Annexin A1 and Glucocorticoids As Effectors of the Resolution Of

PersPectives

55. Sharpless, N. E. & DePinho, R. A. How stem cells age Acknowledgements facilitates the control of several homeostatic and why this makes us grow old. Nature Rev. Mol. Cell Work from our laboratory was supported by grant AG21450 Biol. 8, 703–713 (2007). from the National Institutes of Health, USA. processes that are crucial for health. During 56. Krishnamurthy, J. et al. Ink4a/Arf expression is a inflammation, increased levels of cortisol biomarker of aging. J. Clin. Invest. 114, 1299–1307 (2004). DAtABAsEs dampen local and systemic inflammatory 57. Campisi, J. Cancer and ageing: rival demons? Nature entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query. events, thereby favouring proper resolution Rev. Cancer 3, 339–349 (2003). fcgi?db=gene 4,5 58. Signer, R. A. J., Montecino‑Rodriguez, E., Witte, O. N. cD28 | CDKN2A | FGFr2 | growth hormone | iGF1 | iL-6 | iL-7 | of the inflammatory response . These & Dorshkind, K. Aging and cancer resistance in KGF fundamental pathophysiological functions lymphoid progenitors are linked processes conferred by p16Ink4a and Arf. Genes Dev. 22, 3115–3120 FURtHER inFoRMAtion of glucocorticoids are achieved through Kenneth Dorshkind’s homepage: http://faculty.uclaaccess. (2008). ucla.edu/institution/personnel?personnel_id=45615 many molecular mechanisms, which can be 59. Rossi, D. J. et al. Cell intrinsic alterations underlie http://www.cdc.gov/nchs/data/hus/ hematopoietic stem cell aging. Proc. Natl Acad. Sci. Us census Bureau: broadly divided into genomic mechanisms hus07.pdf#fig01 USA 102, 9194–9199 (2005). (involving transactivation or transrepres- 60. Montecino‑Rodriguez, E., Clark, R. & Dorshkind, K. World Health Organization: http://www.who.int/en/ Effects of insulin‑like growth factor administration seNieUr Protocol: http://www.medizin.uni-tuebingen.de/ sion of gene transcription) and non- eucambis/home/senieur.html and bone marrow transplantation on thymopoiesis genomic mechanisms (that are rapid and in aged mice. Endocrinology 139, 4120–4126 All lInks ArE AcTIvE In ThE onlInE pdf (1998). independent of de novo protein synthesis) (BOX 1). In this Opinion article, we focus on one downstream mediator of glucocorticoids, the 37 kDa protein annexin A1 (also known oPinion as lipocortin 1; encoded by ANXA1). We review the recent evidence indicating that glucocorti coids regulate the synthesis and Annexin A1 and glucocorticoids function of annexin A1 (reFs 4,6), possibly through a combination of both genomic as effectors of the resolution of and non-genomic processes, depending on the cell type and the time of induction. inflammation In addition, we describe the emerging data showing that glucocorticoids can differentially affect the annexin A1 path- Mauro Perretti and Fulvio D’Acquisto way in cells of the innate and adaptive Abstract | Glucocorticoids are widely used for the management of inflammatory immune system, in which this pathway can have opposing effects. We propose that diseases. Their clinical application stems from our understanding of the inhibitory the annexin A1 pathway is an important effect of the corticosteroid hormone cortisol on several components of the immune mediator of the anti-inflammatory effects system. endogenous and exogenous glucocorticoids mediate their multiple of glucocorticoids. anti-inflammatory effects through many effector molecules. in this Opinion article, we focus on the role of one such effector molecule, annexin A1, and summarize the Effects in innate immunity recent studies that provide insight into its molecular and pharmacological functions Annexin A1 is a member of a superfamily of annexin proteins that bind acidic phospho- in immune responses. in addition, we propose a model in which glucocorticoids lipids with high affinity in the presence regulate the expression and function of annexin A1 in opposing ways in innate and of Ca2+ (reF. 7). There are 13 mammalian adaptive immune cells to mediate the resolution of inflammation. annexin proteins, each of which has specific biological functions. Similarly to other Inflammation is a primordial response that components of the inflammatory response annexin proteins, annexin A1 is expressed functions to protect the host against invasion is crucial for restoring tissue structure and in resting cells and binds to acidic phospho- by pathogens or exposure to xenobiotics. homeostasis. Elucidation of their effects on lipids in the presence of Ca2+; original studies However, overly aggressive or prolonged immune cells could lead to the identification of annexin A1 therefore focused on its role inflammatory responses can be detrimental of the molecular target (or targets) of a given in granule fusion and exocytosis, in which it to the host. Therefore, higher organisms mediator, thereby prompting innovative was shown to promote the fusion of vesicle have evolved mechanisms to ensure that drug discovery for the treatment of chronic membranes with plasma membranes in the inflammatory response is limited in inflammatory conditions2,3. reconstituted systems7. time and space. Many endogenous anti- Glucocorticoids are the first class of inflammatory and pro-resolving mediators endogenous anti-inflammatory mediators The actions of annexin A1. In resting con- function to counteract the properties of pro- that have been successfully used for thera- ditions, human and mouse neutrophils, inflammatory factors and to ensure a prompt peutic purposes; budesonide and beclo- monocytes and macrophages constitutively resolution of inflammation. The concept methasone are widely used for the treatment contain high levels of annexin A1 in their that inflammation is resolved in a time- and of asthma, prednisolone is used for rheu- cytoplasm8–10. Following cell activation space-specific manner is emerging, such that ma toid arthritis and other auto immune (for example, by neutrophil adhesion to different outcomes can be produced accord- diseases, and mometasone and hydrocorti- endothelial-cell monolayers), annexin A1 ing to the stage or site at which a given sone are used for eczema and psoriasis. In is promptly mobilized to the cell surface pathway or mediator becomes operative1,2. healthy individuals, the circadian release and secreted11. The molecular mechanisms The action of anti-inflammatory and pro- of glucocorticoids (such as cortisol and that are responsible for this rapid secretion resolving mediators and pathways on several corticosterone) from the adrenal glands are cell specific. In macrophages, the

