Evaluation of caged compounds for use in studies of u.v.-induced DNA repair. Rosalind A. ... Since the nucleotide has 32P in the a- position the break is trapped ...
Biochem. J. (1990) 266, 891-895 (Printed in Great Britain)
891
Kinetics and mechanism of DNA repair Evaluation of caged compounds for use in studies of u.v.-induced DNA repair Rosalind A. MELDRUM,* Sydney SHALLt and Christopher W. WHARTON*$ * Department of Biochemistry, University of Birmingham, P.O. Box 363, Birmingham B15 2TT, U.K., and tCell and Molecular Biology Laboratory, School of Biological Sciences, University of Sussex, Brighton BNI 9QG, U.K.
Experiments are described in which the feasibility of using caged dideoxy and other nucleoside triphosphate analogues for trapping breaks induced by u.v. radiation damage to mammalian cell DNA is evaluated. These nucleotide analogues that have a photolabile 1-(2-nitrophenyl)ethyl-protecting group attached to the yphosphate are placed in situ by permeabilizing cells by exposure to hypo-osmotic medium. The nucleoside triphosphate is released from the cage by a 351 nm u.v. laser pulse whence it may incorporate in the growing chain of DNA induced by the excision-repair process and terminate chain elongation. If the photoreleased dideoxynucleoside triphosphate is isotopically labelled in the a-phosphate position the break is trapped and labelled. Incorporation of radioactivity into trichloroacetic acid insoluble material in these experiments confirms their potential for use in studies of the kinetics of mammalian cell DNA repair. INTRODUCTION Several methods exist that measure single-strand breaks in DNA: alkaline elution (Fornace et al., 1976), nucleoid sedimentation (Cook & Brazell, 1975), sedimentation in alkaline sucrose gradients (McGrath & Williams, 1966), alkaline denaturation-hydroxyapatite chromatography (Ahnstrom & Evardsson, 1974) and nitrocellulose filter binding (Center & Richardson, 1970). All these methods suffer notable disadvantages and are limited in accuracy and sensitivity. Methods that involve sedimentation in sucrose gradients may introduce a small number of breaks initially during cell lysis and the very early stages of centrifugation as well as being complicated by the possibility of sedimentation of gel complexes of molecules. Nucleoid sedimentation may measure not necessarily breaks but changes in chromatin structure in the form of relaxation of coiling of DNA. Alkaline elution and hydroxapatite chromatography involve technical manipulations that could introduce hydrodynamic shearing of DNA and are time-consuming procedures. Nitrocellulose membrane filtration will not trap very small pieces of DNA (i.e. 100 bases) and so may underestimate the number of breaks. The number of breaks required for these methods to detect repair is higher than desirable. Alkaline elution, detecting a minimum of 300 breaks per cell (calculated from Kohn et al., 1981), has been considered the most sensitive, but more recently less than 100 breaks per cell appear to have been detected by the hydroxyapatite chromatography method (Squires & Johnson, 1988). Estimates of initial velocities derived from progress curves have been used to determine apparent Michaelis kinetic parameters of DNA-repair synthesis by the use of inhibitors of DNA-replication synthesis (Squires et al., 1982). However, the inhibition of replication synthesis, which probably yields an underestimate of breaks, may
lead to alterations of the nucleotide substrate concentrations normally available when repair synthesis is carried out. None of the above methods can measure repair rates and kinetics in the early stages of repair. In the present paper we describe a novel method that will enable estimation of repair kinetics to be made in time courses of seconds or less by trapping and measuring singlestrand breaks in DNA of mammalian cells. A 32P-labelled dideoxynucleoside triphosphate with a photolabile protecting group is placed in situ by permeabilizing cells. The caged dideoxynucleoside triphosphate may then be photolysed at chosen times during the process of repair in DNA following damage by various agents. When the caged dideoxynucleoside triphosphate is photolysed by a laser pulse of 351 nm u.v. light the photolysed nucleoside triphosphate is incorporated in the growing chain of DNA, which is catalysed probably by polymerase /8 or a in repair synthesis. The lack of the hydroxy group on the dideoxy form leads to termination of chain elongation. Since the nucleotide has 32P in the aposition the break is trapped and labelled. Chain termination by dideoxynucleotide as used in vitro by the Sanger sequencing methods (Sanger et al., 1977) forms the basis of the break trapping method. Photolysed caged ATP has been used to good effect by Trentham and co-workers in the study of kinetics of muscle-cell contractile mechanisms (McCray et al., 1980). The experiments described here demonstrate the feasibility of the method. MATERIALS AND METHODS Cell culture AA8 Chinese-hamster fibroblasts and PY3T3 Swissmouse transformed fibroblasts were grown in monolayers in 30 mm plastic Petri dishes (Nunc) in Dulbecco's
Abbreviations used: ddATP, dideoxyadenosine triphosphate; ddTTP, dideoxyribosylthymine triphosphate; ddT, dideoxyribosylthymine; araCTP, arabinosylcytosine triphosphate; araC, arabinosylcytosine. $ To whom correspondence should be addressed.
