Hydrothermal treatment of chestnut shells (Castanea sativa) to produce oligosaccharides and antioxidant compounds

1 Hydrothermal treatment of chestnut shells (Castanea sativa) to produce oligosaccharides 1 and antioxidant compounds 2 Beatriz Gullón1, Gemma Eibes*1, Izaskun Dávila2, María Teresa Moreira1, Jalel Labidi2, Patricia 3 Gullón2 4 1Department of Chemical Engineering, Institute of Technology, Universidade de Santiago de Compostela, 5 15782 Santiago de Compostela, Spain 6 2Chemical and Environmental Engineering Department, University of Basque Country, 20018 San 7 Sebastián, Spain 8 *E-mail address: [email protected] 9 Abstract 10 Hydrothermal treatment is an environmentally friendly technology that allows the solubilisation 11 of hemicellulosic oligosaccharides with potential for their use as prebiotics. The purpose of this 12 study was to solubilize oligosaccharides and antioxidant compounds from chestnut shells by a 13 hydrothermal processing. The highest content of oligosaccharides (18.3 g/L), with a relatively 14 low level of monosaccharides (2.4 g/L) and degradation products (0.5 g/L) was obtained at 180 15 ºC (severity of 3.08). In addition, the liquors presented a high content of phenolic and flavonoid 16 compounds with good antioxidant properties. The GC-MS revealed that the most abundant 17 phenolic compound was pyrogallol (13.2%). The molecular weight distribution of the 18 solubilization products showed that a 26.5% presented an apparent Mw of 6077 g/mol and a 19 73.5% presented an apparent Mw of 586 g/mol with a high polydispersity index. MALDI-TOF, 20 FTIR, and TGA analyses revealed structural information of these compounds and their thermal 21 stability. 22 23 Keywords: autohydrolysis, oligosaccharides, antioxidant activity, structural characterization, 24 chestnut shells 25 26 27 28 2 1. Introduction 29 Changes in the lifestyle of the first world society are bringing about a shift in eating 30 habits and trends in food manufacture and consumption, which have an impact on health, 31 environment and society (Cencic & Chingwaru, 2010; Isanga, & Zhang, 2007). In this context, 32 the demand for fast and ready-made food is constantly increasing. Opposite to this trend, 33 consumers are increasingly attentive to food safety, quality and health-related issues (Cencic & 34 Chingwaru, 2010). In this sense, the growing awareness of the relationship between health and 35 nutrition has fostered the search for and isolation of bioactive substances as a way to counteract 36 unbalanced diets (Antov & Ðordevic, 2017; Cencic & Chingwaru, 2010). In particular, there is 37 a growing interest in the search of new sources to obtain bioactive compounds. 38 In this perspective, undervalued by-products generated from the industrial processing of 39 agro-industrial feedstocks can potentially be exploited as an inexpensive and renewable source 40 of bio-compounds (Mandelli et al., 2014). Therefore, the utilization of agro-industrial by-41 products needs to find suitable approaches with a double beneficial effect: 1) waste 42 management; 2) valorization of added value products (Morana et al., 2017). 43 Among the processing industry, in recent decades, the chestnut (Castanea sativa) 44 industry has grown significantly in Europe, especially for the production of marron glace and 45 chestnut flour, the latter used as ingredient in gluten-free diets (Vella, Laratta, La Cara, & 46 Morana, 2017). During the chestnut peeling process, the chestnut shell is removed, a fraction 47 that represents between 10 and 15% of the weight of the whole chestnut (Vázquez, Mosquera, 48 Freire, Antorrena, & González-Álvarez, 2012). Chestnut shells are lignocellulosic materials 49 comprised of major constituents: cellulose, hemicelluloses and lignin (González-López, Moure, 50 Domínguez, & Parajó, 2012), with different types of functional groups which can be binding 51 sites for ion exchange and complexation reactions (Vázquez et al., 2012). 52 Nowadays, chestnut shells are currently used as fuel, but their extracts have been 53 reported to have antioxidant activity due to their high polyphenolic content (Vázquez et al. 54 2008). In addition, the moderate amount of hemicelluloses (González-López et al., 2012) can be 55 a remarkable substrate for obtaining non digestible oligosaccharides (NDO) described as 56 3 potential prebiotic substrates, but to our best knowledge, their exploitation for this application 57 has not yet been considered. 58 A prebiotic is “a selectively fermented ingredient, or a fiber that allows for specific 59 changes in both the composition and/or activity of the gastrointestinal microbiota, conferring 60 benefits on the well-being and health of host” (Roberfroid et al., 2010). Moreover, prebiotics 61 promote other indirect effects such as immunological, anti-inflammatory, anticancer, 62 antiallergic action and also improve intestinal function and bioavailability of calcium (Aachary 63 & Prapulla, 2011; Azevedo-Carvalho, de Oliva Neto, da Silva, & Pastore, 2013). Among the 64 different oligosaccharides, xylooligosacharides (XOS) provide additional benefits beyond 65 prebiotic properties: moderate degree of sweetness, effect on starch retrogradation, stability of 66 pH and temperature and improved nutritional and sensory properties of food (Voragen, 1998). 