66 Clin Pathol 1992;45:616-620

Epstein Barr virus (EBV) encoded small RNAs:Targets for detection by in situ hybridisation witholigonucleotide probes

G Khan, P J Coates, H 0 Kangro, G Slavin

AbstractAims: To develop a rapid, sensitive, andspecific non-isotopic in situ hybridisation(NISH) procedure for the detection ofEpstein-Barr virus in formalin fixed, par-affin wax embedded tissues.Methods: Two low molecular weightRNAs, designated EBER-1 and EBER-2(Epstein-Barr encoded RNA), were used:celis latently infected with EBV secretelarge amounts of EBERs. The methoduses digoxigenin labelled anti-senseoligonucleotides, corresponding tosequences in EBER-1 and EBER-2.Results: The use of these probes, in con-junction with high temperature micro-wave denaturation, ensured that the tech-nique was considerably more sensitivethan other in situ hybridisation tech-niques for detecting EBV. Furthermore,the hybridisation signal was morpholog-ically distinct in that only the nucleus andnot the nucleolus give a positive signal. Nocross-hybridisation was observed withcells infected with other lymphotropicherpes viruses.Conclusion: The sensitivity, simplicity,and rapidity of this technique make itideal for diagnostic use, and for studiesinvestigating the role of this virus inneoplastic disease.

Epstein-Barr virus (EBV) is a large (172kilobase), double stranded DNA virus, des-ignated the y herpes virus. This ubiquitoustransforming virus"2 is known to infect bothepithelial3 and lymphoid cells."2 More than90% of the world population is seropositive forEBV.4

In less than three decades since the discov-ery of EBV in a Burkitt's lymphoma cellculture,5 this virus has been shown to beassociated with a whole range of benign andmalignant human diseases. Typically, EBV hasbeen associated with two geographicallyrestricted malignancies, Burkitt's lymphomaand nasopharyngeal carcinoma,67 and it isknown to be an aetiological agent of infectiousmononucleosis.8 More recently, EBV has beenimplicated in the aetiology of post-transplanta-tion lymphoproliferative disorders,9 Hodgkin'sdisease,'0 brain lymphomas," oral hairy leuco-plakia,'2 andT cell lymphomas.'"The ubiqui-tous nature of EBV and its diverse associationwith human diseases emphasises the need forsensitive, specific, and rapid techniques whichcan detect the presence of the virus at the

cellular level. Several techniques have beenused for showing the presence of EBV, includ-ing serology,'4 Southern blotting,'5 polymer-ase chain reaction'1 16 and in situ hybrid-isation.'0-12 However, of these only in situhybridisation allows the virus to be localised atthe cellular level. In recent years it has beenmodified to give very sensitive results, evenwith non-isotopically labelled probes. Usingdigoxigenin labelled BamHl W fragmentprobe, the internal repeat sequence of EBV,single copies of EBV can be detected ininfected cells.'7 18 The interpretation of theresults, however, can be very difficult if only afew copies of EBV are present, because of thedot pattern of the hybridisation signal.EBV infection can follow one of two major

pathways; either a latent cycle or a productiveone, depending on the cell type infected. In Blymphocytes EBV generally establishes latentinfection during which at least eight virallyencoded proteins are expressed."' In addition,two non-polyadenylated non-coding smallRNAs, referred to as Epstein Barr encodedRNAs (EBER-1 and EBER-2), are also pres-ent."' In latently infected cells EBERs are byfar the most abundant gene transcripts (107copies per cell) 20-22 These small RNAs, 166and 172 nucleotides, respectively, are locatedin the nucleus23 where they are bound tocellular La (lupus antigen) protein20 andEBER associated protein (EAP).24 HenceEBERs exist as ribonucleoproteins, complexedwith at least two cellular proteins, and exhibitstable secondary structure.20 22 24 Despite thehigh abundance of EBERs in infected cells,their function remains elusive.22 23 25The detection of EBERs by in situ hybrid-

isation using cloned EBER-1 riboprobe hasrecently been described and shown to be apromising technique for the detection ofEV 26 27

