The cytokines of the IL-1 superfamily are proteins of about 17-18 kDa, with a typical β-trefoil pyramidal barrel structure composed by six two-stranded β hairpins, a structure common to many proteins with different functions throughout evolution, from bacteria to mammals, in particular for carbohydrate recognition (ricin-type lectins) and as toxins (30–34). In humans (and most mammals), eleven IL-1-like cytokines have been described, some with inflammatory activity and some with anti-inflammatory functions. From evolutionary analysis (see below), the ancestral IL-1 family encompasses IL-1β (the prototypical IL-1 cytokine) and other eight cytokines derived from it (IL-1α, IL-1Ra, IL-36Ra, IL-36α, IL-36β, IL-36γ, IL-37 and IL-38) (35, 36). IL-1β, IL-1α and the three IL-36 isoforms have largely overlapping inflammatory activities, with IL-1β and IL-1α binding to the same receptor (35, 36). However, IL-1α is a moonlighting protein, with a number of diverse functions attributed to its long intracellular form, which differ from the inflammatory functions of the extracellular shorter cytokine (37, 38). The three IL-36 isoforms (which all bind to the same receptor) have a range of tissue-specific inflammatory activities and are differentially present in different tissues (39, 40). Other members of the family have antagonistic or inhibitory activity: IL-1Ra is a receptor antagonist that binds in an inactive fashion to the same receptor as IL-1β and IL-1α, thereby antagonizing them (41); IL-36Ra is a receptor antagonist for the three IL-36 isoforms (34); IL-37 is an anti-inflammatory cytokine with a wide range of targets, possibly binding to the IL-18 receptor and to the inhibitory receptor IL-1R8 (35, 42–44); IL-38 was also reported to have anti-inflammatory activity (although its functions are not fully clarified) (45–47).

In addition to the ancestral family, the broader IL-1 superfamily also includes IL-18 and IL-33, which are not evolutionarily related, but have substantial structural and functional similarities with IL-1 (48). Both cytokines have multiple activities that are not strictly inflammatory and that are largely involved in immune metabolism. IL-18 is present in circulation at measurable levels (implying that it is not directly inflammatory), it induces the production of IFN-γ in T and NK cells in concert with IL-12, and it contributes to the differentiation of Th1 cells (49, 50). In its extracellular form, IL-33 is a potent activator of Th2 cells, as well as mast cells, basophils and eosinophils, thereby contributing to type 2 inflammation (51–53). In addition, intracellular IL-33 localises to the nucleus and presumably binds to DNA, although with unknown effects on gene expression, and it was reported able to interact with NFκB to decrease NFκB-dependent inflammatory gene transcription (51–53).

The receptors that bind the IL-1 superfamily cytokines are also structurally very similar, and are ten transmembrane proteins, which display an extracellular portion encompassing three Ig-like domains and a cytoplasmic portion that contains a Toll-IL-1R (TIR) signaling domain. The IL-1R complex that allows for cell activation is typically composed by a ligand-binding chain, which engages with the cytokine, and an accessory chain that associates with the dimeric ligand/receptor complex. The approximation of the two homologous intracellular TIR domains is the event that initiates signaling and eventual cell activation (54). The prototypical IL-1R is IL-1R1, the ligand-binding receptor for IL-1β and IL-1α. The accessory chain for IL-1R1 is IL-1R3, a promiscuous accessory chain that participates to several IL-1R complexes. Another ligand binding chain for IL-1β and IL-1α is the IL-1R2, which acts as a decoy receptor because it lacks the intracellular TIR domain and therefore can sequester ligands and IL-1R3 into an inactive receptor complex. The receptor antagonist IL-1Ra can inhibit binding of the active ligands IL-1β and IL-1α to IL-1R1. Other receptor complexes are similarly composed. The receptor for IL-36 encompasses the ligand binding chain IL-1R6 and the accessory chain IL-1R3, while the IL-33 receptor complex is composed by IL-1R4 and IL-1R3. The receptor complex for IL-18 includes the ligand binding chain IL-1R5 and a unique accessory protein, IL-1R7.

