All protocols for animal care and use were approved by the Animal Care Committee of the Shanghai Cancer Institute and were in accordance with regulations for the Administration of Affairs Concerning Experimental Animals and National Institutes of Health Guidelines. All of the mice were housed in pathogen-free (SPF) animal facilities under a standard 12-h-light/12-h-dark cycle. Animals received free access to water and commercial mouse chow throughout the present study. Mice were sacrificed by cervical dislocation.

Hepatocyte-specific Lass2 KO mice were generated using the Cre/LoxP system for site-specific excisional DNA recombination. The background of the transgenic mice was C57BL/6J, and mice were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). Mice carrying the floxed second exon of Lass2 (Lass2 LoxP⁺/LoxP⁻) were obtained from the Shanghai Research Centre for Model Organisms. Albumin-Cre transgenic mice expressing Cre recombinase under control of the hepatocyte-specific albumin promoter were purchased from the Jackson Laboratory. Lass2 LoxP⁺/LoxP⁻ were crossbred with Albumin-Cre transgenic mice to generate heterozygous transgenic mice, and the siblings were crossbred with each other to generate Lass2 LoxP⁺/LoxP⁺ and Alb-Cre⁺ mice. The DNA of the mouse tail was extracted using the QIAamp DNA Blood kit (Qiagen Inc., Germany), and genotypes were identified by PCR assay. PCR was performed using the Takara Taq system (Takara Bio, Inc., Otsu, Japan) using sense, 5′-GGCTTC GTGTTGGTCTTCTGA-3′ and antisense, 5′-ATTCCCTGGC ATCCACCTTTC-3′ for Lass2 LoxP primer pairs; and sense, 5′-GCGGTCTGGCAGTAAAAACTATC-3′ and antisense, 5′-GTGAAACAGCATTGCTGTCACTT-3′ for Cre primer pairs. Lass2 LoxP⁺/LoxP⁺ and Alb-Cre⁺ mice were used as Lass2 KO mice. Wild-type control was Lass2 LoxP⁻/LoxP⁻ and Alb-Cre⁺ mice. When mice were sacrificed after DEN treatment at week 23 or 40, the genotypes of the brain, liver and kidney tissues were identified by PCR assay with the paired primers as sense, 5′-TATGTAGCCAGAGTCCAAGGC-3′ and antisense, 5′-CACTATTGACCAGGCGAGGAG-3′, western blotting and anti-Lass2 immunohistology (IHC) in order to confirm the hepatocyte-specific deletion of Lass2.

Groups of 14-day-old Lass2 KO and WT mice were administered a single intraperitoneal injection of the genotoxic hepatocarcinogen DEN (Sigma) dissolved in PBS at a dose of 20 mg/kg body weight or saline (no DEN). Mice were sacrificed respectively at week 23 and 40 following DEN or saline injections. The livers were weighed, the diameter of tumours ≥1 mm on the surface of the livers were enumerated and the sizes of tumours were measured. Part of the liver tissues was fixed in AAF (100% alcohol 85 ml, acetic acid 5 ml, formalin 10 ml) for histological study, and the remaining part was frozen at -80°C until use.

The fixed liver, kidney and brain tissues were embedded in paraffin, sectioned into 4-μm slices and anti-Lass2 immunohistochemical staining was performed to ascertain hepatocyte-specific deletion of Lass2. The liver sections were H&E-stained for examination of morphological alteration after DEN treatment. Moreover, IHC staining of proliferating cell nuclear antigen (PCNA) and 5-ethynyl-2′-deoxyuridine (EdU) staining were performed for examining the proliferation, whereas terminal deoxynucleotidyl-transferase-mediated dUTP-biotin nick end-labeling (TUNEL) assay was carried out for assessment of apoptosis in the DEN-treated livers.

