Interleukin-12: Biological Properties and Clinical Application Free

Interleukin-12 (IL-12) is recognized as a master regulator of adaptive type 1, cell-mediated immunity, the critical pathway involved in protection against neoplasia and many viruses. This is supported by the analysis of numerous animal (1, 2) and human clinical studies that attribute improved clinical outcome (3) and mechanisms of IL-12–based therapy (4) to strong type 1 responses in situ. Since the initial preclinical and clinical studies of IL-12, done over a decade ago, basic and translational science studies have contributed to the greater understanding of IL-12 immunobiology. In addition to its noted effects in the priming of T helper 1 (TH1) cell responses and IFN-γ production by T and natural killer (NK) cells, more recent studies support its critical role as a third signal for CD8⁺ T cell differentiation (5, 6), and its ability to serve as an important factor in the reactivation and survival of memory CD4⁺ T cells (7). This is particularly relevant in the repolarization of CD4⁺ T cells from dysfunctional antitumor TH2 into TH1 cells in the cancer setting (8). Here, we review the immunomodulatory and antiangiogenic functions of IL-12, as well as the results of preclinical and clinical studies in which IL-12 was used as an anticancer agent.

IL-12 was independently discovered by Trinchieri and colleagues (in 1989) and by Gately and colleagues (in 1990) as “natural killer–stimulating factor” and as “cytotoxic lymphocyte maturation factor”, respectively (9, 10). It was identified as a heterodimeric cytokine composed of two covalently linked p35 and p40 subunits and was present in the supernatant of phorbol-ester–induced EBV-transformed B cell lines (9, 10). Initial characterization of its biological activities revealed that IL-12, when added to human peripheral blood lymphocytes, induced IFN-γ production, increased NK cell cytotoxicity as well as T cell proliferation in response to mitogenic lectins and phorbol diesters (see ref. 11 for review). Subsequent studies indicated that IL-12 could boost the generation of cytotoxic T cells by promoting the transcription of genes encoding cytolytic factors including perforin and granzymes (11). In 1993, Hsieh et al. (12) discovered that IL-12, produced by macrophages in response to microbial pathogens, was a key cytokine in TH1 T cell differentiation. This finding established the central role of IL-12 in a pathway in which innate immune cells drove the adaptive immune response, polarizing naïve CD4⁺ cells towards the TH1 phenotype. The general model on the biological role of IL-12 predicts that this cytokine is required for resistance to bacterial and intracellular parasites, as well as for the establishment of organ-specific autoimmunity (11). According to such a model, IL-12, produced by activated hematopoietic phagocytic cells (monocytes, macrophages, and neutrophils), by dendritic cells (DC), and by the recently identified IFN-producing killer DC lineage (13), acts as a critical regulator of cell-mediated responses. The biological functions of IL-12 are mediated by the IL-12 receptor (14) composed of two chains (β1 and β2). Triggering of the receptor activates the JAK-STAT signaling pathway, with STAT4 being the predominant mediator of cellular responses activated by IL-12 (14).

The production of the p35 and p40 subunits, as well as of the biologically active IL-12 heterodimer (i.e., IL-12p70) by phagocytic cells, require specific “priming signals” from bacterial products, and “amplification” signals dependent on cytokines produced by T cells or DC subtypes, and on cell-cell interactions. Production of the IL-12 p40 subunit by myeloid cells requires Toll-like receptor engagement by damage-associated molecular pattern molecules or pathogen-associated molecular pattern molecules, but T cell–derived cytokines such as IFN-γ are required for the optimal production of both p35 and p40 subunits (15). IL-12 production by conventional DCs in response to TLR9 agonists (CpG) requires IL-15 production leading to up-regulation of CD40 on such APCs (16). This in turn allows plasmacytoid DCs to promote IL-12 production in conventional DCs, through CD40L-CD40 interaction (16). At the T-DC interface, T cells contribute to amplify IL-12 production by DCs, in response to bacterial priming, through cell-cell interactions mediated by CD40L on activated T cells and CD40 on DCs (17). In addition, recent results indicate that IL-12 production by mature DCs requires a signaling event mediated by the tumor necrosis factor-α intracellular domain that is released upon intramembrane proteolysis by SPPL peptidases (18). In turn, the tumor necrosis factor-α intracellular domain can signal to the nucleus for IL-12 production (18). Furthermore, simultaneous stimulation of human monocytes with TLR4 and TLR8 agonists leads to IL-12p70 production, even in the absence of overt T cell help (19). Thus, DCs have a “combinatorial code” for optimal IL-12 production dependent on TLR3 and TLR4 acting in synergy with TLR7, TLR8, or TLR9 (20). Such “coding” may allow DCs to express an efficient TH1 polarizing program upon appropriate stimulation by pathogen- or damage-associated molecular pattern molecules.