67 All of these properties make xylooligosaccharides suitable compounds to be incorporated into 68 food (Ayyappan et al. 2016). 69 The production of NDO from agricultural residues has received much attention due to 70 their high availability and low cost. Between the different techniques that could be used to 71 obtain NDO, hydrothermal treatment, also called autohydrolysis or liquid hot water, is a low 72 cost and environmentally friendly technology that allows the solubilization of compounds to be 73 used as prebiotics (Quitain, Sato, Daimon, & Fujie, 2003), avoiding the use of chemicals. 74 Oligosaccharides obtained by hydrothermal treatment from different raw materials has been 75 exploited in recent years. Gullón et al. (2008) and Dávila, Gordobil, Labidi, and Gullón (2016) 76 obtained xylooligosaccharides from rice husks and vine shoots by an autohydrolysis treatment, 77 respectively, while Gullón, Yáñez, Alonso, and Parajó (2010), Rostro et al. (2014) and Rico, 78 Gullón, Alonso, Parajó and Yáñez (2018) reported the production of oligosaccharides from rye 79 straw, maize pericarp or peanut shells, respectively. The liquors from the hydrothermal process, 80 in addition to the presence of oligosaccharides, have also been reported to contain compounds 81 with antioxidant activity. Gullón et al. (2017) studied the influence of the severity of the 82 hydrothermal treatment on the antioxidant activity of autohydrolysis liquors from vine shoots. 83 Rico et al. (2018) also studied the antioxidant properties of the solubilized products from peanut 84 4 shells, while Moure, Conde, Falqué, Domínguez, and Parajó (2014) evaluated the purified 85 extracts of autohydrolysis liquors of chestnut burs. 86 The goal of this work was the solubilization of oligosaccharides and antioxidant 87 compounds from chestnut shells by a hydrothermal processing, since it has not been exploited 88 yet. The chestnut shells were hydrothermally processed under different severities to determine 89 the conditions that allowed maximum solubilization of oligosaccharides. The effect of treatment 90 severity on the composition of the liquid and solid phases resulting from hydrothermal 91 treatments, as well as the antioxidant activity of autohydrolysis liquors were analyzed. 92 Solubilized hemicellulose under optimal autohydrolysis conditions were characterized 93 by techniques such as FTIR, TGA, HPSEC and MALDI-TOF. The GC-MS analysis of ethyl 94 acetate extracts allowed the identification of compounds derived from sugar and lignin or 95 extraction-derived substances that provide antioxidant activity to the liquor. This approach 96 could be the first stage of a new integrated biorefinery for chestnut shells that aims at a suitable 97 revalorization strategy for this unexploited waste biomass. 98 2. Materials and methods 99 2.1. Raw material and chemicals 100 The chestnut shells used in this work were supplied by a chestnut processing plant 101 (Cuevas & Cia S.A., San Cibrao das Viñas, Spain) after being obtained by a slow steam peeling 102 process. The collected shells were air dried, milled and sieved to achieve a particle size of less 103 than 0.4 mm. The milled shells were mixed to acquire a single batch and avoid aliquot 104 variations. The lot was stored at room temperature in a dark and dry place until its further use. 105 Sulphuric acid (95-97%), glucose (≥99.5%), xylose (≥99%), arabinose (≥99%), furfural 106 (99%), hydroxymethylfurfural (≥98%), ethyl acetate (99.8%), gallic acid (≥98%), rutin (95%), 107 trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid), Folin-Ciocalteu reagent, 108 ABTS (2,2'-azino-di(3-ethylbenzothiazoline-6-suslfonic acid; 98%), TPTZ (2,4,6-tri(2-pyridyl)-109 S-triazine; 99%), DPPH (2,2-diphenyl-1-picrylhydrazyl; 97%), sodium carbonate, sodium 110 acetate 3-hydrate, potassium persulfate, acetic acid, hydrochloric acid and iron(III) chloride 111 5 hexahydrate were obtained from Sigma-Aldrich (Barcelona, Spain), pullulan polysaccharides 112 (≥98%) were purchased from Varian (England). 113 2.2. Characterization of the raw material 114 Milled chestnut shells were subjected to moisture (TAPPI T264- om-88), ash (TAPPI 115 T244-om-93) and ethanol-toluene extractives (TAPPI T204 cm-97) determination. The chestnut 116 shells without extractives were milled and sieved to obtain a particle size of less than 0.25 mm 117 in order to subject them to a quantitative acid hydrolysis with 72% (w/w) H2SO4 (TAPPI T-249-118 cm-09) for the determination of the hemicellulosic, glucan and lignin content. The solid phases 119 recovered by filtration after oven-drying were considered as Klason lignin, while the liquid 120 phases were analyzed by High Performance Liquid Chromatography (HPLC) for the 121 determination of monosaccharides (glucose, xylose and arabinose), galacturonic acid and acetic 122 acid and degradation products: furfural (F) and hydroxymethylfurfural (HMF). A Jasco LC Net 123 II/ADC chromatograph equipped with a refractive index detector was used for these analyses. 