Coffee reduces liver damage in a rat model of steatohepatitis: The underlying mechanisms and the role of polyphenols and melanoidins



This article is corrected by:

  1. Errata: Correction Volume 53, Issue 4, 1416, Article first published online: 7 April 2011

  • Potential conflict of interest: Nothing to report.


Epidemiological data associate coffee consumption with a lower prevalence of chronic liver disease and a reduced risk of elevated liver enzyme levels (γ glutamyl transpeptidase and alanine aminotransferase), advanced liver disease and its complications, and hepatocellular carcinoma. Knowledge of the mechanisms underlying these effects and the coffee components responsible for these properties is still lacking. In this study, 1.5 mL/day of decaffeinated coffee or its polyphenols or melanoidins (corresponding to approximately 2 cups of filtered coffee or 6 cups of espresso coffee for a 70-kg person) were added for 8 weeks to the drinking water of rats who were being fed a high-fat, high-calorie solid diet (HFD) for the previous 4 weeks. At week 12, HFD + water rats showed a clinical picture typical of advanced nonalcoholic steatohepatitis compared with control rats (normal diet + water). In comparison, HFD + coffee rats showed: (1) reduced hepatic fat and collagen, as well as reduced serum alanine aminotransferase and triglycerides; (2) a two-fold reduced/oxidized glutathione ratio in both serum and liver; (3) reduced serum malondialdehyde (lipid peroxidation) and increased ferric reducing antioxidant power (reducing activity); (4) reduced expression of tumor necrosis factor α (TNF-α), tissue transglutaminase, and transforming growth factor β and increased expression of adiponectin receptor and peroxisome proliferator-activated receptor α in liver tissue; and (5) reduced hepatic concentrations of proinflammatory TNF-α and interferon-γ and increased anti-inflammatory interleukin-4 and interleukin-10. Conclusion: Our data demonstrate that coffee consumption protects the liver from damage caused by a high-fat diet. This effect was mediated by a reduction in hepatic fat accumulation (through increased fatty acid β-oxidation); systemic and liver oxidative stress (through the glutathione system); liver inflammation (through modulation of genes); and expression and concentrations of proteins and cytokines related to inflammation. (HEPATOLOGY 2010;52:1652-1661)

Nonalcoholic fatty liver disease (NAFLD) is considered the hepatic manifestation of the metabolic syndrome and is associated with its clinical features, including visceral obesity, dislipidemia, and type 2 diabetes.1 NAFLD has high prevalence in the general population, and it can evolve into nonalcoholic steatohepatitis (NASH), cirrhosis, and complications such as liver failure and hepatocellular carcinoma.2, 3

A model in which rats without genetic modifications are given a high-fat, high-calorie diet is the gold standard for studying the pathogenetic factors involved in NASH.4 NASH is a chronic inflammatory state in which ROS and several immunomodulatory factors contribute to liver injury. It is well documented that many natural substances, such as common foods and beverages, may counteract the progression of NAFLD toward NASH.5

Coffee is the most consumed beverage worldwide, mainly due to the psychoactive properties of caffeine and despite some beliefs that its consumption may have negative health consequences.

In the last two decades, coffee has been studied for its beneficial effects on human health.6 Epidemiology studies from different countries have clearly associated coffee consumption with a lower prevalence of chronic liver disease and have found an inverse association of coffee intake (>2 cups/day) with the risk of elevated γ glutamyl transpeptidase or of alanine aminotransferase (ALT) levels.7-9 In cohort studies in patients with alcoholic and nonalcoholic cirrhosis, the association between coffee intake and positive modulation of liver enzymes was found to be even stronger in alcohol consumers.10, 11 Finally, coffee has been reported to reduce the risk of advanced liver disease and its complications12-14 as well as hepatocellular carcinoma.15-17 Despite such evidence, the knowledge of the mechanisms underlying the protection of coffee on subset of liver diseases and on their progression from fatty liver to fibrosis, as well as the individuation of coffee components responsible for these properties, are still lacking.

Of the many bioactive molecules of coffee, clinical studies have focused almost exclusively on caffeine. However, coffee contains several other biologically active substances whose relative concentration may be highly different depending on the type of coffee used as well as the brewing process performed. Polyphenols and melanoidins play a major role due to their concentrations as well as their numerous health properties (including antioxidant, anti-inflammatory, and prebiotic properties; see review by Wang and Ho18).

