Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Comprehensive molecular profiling of lung adenocarcinoma

Comprehensive molecular profiling of lung adenocarcinoma Frequency (%) OPEN ARTICLE doi:10.1038/nature13385 Comprehensive molecular profiling of lung adenocarcinoma The Cancer Genome Atlas Research Network* Adenocarcinoma of the lung is the leading cause of cancer death worldwide. Here we report molecular profiling of 230 resected lung adenocarcinomas using messenger RNA, microRNA and DNA sequencing integrated with copy number, methylationandproteomicanalyses.Highratesofsomaticmutationwereseen(mean8.9mutationspermegabase).Eighteen genes were statistically significantly mutated, including RIT1 activating mutations and newly described loss-of-function MGAmutationswhich aremutually exclusivewithfocalMYCamplification.EGFR mutations weremorefrequentinfemale patients, whereas mutations in RBM10 were more common in males. Aberrations in NF1, MET, ERBB2 and RIT1 occurred in 13% of cases and were enriched in samples otherwise lacking an activated oncogene, suggesting a driver role for these events in certain tumours. DNA and mRNA sequence from the same tumour highlighted splicing alterations driven by somatic genomic changes, including exon 14 skipping in MET mRNA in 4% of cases. MAPK and PI(3)K pathway activity, when measured at the protein level, was explained by known mutations in only a fraction of cases, suggesting additional, unexplained mechanisms of pathway activation. These data establish a foundation for classification and further investi- gations of lung adenocarcinoma molecular pathogenesis. 11–13 Lung cancer is the most common cause of global cancer-related mor- of passenger events per tumour genome . Our efforts focused on com- tality, leading to over a million deaths each year and adenocarcinoma is prehensive, multiplatform analysis of lung adenocarcinoma, with atten- tion towards pathobiology and clinically actionable events. its most common histological type. Smoking is the major cause of lung adenocarcinoma but, as smoking rates decrease, proportionally more Clinical samples and histopathologic data cases occur in never-smokers (defined as less than 100 cigarettes in a life- time).Recently, molecularly targeted therapies have dramaticallyimproved We analysed tumour and matched normal material from230 previously treatment for patients whose tumours harbour somatically activated onco- untreated lung adenocarcinoma patients who provided informed con- genes such as mutant EGFR or translocated ALK,RET, or ROS1 (refs 2–4). sent (Supplementary Table 1). All major histologic types of lung ade- Mutant BRAF and ERBB2 (ref. 5) are also investigational targets. How- nocarcinoma were represented: 5% lepidic, 33% acinar, 9% papillary, ever, mostlung adenocarcinomas either lack an identifiable driver onco- 14% micropapillary, 25% solid, 4% invasive mucinous, 0.4% colloid and gene, or harbour mutations in KRAS and are therefore still treated with 8% unclassifiable adenocarcinoma (Supplementary Fig. 1) . Median conventional chemotherapy. Tumour suppressor gene abnormalities, follow-up was 19 months, and 163 patients were alive at the time of last such as those in TP53 (ref. 6), STK11 (ref. 7), CDKN2A , KEAP1 (ref. 9), follow-up. Eighty-onepercent of patients reported pastor present smok- and SMARCA4 (ref. 10) are also common but are not currently clinically ing. Supplementary Table 2 summarizes demographics. DNA, RNA and actionable. Finally, lung adenocarcinoma shows high rates of somatic protein were extracted from specimens and quality-control assessments mutation and genomic rearrangement, challenging identification of all were performed as described previously . Supplementary Table 3 sum- but the most frequent driver gene alterations because of a large burden marizes molecular estimates of tumour cellularity . Transversion high Transversion low Figure 1 | Somatic mutations in lung a b Gender Number of mutations Number of mutations adenocarcinoma. a, Co-mutation plot from whole Smoking status 150 100 50 0 020 40 60 Male Female NA Ever-smoker Never-smoker exome sequencing of 230 lung adenocarcinomas. TP53 TP53 46 KRAS Data from TCGA samples were combined with KRAS 33 EGFR KEAP1 17 STK11 previously published data for statistical analysis. STK11 17 KEAP1 EGFR 14 NF1 Co-mutation plot for all samples used in the SMARCA4 NF1 11 RBM10 BRAF 10 statistical analysis (n5 412) can be found in PIK3CA SETD2 9 RB1 RBM10 8 Supplementary Fig. 2. Significant genes with a U2AF1 MGA 8 ERBB2 MET 7 corrected P value less than 0.025 were identified ARID1A 7 PIK3CA 7 using the MutSig2CV algorithm and are ranked c Males Females SMARCA4 6 in order of decreasing prevalence. b, c,The RB1 4 Number of mutations Number of mutations CDKN2A 4 60 40 20 0 0 20 40 60 differential patterns of mutation between samples U2AF1 3 EGFR RIT1 2 classified as transversion high and transversion low STK11 Missense Splice site Frameshift SMARCA4 Nonsense In-frame indel samples (b) or male and female patients (c) are RBM10 shown for all samples used in the statistical analysis (n5 412). Stars indicate statistical significance Q < 0.05 Missense Frameshift P < 0.05 Splice site In-frame indel using the Fisher’s exact test (black stars: q, 0.05, Nonsense Other non-synonymous grey stars: P, 0.05) and are adjacent to the sample Transversions Transitions Indels, other set with the higher percentage of mutated samples. *A list of authors and affiliations appears at the end of the paper. 31 J U LY 2014 | V O L 5 1 1 | N ATU RE | 5 4 3 ©2014 Macmillan Publishers Limited. All rights reserved Percentage Normalized RNA-seq read coverage RESEARCH ARTICLE Somatically acquired DNA alterations MDM2,KRAS,EGFR,MET,CCNE1,CCND1,TERC and MECOM (Sup- plementary Table 6), as previously described ,8q24near MYC, and a We performed whole-exome sequencing (WES) on tumour and germ- novel peak containing CCND3 (Supplementary Table 6). The CDKN2A lineDNA,withameancoverageof97.63 and 95.83, respectively, as per- locus was the most significant deletion (Supplementary Table 6). Sup- formed previously . The mean somatic mutation rate across the TCGA plementary Table 7 summarizes molecular and clinical characteristics cohort was 8.87 mutations per megabase (Mb) of DNA (range: 0.5–48, by sample. Low-pass whole-genome sequencing on a subset (n5 93) of median: 5.78). The non-synonymous mutation rate was 6.86 per Mb. MutSig2CV identified significantly mutated genes among our 230 the samples revealed an average of 36 gene–gene and gene–inter-gene cases along with 182 similarly-sequenced, previously reported lung adenocarcinomas . Analysis of these 412 tumour/normal pairs high- lighted 18 statistically significant mutated genes (Fig. 1a shows co-mutation a Exon Exon 13 20 plot of TCGA samples (n5 230),Supplementary Fig. 2 shows co-mutation EML4–ALK plot of all samples used in the statistical analysis (n5 412) and Sup- plementary Table 4 contains complete MutSig2CV results, which also EML4–ALK appear on the TCGA Data Portal along with many associated data files (https://tcga-data.nci.nih.gov/docs/publications/luad_2014/). TP53 was EML4–ALK commonly mutated (46%). Mutations in KRAS (33%) were mutually 11 12 TRIM33–RET exclusive with those in EGFR (14%). BRAF was also commonly mutated (10%), as were PIK3CA (7%), MET (7%) and the small GTPase gene, RIT1 CCDC6–RET (2%). Mutations in tumour suppressor genes including STK11 (17%), 10 34 KEAP1 (17%), NF1 (11%),RB1 (4%) and CDKN2A (4%) were observed. EZR–ROS1 Mutations in chromatin modifying genes SETD2 (9%), ARID1A (7%) and SMARCA4 (6%) and the RNA splicing genes RBM10 (8%) and U2AF1 CD74–ROS1 (3%) were also common. Recurrent mutations in the MGA gene (which 31 35 encodes a Max-interacting protein on the MYC pathway ) occurred in CLTC–ROS1 14 32–34 8% of samples. Loss-of-function (frameshift and nonsense) mutations SLC34A2–ROS1 inMGA were mutually exclusive with focal MYC amplification (Fisher’s exact test P5 0.04), suggesting a hitherto unappreciated potential mech- Portion of original transcripts not in fusion transcript: anism of MYC pathway activation. Coding single nucleotide variants and Normalized, exonic mRNA expression: Low High indel variants were verified by resequencing at a rate of 99% and 100%, respectively (Supplementary Fig. 3a, Supplementary Table 5). Tumour TCGA-99-7458 Exon 14 skipping Number of samples purity was not associated with the presence of false negatives identified None in the validation data (P5 0.31; Supplementary Fig. 3b). (0% skipping) 199 0 0 0 Past or present smoking associated with cytosine to adenine (C.A) TCGA-75-6205 1 11 nucleotide transversions as previouslydescribed both in individual genes Intermediate 12,13 and genome-wide .C. A nucleotide transversion fraction showed (60–80% skipping) 0 1 1 1 two peaks; this fraction correlated with total mutation count (R 5 0.30) and inversely correlated with cytosine to thymine (C. T) transition fre- 27 TCGA-44-6775 Full quency (R 5 0.75) (Supplementary Fig. 4). We classified each sample (90–100% skipping) 0 5 1 0 (Supplementary Methods) into one of two groups named transversion- high (TH,n5 269), and transversion-low (TL,n5 144). The transversion- Y1003 high group was strongly associated with past or present smoking (P , 13 14 15 216 MET mutations 2.23 10 ), consistent with previous reports . The transversion-high and transversion-low patient cohorts harboured different gene mutations. Observed splicing across all tumours Whereas KRAS mutations were significantly enriched in the transversion- (total events = 29,867) high cohort(P5 2.13 10 ),EGFRmutations weresignificantlyenriched 26 Associated with U2AF1 S34F mutation in the transversion-low group (P5 3.33 10 ). PIK3CA and RB1 muta- (total events = 129; q value < 0.05 ) tions were likewise enriched in transversion-low tumours (P, 0.05). 0.0 0.2 0.4 0.6 0.8 1.0 Additionally, the transversion-low tumours were specifically enriched Proportion for in-frame insertions in EGFR and ERBB2 (ref. 5) and for frameshift Cassette exon Coordinate cassette exons Mutually exclusive exon *P < 0.001 indels in RB1 (Fig. 1b). RB1 is commonly mutated in small-cell lung carcinoma (SCLC). We found RB1 mutations in transversion-low ade- Alternative 5′ splice site Alternative 3′ splice site Alternative first exon Alternative last exon nocarcinomas were enriched for frameshift indels versus single nucleotide 20,21 substitutions compared to SCLC (P, 0.05) suggesting a mutational Figure 2 | Aberrant RNA transcripts in lung adenocarcinoma associated mechanism in transversion-low adenocarcinoma that is probably dis- with somatic DNA translocation or mutation. a, Normalized exon level RNA tinct from smoking in SCLC. expression across fusion gene partners. Grey boxes around genes mark the Gender is correlated with mutation patterns in lung adenocarcinoma . regions that are removed as a consequence of the fusion. Junction points of the Onlyafractionofsignificantlymutatedgenesfromthecompletesetreported fusion events are also listed in Supplementary Table 9. Exon numbers refer to reference transcripts listed in Supplementary Table 9. b, MET exon 14 in this study (Fig. 1a) were enriched in men or women (Fig. 1c). EGFR skipping observed in the presence of exon 14 splice site mutation (ss mut), mutations were enriched in tumours from the female cohort (P5 0.03) splice site deletion (ss del) or a Y1003* mutation. A total of 22 samples had whereas loss-of-function mutations within RBM10, an RNA-binding pro- 23 insufficient coverage around exon 14 for quantification. The percentage tein located on the X chromosome were enriched in tumours from men skipping is (total expression minus exon 14 expression)/total expression. (P5 0.002). When examining the transversion-high group, 16 out of 21 c, Significant differences in the frequency of 129 alternative splicing events in RBM10 mutations were observed in males (P5 0.003, Fisher’s exact test). mRNA from tumours with U2AF1 S34F tumours compared to U2AF1 WT Somatic copy number alterations were very similar to those previ- tumours (q value ,0.05). Consistent with the function of U2AF1 in 39 splice ously reported for lung adenocarcinoma (Supplementary Fig. 5, Sup- site recognition, most splicing differences involved cassette exon and alternative 39 splice site events (chi-squared test, P, 0.001). plementary Table 6). Significant amplifications included NKX2-1, TERT, 5 4 4 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved WT ss mut ss del Y1003* Frequency (%) ARTICLE RESEARCH rearrangements per tumour. Chromothripsis occurred in six of the Candidate driver genes 93 samples (6%) (Supplementary Fig. 6, Supplementary Table 8). Low- The receptor tyrosine kinase (RTK)/RAS/RAF pathway is frequently pass whole genome sequencing-detected rearrangements appear in mutated in lung adenocarcinoma. Striking therapeutic responses are Supplementary Table 9. often achieved when mutant pathway components are successfully inhib- ited. Sixty-two per cent (143/230) of tumours harboured known activating Description of aberrant RNA transcripts mutations in known driver oncogenes, as defined by others . Cancer- Gene fusions, splice site mutations or mutations in genes encoding splic- associated mutations in KRAS (32%, n5 74), EGFR (11%, n5 26) and ing factors promote or sustain the malignant phenotype by generating BRAF (7%, n5 16) were common. Additional, previously uncharac- aberrant RNA transcripts. Combining DNA with mRNA sequencing terized KRAS, EGFR and BRAF mutations were observed, but were not enabled us to catalogue aberrant RNA transcripts and, in many cases, classified as driver oncogenes for the purposes of our analyses (see Sup- to identify the DNA-encoded mechanism for the aberration. Seventy- plementary Fig. 9a for depiction of all mutations of known and unknown five per cent of somatic mutations identified by WES were present in the significance); explaining the differing mutation frequencies in each gene RNA transcriptome when the locus in question was expressed (minimum between this analysis and the overall mutational analysis described above. 