568
Chem. Res. Toxicol. 1997, 10, 568-574
Frameshift Mutagenesis Induced in Escherichia coli after in Vitro Treatment of Double-Stranded DNA with Methylene Blue plus White Light: Evidence for the Involvement of Lesion(s) Other than 8-Oxo-7,8-dihydro-2′-deoxyguanosine Je´roˆme Wagner and Robert P. P. Fuchs* Cance´ rogene` se et Mutagene` se Mole´ culaire et Structurale, Unite´ Propre de Recherche (#9003) du Centre National de la Recherche Scientifique, IRCAD, Hopitaux Universitaires, BP 426, 67091 Strasbourg Cedex, France Received October 2, 1996X
By means of specific mutation assays, we show here that in vitro treatment of double-stranded plasmid DNA with methylene blue and white light efficiently promotes frameshift mutagenesis in Escherichia coli. The assays detect either -1 or -2 frameshift mutations within previously characterized hot spot sequences for frameshift mutagenesis induced by the chemical carcinogen N-2-acetylaminofluorene, namely, short runs of contiguous guanines and alternating GpC sequences, respectively. The SOS and umuDC dependences of these mutagenic processes have been investigated. Both -1 and -2 frameshift mutagenesis are increased when the host SOS functions are induced. However, and although functional UmuDC proteins are required for maximal mutation induction, the inducibility of both -1 and -2 frameshift mutagenesis is partially independent upon the integrity of the umuDC operon. In addition, results obtained using plasmids with a site specifically located 7,8-dihydro-8-oxo-2′-deoxyguanosine (8-oxo-dGuo) residue show that this lesion, the major methylene blue plus light induced lesion characterized so far, is inefficient in promoting frameshift mutagenesis. Together, these results led us to conclude that methylene blue plus light treatment of DNA induces, at relatively high rates, lesion(s) other than 8-oxo-dGuo, that efficiently promote(s) frameshift mutagenesis in E. coli.
Introduction Reactive oxygen species, generated during normal aerobic metabolism and oxidative stress, exert deleterious effects to the cell and are involved in different human degenerative diseases including cancer (see ref 1 for review). One of the threats thus imposed to the cell is the generation of various DNA damage and the resulting mutations. The analysis of the mutational spectra derived from several forward mutation assays reveals that oxidative treatment of DNA induces a large excess of base substitutions over the other class of point mutations, i.e., frameshift mutations (2-9). However, frameshift mutations were shown to occur at non negligible levels in two studies using singlet oxygen (1O2) as the oxidative agent (2, 3). White light irradiation of DNA solutions in the the presence of methylene blue (MB+light)1 specifically damages DNA at guanine sites (10, 11), at least partly via the generation of singlet oxygen in the solution (type II mechanism; 12-14). When free deoxyguanosine or DNA is submitted to * Corresponding author. Telephone: 33 3 88 11 90 03. Fax: 33 3 88 11 90 98. E-mail:
[email protected]. X Abstract published in Advance ACS Abstracts, April 15, 1997. 1 Abbreviations: MB+light, methylene blue plus white light; 8-oxodGuo, 7,8-dihydro-8-oxo-2′-deoxyguanosine; AAF, N-2-acetylaminofluorene; FapyGua, 2,6-diamino-4-hydroxy-5-formamidopyrimidine; cyanuric acid, 1-(2-deoxy-β-D-erythro-pentafuranosyl)cyanuric acid; 4-OH-8-oxo-dGuo, 4R* and 4S* diastereomers of 4-hydroxy-8-oxo-4,8dihydro-2′-deoxyguanosine; dZ, 2,2-diamino-4-[(2-deoxy-β-D-erythropentafuranosyl)amino]-5-(2H)-oxazolone; dIz, 2-amino-5-[(2-deoxy-βD-erythro-pentafuranosyl)amino]-4H-imidazol-4-one; AP sites, abasic sites; ssb, single-strand break(s); ApR, ampicillin resistance; KnR, kanamycin resistance; TcR, tetracycline resistance; Tcs, tetracycline sensitivity.
