Reassociation of human lymphoblastoid cell DNA repair replicated following methyl methanesulfonate treatment

Chem.-Biol. Interactions, 39 (1982) 77--88 Elsevier/North-Holland Scientific Publishers Ltd.

77

REASSOCIATION OF HUMAN LYMPHOBLASTOID CELL DNA REPAIR REPLICATED FOLLOWING METHYL METHANESULFONATE TREATMENT

M.L. MELTZ a, NANCY J. WHITTAMb and W.H. THORNBURG c aDepartment of Radiology, University o f Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284, bgouthwest Foundation for Research and Education, San Antonio, TX 78284 and CDepartment of Pharmacology, Universiky of Arizona Health Science Center, Tucson, A Z 85724 (U.S.A.) (Received February 2nd, 1981) (Revision received July l l t h , 1981) (Accepted July 28th, 1981)

SUMMARY

The reassociation rates o f repair replicated DNA of two human lymphoblastoid cell lines, the WIL2-A3 'normal' line and the RAJI line of Burkitt's lymphoma, were examined using the DNA/DNA 'Cot' hybridization technique. The cells were treated with methyl methanesulfonate (MMS), an alkylating agent and mutagen, to induce the repair. The incorporated repair replication radioactivity in highly repetitive sequences of WIL2-A3 cell DNA reassociates as expected for a randomly distributed incorporation. The reassociation of repair radioactivity in sequences of fewer numbers of copies, however, is less than expected for a random distribution. It is less than that occurring for semiconservatively synthesized DNA of WIL2-A3 cells co-incubated with the repair labeled DNA as an internal control. The observed difference could be due to an over-representation of repair replication radioactivity in DNA sequences with fewer copies. It is unlikely to be due to residual alkali labile damage resulting from MMS treatment, since a similar difference was not observed when semiconservatively labeled DNA from cells which had been treated with MMS for the same time and at the same concentration as in the repair experiments was substituted for repair replicated DNA in the reassociation reactions. Other possible causes of the apparent difference in the reassociation rates observed ate discussed.

Abbreviations: BrUdR, 5-bromodeoxyuridine; FUdR, 5-fluorodeoxyuridine; MMS, methyl methanesulfonate; NA-AAF, N-acetoxy-2-acetylamino-fluorene; PBS, phosphate buffered saline; SSC, standard saline citrate (0.15 M sodium chloride + 0.015 M sodium citrate). 0009--2797/82/0000--0000/$02.75 © Elsevier/North-Holland Scientific Publishers Ltd.

78 INTRODUCTION

The DNA/DNA 'Cot' hybridization technique of Britten and Kohne [1] measures the rate of reassociation of isolated and denatured DNA in solution. This allows t h e characterization of DNA into classes based on the extent of hybridization which occurs as a function of the product of the time of incubation and the concentration of DNA present. Those DNA molecules (approx. 10--15%, Figs. 5 and 6) reassociating rapidly (Cot less than 10 -1 tool nucleotides- s/1 under the reaction conditions employed herein) are designated highly repetitive DNA, those reassociating more slowly (Cotvalues of 10-1-102 tool nucleotide, s/l) are called intermediate repetitive DNA (approx. 20%) and those requiring extended incubations at high concentrations before reassociation occurs (Cot greater than 102 tool nucleotides • s/l) are designated unique DNA (approx. 65--70%). The initial damage of DNA caused by either UV light or chemical treatment may or may not be distributed proportionally (uniformly) in the three repetitive classes. Likewise the repair replication resulting after this damage occurs may or may not be proportionally distributed in the three repetitive classes. Before either of these possibilities is explored, an initial question which can be investigated is the nature of the distribution of the damage and its repair as reflected by the sequence distribution of the repair replication radioactivity in preexisting cell DNA. Preliminary studies in our laboratory indicated that repair occurring during MMS treatment, in contrast to UV light [2--4], N-acetoxy.2-acetyl~minofiuorene (NA-AAF) [3] or 7-bromomethylbenz[a]anthracene [3] induced DNA repair, might show a non-uniform distribution. Our investigation of this possibility, after treatment with a high concentration of MMS to induce sufficient incorporation of repair radioactivity so as to be able to perform Cot studies at low DNA concentrations, is described below. MATERIALS AND METHODS Chemicals

