Acyclic retinoid and angiotensin-II receptor blocker exert a combined protective effect against diethylnitrosamine-induced hepatocarcinogenesis in diabetic OLETF rats

Eight-week-old male diabetic Otsuka Long-Evans Tokushima Fatty (OLETF) rats and littermate Long-Evans Tokushima Otsuka (LETO) rats were procured from the Otsuka Pharmaceutical Co. (Tokushima, Japan). The rats were housed in stainless steel mesh cages under controlled conditions (temperature: 23 °C ± 3 °C, relative humidity: 50% ± 20%, 10–15 air changes/h, illumination: 12 h/d). The rats were permitted ad libitum access to tap water throughout the study period. Peretinoin (NIK-333), as an ACR, was provided by the Kowa Company Ltd. (Nagoya, Japan), and losartan potassium, as an ARB, was purchased from Merck Ltd. (Tokyo, Japan).

The experimental design has been presented in Fig. 1a. The OLETF rats were administered intraperitoneal injection of 200 mg/kg diethylnitrosamine (DEN) (Tokyo Kasei Kogyo Co., Ltd., Tokyo, Japan) and were divided into the following four treatment groups (n = 10 each): vehicle (soybean oil), ACR (peretinoin, 40 mg/kg), ARB (losartan, 30 mg/kg), and ACR (40 mg/kg) + ARB (30 mg/kg), administered through daily oral gavage. As a negative control, LETO rats (n = 10) were also injected with DEN and treated with the vehicle. All experimental rats received 2/3 partially hepatectomies 3 weeks after the DEN injection for the promotion of hepatocarcinogenesis as described previously [9, 22]. All the rats were sacrificed 9 wk. after DEN injection. At the end of the experiment, the rats were anesthetized, their abdominal cavities were opened, blood samples were drawn via an aortic puncture, and livers were harvested for histological findings. Serum biological markers were assessed using routine laboratory methods. IR and insulin sensitivity were evaluated using the Homeostasis model assessment-Insulin Resistance (HOMA-IR) and the Quantitative Insulin Sensitivity Check Index (QUICKI), respectively, as described previously [9]. All the animal procedures were performed as per the recommendations of the Guide for Care and Use of Laboratory Animals (National Research Council). The study was approved by the animal facility committee of Nara Medical University.

In order to determine the intrahepatic triglyceride content, the liver tissue homogenates were extracted using 2:1 (vol/vol) chloroform/methanol. Chloroform/methanol was added to the homogenate, and the mixture was then shaken for 15 min. Following centrifugation at 14,000 rpm for 10 min, the organic layer was collected. This extraction was repeated thrice. The collected sample was dried and resuspended in 1% Triton X-100/ethanol. The measurement was conducted using Triglyceride E-test Wako (Wako Pure Chemical Industries, Ltd., Osaka, Japan). The hepatic 4-Hydroxynonenal (4-HNE) content was measured using the 4 Hydroxynonenal ELISA Kit (MyBioSource, San Diego, CA, USA) as per the manufacturer’s protocol.

For determining the liver 8-hydroxy-2′-deoxyguanosine (8-OHdG) levels, DNA was extracted from the frozen liver tissue using a DNeasy mini kit (Qiagen, Tokyo, Japan) and purified as described previously [23]. The 8-OHdG level in the extracted DNA solution was determined using the highly sensitive 8-OHdG check ELISA kit (Nikken Seil Co., Ltd.). The 4-HNE and 8-OHdG levels were indicated as a ratio to those of the negative control group.

The liver sections were fixed with 10% formalin and embedded in paraffin. Then, 5-μm paraffin sections were stained with hematoxylin and eosin. For detecting the pre-neoplastic lesions, immunohistostaining was performed using anti-placental glutathione S-transferase (GST-P) polyclonal antibody (Medical & Biological Laboratories Co., Nagoya, Japan). Semi-quantitative analyses of the size of the GST-P positive pre-neoplastic lesions were conducted using the NIH ImageJ software (http://​imagej.​nih.​gov/​ij/​).

