311

Mutation Research, 33 (1975) 311--320

© Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands

REPAIR BY HUMAN CELLS OF ADENOVIRUS-2 DAMAGED BY PSORALEN PLUS NEAR U L T R A V I O L E T LIGHT TREATMENT

RUFUS S. DAY, III, ALFRED S. GIUFFRIDA and C. WESLEY DINGMAN Chemistry Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20014 (U.S.A.)

(Received April 4th, 1975) (Accepted July 28th, 1975}

Summary Adenovirus-2 is damaged by t r e a t m e n t with psoralen plus near-ultraviolet (UV) light as shown by reduced ability to infect human fibroblasts. The apparent sensitivity of the virus to this t r e a t m e n t depended upon the strain of cells used. The virus was 3--4 times more sensitive to the t r e a t m e n t when infecting xeroderma pigmentosum (XP complementation groups A or D) fibroblasts than when infecting normal fibroblasts. DNA extracted from virus preparations that had undergone such t r e a t m e n t was analyzed for treatment-induced crosslinks by gel electrophoresis and sedimentation in alkaline sucrose gradients. The fraction of adenovirus DNA molecules remaining non-crosslinked after treatment was f o u n d to correlate with the survival of the virus in normal fibroblasts. This result showed t h a t the psoralen plus near-UV treatment gave rise to non-crosslink lesions (presumably psoralen-DNA mono-adducts), that were repairable by normal but not by XP fibroblasts, and suggested the possibility that normal fibroblasts cannot repair this type of crosslink in the DNA of an infecting adenovirion.

Introduction Interstrand crosslinks in DNA (of the type DNA strand 1-psoralen-DNA strand 2) are made when DNA solutions containing the furocoumarin psoralen are irradiated with near-ultraviolet (UV) light (300--380 nm) [4,6]. Such irradiation also produces psoralen mono-addition products to DNA (of the type DNA strand-psoralen) in higher yields than crosslinks [2]. Cole [5] has reported t h a t about 65 psoralen plus near-UV produced crosslinks are required to reduce the colony forming ability of wild-type E s c h e r i c h i a coli to 1/e (1 hit) whereas only 16, 7, and 1.3 crosslinks are required to pro-

312 duce the same effect in E. coli uvrA, recA, and uvrA recA mutants, respectively. However, when phage ~ is treated with psoralen plus near-UV, the 1/e survival occurs at approx. 1 crosslink/genome when infecting wild-type E. coli or its uvrA mutant. This finding indicated the possibility that E. coli enzymes, although they may repair crosslinks in their own DNA, cannot repair similar damage in the DNA of an infecting phage [5]. Using psoralen plus near-UV treatment and other crosslinking agents, Sasaki and Tonomura [20] have shown that chromosome alterations in lymphocytes from patients having Fanconi's anemia are produced more frequently than in cells from normal donors. This suggests the possibility that normal human cells can repair psoralen plus near-UV produced crosslinks but that cells from patients with Fanconi's anemia are deficient in this process. The data to be reported here suggest the possibility that psoralen plus near-UV produced crosslinks in adenovirus 2 DNA are n o t repaired by host cell enzymes. This interpretation is based upon the fact that the plot of the fraction of uncrosslinked DNA molecules parallels the survival curve of psoralen plus near-UV treated adenovirus when infecting normal or Fanconi's anemia fibroblasts. However , the slope of the survival curve of similarly treated adenovirus was 3--4-fold greater when using XP fibroblasts. This difference could be due to the inability of XP fibroblasts to repair mono-addition products in psoralen plus near-UV treated adenovirus. XP fibroblasts have previously been shown to have a high sensitivity to psoralen plus near-UV induced damage to their own DNA [1]. Methods and materials Virus preparation and assay Adenovirus-2 obtained through Dr. James Rose, NIH, was plaque purified on human fibroblast monolayers,~ and grown out on HeLa cells to obtain high titer stocks (~ 3 X 107 plaque forming units (PFU) per ml as titered on human fibroblast monolayers) [8]. [3H] thymidine labelled virus was prepared as follows: 30 plates of confluent HeLa cells (grown in Dulbecco's modification of Eagle's minimal essential medium containing 10% fetal calf serum) were infected with 1 ml of 1/50 dilution of stock virus in basal medium Eagle (BME) containing 1% fetal calf serum, and 50 units penicillin plus 50 pg streptomycin per ml. After a 4-h adsorption period, the liquid was removed from the plates and replaced with 10 ml of the following serum-less medium: to 500 ml BME were added 5 ml each of 100 X BME amino acids, 100 X BME vitamins, 7.2 X 10 -3 M thymidine (unlabelled), 2.2 X 10 -2 M deoxycytidine and a solution of 5000 units penicillin plus 5000 pg streptomycin per ml followed by 2.5 ml of 0.06 M arginine, 10 ml of a solution containing 21.3 g BES, 35.7 g HEPES, 25.2 g HEPES in 200 ml of water, pH 7.6 (Eagle's buffer for general use [14] ), 0.5 ml of 0.5 X 10 -3 M fluorodeoxyuridine, and 3.0 ml of [3H] thymidine (22.4 Ci/mM; 1 mCi/ml). Seven days later the detached infected cells were collected by centrifugation, suspended in 7 ml Tris buffer {0.01 M Tris, pH 8.0} and subjected to 3 cycles of freeze-thawing (dry ice--ethanol and 37°C). Virus was purified through two CsC1 equilibrium density gradients according to the procedure of

