Pectin from sunflower by-products obtained by ultrasound: Chemical characterization and in vivo evaluation of properties in inflammatory bowel disease

International Journal of Biological Macromolecules 246 (2023) 125505 Available online 22 June 2023 0141-8130/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Pectin from sunflower by-products obtained by ultrasound: Chemical characterization and in vivo evaluation of properties in inflammatory bowel disease Nerea Mu˜ noz-Almagro a , Jose Alberto Molina-Tijeras b , Antonia Montilla a , * , Teresa Vezza b , María S´ anchez-Milla c , Fabi´ an Rico-Rodríguez d , Mar Villamiel a a Instituto de Investigaci´ on en Ciencias de la Alimentaci´ on (CIAL) (CSIC-UAM) CEI (CSIC+UAM), Nicol´ as Cabrera, 9, Campus de la Universidad Aut´ onoma de Madrid, 28049 Madrid, Spain b Departamento de Farmacología, Centro de Investigaci´ on Biom´ edica (CIBM), Facultad de Farmacia, Universidad de Granada, 18071 Granada, Spain c Departamento de Química Org´ anica y Química Inorg´ anica, Universidad de Alcal´ a, Campus Universitario, E-28871 Alcal´ a de Henares, Spain d Facultad de Ingeniería, Programa de Ingeniería de Alimentos, Universidad de Cartagena, Cartagena de Indias - Colombia ARTICLE INFO Keywords: Sunflower pectin Structural characterization Galacturonic acid Degree of methyl-esterification Low-methoxyl pectin Anti-inflammatory effect ABSTRACT Inflammatory bowel disease (IBD) is a public health challenge and the use of pectin for symptom amelioration is a promising option. In this work, sunflower pectin has been extracted without (CHP) and with assistance of ultrasound (USP) using sodium citrate as a food-grade extracting agent. At optimal conditions (64 ◦C, 23 min) the highest yield was obtained with ultrasound application (15.5 vs. 8.1 %). Both pectins were structurally characterized by 1 H NMR, HPSEC-ELSD, FT-IR and GC-FID. Unlike CHP, USP showed a lower molecular weight, higher galacturonic acid, lower degree of methyl-esterification and, overall, higher viscosity. These characteristics could affect the anti-inflammatory activity of pectins, evaluated using DSS-induced IBD model mice. So, USP promoted the defence (ICAM-1) and repair of the gastrointestinal mucosa (TFF3, ZO-1) more effectively than CHP. These results demonstrate the potential amelioration of acute colitis in IBD mice through USP supplementation. Taking into account the biomarkers analysed, these results demonstrate, for the first time, the positive impact of sunflower pectin extracted by ultrasound under very soft conditions on inflammatory bowel disease that might open up new possibilities in the treatment of this serious pathology. 1. Introduction Inflammatory bowel disease (IBD) (ulcerative colitis and Crohn's disease) has emerged as one of the major chronic disorders in the world [1,2]. Current therapies (biological agents, antibiotics, immunosuppressive and anti-inflammatory drugs) involve unpredictable side effects [1,3,4]. Thus, there is a growing interest in the development of effective treatments that repair the intestinal barrier and mitigate symptoms. In this sense, pectins seem to be a promising polysaccharide as an adjuvant in the treatment of IBD, exhibiting anti-inflammatory effects in vitro and in vivo [5–7]. Pectins are heteropolysaccharides composed mainly of a linear polymer of (1 → 4) linked galacturonic acid (GalA), known as homogalacturonan (HG) which can be partially methyl esterified in C-6. This domain can be interrupted by rhamnose residues that bear side chains (arabinans, galactans and arabinogalactans) to form rhamonogalacturonan-I (RG-I) [8]. Pectins are classified as highmethoxyl pectins (HMP; degree of methyl-esterification (DM) >50 %) or low-methoxyl pectins (LMP, DM <50 %). The former derive almost exclusively from by-products from juice manufacturing (citrus peel and apple pomace), being the most used commercial pectins [9]. The industry produces artificial LMP by acid, alkaline and/or enzymatic deesterification methods that are costly for large-scale commercial production. An alternative to de-esterification processes is the use of cheap unconventional sources of natural LMP such as sunflower heads (SH) that can contain up to 20 % of this polysaccharide [9,10]. Unlike HMP, LMP has a net negative charge from the free carboxyl