Introduction

Toxic liver injury may lead to acute liver failure, resulting in organ dysfunction. Numerous drugs and toxic substances could cause hepatic damage, with the severity of the changes proportional to the duration of the toxic exposure1. Carbon tetrachloride (CCl4) poisoning is one of the most commonly used models of acute liver damage. The hepatotoxic effects of CCl4 were attributed to the excessive production of free radicals2. Previous studies have shown that natural compounds with antioxidant activity could ameliorate CCl4-induced liver damage, thus preventing acute liver failure3,4.

Rutin is a naturally occurring flavonol consisting of aglycone quercetin and a rutinoside moiety in position 3 of the C ring (Figure 1). These widespread flavonoids, commonly found in various foods5, exert numerous biochemical and pharmacological activities, such as antioxidant6, anti-inflammatory7 and antitumor activities8. However, the pharmacological effects of rutin and its aglycone may differ, suggesting that the presence of the rutinoside moiety is crucial for some of the protective effects of rutin. In several studies, rutin exerted anti-inflammatory activity, whereas quercetin was either not effective against or actually aggravated the inflammatory response both in vivo and in vitro9,10. However, other authors demonstrated that quercetin acted as a strong inhibitor of inflammation in an experimental model of rat colitis11. Interestingly, rutin was an effective anti-inflammatory agent in chronic inflammatory conditions, such as adjuvant arthritis, whereas quercetin, but not rutin, potently suppressed acute inflammation and reduced carrageenan-induced paw edema7. Similarly, the cytochromes CYP1A1 and CYP1B1 were strongly inhibited by quercetin, while rutin exerted no inhibition or only weak inhibitory potential12. Thus, the pharmacological activities of rutin and quercetin may differ substantially, which could be attributed to the presence of a sugar moiety in position 3 of the C ring.

Figure 1
figure 1

The chemical structures of rutin and quercetin.

Previous investigations showed that rutin and quercetin could ameliorate chemically induced liver damage in rodents13,14. The objective of this study was to elucidate the molecular mechanisms of the hepatoprotective activity of rutin against acute toxic liver damage in mice and compare this activity to its aglycone quercetin.

Materials and methods

Chemicals and antibodies

Rutin, quercetin, olive oil, bovine superoxide dismutase (SOD), xanthine, xanthine oxidase, cytochrome c, ethylenediaminetetraacetic acid (EDTA), bovine serum albumin (BSA), bovine Cu/Zn superoxide dismutase (Cu/Zn SOD), glutathione (GSH), glutathione reductase (GR), reduced nicotinamide adenine dinucleotide phosphate (NADPH), metaphosphoric acid, 5,5'-dithiobis-(2-nitrobenzoic acid) (DTNB), dimethyl sulfoxide (DMSO), ammonium molybdate, 2,2-diphenyl-1-picryl-hydrazyl (DPPH·), hydrogen peroxide, potassium ferricyanide, Trolox, sodium acetate, Tris(hydroxymethyl)aminomethane (Tris), Tween 20 and Entelan were purchased from Sigma-Aldrich (Taufkirchen, Germany). Trichloroacetic acid was purchased from Acros Organics (Geel, Belgium). Iron (III) chloride was obtained from Riedel-de Haën (Seelze, Germany). The radioimmunoprecipitation assay (RIPA) buffer (sc-24948) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). CCl4, ethanol, sodium phosphate and sulfuric acid were purchased from Kemika (Zagreb, Croatia). The diagnostic kits for the determination of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were from Dijagnostika (Sisak, Croatia). Mouse monoclonal antibodies to tumor necrosis factor-alpha (TNF-α) (ab1793) and 3-nitrotyrosine (3-NT) (ab78163) and rabbit polyclonal antibodies to nuclear factor-kappa B (NF-κB) p65 (ab7970), cyclooxygenase (COX)-2 (ab15191), inducible nitric oxide synthase (iNOS) (ab3523), heme oxygenase (HO)-1 (ab13243), NF-E2-related factor 2 (Nrf2) (ab31163) and TGF-β1 (ab92486) were purchased from Abcam (Cambridge, UK). The DAKO EnVision+ System was from DAKO Corporation (Carpinteria, CA, USA).

