Single tumor cell preparation and sphere-forming culture
Human primary cells were separated with a Tumor Cell Isolation Kit, Human (Miltenyi Biotec, #130-095-929). The enzyme mix (2.2 ml of DMEM, 100 µl of Enzyme H, 50 µl of Enzyme R, 12.5 µl of Enzyme A) and 0.05–0.2 g of tumor samples with the fat, fibrous areas and necrotic areas removed were prepared. The tumor was cut into small pieces of 2–4 mm. The tissue pieces were transferred into a gentleMACS C tube containing the enzyme mix. Dissociation was initiated by running the gentleMACS program 37C_h_TDK_3. The sample was resuspended, and the cell suspension was filtered through a MACS SmartStrainer (40 µm or 70 µm) placed on a 50 ml tube. The cells and MACS SmartStrainer (40 µm or 70 µm) were washed with 20 ml of DMEM. The mouse primary cells were separated with a Tumor Cell Isolation Kit, Mouse (Miltenyi Biotec, 130-110-187). The enzyme mix (2.35 ml of DMEM, 100 µl of Enzyme D, 50 µl of Enzyme R, 12.5 µl of Enzyme A) and 0.04–1 g of tumor samples with the fat, fibrous areas and necrotic areas removed were prepared. The next steps were the same as described above. The cell suspension was centrifuged at 300g for 7 min and resuspended in DMEM/F12.
Cells were suspended in serum-free advanced DMEM/F12 supplemented with 100 IU/ml penicillin, 100 µg/ml streptomycin, 20 ng/ml human recombinant epidermal growth factor (hrEGF), 20 ng/ml human recombinant basic fibroblast growth factor (hrbFGF), 1% nonessential amino acids, 1% GlutaMAX, 2% B27 supplement (Invitrogen, USA), and 1% N2 supplement (Invitrogen, Carlsbad, CA, USA). Cells were subsequently cultured in ultra-low attachment 6-well plates (Corning Inc., Corning, NY, USA) at a density of no more than 5,000 cells/well.
LDA
Single-cell suspensions of primary HCC cells from the DEN model in Scarb2CreERT2-EGFP mice were sorted to obtain Scarb2EGFP+ or Scarb2EGFP- cells, which were then injected into BALB/c nude mice. Single-cell suspensions of primary HCC cells from the spontaneous HCC model in CreAlbMyc, CreAlbScarb2F/+Myc or CreAlbScarb2F/FMyc mice were sorted to obtain HCC cells, which were then injected into BALB/c nude mice. Single-cell suspensions of HCC cells from subcutaneous HCC tumors were sorted to obtain HCC cells, which were then injected into BALB/c nude mice. The outgrowths were analyzed at 8 weeks post transplantation. The tumorigenicity of the transplanted cell suspension was calculated using an extreme limiting dilution assay (ELDA).
Mouse models of tumor growth and metastasis
To evaluate the effect of combined sorafenib treatment and SCARB2 knockout on the inhibition of tumor growth and metastasis, BALB/c nude mice (5 to 6 weeks old, male) were earmarked before grouping and were then randomly separated into four groups (n = 8 per group) in a blinded manner by an independent person. The tumor growth model in BALB/c-nude mice was established by subcutaneous (s.c.) injection of 2×106 HepG2 cells or HepG2 SCARB2cas9 cells diluted with 100 μl of Matrigel (Corning, 354230) at a 1:1 ratio. Tumor growth was monitored, and the tumor size was measured using calipers. To establish the mouse model of experimental metastasis, BALB/c-nude mice were injected via the tail vein with 2 ×106 HCCLM3 cells or HCCLM3 SCARB2cas9 cells. To evaluate the survival rate, these mice were monitored for more than 3 months. The number of lung metastases was determined.
