Transduction of murine colon carcinoma cells with interleukin-15 gene

© 2000 Nature America, Inc. 0929-1903/00/$15.00/⫹0 www.nature.com/cgt

Transduction of murine colon carcinoma cells with interleukin-15 gene induces antitumor effects in immunocompetent and immunocompromised hosts Kentaro Tasaki,1,2 Yu Yoshida,1,3 Motohiro Miyauchi,1 Tomoko Maeda,1,2 Keizo Takenaga,4 Teruo Kouzu,5 Takehide Asano,3 Takenori Ochiai,3 Shigeru Sakiyamna,6 and Masatoshi Tagawa1 1

Division of Pathology, Chiba Cancer Center Research Institute, Chiba, Japan; Departments of 2Surgery (II) and 3Medicine (I), Chiba University School of Medicine, Chiba, Japan; 4Division of Chemotherapy, Chiba Cancer Center Research Institute, Chiba, Japan; 5Department of Endoscopic Diagnostics and Therapeutics, Chiba University, School of Medicine, Chiba, Japan; and 6Chiba Cancer Center Research Institute, Chiba, Japan. We examined the antitumor effects caused by murine colon carcinoma cells (Colon 26) transduced with interleukin-15 (IL-15) gene. Although the in vitro proliferation rate of IL-15-secreting Colon 26 (Colon 26/IL-15) cells was not different from that of wild-type (wt) cells, small subcutaneous tumors of Colon 26/IL-15 cells that developed in syngeneic immunocompetent mice regressed spontaneously in contrast to tumors of wt cells. The mice that had eliminated tumors of Colon 26/IL-15 cells rejected wt cells when subsequently challenged. The survival of the mice that had been inoculated intraperitoneally with Colon 26/IL-15 cells was significantly prolonged compared with that of the mice injected with wt cells. However, in an experimental lung metastasis model, the survival of the mice inoculated with Colon 26/IL-15 cells remained the same as that of the mice inoculated with wt cells. The inoculation of Colon 26/IL-15 cells into immunocompromised nude or severe combined immunodeficient mice produced tumors, but the survival of the immunocompromised mice was significantly longer than that of the mice inoculated with wt cells. The nude mice inoculated with Colon 26/IL-15 cells also survived longer than the severe combined immunodeficient mice with Colon 26/IL-15 cells. Depletion of natural killer cells in nude mice with anti-asialo GM1 antibody did not influence the survival of the mice injected with Colon 26/IL-15 cells. Immunohistological examination revealed that CD31⫹ cells migrated into tumors of Colon 26/IL-15 cells that developed in immunocompetent and immunocompromised mice. Taken together, our results indicate that an inoculation of IL-15-producing tumor cells can produce antitumor effects that are mediated by a variety of immunocompetent cells. Cancer Gene Therapy (2000) 7, 255–261

Key words: Colon carcinoma; interleukin-15; gene therapy; protective immunity; gene therapy.

I

nduction of systemic immunity to tumor cells is one of the therapeutic strategies for cancer. Recent studies have shown that the expression of cytokine gene(s) in tumor cells can elicit an immune response to the cytokine producers; consequently, the tumor cells can be eliminated from host animals.1 Although the precise mechanisms regarding the rejection of tumors have not been characterized fully,2 clinical trials using cytokineproducing tumor cells as a tumor vaccine have been approved.3 Among the cytokines tested, interleukin-2 (IL-2) has been extensively studied in various experi-

Received August 17, 1998; accepted May 8,1999. Address correspondence and reprint requests to Dr. Masatoshi Tagawa, Division of Pathology, Chiba Cancer Center Research Institute, 666-2 Nitona, Chuo-ku, Chiba 260-8717 Japan. E-mail address: [email protected]

Cancer Gene Therapy, Vol 7, No 2, 2000: pp 255–261

mental models, and the possibility of vaccination using IL-2 producer cells is under investigation.4 IL-15 is a novel cytokine that promotes the growth of T cells,5,6 natural killer (NK) cells,7–10 and B cells.11 Although it displays no sequence homology with IL-2,12 IL-15 requires both the ␤- and ␥-chain of the IL-2 receptor for its binding and signal transduction.13 Therefore, IL-15 and IL-2 may have an overlapped signaling system.14 Recently, a human IL-15-specific ␣ subunit gene has been cloned,15 and this subunit is responsible for the high-affinity binding of IL-15.15 Cellular distribution of the ␣ subunit is much wider than that of the IL-2 receptor ␣ subunit.15 This finding suggests that there is a broader target cell range for IL-15 and differential biological significance between IL-2 and IL-15. In this study, we tested whether the expression of IL-15 gene in tumor cells could evoke immune responses in immunocompetent and immunocompromised host animals. 255