MethodsPREPARATION OF DNA PROBES

The BamHl W fragment of EBV, which is the3 1 kilobase internal repeat sequence, wascloned into the plasmid vector pAT153 andsubsequently used without purification fromthe vector. For oligonucleotide EBER probes,two sequences of 30 nucleotides each, corre-sponding to positions 90-119 and 82-111 ofEBER-1 and EBER-2, respectively, were selec-ted from their published sequence28 and syn-thesised using the 391 DNA synthesiser(Applied Biosystems). The sequences weredesigned to have high G-C content (60%) to

Departments ofHistopathology andVirology, StBartholomew'sHospital, WestSmithfield, LondonECIA 7BEG KhanP J CoatesH 0 KangroG SlavinCorrespondence to:Dr G KhanAccepted for publication28 November 1991

616

on Septem

ber 7, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.45.7.616 on 1 July 1992. D

ownloaded from

Detection ofEBV using oligonucleotide probes

ensure the formation of stable hybrids withtheir targets. The oligonucleotide sequenceswere also selected to be complementary totarget regions in EBER-1 and EBER-2, whichare not bound to cellular proteins and do notcontain significant secondary structure, there-by facilitating the efficiency of hybridisation. A30 mer oligonucleotide, unrelated to EBERsequence, but having similar G-C content(60%), was also synthesised and labelled as forEBER probes. This was used as a negativecontrol.

All probes were labelled with digoxigenin-1 1-dUTP (Boehringer Mannhein, Lewes, Sus-sex). The BamHl W fragment probe was

labelled by nick translation using a commer-

cially available kit (Gibco, Paisley, Scotland),following the manufacturer's protocol. Thesynthetic oligonucleotide probes were end-labelled by tailing with terminal transferase, as

described.29 Briefly, the labelling reaction was

set up by combining the following compoundson ice: I,ul oligonucleotide (1 ,g/,); 4 jul of 5x concentrated tailing buffer (TRIS-HCI 125mmol/l, pH 6-6, potassium cacodylate 1 mol/l,bovine serum albumin 1-25 mg/ml); 2 julCocl2 (50 mmol/l); 1 jul Dig-dUTP labellingmixture (Dig-dUTP 1 mmol/l, dATP 9 mmol/1, TRIS-HCl 10 mmol/l, pH 7 5). The reactionvolume was increased to 17 pl with steriledistilled water and 3 1ul (55 Units) of terminaltransferase (Stratech Scientific Ltd, Luton,UK) was added. The mixture was incubatedfor 30 minutes at 37°C. The reaction was

terminated by adding 1 pl of 0-2 M EDTA andthe labelled oligonucleotides were purified byrepeated ethanol precipitation in the presenceof 100 ,ug salmon sperm DNA.

Before ethanol precipitation of the labelledoligonucleotide probes, 2-5 jul (125 ng) of eachprobe was aliquoted out for the determinationof the tail length and the labelling efficiency byelectrophoresis through a 20% polyacryamidegel. Gels were stained with ethidium bromide,blotted on to a nitrocellulose membrane, andthe labelled probe detected as described.29

CELL CULTURE ANDMSSUES

Human cytomegalovirus and human herpesvirus 6 (HHV-6) were cultured in humanforeskin fibroblasts and HSB-2 cell line,respectively. B95-8 (EBV producer cell line),Namalwa (Burkitt's lymphoma cell line con-

taining two copies of EBV), EBV transformedlymphoblastoid cell line, and Ramos (EBVnegative cell line) were also grown in cellsuspension culture. The cells were fixed in 10%formol-saline overnight at room temperature,pelleted by centrifugation, and embedded in0-5 ml of 2% agar. The agar pellets were

dissected into small blocks and processed toparaffin wax on an automatic tissue processor,according to standard laboratory procedure.

Formalin fixed, paraffin wax embedded tis-sues were retrieved from the departmentalfiles. These consisted of five cases of undiffer-entiated nasopharyngeal carcinoma, five cases

of Hodgkin's disease, and one case of non-

Hodgkin's lymphoma. Except for two cases ofHodgkin's disease, all other cases were positive

for EBV by polymerase chain reaction (PCR)or NISH or both. 16 17 Sections of normal tonsilwere used as negative controls in addition tofibroblast and HSB-2 cell cultures.

