Vitis vinifera polyphenols from seedless black fruit act synergistically to suppress hepatotoxicity by targeting necroptosis and pro-fibrotic mediators

Human is subjected from his surrounding to various hepatotoxins, which aggravates his liver. Nowadays, natural polyphenols have attracted great interest in health improvement, especially liver health. The present research, therefore, assessed the hepatotherapeutic potency of the isolated polyphenols (VVF1) from seedless (pulp and skin) black Vitis vinifera (VV) against CCl4-induced hepatotoxicity in vitro and in vivo. Further, VVF1 was fractionated into resveratrol-enriched (VVF2) and phenolics-enriched (VVF3) fractions to study (in vitro) the possible synergism of their coexistence. The highest content of phenolics in VVF1 displayed in vitro synergistic antioxidant and anti-hepatotoxic activities comparing to VVF2, VVF3, and silymarin (SM, reference drug). More importantly, it exhibited multiple in vivo regulatory functions via diminishing oxidative stress and inflammation, which in turn decreased necroptosis and pro-fibrotic mediators (mixed lineage kinase domain-like protein (MLKL), collagen type I alpha 1 chain (COL1A1), and transforming growth factor (TGF)-β1). In addition to these novel findings, VVF1 had higher anti-hepatotoxic potency than that of SM in most of the studied parameters. The histopathological analysis confirmed the improving role of VVF1 in the serious hepatic damage induced by CCl4. Thus, the synergistic functions of VVF1 polyphenols could be a promising new anti-hepatotoxic agent for targeting both necroptotic and profibrotic mediators.

Antioxidant activities of VVfs. Figure 1A illustrates that VVF1 had the highest total antioxidant capacity (3624 mg ascorbic acid "Asc" eq/g extract) in comparison with VVCE, other VVFs, and silymarin (SM; a standard drug for hepatotoxicity). Based on low IC 50 indicating high antioxidant activity, VVF1 revealed the strongest radical scavenger for DDPH, hydroxyl and superoxide anion radicals at the lowest IC 50 (3.6, 35.78 and 30.07 µg/mL, respectively) compared to other VVFs. The radical scavenging activity of VVF1 was not statistically significant with that of VVCE. Also, no significant difference was recorded between VVF1 and Asc (antioxidant marker) for hydroxyl and superoxide anion radical scavenging activity. Regarding DPPH scavenging potential, VVF1 was significantly (p < 0.0005) higher than Asc (Fig. 1B,C). Moreover, VVF1 exhibited the highest anti-lipid peroxidation  Table 1. The yield and phenolic content of Vitis vinifera crude extract (VVCE) and its fractions (F), their safe concentrations (EC 100 ) on the normal hepatocytes with their effective doses (ED) against the CCl 4 -induced in vitro hepatotoxicity as well as the combination index (CI) values of VVF1. Results are presented as mean ± SE (n = 3). VVF1 was compared with VVCE, VVF2, VVF3 and SM and considered significantly different at *p < 0.05, **p < 0.005, ***p < 0.0005. 4-HCA, 4-hydroxycinnamic acid; RU, Rutin; SM, Silymarin; ED 50 and ED 100, 50% and 100% therapeutic response against the CCl 4 -induced in vitro hepatotoxicity, respectively; DPPH, 2,2-diphenyl-1-picrylhydrazyl.
www.nature.com/scientificreports www.nature.com/scientificreports/ activity by preventing linoleate oxidation then bleaching β-carotene at the lowest IC 50 (0.005 mg/mL). This value was not statistically significant with BHT (standard antioxidant) and VVF2 as shown in Fig. 1D.
The synergism between VVF2 and VVF3 for the studied antioxidant activities was declared by different methods, the CI values at different inhibition levels (50%, 75%, 90%, Table 1), Fa-CI plot (Fig. 1E), as well as isobologram plot (Supplementary Fig. 2A-D). The values of CI were lower than 1 at the studied inhibition levels for all the tested antioxidant assays. Isobolograms were performed, in particular at 50%, 75%, and 90% inhibition levels of Fa. Hence the single-agent Fa value corresponds in our study to the IC 50 value, the isobologram at 50% inhibition for the combination provided a direct comparison with single-agent treatment and synergy refers to a lowering of IC 50 equivalent (left-shift). While 75% and 90% isobologram refers to the combination at a high effect level. The findings of both the Fa-CI and the isobologram plots showed that the combination of VVF2 and VVF3 had strongly synergistic anti-DDPH activity and that the least amount of synergism was observed with the hydroxyl radical. This antioxidant synergism between water-soluble polyphenolic and ethanol-soluble polyphenolic constituents of VVF1 resulted in its powerful antioxidant efficacy (Fig. 1A-E).
