Chem. Res. Toxicol. 1996, 9, 277-283
277
Relative Contribution of Cytosine Deamination and Error-Prone Replication to the Induction of Propanodeoxyguanosine f Deoxyadenosine Mutations in Escherichia coli Stephen P. Fink, G. Ramachandra Reddy, and Lawrence J. Marnett* A. B. Hancock, Jr., Memorial Laboratory for Cancer Research, The Vanderbilt Cancer Center, Department of Biochemistry, Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-0146 Received April 12, 1995X
The role of cytosine deamination as a possible mechanism for PdGfA transitions induced by propanodeoxyguanosine (PdG) was investigated by site-specific mutagenesis techniques. PdG was placed at position 6256 in the (-)-strand of M13MB102 by ligating the oligodeoxynucleotide 5′-GGT(PdG)TCCG-3′ into a gapped-duplex derivative of the vector. Unmodified and PdG-modified M13MB102 genomes containing either uracil or thymine in the (+)-strand were transformed into Escherichia coli strains differing in their ability to excise uracil bases from DNA. After replication of the site specifically modified M13MB102, base-pair substitutions were detected by in situ hybridization using [32P]-labeled probes containing each of the possible mismatched bases opposite position 6256 in the (+)-strand. The ratio of PdGfA and PdGfT was unchanged in strains defective in the repair of uracil residues, which suggests that uracil is not an intermediate in the generation of PdGfA mutations. Similar results were obtained when PdG-M13MB102 was incubated for 14 days prior to transformation in an attempt to increase the extent of deamination. As a control experiment to test the sensitivity of the assay to detect deaminations opposite PdG, uracil-containing M13MB102 with a PdG‚T mismatch at position 6256 was transformed into E. coli JM105. Hybridization analysis indicated that approximately 80% of the phage plaques generated after genome replication contained T in the (+)-strand at position 6256. Thus, any deamination of cytosine to uracil would have been easily detected. Adducted and unadducted genomes were also transformed into E. coli LM114 or LM113, which carry a mutant umuD or umuC gene, respectively. Significant and comparable reductions in PdGfA and PdGfT were observed, suggesting that both mutations require the active participation of the UmuD and the UmuC proteins in the replication complex. The results of our experiments suggest that the PdGfA mutations induced by PdG are not caused by cytosine deamination, but arise coincident with PdGfT mutations during replication of the PdG-containing genomes. Also, the uracil-containing (+)-strand does not appear to be degraded, as is commonly assumed in site-specific mutagenesis experiments, and serves as a template for DNA synthesis when replication of the (-)-strand is blocked by an adduct such as PdG. There is growing interest in endogenously produced electrophiles and their role in mutagenesis as a result of the identification of a number of DNA adducts derived from cellular metabolites in humans and animals (1-3). For example, malondialdehyde, an endogenous product of lipid peroxidation and eicosanoid metabolism, is carcinogenic in rats, mutagenic in several bacterial and mammalian mutation assays, and mutagenic in an M13 forward mutation assay (4-7). The major product of the reaction of malondialdehyde with deoxynucleosides and/ or DNA at neutral pH is a 1,N2-pyrimidopurinone derivative of guanine abbreviated M1G (pyrimido[1,2-a]purin10(3H)-one)1 (8, 9). Recently, M1G has been detected at significant levels in liver DNA of rats and humans (10, 11). * Address correspondence to this author at the Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, TN 37232-0146. Telephone: 615-343-7329; FAX: 615-343-7534. X Abstract published in Advance ACS Abstracts, December 15, 1995. 1 Abbreviations: M G, pyrimido[1,2-a]purin-10(3H)-one; X-gal, 5-bro1 mo-4-chloro-3-indolyl-β-D-galactopyranoside; IPTG, isopropyl β-Dthiogalactoside; PdG, 1,N2-propano-2′-deoxyguanosine; TE, 10 mM Tris-1 mM EDTA buffer (pH 7.8).
0893-228x/96/2709-0277$12.00/0
An important goal in our laboratory is the elucidation of the mutagenic potential of M1G. Until recently, the instability of M1G to the deprotection conditions of oligonucleotide synthesis precluded site-specific mutagenesis experiments to test its mutagenic potency (12). We have used the structural analog propanodeoxyguanosine (PdG) as a model for M1G and recently reported that PdG placed at position 6256 in the (-)-strand of the recombinant bacteriophage M13MB102 leads to PdGfA and PdGfT mutations, following replication in E. coli (13, 14). Studies by Grollman and colleagues using a single-stranded shuttle vector with a site-specific PdG adduct also resulted in PdGfA and PdGfT mutations, but the ratio of PdGfA/PdGfT was much lower than the ratio our laboratory observed using double-stranded vectors (15). © 1996 American Chemical Society
278 Chem. Res. Toxicol., Vol. 9, No. 1, 1996
Fink et al.
