Therapy-Induced Evolution of Human Lung Cancer Revealed by Single-Cell RNA Sequencing

Abstract

Lung cancer, the leading cause of cancer mortality, exhibits heterogeneity that enables adaptability, limits therapeutic success, and remains incompletely understood. Single-cell RNA sequencing (scRNA-seq) of metastatic lung cancer was performed using 49 clinical biopsies obtained from 30 patients before and during targeted therapy. Over 20,000 cancer and tumor microenvironment (TME) single-cell profiles exposed a rich and dynamic tumor ecosystem. scRNA-seq of cancer cells illuminated targetable oncogenes beyond those detected clinically. Cancer cells surviving therapy as residual disease (RD) expressed an alveolar-regenerative cell signature suggesting a therapy-induced primitive cell-state transition, whereas those present at on-therapy progressive disease (PD) upregulated kynurenine, plasminogen, and gap-junction pathways. Active T-lymphocytes and decreased macrophages were present at RD and immunosuppressive cell states characterized PD. Biological features revealed by scRNA-seq were biomarkers of clinical outcomes in independent cohorts. This study highlights how therapy-induced adaptation of the multi-cellular ecosystem of metastatic cancer shapes clinical outcomes.

Full text

Article
Therapy-Induced Evolution of Human Lung Cancer
Revealed by Single-Cell RNA Sequencing
Graphical Abstract
Highlights
dscRNA-seq is feasible in metastatic human NSCLCs and
reveals a rich tumor ecosystem
dIndividual tumors and cancer cells exhibit substantial
molecular diversity
dCancer and tumor microenvironment cells exhibit marked
therapy-induced plasticity
dscRNA-seq of metastatic NSCLCs unveils new opportunities
to improve clinical outcomes
Authors
Ashley Maynard, Caroline E. McCoach,
Julia K. Rotow, ..., Collin M. Blakely,
Spyros Darmanis, Trever G. Bivona
Correspondence
collin.blakely@ucsf.edu (C.M.B.),
spyros.darmanis@czbiohub.org (S.D.),
trever.bivona@ucsf.edu (T.G.B.)
In Brief
Analysis of metastatic lung cancer
biopsies before and after targeted
therapy reveals molecular and immune
adaptations that shape clinical outcomes.
Progressive
Disease
Specific Features
Residual Disease
Specific Features
Shared
Features
scRNAseq
Patient Biopsy
Maynard et al., 2020, Cell 182, 1232–1251
September 3, 2020 ª2020 Elsevier Inc.
https://doi.org/10.1016/j.cell.2020.07.017 ll

Article
Therapy-Induced Evolution of Human Lung
Cancer Revealed by Single-Cell RNA Sequencing
Ashley Maynard,
1,15
Caroline E. McCoach,
2,3,15
Julia K. Rotow,
4,16
Lincoln Harris,
1,16
Franziska Haderk,
2,3,5,16
D. Lucas Kerr,
2,16
Elizabeth A. Yu,
2
Erin L. Schenk,
6
Weilun Tan,
1
Alexander Zee,
1,7
Michelle Tan,
1
Philippe Gui,
2,3
Tasha Lea,
3
Wei Wu,
2
Anatoly Urisman,
8
Kirk Jones,
8
Rene Sit,
1
Pallav K. Kolli,
9
Eric Seeley,
2
Yaron Gesthalter,
2
Daniel D. Le,
1
Kevin A. Yamauchi,
1
David M. Naeger,
10,11
Sourav Bandyopadhyay,
3,12
Khyati Shah,
12
Lauren Cech,
2
Nicholas J. Thomas,
2
Anshal Gupta,
2
Mayra Gonzalez,
2
Hien Do,
2
Lisa Tan,
2
Bianca Bacaltos,
2
Rafael Gomez-Sjoberg,
1
Matthew Gubens,
2,3
Thierry Jahan,
2,3
Johannes R. Kratz,
13
David Jablons,
13
Norma Neff,
1
Robert C. Doebele,
6
Jonathan Weissman,
5,14
Collin M. Blakely,
2,3,
*Spyros Darmanis,
1,
*and Trever G. Bivona
2,3,5,17,
*
1
Chan Zuckerberg Biohub, San Francisco, CA 94158, USA
2
Department of Medicine, University of California, San Francisco, San Francisco, CA 94158, USA
3
Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA 94158, USA
4
Lowe Center for Thoracic Oncology, Dana-Farber Cancer Institute, Boston, MA 02215, USA
5
Department of Cellular and Molecular Pharmacology, University of California, San Francisco, San Francisco, CA 94158, USA
6
Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO 80045, USA
7
Department of Biomolecular Engineering, University of California, Santa Cruz, Santa Cruz, CA 95064, USA
8
Department of Pathology University of California, San Francisco, San Francisco, CA 94143, USA
9
Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA 94143 USA
10
Denver Health Medical Center, Denver, CO 80204, USA
11
Department of Radiology, University of Colorado, Aurora, CO 80045, USA
12
Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, San Francisco, CA 94143, USA
13
Department of Surgery, University of California, San Francisco, CA 94143, USA
14
Howard Hughes Medical Institute, University of California, San Francisco, CA 94143, USA
15
These authors contributed equally
16
These authors contributed equally
17
Lead Contact
*Correspondence: collin.blakely@ucsf.edu (C.M.B.), spyros.darmanis@czbiohub.org (S.D.), trever.bivona@ucsf.edu (T.G.B.)
https://doi.org/10.1016/j.cell.2020.07.017
SUMMARY
Lung cancer, the leading cause of cancer mortality, exhibits heterogeneity that enables adaptability, limits
therapeutic success, and remains incompletely understood. Single-cell RNA sequencing (scRNA-seq) of
metastatic lung cancer was performed using 49 clinical biopsies obtained from 30 patients before and during
targeted therapy. Over 20,000 cancer and tumor microenvironment (TME) single-cell profiles exposed a rich
and dynamic tumor ecosystem. scRNA-seq of cancer cells illuminated targetable oncogenes beyond those
detected clinically. Cancer cells surviving therapy as residual disease (RD) expressed an alveolar-regenera-
tive cell signature suggesting a therapy-induced primitive cell-state transition, whereas those present at on-
therapy progressive disease (PD) upregulated kynurenine, plasminogen, and gap-junction pathways. Active
T-lymphocytes and decreased macrophages were present at RD and immunosuppressive cell states char-
acterized PD. Biological features revealed by scRNA-seq were biomarkers of clinical outcomes in indepen-
dent cohorts. This study highlights how therapy-induced adaptation of the multi-cellular ecosystem of met-
astatic cancer shapes clinical outcomes.
INTRODUCTION
Heterogeneity is a property of many biological systems and dis-
eases such as cancer. Biological plasticity in cancer cells is one
form of heterogeneity that allows for early adaptation to treatment
and limits the success of precision approaches for cancer treat-
ment (Xue et al., 2017;Yuan et al., 2019). In addition to cancer-
cell intrinsic heterogeneity, cells within the tumor microenviron-
ment (TME) further contribute to tumor heterogeneity in a cancer
cell extrinsic manner. While these tumor compartments and tumor
heterogeneity have been characterized in many cancer subtypes
(Alexandrov et al., 2013;Brannon et al., 2014;Gerlinger et al.,
2012;Hata et al., 2016;Lawrence et al., 2013;Lee et al., 2014;Vi-
gnot et al., 2015), our understanding of how these properties
evolve and interact longitudinally in response to systemic treat-
ment remains incomplete, particularly in metastatic tumors.
ll
1232 Cell 182, 1232–1251, September 3, 2020 ª2020 Elsevier Inc.

Many oncogene-driven cancers such as those with alterations
in EGFR,ALK,ROS1, and BRAF are treated with targeted ther-
apies against the cognate oncoprotein. This has led to improve-
ments in the clinical outcomes of metastatic solid cancers such
as lung cancer and melanoma as well as hematologic malig-
nancies (Flaherty et al., 2012;Mok et al., 2009;Shaw et al.,
2013). However, tumors typically respond incompletely and
then regrow after acquiring drug resistance. Bulk tumor sam-
pling after progression on targeted therapy has identified resis-
tance mechanisms and demonstrated that tumors become
increasingly molecularly heterogeneous following treatment
(Blakely et al., 2017;Camidge et al., 2014;McCoach et al.,
2018;Rotow and Bivona, 2017).
Single-cell RNA sequencing (scRNA-seq) is one approach to
dissect the heterogeneity of complex biological systems (Chung
et al., 2017;Darmanis et al., 2017;Tirosh et al., 2016). There is
currently a paucity of single-cell studies that sample metastatic
malignancies and prior scRNA-seq studies of metastatic disease
largely focused on single treatment time points (Chung et al.,
2017;Darmanis et al., 2017;Lambrechts et al., 2018;Patel
et al., 2014;Tirosh et al., 2016;Wang et al., 2019;Zhang et al.,
2019). This is due, in part, to challenges associated with obtain-
ing high-quality samples of metastatic human tumors, particu-
larly at multiple treatment time points.
By developing a custom pipeline, we performed scRNA-seq
analyses on advanced-stage NSCLC samples that were ob-
tained from patients before initiating systemic targeted therapy
(TKI naive [TN]), at the residual disease (RD) state, which includes
samples taken at any time during treatment with targeted ther-
apy while the tumor was regressing or stable by clinical imaging
(RD), and upon subsequent progressive disease as determined
by clinical imaging, at which point the tumors showed acquired
drug resistance (progression [PD]).
RESULTS
scRNA-seq Analysis of Advanced-Stage NSCLCs during
Targeted Therapy
We used scRNA-seq to profile 49 samples (45 lung adenocarci-
nomas, 1 squamous cell carcinoma, and 3 tumor adjacent tis-
sues [TATs]) (Figure 1A), corresponding to 30 individual patients.
We used a customized workflow to isolate viable single cells pri-
marily from small tissue samples as well as surgical resections
(Figure 1B). Samples were categorized into three separate time
points (TN, RD, or PD) and further subcategorized by oncogenic
driver (Figure 1C). Collection time for RD samples is illustrated in
Figure S1A. Additional sample details and patient demographics
are included in Table S1.
Gene-expression profiles of 23,261 cells were retained after
quality control filtering. Following gene-expression normalization,
we performed principal-component analysis (PCA) and clustered
cells using graph-based clustering on the informative PCA space
(n = 20). The resulting cell clusters were annotated as immune,
stromal (fibroblasts, endothelial cells, and melanocytes), or
epithelial cells (Figure 1D) by established marker genes (Lam-
brechts et al., 2018;Schiller et al., 2019;Tabula Muris et al.,
2018;Treutlein et al., 2014)(Table S2). Epithelial cells (n = 5,581)
were subsetted and re-clustered into 26 discrete epithelial clus-
ters (Figure S1B). The number of cellsfor each cell type and the an-
alyses that each sample was utilized for are detailed in Table S1.
Clustering-Based Copy-Number Variation Resolves
Cancer from Non-cancer Epithelial Cells
Given the association between cancer and large-scale chromo-
somal alterations, we utilized copy-number variation (CNV) from
RNA expression to classify epithelial cells as either cancer or
non-cancer (Patel et al., 2014;Puram et al., 2017;Tickle et al.,
2019;Tirosh et al., 2016;Venteicher et al., 2017); compared to
fibroblasts and endothelial cells (controls), cancer cells dis-
played larger changes from relative expression intensities across
the genome (Figure S1C). Three TAT samples were included in
this analysis and the majority of cells originating from these sam-
ples were classified as non-cancerous (Table S1). We compared
the average CNV score of samples among treatment time points
(TN, RD, PD) and found it to be consistent. The non-cancer
epithelial cell clusters (n = 16) were further annotated into cell
subtypes (Figures S1D and S1E).
As noted by others (Zhang et al., 1997), we found that cancer
cells expressed an elevated number of unique genes compared
to non-cancer cells (Figure S1F). The difference in the number of
uniquely expressed genes was not explained by sequencing
depth (Pearson correlation = 0.19).
Cancer cells were identified in 44 of the original 49 tumor bi-
opsy samples including a small fraction of cells originating
from each of the TAT samples (0.57%–1.8% of total TAT ob-
tained cells). Given that TAT cells may represent an intermediate
cellular state between normal and cancer cells, their presence at
low frequency is unsurprising and has been described previously
(Aran et al., 2017;Kadara et al., 2014).
All cancer cells (n = 3,754) were re-clustered, resulting in 25
unique clusters (Figures S2A and S2B). For each of the 25
clusters, we calculated the number of cells of the highest
contributing individual patient over the total number of cells for
that cluster for both non-cancer and cancer epithelial cells (pa-
tient occupancy) (Figures S2C–S2E). The majority of cancer
cell clusters were patient specific, having high patient occu-
pancy scores, similar to prior reports (Chung et al., 2017;Darma-
nis et al., 2017;Jerby-Arnon et al., 2018;Neftel et al., 2019;
Puram et al., 2017;Tirosh et al., 2016). Conversely, non-cancer
cell types exhibited lower patient occupancy (Figure S2E).
Thus, patient-specific malignant cell clustering reflected the
unique molecular signatures of an individual patient’s tumor
rather than technical artifact.
scRNA-seq Analysis Reveals a Rich Complexity of
Expressed Gene Alterations in Cancer Cells
Additional genetic alterations can co-exist with a primary target-
able oncogenic ‘‘driver’’ alteration (e.g., oncogenic EGFR,ALK,
KRAS) and may help promote tumor progression and limit ther-
apy response (Blakely et al., 2017;Kim et al., 2019;Scheffler
et al., 2019;Yang et al., 2019). We queried scRNA-seq tran-
scripts from each cancer cell to identify somatic alterations (Fig-
ures 2A–2C; Figures S2G–S2I). In the 44 of 49 biopsy samples
that contained cancer cells, we identified 20 samples harboring
the clinically known oncogenic driver (Figure 2B; Table S3). This
percentage is consistent with the potential drop-out occurrence
ll
Cell 182, 1232–1251, September 3, 2020 1233
Article

in scRNA-seq analyses (Kharchenko et al., 2014). In 24 samples
where we did not identify the clinically known oncogenic driver,
no cells expressed the gene of interest, thus not allowing muta-
tion detection for that gene (Figure S2H; Table S3). In 11 of the 20
samples (55%) where we identified the clinically actionable
oncogene, we also identified an additional oncogenic alteration
that was not detected in clinical-grade bulk nucleic acid testing
of tumor from the same patient (i.e., occult genetic alterations)
(Figure 2B; Figure S2H). An example is sample LTS47. This tu-
mor was determined to harbor an EML4-ALK oncogenic gene re-
arrangement by clinical-grade bulk DNA analysis. scRNA-seq
additionally revealed that this sample contained cancer cells
harboring KRAS G13D and KRAS G12C occult mutations (Fig-
ure S2F). Given the potential of dropout in scRNA-seq data,
we cannot conclude whether the ALK fusion and the KRAS
mutations co-exist within the same cell. However, neither popu-
lation of KRAS mutant cells showed evidence of the ALK gene
rearrangement. This sample was obtained from the patient after
multiple lines of therapy, which could have allowed for evolution
of multiple mechanisms of resistance (Doebele et al., 2012;
Hrustanovic and Bivona, 2015;Shaw and Engelman, 2013).
Loss of an oncogenic driver is also a mechanism of resistance
(Lovly et al., 2017;Tabara et al., 2012;Xu et al., 2018a), although
given the limitation of scRNA-seq we were not able to determine
whether this mechanism of resistance applies to this case.
We also queried scRNA-seq data for mutations from the COS-
MIC (Catalogue of Somatic Mutations In Cancer) lung adenocar-
cinoma tier 1 mutations (Table S2), (Forbes et al., 2017;Shihab
et al., 2015). Many of the mutations we identified had not been
previously reported by the clinical-grade assay conducted on a
patient’s tumor despite having been included in the clinical panel
(Figure 2C; Figure S2I; Tables S2 and S3). Though this may
reflect differences in biopsy technique or tumor clonality at the
time of clinical testing, these results also demonstrate that clin-
ical-grade bulk DNA-based testing may underestimate tumor
heterogeneity.
A
B
CD
Figure 1. Patient Characteristics and Experimental Overview
(A) Consort diagram. 56 biopsies were processed, 49 samples passed quality control.
(B) Tissue processing pipeline for scRNA-seq. Patient samples were disaggregated into single cells and sorted into microtiter plates using FA CS.cDNA synthesis
was performed using the Smart-seq2 protocol, and libraries were sequenced on Illumina platforms.
(C) Circle plot of the clinically identified oncogenic driver (outer circle) and treatment time point (inner circle) for each sample.
(D) t-stochastic neighbor embedding (t-SNE) plot of all cells colored by their cellular identity (epithelial cells [n = 5,581], immune cells [n = 13,431], stromal cells
[n = 4,249]).
See also Figure S1 and Tables S1 and S2.
ll
1234 Cell 182, 1232–1251, September 3, 2020
Article

To assess the clinical outcomes of patients harboring multiple
oncogenic alterations and to determine the broader translational
impact of our findings, we utilized the MSK-Impact NSCLC data-
set (Zehir et al., 2017). Those patients whose tumor showed
greater than or equal to 2 mutations from the tier 1 COSMIC mu-
tation set detected in the scRNA-seq profiling (mutation high)
had significantly lower overall survival (OS) compared to those
patients whose tumor had less than 2 COSMIC tier 1 mutations
AB C
Figure 2. scRNA-seq Infers Patient Mutational Status and Reveals a Complex Mutational Landscape in Cancer Cells
(A) Clinical characteristics of the 44 NSCLC samples in which at least one cancer cell was identified. Co lumns indicate clinically identified mutated gene, treatment
response time point (TN, RD, PD), biopsy site, and primary or metastatic sample origin, respectively.
(B and C) Cancer cell mutational landscape for each patient sample as determined by scRNA-seq represented as a binarized heatmap across driver genes (B) and
COMSIC tier 1 genes (C). Red indicates the presence of mutation while blue indicates that no mutation was identified.
See also Figure S2 and Tables S2 and S3.
ll
Cell 182, 1232–1251, September 3, 2020 1235
Article

A
E
K
N
Q
L
O
R
M
P
S
FG
I
H
J
BCD
(legend on next page)
ll
1236 Cell 182, 1232–1251, September 3, 2020
Article