Vol. 266
R. A.
892
medium (Gibco) supplemented with 10 % (v/v) foetalcalf serum (Gibco). The cells were initially seeded at 2 x 104 cells/ml and subcultured every 2-3 days. One 30 mm dish contains 106 cells at confluence. HL60 suspension cells from a human promyelocytic-leukaemia cell line were maintained at 2 x 105-4 x 105/ml on RPMI medium (Gibco) containing 10 % (v/v) foetal-calf serum in plastic culture flasks (Nunc). HL60 cells were seeded at 106 cells per dish into Petri dishes for irradiations. Cell permeabilization Cells were permeabilized by the hypo-osmotic-shock method of Reinhard et al. (1977). Cells were washed three times with ice-cold Puck's saline (containing 5.4 mM-MgCl2, 137 mM-NaCl, 4.2 mM-NaHCO3 and 5.5 mM-glucose), then covered with 1.0 ml of hypoosmotic buffer [9 mM-Hepes buffer, pH 7.8, containing 5.0 mM-dithiothreitol, 1.0 mM-EGTA, 4.5 % (w/v) dextran (average Mr 110000) and 4.5 mM-MgCl2]. The dishes were kept at 4 °C until the cells were permeabilized. Permeabilization was monitored by Trypan Blue dye inclusion. When 90-100 % of the nuclei were seen to take up the stain the buffer was made iso-osmotic by addition of 200 1l of high-salt solution [195 mM-Hepes buffer, pH 7.8, containing 780 mM-KCl, 7.0 mM-EGTA, 1.5 % (w/v) dextran and 1.35 mM-sucrose]. Nucleotides were added in the following concentrations: 120 ,sM-CTP, -UTP and -GTP, and 250 ,M-dTTP, -dATP, -dCTP and -dGTP; 10 mM-phosphoenolpyruvate was also added. The appropriate deoxynucleoside triphosphate was omitted when the caged form of the dideoxynucleoside triphosphate was added in order to avoid competitive incorporation. Laser irradiation of cells Petri dishes (35 mm diameter) containing a monolayer of about 106 cells were subjected to 351 nm u.v. irradiation from a XeF-eximer laser (Lamda Physik EMG 150 amplifier section) to photolyse the caged compound. The radiation was delivered as 10 ns pulses at 10 Hz. To induce damage in DNA, cells were subjected to 248 nm u.v. irradiation from a laser (Lamda Physik EMG 150 oscillator section) in the form of 10 ns pulses. [For optics and energy measurements see the preceding paper (Meldrum et al. (1990)).] The growth medium was removed from the cells before irradiation in order to avoid absorption of u.v. light by proteins, and the cells were irradiated in Puck's saline. Fresh growth medium was re-applied after irradiation for incubation of cells for 2 h or more. Precipitation and filter extraction of cellular DNA Reactions were quenched by adding 5 M-NaOH to cell suspensions to give a final concentration of 0.5 M-NaOH. Trichloroacetic acid (20 %, w/v) was added to give a final concentration of 10 % (w/v) to precipitate the acidinsoluble fraction. The precipitate was then collected by filtration on glass-fibre filters (Whatman GC/F) for scintillation counting. RESULTS Effects of 351 nm radiation on cellular DNA synthesis It was first necessary to establish that 351 nm u.v. laser irradiation did not damage cells in such a way as would render the technique invalid. AA8 Chinese-hamster ovary
Meldrum, S. Shall and C. W. Wharton
700 a) c
.F
(b) 248 nm
600
c
500 _
en
j.'Ei-.
