Interleukin 12: Basic Biology and Potential Applications in Cancer Treatment

Interleukin 12 is a heterodimeric cytokine that has potent effects on innate and adaptive immunity. Interleukin 12 induces interferon γ secretion by T cells and natural killer cells, enhances the proliferation of activated T cells and natural killer cells, augments the cytolytic activity of cytotoxic T lymphocytes and natural killer cells, and supports the differentiation of Th1 helper effector cells. Interleukin 12 stimulates in vitro antitumor activity of lymphocytes from patients with cancer and in vivo antitumor activity in many murine tumor models. Current data indicate that CD4 T cells, CD8 T cells, natural killer cells and interferon γ may contribute to the antitumor effects of interleukin 12 therapy. However, further investigation is required to elucidate the precise mechanisms involved in the antitumor activity of interleukin 12. The Oncologist 1996;1:88-97


INTRODUCTION
Interleukin 12 (IL-12) is a heterodimeric cytokine that was independently identified by two groups. Trinchieri and colleagues isolated a cytokine called natural killer cell stimulatory factor (NKSF) from the supernatants of B lymphoblastoid cell lines and subsequently cloned cDNA encoding this cytokine in collaboration with investigators at Genetics Institute (Cambridge, MA) [1, 2]. Gately and colleagues at Hoffmann-LaRoche (Nutley, NJ) isolated cDNA encoding a cytokine they had identified and called cytotoxic lymphocyte maturation factor (CLMF) [3,4]. The deduced amino acid sequences of NKSF and CLMF were found to be identical and the cytokine was thereafter designated IL-12. IL-12 has been shown to exert potent immunostimulatory effects on certain helper T cells, as well as on cytotoxic T lymphocytes (CTL) and natural killer (NK) cells. Preclinical studies suggest that IL-12 may prove useful in the treatment of several human diseases, including atopic and allergic conditions, HIV infection and cancer. This article will focus on the possible role of IL-12 in the treatment of malignant disease. A comprehensive summary of the biology of IL-12 is beyond the scope of this article; topics not dealt with here have been presented in other recent reviews [5][6][7][8].

STRUCTURE OF IL-12 AND THE IL-12 RECEPTOR
Human IL-12 is a disulfide-linked heterodimer with an apparent molecular mass of 70 kDa; it is composed of two subunits with masses of 40 and 35 kDa [1, 3]. The p40 chain is homologous to several cytokine receptors, including the IL-6 receptor and the receptor for ciliary neurotrophic factor [9,10]. In contrast, p35 is homologous to other cytokines, including IL-6, G-CSF and chicken myelomonocytic growth factor [11]. Neither p40 alone nor p35 alone appears to have significant biologic activity, and a combination of soluble p40 with soluble p35 exhibits bioactivity that is markedly inferior to that of heterodimeric IL-12 [2,4,5].
IL-12 receptors (IL-12R) have been demonstrated on resting and activated NK cells, activated CD4 T cells and activated CD8 T cells [12][13][14]. However, IL-12R have not been detected on resting T cells, resting monocytes, or on resting or activated B cells. Scatchard analysis of radiolabeled IL-12 binding to activated T cells demonstrates three distinct binding sites with high, intermediate and low affinity [15]. Molecular cloning of one component of the IL-12R has been reported [15]. This IL-12R subunit is a member of the hematopoietic cytokine receptor superfamily and is most similar to gp130, a common signaling component of the receptors for IL-6, IL-11, leukemia inhibitory factor and oncostatin M. The available data, however, suggest that the high affinity IL-12R contains one or more additional subunits. Indeed, identification of another putative component of the IL-12R has recently been presented (U. Gubler, presentation at New York Academy of Sciences Conference, November 1995).
Robertson, Ritz 89 Relatively little has been published regarding signal transduction after binding of IL-12 to its cell surface receptor. IL-12 induces rapid tyrosine phosphorylation of mitogen activated protein (MAP) kinase species in activated T cells [16]. However, MAP kinase phosphorylation was not detected in resting T cells or NK cells stimulated with IL-12 [16,17]. Janus (JAK) kinases and the Stat family of signal transducer and transcription factors have been implicated in signal transduction induced by a variety of hematopoietic cytokines [18,19]. IL-12 stimulates tyrosine phosphorylation of specific JAK kinases in activated T cells and NK cells [20], and of Stat3 and Stat4 in some CD4 T cell clones [21,22]. Moreover, IL-12 induces rapid tyrosine phosphorylation of the src family tyrosine kinase lck in resting NK cells [17,23]. The precise signaling sequence generated by IL-12 stimulation of T cells or NK cells remains to be elucidated.