groups of the GalA residues, which interact with positively charged amino acids, located at * Corresponding author at: Instituto de Investigaci´ on en Ciencias de la Alimentaci´ on (CIAL) (CSIC-UAM), C/ Nicol´ as Cabrera 9, Campus de la Universidad Aut´ onoma de Madrid, E-28049 Madrid, Spain. E-mail address: [email protected] (A. Montilla). Contents lists available at ScienceDirect International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac https://doi.org/10.1016/j.ijbiomac.2023.125505 Received 8 March 2023; Received in revised form 8 June 2023; Accepted 20 June 2023 International Journal of Biological Macromolecules 246 (2023) 125505 2 the C- and N-termini of mucin. In this way, LMP penetrate more easily the mucin layer of the intestinal wall by direct interaction with the intestinal epithelia, stimulating small intestinal mucin secretion [11–13]. Industrially, pectin extraction requires the use of strong acids during long periods at high temperatures, which is time- and energy-consuming and leads to pectin degradation [14]. However, the use of sodium citrate as an extraction agent can favour the quality requirements for foodgrade products, and also contribute to reducing the environmental risks associated with treating chemical effluents [15]. In addition, Ultrasound Assisted Extraction (UAE) has received considerable attention due to its numerous advantages including reductions in energy costs and, mostly, increasing the yield and quality of the recoverable pectin with a shorter extraction time [16]. The mechanism of this technique is based on the periodic production and collapse of numerous cavitation bubbles that cause cell disruption, contributing to the higher penetration of the solvent into the target material [17,18]. Besides, this cavitation effect can cause, among other effects, the breakage of glycosidic linkages, reducing the molecular weight of the pectin (Mw), and consequently, facilitating its metabolism in the colon, resulting in an increase in its bioactivity [19,20]. Ponmurugan et al. [14] studied the recovery of pectin from SH by UAE, focusing on the process variables used in the optimization of sunflower pectin extracted with ultrasound (US), although its chemical composition was scarcely characterized, and its biological properties were not evaluated. In the present study, the anti-inflammatory activity of sunflower pectins obtained with US was evaluated in mice with colitis induced by dextran sulfate sodium (DSS). Moreover, an effort to elucidate the relationship between pectin structure, its rheological properties and its bioactivity was made. Therefore, a structural analysis of obtained pectins, using different analytical techniques, was carried out. In addition, we optimized the extraction parameters of US through the response surface methodology (RSM) in order to reach the maximum pectin yield. In this way, we propose that these pectins could be ready-to-operate through a scaling process for their industrialization, without endangering the environment. 2. Materials and methods 2.1. Materials and chemicals Sunflower (Helianthus annuus) heads (SH) without seeds (Fig. 1 Supplementary) were harvested in Albacete (Spain) and supplied as a generous gift by Syngenta (Madrid, Spain). The SH samples were grinded and sieved (particle size ≤100 μ m). Standard monosaccharides (glucose, mannose, rhamnose, arabinose, galactose, GalA and xylose), Pullulan Standard Set (0.34–805 kDa), β-phenyl-glucoside, hexamethyldisilazane, trifluoroacetic acid, sodium citrate tribasic dihydrate, citric acid monohydrate and RNAlater® were purchased from Sigma Aldrich (St. Louis, MO, USA). Carbon adhesive tabs and aluminium mount were acquired from Aname (Madrid, Spain). Deuterated water was obtained from Symta (Madrid, Spain). Ammonium acetate was acquired from Panreac Applichem (Darmstadt, Germany) and ethanol 96 % was obtained from VWR (Barcelona, Spain). DSS (36–50 kDa) was obtained from MP Biomedicals (Santa Ana, CA, USA). Tri-Reagent® was provided from Thermo Fisher Scientific (Invitrogen, USA). The oligo (dT) primers (Promega, Southampton, UK) and KAPA SYBRsFAST qPCR Master Mix (Kapa Biosystems, Inc., Wilmington, MA, USA) were used to perform the qPCR analyses. 