Mapping of the varicella zoster virus deoxypyrimidine kinase gene and preliminary identification of its transcript

The varicella-zoster virus (VZV) deoxypyrimidine kinase (dPK) gene was mapped by transfection of cloned viral DNA fragments into thymidine kinase-deficient mouse L (LTK-) cells and subsequent biochemical transformation of these cells to the LTK+ phenotype. Such transforming activity was limited to the BamHI-H and EcoRI-D fragments of the VZV genome, which overlap by 2.2 kb between map units 0.50 and 0.52. Biochemically transformed cells were shown to contain a high copy number of viral DNA sequences that had integrated into the cellular DNA. Extracts of these cells showed a higher level of dPK activity than did extracts of parental LTK- cells. With the use of Northern hybridization analysis of transformed and VZV-infected cell RNAs, it was possible to tentatively assign a 1.8-kb transcript to the VZV dPK.


INTRODUCTION
Three human herpes viruses direct the synthesis of deoxypyrimidine kinases (dPK) during their replicative cycles. The enzymes encoded by herpes simplex virus types 1 and 2 (HSV-1, HSV-2) and varicella-zoster virus (VZV) have been characterized extensively. The initial mapping of the HSV thymidine kinase (TK) genes and much of our understanding of them were facilitated by the biochemical transformation to LTK+ of mouse L cells deficient in thymidine kinase (LTK-) by inactivated virions or purified HSV DNA sequences (Munyon et al., 1971;Bacchetti and Graham, 1977;Wigler et al., 1977Wigler et al., , 1978Maitland and McDougall, 1977;Pellicer et ab, 1978;McDougall et al, 1980;Reyes et al., 1982). Through a judicious selection of sequentially smaller viral DNA fragments for such transformation experiments, it was possible to localize the HSV-1 and 1 Author to whom requests for reprints should be addressed at: Medical Virology Section, Laboratory of Clinical Investigation, NIAID, NIH, Building 10, Room llN-113, Bethesda, Md. 20892.
HSV-2 TK genes with some precision. DNA sequence analysis has demonstrated that each of the genes encoding the HSV-1 and HSV-2 TKs consists of an open reading frame of 1128 nucleotides capable of encoding a protein of 376 amino acids (MeKnight, 1980;Wagner et a& 1981;Swain and Galloway, 1983). Three polypeptide products of the HSV-1 gene with molecular weights (MW) of 38,000,39,000, and 43,000 have been identified by in vitro and in vivo pulse-chase experiments (Marsden et al., 1983).
VZV is also known to induce an enzyme with dPK activity in infected cells (Dobersen et al., 1976;Ogino et al., 1977;Hackstadt and Mallavia, 1978;Cheng et aZ., 1979), and cell-associated VZV has been used to biochemically transform LTK-cells (Yamanishi et al, 1981). Studies of dPK activity in VZV-infected cells suggest that this activity is associated with a 35,000 MW polypeptide that may exist as a 70,000 MW protein composed of nearly identical subunits (Lopetegui et al, 1983;Shiraki et al, 1985).
In this report, we describe the stable transfer of VZV dPK activity to LTK-cells by transfection of cloned viral DNA re-striction fragments. This process has allowed the mapping of the VZV dPK gene, the preliminary enzymatic characterization of its product, and the tentative identification of its primary transcript. MATERIALS AND METHODS Cell culture. The LTK-cells (a gift from M. Rao and R. Klausner) were propagated in Eagle's minimum essential medium (EMEM) (Quality Biological, Inc., Gaithersburg, Md.) supplemented with 10% fetal calf serum, penicillin G at 100 units/ml, streptomycin at 100 pg/ml, and bromodeoxyuridine (BUdR) (Sigma Biochemicals, St. Louis, MO.) at 20 pg/ml. Prior to transfection, the cells were passed at least twice in the absence of BUdR. LTK+ transformants were selected and grown in EMEM supplemented with serum, antibiotics, and HAT (hypoxanthine, 15 pg/ml; aminopterin, 1 pg/ml; and thymidine, 5 pg/ml) (Littlefield, 1963). Vero cells (African green monkey kidney cells) were grown in a mixture of EMEM and medium 199 (1:l) (Quality Biological, Inc.) also supplemented with 10% fetal calf serum and antibiotics. Viral and cellular DNA purijcations. EcoRI and BamHI digests of whole VZV DNA were cloned into XgtWES * XB as previously described (Straus et al., 1982). The digests were further subcloned into pBR 322, pBR 325, and PUC8 and then grown in LE392, HBlOl, and JM83. Plasmid DNAs were prepared by the method of Birnboim and Doly (1979) and further purified with two isopycnic bandings in cesium chloride followed by ethanol precipitation.
