Targeting IL-10 Family Cytokines for the Treatment of Human Diseases

Abstract

Members of the interleukin (IL)-10 family of cytokines play important roles in regulating immune responses during host defense but also in autoimmune disorders, inflammatory diseases, and cancer. Although IL-10 itself primarily acts on leukocytes and has potent immunosuppressive functions, other family members preferentially target nonimmune compartments, such as tissue epithelial cells, where they elicit innate defense mechanisms to control viral, bacterial, and fungal infections, protect tissue integrity, and promote tissue repair and regeneration. As cytokines are prime drug targets, IL-10 family cytokines provide great opportunities for the treatment of autoimmune diseases, tissue damage, and cancer. Yet no therapy in this space has been approved to date. Here, we summarize the diverse biology of the IL-10 family as it relates to human disease and review past and current strategies and challenges to target IL-10 family cytokines for clinical use.

Interleukin (IL)-10, a cytokine with pleiotropic immunosuppressive functions, is also the founding member of the IL-10 family of cytokines (Fig. 1). In addition to IL-10 itself, this group of cytokines encompasses IL-19, IL-20, IL-22, IL-24, and IL-26, which are collectively referred to as the IL-20 subfamily, as well as the more distantly related members IL-28A, IL-28B, and IL-29, also known as the interferon (IFN)-λ family or type III IFNs (Pestka et al. 2004; Ouyang et al. 2011; Rutz et al. 2014).

Figure 1.

Interleukin (IL)-10 family cytokines and their receptors.

IL-10 was initially described as a secreted cytokine synthesis inhibitory factor (CSIF) produced by T helper (Th)2 T-cell clones with the ability to inhibit the production of Th1 cytokines (Fiorentino et al. 1989). It has since been found that IL-10 is expressed by a wide variety of cell types of both the innate and the adaptive arms of the immune system, including macrophages, monocytes, dendritic cells (DCs), mast cells, eosinophils, neutrophils, natural killer (NK) cells, CD4⁺ and CD8⁺ T cells, and B cells (Moore et al. 2001; Ouyang et al. 2011). IL-20 subfamily cytokines are also produced mainly by immune cells, such as myeloid cells and lymphocytes (Rutz et al. 2014). Myeloid cells are the primary source for IL-19, IL-20, and IL-24 (Wolk et al. 2002). Epithelial cells, the main target cells of IL-20 family cytokines, also produce IL-19, IL-20, and IL-24 in response to cytokines secreted by immune cells (Hunt et al. 2006; Sa et al. 2007; Wolk et al. 2009b). T cells are a primary source for IL-22, IL-24, and IL-26. Additionally, IL-22 is produced by various other lymphoid populations, including γδ-T cells, NK cells, and innate lymphoid cells (ILCs) (Rutz et al. 2013, 2014). Finally, both leukocytes and epithelial cells are major sources of IFN-λ family cytokines (Fig. 1) (Kotenko et al. 2003; Sheppard et al. 2003; Uzé and Monneron 2007).

Although the biological functions of the other IL-10 family cytokines are quite distinct from IL-10 itself, all family members share significant structural homology, having evolved through gene duplication. Most IL-10 family cytokines form homodimers, whereas some family members, such as IL-22, exist in a monomeric form. IL-10 family cytokines signal through heterodimeric receptors, composed of class II receptor α- and β-subunits (Pestka et al. 2004). The prototypical class II receptor structure consists of tandem β–sheet-rich immunoglobulin (Ig)-like domains with fibronectin type III connectivity. The α-receptor subunits show higher affinity for the cytokine ligand than the β-subunits. Interestingly, the receptor-binding mode is virtually identical for monomeric or dimeric IL-10 family cytokines (Pestka et al. 2004).

The shared usage of common receptor subunits is a defining feature of the IL-10 cytokine family (Fig. 1). All members bind either the IL-10RB or IL-20RB β-receptor subunits in combination with varying α-subunits. IL-10 uniquely uses IL-10RA, whereas IL-19, IL-20, and IL-24 use IL-20RA as receptor α-subunits. IL-22 binds IL-22RA1, which can also be bound by IL-20 and IL-24, collectively defining the IL-20 subfamily. IL-28A, IL-28B, and IL-29, on the other hand, use a unique IL-28RA subunit. In addition to these membrane-bound receptors, a soluble form of the IL-22 receptor (IL-22BP or IL-22RA2) with homology to the extracellular domain of IL-22RA1, binds IL-22 with high affinity and blocks its activity (Ouyang et al. 2011; Rutz et al. 2014).

IL-10 family receptors signal through Janus tyrosine kinases (JAKs) and signal transducers and activators of transcription (STATs). Receptor α-subunits are constitutively associated with Jak1, whereas Jak2 or Tyk2 are bound to the β-subunits. Ligand binding initiates recruitment and phosphorylation of STATs, which in turn form homo- and heterodimers that translocate into the nucleus to induce transcription. IL-10 and the IL-20 subfamily cytokines signal primarily through STAT3 and STAT1, whereas IL-28A, IL-28B, and IL-29 activate the ISGF3 complex (Pestka et al. 2004; Ouyang et al. 2011; Rutz et al. 2014).

Distinct receptor expression patterns drive the diverse biology of IL-10 family cytokines (Fig. 1). The IL-10RB β-subunit is ubiquitously expressed throughout the body, whereas expression of the IL-20RB β-subunit is more restricted. IL-10RA is mainly expressed on leukocytes. In contrast, IL-20 subfamily receptors show a rather restricted expression pattern. In particular the α-receptor subunits, IL-20RA and IL-22RA1, are preferentially expressed on epithelial cells and fibroblasts, but absent from hematopoietic cells (Aggarwal et al. 2001; Blumberg et al. 2001; Wolk et al. 2002). IL-22RA1 is highly expressed in the skin, pancreas, kidney, lung, intestine, and liver. IL-20RA is expressed in the skin, lung, ovary, testes, and placenta, and at lower levels in the intestine and liver (Rutz et al. 2013, 2014).

TARGETING IL-10 FOR THE TREATMENT OF HUMAN AUTOIMMUNE DISEASES

As a major immune regulatory cytokine, IL-10 can be produced by many leukocyte subsets and is under the control of various signal transduction pathways and transcriptional networks (Saraiva and O’Garra 2010; Rutz and Ouyang 2011, 2016). According to the expression of its receptor, IL-10 acts on many cells of the immune system (Fig. 2), where it has profound anti-inflammatory functions (Moore et al. 2001; Rutz and Ouyang 2011). IL-10 mainly targets antigen-presenting cells (APCs), such as monocytes and macrophages, and inhibits their release of proinflammatory cytokines, such as tumor necrosis factor α (TNF-α), IL-1β, IL-6, IL-8, granulocyte colony-stimulating factor (G-CSF), and granulocyte macrophage colony-stimulating factor (GM-CSF), as well as chemokines, including MCP1, IL-8, and IP-10 (de Waal-Malefyt et al. 1991a; Fiorentino et al. 1991; Moore et al. 2001). IL-10 also interferes with antigen presentation by reducing the expression of major histocompatibility complex (MHC)-II and costimulatory and adhesion molecules (de Waal-Malefyt et al. 1991b; Willems et al. 1994; Creery et al. 1996). Furthermore, IL-10 suppresses cytokines, such as IL-12 and IL-23, which are required for CD4⁺ T-cell differentiation (D’Andrea et al. 1993; Schuetze et al. 2005). Similarly, IL-10 attenuates the production of inflammatory mediators, including cytokines and chemokines from neutrophils (Cassatella et al. 1993; Kasama et al. 1994). In addition, IL-10 can act directly on T cells to inhibit both their proliferation and cytokine production, and to induce nonresponsiveness or anergy (Groux et al. 1996). However, IL-10 also has stimulatory effects on CD8⁺ T cells, and augments their proliferation and cytotoxic activity (Groux et al. 1998; Mumm et al. 2011). It enhances the survival of human B cells and promotes B-cell proliferation (Levy and Brouet 1994; Itoh and Hirohata 1995) and contributes to the differentiation of B cells and their production and isotype switch of antibodies (Rousset et al. 1992).

Figure 2.

Cellular sources and target cells of interleukin (IL)-10 family cytokines. NK, Natural killer; ILCs, innate lymphoid cells.

The Role of IL-10 in Inflammatory Bowel Disease

Given its multiple anti-inflammatory functions, it is not surprising that IL-10 exerts essential regulatory roles in many human autoimmune diseases. Inflammatory bowel disease (IBD) comprises ulcerative colitis (UC) and Crohn’s disease (CD), both of which show uncontrolled inflammation in the intestinal tract but differ in pathophysiology. Mice deficient in either IL-10 or the IL-10 receptor α or β chains develop spontaneous colitis (Kühn et al. 1993), which is dependent on the presence of the intestinal microbiota and involves the up-regulation of IL-23 (Sellon et al. 1998; Yen et al. 2006). Exogenously provided recombinant IL-10 can delay and attenuate colitis development in these IL-10-deficient mice. In addition, IL-10 administration showed therapeutic value in several other colitis models (Hagenbaugh et al. 1997; Steidler et al. 2000).

