Resistance to cellular immune response in AKR leukemias

Eur. J. Immunol. 1986.16: 753-759 Immune resistance of AKR leukemias 753

Andreas Schafer and Resistance to cellular immune response in AKR leukemias* Wilhelm Schmidt

Institut fur Medizinische Virologie, Justus-Liebig-Universitat Giessen, Giessen Spontaneous AKR leukemias express murine leukemia virus (MuLV) gag and env

gene-encoded structural proteins on their cell surface. Cytotoxic T lymphocytes (CTL) induced in AKR mice by syngeneic leukemia 369 which expressed high amounts of H-2 antigens recognized viral gag polyproteins in association with H-2K antigens as target antigens. H-2K-negative leukemias were resistant to lysis by AKR/ Gross MuLV-specific CTL and did not induce a cellular immune response. However, they became susceptible after stimulation with interferon. H-2K-positive leukemias induced CTL which were cytotoxic for 369 cells. However, the majority of H-2K- positive leukemias was not lysed by CTL induced by autologous immunization. These leukemias were also resistant to lysis by anti-369 CTL, but could restimulate AKR/ Gross-specific CTL in vitro, and were susceptible to lysis by H-2Kk-restricted CTL against AKR minor histocompatibility antigens. Thus, there could be specific defects of the H-2Kk antigens of these tumors. However, there were also qualitative and quantitative differences in antigenic determinants of the gag target antigens in these leukemias. Therefore, in addition to quantitative reduction of the H-2K restriction elements, qualitative alterations of H-2 antigens or of the viral target antigens may impair T cell cytotoxicity and thus influence leukemogenesis of AKR spontaneous leukemia.

1 Introduction

Cell surface antigens encoded by the major histocompatibility complex (MHC) play an important role in the regulation of T cell immune responses [l]. The major and possibly exclusive function of the MHC class I molecules is to guide the activity of T lymphocytes. These molecules are recognized by cytotoxic T lymphocytes (CTL) in association with viral or tumor-associated antigens and are necessary for T cell activation as well as effector function [2]. Class I molecules are glycoproteins, composed of a heavy chain with a molecular weight of 45 000 and the noncovalently associated P2-micro- globulin with a molecular weight of 12000 [3]. In the mouse the heavy chains of the MHC class I molecules are encoded by the highly polymorphic H-2K, H-2D or H-2L loci. They give rise to the polymorphic H-2K and H-2D, H-2L molecules which are expressed on the surface of virtually all somatic cells. Tumor cells frequently show quantitative alterations in their expression of H-2 molecules compared to celis from the corresponding normal tissues 14-81. It has been demonstrated that reduced expression of one or several H-2 molecular species may allow tumor cells to grow and escape surveillance by CTL [9-111. Loss of other H-2 allelic products can facilitate metastatic spread of certain tumors [5, 121. Restoration of H-2 antigen expression by in vitro transfection of cloned MHC class I genes restored susceptibility to CTL lysis and altered their tumorigenicity and metastatic properties in vivo [13, 141.

[I 53491

* Supported by the Deutsche Forschungsgemeinschaft (SFB 47, Viro- logie). This work is part of the doctorial thesis submitted by A. S. for the degree of Dr. med.

Correspondence: Wilhelm Schmidt, Institut fur Zellbiologie (Tumor- forschung), Klinikum Essen, Hufelandstr. 55, D-4300 Essen 1, FRG

Abbreviations: FCS: Fetal calf serum MHC: Major histocompatibility complex CTL: Cytotoxic T lymphocytes mAb: Monoclonal antibody(ies) MuLV: Murine leukemia virus IFN: Interferon

Thus, the appearance of H-2 loss variants may give a selective advantage for tumor progression. In AKR spontaneous leukemias the H-2K antigen is necessary for the recognition of viral antigens and lysis of tumor cells by CTL [9]. A large portion of AKR primary leukemias and tissue culture-adapted tumor lines showed reduced H-2K antigens and could not be killed by Gross MuLV-specific CTL [9, 151. Thus, loss of H- 2K on AKR leukemias may be important for their pathogenic- ity. However, an equally large portion of primary leukemias did express H-2K, and tumor cells grew in vivo and killed the animals. This suggested that additional mechanisms might be responsible for the escape of these tumors from immune surveillance. In this report we show that H-2K' leukemias induced CTL but could not be lysed by these effector cells.