Cytokines such as IL-10 and transforming growth factor-β1 can negatively regulate IL-12 production. Both IL-10 and transforming growth factor-β1 suppress transcription of the IL-12 p40 subunit, thus limiting the amount of biologically active p70 heterodimer (21, 22). In addition, the cytokine formed by homodimerization of the IL-12 p40 subunit [i.e., IL-12(p40)₂] acts as an antagonist of IL-12p70 biological activity by binding to the β1 subunit of the IL-12 receptor (23), although this mechanism may act primarily in mouse cells and not in human cells, due to the low affinity of the human p40 homodimer for the IL-12 receptor. Furthermore, recent evidence indicates that the IL-12(p40)₂ homodimer can have agonist activity, as suggested by its ability to promote DC migration to draining lymph nodes and naïve T cell activation in response to Mycobacterium tuberculosis infection (24).

Early models placing IL-12 at the beginning of the TH1 T cell differentiation pathway have been progressively revised, thanks to the emerging evidence supporting the role of T-bet transcription factor in TH1 commitment (25). In 2001, Mullen et al. (26) reported that, in the mouse, T-bet could specify TH1 effector fate without the involvement of IL-12. Such TH1 fate determination was achieved by the effects of T-bet on chromatin remodeling of IFN-γ alleles and by the induction of IL-12Rβ2 expression. This led to a revised TH1 differentiation model (27) placing T-bet upstream of IL-12. According to such a revised model, T-bet expression was induced in naïve T cells by signals derived from innate cells (i.e., NK cell–derived IFN-γ, signaling through STAT1), whereas IL-12 maintained a role as a survival signal acting on cells with a predetermined TH1 fate (27). Recent results further clarified that the contribution of T-bet to TH1 fate determination depends on its activity as a negative regulator of GATA-3, the main regulator of T cell commitment to the TH2 lineage, rather than as solely a positive regulator of the IFN-γ gene (28).

The discovery of new IL-12 family members, IL-23 and IL-27 (29), has prompted a further revision of the role of IL-12 in the differentiation of TH lineages and in some T cell–dependent autoimmune and inflammatory diseases. IL-23 shares the p40 chain with IL-12, but such subunit associates with a p19 chain (29). IL-27 is a heterodimeric cytokine which is composed of the EBV-induced molecule 3 that associates with the IL-27 p28 chain (29). Similar to IL-12, IL-23 and IL-27 are produced predominantly by macrophages and dendritic cells and affect IFN-γ production by T and NK cells (29). IL-23 serves to promote TH4/TH₁₇ cells producing the cytokines IL-6, IL-17, IL-22, and IL-25. All three cytokines (IL-12, IL-23, and IL-27) seem to play roles in the priming and re-activation of polarized T cell responses. Both IL-12p70 and IL-27 can exert effects in the priming of TH1 CD4⁺ T cell responses (30, 31), with recent data suggesting a critical role for T regulatory cell survival. IL-12 and IL-23 enhance the responses of memory T cells (32), whereas IL-27 engagement with its receptor seems to limit inflammatory T cell responses (33). Furthermore, IL-27 suppresses the development of TH4/TH₁₇ cells (34), whereas contributing to TH1 development by promoting the expression of T-bet and IL-12Rβ2 (29).

