Introduction
Obesity is recognised as a medical condition characterised by the accumulation of excess body fat and is associated with a chronic state of low-grade inflammation. Low-grade inflammation in the adipose tissue (AT) involving both the adaptive and the innate immune system is reflected by a cytokine-induced acute-phase response, including elevated levels of TNF-α, IL-6 and C-reactive protein as part of an immune response of both innate and adaptive origin. A key role in the development of metabolic abnormalities has been assigned to M1-polarised macrophages that accumulate in the AT together with other immune cells [
1‐
3].
A role for dendritic cells (DCs) in promoting macrophage infiltration to AT in obesity has been suggested [
4], and an increased frequency of dysfunctional DCs in obese mice and humans supports this notion [
4‐
6]. However, a full understanding of the role of DCs in this process is still lacking.
The Toll-like receptor (TLR)-4 (expressed on the surface of macrophages, DCs and other immune and non-immune cells) has been reported to respond to the increased levels of endogenous lipids found in obese individuals [
7]. In line with this, lipid-lowering agents acting through a TLR-4-mediated mechanism can reduce the production of or the responsiveness to IFN-β and reduce the proinflammatory actions of AT macrophages [
8]. IFN-β is a member of the type I IFN family of pleiotropic cytokines that is critical in the defence against viral infections and able to modulate both innate and adaptive immunity [
9]. IFN-α/β can exert their effects by binding to their cognate receptor, the type I IFN receptor (IFNAR) complex, which stimulates the Janus kinase (JAK) signal transducer and activator of transcription (STAT) signalling pathway, leading to the transcription of several IFN-stimulated genes [
10]. While antibody-mediated neutralisation of IFN-β leads to a decrease in the mRNA expression of proinflammatory genes, including monocyte chemoattractant protein (MCP)-1, inducible nitric oxide synthase and IL-6, recombinant IFN-β protein has the opposite effect [
11]. In addition, IFN-β has been reported to be necessary for maintaining the TLR-mediated MCP-1 production in macrophages [
12], facilitating the recruitment of macrophages to sites of inflammation [
13] and modulating the inflammatory and metabolic effects of diet-induced obesity (DIO) [
14]. Thus, IFN-β induces a feedback activation mechanism, thereby contributing to macrophage-mediated inflammatory responses.
Type I IFNs are constitutively expressed in low quantities by many tissues and cells of the body in the absence of viral infection [
15], supporting a potential physiological role of IFN-α/β as the initial primers of immune function as a means of maintaining immune homeostasis. A role has been proposed for type I IFN in the activation of resident macrophages and the recruitment of proinflammatory M1 macrophages to the AT and liver during obesity.
Plasmacytoid DCs (pDCs) are a unique immune cell population involved in both innate and adaptive immunity. They sense single-stranded RNA and microbial DNA through endosomal TLR-7 and TLR-9, respectively, initiating a myeloid differentiation protein 88 (MyD88)-dependent signalling cascade [
16], leading to downstream activation of IRF7 and robust type I IFN responses. The pDCs are responsible for the vast majority of secreted type I IFN that promotes DC maturation, natural killer cell-mediated cytotoxicity and Th1 differentiation, and plays a crucial role in protection against viral and bacterial threats [
16]. The induction and subsequent nuclear relocation of the transcription factor IRF-7 is essential for type I IFN expression in pDCs [
17]. In comparison, DCs and macrophages from IRF-7 knockout mice were still able to produce normal levels of IFN-β [
18].
Studies have shown that IRF-7 deficiency prevents DIO and insulin resistance in mice [
19], and that IRF-7 expression is upregulated in the arteries of obese rats [
20]. In addition, murine gene expression analyses have reported significant upregulation of IFN-α/β genes in both the visceral and subcutaneous AT during obesity [
21], suggesting a role for pDC-derived type I IFN in obesity development. In support of this, elevated frequencies of pDCs were observed in the liver and AT of obese mice [
4] and humans [
22].
Here, we investigated whether DIO and the associated metabolic abnormalities are dependent on type I IFN signalling. We also analysed pDC, the major type I IFN-producing cellular subset, activity accumulated in the liver and AT during DIO and the effects of induced pDC deficiency on the development of obesity and metabolic abnormalities. We hypothesised that pDCs, through their IFN-producing capacity, not only play significant roles in obesity-induced low-grade inflammation and type 2 diabetes development but also in the early stages of obesity development.