Immunoprecipitation, immunoblotting, immunostaining and HCC tissue microarray analysis
Co-IP was performed as described previously. Cells were harvested and lysed for 30 min in co-IP buffer supplemented with a complete protease inhibitor (Cell Signaling Technology, 5817), 1 mM trichostatin A (Selleck, S1045), and 5 mM nicotinamide (MedChemExpress, HYB0150) on ice. Centrifugation was performed to obtain the supernatant, which was then incubated first with the indicated antibodies at 4°C overnight, and then with Protein A/G Plus-Agarose (Santa Cruz Biotechnology, TX, USA) at 4°C for 2 hr. Soluble lysates were incubated with the indicated anti-Myc magnetic beads (Bimake.com, B26302), anti-Flag affinity gel (Bimake.com, B23102) or anti-HA affinity gel (Bimake.com, B23302) at 4°C overnight. Complexes were eluted from the beads and were then boiled for 10 min. The precipitated proteins were subjected to SDS-PAGE and immunoblotting with the corresponding antibodies. For immunoblot analysis, proteins were extracted from cells and liver tissues using RIPA buffer (Cell Signaling Technology, MA, USA). The protein concentrations were determined with a BCA Protein Assay Kit. Protein extracts were separated by SDS-PAGE, transferred onto PVDF membranes, and subjected to immunoblot analysis. Images of the Western blots were acquired with a Tanon 5200 chemiluminescent imaging system (Tanon, Shanghai, Beijing).
For immunofluorescence 42 and colocalization assays, cells were seeded in 12-well plates and processed differently according to the experimental requirements. Next, the cells were briefly washed with PBS and fixed with 3.7% formaldehyde in PBS for 10 min, washed three times with PBS and permeabilized with 0.5% Triton X-100 for 15 min. Then, the cells were washed with PBS three times and blocked with 3% bovine serum albumin (BSA) for 1 hr at 37°C. Samples were incubated with primary antibodies overnight at 4°C and with secondary antibodies for 2 hr. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) in blocking buffer. Images were acquired using a confocal microscope (Olympus Microsystems, CA, USA). Quantitative image analysis was performed with Imaris 9.3.1 software. The Pearson correlation coefficient was used to analyze colocalization between two target proteins.
For immunohistochemical analysis, tissue sections were deparaffinized in xylene and rehydrated through a graded alcohol series and distilled water. Antigen retrieval was carried out in a microwave with citrate buffer (10 mM sodium citrate buffer, pH 6.0) at a subboiling temperature for 15 min. Sections were permeabilized with 0.5% Triton X-100 in PBS for 20 min. Endogenous peroxidase activity was blocked with 3% H2O2 solution for 10 min, and the sections were then washed three times with PBS. Blocking buffer (3% BSA/PBS) was added to the sections and incubated for 30 min. The sections were then incubated with the indicated primary antibodies at 4°C overnight. After washing three times, the sections were incubated for 30 min with the corresponding secondary antibodies at room temperature. Signals were detected with freshly made DAB substrate solution (ZSGB-BIO Company, Beijing, China). The sections were then counterstained with hematoxylin, dehydrated, and mounted with coverslips. Images were acquired using an Olympus DP72 microscope (Olympus Microsystems, CA, USA) and analyzed with Image-Pro Plus 5.1.