S0893-228x(96)00169-5 CCC: $14.00
MB+light treatment, nine different DNA alterations have been identified (10, 13, 15-18). In double-stranded DNA, 7,8-dihydro-8-oxo-2′-deoxyguanosine (8-oxo-dGuo) is the major MB+light-induced lesion identified so far (14, 18, 19), but the formation of other potent mutagenic and/or lethal lesion(s) is suspected. In the present study, we investigate the genetic requirements and efficiency of frameshift mutagenesis induced by MB+light within specific target DNA sequences, namely, monotonous run of guanines and alternating GpC sequences. These sequences were previously identified as frameshift mutation hot spots for the chemical carcinogen N-2-acetylaminofluorene (AAF; 20, 21), that forms covalent adducts at the C8 position of guanine (22) and strongly blocks DNA replication (23, 24). Frameshift mutations occurring within these two sequence contexts are currently modeled as “incorporation-slippage” events that imply the elongation, by the Escherichia coli replication machinery, of so-called “Slipped Mutagenic Intermediates” (25-27). Here, we found that MB+light-induced lesion(s) efficiently promote(s) both -1 and -2 frameshift mutagenesis in a SOS-dependent manner. In addition, data obtained with site specifically monomodified 8-oxo-dGuo vectors demonstrate that this lesion is inefficient in promoting frameshift mutagenesis, implying thus one or several other oxidative lesions as potent frameshift mutagens.
Experimental Procedures Chemicals, Bacterial Strains, and Plasmids. Methylene blue was purchased from Sigma Chemical Co. E. coli strains used in this study are: AB1157 (wild type, 28); GY8347, a
© 1997 American Chemical Society
Frameshift Mutagenesis by MB+Light generous gift from Dr. R. Devoret (Orsay, France), as AB1157 but ∆umuDC (29); BH1040 (supE endA sbcB15 hsdR4 rpsL thi∆lac-proAB mutY::KnR fpg-1::KnR, X::TcR)/F′ (traD36 proAB+ lacIq lacZ∆M15), a generous gift from Dr. S. Boiteux (Saclay, France), is deficient in the repair of 8-oxo-dGuo; BH400, as AB1157 but uvrA6, fpg-1::KnR. Plasmids used in the mutagenesis assays are derivatives of pBR322: pTcs6(G), used to detect -1 frameshift mutagenesis, contains an addition of one guanine residue at positions 536-540 in the tet gene of pBR322 (ApR, Tcs, 21); pTcs2(GC), which contains an additional GpC at the HaeIII site (596-599) in the tet gene of pBR322 (ApR, Tcs; 30) and is used to detect -2 frameshift mutagenesis. pTcs3(GC) contains an additional GpC at the NarI site (434-439) in the tet gene of pBR322 (ApR, TcS; 31) is also used to detect -2 frameshift mutagenesis. In Vitro Treatment of DNA. (A) Methylene Blue + White Light Treatment. Plasmid DNA modification with methylene blue plus visible light was carried out essentially as described previously (19). Briefly, 10 µg of plasmid DNA [CsCl grade, 100 ng/µL in 10 mM Tris (pH 8.0), 1 mM EDTA, and 10 mM MgCl2] in a 96-well microtiter plate placed on ice is exposed to white light after addition of 10 µL of a fresh methylene blue dilution (0, 20, 200, and 2000 µM). Light, filtered by a 2 mm thick water layer, is provided for 15 min by a 100 W bulb positioned at a distance of 11 cm above the sample. After irradiation, DNA was ethanol-precipitated 3-4 times and finally resuspended in TE buffer [10 mM Tris (pH 8.0), 1 mM EDTA]. Control DNAs were treated as described above, but light was omitted. The number of single-strand breaks (ssb) induced by the MB+light treatment has been estimated by determining the relative proportions of plasmid DNA forms I and II and assuming a Poisson distribution. Thus, ssb per plasmid molecule (ssb/pl) were calculated according to the equation ssb/pl ) -ln(proportion of form I remaining after treatment). (B) Induction of AP Site. AP sites were introduced by incubating CsCl grade plasmid DNA (10 µg) at pH 4 (10 mM sodium citrate, 100 mM KCl, pH 4) at 70 °C for increasing time periods (0, 1.5, 3, 5, and 7 min). After incubation, the DNA (500 ng) was treated with exonuclease III (New England BioLabs, 8 units) for 40 min at 37 °C in reaction buffer (50 mM Tris-HCl, pH 8, 0.05 µM DTT, 0.1 