Determinants of the Host Range of Feline Leukaemia Viruses

J. gen. Virol. (1973), 20, I69-t75 Printed in Great Britain 169 Determinants of the Host Range of Feline Leukaemia Viruses By O. JARRETT, HELEN M. LAIRD AND D. HAY University of Glasgow, Leukaemia Research Unit, Bearsden, Glasgow, Scotland G6I I QH (Accepted 8 March 1973) SUMMARY Feline leukaemia viruses of subgroups B and C multiply in human and canine cells, while subgroup A viruses do not. This host range restriction is determined by the virus envelope and operates at the level of virus entry into the cell. Subgroup A virus genomes are expressed and replicated when they are introduced within B subgroup envelopes into human or dog cells. Therefore, since they are phenotypic mixtures of A and B subgroup viruses, the majority of feline leukaemia virus isolates will infect human and canine cells with the subsequent production of FeLV of each subgroup. INTRODUCTION Knowledge of the determinants of the host range of RNA tumour viruses is important in view of possible associations between leukaemia viruses and leukaemia in man, particularly when viruses of some species are known to multiply in human cells. For example, the majority of isolates of feline leukaemia virus (FeLV) grow in human cells as well as in cat cells (Jarrett, Laird & Hay, 1969, O'Connor & Fischinger, 197o; Jarrett, 1971) and feline sarcoma viruses multiply in, and transform, human cells (Chang, Golden & Harrold, 197o; Sarma et al. 197o). Recently, however, we described an isolate of FeLV which did not replicate in human cells. This virus was a member of FeLV subgroup A (Jarrett, Laird & Hay, I97Z), as defined by Sarma & Log (1971). Here we report the results of experiments to investigate whether this host range restriction is a general property of FeLV subgroup A viruses, and to determine the factors necessary for FeLV isolates to initiate successful infections in human and canine cells. METHODS Cells. Feline embryonic (FE) cells of the FEA strain were derived from whole embryos and were propagated using Eagle's minimal essential medium with IO % foetal bovine serum (EFB). In these experiments, cells between the eighth and twentieth subcultures were used. The human embryonic lung (HEL) cells used here were either derived from primary cultures or from the Flow aooo cell line (Flow Laboratories, Ltd.) and were grown in EFB. A suspension culture derived from normal canine thymus cells (CT 45S) which was originally isolated by Dr J. Mitchell was obtained from Dr M. B. Essex. These cells were grown in stationary suspension in medium consisting of equal proportions of Leibovitz L-15 and McCoy's 5a media supplemented with 15 % foetal bovine serum. Their doubling time was about 24 h and they were maintained in their growth phase at cell concentrations between 7 lo 5 and 2 IO cells/ml. Viruses. The feline leukaemia viruses FeLV-I, 4, 5, 6, 9, IO, 13, 34, 38, 4o and 41 were

17o o. JARRETT, H.M. LAIRD AND D. HAY isolated in our laboratory from cats with naturally occurring leukaemia by previously described methods (Jarrett et al I968). FeLV-A, B and C were the gift of Dr P. S. Sarma. Each virus isolate was grown in FE cells, care being taken to maintain the cells in exponential growth. FeLV pseudotypes of the Moloney strain of routine sarcoma virus - MSV (FeLV) - were prepared after the method of Fischinger & O'Connor (t969). A mixture of 2 ml of MSV containing 2 x Io 6 focus forming units (f.f.u.) and IO ml of the culture fluid from FeLVinfected FE cells was centrifuged at 3500o rev/min for 60 min in a Beckman SW4o rotor. The pellet was resuspended in EFB and was used to infect FE cells in conditions described below for the assay of MSV (FeLV). Cells were infected with diluted virus and after six days single transformed foci were picked out in a pipette and were transferred to fresh cultures of FE cells together with I ml of homologous leukaemia virus. The ceils were subcultured every 3 to 4 days until transformation was widespread, at which time the culture fluids were harvested, centrifuged at 2o0o rev/min for IO rain and stored at -6o C. Virus assay. In the MSV (FeLV) virus assay 5 cm plastic plates, seeded 24 h previously with 3 x Io 5 FE cells, were inoculated with o.5 ml of dilutions of virus in EFB. After an incubation period of 9o min at 37 C the inoculum was replaced with a mixture of 4 ml of EFB and I ml of appropriate FeLV at a concentration known to permit optimum expression of f.f.u, of sarcoma virus. The medium was changed after 3 days and on the sixth day the transformed foci were counted as before (Jarrett et al. 1972). When HEL cells were used