HYBRIDISATIONIn situ hybridisation using BamHl Wfragmentwas carried out, as previously described.'7 Asimilar approach was adopted for EBERhybridisation. In initial experiments a range ofvariables were evaluated for their effects on thedetection ofEBERs. These included stringencyconditions for hybridisation and post-hybrid-isation washes, the temperature of denatura-tion, the degree of proteinase K digestion andthe concentration of probes.

Briefly, 5 um sections on sialinised slideswere heated for one hour at 70°C, dewaxed inxylene, and the endogenous peroxidase activityblocked with 05% hydrogen peroxide in meth-anol. The sections were digested with protei-nase K inTRIS-buffered saline (pH 7 6) for 15minutes at 37°C, dehydrated, and dried. Thelabelled probes were diluted to a concentrationof 0 I pg/ml in hybridisation medium (50%formamide, 5% dextran sulphate, 2 x SSC).Depending on the size of the section, 8-15 pulof the diluted probe was spotted on to thetissue sections and a coverslip placed on top.Probe and target EBER RNA were simul-taneously denatured at about 1 04'C in amicrowave oven (Toshiba ER-665) for eightminutes (one minute at medium setting fol-lowed by seven minutes at low setting).'830The sections were subsequently hybridised at42°C overnight in a hot air oven.The non-specific or unbound probe was

removed by post-hybridisation washes consist-ing of two 10 minute washes in each of thefollowing: 2 x SSC (sodium citrate, sodiumchloride) at room temperature, 0 1 x SSC at55°C, and finally 2 x SSC at room tem-perature. The hybridisation signal was detectedby a three layer ABC-peroxidase technique(Vector Laboratories, Peterborough, England)using mouse anti-digoxin monoclonal anti-body (Sigma, UK; Poole, Dorset). Diamino-benzidine tetrahydrochloride (DAB) was usedas the chromogen. 7

ResultsEVALUATING OPTIMAL CONDITONS FOR EBER

HYBRIDISATIONFrom polyacrylamide gel electrophoresis andthe subsequent blot, the oligonucleotides wereshown to have incorporated dig-dUTP and tohave a tail length in the range of 50-100nucleotides.Stringency: The melting temperature for theoligonucleotide probes was estimated to be60-62°C, according to Maniatis.3' Using thisfigure as a guideline, post-hybridisation washesat various levels of stringency were performed,including 2 x SSC at room temperature; 0 1x SSC at 42°C, 0.1 x SSC at 60°C, and 0-1x SSC at 70°C. From the results of theseexperiments we chose 0-1 x SSC at 550C forthe post-hybridisation washes. At this strin-gency condition, the hybridisation signal was

617

on Septem

ber 7, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.45.7.616 on 1 July 1992. D

ownloaded from

Khan, Coates, Kangro, Slavin

seen only in EBV positive cell lines. At lowerstringency conditions, weak non-specifichybridisation was present in HSB-2 cells,although not in other cell lines.Denaturation: Probe and target EBER RNAwere denatured at 70°C in a hot air oven foreight minutes or at about 104°C in a micro-wave (Toshiba), also for eight minutes. Hightemperature microwave denaturation gave astronger signal.Digestion: Tissue sections were digested at arange of proteinase K concentrations(10-1000 pg/ml). All the tissues examined inthis study required digestion: at concentrationsless than 100 ,ug/ml for optimal EBER signaldetection. This is about one fifth to one tenthof that required for optimal BamHl W probehybridisation. With no digestion, the signal wascompletely abolished for BamHl W probe andconsiderably reduced for EBER detection.Probe concentration: A range of EBER- 1 andEBER-2 probe concentrations were evaluated(10-1000 ng/ml), both individually and incombination. A positive signal could easily beseen at concentrations as low as 10 ng/ml inEBV positive cell lines and nasopharyngealcarcinomas. The combination of the twoEBER probes gave a stronger signal than eitherof the individual probes. The optimal signalwas seen at 100 ng/ml and beyond this theintensity of the signal did not increase. At aconcentration of 1000 ng/ml, weak stainingwas observed in small infiltrating lymphocytesin Hodgkin's disease tissues, although not innasopharyngeal carcinoma. A similar patternof weak staining was also seen with a com-pletely unrelated oligonucleotide used as anegative control. This suggests that this weakstaining in small lymphocytes is non-specific.In keeping with this, the weak staining could beblocked by including 50 ,ug/ml of the unrelatedunlabelled 30 mer in the EBER hybridisationmix.

