THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No Issue of January 15, pp. 1269-1275,1991 Printed in U.S.A.

The Role of N3-Ethyldeoxythymidine in Mutagenesis and Cytotoxicity by Ethylating Agents*

(Received for publication, April 26, 1990)

Peter C. GrevattS, Jean M. Donahue§, and Opinder S. Bhanotn From the Department of Environmental Medicine, New York University Medical Center, New York, New York 10016

The significance of DNA ethylation at the central hydrogen-bonding site (N3) of thymine was investi- gated using an in vitro DNA replication system. The system utilized a primed template in which the 3’-end of the primer is eight nucleotides away from N3- ethyldeoxythymidine (N3-Et-dT), present at template position 26 from the 3’-end. The 34-nucleotide tem- plate corresponds to a specific DNA sequence at gene G of bacteriophage 4x174. DNA synthesis products were quantitated by electrophoretic separation and autoradiography. At 10 PM dNTP and 0.5 mM Mn2+, N3-Et-dT blocked DNA synthesis by Escherichia coli polymerase I (Klenow fragment): 60% after incorpo- rating a nucleotide opposite N3-Et-dT (incorporation- dependent blocked product) and 39% 3‘ to N3-Et-dT. DNA replication past the lesion (post-lesion synthesis) was negligible. Post-lesion synthesis increased using higher concentrations of dNTP, reaching 68% at 200 PM dNTP. DNA sequencing revealed that dA was in- corporated opposite N3-Et-dT in the incorporation- dependent blocked product. In the post-lesion synthesis product, dT was exclusively incorporated opposite N3- Et-dT. Formation of the N3-Et-dT*dA base pair at the replication fork terminated DNA synthesis, while the N3-Et-dT-dT base pair formed at the 3‘-end of the growing chain was extended, leading to an A-T -., T.A transversion mutation. The results suggest a dual role for the N3-Et-dT lesion, contributing in part to the cytotoxicity and mutagenicity of ethylating agents. These studies provide a basis for understanding the activation of oncogene neu by A* T + T-A transversion mutation in rat neuroblastomas induced by N-ethyl-N- nitrosourea.

Most alkylating agents are potent mutagens and carcino- gens. Mutagenesis by these agents is a concerted biological response to covalent modification of DNA involving DNA replication and repair (Singer and Grunberger, 1983; Saffhill et al., 1985). All oxygen and most nitrogen atoms in DNA can be alkylated, but the detailed biochemistry of mutagenesis is

* This work was supported in part by National Science Foundation Grant DMB-8607556, National Institute of Environmental Health Sciences Grant ES 00260, and a Special Institutional Grant SIG 9A from the American Cancer Society. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Supported by National Institute of Environmental Health Sci- ences Training Grant 2T 32 ES 07081 and recipient of a Helen Q. Woodard Graduate Scholar Award.

I Supported by National Institute of Environmental Health Sci- ences Training Grant 2T 32 ES 07081 and a recipient of a Shell Graduate Scholar Award.

T To whom correspondence should be addressed.

known for only DNA alkylation products a t the 0-6 position of guanine (06-alkyl-dG’) and the 0 - 4 position of thymine (04-alkyl-dT) (Singer, 1986; Basu and Essigmann, 1988).

There appear to be two main pathways of alkylation-in- duced mutagenesis, direct base mispairing during DNA rep- lication and misrepair (Walker, 1984). The major pre-muta- genic lesion, O‘j-alkyl-dG, produces G. C + A - T transition mutations by mispairing with d T (Loechler et al., 1984; Bhanot and Ray, 1986), while 04-alkyl-dT induces A .T + G. C transitions by mispairing with dG (Preston et al., 1986). 0’- Alkyl-deoxythymidine (02-alkyl-dT) may be a potential mis- pairing lesion (Singer et al., 1983; Saffhill et al., 1985) but this lesion also inhibits i n uitro DNA replication on both synthetic (Saffhill et al., 1985) and natural DNA substrates (Grevatt and Bhanot, 1990). The extent and precise nature of the contribution of O*-alkyl-dT to alkylation mutagenesis is not fully established. The transition mutations constitute a major portion of the mutations induced by treatment of wild-type Escherichia coli with most simple alkylating agents (Miller, 1983).

Misrepair mutagenesis involves the error prone processing of noncoding lesions. The noncoding lesions may be adducts that modify normal base pairing positions, such as l-alkyl- purines and 3-alkylpyrimidine, or lesions that may sterically interfere with base pairing, such as 3-alkyladenine and @- alkyl-deoxythymidine adducts or apurinic/apyrimidinic sites formed through spontaneous loss (Lawley and Brooks, 1963; Drinkwater et al., 1980) or removal by glycosylases (Lindahl and Sedgwick, 1988) of alkylated bases. All noncoding lesions impede the progress of in vitro DNA replication of synthetic DNA substrates (Strauss, 1985) and may require the “error prone” functions of the SOS response to produce mutations. The SOS response (Witkin, 1976; Little and Mount, 1982; Walker, 1984) includes both accurate and mutagenic DNA repair activities. Mutagenic processing of DNA lesions is associated with the umuDC operon. Defects in this operon eliminate mutagenesis by many agents without affecting the other functions of the SOS response. N-Methyl-N‘-nitro-N- nitrosoguanidine induced-urn&+-dependent mutagenesis was significantly enhanced in alkA bacteria (Foster and Eisen- stadt, 1985), indicating that the premutagenic lesions may be substrates for alkA gene product. Transversion mutations a t A .T base pairs dominated the umu-dependent mutagenesis. Among the substrates for the alkA gene product (Evensen and Seeberg, 1982; McCarthy et al., 1984), 3-methyladenine was argued to be the premutagenic lesion responsible for alkA- umuC+-dependent A. T transversion mutations.

