Local versus systemic interleukin-2: Tumor formation by wild ... - Nature

2 downloads 29 Views 301KB Size Report
admixtures of IL-2-secreting and B7-1-expressing K1735 cells formed fewer tumors than ... Key words: Retrovirus; B7-1; interleukin-2; gene therapy; K1735.
© 2000 Nature America, Inc. 0929-1903/00/$15.00/⫹0 www.nature.com/cgt

Local versus systemic interleukin-2: Tumor formation by wild-type and B7-1-positive murine melanoma cells Amanda L. Barnard,1 Farzin Farzaneh,2 Joop Ga ¨ken,2 and David Darling2 1

Institute of Virology and Immunoprophylaxis, Mittelhaeusern, Switzerland; and 2Department of Molecular Medicine, The Rayne Institute, King’s College School of Medicine and Dentistry, London, United Kingdom. Modification of murine K1735 melanoma cells to express the immune costimulator B7-1 had no effect on tumor formation in syngeneic mice. In contrast, ⬍40% of mice inoculated with K1735 cells modified to secrete murine interleukin-2 (IL-2) formed tumors, and no tumors formed when the K1735 cells coexpressed both murine IL-2 and B7-1. However, administration of systemic recombinant human IL-2 had no detectable effect on the formation of tumors by the B7-1-expressing K1735 cells. By contrast, admixtures of IL-2-secreting and B7-1-expressing K1735 cells formed fewer tumors than either cell type alone. Murine IL-2 was effective only when secreted locally, because the IL-2-secreting cells inoculated into the right flank did not affect the growth of the B7-1-expressing cells inoculated into the opposite flank. Cancer Gene Therapy (2000) 7, 207–214

Key words: Retrovirus; B7-1; interleukin-2; gene therapy; K1735.

A

lthough K1735 cells have been shown to be highly immunogenic using very stringent criteria,1 any primary immune response generated does not result in tumor rejection. This has been attributed to lack of a second signal (or costimulation), which T cells require for activation.2 Naive T cells with the T-cell receptordetermined capability to recognize major histocompatibility complex (MHC)-presented tumor antigens will not be activated and may be anergized in the absence of costimulation.2 Support for this observation can be found in the inability of K1735 cells engineered to express B7-1 (the costimulatory counterreceptor to CD28 found on T cells) to form tumors in syngeneic mice.3 Interleukin-2 (IL-2) is capable of preventing anergy and can even reverse it in some cases.2 IL-2 alone has been used in many different forms for cancer therapy (reviewed in Refs. 4 and 5). Unfortunately, although easily delivered systemically, the doses required to induce an effective antitumoral response can cause unpleasant side effects.6 Although B7-1 or IL-2 alone in some animal models each have little effect, the dual transduction of tumor cells to produce both B7-1 and IL-2 can be highly effective in preventing tumor formation.7,8 Using the K1735 tumor model, we have investigated whether the coexpression of B7-1 and IL-2 is more efficient in preventing tumor formation than either one alone. In addition, if IL-2 can be administered systemiReceived October 20, 1998; accepted April 3, 1999. Address correspondence and reprint requests to Dr. Farzin Farzaneh, Department of Molecular Medicine, The Rayne Institute, King’s College School of Medicine and Dentistry, 123 Coldharbour Lane, London SE5 9NU, United Kingdom.

Cancer Gene Therapy, Vol 7, No 2, 2000: pp 207–214

cally in combination with tumor cells expressing B7-1, then the IL-2 dosage required may be reduced and the side effects may be minimized. We followed a protocol for treatment of tumors using a murine model of systemic human IL-2 (hIL-2) administration (combined with IL-12-expressing murine MCA-105 carcinoma cells)9 and applied it to K1735 mouse melanoma with and without B7-1 expression. We have determined that even with a partially toxic dose of recombinant hIL-2 (rhIL-2), no effect was seen on the formation of K1735 tumors. These studies also show that IL-2 must be secreted locally to the B7-1-expressing cells but need not be expressed by the same cell.

