lymphocytes

Proc. Natl. Acad. Sci. USA

Vol. 73, No. 10, pp. 3714-3718, October 1976 Microbiology

Replication of murine leukemia virus in bone marrow-derived lymphocytes (mouse oncornavirus/mouse cells/syncitial assay with XC cells/lipopolysaccharide stimulation)

NANCY H. RUDDLE*, MARTINE Y. K. ARMSTRONG*, AND FRANK F. RICHARDSt * Department of Epidemiology and Public Health; and t Department of Internal Medicine, Yale University School of Medicine, New Haven, Connecticut 06510 Communicated by Dorothy M. Horstmann, July 19, 1976

MATERIALS AND METHODS Tissue Culture Lines. Viral growth was assayed in BALB/c mouse embryo fibroblasts (for B-tropic virus)t, NIH Swiss mouse embryo fibroblasts (for N-tropic virus)t, and SC-1 (for N-tropic and B-tropic virust) (6). SC-i and XC cells were obtained through the generosity of Drs. Janet Hartley and Wallace P. Rowe at the National Institutes of Health. All cell lines except SC-i were maintained in minimal essential medium with Earle's salts (MEM-E) supplemented with 10% heat-inactivated fetal calf serum, penicillin (50 units/ml), and streptomycin (50 ,tg/ml). SC-1 was maintained in McCoy's 5A medium supplemented with 10% heat-inactivated fetal calf serum and 1% penicillin-streptomycin-neomycin. Virus. Murine leukemia virus WN 1802 B, pool 1897, was used. This is a naturally occurring B-tropic virus isolated from the spleen of an 18-month-old BALB/c mouse by Hartley and Rowe (7). The virus was maintained through infection of BALB/c fibroblasts in tissue culture. Virus Assay. The UVXC test for ecotropic MuLV developed by Rowe, Pugh, and Hartley (8), and performed routinely in our laboratory as described previously (5), was used to assay virus solutions or lymphoid cells which contained virus. Preparation of Lymphoid Cell Cultures. Cells from BALB/cN and (BALB/cJ X A/J) F1 (CAF1) mice 2 to 4 months of age were used. Both strains are of the Fv-1 bb genotype whose fibroblasts support B-tropic virus replication. These mice were obtained from Jackson Laboratory, Bar Harbor, Maine, or from a colony maintained at Yale University. BALB/c nu/nu mice were obtained from ARS Sprague-Dawley. Spleen, thymus, and lymph node cells were prepared as described previously (9). For preparation of bone marrow cells, the ends of femurs and humeri were cut and the marrow flushed out by use of Hanks' balanced salt solution (HBSS), with a 27-gauge needle on a 1 ml tuberculin syringe. Clumps were dispersed with gentle pipetting. Medium RPMI 1640 supplemented with L-glutamine (2 mM), antibiotic-antimycotic solution (penicillin 100 units/ml, fungizone 0.25 ug/ml, and streptomycin 100 ttg/ml) and 5% heat-inactivated fetal calf serum was used. Nucleated cells (2 X 106) were dispensed in 1 ml cultures in 16 X 100 mm culture tubes (Falcon Plastics). Mitogen solutions included were phytohemagglutinin (PHA) 18 ,gg/ml, concanavalin A (Con A) 5 ,gg/ml, and bacterial lipopolysaccharide (LPS) 10,gg/ml. They were obtained from Difco, Calbiochem, and Difco, respectively. Cultures were incubated at 370 in a humidified atmo-

