Reprinted from Proc. Nat. Acad. Sac. USA Vol. 69, NO. 7, pp. 1727-1731, July 1972 Viral-Related RNA in Hodgkins' Disease and Other Human Lymphomas (DNA-RNA hybridization/Rauscher murine leukemia virus/cancer/neoplasia/tumor) R. HEHLMANN, D. KUFE, AND S. SPIEGELMAN Institute of Cancer Research, College of Physicians and Surgeons, Columbia University, New York, N.Y. 10032 Contributed by S. Spiegelman, April 24, 1972 ABSTRACT Molecular hybridization with radioactively labeled DNA complementary to the RNA of the Rauscher leukemia virus was used to probe for homologous RNA in human lymphomas. 22 of 32 specimens contained RNA possessing homology to the RNA of the mouse leukemia virus, but not to that of the unrelated viruses causing mammary tumors in mice or myeloblastosis in chickens. Normal adult and fetal tissues failed to show significant levels of the leukemia-specific RNA. It appears that human lymphomas contain RNA sequences homologous to those found in a viral agent known to cause leukemia and lym- phomas in an experimental animal. The fact that human leukemias and sarcomas also contain this type of RNA further emphasizes a remarkable similarity between the corresponding neoplasias of murine and human origin. We have used molecular hybridization to detect virus- specific RNA in tumors (l), and have found that corre- sponding neoplasias of murine and human origin exhibit remarkable similarities. Thus, human breast carcinomas contained (2) RNA possessing sequence homology to that of mouse mammary tumor virus (MMTV). This type of RNA was unique to the malignant adeno- and medullary-car- cinomas, being undetectable in normal breast tissue and in such benign pathologies as fibrocystic disease and fibroade- noma. In keeping with the known unrelatedness of the leuke- mogenic and mammary tumor viruses, we found that breast cancer RNA did not hydridize to DNA complementary to the RNA of Rauscher leukemia virus (RLV). Finally, and most compelling, was the demonstration that human leukemic cells (3) and human sarcomas (4) both contain RNA showing homology to that of Rauscher leukemia virus, and not to that of mouse mammary tumor virus. It was of obvious interest to pursue this intriguing concordance between the neoplasias of mice and men. From the viewpoint of etiology and cellular pathology, lymphomas in mice are linked to the leukemias and sarcomas. In addition, it may be noted that some human lymphomas are accompanied by the clinical appearance of a peripheral leukemia. In any event, if human neoplasias parallel those observed in mice it might be expected that human lymphomas would, like the leukemias and sarcomas, contain RNA uniquely homologous to that of a mouse leukemia agent. We report here a confirmation of this expectation. Human lymphomas, including Hodgkins' disease, lymphosarcomas, and reticulum cell sarcomas contain RNA exhibiting homology to the RNA of Rauscher leukemia virus, but not to the un- related RNAs of either mouse mammary tumor virus or of avian myeloblastosis virus (AhIV) . Abbreviations: RLV, Rauscher (murine) leukemia virus; pRNA, cytoplasmic (polysomal) RNA; AMV, avian myelo- blastosis virus; MMTV, mouse mammary tumor virus. METHODS AND MATERIALS Preparation of Nucleic Acids. The preparation of labeled DNA is exemplified in the case of RLV-13H]DNA. A 1-ml reaction mixture, containing 100 pg of viral protein prepared from virus purified from plasma (5), 50 mM Tris.HC1 (pH 8.3), 40 mM KC1, 6 mM MgC12, 2.5 mM dithiothreitol, 0.00125% Nonidet (NP-40), 100 mM (each) of dGTP, dATP, and dCTP, and 5 X lo4 pmol of [3H]TTP (8000 cpm/pmol), was incubated at 37" for 180 min. After the addition of sodium dodecyl sulphate to 0.5%, the nucleic acid was extracted with an equal volume of phenol-cresol, and the [3H]DNA product was purified by passage through a Sephadex G-50 column, followed by incubation in 0.5 M NaOH at 43" for 24 hr to remove viral RNA. MMTV-[3H]DNA and AMV- [aH]DNA were prepared in a similar manner. For the preparation of cytoplasmic RNA (designated as pRNA), the tumors were disrupted with a Silverson homoge- nizer at 4" in 2 volumes of 5% sucrose in TNM buffer (0.01 M Tris.HC1, pH 7.4-0.15 M NaC1-2 mM MgC12). The