the replication of viral and cellular dna in human herpesvirus 6-infected cells

12
VIROLOGY 175, 199-210 (1990) The Replication of Viral and Cellular DNA in Human Herpesvirus 6-Infected Cells DAR10 DI LUCA,**’ GEORGE KATSAFANAS,* ERIC C. SCHIRMER,* N. BALACHANDRAN,t AND NIZA FRENKEL*,’ *Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, NIHflwinbrook Il. 1244 1 Park/awn Drive, Rockville, Maryland 20852; and tDepartment of Microbiology, University of Kansas Medical Center, Kansas City, Kansas 66 103 Received July 5, 1989; accepted November 15. 1989 Human herpesvirus 6 (HHV-6) is a newly identified lymphotropic herpesvirus. We have analyzed viral and host DNA replication in peripheral blood lymphocytes infected in the absence of drugs or infected in the presence of phospho- noacetic acid (PAA) or acyclovir (ACV). The results revealed the following: (i) Infection with HHV-6 resulted in the shutoff of host DNA replication. (ii) PAA at concentrations of 100 and 300 pglml significantly reduced virus replication. The drug inhibited viral DNA replication, whereas host cell DNA replication was not affected. This strongly suggests that HHV-6 encodes a PAA sensitive viral DNA polymerase. (iii) ACV at 20 PM did not interfere with virus production and virus spread. ACV at 100 PM only partly interfered with virus replication, whereas at 400 PM the block was more complete. Viral DNA replication was not affected by ACV at 20 PM. However, approximately 60 and 85% inhibition in viral DNA replication was observed in the presence of 100 and 400 PM of ACV. (iv) Assays for viral thymidine kinase (TK) revealed no significant increase in TK activity, whereas increased TK activity was noted following infection of the same peripheral blood lymphocytes with herpes simplex virus. Thus, either HHV-6 does not encode a tk enzyme which can phosphorylate ACV or the inefficient block may reflect lower sensitivity of the HHV-6 DNA polymerase to the drug. 0 1990 Academic Press, Inc. INTRODUCTION Salahuddin et a/. (1986) described the isolation of a novel herpesvirus from lymphocytes of patients with acquired immunodeficiency syndrome (AIDS) and from patients with lymphoproliferative disorders. Additional isolates have been obtained (Tedder et al., 1987; Lo- pez et al., 1988; Agut et al., 1988; Yamanishi et al., 1988; Becker et al., 1989) and recent surveys have shown that more than 80-90% of humans develop an- tibodies to the virus early in life (Saxinger et a/., 1988; Okuno et al., 1989). Yamanishi et al. (1988) presented evidence that human herpesvirus 6 (HHV-6) is the causative agent of exanthem subitum (roseola infan- turn). However, it is as yet unclear whether this virus is responsible for any other human disease(s). We have studied the sensitivity of HHV-6 to phos- phonoacetic acid (PAA) and acyclovir (ACV). Herpesvi- rus DNA polymerases are inhibited by PAA, at concen- trations which do not inhibit the cellular DNA polymer- ases (Leinbach et a/., 1976; Purifoy and Powell, 1977). PAA was shown to bind to the DNA polymerase at the pyrophosphate binding site (Leinbach et al., 1976). ACV is an acyclic analog of deoxyguanosine. It re- quires phosphorylation to its active form (ACV-triphos- phate). Herpes simplex virus (HSV) and varicella zoster ’ Present address: Institute of Microbiology, University of Ferrara. Ferrara, Italy. ’ To whom requests for reprints should be addressed. virus (VZV) are inhibited even at low concentrations of the drug. Higher concentrations of the drug are re- quired for the inhibition of human cytomegalovirus (HCMV) and Epstein Barr virus (EBV) replication (Dor- sky and Crumpacker, 1987). The selective effect of ACV on HSV and VZV replica- tion can be attributed to properties of two viral en- zymes. (i) The viral thymidine kinase (TK) phosphory- lates ACV more efficiently than the cellular enzyme. (ii) The viral DNA polymerase exhibits a higher affinity for ACV triphosphate than the corresponding cellular en- zyme. ACV is being used in clinical practice as an effective anti-herpetic agent, whereas PAA is not used as an anti-viral drug due to its in viva toxicity. Both drugs have been employed in genetic and biochemical studies of herpesvirus DNA polymerases and viral TK enzymes. In an attempt to begin the characterization of genetic markers in the HHV-6 genome we have focused our attention on the effect of these drugs on HHV-6 replica- tion. Our studies have shown that PAA inhibited effi- ciently HHV-6 DNA replication. In contrast, ACV exhib- ited an effect only at very high concentrations. MATERIALS AND METHODS Virus and cells HHV-6, strain 229 (Lopez et al., 1988) was obtained from Dr. C. Lopez (CDC, Atlanta). Peripheral blood lym- phocytes (PBL) were purified from huffy coats of 199 0042.6822/90 $3.00 CopyrIght 0 1990 by Academic Press. Inc. All rights of reproductrw I” any form reserved

Upload: dario-di-luca

Post on 25-Aug-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

VIROLOGY 175, 199-210 (1990)

The Replication of Viral and Cellular DNA in Human Herpesvirus 6-Infected Cells

DAR10 DI LUCA,**’ GEORGE KATSAFANAS,* ERIC C. SCHIRMER,* N. BALACHANDRAN,t AND NIZA FRENKEL*,’

*Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, NIHflwinbrook Il. 1244 1 Park/awn Drive, Rockville, Maryland 20852; and tDepartment of Microbiology, University of Kansas Medical Center, Kansas City, Kansas 66 103

Received July 5, 1989; accepted November 15. 1989

Human herpesvirus 6 (HHV-6) is a newly identified lymphotropic herpesvirus. We have analyzed viral and host DNA replication in peripheral blood lymphocytes infected in the absence of drugs or infected in the presence of phospho- noacetic acid (PAA) or acyclovir (ACV). The results revealed the following: (i) Infection with HHV-6 resulted in the shutoff of host DNA replication. (ii) PAA at concentrations of 100 and 300 pglml significantly reduced virus replication. The drug inhibited viral DNA replication, whereas host cell DNA replication was not affected. This strongly suggests that HHV-6 encodes a PAA sensitive viral DNA polymerase. (iii) ACV at 20 PM did not interfere with virus production and virus spread. ACV at 100 PM only partly interfered with virus replication, whereas at 400 PM the block was more complete. Viral DNA replication was not affected by ACV at 20 PM. However, approximately 60 and 85% inhibition in viral DNA replication was observed in the presence of 100 and 400 PM of ACV. (iv) Assays for viral thymidine kinase (TK) revealed no significant increase in TK activity, whereas increased TK activity was noted following infection of the same peripheral blood lymphocytes with herpes simplex virus. Thus, either HHV-6 does not encode a tk enzyme which can phosphorylate ACV or the inefficient block may reflect lower sensitivity of the HHV-6 DNA polymerase to the drug. 0 1990 Academic Press, Inc.