PRODUCTION OF IL-12
Regulation of IL-12 production is rendered complex by the heterodimeric structure of this cytokine. Messenger RNA for the p35 subunit is constitutively expressed by many different cell types, whereas the message for p40 is highly restricted and appears to be expressed only by cells that produce heterodimeric IL-12 [24,25]. Cells that secrete biologically active p40/p35 heterodimers usually also secrete free p40 chains in approximately 10-fold excess. By contrast, secretion of significant amounts of p35 in the absence of p40 has not been reported. In vitro experiments have shown that p40/p40 homodimers can block binding of p40/p35 heterodimers to the IL-12R and can act as competitive antagonists of IL-12 in various bioassays [26,27]. However, it is unclear what fraction, if any, of physiologically secreted free p40 chains exists as potentially antagonistic homodimers.
Some Epstein-Barr virus-transformed B cell lines constitutively produce IL-12. Phorbol esters strongly augment IL-12 production by B lymphoblastoid cell lines [24]. Several myeloid leukemia and epidermoid carcinoma cell lines can also produce IL-12 after activation [25,28]. However, the major physiologic source of IL-12 appears to be antigen-presenting cells, including monocytes and macrophages, B cells and dendritic cells [24,29]. Neutrophils, mast cells and keratinocytes have also been shown to produce IL-12 under certain experimental conditions, but production of IL-12 by T cells or NK cells has not been convincingly demonstrated [28,[30][31][32]. In contrast to B lymphoblastoid cell lines, normal hematopoietic cells produce little IL-12 in response to phorbol ester. Rather, secretion of IL-12 by normal cells appears to be most effectively stimulated by certain microorganisms [5,24]. Endotoxin is a particularly potent inducer of IL-12; the other microbial products that stimulate IL-12 production have not been well characterized. Interferon γ (IFN-γ) can enhance and IL-10 can inhibit IL-12 production in response to endotoxin or other stimuli [25,33].
IL-12 does not induce the proliferation of resting T cells but can support the growth of T cells that have been activated by mitogens or specific antigens [54][55][56]. IL-12 can enhance the proliferation of activated CD4 and CD8 α/β T cells as well as γ/δ T cells. Similar to IFN-γ production, T cell proliferation in response to IL-12 is augmented by signals delivered through the CD2 and CD28 surface antigens [40][41][42]. The effects of IL-12 on NK cell proliferation are complex. IL-12 induces very little proliferation of resting NK cells and, unlike IL-2, cannot support the growth of human NK cells in vitro [51,[56][57][58]. Moreover, IL-12 inhibits optimal NK cell proliferation in response to IL-2 [56,57]. IL-12 can also inhibit IL-2-induced proliferation of γ/δ T cells [56]. Nevertheless, IL-12 upregulates IL-2 receptor expression on NK cells [51,57,59] and enhances NK cell proliferation in response to suboptimal concentrations of IL-2 [51,56].