SPECIFICITY

Having optimised the EBER hybridisationconditions, various EBV positive and negativecell lines were tested to determine the specifici-ty of the technique. Very strong nuclear hybrid-isation signal was seen in the EBV positiveB95-8 cell line (not shown). A weaker, butnevertheless clearly positive signal was alsoseen in Namalwa and EBV positivelymphoblastoid cell lines. Furthermore, thehybridisation signal was exclusively present inthe nucleus, sparing the nucleolus. The inten-sity of the signal varied from cell to cell in thesame section. The EBV negative cell line,Ramos, and sections of normal tonsil wereconsistently negative, as were HSB-2 andfibroblast cell cultures infected with HHV-6and cytomegalovirus, respectively. To deter-mine if the hybridisation signal observed usingEBER probes was specific, a competitiveinhibition assay was set in which the EBERsignal was found to be inhibited if a 1000-foldexcess of unlabelled probe was included in thehybridisation mix. Further confirmation of thespecificity of the EBER oligonucleotide probeswas obtained by treating sections from Hodg-

kin's disease and nasopharyngeal carcinomatissue with RNase'7 before hybridisation. Nosignal was seen in any of the RNase treatedpreparations, while the corresponding un-treated sections were strongly positive(figs 1A and B).

TUMOUR TISSUES

Except for two cases of Hodgkin's disease, allthe cases studied were known to be positive forEBV DNA by PCR, or NISH, or both."6 17The hybridisation signal using the EBERprobes was very intense with a diffuse patternof staining confined to the nucleus but sparingthe nucleolus (figs 1A and 2A). Comparing thesensitivity of EBER hybridisation and theBamHl W probe hybridisation method, theformer was found to be much more sensitiveon all the tissues tested. The hybridisationsignal seen using BamHl W probe was muchmore difficult to discern. These signals are seenas small dots within the nucleus and required ahigh power objective lens for identification (fig2B). With BamHl W probe, a high degree ofdigestion may often be needed for optimalsignal detection and inevitably this results in aloss of tissue morphology. Additionally, inthose cases which were EBV positive, almostall morphologically malignant cells gave apositive signal with the EBER probes, incontrast to BamHl Wprobe, which resulted inonly a proportion of malignant cells giving apositive signal. In both methods the signal wasconfined to cells showing the morphologicalcharacteristics of malignancy and was not seenin the infiltrating lymphocytes. There wasconsiderable variation in the intensity of thehybridisation signal from one case to anotherand even from cell to cell in the same section.Staining seen in cases of nasopharyngeal carci-noma was consistently stronger than that seenin cases of Hodgkin's or non-Hodgkin's lym-phoma.

DiscussionLocalising the presence of EBV to malignantcells is important for determining any possibleaetiological role of the virus in tumorigenesis.Techniques such as Southern blotting and thepolymerase chain reaction are sensitive butlack this basic criterion. In situ hybridisationfulfils this objective, and with EBV it has beenrefined over the years to give very sensitiveresults. The presence of single copies of EBVcan be detected using a non-isotopicallylabelled BamHl W fragment probe on celllines.'7 18 However, there are several draw-backs with this technique. Not only is thepositive signal difficult to discern due to thesmall dot pattern of hybridisation, but thetissue morphology may also be poor as a resultof the high proteinase K digestion that is oftenrequired. Additionally, the preparation of theBamHl W fragment probe is in itself timeconsuming and expensive. Moreover, the sen-sitivity is questionable on archival materialwith both isotopic and non-isotopic meth-ods.18 32Here we report a novel approach for the