(mutation low) (p < 0.01; Figure S2J). Thus, scRNA-seq analysis
can provide increased granularity into cancer cell genomic het-
erogeneity and provides insight into the transcriptionally repre-
sented mutational landscape.
Transcriptional Differences between TN and RD Cancer
Cells Detected by scRNA-Seq Analysis Reveal Cell-
State-Specific Biological Programs
We hypothesized that defining the biological programs activated
in cancer cells during therapy response may identify signaling
pathways that promote adaptation and survival of cancer cells
that comprise RD during initial treatment. We compared the tran-
scriptional profiles of individual cancer cells obtained from tumor
samples from TN to RD (Table S4) and focused on the 629 signif-
icantly (p < 0.05) upregulated genes in RD cancer cells as a proxy
for evidence of pathway activation. We found numerous genes
associated with cancer-associated pathways (Table S5). Impor-
tantly, we found that RD cancer cells expressed decreased pro-
liferation marker genes compared to TN and PD, consistent with
the expectation that during targeted treatment persisting cancer
cells are generally less proliferative (Figure S3A) (Hsiao
et al., 2019).
Interestingly, we identified an alveolar cell gene-expression
signature composed of 17 established gene markers of alveolar
cells (Vieira Braga et al., 2019;Wade et al., 2006)thatshowed
significantly increased expression in RD versus TN time points
(p < 0.0001; Figure 3A; Figure S3B; Table S2). Alveolar cells are
comprised of alveolar type 1 (AT1) and type 2 (AT2) subtypes
and form the lining of the lung alveoli. AT2 cells produce surfac-
tants and can act as stem-like progenitor cells, which become
active and proliferate in the setting of diverse types of lung injury
and are suspected to be the cell of origin in oncogene-driven
lung cancers (Desai et al., 2014;Hanna and Onaitis, 2013;Nabhan
et al., 2018). AT1 cells are the dominant population in alveoli and
mediate gas exchange and, when injured or dying, can release
proliferation and regenerative signals (Desai et al., 2014). AT1 cells
contain two population subtypes HOPX
+
/IGFBP2
+
and HOPX
+
/
IGFBP2
–
, the latter representing the cell population which main-
tains cellular plasticity and can proliferate as well as trans-differen-
tiate into AT2 cells allowing for tissue regeneration after injury
(Wang et al., 2018).The alveolar signature we detected in the can-
cer cells at RD includes both AT1- and AT2-associated genes (Ta-
ble S2), including AQP4,SFTPB/C/D,CLDN18,FOXA2,NKX2-1,
and PGC for AT2 cells (Desai et al., 2014;Liu et al., 2003;Nabhan
et al., 2018;Wade et al., 2006;Xu et al., 2016;Zhou et al., 2018)
and AGER,HOPX,andIGFBP2 for AT1 cells (Nabhan et al.,
2018;Serveaux-Dancer et al., 2019;Figure S3B). Additionally,
the alveolar cell state we identified in cancer cells was not derived
from mis-annotated non-cancer alveolar cells within our cancer
cell populations (Figure S3C).
We validated the activation of the alveolar cell signature at RD
using orthogonal approaches. First, we used an established pre-
clinical model consisting of patient-derived EGFR mutant
NSCLCs (PC9) (Lee et al., 1985) to develop analogs of the TN,
RD, and PD clinical states. Using RT-PCR, we measured the
expression of NKX2-1, a hallmark alveolar cell signature gene
upregulated in RD clinical samples and found a significantly
higher (p < 0.001) expression in the persister state cells
compared to control and acquired resistance state cells (Fig-
ure 3B). This suggests that the alveolar signature identified
from the clinical scRNA-seq analysis can be reproduced under
controlled conditions in vitro. Furthermore, immunohistochem-
ical (IHC) analysis showed induction of AQP4 protein expression,
another marker of the alveolar cell signature, at the plasma
Figure 3. Differential Gene-Expression Analysis between Treatment Time Points Reveals Treatment Stage-Specific Transcriptional
Signatures
(A) Boxplots showing the expression level of the alveolar signature across treatment time points as well as non-cancerous AT2 cells from our cohort. ***p < 0.001.
(B) Fold-change expression of NKX2-1 as quantified by RT-PCR in EGFR mutant PC9 cells after specified treatment duration (see STAR Methods), ***p < 0.001.
(C) Representative IHC images of TN, RD, and PD tumor tissue sections stained for AQP4 demonstrating increased expression at the RD time point. Red arrows
indicate cancer cells of interest. Scale bars correspond to 50 mm.
(D) Kaplan-Meier plot of the relationship between the alveolar signature and patient OS within the TGCA dataset. Patients were stratified by signature expressio n
quartile (Q1 = 128, Q2 = 127, Q3 = 128, Q4 = 127), where Q1 is the lowest expression and Q4 is the highest expression.
(E and F) Representative IHC images of TN, RD, and PD tumor tissue sections stained for SU SD2 (E) and CTNNB1 (F) demonstrating increased expression at the
RD time point. Red arrows indicate example regions of interest. Scale bars correspond to 50 mm.
(G–J) Treatment response upon inhibition of b-catenin activity in EGFR mutant PC9 cells and ALK fusion-positive H3122 cells. Relative viability is shown as
percent confluency compared to DMSO control. PC9 cells were treated with XAV-939 (G) or PRI-724 (H) with or without the combination of 50nM osimertinib.
H3122 were treated with XAV-939 (I) or PRI-724 (J) with or without the combination of 50 nM alectinib. p values were calculated for all end points (day 6) values
compared to single agent TKI. Error bars represent mean ±standard error of the mean (SEM), n = 4 technical replicates. *p < 0.05, **p < 0.01.
(K) Boxplots showing the expression levels of the kynurenine signature expression across different treatment time points. ***p < 0.001.
(L) Fold-change expression of QPRT as quantified by RT-PCR in PC9 cells after treatment with osimertinib as in (B) (see STAR Methods) (AR), *p < 0.05.
(M) Kaplan-Meier plot of the relationship between the kynurenine signature and patient OS within the TGCA dataset. Patients were stratified by signature
expression quartile (Q1 = 128, Q2 = 127, Q3 = 128, Q4 = 127), where Q1 is the lowest expression and Q4 is the highest expression.
(N) Boxplot showing the expression levels of the plasminogen activation pathway signature across different treatment time points.
(O) Boxplot showing the expression levels of the SERPINE1 across different treatment time points.
(P) Kaplan-Meier plot of the relationship between the plasminogen activating pathway signature and patient OS within the TGCA dataset, respectively. Patients
were stratified by signature expression quartile (Q1 = 128, Q2 = 127, Q3 = 128, Q4 = 127), where Q1 is the lowest expression and Q4 is the highest expression.
(Q) Kaplan-Meier plot of the relationship between SERPINE1 expression and patient OS within the TGCA dataset, respectively. Patients were stratified by
signature expression quartile (Q1 = 128, Q2 = 127, Q3 = 128, Q4 = 127), where Q1 is the lowest expression and Q4 is the highest expression.
(R) Boxplot showing the expression levels of the gap-junction signature across treatment time points.
(S) Kaplan-Meier plots of relationships between the gap-junction signature and patient OS within the TGCA dataset. Patients were stratified by signature
expression quartile (Q1 = 128, Q2 = 127, Q3 = 128, Q4 = 127), where Q1 is the lowest expression and Q4 is the highest expression.
See also Figures S3,S4, and S5 and Tables S2,S4,S5, and S6.
ll
Cell 182, 1232–1251, September 3, 2020 1237
Article

membrane of RD clinical samples compared to both TN and PD
clinical samples (Figure 3C; Figure S3D).
We tested whether the alveolar signature is a clinically
relevant biomarker of patient survival in the TCGA lung adeno-
carcinoma bulk RNA-seq dataset (https://www.cancer.gov/
about-nci/organization/ccg/research/structural-genomics/tcga,
Cancer Genome Atlas Research et al., 2013). We found a signif-
icant (p < 0.0001) association between high expression of our
alveolar signature in these tumors and improved patient OS
when compared to patients whose tumors showed a lower alve-
olar expression signature (Figure 3D; Table S6).
These findings support the assertion that there is a distinct
alveolar gene-expression signature characterizing RD cancer
cells, associated with improved patient survival. A plausible
model is that the identified alveolar signature that is activated
in RD cancer cells reflects a cell injury and repair signature, remi-
niscent of non-cancerous AT1 and AT2 cells (Nabhan et al.,
2018;Wang et al., 2018). Increased expression of this signature
could lead to repair and escape of cell death during treatment to
support cancer cell persistence, while at the same time consti-
tuting a less aggressive malignant state. This is consistent with
the notion that RD represents a persister cell state observed in
preclinical models of slow-cycling cancer cells that survive
without rapid proliferation (as in Figure S3A), as a prelude to
the onset of aggressive tumor progression upon absolute drug
resistance (Hata et al., 2016).
The molecular details of the alveolar and cell injury repair signa-
ture are notable. In our RD cohort, the WNT/b-catenin-associated
pathway genes SUSD2 and CAV1 exhibited increased expression
(Table S5). We used IHC analysis of both SUSD2 and CTNNB1
(b-catenin) protein expression (Figures S3E and S3F) to validate
the observed transcriptional changes. In agreement with our
scRNA-seq findings, we found significantly increased membrane
SUSD2 and significantly increased nuclear CTNNB1 (b-catenin) in
the RD state compared with both TN and PD. Additionally, the
comparison of nuclear localization of CTNNB1 in a unique series
of paired TN and RD samples obtained from EGFR(AZ) or
ROS1(NC) patients treated with neoadjuvant TKI on one of two
clinical trials (osimertinib: NCT03433469 or crizotinib:
NCT03088930) is shown in Figure S3G. SUSD2 is an activated
downstream target of the WNT pathway (Umeda et al., 2018;Xu
et al., 2018b), while CAV1 can promote nuclear localization of
b-catenin (CTNNB1) and transcriptional activation of the WNT/
b-catenin pathway (Yu et al., 2014). In NSCLCs, the WNT/b-cate-
nin signaling pathway contributes to tumorigenesis (Juan et al.,
2014;Nakayama et al., 2014;Pacheco-Pinedo et al., 2011), repair,
and regeneration after cell injury (Huch et al., 2013;Tammela et al.,
2017). The self-renewal and injury response in AT2 cells specif-
ically can utilize the WNT/b-catenin signaling pathway (Nabhan
et al., 2018;Stewart, 2014). Additionally, in EGFR mutant NSCLC
activation of the WNT/b-catenin pathway may limit EGFR inhibitor
response and may promote survival of a persister cell population
during EGFR inhibitor therapy in vitro (Arasada et al., 2018;Blakely
et al., 2017;Casa
´s-Selves et al., 2012;Nakayama et al., 2014).
Overall, the RD state is characterized by signals of cellular injury
and survival, which act, in part, through the WNT/b-catenin
pathway, which may be therapeutically targetable(Krishnamurthy
and Kurzrock, 2018).
The clinical data suggest that WNT/b-catenin activation is
engaged early during treatment to facilitate the development of
RD and drug tolerant persister cells during primary EGFR or
ALK targeted therapy. To further explore the therapeutic poten-
tial of the WNT pathway findings, we utilized patient-derived PC9
cells as an EGFR mutant NSCLC model and H3122 cells as a
model for ALK fusion-driven NSCLCs. We tested the hypothesis
that upfront blockade of WNT/b-catenin signaling together with
oncogenic EGFR or ALK would decrease the number of cells
surviving initial treatment and increase the depth of response
from the outset of therapy. Parental cells were treated with an
IC
50
(inhibitor concentration yielding a 50% decrease in cell
number) dose of the appropriate EGFR or ALK TKI (osimertinib
or alectinib, respectively). Two different WNT/b-catenin pathway
inhibitors XAV939 and PRI-724 in four previously reported con-
centrations or combination therapy thereof were tested. Our
in vitro results support our hypothesis by demonstrating that
the upfront inhibition of the WNT/b-catenin pathway in combina-
tion with the cognate TKI led to a significant and dose dependent
decrease in cell confluency and increased depth of response
(Figures 3G–3J; Figures S3H–S3O).
Transcriptional Differences between TN and PD Cancer
Cells Reveal Immune Modulation and Cellular Invasion
as Key Features of Cancer Progression
When comparing cancer cells from TN and PD samples, we
found 901 differentially upregulated genes in PD cancer cells
(Table S4). Within those genes, we identified genes involved in
the kynurenine pathway and multiple genes and pathways asso-
ciated with oncogenesis and inflammation (Table S5).
We observed a significant (p < 0.0001) increase in the expres-
sion of IDO1,KYNU,andQPRT genes involved in the kynurenine
pathway, in PD and TN cancer cells (Figure 3K; Figure S3P;
Table S2) Expression of these genes can result in immunosup-
pressive behavior (Triplett et al., 2018) indicating that cancer
cells within PD tumors may directly inhibit the activity of the im-
mune system. The identification of this pathway as a mediator of
immune suppression within PD tumors has important potential
therapeutic implications, as IDO1 is upregulated in many can-
cers (Cheong and Sun, 2018;Hornya
´k et al., 2018;Liu et al.,
2018). Multiple clinical trials have attempted to block this
pathway using IDO1 inhibitors as a monotherapy as well as in
combination with immune checkpoint inhibitors or hormone
therapy (Ricciuti et al., 2019), albeit with limited success.
QPRT also exhibited increased expression (p < 0.05) specif-
ically at acquired resistance of EGFR inhibitor treatment (i.e.,
the analog of clinical PD) in our in vitro model using PC9 cells
(Figure 3L), reinforcing the assertion that this pathway is indica-
tive of cancer progression under the selective pressure of
treatment.
To further demonstrate the clinical relevance of the kynure-
nine pathway, we again used the TCGA lung adenocarcinoma
RNA-seq dataset. Higher tumor expression of this signature
was a biomarker of worse OS in patients (p < 0.05; Figure 3M;
Table S6). This is consistent with the notion that activation of
this pathway leads to immunosuppression and an inability of
the immune system to effectively surveil and eradicate can-
cer cells.
ll
1238 Cell 182, 1232–1251, September 3, 2020
Article

scRNA-seq Profiles of Cancer Cells Change from RD
to PD
We compared cancer cells from RD and PD patient samples to
elucidate the differences that occur during the outgrowth of PD
from RD and found a total of 2,182 genes which had significantly
(p < 0.001) increased expression in either RD or PD (N
RD
= 1,121,
N
PD
= 1,061) (Table S4). Among the differentially overexpressed
genes at RD were genes associated with the alveolar cell signa-
ture, cell growth, differentiation, cell motility, and tumor suppres-
sion (Table S5). RD cancer cells overexpress surfactant genes
(SFTPB/C/D and SFTA3), which are part of the alveolar cell
signature (Figure 3A; Figure S3B) (Desai et al., 2014;Treutlein
et al., 2014;Wang et al., 2018). Furthermore, NKX2-1 and NFIX
were overexpressed in RD cancer cells and are associated
with decreased cell motility (Ge et al., 2018;Rahman et al.,
2017;Winslow et al., 2011). Low expression of NKX2-1 leads
to loss of differentiation and enhanced tumor seeding ability
(Winslow et al., 2011). The collective findings arising from this
and the previous RD cancer cell analyses suggest that an
injury-repair and regenerative cell state may promote cancer
cell indolence, increased tumor control, and improved clinical
outcomes.
By contrast, PD cancer cells differentially overexpressed
genes associated with invasion, cell-to-cell communication, dif-
ferentiation, and immune modulation (Table S5). Several genes
in the plasminogen activation pathway were significantly overex-
pressed (ANXA2,PLAT,PLAUR,PLAU) (Figure 3N) along with
the plasminogen inhibitor SERPINE1 (PAI1)(p < 0.0001, Fig-
ure 3O; Figure S3Q).ANXA2 and PLAUR are the receptor pro-
teins in the plasminogen activation cascade and involved in
inflammation, angiogenesis, invasion, and metastasis, via degra-
dation of the extracellular matrix (Kubala et al., 2018;Zhu et al.,
2017). Signaling is initiated when ANXA2 or PLAU binds to PLAT
(uTa) or PLAU (uPa), respectively. Plasminogen is then degraded
to plasmin through the activity of PLAT and/or PLAU leading to
activation of metalloproteinases and degradation of fibrin. SER-
PINE1 shows increased expression in a number of cancer sub-
types and plays important roles in cell adhesion, invasion, tumor
vascularization, radio-resistance, and immunosuppression (Ku-
bala et al., 2018;Zhu et al., 2017). High expression of the plas-
minogen activation signature correlated with worse patient OS
(p < 0.01) within the TCGA lung adenocarcinoma RNA-seq data-
set and cohort (Figure 3P; Table S6). Similarly, in this indepen-
dent dataset high expression of SERPINE1 was associated
with worse OS (p < 0.05) (Figure 3Q; Table S6). EGFR inhibitor
therapy can induce expression of SERPINE1 and EGFR mutant
patients with greater than 2-fold induction of SERPINE1 (PAI1)
plasma levels during EGFR inhibitor treatment demonstrated
shorter progression-free survival (Arasada et al., 2018). Collec-
tively, our scRNA-seq findings shed light on the clinical rele-
vance and potential role of the plasminogen activation cascade
in inferior clinical outcomes and targeted therapy resistance.
Additionally, we found several gap-junction proteins differen-
tially overexpressed in PD cancer cells compared to RD cancer
cells (p < 0.0001, Figure 3R; Table S2). Gap-junction proteins
(e.g., connexins) are integral membrane proteins that allow for
cytosolic exchange of ions, metabolites and secondary messen-
gers between cells (Aasen et al., 2016;Sinyuk et al., 2018). While
some have been identified as tumor suppressors, we found that
high expression of GJB2/3/5 (Figure S3R; Table S6) was linked to
worse survival in the TCGA lung adenocarcinoma RNA-seq da-
taset (p < 0.001) (Figure 3S). These collective findings suggest
a pro-tumor effect not only in our cohort but also in NSCLCs
more generally.
Within cancer cells, we identified a rich complexity of clinically
relevant, expressed mutations that may impact therapy
response. Furthermore, evaluation of transcriptional profiles of in-
dividual cancer cells across different treatment time points iden-
tified several clinically relevant cell-state changes (Figure S4A).
We found that the identified treatment time point signatures
largely persist irrespective of the type of oncogenic driver muta-
tion (EGFR or ALK)(Figures S4B–S4K) and of biopsy site (primary
or metastatic) (data not shown). While we found these cancer cell
signatures are robust, it is important to acknowledge that there is
patient heterogeneity among samples (Figures S4L–S4O).
Longitudinal scRNA-seq Analysis of an Individual
Patient’s Tumor during Treatment
Obtaining consecutive clinical tumor biopsies from individual
advanced-stage lung cancer patients before and during treat-
ment is challenging given that most tumors regress by 50% or
greater, albeit incompletely, during TKI treatment (Camidge
et al., 2019;Soria et al., 2018). Nevertheless, we obtained sam-
ples from the same primary tumor site from 3 treatment time
points from a patient (TH226) whose tumor contained a standard
EGFR exon 19 deletion oncogenic mutation and was treated with
the EGFR inhibitor osimertinib (Figures S5A–S5C). In all 3 bi-
opsies, we identified by scRNA-seq RNA expression of the
EGFR exon 19 driver mutation in the cancer cells and several
other mutations of interest (Figure S5D).
When comparing TH226 to the rest of the scRNA-seq dataset,
we found overlapping differentially expressed genes and signa-
tures (Table S4;Figures S5E–S5H). Intriguingly, we also found
numerous genes associated with squamous cell differentiation
(KRT16,KRT14,KRT6A,KRT5,CLCA2,PKP1,ANXA8,DSG3)
overexpressed at PD compared to TN and RD time points (p <
0.0001, Figure S5I, Tables S4 and S5)(Ben-Hamo et al., 2013;
Chao et al., 2006;Goodwin et al., 2017). This is particularly inter-
esting given that the patient’s lung tumor biopsy at PD demon-
strated a histologic shift to squamous cell carcinoma from that
of prior biopsies that showed pure adenocarcinoma histology
(Figure S5C). Histologic transformation to squamous cell carci-
noma is a mechanism of EGFR inhibitor resistance in EGFR
mutant NSCLCs (Izumi et al., 2018;Jukna et al., 2016). Thus,
scRNA-seq has the power to provide a high resolution, gene
and pathway level view of biological and histological plasticity
that arises during cancer drug treatment.
Inversion of Myeloid and Lymphoid Infiltration within the
TME at Progressive Disease Compared to RD
We next addressed the evolution of the TME during targeted
treatment. Immune cells (n = 13,431) were clustered and anno-
tated (Figure 4A; Tables S1 and S4). In contrast to clusters of
cancer cells, which were predominantly patient specific (Figures
S2C and S2D), immune cell-type clusters showed low patient
occupancy (Figure 4B). This is consistent with the expectation
ll
Cell 182, 1232–1251, September 3, 2020 1239
Article

A
C
D
EF
BFigure 4. Changes in the Composition of the
Tumor Microenvironment within Each Tumor
(A) t-SNE plot of all immune cells colored by immune
cell type.
(B) Patient occupancy for each immune cell type.
(C) Fractional changes for each immune cell type
across the three treatment states. Error bars indi-
cate the 95% confidence interval for the calculated
relative frequencies. *p < 0.01 using a chi-square
test of independence.
(D) Representative in situ immunofluorescence im-
ages of changes from TN to RD and TN to PD in
tumor tissue sections from two separate samples at
two separate time points; AZ003 (TN and RD),
TH281 (TN and PD). Scale bars correspond to
50 mm.
(E) Quantification of fractional changes of macro-
phages across treatment time points from the im-
ages in (D) and Figure S5F.
(F) Quantification of fractional changes of
T-cells across treatment time points from the im-
ages in (D) and Figure S5F.
See also Figure S6.
ll
1240 Cell 182, 1232–1251, September 3, 2020
Article

