High frequency of homoplasmic mitochondrial DNA mutations in human tumors can be explained without selection.

Researchers in several laboratories have reported a high frequency of homoplasmic mitochondrial DNA (mtDNA) mutations in human tumors. This observation has been interpreted to reflect a replicative advantage for mutated mtDNA copies, a growth advantage for a cell containing certain mtDNA mutations, and/or tumorigenic properties of mtDNA mutations. We consider another possibility-that the observed homoplasmy arose entirely by chance in tumor progenitor cells, without any physiological advantage or tumorigenic requirement. Through extensive computer modeling, we demonstrate that there is sufficient opportunity for a tumor progenitor cell to achieve homoplasmy through unbiased mtDNA replication and sorting during cell division. To test our model in vivo, we analyzed mtDNA homoplasmy in healthy human epithelial tissues and discovered that the model correctly predicts the considerable observed frequency of homoplasmic cells. Based on the available data on mitochondrial mutant fractions and cell division kinetics, we show that the predicted frequency of homoplasmy in tumor progenitor cells in the absence of selection is similar to the reported frequency of homoplasmic mutations in tumors. Although a role for other mechanisms is not excluded, random processes are sufficient to explain the incidence of homoplasmic mtDNA mutations in human tumors.

Molecular classification of cancer: class discovery and class prediction by gene expression monitoring.

Although cancer classification has improved over the past 30 years, there has been no general approach for identifying new cancer classes (class discovery) or for assigning tumors to known classes (class prediction). Here, a generic approach to cancer classification based on gene expression monitoring by DNA microarrays is described and applied to human acute leukemias as a test case. A class discovery procedure automatically discovered the distinction between acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) without previous knowledge of these classes. An automatically derived class predictor was able to determine the class of new leukemia cases. The results demonstrate the feasibility of cancer classification based solely on gene expression monitoring and suggest a general strategy for discovering and predicting cancer classes for other types of cancer, independent of previous biological knowledge.

Mutational spectra of a 100-base pair mitochondrial DNA target sequence in bronchial epithelial cells: a comparison of smoking and nonsmoking twins.

Seventeen separate mitochondrial hot spot mutations in a 100-bp target sequence (mitochondrial bp 10,030-10,130) were detected and measured in bronchial epithelial cell samples isolated from smokers and nonsmokers. Among the individuals sampled were three pairs of monozygotic twins in which one twin had never smoked and had a nonsmoking spouse, and the other had a smoking history of >10 pack-years. Individual point mutations present at frequencies as low as 10(-6) were detected. Partially denaturing electrophoresis was used to separate mutant from nonmutant sequences on the basis of their melting temperatures, and the target sequence was subsequently amplified via high-fidelity PCR with Pfu DNA polymerase. Tests were performed to determine whether mismatch intermediates or DNA adducts present in the cellular DNA were converted to mutants during PCR. Hot spot mutations were clearly observed in bronchial epithelial cells, and the same hot spots were observed consistently in different samples. Significant numerical variability in the mutant fractions for individual mutants was observed among samples and are ascribed to unequal mitochondrial segregation in stem and transition cells. The mutational spectra in smokers' samples did not differ significantly from the mutational spectra in nonsmokers' samples for this 100 bp of mitochondrial DNA. No smoking-specific hot spots were detected. The overall mutant fractions in smokers' samples were not elevated compared to those of nonsmokers. As much variability was observed between two samples from the same individual's lung as between a sample from a smoker and a sample from a nonsmoker. These findings demonstrate that inhaled tobacco smoke does not induce prominent point mutations in this 100-bp target mitochondrial sequence in smokers' bronchial epithelial cells. Endogenous factors (e.g., DNA replication errors or DNA damage by endogenous reactive chemicals) are suggested to be more likely to represent the most important contributors to mitochondrial mutagenesis.

Chemically induced mutations in mitochondrial DNA of human cells: mutational spectrum of N-methyl-N'-nitro-N-nitrosoguanidine.