mg/mL BSA, and 5 mM CaCl2). The number of AP sites per plasmid molecule was assessed using the procedure described above for the determination of the number of ssb per plasmid. Measurement of MB+Light-Induced 8-Oxo-dGuo. The amount of 8-oxo-dGuo and the amount of dGuo present in the MB+light-modified DNA were determined essentially as described elsewhere (14). Briefly, 20 µg of modified plasmid DNA was resuspended in 500 µL of digestion buffer [40 mM TrisHCl (pH 8.5), 10 mM MgCl2] and submitted to digestion for 2 h at 37 °C using an enzymatic cocktail (all purchased from Boehringer Mannheim) containing 4 units of DNase I, 4 × 10-3 unit of spleen phosphodiesterase, 9 × 10-3 unit of snake venom phosphodiesterase (Crotalus durissus), and 0.5 unit of alkaline phosphatase from calf intestine. The equivalent of 2 µg of digested DNA was chromatographed on a Beckman Ultrasphere C18 5 µm, analytical column (250 mm × 4.6 mm i.d.) using a Waters Model 590 pump. The mobile phase was 0.2% (v/v) triethylamine acetate (pH 6), 5% (v/v) methanol. The flow rate for analysis was 1 mL/min. Detection was performed through two detectors set in series: an LKB Model 2151 UV detector (Pharmacia LKB Biotechnology, Uppsala, Sweden) set at 260 nm followed by a Model LC-2A amperometric detector (Bioanatycal Systems, West Lafayette, IN) at an electrode potential setting of +0.68 V. Pure 8-oxo-dGuo, a kind gift from J. Cadet, and dGuo (Sigma Chemical Co.) were used to established standard curves and to unequivocally ascribe the peaks observed during HPLC analysis of our samples. In Vivo Reversion Assay. The in vitro/in vivo mutation assay (32) involves the in vitro treatment of the plasmid containing the mutation target with methylene blue plus white light (MB+light) followed by its introduction into bacteria. The recipient bacteria are either SOS-induced or noninduced. SOS induction is achieved by UV irradiation (at 30 J/m2 for the wild type, ∆umuDC, and mutY,fpg strains and at 5 J/m2 for the uvrA,
Chem. Res. Toxicol., Vol. 10, No. 5, 1997 569 fpg strain) of a log-phase bacterial culture followed by incubation for 30 min at 37 °C to allow expression of the SOS functions (20). The bacteria are then transformed by electroporation (33) with the MB+light-modified plasmids. After an incubation time of 30 min, appropriate dilutions of the bacterial suspensions were plated on ampicillin-containing plates to measure the surviving transformant fraction. The revertant frequency was calculated as the ratio of revertants (TcR, recovered on tetracycline-containing plates) to the total number of surviving bacteria (ApR, numbered on ampicillin-containing plates) after a 3 h expression period which allows the optimal recovery of TcR revertants (32). 8-Oxo-dGuo Site Specifically Monomodified Vectors and Mutagenesis Tests. pUC-derived plasmids containing a single 8-oxo-dGuo lesion and the corresponding control plasmids were constructed and purified as described elsewhere (34, 35) using the gapped duplex methodology. Oligonucleotides containing a single 8-oxo-dGuo residue were chemically synthesized (36). The lesion was located either at the third guanine of a NarI restriction site [construct pUC-Nar(Ox3), 5′-ATCACCGGCGoCCACA-3′, Go ) 8-oxodGuo] or at the third guanine of a SmaI site [construct pUC-3G(Ox3), 5′-ATACCCGGGoACATC3′]. Control plasmids were constructed using the corresponding unmodified oligonucleotides. The specific mutation assays were carried out as described previously (25, 37): the monomodified plasmids are introduced in competent BH1040 cells by electroporation (33) and plated on LB agar containing ampicillin (100 mg/mL), 5-bromo-4-chloro-3-indolyl β-D-pyranoside (X-Gal, 50 mg/L), and isopropyl β-D-thiogalactopyranoside (IPTG, 60 mg/ L). -1 or -2 mutations occurring within the polylinker region of the plasmids pUC-3G(Ox3) or pUC-Nar(Ox3), respectively, restore the reading frame of the plasmid-encoded 5′-terminal fragment of the lacZ gene. These mutants express a functional β-galactosidase, and, in an R-complementation host, yield blue transformants on X-Gal/IPTG medium.
