Proc. Nail. Acad. Sci. USAVol. 88, pp. 5922-5926, July 1991Medical Sciences

Differentiation between two distinct classes of viruses nowclassified as human herpesvirus 6ERIC C. SCHIRMER*, LINDA S. WYATT*, KoICHI YAMANISHIt, WILLIAM J. RODRIGUEZ*,AND NIZA FRENKEL*§*Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892; tDepartment ofVirology, Research Institute for Microbial Diseases, Osaka University, Suita, Osaka, Japan; and tDepartment of Infectious Diseases, Children's Hospital,Washington, DC 20011

Communicated by Bernard Roizman, April 2, 1991 (received for review February 25, 1991)

ABSTRACT Human herpesvirus 6 (HHV-6) causes exan-them subitum (ES, roseola infantum), a childhood diseasecharacterized by high fever and skin rash. We have analyzedrestriction enzyme cleavage patterns of the DNAs of ES virusisolates from Japan and the United States. The patterns of allthe ES viral DNAs were highly conserved, except for variablesequences within the terminal repeat sequences. They resem-bled closely the restriction enzyme patterns of the Z29 strain ofHHV-6 but were distinct from those of the U1102 strain. Thatall ES isolates were closely related whereas the U1102 patternswere very different suggests that the U1102 strain represents adistinct virus. Moreover, the ES isolates all resembled the Z29strain and not the U1102 strain with respect to reactivity withHHV-6 monoclonal antibodies. These rindings provide evi-dence for the existence of two distinct classes of virusespreviously classified as HHV-6. Whereas the Z29-like virusesare involved in ES infections, the association of the U1102-likeviruses with human disease has yet to be determined.

Human herpesvirus 6 (HHV-6) was initially isolated fromperipheral blood lymphocytes (PBLs) of AIDS patients andpatients with lymphoproliferative disorders (1-6). The viruswas found to be ubiquitous in humans, with seroconversionoccurring during the first 2 years of life (7-10). Recent reportshave shown it to be present also in saliva of healthy adults (7,11-13). HHV-6 is the causative agent of exanthem subitum(ES, roseola infantum), a childhood disease characterized byspiky fever and skin rash (10, 14). Potential association ofHHV-6 with serious disease has been reported in immuno-suppressed patients following liver, heart, and kidney trans-plantations (15-18). Further, it has been suggested that thevirus might be associated with cases ofinfant hepatitis (19, 20).The U1102 strain, which was isolated from an AIDS patient

in Uganda (2), and the Z29 strain, which was isolated from anAIDS patient in Zaire (4), exhibit variations in a number ofproperties. Although both strains can be propagated in freshcord blood lymphocytes (CBLs) and PBLs, they differ intheir ability to replicate in continuous cell lines (2, 21, 22) andin their reactivity with a number of monoclonal antibodies(mAbs) (21) produced by Balachandran et al. (23) againstHHV-6 strain GS (1). The two strains exhibit pronouncedrestriction enzyme polymorphism, although all the U1102and Z29 clones tested cross-hybridized with the heterologousDNAs (24). Finally, restriction enzyme mapping analysesongoing in several laboratories have revealed pronounceddifferences (R. Honess and colleagues, National Institute forMedical Research, Mill Hill, London, personal communica-tion; P. Pellet and colleagues, Centers for Disease Control,personal communication; and R. M. Danovich, E.C.S., andN.F., unpublished results).

HHV-7 has been isolated from CD4' T cells of a healthyindividual (24). While it appears to be related to HHV-6, thetwo viruses exhibit very different restriction enzyme pat-terns. Moreover, hybridization tests using HHV-6 U1102 andZ29 probes revealed either no homology to HHV-7 DNA oronly limited homology, even with large probes, ranging insize from 2 to 12 kilobases (kb) (24).The present study was undertaken to characterize ES

isolates. We found that all the ES isolates resembled theHHV-6 strain Z29 and not the U1102 strain with respect toDNA structure and antigenic reactivity. These results raisequestions regarding the classification of HHV-6 isolates.

