Cross-talk between iron homeostasis and intestinal inflammation

Abstract

Recent publications from my laboratory have highlighted the important influence of altered iron homeostasis on the inflammatory response to intestinal bacteria. Here, I provide commentary on one of those papers, “Selective modulation of TLR4-activated inflammatory responses by altered iron homeostasis in mice”, which was published in the Journal of Clinical Investigation in November, 2009. It describes experiments that point to a previously unappreciated role for intracellular iron in the regulation of Toll-like receptor 4 signaling, and also demonstrates the potential therapeutic application of this information in a novel anti-inflammatory strategy based on manipulating iron balance. Our findings indicate that further investigation of the cross-talk between iron homeostasis and inflammation will yield new insights into the pathogenesis of chronic inflammatory diseases and may suggest new treatment approaches for these conditions.

Iron plays important roles in a number of biologic processes, including the response to microbial infection. My laboratory has been investigating the effects of altered iron homeostasis on innate immunity, prompted in part by the association of several disorders of iron metabolism with susceptibility to infection.^1–3 To address this issue experimentally, we have been studying the responses of Hfe knock-out mice, a model of human type I (HFE-associated) hereditary hemochromatosis.^4,5 This iron overload condition is caused by mutations in the HFE gene, which encodes an atypical class I MHC protein expressed on the surface of hepatocytes.⁵ The normal function of the HFE protein is to respond to elevated serum iron levels by upregulating expression of hepcidin, a peptide secreted by the liver that binds to the enterocyte and macrophage iron exporter ferroportin (FPN) and induces its lysosomal degradation. Hepcidin-mediated FPN downregulation inhibits absorption of dietary iron in the duodenum as well as the release of iron recycled from effete erythrocytes destroyed by the phagocytes of the reticuloendothelial system. Hepcidin expression is sensitive to a number of exogenous cues. In addition to the HFE-dependent increase in expression when circulating iron is high, hepcidin is downregulated when iron levels are low or when the requirement for iron is increased. Hepcidin is also upregulated in response to inflammatory cytokines such as IL-6, a response that is thought to have protective value by decreasing iron availability to infectious pathogens. Thus, by modulating FPN-mediated release of iron into the circulation in response to systemic iron levels and requirements as well as other signals, hepcidin functions as a central regulator of iron homeostasis (Fig. 1).⁶ In the absence of functional HFE, hepcidin levels are abnormally low, and the consequent increase in FPN expression leads to excessive release of iron from macrophages and duodoenal enterocytes, elevated serum iron and pathologic deposition of the metal in the liver, pancreas, myocardium and other tissues.⁴ Individuals with HFE-associated hemochromatosis are prone to disseminated infection with several enteric bacterial pathogens, including Vibrio, Yersinia and Salmonella, but the exact basis for this susceptibility is not well understood.^2,7

Figure 1.

Hepcidin-mediated regulation of systemic iron homeostasis. Hepcidin controls entry of iron into the circulation by modulating expression of FPN on phagocytes that recycle iron from aged erythrocytes and on duodenal enterocytes that absorb dietary iron from the intestinal lumen. Hepcidin expression in the liver is regulated by systemic iron levels and requirements, and by inflammatory signals. Green arrows indicate activating responses, red lines designate inhibitory effects, and black arrows indicate FPN-mediated iron efflux.

When we compared the responses of wild-type and Hfe-deficient mice to oral infection with virulent Salmonella typhimurium, the most striking result was the attenuation of intestinal inflammation in the mutant animals, an abnormality that was associated with elevated tissue pathogen burden.⁸ We investigated the basis for the weak inflammatory response to Salmonella since it was a likely contributor to the poor control of bacterial replication in the mutant animals. Our studies revealed that peritoneal macrophages isolated from the Hfe-deficient mice produced significantly reduced amounts of the pro-inflammatory cytokines TNFα and IL-6 in response to stimulation with either Salmonella or LPS. Subsequent experiments showed that the abnormality of cytokine biosynthesis was at the level of mRNA translation and that it was probably caused by the low intracellular levels of free iron in the Hfe knock-out macrophages. These observations provided a mechanistic explanation for earlier reports of decreased LPS-induced TNFα production by monocytes from patients with hemochromatosis,⁹ and also indicated that iron had a hitherto unrecognized role in the regulation of cytokine mRNA translation. They suggested further that impaired innate immunity could be an important factor in the susceptibility to infection in disorders of iron metabolism such as hemochromatosis.

