May 8, 1979 - (C3H/An X B6-Lyt-2.1)Fl aERLb (B6). C58 aC58.CE-Lyt-3.2:DS. B6 aMeth A (BALB/c). B6 aSL2 (DBA/2). Antigen detected. Alloantigens: Lyt-2.2.
Proc. Natl. Acad. Sci. USA Vol. 76, No. 7, pp. 3486-3490, July 1979 Immunology
Definition of a unique cell surface antigen of mouse leukemia RL61 by cell-mediated cytotoxicity (T lymphocyte/5tCr release assay/competitive inhibition/antibody blocking/tumor immunology)
EIICHI NAKAYAMA*, HIROSHI SHIKUt, TOSHITADA TAKAHASHIt, HERBERT F. OETTGEN*, AND LLOYD J. OLD* *Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021; tFirst Department of Internal Medicine, Nagoya University School of Medicine, Nagoya, Japan; and tLaboratory of Experimental Pathology, Aichi Cancer Center Research Institute, Nagoya, Japan Contributed by Lloyd J. Old, May 8, 1979
ABSTRACT BALB/c x-ray-induced leukemia RLfl is strongly immunogenic for (BALB/c X C57BL/6)F1 mice. Transplants of RLd1 regressed after initial growth, and after tumor regression mice could resist repeated inocula of 107 R IM cells. Spleen cells from immunized mice after in vitro stimulation with 1L11 were cytotoxic for RLd1 cells in 3-hr 5tCr assays. Pretreatment of immune spleen cells with Thy-1, Lyt-2, or Lyt-3 antisera and complement eliminated cytotoxic activity, indicating that effector cells for RL61 lysis are T cells. Tests with other target cells showed little or no cytotoxicity. Analysis of the specificity of T-cell killing of RL61 by competitive inhibition assays with unlabeled cels indicated that only RL1 could inhibit killing; other BALB/c tumors (13 x-ray or murine leukemia virus-induced leukemias and three myelomas) failed to inhibit lysis of RL11. A range of alloantisera and heteroantisera were tested for their capacity to block lytic activity in the absence of added complement. H-2d antisera and Lyt-2 and -3 antisera blocked lysis, the latter at the level of the effector cell. Antisera to other cell surface alloantigens, murine leukemia virus-related antigens, and immunoglobulins did not block 1111 lysis. Thus, T cells from mice immunized against RL1 recognize an individually distinct or unique antigen that does not appear to be related to any of the serologically defined cell surface determinants of R121. In its restriction to a single leukemia, the RL81 antigen resembles the individually distinct antigens of chemically induced tumors and other tumor types of rodents. Serological analysis has revealed several categories of cell surface antigens on murine leukemia cells (1). In addition to antigens coded for by genes in the H-2 complex, the predominant categories are differentiation alloantigens (Thy-1, Lyt series), antigens related to murine leukemia viruses (MuLV) [GIX, GCSA, G(RADA1), etc.], and antigens belonging to the TL system. As yet, antigens that can be considered leukemia-specific by virtue of their absolute restriction to leukemia cells have not been identified in the mouse. Certain TL antigens (TL. 1 and 2) have long been regarded as leukemia-specific in strains, such as C57BL/6 (B6), that do not express these antigens during normal life. However, recent evidence shows that TL.1 and TL.2 are expressed during the preleukemic period of x-ray leukemogenesis before the appearance of overt leukemia (2), and this is also true for the TL.4 component, which, unlike other TL components, is never expressed as a normal differentiation antigen. In contrast to this impressive body of information about serologically defined surface antigens of leukemia cells, we know little about the non-H-2 antigens that cytotoxic T cells recognize on leukemia cells. In vitro analysis of cellular immunity to leukemia cells in syngeneic or semisyngeneic mice has focused primarily on MuLV-related surface antigens, and The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
