Run Of Guanines Research Articles

G-quadruplex propensity in H. neanderthalensis, H. sapiens and Denisovans mitochondrial genomes.

Current methods of processing archaeological samples combined with advances in sequencing methods lead to disclosure of a large part of H. neanderthalensis and Denisovans genetic information. It is hardly surprising that the genome variability between modern humans, Denisovans and H. neanderthalensis is relatively limited. Genomic studies may provide insight on the metabolism of extinct human species or lineages. Detailed analysis of G-quadruplex sequences in H. neanderthalensis and Denisovans mitochondrial DNA showed us interesting features. Relatively similar patterns in mitochondrial DNA are found compared to modern humans, with one notable exception for H. neanderthalensis. An interesting difference between H. neanderthalensis and H. sapiens corresponds to a motif found in the D-loop region of mtDNA, which is responsible for mitochondrial DNA replication. This area is directly responsible for the number of mitochondria and consequently for the efficient energy metabolism of cell. H. neanderthalensis harbor a long uninterrupted run of guanines in this region, which may cause problems for replication, in contrast with H. sapiens, for which this run is generally shorter and interrupted. One may propose that the predominant H. sapiens motif provided a selective advantage for modern humans regarding mtDNA replication and function.

Identification and characterization of a new G-quadruplex forming region within the kRAS promoter as a transcriptional regulator

kRAS is one of the most prevalent oncogenic aberrations. It is either upregulated or mutationally activated in a multitude of cancers, including pancreatic, lung, and colon cancers. While a significant effort has been made to develop drugs that target kRAS, their clinical activity has been disappointing due to a variety of mechanistic hurdles. The presented works describe a novel mechanism and molecular target to downregulate kRAS expression — a previously undescribed G-quadruplex (G4) secondary structure within the proximal promoter acting as a transcriptional silencer. There are three distinct guanine-rich regions within the core kRAS promoter, including a previously examined region (G4near). Of these regions, the most distal region does not form an inducible and stable structure, whereas the two more proximal regions (termed near and mid) do form strong G4s. G4near is predominantly a tri-stacked structure with a discontinuous guanine run incorporated; G4mid consists of seven distinct runs of continuous guanines and forms numerous competing isoforms, including a stable three-tetrad stacked mixed parallel and antiparallel loop structures with longer loops of up to 10 nucleotides. Comprehensive analysis of the regulation of transcription by higher order structures has revealed that the guanine-rich region in the middle of the core promoter, termed G4mid, is a stronger repressor of promoter activity than G4near. Using the extensive guanine-rich region of the kRAS core promoter, and particularly the G4mid structure, as the primary target, future drug discovery programs will have potential to develop a potent, specifically targeted small molecule to be used in the treatment of pancreatic, ovarian, lung, and colon cancers.

Solvent Effects on Ionization Potentials of Guanine Runs and Chemically Modified Guanine in Duplex DNA: Effect of Electrostatic Interaction and Its Reduction due to Solvent

We examined the ionization potential (IP) corresponding to the free energy of a hole on duplex DNA by semiempirical molecular orbital theory with a continuum solvent model. As for the contiguous guanines (a guanine run), we found that the IP in the gas phase significantly decreases with the increasing number of nucleotide pairs of the guanine run, whereas the IP in water (OP, oxidation potential) only slightly does. The latter result is consistent with the experimental result for DNA oligomers in water. This decrease in the IP is mainly due to the attractive electrostatic interaction between the hole and a nucleotide pair in the duplex DNA. This interaction is reduced in water, which results in the small decrease in the IP in water. This mechanism explains the discrepancy between the experimental result and the previous computational results obtained by neglecting the solvent. As for the chemically modified guanine, the previous work showed that the removal of some solvent (water) molecules due to the attachment of a neutral functional group to a guanine in a duplex DNA stabilizes the hole on the guanine. One might naively have expected the opposite case, since a polar solvent usually stabilizes ions. This mechanism also explains this unexpected stabilization of a hole as follows. When some water molecules are removed, the attractive electrostatic interaction stabilizing the hole increases, and thus, the hole is stabilized. In order to design the hole energetics by a chemical modification of DNA, this mechanism has to be taken into account and can be used.

