Sulfur fertilization influences the sulphur species composition in Allium sativum: sulfomics using HPLC-ICP-MS/MS-ESI-MS/MS

Transcription

1 Electronic upplementary Material (EI) for Metallomics. This journal is The Royal ociety of Chemistry 2017 Electronic supplement for ulfur fertilization influences the sulphur species composition in Allium sativum: sulfomics using HPLC-ICP-M/M-EI-M/M Andrea Raab, Marilena Ronzan, Joerg Feldmann TELA (Trace Element peciation Laboratory), University of Aberdeen, Chemistry, Meston Walk, Aberdeen, AB243UE, cotland, UK Page number: Hoagland solution 2. Details of compounds (M spectra, EIC+ICP-M/M trace, fragmentation by M/M) Figures Cycloalliin, aliin & isoalliin 2.2. Methiin, propiin and methionine allyl-cysteine (AC, deoxyalliin) and derivatives 2.4. γ-glutamyl--allyl-cysteine (GA and γ-glutamyl--1-propenylcysteine (GP 2.5. γ-glutamyl--2-propenylcysteine sulfoxide GAC() and γ- glutamyl--1-propenylcysteine sulfoxide GPC() 2.6. γ-glutamyl--methyl-cysteine (GM and γ-glutamylhomocysteine (C264) 2.7. Glutathione (GH and GG) propenylmercaptoglutathione, -2- propenylmercaptoglutathione and -propylmercaptoglutathione (C379) 2.9. xidised forms of -1-propenylmercaptoglutathione and -2- propenylmercaptoglutathione (C395) Phytochelatin 2 (PC-2) and derivatives Precursor molecules for -alk(en)ylcysteine sulfoxides Amount of sulphur associated with non-resolved chromatographic peaks 3. Results from principal component analysis and Cluster analysis 3.1. Loading plot (PCA) 3.2. Dendrogram 4. Example of ICP-M/M traces for all plant parts fertilized with different amount of sulphur 5. Example of peak area integration 1 P age

2 1. Composition of Hoagland solution ulphur level 0.5 mm: 1.25 mm KN 3, 1.50 mm Ca(N 3 ) 2, 0.75 mm Mg 4, 0.50 mm KH 2 P 4, 50 µm KCl, 50 µm H 3 B 3, 10 µm Mn 4, 2.0 µm Zn 4, 1.5 µm Cu 4, 0.5 µm Na 2 Mo 4.2H 2 0, 0.1 mm Na 2 3 i, and 72 µm Fe- EDTA ph 6.0 ulphur level 0.1 mm: Plant group exposed to 0.1 M sulphur: Mg 4 was replaced by 0.75 mm MgCl 2 ulphur level 2 mm: Plant group exposed to 2 mm sulphur: Mg 4 2 mm Mg 4 was used instead of 0.75 mm 2 P age

3 2. Details of compounds (EI-M spectra, EIC+ICP-M/M trace, fragmentation by EI-M/M) Figures 1-26 The following tables summarize the molecular structure, molecular weight, theoretical and measured [M+H] + and retention time of the compounds and contain the average compound or sulphur concentrations Cycloalliin, aliin & isoalliin The identities of cyclo-alliin, alliin and isoalliin were deduced by comparison of their q-tf-ei-m/m spectra. E-M/M spectra of the q-tf instrument for alliin contained predominantly signals at m/z (C 3 H 7 N 3 ) and m/z (C 3 H 6 N 2 ) (Fig. 2). Isoalliin (Fig. 3) showed additionally a fragment at m/z identical to one of the major fragments of cyclo-alliin (C 5 H 8 N). The different position of the double bond compared to alliin allows stabilisation of isoalliinfragments by cyclisation. The other major fragments of cyclo-alliin were m/z (loss of water) and m/z (C 5 H 7 2 ) (Fig. 1). Table 1 Alliin and isoalliin content in root and bulb in mmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), cylcoalliin; in brackets compound average in % of total sulphur, database ID: Chempider. H 3 C Cycloalliin* Alliin Isoalliin N H H C H 2 Molecular formula C 6 H 11 N 3 C 6 H 11 N 3 C 6 H 11 N 3 MW (g mol -1 ) [M+H] + (theoretical) [M + H] + (measured) RT (min) Database ID Root (0.1 mm ) 14 ± 15 a (7.1 %) 4.1 ± 3.4 a (2.0 %) Root (0.5 mm ) 19 ± 16 b (6.7%) 4.3 ± 4.3 b (1.5 %) Root (2 mm ) 79 ± 41 a,b (17%) 30 ± 27 a,b (6.4%) Bulb (0.1 2 mm ) 21 ± 8.1 (16 %) 3.7 ± 2.1 (2.8 %) *cycloalliin + methiin + methylcysteine, amount of associated sulphur see table 13 a,b : statistically significant difference between groups by ne Way ANVA p < H C H 3 2 H 3 P age

