Additional materials of the author at
CSHL Institutional Repository

Integrated Computational Pipeline for Single-Cell Genomic Profiling.
Chorbadjiev, L. and Kendall, J. and Alexander, J. and Zhygulin, V. and Song, J. and Wigler, M. and Krasnitz, A. (2020) JCO Clinical Cancer Informatics, 4 pp. 464-471.

Purpose: Copy-number profiling of multiple individual cells from sparse sequencing may be used to reveal a detailed picture of genomic heterogeneity and clonal organization in a tissue biopsy specimen. We sought to provide a comprehensive computational pipeline for single-cell genomics, to facilitate adoption of this molecular technology for basic and translational research. Materials and methods: The pipeline comprises software tools programmed in Python and in R and depends on Bowtie, HISAT2, Matplotlib, and Qt. It is installed and used with Anaconda. Results: Here we describe a complete pipeline for sparse single-cell genomic data, encompassing all steps of single-nucleus DNA copy-number profiling, from raw sequence processing to clonal structure analysis and visualization. For the latter, a specialized graphical user interface termed the single-cell genome viewer (SCGV) is provided. With applications to cancer diagnostics in mind, the SCGV allows for zooming and linkage to the University of California at Santa Cruz Genome Browser from each of the multiple integrated views of single-cell copy-number profiles. The latter can be organized by clonal substructure or by any of the associated metadata such as anatomic location and histologic characterization. Conclusion: The pipeline is available as open-source software for Linux and OS X. Its modular structure, extensive documentation, and ease of deployment using Anaconda facilitate its adoption by researchers and practitioners of single-cell genomics. With open-source availability and Massachusetts Institute of Technology licensing, it provides a basis for additional development by the cancer bioinformatics community.

Novel Insights Into Breast Cancer Copy Number Genetic Heterogeneity Revealed by Single-Cell Genome Sequencing.
Baslan, T. and Kendall, J. and Volyanskyy, K. and McNamara, K. and Cox, H. and D'Italia, S. and Ambrosio, F. and Riggs, M. and Rodgers, L. and Leotta, A. and Song, J. and Mao, Y. and Wu, J. and Shah, R. and Gularte-Merida, R. and Chadalavada, K. and Nanjangud, G. and Varadan, V. and Gordon, A. and Curtis, C. and Krasnitz, A. and Dimitrova, N. and Harris, L. and Wigler, M. and Hicks, J. (2020) Elife, 9 pp. e51480.

Copy number alterations (CNAs) play an important role in molding the genomes of breast cancers and have been shown to be clinically useful for prognostic and therapeutic purposes. However, our knowledge of intra-tumoral genetic heterogeneity of this important class of somatic alterations is limited. Here, using single-cell sequencing, we comprehensively map out the facets of copy number alteration heterogeneity in a cohort of breast cancer tumors. Ou/var/www/html/elife/12-05-2020/backup/r analyses reveal: genetic heterogeneity of non-tumor cells (i.e. stroma) within the tumor mass; the extent to which copy number heterogeneity impacts breast cancer genomes and the importance of both the genomic location and dosage of sub-clonal events; the pervasive nature of genetic heterogeneity of chromosomal amplifications; and the association of copy number heterogeneity with clinical and biological parameters such as polyploidy and estrogen receptor negative status. Our data highlight the power of single-cell genomics in dissecting, in its many forms, intra-tumoral genetic heterogeneity of CNAs, the magnitude with which CNA heterogeneity affects the genomes of breast cancers, and the potential importance of CNA heterogeneity in phenomena such as therapeutic resistance and disease relapse.

Single-Chromosomal Gains Can Function as Metastasis Suppressors and Promoters in Colon Cancer.
Vasudevan, A. and Baruah, P.S. and Smith, J.C. and Wang, Z. and Sayles, N. M. and Andrews, P. and Kendall, J. and Leu, J. and Chunduri, N.K. and Levy, D. and Wigler, M. and Storchova, Z. and Sheltzer, J. M. (2020) Developmental Cell, 52(4) pp. 413-428.

High levels of cancer aneuploidy are frequently associated with poor prognosis. To examine the relationship between aneuploidy and cancer progression, we analyzed a series of congenic cell lines that harbor single extra chromosomes. We found that across 13 different trisomic cell lines, 12 trisomies suppressed invasiveness or were largely neutral, while a single trisomy increased metastatic behavior by triggering a partial epithelial-mesenchymal transition. In contrast, we discovered that chromosomal instability activates cGAS/STING signaling but strongly suppresses invasiveness. By analyzing patient copy-number data, we demonstrate that specific aneuploidies are associated with distinct outcomes, and the acquisition of certain aneuploidies is in fact linked with a favorable prognosis. Thus, aneuploidy is not a uniform driver of malignancy, and different aneuploidies can uniquely influence tumor progression. At the same time, the gain of a single chromosome is capable of inducing a profound cell state transition, thereby linking genomic plasticity, phenotypic plasticity, and metastasis.

Multiplex Accurate Sensitive Quantitation (MASQ) With Application to Minimal Residual Disease in Acute Myeloid Leukemia.
Moffitt, A. B. and Spector, M. S. and Andrews, P. and Kendall, J. and Alexander, J. and Stepansky, A. and Ma, B. and Kolitz, J. and Chiorazzi, N. and Allen, S.L. and Krasnitz, A. and Wigler, M. and Levy, D. and Wang, Z. (2020) Nucleic Acids Research, 48(7) pp. e40.

Measuring minimal residual disease in cancer has applications for prognosis, monitoring treatment and detection of recurrence. Simple sequence-based methods to detect nucleotide substitution variants have error rates (about 10-3) that limit sensitive detection. We developed and characterized the performance of MASQ (multiplex accurate sensitive quantitation), a method with an error rate below 10-6. MASQ counts variant templates accurately in the presence of millions of host genomes by using tags to identify each template and demanding consensus over multiple reads. Since the MASQ protocol multiplexes 50 target loci, we can both integrate signal from multiple variants and capture subclonal response to treatment. Compared to existing methods for variant detection, MASQ achieves an excellent combination of sensitivity, specificity and yield. We tested MASQ in a pilot study in acute myeloid leukemia (AML) patients who entered complete remission. We detect leukemic variants in the blood and bone marrow samples of all five patients, after induction therapy, at levels ranging from 10-2 to nearly 10-6. We observe evidence of sub-clonal structure and find higher target variant frequencies in patients who go on to relapse, demonstrating the potential for MASQ to quantify residual disease in AML.

Copolymerization of single-cell nucleic acids into balls of acrylamide gel.
Li, S. and Kendall, J. and Park, S. and Wang, Z. and Alexander, J. and Moffitt, A. and Ranade, N. and Danyko, C. and Gegenhuber, B. and Fischer, S. and Robinson, B. D. and Lepor, H. and Tollkuhn, J. and Gillis, J. and Brouzes, E. and Krasnitz, A. and Levy, D. and Wigler, M. (2020) Genome Res, 30(1) pp. 49-61.

We show the use of 5'-Acrydite oligonucleotides to copolymerize single-cell DNA or RNA into balls of acrylamide gel (BAGs). Combining this step with split-and-pool techniques for creating barcodes yields a method with advantages in cost and scalability, depth of coverage, ease of operation, minimal cross-contamination, and efficient use of samples. We perform DNA copy number profiling on mixtures of cell lines, nuclei from frozen prostate tumors, and biopsy washes. As applied to RNA, the method has high capture efficiency of transcripts and sufficient consistency to clearly distinguish the expression patterns of cell lines and individual nuclei from neurons dissected from the mouse brain. By using varietal tags (UMIs) to achieve sequence error correction, we show extremely low levels of cross-contamination by tracking source-specific SNVs. The method is readily modifiable, and we will discuss its adaptability and diverse applications.

Damaging de novo mutations diminish motor skills in children on the autism spectrum.
Buja, A. and Volfovsky, N. and Krieger, A. M. and Lord, C. and Lash, A. E. and Wigler, M. and Iossifov, I. (2018) Proc Natl Acad Sci U S A, 115(8) pp. E1859-E1866.

In individuals with autism spectrum disorder (ASD), de novo mutations have previously been shown to be significantly correlated with lower IQ but not with the core characteristics of ASD: deficits in social communication and interaction and restricted interests and repetitive patterns of behavior. We extend these findings by demonstrating in the Simons Simplex Collection that damaging de novo mutations in ASD individuals are also significantly and convincingly correlated with measures of impaired motor skills. This correlation is not explained by a correlation between IQ and motor skills. We find that IQ and motor skills are distinctly associated with damaging mutations and, in particular, that motor skills are a more sensitive indicator of mutational severity than is IQ, as judged by mutational type and target gene. We use this finding to propose a combined classification of phenotypic severity: mild (little impairment of either), moderate (impairment mainly to motor skills), and severe (impairment of both IQ and motor skills).

Partial bisulfite conversion for unique template sequencing.
Kumar, V. and Rosenbaum, J. and Wang, Z. and Forcier, T. and Ronemus, M. and Wigler, M. and Levy, D. (2018) Nucleic Acids Res, 46(2) pp. e10.

We introduce a new protocol, mutational sequencing or muSeq, which uses sodium bisulfite to randomly deaminate unmethylated cytosines at a fixed and tunable rate. The muSeq protocol marks each initial template molecule with a unique mutation signature that is present in every copy of the template, and in every fragmented copy of a copy. In the sequenced read data, this signature is observed as a unique pattern of C-to-T or G-to-A nucleotide conversions. Clustering reads with the same conversion pattern enables accurate count and long-range assembly of initial template molecules from short-read sequence data. We explore count and low-error sequencing by profiling 135 000 restriction fragments in a PstI representation, demonstrating that muSeq improves copy number inference and significantly reduces sporadic sequencer error. We explore long-range assembly in the context of cDNA, generating contiguous transcript clusters greater than 3,000 bp in length. The muSeq assemblies reveal transcriptional diversity not observable from short-read data alone.

Utility of single cell genomics in diagnostic evaluation of prostate cancer.
Alexander, J. and Kendall, J. and McIndoo, J. and Rodgers, L. and Aboukhalil, R. and Levy, D. and Stepansky, A. and Sun, G. and Chobadjiev, L. and Riggs, M. and Cox, H. and Hakker, I. and Nowak, D. G. and Laze, J. and Llukani, E. and Srivastava, A. and Gruschow, S. and Yadav, S. S. and Robinson, B. D. and Atwal, G. and Trotman, L. C. and Lepor, H. and Hicks, J. B. and Wigler, M. and Krasnitz, A. (2018) Cancer Res, 78(2) pp. 348-358.

A distinction between indolent and aggressive disease is a major challenge in diagnostics of prostate cancer. As genetic heterogeneity and complexity may influence clinical outcome, we have initiated studies on single tumor cell genomics. In this study, we demonstrate that sparse DNA sequencing of single cell nuclei from prostate core biopsies is a rich source of quantitative parameters for evaluating neoplastic growth and aggressiveness. These include the presence of clonal populations, the phylogenetic structure of those populations, the degree of the complexity of copy number changes in those populations, and measures of the proportion of cells with clonal copy number signatures. The parameters all showed good correlation to the measure of prostatic malignancy, the Gleason score, derived from individual prostate biopsy tissue cores. Remarkably, a more accurate histopathological measure of malignancy, the surgical Gleason score, agrees better with these genomic parameters of diagnostic biopsy than it does with the diagnostic Gleason score and related measures of diagnostic histopathology. This is highly relevant since primary treatment decisions are dependent upon the biopsy and not the surgical specimen. Thus, single cell analysis has the potential to augment traditional core histopathology, improving both the objectivity and accuracy of risk assessment and inform treatment decisions.

Measuring shared variants in cohorts of discordant siblings with applications to autism.
Ye, K. and Iossifov, I. and Levy, D. and Yamrom, B. and Buja, A. and Krieger, A. M. and Wigler, M. (2017) Proc Natl Acad Sci U S A, 114(27) pp. 7073-7076.

We develop a method of analysis [affected to discordant sibling pairs (A2DS)] that tests if shared variants contribute to a disorder. Using a standard measure of genetic relation, test individuals are compared with a cohort of discordant sibling pairs (CDS) to derive a comparative similarity score. We ask if a test individual is more similar to an unrelated affected than to the unrelated unaffected sibling from the CDS and then, sum over such individuals and pairs. Statistical significance is judged by randomly permuting the affected status in the CDS. In the analysis of published genotype data from the Simons Simplex Collection (SSC) and the Autism Genetic Resource Exchange (AGRE) cohorts of children with autism spectrum disorder (ASD), we find strong statistical significance that the affected are more similar to the affected than to the unaffected of the CDS (P value approximately 0.00001). Fathers in multiplex families have marginally greater similarity (P value = 0.02) to unrelated affected individuals. These results do not depend on ethnic matching or gender.

Early Detection of Cancer in Blood Using Single-Cell Analysis: A Proposal.
Krasnitz, Alexander and Kendall, Jude and Alexander, Joan and Levy, Dan and Wigler, Michael (2017) Trends in Molecular Medicine, 23(7) pp. 594-603.

Here, we explore the potential of single-cell genomic analysis in blood for early detection of cancer; we consider a method that screens the presence of recurrent patterns of copy number (CN) alterations using sparse single-cell sequencing. We argue for feasibility, based on in silico analysis of existing single-cell data and cancer CN profiles. Sampling procedures from existing diploid single cells can render data for a cell with any given profile. Sampling from multiple published tumor profiles can interrogate cancer clonality via an algorithm that tests the multiplicity of close pairwise similarities among single-cell cancer genomes. The majority of common solid cancers would be detectable in this manner. As any early detection method must be verifiable and actionable, we describe how further analysis of suspect cells can aid in determining risk and anatomic origin. Future affordability rests on currently available procedures for tumor cell enrichment and inexpensive methods for single-cell analysis.

De novo indels within introns contribute to ASD incidence.
Munoz, Adriana and Yamrom, Boris and Lee, Yoon-ha and Andrews, Peter and Marks, Steven and Lin, Kuan-Ting and Wang, Zihua and Krainer, Adrian R. and Darnell, Robert B. and Wigler, Michael and Iossifov, Ivan (2017) bioRxiv, pp. 137471.

Copy number profiling and whole-exome sequencing has allowed us to make remarkable progress in our understanding of the genetics of autism over the past ten years, but there are major aspects of the genetics that are unresolved. Through whole-genome sequencing, additional types of genetic variants can be observed. These variants are abundant and to know which are functional is challenging. We have analyzed whole-genome sequencing data from 510 of the Simons Simplex Collections quad families and focused our attention on intronic variants. Within the introns of 546 high-quality autism target genes, we identified 63 de novo indels in the affected and only 37 in the unaffected siblings. The difference of 26 events is significantly larger than expected (p-val = 0.01) and using reasonable extrapolation shows that de novo intronic indels can contribute to at least 10% of simplex autism. The significance increases if we restrict to the half of the autism targets that are intolerant to damaging variants in the normal human population, which half we expect to be even more enriched for autism genes. For these 273 targets we observe 43 and 20 events in affected and unaffected siblings, respectively (p-value of 0.005). There was no significant signal in the number of de novo intronic indels in any of the control sets of genes analyzed. We see no signal from de novo substitutions in the introns of target genes.

Indel variant analysis of short-read sequencing data with Scalpel.
Fang, H. and Bergmann, E. A. and Arora, K. and Vacic, V. and Zody, M. C. and Iossifov, I. and O'Rawe, J. A. and Wu, Y. and Jimenez Barron, L. T. and Rosenbaum, J. and Ronemus, M. and Lee, Y. H. and Wang, Z. and Dikoglu, E. and Jobanputra, V. and Lyon, G. J. and Wigler, M. and Schatz, M. C. and Narzisi, G. (2016) Nat Protoc, 11(12) pp. 2529-2548.

As the second most common type of variation in the human genome, insertions and deletions (indels) have been linked to many diseases, but the discovery of indels of more than a few bases in size from short-read sequencing data remains challenging. Scalpel ( is an open-source software for reliable indel detection based on the microassembly technique. It has been successfully used to discover mutations in novel candidate genes for autism, and it is extensively used in other large-scale studies of human diseases. This protocol gives an overview of the algorithm and describes how to use Scalpel to perform highly accurate indel calling from whole-genome and whole-exome sequencing data. We provide detailed instructions for an exemplary family-based de novo study, but we also characterize the other two supported modes of operation: single-sample and somatic analysis. Indel normalization, visualization and annotation of the mutations are also illustrated. Using a standard server, indel discovery and characterization in the exonic regions of the example sequencing data can be completed in approximately 5 h after read mapping.

MUMdex: MUM-based structural variation detection.
Andrews, Peter A. and Iossifov, Ivan and Kendall, Jude and Marks, Steven and Muthuswamy, Lakshmi and Wang, Zihua and Levy, Dan and Wigler, Michael (2016) bioRxiv, pp. 78261.

Motivation: Standard genome sequence alignment tools primarily designed to find one alignment per read have difficulty detecting inversion, translocation and large insertion and deletion (indel) events. Moreover, dedicated split read alignment methods that depend only upon the reference genome may misidentify or find too many potential split read alignments because of reference genome anomalies. Methods: We introduce MUMdex, a Maximal Unique Match (MUM)-based genomic analysis software package consisting of a sequence aligner to the reference genome, a storage-indexing format and analysis software. Discordant reference alignments of MUMs are especially suitable for identifying inversion, translocation and large indel differences in unique regions. Extracted population databases are used as filters for flaws in the reference genome. We describe the concepts underlying MUM-based analysis, the software implementation and its usage. Results: We demonstrate via simulation that the MUMdex aligner and alignment format are able to correctly detect and record genomic events. We characterize alignment performance and output file sizes for human whole genome data and compare to Bowtie 2 and the BAM format. Preliminary results demonstrate the practicality of the analysis approach by detecting de novo mutation candidates in human whole genome DNA sequence data from 510 families. We provide a population database of events from these families for use by others.

SMASH, a fragmentation and sequencing method for genomic copy number analysis.
Wang, Z. and Andrews, P. and Kendall, J. and Ma, B. and Hakker, I. and Rodgers, L. and Ronemus, M. and Wigler, M. and Levy, D. (2016) Genome Res, 26(6) pp. 844-51.

Copy number variants (CNVs) underlie a significant amount of genetic diversity and disease. CNVs can be detected by a number of means, including chromosomal microarray analysis (CMA) and whole-genome sequencing (WGS), but these approaches suffer from either limited resolution (CMA) or are highly expensive for routine screening (both CMA and WGS). As an alternative, we have developed a next-generation sequencing-based method for CNV analysis termed SMASH, for short multiply aggregated sequence homologies. SMASH utilizes random fragmentation of input genomic DNA to create chimeric sequence reads, from which multiple mappable tags can be parsed using maximal almost-unique matches (MAMs). The SMASH tags are then binned and segmented, generating a profile of genomic copy number at the desired resolution. Because fewer reads are necessary relative to WGS to give accurate CNV data, SMASH libraries can be highly multiplexed, allowing large numbers of individuals to be analyzed at low cost. Increased genomic resolution can be achieved by sequencing to higher depth.

Corrigendum: Genome-wide copy number analysis of single cells.
Baslan, T. and Kendall, J. and Rodgers, L. and Cox, H. and Riggs, M. and Stepansky, A. and Troge, J. and Ravi, K. and Esposito, D. and Lakshmi, B. and Wigler, M. and Navin, N. and Hicks, J. (2016) Nat Protoc, 11(3) pp. 616.

Interactive analysis and assessment of single-cell copy-number variations.
Garvin, T. and Aboukhalil, R. and Kendall, J. and Baslan, T. and Atwal, G. S. and Hicks, J. and Wigler, M. and Schatz, M. C. (2015) Nat Methods, 12 pp. 1058-1060.

We present Ginkgo (, a user-friendly, open-source web platform for the analysis of single-cell copy-number variations (CNVs). Ginkgo automatically constructs copy-number profiles of cells from mapped reads and constructs phylogenetic trees of related cells. We validated Ginkgo by reproducing the results of five major studies. After comparing three commonly used single-cell amplification techniques, we concluded that degenerate oligonucleotide-primed PCR is the most consistent for CNV analysis.

Low load for disruptive mutations in autism genes and their biased transmission.
Iossifov, I. and Levy, D. and Allen, J. and Ye, K. and Ronemus, M. and Lee, Y. H. and Yamrom, B. and Wigler, M. (2015) Proc Natl Acad Sci U S A,

We previously computed that genes with de novo (DN) likely gene-disruptive (LGD) mutations in children with autism spectrum disorders (ASD) have high vulnerability: disruptive mutations in many of these genes, the vulnerable autism genes, will have a high likelihood of resulting in ASD. Because individuals with ASD have lower fecundity, such mutations in autism genes would be under strong negative selection pressure. An immediate prediction is that these genes will have a lower LGD load than typical genes in the human gene pool. We confirm this hypothesis in an explicit test by measuring the load of disruptive mutations in whole-exome sequence databases from two cohorts. We use information about mutational load to show that lower and higher intelligence quotients (IQ) affected individuals can be distinguished by the mutational load in their respective gene targets, as well as to help prioritize gene targets by their likelihood of being autism genes. Moreover, we demonstrate that transmission of rare disruptions in genes with a lower LGD load occurs more often to affected offspring; we show transmission originates most often from the mother, and transmission of such variants is seen more often in offspring with lower IQ. A surprising proportion of transmission of these rare events comes from genes expressed in the embryonic brain that show sharply reduced expression shortly after birth.

Optimizing sparse sequencing of single cells for highly multiplex copy number profiling.
Baslan, T. and Kendall, J. and Ward, B. and Cox, H. and Leotta, A. and Rodgers, L. and Riggs, M. and D'Italia, S. and Sun, G. and Yong, M. and Miskimen, K. and Gilmore, H. and Saborowski, M. and Dimitrova, N. and Krasnitz, A. and Harris, L. and Wigler, M. and Hicks, J. (2015) Genome Research, 25(5) pp. 714-724.

Genome-wide analysis at the level of single cells has recently emerged as a powerful tool to dissect genome heterogeneity in cancer, neurobiology, and development. To be truly transformative, single-cell approaches must affordably accommodate large numbers of single cells. This is feasible in the case of copy number variation (CNV), because CNV determination requires only sparse sequence coverage. We have used a combination of bioinformatic and molecular approaches to optimize single-cell DNA amplification and library preparation for highly multiplexed sequencing, yielding a method that can produce genome-wide CNV profiles of up to a hundred individual cells on a single lane of an Illumina HiSeq instrument. We apply the method to human cancer cell lines and biopsied cancer tissue, thereby illustrating its efficiency, reproducibility, and power to reveal underlying genetic heterogeneity and clonal phylogeny. The capacity of the method to facilitate the rapid profiling of hundreds to thousands of single-cell genomes represents a key step in making single-cell profiling an easily accessible tool for studying cell lineage.

Quantitative multigene FISH on breast carcinomas identifies der(1;16)(q10;p10) as an early event in luminal A tumors.
Rye, I. H. and Lundin, P. and Maner, S. and Fjelldal, R. and Naume, B. and Wigler, M. and Hicks, J. and Borresen-Dale, A. L. and Zetterberg, A. and Russnes, H. G. (2015) Genes, Chromosomes & Cancer, 54(4) pp. 235-48.

In situ detection of genomic alterations in cancer provides information at the single cell level, making it possible to investigate genomic changes in cells in a tissue context. Such topological information is important when studying intratumor heterogeneity as well as alterations related to different steps in tumor progression. We developed a quantitative multigene fluorescence in situ hybridization (QM FISH) method to detect multiple genomic regions in single cells in complex tissues. As a "proof of principle" we applied the method to breast cancer samples to identify partners in whole arm (WA) translocations. WA gain of chromosome arm 1q and loss of chromosome arm 16q are among the most frequent genomic events in breast cancer. By designing five specific FISH probes based on breakpoint information from comparative genomic hybridization array (aCGH) profiles, we visualized chromosomal translocations in clinical samples at the single cell level. By analyzing aCGH data from 295 patients with breast carcinoma with known molecular subtype, we found concurrent WA gain of 1q and loss of 16q to be more frequent in luminal A tumors compared to other molecular subtypes. QM FISH applied to a subset of samples (n = 26) identified a derivative chromosome der(1;16)(q10;p10), a result of a centromere-close translocation between chromosome arms 1q and 16p. In addition, we observed that the distribution of cells with the translocation varied from sample to sample, some had a homogenous cell population while others displayed intratumor heterogeneity with cell-to-cell variation. Finally, for one tumor with both preinvasive and invasive components, the fraction of cells with translocation was lower and more heterogeneous in the preinvasive tumor cells compared to the cells in the invasive component. (c) 2014 The Authors Genes, Chromosomes & Cancer Published by Wiley Periodicals, Inc.

Simple genetic models for autism spectrum disorder.
Mukhopadhyay, Swagatam and Wigler, Michael and Levy, Dan (2015) bioRxiv, pp. 17301.

To explore the interplay between new mutation, transmission, and gender bias in genetic disease requires formal quantitative modeling. Autism spectrum disorders offer an ideal case: they are genetic in origin, complex, and show a gender bias. The high reproductive costs of autism ensure that most strongly associated genetic mutations are short-lived, and indeed the disease exhibits both transmitted and de novo components. There is a large body of both epidemiologic and genomic data that greatly constrain the genetic mechanisms that may contribute to the disorder. We develop a computational framework that assumes classes of additive variants, each member of a class having equal effect. We restrict our initial exploration to single class models, each having three parameters. Only one model matches epidemiological data. It also independently matches the incidence of de novo mutation in simplex families, the gender bias in unaffected siblings in simplex populations, and rates of mutation in target genes. This model makes strong and as yet not fully tested predictions, namely that females are the primary carriers in cases of genetic transmission, and that the incidence of de novo mutation in target genes for families at high risk for autism are not especially elevated. In its simplicity, this model does not account for MZ twin concordance or the distorted gender bias of high functioning children with ASD, and does not accommodate all the known mechanisms contributing to ASD. We point to the next steps in applying the same computational framework to explore more complex models.

Stable heteroplasmy at the single-cell level is facilitated by intercellular exchange of mtDNA.
Jayaprakash, A. D. and Benson, E. K. and Gone, S. and Liang, R. and Shim, J. and Lambertini, L. and Toloue, M. M. and Wigler, M. and Aaronson, S. A. and Sachidanandam, R. (2015) Nucleic Acids Research, 43(4) pp. 2177-2187.

Eukaryotic cells carry two genomes, nuclear (nDNA) and mitochondrial (mtDNA), which are ostensibly decoupled in their replication, segregation and inheritance. It is increasingly appreciated that heteroplasmy, the occurrence of multiple mtDNA haplotypes in a cell, plays an important biological role, but its features are not well understood. Accurately determining the diversity of mtDNA has been difficult, due to the relatively small amount of mtDNA in each cell (<1% of the total DNA), the intercellular variability of mtDNA content and mtDNA pseudogenes (Numts) in nDNA. To understand the nature of heteroplasmy, we developed Mseek, a novel technique to purify and sequence mtDNA. Mseek yields high purity (>90%) mtDNA and its ability to detect rare variants is limited only by sequencing depth, providing unprecedented sensitivity and specificity. Using Mseek, we confirmed the ubiquity of heteroplasmy by analyzing mtDNA from a diverse set of cell lines and human samples. Applying Mseek to colonies derived from single cells, we find heteroplasmy is stably maintained in individual daughter cells over multiple cell divisions. We hypothesized that the stability of heteroplasmy could be facilitated by intercellular exchange of mtDNA. We explicitly demonstrate this exchange by co-culturing cell lines with distinct mtDNA haplotypes. Our results shed new light on the maintenance of heteroplasmy and provide a novel platform to investigate features of heteroplasmy in normal and diseased states.

Interactive analysis and quality assessment of single-cell copy-number variations.
Garvin, Tyler and Aboukhalil, Robert and Kendall, Jude and Baslan, Timour and Atwal, Gurinder S and Hicks, James and Wigler, Michael and Schatz, Michael (2015) Nature Methods, 12 pp. 1058-1060.

Single-cell sequencing is emerging as a critical technology for understanding the biology of cancer, neurons, and other complex systems. Here we introduce Ginkgo, a web platform for the interactive analysis and quality assessment of single-cell copy-number alterations. Ginkgo fully automates the process of binning, normalizing, and segmenting mapped reads to infer copy number profiles of individual cells, as well as constructing phylogenetic trees of how those cells are related. We validate Ginkgo by reproducing the results of five major single-cell studies, and discuss how it addresses the wide array of biases that affect single-cell analysis. We also examine the data characteristics of three commonly used single-cell amplification techniques: MDA, MALBAC, and DOP-PCR/WGA4 through comparative analysis of 9 different single-cell datasets. We conclude that DOP-PCR provides the most uniform amplification, while MDA introduces substantial biases into the analysis. Furthermore, given the same level of coverage, our results indicate that data prepared using DOP-PCR can reliably call CNVs at higher resolution than data prepared using either MALBAC or MDA. Ginkgo is freely available at November 11, 2014.Accepted November 12, 2014.© 2014, Published by Cold Spring Harbor Laboratory PressThis pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at

The contribution of de novo coding mutations to autism spectrum disorder.
Iossifov, I. and O'Roak, B. J. and Sanders, S. J. and Ronemus, M. and Krumm, N. and Levy, D. and Stessman, H. A. and Witherspoon, K. T. and Vives, L. and Patterson, K. E. and Smith, J. D. and Paeper, B. and Nickerson, D. A. and Dea, J. and Dong, S. and Gonzalez, L. E. and Mandell, J. D. and Mane, S. M. and Murtha, M. T. and Sullivan, C. A. and Walker, M. F. and Waqar, Z. and Wei, L. and Willsey, A. J. and Yamrom, B. and Lee, Y. H. and Grabowska, E. and Dalkic, E. and Wang, Z. and Marks, S. and Andrews, P. and Leotta, A. and Kendall, J. and Hakker, I. and Rosenbaum, J. and Ma, B. and Rodgers, L. and Troge, J. and Narzisi, G. and Yoon, S. and Schatz, M. C. and Ye, K. and McCombie, W. R. and Shendure, J. and Eichler, E. E. and State, M. W. and Wigler, M. (2014) Nature, 515(7526) pp. 216-221.