Reducing power assay

The reducing power of the samples was determined by the method of Oyaizu15, as described previously16. Briefly, an aliquot of the sample (1.0 mL) at various concentrations (1.25–100 μg/mL) was mixed with phosphate buffer (0.2 mol/L, pH 6.6, 2.5 mL) and 1% potassium ferricyanide (2.5 mL). The mixture was incubated at 50 °C for 20 min. After the addition of 10% trichloroacetic acid (2.5 mL), the mixture was centrifuged at 1000×g for 10 min. The supernatant (2.5 mL) was mixed with distilled water (2.5 mL) and 0.1% iron (III) chloride (0.5 mL), and the absorbance was measured at 700 nm using an appropriate blank. All experiments were performed in triplicate. Trolox was used as a reference.

DPPH radical scavenging assay

The free radical scavenging activity of the samples was measured using the stable DPPH· radical according to the method of Blois17, as described previously18. Briefly, a 0.1 mmol/L solution of DPPH· in ethanol was prepared, and this solution (0.5 mL) was added to sample solution in ethanol (1.5 mL) at different concentrations (0.39–50 μg/mL). After the reaction was performed in the dark at room temperature for 30 min, the absorbance was measured at 517 nm. The capability to scavenge the DPPH· radical was calculated using the following equation: (%)=[(A0A1)/A0]×100, where A0 is the absorbance of the control reaction and A1 is the absorbance in the presence of the sample, corrected for the absorbance of the sample itself. All experiments were performed in triplicate. Trolox was used as a reference.

Total antioxidant capacity assay

The total antioxidant capacity of rutin and quercetin was evaluated by the phosphomolybdenum method according to the procedure of Prieto et al19, as described previously16. Briefly, the sample was dissolved in ethanol, and an aliquot of the solution (0.3 mL) was combined in a vial with the reagent solution (0.6 mol/L sulfuric acid, 28 mmol/L sodium phosphate and 4 mmol/L ammonium molybdate, 2.7 mL). The vials were capped and incubated in a water bath at 95 °C for 90 min. After the mixture was cooled to room temperature, the absorbance was measured at 695 nm against a blank. All experiments were performed in triplicate. The antioxidant capacity of the sample was expressed as Trolox equivalents, utilizing a calibration curve of Trolox in the concentration range from 0.78 to 100 μg/mL.

Nitric oxide radical scavenging assay

The nitric oxide (NO·) scavenging activity of the samples was determined according to the method described by Rai et al20 with a slight modification. The NO· generated from sodium nitroprusside in an aqueous solution at physiological pH interacts with oxygen to produce nitrite ions, which were measured by the Griess reaction. Equal volumes of 10 mmol/L sodium nitroprusside in phosphate buffered-saline (pH 7.4) were mixed with different concentrations of the sample (0.39–50 μg/mL) and incubated at 25 °C for 150 min. After the incubation, 1.0 mL of the reaction mixture was mixed with 1% sulfanilamide (0.5 mL). After 5 min, 0.1% naphthylethylenediamine dihydrochloride (0.5 mL) was added, the solution was mixed, and the absorbance of a pink-colored chromophore was measured at 540 nm against the corresponding blank solution. Trolox was used as a standard. All experiments were performed in triplicate. The NO· scavenging activity was expressed as the percentage of inhibition according to the following equation: (%)=[(A0A1)/A0]×100, where A0 is the absorbance of the control without a sample and A1 is the absorbance in the presence of the sample.

Animals

Male BALB/cN mice, 10–12 weeks old and weighing 23–25 g, were obtained from our breeding colony. The animals were housed under standard environmental conditions and had free access to tap water and a standard rodent diet (pellet, type 4RF21 GLP, Mucedola, Italy). All experimental procedures were approved by the Ethical Committee of the Medical Faculty, University of Rijeka.