of finding common immune cell phenotypes across patients and
samples.
We compared the immune cell composition across all 3 time
points, expressed as the correlation between fractional immune
cell abundance vectors. The immune composition within RD was
the most dissimilar from the other two treatment states (r = 0.78
versus TN samples, r = 0.82 versus PD samples, Pearson’s
correlation coefficient) (Figure S6A). Across all treatment time
points, T cells and macrophages were the dominant cell popula-
tions and demonstrated an inversion in relative abundance dur-
ing tumor response and resistance to treatment, a finding we
examined further as described below (Figure 4C). T cells
comprised a larger fraction of all immune cells within the TME
at RD compared to TN or PD samples (27% T cells TN, 46%
RD, 31% PD). Macrophage infiltration followed the inverse
pattern, with a decrease in macrophages at RD compared to
TN and PD (37% macrophages TN, 21% RD, 37% PD).
In 2 patients, we examined immune cells from available
matched tumor biopsies obtained at different treatment time
points (TH226 and TH266, Figures S6B and S6C, respectively).
In 2 tumor biopsies available for patient TH266, both macro-
phages and T cells showed reduction in the fraction of macro-
phages and an increase in the fraction of T cells from TN to
RD, findings which match the entire cohort (Figure S6D).
TH226 exhibited a similar pattern with the fraction of macro-
phages decreasing at RD after initiation of treatment and
increasing again at PD (Figure S6E). We validated our findings
on tissue samples using immunofluorescence staining (Figures
4D–4F; Figure S6F). Additionally, we deconvoluted TCGA bulk
transcriptome data for NSCLCs into fractions of immune cells
types (see STAR Methods) and found that TCGA samples with
high fractions of macrophages had significantly worse OS (p <
0.01) (Figure S6G). This supports the clinical relevance of our ob-
servations and is consistent with prior reports associating
macrophage infiltration with poor prognosis in patients who un-
dergo surgical resection of early-stage NSCLCs (Chen et al.,
2005;Zhang et al., 2011), as well as with worse progression-
free survival during EGFR TKI therapy (Chung et al., 2012).
These findings are particularly intriguing given their similarity
to melanoma tumors treated with PD-1 inhibitor (Riaz et al.,
2017), albeit here in the distinct context of oncoprotein-targeted
therapy in lung cancer. Specifically, an increase in the number of
CD8
+
T cells and natural killer (NK) cells and a decrease in M1
macrophages were observed in melanoma during PD-1 inhibi-
tion. There may be common responses in NK/T cells and macro-
phages during treatment across different tumor histologies and
treatments. Hence, conserved approaches to targeting RD
across different cancer subtypes and therapeutic modalities
may exist, an area for future investigation.
An IDO1-Expressing Macrophage Population Is
Enriched at PD
Macrophages from lung tumor biopsies (n = 1,379) were clus-
tered into 5 distinct groups (Figure S7A) followed by differential
gene expression in each resulting cluster (Figure S7B; Table
S4). In addition, we calculated the fraction of cells originating
from each of the three treatment groups in each of the 5 macro-
phage clusters (Figure 5A).
Cluster MF0, which was slightly enriched in TN cells, was char-
acterized by expression of genes associated with an immuno-
suppressive phenotype (C1QC,GPNMB,APOE,TREM2,
FOLR2)(Cochain et al., 2018;Zhang et al., 2019;Zhou et al.,
2017)(Figure 5B; Figure S7B). Clusters MF1 and MF3 were
enriched at RD. Macrophage cluster MF1 expressed features
associated previously with tumor-infiltrating myeloid derived
suppressor cells (FCN1,VCAN,S100A8,S100A9)(Zhang
et al., 2019) and with THBS1 and PTX3, which are associated
with resolution of inflammation, wound healing, and with inhibi-
tion of IL-1b(Bouhlel et al., 2007;Faz-Lo
´pez et al., 2016;Marti-
nez and Gordon, 2014;Puig-Kro
¨ger et al., 2009;Shiraki et al.,
2016;Stein et al., 2016;Zhang et al., 2019)(Figure 5B; Fig-
ure S7B). Cluster MF3 expressed genes associated with pro-in-
flammatory response to tissue damage (CLEC2D,IL7R,OGT)
and with promoting inflammatory signaling (FYN,DUSP4,
FOXO1)(Bao et al., 2018;Fan et al., 2010;Lai et al., 2020;Mkad-
dem et al., 2017;Moriwaki and Asahi, 2017). Macrophages in this
cluster express CCL5, a cytokine that has previously been asso-
ciated with promotion of residual HER2
+
breast cancer survival
following HER2-targeted therapy (Walens et al., 2019). Cluster
MF4 consisted of proliferating myeloid cell populations
(TOP2A,MKI67) and did not significantly differ between groups.
Macrophages at PD were overrepresented in group MF2 (Fig-
ure 5A) and expressed pro-inflammatory cytokines CXCL9,
CXCL10, and CXCL11 (Figure 5B; Figure S7B), which favor
lymphocyte recruitment into the TME (Nagarsheth et al., 2017).
Top differentially expressed genes in this population also
included the guanylate-binding family proteins GBP1 and
GBP5, which are induced in IFN-g-activated macrophages and
promote inflammatory signaling within the innate immune sys-
tem via inflammasome assembly (Shenoy et al., 2012)(Figure 5B;
Figure S7B). Despite the expression of pro-inflammatory genes
within the MF2 macrophages, the top differentially expressed
gene within this group of PD-specific macrophages was IDO1
(Figure 5B). IDO1 is induced by inflammation within the TME
and promotes a tolerogenic environment through immunosup-
pressive myeloid cell populations, regulatory T cell (Treg) differ-
entiation, and an immunosuppressive cytokine milieu (Munn and
Mellor, 2016).
An Immunosuppressive T Cell Phenotype Is
Predominant within the TME at PD
T cells and NK cells (n = 2,226) were analyzed in the same
manner as macrophages and resulted in 5 distinct T/NK cell pop-
ulations (Figure 5C; Figure S7C). These included two populations
(TC0, TC4) enriched in TN samples and 2 populations (TC1, TC2)
enriched at PD (Figures 5C and 5D; Figure S7D). There was a
high overall fraction of T cells in RD tumors (Figure 4C), and there
was no single T cell cluster that demonstrated an excess of
T cells in RD (Figure 5C).
Both TN and PD T cells demonstrated a relative decrease in
T cell infiltration (Figure 4C). The T cells which were present in
the TN state were enriched for T cell populations TC4 and TC0.
TC4 expressed markers consistent with a natural killer (NK) or
natural killer T cell (NKT) phenotype, including NK cell markers
(KIR2DL3,FCGR3A) as well as moderate expression of T cell
markers (CD3,CD8). TC0 reflected a naive-like CD8
+
phenotype
ll
Cell 182, 1232–1251, September 3, 2020 1241
Article

log10(CPM) log10(CPM)
MF0
MF1
MF2
MF3
MF4
MF0
MF1
MF2
MF3
MF4
MF0
MF1
MF2
MF3
MF4
TC0
TC1
TC2
TC3
TC4
TC0
TC1
TC2
TC3
TC4
TC0
TC1
TC2
TC3
TC4
A
C
E
B
D
Figure 5. Immune Cell Subpopulations
Demonstrate Unique Transcriptional Profiles
within Each Treatment Time Point
(A) Fraction of cells belonging to each treatment
stage for each macrophage cluster in Figure S6.
Error bars indicate the 95% confidence interval for
the calculated relative frequencies. *p < 0.01 using
chi-square test of independence.
(B) Violin plots showing the expression level distri-
bution of notable individual genes.
(C) Fraction of cells belonging to each treatment
stage for each T cell cluster in Figure S6. Error bars
indicate the 95% confidence interval for the calcu-
lated relative frequencies. *p < 0.01 using chi-
square test of independence.
(D) Violin plots showing the expression level distri-
bution of notable individual genes.
(E) Graphical summary of immune microenviron-
ment changes across treatment time points.
See also Figure S7 and Tables S4 and S5.
ll
1242 Cell 182, 1232–1251, September 3, 2020
Article

with expression of CCR7,IL7R, and SELL (Figure 5D; Figure S7D)
(van der Leun et al., 2020). While overall T cell infiltrate remained
limited at PD (Figure 4C), there was relative enrichment for T cell
phenotypes with immunosuppressive features, including T cell
clusters TC1 and TC2 (Figure 5C). TC1 was identified as a
T cell cluster with a dysfunctional or exhausted phenotype, char-
acterized by expression of the inhibitory receptors PDCD1
(which encodes for the PD-1 protein) and CTLA4 (Wherry and
Kurachi,2015)(Figure 5D). TC2 was composed of Treg cells (ex-
pressing FOXP3,IL2RA). Consistent with a relatively immuno-
suppressive environment, there was additionally reduction at
PD in infiltration by the NK/NKT cell cluster TC4.
Tumor biopsies obtained at RD revealed the presence of a
more pro-inflammatory, ‘‘hot,’’ TME, which was absent in TN
or PD biopsy samples as manifested by increased overall pro-
portion of T cells and reduced infiltration by regulatory T cells
(TC2) (Figure 5E). Compared to the PD state, at RD there were
fewer dysfunctional T cells (TC1) and greater NK/NKT cell
(TC4) infiltration (Figure 5D; Figure S7D). A population of prolifer-
ating tumor infiltrating T cells was shared across all treatment
states (TC3) and was slightly enriched within PD samples. This
T cell population was characterized by both cytotoxic pheno-
types (CD8,GZYMB) and PDL1/CTLA4 expression and may
reflect a pre-dysfunctional cytotoxic T cell population (van der
Leun et al., 2020).
In summary, both theTN and PD TME were characterizedby the
relative predominance of macrophage over T cell infiltration; how-
ever, the phenotypic characteristics of these infiltrating immune
cells differ between the two groups. At PD, there was infiltration
by an IDO1
+
macrophage population, of proliferating regulatory
T cells, and of dysfunctional T cells which were minimally present
at TN and RD. In contrast,the TN state was characterized by a pre-
dominance of more classically immunosuppressive M2-like mac-
rophages (Sica et al., 2008). By distinction, in RD there was
increased infiltration of T cell populations without dysfunctional
or immunosuppressive gene-expression patterns and decreased
immunosuppressive macrophage infiltration (Figure 5E).
DISCUSSION
There remains an incomplete catalog of single-cell transcrip-
tional data that can be used to understand cell states and the
therapy-induced evolution of biological heterogeneity of dis-
eases such as cancer, particularly for advanced-stage solid ma-
lignancies. Patients with metastatic disease do not routinely
receive surgical resection as part of their treatment. Thus, tech-
niques for single-cell profiling that require larger amounts of tis-
sue are not suitable for the interrogation of tissue samples from
metastatic disease (Lambrechts et al., 2018;Schelker et al.,
2017). Our scRNA-seq analyses of advanced-stage NSCLC bi-
opsies obtained at different treatment time points from patients
elucidate the rich mutational and transcriptional diversity within
individual tumor samples and the dynamic changes in the tran-
scriptional profiles of cancer cells and the TME composition dur-
ing treatment. Our findings provide a roadmap that highlights the
underlying cellular ecosystem and mechanisms that can inform
efforts to better treat oncogene-driven cancers. Our study offers
a rare view of the clinically relevant biological processes that
characterize RD, which is a treatment phase that is infrequently
captured in human solid malignancies.
Our scRNA-seq data revealed widespread intra-tumoral het-
erogeneity in oncogenic alterations that are expressed in cancer
cells (Figure 2) by demonstrating expression of not only the puta-
tive oncogenic driver but also additional oncogenic mutations
(Figure 6, #1). This provides a potential explanation for why com-
plete responses to treatment are rare. Tumors harbor the appro-
priate genetic framework and evolutionary playbook to evolve
resistance. These ‘‘hard-wired’’ properties can go undetected
by current bulk sampling analysis. Tumor resilience and evolution
during therapy are bolstered by the therapy-induced transcrip-
tional plasticity that we demonstrated by scRNA-seq profiling.
We uncovered transcriptional signatures specific to different
treatment time points and clinical states (Figures 3 and Figure 6,
#3, #5, #6, and #7). The majority of these signatures were bio-
markers of significantly worse OS and were most pronounced
at PD. Conversely, we found the alveolar cell signature was en-
riched at RD and was associated with improved survival. This
signature exhibited features consistent with cellular plasticity
and injury response, perhaps indicating a treatment-induceda-
daptive phenotype that permits the survival of cancer cells, albeit
in a less aggressive state (Wang et al., 2018). Our data also high-
light a connection from the alveolar cell signature to the WNT/
b-catenin pathway as a mechanism of injury response and
regeneration. Though the WNT/b-catenin pathway is potentially
therapeutically targetable, (Krishnamurthy and Kurzrock, 2018),
it will be critical to determine how to best modulate this pathway
to impact residual cancer cell survival for clinical benefit.
A general principle our data highlight is that by employing tar-
geted treatments that take advantage of specific cell states we
may be able to engineer cancer (or TME) cell fate(s) to improve
therapeutic responses in metastatic solid malignancies. If de-
ployed at the appropriate time, treatments that target liabilities
of a specific cell state or prevent further adaptation may help
improve patient survival by constraining continued tumor evolu-
tion toward complete drug resistance (Table 1).
The recognition of targetable oncogene-driven NSCLCs as a
subset of lung cancer that is poorly responsive to current check-
point inhibitor immunotherapies (Gainor et al., 2016;Mazieres
et al., 2019) necessitates an i mproved understanding of t he immu-
nologic milieu in this patient population. We found a relatively low
T cell infiltration in the TME of TN and PD patients (Figures 5Cand
5D), consistent with prior reports of low cytotoxic T cell infiltration
in treatment-naive EGFR mutant NSCLCs (Gainor et al., 2016)and
with an association between EGFR activation and an immunosup-
pressive phenotype in preclinical models (Akbay et al., 2013;
Jiang et al., 2019). Our results uncovered an induction of a more
inflammatory phenotype during RD on targeted therapy, hall-
marked by infiltration of T cells (Figure 6,#4)anddecreased
infiltration of immunosuppressive macrophages (Figure 5E). This
inflammatory state may represent a complement to the alveolar
cell, injury repair, and regenerative state present in the cancer
cell compartment (described above), with the potential for cross-
talk between the cancer cells and TME. These TME changes were
transient, as at PD there was enrichment for IDO1-expressing
macrophages, regulatory T cells, and other immunosuppressive
T cell populations. These are all features of an environment hostile
ll
Cell 182, 1232–1251, September 3, 2020 1243
Article

to the establishment of an effective immunologic response (Fig-
ure 6, #9 and #10). These findings confirm those reported within
preclinical models of EGFR mutant NSCLCs, which demonstrated
the potential for a transient immunostimulatory effect after initial
EGFR TKI exposure despite the immunosuppressive phenotype
observed following longer-term TKI treatment (Dominguez et al.,
2016). The induction of a more immunostimulatory phenotype
during targeted therapy (i.e., in RD) may offer a window-of-oppor-
tunity to introduce novel TME target-based combination therapies
earlier during treatment, perhaps around the time of RD in the
context of a more favorable TME to increase initial response
and consolidate the anti-tumor responsein a multi-modal thera-
peutic approach.
Given that cancer cell signaling and the TME are linked, there
may be treatment strategies that target both compartments
Figure 6. scRNA-seq Profiles Reveal Clin-
ical-State Specific Features of the Tumor
Cellular Ecosystem
Features common to all time points are shown in the
top-left quadrant and include the presence of mul-
tiple oncogenic drivers (1). Features shared in RD
and PD are shown in the top-right quadrant and
include various invasive signaling pathways (2).
Features unique at RD, shown in the bottom-right
quadrant, include the Alveolar signature (3) and
increased T cell fraction (4). Features unique to PD,
shown in the bottom-left quadrant, include upre-
gulation of the plasminogen activation pathway (5),
expression of gap-junction proteins (6), loss of tu-
mor suppressor genes (7), expression of pro-in-
flammatory chemokines (8), increased Treg popu-
lation (9), and increased kynurenine signature
expression (10).
concurrently. The kynurenine pathway is
one example. We identified increased ky-
nurenine pathway activation in cancer
cells and myeloid cells at PD (Figure 6,
#5). IDO1, as a rate-limiting enzyme in
the kynurenine pathway, can influence
diverse components of the TME including
T cell and myeloid cell populations as
well as angiogenesis in favor of immuno-
suppression (Munn and Mellor, 2016).
The use of IDO1 inhibitors as part of a
combination immunotherapy strategy
with PD1/PDL1 (programmed cell death
protein 1/programmed death-ligand 1)
checkpoint inhibitors showed promise
in early-phase studies (Siefker-Radtke
et al., 2018),yet failed to demonstrate
improved outcomes in advanced-stage
melanoma (Long et al., 2018). We demon-
strated distinct evolving TME states,
suggesting that there may be a window-
of-opportunity at which point kynurenine
pathway inhibitors may be more effective
(Figure 6;Table 1).
The scRNA-seq dataset presented here demonstrates the
feasibility of performing scRNA-seq on tumor biopsies obtained
longitudinally at clinically relevant time points during active tar-
geted treatment of advanced-stage solid malignancy patients.
The data provide a foundation to develop strategies for the elim-
ination or neutralization of RD to induce more durable responses
for patients with advanced-stage NSCLCs and potentially other
solid malignancies across different therapeutic modalities.
Limitation of Study
Limitations of our study include the number and diversity of cells
and genotypes of individual tumor biopsies due to the use of
small needle biopsies or fluid collections versus larger surgical
resections. Due to the real-world challenges of tissue acquisi-
tion, we acquired matched samples from a small number of
ll
1244 Cell 182, 1232–1251, September 3, 2020
Article

individual patients. Because RD is not sampled during standard
treatment, there were fewer samples at this disease state.
Single-cell-derived transcriptomes are relatively sparse as a
consequence of a combination of factors including transcrip-
tional stochasticity, rarity of sampling mRNA molecules, uneven
amplification of mRNA molecules during cDNA synthesis and li-
brary preparation, and sparse read coverage of library molecules
(Borel et al., 2015;Deng et al., 2014;Fan et al., 2018). These chal-
lenges limit our ability to perform a saturating mutation analysis
using single-cell data.
STAR+METHODS
Detailed methods are provided in the online version of this paper
and include the following:
dKEY RESOURCES TABLE
dRESOURCE AVAILABILITY
BLead Contact
BMaterials Availability
BData and Code Availability
dEXPERIMENTAL MODEL AND SUBJECT DETAILS
BHuman Subjects
BCell Lines
dMETHOD DETAILS
BPatient population
BSample preparation of cores and resections
BSample preparation of thoracentesis samples
BLysis plate preparation
BFACS sorting
BcDNA synthesis and library preparation
BLibrary sequencing
BSequencing libraries from 384-well plates
BImmunohistochemistry
BMultiplex Immunofluorescence
BRT PCR in vitro system gene expression
BWnt/b-catenin inhibition
dQUANTIFICATION AND STATISTICAL ANALYSIS
BAlignment and gene counts
BGeneral clustering
BEpithelial subset analysis
BCancer cell subset analysis
BSurvival analysis of cancer gene signatures
BAnalysis of immunohistochemistry
BMutation detection from scRNaseq
BFusion detection from scRNaseq
BMutational analysis of tumor cells
BSurvival analysis within the MSK-Impact data
BGeneral immune analysis
BAnalysis of multiplex Immunofluorescence
BImmune survival analysis within the TCGA
BAnalysis of RT PCR assay
BAnalysis of Wnt/b-catenin inhibition
dADDITIONAL RESOURCES
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/10.1016/j.
cell.2020.07.017.
ACKNOWLEDGMENTS
This project is supported by the NIH/NCI U54CA224081, R01CA204302,
R01CA211052, R01CA231300, R01CA169338, and U01CA217882 (to T.G.B.),
the Van Auken Foundation (to T.G.B. and C.M.B.), and Novartis Pharmaceuti-
cals (to T.G.B), Pfizer, as well as the University of California Cancer League
(United States) (to C.E.M), AstraZeneca (United Kingdom), The Damon Runyon
Cancer Research Foundation P0528804 (United States), Doris Duke Charitable
Foundation P2018110 (United States), V Foundation P0530519 (United States),
and NIH/NCI R01CA227807 (to C.M.B.), F.H. was supported by the Mildred
Scheel postdoctoralfellowship from the German CancerAid. E.A.Y is supported
by T32 HL007185 from the NHLBI. E.L.S is supported by K12 CA086913. Spe-
cial thanks to Bing Wu and Lillian Cohn for their insights and support.
AUTHOR CONTRIBUTIONS
Conceptualization, T.G.B., C.M.B., R.C.D., J.W., S.D., C.E.M., and A.M.;
Methodology, T.G.B., C.M.B., R.C.D., S.D., C.E.M., A.M., J.K.R., F.H., and
Table 1. Table of Common and Unique Features in Different
Treatment Time Points and Possible Therapeutic Approaches
ID Feature
Therapeutic
Approach References
1 Multiple
targetable
mutations
Combination
targeted
therapy
McCoach and
Bivona, 2019
2 Invasion
pathways
Targeted
inhibition
Rahman et al., 2017;
Zhang et al., 2016, 2019
3 Increased
immuno-
stimulatory
T cells
Immune
system
modulation
Ha et al., 2019; Jenkins
et al., 2018; Souza-Fonseca-
Guimaraes et al., 2019;
Valkenburg et al., 2018
4 Alveolar
signature
Targeted
inhibition
Nabhan et al., 2018;
Zhang et al., 2019
5 Plasminogen
activation
signature
Targeted
inhibition
Mahmood et al., 2018;
Zhang et al., 2019
5SERPINE1
signature
Targeted
inhibition,
immune
system
modulation
Placencio and
DeClerck, 2015
6 Gap-junction
signature
Targeted
inhibition
Mulkearns-Hubert et al.,
2019; Wu and Wang, 2019
7 Loss of tumor
suppressors
Targeting
acquired
vulnerabilities
Ding et al., 2019
8 Increased pro-
inflammatory
chemokines
Immune
system
modulation
Tokunaga et al., 2018
9 Increased
tregs
Immune
system
modulation
Tanaka and
Sakaguchi, 2019
10 Kynurenine
signature
Targeted
inhibition,
immune
system
modulation
Labadie et al., 2019
ll
Cell 182, 1232–1251, September 3, 2020 1245
Article