IBD has a strong genetic association with the IL-10 pathway. Genome-wide association studies (GWAS) linked the IL10 locus with the susceptibility to both UC and CD (Franke et al. 2008, 2010). Furthermore, individuals carrying missense mutations in the IL10, IL10RA, or IL10RB genes develop very early-onset UC or neonatal CD (Glocker et al. 2009; Engelhardt et al. 2013; Shim et al. 2013). Restoring IL-10RA or IL-10 expression by hematopoietic stem-cell transplantation rapidly alleviates clinical symptoms (Engelhardt et al. 2013).

As mentioned earlier, IL-10 represses several key proinflammatory cytokines, including TNF-α and IL-23, which are clinically validated to participate in the pathogenesis of IBD. TNF-blocking agents, including anti-TNF antibodies, such as infliximab, adalimumab, and certolizumab, have shown good clinical efficacy in the treatment of both CD and UC, and are currently the standard of care for patients with moderate-to-severe disease (Hanauer et al. 2002, 2006; Järnerot et al. 2005; Sandborn et al. 2007). Recently, an anti-IL-12/IL-23 antibody, ustekinumab, also showed promising efficacy in CD (Sandborn et al. 2008; Feagan et al. 2016). Collectively, these findings provide a strong scientific rationale for targeting IL-10 for the therapy of IBD.

Clinical Experience in Treating Inflammatory Bowel Disease with Recombinant IL-10

Recombinant IL-10 has been tested in the clinic for the treatment of IBD and other inflammatory diseases (Table 1). The first recombinant human (rhu)IL-10 (Tenovil) tested in clinical trials was produced in a genetically engineered Escherichia coli strain. This rhuIL-10 was tolerated in multiple dose toxicity studies in mice and monkey (Rosenblum et al. 2002). It was also tolerated up to 100 μg/kg in a single intravenous dose in healthy volunteers (Chernoff et al. 1995; Huhn et al. 1996). The pharmacokinetics of rhuIL-10 are similar to many other cytokines, with serum half-life ranging from 2.3 to 3.7 h (Huhn et al. 1996). A single dose of either intravenously or subcutaneously administered rhuIL-10 induced a transient increase in neutrophils and monocytes and a reduction in lymphocytes, especially at higher doses (Chernoff et al. 1995; Huhn et al. 1996, 1997). A modest decrease in circulating platelets was also observed (Huhn et al. 1997). Supporting the immunosuppressive functions of IL-10, the production of TNF-α and IL-1β was significantly reduced from lipopolysaccharide (LPS)-stimulated peripheral blood cells isolated from rhuIL-10-treated volunteers (Huhn et al. 1997). In addition, although rhuIL-10 increased the expression of FcγRI on monocytes and neutrophils, the expression of activation markers IL-2Rα and HLA-DR on T cells was inhibited (Dejaco et al. 2000).

Table 1.

Clinical trials targeting interleukin (IL)-10

Intervention	Indication	Clinical stage	Sponsor
Tenovil (rhuIL-10)	Crohn’s disease	Phase I/II N/A	Schering-Plough Research Institute
Dekavil (F8-IL-10)	Rheumatoid arthritis	Phase II NCT02076659	Philogen/Pfizer
		Phase II NCT02270632
Tenovil TM (IL-10)	Acute pancreatitis	Phase II NCT00040131 (terminated)	Merck Sharp & Dohme
IL-10	Psoriasis	Phase II NCT00001797	National Cancer Institute
Prevascar (rhuIL-10)	Cicatrix, wound healing	Phase II NCT00984646	Renovo
AG011 (engineered Lactococcus lactis secreting human IL-10)	Ulcerative colitis	Phase I/II NCT00729872	ActoGeniX N.V.
AM0010: PEGylated human IL-10	Solid tumors/pancreatic cancer	Phase I NCT02009449	ARMO BioSciences
		Phase III NCT02923921
BT063 (antibody to neutralize IL-10)	Systemic lupus erythematosus	Phase II NCT02554019	Biotest

Data source: clinicaltrials.gov.

rhu, Recombinant human.

Because rhuIL-10 was relatively well tolerated, it was tested in multiple clinical trials in IBD patients (Table 1). In the first reported trial, 46 patients with steroid refractory CD were treated with one of five doses of rhuIL-10 (0.5, 1, 5, 10, or 25 μg/kg) or placebo administered intravenously once daily for 7 days. A mild reduction in the Crohn’s Disease Activity Index (CDAI) and an increase in remissions within a 3-week follow-up period were observed in the rhuIL-10-treated groups (van Deventer et al. 1997). A clinical response was observed in a trial in patients with mild-to-moderate CD, who were treated with subcutaneous rhuIL-10 (1, 5, 10, or 20 μg/kg) or placebo once daily for 28 days (Fedorak et al. 2000). Interestingly, in this study, only patients treated with 5 μg/kg, but not higher doses, displayed clinical improvement. However, in a larger multicenter, double-blinded, placebo-controlled prospective study no difference in clinical remission, defined as a reduction in CDAI by more than 150 points, was observed between rhuIL-10 and placebo-treated groups (Schreiber et al. 2000). In this study, patients were subcutaneously dosed with rhuIL-10 (1, 4, 8, or 20 μg/kg) or placebo daily for 28 days, and were followed for an additional 4 weeks after treatment. A significant clinical improvement (reduction of CDAI by more than 100 points) was observed only in patients who received 8 μg/kg rhuIL-10, but not in patients who were treated with the higher dose of 20 μg/kg. In addition, patients with high disease activity responded better (Schreiber et al. 2000). Clinical response was accompanied by a decrease in inflammatory signals as measured by nuclear factor (NF)-κB activation in ileocolic biopsies. In both trials, only the intermediate, but not higher doses of rhuIL-10, induced clinical responses, suggesting a more complex biology of IL-10 in IBD. Finally, rhuIL-10 was also tested in the prevention of relapse after patients underwent a curative colon resection. In this study, both 4 and 8 μg/kg/d did not show significant benefit in comparison to the control-treated group (Colombel et al. 2001).

Although rhuIL-10 was generally tolerated in these studies, the major side effects included a decrease in hemoglobin and thrombocyte counts, which led to significantly more withdrawals in some of the rhuIL-10-treated groups (Buruiana et al. 2010). Adverse events, such as anemia, were observed in a dose-dependent manner. Patients receiving higher doses of rhuIL-10 (>4 μg/Kg) presented progressively and significantly decreased hemoglobin values and platelet counts (Tilg et al. 2002). These symptoms were reversible after discontinuation of therapy in all studies. The mechanism through which IL-10 induces anemia is not clear. The anemia was associated with an altered iron metabolism, as evidenced by increased serum ferritin and soluble transferrin receptor levels (Fedorak et al. 2000; Schreiber et al. 2000; Colombel et al. 2001). An earlier study investigating the effects of the anti-inflammatory cytokines IL-4 and IL-13 had shown that ferritin translation was enhanced by these cytokines in IFN-γ-treated activated macrophages (Weiss et al. 1997). IL-10 may act through a similar mechanism. A study investigating IL-10-induced thrombocytopenia in healthy adult volunteers suggested that IL-10 might affect platelet production indirectly through the inhibition of cytokines produced from monocytes and macrophages (Sosman et al. 2000).

As a general conclusion, systemic administration of rhuIL-10 did not result in significant clinical benefits compared with placebo groups in CD patients (Buruiana et al. 2010). The lack of efficacy may be because of an inability of IL-10 alone to sufficiently suppress all inflammatory mediators or caused by patient-to-patient variabilities in responding to exogenous IL-10. However, an important caveat in these studies is that the local IL-10 concentration in the intestine after systemic dosing may be too low to induce meaningful biological responses.

Therapeutic Potential for IL-10 in Other Autoimmune Diseases

TNF-α production is pathogenic in psoriasis and rheumatoid arthritis (RA), as drugs blocking the TNF-α pathway provide a clear therapeutic benefit in both diseases (Sfikakis 2010). Psoriasis is an inflammatory skin disease characterized by infiltration of epidermis and dermis with leukocytes that produce various inflammatory cytokines, such as TNF-α, IL-17, and IL-22, which stimulate keratinocyte proliferation and promote epidermal hyperplasia (Nestle et al. 2009). Antibodies blocking IL-12/IL-23, TNF-α, or IL-17 have shown good efficacy for the treatment of psoriasis (Kofoed et al. 2015). IL-10 might also provide therapeutic benefit in this disease, given its role in repressing proinflammatory cytokines. Indeed, in multiple clinical studies (Table 1), administration of recombinant IL-10 provided mild-to-moderate benefits for patients with psoriasis (Asadullah et al. 1998). However, further development of IL-10 as a therapy for psoriasis was hindered by its poor in vivo pharmacokinetics and far superior efficacy of other cytokine-blocking therapies.