suspension was centrifuged at 20,000 X g for 15 min at 0" to remove nuclei and cell debris. The supernatant fluid was layered on 20 ml of 25% sucrose containing 200 pg of poly- vinyl sulfate per ml and centrifuged at 4" for 180 min at 180,000 X g in a 60-Ti rotor (Spinco). The pellet was re- suspended in TNM buffer containing 0.5% sodium dodecyl sulfate and the RNA was extracted twice with an equal volume of cresol-phenyl (pH 8.0). The aqueous phase was adjusted to 0.4 M NaC1, and the nucleic acid was precipitated with two volumes of ethanol. The resulting pRNA precipitate was dissolved in minimum volume of 50% formamide-3 mM EDTA. Annealing Reactions Between pRNA and [3H]DNA. Im- mediately before use, [3H]DNA was denatured by incubation for 10 min at 80" in 75% formamide, followed by quick chilling at 0". The annealing mixture usually contained 350 pg of pRNA and 2000 cpm of [3H]DNA in 60 pl of 0.4 M NaC1, adjusted to 50% formamide and pH 7.4. Annealing was for 18 hr at 37", after which 10 ml of Cs2S04 at 50% saturation (density: 1.52 g/ml) was added; the resulting solu- tion was centrifuged at 44,000 rpm for 60 hr at 15" in a 50-Ti rotor (Spinco). Fractions (0.4 nil) were collected through a needle inserted in the bottom of the tube and assayed (3) for acid-precipitable (10% trichloroacetic acid) radioactivity. Pitfalls to be Guarded Against. The experiments described here and in our previous studies on breast cancer (2), the leukemias (3) and the sarcomas (4) were designed to detect in the neighborhood of 0.1 ng or less of complementary RNA in the pRNA preparations being tested. The sensitivity of the method is therefore being pushed towards its limit. Under the 1727 1728 Medical Sciences: Hehlmann et al. A Fraction Number C Proc. Nut. Acad. Sci. USA 69 (197.2) B b x E 0 E D DNA + 10 20 30 Fraction Number FIG. 1. (A-C). CszSOa density profiles of RLV-[3H]DNA hybridized to pRNA obtained from human lymphomas. Polysomal RNA was isolated from tissues of patients with Hodgkins' lymphoma (A), reticulosarcoma (B), and lymphosarcoma (C). 300 pg of polysomal RNA was hybridized to RLV-[3H]DNA in 60-pl volumes, and the reactions were analyzed by Cs2S04 density centrifugation. (D) Comparison of hybridization reactions between RLV-[3H] DNA and various pItNA concentrations of a spleen from a patient with Hodgkins' lym- phoma (.--e) and of a normal human spleen (0--0). The individual annealing reactions were analyzed by CsZSO4 density centrifuga- tion, and the yo DNA hybridized was determined by the counts/lO min of [3H]DNA (corrected for background) banding in the RNA region (between densities 1.62 and 1.69) of the gradients. circumstances, it is necessary to exercise extreme care in the preparation of both the pRNA and the t3H]DNA used as a probe. Complete removal of the contaminating protein from the pRNA is necessary to avoid nonspecific trapping of t3H]- DNA in nonrelevant portions of the density gradient. To insure the interpretability of the results, it is imperative that every purified [3H]DNA preparation used be monitored for adequacy by being subjected to the following tests: (a) It must be shown to band cleanly in the appropriate DNA region of the density gradient to insure that it is free of con- taminating viral RNA; (b) It must be extensively hybridizable . to the homologous viral RNA used as the template in the original synthesis. This can exceed 90% in the presence of excess viral RNA if actinomycin D, at a concentration of 100 pg/ml, is included to prevent DNA-directed DNA synthesis, a possibility particularly prevalent in crude enzyme prepara- tions; and (6) The [3H]DNA product should not hybridize to unrelated viral RNA or to RNA from normal tissues. We customarily use the RNAs from AMV, RLV, and MMTV as test objects for cross-hybridizability checks. A suitable synthetic DNA complementary to any one of them will not hybridize to either of the other two or to pRNA from normal tissue. It is not a trivial matter to obtain nucleic acid preparations that will satisfy all of these criteria. It is, unfortunately, not uncommon to obtain DNA preparations that fail the third requisite of specific hybridizability to its homologous RNA. This