Rationale for IL-37 as a novel therapeutic agent in inflammation

KEYWORDS:

• Acute inflammation
• anti-inflammatory treatment
• chronic inflammation
• cytokine
• disease mechanisms
• immunology
• interleukin 37
• novel therapeutic
• translational research

1. Introduction

An estimated 5–8% of the world’s population is afflicted by severe inflammatory diseases that are among the leading causes of death, particularly in women younger than 65 years in case of autoimmune etiologies. Therefore, the unmet medical need is urgent – and interleukin (IL-) 37 holds substantial promise as a new therapeutic in this field.

IL-37 is an IL-1 family cytokine with powerful anti-inflammatory functions that govern principal pathways of innate and adaptive immunity and inflammation [1,2], including signaling by pattern recognition receptors such as Toll-like receptors (TLRs), cytokines such as IL-1 and tumor necrosis factor (TNF), and type 1-, 2-, and 3-polarized adaptive immune responses. Such broad actions are rare in cytokine biology; only IL-10 is known to have comparable scope and potency. Due to these powerful functions, IL-37 has aptly been called a ‘peacemaker’ in a Faculty of 1000 review.

Of the five IL37 mRNA isoforms a-e,IL37c and e lack exon 4, which encodes the first three β-strands of the IL-37 protein [3]. IL-37b, comprising exons 1, 2 and 4–6 is the most abundantly expressed isoform in blood cells [4], and by far the most studied. Extracellular IL-37 attenuates inflammation at picomolar concentrations [3] utilizing its receptor IL-1R8:IL-18Rα [5]. Data on expression of this heterodimer, particularly at protein level, are sparse.

In addition, IL-37 employs an intracellular pathway involving the transcriptional modulator Smad3 [1], and via both mechanisms works intracellular switches to dampen inflammation, including transcription factors such as NF-κB and AP-1, second messengers such as STATs and MAP-kinases, and the metabolic checkpoint kinases mTOR and AMPK [1,5,6]. By inhibiting mTOR and augmenting AMPK, IL-37 induces a pseudo-starvational state, which constitutes an additional anti-inflammatory mechanism aimed at achieving and maintaining metabolic homeostasis. Moreover, IL-37 suppresses IL-1β and IL-18 processing by the NLRP3 and AIM2 inflammasomes [7]. Notably, many of the mediators reined in by IL-37 are themselves targets of inhibitory drugs in clinical use (e.g. IL-17 in psoriasis; TNF and mTOR in autoimmune disease). Although there is no murine homologue of IL-37, recombinant (rec) and transgenically expressed human IL-37 effectively protect mice from various inflammatory diseases [1,3,8–10]. Given the clinical focus on several targets downstream of IL-37, its compelling characteristics such as multifaceted mechanisms of action and its success in preclinical models, IL-37-based drugs hold great potential.

2. Generating IL-37-based drugs

As a powerful roadblock of inflammatory responses, IL-37 function must be tightly regulated to avoid excessive suppression of innate and adaptive immunity. For example, in mice transgenic for IL-37 (IL-37tg), only little IL-37 protein is detectable under steady-state conditions despite the transgene’s constitutively active CMV promoter; abundance only increases upon inflammatory challenge [1,5,8]. In addition to an mRNA instability element, such regulation is achieved by homodimerization. Cytokine multimerization is common, but in most cases enables signaling. In contrast, the bioactivity of the head-to-head homodimer that is readily assembled by IL-37 molecules even at low concentrations is substantially lower than that of IL-37 monomers [3,11]. This rare structure–function relationship can be exploited by mutating tyrosine in position 85 to alanine, thus preventing dimerization and generating recIL-37 with considerably enhanced bioactivity [3]. To further improve drug-like properties, this IL-37 variant can be fused with an Fc-protein [12].

3. IL-37 in disease

3.1. Interpreting the findings

Since the discovery of its function and its renaming in 2010 [1], the number of reports on IL-37 has skyrocketed from less than 10 to more than 550 today. Notably, many of these publications focus on measuring IL-37 in various settings, while surprisingly few elaborate on mechanisms of action or directly explore therapeutic potential in vivo (Table 1). Interpreting changes in serum or tissue abundance is not straightforward for an anti-inflammatory mediator; in fact, we are often asked what to make of correlations between IL-37 and activity of a disease. Isn’t it counterintuitive to suggest treating an illness with IL-37 when its abundance increases with worsening of that illness?

Table 1. Data on IL-37 in diseases.

