Somatic microindels in human cancer: the insertions are highly error-prone and derive from nearby but not adjacent sense and antisense templates.

Somatic microindels (microdeletions with microinsertions) have been studied in normal mouse tissues using the Big Blue lacI transgenic mutation detection system. Here we analyze microindels in human cancers using an endogenous and transcribed gene, the TP53 gene. Microindel frequency, the enhancement of 1-2 microindels and other features are generally similar to that observed in the non-transcribed lacI gene in normal mouse tissues. The current larger sample of somatic microindels reveals recurroids: mutations in which deletions are identical and the co-localized insertion is similar. The data reveal that the inserted sequences derive from nearby but not adjacent sequences in contrast to the slippage that characterizes the great majority of pure microinsertions. The microindel inserted sequences derive from a template on the sense or antisense strand with similar frequency. The estimated error rate of the insertion process of 13% per bp is by far the largest reported in vivo, with the possible exception of somatic hypermutation in the immunoglobulin gene. The data constrain possible mechanisms of microindels and raise the question of whether microindels are 'scars' from the bypass of large DNA adducts by a translesional polymerase, e.g. the 'Tarzan model' presented herein.

A mouse model for nonsense mutation bypass therapy shows a dramatic multiday response to geneticin.

Aminoglycosides can bypass nonsense mutations and are the prototypic agents for translational bypass therapy (TBT). Initial results demonstrate the need for more potent drugs and an in vivo model system for quantitative assessment of TBT. Herein, we present an in vivo system for evaluating the efficacy of premature stop codon management therapies: in vivo quantitative stop codon management repli-sampling TBT efficacy assay (IQSCMaRTEA). Application of IQSCMaRTEA reveals that geneticin is much more efficacious in vivo than gentamicin. Treatment with geneticin elicits a multiday response, and residual F9 antigen can be detected after 3 weeks. These data demonstrate the utility of IQSCMaRTEA for evaluating drugs that bypass nonsense mutations. In addition, IQSCMaRTEA may be helpful for testing inhibitors of nonsense-mediated decay, as stop codon management therapy will sometimes require inhibition of nonsense-mediated decay and translational bypass of the nonsense mutation. Furthermore, geneticin, its metabolites, or better tolerated analogues should be evaluated as a general treatment with multiday response for severe genetic disease caused by nonsense mutation.

Evidence for mutation showers.

Mutants in the Big Blue transgenic mouse system show spontaneous clustered multiple mutations with unexpectedly high frequency, consistent with chronocoordinate events. We tested the prediction that the multiple mutations seen within the lacI mutation target sometimes occur in the context of chronocoordinate multiple mutations spanning multiple kilobases (mutation showers). Additional sequencing of mutants was performed in regions immediately flanking the lacI region (total of 10.7 kb). Nineteen additional mutations were found outside the lacI region ("ectomutations") from 10 mutants containing two or more lacI mutations, whereas only one ectomutation was found in 130 mutants with a single mutation (P < 0.0001). The mutation showers had an average of approximately one mutation per 3 kb. Four mutants showed closely spaced double mutations in the new sequence, and analysis of the spacing between these mutations revealed significant clustering (P = 0.0098). To determine the extent of the mutation showers, regions (8.5 kb total) remote from the lacI region (approximately 16-17 kb away) were sequenced. Only two additional ectomutations were found in these remote regions, consistent with mutation showers that generally do not extend more than approximately 30 kb. We conclude that mutation showers exist and that they constitute at least 0.2% and possibly 1% or more of mutational events observed in this system. The existence of mutation showers has implications for oncogenesis and evolution, raising the possibilities of "cancer in an instant" and "introns as sponges to reduce the deleterious impact of mutation showers."

Database of somatic mutations in EGFR with analyses revealing indel hotspots but no smoking-associated signature.

We created an Epidermal Growth Factor Receptor (EGFR) Mutation Database (http://www.cityofhope.org/cmdl/egfr_db) that curates a convenient compilation of somatic EGFR mutations in non-small-cell lung cancer (NSCLC) and associated epidemiological and methodological data, including response to the tyrosine kinase inhibitors Gefitinib and Erlotinib. Herein, we analyze 809 mutations collected from 26 publications. Four super hotspots account for 70% of reported mutations while two-thirds of 131 unique mutations have been reported only once and account for only 11% of reported mutations. Consistent with strong biological selection for gain of function, the reported mutations are virtually all missense substitutions or in-frame microdeletions, microinsertions, or microindels (colocalized insertion and deletion with a net gain or loss of 1-50 nucleotides). Microdeletions and microindels are common in a region of exon 19. Microindels, which account for 8% of mutations, have smaller inserted sequences (95% are 1 to 5 bp) and are elevated 16-fold relative to mouse somatic microindels and to human germline microindels. Microdeletions/microindels are significantly more frequent in responders to Gefitinib or Erlotinib (P = 0.003). In addition, EGFR mutations in smokers do not carry signatures of mutagens in cigarette smoke. Otherwise, the mutation pattern does not differ significantly with respect to gender, age, or tumor histology. The EGFR Mutation Database is a central resource of EGFR sequence variant data for clinicians, geneticists, and other researchers. Authors are encouraged to submit new publications with EGFR sequence variants to be included in the database or to provide direct submissions via The WayStation submission and publication process (http://www.centralmutations.org).

