Published online before print March 17, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

,1
* Griffith University, Queensland, Australia; and
Institute for Molecular Bioscience, CRC for Chronic Inflammatory Diseases and ARC Special Research Centre for Functional and Applied Genomics, The University of Queensland, Australia
1 Correspondence: IMB, The University of Queensland, St. Lucia, Queensland, Australia 4072. E-mail: D.Hume{at}imb.uq.edu.au
|
|
|---|
Key Words: Toll-like receptor lipopolysaccharide macrophage
|
|
|---|
There are many parallels between the antisocial behavior of cancer cells in a malignancy and the cells in a chronic inflammatory lesion. Inflammation is the process of recruitment and activation of cells of the acquired and innate immune system in response to infection, trauma, or injury. The normal course of inflammation is directed toward removal of the inducing agent and repair of the damage, followed by elimination of the inflammatory cells (reviewed in ref. [2 ]). This process is critical to host defense and homeostasis.
Inflammation must be tightly regulated; it is the classical two-edged sword. Too little, and one is overwhelmed by infection or injury; too much, and the disease symptoms themselves become overwhelming. Chronic inflammatory diseases represent a massive burden on our society [2 ]. Even that modern scourge, obesity-associated (type II) diabetes, has been recognized as a chronic inflammatory disease [3 ]. There is a great deal of evidence for genetic variation amongst humans and experimental animals in susceptibility to infectious disease on the one hand and inflammatory pathology on the other. Genetic susceptibility in humans to diseases such as emphysema, juvenile arthritis, psoriasis, and chronic inflammatory bowel disease maps to many distinct loci {more than 450 distinct loci listed in the Online Mendelian Inheritance in Man (OMIM) database [4 ]}, suggesting that the outcome of the inflammatory process is controlled by diverse gene products. In this mini-review, we develop the idea that chronic inflammatory disease is commonly a failure of resolution rather than induction of the inflammatory process and that many of the genes that control susceptibility to such disease are analogous to tumor suppressor genes. We will call them inflammation suppressor genes. The review is focused on the activation of macrophages, key cells in the innate immune system, and in the initiation and resolution of inflammation, but the arguments are equally relevant to all cell types involved in immune defense.
|
|
|---|
(TNF-
), interleukin (IL)-1, and IL-6, are central to the initiation and maintenance of inflammation [2
]. The archetypal inducers of the transcription of proinflammatory genes are the products of microorganisms, many of which act through the Toll-like receptor (TLR) family of pathogen pattern recognition receptors. The most studied TLR ligand is the lipopolysaccharide (LPS), a component of the cell wall of gram-negative bacteria, recognition of which requires TLR4. LPS stimulation of macrophages initiates a complex signaling cascade. Recruitment of adaptor proteins such as MyD88 to the TLR4 receptor activates a signaling cascade, which culminates in nuclear translocation of the transcription factor complex, nuclear factor (NF)-
B (Fig. 1
). Many other pathways, acting in series or in parallel, are required for the complete response to LPS [5
]. In particular, the major subclasses of MAPK cascades [the extracellular signal-regulated kinases 1 and 2, jun kinase, and p38/stress-activated kinases] all contribute to the final transcriptional outcome.
![]() View larger version (36K): [in a new window] |
Figure 1. Cartoon of the signaling cascades activated by the LPS-TLR4 receptor complex. Inflammation suppressors are listed beside the diagram. MAPK, Mitogen-activated protein kinase; IRF3, interferon-regulatory factor 3; LBP, LPS-binding protein; TICAM, Toll-IL-1 receptor (IL-1R) homology domain-containing adapter molecule; IRAK2, IL-1R-associated kinase 2; Jip3, c-Jun NH2-terminal kinase (JNK)-interacting protein 3; MEKK4, MAPK kinase (MEK) kinase 4; Tollip, Toll-interacting protein; TRAF6, TNF receptor (TNFR)-associated factor 6; TBK1, TNFR-associated NF- B kinase-binding kinase 1; Tak1, transforming growth factor-ß (TGF- ß)-activated kinase 1; TAB1, Tak1-binding protein 1; IKK, inhibitor of B (I B) kinase; ELK-1, expressed sequence tag-like transcription factor 1; ATF2, activating transcription factor 2; AP-1, activator protein-1; ITIM, immunoreceptor tyrosine-based inhibitory motif; MMP, matrix metalloproteinase; GNS, glucosamine (N-acetyl)-6-sulfatase; card15, caspase-activating and recruitment domain-15; Nod2, nucleotide-binding oligomerization domain 2; Slpi, secretory leukocyte protease inhibitor; Timps, tissue inhibitors of metallo proteinases; MAL, myD88 adaptor-like protein; ECSIT, evolutionarily conserved signaling intermediate in Toll pathway.