MATERIALS AND METHODSVirus Isolation and Propagation. The Z29 strain of HHV-6

was obtained from Carlos Lopez (Centers for Disease Con-trol). It was propagated in our laboratory in human PBLs, asdescribed (25). J JHAN cells and the U1102 strain of HHV-6were obtained from Robert W. Honess (National Institute forMedical Research, Mill Hill, London). The U1102 virus waspropagated in J JHAN cells in RPMI 1640 medium containing10%6 fetal bovine serum. Strains HST, KBT, KSM, KWG, andSUZ were isolated at the University ofOsaka, Japan, from ESpatients during the acute phase of the disease. The isolateswere propagated (14) first in CBLs and were then transferredto the National Institutes of Health, Bethesda, MD, wherethey were used to infect phytohemagglutinin-preactivatedPBLs, for the DNA and immunofluorescence analyses.

Strains VW and AW were isolated at the National Insti-tutes of Health, from ES patients in Washington, DC. PBLsobtained from patients during the acute phase of the diseasewere cultured in RPMI 1640 medium containing 10%o fetalbovine serum, 5 Ag of phytohemagglutinin per ml, and 32units of interleukin 2 (Advanced Biotechnology, SilverSpring, MD) per ml. After the appearance of cytopathiceffect, each of the viruses was propagated in two separateCBL cultures, yielding virus stocks that were propagatedindependently from the patients' PBLs.DNA Analyses. DNA was prepared from infected cells

labeled with [32P]orthophosphate (25). For the hybridiza-tions, infected-cell DNAs were digested with restrictionenzymes, electrophoresed in 0.7% agarose gels, and blottedonto Nytran (Schleicher & Schuell). The probes were labeledin vitro using a 32P-tagged random primer or using digoxi-genin-substituted DNA coupled with alkaline phosphatase-conjugated anti-digoxigenin (Genius system, BoehringerMannheim). The pHD9 probe was a gift of Robert W.Honess. It contained the 8.5-kb HindIII fragment of U1102

Abbreviations: HHV, human herpesvirus; ES, exanthem subitum;CBL, cord blood lymphocyte; PBL, peripheral blood lymphocyte;mAb, monoclonal antibody.§To whom reprint requests should be addressed at: Laboratory ofViral Diseases, National Institutes of Health, Building 4, Bethesda,MD 20892.

5922

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 8,

202

0

Proc. Natl. Acad. Sci. USA 88 (1991) 5923

DNA. The pNF1022 probe contained a 3.9-kb Sal I fragmentof U1102 DNA mapped to the junction between the uniqueand terminal repeat sequences.Immunofluorescence Assay. Acetone-fixed infected cells

were tested by indirect immunofluorescence assay (21) usingmAbs derived and characterized by Balachandran et al. (23)for the GS strain ofHHV-6 (1). The mAbs were obtained fromN. Balachandran (University of Kansas, Medical Center,Kansas City, KS).

RESULTSAnalyses of Viral DNA from Five Japanese ES Isolates. First

we examined the DNAs of five ES isolates from Japan.32P-labeled DNAs were prepared from cells infected with theES virus isolates, two variants (A and B) segregated from theHHV-6 strain Z29, the U1102 strain of HHV-6, and the RKstrain of HHV-7. Restriction enzyme patterns of the DNAsare exemplified in Figs. 1 and 2 and can be summarized asfollows. (i) The cleavage patterns of the DNAs from the ESvirus isolates were different from those of HHV-7 DNA. Asbefore (24), HHV-6 and HHV-7 DNAs exhibited pronounceddifferences in their restriction patterns. For example, in theSal I and Sst II patterns shown in Figs. 1 and 2, HHV-7 DNAgenerated few large fragments whereas digestion of the ESDNAs or Z29 DNA or U1102 DNA yielded multiple smallerfragments. (ii) The restriction patterns of the five ES isolatesclosely resembled those of the Z29 strain whereas very fewor no DNA fragments of the five ES isolates comigrated withDNA fragments of the U1102 strain ofHHV-6. (iii) While therestriction patterns of the ES isolates and Z29 DNA weregenerally similar, two types of variations were noted. Thefirst corresponded to variability in a single strain reflectingthe loss or gain of a fragment in that strain. For example, inthe HindIII patterns, a 4-kb fragment was present in all ESisolates except strain HST. Conversely, a 2-kb fragment waspresent in the Z29 strain but was absent from the ES viruspatterns (Fig. 1, asterisks).The second type of variations reflected the presence of

hypervariable regions in the HHV-6 genome. This hypervari-ability was clearly apparent in the fragments denoted heter-

Eco RI

1|m | =

I_

-*'

...~ ~ ~ ~ ~ ~ .j.