2 Materials and methods

2.1 Mice

Inbred strains of mice BALB.K, B1O.BR (both H-2k), BIO.A (H-2a), C3H (H-2k), C3H.OH (H-202), AKR (H-2k) and (AKR X BIO.BR)F1 were bred in our colony and used when 6-8 weeks old.

2.2 Cell lines

AKR thymomas were kindly provided by Dr. P. Krammer, Heidelberg, FRG. They had been in tissue culture for 2-61 passages at the time of receipt and were kept in the culture for a further 30 passages. After this new aliquots of cells, frozen immediately upon receipt, were used. Leukemia K-GV induced in BALB.K mice with AKlUGross MuLV was kindly provided by F. Lilly, New York [6]. All the cell lines were grown in RPMI 1640 medium (Gibco, Grand Island, NY), supplemented with 10% fetal calf serum (FCS; Seromed, Munchen, FRG), 2 mM L-glutamine and 5 X M 2-mercaptoethanol.

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1986 0014-2980/86/0707-0753$02.50/0

754 A. Schafer and W. Schmidt

3

Eur. J. Immunol. 1986.16: 753-759

alli 121 365 L39

2.3 Interferon treatment

AKR leukemic cells were stimulated in vitro for 20 h with lo4 unitdm1 of mouse interferon (IFN) type I (a/p), which was obtained by the generosity of Dr. A. Senik, Paris, France, at a specific activity of lo8 units/ml.

2.4 Antisera and monoclonal antibodies (mAb)

mAb 100-5 (H-2Kk), 100-27 (H-2Kk, H-2Dk) and 100-30 (H- 2Kk, H-2Dk) were obtained from H. Lemke, Koln [16]; mAb 15-5-5 (H-2Dk) from K. Ozato, Bethesda, MD [17]; mAb against Thy-1.1 and Thy-1.2 were provided by D. Lane, Lon- don; mAb against Lyt-1.2, Lyt-2.1 and Lyt-2.2 were purchased from NEN Chemicals, Dreieich, FRG; anti-viral mAb 60-35 (mouse), 35-270 (rat) and 42-135 (rat) against AKR MuLV p30 and 35-56 (rat) against gp70 were kindly provided by U. Hammerling, New York, and mAb BBF39.27, BBF59.13 and BBF 59.15 specific for MuLV p30 [ 181 were kindly provided by F. Plata, Pans, France. Goat antisera against rat IgG were purchased from Dinova, Hamburg, FRG.

2.5 Mixed lymphocyte and mixed lymphocyte-tumor cell

For allogeneic stimulation 20 X lo6 Ficoll-purified spleen cells of mice immunized 2-3 weeks previously and 10 x lo6 mitomycin A-treated stimulator spleen cells were cultured in Falcon flasks for 5 days in 10 ml of RPMI 1640 (Gibco, Bonn, FRG) medium supplemented with 10% FCS and L-glutamine. For syngeneic tumor immunizations female AKR or (AKR X BIO.BR)F1 mice were injected i.p. with 10 X lo6 live AKR 369 cells or with 20 x lo6 irradiated (4000 rds) AKR leukemic cells. Three weeks later spleen cells were Ficoll purified and restimulated in vitro for 4 days with mitomycin C- treated AKR leukemic cells or AKR 369 cells at a responder to stimulator ratio of 10: 1. Then 10% interleukin 2 (IL2)- containing supernatant from EL4 cells stimulated with phor- bol-ester was added to some cultures to expand the effector population further [19].

For some experiments female AKWJ mice were immunized with 339 or 369 tumor cells grown in RPMI 1640 medium supplemented with 1% AKR normal mouse serum; spleen cells were restimulated in vitro for 4 days in medium supplemented with 0.5% normal mouse serum and 2-mercaptoethanol in the presence of mitomycin C-treated tumor cells grown in medium with mouse serum.

cultures

2.6 Cell-mediated cytotoxicity

Cytotoxic effectors restimulated by in vitro mixed lymphocyte or lymphocyte-tumor cell culture were assayed with 'lCr- labeled tumor target cells as described [9]. The percent specific %r release was calculated for each effector-target cell ratio according to the formula (E-S)/(T-S) X 100, where E was the experimental 'lCr release, S, the spontaneous ' k r release detected after incubation of the target cells in culture medium alone, and T, the maximum "Cr release obtained after incubation of the target cells in 0.5% Triton X-100 for the duration of the assay. All effector-target ratios were carried out in tripli- cate and exhibited less than 10% variation. Spontaneous

release values varied between 5 and 22% of total release values. For measuring the cytotoxic activity of effector in presence of unlabeled target cells, lo4 %r-labeled 369 tumor cells were mixed with varying numbers of unlabeled tumor cells and 1 X lo' effector cells from IL 2-expanded mixed lymphocyte- tumor cell cultures. The total volume was again 0.2 ml and incubation was for 4 h at 37°C.