MATERIALS AND METHODS

Tumor cell lines K1735 is a murine melanoma induced in a mammary tumor virus-negative female C3H/HeN (H-2k) mouse by initiation with ultraviolet irradiation and promotion with repeated croton oil treatment.10,11 Although they are considered to be immunogenic,1 different lines with distinct immunological properties have been generated from the original cell line. The K1735 lines used in these studies (obtained from Professor Ian Hart, St. Thomas’ Hospital, London, UK) are MHC class II-negative and express low levels of MHC class I, which can be up-regulated after interferon-␥ treatment (data not shown). Cells were cultured in Dulbecco’s modified Eagle’s medium containing 10% heat-inactivated fetal calf sera, 2 mM Lglutamine, 200 U/mL penicillin, 200 ␮g/mL streptomycin, and 1⫻ minimum essential medium nonessential amino acids (Sigma, Poole, UK). All cell lines and clones routinely tested negative for mycoplasma (Gen-Probe, San Diego, Calif).

207

208

BARNARD, FARZANEH, GA¨KEN, ET AL: LOCAL/SYSTEMIC IL-2: MURINE MELANOMA TUMOR GROWTH

Vectors and gene transduction protocols The retroviral vector producer cell line GP⫹envAM12-pBabeNeo-murine.IL-2 was a generous gift from Professor Mary Collins (Department of Immunology, University College, London, UK); the pBabe and pM3PSVHygro-mB7-1 vectors have already been described previously.8,12 Cells were cultured in Dulbecco’s modified Eagle’s medium containing 10% heat-inactivated fetal calf sera, 2 mM L-glutamine, 200 U/mL penicillin, and 200 ␮g/mL streptomycin (all from Sigma). For infection. Medium from confluent GP⫹envAM12-pBabeNeo-murine.IL-2 cells was harvested, filtered using a 0.45-␮M sterile filter, and applied with polybrene (4 ␮g/mL) to 24-hourold cultures of K1735 initiated at 2 ⫻ 106/100-mm culture dish. This infection was repeated 4 hours later; after an additional 48 hours, infected cells were selected with fresh medium containing 500 ␮g of active G418/mL. Resistant colonies were ring-cloned and expanded and screened by enzyme-linked immunosorbent assay (ELISA) for murine IL-2 secretion; one clone KIL-2-2 was chosen for further studies. For transfection. A total of 5 ⫻ 105 K1735 or KIL-2-2 cells were cultured for 24 hours and then transfected with 30 ␮g of supercoiled plasmid DNA of either pM3PSVHygro-mB7-1 or the empty pM3PSVHygro vector by standard calcium phosphate coprecipitation. At 2 days posttransfection, selection was initiated with 300 ␮g/mL hygromycin B (Calbiochem, La Jolla, Calif). Resistant colonies were ring-cloned and screened by ELISA for murine IL-2 secretion and screened by flow cytometry for determination of murine B7-1 expression.

Flow cytometry K1735 cells were removed from the substratum with 0.02% ethylenediaminetetraacetic acid in phosphate-buffered saline (PBS) (Ca2⫹/Mg2⫹-free unless otherwise stated); next, 1 ⫻ 106 cells were washed in PBS and resuspended in either 50 ␮L of 10 ␮g/mL fluorescein isothiocyanate-conjugated hamster antimouse B7-1 (CD80) antibody (clone 16-10A1, PharMingen, San Diego, Calif), or PBS alone. After 30 minutes of incubation at 4°C, cells were washed twice in PBS and resuspended in PBS for analysis. Cells were analyzed using a FACScan flow cytometer (Becton Dickinson, Mountain View, Calif) equipped with a 15-mW argon laser, and 2000 events were collected using a logarithmic data mode setting for fluorescence intensity over 4-log decades. Expression of murine B7-1 was assayed before every subcutaneous (s.c.) implantation of the cells.