2.2. Pectin extraction Before extraction, 7.5 g of ground SH powder were washed with 300 mL MilliQ water at room temperature in order to remove impurities. Afterward, the mixture was centrifuged at 3700g (Heraeus Multifuge 3SR Plus, Termo Scientific, Massachusetts, USA) for 10 min. The wet solid was mixed with 0.74 % (w/v) sodium citrate giving an original solid/liquid ratio 1:20 w/v; this mixture was adjusted with citric acid at pH 3.3. These conditions were selected based on a previous study [21] whose parameters were optimized by a response surface methodology (RSM) taking into account the yield of extraction. The extraction method was assisted by US using an ultrasonic bath with internal dimensions of 24.0 cm ×14.0 cm ×10.0 cm, a capacity of 3 L and a frequency of 45 kHz (Sonica Sweep System EP 2200, Soltec, Milan, Italy). The US power used was determined using the following equation (Eq. (1)) Power density =mCp dT dt (1) where m is the water mass (kg), Cp is the specific heat of water [J/(kg ◦C s)] and dT/dt is the temperature change along with the whole time range (◦C/s) determined by a polynomial curve fitting. Power density used was 0.17 W/cm 3 . After the extraction reactions with sodium citrate with the assistance of US, samples were centrifuged and the supernatants were treated with ethanol to precipitate pectins. After two cycles of washing with acidulated ethanol, pectins were freeze-dried for further analysis. The pectin yield was calculated according to the formula: Pectin yield (%) = m(g) m0(g)×100 (2) where m is the weight of dried pectin (g) and m 0 is the initial weight of SH (g). 2.3. Statistical experimental design and analysis Optimization of pectin extraction from SH was done applying a Centred Composite Rotatable Design (CCRD) using “Design-Expert®10 trial version” (Stat-Ease Inc., Minneapolis, USA). The influence of temperature (X 1 , 40–60 ◦C) and time (X 2 , 40–120 min) was studied obtaining surface responses for pectin yield. CCRD consist of four axial, 4 factorial and 5 central points, and totalizing 13 trials for each type of extraction (Table 1) at the 0.05 significance level. The experiments were performed in a randomized order to avoid systematic errors. The quadratic model for predicting the optimal point was expressed as follows: Yi =β0+βiXi+βjX2 i+βijXiXj(3) Table 1 Centered composite design with independent variables, temperature (◦C, X 1 ) and time (min, X 2 ) for yield of sunflower pectin. General conditions 0.74 % (w/ v) sodium citrate, pH 3.3, solid/liquid ratio 1:20 w/v; using an ultrasonic bath (45 kHz). Experiments Design points Temperature (◦C) Time (min) Pectin yield (%) 1 Factorial point 40.0 40.0 2.0 2 Factorial point 60.0 40.0 8.7 3 Factorial point 40.0 120.0 8.7 4 Factorial point 60.0 120.0 7.9 5 Axial (star) point 35.9 80.0 1.4 6 Axial (star) point 64.1 80.0 13.1 7 Axial (star) point 50.0 23.4 2.5 8 Axial (star) point 50.0 136.6 2.0 9 Centre (middle) point 50.0 80.0 1.5 10 Centre (middle) point 50.0 80.0 1.6 11 Centre (middle) point 50.0 80.0 1.4 12 Centre (middle) point 50.0 80.0 1.4 13 Centre (middle) point 50.0 80.0 1.3 N. Mu˜ noz-Almagro et al. International Journal of Biological Macromolecules 246 (2023) 125505 3 where Y i is the predicted response for each experience (i =1–13) and each experimental response (X 1–2 ), β 0 is the intercept, β i , are linear coefficients, βii are squared coefficients, and βij are the interaction coefficients for each independent variable. The quality of fit of the secondorder model equation was expressed by the coefficient of determination R 2 . Also, R 2 -Adjusted (percentage of explained variance) and R 2 -Pre- dicted (indicator of how well the regression model predicts responses for new observations) were taken into account to evaluate uniformity of the model. Statistical significance was determined by P-value. The significance of the regression coefficients was tested by t-value. 