The aims of this study were to evaluate the effects of coffee and its constituents—namely polyphenol and melanoidin fractions—on the progression of NASH in a rat model of a high-fat diet and to shed some light on the mechanisms underlying these effects.


adipo-R2, adiponectin receptor 2; ALT, alanine aminotransferase; FRAP, ferric reducing antioxidant power; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GSH, reduced glutathione; GSSG, oxidized glutathione; HFD, high-fat diet; IFN-γ, interferon-γ; IL-interleukin; MDA, malondialdehyde; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; PPAR, peroxisome proliferator-activated receptor; TGF-β, transforming growth factorβ; TNF-α, tumor necrosis factor α; tTG, tissue transglutaminase.

Materials and Methods

Preparation of Coffee-Based Beverages.

Coffee-based beverages were prepared by filtering on a paper filter a mix of boiling water and decaffeinated coffee powder (4:1 v:w) (Illy Caffë, Trieste, Italy). Filtered coffee was portioned and stored at −20°C until used.

Melanoidins were extracted from filtered decaffeinated coffee by way of dialysis against milli-Q water through a 12- to 14-kDa cutoff membrane for 1 week at 4°C. The nondialyzed fraction was collected and reconstituted to the starting volume with milli-Q water. The product was aliquoted and stored at −20°C until use as a melanoidin-based beverage.

Polyphenols were extracted from decaffeinated coffee powder by way of solid/liquid extraction with ethanol (1:4 wt/vol). Filtered ethanol was dried under a vacuum at room temperature and recovered by the container dissolving the dry extract with the same starting volume of a 0.5% ethanol aqueous solution. Aliquots of the polyphenol-based beverage were stored at −20°C until use.

The above preparations were diluted daily in 20 mL water to afford rats a daily dosage of 1.5 mL. The final beverages consumed by rats per 100 mL had the following composition: coffee, 280 mg and 150 mg of polyphenols and melanoidins, respectively; polyphenol-based beverage, 280 mg of polyphenols and no melanoidins; and melanoidin-based beverage, 150 mg melanoidins and no polyphenols.

Animal Model of NASH.

Male Wistar rats weighing 180-220 g were randomly housed in wire-bottomed cages. Animals were obtained from Harlan Italy (Udine, Italy) and maintained under controlled temperature conditions of 22 ± 1°C with a 12-hour light/dark cycle and free access to water. NASH was induced by way of a high-fat diet (HFD) for 3 months as described by Svegliati-Baroni et al.4; the diet composition was: 58% of energy derived from fat, 18% from protein, and 24% from carbohydrates; 5.6 kcal/g (Harlan Italy). Control rats were fed a standard pellet diet (5% of energy derived from fat, 18% from proteins, and 77% from carbohydrates [3.3 kcal/g]) for 3 months.


Thirty rats were divided into five groups. Animals in four groups were fed an HFD for 3 months to develop NASH. Over the same period, the fifth group of animals was fed a standard diet (normal chow) and represented the healthy control group. Beginning with the second month, the four HFD-fed groups were differentiated for beverages: they drank decaffeinated coffee, a solution of coffee polyphenols or melanoidins, or water. Healthy control rats drank water throughout the study period.

The amount of coffee or coffee fractions was administered at a daily dose corresponding to 6 cups of espresso coffee or 2 cups of filtered coffee for a person weighing 70 kg. The experimental protocol was approved by the University's ethics committee. Food intake was recorded daily, and body weight was measured weekly. All animals were sacrificed 3 months after beginning the diet.

All animals received humane care according to the criteria outlined in the National Institutes of Health Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication 86-23; revised 1985).

Serum Analyses.

Serum aspartate aminotransferase, ALT, alkaline phosphatase, γ glutamyl transpeptidase, and total, high-density lipoprotein, and low-density lipoprotein cholesterol and triglyceride concentrations were determined to assess liver function using a Modular Autoanalyzer (Roche Diagnostics, GmbH, Mannheim, Germany).

Serum markers of antioxidant status such as total, reduced (GSH) and oxidized (GSSG) glutathione as well as ferric reducing antioxidant power (FRAP) were measured according to the procedures reported below.