INTRODUCTION

Salahuddin et a/. (1986) described the isolation of a novel herpesvirus from lymphocytes of patients with acquired immunodeficiency syndrome (AIDS) and from patients with lymphoproliferative disorders. Additional isolates have been obtained (Tedder et al., 1987; Lo- pez et al., 1988; Agut et al., 1988; Yamanishi et al., 1988; Becker et al., 1989) and recent surveys have shown that more than 80-90% of humans develop an- tibodies to the virus early in life (Saxinger et a/., 1988; Okuno et al., 1989). Yamanishi et al. (1988) presented evidence that human herpesvirus 6 (HHV-6) is the causative agent of exanthem subitum (roseola infan- turn). However, it is as yet unclear whether this virus is responsible for any other human disease(s).

We have studied the sensitivity of HHV-6 to phos- phonoacetic acid (PAA) and acyclovir (ACV). Herpesvi- rus DNA polymerases are inhibited by PAA, at concen- trations which do not inhibit the cellular DNA polymer- ases (Leinbach et a/., 1976; Purifoy and Powell, 1977). PAA was shown to bind to the DNA polymerase at the pyrophosphate binding site (Leinbach et al., 1976). ACV is an acyclic analog of deoxyguanosine. It re- quires phosphorylation to its active form (ACV-triphos- phate). Herpes simplex virus (HSV) and varicella zoster

’ Present address: Institute of Microbiology, University of Ferrara. Ferrara, Italy.

’ To whom requests for reprints should be addressed.

virus (VZV) are inhibited even at low concentrations of the drug. Higher concentrations of the drug are re- quired for the inhibition of human cytomegalovirus (HCMV) and Epstein Barr virus (EBV) replication (Dor- sky and Crumpacker, 1987).

The selective effect of ACV on HSV and VZV replica- tion can be attributed to properties of two viral en- zymes. (i) The viral thymidine kinase (TK) phosphory- lates ACV more efficiently than the cellular enzyme. (ii) The viral DNA polymerase exhibits a higher affinity for ACV triphosphate than the corresponding cellular en- zyme.

ACV is being used in clinical practice as an effective anti-herpetic agent, whereas PAA is not used as an anti-viral drug due to its in viva toxicity. Both drugs have been employed in genetic and biochemical studies of herpesvirus DNA polymerases and viral TK enzymes. In an attempt to begin the characterization of genetic markers in the HHV-6 genome we have focused our attention on the effect of these drugs on HHV-6 replica- tion. Our studies have shown that PAA inhibited effi- ciently HHV-6 DNA replication. In contrast, ACV exhib- ited an effect only at very high concentrations.

MATERIALS AND METHODS

Virus and cells

HHV-6, strain 229 (Lopez et al., 1988) was obtained from Dr. C. Lopez (CDC, Atlanta). Peripheral blood lym- phocytes (PBL) were purified from huffy coats of

199 0042.6822/90 $3.00 CopyrIght 0 1990 by Academic Press. Inc. All rights of reproductrw I” any form reserved

200 DI LUCA ET AL.

healthy individuals by centrifugation through Lympho- cyte Separation Media (Organon Teknika, Malvern, PA). The cells were cultured for 5 days in medium RPM1 1640 containing 10% inactivated fetal calf serum and 10 pg/ml of phytohemagglutinin (PHA) (Difco, Detroit, Ml). Virus stocks were produced by passaging the virus at low multiplicity of infection (m.o.i.) in PBLs using in- fection medium. Infection medium, described by Lopez et a/. (1988) consisted of RPMI 1640, supplemented with 5 pg/ml PHA, 5 pg/ml gentamycin and 5% delecti- nated interleukin-2 (Advanced Biotechnologies, Silver Spring, MD).

Virus titers were determined in PBLs by limiting dilu- tions of the infected cell cultures. Infection was deter- mined by the scoring for cytopathic effect (CPE) 2 weeks later.

For the preparation of the mock-infected inoculum, an aliquot of the same PBLs which were used for the preparation of the virus stock was incubated in infec- tion medium, and processed in parallel with the virus stock.

Infection in the absence or presence of drugs

Infection was performed at a multiplicity of 0.1 infec- tious units (i.u.) per cell. For virus adsorption, 10” virus i.u. per milliliter were added to 10’ cells/ml. Two hours later, the infected cells were diluted with fresh infection medium to attain a final concentration of 1 O6 cells per milliliter.

For cultures which were treated with drugs from the time of infection, the cells were mixed with the appro- priate concentrations of PAA (Sigma, St Louis, MO) or ACV (Zovirax, Burroughs Wellcome, Research Triangle Park, NC), 1 hr prior to the addition of virus. For the cultures in which the drugs were added on the 3rd day of infection, PAA and ACV were added to aliquots of the infected cells at 61 hr post infection (p.i.) in the labeling medium as described below. Incubation con- tinued for 4 hr prior to the addition of [“PI- orthophosphate.

Analysis of drug toxicity

To determine the level of drug toxicity, duplicate un- infected cell cultures were incubated in the presence of the drugs, and aliquots removed for analysis 24 and 90 hr later. For the determination of cell counts and cell sizes, the cells were analyzed in a Coulter Counter ZM equipped with a Coulter Channelyzer 256 (Coulter Electronics, Hialeah, FL).

lmmunofluorescence

For indirect immunofluorescence assays (IFA), the cells were fixed with acetone and stained with mono-

clonal antibody (mAb) 9A5D12 (Balachandran et a/., 1989). The slides were then stained with fluorescein isothiocyanate-conjugated rabbit anti-mouse immuno- globulins (Dako, Santa Barbara, CA). Pictures of the same microscope fields were taken in visible and in uv light in order to determine the percentage of fluores- cent cells. For the actual counting, areas where individ- ual cells within aggregates could be clearly delineated were chosen. At least 1200 cells were counted per sample, at each time point shown in Fig. 6.