313 Green and Pina [16] except t h a t CsC1 was used instead of RbC1. The virus band (p = 1.34), dialyzed against 3 changes of Tris buffer and brought to a volume of 1 ml with Tris buffer, contained about 3.8 X 10 s cpm. Subsequent titration on h u m a n fibroblast monolayers showed that the preparation contained about 4 X 108 PFU on h u m a n fibroblasts. The preparation contained about 800 pg of DNA by spectrophotometric measurement of the isolated DNA (see below). The DNA had a specific activity of 1260 dpm/pg, or 0.03 distintegrations per year per double stranded DNA molecule of 25 X 106 MW, so that DNA degradation due to radioactive decay played little role in the results presented here.

Human fibro blasts The origins of the normal fibroblast strains KD and ND have been published [8,9] as have those of XP strains XP12BE, XP6BE, and X P I L O (see also ref [1] ), these latter having previously been designated 1223, 1157 and 1201 respectively [8,9]. The CRL 1196 fibroblast strain from a Fanconi's anemia patient was obtained from the American Type Culture Collection (Rockville, Md.). Plaque assay All plaque assays were done as described [8] except that the second agar overlay was 10 instead of 5 ml, and the third 5 ml overlay was omitted. Psoralen plus near-UV treatment In separate experiments non-labelled and labelled virus suspensions were irradiated with near-UV light both in the presence and absence of psoralen. Non-labelled stock virus was diluted 1 : 100 into PBS either presaturated with psoralen (Trimethylpsoralen, a kind gift of Paul B. Elder Co., Bryan, Ohio, 10 mg/100 ml PBS) at 46°C overnight or into PBS without psoralen. In the case of labelled virus, 0.9 ml purified virus suspension was diluted with 9.0 ml psoralen-saturated Tris buffer (10 mg/100 ml buffer), and 0.1 ml virus suspension was diluted with 0.9 ml Tris buffer with no psoralen as a control. Nonlabelled virus was irradiated in 2 ml quantities in 60 mm (Falcon 3006) culture dishes (with tops on), labelled virus in 1 ml quantities in 35 mm (Falcon 3005) culture dishes (tops on). (The culture dish tops have an absorbance (against air) of 0.13 at 360 nm, 0.16 at 340, 0.44 at 300, 1.8 at 290, and > 3 at 280, 270 and 260 nm and so while these tops pas s near-UV well they are efficient absorbers of any short wavelength UV given off by the near-UV light source). Irradiations were done using a G.E. F15T8 black light bulb at an incident dose rate of about 6W/m 2 (measured with a Blak-Ray ultraviolet meter, model J-221 (Ultraviolet Products, San Gabriel, Calif.). The samples were agitated using an A.H. Thomas swirler and kept at room temperature during irradiation. All virus suspensions prepared for irradiation had absorbances of ~ 0.005 per mm at 340 nm. Isolation o f DNA and heat denaturation An aliquot (0.25 ml) of each irradiated or non-irradiated 3H-labelled virus preparation (psoralen plus near-UV treated and controls) was mixed with an equal volume of 0.01 M Na3EDTA, 0.01 M Tris--HC1 (pH 8.0) containing 2% sodium dodecyl sulfate and 200 pg/ml Proteinase K (EM Laboratories, Inc.)

