Adult Puzzle Books

Ethical approval statement

This research and all study protocols have been approved and comply with the Peter MacCallum Animal Experimental Ethics Committee (AEC) ethical regulations regarding the use of animals. Studies using human peripheral blood mononuclear cells (PBMCs) from healthy donors was approved by the Peter MacCallum Cancer Centre Human Research Ethics committee. Informed consent was obtained from the Australian Red Cross.

Animal models

C57BL/6 wild-type mice and C57BL/6 human-HER2 transgenic mice⁴⁷ were bred in the Peter MacCallum Cancer Centre animal facility. NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ (NSG) mice were either bred at the Peter MacCallum Cancer Centre or obtained from Australian BioResources. Mice used in experiments were between 6 to 16 weeks of age and were housed in PC2 specific pathogen-free conditions and a minimum of 3 mice per group were used in each experiment. Mice were randomized prior to treatment according to tumour size to ensure all groups had equivalent tumour burden prior to therapy. Experiments were not blinded as the same investigators performed and analysed experiments and so blinding was not possible. Experiments were approved by the Animal Experimentation Ethics Committee no. E582, E671 and E693 and all experiments complied with the ethical endpoints stated in the approved projects, including maximum tumour size.

Cell lines

The mouse MC38 colon adenocarcinoma cell line was provided by J. Schlom. The mouse breast carcinoma cell line E0771 was obtained from R. Anderson. The parental MC38 and E0771 tumour cell lines were retrovirally transduced with a mouse stem cell virus (MSCV) vector to express a truncated human HER2 antigen that lacks intracellular signalling components. Transduced tumour cell lines are referred to as MC38-HER2 and E0771-HER2. OVCAR-3 and MCF7 tumour cells were obtained from the American Type Culture Collection. PCR analysis was used to verify that tumour lines were negative for Mycoplasma.

Retroviral vector packaging cell lines PA317 and GP+E-86 were obtained from American Type Culture Collection (ATCC). The GP+E-86 and tumour cell lines were maintained in RPMI medium (Gibco Life Technologies) supplemented with 10% heat-inactivated fetal bovine serum, 1 mM sodium pyruvate, 2 mM glutamine, 0.1 mM non-essential amino acids, 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin. These cells were maintained at 37 °C in a humidified incubator with 5% CO₂. The PA317 cell line was maintained in DMEM (Gibco) supplemented with 2 mM glutamine and 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin and was maintained in a humidified incubator at 37 °C with 10% CO₂.

Reagents and cytokines

Mouse IFNγ antibody (H22, IgG, BE0312) and the isotype control (2A3 clone, IgG2a, BE0254) antibody were purchased from BioXcell. The cytokine IL-2 was obtained from the National Institutes of Health and purchased from Peprotech. IL-7 and IL-15 were purchased from Peprotech. Where indicated, CAR T cells were stimulated with an anti-idiotype antibody that was custom made.

Generation of retroviral packaging lines for the transduction of primary mouse T cells

cDNA encoding mouse TCF7, FOXO1 (wild-type), FOXO1-ADA, ID3 and JUN were cloned into the mouse stem cell virus (MSCV) vector encoding either an mCherry marker gene or truncated (lacks cell signalling components) human nerve growth factor receptor (NGFR). The viral packaging GP+E-86 cell line that produces the anti-HER2 CAR retrovirus was generated as previously described⁴⁸. The anti-HER2 CAR construct was comprised of an extracellular scFv specific for human HER2, an extracellular CD8 hinge region, a CD28 transmembrane domain and an intracellular CD3ζ domain. GP+E-86 cell lines encoding both the anti-HER2 CAR and a transcriptional regulator were generated and the resulting anti-HER2 CAR packaging cells were sorted based on NGFR or mCherry expression by flow cytometry. Supernatants from these cells were used to transduce primary mouse T cells as previously described¹² and following transduction, CAR T cells were maintained in supplemented RPMI medium with IL-7 (200 pg ml⁻¹), IL-15 (10 ng ml⁻¹) and β-mercaptoethanol (50 μM).

Generation of lentivirus for the transduction of human T cells

Lentiviral packaging plasmids (pCMV-VSV-G, pMDLg/pRRE, pRSV-Rev) and plasmid vectors encoding a second-generation Lewis Y CAR and either wild-type FOXO1, FOXO1-ADA or mCherry were purchased from GenScript. In brief, packaging plasmids and transgene plasmids were transfected into HEK293T cells. Across the following 3 days, cell culture supernatants were collected, pooled and centrifuged with Lenti-X-Concentrator (Takara Bio) to concentrate lentivirus. Lentivirus was used to transduce human T cells activated with OKT3 (30 ng ml⁻¹) and IL-2 (600 IU ml⁻¹) for 48 h by adding virus directly to cell cultures at a multiplicity of infection of 0.5 in Lentiboost (Sirion).