23 Box 1 | the molecular mechanisms by which glucocorticoids affect gene expression wild-type neutrophils . These data are in agreement with an earlier study in which genomic effects glucocorticoid treatment increased the levels Glucocorticoids have numerous effects on gene expression, such that it has been proposed that of annexin A1 in circulating leukocytes and ~1% of the genome might be modulated by these drugs4. The glucocorticoid receptor is retained in an inactive state by several chaperone proteins, including heat-shock protein 70 (HSP70) and inhibition of annexin A1 prolonged leuko- HSP90, and following ligand binding, the glucocorticoid–receptor complex translocates to the cyte migration time in the post-capillary 24 nucleus, where it modulates gene expression through transrepression or transactivation. venules of the hamster cheek pouch . Transrepression occurs when the glucocorticoid–receptor complex interacts with transcription Further investigation of Anxa1–/– mice has factors: it can sequester them in the cytoplasm, promote their degradation or inhibit them through shown that these mice develop more-severe other mechanisms. Through transrepression, the glucocorticoid–receptor complex prevents the joint disease (characterized by increased transcription of the target genes, including pro-inflammatory cytokines, growth factors, adhesion destruction of the cartilage and a thicker 50 molecules, nitric oxide, prostanoids and other autacoids . In addition, glucocorticoids transrepress pannus at day seven) than wild-type mice in the transcription of nuclear factor-κB (NF-κB)-induced pro-inflammatory genes by inducing the a model of monoarticular antigen-induced activation of histone deacetylase 2, which reduces the access of NF-κB to the promoter regions arthritis25,26. Increased levels of mRnA of its target genes70. Transactivation involves the direct binding of the glucocorticoid–receptor complex to specific encoding interleukin-1 (Il-1) and Il-6 were 26 nucleotide sequences (termed glucocorticoid-response elements) to promote the transcription also detected in the joints of these mice , of genes, such as interleukin-10, β-adrenergic receptor, interleukin-1-receptor antagonist and which is consistent with the known require- dual-specificity protein phosphatase 1 (DUSP1), many of which have anti-inflammatory functions41. ment for Il-6 in this model of arthritis27. In The anti-inflammatory effects of glucocorticoid-mediated transactivation can also be reinforced addition, Anxa1–/– lung fibroblasts showed indirectly; for example, DUSP1 downregulates mitogen-activated-protein-kinase signalling and higher Il-1-induced Il-6 production, an thereby inhibits cytokine synthesis. Finally, it is possible that annexin A1 (encoded by ANXA1) effect that was shown to be an indirect conse- functions as a mediator of some of these glucocorticoid effects, as suggested by decreased levels quence of decreased levels of dual-specificity –/– 28 of DUSP1 expression in Anxa1 lung fibroblasts (see the main text for more details). protein phosphatase 1 (DUSP1; also known non-genomic effects as MKP1) in these cells28. The potential Glucocorticoids can also have rapid effects, which occur within a few minutes, by modulating involvement of DUSP1 in the mechanism of the degree of activation and responsiveness of target cells. These actions do not require de novo protein synthesis and are referred to as non-genomic effects71. Initially reported in monocytes, the action of annexin A1 is reminiscent of what non-genomic actions of glucocorticoids have been recently described in T cells72 and platelets73. has been observed following glucocorticoid However, it is unclear how non-genomic effects contribute to the overall therapeutic efficacy of treatment, in which the induction of DUSP1 glucocorticoids in controlling vascular inflammatory pathologies. expression leads to the inhibition of mitogen- activated protein kinase signalling and thereby to a reduction in cytokine signalling and cell activation (BOX 1). Overactivation of ATP-binding cassette (ABC) transporter an inflammatory response had an important intracellular signalling pathways that induce system has been shown to be responsible for anti-inflammatory effect, as revealed by the release of effector molecules might also the secretion of annexin A1 (reF. 12), whereas analysis of the mesenteric microcircula- contribute to the observed lethality from in pituitary cells the phosphorylation of tion. This effect was associated with the endotoxic shock in Anxa1–/– mice following annexin A1 at the amino-terminal serine 27 detachment of adherent neutrophils from the administration of an otherwise sub- residue is required13. In human neutrophils, the vascular wall, which resulted in a reduc- lethal dose of lipopolysaccharide (lPS)29. a high proportion (>60%) of cytoplasmic tion in the number of cells that migrated Compared with wild-type mice, lPS-induced annexin A1 is stored in gelatinase granules9,14 into the subendothelial-matrix tissue19 endotoxic shock in Anxa1–/– mice was asso- and can be rapidly mobilized following (FIG. 2). These in vivo effects are supported ciated with a delayed and more prolonged exposure of the cells to weak activating sig- by in vitro analyses showing that annexin increase in the levels of tumour-necrosis nals, such as low concentrations of chemo- A1 or annexin A1-derived peptides inhibit factor, Il-1 and Il-6 in the blood, as well as attractants or adhesion to endothelial-cell neutrophil adhesion to and transmigra- with increased production of these cytokines monolayers11,15. Cell activation generally tion through endo thelial-cell monolay- by peritoneal macrophages. leads to the relocation of annexin A1 to the ers20,21. Quantification of the interaction of In summary, exogenous and endogenous outside leaflet of the plasma membrane, to neutrophils with human umbilical-vein annexin A1 counter-regulate the activities which it is tethered in a calcium-dependent endothelial-cell monolayers under flow of innate immune cells, in particular manner. In addition, extracellular concentra- has recently confirmed these observations, extra vasation and the generation of pro- tions of Ca2+ of ≥1 mM lead to a conforma- showing that low concentrations of annexin inflammatory mediators, and this ensures tion change in the protein, such that the A1 (10 nM) significantly inhibit neutrophil that a sufficient level of activation is reached n-terminal region that is otherwise buried adhesion22. The observed inhibitory activi- but not exceeded. becomes exposed and the active form of ties of recombinant annexin A1 or annexin annexin A1 is generated7,16 (FIG. 1). A1-derived peptides are consistent with the ALXR: the annexin A1 receptor. Annexin Several studies using experimental phenotype of Anxa1–/– neutrophils, which A1 signals through a seven-membrane- models of inflammation have shown that show increased transmigration in the cre- spanning G-protein-coupled receptor administration of annexin A1, or of the bio- master microcirculation of Anxa1–/– mice23. (GPCR)30 known as formyl peptide recep- active peptides that comprise its n-terminal This phenotype is probably neutrophil- tor 2 (FPR2; also known as AlXR in region, to mice results in potent inhibition intrinsic because Anxa1–/– neutro phils show humans), which is also the receptor for the 17,18 19 of neutrophil trafficking . In one study , higher chemotaxis and CD11b upregula- anti-inflammatory molecule lipoxin A4 annexin A1 administration to mice during tion in response to chemoattractants than (BOX 2). Here, we use the term AlXR, to