Other receptor-like chains included in the family have peculiar characteristics. The orphan receptors IL-1R9 and IL-1R10 are mainly expressed in the brain and do not have known ligands (54), although it has been proposed that IL-38 may bind IL-1R9 (46). The unique receptor IL-1R8 (also known as SIGIRR) is characterized by the presence of a single Ig-like domain in its extracellular portion and has the role of inhibiting receptor-mediated activation. Proposed mechanisms include interference with receptor complexes both at the extracellular and intracellular level (in this case active also in inhibiting TLR signaling), and acting as co-receptor for inhibitory ligands (hypothesized in the case of IL-37 bound to IL-1R5) (43). In teleost fish, another truncated receptor has been described, DIGIRR, which displays two Ig-like extracellular domains. DIGIRR is alternative to SIGIRR (same chromosomal location) and has similar anti-inflammatory capacity as SIGIRR, leading to hypothesize that it is a transition from between the classical receptors with 3 Ig domains and the single Ig domain SIGIRR (55).

A list of the human IL-1 and IL-1R molecules and genes, with some of the different names they are known with, is reported in Table 1 , while Table 2 summarizes the main known ligand-receptor interactions.

The evolutionary history of IL-1 and IL-1R offers very interesting implications and can help us understanding more about IL-1 functions. As mentioned above, the structure of IL-1 superfamily cytokines is quite common (β-trefoil barrel), as it can be found in molecules with different recognition and defensive functions across evolution, from bacteria and fungi (e.g., pore-forming toxins and carbohydrate binding proteins) to plants (many lectins and toxins, Kunitz protease inhibitors), invertebrates (chitin-binding proteins and other lectins) and vertebrates (R-type lectins and domains in mannose and other receptors) (30–34). IL-1-like activities were extensively described in invertebrates (echinoderms, tunicates, insects) (56–66). However, no molecular characterization of invertebrate IL-1-like molecule(s) is at present available, and no gene orthologues were found in any clade besides vertebrates, based on the currently available genomes. This implies that IL-1 cytokines in vertebrates have taken functions that in invertebrates are performed by different molecules (58, 59, 63, 66). From evolutionary and chromosomal analysis, the ancestral proto-IL-1β gene appeared about 420 million years ago, i.e., coincidental with the emergence of the vertebrate subphylum (48), encoding for a protein that adopted the β-trefoil barrel structure common to many invertebrate defensive molecules.

From chromosomal location and surrounding gene positioning, three separate IL-1 families can be identified, i.e., IL-1, IL-18 and IL-33 (48). While the IL-1 and IL-18 families are present in all vertebrate species, IL-33 only appeared in mammals. From the proto-IL-1β gene are derived all other cytokines of the ancestral IL-1 family, which are orthologues generated by gene duplication. Of them, only IL-1Ra is present in all vertebrates (independently evolved in fish, derived from proto-IL-1β gene duplication in other vertebrates), whereas IL-1α, the IL-36 group, IL-37 and IL-38 appeared much more recently (between 320 and 160 million years ago), as they are only present in mammals (48), similar to IL-33 ( Figure 1 ). Why all the cytokines of the ancestral IL-1 family only appeared in mammals is a matter of speculation, although they have likely developed for performing specific defensive/immunostimulating functions linked to the peculiar characteristics of mammalian physiology and anatomy (large brain, fetal development, lactation, hair-covered skin, complex anatomy).