The antihepatotoxicity potency of VVF1 among VV fractions (in vitro). Table 1 illustrates that the safe doses (EC 100 ) of VVCE and VVFs were 2 to 3 mg/mL on rat hepatocytes. These doses were used to investigate the therapeutic effect of VVFs on CCl 4 -induced hepatotoxicity in comparison with the standard drug (SM). The treatment of the CCl 4 -exposed hepatocytes with VVCE and its fractions showed that VVF1 diminished Figure 1. Antioxidant content and efficacy of VVF1 with a combination index (CI) plot of its constituents for radical scavenging and anti-lipid peroxidation activities. (A) Total antioxidant capacity (TAC), (B) DPPH scavenging activity, (C) hydroxyl and superoxide anion radical scavenging activities as well as (D) anti-lipid peroxidation activity of VV crude extract (VVCE), VVF1 and its fractions (VVF2 and VVF3) in comparison with silymarin (SM), Asc and butylated hydroxytoluene (BHT). (E) CI graph of antioxidant synergism of VVF2 and VVF3 (VVF1's constituents) including, DPPH, hydroxyl, and superoxide radical scavenging activities and inhibitory activity of lipid peroxidation. VVF1 was compared with VVCE, VVF2, VVF3, SM, Asc, and BHT and considered significantly different at *p < 0.05, **p < 0.005, ***p < 0.0005. hepatotoxicity by 50% at the lowest dose (ED 50 = 0.161 mg/mL) compared to ≥0.4 mg/mL in case of VVCE, other tested VVFs, and SM. Further, VVF1 had the lowest estimated therapeutic dose (ED 100 = 0.926 mg/mL) for complete inhibition of hepatotoxicity while it was needed above 1.7 mg/mL of VVCE, other tested VVFs or SM for reaching the same effect using MTT assay (Table 1). Additionally, the treatment of CCl 4 -exposed hepatocytes with VVF1 was able to maintain normal morphology of hepatocytes which was near to that of healthy untreated control cells ( Fig. 2A). Meanwhile, CCl 4 -exposed hepatocytes had severe damage in their spindle shape (cell rounding with cytoplasmic swelling). This indicated the incidence of necrosis that was also assured by the red fluorescence of their nuclei in contrast to green fluorescence nuclei of healthy control cells after incubation with dual nuclear staining of acridine orange and ethidium bromide (Fig. 2B). The 72 h treatment with VVF1 or VVCE did not show any red swollen nuclei, referring to halt necroptosis. Whereas the treatment of CCl 4 -exposed hepatocytes with other VVFs or SM still had few reddish or yellowish-orange nuclei of late or early necroptotic cells, respectively (Fig. 2B).
For a more accurate estimation of the anti-necroptotic effect of VVF1, the treated hepatocytes were stained with annexin-propidium iodide (PI) then analyzed using flow cytometry for detecting the percentage of double-positive annexin-PI stained cell populations. Figure 2C(I,II) shows that 39.64% of CCl 4 -exposed hepatocytes were necroptotic population. Hence, the induction of necroptosis in CCl 4 -exposed hepatocytes was evidenced by an increase in cell size ( Fig. 2A,B) and an abnormally high percentage of double-positive annexin-PI stained cells. Also, Fig. 2C(I,II) illustrates that the lowest percentage (p < 0.0005) of necroptotic populations (2.68 ± 0.06%) were recorded in VVF1-treated CCl 4 -exposed hepatocytes in comparison with 9.35-20.99% in VVCE, VVF2, VVF3, and SM. The high necroptotic percentage in CCl 4 -exposed hepatocytes was clarified by The phase-contrast microscopic images of necrotic hepatocytes exposed to CCl 4 without any following treatment in comparison with the SM-, VVCE-and its fraction-treated hepatocytes which previously incubated with CCl 4 . (B) Fluorescence images of acridine orange and ethidium bromide nuclear staining of CCl 4 -exposed hepatocytes before and after treatment with VVCE and its fractions as well as SM. Green, yellow and orangered fluorescences refer to healthy alive, early necroptotic and late necroptotic cells. (CI,II) Annexin V/PI flow charts with quantification histogram of the percentages of double-positive Annexin V/PI-stained necroptotic and (DI,II) flow cytometric charts of fluorescence oxidized form of DCF diacetate (indicator of reactive oxygen species "ROS") with quantitative histogram for the percentage of intracellular ROS for the untreated and treated necrotic hepatocytes. C, control untreated healthy hepatocytes; CCl 4 , hepatocytes were exposed to CCl 4 without any treatment; CCl 4 -SM, CCl 4 -VVCE, CCl 4 -VVF1, CCl 4 -VVF2, and CCl 4 -VVF3 hepatocytes were exposed for 36 h to CCl 4 then treated for 72 h with SM, VVCE, VVF1, VVF2 and VVF3, respectively. Data are presented as Mean ± SE. CCl 4 -VVF1 was compared with C, CCl 4 , V, and CCl 4 -SM and considered significantly different at *p < 0.05, **p < 0.005, ***p < 0.0005. elevating the percentage of the dichlorofluorescein (DCF) fluorescence from 4.519 ± 0.174% of healthy control cells to be 58.55 ± 2.28% after 72 h of CCl 4 exposure (Fig. 2DI,II). VVF1 exhibited the highest efficiency to reduce this percentage that reflected the generation of reactive oxygen species (ROS) to about 5% in comparison with VVCE, VVF2, VVF3, and SM (11.32, 19.065, 10.915, and 11.91%, respectively).
Moreover, the estimated CI of VVF1 for blocking the induction of necroptosis and generation of the fluorescent DCF was <1 (0.166 ± 0.005 and 0.479 ± 0.02, respectively) confirming the high synergism between VVF2 and VVF3 constituents of VVF1.
The therapeutic effect of VVF1 on CCl 4 -induced hepatotoxicity (in vivo). Based on the above-mentioned in vitro results, VVF1 was selected to investigate its anti-hepatotoxicity efficacy using an animal model. The induction of hepatotoxicity using CCl 4 was followed by treatment with VVF1 (CCl 4 -VVF1 group) comparing with the standard drug (CCl 4 -SM group) as elucidated in Fig. 3.