In a follow-up paper that was recently published in the Journal of Clinical Investigation, we delved deeper into the mechanism underlying the abnormal inflammatory response to Salmonella in Hfe deficient mice.¹⁰ We first established that the impaired cytokine expression observed in the Hfe knock-out macrophages was not a cell-autonomous phenomenon. Rather, it depended on the low hepcidin environment from which the cells were taken. This idea was substantiated by our observation that LPS-induced upregulation of TNFα and IL-6 was significantly enhanced in both wild-type and Hfe-deficient macrophages by pre-treating the cells with hepcidin. Based on the importance of Toll-like receptor (TLR) 4 in the macrophage response to both LPS and Salmonella,^11–14 we then examined the activation of the 2 major signal transduction pathways downstream of this receptor. These pathways are both required for induction of pro-inflammatory cytokine expression and are dependent on 2 separate sets of adaptor proteins—Mal (MyD88 adaptor-like protein) and MyD88 (myeloid differentiation factor 88) and TRAM (TRIF-related adaptor molecule) and TRIF (TIR domain-containing adaptor inducing interferon (IFN)β), respectively—that interact with the TLR4 cytoplasmic domain (Fig. 2).¹⁵ We found that LPS-induced (TLR4-mediated) activation of MyD88-dependent responses, including activation of NFκB, the ERK kinase and the p38 kinase, was normal in the Hfe-deficient macrophages. In keeping with intact Mal/MyD88 function, TLR2-mediated upregulation of TNFα and IL-6 expression, which requires Mal/MyD88-dependent but not TRAM/TRIF-dependent signals, was also intact in the knock-out cells. These findings raised the possibility that the impaired LPS-induced inflammatory cytokine expression in the Hfe-deficient macrophages was caused by an abnormality in the TRAM/TRIF pathway. Consistent with this idea, TLR4-activated upregulation of IFN β expression, a TRAM/TRIF-dependent response, was significantly reduced in the knock-out macrophages. Surprisingly, upregulation of IFNβ expression following activation of TLR3 by the ligand poly(I:C), a response that requires TRIF but not TRAM, was normal in the Hfe-deficient macrophages. Furthermore, when intracellular iron levels were reduced in wild-type macrophages by treatment with a membrane permeable iron chelator, IFNβ production in response to LPS, but not poly(I:C), was inhibited. This pattern of abnormalities, together with what is known about TLR signal transduction mechanisms, allowed us to conclude that Hfe deficiency and the associated reduction in hepcidin and intra-macrophage iron levels impaired TLR4 signaling at an early step specific to the TRAM/TRIF pathway, probably proximal to TRAM (Fig. 2).¹⁵ We have not yet determined whether low intracellular iron levels have effects on responses activated by receptors other than TLRs 2, 3 and 4.

Figure 2.

Effects of Hfe deficiency and the associated low hepcidin and intracellular iron levels on macrophage TLR4 signaling. Based on our analysis of TLR2-, TLR3- and TLR4-activated responses, the impaired LPS-induced production of TNFα, IL-6 and IFNβ in the Hfe-deficient macrophages is likely to be caused by an abnormality in an early step of TLR4 signal transduction specific to the TRAM/TRIF pathway (indicated by the red bracket). Green arrows indicate activating responses, red lines designate inhibitory effects, and black lines represent protein-protein interactions.