Two human liver cancer cell lines Huh7 (JCRB0403) and HLE (JCRB0404) and human umbilical vascular endothelial cells (HUVECs) (IFO50271) were obtained from the Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan). Cells were cultured in Dulbecco’s modified Eagle’s medium (Thermo Fisher Scientific K.K., Kanagawa, Japan) supplemented with 10% fetal bovine serum and 1% ampicillin/streptomycin.

Cell proliferation was assessed using water-soluble tetrazolium salt − 1 assay as per the manufacturer’s protocol. Initially, for the validation of the inducible effects of insulin on cell proliferation, the cells were treated with different dilutions of recombinant human insulin (rh-insulin) (0, 5, 10, 20, 30, and 40 μU/mL) (Gibco, Tokyo, Japan). Thereafter, we compared the effects of ACR (peretinoin, 10 μM), ARB (losartan, 10 μM), and the combined treatment (each 10 μM) on cell proliferation of Huh7, HLE, and HUVECs that were cultured under high concentrations of D-(+)-glucose (Nacalai Tesque, Kyoto, Japan) and rh-insulin. The cells were seeded on 96-well uncoated plastic dishes at a density of 5 × 10⁴ cells/mL, and each treatment was started 12 h after seeding and incubated for another 24 h. Three independent experiments were performed; the average values were calculated from five replicates of each experiment.

Total RNA was extracted from the cultured human liver cancer cells, HUVECs, or frozen liver tissues using the RNeasy mini kit (QIAGEN, Tokyo, Japan) as per the manufacturer’s instructions. Total RNA from each sample was reverse transcribed into cDNA using the high-capacity RNA-to-cDNA kit (Applied Biosystems Inc., Foster City, Calif., USA) as per the manufacturer’s instructions. The mRNA expression was assessed with quantitative real-time PCR using the StepOnePlus real-time PCR system (Applied Biosystems, Foster City, USA) using Fast SYBR Green master mix (Applied Biosystems). Relative gene expression was measured with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal control. The relative amount of target mRNA in each cycle was evaluated by applying a threshold cycle to the standard curve. The mRNA levels were indicated as a ratio to the negative control. All primers were purchased from Sigma Aldorich (Additional file 1: Table S1).

Whole cell lysates were prepared from 10⁶ cultured human liver cancer cells using T-PER Tissue Protein Extraction Reagent supplemented with proteinase and phosphatase inhibitors (all Thermo Scientific, Rockford, IL, USA). Fifty micrograms of whole cell lysates were separated by SDS-PAGE; thereafter, they were transferred to a PVDF membrane that was subsequently blocked with 5% bovine serum albumin in Tris-buffered saline + Tween–20 for 1 h. Thereafter, each membrane was incubated overnight with an antibody against p44/42 mitogen-activated protein kinase (MAPK) [extracellular signal-regulated kinase (ERK) 1/2], Phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204), Caspase-3, Cleaved caspase-3, and β-Actin (Cell Signaling Technology); washed; and incubated with Amersham ECL IgG and HRP-linked F(ab)2 fragment (GE Healthcare Life Sciences, Piscataway, NJ, USA; 1:5000 dilution). Finally, each membrane was developed using Clarity Western ECL Substrate (BIORAD, Hercules, CA, USA).

In vitro angiogenesis was defined as the formation of capillary-like structures in HUVECs co-cultures with human diploid fibroblasts, as previously described [9]. In brief, 24-well tissue culture plates were coated with Matrigel (Collaborative Biochemical Products, Bedford, MA, USA; 150 mL/well) and allowed to set at 37 °C for 30 min. Thereafter, 2.5 × 10⁴ HUVECs were added to each well and incubated at 37 °C for 20 h under 5% CO₂ atmosphere and different conditions. Semi-quantitative analysis of tubule formation was performed using NIH ImageJ software (http://​imagej.​nih.​gov/​ij/​).

The data were subjected to Student’s t-test or one-way analysis of variance followed by Bonferroni’s multiple-comparison test, as appropriate. Bartlett’s test was used for determining the homology of variance. Statistical analyses were performed using Prism, version 6.04 (GraphPad Software, La Jolla, CA, USA). All tests were two-tailed, and P-values < 0.05 were considered statistically significant.