Whole exome sequencing has proven to be a powerful tool for understanding the genetic architecture of human disease. Here we apply it to more than 2,500 simplex families, each having a child with an autistic spectrum disorder. By comparing affected to unaffected siblings, we show that 13% of de novo missense mutations and 43% of de novo likely gene-disrupting (LGD) mutations contribute to 12% and 9% of diagnoses, respectively. Including copy number variants, coding de novo mutations contribute to about 30% of all simplex and 45% of female diagnoses. Almost all LGD mutations occur opposite wild-type alleles. LGD targets in affected females significantly overlap the targets in males of lower intelligence quotient (IQ), but neither overlaps significantly with targets in males of higher IQ. We estimate that LGD mutation in about 400 genes can contribute to the joint class of affected females and males of lower IQ, with an overlapping and similar number of genes vulnerable to contributory missense mutation. LGD targets in the joint class overlap with published targets for intellectual disability and schizophrenia, and are enriched for chromatin modifiers, FMRP-associated genes and embryonically expressed genes. Most of the significance for the latter comes from affected females.

Facilitated sequence counting and assembly by template mutagenesis.
Levy, D. and Wigler, M. (2014) Proceedings of the National Academy of Sciences of the United States of America, 111(43) pp. E4632-E4637.

Presently, inferring the long-range structure of the DNA templates is limited by short read lengths. Accurate template counts suffer from distortions occurring during PCR amplification. We explore the utility of introducing random mutations in identical or nearly identical templates to create distinguishable patterns that are inherited during subsequent copying. We simulate the applications of this process under assumptions of error-free sequencing and perfect mapping, using cytosine deamination as a model for mutation. The simulations demonstrate that within readily achievable conditions of nucleotide conversion and sequence coverage, we can accurately count the number of otherwise identical molecules as well as connect variants separated by long spans of identical sequence. We discuss many potential applications, such as transcript profiling, isoform assembly, haplotype phasing, and de novo genome assembly.

Accurate de novo and transmitted indel detection in exome-capture data using microassembly.
Narzisi, G. and O'Rawe, Jason and Iossifov, I. and Fang, Han and Lee, Y. H. and Wang, Zihua and Wu, Yiyang and Lyon, Gholson J. and Wigler, M. H. and Schatz, M. C. (2014) Nature Methods, 11(10) pp. 1033-1036.

We present an open-source algorithm, Scalpel (, which combines mapping and assembly for sensitive and specific discovery of insertions and deletions (indels) in exome-capture data. A detailed repeat analysis coupled with a self-tuning k-mer strategy allows Scalpel to outperform other state-of-the-art approaches for indel discovery, particularly in regions containing near-perfect repeats. We analyzed 593 families from the Simons Simplex Collection and demonstrated Scalpel's power to detect long (≥30 bp) transmitted events and enrichment for de novo likely gene-disrupting indels in autistic children.

Rapid phenotypic and genomic change in response to therapeutic pressure in prostate cancer inferred by high content analysis of single circulating tumor cells.
Dago, A. E. and Stepansky, A. and Carlsson, A. and Luttgen, M. and Kendall, J. and Baslan, T. and Kolatkar, A. and Wigler, M. and Bethel, K. and Gross, M. E. and Hicks, J. and Kuhn, P. (2014) PLoS One, 9(8) pp. e101777.

Timely characterization of a cancer's evolution is required to predict treatment efficacy and to detect resistance early. High content analysis of single Circulating Tumor Cells (CTCs) enables sequential characterization of genotypic, morphometric and protein expression alterations in real time over the course of cancer treatment. This concept was investigated in a patient with castrate-resistant prostate cancer progressing through both chemotherapy and targeted therapy. In this case study, we integrate across four timepoints 41 genome-wide copy number variation (CNV) profiles plus morphometric parameters and androgen receptor (AR) protein levels. Remarkably, little change was observed in response to standard chemotherapy, evidenced by the fact that a unique clone (A), exhibiting highly rearranged CNV profiles and AR+ phenotype was found circulating before and after treatment. However, clinical response and subsequent progression after targeted therapy was associated with the drastic depletion of clone A, followed by the sequential emergence of two distinct CTC sub-populations that differed in both AR genotype and expression phenotype. While AR- cells with flat or pseudo-diploid CNV profiles (clone B) were identified at the time of response, a new tumor lineage of AR+ cells (clone C) with CNV altered profiles was detected during relapse. We showed that clone C, despite phylogenetically related to clone A, possessed a unique set of somatic CNV alterations, including MYC amplification, an event linked to hormone escape. Interesting, we showed that both clones acquired AR gene amplification by deploying different evolutionary paths. Overall, these data demonstrate the timeframe of tumor evolution in response to therapy and provide a framework for the multi-scale analysis of fluid biopsies to quantify and monitor disease evolution in individual patients.

The contribution of de novo and rare inherited copy number changes to congenital heart disease in an unselected sample of children with conotruncal defects or hypoplastic left heart disease.
Warburton, D. and Ronemus, M. and Kline, J. and Jobanputra, V. and Williams, I. and Anyane-Yeboa, K. and Chung, W. and Yu, L. and Wong, N. and Awad, D. and Yu, C. Y. and Leotta, A. and Kendall, J. and Yamrom, B. and Lee, Y. H. and Wigler, M. and Levy, D. (2014) Human Genetics, 133(1) pp. 11-27.

Congenital heart disease (CHD) is the most common congenital malformation, with evidence of a strong genetic component. We analyzed data from 223 consecutively ascertained families, each consisting of at least one child affected by a conotruncal defect (CNT) or hypoplastic left heart disease (HLHS) and both parents. The NimbleGen HD2-2.1 comparative genomic hybridization platform was used to identify de novo and rare inherited copy number variants (CNVs). Excluding 10 cases with 22q11.2 DiGeorge deletions, we validated de novo CNVs in 8 % of 148 probands with CNTs, 12.7 % of 71 probands with HLHS and none in 4 probands with both. Only 2 % of control families showed a de novo CNV. We also identified a group of ultra-rare inherited CNVs that occurred de novo in our sample, contained a candidate gene for CHD, recurred in our sample or were present in an affected sibling. We confirmed the contribution to CHD of copy number changes in genes such as GATA4 and NODAL and identified several genes in novel recurrent CNVs that may point to novel CHD candidate loci. We also found CNVs previously associated with highly variable phenotypes and reduced penetrance, such as dup 1q21.1, dup 16p13.11, dup 15q11.2-13, dup 22q11.2, and del 2q23.1. We found that the presence of extra-cardiac anomalies was not related to the frequency of CNVs, and that there was no significant difference in CNV frequency or specificity between the probands with CNT and HLHS. In agreement with other series, we identified likely causal CNVs in 5.6 % of our total sample, half of which were de novo.

The role of de novo mutations in the genetics of autism spectrum disorders.
Ronemus, M. and Iossifov, I. and Levy, D. and Wigler, M. (2014) Nature Reviews Genetics, 15(2) pp. 133-141.

The identification of the genetic components of autism spectrum disorders (ASDs) has advanced rapidly in recent years, particularly with the demonstration of de novo mutations as an important source of causality. We review these developments in light of genetic models for ASDs. We consider the number of genetic loci that underlie ASDs and the relative contributions from different mutational classes, and we discuss possible mechanisms by which these mutations might lead to dysfunction. We update the two-class risk genetic model for autism, especially in regard to children with high intelligence quotients.

Role of SWI/SNF in acute leukemia maintenance and enhancer-mediated Myc regulation.
Shi, J. and Whyte, W. A. and Zepeda-Mendoza, C. J. and Milazzo, J. P. and Shen, C. and Roe, J. S. and Minder, J. L. and Mercan, F. and Wang, E. and Eckersley-Maslin, M. A. and Campbell, A. E. and Kawaoka, S. and Shareef, S. and Zhu, Z. and Kendall, J. and Muhar, M. and Haslinger, C. and Yu, M. and Roeder, R. G. and Wigler, M. A. and Blobel, G. A. and Zuber, J. and Spector, D. L. and Young, R. A. and Vakoc, C. R. (2013) Genes and Development, 27(24) pp. 264-273.

Cancer cells frequently depend on chromatin regulatory activities to maintain a malignant phenotype. Here, we show that leukemia cells require the mammalian SWI/SNF chromatin remodeling complex for their survival and aberrant self-renewal potential. While Brg1, an ATPase subunit of SWI/SNF, is known to suppress tumor formation in several cell types, we found that leukemia cells instead rely on Brg1 to support their oncogenic transcriptional program, which includes Myc as one of its key targets. To account for this context-specific function, we identify a cluster of lineage-specific enhancers located 1.7 Mb downstream from Myc that are occupied by SWI/SNF as well as the BET protein Brd4. Brg1 is required at these distal elements to maintain transcription factor occupancy and for long-range chromatin looping interactions with the Myc promoter. Notably, these distal Myc enhancers coincide with a region that is focally amplified in approximately 3% of acute myeloid leukemias. Together, these findings define a leukemia maintenance function for SWI/SNF that is linked to enhancer-mediated gene regulation, providing general insights into how cancer cells exploit transcriptional coactivators to maintain oncogenic gene expression programs.

Cold Spring Harbor Laboratory Annual Reports Excerpts: Wigler Laboratory 1979-2011.
Wigler, M. H. (2013) Cold Spring Harbor Laboratory.

Target inference from collections of genomic intervals.
Krasnitz, A. and Sun, G. and Andrews, P. and Wigler, M. (2013) Proceedings of the National Academy of Sciences of the United States of America, 110(25) pp. E2271-E2278.

Finding regions of the genome that are significantly recurrent in noisy data are a common but difficult problem in present day computational biology. Cores of recurrent events (CORE) is a computational approach to solving this problem that is based on a formalized notion by which "core" intervals explain the observed data, where the number of cores is the "depth" of the explanation. Given that formalization, we implement CORE as a combinatorial optimization procedure with depth chosen from considerations of statistical significance. An important feature of CORE is its ability to explain data with cores of widely varying lengths. We examine the performance of this system with synthetic data, and then provide two demonstrations of its utility with actual data. Applying CORE to a collection of DNA copy number profiles from single cells of a given tumor, we determine tumor population phylogeny and find the features that separate subpopulations. Applying CORE to comparative genomic hybridization data from a large set of tumor samples, we define regions of recurrent copy number aberration in breast cancer.

The Cancer Stem Cell: Cell Type or Cell State?.
Donnenberg, A. D. and Hicks, J. B. and Wigler, M. and Donnenberg, V. S. (2013) Cytometry Part A, 83A(SI 1) pp. 5-7.

Broad applications of single-cell nucleic acid analysis in biomedical research.
Wigler, M. (2012) Genome Med, 4(10) pp. 79.

Rare De Novo Germline Copy-Number Variation in Testicular Cancer.
Stadler, Z. K. and Esposito, D. and Shah, S. and Vijai, J. and Yamrom, B. and Levy, D. and Lee, Y. H. and Kendall, J. and Leotta, A. and Ronemus, M. and Hansen, N. and Sarrel, K. and Rau-Murthy, R. and Schrader, K. and Kauff, N. and Klein, R. and Lipkin, S. and Murali, R. and Robson, M. and Sheinfeld, J. and Feldman, D. and Bosl, G. and Norton, L. and Wigler, M. and Offit, K. (2012) The American Journal of Human Genetics, 91(2) pp. 379-383.

Although heritable factors are an important determinant of risk of early-onset cancer, the majority of these malignancies appear to occur sporadically without identifiable risk factors. Germline de novo copy-number variations (CNVs) have been observed in sporadic neurocognitive and cardiovascular disorders. We explored this mechanism in 382 genomes of 116 early-onset cancer case-parent trios and unaffected siblings. Unique de novo germline CNVs were not observed in 107 breast or colon cancer trios or controls but were indeed found in 7% of 43 testicular germ cell tumor trios; this percentage exceeds background CNV rates and suggests a rare de novo genetic paradigm for susceptibility to some human malignancies.

Cooperating Tumor Suppressor Genes on Chromosome 8p Predict Survival Outcome in Hepatocellular Carcinoma.
Kitzing, T. and Xue, W. and Roessler, S. and Krasnitz, A. and Schultz, N. and Wang, X. W. and Wigler, M. and Lowe, S. W. (2012) European Journal of Cancer, 48 pp. S139-S139.

Genome-wide copy number analysis of single cells.
Baslan, T. and Kendall, J. T. and Rodgers, L. and Cox, H. and Riggs, M. and Stepansky, A. and Troge, J. E. and Ravi, K. and Esposito, D. and Lakshmi, B. and Wigler, M. and Navin, N. E. and Hicks, J. B. (2012) Nat Protoc, 7(6) pp. 1024-41.

Copy number variation (CNV) is increasingly recognized as an important contributor to phenotypic variation in health and disease. Most methods for determining CNV rely on admixtures of cells in which information regarding genetic heterogeneity is lost. Here we present a protocol that allows for the genome-wide copy number analysis of single nuclei isolated from mixed populations of cells. Single-nucleus sequencing (SNS), combines flow sorting of single nuclei on the basis of DNA content and whole-genome amplification (WGA); this is followed by next-generation sequencing to quantize genomic intervals in a genome-wide manner. Multiplexing of single cells is discussed. In addition, we outline informatic approaches that correct for biases inherent in the WGA procedure and allow for accurate determination of copy number profiles. All together, the protocol takes approximately 3 d from flow cytometry to sequence-ready DNA libraries.

A cluster of cooperating tumor-suppressor gene candidates in chromosomal deletions.
Xue, W. and Kitzing, T. and Roessler, S. and Zuber, J. and Krasnitz, A. and Schultz, N. and Revill, K. and Weissmueller, S. and Rappaport, A. R. and Simon, J. and Zhang, J. and Luo, W. J. and Hicks, J. and Zender, L. and Wang, X. W. and Powers, S. and Wigler, M. and Lowe, S. W. (2012) Proceedings of the National Academy of Sciences of the United States of America, 109(21) pp. 8212-8217.

The large chromosomal deletions frequently observed in cancer genomes are often thought to arise as a "two-hit" mechanismin the process of tumor-suppressor gene (TSG) inactivation. Using a murine model system of hepatocellular carcinoma (HCC) and in vivo RNAi, we test an alternative hypothesis, that such deletions can arise from selective pressure to attenuate the activity of multiple genes. By targeting the mouse orthologs of genes frequently deleted on human 8p22 and adjacent regions, which are lost in approximately half of several other major epithelial cancers, we provide evidence suggesting that multiple genes on chromosome 8p can cooperatively inhibit tumorigenesis in mice, and that their cosuppression can synergistically promote tumor growth. In addition, in human HCC patients, the combined down-regulation of functionally validated 8p TSGs is associated with poor survival, in contrast to the down-regulation of any individual gene. Our data imply that large cancer-associated deletions can produce phenotypes distinct from those arising through loss of a single TSG, and as such should be considered and studied as distinct mutational events.

De Novo Gene Disruptions in Children on the Autistic Spectrum.
Iossifov, I. and Ronemus, M. and Levy, D. and Wang, Z. H. and Hakker, I. and Rosenbaum, J. and Yamrom, B. and Lee, Y. H. and Narzisi, G. and Leotta, A. and Kendall, J. and Grabowska, E. and Ma, B. C. and Marks, S. and Rodgers, L. and Stepansky, A. and Troge, J. and Andrews, P. and Bekritsky, M. and Pradhan, K. and Ghiban, E. and Kramer, M. and Parla, J. and Demeter, R. and Fulton, L. L. and Fulton, R. S. and Magrini, V. J. and Ye, K. and Darnell, J. C. and Darnell, R. B. and Mardis, E. R. and Wilson, R. K. and Schatz, M. C. and McCombie, W. R. and Wigler, M. (2012) Neuron, 74(2) pp. 285-299.

Exome sequencing of 343 families, each with a single child on the autism spectrum and at least one unaffected sibling, reveal de novo small indels and point substitutions, which come mostly from the paternal line in an age-dependent manner. We do not see significantly greater numbers of de novo missense mutations in affected versus unaffected children, but gene-disrupting mutations (nonsense, splice site, and frame shifts) are twice as frequent, 59 to 28. Based on this differential and the number of recurrent and total targets of gene disruption found in our and similar studies, we estimate between 350 and 400 autism susceptibility genes. Many of the disrupted genes in these studies are associated with the fragile X protein, FMRP, reinforcing links between autism and synaptic plasticity. We find FMRP-associated genes are under greater purifying selection than the remainder of genes and suggest they are especially dosage-sensitive targets of cognitive disorders.

Reducing system noise in copy number data using principal components of self-self hybridizations.
Lee, Y. H. and Ronemus, M. and Kendall, J. and Lakshmi, B. and Leotta, A. and Levy, D. and Esposito, D. and Grubor, V. and Ye, K. and Wigler, M. and Yamrom, B. (2012) Proceedings of the National Academy of Sciences of the United States of America, 109(3) pp. E103-E110.

Genomic copy number variation underlies genetic disorders such as autism, schizophrenia, and congenital heart disease. Copy number variations are commonly detected by array based comparative genomic hybridization of sample to reference DNAs, but probe and operational variables combine to create correlated system noise that degrades detection of genetic events. To correct for this we have explored hybridizations in which no genetic signal is expected, namely "self-self" hybridizations (SSH) comparing DNAs from the same genome. We show that SSH trap a variety of correlated system noise present also in sample-reference (test) data. Through singular value decomposition of SSH, we are able to determine the principal components (PCs) of this noise. The PCs themselves offer deep insights into the sources of noise, and facilitate detection of artifacts. We present evidence that linear and piecewise linear correction of test data with the PCs does not introduce detectable spurious signal, yet improves signal-to-noise metrics, reduces false positives, and facilitates copy number determination.

Dosage-dependent phenotypes in models of 16p11.2 lesions found in autism.
Horev, G. and Ellegood, J. and Lerch, J. P. and Son, Y. E. and Muthuswamy, L. and Vogel, H. and Krieger, A. M. and Buja, A. and Henkelman, R. M. and Wigler, M. H. and Mills, A. A. (2011) Proceedings of the National Academy of Sciences of the United States of America, 108(41) pp. 17076-81.

Recurrent copy number variations (CNVs) of human 16p11.2 have been associated with a variety of developmental/neurocognitive syndromes. In particular, deletion of 16p11.2 is found in patients with autism, developmental delay, and obesity. Patients with deletions or duplications have a wide range of clinical features, and siblings carrying the same deletion often have diverse symptoms. To study the consequence of 16p11.2 CNVs in a systematic manner, we used chromosome engineering to generate mice harboring deletion of the chromosomal region corresponding to 16p11.2, as well as mice harboring the reciprocal duplication. These 16p11.2 CNV models have dosage-dependent changes in gene expression, viability, brain architecture, and behavior. For each phenotype, the consequence of the deletion is more severe than that of the duplication. Of particular note is that half of the 16p11.2 deletion mice die postnatally; those that survive to adulthood are healthy and fertile, but have alterations in the hypothalamus and exhibit a "behavior trap" phenotype-a specific behavior characteristic of rodents with lateral hypothalamic and nigrostriatal lesions. These findings indicate that 16p11.2 CNVs cause brain and behavioral anomalies, providing insight into human neurodevelopmental disorders.

Rare De Novo Variants Associated with Autism Implicate a Large Functional Network of Genes Involved in Formation and Function of Synapses.
Gilman, S. R and Iossifov, I. and Levy, D. and Ronemus, M. and Wigler, M. H. and Vitkup, D. (2011) Neuron, 70(5) pp. 898-907.

Summary Identification of complex molecular networks underlying common human phenotypes is a major challenge of modern genetics. In this study, we develop a method for network-based analysis of genetic associations (NETBAG). We use NETBAG to identify a large biological network of genes affected by rare de novo CNVs in autism. The genes forming the network are primarily related to synapse development, axon targeting, and neuron motility. The identified network is strongly related to genes previously implicated in autism and intellectual disability phenotypes. Our results are also consistent with the hypothesis that significantly stronger functional perturbations are required to trigger the autistic phenotype in females compared to males. Overall, the presented analysis of de novo variants supports the hypothesis that perturbed synaptogenesis is at the heart of autism. More generally, our study provides proof of the principle that networks underlying complex human phenotypes can be identified by a network-based functional analysis of rare genetic variants.

Rare De Novo and Transmitted Copy-Number Variation in Autistic Spectrum Disorders.
Levy, D. and Ronemus, M. and Yamrom, B. and Lee, Y. H. and Leotta, A. and Kendall, J. T. and Marks, S. and Lakshmi, B. and Pai, D. and Ye, Kenny and Buja, Andreas and Krieger, Abba and Yoon, S. and Troge, J. E. and Rodgers,  L. and Iossifov, I. and Wigler, M. H. (2011) Neuron, 70(5) pp. 886-897.

To explore the genetic contribution to autistic spectrum disorders (ASDs), we have studied genomic copy-number variation in a large cohort of families with a single affected child and at least one unaffected sibling. We confirm a major contribution from de novo deletions and duplications but also find evidence of a role for inherited "ultrarare" duplications. Our results show that, relative to males, females have greater resistance to autism from genetic causes, which raises the question of the fate of female carriers. By analysis of the proportion and number of recurrent loci, we set a lower bound for distinct target loci at several hundred. We find many new candidate regions, adding substantially to the list of potential gene targets, and confirm several loci previously observed. The functions of the genes in the regions of de novo variation point to a great diversity of genetic causes but also suggest functional convergence.

DNA methylation patterns in luminal breast cancers differ from non-luminal subtypes and can identify relapse risk independent of other clinical variables.
Kamalakaran, S. and Varadan, V. and Giercksky Russnes, H. E. and Levy, D. and Kendall, J. T. and Janevski, A. and Riggs, M. and Banerjee, N. and Synnestvedt, M. and Schlichting, E. and Kåresen, R. and Shama Prasada, K. and Rotti, H. and Rao, R. and Rao, L. and Eric Tang, M. H. and Satyamoorthy, K. and Lucito, R. and Wigler, M. H. and Dimitrova, N. and Naume, B. and Borresen-Dale, A. L. and Hicks, J. B. (2011) Molecular Oncology, 5(1) pp. 77-92.

The diversity of breast cancers reflects variations in underlying biology and affects the clinical implications for patients. Gene expression studies have identified five major subtypes- Luminal A, Luminal B, basal-like, ErbB2+ and Normal-Like. We set out to determine the role of DNA methylation in subtypes by performing genome-wide scans of CpG methylation in breast cancer samples with known expression-based subtypes. Unsupervised hierarchical clustering using a set of most varying loci clustered the tumors into a Luminal A majority (82%) cluster, Basal-like/ErbB2+ majority (86%) cluster and a non-specific cluster with samples that were also inconclusive in their expression-based subtype correlations. Contributing methylation loci were both gene associated loci (30%) and non-gene associated (70%), suggesting subtype dependant genome-wide alterations in the methylation landscape. The methylation patterns of significant differentially methylated genes in luminal A tumors are similar to those identified in CD24 + luminal epithelial cells and the patterns in basal-like tumors similar to CD44 + breast progenitor cells. CpG islands in the HOXA cluster and other homeobox (IRX2, DLX2, NKX2-2) genes were significantly more methylated in Luminal A tumors. A significant number of genes (2853, p < 0.05) exhibited expression-methylation correlation, implying possible functional effects of methylation on gene expression. Furthermore, analysis of these tumors by using follow-up survival data identified differential methylation of islands proximal to genes involved in Cell Cycle and Proliferation (Ki-67, UBE2C, KIF2C, HDAC4), angiogenesis (VEGF, BTG1, KLF5), cell fate commitment (SPRY1, OLIG2, LHX2 and LHX5) as having prognostic value independent of subtypes and other clinical factors. © 2010 Federation of European Biochemical Societies.

Tumour evolution inferred by single-cell sequencing.
Navin, N. E. and Kendall, J. T. and Troge, J. E. and Andrews, P. and Rodgers,  L. and McIndoo, J. and Cook, K. and Stepansky,  A. and Levy, D. and Esposito, D. and Muthuswamy, L. and Krasnitz, A. and McCombie, W. R. and Hicks, J. B. and Wigler, M. H. (2011) Nature, 472(7341) pp. 90-94.

Genomic analysis provides insights into the role of copy number variation in disease, but most methods are not designed to resolve mixed populations of cells. In tumours, where genetic heterogeneity is common, very important information may be lost that would be useful for reconstructing evolutionary history. Here we show that with flow-sorted nuclei, whole genome amplification and next generation sequencing we can accurately quantify genomic copy number within an individual nucleus. We apply single-nucleus sequencing to investigate tumour population structure and evolution in two human breast cancer cases. Analysis of 100 single cells from a polygenomic tumour revealed three distinct clonal subpopulations that probably represent sequential clonal expansions. Additional analysis of 100 single cells from a monogenomic primary tumour and its liver metastasis indicated that a single clonal expansion formed the primary tumour and seeded the metastasis. In both primary tumours, we also identified an unexpectedly abundant subpopulation of genetically diverse 'pseudodiploid' cells that do not travel to the metastatic site. In contrast to gradual models of tumour progression, our data indicate that tumours grow by punctuated clonal expansions with few persistent intermediates.

Genomic architecture characterizes tumor progression paths and fate in breast cancer patients.
Russnes, H. G. and Vollan, H. K. M. and Lingjærde, O. C. and Krasnitz, A. and Lundin, P. and Naume, B. and Sørlie, T. and Borgen, E. and Rye, I. H. and Langerød, A. and Chin, S. F. and Teschendorff, A. E. and Stephens, P. J. and Månér, S. and Schlichting, E. and Baumbusch, L. O. and Kåresen, R. and Stratton, M. P. and Wigler, M. H. and Caldas, C. and Zetterberg, A. and Hicks, J. B. and Børresen-Dale, A. L. (2010) Science Translational Medicine, 2(38)

Distinct molecular subtypes of breast carcinomas have been identified, but translation into clinical use has been limited. We have developed two platform-independent algorithms to explore genomic architectural distortion using array comparative genomic hybridization data to measure (i) whole-arm gains and losses [whole-arm aberration index (WAAI)] and (ii) complex rearrangements [complex arm aberration index (CAAI)]. By applying CAAI and WAAI to data from 595 breast cancer patients, we were able to separate the cases into eight subgroups with different distributions of genomic distortion. Within each subgroup data from expression analyses, sequencing and ploidy indicated that progression occurs along separate paths into more complex genotypes. Histological grade had prognostic impact only in the luminal-related groups, whereas the complexity identified by CAAI had an overall independent prognostic power. This study emphasizes the relation among structural genomic alterations, molecular subtype, and clinical behavior and shows that objective score of genomic complexity (CAAI) is an independent prognostic marker in breast cancer.

Tracing Tumor Lineage and Progression through Genomic Copy Number Profiling at the Single Cell Level.
Hicks, J. B. and Navin, N. E. and Kendall, J. T. and Levy, D. and Wigler, M. H. (2010) Annals of Oncology, 21(S4) pp. iv49.

Inferring tumor progression from genomic heterogeneity.
Navin, N. E. and Krasnitz, A. and Rodgers, L. and Cook, K. and Meth, J. L. and Kendall, J. T. and Riggs, M. and Eberling, Y. and Troge, J. E. and Grubor, V. and Levy, D. and Lundin, P. and Månér, S. and Zetterberg, A. and Hicks, J. B. and Wigler, M. H. (2010) Genome Research, 20(1) pp. 68-80.

Cancer progression in humans is difficult to infer because we do not routinely sample patients at multiple stages of their disease. However, heterogeneous breast tumors provide a unique opportunity to study human tumor progression because they still contain evidence of early and intermediate subpopulations in the form of the phylogenetic relationships. We have developed a method we call Sector-Ploidy-Profiling (SPP) to study the clonal composition of breast tumors. SPP involves macro-dissecting tumors, flow-sorting genomic subpopulations by DNA content, and profiling genomes using comparative genomic hybridization (CGH). Breast carcinomas display two classes of genomic structural variation: (1) monogenomic and (2) polygenomic. Monogenomic tumors appear to contain a single major clonal subpopulation with a highly stable chromosome structure. Polygenomic tumors contain multiple clonal tumor subpopulations, which may occupy the same sectors, or separate anatomic locations. In polygenomic tumors, we show that heterogeneity can be ascribed to a few clonal subpopulations, rather than a series of gradual intermediates. By comparing multiple subpopulations from different anatomic locations, we have inferred pathways of cancer progression and the organization of tumor growth. © 2010 by Cold Spring Harbor Laboratory Press.

Functional identification of tumor-suppressor genes through an in vivo RNA interference screen in a mouse lymphoma model.
Bric, A. and Miething, C. and Bialucha, C. U. and Scuoppo, C. and Zender, L. and Krasnitz, A. and Xuan, Z. and Zuber, J. and Wigler, M. H. and Hicks, J. B. and McCombie, W. R. and Hemann, M. T. and Hannon, G. J. and Powers, S. and Lowe, S. W. (2009) Cancer Cell, 16(4) pp. 324-35.

Short hairpin RNAs (shRNAs) capable of stably suppressing gene function by RNA interference (RNAi) can mimic tumor-suppressor-gene loss in mice. By selecting for shRNAs capable of accelerating lymphomagenesis in a well-characterized mouse lymphoma model, we identified over ten candidate tumor suppressors, including Sfrp1, Numb, Mek1, and Angiopoietin 2. Several components of the DNA damage response machinery were also identified, including Rad17, which acts as a haploinsufficient tumor suppressor that responds to oncogenic stress and whose loss is associated with poor prognosis in human patients. Our results emphasize the utility of in vivo RNAi screens, identify and validate a diverse set of tumor suppressors, and have therapeutic implications.

High definition profiling of mammalian DNA methylation by array capture and single molecule bisulfite sequencing.
Hodges, E. and Smith, A. D. and Kendall, J. T. and Xuan, Z. and Ravi, K. and Rooks, M. and Zhang, M. Q. and Ye, K. and Bhattacharjee, A. and Brizuela, L. and McCombie, W. R. and Wigler, M. H. and Hannon, G. J. and Hicks, J. B. (2009) Genome Research, 19(9) pp. 1593-1605.