failure may stem from slight contamination with cellular DNA in the synthesis mixture, in which case the product will contain labeled DNA that will hybridize to both normal and tumor RNA. Another source of confusion is the possible complementary copying of long stretches of adenylate residues that are present in oncogenic RNAs (6, 7) and in normal cellular messages (8, 9). [3H]DNA with the resulting blocks of T residues will hybridize to any RNA containing correspond- ing blocks of A. This difficulty is greatly magnified by the use of oligo(dT) as a primer to enhance synthesis, a procedure that should be avoided when specific probes are made. One Proc. Wat: had. Sci. USA 69 (1972) Viral-Related RNA in Hodgkins' Disease 1729 can readily test for the occurrence of this type of pairing by esamining the effect of prehybridizing with unlabeled oligo(rU). Every [3H]DNA used as a probe in the present and previous (14) studies successfully passed the three tests outlined above. We cannot overemphasize the need for exercise of the precautions noted in investigations along these lines. Unless the materials used satisfy the criteria stipulated, the resulting esperiments are likely to generate more confusion than information. RESULTS We first illustrate (Fig. 1A-C) representative outcomes of hybridizations between RLV [ 3H]DNA and the pRNA pre- pared from three types of human lymphomas. Between 2 and 6% of the [3H]DNA is shifted to the RNA region of the Csy SO1 gradient. The position of the complex implies that the RNA is much larger than the DNA and, thus, determines the density of the hybrid structure. The amounts of RLV[3H]- DNA involved in the complexes are comparable to those we haye seen in similar experiments with pRNA from human sarcomas (4) and human leukemic white blood cells (3). Fig. 1D compares the responses of pRNA preparations from spleens of normal humans and those with Hodgkins' disease at various concentrations of pRNA. No detectable reaction is obtained with normal spleen pRNA, even at the highest con- centration tested, which was in excess of 7 mg/ml. Within the same range, the pRNA derived from the spleen of the patient with Hodgkins' disease complexed with somewhat more than 2y0 of the RLVj3H]DNA, with no evidence of saturation at 6.7 mg of pRNA per ml. Since it is unwieldy to detail the Cs2SO4 gradient profile of every sample tested, we have adopted a more convenient recording of our data. After correction for background, the tritium counts in the density region of RNA (1.62-1.69 g/ml) were taken as a measure of the amount of [3H]DNA com- plexed to RNA. To attain the statistical accuracy desired, 10-min counts (cplOm) were taken on each sample. The con- vention was adopted that hybridization was negative if the sum of the cplOm in the RNA density region was less than three standard deviations above the mean background in the same region. Table 1 lists the lymphoma samples tested for RNA related to the DNA of RLV. The results are recorded in counts per 10 min (corrected for background) in the RNA density region, and as multiples of the mean background standard deviation. Included are 24 Hodgkins' lymphomas, three reticulum cell sarcomas, and five Iymphosarcomas. Of the 32 lymphomas, 22 (or 69Oj,) gave clear indications of containing RNA ho- mologous to the RNA of RLV. This result is in sharp contrast to the results obtained with pRNA from normal adult and fetal tissues (Table 2). None of the 48 normal preparations tested gave a reaction that could be accepted unambiguously as positive. Fig. 2 provides a convenient pictorial summary as multiples of the mean background standard deviation. The fact that 69% of the pRNAs derived from lymphomas yielded positive hybridizations with RLV-DNA, whereas none of the 48 control tissues (Table 2, Fig. ID) was positive, already argues for the specific significance of the positive reac- tions. Further support for this conclusion can be provided by the use of [3H]DNA complementary to irrelevant viral RNAs in hybridizations with lymphoma pRNAs positive for a reaction with RLV-DNA. We have shown (manuscript in preparation) that RLV-DNA and RNA do not crosshybridize significantly with the corresponding