Mutagenesis by N-ethyl-N-nitrosourea (ENU) revealed that both SOS-independent and SOS-dependent pathways

The abbreviations used are: N3-Et-dT, N3-ethyldeoxythymidine; KfPolI, Klenow fragment of E. coli polymerase I; ENU, N-ethyl-N- nitrosourea.

1269

1270 Significance of DNA Damage at Thymine N3

process premutagenic ENU-induced lesions to yield G - C + A.T transition mutations, but the SOS processing increases the efficiency of fixation of this type of mutation (Eckert et al., 1989). In addition, SOS processing significantly increased the frequency of A . T + C.G transversion mutations. Fre- quent induction of transversion mutations at A . T base pairs is also observed with N-methyl-N-nitrosourea in SOS-in- duced E. coli (Cuoto et al., 1989). Alkylated thymidine lesions, which normally block DNA replication, have been implicated in A . T transversion mutagenesis in SOS-induced E. coli (Eckert et al., 1989), however, the specific lesions contributing to this mutagenesis have not been fully identified.

Alkylation at the N3 position of dT interferes with its base- pairing ability. The presence of N3-methyldeoxythymidine (N3-Me-dT) in synthetic polymers inhibited DNA synthesis by DNA polymerases (Saffhill, 1985; Huff and Topal, 1987) with no nucleotide incorporation opposite the lesion above the background for polymerase infidelity in uitro (Loeb and Kunkel, 1982). Using the natural metal ion Mg2+, the N3-Et- dT present at a single site in the bacteriophage (PX174 DNA sequence blocked DNA synthesis by the Klenow fragment of E. coli polymerase I (KfPolI) (Grevatt et al., 1990). The synthesis was predominantly blocked immediately 3’ to the N3-Et-dT adduct and to a lesser extent at the adduct after incorporating dA opposite it (Bhanot et al., 1990).

In this paper we report that in the presence of mutagenic metal ion Mn”, KfPolI incorporated dA and dT opposite the noncoding N3-Et-dT lesion present in a template correspond- ing to the PX174 DNA sequence. After incorporation of dA, the formation of the N3-Et-dT. dA base pair at the replication fork was resistant to further elongation causing the DNA synthesis to terminate. The N3-Et-dT.dT base pair at the 3‘-end of the growing chain was able to extend leading to an A. T + T. A transversion mutation. The results suggest a dual role for the noncoding N3-Et-dT lesion. Since N3-Et-dT inhibits DNA synthesis, it may play a role in determining some of the cytotoxicity of the ethylating agents. The lesion may also contribute to the A. T + T . A transversion muta- tions induced by ethylating agents. These findings provide a basis for understanding the mechanisms of activation of on- cogene neu by an A-T -+ T . A transversion mutation in rat neuroblastomas induced by ENU (Bargmann et al., 1986; Perantoni et al., 1987).

MATERIALS AND METHODS

Ultra-pure grade dNTP and adenosine 5”triphosphate (ATP) were purchased from Pharmacia P-L Biochemicals. [-y-32P]ATP was ob- tained from Du Pont-New England Nuclear. E. coli polymerase I (Klenow fragment) and Tq DNA ligase were purchased from Boehrin- ger Mannheim. One unit of KfF’olI (as reported by the supplier) incorporates 10 nmol of total nucleotides into acid-precipitable ma- terial in 30 min a t 37 ‘C using poly[d(A.T)] as the primed template. T, polynucleotide kinase was obtained from New England Biolab. Ultra-pure electrophoresis reagents were from Bio-Rad, while all chemicals used in the Maxam-Gilbert sequencing reactions were from Aldrich. All other chemicals were of high grade quality and purchased from different sources.

AAGTTT*AAAACATG (T* = N3-Et-dT), containing the N3-Et-dT Site-modified Oligonucleotide-A 17-nucleotide oligomer, pTAA-

adduct a t a single preselected site, was synthesized on a solid support using phosphite chemistry (Caruthers, 1985). The N3-Et-dT adduct was introduced at the desired site during synthesis by the use of N3- ethyl-5’-0-(4,4‘-dimethoxytrityl)-2’-deoxythymidine-3’-[(2-cyano- ethyl)-N,N-diisopropylphosphoramidite] (Bhanot et al., 1990). The N3-Et-dT adduct is stable in the conditions encountered during oligonucleotide synthesis.

The deprotected oligonucleotide was purified by thin layer chro- matography on silica gel followed by electrophoresis on 20% poly- acrylamide, 8 M urea gel. The purified oligomer was phosphorylated at the 5’-end and fully characterized for its purity, expected DNA

sequences, and presence of N3-Et-dT in over 99.5% of the oligonu- cleotide molecules. The details of the synthesis of the site-modified oligomer and its characterization have been described (Bhanot et al., 1990).

Formation of the Site-modified Primed Template-The primed template, with the 3‘ terminus of the primer eight nucleotides away from the N3-Et-dT lesion in the template, was prepared by annealing about a 2-fold molar excess of complementary 5’-32P-labeled (3,000 Ci/mmol) 17-nucleotide primer to the 34-nucleotide site-modified template. The experimental details, including construction of the 34- nucleotide template containing a single N3-Et-dT adduct at position 26 from the 3’-end, and formation and separation of the primed template, have been described (Bhanot et al., 1990).

The formation of the primed template as checked by electropho- resis on a non-denaturing 12% polyacrylamide gel was >80%. A single batch of primed template was used in all DNA replication studies with N3-Et-dT.

A 34-nucleotide template, containing dT in place of N3-Et-dT, was constructed for use as a control in the DNA replication studies.

DNA Polymerase Reaction-The polymerase reactions using Kf- Pol1 were conducted on 0.05-0.1 pmol of the primed template essen- tially as described (Bhanot et al., 1990). The concentration of dNTPs and Mn2’ ions, and incubation time and temperature were varied depending upon the type of study conducted. The exact conditions used are documented in the figure legends. The DNA synthesis products present in the polymerization reaction were analyzed as described (Bhanot et al., 1990).

