228
E. Gomard, J.P. Levy, F. Plata et al.
Elisabeth Gomard, J.P. Levy, F. Plata, Yvette Henin, Vdronique Duprez, A. Bismuth and T. Reme Laboratoire d‘lmmunologie et de Virologie des Tumeurs, INSERM U 152, Hgpital Cochin, Paris
Eur. J. Immunol. 1978.8: 228-236
Studies on the nature of the cell surface antigen reacting with cytolytic T lymphocytes in murine oncornavirusinduced tumors* The nature of the target antigen, expressed on murine sarcoma virus (MSV) and related murine tumors which reacts with T killer lymphocytes, remains ill-defined. The experiments reported here show that: (a) the previously described H-2 restriction phenomenon is found under all experimental conditions including 3-4 and 16-20-h chromium release tests. With 16-20 h tests and highly efficient T lymphocytes however, quantitative methods are necessary to demonstrate the H-2 restriction. These results support the hypothesis that H-2 molecules may be determinant in the structure and/or in the function of the cytolytic T lymphocyte (CTL) reactive antigen. (b) Under syngeneic conditions (ie.using H-2-identical immune lymphocytes, stimulators and target cells), the pattern of specificities recognized by anti-MSV or antiFriend T killer cells on 20 different lymphomas suggests that the main reactive antigen is an “FMR-like” substance. Identical conclusions were drawn from competition experiments. (c) Blockings were obtained by pre-incubation of the target cells with a goat anti-gp70, suggesting a possible role of the viral gp70 in the antigen recognized. However, this could be due to nonspecific reactions as two other anti-gp70 sera as well as with antisera directed against the viral componentsgp45, pr60, p30, p15, p12 and p10 did not block. The CTL/tumor cell interaction was not inhibited by virion-associated antigens added to the medium. Lysostrip and co-capping experiments have failed to reveal an association between H-2 and gp70. The nature of the viral protein bearing the “FMR-like” substance therefore remains to be established.
1 Introduction
Murine sarcoma virus (MSV)-induced tumors have been extensively used as models in tumor immunology and more especially for the study of cell-mediated immune reactions [ 11. In earlier experiments it has been observed that the tumor cell surface antigen(s) reacting with cytolytic T lymphocytes (CTL), was not identical with serologically defined antigens previously described [2, 31. For technical reasons, the exact nature of the CTL-reacting antigens has however been difficult to establish, and it still remains t o be elucidated. Two different problems may account for this: (a) the H-2 restriction of CTL activity, and (b) the fact that several different antigens have been described by different groups as the “CTLreacting” specificity. The existence of an H-2 restriction in the interaction between anti-MSV CTL and the related target cells [4] means that (a) it is difficult to compare the antigenic spe-
[I 19191
cificities involved when tumor cells differing for H-2 are tested in cell-mediated reactions. (b) If the “altered self” or “composed antigen” hypothesis [5-81 is correct, one must suppose that the antigen recognized by anti-MSV CTL is a complex structure involving both H-2 and viral components [4, 91 as suggested for other systems [5-81. Discrepancies have recently been reported concerning the level of H-2 restriction: while appearing strong in our experiments [9, 101 and in those of some other groups, using different oncornavirus-associated tumors [ 1 1- 131, it was described by others as an “H-2 preference rather than an H-2 restriction” [ 14- 161. Equally, the nature of the viral antigenic specificities involved has been the subject of contradictory reports [ 171. It has been described as (a) an antigen associated with the major internal viral protein p30 [ 181, as otherwise observed in rats reacting against the transplanted C58(NT)D Gross lymphoma [ 181, (b) a MEVSA1 antigen coded for by an endogenous virus and not by the transforming MSV agent [3]. Indications have been obtained from some experiments that a xenotropic endogeneous virus may be concerned [ 191. (c) The serologically defined Friend, Moloney, Rauscher (FMR) antigen or a closely related specificity [20].
* This work was supported by contracts from INSERM and DGRST. Correspondence: Elisabeth Gomard, Laboratoire d’Immunologie et de Virologie des Tumeurs, INSERM U 152, Hepita1 Cochin, 27 Rue du Faubourg Saint Jacques, F-75674 Paris CBdex 14, France Abbreviatiom: MLV: Moloney leukemia virus MSV: Murine sarcoma virus FLV: Friend leukemia virus TLV: Tennant leukemia virus RLV: Rauscher leukemia virus FMR antigen: Friend, Moloney, Rauscher antigen VCSA: Viral cell surface antigen B6: C57BL/6 mice (XL: Cytolytic T lymphocytes GCSA: Gross virus-induced antigen CRT: Chromium release test L/T ratio: Lymphoid to target cell ratio
We here report results of systematic experiments showing that (a) a strong H-2 barrier was constantly in evidence in the MSV system, as in other murine type C virus tumors, provided that quantitative experiments were carried out. (b) The major viral cell surface antigen (VCSA), recognized by the anti-MSV or anti-Friend leukemia virus (FLV) CTL, showed a specificity similar t o that of the well-known FMR antigen. The exact nature of this VCSA however, remains questionable.
Murine sarcoma virus antigen reacting with T lymphocytes
Eur. J. Immunol. 1978.8: 228-236
229
2 Materials and methods
2.3 Immunization
2.1 Viruses
2.3.1 Anti-virus immune lymphocytes
One to two-month-old C57BL/6 (B6), BALB/c, BALB.B and DBA/2 mice were obtained from our own colonies. The murine sarcoma virus (MSV) Moloney isolate was maintained by regular acellular transmission in newborn B6 or BALB/c mice. The Moloney (MLV), Friend (FLV), Tennant (TLV) and Rauscher leukemia virus (RLV) were maintained in vivo in adult BALB/c mice.
Anti-MSV were taken from the spleens of adult B6 or BALB/c inoculated 10-15 days previously with 0.2 ml of a lo-' dilution of MSV. Anti-FLV or anti-RLV lymphocytes were obtained from the spleen of adult B6 mice inoculated i.p. 10-20 days previously with 0.1 ml of a lo-' dilution of FLV or RLV Secondarily stimulated lymphocytes were spleen cells from MSV regressor animals taken 25-80 days after virus inoculation and restimulated in vitro by syngeneic lymphoma cells as described in Sect. 2.3.2.