DNA methylation stabilizes developmentally programmed gene expression states. Aberrant methylation is associated with disease progression and is a common feature of cancer genomes. Presently, few methods enable quantitative, large-scale, single-base resolution mapping of DNA methylation states in desired regions of a complex mammalian genome. Here, we present an approach that combines array-based hybrid selection and massively parallel bisulfite sequencing to profile DNA methylation in genomic regions spanning hundreds of thousands of bases. This single molecule strategy enables methylation variable positions to be quantitatively examined with high sampling precision. Using bisulfite capture, we assessed methylation patterns across 324 randomly selected CpG islands (CGI) representing more than 25,000 CpG sites. A single lane of Illumina sequencing permitted methylation states to be definitively called for >90% of target sties. The accuracy of the hybrid-selection approach was verified using conventional bisulfite capillary sequencing of cloned PCR products amplified from a subset of the selected regions. This confirmed that even partially methylated states could be successfully called. A comparison of human primary and cancer cells revealed multiple differentially methylated regions. More than 25% of islands showed complex methylation patterns either with partial methylation states defining the entire CGI or with contrasting methylation states appearing in specific regional blocks within the island. We observed that transitions in methylation state often correlate with genomic landmarks, including transcriptional start sites and intron-exon junctions. Methylation, along with specific histone marks, was enriched in exonic regions, suggesting that chromatin states can foreshadow the content of mature mRNAs.

High resolution microarray copy number analysis (array CGH) suggests that determination of HER2 amplification by FISH (FISH plus ) is inaccurate in human breast cancer specimens that are HER2 2+by immunohistochemistry (IHC2+).
McArthur, H. L. and Brogi, E. and Patil, S. and Wigler, M. H. and Norton, L. and Hicks, J. B. and Hudis, C. A. (2009) European Journal of Cancer Supplements, 7(2) pp. 88-88.

Topo2a and Her2 Co-Amplification Is Uncommon by High Resolution Representational Oligonucleotide Microarray Analysis (Roma) in Human Breast Cancer.
Hicks, J. B. and McArthur, H. and Tan, L. K. and Patil, S. and Wigler, M. H. and Hudis, C. and Norton, L. (2009) Annals of Oncology, 20(S2) pp. ii22-23.

Molecular subclassification of breast carcinomas based on aCGH reveals two major groups of tumors with common genomic, transcriptomic and clinicopathological characteristics.
Russnes, H. G. and Vollan, H. K. M. and Sorlie, T. and Krasnitz, A. and Zetterberg, A. and Naume, B. and Borgen, E. and Nesland, J. M. and Wigler, M. H. and Borresen-Dale, A. L. and Hicks, J. B. (2009) APMIS, 117(3) pp. 238-239.

Novel genomic alterations and clonal evolution in chronic lymphocytic leukemia revealed by representational oligonucleotide microarray analysis (ROMA).
Grubor, V. and Krasnitz, A. and Troge, J. E. and Meth, J. L. and Lakshmi, B. and Kendall, J. T. and Yamrom, B. and Alex, G. and Pai, D. and Navin, N. E. and Hufnagel, L. A. and Lee, Y. H. and Cook, K. and Allen, S. L. and Rai, K. R. and Damle, R. N. and Calissano, C. and Chiorazzi, N. and Wigler, M. H. and Esposito, D. (2009) Blood, 113(6) pp. 1294-1303.

We examined copy number changes in the genomes of B cells from 58 patients with chronic lymphocytic leukemia (CLL) by using representational oligonucleotide microarray analysis (ROMA), a form of comparative genomic hybridization (CGH), at a resolution exceeding previously published studies. We observed at least 1 genomic lesion in each CLL sample and considerable variation in the number of abnormalities from case to case. Virtually all abnormalities previously reported also were observed here, most of which were indeed highly recurrent. We observed the boundaries of known events with greater clarity and identified previously undescribed lesions, some of which were recurrent. We profiled the genomes of CLL cells separated by the surface marker CD38 and found evidence of distinct subclones of CLL within the same patient. We discuss the potential applications of high-resolution CGH analysis in a clinical setting. (Blood. 2009; 113: 1294-1303)

High resolution representational oligonucleotide microarray analysis (ROMA) suggests that TOPO2 and HER2 co-amplification is uncommon in human breast cancer.
McArthur, H. L. and Tan, L. K. and Patil, S. and Wigler, M. H. and Hudis, C. A. and Hicks, J. B. and Norton, L. (2009) Cancer Res, 69(2, Sup) pp. 163S-163S.

Background Clinical studies relating TOPO2A expression or amplification by FISH have yielded inconsistent results regarding prediction of benefit from various drug therapies, particularly in the setting of HER2 amplification by FISH. Because HER2 and TOPO2A are relatively small genes that are close to each other, and because the relevant commercially-available FISH probes hybridize to DNA sequences beyond the genes of interest, there is significant potential for false positive results. We therefore conducted an exploratory study comparing HER2 and TOPO2A amplification by ROMA and FISH. Methods: Forty-two archived formalin-fixed paraffin-embedded primary breast cancer specimens were pre-selected to contain 36 HER2 amplified and 6 HER2 non-amplified cases by FISH. These specimens were evaluated for HER2 and TOPO2A amplification by FISH at Memorial Sloan-Kettering Cancer Center and by ROMA at Cold Spring Harbor Laboratory in a double-blinded experiment. Two HER2 amplified by FISH specimens proved inevaluable for technical reasons. The results for the remaining 40 evaluable specimens were then compared. Results: Of the 40 evaluable specimens, the 6 cases selected as HER2 non-amplified by FISH were HER2 non-amplified by ROMA and TOPO2A non-amplified by both FISH and ROMA. Thirty-two specimens were HER2 amplified by both FISH and ROMA. Two specimens with HER2 FISH results of 2.2 and 2.4 did not demonstrate focal amplification by ROMA. Twenty-five (74%) of the 34 specimens with HER2 amplification by FISH were TOPO2A non-amplified by both FISH and ROMA (Table 1). However, only 2 (22%) of the 9 specimens with HER2 and TOPO2A co-amplification by FISH demonstrated TOPO2A amplification by ROMA. Conclusions: Our results suggest that by ROMA, the determination of HER2 status (amplified and non-amplified) and TOPO2A non-amplification by FISH is accurate. The absence of TOPO2A amplification in HER2 non-amplified breast cancer is consistent with published reports. We also confirmed the published frequency of HER2 and TOPO2A co-amplification by FISH. However, we found that co-amplification of HER2 and TOPO2A by ROMA is uncommon, which suggests that commercially-available FISH probes over-estimate the frequency of TOPO2A amplification in HER2 amplified cases. Studies evaluating the clinical implications of these findings are underway.

Classification method for microarray probe selection using sequence, thermodynamics and secondary structure parameters.
Gupta, L. and Kumar, S. and Singh, R. and Shaik, R. and Dimitrova, N. and Gorthi, A. and Lakshmi, B. and Pai, D. and Kamalakaran, S. and Zhao, X. and Wigler, M. H. (2008) 19th International Conference on Pattern Recognition,

Probe design is the most important step for any microarray based assay. Accurate and efficient probe design and selection for the target sequence is critical in generating reliable and useful results. Several different approaches for probe design are reported in literature and an increasing number of bioinformatics tools are available for the same. However, based on the reported low accuracy, determining the hybridization efficiency of the probes is still a big computational challenge. Present study deals with the extraction of various novel features related to sequence composition, thermodynamics and secondary structure that may be essential for designing good probes. A feature selection method has been used to assess the relative importance of all these features. In this paper, we validate the importance of various features currently used for designing an oligonucleotide probe. Finally, a classification methodology is presented that can be used to predict the hybridization quality of a probe.

An Oncogenomics-Based In Vivo RNAi Screen Identifies Tumor Suppressors in Liver Cancer.
Zender, L. and Xue, W. and Zuber, J. and Semighini, C. P. and Krasnitz, A. and Ma, B. and Zender, P. and Kubicka, S. and Luk, J. M. and Schirmacher, P. and McCombie, W. R. and Wigler, M. H. and Hicks, J. B. and Hannon, G. J. and Powers, S. and Lowe, S. W. (2008) Cell, 135(5) pp. 852-864.

Summary Cancers are highly heterogeneous and contain many passenger and driver mutations. To functionally identify tumor suppressor genes relevant to human cancer, we compiled pools of short hairpin RNAs (shRNAs) targeting the mouse orthologs of genes recurrently deleted in a series of human hepatocellular carcinomas and tested their ability to promote tumorigenesis in a mosaic mouse model. In contrast to randomly selected shRNA pools, many deletion-specific pools accelerated hepatocarcinogenesis in mice. Through further analysis, we identified and validated 13 tumor suppressor genes, 12 of which had not been linked to cancer before. One gene, XPO4, encodes a nuclear export protein whose substrate, EIF5A2, is amplified in human tumors, is required for proliferation of XPO4-deficient tumor cells, and promotes hepatocellular carcinoma in mice. Our results establish the feasibility of in vivo RNAi screens and illustrate how combining cancer genomics, RNA interference, and mosaic mouse models can facilitate the functional annotation of the cancer genome.

High-Resolution Array-Based Comparative Genome Hybridization (CGH) Identifies Novel and Recurrent Regions in CLL.
Grubor, V. and Krasnitz, A. and Troge, J. E. and Meth, J. L. and Lakshmi, B. and Kendall, J. T. and Yamrom, B. and Alex, G. and Pai, D. and Navin, N. E. and Hufnagel, L. A. and Lee, Y. H. and Cook, K. and Allen, S. L. and Rai, K. R. and Damle, R. and Calissano, C. and Chiorazzi, N. and Wigler, M. H. and Esposito, D. (2008) Blood, 112(11) pp. 717-717.

Although we have some understanding of the genetic abnormalities occurring in B-cell chronic lymphocytic leukemia (CLL) and their association with clinical outcomes, there is an incomplete comprehension of all of the mutations contributing to disease development and progression. In most abnormalities previously observed, it has been difficult to pinpoint specific candidate genes, reflecting the inadequacy of present tools for assessing chromosomal damage.

Molecular subclassification of breast carcinomas based on aCGH, gene expression, IHC and ploidy - relevance for clinical outcome.
Russnes, H. G. and Sorlie, T. and Krasnitz, A. and Zetterberg, A. and Naume, B. and Borgen, E. and Nesland, J. M. and Wigler, M. H. and Borresen-Dale, A. L. and Hicks, J. B. (2008) European Journal of Cancer Supplements, 6(9) pp. 9-9.

DLC1 is a chromosome 8p tumor suppressor whose loss promotes hepatocellular carcinoma.
Xue, W. and Krasnitz, A. and Lucito, R. and Sordella, R. and Van Aelst, L. and Cordon-Cardo, C. and Singer, S. and Kuehnel, F. and Wigler, M. H. and Powers, S. and Zender, L. and Lowe, S. W. (2008) Genes & Development, 22(11) pp. 1439-44.

Deletions on chromosome 8p are common in human tumors, suggesting that one or more tumor suppressor genes reside in this region. Deleted in Liver Cancer 1 (DLC1) encodes a Rho-GTPase activating protein and is a candidate 8p tumor suppressor. We show that DLC1 knockdown cooperates with Myc to promote hepatocellular carcinoma in mice, and that reintroduction of wild-type DLC1 into hepatoma cells with low DLC1 levels suppresses tumor growth in situ. Cells with reduced DLC1 protein contain increased GTP-bound RhoA, and enforced expression a constitutively activated RhoA allele mimics DLC1 loss in promoting hepatocellular carcinogenesis. Conversely, down-regulation of RhoA selectively inhibits tumor growth of hepatoma cells with disabled DLC1. Our data validate DLC1 as a potent tumor suppressor gene and suggest that its loss creates a dependence on the RhoA pathway that may be targeted therapeutically.

Linkage, Association, and Gene-Expression Analyses Identify CNTNAP2 as an Autism-Susceptibility Gene.
Alarcon, M. and Abrahams, B. S. and Stone, J. L. and Duvall, J. A. and Perederiy, J. V. and Bomar, J. M. and Sebat, J. and Wigler, M. H. and Martin, C. L. and Ledbetter, D. H. and Nelson, S. F. and Cantor, R. M. and Geschwind, D. H. (2008) Am J Hum Genet, 82(1) pp. 150-159.

Autism is a genetically complex neurodevelopmental syndrome in which language deficits are a core feature. We describe results from two complimentary approaches used to identify risk variants on chromosome 7 that likely contribute to the etiology of autism. A two-stage association study tested 2758 SNPs across a 10 Mb 7q35 language-related autism QTL in AGRE (Autism Genetic Resource Exchange) trios[1] and [2] and found significant association with Contactin Associated Protein-Like 2 (CNTNAP2), a strong a priori candidate. Male-only containing families were identified as primarily responsible for this association signal, consistent with the strong male affection bias in ASD and other language-based disorders. Gene-expression analyses in developing human brain further identified CNTNAP2 as enriched in circuits important for language development. Together, these results provide convergent evidence for involvement of CNTNAP2, a Neurexin family member, in autism, and demonstrate a connection between genetic risk for autism and specific brain structures.

Recurrent DNA copy number variation in the laboratory mouse.
Egan, C. M. and Sridhar, S. and Wigler, M. H. and Hall, I. M. (2007) Nat Genet, 39(11) pp. 1384-9.

Different species, populations and individuals vary considerably in the copy number of discrete segments of their genomes. The manner and frequency with which these genetic differences arise over generational time is not well understood. Taking advantage of divergence among lineages sharing a recent common ancestry, we have conducted a genome-wide analysis of spontaneous copy number variation (CNV) in the laboratory mouse. We used high-resolution microarrays to identify 38 CNVs among 14 colonies of the C57BL/6 strain spanning approximately 967 generations of inbreeding, and we examined these loci in 12 additional strains. It is clear from our results that many CNVs arise through a highly nonrandom process: 18 of 38 were the product of recurrent mutation, and rates of change varied roughly four orders of magnitude across different loci. Recurrent CNVs are found throughout the genome, affect 43 genes and fluctuate in copy number over mere hundreds of generations, observations that raise questions about their contribution to natural variation.

Copy-Number Variants in Patients with a Strong Family History of Pancreatic Cancer.
Lucito, R. and Suresh, S. and Walter, K. and Pandey, A. and Lakshmi, B. and Krasnitz, A. and Sebat, J. and Wigler, M. H. and Klein, A. P. and Brune, K. and Palmisano, E. and Maitra, A. and Goggins, M. and Hruban, R. H. (2007) Cancer Biol Ther, 6(10) pp. 1592-9.

Copy-number variants such as germ-line deletions and amplifications are associated with inherited genetic disorders including familial cancer. The gene or genes responsible for the majority of familial clustering of pancreatic cancer have not been identified. We used representational oligonucleotide microarray analysis (ROMA) to characterize germ-line copy number variants in 60 cancer patients from 57 familial pancreatic cancer kindreds. Fifty-seven of the 60 patients had pancreatic cancer and three had nonpancreatic cancers (breast, ovary, ovary). A familial pancreatic cancer kindred was defined as a kindred in which at least two first-degree relatives have been diagnosed with pancreatic cancer. Copy-number variants identified in 607 individuals without pancreatic cancer were excluded from further analysis. A total of 56 unique genomic regions with copy-number variants not present in controls were identified, including 31 amplifications and 25 deletions. Two deleted regions were observed in two different patients, and one in three patients. The germ-line amplifications had a mean size of 662 Kb, a median size of 379 Kb (range 8.2 Kb to 2.5 Mb) and included 425 known genes. Examples of genes included in the germ-line amplifications include the MAFK, JunD and BIRC6 genes. The germ-line deletions had a mean size of 375Kb, a median size 151 Kb (range 0.4 Kb to 2.3 Mb) and included 81 known genes. In multivariate analysis controlling for region size, deletions were 90% less likely to involve a gene than were duplications (p < 0.01). Examples of genes included in the germ-line deletions include the FHIT, PDZRN3 and ANKRD3 genes. Selected deletions and amplifications were confirmed using real-time PCR, including a germ-line amplification on chromosome 19. These genetic copy-number variants define potential candidate loci for the familial pancreatic cancer gene.

A unified genetic theory for sporadic and inherited autism.
Zhao, X. and Leotta, A. and Kustanovich, V. and Lajonchere, C. and Geschwind, D. H. and Law, K. and Law, P. and Qiu, S. and Lord, C. and Sebat, J. and Ye, K. and Wigler, M. H. (2007) Proc Natl Acad Sci U S A, 104(31) pp. 12831-12836.

Autism is among the most clearly genetically determined of all cognitive-developmental disorders, with males affected more often than females. We have analyzed autism risk in multiplex families from the Autism Genetic Resource Exchange (AGRE) and find strong evidence for dominant transmission to male offspring. By incorporating generally accepted rates of autism and sibling recurrence, we find good fit for a simple genetic model in which most families fall into two types: a small minority for whom the risk of autism in male offspring is near 50%, and the vast majority for whom male offspring have a low risk. We propose an explanation that links these two types of families: sporadic autism in the low-risk families is mainly caused by spontaneous mutation with high penetrance in males and relatively poor penetrance in females; and high-risk families are from those offspring, most often females, who carry a new causative mutation but are unaffected and in turn transmit the mutation in dominant fashion to their offspring.

Strong Association of De Novo Copy Number Mutations with Autism.
Sebat, J. and Lakshmi, B. and Malhotra, D. and Troge, J. E. and Lese-Martin, C. and Walsh, T. and Yamrom, B. and Yoon, S. and Krasnitz, A. and Kendall, J. T. and Leotta, A. and Pai, D. and Zhang, R. and Lee, Y. H. and Hicks, J. B. and Spence, S. J. and Lee, A. T. and Puura, K. and Lehtimaki, T. and Ledbetter, D. and Gregersen, P. K. and Bregman, J. and Sutcliffe, J. S. and Jobanputra, V. and Chung, W. and Warburton, D. and King, M. C. and Skuse, D. and Geschwind, D. H. and Gilliam, T. C. and Ye, K. and Wigler, M. H. (2007) Science, 316(5823) pp. 445-449.

We tested the hypothesis that de novo copy number variation (CNV) is associated with autism spectrum disorders (ASDs). We performed comparative genomic hybridization (CGH) on the genomic DNA of patients and unaffected subjects to detect copy number variants not present in their respective parents. Candidate genomic regions were validated by higher-resolution CGH, paternity testing, cytogenetics, fluorescence in situ hybridization, and microsatellite genotyping. Confirmed de novo CNVs were significantly associated with autism (P = 0.0005). Such CNVs were identified in 12 out of 118 (10%) of patients with sporadic autism, in 2 out of 77 (3%) of patients with an affected first-degree relative, and in 2 out of 196 (1%) of controls. Most de novo CNVs were smaller than microscopic resolution. Affected genomic regions were highly heterogeneous and included mutations of single genes. These findings establish de novo germline mutation as a more significant risk factor for ASD than previously recognized.

Novel patterns of genome rearrangement and their association with survival in breast cancer.
Hicks, J. B. and Krasnitz, A. and Lakshmi, B. and Navin, N. E. and Riggs, M. and Leibu, E. and Esposito, D. and Alexander, J. and Troge, J. E. and Grubor, V. and Yoon, S. and Wigler, M. H. and Ye, K. and Borresen-Dale, A. L. and Naume, B. and Schlicting, E. and Norton, L. and Hagerstrom, T. and Skoog, L. and Auer, G. and Månér, S. and Lundin, P. and Zetterberg, A. (2006) Genome Research, 16(12) pp. 1465-79.

Representational Oligonucleotide Microarray Analysis (ROMA) detects genomic amplifications and deletions with boundaries defined at a resolution of approximately 50 kb. We have used this technique to examine 243 breast tumors from two separate studies for which detailed clinical data were available. The very high resolution of this technology has enabled us to identify three characteristic patterns of genomic copy number variation in diploid tumors and to measure correlations with patient survival. One of these patterns is characterized by multiple closely spaced amplicons, or "firestorms," limited to single chromosome arms. These multiple amplifications are highly correlated with aggressive disease and poor survival even when the rest of the genome is relatively quiet. Analysis of a selected subset of clinical material suggests that a simple genomic calculation, based on the number and proximity of genomic alterations, correlates with life-table estimates of the probability of overall survival in patients with primary breast cancer. Based on this sample, we generate the working hypothesis that copy number profiling might provide information useful in making clinical decisions, especially regarding the use or not of systemic therapies (hormonal therapy, chemotherapy), in the management of operable primary breast cancer with ostensibly good prognosis, for example, small, node-negative, hormone-receptor-positive diploid cases.

Identification of alterations in DNA copy number in host stromal cells during tumor progression.
Pelham, R. J. and Rodgers, L. and Hall, I. M. and Lucito, R. and Nguyen, K. C. Q. and Navin, N. E. and Hicks, J. B. and Mu, D. and Powers, S. and Wigler, M. H. and Botstein, D. (2006) Proc Natl Acad Sci U S A, 103(52) pp. 19848-19853.

The interactions between cancer cells and the surrounding host stromal tissue play a critical role in tumor progression and metastasis, but the molecular nature of this relationship remains largely uncharacterized. Furthermore, although genetic changes of neoplastic cells in tumors contribute significantly to tumor progression, it is not known whether similar changes occur in the adjacent host stromal microenvironment and whether they contribute to or inhibit tumorigenesis. To address this question in an unbiased and genome-wide manner, we applied high-resolution DNA copy number analysis to murine stromal DNA isolated from human xenograft tumors that were formed in immunodeficient mice. We show that numerous amplifications and deletions are found within the host stromal microenvironment, suggesting that alterations in host DNA copy number can occur and may play a significant role in modifying tumor–stromal interactions.

High-resolution analysis of genome copy number variation in autism.
Sebat, J. and Lakshmi, B. and Troge, J. and Martin, C. and Spence, S. and Ledbetter, D. and Gilliam, T. C. and Ye, K. and Geschwind, D. and Sutcliffe, J. and Wigler, M. H. (2006) International Journal of Developmental Neuroscience, 24(8) pp. 474.

Genome Analysis of CLL by Representational Oligonucleotide Microarray Analysis (ROMA).
Grubor, V. D. and Troge, J. E. and Meth, J. L. and Lakshmi, B. and Yamron, B. and Hufnagel, L. A. and Lee, Y. H. and Kendall, J. T. and Pai, D. and Lee, A. and Gregersen, P. and Yancopoulos, S. and Allen, S. and Rai, K. R. and Chiorazzi, N. and Wigler, M. H. and Esposito, D. (2006) ASH Annual Meeting Abstracts, 108(11) pp. 2085.

The transforming events that lead to B-cell chronic lymphocytic leukemia (B-CLL) are unknown, and although genetic abnormalities appear to promote disease progression, none of these is seen in every patient. Therefore, we have applied a highly sensitive and specific method to probe, in greater depth, for genetic changes that can occur in B-CLL, and to explore if these correlate with disease development or progression. Utilizing ROMA to measure changes in gene copy number, we screened the entire genome of over two dozen B-CLL cases to identify characteristic and novel chromosomal changes. DNA from leukemic cells of patients was compared to the normal DNA from their unaffected neutrophils and to an unrelated, normal human control. By comparing normal DNA from each patient to an unrelated normal, we avoid inadvertently mistaking common copy number variation between people as possible candidates for lesions in B-CLL. ROMA was performed on both 85,000 and 390,000 probe microarrays enabling us to screen for far smaller lesions than are detectable by routine techniques, as well as resolve regions of interest, possibly pinpointing important genes in the etiology and progression of B-CLL. Data were also obtained by using the Illumina platform. These two methods (ROMA and Illumina), while not identical, are largely in agreement. We have observed virtually all the major cytogenetic imbalances previously reported, and in many cases to a higher resolution. Although there are fewer lesions present in B-CLL than in more advanced stages of solid cancers, many early stage cancers exhibit only a few common lesions. The most frequent deletion and often the sole abnormality found in B-CLL occurs at chromosome 13q13.4. Two micro-RNAs, miR-15a and miR-16-1, have been implicated in the smallest region of this deletion. ROMA analysis has further delineated the deletions on 13q, suggesting that there may be two epicenters, one encompassing these micro-RNA genes and one encompassing a neighboring region that does not span miR-15a or miR-16-1. A novel finding is the existence in certain patients (5/26) of numerous single probe deletions or amplifications. Since we compare B-CLL to the normal DNA from the same patient, and to a known control, even single probe events can be relevant. Many B-CLLs display single probe aberrations and these events are currently under investigation. In addition, we have observed two patients with major genomic instability, as evidenced by large and multiple regions that change copy number. The analysis of large numbers of B-CLL cases is essential to identifying and understanding the ongoing progression of genetic lesions and the role they play in clinical outcome. Initial results clearly indicate the need to obtain more data to identify critically mutated genes within the leukemic cells that cause them to become more aggressive. The identification and characterization of these genetic changes may be used as stratification tools for new diagnostic and therapeutic approaches to this currently incurable disease.

PROBER: oligonucleotide FISH probe design software.
Navin, N. E. and Grubor, V. and Hicks, J. B. and Leibu, E. and Thomas, E. and Troge, J. and Riggs, M. and Lundin, P. and Månér, S. and Sebat, J. and Zetterberg, A. and Wigler, M. H. (2006) Bioinformatics, 22(19) pp. 2437-8.

PROBER is an oligonucleotide primer design software application that designs multiple primer pairs for generating PCR probes useful for fluorescence in situ hybridization (FISH). PROBER generates Tiling Oligonucleotide Probes (TOPs) by masking repetitive genomic sequences and delineating essentially unique regions that can be amplified to yield small (100-2000 bp) DNA probes that in aggregate will generate a single, strong fluorescent signal for regions as small as a single gene. TOPs are an alternative to bacterial artificial chromosomes (BACs) that are commonly used for FISH but may be unstable, unavailable, chimeric, or non-specific to small (10-100 kb) genomic regions. PROBER can be applied to any genomic locus, with the limitation that the locus must contain at least 10 kb of essentially unique blocks. To test the software, we designed a number of probes for genomic amplifications and hemizygous deletions that were initially detected by Representational Oligonucleotide Microarray Analysis of breast cancer tumors. AVAILABILITY:

Mouse genomic representational oligonucleotide microarray analysis: detection of copy number variations in normal and tumor specimens.
Lakshmi, B. and Hall, I. M. and Egan, C. and Alexander, J. and Leotta, A. and Healy, J. and Zender, L. and Spector, M. S. and Xue, W. and Lowe, S. W. and Wigler, M. H. and Lucito, R. (2006) Proc Natl Acad Sci U S A, 103(30) pp. 11234-9.

Genomic amplifications and deletions, the consequence of somatic variation, are a hallmark of human cancer. Such variation has also been observed between "normal" individuals, as well as in individuals with congenital disorders. Thus, copy number measurement is likely to be an important tool for the analysis of genetic variation, genetic disease, and cancer. We developed representational oligonucleotide microarray analysis, a high-resolution comparative genomic hybridization methodology, with this aim in mind, and reported its use in the study of humans. Here we report the development of a representational oligonucleotide microarray analysis microarray for the genomic analysis of the mouse, an important model system for many genetic diseases and cancer. This microarray was designed based on the sequence assembly MM3, and contains approximately 84,000 probes randomly distributed throughout the mouse genome. We demonstrate the use of this array to identify copy number changes in mouse cancers, as well to determine copy number variation between inbred strains of mice. Because restriction endonuclease digestion of genomic DNA is an integral component of our method, differences due to polymorphisms at the restriction enzyme cleavage sites are also observed between strains, and these can be useful to follow the inheritance of loci between crosses of different strains.

Identification and validation of oncogenes in liver cancer using an integrative oncogenomic approach.
Zender, L. and Spector, M. S. and Xue, W. and Flemming, P. and Cordon-Cardo, C. and Silke, J. and Fan, S. T. and Luk, J. M. and Wigler, M. H. and Hannon, G. J. and Mu, D. and Lucito, R. and Powers, S. and Lowe, S. W. (2006) Cell, 125(7) pp. 1253-67.

The heterogeneity and instability of human tumors hamper straightforward identification of cancer-causing mutations through genomic approaches alone. Herein we describe a mouse model of liver cancer initiated from progenitor cells harboring defined cancer-predisposing lesions. Genome-wide analyses of tumors in this mouse model and in human hepatocellular carcinomas revealed a recurrent amplification at mouse chromosome 9qA1, the syntenic region of human chromosome 11q22. Gene-expression analyses delineated cIAP1, a known inhibitor of apoptosis, and Yap, a transcription factor, as candidate oncogenes in the amplicon. In the genetic context of their amplification, both cIAP1 and Yap accelerated tumorigenesis and were required to sustain rapid growth of amplicon-containing tumors. Furthermore, cIAP1 and Yap cooperated to promote tumorigenesis. Our results establish a tractable model of liver cancer, identify two oncogenes that cooperate by virtue of their coamplification in the same genomic locus, and suggest an efficient strategy for the annotation of human cancer genes.

Validation of S. Pombe sequence assembly by microarray hybridization.
West, J. and Healy, J. and Wigler, M. H. and Casey, W. and Mishra, B. (2006) J Comput Biol, 13(1) pp. 1-20.

We describe a method to make physical maps of genomes using correlative hybridization patterns of probes to random pools of BACs. We derive thereby an estimated distance between probes, and then use this estimated distance to order probes. To test the method, we used BAC libraries from Schizzosaccharomyces pombe. We compared our data to the known sequence assembly, in order to assess accuracy. We demonstrate a small number of significant discrepancies between our method and the map derived by sequence assembly. Some of these discrepancies may arise because genome order within a population is not stable; imposing a linear order on a population may not be biologically meaningful.

Assessment of schizophrenia genomes for DNA copy number alterations.
Sachs, N. A. and Holmes, S. E. and Cowell, J. K. and Conroy, J. and Nowak, N. J. and McQuaid, D. and Rossi, M. R. and Gaile, D. P. and Christian, S. L. and Wigler, M. H. and Sebat, J. and Ross, C. A. and DeLisi, L. E. and Margolis, R. L. (2005) American Journal of Medical Genetics Part B-Neuropsychiatric Genetics, 138B(1) pp. 90-91.

Application of ROMA (representational oligonucleotide microarray analysis) to patients with cytogenetic rearrangements.
Jobanputra, V. and Sebat, J. and Troge, J. and Chung, W. and Anyane-Yeboa, K. and Wigler, M. H. and Warburton, D. (2005) Genetics in Medicine, 7(2) pp. 111-118.

