Down-regulation of BMAL1 by MiR-494-3p Promotes Hepatocellular Carcinoma Growth and Metastasis by Increasing GPAM-mediated Lipid Biosynthesis

The circadian clock confers daily rhythmicity to many crucial biological processes and behaviors and its disruption is closely associated with carcinogenesis in several types of cancer. Brain and muscle arnt-like protein 1 (BMAL1) is a core circadian rhythm component in mammals and its dysregulation has been shown to contribute to aberrant metabolism in human diseases. However, the biological functions of BMAL1, especially its involvement in aberrant lipid metabolism in hepatocellular carcinoma (HCC), remain elusive. In the present study, we found that BMAL1 was frequently down-regulated in HCC cells mainly due to the up-regulation of miR-494-3p. Down-regulation of BMAL1 was significantly associated with poor survival in HCC patients. BMAL1 down-regulation promoted HCC cell growth and metastasis both in vitro and in vivo. Mechanistically, through cooperating with EZH2, BMAL1 transcriptionally suppressed the expression of glycerol-3-phosphate acyltransferase mitochondrial (GPAM), a key enzyme involved in the regulation of lipid biosynthesis, leading to reduced levels lysophosphatidic acid (LPA), which have long been known as mediator of oncogenesis. Particularly, treatment with SR8278, an activator of BMAL1, exhibited a therapeutic effect on HCC in vitro and in vivo. In conclusion, BMAL1 plays a critical anti-oncogenic role in HCC, providing strong research evidence for BMAL1 as a prospective target for HCC therapy.

Total proteins were purified from HCC cell lines or tissues as described previously [2]. Protein concentrations were determined via the BCA assay (Bio-Rad Laboratories). Equivalent protein amounts were resolved on SDS-polyacrylamide gels, electroblotted onto PVDF membranes and incubated overnight at 4 °C in the presence of specific primary antibodies. Relative protein expressions of target genes were determined after incubation at room temperature for 2 h in the presence of appropriate secondary horseradish-peroxidase-labeled antibodies. Visualization was performed via an enhanced chemiluminescence assay. The primary antibodies in this assay and their working concentrations are shown in Supplementary Table 2. H&E and immunohistochemistry staining H&E and immunohistochemistry staining were performed as previously described [3]. Immunostains were scored based on positive staining cell proportions and staining intensity [3]. Primary antibodies in this assay and their working concentrations are specified in Supplementary Table 2.

Flow cytometry analysis for cell cycle and apoptosis
For cell cycle analysis, HCC cells with different treatments were fixed in 70% ethanol at 4°C overnight, stained with propidium iodide (BestBio, shanghai, China) at room temperature for 30 min, and assessed by flow cytometry (Beckman, Fullerton, CA). For cell apoptosis analysis, the FITC-Annexin V and PI Apoptosis Kit (F6012, US EverbrightInc) was used. A total of 5 μl ANXA5-FITC and 5 μl PI were introduced into HCC cells and incubated at room temperature for 20 min in the dark.
Cells were rinsed thrice using PBS and analyzed by flow cytometry (Beckman, Fullerton, CA).

Wound-healing, cell migration and matrigel invasion assays
To determine cell migration abilities, a pipette tip was used to scratch the middle of wells when cells in the 6-well plates had achieved a 85% confluence. Imaging of wound-closures was performed using a light Olympus microscope at 0 and 48 h after scratching. Relative migrations of differentially treated HCC cells were determined using the Image J software. For matrigel invasion assessment, transwell chambers coated with the matrigel matrix (BD Science) were used. Briefly, 1 × 10 5 cells were loaded into the upper chamber of each well. After 48 h of incubation in a 5% CO 2 atmosphere at 37°C, cells that had invaded the lower chamber were fixed in 4% formaldehyde for 10 min and stained with 0.1% crystal violet for 10 min at room temperature. The number of invaded cells in each group was determined by light microscopy.

Reporter plasmids and site-directed mutagenesis
Promoter sequences of GPAM were abstracted from the UCSC Genome Browser.

Immunofluorescence
For double immunofluorescence staining of BMAL1 and EZH2, HCC cells were plated in the culture dish and incubated with primary BMAL1 and EZH2 antibodies at 4℃ overnight. After rinsing twice in PBS, they were respectively incubated with 488-labelled goat anti-rabbit or 594-labelled goat anti-mouse secondary antibodies.