Table 1. Bacterial Strains Used strain
genotype
supE, thi, rspL, endA, sbcB15, hsdR4, ∆(lac-proAB) [F′, traD36, proAB, lacIQZ∆M15] dut1, ung1, thi, relA1/pCJ105[F′CmR] supE, thi, ∆(lac-proAB) [F′, traD36, proAB, lacIQZ∆M15] thr-1, ara-14, leuB6, ∆(gpt-proA)62, lacY1, tsx-33, supE44, galK2, λ, rac-, hisG4, rfbD1, mgl-51, rspL31, kdgK51, xyl-5, mtl-1, argE3, thi-1 AB1157 umuC122 as AB1157 but umuC122 AB1157 umuD44 as AB1157 but umuD44 LM113 as AB1157 umuC122 but [F′, traD36, proAB, lacIQZ∆M15] LM114 as AB1157 umuD44 but [F′, traD36, proAB, lacIQZ∆M15]
JM105 CJ236 JM101 AB1157
uracil. If uracil is an intermediate in the genesis of PdGfA, the frequency of this mutation should increase because of the longer half-life of the uracil residues. Likewise, preincubating the adducted genomes for several days prior to transformation should increase the extent of deamination of the Hoogsteen base pair and increase the frequency of PdGfA. In fact, neither experiment altered the ratio of PdGfA/PdGfT, suggesting deamination is not important in the genesis of PdGfA. However, transformation of PdG-M13MB102 into strains defective in the umuD or umuC genes yielded significant and comparable reductions in PdGfA and PdGfT consistent with a common pathway of mutagenesis related to error-prone adduct bypass.
Materials and Methods
Figure 1. Deamination of cytosine opposite PdG as a mechanism for PdGfA mutations.
The increase in the PdGfT mutation frequency is consistent with the replication of blocking lesions according to the “A rule” in which deoxyadenosine residues are inserted opposite the blocking lesion (16). However, the mechanism for the PdGfA mutations is not clear. NMR studies by Stone and colleagues of the structure of an 11-mer duplex containing PdG opposite C indicate that PdG forms a Hoogsteen base pair with a protonated cytosine (17). Work by Shapiro and Klein (18) on the deamination of cytidine and cytosine in acidic buffer solutions indicates that one of the intermediates in deamination is cytosine protonated at N-3. Furthermore, studies by Shaw and colleagues of the rates of deamination of thymine-cytosine and cytosine-cytosine mispairs in M13mp2 DNA molecules indicate that the mispairs undergo a higher rate of deamination to thymine-uracil and cytosine-uracil than DNA molecules containing cytosine residues paired to guanine (19). The rates of deamination are close to the rates of deamination of cytosine in single-stranded DNA. Therefore, one possibility for the genesis of PdGfA mutations is that the Hoogsteen pairing between the PdG adduct and the protonated cytosine increases the rate of deamination to uracil. The uracil would then serve as a template for incorporation of adenine as part of the nucleotide excision repair process (Figure 1).2 We tested this hypothesis by transforming PdGM13MB102 into E. coli strains defective in the repair of 2