2.7 Antibody treatment of effectors

For the determination of differentiation antigens effector cells in 1 ml of RPMI 1640 medium were incubated for 30 min at 4°C with mAb against Thy-1, Lyt-1 or Lyt-2 at a final dilution of Then the cells were centrifuged, resuspended in 1 ml of 1 : 10 diluted, rabbit complement preabsorbed with AKR lymphocytes and incubated for 30 min at 37 "C, The surviving cells were washed and used as effector population in cell-mediated cytotoxicity assays.

2.8 Estimation of viral antigens by antibody binding

Two x lo' leukemia cells were incubated in triplicates with 50 p1 mAb dilution for 1 h with shaking. Excess antibody was removed by washing with modified Eagle's medium, and cells which were incubated with rat mAb were further incubated for 1 h with goat anti-rat IgG antiserum. After washing away excess antibodies, cells were incubated with 0.5 pCi = 18.5 kBq of '251-labeled protein A (spec. act. 70 pCi/pg). After washing, cell-bound radioactivity was measured and cor- rected for nonspecific binding of 1251-labeled protein A in the absence of first mAb.

3 Results

3.1 MHC class I antigen expression

Initially the AKR leukemias were typed with monoclonal anti- H-2 mAb in radiobinding assays and the results were corrobo- rated by immunoprecipitation after radiolabeling of the tumor cells [20]. They could be divided into three groups (Fig. 1):

Tumor cells

- 3 L l 389 110

Figure 1. H-2Kk antigen expression of AKR leukemia cells. The tumor cells indicated were incubated with H-2Kk-specific mAb 100-5 and after washing cell-bound antibody was estimated by binding of '251-labeled protein A.


those with high H-2 expression (e .g . 339, 369), leukemias with intermediate (e.g. 365, 424, 439) and those with low H-2 expression (e.g. 110, 344, 389). In the latter group, H-2Kk antigens were extremely low or undetectable, whereas H-2Dk were still present. To confirm the availability of H-2K antigens on the cell surface for T cell recognition we tested the lysis of the leukemic cells by alloreactive Kk-specific effectors. As shown in Fig. 2, all H-2K' leukemias were lysed by the anti- H-2K (C3H.OH anti-C3H) effectors according to their degree of H-2K expression, whereas the H-2K- tumors 344, 389 and 110 were not lysed. In contrast, all leukemic cells were lysed by anti-Dk (BIO.A anti-B1O.BR) or anti-Qalb (B1O.BR anti- C3H) effector cells (not shown).

369 L2L

Em7

I: 1

365 L39 339 314 389 1x) C3HOH-BlRA- blasts blasts Kkneg Kkpos

Figure 2. Lysis of AKR leukemia cells and control lymphoblasts by alloreactive H-2Kk-specific CTL. Effector cells were C3H.OH anti- C3H spleen cells, restimulated in vitro for 5 days. The columns repre- sent the specific 51Cr release at the effectorltarget ratios 30, 10, 3 for the targets indicated; the standard deviation is indicated for each bar. The spontaneous release was between 14% and 22%.

3 / 5 1 439

Figure 3. Immunogenicity of AKR leukemia cells in syngeneic mice. AKWJ mice were immunized in vivo and restimulated in vitro with the leukemic cells indicated in the figure and as described in Sect. 2.7. After 4 days effector cells were assayed in a 4-h Cr-release assay with 369 leukemia cells (x) and with the tumor cells used for immunization and indicated in the figure (0) as targets. Cytotoxic activities at effectorltarget ratios 30, 10 and 3 are shown.