2.2 Cells Table 1 summarizes the main data concerning the transplanted lymphomas used in the chromium release test. All of them were regularly maintained by intraperitoneal (i.p.) inoculation of l o 6 cells in adult syngeneic mice, except for the HFL/B lymphoma which was maintained in vitro. High virus producer rabbit (SIRC) or mink cells infected by murine type C xenotropic viruses were used in some competition experiments in comparison with the homologous noninfected cell line. A murine xenotropic virus producer cell line Ha-NZB (H-Zd) was also tested. Table 1. Tumor cells
Tumor
MBL2 RBLS
FBL3 GiL 11 A13 MBBl
HFL/B LSTRA T2 BCl
RC18 RC19 EdG2 EOKl EAKR EL4 P815 L1210 C136 RLldC
Origin
Inducer agents Mouse Known tumor- H-2 strain associated cell haplcsurface anti- type gens
Herbermana) MLV B6 Herberman RLV B6 Herberman FLV B6 Our lab. Graffi virus B6 Chesebrob) TLV BALB/B Our lab. MLV BALB.B Lillyc) FLV BALB.B Herberman MLV BA LBlc Chesebro FLV BALB/c Chesebro TLV BALB/c Our lab. R LV BALB/c Our lab. RLV BALB/c Boyd Grossvinis B6 DorBe) AKRvirus B6 DorB AKR virus B6 Kleinf) Dimethyl benz- B6 anthracene Our lab. Methylcholan- DBA/2 trene Our lab. Spontaneous DBA/2 Haran x rays BALB/c Gheiad Boyse x rays BALB/c
FMR FMR FMR FMR FMR FMR FM R FMR FMR FMR FMR FMR GCSA(a) GCSA (a) GCSA (a) L,E
b b b b b b b d
d d d d
b b b
d d
x1
d
R.B. Herberman, Laboratory of Tumor and Cellular Immunology, NCI, Bethesda, MD 20014, USA. B. Chesebro, National Institute of Allergy and Infectious Diseases, Rocky Mountain Laboratory, Hamilton, MT 59840, USA. F. Lilly, Albert Einstein College, Department of Genetics, Yeshiva University, Bronx, NY 10461, USA. E.A. Boyse, Sloan Kettering Institute for Cancer Research, New York 10021, USA. J.F. DorB, Institut Gustave Roussy, Villejuif, France. G. Klein, Institute for Tumor Biology, Karolinska Institutet, Stockholm 60, Sweden. N. Haran Ghera, Department of Chemical Immunology, The Weizman Institute of Science, Rehovot, Israel.
2.3.2 Anti-MBLZ, LSTRA, RLldC or EOKl immune lymphocytes These were taken from the spleen of syngeneic adult mice inoculated i.p. 3-4 weeks previously with 2 x lo'irradiated (5000 rads) tumor cells and restimulated in vitro with the corresponding tumor cells. The restimulation was performed in 25 cm2 culture flasks (Costar, Prosciences, France) by mixing 30 x 1O6 immune spleen cells with 2 x 1O6 irradiated tumor cells in 20 ml of RPMI 1640 supplemented with penicillin, streptomycin, 5 x M 2-mercaptoethanol and 10 % fetal calf serum. After 6 days incubation at 37 "C in a humidified atmosphere containing a 5 7% COz-air mixture, the cells were collected, washed once and resuspended in test medium. Previous experiments have shown that this procedure provides optimal conditions to obtain specific CTL. Normal spleen cells from mice of the same inbred strain were prepared in parallel as control.
2.4 The chromium release test (CRT) CRT was performed as previously described [2], except that U-shaped microtiter plates (Linbro Chemicals, New Haven, CT) were used. Each well received 0.1 ml of SICr-labeled cells ( 1O4 cells) and 0.1 ml of normal or immune lymphocytes at effector t o target cell ratios from 100: 1 to 3: 1. Triplicate cultures were incubated for 3- 18 h at 37 OC. Supernatants were harvested using a Titertek apparatus (Skatron AS., Lierbyen, Norway), and the absorbing elements in counting vials were placed in a counter (Packard Instrument Co., Downer's Grove, IL). The specific chromium release (SCR) was expressed as percentage of chromium release in the presence of immune lymphocytes minus percentage of chromium release in the presence of normal lymphocytes of the same line. The frequency of cytolytic effector cells (Lytic Units/l 0 6 )was calculated as: 1 n6
LU/106 =
1"-
_.
Ratio giving 50 % SCR x number of celldwell
2.5 Competition experiments These were performed by adding increasing amounts of unlabeled competitor tumor cells to the CRT. The ratio of cold targets t o labeled targets varied from 100: 1 to 12.5: 1. Results are expressed as the % chromium release in the presence of immune cells minus the % release in the presence of normal cells, the same number of competitive cells being present in both case.
E. Gomard, J.P. Levy, F. Plata et al.
230
Eur. J. ImmunoL 1978.8: 228-236
26 Blocking by antiserum or viral components Target cells ( lo4 in 0.05 ml of medium) were incubated for 20 min at 4 OC with 0.05 ml of undiluted inactivated immune serum, then used in the CRT as described above, without further washing. Controls were always run in parallel, using normal serum of the same inbred strain o r animal species. The final activity was calculated by comparing the chromium release of target cells in the presence of immune and normal lymphocytes. The main characteristics of the antisera used are summarized in Table 2. Blocking by viral components was performed using either intact or ether-disrupted viruses at a final concentration of 1 or 5 pg of viral protein per well. Viruses were incubated for 30 min with the attacker cells. Target cells were then added without further washing and the CRT performed as described in Sect. 2.4. Table 2 Antisera Origin
Scrum
B6 anti-MBL2 B10.D2 anti-B6
Our lab. Our lab.
Specificities Anti-FMR Anti-H-2b)
W u anti-CSB(NT)D Guiboutb) Anti-GCSA(a) An ti-GCSA (b) Ant i-GIX Anti-GLV gp70 Anti-GLV p30
SWu anti-AT124
Our Idb.
Anti-xenotro-
Reciprocal Retitera) marks
32
1600 20
Cytotoxic 1:500 for EdG2 target celli
2
pic virus
€ischer anti-MSV
( h d t
anti-RLV p10 p12
p15 p30 gp45 pr60 gp70 1 gp70 2 gp70 3
Grubcrc)
Gmbcr Griibcr Gruber Grubcr Grubcr Grubcr Grubcr Gildend) Augu\t')
Anti-FMR 400 Anti-MLV gp70 Anti-MLV p30
Anti-p 10 An ti-p 1 2 Anti-p 15 Anti-p 30 Aii ti-gp45 Ant i-p 26 0 Ant I-gp70 Anti-gp70 An ti-gp70
etatox1c 1 :40 for EdG2 target cells
4
carried out by the lysostrip assay [22], using anti-Hi-2and complement treatment after capping of viral polypeptides. Here again, reverse experiments were performed using antiviral polypeptides and complement after H-2 capping.