53) (Supplementary Fig. 7a) similar to prior analyses . Previously iden- We also identified known activating ERBB2 in-frame insertion and point tified fusions involving ALK (3/230 cases), ROS1 (4/230) and RET mutations (n5 5) , as well as mutations in MAP2K1 (n5 2), NRAS and (2/230) (Fig. 2a, Supplementary Table 10), all occurred in transversion- HRAS (n5 1 each). RNA sequencing revealed the aforementioned MET low tumours (P5 1.853 10 , Fisher’s exact test). exon 14 skipping (n5 10) and fusions involving ROS1 (n5 4), ALK MET activation can occur by exon 14 skipping, which results in a (n5 3) and RET (n5 2). We considered these tumours collectively as stabilized protein . Ten tumours had somatic MET DNA alterations oncogene-positive, as they harboured a known activating RTK/RAS/ with MET exon 14 skipping in RNA. In nine of these samples, a 59 or RAF pathway somatic event. DNA amplification events were not con- 39 splice site mutation or deletion was identified . MET exon 14 skip- sidered to be driver events before the comparisons described below. ping was also found in the setting of a MET Y1003* stop codon muta- We sought to nominate previously unrecognized genomic events that tion (Fig. 2b, Supplementary Fig. 8a). The codon affected by the Y1003* might activate this critical pathway in the 38% of samples without a mutation is predicted to disrupt multiple splicing enhancer sequences, RTK/RAS/RAF oncogene mutation. Tumour cellularity did not differ but the mechanism of skipping remains unknown in this case. between oncogene-negative and oncogene-positive samples (Supplemen- S34F mutations in U2AF1 have recently been reported in lung ade- tary Fig. 9b). Analysis of copy number alterations using GISTIC identified nocarcinoma but their contribution to oncogenesis remains unknown. unique focal ERBB2 and MET amplifications in the oncogene-negative S34F Eight samples harboured U2AF1 . We identified 129 splicing events subset (Fig. 3a, Supplementary Table 6); amplifications in other wild-type S34F strongly associated with U2AF1 mutation, consistent with the role of proto-oncogenes, including KRAS and EGFR, were not significantly U2AF1 in 39-splice site selection . Cassette exons and alternative 39 splice different between the two groups. sites were most commonly affected (Fig. 2c, Supplementary Table 11) . We next analysed WES data independently in the oncogene-negative Among these events, alternative splicing of the CTNNB1 proto-oncogene and oncogene-positive subsets. We found that TP53, KEAP1, NF1 and was strongly associated with U2AF1 mutations (Supplementary Fig. 8b). RIT1 mutations were significantly enriched in oncogene-negative tumours Thus, concurrent analysis of DNA and RNA enabled delineation of (P, 0.01; Fig. 3b, Supplementary Table 12). NF1 mutations have previ- both cis and trans mechanisms governing RNA processing in lung ously been reported in lung adenocarcinoma , but this is the first study, adenocarcinoma. to our knowledge, capable of identifying all classes of loss-of-function a b –16 0.6 Oncogene-positive Oncogene-positive 0.5 Oncogene-negative Oncogene-negative –8 0.4 MET ERBB2 0.3 –4 0.2 –2 0.1 0.0 0.1 TP53 KEAP1 NF1 RIT1 HRAS (0.4%) NRAS (0.4%) Chromosome c d RET fusion (0.9%) Previously MAP2K1 (0.9%) Oncogene-positive oncogene-negative ALK fusion (1.3%) (62%, n = 143) (13%, n = 31) RIT1 (2.2%) ROS1 fusion (1.7%) KRAS ERBB2 amp (0.9%) ERBB2 (1.7%) MET amp (2.2%) MET ex14 (4.3%) EGFR 11 NF1 BRAF 7 (8.3%) ROS1/ALK/RET 4 BRAF MAP2K1 / HRAS / NRAS (7.0%) MET None EGFR ERBB2 (24.4%) (11.3%) RIT1 2 KRAS NF1 (32.2%) Amplification Fusion Missense mutation Exon skipping In-frame indel Nonsense mutation / frameshift indel / splice-site mutation Figure 3 | Identification of novel candidate driver genes. a, GISTIC analysis adenocarcinoma. Not shown are the 63 tumours lacking an identifiable driver of focal amplifications in oncogene-negative (n5 87) and oncogene-positive lesion. Only canonical driver events, as defined in Supplementary Fig. 9, and (n5 143) TCGA samples identifies focal gains of MET and ERBB2 that are proposed driver events, are shown; hence not every alteration found is specific to the oncogene-negative set (purple). b, TP53, KEAP1, NF1 and RIT1 displayed. d, New candidate driver oncogenes (blue: 13% of cases) and known mutations are significantly enriched in samples otherwise lacking oncogene somatically activated drivers events (red: 63%) that activate the RTK/RAS/RAF mutations (adjusted P, 0.05 by Fisher’s exact test). c, Co-mutation plot of pathway can be found in the majority of the 230 lung adenocarcinomas. variants of known significance within the RTK/RAS/RAF pathway in lung 31 J U LY 2014 | V O L 5 1 1 | N ATU RE | 5 4 5 ©2014 Macmillan Publishers Limited. All rights reserved FDR q Per cent mutated MAPK pathway score mTOR pathway score RESEARCH ARTICLE NF1 defects and to statistically demonstrate thatNF1 mutations, as well Recurrent alterations in key pathways as KEAP1 and TP53 mutations are enriched in the oncogene-negative Recurrent aberrations in multiple key pathways and processes charac- subset of lung adenocarcinomas (Fig. 3c). All RIT1 mutations occurred terize lung adenocarcinoma (Fig. 4a). Among these were RTK/RAS/ in the oncogene-negative subset and clustered around residue Q79 (homol- RAF pathway activation (76% of cases), PI(3)K-mTOR pathway activa- ogous to Q61 in the switch II region of RAS genes). These mutations tion (25%), p53 pathway alteration (63%), alteration of cell cycle regu- transform NIH3T3 cells and activate MAPK and PI(3)K signalling , lators (64%, Supplementary Fig. 10), alteration of oxidative stress pathways supporting a driver role for mutant RIT1 in 2% of lung adenocarcinomas. (22%, Supplementary Fig. 11), and mutation of various chromatin and This analysis increases the rate at which putative somatic lung adeno- RNA splicing factors (49%). carcinoma driver events can be identified within the RTK/RAS/RAF We then examined the phenotypic sequelae of some key genomic pathway to 76% (Fig. 3d). events in the tumours in which they occurred. Reverse-phase protein arrays provided proteomic and phosphoproteomic phenotypic evidence of pathway activity. Antibodies on this platform are listed in Supplemen- EGFR ERBB2 MET ALK RET ROS1 tary Table 13. This analysis suggested that DNA sequencing did not 11% 3% 7% 1% <1% 2% identify all samples with phosphoprotein evidence of activation of a PTEN given signalling pathway. For example, whereas KRAS-mutant lung ade- 3% PIK3CA KRAS NRAS 4% nocarcinomas had higher levels of phosphorylated MAPK than KRAS 32% <1% PIK3R1 NF1 <1% wild-type tumours had on average, many KRAS wild-type tumours dis- 11% HRAS RIT1 <1% 2% STK11 AKT1 played significant MAPK pathway activation (Fig. 4b, Supplementary 17% 1% Fig. 10). The multiple mechanisms by which lung adenocarcinomas Per cent of cases (%) BRAF achieve MAPK activation suggest additional, still undetected RTK/RAS/ 7% AMPK TSC1/2 RAF pathway alterations. Similarly, we found significant activation of 50 0 100 Inactivated Activated mTOR and its effectors (p70S6kinase, S6, 4E-BP1) in a substantial frac- MAP2K1 Activation Inhibition tion of the tumours (Fig. 4c). Analysis of mutations in PIK3CA and <1% MTOR 33 STK11, STK11 protein levels, and AMPK and AKT phosphorylation led to the identification of three major mTOR patterns in lung adeno- CDKN2A carcinoma: (1) tumours with minimal or basal mTOR pathway activa- Proliferation, cell survival, translation 43% tion, (2) tumours showing higher mTOR activity accompanied by either KEAP1 CUL3 MDM2 ATM CCND1 CDK4 CCNE1 STK11-inactivating mutation or combined low STK11 expression and 19% <1% 8% 9% 4% 7% 3% low AMPK activation and (3) tumours showing high mTOR activity accompanied by either phosphorylated AKT activation, PIK3CA muta- NFE2L2 TP53 RB1 tion, or both. As with MAPK, many tumours lack an obvious underlying 3% 46% 7% genomic alteration to explain their apparent mTOR activation. Oxidative Proliferation, Cell cycle stress response cell survival progression Molecular subtypes of lung adenocarcinoma ARID1A ARID1B Broad transcriptional and epigenetic profiling can reveal downstream 7% 6% consequences of driver mutations, provide clinically relevant classifica- ARID2 SMARCA4 SETD2 U2AF1 RBM10 tion and offer insight into tumours lacking clear drivers. Prior unsuper- 7% 6% 9% 4% 9% vised analyses of lung adenocarcinoma gene expression have used varying 34–37 nomenclature for transcriptional subtypes of the disease .Tocoor- Nucleosome Histone RNA splicing / remodelling methylation processing dinate naming of the transcriptional subtypes with the histopathological , anatomic and mutational classifications of lung adenocarcinoma, we MAPK pathway propose an updated nomenclature: the terminal respiratory unit (TRU, KRAS mut n n = 53 = n = 128 formerly bronchioid), the proximal-inflammatory (PI, formerly squa- p-JNK p-MAPK moid), and the proximal-proliferative (PP, formerly magnoid) transcrip- p-MEK1 tional subtypes (Fig. 5a). Previously reported associations of expression p-p38 34,36,39 p-p90RSK –5 signatures with pathways and clinical outcomes were observed (Sup- p-Shc plementary Fig. 7b) and integration with multi-analyte data revealed P < 0.01 p-c-Raf –10 Subtype statistically significant genomic alterations associated with these tran- Pathway KRAS KRAS scriptional subtypes. The PP subtype was enriched for mutation of KRAS, score mut wt along with inactivation of the STK11 tumour suppressor gene by chro- PI(3)K pathway mosomal loss, inactivating mutation, and reduced gene expression. In PIK3CA STK11 High Low mut mut p-AKT p-AMPK Unaligned ** contrast, the PI subtype was characterized by solid histopathology and ( n = 9) ( n = 42) ( n = 35) ( n = 21) ( n = 74) ** PIK3CA mut Figure 4 | Pathway alterations in lung adenocarcinoma. a, Somatic STK 11 mut PTEN loss alterations involving key pathway components for RTK signalling, mTOR signalling, oxidative stress response, proliferation and cell cycle progression, STK 11/LKB1 p-AMPK nucleosome remodelling, histone methylation, and RNA splicing/processing. p-AKT b, c, Proteomic analysis by RPPA (n5 181) P values by two-sided t-test. p-mTOR –1 p-4E-BP1 Box plots represent 5%, 25%, 75%, median, and 95%. PP, proximal –2 p-p70S6K proliferative; TRU, terminal respiratory unit; PI, proximal inflammatory. p-S6 c, mTOR signalling may be activated, by either Akt (for example, via PI(3)K) or Subtype Pathway inactivation of AMPK (for example, via STK11 loss). Tumours were separated score into three main groups: those with PI(3)K-AKT activation, through either PIK3CA activating mutation or unknown mechanism (high p-AKT); those PI3K-Akt branch active LKB1-AMPK inactive with LKB1-AMPK inactivation, through either STK11 mutation or unknown Protein expression Expression subtype Pathway signature *P < 0.01 mechanism with low levels of LKB1 and p-AMPK; and those showing none Low High PP TRU PI Low High **P < 0.001 of the above features. 5 4 6 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved PIK3CA mut High p-AKT STK11 mut Low p-AMPK Unaligned ARTICLE RESEARCH a Expression subtypes c Integrated subtypes 16 2 3 4 5 Expression subtype Proximal proliferative Proximal inflammatory Terminal respiratory unit DNA methylation subtype mut STK11 CN del p16 methylation under expr. Ploidy mut KEAP1 Non-silent mutation rate mut KRAS Purity mut TP53 mut NF1 methylation p16 Fusions mut EGFR over expr. TTF-1 Mutation total 8 Ploidy Purity CpG T % 11 Never-smoker Female Histology 17 b DNA methylation subtypes DNA methylation subtype Expression subtype Expression, ploidy, purity, mutation rates CIMP-high CIMP-intermediate CIMP-low Normal CIMP-high Proximal proliferative CIMP-intermediate Proximal inflammatory CIMP-low Terminal respiratory unit Low High GATA4 (TRU) SFRP1 GATA5 Integrated subtype Fusion Histology 1 Solid iClust1 4 iClust4 ALK WIF1 Acinar 2 iClust2 5 iClust5 ROS1 Lepidic GATA2 3 iClust3 iClust6 RET Papillary/Micropapillary DNA copy number DNA methylation CDKN2A Mucinous RASSF1 Other –1.0 0 1.0 01.0 SOX17 HOXD1 Mutation Smoking status Gender HOXA9 Mutant Never-smoker Female Concurrent p16 methylation and SETD2 mutation HIC1 Figure 5 | Integrative analysis. a–c, Integrating unsupervised analyses of 230 iCluster analysis (c). All displayed features are significantly associated with lung adenocarcinomas reveals significant interactions between molecular subtypes depicted. The CIMP phenotype is defined by the most variable CpG subtypes. Tumours are displayed as columns, grouped by mRNA expression island and promoter probes. subtypes (a), DNA methylation subtypes (b), and integrated subtypes by co-mutation of NF1 and TP53. Finally, the TRU subtype harboured the and CDKN2A methylation. Tumours with low CDKN2A expression majorityof theEGFR-mutatedtumours as wellas thekinase fusion express- due to methylation (rather than due to mutation or deletion) had lower ing tumours. TRU subtype membership was prognostically favourable, ploidy, fewer overall mutations (Fig. 5c) and were significantly enriched for SETD2 mutation, suggesting an important role for this chromatin- as seen previously (Supplementary Fig. 7c). Finally, the subtypes exhib- ited different mutation rates, transition frequencies, genomic ploidy pro- modifying gene in the development of certain tumours. Integrative clustering of copy number, DNA methylation and mRNA files, patterns of large-scale aberration, and differed in their association with smoking history (Fig. 5a). Unsupervised clustering of miRNA expression data found six clusters (Fig. 5c). Tumour ploidy and mutation sequencing-derived or reverse phase protein array (RPPA)-derived data rate are higher in clusters 1–3 than in clusters 4–6. Clusters 1–3 frequently also revealed significant heterogeneity, partially overlapping with the harbour TP53 mutations and are enriched for the two proximal tran- mRNA-based subtypes, as demonstrated in Supplementary Figs 12 and 13. scriptional subtypes. Fisher’s combined probability tests revealed signi- Mutations in chromatin-modifying genes (for example, SMARCA4, ficant copy number associated gene expression changes on 3q in cluster one, 8q in cluster two, and chromosome 7 and 15q in cluster three (Sup- ARID1A and SETD2) suggest a major role for chromatin maintenance in lung adenocarcinoma. To examine chromatin states in an unbiased plementary Fig. 15). The low ploidy and low mutation rate clusters four manner, we selected the most variable DNA methylation-specific probes and five contain many TRU samples, whereas tumours in cluster 6 have in CpG island promoter regions and clustered them by methylation inten- comparatively lower tumour cellularity, and few other distinguishing sity (Supplementary Table 14). This analysis divided samples into two molecular features. Significant copy number-associated gene expres- distinct subsets: a significantly altered CpG island methylator phenotype- sion changes are observed on 6q in cluster four and 19p in cluster five. high (CIMP-H(igh)) cluster and a more normal-like CIMP-L(ow) group, The CIMP-H tumours divided into a high ploidy, high mutation rate, proximal-inflammatory CIMP-H group (cluster 3) and a low ploidy, low with a third set of samples occupying an intermediate level of methy- lation at CIMP sites (Fig. 5b). Our results confirm a prior report and mutation rate, TRU-associated CIMP-H group (cluster 4), suggesting that the CIMP phenotype in lung adenocarcinoma can occur in markedly provide additional insights into this epigenetic program.CIMP-H tumours often showed DNA hypermethylation of several key genes: CDKN2A, different genomic and transcriptional contexts. Furthermore, cluster four is enriched for CDKN2A methylation and SETD2 mutations, sug- GATA2, GATA4, GATA5, HIC1, HOXA9, HOXD13, RASSF1, SFRP1, SOX17 and WIF1 among others (Supplementary Fig. 14). WNT pathway gestingan interaction between somatic mutationof SETD2 andderegulated chromatin maintenance in this subtype. Finally, cluster membership genes are significantly over-represented in this list (P value5 0.0015) suggesting that this is a key pathway with an important driving role was significantly associated with mutations in TP53, EGFR and STK11 within this subtype. MYC overexpression was significantly associated (Supplementary Fig. 15, Supplementary Table 6). with the CIMP-H phenotype as well (P5 0.003). Conclusions Although we did not find significant correlations between global DNA methylation patterns and individual mutations in chromatin remodel- We assessed themutation profiles, structural rearrangements, copy number ling genes, there was an intriguing association between SETD2 mutation alterations, DNA methylation, mRNA, miRNA and protein expression 31 JU LY 2 0 14 | V OL 5 1 1 | N AT URE | 5 4 7 ©2014 Macmillan Publishers Limited. All rights reserved DNA copy number RESEARCH ARTICLE 13. Govindan, R. et al. Genomic landscape of non-small cell lung cancer in smokers of 230 lung adenocarcinomas. In recent years, the treatment of lung and never-smokers. Cell 150, 1121–1134 (2012). adenocarcinoma has been advanced by the development of multiple 14. Travis, W. D., Brambilla, E. & Riely, G. J. New pathologic classification of lung therapies targeted against alterations in the RTK/RAS/RAF pathway. We cancer: relevance for clinical practice and clinical trials. J. Clin. Oncol. 