IL-10 has a dual role in RA, a disease characterized by synovitis associated with bone and cartilage loss, production of rheumatoid factor and autoantibodies, and systemic inflammation. Although IL-10 represses pathogenic cytokines, such as IL-6 and TNF-α, it also stimulates B cells and drives autoantibody production. Recombinant IL-10 has proven to be relatively safe in clinical trials in human (Chernoff et al. 1995). A new modified IL-10 that can be targeted to the site of inflammation and has better in vivo pharmacokinetics did alleviate disease severity in preclinical RA models (Trachsel et al. 2007). Currently it is being tested in RA patients (Table 1) (Galeazzi et al. 2014).

A pathogenic function for IL-10 has been proposed for type 1 diabetes (T1D) and multiple sclerosis (MS) (Asadullah et al. 2003; Saxena et al. 2015). However, given rather inconsistent results in preclinical models (Wogensen et al. 1994; Cannella et al. 1996; Nagelkerken et al. 1997; Zheng et al. 1997), targeting IL-10 in MS or T1D has not been tested in the clinic.

Systemic lupus erythematosus (SLE) is another autoimmune disease in which IL-10 might have a pathogenic role, as it is a growth factor for human B cells, promotes antibody production, class switching, and plasma cell differentiation (Rousset et al. 1992, 1995). Elevated IL-10 levels have been reported in SLE patients (Koenig et al. 2012). In a small clinical trial, six SLE patients were treated with an IL-10 blocking antibody for 21 days and followed over 6 months (Llorente et al. 2000). Although an improvement of symptoms was observed that allowed a reduction in steroid use, the lack of a control arm demands caution in interpreting the data. A placebo-controlled phase II clinical trial evaluating a neutralized anti-IL-10 antibody in SLE is currently ongoing (Table 1). In all autoimmune disorders, however, the potential risk of developing IBD when blocking the IL-10 pathway needs further evaluation.

In conclusion, although there is strong scientific rationale supporting the use of recombinant IL-10 as a therapy for various autoimmune and inflammatory diseases, the limited clinical experience thus far has not revealed clear clinical benefit.

Strategies for Improved Local Exposure and Pharmacokinetic Properties of IL-10

There are a number of possible explanations for the lack of efficacy of recombinant IL-10 therapy in patients. First, IL-10 is a pleotropic cytokine with both immune-repressive and immune-stimulatory properties. For example, whereas IL-10 represses IL-12 production from macrophages and limits Th1 differentiation and IFN-γ production induced by IL-12, it also promotes proliferation and activation of CD8⁺ T cells and their IFN-γ production. Indeed, increased IFN-γ production has been observed in healthy volunteers and IBD patients following rhuIL-10 administration (Lauw et al. 2000; Tilg et al. 2002). By lowering the serum concentration of the administered IL-10 or by reducing its activity through protein engineering, it might be possible to overcome this problem because myeloid cells, the main therapeutic targets, have higher IL-10 receptor expression and are thus more sensitive to IL-10 (Moore et al. 2001). However, to enable this strategy, the poor pharmacokinetic properties of rhuIL-10 have to be improved to achieve sufficient coverage at lower doses. Several approaches, such as using PEGylated IL-10 or IL-10-Fc fusion proteins, are being pursued (Schwager et al. 2009; Naing et al. 2016). Another concern is that systemic delivery of rhuIL-10 will not result in sufficiently high local concentrations in the inflamed tissues. Several local delivery strategies, such as ectopic expression of IL-10 by commensal bacteria or IL-10/antibody fusion proteins, have been devised and are being tested (Table 1).

Engineered Bacteria for Intestinal Delivery of IL-10

To increase the concentration of IL-10 in the gut while minimizing systemic exposure, local delivery of IL-10 by engineered bacterial strains has been considered (Steidler et al. 2000, 2003; Miyoshi et al. 2004; Innocentin et al. 2009; del Carmen et al. 2011). In this strategy, the Il10 gene is inserted into the genome of Lactococcus lactis, a widely used dairy microbe, and intragastrically administered to mice. The local delivery of mouse IL-10 substantially reduces histology scores in both dextran sulfate sodium (DSS)-induced and spontaneous colitis in IL-10-deficient mice, at a significantly lower therapeutic dose (Steidler et al. 2000). This strategy provides several potential benefits. First, because oral delivery of IL-10 is not a viable option as cytokines are extremely sensitive to the acidic environment of the stomach, the engineered bacteria are able to produce recombinant IL-10-locally in the intestine (Steidler et al. 2000). To further minimize degradation of IL-10 in the gastrointestinal track, an L. lactis strain expressing fibronectin-binding protein A from Staphylococcus aureus was engineered to carry an IL-10-expression vector, which can efficiently deliver recombinant DNA to human epithelial cells, thereby triggering IL-10 expression in situ (Innocentin et al. 2009). IL-10 produced locally is effective in ameliorating gut inflammation in both DSS-induced and trinitrobenzene sulfonic acid (TNBS)-mediated IBD mouse models (del Carmen et al. 2014; Zurita-Turk et al. 2014). Second, this strategy minimizes systemic exposure and toxicity. For example, L. lactis is a nonpathogenic and noninvasive Gram⁺ bacterium that survives in the digestive tract of animals or humans (Drouault et al. 1999). To control potential transgene escape into the environment, a biologically contained L. lactis has been developed in which the thymidylate synthase gene (thyA) has been replaced by the human IL10 gene. This strain is unable to grow in the absence of thymidine or thymine (Steidler et al. 2003). Lastly, the expression of IL-10 in bacteria can be controlled through further engineering. For example, a xylose-inducible expression system has been used to limit IL-10 production by L. lactis to the mucosal epithelium. Anti-inflammatory effects of milk fermented with this bacteria were observed in an acute TNBS-induced IBD mouse model (Miyoshi et al. 2004; del Carmen et al. 2011). Another inducible system uses stress-inducible controlled expression (SICE) based on a stress-inducible promoter (pGroESL) that allows the production of murine IL-10 locally at mucosal surfaces (Benbouziane et al. 2013). Here, IL-10 is produced locally without the need for extrinsic inducers. IL-10 levels achieved through this approach result in significant protective effects as measured by gut permeability, immune responses, and gut function in a mouse IBD model (Martin et al. 2014).

In a first human trial with biologically contained L. lactis expressing human IL-10, patients with moderate-to-severe CD were dosed twice daily for 7 days (Table 1). The results showed that the approach was both safe and controllable. The genetically modified L. lactis could only be recovered from feces with the addition of thymidine. However, the study was conducted in a small patient population, and clinical outcomes revealed no statistically significant differences between those individuals who received the IL-10-expressing strain and those patients in the control group (Steidler et al. 2003). For this strategy to succeed, several parameters need to be optimized further. A sufficient amount of IL-10 has to be produced at the site of inflammation over a prolonged period of time with minimal variations across the patient population. IL-10 needs to remain intact and functional under the harsh conditions in the gastrointestinal tract, so that biologically meaningful amounts of IL-10 can cross the intestinal barrier and act on inflammatory leukocytes. The development of sensitive methods to assess the pharmacokinetics and pharmacodynamics of bacterial IL-10, both locally and systemically, will be key to the success of this approach in patients.

Antibody-Conjugated IL-10 for Improved Targeted Delivery and Pharmacokinetics

To improve the pharmacokinetic properties of IL-10 and to deliver it in a more targeted way to the site of inflammation, antibody-mediated delivery strategies have been conceived. Four antibodies, L19 (binding the extra-domain B of fibronectin), F8 (binding the extra-domain A of fibronectin), G11 (binding the extra-domain C of tenascin-C), and F16 (binding the extra-domain A1 of tenascin-C), which specifically recognize angiogenic but not normal tissue, were considered as fusion partners for IL-10 (Trachsel et al. 2007; Schwager et al. 2009). L19 selectively accumulates at sites of inflammation in a collagen-induced arthritis (CIA) mouse model (Trachsel et al. 2007). As a proof-of-concept, IL-10 as well as the proinflammatory cytokines IL-2 and TNF-α were fused to the carboxyl terminus of a single-chain Fv fragment of L19. As expected, the L19-IL-10 fusion protein was found to be enriched at the site of angiogenesis. L19-IL-2 and L19-TNF-α exacerbate disease in the CIA model, whereas L19-IL-10 significantly reduces disease severity (Trachsel et al. 2007). More importantly, the repression of inflammation by L19-IL-10 is more profound than the effect mediated by a systemically distributed fusion protein of a control antibody single chain fused to IL-10 (Trachsel et al. 2007). In a comparative study, the four antibodies were used to stain synovial tissues of RA patients (Schwager et al. 2009). F8 showed the strongest staining. The corresponding IL-10 immunocytokine, F8-IL-10 (Dekavil), also showed efficacy in the CIA model (Doll et al. 2013). The combination of a murine F8-mIL-10 and a TNF-blocking muTNFR-Fc fusion protein displayed superior efficacy compared with either agent alone (Doll et al. 2013). Interestingly, F8-IL-10 also showed efficacy in endometriosis in a syngeneic mouse model and in a chronic cardiac allograft model (Schwager et al. 2011; Franz et al. 2015).