1 2 6 8 9 10 11

-36.0

- 16.7

- . 75

- 5.7

- 3.9

- 1.6

15 17 18

FIG. 2. EcoRI and Sst II digests of ES isolates from Japan, theA and B variants of HHV-6 strain Z29, HHV-6 strain U1102, andHHV-7 strain RK. Het denotes the fragments showing hypervari-ability.

ogeneous (Het) in Figs. 1 and 2, as well as in the Kpn I, DraI, Pst I, BamHI, Bgl II, and Ssp I patterns (data not shown).The sizes of these fragments varied in different strains andthey differed also in the A and B variants of the Z29 virusstock. In addition, the Het fragments exhibited variable

Sal I Xba I Hind IIIN~~~~~~~~~

F 2;2Z29 2I O 19 1 F Z29 = r:mAB

nAIB

Im (n3cII iKm fI I i.

0 M * -e 1e i0 Il ge |e cn D

- 20.0

- 13.3

* 8.5- 7.3

- 4.8N1 4.0 FIG. 1. Sal I, Xba I, and HindIII digests of ES

isolates from Japan, the A and B variants ofHHV-6strain Z29, HHV-6 strain U1102, and HHV-7 strain

i4* " -2.8 RK. Het denotes the hypervariability in the HindIIIpatterns. Fragment sizes (kb) shown refer to Z29HindIII fragments in lane 25. These sizes were

-2.0 obtained from a separate gel in which HHV-6 (Z29)digested DNA was analyzed along with Bgl II andBamHI digests of herpes simplex virus 1 (Justin)DNA, serving as size markers (data not shown).The arrow denotes the 8.5-kb HindIII fragment ofU1102 DNA. The 4.0- and 2.0-kb fragments vari-able in the HST and Z29 patterns are denoted with

24 26 27 asterisks.

Het 060 ii

4ft a .. ._m _.40*

_, I

I q I

19 4 s

*.. V.

196 8 9 10

.Jb

.415 1718

r - .

Medical Sciences: Schirmer et al.

I.

.1j..,

0

a

Dow

nloa

ded

by g

uest

on

Dec

embe

r 8,

202

0

5924 Medical Sciences: Schirmer et al.

relative molarities in different DNA preparations of the samestrain. These fragments map within the terminal repeats, andthe variable molarities reflect different proportions of con-catemeric and unit-length genomes (unpublished data).

Analyses of ES Isolates from the United States. To testwhether the close resemblance of the ES isolates reflected theprevalence of certain HHV-6 strains in Japan, we analyzedtwo additional ES isolates from Washington, DC. To ascertainthat the isolates represented viruses in the original ES samplesrather than passenger viruses resident in CBL carrier cultures,each isolate was propagated in two different CBL lineages,generating two independently propagated viral stocks. Viruspropagation in fresh CBLs was necessary because the twoisolates failed to propagate in continuous cell lines.For the DNA analyses, CBL were infected with the

duplicate stocks (I and II) of the ES isolates from the UnitedStates, the KWG ES strain from Japan, the HHV-6 strainsZ29 and U1102, and the HHV-7 strain RK. The results areexemplified in Fig. 3 and can be summarized as follows. (i)The cleavage patterns were not affected by virus propagationin CBLs and identical patterns were observed with theduplicate virus stocks (I and II). (ii) The patterns of theAmerican ES isolates were very similar to the Japanese ESisolates. They also resembled the Z29 patterns and not theU1102 patterns. (iii) The two virus strains from the UnitedStates exhibited variability in the Het fragments, in similaritywith the virus isolates from Japan. The variability appearedto be an inherent property of the strain rather than to havearisen during virus propagation in CBLs.

Hybridizations with Probes Representing Unique and Reit-erated Sequences. Unlabeled DNAs were next analyzed byblot hybridizations using two probes chosen to elucidate thevariability of HHV-6 strains and to confirm their distinctionfrom HHV-7. The pHD9 probe, containing an 8.5-kb HindIIIfragment of U1102, represents unique sequences of HHV-6DNA and, as expected, it hybridized to an 8.5-kb HindIllfragment of U1102 DNA (Fig. 4, lane 10). In contrast, thepHD9 probe hybridized to two HindIII fragments (5.4 and 20kb) of Z29 and the seven ES isolates. The pHD9 sequencesdid not hybridize to HHV-7 DNA (Fig. 4, lane 11) confirmingour previous observation that this probe was specific forHHV-6 DNA (24).