Discussion
In the current study, we present evidence that type I IFN plays a critical role in the development of obesity and diabetes in the DIO mouse model of type 2 diabetes. Using IFNAR
−/− mice, we demonstrated that the absence of type I IFN signalling protected them from the development of DIO and diabetes. Since pDCs, a major source of type I IFN [
16,
28], have been reported to accumulate in the AT and liver of obese mice [
4], we reasoned that this cellular subset could constitute a key component in the type I IFN-mediated infiltration of proinflammatory M1 macrophages during obesity development. In agreement with this hypothesis, we demonstrated that pDC-deficient E2-2.cre
+ mice displayed a similar resistance to developing DIO and insulin resistance as observed in IFNAR
−/− mice.
Type I IFNs are prominent in viral infections, where they promote host defence mechanisms. However, a role of the pleiotropic cytokine family in several biological processes has also been proposed [
8,
11]. Type I IFNs have previously been indirectly implicated in obesity development based on the analysis of the IRF-7
−/− mouse [
19]. IRF-7 plays an important role in innate immunity through the regulation of IFN-α/β secretion. These IFNs can further activate cells through interactions with IFNAR, thereby inducing additional IRF-7 expression, which is required for the type I IFN positive feedback loop. A study by Wang et al [
19] suggests that improved energy expenditure protects IRF-7
−/− mice from developing DIO. In agreement with their findings, our data suggest that the protection from weight gain in the absence of type I IFN signalling is a consequence of altered immune–metabolism interplay, resulting in increased energy metabolism. This is in line with the notion that the immune system may play an important role in maintaining energy balance.
The pDCs are professional type I IFN producers responsible for the vast majority of IFN-α/β secretion [
29]. While pDCs are most well-known for their antiviral activity and ability to secrete vast amounts of IFN-I in response to TLR recognition of double-stranded RNA and CpG motifs, these cells have also been implicated in several disease conditions associated with systemic inflammation, including autoimmune diabetes [
30,
31]. An abundance of nucleic acid ligands and a critical role of TLR-9 in obesity-associated inflammation has also been reported previously both in animal models and humans [
22,
32‐
35]. The finding that IFNAR
−/− mice are protected from DIO highlights a possible role of pDCs in obesity-associated inflammation. While the pDCs have previously been found at an elevated frequency in the AT and liver of obese mice [
4], human studies have found decreased [
36‐
38] or unaltered [
30,
39] numbers of circulating pDCs in obese individuals with type 2 diabetes compared with lean individuals but have been unable to investigate the possible relocation of these cells. The data obtained from the analysis of the mouse model of DIO reported here and previously by Stefanovic-Racic et al [
4] supports the notion of a recruitment and activation of pDCs in obese VAT possibly, as has been previously suggested, mediated by the adipokine chemerin [
22,
40]. However, in light of conflicting results [
6], further studies are needed in order to elucidate this possibility.
The accumulation of pDCs in the liver during obesity development may suggest a role of this subset in this organ. Similar to the AT, the accumulation of pDCs in the liver was correlated with an increase in the number of proinflammatory macrophages. Even at steady state, pDCs are highly abundant in the liver, accounting for one-third of the DC population. While the precise reason for this accumulation is unknown, it is likely to be part of the requirement for the optimal defence mechanisms of the liver, as it encounters an array of substances through processed blood from the gastrointestinal tract. This includes gut-derived commensal bacterial products [
41,
42], of which the composition is affected by what is ingested. As pDCs are known to play an important role in the defence against invading microbes, it is plausible that even a swift change in the nutrient composition could induce their TLR-mediated production of proinflammatory cytokines [
43]. Indeed, studies have shown that the gut microbiome rapidly responds to dietary changes in both humans [
44] and mice [
45,
46], which has been proposed to contribute to obesity.
To directly address the pDCs as possible mediators of obesity development through their type I IFN-producing ability, we used a mouse with a conditional knockout of E2-2, the E2-2.cre
+ mouse, resulting in a specific block in the development of pDCs [
23]. We found that these mice were protected from DIO development, comparable to the IFNAR
−/− mice. In addition, the mice were also protected from developing insulin resistance and showed mildly improved glucose tolerance. Taken together, our results suggest a key role for pDC-derived type I IFN in obesity development. This concurs with, and gives genetic evidence for, a model recently presented by Ghosh et al [
22] suggesting that low-grade inflammation associated with obesity is driven by type 1 IFN produced by pDCs recruited to VAT and leading to proinflammatory polarisation of adipose-resident macrophages.
Acknowledgements
Thanks to B. Johansson-Lindbom and M. Dahlgren (Lund University, Sweden) for providing the IFNAR−/− mice Thanks to U. Axling (Lund University) for assisting in the use of the body composition scanner, and Å. Larefalk (Umeå University, Umeå, Sweden), A. Deronic and A.-C. Selberg (Lund University) for help with the DIO experiments.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.