For immunoblotting, the following antibodies were used: anti-GAPDH (ZSGB-BIO TA-08, 1:2000), anti-MYC (D3N8F) (CST, #13987S, 1:1000), anti-MYC (CST, #9402S, 1:1000), anti-Max (S20) (CST, #4739, 1:1000), anti-acetylated lysine (CST, #9441, 1:1000), anti-HDAC3 (7G6C5) (CST, #3949, 1:1000), anti-SCARB2 (Abcam, ab176317, 1:1000), anti-phospho-MYC (Thr58) (E4Z2K) (CST, #46650, 1:1000), anti-phospho-MYC (Ser62) (E1J4K) (CST, #13748, 1:1000), anti-GCN5 (Abcam, ab217876, 1:1000), anti-p44/42 MAPK (Erk1/2) (CST, #4695, 1:1000), anti-CDK7 (CST, #2090, 1:1000), anti-BRD4 (CST, #13440, 1:1000), anti-KAT5 (Abcam, ab151432, 1:1000), anti-Cyclin A2 (E1D9T) (CST, #91500S, 1:1000), anti-Cyclin D (CST, #2922, 1:1000), anti-Cyclin E1 (D7T3U) (CST, #20808, 1:1000), anti-p300 (D8Z4E) (CST, #86377, 1:1000), anti-CDK4 (D9G3E) (CST, #12790, 1:1000), anti- N-Cadherin (D4R1H) (CST, #13116, 1:1000), anti-eIF3A (D51F4) (CST, #3411, 1:1000), anti-eIF4E (CST, #9742, 1:1000), anti-IRP2 (D6E6W) (CST, #37135, 1:1000), anti-phospho-IRS-1 (Ser1101) (CST, #2385, 1:1000), anti-PKM1/2 (CST, #3186, 1:1000), anti-Myc-tag (MBL, #562, 1:1000), anti-GFP-tag (MBL, #598, 1:1000), anti-DDDDK-tag (MBL, PM020, 1:1000), anti-HA-tag (MBL, #561, 1:1000), and anti-His-tag (MBL, PM032, 1:1000). For immunofluorescence and immunohistochemistry, the following antibodies were used: anti-MYC (R&D, AF3696, 1:100), anti-MYC (R&D, MAB3696, 1:100), anti-MYC (Novus, NB600-302, 1:100), anti-SCARB2 (Thermo Fisher, PA5-20540, 1:100), anti-HDAC3 (Novus, NB500-126, 1:100), anti-GFP (Abcam, ab13970, 1:100), anti-RFP (Abcam, ab62341, 1:100), anti-HNF-4-alpha (Abcam, ab199431, 1:100), anti-Albumin (Abcam, ab207327, 1:100), Alexa Fluor 488 (Thermo Fisher, R37114, 1:200), Alexa Fluor 488 (Thermo Fisher, R37118, 1:200), Alexa Fluor 488 (Abcam, ab150173), Alexa Fluor 555 (Thermo Fisher, A-31572,1:200), Alexa Fluor 555 (Thermo Fisher, A-31570, 1:200), Alexa Fluor 555 (Thermo Fisher, A-21432, 1:200), Alexa Fluor 647 (Thermo Fisher, A-31571, 1:200), and Alexa Fluor 647 (Thermo Fisher, A-31573, 1:200). For ChIP-seq, the following antibody was used: anti-MYC (D3N8F) (CST, #13987S, 10 μg/ChIP).
Flow cytometry
Primary cell suspensions from liver tissue with or without Scarb2 deletion were directly labeled. Fluorescently labeled antibodies against the following surface proteins were used for mouse cell staining: APC anti-mouse F4/80 (Cat. 123116, BioLegend), APC/Cyanine7 anti-mouse CD45 (Cat. 103116, BioLegend), PE anti-mouse CD31 (Cat. 102407, BioLegend), Percp/Cyanine5.5 anti-mouse CD68 (Cat. 137010, BioLegend), PE anti-mouse CD24 (101807, Biolegend), PE/Cyanine7 anti-mouse CD133 (Cat. 141209, BioLegend) and APC anti-mouse Ep-CAM (Cat. 118214, BioLegend). Then, cells were washed, and data were acquired using a FACSCanto II flow cytometer 40 and analyzed with FCS EXPRESS software.
Cell Counting Kit-8 (CCK-8) assay
For the CCK-8 assay, HepG2 and HCCLM3 cells with or without SCARB2 deletion were seeded at a density of 1000 cells/well in 96-well plates. Cells were cultured for 1, 2, 3, 4 or 5 days. To generate the time effect curve of sorafenib treatment, CTRLcas9 and SCARB2cas9 cells were seeded in 96-well plates at a density of 2000 cells/well. Vehicle (DMSO, Merck, D2650) or sorafenib (15 μM)-treated cells were cultured for 1, 2, 3, 4 or 5 days. To generate the dose effect curve of sorafenib treatment, vehicle (DMSO)- or sorafenib (2, 4, 8, 16, 32, 62.5, 125, 250, 500 μM)-treated cells were cultured for 24 hr. To evaluate the antiproliferative effects of combined treatment with PMB and sorafenib, HepG2, H22, Hepa 1-6, Huh7, Hep3B and HCCLM3 cells were seeded in 96-well plates at a density of 10000 cells/well. Cells were treated with sorafenib (15 μM) and an appropriate concentration of inhibitor and cultured for 24 hr. Subsequently, 10 μl of CCK-8 solution (Dojindo, CK04) was added to each well, and the plates were incubated at 37°C for 2 hr. Finally, the absorbance was measured at 450 nm.