Various DNA synthesis products in amounts sufficient for DNA sequencing were prepared using 5-10 pmol of the primed template in the presence of the desired dNTP concentration and 0.5 mM Mn”. The products were isolated as described (Bhanot et al., 1990). DNA sequencing was performed by the modified Maxam-Gilbert procedure (Maxam and Gilbert, 1980).

RESULTS

Template Construction Containing the N3-Et-dT Lesion at a Single Preselected Site-A 34-nucleotide long template, containing a single N3-Et-dT adduct at template position 26 from the 3‘-end, was constructed by ligating a 17-nucleotide- long site-modified primer to a second 17-mer in the presence of an 18-nucleotide long complementary oligonucleotide to hold the primers together. The DNA sequence of the resulting 34-mer template corresponded to bacteriophage (PX174 (-) strand sequences from position 2377 to 2410 and contained the N3-Et-dT adduct at position 2402. This position corre- sponds to the second nucleotide in the third codon of (PX gene G. To facilitate a comparison with the in vitro DNA replica- tion studies, these sequences will be used to study the in vivo mutagenic potential, specificity, and frequency of the N3-Et- dT lesion, using a qX174-based site-specific mutagenesis sys- tem (Bhanot and Ray, 1986) in a separate investigation.

I n Vitro DNA Replication System-The i n vitro DNA rep- lication system assays the specificity and frequency of DNA polymerase-induced incorporation of a nucleotide opposite the N3-Et-dT lesion. The in vitro replication system utilizes the primed-template complex shown below. The primed-tem- plate consists of a complementary 32P-labeled (3,000 Ci/ mmol) 17-nucleotide primer hybridized to the site-modified 34-nucleotide template. In this primed-template system, the 3’ terminus of the primer is eight nucleotides away from the N3-Et-dT lesion. This system represents a “running start” for N3-Et-dT in DNA replication (Mendelman et al., 1989) in that synthesis occurs prior to the polymerase reaching the N3-Et-dT lesion.

17-mer primer 5’-:’2P-TAAT

34-mer template 3’-ATTAGTACAAAAT*TTGAAAAT5’ I I I

17 26 34

T * = N3-Et-dT

Significance of DNA Damage at Thymine N3 1271

In this DNA replication system, the primer is extended by the DNA polymerase until the N3-Et-dT lesion is encoun- tered. All of the following possibilities are feasible. First, the progress of the polymerase is blocked one nucleotide 3’ to the encountered lesion. No nucleotide is incorporated opposite the lesion and a 25-nucleotide product (hereafter referred to as “preincorporation blocked product”) is accumulated. Sec- ond, DNA synthesis is stopped at the N3-Et-dT adduct after incorporating a nucleotide opposite the adduct. A 26-nucleo- tide product (hereafter referred to as “incorporation-depend- ent blocked product”) is produced. Finally, the adduct does not represent a block to DNA replication and the synthesis proceeds to the 5’ terminus of the single-stranded template yielding a 34-nucleotide product (hereafter referred to as “post-lesion synthesis product”).

The products of DNA synthesis on the site-modified tem- plate were analyzed by separation on 20% polyacrylamide, 8 M urea sequencing gels (Bhanot et al., 1990). Since the ‘*P- end-labeled primer was used to prime the DNA synthesis, each synthesis product is only labeled once at the 5‘-end. Therefore, the abundance of a particular synthesis product is directly proportional to the radioactivity present in the band after gel electrophoresis. The radioactivity associated with the individual product bands was measured and used to calculate the percentage of DNA synthesis products in the polymeri- zation reaction.

The percent extension of the primer was calculated from the radioactivity present in all product bands and the primer band, using the equation given below:

% Primer extension = [(radioactivity in all product bands)/(total ra- dioactivity in the product and primer bands) - (radioactivity due to non-annealed primer)] x 100

In the formation of the site-modified primed template, the primer was used in excess. The percent of the non-annealed primer was estimated by electrophoresis on 12% polyacryl- amide non-denaturing gels and used to calculate the radioac- tivity due to the non-annealed primer in the polymerization reaction.

The extension of the primed template was essentially quan- titative (>95%) in the presence of M e or Mn2+, and low (10 PM) or high (200 p ~ ) dNTPs in 15 min. Data are presented in relative percentages rather than in absolute yields. While serving the same purpose, this focuses data analysis on the outcome of KfPolI encountering the N3-Et-dT lesion in the template. Furthermore, the ratio of the various DNA synthesis products mediated by the N3-Et-dT lesion are not expected to change even if the extension of the primer is inhibited at the initiation stage, and it is far easier to estimate the relative percentage of DNA synthesis products.

Inhibition of DNA Replication by the N3-Et-dT ksion- Using the primed-template system, we found that the N3-Et- dT lesion, present at a single preselected site in the DNA template, blocked DNA replication by KfPolI in the presence of Mn2+ and 10 PM dNTP (Fig. lB, lane 1). In the control primed-template (containing dT in place of N3-Et-dT), the DNA synthesis proceeded to the 5’ terminus of the template without interruption (Fig. lA, lanes 2 and 3 ) , producing a 34- nucleotide product. The results indicate that the N3-Et-dT lesion was responsible for blocking DNA replication when the site-modified template was used. DNA synthesis was termi- nated preceding and opposite the N3-Et-dT lesion. The 25- nucleotide preincorporation blocked product was obtained in a 39% yield. DNA sequencing of this product confirmed the DNA synthesis termination site to be one nucleotide 3’ to the N3-Et-dT lesion (data not shown). A similar DNA synthesis

