A molecular signature of metastasis in primary solid tumors

Metastasis is the principal event leading to death in individuals with cancer, yet its molecular basis is poorly understood. To explore the molecular differences between human primary tumors and metastases, we compared the gene-expression profiles of adenocarcinoma metastases of multiple tumor types to unmatched primary adenocarcinomas. We found a gene-expression signature that distinguished primary from metastatic adenocarcinomas. More notably, we found that a subset of primary tumors resembled metastatic tumors with respect to this gene-expression signature. We confirmed this finding by applying the expression signature to data on 279 primary solid tumors of diverse types. We found that solid tumors carrying the gene-expression signature were most likely to be associated with metastasis and poor clinical outcome (P < 0.03). These results suggest that the metastatic potential of human tumors is encoded in the bulk of a primary tumor, thus challenging the notion that metastases arise from rare cells within a primary tumor that have the ability to metastasize.     

An oncogenic KRAS2 expression signature identified by cross-species gene-expression analysis

Using advanced gene targeting methods, generating mouse models of cancer that accurately reproduce the genetic alterations present in human tumors is now relatively straightforward. The challenge is to determine to what extent such models faithfully mimic human disease with respect to the underlying molecular mechanisms that accompany tumor progression. Here we describe a method for comparing mouse models of cancer with human tumors using gene-expression profiling. We applied this method to the analysis of a model of Kras2-mediated lung cancer and found a good relationship to human lung adenocarcinoma, thereby validating the model. Furthermore, we found that whereas a gene-expression signature of KRAS2 activation was not identifiable when analyzing human tumors with known KRAS2 mutation status alone, integrating mouse and human data uncovered a gene-expression signature of KRAS2 mutation in human lung cancer. We confirmed the importance of this signature by gene-expression analysis of short hairpin RNA-mediated inhibition of oncogenic Kras2. These experiments identified both a pattern of gene expression indicative of KRAS2 mutation and potential effectors of oncogenic KRAS2 activity in human cancer. This approach provides a strategy for using genomic analysis of animal models to probe human disease.     

Chromosomal instability in ulcerative colitis is related to telomere shortening

Ulcerative colitis, a chronic inflammatory disease of the colon, is associated with a high risk of colorectal carcinoma that is thought to develop through genomic instability. We considered that the rapid cell turnover and oxidative injury observed in ulcerative colitis might accelerate telomere shortening, thereby increasing the potential of chromosomal ends to fuse, resulting in cycles of chromatin bridge breakage and fusion and chromosomal instability associated with tumor cell progression. Here we have used quantitative fluorescence in situ hybridization to compare chromosomal aberrations and telomere shortening in non-dysplastic mucosa taken from individuals affected by ulcerative colitis, either with (UC progressors) or without (UC non-progressors) dysplasia or cancer. Losses, but not gains, of chromosomal arms and centromeres are highly correlated with telomere shortening. Chromosomal losses are greater and telomeres are shorter in biopsy samples from UC progressors than in those from UC non-progressors or control individuals without ulcerative colitis. A mechanistic link between telomere shortening and chromosomal instability is supported by a higher frequency of anaphase bridges--an intermediate in the breakage and fusion of chromatin bridges--in UC progressors than in UC non-progressors or control individuals. Our study shows that telomere length is correlated with chromosomal instability in a precursor of human cancer.     

Genome-wide survey of recurrent HBV integration in hepatocellular carcinoma

To survey hepatitis B virus (HBV) integration in liver cancer genomes, we conducted massively parallel sequencing of 81 HBV-positive and 7 HBV-negative hepatocellular carcinomas (HCCs) and adjacent normal tissues. We found that HBV integration is observed more frequently in the tumors (86.4%) than in adjacent liver tissues (30.7%). Copy-number variations (CNVs) were significantly increased at HBV breakpoint locations where chromosomal instability was likely induced. Approximately 40% of HBV breakpoints within the HBV genome were located within a 1,800-bp region where the viral enhancer, X gene and core gene are located. We also identified recurrent HBV integration events (in ≥ 4 HCCs) that were validated by RNA sequencing (RNA-seq) and Sanger sequencing at the known and putative cancer-related TERT, MLL4 and CCNE1 genes, which showed upregulated gene expression in tumor versus normal tissue. We also report evidence that suggests that the number of HBV integrations is associated with patient survival.     

