S -Alk(en)yl-Cysteine Sulfoxides in Allium Species Are Excellent Acrolein Scavengers: Implications for Secondary Antioxidants in Plants

33 Reactive carbonyl species (RCS), such as acrolein (Acr), are generated via the degradation of 34 lipid peroxides and function both as signaling molecules and as cytotoxins downstream of 35 reactive oxygen species (ROS). The aim of this study was to identify RCS-scavenging 36 compounds, acting as secondary antioxidants, in plants. We established an HPLC-based 37

to sensitively evaluate the RCS-scavenging ability of compounds, using Acr as a 38 representative RCS. Among 80% ethanol extracts of 46 angiosperm species, the highest 39 scavenging activities were observed in garlic, spinach, avocado, broccoli, and lotus, 40 representing a diverse range of plant families. The active ingredient from garlic cloves was 41 identified as S-allyl cysteine sulfoxide (alliin), an amino acid typical of Allium species, which 42 trapped up to two Acr molecules at its amino group. Notably, S-(1E)-propenyl cysteine 43 sulfoxide (isoalliin), a structural analog of alliin and a major constituent of onion, showed 44 even stronger Acr-scavenging activity. Isoalliin outperformed known Acr scavengers such as 45 carnosine, epigallocatechin gallate, resveratrol, and hesperetin. These findings highlight 46 isoalliin as a representative plant-derived RCS scavenger and emphasize the importance of S-47 alk(en)yl cysteine sulfoxides as potent Acr-scavenging compounds in plants. This study 48 further suggests that structurally diverse RCS scavengers remain to be discovered among 49 specialized metabolites in the plant kingdom. 50 51

acrolein, alliin, isoalliin, methiin, reactive carbonyl species 52 53 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 4 Plant life is exposed to potential threats from reactive oxygen species (ROS). During 54 photosynthesis in plants, singlet oxygen (1O2) and superoxide radical (O2–) are constantly 55 produced in chloroplasts (Asada 2006). Furthermore, the H2O2 generation flux in 56 peroxisomes in illuminated leaves of C3 plants reaches 25% of the flux of CO2 fixation on 57 electron base. This is approximately 50 times greater than the ROS generation flux associated 58 with mitochondrial respiration (Foyer and Noctor, 2003). When plants suffer environmental 59 stress, i.e. any changes in the surrounding environmental conditions that are not favorable for 60 a plant, ROS production in the cells are increased due to the disturbance of metabolism (for 61 example, the rate of CO2 fixation in leaves decreases at low temperatures or under drought). 62 ROS oxidize biomolecules such as proteins, nucleic acids and lipids, thereby impairing their 63 functions and damaging cells. As a defense system against ROS, plant cells are richly 64 equipped with multiple types of ROS-scavenging enzymes and antioxidants, thereby keeping 65 ROS concentrations low enough. 66 ROS exert their biological functions also via the secondary products 'reactive carbonyl 67 species (RCS)'. RCS is a collective name of the α,β-unsaturated aldehydes and ketones such 68 as acrolein (Acr) and 4-hydroxy-(E)-2-nonenal (HNE) that are formed via the degradation of 69 lipid peroxides (Esterbauer et al. 1991, Aldini et al. 2005, Mano 2012). An RCS molecule 70 readily forms an adduct with a protein at the thiol-, amino-, or imidazole group via Michael 71 addition of its β-carbon (Esterbauer et al. 1991), and cause alterations of the protein function 72 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 5 (Wible and Sutter 2017). In plants, more than a dozen types of RCS occur (Schauenstein et 73 al. 1977, Mano 2012) and their levels are increased under oxidative stress conditions (Yin et 74 al. 2010, Yamauchi et al. 2012, Biswas and Mano 2015, Roach et al. 2018, Sultana et al. 75 2022, 2024). Rises of RCS levels have significant physiological impacts. For example, the 76 genetic enhancement of the RCS-scavenging enzyme 2-alkenal reductase (AER) (Mano et al. 77 2002, 2005) in transgenic plants suppressed the increases in RCS levels due to salt stress 78 (Sultana et al. 2024), aluminium stress (Yin et al. 2010) and light stress (Mano et al. 2010) 79 and thereby conferred stress tolerance. Vice versa, the genetic deficiency of aldehyde 80 alkenal/one reductase (Yamauchi et al. 2012) and aldehyde oxidase (Srivastava et al. 2017, 81 Nurbekova et al. 2021) made the plants more susceptible to oxidative stress or senescence. 82 The plant defense system against RCS toxicity consists of several types of enzymes and small 83 molecules, as the system against ROS. As for enzymatic defense, several classes of enzymes 84 with distinct RCS-detoxifying reactions have been characterized such as AER (Mano et al. 85 2002, 2005), glutathione transferase tau isozymes (Mano et al. 2017, 2019a,b) and aldehyde-86 scavenging enzymes such as aldo-keto reductase (Yamauchi et al. 2011), aldehyde 87 dehydrogenase (Sunker et al. 2003) and aldehyde oxidase (Srivastava et al. 2017, Nurbekova 88 et al. 2021). 89 RCS-scavenging small molecules also occur in plants. Thiol compounds such as cysteine 90 (Cys) and the reduced form of glutathione are excellent RCS scavengers (Esterbauer et al. 91 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 6 1975). Various polyphenols also have the RCS scavenging ability. Zhu et al. (2009) 92 compared the RCS scavenging ability of a dozen tea polyphenols in vitro and found that nine 93 of them, including epigallocatechin gallate (EGCG), showed a high Acr-scavenging activity. 94 The reaction of EGCG and Acr starts with a nucleophilic attack of the carbon at position 8 on 95 the EGCG A-ring on the b-carbon of Acr, thus forming a Michael adduct. Subsequent aldol 96 condensation produces a ring structure (Huang, et al., 2020). One EGCG molecule can bind 97 up to three Acr molecules (Sugimoto, et al., 2021). Acr-scavenging reactions have been also 98 characterized for other polyphenols and their glycosides such as resveratrol, hesperetin 99 (Wang, et al., 2015), ferulic acid (Tao, et al., 2019), pelargonidin (Colzani, et al., 2018), 100 myricetin (Zhang, et al., 2020) and cyanidin-3-O-glucoside (Song, et al., 2021). 