nucleic acids of either AMV or MMTV. If the annealing reaction is specific, one would not expect a lymphoma pRNA positive for a reaction with RLV- DNA to show any ability to hybridize either with MMTV- Test for viral-specijic RNA in human lymphomas TABLE 1. RNA- region Rea* Lymphomas cplOm cplOm/S tion Hodgkins' lymphomas 1417 (S)* 1/145 L (L) 1/145 Sp (S) 726 (L) 476 (S) 821 (LN) 731 (8) 1413 (S) 622 (LN) 1413 (LN) 1217420 (8) 1121260(S) 1241240 (S) 1238972(S) 93 (SI 030893 (8) 1059649 (S) 1245268 (S) 1244787 (8) 1241721 (S) 1239151 (S) 229 (S) 254 (S) L (8) Lymphoma 24317 (LN) Histiocytic Reticulum cell Reticulum cell sarcomas lymphoma (LN) sarcoma 24282 (LN) Lymphosarcomas Lymphosarcoma Hz (Ly) Lymphosarcoma Hv (Ly) Lymphosarcoma Ev (Ly) Lymphosarcoma 101 (Ly) Lymphosarcoma 241 (LN) 292 269 223 306 166 52 609 66 48 83 600 649 50 285 32 447 463 265 274 296 419 289 318 429 88 438 565 37 1 515 387 1217 178 3.40 3.13 2.59 3.56 1.93 0.60 7.08 0.77 0.56 0.97 6.98 7.55 0.58 3.31 0.37 5.20 5.38 3.10 3.20 3.44 4.87 3.36 3.70 5.00 1.02 5.10 6.57 4.31 6.00 4.50 14.90 2.07 + + + - - - + - - - + + + + + + + + + + + + - - - + + + + + + - 32 Tumors tested, 69% positive. Results of hybridization reactions between RLV-[3H] DNA and pRNA isolated from human lymphomas. 200-1000 pg of pRNA in each sample were hybridized to 2000 cpm of RLV- [3H] DNA, and the reactions were analyzed by CslSOn equilibrium centrifugation. The amount of DNA banding in the RNA region of the gradient (between densities 1.62 and 1.69) was then determined. The results are expressed as cplOm (corrected for background) banding in the RNA region for each RNA sample tested, and 515 multiples of S, the operational standard deviation (see text and legend to Fig. 2). The annealing reaction is con- sidered positive only if the cpl0m per RNA region is greater than 38, thus providing 99.9% confidence statistically. *The tissue tested is given parenthetically: S, spleen; L, liver; LN, lymph node; Ly, lymphocyte. 1730 Medical Sciences: Hehlmann et al. Proc. Nut. Acad, Sci. USA 69 (1972) TABLE 2. Test for viral-specific RNA in normal human tissues Tissue Normal white blood cells Phytohemagglutinin- Phytohemagglutinin- stimulated lymphocytes (39) stimulated lymphocytes (44) 10-14 (LN)* 24192 (LN) 24207 (LN) 24196 (LN) 24193 (LN) 24282 (LN) 24257 (LN) 24300 (LN) 24317 (LN) 24318 (LN) 24207 (S) 24238 (S) 24239 (8) 24240 (8) 24297 (8) 24300 (S) 24302 (S) 1072584 (8) 24318 (S) 24314 (S) W.R. (L) 24206 (L) 24207 (L) Intestine 24207 Fetal lung Fetal liver (12 week) Fetal liver (14 week) Fetal liver (15 week) Fetal liver (16 week) Cell line NC37 (Ly) Placenta 28005 24204 (L) 24205 (L) Intestine 24205 Intestine 24204 Striated muscle 24204 Striated muscle 24205 Fetal limbs (16 week) Fetal limbs (14 week) Fetal limbs (14 week) Fetal limbs (16 week) Fetal limbs (16 week) Fetal limbs (24 week) 64ym (8) 24241 (S) RNA- region cplOm 68 212 24 1 74 249 253 253 254 105 85 39 176 130 226 175 203 180 214 34 127 70 128 116 133 247 242 61 209 232 207 238 62 64 73 13 1 84 42 88 43 124 88 144 86 116 115 __ - - - cplOm/S 0.80 2.50 2.80 0.90 2.90 2.90 2.90 2.95 1.22 0.99 0.0 0.45 2.05 1.50 2.60 2.03 2.36 2.09 2.49 0.40 1.48 0.0 0.81 1.49 0.0 1.40 1.60 2.90 2.80 0.70 2.40 2.70 2.40 2.80 1.17 1.21 1.38 2.48 1.59 0.80 1.66 0.81 2.34 1.66 2.72 1.62 2.19 2.17 48 Samples tested, 0% positive. Results of hybridization reactions between RLV-[3H] DNA and pRNA isolated from normal human tissues. 200-1000 pg of pltNA in each sample were annealed to 2000 cpm of RLV- [3H]DNA, and the reactions were subjected to CszSOa equilib- rium centrifugation. The data are expressed as cplOm (cor- rected for background) that bands in the RNA region of the gradient, and as multiples of the standard deviation (S), as de- scribed in Fig. 2 and Table l. * See Table 1. DNA or AMV-DNA. Fig. 3A and B show that these expecta- tions are realized. Two Hodgkins' lymphoma pRNAs, positive for reactions with RLV-DNA, show no significant reactions with either AMV-DNA or MMTV-DNA. DISCUSSION The absence of positive reactions with the pRNA preparations from certain of the lymphomas may raise questions of uni- versality in the minds of some. However, it must