EFFECTS ON CYTOTOXIC LYMPHOCYTES
Cytotoxic lymphocytes are operationally defined by their ability to kill other cells. Two major types of cytotoxic lymphocytes have been identified: CTL and NK cells [60]. CTL usually differentiate from activated CD8 T cells, although CD4 T cells can become killer cells under some circumstances. Target cell lysis by CTL is triggered by interactions between T cell receptor heterodimers on the CTL and antigenic peptides bound to major histocompatibility (MHC) glycoproteins on the target cell. CTL are thus described as antigen-specific and MHC-restricted. In contrast, NK cells do not express T cell receptors and their cytotoxicity is not limited to target cells bearing syngeneic MHC antigens [61,62]. NK cells can lyse certain malignant and virus-infected target cells in a spontaneous, antibodyindependent process known as natural killing and can lyse antibody-coated target cells by a process known as antibody-dependent cellular cytotoxicity (ADCC). The NK cell receptor for ADCC consists of oligomeric complexes of CD16, which binds to the Fc portion of IgG, and ς family members [63]. Although several candidate NK receptors have been described [64][65][66][67], the receptors that trigger natural killing have not been unequivocally defined. NK cells also express several different receptors for class I MHC molecules that can deliver negative signals that inhibit cytolysis [68,69]. The relative sensitivity of target cells to natural killing therefore appears to reflect their relative expression of ligands that engage triggering and inhibitory NK cell receptors. CTL and NK cells may act in a complementary fashion to destroy cells that have been infected by intracellular pathogens or that have undergone malignant transformation [68,69].
CTL and NK cells are potently stimulated by IL-12, accounting for its former designation as CLMF and NKSF. IL-12 enhances the generation of CTL specific for viral antigens and alloantigens both in vitro and in vivo [43,[70][71][72]. IL-12 supports the proliferation and differentiation of activated CD8 T cells into CTL effectors and stimulates the cytolytic activity of fully differentiated CTL. Augmentation of CTL responses by IL-12 thus appears to be due to increases in the number of CTL and in the cytotoxicity of CTL on a per-cell basis. The cytolytic activity of NK cells is also potently augmented in vitro and in vivo by IL-12 [43,51,57,70,73]. IL-12 enhances the in vitro lysis of NKsensitive and NK-resistant tumor cells, antibody-coated target cells and virus-infected cells [57,73,74]. IL-12 stimulates increased expression of cytolytic effector molecules, such as perforin and serine esterases, that are common to both CTL and NK cells [53,75,76]. Moreover, IL-12 induces the upregulation of several adhesion molecules involved in binding of NK cells to their target cells [57,59]. Optimal concentrations of IL-2 generally augment NK cell cytotoxicity to a greater degree than optimal concentrations of IL-12; however, the concentration of IL-12 required for an optimal effect is approximately 100-fold lower on a molar basis compared to that of IL-2.

EFFECTS ON HELPER T CELLS
Activated CD4 T cells can differentiate into two major types of helper effector cell, T helper type 1 (Th1) or Th2 [77]. Th1 cells secrete predominantly IL-2, IFN-γ and TNF, and support the differentiation of activated CD8 T cells into CTL and of monocytes into macrophages. Th1 cytokines also stimulate the cytolytic activity of NK cells and cause activated B cells to produce antibodies of isotypes that optimally induce ADCC by NK cells and macrophages. Thus, Th1 cells promote cell-mediated immunity to intracellular pathogens, allografts and malignant tumor cells. In contrast, Th2 cells produce IL-4, IL-5 and IL-10, which inhibit the activation of monocytes and NK cells and can cause CD8 T cells to differentiate into noncytotoxic "suppressor" cells. Thus, Th2 cytokines tend to inhibit cell-mediated immunity while promoting eosinophilia and IgE production, which are characteristic of immune responses to allergens and helminths. IL-12 promotes the differentiation of activated CD4 T cells into Th1 effector cells [78]. IL-12 appears to support Th1 development both by direct effects on activated CD4 T cells and indirectly through the induction of IFN-γ secretion [79][80][81][82]. IL-12 also stimulates Th1 effector cell proliferation and cytokine production. IL-12 may thus enhance cell-mediated immunity by its combined effects on the differentiation and function of activated CD4 T cells, CD8 T cells and NK cells.