they were seeded at a concentration of 2 x lo 5 cells/plate. Interference test. FE cells were grown for 21 days after virus inoculation to ensure that all cells were infected. Then 5 x lo 5 cells were seeded into 5 cm plates in 4 ml EFB and 24 h later the culture was inoculated with approximately 5oo f.f.u, of MSV (FeLV) of appropriate subgroup in the same conditions as in the MSV (FeLV) assay. The medium was replaced on the third day. Cultures in which no transformation was observed after 6 days were considered to show positive interference. Detection of virus. The techniques of labelling cells with [3H]-uridine and electron microscopic examination of cell cultures were as described previously (Jarrett et al. 1972 ). RESULTS Restricted host range of FeLV of subgroup A We have shown that an FeLV isolate, FeLV-I, has a restricted host range in that it is unable to infect human cells, and that this virus belongs to the A subgroup (Jarrett et al. I972). The suggestion that there is a correlation between host range and virus subgroup was examined. Human embryonic lung cells were inoculated with FeLV isolates known to be members of subgroup A (FeLV-I, 13, 34, 38 and A). Each inoculum, which had been grown in FE cells, was frozen and thawed at Ieast once before use to ensure that there was no contamination of HEL cells with live cat cells which might continue to release virus. When the HEL cells were examined 21 days later in the electron microscope there was no evidence of virus growth, as shown in Table 1. By contrast, it was found that FeLV stocks which contained viruses of B or C subgroups alone (FeLV-B and FeLV-C), or were mixtures of viruses of subgroups A and B (FeLV-4, 5, 6, 9, lo, 41 and 43), were able to replicate in HEL cells. An identical pattern was observed when suspension cultures of canine thymic cells were inoculated with representative viruses of each subgroup.

Feline leukaemia viruses t 7 I Table r. Host range o f feline leukaemia viruses Virus subgroup A B C AB Growth in the~e cells :* FeLV r ~ isolate feline human canine t -}- -- -- A + - - I3 + -- _ 34 + -- 38 + - B + + + C + + + 4 + + 5 + + + 6 + + 9 + + to + + 4I + + 43 + + * Where no result is given, the virus-cell interaction was not tested. Table 2. Quantity of different subgroups of FeL V produced in feline and human cells Virus Cell PHl-ct/min incorporated ct/min in HEL cells type Subgroup type into virus particles ct/min in FE cells FeLV- I A HEL 6 I FE 1537 FeLV-B B HEL 213 t t '9 FE i I 2o FeLV-C C HEL 648 0'5 FIE 1299 FeLV-5 AB HEL 2352 I "4 FE t 644 HEL 5 I FE 3I Eight cultures of different primary HEL cells and one HEL cell line were inoculated with FeLV-I, and none of these supported virus replication. In a further experiment to determine if cat cells which were non-permissive for FeLV- ~ were common, the virus grew readily in cell cultures from each of 6 feline embryos which was tested. From electron microscopic examination it appeared that in certain situations a larger quantity of virus was produced in human cells than in feline cells. A quantitative comparison revealed that this was so. Cells, inoculated 3 weeks previously with FeLV, were grown in medium containing [~H]-uridine for 24 h and the amount of virus which was released into the culture fluid was measured by the quantity of [3HI-incorporated into particles with a buoyant density between I.I3 and I.I8 g/cm ~, as described previously (Jarrett et al. I972 ). The results in Table 2 indicate that FeLV which contained B subgroup virus was released in greater amounts from HEL cells than from FE cells, but that the reverse was the case for ceils infected with FeLV-C. This experiment also confirmed that FeLV-I was not released from HEL cells. The amount of radioactivity at these densities was essentially similar in HEL cells inoculated with FeLV- t and in uninfected HEL and FE cells. I2 VIR 20

I72 O. JARRETT, H.M. LAIRD AND D. HAY Table 3. Efficiency of plating of MS V (FeLV) of different subgroups on feline and human cells Virus titre (f.lu./ml) on: r ~ ' - - ~ Titre on HEL cells Virus Subgroup FE ceils HEL cells Titre on FE cells MSV (FeLV-r) A 2'o x lo 3 o o MSV (FeLV-B) B 2"8 x IO 4 I-2 x io 5 4'3 MSV (FeLV-C) C 3"o x io 4 3"0 x ]o 3 o'1 MSV (FeLV-4) AB 5"6 x 1o 3 1.2 lo 3 o.2 MSV(FeLV-5) AB l'i 1o 5 4"2 x io 4 0"4 MSV (FeLV-9 AB 2.2 x IO 4 4"4 Io~ 0.2 Table 4. Interference patterns among feline leukaemia viruses Preinfecting Challenge virus : FeLV-type of MSV (FeLV): virus ~ ~ ~- ---* FeLV-I FeLV-B FeLV-C FeLV-5 FeLV-5H FeLV-I +.... FeLV-B -- + -- -- -- FeLV-C - - + - - FeLV-5 + + - + + FeLV-5F + + - + + FeLV-sH + + - + + FeLV-sC + + - + NT + denotes positive interference to a challenge of 500 f.f.u, of the appropriate MSV (FeLV). NT = Not tested. Determinants of the host range of FeLV The results of experiments in which feline and human ceils were exposed to pseudotypes of MSV with the envelope properties of various feline leukaemia viruses indicated that the host range of FeLV is determined by the virus envelope. MSV (FeLV) viruses multiply in, and morphologically transform feline cells (Fischinger & O'Connor, I969; Sarma & Log, I97I ; Jarrett et al. I972 ). However, like their leukaemia virus counterparts, the subgroup A sarcoma viruses, MSV (FeLV-I) and MSV (FeLV-A), did not infect HEL cells. Transformation was not observed when the cells were exposed to 2oo0 f.f.u, of MSV (FeLV-~) or 4ooo f.f.u, of MSV (FeLV-A) alone or together with optimum amounts of FeLV-B as a permissive helper virus. Treatment of the cells for 2 h prior to virus adsorption with 25 #g]ml of DEAE-dextran did not remove this restriction. In contrast, MSV (FeLV) stocks which contained B or C subgroup viruses, either alone or in mixtures with A subgroup virus, transformed and multiplied in HEL ceils, as is indicated in Table 3. MSV (FeLV-B) plated four times more efficiently on HEL cells than on FE cells, while HEL cells appeared to be ~o times less susceptible than FE cells to infection with MSV (FeLV-C). The e.o.p, on HEL cells of three different MSV (FeLV) stocks which were mixtures of A and B subgroup viruses was between 20 and 40 ~ of that on FE cells. From these results it appeared that the virus envelope was an important factor in the ability of FeLV to infect human cells. Infection of cells with phenotypic mixtures The following experiments in which HEL cells were exposed to FeLV-5, which is a phenotypic mixture of A and B subgroup viruses, revealed that there was no restriction on the multiplication of subgroup A virus genome once they had entered human cells.

Feline leukaemia viruses I73 Table 5. Production of FeLV- 5 in feline and human cells FeLV-5-infected cell type f [3H]-ct/min incorporated into virus on post-infection days: 7 14 2I FE 2935 (45)* 345I (93) 2186 (IO7) HEL 63o (IO) 31o6 (83) 3128 (I53) FE control 6545 3728 2035 * Figures in parentheses represent the ct/min as the percentage of the ct/min produced on each occasion by the control culture of FE cells chronically infected with FeLV-5. The subgroup composition of the virus produced by FeLV-5-infected HEL, CT and FE cells (designated FeLV-5H, FeLV-5C and FeLV-5F respectively) was examined by interference tests. Fluids from these cultures were harvested 2I days after inoculation, were frozen and thawed and were inoculated into fresh FE cells. After 2I days in culture these cells were challenged with 5oo f.f.u, of MSV (FeLV) of subgroups A, B or C. The results, in Table 4, show that the virus produced in the original HEL, CT and FE cells was, like the parental FeLV-5, a mixture of A and B subgroup viruses. Therefore, virus of subgroup A multiplied in the human and canine cells. Confirmation of this result was obtained in reciprocal experiments which showed that the MSV (FeLV) pseudotype with the envelope properties of FeLV-5H, MSV (FeLV-5H), had an interference pattern like that of MSV (FeLV-5) which was consistent with it containing both A and B subgroup viruses. As shown in Table 5, a longer period of time was required for FeLV-5 to produce maximum amounts of virus in infected HEL cultures compared with similarly infected FE cultures. It is presumed that this finding was due to only a small proportion of the HEL cells being infected initially since MSV (FeLV-5) has a low e.o.p, on HEL cells, as indicated in Table 3- However, when the infection of HEL cells with FeLV-5 was well-established, i.e. 3 weeks after inoculation, more virus was released from the human cells than from the cat cells. DISCUSSION All feline leukaemia viruses, of whatever subgroup, or mixture of subgroups, which have been tested so far have grown in feline cells; indeed, this is the means by which they are usually isolated from cats. It is not known yet whether, as in avian leukosis systems, there are cats which are genetically resistant to infection by viruses of a particular subgroup of FeLV, nor is it likely that this information will be easily obtained owing to the absence of sufficiently inbred strains of cats. In our animal experiments, which have involved over 9o0 randomly-bred cats, almost every animal inoculated with representative viruses of subgroups A, B and C, as well as with mixtures of A and B subgroup viruses, has shown evidence of virus infection. Also, there is no obvious breed resistance since FeLV-associated haematopoietic tumours in cats are recognized in all common breeds of cats (Brodey et al. I97o). It appears, though, that viruses of the FeLV subgroups A, B and C differ in their interspecies host range, at least in vitro. The results of the experiments described here indicate that FeLV-B and C subgroup viruses grow in human and canine cells but that A subgroup viruses do not. The restricted host range is probably a general property of subgroup A viruses since the five isolates which were tested in this study did not grow in HEL cells, nor did one of these isolates, FeL -I, grow in eight separate HEL cell cultures. While only one I~-2