High-throughput oncogene mutation profiling in human cancer

Systematic efforts are underway to decipher the genetic changes associated with tumor initiation and progression. However, widespread clinical application of this information is hampered by an inability to identify critical genetic events across the spectrum of human tumors with adequate sensitivity and scalability. Here, we have adapted high-throughput genotyping to query 238 known oncogene mutations across 1,000 human tumor samples. This approach established robust mutation distributions spanning 17 cancer types. Of 17 oncogenes analyzed, we found 14 to be mutated at least once, and 298 (30%) samples carried at least one mutation. Moreover, we identified previously unrecognized oncogene mutations in several tumor types and observed an unexpectedly high number of co-occurring mutations. These results offer a new dimension in tumor genetics, where mutations involving multiple cancer genes may be interrogated simultaneously and in 'real time' to guide cancer classification and rational therapeutic intervention.     

Chfr is required for tumor suppression and Aurora A regulation

Tumorigenesis is a consequence of loss of tumor suppressors and activation of oncogenes. Expression of the mitotic checkpoint protein Chfr is lost in 20-50% of primary tumors and tumor cell lines. To explore whether downregulation of Chfr contributes directly to tumorigenesis, we generated Chfr knockout mice. Chfr-deficient mice are cancer-prone, develop spontaneous tumors and have increased skin tumor incidence after treatment with dimethylbenz(a)anthracene. Chfr deficiency leads to chromosomal instability in embryonic fibroblasts and regulates the mitotic kinase Aurora A, which is frequently upregulated in a variety of tumors. Chfr physically interacts with Aurora A and ubiquitinates Aurora A both in vitro and in vivo. Collectively, our data suggest that Chfr is a tumor suppressor and ensures chromosomal stability by controlling the expression levels of key mitotic proteins such as Aurora A.     

Genome-wide retroviral insertional tagging of genes involved in cancer in Cdkn2a-deficient mice

We have used large-scale insertional mutagenesis to identify functional landmarks relevant to cancer in the recently completed mouse genome sequence. We infected Cdkn2a(-/-) mice with Moloney murine leukemia virus (MoMuLV) to screen for loci that can participate in tumorigenesis in collaboration with loss of the Cdkn2a-encoded tumor suppressors p16INK4a and p19ARF. Insertional mutagenesis by the latent retrovirus was synergistic with loss of Cdkn2a expression, as indicated by a marked acceleration in the development of both myeloid and lymphoid tumors. We isolated 747 unique sequences flanking retroviral integration sites and mapped them against the mouse genome sequence databases from Celera and Ensembl. In addition to 17 insertions targeting gene loci known to be cancer-related, we identified a total of 37 new common insertion sites (CISs), of which 8 encode components of signaling pathways that are involved in cancer. The effectiveness of large-scale insertional mutagenesis in a sensitized genetic background is demonstrated by the preference for activation of MAP kinase signaling, collaborating with Cdkn2a loss in generating the lymphoid and myeloid tumors. Collectively, our results show that large-scale retroviral insertional mutagenesis in genetically predisposed mice is useful both as a system for identifying genes underlying cancer and as a genetic framework for the assignment of such genes to specific oncogenic pathways.     

A module map showing conditional activity of expression modules in cancer

DNA microarrays are widely used to study changes in gene expression in tumors, but such studies are typically system-specific and do not address the commonalities and variations between different types of tumor. Here we present an integrated analysis of 1,975 published microarrays spanning 22 tumor types. We describe expression profiles in different tumors in terms of the behavior of modules, sets of genes that act in concert to carry out a specific function. Using a simple unified analysis, we extract modules and characterize gene-expression profiles in tumors as a combination of activated and deactivated modules. Activation of some modules is specific to particular types of tumor; for example, a growth-inhibitory module is specifically repressed in acute lymphoblastic leukemias and may underlie the deregulated proliferation in these cancers. Other modules are shared across a diverse set of clinical conditions, suggestive of common tumor progression mechanisms. For example, the bone osteoblastic module spans a variety of tumor types and includes both secreted growth factors and their receptors. Our findings suggest that there is a single mechanism for both primary tumor proliferation and metastasis to bone. Our analysis presents multiple research directions for diagnostic, prognostic and therapeutic studies.     

XPF nuclease-dependent telomere loss and increased DNA damage in mice overexpressing TRF2 result in premature aging and cancer

TRF2 is a telomere-binding protein that has a role in telomere protection. We generated mice that overexpress TRF2 in the skin. These mice had a severe phenotype in the skin in response to light, consisting of premature skin deterioration, hyperpigmentation and increased skin cancer, which resembles the human syndrome xeroderma pigmentosum. Keratinocytes from these mice were hypersensitive to ultraviolet irradiation and DNA crosslinking agents. The skin cells of these mice had marked telomere shortening, loss of the telomeric G-strand overhang and increased chromosomal instability. Telomere loss in these mice was mediated by XPF, a structure-specific nuclease involved in ultraviolet-induced damage repair and mutated in individuals with xeroderma pigmentosum. These findings suggest that TRF2 provides a crucial link between telomere function and ultraviolet-induced damage repair, whose alteration underlies genomic instability, cancer and aging. Finally, we show that a number of human skin tumors have increased expression of TRF2, further highlighting a role for TRF2 in skin cancer.     