F8-IL-10 was combined with methotrexate in a preclinical toxicity study in cynomolgus monkeys. In this study, 180 μg/kg (equivalent to 60 μg/kg IL-10) of F8-IL-10 was administered subcutaneously three times a week for 8 weeks (Schwager et al. 2009). No major toxicity findings, other than transient regenerative anemia, have been reported. In pharmacokinetic measurements, about 20 ng/ml of F8-IL-10 was detected in the serum 3 h after subcutaneous infection with undetectable levels after 24 h. In a phase Ib study, 24 RA patients were dosed subcutaneously with 6 μg/kg to 300 μg/kg F8-IL-10 in combination with methotrexate every week for 8 weeks. No major side effects have been reported, except for one case of progressive anemia in the 160-μg/kg dose group (Galeazzi et al. 2014). Among the 23 patients evaluated in this study, 15 patients achieved American College of Rheumatology (ACR)20 improvement, seven patients achieved ACR50, and three achieved ACR70 responses. Impressively, two patients experienced long-lasting remission (ACR70 maintained for more than 1 year after last dose) (Galeazzi et al. 2014). Placebo-controlled phase II clinical trials in RA patients are currently ongoing (Table 1).

TARGETING IL-10 AND IL-24 FOR THE TREATMENT OF CANCER

A Potential Role for IL-10 in Cancer

The recent success of therapies aimed at modulating the immune system to elicit antitumor immunity has completely changed the landscape of cancer therapies (Sharma and Allison 2015). Anti-PD1/PD-L1 therapies are now being considered as first-line treatment for major cancer types, such as non-small-cell lung cancer (NSCLC), at least in subsets of patients (Topalian et al. 2012, 2015; Borghaei et al. 2015; Motzer et al. 2015; Robert et al. 2015). However, many cancer patients have yet to benefit from novel cancer immunotherapies. To further unleash the immune system and promote antitumor immunity, targeting additional checkpoints, the indoleamine 2,3-dioxygenase (IDO) pathway and other strategies are being considered (Chen and Mellman 2017; Sharma et al. 2017). IL-10 has been detected in the tumor microenvironment of many cancer types, and has been correlated with poor prognosis (Nemunaitis et al. 2001; O’Garra et al. 2008; Mannino et al. 2015). Based on its strong immunosuppressive functions, especially in inhibiting IL-12 production and Th1 differentiation, IL-10 has been considered as a target for cancer immunotherapy. However, evidence for both tumor-promoting and tumor-repressing functions of IL-10 have been presented (Groux et al. 1998). On the one hand, IL-10 represses cytotoxic T-cell activation by down-regulating MHC expression on cancer cells and on professional APCs, thereby preventing the recognition of cancer cells by antigen-specific T cells (Adris et al. 1999; Steinbrink et al. 1999). IL-10 also inhibits IL-12 production from APCs, a cytokine that strongly promotes Th1 differentiation and cytotoxicity. On the other hand, high doses of IL-10 enhance the proliferation of CD8⁺ T cells and their cytotoxic activity (Groux et al. 1998; Mumm et al. 2011). Certain inflammatory cytokines and conditions may promote tissue damage and oncogenesis. For example, UC is associated with an increased risk of colon cancer (Sturlan et al. 2001; Huang et al. 2006). IL-10 may repress these inflammatory conditions, thus preventing subsequent oncogenesis. It will therefore be necessary to evaluate a potential therapeutic intervention by either inhibiting or promoting the IL-10 pathway on a case-by-case basis in specific cancer types and patient subpopulations.

The role of IL-10 in various cancer models has been examined since it was first discovered (Asadullah et al. 2003; Vicari and Trinchieri 2004). In an IL-2 promoter-driven IL-10 transgenic mouse model, elevated expression of IL-10 results in failure to control an immunogenic tumor, an effect that can be reversed by administering an anti-IL-10 antibody (Hagenbaugh et al. 1997). Similarly, IL-10-deficient mice are resistant to UV-induced skin carcinogenesis, a finding that is correlated with a potent Th1 response in these mice (Loser et al. 2007). In contrast, in a chemically induced skin carcinoma model, IL-10 deficiency results in significantly increased tumor burden and accelerated mortality (Mumm et al. 2011). In tumor-bearing mouse mammary tumor virus–polyoma middle T oncogene (MMTV-PyMT) mice, antibody blockade of IL-10RA combined with paclitacel chemotherapy provides significant therapeutic benefit (Ruffell et al. 2014). On the other hand, the administration of PEGylated murine IL-10 in both a Her2 transgenic mouse breast cancer model and in a squamous cell carcinoma model can suppress tumor growth, supporting the rationale for IL-10 therapy in these cancers (Mumm et al. 2011). IL-10 induces antigen-specific CD8⁺ T-cell responses and increases IFN-γ production from these cells. Furthermore, the antitumor activity of PEGylated IL-10 is dependent on IL-10 receptor expression on CD8⁺ T cells, further supporting the notion of these cells being direct IL-10 targets in this setting (Emmerich et al. 2012).

Clinical Experience with IL-10 Therapies in Cancer

Based on the potential antitumor effects of IL-10, a PEGylated human IL-10 (PEG-rhuIL-10) was engineered for clinical use (Naing et al. 2016). Whereas PEG-rhuIL-10 inhibited IFN-γ production in peripheral blood mononuclear cells (PBMCs) in vitro, it stimulated IFN-γ secretion from CD8⁺ T cells. PEG-rhuIL-10 also promoted perforin and granzyme B production from CD8⁺ T cells, supporting its role in enhancing cytotoxicity (Naing et al. 2016). This molecule has recently been tested in clinical trials in cancer patients (Table 1). The PEG-rhuIL-10 has a prolonged half-life and exposure compared with rhuIL-10. In this study, patients were treated with one of six different doses (1, 2.5, 5, 10, 20, and 40 μg/kg) of PEG-rhuIL-10. In addition to increased IFN-γ levels in the serum, elevated IL-18 and reduced TGF-β levels were observed in treated patients, supporting the immune stimulatory function of this molecule in vivo. Interestingly, some partial clinical responses have been observed in this small cohort of patients (Naing et al. 2016). A phase III clinical trial investigating PEG-rhuIL-10 in combination with FOLFOX in metastatic pancreatic cancer patients is currently ongoing (Table 1).

Antitumor Activity Elicited by Ectopic Expression of IL-24 in Cancer Cells

IL-24 has been ascribed unique antitumor activity when intrinsically expressed in tumor cells. Although the primary cellular sources of IL-24 include T cells and myeloid cells (Fig. 2), IL-24 was first discovered with elevated expression levels in terminally differentiated human melanoma cells treated with recombinant IFN-β and the protein kinase C (PKC) activator mezerein (Jiang et al. 1995). Enhanced IL-24 expression causes an irreversible growth arrest and suppression of tumorigenic properties. Similar results were obtained in a variety of human cancer cells, including lung, colorectal, breast, and prostate cancer (Emdad et al. 2009; Whitaker et al. 2012). Interestingly, IL-24 is also readily induced in some normal epithelial and fibroblast cells with insignificant effects on growth and survival (Jiang et al. 1996; Su et al. 1998). IL-24 was categorized as an IL-10 cytokine family member based on its conserved genomic structure, chromosome localization, and cytokine-like properties (Caudell et al. 2002; Sauane et al. 2003). It shares the same receptor complex with IL-20 (Rutz et al. 2014). However, no direct antitumor effects have been reported for any other family member. Only ectopic expression of IL-24 in tumor cells, but not its addition to the culture media, shows antitumor activity in cultured cancer cells (Whitaker et al. 2012). Several potential mechanisms for these surprising findings have been discussed, including endoplasmic reticulum stress, the unfolded protein response, apoptosis, autophagy, and the production of reactive oxygen species, seemingly independent of IL-24 receptor signaling (Whitaker et al. 2012). However, whether the antitumor activities of IL-24 are linked to its properties as a cytokine is still a matter of debate (Kreis et al. 2007, 2008).

The therapeutic potential of IL-24 for cancer treatment has been evaluated in clinical trials (Table 2). Efficacy was observed in a phase I/II clinical trial in patients with multiple advanced cancers following intratumoral injection of INGN 241, an adenovirus expressing IL-24 (Cunningham et al. 2005; Tong et al. 2005). Up to 80% of tumor cells at the injection sites showed positive terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, and 67% of tumors showed reduced Ki-67 staining postinjection. A transiently elevated level of CD8⁺ T cells and cytokines IL-6, IL-10, and TNF-α were noticed in peripheral blood, consistent with increased tumor apoptosis. The intratumoral injection of INGN 241 was well tolerated with some mild side effects, including pain and erythema at the injection sites. Only one of the patients suffered from grade 3 serious adverse events (SAEs) and discontinued the trial. It is unclear, however, whether the observed activity in this study was mediated solely by IL-24, or by the immune response elicited by the virus, or both.

