 O 1 ). Compound 16, cinnamoyl The MS 2 spectrum of compound 16 ( 263, retention time = 7.8 min) that eluted in the chromatograms of c4h mutants, showed a base peak at 147. The neutral loss of 116 Da leading to this base peak is typical for the loss of malic that is esterified via the 2-O ic function, as in sinapoyl and feruloyl (Mir Derikvand et al., 28). Further neutral loss of 44 Da (CO 2 ) and Supplemental Figure 4 page 9/15 14

15 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc the exact (C 13 H 11 O 6 - ), hint compound 16 to be cinnamoyl. The identity is further supported by the UV/Vis absorption spectrum at 7.8 min that had an absorption maximum at 28.8 nm (see insert) and is indicative for an esterified cinnamic (compounds 5). Note that, since the phenolic function is absent in cinnamoyl (compound 16), no 133 can be seen in the corresponding MS 2 spectrum. Compound 17, p-coumaroyl Another malic conjugated phenylpropanoic eluted at 3.6 min in the chromatograms of 4cl1 mutants. The parental ion of 279 gave rise to 163 upon elimination of malic (-116 Da). Further decarboxylation (-44 Da) resulted in 119. The is expelled (as an ion) onto delocalisation of the negative charge on the phenolic functionas 133. The exact mass of 17 is (C 13 H 11 O 7 - ) and that of the fragment (C 9 H 7 O 3 - ). Taken together, this compound is tentatively identified as p-coumaroyl. Compound 18, caffeyl 3/ The compound with a retention time of 5.7 and 7. min and a of 327in the chromatogram of ccoaomt1 mutants, dissociated to the base peak at 165, indicating the presence of a ( Da). The daughter ion at 147 is due to a water loss from the aglycone. The exact of the parental - ion ( ) and the fragment ( ) led to the chemical formula of C 15 H 19 O 8 and C 9 H 9 O - 3, respectively. Given the chemical formula and the biological origin, the aglycone is thus most likely caffeyl. Based on the MS 2 spectrum, the position of the is ambiguous, but the biological nature of the sample suggest the to be on the 4-O-position (as in coniferin and syringin, compounds 19 and 2, respectively), in at least one of the retention times (i.e., 5.7 or 7. min). However, it must be noted that 3-O-hexosylated caffeic was unambiguously determined by NMR to be present in CCoAOMT-downregulated poplar (Meyermans et al., 2) and that a glucosyltransferase in Arabidopsis has been described to specific transfer a to the 3-hydroxyl position of caffeic (Lim et al., 23). In conclusion, compound 18 is likely caffeyl 3/. Compound 19, coniferin Compound 19 eluted as an acetic adduct with 41 at 7. min. After loss of acetic (- 6 Da), the neutral loss of 162 Da () is followed by neutral losses of 15 Da (CH 3 ) and 18 Da (H 2 O). Exact masses were (C 18 H 25 O 1 - ), (C 16 H 21 O 8 - ) and (C 1 H 11 O 3 - ). The UV/Vis spectrum at 7. min showed an absorption maximum at 257 nm and a shoulder around 29 nm. Both mass-spectral and UV/Vis data of compound 19 matched the standard of coniferin, obtained from Sally Ralph and Professor Noritsugu Terashima (Terashima et al., 1995). Coniferin is a known metabolite of Arabidopsis. Note that the compound is eluting as an acetate adduct because it lacks ic hydrogens (e.g. as in phenolic or carboxylic functions). Compound 2, syringin Supplemental Figure 4 page 1/15 15