The chemicals employed in this study included: hydroxylapatite Bio-Gel HTP (Bio-Rad Laboratories, Richmond, CA); ethylenediamine tetraacetic acid, disodium salt (as well as trisodium salt and free acid) (Sigma, St. Louis, MO); cesium chloride, optical grade (Schwarz/Mann, Orangeburg, NY); cesium sulphate, 99.99 + % (Column One, Inc., Ann Arbor, MI); 5-bromodeoxyuridine (BrUdR), A grade (Calbiochem, San Diego, CA); 5-fiuomdeoxyuridine (FUdR) (Hoffmann-La Roche, Inc., Nutley, NJ); hydroxyurea, A grade (Calbiochem); MMS (Eastman, Rochester, NY); ribonuclease A, phosphate free (Worthington Biochemical Corp., Freehold, NJ); pronase (Calbiochem); and E. coli DNA (Worthington Biochemical Corp.). Radiochemicals

The radiochemicals employed included: [methyl-14C]thymidine, 54--57

79 mCi/mmol; [2-14C]thymidine, 45 mCi/mmol; [6-3H]5-bromo-2'
Cell line maintenance The WIL2-A3 human lymphoblastoid cell line employed in this study was maintained in Minimal Essential Medium (Earle's) with 10% fetal calf serum supplemented with sodium pyruvate (1 mM), L~lutamine (4 raM), twice the normal amount Of non-essential amino acids and the antibiotics penicillinstreptomycin-fungizone (Grand Island Biological Company, Grand Island, NY). The cells were maintained in 125- or 250-ml capped tissue culture flasks on a New Brunswick rotary shaker in a 37°C warm room. The WIL2-A3 cells were generously supplied by Drs. Frank Dixon and Richard Lerner of Scripps Institute (La Jolla, CA). The RAJI and EB-3 lines of Burkitt's lymphoma, similarly maintained, were obtained from the American Type Culture Collection. Procedures for treatment and repair replication labelling The repair replication protocol employed is basically that previously described [2,5]. Cells were cultured to a concentration of approx. 700 000 cells/ml in 50 ml of medium in capped 250-ml Erlenmeyer flasks. The medium in the flasks was adjusted to 1 /zM FUdR and 5 /~M BrUdR and incubated for 1-h. At the end of this 1-h preincubation period, the cells were transferred into sterile conical centrifuge tubes (Corning) and centrifuged at 800 rev./min for 8 rain at room temperature in an International PR~J centrifuge. The supernatant was aspirated and the cellswere resuspended in 10 ml of medium containing 5 mM MMS, I #M~ FUdR, 1 mM hydroxyurea and 50/zCi/ml of [3H]BrUdR (13--14 Ci/mmol in sterile aqueous solution). After 3 h of incubation, the cells were centrifuged, washed by centrifugation with fresh medium containing BrUdR (5 ~uM) and FUdR (1 ~uM) and finally resuspended in this BrUdR-FUdR medium for 1 h additional incubation. The cells were then centrifuged, the medium aspirated, and the pellet of cells f~ozen in ethanol~lry ice. UV light irradiation procedure For the UV light irradiation experiment with EB-3 cells, the repair replication protocol described above was modified as follows. During the 1-h preincubation with BrUdR and FUdR, 5 #g/ml of dextran sulfate was present. At the end of the 1-h preincubation, the cells were washed twice by centrifugation with phosphate-buffered saline containing 5/~g/ml dextran sulfate and finally resuspended in the phosphate buffered saline (PBS)~lextran sulfate. The dextran sulfate was present to reduce lymphoblastoid cell attachment to the plastic surfaces during the procedures. The suspended cells in 2-ml volumes were distributed into 100 mm plastic Optilux petri dishes (Falcon) and irradiation was performed in the specially built UV irradiation device [6]. After irradiation, the cells were transferred into sterile tubes,

80

centrifuged and the PBS aspirated. The cells were then resuspended in repair replication labelling medium and subsequently treated as in the standard protocol.

