The DNA of UV-irradiated normal and excision-deficient mammalian cells undergoes relaxation in an initial stage of DNA repair

123

Mutation Research, 165 (1986) 123-128 DNA Repair Reports Elsevier

MTR06138

The DNA of UV-irradiated normal and excision-deficient mammalian cells undergoes relaxation in an initial stage of DNA repair David L. Mitchell, Judith M. Clarkson and Gerald M. Adair University of Texas System Cancer Center, Science Park -- Research Division, Smithville, TX 78957 (u.S.A.)

(Received2 April 1985) (Revisionreceived10 September1985) (Accepted 17 September1985)

Summary Using a radioimmunoassay specific for Pyr(6-4)Pyo photoproducts, we have demonstrated the removal of these lesions from denaturated DNA isolated from UV-irradiated Chinese hamster ovary cells at various times post irradiation. When assayed undenatured, these same DNA samples, which are initially 10-20 times less capable of binding antibody, show a substantial increase in binding capacity during the first few hours of repair. At 3 h post irradiation the difference between native and heat-denatured DNA samples is negligible, indicating that all of the residual lesions are contained in a single-stranded (relaxed) configuration. This relaxation also occurs in UV-hypersensitive cell lines, that are deficient in the ability to remove Pyr(6-4)Pyo photoproducts. Novobiocin, an inhibitor of topoisomerase II, prevents both the initial increase in binding and the subsequent excision of the antibody-binding sites.

In order to relate DNA damage and repair to cell survival and mutation induction, detailed analyses of the molecular events responsible for the removal of different types of damage are required. Recently we have developed a very sensitive radioimmunoassay (RIA) that has allowed us to study the excision of specific photoproducts (Mitchell et al., 1985a, b). The assay, which uses a polyclonal antiserum raised in rabbits against UV-irradiated DNA (Mitchell and Clarkson, 1981), measures the capacity of DNA samples to inhibit antibody binding to a labelled antigen. When the labelled antigen consists of poly(dA) : poly(dT) that has been irradiated at 254 nm (40 k j/m2), the assay is specific for Pyr(6-4)Pyo photoproducts. Typically, the DNA isolated from mammalian cells is denatured prior to assay and a rapid loss of Pyr(6-4)Pyo photoproducts from UV-irradiated repair-competent cell lines is observed (Clarkson

et al., 1983; Mitchell et al., 1985b). In this paper we describe the results obtained when the native DNA is assayed. These data have led us to conclude that assaying samples under these conditions reveals an initial relaxation step in the repair of DNA and that this is not the defective step in UV-hypersensitive Chinese hamster ovary (CHO) cells. Materials and methods Preparation o f sample D N A

Chinese hamster ovary (CHO) cells were maintained in monolayer cultures in a-modified MEM medium (Gibco) with 10% foetal calf serum. Cells were plated 2 days prior to the experiment in medium containing 0.05 /xCi/ml [14C]thymidine. Irradiation was carried out with two 15-W General Electric germicidal lamps emitting predominantly

0167-8817/86/$03.30 © 1986 ElsevierSciencePublishersB.V.(BiomedicalDivision)

124 254-nm light at a fluence of 0.5 j / m 2 / s e c . Following irradiation the cells were either harvested immediately (for a standard curve) or incubated for various repair periods. The DNA was isolated by lysing the cell pellets with SDS (0.01% final concentration) followed by pronase digestion (50 /~g/ml) for 45 min at 37°C in 4 ml TBS (10 mM Tris, pH 7.5, 150 mM NaC1). EDTA was added to a final concentration of 10 mM as indicated in the text. The mixture was then sheared with a VirTis '45' homogeniser (yielding DNA of approx. 6 x 10 6 dalton) and the protein extracted with an equal volume of chloroform/isoamyl alcohol (24: 1). Following ethanol precipitation, the DNA was redissolved in TBS at a concentration of 50-100 /~g/ml.