124 20 µL of the samples were eluted with a flow rate of 0.6 mL/min of H2SO4 0.005 M through an 125 Aminex HPX-87H 300 x 7.8 mm (Bio-Rad Laboratories, USA) column at 50 ºC. All analyses 126 were performed in duplicate. 127 2.3. Hydrothermal processing of the chestnut shells 128 The autohydrolysis treatments of the milled chestnut shells were carried out in non-129 isothermal regimen at different temperatures (from 170-220 ºC). This residue was mixed with 130 water in a liquid/solid ratio of 8 kg/kg (oven dried basis) in a 1.5 L stainless steel reactor using a 131 Parr PID controller to control temperature. Once the reactor was cooled down, the liquid and 132 solid phases were separated by filtration, the liquid phases being stored at 4 ºC until their later 133 use and the solid phases dried at room temperature after being washed. 134 With the purpose of facilitating the comparison of working conditions, the severity (S0) 135 of treatments was determined. This parameter, which is defined as the logarithm of the severity 136 factor (R0), considers the effect caused by time and temperature throughout the non-isothermal 137 6 treatment, including both the heating and cooling period. This parameter is expressed by the 138 following equation (Rico et al., 2018): 139 𝑆𝑆 0 =𝑙𝑙𝑙𝑙𝑙𝑙𝑅𝑅 0 =𝑙𝑙𝑙𝑙𝑙𝑙�𝑅𝑅 0𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝑙𝑙 +𝑅𝑅 0𝐶𝐶𝑙𝑙𝑙𝑙𝑙𝑙𝐻𝐻𝐻𝐻𝑙𝑙 �=𝑙𝑙𝑙𝑙𝑙𝑙�� 𝐻𝐻 �𝑇𝑇(𝐻𝐻)−𝑇𝑇𝑅𝑅𝐻𝐻𝑅𝑅 𝜔𝜔� 𝑑𝑑𝐻𝐻+� 𝐻𝐻 �𝑇𝑇´(𝐻𝐻)−𝑇𝑇𝑅𝑅𝐻𝐻𝑅𝑅 𝜔𝜔� 𝑑𝑑𝐻𝐻 𝐻𝐻𝐹𝐹 𝑇𝑇𝑇𝑇𝐻𝐻𝑇𝑇 𝑇𝑇𝑇𝑇𝐻𝐻𝑇𝑇 0 � [1] where tMax is the time (min) required to achieve the maximum temperature of each 140 hydrothermal treatment (TMax, ºC); tF is the time (min) of the entire heating–cooling cycles; T(t) 141 and T’(t) (ºC) are the temperature profiles in the heating and cooling processes, respectively, 142 and ɷ and TRef are parameters whose values have been reported in the literature (ω = 14.75 ºC; 143 TRef = 100 ºC). 144 2.4. Chemical characterization of the spent solids from hydrothermal treatment 145 The spent solids from the hydrothermal treatments after air-drying were subjected to 146 gravimetric and moisture analyses to determine the solid yield and the solubilization of the raw 147 material. The composition of the pre-treated solids was analyzed by a quantitative acid 148 hydrolysis, as described for the raw material in Section 2.2. 149 2.5. Chemical characterization of the liquid phases of hydrothermal treatment 150 The content of monosaccharides (glucose, xylose and arabinose), acetic and 151 galacturonic acids and degradation products (furfural and hydroxymethylfurfural) in the liquid 152 phase was analyzed by HPLC as described in Section 2.2. In addition, aliquots of each treatment 153 were subjected to post-hydrolysis acid treatment (4% H2SO4 at 121 ºC for 30 min) and the 154 reaction products were quantified by HPLC using the same methodology as described in Section 155 2.2. The increase in the concentrations of monosaccharides, galacturonic and acetic observed 156 during the quantitative post-hydrolysis makes it possible to determine the concentration of 157 oligomers and their degree of substitution by acetyl and galacturonic groups. Oligosaccharides 158 (OS) were expressed as monosaccharide equivalents. 159 An aliquot of the various liquors was oven-dried at 105±2 °C until constant weight, in 160 order to determine the content of non-volatile compounds (NVC). The quantification of non-161 7 volatile compounds that were not saccharides (called other non-volatile compounds, ONVC) 162 present in the autohydrolysis liquors was carried out by the difference between NVC and 163 saccharides (considering monosaccharides, OS and OS substituents) (Gullón et al., 2010). 164 Apart from the components determined above, the acid soluble lignin (ASL) solubilized 165 by the hydrothermal treatment was quantified (TAPPI UM 250 um-83 method). An aliquot of 166 the autohydrolysis liquors obtained at different temperatures was diluted with H2SO4 1 M until 167 the absorbance measured at 205 nm with an UVmini-1240 spectrophotometer (Shimadzu 168 Corporation) was between 0.1 and 0.8. 169 [2] 170 Where Abs250nm is the absorption (relative to 1 M H2SO4 at 205 nm), DF is the dilution 171 factor, VF is the volume of the hydrothermally liquor, ε is the absorptivity of lignin at this 172 wavelength (110 L/g·cm), DMi is the weight of the raw material used in the autohydrolysis (g as 173 100% dry matter) and l (cm) is the length of solution the light passes through. 174 2.6. Total phenolic content (TPC) and total flavonoid content (TFC) determination 175 The liquors of chestnut shells were evaluated for total phenolic content (TPC) using the 176 Folin-Ciocalteau method (Singleton and Rossi, 1965) and expressed as g of gallic acid 177 equivalents (GAE)/L of autohydrolysis liquors. The total flavonoid content (TFC) of liquors 178 was quantified by the colorimetric method of aluminum chloride described by Blasa et al. 179 (2005). TFC was recorded in g of rutin equivalents (RE)/L of autohydrolysis liquors. All 180 measurements were made in triplicate. 