Selection of patients and 2DOs

Eight patients with distant metastases or recurrences that could be evaluated using computed tomography were selected. All patients underwent baseline imaging within 4 weeks before anticancer drug administration. The tumor volume and reduction rate were measured according to RECIST guidelines42. 2DOs were established from primary CRC specimens and cultured according to a previous report20 and stocked at our laboratory cell bank. Briefly, CRC tissue from resected specimens was minced into 1-mm pieces and dissociated with 1mg/mL of collagenase (C6885; Sigma-Aldrich, St. Louis, MO, USA). Filtered cell pellets between 20m and 200m were seeded in plates coated with iMatrix-511 (Takara Bio Inc., Kusatsu, Japan) and cultured in medium containing 10ng/mL of basic fibroblast growth factor (ThermoFisher Scientific, Waltham, MA, USA) and 2ng/mL of transforming growth factor beta (R&D Systems Inc., Minneapolis, MN, USA) to maintain heterogeneous primary culture cells. Sixteen 2DOs with stable culture and drug sensitivity on testing, including eight 2DOs from patients with distant metastases or recurrences, were selected for further analysis.

The human colorectal tumor cell lines, HCT116, gifted by Dr. Bert Vongelstein (Johns Hopkins University, Baltimore, MD, USA), and HT29 (EC91072201, ECACC), were cultured in Dulbeccos modified Eagles medium supplemented with 10% fetal bovine serum (ThermoFisher Scientific), 1% GlutaMAXI (ThermoFisher Scientific), and 1% penicillin/streptomycin/amphotericin B (Wako Pure Chemical Industries, Osaka, Japan). Cells were incubated at 37C in a humidified atmosphere containing 5% CO2. Cells were harvested using 0.25% Trypsin-EDTA (ThermoFisher Scientific) for further analysis.

Cellartis human iPS cell line 12 (ChiPSC12) cells (Takara Bio) were cultured in the Cellartis DEF-CS 500 Culture System (Takara Bio). Cells were incubated at 37C in a humidified atmosphere containing 5% CO2. Cells were harvested using Accutase (Innovative Cell Technologies, Inc., San Diego, CA) for further analysis.

The expression of proteins in cells was determined using flow cytometry. Cultured cells were dissociated with Accutase (Nacalai Tesque Inc., Kyoto, Japan). CTCs were isolated from clinical blood samples using OncoQuick (Greiner BioOne, Frickenhausen, Germany) according to the manufacturers protocol. Cells were stained with antibodies targeting EpCAM, CD133, CD44, CD41, CD45, and LGR5 (Supplementary TableS2). For detecting POU5F1, a True-Nuclear Transcription Factor Buffer Set (424401; BioLegend) was used. After staining cell surface proteins, cells were fixed and stained with antibodies for POU5F1, according to the manufacturers protocol. Relative fluorescent intensities were measured with an SH800 cell sorter (SONY, Tokyo, Japan) and cell morphology and staining locations were also measured with a FlowSight imaging flow cytometer (Merck-Millipore, Darmstadt, Germany). 7-AAD (Miltenyi Biotec, San Diego, CA, USA) was used to analyze living cells. A dimensionality reduction step in two dimensions was performed using t-distributed stochastic neighbor embedding (t-SNE) to visualize high-dimensional data of stem cell marker expression. Data were analyzed using FlowJo 10.2 software (FlowJo LLC, Ashland, OR, USA).

Anticancer drug sensitivity was examined in sixteen 2DOs within 510 passages. Drugs and their concentrations in clinical drug assays are listed in Supplementary TableS3. The number of viable cells in each well was measured using a Cell Counting Kit-8 (Dojindo Laboratories, Kumamoto, Japan) before drug administration and 96h after drug administration. Cell proliferation in DMSO and distilled water, which were used to dilute each drug, were used as controls. The ratio of the number of living cells after administering the drug to the control is shown. Three independent experiments were performed and the average is shown. The formula used for calculation was as follows: 100Cont. 0h cell num.Drug 96h cell num./{(Cont. 96h cell num.Cont. 0h cell num.) Drug 0h cell num.}

Regarding the sensitivity of each anticancer drug, a dimensionality reduction step in two dimensions was performed using t-SNE to visualize high-dimensional data for 21 drugs in a low-dimensional space. The statistical analyses were performed using R 3.6.3 (R Core Team, 2018), with the data.table (v1.12.8; Dowle & Srinivasan43), t-SNE (Krijthe44), and ggplot2 (Wickham45) packages.