Radioimmunoassay The details of the assay have been described previously (Mitchell and Clarkson, 1981; Mitchell et al., 1982). Briefly, antiserum was raised in rabbits by injection of UV-irradiated (approx. 200 kJ/mZ), denatured calf thymus DNA coupled to methylated bovine serum albumin. In the assay, sample DNA competes with 32p-labelled antigen for antibody-binding sites in a reaction mixture containing 0.15% gelatin (Sigma) and 5 ~ g / m l herring sperm DNA. Poly(dA):poly(dT) (Boehringer-Mannheim), labelled to 5 x 108 c p m / ~ g with thymidine 5'[a-32p]triphosphate by nick translation (Maniatis et al., 1976) received 40 k J / m 2 254-nm UV light. Antiserum was added at a concentration that yielded 30-60% binding to labelled antigen (3 x 10-5). Following incubation with the antiserum for 1-2 h at 37°C, the immune complex was precipitated with goat anti-rabbit immunoglobulin (Calbiochem) and carrier 3'globulin. The pellet was dissolved in NCS solubiliser (Amersham), mixed with Scintiverse (Fisher) and the 32p quantified on a Packard liquid scintillation counter with the discriminators set such that 14C counts were excluded. Chromatography of sample DNA on Sephacryl S-300 Sephacryl S-300 (Superfine) (Pharmacia) was packed in a 1-cm column to 30 cm and equilibrated with TBS at a flow rate of 0.35 ml/min. The void volume was determined with blue dextran and the molecular weight exclusion limit with

Alu 1-restricted q~X174. DNA eluting in the void volume was heat-denatured and, after adding bromophenol blue and glycerol, electrophoresed in 2% agarose (BRL) with Tris-acetate buffer (40 mM Tris, pH 7.8, 5 mM NaOAc, 1 mM EDTA, 0.5 /~g/ml ethidium bromide). An exclusion limit of 119-109 bases was determined by loss of most of these bands and all smaller DNAs from the void volume, in agreement with published reports (Sofer et al., 1984). Sample DNA, prelabelled to a specific activity of about 1000 cpm//~g 14C, was isolated from cells harvested at 0 or 45 min post irradiation. About 50000 cpm in 0.5 ml were loaded onto the column and 1-ml fractions collected. An aliquot was counted to determine the elution profile of the DNA. The remainder was used as competitive inhibitor in a RIA. Results

We have recently demonstrated that a polyclonal antiserum can be used to specifically assay the repair of Pyr(6-4)Pyo photoproducts in mammalian cell DNA (Mitchell et al., 1985a, b). Using this assay system, the relative affinity of antibodybinding sites in double- versus single-stranded DNA from repairing cells has been determined (Fig. 1). Half of each sample was assayed as native DNA while the other half was heat-denatured prior to assay. Dose-response curves generated at the time of irradiation (0 h) (inset, Fig. 1) indicate that heat-denatured UV-DNA has about 10 times the binding capacity of native UV-DNA. By extrapolating the percent inhibition determined for each repair time from both sets of data through the dose-response curve for heat-denatured UVDNA, the relative dose equivalents can be calculated. As previously published (Mitchell et al., 1985b), when heat-denatured UV-DNA is used for these assays 75% of the Pyr(6-4)Pyo-specific binding sites are lost within 3 h. When native DNA is used as a competitive inhibitor, approximately 10fold less binding of antibody is observed initially, but the relative binding capacity of the DNA increases dramatically during the first few hours post irradiation. There is no difference in antibody binding to Pyr(6-4)Pyo photoproducts in heat-denatured versus native DNA by 2-3 h. Repair curves for Pyr(6-4)Pyo photoproducts

125 10'-~

I-'\ A

1o07

~:8

'o

,o1 ~:6=,

A -| , ,~ 10 I" \\ ooze ,J~-,

~

assays, it is apparent that b o t h mutants are deficient in the repair of P y r ( 6 - 4 ) P y o photoproducts. N o decrease in antibody-binding sites is detectable in UVL-10, while UVL-1 appears to remove about 40% of these sites by 24 h post irradiation. W h e n native D N A samples are used for these assays, increased a n t i b o d y binding is observed as a function of time post irradiation for all 3 cell lines. At 24 h post irradiation UVL-10 cells, which are unable to excise any P y r ( 6 - 4 ) P y o photoproducts, show antibody-binding levels in native D N A that are comparable to those in heat-denatured sampies. Since the only change in D N A k n o w n to increase antibody binding is the transition from a double- to single-stranded configuration (Mitchell and Clarkson, 1981; Ley, 1983; Mitchell et al., 1985a), the dramatic increase in relative binding that we observe during the first 2 h post irradiation m a y reflect localised unwinding of doublestranded D N A at the d a m a g e d sites prior to excision. If so, this increase in binding m a y be inhibited by novobiocin, a k n o w n inhibitor of eucaryotic topoisomerase II (Liu et al., 1980). We have previously shown by r a d i o i m m u n o a s s a y of heat-denatured D N A that high concentrations of novobiocin inhibit the removal of antibody-binding sites from