400
00 Ea 300 C
-.,
0
T,
200
0 C
100
0
I
I
0.1 1 10 100 Dose (kJ/m2)
I
I
10 100 1 0.1 Dose (J/m2)
Fig. 1. Incorporation of [Me-3Hlthymidine during 24 h after irradiation of AA8 cells (106 cells per dish) with (a) 351 nm u.v. light and (b) with 248 unm u.v. light For experimental details see the Materials and methods and the Results sections.
fibroblasts were subjected to doses of 0-110 kJ/m2 351 nm u.v. irradiation in the form of 10 ns 10 Hz pulses of approx. 25 mJ per pulse. In order to provide a comparison, cells were subjected to 248 nm u.v. irradiation, which is known to cause considerable damage by the induction of pyrmidine dimer formation (Setlow & Carrier, 1964; Boyce & Howard-Flanders, 1964). Cells were irradiated with doses of 0-10 J/cm2 in the form of 10 ns pulses (1-40) of 248 nm u.v. laser irradiation, then incubated at 37 °C in fresh growth medium for 24 h with 1.0 #uCi of [Me-3H]thymidine/ml to measure DNA synthesis. Fig. 1(a) shows the incorporation of [Me-3H]thymidine during 24 h after 351 nm and 248 nm irradiation. There is some enhancement of DNA synthesis in 24 h after 351 nm radiation doses of 0.1 and 1.1 kJ/m2. DNA synthesis in 24 h after 351 nm radiation doses higher than 11 kJ/m2 is diminished, showing that these levels of irradiation are damaging to the cells. The dose used for optimal photoactivation of caged ddATP is approx. 1 kJ/m2 (see Meldrum et al., 1990). A similar effect was seen on the incorporation of [Me-3H]thymidine during 2 h after irradiation with 351 nm u.v. light (results not shown). By using equilibrium sedimentation of bromodeoxyuridine-substituted DNA (Smith & Hanawalt, 1976) the enhanced [Me-3H]thymidine uptake has been shown to be in replicated DNA rather than in repairsynthesized DNA (R. A. Meldrum, unpublished work). Since repair synthesis is not affected and the effect of the dose required for photolysis of the caged reagent is not large, it does not appear that this will have any significant effect on the measurement of repair-synthesis kinetics. Fig. l(b) shows by comparison the severe damaging effect of 248 nm-wavelength u.v. irradiation. Chain termination by dideoxyribosylthymine (ddT) during repair of u.v.-induced damage The action of ddT as a terminator of DNA chain elongation in repair synthesis was tested in AA8 Chinesehamster ovary fibroblasts and HL60 human leukaemia cells. Arabinosylcytosine (araC) (10 CM) and hydroxyurea (10 mM) were applied to the cells in full medium for 1990
DNA repair: evaluation of caged dideoxynucleoside triphosphates
30 min before irradiation to inhibit replicative DNA synthesis. ddT was applied to the unpermeabilized cells. ddT is incorporated into DNA by polymerase /1, which has been proposed to be principally involved in repair synthesis (Bertazzoni et al., 1976; Hubscher et al., 1979), although it is also incorporated by polymerase a (Dresler & Kimbro, 1987). In cells where replication synthesis was inhibited by araC and hydroxyurea, incorporation of [Me-3]thymidine into DNA repair synthesis, induced by 248 nm u.v. damage, was inhbited by ddT (Fig. 2). Experiments designed to examine the effect of ddT on replication DNA synthesis seem to reveal a different involvement of DNA polymerases a and ,? and/or a in different cell types. ddT diminished the incorporation of [Me-3H]thymidine into DNA in HL60 leukaemia cells but did not decrease the incorporation into DNA replication synthesis in AA8 fibroblasts (Table 1). AraC greatly diminished the incorporation of [Me-3H]thymidine into DNA in HL60 cells, and araC with ddT almost wholly inhibited incorporation. AraC inhibits primarily the activity of polymerase oc. These results suggest that polymerase ,? or polymerase a (Dresler & Kimbro, 1987) as well as polymerase a is involved in HL60-cell replication synthesis, but only polymerase a is involved in AA8-fibroblast replication synthesis.
a) _ 3
72 E X
+
-2 i fi
~
~~~I
-
0 0>i
0C_
l +.+
x
ddT 248 nm (20 J/m2) Hydroxyurea/araC
+
-
+
-
+
_
-
+
+
+
+
+
0;
Fig. 2. Inhibition by 1 mM-ddT of incorporation of IMe3Hjthymidine into AA8 cells Replication synthesis was arrested by incubation of the cells with hydoxyurea (10 mM) and araC (10 /,M) for 30 min before irradiation.