314 and incubated for 30 min at 50°C. After incubation the mix was brought to r o o m temperature and treated with 0.5 ml of 90% phenol by gentle mixing for 10 min. After chilling to 0°C to facilitate the separation of the aqueous and phenol layers, the aqueous phase was dialyzed overnight at 4°C against 0.05 M NaC1, 2 mM Na3EDTA, pH 7.0. This isolation of adenovirus-2 DNA is a modification of that described by Trilling and Axelrod [ 2 3 ] . The use of a protease was found to be absolutely necessary, however, if good yields of linear DNA were desired. Without protease treatment most of the DNA obtained showed very little electrophoretic mobility in our gels (and probably existed as circles [19] ). We found proteinase K superior to CaC12-preserved, heat-treated pronase [10] in terms of minimizing evidence of single-strand nicking during the extraction of DNA. Electron microscopy of the isolated DNA revealed that the majority of the molecules were linear double-strands having lengths of about 12 pm; no effect of the psoralen plus near-UV treatment was observable by this technique (T. Kakefuda, personal communication). Heat denaturation of the DNA samples was accomplished by heating at 100°C for 5 min and quenching in ice immediately prior to electrophoretic analysis.

Gel electrophoresis In all experiments reported here, gel electrophoresis of samples containing 1.5 to 3.0 pg DNA was done at 20°C, using a voltage gradient of 8.8 V/cm in 6 mm thick slab gels composed of 1.7% acrylamide and 0.5% agarose [15,17]. After staining [17] gels were cut and counted as described previously [12]. Zone sedimentation in alkaline sucrose gradients Zone sedimentation in alkaline sucrose gradients was performed using a 11.5 ml gradient of 4% to 20% sucrose and 0.1 N to 0.3 N NaOH containing 0.9 M NaC1 and 5 mM Na3 EDTA in polyallomer tubes. Centrifugation of samples containing 4 to 6 pg DNA was performed at 5°C and 35,000 rpm in the SW40 Spinco rotor and 0.4 ml fractions were collected directly into liquid scintillation vials containing 1.0 ml water. After the addition of 10 ml of Hydromix {Yorktown Research) and mixing, the samples were counted as previously described [ 1 2 ] . Sedimentatiorl coefficients ($20 w) were determined as described [11] and molecular weights calculated using the formula [21] : $20 w = 0.0528 M °'4°° Results and discussion

Inactivation of adenovirus-2 plaque forming ability by psoralen plus near-UV light treatment Initially we tested the abilities of fibroblasts cultured from normal persons and from xeroderma patients to repair psoralen plus near-UV treated adenovirus. The results of t w o experiments are shown in Fig. 1. The X P I L O strain belongs to XP DNA repair complementation group A [18] .the XP4BE strain is from an XP "variant" [3] (having 60--70% of the host cell reactivation of UV (254 nm) irradiated adenovirus-2 shown by normal cells [9] ). ND is a normal strain [8].

315

O FL-

i

i

i

i

i

i

1.

< L.9

-13

0.1

2 0

z

O.Ol

z

_o