A DNA probe prepared for use in Southern and Northern blot analyses consisted of the 2.2kb overlapping fragment between BamHI-H and EcoRI-D. Twenty micrograms of BamHI-H DNA in pBR 322 was digested with BamHI and EcoRI, and the bands were resolved on a 0.8% agarose gel. The 2.2-kb fragment was electro-eluted from the gel and purified through a Nacs-Prepac column [Bethesda Research Laboratories (BRL), Gaithersburg, Md.]. Cellular DNA was prepared from frozen cell pellets by overnight digestion with proteinase K (200 pg/ml) (Sigma) in the presence of 1% sodium dodecyl sulfate, 50 mM ethylenediamine tetraacetic acid (EDTA), and 10 mM Tris-HCl (pH 7.4) at 37". This mixture was extracted twice with phenol and twice with chloroform; it was then precipitated overnight at -20" in 0.4 M sodium acetate with 70% ethanol. The pellet was resuspended in TE buffer (10 mM Tris-HCl and 1 mMEDTA, at pH 7.4) and digested with a RNase A (20 pg/ml) (Worthington, Freehold, N. J.) at 37" for 2 hr. The DNA was reextracted with phenol and chloroform, reprecipitated in ethanol, and dialyzed overnight against TE buffer. Following an additional ethanol precipitation, the final DNA pellet was resuspended in TE buffer.
DNA transfection. Plasmid DNA fragments (0.5-10 pg) were transfected into LTK-cells in 60-mm dishes (containing approximately 1 X lo6 cells) using the calcium phosphate technique of Graham and van der Eb (1973) as modified by . Calf thymus DNA was used as carrier to achieve a final DNA concentration of 20 pg/ml. The cell monolayers were exposed to the calcium phosphate-DNA mixture (0.5 ml/dish) for 30 min at room temperature before fresh medium was added. The cultures were fed 24 hr later with HAT-containing selective media. Control dishes treated with either carrier DNA alone or no DNA were included with each experiment. As a positive control for TK transformation, we used the HSV-1 TK gene cloned into pBR 322 (HSV-106) (BRL). Colonies were isolated with the use of 5-mm filter paper discs (Whatman, No. 1) that had previously been soaked in 0.05% trypsin with 0.1% EDTA (Quality Biological). The discs were placed over the colonies for 5 min and then transferred into dishes containing fresh HAT medium. The dishes were fed with fresh HAT medium every 4 to 5 days. DNA hybridization. Cellular DNA was digested with BamHI, EcoRI, SmaI, P&I, PvuII, S&I, HindIII, or XbaI restriction endonucleases (BRL) according to the manufacturer's instructions.
The cellular DNA was electrophoresed through 1% agarose horizontal slab gels, denatured in situ, and transferred to nitrocellulose paper (Southern, 1975). Cloned VZV DNA fragments were labeled in vitro with [a-32P]deoxycytidine triphosphate (dCTP) (sp act, >3000 Ci/mmol) (Amersham, Arlington Heights, Ill.) to sp act of 0.5-3.0 X lo* cpm/pg using Nick Translation Kits (BRL). Autoradiography was accomplished at -70" with Kodak XAR-5 film in casettes fitted with DuPont Lightning-Plus intensifying screens for 2 to 16 hr. Relative DNA concentrations were estimated by densitometry. DNA fragment sizes were determined by comparison to bacteriophage lambda DNA digested with Hind111 (BRL).

RNA was isolated from VZV-infected Flow 5000 cells (Flow Laboratories,
McLean, Va.), LTK-cells, or VZV-dPK transformants. Briefly, each cell line was grown to confluence in two 150-cm2 tissue culture flasks, washed three times in ice cold phosphate-buffered saline (PBS) (pH 7.4), and lysed in situ with 4 M guanidinium thiocyanate. The lysate was layered onto a 5.7 Mcesium chloride cushion and centrifuged in an SW41 rotor at 33,000 rpm for 25 hr (Chirgwin et al., 1979;Glisen et ah, 1974).

