Key words: cytotoxicity, gene transfer, interleukin-2, mesothelioma, tumour immunology. Introduction. The induction of effective cytotoxic immune responses to.
Immunology and Cell Biology (1997) 75, 356-359
Brief Communication
The induction of immune responses to murine malignant mesothelioma by IL-2 gene transfer CLEMENT C LEONG,i-^ JULIA V MARLEY,' SUZANNE LOH,' BRUCE WS ROBINSON- atid MICHAEL J GARLEPP'^ Australian Neuromuscular Research Institute and -Department of Medicine. University of Western Australia. Elizabeth II Medical Centre. Nedlands, Western Australia, Australia
Queen
S u m m a r y Stable IL-2 transfectanl clones have been derived frotn two non-immunogenic murine malignant tnesothelioma (MM) cell lines to investigate the induction of protective antitumour immunity to MM. AC29IL-2 transfectant clones grew at a slower rate in vivo than the parental cell line or a transfectant control clone but all inoculated mice developed tumours despite the continued ability of the tumour cells to express IL-2. Tumour development after inoculation of ABl-IL 2 transfectants varied, the degree of in vivo inhibition (40-100%) being directly related to the rate of IL-2 secretion of the transfectants. When mice which had rejected the ABl-lL-2 transfectants were challenged with parental ABl cells, a proportion (16-709'f) of mice from each group remained tumour free at least 45 days after challenge (naive mice developed tumours within 26 days). The inhtbttion of growth ofthe initial inoculum of ABl-IL-2 transfectants was independent of CD4-' and CD8+ cells, consistent with the demonstration of non-specific cytotoxic activity by splenocytes from mice inoculated with the IL-2 transfectants. These data suggest that IL-2 expression by MM cells is capable of generating in vivo immunity to the tumour. This immunity may be relatively weak or may be subject to downregulation so that consistent rejection of unmodified tumour cells is not achieved. Genetic modification with combinations of genes, including lL-2 and B7-1. will be necessary for reliable generation of protective immunity to MM. Key words: cytotoxicity, gene transfer, interleukin-2, mesothelioma, tumour immunology.
Introduction The induction of effective cytotoxic immune responses to cancer cells would provide an ideal adjunct to the treatment of many cancers. There is abundant evidence in the literature to suggest that immune responses can be generated against antigens on tumour cells' but most tumours are poor stimulators of T cell activation. The genetic modification of tumours by introduction of genes encoding immunologically relevant molecules has been used in many systems in an effort to stimulate antitumour immunity.--^ A variety of genes has been used but that which has been most commonly used, and which has been subject to the most extensive clinical trial, is lL-2.'' It has been clear for several years that IL-2 is a key cytokine necessary for the activation of both CD4* and CD8+ T cells as well as NK cells.'^ Furthermore, lL-2 in high concentrations promotes the development of lymphokineactivated killer (LAK) cells from lymphoid cells and these LAK cells display high levels of cytotoxic activity toward a variety of target cells in a non-MHC restricted manner.^ Systemic IL-2 has been used in a number of clinical trials, with or without concomitant administration of LAK cells, for the Correspondence: Dr Michael J Garlepp, Australian Neuromuscular Research Institute, QEH Medical Centre, Verdun Street, Nedlands, WA 6009. Australia. Email: Received 28 October 1996; accepted 30 January 1997.
treatment of a variety of cancers."^^ The utility of systemic IL-2 as a therapeutic agent is, however, limited by the severe side effects which accompany its administration'^ and the difticulty in maintaining high concentrations at the site of the tumour. Expression of IL-2 in the local tumour environment may facilitate activation of tumour antigen-specific T cells in the absence of systemic side effects. We have undertaken a programme aimed at defining the genetic modifications necessary to induce an immune response to malignant mesothelioma (MM). This cancer is increasing in prevalence in Australia and elsewhere, as a consequence of the previous widespread use of asbestos and the long latency period for tumour development," and is resistant to all conventional forms of therapy.'^' A clinical trial of inttapleural IL-2 in MM was associated with unacceptable local and systemic side effects;'^ however, some data suggest that MM may be susceptible to immunotherapy. Trials of systemic IFN-a have produced significant reductions in tumour bulk m around 16% of patients"'" and intrapleural IFN-v has been shown to be effective against small tumour loci." A phase I clinical trial in whieh the hutnan IL-2 gene in a recombinant vaccinia virus construct is being directly adminislered to mesotheliomas is currently in progress (BWS Robinson and JA Davidson, unpubl. data, 1996). We have taken advantage of a murine model of MM'- to determine whether expression of IL-2 after transfection into
351
IL-2 gene transfer in mesothelioma
MM cells can generate an antitumour response. The murine cell lines we have used are non-immunogenic, in contrast to many of the experimental tumours used in gene transfer experiments,'* '•* and their phenotypic and biological properties closely parallel those of their human counterparts^ so that data obtained is likely to be relevant to the human cancer. We report that IL-2-expressing transfectant MM clones can generate antitumour immune responses but that there is variation between cell lines and it is likely that signals additional to IL-2 will be necessary to ensure protective immunity.