The second probe employed in the hybridization analyseswas pNF1022 containing a 3.9-kb Sal I fragment of U1102DNA. This probe maps at the boundary of the left terminalrepeat and unique sequences. It hybridizes to fragmentsarising from the right and left genomic termini, as well as tofragments arising from junctions of HHV-6 DNA concate-mers (unpublished results). As seen in Fig. 4, this probehybridized to the variable EcoRI Het fragments in the 7 ESisolates, Z29 A and B variants, and the U1102 virus. It alsohybridized to HHV-7 DNA, confirming that HHV-7 containssequences homologous to this probe, as previously shown(24). We conclude that the unique sequences contained in thepHD9 insert represented relatively nonvariant ES fragments,whereas sequences of the 3.9-kb insert of pNF1022 wereoverlapping in part with the hypervariable fragments.

Antigenic Relatedness of the HHV-6 ES Isolates. Cellsinfected with the five ES isolates from Japan, the two ESisolates from the United States, and the HHV-6 strains Z29and U1102 were fixed with acetone and tested by indirectimmunofluorescence assays using mAbs that were derivedand characterized by Balachandran et al. (23). All the mAbsreacted with the U1102-infected cells, as previously reported(21), whereas the mAb 4A6, directed against the 180-kDaprotein p180, and the 4 mAbs 2D6, 13D6, 2D4, and 3B5,directed against an 82- to 105-kDa glycoprotein, gp82-105(23), did not react with the cells infected with the ES isolatesand the Z29 strain. The ES isolates and the Z29 straininteracted with the remaining four mAbs, including 6A5D5and 7A2, specific for distinct glycoproteins; 12B3G4, specificfor the major capsid antigen; and 9A5D12, which recognizes41- and 110-kDa proteins (23). Thus, the seven ES strainsresembled the Z29 strain and not the U1102 strain withrespect to their interactions with the mAbs (Table 1).

DISCUSSIONThese studies revealed significant differences between theES/Z29 and the U1102 strains. These results form the basisfor the reevaluation of the relatedness of the Z29-like andU1102-like strains and their classification as a single group ofviruses. In this context it is useful to review current infor-mation regarding the properties of these HHV-6 strains.Although the GS strain, the first HHV-6 isolate (1), was not

- 2.0 FIG. 3. Sal I, Sst II, EcoRI, andHindill digests of the ES isolates VWand AW from the United States (I andII were independently propagated), theKWG ES isolate from Japan, HHV-6strains Z29 and U1102, and HHV-7strain RK.

Proc. Nat'L Acad Sci. USA 88 (1991)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 8,

202

0

Proc. Natl. Acad. Sci. USA 88 (1991) 5925

Hindill pHD9II I -

2 a N Z29

20.0 - 40- - %' - am 4

EcoRi pNF1022.1. . . . . ...

C, . 4 -

4,~~~~~~~~4

8.5-^ ..

5.4- _ _ _

f

11 12 22

- 27.1- 18.3

- 10.8

- 7.9

- 5.8

FIG. 4. Hybridization of the U1102-derived- 4.4 probes pHD9 and pNF1022 (which contains se-

quences from the direct repeat of HHV-6) ontoHindIlI and EcoRI digest patterns, respectively, ofES isolates from the United States (VW and AW)and Japan (HST, KBT, KSM, KWG, SUZ), the A

- 2.7 and B variants of HHV-6 strain Z29, HHV-6 strainU1102, and HHV-7 strain RK. The pHD9 probe didnot hybridize to HHV-7 DNA (lane 11) whereas thepNF1022 probe did (lane 22). The same amount ofHHV-7 infected-cell DNA was loaded in both lanes,and by ethidium bromide staining both lanes con-

- 1.4 tained sufficient DNA for hybridizations equivalentto DNAs in the HHV-6 lanes. MW, "molecularweight" markers (sizes in kilobases at right).

included in the analyses, the criteria listed below indicate thatGS is closely related to U1102.

Variations in Growth Properties. Although the Z29 andU1102 strains can be efficiently propagated in fresh CBLs orPBLs, they differ in their ability to replicate in continuous celllines. Thus, U1102 as well as the GS strain can be propagatedin continuous cell lines, whereas Z29 is more restrictive (2,21, 22, 26). It remains to be seen whether this reflects inherentdifferences in growth properties of the viruses or the selec-tion of variants capable of growth in the cell lines.