CRISPR–Cas9 editing of CAR T cells

CRISPR–Cas9 editing of mouse CAR T cells was performed as described¹². Per 20 × 10⁶ naive splenocytes or 1 × 10⁶ activated human PBMCs, 270 pmol single guide RNA (sgRNA) (Synthego) and 37 pmoles recombinant Cas9 were combined and incubated for 10 min to generate Cas9–sgRNA ribonuclear protein (RNP). Cells were resuspended in 20 µl P3 buffer (Lonza), combined with RNP and electroporated with a 4D-Nucleofector (Lonza) with pulse code E0115 or DN100 for human and mouse T cells, respectively. Prewarmed medium was then added to cells for 10 min prior to activation and transduction of mouse T cells or immediate transduction of human T cells. sgRNA sequences used were as follows: Foxo1 guide 1: 5′-CACCUGGGGCGCUUCGGCCA-3′, guide 2: 5′-CCACUCGUAGAUCUGCGACA-3′. FOXO1 guide 1 5′-CACCUGAGGCGCCUCGGCCA-3′.

In vitro re-stimulation assay

Tumour cell targets were co-cultured with CAR T cells at a 1:1 ratio for 24 h. After overnight incubation, supernatants were collected and an equivalent number of tumour cells were reseeded into the incubations for another 24 h. This process was repeated one final time before cells were collected for analysis by flow cytometry and supernatants were analysed by cytometric bead array (CBA) using either mouse or human cytokine Flex sets (BD Biosciences) according to the manufacturer’s instructions.

Flow cytometry and cell sorting

For flow cytometric analysis Fc receptor block (2.4G2 diluted 1:50 from hybridoma supernatant in FACS buffer) was added to cells for 10 min at 4 °C. Cells were stained with 50 μl fluorochrome-conjugated antibody cocktails and incubated for 30 min in the dark at 4 °C. For intracellular staining, cells were fixed and permeabilized using the eBioscience FoxP3/Transcription Factor Staining Buffer Set (Thermo Fisher) according to the manufacturer’s instructions. Samples were quantified using counting beads (Beckman Coulter; 20 μl per sample) using the following formula: number of beads per sample/bead events × cell events of interest. Cells were analysed on a BD LSRFortessa or BD FACSymphony (BD Biosciences) and data were analysed using Flowjo (TreesStar) or OMIQ (https://www.omiq.ai/). Cells were sorted using a BD FACSAria Fusion. See Supplementary Fig. 1 for the list of antibodies used for flow cytometry. Examples of the gating strategy used to identify mouse and human CAR T cells ex vivo is shown in Supplementary Fig. 1.

Treatment of mice with CAR T cells

C57BL/6 human-HER2 transgenic mice were injected with 2 × 10⁵ E0771-HER2 breast carcinoma cells orthotopically into the mammary fat pad 5–7 days prior to treatment or subcutaneously with 2.5 × 10⁵ MC38-HER2 colon adenocarcinoma cells 5 days prior to treatment. After tumours were established, mice bearing E0771-HER2 or MC38-HER2 tumours were preconditioned with 4 Gy or 0.5 Gy total body irradiation respectively. Mice were then treated with intravenous doses of 1 × 10⁷ CAR T cells on 2 consecutive days and one dose of IL-2 (50,000 IU per dose) with the first dose of CAR T cells, followed by two doses of IL-2 each day on the next 2 consecutive days. Tumour area was measured every 2–3 days following treatment. For IFNγ-blockade experiments, mice were dosed with 250 μg of anti-IFNγ or isotype control antibody 2A3 on days 0, 1 and 7 following CAR T cell treatment.

For experiments utilizing human anti-Lewis Y CAR T cells, NSG mice were injected with 5 × 10⁶ OVCAR-3 tumour cells. Once tumours were established, at day 10–15 post injection, mice were treated with 1 Gy total body irradiation and intravenously treated with 2–5 × 10⁶ Flag⁺ CAR T cells. Mice were treated with IL-2 as per experiments in the C57BL/6 human-HER2 transgenic model.