Ten percent formalin-fixed paraffin-embedded sections of liver samples were divided into 4-μm sections by using routine techniques and mounted onto slides with coverslips. Representative sections of each fixed sample were stained with standard hematoxylin-eosin and Sirius red/fast green according to standard protocols. All histological analyses were performed by an experienced histopathologist in a blinded manner.

Liver Collagen Content.

For the detection and quantification of collagen, liver sections were stained with Picrosirius red solution. The extent of liver fibrosis was determined as the proportion of Picrosirius-stained area in each section. For each rat, 64 fields of a constant raster of 31 mm2 were analyzed at 100× final magnification. For semiautomated morphometry, a Sony 3CCD (model DXC-950P) videomicroscope equipped with a motor stage and the Quantimed 500MC (Leica, Germany) software were used.


To detect the immunohistochemical localization of adiponectin receptor 2 (adipo-R2), sections from formalin-fixed, paraffin-embedded specimens were deparaffinized and rehydrated in decreasing concentrations of ethyl alcohol. The detailed procedure, including antibody used and all material specificities and provenience, is provided in the Supporting Information.

Protein Extract Preparation and Western Blot Analysis.

Western blot analyses in tissue lysates prepared and quantified for protein content were performed as described in the Supporting Information.

RNA Extraction and Reverse-Transcription Polymerase Chain Reaction.

Semiquantitative reverse-transcription polymerase chain reaction amplification of messenger RNA liver extracts was performed using the procedure and primers described in the Supporting Information.

Analysis of Antioxidant Status Markers.

Markers of antioxidant status such as total, reduced (GSH), and oxidized (GSSG) glutathione in serum and liver samples; glutathione transferase activity in the liver; and plasma FRAP were measured according to the protocols described in the Supporting Information.

Analysis of Inflammatory Markers.

Tumor necrosis factor α (TNF-α), interferon-γ (IFN-γ), interleukin (IL)-1α, IL-1β, IL-2, IL-4, IL-6, and IL-10 were quantified in liver samples through the xMAP technology developed by Luminex (Austin, TX) and using a rat multiplex bead-based assay Bio-Plex Suspension Array System (Bio-Rad Laboratories, Hercules, CA) according to the manufacturer's instructions. Data were analyzed using Bio-Plex Manager version 3.0 (Bio-Rad Laboratories) with five parameter logistic regression algorithm curve fits. Detection limits for the cytokines were 2-32,000 pg/mL. A detailed protocol for tissue preparation is described in the Supporting Information.

Liver cytokine concentrations in rats belonging to the different groups are expressed as the mean ± SE. Serum c-reactive protein concentration was also measured using a rat c-reactive protein enzyme-linked immunosorbent assay kit obtained from BD Biosciences (San Diego, CA) following the manufacturer's instructions.

Statistical Analysis.

Groups of data were compared using analysis of variance followed by Tukey's multiple comparison tests. P < 0.05 was considered statistically significant.


Weight Gain.

The mean daily energy intakes of HFD-fed rats were 15% higher than those of control rats (75 ± 4 kcal/die versus 65 ± 2 kcal/die). A 22% average higher weight gain in HFD-fed rats versus control rats (365 ± 4 g versus 300 ± 3 g, respectively) was recorded from the 10th week of feeding through the end of the study. No difference in weight gain was found among HFD-fed rats drinking coffee, polyphenols, melanoidins or water.

Serum Biochemical Parameters.

No significant differences were found among treatment groups for the concentrations of aspartate aminotransferase, alkaline phosphatase, and γ glutamyl transpeptidase. Total cholesterol was not statistically different in HFD + water versus control rats (67.8 ± 4.9 mg/dL versus 51.2 ± 2.0 mg/dL [P value not significant]) and a concentration close to that of control animals was found in the HFD + coffee group (56.3 ± 2.6 mg/dL [P value not significant]). Serum concentrations of high-density lipoprotein and low-density lipoprotein cholesterol were not different in HFD + water versus control rats (19.4 ± 2.3 mg/dL versus 16.6 ± 3.6 mg/dL and 10.5 ± 6.3 mg/dL versus 13.4 ± 8.2 mg/dL, respectively [P value not significant]) as well as versus HFD + coffee (14.6 ± 2.9 mg/dL and 16.5 ± 0.7 mg/dL, respectively), HFD + polyphenols (19.0 ± 3.3 mg/dL and 12.0 ± 1.4 mg/dL, respectively), and HFD + melanoidins (15.8 ± 3.6 mg/dL and 16.0 ± 1.3 mg/dL, respectively).