End-stacking of copper cationic porphyrins on parallel-stranded guanine quadruplexes

Nucleic acids that contain multiple sequential guanines assemble into guanine quadruplexes (G-quadruplexes). Drugs that induce or stabilize G-quadruplexes are of interest because of their potential use as therapeutics. Previously, we reported on the interaction of the Cu(2+) derivative of 5,10,15,20-tetrakis(1-methyl-4-pyridyl)-21H,23H-porphine (CuTMpyP4), with the parallel-stranded G-quadruplexes formed by d(T(4)G( n )T(4)) (n = 4 or 8) (Keating and Szalai in Biochemistry 43:15891-15900, 2004). Here we present further characterization of this system using a series of guanine-rich oligonucleotides: d(T(4)G( n )T(4)) (n = 5-10). Absorption titrations of CuTMpyP4 with all d(T(4)G( n )G(4)) quadruplexes produce approximately the same bathochromicity (8.3 +/- 2 nm) and hypochromicity (46.2-48.6%) of the porphyrin Soret band. Induced emission spectra of CuTMpyP4 with d(T(4)G( n )T(4))(4) quadruplexes indicate that the porphyrin is protected from solvent. Electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry revealed a maximum porphyrin to quadruplex stoichiometry of 2:1 for the shortest (n = 4) and longest (n = 10) quadruplexes. Electron paramagnetic resonance spectroscopy shows that bound CuTMpyP4 occupies magnetically noninteracting sites on the quadruplexes. Consistent with our previous model for d(T(4)G(4)T(4)), we propose that two CuTMpyP4 molecules are externally stacked at each end of the run of guanines in all d(T(4)G( n )T(4)) (n = 4-10) quadruplexes.

Parallel-Stranded Guanine Quadruplex Interactions with a Copper Cationic Porphyrin

G-quadruplexes are formed by association of DNA strands containing multiple contiguous guanines. The capability of drugs to induce formation of or stabilize G-quadruplexes is an active area of investigation. We report the interactions of CuTMpyP4, the Cu(2+) derivative of 5,10,15,20-tetrakis(1-methyl-4-pyridyl)-21H,23H-porphine, with the parallel-stranded G-quadruplexes formed by d(T(4)G(4)T(4)) (1) and d(T(4)G(8)T(4)) (3). Absorption titrations of CuTMpyP4 with (1)(4) or (3)(4) cause both bathochromicity and hypochromicity of the porphyrin Soret band, with larger changes observed for the longer oligonucleotide. An approximate binding constant for (1)(4) and CuTMpyP4 according to the Scatchard model is 5.6 x 10(6) M(-)(1) in terms of quadruplexes and according to the McGhee-von Hippel model is 1.3 x 10(6) M(-)(1) in terms of potential binding sites. An approximate binding constant for (3)(4) and CuTMpyP4 according to the Scatchard model is 5.2 x 10(7) M(-)(1) in terms of quadruplexes and in terms of the McGhee-von Hippel model is 2.4 x 10(6) M(-)(1) in terms of potential binding sites. The site size for CuTMpyP4 and (1)(4) is four using the McGhee-von Hippel model. We find a 2:1 binding stoichiometry for CuTMpyP4 and (1)(4) and a 3:1 binding stoichiometry for CuTMpyP4 and (3)(4) using the method of continuous variation analysis. Induced emission spectra of CuTMpyP4 with (1)(4) or (3)(4) indicate a mode of binding in which the ligand is protected from the solvent. Electron paramagnetic resonance spectra of CuTMpyP4 with added oligonucleotide show an increase in the Cu-N superhyperfine coupling constant as the length of the oligonucleotide increases. On the basis of these data, we propose that for both (1)(4) and (3)(4), CuTMpyP4 molecules externally stack at each end of the run of guanines, similar to other planar G-quadruplex ligands. For (3)(4), our data are consistent with intercalation of a CuTMpyP4 molecule into the G-quadruplex.

Translational compensation of a frameshift mutation affecting herpes simplex virus thymidine kinase is sufficient to permit reactivation from latency.