|
There is evidence for a dual role for some proinflammatory compounds that are also necessary for resolution of an inflammatory response. Matrix metalloproteinases (MMP) and glycosoaminoglycan (GAG) inhibitors, for example, act to degrade matrix and permit cellular infiltrates but also act on the inflammatory cells themselves to inhibit signaling through surface receptors and accessory proteoglycans. MMP14 and the GAG inhibitor GNS are expressed late in the acute LPS response. The GNS substrate keratin is itself induced in the first wave of transcriptional responses to LPS [6 ]. The late response gene p47phox, necessary for superoxide production and oxidative burst, also plays a role in limiting local inflammation. Mice homozygous for the p47phox null allele (p47phox/) have exacerbated inflammatory lesions and poor resolution of experimentally induced arthritis [7 ].
|
|
|---|
B, such as IkB
and inhibitors of IL-1R (IL-1RA). An expanding class of immune inhibitory receptors are immunoreceptors with an ITIM, best characterized as antagonists of immunoglobulin G Fc receptors [8
, 9
]. Lyn kinase is a classic example of an ITIM and together with its substrate Hs1, is induced in the late response set of LPS-inducible transcripts [6
]. Lyn kinase has a demonstrated role in reducing sensitivity of macrophages to growth factors and inhibiting macrophage survival, where mutations in Lyn kinase lead to hyper-B cell inflammatory responses and myeloid leukemia [10
].
Targeted mutations in many of these inducible repressors cause spontaneous inflammation or hypersensitivity to the toxic effects of LPS administration or inflammatory stimuli. For example, deletion of the adenosine-uridine-rich element-binding proteins T cell intracellular antigen-1 and tristetraprolin, which promote the degradation of early-phase, proinflammatory mRNAs, leads to LPS hypersensitivity and susceptibility to experimental arthritis [11
]. Nathan [2
] has previously summarized mutations that generate spontaneous inflammation in mice or humans. Of the loci discussed, E3 ubiquitin ligase (and other components of the proteosome pathway), growth arrest and DNA damage-inducible 45ß, and haem oxygenase 1 (HO-1) are amongst the highly LPS-inducible genes in our studies [6
]. The repressive role of HO-1 might infer that the induction of the transcription factor hypoxia-inducible factor 1
(HIF-1
) by LPS in macrophages [12
] is also a part of the resolution process, and other HIF-1
targets may also be inflammation-suppressor genes.
From the viewpoint of human inflammatory disease, two such inducible repressors are of particular interest. The caspase recruitment domain family member CARD15/Nod2 was characterized as a disease-susceptibility locus involved in Crohns colitis, the multisystem inflammatory disease Blau syndrome, and some forms of psoriasis [4
]. Subsequent discoveries implicated the gene in macrophage recognition of muramyl dipeptide, implying that the mutant phenotype could be a result of failure to respond to a particular challenge. Two recent papers [13
, 14
] demonstrate that Nod2 is actually an inducible repressor of TLR signaling, providing a much more plausible explanation of the dysregulated, inflammatory response association with loss-of-function. Mutations in a structurally related protein, cryopyrin, also lead to multisystem inflammatory disease [2
]. Like Nod2, this protein acts as a LPS-inducible feedback repressor of NF-
B activation in macrophages [15
].
