Spent coffee grounds, an innovative source of colonic fermentable compounds, inhibit inflammatory mediators in vitro

Accepted Manuscript Spent coffee grounds, an innovative source of colonic fermentable compounds, inhibit inflammatory mediators in vitro Dunia Maria López-Barrera, Kenia Vázquez-Sánchez, Ma. Guadalupe Flavia Loarca-Piña, Rocio Campos-Vega PII: DOI: Reference:

S0308-8146(16)30866-4 http://dx.doi.org/10.1016/j.foodchem.2016.05.175 FOCH 19324

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

27 January 2016 11 May 2016 28 May 2016

Please cite this article as: López-Barrera, D.M., Vázquez-Sánchez, K., Loarca-Piña, a.G.F., Campos-Vega, R., Spent coffee grounds, an innovative source of colonic fermentable compounds, inhibit inflammatory mediators in vitro, Food Chemistry (2016), doi: http://dx.doi.org/10.1016/j.foodchem.2016.05.175

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Spent coffee grounds, an innovative source of colonic fermentable compounds, inhibit inflammatory mediators in vitro

Running title: Spent coffee grounds inhibit inflammatory mediators in vitro

Dunia Maria López-Barrera, Kenia Vázquez-Sánchez, Ma. Guadalupe Flavia LoarcaPiña, Rocio Campos-Vega*

Programa de Posgrado en Alimentos del Centro de la Republica (PROPAC), Research and Graduate Studies in Food Science, School of Chemistry, Universidad Autónoma de Querétaro, 76010 Santiago de Querétaro, Qro, México

* Corresponding author. E-mail address: [email protected] (R. Campos-Vega).

1

ABSTRACT Spent coffee grounds (SCG), rich in dietary fiber can be fermented by colon microbiota producing short-chain fatty acids (SCFAs) with the ability to prevent inflammation. We investigated SCG anti-inflammatory effects by evaluating its composition, phenolic compounds, and fermentability by the human gut flora, SCFAs production, nitric oxide and cytokine expression of the human gut fermented-unabsorbed-SCG (hgf-NDSCG) fraction in LPS-stimulated RAW 264.7 macrophages. SCG had higher total fiber content compared with coffee beans. Roasting level/intensity reduced total phenolic contents of SCG that influenced its colonic fermentation. Medium roasted hgf-NDSCG produced elevated SCFAs (61:22:17, acetate, propionate and butyrate) after prolonged (24 h) fermentation, suppressed NO production (55%) in macrophages primarily by modulating IL-10, CCL-17, CXCL9, IL-1β, and IL-5 cytokines. SCG exerts anti-inflammatory activity, mediated by SCFAs production from its dietary fiber, by reducing the release of inflammatory mediators, providing the basis for SCG use in the control/regulation of inflammatory disorders. The results support the use of SGC in the food industry as dietary fiber source with health benefits. Chemical compounds studied in this article Acetic acid (PubChem CID: 176); propionic acid (PubChem CID: 1032); butyric acid (PubChem CID: 264); chlorogenic acid (PubChem CID: 1794427); nitric oxide (PubChem CID: 145068); gallic acid (PubChem CID: 370); catechin (PubChem CID: 9064); caffeine (PubChem CID: 2519); caffeic acid (PubChem CID: 689043); rutin (PubChem CID: 5280805)

2

Keywords: Spent coffee grounds; in vitro gastrointestinal digestion, in vitro colonic fermentation; inflammation; nitric oxide; cytokines.

Abbreviations: ABTS, 2,2′-azinobis (3-ethylbenzothiazoline-6-sulfonic acid); AC, Antioxidant capacity; CB, Coffee beans; CBDR, Coffee bean dark roasted; CBMR, Coffee bean medium roasted; DF, Dietary fiber; DPPH, 2,2-diphenyl-1-picrylhydrazyl; DR, Dark roasted; FBS, Fetal bovine serum; GI, Gastrointestinal; GO, Gene ontology; hgf-NDSCG, Human gut fermented-unabsorbed-SCG; HPLC-DAD, High-performance liquid chromatography-diode array detection;ICO, International Coffee Organization; IDF, Insoluble dietary fractions; IL, Interleukin; KEGG, Kyoto encyclopedia of genes and genomes; LPS, Lipopolysaccharides; MR, Medium roasted; MTT, Methylthiazolyltetrazolium;NDSCG, Unabsorbed spent coffee grounds; NO, Nitric oxide; SCDF, Spent coffee-dietary fiber; SCFAs, Short-chain fatty acids; SCG, Spent coffee grounds;SCGDR, Spent coffee grounds dark roasted; SCGMR, Spent coffee grounds medium roasted; SDF, Soluble dietary fractions; TDF, Total dietary fiber; TNF-α, Tumor necrosis factor-α.

3

1. Introduction In 2012-13, 146 million bags (60 kg each) of coffee provided an estimated 4 billion of coffee cups consumed daily in the world (ICO, 2014), generating large quantity of byproducts during processing. Spent coffee grounds (SCG), the residue obtained after brewing, are rich in polysaccharides, proteins, phenolic compounds, melanoidins and dietary fiber (DF) (Campos-Vega, Loarca-Piña, Vergara-Castañeda & Oomah, 2015a). Spent coffee-dietary fiber (SCDF) with 84% insoluble and 16% soluble dietary fiber relative to total dietary fiber, exhibit antioxidant properties and can be categorized as antioxidant dietary fiber (Campos-Vega, Loarca-Piña, Vergara-Castañeda & Oomah, 2015a; Campos-Vega et al., 2015b). These dietary fibers strongly associated/bound to tannins are metabolized primarily during colonic fermentation and subsequently absorbed and distributed to tissues with potential health benefits (Campos-Vega et al., 2015b). Fermentation of dietary fiber from SCDF with human gut flora can potentially release short chain fatty acids (SCFAs), similar to those of nondigestible bean fraction that modulates genes and proteins involved in the anti-inflammatory process/pathway (Campos-Vega et al., 2012). SCFAs affect different cells involved in the inflammatory and immune responses. They modulate the function of leukocytes (e.g., production of inflammatory mediators and ability of leukocytes to migrate) and induce apoptosis in lymphocytes, macrophages and neutrophils (Vinolo, Rodrigues, Nachbar & Curi, 2011). The physiological potential and health benefits of SCG have not been adequately investigated although it is a rich antioxidant and DF (~ 60%) source, and has been proposed as a healthy bakery ingredient for the general population and for people with special nutritional requirements (Galanakis et al., 2015; Campos-Vega et al., 2015b). This study aimed at evaluating the nutraceutical and anti-inflammatory activity of bioactive

4

compounds from unabsorbed spent coffee grounds (NDSCG) fraction. In vitro gastrointestinal assay and colonic fermentation were used to determine the fate of SCFAs and the anti-inflammatory activity was evaluated on lipopolysaccharide (LPS)-induced mouse RAW 264.7 macrophages.