most studies have been directed to understanding the nature of the immune reaction and its relationship to leukemia growth and rejection, rather than to a strict definition of the cell surface antigens being recognized. With advances in methodology for the study of T cells in vitro, the specificity of cell-mediated reactions can be analyzed with increasing precision, and techniques are now available (3, 4) that provide the sort of information about specificity in T-cell assays that absorption tests provide for defining the specificity of serological reactions. In the present study, we have investigated the specificity of the cytotoxic T-cell response to the x-ray-induced BALB/c leukemia RL61. This leukemia was selected for these initial studies because of its strong immunogenicity in H-2-compatible F1 mice (5) and because its surface phenotype has been the object of considerable serological study (1, 5). MATERIALS AND METHODS Mice. Inbred and F1 hybrid mice were obtained from our breeding colony or from the Jackson Laboratory. Derivation of the B6 Lyt congenic stocks is described in ref. 6. Tumors. RL1 is an x-ray-induced leukemia of BALB/c origin (5). RL63, RL93, RL84, RL94, RL86, RL96, RL97, RL98, and RL99 are BALB/c leukemias induced by x-ray in 1977 and obtained from Ronald Ellis, Sloan-Kettering Institute. BALBRV1, BALBRV2, and BALBRV3 leukemias (7) were induced by injection of neonatal BALB/c mice with radiation leukemia virus and were obtained from Elisabeth Stockert, Sloan-Kettering Institute. Other tumors have been described (8-11). Antisera. The Lyt and other antisera used in this study (see Table 5) have been described (5, 12). Immunization of Mice with RL1. (BALB/c X C57BL/ 6)F1 (CB6F,) mice were immunized by intradermal injection of 5 X 105 viable RL61 cells. Viability of the cell suspension, as judged by exclusion of trypan blue, approached 100%. The inoculum grew to a tumor of 5-10 mm in diameter in 2 weeks, and then regressed within 3-4 weeks. Spleen cells for in vitro sensitization were harvested from immunized mice within 2 months after the tumors had regressed. In Vitro Sensitization of Spleen Cells. Responder spleen cells (40 X 106) were cultured with 4 X 106 stimulator RL61 cells (irradiated with 10,000 rads; 1 rad = 0.02 J/kg) in Falcon 3013 flasks. The culture medium was Eagle's minimum essential medium supplemented with 10% heat-inactivated fetal calf serum, 2 mM glutamine, 1% nonessential amino acids, 100 units of penicillin per ml, 100 ,tg of streptomycin per ml, and 50 ,gM 2-mercaptoethanol (complete medium). The cells were harvested after 6 days of culture and used as effector cells in assays for cell-mediated cytotoxicity. Abbreviations: B6, C57BL/6; CB6F1, (BALB/c X C57BL/6)FI; MuLV, murine leukemia virus. 3486
Immunology: Nakayama et al.
Proc. Nati. Acad. Sci. USA 76 (1979)
b10 20 30 Days after inoculation of 5 X
105 BALB/cRLdl cells
FIG. 1. Growth of BALB/c leukemia RL61 in BALB/c (Left) and CB6F1 (Right) male mice 10-12 weeks of age. t indicates death from progressive RL61 growth.