DNA labeling, extraction, and analysis

For labeling with [“‘PIPi, the cells were preincubated from 61 to 65 hr p.i. in labeling medium containing the same ingredients as infection medium, except that phosphate-free RPMI 1640 was used and the amount of serum was reduced to 1%. The cells were then la- beled from 65 to 86 hr p.i., using labeling medium con- taining 20 &i/ml of [32P]orthophosphate (Amersham, Arlington Heights, IL). For the preparation of DNA, cells were lysed in the presence of sodium dodecyl sulfate and proteinase K, as described by Locker and Frenkel (1979). DNA was purified by extractions with phenol, chloroform-2% isoamyl alcohol, and ether. The DNA was then precipitated with ethanol and resuspended in TE buffer (10 mlLlTris, pH 7.5, 1 mM EDTA). The DNA was cleaved with Sal1 or BarnHI (Biolabs, Beverly, MA) and electrophoresed in 0.6% agarose gels. Following ethidium bromide staining and photography, the gels were dried and exposed to X-ray film with intensifying screens. To calculate the residual synthesis in the presence of the drugs, the relative intensities of several virus bands were compared in the various lanes of Fig. 8. The quantitation of band intensities was performed using a LKB Ultroscan XL densitometer.

Preparation of HSV DNA size markers

32P-labeled HSV-infected cell DNA was prepared as previously described (Locker and Frenkel, 1979; Deiss and Frenkel, 1986). Vero cells were infected with 2 PFU of HSV-1 (Justin) per cell. The DNA was digested with BarnHI and Bglll, thus providing a wide range of size markers (Locker and Frenkel, 1979).

Thymidine kinase assays

For the TK assay, cells were either mock infected (with mock PBL inoculum) or were infected with 10 plaque forming units (PFU) of HSV-1 (Justin) per cell, or with 0.2 i.u. of HHV-6 per cell. The TK assay was performed as previously described (Leiden et al., 1976; Posteta/., 1981).

FIG. 1. HHV-6 cytopathic effect. Peripheral blood lymphocytes showing the characteristic CPE induced by HHV-6 infection. (A) Mock-infected cells. (B) Balloonrng cells within aggregates of cells. At late stages p.i. these large ballooning cells burst leaving “chewed up” aggregates, such as the aggregates in C. A membranous boundary surrounding these aggregated cells becomes visible at later stages p.i. Some of the cells may have fused into syncitia. (D) Syncrtial cells, each containing several nuclei, as apparent from electron microscopy studies (R. Roffman and N Frenkel, unpublrshed results). Origrnal magnification 200x

RESULTS

infection of peripheral blood cells with HHV-6 strain 229

The 229 strain which was isolated by Lopez et al. (1988) does not grow to high titers in continuous cell lines (Lopez et a/., 1988; Wyatt, Katsafanas, Schirmer, and Frenkel, unpublished results). Consequently, for the studies described in this paper we have employed human PBLs. Figure 1 shows the typical CPE observed in PBLs following infection with the 229 strain. Charac- teristic CPE included ballooning cells (Fig. 1 B), aggre- gates of infected cells which appear to be in the pro- cess of cell fusion (Fig. 1C). The large cells shown in Fig. 1 D are multinucleated syncytial cells as shown by electron microscopic analyses (E. Roffman and N. Frenkel, unpublished results). The characteristic CPE exhibited by HHV-6 (Z29))infected PBLs is similar to that described for the strain originally isolated by Sala- huddin and co-workers (Ablashi et a/., 1988; Salahud- din et al., 1986).

CPE in cells infected in the presence of PAA and acyclovir

To determine whether HHV-6 is sensitive to PAA and ACV, replicate cell cultures were infected with 0.1 i.u. per ceil of HHV-6 (229) in the absence of drugs, or in the presence of 100 and 300 pglml of PAA, or 20,100, and 400 PM of ACV. A mock-infected inoculum was prepared in parallel to the virus inoculum from the same PBLs in which the virus stock was prepared. Fig- ure 2 shows the apparent CPE observed on Day 4 post- infection. The data can be summarized as follows: (i) PAA, at 100 pg/ml, and even more so at 300 pg/ml, exhibited low levels of toxicity characterized by limited cell fragmentation. In contrast, ACV at 20 PM, as well as 100 PM, resembled the untreated mock-infected cells. With ACV at 400 &I, some toxicity was ob- served. (ii) In the culture infected in the absence of drugs, the characteristic HHV-6 CPE was first ob- served by the 2nd day p.i. (iii) The infection did not pro- gress in the presence of 100 or 300 pg of PAA per milli- liter. The presence of a few large infected cells was

202 DI LUCA ET AL.

HHV-6

NO DRUG ACV 20 pM

PAA 100 p.g/ml ACV 100 pM

PAA 300 pg/ml ACV 400 pM

NO DRUG

PAA 100 pg/ml

PAA 300 pg/ml

ACV 20 j.tM

ACV 100 pM

ACV 400 pM

FIG. 2. Morphology of mock-infected or infected cells treated with PAA and ACV. Microphotographs of cells which were mock infected or infected with HHV-6 (229) in the presence or absence of drugs as described in the text.

noted. However, these cells could have represented the virus inoculum and their proportion in the cell popu- lation (approximately 10%) was consistent with this in- terpretation (iv) In the cultures treated with 20 or 100 &I ACV the infection was apparent, as judged by the appearance of large individual infected cells. However, the CPE, as judged by cell ballooning and syncytia for- mation, was delayed and reduced. At 400 @W some toxicity was noted in the infected culture as judged by apparent cell fragmentation.

Quantitative evaluation of drug toxicity

To obtain a more quantitative estimate of drug toxic- ity, cell numbers and cell sizes in the mock-infected cells were analyzed, using a Coulter Counter. The re- sults are summarized in Table 1 and in Fig. 3. ACV at 20 @W did not exhibit any toxic effect. PAA at 100 PgI ml and ACV at 100 NM exhibited low levels of toxicity, as judged by reduced cell numbers and cell fragmenta- tion. More pronounced toxity and cell fragmentation were observed in the presence of PAA at 300 pg/ml, and ACV, at 400 pM.

Virus yields in the presence of the drugs

To determine the effect of the drugs on the yield of infectious virus, aliquots of the infected cultures were

titered at daily intervals, starting from Day 2 of infection. The results are shown in Fig. 4. By Day 5, virus yields in the cultures infected in the absence of drugs ap- proached lo8 i.u./ml per lo6 cells. In contrast, in the presence of PAA, by 5 days p.i., infectious virus titers had dropped several logs below 1 O5 i.u., which was the amount of input virus per milliliter per lo6 cells. Thus, the input virus was gradually inactivated. Because the drop of infectious virus in the presence of PAA at 100 pg/ml was slower than that in the presence of 300 pgl ml, it can be concluded that, albeit inefficiently, low lev- els of infectious virus were produced in the presence of 100 pglml PAA. At 300 rglml, the drug was more effective in inhibiting virus production.