All the rats were survived until the end of experiments. Physical and serological data for all the experimental groups of OLETF and LETO rats are presented in Table 1.

According to the underlying characteristics, OLETF rats exhibited obesity, hypertriglyceridemia, hyperglycemia, and hyperinsulinemia along with mild elevation of aspartate and alanine aminotransferase (AST and ALT) as compared to the LETO rats. In the OLETF rats, HOMA-IR was notably higher. Correspondingly, QUICKI, a surrogate marker of insulin sensitivity, was lower than that in the LETO rats (Fig. 1b). Meanwhile, there was no significant difference between the two groups in terms of the serum albumin and total bilirubin levels. Among the OLETF groups, treatment of ACR and/or ARB reduced hypertriglyceridemia; however, the body weight, liver function, and glycemic status remained unchanged (Table 1 and Fig. 1b).

Initially, in order to determine whether IR stimulates the augmentation of hepatocarcinogenesis, we conducted a histological comparison of the number and size of the DEN-induced GST-P positive pre-neoplastic lesions (GST-P⁺ PNL) between the LETO and OLETF rats (Fig. 1c). Semi-quantitative analyses revealed that both the number and size of hepatic GST-P⁺ PNL in the OLETF rats were higher than those in the LETO rats, indicating that IR promotes hepatocarcinogenesis (Fig. 1d). Then, we evaluated the effects of ACR and/or ARB on the IR-induced exacerbation of hepatocarcinogenesis. In non-diabetic LETO rats, either ACR or ARB significantly decreased the number and size of hepatic GST-P⁺ PNL in accordance with the previous reports (Additional file 2: Figure S1a, b). Interestingly, these suppressive effects were observed in diabetic OLETF rats with greater magnitude than those in LETO rats (Fig. 1c, d). Moreover, the individual administration of ACR or ARB demonstrated weaker effects than the combined administration of ACR and ARB (Fig. 1c, d).

Several reports have shown that ACR can inhibit HCC cell growth by increasing the cellular levels of p21^CIP1 and concomitantly decreasing the cyclin D1 levels [24]. Thus, we evaluated the mRNA expressions of these molecules in the liver of DEN-treated rats during tumorigenesis. Quantitative real-time RT-PCR analysis indicated that OLETF rats showed decreased Cdkn1a and increased Ccnd1 mRNA levels compared to LETO rats; further, these altered expressions were inhibited by ACR treatment (Fig. 2a). In contrast, ARB treatment had no influence on these mRNA levels. Subsequently, we assessed the effects of ACR and ARB on in vitro cell growth in human liver cancer cells under both normal and IR condition. First, we examined the effects of both agents on the proliferation of two human liver cancer lines, Huh7 and HLE, under the normal culture condition. ACR inhibited cell proliferation of both of two lines, while ARB did not (Additional file 3: Figure S2a). Next, to evaluate the effect on cell growth under IR condition, we examined the inducible activity of rh-insulin on cell proliferation for optimizing the IR-mimetric condition. As shown in Fig. 2b, rh-insulin induced cell proliferation of two human liver cancer lines in a dose-dependent manner. Thereafter, we generated IR-mimetic conditional media by modulating higher concentrations of D-(+)-glucose (199.6 mg/dl) and rh-insulin (33.2 μU/ml), almost corresponding to the serum level in the OLETF rats for evaluating the effects of ACR and ARB on cell proliferation under the IR condition. It is noteworthy that ACR treatment considerably inhibited IR-stimulated cell proliferation of both the liver cancer lines (Fig. 2c). Consistent with the in vivo results, IR condition stimulated the downregulation of CDKN1A and upregulation of CCDN1. These alteration in mRNA levels were inhibited by ACR treatment (Fig. 2d). Moreover, ACR suppressed ERK1/2 phosphorylation and facilitated caspase-3 cleavage in both human liver cancer lines (Fig. 2e). These in vitro findings support that ACR directly inhibits cell proliferation by inducing cell cycle arrest and apoptosis in liver cancer cells under the IR condition. There was no difference in the degree of inhibitory effect of ACR on liver cancer cell growth between normal culture condition and IR condition (Fig. 2c and Additional file 3: Figure S2a). As in the DEN-treated liver tissue, ARB treatment did not directly influence cell growth or survival in liver cancer lines even under the IR condition.