Paracrine In addition to sharing a receptor, a com- activation mon function for annexin A1 and lipoxin A4 has been proposed recently to explain why germ-free mice show inflammatory hypo- responsiveness39. Compared with wild-type ABC Monocyte transporter mice, germ-free mice showed increased b P levels of endogenous annexin A1 and a lipoxin A , and this was associated with Autocrine Juxtacrine 4 activation Granule increased levels of the anti-inflammatory c activation cytokine Il-10 in the gut; the increased Annexin A1 Il-10 levels were responsible for the reduced ALXR inflammation observed in these mice. A role for endogenous Il-10 in the anti- inflammatory effects of annexin A1 is also consistent with the observation that exposure Glucocorticoid Glucocorticoid Neutrophil induction of induction of of macrophages to annexin A1-derived ANXA1 expression ALXR expression bioactive peptides in vitro induces Il-10 secretion40. Moreover, the finding that Figure 1 | mobilization of annexin A1 in activated cells and its potential mode of action. gluco corticoids also mediate some of their Following cell activation, such as through adhesion to endothelial-cell monolayers,Nature Re intracellularviews | Immunolog annexiny anti-inflammatory effects through the A1 is mobilized to the plasma membrane. Depending on the cell type, annexin A1 is then externalized induction of expression of Il-10 (reF. 41) and and/or secreted through one of three mechanisms: through the activation of the ATP-binding cassette DUSP1 (as discussed above) supports the 12 (ABc) transporter (a); through phosphorylation of its amino-terminal serine 27 residue followed idea that annexin A1 and gluco corticoids by membrane localization to specific lipid domains13 before it moves to the outer leaflet of the have overlapping activities in the regulation plasma membrane and is secreted (b); or through fusion of annexin A1-containing granules with the plasma membrane, followed by release of annexin A1 (c). In the presence of ≥1 mM Ca2+, extra- of inflammatory responses. However, the cellular annexin A1 undergoes a conformational change that leads to exposure of the N-terminal region contribution made by Il-10 and DUSP1 and binding to its receptor ALXr (also known as FPr2). Annexin A1 can function in an autocrine, para- to the range of biological actions that are crine and juxtacrine (involving cell–cell contact) manner to activate ALXr signalling. The juxtacrine elicited by annexin A1 and glucocorticoids interaction between annexin A1 that is tethered on the surface of the producing cell and ALXr on the remains to be determined. target cell might be the most plausible mechanism of action in inflammatory conditions, as ‘activated’ annexin A1 could bind to acidic phospholipids at plasma ca2+ concentration of ≥1 mM and could there- Glucocorticoids induce annexin A1 and fore be presented to target cells that express ALXr. The annexin A1–ALXr pathway can be manipulated ALXR expression. Administration of gluco- by glucocorticoids, which induce the expression of the gene encoding annexin A1, ANXA1, as well corticoids to healthy human volunteers leads as ALXR, by innate immune cells, thereby increasing the effects of this anti-inflammatory circuit. to an increase in the levels of annexin A1 expression by circulating monocytes and neutrophils42. The levels of annexin A1 are reflect the biological function of this receptor rapid increase in the concentration of intra- also regulated by glucocorticoids during instead of its structural homology with FPR1. cellular calcium32,35. Blocking of AlXR with disease. More specifically, leukocytes from Human AlXR belongs to a small family of a monoclonal antibody prevented annexin patients with Cushing’s disease (which is receptors consisting of three members (FPR1, A1-induced inhibition of human neutrophil associated with high levels of cortisol) have 32 AlXR and FPR3) that are coupled to Gi pro- transmigration and adhesion to higher intracellular levels of annexin A1, teins and are expressed by several cell types, endothelial-cell monolayers under flow22. and leukocytes from patients with Addison’s including human neutrophils, monocytes, As AlXR-deficient mice have yet to be disease (which is associated with low levels macrophages, endothelial cells and epithelial generated and FPR1 is structurally related to of cortisol) have lower intracellular levels of cells31,32. Annexin A1 and peptides derived AlXR, Fpr1–/– mice and pharmacological pan- annexin A1 than healthy controls10. However, from its n-terminal region compete with receptor antagonists (that can block members the mechanism by which glucocorticoids lipoxin A4 and with another AlXR ligand, of the FPR family in mice) have been used to regulate annexin A1 has not been defined. serum amyloid protein A30,33, for binding to study the effects of annexin A1 in the absence The ANXA1 promoter does not seem to AlXR. Intriguingly, peptides derived from of signalling through FPRs. In a model of contain a canonical glucocorticoid-response the annexin A1 n-terminal region have been peritonitis, the inhibition of leukocyte element, but it does contain a partial consen- shown in vitro to activate all three receptors migration by exogenous annexin A1 was par- sus-binding site that mediates responsiveness of the FPR family34. However, the biological tially reduced in Fpr1–/– mice37. By contrast, to Il-6 (reF. 43), which suggests that gluco- significance of this finding is not clear, as the the increased neutrophil detachment that corticoids indirectly regulate the expres- generation of short and bioactive annexin A1 was observed in the mesenteric microcircula- sion of annexin A1. However, more work is fragments in vivo has not been proven. tion following the administration of annexin needed to determine the mechanisms that

Downstream signalling following A1-derived peptides or lipoxin A4 remained are involved in the cell-specific induction of annexin A1-induced AlXR activation intact in Fpr1–/– mice, but was abrogated annexin A1 expression by glucocorticoids, involves transient phosphorylation of by the administration of the pan-receptor as it remains possible that glucocorticoid- extracellular-regulated kinase 1 (ERK1; also antagonist38. This suggests that mouse AlXR response elements might be present in other known as MAPK3) and ERK2 (also known is probably the main receptor for annexin A1; regions of the ANXA1 promoter that are yet as MAPK1)22,33, which is associated with a the same is likely to apply in humans. to be explored.

Despite the probable involvement of many transcription factors and the differential effects of glucocorticoids in distinct Glucocorticoids Annexin A1 and its bioactive peptides ↑ Neutrophil detachment ↑ Neutrophil detachment cell types, a link between glucocorticoid ↓ Neutrophil transmigration ↓ Neutrophil transmigration administration and annexin A1 expres- ↓ Endothelial-cell adhesion molecule ↑ L-selectin shedding sion has been confirmed in Anxa1–/– mice expression ↑ L-selectin shedding bearing a LacZ reporter construct under ↑ Neutrophil annexin A1 expression 44 Detachment the control of the Anxa1 promoter . Blood vessel Treatment of these mice with dexametha- Neutrophil sone increased the activity of the AnxA1 promoter in circulating neutrophils and monocytes, leading to a significant increase Transmigration in LacZ expression as early as 2 hours after Endothelial cell dexamethasone injection45. The timing of the increase in promoter activity in these cells following dexamethasone injection Monocyte Antibacterial Inflammatory activities correlates with that of annexin A1 expres- exudate sion in cells from wild-type mice. This further supports the idea that annexin A1 Annexin A1 expression is regulated by glucocorticoids Macrophage (as shown in the human studies), although Phagocytosis Apoptotic the molecular mechanism underlying this neutrophil process remains unclear. Glucocorticoids Annexin A1 Glucocorticoids have also recently been ↓ Neutrophil apoptosis ↑ Neutrophil apoptosis shown to induce the expression of AlXR. ↑ Phagocytosis of apoptotic neutrophils ↑ ANXA1 expression Incubation of human monocytes with ↑ Macrophage annexin A1 expression (neutrophils and macrophages) dexa methasone or other synthetic gluco- ↑ Macrophage ALXR expression ↑ Phagocytosis of apoptotic corticoids induced the de novo synthesis ↑ Annexin A1 secretion neutrophils of AlXR, leading to an increase in mRnA Figure 2 | leukocyte trafficking and fate in inflammation: annexin A1–AlXr checkpoints. levels after 4–6 hours and protein levels after During an inflammatory response, leukocytes, such as neutrophils and monocytes, are recruited to the 46 Nature Reviews | Immunology 12–24 hours . A microarray study of human site of inflammation. Glucocorticoids control the interaction of leukocytes with the vessel wall by affect- monocytes that had been treated with the ing both the level of leukocyte responsiveness and endothelial-cell reactivity. Glucocorticoid-induced glucocorticoid fluticasone for 16 hours upregulation of the expression of annexin A1 by circulating leukocytes is one possible mechanism by showed that FPR1 mRnA (which encodes which glucocorticoids might inhibit leukocyte responses. Pharmacological administration of annexin a receptor that is structurally related to A1 can also mimic the effect of glucocorticoids, resulting in the detachment of adherent cells and the AlXR) was upregulated, but ALXR mRnA inhibition of neutrophil transmigration. shedding of L-selectin could be one of the molecular mecha- 20 was not47. However, it is worth noting that nisms that mediate this glucocorticoid- and annexin A1-mediated effect on leukocytes . in the inflam- fluticasone-mediated induction of ANXA1 matory exudate, monocytes differentiate into macrophages and the lifespan of neutrophils is prolonged by 3–5 days, which allows sufficient time for them to exert their antibacterial and phagocytic actions. mRnA was also not detected under the in addition, the intracellular concentration of annexin A1 is restored in the exudate by neutrophils conditions used in this study. Another study through de novo protein synthesis. At the end of this period, the neutrophils die by apoptosis, which showed that dexamethasone treatment led limits the release of annexin A1 and toxins, and prevents further tissue damage. Following glucocorti- to the upregulation of AlXR expression coid therapy, the secretion of annexin A1 by macrophages is also increased. in addition, the production by human neutrophils, as well as by skin of annexin A1 by other cell types, such as mast cells and monocytes, promotes the clearance of apop- tissues in a mouse model of dermatitis48. totic neutrophils and cell bodies by macrophages. it is unclear whether this effect is mediated by ALXr