101 Amino compounds also have RCS-scavenging potential. The primary amino group in g-102 aminobutyric acid (GABA), L-Ala and L-Ser serves as a Michael acceptor to bind up to two 103 molecules of Acr (Jiang et al. 2020a, Zou et al. 2021). Recently it was reported that S-allyl 104 cysteine sulfoxide (alliin), an amino acid typical of Allium species, can scavenge Acr at a 105 high speed although the reaction mechanism has yet to be clarified (Uemura et al. 2023). 106 Theophylline, a purine derivative found in tea, forms a Michael adduct between its secondary 107 amine and the b-carbon of Acr at a high rate (Jiang, et al., 2020b). Interestingly, animals 108 contain RCS-scavenging dipeptides carnosine, homocarnosine, and anserine (Aldini et al. 109 2002) in brain and muscle. The amino and imidazole groups of these histidine-containing 110 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 7 dipeptides react as Michael acceptors (Aldini et al., 2005), allowing them to be excellent RCS 111 scavengers (Aldini et al., 2002; Liu et al., 2003; Carini et al., 2003). Exogenous application 112 of these dipeptides to plant cells or seedlings suppressed the RCS levels and mitigated 113 oxidative injury, without affecting ROS levels (Biswas and Mano, 2015; Sultana et al., 2022), 114 showing that RCS-scavenging compounds can protect plants as the "secondary antioxidants". 115 In this study, we aimed at discovering RCS-scavenging compounds in plants. Employing Acr 116 as a representative RCS, we evaluated RCS-scavenging ability of dozens of angiosperms. 117 Garlic cloves showed the highest content of Acr-scavenging ability, which was attributed to 118 alliin, a garlic-typical amino acid. We found that isoalliin and methiin, the cysteine 119 derivatives relative to alliin, also showed excellent Acr-scavenging ability and revealed the 120 reaction mechanism. Potential of plant specialized compounds as the secondary antioxidants 121 is discussed. 122 123

124 Chemicals Alliin and methiin were purchased from LKT Laboratories (Saint Paul, MN). 125 Carnosine, fluorescein, and HNE dimethyl acetal (HNE -DMA) were purchased from Sigma-126 Aldrich Japan (Tokyo), and isoalliin from Nagara Science Co. (Gifu, Japan). Anserine nitrate 127 was purchased from Fujifilm Wako Pure Chemical (Osaka, Japan). 2,4-128 Dinitrophenylhydrazine ( DNPH), purchased from Fujifilm Wako, was purified by 129 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 8 recrystallization in acetonitrile and kept in acetonitrile until use (Mano et al. 2022). All other 130 chemicals were of analytical grade. Acr was prepared on the day of use by hydrolysis of the 131 acetal, as follows. Acrolein dimethyl acetal (Tokyo Chemical Industry, Tokyo , Japan ) was 132 dissolved in 100 mM HCl aqueous solution at 40˚C for 40 min. The solution was then 133 neutralized with NaHCO3 and further incubated at 25˚C for 2 h. The concentration of Acr was 134 determined from the absorbance at 215 nm with an extinction coefficien t of 15.0 mM–1 cm–1 135 (in water). 136 Plant Materials Bamboo ( Phyllostachys heterocycla) shoots and yuzu ( Citrus junos ) fruits 137 were harvested from Yamaguchi University’s experimental field. Following materials of fresh 138 plants were purchased from grocery stores in Yamaguchi (Japan) between April and December 139 2018: Apple ( Malus domestica ) fruits, asparagus ( Asparagus officinalis ) shoots, avocado 140 (Persea americana ) fruits, bell pepper ( Capsicum annuum 'Grossum') fruits, bitter melon 141 (Momordica charantia var. pavel) fruits, broccoli (Brassica oleracea var. italica) buds, carrot 142 (Daucus carota subsp. sativus) roots, cherry ( Cerasus avium ) fruits, chestnuts ( Castanea 143 crenata), Chinese yam ( Dioscorea polystachya) roots, common ginger ( Zingiber officinale) 144 rhizomes, edible burdock ( Arctium lappa ) roots, eggplant ( Solanum melongena ) fruits, fig 145 (Ficus carica) fruits, garlic (Allium sativum) cloves, grape (Vitis labruscana 'Pione') fruits, 146 green onion ( Allium fistulosum ) leaves, guava ( Psidium guajava ) fruits, Haskap berry 147 (Lonicera caerulea var. emphyllocalyx) fruits, Indian gooseberry (Phyllanthus emblica) fruits, 148 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 9 Japanese apricot ( Prunus mume ) fruits, Japanese butterbur ( Petasites japonicus ) stalks, 149 Japanese ginger (Zingiber mioga) shoots, Jun-sai (Bransenia shreberi) leaves, Kaki persimmon 150 (Diospyros kaki) fruits, kiwifruit (Actinidia deliciosa) fruits, Korean lettuce ( Lactuca sativa) 151 leaves, loquat ( Rhaphiolepis bibas ) fruits, lotus ( Nelumbo nucifera ) rhizomes, Mandarin 152 orange ( Citrus unshiu ) fruits, mulkhiya ( Corchorus olitorius ) leaves, okra ( Abelmoschus 153 esculentus) fruits, onion ( Allium cepa ) bulbs, parsley ( Petroselinum crispum ) leaves, pea 154 (Pisum sativum ) seeds, perilla ( Perilla frutescens var. crispa) leaves, pitaya ( Hylocereus 155 undatus) fruits, plum ( Prunus salicina ) fruits, spinach ( Spinacia oleacea ) leaves, star fruit 156 (Averrhoa carambola ) fruits, strawberry ( Fragaria × ananassa) achenes, taro ( Colocasia 157 esculentai) tubers, watermelon ( Citrullus lanatus ) fruits, and white radish ( Raphanus 158 sativus var. hortensis) roots. 159 Plant Extract Preparation An edible portion (10 g) taken from a fresh sample was 160 homogenized in 80% (v/v) ethanol (20 mL) using a kitchen blender and filtered through a layer 161 of Miracloth (Sigma -Aldrich Japan). The filtrate was acidified with HCl (10 mM final) to 162 facilitate protein precipitation. It was centrifuged after 10 minutes of incubation at 25 ˚C, and 163 the resulting supernatant was neutralized with sodium bicarbonate. Th e volume of 164 deproteinized extract was measured. 