Variations in Viral DNA. The full assessment of the geneticrelatedness ofZ29 and U1102 awaits the complete nucleotidesequencing of their genomes. At present, restriction enzymepatterns offer a valuable tool in studies to distinguish the twostrains molecularly and epidemiologically, as has been thecase in studies of the classification and epidemiology of otherherpesviruses (e.g., refs. 27 and 28). In our studies ofHHV-7(24), we found that although Z29 and U1102 exhibited re-striction enzyme polymorphism, they were closely relatedinasmuch as clones generated from either strain cross-hybridized with the heterologous strain. By comparison,large clones of the U1102 or Z29 DNAs exhibited either nohomology or only partial homology to HHV-7 DNA, definingthis virus as a separate herpesvirus. Further studies (L.S.W.,E.C.S., R. M. Danovich, and N.F., unpublished results)

have shown that HHV-6 and HHV-7 are immunologically andmolecularly distinct. In the present study we have- comparedseven ES isolates from Japan and the United States with theZ29 and U1102 strains. Our analysis confirmed and extendedprevious reports by Pellett et al. (29) and Yamanishi et al. (13)that ES strains from Japan were similar to Z29. Especiallystriking, however, was the observation that all ES/Z29isolates were dlosely related, whereas cleavage of the U1102DNA yielded many fragments that did not comigrate with theZ29-like virus bands. We conclude that the differences be-tween the Z29 and U1102 strains cannot be explained bypronounced restriction enzyme polymorphism of HHV-6strains, inasmuch as the ES/Z29 HHV-6 strains were wellconserved, yet the U1102 strain was clearly distinct.

Heterogeneity in the fragments representing the terminalsequences of variants present in the Z29 virus stock was firstobserved by Pellett et al. (as quoted in ref. 30). In our ownstudies we have segregated such Z29 variants and havecharacterized structural features responsible for these vari-ations (unpublished results). Several elements contribute tothe heterogeneity: (i) stable and inherent variations in the sizeof the terminal repeats of different virus strains; (ii) differentratios of cleaved and concatemeric genomes present in dif-ferent preparations of viral DNA within a given strain; and

Table 1. Reactivity of HHV-6 mAbs with cells infected by various HHV-6 strains

Reacting Reactivity with infected cells

mAb protein* HST KBT KSM KWG SUZ VW AW Z29 U1102

2D6 gp82-105 0 0 0 0 0 0 0 0 +13D6 gp82-105 0 0 0 0 0 ND ND 0 +2D4 gp82-105 0 0 0 0 0 ND ND 0 +3B5 gp82-105 0 0 0 0 0 ND ND 0 +6A5D5 gpll6,-64,-54 + + + + + + + + +7A2 gp102 + + + + + + + + +9A5D12 p41,-110 + + + + + + + + +12B3G4 p135 + + + + + + + + +4A6 p180 0 0 0 0 0 0 0 0 +

The Z29 and U1102 strains of HHV-6 were tested with ascitic fluids at dilutions of 1:10, 1:100, and 1:1000. The remaining HHV-6 strains weretested with the following dilutions: 1:100 for 2D6, 13D6, 2D4, 3B5, 6A5D5, 12B3G4, and 4A6; 1:1000 for 7A2 and 9A5D12. +, Reactivity; 0,no reactivity; ND, not done.*Ref. 23.

Medical Sciences: Schirmer et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 8,

202

0

5926 Medical Sciences: Schirmer et al.

(iii) additional diffuse low-molarity bands whose nature isunknown.

Antigenic Relatedness and Epidemiology of the Z29-Like andU1102-Like Strains. Assessment of the prevalence of Z29-likeand U1102-like viruses awaits the derivation of reagentswhich enable unambiguous distinction between these strains.Our study has suggested the utility of the pHD9 probe ofU1102, which was derived by Robert W. Honess and co-workers, or its equivalent clone pZVH14, derived by Josephset al. (31) from the GS strain, as a potential marker for theclassification of HHV-6-related isolates. Thus, the pHD9probe is specific for HHV-6 and had no detectable homologyin the HHV-7 genome (24). It hybridized to a single HindIIIfragment (8.5 kb) of U1102 DNA but to two HindIII frag-ments (5.4 and 20 kb) in the ES/Z29 isolates. In fact,variations in restriction enzyme patterns of viral genomeshave been previously noted in studies employing the 8.5-kbpZVH14 hybridization (6,32-34). The entire genome patternshave not been investigated in these studies. It remains to beseen whether the presence of an 8.5-kb HindIll band or 5.4-and 20-kb fragments can be taken as a universal characteristicof the U1102 and Z29/ES like viruses. Furthermore, themAbs derived by Balachandran et al. (23) can also be used totest the type of HHV-6, as clearly reviewed in Table 1.