A B C 1 2 9 1 2 1 2

“ -34 m -34

Z “ 2 5 2 -26

* -17 -17 r, -17

- I . I

FIG. 1. Analysis of DNA products synthesized by KfPolI on a primed template. The primed template contained dT or an N3- Et-dT adduct at template position 26 from the 3’-end. The primed template (0.05 pmol) was incubated at 37 “C for 15 min with 0.2 units of KfPolI in the presence of 0.5 mM Mn2+. Products were analyzed by gel electrophoresis and autoradiography. A 25-nucleotide product corresponds to the preincorporation (adjacent to N3-Et-dT) blocked product; 26-nucleotide product depicts incorporation (opposite N3- Et-dT)-dependent blocked product; 34-nucleotide post-lesion synthe- sis product represents no block during DNA replication with the synthesis proceeding to the 5‘-end of the 34-nucleotide template. A, control primed template containing normal dT at position 26 from the 3’-end; lane I, standard 34-nucleotide marker; lane 2, polymeri- zation in the presence of 10 p~ dNTP; lane 3, polymerization in the presence of 200 p~ dNTP. B, primed template containing an N3-Et- dT adduct a t position 26 from the 3’-end; lane I, polymerization in the presence of 10 p~ dNTP; lane 2, standard 25-nucleotide marker. C, primed template is the same as in B ; lane I , standard 26-nucleotide marker; lane 2, polymerization in the presence of 200 p~ dNTP.

block by N3-Et-dT was obtained under identical conditions using M e in place of Mn2+ except that the preincorporation blocked product was obtained in >90% yield (Grevatt et al., 1990, Bhanot et al., 1990). The results are consistent with the published reports that N3-Et-dT randomly present in syn- thetic polymers impedes DNA synthesis by KfPolI and avian myeloblastosis virus-reverse transcriptase (Saffhill, 1985; Huff and Topal, 1987).

KfPolI was also able to incorporate a nucleotide opposite the N3-Et-dT lesion. This was evident from the formation of a 26-nucleotide incorporation-dependent blocked product in a 60% yield. The DNA synthesis was terminated after incor- porating a nucleotide opposite N3-Et-dT, suggesting that the newly formed base pair at the 3‘-end of the growing chain is inhibitory to DNA replication. The formation of the post- lesion synthesis product in negligible amounts supported this hypothesis. The results demonstrate that the N3-Et-dT le- sion, present in a natural DNA template, provides a strong block to DNA replication by KfPolI with the DNA synthesis terminating exclusively at and immediately 3‘ to the lesion.

DNA sequencing of the incorporation-dependent blocked product revealed that in the presence of Mn2+, dA was incor- porated opposite N3-Et-dT (Fig. 2). The results are consistent with earlier studies of DNA replication in the absence of an “instructive” base-pairing template (Strauss et al., 1982) that dA is preferentially incorporated opposite noncoding lesions during DNA replication.

Degradation of the 17-nucleotide primer by the associated 3’ + 5’ exonuclease activity of KfPolI was observed (Fig. 1). Degradation of the DNA synthesis products was not detected. This indicates that only those primer molecules that are not in the process of elongation are subject to degradation, con- sistent with earlier findings that KfPolI degrades oligomers in the absence of template + dNTPs (Bhanot et al., 1979).

The percentage of DNA synthesis products formed in the presence of Mn2+ as the enzyme cofactor was not affected by

1272 Significance of DNA Damage at Thymine N3 T

C C G G A

- A

FIG. 2. DNA sequence analysis of incorporation-dependent blocked product by Maxam-Gilbert procedure. The presence of a band only in the dA-specific lane at position 26 indicates exclusive incorporation of dA opposite N3-Et-dT. N3-Et-dT is present in the template at position 26 from the 3'-end.

0 1 2 3 4 5 Mn++ concentration (mM)

FIG. 3. Effect of Mn2+ concentration on the formation of DNA products synthesized by KfPolI on the primed template containing a single N3-Et-dT adduct. The primed template (0.05 pmol) was incubated at 37 "C for 15 min with 0.2 units of KfPolI in the presence of 10 p~ dNTP and varying concentrations of Mn2+. The products were analyzed as described (Bhanot et al., 1990). 0 ,25- nucleotide preincorporation-dependent blocked product; A, 26-nu- cleotide incorporation-dependent blocked product; 0, 34-nucleotide post-lesion synthesis product.

temperature (16-37 "C) or by reaction time (15-120 min) (data not shown). The results suggest that the synthesis termination bands seen in our experiments represent perma- nent synthesis blocks by the N3-Et-dT lesion and not merely pause sites.

The effect of MnZ+ concentrations on the accumulation of DNA synthesis-blocked products was significant. Fig. 3 rep- resents the effect of Mn2+ concentration on the percent of DNA products synthesized by KfPolI on the site-modified primed template. Each point represents an average of three experiments, with a standard deviation of 3.5 to 4.1% for the preincorporation blocked product, 2.4-4.3% for the incorpo- ration-dependent blocked product, and 0.4-0.6% for the post- lesion synthesis product. When the concentration of Mn2+ was increased from 0.5 to 5 mM, the incorporation-dependent blocked product (26 nucleotides) was decreased from 60 to 35%, while the preincorporation-blocked product (25 nucleo- tides) was increased from 39 to 64% (Fig. 3). The post-lesion synthesis product was not affected and remained at ~ 1 % . The results indicate that higher concentrations of Mn'" may mod-

ify polymerase and 3' + 5' exonuclease activities of KfPolI. A change in the ratio of the polymerase and the exonuclease activities can affect the accumulation of the 25-nucleotide preincorporation blocked product in two ways. First, the ratio of the modified polymerase and exonuclease activities may not facilitate incorporation of a nucleotide opposite N3-Et- dT. Second, the formation of a 25-mer may be enhanced by increased exonuclease editing of the 26-mer product. The presence of an abnormal base pair, N3-Et-dT.dA, at the 3'- end stalls the DNA replication and may provide a greater opportunity for proofreading to occur. All of the following experiments were performed in the presence of 0.5 mM Mn". Whenever EDTA was present in the reaction mixture, the Mn'+ concentration was increased to compensate for the EDTA concentration.