Returning the hepatic redox stress balance and suppression of necroptotic and fibrotic driving forces by VVF1. Oxidative stress and inflammation are cross-linked driving forces of necroptosis-dependent hepatotoxicity. Figure 4A shows CCl 4 induced hepatic oxidative stress that was clarified by the significant elevation of ROS (124.18 ± 0.628 mM H 2 O 2 eq/g tissue) and NO (236.54 ± 10.79 nmol/g tissue) leading to an excess generation of lipid peroxide products (1473.4 ± 12.50 nmol/g). This is associated with 2 folds enhancement of myeloperoxidase (MPO) activity as well as suppression of enzymatic antioxidant activities (superoxide dismutase "SOD" and glutathione peroxidase "GPX") and the GSH level by 1.7, 5.3 and 2.3 folds, respectively compared to the control (C) (Fig. 4B). Thus, hepatic total antioxidant capacity (TAC) was significantly lowered in the CCl 4 -injected rat (CCl 4 group) by 3 folds than C group. Meanwhile, the treatment of CCl 4 -injected rat group with VVF1 (CCl 4 -VVF1 group) diminished the prooxidant parameters (p < 0.0005), including ROS, NO, TBARS and MPO activity by 54.91%, 66.58%, 85.29%, and 37.77%, respectively, compared to the CCl 4 group (Fig. 4A). Furthermore, VVF1 was able to enhance the hepatic antioxidant system (SOD, GPX, and GSH) by ≥2 folds that also corroborated by 2.6 folds increase in the TAC level relative to the CCl 4 group (Fig. 4B). VVF1 was significantly exhibited a higher potency to halt ROS, NO, and TBARS production and to ameliorate SOD and GPX as well as TAC than standard drug (SM). However, there was no significant difference was observed between the effect of VVF1 and SM on MPO and GSH. On the other hand, injection with a CCl 4 vehicle (V group) caused non-significant alteration of the oxidative stress parameters compared to the control rats. Regarding the change in liver weight relative to body weight (b.w.), there was no significant difference was recorded between all the studied groups, data not shown. Experimental design with animal group classification. Induction of hepatotoxicity was done by injection CCl 4 group with CCl 4 twice (Sunday "S" and Wednesday "W")/week for 3 weeks. After induction of hepatotoxicity, CCl 4 -VVF1 and CCl 4 -SM rat groups were treated with VVF1 and SM, respectively for 10 days. Two healthy rat groups (VVF1 and SM) were orally injected with VVF1 and SM daily for 10 days in comparison with control untreated healthy group (C) and another group (V) were injected with the vehicle of CCl 4 (olive oil) twice/week for 3 weeks. Figure 5A illustrates that CCl 4 significantly upregulated the necroptotic protein MLKL by 92.6 folds comparing to C group while this elevation was repressed to 4.85 and 11.87 folds after the treatment with VVF1 and SM, respectively. This result suggested that VVF1 has shown a higher efficacy in alleviating the elevation in MLKL than SM.
The injection of VVF1 or SM to healthy rats (VVF1 group or SM group, respectively) did not cause any abnormal changes in the hepatic redox status or any elevation in the parameters of inflammation, necroptosis or fibrosis. More interestingly, the injection of healthy rats with VVF1 resulted in a significant increase in hepatic GSH in the C group (Figs. 4 and 5). In this study, it was noted that the injection of rats with a CCl 4 vehicle (olive oil, V group) did not show a strong effect on the most tested parameters. This declares that the hepatotoxic effect of CCl 4 was mostly attributed to its potency alone (Figs. 4 and 5).
preserving liver morphology, architecture, and functions in ccl 4 -VVF1 group. Histopathological analysis of rat livers (brown color liver) of CCl 4 group confirmed the incidence of necroinflammation with dense fibrosis bands that accompanied with steatosis (accumulation of fat vacuoles) comparing to only congestion in case of V group. Meanwhile, liver tissues of CCl 4 -VVF1 group, VVF1 group or SM group showed normal liver -induced hepatotoxicity with VVF1 and SM, respectively; VVF1 and SM, healthy rats were treated with VVF1 and SM, respectively. CCl 4 -VVF1 was compared with C, CCl 4 , V, and CCl 4 -SM while VVF1 was compared with C and SM. These comparisons were considered significantly different at *p < 0.05, **p < 0.005, ***p < 0.0005. (2020) 10:2452 | https://doi.org/10.1038/s41598-020-59489-z www.nature.com/scientificreports www.nature.com/scientificreports/ image (dark reddish-brown color) and healthy hepatocytes like C group with no signs of necroinflammation, steatohepatitis or fibrosis. Also, the CCl 4 -SM group had normal liver (slight reddish-brown color) but with dilated sinusoids (Fig. 6).
Additionally, Table 2 illustrates that the CCl 4 group exhibited a defect in liver function which was indicated by lowering blood albumin and increasing blood cholesterol compared to C group. However, treatment with VVF1 normalized blood levels of albumin and cholesterol. This enhancing effect of VVF1 on liver function was higher than that of SM (p > 0.05). There was no significant difference between all experimental groups regarding alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activity (Table 2).

Discussion
Herbal products have been used as drugs for thousands of years and recorded from many ancient cultures as historical evidence. VV is one of those herbs that are rich in phenolic compounds with other constituents. We prepared three different phenolic fractions from the black VV seedless fruit (pulp and skin) crude extract (VVCE), VVF1, VVF2, and VVF3. The results found that VVF1, which represents the highest yield, had the greatest phenolic content compared to VVF2 and VVF3. Phenolic compounds are widely known for their successful antioxidant activities through different mechanisms such as electron or hydrogen donors to prevent the chain radical reaction 13 . Therefore, VVF1, among the prepared fractions of VV, had the strongest antioxidant activities and was able to scavenge different types of free radicals as shown in Fig. 1. Moreover, VVF1 showed greater antioxidant activity than BHT or Asc and also SM. In addition, VVF1 exhibited synergistic antiradical activities against hydroxyl, superoxide, and DPPH radicals as well as anti-lipid peroxidation (β-carotene-linoleate bleaching inhibition activity). This synergistic antioxidant effect of VV polyphenols has mediated the powerful efficiency of VVF1 among VVFs. The HPLC analysis revealed different types of phenolic compounds in VVF1, some of which demonstrated their synergism due to their coexistence or combination. The previous studies revealed synergistic antioxidant activities of mixing some phenolic compounds such as vanillic, gallic, and chlorogenic acids 14 as well as catechin and resveratrol 15 . Furthermore, the combination of rosmarinic acid and quercetin, rosmarinic acid and caffeic acid and the mixture of caffeic acid, ferulic acid, and epigallocatechin-3-gallate demonstrated synergistic antioxidant actions 16 . VVF1 showed also more potent antioxidant activities than the crude extract (VVCE). , collagen type I alpha 1 chain (COL1A1) and transforming growth factor (TGF)-β1 in rat liver tissues with its heat map distribution that represents the relative gene expressions of CCl 4 , CCl 4 -VVF1 and CCl 4 -SM groups, the color distributed from blue (downregulated genes) to red (upregulated genes). C, control untreated healthy rat group; V, olive oil (vehicle of CCl 4 )-injected rats; CCl 4 , CCl 4 -induced hepatotoxicity group; CCl 4 -VVF1 and CCl 4 -SM; the treatment of CCl 4 -induced hepatotoxicity with VVF1 and SM, respectively; VVF1 and SM, healthy rats were treated with VVF1 and SM, respectively. CCl 4 -VVF1 was compared with C, CCl 4 , V, and CCl 4 -SM while VVF1 was compared with C and SM. These comparisons were considered significantly different at *p < 0.05, **p < 0.005, ***p < 0.0005.