TABLE 1

CELLCOUNTSIN MOCK-INFECTEDCULTURESTREATEDWITH DRUGS

Treatment 24 hr

(cells/ml) 90 hr

(cells/ml)

No drug PAAlOOfig/ml

PAA 300 pg/ml ACV 20pM ACVlOO cLM ACV400 pM

1.1 x lo6 1.2 x lo6

1.1 x lo6 1.0x lo6

9.5 x 1 o5 7.5 x 105 1.1 x 106 1.2 x 106 1.0x 106 9.9 x lo5 8.4X lo5 7.3 x lo5

VIRAL AND CELLULAR DNA REPLICATION IN HHV-6-INFECTED CELLS 203

24 HOURS p.i. 90 HOURS p.i.

ACV 20 jlM

FIG. 3. Analysis of morphologrcal alterations of peripheral blood lymphocytes treated with different concentrations of PAA or ACV. PBLs were treated with PAA or ACV for 24 or 90 hr, and cell sizes determined in a Coulter Counter. The effects of PAA at 300 FM and ACV at 400 PM are noticeable.

ACV at 20 j&I did not reduce the yields of infectious virus, and 1 O8 infectious units per 1 O6 cells were pro- duced by Day 5 p.i., similar to virus yield in the absence of drugs. At 100 PM of ACV, there was a gradual in- crease in infectious virus titers, reaching 1 O6 i.u./ml by Day 5 p.i. Thus, although accompanied by some virus production, the drug decreased infectious virus yields by approximately 2 logs. A hundredfold reduction in vi- rus yield should be considered as maximal, inasmuch as the infection was performed at 0.1 i.u./cell and the titers reached in the control cells, infected in the ab- sence of drugs, undoubtedly reflected several rounds of replication.

In the cultures infected in the presence of 400 PM ACV, virus titers decreased somewhat by Day 4 p.i., reaching lo4 i.u./ml on Day 5 p.i. It is reasonable to assume that the rate of inactivation of the input virus was the same in the cultures infected in the presence of 300 hg/ml PAA and 400 @I ACV. Because the re- duction in virus titer was slower in the presence of 400 &I ACV, we can conclude that some virus was produced, even in the presence of this high concentration of ACV.

Effect of PAA and ACV on virus spread Using the m.o.i. of 0.1 i.u./cell we expected that ap-

proximately 10% of the cells would become infected

during the first cycle of infection. To determine whether PAA and ACV affected the progression of virus infec- tion, aliquots of the infected cultures shown in Fig. 2 were removed on Days 2, 3, and 5 p.i. The cells were tested for the presence of viral antigens by IFA, using the mAb 9A5D12, which reacts with two related poly- peptides of sizes 1 1 OK and 41 K (Balachandran et al., 1989). Preliminary studies (Balachandran, unpublished results) have shown that the 41 K polypeptide is virion associated, and the 11 OK may be an aggregated form of the 41 K polypeptide.

Representative examples of the IFA tests are shown in Fig. 5 for Day 6 p.i. No immunofluorescence was observed with the mock-infected cells. A fraction of the infected cells exhibited bright nuclear fluorescence, whereas in the remaining cells there was a diffuse cy- toplasmic immunofluorescence. This pattern is consis- tent with the mAb targeting a capsid component. Such protein could accumulate in the nucleus during capsid assembly, and could then be translocated to the cyto- plasm during virion maturation and egress. The frac- tions of fluorescent cells were quantified for each time point, as described under Materials and Methods. The data are summarized in Fig. 6. In the culture infected in the absence of added drugs, 94% of the cells were

204 DI LUCA ET AL.

no drugs

FIG. 4. Infectious virus yields in

PAA,lOO ug/ml

PAA, ug/ml

o! -, . , . , . , . , . , 0 1 2 3 4 5 6

Days

6

0 1 2 3 4 5 6

Days

cells infected in the presence of the drugs. Virus titers were determined by terminal dilution assays.

infected by Day 4 p.i. PAA at 100 and 300 pg/ml re- duced the fraction of fluorescent cells to 18 and 13%, respectively. From the ratio of fresh PBLs to infected PBLs constituting the inoculum virus it could be calcu- lated that approximately 10% of the immunofluoresc- ing cells represented cells in the virus inoculum. Fur- thermore, the fraction of immunofluorescing cells did not increase from Days 3 to 6 of infection in the pres- ence of 100 and 300 pg PAA per milliliter. We conclude on the basis of these results that PAA inhibited virus replication and spread.

ACV, at 20 PM, did not inhibit virus spread. Thus, 85 and 96% of the cells reacted with the mAb by Days 3 and 4 p.i. A slight delay in the progression of infection was noted in the presence of 100 till/l ACV. However, by Day 6 p.i., 94% of the cells were positive. In con- trast, significant inhibition of virus spread was ob- served in those cultures which were infected in the presence of 400 PM ACV, and the fraction of immuno- fluorescent cells did not increase above 30%.

Analyses of viral and host DNA replication To analyze the effect of the drugs on viral DNA repli-

cation, replicate cell cultures were mock infected or in- fected in the absence or presence of PAA or ACV. The drugs were added starting 1 hr before infection. The cells were then labeled with [32P]orthophosphate, from 65 to 86 hr p.i. Total infected cell DNA was prepared and digested with the Sal1 and BarnHI restriction en- zymes. Figure 7 shows the digestion patterns. The re- sults of these analyses revealed the following: (i) The SalI enzyme did not cleave the host cell DNA (lanes 3 to 8) due to its methylation. In contrast, HHV-6 DNA was cleaved with &/I, yielding fragments ranging in size from 1 to more than 30 kb (e.g., lane 9). In this respect HHV-6 DNA was similar to HSV DNA which is also selectively not methylated during productive infec- tion while the cell DNA is resistant to Sal1 due to its methylation (Deiss and Frenkel, 1986).

Ethidium bromide staining of the gel (data not shown) revealed similar amounts of S&l-resistant (i,e.