Results Survival of MB+Light-Treated DNA in E. coli. Survival of MB+light-treated pTcs2(GC) and pTcs(6G) plasmids was determined in SOS-induced (SOS+) and noninduced (SOS-) wild-type and ∆umuDC-competent Escherichia coli cells (Figure 1a). Transformation of wild-type cells with DNA treated with 2, 20, and 200 µM MB+light results in the loss of about 20, 95, and 98% of viability, respectively. The lesion(s) responsible for the observed toxicity is (are) generated only when the DNA solutions are exposed to both MB and light. When light was omitted (MB-light), no decrease in survival was observed (data not shown), thus excluding the possibility that toxicity is mediated by noncovalent intercalation of MB that would remain bound to the DNA despite the extensive purification steps (see Experimental Procedures). Induction of the SOS functions in the wild type strain results in a 2.5-4-fold increase in viability at MB concentrations of 20 and 200 µM, respectively. When ∆umuDC mutant cells are used, the survival of the double-stranded DNA is similar to the survival observed in noninduced wild-type cells and is not increased when the SOS response is induced. According to the SOS hypothesis (38), these results suggest that the slight gain in viability observed after induction of the SOS functions needs a functional umuDC operon whose products are known to facilitate the bypass of blocking lesions (39). In a strain defective in both nucleotide excision repair (uvrA) and the 8-oxo-dGuo glycosylase (fpg), a severe reduction in survival (≈10-fold decrease as compared to the wild-type strain) is observed at a dose of 10 µM MB (Figure 1b). An SOS-induced reactivation factor similar to that measured in the wild-type strain is also observed (Figure 1b). These results suggest that both Fpg and
570 Chem. Res. Toxicol., Vol. 10, No. 5, 1997
Figure 1. Survival of methylene blue plus light (MB+light)treated DNA in repair-proficient (a) and repair-deficient (b) Escherichia coli strains. pTcs2(GC) and pTcs6(G) plasmids were illuminated for 15 min in the presence of increasing amounts of MB and transfected into SOS-noninduced (open symbols) or SOS-induced (closed symbols) competent host cells. Each point represents the mean value of at least two independent experiments. Triangles, AB1157 (wild type); circles, GY8347 (∆umuDC); squares, BH400 (fpg, uvr).