256

TASAKI, YOSHIDA, MIYAUCHI, ET AL: ANTITUMOR EFFECT BY EXPRESSED IL-15 GENE

MATERIALS AND METHODS

Flow cytometry

Cells and animals

Transduced and wild-type (wt) cells were incubated with fluorescein isothiocyanate-conjugated monoclonal anti-H-2Kd or anti-H-2Dd antibody (Ab) (PharMingen) at room temperature for 20 minutes in the presence of 0.1% (wt/vol) sodium azide. As a control, these cells were reacted with fluorescein isothiocyanate-conjugated rabbit anti-mouse immunoglobulin Ab. The stained cells were analyzed by FACScan (Becton Dickinson, Mountain View, Calif) with CellQuest software (Becton Dickinson).

BALB/c and BALB/c nu/nu mice (6- to 8-week-old females) were purchased from the Shizuoka Laboratory Animal Center (Hamamatsu, Japan); BALB/c scid/scid mice (6- to 8-week-old females) were obtained from Nippon Clea (Tokyo, Japan). Colon 26, a carcinogen-induced undifferentiated adenocarcinoma cell line,16,17 was provided by Dr. J. Hamuro (Ajinomoto, Tokyo, Japan). Both ecotropic ␺2 and amphotropic PA317 packaging cells were obtained from the American Type Culture Collection (Manassas, Va). These cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf sera.

Isolation of murine IL-15 cDNA

Animal studies Wt or Colon 26/IL-15 cells (1 ⫻ 106) were injected subcutaneously (s.c.) into BALB/c, BALB/c nu/nu, or BALB/c scid/scid mice. These cells were also inoculated intraperitoneally (i.p.) or intravenously (i.v.). The mice that rejected Colon 26/IL-15 cells were challenged with wt cells (1 ⫻ 106) s.c. or i.v. Mice were assessed for survival, and analysis was conducted by the Kaplan-Meier test. Statistical analysis was performed using the log-rank test. Tumor volume was calculated according to the following formula: (1/2 ⫻ length ⫻ width2). In an experiment for the depletion of asialo GM1-positive cells, BALB/c nude mice were injected i.p. with anti-asialo GM1 Ab (100 ␮g/mice, Wako, Osaka, Japan) 1 day before the inoculation of tumor cells.20

The reverse transcription-based polymerase chain reaction (PCR) method was used to clone murine IL-15 cDNA. Poly(A⫹) mRNA was extracted from lipopolysaccharide (10 ␮g/mL)-stimulated spleen cells of BALB/c mice, and synthesized first-strand cDNAs were amplified with two primers, 5⬘-ATTACTCGAGCCATAGCCAGCTCATCTTCAAC-3⬘ (as a 5⬘ primer) and 5⬘-ACTACTCGAGAGCAGGTGGAGGTACCTTAATAAC-3⬘ (as a 3⬘ primer). Amplification was performed according to the manufacturer’s recommendations (Perkin-Elmer Cetus, Norwalk, Conn) and consisted of 30 cycles under the following conditions: 1 minute at 94°C for denaturation, 1 minute at 59°C for primer annealing, and 2 minutes at 72°C for primer extension. The sequence of the PCR product was confirmed to be identical with the published sequence of C57BL/6 mice-derived IL-15 cDNA18 except for two nucleotides in the 3⬘-untranslational region. The nucleotide substitutions are probably due to the polymorphism between the mouse strains.

Sections (4- or 5-␮m thickness) from frozen tumors were reacted with anti-CD11b, anti-CD90, or anti-CD31 Ab (PharMingen). Next, they were reacted with biotinylated secondary Ab. The staining using avidin-conjugated horseradish peroxidase was performed according to the manufacturer’s instructions (Vector Laboratories, Burlingame, Calif).