4 H 3 C A) N H H C 6 H 11 N 3 relative intensity ICP-M/M m/z 178 relative intensity EI-M CH 3 2 -CH 4 N -H 2 Fig. 1 Cycloalliin; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 4 P age

5 A) N H 2 C 6 H 11 N 3 CH 2 H relative intensity ICP-M/M m/z relative intensity EI-M -C 3 H 5 -C 3 H 6-2 Fig. 2 Alliin; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 5 P age

6 A) N H 2 CH 3 H relative intensity ICP-M/M m/z 178 relative intensity EI-M C 6 H 11 N CH 3 2 -C 3 H C 3 H 6-2 Fig. 3 Isoalliin; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 6 P age

7 2.2. Methiin, propiin and methionine It was not possible to confirm the identity of methiin other than from the accurate m/z ratio (Δppm = 0.54). The signal was not intense enough in any of the samples for fragmentation by either rbitrap EI-M or q-tf-ei-m (Fig. 4). The retention time is similar to those published by others. Cycloalliin (Table 1) and methylcysteine co-eluted with methiin. Propiin (Fig. 5) eluted together with glutathione (Fig. 17) and methionine (Fig. 6). In addition to the accurate [M+H] +, propiin and methionine showed the expected fragmentation patterns, with propiin fragmenting to m/z (loss of CH 2 2 ), m/z (loss of CH 3 3 ) and m/z (loss of 2 and C 3 H 7 ) (Fig. 5). Methionine fragments of m/z (loss of C 2 ) m/z (loss of 2 ) (Fig. 6) confirmed its presence together with its accurate mass. Table 2 methiin propiin and methionine, database ID: Chempider. C H 3 Methiin* Propiin + Methionine + 2 C H 3 2 C H 3 H H Molecular formula C 4 H 9 N 3 C 6 H 13 N 3 C 5 H 11 N 2 MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) Root (0.5 mm ) Root (2 mm ) Bulb (0.1 2 mm ) *cycloalliin + methiin + methylcysteine, amount of associated sulphur see table 13 + propiin + glutathione + methionine, amount of associated sulphur see table 13 H 2 7 P age

8 A) C H 3 2 H C 4 H 9 N 3 relative intensity ICP-M/M m/z 152 relative intensity EI-M Fig. 4 Methiin; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel no fragmentation; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 8 P age

9 A) N H 2 C 6 H 13 N 3 CH 3 H relative intensity ICP-M/M m/z 180 relative intensity EI-M CH 2 2 -CH 3 3 -C 3 H Fig. 5 Propiin; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 9 P age

10 A) C 5 H 11 N 2 H H 2 N H 3 C relative intensity ICP-M/M m/z 150 relative intensity EI-M C 2-2 Fig. 6 Methionine; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 10 P age