Scientific RepoRtS |
(2020) 10:2452 | https://doi.org/10.1038/s41598-020-59489-z www.nature.com/scientificreports www.nature.com/scientificreports/ This may be due to the enrichment of this fraction mainly with phenolic compounds and/or the presence of certain antagonistic interactions between different constituents in VVCE. Hence, certain previous studies found that the combination of caffeic acid and α-tocopherol exerted antagonistic antioxidant effect 15 .
The present study was in vitro evaluated the potential role of the three VV prepared fractions against CCl 4 -induced hepatotoxicity. The CCl 4 is a well-known hepatotoxin that caused hepatic oxidative stress, inflammation, necroptosis, fibrosis as well as liver cancer in rats 17 . CCl 4 metabolizes by the hepatic Cytochrome P4502E1 (CYP2E1) producing trichloromethyl free radical (CCl3 * ) that combines with oxygen to generate the more reactive trichloromethyl peroxyl radical (CCl3OO * ) 18 . Therefore, the level of ROS was elevated in the hepatocytes after their exposure to CCl 4 (Fig. 2DI,II) that resulted in oxidative damage-dependent necroptosis promoting liver injury. The latter was assured by the increasing size of hepatocytes and the loss of their normal spindle shape as well as red fluorescence of AO/EB-stained swollen nuclei with a high percentage of annexin-PI stained necroptotic cells (39.64%) as shown in Fig. 2A The treatment with SM or VVFs clearly improved this damage as appeared under the light and fluorescence microscope ( Fig. 2A,B). Moreover, VVF1 showed the most potent efficacy in suppressing the oxidant-mediated necroptosis compared to the other two fractions (p < 0.005), crude extract (p < 0.05), and SM (p < 0.005). This may be due to its strongest antiradical activity (Fig. 1) that enable it to scavenge the most generated free radicals of CCl 4 metabolism in hepatocytes compared to other VVFS and SM (Fig. 2D) and in turn, prohibited the . Liver morphology and histopathological images confirming the therapeutic potency of VVF1 as an anti-hepatotoxic agent. C, VVF1 and SM groups showing healthy liver (reddish-brown liver color). Congestion (black arrow) in V group and necrosis (green arrows) with dense fibrous bands (red arrows) as well as steatosis with clusters of inflammatory cells (blue arrows) in the CCl 4 group (brown liver color with the abnormal white color area "yellow arrow"). No necroinflammation, steatosis, and fibrosis were shown in CCl 4 -VVF1 and CCl 4 -SM (with dilated sinusoids), magnification x200. C, control untreated healthy rat group; V, olive oil (vehicle of CCl 4 )-injected rats; CCl 4 , CCl 4 -induced hepatotoxicity group; CCl 4 -VVF1 and CCl 4 -SM; the treatment of CCl 4 -induced hepatotoxicity with VVF1 and SM, respectively; VVF1 and SM, healthy rats were treated with VVF1 and SM, respectively.  Table 2. Liver function markers and cholesterol level in the sera of rats in all the studied groups. Results are presented as Mean ± SE (n = 7). CCl 4 -VVF1 was compared with C, CCl 4 , V, and CCl 4 -SM while VVF1 was compared with C and SM. These comparisons were considered significantly different at *p < 0.05, **p < 0.005, ***p < 0.0005. ALT, alanine aminotransferase; AST, aspartate aminotransferase; SM, Silymarin; VVF1, Vitis vinifera fraction 1; V, vehicle.
Scientific RepoRtS | (2020) 10:2452 | https://doi.org/10.1038/s41598-020-59489-z www.nature.com/scientificreports www.nature.com/scientificreports/ necroptosis-dependent cellular damage. These findings were consistent with the previous studies that proved the significance of phenolic-rich extracts in avoiding CCl 4 -induced hepatic damage 19 . Furthermore, VVF1 exhibited a synergistic (CI ˂ 1) effect in lowering the ROS level and necrotic hepatocytes, which may be related to the synergistic antioxidant action of its phenolic contents as discussed above.
The current study also investigated the therapeutic effect of the most potent VV fractions (VVF1) on the CCl 4 -induced hepatotoxicity in rats to confirm the in vitro outcomes. The injection of CCl 4 in rats resulted in ROS production in liver tissue, which deactivated with the cellular antioxidants to maintain its threshold level. After multiple injections of CCl 4 , the elevated ROS level, altering the redox state, and damaging the cellular macromolecules were observed. The membrane lipids are the most susceptible to ROS damage causing lipid peroxidation that mediated cellular oxidative damage. The antioxidant defense system including GSH, SOD, and GPX was depleted after raising lipid peroxidation and ROS 20 causing a reduction in the hepatic TAC level. This imbalance between the ROS level and the antioxidant defense system has led to oxidative stress conditions in the hepatic tissue. The present findings were in accordance with the previous study of Adewale et al. 21 .