At this time, we can only speculate about how low intracellular iron levels influence TLR4 signaling via the TRAM/TRIF pathway. Existing data do not suggest direct involvement of iron in the function of the proteins that are currently known to play a role in TLR4 activation and signal transduction. It is possible, however, that iron may influence such proteins indirectly by affecting their expression, post-translational modification or sub-cellular location. This is an idea that we are currently investigating. Alternatively, it is conceivable that future studies will reveal the participation of proteins directly regulated by iron in TLR4 function. It should also be kept in mind that the effects of Hfe deficiency on TLR4 signaling may involve mechanisms that are independent of changes in intracellular iron. Recent work has shown that the binding of hepcidin to FPN leads to the activation of the Jak2 kinase.¹⁶ Although the main consequence of this activation appears to be the phosphorylation of tyrosine residues in a cytosolic loop of FPN and the resultant internalization of the transporter, it is formally possible that Jak2 could influence activation of TLR4. The low levels of hepcidin associated with Hfe deficiency could then impact on TLR4 function as a result of changes in Jak2 activation status rather than via alterations in intracellular iron.

Regardless of the exact mechanism involved in the impaired TLR4 signaling, our findings indicated that low hepcidin expression was associated with attenuated inflammation and, conversely, that increased hepcidin levels promoted pro-inflammatory responses in macrophages. Furthermore, inflammatory conditions, including inflammatory bowel disease (IBD), lead to IL-6-mediated upregulation of hepcidin expression and a consequent hypoferremia that contributes to the development of the anemia of chronic disease.^17,18 We reasoned, therefore, that blocking hepcidin expression in such conditions might help both to reduce the unwanted inflammation and to correct the anemia. The second part of our recently published work was directed at testing this idea.¹⁰ Our strategy for inhibiting hepcidin expression was based on recent discoveries that demonstrated the essential role played by bone morphogenetic proteins (BMPs) in regulating hepcidin levels, and that identified the hepatocyte cell surface protein hemojuvelin (HJV) as a BMP co-receptor.^19–23 These studies showed that hepcidin expression could be inhibited, both in vitro and in vivo, by treatment with either a small organic compound dorsomorphin, which blocked intracellular signals activated by BMPs, or by a soluble, recombinant HJV.Fc fusion protein, which prevented the interaction of BMPs with their receptor.

We tested the in vivo effects of dorsomorphin, a related small molecule inhibitor LDN-193189 and HJV.Fc in both the mouse model of Salmonella-induced enterocolitis and in piroxicam-induced colitis in IL-10 knock-out mice, a model of chronic intestinal inflammation resembling human IBD.²⁴ Each of the reagents was able to inhibit the upregulation of hepcidin associated with the induction of intestinal inflammation. Gratifyingly, the inhibitors also significantly reduced the severity of intestinal inflammation in both colitis models. Although we cannot completely exclude the possibility that the suppression of inflammation was unrelated to the inhibition of hepcidin expression, the fact that it was seen with 2 structurally distinct categories of molecules that blocked hepcidin upregulation by different mechanisms suggests that it is unlikely to be an off-target effect. Dorsomorphin also partially reversed the decrease in serum iron levels associated with Salmonella enterocolitis, but the impact of LDN-193189 and HJV.Fc on the hypoferremia of piroxicaminduced colitis was not obvious. We are currently evaluating additional hepcidin inhibiting reagents in the piroxicam/IL-10 knock-out colitis model.

What are the implications of our findings? First and foremost, they indicate that hepcidin has a previously unrecognized role in the inflammatory response, probably as a result of its ability to bring about changes in intra-macrophage iron levels. This idea is supported by the clear reduction of in vitro and in vivo inflammatory responses associated with the low hepcidin levels seen in Hfe-deficient mice,⁸ and by the similar anti-inflammatory effects that follow inhibition of hepcidin expression in wild-type mice.¹⁰ It is also supported by our observation that exogenous hepcidin promotes inflammatory cytokine production in macrophages.¹⁰ Since many inflammatory conditions lead to increased expression of hepcidin,¹⁷ the cross-regulation between hepcidin and inflammation constitutes a positive feedback loop that may perpetuate the inflammatory state (Fig. 3). The involvement of hepcidin in such a vicious cycle would seem to be maladaptive. However, it is likely that the response evolved to deal with infection, and in that context an increase in hepcidin levels would have the dual advantages of helping to sequester iron from microbial pathogens and of “revving up” macrophage inflammatory responses that are involved in eliminating the pathogens.