The supernatant
was decanted, and the RNA pellet was further purified as previously described (Ostrove et aZ., 1985). Aliquots of cellular RNA were heated in 50% formamide, 6% formaldehyde, and 1X electrophoresis buffer (20 mM sodium [3-IN-morpholino-propane-sulfonic acid [MOPS], 5 mM sodium acetate, and 1 mM EDTA at pH 7.0) for 10 min. RNA samples (4-10 pg/lane) were electrophoresed at 35 mA for 16 hr through 6% formaldehyde/ 1.5% agarose horizontal slab gels (Rave et al., 1979). The transfer of RNA to nitrocellulose and the hybridizations were carried out as previously described (Ostrove et al., 1985). RNA sizes were determined by comparison to single-stranded DNA markers. dPK assay. The various cell lines to be used for both the TK and the deoxycytidine kinase activity studies were grown to confluence in 60-mm dishes, washed three times in PBS (pH 7.4), and scraped from the dishes. The cells were pelleted at 1500 rpm for 10 min in microcentrifuge tubes, and the supernatant was completely re-moved. The cell pellets were stored at -70". Pellets of each cell type to be tested were thawed on ice, and 100 ~1 of lysis buffer was added (10 mM Tris-HCl at pH 8.0,160 mM KCl, 250 mM sucrose, 1.5 mM MgC12, 0.5% NP-40, and 3 mM2-mercaptoethanol). The samples were vortexed twice (2 set each) and incubated on ice for 10 min. The samples were then vortexed vigorously for 5 see and spun for 10 min in a microcentrifuge (13,800 9). The supernatants were tested for enzyme activity (modification of Kit et ab, 1973;Kit and Leung, 1974 Yamanishi et ah, 1981). Fiftymicroliter samples of the reaction mixtures were spotted onto DE81 filters (Whatman, 2.4 cm), given two 60-min washes in distilled water, dipped five times in 95% ethanol, and dried. The filters were placed in Aquafluor (New England Nuclear, Boston, Mass.) and counted in a liquid scintillation counter. Enzyme activities were corrected for the amount of protein in the extracts. Protein concentrations were determined by the method of Bradford (1976).

RESULTS
Sixteen recombinant VZV DNA BamHI and EcoRI fragments representing 75% of the viral genome (map units 0.18-0.66 and 0.73-1.00, except for about 500 bp of the unique short segment (Us) were transfected into LTK-cells. The induction of stable biochemical transformants expressing TK activity was determined by formation of LTKf clones under HAT selection. This activity was occasionally observed with large amounts of calf thymus DNA and was seen with the HSV TK-containing recombinant (HSV-106) and with two fragments of VZV DNA: BamHI-H (map units 0.47-0.52) and EcoRI-D (map units 0.50-0.62). These fragments are known to overlap and share a 2.2-kb region of the genome (map units 0.50-0.52) (Straus et aZ., 1982) (Fig. 1). VZV-transformed LTK+ colonies developed 18 to 25 days after transfection and arose from a background of surviving cells. The majority of such surviving cells failed to proliferate into clones, perhaps demonstrating only transient expression of TK activity. No such background was seen in dishes transfected with carrier DNA alone. The VZV DNA-transformed cells resembled the parental LTK-cells in terms of both morphology and growth rate. HSV-transformed colonies appeared more quickly (at approximately 14 days after transfection) and exhibited more rapid growth than VZV transformants. The transforming frequency of both BamHI-H and EcoRI-D was 0.2-0.4 colonies/dish/pg of DNA transfected, which was approximately loo-fold lower than that seen with HSV TK genes under similar conditions. No LTK+ revertants were seen in LTKcells not transfected with foreign DNA, consistent with the known reversion frequency of less than lo-' for these cells (Wigler et ab, 1978). Three stable clones resulting from the transfection of BamHI-H (BHl and BH2) and EcoRI-D (EDl) were established and have been maintained in HAT-containing media for more than 25 passages. Between passages 6 and 8, the DNAs from these biochemically transformed cell lines were purified, and the presence within them of viral or plasmid sequences was determined by Southern hybridizations with purified VZV BamHI-H, HSV-106, or the 2.  and BH2 following cleavage with BamHI, EcoRI, ,%I, HindIII, SmaI, P&I, or PvuII revealed fragments of the sizes predicted from the restriction map (see Fig. 1). This suggests that most of the transfected DNAs were incorporated intact into the cells. Digestion of BH2 DNA with XbaI, an enzyme that does not cleave the VZV BamHI-H fragment or within pBR 322, and subsequent hybridization with a homologous VZV DNA probe revealed a single band of very high molecular weight (>23 kb) (Fig. 3, lane 3). Further analysis of transformed cell DNAs with enzymes that cleave within the VZV inserts or pBR 322 revealed multiple bands of higher and lower molecular weights than those of the transfected DNA fragments. The combined data indicate that much of the transfected DNA is present as tandem repeats with integration into the cellular DNA in multiple sites.
The HSV DNA transformants contained a low copy number per cell of the HSV TK gene and its associated plasmid DNA. With the conditions employed, no cross-hybridization was seen between the HSV TK sequences and the VZV sequences. Transfor-mants that arose from transfections with calf thymus DNA alone contained no sequences homologous to VZV, HSV, or pBR 322 DNAs (data not shown).