3 Results 3.1 Antigenic specificity observed in CRT using primary or secondary immune effector cells
It has previously been established that the attacker cells in the different systems tested were T cells, and this was regularly verified'by abrogation of activity with anti-Thy-1.2 serum and complement. As this point has been extensively documented [23], it will not be mentioned further in this report. Table 3 summarizes the results of experiments in which in vivo primed B6 anti-MSV or anti-FLV and BALB/c anti-MSV effector cells were tested against 20 different 51Crlabeled tumor cells including 7 FMR+ H-2b, 4 FMR- H-2b, 5 FMR' H-2d and 4 FMR- H-2d lymphomas. All the results w:re obtained in a 16 to 18-h CRT. The H-2b CTL very efficiently lysed FMR' H-2b targets, but were devoid of activity or only marginaly effective against tumor cells belonging to the three other groups. Similarly, BALB/c CTL were only cytolytic for FMR' H-2d tumor cells. In individual experiments, EL4 cells sometimes reacted with H-2b anti-MSV or anti-FLV CTL. These results have, however, been discarded since in every case, control experiments revealed that the sensitive EL4 cells were also sensitive t o anti-FMR antibodies and complement and produced C-type viruses. This further supports the idea that an "FMR-like" antigen is involved in Table 3. Cytolytic activity of anti-MSV or anti-FLV T lymphocytes in the CRT Target cell?
25 < 2 32 ND < 2 1200 640 800
a) Reciprocal titer in complement-dependent cytotoxicity or immune
fluorescence on RBLS or MBL2 target cells. b) C. Guibout, Institut Gustave Roussy, Villejuif, France. c) J. Gruber, NCI, Bethesda, MD 20014, USA d) RV. Gilden, Frederick Cancer Research Center, Frederick, MD 21701, USA.
e) J.T. August, Department of Pharmacology and Experimental Therapeutics, The Johns Hopkins University, School of Medicine, Baltimore, MD 21205, USA.
MBL2 RBLS FBL3 GiLl 1 A13 MBBl
HFLIB
18-44b) 15-50 17-37 20-52 16-51 13-33 18--47
LSTRA T2 BC 1 RC18 RC19
0-12 0-1 1 0-13 0- 7 0- 6
0-10 0-13 0- 15 0- 5 0- 7
EdG2 EYKl EAKR EL4 P815
00000000-
6 5 6 4
0000-
5 3 4 6
0000-
L1210 2.7 Co-capping and lysostrip experiments
The co-migration of H-2 and viral polypeptides on the host cell surface was tested by immunofluorescence, using the co-capping method [21]. The experiments were performed by either first capping for H-2 and then labeling for anti-viral polypeptides, or vice versa. In addition, experiments were
H-2 VCSA Attacker cells~) h a p b FMR GCSA(d) B6 dnti-MSV B6 a m - r L V BALB/c anti-MSV
C136 RL~&
17-47
16--52 13-43
15-50 18-48 15-35 12-54
3
0 -10 0- 9 0- 6 0-11 0 10 0-12 0- 8
20-49 15-38 17-50 5-20 6-25
5
0000-
4 5 3 2
2 4 3 2
0000-
4 5 3 4
4
3
a) Immune CTL were taken from in vivo primed mice 10-15 days
after virus infection. b) Minimum and maximum percent specific chromium release observed using a 1OO:l L/T cell ratio and 16-18 h incubation periods, in repeated experiments.
Eur. J. Immunol. 1978.8: 228- 236
Murine sarcoma virus antigen reacting with T lymphocytes
cell-medialed cytolysis. Identical results were found with B6 anti-RLV CTL and BALB/c o r B6 anti-syngeneic FMR’ lymphoma cells. The same pattern of specific activity was also detected (results not shown) with anti-MSV CTL restimulated in vitro by syngeneic lymphoma cells and tested in a 3 t o 4-h CRT.
3.2 Quantitative studies of the H-2 restriction These experiments were realized t o quantify and compare more precisely the H-2 restriction in 3-4 and 16- 18-h CRT, since it ha:; been reported that the H-2 restriction disappears in the 16- 18-h test [ 14, 161. After 3-4 h incubation, restriction was regularly evident and practically n o activity against allogeneic target cells found. However, the restriction was frequently unclear after 16-- 18 h incubation when high (1 00: 1) lymphoid to target cell (L/T) ratios were considered (Table 4). Nevertheless, allogeneic activity always fell sharply with decreasing LIT ratios, and in terms of lytic units/106 effector cells, the syngeneic activity was always 10 t o 50 times greater than the allogeneic activity (Table 4).
3.3 Antigenic specificities observed in competition experiments The 20 tumor cells used in the experiments listed in Table 3 were also rested as unlabeled competitors in the interaction between anti-MSV-CTL and labeled syngeneic target cells. Fig. 1 shows the results of experiments performed with in vivo primed CTL. Due t o the relatively weak activities, these tests were performed with an 18-h incubation period and 100: 1 L/T ratios. Fig. 2 shows experiments performed with CTL produced after secondary stimulation b y syngeneic tumor cells.
231
Under these conditions cytolytic activity was very high, and the tests were performed with incubation periods of only 3 h and 30:l L/T ratios. In both cases the results were similar: in the range of 12.5: 1 t o 50: 1 cold t o labeled cell ratios only syngeneic FMR’ cells competed, irrespective of experimental conditions. Not all t h e competitor cells used are shown in Figs. 1 and 2 ; however, similar results were obtained in all cases. A weak level of nonspecific competition was frequently observed when the competitor: labeled cell ratio reached 100: 1, and regularly at higher ratios, i.e. normal, unlabeled lymphocytes also competed (not shown). Among tumor cells which failed t o compete, several lymphomas figured notably bearing t h e GCSA(a) Gross virusinduced antigen [24], the X1 [25], L [26] or E [27] antigens, or embryonic specificities [ 281. Since other groups have reported that xenotropic virus-producing SIRC cells can compete for anti-MSV CTL [ 191, additional experiments were done using as competitors xenotropic virus producer rabbit (SIRC) and mink cells as well as the syngeneic producer H-2d Ha-NZB cells, which might b e a still better competitor since its bears the same H-2 as BALB/c CTL. As shown in Fig. 3a, the HaNZB cells and the xenotropic virus-infected mink cells d o not compete. The xenotropic virus producer SIRC cells apparently competed, but similar results were also found with uninfected control cells. Moreover, t h e same kind of supposed nonspecific competition has been observed when infected or uninfected SIRC cells were tested in t h e unrelated B6 antiBALB/c CRT (Fig. 3b). We can conclude that (a) our line of SIRC cells cannot be used for such experiments and (b) n o information is available t o support the idea that xenotropic virus-induced cell surface antigens play a role as target antigens for anti-MSV CTL in our experimental conditions.