D.L.K.; Software Programming, A.M., L.H., S.D., and W.W.; Validation, F.H.,
D.L.K., L.C., C.E.M., A.M., S.D., P.G., E.L.S., E.A.Y., J.K.R., S.B., and K.S.
Formal Analysis, A.M., L.H. S.D., A.Z., W.T., M.T., R.S., K.A.Y., C.E.M.,
W.W., J.K.R., E.A.Y., D.L.K., and F.H.; Resources, C.E.M., C.M.B., J.K.R.,
A.M., S.D., N.N., T.L., A.U., K.J., P.K.K., E.S., Y.G., D.M.N., N.J.T., A.G., M.
Gonzalez, H.D., L.T., B.B., M. Gubens, T.J., J.R.K., D.J., E.L.S., and E.A.Y.;
Data Curation Management, A.M., C.E.M., and S.D.; Writing – Original Draft,
C.E.M., A.M., J.K.R., and S.D.; Writing – Revisions and Editing, all authors;
Visualization, A.M., C.E.M., J.K.R. S.D., C.M.B., and T.G.B.; Supervision,
S.D., C.M.B., T.G.B., J.W., R.C.D., R.G., and N.N.; Project Administration,
S.D., C.M.B., and T.G.B.; Funding Acquisition, S.D., C.M.B., T.G.B.,
and C.E.M.
DECLARATION OF INTERESTS
C.E.M., advisory board–Genentech; honoraria–Novartis, Guardant, Research
Funding: Novartis, Revolution Medicines; J.K.R., advisory board:
AstraZeneca, consulting: Takeda; E.L.S., employee – editorial contributor,
Elsevier, PracticeUpdate.com; speakers fees: Takeda, Roche/Genentech,
Physicians’ Education Resource, Medscape; Consultant: AbbVie; R.G.S.,
stock ownership in Celgene Corporation (Bristol-Myers Squibb); IP licensing:
Newomics; S.B. consults with and/or receives research funding from Pfizer,
Ideaya Biosciences and Revolution Medicines; M.G., research funding (to
institution) for Celgene, Merck, Novartis, OncoMed, Roche; R.C.D., Advisory
Board: Rain Therapeutics, Blueprint Medicines, Anchiano, Green Peptide,
Genentech/Roche, Bayer, AstraZeneca; Intellectual Property Licensing: Rain
Therapeutics, Foundation Medicine, Abbott Molecular, Black Diamond, Pearl
River, Voronoi; Stock Ownership: Rain Therapeutics; J.W., Scientific Advisory
Board member for Tenaya Therapeutics and Amgen. Founder and Consultant
of KSQ Therapeutics and Maze Therapeutics. Venture Partner of 5AM Ven-
tures; C.M.B., Consulting: Amgen, Foundation Medicine, Blueprint Medicines,
Revolution Medicines; Research Funding: Novartis, AstraZeneca, Takeda;
Institutional Research Funding: Mirati, Spectrum, MedImmune, Roche;
T.G.B., Advisor to Novartis, Astrazeneca, Revolution Medicines, Array/Pfize r,
Springworks, Strategia, Relay, Jazz, Rain and receives research funding
from Novartis and Revolution Medicines and Strategia.
Received: December 9, 2019
Revised: May 4, 2020
Accepted: July 13, 2020
Published: August 20, 2020
REFERENCES
Aasen, T., Mesnil, M., Naus, C.C., Lampe, P.D., and Laird, D.W. (2016). Gap
junctions and cancer: communicating for 50 years. Nat. Rev. Cancer 16,
775–788.
Akbay, E.A., Koyama, S., Carretero, J., Altabef, A., Tchaicha, J.H., Christen-
sen, C.L., Mikse, O.R., Cherniack, A.D., Beauchamp, E.M., Pugh, T.J., et al.
(2013). Activation of the PD-1 pathway contributes to immune escape in
EGFR-driven lung tumors. Cancer Discov. 3, 1355–1363.
Alexandrov, L.B., Nik-Zainal, S., Wedge, D.C., Aparicio, S.A.J.R., Behjati, S.,
Biankin, A.V., Bignell, G.R., Bolli, N., Borg, A., Børresen-Dale, A.-L., et al.;
Australian Pancreatic Cancer Genome Initiative; ICGC Breast Cancer Con-
sortium; ICGC MMML-Seq Consortium; ICGC PedBrain (2013). Signatures
of mutational processes in human cancer. Nature 500, 415–421.
Anders, S., Pyl, P.T., and Huber, W. (2015). HTSeq–a Python framework to
work with high-throughput sequencing data. Bioinformatics 31, 166–169.
Aran, D., Camarda, R., Odegaard, J., Paik, H., Oskotsky, B., Krings, G., Goga,
A., Sirota, M., and Butte, A.J. (2017). Comprehensive analysis of normal adja-
cent to tumor transcriptomes. Nat. Commun. 8, 1077.
Arasada, R.R., Shilo, K., Yamada, T., Zhang, J., Yano, S., Ghanem, R., Wang,
W., Takeuchi, S., Fukuda, K., Katakami, N., et al. (2018). Notch3-dependent
b-catenin signaling mediates EGFR TKI drug persistence in EGFR mutant
NSCLC. Nat. Commun. 9, 3198.
Bao, C., Wang, B., Yang, F., and Chen, L. (2018). Blockade of Interleukin-7 Re-
ceptor Shapes Macrophage Alternative Activation and Promotes Functional
Recovery After Spinal Cord Injury. Neuroscience 371, 518–527.
Ben-Hamo, R., Boue, S., Martin, F., Talikka, M., and Efroni, S. (2013). Classi-
fication of lung adenocarcinoma and squamous cell carcinoma samples based
on their gene expression profile in the sbv IMPROVER Diagnostic Signature
Challenge. Syst. Biomed. 1, 268–277.
Blakely, C.M., Watkins, T.B.K., Wu, W., Gini, B., Chabon, J.J., McCoach, C.E.,
McGranahan, N., Wilson, G.A., Birkbak, N.J., Olivas, V.R., et al. (2017). Evolu-
tion and clinical impact of co-occurring genetic alterations in advanced-stage
EGFR-mutant lung cancers. Nat. Genet. 49, 1693–1704.
Borel, C., Ferreira, P.G., Santoni, F., Delaneau, O., Fort, A., Popadin, K.Y., Gar-
ieri, M., Falconnet, E., Ribaux, P., Guipponi, M., et al. (2015). Biased allelic
expression in human primary fibroblast single cells. Am. J. Hum. Genet.
96, 70–80.
Bouhlel, M.A., Derudas, B., Rigamonti, E., Die
`vart, R., Brozek, J., Haulon, S.,
Zawadzki, C., Jude, B., Torpier, G., Marx, N., et al. (2007). PPARgamma acti-
vation primes human monocytes into alternative M2 macrophages with anti-in-
flammatory properties. Cell Metab. 6, 137–143.
Brannon, A.R., Vakiani, E., Sylvester, B.E., Scott, S.N., McDermott, G.,
Shah, R.H., Kania, K., Viale, A., Oschwald, D.M., Vacic, V., et al. (2014).
Comparative sequencing analysis reveals high genomic concordance be-
tween matched primary and metastatic colorectal cancer lesions. Genome
Biol. 15, 454.
Camidge, D.R., Pao, W., and Sequist, L.V. (2014). Acquired resistance to
TKIs in solid tumours: learning from lung cancer. Nat. Rev. Clin. Oncol. 11,
473–481.
Camidge, D.R., Dziadziuszko, R., Peters, S., Mok, T., Noe, J., Nowicka, M.,
Gadgeel, S.M., Cheema, P., Pavlakis, N., de Marinis, F., et al. (2019). Updated
Efficacy and Safety Data and Impact of the EML4-ALK Fusion Variant on the
Efficacy of Alectinib in Untreated ALK-Positive Advanced Non-Small Cell
Lung Cancer in the Global Phase III ALEX Study. J. Thorac. Oncol. 14,
1233–1243.
Cancer Genome Atlas Research, N., Weinstein, J.N., Collisson, E.A., Mills,
G.B., Shaw, K.R., Ozenberger, B.A., Ellrott, K., Shmulevich, I., Sander, C.,
and Stuart, J.M. (2013). The Cancer Genome Atlas Pan-Cancer analysis proj-
ect. Nat Genet 45, 1113–1120.
Casa
´s-Selves, M., Kim, J., Zhang, Z., Helfrich, B.A., Gao, D., Porter, C.C.,
Scarborough, H.A., Bunn, P.A., Jr., Chan, D.C., Tan, A.C., and DeGregori, J.
(2012). Tankyrase and the canonical Wnt pathway protect lung cancer cells
from EGFR inhibition. Cancer Res. 72, 4154–4164.
Cerami, E., Gao, J., Dogrusoz, U., Gross, B.E., Sumer, S.O., Aksoy, B.A., Ja-
cobsen, A., Byrne, C.J., Heuer, M.L., Larsson, E., et al. (2012). The cBio cancer
genomics portal: an open platform for exploring multidimensional cancer ge-
nomics data. Cancer Discov. 2, 401–404.
Chao, A., Wang, T.-H., Lee, Y.-S., Hsueh, S., Chao, A.-S., Chang, T.-C., Kung,
W.-H., Huang, S.-L., Chao, F.-Y., Wei, M.-L., and Lai, C.H. (2006). Molecular
characterization of adenocarcinoma and squamous carcinoma of the
uterine cervix using microarray analysis of gene expression. Int. J. Cancer
119, 91–98.
Chen, J.J., Lin, Y.C., Yao, P.L., Yuan, A., Chen, H.Y., Shun, C.T., Tsai, M.F.,
Chen, C.H., and Yang, P.C. (2005). Tumor-associated macrophages: the dou-
ble-edged sword in cancer progression. J. Clin. Oncol. 23, 953–964.
Cheong, J.E., and Sun, L. (2018). Targeting the IDO1/TDO2-KYN-AhR
Pathway for Cancer Immunotherapy - Challenges and Opportunities. Trends
Pharmacol. Sci. 39, 307–325.
Chung, F.T., Lee, K.Y., Wang, C.W., Heh, C.C., Chan, Y.F., Chen, H.W., Kuo,
C.H., Feng, P.H., Lin, T.Y., Wang, C.H., et al. (2012). Tumor-associated mac-
rophages correlate with response to epidermal growth factor receptor-tyro-
sine kinase inhibitors in advanced non-small cell lung cancer. Int. J. Cancer
131, E227–E235.
Chung, W., Eum, H.H., Lee, H.-O., Lee, K.-M., Lee, H.-B., Kim, K.-T., Ryu, H.S.,
Kim, S., Lee, J.E., Park, Y.H., et al. (2017). Single-cell RNA-seq enables
ll
1246 Cell 182, 1232–1251, September 3, 2020
Article

comprehensive tumour and immune cell profiling in primary breast cancer.
Nat. Commun. 8, 15081.
Cochain, C., Vafadarnejad, E., Arampatzi, P., Pelisek, J., Winkels, H., Ley, K.,
Wolf, D., Saliba, A.E., and Zernecke, A. (2018). Single-Cell RNA-Seq Reveals
the Transcriptional Landscape and Heterogeneity of Aortic Macrophages in
Murine Atherosclerosis. Circ. Res. 122, 1661–1674.
Darmanis, S., Sloan, S.A., Zhang, Y., Enge, M., Caneda, C., Shuer, L.M., Hay-
den Gephart, M.G., Barres, B.A., and Quake, S.R. (2015). A survey of human
brain transcriptome diversity at the single cell level. Proc. Natl. Acad. Sci.
USA 112, 7285–7290.
Darmanis, S., Sloan, S.A., Croote, D., Mignardi, M., Chernikova, S., Samgha-
babi, P., Zhang, Y., Neff, N., Kowarsky, M., Caneda, C., et al. (2017). Single-
Cell RNA-Seq Analysis of Infiltrating Neoplastic Cells at the Migrating Front
of Human Glioblastoma. Cell Rep. 21, 1399–1410.
Davidson-Pilon, C., Kalderstam, J., Jacobson, N., Zivich, P., Kuhn, B., William-
son, M., Reed, S., Abdeali, J.K., Fiore-Gartland, A., Datta, D., et al. (2019).
CamDavidsonPilon/lifelines: v0.24.8 https://doi.org/10.5281/zenodo3833188.
https://github.com/CamDavidsonPilon/lifelines/.
Deng, Q., Ramsko
¨ld, D., Reinius, B., and Sandberg, R. (2014). Single-cell RNA-
seq reveals dynamic, random monoallelic gene expression in mammalian
cells. Science 343, 193–196.
DePristo, M.A., Banks, E., Poplin, R., Garimella, K.V., Maguire, J.R., Hartl, C.,
Philippakis, A.A., del Angel, G., Rivas, M.A., Hanna, M., et al. (2011). A frame-
work for variation discovery and genotyping using next-generation DNA
sequencing data. Nat. Genet. 43, 491–498.
Desai, T.J., Brownfield, D.G., and Krasnow, M.A. (2014). Alveolar
progenitor and stem cells in lung development, renewal and cancer. Nature
507, 190–194.
Ding, H., Zhao, J., Zhang, Y., Yu, J., Liu, M., Li, X., Xu, L., Lin, M., Liu, C., He, Z.,
et al. (2019). Systematic Analysis of Drug Vulnerabilities Conferred by Tumor
Suppressor Loss. Cell Rep 27, 3331–3344.e3336.
Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut,
P., Chaisson, M., and Gingeras, T.R. (2013). STAR: ultrafast universal RNA-seq
aligner. Bioinformatics 29, 15–21.
Doebele, R.C., Pilling, A.B., Aisner, D.L., Kutateladze, T.G., Le, A.T.,
Weickhardt, A.J., Kondo, K.L., Linderman, D.J., Heasley, L.E., Franklin,
W.A., et al. (2012). Mechanisms of resistance to crizotinib in patients
with ALK gene rearranged non-small cell lung cancer. Clin. Cancer Res.
18, 1472–1482.
Dominguez, C., Tsang, K.Y., and Palena, C. (2016). Short-term EGFR blockade
enhances immune-mediated cytotoxicity of EGFR mutant lung cancer cells:
rationale for combination therapies. Cell Death Dis. 7, e2380.
Enge, M., Arda, H.E., Mignardi, M., Beausang, J., Bottino, R., Kim, S.K., and
Quake, S.R. (2017). Single-Cell Analysis of Human Pancreas Reveals Tran-
scriptional Signatures of Aging and Somatic Mutation Patterns. Cell 171,
321–330.
Fan, W., Morinaga, H., Kim, J.J., Bae, E., Spann, N.J., Heinz, S., Glass, C.K.,
and Olefsky, J.M. (2010). FoxO1 regulates Tlr4 inflammatory pathway signal-
ling in macrophages. EMBO J. 29, 4223–4236.
Fan, J., Lee, H.-O., Lee, S., Ryu, D.-e., Lee, S., Xue, C., Kim, S.J., Kim, K., Bar-
kas, N., Park, P.J., et al. (2018). Linking transcriptional and genetic tumor het-
erogeneity through allele analysis of single-cell RNA-seq data. Genome Res.,
gr.228080.228117.
Faz-Lo
´pez, B., Morales-Montor, J., and Terrazas, L.I. (2016). Role of Macro-
phages in the Repair Process during the Tissue Migrating and Resident Hel-
minth Infections. BioMed Res. Int. 2016, 8634603.
Fedchenko, N., and Reifenrath, J. (2014). Different approaches for interpreta-
tion and reporting of immunohistochemistry analysis results in the bone tissue -
a review. Diagn. Pathol. 9, 221.
Finak, G., McDavid, A., Yajima, M., Deng, J., Gersuk, V., Shalek, A.K.,
Slichter, C.K., Miller, H.W., McElrath, M.J., Prlic, M., et al. (2015). MAST:
a flexible statistical framework for assessing transcriptional changes and
characterizing heterogeneity in single-cell RNA sequencing data. Genome
Biol. 16, 278.
Flaherty, K.T., Infante, J.R., Daud, A., Gonzalez, R., Kefford, R.F., Sosman, J.,
Hamid, O., Schuchter, L., Cebon, J., Ibrahim, N., et al. (2012). Combined BRAF
and MEK inhibition in melanoma with BRAF V600 mutations. N. Engl. J. Med.
367, 1694–1703.
Forbes, S.A., Beare, D., Boutselakis, H., Bamford, S., Bindal, N., Tate, J., Cole,
C.G., Ward, S., Dawson, E., Ponting, L., et al. (2017). COSMIC: somatic cancer
genetics at high-resolution. Nucleic Acids Res. 45 (D1), D777–D783.
Gainor, J.F., Shaw, A.T., Sequist, L.V., Fu, X., Azzoli, C.G., Piotrowska, Z.,
Huynh, T.G., Zhao, L., Fulton, L., Schultz, K.R., et al. (2016). EGFR Mutations
and ALK Rearrangements Are Associated with Low Response Rates to PD-1
Pathway Blockade in Non-Small Cell Lung Cancer: A Retrospective Analysis.
Clin. Cancer Res. 22, 4585–4593.
Gao, J., Aksoy, B.A., Dogrusoz, U., Dresdner, G., Gross, B., Sumer, S.O., Sun,
Y., Jacobsen, A., Sinha, R., Larsson, E., et al. (2013). Integrative analysis of
complex cancer genomics and clinical profiles using the cBioPortal. Sci.
Signal. 6, pl1.
Ge, J., Dong, H., Yang, Y., Liu, B., Zheng, M., Cheng, Q., Peng, L., and Li, J.
(2018). NFIX downregulation independently predicts poor prognosis in lung
adenocarcinoma, but not in squamous cell carcinoma. Future Oncol. 14,
3135–3144.
Gerlinger, M., Rowan, A.J., Horswell, S., Math, M., Larkin, J., Endesfelder, D.,
Gronroos, E., Martinez, P., Matthews, N., Stewart, A., et al. (2012). Intratumor
heterogeneity and branched evolution revealed by multiregion sequencing.
N. Engl. J. Med. 366, 883–892.
Goodwin, J., Neugent, M.L., Lee, S.Y., Choe, J.H., Choi, H., Jenkins, D.M.R.,
Ruthenborg, R.J., Robinson, M.W., Jeong, J.Y., Wake, M., et al. (2017). The
distinct metabolic phenotype of lung squamous cell carcinoma defines selec-
tive vulnerability to glycolytic inhibition. Nat. Commun. 8, 15503.
Ha, D., Tanaka, A., Kibayashi, T., Tanemura, A., Sugiyama, D., Wing, J.B., Lim,
E.L., Teng, K.W.W., Adeegbe, D., Newell, E.W., et al. (2019). Differential control
of human Treg and effector T cells in tumor immunity by Fc-engineered anti-
CTLA-4 antibody. Proc Natl Acad Sci 116, 609–618.
Haas, B.J., Dobin, A., Li, B., Stransky, N., Pochet, N., and Regev, A. (2019). Ac-
curacy assessment of fusion transcript detection via read-mapping and de
novo fusion transcript assembly-based methods. Genome Biol. 20, 213.
Haderk, F., Olivas, V., and Bivona, T.G. (2019). Immunohistochemistry to
Study YAP in Human Tissue Samples. Methods Mol. Biol. 1893, 89–95.
Hanna, J.M., and Onaitis, M.W. (2013). Cell of origin of lung cancer.
J. Carcinog. 12,6.
Hata, A.N., Niederst, M.J., Archibald, H.L., Gomez-Caraballo, M., Siddiqui,
F.M., Mulvey, H.E., Maruvka, Y.E., Ji, F., Bhang, H.E., Krishnamurthy Radhak-
rishna, V., et al. (2016). Tumor cells can follow distinct evolutionary paths to
become resistant to epidermal growth factor receptor inhibition. Nat. Med.
22, 262–269.
Hornya
´k, L., Dobos, N., Koncz, G., Kara
´nyi, Z., Pa
´ll, D., Szabo
´, Z., Halmos, G.,
and Sze
´kvo
¨lgyi, L. (2018). The Role of Indoleamine-2,3-Dioxygenase in Cancer
Development, Diagnostics, and Therapy. Front. Immunol. 9, 151.
Hrustanovic, G., and Bivona, T.G. (2015). RAS-MAPK in ALK targeted therapy
resistance. Cell Cycle 14, 3661–3662.
Hsiao, C.J., Tung, P., Blischak, J.D., Burnett, J., Barr, K., Dey, K.K., Stephens,
M., and Gilad, Y. (2019). Characterizing and inferring quantitative cell cycle
phase in single-cell RNA-seq data analysis. Genome Res. 30, 611–621.
Huch, M., Bonfanti, P., Boj, S.F., Sato, T., Loomans, C.J.M., van de Wetering,
M., Sojoodi, M., Li, V.S.W., Schuijers, J., Gracanin, A., et al. (2013). Unlimited
in vitro expansion of adult bi-potent pancreas progenitors through the Lgr5/R-
spondin axis. EMBO J. 32, 2708–2721.
Izumi, H., Yamasaki, A., Ueda, Y., Sumikawa, T., Maeta, H., Nakamoto, S., and
Shimizu, E. (2018). Squamous Cell Carcinoma Transformation from EGFR-
mutated Lung Adenocarcinoma: A Case Report and Literature Review. Clin.
Lung Cancer 19, e63–e66.
ll
Cell 182, 1232–1251, September 3, 2020 1247
Article