Figure 3.

A hepcidin-mediated positive feedback loop perpetuates inflammation. Inflammatory conditions stimulate hepcidin upregulation via cytokines such as IL-6, TNFα and IL-1. The increased levels of hepcidin lead to FPN downregulation and a consequent rise in intra-macrophage iron, which, in turn, promotes macrophage production of TNFα and IL-6, thus exacerbating the inflammation.

The second important implication of our work is that inhibition of hepcidin expression or function may represent a new approach to suppressing unwanted inflammation based on interrupting the vicious cycle illustrated in Figure 3. Our initial application of this strategy in a mouse model of IBD has yielded very encouraging results,¹⁰ but validation in additional models of chronic inflammation will be required. This is one of the goals of experiments that are currently in progress in our laboratory. In addition to reagents such as dorsomorphin, LDN-193189 and HJV.Fc that act by suppressing hepcidin expression, inhibitors of hepcidin-FPN interactions would also be predicted to have anti-inflammatory effects and merit testing in models of inflammation. Besides the potential beneficial effects on inflammation that our work suggests, inhibition of hepcidin expression or function would have the added advantage of helping to correct the anemia that is often associated with long-standing inflammatory conditions.¹⁷ This so-called anemia of chronic disease is at least partly caused by abnormal elevation of hepcidin levels, which leads to downregulation of FPN expression and consequent decreases of iron absorption from the gut and iron release from phagocytes. The anemia can be a significant factor contributing to morbidity in illnesses such as IBD.²⁵ Treatment strategies that block hepcidin-mediated downregulation of FPN would restore normal iron homeostasis in these conditions, improve erythropoiesis and raise the general quality of life.

Finally, our studies raise the intriguing possibility that alterations in iron metabolism may influence susceptibility to chronic inflammatory conditions such as IBD. Although none of the genes that have been implicated in the pathogenesis of Crohn's disease and ulcerative colitis²⁶ is directly involved in the regulation of iron homeostasis, there are some potential areas of overlap. For instance, there is a well established link between IBD and the MHC, a locus that contains a large number of genes, one of which happens to be HFE. Similarly, two of the IBD susceptibility loci identified in genome-wide association studies encompass the JAK2 and STAT3 genes, which encode proteins involved in a number of biological processes, including FPN downregulation and hepcidin expression, respectively.^6,16 The significance of these tantalizing connections remains to be determined.

Our observations have shed new light on important interactions between iron homeostasis and inflammation, particularly with regard to the role played by hepcidin in both of these processes. Further characterization of these interactions and their involvement in the pathogenesis of chronic inflammatory conditions is likely to be a fruitful area for future investigations, one that may lead to the development of new anti-inflammatory therapies.

Abbreviations

BMP

bone morphogenetic protein

FPN

ferroportin

HJV

hemojuvelin

IBD

inflammatory bowel disease

IFN

interferon

Mal

MyD88 adaptor-like protein

MyD88

myeloid differentiation factor 88

TLR

Toll-like receptor

TRAM

TRIF-related adaptor molecule

TRIF

TIR domain-containing adaptor inducing IFNβ

Addendum to: Wang L, Harrington L, Trebicka E, Sh HN, Kagan JC, Hong CC, et al. Selective modulation of TLR4-activated inflammatory responses by altered iron homeostasis in mice. J Clin Invest. 2009;119:3322–3328. doi: 10.1172/JCI39939.