31, 992–1001 (2013). nominate amplifications in MET and ERBB2 as well as mutations of 15. The Cancer Genome Atlas Research Network Comprehensive genomic NF1 and RIT1 as driver events specifically in otherwise oncogene-negative characterization of squamous cell lung cancers. Nature 489, 519–525 lung adenocarcinomas. This analysis increases the fraction of lung ade- (2012). nocarcinoma cases with somatic evidence of RTK/RAS/RAF activation 16. Carter, S. L. et al. Absolute quantification of somatic DNA alterations in human cancer. Nature Biotechnol. 30, 413–421 (2012). from 62% to 76%. While all lung adenocarcinomas may activate this 17. Cibulskis, K. et al. Sensitive detection of somatic point mutations in impure and pathway by some mechanism, only a subset show tonic pathway acti- heterogeneous cancer samples. Nature Biotechnol. 31, 213–219 (2013). vation at the protein level, suggesting both diversity between tumours 18. Lawrence, M. S. et al. Discovery and saturation analysis of cancer genes across 21 tumour types. Nature 505, 495–501 (2014). with seemingly similar activating events and as yet undescribed mech- 19. Hurlin, P. J., Steingrimsson, E., Copeland, N. G., Jenkins, N. A. & Eisenman, R. N. anisms of pathway activation. Therefore, the current study expands the Mga, a dual-specificity transcription factor that interacts with Max and contains a range of possible targetable alterations within the RTK/RAS/RAF path- T-domain DNA-binding motif. EMBO J. 18, 7019–7028 (1999). way in general and suggests increased implementation of MET and 20. Peifer, M. et al. Integrative genome analyses identify key somatic driver mutations of small-cell lung cancer. Nature Genet. 44, 1104–1110 (2012). ERBB2/HER2 inhibitors in particular. Our discovery of inactivating 21. Rudin, C. M. et al. Comprehensive genomic analysis identifies SOX2 as a mutations of MGA further underscores the importance of the MYC frequently amplified gene in small-cell lung cancer. Nature Genet. 44, 1111–1116 pathway in lung adenocarcinoma. (2012). 22. Tokumo, M. et al. The relationship between epidermal growth factor receptor This study further implicates both chromatin modifications and splic- mutations and clinicopathologic features in non-small cell lung cancers. ing alterations in lung adenocarcinoma through the integration of DNA, Clin. Cancer Res. 11, 1167–1173 (2005). transcriptome and methylome analysis. We identified alternative splic- 23. Coleman, M. P. et al. A novel gene, DXS8237E, lies within 20 kb upstream of UBE1 ing due to both splicing factor mutations in trans and mutation of splice in Xp11.23 and has a different X inactivation status. Genomics 31, 135–138 (1996). sites in cis, the latter leading to activation of the MET gene by exon 14 24. Weir, B. A. et al. Characterizing the cancer genome in lung adenocarcinoma. skipping. Cluster analysis separated tumours based on single-gene driver Nature 450, 893–898 (2007). events as well as large-scale aberrations, emphasizing lung adenocarci- 25. Stephens, P. J. et al. Massive genomic rearrangement acquired in a single catastrophic event during cancer development. Cell 144, 27–40 (2011). noma’s molecular heterogeneity and combinatorial alterations, includ- 26. Kong-Beltran, M. et al. Somatic mutations lead to an oncogenic deletion of Met in ing the identification of coincident SETD2 mutations and CDKN2A lung cancer. Cancer Res. 66, 283–289 (2006). methylation in a subset of CIMP-H tumours, providing evidence of a 27. Seo, J. S. et al. The transcriptional landscape and mutational profile of lung somatic event associated with a genome-wide methylation phenotype. adenocarcinoma. Genome Res. 22, 2109–2119 (2012). 28. Wu, S., Romfo, C. M., Nilsen, T. W. & Green, M. R. Functional recognition of These studies provide new knowledge by illuminating modes of geno- the 39 splice site AG by the splicing factor U2AF . Nature 402, 832–835 mic alteration, highlighting previously unappreciated altered genes, and (1999). enabling further refinement in sub-classification for the improved per- 29. Brooks, A. N. et al. A pan-cancer analysis of transcriptome changes associated with somatic mutations in U2AF1 reveals commonly altered splicing events. PLoS ONE sonalization of treatment for this deadly disease. 9, e87361 (2014). 30. Pao, W. & Hutchinson, K. E. Chipping away at the lung cancer genome. Nature Med. METHODS SUMMARY 18, 349–351 (2012). All specimens were obtained from patients with appropriate consent from the rele- 31. Beroukhim, R. et al. Assessing the significance of chromosomal aberrations in cancer: methodology and application to glioma. Proc. Natl Acad. Sci. USA 104, vant institutional review board. DNA and RNA were collected from samples using 20007–20012 (2007). the Allprep kit (Qiagen). We used standard approaches for capture and sequencing of 32. Berger, A. H. et al. Oncogenic RIT1 mutations in lung adenocarcinoma. Oncogene exomes from tumour DNA and normal DNA and whole-genome shotgun sequenc- http://dx.doi.org/10.1038/onc.2013.581 (2014). ing. Significantly mutated genes were identified by comparing them with expectation 33. Creighton, C. J. et al. Proteomic and transcriptomic profiling reveals a link between models based on the exact measured rates of specific sequence lesions . GISTIC the PI3K pathway and lower estrogen-receptor (ER) levels and activity in ER analysis of the circular-binary-segmentedAffymetrix SNP 6.0 copy numberdata was breast cancer. Breast Cancer Res. 12, R40 (2010). 34. Wilkerson, M. D. et al. Differential pathogenesis of lung adenocarcinoma subtypes used to identify recurrent amplification and deletion peaks . Consensus clustering involving sequence mutations, copy number, chromosomal instability, and approaches were used to analyse mRNA, miRNA and methylation subtypes using methylation. PLoS ONE 7, e36530 (2012). previous approaches . The publication web page is (https://tcga-data.nci.nih.gov/ 35. Beer, D. G. et al. Gene-expression profiles predict survival of patients with lung docs/publications/luad_2014/). Sequence files are in CGHub (https://cghub.ucsc.edu/). adenocarcinoma. Nature Med. 8, 816–824 (2002). 36. Hayes, D. N. et al. Gene expression profiling reveals reproducible human lung Received 11 June 2013; accepted 22 April 2014. adenocarcinoma subtypes in multiple independent patient cohorts. J. Clin. Oncol. 24, 5079–5090 (2006). Published online 9 July 2014. 37. Bhattacharjee, A. et al. Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses. Proc. Natl Acad. 1. Paez, J. G. et al. EGFR mutations in lung cancer: correlation with clinical response to Sci. USA 98, 13790–13795 (2001). gefitinib therapy. Science 304, 1497–1500 (2004). 38. Travis, W. D. et al. International association for the study of lung cancer/American 2. Kwak, E. L. et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung Thoracic Society/European Respiratory Society international multidisciplinary cancer. N. Engl. J. Med. 363, 1693–1703 (2010). classification of lung adenocarcinoma. J. Thoracic Oncol. 6, 244–285 (2011). 3. Bergethon, K. et al. ROS1 rearrangements define a unique molecular class of lung 39. Yatabe, Y., Mitsudomi, T. & Takahashi, T. TTF-1 expression in pulmonary cancers. J. Clin Oncol. 30, 863–870 (2012). adenocarcinomas. Am. J. Surg. Pathol. 26, 767–773 (2002). 4. Drilon, A. et al. Response to cabozantinib in patients with RET fusion-positive lung 40. Shinjo, K. et al. Integrated analysis of genetic and epigenetic alterations adenocarcinomas. Cancer Discov. 3, 630–635 (2013). reveals CpG island methylator phenotype associated with distinct clinical 5. Stephens, P. et al. Lung cancer: intragenic ERBB2 kinase mutations in tumours. characters of lung adenocarcinoma. Carcinogenesis 33, 1277–1285 Nature 431, 525–526 (2004). (2012). 6. Takahashi, T. et al. p53: a frequent target for genetic abnormalities in lung cancer. 41. Mo, Q. et al. Pattern discovery and cancer gene identification in integrated cancer Science 246, 491–494 (1989). genomic data. Proc. Natl Acad. Sci. USA 110, 4245–4250 (2013). 7. Sanchez-Cespedes, M. et al. Inactivation of LKB1/STK11 is a common event in 42. Lawrence, M. S. et al. Mutational heterogeneity in cancer and the search for new adenocarcinomas of the lung. Cancer Res. 62, 3659–3662 (2002). INK4 cancer-associated genes. Nature 499, 214–218 (2013). 8. Shapiro, G. I. et al. Reciprocal Rb inactivation and p16 expression in primary lung cancers and cell lines. Cancer Res. 55, 505–509 (1995). Supplementary Information is available in the online version of the paper. 9. Singh, A. et al. Dysfunctional KEAP1–NRF2 interaction in non-small-cell lung cancer. PLoS Med. 3, e420 (2006). Acknowledgements This study was supported by NIH grants: U24 CA126561, 10. Medina, P. P. et al. Frequent BRG1/SMARCA4-inactivating mutations in human U24 CA126551, U24 CA126554, U24 CA126543, U24 CA126546, U24 lung cancer cell lines. Hum. Mutat. 29, 617–622 (2008). CA137153, U24 CA126563, U24 CA126544, U24 CA143845, U24 CA143858, U24 11. Ding, L. et al. Somatic mutations affect key pathways in lung adenocarcinoma. CA144025, U24 CA143882, U24 CA143866, U24 CA143867, U24 CA143848, Nature 455, 1069–1075 (2008). U24 CA143840, U24 CA143835, U24 CA143799, U24 CA143883, U24 CA143843, 12. Imielinski, M. et al. Mapping the hallmarks of lung adenocarcinoma with massively U54 HG003067, U54 HG003079 and U54 HG003273. We thank K. Guebert and parallel sequencing. Cell 150, 1107–1120 (2012). L. Gaffney for assistance and C. Gunter for review. 5 4 8 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved ARTICLE RESEARCH 11 11 11 11 11 Author Contributions The Cancer Genome Atlas Research Network contributed Prins ,J.S.Marron , Joel S. Parker , D. Neil Hayes , Charles M. Perou ; University collectively to this study. Biospecimens were provided by the tissue source sites and of Kentucky Jinze Liu ; The USC/JHU Epigenome Characterization Center Leslie 6 6 17 17 processed by the biospecimen core resource. Data generation and analyses were Cope , Ludmila Danilova , Daniel J. Weisenberger ,Dennis T.Maglinte , Philip H. 17 17 17 17 performed by the genome sequencing centres, cancer genome characterization Lai , Moiz S. Bootwalla , David J. Van Den Berg , Timothy Triche Jr , Stephen B. 6 17 centres and genome data analysis centres. All data were released through the data Baylin , Peter W. Laird coordinating centre. The National Cancer Institute and National Human Genome Research Institute project teams coordinated project activities. We also acknowledge 2 Genome data analysis centres: The Eli & Edythe L. Broad Institute Mara Rosenberg , the following TCGA investigators who made substantial contributions to the project: 12 12 2 2 2 2 Lynda Chin , Jianhua Zhang ,Juok Cho ,Daniel DiCara , David Heiman , Pei Lin , E. A. Collisson (manuscript coordinator); J. D. Campbell, J. Chmielecki, (analysis 2 2 2 2 2 William Mallard , Douglas Voet , Hailei Zhang , Lihua Zou , Michael S. Noble , coordinators); C. Sougnez (data coordinator); J. D. Campbell, M. Rosenberg, W. Lee, 2 2 2 2 Michael S. Lawrence , Gordon Saksena ,Nils Gehlenborg , Helga Thorvaldsdottir , J. Chmielecki, M. Ladanyi, and G. Getz (DNA sequence analysis); M. D. Wilkerson, 2 2 2 2,9,10 Jill Mesirov , Marc-Danie Nazaire , Jim Robinson , Gad Getz ; Memorial A. N. Brooks, and D. N. Hayes (mRNA sequence analysis); L. Danilova and L. Cope (DNA 4 4 4 Sloan-Kettering Cancer Center William Lee , B. Arman Aksoy , Giovanni Ciriello , methylation analysis); A. D. Cherniack (copy number analysis); M. D. Wilkerson and 1 4 4 4 Barry S. Taylor , Gideon Dresdner , Jianjiong Gao ,Benjamin Gross , Venkatraman E. A. Hadjipanayis (translocations); N. Schultz, W. Lee, E. A. Collisson, A. H. Berger, 4 4 4 4 4 4 Seshan ,MarcLadanyi , Boris Reva ,Rileen Sinha ,S.OnurSumer , Nils Weinhold , J. Chmielecki, C. J. Creighton, L. A. Byers and M. Ladanyi (pathway analysis); A. Chu and 4 4 4 Nikolaus Schultz , Ronglai Shen , Chris Sander ; University of California, Santa Cruz/ A. G. Robertson (miRNA sequence analysis); W. Travis and D. A. Wigle (pathology and 18 18 18 18 Buck Institute Sam Ng ,Singer Ma , Jingchun Zhu , Amie Radenbaugh ,Joshua clinical expertise); L. A. Byers and G. B. Mills (reverse phase protein arrays); S. B. Baylin, 18 21 21 18,22 M. Stuart , Christopher C. Benz , Christina Yau & David Haussler ; Oregon R. Govindan and M. Meyerson (project chairs). Health & Sciences University Paul T. Spellman ; University of North Carolina, 11 11 11 Author Information The primary and processed data used to generate the analyses Chapel Hill Matthew D. Wilkerson , Joel S. Parker , Katherine A. Hoadley , Patrick K. 11 11 11 presented here can be downloaded by registered users from The Cancer Genome Atlas Kimes , D. Neil Hayes , Charles M. Perou ; The University of Texas MD Anderson 12 12 12 12 at (https://tcga-data.nci.nih.gov/tcga/tcgaDownload.jsp). All of the primary sequence Cancer Center Bradley M. Broom ,JingWang ,YilingLu , Patrick Kwok Shing Ng , 12 12 12 12 files are deposited in cgHub and all other data are deposited at the Data Coordinating Lixia Diao , Lauren Averett Byers , Wenbin Liu , John V. Heymach , 12 12 12 12 Center (DCC) for public access (http://cancergenome.nih.gov/), (https:// Christopher I. Amos , John N. Weinstein , Rehan Akbani , Gordon B. Mills cghub.ucsc.edu/) and (https://tcga-data.nci.nih.gov/docs/publications/luad_2014/). Reprints and permissions information is available at www.nature.com/reprints. The Biospecimen core resource: International Genomics Consortium Erin Curley , authors declare no competing financial interests. Readers are welcome to comment on 24 24 24 24 24 Joseph Paulauskis , Kevin Lau , Scott Morris ,Troy Shelton , David Mallery , the online version of the paper. Correspondence and requests for materials should be 24 24 Johanna Gardner ,Robert Penny addressed to M.M. ([email protected]). Tissue source sites: Analytical Biological Service, Inc. Charles Saller , Katherine This work is licensed under a Creative Commons Attribution- 25 14 Tarvin ; Brigham & Women’s Hospital William G. Richards ; University of Alabama NonCommercial-ShareAlike 3.0 Unported licence. The images or other 26 26 at Birmingham Robert Cerfolio , Ayesha Bryant ; Cleveland Clinic: third party material in this article are included in the article’s Creative Commons licence, 27 27 27 Daniel P. Raymond ,Nathan A.Pennell , Carol Farver ; Christiana Care unless indicated otherwise in the credit line; if the material is not included under the 28 28 28 28 Christine Czerwinski , Lori Huelsenbeck-Dill , Mary Iacocca , Nicholas Petrelli , Creative Commons licence, users will need to obtain permission from the licence holder 28 28 28 29 Brenda Rabeno , Jennifer Brown , Thomas Bauer ; Cureline Oleg Dolzhanskiy , to reproduce the material. To view a copy of this licence, visit http://creativecommons. 29 29 29 29 Olga Potapova ,DaniilRotin ,OlgaVoronina , Elena Nemirovich-Danchenko , org/licenses/by-nc-sa/3.0 29 30 30 Konstantin V. Fedosenko ; Emory University Anthony Gal , Madhusmita Behera , 30 30 31 Suresh S. Ramalingam , Gabriel Sica ; Fox Chase Cancer Center Douglas Flieder , 31 31 32 32 Jeff Boyd , JoEllen Weaver ; ILSbio Bernard Kohl , Dang Huy Quoc Thinh ; 33 34 Indiana University George Sandusky ; Indivumed Hartmut Juhl ; John Flynn 35,36 6 The Cancer Genome Atlas Research Network Hospital Edwina Duhig ; Johns Hopkins University Peter Illei , Edward 6 6 6 6 6 Gabrielson , James Shin , Beverly Lee , Kristen Rodgers , Dante Trusty ,Malcolm V. 1 2 6 37 37 Disease analysis working group Eric A. Collisson , Joshua D. Campbell , Angela N. Brock ; Lahey Hospital & Medical Center Christina Williamson , Eric Burks , 37 37 37 2,3 2 4 2 5 Kimberly Rieger-Christ , Antonia Holway , Travis Sullivan ; Mayo Clinic Dennis A. Brooks , Alice H. Berger , William Lee , Juliann Chmielecki , David G. Beer ,Leslie 16 16 16 6 7 6 8 2,9,10 Wigle , Michael K. Asiedu , Farhad Kosari ; Memorial Sloan-Kettering Cancer Cope , Chad J. Creighton , Ludmila Danilova ,Li Ding , Gad Getz ,Peter S. 4 4 4 4 2 11 2 6 Center William D. Travis , Natasha Rekhtman , Maureen Zakowski , Valerie W. Rusch ; Hammerman , D. Neil Hayes , Bryan Hernandez , James G. Herman ,John V. 38 38 38 12 13 9 14 4 NYU Langone Medical Center Paul Zippile , James Suh , Harvey Pass , Chandra Heymach ,Igor Jurisica ,Raju Kucherlapati , David Kwiatkowski , Marc Ladanyi , 38 38 39 15 4 4 12 Goparaju , Yvonne Owusu-Sarpong ; Ontario Tumour Bank John M. S. Bartlett , Gordon Robertson , Nikolaus Schultz , Ronglai Shen , Rileen Sinha , 2 13 4 12 39 39 39 39 Sugy Kodeeswaran ,Jeremy Parfitt , Harmanjatinder Sekhon , Monique Albert ; Carrie Sougnez , Ming-Sound Tsao , William D. Travis , John N. Weinstein , 16 11 15 2 40 40 Penrose St. Francis Health Services John Eckman ,JeromeB.Myers ; Roswell Park Dennis A. Wigle , Matthew D. Wilkerson ,Andy Chu , Andrew D. Cherniack , 9 2 17 17 41 41 41 Angela Hadjipanayis , Mara Rosenberg , Daniel J. Weisenberger , Peter W. Laird , Cancer Institute Richard Cheney , Carl Morrison , Carmelo Gaudioso ; Rush 18 18 18 12 42 42 42 Amie Radenbaugh ,Singer Ma , Joshua M. Stuart , Lauren Averett Byers , University Medical Center Jeffrey A. Borgia , Philip Bonomi ,MarkPool ,MichaelJ. 6 8 2,3 42 43 43 Stephen B. Baylin , Ramaswamy Govindan , Matthew Meyerson Liptay ; St. Petersburg Academic University Fedor Moiseenko , Irina Zaytseva ; Thoraxklinik am Universitatsklinikum Heidelberg, Member of Biomaterial Bank 2 Heidelberg (BMBH) & Biobank Platform of the German Centre for Lung Research Genome sequencing centres: The Eli & Edythe L. Broad Institute Mara Rosenberg , 44 44 45 2 2 2 2 2 (DZL) Hendrik Dienemann , Michael Meister , Philipp A. Schnabel , Thomas R. Stacey B. Gabriel , Kristian Cibulskis ,Carrie Sougnez , Jaegil Kim ,Chip Stewart , 44 46 2 2,19 2 2,9,10 Muley ; University of Cologne Martin Peifer ; University of Miami Carmen Lee Lichtenstein ,Eric S.Lander , Michael S. Lawrence ,Getz ; Washington 47 47 47 8 8 8 Gomez-Fernandez , Lynn Herbert , Sophie Egea ; University of North Carolina University in St. Louis Cyriac Kandoth , Robert Fulton , Lucinda L. Fulton ,Michael D. 11 11 11 11 8 8 8 8 8 Mei Huang , Leigh B. Thorne ,Lori Boice , Ashley Hill Salazar , William K. McLellan , Richard K. Wilson , Kai Ye , Catrina C. Fronick , Christopher A. Maher , 11 11 48 8 8 8 8 Funkhouser , W. Kimryn Rathmell ; University of Pittsburgh Rajiv Dhir ,SamuelA. Christopher A. Miller , Michael C. Wendl , Christopher Cabanski ,LiDing ,Elaine 48 48 48 48 Yousem , Sanja Dacic , Frank Schneider , Jill M. Siegfried ; The University of 8 8 7 Mardis , Ramaswamy Govindan ; Baylor College of Medicine Chad J. Creighton , Texas MD Anderson Cancer Center Richard Hajek ; Washington University School of David Wheeler 8 8 8 Medicine Mark A. Watson , Sandra McDonald , Bryan Meyers ; Queensland Thoracic 35 35 35 35 Research Center Belinda Clarke , Ian A. Yang , Kwun M. Fong , Lindy Hunter , 35 35 Genome characterization centres: Canada’s Michael Smith Genome Sciences Morgan Windsor , Rayleen V. Bowman ; Center Hospitalier Universitaire Vaudois 49 49 50 Centre, British Columbia Cancer Agency Miruna Balasundaram , Yaron S. N. Solange Peters ,IgorLetovanec ; Ziauddin University Hospital Khurram Z. Khan 15 15 15 15 15 Butterfield , Rebecca Carlsen ,Andy Chu , Eric Chuah , Noreen Dhalla ,Ranabir 15 15 15 15 15 Guin , Carrie Hirst , Darlene Lee ,Haiyan I. Li , Michael Mayo , Richard A. 51 51 51 15 15 15 15 Data Coordination Centre Mark A. Jensen ,Eric E. Snyder , Deepak Srinivasan , Moore , Andrew J. Mungall , Jacqueline E. Schein , Payal Sipahimalani , Angela 51 51 51 15 15 15 15 15 Ari B. Kahn ,JulienBaboud , David A. Pot Tam , Richard Varhol , A. Gordon Robertson , Natasja Wye , Nina Thiessen , 12 15 15 Robert A. Holt , Steven J. M. Jones , Marco A. Marra ; The Eli & Edythe L. Broad 2 2,3 2 52 52 Institute Joshua D. Campbell , Angela N. Brooks , Juliann Chmielecki , Project team: National Cancer Institute Kenna R. Mills Shaw , Margi Sheth ,Tanja 2,9,10 2 9 2 2 52 52 52 52 52 Marcin Imielinski , Robert C. Onofrio ,EranHodis ,Travis Zack , Carrie Sougnez , Davidsen ,John A.Demchok , Liming Yang , Zhining Wang , Roy Tarnuzzer , 2 2 2 2 Elena Helman , Chandra Sekhar Pedamallu , Jill Mesirov , Andrew D. Cherniack , Jean Claude Zenklusen ; National Human Genome Research Institute Bradley A. 2 2 2 2 53 53 Gordon Saksena , Steven E. Schumacher , Scott L. Carter , Bryan Hernandez ,Levi Ozenberger , Heidi J. Sofia 2,3,9 2,3,9 2 2,9,10 Garraway , Rameen Beroukhim , Stacey B. Gabriel , Gad Getz , Matthew 2,3,9 Meyerson ; Harvard Medical School/Brigham & Women’s Hospital/MD Anderson 4 41 35 Expert pathology panel William D. Travis , Richard Cheney , Belinda Clarke , 9,14 9,14 12 Cancer Center Angela Hadjipanayis , Semin Lee , Harshad S. Mahadeshwar , 48 36,35 11 6 27 Sanja Dacic , Edwina Duhig , William K. Funkhouser , Peter Illei , Carol Farver , 9,14 12 9 12 12 Angeliki Pantazi , Alexei Protopopov , Xiaojia Ren , Sahil Seth , Xingzhi Song , 4 30 38 13 Natasha Rekhtman , Gabriel Sica ,James Suh & Ming-Sound Tsao 12 9 12 9 9,14 Jiabin Tang ,LixingYang , Jianhua Zhang ,Peng-ChiehChen , Michael Parfenov , 9,14 9,14 12 9,14 Andrew Wei Xu , Netty Santoso , Lynda Chin , Peter J. Park &Raju 9,14 11 1 2 Kucherlapati ; University of North Carolina, Chapel Hill Katherine A. Hoadley , University of California San Francisco, San Francisco, California 94158, USA. The Eli and 11 11 11 11 11 J. Todd Auman ,ShaowuMeng , Yan Shi , Elizabeth Buda , Scot Waring , Edythe L. Broad Institute, Cambridge, Massachusetts 02142, USA. Dana Farber Cancer 11 11 11 11 4 Umadevi Veluvolu , Donghui Tan , Piotr A. Mieczkowski , Corbin D. Jones , Janae Institute, Boston, Massachusetts 02115, USA. Memorial Sloan-Kettering Cancer Center, 11 11 11 11 5 V. Simons , Matthew G. Soloway , Tom Bodenheimer , Stuart R. Jefferys ,Jeffrey New York, New York 10065, USA. University of Michigan, Ann Arbor, Michigan 48109, 11 11 11 11 11 6 7 Roach ,Alan P. Hoyle , Junyuan Wu , Saianand Balu , Darshan Singh ,Jan F. USA. Johns Hopkins University, Baltimore, Maryland 21287, USA. Baylor College of 31 JULY 20 1 4 | V OL 51 1 | N A T U RE | 5 49 ©2014 Macmillan Publishers Limited. All rights reserved RESEARCH ARTICLE 8 33 Medicine,Houston, Texas 77030,USA. Washington University, St. Louis, Missouri 63108, 21620, USA. Indiana University School of Medicine, Indianapolis, Indiana 46202, USA. 9 10 34 35 USA. Harvard Medical School, Boston, Massachusetts 02115, USA. Massachusetts Individumed, Silver Spring, Maryland 20910, USA. The Prince Charles Hospital and General Hospital, Boston, Massachusetts 02114, USA. University of North Carolina at the University of Queensland Thoracic Research Center, Brisbane, 4032, Australia. 12 36 37 Chapel Hill, Chapel Hill, North Carolina 27599, USA. University of Texas MD Anderson Sullivan Nicolaides Pathology & John Flynn Hospital, Tugun 4680, Australia. Lahey 13 38 Cancer Center, Houston, Texas 77054, USA. Princess Margaret Cancer Centre, Toronto, Hospital and Medical Center, Burlington, Massachusetts 01805, USA. NYU Langone 14 39 Ontario M5G 2M9, Canada. Brigham and Women’s Hospital Boston, Massachusetts Medical Center, New York, New York 10016, USA. Ontario Tumour Bank, Ontario 15 16 40 02115,USA. BC Cancer Agency, Vancouver, British Columbia V5Z 4S6, Canada. Mayo Institute for Cancer Research, Toronto, Ontario M5G 0A3, Canada. Penrose St. Francis 17 41 Clinic, Rochester, Minnesota 55905, USA. University of Southern California, Los Health Services, Colorado Springs, Colorado 80907, USA. Roswell Park Cancer 18 42 Angeles, California 90033, USA. University of California Santa Cruz, Santa Cruz, Center, Buffalo, New York 14263, USA. Rush University Medical Center, Chicago, Illinois 19 43 California 95064, USA. Massachusetts Institute of Technology, Cambridge, 60612, USA. St. Petersburg Academic University, St Petersburg 199034, Russia. 20 44 Massachusetts 02142, USA. University of Kentucky, Lexington, Kentucky 40515, USA. Thoraxklinik am Universita¨tsklinikum Heidelberg, 69126 Heidelberg, Germany. 21 22 45 46 Buck Institute for Age Research, Novato, California 94945, USA. Howard Hughes University Heidelberg, 69120 Heidelberg, Germany. University of Cologne, 50931 Medical Institute, University of California Santa Cruz, Santa Cruz, California 95064, USA. Cologne, Germany. University of Miami, Sylvester Comprehensive Cancer Center, 23 24 48 Oregon Health and Science University, Portland, Oregon 97239, USA. International Miami, Florida 33136, USA. University of Pittsburgh, Pittsburgh, Pennsylvania 15213, 25 49 Genomics Consortium, Phoenix, Arizona 85004, USA. Analytical Biological Services, USA. Center Hospitalier Universitaire Vaudois, Lausanne and European Thoracic 26 50 University of Alabama at Birmingham, Ziauddin University Hospital, Inc., Wilmington, Delaware 19801, USA. Oncology Platform, CH-1011 Lausanne, Switzerland. 27 51 Birmingham, Alabama 35294, USA. Cleveland Clinic, Cleveland, Ohio 44195, USA. Karachi, 75300, Pakistan. SRA International, Inc., Fairfax, Virginia 22033, USA. 28 29 52 Christiana Care, Newark, Delaware 19713, USA. Cureline, Inc., South San Francisco, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, 30 31 53 California 94080, USA. Emory University, Atlanta, Georgia 30322, USA. Fox Chase USA. National Human Genome Research Institute, National Institutes of Health, Cancer Center, Philadelphia, Philadelphia 19111, USA. ILSbio, Chestertown, Maryland Bethesda, Maryland 20892, USA. 5 5 0 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved CORRECTIONS & AMENDMENTS CORRIGENDUM doi:10.1038/nature13879 Corrigendum: Comprehensive molecular profiling of lung adenocarcinoma The Cancer Genome Atlas Research Network Nature 511, 543–550 (2014); doi:10.1038/nature13385 In this Article, the surname of author Kristen Rodgers was incorrectly spelled Rogers. This error has been corrected in the HTML and PDF of the original paper. 262|NATURE | VOL514| 9OCTOBER 2014 ©2014 Macmillan Publishers Limited. All rights reserved CORRECTIONS & AMENDMENTS CorreCtion https://doi.org/10.1038/s41586-018-0228-6 Author Correction: Comprehensive molecular profiling of lung adenocarcinoma The Cancer Genome Atlas Research Network Correction to: Nature https://doi.org/10.1038/nature13385, published online 9 July 2014; corrected online 8 October 2014. In this Article, the Supplementary Table 7 iCLUSTER output column included incorrect cluster labels for the integrated subtypes presented in Fig. 5c. These changes affect only the iCLUSTER output column and do not affect the analysis or the conclusions of the work. The authors apologise for the error. Supplementary Table 7 has been corrected online, and the original incorrect table is provided as Supplementary Information to this Amendment for transparency. Supplementary Information is available in the online version of this Amendment. N A TURE | www.nature.com/nature © 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nature Springer Journals

Comprehensive molecular profiling of lung adenocarcinoma

Nature , Volume 511 (7511) – Jul 9, 2014
Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/comprehensive-molecular-profiling-of-lung-adenocarcinoma-P6De7OUn63

References (79)

Publisher
Springer Journals