Purpose: To demonstrate the accuracy and sensitivity of Representational Oligonucleotide Microarray Analysis (ROMA) to describe copy number changes in patients with chromosomal abnormalities. Methods: ROMA was performed using Bg/II digested DNA from two cases with cytogenetically detected deletions and one case with an unbalanced terminal rearrangement detected only by subtelomeric FISH. Hybridization was to an 85,000-probe oligonucleotide microarray, providing an average resolution of 35 kb. FISH was used to confirm some of the ROMA findings. Results: By ROMA, a del(13)(q14.3q21.2) was shown to be noncontiguous, with deletions extending from 53.08 to 61.40 Mb and from 72.88 to 74.83 Mb. The 10-Mb deletion contained only six known genes. FISH confirmed L-he noncontiguous nature of the deletion, as well as a small amplification in 6q that was also found in the patient's mother. A del(4)(q12q21.2) was found by ROMA to be 23 Mb in length, from 58.8 to 81.9 Mb on chromosome 4, in agreement with the cytogenetically assigned breakpoints. ROMA showed that an unbalanced "subtelomeric" rearrangement involved a 6-Mb deletion of 22q and an 8-Mb duplication of 16q. Conclusions: ROMA can define cytogenetic aberrations with extraordinary precision. Unexpected findings included the interrupted nature of the deletion in 13q and the large size of the imbalances in the "subtelomeric" rearrangement. Together with the information from the human genome sequence and proteomics, the ability to define rearrangements with "ultra-high" resolution will improve the ability to provide accurate prognosis both prenatally and postnatally to parents of offspring with chromosomal aberrations.

High-resolution ROMA CGH and FISH analysis of aneuploid and diploid breast tumors.
Hicks, J. B. and Muthuswamy, L. and Krasnitz, A. and Navin, N. E. and Riggs, M. and Grubor, V. and Esposito, D. and Alexander, J. and Troge, J. E. and Wigler, M. H. and Maner, S. and Lundin, P. and Zetterberg, A. (2005) Cold Spring Harb Symp Quant Biol, 70 pp. 51-63.

Combining representational oligonucleotide microarray analysis (ROMA) of tumor DNA with fluorescence in situ hybridization (FISH) of individual tumor cells provides the opportunity to detect and validate a wide range of amplifications, deletions, and rearrangements directly in frozen tumor samples. We have used these combined techniques to examine 101 aneuploid and diploid breast tumors for which long-term follow-up and detailed clinical information were available. We have determined that ROMA provides accurate and sensitive detection of duplications, amplifications, and deletions and yields defined boundaries for these events with a resolution of <50 kbp in most cases. We find that diploid tumors exhibit fewer rearrangements on average than aneuploids, but rearrangements occur at the same locations in both types. Diploid tumors reflect at least three consistent patterns of rearrangement. The reproducibility and frequency of these events, especially in very early stage tumors, provide insight into the earliest chromosomal events in breast cancer. We have also identified correlations between certain sets of rearrangement events and clinically relevant parameters such as long-term survival. These correlations may enable novel prognostic indicators for breast and other cancers as more samples are analyzed.

A versatile statistical analysis algorithm to detect genome copy number variation.
Daruwala, R. S. and Rudra, A. and Ostrer, H. and Lucito, R. and Wigler, M. H. and Mishra, B. (2004) Proceedings of the National Academy of Sciences of the United States of America, 101(46) pp. 16292-16297.

We have developed a versatile statistical analysis algorithm for the detection of genomic aberrations in human cancer cell lines. The algorithm analyzes genomic data obtained from a variety of array technologies, such as oligonucleotide array, bacterial artificial chromosome array, or array-based comparative genomic hybridization, that operate by hybridizing with genomic material obtained from cancer and normal cells and allow detection of regions of the genome with altered copy number. The number of probes (i.e., resolution), the amount of uncharacterized noise per probe, and the severity of chromosomal aberrations per chromosomal region may vary with the underlying technology, biological sample, and sample preparation. Constrained by these uncertainties, our algorithm aims at robustness by using a priorless maximum a posteriori estimator and at efficiency by a dynamic programming implementation. We illustrate these characteristics of our algorithm by applying it to data obtained from representational oligonucleotide microarray analysis and array-based comparative genomic hybridization technology as well as to synthetic data obtained from an artificial model whose properties can be varied computationally. The algorithm can combine data from multiple sources and thus facilitate the discovery of genes and markers important in cancer, as well as the discovery of loci important in inherited genetic disease.

Circular binary segmentation for the analysis of array-based DNA copy number data.
Olshen, A. B. and Venkatraman, E. S. and Lucito, R. and Wigler, M. H. (2004) Biostatistics, 5(4) pp. 557-572.

DNA sequence copy number is the number of copies of DNA at a region of a genome. Cancer progression often involves alterations in DNA copy number. Newly developed microarray technologies enable simultaneous measurement of copy number at thousands of sites in a genome. We have developed a modification of binary segmentation, which we call circular binary segmentation, to translate noisy intensity measurements into regions of equal copy number. The method is evaluated by simulation and is demonstrated on cell line data with known copy number alterations and on a breast cancer cell line data set.

Large-scale copy number polymorphism in the human genome.
Sebat, J. and Lakshmi, B. and Troge, J. E. and Alexander, J. and Young, J. and Lundin, P. and Maner, S. and Massa, H. and Walker, M. and Chi, M. and Navin, N. E. and Lucito, R. and Healy, J. and Hicks, J. B. and Ye, K. and Reiner, A. and Gilliam, T. C. and Trask, B. and Patterson, N. and Zetterberg, A. and Wigler, M. H. (2004) Science, 305(5683) pp. 525-8.

The extent to which large duplications and deletions contribute to human genetic variation and diversity is unknown. Here, we show that large-scale copy number polymorphisms (CNPs) (about 100 kilobases and greater) contribute substantially to genomic variation between normal humans. Representational oligonucleotide microarray analysis of 20 individuals revealed a total of 221 copy number differences representing 76 unique CNPs. On average, individuals differed by 11 CNPs, and the average length of a CNP interval was 465 kilobases. We observed copy number variation of 70 different genes within CNP intervals, including genes involved in neurological function, regulation of cell growth, regulation of metabolism, and several genes known to be associated with disease.

Distribution of short paired duplications in mammalian genomes.
Thomas, E. E. and Srebro, N. and Sebat, J. and Navin, N. and Healy, J. and Mishra, B. and Wigler, M. H. (2004) Proc Natl Acad Sci U S A, 101(28) pp. 10349-54.

Mammalian genomes are densely populated with long duplicated sequences. In this paper, we demonstrate the existence of doublets, short duplications between 25 and 100 bp, distinct from previously described repeats. Each doublet is a pair of exact matches, separated by some distance. The distribution of these intermatch distances is strikingly nonrandom. An unexpectedly high number of doublets have matches either within 100 bp (adjacent) or at distances tightly concentrated approximately 1,000 bp apart (nearby). We focus our study on these proximate doublets. First, they tend to have both matches on the same strand. By comparing nearby doublets shared in human and chimpanzee, we can also see that these doublets seem to arise by an insertion event that produces a copy without markedly affecting the surrounding sequence. Most doublets in humans are shared with chimpanzee, but many new pairs arose after the divergence of the species. Doublets found in human but not chimpanzee are most often composed of almost tandem matches, whereas older doublets (found in both species) are more likely to have matches spaced by approximately 1 kb, indicating that the nearly tandem doublets may be more dynamic. The spacing of doublets is highly conserved. So far, we have found clearly recognizable doublets in the following genomes: Homo sapiens, Mus musculus, Arabidopsis thaliana, and Caenorhabditis elegans, indicating that the mechanism generating these doublets is widespread. A mechanism that generates short local duplications while conserving polarity could have a profound impact on the evolution of regulatory and protein-coding sequences.

Annotating large genomes with exact word matches.
Healy, J. and Thomas, E. E. and Schwartz, J. T. and Wigler, M. H. (2003) Genome Research, 13(10) pp. 2306-2315.

We have developed a tool for rapidly determining the number of exact matches of any word within large, internally repetitive genomes or sets of genomes. Thus we can readily annotate any sequence, including the entire human genome, with the counts of its constituent words. We create a Burrows-Wheeler transform of the genome, which together with auxiliary data structures facilitating counting, can reside in about one gigabyte of RAM. Our original interest was motivated by oligonucleotide probe design, and we describe a general protocol for defining unique hybridization probes. But our method also has applications for the analysis of genome structure and assembly. We demonstrate the identification of chromosome-specific repeats, and outline a general procedure for finding undiscovered repeats. We also illustrate the changing contents of the human genome assemblies by comparing the annotations built from different genome freezes.

Representational oligonucleotide microarray analysis: A high-resolution method to detect genome copy number variation.
Lucito, R. and Healy, J. and Alexander, J. and Reiner, A. and Esposito, D. and Chi, M. Y. and Rodgers, L. and Brady, A. and Sebat, J. and Troge, J. E. and West, J. A. and Rostan, S. and Nguyen, K. C. Q. and Powers, S. and Ye, K. Q. and Olshen, A. and Venkatraman, E. and Norton, L. and Wigler, M. H. (2003) Genome Research, 13(10) pp. 2291-2305.

We have developed a methodology we call ROMA (representational oligonucleotide microarray analysis), for the detection of the genomic aberrations in cancer and normal humans. By arraying oligonucleoticle probes designed from the human genome sequence, and hybridizing with "representations" from cancer and normal cells, we detect regions of the genome with altered "copy number." We achieve an average resolution of 30 kb throughout the genome, and resolutions as high as a probe every 15 kb are practical. We illustrate the characteristics of probes on the array and accuracy of measurements obtained using ROMA. Using this methodology, we identify variation between cancer and normal genomes, as well as between normal human genomes. In cancer genomes, we readily detect amplifications and large and small homozygous and hemizygous deletions. Between normal human genomes, we frequently detect large (100 kb to I Mb) deletions or duplications. Many of these changes encompass known genes. ROMA will assist in the discovery of genes and markers important in cancer, and the discovery of loci that may be important in inherited predispositions to disease.

Genomic amplification and oncogenic properties of the KCNK9 potassium channel gene.
Mu, D. and Chen, L. Y. and Zhang, X. P. and See, L. H. and Koch, C. M. and Yen, C. and Tong, J. J. and Spiegel, L. and Nguyen, K. C. Q. and Servoss, A. and Peng, Y. and Pei, L. and Marks, J. R. and Lowe, S. W. and Hoey, T. and Jan, L. Y. and McCombie, W. R. and Wigler, M. H. and Powers, S. (2003) Cancer Cell, 3(3) pp. 297-302.

Representational difference analysis (RDA) of human breast cancer was used to discover a novel amplicon located at chromosomal region 8q24.3. We examined a series of breast cancer samples harboring amplification of this region and determined that KCNK9 is the sole overexpressed gene within the amplification epicenter. KCNK9 encodes a potassium channel that is amplified from 3-fold to 10-fold in 10% of breast tumors and overexpressed from 5-fold to over 100-fold in 44% of breast tumors. Overexpression of KCNK9 in cell lines promotes tumor formation and confers resistance to both hypoxia and serum deprivation, suggesting that its amplification and overexpression plays a physiologically important role in human breast cancer.

DBC2, a candidate for a tumor suppressor gene involved in breast cancer.
Hamaguchi, M. and Meth, J. L. and von Klitzing, C. and Wei, W. and Esposito, D. and Rodgers, L. and Walsh, T. and Welcsh, P. and King, M. C. and Wigler, M. H. (2002) Proceedings of the National Academy of Sciences of the United States of America, 99(21) pp. 13647-13652.

Genetics. Wild by nature.
Wigler, M. H. and Mishra, B. (2002) Science, 296(5572) pp. 1407-1408.

How do networks of biological molecules, such as signal transduction pathways, operate, and are their behaviors predictable? In a Perspective, Wigler and Mishra explain a new study ( Guet et al.) where the authors have designed a genetic combinatorial network and examined the predictability of its behavior.

The Ras-Byr2RBD complex: Structural basis for Ras effector recognition in yeast.
Scheffzek, K. and Grunewald, P. and Wohlgemuth, S. and Kabsch, W. and Tu, H. and Wigler, M. H. and Wittinghofer, A. and Herrmann, C. (2001) Structure, 9(11) pp. 1043-1050.

Background-The small GTP binding protein Ras has important roles in cellular growth and differentiation. Mutant Ras is permanently active and contributes to cancer development. In its activated form, Ras interacts with effector proteins, frequently initiating a kinase cascade. In the lower eukaryotic Schizosaccharomyces pombe, Byr2 kinase represents a Ras target that In terms of signal-transduction hierarchy can be considered a homolog of mammalian Raf-kinase. The activation mechanism of protein kinases by Ras is not understood, and there is no detailed structural information about Ras binding domains (RBDs) in nonmammalian organisms. Results: The crystal structure of the Ras-Byr2RBD complex at 3 Angstrom resolution shows a complex architecture similar to that observed in mammalian homologous systems, with an interprotein beta sheet stabilized by predominantly polar interactions between the interacting components. The C-terminal half of the Ras switch I region contains most of the contact anchors, while on the Byr2 side, a number of residues from topologically distinct regions are Involved in complex stabilization. A C-terminal helical segment, which is not present in the known mammalian homologous systems and which is part of the auto-inhibitory region, has an additional binding site outside the switch I region. Conclusions: The structure of the Ras-Byr2 complex confirms the Ras binding module as a communication element mediating Ras-effector interactions; the Ras-Byr2 complex is also conserved in a lower eukaryotic system like yeast, which is in contrast to other small GTPase families. The extra helical segment might be involved in kinase activation.

Design of a retroviral-mediated ecdysone-inducible system and its application to the expression profiling of the PTEN tumor suppressor.
Stolarov, J. and Chang, K. and Reiner, A. and Rodgers, L. and Hannon, G. J. and Wigler, M. H. and Mittal, V. (2001) Proceedings of the National Academy of Sciences of the United States of America, 98(23) pp. 13043-13048.

We have engineered the ecdysone-inducible mammalian expression system for general retroviral delivery to cultured mammalian cells. We inducibly expressed PTEN in the glioblastoma cell line, U87MG, lacking this gene. Because nearly all cells are recruited on induction, we find both up- and down-regulated genes by cDNA microarray analysis. The changes we see are similar to those observed after treatment with LY294002, an inhibitor of phosphatidylinositol 3-OH kinase, fully consistent with the model that PTEN antagonizes phosphatidylinositol 3-OH kinase. Both treatments result in suppressed expression of the transforming growth factor (TGF)-beta gene and the genes of the cholesterol biosynthesis pathway. Our results illustrate the power of using a fully inducible expression system in conjunction with cDNA microarray analysis for exploring gene function.

PTEN controls tumor-induced angiogenesis.
Wen, S. H. and Stolarov, J. and Myers, M. P. and Su, J. D. and Wigler, M. H. and Tonks, N. K. and Durden, D. L. (2001) Proceedings of the National Academy of Sciences of the United States of America, 98(8) pp. 4622-4627.

Mutations of the tumor suppressor PTEN, a phosphatase with specificity for 3-phosphorylated inositol phospholipids, accompany progression of brain tumors from benign to the most malignant forms. Tumor progression, particularly in aggressive and malignant tumors, is associated with the induction of angiogenesis, a process termed the angiogenic switch. Therefore, we tested whether PTEN regulates tumor progression by modulating angiogenesis. U87MG glioma cells stably reconstituted with PTEN cDNA were tested for growth in a nude mouse orthotopic brain tumor model. We observed that the reconstitution of wild-type PTEN had no effect on in vitro proliferation but dramatically decreased tumor growth in vivo and prolonged survival in mice implanted intracranially with these tumor cells. PTEN reconstitution diminished phosphorylation of AKT within the PTEN-reconstituted tumor, induced thrombospondin 1 expression, and suppressed angiogenic activity. These effects were not observed in tumors reconstituted with a lipid phosphatase inactive G129E mutant of PTEN, a result that provides evidence that the lipid phosphatase activity of PTEN regulates the angiogenic response in vivo. These data provide evidence that PTEN regulates tumor-induced angiogenesis and the progression of gliomas to a malignant phenotype via the regulation of phosphoinositide-dependent signals.

Detecting gene copy number fluctuations in tumor cells by microarray analysis of genomic representations.
Lucito, R. and West, J. and Reiner, A. and Alexander, J. and Esposito, D. and Mishra, B. and Powers, S. and Norton, L. and Wigler, M. H. (2000) Genome Research, 10(11) pp. 1726-1736.

In this work, we explore the use of representations in conjunction with DNA microarray technology to measure gene copy number changes in cancer. We demonstrate that arrays of DNA probes derived From low-complexity representations can be used to detect amplifications, deletions, and polymorphic differences when hybridized to representations of genomic DNA. The method is both reproducible and verifiable, and is applicable even to microscopic amounts of primary tumors. We also present a mathematical model for array performance that is useful for designing and understanding DNA microarray hybridization protocols. The future applications and challenges of this approach are discussed.

Molecular forceps from combinatorial libraries prevent the farnesylation of Ras by binding to its carboxyl terminus.
Dong, D. L. and Liu, R. and Sherlock, R. and Wigler, M. H. and Nestler, H. P. (1999) Chem Biol, 6(3) pp. 133-41.

INTRODUCTION: Ras is one of the major oncogenes. In order to function properly it has to undergo post-translational processing at its carboxyl terminus. It has been shown that inhibitors of farnesyl transferase, the first enzyme in the processing chain, can suppress the transforming activity of oncogenic Ras. RESULTS: We have identified molecular forceps, branched peptidic molecules, from combinatorial libraries that bind to the carboxyl terminus of Ras and interfere with its farnesylation without inhibiting the farnesyl transferase. The active molecules were selected by a screening against the carboxy-terminal octapeptide of Ras. CONCLUSIONS: The implications of our findings are twofold. First, we demonstrate that it is possible to prevent enzymatic transformations by blocking the enzyme's access to its substrate using a synthetic small molecule to mask the substrate. Second, we show that it is feasible to derive molecules from combinatorial libraries that bind a specific epitope on a protein by selecting these molecules with the isolated peptide epitope.

Genetic evidence for Pak1 autoinhibition and its release by Cdc42.
Tu, H. and Wigler, M. H. (1999) Mol Cell Biol, 19(1) pp. 602-11.

Pak1 protein kinase of Schizosaccharomyces pombe, a member of the p21-GTPase-activated protein kinase (PAK) family, participates in signaling pathways including sexual differentiation and morphogenesis. The regulatory domain of PAK proteins is thought to inhibit the kinase catalytic domain, as truncation of this region renders kinases more active. Here we report the detection in the two-hybrid system of the interaction between Pak1 regulatory domain and the kinase catalytic domain. Pak1 catalytic domain binds to the same highly conserved region on the regulatory domain that binds Cdc42, a GTPase protein capable of activating Pak1. Two-hybrid, mutant, and genetic analyses indicated that this intramolecular interaction rendered the kinase in a closed and inactive configuration. We show that Cdc42 can induce an open configuration of Pak1. We propose that Cdc42 interaction disrupts the intramolecular interactions of Pak1, thereby releasing the kinase from autoinhibition.

The lipid phosphatase activity of PTEN is critical for its tumor supressor function.
Myers, M. P. and Pass, I. and Batty, I. H. and Van der Kaay, J. and Stolarov, J. P. and Hemmings, B. A. and Wigler, M. H. and Downes, C. P. and Tonks, N. K. (1998) Proc Natl Acad Sci U S A, 95(23) pp. 13513-8.

Since their discovery, protein tyrosine phosphatases have been speculated to play a role in tumor suppression because of their ability to antagonize the growth-promoting protein tyrosine kinases. Recently, a tumor suppressor from human chromosome 10q23, called PTEN or MMAC1, has been identified that shares homology with the protein tyrosine phosphatase family. Germ-line mutations in PTEN give rise to several related neoplastic disorders, including Cowden disease. A key step in understanding the function of PTEN as a tumor suppressor is to identify its physiological substrates. Here we report that a missense mutation in PTEN, PTEN-G129E, which is observed in two Cowden disease kindreds, specifically ablates the ability of PTEN to recognize inositol phospholipids as a substrate, suggesting that loss of the lipid phosphatase activity is responsible for the etiology of the disease. Furthermore, expression of wild-type or substrate-trapping forms of PTEN in HEK293 cells altered the levels of the phospholipid products of phosphatidylinositol 3-kinase and ectopic expression of the phosphatase in PTEN-deficient tumor cell lines resulted in the inhibition of protein kinase (PK) B/Akt and regulation of cell survival.

Signaling pathways in Ras-mediated tumorigenicity and metastasis.
Webb, C. P. and Van Aelst, L. and Wigler, M. H. and Woude, G. F. (1998) Proc Natl Acad Sci U S A, 95(15) pp. 8773-8.

The effector domain mutants of oncogenic Ras, V12S35 Ras, V12G37 Ras, and V12C40 Ras were tested for their abilities to mediate tumorigenic and metastatic phenotypes in athymic nude mice when expressed in NIH 3T3 fibroblasts. All mutants displayed comparable tumorigenic properties, but only the mutant that activates the Raf-mitogen-activated protein kinase kinase (MEK)-extracellular regulated kinase (ERK) 1/2 pathway, V12S35 Ras, induced tumors in the experimental metastasis assay. Furthermore, direct activation of the MEK-ERK1/2 pathway in NIH 3T3 cells by mos or a constitutively active form of MEK was sufficient to induce metastasis whereas R-Ras, which fails to activate the ERK1/2 pathway, is tumorigenic but nonmetastatic. The subcutaneous tumors and lung metastases derived from V12S35 Ras-transformed NIH 3T3 cells expressed higher levels of activated ERK1/2 in culture when compared with the parental cellular pool before injection, indicating that selection for cells with higher levels of activated ERK1/2 occurred during tumor growth and metastasis. By contrast, cells explanted from V12G37-Ras or V12C40-Ras-induced tumors did not show changes in the level of ERK1/2 activation when compared with the parental cells. When tumor-explanted cell lines derived from each of the effector domain mutants were passaged one additional time in vivo, all mediated rapid tumor growth, but, again, only cells derived from V12S35 Ras-tumors formed numerous metastatic lesions within the lung. These results show that the metastatic properties of the Ras effector domain mutants segregate, and that, whereas Ras-mediated tumorigenicity can arise independently of ERK1/2 activation, experimental metastasis appears to require constitutive activation of the ERK1/2 pathway.

Genetic analysis using genomic representations.
Lucito, R. and Nakimura, M. and West, J. A. and Han, Y. and Chin, K. and Jensen, K. R. and McCombie, R. W. and Gray, J. W. and Wigler, M. H. (1998) Proceedings of the National Academy of Sciences of the United States of America, 95(8) pp. 4487-92.

Analysis of the genetic changes in human tumors is often problematical because of the presence of normal stroma and the limited availability of pure tumor DNA. However, large amounts of highly reproducible "representations" of tumor and normal genomes can be made by PCR from nanogram amounts of restriction endonuclease cleaved DNA that has been ligated to oligonucleotide adaptors. We show here that representations are useful for many types of genetic analyses, including measuring relative gene copy number, loss of heterozygosity, and comparative genomic hybridization. Representations may be prepared even from sorted nuclei from fixed and archived tumor biopsies.

Rapid isolation of cDNA by hybridization.
Hamaguchi, M. and O'Connor, E. A. and Chen, T. and Parnell, L. and McCombie, R. W. and Wigler, M. H. (1998) Proceedings of the National Academy of Sciences of the United States of America, 95(7) pp. 3764-9.

The isolation of genes from a given genomic region can be a rate-limiting step in the discovery of disease genes. We describe an approach to the isolation of cDNAs that have sequences in common with large genomic clones such as bacterial artificial chromosomes. We applied this method to loci both amplified and deleted in cancer, illustrating its usage in the identification of both oncogenes and tumor suppressor genes, respectively. The method, called rapid isolation of cDNAs by hybridization (RICH), depends on solution hybridization, enzymatic modification, and amplification/selection of sequences present in both cDNA populations and the genomic clones. The method should facilitate the development of transcription maps for large genomic clones, possibly even yeast artificial chromosomes.

Multiple regulatory domains on the Byr2 protein kinase.
Tu, H. and Barr, M. and Dong, D. L. and Wigler, M. H. (1997) Mol Cell Biol, 17(10) pp. 5876-87.

Byr2 protein kinase, a homolog of mammalian mitogen-activated protein kinase/extracellular signal-regulated kinase kinase (MEKK) and Saccharomyces cerevisiae STE11, is required for pheromone-induced sexual differentiation in the fission yeast Schizosaccharomyces pombe. Byr2 functions downstream of Ste4, Ras1, and the membrane-associated receptor-coupled heterotrimeric G-protein alpha subunit, Gpa1. Byr2 has a distinctive N-terminal kinase regulatory domain and a characteristic C-terminal kinase catalytic domain. Ste4 and Ras1 interact with the regulatory domain of Byr2 directly. Here, we define the domains of Byr2 that bind Ste4 and Ras1 and show that the Byr2 regulatory domain binds to the catalytic domain in the two-hybrid system. Using Byr2 mutants, we demonstrate that these direct physical interactions are all required for proper signaling. In particular, the physical association between Byr2 regulatory and catalytic domains appears to result in autoinhibition, the loss of which results in kinase activation. Furthermore, we provide evidence that Shk1, the S. pombe homolog of the STE20 protein kinase, can directly antagonize the Byr2 intramolecular interaction, possibly by phosphorylating Byr2.

P-TEN, the tumor suppressor from human chromosome 10q23, is a dual-specificity phosphatase.
Myers, M. P. and Stolarov, J. P. and Eng, C. and Li, J. and Wang, S. I. and Wigler, M. H. and Parsons, R. and Tonks, N. K. (1997) Proc Natl Acad Sci U S A, 94(17) pp. 9052-7.

Protein tyrosine phosphatases (PTPs) have long been thought to play a role in tumor suppression due to their ability to antagonize the growth promoting protein tyrosine kinases. Recently, a candidate tumor suppressor from 10q23, termed P-TEN, was isolated, and sequence homology was demonstrated with members of the PTP family, as well as the cytoskeletal protein tensin. Here we show that recombinant P-TEN dephosphorylated protein and peptide substrates phosphorylated on serine, threonine, and tyrosine residues, indicating that P-TEN is a dual-specificity phosphatase. In addition, P-TEN exhibited a high degree of substrate specificity, showing selectivity for extremely acidic substrates in vitro. Furthermore, we demonstrate that mutations in P-TEN, identified from primary tumors, tumor cells lines, and a patient with Bannayan-Zonana syndrome, resulted in the ablation of phosphatase activity, demonstrating that enzymatic activity of P-TEN is necessary for its ability to function as a tumor suppressor.

PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer.
Li, J. and Yen, C. and Liaw, D. and Podsypanina, K. and Bose, S. and Wang, S. I. and Puc, J. and Miliaresis, C. and Rodgers, L. and McCombie, R. and Bigner, S. H. and Giovanella, B. C. and Ittmann, M. and Tycko, B. and Hibshoosh, H. and Wigler, M. H. and Parsons, R. (1997) Science, 275(5308) pp. 1943-7.

Mapping of homozygous deletions on human chromosome 10q23 has led to the isolation of a candidate tumor suppressor gene, PTEN, that appears to be mutated at considerable frequency in human cancers. In preliminary screens, mutations of PTEN were detected in 31% (13/42) of glioblastoma cell lines and xenografts, 100% (4/4) of prostate cancer cell lines, 6% (4/65) of breast cancer cell lines and xenografts, and 17% (3/18) of primary glioblastomas. The predicted PTEN product has a protein tyrosine phosphatase domain and extensive homology to tensin, a protein that interacts with actin filaments at focal adhesions. These homologies suggest that PTEN may suppress tumor cell growth by antagonizing protein tyrosine kinases and may regulate tumor cell invasion and metastasis through interactions at focal adhesions.

Identification of Ste4 as a potential regulator of Byr2 in the sexual response pathway of Schizosaccharomyces pombe.
Barr, M. M. and Tu, H. and Van Aelst, L. and Wigler, M. H. (1996) Mol Cell Biol, 16(10) pp. 5597-603.

A conserved MAP kinase cascade is central to signal transduction in both simple and complex eukaryotes. In the yeast Schizosaccharomyces pombe, Byr2, a homolog of mammalian MAPK/ERK kinase kinase and Saccharomyces cerevisiae STE11, is required for pheromone-induced sexual differentiation. A screen for S. pombe proteins that interact with Byr2 in a two-hybrid system led to the isolation of Ste4, a protein that is known to be required for sexual function. Ste4 binds to the regulatory region of Byr2. This binding site is separable from the binding site for Ras1. Both Ste4 and Ras1 act upstream of Byr2 and act at least partially independently. Ste4 contains a leucine zipper and is capable of homotypic interaction. Ste4 has regions of homology with STE50, an S. cerevisiae protein required for sexual differentiation that we show can bind to STE11.

A role for the Ral guanine nucleotide dissociation stimulator in mediating Ras-induced transformation.
White, M. A. and Vale, T. and Camonis, J. H. and Schaefer, E. and Wigler, M. H. (1996) J Biol Chem, 271(28) pp. 16439-42.

Oncogenic Ras transforms cells through the activation of multiple downstream pathways mediated by separate effector molecules, one of which is Raf. Here we report the identification of a second ras-binding protein that can induce cellular transformation in parallel with activation of the Raf/mitogen-activated protein kinase cascade. The Ral guanine nucleotide dissociation stimulator (RalGDS) was isolated from a screen for Ras-binding proteins that specifically interact with a Ras effector-loop mutant, ras(12V,37G), that uncouples Ras from activation of Raf1. RalGDS, like ras(12V, 37G), cooperates synergistically with mutationally activated Raf to induce foci of growth and morphologically transformed NIH 3T3 cells. RalGDS does not significantly enhance MAP kinase activation by activated Raf, suggesting that the cooperativity in focus formation is due to a distinct pathway acting downstream of Ras and parallel to Raf.