Table 1. CTL response in AKWJ mice against syngeneic leukemia 339

Exp. Stimulators") Targets % Spec. "Cr release no. 1" 2" Em: 30 10 3 1

1 339 339 339 0 -2.0 -0.6 0 369 19.3 16.8 3.0 0.6

369 339 0.3 3.6 -0.2 1.5 369 38.3 34.7 14.5 5.0

2 339b' 339 339 9.4 6.5 2.8 1.0 369 50.8 34.1 30.7 20.2

3 369 369 339 14.3 10.6 5.6 3.0 369 60.3 47.1 28.7 6.5

a) For Exp. 1 mice were immunized with 20 X lo6 irradiated 339 cells. Twenty-one days later spleen cells were restimulated in vitro with mitomycin C-treated 339 or 369 cells at a responded stimulator ratio of 10 : 1. E/T = effectorltarget ratio. 1": in vivo immunization; 2": in vitro restimulation.

b) For Expts. 2 and 3 mice were immunized twice at 2-week intervals with mitomycin C-treated leukemia cells grown in RPMI 1640 medium supplemented with 1% AKR normal mouse serum; restimulation in v i m was carried out for 4 days in RPMI 1640 medium supplemented with 0.5% AKR normal mouse serum.

3.2 Immunogenicity of leukemia cells

We have shown in previous experiments that leukemia 369 was highly immunogenic and induced CTL which killed 369 cells in vitro with H-2Kk acting as restricting element [9]. These CTL also lysed another H-2K+ leukemia, 424, but not H-2K- AKR leukemias. To find out whether other H-2' leukemias could also induce CTL, we immunized AKR mice with irradiated leukemia cells and restimulated their spleen cells in vitro with mitomycin C-treated cells from the same leukemias. The resulting effector cells were then tested in a cytotoxicity assay against the immunizing tumor and against 369 as target cells (Fig. 3). All H-2K' leukemias (369, 424, 365, 439, 339) induced CTL which lysed 369 tumor cells whereas H-2K- tumors (344, 389, 110) did not induce CTL. However, CTL induced by leukemias 339, 365 and 439, which were cytotoxic for 369 cells, did not lyse the immunizing tumor cells. Further- more, spleen cells from mice, which were immunized with H- 2Kt leukemia cells, were restimulated in vitro with 369 cells. Again we obtained effectors which were cytotoxic for 369 but not for the tumors used for in vivo priming. The results of an experiment in which 339 tumor cells were used for primary immunization is shown in Table 1 (Expts. 1 and 2) . Similar results were obtained for immunizations with leukemias 365 and 439. The cytotoxic effectors were CTL as shown by their sensitivity to treatment with anti-Thy-1.1 or anti-Lyt-2.1 and complement; in addition, the effectors did not lyse YAC-1 cells which are targets for natural killer cells (not shown).

Since there was the possibility that FCS components at the surface of immunizing cells might lead to activation and secon- dary development of polyclonal cytotoxic T cells with a significant anti-syngeneic component (i.e. FCS-induced kill [21, 221, we performed immunizations and restimulations in vitro with 339 and 369 tumor cells grown for at least 5 passages in RPMI medium supplemented with 1% AKR normal mouse serum; restimulation in vitro was carried out in RPMI 1640 medium supplemented with 0.5% normal mouse serum. Again, in both

756 A. Schafer and W. Schmidt Eur. J. Immunol. 1986.16: 753-759

Table 2. Effect of interferon treatment on lysis of H-2K- cells by anti-369 CTL

Responders Stimulators Targets") % Spec. "Cr release Em: 30 10 3 1

AKR 369 369 344 344 + IFN 389 389 + IFN

(AKR X B1O.BR)FI 369 369 344 344 + IFN 389 389 + IFN

86 67 42 17 9 2 0 3 0

31 33 25 9 1 2 -5 -3

n.d. 7 4 4

78 62 30 13 1 1 7 1 1 32 25 15 6

1 1 -1 -1 a) Target cells were AKR leukemia cells either un- 8 4 6 4 treated or treated with IFN-a/P as described in

Sect. 2.3.

Table 3. Restimulation in vitro of AKR-Gross-MuLV-specific CTL by 339 and 369 tumor cells

Responders Stimulators') Targets % Spec. "Cr release 1" 2" Em: 100 30 10 3

(AKRXBlO.BR)F, 369 369 339 5.3 0.8 0 0

339 339 0.6 0 0 0 369 77.6 61.1 32.7 13.4

369 31.9 15.3 6.7 2.2

BAL9.K K-GV 369 339 3.1 1.2 -0.5 0 369 62.7 57.8 50.5 27.8

339 339 2.8 4.5 0.7 0.3 a) 1": In vivo immunization with life tumor cells. 2": In 369 43.2 41.9 27.1 11.6 vitro restimulation with mitomycin C-treated tumor

cells.

immunizations with 339 and 369 cells we obtained cytotoxic effectors which lysed 369 but not 339 target cells (Table 1, Expts. 2 and 3). This indicated that FCS-induced cytotoxicity does not play a significant role and cannot explain lysis of 369 tumor cells (Fig. 3).