Copyright
Copyright © 2014 by The Author(s)
Subject
Science, Humanities and Social Sciences, multidisciplinary; Science, Humanities and Social Sciences, multidisciplinary; Science, multidisciplinary
ISSN
0028-0836
eISSN
1476-4687
DOI
10.1038/nature13385
Publisher site
See Article on Publisher Site

Abstract

Frequency (%) OPEN ARTICLE doi:10.1038/nature13385 Comprehensive molecular profiling of lung adenocarcinoma The Cancer Genome Atlas Research Network* Adenocarcinoma of the lung is the leading cause of cancer death worldwide. Here we report molecular profiling of 230 resected lung adenocarcinomas using messenger RNA, microRNA and DNA sequencing integrated with copy number, methylationandproteomicanalyses.Highratesofsomaticmutationwereseen(mean8.9mutationspermegabase).Eighteen genes were statistically significantly mutated, including RIT1 activating mutations and newly described loss-of-function MGAmutationswhich aremutually exclusivewithfocalMYCamplification.EGFR mutations weremorefrequentinfemale patients, whereas mutations in RBM10 were more common in males. Aberrations in NF1, MET, ERBB2 and RIT1 occurred in 13% of cases and were enriched in samples otherwise lacking an activated oncogene, suggesting a driver role for these events in certain tumours. DNA and mRNA sequence from the same tumour highlighted splicing alterations driven by somatic genomic changes, including exon 14 skipping in MET mRNA in 4% of cases. MAPK and PI(3)K pathway activity, when measured at the protein level, was explained by known mutations in only a fraction of cases, suggesting additional, unexplained mechanisms of pathway activation. These data establish a foundation for classification and further investi- gations of lung adenocarcinoma molecular pathogenesis. 11–13 Lung cancer is the most common cause of global cancer-related mor- of passenger events per tumour genome . Our efforts focused on com- tality, leading to over a million deaths each year and adenocarcinoma is prehensive, multiplatform analysis of lung adenocarcinoma, with atten- tion towards pathobiology and clinically actionable events. its most common histological type. Smoking is the major cause of lung adenocarcinoma but, as smoking rates decrease, proportionally more Clinical samples and histopathologic data cases occur in never-smokers (defined as less than 100 cigarettes in a life- time).Recently, molecularly targeted therapies have dramaticallyimproved We analysed tumour and matched normal material from230 previously treatment for patients whose tumours harbour somatically activated onco- untreated lung adenocarcinoma patients who provided informed con- genes such as mutant EGFR or translocated ALK,RET, or ROS1 (refs 2–4). sent (Supplementary Table 1). All major histologic types of lung ade- Mutant BRAF and ERBB2 (ref. 5) are also investigational targets. How- nocarcinoma were represented: 5% lepidic, 33% acinar, 9% papillary, ever, mostlung adenocarcinomas either lack an identifiable driver onco- 14% micropapillary, 25% solid, 4% invasive mucinous, 0.4% colloid and gene, or harbour mutations in KRAS and are therefore still treated with 8% unclassifiable adenocarcinoma (Supplementary Fig. 1) . Median conventional chemotherapy. Tumour suppressor gene abnormalities, follow-up was 19 months, and 163 patients were alive at the time of last such as those in TP53 (ref. 6), STK11 (ref. 7), CDKN2A , KEAP1 (ref. 9), follow-up. Eighty-onepercent of patients reported pastor present smok- and SMARCA4 (ref. 10) are also common but are not currently clinically ing. Supplementary Table 2 summarizes demographics. DNA, RNA and actionable. Finally, lung adenocarcinoma shows high rates of somatic protein were extracted from specimens and quality-control assessments mutation and genomic rearrangement, challenging identification of all were performed as described previously . Supplementary Table 3 sum- but the most frequent driver gene alterations because of a large burden marizes molecular estimates of tumour cellularity . Transversion high Transversion low Figure 1 | Somatic mutations in lung a b Gender Number of mutations Number of mutations adenocarcinoma. a, Co-mutation plot from whole Smoking status 150 100 50 0 020 40 60 Male Female NA Ever-smoker Never-smoker exome sequencing of 230 lung adenocarcinomas. TP53 TP53 46 KRAS Data from TCGA samples were combined with KRAS 33 EGFR KEAP1 17 STK11 previously published data for statistical analysis. STK11 17 KEAP1 EGFR 14 NF1 Co-mutation plot for all samples used in the SMARCA4 NF1 11 RBM10 BRAF 10 statistical analysis (n5 412) can be found in PIK3CA SETD2 9 RB1 RBM10 8 Supplementary Fig. 2. Significant genes with a U2AF1 MGA 8 ERBB2 MET 7 corrected P value less than 0.025 were identified ARID1A 7 PIK3CA 7 using the MutSig2CV algorithm and are ranked c Males Females SMARCA4 6 in order of decreasing prevalence. b, c,The RB1 4 Number of mutations Number of mutations CDKN2A 4 60 40 20 0 0 20 40 60 differential patterns of mutation between samples U2AF1 3 EGFR RIT1 2 classified as transversion high and transversion low STK11 Missense Splice site Frameshift SMARCA4 Nonsense In-frame indel samples (b) or male and female patients (c) are RBM10 shown for all samples used in the statistical analysis (n5 412). Stars indicate statistical significance Q < 0.05 Missense Frameshift P < 0.05 Splice site In-frame indel using the Fisher’s exact test (black stars: q, 0.05, Nonsense Other non-synonymous grey stars: P, 0.05) and are adjacent to the sample Transversions Transitions Indels, other set with the higher percentage of mutated samples. *A list of authors and affiliations appears at the end of the paper. 31 J U LY 2014 | V O L 5 1 1 | N ATU RE | 5 4 3 ©2014 Macmillan Publishers Limited. All rights reserved Percentage Normalized RNA-seq read coverage RESEARCH ARTICLE Somatically acquired DNA alterations MDM2,KRAS,EGFR,MET,CCNE1,CCND1,TERC and MECOM (Sup- plementary Table 6), as previously described ,8q24near MYC, and a We performed whole-exome sequencing (WES) on tumour and germ- novel peak containing CCND3 (Supplementary Table 6). The CDKN2A lineDNA,withameancoverageof97.63 and 95.83, respectively, as per- locus was the most significant deletion (Supplementary Table 6). Sup- formed previously . The mean somatic mutation rate across the TCGA plementary Table 7 summarizes molecular and clinical characteristics cohort was 8.87 mutations per megabase (Mb) of DNA (range: 0.5–48, by sample. Low-pass whole-genome sequencing on a subset (n5 93) of median: 5.78). The non-synonymous mutation rate was 6.86 per Mb. MutSig2CV identified significantly mutated genes among our 230 the samples revealed an average of 36 gene–gene and gene–inter-gene cases along with 182 similarly-sequenced, previously reported lung adenocarcinomas . Analysis of these 412 tumour/normal pairs high- lighted 18 statistically significant mutated genes (Fig. 1a shows co-mutation a Exon Exon 13 20 plot of TCGA samples (n5 230),Supplementary Fig. 2 shows co-mutation EML4–ALK plot of all samples used in the statistical analysis (n5 412) and Sup- plementary Table 4 contains complete MutSig2CV results, which also EML4–ALK appear on the TCGA Data Portal along with many associated data files (https://tcga-data.nci.nih.gov/docs/publications/luad_2014/). TP53 was EML4–ALK commonly mutated (46%). Mutations in KRAS (33%) were mutually 11 12 TRIM33–RET exclusive with those in EGFR (14%). BRAF was also commonly mutated (10%), as were PIK3CA (7%), MET (7%) and the small GTPase gene, RIT1 CCDC6–RET (2%). Mutations in tumour suppressor genes including STK11 (17%), 10 34 KEAP1 (17%), NF1 (11%),RB1 (4%) and CDKN2A (4%) were observed. EZR–ROS1 Mutations in chromatin modifying genes SETD2 (9%), ARID1A (7%) and SMARCA4 (6%) and the RNA splicing genes RBM10 (8%) and U2AF1 CD74–ROS1 (3%) were also common. Recurrent mutations in the MGA gene (which 31 35 encodes a Max-interacting protein on the MYC pathway ) occurred in CLTC–ROS1 14 32–34 8% of samples. Loss-of-function (frameshift and nonsense) mutations SLC34A2–ROS1 inMGA were mutually exclusive with focal MYC amplification (Fisher’s exact test P5 0.04), suggesting a hitherto unappreciated potential mech- Portion of original transcripts not in fusion transcript: anism of MYC pathway activation. Coding single nucleotide variants and Normalized, exonic mRNA expression: Low High indel variants were verified by resequencing at a rate of 99% and 100%, respectively (Supplementary Fig. 3a, Supplementary Table 5). Tumour TCGA-99-7458 Exon 14 skipping Number of samples purity was not associated with the presence of false negatives identified None in the validation data (P5 0.31; Supplementary Fig. 3b). (0% skipping) 199 0 0 0 Past or present smoking associated with cytosine to adenine (C.A) TCGA-75-6205 1 11 nucleotide transversions as previouslydescribed both in individual genes Intermediate 12,13 and genome-wide .C. A nucleotide transversion fraction showed (60–80% skipping) 0 1 1 1 two peaks; this fraction correlated with total mutation count (R 5 0.30) and inversely correlated with cytosine to thymine (C. T) transition fre- 27 TCGA-44-6775 Full quency (R 5 0.75) (Supplementary Fig. 4). We classified each sample (90–100% skipping) 0 5 1 0 (Supplementary Methods) into one of two groups named transversion- high (TH,n5 269), and transversion-low (TL,n5 144). The transversion- Y1003 high group was strongly associated with past or present smoking (P , 13 14 15 216 MET mutations 2.23 10 ), consistent with previous reports . The transversion-high and transversion-low patient cohorts harboured different gene mutations. Observed splicing across all tumours Whereas KRAS mutations were significantly enriched in the transversion- (total events = 29,867) high cohort(P5 2.13 10 ),EGFRmutations weresignificantlyenriched 26 Associated with U2AF1 S34F mutation in the transversion-low group (P5 3.33 10 ). PIK3CA and RB1 muta- (total events = 129; q value < 0.05 ) tions were likewise enriched in transversion-low tumours (P, 0.05). 0.0 0.2 0.4 0.6 0.8 1.0 Additionally, the transversion-low tumours were specifically enriched Proportion for in-frame insertions in EGFR and ERBB2 (ref. 5) and for frameshift Cassette exon Coordinate cassette exons Mutually exclusive exon *P < 0.001 indels in RB1 (Fig. 1b). RB1 is commonly mutated in small-cell lung carcinoma (SCLC). We found RB1 mutations in transversion-low ade- Alternative 5′ splice site Alternative 3′ splice site Alternative first exon Alternative last exon nocarcinomas were enriched for frameshift indels versus single nucleotide 20,21 substitutions compared to SCLC (P, 0.05) suggesting a mutational Figure 2 | Aberrant RNA transcripts in lung adenocarcinoma associated mechanism in transversion-low adenocarcinoma that is probably dis- with somatic DNA translocation or mutation. a, Normalized exon level RNA tinct from smoking in SCLC. expression across fusion gene partners. Grey boxes around genes mark the Gender is correlated with mutation patterns in lung adenocarcinoma . regions that are removed as a consequence of the fusion. Junction points of the Onlyafractionofsignificantlymutatedgenesfromthecompletesetreported fusion events are also listed in Supplementary Table 9. Exon numbers refer to reference transcripts listed in Supplementary Table 9. b, MET exon 14 in this study (Fig. 1a) were enriched in men or women (Fig. 1c). EGFR skipping observed in the presence of exon 14 splice site mutation (ss mut), mutations were enriched in tumours from the female cohort (P5 0.03) splice site deletion (ss del) or a Y1003* mutation. A total of 22 samples had whereas loss-of-function mutations within RBM10, an RNA-binding pro- 23 insufficient coverage around exon 14 for quantification. The percentage tein located on the X chromosome were enriched in tumours from men skipping is (total expression minus exon 14 expression)/total expression. (P5 0.002). When examining the transversion-high group, 16 out of 21 c, Significant differences in the frequency of 129 alternative splicing events in RBM10 mutations were observed in males (P5 0.003, Fisher’s exact test). mRNA from tumours with U2AF1 S34F tumours compared to U2AF1 WT Somatic copy number alterations were very similar to those previ- tumours (q value ,0.05). Consistent with the function of U2AF1 in 39 splice ously reported for lung adenocarcinoma (Supplementary Fig. 5, Sup- site recognition, most splicing differences involved cassette exon and alternative 39 splice site events (chi-squared test, P, 0.001). plementary Table 6). Significant amplifications included NKX2-1, TERT, 5 4 4 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved WT ss mut ss del Y1003* Frequency (%) ARTICLE RESEARCH rearrangements per tumour. Chromothripsis occurred in six of the Candidate driver genes 93 samples (6%) (Supplementary Fig. 6, Supplementary Table 8). Low- The receptor tyrosine kinase (RTK)/RAS/RAF pathway is frequently pass whole genome sequencing-detected rearrangements appear in mutated in lung adenocarcinoma. Striking therapeutic responses are Supplementary Table 9. often achieved when mutant pathway components are successfully inhib- ited. Sixty-two per cent (143/230) of tumours harboured known activating Description of aberrant RNA transcripts mutations in known driver oncogenes, as defined by others . Cancer- Gene fusions, splice site mutations or mutations in genes encoding splic- associated mutations in KRAS (32%, n5 74), EGFR (11%, n5 26) and ing factors promote or sustain the malignant phenotype by generating BRAF (7%, n5 16) were common. Additional, previously uncharac- aberrant RNA transcripts. Combining DNA with mRNA sequencing terized KRAS, EGFR and BRAF mutations were observed, but were not enabled us to catalogue aberrant RNA transcripts and, in many cases, classified as driver oncogenes for the purposes of our analyses (see Sup- to identify the DNA-encoded mechanism for the aberration. Seventy- plementary Fig. 9a for depiction of all mutations of known and unknown five per cent of somatic mutations identified by WES were present in the significance); explaining the differing mutation frequencies in each gene RNA transcriptome when the locus in question was expressed (minimum between this analysis and the overall mutational analysis described above. 