Table 2.

Clinical trials targeting interleukin (IL)-20, IL-22, and IL-24

Intervention	Indication	Clinical stage	Sponsor
NNC109-0012: anti-IL-20	Rheumatoid arthritis	Phase I NCT00818064	Novo Nordisk A/S
		Phase I NCT01038674
		Phase II
		NCT01282255
		NCT01636817 (terminated)
		NCT01636843 (terminated)
	Psoriasis	Phase I
		NCT01261767 (terminated)
F-652:	Alcoholic hepatitis	Phase I/II NCT02655510	Mayo Clinic Generon (Shanghai)
IL-22 IgG2-Fc	Acute GVHD	Phase I/II NCT02406651
ILV-094: anti-IL22	HV	Phase I NCT00434746	Pfizer
	HV	Phase I NCT00447681
	Psoriasis	Phase I NCT00563524
	Atopic dermatitis	Phase II NCT01941537	Rockefeller University
	Rheumatoid arthritis	Phase II NCT00883896	Pfizer
ILV-095: anti-IL22	HV	Phase I NCT00822835	Pfizer
	HV	Phase I NCT00822484
	Psoriasis	Phase I NCT01010542 (terminated)
INGN 241 (Ad-mda-7)	Melanoma	Phase II NCT00116363	Introgen Therapeutics

Data source: clinicaltrials.gov.

GVHD, Graft-versus-host disease.

THE IL-20 SUBFAMILY OF CYTOKINES IN HUMAN DISEASE

IL-20 subfamily cytokines mainly facilitate the communication between leukocytes and epithelial cells (Fig. 2), where they function to enhance innate defense mechanisms, wound healing, and tissue repair at epithelial surfaces (Rutz et al. 2014). Although it is evident how these functions can be beneficial to the host during infections, a clear role in host defense has only been established for IL-22. IL-22 is indispensable during infections with extracellular pathogens such as Citrobacter rodentium, Klebsiella pneumonia, or yeast at mucosal surfaces (Ouyang et al. 2011; Rutz et al. 2013; Eidenschenk et al. 2014). Recent studies also reveal important functions of IL-22 in shaping intestinal homeostasis and commensal communities (Sonnenberg et al. 2011). However, IL-20 subfamily cytokines have been widely studied for their roles in inflammation and autoimmunity, most prominently in the context of skin inflammation, with well-defined pathogenic roles in psoriasis, atopic dermatitis (AD) (Sa et al. 2007; Ouyang 2010; Sabat et al. 2013), and RA. Exciting new approaches are aimed at harnessing the tissue-protective and wound-healing functions of IL-22 for the treatment of IBD and diabetic foot ulcer (DFU).

Role of IL-20 Subfamily Cytokines in Psoriasis

As briefly discussed above, psoriasis is a chronic inflammatory disease of the skin characterized by abnormal keratinocyte differentiation and proliferation, leukocyte infiltration into the dermis and epidermis, and increased dilation and growth of blood vessels. The cross talk between keratinocytes and leukocytes is essential during the pathogenesis of psoriasis. Various leukocytes, including T cells, neutrophils, DCs, and macrophages, infiltrate into the skin and contribute to inflammation. IL-19, IL-20, IL-22, and IL-24 are expressed in psoriatic but not in healthy skin (Rømer et al. 2003; Wolk et al. 2004; Otkjaer et al. 2005). More importantly, the receptors for IL-20 subfamily cytokines are highly expressed on keratinocytes (Sa et al. 2007). T cells appear to be the major source of IL-22 in psoriatic skin lesions. T cells isolated from skin lesions produce much higher levels of IL-22 than T cells in circulation (Boniface et al. 2007). In addition, T-cell clones generated from psoriatic tissue are largely CCR6⁺ IL-22⁺, presumably Th17 cells or Th22 cells (Pène et al. 2008; Kagami et al. 2010). Myeloid cells and keratinocytes themselves are potential sources of IL-19, IL-20, and IL-24 (Fig. 1) (Sa et al. 2007; Tohyama et al. 2009; Wolk et al. 2009b).

The role for IL-20 subfamily cytokines in psoriatic skin inflammation has been extensively studied in preclinical models in mice. Transgenic mice that ectopically express IL-20, IL-22, or IL-24 develop skin lesions and histological features similar to those seen in human psoriasis (Blumberg et al. 2001; Wolk et al. 2009a; He and Liang 2010). Direct injection of IL-23 induces ear thickening, acanthosis, and dermal infiltrates, similar to some features in psoriatic skin (Kopp et al. 2003; Chan et al. 2006; Zheng et al. 2007), which are dependent on IL-22 and other IL-20 family cytokines. In addition, injection of IL-22 into the skin of normal mice induces S100 and β-defensin, as well as keratinocyte hyperplasia (Ma et al. 2008), whereas neutralization of IL-22 or IL-23 prevents psoriasis-like symptoms induced by the transfer of CD4⁺CD45RB^hi CD25⁻ cells into severe combined immunodeficiency (SCID) mice (Ma et al. 2008). Furthermore, the treatment of immunodeficient mice engrafted with human psoriatic skin with blocking antibodies to IL-20 or IL-22 resolves the psoriasis condition (Stenderup et al. 2009; Perera et al. 2014).

IL-20 subfamily cytokines induce several factors that either promote or sustain psoriatic lesions. All IL-20 subfamily members, except IL-26, induce the expression of various antimicrobial peptides, including S100 family genes and β-defensin family genes (Boniface et al. 2005; Liang et al. 2006; Wolk et al. 2006; Sa et al. 2007). Furthermore, the IL-20 subfamily cytokines, especially IL-22, regulate genes in keratinocytes and fibroblasts involved in tissue remodeling, including proteases and extracellular matrix proteins, such as MMP1, MMP3, kallikrienes, marapsin, and platelet-derived growth factor (PDGF) (Boniface et al. 2005; Wolk and Sabat 2006; Sa et al. 2007; Li et al. 2009). IL-20 subfamily cytokines promote the re-epithelialization process by acting on epidermal keratinocytes, partly through the induction of keratinocyte growth factor (KGF) and epidermal growth factor (EGF). IL-20 subfamily cytokines also activate proinflammatory responses through the induction of chemokines and cytokines from keratinocytes, including CXCL1, CXCL5, and CXCL7 (Boniface et al. 2005; Sa et al. 2007). The effects of IL-20 subfamily cytokines can be further synergized by other proinflammatory cytokines, such as IL-17A, TNF-α, or IFN-γ (Liang et al. 2006; Tohyama et al. 2009; Guilloteau et al. 2010), suggesting that IL-20 subfamily cytokines in conjunction with other proinflammatory cytokines orchestrate the pathogenesis of psoriasis.

Clinical Experience with Targeting IL-20 Subfamily Cytokines in Psoriasis

Despite the convincing preclinical data suggesting a disease-promoting role for IL-20 subfamily cytokines in psoriasis, neutralizing antibodies targeting either IL-20 or IL-22 did not advance in clinical trials (Table 2). A randomized, placebo-controlled, phase I/IIa dose-escalation trial was conducted to evaluate a fully human IgG4 anti-IL-20 antibody in patients with moderate-to-severe stable chronic plaque psoriasis. However, the trial’s expansion phase was terminated early owing to an apparent lack of Psoriasis Area and Severity Index (PASI) improvement (Gottlieb et al. 2015). Similarly, early-stage clinical trials evaluating two anti-IL22 antibodies (fezakinumab, ILV-095) were ended, presumably because of lack of efficacy. Although the failure of these trials may be simply a result of redundant functions of IL-20 and IL-22, the clinical data suggest that IL-20 subfamily cytokines might not be the key engine driving the inflammatory cascade in psoriasis.

Targeting IL-22 for the Treatment of Atopic Dermatitis

Atopic dermatitis (AD) is the most common inflammatory skin disease, affecting up to 25% of children and up to 3% of the adult population (Williams 2005; Montes-Torres et al. 2015). AD is characterized by itchy, red, and flaky lesions that often occur on bending sides of the limbs. The lesions are infiltrated with immune cells in the dermis and epidermis. Acanthosis, fibrosis, and collagen deposition are observed during the chronic phase. IL-22 has been shown to contribute to the epidermal barrier dysfunction in AD, and also seems to be responsible for the characteristic epidermal hyperplasia (Nograles et al. 2009; Gittler et al. 2012). IL-22 is highly expressed in the affected skin of patients (Wolk et al. 2004). In contrast to psoriasis, the expression of IL-22 in AD is mainly derived from Th22 cells and from IL-22-producing CD8⁺ cells (Nograles et al. 2009). The role of IL-22 in the development and maintenance of AD needs further exploration. However, a phase II clinical trial with an anti-IL-22 antibody (ILV-094) in patients with moderate-to-severe AD is currently ongoing (Table 2). Data from this study is not yet available.