Serum triglyceride and ALT levels were significantly increased in HFD-fed rats compared with controls. A significant reduction of triglycerides was found only in rats treated with coffee or melanoidins (Fig. 1A), whereas a reduction of serum ALT concentration was found with both coffee and the two components (Fig. 1B).

Figure 1.

Mean (A) triglycerides and (B) ALT concentrations of serum samples from rats fed a normal diet and drinking water (control rats) and rats fed an HFD and drinking water (HFD + water), coffee (HFD + coffee), coffee polyphenols (HFD + polyphenols), or coffee melanoidins (HFD + melanoidins). *P < 0.05 versus control rats; **P < 0.05 versus HFD + water.

Histology of Liver Samples.

Hematoxylin-eosin and Sirius red staining in livers of normal rats are shown in Fig. 2A and 2B, respectively. Steatosis affected a large number of hepatocytes, with the presence of diffuse ballooning and foci of inflammatory cell infiltration present throughout the lobule (Fig. 2C). Both the presence of lipid droplets and the inflammatory infiltrate were significantly reduced by coffee, polyphenols, or melanoidins (Fig. 2E). Sirius red staining in HFD rats revealed the presence of red-stained collagen fibers (Fig. 2D) as an index of hepatic fibrosis. The fibrotic septa significantly regressed after intake of coffee, polyphenols, or melanoidins (Fig. 2F).

Figure 2.

Liver histology and collagen staining. Sample slides were stained with hematoxylin-eosin or Sirius red to highlight the collagen fibers. (A) Hematoxylin-eosin and (B) Sirius red staining of normal rat liver. (C) Steatosis affected a large number of hepatocytes with the presence of diffuse ballooning and foci of inflammation throughout the lobule. (D) Red-stained collagen fibers in HFD-fed rats indicate hepatic fibrosis. (E) Coffee, polyphenols, or melanoidins significantly reduced both lipid droplets and inflammatory infiltrate. (F) The fibrotic septa significantly regressed after treatment with either coffee, polyphenols, or melanoidins.

Hepatic Expression of Genes and Proteins Related to Fat Metabolism, Inflammation, and Fibrosis.

According to liver inflammation and collagen deposition in fibrotic septa, as evidenced by histology, TNF-α and tissue transglutaminase (tTG) expressions were higher in HFD-fed rats than in control rats (Fig. 3A,B). Transforming growth factor β (TGF-β), which is activated by tTG during the fibrotic process, was also up-regulated in HFD-fed rats (Fig. 3A). Among the molecules belonging to the peroxisome proliferator-activated receptor (PPAR) family, the α-type isoform is involved in hepatic lipid metabolism and is regulated by adipokines such as adiponectin. Adipo-R2 serves as a receptor for the globular and full-length adiponectin molecule. In HFD-fed rats, both PPAR-α and adipo-R2 were significantly reduced compared with the control group (Fig. 3A,B).

Figure 3.

TNF-α, PPAR-α, tTG, TGF-β, and adipo-R2 gene expression in livers of control rats and HFD-fed rats treated with coffee and its components. (A) Reverse-transcription polymerase chain reaction coamplification of TNF-α/glyceraldehyde 3-phosphate dehydrogenase (GAPDH), PPAR-α/GAPDH, tTG/GAPDH, TGF-β/GAPDH, and adipo-R2/GAPDH. (B) TNF-α, PPAR-α, tTG, TGF-β, and adipo-R2 protein expression in livers of control rats and HFD-fed rats treated with coffee and its components.

TNF-α, TGF-β, and tTG up-regulation was counteracted by treatment with coffee, polyphenols, or melanoidins (the latter to a lesser extent). PPAR-α and adipo-R2 down-regulation in HFD-fed rats might represent a hepatic feature of NASH, indicating that reduced lipid breakdown occurred in this rat model of NASH in addition to an increased fatty acid afflux to the liver. Indeed, coffee or coffee polyphenol treatment associated with an HFD counteracted this down-regulation (Fig. 3A,B).