165 Determination of the Acr-scavenging ability A reaction mixture containing 200 µM Acr, 50 166 mM potassium phosphate pH 6.8, and a plant extract was incubated at 30˚C for 10 min. One 167 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 10 volume of the mixture was then added to 6.35 volumes of acetonitrile containing 4 mM DNPH 168 and 0.4 M formic acid and incubated at 30˚C for 60 min. The resulting hydrazone solution was 169 filtered through a 0.45 µm filter (Millex LH ; Merck, Darmstadt, Germany) and analyzed by 170 HPLC, as follows. The hydrazone solution (10 µL) was injected into a Zorbax ODS column 171 (4.6 ´ 250 mm, 5 µm; Agilent Technologies Japan, Hachioji, Japan) kept at 40˚C and separated 172 in a 1.0 mL min–1 flow of 75% (v/v) acetonitrile in water (isocratic elution) with a high-pressure 173 pump (2690 Separation Module , Waters Corp., Milford, MA ). The hydrazone was detected 174 with a UV-visible detector (2487 Absorbance Detector, Waters) at 370 nm. 175 The decrease rate (v) of Acr concentration in the reaction mixture (in mol Acr consumed s–1 L–176 1) was used to calculate the Acr-scavenging ability of the plant extract per unit volume (in mol 177 Acr s–1 (L extract)–1) by multiplying v by the volume ratio P (reaction mixture/plant extract) as 178 follows: 179 Acr-scavenging ability per unit volume of plant extract (mol s–1 (L extract)–1) 180 = v (mol Acr scavenged L–1 s–1) ´ P 181 The Acr-scavenging ability content in the original sample , as mol Acr consumed s –1 (g 182 sample)–1, was then calculated from the extraction efficiency E (L extract (g sample) –1) as 183 follows: 184 Acr-scavenging ability content (mol s–1 (g sample)–1) 185 = Acr-scavenging ability of the extract (mol s–1 L–1) ´ 1 (L) ´ E (L extract (g sample)–1) 186 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 11 Purification of the Acr-scavenging component from garlic cloves Garlic cloves (40 g) were 187 homogenized in 80% (v/v) ethanol (80 mL) using a kitchen blender, filtered through two layers 188 of Miracloth, and centrifuged to remove insoluble debris. The supernatant was freeze-dried to 189 obtain 7.6 g of material, which was dissolved in distilled water (15 mL) and loaded on to a 190 preparative hydrophilic column (Universal Column Premium Amino 30 µm, 3.0 ´ 16.5 cm; 191 Yamazen Corp., Osaka, Japan) with a medium pressure chromatography system (EPCLC AI -192 580; Yamazen Corp.) equipped with a detector (fixed wavelength at 254 nm). Chromatography 193 was performed at a flow rate of 20 mL/min under the following solvent conditions: solvent A: 194 acetonitrile:10 mM ammonium acetate in water (88:12, v/v); solvent B: acetonitrile:100 mM 195 ammonium acetate in water (1:1, v/v). The gradient program consisted of 0–16 min isocratic 196 elution of 0% B, 16–56 min of a linear gradient from 0% to 100% B, and 56–80 min isocratic 197 elution of 100% B. The f raction size was 20 mL. Each fraction was freeze -dried and then 198 dissolved in 2 mL of distilled water, and its Acr-scavenging ability was determined as 199 described above. The most active fraction was then separated on a Unison UK-Amino Column 200 (3 µm, 4.6×100 mm; Imtakt, Kyoto, Japan) with an HPLC apparatus equipped with an LH-40 201 liquid handler and an SPD -10AVP UV -VIS detector (Shimadzu Corp., Kyoto, Japan). 202 Chromatography was performed at a flow rate of 0.6 mL min–1 with the following multistep 203 gradient elution: 0–10 min isocratic elution of 0% B; 10–35 min of a linear gradient from 0% 204 to 100% B; 35–50 min isocratic elution of 100% B; 50–55 min of a linear gradient from 100% 205 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 12 to 0% B; and, finally, 55–90 min isocratic elution of 0% B. The fraction size was 1.8 mL. The 206 elution profile was monitored at 220 nm for nonspecific detection of organic substances. After 207 40 separation runs (50 µL injection per run), identical fractions were accumulated, freeze-dried, 208 and dissolved in a smaller volume to determine the Acr-scavenging ability. 209 LC-MS/MS analysis Chromatographic separation was performed on a hydrophilic Intrada 210 Amino Acid column (3 µm, 3 × 100 mm; Imtakt) with a Vanquish UHPLC System (Thermo 211 Fisher Scientific, Waltham, MA). A 5 -µL aliquot of the sample was injected into the column 212 and eluted at a flow rate of 0.4 mL min –1 under the following solvent conditions: solvent A: 213 0.3% (v/v) in acetonitrile; solvent B: 0.1 M ammonium formate in water:acetonitrile (80:20, 214 v/v). The gradient program was as follows: 0 –10 min isocratic elution of 15% B; 10 –25 min 215 of a linear gradient from 15% to 60% B; 25–40 min isocratic elution of 60% B; 40–45 min of 216 a linear gradient from 60% to 15% B; and finally 45 –50 min isocratic elution of 15% B. The 217 column temperature was set to 40˚C. Mass spectra were acquired using an Orbitrap Exploris 218 120 mass spectrometer (Thermo Fisher Scientific) equipped with an H -ESI source, with 219 acquisition in positive ion mode. The source parameters were as follows: capillary voltage of 220 3.5 kV in positive ion mode, ion transfer tube temperature of 325˚C, vaporizer temperature of 221 350˚C, sheath gas flow of 50 units, auxiliary gas flow of 10 units, and sweep gas of 1 unit. A 222 full scan was performed in the mass-to-charge ratio (m/z) range of 100–1000, with a resolution 223 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 13 of 120,000 at m/z 200. Tandem MS information was acquired in tMS2 mode with a resolution 224 of 12,000 and HCD collision energies of 15, 30, and 45. 225 226

and Discussion 227 Comparison of Acr-scavenging ability of plants 228 For choosing a starting material suitable for purifying RCS-scavenging compounds, we first 229 collected various plant samples broadly in the angiosperm phylogeny. To obtain samples in 230 such a variety in short time, we obtained them (46 species in total), mainly fruits and vegetables, 231 from local retail stores (see Materials and Methods). We extracted plants with 80% ethanol 232 because we targeted polar compounds rather than polyphenols, which have been extensively 233 investigated (Colzani, et al., 2018; Huang, et al., 2020; Sugimoto, et al., 2021; Tao, et al., 234 2019; Wang, et al., 2015; Zhang, et al., 2020; Zhu, et al., 2009). This extract was deproteinized 235 by acid treatment, then neutralized, and its Acr-scavenging rate was determined. To survey 236 RCS scavengers that can react rapidly with Acr, we determined Acr consumption in the initial 237 10 minutes in our assay. The initial Acr concentration was set at 200 µM, mimicking 238 physiological concentration, which was estimated on the following observation. A typical Acr 239 content in salt -stressed A. thaliana seedlings is 200 nmol (g fresh weight) –1 (Sultana, et al., 240 2022). From this value, Acr concentration of 200 µM in the tissue was calculated on a simple 241 assumption that the Acr molecules occur in the tissue in a homogenous solution of 1 mL 242 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 14 volume per gram tissue weight. The reaction temperature was set at 30˚C to simulate field 243 conditions. 