618

on Septem

ber 7, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.45.7.616 on 1 July 1992. D

ownloaded from

Detection ofEBV using oligonucleotide probes

Figure I A case of undifferentiated nasopharyngeal carcinoma. (A) Hybridisation with 100 ng/ml of digoxigeninlabelled EBER-1 probe; proteinase K digestion 50 pglml. Very strong nuclear staining, sparing the nucleolus is seen onlyin the morphologically malignant cells. (B) Specificity assay; RNase treatment (200 pglml) before hybridisation with100 nglml of EBER-1 probe.

detection of EBV in paraffin wax embeddedtissues, using EBV encoded small RNAs astargets for digoxigenin labelled synthetic oligo-nucleotide probes. We have shown that thistechnique is significantly more sensitive thanother in situ hybridisation methods used forthe detection of EBV, in terms of intensity ofthe signal, the proportion of cells giving apositive signal, and the ease of interpreting apositive result. The hybridisation pattern ismorphologically distinct, staining only thenucleus while sparing the nucleolus. Lowproteinase K digestion is required, hence tissuemorphology is preserved.A similar approach for the detection of EBV

has recently been described by Wu et al, whohave shown that EBERs are expressed inHodgkin's disease26 and nasopharyngeal carci-noma,27 using in situ hybridisation. In theformer report 3H labelled EBER-1 riboprobewas used while in the latter study both radi-olabelled and digoxigenin labelled EBER-1riboprobe was used. The preparation of ribo-probe by cloning EBER sequences in plasmidsand the subsequent in vitro transcription is anexpensive and time consuming process. Ribo-probes are relatively unstable and requireRNase free equipment when handling. Incontrast, we have used short synthetic oligonu-cleotides to show the expression of EBERs incases of Hodgkin's and non-Hodgkin's lym-phoma, and nasopharyngeal carcinoma. Thepreparation of the oligonucleotide probes issimple, relatively inexpensive, and very rapid.

Although the oligonucleotide probes are lesssensitive in themselves, the extremely highEBER copy number in latently EBV infectedcells compensates for this loss in sensitivity.The overall technique is sensitive and theresults easy to interpret due to the intensehybridisation signal, which is morphologicallycharacteristic of EBER distribution.23 Gen-erally all our positive cases showed a diffusepattern of staining with EBER probes, incontrast to the dot pattern of EBER hybrid-isation observed by others.23 27 This apparentdiscrepancy in the two sets of results may bedue to the small size of the oligonucleotideprobes and the high temperature denaturationused in our protocol, both of which mayfacilitate a greater access for the probes tohybridise to the target EBERs. We found thathigh temperature microwave denaturation atabout 104°C for eight minutes was essential forobtaining an optimal signal. This is not alto-gether surprising because it is well known thatEBERs exhibit extensive intramolecular basepairing and a stable secondary structure.22 28

In the cases of nasopharyngeal carcinomaand Hodgkin's lymphoma examined the EBERhybridisation signal was consistently strongerin the former than in the latter. Furthermore,considerable variation in signal intensity wasobserved from cell to cell in the same section,consistent with reported observations.27 Thesefindings indicate that the EBER copy numbervaries from cell to cell which may correlatewith differing EBV copy number.2021 None

C. .?

Figure 2 A case of Hodgkin's disease. (A) Hybridisation with a combination of EBER probes (100 nglml). Digestion:30 yuglml. Strong diffuse pattern of staining seen in morphologically malignant cells. (B) Hybridisation with 1 pglml ofBamHlW probe. Digestion: 100 uglml. Hybridisation signal is seen as smaUl dots in only the large neoplastic cells (oilimmersion).

619

on Septem

ber 7, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.45.7.616 on 1 July 1992. D

ownloaded from

Khan, Coates, Kangro, Slavin

the less, all the samples gave easily inter-pretable signals, emphasising the general appli-cability of this method for the identification ofcells infected with EBV in paraffin wax pro-

cessed tissue sections. EBERs are by no meansuniversally expressed in all such infected cells.Recently it has been shown that EBERs are notexpressed in permissively infected cells inAIDS associated leucoplakia.33 3 As the EBVcopy number in permissively infected cells isrelatively high, the detection of EBV by in situhybridisation in such cases is not a problemand can easily be done using the BamHl Wprobe or oligonucleotide probe to EBVDNA.35

We thank P Crocker for the photography and M Wilks for theoligonucleotide synthesis. This project was funded by the jointresearch board of St Bartholomew's Hospital.

