Characteristics of a novel cell line ZJU-0430 established from human gallbladder carcinoma

This study was performed in accordance with the Declaration of Helsinki of 1975, and the official recommendations of Chinese Community Guidelines, and was approved by the Ethics Committee and Institutional Review Board of the Sir Run Run Hospital. Written informed consent was obtained from the patient.

A 74-year-old male patient with pain in the upper abdomen was admitted in our centre. Gastroscopy showed the presence of a gastric antral ulcer (stage S2), and irregular deep concave ulcer at the gastric angle, and cancer priority. Abdominal computed tomography (CT) also found a 1.5-cm thick wall at the bottom of the gallbladder. A radioimmunoassay showed that the patient’s serum levels of a variety of biomarkers were normal (CA19-9, CA-125, AFP, CEA, and PSA) except for ferritin, which was high (409.3 ng/ml, 30–400 ng/ml). A radical resection of stomach and gallbladder were performed and a pathological examination showed that gastric carcinoma was a poorly differentiated adenocarcinoma derived from signet-ring cell carcinoma, whereas the primary gallbladder cancer was a well differentiated adenocarcinoma.

GBC-SD cells were obtained from the Cell Bank of Type Culture Collection of Chinese Academy of Sciences (Shanghai, China), and maintained in Roswell Park Memorial Institute (RPMI) 1640 medium (Invitrogen, Carlsbad CA, USA) supplemented by 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA), and 1% penicillin/streptomycin and amphotericin B (Invitrogen).

The approach of primary culture was as previously described [28]. Single-cell suspensions were obtained according to the manufacturer’s protocol (Miltenyi Biotec GmbH, Germany). Specifically, the surgically resected gallbladder specimens were immediately transferred to the lab, rinsed several times in cold Dulbecco’s Phosphate Buffered Saline (DPBS) containing 1% penicillin/streptomycin and amphotericin B, cut into small pieces (1 mm³), and transferred into a gentleMACS™ C tube containing the buffer from the Tumor Dissociation Kit (human), and subjected to a gentleMACS™ Tissue Dissociator. Cells were cultured in complete growth RPMI 1640 medium (containing additives) at 37 °C in a humidified incubator containing an atmosphere of 5% CO₂. Fibroblasts were removed by differential trypsinisation as described previously [29, 30]. Serial passages were carried out every 3–4 days, routinely at a ratio of 1:2, and the medium was replaced when colour changed. The ZJU-0430 cell line was cultured for > 100 passages and showed no changes in its morphology.

ZJU-0430 cells were grown in mycoplasma-free complete growth RPMI 1640 medium without antibiotics. Cells were grown to 60% confluence, and washed with 4 ml of DPBS, and then treated with StemPro™ Accutase™ Cell Dissociation Reagent (Invitrogen). Cells were then centrifuged at 300 rcf for 30 s at room temperature and the supernatant removed without disturbing the cell pellet. Next, 0.04% bovine serum albumin fraction V (BSA) solution (Sigma-Aldrich, St. Louis, MO, USA) was added and then centrifugation at 300 rcf for 5 min. The resuspended cells were then filtered through 40 µm cell strainer (Corning Incorporated, Corning, NY, USA) and following Trypan Blue staining, the cell density and viability were determined using haemocytometer.

Single-cell RNA-seq libraries were prepared with Chromium Single cell 3′ Reagent v2 (or v3) Kits according to the manufacturer’s protocol. Single-cell suspensions were loaded on the Chromium Single Cell Controller Instrument (10 × Genomics) to generate single cell gel beads in emulsions (GEMs). In order to do this, ZJU-0430 single cells were suspended in DPBS containing 0.04% weight/volume BSA. About cells were added to each channel with a targeted cell recovery estimate of cells. After generation of GEMs, reverse transcription reactions were performed to generate barcoded full-length cDNA followed by the disruption of emulsions using the recovery agent and cDNA clean up with DynaBeads Myone Silane Beads (Thermo Fisher Scientific, MA, USA). The cDNA was then amplified by PCR with appropriate an appropriate number cycles which depending on the recovery cells. Subsequently, the amplified cDNA was fragmented, end-repaired, A-tailed, index adaptor ligated and library amplification. These libraries were sequenced on the Illumina sequencing platform (HiSeq X Ten) and 150 bp paired-end reads were generated.