Total RNA was extracted using an RNA Purification Kit (Qiagen, Hilden, Germany). TruSeq Stranded mRNA Library Prep (Illumina, San Diego, CA, USA) was used to prepare RNA-seq libraries from the total RNA (1g). Multiplexed libraries were sequenced on an Illumina NextSeq with single-end 75-bp sequencing. RNA-seq data were mapped to the hg38 genome release using the bioinformatic pipeline of the Illumina Base Space Sequence Hub and the Subio software platform (Subio, Inc., Kagoshima, Japan).

The vector, PL-SIN-Oct4-EGFP, kindly provided by James Ellis (Addgene plasmid #21319; http://n2t.net/addgene:21319)22, was used to establish cells expressing EGFP under the OCT4 (POU5F1) promoter. The vector was transfected into 2DOs and cell lines using Lentiviral High Titer Packaging Mix with pLVSIN (Takara Bio). EGFP-positive cells were purified by sorting using a SH800 cell sorter (SONY) at least twice. POU5F1 expression was confirmed by polymerase chain reaction (PCR).

Total RNA was isolated using an RNA Purification Kit (Qiagen). Quantitative assessment was performed by real-time PCR using 100nM universal probe libraries, 0.1 FASTStart TaqMan Probe Master (Roche Diagnostics, Basel, Switzerland) for designed primers, iTaq Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, USA) for commercially available primers, 100nM primers, and 10ng cDNA for cDNA amplification of target genes. Primers are listed in Supplementary TableS4. PCR was performed with 20L of the master mix in each well of a 96-well plate, and signals were detected with the CFX Connect Real-Time PCR Detection System (Bio-Rad). The thermocycler was programmed for one cycle at 95C for 10min, followed by 40 cycles at 94C for 10s, 60 C for 20s, and 72 C for 1s. cDNAs from NTERA-2 cells were used as positive controls.

A subcutaneous model was established to investigate the ability to differentiate from a single sorted cell. A single sorted cell was cultured in a dish for expansion using the 2DO culture methods described above. Accutane-dissociated cells (1106 cells) suspended in Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) were subcutaneously transplanted into the dorsal flanks of 7-week-old, non-obese diabetic/severe combined immunodeficient mice (CLEA, Tokyo, Japan). The average weight was 27g at the start of the experiments. The mice were sacrificed when the tumors reached a diameter of 10mm. For the liver metastasis model, live cells (1106 cells) were sorted by 7-AAD (Miltenyi Biotec) according to EGFP expression using a SH800 cell sorter (SONY) and injected into the spleen. Liver metastasis was assessed every 4 weeks. Mice were sacrificed 8 weeks after injection for the assessment of liver metastases in the POU5F1 expression metastatic ability experiment and 10 weeks after injection in the XAV939 experiment.

Xenograft tumors were fixed in formalin, processed through a series of graded concentrations of ethanol, embedded in paraffin, and sectioned. Sections were stained with hematoxylin and eosin (H&E). Three-dimensional (3D)-formed 2DOs cultured on a NanoCulture plate were collected and centrifuged at 400g for 5min at room temperature. The pellet was consolidated using iPGell (GenoStaff Co., Ltd., Tokyo, Japan) and fixed in formalin. The pellet was processed through a series of graded concentrations of ethanol, embedded in paraffin, sectioned, and stained with H&E.

Xenograft tumors were also fixed in 10% buffered formalin and embedded in paraffin blocks. For cultured 2DOs, 3D-formed 2DOs cultured on an Ultra-Low Attachment Multiple Well Plate (Corning, NY, USA) were collected and centrifuged at 400g for 5min at room temperature. They were embedded in paraffin blocks using iPGell (GenoStaff). A 3-m section was obtained from each block. Sections were deparaffinized, and slides were boiled for 15min. Expressions of CD44, CK20, MUC2, and chromogranin A were quantified using antibodies (Supplementary TableS5). The slides were incubated with a primary antibody for 60min at room temperature and then incubated with a secondary antibody for 30min at room temperature. Slides were mounted in Prolong Gold with DAPI (Invitrogen, Waltham, MA, USA). Mucus production ability was assessed via Alcian blue staining (pH 2.5).