Construction of retrovirus and establishment of IL-15-secreting cells

RESULTS

The retrovirus vector LXSN (provided by Dr. A. D. Miller, Fred Hutchinson Cancer Research Center, Seattle, Wash)19 was used to harbor cloned IL-15 cDNA. The vector DNA with IL-15 cDNA was transfected into ecotropic ␺2 cells using lipofectin reagent (Life Technologies, Gaithersburg, Md); after the drug selection with G418 (400 ␮g/mL, Life Technologies), cell-free supernatants of G418-resistant clones were used as a retrovirus stock. The culture supernatants containing retrovirus were incubated with amphotropic PA317 cells in the presence of 8 ␮g/mL polybrene (Aldrich, Milwaukee, Wis) for infection. Among the G418-resistant PA317 cells, a clone that produced the largest amount of IL-15 mRNA was selected, and the culture supernatants were used for infecting Colon 26 cells. G418 (400 ␮g/mL)-resistant Colon 26 cells were cloned and used for experiments.

Assay for secreted IL-15 An aliquot of the supernatants of retrovirally transduced Colon 26 clones was measured for the secreted amount of IL-15 by enzyme-linked immunosorbent assay (provided by PharMingen, San Diego, Calif). Because recombinant mouse IL-15 was not available, the amounts of IL-15 secreted were determined based on a standard sample obtained from pooled supernatants of IL-15-producing cells. A clone (Colon 26/IL15) that produced the largest amount of IL-15 and a clone (Colon 26/IL-15low) that secreted one-third the amount of Colon 26/IL-15-derived IL-15 were selected.

Immunohistochemical study

In vitro analysis of Colon 26/IL-15 cells We selected both a clone (Colon 26/IL-15) that produced the largest amount of IL-15 and a low IL-15 producer (Colon 26/IL-15low). The in vitro proliferation rates of both IL-15 producers were not different from that of wt cells (data not shown). The expression level of class I antigens (Ags) of major histocompatibility complex was measured by flow cytometry. Mean fluorescence values subtracted by background staining of secondary Ab were calculated. The values of H-2Kd on wt, Colon 26/IL-15, and Colon 26/IL-15low cells were 86.8 (arbitrary units), 54.4; and 54.8, respectively, and those of H-2Dd were 135, 43.1, and 42.7, respectively. The expression levels of the class I Ags of the transduced cells were relatively lower than that of wt cells.

Antitumor effect of Colon 26/IL-15 cells in immunocompetent mice We s.c. inoculated Colon 26/IL-15 cells into syngeneic immunocompetent mice to examine the antitumor effect, if any, achieved by the expression of IL-15 gene. Although the mice injected with Colon 26/IL-15 cells developed small tumors in an early phase, the tumors regressed spontaneously and the mice survived until the

Cancer Gene Therapy, Vol 7, No 2, 2000

257


Figure 1. Tumor growth of wt and Colon 26/IL-15 cells (IL-15) inoculated in BALB/c or BALB/c nude mice The average tumor volume of seven mice per group and the SE bars are shown

end of the observation period (Fig 1 and Table 1). In contrast, wt tumors grew continuously and all of the inoculated mice died. In an experiment of i.p. inoculation, the survival of the mice injected with Colon 26/ IL-15 cells was significantly prolonged compared with that of the mice inoculated with wt cells (Table 1, ␹2 ⫽ 14.6, df ⫽ 1, P ⬍ .01). Some of the mice died due to i.p. disseminations of the tumors, but the rest of the mice survived without any signs of tumor burdens. In a model of experimental lung metastasis caused by i.v. injection of tumor cells, the survival of the mice injected with Colon 26/IL-15 cells was not different from that of the mice injected with wt cells (Table 1, P ⫽ .0790). Therefore, the antitumor effect produced by the expression of IL-15 gene was influenced by inoculation routes. To show that the antitumor effect was dependent upon the amount of IL-15 secreted, we i.p. inoculated an equally mixed population of Colon 26/IL-15 and wt cells