2.4. Structural characterization of sunflower pectin The estimation of Mw distribution of pectin samples was determined according to the method described by Mu˜ noz-Almagro et al. [22]. Samples (50 μ L) were eluted with 0.01 M NH 4 Ac at a flow rate of 0.5 mL/min for 50 min at 30 ◦C in a HPSEC-ELSD. Pectins extracted were hydrolyzed and derivatized forming their thimethylsilyl oximes before GC-FID analysis following the method of Mu˜ noz-Almagro et al. [21]. Analyses were carried out using a DB-5HT capillary column (30 m ×0.32 mm ×0.10 μ m) (J&W Scientific, Folson, California, USA). Oven temperature program was increasing from 150 ◦C to 165 ◦C at 1 ◦C/min, held for 17 min, then increased at a rate of 10 ◦C/min to 200 ◦C and up to 380 ◦C at a heating rate of 50 ◦C/min. Freeze-dried samples were analysed by FT-IR according to the method described by Mu˜ noz-Almagro et al. [21]. The DM of pectin was determined as the average of the ratio of the peak area at 1747 cm −1 (COO-R) over the sum of the peak areas of 1747 cm −1 (COO-R) and 1632 cm −1 (COO − ). Morphological observations of the sunflower pectins extracted without and with UAE at optimal conditions (CHP, USP), were performed using a DSM 950 scanning electron microscope (SEM) (Zeiss Iberia, Madrid, Spain) at 7 kV accelerating voltage, a magnification of 350×and a distance of 10 mm. For Nuclear Magnetic Resonance (NMR) spectroscopy analysis, 1 H NMR spectra were acquired at 30 ◦C. The samples were solubilized in D 2 O and the chemical shifts were expressed as δ (ppm), using this solvent as internal references at 4.64 ppm. NMR spectra were recorded on a Varian Unity VXR-300. 2.5. Rheological properties of sunflower pectins Rheological measurements were done using an AR 2000 rheometer (TA Instruments Ltd., Crawley, U.K.). Data were analysed with the computer software Rheology Advantage (TA Instruments, Waters Co., Ltd). A cone (40 mm diameter, 1◦) was employed for viscosity (0.1–2 % w/v) using a gap size of 85 μ m at 25 ◦C. For steady flow studies, the shear rate changed from 0.1 to 1200 s −1 . Rheological tests were performed in the range 0.05–10 Hz (0.5 % strain). 2.6. Assessment of anti-inflammatory activity in DSS-treated mice 2.6.1. Experimental design C57BL/6 male mice (7–9 weeks old) were obtained from Janvier (St Berthevin Cedex, France). This study was carried out in accordance with the Guide for the Care and Use of Laboratory Animals as promulgated by the National Institutes of Health. The experimental protocol was approved by the Commission of Ethics in Animal Experimentation (Protocol CEEA 2010-286) of the University of Granada (Spain). They were housed in Makrolon cages, maintained under an air-conditioned atmosphere with a 12 h light–dark cycle, and provided with free access to tap water and a standard rodent diet (Panlab A04 diet, Panlab S. A., Barcelona, Spain). Mice (23 ±2 g) were maintained under specific pathogen-free conditions in the facilities of Animal Experimentation Unit of the University of Granada, and were randomly assigned to four groups (n =10): healthy, DSS control, sunflower pectin extracted by UAE at optimal conditions (USP) and its corresponding control heating pectin (CHP), sunflower pectin extracted at optimal conditions without US. The pectins were diluted with water and administered by oral gavage (100 μ L per day) corresponding to a daily dose of 40 mg/kg per mouse. This dose was chosen taking into account previous studies [6]. Induction of colitis was performed 15 days after the start of the treatment with pectins by adding 3 % (w/v) DSS to the drinking water for seven days, following the time schedule showed in Fig. 2 Supplementary. 2.6.2. Macroscopic indicators Weight variation and Disease Activity Index (DAI) were considered to be macroscopic indicators. DAI was assigned as the sum of body weight, the presence of gross blood in the feces and the stool consistency according to the Zhao et al. [23] (Table 1 Supplementary). 2.6.3. In vivo intestinal permeability The day before sacrifice, mice were fasted for 12 h and administered DX-4000–FITC (Sigma-Aldrich) by oral gavage (350 mg/kg body weight). 4 h after administration, just before sacrifice under isoflurane anaesthesia, blood was collected by cardiac puncture and centrifuged at Fig. 1. Response surface plot (2D, right) and contour plot (3D, left) of sunflower pectin yield (%) as a function of significant interaction between extraction time and extraction temperature. N. Mu˜ noz-Almagro et al. International Journal of Biological Macromolecules 246 (2023) 125505 4 3000 rpm for 10 min at 4 ◦C. Plasma was diluted (1:10) in PBS, and FITC- dextran concentration was determined with a fluorescence spectrophotometer (Fluorostart, BMG Lab Technologies) at an emission wavelength of 535 nm and an excitation wavelength of 