244 Fig. 1 summarizes the Acr-scavenging ability on the fresh weight base of the plant materials. 245 Most plant materials showed significant levels of Acr -scavenging ability, thus indicating that 246 many plants contain Acr-scavenging compounds. The highest Acr -scavenging ability content 247 (in nmol Acr scavenged s –1 (g fresh weight) –1) were 1.39 for garlic cloves, 0.99 for spinach 248 leaves, 0.96 for avocado fruits, 0.92 for lotus rhizomes, 0.91 for broccoli buds, 0.82 for 249 chestnuts, 0.70 for bamboo shoots, 0.67 for onion bulbs, and 0.65 for both Chinese yam roots 250 and white radish roots. These plants, when placed in the angiosperm phylogeny tree (Cole, et 251 al., 2019), were distributed to different orders of taxa (Fig. 1). We chose garlic cloves as the 252 starting material for purification to identify an Acr-scavenging compound. 253 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 15 254 255 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 16 Purification of Acr-scavenging components from garlic extract 256 Garlic clove extract was first separated by preparative amino column chromatography 257 (Supplementary Fig. S1A), which showed a single prominent peak of Acr-scavenging ability 258 in fractions #13, #14, and #15. We scaled up the sample amount and the column size , as 259 described in the Materials and Methods section, collected the active fractions, concentrated and 260 purified them using an analytical HPLC column. The resulting f raction #8 had a high Acr-261 scavenging ability (Supplemental Fig. S1B). We then analyzed this fraction by LC-MS/MS. 262 Total ion chromatograph y (m/z 100–1000) revealed two major components: a broad peak 263 eluted at 15.3–16.6 min with m/z 178.0528 and a narrow peak at 18.87 min with m/z 258.1096 264 (Fig. 2A, black trace). 265 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 17 To clarify which compone nt is the Acr scavenger, we mixed fraction #8 with Acr and 266 examined the changes in the total ion chromatogram. The addition of Acr at 2 mM resulted in 267 a significant decrease in t he 15.3–16.6 min peak (Fig. 2A, inset), while that of the 18.87 min 268 peak was unchanged, thus suggesting that the m/z 178.0528 signal represented the Acr 269 scavenger in fraction #8. Addition of Acr to fraction #8 generated two new broad peaks at 8.57 270 (m/z 290.1056) and 13.16 min ( m/z 234.0795) (Fig. 2A, asterisks). The m/z values of these 271 peaks matched hypothetical Acr adducts to the m/z 178.0528 compound; the 13.16 min peak 272 corresponded to a mono-Acr adduct (mass increment 56.0265, matching the molecular weight 273 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 18 of Acr (C3H4O)), and the 8.57 min peak a di-Acr adduct (mass increment 112.0526 )., these 274 were due to the reaction products between the m/z 178.0528 compound and acrolein; the 13.16 275 min peak corresponding to 276 Assuming that the m/z 178.0528 ion represents a [M + H] + ion, the elemental composition 277 of the molecule was deduced as C6H11NO3S. Retrieval of the corresponding compounds in 278 the databases AraCyc (Mueller, et al., 2003), PlantCyc (Plant Metabolic Network (PMN): 279 https://pmn.plantcyc.org/organism-summary?object=PLANT, on www.plantcyc.org, Jan 15, 280 2022), and KEGG (Kanehisa, et al., 2000) using the software package Compound Discoverer 281 3.0 (Thermo Fisher Scientific), resulted in S-allyl-cysteine sulfoxide (alliin; C6H11NO3S, 282 177.22 g mol–1) as the most probable candidate. 283 Alliin is an amino acid sulfoxide characteristic of garlic, accounting for about 30% (w/w) of 284 total free amino acids in garlic cloves (Ueda, et al., 1991) , and is a precursor of allicin, the 285 major compound of garlic odor (Stoll and Seebeck, 1948) . In the LC-MS/MS analysis , 286 authentic L-alliin eluted as a broad peak ranging from 15.2–16.7 min with an m/z of 178.0532 287 (Fig. 2B). The MS2 spectrum of this parent ion ( m/z 178.0532) matched the spectrum of the 288 m/z 178.0528 ion of the Acr scavenger in fraction #8 (Fig. 2A). The addition of Acr to alliin 289 decreased the alliin peak and generated two broad peaks at 8.20 (m/z 290.1056) and 13.02 min 290 (m/z 234.0795) (Fig. 2B), as observed for fraction #8. Thus, the Acr-scavenging component in 291 the purified fraction was identified as alliin. 292 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 19 Reaction mechanism of alliin with Acr Uemura et al. (2023) has reported that alliin has a high 293 capacity to scavenge Acr. Its reaction mechanism with Acr, yet to be elucidated, could be 294 similar to that of an amino compound; i.e. the Michael addition of the b-carbon to the amino 295 nitrogen, and the second addition of Acr to the same nitrogen, followed by the intramolecular 296 aldol condensation to form a nitrogen-containing heterocycle (Carini et al. 2003). When alliin 297 was incubated with Acr, larger MS signals of the mono-Acr adduct (m/z 234.0795, eluted at 298 8.20 min ) and the di-Acr adduct (m/z 290.1056, 13.02 min ) were observed as increasing 299 concentration of Acr (Fig. 3A, B). In addition, another signal at m/z 272.0950 arose in higher 300 Acr concentration (Fig. 3C). On the assumption that it represented a [M + H]+ ion, the atomic 301 composition of the compound was deduced as C12H17NO4S. W e presumed that this signal 302 corresponded to a dehydrated product of the di-Acr adduct. 