into mice. The amount of IL-15 per cells secreted from this population was thereby half of that of Colon 26/ IL-15 cells. We also examined the survival of the mice inoculated with Colon 26/IL-15low cells, whose IL-15 production was one-third that of Colon 26/IL-15 cells. Although the survival of the mice injected with Colon 26/IL-15low cells was not significantly different from that of the mice injected with wt cells (␹2 ⫽ 3.12, df ⫽ 1, P ⫽ .0771), the difference in survival days between the mice injected with Colon 26/IL-15low cells and those injected with the mixed population and the difference between the mice injected with the mixed population and those injected with Colon 26/IL-15 cells were statistically significant (␹2 ⫽ 6.89, df ⫽ 1, P ⬍ .01, ␹ ⫽ 12.4, df ⫽ 1, P ⬍ .01, respectively). Thus, survival of the mice was prolonged depending upon the amounts of IL-15 secreted (Table 1, P ⬍ .01). We examined whether protective immunity could be generated in the mice that had rejected Colon 26/IL-15 cells. The mice were inoculated s.c. or i.v. with wt cells on day 45 after they had received Colon 26/IL-15 cells s.c. Irrespective of injection routes, they survived until the end of the observation period without any signs of tumor development (Table 1). We did not observe any metastatic foci in the lung. Thus, systemic immunity was induced by a s.c. inoculation of Colon 26/IL-15 cells, and the immunity could inhibit both i.p. and lung metastasis.

Reduced antitumor effect of Colon 26/IL-15 cells in immunocompromised mice To understand the cell types required for the antitumor effect produced by Colon 26/IL-15 cells, we s.c. inoculated wt or Colon 26/IL-15 cells into nude or severe combined immunodeficient (SCID) mice. All of the mice that received wt cells developed tumors, and the survival days of the inoculated hosts were not different among immunocompetent and immunocompromised mice (␹2 ⫽ 3.48, df ⫽ 2, P ⫽ .175). Moreover, the survival of the nude mice that were treated with antiasialo GM1 Ab and subsequently injected with wt cells was not different from that of untreated nude mice

Table 1. Induction of an Antitumor Effect and Protective Immunity by the Inoculation of Colon 26/IL-15 Cells Injection site

Challenged cells (injection site) (1 ⫻ 106) (day 45)

n

Survival days (since last injection)

Wt Colon 26/IL-15 Colon 26/IL-15 Colon 26/IL-15

s.c. s.c. s.c. s.c.

— — wt (s.c.) wt (i.v.)

6 6 6 6

27, 29, 30, 30, 49, 54 ⬎100, ⬎100, ⬎100⬎100, ⬎100, ⬎100 ⬎100, ⬎100, ⬎100⬎100, ⬎100, ⬎100 ⬎100, ⬎100, ⬎100⬎100, ⬎100, ⬎100

Wt Colon 26/IL-15 Colon 26/IL-15low Colon 26/IL-15 ⫹ wt

i.p. i.p. i.p. i.p.

— — — —

7 7 7 7

14, 48, 17, 20,

Wt Colon 26/IL-15

i.v. i.v.

— —

7 7

11, 12, 12, 12, 14, 14, 19† 13, 15, 16, 16, 18, 18, 22†

Inoculated cells (1 ⫻ 106)

* ␹2 ⫽ 35.8; df ⫽ 3; P ⬍ .01 (log-rank test). † ␹2 ⫽ 3.09; df ⫽ 1; P ⫽ .0790 (log-rank test).


15, 54, 17, 23,

15, 75, 17, 24,

17, 80, 18, 31,

18, 19, 20* ⬎180, ⬎180, ⬎180* 21, 23, 27* 31, 32, 51*

258


Table 2. Survival of Immunocompromised Hosts Inoculated with Wt or Colon 26/IL-15 Cells and the Influence of AntiAsialo GM1 Ab Treatment Ab Inoculated cells Mice treatment* n Nude Nude

⫺ ⫺

Wt SCID Colon 26/IL-15 SCID Wt Nude Colon 26/IL-15 Nude

⫺ ⫺ ⫹ ⫹

Wt Colon/IL-15

Survival days

7 18, 18, 19, 7 65, 70, 76, 86†¶# 7 12, 22, 22, 7 36, 40, 60, 5 37, 39, 41, 5 46, 46, 70,