For IHC staining, the standard protocol was performed. Briefly, sections were incubated with 3% H₂O₂ to eliminate the endogenous peroxidase activity, then blocked with 5% bovine serum albumin (BSA), and then incubated with rabbit anti-Lass2 polyclonal antibody (1:100) (Beijing Biosynthesis Biotechnology Co., Ltd., Beijing, China) or rabbit anti-PCNA polyclonal antibody (1:100) (ProSci, Inc., Poway, CA, USA) at 4°C overnight, respectively. The sections were washed 3 times with PBS for 5 min each. For the Lass2 assay, the sections were incubated with Alexa Fluor^® 488-conjugated goat anti-rabbit antibody (1:100) (Invitrogen Life Technologies, Carlsbad, CA, USA). Images were captured using a Leica DM LB2 epifluorescence microscope. However, for PCNA detection, the sections were incubated with HRP-conjugated mouse anti-rabbit IgG (1:100) (ProSci, Inc.) for 1 h at 37°C. Color was developed with DAB (KeyGen Biotech, Nanjing, China), and microscopic examination was performed with a bright field microscope (Motic BA210; Xiamen, China).

EdU assay was performed on mouse liver section using the EdU DNA Imaging kit (RiboBio, Guangzhou, China) according to the manufacturer’s instructions. Briefly, 5 mg/kg EdU was intraperitoneally injected into mice 8 h before sacrifice. The mice were sacrificed under ether anaesthesia, and the liver sections were developed as indicated above. The slices were washed in a shaker with glycine (2 mg/ml) for 10 min and then incubated with 0.5% Triton X-100 for 10 min. Following washing with PBS for 2 times, the sections were incubated with Apollo reaction buffer for 10–30 min. Finally, after being washed with 0.5% Triton X-100 for 3 times, methanol twice and PBS for 1 time in due order, the slices were stained with 100 μl Hoechst 33342 (x1) for 30 min at room temperature (RT). Images were obtained using a Leica DM LB2 epifluorescence microscope. The EdU labeling index of the liver tissues was calculated as a percentage of EdU-positive cells to the total of Hoechst 33342-stained cells in each field, counted in 5 random ×100 fields in 5 representative livers from each group.

Cell apoptosis in liver tissues was determined using a TUNEL kit (Beyotime Institute of Biotechnology, Jiangsu, China) according to the manufacturer’s recommendations. Briefly, paraffin-embedded tissue sections were deparaffinized and rehydrated as above. The slides were treated with protease K (1:500, 20 μg/ml) for 15 min at 37°C, washed 3 times with PBS, incubated in TUNEL reaction buffer (Tdt enzyme 2 μl, fluorescent buffer 48 μl) for 60 min at 37°C while protecting from light, and again washed 3 times with PBS, then stained with 100 μl Hoechst 33342 (x1) for 30 min at room temperature, and washed with PBS twice. Slides were observed under a Leica DM LB2 epifluorescence microscope. The apoptotic cell index was expressed as the average of five random areas of a single ×100 field in a specific individual liver tissue, with 5 mice/group and 1 section/mouse.

Tissues were lysed in RIPA lysis buffer containing the protease inhibitor phenylmethanesulfonyl fluoride (both from Beyotime Institute of Biotechnology). The cell extracts were centrifuged at 12,000 × g for 20 min at 4°C, and the supernatants were used for experiments. The protein concentrations were determined with the BCA assay kit (Beyotime Institute of Biotechnology). The equivalent tissue proteins (10 μg/lane) were subjected to electrophoresis and transferred onto PVDF membranes (Millipore, Bedford, MA, USA) via a semi-dry transfer system (Bio-Rad, Hercules, CA, USA). The membranes were blocked with 5% non-fat milk for 1 h at room temperature in TBST (50 mM pH 7.5 Tris, 0.9% NaCl and 0.1% Tween-20) and then incubated with a 1:200 dilution of the primary antibodies (anti-PAI-1, anti-AFP and anti-Smad-4 were purchased from Bioworld Technology, Inc.; anti-Lass2 was from Beijing Biosynthesis Biotechnology Co., Ltd; and anti-β-actin from Santa Cruz Biotechnology, Inc.) overnight at 4°C. The membrane was washed then incubated with secondary antibody for 1 h at room temperature and developed using BeyoECL Plus kit (Beyotime Institute of Biotechnology) for 1 min, then scanned by ChampChemi™ professional (Beijing Sage Creation Science Co., Ltd., Beijing, China).