16 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc The peak with retention time of 8. min and 431 was low in abundance. However, the similarity of its MS 2 spectrum with that of coniferin (compound 19) and the slight delay in retention time in comparison in coniferin, hint compound 2 to be syringin, i.e., sinapyl. The identity was further confirmed via the match of spectral data of compound 2 with the standard syringing obtained from Sally Ralph and Professor Noritsugu Terashima (Terashima et al., 1995). Compound 21, At 9.5 min, a compound of 4 Da eluted. The MS 2 spectrum was similar to that of compound 19 except that all ions were observed at an value that is 2 units lower. Therefore, compound 21 is tentatively identified as. Compound 22, G(8 5)feruloyl At 17.4 min, a hexosylated compound elutes with 533. The aglycone ( 371) further showed losses of -18 (H 2 O) and -3 (CH 2 O) which are the typical small neutral losses for a phenylcoumaran (Morreel et al., 21b). The neutral loss of 44 Da (CO 2 ) might hint that the carboxylic is free, and thus not esterified to the. However, in this compound the loss of 44 Da most likely derive from a charge remote rearrangement, as is observed in compound 23 (G(8 O 4)caffeoyl, where the is most likely on the carboxylic ). Furthermore, the high intesity of the ion resulting from a consecutive loss of 12 and 18 Da ( 395) is also only observed in case when the is esterified (e.g. in compound 9 and not in32). Therefore, compound 22 is tentatively identified as G(8 5)feruloyl. Compound 23, G(8 O 4)caffeoyl A -bearing compound with molecular weights of 52 Da appeared at 17.4 in the chromatogram of ccoaomt1 mutants. The exact mass was , which results in the predicted chemical formula of C 25 H 27 O The similarity between the MS 2 spectrum of compound 23 and that of the benzodioxane G(8 O 4)5-hydroxyferuloyl (Vanholme et al., 21) was noted. The loss of in the MS/MS spectrum delivered the base peak of 457. Additional losses of 6, 9 and 12 Da and the decarboxylation of the aglycone, leading to the ions at 459, 429, 399 and 313, respectively, indicated an acyl moiety. The second generation ions at 339 and 341 arise from the B and (B+2H) moieties after retro cleavage of the benzodioxane ring (Morreel et al., 21b). Further neutral loss 6, 9 and 162 Da from B ( 339) again shows that is attached on the B part of the compound. The fragment A with 179 suggest the A to be derived from coniferyl. However, 179 might in part also be derived from a loss from (B+2H) ( 341). Taken together, compound 23 is the tentatively identified as the benzodioxane G(8 O 4)caffeoyl. Compound 24, G 4-O-glucoside(8 5)G, DCG An acetate adduct of a compound with mass 52 Da eluted at 15.2 min. The MS 2 showed a neutral loss of 162 Da, which indicate the compound to be hexosylated. The subsequent loss of 18 (H 2 O) and 3 (CH 2 O) from the aglycon is typical for a phenylcoumaran structure, whereas the cleavage of the phenylcoumarin bonding structure yields the 1,2 B ion with 221 (Morreel et al., 21b). The latter is Supplemental Figure 4 page 11/15 16