The treatment of CCl 4 -administered rats with VVF1 restored the oxidative damage by diminishing the ROS level, inhibiting lipid peroxidation as well as normalizing the hepatic antioxidant indices and TAC (Fig. 4). These results were in harmony with our in vitro outcomes, which elucidated the potential role of VVF1 in quenching superoxide, hydroxide, and peroxide radicals. This may be owed to the phenolic content of VVF1, which not only scavenged the generated ROS from CCl 4 metabolism but also improved the antioxidant defense system. The enhancing antioxidant abilities of gallic 22 , vanillic 23 , caffeic, syringic, p-coumaric, ferulic, ellagic, and salicylic acids 24 , as well as flavonoids 25 , have been reported previously. In addition, the administration of VVF1 alone without CCl 4 improved the redox state (TAC) of the liver tissue by a significant increase in the level of GSH. This may be due to certain constituents in this fraction, which are capable of increasing the expression of γ-glutamylcysteine synthetase, the rate-limiting enzyme of GSH synthesis, such as flavonoids 25 . Increasing the level of GSH will result in increasing GPX activity (p > 0.05) due to its essential role as enzyme co-substrate 20 . These results are in line with the previous work of Ragab et al. who proved the role of VV seed extract in improving the CCl 4 -induced oxidative stress in rat liver 11 . The current study found an improvement in the antioxidant status of the liver after treatment with SM, but with extremely less potency than VVF1. The antioxidant mechanism of SM was known before and related to its direct ROS scavenging ability, inhibiting ROS-producing enzymes, activating enzymatic and non-enzymatic antioxidants 26 . The ability of SM to overcome the CCl 4 -induced oxidative damage in the liver was reported previously 27,28 . Moreover, the present study found that SM had no significant effect on the antioxidant status of the liver tissue upon its oral administration to normal rats for ten days.
The present research assessed the anti-inflammatory function of VVF1 in CCl 4 -induced hepatic necroptosis in addition to the oxidative stress. Hence, there is strong crosstalk between these two damage effects for the induction of necroptosis-dependent hepatotoxicity. Here, we evaluated certain prooxidant inflammatory parameters, including NF-κB, COX-2, TNF-α, iNOS, NO, and MPO, all of which were raised after CCl 4 injection. ROS generation could modulate the NF-κB response and its target genes, including COX-2, TNF-α, and iNOS. The latter is responsible for the formation of NO that can interact with superoxide radical resulting in the creation of extremely reactive peroxynitrite. However, COX-2 catalyzes the conversion of arachidonic acid to prostaglandin H2 in addition to the generation of superoxide radicals as side products. All of these mediators, alongside TNF-α, which amplify NF-кB, have potentiated CCl 4 damage in hepatic tissue 29 . In addition, CCl 4 was correlated with an increase in MPO activity which magnifies the inflammatory signal and stimulates lipid peroxidation in the presence of halide ions and H 2 O 2 30 . This may be related to its principal role in the formation of hypochlorous acid (HOCl) within the neutrophils from Cl − and H 2 O 2 31 . The formation of HOCl may further contribute to the consumption of GSH owing to the ability of this antioxidant molecule to interact with it 32 . Therefore, MPO not only amplified the inflammatory response within the hepatic tissue but also increased oxidative stress. This study showed that the animals administered olive oil had normal prooxidant inflammatory markers. However, the treatment of CCl 4 -exposed animals with VVF1 considerably decreased all of the studied inflammatory mediators relative to the CCl 4 group. These outcomes were in line with Aouey et al. 33 and may be attributed to the anti-inflammatory potential of its active ingredients, including ferulic, caffeic, p-coumaric, salicylic, and ellagic acids 24 as well as flavonoids 33 . The efficiency of VVF1 was not only related to the presence of these polyphenols but also the synergistic effect between them. Hence, the synergistic anti-inflammatory activities of certain phenolics such as resveratrol and quercetin 34 as well as phenolic-containing extracts have been reported before 35,36 . Moreover, the current research has confirmed the anti-inflammatory action of SM, which has been reported earlier 37 and has shown a lower anti-inflammatory effect than our VVF1. These results proved and clarified the potent anti-inflammatory activity of this grape fraction.
Induction of both oxidative stress and inflammation in hepatic tissue by CCl 4 was considered to be the driving force of necroptosis. This damage effect is a form of necrosis that mediated via death receptors such as TNF, Fas, and TNF-related apoptosis-inducing ligand (TRAIL). This process has happened following the activation of receptor-interacting protein kinase 3 (RIPK3) and MLKL, and the inhibition of caspase 8, which converts extrinsic apoptosis to necrosis. Activation of RIPK3 phosphorylates MLKL leading to its translocation into the plasma membrane inner leaflet, causing perforation of membrane and disruption of the cell integrity. MLKL is significant for the induction of necroptosis and is the one that decides whether the cell is undergoing necroptosis or apoptosis 38 . In the current study, the injection of CCl 4 led to a dramatic increase in the hepatic level of MLKL, which confirms the induction of necroptosis in rat liver tissue. In addition, certain molecules released from necroptotic cells can trigger activation of hepatic stellate cells (HSCs) 39 , which play a key role in liver fibrosis after activation by TGF-β1 and TNF-α 40 . There are two essential markers (COL1A1 and TGF-β1) that imply HSC activation. After activation, they secrete collagen type I with other mediators to promote fibrogenesis 40 . Subsequently, the gene expression of both COL1A1 and TGF-β1 was significantly upregulated (Fig. 5B). Therefore, the injection of CCl 4 to rats in this study induced necroptosis and activation of HSC to promote fibrogenesis. These results were Scientific RepoRtS | (2020) 10:2452 | https://doi.org/10.1038/s41598-020-59489-z www.nature.com/scientificreports www.nature.com/scientificreports/ in harmony with our histopathological outcomes that revealed the presence of hepatic necrosis and steatosis with inflammation (steatohepatitis) and deposition of collagen fibers. Moreover, the serum profile of these rats showed a significant elevation of cholesterol level, which could lead to the accumulation of the lipid droplets in the liver. These results were in accordance with the previous studies 41,42 . The ALT and AST activities showed no significant change compared to the control rats in addition to the dramatic depletion of albumin level. ALT and AST leaked and raised in serum after hepatocyte death then their levels dropped and returned to normal after a few days. However, albumin has a longer half-life in serum, coupled with the capacity of the liver to synthesize it, so its concentration shifts slowly 43 . Therefore, the normal activities of these transaminases with depletion of albumin indicate liver damage and injury in the CCl 4 group. The administration of olive oil to rats in the V group did not significantly alter either the MLKL level or the studied pro-fibrotic mediators and only congestion and dilated sinusoids in its hepatic histopathological assessment. Thus, the main toxicity in CCl 4 -injected rats is related to CCl 4 itself, not to olive oil.