Experimental design

The mice were divided into six groups, each containing five animals. The normal control (group I) received saline, and group II received CCl4 dissolved in olive oil (2 mL/kg, 10% v/v) intraperitoneally (ip). Rutin or quercetin, dissolved in 5% (v/v) DMSO, was administered ip at 10, 50, and 150 mg/kg (groups III, IV, and V, respectively) and 50 mg/kg (group VI), respectively, once daily for five consecutive days. Immediately after the last dose, the mice were given CCl4. The doses of rutin were selected on the basis of our preliminary studies (data not shown), whereas the middle dose of quercetin (50 mg/kg) was used for the comparison with rutin. The mice were sacrificed 24 h after the injection of CCl4. Blood was collected by cardiac puncture, and heparinized plasma was separated for the determination of the ALT and AST activities. The gall bladder was removed, and the liver was carefully excised, washed with saline, blotted dry and divided into samples. The tissue specimens were snap frozen in liquid nitrogen and stored at −80 °C if not used on the same day. The liver samples were used to assess the Cu/Zn SOD activity, GSH and protein concentration and also for immunoblotting. Portions of the livers were immersed in 4% paraformaldehyde for histology and immunohistochemistry.

Determination of hepatotoxicity

The activity of transaminases (ALT and AST) in plasma was measured using a Bio-Tek EL808 Ultra Microplate Reader (BioTek Instruments, Winooski, VT, USA) according to the manufacturer's instructions.

Measurement of oxidative stress

Mouse livers were homogenized in 50 mmol/L phosphate buffer saline (PBS), pH 7.4, using a Polytron homogenizer (Kinematica, Lucerne, Switzerland). The supernatants were separated by centrifugation at 15 000×g for 20 min at 4 °C (Beckman L7-65 Ultracentrifuge, Beckman, Fullerton, CA, USA) and used to measure the Cu/Zn SOD activity and GSH content. The Cu/Zn SOD activity was determined as described previously3. The GSH content was evaluated according to Anderson21, with modifications. Briefly, the supernatants were deproteinized with 1.25 mol/L metaphosphoric acid and centrifuged at 5000×g for 10 min at room temperature (Rotina 420R, Andreas Hettich GmbH, Tuttlingen, Germany). Then, 25 μL of the deproteinized sample was mixed in a cuvette with 700 μL of 0.3 mmol/L NADPH in PBS, 100 μL of 6 mmol/L DTNB and water to give a final volume of 1.0 mL. The reaction was started by the addition of 10 μL of GR (50 units/mL), and the absorbance was monitored at 405 nm for 25 min. The GSH concentration in the samples was determined using the standard curve generated with different GSH solutions under the same condition. The protein content in the liver homogenates was estimated by the Bradford method22.

Histopathology

Paraformaldehyde-fixed tissues were processed routinely, embedded in paraffin, sectioned, deparaffinized and rehydrated using standard techniques23. Hepatocellular necrosis was evaluated by measuring the size of the necrotic area in hematoxylin and eosin (H&E) stained liver sections. The necrotic areas were manually selected, and their size was determined using Cell F v3.1 software (Olympus Soft Imaging Solutions, Münster, Germany).

Immunohistochemistry

For immunohistochemistry, 4-μm thick liver tissue sections were used, as described elsewhere3. Briefly, the slides were deparaffinized using xylene and washed with ethanol. The liver slices were incubated overnight with antibodies against TNF-α (diluted 1:50), NF-κB (diluted 1:1000) and iNOS, 3-NT, Nrf2, and TGF-β1 (diluted 1:100). The detection of antibodies was performed using DAKO EnVision+ System, Peroxidase/DAB kit (DAKO Corporation, Carpinteria, CA, USA). Later, the slides were counterstained with hematoxylin, dehydrated using graded ethanol and xylene and mounted with Entelan. The expression profile and cellular localization of NF-κB, iNOS, Nrf2, and 3-NT were analyzed by light microscopy (Olympus BX51, Tokyo, Japan). The relative staining intensity was analyzed with ImageJ software (National Institutes of Health, Bethesda, MD, USA)23.