3.3 Lysis by H-2Kk-restricted CTL

Resistance of H-2K' leukemias to lysis by CTL induced by autologous immunization prompted us to investigate their sensitivity to lysis by AKR anti-369 CTL and by other H-2Kk- restricted cytotoxic effectors. Fig. 4 demonstrates that H-2' leukemias 339, 365 and 439 were not killed by anti-369 CTL. Only leukemia 369 and to a lesser extent 424 were lysed; the resistance of H-2K- leukemias to lysis corroborates our previous findings [9]. Stimulation of H-2 antigens in H-2- leukemias by treatment with interferon renders them susceptible to lysis by anti-369 CTL (Table 2).

Although leukemia 339, which was comparable in H-2 expression to leukemia 369, was resistant to lysis by AKRIGross- MuLV-specific CTL, it could restimulate spleen cells from mice primed with 369 tumor cells or AKR/Gross-MuLV in vitro (Table 3). These effectors lysed 369 cells in an H-2Kk- restricted manner (not shown), but again, 339 cells could not be killed.

In order to test whether the resistance to lysis of 339 tumor cells was due to structural alterations in the H-2Kk antigens we used 339 cells as stimulators and targets for other H-2Kk- restricted CTL. We immunized BALB.K (H-2k) mice with

AKR (H-2k) lymphocytes and restimulated their spleen cells in vitro with 339 or 369 tumor cells. Table 4 demonstrates that both effector populations were cytotoxic for 339 cells. BALB.K anti-AKR CTL also lysed 365 and 439 cells (not shown). Fig. 5 demonstrates that lysis of 339 cells was indeed H-2Kk restricted. These results indicate that H-2K antigens on 339, as well as on 365 and 439 cells, can function as restriction elements and suggest that resistance of these tumors to lysis by AKR/Gross-MuLV-specific CTL was not due to structural alterations in the H-2K antigens.

100 [

30 10 3 t o r g e t s 369 L21 365 L39 339 3LL 389 110

Figure 4. Resistance of AKR leukemias to lysis by syngeneic AKR anti-369 CTL. AKWJ mice were injected with 10 X lo6 live 369 and regressor mice were restimulated in vitro for 4 days as described [9]. Lysis of different AKR leukemia target cells indicated in the figure is shown at effectodtarget ratios 30, 10 and 3. Spontaneous lysis was between 12% and 17%; the standard deviations are indicated.

Eur. J. Immunol. 1986.16: 753-759

aJ

aJ 0" 80- -

Immune resistance of AKR leukemias 757

369 369 43 42 22 10

l o o t 3.4 Cold target competition experiments

369

Resistance to lysis by anti-AKWGross-MuLV specific CTL of 339 and other H-2K' leukemias could be due to differences in viral antigen expression. We tested this possibility by using AKR leukemia cells as unlabeled competitors for the lysis of SICr-labeled 369 target cells. The results in Fig. 6 demonstrate that H-2K' lysis-resistant leukemias 339,365 and 439 inhibited lysis of 369 targets. 439 cells inhibited even more efficiently than 424 cells which were killed by anti-369 CTL (Fig. 4). Apparently all H-2K' AKR leukemias expressed cell surface structures which are recognized by the antigen receptors of syngeneic 369 specific CTL.