53) (Supplementary Fig. 7a) similar to prior analyses . Previously iden- We also identified known activating ERBB2 in-frame insertion and point tified fusions involving ALK (3/230 cases), ROS1 (4/230) and RET mutations (n5 5) , as well as mutations in MAP2K1 (n5 2), NRAS and (2/230) (Fig. 2a, Supplementary Table 10), all occurred in transversion- HRAS (n5 1 each). RNA sequencing revealed the aforementioned MET low tumours (P5 1.853 10 , Fisher’s exact test). exon 14 skipping (n5 10) and fusions involving ROS1 (n5 4), ALK MET activation can occur by exon 14 skipping, which results in a (n5 3) and RET (n5 2). We considered these tumours collectively as stabilized protein . Ten tumours had somatic MET DNA alterations oncogene-positive, as they harboured a known activating RTK/RAS/ with MET exon 14 skipping in RNA. In nine of these samples, a 59 or RAF pathway somatic event. DNA amplification events were not con- 39 splice site mutation or deletion was identified . MET exon 14 skip- sidered to be driver events before the comparisons described below. ping was also found in the setting of a MET Y1003* stop codon muta- We sought to nominate previously unrecognized genomic events that tion (Fig. 2b, Supplementary Fig. 8a). The codon affected by the Y1003* might activate this critical pathway in the 38% of samples without a mutation is predicted to disrupt multiple splicing enhancer sequences, RTK/RAS/RAF oncogene mutation. Tumour cellularity did not differ but the mechanism of skipping remains unknown in this case. between oncogene-negative and oncogene-positive samples (Supplemen- S34F mutations in U2AF1 have recently been reported in lung ade- tary Fig. 9b). Analysis of copy number alterations using GISTIC identified nocarcinoma but their contribution to oncogenesis remains unknown. unique focal ERBB2 and MET amplifications in the oncogene-negative S34F Eight samples harboured U2AF1 . We identified 129 splicing events subset (Fig. 3a, Supplementary Table 6); amplifications in other wild-type S34F strongly associated with U2AF1 mutation, consistent with the role of proto-oncogenes, including KRAS and EGFR, were not significantly U2AF1 in 39-splice site selection . Cassette exons and alternative 39 splice different between the two groups. sites were most commonly affected (Fig. 2c, Supplementary Table 11) . We next analysed WES data independently in the oncogene-negative Among these events, alternative splicing of the CTNNB1 proto-oncogene and oncogene-positive subsets. We found that TP53, KEAP1, NF1 and was strongly associated with U2AF1 mutations (Supplementary Fig. 8b). RIT1 mutations were significantly enriched in oncogene-negative tumours Thus, concurrent analysis of DNA and RNA enabled delineation of (P, 0.01; Fig. 3b, Supplementary Table 12). NF1 mutations have previ- both cis and trans mechanisms governing RNA processing in lung ously been reported in lung adenocarcinoma , but this is the first study, adenocarcinoma. to our knowledge, capable of identifying all classes of loss-of-function a b –16 0.6 Oncogene-positive Oncogene-positive 0.5 Oncogene-negative Oncogene-negative –8 0.4 MET ERBB2 0.3 –4 0.2 –2 0.1 0.0 0.1 TP53 KEAP1 NF1 RIT1 HRAS (0.4%) NRAS (0.4%) Chromosome c d RET fusion (0.9%) Previously MAP2K1 (0.9%) Oncogene-positive oncogene-negative ALK fusion (1.3%) (62%, n = 143) (13%, n = 31) RIT1 (2.2%) ROS1 fusion (1.7%) KRAS ERBB2 amp (0.9%) ERBB2 (1.7%) MET amp (2.2%) MET ex14 (4.3%) EGFR 11 NF1 BRAF 7 (8.3%) ROS1/ALK/RET 4 BRAF MAP2K1 / HRAS / NRAS (7.0%) MET None EGFR ERBB2 (24.4%) (11.3%) RIT1 2 KRAS NF1 (32.2%) Amplification Fusion Missense mutation Exon skipping In-frame indel Nonsense mutation / frameshift indel / splice-site mutation Figure 3 | Identification of novel candidate driver genes. a, GISTIC analysis adenocarcinoma. Not shown are the 63 tumours lacking an identifiable driver of focal amplifications in oncogene-negative (n5 87) and oncogene-positive lesion. Only canonical driver events, as defined in Supplementary Fig. 9, and (n5 143) TCGA samples identifies focal gains of MET and ERBB2 that are proposed driver events, are shown; hence not every alteration found is specific to the oncogene-negative set (purple). b, TP53, KEAP1, NF1 and RIT1 displayed. d, New candidate driver oncogenes (blue: 13% of cases) and known mutations are significantly enriched in samples otherwise lacking oncogene somatically activated drivers events (red: 63%) that activate the RTK/RAS/RAF mutations (adjusted P, 0.05 by Fisher’s exact test). c, Co-mutation plot of pathway can be found in the majority of the 230 lung adenocarcinomas. variants of known significance within the RTK/RAS/RAF pathway in lung 31 J U LY 2014 | V O L 5 1 1 | N ATU RE | 5 4 5 ©2014 Macmillan Publishers Limited. All rights reserved FDR q Per cent mutated MAPK pathway score mTOR pathway score RESEARCH ARTICLE NF1 defects and to statistically demonstrate thatNF1 mutations, as well Recurrent alterations in key pathways as KEAP1 and TP53 mutations are enriched in the oncogene-negative Recurrent aberrations in multiple key pathways and processes charac- subset of lung adenocarcinomas (Fig. 3c). All RIT1 mutations occurred terize lung adenocarcinoma (Fig. 4a). Among these were RTK/RAS/ in the oncogene-negative subset and clustered around residue Q79 (homol- RAF pathway activation (76% of cases), PI(3)K-mTOR pathway activa- ogous to Q61 in the switch II region of RAS genes). These mutations tion (25%), p53 pathway alteration (63%), alteration of cell cycle regu- transform NIH3T3 cells and activate MAPK and PI(3)K signalling , lators (64%, Supplementary Fig. 10), alteration of oxidative stress pathways supporting a driver role for mutant RIT1 in 2% of lung adenocarcinomas. (22%, Supplementary Fig. 11), and mutation of various chromatin and This analysis increases the rate at which putative somatic lung adeno- RNA splicing factors (49%). carcinoma driver events can be identified within the RTK/RAS/RAF We then examined the phenotypic sequelae of some key genomic pathway to 76% (Fig. 3d). events in the tumours in which they occurred. Reverse-phase protein arrays provided proteomic and phosphoproteomic phenotypic evidence of pathway activity. Antibodies on this platform are listed in Supplemen- EGFR ERBB2 MET ALK RET ROS1 tary Table 13. This analysis suggested that DNA sequencing did not 11% 3% 7% 1% <1% 2% identify all samples with phosphoprotein evidence of activation of a PTEN given signalling pathway. For example, whereas KRAS-mutant lung ade- 3% PIK3CA KRAS NRAS 4% nocarcinomas had higher levels of phosphorylated MAPK than KRAS 32% <1% PIK3R1 NF1 <1% wild-type tumours had on average, many KRAS wild-type tumours dis- 11% HRAS RIT1 <1% 2% STK11 AKT1 played significant MAPK pathway activation (Fig. 4b, Supplementary 17% 1% Fig. 10). The multiple mechanisms by which lung adenocarcinomas Per cent of cases (%) BRAF achieve MAPK activation suggest additional, still undetected RTK/RAS/ 7% AMPK TSC1/2 RAF pathway alterations. Similarly, we found significant activation of 50 0 100 Inactivated Activated mTOR and its effectors (p70S6kinase, S6, 4E-BP1) in a substantial frac- MAP2K1 Activation Inhibition tion of the tumours (Fig. 4c). Analysis of mutations in PIK3CA and <1% MTOR 33 STK11, STK11 protein levels, and AMPK and AKT phosphorylation led to the identification of three major mTOR patterns in lung adeno- CDKN2A carcinoma: (1) tumours with minimal or basal mTOR pathway activa- Proliferation, cell survival, translation 43% tion, (2) tumours showing higher mTOR activity accompanied by either KEAP1 CUL3 MDM2 ATM CCND1 CDK4 CCNE1 STK11-inactivating mutation or combined low STK11 expression and 19% <1% 8% 9% 4% 7% 3% low AMPK activation and (3) tumours showing high mTOR activity accompanied by either phosphorylated AKT activation, PIK3CA muta- NFE2L2 TP53 RB1 tion, or both. As with MAPK, many tumours lack an obvious underlying 3% 46% 7% genomic alteration to explain their apparent mTOR activation. Oxidative Proliferation, Cell cycle stress response cell survival progression Molecular subtypes of lung adenocarcinoma ARID1A ARID1B Broad transcriptional and epigenetic profiling can reveal downstream 7% 6% consequences of driver mutations, provide clinically relevant classifica- ARID2 SMARCA4 SETD2 U2AF1 RBM10 tion and offer insight into tumours lacking clear drivers. Prior unsuper- 7% 6% 9% 4% 9% vised analyses of lung adenocarcinoma gene expression have used varying 34–37 nomenclature for transcriptional subtypes of the disease .Tocoor- Nucleosome Histone RNA splicing / remodelling methylation processing dinate naming of the transcriptional subtypes with the histopathological , anatomic and mutational classifications of lung adenocarcinoma, we MAPK pathway propose an updated nomenclature: the terminal respiratory unit (TRU, KRAS mut n n = 53 = n = 128 formerly bronchioid), the proximal-inflammatory (PI, formerly squa- p-JNK p-MAPK moid), and the proximal-proliferative (PP, formerly magnoid) transcrip- p-MEK1 tional subtypes (Fig. 5a). Previously reported associations of expression p-p38 34,36,39 p-p90RSK –5 signatures with pathways and clinical outcomes were observed (Sup- p-Shc plementary Fig. 7b) and integration with multi-analyte data revealed P < 0.01 p-c-Raf –10 Subtype statistically significant genomic alterations associated with these tran- Pathway KRAS KRAS scriptional subtypes. The PP subtype was enriched for mutation of KRAS, score mut wt along with inactivation of the STK11 tumour suppressor gene by chro- PI(3)K pathway mosomal loss, inactivating mutation, and reduced gene expression. In PIK3CA STK11 High Low mut mut p-AKT p-AMPK Unaligned ** contrast, the PI subtype was characterized by solid histopathology and ( n = 9) ( n = 42) ( n = 35) ( n = 21) ( n = 74) ** PIK3CA mut Figure 4 | Pathway alterations in lung adenocarcinoma. a, Somatic STK 11 mut PTEN loss alterations involving key pathway components for RTK signalling, mTOR signalling, oxidative stress response, proliferation and cell cycle progression, STK 11/LKB1 p-AMPK nucleosome remodelling, histone methylation, and RNA splicing/processing. p-AKT b, c, Proteomic analysis by RPPA (n5 181) P values by two-sided t-test. p-mTOR –1 p-4E-BP1 Box plots represent 5%, 25%, 75%, median, and 95%. PP, proximal –2 p-p70S6K proliferative; TRU, terminal respiratory unit; PI, proximal inflammatory. p-S6 c, mTOR signalling may be activated, by either Akt (for example, via PI(3)K) or Subtype Pathway inactivation of AMPK (for example, via STK11 loss). Tumours were separated score into three main groups: those with PI(3)K-AKT activation, through either PIK3CA activating mutation or unknown mechanism (high p-AKT); those PI3K-Akt branch active LKB1-AMPK inactive with LKB1-AMPK inactivation, through either STK11 mutation or unknown Protein expression Expression subtype Pathway signature *P < 0.01 mechanism with low levels of LKB1 and p-AMPK; and those showing none Low High PP TRU PI Low High **P < 0.001 of the above features. 5 4 6 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved PIK3CA mut High p-AKT STK11 mut Low p-AMPK Unaligned ARTICLE RESEARCH a Expression subtypes c Integrated subtypes 16 2 3 4 5 Expression subtype Proximal proliferative Proximal inflammatory Terminal respiratory unit DNA methylation subtype mut STK11 CN del p16 methylation under expr. Ploidy mut KEAP1 Non-silent mutation rate mut KRAS Purity mut TP53 mut NF1 methylation p16 Fusions mut EGFR over expr. TTF-1 Mutation total 8 Ploidy Purity CpG T % 11 Never-smoker Female Histology 17 b DNA methylation subtypes DNA methylation subtype Expression subtype Expression, ploidy, purity, mutation rates CIMP-high CIMP-intermediate CIMP-low Normal CIMP-high Proximal proliferative CIMP-intermediate Proximal inflammatory CIMP-low Terminal respiratory unit Low High GATA4 (TRU) SFRP1 GATA5 Integrated subtype Fusion Histology 1 Solid iClust1 4 iClust4 ALK WIF1 Acinar 2 iClust2 5 iClust5 ROS1 Lepidic GATA2 3 iClust3 iClust6 RET Papillary/Micropapillary DNA copy number DNA methylation CDKN2A Mucinous RASSF1 Other –1.0 0 1.0 01.0 SOX17 HOXD1 Mutation Smoking status Gender HOXA9 Mutant Never-smoker Female Concurrent p16 methylation and SETD2 mutation HIC1 Figure 5 | Integrative analysis. a–c, Integrating unsupervised analyses of 230 iCluster analysis (c). All displayed features are significantly associated with lung adenocarcinomas reveals significant interactions between molecular subtypes depicted. The CIMP phenotype is defined by the most variable CpG subtypes. Tumours are displayed as columns, grouped by mRNA expression island and promoter probes. subtypes (a), DNA methylation subtypes (b), and integrated subtypes by co-mutation of NF1 and TP53. Finally, the TRU subtype harboured the and CDKN2A methylation. Tumours with low CDKN2A expression majorityof theEGFR-mutatedtumours as wellas thekinase fusion express- due to methylation (rather than due to mutation or deletion) had lower ing tumours. TRU subtype membership was prognostically favourable, ploidy, fewer overall mutations (Fig. 5c) and were significantly enriched for SETD2 mutation, suggesting an important role for this chromatin- as seen previously (Supplementary Fig. 7c). Finally, the subtypes exhib- ited different mutation rates, transition frequencies, genomic ploidy pro- modifying gene in the development of certain tumours. Integrative clustering of copy number, DNA methylation and mRNA files, patterns of large-scale aberration, and differed in their association with smoking history (Fig. 5a). Unsupervised clustering of miRNA expression data found six clusters (Fig. 5c). Tumour ploidy and mutation sequencing-derived or reverse phase protein array (RPPA)-derived data rate are higher in clusters 1–3 than in clusters 4–6. Clusters 1–3 frequently also revealed significant heterogeneity, partially overlapping with the harbour TP53 mutations and are enriched for the two proximal tran- mRNA-based subtypes, as demonstrated in Supplementary Figs 12 and 13. scriptional subtypes. Fisher’s combined probability tests revealed signi- Mutations in chromatin-modifying genes (for example, SMARCA4, ficant copy number associated gene expression changes on 3q in cluster one, 8q in cluster two, and chromosome 7 and 15q in cluster three (Sup- ARID1A and SETD2) suggest a major role for chromatin maintenance in lung adenocarcinoma. To examine chromatin states in an unbiased plementary Fig. 15). The low ploidy and low mutation rate clusters four manner, we selected the most variable DNA methylation-specific probes and five contain many TRU samples, whereas tumours in cluster 6 have in CpG island promoter regions and clustered them by methylation inten- comparatively lower tumour cellularity, and few other distinguishing sity (Supplementary Table 14). This analysis divided samples into two molecular features. Significant copy number-associated gene expres- distinct subsets: a significantly altered CpG island methylator phenotype- sion changes are observed on 6q in cluster four and 19p in cluster five. high (CIMP-H(igh)) cluster and a more normal-like CIMP-L(ow) group, The CIMP-H tumours divided