DNA isolation procedure The procedure for DNA isolation has been described [5] and includes cell homogenization, RNase digestion, pronase digestion and deproteinization steps. Density gradient centrifugation procedures For these experiments, the density gradient procedure initially used [2] was modified [7] to employ two alkali density gradient centrifugations in cesium chloride-cesium sulfate at room temperature for 40 h at 42 000 rev./min in a Beckman~pinco Type 50 Ti Rotor in a Beckman L3-50 centrifuge. After centrifugation, the alkali gradients were fractionated by b o t t o m collection using an ISCO fraction collector with drop counter. The fractions containing the preexisting, non
81 2.0

:E 2 0 0 o. (1)

T! 160a o

•

,

>t-

/' !

.'\

120-

1.5

t

I

)tz

1.0

t-

I

80-

/

o_ r~
t

40-

I

-,

..1

t

/

.J

',

t

, ./ i

i

u "0.5

',

n 0

n-

!

,

;

I

,E,

,5

I

2o

2'5

~2.0

3E I 0 0 . Q. (.)

' f I i

80-

I

I

t

-I.5 ~-

>- 6 0 I.--

-L0

~ 4oo_

z ~ _J <[ (.)

-0.5 C--

,~ 2 0 n~

o /

•

s I~

l

I

!

I

I

l

5

iO

15

20

25

FRACTIONS Fig. 1. Density gradient profiles (alkaline cesium chloride-cesium sulfate) of DNA obtained

from WIL=-A3 cells repair replication labeled in the presence of 5 mM MMS. Upper figure: 1st gradient; lower figure: 2nd gradient.

Profiles for RAJI and EB-3 cell DNA are similar. We found in a series of preliminary experiments that the amount of repair radioactivity incorporated in both WIL2-A3 and RAJI cells upon MMS treatment during a 3-h treatmentlabelling interval is maximal at an MMS concentration of 5 mM. Also, the repair radioactivity incorporated in WIL2-A3 cells during treatment with MMS a t 5 mM reaches a near maximal level by 3 h (unpublished observations). Experimental designs which allowed recovery of a maximal repair replication radioactivity concentration (dpm/~g) were essential for satisfactory performance of the low Cot kinetic reactions. Our initial kinetic studies, performed over the low to intermediate Cot ranges (6.2 X 10-4--1.4 X 101) with semiconservatively synthesized DNA and repair replicated DNA from 5 mM MMS treated WIL2-A3 and RAJI cells are shown in Fig. 2. The reassociation of repair replication radioactivity appears to be less at each Cot point then radioactivity incorporated during semiconservative synthesis; further, there appears to be an increasing divergence of the two curves with increasing Cot in this range.

82 0 ~

I020-

'~ Z r-, a uJ

0 if)

3040-

50

,

t.J rr

O-

~F-Z i,i o rtuJ 13_

10-

i

,

~

~

-

20-

30"

40' 50

I

10-4

i0 -3

!

10-2

I

i0 -I

'1

100

I

101

Cot [(MOLE-NUCLEOTIDES/LITER) X SEC.]

Fig. 2. Kinetic reassociation (Cot) studies performed in the low to intermediate Cot range with semiconservatively synthesized DNA and repair replicated DNA from 5 m~I MMS treated 'WIL~-A3 cells (upper figure) and RAJI cells (lower figure), o - ~ a , DNA repair replication labeled with [3H]BrUdR; o - - - - - o , DNA semiconservatively labeled with [l'C]thymidine. Open symbols indicate average of replicate incubations; closed symbols indicate actual data points.