Herpes simplex virus thymidine kinase is important for reactivation of virus from its latent state and is a target for the antiviral drug acyclovir. Most acyclovir-resistant isolates have mutations in the thymidine kinase gene; however, how these mutations confer clinically relevant resistance is unclear. Reactivation from explanted mouse ganglia was previously observed with a patient-derived drug-resistant isolate carrying a single guanine insertion within a run of guanines in the thymidine kinase gene. Despite this mutation, low levels of active enzyme were synthesized following an unusual ribosomal frameshift. Here we report that a virus, generated from a pretherapy isolate from the same patient, engineered to lack thymidine kinase activity, was competent for reactivation. This suggested that the clinical isolate contains alleles of other genes that permit reactivation in the absence of thymidine kinase. Therefore, to establish whether thymidine kinase synthesized via a ribosomal frameshift was sufficient for reactivation under conditions where reactivation requires this enzyme, we introduced the mutation into the well-characterized strain KOS. This mutant virus reactivated from latency, albeit less efficiently than KOS. Plaque autoradiography revealed three phenotypes of reactivating viruses: uniformly low thymidine kinase activity, mixed high and low activity, and uniformly high activity. We generated a recombinant thymidine kinase-null virus from a reactivating virus expressing uniformly low activity. This virus did not reactivate, confirming that mutations in other genes that would influence reactivation had not arisen. Therefore, in strains that require thymidine kinase for reactivation from latency, low levels of enzyme synthesized via a ribosomal frameshift can suffice.

High-frequency phenotypic reversion and pathogenicity of an acyclovir-resistant herpes simplex virus mutant.

A double-guanine-insertion mutation within a run of guanines in the herpes simplex virus gene encoding thymidine kinase (TK) was previously found in an acyclovir-resistant clinical isolate. This mutation was engineered into strain KOS, and stocks were generated from single plaques. Plaque autoradiography revealed that most plaques in such stocks exhibited low levels of TK activity, while approximately 3% of plaques exhibited high levels of TK activity, indicating a remarkably high frequency of phenotypic reversion. This virus was able to reactivate from latency in mouse ganglia; a fraction of the reactivating virus expressed a high level of TK activity due to an additional G insertion, suggesting that the observed genetic instability contributed to pathogenicity.

Comparison of the mutant frequencies and mutation spectra of three non-genotoxic carcinogens, oxazepam, phenobarbital, and Wyeth 14,643, at the λcII locus in Big Blue® transgenic mice

Oxazepam (OX), a widely used benzodiazepine anxiolytic, phenobarbital (PHE), a drug used for convulsive disorders, and Wyeth 14,643 (WY; [4-chloro-6-(2,3-xylidino)-2-pyrimidinylthio]acetic acid), a hypolipidemic agent, are all hepatocarcinogenic in B6C3F1 mice. They have been classified as “non-genotoxic” carcinogens since they are non-DNA reactive in in vitro assays and are either negative or weakly positive in Salmonella typhimurium (Ames assay). Male B6C3F1 Big Blue® transgenic mice were fed 2500 ppm of OX or PHE or 500 ppm of WY in their diet, while a control group of mice received diet alone for 180 days. The mutant frequency (MF) of cII in the control mice, after correction for clonality, was 6.2 ± 2.8 × 10−5. The MF values for mice fed OX, PHE, and WY were 10.0 ± 3.6 × 10−5 (P < 0.05), 7.9 ± 1.3 × 10−5 (P = 0.1) and 17.4 ± 4.2 × 10−5 (P < 0.01), respectively. The mutation spectrum (MS) at cII from the PHE-fed mice was significantly different (P < 0.05) from that of the control mice even though the MF was not, whereas the MS spectra of mice fed OX (P = 0.4) and WY (P = 0.7) were not significantly different. The PHE-derived spectrum differed from the spontaneous spectrum in the lower occurrence of G:C>C:G transversions (17 vs 1.6%) and the higher incidence of A:T>T:A transversions (3.4 vs 9.5%). Prior to correction for clonal expansion, each treated group exhibited a high incidence of frameshift mutations at the homopolymeric run of guanines at bp 179–184 (OX 21%, PHE 21%, WY 16% of the total mutations); this was not the case with the control group (6%). Even after clonal correction, more than 10% of the mutations were frameshifts in the treated mice, while 5% were frameshifts in the control mice. Despite this hypersensitive region of the gene, our findings suggest that the cII locus is less sensitive than the lacI locus to mutation induction by non-DNA reactive carcinogens.

The processing of a Benzo(a)pyrene adduct into a frameshift or a base substitution mutation requires a different set of genes in Escherichia coli.