303 304 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 20 We have purified this putative dehydrated product of the di -Acr adduct and determined its 305 structure with NMR. The deduced structure corresponded to the Nb-(3-formyl-3,4-306 dehydropiperidino) derivative of alliin (Table 1) , indicating that the amino group of alliin 307 accepted two molecules of Acr, as expected. Specifically, one Acr molecule adds to the amino 308 group (Michael-type reaction), then the resulting secondary amino group undergoes Michael 309 addition with another Acr molecule to form the (Acr)2 adduct, which then under goes aldol 310 condensation to form a nitrogen-containing heterocycle (Figure 4). The MS2 spectra for the 311 Acr adducts support this mechanism ; the spectrum for the mono -Acr adduct contained a 312 product ion at m/z 144.0655, the deduced atomic composition for which (C6H10NO3) matched 313 the fragment generated by cleavage between the sulfur atom and the b-carbon of the amino 314 acid, the same cleavage position as in alliin (Fig. 2B). Similarly, the MS2 spectra of the di-Acr 315 Table 1. 1H NMR (400 MHz) and 13C NMR (500 MHz) chemical shift data of the purified adduct of m/z 272.0950. Deduced structure of the dehydrated product of (Acr)2-alliin adduct and a possible LC-MS/MS fragment ion structure are shown. NMR data Structural formula of the dehydrated product of (Acr) 2- alliin adduct Fragment ion (+) of LC-MS/MS No. δH δC exact mass 271.08728 m/z 182.0810 1 5.52 124.85 2 5.94 125.03 3 3.60, 3.80 54.2 4 3.27, 3.33 49.97 5 3.7 63.68 6 - 174.82 7 2.87, 3.06 45.9 8 2.61 26.81 9 7.18 152.41 10 - 138.06 11 9.35 195.6 12 3.5 45.48 N O OHS O O 1 2 3 4 5 6 7 8 9 10 11 12 N O OH H2C O (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 21 adduct and its dehydrated product contained product ions at m/z 200.0917 (Fig. 3B) and m/z 316 182.0811 (Fig. 3C), which correspond to the fragments at the same cleavage position of the 317 deduced compounds, respectively. 318 319 320 Comparison of the Acr-scavenging ability of S-alk(en)yl cysteine sulfoxides, amino 321 compounds, and polyphenols 322 Plants in the Allium family are rich in alliin-analogous S-alk(en)yl cysteine sulfoxides, such as 323 S-propenylcysteine sulfoxide (isoalliin) in onion (A. cepa ) and S-methylcysteine sulfoxide 324 (methiin) in Oriental garlic (A. tuberosum) (Yamazaki, et al., 2011). Isoalliin and methiin also 325 formed mono-Acr and (Acr)2 adducts and the dehydration product of the (Acr)2 adduct, as did 326 alliin (Supplementary Fig. S3 and S4). The Acr-scavenging rates of these S-alk(en)yl cysteine 327 sulfoxides were 120–150 µM Acr in 30 min, when the reaction was started with 1 mM Acr and 328 1 mM scavenger (Table 2A) . These rates were comparable to, or even higher than, those for 329 carnosine and anserine , Acr-scavenging dipeptides in animals (Carini, et al., 2003; Spaas, et 330 al., 2021). 331 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 22 As compared with broader variety of amino compounds, these S-alk(en)yl cysteine sulfoxides 332 excelled GABA and most of protein-constituting amino acids such as Met, His, Lys, Arg and 333 Gly for the Acr -scavenging rate . Such high Acr-scavenging rate for S-alk(en)yl cysteine 334 sulfoxides appears to be attributed to the presence of sulfoxide close to the amino carbon. 335 Indeed, methionine sulfoxide, in which the sulfoxide is one carbon further away, showed a 336 significantly faster Acr-scavenging rate than other amino acids lacking a sulfoxide although its 337 rate was slower than S-alk(en)yl cysteine sulfoxides (Table 2A). 338 We also compared the S-alk(en)yl cysteine sulfoxides with polyphenols that can scavenge 339 Acr. Isoalliin, methiin and alliin outperformed EGCG in scavenging Acr (Table 2A). To test 340 hydrophobic polyphenols resveratrol and hesperetin (Wang, et al., 2015), a separate experiment 341 was performed in 10% DMSO solution (Table 2B). Reaction time was set to 5 h because these 342 polyphenols scavenged Acr slowly in our preliminary experiments. Under these conditions, 343 alliin scavenged Acr faster than EGCG, resveratrol, and hesperetin. Theophylline, another type 344 of Acr scavenger (Jiang, et al., 2020b) showed an Acr-scavenging rate comparable to that for 345 hesperetin and much slower than that for EGCG and alliin (Table 2B). 346 347 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 23 Table 2 (A) Comparison of Acr-scavenging rate of amino compounds. Acr (1 mM) was 348 incubated with a scavenger (1 mM) in 50 mM potassium phosphate, pH 7.0, at 30˚C for 30 349 min. The remaining Acr was determined by HPLC after derivatization with DNPH, as 350 described in the Materials and Methods section. Acr consumption (concentration in the 351 reaction mixture) in 30 min is shown. (B) Comparison of Acr-scavenging rate for alliin, 352 polyphenols, and theophylline. Acr (1 mM) was incubated with a scavenger (1 mM) in 10% 353 DMSO, 50 mM potassium phosphate, pH 7.0, at 30˚C for 5 h. Acr consumption in 5 h is 354 shown. The values shown are the average and standard error of three independent runs. 355 Different letters in the Statistics column represent significantly different values (P < 0.05, 356 Tukey's test). 357 A Compound Acr consumption (µM in 30 min) Statistics Isoalliin 153.8 ± 6.5 a Methiin 144.5 ± 14.0 ab Alliin 126.8 ± 5.4 bc Carnosine 114.3 ± 1.3 cd Anserine 99.4 ± 5.5 de Methionine sulfoxide 86.0 ± 8.3 e Histidine 63.9 ± 6.0 f Lysine 59.5 ± 8.9 f Phenylalanine 57.5 ± 0.9 f EGCG 45.4 ± 5.2 f Arginine 44.7 ± 6.3 f Histamine 40.1 ± 8.4 fg Methionine 22.7 ± 8.7 gh GABA 10.6 ± 5.9 h b-Alanine 6.5 ± 3.7 h 358 B Compound Acr consumption (µM in 5 h) Statistics Alliin 419.1 ± 3.0 a EGCG 282.0 ± 4.6 b Resveratrol 31.8 ± 1.5 c Hesperetin 18.9 ± 1.0 d Theophylline 17.9 ± 3.6 d (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 24 359 RCS scavenging compounds are present in many plant species Because RCS are secondary 360 toxic products derived from ROS, plants have evolved multiple enzymatic and non-enzymatic 361 mechanisms to mitigate their harmful effects (Mano et al. 2019 a,b). In this study, we 362 established a method to evaluate the RCS-scavenging ability of plant extracts and demonstrated 363 that S-alk(en)yl cysteine sulfoxides, amino acids characteristic of Allium family, are highly 364 efficient Acr scavengers. Among them, isoalliin, the major sulfoxide in onion, exhibited the 365 strongest Acr-scavenging activity, surpassing not only its structural analog alliin but also other 366 well-known plant-derived and animal-derived scavengers. This finding identifies isoalliin as a 367 representative plant RCS scavenger and underscores the contribution of Allium metabolites to 368 the chemical defense system against carbonyl stress. 