2. Materials and methods

2.1. Materials Roasted coffee arabica beans (medium- and dark-roasted), purchased directly from the manufacturer were grown and harvested in the state of Chiapas, México. Butyrate was purchased from Fluka (Sigma-Aldrich Canada Ltd., Oakville, ON, Canada). Vanillin, D(+)-raffinose, and (+)-catechin were obtained from Sigma Chemical Co. (St. Louis, MO, USA). Acetate, propionate and formic acid, and other chemicals were purchased from Sigma Chemical Co and J. T. Baker (México City, México).

2.2. Spent coffee grounds preparation Arabica spent coffee grounds were recovered from the filter of a coffeemaker (MOULINEX, Heliora comfort, México) after 6 min brewing (7:100, w/v) at 90 °C. The SCG were freeze-dried, defatted (Soxhlet extraction, petroleum ether; 6 h) and then stored in the dark until use.

2.3. Proximate composition AOAC procedures were used to determine moisture (method 925.10), lipid (method

5

920.39), ash (method 923.03), and nitrogen (method 920.87) contents of the ground bean samples (coffee beans and SCG) (AOAC, 2002). Moisture was assessed based on weight loss after oven drying at 105 °C until constant weight was reached. Nitrogen content was determined using the micro-Kjeldahl method with sodium sulfate as catalyst. Protein content was calculated as nitrogen x 6.25. Lipid content was obtained from Soxhlet extraction (6 h) with petroleum ether. Ash content was calculated from the weight of the sample after incineration in a muffle furnace at 550 °C for 2 h. Carbohydrate values were obtained by difference.

2.4. Total dietary fiber (TDF) and resistant starch Dietary fiber fractions, containing soluble dietary fractions (SDF), and insoluble dietary fractions (IDF) were determined following the enzymatic-gravimetric method of Shiga et al., (2003). Resistant starch was quantified following the gravimetric method of SauraCalixto et al., (1993) described briefly in our earlier study (Campos-Vega et al., 2009).

2.5. Polyphenols extraction Microwave extraction was performed by the modified method described previously (Campos-Vega et al., 2015b). Spent coffee grounds and 20% ethanol solution (1:9 solid/liquid ratio) were heated in 100 mL Erlenmeyer flasks at 80 W for 20 s in a microwave oven (LG MS-1145KYL); cooled, then re-extracted under the same conditions. Extracts were centrifuged (5000 rpm, 10 min, 4 °C; Hermle Z 323 K, Hermle Labortechnik GmbH), and the recovered supernatant was filtered (45 µm) for HPLC.

2.6. Analysis of polyphenols compounds and caffeine by HPLC-DAD

6

Polyphenols and caffeine were analyzed by HPLC following the method described by Ramírez-Jiménez et al., (2014) with an Agilent 1100 Series HPLC system (Agilent Technologies, Palo Alto, CA, USA) equipped with a binary pump and an automated sample injector. An Agilent reversed-phase Zorbax Eclipse XDB C-18 (5 µm particle size, 250 x 4.6 mm) column was used at 35 °C. The chromatographic separation was performed using a gradient of Milli-Q water acidulated with acetic acid (pH 3.0) (solvent A) and acetonitrile, (solvent B) at 1 mL/min constant flow rate. Detection was accomplished with a diode array detector (DAD) and chromatograms were recorded at 280 nm. Quantification was performed using the external standard method with commercial

standards

of

(+)-catechin,

quercetin,

rutin,

caffeine,

and

caffeic,

chlorogenic, p-coumaric, ellagic, ferulic and gallic acids.

2.7. In vitro gastrointestinal digestion The adapted method of Campos-Vega et al., (2015b) was followed to mimic physiologic conditions. Briefly, four healthy volunteers, who had consumed their last meal at least 90 min prior to the test, were recruited. All participants provided written informed consent prior to participating in the study. The subjects chewed the test products under standardized conditions after brushing their teeth without toothpaste. Spent coffee grounds (1 g) were chewed 15 times for ~15 s. After chewing, the product was expectorated into a beaker containing 5 mL of distilled water. The subjects rinsed their mouths with another 5 mL of distilled water for 60 s. Subsequently, the suspensions of each sample were mixed in a single vessel and an aliquot (10 mL) was adjusted to pH 2 using HCl solution (150 mM, 2·81 ml). Pepsin (0.055 g, Sigma) dissolved in 0.94 ml of 20 mM HCl was added to each sample and incubated for 2 h at 37 ° C. An intestinal extract was prepared 30 min

7

before use by dissolving 3 mg of gall Ox (bile salts) and 2.6 mg pancreatin (enzymatic components including trypsin, amylase and lipase, ribonuclease, and protease) in 5 ml Krebs-Ringer buffer [118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, 11 mM glucose and 2.5 mM CaCl2, pH 6.8]. This solution (5 ml) was added to each sample; the suspension (15 ml) was transferred to a vessel containing an everted gut sac, prepared from male Wistar rats (body wt. 250-300 g, n = 6). Prior to the surgical procedure, the rats were fasted overnight (16–20 h) with water ad libitum. The rats were anesthetized with pentobarbital sodium (60 mg/kg, i.p.). The intestine of the rats was exposed by a midline abdominal incision and a 20-25 cm segment of the proximal rat jejunum was excised and placed in the gasified (CO2) buffer solution Krebs-Ringer at 37 ° C. The intestine was washed with the same buffer and gently everted over a glass rod, divided into segments of approximately 6 cm length, filled on the serosal side with 1 mL of Krebs-Ringer buffer and tied to achieve a final length of approximately 4 cm (this procedure was performed with the intestinal segments submerged in Krebs-Ringer buffer to prevent the loss of tissue viability). The everted gut sac containing 15 ml of the suspension was incubated in an oscillating water bath (80 cycles per min, 37 ° C, 2 h) under anaerobic atmosphere (CO2). After the incubation period the sacs were removed; the sample from the mucosal side (outside) referred as unabsorbed SCG (NDSCG) fraction was subjected to in vitro colonic fermentation. Experiments were performed in triplicate; a blank was prepared using only distilled water instead of sample following the procedure described above.

2.8. In vitro colonic fermentation and SCFAs analysis

8

The human gut flora fermentation method (Campos-Vega et al., 2009) was used to estimate the effects of NDSCG in the colon, representing simulated large bowel model. It provides useful data to form hypotheses for in vitro studies. This study approved by Universidad Autonoma de Queretaro Human Research Internal Committee complied with the National Institutes of Health Guide for Care and Use of Laboratory Animals. All participants signed informed consent. Briefly, fermentations were performed in duplicate in a water bath at 37 ◦C. Raffinose was used as a control for fermentable sugar under the same conditions. Fecal inoculum was prepared from stool supplied by two healthy subjects, who had not consumed antibiotics for at least 3 months and had no history of gastrointestinal diseases. Sterile tubes (15 mL capacity) were filled with 9 mL of sterile basal culture medium containing (g/L): peptone water, 2; yeast extract, 2; NaCl, 0.1; K2HPO4, 0.04; KH2PO4, 0.04; MgSO4.7H2O, 0.01; CaCl2.2H2O, 0.01; NaHCO3, 2; cysteine HCl, 0.5; bile salts, 0.5; tween 80, 2 mL; and hematin, 0.2g (diluted in 5 mL of NaOH). Sealed tubes were maintained under a headspace containing H2 -CO2 -N2 (10 : 10 : 80, by volume), O2 -free for 24 h. Fecal slurries were prepared by homogenizing 2 g of fresh stools with 18 mL of 0.1 M-sodium phosphate buffer, pH 7.0. The tubes containing basal culture medium were inoculated with 1 mL of fecal slurries and the sample (500 mg and 500 µl) was added after inoculation, except for blanks. The samples were shaken in vortex for 30 s and placed in a water bath at 37 ◦C. During fermentation, the pH of the sample and SCFAs production were assessed at 6, 12, and 24 h. Fermentation was stopped by placing the tubes in a freezer at −70 ◦C. The quantification of SCFAs was performed according to Campos-Vega et al., (2009). Two independent experiments were performed in triplicate.