Analysis of immune subsets in tumour, spleen, dLNs and blood

Blood was collected via submandibular or retroorbital bleed into tubes containing EDTA prior to euthanasia. Blood and spleen samples were treated twice or once respectively with ACK lysis buffer before staining for flow cytometry. Tumours were digested in SAFC DMEM medium (Gibco) with 0.01 mg ml⁻¹ DNase (Sigma Aldrich) and 1 mg ml⁻¹ type IV collagenase (Sigma Aldrich) for 30 min at 37 °C. Following digestion, tumour samples were filtered twice through a 70-μm filter to create a single-cell suspension and resuspended in Fc block prior to staining for analysis by flow cytometry. For stimulation of intratumoral CAR T cells to assess cytokine secretion capacity, tumour cell suspensions were resuspended in complete RPMI medium with 10 ng ml⁻¹ phorbol 12-myristate 13-acetate (Abcam), 1 μg ml⁻¹ ionomycin (Abcam), GolgiStop (1:1,500 dilution, BD Biosciences) and GolgiPlug (1:1,000 dilution, BD Biosciences). Samples were incubated for 3 h at 37 °C with 5% CO₂ prior to staining for analysis by flow cytometry. Single-cell suspensions from dLN were created by placing tissue between two pieces of 70-μm filter mesh in 400 μl of FACS buffer and by mechanically digesting using the end of a syringe. The resultant cell suspension was then stained for analysis by flow cytometry. For mitochondrial analysis, isolated cells were stained using Mitotracker Deep Red FM and Mitotracker Green FM (Thermo Fisher) according to the manufacturer’s protocols.

Seahorse assay

A Seahorse XFe24 Bioanalyser (Agilent) was used to determine OCR for indicated CAR T cells prepared from 5 separate donors. Cells were washed in assay medium (XF Base media (Agilent) with glucose (10 mM), sodium pyruvate (1 mM) and l-glutamine (2 mM) (Gibco), pH 7.4 at 37 °C) before being plated onto Seahorse cell culture plates coated with Cell-Tak (Corning) at 4 × 10⁵ cells per well. After adherence and equilibration, cellular OCR and extracellular acidification rates (ECAR) were measured using a Seahorse MitoStress assay (Agilent), with addition of oligomycin (1 μM), carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (1.2 μM) and antimycin A and rotenone (0.5 μM each). Assay parameters were as follows: 3 min mix, no wait, 3 min measurement, repeated 3 times at basal and after each addition. Raw OCR values were normalized to the amount of protein per well, as assessed by a Pierce BCA protein assay (Thermo Fisher) performed as per manufacturer instructions. SRC was calculated as OCR at maximum rate − OCR in basal state.

Gene expression analysis

Following manufacturer’s instructions, RNA-seq libraries were prepared from RNA using the Quant-seq 3′ mRNA-seq Library Prep Kit for Illumina (Lexogen). Single-end, 75 bp RNA-seq was performed via NextSeq (Illumina) and CASAVA 1.8.2 was subsequently used for base calling. Cutadapt v2.1 was used to remove random primer bias and trim 3′ end poly-A-tail derived reads. Quality control was assessed using FastQC v0.11.6 and RNA-SeQC v1.1.8⁴⁹. Sequence alignment against the mouse reference genome mm10 or the human genome hg19 was performed using HISAT2. Finally, featureCounts from the Rsubread software package 2.10.5 was used to quantify the raw reads with genes defined from the respective Ensembl releases⁵⁰. Gene counts were normalized using the TMM (trimmed means of M-values) method and converted into log₂-transformed counts per million (CPM) using the EdgeR package^51,52. The quasi-likelihood F-test statistical test method based on the generalized linear model (glm) framework from EdgeR was used for differential gene-expression comparisons adjusted P values were computed using the Benjamini–Hochberg method. Principal component analysis was performed generated based on the top most variable genes. DEGs were classified as significant based on a false discovery rate cut-off of less than 0.05. For heat maps, the pheatmap R package was used to plot row mean centred and scaled normalized log₂(CPM + 0.5) values. Genes columns or rows were sorted by hierarchical clustering using Euclidean distance and average linkage.