Murine lymphoid cells were infected in vitro ABSTRACT with WN 1802 B, a naturally occurring murine leukemia virus isolated from the spleen of an 18-month-old BALB/c mouse. Normal spleen and bone marrow cells were more susceptible to infection than were cells prepared from thymus and lymph node. Spleen cells from athymic nu/nu mice also could be readily infected with virus. Permissive cells did not ingest iron filings and did not adhere to plastic. Exogenous replication of murine leukemia virus was enhanced in spleen and lymph node cells treated with lipopolysaccharide, a bone marrow-derived lymphocyte mitogen. Conversely, cells treated with the thymus-derived lymphocyte cell mitogens, phytohemagglutinin and concanavalin A, were less capable of supporting murine leukemia virus replication. These studies suggest that the natural host for WN 1802 B is the bone marrow-derived lymphocyte. Murine leukemia virus (MuLV) has been implicated in the development of lymphoid tumors (1). However, analysis of its oncogenicity has been hampered until recently by the absence of appropriate lymphocyte tissue culture systems. Rosenberg et al. (2) have effected transformation of fetal liver cells in vitro by Abelson leukemia virus in association with Moloney leukemia virus. The transformants express immunoglobulin determinants which suggests an origin from a primitive bone marrow-derived lymphocyte (B cell). A correlation between viral replication and the ability to transform a B cell precursor was observed which suggests that systems which are used to explore factors influencing replication of oncornaviruses can be expected to provide insight into malignant transformation. A similar relationship has been observed between the ability of radiation leukemia virus to infect thymocytes in vitro and induce thymomas in vivo (3). Datta et al. (4) infected populations of spleen cells with virus obtained from AKR and B10.A lymphocytes and observed that cells activated in the mixed leukocyte reaction supported more viral replication than normal cells. We report here the results of experiments designed to identify the host cell for another strain of MuLV and to define conditions which influence permissivity in this system. Our interest derives in part from the observation that MuLV titer is correlated with the development of lymphoreticular tumors associated with the graft versus host reaction (5). Our results indicate that the bone marrow-derived lymphocyte is the cell which can be infected most readily in vitro by WN 1802 B, a naturally occurring MuLV. Viral replication is considerably enhanced when cells are activated by the B cell mitogen, lipopolysaccharide.

i The terms N-tropic and B-tropic as used here are concerned with the genotype of the murine fibroblast which supports replication of a particular virus (6). N-tropic viruses replicate preferentially in N-type (Fv-I nnf) mice, as exemplified by the NIH Swiss strain, and B-tropic viruses replicate preferentially in B-type (FV-lbb) mice as exem-

Abbreviations: MuLV, murine leukemia virus; B cells, bone marrowderived lymphocytes; T cells, thymus-derived lymphocytes; LPS, lipopolysaccharide; PHA, phytohemagglutinin; Con A, concanavalin A; HBSS, Hanks', balanced salt solution; MEM-E, minimum essential medium with Earle's salts; PFU, plaque-forming units.

plified by the BALB/c strain.

3714

Microbiology:

Ruddle et al.

Proc. Natl. Acad. Sci. USA 73 (1976)

sphere of 95% air, 5% CO2. Tissue culture media, sera, HBSS, and additives were purchased from Gibco. Virus Infection of Lymphoid Cells. Lymphoid cells were infected at the time of culture initiation or 48 hr later. The supernatant was removed, the cells were resuspended in 0.1 ml of medium, exogenous WN 1802 B (0.3 ml) was allowed to absorb for 1 hr at 40 or 370 with intermittent gentle shaking, and 0.6 ml of fresh medium was added. Agents which enhance MuLV replication, such as DEAE-dextran or polybrene, were not added to lymphoid cultures because we have observed that DEAE-dextran inhibits normal lymphocyte function as assayed by [3H]dT incorporation after PHA stimulation (N. Ruddle, unpublished). Most cultures received 6 X 103 plaque-forming units (PFU) of virus, as determined by assaying virus on DEAE-dextran treated fibroblasts. However, in the absence of polycation, the titer is at least 1 logarithm lower. Four days after addition of virus, supernatants were discarded. The cells were resuspended in 1 ml of MEM-E and placed on monolayers of mouse embryo fibroblasts for an infectious center assay. One week later the fibroblasts were exposed to 1800 ergs/mm2 of ultraviolet light and overlaid with XC cells. In each experiment, a number of parameters were analyzed. The residual amount of initial WN 1802 B virus that remained was measured by adding an entire 4 day culture which contained only virus to BALB/c fibroblasts. Such controls were always negative, due to the rapid inactivation of MuLV at 37'. The release of endogenous MuLV from normal and mitogentreated cells was determined by adding cells from triplicate cultures in three 10-fold dilutions to Swiss, BALB/c, and SC-1 cells. Replication of exogenous WN 1802 B virus was determined by adding infected lymphoid cells to BALB/c fibroblast cultures. RESULTS Three approaches to identify cells capable of supporting exogenous MuLV replication were employed. These were an analysis of the organ distribution of cells capable of being infected by MuLV, analysis of the effect of preferential mitogenic stimulation of different classes of lymphocytes on their ability to support MuLV, and selective elimination of particular cell types by their adherence or phagocytic activity. Preferential infection of spleen and bone marrow cells by exogenous MuLV Organs which contain high proportions of B cells, namely the spleen and bone marrow, were susceptible to infection by MuLV, particularly when virus was added at the time of culture initiation (Table 1). In contrast, thymus cells and lymph node cells were not infected when similarly inoculated. When virus was added to lymphoid cells 2 days after culture initiation, in data not included here, preferential infection of spleen cells was still observed. Thymus cells and lymph node cells were slightly more sensitive to infection at this later time, than when the virus was added at the time of culture initiation. Preferential replication of MuLV in LPS stimulated spleen and lymph node cells from normal and nude