K. A. Johnson and L. J. Marnett, submitted for publication.
Bacterial Strains and M13MB102 Bacteriophage DNA Preparation. Escherichia coli strains JM105, CJ236, LM114, and LM113 were used for transformations (Table 1). JM105 and CJ236 were provided by Dr. R. S. Lloyd (University of Texas, Medical Branch at Galveston). LM114 and LM113, which are defective in SOS mutagenesis, were constructed through conjugation of AB1157 umuD44 or AB1157 umuC122 with E. coli JM101 (New England Biolabs, Beverly, MA) using the procedure described in Benamira et al. (7). AB1157 umuD44 and AB1157 umuC122 were a generous gift from John Battista (Louisiana State University, Baton Rouge). Single-stranded M13MB102 DNA with uracil or thymine in the (+)-strand was isolated as described by Burcham and Marnett (14). Double-stranded M13MB102 was isolated using a Qiagen plasmid kit (Chatsworth, CA). All DNA preparations were quantitated by measuring the absorbance at 260 nm in 10 mM Tris (pH 8.0), and the purity of the DNA was determined by agarose gel electrophoresis. Formation of M13MB102 Gapped-Duplex DNA and Ligation of dG- and PdG-8-mers. Construction of uracil- or thymine-containing M13MB102 gapped-duplex DNA and the ligation of the dG- or PdG-8-mer followed the same procedure as described by Burcham and Marnett (14). Briefly, doublestranded M13MB102 DNA was linearized by treatment with KspI and BssHII (Boehringer Mannheim, Indianapolis, IN) and dialyzed with a 12-fold excess of single-stranded M13MB102 DNA in decreasing concentrations of formamide. Following dialysis, the sample volumes were reduced using a SpeedVac (Savant, Farmingdale, NY) and loaded onto a 0.8% low melting point agarose gel. The gapped-duplex-containing band was excised, and the DNA was recovered by a phenol/LiCl extraction of the gel portions followed by ethanol precipitation (20). PdGand dG-containing 8-mers were phosphorylated using ATP (50 µM final concentration) and T4 polynucleotide kinase (United States Biochemical, Cleveland, OH) prior to use in ligations. For the ligations, gapped-duplex DNA was added to each of the phosphorylated dG- and PdG-containing 8-mers along with T4 DNA ligase (Boehringer Mannheim, Indianapolis, IN) and ATP to a final concentration of 1 mM. The ligation reaction proceeded overnight at 20 °C. The reaction mixtures were then heat denatured, and protein was removed by phenol and chloroform/isoamyl alcohol extractions. The ligation products were then resolved on a 0.8% low melting point agarose gel, and doubly-ligated DNA was excised from the gel and recovered
Cytosine Deamination in PdGfA Mutants
Chem. Res. Toxicol., Vol. 9, No. 1, 1996 279
as outlined above. Formation of G‚T- and PdG‚T-mismatched M13MB102 containing uracil in the (+)-strand was the same as above except during the formamide dialysis a 12-fold excess of uracil-containing single-stranded M13MB102 was used which contained a T at position 6256. SOS Induction, DNA Transformation, and Plating. An overnight culture of bacteria grown in 2× YT medium (16 g of Bacto tryptone, 10 g of Bacto yeast extract, and 5 g of NaCl, pH 7.4, per liter) was diluted 50-fold in 2× YT medium and vigorously shaken at 37 °C until the bacterial density reached an OD600 of 0.6-0.8. The cells were harvested by centrifugation. To induce the SOS response, bacteria were resuspended in an equal volume of 10 mM MgSO4 and treated with 90 J/m2 (JM105, LM114, and LM113) or 60 J/m2 (CJ236) UV radiation using a UVP, Inc., Model UVGL-25 hand-held lamp (UVP, Inc., San Gabriel, CA). UV dose was determined by irradiating cells at increasing times from 0 to 3 min and then plating dilutions of the irradiated cells on 2× YT plates. The optimal UV dose corresponded to a 1-10% survival of the cells as compared to a 0 time exposure. Following irradiation, all further steps were performed in the dark. The cells were allowed to recover during a 30 min incubation at 37 °C in 2× YT medium before transformation. Transformation was carried out by electroporation with a Gibco BRL Cell-Porator E. coli electroporation system (Gibco BRL, Grand Island, NY), rather than with CaCl2-treated cells as previously described by Burcham and Marnett (14). Three microliters of DNA sample (10 ng/µL) was added to 20 µL of SOS-induced JM105 (ung+), CJ236 (ung-), LM114 (umuD44), or LM113 (umuC122) cells. The cell/DNA mixture was placed into a chilled Gibco BRL microelectroporation cuvette, which was then placed into the Cell-Porator single safe and pulsed with 1.5 × 103 V (JM105, LM114, and LM113) or 2.5 × 103 V (CJ236). After the pulse, the cells were immediately mixed with 1 mL of cold SOC medium (2% Bacto tryptone, 0.5% Bacto yeast extract, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, and 20 mM glucose) in the electroporation cuvette, which was kept on ice until plating. Following transformation, the bacteria were plated on 2× YT X-gal (Research Organics, Cleveland, OH) plates (JM105) or 2× YT plates (CJ236, LM114, and LM113) and incubated for 18 h at 37 °C (primary plating). For experiments that required incubation of the sample DNA before transformation, uracil-containing PdG- and dG-modified M13MB102 DNA was aliquoted into sterile microcentrifuge tubes. The tubes were sealed with parafilm, wrapped in aluminum foil, and incubated in the dark at room temperature for 3, 7, or 14 days. Typically, 5 µL DNA samples were incubated for each time point. At the proper time, the DNA was transformed into CJ236 as outlined above. Identification of Mutations. Phage were eluted from the primary plates with 5 mL of sterile TE buffer (10 mM Tris-1 mM EDTA, pH 7.8) for 3 h at room temperature (13). The phage-containing supernatant was pipetted off and centrifuged (700g) to remove cells and agar. Serial dilutions of the eluate were then replated with JM105 and X-gal/IPTG (Research Organics, Cleveland, OH) to give plates that contained 15003000 plaques. The plaques on these plates were then probed for base-pair substitution mutant genotypes using an in situ hybridization technique described by Burcham and Marnett (14). Briefly, the agar plates were chilled for 1 h at 4 °C, and then the bacteriophage DNA was transferred to nitrocellulose filters (Schleicher and Schuell, Keene, NH) as described by Maniatis et al. (21). After allowing the membranes to dry, the filters were baked in vacuo for 2 h at 80 °C. The 13-mer hybridization probes used for the detection of PdGfC, PdGfA, and PdGfT mutations were [5′-32P]-end labeled with T4 polynucleotide kinase and 60 µCi of [γ-32P]ATP (DuPont-New England Nuclear, Boston, MA). After the kinase reaction was stopped, each probe was diluted to 100 µL with water and loaded onto a BioSpin-6 column (Bio-Rad, Richmond, CA) for the separation of phosphorylated probes from unincorporated [γ-32P]ATP.