Replication through a single DNA lesion may give rise to a panel of translesion synthesis (TLS) events, which comprise error-free TLS, base substitutions and frameshift mutations. In order to determine the genetic control of the various TLS events induced by a single lesion, we have chosen the major N2-dG adduct of (+)-anti-Benzo(a)pyrene diol epoxide [(+)-anti-BPDE] adduct located within a short run of guanines as a model lesion. Within this sequence context, in addition to the major event, i.e. error-free TLS, the adduct also induces base substitutions (mostly G --> T transversions) and -1 frameshift mutations. The pathway leading to G --> T base substitution mutagenesis appears to be SOS independent, suggesting that TLS is most probably performed by the replicative Pol III holoenzyme itself. In contrast, both error-free and frameshift TLS pathways are dependent upon SOS-encoded functions that belong to the pool of inducible DNA polymerases specialized in TLS (translesional DNA polymerases), namely umuDC (Pol V) and dinB (Pol IV). It is likely that, given the diversity of conformations that can be adopted by lesion-containing replication intermediates, cells use one or several translesional DNA polymerases to achieve TLS.

Site-specific tamoxifen-DNA adduct formation: lack of correlation with mutational ability in Escherichia coli.

We have mapped sites of tamoxifen adduct formation, in the lacI gene using the polymerase STOP assay, following reaction in vitro with alpha-acetoxytamoxifen and horseradish peroxidase (HRP)/H(2)O(2) activated 4-hydroxytamoxifen. For both compounds, most adduct formation occurred on guanines. However, one adenine, within a run of guanines, generated a strong polymerase STOP site with activated 4-hydroxytamoxifen, and a weaker STOP site with alpha-acetoxytamoxifen at the same location. In Escherichia coli the lac I gene reacted with 4-hydroxytamoxifen was more likely to be mutated (2 orders of magnitude) than when reacted with alpha-acetoxytamoxifen, despite the greater DNA adduct formation by alpha-acetoxytamoxifen. This correlates with the greater predicted ability of activated 4-hydroxytamoxifen adducts to disrupt DNA structure than alpha-acetoxytamoxifen adducts. For lac I reacted with activated 4-hydroxytamoxifen, a hot spot of base mutation was located in the region of the only adenosine adduct. No mutational hot spots were observed with alpha-acetoxytamoxifen. Our data clearly shows a lack of correlation between gross adduct number, as assayed by (32)P-postlabeling and mutagenic potential. These data indicate the importance of minor adduct formation in mutagenic potential and further that conclusions regarding the mutagenicity of a chemical may not be reliably derived from the gross determination of adduct formation.

DNA frayed wires: Differential polymerization of dAnGm oligonucleotides

Oligodeoxyribonucleotides with terminal runs of contiguous guanines, d(AnGm), spontaneously associate into high molecular weight complexes that resolve on polyacrylamide gels as a regular ladder pattern of bands with low mobility. The aggregates, which we call frayed wires, arise from the interaction between the guanine residues of the oligonucleotides; the adenine tracts are single stranded and can take part in Watson–Crick interactions. Oligonucleotides, with different arm-to-stem ratios and total length, readily associate in the presence of Mg2+ to form aggregates consisting of an integer number of strands. The type of the observed aggregates is determined by the length of the guanine run. Oligonucleotides with six guanines form four- and eight-stranded complexes; there is no further polymerization. An increase in the number of guanine residues to 10 and 15 leads to polymerization resulting in a ladder pattern of up to 9 bands and an intense signal at the top of the gel. The relative population of any given species in a frayed wire sample is governed by the guanine stem length and is not affected to any substantial extent by arms up to 40 bases long. The type and concentration of the cation in the solution affect the degree of aggregation, with Na+ and K+ promoting the formation of complexes comprised of 2–4 strands and Mg2+ being the most effective in facilitating polymerization. The electrophoretic behavior of frayed wires was analyzed in the framework of the Ogston theory. The free mobility of frayed wires in the solution is close to the values reported for single-stranded DNA, indicating the equivalence of the charge density of the two conformations. The retardation coefficients for frayed wires arising from a single kind of parent strand increase with the introduction of each additional strand. There is no correlation between the retardation coefficient and the type of parent strand; rather, the magnitude of the retardation coefficient is determined by the total molecular weight of the complex. The values of the retardation coefficients are consistently higher than those for double-stranded DNA and they display much stronger dependence on the total molecular weight. Presumably, the distinct structural and dynamic characteristics of the two conformations account for their different electrophoretic behavior. © 1999 John Wiley & Sons, Inc. Biopoly 49: 287–295, 1999