Cytokine assays Cells were trypsinized and counted, and 2 ⫻ 106 cells were plated out in 12-well microtest plates. After a 24-hour incubation at 37°C, medium was aspirated, cleared by centrifugation, and stored at ⫺20°C for later analysis. Murine IL-2 was detected by ELISA as described previously13 using a monoclonal rat anti-mouse IL-2 (PharMingen clone JES6-1A12) for capture and biotinylated monoclonal rat anti-mouse IL-2 (PharMingen JES6-5H4) for detection. Production of IL2by transduced K1735 cells was determined before each inoculation.

rhIL-2 Proleukin® (Chiron, Harefield, UK) was used for intraperitoneal (i.p.) IL-2 administration for treatment of tumors. The aldesleukin it contains is an rhIL-2 produced in Escherichia

Table 1. Production of IL-2 by K1735 Murine Melanoma Cells* Cells

IL-2 secretion (ng/106/24 hours)

K1735 KB7-14 KIL-2-2 KIL-2-2-B7-3

0 0 91 ⫾ 12 96 ⫾ 18

*Wt and modified K1735 cells were grown in 2 mL of culture medium at a concentration of 106/mL. After 24 hours, the supernatants were removed; IL-2 was assayed as described in Materials and Methods. The mean and SD of 10 separate determinations are shown.

coli, differing from native hIL-2 by two amino acids and the absence of glycosylation.

Inoculation of mice Subconfluent K1735 cell cultures were detached with 0.02% ethylenediaminetetraacetic acid in PBS, washed twice in Hanks’ balanced salt solution (HBSS), and counted; the concentration was adjusted to that required. Unless otherwise stated, 6- to 12-week-old female C3H/HeN mice were inoculated s.c. in the right flank with 0.2 mL of cell suspension (see figure legends for total cell numbers). Tumors developed s.c. in the vicinity of the point of inoculation and were monitored by a ruler measuring maximum diameter. Animals were sacrificed either when tumor diameter was ⬎19 mm or when ulceration was seen. Graphs showing the percentage of survival against time after inoculation represent the formation of a tumor of 20 mm in diameter or ulceration of a tumor; no ulceration was detected in tumors of ⬍15 mm in diameter. Statistical analyses of tumor development data were performed with the SPSS statistical package (SPSS Inc., Chicago, Ill), incorporating Kaplan-Meier survival statistics,14 and the log rank test was applied to determine a ␹2 value. A ␹2 value of ⱕ0.05 (P ⱕ .05) comparing two groups was considered statistically significant.

RESULTS

Expression of IL-2 and B7-1 The IL-2 secreted by the K1735 clones KIL-2-2 and KIL-2-2-B7-3 (murine B7-1-transfected subclone of KIL-2-2) was determined by ELISA; the two clones were shown to be comparable with each other (Table 1). Wild-type (wt) K1735 cells and the B7-1-transfected K1735 clone KB7-14 secreted no detectable murine IL-2. Figure 1 demonstrates that neither wt K1735 cells or the murine IL-2-secreting KIL-2-2 clone expressed detectable levels of B7-1, whereas murine B7-1 was readily detectable; its expression was roughly equivalent in the transfected clones KB7-14 and KIL-2-2-B7-3. These clones were chosen for their comparable levels of B7-1 expression in preference to other clones expressing higher levels of B7-1. In vivo growth characteristics of IL-2- and B7-1-modified K1735 cells The chosen clones were first analyzed in vivo for their ability to form tumors in syngeneic mice. Figure 2 uses

Cancer Gene Therapy, Vol 7, No 2, 2000

BARNARD, FARZANEH, GA¨KEN, ET AL: LOCAL/SYSTEMIC IL-2: MURINE MELANOMA TUMOR GROWTH

209

Figure 1. Flow cytometric analysis of B7-1 expression on K1735 clones. Flow cytometric profiles show B7-1-negative K1735 cells (dotted lines) compared with the solid-line profiles of KB7-14, KIL-2-2, and KIL-2-2-B7-3 after staining with fluorescein isothiocyanate anti-murine CD80 (B7-1) antibody.