Cultured cells were fixed with 4% formaldehyde and blocked. They were incubated with primary antibodies (Supplementary TableS6) overnight at 4C. Cells were incubated with secondary antibodies for 90min. Slides were mounted in Prolong Gold with DAPI (ThermoFisher Scientific) overnight.

The vector, pLV[Exp]-Neo-CMV>DsRed_Express2, was constructed by VectorBuilder, Inc. (Chicago, IL, USA) (Supplementary Fig.S27). This vector was transfected into 2DOs and iPS cells using Lentiviral High Titer Packaging Mix with pLVSIN (Takara Bio). DsRed_Express2-positive cells were selected by antibiotic selection using G418 (10131035; ThermoFisher Scientific) and sorted twice by an SH800 cell sorter (SONY). All cells expressing DsRed-Express2 were detected by an SH800 cell sorter (SONY).

The vector, PL-SIN-Oct4-EGFP, kindly provided by James Ellis (Addgene plasmid #21319)22, and the vector, pMSCV-F-del Casp9.IRES.GFP, kindly provided by David Spencer (Addgene plasmid # 15567)46, were used to establish cells expressing EGFP under the OCT4 (POU5F1) promoter with inducible caspase 9. Sequence-encoding caspase 9 was digested with restriction enzymes, XhoI (R0146S; New England Biolabs, Beverly, MA, USA) and EcoRI-HF (R3101S; New England Biolabs). The DNA fragment of caspase 9 was extracted from E-Gel CloneWel 0.8% (G6500ST; ThermoFisher Scientific) using the E-Gel Power Snap Electrophoresis System (ThermoFisher Scientific) (Supplementary Fig.S28). The fragment was amplified using CloneAmp HiFi PCR Premix (Z9298N; Takara Bio) with designed primers (FW_gaattctgcagtcgatcgagggagtgcaggtgg, RV_ccgcggtaccgtcgacttagtcgagtgcgtagtc). The vector, PL-SIN-Oct4-EGFP, was linearized by a restriction enzyme, SalI-HF (R3138S; New England Biolabs). The amplified fragments and linearized vector were used for the cloning reaction by the In-Fusion HD Cloning Kit (Z9648N; Takara Bio). The transformation procedure was performed using Competent High E. Coli DH5 (TYB-DNA903; Toyobo, Osaka, Japan), and the plasmid was extracted using the Qiagen Plasmid Plus Midi Kit (12945; Qiagen). The nucleotide sequence of the vector was confirmed by Sanger sequencing, performed by GENEWIZ Japan Corp. (Kawaguchi, Japan). Primer extension sequencing was performed using Applied Biosystems BigDye version 3.1, and the reactions were then run on an Applied Biosystem 3730xl DNA Analyzer. The constructed vector was transfected into two 2DOs (603iCC and 25DiCC) using Lentiviral High Titer Packaging Mix with pLVSIN (Takara Bio). EGFP-positive cells were cloned by single-cell sorting using an SH800 cell sorter (SONY). POU5F1 expression was confirmed by PCR, and a decrease in the number of EGFP-positive cells was confirmed by the administration of B/B Homodimerizer (Z5059N; Takara Bio). The mean provirus copy number was 6.05 (1.16, n=6), as measured using the Let-X Provirus Quantitation Kit (Z1239N; Takara Bio).

603iCC-transfected POU5F1-EGFP cells with inducible caspase 9 (4.5104/well) were seeded, and 5M B/B Homodimerizer (Takara Bio) was administered for three days. Four days after the dimerizer was removed, live cells were sorted using an SH800 cell sorter (SONY) as day 7 cells. For cells not treated with a dimerizer, live cells were also sorted as day 0 cells. Single-cell library preparation was performed following the manufacturers instructions for the Chromium Next GEM Single Cell 3 Reagent Kit (v3.1) (10x Genomics, Pleasanton, CA, USA), and the libraries were sequenced on a HiSeq X sequencer (Illumina). To generate a data matrix, the Cell Ranger pipeline (v4.0.0) was applied, and raw reads were aligned to the human reference genome (GRCh 38) using the STAR aligner. For GFP transcript mapping, the GFP sequence (XM_013393261) was added to the reference fastq and gtf files. Data were deposited in Gene Expression Omnibus under the accession number GSE169220.