DNA frayed wires: Differential polymerization of d(AnGm) oligonucleotides

Oligodeoxyribonucleotides with terminal runs of contiguous guanines, d(AnGm), spontaneously associate into high molecular weight complexes that resolve on polyacrylamide gels as a regular ladder pattern of bands with low mobility. The aggregates, which we call frayed wires, arise from the interaction between the guanine residues of the oligonucleotides; the adenine tracts are single stranded and can take part in Watson–Crick interactions. Oligonucleotides, with different arm-to-stem ratios and total length, readily associate in the presence of Mg2+ to form aggregates consisting of an integer number of strands. The type of the observed aggregates is determined by the length of the guanine run. Oligonucleotides with six guanines form four- and eight-stranded complexes; there is no further polymerization. An increase in the number of guanine residues to 10 and 15 leads to polymerization resulting in a ladder pattern of up to 9 bands and an intense signal at the top of the gel. The relative population of any given species in a frayed wire sample is governed by the guanine stem length and is not affected to any substantial extent by arms up to 40 bases long. The type and concentration of the cation in the solution affect the degree of aggregation, with Na+ and K+ promoting the formation of complexes comprised of 2–4 strands and Mg2+ being the most effective in facilitating polymerization. The electrophoretic behavior of frayed wires was analyzed in the framework of the Ogston theory. The free mobility of frayed wires in the solution is close to the values reported for single-stranded DNA, indicating the equivalence of the charge density of the two conformations. The retardation coefficients for frayed wires arising from a single kind of parent strand increase with the introduction of each additional strand. There is no correlation between the retardation coefficient and the type of parent strand; rather, the magnitude of the retardation coefficient is determined by the total molecular weight of the complex. The values of the retardation coefficients are consistently higher than those for double-stranded DNA and they display much stronger dependence on the total molecular weight. Presumably, the distinct structural and dynamic characteristics of the two conformations account for their different electrophoretic behavior. © 1999 John Wiley & Sons, Inc. Biopoly 49: 287–295, 1999

Mutagenicity of 2-amino-3-methylimidazo[4,5-f]quinoline in human lymphoblastoid cells.

2-Amino-3-methylimidazo[4,5-f]quinoline (IQ), a heterocyclic aromatic amine that has been identified in cooked meats and cigarette condensates, is mutagenic in human lymphoblastoid TK6 cells at the thymidine kinase and hypoxanthine-guanine phosphoribosyl transferase (hprt) loci. Treatment of the cells with IQ following activation with either an exogenous metabolizing mixture (S9) or following photoactivation of the azido-derivative of IQ (N3-IQ) showed that the photolytic-derivative of N3-IQ was more active. This observation is consistent with other reports that indicate that the weak mutagenicity of IQ in mammalian cells is caused by the lack of enzymes required for the ultimate activation of the compound within the cells. Two DNA adducts were found by 32P-post-labelling in the cells treated with the photoactivated N3-IQ. The major adduct was identified as N-(deoxyguanosin-8-yl)-2-amino-3-methylimidazo[4,5-f]quinoline (dG-C8-IQ) and the minor adduct as 5-(deoxyguanosin-N2-yl)-2-amino-3-methylimidazo[4,5-f]quinoline (dG-N2-IQ). The ratio of the dG-C8IQ to the dG-N2-IQ adducts was approximately 3:1 and did not significantly change in cultures treated with different concentrations of the mutagen. Approximately 50% of the adducts were removed 9 h after treatment with IQ and <10% of these adducts remained after 24 h. There was no significant preferential repair of either adduct under the experimental conditions used. The identification of 15 mutations induced at the hprt locus (of the 44 mutants analysed) showed IQ to be efficient at inducing single base deletions in a run of guanines. Six single guanine deletions were observed in the run of six guanines in exon III and one deletion of a single guanine was observed in a non-repetitive sequence in exon VI. Other mutations observed were two GC-->TA transversions, two GC-->CG transversions, one AT-->TA transversion and one GC-->AT transition. In addition, two multiple mutations were found. The majority of the identified mutations (12/15) occurred at GC base pairs and suggests either the dG-C8-IQ or the dG-N2-IQ adduct to be the pre-mutagenic lesion.