Role of IL-20 Subfamily Cytokines in Rheumatoid Arthritis

IL-20 subfamily cytokines have been studied extensively for their role in RA. All family members are up-regulated in synovial fluid of RA patients (Ikeuchi et al. 2005; Hsu et al. 2006; Kragstrup et al. 2008; Sakurai et al. 2008; Alanärä et al. 2010; Corvaisier et al. 2012). The corresponding receptors are expressed in synovial tissues (Ikeuchi et al. 2005; Kragstrup et al. 2008; Sakurai et al. 2008; Corvaisier et al. 2012). Increased frequencies of IL-22- and IL-26-producing Th17 cells and Th22 cells are found in peripheral blood and joints of RA patients (Pène et al. 2008; Shen et al. 2009; Leipe et al. 2011; Zhang et al. 2012). IL-19, IL-20, and IL-22 are thought to be pathogenic in RA owing to their ability to enhance the proliferation of synovial cells and to induce proinflammatory cytokines and chemokines, including IL-6, IL-8, and CCL2 (Ikeuchi et al. 2005; Sakurai et al. 2008). Additionally, IL-20 promotes neutrophil chemotaxis (Hsu et al. 2006).

Preclinical data further support a pathogenic role of IL-20 subfamily cytokines in arthritis. Blockade of IL-19 or the administration of a soluble IL-20RA attenuates disease in collagen-induced arthritis in rats (Hsu et al. 2006, 2012).

Based on these data, a neutralizing anti-IL-20 antibody has been evaluated in a phase II clinical study in patients with active RA (Table 2). Sixty-seven patients with RA received either anti-IL-20 (3 mg/kg per week, subcutaneously) or a placebo over 12 weeks with a 13-week follow-up. A significant proportion of patients with seropositive RA receiving anti-IL-20, compared with those receiving placebo, achieved treatment responses according to the American College of Rheumatology 20% (ACR20) (59% vs. 21%), ACR50 (48% vs. 14%), and ACR70 (35% vs. 0%) levels of improvement (Šenolt et al. 2015).

An elevated plasma concentration of IL-22 is associated with disease severity and progression of erosive RA in patients (Leipe et al. 2011; Zhang et al. 2011; da Rocha et al. 2012), suggesting that IL-22 blockade might be equally efficacious in the treatment of RA. However, preclinical arthritis models paint a more complicated picture. Blockade of IL-22 with neutralizing antibodies before disease onset increases the incidence and severity of collagen-induced arthritis in mice, whereas treatment after disease onset is beneficial (Justa et al. 2014). In a different model, anti-IL-22 reduces inflammation and bone erosion, but does not affect overall arthritis severity (Marijnissen et al. 2011).

Nonetheless, a neutralizing anti-IL-22 antibody (ILV-094) has been evaluated in a phase II clinical trial for the treatment of RA (Table 2), but the results of the study have not been published to date.

Therapeutic Potential for IL-22 in Inflammatory Bowel Disease

Genetic studies in IBD not only revealed alterations in pathways regulating inflammation, such as the IL-10 pathway, but also showed a strong link between defects in epithelial innate defense and integrity and disease onset (Xavier and Podolsky 2007; Kaser et al. 2010). The IL-22 receptor, but not the IL-20 receptor, is highly expressed in the gastrointestinal tract (Zheng et al. 2008; Wang et al. 2014), suggesting a predominant role for IL-22 in regulating intestinal epithelial cells during inflammation and host defense (Ouyang 2010). Although more pronounced in CD, UC patients also show increased numbers of IL-22-producing cells in affected tissues (Andoh et al. 2005; Yu et al. 2013). Sources of IL-22 in the intestine comprise Th cells and innate cells, such as Th17, Th22, ILCs, and NK22 cells (Andoh et al. 2005; Geremia et al. 2011). More importantly, changes in IL-22 expression occur during acute phases of disease. In UC patients, a significant reduction in IL-22⁺ CD4⁺ T cells has been observed in actively inflamed tissues compared with normal tissues from UC patients or from healthy controls (Leung et al. 2013). The reduction in IL-22 production might result from an increased expression of TGF-β, as TGF-β represses IL-22 production from T cells (Zheng et al. 2008; Rutz et al. 2011). Furthermore, expression of IL-22BP, acting as a natural IL-22 antagonist, has been detected in mucosal samples from IBD patients (Pelczar et al. 2016). Blockade of the TNF-α pathway can reduce IL-22BP production, suggesting that the enhanced IL-22 activity might be partially attributed to the efficacy of anti-TNF-α therapies in IBD. The beneficial effects of IL-22 in IBD were further supported by a case study of a UC patient (Broadhurst et al. 2010). The reported patient infected himself with the nematode Trichuris trichiura to treat his symptoms. The disease went into remission accompanied by an accumulation of IL-22-producing CD4⁺ T cells in the intestinal mucosa. Goblet cell hyperplasia and increased mucus production were also observed (Broadhurst et al. 2010).

IL-22 exerts its protective effects at mucosal surfaces (Fig. 3A) by increasing the expression of mucus-associated molecules and restoring goblet cells (Sugimoto et al. 2008). It contributes to intestinal immunity by promoting the formation of secondary lymphoid structures, such as colonic patches, in the gut (Ota et al. 2011). Furthermore, IL-22 induces the production of various antimicrobial peptides and enhances innate intestinal defense functions (Zheng et al. 2008; Zenewicz et al. 2013). Finally, IL-22 fosters the repair of epithelial barriers by promoting epithelial cell proliferation and the expansion of intestinal stem cells (Pickert et al. 2009; Huber et al. 2012; Lindemans et al. 2015).

Figure 3.

Therapeutic potential of interleukin (IL)-22-Fc fusion protein in inflammatory bowel disease and diabetic foot ulcer patients. (A) In inflammatory bowel disease, severe epithelium damage and uncontrolled microbiota result in chronic inflammation in the gut. Systemic administration of IL-22-Fc is thought to promote epithelial cell regeneration, stimulate the production of antimicrobial peptides to control the microbiota, enhance the mucin production to restore mucus layer, and boost chemokine secretion to recruit immune cells for host defense. As a consequence, such therapy may reduce gut inflammation and restore epithelial integrity. (B) In diabetic foot ulcer, the cutaneous wound-healing process is significantly impaired, and the wounded skin is often associated with bacterial infections. IL-22-Fc is thought to increase re-epithelialization through directly stimulating kerantinocytes or indirectly promoting epidermal growth factor (EGF) and keratinocyte growth factor (KGF) production, thereby augmenting antimicrobial peptide production, angiogenesis, chemokine secretion to recruit immune cells, and tissue remodeling, with accelerated wound repair. IEL, Intraepithelial lymphocytes; IL, interleukin; TNF, tumor necrosis factor; DCs, dendritic cells; MMPs, matrix metalloproteinases.

Studies in mice have shown a beneficial role for IL-22 in various colitis models. IL-22-deficient mice (Zenewicz et al. 2008) or mice receiving neutralizing anti-IL-22 antibodies (Sugimoto et al. 2008) show increased disease severity, as evidenced by extensive epithelial destruction and inflammation in the colon, increased weight loss, and delayed recovery in a DSS-induced acute colitis model. In a T-cell transfer model of colitis, the transfer of IL-22-deficient naïve CD4⁺ T cells cause more severe colitis than the transfer of wild-type CD4⁺ T cells (Zenewicz et al. 2008). In a C. rodentium–dependent colitis model, in which wild-type mice normally recover from the infection, IL-22-deficient mice do not (Zheng et al. 2008; Ota et al. 2011). Conversely, an increase in IL-22 levels either by administration of recombinant cytokine or hydrodynamic tail vein injection of IL-22-encoding plasmids ameliorates disease in various colitis models (Sugimoto et al. 2008; Zenewicz et al. 2008; Ota et al. 2011).

However, several potential caveats must be considered carefully before moving IL-22 forward as a treatment for IBD. Despite its tissue-protective role, IL-22 is a proinflammatory cytokine that can enhance inflammation, especially in synergy with other cytokines (Rutz et al. 2014). Indeed, a pathogenic role of IL-22 has been described in certain colitis models (Kamanaka et al. 2011). Furthermore, IL-22 targets many other organs such as skin, liver, and lung, and the downstream effects on these organs following systemic administration of IL-22 need to be considered (Park et al. 2011). Lastly, IL-22 activates STAT3 and promotes proliferation of many epithelial cell types. STAT3 is well known for its oncogenic activity in many epithelial tumors (Yu et al. 2014). Many cell types expressing IL-22R have been associated with neoplastic proliferation and IL-22-expressing cells have been detected in carcinomas of the colon (Jiang et al. 2013), stomach (Zhuang et al. 2012), liver (Jiang et al. 2013), and lung (Kobold et al. 2013). Increased colon tumorigenesis is observed when IL-22 is elevated in mice with colitis undergoing carcinogen treatment (Kirchberger et al. 2013) or in mice bearing a genetic alteration (Apc^Min/+) that causes colorectal cancer (Huber et al. 2012). However, studies in transgenic mouse models suggest that long-term ectopic expression of IL-22 itself is insufficient to cause cancer (Park et al. 2011; Wang et al. 2011). Given that chronic inflammation is a risk factor for colon cancer, the beneficial effect of IL-22 in reducing tissue damage and inflammation might actually decrease the risk of early oncogenesis (Huber et al. 2012).