The treatment with VVF1 massively reduced hepatic MLKL relative to rats in the CCl 4 group. This was probably due to its phenolic-related antioxidant and anti-inflammatory activities. As VVF1 can diminish hepatic ROS, which plays a key role in the induction of TNF-mediated hepatic necroptosis thus CCl 4 -induced hepatic cell death, via elevation of MLKL, can be inhibited. No prior study has investigated the influence of VV on MLKL during hepatic necroptosis, so our research is the first to elucidate this point. However, few studies have investigated the influence of phenolic compounds on necroptosis. Recently, the protective effect of certain phenolic-containing extract against necroptosis was explored 44 . In contrast, gallic acid proved its efficiency in the induction of necroptosis-dependent death in the activating HSC and was considered as a new therapeutic strategy for the avoidance of hepatic fibrosis 45 . Therefore, phenolic compounds can perform dual actions during hepatotoxicity by stimulating the necroptosis, especially, in the activated HSCs, thereby halting fibrogenesis and preventing death for other hepatic cells. Subsequently, the phenolic compounds in VVF1 assumed inhibition of hepatic necroptosis and eliminated the activated HSCs by decreased MLKL level and TGF-β1 expression and, in turn, decreased the gene expression of COL1A1. These results were confirmed by the histopathological findings and the serum profile of liver function. The present research also disclosed the enhancing ability of SM for hepatic necroptosis, the depletion of pro-fibrotic mediators, and the improvement of the serum profile that was less potent than VVF1. In addition, the administration of SM or VVF1 without CCl 4 had no toxicity on the liver.

conclusions
The highest polyphenol-enriched fraction of the seedless black VV fruit (VVF1) exhibited the strongest antioxidant-dependent anti-necroptotic impact against CCl 4 -exposed hepatocytes. Furthermore, this study declared, for the first time, the synergistic antioxidant and anti-necroptosis effectiveness of VVF1's constitutes. Based on our best knowledge, no previous studies reported the in vivo efficacy of VVF1 to suppress necroptosis and fibrosis-mediated hepatic damage by the normalization of cellular redox status as well as lowering the pro-necroptotic protein (MLKL), -inflammatory and -fibrotic mediators (TGF-β1 and COL1A1). Additionally, VVF1 has been able to improve liver architecture (no necrosis, fibrosis, steatosis, and inflammation in contrast to the untreated CCl 4 group) and functions. Moreover, these above-mentioned investigations revealed that VVF1 possessed higher anti-hepatotoxicity potentials than the standard SM drug and thus VVF1 considered as a new, effective natural anti-hepatotoxic agent for targeting necroptotic and profibrotic mediators. . Gene JET RNA purification kit, cDNA synthesis kit, and 2X SYBR green master mix kit were supplied from Thermo Fisher Scientific, USA. MLKL and resveratrol ELISA kits were obtained from Cloudclone Corp, USA and GmbH, Aachen, Germany, respectively. ALT and AST, albumin, and cholesterol kits were purchased from Biosystem, Spain. Primers were purchased from Bioneer, Korea. Other chemicals were obtained with a high grade.
Animals. Fifty-nine male Albino rats were purchased from MISR University for Science and Technology with animal welfare (assurance number: A5865-01). Rats were acclimatized under the conventional conditions of about 30 °C with a 12-hour light-dark cycle for two weeks. During this period, animals allowed free access to tap water and a standard commercial diet. All relevant international and/or institutional recommendations for using animals were followed. plant material and preparation of the phenolic fractions. The VV (NCBI:txid29760) was imported from Lebanon and used for the preparation of three phenolic fractions. The black color fruit (seedless pulp with www.nature.com/scientificreports www.nature.com/scientificreports/ the skin) was ground using an electric grinder, then lyophilized (Telstar, Terrassa, Spain) to obtain the powdered VVCE (yield 14.020 ± 0.140 g/100 g grape). Then 50 g of VVCE was extracted twice with 70% ethanol (500 mL for each) using reflux for an hour at 50 °C. After filtration, the obtained extract (VVF1) was distilled to remove ethanol and then cooled for 12 h in a refrigerator (4 °C) for precipitation of the ethanol-soluble components. The precipitate (resveratrol-rich fraction, VVF2) was separated from the filtrate (water-soluble components, VVF3) by centrifugation for 10 min at 4000 rpm. The obtained fractions (VVF1, VVF2, and VVF3) were freeze-dried and stored at −20 °C until used.
Spectrophotometric and chromatographic analysis of the VV polyphenols. The total phenolics and flavonoids in the crude extract (VVCE) and the prepared fractions (VVF1, VVF2, and VVF3) were determined spectrophotometrically. Total phenolics were quantified by the Folin-Ciocalteau method using 4-HCA calibration curve 46 . Total flavonoids were measured at 510 nm after mixing each fraction with 5% sodium nitrite and 10% AlCl 3 and the concentration was calculated using RU standard curve 47 .
Twenty microliters of VVF1 were separated on the Zorbax Eclipse plusC18 column (100 mm × 4.6 mm Agilent Technologies, Palo Alto, CA, USA). The separation was achieved at 284 nm with a flow rate of 0.75 mL/min using a ternary linear elution gradient and a mobile phase of 0.2% H 3 PO 4 , methanol, and acetonitrile. Under similar chromatographic conditions, pure phenolic standards were run to match the retention items 48 .