485 nm. A standard curve was obtained by serial dilutions of FITC-dextran in PBS. 2.6.4. Biochemical markers The expressions of pro-inflammatory cytokines such as tumor necrosis factor (TNF- α ), interleukins (IL-6, IL-1β), inducible nitric oxide synthase (iNOS), monocyte chemoattractant protein-1 (MCP-1), receptors as toll-like receptor 4 (TLR4) as well as barrier intestinal proteins such as trefoil factor peptide 3 (TFF3), intercellular adhesion molecule (ICAM-1), mucin (MUC-1, MUC-2, MUC-3), occluding, villin and zonula occludens-1 (ZO)-1, were evaluated. The colon tissue was longitudinally divided into different fragments and stored at −80 ◦C in RNAlater®. Total RNA from colonic samples was isolated using Tri-Reagent® following the manufacturer's protocol. All RNA samples were quantified with the Thermo Scientific NanoDrop™ 2000 Spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA, USA) and 2 μ g of RNA was reverse transcribed using oligo (dT) primers (Promega). Real-time quantitative PCR was carried out on optical grade 48-well plates in an Eco™ Real-Time PCR System (Illumina, CA, USA) with 20 ng of cDNA, Table 2 Analysis of variance (ANOVA) of the quadratic model for the yield of sunflower pectin. Source Sum of squares DF Mean square F-value P-value Model 5 163.569 32.7138 8.1 0.008 Significant Linear 2 66.35 33.1752 8.21 0.015 Temperature 1 62.98 62.9795 15.58 0.006 Time 1 3.371 3.3708 0.83 0.391 Square 2 83.156 41.5781 10.29 0.008 Temperature * temperature 1 81.545 81.5448 20.18 0.003 Time * time 1 5.936 5.9361 1.47 0.265 2-Way interaction 1 14.062 14.0625 3.48 0.104 Temperature * time 1 14.062 14.0625 3.48 0.104 Error 7 28.288 4.0411 Lack-of-fit 3 28.236 9.412 724 0.000 Significant Pure error 4 0.052 0.013 Total 12 191.857 R 2 85.26 % R 2 adj 74.72 % Table 3 Molecular parameters and sugar composition of pectins extracted from SH under optimum conditions on yield extraction, DM and Mw (mean ±SD). CHP USP Yield (%) 8.1 ±0.2 a 15.5 ±0.4 b GalA (%) 84.2 ±0.6 a 88.0 ±0.9 b Rha (%) 3.8 ±0.1 a 4.0 ±0.2 a Gal (%) 3.9 ±0.1 b 3.1 ±0.1 a Xyl (%) 2.4 ±0.1 b 0.2 ±0.0 a Ara (%) 2.9 ±0.2 a 3.0 ±0.1 a Man (%) 1.2 ±0.1 b 0.8 ±0.2ª Glc (%) 1.6 ±0.2 b 0.9 ±0.3 a Mw (kDa) 776 ±12 b 500 ±8 a DM (%) 42 ±0.5 b 35 ±0.2 a GalA Rha 22.2 ±0.6 a 22.0 ±0.9 a Gal +Ara Rha 1.8 ±0.1 b 1.6 ±0.1 a Note: values with different small case superscript letters (a–b) in the same column within each solvent and each pectin indicate significant differences as estimated by Tukey's test (P <0.05); Xyl: xylose; Ara: arabinose; Rha: rhamnose; Gal: Galactose; Man: mannose; Glc: glucose; GalA: galacturonic acid. Fig. 2. Chromatographic HPSEC-ELSD profiles for sunflower pectin extracted without (CHP) and with ultrasound (USP) at optimal conditions (64 ◦C, 23 min). N. Mu˜ noz-Almagro et al. International Journal of Biological Macromolecules 246 (2023) 125505 5 KAPA SYBR®FAST qPCR Master Mix (Kapa Biosystems), and specific primers at their annealing temperatures (Table 2 Supplementary). In order to normalize mRNA, the expression of the housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), was measured. The relative quantitation of mRNA was calculated using the ΔΔCt method. 2.7. Statistical analysis Data were expressed as mean values ±SD. A Tukey's test to evaluate the differences was performed using IBM® SPSS® Statistics 25 (IBM Corporation, Armonk, NY, USA). 3. Results and discussion 3.1. Optimization and validation of pectin extraction To investigate the influence of time and temperature on the extraction yield of sunflower pectins, a CCRD was carried out with an US bath (45 kHz). In general, the higher the temperature the better the yields of pectin extraction (Table 1). The lowest yields (<1.5 %) were found at temperatures below 50 ◦C, whereas the highest yield values were obtained at the highest temperature (64.1 ◦C). The optimal conditions obtained to reach the highest extraction recovery of pectin from SH were 23 min and 64 ◦C with a desirability of >0.999. In these conditions, the mean experimental pectin yield (15.5 %) was near the predicted value (14.4 %), demonstrating the validation of the optimized conditions. It is noteworthy that the optimal values found for these experiments were located in the edges of the surface of study for all evaluated variables, which suggests that the optimal region Fig. 3. FT-IR spectra of the sunflower pectin extracted without (CHP) and with ultrasound (USP) at optimal conditions (64 ◦C, 23 min). Fig. 4. 