369 It is noteworthy that the presence of a sulfoxide group in close proximity to the amino 370 carbon markedly enhances the Michael acceptor reactivity of the amino group, thereby 371 conferring the high RCS -scavenging capacity observed in S -alk(en)yl cysteine sulfoxides. 372 Consequently, the RCS -scavenging property of a compound appears to be governed by the 373 cooperative action of two or more functional groups, rather than by the reactivity of a single 374 moiety alone. More broadly, our findings suggest that structurally diverse, and as yet 375 unexplored, plant metabolites may play critical roles in mitigating oxidative damage through 376 RCS detoxification. 377 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 25 378 379 ABBREVIATIONS 380 Acr, acrolein; DNP-, dinitrophenylhydrazo-; DNPH, 2,4-dinitrophenylhydrazine; FW, fresh 381 weight; GSH, the reduced form of glutathione; HNE, 4-hydroxy-(E)-2-nonenal; LOOH, lipid 382 peroxide; RCS, reactive carbonyl species; ROS, reactive oxygen species. 383 384 AUTHOR CONTRIBUTION 385 Conceptualization: J. M. and D.S.; Investigation: A.H., N.T.; Methodology: J.M., C.N., A.H., 386 N.T., Y.M., K.M.; Roles/Writing - original draft: A.H., N.T., Writing - review & editing; A.H., 387 J.M., Funding acquisition: J.M., D.S.; Supervision: J.M. 388 389 ACKNOWLEDGMENT 390 The authors would like to express their gratitude to Ryoma Oishi and Suzuka Monden for their 391 technical assistance and to Prof. Toshihiro Murafuji, Yamaguchi University, for discussion. 392 393 FUNDING 394 This work was supported by the Ministry of Agriculture, Forestry and Fisheries of Japan , by 395 KAKENHI Grants (nos. 17K19909 and 20H03278) from the Japan Society for the Promotion 396 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 26 of Science (to J. M.), and by Yamaguchi University Project for Formation of the Core Research 397 Project. This work resulted from the use of research equipment shared in the MEXT Project 398 for promoting public utilization of advanced research infrastructure (Program for supporting 399 the construction of core facilities) Grant Number JPMXS0440400022. 400 401 FIGURE CAPTIONS 402 Fig 1. Acr-scavenging ability contents of plant materials. Plant extracts (80% ethanol) were 403 deproteinized, neutralized, and tested for the Acr-scavenging ability as in Materials and 404 Methods. Average of two runs. The phylogenic tree was constructed using the Angiosperm 405 Phylogeny Poster (Cole et al. 2019) as a guide. 406 Fig. 2. LC-MS/MS evidence that alliin is the Acr-scavenging substance in fraction #8 using a 407 Unizon UK-Amino Column. A: Total ion chromatograms (m/z range 100–1000) of fraction 408 #8 with 2 mM Acr (red trace) and without it (black). The MS/MS spectrum of the m/z 409 178.0528 signal at 15.4 min is shown on the right. Inset: expanded chromatogram of the 15.1 410 min peak. B: Total ion chromatograms (m/z range 100–1000) of 1 mM alliin with 2 mM Acr 411 (red trace) and without it (black). The MS/MS spectrum of the m/z 178.0532 signal at 15.4 412 min (arrowhead) is shown on the right. Asterisks show possible Acr reaction products with 413 the scavenging compound (A) or with alliin (B). 414 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 27 Fig. 3. LC-MS/MS results for alliin-Acr adducts. An authentic alliin preparation (5 mM) was 415 mixed with various concentrations of Acr, incubated for 1 h, and subjected to LC -MS/MS 416 analysis. A: Extracted ion chromatograms of the mono-Acr adduct of alliin (m/z 234.0795). B: 417 Extracted ion chromatograms of the (Acr)2 adduct ( m/z 290.1056). C : Extracted ion 418 chromatograms of the dehydrated products from the di -Acr adduct ( m/z 272.0950). The 419 corresponding parent ion MS/MS spectrum is shown at the bottom of panels A–C. 420 Fig. 4. Possible reactions of alliin with Acr. 421 422 ASSOCIATED CONTENTS 423 Supporting information is available. 424 Two chromatographic results for the purification of Acr-scavenging substances from garlic 425 extract (Fig. S1), MS spectrum of the 15.42 min peak of fraction #8 (Fig. S2), LC-MS/MS 426

for the reaction products of isoalliin and Acr (Fig. S3), and LC-MS/MS results for the 427 reaction products of methiin and Acr (Fig. S4). 428 429 430 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 28

431 G. Aldini, M. Carini, G. Beretta, S. Bradamante, R. M. Facino, Carnosine is a quencher of 4-432 hydroxy-nonenal: through what mechanism of reaction? Biochem. Biophys. Res. 433 Commun. 298 (2002) 699-706, https://doi.org/10.1016/S0006-291X(02)02545-7. 434 G. Aldini, I. Facino, G. Beretta, M. Carini, Carnosine and related dipeptides as quenchers of 435 reactive carbonyl speices: From structural studies to therapeutic perspecitves, 436 BioFactors 24 (2005) 77-87, https://dx.doi.org/10.1002/biof.5520240109. 437 G. Aldini, I. Dalle -Donne, R. M. Facino, A. Milzani , M. Carini, Intervention strategies to 438 inhibit protein carbonylation by lipoxidation-derived reactive carbonyls, Med. Res. Rev. 439 27 (2007) 817-868, https://doi.org/10.1002/med.20073. 440 G. Aldini, M. Orioli, R. Rossoni, F. Savi, P. Braidotti, G. Vistoli, K.-J. Yeum, G. Negrisoli, M. 441 Carini, The carbonyl scavenger carnosine ameliorates dyslipidaemia and renal function 442 in Zucker obese rats. J. Cell. Mol. Med. 15 (2011) 1339-1354, 443 https://doi.org/10.1111/j.1582-4934.2010.01101.x. 444 K. Asada, Production and scavenging of reactive oxygen species in chloroplasts and their 445 functions, Plant Physiol. 141 (2006) 391-396, https://doi.org/10.1104/pp.106.082040. 446 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 29 M. S. Biswas, J. Mano, Lipid peroxide-derived short-chain carbonyls mediate hydrogen 447 peroxide-induced and salt-induced programmed cell death in plants, Plant Physiol. 