We have observed a reproducible mitochondrial mutational spectrum in the MT1 human lymphoblastoid line treated with N-methyl-N'-nitro-N-nitrosoguanidine (MNNG). The MNNG spectrum was distinct from the spontaneous mutational spectrum. However, our ability to observe MNNG-induced mitochondrial mutations above the high level of accumulated spontaneous mutations was dependent on the MT1 phenotype. MT1 cells are markedly resistant to the cytotoxicity but not the mutagenicity of MNNG, presumably as a result of inactivation of both copies of the hMSH6 (GTBP) mismatch repair gene. Thus, we were able to use conditions of treatment that yielded induced mitochondrial mutant fractions beyond the practical limits for human cell experiments in mismatch-proficient human cell lines. In contradistinction, when MT1 cells were treated repeatedly with maximum tolerated concentrations of (+/-) anti-benzo(a)pyrene diol-epoxide, no induced mitochondrial mutations above the spontaneous background were observed. A single dose of 4 microM MNNG (survival, 0.85) induced a mutant fraction of 8 x 10(-3) in the nuclear hypoxanthine-guanine phosphoribosyltrans. ferase gene, and a clear and reproducible pattern of seven MNNG-induced hotspot mutations was observed within the mitochondrial DNA target sequence studied (mitochondrial bp 10,030-10,130). All of the MNNG-induced hotspot mutations were G:C to A:T transitions present at frequencies between 6 x 10(-5) and 30 x 10(-5). Additional experiments supported the conclusion that MNNG-induced hotspot mutations observed were generated in living cells as a result of MNNG treatment and not from mismatch intermediates or DNA adducts converted into mutations during the PCR process.

Statistical analysis of repetitive subcloning by the limiting dilution technique with a view toward ensuring hybridoma monoclonality.

After empirically determining the adequacy of the Poisson equation for describing the distribution of particles in microtiter plate wells under the experimental conditions employed for subcloning hybridoma cells, an equation was developed for determining the probability of monoclonality after repetitive subclonings. A table derived from the equation and based on the percentage of wells showing growth allows the rapid evaluation of the likelihood of monoclonality.

Efficiency of separation of DNA mutations by constant denaturant capillary electrophoresis is controlled by the kinetics of DNA melting equilibrium.

Constant denaturant capillary electrophoresis (CDCE) separation takes place in the heated portion of the capillary where faster-moving, unmelted DNA fragments are in equilibrium with slower-moving, partially melted forms. Within a certain temperature range, the position of the melting equilibrium and thus the average electrophoretic mobility of each mutant is different. The resulting differences in mobility allow sequences containing single base pair point mutations to be separated from each other. We report the results of experiments in which we explored the rules defining separation efficiency by varying the parameters of CDCE. We discovered an unusual peak broadening mechanism. In contrast to most other DNA electrophoresis systems, peak width in CDCE steadily decreases with the square root of the separation speed. Moreover, the peak width displays a sharp maximum at a specific temperature. To account for these observations, we use a model which describes CDCE separation as a random walk. According to this model, peaks in CDCE are broad because the kinetics of the melting equilibrium are slow and therefore the number of random walk steps represented by melting/renaturation transitions is relatively small. In addition to providing a satisfactory interpretation of the data, the model also predicts that separation efficiency will increase as the ionic strength of the running buffer is increased and as the concentration of denaturant in the buffer is decreased. These predictions were verified and were used to establish conditions for high-resolution CDCE suitable for separating complex mixtures of single base pair mutants.

Identification of point mutations in mixtures by capillary electrophoresis hybridization.

We have developed a rapid method for unambiguous identification and mutant fraction determination of individual mutants in mixtures of DNA sequence variants each differing by one or a few nucleotides. This method has applications to such diverse areas as interpretation of mutational spectra, screening of populations for polymorphisms and identification of species in environmental mixtures. In our approach, a mixture of unknown sequences labeled with a fluorescent dye is combined with a set of predetermined sequences (standards) representing the variants to be assayed. Labeling the standards with another dye allows the two sets of variants to be measured independently. Using constant denaturing capillary electrophoresis, the sequence variants are separated as individual peaks on the basis of differential melting equilibria. The unknown sequence variants are initially identified based on co-migration with particular standards. This preliminary identification is verified by hybridization of the unknown variants with the co-migrating standards within the capillary. We demonstrate the use of capillary electrophoresis hybridization to dissect complex mutational spectra of human cells in culture.

Expression analysis with oligonucleotide microarrays reveals that MYC regulates genes involved in growth, cell cycle, signaling, and adhesion.

MYC affects normal and neoplastic cell proliferation by altering gene expression, but the precise pathways remain unclear. We used oligonucleotide microarray analysis of 6,416 genes and expressed sequence tags to determine changes in gene expression caused by activation of c-MYC in primary human fibroblasts. In these experiments, 27 genes were consistently induced, and 9 genes were repressed. The identity of the genes revealed that MYC may affect many aspects of cell physiology altered in transformed cells: cell growth, cell cycle, adhesion, and cytoskeletal organization. Identified targets possibly linked to MYC's effects on cell growth include the nucleolar proteins nucleolin and fibrillarin, as well as the eukaryotic initiation factor 5A. Among the cell cycle genes identified as targets, the G1 cyclin D2 and the cyclin-dependent kinase binding protein CksHs2 were induced whereas the cyclin-dependent kinase inhibitor p21(Cip1) was repressed. A role for MYC in regulating cell adhesion and structure is suggested by repression of genes encoding the extracellular matrix proteins fibronectin and collagen, and the cytoskeletal protein tropomyosin. A possible mechanism for MYC-mediated apoptosis was revealed by identification of the tumor necrosis factor receptor associated protein TRAP1 as a MYC target. Finally, two immunophilins, peptidyl-prolyl cis-trans isomerase F and FKBP52, the latter of which plays a role in cell division in Arabidopsis, were up-regulated by MYC. We also explored pattern-matching methods as an alternative approach for identifying MYC target genes. The genes that displayed an expression profile most similar to endogenous Myc in microarray-based expression profiling of myeloid differentiation models were highly enriched for MYC target genes.

