Methods
Reagents and antibodies
Antibodies used in FACS sorting were: EpCAM: Biolegend 9C4 antibody,
cat. no. 324208, lot B198775, fluorochrome APC. CD45: Biolegend 2D1
antibody, cat. no. 368516, lot B264243, fluorochrome APC/Cy7. CD3: BD
Biosciences SK7 antibody, cat. no. 564001, lot 6109971, fluorochrome
BUV395.
Software versions
Data were collected using Cell Ranger software (10x Genomics) v.2.2.0^32.
Data were analysed using Perl v.5.18.2, R v.3.6.0, and the following
packages and versions in R for analysis: Seurat, 3.0.2^33 ; SingleR, 1.0.1^34 ;
RankProd, 3.11.0^35 ; GSEABase, 1.47.0; limma, 3.41.15^36 ; annotate, 1.63.0;
homologene, 1.4.68.19.3.27; mouse4302.db, 3.2.3; cocor, 1.1-3^37 ; iNEXT,
2.0.19^38 ; and survival, 2.44-1.1^39. Two-dimensional gene expression
maps were generated using coordinates from the UMAP algorithm^40 ,^41 ,
as implemented in the umap-learn module v.0.3.10 in Python v.3.6.0,
run under R using reticulate v.1.13. Figures were produced using the
following packages and versions in R: colorspace, 1.4-1^42 ; RColorBrewer,
1.1-2; pheatmap, 1.0.12; and superheat, 0.1.0^43. External datasets were
obtained using GenomicDataCommons, 1.9.0^44 and GEOquery, 2.53.0^45.
Human research participants
Patient samples were procured from commercial vendors (Discovery
Life Sciences, iSpecimen Inc., Avaden BioSciences and TriMetis Life Sci-
ences) from adult patients undergoing resection surgery. We complied
with all ethical standards of the Roche Ethics Committee. Informed
consent was obtained from all participants.
Tissue dissociation
Fresh surgical samples were separated into tumour and NAT compart-
ments by the reviewing pathologist and shipped overnight to our insti-
tution. Upon arrival, samples were rinsed with PBS until no traces of
blood were visually detected. Subsequently, samples were digested
with a combination of collagenase D (0.5 mg ml−1) and DNase (0.1 mg
ml−1) for 15 min at 37 °C with gentle shaking. Subsequently, samples
were subjected to a gentleMACS dissociator (Miltenyi Biotec), followed
by an additional 10 min incubation at 37 °C.
Peripheral blood mononuclear cell isolation
Peripheral blood mononuclear cells from patient blood samples were
isolated using 50-ml Leucosep tubes (Greiner Bio-One International)
and Ficoll-Paque PLUS (GE Healthcare). Whole blood drawn into
sodium heparin blood collection tubes were diluted three times with
PBS without calcium or magnesium (Lonza). Diluted cell suspensions
were carefully layered on Leucosep tubes and centrifuged for 15 min
at 800g at room temperature. Interphase-containing peripheral blood
mononuclear cells were collected and washed with PBS and subse-
quently centrifuged for 10 min at 250g at room temperature before
further processing.
FACS sorting
Following tissue enzymatic dissociation of tissues, single-cell suspen-
sions were subjected to antibody staining with anti-EPCAM, anti-CD45,
and anti-CD3. Cells were purified by fluorescence-activated cell sorting
(FACS) on a Becton Dickinson FACSAria Fusion cell sorter equipped
with four lasers (405 nm, 488 nm, 561 nm and 638nm). A 70-μm nozzle
running at 70 psi and 90 kHz was used as the setup for each sort session.
FACSDiva (v.8.0.1) and FlowJo (v.10) were used to collect and analyse
the flow cytometry data. Before gating on fluorescence, single cells
were gated using forward scatter (FSC-A) and side scatter (SSC-A) (for
intact cells) and SSC-W/SSC-H and FSC-W/FSC-H (to ensure that only
singlets were sorted). FACS gates were drawn to include only live single
cells based on Calcein Blue AM+ and 7-AAD (Thermo Fisher Scientific).