Invasion assay
Transwell assays were performed using Millicell inserts (8.0 μm, Millipore, Billerica, MA, USA) to evaluate cell invasion. Millicell inserts were precoated with 10 μg/ml fibronectin and Matrigel (1:8; BD Bioscience, Bedford, MA, USA) and then allowed to dry. Millicell inserts were placed into the wells of a 24-well plate containing culture medium supplemented with 10% FBS. Cells (5×104 cells/well) were starved overnight and were then seeded in the upper chambers without FBS culture medium. 12 hours later, the migrated cells were fixed with paraformaldehyde, stained with 0.1% crystal violet and counted.
Real-time PCR and RNA interference
Total RNA was extracted using TRIzol (Invitrogen, CA, USA) according to the manufacturer’s instructions. Total cellular RNA was reverse transcribed using oligo(dT) primers and M-MLV reverse transcriptase (Transgen Biotech, Beijing, China). PCR was performed using a MyCycler thermal cycler and analyzed using LineGene 9600. The PCR primer sequences were as follows: Gapdh forward, 5′-CATCACTGCCACCCAGAAGACTG-3′; Gapdh reverse, 5′-ATGCCAGTGAGCTTCCCGTTCAG-3′; Tfap4 forward, 5′-GACGCGAGATTGCCAACAGCAA-3′; Tfap4 reverse, 5′-TGCTGTCTGCTGGAGAATGGCT-3′; Pld6 forward, 5′-TCTGCCTCTTCGCCTTCTCCAG-3′; Pld6 reverse, 5′-GTAGTCGCAGTCAGTGATGACC-3′; Glut4 forward, 5′-GGTGTGGTCAATACGGTCTTCAC-3′; Glut4 reverse, 5′-AGCAGAGCCACGGTCATCAAGA-3′; Glut1 forward, 5′-GCTTCTCCAACTGGACCTCAAAC-3′; Glut1 reverse, 5′-ACGAGGAGCACCGTGAAGATGA-3′; Cdk4 forward, 5′-CATACCTGGACAAAGCACCTCC-3′; Cdk4 reverse, 5′-GAATGTTCTCTGGCTTCAGGTCC-3′; Ccne1 forward, 5′-AAGCCCTCTGACCATTGTGTCC-3′; Ccne1 reverse, 5′-CTAAGCAGCCAACATCCAGGAC-3′; Ccnd1 forward, 5′-GCAGAAGGAGATTGTGCCATCC-3′; Ccnd1 reverse, 5′-AGGAAGCGGTCCAGGTAGTTCA-3′; Ccna1 forward, 5′-GCTACTGAGGATGGAGCATCTG-3′; and Ccna1 reverse, 5′-CAGCTTCCAGAAGGCTCAGTTC-3′. TIP60 siRNA (sc-37967), GCN5 siRNA (sc-37947), and p300 siRNA (sc-29432) were purchased from Santa Cruz Biotechnology.
ChIP-seq
ChIP assays were performed according to the manufacturer’s protocol using a SimpleChIP®Plus Sonication Chromatin IP Kit (Cell Signaling Technology, Danvers, MA, USA, #56383). In brief, HCCLM3 cells were fixated with 1% formaldehyde (Sigma-Aldrich, F8775) for 8 min, incubated with glycine (50 mM final concentration) for 10 min and washed three times with PBS. After cell lysis and chromatin extraction, chromatin was sonicated into fragments of 100–500 bp using a BioRuptor sonicator (Diagenode) and was then centrifuged at 16,000 × g for 10 min at 4°C. To control for experimental variation, our spike-in data were normalized to a spike-in control consisting of a small amount of chromatin from another species. In brief, H22 cell lysates were sonicated exactly as described above and added to HCCLM3 cells at a 1:100 ratio. The spiked lysates were incubated overnight at 4°C with a ChIP-grade antibody specific for MYC (CST, #13987S, 10 mg/ChIP), which was coupled to magnetic beads. The precipitated material was eluted (input chromatin was used as a control), crosslinking was reversed, and DNA was purified by chloroform/phenol extraction and resuspended in DNA elution buffer. ChIP-seq libraries were generated by sequential DNA fragmentation, end repair, dA tailing, adaptor ligation and PCR amplification. A Qubit® 3.0 A fluorometer was used for quantitation of the ChIP library. An Agilent 2100 Bioanalyzer was used to determine the insert sizes of the library. A StepOnePlusTM Real-Time PCR system was used to assess the molality of the library, which was required to exceed 10 mM for sequencing.