1 H NMR spectra of USP. N. Mu˜ noz-Almagro et al. International Journal of Biological Macromolecules 246 (2023) 125505 6 might be located at high temperatures and short reaction times (Fig. 1), but always taking into account that the maximum value was very close to 60 ◦C. According to the literature, the maximum mechanical effect of US is obtained at temperatures near 60 ◦C, since this appeared to be the maximum temperature for cavitation bubbles to coalesce with vapour bubbles, suggesting that values above this temperature might interfere with the US effect on pectin extraction [24]. Temperature had a significant effect (p <0.05) for linear and interaction in the model parameters, which means it plays an important role in pectin extraction. For all cases, R 2 was above 85 %, which indicates a good fit of data to the model. R 2 -adjusted values were close to R 2 , which is desired for the model. According to the ANOVA results shown in Table 2, temperature had an important effect (p <0.05) on the yield of pectin extraction for linear and quadratic parameters. At the same optimal conditions, sunflower pectin was also obtained but without the use of US (control heating pectin, CHP). Thus, as indicated in Table 3 the yields obtained using sodium citrate at 64 ◦C, 23 min and pH 3.2 without US were 8.1 %, around 47 % lower than the yield obtained under the same conditions but with ultrasonic assistance. The UAE increases the rupture of the sunflower cell wall matrix leading to better interactions between the solvent and extracted material [25]. In addition to that, our results were also above those reported by Ponmurugan et al. [14] who obtained 8.9 % of sunflower pectin by UAE. These discrepancies could be attributed to the different solid-liquid rate used (1:15 vs 1:20). Bayar et al. [25] found that the highest yields are reached by the increase of the liquid-solid ratio, which improves the contact area between the raw material and the extracting agent. In other words, an attenuation of the US intensity occurs with increasing solid particle contents. Fig. 5. Viscosity-shear rate profiles of sunflower pectins at different concentration (T =25 ◦C, 0.1–2 % (w/v)). N. Mu˜ noz-Almagro et al. International Journal of Biological Macromolecules 246 (2023) 125505 7 3.2. Characterization of extracted pectin under optimal selected conditions The compositional and structural characteristics of the pectin recovered with and without the assistance of US are shown in Table 3. As expected, GalA was the most abundant monosaccharide in both samples, being significantly higher in samples extracted by US, in agreement with Dranca and Odorian [26] who stated, in apple pomace, that sonication has a statistically significant influence on the GalA content of the extracted pectin. Regarding the content of GalA and neutral sugars (Table 3), it is evident that the structure of pectin consists mainly of HG, as well as a small part of RG-I (arabinan and arabinogalactan) chains [8]. As a consequence of sonication, hardly any change was observed in the degree of branching (GalA/Rha) and complexity of branch distribution (Ara +Gal/Rha), with only a slight decrease in the latter. Another noteworthy aspect is that the content of the monosaccharides (glucose and mannose) derived from non-pectic polysaccharides such as cellulose and hemicellulose, were considerably lower in the pectin obtained after sonication. Ponmurugan et al. [14] showed similar trends in sunflower pectin recovered by UAE under different extraction conditions. A plausible explanation for this fact could be that the US treatment cleaved the covalent linkages of nonpectic and pectic polysaccharides [14]. This could corroborate that US treatment improves the purity of the obtained pectin in comparison with the traditional method. Moreover, a study on the effect of US on the Mw of sunflower pectins was carried out by HPSEC-ELSD. Fig. 2 depicts the chromatographic profiles of pectin extracted with and without US in the presence of sodium citrate. The pectin distribution in the CHP sample showed the earliest elution and, in the case of USP, a displacement of the peaks to be right could indicate