3.5 Viral antigens recognized by CTL

The role of viral antigens for the differential behavior of AKR leukemic cells as stimulators or targets of syngeneic AKW Gross-MuLV-specific CTL was further elucidated (a) by characterizing the viral antigens recognized by 369-specific CTL, and (b) by the demonstration of differences in expression of viral antigens in various tumors. For this purpose we treated "Cr-labeled 369 target cells with heteroantisera or mAb against viral p30 and gp70 which reacted with the glycosylated gag polyproteins and envelope glycoproteins,

Table4. Lysis of AKR 339 leukemia cells by BALB.K anti-AKR effectors

Stirnula- Targets % spec. "Cr release to& WT: 30 10 3 1

1,252,5 5 10 ~l antiserum

Figure 5. H-2Kk restriction of lysis of 339 leukemia celIs by BALB.K anti-AKR CTL. Effector cells restimulated in vitro with 339 leukemia cells were assayed in a 4-h Cr-release assay on 339 target cells in presence of increasing amounts of culture supernatants containing mAb 100-5 (H, and-Kk), 15-5-5 (A, anti-Dk) or control supernatant (0). Only antibody 100-5 (Kk specific) was inhibitory.

l o o r

E L 4

365 339 4 2 L

\ \ 2 0 1 \L 439

Table 5. Inhibition of AKR-Gross-MuLV-specific CTL by anti-viral antisera and antibodies

Exp.') Antiserum % Spec. W r releaseb) Volume (PI): 10 5 2.5 1.25

1 Control serum Anti-p30 FLV Anti-gp70 FLV 100/5 mAb Control serum" + 0.5 pl 100/5 Anti-p30 FLV + 0.5 pl 10015 Anti-gp70FLV + 0.5 pl 100/5

35/56 (anti-gp70) 35/270 (anti-p30)

BBF39.27 (aoti-p30) BBF59.13 (anti-p30) BBF59.15 (anti-p30)

2 Control ascites

3 Control ascites

84 43 87

-1.5 91 36 88 57 61 64 75 60 53 59

- 30 86

97 41 88

59 56

76 73 71

-

-

-

78 - 64 67 9 2 9 0 15 53 9 2 9 0 29 44 83 77

59 61 63 63

82 - a) Effector cells were (AKR x BIO.BR)Fl anti-369 in Expts. 1 73 86 and 3 and AKR anti-369 in Exp. 2. Effectors were restimu- 65 89 lated in vitro and further expanded in IL2-containing 70 86 medium. The EIT ratio was 10 : 1.

62 -

b) Targets were 51Cr-labeled 369 cells.

A. Schafer and W. Schmidt Eur. J. Immunol. 1986.16: 753-759 758

6 0 - 0

0 x

E, 9

m

TJ

D

- 1 0 5

E

a

i 2 0 z a C

d i 369 L2L 365 L34 339 3 L I 389 110

Figure 7. Presence of MuLV p30 antigenic determinants. AKR leukemia cells indicated in the figure were incubated with p30 specific mAb 60/35 (open bars) and 35/270 (closed bars). Cell-bound antibody was estimated by binding of 'Z51-labeled protein A as described in Sect. 2.8.

respectively, of MuLV present in the tumor cell membranes. The results shown in Table 5 indicate that rabbit anti-p30 of Friend-MuLV (FLV) consistently inhibited lysis of 369 target cells, whereas anti-FLV-gp70 sera did not inhibit. The inhibition of anti-FLV-p30 sera was further enhanced by addition of 100-5 mAb (H-2Kk specific) in small amounts not causing inhibition on their own. Antisera against p30 or gp70 of Moloney- MuLV did not inhibit (not shown). In addition, we used a series of mAb against AKR MuLV p30 or gp70 determinants as inhibitors. Only some anti-p30 mAb produced a weak but consistant inhibition of 369 lysis. Table 5 shows that anti-p30 mAb BBF59.13 and BBF59.15 were weakly inhibitory. Since anti-gp70 reagents consistently failed to inhibit, these data indicate that gag polyproteins, which carry MuLV p30 determinants, might be recognized by AKR anti-369 CTL.

We than determined the amounts of MuLV p30 on AKR leukemic cells in an antibody-binding assay. As shown in Fig. 7, there were considerable differences in the expression of p30 determinants as recognized by different mAb. This indicated that there are distinct differences in the gag polyproteins present on the surface of different leukemic cells. In addition, we have shown that the leukemic cells differed in the expression of ecotropic and xenotropic gp70 determinants as was expected from the different nature of recombination in their Mink cell focus-forming (MCF) viruses (not shown) [23].