ChIP-seq data analysis
ChIP libraries were sequenced on the Illumina HiSeq platform with 150 bp paired-end reads. After quality control, the clean reads were aligned to the human reference genome (Homo sapiens GRCh38.87v2) with Bowtie v2.1.0. Unambiguously mapped reads were retained for subsequent generation of binding profiles, heatmap visualization and peak calling. Model-based Analysis of ChIP-seq (MACS2) version 2 was used to identify regions in ChIP samples in which the signal was enriched over the background signal from the corresponding input sample, and a P value of 10^−9 was used as the cutoff to identify statistically significant peaks. Mapped reads were visualized using Integrative Genomics Viewer (IGV). We next used the murine spike-in to quantitatively normalize different samples. In brief, the total number of reads aligned to the mouse mm9 genome assembly was used as a normalization factor to scale ChIP-seq data sets produced from equal numbers of cells. Bamliquidator (https://github.com/BradnerLab-/pipeline/wiki/bamliquidator, version 1.0) was used to calculate the ChIP-seq read density over a given genomic coordinate. Heatmaps were generated and genome-wide correlation analyses were performed using deepTools2. To visualize density distributions around transcription start sites (TSSs) or ChIP peaks and heatmaps indicating MYC occupancy, plotHeatmap was used. Peaks were annotated using the ‘closestBed’ tool in the BEDTools suite v2.20.1.
Generation of the MYC K148 mutation with the CRISPR/Cas system
The sgRNAs for introducing the MYC K148R knock-in mutation were annealed and ligated into the YKO vector. The gRNA sequences targeting MYC K148R were gRNA1: 5’-CAGCTTCTCTGAGACGAGCTTGG-3’ and gRNA2: 5’-TTTGCGCGCAGCCTGGTAGGAGG. The repair template designed with a homologous genomic flanking sequence was TCATCATCCAGGACTGTATGTGGAGCGGCTTCTCGGCCGCCGCGAAGCTCGTCTCAGAGAGGCTGGCGTCCTACCAGGCTGCGCGCAAAGACAGCGGCAGCCCGAACCCCGCCCGCGGCCA. 1 mg of each sgRNA plasmid was mixed with 1 mg of donor plasmid for transfection into HCCLM3 cells with Lipofectamine 3000 Reagent (Invitrogen, L3000008) according to the manufacturer's instructions. Twelve hours after transfection, HCCLM3 cells were treated with 5 mg/ml puromycin for 24 hr. Then, the puromycin was removed from the cell culture medium, and the cultured cells were sorted into 96-well plates at 1 cell/well. The cells were incubated and expanded for 2–3 weeks, and all clones were further subjected to genomic DNA extraction (TIANamp Genomic DNA Kit, DP304), PCR amplification of the MYC sequence and Sanger sequencing. The sequences of the PCR primers were as follows: MYC-K148R forward, 5'-CTCGTCTCAGAGAGGCTGGCCTCCT-3' and MYC-K148R reverse, 5'-AGGAGGCCAGCCTCTCTGAGACGAG-3'. The correct K148R knock-in cell clones were selected for further experiments.
Structured Illumination Microscopy (SIM)
Cells (~20,000) were grown on coverslips and fixed with formalin after the indicated treatment. Subsequently, the cells were washed in Tris-buffered saline (TBS) and permeabilized for 5 min with 0.5% Triton X-100 in TBS. The reaction was quenched for 10 min with 50 nM glycine in TBS, and blocking was performed for 30 min with 5% normal mouse serum or normal goat serum in a 0.2% gelatin-TBS solution depending on the isotype of the secondary antibody. Subsequently, 30-μl droplets containing the indicated dilutions of antibodies specific for MYC (R&D, AF3696, 1:100), SCARB2 (Thermo Fisher, PA5-20540, 1:100) or HDAC3 (Novus, NB500-126, 1:100) were placed on Parafilm in a dark humidified chamber. Coverslips were placed face down on the droplets and incubated overnight at 4°C. The next day, the coverslips were lifted by adding a small volume (200 μl) of TBS under the coverslip. After washing 3 x 20 min with 1 ml of 0.2% gelatin-TBS, the coverslips were incubated with Alexa Fluor 488 (Thermo Fisher, R37114, 1:200) and Alexa Fluor 555 (Thermo Fisher, A-31572, 1:200) secondary antibodies as described for primary antibody incubation for 1 hr at room temperature. After washing 3 x 10 min with 1 ml of 0.2% TBS-gelatin and 1 wash with regular TBS, the coverslips were mounted using soft-set mounting medium with DAPI (Vectashield) and sealed with nail polish. Images were acquired using a confocal microscope (Olympus Microsystems, CA, USA) or a GE Healthcare 3D structured illumination microscope. Intensity plots of individual pixels taken from a straight line in the indicated immunofluorescence images were generated by the twin slicer tool in Imaris 3D & 4D imaging software (Bitplane AG, Switzerland). Images were cropped and processed in Adobe Photoshop. When comparisons were made between images from the same experiment, all levels were adjusted equally, and the ratio between the levels was not altered.