1 HenleW, DielV, Kohn G, Zur Hausen H, Henle G. Herpes-type virus and chromosome marker in normal leukocytesafter growth with irradiated Burkitt cell. Science1967;157: 1064-5.

2 Pope JH, Horne MK, Scot W. Transformation of fetalhuman leukocytes in vitro by filtrates of a humanleukaemic cell line containing herpes-like virus. IntCancer 1968;3:857-66.

3 Shapiro IM, Volsky DJ. Infection of normal human epith-elial cells by Epstein-Barr virus. Science 1983;219:1225-8.

4 Miller G. Epstein-Barr virus biology, pathogenesis, andmedical aspects. In: Fields BN, Knipe DM, eds. Virology.2nd ed. NewYork: Raven Press, 1990:1921-57.

5 Epstein MA, Achong BG, Barr YM. Virus particles incultured lymphoblasts from Burkitt's lymphoma. Lancet1964;i:702-3.

6 Zur Hausen H, Schutte-Holthausen H, Klein G, et al. EBVDNA in biopsies of Burkitt tumour and anaplasticcarcinoma of the nasopharynx. Nature 1970;228:1056-8.

7 Klein G, Giovanella BC, LindahlT, et al. Direct evidence forthe presence of Epstein-Barr virus DNA and nuclearantigen in malignant epithelial cells from patients withpoorly differentiated carcinoma of the nasopharynx. ProcNadtAcad Sci USA 1974;71:4737-41.

8 Henle G, Henle W, Diehl V. Relation of Burkitt's tumor-associated herpes-type virus to infectious mononucleosis.Proc Nad Acad Sci USA 1968;56:94-101.

9 Randhawa PS, Yousem SA, Paradis IL, et al. The clinicalspectrum, pathology and clonal analysis of Epstein-Barrvirus-associated lymphoproliferative disorders in heart-lung transplant recipients. Am Clin Pathol 1989;92:177-85.

10 Weiss LM, Movahed LA, Warnke RA, Sklar J. Detection ofEpstein-Barr virus genomes in Reed-Stemnberg cells ofHodgkin's disease. N Engl Med 1989;320:502-6.

11 Rouah E, Rogers BB, Wilson DR, Kirkpatrick JB, BuffoneGJ. Demonstration of Epstein-Barr virus in primarycentral nervous system lymphoma by the polymerasechain reaction and in situ hybridization. Hum Pathol1990;21:545-50.

12 Eversole LR, Stone CE, Beckmann AM. Detection of EBVand HPV DNA sequences in oral 'hairy' leukoplakia by insitu hybridization. Med Virol 1988;26:271-7.

13 Jones JF, Shurin M, Abramowsky C. T-cell lymphomascontaining Epstein-Barr viral DNA in patients withchronic Epstein-Barr virus infections. N Engl Med1988;318:733-41.

14 De Schryver A, Friberg S Jr, Klein G, et al. Epstein-Barrvirus-associated antibody patterns in carcinoma of the

post-nasal space. Clin Exp Immunol 1969;5:443-59.15 Su I-J, Hsieh H-C, Un K-H, et al. Aggressive peripheral

T-cell lymphomas containing Epstein-Barr virus DNA: Aclinicopathologic and molecular analysis. Blood 1991;77:799-808.

16 Coates PJ, d'Ardenne AJ, Khan G, Kangro HO, Slavin G.Simplified procedures for applying the polymerase chainreaction to routinely fixed paraffin wax sections. J ClinPathol 199 1;44:115-8.

17 Coates PJ, Mak WP, Slavin G, d'Ardenne AJ. Detection ofsingle copies of Epstein-Barr virus in paraffin waxsections by non-radioactive in situ hybridization. J ClinPathol 1991 ;44:487-9 1.