17 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc derived from a coniferyl coupled via the 5-C and 4-O position. Taking into account the molecular mass and the UV/Vis absorption spectrum at 15.2 min, the aglycone must be G(8 5)G (the standard of G(8-5)G has absorption maxima at and nm (Morreel et al., 24)). The occurrence of the compound as an acetate adduct proves the absence of an ic phenolic function (see e.g. compound19). Therefore, the must be linked at the phenolic function. Therefore, the identity of compound 24 is tentatively resolved as G 4-O-glucoside(8 5)G, a metabolite of Arabidopsis also known as dehydrodiconiferyl glucoside (DCG). Compound 25, G 4-O-glucoside(red8 5)G, IDDDC glucoside Another acetate adduct of a hexosylated compound eluted at 15.4 min with a of 581. The UV/Vis absorption spectrum at 15.4 min showed an absorption maximum at nm. The loss of 6 Da (acetate) and Da (acetate and ) resulted in the aglycone that further fragmented as isodihydrodehydrodiconiferyl (IDDDC, G(red8-5)G) (Morreel et al., 21b). The exact of the molecule and the aglycone was (C 26 H 33 O 11 - ) and , (C 2 H 23 O 6 - ), respectively. The position of the could not exactly be determined,but given the biological nature of the sample and the presence of compound 24, G 4-O-glucoside(8 5)G, in the same samples,the most likely position is the 4-O-position of the 8-coupled unit (unit A). Therefore, compound 25 is called G 4-O-glucoside(red8 5)G. It must be noted that the proposed structure has a free phenolic function, which is not expected to result in the formation of an acetate adduct. Compound 26, G(8 O 4)caffeyl hexoside A compound eluted with 35 and retention time of 11.8 min in the chromatogram of ccoaomt1 mutants. Based on MS 2 similarity with G(8 O 4)5H hexoside (Vanholme et al, 21), the compound was identified as a hexosylated benzodioxane structure.fragmentation of the benzodioxane ring led to second generation daughter ions at 179 and 165 (A - and (B+2H) - ions, respectively) (Morreel et al., 21b). The loss of formadehyde (-3) from the parental ion indicates that the is not on the 9-O-position of the A-unit, but the position of the could further not be resolved. The position of the on the 4-O-position is suggested by the following observations: 1) the peak at 325, most likely derive from loss of ( Da) followed by water (-18), and not from 9-Ohexosylated B - ion as the loss of the latter would result in a peak at 163, which is below detection limit. Moreover, 2) the rather low intensity of the A - fragment ( 179) is atypical for benzodioxane structures with free fenolic function. On the other hand, following observations plaid against a 4-O-hexosylation but rather a 9-O-hexosylation on the B-unit: 1) With the absence of a phenolic function the molecule would most likely elute as the acetate adduct (see e.g. compound 9). Furthermore, 2) the loss of water and formaldehyde (-18 Da and -3 Da respectively) results most likely via the charge driven electron delocalistion of the free phenolic function (Morreel et al., 21a, b). Therefore, compound 26 is called G(8 O 4)caffeyl hexoside, with the on the 4-O-G or 9- O-caffeyl position (i.e., 4-O-hexosyl G(8 O 4)caffeyl or G(8 O 4) 9-O-hexosyl caffeyl ). Compound 27, G 7/9-O-hexoside(8 O 4)G Supplemental Figure 4 page 12/15 17