Immunoblotting

The liver samples were lysed in radioimmunoprecipitation assay (RIPA) buffer containing 50 mmol/L Tris-HCl, pH 7.4, 150 mmol/L NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 2 mmol/L PMSF, 1 mmol/L sodium orthovanadate and 2 μg/mL each of aprotinin, leupeptin and pepstatin. Volumes equivalent to 50 μg of proteins were loaded onto a 12% polyacrylamide gel. After electrophoresis, the gels were blotted onto a polyvinylidene fluoride membrane (Roche Diagnostics GmbH, Mannheim, Germany). After the protein transfer, the membranes were blocked overnight with a milk blocking reagent (Santa Cruz Biotechnology, Santa Cruz, CA, USA) at 4 °C. TNF-α, COX-2, and HO-1 were visualized by the addition of their respective antibodies, followed by peroxidase-labeled goat anti-mouse or anti-rat antibodies (Amersham Pharmacia Biotech, Uppsala, Sweden). β-Actin was used as a control for protein loading. The membranes were washed with Tris-buffered saline containing 0.1% Tween 20 (TBST), pH 7.6, incubated with Amersham ECL Prime (GE Healthcare, Uppsala, Sweden) for 5 min and exposed to X-ray film (Ortho CP-G Plus film, Agfa-Gevaert NV, Belgium) for 1 min.

Statistical analysis

The data were analyzed using StatSoft STATISTICA version 7.1 and Microsoft Excel 2000 software. Differences between the groups were assessed by a one-way ANOVA and Dunnett's post hoc test. The values in the text are mean±standard deviation (SD). For the in vitro studies, the concentration of samples that provide 50% inhibition (IC50) was obtained by interpolation from a linear regression analysis. Differences from P<0.05 were considered statistically significant.

Results

In vitro antioxidant activity

The ability of the tested samples to reduce iron (III) and scavenge DPPH· and NO· was assessed on the basis of their IC50 values. Figure 2A shows the plot of the reducing power of rutin and quercetin compared with Trolox as a reference antioxidant. Quercetin exhibited the most powerful effect (IC50=2.42 μg/mL) compared with rutin (IC50=7.18 μg/mL) and Trolox (IC50=6.66 μg/mL) (P<0.05). Quercetin also showed the most effective DPPH· and NO· radical scavenging activity (IC50=0.56 μg/mL and 4.04±1.03 μg/mL, respectively), which was significantly stronger (P<0.05) than that of both rutin (IC50=2.07 μg/mL and 9.32±1.24 μg/mL, respectively) and Trolox (IC50=1.11 μg/mL and 8.51±1.03 μg/mL, respectively) (Figure 2B and 2C). Similarly, the total antioxidant capacity of quercetin (Figure 2D) was higher than that of rutin (P<0.05).

Figure 2
figure 2

Evaluation of the in vitro antioxidant properties of rutin and quercetin. (A) Reducing power, (B) DPPH· radical scavenging effect, (C) NO· radical scavenging effect and (D) total antioxidant activity of rutin and quercetin. Each value represents the mean±SD from three independent measurements.

Liver weight and plasma activity of transaminases

The relative liver weight in CCl4-intoxicated mice decreased compared with that in the controls. Rutin dose-dependently prevented liver weight loss significantly more than quercetin at the equivalent dose. The plasma AST and ALT activities significantly increased 24 h after CCl4-intoxication (P<0.05). Treatment with rutin decreased the activity of these transaminases in a dose-dependent manner (Table 1). However, quercetin showed a less protective effect against hepatocellular damage than rutin at the equivalent dose.