Total RNA from different liver tissues including the non-DEN-treated, DEN-treated non-tumour region and tumour region from the male Lass2 KO and WT mice (3 mice/group) at week 40 were, respectively, extracted using TRIzol (Invitrogen Life Technologies) and subjected to reverse-transcription using the miScript Reverse Transcription kit (Takara Bio, Inc.) to synthesize cDNA, and comparative quantitative PCR (QT-PCR) was preformed using SYBR-Green (Takara Bio, Inc.) chemistry analysis in quadruplicates. Relative expression was calculated using the comparative threshold cycle (Ct) method. The primers for AFP, PAI-1, TGF-β1, Smad-4, Smad-7, GAPDH are shown in Table I.

All data are presented as means ± standard error of the mean (SEM). Unpaired Student’s t-test was used for statistical analysis. P≤0.05 was considered to indicate a statistically significant result. All QT-PCR and western blot analyses were repeated at least 3 times. All statistical analyses were conducted using the GraphPad Prism 5.

Hepatocyte-specific Lass2 KO mice were produced using the Cre-LoxP system. Lass2-ABRLFn-pBR322 vector carrying 2 LoxP cassettes was inserted into the intron 1–2 and intron 2–3, respectively. The excision of Lass2 was the result of the recombination event by one Cre recombinase enzyme controlled by the upstream Alb-promoter on the Lass2-ABRLFn-pBR322 which was able to identify the sequence between 2 LoxP cassettes (Fig. 1A). Mice carrying a floxed Lass2 allele were mated with Alb-Cre transgenic mice that express the Cre recombinase in hepatocytes. The siblings that carried both floxed Lass2 allele and Alb-Cre were selected by the genotyping of tail DNA and were named as Lass2 KO mice (Fig. 1B, lane 2). To confer hepatocyte-specific Lass2 deletion, PCR assay and anti-Lass2 immunohistology of the liver, kidney and brain tissues were carried out. In Lass2 KO mice, liver tissue displayed the Lass2-knockout band whereas kidney and brain tissues did not (Fig. 1C). The phenotype was verified by western blotting, and the results showed the absence of the Lass2 protein from the liver tissues of Lass2 KO mice when compared with the control mice; no significant difference in brain and kidney tissues was noted between the 2 groups (Fig. 1D). Immunohistology demonstrated that Lass2 signals were undetectable in the livers from Lass2 KO mice whereas they were visible in kidney and brain tissues (Fig. 1E).

The liver carcinogenesis of mice was induced using the classical carcinogen DEN. In the well-established model of DEN-induced liver carcinogenesis, the development of tumour is usually monitored approximately 38–40 weeks following the treatment of DEN. We hypothesized that the loss of Lass2 may accelerate the formation of tumours. Therefore, the mice were sacrificed at two time points: week 23 and 40, respectively, after DEN or saline treatment. Following DEN injection, at week 23, notably, all the KO mice, either male or female, developed tumours in the livers, whereas only 2 out of 7 males and 1 out of 7 females in the WT group developed tumours. The difference in the tumour occurrence between the Lass2 KO and WT group was highly statistically significant (P<0.001). At week 40 following DEN treatment, although almost all mice from both groups developed tumours and no significant difference in tumour occurrence was noted between the 2 groups (Table II), in Lass2 KO group, the indices of liver weight/body weight, the numbers of tumours on the surface of livers and the average sizes of tumours were higher than these values in the control group (Fig. 2A and B). We also found a greater elevation in the AFP level in the Lass2 KO group (both in the tumour and non-tumour regions) when compared with its counterpart at week 40 (Fig. 2C). Histologic analysis of H&E-stained liver sections from Lass2 KO and WT mice revealed that the tumours were hepatocellular adenomas/carcinomas, showing basophilic foci characterized by clearly defined margins, compression of the surrounding parenchyma, thicker trabeculae extensive fatty degeneration and Mallory bodies inside of the tumours (Fig. 2D). Unequivocal evidence of metastases or vascular invasion was not found. The results indicated that the deletion of Lass2 resulted in the earlier occurrence of DEN-induced liver carcinogenesis and accelerated tumour growth.