the combined data from two separate experiments (total animal number per group, n ⫽ 15) and shows the percentage of survival of C3H/HeN females after s.c. inoculation with wt and modified K1735 cells. No difference in growth rate (data not shown) or survival (P ⫽ .575) was detectable between nonmodified wt K1735 cells and those engineered to express B7-1 alone (KB7-14 cells). Animals injected with K1735 cells secreting IL-2 alone (KIL-2-2) all survived for ⬎60 days, a time when all animals inoculated with the wt cells had developed tumors of a sufficient size to require culling. More than 60% of KIL-2-2 inoculated animals remained tumor-free when the experiment was terminated after 156 days. This IL-2-mediated increased rate of survival was significant when compared with either K1735 (P ⬍ .0005) or B7-1-expressing (KB7-14) cells (P ⬍ .0005). The combination of IL-2 and B7-1 expressed by the same cells in KIL-2-2-B7-3 appeared the most effective (P ⫽ .007 compared with KIL-2-2), with none of the animals exhibiting any sign of disease throughout the 150-day course of the experiment. Thus B7-1 and IL-2 together resulted in a total inhibition of tumor growth, IL-2 alone gave some protection, and B7-1 expression alone conferred no protection against tumor formation. Explants of KB7-14-derived tumors cultured in the absence of selection (hygromycin) showed reduced expression of B7-1 (compared with the inoculate), whereas parallel cultures, in the presence of hygromycin, selected a population expressing B7-1 at levels that were comparable with those used to initiate the tumor.

ing (KB7-14) cells. In these studies, rhIL-2 was administered twice daily by i.p. injection at the indicated doses (2,000 – 600,000 international units (IU) per injection). Figure 3a shows the effect on K1735 tumor cell growth of an i.p. injection of rhIL-2 over 5 days, commencing the day of s.c. inoculation with K1735 cells. Figure 3b shows a concurrent experiment with the B7-1-expressing KB7-14 cells. Using doses of 2,000 and 20,000 IU per injection (20,000 and 200,000 IU total over 5 days), no difference to the rate of tumor formation in either K1735 or KB7-14 was detected when compared with HBSS carrier controls. Concluding that the dosage was too low, we performed additional experiments with the same protocol using escalated doses of 60,000 and 600,000 IU per injection on K1735 (Fig 3c) and B7-1-expressing (KB7-14) cells (Fig 3d). Once again, using 60,000 IU per injection, no delay in growth or increased survival was seen in either K1735 or KB7-14 cells. We are confident that a further increase in IL-2 dose would be of little benefit, because two animals in the K1735 group (Fig 3c) and five animals in the KB7-14 group (Fig 3d) died after only eight injections with 600,000 IU per injection. In these groups, the IL-2 was discontinued after only 8 of the 10 planned injections. The remaining animals did exhibit signs of hyperthermia but survived long enough to be culled on the basis of tumor size. Figure 3, c and d, are plotted on the basis of those animals that survived the first 8 days of treatment. Thus, the highest possible dose of systemic IL-2 had no detectable effect tumor growth. In vivo IL-2 production at a distant site is not effective

Systemic administration cannot substitute for in vivo IL-2 production Because local IL-2 production reduced the rate of tumor formation of K1735 cells, especially in conjunction with B7-1, we tested the efficacy of systemic IL-2 administration in preventing the growth of K1753 and B71-express-

Cancer Gene Therapy, Vol 7, No 2, 2000

The lack of effectiveness of systemic hIL-2 in delaying tumor growth could be due to its short half-life in vivo. The IL-2 and putative tumor antigens from the K1735 cells may not have been present at the same site for a long enough time or at high enough concentrations to allow T-cell activation. Providing a lower but more sustained

210

BARNARD, FARZANEH, GA¨KEN, ET AL: LOCAL/SYSTEMIC IL-2: MURINE MELANOMA TUMOR GROWTH

Figure 2. Survival of C3H/HeN mice after inoculation with B7-1- or IL-2-modified K1735 cells. Wt K1735 cells or K1735 cells modified to express B7-1 alone (KB7-14), IL-2 alone (KIL-2-2), or both IL-2 and B7-1 (KIL-2-2-B7-3) were inoculated (3 ⫻ 106 in 0.2 mL HBSS, n ⫽ 15/group) s.c. into the right flanks of female C3H/HeN mice. Tumor growth was monitored, and animals were culled by the criteria described in Materials and Methods.