448 168 (2015) 885–898, https://doi.org/10.1104/pp.115.256834. 449 M. Carini, G. Aldini, G. Beretta, E. Arlandini, R. M. Facino, Acrolein-sequestering ability of 450 endogenous dipeptides: characterization of carnosine and homocarnosine/acrolein 451 adducts by electrospray ionization tandem mass spectrometry, J. Mass Spectr. 38 452 (2003) 996-1006, https://doi.org/10.1002/jms.517. 453 T. C. Cole, H. H. Hilger, P. Stevens, Angiosperm Phylogeny Poster (APP) – Flowering plant 454 systematics, PeerJ Preprints 7 (2019) e2320v6, 455 https://doi.org/10.7287/peerj.preprints.2320v6 456 M. Colzani, L. Regazzoni, A. Criscuolo, G. Baron, M. Carini, G. Vistoli, Y.-M. Lee, S. I. Han, 457 G. Aldini, K.-J. Yeum, Isotopic labelling for the characterisation of HNE-sequestering 458 agents in plant-based extracts and its application for the identification of anthocyanidins 459 in black rice with giant embryo , Free Rad. Res . 52 (2018) 896-906, 460 https://doi.org/10.1080/10715762.2018.1490735. 461 H. Esterbauer, H. Zollner , N. Scholtz, Reaction of glutathione with conjugated carbonyls, Z. 462 Naturforsch. 30c (1975) 466-473, http://doi.org/10.1515/znc-1975-7-808. 463 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 30 H. Esterbauer, R. J. Schaur H. Zollner, Chemistry and biochemistry of 4 -hydroxynonenal, 464 malonaldehyde and related aldehydes , Free Rad. Biol. Med. 11 (1991) 81-128, 465 https://doi.org/10.1016/0891-5849(91)90192-6. 466 C. H. Foyer, G. Noctor, Redox sensing and signalling associated with reactive oxygen in 467 chloroplasts, peroxisomes and mitochondria, Physiol. Plant. 119 (2003) 355 -364, 468 https://doi.org/10.1034/j.1399-3054.2003.00223.x. 469 Q. Huang, Y. Zhu, L. Lv, S. Sang, Translating in vitro acrolein -trapping capacities of tea 470 polyphenol and soy genistein to in vivo situations is mediated by the bioavailability and 471 biotransformation of individual polyphenols, Mol. Nutr. Food Res. 64 (2020) 1900274, 472 https://doi.org/10.1002/mnfr.201900274. 473 K. Jiang, Z. Yin, P. Zhou, H. Guo, C. Huang, G. Zhang, W. Hu, S. Ou, J. Ou, The scavenging 474 capacity of g-aminobutyric acid for acrolein and the cytooxicity of the formed adduct, 475 Food Funct. 11 (2020a) 7736, https://doi.org/10.1039/C9FO02518A. 476 X. Jiang, D. Zhang, Y. Lu, L. Lv, Acrolein-trapping mechanism of theophylline in green tea, 477 coffee, and cocoa: Speedy and successful, J. Agric. Food Chem. 68 (2020b) 9718-9724, 478 https://doi.org/10.1021/acs.jafc.0c03895. 479 M. Kanehisa, S. Goto, KEGG: Kyoto Encyclopedia of Genes and Genomes, Nucleic Acids Res. 480 28 (2000) 27-30, https://doi.org/10.1093/nar/28.1.27. 481 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 31 M. Knieper, A. Veihhause, K.-J. Dietz, Oxylipins and reactive carbonyls as regulators of the 482 plant redox and reactive oxygen species network under stress. Antioxidants 12 (2023) 483 814, https://doi.org/10.3390/antiox12040814. 484 Y. Liu, G. Xu, L. M. Sayre, Carnosine 1nhibits ( E)-4-hydroxy-2-nonenal-induced protein 485 cross-linking:  structural characterization of carnosine−HNE adducts, Chem. Res. 486 Toxicol. 16 (2003) 1589-1597, https://doi.org/10.1021/tx034160a. 487 J. Mano, Reactive carbonyl species: their production from lipid peroxides, action in 488 environmental stress, and the detoxification mechanism, Plant Physiol. Biochem. 59 489 (2012) 90-97, https://doi.org/10.1016/j.plaphy.2012.03.010. 490 J. Mano, Y. Torii, S. Hayashi, K. Takimoto, K. Matsui, K. Nakamura, D. Inzé, E. 491 Babiychuk, S. Kushnir, K. Asada, The NADPH:quinone oxidoreductase P1-z-492 crystallin in Arabidopsis catalyzes the a,b-hydrogenation of 2-alkenals: detoxication 493 of the lipid peroxide-derived reactive aldehydes, Plant Cell Physiol. 43 (2002) 1445-494 1455, https://doi.org/10.1093/pcp/pcf187. 495 J. Mano, E. Belles-Boix, E. Babiychuk, D. Inzé, E. Hiraoka, K. Takimoto, L. Slooten, K. 496 Asada, S. Kushnir, Protection against photooxidative injury of tobacco leaves by 2-497 alkenal reductase. Detoxication of lipid peroxide-derived reactive carbonyls, Plant 498 Physiol. 139 (2005) 1773–1783, https://doi.org/10.1104/pp.105.070391. 499 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 32 J. Mano, K. Tokushige, H. Mizoguchi, H. Fujii, S. Khorobrykh, Accumulation of lipid 500 peroxide-derived, toxic α,β-unsaturated aldehydes (E)-2-pentenal, acrolein and (E)-2-501 hexenal in leaves under photoinhibitory illumination. Plant Biotechnol. 27 (2010) 502 193–197, https://doi.org/10.5511/plantbiotechnology.27.193. 503 J. Mano, A. Ishibashi, H. Muneuchi, C. Morita, H. Sakai, M. S. Biswas, T. Koeduka, S. 504 Kitajima, Acrolein-detoxifying isozymes of glutathione transferase in plants, Planta 505 245 (2017) 255-264, https://doi.org/10.1007/s00425-016-2604-5. 506 J. Mano, S. Kanameda, R. Kuramitsu, N. Matsuura, Y. Yamauchi, Detoxification of reactive 507 carbonyl species by glutathione transferase Tau isozymes, Front. Plant Sci. 10 (2019a) 508 487, https://doi.org/10.3389/fpls.2019.00487. 509 J. Mano, M. S. Biswas, K. Sugimoto, Reactive carbonyl species: A missing link in ROS 510 signaling, Plants 8 (2019b) 391, https://doi.org/10.3390/plants8100391. 511 L. A. Mueller, P. Zhang, S. Y. Rhee, AraCyc: A biochemical pathway database for Arabidopsis. 512 Plant Physiol.,132 (2003) 453-460, https://doi.org/10.1104/pp.102.017236. 513 Z. Nurbekova, S. Srivastava, D. Standing, A. Kurmanbayeva, A. Bekturova, A. Soltabayeva, 514 D. Oshanova, V. Turečková, M. Strnad, M. S. Biswas, J. Mano, M. Sagi, Arabidopsis 515 aldehyde oxidase 3, known to oxidize abscisic aldehyde to ABA, protects leaves from 516 aldehyde toxicity, Plant J. 108 (2021) 1439-1455, https://doi.org/10.1111/tpj.15521. 517 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 33 T. Roach, W. Stöggl, T. Baur, I. Kranner, Distress and eustress of reactive electrophiles and 518 relevance to light stress , Free Rad. Biol. Med., 122 (2018) 65 -73, 519 https://doi.org/10.1016/j.freeradbiomed.2018.03.030. 520 E. Schauenstein, H. sterbauer, H. Zollner, Aldehyde in Biological Systems: Their Natural 521 Occurrence and Biological Activities; Pion, 1977, pp. 589–590. 