38, 38, 40, 42†§㛳 79, 81, 82, 22, 64, 42, 74,

22, 23, 43‡§ 65, 67, 67‡¶ 46㛳** 81#**

* Nude mice were treated with 100 ␮g of anti-asialo GM1 Ab 1 day before the inoculation of tumor cells. † ␹2 ⫽ 14.5; df ⫽ 1, P ⬍ .01 (log-rank test). ‡ ␹2 ⫽ 10.3; df ⫽ 1, P ⬍ .01 (log-rank test). § ␹2 ⫽ 0.002; df ⫽ 1, P ⫽ .968 (log-rank test). ¶ ␹2 ⫽ 10.4; df ⫽ 1, P ⬍ .01 (log-rank test). 㛳 ␹2 ⫽ 2.24; df ⫽ 1, P ⫽ .135 (log-rank test). # 2 ␹ ⫽ 2.34; df ⫽ 1, P ⫽ .126 (log-rank test). ** ␹2 ⫽ 7.07; df ⫽ 1, P ⬍ .01 (log-rank test).

inoculated with wt cells (Table 2, P ⫽ .135). Thus, any specific types of cells were not primarily responsible for the growth suppression of wt cells. The nude mice that were inoculated with Colon 26/IL-15 cells developed tumors, but the growth was markedly retarded compared with that of wt cells (Fig 1). Consequently, the survival days were longer compared with the nude mice injected with wt cells (Table 2, P ⬍ .01). Because Colon 26/IL-15 cells were rejected in syngeneic immunocompetent mice, ␣␤ T cells are required for the rejection of Colon 26/IL-15 cells. In addition, ␥␦ T cells, which develop normally in nude mice, are also involved in the antitumor effect caused by Colon 26/IL-15 cells. In an experiment using SCID mice, the survival of the mice inoculated with Colon 26/IL-15 cells was prolonged compared with that of the mice injected with wt cells (Table 2, P ⬍ .01). NK cells, granulocytes, and macrophages whose functions are intact in SCID mice can contribute to the antitumor effect. The survival days of the nude mice inoculated with Colon 26/IL-15 cells were also longer than those of the Colon 25/IL-15-injected SCID mice (Table 2, P ⬍ .01). Thus, ␥␦ T cells, NK1⫹ T cells,21 and/or B cells that are absent in SCID mice are also involved in the antitumor activity. To examine the involvement of NK cells, nude mice were treated with anti-asialo GM1 Ab and subsequently inoculated with wt or Colon 26/IL-15 cells. The survival of the mice treated with the Ab was not statistically different from that of the Ab-untreated mice, irrespective of cell types inoculated (Table 2, wt; P ⫽ .135; Colon 26/IL-15, P ⫽ .126). However, in Ab-treated group, the mice inoculated with Colon 26/IL-15 cells survived longer than those inoculated with wt cells (Table 2, P ⬍ .01). Therefore, the contribution of asialo GM1-positive NK cells to the antitumor immunity is not significant.

Immunohistochemical examination We histochemically analyzed tumors of Colon 26/IL-15 cells and those of wt cells. In immunocompetent mice, an infiltration of mononuclear cells into Colon 26/IL-15 tumors was more significant than that observed for wt tumors (data not shown). Immunohistochemical staining showed a remarkable migration of CD31⫹ cells into the tumors of Colon 26/IL-15 cells (Fig 2A) but not into wt tumors in immunocompetent mice (Fig 2B). Migration of CD31⫹ cells was also observed in Colon 26/IL-15 tumors developed in nude and SCID mice (Fig 2, C–F), although the level of cell migration was not comparable with that seen in immunocompetent mice. The infiltration of CD11b⫹ cells into the Colon 26/IL-15 tumors that developed in immunocompetent mice was heavier than that into wt tumors (Fig 2, G and H). However, the infiltration was not observed in immunocompromised mice irrespective of the inoculated cells. The accumulation of CD90⫹ cells into Colon 26/IL-15 tumors in immunocompetent mice was more significant than that into wt tumors (Fig 2, I and J), but the accumulation level was not markedly different between the Colon 26/IL-15 and wt tumors that developed in immunocompromised mice (data not shown). DISCUSSION In this study, we examined the antitumor effect of the IL-15 secreted from tumor cells. Syngeneic immunocompetent mice that were inoculated s.c. with Colon 26/IL-15 cells developed small tumors; however, these tumors regressed spontaneously. Protective immunity was generated in the mice that had rejected Colon 26/IL-15 cells, and this systemic immunity was effective for eradicating wt cells subsequently challenged even in the experimental lung metastasis model (see below). The protective immunity was also tumor-specific, because the mice that had rejected Colon 26/IL-15 cells developed tumors of syngeneic fibrosarcoma Meth A cells or lymphoma RL1 cells (data not shown). Routes of inoculation influenced the antitumor effect caused by Colon 26/IL-15 cells. Colon 26/IL-15 cells inoculated s.c. were completely rejected, whereas i.v. injected Colon 26/IL-15 cells generated multiple lung metastatic foci. The number of lung foci was the same as that seen in mice injected with wt cells (data not shown); consequently, the survival of the mice injected with Colon 26/IL-15 cells was not statistically different from that of the mice injected with wt cells (Table 1). Injection of smaller numbers of cells also produced lung foci, and the numbers of foci between the Colon 26/IL-15 and wt cell-injected groups were not different irrespective of the numbers of cells injected (data not shown). The antitumor effect induced by i.p. injection was intermediate, because some of the mice could survive without i.p. dissemination (Table 1). Differential antitumor effects depending upon the inoculation routes were not characterized in this study. We speculate that in the s.c. inoculation, professional Ag-presenting cells such as