18 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc A hexosylated dilignol eluted at 8.2 min with a 537. The aglycone fragmentation (-18, -3 and -48 Da) was reminiscent to a 8 O 4 coupling, whereas the of the aglycone (375) further suggested the aglycone to be G(8 O 4)G. The position of the was not completely resolved, but the absence of neutral losses -18 (M-H + -H 2 O) -, -3 (M-H + - CH 2 O) - and the combination of -18 and -3 (M- H + -H 2 O-CH 2 O) - from the parental ion (as compared to compound 29) suggest the to be on the 8- coupled G-unit. Because a on the 4-O-position is expected to result in the elution of the acetate adduct (as in compound 9), the is most likely on the 7-O or 9-O-position of the 8-coupled G-unit. Compound 27is therefore called G 7/9-O-hexoside (8 O 4)G. Compound 28, G (8 O 4)G Another hexosylated G(8 O 4)G eluted at 1.3 min with 597. The mass 597 was not fragmented but a co-eluting structure of 375, derived from a neutral loss of 162 () and 6 (acetic ) showed that 597 was hexosylated. Further fragmentation of 375 was typical for G(8-O- 4)G (Morreel et al., 21b). Also the UV/Vis absorption spectrum with absorption maximum at hints the aglycon to be G(8 O 4)G (Morreel et al., 24). The high abundance of the combined water (- 18) and formaldehyde (-3) loss, suggest the (8 O 4) structure to be in the threo-position. As the molecule elutes as an acetate adduct, the is most likely coupled at the phenolic function (see, e.g., compound 9). The tentative identity of compound 28 is thus G (8 O 4)G. Compound 29, G(8 O 4)G 9-O-hexoside A third hexosylated G eluted at 15.6 ( 537). The structure of the aglycone is resolved as G according known fragmentation rules (Morreel et al., 21a; b). In contrast to compound 28, the relative low abundance of a combined water (-18) and formaldehyde (-3) loss, suggest the (8-O- 4) structure to be in the erythro-position. The must be on the 9-O-position of the O-4 coupled coniferyl because the phenolic function and the two aliphatic functions on the O-4 coupled coniferyl are needed during the electron delocalization that lead to the loss of water and formaldehyde. Therefore, compound 29 is tentatively identified as G(8 O 4)G 9-O-hexoside. Compound 3, G (8 O 4)S(8 5)G At 19.2 min a compound with 85 eluted. At the same retention time a compound with 583 is found. The mass difference between both compounds (222 Da) is typically the difference of a (162 Da) and acetic (6 Da). Therefore, compound 3 must be hexosylated. The acetate adduct suggest the compound to have no free acedic hydrogens (no phenolicnor carboxylic functions). As the MS 2 spectrum of the aglycone ( 583) is identical to the MS 2 of G(8 O 4)S(8 5)G (Morreel et al. 21a), compound 3 is tentatively identified as G (8 O 4)S(8 5)G. Compound 31, p-coumaric A compound with 325 eluted at 3.6 minutes in the chromatograms of 4cl1 mutants.although no MS 2 spectrum was available, in-source fragmentation led to the neural loss of 162 Da, resulting in the co-elution of 163. As a loss of 162 is the base peak in 4-O-hexosylated phenylpropanoic s (see Supplemental Figure 4 page 13/15 18

19 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc e.g. (Dauwe et al., 27), sinapic (compound 32)) and given the biological origin of the samples, compound 31 is most likely p-coumaric. This identity is further supported by the UV/Vis absorption spectrum at 3.6 min that showed an absorption maximum at 39.4 nm and is comparable with the UV/Vis absorption spectrum of p-coumaric (compound 6). Compound 32, dihydro-p-coumaric + At 4.8 min a compound with 327 eluted in the chromatograms of 4cl1 mutants. The base peak in the MS 2 spectrum ( 165) is the result of a loss.further neutral loss