11 2.3. -allyl-cysteine (AC, deoxyalliin) and derivatives Methyl-cysteine showed a fragment at m/z (loss of 2 ) (Fig. 7) and no other significant fragments due to settings of the mass range for both rbitrap EI- M and q-tf-ei-m. -allyl-cysteine showed fragments at m/z (amine loss), m/z (loss of C 2 ) and m/z (loss of 2 and C 2 ) (Fig. 8). AC was also identified in garlic by Yamazaki et al. 1 eluting after GMC identical to the retention behaviour under the conditions used here. -propyl-cysteine did not occur in either bulbs or roots. Table 3 -allyl-cysteine in bulb and root in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), methyl cysteine, database ID: Chempider. -allyl-cysteine -propyl-cysteine Methylcysteine* C H 2 2 H Molecular formula C 6 H 12 N 2 C 6 H 14 N 2 C 4 H 9 N 2 MW (g mol -1 ) [M + H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) 325 ± Root (0.5 mm ) 256 ± Root (2 mm ) 414 ± Bulb (0.1 2 mm ) 143 ± *cycloalliin + methiin + methylcysteine, amount of associated sulphur see table 13 C H 3 2 H C H 3 2 H 11 P age

12 A) C H 3 2 H C 4 H 9 N 2 relative intensity ICP-M/M m/z 136 relative intensity EI-M Fig. 7 Methyl-cysteine; panel A) structure; panel B: EIC of m/z + ICP- M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 12 P age

13 A) N H 2 H relative intensity ICP-M/M m/z 162 relative intensity EI-M CH 2 C 6 H 12 N C 2 -C 2 Fig. 8 -allyl-cysteine; panel A) structure; panel B: EIC of m/z + ICP- M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 13 P age

14 2.4. γ-glutamyl--allyl-cysteine (GA and γ-glutamyl--1- propenyl-cysteine (GP GAC and GPC were identified by their accurate m/z ratio and their fragmentation pattern. The assignment of the allyl respectively the propenyl-group to either compound was done based on comparison with published retention times, which all showed that compounds containing the allyl-group elute before the related compound containing a propenyl-group. The main fragment for both compounds was m/z (C 6 H 9 2 ) followed by m/z (C 6 H 12 N 2, loss of γ-glu) and m/z (C 5 H 8 N 3, loss of propenyl/allyl-cysteine) (Fig. 9 and 10). Identification and quantification of both compounds in bulbs was uncomplicated. Both compounds showed clear signals in the ICP-M/M trace and the EI-M extracted ion chromatogram (Fig. 11b). GAC in root was similarly clear to identify. GPC identification/quantification was more complicated, since the extracted ion chromatogram from roots showed two clearly separated signals at m/z (with the same fragmentation pattern) at RT 17.0 and 17.6 min overlapping one broad sulphur signal from the ICP-M/M (Fig. 10 and 11a). The extracted ion chromatogram of bulbs showed one dominant signal at 17.6 min associated with one sulphur signal from the ICP-M/M. It was not possible to identify any other coeluting sulphur compounds in root extracts and the existence of this double peak in the EIC remained unexplained (Fig. 11 comparison of bulb and root). Table 4 GAC and GPC in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), database ID: Chempider. GAC GPC H 2 C H H 3 C H Molecular formula C 11 H 19 N 2 5 C 11 H 19 N 2 5 MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) 70 ± ± 50 a Root (0.5 mm ) 61 ± ± 59 b Root (2 mm ) 78 ± ± 97 a,b Bulb (0.1-2 mm ) 40 ± ± 660 a,b : statistically significant difference between groups by ne Way ANVA p < H 2 H 14 P age

15 A) H 2 N H H relative intensity ICP-M/M m/z 291 relative intensity EI-M C 11 H 19 N CH 2 -C 5 H 9 N 2 3 -allyl/propenyl-cys -γ-glu Fig. 9 GAC; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 15 P age

16 A) H H 2 N H relative intensity ICP-M/M m/z 291 relative intensity EI-M C 11 H 19 N 2 5 CH C 5 H 9 N 2 3 -allyl/propenyl-cys -γ-glu CH 2-2 Fig. 10 GPC; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 16 P age

17 A) relative intensity ICP-M/M m/z 291 relative intensity EI-M m/z 291 relative intensity ICP-M/M relative intensity EI-M Fig. 11 Comparison of EIC (GAC and GP and ICP-M/M trace for root (panel A) and bulb (panel b). 17 P age