DNA Synthesis Past the N3-Et-dT Lesion-DNA synthesis past the N3-Et-dT lesion (post-lesion synthesis) was sensitive to the dNTP concentration (Fig. lC, lane 2, and Fig. 4). Fig. 4 represents the effect of dNTP concentration on the per- centage of various DNA synthesis products. Each point was measured three times and the standard deviations ranged from 1.1 to 3.1% for the preincorporation-blocked product, 2.2-4.1% for the incorporation-dependent-blocked product, and 0.4-3.4% for the post-lesion synthesis product. When the dNTP concentration was increased from 10 to 200 PM, the 34-nucleotide post-lesion synthesis product was increased from ~1 to 68% (Fig. 4). The incorporation-dependent blocked product was decreased from 60 to 25%, while the preincorporation blocked product was lowered from 39 to 7% (Fig. 4). The presence of high concentrations of dNTPs in- creased the extension of the preincorporation blocked product to the incorporation-dependent blocked product, a major part of which is then extended to the 5' terminus of the template yielding a 34-nucleotide post-lesion synthesis product. In addition to the 34-nucleotide product, the post-lesion synthe- sis product contained 33-nucleotide ( ~ 3 5 % ) and 35-nucleotide ( 4 % ) products (Fig. lC, lane 2). These products have iden- tical sequences as compared with the 34-nucleotide product. The last nucleotide (dA) in the 33-mer was missing from the 3'-end. The nucleotide present at the 3'-end of the 35-nucleo- tide product was not investigated.

Both the post-lesion synthesis product and the incorpora- tion-dependent blocked product isolated from the same ex- periment were sequenced by the modified Maxam-Gilbert procedure. The post-lesion synthesis product exclusively con-

0 40 80 120 160 200 dNTP concentmtion(pM)

FIG. 4. Effects of dNTP concentration on the formation of DNA products synthesized by KfPolI on the primed template containing a single N3-Et-dT adduct. The primed template (0.05 pmol) was incubated at 37 "C for 15 min with 0.2 units of KfPolI in the presence of 0.5 mM Mn'+ and varying concentrations of dNTP. The products were analyzed as described (Bhanot et al. 1990). 0, 25- nucleotide preincorporation blocked product; A, 26-nucleotide incor- poration-dependent blocked product; 0, 34-nucleotide post-lesion synthesis product.

Significance of DNA Damage at Thymine N3 1273 T A

C C G G

T-

FIG. 5. DNA sequence analysis of translesion synthesis product by Maxam-Gilbert procedure. Presence of a band only in the dT-specific lane at position 26 indicates exclusive incorporation of dT opposite N3-Et-dT. N3-Et-dT is present in the template at position 26 from the 3’-end.

tained dT as the nucleotide incorporated opposite the N3-Et- dT lesion (Fig. 5). DNA sequence analysis of the incorpora- tion-dependent blocked product revealed incorporation of dA opposite N3-Et-dT. The results demonstrate that in the pres- ence of 200 p~ dNTPs and 0.5 mM Mn2+ both dA and dT are incorporated opposite the N3-Et-dT lesion. After incorpora- tion of dA, the N3-Et-dT-dA base pair at the 3‘-end of the growing chain is resistant to further extension and the DNA synthesis is terminated, yielding a 26-nucleotide incorpora- tion-dependent blocked product. Incorporation of dT opposite the N3-Et-dT lesion allows extension of the growing chain to yield the post-lesion synthesis product.

DISCUSSION

The studies described in this paper suggest a dual role for the N3-Et-dT lesion, which is formed both in vitro and in vivo by exposure of DNA to ethylating agents (Singer and Grunberger, 1983). The N3-Et-dT lesion blocks DNA repli- cation in vitro and may terminate DNA synthesis in vivo, contributing to the cytotoxicity of the alkylating agents. In- corporation of dT opposite this lesion in the post-lesion synthesis product suggests a promutagenic role for N3-Et-dT leading to A. T + T e A transversion mutation.

We have shown that in the presence of Mg2+ as a divalent cation and low dNTP concentration (10 p ~ ) , in vitro DNA synthesis by KfPolI on a natural DNA template containing N3-Et-dT at a single preselected site was predominantly (>go%) blocked immediately 3‘ to the N3-Et-dT lesion (Gre- vatt et al., 1990; Bhanot et al., 1990). Here we report that when 0.5 mM Mn2+ was substituted for M F in the polymer- ization reaction, a nucleotide was incorporated opposite N3- Et-dT with subsequent termination of the DNA synthesis. This resulted in the accumulation of a 26-nucleotide incor- poration-dependent blocked product (Fig. lB, lane 1 ). No appreciable post-lesion-synthesis product was produced under these conditions. The results reflect an impairment of the nascent chain extension, presumably due to the nonpairing of the incoprorated nucleotide with N3-Et-dT. This is con-

sistent with the interference in hydrogen bond formation due to the presence of the ethyl group in N3-Et-dT at a base- pairing site. Although the N3-Et-dT adduct is formed in DNA to a limited extent (Beranek et al., 1980; Singer and Grunber- ger, 1983), the adduct may play a partial role in determining the cytotoxic properties of the ethylating agents.

DNA sequencing of the incorporation-dependent blocked product revealed that dA is incorporated opposite N3-Et-dT (Fig. 2). The results indicate that the insertion of nucleotides a t the non-instructional lesion, N3-Et-dT, does not occur at random. The nucleotide specificity observed can be under- stood as a reflection of the polymerase affinity for purines over pyrimidines under template-free conditions (Englund et al., 1969). Our data are consistent with results reported for other non-instructional lesions and support the hypothesis that in the absence of instruction from the template, DNA polymerases tend to insert a purine (predominantly dA) op- posite the non-instructional lesion (Strauss et al., 1982).