Considerable evidence suggests that neovascularization plays a crucial role in hepatocarcinogenesis, especially in the presence of IR; therefore, we then investigated angiogenesis in the liver tissue of DEN-treated rats [9]. Quantitative real-time RT-PCR analysis revealed that mRNA levels of Vegfa and Pecam1, representative molecules associated with angiogenesis, were raised in the liver of OLETF rats (Fig. 3a). ACR or ARB treatment suppressed the upregulation of Vegfa and Pecam1 mRNA levels, while the combination treatment exerted a much stronger suppressive effect than each monotherapy (Fig. 3a). We further evaluated the in vitro cell growth of the vascular endothelial cell lines, HUVEC, to validate the effects of both the agents on the angiogenic activity. Similar to the experiment for liver cancer cells, we first assessed the effects of ACR and ARB on the proliferation of HUVEC cultured under normal condition. As shown in Additional file 3: Figure S2b, both ACR and ARB suppressed HUVEC proliferation. We next confirmed that rh-insulin could induce HUVEC proliferation in a dose-dependent manner (Fig. 3b). IR-mimetic condition markedly promoted the proliferation of HUVECs and liver cancer cells. Further, both ACR and ARB inhibited IR-induced HUVEC proliferation with single treatment (Fig. 3c). Comparing the magnitude of these antiproliferative effects of ACR and ARB, there was no difference between normal culture condition and IR condition. It is noteworthy that the combination treatment exerted a more potent antiproliferative effect on HUVECs (Fig. 3c). These effects coincided with the downregulation of VEGFA and PECAM1 mRNA levels in HUVEC cultured under the IR-mimic condition (Fig. 3d). Moreover, we examined whether both agents affect endothelial tubular formation. As shown in Fig. 3e, ARB strongly inhibited tubular formation, and ACR significantly suppressed EC tubular formation; further, combined treatment with ACR and ARB showed additive inhibitory effect on tubular formation compared with each single treatment. Thus, an anticancer effect of ACR and ARB is at least partially involved in their antiangiogenic activities.

As reactive oxygen species are critical for the development of hepatocarcinogenesis, we assessed the effect of ACR and ARB on lipid peroxidation and oxidative DNA damage. First, we measured the hepatic content of triglycerides in the DEN-treated rats. The hepatic triglyceride level was higher in the OLETF rats compared to that in the LETO rats and was lowered following ACR treatment (Fig. 4a). Similarly, ARB treatment also decreased the triglyceride levels in the liver of the OLETF rats, while combined treatment showed greater reduction than each single treatment (Fig. 4a). Corresponding to the decrease in hepatic triglycerides, the levels of hepatic 4-HNE, a lipid peroxidation marker, were lowered in ACR- and/or ARB- treated OLETF rats (Fig. 4b). Thereafter, we evaluated the level of hepatic 8-OHdG, an indicator of oxidative DNA damage. As shown in Fig. 4c, hepatic 8-OHdG level was lower in ACR-treated OLETF rats than in the vehicle-treated rats, while there was no significant change in the ARB-treated rats. In addition to oxidative stress, an inflammatory state based on IR and lipid accumulation also plays a crucial role in obesity-associated hepatocarcinogenesis. Thus, we further examined the effects of ACR and ARB on the levels of the proinflammatory cytokines, Tnfa, Il6, Il1b, Ccl2, and Serpine1, as well as the mRNA levels in the liver of DEN-treated OLETF rats. Both treatments (ACR and ARB) lowered the hepatic mRNA levels of Tnfa, Il6, and Il1b (Fig. 4d, f). The hepatic expressions of these proinflammatory cytokines were lower following combined treatment than after each single treatment (Fig. 4d, f). Nevertheless, neither agent altered the hepatic mRNA levels of Ccl2 and Serpine1 (Fig. 4g, h).