4 Discussion

The results presented here indicate that certain H-2' AKR leukemias are resistant to immune attack by syngeneic Gross- MuLV-specific T cells in spite of their induction of potent cytolytic effector cells. In vivo immunization and in vitro restimulation of syngeneic AKR spleen cells with these H-2' leukemias induced CTL which lysed AKR 369 tumor cells, but not the immunizing tumor cells. In contrast, H-2K- leukemias did not induce CTL nor were they lysed by AKWGross- MuLV-specific effectors ([9], and Fig. 3). Thus, in addition to reduced H-2K antigen expression, there exist additional mechanism(s) in AKR leukemias which escape T cell surveil-

lance. However, it is clear from our results that the AKR leukemias in our experiments did not induce suppressor cells.

There was no evidence that these leukemias were generally resistant to CTL lysis. We have shown that they were killed by alloreactive H-2Kk-specific CTL, by H-Zunrestricted Qalb- specific CTL and by H-Zrestricted CTL. We have also shown that leukemia 339, which was resistant to lysis by syngeneic Gross-MuLV-specific CTL, could nevertheless be killed by H- 2Kk-restricted BALB.K anti-AKR CTL (Fig. 5). This indicated that H-2Kk antigens on 339 leukemia cell surfaces were functioning as restriction elements for AKR minor histocompatibility antigens and suggested that resistance to lysis was not due to major changes in the MHC antigens involved in associative recognition (H-2Kk). Furthermore, lysis of 369 tumor cells was not due to FCS-induced kill [21, 221 because CTL specifically lytic for 369 but not other H-2K' leukemias were also obtained when in vivo immunizations and in vitro restimulations were done with tumor cells grown in medium supplemented with mouse serum.

A dissociation of immunogenicity and resistance to lysis has also been reported by Plata et al. [24] and Greenberg et al. [25]. There are several possible explanations for these findings. (a) There may be minor structural alterations in the H- 2Kk antigens of some leukemias (e.g. 339) which affect their association with Gross/AKR-MuLV-specific antigens but not others, e.g. AKR minor histocompatibility antigens. CTL sensitized by macrophage-processed viral antigens presented on a correct H-2Kk would kill 369 (also with native H-2K) but not those leukemias with altered H-2K. However, there may still be sufficient interaction of these leukemias with CTL to allow inhibition of lysis of 369 tumor cells. (b) Alternatively, AKR leukemias may express MHC class I molecules distinct from but cross-reactive with H-2Kk antigens. These could sensitize CTL, but only 369 and to a lesser extent 424 tumor cells would express sufficient amounts to act as targets. (c) CTL sensitized by macrophage-processed viral antigens could be unable to recognize native viral antigens on the surface of certain leukemia cells. This explanation implies the assumption of differences in viral target antigens recognized. We have indica- tions that the glycosylated gag polyproteins which carry the p30 determinants on the cell surface [26, 271 may be the targets recognized by Gross-AKRMuLV-specific CTL. These viral antigens were also identified as target antigens for CTL specific for Gross-MuLV-induced leukemias in H-2b mice [18, 281 and for Moloney (M)-MuLV-induced leukemias [29]. However, a subset of M-MuLV-specific CTL also recognized gp70 in cells transfected with cloned DNA [30].

Radiobinding data indicated that there are indeed qualitative and quantitative differences in the expression of p30 determinants on different leukemic cells (Fig. 7). Since recombination events which lead to the formation of leukemogenic MCF viruses are not known to affect the viral gag region [23, 311, these differences may be due to differences in glycosylation and processing.

It is not known whether tumor antigens are processed by mac- rophages. If processing does indeed occur, antigens from H- 2K- leukemias should also be presented in an antigenic form on macrophage surfaces and induce AKWGross-MuLV- specific CTL. Yet we did not find CTL induction after immunization with H-2K- leukemias.


Quantitative reduction of viral gene expression was reported in Friend virus transformed cell lines to affect antigenicity of the malignant cells [32]. Quantitative reduction of H-2 antigens was shown in several viral tumors to allow immune escape [9-111. In AKR spontaneous leukemias in addition to quantitative variations either qualitative alterations of the H-2 restriction elements or of the viral target antigens may allow the evasion of tumor cell surveillance and influence leukemogenesis. To distinguish the molecular events underly- ing resistance to CTL lysis a detailed analysis of H-2K antigens and viral gag polyproteins has to be performed.

Received November 18, 1985; in revised form March 21, 1986.