Table 4. Cytolysis of H-2d or H-2b FMR’ tumor cells by secondary stimulated anti-MSV-CTL. Comparison of the activities in short and long incubation CRT Attacker cells Target cells
L/T ratios 30:l 1O:l 3:l (Spec. 5lCr release)
LUllOab)
Incubation timea) (h)
1OO:l
4 4 18 18
62a) 14 82 69
53 6 78 52
45 0 76 38
1OO:l
20:l
4: 1 0.8:l 0.2:1
57 57 0 0 92 95 22 28 6 4 0 87 35 24 14 100
24 19 0 0 71 94 11 39 4 2 0 75 39 2 18 92
1:l
Experiment 1 B6(H-2b)C) anti-MSV B6(H-2b) anti-MSV
MBL2 LSTRA MBLZ LSTRA
(H-2b) (H-2”) (H-2b) (H-2d)
37 0 65 16
17 0 49 0
5 0.1 100 5
Experinient 2 B6(H-2b) anti-MSV
RBLS MBLZ L1210 LSTRA RBL5 MBL2 L1210 LSTRA BALBlc(H-2d) RBLS anti-MSVC) EL4 L1210 ESTRA RBL5 EL4 L1210 LSTRA
(H-2”) (H-D) (H-2d) (H-2d) (H-2b) (H- 2b) (H-2“) (H-2d) (H-2”) (H-2b) (H-2“) (H-2d) (H-2b) (H-2”) (H-2d) (H-2d)
3.5 3.5 3.5 3.5 16 16 16 16 3.5 3.5 3.5 3.5 16 16 16 16
8 4 0 0 51 53 4 0 0 0
0 20
17 5 10 57
0 1 0
0 46 32 8 0 0 0 0
3 4 0 5 25
0
1.3
0 0 0
0 0
1.3
5 0 0 0
37
0 0 0 0
0 0 10
0 0 2 5
2 0. I 0.05 35.7
31
0.03 3.1 0 a) Incubation time in the CRT. b) Lytic units/l06 lymphoid cells calculated as described in Sect. 2.4. c) Restimulated by syngeneic FMR’ cells Control cells co-cultivated with normal syngeneic lymphocytes were inactive in the CRT.
232
E. Gomard, J.P. Levy, F. Plata et aL
Eur. J. Immunol. 1978.8: 228-236
BC 1
EU62 12 EAKR
-! I , L al
,
1.25 x 105
._E,
MBL2 HFllB M8El A13 F813
,
2.5 x105
5 x105
: E
1
1.25~10~
2.5~10~
5 x 105
0 r L
30
+ AC18
lo -
LSIRA
72 0
11191911
ec1
Number ot unlabeled competitor cells
1.25~10~
[I1919.21
Ffgure 1. Inhibition of CRT (18 h) mediated by in viVo primed antiMSV attacker cells by unlabeled tumor cells. (a) Attacker cells: B6 anti-MSV (2 x 106) - Target cells: MBL2 (lo4); (b) attacker cells: BALB/c anti-MSV (2 x 106) - Target cells: LSTRA (104).
2 5 ~ 1 0 ~ Number o t unkabeled competitor c e k
s x lo5
3.4 Failure to lyse FMR' cells by CTL directed against FMR- lymphomas
Figure 2. Inhibition of CRT (4 h) mediated by in vivo primed antiMSV-CTL restimulated in vifro by syngeneic lymphoma cells by unlabeled tumor cells. (a) Attacker cells: B6 anti-MSV restimulated by MBL2 (3 x lo5) - Target cells: MBL2 (104); (b) attacker cells: BALB/c anti-MSV restimulated by LSTRA (3 x 105) - Target cells: LSTRA (104).
In order t o carry out these experiments, advantage was taken of t w o different FMR- systems i n which syngeneic CTL can be obtained by secondary stimulation. The EVKl t u m o r of B6 mice was originally induced b y t h e AKR endogeneous virus [ 2 9 ] . The R L l d C leukemia is a n x ray-induced BALB/c leukemia bearing the X1 antigen, due probably t o a n endo-
geneous virus [ 2 5 ] . Table 5 shows that anti-EgK1 B6 CTL lysed E g K l and EAKR but not other B6 lymphomas. These results were confirmed b y competition experiments (see Fig. 4). Anti-RLldC CTL also lysed only R L l d C cells, and it has been shown elsewhere that FMR' cells d o not compete in this system (Duprez e t al., submitted for publication).
Table 5. Effects of anti-EVK1 and anti-RLldC CTL on various lymphoma cells Target cells (H-2b) 1OO:l
B6 anti-MBL2 CTLa) 30:l 1O:l 3:l LU/106b) (spec. 5 1 ~ releaselc) 1
1OO:l
B6 anti-EpK1 CTLa) 30:l 1O:l 3:l LU/106 1 (Spec. 5 1 ~ releasejc)
Experiment 1 MBL2 RBL5 EpKl EAKR
53 62 17 6
41 48 6 0
24 32 0 0
11 19 0 0
1.4 3.0 < 0.1 < 0.1
BALB/c anti-LSTRA CTLa)
15 9 60 57
10 7 47 43
0 0
0 0
< 0.1 < 0.1
35 36
20 19
2.5 1.8 a) Immune lymphocytes were spleen cells
(B6 x BALB/c)Fl anti-RLldC CTLa)
