J. Mol. Biol. (1990) 212, 433-436

COMMUNICATIONS

Mutational Hotspot Variability in an Ultraviolet-treated Shuttle Vector Plasmid Propagated in Xeroderma Pigmentosum and Normal Human Lymphoblasts and Fibroblasts Saraswathy

Seetharam,

Kenneth

H. Kraemer,

Haywood

L. Waters

Laboratory

of Molecular Carcinogenesis Cancer Institute Building 3’7, Room 3324 Bethesda, MD 20892, U.S.A. National

and Michael

M. Seidman

Otsulca America Pharmaceutical, Inc. 9900 Medical Center Drive Rockville, MD 20850, U.S.A. (Received 9 August

1989; accepted 1 December 1989)

The mutagenesis shuttle vector, pZ189, was treated with ultraviolet (u.v.) radiation ?ln vitro and passed through a DNA repair-deficient lymphoblastoid cell line derived from a patient with xeroderma pigmentosum complementation group A (XP-A) (XP12BE(EBV)) and a DNA repair-proficient lymphoblastoid cell line (GM606(EBV)). After U.V. treatment, plasmid survival was lower and mutation frequency higher with the XP-A cells mirroring the survival and mutagenesis of the host cells. The nature of the mutations in the suppressor tRNA marker gene was determined by direct sequence analysis. The G. C to A. T transition was the dominant (85%) base substitution mutation with the XP lymphoblasts and was the major (56%) base substitution mutation with the repair-proficient lymphoblasts. We found a G. C to A. T transition mutational hotspot with the XP lymphoblasts not seen in our previous experiments with fibroblasts from the same patient. Comparison of the data presented here with our results with DNA repair-deficient and DNA repair-proficient fibroblasts suggests that hotspot variability is not due to genetic polymorphism or repair capacity of the cells. Instead it appears that cellular factors can influence the probability of mutagenesis of modified DNA at particular sites.

mentation group A (Robbins et al., 1974; Cleaver & Kraemer, 1989). The most frequent mutation with both lines was the G. C to A. T transition, in agreement with studies in many other systems (Drake, 1963; Howard & Tessman, 1964; Miller, 1985; Hauser et al., 1986; Keyse et al., 1988; Hsia et al.; 1989; Drobetsky et al., 1987). There were G.C to A. T hotspots at the same positions with the two cell lines and a major G. C to A. T hotspot in the XP cells that was not seen with the repair-proficient cells. We have observed different hotspot patterns in experiments with other cell lines (Seetharam et al., 1987, 1989). Since both repair-proficient #and XP cells were used, variable repair of modified sites

We have been studying ultraviolet (u.v.7) mutagenesis in DNA repair-deficient and DNA repairproficient human cell lines (Bredberg et al., 1986; Xeetharam et al., 1987; 1989; Brash et al., 1987). We used the pZ189 shuttle vector system to determine the nature and location of mutations in the suppressor tRNA marker gene (Seidman et al., 1985; Kraemer 85 Seidman, 1989). Our initial experiments (Bredberg et al., 1986) were with skin fibroblast cell lines derived from a normal individual and from a patient with the DNA repair-deficient, cancer-prone disorder, xeroderma pigmentosum (XP), completabbreviations used: u.v., ultraviolet; XP, xeroderma pigmentosum. 0022-2S36/90/070433%04

$03.00/O

433

0

1990 Academic

Press Limited

X. Xeetharam

434

et al.

could have influenced the probability of mutagenesis at particular sites. The cell lines were derived from different donors and this may have contributed to the hotspot variability. We have attempted to address these possibilities by determining the mutagenic spectrum for the u.v.-treated shuttle vector passed through a repair-proficient lymphoblastoid cell line and a lymphoblastoid cell line derived from the same XP complementation group A patient (XPlSBE) from whom the fibroblasts in our ea,rlier experiment (Bredberg et al., 1986) were established. Plasmid