Jenkins, R.W., Barbie, D.A., and Flaherty, K.T. (2018). Mechanisms of resis-
tance to immune checkpoint inhibitors. British Journal of Cancer 118, 9–16.
Jerby-Arnon, L., Shah, P., Cuoco, M.S., Rodman, C., Su, M.J., Melms, J.C.,
Leeson, R., Kanodia, A., Mei, S., Lin, J.R., et al. (2018). A Cancer Cell Program
Promotes T Cell Exclusion and Resistance to Checkpoint Blockade. Cell 175,
984–997.
Jiang, L., Guo, F., Liu, X., Li, X., Qin, Q., Shu, P., Li, Y., and Wang, Y. (2019).
Continuous targeted kinase inhibitors treatment induces upregulation of PD-
L1 in resistant NSCLC. Sci. Rep. 9, 3705.
Juan, J., Muraguchi, T., Iezza, G., Sears, R.C., and McMahon, M. (2014).
Diminished WNT -> b-catenin -> c-MYC signaling is a barrier for malignant pro-
gression of BRAFV600E-induced lung tumors. Genes Dev. 28, 561–575.
Jukna, A., Montanari, G., Mengoli, M.C., Cavazza, A., Covi, M., Barbieri, F.,
Bertolini, F., and Rossi, G. (2016). Squamous Cell Carcinoma ‘‘Transforma-
tion’’ Concurrent with Secondary T790M Mutation in Resistant EGFR-Mutated
Adenocarcinomas. J. Thorac. Oncol. 11, e49–e51.
Kadara, H., Fujimoto, J., Yoo, S.Y., Maki, Y., Gower, A.C., Kabbout, M., Gar-
cia, M.M., Chow, C.W., Chu, Z., Mendoza, G., et al. (2014). Transcriptomic ar-
chitecture of the adjacent airway field cancerization in non-small cell lung can-
cer. J. Natl. Cancer Inst. 106, dju004.
Kassambara, A., Kosinsk, M., and Biecek, P. (2019). survminer: Drawing Sur-
vival Curves using ‘ggplot20, R.p.v. 0.4.7. https://CRANR-projectorg/
package=survminer.
Kharchenko, P.V., Silberstein, L., and Scadden, D.T. (2014). Bayesian
approach to single-cell differential expression analysis. Nat. Methods 11,
740–742.
Kim, Y., Lee, B., Shim, J.H., Lee, S.-H., Park, W.-Y., Choi, Y.-L., Sun, J.-M.,
Ahn, J.S., Ahn, M.-J., and Park, K. (2019). Concurrent Genetic Alterations Pre-
dict the Progression to Target Therapy in EGFR-Mutated Advanced NSCLC.
J. Thorac. Oncol. 14, 193–202.
Kolde, R. (2019). Pretty Heatmaps, R.p.v. 1.0.12. https://CRANR-projectorg/
package=pheatmap.
Krishnamurthy, N., and Kurzrock, R. (2018). Targeting the Wnt/beta-catenin
pathway in cancer: Update on effectors and inhibitors. Cancer Treat. Rev.
62, 50–60.
Kubala, M.H., Punj, V., Placencio-Hickok, V.R., Fang, H., Fernandez, G.E.,
Sposto, R., and DeClerck, Y.A. (2018). Plasminogen Activator Inhibitor-1 Pro-
motes the Recruitment and Polarization of Macrophages in Cancer. Cell Rep.
25, 2177–2191.
Labadie, B.W., Bao, R., and Luke, J.J. (2019). Reimagining IDO Pathway Inhi-
bition in Cancer Immunotherapy via Downstream Focus on the Tryptophan-
Kynurenine-Aryl Hydrocarbon Axis. Clin Cancer Res 25, 1462–1471.
Lai, J.J., Cruz, F.M., and Rock, K.L. (2020). Immune Sensing of Cell Death
through Recognition of Histone Sequences by C-Type Lectin-Receptor-2d
Causes Inflammation and Tissue Injury. Immunity 52, 123–135.
Lambrechts, D., Wauters, E., Boeckx, B., Aibar, S., Nittner, D., Burton, O., Bas-
sez, A., Decaluwe
´, H., Pircher, A., Van den Eynde, K., et al. (2018). Phenotype
molding of stromal cells in the lung tumor microenvironment. Nat. Med. 24,
1277–1289.
Lawrence, M.S., Stojanov, P., Polak, P., Kryukov, G.V., Cibulskis, K., Siva-
chenko, A., Carter, S.L., Stewart, C., Mermel, C.H., Roberts, S.A., et al.
(2013). Mutational heterogeneity in cancer and the search for new cancer-
associated genes. Nature 499, 214–218.
Lee, Y.C., Saijo, N., Sasaki, Y., Takahashi, H., Sakurai, M., Ishihara, J., Hoshi,
A., Chen, K.M., and Hamburger, A.W. (1985). Clonogenic patterns of human
pulmonary adenocarcinoma cell lines (PC-9, PC-13 and PC-14) and how
they influence the results of test for chemosensitivity to cisplatin in the human
tumor clonogenic assay. Jpn. J. Clin. Oncol. 15, 637–644.
Lee, S.Y., Haq, F., Kim, D., Jun, C., Jo, H.-J., Ahn, S.-M., and Lee, W.-S.
(2014). Comparative genomic analysis of primary and synchronous metastatic
colorectal cancers. PLoS ONE 9, e90459.
Liu, H., Hooper, S.B., Armugam, A., Dawson, N., Ferraro, T., Jeyaseelan, K.,
Thiel, A., Koukoulas, I., and Wintour, E.M. (2003). Aquaporin gene expression
and regulation in the ovine fetal lung. J. Physiol. 551, 503–514.
Liu, M., Wang, X., Wang, L., Ma, X., Gong, Z., Zhang, S., and Li, Y. (2018). Tar-
geting the IDO1 pathway in cancer: from bench to bedside. J. Hematol. Oncol.
11, 100.
Long, G.V., Dummer, R., Hamid, O., Gajewski, T., Caglevic, C., Dalle, S., Ara-
nce, A., Carlino, M.S., Grob, J.-J., Kim, T.M., et al. (2018). Epacadostat (E) plus
pembrolizumab (P) versus pembrolizumab alone in patients (pts) with unre-
sectable or metastatic melanoma: Results of the phase 3 ECHO-301/KEY-
NOTE-252 study. Lancet Oncol. 20, 1083–1097.
Lovly, C.M., Iyengar, P., and Gainor, J.F. (2017). Managing Resistance to
EFGR- and ALK-Targeted Therapies. Am. Soc. Clin. Oncol. Educ. Book 37,
607–618.
Mahmood, N., Mihalcioiu, C., and Rabbani, S.A. (2018). Multifaceted Role of
the Urokinase-Type Plasminogen Activator (uPA) and Its Receptor (uPAR):
Diagnostic, Prognostic, and Therapeutic Applications. Front Oncol 8,24.
Maier, M.J. (2015). Companion Package to the Book ‘‘R: Einfu
¨hrung durch an-
gewandte Statistik’’, R.p.v. 0.9.3. http://CRANR-projectorg/package=REdaS.
Martinez, F.O., and Gordon, S. (2014). The M1 and M2 paradigm of macro-
phage activation: time for reassessment. F1000Prime Rep. 6,13.
Mazieres, J., Drilon, A., Lusque, A., Mhanna, L., Cortot, A.B., Mezquita, L.,
Thai, A.A., Mascaux, C., Couraud, S., Veillon, R., et al. (2019). Immune check-
point inhibitors for patients with advanced lung cancer and oncogenic driver
alterations: results from the IMMUNOTARGET registry. Ann. Oncol. 30,
1321–1328.
McCoach, C.E., and Bivona, T.G. (2019). Engineering Multidimensional Evolu-
tionary Forces to Combat Cancer. Cancer Discov 9, 587–604.
McCoach, C.E., Blakely, C.M., Banks, K.C., Levy, B., Chue, B.M., Raymond,
V.M., Le, A.T., Lee, C.E., Diaz, J., Waqar, S.N., et al. (2018). Clinical Utility of
Cell-Free DNA for the Detection of ALK Fusions and Genomic Mechanisms
of ALK Inhibitor Resistance in Non-Small Cell Lung Cancer. Clin. Cancer
Res. 24, 2758–2770.
McGinnis, C.S., Murrow, L.M., and Gartner, Z.J. (2019). DoubletFinder:
Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest
Neighbors. Cell Syst. 8, 329–337.
Mkaddem, S.B., Murua, A., Flament, H., Titeca-Beauport, D., Bounaix, C., Da-
nelli, L., Launay, P., Benhamou, M., Blank, U., Daugas, E., et al. (2017). Lyn and
Fyn function as molecular switches that control immunoreceptors to direct ho-
meostasis or inflammation. Nat. Commun. 8, 246.
Mok, T.S., Wu, Y.-L., Thongprasert, S., Yang, C.-H., Chu, D.-T., Saijo, N., Sun-
paweravong, P., Han, B., Margono, B., Ichinose, Y., et al. (2009). Gefitinib or
carboplatin-paclitaxel in pulmonary adenocarcinoma. N. Engl. J. Med. 361,
947–957.
Moriwaki, K., and Asahi, M. (2017). Augmented TME O-GlcNAcylation Pro-
motes Tumor Proliferation through the Inhibition of p38 MAPK. Mol. Cancer
Res. 15, 1287–1298.
Mulkearns-Hubert, E.E., Torre-Healy, L.A., Silver, D.J., Eurich, J.T., Bayik, D.,
Serbinowski, E., Hitomi, M., Zhou, J., Przychodzen, B., Zhang, R., et al. (2019).
Development of a Cx46 Targeting Strategy for Cancer Stem Cells. Cell Rep 27,
1062–1072.e1065.
Munn, D.H., and Mellor, A.L. (2016). IDO in the Tumor Microenvironment:
Inflammation, Counter-Regulation, and Tolerance. Trends Immunol. 37,
193–207.
Nabhan, A.N., Brownfield, D.G., Harbury, P.B., Krasnow, M.A., and Desai, T.J.
(2018). Single-cell Wnt signaling niches maintain stemness of alveolar type 2
cells. Science 359, 1118–1123.
Nagarsheth, N., Wicha, M.S., and Zou, W. (2017). Chemokines in the cancer
microenvironment and their relevance in cancer immunotherapy. Nat Rev Im-
munol 17, 559–572.
Nakayama, S., Sng, N., Carretero, J., Welner, R., Hayashi, Y., Yamamoto, M.,
Tan, A.J., Yamaguchi, N., Yasuda, H., Li, D., et al. (2014). b-catenin contributes
ll
1248 Cell 182, 1232–1251, September 3, 2020
Article

to lung tumor development induced by EGFR mutations. Cancer Res. 74,
5891–5902.
Neftel, C., Laffy, J., Filbin, M.G., Hara, T., Shore, M.E., Rahme, G.J., Richman,
A.R., Silverbush, D., Shaw, M.L., Hebert, C.M., et al. (2019). An Integrative
Model of Cellular States, Plasticity, and Genetics for Glioblastoma. Cell 178,
835–849.
Newman, A.M., Liu, C.L., Green, M.R., Gentles, A.J., Feng, W., Xu, Y., Hoang,
C.D., Diehn, M., and Alizadeh, A.A. (2015). Robust enumeration of cell subsets
from tissue expression profiles. Nat Methods 12, 453–457.
Pacheco-Pinedo, E.C., Durham, A.C., Stewart, K.M., Goss, A.M., Lu, M.M.,
Demayo, F.J., and Morrisey, E.E. (2011). Wnt/b-catenin signaling accelerates
mouse lung tumorigenesis by imposing an embryonic distal progenitor pheno-
type on lung epithelium. J. Clin. Invest. 121, 1935–1945.
Patel, A.P., Tirosh, I., Trombetta, J.J., Shalek, A.K., Gillespie, S.M., Wakimoto,
H., Cahill, D.P., Nahed, B.V., Curry, W.T., Martuza, R.L., et al. (2014). Single-
cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma.
Science 344, 1396–1401.
Picelli, S., Bjo
¨rklund, A.K., Faridani, O.R., Sagasser, S., Winberg, G., and
Sandberg, R. (2013). Smart-seq2 for sensitive full-length transcriptome
profiling in single cells. Nat. Methods 10, 1096–1098.
Picelli, S., Faridani, O.R., Bjo
¨rklund, A.K., Winberg, G., Sagasser, S., and
Sandberg, R. (2014). Full-length RNA-seq from single cells using Smart-
seq2. Nat. Protoc. 9, 171–181.
Placencio, V.R., and DeClerck, Y.A. (2015). Plasminogen Activator Inhibitor-1
in Cancer: Rationale and Insight for Future Therapeutic Testing. Cancer Res
75, 2969–2974.
Puig-Kro
¨ger, A., Sierra-Filardi, E., Domı
´nguez-Soto, A., Samaniego, R., Cor-
cuera, M.T., Go
´mez-Aguado, F., Ratnam, M., Sa
´nchez-Mateos, P., and Corbı
´,
A.L. (2009). Folate receptor beta is expressed by tumor-associated macro-
phages and constitutes a marker for M2 anti-inflammatory/regulatory macro-
phages. Cancer Res. 69, 9395–9403.
Puram, S.V., Tirosh, I., Parikh, A.S., Patel, A.P., Yizhak, K., Gillespie, S.,
Rodman, C., Luo, C.L., Mroz, E.A., Emerick, K.S., et al. (2017). Single-Cell
Transcriptomic Analysis of Primary and Metastatic Tumor Ecosystems in
Head and Neck Cancer. Cell 171, 1611–1624.
Python. (2015). Python: A dynamic, open source programming language, Py-
thon Software Foundation.
R. (2013). R: A language and environment for statistical computing, R Founda-
tion for Statistical Computing.
Rahman, N.I.A., Abdul Murad, N.A., Mollah, M.M., Jamal, R., and Harun, R.
(2017). NFIX as a Master Regulator for Lung Cancer Progression. Front. Phar-
macol. 8, 540.
Riaz, N., Havel, J.J., Makarov, V., Desrichard, A., Urba, W.J., Sims, J.S., Hodi,
F.S., Martı
´n-Algarra, S., Mandal, R., Sharfman, W.H., et al. (2017). Tumor and
Microenvironment Evolution during Immunotherapy with Nivolumab. Cell 171,
934–949.
Ricciuti, B., Leonardi, G.C., Puccetti, P., Fallarino, F., Bianconi, V., Sahebkar,
A., Baglivo, S., Chiari, R., and Pirro, M. (2019). Targeting indoleamine-2,3-di-
oxygenase in cancer: Scientific rationale and clinical evidence. Pharmacol.
Ther. 196, 105–116.
Rotow, J., and Bivona, T.G. (2017). Understanding and targeting resistance
mechanisms in NSCLC. Nat. Rev. Cancer 17, 637–658.
RStudio (2015). RStudio: Integrated Development for R, RStudio, Inc.
Scheffler, M., Ihle, M.A., Hein, R., Merkelbach-Bruse, S., Scheel, A.H., Siema-
nowski, J., Bra
¨gelmann, J., Kron, A., Abedpour, N., Ueckeroth, F., et al. (2019).
K-ras Mutation Subtypes in NSCLC and Associated Co-occuring Mutations in
Other Oncogenic Pathways. J. Thorac. Oncol. 14, 606–616.
Schelker, M., Feau, S., Du, J., Ranu, N., Klipp, E., MacBeath, G., Schoeberl, B.,
and Raue, A. (2017). Estimation of immune cell content in tumour tissue using
single-cell RNA-seq data. Nat. Commun. 8, 2032.
Schiller, H.B., Montoro, D.T., Simon, L.M., Rawlins, E.L., Meyer, K.B., Strunz,
M., Vieira Braga, F.A., Timens, W., Koppelman, G.H., Budinger, G.R.S., et al.
(2019). The Human Lung Cell Atlas: A High-Resolution Reference Map of the
Human Lung in Health and Disease. Am. J. Respir. Cell Mol. Biol. 61, 31–41.
Serveaux-Dancer, M., Jabaudon, M., Creveaux, I., Belville, C., Blondonnet, R.,
Gross, C., Constantin, J.-M., Blanchon, L., and Sapin, V. (2019). Pathological
Implications of Receptor for Advanced Glycation End-Product (AGER) Gene
Polymorphism. Dis. Markers 2019, 2067353.
Shaw, A.T., and Engelman, J.A. (2013). ALK in lung cancer: past, present, and
future. J. Clin. Oncol. 31, 1105–1111.
Shaw, A.T., Kim, D.-W., Nakagawa, K., Seto, T., Crino
´, L., Ahn, M.-J., De Pas,
T., Besse, B., Solomon, B.J., Blackhall, F., et al. (2013). Crizotinib versus
chemotherapy in advanced ALK-positive lung cancer. N. Engl. J. Med. 368,
2385–2394.
Shenoy, A.R., Wellington, D.A., Kumar, P., Kassa, H., Booth, C.J., Cresswell,
P., and MacMicking, J.D. (2012). GBP5 promotes NLRP3 inflammasome as-
sembly and immunity in mammals. Science 336, 481–485.
Sherry, S.T., Ward, M.H., Kholodov, M., Baker, J., Phan, L., Smigielski, E.M.,
and Sirotkin, K. (2001). dbSNP: the NCBI database of genetic variation. Nu-
cleic Acids Res. 29, 308–311.
Shihab, H.A., Rogers, M.F., Gough, J., Mort, M., Cooper, D.N., Day, I.N.,
Gaunt, T.R., and Campbell, C. (2015). An integrative approach to predicting
the functional effects of non-coding and coding sequence variation. Bioinfor-
matics 31, 1536–1543.
Shiraki, A., Kotooka, N., Komoda, H., Hirase, T., Oyama, J.-I., and Node, K.
(2016). Pentraxin-3 regulates the inflammatory activity of macrophages. Bio-
chem. Biophys. Rep. 5, 290–295.
Sica, A., Larghi, P., Mancino, A., Rubino, L., Porta, C., Totaro, M.G., Rimoldi,
M., Biswas, S.K., Allavena, P., and Mantovani, A. (2008). Macrophage polari-
zation in tumour progression. Semin. Cancer Biol. 18, 349–355.
Siefker-Radtke, A.O., Necchi, A., Park, S.H., Garca-Donas, J.s., Huddart, R.A.,
Burgess, E.F., Fleming, M.T., Rezazadeh, A., Mellado, B., Varlamov, S., et al.
(2018). First results from the primary analysis population of the phase 2 study
of erdafitinib (ERDA; JNJ-42756493) in patients (pts) with metastatic or unre-
sectable urothelial carcinoma (mUC) and < i > FGFR < /i > alterations
(FGFRalt). JCO 36, 4503, 4503.
Sinyuk, M., Mulkearns-Hubert, E.E., Reizes, O., and Lathia, J. (2018). Cancer
Connectors: Connexins, Gap Junctions, and Communication. Front. Oncol.
8, 646.
Soria, J.C., Ohe, Y., Vansteenkiste, J., Reungwetwattana, T., Chewaskulyong,
B., Lee, K.H., Dechaphunkul, A., Imamura, F., Nogami, N., Kurata, T., et al.;
FLAURA Investigators (2018). Osimertinib in Untreated EGFR-Mutated
Advanced Non-Small-Cell Lung Cancer. N. Engl. J. Med. 378, 113–125.
Souza-Fonseca-Guimaraes, F., Cursons, J., and Huntington, N.D. (2019). The
Emergence of Natural Killer Cells as a Major Target in Cancer Immunotherapy.
Trends Immunol 40, 142–158.
Stein, E.V., Miller, T.W., Ivins-O’Keefe, K., Kaur, S., and Roberts, D.D. (2016).
Secreted Thrombospondin-1 Regulates Macrophage Interleukin-1bProduc-
tion and Activation through CD47. Sci. Rep. 6, 19684.
Stewart, D.J. (2014). Wnt signaling pathway in non-small cell lung cancer.
J. Natl. Cancer Inst. 106,djt356.
Stuart, T., Butler, A., Hoffman, P., Hafemeister, C., Papalexi, E., Mauck, W.M.,
3rd, Hao, Y., Stoeckius, M., Smibert, P., and Satija, R. (2019). Comprehensive
Integration of Single-Cell Data. Cell 177, 1888–1902.
Tabara, K., Kanda, R., Sonoda, K., Kubo, T., Murakami, Y., Kawahara, A.,
Azuma, K., Abe, H., Kage, M., Yoshinaga, A., et al. (2012). Loss of activating
EGFR mutant gene contributes to acquired resistance to EGFR tyrosine kinase
inhibitors in lung cancer cells. PLoS ONE 7, e41017.
Tabula Muris, C.; Tabula Muris Consortium; Overall coordination; Logistical
coordination; Organ collection and processing; Library preparation and
sequencing; Computational data analysis; Cell type annotation; Writing group;
Supplemental text writing group; Principal investigators (2018). Single-cell
transcriptomics of 20 mouse organs creates a Tabula Muris. Nature 562,
367–372.
ll
Cell 182, 1232–1251, September 3, 2020 1249
Article

