Fresh mature mulberry variety PR-01 fruits was obtained in May 2016 from the Institute of Agricultural Product Quality, Fujian Agriculture and Forestry University (Fujian, China) and stored at −80˚C on the same day.

All chemicals were ordered from Sigma-Aldrich (Shanghai, China) or Beijing Dingguo Changsheng Biotech Co., Ltd. (Beijing, China). TRIzol reagent, HiScript II Q RT SuperMix for qPCR (+cDNA wiper) and AceQTM qPCR SYBRR Green Master Mix were purchased from Takara Co., Ltd. (Dalian, China). Rabbit polyclonal antibodies nuclear factor-kappaB (NF-κB), cyclooxygenase-2 (COX-2), tumor necrosis factor alpha (TNF-α), Kelch-like ECH-associated protein 1 (Keap1), NAD(P)H:quinine oxidoreductase1 (NQO1), heme oxygenase-1 (HO-1) and Nrf2 were purchased from Santa Cruz Biotechnology or Cell Signaling (Danvers, MA, US).

The PR-01 mulberry anthocyanin rich-extract was extracted as previously described [17] . Freeze mulberry fruits were dried in an ALPHA 1-2LD PLUS lyophilizer (Beijing, China), then ground and sifted for homogenization and stored at −80˚C. The finedly ground mulberries (3 kg) were extracted with 300 L of methanol-hydrochloric acid (999:1) at 37˚C under the condition of ultrasonic power 200 w and frequency 40 kHz for 30 min. This process was repeated three consecutive times. The extracts is then mixed and evaporated into syrup using RE-2000E rotatory evaporator (Xi’an Taikang Biotechnology Co., Ltd., Xi’an, China). The syrup was dissolved in 1000 mL of 0.01 mol/L HCl, and was defatted with 1200 mL of ethyl acetate. The water solution was subjected to column chromatography of Amberlite XAD-7 orderly eluted with water and methanol-formic acid (9:1) [18] [19] . The methanol-formic acid elution solution was evaporated into syrup and freeze-dried to yield 403 g of MAEs. The MAEs stored at −80˚C for further use.

Reference to previous study [20] , the MAEs was diluted 10-fold with hydrochloric acid-sodium chloride of pH 1.0 and pH 4.5, respectively. The absorbance of MAEs diluents was measured at wavelengths of 520 nm and 700 nm by 755b UV-Vis spectrophotometer (Xi’an Heb Biotechnology Co., Ltd, Xi’an, China), respectively. The absorbance (A) was calculated by the Equation (1):

A = (A520 nm − A700 nm) pH 1.0 − (A520 nm − A700 nm) pH 4.5 (1)

The total anthocyanin content of MAEs was calculated according to the formula derived by Wrolstad et al. [21] :

Total anthocyanin content (mg・g⁻¹) = A × M W × D F × V × 1000 ε × 1 × M (2)

where MW is the molecular weight of cyclamine-3-glucoside (449.2), DF is the dilution, ε is the molar absorptivity (29,600), l is the optical path (1.0 cm), V is the total volume of MAEs diluents (1.0 ml), and M (0.01 g) is the weight of the MAEs.

Quantifications of anthocyanin monomers were performed on a LC-MS/MS (Thermo Fisher Scientific Inc., SAN JOSE, Calif.) with a MRM mode. The LC-MS analysis was carried out using a LC-MS system comprising of a LC (2010, Finnigan, USA) and LCQ Fleet Ion Trap LC/MS (LCQ Fleet, Thermo Fisher Scientific). Mass spectrum condition: positive ion scanning (ESI⁺, m/z 200 ~ 1200), the collision induced dissociation voltage was 150 V; capillary voltage: 3.8 kV; taper hole voltage: 30 v. Voltage of photoelectric multiplier: 650 V; Ion source temperature: 120˚C; Take off the solvent temperature: 250˚C.

Separation of anthocyanins in MAEs was on a 2.1 × 150 mm, 1.7 μm ACQUITY UPLC BEH C18 column (Waters, UK) analyzed by UPLC using a ACQUITY UPLC H-Class system with a 2996 PDA detector (Waters, Milford, MA, USA). Anthocyanin standards (Sigma-Aldrich, Shanghai, China) were used for quantitative determination.