IL-2 dose in combination with tumor antigens may be more effective at inhibiting subsequent K1735 cell growth at a distant site. Figure 4 shows an experiment designed to test this possibility . Either K1735 cells (Fig 4a) or KB7-14 cells (Fig 4b) were injected s.c. in the right flank, immediately

followed by a s.c. injection into the left flank of KIL-2-2, KIL-2-2-B7-3, or HBSS alone. Tumor growth was monitored in both flanks for 160 days, and all animals were culled on the basis of tumor formation in the right flank (K1735 or KB7-14); only a few small tumor seeds were ever

Cancer Gene Therapy, Vol 7, No 2, 2000

BARNARD, FARZANEH, GA¨KEN, ET AL: LOCAL/SYSTEMIC IL-2: MURINE MELANOMA TUMOR GROWTH

211

Figure 3. Survival of C3H/HeN mice after inoculation with K1735 or KB7-14 cells and rhIL-2 treatment. Female C3H/HeN mice were inoculated s.c. in the right flank with 3 ⫻ 106 (n ⫽ 7/group) K1735 (a,c) or KB7-14 cells (b,d). Mice were injected upon inoculation and twice daily for 4 or 5 days (8 or 10 times in total) with 0, 2,000, or 20,000 IU of rhIL-2 in 0.2 mL of HBSS (a,b) or with 0, 60,000, or 600,000 IU of rhIL-2 in 0.2 mL of HBSS (c,d). Several mice (K1735, two mice; KB7-14, five mice) died after receiving eight injections of the highest hIL-2 dose; the remaining animals were not given further injections. Tumor growth was monitored, and animals were culled by the criteria described in Materials and Methods.

present in the left (KIL-2-2 or KIL-2-2-B7-3) flanks. No effect on right flank tumor formation was seen as a result of the concurrent injection on the left. Thus, IL-2 or IL-2 plus B7 in the presence of K1735 cells, although effective in inhibiting growth in the left flank, was not effective at inhibiting growth in the right flank.

Inhibition of tumor growth does not require expression of IL-2 and B7-1 by the same cells Both systemic IL-2 and the inhibition of tumor cell growth at a distant site failed to inhibit K1735 tumor growth. Thus, it was possible that even IL-2 expressed at the site of tumor formation would fail to inhibit the growth of those cells not secreting IL-2. Admixtures of K1735 and KIL-2-2 cells were inoculated into groups of C3H/HeN mice to observe whether local production of IL-2 had an effect on tumor cells that did not themselves secrete IL-2. Figure 5 demonstrates that a s.c. inoculation with an admixture of 1.5 ⫻ 106 KIL-2-2 cells with 1.5 ⫻ 106 K1735 cells results in tumor formation in ⬎80% of animals, whereas an admixture of 1.5 ⫻ 106 KIL-2-2 cells with 1.5 ⫻ 106 KB7-14 cells reduced this (P ⫽ .008) to only 30%, coupled with a slower growth rate (data not shown). This finding indicates that IL-2 is not effective against wt K1735, even when locally produced from a KIL-2-2 cell. However, IL-2 is effective if the target cell expresses B7-1 in a situation in which B7-1 alone is not able to inhibit tumor growth. Thus,

Cancer Gene Therapy, Vol 7, No 2, 2000

IL-2-expressing cells may be inhibited in their growth through nonspecific mechanisms; however, in the presence of B7-1, a more specific response is elicited. This point is further reinforced by the admixture of K1735 with KIL-2-2-B7-3, where the formation of tumors by the K1735 cells is significantly reduced compared with those admixed with KIL-2-2 alone (P ⫽ .014). DISCUSSION In this study, we have shown that K1735 cells expressing low levels of B7-1 alone formed tumors at the same rate as nonmodified K1735 cells (Fig 2), whereas those expressing IL-2 formed tumors in only 40% of animals. Our B7-1 findings are therefore in accordance with the work of Chen et al,15 who showed that expression of B7-1 alone was not sufficient to prevent K1735 tumor formation. However, this observation is in contrast to other reports3 indicating that expression of B7-1 alone expressed on K1735 cells is very effective at preventing the same K1735 tumor formation. Whether this discrepancy results from differences in the source of cells, in the method of propagation, or in the level of B7-1 expression achieved is unclear. The K1735 cells used in the study by Townsend and Allison3 were MHC class IIpositive, but we have found no evidence of MHC class II and only a low level of expression of MHC class I in our cells. In the two previous studies mentioned above, the