Tammela, T., Sanchez-Rivera, F.J., Cetinbas, N.M., Wu, K., Joshi, N.S., Helen-
ius, K., Park, Y., Azimi, R., Kerper, N.R., Wesselhoeft, R.A., et al. (2017). A Wnt-
producing niche drives proliferative potential and progression in lung adeno-
carcinoma. Nature 545, 355–359.
Tanaka, A., and Sakaguchi, S. (2019). Targeting Treg cells in cancer immuno-
therapy. Eur J Immunol 49, 1140–1146.
Tate, J.G., Bamford, S., Jubb, H.C., Sondka, Z., Beare, D.M., Bindal, N., Bout-
selakis, H., Cole, C.G., Creatore, C., Dawson, E., et al. (2019). COSMIC: the
Catalogue Of Somatic Mutations In Cancer. Nucleic Acids Res. 47 (D1),
D941–D947.
Therneau, T. (2015). A Package for Survival Analysis in R, R.p.v. 3.1-12. https://
CRANR-projectorg/package=survival.
Tickle, T., Tirosh, I., Georgescu, C., Brown, M., and Haas, B. (2019). inferCNV
of the Trinity CTAT Project. Klarman Cell Observatory (Broad Institute of MIT
and Harvard). https://github.com/broadinstitute/inferCNV.
Tirosh, I., Izar, B., Prakadan, S.M., Wadsworth, M.H., 2nd, Treacy, D., Trom-
betta, J.J., Rotem, A., Rodman, C., Lian, C., Murphy, G., et al. (2016). Dissect-
ing the multicellular ecosystem of metastatic melanoma by single-cell RNA-
seq. Science 352, 189–196.
Tokunaga, R., Zhang, W., Naseem, M., Puccini, A., Berger, M.D., Soni, S.,
McSkane, M., Baba, H., and Lenz, H.-J. (2018). CXCL9, CXCL10, CXCL11/
CXCR3 axis for immune activation – A target for novel cancer therapy. Cancer
Treatment Reviews 63, 40–47.
Treutlein, B., Brownfield, D.G., Wu, A.R., Neff, N.F., Mantalas, G.L., Espinoza,
F.H., Desai, T.J., Krasnow, M.A., and Quake, S.R. (2014). Reconstructing line-
age hierarchies of the distal lung epithelium using single-cell RNA-seq. Nature
509, 371–375.
Triplett, T.A., Garrison, K.C., Marshall, N., Donkor, M., Blazeck, J., Lamb,
C., Qerqez, A., Dekker, J.D., Tanno, Y., Lu, W.-C., et al. (2018). Reversal
of indoleamine 2,3-dioxygenase-mediated cancer immune suppression by
systemic kynurenine depletion with a therapeutic enzyme. Nat. Biotechnol.
36, 758–764.
Umeda, S., Kanda, M., Miwa, T., Tanaka, H., Tanaka, C., Kobayashi, D., Sue-
naga, M., Hattori, N., Hayashi, M., Yamada, S., et al. (2018). Expression of su-
shi domain containing two reflects the malignant potential of gastric cancer.
Cancer Med. 7, 5194–5204.
Valkenburg, K.C., de Groot, A.E., and Pienta, K.J. (2018). Targeting the tumour
stroma to improve cancer therapy. Nat Rev Clin Oncol 15, 366–381.
van der Leun, A.M., Thommen, D.S., and Schumacher, T.N. (2020). CD8
+
T cell
states in human cancer: insights from single-cell analysis. Nat. Rev. Cancer 20,
218–232.
Venteicher, A.S., Tirosh, I., Hebert, C., Yizhak, K., Neftel, C., Filbin, M.G., Hov-
estadt, V., Escalante, L.E., Shaw, M.L., Rodman, C., et al. (2017). Decoupling
genetics, lineages, and microenvironment in IDH-mutant gliomas by single-
cell RNA-seq. Science 355, eaai8478.
Vieira Braga, F.A., Kar, G., Berg, M., Carpaij, O.A., Polanski, K., Simon, L.M.,
Brouwer, S., Gomes, T., Hesse, L., Jiang, J., et al. (2019). A cellular census
of human lungs identifies novel cell states in health and in asthma. Nat.
Med. 25, 1153–1163.
Vignot, S., Lefebvre, C., Frampton, G.M., Meurice, G., Yelensky, R., Palmer,
G., Capron, F., Lazar, V., Hannoun, L., Miller, V.A., et al. (2015). Comparative
analysis of primary tumour and matched metastases in colorectal cancer pa-
tients: evaluation of concordance between genomic and transcriptional pro-
files. Eur. J. Cancer 51, 791–799.
Wade, K.C., Guttentag, S.H., Gonzales, L.W., Maschhoff, K.L., Gonzales, J.,
Kolla, V., Singhal, S., and Ballard, P.L. (2006). Gene induction during differen-
tiation of human pulmonary type II cells in vitro. Am. J. Respir. Cell Mol. Biol.
34, 727–737.
Walens, A., DiMarco, A.V., Lupo, R., Kroger, B.R., Damrauer, J.S., and Alvarez,
J.V. (2019). CCL5 promotes breast cancer recurrence through macrophage
recruitment in residual tumors. eLife 8, e43653.
Wang, Y., Tang, Z., Huang, H., Li, J., Wang, Z., Yu, Y., Zhang, C., Li, J., Dai, H.,
Wang, F., et al. (2018). Pulmonary alveolar type I cell population consists of two
distinct subtypes that differ in cell fate. Proc. Natl. Acad. Sci. USA 115,
2407–2412.
Wang, J., Xu, R., Yuan, H., Zhang, Y., and Cheng, S. (2019). Single-cell RNA
sequencing reveals novel gene expression signatures of trastuzumab treat-
ment in HER2+ breast cancer: A pilot study. Medicine (Baltimore) 98, e15872.
Wherry, E.J., and Kurachi, M. (2015). Molecular and cellular insights into T cell
exhaustion. Nat. Rev. Immunol. 15, 486–499.
Wickham, H. (2016). ggplot2: Elegant Graphics for Data Analysis (Springer-
Verlag).
Wickham, H., Francois, R., Henry, L., and Muller, K. (2020). dplyr: A Grammar
of Data Manipulation. R package, R.p.v. 0.8.5. https://CRANR-projectorg/
package=dplyr.
Winslow, M.M., Dayton, T.L., Verhaak, R.G.W., Kim-Kiselak, C., Snyder, E.L.,
Feldser, D.M., Hubbard, D.D., DuPage, M.J., Whittaker, C.A., Hoersch, S.,
et al. (2011). Suppression of lung adenocarcinoma progression by Nkx2-1. Na-
ture 473, 101–104.
Xu, Y., Mizuno, T., Sridharan, A., Du, Y., Guo, M., Tang, J., Wikenheiser-Bro-
kamp, K.A., Perl, A.T., Funari, V.A., Gokey, J.J., et al. (2016). Single-cell RNA
sequencing identifies diverse roles of epithelial cells in idiopathic pulmonary
fibrosis. JCI Insight 1, e90558.
Xu, J., Zhao, X., He, D., Wang, J., Li, W., Liu, Y., Ma, L., Jiang, M., Teng, Y.,
Wang, Z., et al. (2018a). Loss of EGFR confers acquired resistance to
AZD9291 in an EGFR-mutant non-small cell lung cancer cell line with an
epithelial-mesenchymal transition phenotype. J. Cancer Res. Clin. Oncol.
144, 1413–1422.
Wu, J.-I., and Wang, L.-H. (2019). Emerging roles of gap junction proteins con-
nexins in cancer metastasis, chemoresistance and clinical application. J Bio-
med Sci 26,8.
Xu, Y., Miao, C., Jin, C., Qiu, C., Li, Y., Sun, X., Gao, M., Lu, N., and Kong, B.
(2018b). SUSD2 promotes cancer metastasis and confers cisplatin resistance
in high grade serous ovarian cancer. Exp. Cell Res. 363, 160–170.
Xue, Y., Hou, S., Ji, H., and Han, X. (2017). Evolution from genetics to pheno-
type: reinterpretation of NSCLC plasticity, heterogeneity, and drug resistance.
Protein Cell 8, 178–190.
Yang, J.-J., Zhang, C., Zhao, J., Dai, P., Lin, G., Han, L., Chen, C., Guan, Y.-F.,
Yang, L., Yi, X., et al. (2019). ALK resistance mutations and co-occurring ge-
netic alterations to the ALK tyrosine kinase inhibitors in lung cancer. JCO 37,
e20675, e20675.
Yu, H., Shen, H., Zhang, Y., Zhong, F., Liu, Y., Qin, L., and Yang, P. (2014).
CAV1 promotes HCC cell progression and metastasis through Wnt/b-catenin
pathway. PLoS ONE 9, e106451.
Yuan, S., Norgard, R.J., and Stanger, B.Z. (2019). Cellular Plasticity in Cancer.
Cancer Discov. 9, 837–851.
Zehir, A., Benayed, R., Shah, R.H., Syed, A., Middha, S., Kim, H.R., Srinivasan,
P., Gao, J., Chakravarty, D., Devlin, S.M., et al. (2017). Mutational landscape of
metastatic cancer revealed from prospective clinical sequencing of 10,000 pa-
tients. Nat. Med. 23, 703–713.
Zhang, L., Zhou, W., Velculescu, V.E., Kern, S.E., Hruban, R.H., Hamilton, S.R.,
Vogelstein, B., and Kinzler, K.W. (1997). Gene expression profiles in normal
and cancer cells. Science 276, 1268–1272.
Zhang, Y., Xie, Z.-Y., Guo, X.-T., Xiao, X.-H., and Xiong, L.-X. (2019). Notch and
breast cancer metastasis: Current knowledge, new sights and targeted ther-
apy. Oncol Lett 18, 2743–2755.
Zhang, B., Yao, G., Zhang, Y., Gao, J., Yang, B., Rao, Z., and Gao, J. (2011).
M2-polarized tumor-associated macrophages are associated with poor prog-
noses resulting from accelerated lymphangiogenesis in lung adenocarcinoma.
Clinics (Sa
˜o Paulo) 66, 1879–1886.
Zhang, D., Dai, Y., Cai, Y., Suo, T., Liu, H., Wang, Y., Cheng, Z., and Liu, H.
(2016). PEBP4 promoted the growth and migration of cancer cells in pancre-
atic ductal adenocarcinoma. Tumour Biol 37, 1699–1705.
Zhang, Q., He, Y., Luo, N., Patel, S.J., Han, Y., Gao, R., Modak, M., Carot ta, S.,
Haslinger, C., Kind, D., et al. (2019). Landscape and Dynamics of Single Im-
mune Cells in Hepatocellular Carcinoma. Cell 179, 829–845.
ll
1250 Cell 182, 1232–1251, September 3, 2020
Article

Zhou, L., Zhuo, H., Ouyang, H., Liu, Y., Yuan, F., Sun, L., Liu, F., and Liu, H.
(2017). Glycoprotein non-metastatic melanoma protein b (Gpnmb) is highly ex-
pressed in macrophages of acute injured kidney and promotes M2 macro-
phages polarization. Cell. Immunol. 316, 53–60.
Zhou, B., Flodby, P., Luo, J., Castillo, D.R., Liu, Y., Yu, F.-X., McConnell, A.,
Varghese, B., Li, G., Chimge, N.-O., et al. (2018). Claudin-18-mediated YAP
activity regulates lung stem and progenitor cell homeostasis and tumorigen-
esis. J. Clin. Invest. 128, 970–984.
Zhu, C., Shen, H., Zhu, L., Zhao, F., and Shu, Y. (2017). Plasminogen Activator
Inhibitor 1 Promotes Immunosuppression in Human Non-Small Cell Lung Can-
cers by Enhancing TGF-B1 Expression in Macrophage. Cell. Physiol. Bio-
chem. 44, 2201–2211.
Liu, J., Lichtenberg, T., Hoadley, K.A., Poisson, L.M., Lazar, A.J., Cherniack,
A.D., Kovatich, A.J., Benz, C.C., Levine, D.A., Lee, A.V., et al.(2018). An Inte-
grated TCGA Pan-Cancer Clinical Data Resource to Drive High-Quality Sur-
vival Outcome Analytics. Cell 173, 400-416 e411.
ll
Cell 182, 1232–1251, September 3, 2020 1251
Article

STAR+METHODS
KEY RESOURCES TABLE
REAGENT or RESOURCE SOURCE IDENTIFIER
Antibodies
CD45-FITC Miltenyi Biotec Cat# 1300-080-202; RRID: AB_244234
CTNNB1 rabbit monoclonal Cell Signaling Technology Cat# 8480S; RRID: AB_11127855
AQP4 rabbit monoclonal Cell Signaling Technology Cat# 59678; RRID: AB_2799571
SUSD2 rabbit polyclonal Sigma Aldrich Cat# HPA004117; RRID: AB_1857674
PD-L1 Cell Signaling Technologies Clone E1L3N; RRID: AB_2687655
CD68 Dako Clone KP1; RRID: AB_578703
IDO Cell Signaling Technologies Clone D5J4E; RRID: AB_2636818
HLA-DR Abcam Clone CR3/43; RRID: AB_443647
CD14 Abcam Clone SP192;
Cytokeratin Dako Polyclonal Z0622; RRID: AB_2650434
CD3 Leica Clone LN10; RRID: AB_10554454
PD-1 Abcam Clone NAT105; RRID: AB_881954
CD8 Dako Clone C8/144B; RRID: AB_2075537
FoxP3 Abcam Clone 236A/E7; RRID: AB_445284
Biological Samples
Thoracentesis, resection, tumor adjacent
tumor and core biopsy samples
University of California San Francisco N/A
Chemicals, Peptides, and Recombinant Proteins
DMEM GE Life Sciences Cat# SH30081.01
Collagenase Type 2 Worthington Biochemical Cat# LS004176
RBC lysis buffer Thermo Fisher Scientific Cat# A1049201
FBS Omega Scientific, Inc. Cat# FB-11
Running buffer Miltenyi Biotec Cat# 130-091-221
BSA Miltenyi Biotec Cat# 130-091-221
Hoechst stain Thermo Fisher Scientific Cat# H3570
PI Life Technologies Cat# P3566
Sytox Blue Thermo Fisher Scientific Cat# S34867
Recombinant RNase Inhibitor Takara Bio Cat# 2313B
Triton
TM
X-100 Sigma Cat# 93443
dNTP mix Thermo Fisher Cat# R0193
ERCC RNA spike-in mix Thermo Fisher Cat# 4456740
SMARTScribe Reverse Transcriptase Takara Bio Cat# 639538
First-Strand Buffer Takara Bio Cat# 639538
DTT Bioworld Cat# 40420001-1
Betaine Sigma Cat# B0300
MgCl
2
Sigma Cat# M1028
KAPA HiFi HotStart ReadyMix Kapa Biosystems Cat# KK2602
Lambda Exonuclease NEB Cat# M0262L
Tris-HCl Thermo Fisher Cat# 15568025
qPCR Kapa Biosystems Cat# KK4923
Osimertinib Selleck Chemicals Cat# S7297
Alectinib Selleck Chemicals Cat# S2762
XAV-939 Selleck Chemicals Cat# S1180
(Continued on next page)
ll
e1 Cell 182, 1232–1251.e1–e10, September 3, 2020
Article

Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
PRI-724 Selleck Chemicals Cat# S8262
Sodium Bicarbonate Millipore Sigma Cat# S5761
Bovine Serum Albumin VWR International Cat# 97061-422
Tween-20 Millipore Sigma Cat# P9416
Sodium Azide Millipore Sigma Cat# S2002
Xylenes Fisher Chemical Cat# X5-4
Hematoxylin solution VWR International Cat# 95057-844
Target Retrieval solution, Citrate Agilent Dako Cat# S169984-2
Richard-Allan Scientific Cytoseal 60 Thermo Fisher Scientific Cat# 8310-16
Epitope Retrieval Solution 1 Lecia Cat# AR9961
Epitope Retrieval Solution 2 Lecia Cat# AR9640
Antibody Diluent Akoya Biosciences Cat# ARD1001EA
Opal Polymer HRP Ms + Rb Akoya Biosciences Cat# ARH1001EA
BOND Wash Solution Lecia Cat# AR9590
DAPI Akoya Biosciences Cat# FP1490
ProLong Diamond Antifade Mountant Thermo Fisher Scientific Cat# P36961
Critical Commercial Assays
Nextera XT Library Sample Preparation kit Illumina Cat# FC-131-1096
NextSeq 500/550 Hi Output Kit v2.5
(300 cycle)
Illumina Cat# 20024908
Novaseq S2 (300 cycle) Illumina Cat# 20012860
RT2 Profiler PCR array QIAGEN Cat# CLAH34795
Fragment analyzer kit Agilent Cat# DNF-474-0500
Tapestation D5000 Kit Agilent Cat# 5067-5593
Tapestation D5000 Tapes Agilent Cat# 5067-5592
EnVision+ Dual Link Kit Agilent Dako Cat# K406511-2
Opal IHC Multiplex Assay Perkin Elmer Cat# NEL801001KT
Deposited Data
MSK-Impact (Cerami et al., 2012;Gao et al., 2013)https://www.mskcc.org/msk-impact
TCGA TCGA Research Network; (Liu et al., 2018)https://www.cancer.gov/about-nci/
organization/ccg/research/
structural-genomics/tcga
dbSNP (Sherry et al., 2001)https://www.ncbi.nlm.nih.gov/variation/
docs/human_variation_vcf/
COSMIC (Catalogue of Somatic Mutations
in Cancer)
(Tate et al., 2019)https://cancer.sanger.ac.uk/cosmic/
download
Normal AT2 single-cell gene counts (Vieira Braga et al., 2019) GEO-GSE130148
scRNaseq NSCLC This study BioProject- PRJNA591860
Experimental Models: Cell Lines
PC9 cells ATCC N/A
H3122 cells ATCC N/A
Oligonucleotides
Oligo-dT
30
VN-50AAGCAGTGGT
ATCAACGCAGAGTACT
30
VN-30
IDT N/A
TSO-50AAGCAGTGGTA
TCAACGCAGACTACATrGrG+G-30
Exiqon N/A
IS PCR primer-
50AAGCAGTGGTATCAACGCAGAGT-30
IDT N/A
(Continued on next page)
ll
Cell 182, 1232–1251.e1–e10, September 3, 2020 e2
Article

Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
Software and Algorithms
R (R, 2013) https://www.r-project.org/
bcl2fastq Illumina https://support.illumina.com/sequencing/
sequencing_software/
bcl2fastq-conversion-software.html;
RRID:SCR_015058
STAR (Dobin et al., 2013)https://github.com/alexdobin/STAR
HTSEQ (Anders et al., 2015)https://htseq.readthedocs.io/en/master/
Rstudio (RStudio 2015)https://rstudio.com/
Seurat v3.0 (Stuart et al., 2019)https://github.com/satijalab/seurat
DoubletFinder (McGinnis et al., 2019)https://github.com/chris-mcginnis-ucsf/
DoubletFinder
inferCNV (Tickle et al., 2019)https://github.com/broadinstitute/
inferCNV.).
MAST (Finak et al., 2015)https://github.com/RGLab/MAST
pheatmap (Kolde, 2019)https://cran.r-project.org/web/packages/
pheatmap/index.html
dyplr (Wickham et al., 2020)https://cran.r-project.org/web/packages/
dplyr/index.html
ggplot2 (Wickham, 2016)https://cran.r-project.org/web/packages/
ggplot2/index.html
Reflow GRAIL https://github.com/grailbio/reflow
survival (Therneau, 2015)https://cran.r-project.org/web/packages/
survival/index.html
survminer (Kassambara et al., 2019)http://cran.r-project.org/web/packages/
surviminer/index.html
GATK HaplotypeCaller (DePristo et al., 2011) broadinstitute/gatk:4.0.11.0
fathmm (Shihab et al., 2015)https://github.com/HAShihab/fathmm
STAR-fusion (Haas et al., 2019)https://github.com/STAR-Fusion/
STAR-Fusion/wiki,
cerebra (Unpublished data) https://pypi.org/project/cerebra/
lifelines (Davidson-Pilon et al., 2019)https://github.com/CamDavidsonPilon/
lifelines/
Python v3.4 (Python, 2015)https://python.org/
REdaS (Maier, 2015) N/A
Phenochart v1.0.8 Perkin Elmer N/A
inForm v2.4.8 Akoya N/A
CellInsight Thermo Fisher Scientific N/A
BioRender BioRender N/A
Adobe Illustrator Adobe N/A
Other
Mosquito 384w Spool of 4.5mm tips TTP Labtech Cat# 4150-03010
Mantis Low Volume Chip Fisher Scientific Cat# NC1491372
Mantis High Volume Chip Fisher Scientific Cat# NC1491373
RNeasy Mini Kit QIAGEN Cat# 74104
Bioanalyzer RNA 6000 Pico kit Agilent Cat# 5067-1514
Qubit RNA HS Assay kit Thermo Fisher Scientific Cat# Q32852
First Strand Synthesis Kit QIAGEN Cat# 330401
AMPure beads Fisher Cat# A63881
(Continued on next page)
ll
e3 Cell 182, 1232–1251.e1–e10, September 3, 2020
Article