Pretreatment of Effector Cells with Thy-i and Lyt-2,3 Antisera and Complement. The method has been described (12). Cell-Mediated Cytotoxicity Assay. Target cells were labeled with 5'Cr by incubating 5 X 106 tumor cells with 100 jACi of Na251CrO4 (New England Nuclear) in 0.5 ml of complete medium for 30 min in a 37°C water bath. After labeling, the cells were washed twice in complete medium before use. In direct assays, 2 X 104 labeled target cells in 100 ,ul were incubated with serial dilutions of the effector cell suspension (in 100 ,Al), resulting in a range of effector cell to target cell ratios. In competitive inhibition assays, 2 X 104 labeled RL61 cells in 100
Al were incubated with different numbers of unlabeled inhibitor cells and 50 X 104 effector cells, each in 50 A1I of medium. The mixtures were then centrifuged at 200 X g for 1 min before incubation. In antibody blocking assays, serially diluted antiserum (50 ,ul), effector cell suspension (50,ul), and 2 X 104 labeled RL61 cells (50 ,l) were incubated together; a fixed effector cell to target cell ratio was used in these studies. Assays were performed in duplicate in microtiter plates (Dynatech Laboratories, Alexandria, VA). The plates were incubated for 3 hr at 37°C in an atmosphere of 5% CO2. The supernatant was then removed from each well by Titertek (Flow Laboratories, Rockville, MD) and assayed for radioactivity in a Packard 5210 scintillation counter. Percent specific lysis was calculated by using the following equation: (a - b/c - b) X 100, where a = cpm in the supernatant of target cells mixed with effector cells, b = cpm in the supernatant of target cells incubated alone, and c = cpm after lysis of target cells with Nonidet P-40. RESULTS Growth of RL61 in Syngeneic and F1 Hybrid Mice. Fig. 1 illustrates the growth of RL61 in BALB/c and CB6F1 male mice after intradermal injection of 5 105 RL51 cells. In BALB/c mice (Fig. 1 left), there was progressive growth of X
RL81, and mice generally died 4-5 weeks after RL81 transplantation. In CB6F1 mice (Fig. 1 right), initial RL61 growth was followed by tumor regression; subsequent tumor recurrence was exceedingly rare. CB6F1 mice that had rejected RL61 resisted further challenge with inocula of 107-108 RL61 cells. Generation of Cytotoxic Effector Cells Reactive with RL61. Table 1 summarizes the results of one of several tests aimed at defining conditions for optimal sensitization of CB6F1 spleen cells against RLU1 cells. Spleen cells were obtained from (i) normal CB6F1 mice, (ii) CB6F1 mice 1-2 weeks after RL61 regression, and (iii) CB6F1 mice 15 days after inoculation of RL51 cells (at a time when transplantable RL6L cells can still be isolated). The cytotoxic activity of these cells for RL61 was tested before and after in vitro sensitization with irradiated RL61 cells. In vivo sensitization alone resulted in cells with no cytotoxic activity. In vitro sensitization alone resulted in cells with a low level of cytotoxic activity. Spleen cells sensitized by RL81 both in vivo and in vitro showed maximal reactivity. Spleen cells from CB6F1 mice during the later phase of RL61 rejection (15 days) were no less reactive than spleen cells from mice that had rejected RL81 1-2 weeks previously. Cytotoxic effector cells could also be generated by in vitro sensitization of spleen cells from BALB/c and (BALB/c X A)F1 mice 15 days after RL61 implantation. In contrast to the strong immunogenicity of RL61 in CB6F1 mice, RL61 grows progressively in BALB/c and (BALB/c X A)F1 mice. Thus, cytotoxicity for RL81 can be generated in susceptible as well as resistant mice.
Because highly active effector cells could be reproducibly generated by in vitro sensitization of spleen cells from CB6F1 mice after RL61 rejection, these cells were chosen for additional characterization. T-Cell Characteristics of Cytotoxic Effector Cells for RL61. The cytotoxic activity of sensitized CB6F1 spleen cells could be abolished by pretreatment with Thy-1.2, Lyt-2.2, or Lyt-3.2 antisera and complement. The specificity of these reactions was established by tests with antisera to allelic Thy-i and Lyt-2 and -3 products and by absorption tests with thymocytes from mice expressing the corresponding or alternative Thy-i and Lyt-2 and -3 alleles. Specificity of CB6F1 Effector Cells Generated against RL61: Direct Assays with Other Target Cells. Table 2 summarizes tests with T-cell leukemias of BALB/c, A strain, C57BL, and AKR origin and BALB/c myelomas. The predominant reactivity of CB6F, effector cells sensitized against RL61 was directed against RL61. At higher effector cell to target cell ratios, some cytotoxicity for MOPC-70A, MPC-il, and RADA1 was observed. In each instance, however, cytotoxicity for these cells was much weaker than cytotoxicity for RL61.
Table 1. Generation of cytotoxic effector cells reactive with BALB/c leukemia RLl1
Source of effector cells
Method of sensitization In vivo* In vitro
CB6F1 A B +
(BALB/c X A)F1
Specific lysis, %
28 84 92
27 78 86
10 70 83
13 66 80
BALB/c B + 70 62 43 25 * A, spleen cells obtained 1-2 weeks after rejection of RL61; B, spleen cells obtained 15 days after injection of RL61. Effector cell to target cell ratio.