212

BARNARD, FARZANEH, GA¨KEN, ET AL: LOCAL/SYSTEMIC IL-2: MURINE MELANOMA TUMOR GROWTH

Figure 5. Survival of C3H/HeN mice inoculated with admixtures of K1735 cells with B7-1-positive and IL-2-producing subtypes. Female C3H/HeN mice were inoculated s.c. in the right flank with 1:1 admixtures of cells totaling 3 ⫻ 106 K1735 melanoma cells per inoculation (n ⫽ 7/group). Tumor growth was monitored, and animals were culled by the criteria described in Materials and Methods. Surviving mice were all tumor-free for the course of the study. Figure 4. Survival of C3H/HeN mice after inoculation with K1735 or KB7-14 cells the right flank and IL-2-producing cells in the left flank. Female C3H/HeN mice were inoculated s.c. in the right flank with either 106 K1735 cells (a) or 106 KB7-14 cells (b) in 0.2 mL of HBSS (n ⫽ 7/group). Animals also received a concurrent s.c. left flank inoculation of either 0.2 mL of HBSS alone, 106 KIL-2-2 cells in 0.2 mL of HBSS, or 106 KIL-2-2-B7-3 cells in 0.2 mL of HBSS. Tumor growth was monitored, and animals were culled by the criteria described in Materials and Methods; only a few small seeds were present in the left flank at the cull.

cells were propagated either in vitro15 or in nude mice,3 which may be significant. However this may not be a sufficiently good explanation, as the same cells from the same source, whether propagated in nude mice3 or independently B7-1 transduced and propagated in vitro, exhibit essentially the same rate of tumor formation in animals.3,16 Finally, the level of B7-1 expression in these different studies may be a key factor, as this has been found to be a crucial determinant of its ability to prevent tumor formation in studies on another murine melanoma model.17 It also appears from the low level of B7-1

in explants of KB7-14 cells that reduced expression of B7-1 can be advantageous to growth in vivo. The fact that cells expressing B7-1 at levels comparable with that of the inoculate can be reselected in vitro indicates that they continue to survive in vivo. In a model of autologous gene therapy, a low level of expression of B7-1 may be the most realistic approach, because low transduction efficiency in primary tumor cells will be coupled with limited scope for the selection/ expansion of high expresser clones. Low B7-1 and finite supplies of autologous tumor cells thus require a determination of the best possible combination of immunomodulators. A number of tumor models have shown that the combination of B7-1 with IL-2 is more effective than either one alone in preventing outgrowth of inoculated tumor cells.7,8 The reasons for this are still unclear, which is perhaps not surprising given the plethora of possible effects of IL-2, including natural killer cell proliferation,18 prevention and reversal of T-cell anergy,19,20 and mimicry of delayed-type hypersensitivity.21 A major hurdle for B7-1-mediated immune gene

