562

Biochimica et Biophysica Acta, 4 7 5 ( 1 9 7 7 ) 5 6 2 - - 5 7 0 © E l s e v i e r ] N o r t h - H o l l a n d B i o m e d i c a l Press

BBA 98900

INTERACTION OF C A F F E I N E WITH THE DNA OF CHINESE HAMSTER CELLS

F. T O N D E U R

* a n d J. R O M M E L A E R E

**

Laboratoire de Biophysique et Radiobiologie, Universitd Libre de Bruxelles, rue des Chevaux 67, B 1640 Rhode St Gen~se (Belgium) (Received A u g u s t 3rd, 1976)

Summary Incubation of Chinese hamster cells with labelled caffeine leads to transfer of radioactivity to DNA. This association occurs during the S phase of the cell cycle and involves .parental as well as newly synthesised strands. The replacement of thymidine by BrdUrd prevents the incorporation radioactivity from caffeine into the DNA strands containing BrdUrd. Thymine is the only base which becomes labelled and data suggesting the participation of methyl groups of caffeine in the biosynthesis of thymine are presented. Ultraviolet irradiation increases the incorporation of radioactivity from caffeine to DNA.

Introduction Caffeine (1,3,7-trimethylxanthine) is known for its toxic, mutagenic and radiosensitising action [ 1 ]. Several authors have studied in vitro the interaction of caffeine with DNA: T'so et al. [2] have shown that caffeine binds to DNA by h y d r o p h o b i c bonds and by intercalation between the purine and pyrimidine bases; T'so and Lu [3] found that the affinity constant caffeine for DNA is much higher for denatured than for native DNA. D o m o n et al. [4] observed that binding of caffeine to DNA increases when DNA is partially denatured by ultraviolet light. By analogy to its action on DNA in vitro, caffeine could act in vivo preferen* Boursier I.R.S.I.A. ** Aspirant Fonds National de la Recherche Scientifique. Abbreviations: Thy, thymine; dThy, thymidine; Urd, uridine; BrdUrd, bromodeoxyuridine; FldUrd, fluorodeoxyuridine.

563 tially at m o m e n t s of the cell cycle when D N A i s locally denatured, as during its semi-conservative replication [5--7]. In fact, Lehman [8] observed that in mouse L cells, the DNA fibres newly synthesized in the presence of caffeine are shorter than normal and suggests that the c o m p o u n d slows down the progression of the replicative forks. On the other hand, Wragg et al. [9] demonstrated that caffeine inhibits the activity of DNA polymerase in mammalian cell extracts. Cleaver and Thomas [10] have observed in ultraviolet-irradiated Chinese hamster cells that caffeine in addition inhibits postreplicative closing of gaps formed in the newly synthesised DNA strands opposite nonexcised pyrimidine dimers. Buhl and Regan [11] also noticed that in presence of caffeine, DNA replication is delayed in ultraviolet-irradiated cells from Xeroderma pigmentosum patients. There might be increased binding of caffeine in primary or secondary single-stranded regions induced in the DNA of irradiated cells. The first purpose of this work is to test whether caffeine binding sites exist in the DNA of cultured Chinese hamster cells, whether they appear during the S phase, and whether their appearance can be induced by ultraviolet light. Cornisch and Christman [12] have shown that caffeine is catabolized in humans to various methylated derivatives to xanthine and to derivatives of uric acid. Our next purpose is to analyse the hypothetical binding of caffeine on DNA to check whether caffeine could possibly be a m e t h y l d o n o r for the purines in providing methyl groups to enter the 1-carbon pool or for the pyrimidines, either for the synthesis of dTMP from dUMP or for the synthesis of 5-methylcytosine from cytosine. In this last case methylation of cytosine would occur after DNA replication through the action of methylases; in mammalian cells there is a lag of approximately 2 min between DNA synthesis and the beginning of methylation [13, 14]. Materials and Methods

Cell cultures. Chinese hamster V79 cells are cultures in monolayer in F12 medium [15]. The absence of mycoplasma (PPLO) is checked by 3 methods [16--181. The cells become contact inhibited in Go [19] after 50 h when seeded at a cell density of l 0 s cells/cm 2. A partial synchronisation is observed after release from this Go block. The mitotic index and [3H]thymidine (dThy) uptake detected by autoradiography [20] as a function of time are represented in Fig. 1. A maximum of + 80% of cells in S phase occurs approximately 8 h after Go release (Fig. 1). Extraction o f DNA and CsC1 density gradient centrifugation. The cells are rinsed 3 times with cold Eagle medium and trypsinised. After 3 extra "rinses", the pellet is resuspended in 10 -2 M Tris • HC1, 10 -3 M EDTA, pH 7.5, and the DNA is extracted by successive treatment with SDS (1%, 20 min at 4°C with shaking), ribonuclease (50 pg/ml, 90 min, 37°C), pronase (500 pg/ml, 90 min, 37°C). The lysate is then layered on top of a CsC1 gradient (9 ml; final density, 1.7 g/cm3). After centrifugation (35 000 rev./min, 48 h, 15°C) in a Beckman 40 rotor, the absorbances are monitored at 260 nm (A260) and fractions are collected in an ISCO fractionator.