These results can be compared to those of a kinetic experiment with DNA from UV light irradiated (16 s of 9.4 ergs/mm2/s; 15 J/M 2) EB-3 Burkitt's lymphoma cells which were repair replication labeled as described above for 3 h immediately postirradiation (Fig. 3). There is no evidence after UV irradiation of delayed reassociation of the repair replication radioactivity in the low to intermediate Cot ranges (1.5 X 1 0 - 4 - 3 × 10°), as compared to semiconservatively synthesized DNA. To further confirm and repeat the initial observations, an additional treatment-labeUing experiment was performed with MMS treatment (5 raM) of WIL2-A3 cells and the kinetic reaction of the DNA with repair replication radioactivity performed from low to high C0t-values (6.1 × 1 0 - 4 - 7 . 3 × 103), which includes the reassociation range o f unique DNA sequences. As can be seer in Fig. 4, there appears,to be an increasing difference in reassociation of repair replication radioactivity with increasing Cot, the greatest difference (17%) being at the highest Cot (7.25 × 103).

83 <

O-

fo-

"-,-..... r-

Z LLI (.~ n-"

30-

40

J

l

IO - 3

10-4

I0- Z

J

I

I

IO- l

i0 o

IO I

Cot [(MOLE-NUCLEOTIDES/LITER)X SEC.] Fig. 3. Kinetic reassociation study performed in the low to intermediate C#t range with repair replicated DNA from UV-irradiated EB-3 cells exposed to 15 J/M ~. o ~ o , DNA repair replication labeled with [ 3H ]BrUdR; o-- -- --% DNA semiconservatively labeled with [ 14C]thymidine. Open symbols indicate average of replicate incubations; closed symbols indicate actual data points. O-

I0-

,~ Z o Q w

205O-

0 4Ou) if) w e:::

50-

t--Z ILl ¢.) n-" W n

60-

70-

80I

iO-4

10-3

I

10-2

|

10-~

i

10o

;O~

I JO 2

!

103

I JO 4

Cot [(MOLE-NUCLEOTIDES/LITER) X SEC.] Fig. 4. Kinetic reassociation study performed over the low to high Cot range with repair replicated DNA from 5 mM MMS treated WIL=-A3 cells, o o, DNA repair replication labeled with [ZH]BrUdR; o - - - - - - - o , DNA semiconservatively labeled with [ ' 4 C ] t h y m i d ine. Open symbols indicate average o f replicate incubations; closed symbols indicate actual data points.

84 0-

0

20Z C3 LU

30-

re

50-

L) L~ EL

70-

80-

,

"o

,b-.

,b

,;.,

,b'

,b

,b

Cot [(MOLE-NUCLEOTIDES/LITER)X SEC] Fig. 5. Kinetic reassociation study performed with either [SH ]thymidine labelled (~ .~ ) or ["C ]thyrnidine labelled (o-------o) semiconservatively synthesized DNA. The former was subjected to two alkali cesium chloride-cesium sulfate density gradient centrifugations. Open symbols indicate average of replicate incubations; closed symbols indicate actual data points, ti indicates that the data came from incubations with high concentrations of WIL~-A3 DNA, as compared to the points at similar C0t-values from incubations with low concentrations of WIL~-A3 DNA (overlap region).