9

2.9. Macrophage cell culture and cell viability assay The murine macrophage cell line RAW 267.4 was obtained from the American Type Culture Collection (Manassa, VA, USA) and maintained in DMEM medium containing 10% fetal bovine serum (FBS) and 1% antibiotics of penicillin/streptomycin at 37 °C under 5% CO2. Cytotoxicity was analyzed using methylthiazolyltetrazolium (MTT) assay. RAW 264.7 cells were seeded at 5 × 104 cells/mL densities on 96-well plates (Nunc, Roskilde, Denmark), and allowed to grow in DMEM medium (24 h). Subsequently, the culture medium was discarded, replenished with a mixture of LPS (1 µg/mL) and various concentrations of human gut flora fermented NDSCG (hgf-NDSCG), and then incubated for another 24 h. Filtered MTT solution in serum-free DMEM was added to each well (0.5 mg MTT/mL) prior to cell incubation (37 °C, 2 h) and unreacted dye was removed. The insoluble MTT formazan crystals were allowed to dissolve in DMSO at room temperature for 15 min, and the absorbance (570 nm) of each sample was measured. The viability of the cells was calculated using the following equation: Viability (%) = (absorbance sample)/(absorbance control) X 100

2.10. Nitric oxide (NO) measurement Briefly, cells were incubated with medium alone, or with LPS and the hgf-NDSCG for 24 h. To each well was added 100 µL of Griess reagent [1% sulfanilamide in 5% phosphoric acid and 0.1% N-(1-naphthyl) ethylenediamine dihy-drochloride in water] and then incubated for 15 min in the dark. The total amount of nitrite present was calculated based on sample absorbance at 570 nm. The cells were incubated with LPS (1 µg/mL) alone as

10

the positive control and PBS as the blank control. The amount of NO synthesized in response to LPS stimulation was calculated using the following equation: NO synthesis (%) = [(absorbance of hgf-NDSCG-treated sample - absorbance of DMEM-treated sample)/absorbance of LPS-treated sample - absorbance of DMEM-treated sample)] X 100 %.

2.11. Proteins/cytokines array RAW264.7 cells were seeded at a density of 1 × 106 cells in a 75 cm2 canted neck flask and treated either with 100 ng/ml LPS alone or hgf-NDSCG (LC50) for 24 h. Conditioned media was assayed using Proteome Profiler Array Mouse Cytokine array panel A (ARY006, R and D Systems Inc., Minneapolis, MN) following the manufacturer’s protocol. Protein expression was detected using 1:1 provided chemiluminescent reagents A and B and visualized using a GelLogic 4000 Pro Imaging System (Carestream Health, Inc., Rochester, NY). Data were expressed as fold change relative to control (LPS-stimulated RAW 264.7 cells, without treatment). Additionally, the protein interaction network from the STRING database was used (Von Mering et al., 2007). Gene ontology (GO) was used to categorize biological process, cellular component and molecular function using Kyoto encyclopedia of genes and genomes (KEGG) database pathways.

2.12. Statistical analysis Results were expressed as mean ± standard error of at least two independent experiments and analyzed by ANOVA. Statistical significance was determined using Tukey’s t test (α = 0.05) with JMP version 7.0 software.

11

3. Results and discussion

3.1. Chemical composition Carbohydrates, the most abundant coffee bean (CB) constituent increased (6.8%) on roasting from medium to dark (Table 1) due to polysaccharide depolymerization (Campos-Vega et al., 2015a) and/or increase in galactomannans, the main polysaccharides in roasted coffee with increasing degree of roast (Simões, Nunes, Domingues & Coimbra, 2013). Also, it has been suggested that, during roasting, oligomers and especially monomers are converted very rapidly into Maillard and pyrolysis products (Oosterveld et al., 2003). Furthermore, molecular weight of the brown material linked to galactomannans is higher in the dark-roasted than in the light-roasted coffee (Nunes et al., 2006). Similarly, carbohydrate content of SCG was higher (13.4 %) from dark roast (DR) than from medium roast (MR) reflecting the higher amount of insoluble polysaccharides (9.4% vs 3.1% for DR and MR, respectively) bound to the SCG matrix, and the possible supports of low molecular weight brown compounds by the high molecular weight galactomannans (Nunes et al., 2006). Carbohydrates accounted for 60% of the dry weight of SCG from medium roasted coffee similar to those reported by Simões et al. (2013). However, additional studies are needed since melanoidins may be the predominant carbohydrates,

while

glucose, and consequently starch and resistant starch, could only be minor components (Nunes et al., 2007). Total dietary fiber (mostly insoluble fiber ≥ 95%) comprised 83, 95, 77 and 86 % of the total carbohydrates of MR, SCGMR, DR, and SCGDR, respectively. Despit e the fact that DF was unaffected by roasting intensity, (galacto)mannans degraded only moderately during roasting, and those remaining in the bean showed no evidence of molecular weight changes even after a dark roast. In contrast, arabinogalactans

12

are depolymerized after a light roast both by galactan backbone fission and arabinose loss from the side chains. Furthermore, roasting minimally affects the covalent link between the coffee bean arabinogalactans and protein (Redgwell, Trovato, Curti & Fischer, 2002). Resistant starch, often considered as dietary fiber increased, (without attaining statistical significance) with roasting intensity (18.8%) from MR to DR, thereby inducing similar increase (10.7%) in their corresponding SCG. The thermal treatment (roasting) potentially induced interactions resulting in modified starch formation resisting enzymatic action and increasing TDF and their fractions (Lintas & Cappeloni, 1988). The present data constitute the first report of resistant starch in roast coffee and SCG. Lipid content remained stable during roasting and SCG preparation and the SCG values were within previously reported range (Campos-Vega et al., 2015a). Protein content was generally higher than those reported in the literature (Campos-Vega et al., 2015a), except for SCGDR and decreased with roasting and SCG preparation, probably due to the wellknown involvement of proteins in melanoidin formation during coffee roasting. Ash content also decreased with roasting and the SCG values were similar to those reported recently by Jiménez-Zamora et al., (2015). Therefore, SCG retained high amounts of protein (15.8 and 11.5 %), and to a lesser extent, minerals (1.8 and 1.1 %), with high proportions of lipids (15.1 and 15.3 %) from the medium than the dark roasted beans.