into a high ploidy, high mutation rate, proximal-inflammatory CIMP-H group (cluster 3) and a low ploidy, low with a third set of samples occupying an intermediate level of methy- lation at CIMP sites (Fig. 5b). Our results confirm a prior report and mutation rate, TRU-associated CIMP-H group (cluster 4), suggesting that the CIMP phenotype in lung adenocarcinoma can occur in markedly provide additional insights into this epigenetic program.CIMP-H tumours often showed DNA hypermethylation of several key genes: CDKN2A, different genomic and transcriptional contexts. Furthermore, cluster four is enriched for CDKN2A methylation and SETD2 mutations, sug- GATA2, GATA4, GATA5, HIC1, HOXA9, HOXD13, RASSF1, SFRP1, SOX17 and WIF1 among others (Supplementary Fig. 14). WNT pathway gestingan interaction between somatic mutationof SETD2 andderegulated chromatin maintenance in this subtype. Finally, cluster membership genes are significantly over-represented in this list (P value5 0.0015) suggesting that this is a key pathway with an important driving role was significantly associated with mutations in TP53, EGFR and STK11 within this subtype. MYC overexpression was significantly associated (Supplementary Fig. 15, Supplementary Table 6). with the CIMP-H phenotype as well (P5 0.003). Conclusions Although we did not find significant correlations between global DNA methylation patterns and individual mutations in chromatin remodel- We assessed themutation profiles, structural rearrangements, copy number ling genes, there was an intriguing association between SETD2 mutation alterations, DNA methylation, mRNA, miRNA and protein expression 31 JU LY 2 0 14 | V OL 5 1 1 | N AT URE | 5 4 7 ©2014 Macmillan Publishers Limited. All rights reserved DNA copy number RESEARCH ARTICLE 13. Govindan, R. et al. Genomic landscape of non-small cell lung cancer in smokers of 230 lung adenocarcinomas. In recent years, the treatment of lung and never-smokers. Cell 150, 1121–1134 (2012). adenocarcinoma has been advanced by the development of multiple 14. Travis, W. D., Brambilla, E. & Riely, G. J. New pathologic classification of lung therapies targeted against alterations in the RTK/RAS/RAF pathway. We cancer: relevance for clinical practice and clinical trials. J. Clin. Oncol. 31, 992–1001 (2013). nominate amplifications in MET and ERBB2 as well as mutations of 15. The Cancer Genome Atlas Research Network Comprehensive genomic NF1 and RIT1 as driver events specifically in otherwise oncogene-negative characterization of squamous cell lung cancers. Nature 489, 519–525 lung adenocarcinomas. This analysis increases the fraction of lung ade- (2012). nocarcinoma cases with somatic evidence of RTK/RAS/RAF activation 16. Carter, S. L. et al. Absolute quantification of somatic DNA alterations in human cancer. Nature Biotechnol. 30, 413–421 (2012). from 62% to 76%. While all lung adenocarcinomas may activate this 17. Cibulskis, K. et al. Sensitive detection of somatic point mutations in impure and pathway by some mechanism, only a subset show tonic pathway acti- heterogeneous cancer samples. Nature Biotechnol. 31, 213–219 (2013). vation at the protein level, suggesting both diversity between tumours 18. Lawrence, M. S. et al. Discovery and saturation analysis of cancer genes across 21 tumour types. Nature 505, 495–501 (2014). with seemingly similar activating events and as yet undescribed mech- 19. Hurlin, P. J., Steingrimsson, E., Copeland, N. G., Jenkins, N. A. & Eisenman, R. N. anisms of pathway activation. Therefore, the current study expands the Mga, a dual-specificity transcription factor that interacts with Max and contains a range of possible targetable alterations within the RTK/RAS/RAF path- T-domain DNA-binding motif. EMBO J. 18, 7019–7028 (1999). way in general and suggests increased implementation of MET and 20. Peifer, M. et al. Integrative genome analyses identify key somatic driver mutations of small-cell lung cancer. Nature Genet. 44, 1104–1110 (2012). ERBB2/HER2 inhibitors in particular. Our discovery of inactivating 21. Rudin, C. M. et al. Comprehensive genomic analysis identifies SOX2 as a mutations of MGA further underscores the importance of the MYC frequently amplified gene in small-cell lung cancer. Nature Genet. 44, 1111–1116 pathway in lung adenocarcinoma. (2012). 22. Tokumo, M. et al. The relationship between epidermal growth factor receptor This study further implicates both chromatin modifications and splic- mutations and clinicopathologic features in non-small cell lung cancers. ing alterations in lung adenocarcinoma through the integration of DNA, Clin. Cancer Res. 11, 1167–1173 (2005). transcriptome and methylome analysis. We identified alternative splic- 23. Coleman, M. P. et al. A novel gene, DXS8237E, lies within 20 kb upstream of UBE1 ing due to both splicing factor mutations in trans and mutation of splice in Xp11.23 and has a different X inactivation status. Genomics 31, 135–138 (1996). sites in cis, the latter leading to activation of the MET gene by exon 14 24. Weir, B. A. et al. Characterizing the cancer genome in lung adenocarcinoma. skipping. Cluster analysis separated tumours based on single-gene driver Nature 450, 893–898 (2007). events as well as large-scale aberrations, emphasizing lung adenocarci- 25. Stephens, P. J. et al. Massive genomic rearrangement acquired in a single catastrophic event during cancer development. Cell 144, 27–40 (2011). noma’s molecular heterogeneity and combinatorial alterations, includ- 26. Kong-Beltran, M. et al. Somatic mutations lead to an oncogenic deletion of Met in ing the identification of coincident SETD2 mutations and CDKN2A lung cancer. Cancer Res. 66, 283–289 (2006). methylation in a subset of CIMP-H tumours, providing evidence of a 27. Seo, J. S. et al. The transcriptional landscape and mutational profile of lung somatic event associated with a genome-wide methylation phenotype. adenocarcinoma. Genome Res. 22, 2109–2119 (2012). 28. Wu, S., Romfo, C. M., Nilsen, T. W. & Green, M. R. Functional recognition of These studies provide new knowledge by illuminating modes of geno- the 39 splice site AG by the splicing factor U2AF . Nature 402, 832–835 mic alteration, highlighting previously unappreciated altered genes, and (1999). enabling further refinement in sub-classification for the improved per- 29. Brooks, A. N. et al. A pan-cancer analysis of transcriptome changes associated with somatic mutations in U2AF1 reveals commonly altered splicing events. PLoS ONE sonalization of treatment for this deadly disease. 9, e87361 (2014). 30. Pao, W. & Hutchinson, K. E. Chipping away at the lung cancer genome. Nature Med. METHODS SUMMARY 18, 349–351 (2012). All specimens were obtained from patients with appropriate consent from the rele- 31. Beroukhim, R. et al. Assessing the significance of chromosomal aberrations in cancer: methodology and application to glioma. Proc. Natl Acad. Sci. USA 104, vant institutional review board. DNA and RNA were collected from samples using 20007–20012 (2007). the Allprep kit (Qiagen). We used standard approaches for capture and sequencing of 32. Berger, A. H. et al. Oncogenic RIT1 mutations in lung adenocarcinoma. Oncogene exomes from tumour DNA and normal DNA and whole-genome shotgun sequenc- http://dx.doi.org/10.1038/onc.2013.581 (2014). ing. Significantly mutated genes were identified by comparing them with expectation 33. Creighton, C. J. et al. Proteomic and transcriptomic profiling reveals a link between models based on the exact measured rates of specific sequence lesions . GISTIC the PI3K pathway and lower estrogen-receptor (ER) levels and activity in ER analysis of the circular-binary-segmentedAffymetrix SNP 6.0 copy numberdata was breast cancer. Breast Cancer Res. 12, R40 (2010). 34. Wilkerson, M. D. et al. Differential pathogenesis of lung adenocarcinoma subtypes used to identify recurrent amplification and deletion peaks . Consensus clustering involving sequence mutations, copy number, chromosomal instability, and approaches were used to analyse mRNA, miRNA and methylation subtypes using methylation. PLoS ONE 7, e36530 (2012). previous approaches . The publication web page is (https://tcga-data.nci.nih.gov/ 35. Beer, D. G. et al. Gene-expression profiles predict survival of patients with lung docs/publications/luad_2014/). Sequence files are in CGHub (https://cghub.ucsc.edu/). adenocarcinoma. Nature Med. 8, 816–824 (2002). 36. Hayes, D. N. et al. Gene expression profiling reveals reproducible human lung Received 11 June 2013; accepted 22 April 2014. adenocarcinoma subtypes in multiple independent patient cohorts. J. Clin. Oncol. 24, 5079–5090 (2006). Published online 9 July 2014. 37. Bhattacharjee, A. et al. Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses. Proc. Natl Acad. 1. Paez, J. G. et al. EGFR mutations in lung cancer: correlation with clinical response to Sci. USA 98, 13790–13795 (2001). gefitinib therapy. Science 304, 1497–1500 (2004). 38. Travis, W. D. et al. International association for the study of lung cancer/American 2. Kwak, E. L. et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung Thoracic Society/European Respiratory Society international multidisciplinary cancer. N. Engl. J. Med. 363, 1693–1703 (2010). classification of lung adenocarcinoma. J. Thoracic Oncol. 6, 244–285 (2011). 3. Bergethon, K. et al. ROS1 rearrangements define a unique molecular class of lung 39. Yatabe, Y., Mitsudomi, T. & Takahashi, T. TTF-1 expression in pulmonary cancers. J. Clin Oncol. 30, 863–870 (2012). adenocarcinomas. Am. J. Surg. Pathol. 26, 767–773 (2002). 4. Drilon, A. et al. Response to cabozantinib in patients with RET fusion-positive lung 40. Shinjo, K. et al. Integrated analysis of genetic and epigenetic alterations adenocarcinomas. Cancer Discov. 3, 630–635 (2013). reveals CpG island methylator phenotype associated with distinct clinical 5. Stephens, P. et al. Lung cancer: intragenic ERBB2 kinase mutations in tumours. characters of lung adenocarcinoma. Carcinogenesis 33, 1277–1285 Nature 431, 525–526 (2004). (2012). 6. Takahashi, T. et al. p53: a frequent target for genetic abnormalities in lung cancer. 41. Mo, Q. et al. Pattern discovery and cancer gene identification in integrated cancer Science 246, 491–494 (1989). genomic data. Proc. Natl Acad. Sci. USA 110, 4245–4250 (2013). 7. Sanchez-Cespedes, M. et al. Inactivation of LKB1/STK11 is a common event in 42. Lawrence, M. S. et al. Mutational heterogeneity in cancer and the search for new adenocarcinomas of the lung. Cancer Res. 62, 3659–3662 (2002). INK4 cancer-associated genes. Nature 499, 214–218 (2013). 8. Shapiro, G. I. et al. Reciprocal Rb inactivation and p16 expression in primary lung cancers and cell lines. Cancer Res. 55, 505–509 (1995). Supplementary Information is available in the online version of the paper. 9. Singh, A. et al. Dysfunctional KEAP1–NRF2 interaction in non-small-cell lung cancer. PLoS Med. 3, e420 (2006). Acknowledgements This study was supported by NIH grants: U24 CA126561, 10. Medina, P. P. et al. Frequent BRG1/SMARCA4-inactivating mutations in human U24 CA126551, U24 CA126554, U24 CA126543, U24 CA126546, U24 lung cancer cell lines. Hum. Mutat. 29, 617–622 (2008). CA137153, U24 CA126563, U24 CA126544, U24 CA143845, U24 CA143858, U24 11. Ding, L. et al. Somatic mutations affect key pathways in lung adenocarcinoma. CA144025, U24 CA143882, U24 CA143866, U24 CA143867, U24 CA143848, Nature 455, 1069–1075 (2008). U24 CA143840, U24 CA143835, U24 CA143799, U24 CA143883, U24 CA143843, 12. Imielinski, M. et al. Mapping the hallmarks of lung adenocarcinoma with massively U54 HG003067, U54 HG003079 and U54 HG003273. We thank K. Guebert and parallel sequencing. Cell 150, 1107–1120 (2012). L. Gaffney for assistance and C. Gunter for review. 5 4 8 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved ARTICLE RESEARCH 11 11 11 11 11 Author Contributions The Cancer Genome Atlas Research Network contributed Prins ,J.S.Marron , Joel S. Parker , D. Neil Hayes , Charles M. Perou ; University collectively to this study. Biospecimens were provided by the tissue source sites and of Kentucky Jinze Liu ; The USC/JHU Epigenome Characterization Center Leslie 6 6 17 17 processed by the biospecimen core resource. Data generation and analyses were Cope , Ludmila Danilova , Daniel J. Weisenberger ,Dennis T.Maglinte , Philip H. 17 17 17 17 performed by the genome sequencing centres, cancer genome characterization Lai , Moiz S. Bootwalla , David J. Van Den Berg , Timothy Triche Jr , Stephen B. 6 17 centres and genome data analysis centres. All data were released through the data Baylin , Peter W. Laird coordinating centre. The National Cancer Institute and National Human Genome Research Institute project teams coordinated project activities. We also acknowledge 2 Genome data analysis centres: The Eli & Edythe L. Broad Institute Mara Rosenberg , the following TCGA investigators who made substantial contributions to the project: 12 12 2 2 2 2 Lynda Chin , Jianhua Zhang ,Juok Cho ,Daniel DiCara , David Heiman , Pei Lin , E. A. Collisson (manuscript coordinator); J. D. Campbell, J. Chmielecki, (analysis 2 2 2 2 2 William Mallard , Douglas Voet , Hailei Zhang , Lihua Zou , Michael S. Noble , coordinators); C. Sougnez (data coordinator); J. D. Campbell, M. Rosenberg, W. Lee, 2 2 2 2 Michael S. Lawrence , Gordon Saksena ,Nils Gehlenborg , Helga Thorvaldsdottir , J. Chmielecki, M. Ladanyi, and G. Getz (DNA sequence analysis); M. D. Wilkerson, 2 2 2 2,9,10 Jill Mesirov , Marc-Danie Nazaire , Jim Robinson , Gad Getz ; Memorial A. N. Brooks, and D. N. Hayes (mRNA sequence analysis); L. Danilova and L. Cope (DNA 4 4 4 Sloan-Kettering Cancer Center William Lee , B. Arman Aksoy , Giovanni Ciriello , methylation analysis); A. D. Cherniack (copy number analysis); M. D. Wilkerson and 1 4 4 4 Barry S. Taylor , Gideon Dresdner , Jianjiong Gao ,Benjamin Gross , Venkatraman E. A. Hadjipanayis (translocations); N. Schultz, W. Lee, E. A. Collisson, A. H. Berger, 4 4 4 4 4 4 Seshan ,MarcLadanyi , Boris Reva ,Rileen Sinha ,S.OnurSumer , Nils Weinhold , J. Chmielecki, C. J. Creighton, L. A. Byers and M. Ladanyi (pathway analysis); A. Chu and 4 4 4 Nikolaus Schultz , Ronglai Shen , Chris Sander ; University of California, Santa Cruz/ A. G. Robertson (miRNA sequence analysis); W. Travis and D. A. Wigle (pathology and 18 18 18 18 Buck Institute Sam Ng ,Singer Ma , Jingchun Zhu , Amie Radenbaugh ,Joshua clinical expertise); L. A. Byers and G. B. Mills (reverse phase protein arrays); S. B. Baylin, 18 21 21 18,22 M. Stuart , Christopher C. Benz , Christina Yau & David Haussler ; Oregon R. Govindan and M. Meyerson (project chairs). Health & Sciences University Paul T. Spellman ; University