TSLC1 is a tumor-suppressor gene in human non-small-cell lung cancer

The existence of tumor-suppressor genes was originally demonstrated by functional complementation through whole-cell and microcell fusion. Transfer of chromosome 11 into a human non-small-cell lung cancer (NSCLC) cell line, A549, suppresses tumorigenicity. Loss of heterozygosity (LOH) on the long arm of chromosome 11 has been reported in NSCLC and other cancers. Several independent studies indicate that multiple tumor-suppressor genes are found in this region, including the gene PPP2R1B at 11q23-24 (ref. 7). Linkage studies of NSCLC are precluded because no hereditary forms are known. We previously identified a region of 700 kb on 11q23.2 that completely suppresses tumorigenicity of A549 human NSCLC cells. Most of this tumor-suppressor activity localizes to a 100-kb segment by functional complementation. Here we report that this region contains a single confirmed gene, TSLC1, whose expression is reduced or absent in A549 and several other NSCLC, hepatocellular carcinoma (HCC) and pancreatic cancer (PaC) cell lines. TSLC1 expression or suppression is correlated with promoter methylation state in these cell lines. Restoration of TSLC1 expression to normal or higher levels suppresses tumor formation by A549 cells in nude mice. Only 2 inactivating mutations of TSLC1 were discovered in 161 tumors and tumor cell lines, both among the 20 primary tumors with LOH for 11q23.2. Promoter methylation was observed in 15 of the other 18 primary NSCLC, HCC and PaC tumors with LOH for 11q23.2. Thus, attenuation of TSLC1 expression occurred in 85% of primary tumors with LOH. Hypermethylation of the TSLC1 promoter would seem to represent the 'second hit' in NSCLC with LOH.     

Exome sequencing of gastric adenocarcinoma identifies recurrent somatic mutations in cell adhesion and chromatin remodeling genes

Gastric cancer is a major cause of global cancer mortality. We surveyed the spectrum of somatic alterations in gastric cancer by sequencing the exomes of 15 gastric adenocarcinomas and their matched normal DNAs. Frequently mutated genes in the adenocarcinomas included TP53 (11/15 tumors), PIK3CA (3/15) and ARID1A (3/15). Cell adhesion was the most enriched biological pathway among the frequently mutated genes. A prevalence screening confirmed mutations in FAT4, a cadherin family gene, in 5% of gastric cancers (6/110) and FAT4 genomic deletions in 4% (3/83) of gastric tumors. Frequent mutations in chromatin remodeling genes (ARID1A, MLL3 and MLL) also occurred in 47% of the gastric cancers. We detected ARID1A mutations in 8% of tumors (9/110), which were associated with concurrent PIK3CA mutations and microsatellite instability. In functional assays, we observed both FAT4 and ARID1A to exert tumor-suppressor activity. Somatic inactivation of FAT4 and ARID1A may thus be key tumorigenic events in a subset of gastric cancers.     

Network modeling links breast cancer susceptibility and centrosome dysfunction

Many cancer-associated genes remain to be identified to clarify the underlying molecular mechanisms of cancer susceptibility and progression. Better understanding is also required of how mutations in cancer genes affect their products in the context of complex cellular networks. Here we have used a network modeling strategy to identify genes potentially associated with higher risk of breast cancer. Starting with four known genes encoding tumor suppressors of breast cancer, we combined gene expression profiling with functional genomic and proteomic (or 'omic') data from various species to generate a network containing 118 genes linked by 866 potential functional associations. This network shows higher connectivity than expected by chance, suggesting that its components function in biologically related pathways. One of the components of the network is HMMR, encoding a centrosome subunit, for which we demonstrate previously unknown functional associations with the breast cancer-associated gene BRCA1. Two case-control studies of incident breast cancer indicate that the HMMR locus is associated with higher risk of breast cancer in humans. Our network modeling strategy should be useful for the discovery of additional cancer-associated genes.     