3.2. Characterization of polyphenols by HPLC-DAD Chlorogenic and gallic acids were the most abundant phenolics in roasted coffee and SCG (Table 2), accounting for ≥80% of the total phenolics, except for SCDR (68%), where quercetin contributed 22% of SCDR total phenolics. Catechin and quercetin accounted for less than 10% of the total phenolics except for MR and SCDR, respectively.

13

Caffeic, p-coumaric, ellagic, ferulic acids and rutin were detected in very small amounts (< 3% of the total phenolics). Ferulic acid and rutin were also present in coffee pulp (Ramírez-Martínez, 1988). Increased coffee roasting intensity (from medium to dark) reduced chlorogenic (6%), gallic (15%), p-coumaric (33%), ferulic (14%) acids, quercetin (14%) and catechin (60%); and increased ellagic acid (50%) without affecting caffeic acid and rutin contents. The chlorogenic acid reduction with increased roasting due to isomerization and/or loss corresponds to those reported for Colombian coffee (Cho et al., 2014). SCG from medium roast retained the highest amount of polyphenols (67% chlorogenic, 69% gallic, 57% ferulic, 50% ellagic, 33% p-coumaric, 23% caffeic acids, 40% catechin, and 30% rutin). Furthermore, ferulic acid and quercetin in SCGDR increased considerably (67%) relative to dark roast coffee; a smaller increase (37%) was also observed in quercetin of SCGMR relative to medium roast coffee. This increase

in

quercetin

and

ferulic

acid

concentrations

probably reflects their

release/cleavage from bound to free form in agreement with recent report on SCG (Monente, Ludwig, Irigoyen, De Peña & Cid, 2015). The the total polyphenol content reduction observed in this work, could explain our previous results (Campos-Vega et al., 2015b), where ABTS radical scavenging capacity decreased with increased roasting duration, whereas the higher DPPH antioxidant activity of SCGDR may be supported by the unit contribution of the phenolic compounds to the antioxidant activity and/or other compounds, like melanoidins, that also exhibit considerable antiradical activity. Further studies are needed to identify the main contributors of the reported antioxidant activity.

3.3. Effect of NDSCG on pH and short-chain fatty acids production during in vitro

14

fermentation Colonic fermentation started at neutral pH (7.4) inducing SCFAs dissociation resulting in significant (p < 0.05) pH reduction of all samples after 6 h (Table 3). Fermentation time had no significant effect on pH, independent of changes in total SCFAs production. The net total SCFAs generated by NDSC after 6, 12, and 24 h of fermentation was significantly lower compared to those from raffinose, an observation similar to those reported for coffee fractions (Reichardt et al., 2009) where glucose was used for comparison. It suggests that NDSC substrates are more difficult to break down due to their complex polysaccharide structures. Fermentation of NDSC from medium roasted coffee displayed higher total SCFAs production after 12 and 24 h compared to those from dark roasted coffee, probably reflecting relative changes in monosaccharides with increase in degree of roast. However, fermentability index (%)[(SCFAs(sample)/SCFAs(raffinose)) × 100)] was highest and lowest after 6 and 24 h, respectively with SCGDR displaying higher values than SCGMR (74.8 and 66.6% after 6 h; 44.4 and 34.4% after 24 h). This reflects the increase in condensed tannins and flavonoids released during colonic fermentation in our previous study (Campos-Vega et al., 2015b). The comparatively higher fermentability index of SCGDR may be associated with its distinct competitive microbiota affinity induced by subtle changes in polysaccharide structure/composition. The fermentability index of SC after 24 h was half the reported polysaccharide degradation (75 –90%) of coffee fractions determined by phenol-sulfuric acid method (Reichardt et al., 2009) with no distinct difference in roasting intensity. Acetic acid was the predominant SCFAs produced during SCG fermentation, followed by propionic and butyric acids. SCGDR produced the highest level of SCFAs after 6 h fermentation with the same molar ratios of acetate, propionate and butyrate (70:17:13) as SCGMR. Longer fermentation time changed the

15

molar ratios of acetate, propionate and butyrate (69:17:14 and 61:22:17 for SCGMR; 66:18:16 and 51:28:21 for SCGDR after 12 and 24 h, respectively). The higher relative propionate production particularly after 24 h fermentation is consistent with the strong increase in propionate resulting from coffee arabinogalactans and galactomannans degradation (Reichardt et al., 2009). Fermentation time also increased molar proportions of butyrate (13-17 and 13-21 % of total SCFAs for SCGMR and SCDR, respectively), thereby exerting beneficial health effects by inhibiting tumor cell proliferation and inducing apoptosis (Campos-Vega et al., 2013). This is the first report of SCFAs production from SCG.

3.4. Effect of hgf-NDSCG (medium roasted) on RAW 264.7 cells viability, LPSinduced NO production and multiple cytokines The hgf-NDSCG from SCGMR was investigated for its anti-inflammatory effects on nitric oxide production in LPS-activated RAW 264.7 macrophages because of its high total phenolic acids and SCFAs. The amount of nitrite accumulated in the culture medium was estimated using Griess reagent as an index of nitric oxide. Nitrite production induced by LPS was significantly (p < 0.05) inhibited at > 10 % hgf-NDSCG in a dose-dependent manner (Fig. 1), but unaffected at 5 and 7.5 %. Approximately 50 % of the nitrite was suppressed with 50 % hgf-NDSCG treatment (Fig. 1); an effect similar to the 54 % inhibition observed with chlorogenic acid (4mM) on LPS induced nitrite production in microglia (Shen et al., 2012). Nevertheless, the amount of chlorogenic acid in hgf-NDSCGLC50 was only 0.0011 mM (unpublished data), suggesting that other compounds may be contributing to this effect. Similar dose-dependent suppression of NO production in Raw 264.7-LPS stimulated macrophage has been reported for kahweol (Shen et al., 2012), a

16

coffee diterpene not evaluated in this study. Caffeic acid, a potent antioxidant present in SCGMR also reduced nitrite concentration in LPS stimulated RAW 264.7 macrophage cells, inhibited JNK1/2 and p38 MAPK phosphorylation and signaling molecules involved in inflammation (Búfalo et al., 2013). However, in this work, caffeic acid was found under the limit of detection during in vitro colonic fermentation (unpublished work).

Since an impressive number of papers have described the effects of SCFAs on inflammatory diseases in recent years, we evaluated the effect of the synthetic mixture of the three main types of SCFAs found in the hgf-NDSCG-LC50 (equivalent to 7.35, 2.64 and 1.97 mmol/L of acetic, propionic, and butyric acids, respectively) on NO production. The synthetic SCFAs mixture was more effective in NO inhibition (72.6 %) than hgf-NDSCGLC50, suggesting that SCFAs in the hgf-NDSCG are mainly responsible for the observed effect, and that other compounds could be acting as antagonists, limiting SCFAs effect . Individual compounds, sodium butyrate, propionate and acetate also inhibited NO production (≈60 %) in LPS-stimulated RAW264.7 cells, at lower concentrations (1-1200 µmol/L) (Liu et al., 2012), thereby supporting our results.