259

Figure 2. Immunohistochemical examination of Colon 26/IL-15 tumors (A,C,E,G,I) and wt tumors (B,D,F,H,J) that developed in BALB/c (A,B,G–J), BALB/c nu/nu (C,D), or BALB/c scid/scid (E,F) mice. The sections were reacted with anti-CD30 Ab (A–F), anti-CD11b Ab (G,H), or anti-CD90 Ab (I,J). Magnification is ⫻100.


260


dendritic cells migrated more efficiently into the tumor cells than in other experimental models. This enhanced Ag processing of putative tumor Ag(s) facilitates the generation of systemic immunity. It has been shown in our previous studies and in other studies that the generation of antitumor immunity by i.v. injection was more difficult than that by s.c. inoculation.22,23 Experiments using immunocompromised mice showed that the antitumor effect caused by Colon 26/ IL-15 cells was influenced by the immunocompetency of the host. Recent investigations stress the importance of IL-15 in the generation of NK cells9 and in the enhancement of the cytotoxic activity of NK cells.7 IL-15 also plays an important role in the production of interferon-␥ (IFN-␥) from NK cells.24 In addition, IL-15 can induce lymphokine-activated killer cells and cytotoxic T cells.7,12 In the present study, however, we showed that the depletion of asialo GM1-positive NK cells from nude mice did not affect the survival of the mice inoculated with Colon 26/IL-15 cells (Table 2). Immunohistochemical staining did not reveal DX5 (pan-NK marker)positive cells25 among the cells infiltrated into the Colon 26/IL-15 tumors that developed in immunocompetent and immunocompromised mice (data not shown). Consequently, the contribution of NK cells to the IL-15mediated antitumor effect might not be significant in this study. A comparison of survival between Colon 26/IL15-injected immunocompetent and nude mice indicated a pivotal role of ␣␤ T cells in this immune response. Because the survival of the SCID mice inoculated with Colon 26/IL-15 cells was shorter than that of the nude mice, ␥␦ T cells, NK1⫹ T cells, and/or B cells that are not generated in SCID mice also play a certain role in the antitumor effect. In fact, IL-15 is a growth factor for ␥␦ T cells and stimulates the production of IFN-␥ from ␥␦ T cells.6 IL-15 also has the ability to induce B-cell proliferation.11 In contrast, the expression of the IL-15 receptor on NK1⫹ T cells and the biological activities of IL-15 on NK1⫹ T cells have not been analyzed. The survival of the SCID mice inoculated with Colon 26/ IL-15 cells was prolonged in comparison with that of SCID mice injected with wt cells. This observation suggests that granulocytes and/or macrophages are also involved in the IL-15-mediated antitumor effect. The immunohistochemical study revealed that CD31⫹ cells, which were distinguished from CD31⫹ vascular endothelium, infiltrated into the tumors of Colon 26/IL-15 cells that developed not only in immunocompetent mice but in immunocompromised mice (Fig 2, A, C, and E). In contrast, the migration of CD11b⫹ cells and CD90⫹ cells into Colon 26/IL-15 tumors, which was observed in immunocompetent mice (Fig 2, G and I), was not evident in immunocompromised mice (data not shown). Infiltrated cells cannot be directly linked with their roles in defense mechanisms; however, the present study suggests that CD31⫹ granulocytes and/or macrophages contribute to the IL-15-mediated antitumor effects in immunocompromised hosts and may also play a role in Ag presentation in immunocompetent hosts. The translational regulation of the murine IL-15 gene