As shown by immunohistochemistry (Fig. 4), adipo-R2, which was widely distributed in control rats, was scarcely represented in HFD-fed rats and was restored by coffee or coffee polyphenols. This finding suggests that both coffee and its polyphenols may account for the reduced deposition of cholesterol and triglycerides in the liver.

Figure 4.

Immunohistochemistry of adipo-R2. (A) Control rats. (B) HFD-fed rats. (C) HFD + coffee rats.

Antioxidant Status.

The biomarkers of antioxidant status measured in both serum and liver samples from each group are shown in Table 1. The data showed that in both serum and livers, HFD-fed rats always had significantly higher concentrations of GSSG than control rats. Coffee, polyphenols, or melanoidins reduced GSSG concentrations in HFD-fed rats drinking coffee compared with those drinking water. In particular, a clear effect of the three coffee beverages on GSSG reduction (P < 0.05 versus HFD + water) was recorded in serum samples, whereas it was not significant in the livers of melanoidin-treated rats (0.82 ± 0.14 nmol/mg protein versus 1.02 ± 0.09 nmol/mg protein). In contrast, GSH levels were not significantly altered by an HFD except for a decrease in the livers of rats drinking melanoidins. Among HFD-fed rats, a significant increase of systemic GSH was associated with polyphenol treatment, whereas a slight reduction was associated with melanoidins (116.65 ± 11.43 μM and 64.32 ± 7.82 μM versus 72.99 ± 12.66 μM, respectively).

Table 1. Serum and Liver Biomarkers of Rat Antioxidant Status at the End of Each Dietary Treatment
 Control RatsHFD + WaterHFD + CoffeeHFD + PolyphenolsHFD + Melanoidins
  • Rats were fed an HFD for 3 months and administered coffee, coffee polyphenols, or coffee melanoidins beginning with the second month of dietary treatment. Data are presented as the mean ± SE (n = 6 per treatment group).

  • *, †

    P < 0.05 versus control.

  • P < 0.05 versus HFD + water.

 Total GSH (μM)146.78 ± 24.96129.44 ± 8.65112.93 ± 5.16*,149.65 ± 12.14105.15 ± 6.01*,
 GSSG (μM)20.33 ± 4.8456.44 ± 5.16*28.33 ± 0.27*,33.00 ± 2.89*,40.83 ± 0.67*,
 GSH (μM)94.75 ± 20.9972.99 ± 12.6684.60 ± 5.33116.65 ± 11.4364.32 ± 7.82
 Malondialdehyde (μM)1.47 ± 0.122.03 ± 0.14*1.50 ± 0.09*,1.62 ± 0.081.83 ± 0.14*
 FRAP (mM TE)0.39 ± 0.010.32 ± 0.01*0.27 ± 0.03*,0.36 ± 0.020.29 ± 0.02*
 Total GSH (nmol/mg protein)0.80 ± 0.071.55 ± 0.13*0.93 ± 0.06*,1.12 ± 0.101.11 ± 0.11
 GSSG (nmol/mg protein)0.26 ± 0.041.02 ± 0.09*0.44 ± 0.09*,0.57 ± 0.09*,0.82 ± 0.14*
 GSH (nmol/mg protein)0.53 ± 0.070.53 ± 0.150.50 ± 0.110.55 ± 0.050.29 ± 0.12*,
 GST activity (nmol/min/mg protein)3.01 ± 0.113.01 ± 0.092.55 ± 0.05*,3.10 ± 0.143.10 ± 0.06

The value of the GSH/GSSG ratio, indicating the effective grade of systemic and organ oxidative state regulated by a glutathione-based antioxidant system, showed that: (1) HFD-fed rats always had a lower GSH/GSSG ratio than the ones found in control rats (due to the increase of systemic and liver GSH oxidation by HFD); (2) GSH/GSSG in HFD-fed rats and drinking experimental beverages were always higher than the ones from rats belonging to the HFD + water group except in livers of rats from HFD + melanoidins (due to high reduction of liver GSH in this group); (3) among HFD-treated rats, those drinking polyphenols or coffee showed activated protection by an endogenous GSH antioxidative system in serum and liver, respectively.