893
Permeabilization of cells Permeabilization can be problematical and is well known to produce variable and inconsistent results. We found that the transformed PY3T3 fibroblasts were well permeabilized (95-100 %) in hypo-osmotic solution in 5-10 minutes. HL60 cells, a human promyelocyticleukaemia cell line, were also permeabilized in a similar period of time in hypo-osmotic solution. AA8 and nontransformed 3T3 fibroblasts were less easily permeabilized by hypo-osmotic buffer. The time course of hypo-osmotic permeabilization, although suitable for the experiments that test the feasibility of the method, is too slow to enable measurements of repair kinetics to be made. It was confirmed by the nucleoid sedimentation technique (Cook & Brazell, 1975) that permeabilization by hypo-osmotic solution did not introduce breaks into DNA. Incorporation of photolysed caged lx-32PlddATP The main feature of the method was shown to work when significant enhancement of 32p uptake occurred when PY3T3 cells were subject to 248 nm followed by 351 nm laser irradiation after permeabilization and loading with the caged reagent. For this experiment [a-32P]dATP was caged (see Meldrum et al., 1990), as [a_-32P]ddTTP is not yet commercially available. In this instance the reagent was 70 % pure with respect to the caged form. Fig. 3 shows the results obtained when cells loaded with radiolabelled reagent were subjected to 248 nm and/or 351 nm irradiation. Uptake in unirradiated permeabilized cells would be due to incorporation of the contaminating uncaged reagent by DNA replication synthesis. Uptake of [a-32P]ddATP was increased by about 40 % in cells that had been damaged by 248 nm u.v. irradiation. When caged reagent was photoactivated by 351 nm irradiation, uptake due to DNA replication synthesis increased by the expected amount. This was 50 % higher than the uptake when the caged reagent was not photoactivated. When repair synthesis was promoted by 248 nm u.v. irradiation and optimum uptake of [a-32P]ddATP was allowed by photoactivation of the caged compound by 351 nm irradiation, the 32p incorporation increased 4fold over that of undamaged cells in which the caged reagent was not activated and 3-fold over that of damaged cells with unactivated caged reagent. A 1 J dose of 351 nm irradiation was applied to the cells. It is calculated that approx. 50 % of caged compound is released at this dose of u.v. irradiation [see the preceding paper (Meldrum et al., 1990) for details of photoactivation].
Table 1. Effects of different inhibitors on the incorporation of IMe-3Hlthymidine (1 jCi/ml) into AA8 cells and HL60 cells For experimental details see the text. Abbreviation: N.D., not determined.
I-' x Incorporation of [Me-3H]thymidine (d.p.m./min per 106 cells)
Vol. 266
Cell type
No inhibitor
AraC (10 fM)
ddT (1 mM)
AraC + ddT
AA8 HL60
14.6+0.68 18.5+3.01
4.5 +0.03 0.6+0.21
14.0 +0.23 3.7+ 1.93
N.D. 0.24+0.03
R. A. Meldrum, S. Shall and C. W. Wharton
894
Io( __r-
22 a1) '.-
20
C.)
0o c c
o
18
o cl
X -6 10 _-
a)
0)
X16
V = 2 )2 en O
>
o 0. 0
_-
cV x
14
or.
r Ic q)
Q-s gfa
_
z
0 .o
.° (L 1 0
_
351 nm
0
_-
12 _610
#.
Fig. 4. Inhibition of IMe-3HIdTTP incorporation into AA8 cells (A) or HL60 cells (@) by photolysed caged araCTP Caged araCTP (1 mM) was added to permeabilization buffer with nucleotides. The araCTP was left inactive or activated by a 1 kJ/m2 dose of 351 nm u.v. laser irradiation.