Alternate splicing events regulate the production of soluble forms of membrane-bound proteins such as receptors [16
], controlling the capacity of a cellular microenvironment to respond to a particular ligand. Alternate splicing of key signaling molecules analogously produces activating and repressor protein variants. At a higher level in the TLR pathway, there is a LPS-inducible splice variant of the key adaptor protein, MyD88, which binds activated TLRs but fails to recruit IRAK4 and thereby acts as a dominant-negative repressor [17
]. A macrophage-expressed and LPS-inducible IRAK isoform, IRAK-M [18
], as well as a LPS-inducible isoform of IRAK2 [19
] act to repress activation of NF-
B. Within the parallel MAPK pathways, the inactivating enzyme MAPK phosphatase-1 [20
] and the AP-1 transcriptional repressor, b-ATF [21
], are substantially LPS-inducible.
It is beyond the scope of this short review to identify and discuss the function of all of the late-inducible genes in the LPS-inducible gene set in macrophages. Two classes deserve mention. First, there is a large set of negative transcriptional regulators, including histone and DNA-modifying enzymes, which we infer eventually cause irreversible silencing of the proinflammatory genes. Second, many of the most LPS-inducible genes, such as Fas, myeloid cell leukemin sequence 1 (EAT/Mcl-1), and the kinase proviral integration site 1 (PIM-1) [6 ], are pro- or antiapoptotic. Macrophages in an inflammatory lesion probably go through two phases: first, needing to resist the toxic environment that is created by their peers and by pathogen and tissue damage and subsequently, being required to die and be cleared to permit the replacement of the tissue or scar formation. Just as failure of cell death can predispose to malignancy, inappropriately prolonged life of macrophages is likely to lead to inflammatory pathology. Indeed, mutations in the Fas cell death gene cause spontaneous, multiorgan inflammation [2 ].
|
|
|---|
and cyclooxygenase 2 are proven anti-inflammatory drug targets [2
]. The expression of late-response genes would likely predominate in disease tissue with the early-response genes appearing only transiently. The development of new disease therapies based on targets identified on expression profiling of disease tissue may be quite misleading, as targeting late-response genes, even those overtly proinflammatory players such as p47phox, could actually exacerbate disease. The intellectual question that arises from this discussion is: Why is the system so complicated? It seems like overkill to turn off every possible switch. The trite answer is that there needs to be tight control to ensure that the response to the wide range of challenges is appropriate. But if that is the case, why does mutation of any one negative-feedback regulator generate an overt pathology and dysregulation? Surely the system would be more robust. Could there be an advantage to a lack of robustness? If an animal population responded the same way to an infectious challenge, they would likely be wiped out. So, it is not surprising that all the components of the LPS receptor complex, including TLR4, are polymorphic in humans [22 ]. Polymorphism is also evident in inflammation suppressors. For example, many distinct alleles of the Nod2 gene have been found amongst inbred mouse strains in keeping with the polymorphism in humans [23 ]. As there is such a diversity of pathogens, an individual has the greatest chance of survival if he is maximally heterozygous at many different loci that control the inflammatory process and consequentially, susceptible to many different pathogens.
Tumor suppressor genes require two-hits for loss-of-function leading to a malignancy. For immune-suppressor genes, however, selection for genetic heterozygosity can only operate if there is a heterozygous phenotype; that is, carrying two alleles of an inflammation suppressor gene should provide more chances of generating an adequate immune response than inheriting a single version of that gene. One mechanism analogous to loss-of-heterozygosity of tumor suppressor genes would be hemizygous expression of a functional polymorphism. Indeed, monoallelic expression of some immune genes, such as TLR4 [24 ], is known to generate two populations of macrophages expressing different functional alleles.