Materials and methods The cell lines ABl and AC29 have been derived from mesotheliomas produced by i.p. inoculation of croeidolite asbestos into BALB/c and CBA miee, respectively.''' The YAC-I cell line was obtained from American Type Culture Collection (Rockvillc, MD, USA). The murine lL-2 gene in the BCMGNeo vector was supplied by Basel Institute for Immunology (Basel, Switzerland).'^ Stable transtectant clones were derived from the cell lines by lipofection. followed by selection in the presence of G4I8 (Gibco. Grand Island. NY. USA) and limiting dilution cloning.'" IL-2 in the supcmatanis nt transtectan; clones, derived by culture of 1 x 10-'' cells in 1 mL of RPMI plus 5% FCS for 4S h. was measured by ELISA using mAb to murine IL-2 (Pharmingen, San Diego, CA, USA). Recombinant murine IL-2 (Pharmingen) was used to derive a standard curve and the concentrations were expressed in International Units per mL. In primary tumour challenge cxpcrimcnt.s, 6-8 week old female BALB/c or CBA mice (8-!0 mice per group) were inoculated s.c. m one rear Hank with .^ x 10^ cells. Mice which rejected the primary inoLulum were subjected to secondary challenge with I x 10^ parental lumour cells in the opposite flank. Groups of five mice were subjected to depletion of CD4^ CD8* or both CD4+ and CD8* cells by i.p. inoculation of mAh YTS-191.1 (anli-CD4), YTS-169.4 (anii-CDS)'' or a cocktail of both mAb at 0.5 mg/mL, 6 and 4 days prior to tumour inoculation and weekly thereafter."' One week following the initial depletion, the splenic mononuciear population contained less than 47f and 6% residual CD4* and CD8* cells, respectively.'^ Cytotoxicity assays were performed using splenocytes derived from mice 8 or If) days after i.p. inoculalion of 2 x 10'' live iransfectant cells. RBC-free splenocytes were cncultured with irradiated (20 000 rad) immunizing cells and cyltitoxic activity (oward a range of target celts was determined using a ^'Cr-release assay.'" Percentage lysis was calculated after subtraction of spontaneous ^'Cr release by target cells trom that released in the presence of effector cells and from target cells lysed with 2 mol/L HCI (maximum release).'"
Results Growth of IL-2 transfectants in tiaivc mice Four AC29-IL-2 transfectant clones, secreting 1.3-70 U/mL of IL-2 per 48 h per 10'' cells, were inoculated inlo CBA miec. Tumours developed in all mice. The rate of growth of the IL-2 transfectants was slower than that of lhe parental cell line but the time of appearance of the transfectant tumours was not significantly different to that of the parental cell line (Fig. la). Five ABI-lL-2 transfectant clones, secreting 10-26 U/mL of IL-2 per 4S h per 10'' ceils, respectively, were inoculated into BALB/c miee. A proportion of mice
(b) 100
a
•
o --^
• — •
•—-—•
•
a
•
D
n
25
30
D
80-
10
15
20
35
Time (days)
Ki)>urc 1 (a) Development of tumours in CBA mice after iniKulation of AC29-IL 2 transfectant clones (1 x HI"* cells), ( • ) parental AC29 or ( • ) a transfectant control (AC24-nen), tn vitro rL-2 pruduction by clones ( • ) AC29-1L-2-9, (K) AC29-1L 2-91, (O) AC29-IL-2-IS and ( • ) AC29-1L 2-43 was 1.3. 1.4. 27 and 70 U/mL per 10'' cells per 4S h, respectively, (b) The rate of tununir development in naive BALB/c mice after inoculation of ABl-lL-2 transfectant clones (1 X 10^ celts) or ( • ) parental ABl cells. In vitro IL-2 pruduction by clones ( • ) ABl-lL-2-2. < • ) ABl-IL-2-3, (%) ABllL-2-4, iC ) ABI-IL-2-5 and ( • ) ABl-lL-2-6 was 10. 14, 14, 17 and 26 U/mL per 10'' cells per 48 h, respectively, (c) Tumour development in BALB/c mice from Fig. Ic which had rejected the ADl-IL2 transfeciant clones. Mice were challenged with 1 x 10^ parental ABl cells injected subcutaneously and tumour development v^'as compared with Ihat in ( • ) naive mice (n = 8). ( • ) ABl-IL-2-2 (/I = 4); ( • ) AB1-1L-2-J (/i = 7): («) ABI-lL-2-4 (n = 9); (O) ABlIL-2-5 (n = 9); ( • ) ABl-IL-2-6 (/i = 10).