,"

~

\

REPAIR TIME (h)

Fig. 1. The repair of Pyr(6-4)Pyo photoproducts in D N A assayed single- and double-stranded. CHO cells, prelabelled with [14C]thymidine, were irradiated with 10 J/m 2 254-nm UV light and incubated at 37°C for various lengths of time. The DNA was isolated and 10/Lg of native ( ) or heat-denatured (. . . . . . ) sample were competed against 10 pg 32p-labelled poly(dA): poly(dT) (5000 cpm) which had been irradiated with 40 kJ/m 2 254-nm light. The antiserum was diluted to a final concentration of 3 × 10 -5, yielding 30-50% binding of the labelled antigen, The number of antibody-binding sites (dose equivalents) was calculated for each repair time from the dose-response curve for denatured DNA (inset). in two UV-hypersensitive m u t a n t cell lines derived f r o m C H O - A T 3 - 2 are shown in Fig. 2. W h e n heat-denatured D N A samples are used for these

0

~0~0

\

o

0

.,.

o

//.

A,3-2

lb

2b

J

e

UVL- 1

1()

2'0

~

UVL- 10

1'0

2'0

REPAIR TIME (h)

Fig. 2. The repair of Pyr(6-4)Pyo photoproducts in UV-hypersensitive cell lines. CHO-AT3-2 cells and two UV-hypersensitive cell lines derived from them were tested for their ability to repair Pyr(6-4)Pyo photoproducts as in Fig, 1. The isolated DNA was assayed native ( ~ ) or denatured (. . . . . . ) and the dose equivalents calculated from a dose-response curve for denatured AT3-2 DNA.

126 o o

lO

o.

\

E

3

~

....

\o

~'~ 8

40 A

A 10"

"'\O

o"~'~8'~

~

~o

20

0 "\. 0

6¸

~O

N

6'

10-

\ 0

0" 4ILl

[ i,~ . .....

I

e__s

I 70-

o CI 2

• 0

2

4

6 o

2

I

~4

6

A 1601 A ~,-j • ° v c~

z 120-

~

.

.

50

E 40-

40

60 0

z

30-

30

I

20

//

2010-

0

,../

',!

1 10 ......... ;o FRACTION NUMBER

0

?

10 |

•

0

Fig. 5. Size distribution of DNA containing Pyr(6-4)Pyo-binding sites fractionated on Sephacryl S-300. CHO cells, prelabelled to a specific activity of 1000 cpm 14C/#g and irradiated with 20 J / m 2, were harvested at 0 (panel A) and 45 (panel B) rain post irradiation. Following DNA isolation (in the absence of EDTA), about 50000 cpm 14C were loaded and chromatographed in TBS at a flow rate of 0.3 ml/min. 1-ml fractions were collected and the 14C profile of the eluting DNA determined (. . . . . . ). 0.2-ml aliquots from each fraction (2000 cpm in the peak fraction) were competed against UV-irradiated poly(dA) : poly(dT) as in Fig. 1.

12" i

> m ~ 80-

8-

ILl if) 0 ~ 40-

4-

0

• B

16"

.

_o 50-

I--

i r r a d i a t e d cells ( C l a r k s o n a n d Mitchell, 1983). I n Fig. 3 d a t a are s h o w n for D N A i s o l a t e d f r o m cells i n c u b a t e d i n the p r e s e n c e of 0.3 ( p a n e l A) or 3 ( p a n e l B) m M n o v o b i o c i n a n d a s s a y e d single- or 20-~{

60 E

z

Fig. 3. Inhibition of Pyr(6-4)Pyo photoproduct repair by novobiocin in DNA assayed single- and double-stranded. CHO cells were prelabelled, irradiated as in Fig. 1 and incubated for various lengths of time in media containing 0.3 (panel A) or 3 (panel B) mM novobiocin. Isolated DNA was assayed native ( ) or denatured (. . . . . . ) as in Fig. 1 and dose equivalents calculated from a dose-response curve for denatured DNA.