We think these data support the hypothesis (also known as FPr2), although other ALXr ligands, such as lipoxin A4, can also promote the clearance that the upregulation of specific anti- of apoptotic cells. Phagocytosis of apoptotic neutrophils by macrophages leads to the transduction of inflammatory GPCRs (such as AlXR) is anti-inflammatory signals, including those mediated by transforming growth factor-β. involved in a mechanism that mediates the anti-inflammatory effects of glucocorticoids. In addition, we propose that the link the availability of annexin A1 at the cell In conclusion, we believe that gluco- between glucocorticoids and the expression surface (where its receptor is located) corticoids positively regulate the annexin of annexin A1 and AlXR involves a genomic through non-genomic mechanisms. A1–AlXR pathway, thereby assuring an mechanism, and therefore requires de novo They then (within 2–4 hours) upregulate appropriate level of activation of innate protein synthesis, although this needs to be the expression of the genes that encode immune cells while limiting the duration further substantiated experimentally. annexin A1 and AlXR through genomic of the pro-inflammatory response. Finally, the finding that glucocorti- mechanisms. In the context of resolution coids can also cause a rapid non-genomic of inflammation, recent work indicates Effects in adaptive immunity mobilization and secretion of annexin A1 that the interaction between annexin A1 The widespread use of glucocorticoids as by target cells49 has led to the hypothesis and AlXR has a role in controlling leuko- a treatment for inflammatory diseases is that, during an inflammatory response, cyte apoptosis and leukocyte clearance by largely due to their ability to simultaneously glucocorticoids first rapidly increase macrophages (BOX 3). block both innate and adaptive immune

Box 2 | Pharmacophysiology of ALXR concentrations, on T cells that had been stimulated with CD3- and CD28-specific The receptor for lipoxin A (ALXR; also known as FPR2), which is also thought to be the receptor 58 4 antibodies . Exposure to annexin A1 had no for annexin A1, is a seven-membrane-spanning G-protein-coupled receptor (GPCR) that is effect on naive T cells, as they express neg- structurally related to formyl peptide receptor 1 (FPR1)31. ALXR can recognize various different ligible levels of AlXR. However, following types of ligand, including several short-lived lipids, such as lipoxin A4 and its stable analogues, as well as proteins and peptides, such as serum amyloid protein A (SAA) and those derived from viral stimulation with CD3- and CD28-specific glycoproteins74,75. Although it was originally cloned from monomyelocytic cells, ALXR is expressed antibodies, annexin A1 led to the upregula- by a wide range of cells, including epithelial cells, endothelial cells, T cells and cells of the central tion of AlXR expression on the surface of nervous system (CNS). In the CNS, ALXR is expressed by microglial cells, in which it mediates the the T cells in a time- and concentration- activating and pro-inflammatory properties of amyloid-β peptide75. dependent manner. Interestingly, this The long list of ALXR agonists underlies the versatility of this GPCR and its ability to bind a wide occurred in parallel with the mobilization range of molecules (such as lipids, proteins and peptides)76. It is probable that different receptor and release of endogenous annexin A1 regions, and perhaps conformations, might be required for, or are induced after, binding to each (reF. 58). Upregulation of the annexin distinct agonist77. Consistent with this hypothesis, a comparison of lipoxin A - and SAA-mediated 4 A1–AlXR pathway in T cells following treat- responses in human neutrophils33 showed that SAA activated the pro-inflammatory transcription ment with annexin A1 led to increased and factor nuclear factor-κB, whereas lipoxin A inhibited the effect of SAA. In addition, many ALXR 4 prolonged phosphorylation of AKT and ERK agonists, but not lipoxin A4, can induce a rapid and transient increase in the concentration of intracellular calcium. after the cells were stimulated with CD3- and CD28-specific antibodies. This is in agree- Both annexin A1 and lipoxin A4 induce rapid phosphorylation of extracellular-signal-regulated kinase31. More recent studies have used genomic and proteomic approaches to determine the ment with the observation that T cells from downstream targets of annexin A1-induced signalling. For example, CC-chemokine receptor 10 Anxa1–/– mice show a lower (suboptimal) (CCR10) has been shown to be one such target in an epithelial cell line and in cells from the mouse signalling response following stimulation trachea and ileum78. In addition, annexin A1 has been shown to upregulate IL1RA (interleukin-1 than wild-type mice59. The increase in ERK 79 receptor antagonist) and downregulate CCR2 among the genes it regulates in human monocytes . activation by annexin A1 also occurs in Small chemical ligands that function as selective agonists for ALXR but not FPR1 have recently neutro phils and other cells in the innate been developed80. Importantly, these small molecules showed anti-inflammatory properties in immune system. However, the consequences mice80, which supports the hypothesis that ALXR is an important anti-inflammatory GPCR during acute inflammation. of this increase differ between cell types — ERK and AKT activation in T cells leads to cell proliferation, whereas ERK activation in responses50. With regard to the inhibition of Indeed, naive T cells contain ~100-fold less neutrophils leads to a loss of cell adhesion adaptive immune responses, glucocorticoids ANXA1 mRnA and protein than neutrophils and can have an anti-inflammatory effect. have marked effects on T-cell activation and and macrophages42; however, annexin A1 The divergent outcomes of the ERK signal- differentiation pathways. In naive T cells, glu- expression is upregulated in TH cells following ling pathways suggest that many checkpoints cocorticoids negatively modulate early T-cell activation and differentiation (F.D’A., unpub- exist to control cell behaviour and responses60, receptor (TCR)-initiated signalling events51,52 lished observations). The first evidence for and highlight the importance of cell type and and inhibit the activation of key downstream a role of annexin A1 in T-cell function came of the microenvironment in determining the transcription factors, such as activator pro- from the studies of its effects, at nanomolar biological function of a molecule. tein 1 (AP1), nuclear factor-κB (nF-κB) and nuclear factor of activated T cells (nFAT)53. + In vitro studies using isolated CD4 T cells Box 3 | Annexin A1 and ALXR in leukocyte apoptosis and in vivo studies in mice have shown that Endogenous glucocorticoids control the rate of clearance of apoptotic neutrophils by glucocorticoids promote T helper 2 (TH2)- type responses directly and indirectly54–56, an macrophages, as shown by the delay in apoptotic-cell clearance in the absence of 11-β- effect that would explain their therapeutic hydroxysteroid dehydrogenase type 1, an enzyme that is required to convert endogenous glucocorticoids into bioactive corticosterone81. Such a mechanism is consistent with the efficacy in T 1-cell-driven autoimmune dis- H well-known effects of exogenous glucocorticoids in inhibiting neutrophil apoptosis, while inducing eases. Glucocorticoids are also widely used eosinophil and lymphocyte apoptosis, and promoting the non-inflammatory removal of apoptotic for the treatment of TH2-cell-driven diseases, cells by macrophages82. such as asthma, in which they act by suppress- Three main effects of annexin A1 in leukocyte apoptosis have been described. First, exogenous ing the production of cytokines that propagate administration of annexin A1 promotes human neutrophil apoptosis owing to transient calcium inflammation in these diseases, such as Il-5 fluxes and dephosphorylation of the pro-apoptotic protein BCL-2-antagonist of cell death (BAD)35. and Il-13, while increasing the synthesis of Interestingly, glucocorticoids inhibit neutrophil apoptosis, revealing a surprising converse effect to anti-inflammatory cytokines, such as Il-10 that of annexin A1 (discussed in reF. 6). Second, endogenous annexin A1 is released from apoptotic (reFs 56,57). Recent work has identified new neutrophils and acts on macrophages to promote the phagocytosis and removal of the apoptotic 83 effects of annexin A1 on T cells; surprisingly, cells . Third, glucocorticoid-treated macrophages secrete annexin A1, which then acts in a autocrine or paracrine manner to increase the engulfment of apoptotic neutrophils84. The however, these effects are the opposite to molecular mechanisms that are responsible for these effects of annexin A1 have not been fully those exerted by glucocorticoids. elucidated and may involve the activation of ALXR (also known as FPR2) and/or binding to phosphatidylserine, a phospholipid that is externalized on the outer leaflet of the plasma Annexin A1 and T‑cell activation. Until membrane of apoptotic cells that can interact with a specific receptor on the phagocyte, recently, little was known about how annexin thereby favouring the engulfment of apoptotic cells. A1 influences the adaptive immune response, Although these observations need to be confirmed in vivo, they suggest that annexin A1 might possibly because T cells express lower levels alter the fate of the extravasated neutrophils and promote their removal by macrophages, thereby of annexin A1 (reF. 42) than myeloid cells8. contributing to the resolution of inflammation.