Z29-like strains have been isolated from AIDS patients,including the Z29 virus itself, and ES patients (refs. 13 and 29and this study). The ES isolates obtained by Kikuta et al. (32)were most likely Z29-like, inasmuch as HindIII digestion ofDNAs of these viruses yielded two fragments that hybridizedwith pZVH14 rather than a single fragment. A number ofstrains recently isolated from healthy individuals, most likelyby reactivation from latency in vitro, all resembled the Z29virus (N.F., G. Katsafanas, E.C.S., and L.S.W., unpub-lished data). It appears that the viruses infecting youngchildren are Z29-like and that these viruses persist in adultsto be potentially reactivated later. It is unclear whetherinfection with Z29-like virus precludes infection with theU1102-like strains and whether the high prevalence of anti-bodies in the human population represents both viruses. Infact, the prevalence of U1102-like virus infection in thehuman population is as yet undetermined and further work isrequired to derive specific reagents that will allow serologicaldistinction between the Z29- and U1102-like strains. Finally,further consideration should be given to the reclassificationof HHV-6 strains; a proposal to this effect has been submittedto the Herpesvirus Study Group of the International Com-mittee for the Taxonomy of Viruses.This paper is dedicated to the memory of Robert W. Honess

(National Institute for Medical Research, Mill Hill, London). Wewish to acknowledge his numerous contributions to our studies oftheT-lymphotropic herpesviruses. We thank N. Balachandran (Univer-sity of Kansas Medical Center) for the gift ofmAbs. We thank GeorgeKatsafanas for his skillful technical assistance. We thank Robert M.Danovich for helpful discussions.

1. 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. & Gallo, R. C. (1986)Science 234, 596-601.

2. Downing, R. G., Sewankambo, N., Serwadda, D., Honess, R.,Crawford, D., Jarrett, R. & Griffin, B. E. (1987) Lancet ii, 390.

3. Tedder, R. S., Briggs, M., Cameron, C. H., Honess, R., Rob-ertson, D. & Whittle, H. (1987) Lancet ii, 390-392.

4. Lopez, C., Pellett, P., Stewart, J., Goldsmith, C., Sanderlin,K., Black, J., Warfield, D. & Feorino, P. (1988) J. Infect. Dis.157, 1271-1273.

5. Agut, H., Guetard, D., Collandre, H., Dauguet, C., Montag-nier, L., Miclea, J.-M., Baurmann, H. & Gessain, A. (1988)Lancet i, 712.

6. Becker, W. B., Engelbrecht, S., Becker, M. L. B., Piek, C.,Robson, B. A., Wood, L. & Jacobs, P. (1989) Lancet i, 41.

7. Levy, J. A., Ferro, F., Greenspan, D. & Lennette, E. T. (1990)Lancet i, 1047-1050.

8. Saxinger, C., Polesky, H., Eby, N., Grufferman, S., Murphy,R., Tegtmeir, G., Parekh, V., Memon, S. & Hung, C. (1988) J.Virol. Methods 21, 199-208.

9. Okuno, T., Takahashi, K., Balachandra, K., Shiraki, K.,Yamanishi, K., Takahashi, M. & Baba, K. (1989) J. Clin.Microbiol. 27, 651-653.

10. Ueda, K., Kusuhara, K., Hirose, M., Okada, K., Miyazaki, C.,Tokugawa, K., Nakayama, M. & Yamanishi, K. (1989) J.Infect. Dis. 159, 750-752.

11. Pietroboni, G. R., Harnett, G. B., Bucens, M. R. & Honess,R. W. (1988) Lancet i, 1059.

12. Harnett, G. B., Farr, T. J., Pietroboni, G. R. & Bucens, M. R.(1990) J. Med. Virol. 30, 128-130.

13. Yamanishi, K., Kondo, T., Kondo, K., Hayakawa, Y., Kido,S., Takahashi, K. & Takahashi, M. (1990) in Immunobiologyand Prophylaxis ofHuman Herpesvirus Infections, eds. Lopez,C., Mori, R., Roizman, B. & Whitley, R. J. (Plenum, NewYork), pp. 29-37.