The incorporation of a nucleotide opposite N3-Et-dT in the presence of Mn2+ may be due to the interaction of Mn2+ with the primer-template-enzyme complex, possibly altering the conformation at the active site of the polymerase and the 3’ 4 5’-exonuclease activities. Whether the effects of Mn2+ are primarily on the enzyme conformation or secondary to the changes in the DNA structure is not fully established. At low concentrations, Mn2+ has been shown to result in decreased specificity of base selection by the E. coli DNA polymerase I and decreased specificity of proofreading by the 3 ’ + 5 ’ - exonuclease activity (El-Deiry et al., 1988). Increased sensi- tivity of the 3’ + 5’-exonuclease activity to inhibition by the nucleoside 5’-monophosphate has also been observed in the presence of Mn2+ (El-Deiry et al., 1988).

Post-lesion synthesis by KfPolI is sensitive to dNTP con- centration (Fig. 4). The synthesis was increased from 4 to 68% when the concentration of dNTPs was increased from 10 to 200 p M in the presence of 0.5 mM Mn2+. Evidently the high dNTP concentration increased the rate of polymeriza- tion of the next correct nucleotide (“next nucleotide” effect) following the mispair at N3-Et-dT, protecting the unpaired nucleotide from excusion by the 3’ + 5’ exonuclease proof- reading activity (Ninio, 1975). The resulting nascent chains, with the correct base pair at the 3’ terminus, were then extended to completion. Significant post-lesion synthesis sug- gests that Mn2+ may also be contributing to the stability of the “frayed primer terminus and the polymerase active site complex enhancing the addition of the next nucleotide. The major post-lesion synthesis product (34-nucleotide) was iden- tical in chain length to the template. Products containing one base less (-1) and one base more (+I) than the template were also formed (Fig. IC, lane 2). The +1 product was formed in a very low yield and may have resulted from the blunt-end addition of a nucleotide to the synthesized 34-nucleotide duplex. Blunt-end addition reactions catalyzed by KfPolI have been observed (Clark et al., 1987). The formation of the -1 product may reflect kinetic factors related to slowing of the chain extension and has been observed in other studies (Takeshita et al., 1987). Failure to detect -1 and +1 addition products in the control (Fig. lC, lanes 2 and 3) suggests that the formation of these products may be stimulated by the presence (in the template) of a DNA lesion that impedes the progress of DNA synthesis.

DNA sequencing of the post-lesion synthesis products re- vealed that dT was exclusively incorporated opposite N3-Et- dT (Fig. 5). The incorporation of dT opposite N3-Et-dT suggests a promutagenic role for the N3-Et-dT lesion leading to transversions at A - T base pairs. DNA sequencing of the

1274 Significance of DNA Damage at Thymine N3

incorporation-dependent blocked product from the same ex- periment (200 PM dNTPs and 0.5 mM Mn2+) revealed that dA is incorporated opposite N3-Et-dT. Identical results were obtained from a similar product synthesized at low dNTP concentration (10 PM). The absence of dA corresponding to N3-Et-dT in the post-lesion synthesis product (Fig. 5 ) sug- gests that the N3-Et-dT.dA base pair is energetically and sterically unfavored for extension by KfPolI. When dA is incorporated opposite N3-Et-dT, the nascent chains contain- ing the N3-Et-dT.dA basepair at the 3' terminus are not extended, perhaps due to dissociation of the polymerase from the primer-template-enzyme complex. Alternatively, when d T is incorporated opposite N3-Et-dT, the nascent chains con- taining the N3-Et-dT-dT pair at the 3' terminus are able to extend, yielding a post-lesion synthesis product. This is evi- dent from the exclusive incorporation of d T opposite N3-Et- d T in the post-lesion synthesis product and the absence of d T incorporation corresponding to the lesion in the incorpo- ration-dependent blocked product. The mechanisms for ex- tension of the N3-Et-dT.dT base pair and blockage by the N3-Et-dT - dA base pair are not known. Physicochemical stud- ies to investigate the conformation of these base pairs are in progress.

Since the post-lesion synthesis at 10 PM dNTP and 0.5 mM Mn'+ is negligible ( ~ 1 % ) and dA is incorporated opposite N3- Et-dT in the incorporation-dependent blocked product under these conditions, it appears that at low dNTP concentration, dATP competes with dTTP and other nucleotides for incor- poration opposite N3-Et-dT in accordance with the " A rule in mutagenesis. However, at a high dNTP concentration (200 PM), d T is also incorporated opposite N3-Et-dT. The relative net incorporation of dT and dA at high dNTP concentration, which corresponds to the post-lesion synthesis product and the incorporation-dependent blocked product, is 68 and 25%, respectively. The results suggest that the N3-Et-dT lesion may be more mutagenic under conditions that affect the nucleotide pools in a cell (Das et al., 1983; Volkin e t al., 1983).

As an alternative to miscoding, the correct coding by a transiently misaligned template primer has been shown to produce base substitution mutations at certain DNA se- quences (Kunkel and Alexander, 1986; Kunkel and Soni, 1988). Fig. 6 illustrates the incorporation of d T (pathway A) and dA (pathway B) opposite N3-Et-dT through correct cod- ing of the transiently misaligned template primer sequences during DNA replication. However, many DNA synthesis prod- ucts, that could be formed by the addition of two nucleotides before realignment or two cycles of slippage, were not detected in DNA replication with KfF'olI (this paper) or a highly processive T7 DNA polymerase.' These studies suggest that addition of dT and dA opposite N3-Et-dT may occur by direct incorporation opposite the lesion, rather than through tran- sient-misalignment.