IL-1R-like genes are present in invertebrates and are predicted to encode proteins with a typical organisation encompassing an extracellular Ig-like domain, a transmembrane region and an intracellular signalling domain. Several IL-1R-like orthologous genes, with homologies to the vertebrate genes, have been identified in Cnidaria, Protostomia and Deuterostomia, implying important yet unidentified functions (67–69). In addition, a significant homology can be observed between the intracellular portion of IL-1R and the Toll/Toll-like receptors (TLR), the most represented innate/developmental receptors in invertebrates, in particular in the signalling domain that has been therefore dubbed TIR (from Toll-IL-1R) (70, 71). Thus, IL-1R molecules are hybrid transmembrane receptors with the same signalling domain as TLR but with a distinct extracellular portion displaying Ig-like domains. While Ig domains are present in many invertebrate proteins, their inclusion into Ig-like receptors has expanded from pre-vertebrate chordata (68, 69, 72) to vertebrates giving rise to a plethora of receptors with high interaction capacity and specificity [including antigen-recognising T and B cell receptors; (73-75)]. In this perspective, the presence of Ig-like domains in the binding portion of IL-1R molecules is predictive of high binding specificity, at variance with the broad pattern-recognition capacity of TLR and other innate receptors.

The sequence similarity and the chromosomal anatomy of IL-1R genes in vertebrates suggest that most IL-1R molecules belong to a single family derived from duplication of an ancestral IL-1R gene. These are the two IL-1 receptor genes IL1R1 and IL1R2, the IL-33 receptor gene IL1RL1, the IL-18 binding and accessory chain genes IL18R1 and IL18AP, the IL-36 receptor gene IL1RL2 and, by further duplication from one of the IL-1R genes, IL1RAPL2 and (from it) IL1RAPL1, which is only present in birds and mammals (48). The only two IL-1R genes that do not seem to belong to the IL-1R family are the orphan receptor gene SIGIRR and the IL-1R3 gene IL1RAP, possibly derived from IL18RAP (48). SIGIRR shares low sequence homology and no chromosomal anatomy with the other genes, suggesting no common ancestry. As mentioned above, SIGIRR is likely the result of a gradual loss of extracellular Ig domains from a distinct IL-1R-like ancestral gene. IL-1R molecules are present in all vertebrates, except IL1RAPL1, which is absent in fish and only appeared in Tetrapoda about 365 million years ago ( Figure 1 ). Given the presence of the intracellular TIR domain, the signalling pathways of IL-1R and TLR are similar, involving an overlapping range of downstream signalling and regulatory molecules, and leading to activation of overlapping immune defensive responses.

Notably, the appearance of receptors preceded that of ligands. This is obvious in invertebrates, in which IL-1R are present in the absence of IL-1-like molecules, but also in vertebrates, as in the case of IL-1R4 for IL-33 and IL-1R6 for the IL-36 group. The notion that receptors were present long before their cytokine ligands underlines their different original functions and their re-use in mammals for additional, IL-1-mediated effects. That IL-1R have appeared for functions that do not require IL-1 ligand binding is supported by the case of the orphan receptors IL-1R10 and IL-1R9, which do not have known ligands and that have non-immune functions in the brain (54).

Despite the common notion that IL-1 is an innate/inflammatory cytokine, the observation that IL-1 appeared in vertebrates, coincidental with the development of adaptive immunity, leads to hypothesize that its functions are not exclusively linked to innate immunity. The co-existence of innate and adaptive immunity in vertebrates likely led to a progressive “simplification” of the innate immune defensive mechanisms, which are particularly complex in several invertebrates (67). For instance, TLR and NLR genes decreased from over 200 in sea urchin to 10-20 in human beings. In this context, IL-1 family cytokines may have developed for taking a double role, to maintain a functional innate immunity that could also amplify the adaptive immune responses.

IL-1 family cytokines are considered innate immune factors because their production does not depend on cognate signalling and their effects are likewise non-specific. In the case of the prototypical cytokine IL-1β, its production is inducible, i.e., the cytokine is produced upon stimulation, and there is no constitutive production in quiescent conditions, implying that its main role is in the reaction to triggering agents, typically infective microorganisms. Since the IL-1 receptors are receptors that signal through a TLR-like pathway, being TLR the main innate defensive receptors in invertebrate immunity, it is reasonable to say that IL-1β is a factor involved in immune defense, and we can therefore suggest that its main physiological role is inducing/participating to the innate immune protection from invasion and damage. However, the fact that IL-1 family cytokines are only present in vertebrates further suggests that their defensive activity goes beyond innate immunity and that they have a role in helping/amplifying adaptive immune responses.