of 44 Da (CO 2 ) suggests the aglycone to have a carboxylic function. The exact of the parental ion ( ) and the aglycone ( ), led to the prediction of the formulasof C 15 H 19 O 8 - and C 9 H 9 O 3 -, respectively. These correspond with the composition of a hexosylated dihydro-p-coumaric. Two possibilities for the position of the are plausible, i.e., esterified on the 9-O-position or etherified on the 4-Oposition. Therefore, compound 32 was called dihydro-p-coumaric +. Compound 33, dihydrocaffeic + Compound 34, eluted at 4.6 min with 343. The exact mass was (C 15 H 19 O 9 - ). The reasoning for the identification was similar to that of compound 32 (dihydro-p-coumaric 4-Ohexoside). The fragmentation of the aglycone similar to that of the chemical standard dihydrocaffeic (Sigma-aldrich, Steinheim, Germany) and led to the partial elucidation of compound 34 to be dihydrocaffeic +, with on the 9-O-, 4-O- or 3-O- position. Compound 34, dihydro+ A third hexosylated dihydrophenylpropanoic eluted at 5.7 min with a 357. The exact mass was determined to be (C 16 H 21 O 9 - ). The UV/Vis absorption spectrum at 5.7 min showed a maximum at nm. The reasoning for the structure of compound 35 was again similar to that of compound 32 (dihydro-p-coumaric ), and led to the tentative identification of compound 35 to be dihydro+hexoside, with on the 9-O- or 4-O- position. References: Dauwe, R., Morreel, K., Goeminne, G., Gielen, B., Rohde, A., Van Beeumen, J., Ralph, J., Boudet, A.-M., Kopka, J., Rochange, S.F., Halpin, C., Messens, E., and Boerjan, W. (27). Molecular phenotyping of lignin-modified tobacco reveals associated changes in cell wall metabolism, primary metabolism, stress metabolism and photorespiration. Plant J. 52, Fraser, C.M., Thompson, M.G., Shirley, A.M., Ralph, J., Schoenherr, J.A., Sinlapadech, T., Hall, M.C., and Chapple, C. (27). Related Arabidopsis serine carboxypeptidase-like sinapoylglucose acyltransferases display distinct but overlapping substrate specificities. Plant Physiol. 144, Lim, E.K., Higgins, G.S., Li, Y., and Bowles, D.J. (23). Regioselectivity of glucoylation of caffeic acd by a UDPglucose:glucosyltransferase is maintained in planta. Biochem. J. 373, Meyermans, H., Morreel, K., Lapierre, C., Pollet, B., De Bruyn, A., Busson, R., Herdewijn, P., Bart Devreese, B., Van Beeumen, J., Marita, J.M., Ralph, J., Chen, C., Burggraeve, B., Van Montagu, M., Messens, E., and Boerjan W., (2) Modifications in Lignin and Accumulation of Phenolic Glucosides in Poplar Xylem upon Down- Supplemental Figure 4 page 14/15 19

20 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc regulation of Caffeoyl-Coenzyme A O-Methyltransferase, an Enzyme Involved in Lignin Biosynthesis. J. Biol. Chem. 275, Mir Derikvand, M., Berrio Sierra, J., Ruel, K., Pollet, B., Do, C.-T., Thévenin, J., Buffard, D., Jouanin, L., and Lapierre, C. (28). Redirection of the phenylpropanoid pathway to feruloyl in Arabidopsis mutants deficient for cinnamoyl CoA reductase 1. Planta 227, Morreel, K., Ralph, J., Kim, H., Lu, F., Goeminne, G., Ralph, S., Messens, E. andboerjan, W. (24), Profiling of Oligolignols Reveals Monolignol Coupling Conditions in Lignifying Poplar Xylem. Plant Physiol. 136, Morreel, K., Dima, O., Kim, H., Lu, F., Niculaes, C., Vanholme, R., Dauwe, R., Goeminne, G., Inzé, D., Messens, E., Ralph, J., and Boerjan, W. (21a). Mass spectrometry-based sequencing of lignin oligomers. Plant Physiol. 