14. Yamanishi, K., Okuno, T., Shiraki, K., Takahashi, M., Kondo,T., Asano, Y. & Kurata, T. (1988) Lancet i, 1065-1067.

15. Kurata, T., Iwasaki, T., Sata, T., Wakabayashi, T., Yamagu-chi, K., Okuno, T., Yamanishi, K. & Takei, Y. (1990) inImmunobiology and Prophylaxis ofHuman Herpesvirus Infec-tions, eds. Lopez, C., Mori, R., Roizman, B. & Whitley, R. J.(Plenum, New York), pp. 39-47.

16. Okuno, T., Higashi, K., Shiraki, K., Yamanishi, K., Taka-hashi, M., Kokado, Y., Ishibashi, M., Takahara, S., Sonoda,T., Tanaka, K., Baba, K., Yabuuchi, H. & Kurata, T. (1990)Transplantation 49, 519-522.

17. Irving, W. L., Cunningham, A. L., Keogh, A. & Chapman,J. R. (1988) Lancet ii, 630-631.

18. Ward, K. N., Gray, J. J. & Efstathiou, S. (1989) J. Med. Virol.28, 69-72.

19. Tajiri, H., Nose, O., Baba, K. & Okada, S. (1990) Lancet i, 863.20. Asano, Y., Yoshikawa, T., Suga, S., Yazaki, T., Kondo, K. &

Yamanishi, K. (1990) Lancet i, 862-863.21. Wyatt, L. S., Balachandran, N. & Frenkel, N. (1990) J. Infect.

Dis. 162, 852-857.22. Black, J. B., Sanderlin, K. C., Goldsmith, C. S., Gary, H. E.,

Lopez, C. & Pellett, P. E. (1989) J. Virol. Methods 26, 133-146.23. Balachandran, N., Amelse, R. E., Zhou, W. W. & Chang,

C. K. (1989) J. Virol. 63, 2835-2840.24. Frenkel, N., Schirmer, E. C., Wyatt, L. S., Katsafanas, G.,

Roffman, E., Danovich, R. M. & June, C. H. (1990) Proc. Natl.Acad. Sci. USA 87, 748-752.

25. DiLuca, D., Katsafanas, G., Schirmer, E. C., Balachandran,N. & Frenkel, N. (1989) Virology 175, 199-210.

26. Ablashi, D. V., Lusso, P., Hung, C.-L., Salahuddin, S. Z.,Josephs, S. F., Llana, T., Kramarsky, B., Biberfeld, P., Mark-ham, P. D. & Gallo, R. C. (1988) Int. J. Cancer 42, 787-791.

27. Buchman, T. G., Simpson, T., Nosal, C. & Roizman, B. (1980)Ann. N. Y. Acad. Sci. 354, 279-290.

28. Roizman, B. & Tognon, M. (1983) in Current Topics in Micro-biology and Immunology, eds. Cooper, M., Hofschneider,P. H., Koprowski, H., Melchers, F., Rott, R., Schweiger,H. G., Vogt, P. & Zinkernagel, R. (Springer, Berlin), Vol. 104,pp. 273-286.

29. Pellett, P. E., Lindquester, G. J., Feorino, P. & Lopez, C.(1990) in Immunobiology and Prophylaxis ofHuman Herpes-virus Infections, eds. Lopez, C., Mori, R., Roizman, B. &Whitley, R. J. (Plenum, New York), pp. 9-18.

30. Lopez, C. & Honess, R. W. (1990) in Virology, eds. Fields,B. N. & Knipe, D. M. (Raven, New York), pp. 2055-2062.

31. Josephs, S. F., Salahuddin, S. Z., Ablashi, D. V., Schachter,F., Wong-Staal, F. & Gallo, R. C. (1986) Science 234, 601-603.

32. Kikuta, H., Lu, H., Matsumoto, S., Josephs, S. F. & Gallo,R. C. (1989) J. Infect. Dis. 160, 550-551.

33. Jarrett, R. F., Gallagher, A., Gledhill, S., Jones, M. D., Teo,I. & Griffin, B. E. (1989) Lancet i, 448-449.

34. Josephs, S. F., Ablashi, D. V., Salahuddin, S. Z., Kramarsky,B., Franza, B. R., Pellett, P., Buchbinder, A., Memon, S.,Wong-Staal, F. & Gallo, R. C. (1988) J. Virol. Methods 21,179-190.

Proc. Natl. Acad Sci. USA 88 (1991)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 8,

202

0