ELISA for determination of lipid levels
The amounts of glycerolipid synthesis products (LPA, PA, DAG, TAG) in HCC cells were measured using the avidin biotin system (ABS) antibody sandwich ELISA method. The LPA, PA, DAG, and TAG kits (human, 96-well) were purchased from mlbio (shanghai). Based on the instructions, intracellular components of HCC cells (1×10 6 cells/ml) were extracted and quantitated by ELISA. Assays were performed as described [4,5].

In vivo tumorigenicity and metastatic assays
To assess the in vivo tumor growth, 1x10 7 HCC cells with varied treatments were subcutaneously administered into the flanks of four to five-week-old male BALB/c nude mice (n=6 per group). Then, tumor volumes were determined using a Vernier caliper every week. At 5 weeks after cell injections, mice were sacrificed, tumors For the in vivo metastatic assay, 5×10 6 HCC cells with varied treatments were intravenously administered into four to five-week-old male BALB/c nude mice via their tail veins (n=6 per group). Mice were sacrificed at two months after cell administration and their lungs collected for H&E staining. Metastatic tumor nodules in lungs were counted. SR8278 at dose of 0.5 mg/mice was administered into each mouse twice a week by intraperitoneal injection after operation.

Silencing and forced expressions of target genes
To transiently silence BMAL1, small interference RNAs (siRNAs) targeting BMAL1 were transfected into HCC cells using the lipofectamine 2000 reagent (Invitrogen) as per the manufacturer's instructions. A pSilencer™ 3.1-H1 puro vector (Ambion) was used to construct the shRNA targeting BMAL1. To construct the BMAL1 over-expression vector, the BMAL1 coding sequence was amplified and cloned into a pcDNA TM 3.1(C) vector (Invitrogen).

Luciferase assay
Luciferase assays were performed as previously described [1]. The HCC cells were co-transfected with 4 μg of GPAM promoter constructs and Renilla luciferase expressing control vector and grown for 48 h. Cells were lysed using the ice-cold lysis buffer after which luciferase enzyme activities were assessed by Dual Luciferase Reporter Assay (Promega, E1910), as instructed by the manufacturer. A Luminoscan Ascent MicroplateLuminometer (Thermo Scientific) was used to determine the relative light units. Luciferase activities were standardized to renilla luciferase enzyme activities.

Chromatin immunoprecipitation (ChIP)-PCR assay
The ChIP assay was performed using a ChIP assay kit (Cell Signaling, #9005), as per the manufacturers' protocol. The HCC cells were fixed in 1% formaldehyde and disrupted in the lysis buffer. Subsequently, samples were sonicated to degrade the nuclear membranes and the supernatants collected. Then, immune-precipitations of chromatin by an anti-BMAL1 antibody (1:100) or equivalent quantities of normal rabbit IgG were performed. DNA was isolated and PCR amplified using primer pairs within the modulatory site of GPAM, sequences of which are shown in Supplementary Table 1.

Co-immunoprecipitation (Co-IP) assay
For co-IP assays, HCC cells were lysed in ice-cold lysis buffer and incubated overnight at 4°C with 25 μl protein A beads (Santa Cruz) enriched with anti-BMAL1 (1:500) and anti-GPAM (1:300) antibodies. After rinsing 3 times using cold washing buffer, immunoprecipitated samples were eluted by heating in a loading buffer at 100 °C for 5 mins, followed by immunoblotting (WB) for analysis.

Statistical analysis
Data are shown as mean ± SEM. The SPSS software (17.0 version, Chicago, IL) was used for analyses, with p< 0.05 as the threshold for significance (*). Overall and recurrence-free survival curves were computed using the Kaplan-Meier approach.
Two-tailed student's t-test and one-way ANOVA followed by Tukey's post-hoc test were used for comparisons of means between and among groups, respectively.
Relationships between measured variables were determined by Spearman rank correlation analysis. Figure S1: Western blot analysis for BMAL1 expressions in 30 paired HCC and adjacent non-tumor tissues.       and Hep3B cells exposed to SR1078. (K and L) Scratch-wound-healing and matrigel invasion assays for MHCC97H and Hep3B cells treated with SR1078. Tables   Table S1: Sequences of primers, siRNAs and miRNAs used in this study.