be recognized that a negative outcome cannot be accepted as evidence for the nonexistence of the relevant RNA, whether the tissue being tested is neoplastic or normal. The sensitivity of the method is, at present, principally limited by the amount of RNA that can be dissolved in the annealing mixture, a diffi- culty that will be greatly obviated by the development of a suitable enrichment procedure for the pertinent RNA frac- tion. As in the instances of the leukemias (3) and sarcomas (4), we can at best conclude that the probability of finding RNA homologous to RLV-RNA is much greater in human lymphoma cells than in normal tissues. Indeed, if the provirus (10) or oncogene (11) hypotheses are valid, some part of this oncogenic information could be expressed and detected with the aid of more sensitive tests in certain normal tissues that have been reported (12) to exhibit group-specific antigens of the mammalian leukemogenic viruses. Tumor Tissue Normal Human Tissues FIG. 2. Results of hybridization reactions with RLV-[3Hl- DNA and pRNA from human lymphomas and normal human cells. The normal tissues tested are: normal white blood cells (X 1, phytohemagglutinin-stimulated lymphocytes (El), a human lymphocyte cell line, NC37 (@), lymph nodes, spleens; other adult tissues: liver (A), intestine (0), and striated muscle (G); and fetal tissues: liver (A), lung (V), limbs (0), and placenta (0). The re- actions were centifuged in Cs2SOa equilibrium density gradients. The amount of [3H]DNA, expressed as cpl0m corrected for back- ground, banding in the density region of RNA (between densities 1.62 and 1.69), was determined for each reaction. An operational mean and standard deviation (8) were then determined for each counter by the total cplOm of three tubes (e.g., 2,3,4) of each of 60 gradients. The number of [3H]DNA cpl0m corrected for back- ground banding in the RNA region of the gradient was then divided by the appropriate operational standard deviation. Any reaction with cpl0m in the RNA region less than 3S is con- sidered negative, thus providing 9(3.9cj, confidence that those re- actions retained as positive (greater than 3s) are significant. Proc. Nat. Acad. Sci. USA 69 (1972) Viral-Related RNA in Hodgkins' Disease 1731 1.7 1.5 Z 1.4 pI 1.3 50 FIG. 3. Cs2SO4 equilibrium gradient centrifugation of (A) AMV-[3H]DNA and (B) MMTV-[3H]DNA hybridized, each to 400 pg of Hodgkins' lymphoma pRNA. Hybridization conditiom and Cs2SO4 gradient analysis are detailed in Methods. The existence of viral-related RNA in human lymphonias does not of course establish a viral etiology for this disease. One must now perform experiments designed to answer the following questions: (i) How large is the RNA being detected? (ii) How much homology does it in fact have to the RLV- RNA? and (iii) Is the viral-related RNA associated with a reverse transcriptase (RNA-instructed DNA polymerase) and is it located in structures characteristic of incomplete or com- plete virus particles? The requisite techniques have been de- veloped to answer these questions with respect to the human rieoplasias we have studied, and the experiments are underway. Discovery of RNA homologous to that of the murine leukemia virus in human lymphomas immediately raises the issue of Burkitt's disease, a malignant lymphoma found in children living in a geographical belt extending across Central Africa. There exists serological (13) and nucleic acid hybridiaa- tion (14) evidence linking this disease to the Epstein-Barr virus (EBV), a herpes-like virus that contains DNA detected in (15), and isolated (16), from Burkitt's lymphoma cells grown in culture. Since a causal relation between Epstein- Barr virus and Burkitt's tumor has not been established, it is pertinent to inquire whether this lymphoma, like the others examined here, also contains RLV-related RNA. In many ways the most noteworthy features of our studies of human neoplasias emerge when they are compared with one another, and to this end a composite of the results we have thus far accumulated is presented in Table 3. Breast cancer TABLE 3. Homologies among human neoplastic RNL4s and animal tumor viral RNAs Human neoplastic RNAs Breast Leu- Lym- Viral