18 Teo CG, Griffin BE. Visualization of single copies of theEpstein-Barr virus genome by in situ hybridization. AnalBiochem 1990;186:78-85.

19 Kieff E, Liebowitz D. Epstein-Barr virus and its replication.In: Fields BN, Knipe DM, eds. Virology. 2nd ed. NewYork: Raven Press, 1990:1889-920.

20 Lerner MR, Andrews NC, Miller G, Steitz JA. Two smallRNAs encoded by Epstein-Barr virus and complexedwith protein are precipitated by antibodies from patientswith systemic lupus erythematosis. Proc Natl Acad SciUSA 1981;78:805-9.

21 Arrand JR, Rymo L. Characterization of the major Epstein-Barr virus-specific RNA in Burkitt lymphoma-derivedcells. Jf Virol 1982;41:376-89.

22 Glickman JN, Howe JG, Steitz JA. Structural analysis ofEBER- 1 and EBER-2 ribonucleoprotein particles presentin Epstein-Barr virus-infected cells. J Virol 1988;62:902-11.

23 Howe JG, Steitz JA. Localization of Epstein-Barr virus-encoded small RNAs by in situ hybridization. Proc NadAcad Sci USA 1986;83:9006-10.

24 Toczysky DPW, Steitz JA. EAP, a highly conserved cellularprotein associated with Epstein-Barr virus small RNAs(EBERs). EMBOJ3 1991;10:459-66.

25 Bhat RA, Thimmappaya B. Two small RNAs encoded byEpstein-Barr virus can functionally substitute for virus-associated RNAs in the lytic growth of adenovirus 5. ProcNadtAcad Sci USA 1983;80:4789-93.

26 Wu T-C, Mann RB, Charache P, et al. Detection of EBVgene expression in Reed-Stemnberg cells of Hodgkin'sdisease. In J Cancer 1990;46:801-4.

27 WuT-C, Mann RB, Epstein JI, et al. Abundant expression ofEBER 1 small nuclear RNA in nasopharyngeal carcin-oma. Am J Pathol 1991;138:1461-9.

28 Howe JG, Shu M-D. Isolation and characterization of thegenes for two small RNAs of herpesvirus papio and theircomparison with Epstein-Barr virus-encoded EBERRNAs. J Virol 1988;62:2790-8.

29 Schmitz GG,WalterT, Seibel R, Kessler C. Nonradioactivelabeling of oligonucleotides in vitro with the haptendigoxigenin by tailing with terminal transferase. AnalBiochem 1991;192:222-31.

30 Coates PJ, Hall PA, Buttler MG, d'Ardenne AJ. Rapidtechnique ofDNA-DNA in situ hybridization on formalinfixed tissue sections using microwave irradiation. J ClinPathol 1987;40:865-9.

31 Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: alaboratory manual. 2nd edn. Cold Spring Harbor: ColdSpring Harbor Laboratory Press, 1989:11-46.

32 Hamilton-Dutoit SJ, Delecluse HJ, Lenoir G, Pallesen G.Detection of Epstein-Barr virus genomes in AIDS relatedlymphomas: Sensitivity and specificity of in situ hybrid-ization compared with Southern blotting. JT Clin Pathol1991;44:676-80.

33 Young LS, Brooks L, Gilligan K, et al. Epstein-Barr virusreplication and gene expression in oral 'hairy' leukopla-kia. In: Ablashi DV, ed. Epstein-Barr virus and humandiseases 1988. Clifton, New Jersey: Humana Press,1989:329-31.

34 Gilligan K, Rajadurai P, Resnick L, Raab-Traub N. Epstein-Barr virus small nuclear RNAs are not expressed inpermissively infected cells in AIDS-associated leuko-plakia. Proc Nad Acad Sci USA 1990;87:8790-4.

35 Montone KT, Budgeon LR, Brigati DJ. Detection ofEpstein-Barr virus genomes by DNA hybridization with aterminally biotin-labelled synthethic oligonucleotideprobe from the EBV NOT 1 and PST 1 tandem repeatregions. Modern Pathol 1990;3:89-96.

620

on Septem

ber 7, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.45.7.616 on 1 July 1992. D

ownloaded from