181 2.7. Antioxidant activity 182 Antioxidant activity was evaluated using three complementary methods: DPPH (α,α-183 Diphenyl-β-picrylhydrazyl radical scavenging assay), ABTS (2,2-azino-bis-3-184 ethylbenzothiazoline-6-sulphonic acid) and FRAP (ferric reducing antioxidant power) according 185 to the procedures described in the literature by Gullón et al. (2017). For all antioxidant activity 186 8 assays, Trolox was used as standard and results were expressed as g of Trolox equivalents 187 (TE)/L of autohydrolysis liquors as mean of three replicates. 188 2.8. Chemical and structural characterization of the solubilization products from 189 hemicelluloses of chestnut shells 190 Solubilized oligosaccharides at the optimal temperature of autohydrolysis were freeze-191 dried and subjected to different analytical techniques to obtain detailed information on their 192 chemical and structural characteristics. 193 2.8.1. Fourier transform infrared spectroscopy (FTIR) 194 The chemical groups and bonding arrangement of constituents present in the 195 oligosaccharides were determined by FTIR in a PerkinElmer Spectrum Two FT-IR spectrometer 196 working in transmission mode with a resolution of 4 cm-1 and accumulating a total of 8 scans. 197 2.8.2. Thermogravimetric analysis (TGA) 198 Thermal stability was evaluated by a TGA/SDTA RSI 851 Mettler Toledo analyzer. 199 Between 3 and 5 mg of freeze-dried liquors were tested under nitrogen atmosphere at a heating 200 rate of 10 ºC/min from 25 ºC to 800 ºC. 201 2.8.3. Molecular weight distribution analysis 202 The molecular weight distribution of oligosaccharides was estimated by High 203 Performance Size Exclusion Chromatography (HPSEC) using a Jasco LC Net II/ADC 204 chromatograph equipped with a refractive index (RI) detector was used. 40 µL of the sample 205 was eluted with a flow rate of 0.6 mL/min of 0.005 N H2SO4 at 40 ºC through a Varian Polymer 206 Laboratories Aquagel-OH mixed-H 8 µm column. The HPSEC calibration was carried out with 207 pullulan polysaccharides with different molecular weights (between 180 and 805000 Da). 208 209 210 9 2.8.4. Matrix assisted laser desorption/ionization-time of flight mass spectroscopy (MALDI-211 TOF MS) 212 The absolute masses of OS were determined by MALDI-TOF MS using an Ultraflex III 213 TOF/TOF mass spectrometer equipped with a Smartbeam® laser (Bruker Daltonics). 214 Measurements were performed in reflectron operating mode and positive polarity. The 215 acceleration voltage was set at 25 kV and a total of 1200 laser shots per spot were automatically 216 acquired. Sample preparation was carried according to the procedure described in Gullón et al. 217 (2014) using 2,5-dihydroxybenzoic acid (DHB) as matrix. 218 2.8.5. Qualitative Analysis of the Ethyl Acetate Soluble Fraction 219 The non-saccharide components present in the liquor obtained under optimum 220 conditions were analyzed by carrying a liquid-liquid extraction with ethyl acetate (EAc). A 221 single extraction stage was carried out by stirring a mixture of the liquor with EAc using a 222 liquor-solvent ratio of 1:3 (v/v) for 15 min. The immiscible phases were separated by 223 decantation, with the organic phase vacuum evaporated at 40 ºC to eliminate the solvent and the 224 dissolved volatile compounds. These extracts were dissolved in EAc and analyzed by Gas 225 Chromatography-Mass Spectrometry (GC-MS) using an Agilent Technologies 7890A gas 226 chromatograph (GC) coupled to an Agilent Technologies 5975C mass spectrometer (MS). 1 µL 227 of the sample was introduced in the GC in split mode and using a flow of 1 mL/min of He it was 228 passed through a column of 30 m x 0.25 mm x 0.25 µm thickness HP-5MS (5% 229 phenylmethylpolysiloxane). The separation method used was the one reported by Gullón et al. 230 (2017). The identification of the compounds was carried out by comparing their mass spectra 231 with those of the National Institute of Standards (NIST) library database and with compounds 232 reported in the literature. Molar peak areas were calculated for compounds with a peak area 233 greater than 0.4% and to determine the relative abundance of the compounds, the sum of the 234 molar peak area was normalized to 100%. 235 236 237 16 Figure 1. Effect of hydrothermal treatment temperature on the composition of liquors: (a) oligosaccharides (b) monosaccharides and (c) degradation compounds. a) b) c) 17 Another aspect evaluated in this work was the antioxidant potential of the liquors from chestnut shells. Table 3 shows the characterization of the different autohydrolysis liquors in terms of antioxidant phenolic compounds. As for total phenolic content (TPC), it reached a maximum concentration at 180 ºC (4.69 g GAE/L, corresponding to 3.9 g GAE/100 g chestnut shell), coinciding with the maximum XOS yield, and then started to decrease, probably due to degradation reactions. A few research works have described the gradual increase in phenolic content with temperature during hydrothermal treatments and was related to the release of phenolic compounds linked to oligosaccharides and the partial depolymerization of lignin (Conde et al., 2011; Gullón et al., 2017). The TPC values