Whole Exome and Transcriptome Analyses Integrated with Microenvironmental Immune ... · Research...

13
Research Article Whole Exome and Transcriptome Analyses Integrated with Microenvironmental Immune Signatures of Lung Squamous Cell Carcinoma Jeong-Sun Seo 1,2,3,4 , Ji Won Lee 2,3 , Ahreum Kim 2,3 , Jong-Yeon Shin 2,4 , Yoo Jin Jung 5 , Sae Bom Lee 5 , Yoon Ho Kim 5 , Samina Park 6 , Hyun Joo Lee 6 , In-Kyu Park 6 , Chang-Hyun Kang 6 , Ji-Young Yun 2,4 , Jihye Kim 2,4 , and Young Tae Kim 2,5,6 Abstract The immune microenvironment in lung squamous cell carci- noma (LUSC) is not well understood, with interactions between the host immune system and the tumor, as well as the molecular pathogenesis of LUSC, awaiting better characterization. To date, no molecularly targeted agents have been developed for LUSC treatment. Identication of predictive and prognostic biomarkers for LUSC could help optimize therapy decisions. We sequenced whole exomes and RNA from 101 tumors and matched noncancer control Korean samples. We used the information to predict subtype-specic interactions within the LUSC microenvironment and to connect genomic alterations with immune signatures. Hierarchical clustering based on gene expression and mutational proling revealed subtypes that were either immune defective or immune competent. We analyzed inltrating stromal and immune cells to further characterize the tumor microenviron- ment. Elevated expression of macrophage 2 signature genes in the immune competent subtype conrmed that tumor-associated macrophages (TAM) linked inammation and mutation-driven cancer. A negative correlation was evident between the immune score and the amount of somatic copy-number variation (SCNV) of immune genes (r ¼0.58). The SCNVs showed a potential detrimental effect on immunity in the immune-decient subtype. Knowledge of the genomic alterations in the tumor microenvi- ronment could be used to guide design of immunotherapy options that are appropriate for patients with certain cancer subtypes. Cancer Immunol Res; 6(7); 84859. Ó2018 AACR. Introduction Lung cancer is the second leading cause of death in Korea. The most common type of primary lung cancer, lung adenocarcino- ma, has been characterized at the molecular level (1, 2). Lung squamous cell carcinoma, which accounts for 30% of all lung cancers (3), is not well characterized due to poor understanding of the cancer's genomic evolution (4) and the antitumor activity of immune cells (5, 6). Genomic alterations in the tumor charac- terize various stages of cancer progression. Immune defenses, on the other hand, are governed by tumor stroma, including base- ment membrane, extracellular matrix, vasculature, and cells of the immune system (79). Most cells in tumor stroma have some capacity to suppress a tumor, although this capacity changes as the cancer progresses; invasion and metastasis can follow (1013). Immune and stromal characteristics have emerged as prognos- tic and predictive factors that could be used to guide a person- alized approach in cancer immunotherapy (14, 15). Analyses of genomic alterations, especially somatic mutations, have been used to predict response to immunotherapy (16, 17). Here, we used genomic and transcriptomic analysis to dene molecular subtypes of tumors with immune responses. We show that genomic alterations affect the tumor microenvironment and tumor development in a subtype-specic manner. The data show how genomic alterations and tumor microenvironment impact cancer proliferation and invasion, and how predicted roles of immune cells and their interactions with cancer cells in lung squamous cell carcinoma (LUSC) might affect cancer therapy and patient survival. Materials and Methods RNA and whole-exome sequencing All protocols of this study were approved by the Institutional Review Board of Seoul National University Hospital (IRB No:1312-117-545). One hundred and one cases of lung squamous cell cancer samples, taken between 2011 and 2013, were included. Of these 101 patients we excluded two patients, a patient treated with one cycle of weekly docetaxel 65 mg and cisplatin 48 mg regimen preoperatively, and another patient who died of massive pulmonary embolism at 16 days after operation, from subsequent survival analysis. All the tumor and matched adjacent noncancer 1 Precision Medicine Center, Seoul National University Bundang Hospital, Seongnamsi, Korea. 2 Genomic Medicine Institute (GMI), Medical Research Center, Seoul National University, Seoul, Republic of Korea. 3 Department of Biomedical Sciences, Seoul National University College of Medicine, Seoul, Republic of Korea. 4 Macrogen Inc., Seoul, Republic of Korea. 5 Seoul National University Cancer Research Institute, Seoul, Republic of Korea. 6 Department of Thoracic and Cardiovascular Surgery, Seoul National University Hospital, Seoul, Republic of Korea. Note: Supplementary data for this article are available at Cancer Immunology Research Online (http://cancerimmunolres.aacrjournals.org/). J.-S. Seo, J.W. Lee, A. Kim, and J.-Y. Shin contributed equally to this article. Corresponding Authors: Jeong-Sun Seo, Precision Medicine Center, Seoul National University Bundang Hospital, Seongnamsi 13605, Korea. Phone: 82- 31-600-3001; E-mail: [email protected]; and Young Tae Kim, Department of Thoracic and Cardiovascular Surgery, Seoul National University Hospital, Seoul 03080, Republic of Korea. Phone: 82-22-072-3161; Email: [email protected] doi: 10.1158/2326-6066.CIR-17-0453 Ó2018 American Association for Cancer Research. Cancer Immunology Research Cancer Immunol Res; 6(7) July 2018 848 on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

Transcript of Whole Exome and Transcriptome Analyses Integrated with Microenvironmental Immune ... · Research...

  • Research Article

    Whole Exome and Transcriptome AnalysesIntegrated with Microenvironmental ImmuneSignatures of Lung Squamous Cell CarcinomaJeong-Sun Seo1,2,3,4, Ji Won Lee2,3, Ahreum Kim2,3, Jong-Yeon Shin2,4,Yoo Jin Jung5, Sae Bom Lee5, Yoon Ho Kim5, Samina Park6, Hyun Joo Lee6,In-Kyu Park6, Chang-Hyun Kang6, Ji-Young Yun2,4, Jihye Kim2,4, and Young Tae Kim2,5,6

    Abstract

    The immune microenvironment in lung squamous cell carci-noma (LUSC) is not well understood, with interactions betweenthe host immune system and the tumor, as well as the molecularpathogenesis of LUSC, awaiting better characterization. To date,no molecularly targeted agents have been developed for LUSCtreatment. Identification of predictive and prognostic biomarkersfor LUSC could help optimize therapy decisions. We sequencedwhole exomes andRNA from101 tumors andmatchednoncancercontrol Korean samples. We used the information to predictsubtype-specific interactions within the LUSCmicroenvironmentand to connect genomic alterations with immune signatures.Hierarchical clustering based on gene expression and mutationalprofiling revealed subtypes that were either immune defective or

    immune competent. We analyzed infiltrating stromal andimmune cells to further characterize the tumor microenviron-ment. Elevated expression ofmacrophage 2 signature genes in theimmune competent subtype confirmed that tumor-associatedmacrophages (TAM) linked inflammation and mutation-drivencancer. A negative correlation was evident between the immunescore and the amount of somatic copy-number variation (SCNV)of immune genes (r ¼ �0.58). The SCNVs showed a potentialdetrimental effect on immunity in the immune-deficient subtype.Knowledge of the genomic alterations in the tumor microenvi-ronment could be used to guide design of immunotherapyoptions that are appropriate for patients with certain cancersubtypes. Cancer Immunol Res; 6(7); 848–59. �2018 AACR.