Cell viability evaluated by the MTT assay remained stable at 5-50% hgf-NDSCG and in the presence of the evaluated SCFAs mixture (Fig. 1), demonstrating that such concentrations exhibited strong anti-inflammatory potential without affecting cell viability. LPS induced RAW 264.7 cell cultures incubated with hgf-NDSCG50 were evaluated as in vitro models to explore the underlying anti-inflammatory mechanisms with regards to regulation/expression of cytokines, chemokines and immune responses. Twenty-five of the 40 cytokines were up-regulated (>1-fold), 10 suppressed, and 5 (IFN-γ, M-CSF and 3 CXC

17

chemokines:

CXCL1/KC,

CXCL10/IP-10/CRG,

and

CXCL11/I-TAC)

remained

unchanged (< ± 1-fold) (Table 4). Both CXCL10 and CXCL11 are IFN-γ-inducible chemokines. Cytokines with +/-2-fold change were considered to be differently expressed with CXCL9/MIG and IL-5 overexpressed/activated, and IL-1β/IL-1F2, CCL17/TARC and IL-10 suppressed (associated with the control cells-i.e., down-regulated in the hgfNDSCG50 treatment). NO inhibition/scavenged by hgf-NDSCG50 down regulated (-3.79 fold) the immunosuppressive cytokine IL-10, a key inhibitor of several pro-inflammatory cytokines [IL-1β, IL-6, granulocyte-macrophage colony stimulating factor (GM-CSF), tumor necrosis factor α (TNF-α)] and chemokines IL-2 and IL-3 produced in response to LPS activated macrophages (Zdanov et al., 1995). IL-1β, an important mediator in the early immune response, was also suppressed (-2.57 fold); abnormally elevated levels of this macrophage-derived cytokine has been reported in type 2 diabetic obese subjects (Dalmas et al., 2014). CCL17, a Th2 lymphocyte selective chemokine was drastically reduced/suppressed (-3.21-fold), revealing its chemoattractant activity and concurrent regulation of other chemokines (IL-1β, IL-10, IL-6 and TNF-α) during allergic response of immune-mediated diseases (Katakura et al., 2004). IL-5, another Th2 cytokine involved in allergic response was up-regulated (2.18 fold); this increase was similar to IL-5 secretion (2.1 fold) by long chain n-3 polyunsaturated fatty acids on LPS stimulated splenic/peritoneal B cells with no effect on IL-13 level (Teague et al., 2015). Although IL-5 protects against parasitic infection, it is nevertheless implicated in the pathogenesis of several diseases; therefore, its overexpression could adversely affect inflammatory lung diseases in humans (Teague et al., 2015). The increase in CXCL9/MIG-expression (3.23 fold) reflects an immune response induced by hgf-NDSCG50 similar to the potent antibacterial function of CXCL9 demonstrated by epithelial recognition of group A

18

Streptococcus pyrogenes (Eliasson, 2010). CXCL9 was the only Th1-type immune response, since other cytokines from the same family CXCL10 (0.02 fold) and CXCL11 (-0.07) was unaffected by hgf-NDSCG50. The upregulation of CXCL9 has been proposed as part of a general defense system against colonizing pathogenic bacterial infections on mucosal surfaces such as pharyngeal epithelium (Eliasson, 2010). The cytokines IL-4, IL-7, IL-13, IL-17, IL-23 and IL-27 were similarly activated (1.06 to 1.28 fold) probably by the same Stat proteins that contribute to the redundant cytokine actions also known as cytokine pleiotrophy (Lin et al., 1995). Our results correspond with those from several studies that have associated the inflammatory responses of coffee and/or its key constituents to disease pathology. For example, coffee phenolics (chlorogenic, caffeic and ferulic acids) and caffeine exerted biological activities highlighting possible link to depression through reduction in proinflammatory mediators, TNF-α, IL-1β, IL-6, IFN-γ, probably via the JNK1/2 and p38MAPK signaling pathway (Hall et al., 2015 and references therein). Liver damage in a rat model of steatohepatitis was protected by coffee polyphenols due to reduced expression of proinflammatory TNF-α, IL-1α, IL-1β and increased anti-inflammatory interleukins (IL-4, IL-6 and IL-10) (Vitaglione et al., 2010). Chlorogenic acid, the major coffee polyphenol, was purported to exhibit antiinflammatory effects in mice fed high-fat diets given decaffeinated coffee through significant reduction in cytokine IL-1β expression produced in macrophages (Fukushima et al., 2009). IL-2 and IL-4 were dose dependently reduced in a mouse model by caffeine (Pohanka, 2015). Arabinogalactan, the major coffee bean polysaccharide exerted immunomodulation by stimulating mouse immunocytes and enhancing Th1 (IL-12p40) immune response (Gotoda et al., 2012). Human studies showed contrasting effects of coffee on inflammatory response with IL-6 increase in subjects consuming 200 ml

19

coffee/day in one study, while others revealed no correlation (Frost-Meyer & Logomarsino, 2012 and references therein). Proteins from the STRING network analysis (Fig. 2) clustered the anti-inflammatory response of hgf-NDSCG50-stimulated macrophages along 4 KEGG pathways: “cytokine-cytokine receptor interaction”: this pathway play a role in health and are crucial during immunological and inflammatory responses in disease. Cytokine interactions can result in additive, antagonist, or synergistic activities in maintaining physiological functions such as feeding, body temperature, and sleep, as well as in anorectic, pyrogenic, and somnogenic neurological manifestations of acute and chronic disease (Turrin & Plata-Salamán 2000); “inflammatory bowel disease (IBD)”: the imbalance between pro-inflammatory and anti-inflammatory cytokines that occurs in IBD impedes the resolution of inflammation and instead leads to disease perpetuation and tissue destruction. Recent studies suggest a regulatory cytokines network that has important implications for disease progression; f o r e x a m p l e i n “rheumatoid arthritis” the induction of autoimmunity, chronic inflammation and thereby joint damage is favored by an imbalance between pro- and anti-inflammatory cytokine activities (McInnes & Schett, 2007); and “chemokine signaling pathway” (Fig. 2 A). The biological process according to the gene ontology differentiated the proteins into immune response, positive regulation of the immune system, and inflammatory response. The corresponding GO molecular functions were: cytokine activity, cytokine receptor binding and chemokine activity with all evaluated proteins as part of the extracellular space/region. In this work, bioinformatics analysis was performed to elucidate the network between the identified differential proteins and inflammation function proteins. Figure 2 B illustrates the network edges representing the functional associations (evidence view), confirming the presence of well-known proteins and their functional relationships. Only two of the differentially

20

expressed proteins have binding function. IL10 was found to interact with CCL17 and substantially up-regulated in LPS-stimulated RAW 264.7 cells. However, IL10 network interact with STAT6, SMAD4 and SMAD3, and a large body of evidence supports the role for STAT and SMAD proteins downstream of cytokine receptors in regulating IL-10 expression in Th cells, although this may be tightly linked to the each Th cell subset differentiation (Gabryšová et al., 2014). In addition, IL-1β interact with NFKB1, a pleiotropic transcription factor present in almost all cell types and is the endpoint of a series of signal transduction events initiated by a vast array of stimuli related to many biological processes such as inflammation and with Caspase 1, a thiol protease that cleaves IL-1 beta releasing the mature cytokine involved in various inflammatory processes (Von Mering et al., 2007).