174 O. JARRETT, H.M. LAIRD AND D. HAY B subgroup virus was tested and found to grow in HEL and canine cells, seven other isolates which contained B in addition to A subgroup viruses also successfully infected these cells. The virus of subgroup C which was tested behaved qualitatively like virus of the B subgroup. As with the avian and routine RNA tumour viruses (Hanafusa & Hanafusa, I966; Huebner et al. I966), the factor which determines host range resides in the virus envelope. Thus, FeLV pseudotypes of MSV have the same host range as their parental leukaemia viruses, and the pseudotypes representing each FeLV subgroup differ only, as far as is known, in their envelope properties. Apparently, viruses of subgroup A cannot enter human or canine cells due to their A-specific envelopes. However, the results obtained using a phenotypically mixed population of A and B subgroup viruses indicate that subgroup A virus genomes can enter human cells within subgroup B virus envelopes. In such mixtures the proportion of particles which can enter human cells is apparently low. This is evident from the prolonged time taken to establish maximum production of virus in FeLV-5 infected HEL cells compared with FE cells (Table 4); and from the result, shown in Table 3, that the e.o.p, of MSV (FeLV-5) on HEL cells was only 40 ~ of that on cat cells. Similar results were obtained with MSV (FeLV-4) and MSV (FeLV-5). However, when FeLV-5 production was at a maximum rate more virus was released from HEL cells than from FE cells. Sarma et al. 097o) reported what is now recognized as an analogous situation in which the titre of the Gardner-Arnstein strain of feline sarcoma virus (FeSV) was Io times greater when assayed on FE cells than on whole human embryonic cells. Also, the endogenous leukaemia virus which is present in this sarcoma virus isolate was produced in greater amounts in human cells than in cat cells after I4 days in culture. This strain of FeSV was subsequently shown to contain both A and B subgroup viruses (Sarma et al. I97I ; Sarma, Sharar & McDonough, I972). The present results support the observation that FeLV-5 consists of about Io ~ of viruses with essentially A subgroup envelopes; Io ~ with B subgroup envelopes, and the remainder with mixed AB envelopes (O. Jarrett, unpublished observations). It is proposed that it is only those particles with substantially B type envelopes which can infect human cells. The apparent discrepancy between the relative e.o.p, of MSV (FeLV-5) on human cells compared with cat cells (40 ~o) and the proportion of MSV (FeLV-5) with B envelopes (IO ~) is assumed to be due to the fact that MSV (FeLV) viruses with subgroup B envelopes plate 2 to 4 times more efficiently on HEL cells than on FE cells, as shown in Table 3 for MSV (FeLV-B). An as yet unknown proportion of the particles in FeLV-5 which have subgroup B envelopes must contain genomes of subgroup A viruses, since the progeny from FeLV-5 infected human and canine cells, like the parental FeLV-5, contains viruses of both subgroups. Therefore, human and canine cells are permissive for the expression and replication of subgroup A virus RNA. Such expansion of the host range of RNA tumour viruses which are normally excluded from cells of a particular genotype has been recognized before for avian leukosis viruses (Vogt, I967). Our results indicate that since most feline leukaemia viruses which are isolated from cats are mixtures of A and B subgroup viruses (Sarma & Log, 1973; O. Jarrett, unpublished observations), they are capable of multiplying in human and canine cells, and that FeLV of each subgroup will be produced in these cells. This work was supported by grants from the Cancer Research Campaign and the Lady Tata Memorial Trust. The authors thank Dr P. S. Sarma and Dr M. B. Essex for their generous gifts of viruses and cells, Professor W. F. H. Jarrett for helpful discussions and M. Golder for excellent technical assistance.