Table 1 Effect of CCl4, rutin and quercetin on relative liver weight, plasma transaminase activity and oxidative stress markers in mice liver.

Effect of rutin and quercetin on hepatic oxidative stress

Our results showed that CCl4 administration induced oxidative stress in mouse livers. The Cu/Zn SOD activity and GSH concentration were significantly lower compared with the control group (Table 1) (P<0.05). Treatment with rutin elevated the Cu/Zn SOD activity and GSH concentration in a dose-dependent manner. Quercetin at 50 mg/kg significantly attenuated the decrease in the oxidative stress markers in CCl4-intoxicated mice, restoring Cu/Zn SOD activity more potently than rutin at 50 mg/kg (P<0.05).

Amelioration of histopathological changes by rutin and quercetin in the liver

Representative images of histological sections from the experimental groups are shown in Figure 3. The livers of the control mice showed a normal morphology and architecture (Figure 3A). In the CCl4-intoxicated mice, severe hepatic damage with a massive centrilobular necrosis was detected (Figure 3B). In the livers of mice treated with rutin at 10 mg/kg, no histological improvement was found (Figure 3C). Rutin at 50 mg/kg markedly reduced the size of the hepatic necrotic areas (Figure 3D), whereas the high-dose rutin (150 mg/kg) almost completely prevented hepatocellular damage (Figure 3E). In mice treated with quercetin at 50 mg/kg, larger necrotic areas were present than in the livers of mice treated with rutin at the equivalent dose (Figure 3F and 3G).

Figure 3
figure 3

Representative hematoxylin and eosin stained sections of livers from experimental mice. (A) Control group. (B) CCl4-intoxicated mice. (C–F) Mice treated with CCl4 and protected with rutin (R) at 10 mg/kg, 50 mg/kg and 150 mg/kg and with quercetin (Q) at 50 mg/kg. Original magnification ×100, insets ×400. (G) Measurement of the size of the necrotic areas (n=5). bP<0.05, CCl4-intoxication vs control. eP<0.05, rutin and quercetin treatment vs CCl4-intoxication. hP<0.05, 50 mg/kg rutin vs 50 mg/kg quercetin.

Amelioration of hepatic inflammation by rutin and quercetin

To determine whether rutin or quercetin could reverse the acute liver inflammation induced by CCl4, we analyzed proteins involved in the proinflammatory response, such as NF-κB, TNF-α, and COX-2. The livers of control mice were NF-κB immunonegative (Figure 4A). In contrast, strong NF-κB immunoreactivity was detected in the CCl4-treated mice (Figure 4B). Immunohistochemical analysis revealed NF-κB p65 localization in both the cytoplasm and the nucleus of hepatocytes and in Kupffer cells. The low-dose rutin treatment (10 mg/kg) did not substantially affect NF-κB immunopositivity (Figure 4C). However, the higher doses of rutin, 50 and 150 mg/kg, gradually decreased NF-κB expression and prevented the accumulation of NF-κB p65 in the nuclei (Figure 4D and 4E). Quercetin had a more pronounced effect on NF-κB suppression than rutin at the equivalent dose (Figure 4F and 4G). CCl4 administration also increased the hepatic levels of TNF-α and COX-2 (Figure 5). Rutin dose-dependently reduced both TNF-α and COX-2 expression compared with the CCl4-treated mice. Similarly to NF-κB, quercetin more effectively suppressed TNF-α and COX-2 expression than rutin at the equivalent dose.

Figure 4
figure 4

NF-κB immunostaining in liver sections. (A) Control group. (B) CCl4-intoxicated mice. (C–F) Mice treated with CCl4 and protected with rutin (R) at 10, 50, and 150 mg/kg and with quercetin (Q) at 50 mg/kg. The arrows show NF-κB p65-positive hepatocyte nuclei, and the arrowheads show NF-κB p65-positive Kupffer cells. Original magnification ×100, insets ×400. Representative results from five mice. (G) Measurement of NF-κB immunostaining intensity (n=5). bP<0.05, CCl4-intoxication vs control; eP<0.05, rutin and quercetin treatment vs CCl4-intoxication; hP<0.05, 50 mg/kg rutin vs 50 mg/kg quercetin.