    IntroductionLung cancer is the second leading cause of death in Korea. The

    most common type of primary lung cancer, lung adenocarcino-ma, has been characterized at the molecular level (1, 2). Lungsquamous cell carcinoma, which accounts for 30% of all lungcancers (3), is notwell characterized due to poor understanding ofthe cancer's genomic evolution (4) and the antitumor activity ofimmune cells (5, 6). Genomic alterations in the tumor charac-terize various stages of cancer progression. Immune defenses, onthe other hand, are governed by tumor stroma, including base-mentmembrane, extracellularmatrix, vasculature, and cells of the

    immune system (7–9). Most cells in tumor stroma have somecapacity to suppress a tumor, although this capacity changes as thecancer progresses; invasion and metastasis can follow (10–13).

    Immune and stromal characteristics have emerged as prognos-tic and predictive factors that could be used to guide a person-alized approach in cancer immunotherapy (14, 15). Analyses ofgenomic alterations, especially somatic mutations, have beenused to predict response to immunotherapy (16, 17). Here, weused genomic and transcriptomic analysis to define molecularsubtypes of tumors with immune responses. We show thatgenomic alterations affect the tumor microenvironment andtumor development in a subtype-specific manner. The data showhow genomic alterations and tumor microenvironment impactcancer proliferation and invasion, and how predicted roles ofimmune cells and their interactions with cancer cells in lungsquamous cell carcinoma (LUSC) might affect cancer therapyand patient survival.

    Materials and MethodsRNA and whole-exome sequencing

    All protocols of this study were approved by the InstitutionalReview Board of Seoul National University Hospital (IRBNo:1312-117-545).

    One hundred and one cases of lung squamous cell cancersamples, taken between 2011 and 2013, were included. Of these101 patients we excluded two patients, a patient treated with onecycle of weekly docetaxel 65 mg and cisplatin 48 mg regimenpreoperatively, andanother patientwhodiedofmassivepulmonaryembolism at 16 days after operation, from subsequent survivalanalysis. All the tumor and matched adjacent noncancer

    1Precision Medicine Center, Seoul National University Bundang Hospital,Seongnamsi, Korea. 2Genomic Medicine Institute (GMI), Medical ResearchCenter, Seoul National University, Seoul, Republic of Korea. 3Department ofBiomedical Sciences, Seoul National University College of Medicine, Seoul,Republic of Korea. 4Macrogen Inc., Seoul, Republic of Korea. 5Seoul NationalUniversity Cancer Research Institute, Seoul, Republic of Korea. 6Departmentof Thoracic and Cardiovascular Surgery, Seoul National University Hospital,Seoul, Republic of Korea.

    Note: Supplementary data for this article are available at Cancer ImmunologyResearch Online (http://cancerimmunolres.aacrjournals.org/).

    J.-S. Seo, J.W. Lee, A. Kim, and J.-Y. Shin contributed equally to this article.

    Corresponding Authors: Jeong-Sun Seo, Precision Medicine Center, SeoulNational University Bundang Hospital, Seongnamsi 13605, Korea. Phone: 82-31-600-3001; E-mail: [email protected]; and Young Tae Kim, Department ofThoracic and Cardiovascular Surgery, Seoul National University Hospital, Seoul03080, Republic of Korea. Phone: 82-22-072-3161; Email: [email protected]

    doi: 10.1158/2326-6066.CIR-17-0453

    �2018 American Association for Cancer Research.

    CancerImmunologyResearch

    Cancer Immunol Res; 6(7) July 2018848

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://crossmark.crossref.org/dialog/?doi=10.1158/2326-6066.CIR-17-0453&domain=pdf&date_stamp=2018-6-11http://cancerimmunolres.aacrjournals.org/

  • control tissue specimens were grossly dissected immediate aftersurgery and preserved in liquid nitrogen. Data on clinicalfeatures such as smoking history, pathologic TNM stage, tumorsize, and degree of differentiations were collected (Table 1;Supplementary Table S1). For RNA-seq, we extracted RNA fromtissue using RNAiso Plus (Takara Bio Inc.), followed by puri-fication using RNeasy MinElute (Qiagen). RNA was assessed forquality and was quantified using an RNA 6000 Nano LabChipon a 2100 Bioanalyzer (Agilent Technologies). The RNA-seqlibraries were prepared as previously described (18).

    For whole-exome sequencing, genomic DNAwas extracted and3 mg from each sample was sheared and used for the constructionof a paired-end sequencing library as described in the protocolprovided by Illumina. Enrichment of exonic sequences was thenperformed for each library using the SureSelect Human All Exon50Mb Kit (Agilent Technologies) following the manufacturer'sinstructions.

    Libraries for RNA and whole-exome sequencing weresequenced with Illumina TruSeq SBS Kit v3 on a HiSeq 2000sequencer (Illumina Inc.) to obtain 100-bp paired-end reads. Theimage analysis and base calling were performed using the Illu-mina pipeline (v1.8) with default settings.

    RNA-seq analysisTo characterize the LUSC transcriptome profile in cancer and

    noncancer control cells, we performedRNA-seq for 101 LUSC andmatched noncancer control samples. Total RNA extracted fromlung specimens and depleted of ribosomal RNAwas sequenced atthe desired depth (100�) on RNA-Seq (Illumina HiSeq). Thereads were aligned to the human genome (version GRCh37) withthe Spliced Transcripts Alignment to a Reference (STAR) align-

    ment software. The preprocessing pipeline on the GTAK websitewas followed (19). The raw read counts were generated usingHTSeq-count for each annotated gene.

    Unsupervised subtype clusteringWith the Ensembl gene set, the number of raw reads aligned to

    each genewas computed byHT-seq count andwas normalized bythe Variance Stabilizing Data (VSD) method with use of the Rpackage DEseq2. The variance for each gene was calculated, andthe top 1,000 genes by variance were selected for PCA analysis(20). PCA analysis using the 1,000 most variable genes wasconducted with all tumor and noncancer control samples. Sam-ples were clustered based on principal components into threegroups noncancer control with 95% confidence interval by hier-archical clustering methods as implemented in the R package rgl(21). When analyzing RNA sequencing data, batch effects shouldbe considered if experimental conditions and library preparationvaried. All of our sampleswere processed in the samebatches, thusadditional batch-effect corrections were not necessary (22).