Experiment 2 LSTRA T2 RLI&
65 59 12
48 43 8
32 21 0
21 12 0
2.8 2.0 < 0.1
24 16 63
14 8 48
10 0 29
2 0 11
E. Gomard, J.P. Levy, F. Plata et al.
Elisabeth Gomard, J.P. Levy, F. Plata, Yvette Henin, Vdronique Duprez, A. Bismuth and T. Reme Laboratoire d‘lmmunologie et de Virologie des Tumeurs, INSERM U 152, Hgpital Cochin, Paris
Eur. J. Immunol. 1978.8: 228-236
Studies on the nature of the cell surface antigen reacting with cytolytic T lymphocytes in murine oncornavirusinduced tumors* The nature of the target antigen, expressed on murine sarcoma virus (MSV) and related murine tumors which reacts with T killer lymphocytes, remains ill-defined. The experiments reported here show that: (a) the previously described H-2 restriction phenomenon is found under all experimental conditions including 3-4 and 16-20-h chromium release tests. With 16-20 h tests and highly efficient T lymphocytes however, quantitative methods are necessary to demonstrate the H-2 restriction. These results support the hypothesis that H-2 molecules may be determinant in the structure and/or in the function of the cytolytic T lymphocyte (CTL) reactive antigen. (b) Under syngeneic conditions (ie.using H-2-identical immune lymphocytes, stimulators and target cells), the pattern of specificities recognized by anti-MSV or antiFriend T killer cells on 20 different lymphomas suggests that the main reactive antigen is an “FMR-like” substance. Identical conclusions were drawn from competition experiments. (c) Blockings were obtained by pre-incubation of the target cells with a goat anti-gp70, suggesting a possible role of the viral gp70 in the antigen recognized. However, this could be due to nonspecific reactions as two other anti-gp70 sera as well as with antisera directed against the viral componentsgp45, pr60, p30, p15, p12 and p10 did not block. The CTL/tumor cell interaction was not inhibited by virion-associated antigens added to the medium. Lysostrip and co-capping experiments have failed to reveal an association between H-2 and gp70. The nature of the viral protein bearing the “FMR-like” substance therefore remains to be established.
1 Introduction
Murine sarcoma virus (MSV)-induced tumors have been extensively used as models in tumor immunology and more especially for the study of cell-mediated immune reactions [ 11. In earlier experiments it has been observed that the tumor cell surface antigen(s) reacting with cytolytic T lymphocytes (CTL), was not identical with serologically defined antigens previously described [2, 31. For technical reasons, the exact nature of the CTL-reacting antigens has however been difficult to establish, and it still remains t o be elucidated. Two different problems may account for this: (a) the H-2 restriction of CTL activity, and (b) the fact that several different antigens have been described by different groups as the “CTLreacting” specificity. The existence of an H-2 restriction in the interaction between anti-MSV CTL and the related target cells [4] means that (a) it is difficult to compare the antigenic spe-
[I 19191
cificities involved when tumor cells differing for H-2 are tested in cell-mediated reactions. (b) If the “altered self” or “composed antigen” hypothesis [5-81 is correct, one must suppose that the antigen recognized by anti-MSV CTL is a complex structure involving both H-2 and viral components [4, 91 as suggested for other systems [5-81. Discrepancies have recently been reported concerning the level of H-2 restriction: while appearing strong in our experiments [9, 101 and in those of some other groups, using different oncornavirus-associated tumors [ 1 1- 131, it was described by others as an “H-2 preference rather than an H-2 restriction” [ 14- 161. Equally, the nature of the viral antigenic specificities involved has been the subject of contradictory reports [ 171. It has been described as (a) an antigen associated with the major internal viral protein p30 [ 181, as otherwise observed in rats reacting against the transplanted C58(NT)D Gross lymphoma [ 181, (b) a MEVSA1 antigen coded for by an endogenous virus and not by the transforming MSV agent [3]. Indications have been obtained from some experiments that a xenotropic endogeneous virus may be concerned [ 191. (c) The serologically defined Friend, Moloney, Rauscher (FMR) antigen or a closely related specificity [20].
* This work was supported by contracts from INSERM and DGRST. Correspondence: Elisabeth Gomard, Laboratoire d’Immunologie et de Virologie des Tumeurs, INSERM U 152, Hepita1 Cochin, 27 Rue du Faubourg Saint Jacques, F-75674 Paris CBdex 14, France Abbreviatiom: MLV: Moloney leukemia virus MSV: Murine sarcoma virus FLV: Friend leukemia virus TLV: Tennant leukemia virus RLV: Rauscher leukemia virus FMR antigen: Friend, Moloney, Rauscher antigen VCSA: Viral cell surface antigen B6: C57BL/6 mice (XL: Cytolytic T lymphocytes GCSA: Gross virus-induced antigen CRT: Chromium release test L/T ratio: Lymphoid to target cell ratio
We here report results of systematic experiments showing that (a) a strong H-2 barrier was constantly in evidence in the MSV system, as in other murine type C virus tumors, provided that quantitative experiments were carried out. (b) The major viral cell surface antigen (VCSA), recognized by the anti-MSV or anti-Friend leukemia virus (FLV) CTL, showed a specificity similar t o that of the well-known FMR antigen. The exact nature of this VCSA however, remains questionable.
Murine sarcoma virus antigen reacting with T lymphocytes
Eur. J. Immunol. 1978.8: 228-236
229
2 Materials and methods
2.3 Immunization
2.1 Viruses
2.3.1 Anti-virus immune lymphocytes
One to two-month-old C57BL/6 (B6), BALB/c, BALB.B and DBA/2 mice were obtained from our own colonies. The murine sarcoma virus (MSV) Moloney isolate was maintained by regular acellular transmission in newborn B6 or BALB/c mice. The Moloney (MLV), Friend (FLV), Tennant (TLV) and Rauscher leukemia virus (RLV) were maintained in vivo in adult BALB/c mice.
Anti-MSV were taken from the spleens of adult B6 or BALB/c inoculated 10-15 days previously with 0.2 ml of a lo-' dilution of MSV. Anti-FLV or anti-RLV lymphocytes were obtained from the spleen of adult B6 mice inoculated i.p. 10-20 days previously with 0.1 ml of a lo-' dilution of FLV or RLV Secondarily stimulated lymphocytes were spleen cells from MSV regressor animals taken 25-80 days after virus inoculation and restimulated in vitro by syngeneic lymphoma cells as described in Sect. 2.3.2.