The findings reported in this paper provide additional sup- port for the A rule in mutagenesis. When cells are confronted with non-instructional lesions, they tend to insert dA. If the non-instructional lesion is a modified thymidine such as N3- Et-dT, the incorporation of dA opposite the modification is not mutagenic. The formation of the N3-Et-dT. dA base pair at the growing end of the chain is resistant to further exten- sion in vitro and terminates the DNA synthesis. Under normal cellular conditions, replication past the N3-Et-dT. dA base pair may either not occur or occur with low efficiency, as observed in the bypass of apurinic/apyrimidinic sites and UV lesions by the DNA polymerase 111 holoenzyme in vitro (Liv- neh, 1986; Hevroni and Livneh, 1988). The results suggest

P. C. Grevatt, J. M. Donahue, 0. S. Bhanot, unpublished results.

Primer strand 5"T-T-T-T

Template strand 3*-A-A-A-A-T*-l-l-

I Misalignment I I

T / \

-A-A-A-A-T*-T-T- -T-T T

Incorporation

V T

/ \

-A-A-A-A-T*-T-T- -T-T T-T

I Real ignwnt V

-A-A-A-A-T*-T-T- "T-T-T-T-T

1 V

-T-T-T-T -A A-A-T*-T-T-

\ /

Incorporation

-A A-A-T*-T-T- -T-T-T-T- A

\ /

* I Realignment v

-T-T-T-T-A -A-A-A-A-T*-T-T-

T* = N3-Et-dT

FIG. 6. Outline of nucleotide incorporation opposite N3-Et- dT by transient misalignment of primer template. Puthwuy A shows the incorporation of dT opposite N3-Et-dT. During the stalling of KPolI by N3-Et-dT, dT (in the run of four dTs in the primer) is looped out with subsequent extension by incorporation of the correct dT opposite template dA 3' to the N3-Et-dT lesion. Realignment, before additional incorporation can occur, creates a replication com- plex with the N3-Et-dT.dT base pair at the replication fork. This base pair is capable of extension and yields a 34-nucleotide post- lesion synthesis product with dT corresponding to N3-Et-dT. Puth- way B depicts the incorporation of dA opposite N3-Et-dT. The dA within the run of four dAs in the template is looped out, creating a replication complex with the N3-Et-dT.dA base pair at the growing end of the chain after correct incorporation of dA opposite dT 5' to the N3-Et-dT lesion and subsequent realignment. The N3-Et-dT. dA base pair is resistant to extension and terminates DNA synthesis, producing a 26-nucleotide incorporation-dependent blocked product with dA corresponding to N3-Et-dT.

that the N3-Et-dT may contribute in part to the cytotoxicity of ethylating agents.

The fixation of DNA lesions induced in E. coli by alkylating agents occurs by both SOS-dependent and SOS-independent pathways (Eckert and Drinkwater, 1987; Cuoto et al., 1989). While transition mutations are the predominant type of base substitution mutations observed in SOS-independent muta- genesis (Miller, 1983), transversion mutations at A ' T base pairs represent an important component of alkylation muta- genesis under conditions of SOS processing. A significant number of A. T + T. A and A. T + C . G transversior, muta- tions were observed after N-methyl-N'-nitro-N-nitrosoguan- idine treatment of alkA E. coli, which are more sensitive to SOS induction by N-methyl-N'-nitro-N-nitrosoguanidine (Foster and Eisenstadt, 1985). A .T transversions were also produced when SOS-induced E. coli was transfected with N- methyl-N-nitrosurea-modified DNA treated with @-methyl- guanine DNA-methyltransferase to correct for 06-Me-dG and 04-Me-dT lesions (Cuoto et al., 1989). 0'- and/or @-ethylthy- mine lesions have been correlated with A. T + C . G transver- sion mutations produced in ENU-treated Salmonella typhi- murium (uurB-) strains carrying the plasmid pKMlOl (Zie- lenska et al., 1988). Frequent induction of A .T + C.G transversions was also observed when SOS-induced E. coli was treated with ENU (Eckert e t al., 1989). Alkylated DNA

Significance of D N A Damage at Thymine N 3 1275

lesions that normally block DNA replication have been sug- gested to be responsible for transversion mutations at A . T base pairs in SOS-induced bacteria. The importance of A . T transversion mutations in mammalian systems has been dem- onstrated (Popp et al., 1983; Lewis et al., 1985; Bargmann et al., 1986; Perantoni et al., 1987; Stowers et al., 1988; Eckert et al., 1988).

The induction of A. T mutations by alkylating agents in mammalian systems or through the SOS-dependent pathway in bacteria suggests that dA or dT adducts and/or breakdown products of these adducts are responsible for transversion mutations at the A. T base pairs. The studies reported in this paper implicate the N3-Et-dT lesion in A . T + T .A trans- version mutagenesis. These results provide a basis for under- standing the mechanisms of activation of oncogene neu by an A . T + T . A transversion event in tumors induced by trans- placental exposure of rats to ENU (Bargmann et al., 1986; Perantoni et al., 1987).

The misincorporation of dT opposite N3-Et-dT suggests that the role of Mn2+ ions in modifying the fidelity of KfPolI in DNA replication in uitro may be comparable to the SOS- induced functions in bacteria. Inside the cell the SOS-induced proteins may alter the fidelity of the replication complex, facilitating the incorporation of dT opposite N3-Et-dT and subsequent extension. This hypothesis suggests that N3-Et- dT may be a promutagenic lesion capable of inducing A. T + T. A transversion mutations under SOS-induced conditions. This is being investigated in SOS-induced bacteria using a PX174-based site-specific mutagenesis system (Bhanot and Ray, 1986).

REFERENCES

Bargmann, C. I., Hung, M.-C., and Weinherg, R. A. (1986) Cell 45,

Basu, A. K., and Essigmann, J . M. (1988) Chem. Res. Toxicol. 1 , 1-

Beranek, D. T., Weis, C. C., and Swenson, D. H. (1980) Carcinogenesis

Bhanot, 0. S., and Ray, A. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,

Bhanot, 0. S., Khan, S. A., and Chambers, R. W. (1979) J. B i d .