RESOURCE AVAILABILITY
Lead Contact
Further information and requests for resources should be directed to and will be fulfilled by the Lead Contact, Trever Bivona (https://
cran.r-project.org/web/packages/REdaS/index.html).
Materials Availability
This study did not generate new unique reagents.
Data and Code Availability
The dataset generated during this study is available as an NCBI BioProject #PRJNA591860. All code used to generate the results of
this study can be found on github at czbiohub/scell_lung_adenocarcinoma and czbiohub/cerebra. The below methods reference
specific code notebooks (script xx) available at czbiohub/scell_lung_adenocarcinoma to analyze data.
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Human Subjects
All patients gave informed consent for collection of clinical correlates, tissue collection, research testing under Institutional Review
Board (IRB)-approved protocols (CC13-6512 and CC17-658, NCT03433469). Patient demographics are listed in Table S1. Patient
studies were conducted according to the Declaration of Helsinki, the Belmont Report, and the U.S. Common Rule.
Cell Lines
PC9 (EGFR
exon19del
) and H3122 (EML4-ALK
v1
) cells were purchased from ATCC and grown in a 5% CO
2,
humidified atmosphere at
37C. Cultures were maintained using RPMI 1640 medium (GE Healthcare) supplemented with 10% (v/v %) fetal bovine serum
(VWR), 100 IU/mL penicillin and 100 mg/mL streptomycin (GIBCO).
METHOD DETAILS
Patient population
Formalin-fixed paraffin embedded (FFPE), frozen, and fresh tissue samples were obtained according to the safety standards of the
interventional radiologist, pulmonologist, or surgeon. Demographic and clinical history for each patient was obtained from chart re-
view and is listed in Table S1. Days until progression were determined based on imaging studies which demonstrated definitive
growth of a known tumor site or new extra-CNS metastatic deposits. Residual disease state was determined by serial imaging
demonstrating continued reduction or stability tumor with no evidence of progression. Complete details of each patient sample
acquisition are outlined in Table S1 and Figure S1A. Additionally, the timing of each sample acquisition is shown in Figure S1A.
Sample preparation of cores and resections
Tissue was first cut into small pieces and placed into a 1.5 mL tube (or multiple tubes if necessary). 1.5 mL of collagenase buffer (10mL
DMEM (GE Life Sciences, SH30081.01), 0.20 g Collagenase Type 2 (Worthington Biochemical, LS004176)) was added to the tube
and the sample was digested for 30 minutes at 37C, shaking in a thermomixer @ 800-1000 rpm. The sample was manually agitated
by pipetting up and down 15 times then returned to the thermomixer for 25 minutes. After incubation, the sample was removed from
the thermomixer, agitated again by pipetting the sample up and down 15 times before passing the sample through a 100-micron filter
(Fisherbrand, 22363548) into a new 15 mL falcon tube. The filter was washed with 1-2 mL of collagenase buffer before the sample was
spun in the centrifuge at 500xg for 10 minutes. If the resulting cell pellet was red, 0.5 mL RBC lysis buffer (Thermo Fisher Scientific,
A1049201) was added to sample tubes and allowed to sit at room temperature for 3 minutes before quenching with 1.0 mL DMEM
(GE Life Sciences, SH30081.01) + 6% FBS (Omega Scientific, Inc, FB-11) and spun in the centrifuge at 500xg for 5 minutes. Remain-
ing cells were stained with 10 ml CD45-FITC (Miltenyi Biotec, 130-080-202) and 1 ml of Hoechst stain (Thermo Fisher Scientific,
Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
100-micron filter Fisherbrand Cat# 22363548
FACS tube Falcon Cat# 14-956-3C
384-well hard-shell PCR plates BioRad Cat# HSP3901
Fisherfinest Premium Cover Glasses
(50 324 mm)
Fisher Scientific Cat# 12-548-5M
ll
Cell 182, 1232–1251.e1–e10, September 3, 2020 e4
Article

H3570). Samples incubated on ice in the dark for 20 minutes. 1mL of FACS Buffer (1:20 dilution of BSA (Miltenyi Biotec, 130-091-221)
in Running Buffer (Miltenyi Biotec, 130-091-221)) was then added to the stained cells and spun at 500xg for 10 minutes before aspi-
rating off supernatant. Cells were resuspended with 0.5 mL of FACS Buffer. PI (Life Technologies, P3566) was added immediately
prior to sorting.
Sample preparation of thoracentesis samples
Cells were filtered through a 100 mm strainer (Fisherbrand 22363548), pelleted (500xg, 5 min, 4C), and resuspended in FACS buffer.
Cells were then stained with CD45-FITC (Miltenyi Biotec, 130-080-202) for 20 min at 4C in the dark. Cells were then pelleted (500xg,
5 min, 4C) and resuspended in FACS buffer before being transferred to a FACS tube (Falcon 14-956-3C). Sytox Blue dead cell stain
(Thermo Fisher Scientific, S34867) was added immediately prior to sorting.
Lysis plate preparation
Lysis plates were created by dispensing 0.4 mL lysis buffer (0.5U Recombinant RNase Inhibitor (Takara Bio, 2313B), 0.0625% Tri-
ton
TM
X-100 (Sigma, 93443-100ML), 3.125 mM dNTP mix (Thermo Fisher, R0193), 3.125 mM Oligo-dT
30
VN (IDT,
50AAGCAGTGGTATCAACGCAGAGTACT
30
VN-30) and 1:600,000 ERCC RNA spike-in mix (Thermo Fisher, 4456740)) into 384-well
hard-shell PCR plates (Biorad HSP3901) using a Tempest liquid handler (Formulatrix). All plates were then spun down for 1 minute
at 3220xg and snap frozen on dry ice. Plates were stored at 80C until used for sorting.
FACS sorting
Cells were sorted into 384-well plates using SH800S (Sony) sorter. Cells were sorted using the ‘‘Ultra purity’’ setting on the sorter. For
a typical sort, a FACs tube containing 0.3-1ml the pre-stained cell suspension was vortexed gently and loaded onto the FACS ma-
chine. A small number of cells were flowed at low pressure to check cell concentration and amount of debris. Then the pressure was
adjusted, flow was paused, the first destination plate was unsealed and loaded. Single-cell sorting was done where half the plate was
sorted for CD45+/PI-/Hoechst+ while the second half was sorted for CD45-/PI-/Hoechst+. Immediately after sorting, plates were
sealed with a pre-labeled aluminum seal, centrifuged and flash frozen on dry ice.
cDNA synthesis and library preparation
cDNA synthesis was performed using the Smart-seq2 (Picelli et al., 2013;Picelli et al., 2014;Tabula Muris et al., 2018). Briefly, 384-
well plates containing single-cell lysates were thawed on ice followed by first strand synthesis. 0.6 mL of reaction mix (16.7 U/ml
SMARTScribe Reverse Transcriptase (Takara Bio, 639538), 1.67 U/ml Recombinant RNase Inhibitor (Takara Bio, 2313B), 1.67X
First-Strand Buffer (Takara Bio, 639538), 1.67 mM TSO (Exiqon, 50-AAGCAGTGGTATCAACGCAGACTACATrGrG+G-30), 8.33 mM
DTT (Bioworld, 40420001-1), 1.67 M Betaine (Sigma, B0300-5VL), and 10 mM MgCl
2
(Sigma, M1028-10X1ML)) was added to
each well using a Tempest liquid handler or Mosquito (TTP Labtech). Reverse transcription was carried out by incubating wells on
a ProFlex 2x384 thermal-cycler (Thermo Fisher) at 42C for 90 min and stopped by heating at 70C for 5 min.
Subsequently, 1.5 mL of PCR mix (1.67X KAPA HiFi HotStart ReadyMix (Kapa Biosystems, KK2602), 0.17 mM IS PCR primer (IDT,
50-AAGCAGTGGTATCAACGCAGAGT-30), and 0.038 U/ml Lambda Exonuclease (NEB, M0262L)) was added to each well with a
Mantis liquid handler (Formulatrix) or Mosquito, and second strand synthesis was performed on a ProFlex 2x384 thermal-cycler
by using the following program: 1. 37C for 30 minutes, 2. 95C for 3 minutes, 3. 23 cycles of 98C for 20 s, 67C for 15 s, and
72C for 4 minutes, and 4. 72C for 5 minutes.
The amplified product was diluted with a ratio of 1-part cDNA to 10 parts 10mM Tris-HCl (Thermo Fisher, 15568025). 0.6 mLof
diluted product was transferred to a new 384-well plate using the Viaflow 384 channel pipette (Integra). Illumina sequencing libraries
were prepared as described in Darmanis et al. (2015). Briefly, tagmentation was carried out on double-stranded cDNA using the Nex-
tera XT Library Sample Preparation kit (Illumina, FC-131-1096). Each well was mixed with 0.8 mL Nextera tagmentation DNA buffer
(Illumina) and 0.4 mL Tn5 enzyme (Illumina), then incubated at 55C for 10 min. The reaction was stopped by adding 0.4 mL ‘‘Neutralize
Tagment Buffer’’ (Illumina) and spinning at room temperature in a centrifuge at 3220xg for 5 min. Indexing PCR reactions were per-
formed by adding 0.4 mLof5mM i5 indexing primer, 0.4 mLof5mM i7 indexing primer, and 1.2 mL of Nextera NPM mix (Illumina). All
reagents were dispensed with the Mantis or Mosquito liquid handlers. PCR amplification was carried out on a ProFlex 2x384 thermal
cycler using the following program: 1. 72C for 3 minutes, 2. 95C for 30 s, 3. 12 cycles of 95C for 10 s, 55C for 30 s, and 72C for
1 minute, and 4. 72C for 5 minutes.
Library sequencing
Following library preparation, wells of each library plate were pooled using a Mosquito liquid handler. Pooling was followed by two
purifications using 0.7x AMPure beads (Fisher, A63881). Library quality was assessed using high sensitivity capillary electrophoresis
on a Fragment Analyzer (Agilent) or Tapestation (Agilent), and libraries were quantified by qPCR (Kapa Biosystems, KK4923) on a
CFX96 Touch Real-Time PCR Detection System (Biorad). Plate pools were normalized to 2 nM and equal volumes from library plates
were mixed together to make the sequencing sample pool.
ll
e5 Cell 182, 1232–1251.e1–e10, September 3, 2020
Article

Sequencing libraries from 384-well plates
Libraries were sequenced on the NextSeq or NovaSeq 6000 Sequencing System (Illumina) using 2 3100bp paired-end reads and
238bp or 2 312bp index reads. NextSeq runs used high output kits, whereas NovaSeq runs used either a 200 or 300-cycle kit
(Illumina, 20012860). PhiX control library was spiked in at 1%.
Immunohistochemistry
All specimens were acquired from individuals with NSCLC as noted above. 4-micron thick formalin-fixed paraffin embedded (FFPE)
human tissue sections were processed using previously published method (Haderk et al., 2019) and Agilent-Dako manufacturer rec-
ommendations were followed for antigen retrieval. All wash steps were performed at room temperature for three minutes each, unless
otherwise noted. Briefly, slides were deparaffinized in xylenes (2 washes, 5 min each), and rehydrated in graded dilutions of aqueous
ethanol (2 washes in 100% EtOH; 2 washes in 95% EtOH; 1 wash in 70% EtOH). Slides were washed once in ddH2O before being
placed in an antigen target retrieval solution, 1x pH 6.1 Citrate retrieval solution (Dako) and pressure cooked using one cycle (2 hours)
for antigen retrieval. Slides were allowed to cool to room temperature, washed three times with 1x PBS, then the tissue was blocked
for endogenous peroxidase activity for 10 minutes using 0.3% H2O2. Slides were washed three times with 1x PBS, then incubated for
1 hour in a prepared protein blocking buffer solution (1X PBS containing 1% (w/v) BSA, 15 mM sodium azide, 0.05% (w/v) Tween-20).
Slides were incubated overnight at 4C with either b-catenin (CTNNB1) rabbit monoclonal antibody (#8480S, Cell Signaling Technol-
ogy, 1:100 dilution), AQP4 rabbit monoclonal antibody (#59678, Cell Signaling Technology, 1:100 dilution), or SUSD2 rabbit poly-
clonal antibody (HPA004117, Sigma Aldrich, 1:400 dilution). The following morning, the slides were washed three times with 1x
PBS, incubated using commercial anti-rabbit and anti-mouse labeled polymer-HRP solution (Agilent Dako) for 30 minutes. Slides
were washed three times in 1x PBS before incubation with freshly prepared 3,3-diaminobenzidine chromogen solution (Agilent
Dako) for < 1 minute. Slides were washed twice in ddH2O and were counterstained using a commercial hematoxylin solution
(VWR Biosciences). Excess dye was removed using three washes in ddH2O, and the hematoxylin was developed by incubating
for 1 minute in 0.1% (w/v) sodium bicarbonate solution, and washed once in ddH2O. Tissues were dehydrated in aqueous ethanol
(2 washes in 95% EtOH; 2 washes in 100% EtOH) and incubated in xylene for 5 minutes before being coverslipped. Stained slides
were digitized using an Aperio ScanScope XT Slide Scanner (Leica Biosystems) using a 40X objective.
Multiplex Immunofluorescence
Multiplex immunofluorescence staining was performed on the Opal IHC Multiplex Assay (NEL821001KT, Akoya Biosciences).
Sequential 4 micron sections mounted on glass slides were sequentially stained for panel 1: PD-L1 (clone E1L3N, dilution 1:50,
Cell Signaling Technologies), CD68 (clone KP1, dilution 1:500, Dako), IDO (clone D5J4E, dilution 1:100, Cell Signaling Technologies),
HLA-DR (clone CR3/43, dilution 1:250, Abcam), CD14 (clone SP192, dilution 1:100, Abcam), and cytokeratin (polyclonal Z0622, dilu-
tion 1:250, Dako); or panel 2: CD3 (clone LN10, Leica), PD-1 (clone NAT105, dilution 1:100, Abcam), CD14, CD8 (clone C8/144B,
dilution 1:100, Dako), FoxP3 (clone 236A/E7, dilution 1:200, Abcam), and cytokeratin on a Bond RX autostainer (Leica Biosystems).
Slides were dewaxed (Leica), heat treated in Epitope retrieval solution 1 or 2 (AR9961/AR9640, Lecia) buffer depending on the anti-
body for 20 min at 93C, blocked in Antibody (Ab) Diluent (ARD1001EA, Akoya Biosciences), incubated for 30 min with the primary Ab,
10 min with horseradish peroxidase (HRP)-conjugated secondary polymer (anti-rabbit and anti-mouse, ARH1001EA, Akoya Biosci-
ences), and 10 min with HRP-reactive OPAL fluorescent reagents (NEL821001KT, Akoya Biosciences). Slides were washed between
staining steps with Bond Wash (AR9590, Leica) and stripped between each round of staining with heat treatment in antigen retrieval
buffer. After the final heat treatment in antigen retrieval buffer, the slides were stained with spectral DAPI (FP1490, Akoya Biosci-
ences), and coverslipped with Prolong Diamond mounting media (P36961, Thermo Fisher). Whole slide scans were acquired using
the 10x objective via the Vectra imaging system (Perkin Elmer, version 3.0).
RT PCR in vitro system gene expression
For validation of candidate gene expression via a RT2 Profiler PCR array (QIAGEN, CLAH34795), human lung cancer PC9 cells (5 3
10
5
) were treated for 48 hours (day 2) with DMSO (TN) or for 7 and 19 days with 2mM Osimertinib (Selleck Chemicals, S7297) with
replenishment of drug every 3-4 days (Persister cells that evade drug-induced apoptosis by being in a low- to no-proliferative state,
in patients this corresponds to the RD state), respectively. PD samples were derived from an acquired resistant PC9 cell line (Osi-
mertnib IC
50
=89mM), that was generated by continuous treatment with 2mM Osimertinib with replenishment of drug every 3-
4 days and presented active proliferation under drug at which time they were considered to be resistant and in the PD state. RNA
was extracted via RNeasy Mini Kit (QIAGEN, 74104). RNA quality was confirmed as RIN > 7.5 via Bioanalyzer RNA 6000 Pico kit (Agi-
lent, 5067-1514) and RNA was quantified via Qubit RNA HS Assay kit (Thermo Fisher Scientific, Q32852). A total of 400ng of RNA was
reverse transcribed using the First Strand Synthesis Kit (QIAGEN, 330401) and then loaded into a custom 384 well RT2 profiler array
(QIAGEN, CLAH34795).
Wnt/b-catenin inhibition
Small molecule inhibitors were all purchased commercially from Selleck Chemicals, and included Osimertinib (S7297), Alectinib
(S2762), XAV-939 (S1180), and PRI-724 (S8262). Dimethyl sulfoxide (DMSO) (Fisher Scientific) was used to dissolve small molecule
inhibitors according to manufacturer’s recommendations for use in in vitro experiments. PC9 and H3122 cells (5 310
3
) were seeded
ll
Cell 182, 1232–1251.e1–e10, September 3, 2020 e6
Article

in 96-well plate format (mclear CellStar, Greiner) and rested for 24 hours before treatment. Treatment included: i) DMSO, ii) tyrosine
kinase inhibitors (TKI) Osimertinib (PC9 cells) or Alectinib (H3122 cells), iii) Wnt/b-catenin inhibitors PRI-724 or XAV-939, and v) indi-
cated combination therapies of TKI and Wnt/ b-catenin inhibitors. All conditions were plated in technical quadruplicate and cells were
retreated every 3 days. At each imaging interval, cellular nuclei were stained with Hoechst 33342 (Thermo Fisher Scientific) and
scanned using a CellInsight High-Content Microscope (Version 6.4.3 Build 7204, Thermo Fisher Scientific) with a 4X objective.
QUANTIFICATION AND STATISTICAL ANALYSIS
Alignment and gene counts
Sequences from the Illumina sequencing were demultiplexed using bcl2fastq version 2.19.0.316 (Illumina). Reads were aligned using
the hg38 genome using STAR version 2.5.2b (Dobin et al., 2013) with parameters TK. Gene counts were produced using HTSEQ
version 0.6.1p1 (Anders et al., 2015) with default parameters except stranded was set to false and mode was set to intersection-
nonempty.
General clustering
Standard procedures for filtering, variable gene selection, dimensionality reduction, and clustering were performed using the Seurat
v3 (Stuart et al., 2019) in RStudio (RStudio, 2015) using R (R, 2013), where cells with fewer than 500 genes and 50,000 reads were
excluded. We used DoubletFinder (McGinnis et al., 2019) to identify potentially sorted doublet cells. 218 doublets were excluded from
further analysis. Samples with less than 10 total cells were filtered from the analysis. Counts were log-normalized, then scaled by
linear regression against the number of reads. Variable genes (Ngenes = 2,000) were selected using a threshold for dispersion,
with z-scores normalized by expression level. The variable genes were projected onto a low-dimensional subspace using principal
component analysis. The number of principal components (Npcs) were selected based on inspection of the plot of variance explained
(Npcs = 20). A shared-nearest-neighbors graph was constructed based with metric the Euclidean distance in the low-dimensional
subspace. Cells were visualized using a 2-dimensional tSNE on the same distance metric (Res = 0.5, Kparam = 30, script 03).
Cell types were assigned to each cluster of cells using the abundance of known marker genes (Table S2, script S01-03 and
script NI01).
Epithelial subset analysis
Cells previously annotated as epithelial (n = 5,581) were subset and re-clustered using methods described above and the following
parameters: Ngenes = (2,000), Npcs = 20, Res = 0.7, Kparam = 30 (script NI02). Malignant epithelial cells were identified using in-
ferCNV (Tickle et al., 2019). inferCNV which works by finding cells with large copy number variations as determined by sorting
expressed genes by their chromosomal location and applying a moving average, a sliding window of 100 genes within each chro-
mosome, to the relative expression values (Patel et al., 2014;Puram et al., 2017;Tirosh et al., 2016). All epithelial cells as well as
300 fibroblasts and 300 endothelial cells were used as input (script NI03). An additional 500 fibroblasts and 500 endothelial cells
were used as reference controls. We scored each cell for the extent of CNV signal and plotted cells on a dendrogram which was
then cut at the highest point in which all the spiked in endothelial and fibroblasts cells belonged to one cluster (k = 6, one fibroblast
control was misassigned). All cells that clustered together with spiked in controls were labeled ‘‘nontumor,’’ whereas the remaining
two clusters were labeled as ‘‘tumor.’’
Noncancerous epithelial cells (n = 1,827), as determined as those cells lacking large chromosomal aberrations from InferCNV anal-
ysis, were subset and re-clustered using the following parameters: Ngenes = (2,000), Npcs = 20, Res = 0.5, Kparam = 20 (script NI05).
Cell types were assigned to each cluster of cells using the abundance of known marker genes (Table S2) and differentially expressed
genes as found by using the Seurat function FindAllMarkers using the default Wilcoxon rank sum test.
Cancer cell subset analysis
Cancerous epithelial cells (n = 3,754), as determined as those cells harboring large chromosomal aberrations from InferCNV analysis,
were subset and re-clustered using the following parameters: Ngenes = 2,000, Npcs = 20, Res = 0.9, Kparam = 10 (script NI04). We
found the differences in gene expression between the three treatment time points (TN, RD, and PD) using the Seurat function Find-
Markers using the MAST test (Finak et al., 2015) and sample_name as the latent variable. Three separate tests were used to ascertain
the differences between: 1) TN and RD, 2) TN and PD and 3) RD and PD (Table S5). Resulting differential gene lists were then filtered to
limit patient specific effects. This is achieved by setting a threshold for non-zero expressing cells per patient (RD = 3 of RD patients
and PD = 6 of PD patients) and removing differentially expressed genes explained by less than the thresholds set. The top 100 genes
from each comparison were manually curated to evaluate for pathway activation. Decreased expression could indicate lack of detec-
tion due to the stochasticity of scRNaseq and thus for analysis of activated pathways we focused on upregulated genes. Gene sig-
natures (Table S2) were compiled using differential expressed as well as known cell marker genes. Specifically, the alveolar signature
is made of differentially expressed AT1/AT2 genes among the cancer cell time point comparisons as well has additional known AT1/
AT2 genes (Vieira Braga et al., 2019;Wade et al., 2006). The remaining signatures were identified directly from top differentially ex-
pressed genes.
ll
e7 Cell 182, 1232–1251.e1–e10, September 3, 2020
Article