18 2.5. γ-glutamyl--2-propenylcysteine sulfoxide GAC() and γ- glutamyl--1-propenylcysteine sulfoxide GPC() The main fragments of oxidised GAC and GPC were m/z (γ-glu) and m/z (Fig. 12 and 13). Also the loss of γ-glu at m/z was present in both M spectra, careful comparison of the fragment spectra showed that m/z (loss of C 8 H 12 N 3 ), m/z (C 3 H 6 N 2 ) and m/z (loss of C 4 H 7 ) were more intense in GAC() than GPC() (Fig. 14). Table 5 GAC() and GPC() in root and bulb in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), database ID: Chempider. H 2 C GAC() H 3 C GPC() H H H H Molecular formula C 11 H 19 N 2 6 C 11 H 19 N 2 6 MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) 703 ± 398 a 836 ± 635 b Root (0.5 mm ) 1236 ± ± 1317 Root (2 mm ) 3065 ± 2471 a 1900 ± 856 b Bulb (0.1 2 mm ) 400 ± ± 241 a : statistically significant difference between groups by ne Way ANVA p < b : statistically significant difference between groups by ne Way ANVA p < P age

19 A) H 2 N H H relative intensity ICP-M/M m/z 307 relative intensity EI-M C 11 H 19 N CH 2 -C 6 H 10 N 3 -CH C 3 H 5 -C 3 H 5 -C 4 H 7 + H -C 3 H 5 Fig. 12 GAC(); panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 19 P age

20 A) H 2 N H C 11 H 19 N 2 6 H 32- m/z 307 relative intensity ICP-M/M relative intensity EI-M CH 3 -C 6 H 10 N 3 -CH C 3 H 5 -C 3 H 5 -C 3 H 5 Fig. 13 GPC(); panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 20 P age

21 A) Fig. 14 Details of q-tf EI-M/M spectra of GAC() and GPC(), main differences in occurring fragments marked by cycles. 21 P age

22 2.6. γ-glutamyl--methyl-cysteine (GM and γ-glutamylhomocysteine (C264) Three chromatographic peaks with the same molecular composition at m/z (GM were clearly detectable in the extracted ion chromatogram. The detailed study of the EI-M/M fragmentation patterns showed that one of the minor signals at 12.5 min was an in-source fragment of C393, whereas the signal at ca. 8.4 min was a genuine compound with the same m/z but a different fragmentation pattern not yet described in the literature (C264). GMC showed a main fragment at m/z (loss of C 5 H 9 N 2 3 ) (Fig. 15) and C264 s main fragment was m/z (loss C 5 H 8 N 3 ) (Fig. 16). To fragment this way C264 is likely to be γ-glu- HCy (2-amino-5-[(1-carboxy-3-sulfanylpropyl)amino]-5-oxopentanoic acid). Table 6 GMC and γ-glu-hcy in root and bulb in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), database ID: Chempider. GMC C264 (γ-glu-hcy) C H 3 H H H Molecular formula C 9 H 17 N 2 5 C 9 H 17 N 2 5 MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) 699 ± 750 a 288 ± 207 Root (0.5 mm ) 1863 ± 1758 a 278 ± 222 Root (2 mm ) 1695 ± ± 69 Bulb (0.1 2 mm ) 79 ± ± 117 a : statistically significant difference between groups by ne Way ANVA p < H 2 H 22 P age

23 A) H H 2 N H C 9 H 17 N m/z 265 relative intensity ICP-M/M relative intensity EI-M CH C 5 H 9 N 2 3 -C 4 H 8 N 2 -CH 3 + CH CH 2 Fig. 15 GMC; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 23 P age

24 A) H H 2 N H C 9 H 17 N 2 5 H 32- m/z 265 relative intensity ICP-M/M relative intensity EI-M C 5 H 8 N 3 -C 5 H 9 N 2 3 -C 4 H 5 N 2 -CH H Fig. 16 C264 (γ-glu-hcy); panel A) structure; panel B: EIC of m/z + ICP- M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 24 P age