200-

70

60-

•

Repair Time (h)

0

t4 °

ti,

30-

8-

"\

=3

\.g

b,

0

REPAIR TIME (h) Fig. 4: Effect of DNA isolation conditions on the binding of double-stranded DNA from repairing cells. CHO cells, prelabelled with [14C]thymidine, were irradiated with 20 J / m 2 UV light and incubated for various repb.ir times. The DNA was purified as in Fig. 1 ( ) or was isolated in buffer containing l0 mM EDTA (. . . . . . ). These samples, together with a dose-response curve, were assayed native (panel A) or denatured (panel B) as in Fig. 1.

d o u b l e - s t r a n d e d . A t the lower c o n c e n t r a t i o n o n l y 70% of the l e s i o n s are r e m o v e d at 6 h a n d the i n c r e a s e d b i n d i n g of the n a t i v e s a m p l e s is r e d u c e d b o t h in rate a n d m a g n i t u d e . A t the h i g h e r c o n c e n t r a t i o n , however, a l m o s t total i n h i b i t i o n of b o t h the e x c i s i o n a n d the r e l a x a t i o n is observed. W h e n the D N A f r o m the r e p a i r s a m p l e s is i s o l a t e d i n the p r e s e n c e of 10 m M E D T A (Fig. 4) a n d a s s a y e d d o u b l e - s t r a n d e d , the 6 - f o l d i n c r e a s e i n b i n d i n g a s s o c i a t e d with early r e p a i r is signific a n t l y i n h i b i t e d . D N A a s s a y e d after heat den a t u r a t i o n shows n o s i g n i f i c a n t d i f f e r e n c e in the k i n e t i c s of loss of a n t i b o d y - b i n d i n g sites with or w i t h o u t E D T A i n the i s o l a t i o n buffer. T h u s , the i n c r e a s e i n b i n d i n g o b s e r v e d early in r e p a i r s a m -

127

~

the total binding is associated with DNA fragments undetected as 14C counts. Fragments from Alu 1-restricted ~X174 were used as molecular weight markers to demonstrate an exclusion limit of 119 bases. Hence, a substantial portion of the increased binding observed in DNA from cells incubated for 0.5-2 h post irradiation and isolated in the absence of EDTA is associated with small molecular weight DNA.

UVL

±EDTA

~

45' repair

DNA Isolation

Discussion +EDTA

Low affinity binding site

(double-stranded) DNA isolation

-EDTA

High affinity binding site (single-stranded)

Fig. 6. Model for the productionof high-affinitybindingsites in DNA fromcells repairingUV photodamage. pies assayed double-stranded is manifested by isolating the DNA in the absence of EDTA, a nuclease inhibitor. In Fig. 5 the size distribution of the DNA associated with antibody-binding sites is illustrated for double-stranded DNA isolated from cells 0 and 45 min post irradiation, The samples were loaded on Sephacryl S-300 equilibrated in TBS and 1 ml fractions collected. The amount of 14C in each fraction was determined and equal volumes from each fraction used as competitive inhibitor in a Pyr(6-4)Pyo-specific RIA. In panel A it can be seen that the binding capacity (% inhibition) of DNA from cells harvested at the time of irradiation closely resembles the profile of the DNA itself (i.e. the UV-irradiated DNA and its associated binding sites elute together in the void volume). When the DNA from cells which have been allowed to repair for a short time (45 min) is chromatographed, a quite different profile for binding activity is observed (panel B). Whereas most of the DNA elutes in the void volume, as in panel A, the binding capacity is seen to increase both in the void volume and in the DNA retarded by the column. In fact, a significant proportion of

Several reports have indicated that repair begins with the relaxation of supercoiled DNA. This process has been visualised by nucleoid sedimentation in neutral sucrose gradients and indicates that relaxation of the DNA is completed within the first hour post irradiation, returning to its original conformation within 4 h (Cook and Brazell, 1976; Mattern et al., 1982). This relaxation of supercoiled nucleoids following UV irradiation can be inhibited by novobiocin (Mattern et al., 1982), a drug known to inhibit a type II eucaryotic DNA topoisomerase (Liu et al., 1980). The increase in antibody binding that we observe in undenatured DNA samples as a function of repair time has similar kinetics (Fig. 1). It is interesting that this effect can be inhibited by the DNA repair inhibitor novobiocin (Fig. 3), yet is evidenced to different degrees by UV-hypersensitive cell lines (Fig. 2). Since denaturation of the sample DNA is the only conformational change known to us that increases the binding capacity of the sample DNA, we suggest that the observed increase in binding as a function of repair time in native DNA samples may result from the localised unwinding of double-stranded DNA in response to damage by UV irradiation. The inability of the UV-hypersensitive CHO cells to completely remove Pyr(6-4)Pyo photoproducts does not result from the inhibition of this early stage in DNA repair. It is reasonable to assume that the relaxation of supercoiling and presumed unwinding of the damaged site is dependent upon DNA-protein interactions. It is very unlikely that the integrity of such a structure would be maintained after deproteinisation and organic extraction during the isolation procedure. In Fig. 6, a model is described in which high-affinity binding sites produced early in