Brunda et al. (35), as well as Tahara et al. (36) and Nastala et al. (37) showed that IL-12 has potent in vivo antitumor and antimetastatic activity against murine tumors. Interestingly, the efficacy of IL-12 was greatly reduced, but not abolished, in immune-incompetent mice. By building on these observations, in 1995, Folkman and colleagues discovered the potent antiangiogenic properties of IL-12 (38). They found that IL-12 treatment inhibited basic fibroblast growth factor–induced corneal neovascularization in both immunocompetent and immunodeficient mice (38). Suppression of angiogenesis by IL-12 was dependent on its ability to induce IFN-γ expression (see Fig. 1 for an outline of IL-12 effects on angiogenesis). Accordingly, administration of IFN-γ reproduced the antiangiogenic effects promoted by IL-12 (38). Trinchieri and colleagues (39) injected neoplastic cells engineered to secrete murine IL-12 or IL-18 in syngeneic mice and found significant protection against the growth of a concurrent tumor, not secreting cytokines, and injected at a distant site. This effect was shown to be mediated by inhibition of angiogenesis by IL-12 and IL-18 released by transfected neoplastic cells, rather than by the promotion of antitumor immunity (39). In a related article, the same group also showed that inhibition of tumor growth by IL-12 or IFN-γ required an intact signaling from IFN-γ receptors expressed in neoplastic cells. This indicated that IL-12 could inhibit tumor growth by inducing neoplastic cells to produce antiangiogenic factors (40). Two of the most relevant factors were soon identified as the IFN-γ–inducible genes IP-10 and Mig (41, 42). Initial results obtained by Sgadari et al. (43) showed that intratumor delivery of Mig into Burkitt's tumors, growing subcutaneously in nude mice, led to tumor necrosis associated with vascular damage. Subsequently, these authors found that treatment in vivo with IL-12 led to the expression of IP-10 and Mig genes in tumor cells (42). In addition, administration of neutralizing antibodies to IP-10 and Mig substantially reduced the antitumor effects of IL-12 (42).

The available evidence puts IFN-γ and lymphocyte-endothelial cells cross-talk at the beginning of the process of inhibition of angiogenesis by this cytokine (Fig. 1). Dias et al. (44) found that IFN-γ, induced by IL-12, reduced tumor cell production of vascular endothelial growth factor. Moreover, IL-12 treatment reduced the production of metalloproteases playing a role in matrix remodeling, a process required during neoangiogenesis (45). Furthermore, IL-12–induced IFN-γ reduces activation of integrin αVβ3 on endothelial cells (46), leading to decreased endothelial cell adhesion and survival (46). The relevance of lymphocyte-endothelial cells cross-talks has been investigated by Strasly et al. (47). They found that IL-12 activates an antiangiogenic program in lymphocytes that leads to the production of IP-10 and Mig (Fig. 1). These factors in turn negatively modulate the cycle of endothelial cells, the production of matrix metalloproteinase-9, the ability of endothelial cells to adhere to vitronectin and to up-regulate intercellular adhesion molecule 1 and vascular cell adhesion molecule-1 expression (47). The antiangiogenic program activated in lymphocytes by IL-12 can also directly affect gene expression in neoplastic cells. In fact, up-regulation of STAT1, IRF-1, LMP2, LMP7, Mig, monocyte chemoattractant protein 1, and angiopoietin 2 genes, with down-modulation of vascular endothelial growth factor, has been documented in neoplastic cells exposed to soluble factors released by IL-12–stimulated lymphocytes (48). Further studies have also shown that NK cells are important effectors of the antiangiogenic activity of IL-12. In fact, NK cells accumulate at sites of IL-12–mediated inhibition of angiogenesis and are cytolytic for endothelial cells at such sites (49).