	Disease	In vivo evidence	MoA	Abundance in humans
Gastrointestinal	Inflammatory bowel disease	IL-37tg mice protected from DSS colitis [8].	Inhibition of innate and adaptive responses [8].	Elevated in inflamed intestinal mucosa (adult and pediatric) [18].
	Necrotizing enterocolitis	IL-37tg and recIL-37 protective in murine NEC model [17].	Amelioration of innate (e.g. lower IL-1β and IL-36) and adaptive inflammation (e.g. lower IL-17); restoration of microbial homeostasis [17].	Deficiency of IL-37 and IL-1R8 in affected intestinal mucosa and circulating monocyte subsets, but not lamina propria [17].
Musculo-skeletal	Rheumatoid arthritis	RecIL-37 suppressed joint and systemic inflammation in murine arthritis [13].	Rec IL-37 reduced TNF, IL-6, and IL-17 in RA-patient PBMCs [19], and suppressed Th17 cell proliferation, but not Th17 differentiation [15].	IL-37 increased in affected synovial and blood cells [1]; correlation of circulating IL-37 with bone loss and disease activity [14].
	Gout	RecIL-37 ameliorated urate crystal-induced joint inflammation in mice [20].	Inhibition of synovial IL-1β, IL-6, and cell influx in mice; reduction of ROS and IL-8 in human blood cells [20].	Genetic variants in IL37 increase susceptibility to gout [20].
Multi-organ	Systemic lupus erythematosus	IL-37-expressing MSCs reduced murine SLE severity and improved survival [21].	Rec IL-37 reduced TNF, IL-1β, IL-6, and IL-17, but not IL-10, in SLE-patient blood cells [15].	Conflicting evidence: Increase in serum IL-37 and correlation with disease activity in one study [22], but not another [16].
	CRS, e.g. in Covid-19, sepsis, CAR-T and conventional chemotherapy	RecIL-37 attenuated lung inflammation and tissue injury in murine SARS-CoV-2 infection [23]. IL-37tg mice protected in models mimicking aspects of sepsis [1,10], including endotoxic shock [3].	IL-37 ameliorates inflammation driven, e.g. by IL-1 and NLRP3 [7], both important e.g. in Covid [24], lowers cytokines that contribute to CRS such as IFNγ, IL-6, and IL-8 [1].	Increased serum IL-37 early in Covid correlated with reduced pro-inflammatory cytokines and improved clinical outcomes, such as faster viral clearance and chest CT improvement [23]. No interference with host’s anti-viral type I interferon responses [23].
Lung	Asthma	Intranasal recIL-37 attenuated airway inflammation and remodeling in murine ovalbumin model and restored Th1/2 balance [9,25].	Ratios of IL-1β and IL-33 to IL-37 skewed in favor of the pro-inflammatory and type 2-polarizing cytokines in asthma patients [25].	IL-37 decreased in patients’ PBMCs and sputum, which was inversely correlated with disease severity [9,25].
IRI	Stroke, myocardial infarction, acute kidney injury	IL-37tg mice protected from stroke-associated inflammatory brain injury, motor impairment and lung infection [26]. RecIL-37 protected mice from myocardial [27] and renal [28] IRI.	RecIL-37 ameliorated hypoxia-induced pro-inflammatory cytokine responses [28].	IL-37 increased in the plasma of patients with ischemic stroke [26].
Cardiovascular and metabo-lism	For example atherosclerosis, diabetes mellitus, inflammation-related fatigue	IL-37 expression or treatment ameliorated murine cardiovascular inflammation, e.g. including prevention of atherosclerosis in diabetic mice [29]. RecIL-37 increased exercise performance via IL-1R8 and AMPK [30], as well as obesity-related inflammation and insulin resistance [31].	IL-37 skewed metabolism toward pseudo-starvation [5,6], favorably altered murine oxLDL-induced DC maturation [29], increased muscle AMPK and oxidative phosphorylation [30], and reduced insulin resistance and cytokine release from adipose tissue [31].	IL-37 increased in human atherosclerotic plaques compared to healthy anterior tibial arteries [29].

MoA, mechanism of action; CRS, cytokine release syndrome; IRI, ischemia-reperfusion injury; ROS, reactive oxygen species; MSCs, mesenchymal stem cells; oxLDL, oxidized low-density lipoprotein.

To justify clinical trials, which provide top-level evidence, relevance of a mediator to disease pathogenesis can be demonstrated using preclinical in vivo blocking experiments (second-level evidence). Evidence that can be considered third-level includes ex vivo-blockade in cultured cells as in [1], or treatment with a mediator (e.g. recIL-37 [5,13] or via transgenic expression [1,8]). The absence of IL-37 in mice renders in vivo blockade challenging; thus, most available mechanistic and translational evidence on this cytokine falls in the last category. But what if not even this third level of evidence is available, and measurements of abundance is all we have?