Extracts from the transformed cell lines BHl, BH2, and ED1 as well as from a cell line transformed with the HSV-1 TK gene (HSV), LTK-cells, and Vero cells were assayed for kinase activity. BHl, BH2, and ED1 cells demonstrated a 70-to 115-fold greater TK activity than the parental LTK-cells (Table 1). The abilities of these cellular extracts to phosphorylate deoxycytidine were compared to their abilities to phosphorylate thymidine. It was observed that all cell lines transformed with VZV DNA showed greater deoxycytidine kinase activity than TK activity in accord with prior reports of the substrate preferences of purified VZV dPK (Dobersen et ah, 1976;Ogino et al, 1977;Fyfe et al., 1982) ( Fig. 4). The opposite pattern was seen with extracts from cells transformed with HSV TK and Vero cells, also in accord with prior observations with HSV TK (Fyfe et al., 1978).
Northern hybridization analysis of the cellular RNA extracted from BHl, BH2, EDl, and LTK-cells demonstrated the transcription of VZV-specific mRNAs in the BHl, BH2, and ED1 transformed cell lines. To identify the messages most likely to be associated with the transfected dPK activity, total cellular RNA from these cell lines were hybridized to the in vitro CX-~~Plabeled, gel purified 2.2-kb DNA fragment representing the region of overlap between the BumHI-H and EcoRI-D fragments. The VZV-specific messages from these transformed lines were then compared to those detected by the hybridization of VZV-infected Flow 5000 cell RNA using the same 2.2-kb overlap-fragment probe (Fig. 5)  fragments into mouse LTK-cells followed by selection in HAT medium, we were able to screen a large portion of the VZV genome for sequences capable of inducing biochemical transformation to the LTK+ phenotype. We identified two separate fragments-BamHI-H and EcoRI-D-that exhibit transforming activity. These fragments share a 2.2-kb region of DNA, one that is large enough to encode a protein with a MW of greater than 35,000, the presumed size of the VZV dPK (Shiraki et al., 1985), and is larger than the minimal corresponding fragment of HSV DNA needed to transfer TK activity in similar systems (Reyes et ah, 1982). Although LTK' transformants were noted following transfections with calf thymus DNA alone, these cells exhibited very low TK activity and contained no viral RNA or DNA sequences.
The region of VZV DNA associated with TK transformation lies within the long unique (UL) segment of the genome, as does the HSV TK gene in its genome. Its position, at map units 0.50-0.52, is 0.31 map units in from the &-internal repeat (IRL) junction, and is similar to the distance of the HSV TK gene from the HSV terminal repeat (TR,,) sequence. This finding is consistent with A. Davison's impression from cross-hybridization studies that the organization of the VZV genome is similar to that of HSV except for an inversion of part of the UL sequence (personal communication).
The dPK activities of the VZV-transformed cell lines were markedly greater than that of the parental LTK-cell line. Limited biochemical characterization of this activity showed that deoxycytidine was phosphorylated more rapidly than thymidine. Such findings are consistent with the reported behavior of the enzyme recovered from VZV-infected cells; however, they are the opposite of those reported for the HSV TK and those we observed in extracts prepared from HSV TKtransformed cells (Dobersen et al., 1976;Ogino et ab, 1977;Fyfe et ab, 1978Fyfe et ab, , 1982. The VZV-transformed cell lines contained extremely high copy numbers of the transfected DNA, yet produced dPK in amounts similar to that in the HSV-transformed cell line, where copy number is usually low Wigler et al, 1978). Some possible explanations for these findings are that the VZV dPK promoter is a weak one or that additional VZV gene products are necessary for efficient VZV dPK gene expression.
We were further able to show that three independently derived VZV DNA-transformed cell lines expressed mRNAs homologous to the segment of VZV DNA associated with the dPK-transforming activity. Only the 1.8-kb message detected in the BHl and BH2 transformants comigrated with an abundant message expressed from this region of the genome in infected cells. The ED1 line, however, contained an abundant message of 2.2 kb in length. The 2.2kb transcript may originate from a fusion of the VZV dPK sequence with additional nonviral sequences and be initiated by a cellular or plasmid promoter. The 2.2-kb transcript contains pBR 322 sequences as determined by Northern blot analysis (data not shown). If the promoter for the VZV dPK gene were to lie to the left of the EcoRI site within the BarnHI-H fragment and the body of the message (open reading frame) were to lie to the right of the EcoRI site, then expression of the dPK by the EcoRI-D fragment would require initiation by a nonviral promoter.
Consistent with these hypotheses are recent analyses of the VZV DNA sequence in this region of the genome. A gene with limited sequence homology to the HSV TK gene has been found in which the promoter is upstream and to the left of the adjacent EcoRI site (A. Davison, personal communication). Thus, we believe the VZV dPK gene lies between map units 0.50 and 0.52 in the UL of the VZV genome and encodes a primary 1.8-kb transcript.