To ensure that we were not misclassifying healthy AT2 cells as cancer cells, we compared the expression levels of our combined
alveolar gene signature between the three time points (TN, RD, PD) and non-cancer AT2 cells from our dataset as well as additional
non-cancer AT2 cells from an external dataset (Vieira Braga et al., 2019). Non-cancer AT2 cells from our dataset were more similar to
the external AT2 cells than any of our cancer cells across all time points (average spearman correlation coefficient = 0.65, 0.10, 0.24,
0.19, for non-cancer AT2 cells, and TN, RD, PD cancer cells respectively).
Cancer cells from EGFR and ALK driven tumor samples were subset separately. We compared all three treatment time points (TN,
RD, PD) for EGFR patients where we only compared two treatment time points for ALK (TN, PD) as only one ALK+ driven sample
represented the RD time point. We then compared the five cancer cell signatures derived from the grouped analysis (alveolar, kynur-
enine, plasminogen activation, SERPINE1, and gap junction). Pairwise wilcoxon tests were calculated between each treatment time
point (TN, RD, and PD).
To understand the PD sample heterogeneity all cancer cells from PD samples were subset. Each sample’s average expression of
genes included in gene signatures (alveolar, kynurenine, plasminogen activation, SERPINE1, and gap junction, Figures S3B and S3F–
S3H) and overall signature score was calculated and plotted using the R pheatmap package (Kolde, 2019)
Longitudinal analysis of a single patient was done by subsetting all cells originating from patient TH226. As above, the differences in
gene expression between the three treatment time points (TN, RD, and PD) was found by applying the Seurat function FindMarkers
using the MAST test (Finak et al., 2015) with sample_name as the latent variable. Three separate tests were used to ascertain the
differences between: 1) TN and RD, 2) TN and PD and 3) RD and PD (script NI07-08, Table S4)
Survival analysis of cancer gene signatures
TCGA LUAD data were downloaded from https://xenabrowser.net/datapages/. Metadata was downloaded from An Integrated
TCGA Pan-Cancer Clinical Data Resource Liu et al., 2018. Mean expression of each cancer cell expression signature (alveolar, ky-
nurenine, plasminogen activating, SERPINE1, and gap junction) was calculated per TCGA sample. TCGA samples were then split by
quartile groups. Quartiles were plotted using library packages survival (Therneau, 2015) and survminer (Haas et al., 2019) in R (script
NI10). Log rank p values are reported for each signature across four expression quartiles. Cox hazard regression model was
computed for comparison of quartile 1 (low expressors) versus quartile 4 (high expressors) for all signatures.
Analysis of immunohistochemistry
Tumor populations were annotated, then scored in a blinded, randomized analysis by a clinical pathologist for percent tumor pos-
itivity and subcellular staining intensity at the membrane, cytosolic, and nuclear compartments. SUSD2 membrane staining was
graded by two reviewers in a blinded, randomized fashion using the slides annotated for tumor presence. Staining intensity was
graded as negative, weak, intermediate, or strong and received scores of 0, 1, 2, or 3 respectively. Percent tumor positivity coefficient
was graded as 0, negative; 1, less than 10% immunopositive; 2, between 10%–50% immunopositive; 3, between 51%–80% immu-
nopositive; 4, greater than 80% immunopositive. Calculation of immunoreactivity scores was performed by multiplying the staining
intensity score (0-3) with the percent tumor positive coefficient (0-4) to yield a value between 0 and 12 (Fedchenko and Reifen-
rath, 2014).
Mutation detection from scRNaseq
Alignment bams for all non-immune cells (stroma and epithelial) were passed to GATK HaplotypeCaller (DePristo et al., 2011) which
was run from the latest available Docker container (broadinstitute/gatk:4.0.11.0) using the following options:
d–disable-read-filter MappingQualityReadFilter
d–disable-read-filter GoodCigarReadFilter
d–disable-read-filter NotSecondaryAlignmentReadFilter
d–disable-read-filter MappedReadFilter
d–disable-read-filter MappingQualityAvailableReadFilter
d–disable-read-filter NonZeroReferenceLengthAlignmentReadFilter
d–disable-read-filter NotDuplicateReadFilter
d–disable-read-filter PassesVendorQualityCheckReadFilter
d–disable-read-filter WellformedReadFilter
Disabling these specific read filters proved necessary for scRNaseq, as inherent low-coverage causes the vast majority of reads to
be flagged for removal otherwise. The full human variant set (dbSNP) was downloaded from NCBI (https://www.ncbi.nlm.nih.gov/
variation/docs/human_variation_vcf/), and every variant call was assessed for its presence/absence in the human variant database.
dbSNP is a public, living catalog of 674 million human somatic SNPs and indels that have been reported by peer-reviewed publi-
cations (Sherry et al., 2001).
Cloud-based parallelization of HaplotypeCaller jobs was achieved with Reflow, a workflow engine for distributed, incremental data
processing in the cloud (GRAIL, https://github.com/grailbio/reflow). HaplotypeCaller outputs a separate variant calling format file
(VCF) for each cell, which were processed with the python package cerebra (https://github.com/czbiohub/cerebra). Variants found
ll
Cell 182, 1232–1251.e1–e10, September 3, 2020 e8
Article

in dbSNP were removed, not to be included in further analysis. We reasoned that by removing ‘common’, population-level variants,
we could better hone-in on disease specific variation.
In addition to scRNaseq reads, we obtained bulk DNA reads from peripheral blood for the majority of our patients (with the excep-
tion of three). These PBMC reads were run through HaplotypeCaller to establish ‘germline’ mutation profiles for each of our patients.
Germline mutations were then subtracted out from each of that patient’s single cell VCFs. This filtering step was omitted for the three
patients for which we did not obtain peripheral blood, however, these single cell VCFs were still passed through our dbSNP filter.
We also applied a fathmm filter to all cells (Shihab et al., 2015). fathmm takes a machine learning approach to predict the likelihood
of a given SNP to be pathogenic, integrating ENCODE annotations for things like transcription factor binding sites, histone modifi-
cations, cross-species sequence alignment and conservation scores, etc. Only variants computationally predicted to be pathogenic
were included in our analysis, i.e., those variants with a fathmm score > 0.7.
The remaining variants were then filtered through the COSMIC (Catalogue of Somatic Mutations in Cancer) complete mutation–
genome screens database (Tate et al., 2019)(https://cancer.sanger.ac.uk/cosmic/download). Only SNPs/indels associated with
‘Lung’ as per their COSMIC annotation were kept. Variant calls were mapped to their corresponding genes, and per-patient / per-
sample mutational profiles were established. We used the ERCCs spiked into each cell sample as a negative control for false positive
mutations, which can arise due to technical artifacts such as PCR errors. We found the median false positive mutation rate to be
0.000256% per (Enge et al., 2017).
Fusion detection from scRNaseq
Fusion transcripts were detected with STAR-fusion (Haas et al., 2019)(https://github.com/STAR-Fusion/STAR-Fusion/wiki) version
1.6.0, run from a Docker container (trinityctat/ctatfusion:1.5.0). The following options were used: –FusionInspector validate,–exam-
ine_coding_effect,–denovo_reconstruct. Distributed processing of STAR-fusion jobs was accomplished with Reflow. Output files
were processed with cerebra, then combined with variant calls to create per-cell and per-sample summary tables.
Mutational analysis of tumor cells
Mutation information from cerebra outputs were summarized by sample. Coverage information was provided by a secondary output
from cerebra summarized by sample and gene. Where all cells are summarized by sample and all fathmm (Shihab et al., 2015) filtered
ROIs are summarized by corresponding gene (script NI06). Plots were generated using the R pheatmap package (Kolde, 2019). Two
comprehensive tables, Tables S3 and S7, detail mutations and fusions per cell.
Survival analysis within the MSK-Impact data
MSK-Impact data was downloaded from cBioPortal (Cerami et al., 2012;Gao et al., 2013) (and subset to only NSCLC samples MSK-
Impact data was subset to only those mutations that were also found in the scRNaseq dataset of mutations (n = 141 unique muta-
tions)). We stratified MSK-Impact samples by those with greater than or equal to 2 mutations from the tier one COSMIC mutations
found in the scRNaseq dataset (mutation high), and those less than 2 mutations (mutation low) (Figure 2D). Kaplan-Meier plots were
visualized with the lifelines package (Davidson-Pilon et al., 2019) in python version 3.4 (Python, 2015) (script NI12).
General immune analysis
All cells annotated as immune (n = 13,431) were subset and clustered as described above (script IM01) using the following param-
eters (Ngenes = 2000, Npc = 20, Res = 0.7). The resulting 18 clusters were assigned to different major immune cells types using a list
of curated gene markers (Table S2) and by manual curation of differentially expressed genes for each cluster (Table S4). The different
cell types and number of cells belonging to each type are described in the main text.
To assess changes in fractional abundance of different immune cell populations we used all cells though excluded thoracentesis
and brain samples due to difference in the immune makeup of these tumor environments which would skew the data. The function
freqCI from the R package REdaS (Maier, 2015) (script IM02) was used to calculate confidence intervals for relative frequencies.
Macrophages (n = 1,379) and T cells (n = 2,226) from lung biopsies were subset and clustered as described above (script IM03 and
IM04 respectively) using the following parameters for MFs (Ngenes = 2000, Npc = 10, res = 0.3) and T cells (Ngenes = 2000, Npc = 10,
res = 0.3). The resulting clusters are discussed in the main text and the lists of differentially expressed genes are provided (Table S4).
We repeated this analysis where we subset the data to only patients with multiple biopsies and sufficient cells (TH226 and TH266)
(script IM05).
Analysis of multiplex Immunofluorescence
Three to six regions from each slide containing tumor and stroma were selected utilizing Phenochart (v1.0.8, Perkin Elmer) for high
resolution multispectral acquisition on the Vectra system at 20X magnification. The images were analyzed with inForm software
(v2.4.8, Akoya) to unmix adjacent fluorochromes, subtract autofluorescence, segment the tissue into tumor and stroma regions,
segment the cells into nuclear, cytoplasmic, and membrane compartments, and to phenotype the cells according to morphology
and cell marker expression. Fractions of macrophage and T cell populations were calculated as: (population of interest) / (macro-
phage + T cell populations) and plotted using ‘ggplot20(Wickham, 2016)inR.
ll
e9 Cell 182, 1232–1251.e1–e10, September 3, 2020
Article

Immune survival analysis within the TCGA
As with the survival analysis using cancer cell gene signatures, we used the downloaded TCGA LUAD dataset and metadata to
access patient survival outcomes as they pertain to the fractional changes of immune populations within a given tumor. We used
CIBERSORT Newman et al., 2015 to deconvolute the bulk TCGA samples into relative fractions of immune cell populations as deter-
mined by using the LM22 reference. The total macrophage population was found by combining fractions for Monocytes, Macropha-
ges.M0, Macrophages.M1, and Macrophages.M2. The total T cell population was found by combining fractions of T.cells.CD8,
T.cells.CD4.naive, T.cells.CD4.memory.resting, T.cells.CD4.memory.activated, T.cells.follicular.helper, T.cells.regulatory.Tregs,
T.cells.gamma.delta, NK.cells.resting, and NK.cells.activated. TCGA samples were then split by quartile groups. Quantiles were
plotted using library packages survival (Therneau, 2015) and survminer (Kassambara et al., 2019) in R (script NI10). Log rank p values
are reported across four expression quartiles. Cox hazard regression model was computed for comparison of quartile 1 (low expres-
sors) versus quartile 4 (high expressors).
Analysis of RT PCR assay
Fold Change was calculated by determining the ratio of mRNA levels to control (day 2) values using the delta threshold cycle (Ct)
method (DCt). A t test was used to find the significance of change between baseline (day 2) and treated time points (days 7, 19
and 70) based on normalized Cts to baseline (script NI14). Plots were made using ‘ggplot20(Wickham, 2016)inR.
Analysis of Wnt/b-catenin inhibition
Analysis was performed using CellInsight (Thermo Fisher Scientific) companion software across technical quadruplicates. The
DMSO treated condition and single agent Wnt/b-catenin inhibitors reached confluency after 3 days in culture (100% maximum
cut-off value). Significance values were calculated using a Student’s t test calculated at treatment endpoints (day 6).
ADDITIONAL RESOURCES
Detailed protocols for single cell dissociation of small tumor biopsies (https://doi.org/10.17504/protocols.io.65rhg56) and high
throughput smartseq2 libraries (https://doi.org/10.17504/protocols.io.2uwgexe) are available at protocols.io.
ll
Cell 182, 1232–1251.e1–e10, September 3, 2020 e10
Article

Supplemental Figures
(legend on next page)
ll
Article

Figure S1. Related to Figure 1
(A) Bar plots of the time interval between treatment start and tissue acquisition for PD and RD tumor samples (B) t-SNE of all epithelial cells (n = 5,581), numbers
correspond to individual clusters. (C) Inferred large-scale copy number variations (CNVs) help identify cancer (pink) and non-cancer cells (purple). Epithelial and
spike in control cells are included in the x axis and chromosomal regions on the y axis. Amplifications (red) or deletions (blue) were inferred by averaging
expression over 100-gene stretches on the respective chromosomes. (D) Bar plot of cell counts for annotated epithelial cells. (E) Bar plot of the number of unique
genes across all annotated epithelial cell types. (F) Bar plot of unique gene count of cancer versus non-cancer epithelial cells.
ll
Article

(legend on next page)
ll
Article

Figure S2. Related to Figure 2
(A) t-SNE plot of 3,754 cancer cells from 44 samples, numbers indicate individual clusters. (B) Circle plot illustrating the clinically identified oncogenic driver (out er
circle) and time points (inner circle) of each biopsy, only for cancer cells. C) Density distribution of cluster occupancy of cancer (red) and non-cancer (blue)
epithelial cell clusters, calculated as the percentage of the highest contributing individual patient over the total number of cells for that cluster. (D) tSNE of cancer
cells colored by patient. (E) tSNE of non-cancer epithelial cells colored by patie nt. (F) Illustration of heterogeneity of primary driver mutated cancer cells found in
exemplary sample LTS47. (G) Clinical characteristics of the 44 NSCLC samples in which at least one cancer cell was identified. Columns indicate clinically
identified mutated gene, treatment response time point (TN, RD, PD), biopsy site, and primary or metastatic sample origin, respectively. (H) Cancer cell mutational
landscape for each patient sample as determined by scRNaseq represented as a heatmap. Color indicates the number of mutant reads for each genomic region
and sample divided by the total number of reads for that region in that sample, NC:No Coverage over the specific genomic region. (I) Mutational landscapeof
COSMIC tier 1 genes. Color indicates the number of mutant reads for each genomic region and sample divided by the total number of reads for that region in that
sample, NC:No Coverage over the specific genomic region. (J) Kaplan-Meier plot showing overall survival of 1269 NSCLC patients within the MSK-Impact
dataset. Patients were stratified by high (> = 2) and low (< 2) mutations from the 141 mutations that are present in both the MSK-Impact dataset and panel (I).
ll
Article

(legend on next page)
ll
Article

Figure S3. Related to Figure 3
(A) Dot plot of the relative expression of established cellular proliferation genes (x axis) across treatment time points (y axis). The color intensity scale reflects the
average gene expression and the size scale indicates the number of cells expressing the gene within that treatment time point. Applying grouped, pairwise
comparisons of treatment time points of the average scaled expression of all genes demonstrated significantly different expressio n (p < 0.0001) in all com-
parisons. (B) Heatmap showing the expression of genes in the alveolar signature. Cells are grouped by treatment time point. (C) Boxplot of Spearman correlations
of cancer cells from all treatment time points and healthy AT2 cells to an external reference of healthy AT2 cells. Non-cancer AT2 cells from our dataset were more
similar to the external, healthy AT2 cells than any of our cancer cells across all time points (mean r= 0.65, 0.10, 0.24, 0.19, for healthy AT2 cells, and TN, RD,
PD cancer cells, respectively). *** indicates a p value < 0.001 (D-F) Immunoreactivity score (IRS) for membrane AQP4 (D), membrane SUSD2 (E), and nuclear
CTNNB1 (F) across all time points. (G) Pairwise comparison of nuclear CTNNB1 IRS for a subset of patients receiving neoadjuvant TKI treatment prior to surgical
removal of tumors, allowing for controlled, matched sample pairs at TN and RD treatment time points. Samples with AZ identifiers refer to patients with EGFR
mutant NSCLC receiving neo-adjuvant osimertinib treatment. Sample with NC identifier refer to patient with ROS1 fusion-positive NSCLC receiving neo-adjuvant
crizotinib treatment. (H-O) High content microscopy screening of EGFR mutant PC9 cells and ALK fusion-positive H3122 cells showing treatment response to TKI
in presence or absence of WNT/b-catenin inhibition. In comparison to DMSO control, upper panel (H-K) shows single agent treatment, lower panel (L-O) shows
combinational treatment of TKI WNT/b-catenin inhibitors. Two WNT/b-catenin inhibitors have been tested, XAV-939 (H,I,L,M) and PRI-724 (J,K,N,O). Values are
shown as percent confluency, with maximum cutoff for full well confluency (100%). pvalues are calculated for all end points (day 6) values compared to single
agent TKI. Error bars represent mean ±standard error of the mean (SEM), n = 4 technical replicates. (P-R) Heatmaps showing the expression of genes within each
signature (kynurenine, SERPINE1/plasminogen activation, and gap junction, respectively) grouped by treatment time point.
ll
Article

(legend on next page)
ll
Article

Figure S4. Related to Figure 3
(A) Graphical summary of cancer cell expression changes across treatment time points. RD features include (1) Alveolar signature, and (2) various RD specific
invasive signaling pathways. PD features include: (3) kynurenine signature, (4) plasminogen activation and SERPINE1 signatures, (5) gap junction proteins, (6)
expression of pro-inflammatory chemokines, (7) loss of tumor suppressor genes, and (8) various PD specific invasive signaling pathway s. (B-F) Boxplots of
pathway signature changes (alveolar, kynurenine, plasminogen activating, SERPINE1, and gap junction, respectively) across treatment time points within only
EGFR mutant cancer cells (*** indicates a p value < 0.0001). (G-K) Boxplots of pathway signature changes (alveolar, kynurenine, plasminogen activating,
SERPINE1, and gap junction, respectively) across treatment time points within only ALK cancer cells (*** indicates a p value < 0.0001). (L-O) Heatmap of sample
average expression with PD only cancer cells for each cancer derived signature gene (alve olar, kynurenine, plasminogen activating/SERPINE1, and gap junction,
respectively).
ll
Article

Figure S5. Related to Figure 3
(A, B, C) Longitudinal timeline of patient treatment, (A) Chest CT scan at each clinical evaluation time point, (B) Biopsy time point with procedural CTscan,(C)
Hematoxylin and eosin (H&E) stainingfrom treatment naiveand progression time points demonstrating adenocarcinoma and squamous cellcarcinoma, respectively,
scale bar indicates 50 mm. (D) Heatmap of mutationstate in clinical driver and a subset of COSMIC tier 1 mutated genes (displayed COSMICtier 1 mutations occur in
at least two out of three samples). Color red indicates the presence of mutation whereas color blue indicates no presence of mutation. (E-I) Boxplots of pathway
signature changes (alveolar, plasminogen activating, SERPINE1, gap junction and squamous histology, respectively) across treatment time points.
ll
Article

Figure S6. Related to Figure 4
(A) Pairwise Pearson correlations between each treatment group’s immune cell compositions which corresponds to the fraction of each immune cell type’s
abundance in the total immune cell population. (B) Total immune cells for each biopsy of patient TH266. (C) Total immune cells for each biopsy of patient TH226.
(D) Fraction of each immune cell subtype for the two biopsies of patient TH266. Error bars indicate the 95% confidence interval for the calculated relative fre-
quencies. Asterisks next to the title of each cell type indicate significance (p < 0.01) when using a chi-square test of independence. Titles of non-significant cell
types are colored red and lack an asterisk. (E) Fraction of each immune cell sub-type for the three biopsies of patient TH226. Error bars indicate the 95%
confidence interval for the calculated relative frequencies. Asterisks next to the title of each cell type indicate significance (p < 0.01) when using a chi-square test
of independence. Titles of non-significant cell types are colored red and lack an asterisk. (F) Representative in situ immunofluorescence images from two patient s
with matched samples at different treatment time points, demonstrating fractional changes in the immune populations of macrophages and T cells. Scale bars
correspond to 50 microns. (G) Kaplan-Meier plot of deconvoluted TCGA lung adenocarcinoma data showing the relation between OS and the fraction of
macrophages for each patient. Patients were stratified by high and low macrophage fraction.
ll
Article

(legend on next page)
ll
Article

Figure S7. Related to Figure 5
(A) t-SNE plot of all lung-derived macrophage cells. (B) Heatmap showing the expression level of the top 10 differentially expressed genes for each
macrophage cluster. (C) t-SNE plot of all lung-derived T cells. (D) Heatmap showing the expression level of the top 10 differentially expressed genes for each T cell
cluster.
ll
Article