The antitumor and antimetastatic activities of IL-12 have been extensively shown in murine models including melanomas, mammary carcinomas, colon carcinoma, renal carcinoma, and sarcomas (see ref. 50 for review). Some of these studies have addressed the issue of local IL-12 production versus systemic delivery (i.e., intraperitoneally). Production of IL-12 at the tumor site (by neoplastic cells engineered to release IL-12 by appropriate expression vectors) induces the rejection of neoplastic cells by CD8⁺ T cells associated with macrophage infiltration, vessel damage, and necrosis (51). Interestingly, the cure rates of mice bearing established tumors were much higher following i.p. administration of rIL-12 when compared to vaccination with tumors releasing IL-12 (51). Improved antitumor effects have been shown when IL-12 was administered with other cytokines such as IL-2 and IL-18 (39) or with neoplastic cells expressing costimulatory molecules (52). Analysis of the immune mechanisms activated by IL-12 in these preclinical models has suggested the role of several subsets, including NK cells, CD4⁺ and CD8⁺ T cells, and CD3⁺ CD56⁺ NK-T cells expressing the Vα14 invariant T cell receptor (35, 53).

The activity of IL-12 has been investigated in patients with advanced solid tumors and hematologic malignancies (54–75), as either monotherapy (Table 1), or in combination with other therapies (Table 2). With the exception of the results obtained in cutaneous T cell lymphoma variants (59, 68), in AIDS-related Kaposi sarcoma (67) and non–Hodgkin's lymphoma (69), efficacy was minimal, with an objective response rate ranging between 0% and 11% (Tables 1 and 2). In the first published trial, Atkins and colleagues (54) enrolled 40 patients, including 20 with renal cancer and 12 with melanoma, in a phase I dose escalation study of i.v. administered recombinant human IL-12 (rHuIL-12). A transient complete response in a patient with melanoma and one partial response in a patient with renal cell cancer were documented. Bajetta and colleagues enrolled 10 pretreated patients with advanced melanoma in a pilot study (56). The patients received a fixed dose of rHuIL-12 (0.5 μg/kg) on days 1, 8, and 15 for two sequential 28-day cycles. No partial or complete responses were documented, but tumor shrinkage involving subcutaneous metastases, superficial adenopathy, and hepatic metastases was observed. Immune monitoring of these patients, by Mortarini et al. (57), indicated that IL-12 administration induced a striking burst, in the periphery, of HLA-restricted CTL precursors directed to autologous tumors and to an immunogenic tumor-associated antigen (Melan-A/Mart-1_26-35 peptide). Interestingly, infiltration of neoplastic tissue by CD8⁺ T cells with a memory and cytolytic phenotype was identified by immunohistochemistry in eight of eight posttreatment metastatic lesions, but not in five of five pretreatment metastatic lesions from three patients (57). These results provided the first evidence that rHuIL-12 can boost the frequency of circulating antitumor CTL precursors in tumor patients and promote infiltration of neoplastic lesions by CD8⁺ memory T cells in a clinical setting.

Gollob et al. (60), using an administration schedule based on twice-weekly injections for 6 weeks, found raised blood levels of IFN-γ, IL-15, and IL-18 in treated patients. Interestingly, whereas IFN-γ and IL-15 induction was attenuated during the first cycle in patients with disease progression, patients with tumor regression or stable disease showed constant higher levels of IFN-γ, IL-15, and IL-18 (60). In 2001, Motzer et al. (61) reported the results of a randomized phase II trial on 46 patients comparing rHuIL-12 IFN-α in patients with previously untreated and advanced renal cell carcinoma. Only 2 of 30 patients treated with rHuIL-12 achieved a partial response.

In order to avoid the IL-12–related toxicity, observed after systemic administration, locoregional treatment with IL-12 was investigated in two small clinical trials. Lenzi et al. (62) administered IL-12 to 29 previously treated patients with peritoneal carcinomatosis from various abdominal cancers through an indwelling peritoneal catheter. Two patients (one with ovarian cancer and one with surgically pretreated mesothelioma and with small residual disease) had no remaining disease at laparoscopy, eight patients had stable disease and the remainder had progressive disease (62). In a phase I study, Weiss et al. (63) evaluated intravesical IL-12 administration in patients with recurrent superficial transitional cell carcinoma of the bladder that had failed at least one prior intravesical therapy, or had at least two recurrences of low-grade tumors. Each patient received intravesical IL-12 weekly for 6 weeks. The treatment was well-tolerated, but no objective responses were observed (63).