Cancer Gene Therapy, Vol 7, No 2, 2000

BARNARD, FARZANEH, GA¨KEN, ET AL: LOCAL/SYSTEMIC IL-2: MURINE MELANOMA TUMOR GROWTH

therapy of cancer is simply supplying sufficient B7-1expressing cells; the additional requirement for IL-2 represents an added complication. Although the engineered secretion of IL-2 alone by tumor cells22,23 can be effective in preventing tumor formation, this is not always the case. In other models, IL-2 secretion must be combined with other immunomodulators to be fully effective.8,22 Systemic administration of IL-2 may more rigorously require additional cytokine/costimulation, having little or no effect by itself24,25 and being more effective in combination with tumor cells secreting other cytokines.9 Our results indicate that i.p. injected and therefore systemic rhIL-2 has no effect on the growth of either wt or B7-1-expressing K1735 tumor cells injected s.c. (Fig 3). This is the case even when the dose of IL-2 is so high as to be toxic, causing a significant number of deaths in the animals. Although we can be confident that we were using a dosage that was as high as the animals could tolerate, we have noticed that equivalent doses have been reported to cause few side effects in other mouse strains such as C57BL/6.9 We cannot be sure that systemic murine IL-2 would behave in the same way; however, because mouse IL-2 and hIL-2 are largely interchangeable and hIL-2 is found to be effective either systemically9 or when expressed from tumor cells in murine models,26 we believe that the source of IL-2 is not responsible for its lack of effect. A more likely explanation is that even when using the highest doses, the level of IL-2 in the microenvironment of the tumor was not sufficiently high for a sufficiently long time. The dose regimen was limited to a period of 5 days, during which time an adequate blood supply to the s.c. tumor cell inoculation may not have been available. However, this did not appear to interfere with the effectiveness of a combination of systemic IL-2 and locally expressed IL-12 in another tumor model.9 It is useful to point out, however, that even in cases in which tumor cell secretion of IL-2 only delays tumor formation, explants from outgrowing tumor cells no longer secreted IL-2 in culture (Ref. 13 and our unpublished results). Tumor formation is not blocked in those cells that have stopped secreting IL-2; thus, these cells are not always susceptible to the antitumoral effect in the local environment induced by the initial majority of the IL-2-secreting cells. The lack of effectiveness of systemic rhIL-2 prompted us to look at the effect on tumor formation by K1735 and KB7-14 cells of parallel opposite flank injections of non(or minimally) tumor-forming K1735 subtypes (KIL-2-2 and KIL-2-2-B7-3) (Fig 4). In this case, the murine IL-2 would not be expected to act systemically, as the systemic concentration would be too low and unlikely to have any effect.24 We wished to test whether supplying the conditions effective for inhibition of localized tumor growth could then affect cells on the opposite flank. Even when using a cell subtype that was unable to form tumors (KIL-2-2-B7-3), no detectable effect was seen on either K1735 or B7-1-expressing cells in the other flank. This is a rigorous test, as any antitumoral activity stimulated would be expected to be systemic and there-

Cancer Gene Therapy, Vol 7, No 2, 2000

213

fore extend to another tumor site (the opposite flank). In this sense, vaccination before tumor inoculation would be a much less strict assay for the potential therapeutic activity of the tumor cell vaccine. To determine whether there was an effect in a less rigorous model, we injected the non-tumor-forming cells in the same site as those with the potential to be tumorigenic (Fig 5). In these experiments, IL-2 (from KIL-2-2) was not effective in preventing tumor formation of K1735 cells, even when injected as an admixture at the same site. The effectiveness of IL-2 was dependent upon the presence of B7-1 (compare KIL-2-2 plus K1735 with KIL-2-2-B7-3 plus K1735). However, B7-1 could be expressed either on the ultimate target cell (KB7-14) or on the cell providing the “treatment” (KIL-2-2-B7-3). In conclusion, we find that IL-2 must be provided locally and concurrently with B7-1 to inhibit the growth of tumor-forming cells. In this model, IL-2 and B7-1 need not be expressed by the same cell; this may be of significance for autologous gene therapy if two populations of cells, each with only one immunomodulatory modification, could be placed and retained at one site of vaccination. REFERENCES 1. Donawho C, Kripke ML. Immunogenicity and cross-reactivity of syngeneic murine melanomas. Cancer Commun. 1990;2:101–107. 2. Schwartz RH. A cell culture model for T lymphocyte clonal anergy. Science. 1990;248:1349 –1356. 3. Townsend SE, Allison JP. Tumor rejection after direct costimulation of CD8⫹ T cells by B7-transfected melanoma cells. Science. 1993;259:368 –370. 4. Maas RA, Dullens HFJ, Den Otter W. Interleukin-2 in cancer treatment: disappointing or (still) promising? Cancer Immunol Immunother. 1993:36:141–148. 5. Oppenheim MH, Lotze MT. Interleukin 2: solid-tumor therapy. Oncology. 1994;51:154 –169. 6. Rosenberg SA, Mule JJ, Spiess PJ, et al. Regression of established pulmonary metastases and subcutaneous tumor mediated by the systemic administration of high-dose recombinant interleukin 2. J Exp Med. 1985;161:1169 – 1188. 7. Salvadori S, Gansbacher B, Wernick I, Tirelli S, Zier K. B7-1 amplifies the response to interleukin-2-secreting tumor vaccines in vivo, but fails to induce a response by naive cells in vitro. Hum Gene Ther. 1995;6:1299 –1306. 8. Ga¨ken JA, Hollingsworth SJ, Hirst WJR, et al. Irradiated NC adenocarcinoma cells transduced with both B7.1 and interleukin-2 induce CD4⫹-mediated rejection of established tumors. Hum Gene Ther. 1997;8:477– 488. 9. Pappo I, Tahara H, Robbins PD, et al. Administration of systemic or local interleukin-2 enhances the antitumor effects of interleukin-12 gene-therapy. J Surg Res. 1995:58: 218 –226. 10. Kripke ML. Latency, histology, and antigenicity of tumors induced by ultraviolet light in three inbred mouse strains. Cancer Res. 1977;37:1395–1400. 11. Kripke ML. Speculations on the role of ultraviolet radiation in the development of malignant melanoma. J Natl Cancer Inst. 1979;63:541–548. 12. Morgenstern JP, Land H. Advanced mammalian gene