of North Carolina, 11 11 11 Author Information The primary and processed data used to generate the analyses Chapel Hill Matthew D. Wilkerson , Joel S. Parker , Katherine A. Hoadley , Patrick K. 11 11 11 presented here can be downloaded by registered users from The Cancer Genome Atlas Kimes , D. Neil Hayes , Charles M. Perou ; The University of Texas MD Anderson 12 12 12 12 at (https://tcga-data.nci.nih.gov/tcga/tcgaDownload.jsp). All of the primary sequence Cancer Center Bradley M. Broom ,JingWang ,YilingLu , Patrick Kwok Shing Ng , 12 12 12 12 files are deposited in cgHub and all other data are deposited at the Data Coordinating Lixia Diao , Lauren Averett Byers , Wenbin Liu , John V. Heymach , 12 12 12 12 Center (DCC) for public access (http://cancergenome.nih.gov/), (https:// Christopher I. Amos , John N. Weinstein , Rehan Akbani , Gordon B. Mills cghub.ucsc.edu/) and (https://tcga-data.nci.nih.gov/docs/publications/luad_2014/). Reprints and permissions information is available at www.nature.com/reprints. The Biospecimen core resource: International Genomics Consortium Erin Curley , authors declare no competing financial interests. Readers are welcome to comment on 24 24 24 24 24 Joseph Paulauskis , Kevin Lau , Scott Morris ,Troy Shelton , David Mallery , the online version of the paper. Correspondence and requests for materials should be 24 24 Johanna Gardner ,Robert Penny addressed to M.M. ([email protected]). Tissue source sites: Analytical Biological Service, Inc. Charles Saller , Katherine This work is licensed under a Creative Commons Attribution- 25 14 Tarvin ; Brigham & Women’s Hospital William G. Richards ; University of Alabama NonCommercial-ShareAlike 3.0 Unported licence. The images or other 26 26 at Birmingham Robert Cerfolio , Ayesha Bryant ; Cleveland Clinic: third party material in this article are included in the article’s Creative Commons licence, 27 27 27 Daniel P. Raymond ,Nathan A.Pennell , Carol Farver ; Christiana Care unless indicated otherwise in the credit line; if the material is not included under the 28 28 28 28 Christine Czerwinski , Lori Huelsenbeck-Dill , Mary Iacocca , Nicholas Petrelli , Creative Commons licence, users will need to obtain permission from the licence holder 28 28 28 29 Brenda Rabeno , Jennifer Brown , Thomas Bauer ; Cureline Oleg Dolzhanskiy , to reproduce the material. To view a copy of this licence, visit http://creativecommons. 29 29 29 29 Olga Potapova ,DaniilRotin ,OlgaVoronina , Elena Nemirovich-Danchenko , org/licenses/by-nc-sa/3.0 29 30 30 Konstantin V. Fedosenko ; Emory University Anthony Gal , Madhusmita Behera , 30 30 31 Suresh S. Ramalingam , Gabriel Sica ; Fox Chase Cancer Center Douglas Flieder , 31 31 32 32 Jeff Boyd , JoEllen Weaver ; ILSbio Bernard Kohl , Dang Huy Quoc Thinh ; 33 34 Indiana University George Sandusky ; Indivumed Hartmut Juhl ; John Flynn 35,36 6 The Cancer Genome Atlas Research Network Hospital Edwina Duhig ; Johns Hopkins University Peter Illei , Edward 6 6 6 6 6 Gabrielson , James Shin , Beverly Lee , Kristen Rodgers , Dante Trusty ,Malcolm V. 1 2 6 37 37 Disease analysis working group Eric A. Collisson , Joshua D. Campbell , Angela N. Brock ; Lahey Hospital & Medical Center Christina Williamson , Eric Burks , 37 37 37 2,3 2 4 2 5 Kimberly Rieger-Christ , Antonia Holway , Travis Sullivan ; Mayo Clinic Dennis A. Brooks , Alice H. Berger , William Lee , Juliann Chmielecki , David G. Beer ,Leslie 16 16 16 6 7 6 8 2,9,10 Wigle , Michael K. Asiedu , Farhad Kosari ; Memorial Sloan-Kettering Cancer Cope , Chad J. Creighton , Ludmila Danilova ,Li Ding , Gad Getz ,Peter S. 4 4 4 4 2 11 2 6 Center William D. Travis , Natasha Rekhtman , Maureen Zakowski , Valerie W. Rusch ; Hammerman , D. Neil Hayes , Bryan Hernandez , James G. Herman ,John V. 38 38 38 12 13 9 14 4 NYU Langone Medical Center Paul Zippile , James Suh , Harvey Pass , Chandra Heymach ,Igor Jurisica ,Raju Kucherlapati , David Kwiatkowski , Marc Ladanyi , 38 38 39 15 4 4 12 Goparaju , Yvonne Owusu-Sarpong ; Ontario Tumour Bank John M. S. Bartlett , Gordon Robertson , Nikolaus Schultz , Ronglai Shen , Rileen Sinha , 2 13 4 12 39 39 39 39 Sugy Kodeeswaran ,Jeremy Parfitt , Harmanjatinder Sekhon , Monique Albert ; Carrie Sougnez , Ming-Sound Tsao , William D. Travis , John N. Weinstein , 16 11 15 2 40 40 Penrose St. Francis Health Services John Eckman ,JeromeB.Myers ; Roswell Park Dennis A. Wigle , Matthew D. Wilkerson ,Andy Chu , Andrew D. Cherniack , 9 2 17 17 41 41 41 Angela Hadjipanayis , Mara Rosenberg , Daniel J. Weisenberger , Peter W. Laird , Cancer Institute Richard Cheney , Carl Morrison , Carmelo Gaudioso ; Rush 18 18 18 12 42 42 42 Amie Radenbaugh ,Singer Ma , Joshua M. Stuart , Lauren Averett Byers , University Medical Center Jeffrey A. Borgia , Philip Bonomi ,MarkPool ,MichaelJ. 6 8 2,3 42 43 43 Stephen B. Baylin , Ramaswamy Govindan , Matthew Meyerson Liptay ; St. Petersburg Academic University Fedor Moiseenko , Irina Zaytseva ; Thoraxklinik am Universitatsklinikum Heidelberg, Member of Biomaterial Bank 2 Heidelberg (BMBH) & Biobank Platform of the German Centre for Lung Research Genome sequencing centres: The Eli & Edythe L. Broad Institute Mara Rosenberg , 44 44 45 2 2 2 2 2 (DZL) Hendrik Dienemann , Michael Meister , Philipp A. Schnabel , Thomas R. Stacey B. Gabriel , Kristian Cibulskis ,Carrie Sougnez , Jaegil Kim ,Chip Stewart , 44 46 2 2,19 2 2,9,10 Muley ; University of Cologne Martin Peifer ; University of Miami Carmen Lee Lichtenstein ,Eric S.Lander , Michael S. Lawrence ,Getz ; Washington 47 47 47 8 8 8 Gomez-Fernandez , Lynn Herbert , Sophie Egea ; University of North Carolina University in St. Louis Cyriac Kandoth , Robert Fulton , Lucinda L. Fulton ,Michael D. 11 11 11 11 8 8 8 8 8 Mei Huang , Leigh B. Thorne ,Lori Boice , Ashley Hill Salazar , William K. McLellan , Richard K. Wilson , Kai Ye , Catrina C. Fronick , Christopher A. Maher , 11 11 48 8 8 8 8 Funkhouser , W. Kimryn Rathmell ; University of Pittsburgh Rajiv Dhir ,SamuelA. Christopher A. Miller , Michael C. Wendl , Christopher Cabanski ,LiDing ,Elaine 48 48 48 48 Yousem , Sanja Dacic , Frank Schneider , Jill M. Siegfried ; The University of 8 8 7 Mardis , Ramaswamy Govindan ; Baylor College of Medicine Chad J. Creighton , Texas MD Anderson Cancer Center Richard Hajek ; Washington University School of David Wheeler 8 8 8 Medicine Mark A. Watson , Sandra McDonald , Bryan Meyers ; Queensland Thoracic 35 35 35 35 Research Center Belinda Clarke , Ian A. Yang , Kwun M. Fong , Lindy Hunter , 35 35 Genome characterization centres: Canada’s Michael Smith Genome Sciences Morgan Windsor , Rayleen V. Bowman ; Center Hospitalier Universitaire Vaudois 49 49 50 Centre, British Columbia Cancer Agency Miruna Balasundaram , Yaron S. N. Solange Peters ,IgorLetovanec ; Ziauddin University Hospital Khurram Z. Khan 15 15 15 15 15 Butterfield , Rebecca Carlsen ,Andy Chu , Eric Chuah , Noreen Dhalla ,Ranabir 15 15 15 15 15 Guin , Carrie Hirst , Darlene Lee ,Haiyan I. Li , Michael Mayo , Richard A. 51 51 51 15 15 15 15 Data Coordination Centre Mark A. Jensen ,Eric E. Snyder , Deepak Srinivasan , Moore , Andrew J. Mungall , Jacqueline E. Schein , Payal Sipahimalani , Angela 51 51 51 15 15 15 15 15 Ari B. Kahn ,JulienBaboud , David A. Pot Tam , Richard Varhol , A. Gordon Robertson , Natasja Wye , Nina Thiessen , 12 15 15 Robert A. Holt , Steven J. M. Jones , Marco A. Marra ; The Eli & Edythe L. Broad 2 2,3 2 52 52 Institute Joshua D. Campbell , Angela N. Brooks , Juliann Chmielecki , Project team: National Cancer Institute Kenna R. Mills Shaw , Margi Sheth ,Tanja 2,9,10 2 9 2 2 52 52 52 52 52 Marcin Imielinski , Robert C. Onofrio ,EranHodis ,Travis Zack , Carrie Sougnez , Davidsen ,John A.Demchok , Liming Yang , Zhining Wang , Roy Tarnuzzer , 2 2 2 2 Elena Helman , Chandra Sekhar Pedamallu , Jill Mesirov , Andrew D. Cherniack , Jean Claude Zenklusen ; National Human Genome Research Institute Bradley A. 2 2 2 2 53 53 Gordon Saksena , Steven E. Schumacher , Scott L. Carter , Bryan Hernandez ,Levi Ozenberger , Heidi J. Sofia 2,3,9 2,3,9 2 2,9,10 Garraway , Rameen Beroukhim , Stacey B. Gabriel , Gad Getz , Matthew 2,3,9 Meyerson ; Harvard Medical School/Brigham & Women’s Hospital/MD Anderson 4 41 35 Expert pathology panel William D. Travis , Richard Cheney , Belinda Clarke , 9,14 9,14 12 Cancer Center Angela Hadjipanayis , Semin Lee , Harshad S. Mahadeshwar , 48 36,35 11 6 27 Sanja Dacic , Edwina Duhig , William K. Funkhouser , Peter Illei , Carol Farver , 9,14 12 9 12 12 Angeliki Pantazi , Alexei Protopopov , Xiaojia Ren , Sahil Seth , Xingzhi Song , 4 30 38 13 Natasha Rekhtman , Gabriel Sica ,James Suh & Ming-Sound Tsao 12 9 12 9 9,14 Jiabin Tang ,LixingYang , Jianhua Zhang ,Peng-ChiehChen , Michael Parfenov , 9,14 9,14 12 9,14 Andrew Wei Xu , Netty Santoso , Lynda Chin , Peter J. Park &Raju 9,14 11 1 2 Kucherlapati ; University of North Carolina, Chapel Hill Katherine A. Hoadley , University of California San Francisco, San Francisco, California 94158, USA. The Eli and 11 11 11 11 11 J. Todd Auman ,ShaowuMeng , Yan Shi , Elizabeth Buda , Scot Waring , Edythe L. Broad Institute, Cambridge, Massachusetts 02142, USA. Dana Farber Cancer 11 11 11 11 4 Umadevi Veluvolu , Donghui Tan , Piotr A. Mieczkowski , Corbin D. Jones , Janae Institute, Boston, Massachusetts 02115, USA. Memorial Sloan-Kettering Cancer Center, 11 11 11 11 5 V. Simons , Matthew G. Soloway , Tom Bodenheimer , Stuart R. Jefferys ,Jeffrey New York, New York 10065, USA. University of Michigan, Ann Arbor, Michigan 48109, 11 11 11 11 11 6 7 Roach ,Alan P. Hoyle , Junyuan Wu , Saianand Balu , Darshan Singh ,Jan F. USA. Johns Hopkins University, Baltimore, Maryland 21287, USA. Baylor College of 31 JULY 20 1 4 | V OL 51 1 | N A T U RE | 5 49 ©2014 Macmillan Publishers Limited. All rights reserved RESEARCH ARTICLE 8 33 Medicine,Houston, Texas 77030,USA. Washington University, St. Louis, Missouri 63108, 21620, USA. Indiana University School of Medicine, Indianapolis, Indiana 46202, USA. 9 10 34 35 USA. Harvard Medical School, Boston, Massachusetts 02115, USA. Massachusetts Individumed, Silver Spring, Maryland 20910, USA. The Prince Charles Hospital and General Hospital, Boston, Massachusetts 02114, USA. University of North Carolina at the University of Queensland Thoracic Research Center, Brisbane, 4032, Australia. 12 36 37 Chapel Hill, Chapel Hill, North Carolina 27599, USA. University of Texas MD Anderson Sullivan Nicolaides Pathology & John Flynn Hospital, Tugun 4680, Australia. Lahey 13 38 Cancer Center, Houston, Texas 77054, USA. Princess Margaret Cancer Centre, Toronto, Hospital and Medical Center, Burlington, Massachusetts 01805, USA. NYU Langone 14 39 Ontario M5G 2M9, Canada. Brigham and Women’s Hospital Boston, Massachusetts Medical Center, New York, New York 10016, USA. Ontario Tumour Bank, Ontario 15 16 40 02115,USA. BC Cancer Agency, Vancouver, British Columbia V5Z 4S6, Canada. Mayo Institute for Cancer Research, Toronto, Ontario M5G 0A3, Canada. Penrose St. Francis 17 41 Clinic, Rochester, Minnesota 55905, USA. University of Southern California, Los Health Services, Colorado Springs, Colorado 80907, USA. Roswell Park Cancer 18 42 Angeles, California 90033, USA. University of California Santa Cruz, Santa Cruz, Center, Buffalo, New York 14263, USA. Rush University Medical Center, Chicago, Illinois 19 43 California 95064, USA. Massachusetts Institute of Technology, Cambridge, 60612, USA. St. Petersburg Academic University, St Petersburg 199034, Russia. 20 44 Massachusetts 02142, USA. University of Kentucky, Lexington, Kentucky 40515, USA. Thoraxklinik am Universita¨tsklinikum Heidelberg, 69126 Heidelberg, Germany. 21 22 45 46 Buck Institute for Age Research, Novato, California 94945, USA. Howard Hughes University Heidelberg, 69120 Heidelberg, Germany. University of Cologne, 50931 Medical Institute, University of California Santa Cruz, Santa Cruz, California 95064, USA. Cologne, Germany. University of Miami, Sylvester Comprehensive Cancer Center, 23 24 48 Oregon Health and Science University, Portland, Oregon 97239, USA. International Miami, Florida 33136, USA. University of Pittsburgh, Pittsburgh, Pennsylvania 15213, 25 49 Genomics Consortium, Phoenix, Arizona 85004, USA. Analytical Biological Services, USA. Center Hospitalier Universitaire Vaudois, Lausanne and European Thoracic 26 50 University of Alabama at Birmingham, Ziauddin University Hospital, Inc., Wilmington, Delaware 19801, USA. Oncology Platform, CH-1011 Lausanne, Switzerland. 27 51 Birmingham, Alabama 35294, USA. Cleveland Clinic, Cleveland, Ohio 44195, USA. Karachi, 75300, Pakistan. SRA International, Inc., Fairfax, Virginia 22033, USA. 28 29 52 Christiana Care, Newark, Delaware 19713, USA. Cureline, Inc., South San Francisco, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, 30 31 53 California 94080, USA. Emory University, Atlanta, Georgia 30322, USA. Fox Chase USA. National Human Genome Research Institute, National Institutes of Health, Cancer Center, Philadelphia, Philadelphia 19111, USA. ILSbio, Chestertown, Maryland Bethesda, Maryland 20892, USA. 5 5 0 | N ATU R E | V OL 51 1 | 31 J U LY 2 0 14 ©2014 Macmillan Publishers Limited. All rights reserved CORRECTIONS & AMENDMENTS CORRIGENDUM doi:10.1038/nature13879 Corrigendum: Comprehensive molecular profiling of lung adenocarcinoma The Cancer Genome Atlas Research Network Nature 511, 543–550 (2014); doi:10.1038/nature13385 In this Article, the surname of author Kristen Rodgers was incorrectly spelled Rogers. This error has been corrected in the HTML and PDF of the original paper. 262|NATURE | VOL514| 9OCTOBER 2014 ©2014 Macmillan Publishers Limited. All rights reserved CORRECTIONS & AMENDMENTS CorreCtion https://doi.org/10.1038/s41586-018-0228-6 Author Correction: Comprehensive molecular profiling of lung adenocarcinoma The Cancer Genome Atlas Research Network Correction to: Nature https://doi.org/10.1038/nature13385, published online 9 July 2014; corrected online 8 October 2014. In this Article, the Supplementary Table 7 iCLUSTER output column included incorrect cluster labels for the integrated subtypes presented in Fig. 5c. These changes affect only the iCLUSTER output column and do not affect the analysis or the conclusions of the work. The authors apologise for the error. Supplementary Table 7 has been corrected online, and the original incorrect table is provided as Supplementary Information to this Amendment for transparency. Supplementary Information is available in the online version of this Amendment. N A TURE | www.nature.com/nature © 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

Journal

Nature – Springer Journals

Published: Jul 9, 2014

There are no references for this article.