VIRAL AND CELLULAR DNA REPLICATION IN HHV-6-INFECTED CELLS 205

ACV, 100 pM

FIG. 5. lmmunofluorescence of HHV-6-infected cells treated with PAA or ACV. PBLs were either mock infected or infected in the ab- sence of drugs or in the presence of drugs from the time of infection. Aliquots of the cultures were taken daily and tested by IFA with mAb 9A5D12. Shown are data from the 6th day p-i.

host) DNA in all the lanes. Thus, the lanes in Fig. 7 rep- resent equal numbers of cells and similar DNA recover- ies in the various samples.

The synthesis of host DNA was strongly inhibited in the cells which were infected in the absence of drugs

(Fig. 7, lane 9) as well as in the cells which were in- fected in the presence of 20 pM ACV (lane 12). The reduced amount of 32P-labeled S&/l-resistant DNA in lanes 9 and 12 represented reduced synthesis of host DNA during the labeling interval. It did not represent demethylation of the cell DNA, because the SalI resis- tant band was clearly visible in the ethidium bromide pattern of the gel (data not shown). Thus, HHV-6 ex- presses a host DNA shutoff function, at least at late times p.i.

PAA at 100 and 300 pg/ml, and ACV at 20,100, and 400 pM, did not inhibit cell DNA replication in the mock-infected cells (Fig. 7, compare lanes 4-8 with lane 3).

PAA at 100 and 300 pg/ml reduced significantly the amount of 32P-labeled HHV-6 DNA. ACV, at 20 pM, did not significantly inhibit the synthesis of viral DNA. Re- duced synthesis was observed in the presence of 100

100

80 1 UJ

g 60

r Nodrug

2 3 4 5 6 7

Days P.I.

Ac~;~ookM \ Nodrug

O 40 5

s 20 - ACV400~M

2 3 4 5 6 7

Days P.I. FIG. 6. Analysis of virus spread in the presence of different concen-

trattons of PAA and ACV, as determrned by IFA. The fraction of immu- nofluorescent cells was determined from photographs of the same fields with visible and fluorescent light, as described under Materials and Methods.

206 DI LUCA ET AL.

- 36.6 - 24.3

10.8

9.1

7.5

6.6

5.9

4.9

3.8 3.4

2.7

2.1

- 10.5

- 6.1

- 5.7 - 5.2

- 4.6

12 3 8 9 14 15 20 21 26 27 28

FIG. 7. Autoradiogram of the Sal1 and BamHl patterns of 32P-labeled DNA extracted from PBLs infected with HHV-6 in the presence of drugs. Mock-infected or infected cells were treated with different concentrations of PAA and ACV 1 hr before infection, and were labeled with [32P]orthophosphate, from 65 to 86 hr p.i. Total cell DNA was extracted and digested with SalI or BamHl restriction enzymes.

PM ACV. However, substantial inhibition of the incor- poration of [32P]P8 into viral DNA was observed only in the cells which were infected in the presence of 400 pM ACV.

32P-labeled host DNA was synthesized in the cul- tures which were infected in the presence of PAA and in the cultures which were infected in the presence of 100 and 400 pMACV. The lack of inhibition may reflect the fact that only a small fraction of the cells was in- fected under these conditions, as revealed by the im- munofluorescence data shown in Fig. 6.

Inspection of the BarnHI digests (Fig. 7, lanes 15- 26) yielded the same conclusions as above, regarding viral DNA synthesis. With regards to cell DNA synthe- sis, the BarnHI digests contained (as expected) an un- resolved continuous array of fragments which, unlike the SalI patterns, could not be readily quantitated. However, the BarnHI-digested DNAs from the mock- infected cultures treated with drugs contained a faint distinct band (Fig. 7, lanes 16-20, arrow next to lane 20) which was not visible in the BarnHI-digested DNA from the cells infected in the absence of drugs (lane 15). This band, most likely, represented stress DNA synthesis as described previously for cells treated with drugs (Smith and Vinograd, 1972). Interestingly, a mi-

nor band, comigrating in the same position, was clearly visible in the BarnHI patterns of the DNA samples ex- tracted from the infected cells (lanes 21-26). This re- sult suggests that HHV-6 infection induces stress DNA replication.

The reduced synthesis of 32P-labeled viral DNA in the presence of PAA and ACV could reflect a direct inhibi- tion of viral DNA replication by the drugs, through the interaction with the viral DNA polymerase. Alterna- tively, it could reflect an indirect effect on viral DNA rep- lication. For example, the drugs could inhibit the ex- pression of genes (viral or cellular) which are required for viral DNA replication. Furthermore, because the m.o.i. used in the experiment was 0.1 i.u./cell, the ma- jority of the cells were not infected in the first round of infection. Thus, the reduced synthesis of 32P-labeled viral DNA could have resulted from the inhibition of a late event essential for virus production and virus spread to the remaining uninfected cells.

To determine whether PAA and ACV had a direct effect on viral DNA replication, a second experimental design was undertaken. Duplicate aliquots of the mock-infected or infected cultures were incubated for 61 hr p.i. in the absence of drugs, permitting virus spread to the entire culture, and the expression of viral

VIRAL AND CELLULAR DNA REPLICATION IN HHVWNFECTED CELLS 207

36.8 24.3

10.8

9.1

7.5

6.6

5.9

4.9

4.6

3.8 3.4 3.2

12 3 6 9 14 15 16

FIG. 8. Autoradiogram of the Salt and BarnHI patterns of “P-labeled DNA extracted from PBLs infected with HHV-6 in the absence of drugs. Duplicate cultures were not treated (lanes 3. 9, 16, 22) or were treated with PAA or ACV at the indicated concentrations, from 61 to 85 hr p.i. The cells were labeled with [32P]P, from 65 to 86 hr p.t

DNA replication functions. At that time, aliquots of the mock-infected and infected cultures were treated with PAA (100 and 300 fig/ml) or ACV (20, 100, and 400 PM) for 4 hr. The cultures were then labeled (from 65 to 86 hr p.i.) with [32P]Pi. DNAs were harvested in paral- lel to the samples which have already been described above (Fig. 7). Analyses of these DNAs with the Sell and BarnHI enzymes are shown in Fig. 8. The results revealed the following: (i) No significant inhibition of host cell DNA synthesis was observed in the mock-in- fected cultures which were treated with the drugs (lanes 3-8). The slight reduction in host DNA replica- tion in the presence of PAA at 300 pg/ml (lane 5) and ACV at 400 PM (lane 8) was caused by an imperfect recovery of DNA during the extraction or gel loading procedures. The ethidium bromide pattern of the gel (data not shown) revealed that this was the case.