564 100

&

ocye60 ~ 0~

20 0

o".o"o I

2

i

I

I

6

~

I0

I

I

I

I/*

Time

(hours}

Fig. 1. S y n c h r o n i s a t i o n of cells a f t e r G O release. A t v a r i o u s t i m e s a f t e r s u b c u l t u r e , the cells arc p u l s e d ( 3 0 m i n ) w i t h [ 3 H ] d T h y ( 2 5 p C i / m l ) . A f t e r f i x a t i o n ( e t h a n o l / a c e t i c acid, 3 : 1, 20 rain, 4 ° C ) t h e y are c o v e r e d w i t h n f o r d L 4 e m u l s i o n , e x p o s e d f o r 12 d a y s , d e v e l o p e d a n d s t a i n e d w i t h U n n a (20 rain). • • , p e r c e n t a g e of labelled cells; o . . . . . . G p e r c e n t a g e o f cells in mitosis.

For alkaline CsC1 gradients, 0.025 ml of 0.1 M KOH are added to 9 ml of CsC1 solution {final density, 1.750 g/ml). In neutral gradients, the b u o y a n t density observed for native light DNA was 1.70 g/ml. In alkaline conditions, light strands and strands substituted with b r o m o d e o x y u r i d i n e (BrdUrd) banded at densities of approx. 1.75 g/ml and 1.85 g/ml, respectively. The high degree (>90%) of replacement of dThd sites by BrdUrd in the heavy strand was expected since BrdUrd incorporation occurred in the presence of fluorodeoxyuridine (FldUrd), an inhibitor of thymidylate synthesis. Radioactivity determination. The radioactivity of the fraction is determined in a liquid scintillation counter after addition to scintillation liquid in Triton X100 [21]. Deoxyribonuclease treatment. The fractions from the CsC1 gradients are dialyzed overnight in 0.1 M Tris, 3 . 1 0 - 3 M magnesium acetate, pH 7.5 and then treated with Worthington bovine pancreatic deoxyribonuclease (40 pg/ml) for 1 h at 37°C. Chromatography. Fractions containing DNA from the neutral gradient are dialyzed for 24 h at 4°C against Tris buffer 10-2M pH 7.5 and precipitated with perchloric acid [22]. The DNA is hydrolyzed in formic acid [23] and chromatographed in two dimensions on thin cellulose layers (Polygram cell MN 300) first with butanol/acetic acid/H20 (80 : 12 : 30, v/v) then with methanol/ HCl/H20 (70 : 20 : 10, v/v). The plates are subdivided in 16 squares, each of which is scraped and eluted in 0.1 M HC1 for 24 h at 37°C. Spectra of the eluates are analysed in a Carry 14 spectrophotometer; radioactivity is determined after addition of toluene/Triton X100 scintillation solution in a Intertechnique counter [ 21]. Ultraviolet irradiation. Monolayers cells are " d r y " irradiated in Petri dishes with a low energy ultraviolet source giving its maximum o u t p u t at 253.7 nm; doses are determined with a Latajet dosimeter [24]. Radioisotopes and chemicals. From The Amersham Radiochemical Centre were the following obtained: [3H] Caffeine, 1 Ci/m, generally labelled in 1,3 and 7 methyl groups and the 8-H on the xanthine ring; [3H]thymidine, 5 Ci/mmol; [14C]thymidine, 59 Ci/mol; ['4C]bromodeoxyuridine, 57 Ci/mol; [2-'4C]uridine, 62 Ci/mol. Fluorodeoxyuridine was a gift of Hoffman-Laroche

565 and aminopterine and hypoxanthine were obtained from K & K Lab and Calbiochem, respectively. Results

Asynchronous cells The DNA of synchronous cells grown in [3H]caffeine (10 pCi/ml) for 36 h is analysed by density gradient centrifugation. Fig. 2 shows that radioactivity originally bound to caffeine is found superimposed on the DNA peak, although it is obvious from the counts spread through the gradient that a large fraction of the total counts are not in bound form. When isolated, treated with DNAase and recentrifuged, both peaks disappear jointly. As a control, [3H]caffeine was added in vitro to a lysate of untreated cells, and in this case no caffeine radioactivity binds preferentially to DNA (Fig. 3). DNA-caffeine association in S phase cells (a) Cells are blocked by contact inhibition (Go) and labelled for 30 min with Jail]caffeine (10 pCi/ml) and [14C]dTdhd (3 pCi/ml). A second population is similarly treated after being preincubated with [all]caffeine for 9 h after release of contact inhibition (the majority of cells are then in S phase). DNA is extracted after the pulse and the radioactivity distribution found is shown in Figs. 4a and b. The radioactivity coming from caffeine and thymidine is associated with the DNA only in the preincubated cells (Fig. 4b); the peaks of aH and ~4C radioactivities are displaced towards absorbance values higher than the major absorbance band; this displacement is not found when thymidine is pulsed in the absence of caffeine. This suggest either that caffeine is specifically associated with replicating DNA and maintains it temporarily in a more denatured configuration [25], or that DNA of greater density is being synthesised in the presence of caffeine. (b) Knowing that during S phase the radioactivity of caffeine associates with DNA, it was asked whether the drug interacts preferentially with the parental

6OO ._c E

,oo

0.6

1 o

.< 21111 "r "

0.2

~ 0.4