Evaluation of the FHIT gene in colorectal cancers.
Thiagalingam, S. and Lisitsyn, N. A. and Hamaguchi, M. and Wigler, M. H. and Willson, J. K. and Markowitz, S. D. and Leach, F. S. and Kinzler, K. W. and Vogelstein, B. (1996) Cancer Res, 56(13) pp. 2936-9.

A variety of studies suggests that tumor suppressor loci on chromosome 3p are important in various forms of human neoplasia. Recently, a chromosome 3p14.2 gene called FHIT was discovered and proposed as a candidate tumor suppressor gene in colorectal and other cancers. We evaluated the FHIT gene in a panel of colorectal cancer cell lines and xenografts, which allowed a comprehensive mutational analysis. A transcript containing the complete coding sequence was found to be expressed at robust levels in 29 of 31 cancers tested. The complete sequence of the coding region of the gene was determined and found to be normal in all 29 of these cases. These studies suggest either that FHIT is inactivated by an unusual mechanism or that it plays a role in relatively few colorectal tumors.

Oncogenic Ras activation of Raf/mitogen-activated protein kinase-independent pathways is sufficient to cause tumorigenic transformation.
Khosravi-Far, R. and White, M. A. and Westwick, J. K. and Solski, P. A. and Chrzanowska-Wodnicka, M. and Van Aelst, L. and Wigler, M. H. and Der, C. J. (1996) Mol Cell Biol, 16(7) pp. 3923-33.

Substantial evidence supports a critical role for the activation of the Raf-1/MEK/mitogen-activated protein kinase pathway in oncogenic Ras-mediated transformation. For example, dominant negative mutants of Raf-1, MEK, and mitogen-activated protein kinase all inhibit Ras transformation. Furthermore, the observation that plasma membrane-localized Raf-1 exhibits the same transforming potency as oncogenic Ras suggests that Raf-1 activation alone is sufficient to mediate full Ras transforming activity. However, the recent identification of other candidate Ras effectors (e.g., RalGDS and phosphatidylinositol-3 kinase) suggests that activation of other downstream effector-mediated signaling pathways may also mediate Ras transforming activity. In support of this, two H-Ras effector domain mutants, H-Ras(12V, 37G) and H-Ras(12V, 40C), which are defective for Raf binding and activation, induced potent tumorigenic transformation of some strains of NIH 3T3 fibroblasts. These Raf-binding defective mutants of H-Ras induced a transformed morphology that was indistinguishable from that induced by activated members of Rho family proteins. Furthermore, the transforming activities of both of these mutants were synergistically enhanced by activated Raf-1 and inhibited by the dominant negative RhoA(19N) mutant, indicating that Ras may cause transformation that occurs via coordinate activation of Raf-dependent and -independent pathways that involves Rho family proteins. Finally, cotransfection of H-Ras(12V, 37G) and H-Ras(12V, 40C) resulted in synergistic cooperation of their focus-forming activities, indicating that Ras activates at least two Raf-independent, Ras effector-mediated signaling events.

Stimulation of membrane ruffling and MAP kinase activation by distinct effectors of RAS.
Joneson, T. and White, M. A. and Wigler, M. H. and Bar-Sagi, D. (1996) Science, 271(5250) pp. 810-812.

The RAS guanine nucleotide binding proteins activate multiple signaling events that regulate cell growth and differentiation. In quiescent fibroblasts, ectopic expression of activated H-RAS (H-RAS(V12), where V12 indicates valine-12) induces membrane ruffling, mitogen-activated protein (MAP) kinase activation, and stimulation of DNA synthesis. A mutant of activated H-RAS, H-RAS(V12C40) (where C40 indicates cysteine-40), was identified that was defective for MAP kinase activation and stimulation of DNA synthesis, but retained the ability to induce membrane ruffling. Another mutant of activated H-RAS, H-RAS(V12S35) (where S35 indicates serine-35), which activates MAP kinase, was defective for stimulation of membrane ruffling and induction of DNA synthesis. Expression of both mutants resulted in a stimulation of DNA synthesis that was comparable to that induced by H-RAS(V12). These results indicate that membrane ruffling and activation of MAP kinase represent distinct RAS effector pathways and that input from both pathways is required for the mitogenic activity of RAS.

New polymorphic markers in the vicinity of the pearl locus on mouse chromosome 13.
Xu, H. P. and Yanak, B. L. and Wigler, M. H. and Gorin, M. B. (1996) Mamm Genome, 7(1) pp. 16-9.

We have used a Mus domesticus/-Mus spretus congenic animal that was selected for retention of Mus spretus DNA around the pearl locus to create a highly polymorphic region suitable for screening new markers. Representation difference analysis (RDA) was performed with either DNA from the congenic animal or C57BL/6J as the driver for subtraction. Four clones were identified, characterized, and converted to PCR-based polymorphic markers. Three of the four markers equally subdivide a 10-cM interval containing the pearl locus, with the fourth located centromeric to it. These markers have been placed on the mouse genetic map by use of an interspecific backcross panel between Mus domesticus (C57BL/6J) and Mus spretus generated by The Jackson Laboratory.

Shk1, a homolog of the Saccharomyces cerevisiae Ste20 and mammalian p65PAK protein kinases, is a component of a Ras/Cdc42 signaling module in the fission yeast Schizosaccharomyces pombe.
Marcus, S. and Polverino, A. and Chang, E. and Robbins, D. and Cobb, M. H. and Wigler, M. H. (1995) Proc Natl Acad Sci U S A, 92(13) pp. 6180-4.

We describe a protein kinase, Shk1, from the fission yeast Schizosaccharomyces pombe, which is structurally related to the Saccharomyces cerevisiae Ste20 and mammalian p65PAK protein kinases. We provide genetic evidence for physical and functional interaction between Shk1 and the Cdc42 GTP-binding protein required for normal cell morphology and mating in S. pombe. We further show that expression of the STE20 gene complements the shk1 null mutation and that Shk1 is capable of signaling to the pheromone-responsive mitogen-activated protein kinase cascade in S. cerevisiae. Our results lead us to propose that signaling modules composed of small GTP-binding proteins and protein kinases related to Shk1, Ste20, and p65PAK, are highly conserved in evolution and participate in both cytoskeletal functions and mitogen-activated protein kinase signaling pathways.

Mutations in the SHR5 gene of Saccharomyces cerevisiae suppress Ras function and block membrane attachment and palmitoylation of Ras proteins.
Jung, V. and Chen, L. and Hofmann, S. L. and Wigler, M. H. and Powers, S. (1995) Mol Cell Biol, 15(3) pp. 1333-42.

We have identified a gene, SHR5, in a screen for extragenic suppressors of the hyperactive RAS2Val-19 mutation in the budding yeast Saccharomyces cerevisiae. SHR5 was cloned, sequenced, and found to encode a 23-kDa protein not significantly homologous to other proteins in the current data bases. Genetic evidence arguing that Shr5 operates at the level of Ras is presented. We tested whether SHR5, like previously isolated suppressors of hyperactivated RAS2, acts by affecting the membrane attachment and/or posttranslational modification of Ras proteins. We found that less Ras protein is attached to the membrane in shr5 mutants than in wild-type cells and that the Ras proteins are markedly underpalmitoylated, suggesting that Shr5 is involved in palmitoylation of Ras proteins. However, shr5null mutants exhibit normal palmitoyltransferase activity measured in vitro. Further, shr5null mutations attenuate Ras function in cells containing mutant Ras2 proteins that are not palmitoylated or farnesylated. We conclude that SHR5 encodes a protein that participates in the membrane localization of Ras but also interacts in vivo with completely unprocessed and cytosolic Ras proteins.

Multiple Ras functions can contribute to mammalian cell transformation.
White, M. A. and Nicolette, C. and Minden, A. and Polverino, A. and Van Aelst, L. and Karin, M. and Wigler, M. H. (1995) Cell, 80(4) pp. 533-41.

We have developed a generalized approach, using two hybrid interactions, to isolate Ha-Ras effector loop mutations that separate the ability of Ha-Ras to interact with different downstream effectors. These mutations attenuate or eliminate Ha-ras(G12V) transformation of mammalian cells, but retain complementary activity, as demonstrated by synergistic induction of foci of growth-transformed cells, and by the ability to activate different downstream components. The transformation defect of Ha-ras(G12V, E37G) is rescued by a mutant, raf1, that restores interaction. These results indicate that multiple cellular components, including Raf1, are activated by Ha-Ras and contribute to Ha-Ras-induced mammalian cell transformation.

Comparative genomic analysis of tumors: detection of DNA losses and amplification.
Lisitsyn, N. A. and Lisitsina, N. M. and Dalbagni, G. and Barker, P. and Sanchez, C. A. and Gnarra, J. and Linehan, W. M. and Reid, B. J. and Wigler, M. H. (1995) Proc Natl Acad Sci U S A, 92(1) pp. 151-5.

We demonstrate the use of representational difference analysis for cloning probes that detect DNA loss and amplification in tumors. Using DNA isolated from human tumor cell lines to drive hybridization against matched normal DNA, we were able to identify six genomic regions that are homozygously deleted in cultured cancer cells. When this method was applied in the reverse way, using normal DNA to drive hybridization against tumor cell DNA, we readily isolated probes detecting amplification. Representational difference analysis was also performed on DNAs derived from tumor biopsies, and we thereby discovered a probe detecting very frequent homozygous loss in colon cancer cell lines and located on chromosome 3p.

Representational difference analysis in detection of genetic lesions in cancer.
Lisitsyn, N. and Wigler, M. H. (1995) Methods Enzymol, 254 pp. 291-304.

This chapter presents a global approach to the analysis of the sequence differences between the genomes of the cancer and normal cells that uses a method called "representational difference analysis (RDA)." In principle, RDA can be used to derive probes for genomic losses, rearrangements, amplifications, point mutations, and pathogenic organisms found within the cancer cell. RDA belongs to the general class of DNA subtractive methodologies. These methodologies all have in common that one DNA population (the "driver") is hybridized in excess against a second population (the "tester") to remove common hybridizing sequences, thereby enriching "target" sequences unique to the tester. RDA is a powerful, versatile, but complex procedure. To perform it properly requires not only diligence at the bench but also a clear headed understanding of how to use it. This chapter addresses both of these aspects. The chapter also discusses the application of RDA and provides detailed protocols. Pure subtractive methodologies have limited usefulness in the analysis of complex genomes. This is because the enrichment required to purify target sequences is very high and because the sequence complexity of DNA from higher organisms is too great for single copy sequences to hybridize sufficiently to completion. RDA combines subtractive enrichment with two further elements: kinetic enrichment and representation. The RDA procedure consists of the preparation of amplicons followed by multiple cycles of hybridization/amplification. Because RDA is a complex procedure, the inclusion of controls is controlled. Care and thought must be given to the source of DNA for analysis, and this is also discussed.

Cooperative interaction of S. pombe proteins required for mating and morphogenesis.
Chang, E. C. and Barr, M. and Wang, Y. and Jung, V. and Xu, H. P. and Wigler, M. H. (1994) Cell, 79(1) pp. 131-41.

We isolated two S. pombe genes, scd1 and scd2, that are required for normal morphology and mating. scd1 and scd2 are homologous to CDC24 and BEM1, respectively, of S. cerevisiae. Epistasis analyses indicate that scd2 and ras1 converge upon scd1, which, in turn, interacts with cdc42sp, a RHO-like GTPase. Studies with the yeast two-hybrid system indicate that scd2 forms complexes with both scd1 and cdc42sp. Furthermore, biochemical studies indicate that the interaction between scd1 and scd2 is direct. The yeast two-hybrid data further suggest that scd1, scd2, cdc42sp, and ras1, in its GTP-bound state, act cooperatively to form a protein complex.

Complexes between STE5 and components of the pheromone-responsive mitogen-activated protein kinase module.
Marcus, S. and Polverino, A. and Barr, M. and Wigler, M. H. (1994) Proc Natl Acad Sci U S A, 91(16) pp. 7762-6.

We present genetic evidence for complex formation of STE5 and the STE11, STE7, and FUS3 protein kinases, the pheromone-responsive mitogen-activated protein kinase module of Saccharomyces cerevisiae. Interaction between STE5 and STE11 is not dependent on STE7, and interaction between STE5 and STE7 does not require STE11. The N-terminal regulatory domain of STE11 is both necessary and sufficient for interaction with STE5. Interaction between STE7 and STE11 is bridged by STE5, suggesting the formation of a multiprotein complex. We also demonstrate biochemical interaction between STE5 and STE11 by using a combination of bacterially expressed fusion proteins and extracts prepared from yeast. Our results suggest that STE5 is a scaffolding protein that facilitates interactions between components of the pheromone-responsive mitogen-activated protein kinase module. We further propose that such scaffolding proteins serve to inhibit cross-talk between functionally unrelated mitogen-activated protein kinase modules within the same cell.

Two types of RAS mutants that dominantly interfere with activators of RAS.
Jung, V. and Wei, W. and Ballester, R. and Camonis, J. and Mi, S. and Van Aelst, L. and Wigler, M. H. and Broek, D. (1994) Mol Cell Biol, 14(6) pp. 3707-18.

In the fission yeast Schizosaccharomyces pombe, ras1 regulates both sexual development (conjugation and sporulation) and cellular morphology. Two types of dominant interfering mutants were isolated in a genetic screen for ras1 mutants that blocked sexual development. The first type of mutation, at Ser-22, analogous to the H-rasAsn-17 mutant (L. A. Feig and G. M. Cooper, Mol. Cell. Biol. 8:3235-3243, 1988), blocked only conjugation, whereas a second type of mutation, at Asp-62, interfered with conjugation, sporulation, and cellular morphology. Analogous mutations at position 64 of Saccharomyces cerevisiae RAS2 or position 57 of human H-ras also resulted in dominant interfering mutants that interfered specifically and more profoundly than mutants of the first type with RAS-associated pathways in both S. pombe or S. cerevisiae. Genetic evidence indicating that both types of interfering mutants function upstream of RAS is provided. Biochemical evidence showing that the mutants are altered in their interaction with the CDC25 class of exchange factors is presented. We show that both H-rasAsn-17 and H-rasTyr-57, compared with wild-type H-ras, are defective in their guanine nucleotide-dependent release from human cdc25 and that this defect is more severe for the H-rasTyr-57 mutant. Such a defect would allow the interfering mutants to remain bound to, thereby sequestering RAS exchange factors. The more severe interference phenotype of this novel interfering mutant suggests that it functions by titrating out other positive regulators of RAS besides those encoded by ste6 and CDC25.

Conservation of Ste20-Related Protein-Kinases in Eukaryotes.
Marcus, S. and Polverino, A. and Robbins, D. and Cobb, M. and Wigler, M. H. (1994) Faseb Journal, 8(7) pp. A1431-A1431.

Direct isolation of polymorphic markers linked to a trait by genetically directed representational difference analysis.
Lisitsyn, N. A. and Segre, J. A. and Kusumi, K. and Lisitsyn, N. M. and Nadeau, J. H. and Frankel, W. N. and Wigler, M. H. and Lander, E. S. (1994) Nat Genet, 6(1) pp. 57-63.

We describe a technique, genetically directed representational difference analysis (GDRDA), for specifically generating genetic markers linked to a trait of interest. GDRDA is applicable, in principle, to virtually any organism, because it requires neither prior knowledge of the chromosomal location of the gene controlling the trait nor the availability of a pre-existing genetic map. Based on a subtraction technique described recently called representational difference analysis, GDRDA uses the principles of transmission genetics to create appropriate Tester and Driver samples for subtraction. We demonstrate the usefulness of GDRDA by, for example, successfully targeting three polymorphisms to an interval of less than 1 cM of the mouse nude locus of chromosome 11.

Concerted action of RAS and G proteins in the sexual response pathways of Schizosaccharomyces pombe.
Xu, H. P. and White, M. and Marcus, S. and Wigler, M. H. (1994) Mol Cell Biol, 14(1) pp. 50-8.

We have shown that the expression of mam2, the gene encoding the Schizosaccharomyces pombe P-factor pheromone receptor, is dependent upon components of the pheromone signal transduction pathway, including Ras1, Gpa1, Byr1 and Byr2, each of which is required for both conjugation and sporulation. Studies of the expression of mam2 in mutant S. pombe cells confirm previous conclusions, based on the ability of cells to sporulate, that the Byr1 protein kinase acts downstream of the Byr2 protein kinase and that both act downstream of Ras1, the S. pombe RAS homolog, and Gpa1, the G alpha component that mediates the occupancy of the mam2 receptor. In addition, our present studies show that Ras1 and Gpa1 each act downstream from the other and hence act in concert. The Spk1 kinase, which is required for conjugation and sporulation and which is a structural and functional homolog of the vertebrate MAP kinases, is not required for mam2 expression.

Detection of genetic loss in tumors by representational difference analysis.
Lisitsyn, N. A. and Leach, F. S. and Vogelstein, B. and Wigler, M. H. (1994) Cold Spring Harbor Symposia on Quantitative Biology, 59 pp. 585-7.

A variety of genetic lesions are found in tumors, including DNA losses, point mutations, gene amplifications, and rearrangements (Lasko et al. 1991; Salomon et al. 1991). Frequent losses of both alleles at a given locus or losses of one allele with functional inactivation of the other have been detected in many tumor types. These genetic lesions, manifesting themselves as loss of heterozygosity (LOH) and hemizygous and homozygous deletions, have been found to be the hallmarks of the presence of tumor suppressor genes. Many approaches have been taken in the past to identify these genes, but recently we have developed a new method that is both general and efficient (Lisitsyn et al. 1993, 1995). The method, called representational difference analysis, or RDA, is designed for analyzing the differences between complex but highly related genomes and combines three elements: representation, subtractive enrichment, and kinetic enrichment.

Ras partners.
Van Aelst, L. and White, M. and Wigler, M. H. (1994) Cold Spring Harbor Symposia on Quantitative Biology, 59 pp. 181-186.

The RAS oncogenes were first discovered as the transforming elements of acutely oncogenic retroviruses. Subsequently, cellular RAS genes were found to be frequently activated by mutation in a wide variety of human cancers, providing the first example of a common oncogenic mechanism. As a consequence, Ras proteins have been studied extensively (for review, see Barbacid 1987). RAS genes are found ubiquitously in eukaryotic organisms. They encode low-molecular-weight guanine nucleotide-binding proteins that hydrolyze GTP and localize to the inner surface of the plasma membrane after undergoing elaborate carboxy-terminal processing. Proteins that are involved in processing Ras, or in regulating its activity, e.g., by accelerating guanine nucleotide hydrolysis or exchange, have been largely conserved in evolution (for review, see Wigler 1993).

Complex synthetic chemical libraries indexed with molecular tags.
Ohlmeyer, M. H. J. and Swanson, R. N. and Dillard, L. W. and Reader, J. C. and Asouline, G. and Kobayashi, R. and Wigler, M. H. and Still, W. C. (1993) Proceedings of the National Academy of Sciences of the United States of America, 90(23) pp. 10922-10926.

Combinatorial methods of chemical synthesis allow the creation of molecular libraries having immense diversity. The utility of such libraries is dependent upon identifying the structures of the molecules so prepared. We describe the construction of a peptide combinatorial library, having 117,649 different members, synthesized on beads and indexed with inert chemical tags. These tags are used as a binary code to record the reaction history of each bead. The code can be read directly from a single bead by electron capture capillary gas chromatography. We demonstrate the correct selection of members of the library on the basis of binding to a monoclonal antibody.

A family of human phosphodiesterases homologous to the dunce learning and memory gene product of Drosophila melanogaster are potential targets for antidepressant drugs.
Bolger, G. and Michaeli, T. and Martins, T. and St John, T. and Steiner, B. and Rodgers, L. and Riggs, M. and Wigler, M. H. and Ferguson, K. (1993) Mol Cell Biol, 13(10) pp. 6558-71.

We have isolated cDNAs for four human genes (DPDE1 through DPDE4) closely related to the dnc learning and memory locus of Drosophila melanogaster. The deduced amino acid sequences of the Drosophila and human proteins have considerable homology, extending beyond the putative catalytic region to include two novel, highly conserved, upstream conserved regions (UCR1 and UCR2). The upstream conserved regions are located in the amino-terminal regions of the proteins and appear to be unique to these genes. Polymerase chain reaction analysis suggested that these genes encoded the only homologs of dnc in the human genome. Three of the four genes were expressed in Saccharomyces cerevisiae and shown to encode cyclic AMP-specific phosphodiesterases. The products of the expressed genes displayed the pattern of sensitivity to inhibitors expected for members of the type IV, cyclic AMP-specific class of phosphodiesterases. Each of the four genes demonstrated a distinctive pattern of expression in RNA from human cell lines.

Complex formation between RAS and RAF and other protein kinases.
Van Aelst, L. and Barr, M. and Marcus, S. and Polverino, A. and Wigler, M. H. (1993) Proc Natl Acad Sci U S A, 90(13) pp. 6213-7.

We used a Saccharomyces cerevisiae genetic system to detect the physical interaction of RAS and RAF oncoproteins. We also observed interaction between RAS and byr2, a protein kinase implicated as a mediator of the Schizosaccharomyces pombe ras1 protein. Interaction with RAS required only the N-terminal domains of RAF or byr2 and was disrupted by mutations in either the guanine nucleotide-binding or effector-loop domains of RAS. We observed interaction between MEK (a kinase that phosphorylates mitogen-activated protein kinases) and the catalytic domain of RAF. RAS and MEK also interacted but only when RAF was overexpressed.

Isolation and characterization of a previously undetected human cAMP phosphodiesterase by complementation of cAMP phosphodiesterase- deficient Saccharomyces cerevisiae.
Michaeli, T. and Bloom, T. J. and Martins, T. and Loughney, K. and Ferguson, K. and Riggs, M. and Rodgers, L. and Beavo, J. A. and Wigler, M. H. (1993) Journal of Biological Chemistry, 268(17) pp. 12925-12932.

We have established a highly sensitive functional screen for the isolation of cDNAs encoding cAMP phosphodiesterases (PDEs) by complementation of defects in a Saccharomyces cerevisiae strain lacking both endogenous cAMP PDE genes, PDE1 and PDE2. Three groups of cDNAs corresponding to three distinct human genes encoding cAMP-specific PDEs were isolated from a human glioblastoma cDNA library using this functional screen. Two of these genes are closely related to the Drosophila dunce cAMP-specific PDE. The third gene, which we named HCP1, encoded a novel cAMP-specific PDE. HCP1 has an amino acid sequence related to the sequences of the catalytic domains of all cyclic nucleotide PDEs. HCP1 is a high affinity cAMP-specific PDE (K(m) = 0.2 muM) that does not share other properties of the cAMP-specific PDE family, i.e. extensive sequence homology to the Drosophila dunce cAMP PDE and sensitivity to rolipram and R020-1724. The PDE activity of HCP1 is not sensitive to cGMP or other inhibitors of the cGMP-inhibitable PDEs, such as milrinone. The biochemical and pharmacological properties of HCP1 suggest that it is a member of a previously undiscovered cyclic nucleotide PDE family. Northern blot analysis indicates that high levels of HCP1 mRNA are present in human skeletal muscle.

Human Sos1: a guanine nucleotide exchange factor for Ras that binds to GRB2.
Chardin, P. and Camonis, J. H. and Gale, N. W. and van Aelst, L. and Schlessinger, J. and Wigler, M. H. and Bar-Sagi, D. (1993) Science, 260(5112) pp. 1338-1343.

A human complementary DNA was isolated that encodes a widely expressed protein, hSos1, that is closely related to Sos, the product of the Drosophila son of sevenless gene. The hSos1 protein contains a region of significant sequence similarity to CDC25, a guanine nucleotide exchange factor for Ras from yeast. A fragment of hSos1 encoding the CDC25-related domain complemented loss of CDC25 function in yeast. This hSos1 domain specifically stimulated guanine nucleotide exchange on mammalian Ras proteins in vitro. Mammalian cells overexpressing full-length hSos1 had increased guanine nucleotide exchange activity. Thus hSos1 is a guanine nucleotide exchange factor for Ras. The hSos1 interacted with growth factor receptor-bound protein 2 (GRB2) in vivo and in vitro. This interaction was mediated by the carboxyl-terminal domain of hSos1 and the Src homology 3 (SH3) domains of GRB2. These results suggest that the coupling of receptor tyrosine kinases to Ras signaling is mediated by a molecular complex consisting of GRB2 and hSos1.

Evolution of Ras Signal Transduction Pathways.
Wigler, M. H. and Xu, H. P. and Vanaelst, L. and Jung, V. and Marcus, S. and Polverino, T. and Oneill, K. and Chang, E. and Camonis, J. and Nicolette, C. and White, M. and Ballester, R. (1993) Faseb Journal, 7(7) pp. A1124-A1124.

Analysis of the neurofibromatosis type-1 (NF1) GAP-related domain by site-directed mutagenesis.
Gutmann, D. H. and Boguski, M. and Marchuk, D. and Wigler, M. H. and Collins, F. S. and Ballester, R. (1993) Oncogene, 8(3) pp. 761-769.

The gene for von Recklinghausen neurofibromatosis type 1 (NF1) was recently identified by positional cloning and found to encode a protein with sequence similarity to a family of eucaryotic GTPase-activating proteins (GAPs). Expression of the NF1-GAP-related domain (NF1GRD) has been shown to complement yeast strains deficient in the yeast GAP homologs, IRA1 and IRA2, to interact with human RAS proteins and to accelerate the conversion of ras-GTP to ras-GDP. Further analysis of this region has revealed a number of residues that are highly conserved between members of the GAP family. Mutational analysis of a representative number of these residues produced one of three effects: (1) no change in NF1GRD function. (2) complete disruption of NF1GRD function and (3) intermediate retention of NF1GRD function. One of these mutations at residue 1423 was shown to have reduced ability to negatively regulate ras in yeast, which is interesting in tight of a recent report demonstrating a similar naturally occurring mutation in human malignancies.

Cloning the differences between two complex genomes.
Lisitsyn, N. and Lisitsyn, N. and Wigler, M. H. (1993) Science, 259(5097) pp. 946-51.

The analysis of the differences between two complex genomes holds promise for the discovery of infectious agents and probes useful for genetic studies. A system was developed in which subtractive and kinetic enrichment was used to purify restriction endonuclease fragments present in one population of DNA fragments but not in another. Application of this method to DNA populations of reduced complexity ("representations") resulted in the isolation of probes to viral genomes present as single copies in human DNA, and probes that detect polymorphisms between two individuals. In principle, this system, called representational difference analysis (RDA), may also be used for isolating probes linked to sites of genomic rearrangements, whether occurring spontaneously and resulting in genetic disorders or cancer, or programmed during differentiation and development.

A conserved alternative splice in the von Recklinghausen neurofibromatosis (NF1) gene produces two neurofibromin isoforms, both of which have GTPase-activating protein activity.
Andersen, L. B. and Ballester, R. and Marchuk, D. A. and Chang, E. and Gutmann, D. H. and Saulino, A. M. and Camonis, J. and Wigler, M. H. and Collins, F. S. (1993) Molecular and Cellular Biology, 13(1) pp. 487-495.

Sequence analysis has shown significant homology between the catalytic regions of the mammalian ras GTPase-activating protein (GAP), yeast Ira1p and Ira2p (inhibitory regulators of the RAS-cyclic AMP pathway), and neurofibromin, the protein encoded by the NF1 gene. Yeast expression experiments have confirmed that a 381-amino-acid segment of neurofibromin, dubbed the GAP-related domain (GRD), can function as a GAP. Using the RNA polymerase chain reaction with primers flanking the NF1-GRD, we have identified evidence for alternative splicing in this region of the NF1 gene. In addition to the already published sequence (type I), an alternative RNA carrying a 63-nucleotide insertion (type II) is present in all tissues examined, although the relative amounts of types I and II vary. The insertion is conserved across species but is not present in GAP, IRA1, or IRA2. GenBank searches have failed to identify significant similarity between the inserted sequence and known DNA or protein sequences, although the basic amino acid composition of the insertion shares features with nuclear targeting sequences. Expression studies in yeasts show that despite the partial disruption of the neurofibromin-IRA-GAP homology by this insertion, both forms of the NF1-GRD can complement loss of IRA function. In vivo assays designed to compare the GAP activity of the two alternatively spliced forms of the NF1-GRD show that both can increase the conversion of GTP-bound ras to its GDP-bound form, although the insertion of the 21 amino acids weakens this effect. The strong conservation of this alternative, splicing suggests that both type I and II isoforms mediate important biological functions of neurofibromin.

Functional homology of protein kinases required for sexual differentiation in Schizosaccharomyces pombe and Saccharomyces cerevisiae suggests a conserved signal transduction module in eukaryotic organisms.
Neiman, A. M. and Stevenson, B. J. and Xu, H. P. and Sprague, G. F. and Herskowitz, I. and Wigler, M. H. and Marcus, S. (1993) Molecular Biology of the Cell, 4(1) pp. 107-120.

We present genetic evidence that three presumptive protein kinases of Schizosaccharomyces pombe, byr2, byr1, and spk1 that are structurally related to protein kinases of Saccharomyces cerevisiae, STE11, STE7, and FUS3, respectively, are also functionally related. In some cases, introduction of the heterologous protein kinase into a mutant was sufficient for complementation. In other cases (as in a ste11- mutant of S. cerevisiae), expression of two S. pombe protein kinases (byr2 and byr1) was required to observe complementation, suggesting that byr2 and byr1 act cooperatively. Complementation in S. pombe mutants is observed as restoration of sporulation and conjugation and in S. cerevisiae as restoration of conjugation, pheromone-induced cell cycle arrest, and pheromone-induced transcription of the FUS1 gene. We also show that the S. pombe kinases bear a similar relationship to the mating pheromone receptor apparatus as do their S. cerevisiae counterparts. Our results indicate that pheromone-induced signal transduction employs a conserved set of kinases in these two evolutionarily distant yeasts despite an apparently significant difference in function of the heterotrimeric G proteins. We suggest that the STE11/byr2, STE7/byr1, and FUS3/spk1 kinases comprise a signal transduction module that may be conserved in higher eukaryotes. Consistent with this hypothesis, we show that a mammalian mitogen-activated protein (MAP) kinase, ERK2, can partially replace spk1 function in S. pombe.