As already noted above, by 65 hr p.i., there was a significant shutoff of the synthesis of host DNA. This result also shows that the majority, if not all, of the cells in the cultures were infected by the time of the pulse at 65 hr p.i. Had there been a significant fraction of the cells which were not infected, we would expect these uninfected cells to continue to synthesize cell DNA. Thus, it is reasonable to suggest that at the time of

drug addition the majority of the cells were infected and were in the process of synthesizing viral DNA. There- fore, the amount of 32P-labeled DNA produced during the pulse in the presence of drugs is a direct measure of the inhibitory effect of the drugs on viral DNA replica- tion.

As seen in lane 10, PAA at 100 pg/ml partly inhibited HHV-6 DNA replication. Densitometic scanning of the autoradiogram shown in Fig. 8 revealed approximately 24% residual synthesis. More significant inhibition (ap- proximately 12% residual synthesis) was noted with PAA at 300 pg/ml. ACV at 20 FM had almost no inhibi- tory effect (86% residual synthesis). The inhibitory effect increased somewhat (approximately 40% resid- ual synthesis) at 100 PAIIACV. Finally, significant inhibi- tion was observed at 400 &’ (14% residual synthesis).

Analyses of the DNA samples with the BarnHI en- zyme (Fig. 8, lanes 16-27) corroborated these results. In addition, as seen from inspection of lanes 22 to 27, PAA and ACV did not inhibit the stress DNA synthesis induced by HHV-6. Thus, it is unlikely that the stress DNA replication induced by HHV-6 is mediated by a viral DNA polymerase. If it were, we could expect it to be inhibited by PAA and ACV to the same extent that these drugs inhibit viral DNA replication.

208 DI LUCA ET AL.

TABLE 2

THYMIDINE KINASE ACTIVIP/

Time 0-d Mock HHV-6 HSV-1

6 nd nd 7,851 24 4096 3176 16,216 44 1881 2611 10,361 68 1226 2277 nd

Note. nd, not done. Values are averages of two independent infec- tions for each time point.

Assay for thymidine kinase activity in lymphocytes infected with HHV-6

ACV requires phosphorylation, which in HSV- and VZV-infected cells occurs readily by the viral TK. The finding that 20 and 100 PM of ACV do not efficiently inhibit HHV-6 replication could reflect, at least in part, the absence of a virally encoded thymidine kinase which could efficiently pliosphorylate ACV. In an at- tempt to test this hypothesis we assayed the TK activity in PBLs which were either mock infected or infected with HHV-6 at 0.2 i.u. per cell. Replicate cultures of the same PBLs were also infected with HSV-1 (Justin) at a m.o.i. of 10 PFlJ/cell (titer determined in Vero cells). At 6, 24, 44, and 68 hr p.i., duplicate cultures were har- vested and assayed for TK activity. Inspection of the HSV-infected PBLs did not reveal the typical HSV CPE observed in infected cell monolayers. Nonetheless, the HSV-infected cells differed morphologically from the mock-infected cells. The cells were smaller and exhib- ited fragmented appearance. As expected, the HHV-6- infected cells showed the characteristic CPE and by 68 hr p.i. a great majority of the cells appeared infected.

The results are summarized in Table 2. The mock- infected cells exhibited measurable TK activity. In the HSV-1 -infected cultures, the TK activity increased and peaked by 24 hr p.i. These results indicate that HSV was capable of inducing TK activity in PBLs. In con- trast, in the HHV-6-infected cells, the TK levels resem- ‘bled those in the mock-infected cells. TK activity did not increase from 24 to 68 hr p.i., although a significant fraction of the cells became infected during this inter- val, as judged by CPE. Thus, it is likely that HHV-6 does not encode a thymidine kinase enzyme which can be assayed under conditions which detect the HSV TK.

DISCUSSION

The effect of PAA on HHVS replication

We have examined the sensitivity of HHV-6 replica- tion to PAA and ACV by using a combination of criteria,

including CPE, infectious virus yields, virus spread as measured by IFAs, and viral DNA replication.

Our study revealed that treatment of the cells with 100 and 300 pg/ml PAA caused a significant inhibition in the production of infectious virus and a substantial reduction in virus spread. The drop in virus titers was faster in the cultures treated with the high PAA dose. We concluded, therefore, that some virus was pro- duced in the presence of PAA at 100 pg/ml. These re- sults could be explained in terms of the inhibitory effect of PAA on viral DNA replication. Thus, treatment of the infected cells with 100 and 300 pg PAA per ml reduced viral DNA synthesis to approximately 24 and 12% of viral DNA synthesis in the untreated infected cells. Inhi- bition of viral DNA replication was specific, inasmuch as the same PAA concentrations did not inhibit host cell DNA replication. On the basis of this finding we have concluded that HHV-6 encodes a PAA-sensitive DNA polymerase. Our study confirms and extends two recent reports by other laboratories. First, using human serum in immunofluorescence assays, Streicher et a/. (1988) have found that HHV-6 was sensitive to the drug phosphonoformate, which in similarity to PAA is known to specifically inhibit the DNA polymerases of several herpesviruses. Second, Bapat et al. (1989) have identi- fied a novel DNA polymerase activity in HHV-6-infected cells. The enzyme displayed sensitivity to PAA and ex- hibited distinct salt requirements for maximal activity.

The effect of ACV on HHV-6 replication

Our studies have shown that viral DNA replication was not significantly inhibited by 20 PM of ACV, nor was there any effect on virus spread and the produc- tion of infectious virus. At 100 FM, ACV reduced the level of viral DNA replication to approximately 40% of that in cells which were infected in the absence of the drug. The production of infectious virus (Fig. 4) as well as virus spread (Fig. 6) were somewhat delayed, as ex- pected from the reduced synthesis of viral DNA.

At 400 puMACV, the results were considerably more complex. Thus, the rate of virus DNA replication was reduced to 149/o of the value in the absence of drugs. This inhibition was close to the 12% residual synthesis observed in the presence of 300 pg PAA per milliliter. Yet, the kinetics of infectious virus production were somewhat different. In the presence of 400 PM of ACV, virus titers did not decrease until the 3rd day p.i. and by Day 5 p.i. they reached lo4 i.u./ml. In contrast, in the culture which was infected in the presence of 300 pg PAA per milliliter, virus titers dropped gradually over 5 days of infection, from the input values of 1 O5 to lower than 10’ i.u./ml. Assuming that this drop in titer repre- sents the minimal rate of inactivation of input virus, we

VIRAL AND CELLULAR DNA REPLICATION IN HHV-6-INFECTED CELLS 209

concluded that even at the high ACV concentration there was virus production, at least during the first 3 days of infection. The immunofluorescence data sup- ported this point, inasmuch as 29% of the cells were infected by the 3rd day p.i., as compared to 13Ob of the cells infected in the presence of 300 pg PAA per milliliter. Virus spread did not continue beyond Day 3 of infection in the presence of 400 PM ACV.