Feline Ieukaemia viruses I75 REFERENCES BRODEY, R. S., MCDONOUGH, S. K., FRYE, F. L. & HARDY, W. D. (197O). Epidemiology of feline leukemia (lymphosarcoma). In Comparative Leukemia Research, pp. 333-342. (1969). Edited by R. M. Dutcher. Basel: Karger. CHANG, R. S., GOLDEN, n. D. & HARROLD, a. (I970). Propagation in human cells of a filterable agent from the ST feline sarcoma. Journal of Virology 6, 599-6o 3. FISCHINGER, P. J. & O'CONNOR, T. E. (1969). Viral infection across species barriers: reversible alteration of murine sarcoma virus for growth in cat cells. Science, New York x65, 714-716. HANAFUSA, H. & HANAFUSA, T. (1966). Determining factor in the capacity of RSV to induce tumours in mammals. Proceedings of the National Academy of Sciences of the United States of America $5, 532-538. HUEBNER, g.j., HARTLEY, J.W., ROWE, W. l'., LANE, W.Y. & CARPS, W.I. (I966). Rescue of the defective genome of Moloney sarcoma virus from a noninfectious hamster tumor and the production of pseudotype sarcoma viruses with various murine leukemia viruses. Proceedings of the National Academy of Sciences of the United States of America 56, I 164- I 169- JARRETT, O. (1971). Virology and host range of feline leukemia virus. Journal of the American Veterinary Medical Association x58, Io32-Io39. ~ARRETT, O., LA1RD, H. M., nay, D. & CRIGHTON, G. W. (I968). Replication of cat leukaemia virus in cell cultures. Nature, London 219, 521-522. JARREWT, O., LAIRD, H. M. & HAY, D. (I969). Growth of feline leukaemia virus in human cells. Nature, London 224, 12o8-12o9. JARRETW, O., LAIRD, I~. M. & HAY, D. (I972). Restricted host range of a feline leukaemia virus. Nature, London 238, 220--22 I. O'CONNOR, T.E. & EISCHINGER, P.J. (1970). Physical alterations of a murine leukemia virus complex in mammalian cell cultures. In Comparative Leukemia Research, pp. 250-256 (1969). Edited by R. M. Dutcher. Basel: Karger. SARMA, P. S., BASKAR~ J. F., GILDEN, R. V., GARDNER, M. B. & HUEBNER, R. J. (1971). In vitro isolation and characterization of the GA strain of feline sarcoma virus. Proceedings of the Society./'or Experimental Biology and Medicine x37, 1333-1336. SARMA, P. S., HUEBNER, R. J., BASKER, 3. Y., VERNON, L. & GILDEN, R. V. (I970). Feline leukemia and sarcoma viruses: susceptibility of human cells to infection. Science, New York 168, lo98-i IOO. SARMA, V. S. & LOG, T. 097I). Viral interference in feline leukemia-sarcoma complex. Virology 44,352-358. SARMA, V. S. & LOG, T. (I973). Viral envelope antigens of feline leukemia and sarcoma viruses. Proceedings of the Fifth International Symposium on Comparative Leukemia Research, Padova, (197I). Basel: Karger. SARMA, V. S., SHARAR, A. L. & MCDONOUGH, S. (1972). The SM strain of feline sarcoma virus. Biological and antigenic characterization of the virus. Proceedings of the Society for Experimental Biology and Medicine I4O, 1365-I368. VOGT, v. K. (I967). Phenotypic mixing in the avian tumor virus group. Virology 32, 7o8-717. (Received I4 February 1973)