The content of the resveratrol was determined using a specific ELISA kit. The kit depends on a competitive inhibition enzyme immunoassay technique using resveratrol-specific monoclonal antibody and avidin conjugated Horseradish Peroxidase. Antioxidant activities. The total antioxidant capacity (TAC), β-carotene-linoleate bleaching assay, and antiradical activities (anti-DPPH, superoxide, and hydroxyl radicals) of the crude and prepared phenolic fractions (VVCE, VVF1, VVF2, and VVF3) were tested. The value of IC 50 (50% inhibitory concentration) for each studied radical and the β-carotene-linoleate bleaching was estimated by the GraphPad Instat software version 3.
The TAC of the studied fractions was determined using a mixture of ammonium molybdate (4 mM), sodium phosphate (28 mM), and H 2 SO 4 (0.6 M). After 90 min at 95 °C, the reducing ability of the studied fractions to phosphomolybdate was read at 695 nm using Asc standard curve 49 .
The DPPH scavenging activity was evaluated following the modified standard method of Blois 50 by incubating serial concentrations of VV crude or fractions with freshly prepared 0.004% DPPH (in methanol) at room temperature for 30 min. Then the absorbance of the non-scavenged DPPH was measured at 490 nm. Regarding, the hydroxyl radical scavenging activity was assessed at 510 nm after 60 min incubation of serial dilutions of samples with 9 mmol/L of salicylic acid, FeSO 4 and H 2 O 2 , at 37 °C 51 . The method of McCord and Fridovich 52 was used to determine the scavenging activity of the studied fractions to superoxide anion radical. In this method, serial concentrations of VV or each fraction were incubated with a reaction mixture containing 0.1 M EDTA, 1.5 mM NBT, 0.0015% NaCN, 0.12 mM riboflavin, and 67 mM phosphate buffer (pH 7.8) for 15 min. The absorbance was then recorded at 530 nm.
The β-carotene-linoleate bleaching method measured the anti-lipid peroxidation effect of the samples using β-carotene, linoleic acid, and Tween-80 emulsion 53 . The decrease in the bleaching rate of β-carotene which reflects the ability of each fraction to scavenge the radicals generated from linoleic acid oxidation was recorded. The absorbance (a, b) was read at 490 nm instantly and after 180 min (t), respectively. Then the value of the degradation rate (DR) of each fraction, standard (BHT), and control (without fraction) was calculated using the equation: [DR = ln (a/b) × (1/t)]. The antioxidant ability was determined as % of inhibition using the formula: Antioxidant activity (%) = (DR control − DR fraction /DR control ) × 100.
In vitro evaluation of the most effective anti-hepatotoxic VV fractions. Assessment of the safe concentrations of VV fractions on the isolated hepatocytes. The liver of three male Albino rats (weighing 30-35 g, 2 weeks age) was used to isolate the hepatocytes according to the method of Whitehead and Robinson with some modifications 54 . After hepatocyte isolation using collagenase I and cultivation in William's E medium supplemented with FBS (10%), the cytotoxicity of the phenolic fractions on the hepatocytes was tested using MTT assay 55 . Briefly, the isolated hepatocytes were seeded in 96-well cell culture plate and treated individually with serial concentrations of VVF1, VVF2, VVF3 as well as the standard drug for hepatotoxicity (SM). Also, the untreated cells were considered as a negative control sample. The plates were incubated for 72 h at 5% CO 2, 37 °C and 90% relative humidity in the CO 2 incubator. Then MTT (5 mg/mL) was added to each well and incubated for an additional 4 h. At the end of the incubation period, MTT was replaced by 150 µl of DMSO and the absorbance was read at 570 nm using an ELISA reader (BMG LabTech, Germany). Safe concentration at 100% cell viability (EC 100 ) was calculated using GraphPad InStat software.
In vitro induction of hepatotoxicity and treatment procedure. The hepatotoxicity was induced in the isolated rat hepatocytes by incubation with 0.13 mM CCl 4 using the previous method of Abu-Serie & habashy 56 . After 36 h, cells were incubated with serial dilutions of safe concentration (EC 100 ) of VV fractions and SM for 72 h in 5% CO 2 incubator at 37 °C. The CCl 4 -exposed cells without any further treatments and the normal untreated hepatocytes were used as positive and negative control cells, respectively. Then the cell viability after each treatment compared with the control cells was assessed using MTT as indicated above 55 . The theoretical effective doses of each VV fraction and SM that terminated CCl 4 -induced hepatotoxicity by 100% (ED 100 ) and 50% (ED 50 ) were calculated by the GraphPad Instat software version 3. The further analyses of the VV fractions and SM effectiveness on the CCl 4 -induced hepatotoxicity were performed at the theoretical ED 100 value for each sample. (2020) 10:2452 | https://doi.org/10.1038/s41598-020-59489-z www.nature.com/scientificreports www.nature.com/scientificreports/ Morphological examination of the CCl 4 -damaged hepatocytes after VV treatment. Before and after the treatment of the CCl 4 -exposed hepatocytes with the VV fractions and SM, cells were investigated morphologically using a phase-contrast microscope (Olympus, Japan). In addition, EB and AO dyes (100 µg/mL for each) were used for staining the cells to examine the influence of the VV fractions and SM on CCl 4 -induced hepatocyte death. Then the stained hepatocytes were visualized under the fluorescent phase-contrast microscope (Olympus, Japan).
Flow cytometric analysis of necroptotic hepatocytes and intracellular ROS quantification. The quantification of necroptosis was determined by the flow cytometer using annexin V/PI stain. The untreated and treated hepatocytes were trypsinized and incubated with isothiocyanate (FITC)-labelled annexin V/PI for 15 min. The choice of the most effective anti-hepatotoxic VV fraction was established by quantification of the annexin/PI-stained population in the studied cells. This was done by the flow cytometer using the phycoerythrin emission signal detector (FL2) against the FITC signal detector (FL1).