When compared to previously published reports concerning ACV, our results differ from those of Agut era/. (1988) who reported that HHV-6 infection was de- layed by 10 PM ACV and was completely inhibited by 100 PM ACV. The differences in the results could re- flect altered properties of the HHV-6 isolates used in the two studies, or the use of different experimental procedures. Our study agrees with the recent report by Streicher et a/. (1988) that HHV-6 was incompletely inhibited by ACV at 200 and 300 PM.

There are two nonmutually exclusive explanations for our observation that ACV at lower concentrations failed to inhibit HHV-6 DNA replication. First, HHV-6 might lack the TK enzyme necessary for the phosphor- ylation of ACV. The drug may be phosphorylated at low efficiency by the cellular enzyme. At 400 &I, the con- centration of the phosphorylated product is high enough to affect the viral DNA polymerase, whereas the major cell DNA polymerase which is responsible for the synthesis of the SalI-resistant cell DNA band is not inhibited. Support for this hypothesis comes from our data regarding the lack of detectable increase in TK activity in HHV-6 infected cells,

The second explanation is that although the drug is phosphorylated, the viral DNA polymerase is not inhib- ited by the drug at low concentrations. In this regard it is noteworthy that Streicher er al. (1988) reported that HHV-6 was not inhibited by gancyclovir at 80 and 150 pg/ml. Because, unlike ACV, this drug can be phos- phorylated by cellular enzymes, these data suggest that HHV-6 DNA polymerase may be insensitive to nu- cleoside analogs. Furthermore, Bapat eT a/. (1989) have recently shown that a novel DNA polymerase from HHVGnfected cells exhibited a K, value for ACV triphosphate which was significantly higher than those of HSV-1, HSV-2, HCMV, and EBV.

Regardless of the mechanism, it is noteworthy that the effective ACV concentrations required for efficient inhibi- tion of HHV-6 replication are significantly above the re- ported physiological range of ACV plasma concentrations attainable after oral or even intravenous administration (Bridgen et a/., 198 1). Thus, ACV might not be an effective anti HHV-6 drug as judged from our in vitro study.

HHVS-induced shutoff of host DNA replication HHV-6 induced a strong shutoff of host DNA replica-

tion. By 3 days p.i. with 0.1 i.u./cell the inhibition of host

DNA replication was almost complete. Host DNA repli- cation was not inhibited in the cells which were in- fected in the presence of 100 and 300 pg of PAA per milliliter. Nor was it reduced in the cells treated with 100 and 400 PM of ACV. With respect to the shutoff of host DNA replication, HHV-6 resembles HSV, for which it was shown that infection induces the shutoff of cellu- lar DNA synthesis (Fenwick et al., 1979). In the case of HSV, the shutoff of host DNA replication was co- mapped with the shutoff of host protein synthesis (Fen- wick et al., 1979; Kwong and Frenkel, unpublished re- sults). Furthermore, the shutoff of host protein synthe- sis has been shown to reflect the activity of a generalized mRNA destabilizing function, encoded by the virion host shutoff (vhs) gene (Kwong and Frenkel, 1987). It is at present unknown whether HHV-6 pos- sesses these other shutoff functions.

HHV6-induced stress amplification of cell DNA synthesis

As seen in Figs. 7 and 8, HHV-6 infection induced the amplification of a 14-kb BarnHI fragment. This band comigrated with a host DNA band of similar size pres- ent in mock-infected cells treated with PAA or ACV. Smith and Vinograd (1972) observed that treatment of HeLa cells with cycloheximide resulted in the amplifi- cation of small polydisperse circular cytoplasmic DNA molecules. Amplification of DNA sequences has been reported in response to treatment of cells with drugs that inhibit DNA replication, such as methotrexate (Tisty et al., 1982) and hydroxyurea (Brown eta/., 1983) and in response to treatment with chemical carcino- gens (Kleinberger et al., 1986) and other stress condi- tions, such as hypoxia (Rice eta/., 1986). The synthesis of the BamHl band could reflect stress events induced by HHV-6 and/or result from the inhibition of host mac- romolecular synthesis such as inhibition of host DNA replication. It is of interest to note that PAA did not in- hibit the synthesis of the stress BamHl band, implying that the HHV-6 DNA polymerase was not involved. Fur- thermore, recent studies in our laboratory (B. Avidor and N. Frenkel, unpublished results) have shown that the HHV-6-induced replication is cytoplasmic. Studies designed to further characterize the shutoff as well as the inductive functions of HHV-6 functions are cur- rently in progress.

ACKNOWLEDGMENTS

We thank Dr. Carlos Lopez and Ms. Karen Sanderlin (CDC) for the gift of HHV-6 strain 229 and for communicating to us the initial proce- dures of virus propagation. D.D.L. was the recipient of an A.I.R.C. (Associazione ltaliana per la Ricerca sul Cancro) fellowship. Work performed at the University of Kansas was supported by Public

210 DI LLJCA ET AL.

Health Service Grant Al24224 from the National Institutes of Health and a grant from Wesley Foundation of Wichita, Kansas,

REFERENCES ABLMHI, D. V., JOSEPHS, S. F., BUCHBINDER, A., HELLMAN, K., NAKA-

MURA, S., LLANA, T., Lusso, P., KAPLAN, M., DAHLBERG, J., MEMON, S., IMAM, F.. ABLASHI, K. L., MARKHAM, P. D., KRAMARSKY, 6.. KRUEGER, G. R. F., BIBERFELD, P., WONG-STAAL, F., SALAHIJD~IN, S. Z., and GALLO, R. C. (1988). Human B-lymphotrophic virus (hu- man herpesvirus-6). J. Viral. Methods 21, 29-48.

AGUT, H., GUETARD, D., COLLANDRE, H., DAUGUET, C., MONTAGNIER, L., MICLEA, J. M., BAURMANN, H., and GESSAIN, A. (1988). Concomitant infection by human herpesvirus-6, HTLV-1, and HIV-2. Lancer 1, 712.

BALACHANDRAN, N., AMELSE, R. E., ZHOU, W. W., and CHANG, C. K. (1989). Identification of proteins specific for human herpesvirus-6- infected human T cells. J. Viral. 63, 2835-2840.