4. Conclusion This study described the immunomodulatory response of non-digested SCG fraction fermented by human gut flora. Metabolites produced by colonic fermentation of SCG exhibited strong anti-inflammatory potential by suppressing NO production, and inhibited inflammatory mediators, mainly IL-10, CCL-17, CXCL9, IL-1β and IL-5 cytokines. These effects were mainly induced by S C F A s p r o d u c e d fr o m t he S C G - d ie t a r y f i b e r c o lo n ic m e t a bo l i s m. The activation of RAW 264.7 macrophages by hgfNDSCG probably reflects the key role SCG can play in protecting the onset and/or progression of chronic inflammatory diseases such as inflammatory bowel disease and rheumatoid arthritis. The results support the use of SGC in the food industry as dietary fiber source with health benefits.

21

Acknowledgments This study was supported by Consejo Nacional de Ciencia y Tecnología (CONACYT) Grant No. 242282. In addition, we would like to thank B. Dave Oomah for his kind revision of this manuscript.

Conflict of interest The authors declare that they have no competing interests.

REFERENCES [AOAC] Association of Official Analytical Chemists (2002) Official methods of analysis.17th ed. Arlington, Va.: AOAC.1652-1661. Búfalo, M. C., Ferreira, I., Costa, G., Francisco, V., Liberal, J., Cruz, M. T., ... & Sforcin, J. M. (2013). Propolis and its constituent caffeic acid suppress LPSstimulated pro- inflammatory response by blocking NF-κB and MAPK activation in macrophages. Journal of Ethnopharmacology, 149(1), 84-92. Campos-Vega, R., Loarca-Piña, G., Vergara-Castañeda, H., & Oomah, B. D. (2015a). Spent coffee grounds: A review on current research and future prospects. Trends in Food Science & Technology, 45(1), 24-36. Campos-Vega, R., Oomah, B. D., Loarca-Piña, G., & Vergara-Castañeda, H. A. (2013). Common beans and their non-digestible fraction: cancer inhibitory activity—an overview. Foods, 2(3), 374-392. Campos-Vega, R., Reynoso-Camacho, R., Pedraza-Aboytes, G., Acosta-Gallegos, J.,

22

Guzmán- Maldonado, S., Paredes-López, O., Oomah, B., Loarca-Piña, G. (2009) Chemical composition and in vitro polysaccharide fermentation of different beans (Phaseolus vulgaris L.). Journal of Food Science, 74(7), T59. Campos-Vega, R., Vázquez-Sánchez, K., López-Barrera, D., Loarca-Piña, G., Mendoza-Díaz, S., & Oomah, B. D. (2015b). Simulated gastrointestinal digestion and in vitro colonic fermentation of spent coffee (Coffea arabica L.): Bioaccessibility and intestinal permeability. Food Research International. In Press. Cho, A. R., Park, K. W., Kim, K. M., Kim, S. Y., & Han, J. (2014). Influence of roasting conditions on the antioxidant characteristics of Colombian coffee (Coffea arabica L.) Beans. Journal of Food Biochemistry, 38(3), 271-280. Dalmas, E., Venteclef, N., Caer, C., Poitou, C., Cremer, I., Aron-Wisnewsky, J., ... & Guerre- Millo, M. (2014). T Cell–Derived IL-22 Amplifies IL-1β–Driven Inflammation in Human Adipose Tissue: Relevance to Obesity and Type 2 Diabetes. Diabetes, 63(6), 1966-1977. Eliasson, M., Mörgelin, M., Farber, J. M., Egesten, A., & Albiger, B. (2010). Streptococcus pneumoniae induces expression of the antibacterial CXC chemokine MIG/CXCL9 via MyD88-dependent signaling in a murine model of airway infection. Microbes and Infection, 12(7), 565-573. Frost-Meyer, N. J., & Logomarsino, J. V. (2012). Impact of coffee components on inflammatory markers: A review. Journal of Functional Foods, 4(4), 819-830. Fukushima, Y., Kasuga, M., Nakao, K., Shimomura, I., & Matsuzawa, Y. (2009). Effects of coffee on inflammatory cytokine gene expression in mice fed high-fat

23

diets. Journal of Agricultural and Food Chemistry, 57(23), 11100-11105. Gabryšová, L., Howes, A., Saraiva, M., & O’Garra, A. (2014). The regulation of IL-10 expression. In Interleukin-10 in Health and Disease (pp. 157-190). Springer Berlin Heidelberg.

Galanakis CM, Martinez-Saez N, del Castillo MD, Barba FJ, Mitropoulou VS (2015). Patented & commercialized applications. En C. M. Galanakis, Food Waste Recovery: Processing technologies and industrial techniques (pp.339-362). Elsevier – Academic Press, [ISBN: 978-0-12-800351-0]. Gotoda, N., Iwai, K., Furuya, K., Ueda, T., Fukunaga, T., Kimura, R., & Takagi, M. (2007).

Arabinogalactan

isolated

from

coffee

beans

indicates

immunomodulating properties. In 21st International Conference on Coffee Science, Montpellier, France, 11-15 September, 2006. (pp. 116-120). Association Scientifique Internationale du Café (ASIC). Hall, S., Desbrow, B., Anoopkumar-Dukie, S., Davey, A. K., Arora, D., McDermott, C., ... & Grant, G. D. (2015). A review of the bioactivity of coffee, caffeine and key coffee constituents on inflammatory responses linked to depression. Food Research International, 76, 626-636. International Coffee Organization (ICO). (2014). World coffee trade (1963 – 2013): A review of the markets, challenges and opportunities facing the sector. Mayo 2014,

International

Coffee

Organization

Sitio

web:

http://www.ico.org/news/icc-111-5-r1e-world-coffee- outlook.pdf Jiménez-Zamora, A., Pastoriza, S., & Rufián-Henares, J. A. (2015). Revalorization of coffee by- products. Prebiotic, antimicrobial and antioxidant properties. LWT-