214

13.

14. 15.

16.

17. 18. 19. 20.

BARNARD, FARZANEH, GA¨KEN, ET AL: LOCAL/SYSTEMIC IL-2: MURINE MELANOMA TUMOR GROWTH

transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res. 1990;18:3587–3596. Hollingsworth SJ, Darling D, Ga¨ken J, et al. The effect of combined expression of interleukin 2 and interleukin 4 on the tumorigenicity and treatment of B16F10 melanoma. Br J Cancer. 1996;74:6 –15. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Statist Assoc. 1958;53:457– 481. Chen L, Ashe S, Brady WA, et al. Costimulation of antitumor immunity by the B7 counterreceptor for the T lymphocyte molecules CD28 and CTLA-4. Cell. 1992;71: 1093–1102. Coughlin CM, Wysocka M, Kurzawa HL, Lee WMF, Trinchieri G, Eck SL. B7-1 and interleukin-12 synergistically induce effective antitumour immunity. Cancer Res. 1995;55:4980 – 4987. Wu TC, Huang AYC, Jaffee EM, Levitsky HI, Pardoll DM. A reassessment of the role of B7-1 expression in tumor rejection. J Exp Med. 1995;182:1415–1421. Henney CS, Kuribayashi K, Kern DE, Gillis S. Interleukin-2 augments natural killer cell activity. Nature. 1981; 291:335–338. Schwartz RH. Costimulation of T lymphocytes: the role of CD28, CTLA-4, and B7/BB1 in interleukin-2 production and immunotherapy. Cell. 1992;71:1065–1068. Becker JC, Brabletz T, Czerny C, Termeer C, Brocker EB. Tumor escape mechanisms from immunosurveillance: induction of unresponsiveness in a specific MHC-restricted

21.

22.

23.

24.

25.

26.

CD4⫹ human T cell clone by the autologous MHC class II⫹ melanoma. Int Immunol. 1993;5:1501–1508. Hara I, Nguyen H, Takechi Y, Gansbacher B, Chapman PB, Houghton AN. Rejection of mouse melanoma elicited by local secretion of interleukin-2: implicating macrophages without T cells or natural killer cells in tumor rejection. Int J Cancer. 1995;61:253–260. Gansbacher B, Zier K, Daniels B, Cronin K, Bannerji R, Gilboa E. Interleukin 2 gene transfer into tumour cells abrogates tumorigenicity and induces protective immunity. J Exp Med. 1990;172:1217–1224. Fearon ER, Pardoll DM, Itaya T, et al. Interleukin-2 production by tumor cells bypasses T helper function in the generation of an antitumor response. Cell. 1990;60: 397– 403. Zatloukal K, Schneeberger A, Berger M, et al. Elicitation of a systemic and protective anti-melanoma immune response by an IL-2-based vaccine: assessment of critical cellular and molecular parameters. J Immunol. 1995:154: 3406 –3419. Eisenthal A, Skornick Y, Merimsky O, et al. Effect of allogeneic tumor cells, interleukin-2 and interleukin-6, on the growth of subcutaneous syngeneic tumors. Cancer Immunol Immunother. 1993;37:233–239. Emtage PCR, Wan Y, Bramson JL, Graham FL, Gauldie J. A double recombinant adenovirus expressing the costimulatory molecule B7-1 (Murine) and human IL-2 induces complete tumor regression in a murine breast adenocarcinoma model. J Immunol. 1998;160:2531–2538.

Cancer Gene Therapy, Vol 7, No 2, 2000