Lee et al. (70) reported a trial in which 48 patients with resected stage III or IV melanoma were immunized with peptides derived from tyrosinase and gp100, with or without s.c. administration of IL-12. IL-12 augmented peptide-specific delayed-type hypersensitivity reactivity to the gp100 antigen in 34 of 40 patients. Moreover, the treatment boosted the gp100-specific and tyrosinase-specific peripheral immune response, as measured by IFN-γ release in 37 of 42 patients. In another trial conducted by Cebon and colleagues (71), rHuIL-12 was administered s.c. or i.v. in two cohorts of stage III or IV melanoma patients expressing Melan-A/Mart-1 in their tumors. Melan-A/Mart-1_26-35 and influenza matrix_58-66 peptides were administered intradermally. Clinical responses were mostly mixed, but one complete response and one stabilization of the disease were achieved in the i.v. arm, one partial response, and five stabilizations were attained in the s.c. arm. Cutaneous delayed-type hypersensitivity reactions were associated with CD4⁺ and CD8⁺ T lymphocyte infiltration capable of lysing a Melan-A/Mart-1 peptide-pulsed target in vitro (71). Peterson et al. (72) immunized 20 pretreated patients with advanced melanoma with PMBC loaded with Melan-A/Mart-1 peptide plus IL-12. Two patients achieved a complete response, five patients a minor or mixed response, and four patients had stable disease. The overall median survival was 12.25 months and seven patients remained alive at the time of data analysis, with all patients followed for more than 12 months (72).

In the study done by Gollob et al. (73), pretreated patients with metastatic renal cell cancer, melanoma, or transitional cell cancer were treated with 6-week cycles of twice-weekly i.v. rHuIL-12 plus IL-2 s.c. There was one partial response and two pathologic responses, all of which occurred in patients with melanoma. When administered at the maximum tolerated dose, IL-2 significantly augmented IFN-γ and IP-10 production by rHuIL-12 and led to a 3-fold expansion of NK cells (73). Alatrash et al. (74) treated 26 patients affected by metastatic melanoma or metastatic renal cell cancer with escalating doses of IL-12 and IFN-α2b. Three patients had a partial response and the median overall survival was 13.8 months (74). More recently, in a limited phase I clinical trial, IL-12 was administered with trastuzumab in patients with Her2⁺ solid tumors (75), but the addition of the cytokine did not seem to enhance the efficacy of this antibody treatment. One patient had a complete response and two patients had stabilization of disease (75).

Systemic administration of IL-12 in patients is limited by toxicity. Based on the promising results obtained in a large series of preclinical IL-12 gene therapy studies (see ref. 76 for review), clinical trials have been designed with the aim of achieving production of the cytokine at the tumor site, whereas maintaining low serum concentrations to reduce systemic toxicity. Kang et al. (77) enrolled seven patients with advanced malignancies accessible from the body surface in a phase I dose-escalation clinical study of peritumoral injection of IL-12–transduced autologous fibroblasts. Weekly injection consisted of sufficient fibroblasts to secrete an estimated 300 ng of IL-12 per 24 h. Transient reductions of tumor sizes were observed at the injected sites in four patients and at noninjected distant sites in one melanoma patient. Overall, five of seven patients showed clinical responses. No clinically significant toxicities were reported (77). In another study, patients with melanoma received s.c. injections at weekly intervals of autologous tumor cells transduced with two independent eukaryotic expression vectors coding for the p35 and p40 subunits of IL-12 (78). Two patients developed delayed-type hypersensitivity reaction against their autologous melanoma cells and one had a minor clinical response (78). Heinzerling et al. (79) did a study in which plasmid DNA encoding human IL-12 was injected into lesions of pretreated patients with advanced malignant melanoma. Intratumoral injection resulted in local responses at the injection site in all but one of the nine treated patients, and led to significant local tumor reduction in five patients. Overall, two instances of “stable disease” and one complete remission were achieved. Real-time quantitative PCR on biopsies taken 24 h after the last injection revealed a more pronounced increase in IL-12, IFN-γ, and IP-10 in responders than in nonresponders (79). Triozzi et al. examined the activity of the intratumoral injection in melanoma patients of two vectors, one encoding the costimulatory molecule B7.1 and the other encoding IL-12. Nine patients received only the B7.1-encoding vector and five received both vectors. No clinical responses were observed (80). Mazzolini et al. (81) described the intratumoral injection of dendritic cells engineered to secrete IL-12 in 17 patients with metastatic gastrointestinal tumors. A partial response was observed in a patient with pancreatic carcinoma and infiltration of CD8⁺ lymphocytes was documented in 3 of 11 biopsies analyzed.