GSEA
We ranked genes by their association with CreAlbScarb2F/FMyc mice vs. CreAlbMyc mice and HepG2 cells vs. PMB treated HepG2 cells using the signal-to-noise metric determined by GSEA according to the log2 fold change values. MYC target gene sets were obtained from a database (http://software.broadinstitute.org/gsea/msigdb/index.jsp).
Mass spectrometry analysis
Whole-cell lysates of HepG2 cells were immunoprecipitated with an anti-MYC antibody (CST, 9402S) or IgG1 isotype control antibody (CST, 5415) and Protein A/G Plus-Agarose (Santa Cruz Biotechnology) overnight at 4°C. Interaction complexes were eluted from the beads by heating at 98°C for 10 min. Mass spectrometric data analysis was performed by Beijing Qinglian Biotech Co., Ltd.
Surface plasmon resonance analysis
A BIAcore S200 system (GE Healthcare) was used to analyze the surface plasmon resonance binding kinetics between SCARB2 and the indicated small molecules. In brief, SCARB2 protein (Sino Biological, China) was immobilized onto channel 2 in a CM5 sensor chip (GE Healthcare) through a standard coupling protocol. To measure the binding kinetics, the indicated small molecule (100~0.098 nM in 2-fold serial dilutions) and a buffer blank for baseline subtraction were sequentially injected, with a regeneration step (glycine, pH 2.5) performed between each cycle. The equilibrium dissociation constant was calculated with BIA evaluation software.
Mouse models to evaluate tumor growth and combination treatment with sorafenib and PMB
To evaluate the effect of combination treatment with sorafenib and PMB (polymyxin B sulfate) on tumor growth, BALB/c nude or NSG mice (5 to 6 weeks old, male) were earmarked before grouping and were then randomly separated into four groups (n=8 per group) in a blinded manner by an independent person. To establish the tumor growth model in BALB/c nude mice, 2 × 106 HepG2 or HCCLM3 cells per mouse diluted with 100 μl of Matrigel (Corning, 354230) at a 1:1 ratio were subcutaneously injected into the right flanks of nude mice. To establish the PDX model, tumor masses (0.1 cm3) were subcutaneously transplanted into NSG mice. Tumors were allowed to grow for 7 days, and mice were then administered vehicle (Kolliphor® HS 15, i.g., once a day), sorafenib (30 mg/kg, i.g., once a day), PMB (25 mg/kg, i.g., once a day), or PMB plus sorafenib for 14 days. During the indicated treatments, tumor burden was evaluated by measuring tumor volumes to determine the therapeutic efficacy in the mouse models and PDX models.
Statistics & Reproducibility
No data were excluded from the analyses. For in vivo cancer studies, mice were randomly assigned into experimental groups. The Investigators were not blinded to allocation during experiments and outcome assessment. Unpaired Student’s t test in SPSS 13.0 software was used to compare parameters between two groups. One-way analysis of variance (ANOVA) was used to compare parameters among more than two groups. Pearson correlation analysis was used to determine correlations between groups. The Kaplan–Meier method was used to analyze survival. All data are expressed as the mean ± standard error of the mean (SEM) values. Generally, all experiments were carried out with n≥3 biological replicates. All statistical tests were two-tailed. P < 0.05 was considered statistically significant. All the data presented has been reviewed by a statistician to ensure scientific rigor and diligence of our data. Further information on research design is available in the Nature Research Reporting Summary linked to this article.