IL-1β can non-specifically amplify T and B lymphocyte responses that, conversely, are triggered by specific antigenic signals. Thus, IL-1β can induce IL-2 secretion and expression of IL-2 receptors on T cells, induce expansion and commitment of naïve and memory CD4+ T cells (in particular Th17) in response to antigen and enhance their functions (76–87), thereby indirectly increasing B cell activation, and it can also increase CD8+ T cell activation and cytocidal capacity (88–90).

Likewise, the IL-1 superfamily member IL-18 is a potent activator of Th1 responses and participates to Th1 differentiation and activation, typically by inducing the production of IFN-γ in concert with IL-12 (49, 80, 91). IL-18 is constitutively produced by many cell types, and mainly released by monocytes and macrophages, and is present at measurable level in the circulation of healthy people, suggesting a homeostatic or surveillance role (91).

The other, most recent IL-1 superfamily member IL-33, present only in mammals, is an inducer of Th2 differentiation and amplification, thereby contributing to the specific defensive reaction against extracellular parasites (e.g., helmints), bacteria and toxins (and to the overall type 2 inflammatory reaction) (51, 80), as well as to the successful establishment of pregnancy (92, 93).

IL-1α is very similar in its effects as IL-1β in its mature extracellular form, including effects on T cell activation. This raises the question why, after duplication from IL-1β, IL-1α has divergently evolved and has been conserved in mammals. The most likely hypothesis is that IL-1α possesses different functions, non-redundant with those of IL-1β. Indeed, IL-1α is not a typical inflammatory cytokine, as its precursor protein (pro-IL-1α) is constitutively expressed in macrophages and many non-immune stromal cells (epithelial cells, endothelial cells, fibroblasts) in resting conditions, and it localises in the nucleus where it acts as transcription factor thereby regulating cellular functions, proliferation, senescence and apoptosis (37, 38, 94, 95). Upon inflammatory stimulation, pro-IL-1α can be cleaved and, while the C-terminal portion is exported outside the cell and acts as a cytokine (binding to the same receptor as IL-1β), the pro-piece can still act as transcription factor and upregulate the expression of inflammation-related genes (38, 95–98). Thus, the moonlighting nature of IL-1α, which has different functions in different conditions and in different compartments, is probably among the reasons why it has been conserved throughout mammalian evolution.

IL-33 shares with IL-1α its dual functions, both as a soluble IL-1-like cytokine active on Th2 cells and as an intracellular transcription factor constitutively present in several stromal cell types (37, 51). At variance with IL-1α, however, the nuclear IL-33 represses gene expression, as it does facilitate chromatin compaction (99), suggesting that IL-33 can act as a strong down-regulator of inflammatory responses and contribute to the re-establishment of tissue homeostasis during the resolution of a defensive reaction. Also, as mentioned above, the fact that IL-33 is only present in mammals suggests its major role in controlling inflammatory reactions during pregnancy, thereby facilitating embryo implantation and development (92, 93).

Based on the above notions, we may consider the IL-1 superfamily as a hybrid cytokine family, which has developed for amplifying and controlling adaptive immune responses in vertebrates but that is non-specific in its induction and effects, as it simultaneously displays typical innate/inflammation-related activities. The IL-1R characteristics contribute to the mixed activities of the IL-1 family cytokines, since the receptors display specific ligand recognition and binding (with their Ig-like extracellular domains) and are differentially expressed on the surface of distinct lymphocyte populations (thereby specifically targeting the IL-1 effects to them), but are typical innate receptors in their downstream TIR-dependent effects.