Figure 2. Diagram of the PdG-adducted uracil- and thyminecontaining M13MB102 genomes used for these studies. Hybridizations were carried out in 100 mm flat-bottomed Pyrex crystallization flasks. Baked filters were placed in each flask and were prehybridized with 30 mL of hybridization media (21) for 2 h at 28 °C as determined previously (14). After prehybridization, the probes were added and the hybridizations were allowed to proceed overnight. The next day, the membranes were washed twice for 20 min at the hybridization temperature in 2× SSC buffer (0.3 M NaCl and 0.03 M sodium citrate, pH 7.0) before they were exposed to X-ray film for 1224 h at -80 °C. To confirm that the probes were binding to the proper basepair substitution mutation, mutant plaques were picked from plates, diluted in 1 mL of TE buffer, replated, and probed as above to eliminate false positives. To determine the exact sequence of the selected mutants, (+)-single-stranded DNA was sequenced using the dideoxynucleotide chain-termination method (21) or by automated sequencing using an Applied Biosystems Model 373A DNA sequencing unit (Applied Biosystems, Foster City, CA) in the DNA Sequencing Shared Resource of the Vanderbilt Cancer Center.
Results Site-specific, PdG- and dG-containing M13MB102 genomes were constructed by the gapped-duplex method (14). Briefly, RF M13MB102 was linearized by KspI/ BssHII followed by formamide dialysis with a 12-fold excess of either uracil- or thymine-containing singlestranded DNA. The resultant product isolated following agarose gel electrophoresis was a gapped-duplex in which the (+)-strand contained either uracil or thymine. The (-)-strand did not contain uracil and had an eight-base gap. An eight-base oligodeoxynucleotide containing either the PdG adduct or dG was ligated into this gap, and the fully ligated modified or unmodified M13MB102 containing uracil or thymine in the (+)-strand was purified by electrophoresis in ethidium bromide-containing agarose gels (Figure 2). The isolated PdG-adducted M13MB102 DNA containing uracil or thymine in the (+)-strand was then electroporated into either JM105, CJ236, LM114, or LM113. The infected cells were plated giving a thick lawn of plaques. This was considered the primary plating. Each plate was eluted with sterile TE buffer to produce a phage stock for mutant quantitation. An aliquot of the stock was replated to yield roughly 2000 plaques per plate. Plaque DNA from the secondary plates was lifted with nitrocellulose membranes and probed by differential hybridization using radiolabeled probes specific for each type of base substitution. Parallel experiments were performed with dG-M13MB102 DNA as a control. In previous experiments, we transformed bacteria using a CaCl2 method (14); transformation efficiency was approximately 800 plaques/ng of DNA.3 For the experiments described in this paper, transformation by electroporation was performed. Electroporation was chosen because it yields reproducible results, and for many cell types higher transformation efficiencies. For these experiments, there was a significant increase in the trans-
280 Chem. Res. Toxicol., Vol. 9, No. 1, 1996
Fink et al.
Table 2. Base-Pair Substitutions Detected in Progeny of dG- and PdG-Modified M13MB102 Replicated in SOS-Induced JM105 or CJ236 Cellsa PdGfC
PdGfA
PdGfT
Table 3. Effect of Preincubation Time on PdGfA and PdGfT Mutationsa incubation time (days)
PdGfA/PdGfT
Uracil-Containing (+)-Strand JM105 dG