25 2.7. Glutathione (GH and GG) The molecular masses of reduced and oxidised glutathione were detectable by both EI-M instruments, but EI-M/M fragments of the single charge molecular ions were only found in the data of the rbitrap (shown in Fig. 17 and 18). The main fragments for GH were m/z (loss of γ-glu), m/z (C 5 H 8 N 3 ) and m/z (loss of Gly) (Fig. 17), whereas GG showed main fragments at m/z (loss of γ-glu) and m/z (loss of 2 γ-glu) (Fig. 18). Table 7 GG in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), GH, database ID: Chempider. H GH + H H H H 2 GG 2 H H Molecular formula C 10 H 18 N 3 6 C 20 H 33 N MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) 174 ± 178 Root (0.5 mm ) 253 ± 137 Root (2 mm ) 275 ± 104 Bulb (0.1 2 mm ) 93 ± 88 + propiin + glutathione + methionine, amount of associated sulphur see table P age

26 H A) H H relative intensity ICP-M/M m/z 308 relative intensity EI-M C 10 H 18 N Glu 5 0 -C 6 H 9 N Gly m/z m/z Fig. 17 GH; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel EI-M-spectra, panel ) rbitrap- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 26 P age

27 A) H 2 N H H relative intensity ICP-M/M m/z 613 relative intensity EI-M H 2 N H H C 20 H 33 N * Glu -Glu m/z Fig. 18 GG; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel M-spectra, panel rbitrap-m/m spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation m/z -Glu + Gly -Gly P age

28 propenylmercaptoglutathione, -2- propenylmercaptoglutathione and -propylmercaptoglutathione Both compounds were described before by Nakabayashi et al. 2 Their main EI- M/M fragments were m/z (loss of C 5 H 7 N 3 ), m/z (loss of C 8 H 13 N 3 ), m/z (C 5 H 10 N 2 ), m/z (C 5 H 8 N 3 ) and m/z (C 3 H 5 2 ). We identified the same M fragments using q-tf-ei-m and rbitrap EI-M as identified by Nakabayashi and co-workers using an FT-ICR-M. The fragmentation spectra for both compounds also showed the by Nakabayashi et al identified characteristic ions for the allyl respectively propenyl compound (m/z (C 5 H 9 N ), Fig. 19, respectively m/z (C 8 H 13 N 2 3, Fig. 20). -propylmercaptoglutathione (C 13 H 23 N ), [M+H] + m/z , measured m/z ) not yet mentioned in the literature was identified by EI-M, eluting shortly after -2-propenylmercaptoglutathione. Its main EI-M/M fragments were m/z (loss of C 5 H 7 N 3 ), m/z (loss of C 8 H 13 N 3 ), m/z (loss of Gly) and m/z (C 5 H 8 N 3 ) (Fig. 21). Table 8-1-propenylmercaptglutathione and -2-propenylmercaptoglutathione in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), - propylmercaptoglutathione, database ID: Chempider propylmercaptoglutathione* propenylmercaptoglutathione propenylmercaptoglutathione H H 2 H H 2 H H 2 CH 2 Molecular formula C 13 H 21 N C 13 H 21 N C 13 H 23 N MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) 245 ± ± 720 Root ( ± ± 724 CH 3 CH 3 28 P age

29 mm ) Root (2 mm ) Bulb (0.1-2 mm ) 344 ± ± ± ± 93 *co-eluting with one isomer of -allyl/propenyl-pc-2, amount of associated sulphur see table P age

30 H H H 2 N A) C 13 H 21 N relative intensity ICP-M/M m/z relative intensity EI-M H 2 C Glu + C 3 H Glu + C 3 H 4 N 3 -C 8 H 13 N Glu +C 3 H 5 -Glu Fig propenylmercaptoglutathione; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI- M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation; cycle marks specific fragment for this isomer. 30 P age