All animal studies were conducted in accordance with Chinese National Guidelines for the Care of Laboratory Animals and approved by the Animal Ethics Committee of the Laboratory Animal Center, Fujian University of Traditional Chinese Medicine (Fujian, China). Four-week-old SPF-grade male Wistar rats were purchased from Shanghai slac laboratory animal Co., Ltd. with animal production license No.: SCXK (HU) 2012-0002. All animals were housed in an SPF barrier system of the Laboratory Animal Center, Fujian University of Traditional Chinese Medicine. Breeding conditions were as follows: temperature 20˚C ~ 25˚C, relative humidity 40% ~ 60%, 12 h day/night alternation.

Rats were fed with a rodent diet and had tap water ad libitum. After a week of adaptive feeding, the rats were randomly divided into five groups. The control group (Control), model group (NDEA), NDEA + low-dose MAEs (MAEs-75), NDEA + high-dose (MAEs-150), and NDEA + cyanidin-3-glucoside (C3G-150). The animal experiment design is showned in Table 1. Body weights were measured weekly. Twenty-four hours after the last MAEs administration, the rats were anesthetized by ip of pentobarbital sodium (40 mg・kg⁻¹) for collection of

blood samples and blood samples were centrifuged at 3000 rpm at 4˚C for 10 min to obtain serum. Livers were fully excised from the animal and accurately weighed [6] . The size and number of liver nodules were measured as described earlier [3] . Liver tissue was collected for biochemical and histopathological analyses.

The levels of serum aspartate transaminase (AST), alanine aminotransferase (ALT), γ-glutamyl transpeptidase (GGT), total bilirubin (TBiL), alkaline phosphatase (ALP), Blood Urea Nitrogen (BUN), and creatinine were detected according to the instructions of commercial kits (Nanjing Jiancheng Bioengineering Engineering Institute, Nanjing, China) using Infinite M200 PRO Varioskan Flash (TECAN, Switzerland). The levels of serum alpha fetal protein (AFP), carcino embryonie antigen (CEA), TNF-α, interleukin-6 (IL-6), interleukin-10 (IL-10) and interferon-γ (IFN-γ) were measured by ELISA kit (CUSABIO Biotech Co., Ltd., Wuhan, China) according to manufacturer’s instructions.

Precisely 1 gram of liver was taken, rinsed with cold PBS, pH 7.4, dried with paper wipe, and placed in a 10-mL centrifuge tube. Then the tissue homogenate was centrifuged at 3500 rpm for 15 min. The supernatant was collected in a 10-mL centrifuge tube (namely 10% liver homogenate), immediately frozen in liquid nitrogen and stored at −20˚C until use [6] . The contents of malondialdehyde (MDA), glutathione (GSH), and the activities of superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GSH-Px) were assayed using standard commercially available kits (Nanjing Jiancheng Bioengineering Engineering Institute, Nanjing, China) according to the manufacturer’s instructions. The serum thiobarbituric acid-reactive substances (TBARS) concentrations were assayed by rat TBARS ELISA kit (QIYI Biological technology Co., Ltd., Shanghai, China). The levels of Glutathione S-transferase (GST) and UDP-glucuro- nosyltransferase 2B1 (UGT2b1) in the liver homogenates were measured with ELISA kit (CUSABIO Biotech Co., Ltd., Wuhan, China).

Hepatic tissues from each group were washed with cold physiological saline and rapidly excised after the rats were sacrificed. The tissue samples were immediately fixed in 4% phosphate-buffered paraformaldehyde for 48 h, then dehydrated in a graded ethanol series, cleared in xylene and embedded in paraffin. Sections (4 μm thick) were stained with hematoxylin and eosin (H&E) according to published methods [22] . The samples were also sectioned and Masson’s trichome staining was performed as previous studies [23] . For histological analysis, the tissue sections were photographed with a microscope (Nikon E200, Nikon Corp., Japan).