The DCFH-DA (5 μM) probe was used for the quantification of the intracellular ROS level 57 . The untreated and treated hepatocytes were incubated in the dark with the fluorescent probe for 30 min at 37 °C. Then cells were trypsinized and suspended in phosphate-buffered saline (PBS) to analyze the fluorescence intensity by flow cytometer (Partec, Germany). The analysis was done at an excitation and an emission wavelength of 488 and 530 nm, respectively.
Combination index (CI) analysis. The combination of VVF2 phenolics with VVF3 phenolics in VVF1 may or may not (additive effect) confer higher (synergistic) or lower (antagonistic) antioxidant and anti-hepatotoxic activity for VVF1. The probable new antioxidant effect can be evaluated using the effect "Fa"-CI and isobologram plots. In addition, the CI values at Fa 0.5 (50% inhibition), Fa 0.75 (75% inhibition), and Fa 0.9 (90% inhibition) were calculated. Both the plots and the CI values were generated by the current software, CompuSyn. However, the anti-hepatotoxic (in vitro) activity (% necroptotic cells and the intracellular ROS) of VVF1 was detected using the CI values that were calculated by dividing the expectable value by the observed value. The CI value may be ˂ 1 (synergistic effect), equal to 1 (additive effect), or ˃ 1 (antagonistic effect) 48 .
In vivo evaluation for the anti-hepatotoxic effect of the most effective VV fraction. Experimental scheme. Fifty-six rats (weighing 140-200 g, 6 weeks' age) were divided randomly into seven groups (eight animals in each). The treatment of animals in each experimental group is illustrated in Fig. 1. Briefly, hepatotoxicity was induced in rats by intraperitoneal (i.p.) injection of 1 ml 50% CCl 4 in olive oil/kg b.w./twice/week for 3 weeks 58 . After these 3 weeks, CCl 4 -VVF1 and CCl 4 -SM rat groups were orally injected with 1.5 g VVF1 and 50 mg SM/kg b.w., respectively, using gavage, daily for 10 days. At the end of the experimental period (day 30), rats were anesthetized and dissected then blood (by cardiac puncture) and liver tissues were collected immediately. The heparinized blood was centrifuged at 6000 rpm (15 min) to separate plasma for assessment of the liver function markers, cholesterol, and TAC. Liver tissues were washed with cold saline (0.9% NaCl), weighted and small portions were fixed in 10% formalin for histopathological investigation. The remaining liver tissue was kept at −80 °C until used in the biochemical and molecular analyses.
Assessment of necroptotic and fibrotic mediators in liver tissues. The necroptotic mediators (oxidative stress and necroinflammation), as well as the fibrotic mediators (TGF-β and COL1A1), were assessed at mRNA and protein levels in rat livers to evaluate the anti-hepatotoxic role of VVF1 in vivo.
Biochemical assessment of hepatic oxidative stress and inflammation-dependent necroptotic mediators. The oxidative stress (cellular redox state disruption)-mediated inflammation in the liver homogenate was assessed by determination of the intracellular ROS, NO, TAC, lipid peroxidation, MPO and the antioxidant indices levels. The homogenates were prepared by homogenizing the liver of each rat in each experimental group in fresh cold PBS (1:10 w/v) then centrifuged at 6000 rpm (4 °C) for 30 min and the clear supernatants were used for the analyses.
The ROS level was assessed using the extremely sensitive DCFH-DA (5 μM) fluorescent probe. In brief, the diluted clear homogenate (2-fold dilution with PBS) of each sample was mixed with an equal volume of diluted DCFH-DA (1000-fold dilution with 10 µM dimethyl sulfoxide). Then the reaction mixtures were incubated, in dark, for 5 min at 37 °C and finally, the fluorescence intensity was measured at 485 (excitation) and 520 nm (emission) to calculate the ROS level using H 2 O 2 standard curve 59 .
NO was determined as nitrite using a Griess reaction that produced colored azo dye with a maximum absorbance at 490 nm 60 . The lipid peroxidation level was examined by TBA reactive substances (TBARS) colorimetric method 61 using TMP calibration curve.
Myeloperoxidase activity was examined colorimetrically using ODD (16.7 mg%) and H 2 O 2 (1.2%) as described previously 62 . The enzyme activity was measured as IU/mg protein, where one IU is defined as the amount of the enzyme that able to degrade 1 μmoL of H 2 O 2 /min at 25 °C.
The antioxidant indices comprising the activities of SOD and GPX were assessed using the pyrogallol autooxidation assay 63 and Rotruck method 64 , respectively. The specific activities of SOD and GPX were determined by dividing the activity of each enzyme by the protein content in the homogenate. The protein level was assessed using biuret method followed the manual protocol of the specific kit. In addition, the level of nonenzymatic antioxidant (GSH) was determined using Ellman's reagent (5, 5′-dithio bis2-nitrobenzoic acid) and its concentration was calculated from the GSH calibration curve 65 .
Histopathological examination. After fixation, liver tissue specimens were processed by following the routine protocol for histopathological investigation. Hence, the samples were embedded in paraffin wax and 5 µm thickness slices were cut and stained with hematoxylin and eosin. The pathological features of liver tissues in all the studied groups were visualized by the phase-contrast microscope then high-resolution images were captured at 200x magnification.
Plasma analyses. The liver function markers, including ALT, AST, albumin as well as cholesterol level were determined spectrophotometrically in plasma samples of rats in all the studied groups using the specific kits.
Statistical analysis. The data are expressed as mean ± SE and the p-value < 0.05 is considered significant. The difference between the mean values of the studied groups was evaluated by one-way analysis of variance (ANOVA) by Tukey's test. Before applying this parametric test, all data were checked for their normal distribution (skewness 0-0.868). The analysis was performed for seven rats using SPSS software version 16. The IC 50 , EC 100 , ED 50 , and ED 100 values were calculated by GraphPad Instate software version 3. In addition, the CompuSyn software (ComboSyn, Inc, Paramus, NJ) accomplished the CI values, Fa-CI plots and isobologram plots for the in vitro antioxidant experiments.

Data availability
All data produced during this study is included in this published article.