2.2 Cells Table 1 summarizes the main data concerning the transplanted lymphomas used in the chromium release test. All of them were regularly maintained by intraperitoneal (i.p.) inoculation of l o 6 cells in adult syngeneic mice, except for the HFL/B lymphoma which was maintained in vitro. High virus producer rabbit (SIRC) or mink cells infected by murine type C xenotropic viruses were used in some competition experiments in comparison with the homologous noninfected cell line. A murine xenotropic virus producer cell line Ha-NZB (H-Zd) was also tested. Table 1. Tumor cells
Tumor
MBL2 RBLS
FBL3 GiL 11 A13 MBBl
HFL/B LSTRA T2 BCl
RC18 RC19 EdG2 EOKl EAKR EL4 P815 L1210 C136 RLldC
Origin
Inducer agents Mouse Known tumor- H-2 strain associated cell haplcsurface anti- type gens
Herbermana) MLV B6 Herberman RLV B6 Herberman FLV B6 Our lab. Graffi virus B6 Chesebrob) TLV BALB/B Our lab. MLV BALB.B Lillyc) FLV BALB.B Herberman MLV BA LBlc Chesebro FLV BALB/c Chesebro TLV BALB/c Our lab. R LV BALB/c Our lab. RLV BALB/c Boyd Grossvinis B6 DorBe) AKRvirus B6 DorB AKR virus B6 Kleinf) Dimethyl benz- B6 anthracene Our lab. Methylcholan- DBA/2 trene Our lab. Spontaneous DBA/2 Haran x rays BALB/c Gheiad Boyse x rays BALB/c
FMR FMR FMR FMR FMR FMR FM R FMR FMR FMR FMR FMR GCSA(a) GCSA (a) GCSA (a) L,E
b b b b b b b d
d d d d
b b b
d d
x1
d
R.B. Herberman, Laboratory of Tumor and Cellular Immunology, NCI, Bethesda, MD 20014, USA. B. Chesebro, National Institute of Allergy and Infectious Diseases, Rocky Mountain Laboratory, Hamilton, MT 59840, USA. F. Lilly, Albert Einstein College, Department of Genetics, Yeshiva University, Bronx, NY 10461, USA. E.A. Boyse, Sloan Kettering Institute for Cancer Research, New York 10021, USA. J.F. DorB, Institut Gustave Roussy, Villejuif, France. G. Klein, Institute for Tumor Biology, Karolinska Institutet, Stockholm 60, Sweden. N. Haran Ghera, Department of Chemical Immunology, The Weizman Institute of Science, Rehovot, Israel.
2.3.2 Anti-MBLZ, LSTRA, RLldC or EOKl immune lymphocytes These were taken from the spleen of syngeneic adult mice inoculated i.p. 3-4 weeks previously with 2 x lo'irradiated (5000 rads) tumor cells and restimulated in vitro with the corresponding tumor cells. The restimulation was performed in 25 cm2 culture flasks (Costar, Prosciences, France) by mixing 30 x 1O6 immune spleen cells with 2 x 1O6 irradiated tumor cells in 20 ml of RPMI 1640 supplemented with penicillin, streptomycin, 5 x M 2-mercaptoethanol and 10 % fetal calf serum. After 6 days incubation at 37 "C in a humidified atmosphere containing a 5 7% COz-air mixture, the cells were collected, washed once and resuspended in test medium. Previous experiments have shown that this procedure provides optimal conditions to obtain specific CTL. Normal spleen cells from mice of the same inbred strain were prepared in parallel as control.
2.4 The chromium release test (CRT) CRT was performed as previously described [2], except that U-shaped microtiter plates (Linbro Chemicals, New Haven, CT) were used. Each well received 0.1 ml of SICr-labeled cells ( 1O4 cells) and 0.1 ml of normal or immune lymphocytes at effector t o target cell ratios from 100: 1 to 3: 1. Triplicate cultures were incubated for 3- 18 h at 37 OC. Supernatants were harvested using a Titertek apparatus (Skatron AS., Lierbyen, Norway), and the absorbing elements in counting vials were placed in a counter (Packard Instrument Co., Downer's Grove, IL). The specific chromium release (SCR) was expressed as percentage of chromium release in the presence of immune lymphocytes minus percentage of chromium release in the presence of normal lymphocytes of the same line. The frequency of cytolytic effector cells (Lytic Units/l 0 6 )was calculated as: 1 n6
LU/106 =
1"-
_.
Ratio giving 50 % SCR x number of celldwell
2.5 Competition experiments These were performed by adding increasing amounts of unlabeled competitor tumor cells to the CRT. The ratio of cold targets t o labeled targets varied from 100: 1 to 12.5: 1. Results are expressed as the % chromium release in the presence of immune cells minus the % release in the presence of normal cells, the same number of competitive cells being present in both case.
E. Gomard, J.P. Levy, F. Plata et al.
230
Eur. J. ImmunoL 1978.8: 228-236
26 Blocking by antiserum or viral components Target cells ( lo4 in 0.05 ml of medium) were incubated for 20 min at 4 OC with 0.05 ml of undiluted inactivated immune serum, then used in the CRT as described above, without further washing. Controls were always run in parallel, using normal serum of the same inbred strain o r animal species. The final activity was calculated by comparing the chromium release of target cells in the presence of immune and normal lymphocytes. The main characteristics of the antisera used are summarized in Table 2. Blocking by viral components was performed using either intact or ether-disrupted viruses at a final concentration of 1 or 5 pg of viral protein per well. Viruses were incubated for 30 min with the attacker cells. Target cells were then added without further washing and the CRT performed as described in Sect. 2.4. Table 2 Antisera Origin
Scrum
B6 anti-MBL2 B10.D2 anti-B6
Our lab. Our lab.
Specificities Anti-FMR Anti-H-2b)
W u anti-CSB(NT)D Guiboutb) Anti-GCSA(a) An ti-GCSA (b) Ant i-GIX Anti-GLV gp70 Anti-GLV p30
SWu anti-AT124
Our Idb.
Anti-xenotro-
Reciprocal Retitera) marks
32
1600 20
Cytotoxic 1:500 for EdG2 target celli
2
pic virus
€ischer anti-MSV
( h d t
anti-RLV p10 p12
p15 p30 gp45 pr60 gp70 1 gp70 2 gp70 3
Grubcrc)
Gmbcr Griibcr Gruber Grubcr Grubcr Grubcr Grubcr Gildend) Augu\t')
Anti-FMR 400 Anti-MLV gp70 Anti-MLV p30
Anti-p 10 An ti-p 1 2 Anti-p 15 Anti-p 30 Aii ti-gp45 Ant i-p 26 0 Ant I-gp70 Anti-gp70 An ti-gp70
etatox1c 1 :40 for EdG2 target cells
4
carried out by the lysostrip assay [22], using anti-Hi-2and complement treatment after capping of viral polypeptides. Here again, reverse experiments were performed using antiviral polypeptides and complement after H-2 capping.