mice Experiments were initiated to determine if preferential stimulation of particular cell types would enhance their ability to support the replication of exogenous WN 1802 B. Spleen and lymph node cells were exposed to mitogen for 48 hr. Mitogen was removed by washing or in the case of Con A cultures,

treatment with a methyl-D-mannoside, and virus was added. Control cultures which received no exogenous virus were also

3715

Table 1. Infection of normal lymphoid cells by WN 1802 B added on day one XC infectious centers/cell culture

Exp. Input MuLV no. PFU/cell 1 2 3 4 5 6

0.001 * 0.009* 0.006t 0.003t 0.003t 0.003t

Spleen

Thymus

Lymph node

marrow

105 445 48 600 1200 352

0 0 0 0 0 11

1 1 0 7 2 0.5

60 135 100 15 595 503

Bone

Virus (0.3 ml) was absorbed for 1 hr to 2 X 106 cells at the time of cell culture initiation. Ninety-six hours later, medium was discarded, the cells were resuspended in 1 ml of MEM-E and added to BALB/c fibroblasts to test for infection by exogenous WN 1802 B. Control cultures which had not received virus were tested for endogenous ecotropic MuLV on NIH Swiss, SC-1, and BALB/c cells. Fibroblasts were irradiated 7 days later with 1800 ergs/mm2 and overlaid with XC cells. The plates were fixed and stained with Giemsa 4 days later. * BALB/c mice. t CAF1 mice.

established to determine whether mitogen treatment would activate endogenous ecotropic virus. Four days after addition of virus (6 days after culture initiation) medium was removed, and the lymphocytes were tested for MuLV content by cocultivation with Swiss, SC-1, and BALB/c fibroblasts. In the presence of the B cell mitogen lipopolysaccharide (LPS), the ability of spleen and lymph node cells to replicate WN 1802 B was greatly enhanced. In the representative experiments in Table 2, virus replication was 16 times greater in normal spleen cells stimulated by LPS than in untreated cells. Replication of virus was also increased 22-fold in LPS stimulated lymph node cells. On the other hand, viral replication was not stimulated in cells treated with the thymus-derived (T) cell mitogen PHA; in fact inhibition was seen in some cases. BALB/c nu/nu mice are congenitally athymic, though their B cell functions are normal. Spleen cells from nude mice can be infected by WN 1802 B. Replication is enhanced in cells treated with LPS, and this enhancement is proportional to the extent of DNA stimulation induced by the mitogen. Neither DNA synthesis nor viral replication are stimulated in nu/nu spleen cells exposed to the T cell mitogen PHA. The stimulatory effects of the B cell mitogen and the inhibitory effect of T cell mitogens on viral replication are probably mediated through an effect on the host cell rather than through a direct effect on the virus, because mitogen was no longer present in large quantities at the time of viral addition as the medium was replaced or treated with inhibitors. Furthermore, when fibroblasts were treated with LPS or PHA, no effect on exogenous MuLV replication was seen (Table 2). A series of experiments performed to investigate the effects of LPS, PHA, and another T cell mitogen, Con A, on both DNA synthesis and WN 1802 B replication are summarized in Table 3. Results of four experiments in which LPS-treated spleen cells were exposed to exogenous MuLV are presented. In three experiments, B-tropic MuLV content of LPS-treated spleen cells was increased 2-fold, 4-fold, and 16-fold over untreated cells. (In one experiment this stimulation was only 1.36-fold.) Stimulation of MuLV replication was even more dramatic in lymph node cells treated with LPS, probably due to the lower base line replication of exogenous MuLV in unstimulated lymph node cells compared to spleen cells, as indicated in Table 1. LPS-

Microbiology: Ruddle et al.

3716


Table 2. Stimulation of WN 1802 B replication in spleen and lymph node cells treated with LPS

DNA synthesist Exp. no.