24

Food Science and Technology, 61(1), 12-18. Katakura, T., Miyazaki, M., Kobayashi, M., Herndon, D. N., & Suzuki, F. (2004). CCL17 and IL-10 as effectors that enable alternatively activated macrophages to inhibit the generation of classically activated macrophages. The Journal of Immunology, 172(3), 1407-1413. Lin, J. X., Migone, T. S., Tseng, M., Friedmann, M., Weatherbee, J. A., Zhou, L., ... & Leonard, W. J. (1995). The role of shared receptor motifs and common Stat proteins in the generation of cytokine pleiotropy and redundancy by IL-2, IL-4, IL-7, IL-13, and IL-15. Immunity, 2(4), 331-339. Lintas, C., & Cappeloni, M. 1988. Content and composition of dietary fiber in raw and cooked vegetables. Food Science and Nutrition, 42, 117–24. Liu, T., Li, J., Liu, Y., Xiao, N., Suo, H., Xie, K., ... & Wu, C. (2012). Short-chain fatty acids suppress lipopolysaccharide-induced production of nitric oxide and proinflammatory cytokines through inhibition of NF-κB pathway in RAW264. 7 cells. Inflammation, 35(5), 1676-1684. McInnes, I. B., & Schett, G. (2007). Cytokines in the pathogenesis of rheumatoid arthritis. Nature Reviews Immunology, 7(6), 429-442. Monente, C., Ludwig, I. A., Irigoyen, A., De Peña, M. P., & Cid, C. (2015). Assessment of Total (Free and Bound) Phenolic Compounds in Spent Coffee Extracts. Journal of Agricultural and Food Chemistry. Nunes, F. M., & Coimbra, M. A. (2007). Melanoidins from coffee infusions. Fractionation, chemical characterization, and effect of the degree of roast. Journal of Agricultural and Food Chemistry, 55(10), 3967-3977.

25

Nunes, F. M., Reis, A., Domingues, M. R. M., & Coimbra, M. A. (2006). Characterization of galactomannan derivatives in roasted coffee beverages. Journal of agricultural and food chemistry, 54(9), 3428-3439. Oosterveld, A., Voragen, A. G. J., & Schols, H. A. (2003). Effect of roasting on the carbohydrate composition of Coffea arabica beans. Carbohydrate Polymers, 54(2), 183-192. Pohanka, M., & Dobes, P. (2013). Caffeine inhibits acetylcholinesterase, but not butyrylcholinesterase. International Journal of Molecular Sciences, 14(5), 98739882. Ramírez-Jiménez, A., Reynoso-Camacho, R., Mendoza-Díaz, S., & Loarca-Piña, G. (2014). Functional and technological potential of dehydrated Phaseolus vulgaris L. Flours. Food Chemistry, 161, 254–260. Ramirez‐Martinez, J. R. (1988). Phenolic compounds in coffee pulp: quantitative determination by HPLC. Journal of the Science of Food and Agriculture, 43(2), 135-144. Redgwell, R. J., Trovato, V., Curti, D., & Fischer, M. (2002). Effect of roasting on degradation and structural features of polysaccharides in Arabica coffee beans. Carbohydrate Research, 337(5), 421-431. Reichardt, N., Gniechwitz, D., Steinhart, H., Bunzel, M., & Blaut, M. (2009). Characterization of high molecular weight coffee fractions and their fermentation by human intestinal microbiota. Molecular Nutrition & Food Research, 53(2), 287-299. Saura-Calixto, F., Goñi, I., Bravo, L., & Mañas, E. (1993). Resistant Starch in foods:

26

modified method for dietary fiber residues. Journal of Food Science, 58(3), 642643. Shen, W., Qi, R., Zhang, J., Wang, Z., Wang, H., Hu, C., ... & Lu, D. (2012). Chlorogenic acid inhibits LPS-induced microglial activation and improves survival of dopaminergic neurons. Brain Research Bulletin, 88(5), 487-494. Shiga, M., Lajolo, M. & Filisetti, M. (2003). Celll wall polysaccharides of common beans (Phaseolus vulgaris L.) Food Science and Technology (Caminas), 23, 141-148. Simões, J., Nunes, F. M., Domingues, M. R., & Coimbra, M. A. (2013). Extractability and structure of spent coffee ground polysaccharides by roasting pre-treatments. CarbohydratePpolymers, 97(1), 81-89. Teague, H., Harris, M., Whelan, J., Comstock, S. S., Fenton, J. I., & Shaikh, S. R. (2015). Short- term consumption of n-3 PUFAs increases murine IL-5 levels but IL-5 is not the mechanistic link between n-3 fatty acids and changes in B cell populations. The Journal of Nutritional Biochemistry. Turrin, N. P., & Plata-Salamán, C. R. (2000). Cytokine–cytokine interactions and the brain. Brain research bulletin, 51(1), 3-9. Vinolo, M., Rodrigues, H., Nachbar, R., & Curi, R. (2011). Regulation of inflammation by short chain fatty acids. Nutrients, 3(10), 858-876. Vitaglione, P., Morisco, F., Mazzone, G., Amoruso, D. C., Ribecco, M. T., Romano, A., ... & D'Argenio, G. (2010). Coffee reduces liver damage in a rat model of steatohepatitis: the underlying mechanisms and the role of polyphenols and melanoidins. Hepatology, 52(5),

27

Von Mering, C., Jensen, L. J., Kuhn, M., Chaffron, S., Doerks, T., Krüger, B., ... & Bork, P. (2007). STRING 7—recent developments in the integration and prediction of protein interactions. Nucleic Acids Research, 35(suppl 1), D358D362. Zdanov, A., Schalk-Hihi, C., Gustchina, A., Tsang, M., Weatherbee, J., & Wlodawer, A. (1995). Crystal structure of interleukin-10 reveals the functional dimer with an unexpected topological similarity to interferon γ. Structure, 3(6), 591-601.

28

Figure(s)

Fig. 1. hgf-NDSCG (medium roasted) effect on cell viability and NO production in lipopolysaccharide-stimulated RAW 264.7 macrophages. hgf-NDSCG= human gut fermented-unabsorbed-SCG fraction from medium roasting, 24 hours. LPS= lipopolysaccharides. Results are the average of 3 independent experiments ± SEM. * means difference between treatments (P ≤ 0.05), Dunnett's test. The effect of hgf-NDSCG was normalized to the blank control (hgf-NDSCG without cells) and to the nontreated control (0%/mL, 100%) cells.

A)

B)

Fig. 2. Functional predicted associations for the evaluated cytokines involved in inflammatory processes. A) Overrepresented KEGG pathways and ontology analysis. B) Network evidence view. The network edges represent the predicted functional associations. Any edge may be drawn with up to 3 differently coloured lines; these lines represent the existence of the three types of evidence used in predicting the associations. A red line indicates the presence of fusion evidence; a green line - neighborhood evidence; a blue line – databases; a black line coexpression.