is influenced by the mechanism of alternative splicing. Deletion of exon 2 and alternatively spliced exon 5 up-regulated the translational product.26 Because our IL-15 cDNA derived from PCR is devoid of exon 2 but contains nonalternatively sliced exon 5, the production of IL-15 in Colon 26 cells can be increased by modifying the cDNA. Secretion of IL-15 is also regulated at the translational or posttranslational level.27 The mechanism of IL-15 secretion is not well characterized; however, a certain amount of IL-15 was detected in the supernatants of Colon 26/IL-15 cells, and its antitumor effect was dependent upon the amount of IL-15 secreted (Table 1). Further investigation is necessary to analyze the regulatory mechanism concerning IL-15 production, because recently identified IL-18, for example, needs caspase I protease to be processed.28,29 In our previous study, we showed an antitumor effect with IL-2-producing Colon 26 cells.30 Inoculation of IL-2-producing Colon 26 cells into syngeneic immunocompetent mice did not result in s.c. tumors, and protective immunity was generated in the mice. However, we did not observe the antitumor effect with IL-2 producers in nude mice. The differential antitumor effects caused by IL-2 or IL-15 producer cells may be due to distinct kinds of cells activated by the cytokines. It has been reported that IL-15 can cooperate with IL-12 or IL-2 to proliferate NK cells and to induce IFN-␥ secretion.31,32 Therefore, it is worthwhile to test the combinatory expression of these cytokines for its efficacy in inducing antitumor effects. ACKNOWLEDGMENTS We thank Masayoshi Hirose for his assistance and Dr. K. Hirose for PCR primers. This work was supported by a grant-in-aid for scientific research and a grant-in-aid for scientific research on priority areas from the Ministry of Education, Science, Sports, and Culture of Japan and by a grant-in-aid from the Ministry of Health and Welfare of Japan.

REFERENCES 1. Tepper RI, Mule´ JJ. Experimental and clinical studies of cytokine gene-modified tumor cells. Hum Gene Ther. 1994;5:153–164. 2. Zier KS, Gansbacher B. The impact of gene therapy on T cell function in cancer. Hum Gene Ther. 1995;6:1259 –1264. 3. Roth JA, Cristiano RJ. Gene therapy for cancer: what have we done and where are we going? J Natl Cancer Inst. 1997;89:21–39. 4. Stingl G. Phase I study to the immunotherapy of metastatic malignant melanoma by a cancer vaccine consisting of autologous cancer cells transfected with the human IL-2 gene. Hum Gene Ther. 1996;7:551–563. 5. Bamford RN, Grant AJ, Burton JD, et al. The interleukin (IL)-2 receptor ␤ chain is shared by IL-2 and a cytokine, provisionally designated IL-T, that stimulates T-cell proliferation and the induction of lymphokine-activated killer cells. Proc Natl Acad Sci USA. 1994;91:4940 – 4944. 6. Nishimura H, Hiromatsu K, Kobayashi N, et al. IL-15 is a



7.

8.

9. 10. 11. 12. 13. 14.

15.

16.

17. 18. 19.