RAS function and protein kinase cascades.
Marcus, S. and Wigler, M. H. and Xu, H. P. and Ballester, R. and Kawamukai, M. and Polverino, A. (1993) Ciba Found Symp, 176 pp. 53-61; discussion 61.

This paper reviews recent progress in understanding the function of RAS in three systems: the budding yeast (Saccharomyces cerevisiae), the fission yeast (Schizosaccharomyces pombe) and Xenopus laevis oocytes. One of the functions of RAS in S. cerevisiae is the stimulation of adenylate cyclase. This leads to the activation of the cAMP-dependent protein kinases--a function that has probably not been conserved in evolution. The immediate function of RAS in S. pombe is not known, but it may lead to the activation of a protein kinase cascade. This cascade has likely been conserved in evolution and linkage between it and RAS can be demonstrated in cell-free extracts from Xenopus oocytes. The Xenopus cell-free system provides a means to test specific hypotheses about RAS function and to isolate targets of RAS.

Oncogenic Ras triggers the activation of 42-kDa mitogen-activated protein kinase in extracts of quiescent Xenopus oocytes.
Shibuya, E. K. and Polverino, A. J. and Chang, E. and Wigler, M. H. and Ruderman, J. V. (1992) Proceedings of the National Academy of Sciences of the United States of America, 89(20) pp. 9831-9835.

Quiescent, full-grown Xenopus oocytes, which are arrested at the G2/M border of meiosis, contain an inactive 42-kDa mitogen-activated protein kinase (p42MAPK) that is activated when oocytes are stimulated to resume the meiotic cell cycle. We have made extracts from these oocytes that respond to four cell cycle activators: oncogenic [Val12]Ras protein, clam cyclins ADELTA60 and BDELTA97, and the phosphatase inhibitor okadaic acid. All four induce the tyrosine phosphorylation and activation of p42MAPK. Both cyclins and okadaic acid, but not [Val12]Ras, also lead to activation of the endogenous cyclin B/cdc2 kinase complexes in extracts of quiescent oocytes. Using extracts prepared from cycloheximide-arrested interphase cells, we show that although p42MAPK activation can occur in response to cyclin-activated cdc2, the Ras-induced activation of p42MAPK occurs without intervening cdc2 activation. Neither the nononcogenic [Gly12]Ras nor [Val12,Arg186]-Ras, a mutant that lacks the C-terminal consensus sequence directing prenylation and subsequent membrane association, is an effective activator of p42MAPK in vitro.

A gene encoding a protein with seven zinc finger domains acts on the sexual differentiation pathways of Schizosaccharomyces pombe.
Xu, H. P. and Rajavashisth, T. and Grewal, N. and Jung, V. and Riggs, M. and Rodgers, L. and Wigler, M. H. (1992) Mol Biol Cell, 3(7) pp. 721-34.

Byr3 was selected as a multicopy suppressor of the sporulation defects of diploid Schizosaccharomyces pombe cells that lack ras1. Like cells mutant at byr1 and byr2, two genes that encode putative protein kinases and that in multiple copies are also suppressors of the sporulation defects of ras1 null diploid cells, cells mutant at byr3 are viable but defective in conjugation. Nucleic acid sequence indicates byr3 has the capacity to encode a protein with seven zinc finger binding domains, similar in structure to the cellular nucleic acid binding protein (CNBP), a human protein that was identified on the basis of its ability to bind DNA. Expression of CNBP in yeast can partially suppress conjugation defects of cells lacking byr3.

SNC1, a yeast homolog of the synaptic vesicle-associated membrane protein/synaptobrevin gene family: genetic interactions with the RAS and CAP genes.
Gerst, J. E. and Rodgers, L. and Riggs, M. and Wigler, M. (1992) Proc Natl Acad Sci U S A, 89(10) pp. 4338-42.

SNC1, a gene from the yeast Saccharomyces cerevisiae, encodes a homolog of vertebrate synaptic vesicle-associated membrane proteins (VAMPs) or synaptobrevins. SNC1 was isolated by its ability to suppress the loss of CAP function in S. cerevisiae strains possessing an activated allele of RAS2. CAP is a component of the RAS-responsive S. cerevisiae adenylyl cyclase complex. The N-terminal domain of CAP is required for full cellular responsiveness to activated RAS proteins. The C-terminal domain of CAP is required for normal cellular morphology and responsiveness to nutrient extremes. Multicopy plasmids expressing SNC1 suppress only the loss of the C-terminal functions of CAP and only in the presence of activated RAS2.

Genetic and biochemical analysis of the adenylyl cyclase-associated protein, cap, in Schizosaccharomyces pombe.
Kawamukai, M. and Gerst, J. and Field, J. and Riggs, M. and Rodgers, L. and Wigler, M. H. and Young, D. (1992) Molecular Biology of the Cell, 3(2) pp. 167-180.

We have identified, cloned, and studied a gene, cap, encoding a protein that is associated with adenylyl cyclase in the fission yeast Schizosaccharomyces pombe. This protein shares significant sequence homology with the adenylyl cyclase-associated CAP protein in the yeast Saccharomyces cerevisiae. CAP is a bifunctional protein; the N-terminal domain appears to be involved in cellular responsiveness to RAS, whereas loss of the C-terminal portion is associated with morphological and nutritional defects. S. pombe cap can suppress phenotypes associated with deletion of the C-terminal CAP domain in S. cerevisiae but does not suppress phenotypes associated with deletion of the N-terminal domain. Analysis of cap disruptants also mapped the function of cap to two domains. The functional loss of the C-terminal region of S. pombe cap results in abnormal cellular morphology, slow growth, and failure to grow at 37-degrees-C. Increases in mating and sporulation were observed when the entire gene was disrupted. Overproduction of both cap and adenylyl cyclase results in highly elongated large cells that are sterile and have measurably higher levels of adenylyl cyclase activity. Our results indicate that cap is required for the proper function of S. pombe adenylyl cyclase but that the C-terminal domain of cap has other functions that are shared with the C-terminal domain of S. cerevisiae CAP.

Alternative Splicing in the Proposed Catalytic Domain of the Type-1 Neurofibromatosis Gene.
Andersen, L. and Ballester, R. and Marchuk, D. and Saulino, A. and Wigler, M. H. and Collins, F. (1991) American Journal of Human Genetics, 49(4) pp. 21-21.

Evidence for a functional link between profilin and CAP in the yeast S. cerevisiae.
Vojtek, A. and Haarer, B. and Field, J. and Gerst, J. and Pollard, T. D. and Brown, S. and Wigler, M. H. (1991) Cell, 66(3) pp. 497-505.

CAP is a component of the S. cerevisiae adenylyl cyclase complex. The N-terminal domain is required for cellular RAS responsiveness. Loss of the C-terminal domain is associated with morphological and nutritional defects. Here we report that cap- cells bud randomly and are defective in actin distribution. The morphological and nutritional defects associated with loss of the CAP C-terminal domain are suppressed by over-expression of PFY, the gene encoding profilin, an actin- and polyphosphoinositide-binding protein. The phenotype of cells lacking PFY resembles that of cells lacking the CAP C-terminal domain. Study of mutated yeast profilins and profilins from Acanthamoeba suggests that the ability of profilin to suppress cap- cells is dependent upon a property other than, or in addition to, its ability to bind actin. This property may be its ability to bind polyphosphoinositides. We propose that CAP and profilin provide a link between growth signals and remodeling of the cellular cytoskeleton.

The Ras Oncoprotein and M-Phase Activity.
Daar, I. and Nebreda, A. R. and Yew, N. and Sass, P. and Paules, R. and Santos, E. and Wigler, M. H. and Vandewoude, G. F. (1991) Science, 253(5015) pp. 74-76.

The endogenous mos proto-oncogene product (Mos) is required for meiotic maturation In Xenopus oocytes, the ras oncogene product (Ras) can induce meiotic maturation and high levels of M-phase-promoting factor (MPF) independent of endogenous Mos, indicating that a parallel pathway to metaphase exists. In addition, Ras, like Mos and cytostatic factor, can arrest Xenopus embryonic cell cleavage in mitosis and maintain high levels of MPF. Thus, in the Xenopus oocyte and embryo systems Ras functions in the M phase of the cell cycle. The embryonic cleavage arrest assay is a rapid and sensitive test for Ras function.

byr2, a Schizosaccharomyces pombe gene encoding a protein kinase capable of partial suppression of the ras1 mutant phenotype.
Wang, Y. and Xu, H. P. and Riggs, M. and Rodgers, L. and Wigler, M. H. (1991) Molecular & Cellular Biology, 11(7) pp. 3554-63.

Schizosaccharomyces pombe contains a single gene, ras1, which is a homolog of the mammalian RAS genes. ras1 is required for conjugation, sporulation, and normal cell shape. ras1 has been previously identified as ste5. We report here a gene we call byr2 that can encode a predicted protein kinase and can partially suppress defects in ras1 mutants. ras1 mutant strains expressing high levels of byr2 can sporulate competently but are still defective in conjugation and abnormally round. byr2 mutants are viable and have normal shape but are absolutely defective in conjugation and sporulation. byr2 is probably identical to ste8. In many respects, byr2 resembles the byr1 gene, another suppressor of the ras1 mutation, which has been identified previously as ste1. Our data indicate that if ras1, byr2, and byr1 act along the same pathway, then the site of action for byr2 is between the sites for ras1 and byr1.

The adenylyl cyclase-encoding gene from Saccharomyces kluyveri.
Young, D. and O'Neill, K. and Broek, D. and Wigler, M. H. (1991) Gene, 102(1) pp. 129-32.

The gene encoding adenylyl cyclase (CYR) from Saccharomyces kluyveri has been cloned. Comparison of the predicted amino acid sequence of this protein with the Schizosaccharomyces pombe and Saccharomyces cerevisiae CYRs revealed homology between different structural and putative functional domains that suggest a high degree of conservation in the function and regulation of these proteins.

Expression of three mammalian cDNAs that interfere with RAS function in Saccharomyces cerevisiae.
Colicelli, J. and Nicolette, C. and Birchmeier, C. and Rodgers, L. and Riggs, M. and Wigler, M. H. (1991) Proc Natl Acad Sci U S A, 88(7) pp. 2913-7.

Saccharomyces cerevisiae strains expressing the activated RAS2Val19 gene or lacking both cAMP phosphodiesterase genes, PDE1 and PDE2, have impaired growth control and display an acute sensitivity to heat shock. We have isolated two classes of mammalian cDNAs from yeast expression libraries that suppress the heat shock-sensitive phenotype of RAS2Val19 strain. Members of the first class of cDNAs also suppress the heat shock-sensitive phenotype of pde1- pde2- strains and encode cAMP phosphodiesterases. Members of the second class fail to suppress the phenotype of pde1- pde2- strains and therefore are candidate cDNAs encoding proteins that interact with RAS proteins. We report the nucleotide sequence of three members of this class. Two of these cDNAs share considerable sequence similarity, but none are clearly similar to previously isolated genes.

CAP is a bifunctional component of the Saccharomyces cerevisiae adenylyl cyclase complex.
Gerst, J. E. and Ferguson, K. and Vojtek, A. and Wigler, M. H. and Field, J. (1991) Molecular & Cellular Biology, 11(3) pp. 1248-57.

CAP, a protein from Saccharomyces cerevisiae that copurifies with adenylyl cyclase, appears to be required for yeast cells to be fully responsive to RAS proteins. CAP also appears to be required for normal cell morphology and responsiveness to nutrient deprivation and excess. We describe here a molecular and phenotypic analysis of the CAP protein. The N-terminal domain is necessary and sufficient for cellular response to activated RAS protein, while the C-terminal domain is necessary and sufficient for normal cellular morphology and responses to nutrient extremes. Thus, CAP is a novel example of a bifunctional component involved in the regulation of diverse signal transduction pathways.

Genetic and biochemical analysis of the adenylyl cyclase of Schizosaccharomyces pombe.
Kawamukai, M. and Ferguson, K. and Wigler, M. H. and Young, D. (1991) Cell Regulation, 2(2) pp. 155-64.

The adenylyl cyclase gene, cyr1, of Schizosaccharomyces pombe has been cloned. We have begun an analysis of the function and regulation of adenylyl cyclase by disrupting this gene and by over-expressing all or parts of this gene in various strains. cyr1- strains are viable and contain no measurable cyclic AMP. They conjugate and sporulate under conditions that normally inhibit wild-type strains. Strains containing the cyr1 coding sequences transcribed from the strong adh1 promoter contain greatly elevated adenylyl cyclase activity, as measured in vitro, but only modestly elevated cAMP levels. Such strains conjugate and sporulate less frequently than wild-type cells upon nutrient limitation. Strains which carry the wild-type cyr1 gene but that also express high levels of the amino terminal domain of adenylyl cyclase behave much like cyr1-strains, suggesting that the amino terminal domain can bind a positive regulator. A protein that copurifies with the adenylyl cyclase of S. pombe cross-reacts to antiserum raised against the S. cerevisiae adenylyl cyclase-associated regulatory protein, CAP.

The Ras Signalling Pathway in Yeast.
Vojtek, A. and Field, J. and Gerst, J. and Wigler, M. H. (1991) Journal of Cellular Biochemistry - Supplement.(15 PART B), pp. 84.

sar1, a gene from Schizosaccharomyces pombe encoding a protein that regulates ras1.
Wang, Y. and Boguski, M. and Riggs, M. and Rodgers, L. and Wigler, M. (1991) Cell Regulation, 2(6) pp. 453-465.

Proper ras1 function is required for normal sexual function in the yeast Schizosaccharomyces pombe. We have found a gene in S. pombe, sar1, that encodes a product capable of regulating ras1 function. sar1 is a member of an expanding family of RAS GTPase-activating proteins (GAPs) that includes mammalian GAP, the yeast Saccharomyces cerevisiae IRA proteins, and the product of the human neurofibromatosis locus, NF1 sar1, like these other proteins, can complement the loss of IRA function in S. cerevisiae. Computer analysis shows that the highest degree of sequence conservation is restricted to a very small number of diagnostic residues represented by the motif Phe-Leu-Arg-X-X-X-Pro-Ala-X-X-X-Pro. We find no evidence that sar1 is required for the effector function of ras1.

The Ways of Ras.
Wigler, M. H. and Ballester, R. and Chang, E. and Colicelli, J. and Field, J. and Gerst, J. and Kawamukai, M. and Michaeli, T. and Nicolette, C. (1991) Proceedings American Association for Cancer Research Annual Meeting, pp. 435.

The NF1 locus encodes a protein functionally related to mammalian GAP and yeast IRA proteins .
Ballester, R. and Marchuk, D. and Boguski, M. and Saulino, A. and Letcher, R. and Wigler, M. H. and Collins, F. (1990) Cell, 63(4) pp. 851-9.

The von Recklinghausen neurofibromatosis locus, NF1, encodes a protein with homology restricted to the catalytic region of the RAS GTPase-activating protein, GAP, and with extensive homology to the IRA1 and IRA2 gene products of the yeast S. cerevisiae. A segment of the NF1 cDNA gene, expressed in yeast, can complement loss of IRA function and can inhibit both wild-type and mutant activated human H-ras genes that are coexpressed in yeast. Yeast expressing the NF1 segment have increased H-ras GTPase-stimulating activity. These studies indicate that the NF1 gene product can interact with RAS proteins and demonstrate structural and functional similarities and differences among the GAP, IRA1, IRA2, and NF1 proteins.

Biological activity of the mammalian RAP genes in yeast.
Xu, H. P. and Wang, Y. and Riggs, M. and Rodgers, L. and Wigler, M. H. (1990) Cell Regulation, 1(10) pp. 763-769.

We have screened expression libraries for mammalian cDNAs capable of suppressing defects in ras1- Schizosaccharomyces pombe. Both the RAP1A and RAP1B genes were identified in this manner. They suppress defects in cell morphology and sporulation, although not conjugation. In contrast, RAP genes do not suppress phenotypes in the yeast Saccharomyces cerevisiae that are deficient in RAS. Indeed, expression of RAP1A appears to antagonize the activated S. cerevisiae RAS2val19 gene. These results indicate that RAP proteins can interact with RAS targets, sometimes productively, sometimes nonproductively.

Oncoproteins. Gaps in Understanding Ras.
Wigler, M. H. (1990) Nature, 346(6286) pp. 696-697.

Mutational mapping of RAS-responsive domains of the Saccharomyces cerevisiae adenylyl cyclase.
Colicelli, J. and Field, J. and Ballester, R. and Chester, N. and Young, D. and Wigler, M. H. (1990) Molecular & Cellular Biology, 10(6) pp. 2539-43.

Large deletion and small insertion mutations in the adenylyl cyclase gene of Saccharomyces cerevisiae were used to map regions required for activation by RAS protein in vitro. The amino-terminal 605 amino acids were found to be dispensable for responsiveness to RAS protein. All other deletions in adenylyl cyclase destroyed its ability to respond to RAS. Small insertion mutations within the leucine-rich repeat region also prevented RAS responsiveness, while other insertions did not.

Cloning and characterization of CAP, the S. cerevisiae gene encoding the 70 kd adenylyl cyclase-associated protein.
Field, J. and Vojtek, A. and Ballester, R. and Bolger, G. and Colicelli, J. and Ferguson, K. and Gerst, J. and Kataoka, T. and Michaeli, T. and Powers, S. and Riggs, M. and Rodgers, L. and Wieland, I. and Wheland, B. and Wigler, M. H. (1990) Cell, 61(2) pp. 319-327.

Adenylyl cyclase from S. cerevisiae contains at least two subunits, a 200 kd catalytic subunit and a subunit with an apparent molecular size of 70 kd, which we now call CAP (cyclase-associated protein). We cloned a cDNA encoding CAP by screening a yeast cDNA expression library in E. coli with antisera raised against the purified protein. The cDNA contained an open reading frame capable of encoding a 526 amino acid protein that is not homologous to any sequences in the current data bases. Adenylyl cyclase activity in membranes from cells that lacked CAP was not stimulated by RAS2 proteins in vitro. These results suggest that CAP is required for at least some aspects of the RAS-responsive signaling system. Mutants lacking CAP had four additional phenotypes that appear to be unrelated to effects of the RAS/adenylyl cyclase pathway: the inability to grow on rich medium (YPD), temperature sensitivity on minimal medium, sensitivity to nitrogen starvation, and a swollen cell morphology.

A method for difference cloning: gene amplification following subtractive hybridization.
Wieland, I. and Bolger, G. and Asouline, G. and Wigler, M. H. (1990) Proc Natl Acad Sci U S A, 87(7) pp. 2720-4.

We describe a procedure for genomic difference cloning, a method for isolating sequences present in one genomic DNA population ("tester") that is absent in another ("driver"). By subtractive hybridization, a large excess of driver is used to remove sequences common to a biotinylated tester, enriching the "target" sequences that are unique to the tester. After repeated subtractive hybridization cycles, tester is separated from driver by avidin/biotin affinity chromatography, and single-stranded target is amplified by the polymerase chain reaction, rendering it double-stranded and clonable. We model two situations: the gain of sequences that result from infection with a pathogen and the loss of sequences that result from a large hemizygous deletion. We obtain 100- to 700-fold enrichment of target sequences.

Mutations of the adenylyl cyclase gene that block RAS function in Saccharomyces cerevisiae.
Field, J. and Xu, H. P. and Michaeli, T. and Ballester, R. and Sass, P. and Wigler, M. H. and Colicelli, J. (1990) Science, 247(4941) pp. 464-7.

The interaction between RAS proteins and adenylyl cyclase was studied by using dominant interfering mutations of adenylyl cyclase from the yeast Saccharomyces cerevisiae. RAS proteins activate adenylyl cyclase in this organism. A plasmid expressing a catalytically inactive adenylyl cyclase was found to interfere dominantly with this activation. The interfering region mapped to the leucine-rich repeat region of adenylyl cyclase, which is homologous to domains present in several other proteins and is thought to participate in protein-protein interactions.

Isolation and Characterization of Genes That Regulate Ras Function in the Yeast Saccharomyces-Cerevisiae.
Ballester, R. and Wigler, M. H. (1990) Yeast, 6(SPEC. ) pp. S212.

Characterization of Ros1 cDNA from a Human Glioblastoma Cell Line.
Birchmeier, C. and O'Neill, K. and Riggs, M. and Wigler, M. H. (1990) Proceedings of the National Academy of Sciences of the United States of America, 87(12) pp. 4799-4803.

We have isolated and characterized a human ROS1 cDNA from the glioblastoma cell line SW-1088. The cDNA, 8.3 kilobases long, has the potential to encode a transmembrane tyrosine-specific protein kinase with a predicted molecular mass of 259 kDa. The putative extracellular domain of ROS1 is homologous to the extracellular domain of the sevenless gene product from Drosophila. No comparable similarites in the extracellular domains were found between ROS1 and other receptor-type tyrosine kinases. Together, ROS1 and sevenless gene products define a distinct subclass of transmembrane tyrosine kinases.

Molecular and Phenotypic Analysis of Cap a Regulatory Element Involved in Mediating Both Ras Action and Amino Acid Balance.
Gerst, J. E. and Field, J. and Ferguson, K. and Wigler, M. H. (1990) Yeast, 6(SPEC. ) pp. S174.

A gene from Schizosaccharomyces-Pombe with homology to Escherichia-Coli RNAse III blocks conjugation and sporulation when overexpressed in wild type cells.
Xu, H. P. and Riggs, M. and Rodgers, L. and Wigler, M. H. (1990) Nucleic Acids Research, 18(17) pp. 5304.

Genetic analysis of mammalian GAP expressed in yeast.
Ballester, R. and Michaeli, T. and Ferguson, K. and Xu, H. P. and McCormick, F. and Wigler, M. H. (1989) Cell, 59(4) pp. 681-6.

We have designed a vector to express the mammalian GAP protein in the yeast S. cerevisiae. When expressed in yeast, GAP inhibits the function of the human H-rasgly12 protein, but not that of the H-rasval12 protein, and complements the loss of IRA1. IRA1 is a yeast gene that encodes a protein with homology to GAP and acts upstream of RAS. Mammalian GAP can therefore function in yeast and interact with yeast RAS. Because expression of GAP complements ira1-mutants, we propose that GAP shares some biochemical functions with IRA1. Other studies indicate that IRA1 controls the level of RAS activity, presumably by regulating GTP hydrolysis. By analogy, we propose that GAP may play a similar role.

Mutants of H-ras that interfere with RAS effector function in Saccharomyces cerevisiae.
Michaeli, T. and Field, J. and Ballester, R. and O'Neill, K. and Wigler, M. H. (1989) Embo J, 8(10) pp. 3039-44.

We report a class of interfering mutants of the human H-ras gene capable of inhibiting phenotypes arising from the expression of the activated RAS2 gene, RAS2val19, in the yeast Saccharomyces cerevisiae. All these mutants encode unprocessed H-ras proteins that remain in the cytoplasm. One of the mutants, H-rasarg186, was examined in detail. H-rasarg186 protein is a competitive inhibitor of RAS2val19 protein. It does not interfere with processing and membrane localization of RAS2val19, nor does it appear to compete with RAS protein for its proposed regulator, the CDC25 protein. By several criteria the RAS2val19 adenylate cyclase interaction is unaffected by H-rasarg186. We infer from our results that H-rasarg186 protein interferes with an alternative function of RAS2val19.

The adenylyl cyclase gene from Schizosaccharomyces pombe.
Young, D. and Riggs, M. and Field, J. and Vojtek, A. and Broek, D. and Wigler, M. H. (1989) Proc Natl Acad Sci U S A, 86(20) pp. 7989-93.

We cloned the adenylyl cyclase gene from the fission yeast Schizosaccharomyces pombe using low-stringency hybridization to the Saccharomyces cerevisiae adenylyl cyclase gene. The Sc. pombe gene encodes a 1692-amino acid-residue protein. The identity of this gene was confirmed by studies of its expression in Sa. cerevisiae. Expression of the carboxyl-terminal region of the Sc. pombe adenylyl cyclase protein will suppress a temperature-sensitive mutation in the Sa. cerevisiae adenylyl cyclase gene. Furthermore, Sa. cerevisiae that lack their endogenous adenylyl cyclase gene and express the carboxyl-terminal region of the Sc. pombe adenylyl cyclase protein have measurable adenylyl cyclase activity. The carboxyl-terminal region of this protein has strong homology with the catalytic domain of the Sa. cerevisiae adenylyl cyclase. Also, Sc. pombe adenylyl cyclase, like Sa. cerevisiae adenylyl cyclase, contains a tandemly repeated motif rich in leucine. Neither yeast protein is particularly homologous to the recently cloned Gs-responsive mammalian adenylyl cyclase [Krupinski, J., Coussen, F., Bakalyar, H. A., Tang, W.-J., Feinstein, P. G., Orth, K., Slaughter, C., Reed, R. R. & Gilman, A. G. (1989) Science 244, 1558-1564].

Isolation and characterization of a mammalian gene encoding a high-affinity cAMP phosphodiesterase.
Colicelli, J. and Birchmeier, C. and Michaeli, T. and O'Neill, K. and Riggs, M. and Wigler, M. H. (1989) Proc Natl Acad Sci U S A, 86(10) pp. 3599-603.

A rat brain cDNA library has been constructed in a Saccharomyces cerevisiae expression vector and used to isolate genes that can function in yeast to suppress the phenotypic effects of RAS2val19, a mutant form of the RAS2 gene analogous to an oncogenic mutant of the human HRAS gene. One cDNA, DPD, was cloned and its genetic and biochemical properties were characterized. A DPD product would share 80% amino acid sequence identity with the Drosophila melanogaster dunce-encoded protein over an extended region. We have shown that the DPD protein is a high-affinity cAMP-specific phosphodiesterase.

Dominant yeast and mammalian RAS mutants that interfere with the CDC25-dependent activation of wild-type RAS in Saccharomyces cerevisiae.
Powers, S. and O'Neill, K. and Wigler, M. H. (1989) Molecular & Cellular Biology, 9(2) pp. 390-5.

Two mutant alleles of RAS2 were discovered that dominantly interfere with wild-type RAS function in the yeast Saccharomyces cerevisiae. An amino acid substitution which caused the dominant interference was an alanine for glycine at position 22 or a proline for alanine at position 25. Analogous mutations in human H-ras also dominantly inhibited RAS function when expressed in yeast cells. The inhibitory effects of the mutant RAS2 or H-ras genes could be overcome by overexpression of CDC25, but only in the presence of wild-type RAS. These results suggest that these mutant RAS genes interfere with the normal interaction of RAS and CDC25 proteins and suggest that this interaction is direct and has evolutionarily conserved features.

Characterization of the ros1-gene products expressed in human glioblastoma cell lines.
Sharma, S. and Birchmeier, C. and Nikawa, J. and O'Neill, K. and Rodgers, L. and Wigler, M. H. (1989) Oncogene Res, 5(2) pp. 91-100.

Most glioblastoma-derived cell lines express an 8.3 kb ros1 transcript and a 280 kD glycoprotein designated gp280ros1, which can be specifically immunoprecipitated with an anti-ROS antibody. This 280 kD protein possesses in vitro autokinase activity and was observed in four independent glioblastoma cell lines. In a fifth glioblastoma cell line, U-118 MG, a smaller ros1 transcript of 4.0 kb was observed. Immunoprecipitation analysis reveals that the U-118 MG expressed a smaller, 116 kD ros1 gene product. cDNA cloning and sequencing of the U-118 MG ros1 transcript indicates it encodes the entire tyrosine kinase domain and two amino acids of the transmembrane domain of ros1 at its 3' end. Sequences at its 5' end likely arise from another gene.

Characterization of the rat mas oncogene and its high-level expression in the hippocampus and cerebral cortex of rat brain.
Young, D. and O'Neill, K. and Jessell, T. and Wigler, M. H. (1988) Proc Natl Acad Sci U S A, 85(14) pp. 5339-42.

The human mas oncogene was originally detected by its ability to transform NIH 3T3 cells. We previously showed that the protein encoded by this gene is unique among cellular oncogene products in that it has seven hydrophobic potential transmembrane domains and shares strong sequence similarity with a family of hormone-receptor proteins. We have now cloned the rat homolog of the mas oncogene, determined its DNA sequence, and examined its expression in various rat tissues. A comparison of the predicted sequences of the rat and human mas proteins shows that they are highly conserved, except in their hydrophilic amino-terminal domains. Our examination of the expression of mas, determined by RNA-protection studies, indicates that high levels of mas RNA transcripts are present in the hippocampus and cerebral cortex of the brain, but not in other neural regions or in other tissues. This pattern of expression and the similarity of mas protein to known receptor proteins suggest that mas encodes a receptor that is involved in the normal neurophysiology and/or development of specific neural tissues.

cAMP-independent control of sporulation, glycogen metabolism, and heat shock resistance in S. cerevisiae.
Cameron, S. and Levin, L. and Zoller, M. and Wigler, M. H. (1988) Cell, 53(4) pp. 555-66.

Genes encoding the regulatory (BCY1) and catalytic (TPK1, TPK2, and TPK3) subunits of the cAMP-dependent protein kinase (cAPK) are found in S. cerevisiae. bcy1- yeast strains do not respond properly to nutrient conditions. Unlike wild type, bcy1- strains do not accumulate glycogen, form spores, or become resistant to heat shock when nutrient limited. We have isolated mutant TPK genes that suppress all of the bcy1- defects. The mutant TPK genes appear to encode functionally attenuated catalytic subunits of the cAPK. bcy1- yeast strains containing the mutant TPK genes respond appropriately to nutrient conditions, even in the absence of CDC25, both RAS genes, or CYR1. Together, these genes encode the known components of the cAMP-generating machinery. The results indicate that cAMP-independent mechanisms must exist for regulating glycogen accumulation, sporulation, and the acquisition of thermotolerance in S. cerevisiae.

Purification of a RAS-responsive adenylyl cyclase complex from Saccharomyces cerevisiae by use of an epitope addition method.
Field, J. and Nikawa, J. and Broek, D. and MacDonald, B. and Rodgers, L. and Wilson, I. A. and Lerner, R. A. and Wigler, M. H. (1988) Molecular & Cellular Biology, 8(5) pp. 2159-65.