that the pectin extracted with US treatment was modified to species of lower Mw (776 vs 500 kDa, Table 3), in agreement with Soria, Villamiel and Montilla [27] who revised the main structural effect of US on polysaccharides, including pectins. FT-IR, 1 H NMR and SEM (Figs. 3–4 and Fig. 3 Supplementary) were applied to elucidate if the different extraction methods used to obtain pectin could affect the functional groups and their bonding configurations. The FT-IR spectra of both sunflower pectins are presented in Fig. 3. The wide and marked absorption band at 3429 cm −1 corresponds to stretching vibration of hydroxyl groups due to inter- and intramolecular hydrogen linkages located in the main chain of GalA [25]. The band that appears on 2928 cm −1 assigned to the vibrations of C – H ( – CH, – CH 2 , – CH 3 ), exhibited the same area in all pectin samples, indicating that the main linear domain (HG) of the USP did not suffer any breakage of glycosidic bond during the sonication. These results can also be corroborated with the peaks 1405, 1239, 1145 and 1052 cm −1 corresponding to the stretching vibrations of C – OH side groups and the C – O – C glycosidic bond vibration, which could be due to the minor presence of glycosidic linkages between sugar units [28]. From the analysis of the DM (Table 3), it is observed that slight but significant changes were registered between the pectin obtained after US treatment and its corresponding control (35 % vs 42 %). These data might indicate that the de-esterification of pectin chains that occurs under US extraction was not along with variations in the main structural HG chain of pectin [29]. The 1 H NMR spectrum corresponding to USP (Fig. 4) showed different doublet around 1.20 ppm corresponding to methyl protons of H-6 rhamnoses. The following peaks appearing at 2.01 and 2.12 ppm corresponded to methyl protons from acetyl group binding [30,31]. Most of the pectin signals appear between 3.0 and 5.5 ppm. 3,4-β-Xylp gave a H2 signal at 3.31 ppm [32]. The signal at 3.45 ppm was assigned to H2 of 1,3,6-β-Galp, H6 of 1,3-β-Galp and H4 of 1,2- α -Rhap. The signals overlapping between 3.60 and 3.80 ppm were attributed to 1,4- β-Galp: H2, H3, H5 and H6; 1,6-β-Galp: H2 and H3 and 1,2- α -Rhap: H3 and H5 [33]. The signal for H2 of 1,4- α -GalpA was observed at 3.83 ppm in the spectrum. The signal at 3.88 ppm was attributed to H5 and H6 of 1,6-β-Galp and H5 of 1,5- α -Araf. The intensity signal at 4.23 was assigned to H4 of 1,4- α -GalpA. At around 5 ppm appear the signals of H1 of 1,4- α -GalpA, 1,5- α -Araf and 1,2- α -Rhap [33]. These results were in agreement with the aforementioned characterization of pectin. The microstructure of CHP and USP analysed by SEM is illustrated in Fig. 2 Supplementary (a–b). As is depicted, both pectins consisted of larger, irregular, smooth and flake-shaped surfaces. A slight and increasing compaction of the structure seems to be observed in USP, probably due to the higher content of GalA that favours the formation of a network of pectin [34]. 3.3. Steady-shear measurements of sunflower pectins As shown in Fig. 5, the apparent viscosity of the studied sunflower pectins depended on the shear rate at different concentrations. CHP and USP solutions showed nearly Newtonian flow behaviour at concentrations inferior to 0.2 and 1 % w/v, respectively. On the contrary, a pseudoplastic flow became dominant when the concentration increased. The viscosity of pectin decreased rapidly with increasing shear rate Fig. 6. Effects of CHP and USP on the macroscopic indicators. (a) weight gain; (b) Disease Activity Index (DAI); (c) intestinal permeability. CHP: control heating pectin. USP: pectin extracted by ultrasound. N. Mu˜ noz-Almagro et al. International Journal of Biological Macromolecules 246 (2023) 125505 8 (from 1 to 100 s −1 ) but decreased less rapidly at higher shear rates (100–1000 s −1 ) at higher concentrations. Unlike the USP sample, low viscosities were observed in CHP at concentrations lower than 0.2 % (w/ v). In general terms, when pectin concentration was above 0.5 % (w/w) in both samples, the initial viscosity was significantly increased. This change indicated that the intermolecular distance had been shortened due to an increase in solid concentration facilitating the intermolecular interactions, especially