Tables Table 1. Chemical composition, fiber and resistant starch content of medium/dark roasted coffee beans and spent coffee grounds. Medium roasted Protein Lipid Carbohydrates Ash Total fiber Soluble fiber Insoluble fiber Resistant starch

Dark roasted

Coffee bean

Spent coffe grounds

Coffee bean

16.5 ± 0.5ª 15.9 ± 0.0a 58.5 ± 0.5d 6.2 ± 0.1ª 48.6 ± 0.5b 2.1 ± 0.1ª 46.5 ± 0.5b 4.8 ± 0.8a

15.8 ± 0.1ª 15.1 ± 0.0c 60.3 ± 0.1c 1.8 ± 0.0c 57.1 ± 0.9ª 1.6 ± 0.1b 55.5 ± 0.9ª 5.6 ± 0.2a

15.1 ± 0.1ª 15.7 ± 0.1ª 62.5 ± 0.2b 4.3 ± 0.0b 48.2 ± 1.6b 2.2 ± 0.2bc 45.9 ± 1.7b 5.7 ± 0.0a

Spent coffe grounds

11.5 ± 0.4b 15.3 ± 0.1b 68.4 ± 0.4ª 1.1 ± 0.0d 58.6 ± 0.6ª 1.5 ± 0.2c 57.1 ± 0.7ª 6.2 ± 0.0a

Results are the average of 3 independent experiments ± SEM and are expressed as percentage (%) per gram of dry basis sample. Protein content was calculated as nitrogen × 6.25. Means in a row with different letters are significantly different (p < 0.05).

Table 2. Polyphenols and caffeine (HPLC-DAD) content of medium/dark roasted coffee beans and spent coffee grounds. Medium Roasted

a

Polyphenolsa

Coffee bean

Gallic acid Chlorogenic acid Catechin Caffeic acid Rutin Ellagic acid p-coumaric acid Ferulic acid Quercetin Total polyphenols Caffeine

3.6 ± 0.06a 8.3 ± 0.3a 1.5 ± 0.1a 0.3 ± 0.0a 0.2 ± 0.1a 0.2 ± 0.03ab 0.03 ± 0.0a 0.007 ± 0.0ab 0.7 ± 0.1ab 14.83 1.0 ± 0.0a

Spent coffee grounds 2.5 ± 0.06b 5.6 ± 0.1b 0.6 ± 0.1b 0.07 ± 0.0b 0.06 ± 0.01b 0.1 ± 0.01bc 0.01 ± 0.0b 0.004 ± 0.0c 0.96 ± 0.1ab 9.9 0.4 ± 0.0b

Dark Roasted

Coffee bean 3.06 ± 0.5ab 7.8 ± 1.0a 0.6 ± 0.01bc 0.3 ± 0.01a 0.2 ± 0.01a 0.3 ± 0.06a 0.02 ± 0.0a 0.006 ± 0.0bc 0.6 ± 0.1b 12.88 0.9 ± 0.0a

Spent coffee grounds 1.3 ± 0.06c 1.8 ± 0.1c 0.3 ± 0.1c 0.03 ± 0.0c 0.06 ± 0.00b 0.06 ± 0.00c 0.01 ± 0.0b 0.01 ± 0.0a 1.0 ± 0.0a 4.57 0.4 ± 0.0b

Results are the average of 3 independent experiments ± SEM and are expressed as mg/per gram of sample. Means in a row with different letters are

significantly different (p < 0.05).

Table 3. Amount of short-chain fatty acids (SCFAs) (mmol/L) and pH in fermented hgf-NDSCG of medium/dark roasted.

Parameter pH

Sample SCGMR

0h

6h

7.4 ± 0.03aA 7.4 ± 0.02aA

6.8 ± 0.1aC 6.8±0.1aC

SCGMR

7.4 ± 0.04aA -

4.9 ± 0.2bB 37.5 ±7.4abA

SCGDR

-

Raffinose

-

42.5 ± 5.1bA 65.7 ± 2.6cA

SCGMR

-

SCGDR

SCGDR Raffinose

Acetate*

Propionate*

Butyrate*

12 h 6.9 ± 0.0aBC 6.9 ± 0.0aBC 4.6 ± 0.2bB 37.6 ± 9.2abA 28.1 ± 6.2abAB

24 h 7.1 ± 0.0aB 6.9 ± 0.0aB 4.5 ± 0.0bB 29.4 ± 4.3cdA

104.3 ± 2.0dC

18.8 ± 3.6aB 77.6 ± 1.8cB

-

9.0 ± 0.6aA 10.1 ± 1.0aA

9.3± 1.2aA 7.5 ± 0.6aA

10.6 ± 0.6aA 10.4 ± 0.4aA

Raffinose

-

8.1 ± 0.4aA

8.0 ± 0.5aA

18.8 ± 1.2bB

SCGMR

-

SCGDR

-

7.2 ± 0.3aA 7.7 ± 0.4aA

7.3 ± 0.5aA 6.7 ± 0.2aA

7.9 ± 0.3aA 7.8 ± 0.2aA

Raffinose

-

6.8 ± 0.2aA

6.5 ± 0.2aA

11.5 ± 0.5bB

-No data. Results are the average of 3 independent experiments ± SEM. The SCFAs production was normalized to the blank control (without hgfNDSCG). Means in the same column for samples in each section with different small letters are different (p < 0.05). Means in the same row for hours with different letters are different (p < 0.05). SCGMR=Spent Coffee Grounds Medium Roasted, SCGDR=Spent Coffee Grounds Dark Roasted.

Table 4. hgf-NDSCG50 (from medium roasted) effect on the expression of cytokine proteins

Cytokine

Fold change

Cytokine

Fold change

IL-2 IL 1ra/IL-1F3 IL 1β/IL-1F2 IL1 α/IL-1F1 INF-γ sICAM-1/CD54 GM-CSF IL-12p70 IL-13 IL-10 IL-6 IL-4 CCL5/RANTES CCL12/MCP-5 TNF-α IL-23 IL-27 CCL11/EOTAXIN G-CSF IL-17

‐1.78 ‐1.09 -2.57 ‐1.29 0.02 1.28 1.15 1.10 1.07 -3.79 1.25 1.06 ‐1.04 ‐1.24 1.23 1.28 1.14 1.14 1.20 1.17

IL-16 IL-7 IL-3 CCL2/JE/MCP-1 CXCL1/KC CXCL11/I-TAC CXCL10/IP-10/CRG-2 TREM-1 TIMP-1 C5/C5A CXCL13/BLC/BCA-1 IL-5 CXCL12/SDF1 CXCL2/MIP2 CCL4/MIP-1β CXCL9/MIG M-CSF CCL17/TARC CCL1/I-309/TCA-3 CCL3/MIP-1α

‐1.39 1.12 1.17 1.22 ‐0.12 ‐0.07 0.02 1.09 1.07 1.20 1.27 2.18 1.93 ‐1.24 1.46 3.23 ‐0.45 -3.21 1.65 1.30

Data are expressed as fold change relative to control (LPS-stimulated RAW 267.4 cells, without treatment). hgf-NDSCG50 = human gut fermented unabsorbed SCG fraction-LC50. The reported fold changes are the average of two independent experiments.

Highlights SCG had higher total fiber and resistant starch content compared with coffee beans SCG can be fermented by colon microbiota producing short-chain fatty acids Human gut fermented-unabsorbed-SCG fraction suppress NO production Hgf-NDSCG modulates IL-10, CCL-17, CXCL9, IL-1β, and IL-5 cytokines SCG could protect the onset and/or progression of chronic inflammatory diseases