The Cell Ranger software pipeline (version 2.2.0) provided by 10 × Genomics was used to demultiplex cellular barcodes, map reads to the genome and transcriptome using the STAR aligner, and down-sample reads as required to generate normalized aggregate data across samples, producing a matrix of gene counts versus cells. We processed the unique molecular identifier (UMI) count matrix using the R package Seurat (version 2.3.4). To remove low quality cells, we filtered those cells which with UMI larger than 8000 and gene lower than 2000. And also we discarded low-quality cells where > 20% of the counts belonged to mitochondrial genes. After applying these QC criteria, 11,083 single cells and 20,118 genes in total remained and were included in downstream analyses. Library size normalization was performed in Seurat on the filtered matrix to obtain the normalized count. Top variable genes across single cells were identified using the method described in Macosko et al. [31]. Principal component analysis (PCA) was performed to reduce the dimensionality on the log transformed gene-barcode matrices of top variable genes. Cells were clustered based on a graph-based clustering approach, and were visualized in 2-dimension using tSNE and UMAP. Likelihood ratio test that simultaneously test for changes in mean expression and in the percentage of expressed cells was used to identify significantly differentially expressed genes between clusters. Enriched biological function were analysis with DAVID website and fgsea packages in R.

All BALB/c nude mice experiments were carried out in accordance with Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines, and were approved by the Zhejiang University Animal Care and Use Committee (ZUACUC) (Project License 11627). To determine the oncogenicity of the ZJU-0430 cell line, cells were subcutaneously injected into the hind flanks of four nude mice (BALB/c nu; 4–6 weeks old) (SLAC Laboratory Animals Company, Shanghai, China), Mice were maintained in laminar flow cabinets under specific pathogen-free conditions. A total of 1 × 10⁶ cells were injected using a 27-gauge needle and the tumours monitored at intervals using digital callipers. After 3 weeks, mice was killed by cervical dislocation under sodium pentobarbital (45 mg/kg) anaesthesia. Tumours were removed, fixed in 10% formalin and subjected to pathological examination and IHC evaluation.

Both the morphology and ultrastructure of the ZJU-0430 cells were examined following seeding into 25-mm³ tissue-culture flasks. The general morphology of the cells was observed daily under a phase-contrast microscope and histopathologically compared to that of the original tumours. Electron microscopy was conducted as previously described [32] using the GBC-SD cell line as the control. Cells were harvested, pelleted, fixed (2.5% glutaraldehyde), post-fixing (2% osmium tetraoxide) and then embedded in EPON resin. Ultrathin sections were stained and examined under a HT7700 microscope (Hitachi, Tokyo, Japan).

The cell growth curve analysis was obtained, as described previously [33]. The ZJU-0430 cells and GBC-SD cells were plated at 5000 cells per well, cultured at 12 h intervals for 3 consecutive days, and cell growth determined using an MTT assay, respectively. The growth curve was plotted, and doubling time was calculated as described (http://​www.​doubling-time.​com/​index.​php).

ZJU-0430 and GBC-SD cells were plated, harvested, washed twice with cold DPBS, and fixed overnight in 70% ethanol at 4 °C, treated for 30 min with RNase A (2 μg/ml), and then with propidium iodide (PI) at room temperature at the dark. The cell cycle distribution was determined using a FACSCanto™ II flow cytometer (Becton–Dickinson, Mount View, VA, USA) and analysed using appropriate software.

Cells at the primary and 100th passage were used for analysis. Karyotyping was performed as described previously [34]. Briefly, cells were treated with 0.02 μg/ml Colcemid for 90 min at 37 °C until exponential proliferation was reached. The cell pellet was resuspended with potassium chloride (0.075 mol/l), incubated for 20 min at 37 °C, fixed in Carnoy’s solution (methanol: acetic acid = 3:1 by volume), and analyzed using trypsin G banding.