Figure 5
figure 5

Western blot analysis of TNF-α, COX-2, and HO-1 in mouse livers. Representative results from five mice. Quantification of band intensity relative to β-actin intensity is shown below the Western blot. bP<0.05, CCl4-intoxication vs control. eP<0.05, rutin and quercetin treatment vs CCl4-intoxication. hP<0.05, 50 mg/kg rutin vs 50 mg/kg quercetin. R, rutin; Q, quercetin.

Effect of rutin and quercetin on hepatic nitrosative stress

The analysis of the hepatic expression of iNOS and the formation of the NO·-dependent product, 3-NT, revealed no immunopositivity in the control mice (Figure 6A and 7A). CCl4 administration induced a strong expression of iNOS and 3-NT (Figure 6B and 7B). Rutin dose-dependently reduced the iNOS (Figure 6C, 6D, and 6E) and 3-NT (Figure 7C, 7D, and 7E) immunopositivity compared with the CCl4-treated mice. However, quercetin was less effective than rutin in the reduction of both iNOS (Figure 6F and 6G) and 3-NT (Figure 7F and 7G) immunoreactivity.

Figure 6
figure 6

Immunohistochemical analysis of iNOS expression in mouse livers. (A) Control group. (B) CCl4-intoxicated mice. (C–F) Mice treated with CCl4 and protected with rutin (R) at 10, 50, and 150 mg/kg and with quercetin (Q) at 50 mg/kg. The arrows show iNOS-positive hepatocyte nuclei, and the arrowheads show iNOS-positive Kupffer cells. Original magnification ×100, insets ×400. Representative results from five mice. (G) Measurement of iNOS immunostaining intensity (n=5). bP<0.05, CCl4-intoxication vs control. eP<0.05, rutin and quercetin treatment vs CCl4-intoxication. hP<0.05, 50 mg/kg rutin vs 50 mg/kg quercetin.

Figure 7
figure 7

Immunohistochemical detection of 3-NT production in mouse livers. (A) Control group. (B) CCl4-intoxicated mice. (C–F) Mice treated with CCl4 and protected with rutin (R) at 10, 50, and 150 mg/kg and with quercetin (Q) at 50 mg/kg. Original magnification ×200. Representative results from 5 mice. (G) Measurement of 3-NT immunostaining intensity (n=5). bP<0.05, CCl4-intoxication vs control. eP<0.05, rutin and quercetin treatment vs CCl4-intoxication. hP<0.05, 50 mg/kg rutin vs 50 mg/kg quercetin.

Induction of the Nrf2/HO-1 pathway in the liver by rutin and quercetin

The hepatic expression of HO-1 (Figure 5), the cytoprotective enzyme that plays a critical role in the response to stressful conditions, and Nrf2, its upstream inducer24, is shown in Figure 8. The livers of control mice were Nrf2 immunonegative (Figure 8A). The livers of CCl4-intoxicated mice were weakly Nrf2 immunopositive (Figure 8B). However, the treatment with rutin resulted in a marked dose-dependent induction of Nrf2 (Figure 8C, 8D, and 8E) in the cytoplasm and nucleus of hepatocytes and Kupffer cells. Nrf2 immunoreactivity in hepatocytes of mice treated with quercetin (Figure 8F) markedly increased compared with the CCl4 group but was less expressed compared with mice receiving rutin at the equivalent dose (Figure 8G). Similarly, rutin dose-dependently increased the hepatic HO-1 levels compared with CCl4-treated mice, whereas quercetin was a less potent up-regulator of HO-1 than rutin at the equivalent dose (Figure 5).