Guanine tetraplex formation by short DNA fragments containing runs of guanine and cytosine

Using CD spectroscopy, guanine tetraplex formation was studied with short DNA fragments in which cytosine residues were systematically added to runs of guanine either at the 5' or 3' ends. Potassium cations induced the G-tetraplex more easily with fragments having the guanine run at the 5' end, which is just an opposite tendency to what was reported for (G+T) oligonucleotides. However, the present (G+C) fragments simultaneously adopted other conformers that complicated the analysis. We demonstrate that repeated freezing/thawing, performed at low ionic strength, is a suitable method to exclusively stabilize the tetraplex in the (G+C) DNA fragments. In contrast to KCl, the repeated freeze/thaw cycles better stabilized the tetraplex with fragments having the guanine run on the 3' end. The tendency of guanine blocks to generate the tetraplex destabilized the d(G5).d(C5) duplex whose strands dissociated, giving rise to a stable tetraplex of (dG5) and single-stranded (dC5). In contrast to d(G3C3) and d(G5C5), repeated freezing/thawing induced the tetraplex even with the self-complementary d(C3G3) or d(C5G5); hence the latter oligonucleotides preferred the tetraplex to the apparently very stable duplex. The tetraplexes only included guanine blocks while the 5' end cytosines interfered neither with the tetraplex formation nor the tetraplex structure.

Analysis of mutations induced by 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) in human lymphoblastoid cells.

2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), a heterocyclic aromatic amine that is formed in abundance in cooked meats, has been found to be mutagenic in human lymphoblastoid TK6 cells at the thymidine kinase and hypoxanthine-guanine phosphoribosyl transferase (hgprt) loci. The mutations induced at the hgprt locus have been analysed. Of the mutations that have been identified, 60% were found in the coding sequence of the gene. Forty percent were in the introns which resulted in aberrant splicing and consequently, leading to exon losses in the mature hprt mRNA. Mutations resulting in a loss of exonIII appeared most frequently followed by losses of exonVI, exonVIII and partial loss of exonIX. All identified mutations occurred at GC base pairs, consistent with the adducts of PhIP that have been found previously and suggesting that the N-(deoxyguanosin-8-yl)-2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine, (dG-C8-PhIP) adduct may be the premutagenic lesion. Most of the mutations are GC-->TA transversions except for a cluster of single base pair deletions in a run of guanines. There appears to be strand bias in the induction of mutations with 85% of the mutations on the non-transcribed strand. Although the number of mutations analysed is limited (54 mutants), there are several sites (positions 166 and 207 of the coding sequence, and the splice acceptor site of exonIII) which are overrepresented. There is a preference for a 5' purine but not a strong bias for 3' A as has been found for other mutagens that form a premutagenic lesion at G. Triplet analysis shows that the triplets, 5'GGA3' and 5'AGG3', where the middle base is mutated are preferred.

DNA adduct-induced stabilization of slipped frameshift intermediates within repetitive sequences: implications for mutagenesis.

Chemical carcinogens such as the aromatic amide 2-acetylaminofluorene (AAF) are known to induce -1 frameshift mutation hotspots at repetitive sequences. This mutagenesis pathway was suggested to involve slipped intermediates formed during replication. To investigate the stability and structure of such intermediates we have constructed DNA duplexes containing single AAF adducts within a run of three guanine residues. The strand complementary to that bearing the AAF adducts contained either the wild-type sequence (homoduplexes) or lacked one cytosine directly opposite the run of guanines containing the AAF adduct and thus modeled the putative slipped mutagenic intermediates (SMIs). The melting temperature of AAF-modified homoduplexes or the unmodified SMI was reduced by approximately 10 degrees C relative to the unmodified homoduplex. Surprisingly, AAF adducts stabilized the SMIs as evidenced by an increase in melting temperature to a level approaching that of the unmodified homoduplex. The chemical probes hydroxylamine and bromoacetaldehyde were strongly reactive toward cytosine residues opposite the adduct in AAF-modified homoduplexes, indicating adduct-induced denaturation. In contrast, no cytosine reactivities were observed in the AAF-modified SMIs, suggesting that the two cytosines were paired with unmodified guanines. Use of diethyl pyrocarbonate to probe the guanine residues showed that all three guanines in the unmodified SMI adopted a transient single-stranded state which was delocalized along the repetitive sequence. However, when an AAF adduct was present, reduced diethyl pyrocarbonate reactivity at guanines adjacent to the adduct in AAF-modified SMIs reflected localization of the bulge to the adducted base. Our results suggest that AAF exerts a local denaturing and destabilizing effect within the homoduplex which is alleviated by the formation of a bulge. The stabilization by the AAF adduct of the SMIs may contribute to the dramatic increase in -1 frameshift mutation frequency induced by AAF adducts in repetitive sequences.