    Differentially expressed gene analysisDifferentially expressed genes of tumor subtypes comparedwith

    noncancer control expression in noncancer control cells weredetermined by the significance criteria (adjusted P < 0.05, |Log2(fold change)|� 1, and basemean� 100) as implemented in the Rpackages DESeq2 and edgeR. The adjusted P value for multipletestingwas calculated byusing the Benjamini–Hochberg correctionfrom the computed P value (23). The centered VSD values of thedifferentially expressed gene list were applied to the array hierar-chical clustering algorism (Cluster 3.0)with uncentered correlationand average linkage (24). The gene expression pattern was visual-ized with use of JAVA treeview. The hierarchical tree by arrays wasgenerated by the clustering process and two types of gene sets indifferentially expressed genes (subtype A-UP and B-DOWN, sub-type A-DOWN and B-UP) were selected and enriched for GeneOntology (GO) gene sets byGene Set Enrichment Analysis (GSEA)in order to determine genes enriched in ranked gene lists.

    Fragments per kilobase million (FPKM) calculation andnormalization

    Raw reads (HTseqcounts) were normalized using FPKM asimplemented in the R package edgeR, and the FPKM values weretransformed to log2 values and adjusted to the median-centeredgene expression values by subtracting the row-wise median fromthe expression values in each row (Cluster 3.0). The centered andlog2-transformed VSD and FPKM expression were used to illus-trate gene expression pattern in a heat map.

    GSEA and network analysisGO analysis of the gene expression data was performed using

    GSEA (v2.24) desktop tools (permutation ¼ 1,000) and visual-ized by the Enrichment Map tool in Cytoscape [P � 0.05, falsediscovery rate (FDR) q value � 0.1, and similarity � 0.5; ref. 25].

    Tumor microenvironment analysisThe fractions of stromal and immune cells in tumor samples

    were estimated by Estimation of STromal and Immune cells inMAlignant Tumours using Expression data (ESTIMATE) scores,with predictions of tumor purity based on the absolute methodpreviously reported (26). The abundance of six infiltratingimmune cell types (B cells, CD4þ T cells, CD8þ T cells,

    Table 1. Clinical data summary

    Korean (n ¼ 101)Patient characteristics Number of patients

    Age at diagnosis, yearsMedian 70Range 35–83

    SexMale 95Female 6

    Smoking statusNever-smoker 12Former smoker 61Current smoker 28

    Median follow-up, months 45Tumor stageI 51II 29III 21IV 0

    T stageT1 27T2 58T3 15T4 1

    N stageN0 62N1 24N2 15

    Recurrancy 31Total LNMedian 30Range 5–66

    Genomic Landscape and Microenvironmental Immune Signature

    www.aacrjournals.org Cancer Immunol Res; 6(7) July 2018 849

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • neutrophils, macrophages, and dendritic cells) in the two sub-types of LUSCwas estimated with the Tumor IMmune EstimationResource (TIMER) algorithm (27).

    Immune score, stromal score, tumor purity, and cell-cycle scoreImmune score, stromal score, and tumor purityweremade using

    ESTIMATE. The gene set for cell-cycle scorewas used to calculate thecell-cycle score from the one in Davoli and colleagues (28).

    Statistical analysesQuantitative data are presented as mean � standard deviation.

    We used R-3.2.3 to perform the statistical analyses. The normality ofthe variableswas testedby the Shapiro–Wilknormality test (29). Fortwo groups, significance (P value) for normally distributed variableswas determined by an unpaired Student t test, and nonnormallydistributed variables were analyzed by the Mann–Whitney U test.For more than two groups, Kruskal–Wallis and one-way ANOVAtests were used for the nonparametric and parametric methods,respectively (30). Statistically significant differences were tested at Pvalues < 0.05. R value was computed by Pearson and distancecorrelation. Survival graph was plotted by the Kaplan–Meier meth-od, and comparison between the subtypes was analyzed with thelog-rank test, andCoxproportional hazardsmodelwas also used forsurvival analysis (SPSS for Windows, version 22; SPSS Inc.).

    Somatic mutation detection using whole-exome sequencingTo find somatic mutations, we performed whole-exome

    sequencing (100�). Reads were aligned to the NCBI humanreference genome (hg19) using BWA (31). Picard was appliedto mark duplicates and we used Genome Analysis Toolkit (GATKIndel Realigner) to improve alignment accuracy. Somatic single-nucleotide variants (SNV) from 101 LUSC samples with matchednoncancer control sampleswere called usingMuTect (32).GATK'sHaplotype Caller was also used for indel detection. All variantswere annotated with information from several databases usingANNOVAR (33).

    Somatic copy-number variant detectionTo detect somatic copy-number variants from whole-exome

    sequencing data, EXCAVATOR analysis was applied (34). GISTICanalysis was used to identify recurrent amplification and deletionpeaks (35).

    Calculation of SCNV levelsWe obtained amplified and deleted copy-number variants

    using EXCAVATOR and GISTIC analyses. Arm and focal SCNVlevels of each patient were respectively calculated by summing thecopy-number changes at each copy-number event. Arm, chromo-some, and focal CNV level were normalized to the mean andstandard deviation among the samples (28).

    Total SCNV ¼X�� Copy-number change at total

    copy-number region��

    Arm-level SCNV ¼X�� Copy-number change at arm

    copy-number region��

    Focal-level SCNV ¼X��Copy-number change at focal

    copy-number region��

    Cell-cycle-related SCNV ¼X��Copy-number change at copy-number

    region of genes used in cell-cycle score��

    Immune-related SCNV ¼X��Copy-number change at copy-number

    region of genes used in immune score��

    Prediction of neoantigensWe predicted neoantigens using the pVAC-Seq pipeline (36).

    We used nonsynonymous mutations to follow the pVAC-seqpipeline. Amino acid changes and transcript sequences wereannotated by variant effect predictor. Epitopes predicted byHLAminer and were filtered by RNA expression (FPKM>1) andcoverage (tumor coverage >10� and noncancer control coverage>5�).

    Availability of data and materialWhole-exome and RNA sequencing data are available under

    the NCBI Sequence Read Archive (SRA) study accession no.SRP114315.

    ResultsIdentification of LUSC subtypes

    In this study, 101 LUSC and matched noncancer controlsamples were used to discover significant differential gene expres-sion, and principal component analysis (PCA) on entire sampleswas performed to identify distinctive clusters based on genevariability between LUSC and noncancer control samples. PCAwith the top 1,000 most variable genes and unsupervised hier-archical clustering with the k-means algorithm were applied toRNA-seq data and distinguished noncancer control from tumorsamples. In contrast, 19 tumor samples overlapped and weremore closely associated with the noncancer control group with95% confidence interval ellipsoids (Fig. 1A; Supplementary Fig.S1). Additional PCAwith 101 tumor samples distinguished the19tumor samples (subtype B) from 82 other major tumor samples(subtype A) at the 95% confidence interval. Thus, PCA revealedtwo molecular subtypes, A and B. These two subtypes were alsodistinguishable in The Cancer Genome Atlas (TCGA) LUSCcohort (n¼ 431) by PCA and unsupervised hierarchical clusteringat 95% confidence interval (Supplementary Fig. S2).