Immunology: Nakayama et al.
Proc. Nati. Acad. Sci. USA 76 (1979)
Table 2. Cytotoxicity of CB6F1 effector cells generated against BALB/c leukemia RLI1: Direct assays with other target cells Specific lysis, % 12 50* 25 6 40 20 10 Exp. no. Target cells Strain of origin 14 56 42 27 I RL81 BALB/c 0 0 0 0 RL86 BALB/c 1 1 5 2 RL98 BALB/c BALBRV3 4 4 0 BALB/c 0
RL1 MOPC-70A MPC-11 WEHI-22
BALB/c BALB/c BALB/c BALB/c
56 12 14 3
43 6 9 5
31 5 7 1
20 2 4 3
RLd1 MOPC-70A MPC-67
BALB/c BALB/c BALB/c
65 8 7
58 7 1
44 3 0
36 3 0
RL81 RADA1 K36
BALB/c A AKR
59 21 4
50 21 8
39 9 5
25 7 0
RLd1 BALB/c EL4 C57BL AKSL2 AKR MOPC-70A BALB/c * Effector cell to target cell ratio.
44 1 2 7
Specificity of CB6F1 Effector Cells Generated against RL81: Competitive Inhibition Assays. Fig. 2 illustrates two competitive inhibition assays with CB6F1 effector cells. Lysis of labeled RLf1 cells was inhibited by unlabeled RL81 cells, but not by other cell types of normal or leukemic origin. Table 3 lists the range of cell types that have been tested in competitive inhibition assays. Without exception, the only cell that inhibited RL81 lysis of CB6F1 effector cells was RL81. Ten xray-induced leukemias of BALB/c origin, the cell type most closely related to RL81, were included in this series; none inhibited RL81 cytotoxicity.
29 0 0 3
1 0 1
Blocking Tests with Alloantisera and Heteroantisera to Cell Surface Antigens of Mouse Lymphoid Cells. As an approach to defining the determinants recognized by CB6F1 effector cells on RL61 cells, an attempt was made to block effector cell cytotoxicity with antisera to cell surface antigens expressed by RL81 cells and other mouse lymphoid cells. (In blocking tests, antisera are tested without the addition of exogenous complement.) Table 4 illustrates three such blocking tests and Table 5 summarizes the tests that have been done. Antisera against H-2d antigens shared by RL61 cells and CB6F1 effector cells showed a modest degree of blocking. Antisera against
Table 3. Specificity of CB6F1 effector cells generated against BALB/c leukemia RLd1: Summary of competitive inhibition assays Cells with inhibitory Cells with no inhibitory activity activity MuLV (RadLV)-induced Spontaneous leukemias: Spleen cells:
C3H/An B6 DBA/2 I RF BALB/c Con A blasts B6 Con A blasts
Thymocytes: AKR B6 BALB/c
Con A, concanavalin A.
ASL1 AKSL2 Pu 5
X-ray-induced leukemias: RLd3 RL93 RLd4 RL94 RLe6 RLQ6 RLQ7 RLQ8 RL99 WEHI-22 RADA1 ERLD
BALB/c BALB/c BALB/c BALB/c BALB/c BALB/c BALB/c BALB/c BALB/c BALB/c A B6
leukemias: BALBRV1 BALBRV2 BALBRV3
BALB/c BALB/c BALB/c
MuLV (Gross)-induced leukemias: B6 EdG2
Chemically induced leukemia: EL4
Myelomas: MOPC-70A MPC-67 HOPC-1
BALB/c BALB/c BALB/c
Immunology: Nakayama et al.