The proteins expression of inflammatory markers in liver tissue was detected by the SABC immunohistochemical method. The main steps are as follows: paraffin section (4 μm thick) was de-waxed and incubated with sodium citrate buffer (pH 6.0) at 37˚C for 5 min (twice), then washed with 0.1 M PBS for 5 min (3 times), incubated with 3% hydrogen peroxide at 37˚C for 10 min, washed with PBS for 5 min (3 times) and placed in a blocking solution at room temperature for 1.5 - 3 h. The sections were incubated with rabbit polyclonal antibodies, NF-κB, COX-2, and TNF-α (Cell Signaling, Danvers, MA, USA) with TBS and Tween 20 overnight at 4˚C. Then sections were washed with PBS and incubated with HRP-labeled sheep-anti-rabbit secondary antibody at 37˚C for 1 - 2 h followed by streptavidin-biotin-peroxidase at room temperature for 30 min. The slides were washed and the immunoprecipitation was visualized by treating with 3, 3'-diaminobenzidine for color development for 25 min. Then slides were counterstained with hematoxylin and the brown color signifying the presence of antigen bound to antibody was detected by light microscopy. For the negative control, TBS was used instead of a primary antibody. From ten randomly selected sections of each slide, 500 cells were counted. The percentage of positive cells for each group was calculated using the Image-Pro Plus 6.0 image analysis system.

Liver tissue was lysated in RIPA lysis buffer containing protease inhibitor cocktail protease inhibitors and phosphatase inhibitors according to the M-PER (R) Mammalian Protein Extraction Reagent (Thermo Scientific, Fair Lawn, NJ, USA). The protein extraction was collected and the concentrations were quantified using a BCA kit (Biotechnology Development Co., Ltd., Beijing, China). The same amount of proteins was separated using 10% SDS-polyacrylamide gel and transferred to polyvinylidence fluoride (PVDF) membranes. After blocked with 5% skim milk for 1 h at room temperature, membranes were incubated with various antibodies against Nrf2 (1:2000), NQO1 (1:1000), HO-1 (1:2000), Keap1 (1:500), NF-κB (1:4000), COX-2 (1:500), TNF-α (1:1000), and β-actin (1:2000) which purchased from Cell Signaling Technology (Danvers, MA, USA) or Santa Cruz Biotechnology (Santa Cruz, CA, USA) overnight at 4˚C. Then membranes were washed and exposed to HRP-conjugated secondary antibodies (1:10,000) at room temperature for 1 h. Immunoreactive bands were detected by enhanced cheiluminescence solution (CUSABIO Biotech Co., Ltd., Wuhan, China) and exposed to X-ray film using the Bio-Rad Chemi Doc XRS imaging System. The immunoreactive bands were visualized and quantified using Quantity One software and normalized to β-actin.

Total RNA was extracted from liver tissue using a TRIzol reagent (Invitrogen, Thermo Fisher Scientific Inc., Beijing, China) according to the manufacturer’s instructions. Real-time PCR was performed using the SYBR Green Kit (Takara Biomedical Technology Co., Ltd., Beijing, China) on the ABI Step One RT-PCR system. Primers designed with Primer Premier 5 and Beacon Designer 8.1 was listed in Table 2.

Data was presented by mean standard deviation (SD). Data from study was dealed with SPSS 21.0 statistical package. Statistical analysis was performed by one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons or Student’s t-test. A P-value of <0.05 was considered statistically significant. The GraphPad Prism 6.01 was used for the graphical evaluations.

The total anthocyanins content of the PR-01 mulberry anthocyanins rich-extract was (718 ± 8.9) mg・100g⁻¹. UPLC-ESI-MS analysis showed the pelargonidin 3-glucoside (P3G), C3R and C3G is the main anthocyanins (Figure 1), which was in agreement with that reported by Song [17] and Huang [24] . UPLC analysis

Ip, intraperitoneal injection; C3G, cyanidin-3-glucoside; MAEs, mulberry anthocyanins extract; NDEA, N-nitrosodiethylamine. Gavage doses of MAEs intervention groups were calculated by C3G contents. The MAEs and C3G were dissolved in distilled water.

showed that the content of P3G, C3R and C3G in the MAEs is (75.4 ± 6.5) mg・g⁻¹, (105.8 ± 4.1) mg・g⁻¹, and (593.0 ± 12.6) mg・g⁻¹, respectively. The mass spectrum of MAEs is shown in Figure 1, while the characterization of chemical constituents of MAEs is presented in Table A1.