In conclusion, strong genetic and mechanistic data support a novel strategy for treating IBD by enhancing barrier function and intestinal innate defense, which is an approach that is differentiated from current therapies that mainly rely on systemic anti-inflammatory agents, such as TNF-α blockers (Ouyang 2010). To modulate the activity of IL-22 and to prolong its half-life, we generated murine IL-22-Fc fusion proteins (Ota et al. 2011; Wang et al. 2014). These molecules have a longer half-life in vivo, are able to induce downstream effector functions similar to IL-22, and protect intestinal integrity in various models.

Tissue Protective and Regenerative Effects of IL-22 in Pancreatitis and Hepatitis

In addition to skin and intestinal epithelial cells, the IL-22 receptor is also expressed in liver, pancreas, lung, kidney, and many other cell types of epithelial origin (Ouyang et al. 2011; Rutz et al. 2014). Recent advances in IL-22 biology revealed profound effects of IL-22 in tissue protection and regeneration in many organs including liver, pancreas, thymus, and skin (Radaeva et al. 2004; Zenewicz et al. 2008; Dudakov et al. 2012; Feng et al. 2012b; Kolumam et al. 2017). Acinar cells in the pancreas have the highest expression level of IL-22 receptor in the body (Xie et al. 2000; Aggarwal et al. 2001; Huan et al. 2016). Pancreatitis, an inflammatory condition, can occur in acute or chronic forms (Braganza et al. 2011; Banks et al. 2013). Excessive alcohol consumption, gallstones, and autoimmune diseases are some of its leading causes. The current treatment for pancreatitis is supportive and no curative therapy is available. A protective role of IL-22 in murine pancreatitis models has been described, suggesting a therapeutic potential for IL-22-based therapies (Feng et al. 2012b; Huan et al. 2016). Elevated IL-22 levels protect mice from cerulein-induced acute pancreatitis, as evidenced by a reduction in serum digestive enzymes, apoptosis, and infiltration with inflammatory cells (Feng et al. 2012b). Increased IL-22 levels inhibit autophagosome formation through STAT3-dependent regulation of Bcl-2 and Bcl-XL (Feng et al. 2012b; Huan et al. 2016). Furthermore, IL-22 up-regulates the expression of Reg3 (Hill et al. 2013; Huan et al. 2016), which protects acinar cells against injury and inflammation. IL-22 has not yet been tested clinically for the treatment of pancreatitis.

Hepatitis is commonly associated with viral infections or excessive alcohol consumption. Patients with chronic hepatitis B virus (HBV) or hepatitis C virus (HCV) infection show increased numbers of IL-22-producing cells and corresponding elevated levels of IL-22 (Dambacher et al. 2008; Park et al. 2011). Administration of recombinant IL-22 protein (Radaeva et al. 2004), delivery of an IL-22-expression construct via hydrodynamic tail vein injection (Pan et al. 2004), or transgenic expression of IL-22 under a liver-specific albumin promoter (Park et al. 2011) protect mice from liver damage in hepatitis models, such as concanavalin A–induced T-cell-mediated hepatitis. In contrast, inhibition of IL-22 by neutralizing antibodies (Radaeva et al. 2004) or genetic ablation of IL-22 (Zenewicz et al. 2007) worsens liver damage in these models. In acute and chronic mouse models of alcohol-induced liver damage, treatment with IL-22 ameliorates alcoholic fatty liver, liver injury, and hepatic oxidative stress (Ki et al. 2010; Xing et al. 2011). The protective effects of IL-22 in the liver are thought to be mediated through STAT3 and the induction of antiapoptotic, mitogenic, and antioxidant pathways in hepatocytes and liver stem cells (Pan et al. 2004; Radaeva et al. 2004; Feng et al. 2012a). Based on these preclinical data, an IL-22-IgG2 fusion protein is currently being tested in the clinic for the treatment of alcoholic hepatitis (Table 2).

A Protective Role for IL-22 in Graft-Versus-Host Disease

Allogeneic hematopoietic stem-cell transplantation holds great promise for the treatment of malignant and nonmalignant hematologic diseases. The major obstacle that hinders successful transplantation is graft-versus-host disease (GVHD) (Ferrara et al. 2009), which is associated with severe morbidity and mortality (Wingard et al. 2011). Both innate and adaptive immune cells are thought to be involved in the pathophysiology of GVHD (Ferrara et al. 2009; Konya and Mjösberg 2015). Interestingly, an increased number of donor-derived ILCs in the blood correlates with reduced severity of GVHD in patients (Munneke et al. 2014). A recent study in mice reported a protective role of IL-22 in GVHD during transplantation of intestinal stem cells (Hanash et al. 2012). Intestinal stem cells and downstream progenitor cells express the IL-22 receptor (Hanash et al. 2012), which can promote epithelial regeneration through STAT3-dependent mechanisms in response to IL-22 secreted from various sources, including ILC3s (Hanash et al. 2012). Furthermore, administration of IL-22 promotes the recovery of intestinal stem cells and intestinal regeneration, and reduces mortality from GVHD in transplanted animals (Hanash et al. 2012). IL-22 deficiency in the recipient, on the other hand, increases tissue damage and mortality as the result of acute GVHD (Hanash et al. 2012). A clinic trial is currently ongoing to evaluate the therapeutic potential of IL-22 in GVHD (Table 2).

Potential Therapeutic Applications for IL-20 Subfamily Cytokines to Enhance Wound Healing in Diabetic Foot Ulcer

A regulated topical wound-healing response restores tissue homeostasis and immune defense mechanisms in the skin (Singer and Clark 1999; Li et al. 2007). The healing process can be divided into several overlapping phases: hemostasis, inflammation, angiogenesis, re-epithelialization, and remodeling or maturation. These phases involve many types of tissue cells, immune cells, and soluble factors, such as cytokines and growth factors. DFU is a chronic wound with impeded healing. As one of the major complications of type II diabetes, diabetic ulcers lead to a significant number of amputations, and increase the morbidity and mortality of diabetes patients (Jeffcoate and Harding 2003; Boulton et al. 2005). The impaired wound healing is in part because of deficiencies in leukocyte recruitment, macrophage function, angiogenesis, extracellular matrix deposition, epidermal barrier function, and fibroblast activities under diabetic conditions (Falanga 2005; Brem and Tomic-Canic 2007). Frequently, wound site infections in DFU patients are difficult to control and further hinder the wound-healing process.

Several IL-20 subfamily cytokines participate in the different stages of the wound-healing response (Fig. 3B) by regulating re-epithelialization, immune cell infiltration, production of growth factors, and tissue remodeling (Rutz et al. 2014). Mice ectopically expressing IL-20 subfamily cytokines develop skin pathogenesis that resembles human psoriasis, which can be viewed as an excessive wound-healing response (Blumberg et al. 2001; Wolk et al. 2009a; He and Liang 2010). All IL-20 subfamily cytokines are induced during normal wound healing in mouse models, and elevated IL-24 levels have been identified in biopsies of human wounded skin (Bosanquet et al. 2012; Kolumam et al. 2017). The receptors for IL-20 subfamily cytokines are expressed primarily on keratinocytes in the skin (Sa et al. 2007). Studies conducted in vitro and in vivo have shown several downstream pathways induced by IL-20 subfamily cytokines to contribute to different stages of wound healing (Wolk and Sabat 2006; Sa et al. 2007; Rutz et al. 2014). All IL-20 subfamily cytokines promote re-epithelialization by stimulating the proliferation of epidermal keratinocytes and by inducing EGF and KGF production from keratinocytes (Sa et al. 2007). They also trigger chemokine production from keratinoytes and thereby enhance inflammatory infiltration of leukocytes, especially macrophages. In addition, all IL-20 subfamily cytokines induce vascular endothelial growth factor (VEGF), an important growth factor that facilitates angiogenesis and neovascularization (Sa et al. 2007; Kolumam et al. 2017). Importantly, they induce the expression of various antimicrobial peptides from keratinocytes to dampen infections that are commonly associated with wounded skin. Finally, IL-20 subfamily cytokines induce the production of many proteases and extracellular matrix proteins from keratinocytes and fibroblasts to mediate scar formation and tissue remodeling. IL-22 is the most potent, followed by IL-24 and IL-20, with IL-19 being the least effective family member with regard to its wound-healing activity. Yet, various IL-20 subfamily members appear to have redundant functions (McGee et al. 2013; Kolumam et al. 2017).