    To identify the pattern of genomic alterations in LUSC, wesequenced the exome of independent noncancer control samples(n ¼ 101) and tumor samples (n ¼ 101). The number of totalmutations in subtype B was less than that in subtype A (P < 0.001by Mann–Whitney U test; Fig. 1B; Supplementary Fig. S3A).Mutations in genes encoding TP53, NAV3, CDH10, KMT2D,NFE2L2, CTNNA3, KEAP1, NTRK3, RB1, NOTCH1, PTEN, FGFR2and EGFR were also identified in the current study (Fig. 1C).Mutational signature combinations in subtype Awere significant-ly different from those in subtype B (P < 0.01 by Mann–WhitneyU test) showing more irregular patterns [Fig. 1D; SupplementaryFig. S3D, (32)]. We compared the burden of somatic copy-num-ber variation (SCNV; ref. 28). Arm-level SCNVs (length > 98% ofa chromosome arm), focal-level SCNVs (length < 98% of achromosome arm), and total SCNVs (all of a chromosome),which comprise the burden of SCNV, were mostly found insubtype A (Fig. 1E and F). To elucidate the effect of genomicalteration on the tumor microenvironment, we also investigated

    Seo et al.

    Cancer Immunol Res; 6(7) July 2018 Cancer Immunology Research850

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • Figure 1.

    Identification of the subtypes of LUSC. A, Principal component analysis (PCA) of the top 1,000 most variable genes among the subtype A, B, and noncancercontrol were performed across all samples. 3D PCA scores plots of top 1,000 of most variable genes was drawn as meshes containing cancer and noncancer controlpoints (left) and subtype A and B points (right) based on K-means clustering (k-means¼ 2) on the first three PCswith 95% confidence interval ellipsoids. B, Ratio ofsomatic synonymous and nonsynonymous mutations in mutations per megabase and subtypes were classified. C, Name of significantly mutated genes (left),distribution of mutations across 101 LUSC, and frequency of significantly mutated genes (right) were plotted. D, The mutational signature revealed by somaticmutations in whole-exome sequencing. E, Arm-level CNV, focal-level CNV, and total level CNV of individual sample displayed in each column. F, Significant,focally amplified (red), and deleted (blue) regions are plotted across chromosomal locations.

    Genomic Landscape and Microenvironmental Immune Signature

    www.aacrjournals.org Cancer Immunol Res; 6(7) July 2018 851

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • major histocompatibility complex (MHC) and tumor-specificantigen (TSA) level, and found that subtype A carried moredeletions of MHC regions than did subtype B (SupplementaryFig. S3C). TSA level was also higher in subtype A than in subtype B(P < 0.001 by Mann–Whitney U test; Supplementary Fig. S3D).Our transcriptome analysis thus defined two LUSC subtypes, Aand B, with different patterns of somatic mutations.

    Identification of differentially expressed genes in subtypesWeanalyzed differential expression of 4,807 genes in subtype A

    and 1,471 genes in subtype B compared with noncancer controlexpression, and matched the significance criteria applied in thisstudy (DESeq2: false discovery rate (FDR) < 0.1, P < 0.05). Wederived a heat map depicting 5,387 differentially expressed genesin subtype A and B by excluding the overlapped genes (Supple-mentary Table S2). Differentially expressed genes were signifi-cantly enriched in several pathways through GO term enrichmentanalysis. Genes for which expression was upregulated were main-ly involved with the cell-cycle andDNA replication (n¼ 1,826) insubtype A andwith immune and defense pathways (n¼ 1,876) insubtype B (Fig. 2A).

    Further functional gene enrichment analysis using GSEA(v2.2.3; ref. 25) proved that these differentially expressed geneswere enriched in pathways related to cytoskeleton, mitosis, cellcycle, and chromatin modification in subtype A and pathwaysrelated to defense response, immune response, cytokine, andmetabolic and biosynthetic process in subtype B. Using networkanalysis with Cytoscape (P � 0.05, FDR q-value � 0.1, andsimilarity � 0.5), we found that the increase in expression andactivation of pathway-related genes in subtype B was linked toregulation of immune cell differentiation and apoptotic process-es, which would increase immune cell abundance. Genes char-acterizing subtype A, on the other hand, were linked to cell-cycleand microtubule pathways that would enhance cancer cell pro-liferation (Fig. 2B). Consistent with GSEA, network analysisshowed that cell cycle and immune system gene sets were differ-entially correlated in each subtype. Both immune response andcell-cycle pathways were differently upregulated in the subtypes,showing an interaction between immune response and somaticmutations.

    The immune landscape of the microenvironment in LUSCTo identify the roles of immunity and distribution of infiltrat-

    ing immune cells in tumor and noncancer control samples, wecomputed stromal and immune scores along with tumor puritybased on the ESTIMATE method (26). The stromal and immunescores in subtype B compared with subtype A were significantlyincreased (stromal core P¼ 9.10� 10�25; immune score P¼ 7.51� 10�19 by unpaired Student t test; Fig. 3A; Supplementary Fig.S4). The high immune scores suggested that recruitment ofimmune cells was more enhanced in subtype B than in A, whichimplicates not only immune cells surrounding the tumor but alsoimmune cells infiltrating the tumor (26). This observation mayreflect the fact that the different portion of infiltrating immuneand stromal cells was intermixed in a dissecting tumor, and thiscan be useful for predicting the degree of tumor purity and tumormicroenvironment (tumor purity P ¼ 9.80 � 10�20 by unpairedStudent t test; Supplementary Fig. S4). As subtype B had moretumor-infiltrating immune cells than did subtype A, cytolyticactivity of subtype B must have been higher. Indeed, the cytolyticactivity score was higher in subtype B than in subtype A (P¼ 2.07

    � 10�6 by unpaired Student t test; Supplementary Fig. S4).Cytolytic activity is ametric of immune-mediated cell destruction,and immune infiltration predicts more cytolytic activity in sub-type B of LUSC (37). Similarly, TCGA LUSC cohort resultsidentified different microenvironmental factors in the two sub-types. Subtype B (n ¼ 328) in the TCGA LUSC cohort wasconsistent with the description of our LUSC subtype B (n ¼ 19;higher immune and stromal scores, lower tumor purity, andhigher CYT score; Supplementary Fig. S5). The TCGA cohortcontained more subtype B than subtype A samples; the Koreanand TCGA cohorts likely had different subtype compositions.