Other downstream signalling responses of molecules that have been shown to be stem from the fact that a single time point can also be modulated (both positively and involved in the regulation of early events in (7 days after challenge) was assessed in the negatively) by exogenous annexin A1 or by TCR signalling has increased enormously, Anxa1–/– mice with antigen-induced arthritis its deletion in T cells. For example, stimula- and in many cases a link has been found and that this disease is monoarticular (it only tion of T cells with annexin A1 together with between the dysfunction of these molecules affects the injected joint). In addition, the CD3- and CD28-specific antibodies increases and the development of autoimmune dis- comparison of two such different effects — the activation of several key transcriptional eases62. Recent work suggests that annexin the pharmacological effects of annexin A1 factors that are downstream of the TCR58: A1 might belong to the group of molecules administration in collagen-induced arthritis AP1, nF-κB and nFAT. Accordingly, that fine-tune TCR signalling (FIG. 3). and the effect of the absence of annexin A1 reduced activation of these transcription Activation of naive T cells in non- expression in antigen-induced arthritis — factors was detected in Anxa1–/– T cells polarizing conditions (TCR stimulation could further confound any conclusions 59 following stimulation . So, in contrast to and Il-2) or TH1-cell-polarizing conditions drawn. It is therefore difficult at this stage to glucocorticoids, these studies indicate that (TCR stimulation and Il-2, Il-12 and Il-4- determine whether either model might be annexin A1 transduces a stimulatory signal specific blocking antibodies) in the presence useful for understanding the role of annexin to promote T-cell activation, probably of exogenous annexin A1 promotes the A1 in human disease. Of interest in this through AlXR (FIG. 3). differentiation of Il-2- and interferon-γ regard, and in agreement with the studies of γ) (IFn -producing TH1 cells and suppresses collagen-induced arthritis, annexin A1 has

Annexin A1 and effector TH‑cell differentia‑ the production of the TH2-type cytokines recently been shown to have an activating tion. Many factors, including the micro- Il-4 and Il-13 (reF. 58). Consistent with these effect on synovial fibroblasts from patients environment where T cells encounter their findings, in vitro differentiation of T cells with rheumatoid arthritis, in which it com- antigen and the affinity of the antigen for from Anxa1–/– mice revealed a reduction plements the actions of tumour-necrosis the TCR, have an important role in the early in the production of TH1-type cytokines factor and contributes to the release of (reF. 63) phase of TCR signalling and therefore in and increased skewing to a TH2-cell pheno- metalloproteinase 1 . More studies in determining the activation of specific gene type59. In agreement with these in vitro data, these and other models are needed to clarify programmes that lead to the differentia- systemic treatment of mice with annexin the role of the annexin A1–AlXR pathway 61 tion of naive T cells into effector TH cells . A1 immediately after immunization with in arthritis and in distinct phases of the Understanding how subtle changes in the collagen exacerbated the severity of collagen- experimental arthritic response. strength of TCR signalling modulate T-cell induced arthritis at disease onset58. However, Further studies of human cells provide differentiation is of particular importance these observations are in contrast to those clues that annexin A1-mediated stimulation for understanding and treating many auto- from studies of Anxa1–/– mice with antigen- of T cells might be relevant to human disease. immune diseases, in which T cells become induced arthritis26, which showed increased First, ANXA1 mRnA and protein levels were responsive to signals that would otherwise be disease severity in the absence of annexin A1. shown to be higher in circulating T cells from ignored. Over the past 20 years, the number These apparently contradictory results could patients with rheumatoid arthritis than those a Physiological condition b Pathological condition c Pathological condition Appropriate T-cell activation T-cell hyperactivation Insufficient T-cell activation (such as in autoimmune disease) (such as immunodeficiency) Annexin A1 TCR

ALXR CD3

MAPK MAPK MAPK MAPK MAPK MAPK MAPK MAPK MAPK

Cytoplasm Transcription factor Nucleus Strength of TCR signal: ++++ Strength of TCR signal: ++++++ Strength of TCR signal: ++

Figure 3 | The annexin A1–AlXr pathway and strength of T-cell strength of Tcr signalling and exert homeostatic control on T-cell activation receptor signalling. engagement of the T-cell receptor (Tcr) triggers under physiological conditions (a). Higher expressionNature Reof vieannexinws | Immunolog A1 duringy several signalling pathways, including the mitogen-activated protein kinase pathological conditions could increase the strength of Tcr signalling through (MAPK) pathway. Activation of this pathway can lead to the expression of the MAPK signalling pathway, thereby causing a state of hyperactivation of genes such as interleukin-2 (iL-2), which is responsible for T-cell activation and T cells (b). By contrast, low levels of annexin A1 expression, which occurs during proliferation. Activation of T cells also results in the release of annexin A1 and pathological immunodeficiency, could be associated with a hyporesponsive the expression of its receptor ALXr. This pathway seems to fine-tune the state of T cells owing to the decreased strength of Tcr signalling (c).

Adrenal for the efficient inhibition of T-cell activa- gland Endogenous Synthetic tion. We propose that this time-dependent glucocorticoids glucocorticoids difference reflects the differential effects of glucocorticoids on annexin A1 expression in these cells. Future investigation will be Innate immune cells Adaptive immune cells needed to determine the molecular mechanisms that are responsible for the differential ↑ Annexin A1 ↓ Annexin A1 modulation of annexin A1 expression by glucocorticoids in the innate and adaptive Monocyte Mast cell immune system. In this regard, a thorough Neutrophil Macrophage T cell B cell and detailed analysis of the transcriptional regulation of ANXA1 in innate and adaptive immune cells might provide some clues, such as the involvement of cell-specific ↓ Neutrophil ↑ Phagocytosis ↓ Mast-cell ↓ T-cell proliferation No effect reported trafficking of apoptotic degranulation ↓ T-cell activation transcription factors. Indeed, we have found ↑ Neutrophil cells by ↓ Mast-cell ↓ TH1-cell differentiation three consensus GATA-binding protein 3 apoptosis macrophages activation ↑ T 2-cell differentiation H (GATA3)-binding sites in the Anxa1 pro-