In establishing the validity of this difference, we were concerned with several steps in the protocol which might generate an artifact. We first explored the effect of two alkali cesium chloride-cesium sulfate gradient centrifugations on the subsequent kinetic reaction of [3H]thymidine labened, semiconservatively synthesized DNA, as compared to [14C]thymidine labelled DNA which was also semiconservatively synthesized, but subjected only to one neutral cesium chloride gradient. The results of the kinetic experiment (Fig. 5) show differences in the two curves at individual C0t-values, but the dispersive trend obvious in the MMS experiments is not apparent. We were further concerned that treatment with 5 mM MMS for 3 h might result in extensive alkali or heat labile damage which could produce artifactual effects when the treated cell DNA was centrifuged in alkali gradients or incubated for long times at 60°C. Such an artifact might be differences in the 'piece sizes' resulting after MMS treatment. We therefore prelabeled DNA with [3H]thymidine by semiconservative synthesis (21-h incubation), and then treated the cells with 5 mM MMS for 3 h. The

85 O-

I0-

20Z C~ C2~ Ld 30-

0 U.I

40-

50-

t--z L~ 60n~ W O.

70-

\ 80-

I0 "4

I

10-3

I

I0 -2

I0-

I

I0 0

I

I0 r

1~)2

I0 31

I041

CO/[(MOLE-NUCLEOTIDES/LITER) X SEC] Fig. 6. Kinetic reassociation study performed with D N A o f WIL~-A3 cells which was semiconservatively labeled for 21 h with [°H]thymidine, f o l l o w e d b y a 3-h treatment with 5 m M MMS. D ~ , [°H]thymidine labelled and MMS-treated D N A ; o o, D N A semiconservatively labeled with [z4 C]thymidine. Open symbols indicate average o f

replicate incubations; closed symbols indicate actual data points, x indicates that the data came from incubations with high concentrations o f WIL~-A3 D N A , as compared to the points at similar C0t-values from incubations with low concentration o f WIL2-A3 D N A (overlap region).

DNA was then isolated, density gradient centrifuged twice in alkali cesium chloride-cesium sulfate and reacted according to the standard protocol. The results of this kinetic experiment (Fig. 6) show differences at individual C0t-values, but the dispersive trend obvious in the MMS experiments is again not apparent. DISCUSSION

Three of the approaches which have been employed to determine whether or not DNA damage and its subsequent repair in mammalian cells occur randomly in the genome include (a) the kinetic analysis of repetitious frequencies by DNA/DNA hybridization, (b) the separation of DNA by density gradient centrifugation into satellite and mainband DNA and (c) micrococcal nuclease digestion of chromatin in isolated nuclei. An early kinetic investigation of whether UV-induced damage and the

86

subsequent repair replication occurred randomly in human cancer cell DNA sequences of all repetitious frequencies or whether a particular class(es) of sequences was either more susceptible to damage and/or more proficient (or deficient) in its repair, indicated a uniform distribution of the UV-induced repair replication radioactivity [2]. A subsequently reported kinetic investigation [3] of UV-induced repair in normal human diploid fibroblasts (WI-38 cells) confirmed the HeLa cell observation [2] and found a similar uniform distribution for incorporated DNA repair radioactivity induced by two chemical carcinogens, NA-AAF and 7-bromomethylbenz [a]anthracene. The second approach (density gradient centrifugation to separate mouse cell satellite DNA from mainband DNA) was used in a study in which BALB/c mouse mammary tissue explants were treated with 3 mM MMS for 3 h or 1.5 and 3 mM N-methyl-N-nitrosourea for 1 h and then repair labeled [9]. A significantly lower incorporation of repair activity was found in the satellite DNA than in the mainband DNA. These observations must be contrasted with an earlier report [10] which indicated that after treatment of BALB/c 3T3 cells with UV light (200 ergs/mm 2) or 10/~M NA-AAF, repair synthesis occurred to the same extent in mainband and satellite DNA. The third approach which has been employed is the use of a micrococcal nuclease to digest chromatin. The initial action appears to be a preferential digestion of repaired regions in internucleosomal DNA [11--13]. After treatment of BALB/c mouse mammary fragments for 3 h with 3 mM MMS and 2 h of repair labelling, nuclei were isolated and then incubated with micrococcal nuclease. The results indicated that repair patches were located in a chromatin subfraction which is preferentially digested by micrococcal nuclease during the initial period of incubation [13,14]. When a similar procedure using staphylococcal nuclease was used to examine the fast repair component occurring during the first 2--3 h after UV exposure (12 J/M 2) of IMR-90 humau diploid fibroblasts, fast-repair synthesis appeared to favor nuclease~ensitive regions [14]. In the studies described herein using the DNA/DNA kinetic approach, the results suggest a relative increase in the appearance of repair replication radioactivity induced by treatment of human cells with MMS at high concentration (5 raM) in DNA sequences of lower repetitious frequency. This indicates that there could be proportionally more repair occurring in sequences of fewer copies. Such a result would be in contrast to those in 'normal' human lymphoblastoid cells, where we have observed a uniform distribution of incorporated repair replication radioactivity induced by UV. The latter observation was previously reported in HeLa cells [2]. While the Cot technique cannot provide information as to whether this is due to more extensive damage occurring in sequences of fewer numbers of copies, it is interesting to note that the results would be in agreement with the observations made of reduced DNA repair radioactivity in mouse satellite DNA after MMS treatment [9] (this fraction is enriched in highly repetitious DNA sequences [ 1] ). The extents of alkylation by MMS in mainband and satellite DNA were found to be similar in that study [9]. While the nuclease digestion studies indicate similarities in the nuclease