Mitochondrial mutational spectra in human cells and tissues.

We have found that human organs such as colon, lung, and muscle, as well as their derived tumors, share nearly all mitochondrial hotspot point mutations. Seventeen hotspots, primarily G --> A and A --> G transitions, have been identified in the mitochondrial sequence of base pairs 10,030-10,130. Mutant fractions increase with the number of cell generations in a human B cell line, TK6, indicating that they are heritable changes. The mitochondrial point mutation rate appears to be more than two orders of magnitude higher than the nuclear point mutation rate in TK6 cells and in human tissues. The similarity of the hotspot sets in vivo and in vitro leads us to conclude that human mitochondrial point mutations in the sequence studied are primarily spontaneous in origin and arise either from DNA replication error or reactions of DNA with endogenous metabolites. The predominance of transition mutations and the high number of hotspots in this short sequence resembles spectra produced by DNA polymerases in vitro.

Mutational spectrometry without phenotypic selection: human mitochondrial DNA.

By first separating mutant from nonmutant DNA sequences on the basis of their melting temperatures and then increasing the number of copies by high-fidelity DNA amplification, we have developed a method that allows observation of point mutations in biological samples at fractions at or above 10-6. Using this method, we have observed the hotspot point mutations that lie in 100 base pairs of the mitochondrial genome in samples of cultured cells and human tissues. To date, 19 mutants have been isolated, their fractions ranging from 4x10-4 down to the limit of detection. We performed specific tests to determine if the observed signals were artefacts arising from contamination, polymerase errors during PCR or DNA adducts created during the procedure. We also tested the possibilities that DNA replication mismatch intermediates, or endogenous DNA adducts that were originally present in the cells, were included with true mutants in our separation steps and converted to mutants during PCR. We show that while most of the mutants behave as double-stranded point mutants in the cells, some appear to arise at least in part from mismatch intermediates or cellular DNA adducts. This technology is therefore sufficient for the observation of the spectrum of point mutations in human mitochondrial DNA and is a tool for discovering the primary causes of these mutations.

Cell-by-cell scanning of whole mitochondrial genomes in aged human heart reveals a significant fraction of myocytes with clonally expanded deletions.

Quantitative information on the cell-to-cell distribution of all possible mitochondrial DNA (mtDNA) mutations in young and aged tissues is needed to assess the relevance of these mutations to the aging process. In the present study, we used PCR amplification of full-length mitochondrial genomes from single cells to scan human cardiomyocytes for all possible large deletions in mtDNA. Analysis of more than 350 individual cells that were derived from three middle-aged and four centenarian donors demonstrates that while most of the cells contain no deletions, in certain cardiomyocytes a significant portion of the mtDNA molecules carried one particular deletion. Different affected cells contained different deletions. Although similar numbers of cells were screened for each donor, these deletion-rich cells were found only in the hearts of old donors, where they occurred at a frequency of up to one in seven cells. These initial observations demonstrate the efficiency of the method and indicate that mitochondrial mutations have the potential to play an important role in human myocardial aging.

Chemosensitivity prediction by transcriptional profiling.

In an effort to develop a genomics-based approach to the prediction of drug response, we have developed an algorithm for classification of cell line chemosensitivity based on gene expression profiles alone. Using oligonucleotide microarrays, the expression levels of 6,817 genes were measured in a panel of 60 human cancer cell lines (the NCI-60) for which the chemosensitivity profiles of thousands of chemical compounds have been determined. We sought to determine whether the gene expression signatures of untreated cells were sufficient for the prediction of chemosensitivity. Gene expression-based classifiers of sensitivity or resistance for 232 compounds were generated and then evaluated on independent sets of data. The classifiers were designed to be independent of the cells' tissue of origin. The accuracy of chemosensitivity prediction was considerably better than would be expected by chance. Eighty-eight of 232 expression-based classifiers performed accurately (with P < 0.05) on an independent test set, whereas only 12 of the 232 would be expected to do so by chance. These results suggest that at least for a subset of compounds genomic approaches to chemosensitivity prediction are feasible.