As shown in Table 3, the body weight of rats in NDEA group was significantly reduced subsequent to NDEA compared with that in control group (P < 0.01), while the weight gains of animals treated with MAEs and C3G-150 were similar(P > 0.05). With regard to average daily feed consumption, the NDEA group consumed a lower amount of the feed than the control group (P < 0.01). However, after treatment with MAEs and C3G, the daily feed intakes were significantly increased than the NDEA group (P < 0.05). Compared to NDEA group, the relative liver weight was remarkably decreased in the MAEs-75 and MAEs-150 groups by 23.64% and 28.37%, respectively after MAEs treatment (P < 0.01).

NDEA causes liver damage and consequently releases liver enzymes in its selected dose [25] . The past research suggested that mulberry water extracts may serve as liver protective agents [4] [25] . In this study, the elevated activity of ALT, AST, TBiL, ALP, and GGT are indicative of poor hepatic function in the NDEA alone treated animals compared to the control animals (Table 4), as reported by Latief et al. [26] . After treated with MAEs, the levels of the elevated liver function enzymes above mentioned were significantly restored in MAEs-75 and MAEs-150 group animals, suggesting that MAEs has an efficacy function on liver protection. Pretreatment with the higher dose of MAEs (150 mg・kg⁻¹) also reversed the NDEA-induced increase in serum BUN and creatinine to normal levels (Table 4). In addition, the serum CEA and AFP levels were obviously higher in NDEA group than that of the control group (P < 0.05), while 75 mg・kg⁻¹ MAEs and 150 mg・kg⁻¹ MAEs treatment remarkably reduced the serum CEA and AFP levels by 25.99%, 39.32%, 32.57%, and 49.24%, respectively compared with the NDEA group (Table 4, P < 0.05). Pretreatment with higher dose of MAEs (150 mg・kg⁻¹) appeared to be more efficient in inhibiting NDEA-induced

Note: ^#Relative liver weights equal liver weight/body weight × 100. Values expressed as mean ± S.D. (n = 8 - 10). Values with different superscript letters (a, b, c, d, e) within cultivar are significantly different.

Values with different superscript letters (a, b, c, d, e) within cultivar are significantly different.

liver damage in rats.

As one of the most important environmental hepatotoxin and carcinogen, NDEA can cause severe liver damage and lead to severe alterations in the lobular architecture and hampers liver functioning [26] . During the experiment, 2 rats died of severe hepatic tumor pathogenesis at 16^th and 18^thweek in the NDEA group. No death was visible in other group animals. The results of this study shown that low-dose multiple injection of NDEA (25 mg・kg⁻¹ of NDEA twice a week) significantly increased the nodule incidence and the apparent HCC incidence (100.0%, 50.0%, respectively) and only a low mortality rate (16.7%) for 18 weeks in the NDEA group, indicating successful induction of liver cancer model. However, there was significant decrease in hepatic nodules incidence, maximum diameter, average number in the MAEs and C3G-150 group rats as compared with NDEA group animals (Table 5, P < 0.05). In particularly, treatment of 150 mg・kg⁻¹ MAEs prior to C3G showed a significant reduction of the nodule incidence (41.7%) compared to the NDEA group animals (100%).

As shown in Table 5, treatment of 150 mg・kg⁻¹ MAEs prior to C3G showed a significant reduction of the apparent HCC incidence (8.3%) compared to the NDEA group animals (50.0%). The liver phenotype of rats in the control group was not significantly changed, while the liver size and structure of NDEA group rats were significantly changed (Figure 2). Apparently the liver structure of injured rats was characterized by cells necrosis, hemorrhage, scars and hepatic nodule like structure, abundant eosinophilic cells, basophilic cells and egg cells hyperplasia obviously accompanied by inflammatory cells infiltration [26] , which were observed in the liver of NDEA group rats. However, MAEs (75 mg・kg⁻¹ or 150 mg・kg⁻¹) and 150 mg・kg⁻¹ C3G administration regress these changes significantly (Figure 2(B)). The results of masson-stained sections showed that rats developed liver fibrosis after NDEA administration. However, the degree of