    Immune cells are important in tumors. Stromal cells influencetumor proliferation and inflammation. In order to evaluatetumor-associated stroma, the effects of activated and normalstroma on two tumor subtypes were described by 50 stromalsignature genes as previously reported (38). The heat map ofLUSC and noncancer control cells using active and normal stromagenes indicated three groups: activated stromal-rich samples(subtype A), normal and activated stromal-rich samples (subtypeB), and normal stromal-rich samples (noncancer control). Meannormalized gene expression of the 25 exemplar activated andnormal stromal genes was higher in subtype B than in subtype A(activated stroma P¼ 0.0247; normal stroma P < 0.01 by Mann–Whitney U test: Fig. 3A; Supplementary Figs. S6A and S6B). Theactivated stromal signature geneswere associatedwith carcinoma-associated fibroblasts (CAF) and were overexpressed in subtype Bas compared with subtype A and noncancer control. Likewise, thenormal stromal genes, which reflect the composition of immunecells, were more highly expressed in subtype B than in subtype A.Overexpression of growth factors and chemokines related toCAFs, such as HGF, FGF7, CXCL12, MMP2, IL6, CCL2, NFkB,mediated tumor promotion and aggressive invasion in subtype B(Supplementary Fig. S6C; ref. 39).

    To analyze tumor microenvironment integration in subtypesA and B, we assessed the correlation of stromal and immunescores as well as tumor purity. There was a positive correlationbetween stromal and immune scores (subtype A, Pearson r,0.79; distance r, 0.78; subtype B Pearson r, 0.46; distance r,0.54); samples with high tumor purity showed low stromal andimmune scores (Fig. 3B). Subtypes A and B differed in theirassociation with stromal and immune scores, suggesting vari-ation in microenvironments and in interactions with stromaland immune cells.

    We estimated (with use of the TIMER algorithm) the abun-dance of tumor-infiltrating immune cells in six cell types (B cells,CD4þ T cells, CD8þ T cells, neutrophils, macrophages, anddendritic cells) to predict immune cell profiling in subtypes AandB.Noncancer control A andnoncancer control B subtypes hadno significant differences. Dendritic cells were themost abundantcells among both tumor and noncancer control cells. Althoughmacrophages were not the most abundant cells in tumors andnoncancer control samples, macrophages seemed to be the mostinfluential in a subtype-specific manner (P ¼ 6.27 � 10�9 byMann–Whitney U test; Figs. 3C and D). The proportion ofmacrophages in subtypeBwas significantly higher than in subtypeA, supporting the previous observations that macrophages arepresent in large numbers in the tumor microenvironment andpromote tumor progression and metastasis (40).

    Intratumoral immune coordination was assessed by analyzingthe correlation between selected immune cell markers. Pairwisecomparisons of immune cell abundance levels were done by

    Seo et al.

    Cancer Immunol Res; 6(7) July 2018 Cancer Immunology Research852

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • Figure 2.

    Molecular subtypes at transcriptomic expression level. A, Heat map depicted 5,387 differentially expressed genes in subtypes A and B, and two distinctclustered genes in differentially expressed genes were selected. Top 10 GO gene sets in either clusters were determined based on the rank of enrichment�log10(Pvalue) of pathwayand thematched significance criteria (P

  • Figure 3.

    The immune landscape of the microenvironment in LUSC. A, The heat map depicted the expression of stromal genes and the tumor microenvironmentalfactor [cytolytic activity (CYT), purity, immune, stromal], which divided into three groups, describing activated stroma-rich samples (subtype A), normal andactivated stroma-rich samples (subtype B), and normal stroma-rich samples (noncancer control) gene expression. B, Scatterplots between stromal and immunescores with tumor purity gradient were shown, and correlation coefficient was indicated by each of the subtypes. The color grading corresponds to the tumorpurity, indexed as shown on the color bar at the bottom right of the figure. C, The abundance of infiltrating B cells, CD4þ T cells, CD8þ T cells, neutrophil,macrophages, and dendritic cells in two subtypes was estimated, and each P value was indicated by each of the subtypes (Mann–Whitney U test and Kruskal–Wallistest). Box represents the median (thick line) and the quartiles (line). D, The median-centered and log2-transformed expression level (log2fpkmþ1) of M1and M2 signature genes in subtypes and noncancer control was box plotted with corresponding Mann–Whitney U and Kruskal–Wallis test results. Boxrepresents the median (thick line) and the quartiles (line).

    Seo et al.

    Cancer Immunol Res; 6(7) July 2018 Cancer Immunology Research854

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • measuring Pearson correlation coefficients (r). The relationshipsimplied by these correlations were visualized as r values (Sup-plementary Fig. S7). Correlation between immune cells wasgreater in tumor cells than in noncancer control cells, and thedegree of correlation was higher in subtype B than in subtype A.Dendritic cells were more correlated with other immune cells insubtype B than in subtype A, and the correlations of dendritic cellswith neutrophil and neutrophil with CD8þ T cells were increasedin subtype B. Dendritic cells and neutrophils seemed to haveimmediate connection with CD8þ T cells, confirming previousfindings that the cross-talk between neutrophils and recruiteddendritic cells accelerates antigen presentation to T cells andgenerates an antigen-specific immune response (41, 42). Elevated

    expression of macrophage 2 signature genes in subtype B, apreviously validated gene set (43), suggests that tumor-associatedmacrophages (TAM) promote tumor growth in subtype B cancer(Fig. 3D; Supplementary Figs. S8A and S8B). We examined thereproducibility of the immune parameters by subtype in both ourLUSC and TCGA LUSC cohorts. All Immune scores, activatedstromal gene expression, and macrophage 2 activity patterns bysubtype were highly reproducible across both cohorts. LUSC andTCGA LUSC cohort subtype B showed higher immune infiltration(Supplementary Fig. S9). The immune microenvironment insubtype B was converted to escapemode during the immunoedit-ing process of LUSC in order to help proliferation of resistantclones in an immunocompetent host. This complexity of subtype-

    Figure 4.

    Impacts of genomic alterations on tumor microenvironments in LUSC. A, The level of total CNVs in cell-cycle–related gene sets is displayed across classes (left),and the level of total CNVs in immune-related gene sets is displayed across classes (right). B, The association between genomic alteration load (thenumber of mutations, the level of CNV) and cell-cycle score is shown (left) and the association between genomic alteration load and immune score is shown (right).C, The correlations between genomic alterations (CNVs, mutations) and immune-related properties were displayed. D, The cell-cycle scores are shownin each sample, and the high score densities and low score densities of cell-cycle score are plotted on the x and y axes. E, The immune scores are shown ineach sample, and the high score densities and low score densities of immune score are plotted on the x and y axes. F, The correlation between CNV level (arm, focal)and cell-cycle score or immune score is displayed across subtype.

    Genomic Landscape and Microenvironmental Immune Signature

    www.aacrjournals.org Cancer Immunol Res; 6(7) July 2018 855

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • specific microenvironments should inform therapeutic strategiesfor treatment of LUSC.