Cells

1

Normal spleen Normal spleen Normal spleen Normal lymph node Normal lymph node Normal lymph node nu/nu spleen nu/nu spleen nu/nu spleen Fibroblasts Fibroblasts Fibroblasts

2

3 4

Mitogen*

cpm

LPS PHA

11,077 60,483 79,550 6,618 47,110 46,768

LPS PHA LPS PHA

102,142 2,169 NT§ NT NT

LPS PHA

95,619

WN 1802 replication

E/C 5.5 7.2 7.1 7.1 18.2 0.3

XC PFU

30,000 490,000 210 285

6,400 165

1,300 29,000 400 210 230 240

E/Ct 16.3 0.007 22.45 0.57 22.3 0.3

1.1 1.1

Cell cultures of BALB/c spleen and lymph node were established with and without mitogen. After 48 hr, medium was removed, 0.3 ml of WN 1802 B (0.004 PFU/cell, Exp. 1 and 0.001 PFU/cell, Exp. 3) was added; virus was absorbed to cells for 1 hr, and then an additional 0.6 ml medium was added. Control cultures that received no exogenous MuLV were also maintained. Four days later, the lymphoid cells were tested for release of endogenous MuLV, and replication of the added WN 1802 B as in the legend accompanying Fig. 1. BALB/c fibroblasts (1 x 105) were infected with 0.1 ml of WN 1802 B (0.002 PFU/cell), irradiated after 1 week, and overlaid with XC cells. DEAE-dextran was not used. No endogenous MuLV was detected in these experiments. * Ten micrograms of LPS or 18 Atg of PHA was added per ml of culture. t DNA synthesis was determined between 48 and 72 hr of cell culture after addition of 1 MC, of [3H]dT and trichloroacetic acid precipitation. E/C is the ratio of mitogen stimulated DNA radioactivity/control DNA radioactivity. t The E/C ratio for WN 1802 B replication was calculated as follows: the number of plaques produced by cells after exposure to both mitogen and WN 1802 B was divided by the number of plaques produced after exposure to virus alone. § Not tested.

stimulated lymph node cells which were subsequently exposed to exogenous MuLV for 4 days, contained 8 times, 10 times, or 22 times more virus than unstimulated cells. These results are in striking contrast to those obtained when cells were treated with T cell mitogens PHA and Con A, and then exposed to MuLV. In one experiment out of eight, virus replication was 1.7-fold higher in PHA treated spleen cells; in two of eight experiments, replication was the same, and in the remainder it was inhibited in mitogen-treated cells. The ability of lymph node cells to support MuLV replication was also uniformly inhibited after PHA or Con A treatment. Table 3. Average stimulation of WN 1802 B replication in normal BALB/c spleen and lymph node cells treated with mitogens

WN 1802 B replication

Cells

Mitogen*

Spleen Spleen Lymph node Lymph node Lymph node

LPS PHA LPS PHA Con A

Average stimulation No. of exp./ total exp.t (E/C) 5.97 0.60 13.61 0.34 0.37

3/4 1/8 3/3 0/5 0/3

The experimental conditions for studying viral replication are those described in the legend accompanying Table 2. No release of endogenous ecotropic MuLV was observed. * Ten micrograms of LPS, 18 g of PHA, or 5 ,g of Con A was added per ml of culture as indicated. t Stimulation of WN 1802 B replication was calculated as in the legend accompanying Table 2. The values obtained in individual experiments were pooled to calculate the average stimulation. The values are the number of experiments with E/C ratios greater than 1.5/total number of experiments.

Release of endogenous ecotropic virus was not stimulated after mitogen treatment in any of the experiments depicted in Tables 2 and 3, nor in several additional experiments: seven in which spleen or lymph node cells were treated with LPS, sixteen with PHA, five with Con A, and seven mixed leukocyte reactions. Of 52 experiments, 47 were negative for MuLV release. In the remaining five cell preparations, there were never more than two plaques of endogenous MuLV, whether or not the cells were treated with mitogen. Though another group has reported activation of XC detectable MuLV in the mixed leukocyte reaction (10), others have not observed ecotropic virus in lymphocytes stimulated by the mixed leukocyte reaction (11) or LPS (12) with the reverse transcriptase assay. However, they did note an increase in enzyme activity due to xenotropic virus (which would not be scored with the XC test). Preferential replication of MuLV in nonadherent and