In preclinical models, topical treatment of wounded skin with recombinant IL-19, IL-20, or Fc-fusion proteins of IL-20, IL-22, and IL-24 accelerates the cutaneous healing process in mice (Sun et al. 2013; Kolumam et al. 2017). The therapeutic potential of IL-22, IL-20, and IL-24 has also been studied in diabetic wound-healing models. In streptozotocin-induced type I diabetic mice, topical treatment with recombinant IL-22 accelerates wound closure compared with control-treated diabetic wounds (Avitabile et al. 2015). Re-epithelialization of wounds treated with IL-22 is significantly improved. Another study examined the therapeutic effect of IL-22R-binding cytokines in diabetic wound repair in db/db mice (Kolumam et al. 2017), a commonly used mouse model for type II diabetes, which shows much slower wound closure compared with other diabetes models (Michaels et al. 2007). Both systemic and topical application of Fc fusion versions of IL-20, IL-22, and IL-24 significantly accelerates wound closure in db/db mice, independent of their metabolic functions (Wang et al. 2014; Kolumam et al. 2017). Gene-expression analysis revealed that IL-22-Fc treatment specifically promotes the activity of critical pathways for re-epithelialization, tissue remodeling, host defense, and fatty acid/lipid metabolism compared with PDGF or VEGF treatment. Many genes that are up-regulated by IL-22-Fc have been reported to participate in cutaneous wound healing, such as Cxcr2, GrhI3, and Klk8 (Kolumam et al. 2017).

THE INTERFERON λ SUBFAMILY IL-28A, IL-28B, IL-29

The genes encoding IL-28A (IFN-λ2), IL-28B (IFN-λ3), and IL-29 (IFN-λ1) were identified based on their homology with genes of the IFN/IL-10 superfamily of cytokines (Kotenko et al. 2003; Sheppard et al. 2003; Fox et al. 2009). Structurally, these three cytokines are more closely related to IL-10 family cytokines, in particular to IL-22 (Gad et al. 2009). They signal through the IL-10R2 chain, also used by other IL-10 family cytokines, but they use a unique IL-28R as their receptor α chain (Kotenko et al. 2003; Sheppard et al. 2003). Despite differences in receptor usage, downstream signaling shows substantial overlap with IFN-α/β (Ouyang et al. 2011; Durbin et al. 2013). Accordingly, IFN-λ and type I IFNs trigger similar biological effects and antiviral responses (Onoguchi et al. 2007; Kotenko et al. 2003). However, substantial differences exist with regard to the magnitude and kinetics of IFN-sensitive gene (ISG) induction. Whereas IFN-α strongly induces ISGs with an early peak followed by rapid down-regulation, IFN-λ triggers a weak but sustained ISG response (Kotenko et al. 2003; Meager et al. 2005; Marcello et al. 2006). More importantly, in contrast to a broad range of cells that respond to type I IFNs, only a few cell types, in particular epithelial cells (Fig. 1), respond to type III IFNs (Sommereyns et al. 2008; Witte et al. 2009). This has led to the idea that IFN-λ and IL-20 subfamily members serve parallel functions in protecting epithelial tissue against viral and bacterial infections, respectively (Gad et al. 2009).

IFN-λ for the Treatment of Viral Infections

Generally speaking, IFN-λs induce relatively weak antiviral responses. However, GWAS identified a role for polymorphisms near the IFNL3 gene in the control of HCV infection, as well as other viruses with epithelial cell tropism. IFN-λs are in fact the dominant IFN subclass produced in the liver of humans, chimpanzees, and in primary human hepatocyte cultures infected with HCV (Park et al. 2012; Thomas et al. 2012). A genetic polymorphism linked to the gene for IL-28B has been associated with clinical response to IFN-α and ribavirin therapy in patients with chronic HCV infection (Ge et al. 2009; Suppiah et al. 2009; Tanaka et al. 2009; Thomas et al. 2009), suggesting cooperation between the IFN-λ and type I IFN responses during host defense. A critical function for IFN-λ has been reported for a variety of preclinical models of pulmonary infections in mice, including influenza A, influenza B, severe acute respiratory syndrome (SARS), coronavirus, and H1N1 influenza virus (Mordstein et al. 2008, 2010). Furthermore, IFN-λs have the ability to inhibit the replication of multiple other viruses, including human immunodeficiency virus (HIV) (Hou et al. 2009), herpes simplex virus type 2 (HSV2) (Ank et al. 2006), cytomegalovirus (CMV) (Brand et al. 2005), and encephalomyocarditis virus (EMCV) (Hou et al. 2009).

IFN-λ is an attractive therapeutic target for the treatment of viral infections. The more restricted expression of IFN-λ receptors predominantly on epithelial cells suggests that IFN-λ, although sharing the same therapeutic advantages, might avoid many of the systemic side effects of IFN-α/β. Furthermore, there is genetic evidence that certain IFNL3/IFNL4 polymorphisms are linked to the clearance of HCV infection and potentially other types of infections (Miller et al. 2009). Type I and II IFNs are currently approved for the treatment of HCV and HBV (Gibbert et al. 2013; Lin and Young 2014).

A PEGylated IFN-λ1 has been evaluated in clinical trials for the treatment of HCV (Table 3). A phase II trial showed that IFN-λ1 was as effective as PEGylated IFN-α with significantly fewer extrahepatic adverse events, such as neutropenia, thrombocytopenia, flu-like symptoms, autoimmune thyroid disease, and pulmonary arterial hypertension (Muir et al. 2014; Fredlund et al. 2015). Several phase III clinical trials are currently ongoing (Friborg et al. 2013).

Table 3.

Clinical trials targeting interleukin (IL)-29

Intervention	Indication	Clinical stage	Sponsor
PEG-rIL-29	Hepatitis C	Phase I NCT00565539 Phase II NCT01001754	ZymoGenetics
PEG-rIL-29	Hepatitis C	Phase III NCT01718158 Phase III NCT01598090 Phase III NCT01754974 Phase III NCT01866930 Phase III NCT01616524	Bristol-Myers Squibb
PEG-rIL-29	Hepatitis D	Phase II NCT02765802	Eiger BioPharmaceuticals

Data source: clinicaltrials.gov.

However, the availability of highly efficacious direct-acting antiviral agents (DAAs) makes it highly unlikely for IFN-λ to play a major role for HCV therapy in the future. Although efficacy will have to be shown in the clinic, IFN-λ could potentially hold promise for the treatment of select infections with epithelial tropism, such as intestinal infection caused by rotaviruses or noroviruses, and respiratory infections caused by viruses, such as influenza virus or coronaviruses.

Role for IFN-λ in Cancer

IFN-λ is induced in the tumor microenvironment, either through viral infections or other mechanisms, and shows antitumor activity (Steen and Gamero 2010; Burkart et al. 2013; Cannella et al. 2014). Depending on receptor expression, IFN-λ induces apoptosis in various murine and human cancer cells. Responsive tumors include neuroendocrine tumors, colorectal/intestinal carcinoma, hepatocellular carcinoma, esophageal carcinoma, lung adenocarcinomas, Burkitt lymphoma, and melanoma (Sato et al. 2006; Zitzmann et al. 2006; Guenterberg et al. 2010; Li et al. 2010; Steen and Gamero 2010). In a mouse model of breast cancer, for example, IFN-λ expression on mammary epithelial cells inversely correlates with tumor growth (Burkart et al. 2013). IFN-λs induce apoptosis in colorectal cancer cells more potently than IFN-α/β or IFN-γ (Li et al. 2008), and show comparable potency in a mouse hepatoma model (Abushahba et al. 2010). Interestingly, the combination of IFN-λ and IFN-α showed more substantial antitumor activity than either IFN alone (Lasfar et al. 2016). Like other IFNs, IFN-λs possess immune stimulatory activities and enhance antitumor immunity through various mechanisms (Numasaki et al. 2007; Li et al. 2010; Steen and Gamero 2010). Accordingly, antitumor activity has also been observed in models in which the tumor cells themselves are not responsive to IFN-λ (Sato et al. 2006). To date, IFN-λ has not been studied clinically as a cancer therapy.

CONCLUDING REMARKS

Over the past three decades, a number of cytokines have been targeted successfully for the treatment of a wide spectrum of human disorders (O’Shea et al. 2014). In fact, since the approval of recombinant IFN-α2b for the treatment of chronic HBV infection, therapies targeting members of most cytokine families, including IL-1, IL-2 γc, IL-6, IL-12, IL-17, and TNF-α, have been developed for clinical use in autoimmune disorders, infectious diseases, and cancer. However, despite their known essential immune regulatory functions in host defense and tissue protection, the IL-10 family of cytokines is currently an exception with no approved therapies for any indication. As we have discussed, part of the reason is that we are still in the early stages of understanding the pleiotropic functions of many cytokines in this family. The degree of functional redundancy within the IL-20 subfamily further complicates targeting these cytokines. However, given the recent advances in our understanding of the diverse biology of IL-10 family cytokines, combined with technical progress in protein engineering, and innovative targeting and delivery strategies, we are increasingly able to modulate these cytokine pathways in a cell-type- and disease-specific manner. Novel therapies targeting IL-10 family cytokines that will improve the lives of patients may therefore be within reach.