survival

and mutation

frequency

The pZ189 plasmid DNA was treated with 254 nm radiation at doses (50 to 1000 JmW2) such that between 0.1 and 2 U.V. photoproducts were formed in the 200 base-pair region carrying the supF marker gene (Protid-Sabljid & Kraemer, 1985). The modified D,IUTA was transfected into the DNA lymphoblastoid repair-proficient and the DNA rep~~defici?~ GM606(EBV); lymphoblastoid cell line XPlBBE(EBV), by use of DEAE dextran (Seetharam et al., 1989). After 48 hours the progeny plasmids were harvested and introduced into the Escherichia coli indicator strain MBM7070. The yield of colonies was determined and the colonies carrying mutant plasmids (white or light blue colonies) were picked and purified. Plasmid survival was reduced more sharply in the XP cells as compared to the repair-proficient cells (Fig. l), which is consistent with the survival of these two cell lines after direct treatment with U.V. (Perera et al., 1986). We found a much steeper increase in mutation frequency with the XP cells than with the repair-proficient cells (Fig. 2). At 150 Jme2 the mutation frequency in the XP cells was nearly 200-fold higher than the background. A similar hypermutability has been observed in experiments in which normal and XP cells were treated directly with U.V. (Maher et al., 1979; DeLuca et al., 1984). These data further support our previous observation that the survival and mutagenesis of the vector reflect the survival and mutagenesis of the cells after exposure to U.V. (Bredberg et al., 1986; Brash et al., 1987; Seetharam et al., 1987; Kraemer et al., 1988). ikfutational

21 0

I 100

200

/ , 1 300 400 500 600 700 800 Ultraviolet dose to plasmid (J mm’)

900

-z

Figure 1. Survival of u.v.-treated pZ189 replicated in DNA repair-deficient and DICTA repair-proficient human lymphoblastoid cell lines. The survival of the plasmid the repair-deficient cells was much lower with (XPlZBE(EBV)) (filled circles, 6 experiments) than with the repair-proficient cells (GM606(EBV)) (open circles. 6 experiments).

and 168 and, to a lesser degree, at 169. There was no hotspot at position 159 with the repair-proficient lymphoblasts in contrast to the XP pattern @J= 0.02; Fisher’s exact test)

analysis

Sequence analysis of 96 independent mutant plasmids from the XP cells and 102 independent mutant plasmids from the repair-proficient cells revealed that the major mutation was the G. C to A. T transition (8.5% with the XP cells and 56% with the repair-proficient cells) as found in our earlier work. The majority of these were single base substitution mutations. In the XP profile this mutation was found in several strong hotspots at base-pairs 123, 156, 159, 168 and, to a lesser degree, at 169. The pattern with the repair-proficient lymphoblasts showed G ’ C to A. T hotspots at base-pairs 123, 156,

I

003 0

1 100

I 200

I 300

1 400

( 5;

Ultrawolet dose to plosm~d (J m?)

Figure 2. MutaGion frequency of u.v.-treated pZl89 replicated in DNA repair-deficient and DPiA repair-proficient human lymphoblastoid cell lines. The mutation frequency of the plasmid was much higher wit,h the repair-deficient cells (XP12BE(EBV)) (filled circles, 24 to 51 samples per point) than with the repair-proficient cells (GLM606(EBV)) (open circles, 16 to 52 samples per point). The error bars indicate the standard error of the mean. Approximately 25,000 colonies were counted at 0 Jme2 with each cell line.

Communications

156

168

169

123

159

Bose-pairs

Figure 3. Proportion of G. C to A. T transition mutations at 5 hotspots in u.v.-treated pZl89 propagated in DNA repair-deficient and DNA repair-proficient human lymphoblasts and fibroblasts. Mutation data are presented for XPlZBE(EBV)’ (a, total 92 G.C to A.T mutants; this report), GM606(EBV) (Q> total 60 G.C to A.T mutants; this report), XPl2BE(SV40) (m, total 220 G.C to A.T mutants; Bredberg et al., 1986; S. Seetharam & M. M. Seidman, unpublished data), GMO637(SV40) ( q , total 59 G.C! to A.T mutants; Bredberg et al., 1986), and XPBBE(SV40) (@, total 57 G.C to A.T mutants; Seetharam et al., 1987).

These data are shown in Figure 3 in comparison with our results with repair-proficient fibroblasts (GMO637(SV40): Bredberg et al., 1986) and XP fibroblasts from complementation groups A (XPlBBE(SV40): Bredberg et aZ., 1986; S. Seetharam & M. M. Seidman, unpublished data) and D (XPGBE(SV40): Seetharam et al., 1987). Base-pair 156 was a hotspot for G. C to A.T transition mutations with all five cell lines. With the fibroblasts it represents more than 25% of these transition mutations while with the lymphoblasts the frequency was lower (14 to 17%). The other four sites each accounted for 8 to 28% of the G. C to A. T transition mutants in two to four cell lines and 5% or less in the other lines. Thus, the hotspot intensity varied with the cell line. The same four sites appeared as hotspots in the XPlSBE lymphoblasts and fibroblasts (base-pairs 156, 168, 169 and 123). However, base-pair 159 was not a hotspot with the XPl2BE fibroblasts but was with the XPl2BE lymphoblasts a hotspot (~=0004) and the XP6BE fibroblasts @