358
CC Leong et al.
(40-100%) inoculated with each clone resisted tutnour development (Fig. Ib). The number of resistant mice increased with increasing capacity of the transfectant clones to secrete IL-2 (Fig. Ib). Representative tumours which developed from each of the transfectant clones were excised, dissociated and culttired in vitro. The cultured tumour cells remained capable of secreting lL-2 (data not shown) indicating that the transfected gene remained functional in the growing tumours.
(a) 60 1 50-40-30--
Secondary challenge with parental ABl Mice which remained tumour free 78 days after inoculation of the ABI-lL-2 clones were challenged with 1 x 10^ parental ABl cells in the opposite flank. Each ofthe mice which had rejected ABl-IL-2-2 (the lowest lL-2 producer) developed tumours within 33 days. Five weeks after challenge, 33-70% of mice in each of the remaining groups were tumour free (Fig. lc). These experiments have been repeated twice using the transfectant clones producing the highest and the lowest levels of IL-2 (ABI-IL-2-6 and ABl-IL-2-2). In each of the repeat experiments rejection of the primary inoculum was complete for ABl-IL-2-6 but varied from 60 to 100% for ABl-IL 2-2. Protection against secondary challenge with ABl was more variable. The percentage of tumour-free animals at 45 days after secondary challenge ranged from 20 to 60% for ABl-IL-2-6 and 0-70% for ABl-IL-2-2. Depletion of CD4^ CD8+ or both CD4^ and CD8^ T cells from naive mice did not enhance in vivo tumour development by the transfectant clone ABl-IL-2-6 (data not shown).
20-10--
5:1
50:1
Cytotoxicity The ability of the transfectant clones to generate tumourspecific cytotoxicity was determined. Splenocytes derived from mice inoculated with one of the AC29 transfectants. AC29-1L-2-43. demonstrated non-specific cytotoxic activity toward YAC-1 cells, ABl and AC29 after 5 days coculture with irradiated IL-2 transfectants (Fig. 2a) but not after coculture with irradiated parental AC29 (data not shown). Spleen cells derived from mice inoculated with either ABl-IL-2-2 or ABML-2-6) displayed non-specific cytotoxicity (Fig. 2b,c) after in vitro coculture with the corresponding irradiated ABl-IL-2 clones but not after cocullurc with irradiated parental cells (data not shown). Splenocytes from naive mice, or mice inoculated with parental ABl, and then cocultured with irradiated IL-2 transfectant cells, displayed some cytotoxic activity toward YAC-1 targets (20-25% lysis at an effector ; target ratio of .50 : 1), but were ineffective against AB1. AC29 or ABI -IL-2 transfectants « 10% lysis at 50 : I). 50:1
Discussion The data presented here indicate that it is possible to generate immune responses against MM cells by inoculation of genetically modified tumour cells. Both of the cell lines used are non-immunogenic, in that inoculation of irradiated cells does not alter the growth rates of parental tumours on subsequent inoculation (C Leong, unpubl. data, 1994),"^ and both consti-
25:1
5:1
Effector : target ratio Figure 2 Cytotoxic activity of splenocytes derived from mice inoculated intraperitoneally with 2 x 10^ live lL-2 transfectant cells 16 days previously and cocultured with the same cells after irradiation (20 000 rads) in vitro for 5 days prior to assay, (a) AC29-IL-2-43, (b) ABl-lL-2-2, (c) ABI-rL-2-6. (•), AC29; (A), ABl; (•), AC29-IL-43; (D), AB1-1L2-2; («), AB1-1L2-6; (*), YAC-L
lL-2 gene transfer in mesothelioma
tutively express high levels of class I MHC. '^'^ The cell lines do, however, vary in their capacity to stimulate resistance to in vivo tutnour growth after inoculation of IL-2 transfectants. AC29-IL-2 transfectants produced tumours in all mice but tumour-take after inoculation of ABl-IL-2 clones was variable and related to the level of IL-2 secretion. Some resistance to the growth of parental ABl cells was induced after rejection of the ABl-IL-2 transfectants but the extent varied between clones and between experiments. The protection against parental cell growth was apparently independent of the level of IL-2 secretion by the ABl-IL-2 transfectant initially inoculated. This may relate to published data which suggest that high levels of IL-2 secretion by transfectants promole non-specific rejection mechanisms at the expense of tumour-specific immunity.'