reported here show that chestnut shells are a rich source of phenolic compounds. In fact, the extraction of this residue with a 2.5% Na2SO3 solution resulted in a total phenolic content of 13.4 g GAE/100 g oven-dried shells (Vázquez et al., 2008). This difference in the results of both studies could indicate that some of the phenolic compounds from chestnut shells could be degraded at the autohydrolysis temperatures evaluated. Nevertheless, the phenolic content determined in the liquor was substantially higher than those obtained by autohydrolysis treatment of different materials. For example, Jesus et al. (2017) reported a maximum of 2.09 g GAE/L in the autohydrolysis liquor of vine pruning residue (S0=4.13), and Gullón et al. (2017) reached a maximum TPC of 2.25 g GAE/100 g vine shoots at higher temperature (T=215 ºC). Conde et al. (2011) evaluated the phenolic content of autohydrolysis liquors of five different lignocellulosic wastes, and only those of chestnut burs liquors had a higher phenolic content (4.4 g GAE/100 g chestnut bur at 240 ºC) than that achieved in the present study. The same trend was observed for the total flavonoid content (TFC), with the maximum also at 180 ºC (4.82 g RE/L, corresponding to 4.0 g RE/100 g chestnut shell). The flavonoid content in this liquor was 3.6 times higher than that reported for the autohydrolysis liquors of vine shoots (1.1 g RE/100 g, 215 ºC; Gullón et al. 2017). Interestingly, the trend in antioxidant activity depended on the methodology used. Moreover, the DPPH method indicated that the 18 antioxidant activity of the liquor remained almost constant in the range 170 ºC to 200 ºC, and at the highest temperatures, activity dropped slightly to 84% of the initial activity (4.6 g TE/L). The ABTS method, on the other hand, showed a continuous decrease of the antioxidant activity with the autohydrolysis temperature, leading to 49% of the initial activity in the harsher conditions. The antioxidant capacity measured by the FRAP assay showed that liquors generated at 170-180 ºC exhibited similar antioxidant activity, and at higher temperatures, decreased to a final relative value of 69%. Contrary to these observations, Gullón et al. (2017) observed a steady increase of the antioxidant activity of the vine shoots autohydrolysis liquors with temperature, consistent with the rise of the TPC levels. However, the maximum antioxidant activities reported in that work were in the range 2.9-4.4 times lower than those achieved here (1.05, 4.45 and 2.68 g TE/100 g vine shoots, for DPPH, ABTS and FRAP assays, respectively). Since the principal objective of the present research was to obtain a liquor with a high content of both XOS and phenolic antioxidants, the optimum temperature that allowed this objective to be accomplished was 180 ºC. At this temperature, the total oligosaccharide content was 18.3 g/L, monosaccharides accounted for 2.4 g/L, acetic acid as the only degradation product was present at 0.5 g/L and the total phenolic content was 4.7 g GAE/L. In this sense, the mass ratio of oligosaccharide to monosaccharide was 7.6, which is more favorable than the value of 5 determined for peanut shells (Rico et al., 2018). 19 Table 3. Effect of temperature of hydrothermal treatment on the TPC, TFC and antioxidant activity (analyzed by the DPPH, ABTS and FRAP methods) of autohydrolysis liquors. TPC: total phenolic content; TFC: total flavonoid content; GAE: gallic acid equivalents; RE: rutin equivalents; TE: Trolox equivalents. Temperature (ºC) TPC (g GAE/L) TFC (g RE/L) DPPH (g TE/L) ABTS (g TE/L) FRAP (g TE/L) 170 4.49 ± 0.08 4.55 ± 0.24 5.60 ± 0.09 16.27 ± 0.09 9.36 ± 0.32 175 4.43 ± 0.04 4.68 ± 0.09 5.50 ± 0.14 15.66 ± 0.14 9.44 ± 0.21 180 4.69 ± 0.01 4.82 ± 0.07 5.35 ± 0.02 15.17 ± 0.25 9.33 ± 0.33 185 4.17 ± 0.19 4.15 ± 0.18 5.14 ± 0.16 13.99 ± 0.36 8.94 ± 0.05 190 3.88 ± 0.12 3.42 ± 0.06 5.38 ± 0.03 13.43 ± 0.23 8.47 ± 0.05 200 3.25 ± 0.08 2.66 ± 0.03 5.31 ± 0.04 9.89 ± 0.21 7.29 ± 0.16 210 2.89 ± 0.03 2.51 ± 0.04 4.92 ± 0.11 8.49 ± 0.26 6.90 ±0.21 215 2.77 ± 0.05 2.44 ± 0.14 4.63 ± 0.15 8.05 ± 0.18 6.72 ±0.13 220 2.57 ± 0.05 2.44 ± 0.03 4.64 ± 0.05 7.80 ± 0.12 6.31 ±0.14 Values are the mean of two hydrothermal treatments ± standard deviations Figure 2 shows the mass balance of the hydrothermal pretreatment performed at 180 ºC. From 100 kg of dried chestnut shells, 15.11 kg of substituted oligosaccharides (sum of GOS, XOS and ArOS), 1.87 kg of monosaccharides and 0.29 kg of antioxidant compounds can be obtained. In addition, 72.8 kg of spent solid (enriched in cellulose and lignin), that can be used as substrate for a variety of marketable chemicals, were also recovered. 20 Figure 2. Material balances of the products obtained from chestnut shells at the reaction temperature of 180 ºC Abbreviations: NVC, non-volatile compounds; ONVC, other non-volatile compounds; S0, severity 3.3. Structural characterization of the solubilization products from chestnut shells hemicellulose A broader study of the structural characteristics of oligosaccharides was carried out to assess the potentiality of the water-soluble compounds as nutraceuticals. The autohydrolysis liquor from the experiment carried out at 180 ºC (S0=3.08) was analyzed by HPSEC, MALDI TOF, FTIR, and TGA to obtain structural and thermal behavior information for these compounds. This liquor was obtained under the autohydrolysis conditions which led to the maximum concentration of oligosaccharides and antioxidant compounds. 