We developed a method for immunoaffinity purification of Saccharomyces cerevisiae adenylyl cyclase based on creating a fusion with a small peptide epitope. Using oligonucleotide technology to encode the peptide epitope we constructed a plasmid that expressed the fusion protein from the S. cerevisiae alcohol dehydrogenase promoter ADH1. A monoclonal antibody previously raised against the peptide was used to purify adenylyl cyclase by affinity chromatography. The purified enzyme appeared to be a multisubunit complex consisting of the 200-kilodalton adenylyl cyclase fusion protein and an unidentified 70-kilodalton protein. The purified protein could be activated by RAS proteins. Activation had an absolute requirement for a guanine nucleoside triphosphate.

SCH9, a gene of Saccharomyces cerevisiae that encodes a protein distinct from, but functionally and structurally related to, cAMP-dependent protein kinase catalytic subunits.
Toda, T. and Cameron, S. and Sass, P. and Wigler, M. H. (1988) Genes Dev, 2(5) pp. 517-27.

A new gene, SCH9, was isolated from Saccharomyces cerevisiae by its ability to complement a cdc25ts mutation. Sequence analysis indicates that it encodes a 90,000-dalton protein with a carboxy-terminal domain homologous to yeast and mammalian cAMP-dependent protein kinase catalytic subunits. In addition to suppressing loss of CDC25 function, multicopy plasmids containing SCH9 suppress the growth defects of strains lacking the RAS genes, the CYR1 gene, which encodes adenylyl cyclase, and the TPK genes, which encode the cAMP-dependent protein kinase catalytic subunits. Cells lacking SCH9 grow slowly and have a prolonged G1 phase of the cell cycle. This defect is suppressed by activation of the cAMP effector pathway. We propose that SCH9 encodes a protein kinase that is part of a growth control pathway which is at least partially redundant with the cAMP pathway.

A mutation in the catalytic subunit of cAMP-dependent protein kinase that disrupts regulation.
Levin, L. R. and Kuret, J. and Johnson, K. E. and Powers, S. and Cameron, S. and Michaeli, T. and Wigler, M. H. and Zoller, M. J. (1988) Science, 240(4848) pp. 68-70.

A mutant catalytic subunit of adenosine 3',5'-monophosphate (cAMP)-dependent protein kinase has been isolated from Saccharomyces cerevisiae that is no longer subject to regulation yet retains its catalytic activity. Biochemical analysis of the mutant subunit indicates a 100-fold decreased affinity for the regulatory subunit. The mutant catalytic subunit exhibits approximately a threefold increase in Michaelis constant for adenosine triphosphate and peptide cosubstrates, and is essentially unchanged in its catalytic rate. The nucleotide sequence of the mutant gene contants a single nucleotide change resulting in a threonine-to-alanine substitution at amino acid 241. This residue is conserved in other serine-threonine protein kinases. These results identify this threonine as an important contact between catalytic and regulatory subunits but only a minor contact in substrate recognition.

Integration of Ras and Adenylate Cyclase Pathways in Yeast.
Wigler, M. H. and Cameron, S. and Powers, S. and Field, J. and Toda, T. and Broek, D. and Nikawa, J. and Michaelis, T. and Colicelli, J. (1988) Journal of Cell Biology, 107(6 PART) pp. 4A.

Studies of RAS function in the yeast Saccharomyces cerevisiae.
Wigler, M. H. and Field, J. and Powers, S. and Broek, D. and Toda, T. and Cameron, S. and Nikawa, J. and Michaeli, T. and Colicelli, J. and Ferguson, K. (1988) Cold Spring Harbor Symposia on Quantitative Biology, 53(2) pp. 649-55.

The three mammalian RAS genes, Ha-ras, Ki-ras, and N-ras, are capable of the malignant transformation of cultured animal cells (Barbacid 1987). Mutations in these genes have been linked to a large number of human cancers (Barbacid 1987). These genes encode closely related proteins that bind guanine nucleotides (Scolnick et al. 1979; Shih et al. 1980; Ellis et al. 1981) and are localized to the inner surface of the plasma membrane (Willingham et al. 1980; Papageorge et al, 1982). Normal RAS proteins also slowly hydrolyze GTP (Gibbs et al. 1984; McGrath et al. 1984; Sweet et al. 1984). These properties are similar to those of the G proteins, which has led to the widespread expectation that RAS proteins, like G proteins, are involved in the transduction of membrane signals that are linked to cellular proliferation or differentiation.

Expression and rearrangement of the ROS1 gene in human glioblastoma cells.
Birchmeier, C. and Sharma, S. and Wigler, M. H. (1987) Proc Natl Acad Sci U S A, 84(24) pp. 9270-4.

The human ROS1 gene, which possibly encodes a growth factor receptor, was found to be expressed in human tumor cell lines. In a survey of 45 different human cell lines, we found ROS1 to be expressed in glioblastoma-derived cell lines at high levels and not to be expressed at all, or expressed at very low levels, in the remaining cell lines. The ROS1 gene was present in normal copy numbers in all cell lines that expressed the gene. However, in one particular glioblastoma line, we detected a potentially activating mutation at the ROS1 locus.

Rigorous feedback control of cAMP levels in Saccharomyces cerevisiae.
Nikawa, J. and Cameron, S. and Toda, T. and Ferguson, K. M. and Wigler, M. H. (1987) Genes & Development, 1(9) pp. 931-937.

We have isolated and characterized normal and mutant alleles of many of the genes of the RAS/adenylyl cyclase pathway of the yeast Saccharomyces cerevisiae. Manipulation of those genes has revealed a system for feedback control that can modulate cAMP levels over at least a 10,000-fold range. The feedback control depends upon the activity of the cAMP-dependent protein kinases and requires the presence of the CDC25 and RAS proteins. The capacity for such dramatic control of cAMP levels raises fundamental questions about the normal mechanism of action of the cAMP signaling system in yeast.

Cloning and characterization of the low-affinity cyclic AMP phosphodiesterase gene of Saccharomyces cerevisiae.
Nikawa, J. and Sass, P. and Wigler, M. H. (1987) Molecular & Cellular Biology, 7(10) pp. 3629-36.

Saccharomyces cerevisiae contains two genes which encode cyclic AMP (cAMP) phosphodiesterase. We previously isolated and characterized PDE2, which encodes a high-affinity cAMP phosphodiesterase. We have now isolated the PDE1 gene of S. cerevisiae, which encodes a low-affinity cAMP phosphodiesterase. These two genes represent highly divergent branches in the evolution of phosphodiesterases. High-copy-number plasmids containing either PDE1 or PDE2 can reverse the growth arrest defects of yeast cells carrying the RAS2(Val-19) mutation. PDE1 and PDE2 appear to account for the aggregate cAMP phosphodiesterase activity of S. cerevisiae. Disruption of both PDE genes results in a phenotype which resembles that induced by the RAS2(Val-19) mutation. pde1- pde2- ras1- ras2- cells are viable.

Human ROS1 and MAS1 oncogenes located in regions of chromosome 6 associated with tumor-specific rearrangements.
Rabin, M. and Birnbaum, D. and Young, D. and Birchmeier, C. and Wigler, M. H. and Ruddle, F. H. (1987) Oncogene Res, 1(2) pp. 169-78.

Oncogenes have been implicated in tumorigenesis based on their localization to chromosomal sites associated with tumor-specific structural rearrangements. We have mapped the human ros1 (formerly mcf3) and mas1 oncogenes to the distal half of chromosome 6q, within a region frequently rearranged in malignant cells. Chromosomal mapping of these two new human transforming genes may help elucidate the involvement of the long arm of chromosome 6 in diverse tumor types.

Cloning and characterization of BCY1, a locus encoding a regulatory subunit of the cyclic AMP-dependent protein kinase in Saccharomyces cerevisiae.
Toda, T. and Cameron, S. and Sass, P. and Zoller, M. and Scott, J. D. and McMullen, B. and Hurwitz, M. and Krebs, E. G. and Wigler, M. H. (1987) Molecular & Cellular Biology, 7(4) pp. 1371-1377.

We have cloned a gene (BCY1) from the yeast Saccharomyces cerevisiae that encodes a regulatory subunit of the cyclic AMP-dependent protein kinase. The encoded protein has a structural organization similar to that of the RI and RII regulatory subunits of the mammalian cyclic AMP-dependent protein kinase. Strains of S. cerevisiae with disrupted BCY1 genes do not display a cyclic AMP-dependent protein kinase in vitro, fail to grow on many carbon sources, and are exquisitely sensitive to heat shock and starvation.

The S. cerevisiae CDC25 gene product regulates the RAS/adenylate cyclase pathway .
Broek, D. and Toda, T. and Michaeli, T. and Levin, L. and Birchmeier, C. and Zoller, M. and Powers, S. and Wigler, M. H. (1987) Cell, 48(5) pp. 789-800.

The gene corresponding to the S. cerevisiae cell division cycle mutant cdc25 has been cloned and sequenced, revealing an open reading frame encoding a protein of 1589 amino acids that contains no significant homologies with other known proteins. Cells lacking CDC25 have low levels of cyclic AMP and decreased levels of Mg2+-dependent adenylate cyclase activity. The lethality resulting from disruption of the CDC25 gene can be suppressed by the presence of the activated RAS2val19 gene, but not by high copy plasmids expressing a normal RAS2 or RAS1 gene. These results suggest that normal RAS is dependent on CDC25 function. Furthermore, mutationally activated alleles of CDC25 are capable of inducing a set of phenotypes similar to those observed in strains containing a genetically activated RAS/adenylate cyclase pathway, suggesting that CDC25 encodes a regulatory protein. We propose that CDC25 regulates adenylate cyclase by regulating the guanine nucleotide bound to RAS proteins.

Guanine nucleotide activation of, and competition between, RAS proteins from Saccharomyces cerevisiae.
Field, J. and Broek, D. and Kataoka, T. and Wigler, M. H. (1987) Molecular & Cellular Biology, 7(6) pp. 2128-2133.

In the yeast Saccharomyces cerevisae, yeast RAS proteins are potent activators of adenylate cyclase. In the present work we measured the activity of adenylate cyclase in membranes from Saccharomyces cerevisae which overexpress this enzyme. The response of the enzyme to added RAS2 proteins bound with various guanine nucleotides and their analogs suggests that RAS2 proteins are active in their GTP-bound form and are virtually inactive in their GDP-bound form. Also, active RAS2 protein is not inhibited by inactive RAS2, suggesting that the inactive form does not compete with the active form in binding to its effector.

Expression in Escherichia coli of BCY1, the regulatory subunit of cyclic AMP-dependent protein kinase from Saccharomyces cerevisiae. .
Johnson, K. E. and Cameron, S. and Toda, T. and Wigler, M. H. and Zoller, M. J. (1987) Journal of Biological Chemistry, 262(18) pp. 8636-8642.

The regulatory (R) subunit of cAMP-dependent protein kinase from the yeast Saccharomyces cerevisiae was expressed in Escherichia coli by engineering the gene for yeast R, BCY1, in an E. coli expression vector that contained a promoter from phage T7. Oligonucleotide-directed mutagenesis was used to create an NdeI restriction site at the natural ATG of the yeast R. This facilitated construction of the T7 expression vector so that the sequence of the protein produced was identical to the natural R subunit. Yeast R was highly expressed in a soluble form. 20 mg of purified yeast R was obtained from 4 liters of E. coli. N-terminal amino acid sequencing revealed that the expressed protein began with the natural sequence. 60% of the molecules contained an N-terminal methionine, and 40% initiated with valine, the second amino acid of yeast R. The protein produced in E. coli migrated on a sodium dodecyl sulfate-polyacrylamide gel with an Mr of 52,000. The yeast R bound 2 mol of cAMP/mol of R monomer with a Kd of 76 nM. The protein was treated with urea to remove bound cAMP. Sedimentation values before and after the urea treatment were identical ( = 5.1). Addition of purified R subunit to a preparation of yeast C subunit (TPK1) rendered catalytic activity cAMP-dependent with an activity ratio of 4.6. The yeast R was autophosphorylated by yeast C to a level of 0.8 mol of phosphate/mol of R monomer. By these criteria, the R subunit produced in E. coli was structurally and functionally identical to the natural yeast R subunit and similar to mammalian type II R subunits.

Yeast Mating Pheromone a-Factor and Its Relationship to Ras Proteins.
Michaelis, S. and Anna, S. S. and Powers, S. and Wigler, M. H. and Herskowitz, I. (1987) Journal of Cellular Biochemistry, pp. 261-261.

Three different genes in Saccharomyces cerevisiae encode the catalytic subunits of the cAMP-dependent protein kinase .
Toda, T. and Cameron, S. and Sass, P. and Zoller, M. and Wigler, M. H. (1987) Cell, 50(2) pp. 277-288.

We have isolated three genes (TPK1, TPK2, and TPK3) from the yeast Saccharomyces cerevisiae that encode the catalytic subunits of the cAMP-dependent protein kinase. Gene disruption experiments demonstrated that no two of the three genes are essential by themselves but at least one TPK gene is required for a cell to grow normally. Comparison of the predicted amino acid sequences of the TPK genes indicates conserved and variable domains. The carboxy-terminal 320 amino acid residues have more than 75% homology to each other and more than 50% homology to the bovine catalytic subunit. The amino-terminal regions show no homology to each other and are heterogeneous in length. The TPK1 gene carried on a multicopy plasmid can suppress both a temperature-sensitive ras2 gene and adenylate cyclase gene.

Induction of Oocyte Maturation by Injection of Oncogenic H-Ras Protein in the Starfish.
Herlands, L. and Kishimoto, T. and Birchmeier, C. and Wigler, M. H. and Koide, S. S. (1986) Zoological Science, 3(6) pp. 1036-1036.

Characterization of two new human oncogenes.
Birchmeier, C. and Young, D. and Wigler, M. H. (1986) Cold Spring Harbor Symposia on Quantitative Biology, 51(2) pp. 993-1000.

The first oncogenes discovered were the transforming genes of the oncogenic viruses (reviewed by Bishop 1985). The subsequent discovery that the oncogenes of retroviruses were derived from normal host cellular genes provided the first direct evidence that cellular genomes contain genes with transforming potential. More recently, the development of techniques for DNA transfer in eukaryotic cells led to the discovery of cellular transforming genes in tumor cells by their ability to induce foci of transformed NIH-3T3 cells (reviewed by Land et al. 1983). Several new oncogenes have been discovered this way, including N-ras (Shimizu et al. 1983), met (Cooper et al. 1984), neu (Bargmann et al. 1986), and possible others (Goubin et al. 1983; Lane et al. 1984; Takahashi et al. 1985).

Characterization of an activated human ros gene.
Birchmeier, C. and Birnbaum, D. and Waitches, G. and Fasano, O. and Wigler, M. H. (1986) Molecular & Cellular Biology, 6(9) pp. 3109-3116.

A human oncogene, mcf3, previously detected by a combination of DNA-mediated gene transfer and a tumorigenicity assay, derives from a human homolog of the avian v-ros oncogene. Both v-ros and mcf3 can encode a protein with homology to tyrosine-specific protein kinases, and both mcf3 and v-ros encode a potential transmembrane domain N terminal to the kinase domain. mcf3 probably arose during gene tranfer from a normal human ros gene by the loss of a putative extracellular domain. There do not appear to be any other gross rearrangements in the structure of mcf3.

RAM, a gene of yeast required for a functional modification of RAS proteins and for production of mating pheromone a-factor.
Powers, S. and Michaelis, S. and Broek, D. and Ann-A, S. S. and Field, J. and Herskowitz, I. and Wigler, M. H. (1986) Cell, 47(3) pp. 413-422.

We have identified a gene (SUPH) of S. cerevisiae that is requried for both RAS function and mating by cells of a mating type. supH is allelic to ste16, a gene required for the production of the mating pheromone a-factor. Both RAS and a-factor coding sequences terminate with the potential acyltransferase recognition sequence Cys-A-A-X, where A is an aliphatic amino acid. Mutations in SUPH-STE16 prevent the membrane localization and maturation of RAS protein, as well as the fatty acid acylation of it and other membrane proteins. We propose the designation RAM (RAS protein and a-factor maturation function) for SUPH and STE16. RAM may encode an enzyme responsible for the modification and membrane localization of proteins with this C-terminal sequence.

Cloning and characterization of the high-affinity cAMP phosphodiesterase of Saccharomyces cerevisiae .
Sass, P. and Field, J. and Nikawa, J. and Toda, T. and Wigler, M. H. (1986) Proceedings of the National Academy of Sciences of the United States of America, 83(24) pp. 9303-9307.

A gene, PDE2, has been cloned from the yeast Saccharomyces cerevisiae that, when present in high copy, reverses the phenotypic effects of RAS2Val19, a mutant form of the RAS2 gene that renders yeast cells sensitive to heat shock and starvation. It has previously been shown that the RAS proteins are potent activators of yeast adenylate cyclase. We report here that PDE2 encodes a high-affinity cAMP phosphodiesterase that shares sequence homology with animal cell phosphodiesterases. These results therefore imply that the effects of RAS2Val119 are mediated through its changes in cAMP concentration.

Isolation and characterization of a new cellular oncogene encoding a protein with multiple potential transmembrane domains .
Young, D. and Waitches, G. and Birchmeier, C. and Fasano, O. and Wigler, M. H. (1986) Cell, 45(5) pp. 711-720.

We have cloned and sequenced a new human oncogene and have named it mas. This oncogene was detected by its tumorigenicity in nude mice using the cotransfection and tumorigenicity assay previously described. The mas oncogene has a weak focus-inducing activity in transfected NIH 3T3 cells. A DNA rearrangement in the 5' noncoding sequence, which occurred during transfection, is probably rsponsible for activation of the mas gene. The cDNA sequence of the mas oncogene reveals a long open reading frame that codes for a 325 amino acid protein. This protein is very hydrophobic and has seven potential transmembrane domains. In this respect, the structure of the mas protein is novel among cellular oncogene products and may reflect a new functional class of oncogenes.

RAS proteins can induce meiosis in Xenopus oocytes .
Birchmeier, C. and Broek, D. and Wigler, M. H. (1985) Cell, 43(3 PART) pp. 615-622.

Injection of human H-ras protein induces maturation of Xenopus oocytes; that is, progression from prophase to metaphase of meiosis. The oncogenic protein encoded by H-rasval12 is nearly a 100-fold more potent than the protein encoded by the wild-type gene. We do not observe any measurable increase or decrease in cyclic AMP concentration in injected oocytes, and the effects of H-ras protein are only partially blocked by cholera toxin. Our results suggest that not all, if any, of the effects of H-rasval12 protein in this system are mediated by adenylate cyclase.

DNA sequence and characterization of the S. cerevisiae gene encoding adenylate cyclase.
Kataoka, T. and Broek, D. and Wigler, M. H. (1985) Cell, 43(2) pp. 493-505.

We have cloned CYR1, the S. cerevisiae gene encoding adenylate cyclase. The DNA sequence of CYR1 can encode a protein of 2026 amino acids. This protein would contain a central region comprised of over twenty copies of a 23 amino acid repeating unit with remarkable homology to a 24 amino acid tandem repeating unit of a trace human serum glycoprotein. Gene disruption and biochemical experiments indicate that the catalytic domain of adenylate cyclase resides in the carboxyl terminal 400 amino acids. Elevated expression of adenylate cyclase suppresses the lethality that otherwise results from loss of RAS gene function in yeast. Yeast adenylate cyclase, made in E. coli, cannot be activated by added RAS protein.

Conservation and divergence of RAS protein function during evolution.
Birchmeier, C. and Broek, D. and Toda, T. and Powers, S. and Kataoka, T. and Wigler, M. H. (1985) Cold Spring Harbor Symposia on Quantitative Biology, 50 pp. 721-725.

The ras genes were first isolated as the transforming genes of Harvey and Kirsten sarcoma virus (Ellis et al. 1981). At least three different ras genes, Ha-ras, Ki-ras, and N-ras, exist in mammals and code for three very similar 21-kD proteins (Shimizu et al. 1983b). The ras proteins are localized in the plasma membrane (Willingham et al. 1980), bind guanine nucleotides (Shih et al. 1980, 1982), and have weak GTPase activity (Gibbs et al. 1984; McGrath et al. 1984; Sweet et al. 1984). A large number of tumor cells contain structurally mutated ras genes that are capable of tumorigenic transformation of NIH-3T3 cells upon DNA-mediated gene transfer (Reddy et al. 1982; Tabin et al. 1982; Taparowsky et al. 1982; Capon et al. 1983; Shimizu et al. 1983a; Yuasa et al. 1983)

Ras Proteins Function Exclusively to Modulate Adenylate-Cyclase Activity in the Yeast Saccharomyces.
Broach, J. and Toda, T. and Kataoka, T. and Powers, S. and Broek, D. and Wigler, M. H. (1985) DNA-a Journal of Molecular & Cellular Biology, 4(1) pp. 64-64.

Differential activation of yeast adenylate cyclase by wild type and mutant RAS proteins .
Broek, D. and Samily, N. and Fasano, O. and Fujiyama, A. and Tamanoi, F. and Northup, J. and Wigler, M. H. (1985) Cell, 41(3) pp. 763-770.

Purified RAS proteins, whether derived from the yeast RAS1 or RAS2 or the human H-ras genes, activate yeast adenylate cyclase in the presence of guanine nucleotides. These results confirm the prediction of earlier genetic and biochemical data and for the first time provide a complete biochemical assay for RAS protein function. A biochemical difference is observed between the RAS2 and RAS2val19 proteins in their ability to activate adenylate cyclase after preincubation with GTP.

Microinjection of Ras Oncogene Proteins and Inhibitory Antibodies Specific for the Ras Oncogene Proteins into Living Normal and Transformed-Cells.
Feramisco, J. and Sullivan, N. and Kamata, T. and Barsagi, D. and Welch, W. and Gross, M. and Sweet, R. and Rosenberg, M. and Sathe, G. and Fasano, O. and Goldfarb, M. and Wigler, M. H. and Nitecki, D. and Clark, R. and Arnheim, N. and McCormick, F. (1985) Federation Proceedings, 44(5) pp. 1768-1768.

Purification and characterization of human H-RAS proteins expressed in Escherichia coli.
Gross, M. and Sweet, R. W. and Sathe, G. and Yokoyama, S. and Fasano, O. and Goldfarb, M. and Wigler, M. H. and Rosenberg, M. (1985) Molecular and Cellular Biology, 5(5) pp. 1015-1024.

The full-length normal and T24 mutant human H-ras proteins and two truncated derivatives of the T24 mutant were expressed efficiently in Escherichia coli. The proteins accumulated to 1 to 5% of total cellular protein, and each was specifically recognized by anti-ras monoclonal antibodies. The two full-length proteins as well as a carboxyl-terminal truncated derivative (deleted for 23 amino acid residues) were soluble upon cell lysis and were purified to 90% homogeneity without the use of denaturants. In contrast, an amino-terminal truncated ras derivative (deleted for 22 amino acid residues) required treatment with urea for its solubilization. The guanine nucleotide binding activity of these four proteins was assessed by a combination of ligand binding on proteins blots, immunoprecipitation, and standard filter binding procedures. The full-length proteins showed similar binding kinetics and a stoichiometry approaching 1 mol of GTP bound per mol of protein. The showed similar binding kinetics and a stoichiometry approaching 1 mol of GTP bound per mol of protein. The carboxyl-terminal truncated protein also bound GTP, but to a reduced extent, whereas the amino-terminal truncated protein did not have binding activity. Apparently, the carboxyl-terminal domain of ras, although important for transforming function, does not play a critical role in GTP binding.

Functional homology of mammalian and yeast RAS genes.
Kataoka, T. and Powers, S. and Cameron, S. and Fasano, O. and Goldfarb, M. and Broach, J. and Wigler, M. H. (1985) Cell, 40(1) pp. 19-26.

Yeast spores lacking endogenous RAS genes will not germinate. If such spores contain chimeric mammalian/yeast RAS genes or even the mammalian H-ras gene under the control of the galactose-inducible GAL10 promoter, they will germinate in the presence of galactose and produce for continued growth and viability. The results indicate that the biochemical function of RAS proteins is essential for vegetative haploid yeast and that this function has been conserved in evolution since the progenitors of yeast and mammals diverged.

Control of Yeast Adenyl-Cyclase by Ras Proteins.
Powers, S. and Kataoka, T. and Toda, T. and Fasano, O. and Wigler, M. H. (1985) Federation Proceedings, 44(3) pp. R8-R8.

In yeast, RAS proteins are controlling elements of adenylate cyclase.
Toda, T. and Uno, I. and Ishikawa, T. and Powers, S. and Kataoka, T. and Broek, D. and Cameron, S. and Broach, J. and Matsumoto, K. and Wigler, M. H. (1985) Cell, 40(1) pp. 27-36.

Saccharomyces cerevisiae strains containing RAS2val19, a RAS2 gene with a missense mutation analogous to one that activates the transforming potential of mammalian ras genes, have growth and biochemical properties strikingly similar to yeast strains carrying IAC or bcy1. Yeast strains carrying the IAC mutation have elevated levels of adenylate cyclase activity. bcy1 is a mutation that suppresses the lethality in adenylate cyclase deficient yeast. Yeast strains deficient in RAS function exhibit properties similar to adenylate cyclase deficient yeast. bcy1 suppresses lethality in ras1- ras2- yeast. Compared to wild-type yeast strains, intracellular cAMP levels are significantly elevated in RAS2val19 strains, significantly depressed in ras2- strains, and virtually undetectable in ras1- ras2- bcy1 strains. Membranes from ras1- ras2- bcy1 yeast lack the GTP-stimulated adenylate cyclase activity present in membranes from wild-type cells, and membranes from RAS2val19 yeast strains have elevated levels of an apparently GTP-independent adenylate cyclase activity. Mixing membranes from ras1- ras2- yeast with membranes from adenylate cyclase deficient yeast reconstitutes a GTP-dependent adenylate cyclase.

New human transforming genes detected by a tumorigenicity assay.
Fasano, O. and Birnbaum, D. and Edlund, L. and Fogh, J. and Wigler, M. H. (1984) Molecular & Cellular Biology, 4(9) pp. 1695-1705.

A sensitive bioassay for transforming genes based on the tumorigenicity of cotransfected NIH3T3 cells in nude mice was developed. The assay differs substantially from the NIH3T3 focus assay. Using it, the transfer of 3 transforming genes from the DNA of MCF-7, a human mammary carcinoma cell line was detected. One of these is N-ras, which is amplified in MCF-7 DNA. The other two, which were called mcf2 and mcf3, do not appear to be related to known oncogenes. Their transfer could not be detected by using the NIH3T3 focus assay. Whether either mcf2 or mcf3 is associated with genetic abnormalities in MCF-7 cells is not yet known.

Genetic analysis of yeast RAS1 and RAS2 genes.
Kataoka, T. and Powers, S. and McGill, C. and Fasano, O. and Strathern, J. and Broach, J. and Wigler, M. H. (1984) Cell, 37(2) pp. 437-445.

We present a genetic analysis of RAS1 and RAS2 of S. cerevisiae, two genes that are highly homologous to mammalian ras genes. By constructing in vitro ras genes disrupted by selectable genes and introducing these by gene replacement into the respective ras loci, we have determined that neither RAS1 nor RAS2 are by themselves essential genes. However, ras1 - ras2 - spores of doubly heterozygous diploids are incapable of resuming vegetative growth. We have determined that RAS1 is located on chromosome XV, 7 cM from ade2 and 63 cM from his3; and RAS2 is located on chromosome XIV, 2 cM from met4 . We have also constructed by site-directed mutagenesis a missense mutant, RAS2val19 , which encodes valine in place of glycine at the nineteenth amino acid position, the same sort of missense mutation that is found in some transforming alleles of mammalian ras genes. Diploid yeast cells that contain this mutation are incapable of sporulating efficiently, even when they contain wild-type alleles.

Genes in Saccharomyces cerevisiae encoding proteins with domains homologous to the mammalian ras proteins.
Powers, S. and Kataoka, T. and Fasano, O. and Goldfarb, M. and Strathern, J. and Broach, J. and Wigler, M. H. (1984) Cell, 36(3) pp. 607-612.

The ras genes, which were first identified by their presence in RNA tumor viruses and which belong to a highly conserved gene family in vertebrates, have 2 close homologues in yeast, detectable by Southern blotting. Both genes (RAS1 and RAS2) were cloned from plasmid libraries and determined the complete nucleotide sequence of their coding regions. They encode proteins with nearly 90% homology to the first 80 positions of the mammalian ras proteins, and nearly 50% homology to the next 80 amino acids. Yeast RAS1 and RAS2 proteins are more homologous to each other, with about 90% homology for the first 180 positions. After this, at nearly the same position that the mammalian ras proteins begin to diverge from each other, the 2 yeast ras proteins diverge radically. The yeast ras proteins, like the proteins encoded by the mammalian genes, terminate with the sequence cysAAX, where A is an aliphatic amino acid. Thus the yeast ras proteins have the same overall structure and interrelationship as the family of mammalian ras proteins. The domains of divergence may correspond to functional domans of the ras proteins. Monoclonal antibody directed against mammalian ras proteins immunoprecipitates protein in yeast cells containing high copy numbers of the yeast RAS2 gene.

Analysis of the transforming potential of the human H-ras gene by random mutagenesis.
Fasano, O. and Aldrich, T. and Tamanoi, F. and Taparowsky, E. and Furth, M. and Wigler, M. H. (1984) Proceedings of the National Academy of Sciences of the United States of America, 81(13) pp. 4008-4012.

Some tumor cells contain mutant ras genes that are capable of transforming NIH 3T3 cells. Those genes that were analyzed arise from the wild-type, non-transforming ras genes by mutations producing single amino acid substitutions at position 12 or 61 of the encoded protein. Random bisulfite-induced mutagenesis was performed on the cloned wild-type human H-ras gene to find if mutations at other positions can activate the transforming potential of that gene. Most mutations are not activating, but mutations that specify single amino acid substitutions at position 12, 13, 59, or 63 if the encoded protein do activate the transforming potential of the H-ras gene. Some, but not all, mutant ras proteins show an altered electrophoretic mobility in NaDodSO4[sodium dodecyl sulfate]/polyacrylamide gels.