To verify that ZJU-0430 cell line was indeed derived from humans and had no-contamination, genomic DNA from ZJU-0430 cell lines was extracted using the Gentra Puregene Cell Kit (Qiagen, Inc., Valencia, CA) according to the manufacturer’s protocol. STR profiling was performed by amplifying eighteen loci simultaneously in a single tube and analyzing by capillary electrophoresis on an ABI Prism^® 3500 × 1 Genetic Analyzer. Data were matched with the American Type Culture Collection (ATCC) STR database (ATCC sales order number: SO0146233).

The cell-culture supernatant was collected and evaluated using the PCR Mycoplasma Test Kit (HuaAn Biotechnology, Hangzhou, Zhejiang, China), following the manufacturer’s instructions. PCR products were separated by electrophoresis in 1% agarose and documented by photography.

The primary and 100-passage cells were collected for flow cytometry analysis using a FACSCanto™ II flow cytometer (Becton–Dickinson, Mount View, VA, USA), respectively. The panel of monoclonal antibodies used included those specific for CD24, CD44, CD29, CD34, CD90, CD117, CD133, CD184, CD326, and CD338. All antibodies and reagents were purchased from BioLegend (San Diego, CA). Data were obtained using FCS Express Software (De Novo Software, Los Angeles, CA). Antigen expression was scored as positive based on a significant shift in staining in comparison to corresponding isotype control.

This procedure was conducted by the Cell Comb™ Scratch Assay according to the manufacturer’s instructions (Merck, Darmstadt, Germany). Briefly, a wound was created in the cell monolayer using cell comb, when the cells reached semi-confluence, and they were cultured thereafter for an additional 72 h at 37 °C and 5% CO₂. Eight scratched fields were randomly chosen for cell counting.

The in vitro migration and invasion assays were performed as previously described [35, 36]. Briefly, 1 × 10⁴ cells were seeded into cell-culture insert wells (Corning Incorporated, Corning, NY, USA) with or without Matrigel-coating (BD Biosciences, Bedford, MA, USA) in RPMI 1640 medium. The extent of migration and invasion was determined after cultured for 48 h. The cells on the lower surface were fixed and stained with Giemsa solution.

A colony formation assay was used to assess the proliferative ability of the ZJU-0430 cells. A single-cell suspension was prepared and plated on 6-well plates (200 cells/well) in triplicate and cultured for 14 days. The colonies were stained with 1% crystal violet for 30 s after fixation with 4% paraformaldehyde for 5 min.

Tissue section slides from original patient (GBC and gastric cancer), ZJU-0430 tumours grown in mice, and cytospins of ZJU-0430 cells were used in this experiment as described previously [37, 38]. For cytospins, ZJU-0430 cells were plated onto glass coverslips, fixed with 4% paraformaldehyde (PFA), blocked with 10% goat serum, incubated with primary antibodies (vimentin, CK-pan, CK7, and CK19 all from cell signaling technology (CST)) at 4 °C overnight, followed by the addition of immunofluorescent labelled antibodies [FITC/PE goat anti-rabbit IgG (H + L)]. For tissue slides, 5 μm sections were fixed onto the slides, dewaxed, rehydrated, heated for antigen retrieval, incubated in 3% hydrogen peroxide to inactivate endogenous peroxidase, blocked with 5% BSA, and then incubated with primary antibodies specific for the following proteins, MUC1, MUC2, MUC4, MUC5AC, MUC6, AFP, Hepatocyte, Glypican-3, GS, CK20, CAD17, CDX2, beta-catenin, STAB 2, vimentin, MMP-2, MMP-9, CK7, and CK19. A 3,3′ diaminobenzidine tetrahydrochloride (DAB) horseradish peroxidase colour development kit (Boster) was used for visualizing target protein and Mayer’s haematoxylin (Boster) was used for counterstaining.