In our previous study, we found that overexpression of Lass2 in a human liver cancer cell line inhibited proliferation of the cells in vitro (6). In the present study, we also found that the average size of tumours in the Lass2 KO group was larger than that of the control group. Therefore, we detected the cell proliferation of liver tissues in male mice at week 23 with EdU assay (Fig. 3A), the expression of cell proliferation marker PCNA with IHC staining (Fig. 3B), and apoptosis with TUNEL assay (Fig. 3C). Quantification revealed an ~8-fold increase in the EdU-positive ratio (P<0.001) and an ~12-fold increase in the PCNA-positive ratio (P<0.001) in Lass2 KO mice when compared with the control. Moreover, TUNEL staining demonstrated that the average TUNEL-positive ratio was dramatically decreased in the liver tissues of Lass2 KO mice when compared with the control (P<0.001) (Fig. 3D). The results suggest that the deletion of Lass2 promoted the proliferation and decreased the apoptosis of the liver tissues.

PAI-1 is a tumour-promoting gene predominantly produced by hepatocytes in liver tissue. We used QT-PCR, western blotting and IHC to compare the expression of PAI-1 in the liver tissues of the Lass2 KO and control mice at week 40 with or without treatment of DEN, including respective tumour regions (TR) and non-tumour regions (NTR). The results indicated, in each paired liver tissues from Lass2 KO and WT mice, markedly increased mRNA and protein levels of PAI-1 in the Lass2 KO mice when compared with the WT counterparts. Without DEN treatment, mRNA levels of PAI-1 increased ~7-fold. After DEN-treatment, before transformation (i.e. in NTR), the mRNA level of PAI-1 in the Lass2 KO liver increased even more than that in the WT counterpart, with an increase of ~20-fold; whereas after transformation, i.e., in the TR, this increase was decreased ~9-fold (Fig. 4A). The alteration of the protein level of PAI-1 was in accordance with that of the mRNA level (Fig. 4B).

We used QT-PCR to detect the mRNA levels of TGF-β1 and Smad-4 and -7, genes reported to be closely related with PAI-1, in Lass2 KO and WT liver tissues. Smad-4 expression was also assessed by western blot analysis. The results demonstrated that without DEN treatment, the deletion of Lass2 caused increases in TGF-β1, Smad-4 and -7, which were parallel to the levels of PAI-1. Following DEN treatment, in each Lass2 KO and WT paired specimen, mRNA levels of TGF-β1 and Smad-4 in the Lass2 KO specimen (including NTR and TR), respectively, were increased when compared to the WT; whereas regarding the mRNA level of Smad-7, a significant decrease was noted in Lass2 KO NTR, and no significant difference was noted in Lass2 KO TR, which exhibited a different pattern of alteration to TGF-β1 and Smad-4 (Fig. 5A–C). Western blotting of Smad-4 showed a similar trend with the result of QT-PCR analysis (Fig. 5D). Taken together, in each pair of Lass2 KO and WT specimens, the TGF-β1, Smad-4 and PAI-1 from the Lass2 KO group was consistently and respectively significantly increased when compared to the WT group, suggesting that upregulation of the TGF-β1-Smad4-PAI-1 occurred in the Lass2 KO mice. It was noted that mRNA levels of TGF-β1, Smad-4 and -7, after DEN induction when compared with these levels before DEN treatment, decreased in several orders of magnitude, the significance of which is yet to be known.