153, Morreel, K., Kim, H., Lu, F.C., Dima, O., Akiyama, T., Vanholme, R., Niculaes, C., Goeminne, G., Inzé, D., Messens, E., Ralph, J., and Boerjan, W. (21b). Mass spectrometry-based fragmentation as an identification tool in lignomics. Anal. Chem. 82, Terashima, N., Ralph S.A. and Landucci L.L. (1995). New facile syntheses of monolignol glucosides; p-glucocoumaryl, coniferin and syringin. Holzforschung, 1995, 5, Vanholme, R., Ralph, J., Akiyama, T., Lu, F., RencoretPazo, J., Kim, H., Christensen, J.H., Van Reusel, B., Storme, V., De Rycke, R., Rohde, A., Morreel, K., and Boerjan, W. (21c). Engineering traditional monolignols out of lignin by concomitant up-regulation of F5H1 and downregulation of in Arabidopsis. Plant J. 64, Supplemental Figure 4 page 15/15 2

21 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc A erythrose-4-phosphate + phospoenol pyruvate gallic DHS DQS DHQD/SD SK EPSPS CS shikimate chorismate CM ICS salicylic glucoside pal1 P PAI IGPS TSA TSB tryptophan AD tyrosine phytoalexins indolacetic PD indolic glucosinolates phenylalanine benzoic benzoyl benzyl PAL1 PAL2 PAL3 PAL4 C4H cinnamic isorhamnetin glycosides isorhamnetin CHS CHI F3H kaempferol glycosides kaempferol dihydrokaempferol quercetin glycosides caffeic 5-hydroxy ADT PDT cinnamoyl p -coumaroyl cinnamoyl p-coumaroyl caffeoyl F3H quercetin dihydroquercetin anthocyanidins and proanthocyanidins diferuloyl feruloyl HHT suberin-esterified F6 H1 coniferin caffeyl feruloyl p-coumaric p-coumaric caffeic 3/ 4CL1 4CL2 4CL3 4CL4 p-coumaroyl-coa HCT C3H1 HCT caffeoyl-coa CCoAOMT1 feruloyl-coa p-hydroxy benzoyl s caffeyl 3/ dihydrocaffeic + disinapoyl glucose SMT SST feruloyl 5-hydroxyferuloyl sinapoyl glucose 5-hydroxyferuloyl sinapoyl SGT CCR1 CCR2 scopolin caffealdehyde 5-hydroxyconiferin dihydroferulic + 5-hydroxyferulic 4/5-O-hexoside vanilloyl F5H1 feruloyl sinapoyl 5-hydroxy 5-hydroxy 5-hydroxyconiferyl sinapic syringoyl syringin sinapoyl sinapic sinapaldehyde coniferyl sinapyl G(8-O-4)caffeoyl G(8-5)feruloyl G(8-O-4) G(8-O-4)5-hydroxy G(8-O-4)caffeyl hexoside (8-5)G (red8-5)g G 9-O-hexoside G - G 7/9-O-hexoside- G - (8-O-4)S(8-5)G G-oligolignols G+5H-mixed 5H-oligolignols lignin G+S-mixed S-oligolignols B erythrose-4-phosphate + phospoenol pyruvate gallic DHS DQS DHQD/SD SK EPSPS CS shikimate chorismate CM ICS salicylic glucoside pal2 P PAI IGPS TSA TSB tryptophan AD tyrosine phytoalexins indolacetic PD indolic glucosinolates phenylalanine benzoic benzoyl PAL1 PAL2 PAL3 PAL4 C4H cinnamic isorhamnetin glycosides isorhamnetin CHS CHI F3H kaempferol glycosides kaempferol dihydrokaempferol quercetin glycosides caffeic 5-hydroxy ADT PDT benzyl cinnamoyl p -coumaroyl cinnamoyl p-coumaroyl caffeoyl F3H quercetin dihydroquercetin anthocyanidins and proanthocyanidins p-coumaroyl dihydrop-coumaric + caffeic diferuloyl feruloyl HHT suberin-esterified F6 H1 coniferin caffeyl feruloyl p-coumaric p-coumaric caffeic 3/ 4CL1 4CL2 4CL3 4CL4 p-coumaroyl-coa HCT C3H1 HCT caffeoyl-coa CCoAOMT1 feruloyl-coa p-hydroxy benzoyl s caffeyl 3/ dihydrocaffeic + disinapoyl glucose SMT SST feruloyl 5-hydroxyferuloyl sinapoyl glucose 5-hydroxyferuloyl sinapoyl SGT CCR1 CCR2 scopolin caffealdehyde 5-hydroxyconiferin dihydroferulic + p-coumaroyl dihydrop-coumaric + caffeic 5-hydroxyferulic 4/5-O-hexoside vanilloyl F5H1 feruloyl sinapoyl 5-hydroxy 5-hydroxy 5-hydroxyconiferyl sinapic syringoyl syringin sinapoyl sinapic sinapaldehyde coniferyl sinapyl G(8-O-4)caffeoyl G(8-5)feruloyl G(8-O-4) G(8-O-4)5-hydroxy G(8-O-4)caffeyl hexoside (8-5)G (red8-5)g G 9-O-hexoside G - G 7/9-O-hexoside- G - (8-O-4)S(8-5)G G-oligolignols 5H-oligolignols G+5H-mixed G+S-mixed lignin S-oligolignols Supplemental Figure 5 page 1/7 21