5 References

1 Zinkernagel, R. M. and Doherty, P. C., Adv. Zmmunol. 1979.27:

2 Klein, J., Juretic, A., Baxevanis, C. H. and Nagy, Z. A., Nature

3 Nathenson, S. G., Uehara, H., Ewenstein, B. M., Kindt, T. J. and

4 Schmidt, W., Atfield, G. and Festenstein, H., Immunogenetics

5 De Baetselier, P., Katzov, S . , Gorelik, E., Feldman, M. and

6 Plata, F., Tilkin, A. F., Levy, J . P. and Lilly, F., J . Exp. Med.

7 Travers, P. J . , Arklie, J. L., Trowsdale, J., Patillo, R. A. and

8 Schrier, P. I., Bernards, R., Vaessen, R. T. M. J., Houweling, A.

9 Schmidt, W. and Festenstein, H., Immunogenetics 1982. 16: 257. 10 Bernards, R., Schrier, P. I., Houweling, A., Bos, J. L., Van der

Erb, A. J., Zylstra, M. and Melief, C. J. M., Nature 1983. 305: 776.

51.

1981. 291: 455.

Coligan, J. E., Annu. Rev. Biochem. 1981. 50: 1025.

1979. 8: 311.

Segal, S. , Nature 1980. 288: 179.

1981. 154: 1975.

Bodmer, W. F., Natl. Cancer Inst. Monogr. 1982. 60: 175.

and Van der Erb, A. J., Nature 1983. 305: 771.

11 Gooding, L. J., J . Immunol. 1982. 129: 1306. 12 Katzov, S. , De Baetselier, P., Tartakovsky, B., Feldman, M. and

Segal, S., J. Natl. Cancer Inst. 1983. 71: 317. 13 Hui, K., Grosveld, F. and Festenstein, H., Nature 1984.311: 750. 14 Wallich, R., Bulbuc, N., Hammerling, G. J., Katzov, S. , Segal, S.

and Feldman, M., Nature 1985. 315: 301. 15 Schmidt, W., Alonzo, A., Leven, L. and Festenstein, H., Trans-

plant. Proc. 1981. 13: 1814. 16 Lemke, H., Hammerling, G. J. and Hammerling, U., Immunol.

Rev. 1979. 47: 175. 17 Ozato, K., Mayer, N. and Sachs, D. H., J . lmmunol. 1980. 124:

533. 18 Plata, F., Kalil, J., Zilber, M. T., Fellous, M. and Levy, D., J .

Immunol. 1983. 131: 2552. 19 Wagner, H., Hardt, E., Rouse, B. T., Rollinghoff, P. M.,

Schneider, P. and Pfizenmaier, K., J . Exp. Med. 1982. 155: 1876. 20 Schmidt, W., Henseling, U., Bevec, D., Alonzo, A. D. R. and

Festenstein, H., Immunogenetics 1985. 22: 483. 21 Levy, R. B., Shearer, G. M., Kun, K. J. and Asofsky, R. M., Cell.

Immunol. 1979. 48: 276. 22 Thom, R. M . , Cell. Zmmunol. 1980. 54: 203. 23 Elder, J., Gantsch, J., Jensen, F., Lerner, R., Hartley, J. and

24 Plata, F., Goodenov, M. M. and Lilly, F., J . Exp. Med. 1980.151:

25 Greenberg, P. D., Cheever, M. A. and Tever, A, , J . Immunol.

26 Ledbetter, J. and Nowinsky, R. C., J . Virol. 1977. 23: 315. 27 Snyder, H. W., Stockert, F. and Fleissner, E., J . Virol. 1977. 23:

28 Abastado, J. P., Plata, F., Morello, D., Daniel-Vedele, F. and

29 Van der Hoorn, F. A., Lahaye, T., Miiller, V., Ogle, M. A. and

30 Flyer, D. C., Burakoff, S. J. and Faller, D. V., Nature 1983. 305:

31 Chattopadhyay, S. K., Cloya, M. W., Linemeyer, D. L., Lander,

32 Freedman, H. A., Lilly, F., Strand, M. and August, J. T., Cell

Rowe, W. P., Proc. Natl. Acad. Sci. USA 1977. 74: 4676.

726.

1981.126: 200.

302.

Kourilsky, P. I., Immunology 1985. 135: 3512.

Ender, H. A , , J . Exp. Med. 1985. 162: 128.

815.

M. R., Rands, E. and Lowy, D. R., Nature 1982. 295: 25.

1978. 13: 33.

Resistance to cellular immune response in AKR leukemias

Documents

Transcript of Resistance to cellular immune response in AKR leukemias