N
C Xn
x
eS
6
w-
4
ii
2_ QL
248 nm 351 nm
+
_+ _
_
_+
~+
+
Fig. 3. Incorporation of photolysed caged Ioc-32PIddATP PY3T3 cells (10 per dish) were irradiated with 248 nm u.v. light (20 J/m2) and/or 351 nm u.v. light (I kJ/M2) in the presence of 1 /Ci of caged [a-32P]ddATP
The enhancement of uptake (2-3-fold) in damaged cells approximately corresponds to the proportional increase of [a-32P]ddATP made available by photoactivation. The increase in uptake by DNA replication synthesis in undamaged cells with maximally activated reagent present, compared with undamaged cells with unactivated agent present, is not as proportionally great as the additional [32P]ddATP made available for incorporation by photoactivation. In an experiment performed on PY3T3 cells in which DNA replication synthesis was arrested as a consequence of confluence of the cells, a 10fold enhancement of [32P]ddATP uptake was seen upon 248 nm irradiation (results not shown). This indicates that, when 100 % caged material is photoactivated in cells in which DNA replication synthesis is minimal, at least a 10-fold enhancement of uptake should be observed upon stimulation of repair synthesis. An enhancement of at least 10-fold incorporation produced by stimulation of DNA repair synthesis will allow good resolution in kinetic studies, provided that the sampling error can be reduced. Inhibition of IMe-3HIdTTP uptake by photolysed caged arabinosylcytosine triphosphate (araCTP) That caged araCTP could be used to study the involvement of polymerase a was demonstrated by the observation that photolysed caged araCTP caused 50 % inhibition of [Me-3H]dTTP uptake in permeabilized cells. AA8 fibroblasts and HL60 cells were permeabilized, and caged araCTP was introduced into the cells with [Me3H]dTTP, along with nucleotides in iso-osmotic medium. During incubation for 1 h at 37 °C incorporation of [Me-
3H]dTTP was diminished in both cell types in samples where the caged araCTP was activated by 351 nm irradiation compared with samples that contained intact caged araCTP (Fig. 4). The presence of inactivated caged nucleoside triphosphate, albeit caged ddATP, did not affect the viability of cells as measured by Trypan Blue dye exclusion or uptake of [Me-3H]thymidine into DNA synthesis. DISCUSSION The experiments described here confirm that a radiolabelled photoactivated caged dideoxynucleoside triphosphate will trap and label breaks that result from u.v. radiation damage to DNA. Since the molecule can be introduced into cells and may be held in an inactive form until photoactivated by a laser pulse, the numbers of DNA breaks present in the cell genome may potentially be measured in very short time courses over the early part of the repair process. This will make possible studies of detailed kinetics of repair mechanisms that have been refractory to investigation by existing methods of measuring repair activity, owing to insensitivity and poor time resolution. In some preliminary experiments we have used electropermeabilization to introduce molecules into mammalian cells, since the time course of this procedure is very short and more advantageous to the potentially very fast action of our break-trapping method. It also appears to be less variable between cell types and to permit resealing of cells (Chu et al., 1987). A general scheme of this procedure is shown in Fig. 5. When permeabilization procedures are standardized for all cell types, our technique should provide a very sensitive and rapid method for studying the kinetics of mammalian cell DNA repair with minimal disruption of the cells' natural mechanistic response to damage. Time-course analysis, made by varying the time gaps between firing of the lasers, will enable deductions to be made concerning the strategy by which DNA is scanned for damage. This methodology will allow consideration whether DNA is scanned for damage by random or processive processes, as discussed by Collins & Squires (1986). Results of experiments using techniques that used inhibitors and measurements made over an interval as 1990
DNA repair: evaluation of caged dideoxynucleoside triphosphates
895 XeF 351 nm laser pulse (1 J/10 cm2)
Re-seal cells
Quench (trichloroacetic acid)
I
Washing and work-up for scintillation counting on filters
KrF 248 nm laser pulse
Permeabilize cells
( < 40 J/m2)
*
Load cells with caged 32P-labelled reagent (1 pCi)
At1 -
N o. of
ms-* s
breaks after At1 /-~alculated from 32p incorporation *4
0
I
At2
-
*
s
~1m
Time -
I
At1 is varied (ms-*s) to give time course of incision in DNA (rate-limiting step); At2 is optimized for each time range
Fig. 5. General experimental scheme for the study of the rapid-phase kinetics of the repair of radiation-induced breaks in mammalian cell DNA At1 is varied (ms -+ s) to give the time course of incision in DNA (rate-limiting step in u.v.-radiation-induced repair). At2 is optimized for each time range.