Telomere dysfunction and evolution of intestinal carcinoma in mice and humans

Telomerase activation is a common feature of advanced human cancers and facilitates the malignant transformation of cultured human cells and in mice. These experimental observations are in accord with the presence of robust telomerase activity in more advanced stages of human colorectal carcinogenesis. However, the occurrence of colon carcinomas in telomerase RNA (Terc)-null, p53-mutant mice has revealed complex interactions between telomere dynamics, checkpoint responses and carcinogenesis. We therefore sought to determine whether telomere dysfunction exerts differential effects on cancer initiation versus progression of mouse and human intestinal neoplasia. In successive generations of ApcMin Terc-/- mice, progressive telomere dysfunction led to an increase in initiated lesions (microscopic adenomas), yet a significant decline in the multiplicity and size of macroscopic adenomas. That telomere dysfunction also contributes to human colorectal carcinogenesis is supported by the appearance of anaphase bridges (a correlate of telomere dysfunction) at the adenoma-early carcinoma transition, a transition recognized for marked chromosomal instability. Together, these data are consistent with a model in which telomere dysfunction promotes the chromosomal instability that drives early carcinogenesis, while telomerase activation restores genomic stability to a level permissive for tumor progression. We propose that early and transient telomere dysfunction is a major mechanism underlying chromosomal instability of human cancer.     

Detectable clonal mosaicism and its relationship to aging and cancer

In an analysis of 31,717 cancer cases and 26,136 cancer-free controls from 13 genome-wide association studies, we observed large chromosomal abnormalities in a subset of clones in DNA obtained from blood or buccal samples. We observed mosaic abnormalities, either aneuploidy or copy-neutral loss of heterozygosity, of >2 Mb in size in autosomes of 517 individuals (0.89%), with abnormal cell proportions of between 7% and 95%. In cancer-free individuals, frequency increased with age, from 0.23% under 50 years to 1.91% between 75 and 79 years (P = 4.8 × 10(-8)). Mosaic abnormalities were more frequent in individuals with solid tumors (0.97% versus 0.74% in cancer-free individuals; odds ratio (OR) = 1.25; P = 0.016), with stronger association with cases who had DNA collected before diagnosis or treatment (OR = 1.45; P = 0.0005). Detectable mosaicism was also more common in individuals for whom DNA was collected at least 1 year before diagnosis with leukemia compared to cancer-free individuals (OR = 35.4; P = 3.8 × 10(-11)). These findings underscore the time-dependent nature of somatic events in the etiology of cancer and potentially other late-onset diseases.     

Nonsense-mediated decay microarray analysis identifies mutations of EPHB2 in human prostate cancer

The identification of tumor-suppressor genes in solid tumors by classical cancer genetics methods is difficult and slow. We combined nonsense-mediated RNA decay microarrays and array-based comparative genomic hybridization for the genome-wide identification of genes with biallelic inactivation involving nonsense mutations and loss of the wild-type allele. This approach enabled us to identify previously unknown mutations in the receptor tyrosine kinase gene EPHB2. The DU 145 prostate cancer cell line, originating from a brain metastasis, carries a truncating mutation of EPHB2 and a deletion of the remaining allele. Additional frameshift, splice site, missense and nonsense mutations are present in clinical prostate cancer samples. Transfection of DU 145 cells, which lack functional EphB2, with wild-type EPHB2 suppresses clonogenic growth. Taken together with studies indicating that EphB2 may have an essential role in cell migration and maintenance of normal tissue architecture, our findings suggest that mutational inactivation of EPHB2 may be important in the progression and metastasis of prostate cancer.     

Constitutional aneuploidy and cancer predisposition caused by biallelic mutations in BUB1B

Mosaic variegated aneuploidy is a rare recessive condition characterized by growth retardation, microcephaly, childhood cancer and constitutional mosaicism for chromosomal gains and losses. In five families with mosaic variegated aneuploidy, including two with embryonal rhabdomyosarcoma, we identified truncating and missense mutations of BUB1B, which encodes BUBR1, a key protein in the mitotic spindle checkpoint. These data are the first to relate germline mutations in a spindle checkpoint gene with a human disorder and strongly support a causal link between aneuploidy and cancer development.     

A functional switch from lung cancer resistance to susceptibility at the Pas1 locus in Kras2LA2 mice

Pulmonary adenoma susceptibility 1 (Pas1) is the major mouse lung cancer susceptibility locus on chromosome 6 (ref. 1). Kras2 is a common target of somatic mutation in chemically induced mouse lung tumors and is a candidate Pas1 gene. M. spretus mice (SPRET/Ei) carry a Pas1 resistance haplotype for chemically induced lung tumors. We demonstrate that the SPRET/Ei Pas1 allele is switched from resistance to susceptibility by fixation of the parental origin of the mutant Kras2 allele. This switch correlates with low expression of endogenous Kras2 in SPRET/Ei. We propose that the Pas1 modifier effect is due to Kras2, and that a sensitive balance between the expression levels of wild-type and mutant alleles determines lung tumor susceptibility. These data demonstrate that cancer predisposition should also be considered in the context of somatic events and could have major implications for the design of human association studies to identify cancer susceptibility genes.