    Impacts of somatic copy-number variants on tumormicroenvironment

    To identify the effect of genomic alterations on tumor micro-environments in LUSC, we investigated the association betweensomatic genomic alterations and the immune score. The SCNVswere significantly higher in subtype A (P < 0.001 by Mann–Whitney test; Fig. 4A) and were negatively correlated with theimmune score (Pearson coefficient: �0.58) and with otherimmune-related properties such as the stromal score (stromalcell infiltration), CYT score (immune cytolytic activity), CD4þ

    T-cell infiltration (Pearson coefficient: �0.42), macrophage infil-tration (Pearson coefficient:�0.39), and dendritic cell infiltration(Pearson coefficient: �0.41; Fig. 4B and C). These results sug-gested that the negative correlation between immune score andSCNVs in subtype A may influence the immune response of

    subtype A. Focal-level SCNVs contributedmore to immune scoresthan arm-level SCNVs (Fig. 4DandE). In an association test, focal-level SCNVsweremore highly correlated with immune score (Fig.4F). These findings indicate that focal-level SCNVs are moreinfluential on the tumor microenvironment and might be bettertargets for immunotherapy. Focal-level SCNVs had a strongercorrelation with LUSC immunity and may make it easier todetermine the target genes for drug interventions than arm-levelSCNVs, because arm-level SCNVs are large genomic defects thatmay affect multiple targets (44, 45). To assess the role of focal-level SCNV in the tumor microenvironment, we investigatedfocal-level SCNVs in immune-related pathways via KEGG enrich-ment analysis. Subtype A had a high prevalence of SCNVs onimmune-related pathways (B-cell receptor signaling pathway,chemokine signaling pathway, T-cell receptor signaling pathway,and Toll-like receptor signaling pathway of KEGG; Fig. 5). Genesharboring SCNV deletions, which lead to loss of function, wereenriched in subtype A only for immune-related pathways of

    Figure 5.

    SCNV in LUSC immune pathway. The diagrams show the genes with SCNV in the four immune-related pathways and the percentage of samples withSCNV in immune-related pathway across classes; copy-number deletion (blue), copy-number amplification (pink), the percentage of samples with SCNV insubtype A (red box), the percentage of samples with SCNV in subtype B (blue box)

    Cancer Immunol Res; 6(7) July 2018 Cancer Immunology Research856

    Seo et al.

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • immune system processes, immune responses, and response tocytokine (Supplementary Fig. S10). These data suggested thatfocal-level SCNVs may drive the low immune response intumor microenvironments and would be good targets to con-trol immunity.

    By analyzing the tumor microenvironment with RNA sequenc-ing and whole-exome sequencing, we defined two immune-related subtypes: subtype A (immune defective) and subtype B(immune competent). The log2 VSD expression pattern betweentumors and noncancer control tissue was consistent over the vali-

    dation gene set with previously studied subtypes (Supplemen-tary Fig. S11; ref. 46). The subtype A expression pattern resembledthat of the classical subtype, in which TP63, AKR1C3, FOXE1, andTXN genes were over-expressed. Subtype B appeared to over-express basal related genes (S100A7, MMP13, and SERPINB3)and secretory related genes (ARHGDIB and TNFRSF14). Exceptfor the secretory-related genes, the other genes were all over-expressed in subtype A relative to subtype B. The expressionpattern of previous validated genes differed between subtypes.Thus, our clustering method with gene selection for subtyping

    Figure 6.

    The expression pattern of immune checkpoints in LUSC. A, The expression of immune checkpoint genes was analyzed between subtypes in our LUSC samples(n ¼ 101) and TCGA LUSC cohort (n ¼ 431), respectively. The comparison of median-centered and log2-transformed expression (log2fpkm) of immunecheckpoint genes was performed between subtypes, and P value was indicated by Mann–Whitney U or unpaired Student t test based on the normality. A two-colorscale was used, with blue indicating low expression values and red representing highly expressed genes. B, The median-centered and log2-transformedexpression level (log2fpkm) of immune checkpoint genes in both cohorts was box plotted with corresponding Kruskal–Wallis test results. Box represents themedian (thick line) and the quartiles (line).

    www.aacrjournals.org Cancer Immunol Res; 6(7) July 2018 857

    Genomic Landscape and Microenvironmental Immune Signature

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • achieved similar results to thoseof theprevious study.Our immunesubtyping approach categorized the LUSC into immune-defectiveand -competent subtypes based on the pattern of infiltratingimmune, stromal cells, immune cell composition. Our data alsosuggest that genomic alterations, especially SCNVs, decreaseimmune-related activities and immune cells in cell proliferation–related subtypes of LUSC, and that activated stroma and TAM leadto the evolution of cancer cells in the immune-related subtype.Although the subtypes differed in immune activation, mutationburden, and SCNV, the subtypes had no significant differences inclinical features or overall survival in both cohorts (our LUSCcohort P value ¼ 0.223; TCGA LUSC cohort P value ¼ 0.54; log-rank test; Supplementary Figs. S12 and S13). It was difficult toevaluate to what extent clinical features were affected by lowimmune activities due to the high SCNV in subtype A and theincreased activity of TAMs and activated stroma in subtype B.

    We found that all immune checkpoint genes were morehighly expressed in subtype B than subtype A in both ourLUSC samples and the TCGA LUSC cohort (Fig. 6A), and twoindependent cohorts had similar expression patterns forimmune checkpoints (Fig. 6B). Both PD-1 and its ligand PD-L1as well as other immune checkpoint genes are highly expressedin subtype B samples. PD-1 and PD-L1 are involved in immunetolerance by preventing stimulation of an immune responseand inducing tumor immune escape (47) and might be valu-able targets for new drugs in subtype B. However, difficultywith inducibility of PD-L1 protein expression and with accu-rate assays complicates their use. More reliable biomarkers areneeded (48).

    DiscussionAlthough the study of PD1 expression for immunotherapy has

    produced benefit in many clinical cases, some patients do notrespond to checkpoint blockade (49). For such nonresponders,tumor-specific antigens with appropriate MHC-binding affinitymight be more useful for immunotherapy (36, 50, 51). Weanalyzed the MHC region and TSA levels for both subtypes andfound that subtype A had more TSA and more deletions of MHC

    regions than did subtype B.However, these results did not explainthe role of the tumor microenvironment. Further study is neededto validate the results. In particular TSA content was predicted butnot validated in the clinical environment.

    Our genetic results thus provide evidence for an immuneresponse to cancer in humans and indicate a mechanism oftumor-intrinsic resistance to cytolytic activity in tumors with ahigh burden of somatic mutations. Analysis of genomic altera-tions and their impact on the tumor microenvironment in asubtype-specific manner might identify patients who could ben-efit from cancer immunotherapies that boost the immune system.

    Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

    Authors' ContributionsConception and design: J.-S. Seo, A. Kim, J.-Y. Shin, Y.T. KimDevelopment of methodology: A. Kim, J.-Y. Shin, J.-Y. Yun, J. KimAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J.-Y. Shin, Y.H. Kim, S. Park, H.J. Lee, I.-K. Park,C.-H. Kang, J.-Y. Yun, J. Kim, Y.T. KimAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J.W. Lee, A. Kim, J.-Y. ShinWriting, review, and/or revision of themanuscript: J.W. Lee, A. Kim, J.-Y. Shin,Y.T. KimAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): J.-S. Seo, J.W. Lee, A. Kim, Y.H. Kim, Y.T. KimStudy supervision: J.-S. Seo, C.-H. Kang, Y.T. Kim

    AcknowledgmentsThis work has been supported by Macrogen, Inc. (grant no. MGR17-01)

    and the National Research Foundation of Korea (NRF) grant funded by theKorea government (MSIP; No. NRF-2014R1A2A2A05003665).

    The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

    Received August 19, 2017; revised December 18, 2017; accepted April 25,2018; published first May 2, 2018.

    References1. Park JY, Jang SH. Epidemiology of lung cancer in Korea: recent trends.

    Tuberc Respir Dis (Seoul) 2016;79:58–69.2. Lee CT. Epidemiology of lung cancer in Korea. Cancer Res Treat 2002;34:

    3–5.3. KimY,HammermanPS, Kim J, Yoon JA, Lee Y, Sun JM, et al. Integrative and

    comparative genomic analysis of lung squamous cell carcinomas in EastAsian patients. J Clin Oncol 2014;32:121–8.

    4. Yates LR, Campbell PJ. Evolution of the cancer genome. Nat Rev Genet2012;13:795–806.

    5. Gajewski TF, Schreiber H, Fu YX. Innate and adaptive immune cells in thetumor microenvironment. Nat Immunol 2013;14:1014–22.

    6. Yang Y. Cancer immunotherapy: harnessing the immune system to battlecancer. J Clin Invest 2015;125:3335–7.

    7. Green S, Dawe DE, Banerji S. Immune signatures of non-small cell lungcancer. J Thorac Oncol 2017;12:913–5.

    8. Schoenhals JE, Seyedin SN, Anderson C, Brooks ED, Li YR, Younes AI, et al.Uncovering the immune tumor microenvironment in non-small cell lungcancer to understand response rates to checkpoint blockade and radiation.Transl Lung Cancer Res 2017;6:148–58.

    9. Safonov A, Jiang T, Bianchini G, Gyorffy B, Karn T, Hatzis C, et al. Immunegene expression is associated with genomic aberrations in breast cancer.Cancer Res 2017;77:3317–24.

    10. Li H, Fan X, Houghton J. Tumor microenvironment: the role of the tumorstroma in cancer. J Cell Biochem 2007;101:805–15.

    11. Vannucci L. Stroma as an active player in the development of the tumormicroenvironment. Cancer Microenviron 2015;8:159–66.

    12. Ramamonjisoa N, Ackerstaff E. Characterization of the tumor microenvi-ronment and tumor-stroma interaction by non-invasive preclinical imag-ing. Front Oncol 2017;7:3.

    13. Siniard RC, Harada S. Immunogenomics: using genomics to personalizecancer immunotherapy. Virchows Arch 2017;471:209–19.

    14. Becht E, de Reynies A, Giraldo NA, Pilati C, Buttard B, Lacroix L, et al.Immune and stromal classification of colorectal cancer is associated withmolecular subtypes and relevant for precision immunotherapy. ClinCancer Res 2016;22:4057–66.

    15. Wang RF. A special issue on cancer immunotherapy. Cell Res 2017;27:1–2.16. Rizvi NA, Hellmann MD, Snyder A, Kvistborg P, Makarov V, Havel JJ,

    et al. Cancer immunology. Mutational landscape determines sensitivityto PD-1 blockade in non-small cell lung cancer. Science 2015;348:124–8.

    17. Palmieri G, Colombino M, Cossu A, Marchetti A, Botti G, Ascierto PA.Genetic instability and increased mutational load: which diagnostic toolbest direct patients with cancer to immunotherapy? J Transl Med2017;15:17.

    Seo et al.

    Cancer Immunol Res; 6(7) July 2018 Cancer Immunology Research858

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • 18. Ju YS, Kim JI, Kim S, Hong D, Park H, Shin JY, et al. Extensive genomic andtranscriptional diversity identified through massively parallel DNA andRNA sequencing of eighteen Korean individuals. Nat Genet 2011;43:745–52.

    19. Soundararajan R, Stearns TM,Griswold AL,Mehta A, Czachor A, FukumotoJ, et al. Detection of canonical A-to-G editing events at 30 UTRs andmicroRNA target sites in human lungs using next-generation sequencing.Oncotarget 2015;6:35726–36.

    20. Bailey P, Chang DK, Nones K, Johns AL, Patch AM, Gingras MC, et al.Genomic analyses identifymolecular subtypes of pancreatic cancer. Nature2016;531:47–52.

    21. Inamura K, Fujiwara T, Hoshida Y, Isagawa T, Jones MH, Virtanen C, et al.Two subclasses of lung squamous cell carcinoma with different geneexpression profiles and prognosis identified by hierarchical clustering andnon-negative matrix factorization. Oncogene 2005;24:7105–13.

    22. Liu Q, Markatou M. Evaluation of methods in removing batch effects onRNA-seq data. Infect Dis Transl Med 2016;2:3–9.

    23. Love MI, Huber W, Anders S. Moderated estimation of fold change anddispersion for RNA-seq data with DESeq2. Genome Biol 2014;15:550.

    24. de HoonMJ, Imoto S, Nolan J, Miyano S. Open source clustering software.Bioinformatics 2004;20:1453–4.

    25. SubramanianA, TamayoP,Mootha VK,Mukherjee S, Ebert BL,GilletteMA,et al. Gene set enrichment analysis: a knowledge-based approach forinterpreting genome-wide expression profiles. Proc Natl Acad Sci U S A2005;102:15545–50.

    26. Yoshihara K, Shahmoradgoli M, Martinez E, Vegesna R, Kim H, Torres-Garcia W, et al. Inferring tumour purity and stromal and immune celladmixture from expression data. Nat Commun 2013;4:2612.

    27. Li B, Severson E, Pignon JC, Zhao H, Li T, Novak J, et al. Comprehensiveanalyses of tumor immunity: implications for cancer immunotherapy.Genome Biol 2016;17:174.

    28. Davoli T, UnoH,Wooten EC, Elledge SJ. Tumor aneuploidy correlates withmarkers of immune evasion and with reduced response to immunother-apy. Science 2017;355:.

    29. Ghasemi A, Zahediasl S. Normality tests for statistical analysis: a guide fornon-statisticians. Int J Endocrinol Metab 2012;10:486–9.

    30. Hazra A, Gogtay N. Biostatistics series module 3: comparing groups:numerical variables. Indian J Dermatol 2016;61:251–60.