↓ Innate immune response ↓ Adaptive immune response moter (F.D’A., unpublished observations). This transcription factor is preferentially Figure 4 | model of glucocorticoid modulation of the annexin A1 pathway in immune regulation. 65 We propose that endogenous and synthetic glucocorticoids can control innate and adaptive immune expressed by T cells but not monocytes and responses by modulating the expression and release of annexin A1. exposure of innate immune cells, is a well-known regulator of TH2-cell dif- 66 such as neutrophils, monocytes, macrophages and mast cells, to glucocorticoids induces the release of ferentiation . Interestingly, glucocorticoids annexin A1. This, in turn, acts in a paracrine or autocrine manner to maintainNatur thee Re levelviews (and | Immunolog duration)y have been reported to inhibit GATA3 tran- of cell activation. in a dynamic system such as the inflammatory reaction, in which spatial and temporal scriptional activity67. We speculate that inhi- factors can influence the final outcome, this could promote the resolution of the inflammatory reaction bition of GATA3 activity by glucocorticoids by actively maintaining homeostatic control of innate immune cells. By contrast, exposure of T cells occurs upstream of their suppressive action to glucocorticoids leads to a reduction in annexin A1 expression and consequently inhibition of T-cell on annexin A1 expression in T cells. activation and the differentiation of T cells towards a T helper 2 (TH2)-cell phenotype. conclusions On the basis of recent studies of the adap- from healthy individuals58. Second, synovial and second, incubation of dexamethasone- tive immune response in Anxa1–/– mice T cells expressed higher levels of annexin A1 pretreated human T cells with recombinant and previous studies of the innate immune than tonsil T cells from the same patient64. annexin A1 for 12 hours almost completely system, we propose a new hypothesis to Therefore, it is tempting to speculate that reversed the suppressive effects of dexametha- explain the link between annexin A1 and the dysregulated expression of annexin A1 sone pretreatment and restored their ability glucocorticoids. The model suggests that in T cells might contribute to the hyper- to proliferate following stimulation with part of the anti-inflammatory effect of gluco- sensitivity of these cells to antigen stimulation CD3- and CD28-specific antibodies64. corticoids on the innate immune system is and to the development of an overexuberant The clinical relevance of these findings mediated through the release of annexin A1 immune response in disease states. is supported by a recent study in humans. and the subsequent activation of AlXR, in Treatment of five patients with rheumatoid an autocrine or paracrine manner, in neutro- Glucocorticoids and annexin A1 in T cells. arthritis with the synthetic glucocorticoid phils and macrophages. By contrast, the The modulation of the annexin A1–AlXR methylprednisolone acetate was shown to immunosuppressive effect of glucocorticoids pathway in innate immune cells by glucocor- reduce the expression of annexin A1 in cir- on the adaptive immune system is medi- ticoids (as described above) prompted us to culating T cells by ≥ 50% (reF. 64). Together, ated through the inhibition of annexin A1 investigate the effects of glucocorticoids on these results suggest that inhibition of expression by T cells (FIG. 4). We speculate annexin A1 expression in T cells. Incubation annexin A1 expression might be one of the that daily treatment with glucocorticoids of mouse and human T cells with dexam- mechanisms responsible for the immuno- or intermittent therapy with high doses of ethasone induced a time-dependent decrease suppressive effects of glucocorticoids on these drugs would induce a constant release in both ANXA1 mRnA and annexin A1 T cells, and that glucocorticoids exert dif- of annexin A1 by innate immune cells, and expression levels64. These results contrast ferent effects on annexin A1 expression in this would have a pro-resolving effect on the with those that were obtained from studies of innate and adaptive immune cells. inflammatory response. In parallel, gluco- innate immune cells, but may explain how the The modulation of annexin A1 expres- corticoids would reduce the co-stimulatory immunosuppressive effects of glucocorticoids sion in T cells by glucocorticoids might also effect of annexin A1 on T-cell activation, can be achieved given the stimulatory effects explain a well-known but poorly understood thereby lowering the strength of the TCR of annexin A1 on T cells. It might therefore aspect of glucocorticoid pharmacology, the signal (FIG. 3) and maintaining these cells in be possible that glucocorticoids inhibit T-cell cell-specific time latency of their beneficial an unresponsive state. activation and proliferation by reducing effects. The inhibitory effects of glucocor- In this model, annexin A1 has important the expression of annexin A1. Two pieces ticoids on macrophages and neutrophils opposing properties during innate and adap- of evidence support this hypothesis: first, requires short incubation times (less than tive immune responses, as it inhibits innate T cells from Anxa1–/– mice proliferate less in 2 hours), whereas much longer incubation immune cells and promotes T-cell activa- response to CD3 and CD28 stimulation59, times (more than 6–12 hours) are needed tion, which is a property that is shared with

other immune regulators, such as galectin 1 20. Walther, A., Riehemann, K. & Gerke, V. A novel ligand 43. Solito, E., de Coupade, C., Parente, L., Flower, R. J. & of the formyl peptide receptor: annexin I regulates Russo‑Marie, F. IL‑6 stimulates annexin 1 expression (reF. 68) and T-cell immunoglobulin domain neutrophil extravasation by interacting with the FPR. and translocation and suggests a new biological role and mucin domain 3 (TIM3)69. A complete Mol. Cell 5, 831–840 (2000). as class II acute phase protein. Cytokine 10, 514–521 21. Zouki, C., Ouellet, S. & Filep, J. G. The anti‑ (1998). understanding of the effects of glucocorti- inflammatory peptides, antiflammins, regulate 44. Hannon, R. et al. Aberrant inflammation and coids and annexin A1 on immune responses, the expression of adhesion molecules on human resistance to glucocorticoids in annexin 1–/– mouse. leukocytes and prevent neutrophil adhesion to FASEB J. 17, 253–255 (2003). and the identification of AlXR as its main endothelial cells. FASEB J. 14, 572–580 (2000). 45. Damazo, A. S., Yona, S., Flower, R. J., Perretti, M. & receptor, should aid the development of 22. Hayhoe, R. P. et al. Annexin 1 and its bioactive Oliani, S. M. Spatial and temporal profiles for anti‑ peptide inhibit neutrophil–endothelium interactions inflammatory gene expression in leukocytes during a selective agonists for the resolution of inflam- under flow: indication of distinct receptor involvement. resolving model of peritonitis. J. Immunol. 176, matory responses and antagonists for the Blood 107, 2123–2130 (2006). 4410–4418 (2006). 23. Chatterjee, B. E. et al. Annexin 1 deficient neutrophils 46. Sawmynaden, P. & Perretti, M. Glucocorticoid control of adaptive immune responses. exhibit enhanced transmigration in vivo and increased upregulation of the annexin‑A1 receptor in leukocytes. responsiveness in vitro. J. Leukoc. Biol. 78, 639–646 Biochem. Biophys. Res. Commun. 349, 1351–1355 Mauro Perretti and Fulvio D’Acquisto are at the (2005). (2006). William Harvey Research Institute, Barts and 24. Mancuso, F., Flower, R. J. & Perretti, M. Leukocyte 47. Ehrchen, J. et al. Glucocorticoids induce differentiation The London School of Medicine and Dentistry, Queen transmigration, but not rolling or adhesion, is of a specifically activated, anti‑inflammatory subtype Mary University of London, London EC1M 6BQ, UK. selectively inhibited by dexamethasone in the hamster of human monocytes. Blood 109, 1265–1274 post‑capillary venule. Involvement of endogenous (2007). Correspondence to M.P. lipocortin 1. J. Immunol. 155, 377–386 (1995). 48. Hashimoto, A., Murakami, Y., Kitasato, H., Hayashi, I. e-mail: [email protected] 25. Frey, O. et al. The role of regulatory T cells in & Endo, H. Glucocorticoids co‑interact with lipoxin A4 antigen‑induced arthritis: aggravation of arthritis via lipoxin A receptor (ALX) up‑regulation. Biomed. doi:10.1038/nri2470 4 after depletion and amelioration after transfer of Pharmacother. 61, 81–85 (2007). 1. Serhan, C. N. & Savill, J. Resolution of inflammation: CD4+CD25+ T cells. Arthritis Res. Ther. 7, 49. Croxtall, J. D., van Hal, P. T., Choudhury, Q., the beginning programs the end. Nature Immunol. 6, R291–R301 (2005). Gilroy, D. W. & Flower, R. J. Different glucocorticoids 1191–1197 (2005). 26. Yang, Y. H. et al. Modulation of inflammation and vary in their genomic and non‑genomic mechanism of 2. Serhan, C. N. et al. Resolution of inflammation: state response to dexamethasone by annexin‑1 in antigen‑ action in A549 cells. Br. J. Pharmacol. 135, 511–519 of the art, definitions and terms. FASEB J. 21, induced arthritis. Arthritis Rheum. 50, 976–984 (2002). 325–332 (2007). (2004). 50. Franchimont, D. Overview of the actions of 3. Gonzalez‑Rey, E., Chorny, A. & Delgado, M. 27. Ohshima, S. et al. Interleukin 6 plays a key role in the glucocorticoids on the immune response: Regulation of immune tolerance by anti‑inflammatory development of antigen‑induced arthritis. Proc. Natl a good model to characterize new pathways of neuropeptides. Nature Rev. Immunol. 7, 52–63 Acad. Sci. USA 95, 8222–8226 (1998). immunosuppression for new treatment strategies. (2007). 28. Yang, Y. H. et al. Annexin 1 negatively regulates IL‑6 Ann. NY Acad. Sci. 1024, 124–137 (2004). 4. Rhen, T. & Cidlowski, J. A. Antiinflammatory action of expression via effects on p38 MAPK and MAPK 51. Van Laethem, F. et al. Glucocorticoids attenuate T cell glucocorticoids — new mechanisms for old drugs. phosphatase‑1. J. Immunol. 177, 8148–8153 receptor signaling. J. Exp. Med. 193, 803–814 N. Engl. J. Med. 353, 1711–1723 (2005). (2006). (2001). 5. Sapolsky, R. M., Romero, L. M. & Munck, A. U. 29. Damazo, A. S. et al. Critical protective role for annexin 1 52. Van Laethem, F., Baus, E., Andris, F., Urbain, J. & How do glucocorticoids influence stress responses? gene expression in the endotoxemic murine Leo, O. A novel aspect of the anti‑inflammatory actions Integrating permissive, suppressive, stimulatory, and microcirculation. Am. J. Pathol. 166, 1607–1617 of glucocorticoids: inhibition of proximal steps of preparative actions. Endocr. Rev. 21, 55–89 (2000). (2005). signaling cascades in lymphocytes. Cell. Mol. Life Sci. 6. Parente, L. & Solito, E. Annexin 1: more than an anti‑ 30. Perretti, M. et al. Endogenous lipid‑ and peptide‑ 58, 1599–1606 (2001). phospholipase protein. Inflamm. Res. 53, 125–132 derived anti‑inflammatory pathways generated with 53. Adcock, I. M. Glucocorticoid‑regulated transcription (2004). glucocorticoid and aspirin treatment activate the factors. Pulm. Pharmacol. Ther. 14, 211–219 (2001).