Data are means ± SDs or n (%).Values with different superscript letters (a, b, c, d) within cultivar are significantly different. *P < 0.05 and **P < 0.01 compared with NDEA group by Fisher’s exact test. ^#Visible nodules (diameter ≥ 1 mm). ^▲Max nodule diameter, mean maximum nodule diameter. ND, not detectable; AHF, altered hepatic foci; HA, hepatic adenoma; HCC, hepatocellular carcinoma.

fibrosis in NDEA group animals was more severe compared to rats treated with MAEs or C3G (Figure 2(C)). Cellular vacuolization was also reduced in the MWEs-treated groups (Figure 2(C)).

These results suggested that MAEs is an effective chemopreventive agent for preventing or delaying NDEA-induced hepatocarcinogenesis in rats, among which the high dose of MAEs has the best effect and is superior to C3G.

Excess reactive oxygen species produced in the processes of NDEA metabolic activation can interfere with the body’s oxidation system, leading to oxidative damage and tissue carcinogenesis, which plays an important role in the pathogenesis of liver cancer [26] . TBARS concentrations were used as markers of oxidative stress [24] . In this study, compared with the control group, the TBARS level of liver tissue in the NDEA group was remarkably increased (P < 0.001) 1.88-fold, which is consistent with earlier reports [13] [27] . Compared with the NDEA group, the hepatic tissue TBARS were significantly reduced in the C3G- 150, MAEs-75 and MAEs-150 groups by 39.01%, 40.85%, and 44.52%, respectively (Table 6, P < 0.001). Interestingly, these results showed that 150 mg・kg⁻¹ MAEs treatment almost restored the TBARS to the normal level. On this basis, we further detected the effect of MAEs on antioxidant capacity. The activities of antioxidant enzymes, GSH, GSH-Px, CAT and SOD in NDEA group were remarkably lower than that in the control group (P < 0.01), while these activities were dramatically promoted by MAEs and C3G (P < 0.05). Among which, the 150 mg・kg⁻¹ MAEs treatment rats had the highest levels of antioxidant enzymes in livers. After NDEA treatment, MDA content was obviously elevated in liver of NDEA group rats, which was approximately 2.44-fold higher than in control group rats (P < 0.01), and significantly decreased after MAEs treatment (Table 6, P < 0.05). Our results showed that NDEA administration changes the antioxidant status, leading to oxidative stress, Whereas MAEs can inhibit lipid peroxidation and enhance antioxidant capacity, so the prevention effect of MAEs against NDEA-induced hepatocarcinogenesis in rats may be closely related to its antioxidant activity. Chen et al. [13] [18] [27] also found that anthocyanins and water extracts from mulberry fruit can scavenge free radicals and shown concentration-dependent antioxidant activity in vitro.

NDEA’s induction increased microsomal phase I metabolizing enzymes with a

The data are the mean ± SD from 8 samples for each group and at least three independent measurements. Values with different superscript letters (a, b, c, d) within cultivar are significantly different. TBARS, thiobarbituric acid-reactive substances; GSH, Glutathione; GSH-Px, glutathione peroxidase; SOD, superoxide dismutase; CAT, catalase; MDA, malondialdehyde.

simultaneous decrease in the Phase II detoxifying enzyme [6] [28] . GST and UGT2b1 play an important role in the detoxification and excretion of toxins, carcinogens, which may actively modulated by Nrf2 and the antioxidant response element [29] . In this study, administration of NDEA, the significant reduction in UGT2b1 and GST activity in livers of NDEA group were observed (Figure 1(A), P < 0.05). Whereas the levels of GST and UGT2b1 in livers of MAEs-75 and MAEs-150 group rats were remarkably increased as compared to the NDEA group. What’s more, the hepatic GST and UGT2b1 levels of MAEs-150 group were obviously higher than that of C3G-150 group (P < 0.05), and closed to the normal level. The result of the qRT-PCR also showed that NDEA induced GST and UGT2b1 mRNA expressions and MAEs repressed these alterations (Figure 3(B)). These results suggested MAEs can inhibit tumor development by stimulating the activity of phase II detoxification enzyme.