Inflammation is vital to combat pathogens; but to regain and maintain healthy homeostasis, subsequent activation of counter-regulatory mechanisms is usually essential. Such activation can be compartmentalized, hence cytokine abundance may differ between compartments within the same individual. To combat the myriad of pathogens, specialized responses have evolved, which in turn require specific pathways and mediators for counter-balance. Besides IL-37, examples for such mediators include IL-1 receptor antagonist, soluble TNF-receptors, IL-10, and IL-18-binding protein. Therefore, it is likely that disease-associated local and/or systemic increases in IL-37 [14,15] are part of a targeted attempt to curtail a specific type of inflammation; however, this attempt is insufficient, at least at the time the disease was investigated. Albeit always speculatively, one might infer that augmenting IL-37 in this scenario could be useful.

The absence of change unfortunately commonly remains unreported (an exception being [16]), which is disappointing because such absence is likely reasonable evidence that a mediator is not involved in the disease process, at least at the time and location of sampling. Local or systemic decreases in IL-37 abundance, occurring for example in necrotizing enterocolitis [17], also require cautious, context-dependent interpretation, as they may indicate targeted dampening of its balancing activities, or a deficiency in IL-37 or the pathways involved in its activation.

3.2. Data in diseases

An overview of data on IL-37 in selected diseases is provided in Table 1.

3.3. Immuno-oncology (IO) – activating or neutralizing IL-37?

The immunosuppressive functions of IL-37 are not dissimilar from those of PD-1 and CTLA-4, both checkpoint inhibitors whose successful targeting has revolutionized medical care of hard-to-treat neoplasms. It is therefore not far-fetched to hypothesize that targeting IL-37 could be an attractive IO-strategy; and both augmenting or blocking its functions represent conceivable approaches in the vast field of cancer, in which every disease type and often even stage represents its own challenge with specific immunological circumstances that require tightly controlled manipulation. Therefore, a lot more research is required before translational decisions can be made.

4. Conclusions and perspectives

We argue that there is strong and robust evidence to support efforts aimed at taking the next translational steps toward deploying the peacemaking capacities of IL-37 in patients suffering from inflammatory or autoimmune disease. The unmet need in these fields is urgent, as existing medications not only have poor efficacy but also cause severe adverse effects and long-term morbidity that rival the damage done by the disease itself. A substantial number of high-quality publications support the conclusion that IL-37 represents an attractive novel treatment option. Stand-alone or combination strategies can be considered.

Having said this, one key question remains: Which clinical scenarios should be prioritized in such translational efforts? Considering the available evidence from the balcony perspective, we observe that the strongest data supporting therapeutic benefits of IL-37 arise from models of acute disease. Examples comprise studies on NEC (3 days), endotoxaemia (24 h), gout (4 h), stroke (1–3 days) or pulmonary inflammation in Covid (5 days), whereas data on longer-term models are sparse. The acute treatment concept is also supported physiologically, i.e. by the multiple auto-regulatory mechanisms designed to quickly shut down IL-37 function, and by the fact that IL-37 is an alarmin (a mediator that is stored intracellularly for example in monocytes, and can be released rapidly upon inflammatory stimulation of these cells [4]). In other words, the immune system quickly activates the IL-37 pathway when it is needed to curtail acute inflammation, then terminates its activities just as quickly. In contrast to models of such acute inflammation, there is a paucity of evidence for IL-37 benefit in models of chronic disease (exceptions including asthma, 12 weeks, and murine intestinal inflammation, 7 days to ~2 months). This discrepancy may be due to a number of reasons other than IL-37 lacking activity in chronic disease (for instance barriers to publication of chronic models, such as greater logistical challenges, cost, and biological variability), and we consider continuing the preclinical exploration of IL-37 in chronic disease – and IO, for that matter – well justified. However, it appears sensible to prioritize moving forward with the first clinical trials designed to deliver the immense therapeutic potential of IL-37 to patients in acute inflammatory disease, including acute flares of chronic illnesses.

Declaration of interest

C A. Nold-Petry and M F. Nold hold three patent families on IL-37, namely PCT/AU2016/050495, PCT/AU2017/051443 (Monash University, Hudson Institute) and PCT/EP2020/087031 (Monash University, Hudson Institute, F. Hoffmann-La Roche). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This work was supported by the National Health and Medical Research Council Investigator Grant Leadership 1 APP1173584 to C A Nold-Petry, and by the Fielding Foundation’s Innovation Award and Fellowship to M F Nold.