However, the evolutionary data suggest a more complex scenario. Many of the IL-1R molecules appeared long before their IL-1 ligands, suggesting that these molecules had/have an original constitutive activity, independent of ligands. In mammals, this hypothesis is supported by some data showing ligand-independent receptor functions, as for instance in the case of IL-1R7 (100) and IL-1R9 (101, 102) in the brain. In the case of the IL-33 receptor IL-1R4, based on structural data, it has been proposed that its activation is brought about by interaction with the accessory chain IL-1R3 in a ligand-independent way, and that the function of IL-33 is merely that of stabilizing IL-1R4 in a conformation that promotes interaction with IL-1R3 (103, 104). We can hypothesize that the IL-1 ligands developed for downregulating and controlling the constitutive functions of such receptors, a strategy adopted by pathogenic microorganisms for inhibiting host receptors with antimicrobial capacity. These ligands would be inverse agonists, i.e., agonists that dampen the activity of constitutively active receptors. Thus, the inverse agonist molecules may have appeared in evolution before the agonist ligands and may be derived from microbial genes/molecules (which, as already mentioned, share the same β-trefoil fold structure as IL-1 family molecules). In the case of microbial inverse agonists, the agonist ligands are the countermeasures created by the host organism to antagonise the microbial inverse agonists, thereby neutralising the pathogen’s escape strategy. Similarly, the IL-1 family agonists may have developed from the inverse agonists, based on modifications of the same structure, to re-activate, control and fine-tune biological functions that are down-regulated by an inverse agonist ( Figure 2 ). IL-1Ra and IL-1β appeared around the same time, i.e., about 420 million years ago with the emergence of vertebrates, underlining their mutual relationship, although we cannot know whether IL-1 emerged to counteract IL-1Ra or vice-versa. In mammals, IL-1R1 does not seem to have ligand-independent functions (which may have been lost during evolution), and IL-1β acts as agonist ligand, whereas IL-1Ra has mainly developed as antagonist ligand. However, it is notable that binding of IL-1β to IL-1R1 does not induce receptor activation, which can only be achieved after the engagement of a second non-ligand binding accessory chain. Thus, IL-1Ra does not counteract IL-1β by dumping IL-1β-induced receptor activation but rather by preventing interaction of the receptor with the accessory chain ( Figure 2 ).

IL-1 family molecules can exert a number of activities that are apparently independent of receptor engagement, as it has been reported in particular for the most ancient cytokines IL-1β and IL-1Ra. In ischemic brain damage, IL-1β can exacerbate damage also in the absence of IL-1R1, whereas IL-1Ra could only antagonise IL-1β in the presence of IL-1R1 (105). Likewise, IL-1β could modulate expression of a number of genes in primary microglial cells independently of IL-1R1, while other genes were only modulated depending on the receptor’s presence (106). In the same system, IL-1α only showed receptor-dependent effects (106). Most interestingly, IL-1Ra was observed having the same agonist activity as IL-1β in the hippocampus, with the IL-1Ra effects being IL-1R1-independent and those of IL-1β IL-1R1-dependent (107). The lack of correlation between receptor binding and activity in IL-1β is further supported by data on the IL-1R1-independent immunostimulatory capacity of the β-bulge surface loop between the 4^th and the 5^th β-strand in the IL-1β structure, and by the use of monoclonal antibodies and IL-1β mutants that showed a dissociation between IL-1R1 binding and different functional activities (108–112).

Notably, reminiscent of the lectin-like activity of β-trefoil proteins, IL-1β displays a lectin-like domain, distinct from the IL-1R1 binding sites, which can enable binding to specific carbohydrates (e.g., heparin, hyaluronic acid, GM4), thereby promoting interaction with glycoproteins on cell membrane and in the extracellular matrix (113–118). This capacity of interacting with glycoproteins not only can promote the tissue-specific localisation of the cytokine, as initially proposed, but it can actually underlie a more complex and more controlled activation mode, entailing the concomitant or differential engagement of receptor and glycoproteins for activation/inhibition of immune functions (114, 115, 119). IL-1α also has a carbohydrate-binding domain, which however is different from that of IL-1β, implying a distinct interaction mode and distinct activity regulation (114, 115, 118, 120), while IL-1Ra does not have such domain (118).

All these observations support the hypothesis that IL-1β may have developed at first as a receptor-independent functional molecule, and that it has been subsequently “re-used” to counteract IL-1Ra effects, but still maintaining both receptor-dependent and independent functions within a complex interplay of regulatory cross-interactions.