If variation in the macrophage activation pathway amongst individuals is desirable, evolution would therefore select against robustness in that pathway. Aside from straightforward gene dosage effects on the amount of protein per cell, heterozygosity of genes involved in feedback control of macrophage activation manifests itself by altering the phenotype of individual macrophages so that each one presents a unique challenge to a potential pathogen. We showed previously that the set of LPS-inducible genes in macrophages varies at the single-cell level in a probabilistic manner [25 ]. This means that individual cells actually induce only a subset of the potential inflammation suppressor genes and most likely, use only one allele. As a consequence, they may switch off only a subset of the parallel pathways activated by the primary stimulus. The family of LPS-stimulated macrophages therefore represents a population of cells with quite distinct states and extents of activation, presenting an infinitely complex target to a pathogen. Where there is allelic variation, the extent of the diversity amongst macrophages becomes even greater.
We suggest that the system of macrophage activation has been under selection pressure to favor diversity of individuals in a population and of macrophages within an individual. The concept of heterozygous advantage is well-established in evolutionary genetics, perhaps the archetype being the advantage conferred by heterozygous sickle cell haemoglobin mutations in resistance to malaria. We propose that allelic diversity in many different inflammation suppressor genes and consequential heterozygosity amongst most individuals in a population contribute to defense against the broadest possible array of pathogens. As with sickle cell disease, the fact that homozygous mutations at particular loci can confer susceptibility to chronic inflammatory disease may be the price we have to pay.
Received December 7, 2004; revised January 17, 2005; accepted January 18, 2005.
|
|
|---|
, cyclooxygenase 2, and inflammatory arthritis Proc. Natl. Acad. Sci. USA 101,2011-2016
Blood 103,1124-1130
B suppressive properties J. Immunol. 171,6329-6333
B activation by alternative splicing of MyD88 Curr. Biol. 12,467-471[CrossRef][Medline]
is required for the induction of mitogen-activated protein kinase phosphatase-1 in lipopolysaccharide-stimulated macrophages J. Immunol. 164,29-37
B kinase complex and NF-
B act as master regulators of lipopolysaccharide-induced gene expression and control subordinate activation of AP-1 Mol. Cell. Biol. 24,6488-6500This article has been cited by other articles:
![]() |
D. A. Hume Macrophages as APC and the Dendritic Cell Myth J. Immunol., November 1, 2008; 181(9): 5829 - 5835. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Varga, J. Ehrchen, A. Tsianakas, K. Tenbrock, A. Rattenholl, S. Seeliger, M. Mack, J. Roth, and C. Sunderkoetter Glucocorticoids induce an activated, anti-inflammatory monocyte subset in mice that resembles myeloid-derived suppressor cells J. Leukoc. Biol., September 1, 2008; 84(3): 644 - 650. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Cao, C. Bao, E. Padalko, and C. J. Lowenstein Acetylation of mitogen-activated protein kinase phosphatase-1 inhibits Toll-like receptor signaling J. Exp. Med., June 9, 2008; 205(6): 1491 - 1503. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. M. Ripoll, K. M. Irvine, T. Ravasi, M. J. Sweet, and D. A. Hume Gpnmb Is Induced in Macrophages by IFN-{gamma} and Lipopolysaccharide and Acts as a Feedback Regulator of Proinflammatory Responses J. Immunol., May 15, 2007; 178(10): 6557 - 6566. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Bhattacharyya, D. E. Brown, J. A. Brewer, S. K. Vogt, and L. J. Muglia Macrophage glucocorticoid receptors regulate Toll-like receptor 4-mediated inflammatory responses by selective inhibition of p38 MAP kinase Blood, May 15, 2007; 109(10): 4313 - 4319. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Ehrchen, L. Steinmuller, K. Barczyk, K. Tenbrock, W. Nacken, M. Eisenacher, U. Nordhues, C. Sorg, C. Sunderkotter, and J. Roth Glucocorticoids induce differentiation of a specifically activated, anti-inflammatory subtype of human monocytes Blood, February 1, 2007; 109(3): 1265 - 1274. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Foti, F. Granucci, M. Pelizzola, O. Beretta, and P. Ricciardi-Castagnoli Dendritic cells in pathogen recognition and induction of immune responses: a functional genomics approach J. Leukoc. Biol., May 1, 2006; 79(5): 913 - 916. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||