EFFECTS ON HEMATOPOIETIC CELLS
IL-12 by itself does not appear to support the growth of immature hematopoietic precursors. Nevertheless, IL-12 can augment in vitro proliferation and/or differentiation of human and murine hematopoietic progenitor cells in response to IL-3, stem cell factor and erythropoietin [83][84][85][86][87]. The synergy of IL-12 with other hematopoietic cytokines appears to be due to direct effects on immature hematopoietic cells. When purified NK cells are added to in vitro cultures of human progenitor cells, however, IL-12 can inhibit formation of myeloid and erythroid colonies [85]. This effect appears be mediated by inhibitory cytokines, including IFN-γ and TNF-α, that are produced by IL-12-activated NK cells.
The effects of IL-12 on hematopoiesis in vivo are complicated. Prolonged administration of IL-12 to healthy mice results in bone marrow hypoplasia and lymphopenia; dosedependent anemia, thrombocytopenia and granulocytopenia also occurred in some studies [43,88,89]. Induction of IFN-γ appears to contribute to the in vivo inhibition of murine hematopoiesis by IL-12 [90]. Lymphopenia, anemia and decreased platelet counts have also been observed after Robertson, Ritz 91 administration of human IL-12 to primates [91,92]. In squirrel monkeys, however, IL-12 induces peripheral blood lymphocytosis and bone marrow hyperplasia [92]. In both mice and monkeys, IL-12 causes splenomegaly and extramedullary hematopoiesis in the spleen. The hematologic effects of IL-12 in vivo are generally rapidly reversible [88,91].

EFFECTS OF IL-12 ON IN VITRO ANTITUMOR ACTIVITY OF LYMPHOCYTES FROM PATIENTS WITH CANCER
IL-12 can augment the cytolytic activity of lymphocytes from patients with cancer [93][94][95][96][97][98]. After overnight incubation in picomolar concentrations of IL-12, peripheral blood lymphocytes of cancer patients can lyse both leukemic and solid tumor cell lines [93,94]. Moreover, IL-12 enhances the lysis of autologous tumor cells by tumor infiltrating lymphocytes from patients with melanoma and ovarian cancer [96,97]. IL-12 also potently augments the antitumor activity of NK cells obtained from patients with hematologic malignancies undergoing allogeneic bone marrow transplantation [93] and patients with solid tumors that have received prolonged infusions of low-dose IL-2 [57,93,99]. IL-2 and IL-12 act additively or synergistically in many of these in vitro experimental systems.

ANTITUMOR EFFECTS OF IL-12 IN HUMAN/SEVERE COMBINED IMMUNODEFICIENCY (SCID) MOUSE MODELS
Mice with SCID lack mature T cells and B cells, and do not reject human tumors or normal human hematopoietic cells. SCID mice have thus been used for in vivo studies of human hematopoiesis and the immunotherapy of human cancers. The growth of pulmonary metastases from human melanoma cells injected intravenously into SCID mice is inhibited by injection of polyclonal activated human NK cells [100]. The antitumor effect of human NK cells in this model is strongly augmented by the systemic administration of human IL-2 and IL-12 in combination. SCID mice injected with a human tumor cell line, U937, rapidly succumb to a disseminated malignancy resembling acute leukemia. Adoptive transfer of a cytotoxic human T cell line prolongs survival of the mice, and this protective effect is enhanced by IL-12 therapy [101]. The SCID mouse studies demonstrate that human cytotoxic lymphocytes can destroy human tumor cells in vivo and that this antitumor activity can be augmented by IL-12.