RNA cancer kemia Sarcoma phoma tumor virus + leukemia virus - sis virus Mouse mammary Itauscher murine Avian myeloblasto- - - - + + + - - - - The results of molecular hybridization between [31-1] DNA com- plementary to the various viral RNAs arid pRNA preparations from the indicated neoplastic tissues. A plus sign indicates that hybridizations were positive and a minus sign that no hybridiza- tion could be detected. contains RNA homologous only to that of the murine mam- mary tumor virus. Leukemias, sarcomas, and lymphomas all contain RNA homologous to that of the Rauscher leukemia virus and not to MMTV-RNA. Finally, none of the human tumors contains RNA detectably related to the RNA of the avian myeloblastosis virus. It is clear that with all four human neoplasias examined, the specificity pattern of the RNA they contain is in complete accord with what has been observed in the corresponding viral-induced malignancies in the mouse. We thank the following for supplying cell material: Dr. George A. Hyman (Presbyterian Hospital, New York City), Dr. J. F. Holland (Roswell Park Memorial Institute, Buffalo, N.Y.) and Dr. L. Dabich (Simpson Memorial Institute, University of Michigan Medical Center, Ann Arbor). We greatly appreciate the excellent technical assistance of Elaine Gordon, Lee Hindin, Jeanne Myers, and Sidney Shinedling. This research was sup- ported by the National Institutes of Health, National Cancer Institute, Special Virus Cancer Program Contract 70-2049 and Research Grant CA-02332. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Axel, R., Schlom, J. & Spiegelman, S. (1972) "Evidence for translation of viral-specific RNA in cells of a mouse mam- mary carcinoma," Proc. Nut. Acad. Sci. USA 69, 535-538. Axel, It., Schlom, J. & Spiegelman, S. (1972) "Presence in human breast cancer of RNA homologous to mouse mam- mary tumor virus RNA," Nature 235, 32-36. Hehlmann, R., Kufe, D. & Spiegelman, S. (1972) "RNA in human leukemic cells related to the RNA of a mouse leukemia virus," Proc. Nut. Acad. Sci. USA 69, 435-439. Kufe, I)., Hehlmann, R. & Spiegelman, S. (1972) "Human sarcomas contain RNA related to the RNA of a mouse leukemia virus," Science 175, 182-185. Kacian, D. L., Watson, K. F., Burny, A. & Spiegelman, S. (1971) "Purification of the RNA-directed DNA polymerase of the avian myeloblastosis virus," Biochim. Biophys. Acta Lai, M. hf. C. & Iluesberg, P. (1972) "Adenylic acid-rich sequence in RNAs of Rous sarcoma virus and Rauscher mouse leukaemia virus," Nature 235, 383-386. Gillespie, D., Marshall, S. & Gallo, It. C. (1972) "RNA of RNA tumour viruses contains poly A," Nature, in press. Darnell, J. E., Philipson, L., Wall, R. & Adesnik, RI. (1971) "Polyadenylic acid sequences: Role in conversion of nu- clear RNA into messenger RNA," Science 174, 507-510. Kacian, D., Spiegelman, S., Bank, A,, Terada, M., Illetafora, S., Dow, L. & >farks, P. (1972) "In vitro synthesis of DNA components of human genes for globins," Nature Neu, Biol. Temin, H. M. (1971) J. Nat. Cancer Inst. 46, 111-VII, guest editorial. Huebner, R. J., Kelloff, G. J., Sarma, P. S., Lane, W. T. & Turner, H. C. (1970) "Group-specific ant,igen expression during embryogenesis of the genome of the C-t'ype RNA tumor virus: Implications for ontogenesis and oncogenesis," Proc. Nut. Acad. Sci. USA 67, 366-376. Hellman, A. & Fowler, A. K. (1971) "Hormone-activated expression of the C-type RNA tumour virus genome," Nature New Biol. 233, 142-144. GunvBn, P., Klein, G., Henle, G., Henle, W. & Clifford, P. (1970) "Epstein-Barr virus in Burkitt's lymphoma and nasopharyngeal carcinoma," Nature 228, 1053-1056. Zur Hausen, H., Schul te-Holthausen, H., Klein, G., Henle, W., Henle, G., Clifford, P. & Santesson, L. (1970) "EBV DNA in biopsies of Burkitt tumors and anaplastic car- cinomas of the nasopharynx," Xature 228, 1056-1058. Epstein, 11. A,, Achong, B. G. & Barr, Y. M. (1964) "Virus particles in cultured lymphoblasts from Burkitt's lym- phoma," Lancet i, 702-703. Epstein, hl. A,, Henle, G., Achong, B. G. & Barr, Y. (1965) "Morphological and biological studies on a virus in cultured lymphoblasts from Burkitt's lymphoma," J. Ezp. Med. 121, 246, 365-383. 235, 167-169. f61-770.