Nrf2-mediated antioxidant, detoxification enzymes and anti-inflammatory signaling are through Nrf2-ARE pathways to protect organisms against cellular damage caused by oxidative stress [28] . Yan et al. [15] reported that consumption of mulberry anthocyanin extract maintained Nrf2, HO-1, and p38 MAPK stimulation that were involved in oxidative stress modulation. The results of this study shown that Nrf2 activation in livers of rats was induced by 150 mg・kg⁻¹ MAEs, accompanied by the stimulation the mRNA expressions of NQO-1, Keap1, and HO-1 (Figure 3(B), P < 0.05). Similar findings were observed in the measurements of Nrf2, Keap1, HO-1, and NQO-1 protein levels. Compared with the control and NDEA groups, treatment of 150 mg・kg⁻¹ MAEs caused significant changes in the protein expressions of Nrf2, Keap1, HO-1, and NQO-1 in the MAEs-150 group (P < 0.05), but less obvious changes were found in MAEs-75 group (Figure 3(C)). Combining these results, it was concluded that MAEs may activate the Nrf2 signaling pathway and induce Nrf2-mediated antioxidant enzymes, further activate the expression of downstream phase II detoxifying GST and UGT2b1, which then accelerated the elimination of the metabolites of NDEA, thereby achieving the goal of detoxification.

Studies showed that the main cause of NDEA-induced liver cancer is inducing chronic inflammatory response and abnormal repair after liver injury [30] [31] . Hassimotto et al. studies showed that administration of wild mulberry extract suppressed carrageenan-induced acute inflammation [7] . As shown in Fig 3A, relative to the NDEA group (85.26 pg・mL⁻¹, 97.60 pg・mL⁻¹), the IL-6 and TNF-α levels were remarkably inhibited by MAEs-75 (46.24 pg・mL⁻¹, 50.08 pg・mL⁻¹), MAEs-150 (37.93 pg・mL⁻¹, 36.75 pg・mL⁻¹), and C3G-150 (62.14 pg・mL⁻¹, 56.62 pg・mL⁻¹). The inhibition of IL-6 and TNF-α showed a dose-dependent relationship with MAEs (75 mg・kg⁻¹ and 150 mg・kg⁻¹). Besides, MAEs treatments

increased anti-inflammatory cytokines IL-10 and IFN-γ in serum after NDEA treatment (Figure 4(A)).

NF-κB, TNF-α, and COX-2 plays an important role in the development of inflammation [13] [24] . Thus, we further determined whether MAEs can reverse the effects on inflammation-related gene (TNF-α, NF-κB, and COX-2) expression by qRT-PCR and Western blot. The result of qRT-PCR analysis revealed

that, MAEs obviously decreased the TNF-α, NF-κB, and COX-2 mRNA expression in the livers of MAEs-75 and MAEs-150 group rats, all of which were remarkably higher than that of C3G treatment rats (Figure 4(B)). These results indicate that MAEs reduced NDEA-induced TNF-α, NF-κB, and COX-2 mRNA expression in a dose-dependent manner. Similar findings were observed in the measurements of the TNF-α, NF-κB, and COX-2 protein levels. Treatment with NDEA improved the protein expression levels of TNF-α, NF-κB, and COX-2 by 2.01-, 4.35-, and 2.81-fold in NDEA group when compared with the control group, respectively (Figure 4(C)). However, administration of MAEs significantly reduced the levels of TNF-α, NF-κB, and COX-2 proteins in MAEs-75 and MAEs-150 group rats (Figure 4(C), P < 0.05). Moreover, immunohistochemistry of NDEA-treated liver sections exhibited TNF-α, NF-κB, and COX-2 positive cells in NDEA group. The amount of TNF-α, NF-κB, and COX-2 positive cells in the liver, which was approximately 6.16-, 4.53-, and 15.92-fold higher in NDEA group rats than in control group rats (P < 0.001), reduced remarkably after MAEs (75 and 150 mg・kg⁻¹) treatment (Figure 5, P < 0.01).

The present study indicated that the NDEA-induced liver cancer may be related to the inflammatory reaction in rats, and MAEs may exert anti-inflammation by regulating the levels of inflammatory cytokines and attenuating the inflammatory

mediators COX-2 through NF-κB inhibition pathway.