Malondialdehyde concentration was significantly higher in rats in the HFD + water group than in control rats (2.03 ± 0.14 μM versus 1.47 ± 0.12 μM) and it returned to control values in rats drinking coffee or polyphenols (1.50 ± 0.09 μM or 1.62 ± 0.08 μM versus 1.47 ± 0.12 μM).

Plasma total antioxidant capacity (FRAP) was significantly reduced by an HFD in rats from all groups, but a significant increase of FRAP was found in rats drinking polyphenols compared with those drinking water (0.36 ± 0.02 mM TE versus 0.32 ± 0.01 mM TE). In contrast, a reduction of FRAP in coffee-treated compared with water-treated rats was found (0.27 ± 0.03 mM TE versus 0.32 ± 0.01 mM TE).

No significant effect of an HFD on liver glutathione transferase activity was recorded (see control rats versus HFD + water rats in Table 1). Reduced activity of the enzyme was found only in coffee-treated rats (2.55 ± 0.05 nmol/min/mg protein versus 3.01 ± 0.11 nmol/min/mg protein or 3.01 ± 0.11 nmol/min/mg protein).

Liver Inflammation.

The concentrations of five proinflammatory and two anti-inflammatory cytokines in liver samples from rats belonging to the experimental and control groups as well as the percentage variations of cytokine concentration of HFD-fed rats versus those from control rats are presented in Table 2 and Fig. 5, respectively. These data indicate the following: (1) The concentrations of IFN-γ and TNF-α were significantly higher in HFD-fed rats than in control rats drinking water, whereas the concentration of IL-6 was lower. No differences were found for IL-1a or IL-1b concentrations. (2) IFN-γ and TNF-α concentrations were reduced by coffee (17% and 42% less abundant in HFD+coffee than in HFD+water, respectively). In contrast, a 26% higher concentration of IL-6 was found after coffee consumption versus water consumption in HFD-fed rats. (3) The concentrations of IL-1a and IL-1b decreased by 24% and 10%, respectively, in rats treated with coffee polyphenols, whereas the concentration of IL-6 increased by 87% compared with HFD-fed rats drinking water. (4) All proinflammatory cytokines (except IL-6 and IFN-γ, which were unchanged) were significantly less abundant in HFD-fed rats drinking melanoidins than those drinking water. The effect of melanoidins was more important in TNF-α, IL-1α, and IL-1b, because reductions of 58%, 31%, and 15%, respectively, were found in melanoidin-drinking versus water-drinking rats. (5) The two anti-inflammatory cytokines (IL-4 and IL-10) were always at higher concentrations in the livers of HFD-fed rats drinking coffee or its fractions than in those of rats drinking water, suggesting that these cytokines are involved in the biochemical pathways contributing to ameliorate tissue inflammation in HFD-fed rats drinking coffee.

Figure 5.

Pattern of cytokines in liver samples. Data are reported as percentage variations of cytokine concentration versus control group (HFD + water). aP < 0.05 versus control group; bP < 0.05 versus HFD + water.

Table 2. Liver Inflammatory Status
Liver (pg/mg Protein)IL-1aIL-1bIL-6IFN-γTNF-αIL-4IL-10
  • Data are presented as the mean ± SE (n = 6 per treatment group).

  • *

    P < 0.05 versus control group.

  • P < 0.05 versus HFD + water.

Control rats709.82 ± 41.36214.73 ± 10.7954.97 ± 3.2558.60 ± 4.2524.00 ± 0.6632.26 ± 5.24900.84 ± 66.10
HFD + water697.32 ± 48.32218.24 ± 6.0945.95 ± 1.74*91.89 ± 8.87*67.65 ± 2.91*30.24 ± 1.131,035.20 ± 33.23*
HFD + coffee720.06 ± 23.44241.18 ± 10.44*57.89 ± 2.3675.90 ± 1.37*,38.95 ± 2.66*,46.97 ± 3.64*,1,152.68 ± 31.35*,
HFD + polyphenols528.78 ± 42.41*,195.49 ± 13.2785.94 ± 10.92*,96.83 ± 7.49*66.45 ± 5.51*54.18 ± 2.59*,1,338.95 ± 68.48*,
HFD + melanoidins478.65 ± 13.8*,186.07 ± 4.12*,40.21 ± 1.09*82.71 ± 2.27*28.25 ± 1.11*,51.39 ± 5.03*,1,002.15 ± 38.89*

The results of a serum c-reactive protein concentration analysis revealed a slight but not significant increase in serum c-reactive protein of HFD-fed rats compared with control rats drinking water. In contrast to liver cytokines, neither coffee nor its components modulated this parameter in this model of NASH, because no difference among treatments was found in HFD-fed rats (HFD + coffee, 291 ± 31.3 ng/mL; HFD + polyphenols, 331 ± 30.7 ng/mL; HFD + melanoidins, 306 ± 33.3 ng/mL; HFD + water, 292 ± 18.0 ng/mL).