Proc. Nati. Acad. Sci. USA 76 (1979)
30 15 7.5 3.8 Tumor cells 180 90 45 22.5 Thymocytes Inhibitor cell to target cell ratio
Inhibitor cells: * None A R Ld4 O RLd1 ° RL94 * RLd3 * RL96 * RL93 RL97
Inhibitor cells: A BALB/c thymocvtes 0 B6 thymocytes * AKR thymocytes *BALBRV2
® None 0 RLcd1 Pu 5
FIG. 2. Competitive inhibition assays with CB6F1 effector cells generated against BALB/c leukemia RL51. Unlabeled inhibitor cells were added at different ratios to 2 X 104 51Cr-labeled R121 target cells and the lysis of RLM1 cells by CB6F1 effector cells was measured. Effector cell to target cell ratio in these tests was 25.
Lyt-2.2 and Lyt-3.2 also blocked cytotoxicity. Past studies (12) have shown that blocking by Lyt-2 and -3 antisera is at the level of the effector cell and not the target cell, a possibility in the case of RLd1 because RL61 cells express Lyt-2 and Lyt-3. Antisera detecting other alloantigens, MuLV-related antigens, immunoglobulin determinants, and the X. 1 antigen (see Discussion) had no blocking activity. DISCUSSION Of the range of normal and malignant cells we examined in competitive inhibition assays, only RLd1 cells inhibited T-cell lysis of RL61. If H-2 restriction is operative in this system as it has been shown to be in a variety of other T-cell lytic systems
Table 4. Blocking tests with alloantisera and heteroantisera to cell surface antigens of mouse lymphoid cells Exp. Specific lysis,* % no. Antiserum 1:8 1:16 1:32 1:4t I aThy-1.2 65 61 64 61 52 53 58 aTL.1,2,3 58 aPC.1 53 66 56 54 29 alH-2d 33 37 51 50 alH-2b 52 52 59 aX.1 57 59 58 57 50 52 60 58 CB6Fj normal serum None 63 II
aLyt-1.1 aLyt-1.2 aLyt-2.1 aLyt-3.2
59 61 55 25 66
57 58 57 26
55 55 58 48
53 60 60 53
III aNTD 54 56 56 api5 55 52 55 57 55 ap30 57 56 agp70 66 64 None 61 a, Anti. * Effector cell to target cell ratio in these tests was 25. t Antiserum dilution.
57 59 58 76
(13), then the inhibition tests with BALB/c cells are most pertinent, and these indicate that the RL61 antigen detected by cytotoxic T cells is not present on normal BALB/c thymocytes or spleen cells or on 17 other BALB/c tumors. Duprez et al. (14) have also studied the specificity of T-cell killing of RL61 and similarly found that other BALB/c tumors (six tested) did not inhibit RL61 lysis. Of the seven allogeneic cell types tested by Duprez et al., only one, the DBA/2 (H-2d) leukemia P815,
Table 5. Blocking tests with alloantisera and heteroantisera to cell surface antigens of mouse lymphoid cells: Summary of results Blocking activity No blocking activity Antiserum Antigen detected Antiserum Antigen detected Alloantigens: Alloantigens: (C3H/An X B6-Lyt-2.1)Fl aERLb (B6) Lyt-2.2 CB6F1 aB6-Lyt-1.1 Lyt-1.1 C58 aC58.CE-Lyt-3.2:DS Lyt-3.2 C3H/An aC3H.CE-Lyt-1.2:DS Lyt-1.2 B6 aMeth A (BALB/c) H-2d B6-H-2k aB6-H-2k.CE-Lyt-2.1:DS Lyt-2.1 B6 aSL2 (DBA/2) H-2d BALB/c aERLD (B6) H-2b (A-Thy-1.1 X AKR-H-2b)Fl aASL1 (A strain) Thy-1.2 (A-TL- X B6)F1 aASL1 (A strain) TL.1,2,3; Qa-1 (DBA/2 X B6)F1 aMOPC-70A (BALB/c) PC.1 Rat (W/Fu X BN)F1 aW/Fu leukemia (C58NT)D Goat aMuLV (AKR)pl5 Rabbit aMuLV (AKR)p30 Goat aMuLV (AKR)gp7O CB6Fj aRLS1
MuL V-related antigens: GIx (and a range of other MuLV-related determinants) MuLV (AKR)pl5 MuLV (AKR)p30 MuLV (AKR)gp7O X.1
Rabbit amouse IgG Rabbit amouse IgM
Immunoglobulin determinants: Mouse IgG Mouse IgM
Normal mouse serum: B6, BALB/c, CB6Fj
ay, Anti; p, protein; gp, glycoprotein.