ANTITUMOR EFFECTS OF SYSTEMICALLY OR LOCALLY ADMINISTERED IL-12 IN ANIMAL MODELS
IL-12 has been shown to have potent antitumor effects in murine models of melanoma, sarcoma, kidney cancer, lung cancer, colon cancer and ovarian cancer [8,47,[102][103][104][105]. Systemic or peritumoral injection of IL-12 can induce complete regression of established tumors, inhibit the formation of distant metastases and substantially prolong the survival of tumor-bearing mice. These studies have identified doses of IL-12 that can induce impressive tumor responses without causing overt toxicity. In some tumor models, mice that had experienced complete responses after IL-12 therapy were subsequently able to reject implants of the same tumor, but not of a different tumor, suggesting that specific antitumor immunity had been established [47,103,104,106]. In models of colon cancer, ovarian cancer, lung cancer, renal cell cancer and melanoma, IL-12 was found to be more effective and/or less toxic than IL-2 [8,47,103,107,108]. Moreover, a combination of IL-2 and IL-12 was more effective than either cytokine alone in models of primary and metastatic renal cell cancer [109]. IL-12 can also augment the graftversus-tumor effect of bone marrow transplantation without promoting graft-versus-host disease [110][111][112].
The mechanisms underlying the antitumor activity of IL-12 are likely to be complex and have not been fully elucidated. Tumor regression in vivo is probably mediated through effects of IL-12 on the host immune system, since IL-12 has not been reported to directly inhibit the growth of malignant cells in vitro [106,107,[113][114][115]. T cells appear to be required for optimal efficacy of IL-12 in several tumor models. Depletion of CD8 T cells partially inhibits, and depletion of both CD4 and CD8 T cells completely eliminates, the antitumor effect of IL-12 in some tumor models [8,47,104,105]. These results suggest that CD8 and CD4 T cells can act in concert to inhibit tumor growth after IL-12 treatment. T cells do not appear to be solely responsible for the antitumor effects of IL-12, however. Indeed, NK cells have been implicated as antitumor effectors during IL-12 therapy for some tumors [102,106,113]. Taken together, the data suggest that no single effector cell is responsible for all of the antitumor activity of IL-12 in vivo. Depending on the experimental system employed, CD4 T cells, CD8 T cells, NK cells, macrophages or other cell types may participate in antitumor responses induced by IL-12.
How IL-12-stimulated effector cells inhibit tumor cell growth is also poorly understood. Activated T cells or NK cells could directly lyse tumor cells or could secrete cytokines that inhibit tumor cell proliferation or that stimulate other antitumor effector cells in vivo. Cytotoxic granule-mediated killing of neoplastic cells does not appear to be absolutely necessary for the antitumor activity of IL-12, because IL-12 therapy is efficacious against B16F10 melanoma in beige mice [102]. Beige mice possess relatively normal numbers of T cells and NK cells, but the granules of their cytotoxic lymphocytes and myeloid cells are defective [116][117][118]. Since IL-12 is known to induce IFN-γ secretion by T cells and NK cells, the possible contribution of IFN-γ to the antitumor effects of IL-12 have been investigated. Neutralization of IFN-γ in vivo was found to strongly inhibit the antitumor effect of IL-12 in murine models of sarcoma, melanoma and renal adenocarcinoma [47,102,104]. IFN-γ may contribute to tumor regression after IL-12 therapy by direct antiproliferative effects on tumor cells or by virtue of its immunomodulatory properties [102]. IFN-γ also inhibits neovascularization in vivo [119] and thus could interfere with tumor angiogenesis. Furthermore, IFN-γ has been shown to mediate the known anti-angiogenic effects of IL-12 [119]. However, antitumor activity of IFN-γ in vivo is substantially inferior to that of IL-12 [44]. IFN-γ production therefore may be necessary but not sufficient for the antitumor effects of IL-12.

ANTITUMOR EFFECTS OF CELLS TRANSDUCED WITH THE IL-12 GENES
Tumor cells or accessory cells genetically engineered to produce IL-12 have also been used in animal models of cancer immunotherapy. Syngeneic or allogeneic fibroblasts transfected with cDNA encoding murine IL-12 have been used to deliver bioactive IL-12 effectively to tumor-bearing mice [114,120,121]. Transduction of the IL-12 genes into tumor cells does not affect their proliferation in vitro but inhibits their ability to grow in immunocompetent mice [115,122,123]. Injection of IL-12-transduced sarcoma cells causes the rejection of contralateral nontransduced tumors placed either simultaneously or three days earlier [122]. Furthermore, mice injected with tumor cells secreting IL-12 can subsequently reject native, untransfected tumors upon rechallenge [122]. Thus, injection of IL-12-transduced tumor cells can induce specific and durable antitumor immunity in vivo. Both T cells and NK cells appear to participate in the immune response to tumors secreting IL-12 [115,122]. A polycistronic retroviral vector containing the human p35 and p40 genes has been described that may be suitable for use in clinical trials of IL-12 gene therapy [124,125].