nonphagocytic cells

Organs which contain cells capable of replicating exogenous MuLV are admixtures of numerous cell types. These include T cells, B cells, hematopoietic cells, fibroblasts, and macrophages. Circumstantial evidence has been presented above which implicates the B cell in WN 1802 B replication. Results of additional experiments (Table 4) indicate that the bulk of MuLV replication in lymphoid cell populations is supported by cells which do not express properties generally associated with macrophages, fibroblasts, or polymorphonuclear leukocytes. Spleen cells, purified of contaminating phagocytic and/or adherent cells by either of two different methods, retained their susceptibility to infection by exogenous WN 1802 B. One method exploited the adherence to plastic plates of "sticky" cells such as fibroblasts, polymorphonuclear leukocytes, and macrophages. The other was based on the phagocytic capability of macrophages. Cells which ingest iron filings can be removed by applying a magnet to the bottom of a tube which contains



Table 4. WN 1802 B infection of spleen cells depleted of macrophages and fibroblasts XC infectious centers/culture

Exp. no.

Depletion method

Untreated spleen cells

Purified cells

Adherent cells

1

Adherence to plastic (2 hr*) Adherence to plastic (48 hr*)

100

440

NT:

1225 2700

1075 3900

26 NT

2

3

Magnett

The protocol for assessing WN 1802 B replication was identical to that accompanying Table 2. There was no release of endogenous

ecotropic MuLV. * Spleen cells (2 X 106) were placed in 60 mm plastic plates, and incubated at 37°. Supernatants ("purified cells") were removed and placed in culture. Virus was added at 48 hr. "Adherent cells" were those which remained attached to the plastic plates. t Spleen cells were incubated with 0.49% carbonyl iron for 30 min. Phagocytic cells were removed by placing a 15 cm Alnico bar magnet at the bottom of the tube. Supernatant cells ("purified cells") were washed and placed in culture. Virus was added at 48 hr. I Not tested.

them. The bulk of virus replication occurred in cells which were nonadherent and nonphagocytic.

DISCUSSION Evidence has been presented that WN 1802 B, an isolate of murine leukemia virus from a BALB/c mouse, is capable of replicating in lymphoid cells in vitro. Though we have not as yet precisely identified the permissive cell on the basis of surface antigen markers, we infer that it is the bone marrow-derived lymphocyte. This conclusion is based on four types of evidence: (i) replication of exogenous WN 1802 B is stimulated in cells treated with LPS, a B cell mitogen, and is inhibited in cells treated with T cell mitogens, PHA and Con A; (0i) spleen cells from nu/nu mice, which lack T cells support WN 1802 B replication; (iii) an organ distribution analysis indicates that cells which can be infected by MuLV are found in high concentrations in bone marrow and spleen (which have correspondingly high concentrations of B cells) and in lower concentrations in thymus and lymph node (with higher proportions of T cells); (iv) the virus preferentially infects nonadherent and nonphagocytic cells. The mechanism by which LPS stimulates WN 1802 B replication in spleen and lymph node cells must be due in part to its mitogenic effect, because it has been amply demonstrated that DNA synthesis is required for MuLV replication in fibroblasts (13). Our experiments demonstrate, however, that mere stimulation of DNA synthesis is not sufficient because T cells stimulated with PHA, Con A, or specific antigen (N. Ruddle, in preparation) were not more permissive than normal cells. This suggests that preferential infection of a particular cell must occur before replication proceeds. The inhibition of WN 1802 B replication in lymphoid cells treated with the T cell mitogens, PHA, and Con A is currently under investigation. One possibility is that interferon is produced by stimulated T cells which inhibits MuLV replication (14); another possibility is that PHA stimulates a population of cells which can bind virus but not replicate it, in-effect diluting the virus concentration for the permissive B cell. It is unlikely