Figure 8
figure 8

Immunohistochemical detection of Nrf2 expression in mouse livers. (A) Control group. (B) CCl4-intoxicated mice. (C–F) Mice treated with CCl4 and protected with rutin (R) at 10, 50, and 150 mg/kg and with quercetin (Q) at 50 mg/kg. The arrows show Nrf2-positive hepatocyte nuclei, and the arrowheads show Nrf2-positive Kupffer cells. Original magnification ×100, insets ×400. Representative results from five mice. (G) Measurement of Nrf2 immunostaining intensity (n=5). bP<0.05, CCl4-intoxication vs control. eP<0.05, rutin and quercetin treatment vs CCl4-intoxication. hP<0.05, 50 mg/kg rutin vs 50 mg/kg quercetin.

Antifibrotic potential of rutin and quercetin in the liver

The livers of control mice did not show substantial TGF-β1 immunopositivity (Figure 9A). In contrast, TGF-β1 was overexpressed in the livers of CCl4-intoxicated mice (Figure 9B). The administration of rutin dose-dependently ameliorated TGF-β1 expression (Figure 9C, 9D, and 9E). However, the hepatic TGF-β1 immunoreactivity in mice treated with quercetin (Figure 9F) was significantly lower than that in mice receiving rutin at the equivalent dose (Figure 9G) (P<0.05).

Figure 9
figure 9

Immunohistochemistry of TGF-β1 expression in mouse livers. (A) Control group. (B) CCl4-intoxicated mice. (C–F) Mice treated with CCl4 and protected with rutin (R) at 10, 50, and 150 mg/kg and with quercetin (Q) at 50 mg/kg. The arrows show TGF-β1-positive hepatocyte nuclei, and the arrowheads show TGF-β1-positive Kupffer cells. Original magnification ×100, insets ×400. Representative results from five mice. (G) Measurement of TGF-β1 immunostaining intensity (n=5). bP<0.05, CCl4-intoxication vs control. eP<0.05, rutin and quercetin treatment vs CCl4-intoxication. hP<0.05, 50 mg/kg rutin vs 50 mg/kg quercetin.

Discussion

The results of the present study suggest the importance of the sugar moiety in position 3 of the C ring for the specific pharmacological activities of rutin and quercetin. The glycosylation of flavonoids reduces their antioxidant activity compared with their corresponding aglycones25. Our results showed that quercetin possessed higher reducing power and DPPH· and NO· free radical scavenging abilities in vitro compared with rutin. Quercetin more potently prevented the decrease in Cu/Zn SOD activity in mouse livers. However, it was less effective in the amelioration of protein nitrosylation, suggesting that the in vitro free radical scavenging ability of compounds does not necessary correlate with their in vivo efficacy, which indicates that absorption and/or metabolism are key modulators of in vivo activity26. Additionally, the hepatic lesions and necrosis induced by CCl4 were markedly reduced in mice treated with rutin, whereas quercetin was less effective. The results of the current study showed that the amelioration of hepatic necrosis was closely related to the reduction of nitrosative but not oxidative stress.

Reactive oxygen species (ROS) generated during CCl4 intoxication may induce NF-κB activation and consequently stimulate the production of cytotoxic and proinflammatory cytokines, such as TNF-α27. Our results showed that rutin, and more potently quercetin, suppressed both NF-κB and TNF-α expression in injured livers. NF-κB is also involved in the regulation of COX-2 and iNOS gene expression by binding to their promoter regions28. COX-2 and iNOS exert a prominent role under inflammatory conditions by producing prostaglandins and NO·, respectively29,30. Thus, the anti-inflammatory effect of quercetin and rutin in CCl4-injured liver could be attributed to the inhibition of the NF-κB pathway, which is in agreement with previous findings31. Additionally, the increased NO· synthesis and superoxide generation could result in the formation of peroxynitrite32 and nitration of protein tyrosine residues33, which actively contribute to the development of hepatic necrosis34. In this study, the activation and nuclear accumulation of NF-κB p65 in CCl4-intoxicated mice coincided with COX-2 and iNOS overexpression, which was prevented by both flavonoids. The nuclear presence of COX-2 and iNOS suggests their involvement in the regulation of some nuclear functions. However, rutin exhibited lower potency in reducing COX-2 expression compared with quercetin, which is in agreement with previous findings35. Interestingly, rutin more effectively suppressed the expression of both iNOS and 3-NT compared with quercetin. The latter coincided with the amelioration of hepatic necrosis. These findings agree with Raghav et al36, who showed a stronger inhibition of iNOS than COX-2 gene expression by rutin in murine macrophages.