Seurat (version 3.2.0)47 was used for quality control and downstream analysis. Poor-quality cells were filtered out using the following parameters: nFeature_RNA 2009000 and percent.mt <10. A total of 6942 cells (control: 3342 cells and day 7: 3602 cells), which passed the quality control, were finally used for further analysis. Mitochondrial genes were filtered by mt.percent (<10). UMAP visualization was used for dimensionality reduction analysis with the following parameters: resolution, 0.5; and perplexity, 20. Marker genes discriminating the different clusters were identified using the FindAllMarkers function (min.pct=0.25 and log[fold-change] >0.25). Pathway enrichment analysis was performed using Enrichr48 (https://maayanlab.cloud/Enrichr/). To construct a single-cell pseudotime trajectory, the Monocle3 (v0.2.2) algorithm was applied (https://cole-trapnell-lab.github.io/monocle3/). After converting the Seurat object using the as.cell_data_set function, the root node was assigned to cluster 4, and the orderCells function was used to assign cells a pseudotime value. To subdivide cells based on their branch in the trajectory, the choose_graph_segments function was applied, and cluster 6 was chosen as an ending node.

Western blot analysis was performed to examine proteins associated with the Wnt/-catenin signaling pathway. Cells were lysed in 50mM TrisHCl (pH 7.6), 1% Nonidet P-40, 150mM sodium chloride, and 0.1mM zinc acetate in the presence of protease inhibitors. Protein concentration was determined by the Lowry method (Bio-Rad), and 20g of each sample was separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis. The gel was transferred electrophoretically onto a polyvinylidene difluoride membrane (Millipore, Billerica, MA, USA). The membrane was blocked with blocking buffer for 1h and then incubated overnight at 4 C with primary antibodies against -catenin (1:1000, 8480, Cell Signaling Technology), Wnt-3a (1:5000, GTX128101, Gene Tex, CA, USA), and HistoneH3 (1:2000, 4499S, Cell Signaling Technology). After a 2-h incubation with the secondary antibody, horseradish peroxidase-conjugated rabbit antibody (1:400, 7074S, Santa Cruz Biotechnology Inc., Dallas, TX, USA), protein bands were visualized using an ECL detection kit (ThermoFisher Scientific) according to the manufacturers instructions.

DNA samples were treated with sodium bisulfite using a bisulfite conversion kit (Zymo Research EZ DNA methylation Kit). After treatment, unmethylated cytosines convert to uracil, while methylated cytosines remain unchanged. Bisulfite-converted DNA samples were analyzed using the Infinium MethylationEPIC BeadChip Kit (Illumina). Bisulfite-converted DNA samples were denatured and neutralized by alkali. The denatured samples were then amplified by whole-genome amplification (37C overnight). Amplified DNA samples were enzymatically fragmented for 1h at 37C in a microsample incubator. 2-Propanol was added to the fragmented DNA samples and precipitated by centrifugation. Precipitated DNA samples were resuspended with hybridization buffer and incubated for 1h at 48C in a hybridization oven. Fragmented and resuspended DNA samples were denatured for 20min at 95C in a microsample incubator. Denatured DNA samples were dispensed onto BeadChips using a TECAN System. The BeadChips were incubated overnight at 48 C in the hybridization oven to hybridize the samples onto the BeadChips. After hybridization, seals were removed from the hybridized BeadChips. Next, unhybridized fragment DNAs were washed away. Labeled nucleotides were added to the washed BeadChips to extend primers which hybridized to the DNA. BeadChips were stained, then coated for protection, and dried. Dried BeadChips were scanned with the iSCAN System. Illumina GenomeStudio software (V2011.1) loaded the signal intensity files of BeadChips, and beta values were decided via normalization and background subtraction. Next, a comparative analysis was executed based on the Illumina Custom Model algorithm, and difference scores for all probes were computed. The markers with signal intensities adequate to distinguish between the signal and background noise were used in subsequent analysis. The markers with high scores (highly methylated and highly unmethylated compared to the reference sample) were extracted, and clustering analysis was conducted.