Further gates were drawn to arrive at CD3+CD45+EpCAM− (for CD3+
selected samples) or CD45+EpCAM− cells (for CD45+ selected samples).
Boundaries between positive and negative cell fractions were drawn
based on single-colour stains. An example gating strategy is shown
in Supplementary Methods.Single-cell RNA-seq and TCR V(D)J clonotype profiling
Sample processing for single-cell gene expression (scRNA-seq) and
T cell receptor V(D)J clonotypes (scTCR-seq) was done using the
Chromium Single Cell 5′ Library and Gel Bead Kit (10x Genomics),
following the manufacturer’s user guide. Cell density and viability of
FACS-sorted CD3+ T cells or CD45+ cells from tumours, NAT and blood
were determined by Vi-CELL XR cell counter (Beckman Coulter). All of
the processed samples had cell viability above 90%. The cell density
was used to impute the volume of single cell suspension needed in the
reverse transcription master mix, aiming to achieve approximately
6,000–10,000 cells per sample. After Gel Bead-in-Emulsion reverse
transcription (GEM-RT) reaction and clean-up, a total of 14 cycles of
PCR amplification were performed to obtain sufficient cDNAs used
for both RNA-seq library generation and TCR V(D)J targeted enrich-
ment followed by V(D)J library generation. TCR V(D)J enrichment was
done per manufacturer’s user guide using Chromium Single Cell V(D)
J Enrichment Kit, Human T cell (10x Genomics). Libraries for RNA-seq
and V(D)J were prepared following the manufacturer’s user guide (10x
Genomics), then profiled using Bioanalyzer High Sensitivity DNA kit
(Agilent Technologies) and quantified with Qubit (Thermo Fisher Scien-
tific) or Kapa Library Quantification Kit (Kapa Biosystems). Single-cell
RNA-seq libraries were sequenced in one lane of HiSeq4000 (Illumina).
Single-cell TCR V(D)J libraries were tagged with a sample barcode for
multiplexed pooling with other libraries, sequenced in both lanes of a
HiSeq2500 machine (Illumina) using Rapid Run mode, and then demul-
tiplexed. All sequencing was done according to the manufacturer’s
specification (10x Genomics).Processing of scTCR-seq data
TCR-seq data for each sample was processed using Cell Ranger soft-
ware with the command ‘cellranger vdj’ using a custom reference set
of 30,727 genes, based on human reference genome GRCh38 and Ref-
Seq gene models, and provided in our software package (see ‘Code
availability’). For each sample, Cell Ranger generated an output file,
filtered_contig_annotations.csv, containing TCR α-chain and β-chain
CDR3 nucleotide sequences for single cells that were identified by
barcodes.
Although Cell Ranger software groups cells according to CDR3 nucle-
otide sequences to obtain clonotypes separately for each sample, our
analysis depends on identifying shared clonotypes across samples. We
therefore re-grouped clonotypes across the tumour, NAT and blood
samples for each patient, requiring that they share all reported α-chain
and β-chain CDR3 consensus nucleotide sequences in common. We
tested our procedure on each sample individually and confirmed that
it grouped clonotypes consistently with the original clonotypes from
the Cell Ranger software.
Clonotypes were assigned a tissue expansion pattern based on their
clone sizes in NAT and tumour as follows. Clonotypes having one cell in
NAT but none in tumour were called NAT singletons, while clonotypes
having more than one cell in NAT but none in tumour were called NAT
multiplets. Conversely, clonotypes having one cell in tumour but none
in NAT were called tumour singletons, and clonotypes having more
than one cell in tumour but none in NAT were called tumour multiplets.
Clonotypes having at least one cell in NAT and at least one cell in tumour
were called dual-expanded clones.
Clonotypes from blood samples were classified according to the
above criteria if they had any cells in NAT or tumour. The remaining
clonotypes, which had cells only in blood and none in NAT or tumour,
were called blood singletons or blood multiplets if they had one cell