The reasons for the limited clinical efficacy of IL-12 as a biological response modifier in cancer patients remain incompletely understood. Early clinical studies (82) showed evidence of an “adaptive response” that down-regulated the pharmacodynamics of IL-12 following the first administration of cytokine. Subsequent studies (56, 57) confirmed that the effects of IL-12 on IFN-γ levels and the frequency of circulating tumor-specific T cells were greatly reduced after the first cytokine administration. These results suggested that the antitumor activity of IL-12 could be progressively inhibited upon continuing administration to patients. Thus, a strategy of “crescendo ma non troppo” (increasing but not too much), by increasing the dose of IL-12 to overcome the tachyphylactic response, was reasonable but never tested. In addition, the immune suppression–dominated microenvironment in advanced tumors (83) is likely a major factor contributing to limit the efficacy of IL-12–based therapies. Indeed, CD4⁺ T cell depletion that presumably removes regulatory T cells has been reported to enhance the antitumor effectiveness of IL-12 in mice (84). However, IL-12 itself does not seem to directly modulate the function or frequencies of CD4⁺CD25⁺FOXP3⁺ regulatory T cells in vitro (85), thus, combinational approaches designed to deplete preexisting regulatory immune cells, or to limit their function, could be used. A broad-based approach could include preconditioning immunodepletion such as low-dose cyclophosphamide (86), followed by an IL-12–based vaccine, or IL-12–based therapy in conjunction with antibodies targeting regulatory T cells (87), like the anti–CTLA-4 monoclonal antibody recently used in clinical studies. As IL-12 may itself contribute to immunoregulation through the compensatory induction of IL-10 (88), circumventing the immunosuppressive effects of IL-10 could also be envisioned to enhance the therapeutic index of IL-12. This could be achieved through blocking IL-10 binding to cellular receptors by coadministration of a soluble IL-10R or anti–IL-10 antibody, or preferably, by obstructing the STAT3 signaling pathway induced by IL-10 using a variety of inhibitors (89). As a corollary, negative feedback pathways mediated by SOCS-1 (90) or SOCS-3 (91) molecules induced by IL-12 signaling could also represent attractive targets for small molecule inhibitors or RNAi.

A further promising area of investigation, with potentially relevant clinical applications, stems from the emerging knowledge on newly discovered IL-12 family members. For example, in murine models, both IL-23 and IL-27 have been used to effectively treat tumors (92, 93). However, due to the role of IL-23 in inflammatory diseases, caution would be justified because such inflammation could conceivably contribute to tumor progression rather than tumor regression (94). One could envision combinatorial approaches using IL-12 and IL-27 to promote naïve T cell responses against tumor antigens, and/or IL-12 and IL-23 to promote memory antitumor responses. Carefully designed combinatorial IL-12 family regimens could potentially exploit a natural mechanism for regulating type 1 immune (re)activation, without triggering chronic inflammation. This could lead to a more durable and effective biological antitumor response than that obtained with IL-12 alone, and hence, greatly improve the therapeutic efficacy of IL-12–based treatments. Further clinical studies could be developed by exploiting the antiangiogenetic program promoted by IL-12 in lymphocytes and in neoplastic cells. To this end, IL-12–based therapy might be combined with other antiangiogenic molecules such as bevacizumab or raf-kinase inhibitors (95).