31 A) H H H 2 N C 13 H 21 N relative intensity ICP-M/M m/z relative intensity EI-M H 3 C -C 8 H 13 N Glu + C 3 H 4 N Gly + C 3 H 5 + -Glu Glu + C 3 H 5 Fig propenylmercaptoglutathione; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI- M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation; cycle marks specific fragment for this isomer. 31 P age

32 A) H H 2 N C 13 H 23 N H relative intensity ICP-M/M m/z relative intensity EI-M H 3 C -C 8 H 15 N Glu + C 3 H 4 N 3 -Glu + C3 H Glu Fig. 21 -propylmercaptoglutathione; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-Mspectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 32 P age

33 2.9. xidised forms of -1-propenylmercaptoglutathione and -2- propenylmercaptoglutathione Just as the oxidised relatives of GAC and GPC are present oxidised forms of - 1 and -2-propenylmercaptoglutathion were present in the extracts of root and bulb. The fragmentation patterns for 3 of the compounds by qtf-ei-m were very similar. Compounds 1a and b showed more pronounced fragments at m/z (C 4 H 7 N 2 3 ) and m/z (C 8 H 11 N ) (Fig. 22 and 23). Compound 2a showed a more pronounced fragment at m/z (C 3 H 6 N 2 ) (Fig. 23). The last of the mentioned fragments may indicate that for this compound the H-group of GH was oxidised. No further identification with regard to the influence of double bond position and oxygen position on retention time and therefore compound characterisation was possible. Table 9 xidised forms of -1-propenylmercaptglutathione and -2- propenylmercaptoglutathione; 4 isomers (forms 1a, 1b, 2a and 2b) were found, double bond and position of -group can vary, compound 1a co-eluted with PC 2 oxidised. um of compounds (1b, 2a and 2b) in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group); quantification of individual compounds in Table 10, database ID: Chempider. -1- propenylmercaptoglutathione oxidised CH 2-2- propenylmercaptoglutathione oxidised CH 3 H H H H 2 2 Molecular formula C 13 H 21 N C 13 H 21 N MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) 11.1, 11.7, 13.6, 14.3 Database ID - - Root (0.1 mm ) 265 ± 125* Root (0.5 mm ) 187 ± 96* Root (2 mm ) 395 ± 219* Bulb (0.1 2 mm ) 85 34* * -1-propenylmercaptoglutathione and -2-propenylmercaptoglutathione compounds 1b + 2a + 2b 33 P age

34 Table 10 xidised forms of -1-propenylmercaptoglutathione and -2- propenylmercaptoglutathione compounds 1b, 2a and 2b in µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group) (mean ± standard deviation, n = 9 per group). compound 1 a (11.3 min)* compound 1b (11.7 min) compound 2a (13.6 min) compound 2b (14.3 min) Root (0.1 mm ) 104 ± ± 3 24 ± 2 Root (0.5 mm ) 63 ± ± 5 26 ± 10 Root (2 mm ) 153 ± ± 6 43 ± 12 Bulb (0.1 2 mm ) 54 ± ± 8 16 ± 5 *PC-2 + oxidised form of -1-propenylmercaptoglutathione or -2-propenylmercaptoglutathione compound 1a, amount of associated sulphur see table P age

35 A) H relative intensity ICP-M/M m/z 396 relative intensity EI-M H 2 N H CH 2 C 13 H 21 N Glu + C 3 H C 8 H 13 N Glu + C 3 H 4 N 3 -Glu + C 4 H 7 2 -C 3 H 5 2 Fig. 22 oxidised forms of -1-propenylmercaptoglutathione and -2- propenylmercaptoglutathione; panel A) structure; panel B: EIC of m/z + ICP- M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 35 P age

36 A) Fig. 23 Details of q-tf EI-M/M spectra of compounds 1a (panel A), 1b (panel and 2a (panel of oxidised forms of -1-propenylmercaptoglutathione and -2-propenylmercaptoglutathione; main differences in occurring fragments marked by cycles. 36 P age