522 X. Song, Y. Lu, Y. Lu, L. Lv, Adduct formation of acrolein with cyanidin-O-glucoside and its 523 degradants/metabolites during thermal processing or in vivo after consumption of red 524 bayberry, J. Agric. Food Chem. 69 (2021) 13143-13154, 525 https://doi.org/10.1021/acs.jafc.1c05727. 526 J. Spaas, W. M. A. Franssen, C. Keytsman, L. Blancquaert, T. Vanmierlo, J. Bogie, B. Broux, 527 N. Hellings, J. van Horssen, D. K. Posa, D. Hoetker, S. P. Baba, W. Derave, B. O. 528 Eijnde, Carnosine quenches the reactive carbonyl acrolein in the central nervous system 529 and attenuates autoimmune neuroinflammation , J. Neuroinflam., 18 (2021) 255, 530 https://doi.org/10.1186/s12974-021-02306-9. 531 S. Srivastava, C. Brychkova, D. Yarmolinsky, A. Soltabayeva, T. Samani, M. Sagi, Aldehyde 532 oxidase 4 plays a critical role in delaying silique senescence by catalyzing aldehyde 533 detoxification, Plant Physiol. 173 (2017) 1977–1997, 534 https://doi.org/10.1104/pp.16.01939. 535 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 34 A. Stoll, E. Seebeck, Über Alliin, die genuine Muttersubstanz des Knoblauchöls. 1. Mitteilung 536 über Allium -Substanzen, Helv. Chim. Acta 31 (1948) 189-210, 537 https://doi.org/10.1002/hlca.19480310140. 538 K. Sugimoto, Y. Matsuoka, K. Sakai, N. Fujiya, H. Fujii, J. Mano, Catechins in green tea 539 powder (matcha) are heat -stable scavengers of acrolein, a lipid peroxide -derived 540 reactive carbonyl species , Food Chem. 355 (2021) 129403, 541 https://doi.org/10.1016/j.foodchem.2021.129403. 542 M. S. Sultana, S. Yamamoto, M. S. Biswas, C. Sakurai, H. Isoai, J. Mano, Histidine-containing 543 dipeptides mitigate salt stress in plants by scavenging reactive carbonyl species, J. Agric. 544 Food Chem. 70 (2022) 11169-11178, https://doi.org/10.1021/acs.jafc.2c03800. 545 M. S. Sultana, C. Sakurai, M. S. Biswas, L. Szabados , J. Mano, Accumulation of reactive 546 carbonyl species in roots as the primary cause of salt stress-induced growth retardation 547 of Arabidopsis thaliana , Physiol. Plant. 176 (2024) e14198, 548 https://doi.org/10.1111/ppl.14198. 549 R. Sunker, D. Bartels, H.-H. Kirch, Overexpression of a stress-inducible aldehyde 550 dehydrogenase gene from Arabidopsis thaliana in transgenic plants improves stress 551 tolerance, Plant J. 35 (2003) 452–464, https://doi.org/10.1046/j.1365-552 313X.2003.01819.x. 553 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 35 Z.-h. Tao, C. Li, X.-f. Zu, Y.-j. Pan, Scavenging activity and mechanism study of ferulic acid 554 against reactive carbonyl species acrolein, J. Zhejiang Univ.-Science B 20 (2019) 868-555 876, https://doi.org/10.1631/jzus.B1900211. 556 Y. Ueda, H. Kawajiri, N. Miyamura, R. Miyajima, Content of some sulfur -containing 557 components and free amino acids in various strains of garlic, Nippon Shokuhin Kogyo 558 Gakkaishi 38 (1991) 429-434, https://doi.org/10.3136/nskkk1962.38.429. 559 T. Uemura, M. Uchida, M. Nakamura, M. Shimekake, A. Sakamoto, Y. Terui, K. Higashi, I. 560 Ishii, K. Kashiwagi, K. Igarashi , A search for acrolein scavengers among food 561 components, Amino Acids 55 (2023) 509 -518, https://doi.org/10.1007/s00726 -023-562 03248-7. 563 W. Wang, Y. Qi, J. R. Rocca, P. J. Sarnoski, A. Jia, L. Gu, Scavenging of toxic A crolein by 564 resveratrol and hesperetin and identification of adducts, J. Agric. Food Chem. 63 (2015) 565 9488-9495, https://doi.org/10.1021/acs.jafc.5b03949. 566 R. S. Wible, T. R. Sutter, Soft cysteine signaling network: the functional significance of 567 cysteine in protein function and the soft acids/bases thiol chemistry that facilitates 568 cysteine modification, Chem. Res. Toxicol. 30 (2005) 729 -762, 569 https://doi.org/10.1021/acs.chemrestox.6b00428. 570 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 36 Y. Yamauchi, A. Hasegawa, A. Taninaka, M. Mizutani, Y. Sugimoto, NADPH-dependent 571 reductases involved in the detoxification of reactive carbonyls in plants, J. Biol. 572 Chem. 286 (2011) 6999–7009, https://doi.org/10.1074/jbc.M110.202226. 573 Y. Yamauchi, A. Hasegawa, M. Mizutani, Y. Sugimoto, Chloroplastic NADPH-dependent 574 alkenal/one oxidoreductase contributes to the detoxification of reactive carbonyls 575 produced under oxidative stress, FEBS Lett. 586 (2012) 1208-1213, 576 https://doi.org/10.1016/j.febslet.2012.03.013. 577 Y. Yamazaki, K. Iwasaki, M. Mikami, A. Yogihashi, Distribution of eleven flavor precursors, 578 S-alk(en)yl-L-cysteine derivatives in seven Allium vegetables, Food Sci. Technol. 579 Res. 17 (2011) 55-62, https://doi.org/10.3136/fstr.17.55. 580 L. Yin, J. Mano, S. Wang, W. Tsuji, K. Tanaka, The involvement of lipid peroxide -derived 581 aldehydes in aluminum toxicity of tobacco roots, Plant Physiol. 152 (2010) 1406–1417, 582 https://doi.org/10.1104/pp.109.151449. 583 D. Zhang, Z. Jiang, L. Xiao, Y. Lu, S. Sang, L. Lv, Mechanistic studies of inhibition on 584 acrolein by myricetin , Food Chem. 323 (2020) 126788, 585 https://doi.org/10.1016/j.foodchem.2020.126788. 586 Q. Zhu, Z.-P. Zheng, K. W. Cheng, J. J. Wu, S. Zhang, Y. S. Tang, K. H. Sze, J. Chen, F. Chen, 587 M. Wang, Natural polyphenols as direct trapping agents of lipid peroxidation -derived 588 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint 37 acrolein and 4 -hydroxy-trans-2-nonenal, Chem. Res. Toxicol. 22 (2009) 1721-1727, 589 https://doi.org/10.1021/tx900221s. 590 Z. Zou, Z. Yin, J. Ou, J. Zheng, F. Liu, C. Huang, S. Ou, Identification of adducts formed 591 between acrolein and alanine or serine in fried potato crips and the cytotoxicity -592 lowering effect of acrolein in three cell lines , Food Chem. 361 (2021) 130164, 593 https://doi.org/10.1016/j.foodchem.2021.130164. 594 595 Declaration of generative AI and AI-assisted technologies in the writing process 596 During the preparation of this work the authors used DeepL, ChatGPT and Google Gemini in 597 order to improve the English. After using this tool/service, the authors reviewed and edited the 598 content as needed and take full responsibility for the content of the publication. 599 600 601 602 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 22, 2025. ; https://doi.org/10.1101/2025.08.18.670978doi: bioRxiv preprint