3.3.1. Determination of the molecular weight distribution of solubilized oligosaccharides The molecular weight of the oligosaccharides is an important characteristic that influences their biological properties (Gullón et al., 2014). In this sense, Gullón et al. (2008) AUTOHYDROLYSIS (S0=3.08) Water (789 kg) Chestnut shells (111 kg of material, equivalent to 100 kg of dried chestnut shells) Filtration Spent solid (72.8 kg) Liquors (827.2 kg) Water and volatile reaction products: 804.98 kg NVC: 22.22 kg Monosaccharides: 1.87 kg Substituted oligosaccharides: 15.11 kg Antioxidant compounds: 0.29 kg ONVC: 4.95 kg 21 suggested that xylooligosaccharides with a degree of polymerization (DP) greater than 4 present potential prebiotic applications. In this work, the molecular weight distribution of the oligosaccharides contained in the autohydrolysis liquors was analyzed by HPSEC (Supplementary Figure S1). Values were calculated based on equivalent RI signals of pullulan standards of known Mw. The oligosaccharides could be divided into two fractions: a minor proportion (26.5%) with higher apparent molecular weight (Mw) of 6077 g/mol (DP: 46.03) and the major fraction (73.5%) with a lower apparent Mw of 586 g/mol (DP: 4.43), which can be regarded as low molecular weight oligomers (Tunc and Van Heiningen, 2011). The apparent Mw of the solubilized oligosaccharides from chestnut shells (S0=3.08) is lower than the one for the oligosaccharides from vine shoots (Dávila et al. 2016). Wang et al. (2016) also obtained oligosaccharides with higher apparent Mw (8430 g/mol) during the autohydrolysis of rapeseed straw carried out with an S0 of 3.26. During the autohydrolysis treatment, depolymerization of solubilized hemicellulosic oligosaccharides of high molecular weight into oligomers of low molecular weight takes place, so the difference of the Mw of the oligosaccharides is a consequence of the structure of the raw material. It should be noted that the high polydispersity index (10.25) of the chestnut shells oligosaccharides is due to the high difference of the Mw of the two fractions. As reported in the literature, oligosaccharides with different average degrees of polymerization may present different prebiotic properties. According to Sanchez et al. (2009), shorter-chain-length molecules were primarily fermented in the proximal colon, whereas longer molecules (DP: 29) reached the distal colon. Hence, oligosaccharides with high degrees of polymerization should be considered to prevent total fermentation during transit through previous compartments (Sanchez et al., 2009). 3.3.2. Matrix Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry For more information on the structural features of the solubilized oligosaccharides at the optimum temperature, a MALDI-TOF MS analysis was carried out (Supplementary Figure S2). Table 4 presents the mass signals in MALDI-TOF spectra, along with the main structures identified (the specific compounds were detected as sodium and potassium adducts). The 22 MALDI-TOF-MS spectra profile showed the presence of different types of oligosaccharides that consisted predominantly of chains of pentoses and hexoses. The identified pentaoligomers (which correspond to xylose residues according the composition analysis of the liquors mentioned above) showed a rich substitution pattern, mainly by acetyl and methylglucuronosyl groups, with DP in the range of 2-12. It is important to note that part of the XOS may be substituted with arabinose units. However, arabinose substitutes cannot be detected by MALDITOF-MS as both xylose and arabinose have the same molecular weight. As for hexose oligomers, these included a series of unsubstituted or acetylated oligomers with DP ranging from 4 to 12. The presence of these oligomers can be explained by the depolymerization of cellulose and the degradation of glucan (Wang et al., 2016). The substitution pattern of oligosaccharides affects their prebiotic potential as confirmed in the literature. In this sense, substituted oligosaccharides can display a different fermentation behavior and bioactivity potential (Kabel, Kortenoeven, Schols & Voragen, 2002). The structural characteristics of oligomeric compounds obtained from different raw materials by hydrothermal treatment have been extensively considered in the literature (Rico et al., 2018; Ruiz et al., 2017; Gullón et al., 2014). However, this is the first time that a detailed structural characterization of oligosaccharides from chestnut shells is reported. 