A routine pathological examination of the original tumour was performed and compared to the similar observations made in the established cell lines. Haematoxylin and eosin (H&E) staining of the primary GBC from the patient’s surgical specimen showed well-differentiated adenocarcinoma polygonal cells that formed ribbons, nests, and sheets (Fig. 1a), characteristic of a classical GBC morphology. The gastric cancer showed a poorly differentiated adenocarcinoma derived from signet-ring cell carcinoma with clusters, beams, or scattered infiltration, with few cytoplasm, nuclear deviation, with obvious atypia and mitosis. The proliferation of interstitial fibrous tissue and capillaries is obvious (Fig. 1b). The established cell lines grew in adherent monolayer pattern with multiple forms; there was no change in the morphology during passage (Fig. 1c). Under the electron microscope, ZJU-0430 cells showed typical epithelial properties, including the presence of numerous organelles (ribosomes, rough endoplasmic reticula, mitochondria), large irregular nuclei with deep sunken caryotheca, microvillus-like projections on the cell surface, and numerous desmosomes forming intercellular connections. In contrast to ZJU-0430 cells, the GBC-SD cells had fewer microvillus-like projections on the cell surface, had sunken caryotheca, and had desmosomes that formed intercellular connections (Fig. 1d).

As showed in Fig. 2a, the ZJU-0430 cell line had a moderately abundant strongly basophilic cytoplasm. The nuclei were ovoid with coarse chromatin and occasional irregular contours. The ZJU-0430 cell line and the original tumour were staining for CK-pan, vimentin, CK7, and CK19. The data showed that there was CK-pan positive staining in both the ZJU-0430 cell line and in the tumour cells in the original tissue. The ZJU-0430 cell line was negative for vimentin expression whereas the mesenchymal cells were positive for vimentin expression (Fig. 2b, c). Both the ZJU-0430 cell line, as well tumour cells in the original tissue, were positive for CK7 and CK19 (Fig. 2d, e). These results confirmed that the ZJU-0430 cell lines were originated from gallbladder epithelium.

Uphoff and Drexler reported that 1–35% cells are prone to mycoplasma contamination in primary early passages and in continuous cell culture [39]. In order to assess if the ZJU-0430 cell line suffered from mycoplasma, mycoplasma contamination was assessed using a PCR assay. As shown in Fig. 3, no bands were detected in samples from the ZJU-0430 cell line, indicating that it is free of mycoplasma contamination.

STR profiles were prepared to confirmed the species of origin of the JU-0430 cell line. Analysis of the ZJU-0430 cell line STR profiles compare to the ATCC and DSMZ databases revealed it to be of human origin. The data also excluded the possibility of cross-contamination with other cell lines, including the HeLa cell line (Table 1, Additional files 1, 2).

The ZJU-0430 cell line was characterized for its ability to proliferate and progress through the cell cycle, the GBC-SD cell lines was used as a control (Fig. 4a). The PDT of the ZJU-0430 cells and the GBC-SD cells were approximately 19.81 h and 33.5 h, respectively. The ZJU-0430 cell line grew shorter than the GBC-SD cell line on the condition of initial cell concentration.

The DNA contents of the ZJU-0430 cell line and GBC-SD cell lines were analysed by flow cytometry. ZJU-0430 cells had higher PI (**P < 0.05) and SPF (**P < 0.05) than GBC-SD cells (Fig. 4b), and therefore have a higher proliferative potential compared with the GBC-SD cells.

Chromosomal abnormalities such as number and structure abnormalities are frequently found as cancer progresses. A cytogenetic analysis of the ZJU-0430 cell line at the primary stage and P100 were performed using banding technology. The results showed both numerical and structural aberrations at different passages. A complex near-triploid clone with a modal number of chromosomes of 73 showed a representative karyotype categorized as: 70–74, XX, − X, + 4, − 5, der (6) t (6: ?) (P22; ?), + 7, − 10, + 14, − 15, + 17, der (17) t (5;17) (p13: p11.2) × 2, − 18, − 19, + 20, + I (21) (q10), +22 × 2, + mar 1, + mar 2 [CP15] (Fig. 4c, e) at P100. The karyotype of primary cells was also investigated, and the results were 69–76, XX, + 1, + 2 × 2, − 5, + 3 × 2, + 4 × 2, + der (5) t (5;17) (q13; q11.2?) × 2, der (6) t (6; ?) (p23; ?), + 7, + 8 × 2, + 9 × 2, + 10, + 11, + 12 × 2, + 13, + 14 × 2, − 16, + 18, + 20, + 21 × 2, + 22 × 2, mar1 × 2, mar2 × 2, [CP5] (Fig. 4c, e). The symbol “?” represent chromosome classification and condition are unknown. These results revealed that ZJU-0430 chromosomal abnormalities included gains, losses, translocations and other events.