3 Results 3.1 Antigenic specificity observed in CRT using primary or secondary immune effector cells
It has previously been established that the attacker cells in the different systems tested were T cells, and this was regularly verified'by abrogation of activity with anti-Thy-1.2 serum and complement. As this point has been extensively documented [23], it will not be mentioned further in this report. Table 3 summarizes the results of experiments in which in vivo primed B6 anti-MSV or anti-FLV and BALB/c anti-MSV effector cells were tested against 20 different 51Crlabeled tumor cells including 7 FMR+ H-2b, 4 FMR- H-2b, 5 FMR' H-2d and 4 FMR- H-2d lymphomas. All the results w:re obtained in a 16 to 18-h CRT. The H-2b CTL very efficiently lysed FMR' H-2b targets, but were devoid of activity or only marginaly effective against tumor cells belonging to the three other groups. Similarly, BALB/c CTL were only cytolytic for FMR' H-2d tumor cells. In individual experiments, EL4 cells sometimes reacted with H-2b anti-MSV or anti-FLV CTL. These results have, however, been discarded since in every case, control experiments revealed that the sensitive EL4 cells were also sensitive t o anti-FMR antibodies and complement and produced C-type viruses. This further supports the idea that an "FMR-like" antigen is involved in Table 3. Cytolytic activity of anti-MSV or anti-FLV T lymphocytes in the CRT Target cell?
25 < 2 32 ND < 2 1200 640 800
a) Reciprocal titer in complement-dependent cytotoxicity or immune
fluorescence on RBLS or MBL2 target cells. b) C. Guibout, Institut Gustave Roussy, Villejuif, France. c) J. Gruber, NCI, Bethesda, MD 20014, USA d) RV. Gilden, Frederick Cancer Research Center, Frederick, MD 21701, USA.
e) J.T. August, Department of Pharmacology and Experimental Therapeutics, The Johns Hopkins University, School of Medicine, Baltimore, MD 21205, USA.
MBL2 RBLS FBL3 GiLl 1 A13 MBBl
HFLIB
18-44b) 15-50 17-37 20-52 16-51 13-33 18--47
LSTRA T2 BC 1 RC18 RC19
0-12 0-1 1 0-13 0- 7 0- 6
0-10 0-13 0- 15 0- 5 0- 7
EdG2 EYKl EAKR EL4 P815
00000000-
6 5 6 4
0000-
5 3 4 6
0000-
L1210 2.7 Co-capping and lysostrip experiments
The co-migration of H-2 and viral polypeptides on the host cell surface was tested by immunofluorescence, using the co-capping method [21]. The experiments were performed by either first capping for H-2 and then labeling for anti-viral polypeptides, or vice versa. In addition, experiments were
H-2 VCSA Attacker cells~) h a p b FMR GCSA(d) B6 dnti-MSV B6 a m - r L V BALB/c anti-MSV
C136 RL~&
17-47
16--52 13-43
15-50 18-48 15-35 12-54
3
0 -10 0- 9 0- 6 0-11 0 10 0-12 0- 8
20-49 15-38 17-50 5-20 6-25
5
0000-
4 5 3 2
2 4 3 2
0000-
4 5 3 4
4
3
a) Immune CTL were taken from in vivo primed mice 10-15 days
after virus infection. b) Minimum and maximum percent specific chromium release observed using a 1OO:l L/T cell ratio and 16-18 h incubation periods, in repeated experiments.
Eur. J. Immunol. 1978.8: 228- 236
Murine sarcoma virus antigen reacting with T lymphocytes
cell-medialed cytolysis. Identical results were found with B6 anti-RLV CTL and BALB/c o r B6 anti-syngeneic FMR’ lymphoma cells. The same pattern of specific activity was also detected (results not shown) with anti-MSV CTL restimulated in vitro by syngeneic lymphoma cells and tested in a 3 t o 4-h CRT.
3.2 Quantitative studies of the H-2 restriction These experiments were realized t o quantify and compare more precisely the H-2 restriction in 3-4 and 16- 18-h CRT, since it ha:; been reported that the H-2 restriction disappears in the 16- 18-h test [ 14, 161. After 3-4 h incubation, restriction was regularly evident and practically n o activity against allogeneic target cells found. However, the restriction was frequently unclear after 16-- 18 h incubation when high (1 00: 1) lymphoid to target cell (L/T) ratios were considered (Table 4). Nevertheless, allogeneic activity always fell sharply with decreasing LIT ratios, and in terms of lytic units/106 effector cells, the syngeneic activity was always 10 t o 50 times greater than the allogeneic activity (Table 4).
3.3 Antigenic specificities observed in competition experiments The 20 tumor cells used in the experiments listed in Table 3 were also rested as unlabeled competitors in the interaction between anti-MSV-CTL and labeled syngeneic target cells. Fig. 1 shows the results of experiments performed with in vivo primed CTL. Due t o the relatively weak activities, these tests were performed with an 18-h incubation period and 100: 1 L/T ratios. Fig. 2 shows experiments performed with CTL produced after secondary stimulation b y syngeneic tumor cells.
231
Under these conditions cytolytic activity was very high, and the tests were performed with incubation periods of only 3 h and 30:l L/T ratios. In both cases the results were similar: in the range of 12.5: 1 t o 50: 1 cold t o labeled cell ratios only syngeneic FMR’ cells competed, irrespective of experimental conditions. Not all t h e competitor cells used are shown in Figs. 1 and 2 ; however, similar results were obtained in all cases. A weak level of nonspecific competition was frequently observed when the competitor: labeled cell ratio reached 100: 1, and regularly at higher ratios, i.e. normal, unlabeled lymphocytes also competed (not shown). Among tumor cells which failed t o compete, several lymphomas figured notably bearing t h e GCSA(a) Gross virusinduced antigen [24], the X1 [25], L [26] or E [27] antigens, or embryonic specificities [ 281. Since other groups have reported that xenotropic virus-producing SIRC cells can compete for anti-MSV CTL [ 191, additional experiments were done using as competitors xenotropic virus producer rabbit (SIRC) and mink cells as well as the syngeneic producer H-2d Ha-NZB cells, which might b e a still better competitor since its bears the same H-2 as BALB/c CTL. As shown in Fig. 3a, the HaNZB cells and the xenotropic virus-infected mink cells d o not compete. The xenotropic virus producer SIRC cells apparently competed, but similar results were also found with uninfected control cells. Moreover, t h e same kind of supposed nonspecific competition has been observed when infected or uninfected SIRC cells were tested in t h e unrelated B6 antiBALB/c CRT (Fig. 3b). We can conclude that (a) our line of SIRC cells cannot be used for such experiments and (b) n o information is available t o support the idea that xenotropic virus-induced cell surface antigens play a role as target antigens for anti-MSV CTL in our experimental conditions.