BAPAT, A. R., BODNER, A. J., TING, R. C. Y., and CHENG. Y. C. (1989). Identification and some properties of a unique DNA polymerase from cells infected with Human B-lymphotropic virus. J. Viral. 63, 1400-1403.

BECKER, W. B., ENGELBRECHT, S., BECKER, M. L. B., PIEK, C., ROBSON, B. A., WOOD, L., and JACOBS, P. (1989). New T-lymphotropic human herpesviruses. Lancer 1,41.

BRIDGEN, D., FIDDIAN, P., ROSLING, A. E., and RAVENSCROFT, T. (1981). Acyclovir-A review of the preclinical and early clinical data of a new antiherpes drug. Antiviral Res. 1, 203-212.

BROWN, P. C., TISTY, T. D., and SCHIMKE, R. T. (1983). Enhancement of methotrexate resistance and dihydrofolate reductase gene am- plification by treatment of mouse 3T6 cells with hydroxyurea. Mol. Cell. Biol. 3, 1097-l 107.

DEISS, L. P., and FRENKEL, N. (1986). Herpes simplex virus amplicon: Cleavage of concatemeric DNA is linked to packaging and involves amplification of the terminally reiterated a sequence. 1. Viral. 57, 933-941.

DORSKY, D. I., and CRUMPACKER, C. S. (1987). Drugs five years later: Acyclovir. Ann. Int. Med. 107, 859-874.

FENWICK, M., MORSE, L. S., and ROIZMAN, B. (1979). Anatomy of her- pes simplexvirus DNA. XI. Apparent clustering of functions affect- ing rapid inhibition of host DNA and protein synthesis. J. Viral. 29, 825-827.

KLEINBERGER, T., ETKIN. S., and LEVI, S. (1986). Carcinogen-mediated methotrexate resistance and dihydrofolate reductase amplifica- tion in Chinese hamster cells. Mol. Cell. Biol. 6, 1958-l 964.

KWONG, A. D., and FRENKEL. N. (1987). Herpes simplex virus-infected cells contain a function(s) that destabilizes both host and viral mRNAs. Proc. Nat/, Acad. Sci. USA 84, 1926-l 930.

KWONG, A. D., KRUPER, J. A., and FRENKEL, N. (1988). Herpes simplex virus virion host shutoff function. J. Viral. 62, 912-921.

LEIDEN, J. M., BUITYAN, R.. and SPEAR, P. G. (1976). Herpes simplex virus gene expression in transformed cells. I. Regulation of the viral thymidine kinase gene in transformed Lcells by products of super- infecting virus. J. Vifol. 20, 413-424.

LEINBACH, S. S., RENO, J. M., LEE, L. F., ISBELL, A. F., and BOEZI, J. A. (1976). Mechanism of phosphonoacetate inhibition of herpesvirus induced DNA polymerase. Biochemistry 15,426-430.

LOCKER, H., and FRENKEL. N. (1979). The BarnHI, Kpnl, and Sal1 re- striction enzyme maps of the DNAs of herpes simplex virus strains Justin and F: Occurrence of heterogeneities in definied regions of the viral DNA. J. Viral. 32, 429-441.

LOPEZ, C., PELLET, P., STEWART, J., GOLDSMITH, C., SANDERLIN, K., BLACK, J., WARFIELD, D., and FEORINO, P. (1988). Characteristics of human herpesvirus-6. J. Infect. Dis. 157, 1271-l 273.

OKUNO, T., TAKAHASHI, K., BALACHANDRA, K., SHIRAKI, K., YAMANISHI, K., TAKAHASHI, M., and BABA, K. (1989). Seroepidemiology of hu- man herpesvirus 6 infection in normal children and adults. J. C/in. Microbial. 27, 651-653.

POST, L. E., MACKEM, S.. and ROIZMAN, B. (1981). Regulation of (Y genes of herpes simplex virus: Expression of chimeric genes pro- duced by fusion of thymidine kinase with (Y gene promoters. Cell 24,555-565.

PURIFOY, D. 1. M., and POWELL, K. L. (1977). Herpes simples virus DNA polymerase as the site of phosphonoacetate sensitivity: Temperature sensitive mutants. J. Viral. 24,470-477.

RICE, G. C., HOY, C., and SCHIMKE, R. (1986). Transient hypoxia en- hances the frequency of diydrofolate reductase gene amplification in Chinese hamster ovary cells. Proc. Nat/. Acad. SC;. USA 83, 5978-5982.

SALAHUDDIN, S. Z., ABLASHI, D. V., MARKHAM, P. D., JOSEPHS. S. F., STURZENEGGER, S., KAPLAN, M., HALLIGAN, G., BIBERFELD, P., WONG- STAAL, F., KRAMARSKY, B., and GALLO. R. C. (1986). Isolation of a new virus, HBLV, in patients with lymphoproliferative disorders. Science 234,596-60 1.

SAXINGER, C., POLESKY, H., EBY, N., GRUFFERMAN, S., MURPHY, R., TEGTMEIR, G., PAREKH, V., MEMON, S., and HUNG, C. (1988). Anti- body reactivity with HBLV (HHV-6) in U.S. populations. J. W-o/. Methods 21, 199-208.

SMITH, C. A., and VINOGRAD, J. (1972). Small polydisperse circular DNA of HeLa cells. 1. Mol. Biol. 69, 163-l 78.

STREICHER, H. Z., HUNG, C. L., ABLASHI, D. V., HELLMAN, K.. SAXINGER, C., FULLEN, J., and SALAHUDDIN, S. Z. (1988). In vitro inhibition of human herpesvirus- by phosphonoformate. 1. viral. Methods 21, 301-304.

TEDDER, R. S., BRIGGS, M., CAMERON, C. H., HONES% R., ROBERTSON, D., and WHITTLE, H. (1987). A novel lymphotropic herpesvirus. Lan- cet 2,390-392.

TISTY, T. D., BROWN, P. C., JOHNSON, R., and SCHIMKE, R. T. (1982). Enhanced frequency of generation of methotrexate resistance and gene amplification in cultured mouse and hamster cell lines. ln “Gene Amplification” (R. T. Schimke, Ed.), pp. 231-238. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

YAMANISHI, K., OKUNO, T., SHIRAKI, K., TAKAHASHI, M., KONDO, T., ASANO, Y., and KURATA, T. (1988). Identification of human herpesvi- rus 6 as a causal agent for exanthem subitum. Lancet 1, 1065- 1067.