Bhanot, 0. S., Grevatt, P. C., Donahue, J . M., Gabrielides, C. N., and

Clark, J . M., Joyce, C. M., and Beardsley, G. P. (1987) J. Mol. Biol.

Cuoto, L. B., Chaudhari, I., Donahue, B. A., Demple, B., and Essig-

Das, S. K., Benditt, E. P., and Loeh, L. A. (1983) Biochem. Biophys.

649-657

18

1,595-606

7348-7352

Chem. 254 , 12684-12693

Solomon, J. J. (1990) Biochemistry, in press

198,123-127

mann, J. M. (1989) J. Bacteriol. 171 , 4170-4177

Res. Commun. 114,458-464 Drinkwater, N. R., Miller, E. C., and Miller, J. A. (1980) Biochemistry

19,5087-5082 Eckert, K. A., and Drinkwater, N. R. (1987) Mutat. Res. 178 , 1-10 Eckert, K. A,, Ingle, C. A., Klinedinst, D., and Drinkwater, N. R.

Eckert, K. A., Ingle, C. A., and Drinkwater, N. R. (1989) Carcinogen-

El-Deiry, W. S., So, A. G., and Downey, K. M. (1988) Biochemistry

(1988) Mol. Carcimgenesis 1, 50-56

esis 10, 2261-2267

27,546-553 Englund, P., Huberman, J., Jovin, T., and Kornberg, A. (1969) J .

Evensen, G., and Seeberg, E. (1982) Nature 296 , 773-775 Foster, P. L., and Eisenstadt, E. (1985) J. Bacterid. 163 , 213-220 Grevatt, P. C., and Bhanot, 0. S . (1990) Proc. Am. Assoc. Cancer Res.

Grevatt, P. C., Donahue, J. M., and Bhanot, 0. S . (1990) Proc. Am.

Hevroni, D., and Livneh, Z. (1988) Proc. Natl. Acad. Sci. U. S. A . 8 5 ,

Huff, A. C., and Topal, M. D. (1987) J. Bid. Chem. 262 , 12843-

Kunkel, T. A., and Alexander, P. S. (1986) J. Biol. Chem. 2 6 1 , 160-

Kunkel, T. A., and Soni, A. (1988) J. Biol. Chem. 263, 14784-14789 Lawley, P. D., and Brookes, P. (1963) Biochem. J. 89, 127-135 Lewis, S. E., Johnson, F. M., Skow, L. C., Popp, D., Barnett, L. B.,

and Popp, R. A. (1985) Proc. Natl. Acad. Sci. U. S. A . 82, 5829- 5831

Lindahl, T., and Sedgwick, B. (1988) Annu. Reu. Biochem. 5 7 , 133- 157

Little, J . W., and Mount, D. W. (1982) Cell 29, 11-22 Livneh, Z. (1986) J. Bid. Chem. 261,9526-9533 Loeb, L. A., and Kunkel, T. A. (1982) Annu. Reu. Biochem. 52, 429-

Loechler, E. L., Green, C. L., and Essigmann, J. M. (1984) Proc. Natl.

Maxam, A. M., and Gilbert, W. (1980) Methods Enzymol. 6 5 , 497-

McCarthy, T., Karran, P., and Lindahl, T. (1984) EMBO J. 3 , 545-

Mendelman, L. V., Boosalis, M. S., Petmska, J., and Goodman, M.

Miller, J . H. (1983) Annu. Reu. Genet. 17, 215-238 Ninio, J. (1975) Biochimie 5 7 , 587-595 Perantoni, A. O., Rice, J. M., Reed, C. D., Watatani, M., and Wenk,

Popp, R. A., Baliff, E. G., Skow, L. C., Johnson, F. M., and Lewis, S.

Preston, B. D., Singer, B., and Loeb, L. A. (1986) Proc. Natl. Acad.

Saffhill, R. (1985) Chem.-Bid. Interact. 5 3 , 121-130 Saffhill, R., Margison, G. P., and O’Connor, P. J . (1985) Biochim.

Singer, B. (1986) Cancer Res. 46,4879-4885 Singer, B., and Grunberger, D. (1983) Molecular Biology of Mutagens

and Carcinogens, Plenum Publishing Co., New York. Singer, B., Sagi, J., and Kusmierek, J . T. (1983) Proc. Natl. Acad.

Sci. U. S. A . 80,4884-4888 Stowers, S. J., Wiseman, R. W., Ward, J. M., Miller, E. C., Miller, J.

A., Anderson, M. W., and Eva, A. (1988) Carcinogenesis 9 , 271- 276

Biol. Chem. 2 4 4 , 3038-3044

31, 100 (ahstr)

Assoc. Cancer Res. 31 , 100 (abstr)

5046-5050

12850

166

457

Acad. Sci. U. S. A. 81, 6271-6275

560

550

F. (1989) J. Biol. Chem. 264 , 14415-14423

M. L. (1987) Proc. Natl. Acad. Sci. U. S. A . 84, 6317-6321

E. (1983) Genetics 105 , 157-167

Sci. U. S. A . 83,8501-8505

B i ~ p h y ~ . Acta 8 2 3 , 111-135

Strauss, B. S. (1985) Cancer Suru. 4 , 493-516 Strauss, B. S., Rahkin, S., Sagher, D., and Moore, P. (1982) Bichimie

Takeshita, M., Chang, C.-N., Johnson, F., Will, S., and Grollman, A.

Volkin, E., Boling, M. E., Lee, W. H., and Jones, M. H. (1983)

Walker, G. (1984) Microbiol. Reu. 48,60-93 Witkin, E. M. (1976) Bacterid. Reu. 40 , 869-907 Zielenska, M., Beranek, D., and Guttenplan, J. (1988) Enuiron. Mol.

64,829-838

P. (1987) J. Bid . Chem. 262, 10171-10181

Biochim. Biophys. Acta 7 5 5 , 217-224

Mutagen. 11 , 473-485