3717

that PHA and Con A are inhibiting MuLV by binding virus, a mechanism which could be inferred from the recent report of Ishizaki and Bolognesi (15). These authors observed inhibition of Rous sarcoma virus infection of fibroblasts after virus was mixed with lectin at concentrations 50 to 200 times greater than those used to stimulate cells in the experiments reported here. Stimulation of virus replication in lymphoid cells after mitogen treatment was first described by Wheelock et al. (16). It has become apparent that each animal virus exhibits a tropism for a particular lymphocyte cell type. For example, viruses which replicate in T cells include vesicular stomatitis virus (17) and herpes- saimiri virus (18), whereas B cells are infected by Epstein-Barr virus (19), cytomegalovirus (20), and the naturally occurring murine leukemia virus, WN 1802 B. Cerny has also reported that Friend leukemia virus is associated with B memory cells in VIvo (21), and replicates in LPS stimulated cells in vitro (22). It will be interesting to determine whether other types of murine leukemia viruses will preferentially replicate in T cells in vitro. Haas and Hilgers (3) have demonstrated that thymocytes can be infected by radiation leukemia virus but did not determine whether replication could be enhanced in stimulated T cells. The studies reported here have broad implications for the understanding of certain types of malignancy, because stimulation of a lymphoid cell by an agent as innocuous as a bacterial cell wall component can markedly enhance its ability to be infected by an oncogenic virus. Such a mechanism may be operative in two systems in which tumor development is associated with lymphocyte stimulation-the increased incidence of lymphoreticular tumors associated with activation of MuLV in the course of the graft versus host reaction in mice (5), and the increased incidence of lymphoid malignancies associated with the chronic immunological stimulus of the kidney transplant (23). We thank Ms. Penny Kauer for excellent technical assistance. This work was supported by the American Cancer Society Institutional Grant IN-31-L-7, American Cancer Society Grant ACS-IM-19, and United States Public Health Service Grants AI-08614, CA-15557, and CA-16038. N.H.R. was a fellow of the American Society, ACS PF846. 1. Gross, L. (1958) Cancer Res. 18, 371-381. 2. Rosenberg, N., Baltimore, D. & Scher, C. D. (1975) Proc. Natl. Acad. Sci. USA 72, 1932-1936. 3. Haas, M. & Hilgers, J. (1975) Proc. Natl. Acad. Sci. USA 72,

3546-3550. 4. Datta, S., Melief, C. S M. & Schwartz, R. S. (1975) J. Natl. Cancer Inst. 55, 425-432. 5. Armstrong, M. Y. K., Ruddle, N. H., Lipman, M. & Richards, F. F. (1973) J. Exp. Med. 137, 1163-1179. 6. Hartley, J. W. & Rowe, W. P. (1975) Virology 65, 128-134. 7. Hartley, J. W., Rowe, W. P. & Huebner, R. J. (1970) J. Virol. 5, 221-225. 8. Rowe, W. P., Pugh, W. E. & Hartley, J. W. (1970) Virology 42, 1136-1139. 9. Ruddle, N. H., Armstrong, M. Y. K. & Richards, F. F. (1974) J. Immunol. 112, 706-715. 10. Hirsch, M., Phillips, S. M., Solnik, C., Black, P. H., Schwartz, R. S. & Carpenter, C. B. (1972) Proc. Natl. Acad. Sci. USA 69, 1069-1072. 11. Sherr, C. J., Lieber, M. M. & Todaro, G. J. (1974) Cell 1, 5558. 12. Moroni, C. & Schumann, G. (1975) Nature 254,60-61. 13. Temin, H. M. & Baltimore, D. (1972) Adv. Virus Res. 17,

129-186. 14. Epstein, L. B., Kreth, H. W. & Herzenberg, L. A. (1974) Cell. Immunol. 12, 407-421.

3718


15. Ishizaki, R. & Bolognesi, D. P. (1976) J. Virol. 17, 132-139. 16. Wheelock, E. F., Toy, S. T. & Stjernholm, R. L. (1971) Prog. Immunol. 1, 787-801. 17. Kano, S., Bloom, B. R. & Howe, M. L. (1973) Proc. Nati. Acad. Sci. USA 70,2299-2303. 18. Wallen, W. C., Neubauer, R. H., Rubin, H. & Cicmanec, J. L. (1973) J. Natl. Cancer Inst. 51, 967-975.

Proc. Natl. Acad. Sci. USA 73 (1976) 19. Jondal, M. & Klein, G. (1973) J. Exp. Med. 138, 1365-1378. 20. Olding, L. B., Jensen, F. C. & Oldstone, M. B. A. (1975) J. Exp. Med. 141, 561-572. 21. Cerny, J. & Waner, E. (1975) J. Immunol. 114,571-580. 22. Cerny, J., Hensgen, P. A., Fistel, S. H. & Demler, L. M. (1976) Int. Jour. of Cancer 18, 189-196. 23. Hoover, R. & Fraumeni, J. F. (1973) Lancet ii, 55-57.