128

repair are manifested by nicking and/or partial digestion of the relaxed DNA by nucleases active during the isolation procedure. One way to test this model is to isolate the sample DNA in the presence of EDTA, an inhibitor of Mg2÷-dependent nucleases. Under these conditions the increased binding to native repaired DNA is not seen (Fig. 4). Hence, the expression of the observed increase appears to depend on conformational changes that predispose the damaged DNA to nuclease activity. Additional data supporting this model are presented in Fig. 5 and demonstrate that the enhanced binding is indeed associated with small molecular weight DNA. It appears that the increase in binding capacity evident 1-3 h post irradiation is a consequence of two phenomena: (1) the repair-dependent relaxation of the DNA at the site of the lesion and (2) nuclease digestion of the DNA during the isolation procedure. The same structural alterations that allow damage-associated DNA to be attacked by nucleases during the isolation procedure may provide access to the damage by repair enzymes and facilitate excision. However, while relaxation may be a prerequisite for the initiation of excision, data from UV-hypersensitive mutants illustrate that it is not determinative for successful repair.

Acknowledgements We wish to thank Dr. J. Ross for helpful suggestions throughout the course of this work and Carrie Haipek for excellent technical assistance. This work was supported by NIH grant CA 19281.

References Clarkson, J.M., and D.L. Mitchell (1983) The effect of various inhibitors of D N A synthesis on the repair of DNA photoproducts, Biochim. Biophys. Acta, 740, 355-361 Clarkson, J.M., D.L Mitchell and G.M. Adair (1983) The use of an immunological probe to measure the kinetics of DNA repair in normal and UV-sensitive mammalian cell lines, Mutation Res., 112, 287-299. Cook, P.R., and I.A. Brazell (1976) Detection and repair of single-strand breaks in nuclear DNA, Nature (London), 263, 679-681. Ley, R.D. (1983) Immunological detection of two types of cyclobutane pyrimidine dimers in DNA, Cancer Res.. 43, 41-45. Liu, L.F., C.C. Liu and B.M. Alberts (1980) Type II DNA topoisomerases: Enzymes that can unknot a topologically knotted DNA molecule via a reversible double-strand break, Cell, 19, 697-707. Maniatis, T., S.G. Kee, A. Efstratiadis and F.C. Kafatos (1976) Amplification and characterisation of a fl-globin gene synthesised in vitro, Cell, 8, 163-182. Mattern, M.R., R.F. Paone and R.S. Day (1982) Eukaryotic DNA repair is blocked at different steps by inhibitors of D N A topoisomerases and of DNA polymerases a and r , Biochim. Biophys. Acta, 697, 6-13. Mitchell, D.L., and J.M. Clarkson (1981) The development of a radioimmunoassay for the detection of photoproducts in mammalian cell DNA, Biochim. Biophys. Acta, 655, 54-60. Mitchell, D.L., R.S. Nairn, J.A. Alvillar and J.M. Clarkson (1982) Loss of thymine dimers from mammalian cell DNA: The kinetics for antibody-binding sites is not the same as that for T4 endonuclease V sites, Biochim. Biophys. Acta, 697, 270-277. Mitchell, D.L., C.A. Haipek and J.M. Clarkson (1985a) Further characterisation of a polyclonal antiserum for D N A photoproducts: the use of different labelled antigens to control .its specificity, Mutation, Res., 146, 129-133. Mitchell, D.L., C.A. Haipek and J.M. Clarkson (1985b) (6-4) photoproducts are removed from the DNA of UV-irradiated mammalian cells more efficiently than cyclobutane pyrimidine dimers, Mutation Res., 143, 109-112. Sofer, G., C. Seitz and M. Lasky (1984) Exclusion limits of DNA restriction fragments on gel filtration media, Am. Biotech. Lab., 2, 38-40.