Use of DNA Polymorphisms to Establish the Linkage Map of the Pth, Beta-Globin, C-Ha-Ras-1, and Insulin Loci on 11p.
Meyers, D. A. and Antonarakis, S. E. and Fearon, E. R. and Mallonee, R. L. and Waber, P. G. and Phillips, J. A. and Kazazian, H. H. and Levine, M. and Kronenberg, H. M. and Wigler, M. H. and Ullrich, A. (1984) Cytogenetics and Cell Genetics, 37(1-4) pp. 540-541.

A product of yeast RAS-2 gene is a guanine nucleotide binding protein.
Tamanoi, F. and Walsh, M. and Kataoka, T. and Wigler, M. H. (1984) Proceedings of the National Academy of Sciences of the United States of America, 81(22) pp. 6924-6928.

Saccharomyces cerevisiae contains 2 genes, RAS1 and RAS2, which show remarkable homology to mammalian ras genes. To characterize these gene products, the RAS2 gene in yeast was expressed using an inducible GAL10 promoter. After labeling with [35S]methionine and immunoprecipitating with a monoclonal antibody Y13-259, which reacts with p21 encoded by mammalian ras genes, a major band having an apparent MW of 41,000 was detected. This band was also identified in cell-free translation products of polyadenylylated RNA extracted from yeast cells grown in the presence of galactose. Crude extracts of cells expressing the RAS2 gene exhibited guanine nucleotide binding activity. This was detected by incubation with [3H]GDP followed by immunoprecipitation with the antibody Y13-259. The binding of labeled GDP was inhibited by a 20-fold excess of GDP, GTP, and, to a lesser extent, by UTP, a characteristic similar to that possessed by the mammalian ras proteins. The activity of the yeast protein differs from that of the mammalian proteins in its strong dependence on temperature. The guanine nucleotide binding activity provides an assay to purify the yeast protein.

Human Transforming Genes.
Wigler, M. H. and Fasano, O. and Taparowsky, E. and Birnbaum, D. and Goldfarb, M. and Fogh, J. (1984) DNA-a Journal of Molecular & Cellular Biology, 3(1) pp. 65-65.

A test of the role of two oncogenes in inherited predisposition to colon cancer.
Barker, D. and McCoy, M. and Weinberg, R. and Goldfarb, M. and Wigler, M. H. and Burt, R. and Gardner, E. and White, R. (1983) Mol Biol Med, 1(2) pp. 199-206.

Inheritance of mutationally altered oncogenes could predispose individuals to the development of specific tumors and account for familial tumor phenotypes. Using adjacent DNA sequence polymorphisms as genetic markers, we have examined two oncogenes, the Kirsten ras2, isolated from a human colon cancer cell line, and the Harvey ras1, isolated from a human bladder cancer cell line, for their role in the genetic etiology of inherited colon cancer in Gardner syndrome. Both oncogene loci have been shown to be unlinked to the Gardner syndrome locus and are, therefore, eliminated as candidates for the Gardner syndrome gene.

Structure of the Ki-Ras Gene of the Human-Lung Carcinoma Cell-Line Calu-1.
Shimizu, K. and Birnbaum, D. and Ruley, M. A. and Fasano, O. and Suard, Y. and Edlund, L. and Taparowsky, E. and Goldfarb, M. and Wigler, M. H. (1983) Nature, 304(5926) pp. 497-500.

The homologue of the viral Kirsten ras (v-Ki-ras) gene found in the human lung carcinoma cell line, Calu-1, has an intron-exon structure similar to that of the human homologue of the viral Harvey ras (v-Ha-ras) gene. A second, potential fourth coding exon is present in the human Ki-ras gene and similar sequences are found in the Kirsten murine sarcoma virus. Cysteine is encoded at the twelfth amino acid position, suggesting that the Calu-1 Ki-ras gene has undergone a mutational activation at the same position as the human Ha-ras gene of the bladder carcinoma cell line, T24. A comparison of their predicted amino acid sequences suggests that ras proteins have a 'constant' region and a 'variable' region. Here we propose a common modular structure for ras gene products in which the variable region forms a physiologically important combining site.

Three human transforming genes are related to the viral ras oncogenes.
Shimizu, K. and Goldfarb, M. and Suard, Y. and Perucho, M. and Li, Y. and Kamata, T. and Feramisco, J. and Stavnezer, E. and Fogh, J. and Wigler, M. H. (1983) Proceedings of the National Academy of Sciences of the United States of America, 80(8) pp. 2112-2116.

Three distinct transforming genes present in human tumor cell lines are all related to the viral oncogenes of Harvey and Kirsten murine sarcoma viruses, designated v-H-ras and v-K-ras, respectively. The transforming gene of a bladder carcinoma cell line has been shown to be a human homolog to v-H-ras [Parada, L. F., Tabin, C. J., Shih, C. & Weinberg, R. A. (1982) Nature (London) 297, 474-478; Santos, E., Tronick, S. R., Aaronson, S. A., Pulciani, S. & Barbacid, M. (1982) Nature (London) 298, 343-347]. The transforming gene common to one colon (SK-CO-1) and two lung carcinoma (SK-LU-1 and Calu-1) cell lines is the same human homolog of v-K-ras as is the transforming gene previously identified in a lung carcinoma cell line Lx-1 [Der, C. J., Krontiris, T. G. & Cooper, G. M. (1982) Proc. Natl. Acad. Sci. USA 79, 3637-3640]. The transforming gene of SK-N-SH neuroblastoma cells is weakly homologous to both v-H-ras and v-K-ras. NIH 3T3 cells transformed with the SK-N-SH transforming gene contain increased levels of a protein serologically and structurally related to the protein products of the v-H-ras and v-K-ras genes. Therefore, it represents a third member of the ras gene family, which we have called N-ras. Based on the homology with the v-ras genes, we have established the orientation of transcription and approximate coding regions of the cloned human K-ras and N-ras genes.

Isolation and preliminary characterization of the transforming gene of a human neuroblastoma cell line.
Shimizu, K. and Goldfarb, M. and Perucho, M. and Wigler, M. H. (1983) Proc Natl Acad Sci U S A, 80(2) pp. 383-387.

DNA from the human neuroblastoma cell line SK-N-SH is capable of inducing foci of transformed NIH 3T3 cells after DNA-mediated gene transfer. using genetic selection with the Escherichia coli sup F gene, we have isolated human sequences from mouse cells responsible for the oncogenic transformation. These sequences are present in all human DNAs surveyed and no gross rearrangements of these sequences are found in SK-N-SH cells. Although clearly distinct from two other human transforming genes present in bladder, lung, and colon carcinoma cell lines, all three transforming gene sequences may be related members of the ras gene family.

Sequence and Structure of the Coding Regions of the Human H-Ras-1 Gene from T-24 Bladder Carcinoma Cells.
Fasano, O. and Taparowsky, E. and Fiddes, J. and Wigler, M. H. and Goldfarb, M. (1983) Journal of Molecular & Applied Genetics, 2(2) pp. 173-180.

c[complementary]DNA was molecularly cloned and sequenced to the transcript of H-ras-1, the transforming gene of the T24 human bladder carcinoma cell line. The transcript derives from at least 5 exons in the H-ras-1 gene, and RNA splicing occurs at sites typical of exon-intron junctions. T24 H-ras-1 RNA has an AUG-initiated open reading frame of 567 nucleotides, which can encode a protein of mass comparable to the apparent MW of the T24 H-ras-1 gene product. The T24 H-ras-1 gene product is nearly identical to v-H-ras p21, the transforming protein encoded by the genome of Harvey sarcoma virus. The implications of this sequence conservation in the structure-function relationships of ras proteins are discussed.

Mutational analysis of the cloned chicken thymidine kinase gene.
Kwoh, T. J. and Zipser, D. and Wigler, M. H. (1983) J Mol Appl Genet, 2(2) pp. 191-200.

We have analyzed the transcription and coding unit of the chicken thymidine kinase (tk) gene. We have constructed a library of mutant chicken tk genes by the in vitro linker insertion method of Heffron et al. A total of 125 mutations within a 3.0 kbp HindIII fragment containing the gene were isolated and mapped. The effect of each mutation upon the thymidine kinase gene was determined by measurement of the transfection efficiency in mouse Ltk- cells. The chicken tk mRNA is about 2 kb and polyadenylated. The direction of transcription was also determined. From these results, we propose a structure for the gene in which at least three small introns separate the amino acid coding region into at least four segments.

Chromosomal Assignment of a Family of Human Oncogenes.
Ryan, J. and Barker, P. E. and Shimizu, K. and Wigler, M. H. and Ruddle, F. H. (1983) Proc Natl Acad Sci U S A, 80(14) pp. 4460-4463.

A family of human transforming genes, previously shown to share homology with the ras family of viral oncogenes, maps to three different human chromosomes. A well-characterized mouse-human hybrid cell panel, combined with Southern blotting, was used in this study. The transforming gene of the T24 bladder carcinoma cell line maps to human chromosome 11. An oncogene isolated from the lung carcinoma cell line SK-Calu-1 maps to human chromosome 12. The third ras-related gene, cloned from SK-N-SH, a neuroblastoma cell line, maps to human chromosome 1.

Structures of 3 Ras Genes Activated in Human-Tumor Cell-Lines.
Shimizu, K. and Taparowsky, E. and Fasano, O. and Suard, Y. and Birnbaum, D. and Ruley, M. A. and Goldfarb, M. and Wigler, M. H. (1983) Journal of Cell Biology, 97(5) pp. A69-A69.

Structure and Activation of the Human N-Ras Gene.
Taparowsky, E. and Shimizu, K. and Goldfarb, M. and Wigler, M. H. (1983) Cell, 34(2) pp. 581-586.

The normal human N-ras gene was cloned. In structure and sequence it closely resembles the human H-ras and K-ras genes. The 3 genes share regions of nucleotide homology and nucleotide divergence within coding sequences and have a common intron/exon structure, indicating that they evolved from a similarly spliced ancestral gene. The N-ras gene of SK-N-SH neuroblastoma cells has transforming activity, while the normal N-ras gene does not, the result of a single nucleotide change substituting Lys for Gln in position 61 of the N-ras gene product. Amino acid substitutions in 2 distinct regions can activate the transforming potential of ras gene products.

Human Transforming Genes.
Wigler, M. H. and Shimizu, K. and Goldfarb, M. and Taparowsky, E. and Suard, Y. and Fasano, O. (1983) DNA-a Journal of Molecular & Cellular Biology, 2(1) pp. 64-64.

Isolation and Preliminary Characterization of a Human Transforming Gene from T24 Bladder-Carcinoma Cells.
Goldfarb, M. and Shimizu, K. and Perucho, M. and Wigler, M. H. (1982) Nature, 296(5856) pp. 404-409.

DNA from T24, a cell line derived from a human bladder carcinoma, can induce the morphological transformation of NIH 3T3 cells. Using techniques of gene rescue to clone the gene responsible for this transformation, we have found that it is human in origin, less than 5 kilobase pairs in size and is homologous to a 1,100-base polyadenylated RNA species found in T24 and HeLa cells. Blot analysis indicates extensive restriction endonuclease polymorphism near this gene, in human DNAs.

Activation of the T24 Bladder-Carcinoma Transforming Gene Is Linked to a Single Amino-Acid Change.
Taparowsky, E. and Suard, Y. and Fasano, O. and Shimizu, K. and Goldfarb, M. and Wigler, M. H. (1982) Nature, 300(5894) pp. 762-765.

Several different transforming genes have been observed in the DNA of a variety of tumours and tumour cell lines of human and rodent origin by the ability of these genes to induce morphological transformation in NIH 3T3 cells1-5. The transforming gene found in a human bladder carcinoma cell line, T24, is H-ras-1, the human homologue of the Harvey sarcoma virus oncogene (v-H-ras)6-9. In the present study we have compared the H-ras-1 genes cloned from T24 and normal human DNA. The H-ras-1 gene cloned from T24 DNA induces transformation in NIH 3T3 cells, while the same gene cloned from normal cellular DNA does not. The functionally significant difference between the transforming and normal genes appears to be a single base mutation, which produces an amino acid change in the sequence of the proteins that the genes encode.

Human-tumor-derived cell lines contain common and different transforming genes.
Perucho, M. and Goldfarb, M. and Shimizu, K. and Lama, C. and Fogh, J. and Wigler, M. H. (1981) Cell, 27(3 Pt 2) pp. 467-76.

We have screened different cultured cell lines established from human tumors for the ability of their DNAs to induce transformed foci in NIH/3T3 cells. Based on restriction endonuclease digestions and the presence of human sequences in mouse transformants, we conclude that five of these human tumor cell lines contain a gene or genes capable of transforming mouse cells and that at least three different transforming genes are present in these five lines. Three cell lines, two derived from lung carcinomas and one derived from a colon carcinoma, transfer the same or closely related human genes. If these transforming genes are mediating the tumorigenic state of the human cells, then our results indicate that overlapping pathways leading to tumorigenesis may arise independently.

Human-tumor-derived cell lines contain common and different transforming genes.
Perucho, M. and Goldfarb, M. and Shimizu, K. and Lama, C. and Fogh, J. and Wigler, M. H. (1981) Cell, 27(3 Pt 2) pp. 467-76.

We have screened different cultured cell lines established from human tumors for the ability of their DNAs to induce transformed foci in NIH/3T3 cells. Based on restriction endonuclease digestions and the presence of human sequences in mouse transformants, we conclude that five of these human tumor cell lines contain a gene or genes capable of transforming mouse cells and that at least three different transforming genes are present in these five lines.

Linkage and expression of foreign DNA in cultured animal cells.
Perucho, M. and Wigler, M. H. (1981) Cold Spring Harb Symp Quant Biol, 45(2) pp. 829-838.

The ability to introduce defined genes into cultured cells radically alters the prospects for understanding differentiation and growth control in higher organisms. Stable transformation of animal cells may be effected by the delivery of DNA as a calcium phosphate coprecipitate (Graham and van der Eb 1973). Since the uptake and expression of DNA is a relatively rare event, transformants are usually isolated by selection for cells that have acquired a new phenotype. For this purpose the thymidine kinase (tk)1 gene of herpes simplex virus type 1 (HSV-1) is often used as the selectable vector with tk+ recipient cells (Wigler et al. 1977). The tk+ transformants are then selected in HAT medium. Nonselectable genes may be stably introduced into cells either by direct ligation to the selectable vector (Mantei et al. 1979) or by unlinked cotransformation (Wigler et al. 1979b).

Linkage and expression of foreign DNA in cultured animal cells.
Perucho, M. and Wigler, M. (1981) Cold Spring Harb Symp Quant Biol, 45 Pt pp. 829-38.

The inheritance of methylation patterns in vertebrates.
Wigler, M. H. (1981) Cell, 24(2) pp. 285-286.

The somatic replication of DNA methylation.
Wigler, M. H. and Levy, D. and Perucho, M. (1981) Cell, 24(1) pp. 33-40.

The hypothesis that DNA methylation patterns are replicated in the somatic cells of vertebrates was tested. M-Hpa II, the modification enzyme from Hemophilus parainfluenzae, was used to methylate bacteriophage .vphi.X174 RF DNA and the cloned chicken thymidine kinase (tk) gene in vitro and then to introduce these DNA and unmethylated controls into tk- cultured mouse cells by DNA-mediated transformation. Twenty-five cell generations later, the state of methylation of transferred DNA was examined by restriction endonuclease analysis and blot hybridization. Methylation at Hpa II sites is replicated by these cultured cells but not with 100% fidelity. Methylation of the cloned chicken tk gene decreases its apparent transformation efficiency relative to unmethylated molecules.

Characteristics of an Sv40-Plasmid Recombinant and Its Movement into and out of the Genome of a Murine Cell.
Hanahan, D. and Lane, D. and Lipsich, L. and Wigler, M. H. and Botchan, M. (1980) Cell, 21(1) pp. 127-139.

A bacterial plasmid carrying the early region of SV40 (pOT) has been stably established in high molecular weight (hmw) DNA of mouse L cells by selection for the herpes virus thymidine kinase (tk) gene. DNA blotting has demonstrated that most cell lines contain multiple discrete copies of pOT, generally with an intact SV40 early region. No free copies of pOT have been detected. Both pOT and tk sequences may be amplified up to 20-200 copies of the SV40 early region. In contrast to the uniform staining pattern normally observed in SV40-transformed lines, indirect immunofluorescence using antiserum to the SV40 T antigen has demonstrated that the expression of the early region is heterogeneous in these cell lines. This fraction expressing T is characteristic of a given cell line, and varies from 0 to 99% positive. Several pOT cell lines have been fused to simian cells, and replicating low molecular weight DNAs were isolated from the heterokaryons. Transformation of E. coli with this DNA demonstrates that pOT can be rescued from hmw DNA in L cells and reestablished as a plasmid in E. coli. Excision is generally precise when pOT is introduced to the murine cells as supercoiled molecule, and imprecise when pOT is introduced in linear form.

Transformation of mammalian cells with an amplifiable dominant acting gene.
Wigler, M. H. and Perucho, M. and Kurtz, D. and Dana, S. and Pellicer, A. and Axel, R. and Silverstein, S. (1980) Proceedings of the National Academy of Sciences of the United States of America, 77(6) pp. 3567-3570.

We have transferred a mutant hamster gene coding for an altered dihydrofolate reductase to wild-type cultured mouse cells by using total genomic DNA from methotrexate-resistant Chinese hamster ovary A29 cells as donor. By demonstrating the presence of hamster gene sequences in transformants we have provided direct evidence for gene transfer. Transformants selected for increased resistance to methotrexate contain increased amounts of the newly transferred gene. We have used this mutant dhfr gene to introduce the Escherichia coli antibiotic resistance plasmid pBR322 into animal cells. Amplification of the dhfr sequences results in amplification of the pBR322 sequences as well. The use of this gene may allow the introduction and amplification of virtually any genetic element in various new cellular environments.

Isolation of the chicken thymidine kinase gene by plasmid rescue.
Perucho, M. and Hanahan, D. and Lipsich, L. and Wigler, M. H. (1980) Nature, 285(5762) pp. 207-210.

The bacterial plasmid pBR322 was used as a vehicle to isolate genes coding for selectable markers from higher eukaryotes. The chicken thymidine kinase (tk) gene was obtained as a 2.2-kilobase EcoRI/HindIII insert in pBR322. The cloned gene transforms tk- animal cells with an efficiency equal to that of the cloned herpes simplex virus-1 tk gene.

Construction of Co Transformation of Human Cell Lines Which Express Segments of the Adenovirus Genome.
Klessig, D. F. and Wigler, M. H. and Grodzicker, T. (1980) Journal of Supramolecular Structure, 9(SUPPL.) pp. 256.

Genetic and Physical Linkage of Exogenous Sequences in Transformed-Cells.
Perucho, M. and Hanahan, D. and Wigler, M. H. (1980) Cell, 22(1) pp. 309-317.

The calcium phosphate precipitation method of Graham and van der Eb (1973) is an efficient means of introducing DNA into cultured animal cells. Cells which incorporate one selectable marker are also likely to incorporate sequences from the carrier DNA. Both selected and unselected markers are found integrated in the high molecular weight nuclear DNA of the host. In the present study, we demonstrate that exogenously acquired sequences are gentically linked, segregating and amplifying coordinately, and that their flanking sequences derive primarily from the carrier species rather than the host species. Based on these results, we propose that, upon transformation, the host cell ligates incorporated DNA into a large concatameric structure which may at times be as large as 2000 kilobases. From blotting data alone we cannot determine whether this structure is chromosomal or extrachromosomal in location.

DNA Mediated Gene Transfer: Theory and Application.
Wigler, M. H. and Perucho, M. and Lipsich, L. and Lane, D. and Klessig, D. and Grodzicker, T. and Botchan, M. and Hanahan, D. (1980) Journal of Supramolecular Structure, 9(SUPPL.) pp. 229.

Introduction and expression of a rabbit Beta-globin gene in mouse fibroblasts.
Wold, B. and Wigler, M. H. and Lacy, E. and Maniatis, T. and Silverstein, S. and Axel, R. (1979) Proceedings of the National Academy of Sciences of the United States of America, 76(11) pp. 5684-5688.

DNA-Mediated Transfer of the Adenine Phosphoribosyltransferase Locus into Mammalian Cells.
Wigler, M. H. and Pellicer, A. and Silverstein, S. and Axel, R. and Urlaub, G. and Chasin, L. (1979) Proc Natl Acad Sci U S A, 76(3) pp. 1373-1376.

In this report, we demonstrate the feasibility of transforming mouse cells deficient in adenine phosphoribosyltransferase (aprt; AMP:pyrophosphate phosphoribosyltransferase, EC to the aprt+ phenotype by means of DNA-mediated gene transfer. Transformation was effected by using unfractionated high molecular weight genomic DNA from Chinese hamster, human, and mouse cells and restriction endonuclease-digested DNA from rabbit liver. The transformation frequency observed was between 1 and 10 colonies per 106 cells per 20 {micro} g of donor DNA. Transformants displayed enzymatic activity that was donor derived as demonstrated by isoelectric focusing of cytoplasmic extracts. These transformants fall into two classes: those that are phenotypically stable when grown in the absence of selective pressure and those that are phenotypically unstable under the same conditions.

Transformation of mammalian cells with genes from procaryotes and eucaryotes.
Wigler, M. H. and Sweet, R. and Sim, G. K. and Wold, B. and Pellicer, A. and Lacy, E. and Maniatis, T. and Silverstein, S. and Axel, R. (1979) Cell, 16(4) pp. 777-785.

We have stably transformed mammalian cells with precisely defined procaryotic and eucaryotic genes for which no selective criteria exist. The addition of a purified viral thymidine kinase (tk) gene to mouse cells lacking this enzyme results in the appearance of stable transformants which can be selected by their ability to grow in HAT. These biochemical transformants may represent a subpopulation of competent cells which are likely to integrate other unlinked genes at frequencies higher than the general population. Co-transformation experiments were therefore performed with the viral tk gene and bacteriophage [Phi]X174, plasmid pBR322 or the cloned chromosomal rabbit [beta]-globin gene sequences. Tk+ transformants were cloned and analyzed for co-transfer of additional DNA sequences by blot hybridization. In this manner, we have identified mouse cell lines which contain multiple copies of 4)X, pBR322 and the rabbit [beta]-globin gene sequences. The [Phi]X co-transformants were studied in greatest detail. The frequency of co-transformation is high: 15 of 16 tk+ transformants contain the [Phi]X sequences. Selective pressure was required to identify co-transformants. From one to more than fifty [Phi]X sequences are integrated into high molecular weight nuclear DNA isolated from independent clones. Analysis of subclones demonstrates that the [Phi]X genotype is stable through many generations in culture. This co-transformation system should allow the introduction and stable integration of virtually any defined gene into cultured cells. Ligation to either viral vectors or selectable biochemical markers is not required.

Biochemical transfer of single-copy eucaryotic genes using total cellular DNA as donor.
Wigler, M. H. and Pellicer, A. and Silverstein, S. and Axel, R. (1978) Cell, 14(3) pp. 725-731.

Previous studies from our laboratories have demonstrated the feasibility of transferring the thymidine kinase (tk) gene from restriction endonuclease-generated fragments of herpes simplex virus (HSV) DNA to cultured mammalian cells. In this study, high molecular weight DNA from cells containing only one copy of the HSV gene coding for tk was successfully used to transform Ltk- cells to the tk+ phenotype. The acquired phenotype was demonstrated to be donor-derived by analysis of the electrophoretic mobility of the tk activity, and the presence of HSV DNA sequences in the recipient cells was demonstrated. In companion experiments, we used high molecular weight DNA derived from tissues and cultured cells of a variety of species to transfer tk activity. The tk+ mouse cells transformed with human DNA were shown to express human type tk activity as determined by isoelectric focusing.

Induction of Plasminogen Activator in Cultured Cells by Macrocyclic Plant Diterpene Esters and Other Agents Related to Tumor Promotion.
Wigler, M. H. and DeFeo, D. and Weinstein, I. B. (1978) Cancer Res, 38(5) pp. 1434-1437.

In vitro systems that are responsive to tumor-promoting agents may facilitate the identification of such agents and the analysis of their mode of action. We have previously reported that the potent tumor promoter phorbol-12-myristate-13-acetate induces the synthesis of the enzyme plasminogen activator in cultured chick embryo fibroblasts. We have, therefore, tested various compounds for their ability to induce plasminogen activator in chicken embryo fibroblasts. Among these, phorbol esters and other macrocyclic diterpene esters isolated from species of the families Euphorbiaceae and Thymelaeaceae were potent inducers of plasminogen activator. These compounds maximally induced enzyme to the same levels, although they differed in their relative molar potencies. Structural requirements for in vitro activity paralleled the requirements for activity in vivo. These results indicate that induction of plasminogen activator is a useful marker for the biologically active macrocyclic diterpene esters. On the other hand, tumor-promoting agents such as anthralin, cantharidin, Tween 60, and tobacco leaf extract failed to induce plasminogen activator.

The transfer and stable integration of HSV thymidine kinase gene into mouse cells.
Pellicer, A. and Wigler, M. H. and Axel, R. and Silverstein, S. (1978) Cell, 14(1) pp. 133-141.

Cell culture studies provide new information on tumour promoters .
Weinstein, I. B. and Wigler, M. H. (1977) Nature, 270(5639) pp. 659-550.

Transfer of Purified Herpes-Virus Thymidine Kinase Gene to Cultured Mouse Cells.
Wigler, M. H. and Silverstein, S. and Lee, L. S. and Pellicer, A. and Cheng, Y. C. and Axel, R. (1977) Cell, 11(1) pp. 223-232.

Treatment of Ltk−, mouse L cells deficient in thymidine kinase (tk), with Bam I restriction endonuclease cleaved DNA from herpes simplex virus-1 (HSV-1) produced tk+ clones with a frequency of 10−6/2 μg of HSV-1 DNA. Untreated cells or cells treated with Eco RI restriction endonuclease fragments produced no tk+ clones under the same conditions. The thymidine kinase activities of four independently derived clones were characterized by biochemical and serological techniques. By these criteria, the tk activities were found to be identical to HSV-1 tk and different from host wildtype tk. The tk+ phenotype was stable over several hundred cell generations, although the rate of reversion to the tk− phenotype, as judged by cloning efficiency in the presence of bromodeoxyuridine, was high (1–5 × 10−3). HSV-1 DNA Bam restriction fragments were separated by gel electrophoresis, and virtually all activity, as assayed by transfection, was found to reside in a 3.4 kb fragment. Transformation efficiency with the isolated fragment is 20 fold higher per gene equivalent than with the unfractionated total Bam digest. These results prove the usefulness of transfection assays as a means for the bioassay and isolation of restriction fragments carrying specific genetic information. Cells expressing HSV-1 tk may also provide a useful model system for the detailed analysis of eucaryotic and viral gene regulation.

Nucleosomes in metaphase chromosomes.
Wigler, M. H. and Axel, R. (1976) Nucleic Acids Res, 3(6) pp. 1463-1461.

Previous studies of the structure of metaphase chromosomes have relied heavily on electron micrography and have revealed the existence of a 10-nm unit fiber that is thought to generate the native 23-30-nm fiber by higher order folding. The structural relationship of these metaphase fibers to the interphase fiber remains obscure. Recent studies on the digestion of interphase chromatin have revealed the existence of a regularly repeating subunit of DNA and histone, the nucleosome that generates the appearance of 10-nm beads connected by a short fiber of DNA seen on electron micrographs. It was therefore of interest to probe the structure of the metaphase chromosome for the presence of nucleosomal subunits. To this end metaphase chromosomes were prepared from colchicine-arrested cultures of mouse L-cells and were subjected to digestion with stayphylococcal nuclease. Comparison of the early and limit digestion products of metaphase chromosomes with those obtained from interphase nuclei indicates that although significant morphologic changes occur within the chromatin fiber during mitosis, the basic subunit structure of the chromatin fiber is retained by the mitotic chromosome

Nucleosomes in metaphase chromosomes.
Wigler, M. H. and Axel, R. (1976) Nucleic Acids Res, 3(6) pp. 1463-1461.

Previous studies of the structure of metaphase chromosomes have relied heavily on electron micrography and have revealed the existence of a 10-nm unit fiber that is thought to generate the native 23-30-nm fiber by higher order folding.

Tumor Promotor Induces Plasminogen Activator.
Wigler, M. H. and Weinstein, I. B. (1976) Nature, 259(5540) pp. 232-233.

INFECTION of chick embryo fibroblasts with Rous sarcoma virus (RSV) induces a cell-specified plasminogen activator 1. Induction occurs with transforming viruses but not with lytic viruses or with oncornaviruses which are not themselves transforming 2. Similarly, many mammalian cell lines and embryo cultures transformed with either viruses or chemical carcinogens may be high producers of plasminogen activator in contrast to their untransformed counterparts 3–6. A correlation has been demonstrated between production of plasminogen activator and various features of the transformed phenotype, such as cell locomotion, morphology and loss of anchorage-dependent growth 4,7. Several established cell lines which are not highly tumorigenic or transformed by the usual criteria are, however, active producers of plasminogen activator; there are examples also of transformed or tumorigenic cell lines which do not produce significant levels of plasminogen activator 3,8–10.

Tumor Promotor Induces Plasminogen Activator.
Wigler, M. H. and Weinstein, I. B. (1976) Nature, 259(5540) pp. 232-233.

Infection of chick embryo fibroblasts with Rous sarcoma virus (RSV) induces a cell-specified plasminogen activator1. Induction occurs with transforming viruses but not with lytic viruses or with oncornaviruses which are not themselves transforming.

Preparative Method for Obtaining Enucleated Mammalian-Cells.
Wigler, M. H. and Weinstein, I. B. (1975) Biochemical and Biophysical Research Communications, 63(3) pp. 669-674.

Mouse L cells can be enucleated in suspension by centrifugation in discontinuous Ficoll density gradients while in the presence of Cytochalasin B. Greater than 50% of the cytoplasts thus obtained attach to glass or plastic and undergo morphologic recovery within 2–4 hours of replating. Protein synthesis in cytoplasts undergoes a biphasic decay from an initial rate of approximately 50% of control nucleated cells. This method can yield up to 5 × 108 cytoplasts with consistently low levels of contamination by nucleated cells (less than 0.2%), and is well suited for obtaining quantitative amounts of cytoplasts or karyoplasts for physiologic or biochemical studies.