37 2.10. Phytochelatin 2 (PC-2) and derivatives The main EI-M/M fragments of oxidised PC-2 were m/z (C 8 H 11 N 2 4 ), m/z (C 10 H 15 N 3 6 ), m/z (loss of γ-glu) and m/z (loss of γ-glu + Gly) (Fig. 24). The yet undescribed, -allyl/propenyl-containing PC-2 (C611) was present in both root and bulb in at least four different isomeric forms (Fig. 24), indicating that the -allyl/propenyl-group can be bound to either H-group of PC2. The main EI-M/M fragments were m/z (loss of γ-glu + Gly), m/z (loss of γ-glu), m/z (loss of γ-glu-cys) and m/z (C 10 H 15 N 3 6 ) (Fig. 25). Table 11 PC-2 oxidised and -allyl/propenyl-pc-2 (mean ± standard deviation, n = 9 per group), database ID: Chempider. H H 2 N PC-2 oxidised* HN H -allyl/propenyl-containing PC-2 + H 2 C N H H H N H H N H 2 H Molecular formula C 18 H 27 N C 21 H 33 N MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID - - Root (0.1 mm ) Root (0.5 mm ) Root (2 mm ) Bulb (0.1-2 mm ) * PC propenylmercaptoglutathione compound1a, amount of associated sulphur see table allyl/propenyl-containing PC-2 + -propymercaptoglutathione, amount of associated sulphur see table P age

38 A) H HN H 2 N H relative intensity ICP-M/M m/z 538 relative intensity EI-M C 18 H 27 N H -Gly + γ-glu-cys -Gly + C 6 H 7 N 4 -Gly + Glu -Glu Fig. 24 oxidised PC-2; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 38 P age

39 A) H HN CH 2 relative intensity ICP-M/M m/z 612 relative intensity EI-M H HN C 21 H 33 N H 2 N H Cys-γ-Glu-Cys(-allyl)-Gly γ-glucys + -ally -Glu + Gly -Glu Fig. 25 -allyl/propenyl-pc-2; panel A) structure; panel B: EIC of m/z + ICP- M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI-M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 39 P age

40 2.11. Precursor molecules for -alk(en)ylcysteine sulfoxides The main fragments of C207 were m/z (C 4 H 7 2 ) and m/z (loss of 3 ) (Fig. 26). For C393 the main fragments were m/z (C 4 H 7 2 ), m/z (γ-glu), m/z (C 6 H 12 N 2 ), m/z (C 9 H 12 N 4 ) and m/z (loss of Gly) (Fig. 27). Table 12 C207 and C393 µmol compound kg -1 d.w. (mean ± standard deviation, n = 9 per group), database ID: Chempider. C207 (2-amino-3-[(2- carboxypropyl)sulfany l]-propanoic acid) C336 (2-amino-5-({1- carboxy-2-[(2- carboxypropyl)sulfany l]ethyl}amino)-5- oxopentanoic acid) C393 (2-amino-5- ({1-[(carboxy- methyl)amino]-3- [(2- carboxypropyl)sulf anyl]-1-oxopropan- 2-yl}amino)-5- oxopentanoic acid) C H 3 H 2 H H 3 C H H H H 3 C H H H Molecular formula C 7 H 13 N 4 C 12 H 20 N 2 7 C 14 H 23 N 3 8 MW (g mol -1 ) [M+H] + (theoretical) [M+H] + (measured) RT (min) Database ID Root (0.1 mm ) 669 ± 305 a - 83 ± 36 c Root (0.5 mm ) 1272 ± 859 b ± 104 d Root (2 mm ) 3214 ± 1827 a,b ± 375 c,d Bulb (0.1 2 mm ) 158 ± ± 216 a,b : statistically significant difference between groups by ne-way ANVA p < 0.05 c,d : statistically significant difference between groups by ne-way ANVA p < P age

41 A) 32- m/z 208 C H 3 H 2 H C 7 H 13 N 4 relative intensity ICP-M/M relative intensity EI-M C 3 H 6 N 2 -CH 2-2 Fig. 26 C207; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 41 P age