The NANOG binding consensus sequence is generally known to be 5TAAT[GT][GT]3 or 5[CG][GA][CG]C[GC]ATTAN[GC]3. Therefore, in the sequence of focus, the CGCCCAGTGTC part is quite similar to the binding sequence. We used Protein Data Bank data, including 4RBO, to predict binding conformations to the NANOG protein with the wild-type sequences or methylated sequence with our original method49. A sufficient amount of water molecules was placed around the complex structures, and thermodynamical sampling was performed under a periodic boundary condition. After stabilizing the complex structure by energy minimization calculations, some molecular dynamics simulations were performed at ~37C (310K) to capture the molecular behavior under the biological environment. After a sufficient thermal equilibration process, the molecular vibrations of the bonding configurations were sampled. All these calculations were performed using the AMBER package. The distributions of the interaction energy between DNA and NANOG protein were calculated by extracting 2000 conformations of complex structures from the trajectory with the abovementioned molecular dynamics simulations. Each binding energy was calculated using Gaussian program packages50 with the AMBER99 Force field level51.

603iCC cells (1104 per well) were seeded into 96-well plates and incubated for 48h. After incubation, cells were exposed to different concentrations of XAV939 (BD248591; BLD Phamatech Ltd., Shanghao, China) for 96h. The percentage of viable cells was determined using a cell counting kit solution (CCK-8; Dojindo Molecular Technologies) according to the manufacturers protocol.

Prior to cancer cell seeding, plates were coated. iPS cell-coated plates were seeded into 12-well plates (2105 iPS cells/well) 2 days prior to seeding. iPS cells were tagged with DsRed-Express by the aforementioned methods. Laminin coatings were prepared using iMatrix-511 (T304, Takara Bio) according to the manufacturers protocol. Sorted POU5F1-positive cells (2105/ well) were seeded on these plates. Medium was prepared with XAV939 (10M) for the XAV939 group and DMSO (0.3%) for the control group. All medium exchanges were performed every other day, and cells in the collected supernatant were analyzed by an SH800 cell sorter (SONY). Cells not expressing DsRED-Express2 were counted as cancer cells.

Stained specimens were analyzed using ImageJ software52. Five independent images were collected for each sample and the areas of protein expression in the samples were measured. The value was normalized by dividing by the number of cells stained with DAPI.

As an evaluation of XAV939, sorted POU5F1-positive cells were directly injected into the spleen of mice (1106 cells). After recovering from anesthesia, mice were randomly allocated to the control (0.3% DMSO that is the final concentration of DMSO in XAV939 group) or XAV939 group (100g/injection/mouse). XAV939 (CS-0494, ChemScene, Monmouth Junction, NJ, USA) was administered by intraperitoneal injection at 1mg/mL (injection volume, 100L) every day for 8 weeks, followed by 2 weeks of observation. Ten weeks after injection, mice were sacrificed for the assessment of metastases. Mouse body weight was measured twice per week, and no weight gain or loss greater than 5% was observed.

The Osaka University Review Board, the OICI Review Board, approved this study, and written informed consent for the study was obtained from all participants according to the ethics guidelines. All ethical regulations relevant to human research participants were followed. The OICI Animal Research Committee approved this study, and we have complied with all relevant ethical regulations for animal use. All experimental protocols were in accordance with the guidelines of the Osaka University, the OICI, and Declaration of Helsinki.

Continuous variables are expressed as the mean with standard error of the mean. The significance of the difference between the two groups was analyzed using the x2 test and Wilcoxons signed rank-sum test. All data were analyzed using JMP software (SAS Institute), R 3.6.3, and Prism 8 (GraphPad Software, San Diego, CA, USA). Results were considered statistically significant at P<0.05.

Originally posted here:
Metastases and treatment-resistant lineages in patient-derived ... - Nature.com