novel growth factor for murine ␥␦ T cells induced by Salmonella infection. J Immunol. 1996;156:663– 669. Carson WE, Giri JG, Lindermann MJ, et al. Interleukin (IL) 15 is a novel cytokine that activates human natural killer cells via components of the IL-2 receptor. J Exp Med. 1994;180:1395–1403. Mro ´zek E, Anderson P, Caligiuri MA. Role of interleukin-15 in the development of human CD56⫹ natural killer cells from CD34⫹ hematopoietic progenitor cells. Blood. 1996;87:2632–2640. Puzanov IJ, Bennett M, Kumar V, et al. IL-15 can substitute for the marrow microenvironment in the differentiation of natural killer cells. J Immunol. 1996;157:4282– 4285. Atedzoe BN, Ahmad A, Menezes J. Enhancement of NK cell cytotoxicity by the human herpesvirus-7 via IL-15 induction. J Immunol. 1997;159:4966 – 4972. Armitage RJ, Macduff BM, Eisenman J, et al. IL-15 has stimulatory activity for the induction of B cell proliferation and differentiation. J Immunol. 1995;154:483– 490. Grabstein KH, Eisenman J, Shanebeck K, et al. Cloning of a T cell growth factor that interacts with the ␤ chain of the interleukin-2 receptor. Science. 1994;264:965–968. Giri JG, Ahdieh M, Eisenman J, et al. Utilization of the ␤ and ␥ chains of IL-2 receptor by the novel cytokine IL-15. EMBO J. 1994;13:2822–2830. Johnston JA, Bacon CM, Finbloon DS, et al. Tyrosine phosphorylation and activation of STAT5, STAT3, and Janus kinases by interleukins 2 and 15. Proc Natl Acad Sci USA. 1995;92:8705– 8709. Anderson DM, Kumaki S, Ahdieh M, et al. Functional characterization of the human interleukin-15 receptor ␣ chain and close linkage of IL15RA and IL2RA genes. J Biol Chem. 1995;50:29862–29869. Corbett TH, Griswold DP Jr, Roberts BJ, et al. Tumor induction relationships in development of transplantable cancers of the colon in mice for chemotherapy assays, with a note on carcinogen structure. Cancer Res. 1975;35:2434 – 2439. Brattain MG, Strobel-Stevens J, Fine D, et al. Establishment of mouse colonic carcinoma cell lines with different metastatic properties. Cancer Res. 1980;40:2142–2146. Anderson DM, Johnson L, Glaccum MB, et al. Chromosomal assignment and genomic structure of IL15. Genomics. 1995;25:701–706. Miller AD, Rosman GJ. Improved retroviral vectors for gene transfer and expression. BioTechniques. 1989;7:980 – 990.


261

20. Farrell HE, Vally H, Lynch DM, et al. Inhibition of natural killer cells by a cytomegalovirus MHC class I homologue in vivo. Nature. 1997;386:510 –514. 21. Bendelac A, Rivera MN, Park S-H, et al. Mouse CD1specific NK1 T cells: development, specificity, and function. Annu Rev Immunol. 1997;15:535–562. 22. Porgador A, Tzehoval E, Vadai E, et al. Combined vaccination with major histocompatibility class I and interleukin 2 gene-transduced melanoma cells synergizes the cure of postsurgical established lung metastases. Cancer Res. 1995;55:4941– 4949. 23. Tasaki K, Tagawa M, Gunji Y, et al. Inhibition of experimental lung metastasis of murine colon carcinoma cells depends on the amount of interleukin-2 secreted from the transduced cells. Anticancer Res. 1998;18:813– 818. 24. Carson WE, Ross ME, Baiocchi RA, et al. Endogenous production of interleukin 15 by activated human monocytes is critical for optimal production of interferon-␥ by natural killer cells in vitro. J Clin Invest. 1995;96:2578 – 2582. 25. Moore TA, von Freeden-Jeffry U, Murray R, Zlotnik A. Inhibition of ␥␦ T cell development and early thymocyte maturation in IL-7⫺/⫺ mice. J Immunol. 1996;157:2366 – 2373. 26. Nishimura H, Washizu J, Nakamura N, et al. Translational efficiency is up-regulated by alternative exon in murine IL-15 mRNA. J Immunol. 1998;160:936 –942. 27. Onu A, Pohl T, Krause H, et al. Regulation of IL-15 secretion via the leader peptide of two IL-15 isoforms. J Immunol. 1997;158:255–262. 28. Ghayur T, Banerjee S, Hugunin M, et al. Caspase-1 processes IFN-␥-inducing factor and regulates LPS-induced IFN-␥ production. Nature. 1997;386:619 – 622. 29. Gu Y, Kuida K, Tsutsui H, et al. Activation of interferon␥-inducing factor mediated by interleukin-1␤-converting enzyme. Science. 1997;275:206 –209. 30. Gunji Y, Tagawa M, Matsubara H, et al. Murine colon carcinoma cells engineered to produce human interleukin-2 induce a tumor-specific anti-tumor response. Int J Cancer. 1996;66:135–139. 31. Warren HS, Kinnear BF, Kastelein RL, et al. Analysis of the costimulatory role of IL-2 and IL-15 in initiating proliferation of resting (CD56dim) human NK cells. J Immunol. 1996;156:3254 –3259. 32. Voss SD, Daley J, Ritz J, et al. Participation of the CD94 receptor complex in costimulation of human natural killer cells. J Immunol. 1998;160:1618 –1626.