    31. Li H, Durbin R. Fast and accurate long-read alignment with Burrows-Wheeler transform. Bioinformatics 2010;26:589–95.

    32. doValle IF, Giampieri E, Simonetti G, Padella A,ManfriniM, Ferrari A, et al.Optimized pipeline ofMuTect and GATK tools to improve the detection ofsomatic single nucleotide polymorphisms in whole-exome sequencingdata. BMC Bioinformatics 2016;17:341.

    33. WangK, LiM,HakonarsonH.ANNOVAR: functional annotationof geneticvariants from high-throughput sequencing data. Nucleic Acids Res 2010;38:e164.

    34. Xu D, Olman V, Wang L, Xu Y. EXCAVATOR: a computer program forefficiently mining gene expression data. Nucleic Acids Res 2003;31:5582–9.

    35. Mermel CH, Schumacher SE, Hill B, Meyerson ML, Beroukhim R, Getz G.GISTIC2.0 facilitates sensitive and confident localization of the targets of

    focal somatic copy-number alteration in human cancers. Genome Biol2011;12:R41.

    36. Hundal J, Carreno BM, Petti AA, Linette GP, Griffith OL, Mardis ER, et al.pVAC-Seq: A genome-guided in silico approach to identifying tumorneoantigens. Genome Med 2016;8:11.

    37. Rooney MS, Shukla SA, Wu CJ, Getz G, Hacohen N. Molecular andgenetic properties of tumors associated with local immune cytolyticactivity. Cell 2015;160:48–61.

    38. Moffitt RA, Marayati R, Flate EL, Volmar KE, Loeza SG, Hoadley KA, et al.Virtual microdissection identifies distinct tumor- and stroma-specific sub-types of pancreatic ductal adenocarcinoma. Nat Genet 2015;47:1168–78.

    39. Banat GA, Tretyn A, Pullamsetti SS, Wilhelm J, Weigert A, Olesch C, et al.Immune and inflammatory cell composition of human lung cancerstroma. PLoS One 2015;10:e0139073.

    40. Pollard JW. Tumour-educated macrophages promote tumour progressionand metastasis. Nat Rev Cancer 2004;4:71–8.

    41. Schuster S, Hurrell B, Tacchini-Cottier F. Crosstalk between neutrophilsand dendritic cells: a context-dependent process. J Leukoc Biol 2013;94:671–5.

    42. TateMD,Brooks AG,ReadingPC,Mintern JD.Neutrophils sustain effectiveCD8(þ) T-cell responses in the respiratory tract following influenzainfection. Immunol Cell Biol 2012;90:197–205.

    43. Shaykhiev R, Krause A, Salit J, Strulovici-Barel Y, Harvey BG, O'Connor TP,et al. Smoking-dependent reprogramming of alveolar macrophage polar-ization: implication for pathogenesis of chronic obstructive pulmonarydisease. J Immunol 2009;183:2867–83.

    44. Beroukhim R, Mermel CH, Porter D, Wei G, Raychaudhuri S, Donovan J,et al. The landscape of somatic copy-number alteration across humancancers. Nature 2010;463:899–905.

    45. Weir BA, Woo MS, Getz G, Perner S, Ding L, Beroukhim R, et al. Charac-terizing the cancer genome in lung adenocarcinoma. Nature 2007;450:893–8.

    46. WilkersonMD, Yin X, Hoadley KA, Liu Y, HaywardMC, Cabanski CR, et al.Lung squamous cell carcinoma mRNA expression subtypes are reproduc-ible, clinically important, and correspond to normal cell types. Clin CancerRes 2010;16:4864–75.

    47. He J, Hu Y, Hu M, Li B. Development of PD-1/PD-L1 pathway in tumorimmune microenvironment and treatment for non-small cell lung cancer.Sci Rep 2015;5:13110.

    48. Tsiatas M, Mountzios G, Curigliano G. Future perspectives in cancerimmunotherapy. Ann Transl Med 2016;4:273.

    49. Roh W, Chen PL, Reuben A, Spencer CN, Prieto PA, Miller JP, et al.Integrated molecular analysis of tumor biopsies on sequential CTLA-4and PD-1 blockade reveals markers of response and resistance. Sci TranslMed 2017;9:eaah3560.

    50. MatsushitaH, VeselyMD, Koboldt DC, Rickert CG, Uppaluri R,Magrini VJ,et al. Cancer exome analysis reveals a T-cell-dependent mechanism ofcancer immunoediting. Nature 2012;482:400–4.

    51. Gubin MM, Zhang X, Schuster H, Caron E, Ward JP, Noguchi T, et al.Checkpoint blockade cancer immunotherapy targets tumour-specificmutant antigens. Nature 2014;515:577–81.

    www.aacrjournals.org Cancer Immunol Res; 6(7) July 2018 859

    Genomic Landscape and Microenvironmental Immune Signature

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/

  • 2018;6:848-859. Published OnlineFirst May 2, 2018.Cancer Immunol Res Jeong-Sun Seo, Ji Won Lee, Ahreum Kim, et al. CarcinomaMicroenvironmental Immune Signatures of Lung Squamous Cell Whole Exome and Transcriptome Analyses Integrated with

    Updated version

    10.1158/2326-6066.CIR-17-0453doi:

    Access the most recent version of this article at:

    Material

    Supplementary

    http://cancerimmunolres.aacrjournals.org/content/suppl/2018/05/02/2326-6066.CIR-17-0453.DC1

    Access the most recent supplemental material at:

    Cited articles

    http://cancerimmunolres.aacrjournals.org/content/6/7/848.full#ref-list-1

    This article cites 49 articles, 7 of which you can access for free at:

    E-mail alerts related to this article or journal.Sign up to receive free email-alerts

    Subscriptions

    Reprints and

    [email protected]

    To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department

    Permissions

    Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

    .http://cancerimmunolres.aacrjournals.org/content/6/7/848To request permission to re-use all or part of this article, use this link

    on June 29, 2021. © 2018 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

    Published OnlineFirst May 2, 2018; DOI: 10.1158/2326-6066.CIR-17-0453

    http://cancerimmunolres.aacrjournals.org/lookup/doi/10.1158/2326-6066.CIR-17-0453http://cancerimmunolres.aacrjournals.org/content/suppl/2018/05/02/2326-6066.CIR-17-0453.DC1http://cancerimmunolres.aacrjournals.org/content/6/7/848.full#ref-list-1http://cancerimmunolres.aacrjournals.org/cgi/alertsmailto:[email protected]://cancerimmunolres.aacrjournals.org/content/6/7/848http://cancerimmunolres.aacrjournals.org/

    /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages false /GrayImageMinResolution 200 /GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages false /MonoImageMinResolution 600 /MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 900 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

    /CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MarksOffset 18 /MarksWeight 0.250000 /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA /PageMarksFile /RomanDefault /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /LeaveUntagged /UseDocumentBleed false >> > ]>> setdistillerparams> setpagedevice