23 Table 4. MALDI-TOF results of the autohydrolysis liquors at 180 º C and suggested structures. m/z Structure m/z Structure 627.19 Pent4AcK 1239.38 Hex6Ac5K 669.19 Pent4Ac2K 1265.4 Pent8Ac4Na 711.11 Pent4Ac3K 1281.39 Pent7Ac3MeUrNa 727.10 Pent3Ac2MeUrK 1287.39 Pent8MeUrNa 741.16 Pent2Ac2Ur2Na 1329.44 Pent8AcMeUrNa 743.18 Pent5AcNa 1337.47 Hex8Na 745.15 Pent4UrNa 1353.46 Hex8K 747.18 Hex4AcK 1371.44 Pent8Ac2MeUrNa 759.2 Pent4MeUrNa 1395.45 Hex8AcK 761.19 Pent4UrK 1397.44 Pent9Ac4Na 785.18 Pent5Ac2Na 1413.46 Pent9Ac4K 787.18 Pent4AcUrNa 1419.45 Pent9MeUrNa 789.16 Pent3Ur2Na 1429.45 Pent7Ac2MeUr2Na 849.24 Pent6K 1455.47 Pent8Ac4MeUrNa 867.25 Hex5K 1461.47 Pent9AcMeUrNa 975.27 Pent6Ac3K 1487.51 Pent10Ac3Na 1011.29 Hex4Pent2AcK 1497.49 Pent8Ac5MeUrNa 1013.32 Hex6Na 1499.54 Hex9Na 1029.32 Hex6K 1515.54 Hex9K 1049.30 Pent7Ac2Na 1529.51 Pent10Ac4Na 1065.32 Pent7Ac2K 1545.52 Pent9Ac3MeUrNa 1091.35 Pent7Ac3Na 1587.53 Pent9Ac4MeUrNa 1107.32 Pent7Ac3K 1593.54 Pent10AcMeUrNa 1113.33 Hex6Ac2K 1635.56 Pent10Ac2MeUrNa 1123.30 Pent5AcMeUr2Na 1677.59 Pent10Ac3MeUrNa 1131.33 Pent4Ac5Ur2Na 1719.58 Pent10Ac4MeUrNa 1133.34 Pent7Ac4Na 1725.57 Pent11AcMeUrNa 1139.35 Hex6Ac3Na 1767.59 Pent11Ac2MeUrNa 1175.39 Hex7Na 1793.62 Pent12Ac4Na 1181.36 Hex6Ac4Na 1809.62 Pent11Ac3MeUrNa 1191.38 Hex7K 1839.67 Pent10Ac3Ur2Na 1197.36 Hex6Ac4K 2001.73 Hex12K 1233.37 Hex7AcK (Pent=pentose; Hex=hexose; Ac=Acetyl group; MeUr, O-methyl-uronic acid; Ur: uronic acid) 24 3.3.3. FTIR Analysis FTIR spectroscopy was used to determine the specific absorption bands of the watersoluble compounds present in the autohydrolysis liquor obtained at 180 ºC. The FTIR spectra (Supplementary Figure S3) of the autohydrolysis liquor of chestnut shells presented typical bands reported for hemicellulosic oligosaccharides (Jiang et al., 2014 and Svärd et al., 2015; Dávila et al., 2016; Rico et al., 2018). The bands at 1021.69 and 895 cm-1 correspond to the C- O, C-C stretching vibration or to the C-OH bending vibration observed in xylan and to the β-(1- 4) glycosidic bonds between sugars respectively. According to Peng et al. (2009), the bands located at 1364.70 cm-1 and 1243.16 cm-1 also could be attributed to the structure of hemicelluloses. The presence of arabinosyl side chains in the sugar structure is confirmed by the band at 1144.51 cm-1, while the band observed at 1719.50 cm-1 indicated the presence of acetyl acids linked to the backbone of oligosaccharides (Dávila et al. 2016). The wide band at 3305.22 cm-1 is ascribed to the O-H stretching of hydroxyl groups and the band observed at 2927.84 cm-1 corresponds to the stretching vibrations of C-H bonds. The band located at 1609.68 cm-1 could be attributed to water bound to the sugar chain, but it could also correspond to syringyl derived linked to oligosaccharides. 3.3.4. Thermal gravimetric analysis (TGA) The thermal stability of the solubilized oligosaccharides under the optimum temperature of autohydrolysis was determined by thermogravimetric analysis, as depicted in Figures 3a and 3b. By combining the information from these two curves, it can be observed that three different degradation processes occurred. The lyophilized liquor showed an initial weight loss of 10.38% below 150 ºC. This event was associated to two degradation processes as can be seen in the derivative thermogravimetric (DTG) curve at 105 and 135 ºC, which could correspond to the evaporation of the absorbed water. The greatest weight loss (43.33%) of the lyophilized liquors took place between 150 and 400 ºC, being two processes of maximum degradation at 239 and 304 ºC. This second stage could be attributed to the decomposition of oligosaccharides, as 25 Demirbas (2000) reported that between 150 and 350 °C, the dehydration and decomposition of the glycosyl units took place. These data show that solubilized oligosaccharides from chestnut shells can withstand high temperatures, suggesting their stability in processes such as cooking, pasteurization, and sterilization in food and drug industries. 0 10 20 30 40 50 60 70 80 90 100 25 125 225 325 425 525 625 725 Weight (%) Temperature (°C) Figure 3. Thermogravimetric analysis (TGA) (a) derivative thermogravimetric (b) curves of autohydrolysis liquors obtained at 180 ºC 25 125 225 325 425 525 625 725 Temperature (°C) b) a) 32 Morana, A., Squillaci, G., Paixão, S.M., Alves, L., La Cara, F., & Moura, P. (2017). Development of an energy biorefinery model for chestnut (Castanea sativa Mill.) shells. Energies, 10, 1504-1069. Moure, A., Conde, E., Falqué, E., Domínguez, H., & Parajó, J. C. (2014). Production of nutraceutics from chestnut burs by hydrolytic treatment. Food Research International, 65, 359- 366. Peng, F., Ren, J. L., Xu, F., Bian, J., Peng, P., & Sun, R. C. (2009). Comparative study of hemicelluloses obtained by graded ethanol precipitation from sugarcane bagasse. Journal of Agricultural and Food Chemistry, 57, 6305-6317. Quitain, A. T., Sato, N., Daimon, H., & Fujie, K. (2003). Qualitative investigation on hydrothermal treatment of Hinoki (Chamaecyparis obtusa) bark for production of useful chemicals. Journal of Agricultural and Food Chemistry, 51, 7926-7929. Rico, X., Gullón, B., Alonso, J. L., Parajó, J. C., & Yáñez, R. (2018). Valorization of peanut shells: Manufacture of bioactive oligosaccharides. Carbohydrate Polymers, 183, 21-28. Roberfroid, M., Gibson, G. R., Hoyles, L., McCartney, A. L., Rastall, R., Rowland, I., et al. (2010). Prebiotic effects: Metabolic