87 sensitivity of the early repaired region of both mouse and human cell DNA after UV and MMS damage, the results of the satellite and kinetic studies indicate that investigations in the future should give separate attention to UV-like damage repair and X-ray-like damage repair [15]. Several comments need to be made about the significance of the increasing dispersion observed between the reassociation of the MMS damaged and repair replicated DNA and semiconservatively synthesized DNA. Although the results of experiments to study the effects on hybridization of the alkali gradient purification procedure (Fig. 5) and of residual alkali labile or heat labile sites due to MMS treatment (Fig. 6) do not show dispersion, they do show differences and it cannot be entirely disproved that some combination of these phenomena is not responsible for the dispersion. The curves in Figs. 5 and 6 do appear, however, to be displaced, rather than increasingly dispersive in nature. An alternative explanation of the results could be that the repair sequences reassociate at the same rate as semiconservatively synthesized-DNA, but the reassoclated DNA is less stable during the subsequent fractionation of single-stranded DNA from double° stranded DNA. Final definition of this possibility would require studies indicating the stability of duplex DNA containing MMS induced repair sequences, as demonstrated for NA-AAF-induced repair sequences [3]. It is interesting to note that after NA-AAF treatment at high concentration, in contrast to the kinetic uniformity observed at lower concentrations, Lieberman and Poirier [3] did note evidence suggestive of a similar shift of repair incorporated radioactivity into less redundant DNA. Finally, if we assume that the dispersion is real, we must be concerned with any attempts to extrapolate to lower MMS concentrations, where such results would be more meaningful in terms of genetic alterations in mammalian cells. In the L5178Y mouse lymphoma cell thymidine kinase mutation assay system described by Clive and Spector [16], a 2-h treatment of those cells in 10% serum with 0.5 mM MMS resulted in a surviving cell population at 48 h after treatment of only 0.15%; the survival would have been less if the colony assay had been performed immediately after treatment (D. Clive, pets. comm.). We have found upon treating L5178Y cells with 5 mM MMS in medium with 10% serum and examining DNA single-strand breakage tising the alkaline elution technique [ 17 ], that massive single-strand breakage is present immediately after treatment (unpublished observation). When we treat these cells for 2 h with 1 mM MMS in medium with 3% serum, not only do we find extensive DNA damage, but the cell concentration of washed and reincubated cells continues to decrease over the 72-h interval post-treatment (unpublished observation). The treatment of 'normal' human lymphoblastoid cells with a high concentration of the mutagen and alkylating agent MMS provides evidence suggestive of proportionally more repair occurring in DNA sequences of fewer numbers of copies. Such a high concentration will cause both extensive cell death and DNA breakage in mammalian cells. Considering the small (albeit repeatable) differences in reassociation kinetics ultimately seen, it is unlikely that the technique employed for these studies would be sensitive