Clinical studies on coffee have focused almost exclusively on caffeine; however, mounting evidence suggests that other coffee components are responsible for its effects, particularly on the liver. In this study, a decaffeinated coffee brew was used in parallel with two of its main components—polyphenol and the high molecular weight polysaccharide fraction melanoidin—in a well-known animal model of NASH.4

A prerequisite to explaining epidemiological evidence by way of an intervention study is to use a coffee dosage in the order of magnitude of its dietary intake. We therefore selected a daily dosage of coffee of about 1.5 mL for this study. This corresponds to about 6 cups/day of espresso or 2 cups/day of filtered coffee for a 70-kg person. Accordingly, the doses for polyphenols and melanoidins were fixed at about 4.2 mg/day of polyphenols and 15 mg/day of melanoidins.

The first evidence of the study was that the administration of coffee and its components at these physiological dosages has a beneficial effect on the liver functions of HFD-fed rats. Histological evaluations of HFD-fed rat livers showed a picture typical of NASH: presence of intrahepatocyte lipid droplets, widespread inflammatory infiltration, perivenular fibrosis, and the formation of porto-central septa. Necrotic damage was also documented by aminotransferase concentrations that were three-fold higher than those of control rats. One consequence of NASH is its evolution toward liver fibrosis, which was present in HFD-fed rats, as evidenced by Sirius red–positive staining and increased expression of tTG. The release into the extracellular matrix of tTG activates latent TGF-β, which increases the tTG expression further.

The biochemical data showed that, compared with HFD-fed rats drinking water, HFD-fed rats drinking coffee or its components had: (1) reduced fat and collagen deposition as well as reduced serum ALT; (2) reduced expression of TNF-α, tTG, and TGF-β and an increased expression of adipo-R2 and PPAR-α in liver tissue; (3) a two-fold GSH/GSSG ratio in both serum and liver tissue; (4) less systemic lipid peroxidation (−18% malondialdehyde concentration in coffee-treated rats); (5) reduced concentrations of proinflammatory cytokines such as TNF-α and IFN-γ and increase of anti-inflammatory ones (IL-4 and IL-10) in liver tissue. These data provide some indications about the mechanisms through which coffee modulates lipid deposition as well as the antioxidant and inflammatory status of rats fed an HFD.

How Coffee Reduces Fat Deposition.

The ability of coffee and chlorogenic acid to suppress body weight gain induced by an HFD in rats has been demonstrated in previous studies.19, 20 To the best of our knowledge, the molecular mechanisms underlying the protection of the liver by coffee are still unknown. The data of this study revealed an up-regulation of PPAR-α gene expression, indicating a higher rate of β-oxidation in the livers of HFD-fed rats that drank coffee or coffee components versus rats that drank water. The increased β-oxidation of fatty acids by PPAR-α in the livers of rats with NASH that drank coffee implies a reduced risk of steatosis progressing toward steatohepatitis and successive fibrosis. This finding is further supported by the down-regulation of tTG and TGF-β in coffee-, polyphenol-, and melanoidin-treated rats compared with water-treated ones (Fig. 3).

TNF-α modulates insulin sensitivity and other metabolic processes at a hepatic level through transcription factors such as PPAR-α, which may regulate lipid metabolism by inducing catabolism of fatty acids, thereby preventing fat deposition and subsequent hepatic damage.21-23

Recently, Cho et al.20 reported that caffeine and chlorogenic acid increased fatty acid β-oxidation activity and PPAR-α expression in the livers of HFD-fed mice compared with controls.

How Coffee Ameliorates Liver Antioxidant Status.