and health benefits. British Journal of Nutrition, 104, S1- S63. Rostro, M., Sánchez-González, M., Rivas, S., Moure, A., Domínguez, H., & Parajó, J. C. (2014). Non-isothermal autohydrolysis of nixtamalized maize pericarp: production of nutraceutical extracts. LWT Food Science and Technology, 58, 550-556. Ruiz, E., Gullón, B., Moura, P., Carvalheiro, F., Eibes, G., Cara, C., et al. (2017). Bifidobacterial growth stimulation by oligosaccharides generated from olive tree pruning biomass. Carbohydrate Polymers, 169, 149-156. Sanchez, J.I., Marzorati, M., Grootaert, C., Baran, M., Van Craeyveld, V., Courtin, C.M., et al. (2009). Arabinoxylan-oligosaccharides (AXOS) affect the protein/carbohydrate fermentation balance and microbial population dynamics of the Simulator of Human Intestinal Microbial Ecosystem. Microbial Biotechnology, 2, 101-113. Sequeiros, A., Gatto, D. A., Labidi, J., & Serrano, L. (2014). Different extraction methods to obtain lignin from almond shell. Journal of Biobased Materials and Bioenergy, 8, 370-376. Singleton, V. L., & Rossi, J. A. (1965). Colorimetric of total phenolics with phosphomolybdic – phosphotungstic acid reagents. American Journal of Enology and Viticulture, 16, 144-158. 33 Squillaci, G., Apone, F., Sena, L. M., Carola, A., Tito, A., Bimonte, M., et al. (2017). Chestnut (Castanea sativa Mill.) industrial wastes as a valued bioresource for the production of active ingredients. Process Biochemistry, In press. DOI: 10.1016/j.procbio.2017.09.017. Sugisawa, H. (1966). The thermal degradation of sugars. Journal of Food Science, 31, 381-385. Surek E., & Buyukkileci, A. O. (2017). Production of xylooligosaccharides by autohydrolysis of hazelnut (Corylus avellana L.) shell. Carbohydrate Polymers, 174, 565-571. Svärd, A., Brannvall, E., & Edlund, U. (2015). Rapeseed straw as a renewable source of hemicelluloses: Extraction, characterization and film formation. Carbohydrate Polymers, 133, 179-186. Tunc, M. S., & Van Heiningen, A. R. P. (2011). Characterization and molecular weight distribution of carbohydrates isolated from the autohydrolysis extract of mixed southern hardwoods. Carbohydrate Polymers, 83, 8-13. Vázquez, G., Fontenla, E., Santos, J., Freire M. S., Gonzàlez-Alvarez, J., & Antorrena, G. (2008). Antioxidant activity and phenolic content of chestnut (Castanea sativa) shell and eucalyptus (Eucalyptus globulus) bark extracts. Industrial Crops and Products, 28, 279-285. Vázquez, G., Mosquera, O., Freire M. S., Antorrena, G., & González-Álvarez, J. (2012). Alkaline pre-treatment of waste chestnut shell from a food industry to enhance cadmium, copper, lead and zinc ions removal. Chemical Engineering Journal, 184, 147-155. Vella, F. M., Laratta, B., La Cara, F., & Morana, A. (2017). Recovery of bioactive molecules from chestnut (Castanea sativa Mill.) by-products through extraction by different solvents. Natural Product Research, 18, 1-11. Voragen, A. G. J. (1998). Technological aspects of functional food-related carbohydrates. Trends in Food Science & Technology, 9, 328-335. Wang, Z. W., Zhu, M. Q., Li, M. F., Wang, J. Q., Wei, Q., & Sun, R. C. (2016). Comprehensive evaluation of the liquid fraction during the hydrothermal treatment of rapeseed straw. Biotechnology for Biofuels, 9,142-158. Yang, H. Y., Song, X. L., Yuan, T. Q., Xu, F., & Sun, R. C. (2011). Fractional characterization of hemicellulosic polymers isolated from Caragana korshinskii Kom. Industrial & Engineering Chemistry Research, 50, 6877-6885. 34 Supplementary Data Hydrothermal treatment of chestnut shells (Castanea sativa) to produce oligosaccharides and antioxidant compounds Beatriz Gullón1, Gemma Eibes*1, Izaskun Dávila2, María Teresa Moreira1, Jalel Labidi2, Patricia Gullón2 1Department of Chemical Engineering, Institute of Technology, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain 2Chemical and Environmental Engineering Department, University of Basque Country, 20018 San Sebastián, Spain *E-mail address: [email protected] 1 10 100 1000 10000 100000 1000000 10000000 0 50000 100000 150000 200000 250000 11 13 15 17 19 21 23 25 Log Molecular weight IR response (Intensity) Time (min) 180 ºC Mw 805000 393000 210000 113000 21700 11300 6000 667 180 Figure S1. Molecular weight distribution of the oligosaccharides contained in the autohydrolysis liquor of chestnut shells at 180 °C. 35 Figure S2. MALDI-TOF mass spectra of the oligosaccharides contained in the autohydrolysis liquor of chestnut shells at 180 °C. 36 3305.22 cm-1 2927.84 cm-1 1719.50 cm-1 1609.68 cm-1 1364.70 cm-1 1243.16 cm-1 1144.51 cm-1 1021.69 cm-1 895.41 cm-1 1 2 Figure S3. FTIR spectra of autohydrolysis liquors from chestnut shells at 180 ºC. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 37 0 100000 200000 300000 400000 500000 600000 700000 800000 3 8 13 18 23 28 33 Intensity Retention time (min) Furfural 5-methyl furfural 2-furoi acid Catchol 5-hydroxymethyl furfural Pyrogallol 3,4-dihydroxybenzaldehyde Not identified Protocatechuic acid Not identified Not identified 22 23 Figure S4. GC-MS chromatogram of acetyl soluble extracts from autohydrolysis liquors from 24 chestnut shells at 180 ºC. 25 26 27