hydrogen bonding with hydroxyl groups [35]. In all cases, USP was more viscous than CHP, which can be interpreted by the higher Mw of CHP in comparison with USP (776 vs 500 kDa). According to Hua et al. [35], the high Mw could be related to a small number of molecules which were too far apart to interact with another one. The fact that USP presented higher viscosity than CHP could be also ascribed to the significantly higher content of GalA and lower DM, indicated above. Hua et al. [35] observed that high Mw and methoxyl content means a small number of molecules, a greater distance between them and strong electrostatic repulsions along the chain, resulting in lower viscosity of pectin extracted from conventional heating. Following a Pearson correlation, Dranca and Oroian [26] found a negative correlation between the viscosity and 1634 cm −1 wavenumber transmittance 0 1 2 3 4 5 6 7 8 Healthy DSS Control CHP USP TNF-α )esaercnidooF( a c c b 0 0.5 1 1.5 2 2.5 Healthy DSS Control CHP USP ICAM (Fold increase) a b b a 0 5 10 15 20 25 Healthy DSS Control CHP USP IL-1β )esaercnIdloF( a c b b 0 2 4 6 8 10 12 Healthy DSS Control CHP USP MCP1 (Fold Increase) a cc b 0 10 20 30 40 50 60 Healthy DSS Control CHP USP IL- )esaercni dloF(6 a c bab 0 0.5 1 1.5 2 2.5 3 Healthy DSS Control CHP USP TFF3 (Fold increase) a cc b 0 1 2 3 4 5 6 7 8 Healthy DSS Control CHP USP )esaercnidloF (SONI a c b b 0 0.5 1 1.5 2 2.5 Healthy DSS Control CHP USP TLR4 (Fold increase) b aa c D C B A E F G H Fig. 7. Effects of CHP and USP on the expression of pro-inflammatory cytokines and chemokines in colonic tissue of the DSS colitis mice. Different letters on the bars indicate statistical differences (p <0.05) among groups. N. Mu˜ noz-Almagro et al. International Journal of Biological Macromolecules 246 (2023) 125505 9 values (r = − 0.9991), which was responsible for (C – – O) stretching vibration of the carboxylate ion. Similarly, Polanco-Lugo et al. [36] observed in citrus pectin high methoxyl group content and lower viscosity. A highly positive correlation between viscosity and GalA content (r =0.992) was also found, since a higher GalA content directly influences the formation of the pectin network [36]. Taking altogether the results indicated above, this might explain why the solutions obtained with the pectin extracted by US were more viscous. 3.4. Assessment of anti-inflammatory activity in DSS-treated mice Based on the structural features mentioned above, USP exhibited a lower Mw which could improve its solubility, facilitating its metabolism in the colon. Moreover, previous studies demonstrated that the presence of neutral sugars ( α -arabinoses and β-galactoses) in pectins is able to decrease pro-inflammatory cytokine expression in mice [6,37]. Given these promising results and the interesting chemical and rheological characteristics of USP, a study on its impact in inflammatory bowel disease was done. First, different macroscopic indicators and symptoms of colitis induced in mice by the oral administration of DSS were investigated. Unlike the healthy control group, both samples showed a significant weight loss that was similar to the DSS control group (Fig. 6a). Pacheco et al. [5] and Sabater et al. [6] also found similar behaviour using citrus by-products and artichoke pectins, respectively, and attributed this fact to the anorexia observed in mice from the first day after the beginning of DSS administration. However, mice which have ingested CHP and USP did not reach the DAI values of the DSS control, which evidenced the protective effect of both pectins against DSS-induced colonic damage (Fig. 6b). As observed in Fig. 6c, during the DSS period intestinal permeability increased significantly in the DSS control group, which could contribute to the entry of commensal bacteria and decrease transepithelial resistance [38]. On the contrary, the concentration of this macroscopic indicator decreased significantly in animals treated with pectins, with the values corresponding to USP being significantly lower than CPH and close to the healthy group. As expected, DSS-induced colitic inflammation was characterized by an altered immune response, which was evidenced by an increasing expression of TNF- α , IL-1β and IL-6 in comparison with healthy mice Fig. 7. (continued). N. Mu˜ noz-Almagro et al.