Much evidence from in vitro and animal studies has indicated that the increase of GSH induced by coffee may be mediated by its ability to activate, through Nrf2/EpRE activity, antioxidant response element–dependent genes encoding antioxidant proteins and phase II detoxifying enzymes, thus playing a role in the prevention of liver carcinogenesis. Among the coffee constituents responsible for these effects, cafestol, kaweol, caffeine, chlorogenic acid, and melanoidins have been considered (for a review, see Tao et al.24 and Paur et al.25). Cafestol, kaweol and caffeine were not present in the beverages used in this study, and the data suggest that chlorogenic acid, the major coffee polyphenol, was primarily responsible for the modulation of serum GSH concentration. In fact, a higher GSH/GSSG ratio was found in samples from rats treated with coffee polyphenols than in those from rats drinking coffee. Thus, coffee consumption guaranteed systemic and liver endogenous antioxidant protection through the glutathione system, mainly due to its polyphenol fraction. However, in this study, the lack of an antioxidative protection in HFD + melanoidin rats was in contrast to the recent findings by Paur et al.,25 who demonstrated that coffee melanoidins induced EpRE activity in EpRE-luciferase mice. The different experimental design (acute versus chronic administration) and the different dosage of coffee melanoidins (50-fold higher in Paur et al. than in the present study) might account for the different results.

How Coffee Reduces Liver Inflammation.

We have demonstrated for the first time that in HFD-fed rats, coffee reduced both the expression and the concentration of liver TNF-α, which plays an important pathogenic role in NASH26 due to its ability to induce oxidative stress.

The effect of coffee in an animal model of insulin resistance or NASH was previously analyzed by gene expression in liver and adipose tissues using complementary DNA microarray.19, 27 Fukushima et al.19 found a down-regulation of IL-1b gene expression in the livers of HFD-fed mice given decaffeinated coffee (1.1% diet), whereas in our study the IL-1b concentration in rat livers was not reduced by coffee consumption, and only a slight effect of polyphenols and melanoidins was recorded (Fig. 5). However, a clear role of coffee melanoidins in reducing inflammation by a 63% inhibition of nuclear factor-κB activation was recently demonstrated in vivo in transgenic nuclear factor-κB/luciferase mice.25

The increase of expression of adipo-R2 in coffee-treated rats, as well as the higher liver concentrations of IL-4 and IL-10 in HFD-fed rats drinking coffee or its fractions compared with HFD-fed rats drinking water, account for the reduced liver inflammation shown using histological parameters. Adiponectin, which has both insulin- sensitizing28 and anti-inflammatory properties,29 can antagonize the effects of TNF-α on NAFLD development.

In this study, we demonstrated in a rat model of NASH that: (1) coffee consumption reduced the rate of fat and collagen deposition in the liver; (2) coffee consumption guaranteed a systemic and liver endogenous antioxidant protection, through glutathione system, mainly due to its polyphenol fraction; (3) consumption of coffee, but not its components, reduced glutathione transferase activity according to ameliorated whole liver status; (4) coffee and polyphenols were associated with an increase of serum-reducing activity and a decrease of lipoperoxydation assessed by malondialdehyde concentration; (5) coffee and its components modulated gene and protein expression of several mediators of inflammation, insulin sensitizers, and hepatic fat β-oxidation according to a reduction of liver inflammation and fat accumulation; and (6) coffee and its components, to different extents, decreased liver concentrations of pro-inflammatory and increased anti-inflammatory cytokines.

Considering the two-hit hypothesis of the pathogenesis of NAFLD and the results obtained in this study, the healthy role of coffee consumption on liver was schematized in Fig. 6. This figure summarizes the two primary findings of this study: (1) coffee may help retard liver damage progression caused by an HFD through reduction of fat accumulation, oxidative stress, and inflammation in the liver; and (2) the modulation of liver functions is triggered by gene expression and concentrations of some important mediators in tissue and/or in the bloodstream.

Figure 6.

Proposed role of coffee consumption within the scheme of the two-hit hypothesis of the pathogenesis of NAFLD. The data found in this study suggest that coffee may help retard liver damage progression mediated by an HFD through reduction of fat accumulation, oxidative stress, and inflammation in the liver. The modulation of liver functions and health may be played by influencing gene and protein expression as well as the concentrations of some important mediators in the tissue and/or in the bloodstream. The specific biomarkers implicated in the protective action of coffee against liver damage progression correspond to the insults to which they were related.