Somatic microindels: analysis in mouse soma and comparison with the human germline.

Microindels, defined as mutations that result in a colocalized microinsertion and microdeletion with a net gain or loss of between 1 and 50 nucleotides, may be an important contributor to cancer. We report the first comprehensive analysis of somatic microindels. Our large database of mutations in the lacI transgene of Big Blue((R)) mice contains 0.5% microindels, 2.8% pure microinsertions, and 11.5% pure microdeletions. There appears to be no age, gender, or tissue-type specificity in the frequency of microindels. Of the independent somatic mutations that result in a net in-frame insertion or deletion, microindels are responsible for 13% of protein expansions and 6% of protein contractions. These in-frame microindels may play a crucial role in oncogenesis and evolution via "protein tinkering" (i.e., modest expansion or contraction of proteins). Four characteristics suggest that microindels are caused by unique mechanisms, not just simple combinations of the same mechanisms that cause pure microinsertions and pure microdeletions. First, microinsertions and microdeletions commonly occur at hotspots, but none of the 30 microindels are recurrent. Second, the sizes of the deletions and insertions in microindels are larger and more varied than in pure microdeletions and pure microinsertions. Third, microinsertions overwhelmingly repeat the adjacent base (97%) while the insertions in microindels do so only infrequently (17%). Fourth, analysis of the sequence contexts of microindels is consistent with unique mechanisms including recruitment of translesion DNA synthesis polymerases. The mouse somatic microindels have characteristics similar to those of human germline microindels, consistent with similar causative mechanisms in mouse and human, and in soma and germline.

Most spontaneous tumors in a mouse model of Li-Fraumeni syndrome do not have a mutator phenotype.

Mutations are the substrate of cancer. Yet, little is known about the degree and nature of mutations in tumors because measurement of mutation load in tumors and normal tissues was generally not possible until the advent of transgenic mouse mutation detection systems. Herein, we present the first analysis of mutation frequency and pattern in thymic tumors from a mouse model of Li-Fraumeni syndrome (p53+/- murine model) using the Big Blue assay with sequencing of all mutants. We also make the first characterization of mutation frequency and pattern in p53-deficient extra-thymic cancers. The data more than triple the literature on all non-mismatch repair deficient tumors for which mutations are identified by sequence analysis, allowing mutation frequency and pattern to be determined. Most tumors had a normal mutation frequency and a normal mutation pattern. Five tumors showed modest increases in mutation frequency (2.3-fold or less). Alterations in mutation patterns were uncommon, tumor-specific and not necessarily associated with increases in mutation frequency. Given the data from two spontaneous tumors (normal mutation frequency with an abnormal pattern in a p53-/- mouse and low mutation frequency in a p53+/+ control mouse), we hypothesize that tumors sometimes can carry a low mutation load. The study was not without certain caveats: mutation load could not be compared between tumor and normal tissue from the same animal; sample sizes for extra-thymic tumor types were small, and only point mutations and deletions, insertions and indels up to 2 kb were detected. However, the data clearly show key differences in tumors from p53+/- mice compared with mismatch repair deficient tumors; a lack of dramatic increase in mutation frequency and absence of a signature of mutation.

Preferential occurrence of 1-2 microindels.

Microindels are unique, infrequent mutations that result in inserted and deleted sequences of different sizes (between one and 50 nucleotides) at the same nucleotide position. Little is known about the mutational mechanisms that are responsible for these mutations. From our database of 6,016 independent somatic mutational events in the lacI gene in Big Blue mice, we assembled the 30 microindels (0.5%) for analysis. Microindels with one nucleotide inserted and two nucleotides deleted (1-2 microindels) accounted for seven (23%) of the microindels observed, with the remaining microindels distributed among 21 other combinations of insertion and deletion sizes. A preferential occurrence of 1-2 microindels (20%) was also observed in human germline transmitted mutations in the Human Gene Mutation Database (HGMD). An examination of the sequence flanking the mouse 1-2 microindels did not reveal obvious site specificity or associated secondary structure. A detailed examination of 1-2 microindels did not reveal the features typical of pure microinsertion and microdeletion events, but rather suggested a unique mutational mechanism. The 1 bp insertion in 1-2 microinsertions, and pure 1 bp insertions show distinct features. The mechanism for 1-2 microindels is not obviously a simple combination of pure microinsertion and microdeletion events. The dramatic enhancement of 1-2 microindels requires explanation. We speculate that certain error-prone polymerases may be responsible for the preferential occurrence of 1-2 microindels in both somatic tissues and germ cells. It is estimated that a human adult carries roughly 400 billion somatic 1-2 microindels with the potential to predispose to cancer.

Spontaneous multiple mutations show both proximal spacing consistent with chronocoordinate events and alterations with p53-deficiency.