Proc. Natl. Acad. Sci. USA 76 (1979)
Immunology: Nakayama et al.
showed partial inhibition. This highly restricted occurrence of the RL31 antigen suggests that it might belong to the family of individually distinct or unique cell surface antigens that were first recognized as tumor-specific transplantation antigens on methylcholanthrene-induced sarcomas of the mouse (15) and that are now known to be present on many types of mouse and rat tumors. These antigens show a remarkable degree of polymorphism, even two tumors induced in the same mouse having different transplantation antigens (16). Until recently, efforts to define these antigens either serologically or by in vitro tests of cell-mediated immunity have met with little success. A major problem in this regard is the frequent and unpredictable occurrence of MuLV-related antigens in tumors of the mouse (17-21), and past studies that have not taken MuLV-related cell surface antigens into account are difficult to interpret. However, even in cases in which this problem has been controlled, it has been difficult to demonstrate antibody to these individually distinct antigens (21). DeLeo et al. (20, 22) have recently identified such antigens on two BALB/c-induced sarcomas (Meth A and CMS4) by cytotoxic antibody produced in syngeneic or semisyngeneic hosts, but both these instances may turn out to be exceptions rather than the rule. In one case, the tumor (Meth A) had a long transplantation history, and in the other, an unusually long period of immunization (>7 months) was required before antibody to CMS4 appeared. One possible explanation for this difficulty in producing antibody against these antigens is that recognition and rejection of these tumors is primarily in the province of cellular immunity and does not result generally, even after extensive immunization, in the formation of antibody. This situation may well be similar to findings with certain H-2 mutant stocks, in which strong incompatibility with the parental strain is easily demonstrable by graft rejection in vivo and cellular immune reaction in vitro, but does not lead to the formation of detectable humoral immunity (23). In addition to the possibility that the RL1 antigen may be related to the individually distinct or unique antigens of sarcomas and other nonlymphoid mouse tumors, extensive antigenic diversity on normal and malignant T cells could also result from the presence of T-cell receptors marked by idiotypic diversity. On leukemias such as RL31, which result from a clonal expansion of a subpopulation of normal T cells presumably expressing the same idiotype as the leukemia, the T-cell receptor would have the appearance of a highly restricted tumor-specific antigen. Exploration of this basis for the RLd1 antigen awaits additional understanding of the nature of the T-cell receptor and its expression on leukemia cells. Serological analysis of sera from CB6F1 mice that had been repeatedly immunized with RL31 has not revealed any antigens that are restricted to RL61. Rather, these sera identified a surface antigen of RL31, designated X.1, that was shared by a number of leukemias, both syngeneic and allogeneic, as well as by normal lymphoid cells from several mouse strains (5). The pattern of X. 1 expression, particularly its occurrence in mouse strains with high leukemia incidence (AKR, C58), suggested that X.1 is related to a subclass of endogenous MuLV, although direct evidence for this supposition is lacking. In view of the serological findings, it seemed possible that X. 1 might be the target for T-cell lysis of RL1, but this does not appear to be the case; T-cell lysis was not inhibited by other BALB/c X. 1+ cells (e.g., RLQ6, RLQ8) or blocked by anti-X.1 sera, indicating that the predominant antigens recognized by humoral and cellular immunity are different. At this point, we do not know whether rejection of RL21 by
CB6F1 mice is due to recognition of X. 1 or recognition of the
individually distinct antigen detected by T-cell lysis, or whether these two antigens that have been defined in titro are involved at all. In this regard, it will be important to determine whether RL81 and more recently derived BALB/c leukemias sharing the X.1 antigen have individually distinct or crossreacting transplantation antigens in rejection tests in F1 mice. In addition, studies of other T-cell leukemias are needed to show whether the detection of an individually distinct antigen on RL31 by T-cell cytotoxicity is a peculiarity of this leukemia or is a general feature of T-cell leukemias of the mouse. We thank Drs. E. Stockert, R. Ellis, F.-W. Shen, and U. Himmerling for providing selected tumors and antisera, and Mr. R. Geismar for excellent technical assistance. The work was supported by grants from the American Cancer Society (IM-162), the National Cancer Institute (CA-08748), the Sherman Fairchild Foundation, and the Junior Society of Memorial Sloan-Kettering Cancer Center. E.N. is recipient of a fellowship from the Cancer Research Institute, Inc. 1. Old, L. J. & Stockert, E. (1977) Annu. Rev. Genet. 11, 127160. 2. 3. 4.