88

enough to confirm such alterations at lower and genetically more interesting concentrations of the mutagen. ACKNOWLEDGEMENTS

This work was supported by Department of Energy Contract No. EY-76-5-05-4761. W e express our appreciation to the Southwest Foundation for Research and Education for the use of their facilities. REFERENCES

1 R.J. Britten and D.E. Kohne, Repeated sequences in D N A , Science, 161 (1968) 529. 2 M.L. Meltz and R.B. Painter, Distribution of repair replication in the HeLa cell genome, Int. J. Radiat. Biol.,23 (1973) 637. 3 M.W. Lieberman and M.C. Poirer, Distribution of deoxyribonucleic acid repair synthesis among repetitive and unique sequences in the human diploid genome, Biochemistry, 13 (1974) 3018. 4 M.L. Meltz, N.J. Whittam and W.H. Thornburg, R a n d o m distribution of highly repetitive and intermediate frequency mouse L-929 cell D N A sequences synthesized after U V lightexposure, Photochem. Photobiol., 27 (1978) 545. 5 M~L. Meltz, D N A repair in baboon alveolar macrophages: a system for assessing biohazardous materials, Environ. Res., 11 (1976) 359. 6 H. Steier and J.E. Cleaver, Exposure chamber for quantitative ultraviolet photobiology, Lab. Pract., 18 (1969) 1295. 7 J.R. Gautschi, B.R. Young and R.B. Painter, Evidence for D N A repair replication in unirradiated mammalian cells- is it an artifact?Biochim. Biophys. Acta, 281 (1972) 324. 8 R.T. Hinegardner, A n improved fluorometric assay for D N A , Anal. Biochem., 39

(1971) 197. 9 W.J. Bodell and M.R. Banerjee, Reduced DNA repair in mouse satellite DNA after treatment with methyl-methanesulfonate, and N-methyl-N-nitrosourea, Nucl. Acids Res., 3 (1976) 1689. 10 M.W. Lieberman and M.C. Poirer, Intragenomal distribution of DNA repair synthesis: repair in satellite and mainband DNA in cultured mouse cells, Proc. Nat. Acad. Sci. U.S.A., 71 (1974) 2461. 11 W.J. Bodell, Nonuniform distribution of DNA repair in chromatin after treatment with methyl methanesulfonate, Nucl. Acids Res., 4 (1977) 2619. 12 M.J. Smerdon, T.D. Tisty and M.W. Lieberman, Distribution of ultraviolet-induced DNA repair synthesis in nuclease sensitive and resistant regions of human chromatin, Biochemistry, 17 (1977) 2377. 13 W.J. Bodell and M.R. Banerjee, The influence of chromatin structure on the distribution of DNA repair synthesis studied by nuclease digestion, Nucl. Acids Res., 6 (1979) 359. 14 M.J. Smerdon and M.W. Lieberman, Nucleosome rearrangement in human chromatin during UV-induced DNA repair synthesis, Proc. Natl. Acad. Sci. U.S.A. 75 (1978) 4238. 15 J.D. Ragan and R.B. Setlow, Two forms o f repair in the DNA of human cells damaged by chemical carcinogens and mutagens, Cancer Res., 34 (1974) 3318. 16 D. Clive and J.F.S. Spector, Laboratory procedure for assessing specific locus mutations of the TK locus in cultured L5178Y mouse l y m p h o m a ceJls, Mutat. Res., 31 (1975) 17. 17 M.L. Meltz and J.T. MacGregor, Activity o f the plant flavanol quercetin in the mouse l y m p h o m a L5178Y TK*/-mutation, DNA single-atrand break, and Balb/c 3T3 chemical transformation assays, Murat Res., 88 (1981) 317.