Table 4. Cytolysis of H-2d or H-2b FMR’ tumor cells by secondary stimulated anti-MSV-CTL. Comparison of the activities in short and long incubation CRT Attacker cells Target cells
L/T ratios 30:l 1O:l 3:l (Spec. 5lCr release)
LUllOab)
Incubation timea) (h)
1OO:l
4 4 18 18
62a) 14 82 69
53 6 78 52
45 0 76 38
1OO:l
20:l
4: 1 0.8:l 0.2:1
57 57 0 0 92 95 22 28 6 4 0 87 35 24 14 100
24 19 0 0 71 94 11 39 4 2 0 75 39 2 18 92
1:l
Experiment 1 B6(H-2b)C) anti-MSV B6(H-2b) anti-MSV
MBL2 LSTRA MBLZ LSTRA
(H-2b) (H-2”) (H-2b) (H-2d)
37 0 65 16
17 0 49 0
5 0.1 100 5
Experinient 2 B6(H-2b) anti-MSV
RBLS MBLZ L1210 LSTRA RBL5 MBL2 L1210 LSTRA BALBlc(H-2d) RBLS anti-MSVC) EL4 L1210 ESTRA RBL5 EL4 L1210 LSTRA
(H-2”) (H-D) (H-2d) (H-2d) (H-2b) (H- 2b) (H-2“) (H-2d) (H-2”) (H-2b) (H-2“) (H-2d) (H-2b) (H-2”) (H-2d) (H-2d)
3.5 3.5 3.5 3.5 16 16 16 16 3.5 3.5 3.5 3.5 16 16 16 16
8 4 0 0 51 53 4 0 0 0
0 20
17 5 10 57
0 1 0
0 46 32 8 0 0 0 0
3 4 0 5 25
0
1.3
0 0 0
0 0
1.3
5 0 0 0
37
0 0 0 0
0 0 10
0 0 2 5
2 0. I 0.05 35.7
31
0.03 3.1 0 a) Incubation time in the CRT. b) Lytic units/l06 lymphoid cells calculated as described in Sect. 2.4. c) Restimulated by syngeneic FMR’ cells Control cells co-cultivated with normal syngeneic lymphocytes were inactive in the CRT.
232
E. Gomard, J.P. Levy, F. Plata et aL
Eur. J. Immunol. 1978.8: 228-236
BC 1
EU62 12 EAKR
-! I , L al
,
1.25 x 105
._E,
MBL2 HFllB M8El A13 F813
,
2.5 x105
5 x105
: E
1
1.25~10~
2.5~10~
5 x 105
0 r L
30
+ AC18
lo -
LSIRA
72 0
11191911
ec1
Number ot unlabeled competitor cells
1.25~10~
[I1919.21
Ffgure 1. Inhibition of CRT (18 h) mediated by in viVo primed antiMSV attacker cells by unlabeled tumor cells. (a) Attacker cells: B6 anti-MSV (2 x 106) - Target cells: MBL2 (lo4); (b) attacker cells: BALB/c anti-MSV (2 x 106) - Target cells: LSTRA (104).
2 5 ~ 1 0 ~ Number o t unkabeled competitor c e k
s x lo5
3.4 Failure to lyse FMR' cells by CTL directed against FMR- lymphomas
Figure 2. Inhibition of CRT (4 h) mediated by in vivo primed antiMSV-CTL restimulated in vifro by syngeneic lymphoma cells by unlabeled tumor cells. (a) Attacker cells: B6 anti-MSV restimulated by MBL2 (3 x lo5) - Target cells: MBL2 (104); (b) attacker cells: BALB/c anti-MSV restimulated by LSTRA (3 x 105) - Target cells: LSTRA (104).
In order t o carry out these experiments, advantage was taken of t w o different FMR- systems i n which syngeneic CTL can be obtained by secondary stimulation. The EVKl t u m o r of B6 mice was originally induced b y t h e AKR endogeneous virus [ 2 9 ] . The R L l d C leukemia is a n x ray-induced BALB/c leukemia bearing the X1 antigen, due probably t o a n endo-
geneous virus [ 2 5 ] . Table 5 shows that anti-EgK1 B6 CTL lysed E g K l and EAKR but not other B6 lymphomas. These results were confirmed b y competition experiments (see Fig. 4). Anti-RLldC CTL also lysed only R L l d C cells, and it has been shown elsewhere that FMR' cells d o not compete in this system (Duprez e t al., submitted for publication).
Table 5. Effects of anti-EVK1 and anti-RLldC CTL on various lymphoma cells Target cells (H-2b) 1OO:l
B6 anti-MBL2 CTLa) 30:l 1O:l 3:l LU/106b) (spec. 5 1 ~ releaselc) 1
1OO:l
B6 anti-EpK1 CTLa) 30:l 1O:l 3:l LU/106 1 (Spec. 5 1 ~ releasejc)
Experiment 1 MBL2 RBL5 EpKl EAKR
53 62 17 6
41 48 6 0
24 32 0 0
11 19 0 0
1.4 3.0 < 0.1 < 0.1
BALB/c anti-LSTRA CTLa)
15 9 60 57
10 7 47 43
0 0
0 0
< 0.1 < 0.1
35 36
20 19
2.5 1.8 a) Immune lymphocytes were spleen cells
(B6 x BALB/c)Fl anti-RLldC CTLa)
Experiment 2 LSTRA T2 RLI&
65 59 12
48 43 8
32 21 0
21 12 0
2.8 2.0 < 0.1
24 16 63
14 8 48
10 0 29
2 0 11