Stockert, E. & Old, L. J. (1977) J. Exp. Med. 146,271-276. Bloom, B. R. & David, J. R. (1976) in In Vitro Methods in CellMediated and Tumor Immunity (Academic, New York). Haas, W. & von Boehmer, H. (1978) in Current Topics in Microbiology and Immunology (Springer, New York), Vol. 84, pp. 1-120. Sato, H., Boyse, E. A., Aoki, T., Iritani, C. & Old, L. J. (1973) J. Exp. Med. 138,593-606. Shen, F.-W., Boyse, E. A. & Cantor, H. (1975) Immunogenetics 2,591-595. Stockert, E., DeLeo, A. B., O'Donnell, P. V., Obata, Y. & Old, L. J. (1979) J. Exp. Med. 149,200-215. Old, L. J., Boyse, E. A. & Stockert, E. (1963) J. Nati. Cancer Inst.
31,977-986. Old, L. J., Boyse, E. A. & Stockert, E. (1965) Cancer Res. 25, 813-819. 10. Takahashi, T., Old, L. J. & Boyse, E. A. (1970) J. Exp. Med. 131, 9.
1325-1341. 11. Wagner, H. & Rollinghoff, M. (1973) J. Exp. Med. 138, 1-15. 12. Nakayama, E., Shiku, H., Stockert, E., Oettgen, H. F. & Old, L. J. (1979) Proc. Natl. Acad. Sci. USA 76, 1977-1981. 13. Zinkernagel, R. M. & Doherty, P. C. (1977) in Contemporary Topics in Immunoblology, ed., Stutman, O. (Plenum, New York), Vol. 7, pp. 179-220. 14. Duprez, V., Gomard, E. & Levy, J. P. (1978) Eur. J. Immunol.
8,650-655. Old, L. J. & Boyse, E. A. (1964) Annu. Rev. Med. 15, 167186. 16. Globerson, A. & Feldman, M. (1964) J. Nati. Cancer Inst. 32, 1229-1243. 17. Old, L. J. & Boyse, E. A. (1965) Fed. Proc. Fed. Am. Soc. Exp. Biol. 24, 1009-1016. 18. Whitmire, C. E., Salerno, R. A., Rabstein, L. S., Huebner, R. J. & Turner, H. C. (1971) J. Natl. Cancer Inst. 47, 1255-1265. 19. Grant, J. P., Bigner, D. D., Fischinger, P. J. & Bolognesi, D. P. (1974) Proc. Natl. Acad. Sci. USA 71,5037-5041. 20. DeLeo, A. B., Shiku, H., Takahashi, T., John, M. & Old, L. J. (1977) J. Exp. Med. 146, 720-734. 21. Brown, J. P., Klitzman, J. M., Hellstrom, I., Nowinski, R. C. & Hellstrom, K. E. (1978) Proc. Natl. Acad. Sci. USA 75, 955958. 22. DeLeo, A. B., Shiku, H., Takahashi, T. & Old, L. J. (1978) in Biological Markers of Neoplasia: Basic and Applied Aspects, ed., Ruddon, R. W. (Elsevier/North-Holland, New York), pp. 2534. 23. Klein, J. (1978) in Advances in Immunology, eds., Dixon, F. J. & Kunkel, H. G. (Academic, New York), Vol. 26, pp. 55-146.