22 Supplemental Data. Vanholme et al. Plant Cell. (212) /tpc C erythrose-4-phosphate + phospoenol pyruvate gallic DHS DQS DHQD/SD SK EPSPS CS shikimate chorismate CM ICS salicylic glucoside c4h P PAI IGPS TSA TSB tryptophan AD tyrosine phytoalexins indolacetic PD indolic glucosinolates phenylalanine benzoic benzoyl PAL1 PAL2 PAL3 PAL4 C4H cinnamic CHS CHI F3H kaempferol glycosides kaempferol dihydrokaempferol quercetin glycosides caffeic 5-hydroxy ADT PDT benzyl isorhamnetin glycosides isorhamnetin cinnamoyl p -coumaroyl cinnamoyl p-coumaroyl caffeoyl F3H quercetin dihydroquercetin anthocyanidins and proanthocyanidins p-coumaroyl dihydrop-coumaric + caffeic diferuloyl feruloyl HHT suberin-esterified F6 H1 feruloyl p-coumaric p-coumaric caffeic 3/ 4CL1 4CL2 4CL3 4CL4 HCT C3H1 HCT p-coumaroyl-coa caffeoyl-coa CCoAOMT1 feruloyl-coa p-hydroxy benzoyl s caffeyl 3/ dihydrocaffeic + CCR1 CCR2 scopolin caffealdehyde 5-hydroxyconiferin coniferin caffeyl disinapoyl glucose SMT SST feruloyl 5-hydroxyferuloyl sinapoyl glucose 5-hydroxyferuloyl sinapoyl SGT dihydroferulic + p-coumaroyl dihydrop-coumaric + caffeic 5-hydroxyferulic 4/5-O-hexoside vanilloyl F5H1 feruloyl sinapoyl 5-hydroxy 5-hydroxy 5-hydroxyconiferyl sinapic syringoyl syringin sinapoyl sinapic sinapaldehyde coniferyl sinapyl G(8-O-4)caffeoyl G(8-5)feruloyl G(8-O-4) G(8-O-4)5-hydroxy G(8-O-4)caffeyl hexoside (8-5)G (red8-5)g G 9-O-hexoside G - G 7/9-O-hexoside- G - (8-O-4)S(8-5)G G-oligolignols G+5H-mixed 5H-oligolignols lignin G+S-mixed S-oligolignols D erythrose-4-phosphate + phospoenol pyruvate gallic DHS DQS DHQD/SD SK EPSPS CS shikimate chorismate CM ICS salicylic glucoside 4cl1 P PAI IGPS TSA TSB tryptophan AD tyrosine phytoalexins indolacetic PDT indolic glucosinolates phenylalanine benzoic benzoyl benzyl PAL1 PAL2 PAL3 PAL4 C4H cinnamic isorhamnetin glycosides isorhamnetin CHS CHI F3H kaempferol glycosides kaempferol dihydrokaempferol quercetin glycosides caffeic 5-hydroxy ADT PDT cinnamoyl p -coumaroyl cinnamoyl p-coumaroyl caffeoyl F3H quercetin dihydroquercetin anthocyanidins and proanthocyanidins diferuloyl feruloyl HHT suberin-esterified F6 H1 coniferin caffeyl feruloyl p-coumaric p-coumaric caffeic 3/ 4CL1 4CL2 4CL3 4CL4 p-coumaroyl-coa HCT C3H1 HCT caffeoyl-coa CCoAOMT1 feruloyl-coa p-hydroxy benzoyl s caffeyl 3/ dihydrocaffeic + disinapoyl glucose SMT SST feruloyl 5-hydroxyferuloyl sinapoyl glucose 5-hydroxyferuloyl sinapoyl SGT CCR1 CCR2 scopolin caffealdehyde 5-hydroxyconiferin dihydroferulic + 5-hydroxyferulic 4/5-O-hexoside vanilloyl F5H1 feruloyl sinapoyl 5-hydroxy 5-hydroxy 5-hydroxyconiferyl sinapic syringoyl syringin sinapoyl sinapic sinapaldehyde coniferyl sinapyl G(8-O-4)caffeoyl G(8-5)feruloyl G(8-O-4) G(8-O-4)5-hydroxy G(8-O-4)caffeyl hexoside (8-5)G (red8-5)g G 9-O-hexoside G - G 7/9-O-hexoside- G - (8-O-4)S(8-5)G G-oligolignols G+5H-mixed 5H-oligolignols lignin G+S-mixed S-oligolignols Supplemental Figure 5 page 2/7 22