Unbiased gene set enrichment analysis was performed using fgsea package on differential expressed genes pre-ranked by fold change with 1,000 permutations (nominal P value cut-off <0.05)⁵³. Reference gene sets were obtained from the MsigDB library for Hallmarks, KEGG (https://www.genome.jp/kegg/kegg1.html), CHEA dataset^54,55,56,57, or based on previously published analyses of glycolysis signature²³, scRNA-seq-derived T cell clusters in patients¹¹.

scRNA-seq data processing and analysis

CAR T cells were co-cultured with MCF7 tumour cells at a 1:1 ratio for 24 h. Fc-receptors were blocked with human Fc Block (BD BioSciences) for 10 min at 4 °C before staining with 50 μl fluorochrome-conjugated antibody cocktail for 30 min in the dark at 4 °C. Samples were labelled with anchor lipid-modified oligo (LMO) (5′-TGGAATTCTCGGGTGCCAAGGgtaacgatccagctgtcact-[lipid]-3), co-anchor LMO (5′-[lipid]-AGTGACAGCTGGATCGTTAC-3′) and sample specific barcodes for 5 min in the dark at 4 °C. CAR⁺ T cells were sorted by FACS and samples were pooled at equal ratios followed by staining with 100 μl TotalSeq-C anti-human CD4 and CD8 (BioLegend) antibody cocktail for 30 min in the dark at 4 °C. scRNA-seq data were generated using the 10x Cell Ranger pipeline (7.1.0) and hg38 genome. Specifically, cellranger multi was used to generate raw feature barcode matrices. Downstream analysis was performed in R (version 4.2.0). Empty droplets were detected and removed from the raw feature barcode matrix using the emptyDrops function from the DropletUtils (version 1.16.0) package and doublets were detected and removed using DoubletFinder (verison 2.0.3). Using Seurat (version 4.3.0), cells with less than 200 features and more than 5% mitochondrial reads were excluded. Standard Seurat data processing and normalization steps were performed: NormalizeData, FindVariableFeatures, ScaleData, RunPCA, RunUMAP, FindNeighbors and FindClusters; clusters with low-quality metrics were removed, and the final resolution was determined using results from the clustree package (version 0.5.0). LMOs were demultiplexed using HTODemux (Seurat). DEGs were calculated using the functions FindAllMarkers (Seurat) using a log₂-transformed fold change threshold of 0.125 and an adjusted P value of less than 0.05, and included the number of counts as a latent variable. Pseudobulk DEGs were detected using the Libra package (version 1.0.0) using the run_de function. Gene-set enrichment was performed using the fgsea package with all expressed genes as the background gene list, which was ranked by average log-transformed fold change detected with FindMarkers using a log₂-transformed fold change threshold of 0 and min.pct parameter set to 0. To perform diffexp analyses and GSEA between individual groups within each cluster, the to_psuedobulk function from Libra was used to pull out pseudobulk count matrix of each replicate pool and clusters. EdgeR and fgsea was then utilized to perform differential expression and gsea analyses of reference gene signatures. The single-cell signature explorer program was utilized for visualization of gene signatures across UMAP plots⁵⁸.

ATAC-seq data analysis

Sequencing files for ATAC-seq experiments were demultiplexed using Bcl2fastq (v2.20) to generate Fastq files. Next, quality control of files were performed using FASTQC (v0.11.5). Adaptor trimming of paired-end reads was performed with NGmerge (v0.3) where required⁵⁹. Alignment of reads to either the reference human (hg38) or mouse (mm10) genome was performed using Bowtie2 (v2.3.3). The resulting SAM files were converted to BAM files using Samtools (v1.4.1) using the view command, which were subsequently sorted and indexed, with potential PCR duplicates marked with Samtools markdup. Peak calling was performed with either MACS2 (v2.1.1) or Genrich (v0.6.0) packages. Annotation of ATAC-seq peaks to proximal genes was performed using either annotatePeaks.pl (Homer, v4.11) or the annotatePeak function from ChIPseeker R package (v1.8.6). BAM files were converted into BigWig files using the bamCoverage function (Deeptools, v3.5.0). BigWig files were then imported into Integrative Genomics Viewer (IGV, v2.7.0) for visualization of specific loci. To generate IGV style track plots from BigWig files, the package trackplot was used⁶⁰. The HOMER makeTagDirectory command was used to generate tag directories, and the findPeaks command was used to identify peaks, with the control tag directory set to respective control groups. Motif discovery using the findMotifsGenome tool and default settings identified de novo motifs from peaks identified. The ChromVAR R package⁶¹ was used to identify enriched motifs from the JASPAR 2022 database⁶², in unstimulated or stimulated groups.

Statistical analysis

Statistical analyses were performed using GraphPad Prism. Analyses performed include paired or unpaired Student’s t-test to compare two datasets, one-way ANOVA to analyse multiple datasets across a single timepoint and two-way ANOVA when analysing multiple sets of data across time.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.