42 A) H 2 N H H H CH 3 C 14 H 23 N 3 8 relative intensity ICP-M/M m/z relative intensity EI-M -Glu + C 3 H 4 N 3 - CH C 3 H 4 N Glu + Gly + C C 10 H 16 N Gly Fig. 27 C393; panel A) structure; panel B: EIC of m/z + ICP-M/M trace of root extract exposed to 2 mm, panel q-tf EI-M-spectra, panel EI- M/M spectra with some annotated fragments, shown is the composition of the loss for the individual fragment calculated from the parent mass; for HPLC + ICP- M/M and EI-M conditions see main text under instrumentation. 42 P age

43 Table 13 Amount of sulphur associated with co-eluting compounds in mmol kg - 1 d.w. at the given retention times Root (0.1 mm ) Root (0.5 mm ) Root (2 mm ) Bulb (0.1 2 mm ) RT 2.2 min RT 3.7 min RT 11.1 min RT 27.3 min PC2 ox. & γ- propylmercaptoglutathione cycloalliin propiin, glutamyl--2- methiin & methinone & propenylcysteine & -allyl/propenylcontaining PC-2 methylcysteine GH sulfoxide 1a 3.5 ± 3.6 (1.7 %) 1.2 ± 0.53 a 0.19 ± ± ± 4.5 (2.2 %) 1.5 ± ± ± ± 1.5 (2.2%) 3.2 ± 2.2 a 0.32 ± ± ± 0.66 (0.67 %) 0.95 ± ± ± P age

44 3. Results from PCA and cluster analysis 3.1. Loading plot Loading Plot allyl/propenyl- GAC Methiin + cyclo + methylcystein -2-PMG GPC() -1-PMG GG 0.2 -allycysteine GPC PC2 0.1 C PC2 GMC Alliin C207-1/2 Propiin + GH + Met C393 GAC() Isoalliin PC Fig. 28 Loading plot for the PCA in which all plants from the three different - fertilization stage (0.1, 0.5 and 2 mm ) by utilisation of all identified low molecular weight -containing metabolites. 44 P age

45 3.2. Dendrogram Cluster analysis: Dendrogram 0.00 imilarity Fig. 29 Cluster analysis in which all plants from the three different - fertilization stage (0.1, 0.5 and 2 mm ) by utilisation of all identified low molecular weight -containing metabolites plants with the different sulphur fertilization (mm ) P age

46 4. Example of HPLC-ICP-M/M traces for all plant parts fertilized with different amount of sulphur root 0.1 mm root 2 mm bulb 0.5 mm root 0.5 mm bulb 0.1 mm bulb 2 mm relative intensity ICP-M/M Fig. 30 comparison of 32 -trace of root and bulb extracts (one each at different fertilizer concentrations); for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. relative intensity ICP-M/M root 0.1 mm root 0.5 mm root 2 mm bulb 0.1 mm bulb 0.5 mm bulb 2 mm Fig. 31 comparison of 32 -trace of root and bulb extracts (one each at different fertilizer concentrations); for HPLC + ICP-M/M and EI-M conditions see main text under instrumentation. 46 P age

47 5. Example of peak area integration Fig. 28 Example of peak area integration using PeakFit. Top panel: measured signal (line) and calculated line (dots) using the described model for calculation in logarithmic y-scale, bottom panel: showing individual peaks (linear y-scale); R 2 = for peak area covered 1 Y. Yamazaki, T. Tokunaga, T. kuno, Quantitative determination of eleven flavor precursors (-alk(en)yl cysteine derivatives) in garlic with HPLC method, Nippon hokuhin Kagaku Kogaku Kaishi, 2005, 52, ; in E. Block, Garlic and other Alliums The Lore and the cience, Royal ociety of Chemistry: Cambridge, U.K., R. Nakabayashi, Y. awada, M. Aoyagi, Y. Yamada, M. Y. Hirai, T. akurai, T. Kamoi, D. D. Rowan, K. aito, Chemical Assignment of tructural Isomers of ulfur-containing Metabolites in Garlic by Liquid Chromatography-Fourier Transform Ion Cyclotron Resonance-Mass pectrometry, J. Nutr., 2016, 146, P age