TGF-β, a pleiotrophic growth factor, has been shown to stimulate fibroblast proliferation and increase the synthesis of ECM proteins37. Previously, we showed that the fibrogenic potential of the CCl4-injured liver is closely related to TGF-β1 expression38. In the current study, TGF-β1 expression markedly increased in both hepatocytes and non-parenchymal cells of the CCl4-intoxicated liver. Although Kupffer cells and HSCs are considered the main sources of TGF-β39, hepatocytes may also produce TGF-β. In cultured hepatocytes, latent TGF-β was rapidly detectable during culture due to demasking of the mature TGF-β by calpains40. The nuclear presence of TGF-β1 could be attributed to increased nuclear membrane permeability during the activation of apoptosis41. Our results showed that treatment with both flavonoids markedly ameliorated the overexpression of TGF-β1, suggesting the reduction of the fibrogenic potential in the liver.

HO-1, the inducible isoform of heme oxygenase, plays a critical role in cell protection against acute and chronic liver injury42. The transcription factor Nrf2 is considered a key regulator of HO-1 expression24. In the current study, hepatic injury was associated with low expression of HO-1 and its upstream inducer Nrf2 but with high expression of NF-κB. The treatments with rutin and, to a lesser extent, quercetin were associated with increased Nrf2 and HO-1 expression in the liver, which coincided with the suppression of NF-κB activation. Most recently, Liu et al43 showed that quercetin acts as an inducer of HO-1 in primary rat hepatocytes. Interestingly, the induction of HO-1 was found in quercetin-treated but not in rutin-treated rat glioma C6 cells44 and in the hydrogen peroxide-induced apoptosis of RAW264.7 macrophages45. However, rutin acted as a strong HO-1 inducer in a rat model of liver ischemia-reperfusion injury46, suggesting a cell type-dependent activation of this enzyme. Additionally, the nuclear translocation of HO-1 could be involved in the regulation of genes responsible for cytoprotection against oxidative stress47.

In conclusion, the results of this study demonstrate that rutin and quercetin can ameliorate acute liver damage by at least four mechanisms: acting as scavengers of free radicals, inhibiting NF-κB activation and the inflammatory response, exerting antifibrotic potential and inducing the Nrf2/HO-1 pathway. The rutinoside moiety in position 3 of the C ring could be responsible for more pronounced protective effects against iNOS induction, nitrosative stress and hepatocellular necrosis. The aglycone quercetin exerted higher antioxidant and anti-inflammatory activities and antifibrotic potential than rutin. The antioxidant actions of rutin and quercetin were partially responsible for their beneficial effects in injured liver tissue. However, the antioxidant properties of these flavonoids cannot solely explain the stronger protective activity of rutin against hepatocellular damage. Thus, the modulation of signaling pathways emerges as an important mode of action of these flavonoids, which should be considered as a specific therapeutic strategy. The application of these flavonoids in medical practice should be further confirmed by conveying preliminary placebo-controlled clinical studies.

Author contribution

Robert DOMITROVIĆ designed the research and wrote the paper; Hrvoje JAKOVAC, Vanja VASILJEV MARCHESI, Sanda VLADIMIR-KNEŽEVIĆ, Olga CVIJANOVIĆ, and Žarko TADIĆ performed the research; Željko ROMIĆ contributed new reagents and analytic tools; and Dario RAHELIĆ analyzed the data.