Nearest neighbor effects on carcinogen binding to guanine runs in DNA

A synthetic DNA fragment was constructed to determine the effect of 5' and 3' neighbors of guanine runs on the binding of chemical carcinogens. Determinations were made on the relative intensity of reactivity between aflatoxin B1 or benzo(a)pyrene and methylnitrosourea or 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea with various guanine positions in an endlabeled DNA fragment of known sequence. After reaction, the fragments were depurinated to produce strand breaks to allow Maxam and Gilbert sequencing for guanine positions. Relative reaction intensities were compared densitometrically. 3' neighbors exerted greater influence on carcinogen binding than did 5' neighbors, the influence extended only to the adjacent guanine and depended upon the chemical nature of the carcinogen. In addition, the presence of one carcinogen adduct in the guanine run influenced the formation of a subsequent adduct when the DNA was exposed to a second carcinogen, and this effect also depended on the nature of the second carcinogen. The results suggest that DNA adduct formation in the presence of multiple carcinogens is more complicated than an additive mechanism would suggest.

Base sequence specificity of three 2-chloroethylnitrosoureas

Chemical modifications of guanine are some of the most common results of interactions of DNA with many carcinogens and anti-cancer drugs, including nitrosoureas, nitrogen mustards, triazenes, polycyclic aromatics, and aflatoxins. The base sequence specificity for alkylation of guanines by three 2-chloroethylnitrosoureas has been determined. Guanines in the midst of a run of guanines are more susceptible than guanines in other base sequences. We have shown that certain 2-chloroethylnitrosoureas (BCNU, CCNU and methyl-CCNU) follow this same pattern. However, the quantitative degree of higher specificity for guanine with guanines as nearest neighbors depended on both the guanine position alkylated and the structure of the alkyl group attached. For example, when hydroxyethylation of runs of guanine occurred at N-7, a 6- to 11-fold increase of alkylation occurred compared to that found in the random base sequences of DNA, while hydroxyethylation at 0–6 increased 1.2 to 3.5-fold and chloroethylation at N-7 was 2- to 4-fold higher than in DNA. Guanines with thymines on both the 3' and 5' sides were much less susceptible, most notably in N-7-hydroxyethylation and N-7-chloroethylation. Since guanine-rich regions are found in regulatory regions of the genome, knowledge concerning the effect of base sequence upon the production of each of the potential DNA lesions is vital to gaining an understanding of the roles of these lesions in the anti-tumor activity of a drug.

Sequence specificity of mutations induced by benzo[a]pyrene-7,8-diol-9,10-epoxide at endogenousaprt gene in CHO cells

We have determined the spectrum of mutations induced by +/--trans-7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a] pyrene (BPDE) at the endogenous aprt locus in an hemizigous Chinese hamster ovary cell line exposed to 0.7 microM BPDE. Southern analysis of 59 independent mutants revealed no major genomic alterations, indicating that gene inactivation was the result of a point mutation. This conclusion was confirmed by the cloning and sequencing of 21 of these mutants. The predominant mutation, the G:CT----T:A transversion, comprised 62% of the spectrum, but other base pair substitutions and frameshifts were recovered. An examination of the target sequences for BPDE mutation revealed that mutations were localized within runs of G:C base pairs. However, approximately half of these G:C runs involved a particular sequence--a run of guanines flanked by adenine residues. Of seven such sites within the coding sequence of aprt, mutations were clustered within five of them. This class of sequence occurs at codon 61 of the human C-Ha-ras 1 protooncogene and may account for the selective activation of this codon by BPDE.