Analysis of spontaneous multiple mutations in normal and tumor cells may constrain hypotheses about the mechanisms responsible for multiple mutations and provide insight into the mutator phenotype. In a previous study, spontaneous doublets in Big Blue mice were dramatically more frequent than expected by chance and exhibited a mutation pattern similar to that observed for single mutations [Mutat. Res. 452 (2000) 219]. The spacing between mutations in doublets was generally closer than expected by chance and the distribution of mutation spacing fit an exponential, albeit with substantial scatter. We now analyze 2658 additional mutants and confirm that doublets are enhanced dramatically relative to chance expectation. The spacing, frequency and pattern of spontaneous doublets and multiplets (domuplets) are examined as a function of age, tissue type, p53-deficiency and neoplasia in the new and combined data. The new and combined data confirm that the distribution of the spacing between mutations in doublets is non-random with the mutations more closely spaced than expected by chance (P < 0.0005; combined data), consistent with temporally coordinate (chronocoordinate) events. An exponential provides an excellent fit to the distribution (R2 = 0.98) and estimates that half of doublets have mutations separated by 120 nucleotides or less (the "half-life of mutation spacing"). We make several novel observations: (i) singlets and doublets show similar overall increases in frequency with age (ii) doublet frequency may be lower in the male germline, consistent with the generally reduced mutation frequency in the male germline (iii) doublet frequencies are elevated in somatic tissues of p53-deficient mice (Li-Fraumini cancer syndrome model; P = 0.005) and (iv) doublets and singlets in tumors from p53-deficient mice have a different mutation pattern (P = 0.007). The observations are consistent with chronocoordinate occurrence of spontaneous doublets and multiplets due to a transient error-prone condition and do not suggest a major role for the recently discovered Y family of error-prone polymerases. The enhancement of doublets in p53-deficient mice may contribute to cancer risk.

Spontaneous mutation in Big Blue mice from fetus to old age: tissue-specific time courses of mutation frequency but similar mutation types.

Transgenic mouse mutation detection systems permit rapid determination of the frequency and type of mutations allowing direct examination of mutational markers for aging, neurodegeneration, and cancer. The Big Blue transgenic mouse mutation detection system was used to determine the frequency and nature of spontaneous mutations versus age in multiple tissue types. Nuclear DNA was extracted from whole fetus at 13.5 days postcoitus (dpc) and from six tissues postbirth (cerebellum, forebrain, thymus, liver, adipose tissue, and male germline) of Big Blue transgenic mice at four ages: 10 days and at 3, 10, and 25 months postbirth. Forty million total plaque-forming units (pfu) were screened. The time course of mutation frequency with age had a significantly different shape in different tissues (P < 10(-6)). By 13.5 dpc, the whole fetus mutation frequency had already started increasing from the theoretical zero at conception to a value that was about one-half the mid-adulthood (3-10 months) average. From 10 days to 3 months, mutation frequency increased significantly in liver (P = 0.007) and showed an increasing trend in cerebellum, forebrain, and thymus. From 3 to 10 months, there was no significant change in mutation frequency in any tissue examined. From 10 to 25 months, the mutation frequency increased significantly in liver (P < 10(-6)) and adipose tissue (P = 0.002), but not in the other tissues examined (cerebellum, forebrain, and male germline). It is of interest that the mutation frequency in the male germline is consistently the lowest, remaining essentially unchanged in old age. The spectrum of mutation types was unaltered with age, tissue type and gender, although, as previously reported, tandem GG-->TT mutations are tissue specific and show significant increases with age and certain hotspots (Buettner VL et al. [1999]: Environ Mol Mutagen 33:320-324; Hill KA et al. [2003]: Mutat Res 534:173-186). The spectrum of mutation types was generally the same for all tissue types, despite the tissue-specific increases in mutation frequency with age. These data provide a useful reference for future studies of endogenous and exogenous mutagenesis.

Mutations in the factor IX gene (F9) during the past 150 years have relative rates similar to ancient mutations.

Pollutants and dietary mutagens have been associated with somatic mutation and cancer, but the extent of their influence on germline mutation is not clear. Since deleterious germline mutations can be transmitted for thousands of years, any influence on germline mutation from the vast increase in man-made chemicals of the past 150 years would be an important public health issue. Observed disease causing mutations in the X-linked factor IX gene (F9) of hemophilia B patients originated predominantly in the past 150 years, since the half-life of these mutations in human populations had been about two generations before effective treatment became available about a generation ago. Recent changes in germline mutational processes may be detected by comparison of the observed hemophilia B causing mutation pattern in F9 with the pattern of neutral polymorphisms which occurred over a much longer period of time. By scanning a total of 1.5 megabases of deep intronic regions of F9 in the genomic DNA from 84 individuals, 42 neutral polymorphisms were found in 23 haplotypes that differed by at least 11 mutations from the ancestral primate haplotype. By sequencing F9 in seven non-human primates, 39 of these polymorphisms were characterized as ancient mutations relative to a unanimous ancestral primate allele. This ancient mutation pattern was compared to the recent pattern of hemophilia B causing mutations. Remarkably, no significant difference was found (P=0.5), suggesting that the vast increase in man-made chemicals during the past 150 years has not had a major impact on the pattern of human germline mutation. This result is consistent with the hypothesis that endogenous processes dominate germline mutation.