7. Gerke, V., Creutz, C. E. & Moss, S. E. Annexins: linking lipoxin A4 receptor. Nature Med. 8, 1296–1302 54. Ramirez, F., Fowell, D. J., Puklavec, M., Simmonds, S. Ca2+ signalling to membrane dynamics. Nature Rev. (2002). & Mason, D. Glucocorticoids promote a TH2 cytokine Mol. Cell Biol. 6, 449–461 (2005). 31. Chiang, N. et al. The lipoxin receptor ALX: potent response by CD4+ T cells in vitro. J. Immunol. 156, 8. Morand, E. F. et al. Detection of intracellular ligand‑specific and stereoselective actions in vivo. 2406–2412 (1996). lipocortin 1 in human leukocyte subsets. Clin. Immunol. Pharmacol. Rev. 58, 463–487 (2006). 55. Ramirez, F. Glucocorticoids induce a Th2 response Immunopathol. 76, 195–202 (1995). 32. Babbin, B. A. et al. Annexin I regulates SKCO‑15 in vitro. Dev. Immunol. 6, 233–243 (1998). 9. Perretti, M. et al. Annexin I is stored within gelatinase cell invasion by signaling through formyl peptide 56. Elenkov, I. J. Glucocorticoids and the Th1/Th2 balance. granules of human neutrophils and mobilised on the receptors. J. Biol. Chem. 281, 19588–19599 (2006). Ann. NY Acad. Sci. 1024, 138–146 (2004). cell surface upon adhesion but not phagocytosis. 33. He, R., Sang, H. & Ye, R. D. Serum amyloid A induces 57. Richards, D. F., Fernandez, M., Caulfield, J. & Cell Biol. Int. 24, 163–174 (2000). IL‑8 secretion through a G protein‑coupled receptor, Hawrylowicz, C. M. Glucocorticoids drive human CD8+ 10. Mulla, A., Leroux, C., Solito, E. & Buckingham, J. C. FPRL1/LXA4R. Blood 101, 1572–1581 (2003). T cell differentiation towards a phenotype with high Correlation between the antiinflammatory protein 34. Dahlgren, C. et al. The synthetic chemoattractant IL‑10 and reduced IL‑4, IL‑5 and IL‑13 production. annexin 1 (lipocortin 1) and serum cortisol in subjects Trp‑Lys‑Tyr‑Met‑Val‑DMet activates neutrophils Eur. J. Immunol. 30, 2344–2354 (2000).