CLINICAL TRIALS OF IL-12 IN CANCER PATIENTS
Based on the encouraging results of IL-12 therapy in animal tumor models and confirmation of an acceptable toxicology profile in nonhuman primates [91,92], a phase I study of recombinant human IL-12 given by bolus i.v. injection to patients with advanced solid tumors was initiated in May, 1994. This study was sponsored by Genetics Institute and was conducted at the New England Medical Center, Dana-Farber Cancer Institute, Indiana University Medical Center and the Pittsburgh Cancer Institute [126]. Cohorts of four to six patients were given a single i.v. injection of IL-12 followed two weeks later by five daily i.v. injections of IL-12 at the same dose. Patients who did not experience tumor progression or dose-limiting toxicity could receive five-day courses of IL-12 every three weeks for a maximum of six cycles. Forty patients were treated in this study, including 20 patients with renal cell cancer and 12 patients with melanoma. Doses of IL-12 as high as 500 ng/kg were given on an outpatient basis with acceptable toxicity. IL-12 therapy was associated with significant immunomodulatory effects, including production of IFN-γ in vivo. Common adverse effects included fever and chills, fatigue, nausea and headache. Transient anemia, leukopenia, thrombocytopenia, hyperglycemia and elevated serum transaminases were also frequently seen. Dose-limiting toxicities, including stomatitis and abnormal liver function tests, occurred in three of four patients treated at the 1000 ng/kg dose level. Objective tumor responses occurred in patients with renal cell cancer and melanoma.
Of 14 patients treated at the 500 ng/kg dose level in the phase I IL-12 trial described above, 10 patients completed 24 five-day treatment cycles without experiencing dose-limiting toxicity. In a subsequent phase II trial in advanced renal cell cancer, however, IL-12 given at a dose of 500 ng/kg i.v. daily for five days was associated with frequent and intolerable adverse effects [127]. This severe toxicity appears to have been due to a minor change in the schedule of IL-12 administration. Unlike patients in the initial phase I trial, patients in the phase II trial were not given a single i.v. dose of IL-12 two weeks before the first five-day cycle of treatment. Subsequent laboratory studies in mice and nonhuman primates have shown that administration of a single i.v. injection of IL-12 makes subsequent multiple-dose cycles tolerable; giving the same multiple-dose regimen without a preceding single dose causes severe toxicity [127]. The mechanisms responsible for this unanticipated and unusual schedule effect are under investigation. Nevertheless, the cause of the severe toxicity encountered in the phase II renal cell study has been elucidated, allowing the moratorium on IL-12 clinical trials to be discontinued. Phase I trials of IL-12 given by s.c. injection are currently in progress in patients with cancer and AIDS. It is hoped that successful completion of these trials will permit the antitumor efficacy of IL-12 to be examined in future phase II studies.

CONCLUSION
IL-12 is a heterodimeric cytokine that has potent effects on innate and adaptive immunity. IL-12 induces IFN-γ secretion by T cells and NK cells, enhances the proliferation of activated T cells and NK cells, augments the cytolytic activity of CTL and NK cells, and supports the differentiation of Th1 helper effector cells. IL-12 stimulates in vitro antitumor Robertson,Ritz 93 activity of lymphocytes from patients with cancer and in vivo antitumor activity in many murine tumor models. Current data indicate that CD4 T cells, CD8 T cells, NK cells and IFN-γ may contribute to the antitumor effects of IL-12 therapy. However, further investigation is required to elucidate the precise mechanisms involved in the antitumor activity of IL-12. The results of preclinical studies suggest several potential strategies for the use of IL-12 in cancer therapy. IL-12 can induce the regression of established, bulky murine tumors, but in most preclinical models, IL-12 is more effective in animals with a smaller tumor burden. Thus, although the safety of IL-12 therapy must be confirmed in phase I trials involving patients with advanced cancer, IL-12 may prove more efficacious in the context of minimal residual disease. Administration of IL-12 as adjuvant therapy could be investigated in patients at high risk for disease recurrence after surgical resection of primary solid tumors and in high-risk patients with hematologic malignancies in complete remission after induction chemotherapy. Similarly, patients with minimal residual disease after autologous or allogeneic bone marrow or peripheral blood stem cell transplantation might be appropriate candidates for treatment in IL-12 clinical trials. Preclinical data also support the use of IL-12 together with other immune manipulations, including the administration of additional immunomodulatory cytokines, adoptive transfer of immunocompetent cells, or vaccination with tumor cells engineered to secrete cytokines or to express costimulatory molecules [100,101,105,106,109,121,128]. Combining IL-12 and IL-2 has produced particularly impressive results in preclinical studies [100,109,121,128]. Properly designed and executed phase II and phase III clinical trials will ultimately be required to define the efficacy of cancer immunotherapy using IL-12.