large as 30 min suggested that the process is random (Collins & Squires, 1986). Measurements made by using the method described here will be more detailed over a shorter time interval and will not be subject to interference of the action of inhibitors. A lag phase might, for instance, indicate a rate-limiting process before incision that may represent topoisomerase involvement. We have also shown indirectly that photolysed caged araCTP will inhibit polymerase oc activity, since [Me3H]dTTP uptake into DNA synthesis was diminished by the activated araCTP molecule. Therefore, assuming that araCTP is stably incorporated into the DNA chain over the time course of the experiment, caged 32P-labelled araCTP could be used to derive kinetic parameters for polymerase a similar to those described above for polymerase , or 6. Previously initial incision rates have been estimated by back-extrapolation to time zero from measurements of break accumulation at times at least 30 min or more after damage. Substrate depletion and enzyme inhibition may occur over these short times. This will lead to underestimates of the initial rates, and hence the shorter the time period of measurement the less the underestimate.
Ahnstrom, G. & Edvarsson, K. A. (1974) Int. J. Radiat. Biol. 26, 493-497
Bertazzoni, U., Stefanini, M., Pedrali Noy, G., Giulotto, E., Nuzzo, F., Falaschi, A. & Spadari, S. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 785-789 Boyce, R. P. & Howard-Flanders, P. (1964) Proc. Natl. Acad. Sci. U.S.A. 51, 293-300 Center, M. S. & Richardson, C. C. (1970) J. Biol. Chem. 245, 6285-6291 Chu, G., Hayakawa, H. & Berg, P. (1987) Nucleic Acids Res. 15, 1311-1326 Collins, A. & Squires, S. (1986) Mutat. Res. 166, 113-119 Cook, P. R. & Brazell, I. A. (1975) J. Cell Sci. 19, 261-279 Dresler, S. L. & Kimbro, S. K. (1987) Biochemistry 26, 26642668 Fornace, A. J., Kohn, K. W. & Kann, H. E. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 39-43 Hiibscher, U., Kuenzle, C. C. & Spadari, S. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 2316-2320 Kohn, K. W., Ewig, R. A. G., Erickson, L. C. & Zwelling, L. A. (1981) in DNA Repair: A Laboratory Manual of Research Procedures (Friedberg, E. C. & Hanawalt, P. C., eds.), vol. 1, part B, pp. 379-401, Marcel Dekker, New York and Basel McCray, J. A., Herbette, L., Kihara, T. & Trentham, D. R. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 7237-7241 McGrath, R. A. & Williams, R. W. (1966) Nature (London) 212, 534-535 Meldrum, R. A., Shall, S., Trentham, D. R. & Wharton, C. W. (1990) Biochem. J. 266, 885-890 Reinhard, P., Burkhalter, M. & Gautschi, J. R. (1977) Biochim. Biophys. Acta 474, 500-511 Sanger, F., Nicklen, S. & Coulsen, A. R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5463-5467 Setlow, R. B. & Carrier, W. L. (1964) Proc. Natl. Acad. Sci. U.S.A. 51, 226-231 Smith, C. & Hanawalt, P. C. (1976) Biochim. Biophys. Acta 432, 336-347 Squires, S. & Johnson, R. T. (1988) Mutat. Res. 193, 181-
Received 21 June 1989/3 October 1989; accepted 12 October 1989
Squires, S., Johnson, R. T. & Collins, A. R. S. (1982) Mutat. Res. 95, 389-404
This work was supported by the S.E.R.C. Laser experiments were conducted at the Laser Support Facility of the Rutherford-Appleton Laboratory, Didcot, Oxon., U.K. We
thank the M.R.C. Radiobiology Unit, Harwell, Berks., U.K., for use of their cell-culture facilities during experiments at the Rutherford-Appleton Laboratory.
REFERENCES
Vol. 266
192