with normal and dysregulated adrenal function. preferentially through the lipoxin A4 receptor. 58. D’Acquisto, F. et al. Annexin‑1 modulates T‑cell J. Clin. Endocrinol. Metab. 90, 557–562 (2005). Blood 95, 1810–1818 (2000). activation and differentiation. Blood 109, 1095–1102 11. Perretti, M. et al. Mobilizing lipocortin 1 in adherent 35. Solito, E. et al. A novel calcium‑dependent pro‑ (2007). human leukocytes downregulates their transmigration. apoptotic effect of annexin 1 on human neutrophils. 59. D’Acquisto, F. et al. Impaired T cell activation and Nature Med. 22, 1259–1262 (1996). FASEB J. 17, 1544–1546 (2003). increased Th2 lineage commitment in annexin‑1‑ 12. Wein, S. et al. Mediation of annexin 1 secretion by a 36. Gao, J.‑L., Chen, H., Filie, J. D., Kozak, C. A. & deficient T cells. Eur. J. Immunol. 37, 3131–3142 probenecid‑sensitive ABC‑transporter in rat inflamed Murphy, P. M. Differential expansion of the (2007). mucosa. Biochem. Pharmacol. 67, 1195–1202 N‑formylpeptide receptor gene cluster in human 60. Ebisuya, M., Kondoh, K. & Nishida, E. The duration, (2004). and mouse. Genomics 51, 270–276 (1998). magnitude and compartmentalization of ERK MAP 13. Solito, E. et al. Post‑translational modification plays an 37. Perretti, M., Getting, S. J., Solito, E., Murphy, P. M. & kinase activity: mechanisms for providing signaling essential role in the translocation of annexin A1 from Gao, J. L. Involvement of the receptor for formylated specificity. J. Cell Sci. 118, 2997–3002 (2005). the cytoplasm to the cell surface. FASEB J. 20, peptides in the in vivo anti‑migratory actions of 61. Boyton, R. J. & Altmann, D. M. Is selection for TCR 1498–1500 (2006). annexin 1 and its mimetics. Am. J. Pathol. 158, affinity a factor in cytokine polarization? Trends 14. Lominadze, G. et al. Proteomic analysis of human 1969–1973 (2001). Immunol. 23, 526–529 (2002). neutrophil granules. Mol. Cell. Proteomics 4, 38. Gavins, F. N., Yona, S., Kamal, A. M., Flower, R. J. & 62. Marti, F., Lapinski, P. E. & King, P. D. The emerging 1503–1521 (2005). Perretti, M. Leukocyte antiadhesive actions of role of the T cell‑specific adaptor (TSAd) protein as an 15. Vong, L. et al. Annexin 1 cleavage in activated annexin 1: ALXR‑ and FPR‑related anti‑inflammatory autoimmune disease‑regulator in mouse and man. neutrophils: a pivotal role for proteinase 3. mechanisms. Blood 101, 4140–4147 (2003). Immunol. Lett. 97, 165–170 (2005). J. Biol. Chem. 282, 29998–30004 (2007). 39. Souza, D. G. et al. The required role of endogenously 63. Tagoe, C. E. et al. Annexin‑1 mediates 16. Rosengarth, A. & Luecke, H. A calcium‑driven produced lipoxin A4 and annexin‑1 for the production TNF‑α‑stimulated matrix metalloproteinase secretion conformational switch of the N‑terminal and core of IL‑10 and inflammatory hyporesponsiveness in from rheumatoid arthritis synovial fibroblasts. domains of annexin A1. J. Mol. Biol. 326, mice. J. Immunol. 179, 8533–8543 (2007). J. Immunol. 181, 2813–2820 (2008). 1317–1325 (2003). 40. Ferlazzo, V. et al. Anti‑inflammatory effects of 64. D’Acquisto, F. et al. Glucocorticoid treatment inhibits 17. Perretti, M. & Flower, R. J. Annexin 1 and the annexin‑1: stimulation of IL‑10 release and inhibition annexin‑1 expression in rheumatoid arthritis CD4+ biology of the neutrophil. J. Leukoc. Biol. 75, of nitric oxide synthesis. Int. Immunopharmacol. 3, T cells. Rheumatology (Oxford) 47, 636–639 25–29 (2004). 1363–1369 (2003). (2008). 18. Lim, L. H. & Pervaiz, S. Annexin 1: the new face of an 41. Clark, A. R. Anti‑inflammatory functions of 65. Caramori, G. et al. Expression of GATA family of old molecule. FASEB J. 21, 968–975 (2007). glucocorticoid‑induced genes. Mol. Cell. Endocrinol. transcription factors in T‑cells, monocytes and 19. Lim, L. H., Solito, E., Russo‑Marie, F., Flower, R. J. & 275, 79–97 (2007). bronchial biopsies. Eur. Respir. J. 18, 466–473 Perretti, M. Promoting detachment of neutrophils 42. Goulding, N. J. et al. Anti‑inflammatory lipocortin 1 (2001). adherent to murine postcapillary venules to control production by peripheral blood leucocytes in 66. Zhou, M. & Ouyang, W. The function role of GATA‑3 inflammation: effect of lipocortin 1. Proc. Natl Acad. response to hydrocortisone. Lancet 335, 1416–1418 in Th1 and Th2 differentiation. Immunol. Res. 28, Sci. USA 95, 14535–14539 (1998). (1990). 25–37 (2003).

67. Jee, Y. K. et al. Repression of interleukin‑5 75. Le, Y. et al. Amyloid β42 activates a G‑protein‑coupled 84. Maderna, P., Yona, S., Perretti, M. & Godson, C. transcription by the glucocorticoid receptor targets chemoattractant receptor, FPR‑like‑1. J. Neurosci. 21, Modulation of phagocytosis of apoptotic neutrophils GATA3 signaling and involves histone deacetylase RC123 (2001). by supernatant from dexamethasone‑treated recruitment. J. Biol. Chem. 280, 23243–23250 76. Fu, H. et al. Ligand recognition and activation of macrophages and annexin‑derived peptide Ac2–26. (2005). formyl peptide receptors in neutrophils. J. Leukoc. J. Immunol. 174, 3727–3733 (2005). 68. Barrionuevo, P. et al. A novel function for galectin‑1 Biol. 79, 247–256 (2006). at the crossroad of innate and adaptive immunity: 77. Le, Y. et al. Identification of functional domains in the Acknowledgements galectin‑1 regulates monocyte/macrophage physiology formyl peptide receptor‑like 1 for agonist‑induced cell Work in our laboratory is funded by the Arthritis Research through a nonapoptotic ERK‑dependent pathway. chemotaxis. FEBS J. 272, 769–778 (2005). Campaign UK, the Wellcome Trust, the British Heart J. Immunol. 178, 436–445 (2007). 78. Rodrigues‑Lisoni, F. C. et al. In vitro and in vivo studies Foundation and the Medical Research Council UK. We apolo‑ 69. Anderson, A. C. et al. Promotion of tissue on CCR10 regulation by annexin A1. FEBS Lett. 580, gize to the many colleagues whose work could not be cited inflammation by the immune receptor Tim‑3 1431–1438 (2006). here owing to space limitations. expressed on innate immune cells. Science 318, 79. Lange, C., Starrett, D., Goetsch, J., Gerke, V. & 1141–1143 (2007). Rescher, U. Transcriptional profiling of human Competing financial interests 70. Ito, K. et al. Histone deacetylase 2‑mediated monocytes reveals complex changes in the expression The authors declare competing financial interests: see web deacetylation of the glucocorticoid receptor enables pattern of inflammation‑related genes in response to version for details. NF‑κB suppression. J. Exp. Med. 203, 7–13 (2006). the annexin A1‑derived peptide Ac1–25. J. Leukoc. 71. Song, I. H. & Buttgereit, F. Non‑genomic Biol. 82, 1592–1604 (2007). glucocorticoid effects to provide the basis for new drug 80. Burli, R. W. et al. Potent hFPRL1 (ALXR) agonists as developments. Mol. Cell. Endocrinol. 246, 142–146 potential anti‑inflammatory agents. Bioorg. Med. DAtABAsEs (2006). Chem. Lett. 16, 3713–3718 (2006). entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query. 72. Lowenberg, M. et al. Glucocorticoids cause rapid 81. Gilmour, J. S. et al. Local amplification of glucocorticoids fcgi?db=gene dissociation of a T‑cell‑receptor‑associated protein by 11 β‑hydroxysteroid dehydrogenase type 1 AKT | ANXA1 | DusP1 | erK1 | erK2 | FPr1 | FPr2 | FPr3 | complex containing LCK and FYN. EMBO Rep. 7, promotes macrophage phagocytosis of apoptotic GATA3 | iL-1 | iL-6 1023–1029 (2006). leukocytes. J. Immunol. 176, 7605–7611 (2006). FURtHER inFoRMAtion 73. Moraes, L. A. et al. Ligand‑specific glucocorticoid 82. Liu, Y. et al. Glucocorticoids promote nonphlogistic Mauro Perretti’s homepage: http://www.whri.qmul.ac.uk/ receptor activation in human platelets. Blood 106, phagocytosis of apoptotic leukocytes. J. Immunol. staff/perretti.html 4167–4175 (2005). 162, 3639–3646 (1999). Fulvio D’Acquisto’s homepage: http://www.whri.qmul.ac.uk/ 74. Badolato, R. et al. Serum amyloid A induces calcium 83. Scannell, M. et al. Annexin‑1 and peptide derivatives staff/D’Acquisto.html mobilization and chemotaxis of human monocytes are released by apoptotic cells and stimulate by activating a pertussis toxin‑sensitive signaling phagocytosis of apoptotic neutrophils by All lInks ArE AcTIvE In ThE onlInE pdf pathway. J. Immunol. 155, 4004–4010 (1995). macrophages. J. Immunol. 178, 4595–4605 (2007).