UvrABC proteins are involved in the removal of toxic and mutagenic lesions, as previously published by Tudek et al. (9). It should be noted that in fpg or uvrA singlemutant strains the survival was similar to that observed in a wild-type strain (data not shown). The putative nature of the DNA lesion(s) responsible for the lethal effect observed after MB+light treatment will be discussed later in this paper (see Discussion). Spontaneous Mutagenesis. Upon induction of the SOS response, the reversion frequency in the absence of MB+light treatment is increased about 10-fold for both pTcs6(G) and pTcs2(GC) plasmids (Figures 2a and 4a). The induction of the SOS response by UV irradiation may enhance the spontaneous mutation frequency by promoting a more efficient elongation of the slipped mutagenic intermediates (25, 40). MB+Light-Induced Mutagenesis within the Contiguous G Sequence. When plasmid pTcs6(G) is introduced into the wild-type strain AB1157, induction of -1 frameshift mutations increases markedly between 0 and
Wagner and Fuchs
Figure 2. MB+light-induced -1 frameshift mutagenesis within a run of six contiguous guanines. (a) pTcs6(G) plasmids were illuminated for 15 min in the presence of increasing amounts of MB and transfected into SOS-noninduced (open symbols) or SOS-induced (closed symbols) AB1157 (wild type) competent host cells. (b) pTcs6(G) plasmids modified with 20 µM MB+light were introduced into SOS-induced or SOS-noninduced AB1157 (wild type) and GY8347 (∆umuDC) competent Escherichia coli cells. Data presented in (b) are the results of at least two independent transformation experiments; standard deviations are shown.
20 µM MB+light and reaches a plateau at 200 µM MB+light (Figure 2a). In SOS noninduced cells (SOS), the 20 µM MB+light treatment increases the frequency of reversion events up to 1000-fold over the spontaneous mutation frequency (0 µM MB+light). Induction of the SOS functions (SOS+) by UV irradiation of the host prior to transformation leads to an additional 3-fold increase in the mutation frequency. The maximal mutation frequency observed here (about 6 × 10-4) reaches a level similar to that previously obtained under the same conditions with plasmid pTcs,6(G) containing an average of about 10 dGuo-C8-AAF adducts (32). We have analyzed the nature of 18 tetracycline-resistant (TcR) revertants by DNA sequencing. The results indicate that ≈ 70% (12/18) were true revertants (loss of a G‚C base pair within the run of 6 G‚C pairs; 21) whereas about 30% (6/18) arise from a -1 frameshift mutation involving
Frameshift Mutagenesis by MB+Light
Chem. Res. Toxicol., Vol. 10, No. 5, 1997 571
Figure 3. MB+light-induced -1 frameshift mutagenesis within a run of six contiguous guanines. pTcs6(G) plasmids were illuminated for 15 min in the presence of increasing amounts of MB and transfected into SOS-induced wild-type AB1157 (triangles) or repair-deficient (uvrA, fpg) BH400 (squares) competent host cells. Data presented are the mean values of at least two independent experiments.
a cytosine residue lying immediately 5′ to the run of six G (5′-CCGGGGGG-3′). Similar results have been previously obtained using the same plasmid randomly modified with AAF adducts (32). We compared -1 mutation frequencies induced by MB+light treatment in wild-type, ∆umuDC, and fpg, uvr cells that were either SOS-induced or noninduced (Figures 2b and 3). The results show that even in the absence of SOS induction and of a functional umuDC locus (Figure 2b), Escherichia coli is able to efficiently convert MB+light-induced lesions into -1 frameshift mutations. Induction of the SOS response further increases -1 frameshift mutagenesis in both wild-type and ∆umuDC strains. Although maximal mutation frequency is obtained in SOS-induced wild-type cells, these results suggest that one (or more) UV-inducible gene(s) product(s), other than umuDC, is (are) involved in -1 frameshift mutagenesis. In SOS-induced cells, a higher mutation frequency was seen at any given level of modification in a uvrA, fpg mutant strain as compared to the corresponding wild-type strain (Figure 3). In fact, in these two strains equal mutation frequencies are found at doses giving rise to equal survival. These results suggest the participation of the Fpg and/or the UvrABC proteins in the removal of the MB+light-induced lesion involved in the induction of -1 frameshift mutagenesis. MB+Light-Induced Mutagenesis at Alternating GC Sequences. When plasmid pTcs2(GC) is introduced into the wild-type strain AB1157, mutagenesis rates increase markedly between 0 and 20 µM MB+light, and, as for the -1 frameshift mutagenesis, a plateau was observed when DNA was treated with 200 µM MB+light (Figure 4a). In SOS-noninduced cells, the 20 µM MB+light treatment increases the frequency of reversion events up to 1000-fold [1 × 10-6/(