^ Variability in the responses to the two cell lines is also manifest after modification by expression of the costimulatory molecule B7-1,"'"* suggesting differences in the inherent capacities of these cell lines to activate immune effector cells. We have been unable to demonstrate in vitro tumourspeciric cytotoxic activity bul have demonstrated the generation of non-specific cytotoxic responses after inoculation of lL-2 transfectant clones derived from both cell lines. The depletion experiments suggested that the inhibition of growth of the primary inoculum of the ABl-IL-2 transfectants was primarily dependent on cells other than CD4* or CD8* T cells. Malignant mesothelioma cell lines tend to be resistant to in vitro lysis by NK cells, although human MM cells are susceptible to LAK cell lysis'^ and, in the data presented here, the parental MM cell lines were lysed by splenocytes derived from mice inoculated with IL-2 transfectant clones after coculture with IL-2 producing clones. Rejection of IL-2 transduced cells in other tumour systems has been shown to be dependent on NK cells and/or CD8* cells and less dependent on CD4'*' T cells,-'"* consistent with the data presented here. The rejection of parental ABl cells after secondary challenge, however, suggests that long-term immunity was generated in many mice, suggesting some CD4* involvement. The demonstration of //( vitro tumour-specific immunity despite rejection of parental tumours on secondary challenge has also been difficult in other systems.-" In our in vitro assays the non-specific responses may have masked any relatively weak tumour-specific cytotoxicity. Although the expression of IL-2 by MM cell lines is capable of modifying the host response to these cell lines, the protection generated is transient. Thus, it seems clear that IL-2 expression alone will not be sufficient to induce effective antitumour immune responses to MM, particularly against established tumours. It will probably be necessary to provide additional signals; for example, B7-I,"' to induce vigorous, long-lasting protective immunity. Nevertheless, the data pre.sented here provide further support for the notion that antigens on MM cells can stimulate tumour-speeifie responses if the appropriate signals are supplied.
359
Acknowledgements This work was supported by grants from the Medical Research Fund of Western Australia, The Cancer Foundation of Western Australia, The State Government Insurance Commission of Western Australia, and the Neuromu.scular Foundation of Western Australia. The authors thiink Dr Hajime Karasuyama (Ba.sel Institute for Immunology, Basel, Switzerland) for the supply of the vector BCMGNeo-mIL-2.
References 1 Ioannides CG, Whiteside TL. Clin. Immunol Immunopathol. 1993; 66: 91-106. 2 Pardoll DM. Annu. Rev. Immunol. 1993; 13: 399-115. 3 Allison JP, Hurwitz AA, Leach DR. Ctirr. Opin. Immunol. 1995; 7: 682-6. 4 Soboi RE, Scanlon KJ. Clinical trials worldwide. In: Sobol RE, Scanlon KJ (eds). The Internet Book of Gene Therapy: Cancer Therapeutics Stamford: Appleton & Lange, 1995; 283-96. 5 Minami Y. Kono T. Miyazaki T. Taniguchi T. .-\nnu Rev. Immunol. 1993; II: 245-68. 6 Grimm BA, Mazumder A, Zhang HZ, Rosenberg SA. J. Exp. Med. 1982: 155: 1823^1. 7 Rosenberg SA. Lotze MT. Muul LM ct al. N. Engl. J. Med. I4S7; 316: 88^-97 8 Sosman JA. Kohlcr PC. Hank J et al. .1. Natl Cancer Insi. 1988; 80: 1451-61 9 Robmson BWS, Davidson JA, Garlepp MJ. The immunology and immunopathology of malignant mesothelioma. In: Kradin RL. Robinson BWS (eds). Ininiunopatholoiiv of Lung Disease. Boston: Buttcrvvorth Hemcmunn. )4%; 491-514. 10 Upham JW, Garlepp MJ, Musk AW. Robinson BWS. Thorax 1995; 50: 887-93. 11 Boutin C. Viatlat JR, Van Zandwijk N et al. Canter 1991: 67: 2033-7. 12 Davis MR. Manning LS. Whitakcr D, Garlepp MJ. Robinson BWS. Int. J. Cancer 1W2: 52: 881-6. 13 Dranoff G. Jaffee E, Lazenby A et al. Pnu . Natt .Acad. Sci. USA 1993: 90: 3539-43. 14. Fakhrai H, Shawler DL, Gjerset R et al. Human Gene Therapy 1995; 6: 591-601. 15 Karasuyama H, Melchers F. Eur. J Immioiol 19S8: 18: 97-104. Ifi Leong C, Marley J, Loh S. Robinson B, Garlepp M. Cancer Gene Therapy 1996; 3: 321-30. 17 Cobhold SP, Jayasunya A, Nash A, Propcro TD, Waldman H. Nature 1984; 312: 548-51. 18 Leong CC. Marley JV, Loh S, Milech N, Robinson BWS, Garlepp MJ. hit. J. Cancer 1997; 17 (in press). 19 Manning LS, Bowman RV, Darby SB, Robinson BWS. Am. Rev. Resp. Dis. 1989; 139: 1369-74. 20 Hui KM, Sim T, Foo TT, Oei A. J. Immunol. 1989; 143: 3835-43.