In order to investigate the stemness of the ZJU-0430 cells, an immunophenotypic analysis of the ZJU-0430 cells was performed by flow cytometry. The immunophenotypes of the primary and 100th passage cells were almost identical. The ZJU-0430 cells were positive with CD24, CD44, CD29, and CD133, partially positive for CD184, and CD326; and negative for CD34, CD90, CD117, and CD338 (Table 2, and Additional file 3: Figure S1). The result show that the ZJU-0430 cells have partial stemness.

In order to investigate the ability of the ZJU-0430 cells to migrate, invade, and proliferate, wound-healing and Transwell (with or without Matrigel) assays were performed. As shown in Fig. 5a, although the scratch wound on ZJU-0430 cells was almost closed after 18 h culture, the scratch wound distance of the GBC-SD cells did not change after 72 h culture. As shown in Fig. 5b, the number of ZJU-0430 cells that invaded the basement membrane (with or without Matrigel) was significantly higher than that of GBC-SD cells.

A colony formation was also performed to examine the ability of ZJU-0430 cells proliferate. As shown in Fig. 5c, ZJU-0430 cells form more compromised tumours than GBC-SD cells.

In order to investigate their in vivo tumorigenic capacity, ZJU-0430 cell suspensions were injected subcutaneously into the flank of nude mice. Tumour development was observed in all of mice and no obvious metastatic lesions were observed. Gross examination revealed that the tumours were round or oval shaped, and often showed mucus production. The atypia of tumour cells was obvious, when arranged with a small strip of cord and viewed under a microscopy they formed different cavities that included foam cells, cancer cells and normal epithelial cells (Fig. 5d, e).

To characterize the original ZJU-0430 cell lines, the original GBC, and the original gastric cancer, a series of markers were examined by IHC technology. As shown in Fig. 6a, ZJU-0430 xenografts of nude mice were positive for the expression of CK-pan, CK7, CK19, MUC5AC, MUC6, and MMP-2, and weakly expressed GS and CK20, there was no expression of AFP, Hepatocyte, Glypican-3, CAD17, CDX2, β-catenin, SATB2, MUC2, vimentin, and MMP-9. In addition, the original GBC tumour showed negatively staining CDX2 (Fig. 6b), and the original gastric cancer showed positive staining for CK20, CAD17, CDX2, and was negative for β-catenin and SATB2. These results confirm the ZJU-0430 cells originate from the original GBC (Additional file 4: Figure S2).

To interpret the characteristic observed during the experimental investigation, we do single cell RNA sequencing analysis on ZJU-0430 cell line. By using Seurat pipeline, we identified 3 main cell clusters (Fig. 7a, Additional file 5: Figure S3a, b). And cell cycle effects did not contribute for the main heterogeneity as cells in different mitotic cycle presented in all 3 clusters (Fig. 7b). All those cell cluster were CD24, CD44, CD29 and CD133 positive and with partial of the cells in the C3 cluster were CD326 negative (Additional file 5: Figure S3c). To expound the characteristic of each cell cluster, we did DEG analysis between each group and using the DEGs for DAVID enrichment analysis. C1 cluster highly expressed genes associated with regulation of Wnt signaling pathway (Fig. 7c, Additional file 5: Figure S3d) and cellular component of extracellular exosome (Additional file 5: Figure S3e). C2 cluster enriched genes involved with apoptotic process (Fig. 7d, Additional file 5: Figure S3f). And C3 cluster expressed less genes and transcripts, which might due to nuclear-transcribed mRNA catabolic process associated with nonsense-mediated decay (Fig. 7e, Additional file 5: Figure S3g). These results indicated that the ZJU-0430 mainly constituted by cells with stemness signature and no too much heterogeneity.